xvii
CHAPTER 1 Chapter Title
Contributors
Numbers in parentheses indicate the volume number and pages on which the a...
54 downloads
1478 Views
80MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
xvii
CHAPTER 1 Chapter Title
Contributors
Numbers in parentheses indicate the volume number and pages on which the authors’ contributions begin.
John P. Bilezikian (2:71) Departments of Medicine and Pharmacology, College of Physicians and Surgeons, Columbia University, New York, New York 10032 Peter V. N. Bodine (1:305) Women’s Health Research Institute, Wyeth-Ayerst Research, Radnor, Pennsylvania 19087 Henry Bone (2:533) Michigan Bone and Mineral Clinic, Detroit, Michigan 48236 Jean-Philippe Bonjour (1:621) Department of Internal Medicine, Division of Bone Diseases (World Health Organization Collaborating Center for Osteoporosis and Bone Diseases), University Hospital, Geneva CH-1211, Switzerland Adele L. Boskey (1:107) Hospital for Special Surgery, Weill Medical College of Cornell University, New York, New York 10021 Mary L. Bouxsein (1:509) Orthopedic Biomechanics Laboratory, Department of Orthopedic Surgery, Beth Israel Deaconess Medical Center, and Harvard Medical School, Boston, Massachusetts 02215 Alan L. Burshell (2:195) Department of Internal Medicine, Alton Ochsner Medical Foundation, Ochsner Clinic, New Orleans, Louisiana 70131 Elizabeth Capezuti (1:795) Nell Hodgson Woodruff School of Nursing, Emory University, Atlanta, Georgia 30322 Dennis R. Carter (1:471) Department of Mechanical Engineering, Biomechanical Engineering Program, Stanford University, Stanford, California 94305; and Rehabilitation Research and Development Center, Veterans Affairs Palo Alto, Palo Alto, California 94304 Jane A. Cauley (1:741) University of Pittsburgh, Pittsburgh, Pennsylvania 15261
M. Arlot (2:501) INSERM Unité 403, Faculté RTH Laennec, Lyon 69372, France Laura K. Bachrach (2:151) Department of Pediatrics, Stanford University School of Medicine, Stanford, California 94305 Daniel Baran (2:229) Departments of Medicine, Orthopedics, and Cell Biology, University of Massachusetts Medical Center, and Merck & Company, Inc., Worcester, Massachusetts 01655 M. Janet Barger-Lux (2:59) Osteoporosis Research Center, Creighton University, Omaha, Nebraska 68131 Elizabeth Barrett-Conner (1:819) Department of Family and Preventive Medicine, University of California, San Diego, La Jolla, California 90293 David J. Baylink (1:405; 2:675) Department of Medicine, Loma Linda University, and the Musculoskeletal Disease Center, Jerry L. Pettis Memorial Veterans Affairs Medical Center, Loma Linda, California 92357 Gary S. Beaupré (1:471) Department of Mechanical Engineering, Biomechanical Engineering Program, Stanford University, Stanford, California 94305; and Rehabilitation Research and Development Center, Veterans Affairs Palo Alto, Palo Alto, California 94304 Belinda R. Beck (1:701) Griffith University, School of Physiotherapy and Exercise Science, Queensland 9726, Australia Daniel D. Bikle (2:237) Department of Medicine, University of California, San Francisco, School of Medicine, and Department of Veterans Affairs, San Francisco Veterans Affairs Medical Center, San Francisco, California 94121
xvii
xviii
CONTRIBUTORS
Jacqueline R. Center (2:169) Bone and Mineral Research Program, Garvan Institute of Medical Research, St. Vincent’s Hospital, Sydney, New South Wales 2010, Australia P. Chavassieux (2:501) INSERM Unité 403, Faculté RTH Laennec, Lyon 69372, France Di Chen (1:373) Department of Medicine and Endocrinology, University of Texas Health Science Center, San Antonio, Texas 78284 Michael Chorov (2:769) Department of Medicine, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts 02215 Roberto Civitelli (2:651) Division of Bone and Mineral Diseases, Washington University School of Medicine, and Barnes-Jewish Hospital, St. Louis, Missouri 63110 Cyrus Cooper (1:557) MRC Environmental Epidemiology Unit, University of Southampton, Southampton General Hospital, Southampton SO16 6YD, England Felicia Cosman (2:577) Regional Bone Center, Helen Hayes Hospital, West Haverstraw, New York 10993 Gilbert J. Cote (1:247) Department of Endocrine Neoplasia and Hormonal Disorders, Division of Internal Medicine, University of Texas M.D. Anderson Cancer Center, Houston, Texas 77030 Ann B. Cranney (2:539) Department of Medicine, University of Ottawa, Ottawa, Ontario, Canada K1Y 1J8 Sarah Dallas (1:373) Department of Medicine and Endocrinology, University of Texas Health Science Center, San Antonio, Texas 78284 Bess Dawson-Hughes (2:545) Jean Mayer USDA Human Nutrition Research Center on Aging, Tufts University, Boston, Massachusetts 02111 Chris de Laet (1:585) Institute for Medical Technology Assessment, Erasmus University, Rotterdam 3000 DR, The Netherlands Arthur A. DeCarlo (2:363) Department of Periodontics, University of Alabama School of Dentistry, Birmingham, Alabama 35294 Pierre D. Delmas (2:459) INSERM Research Unit 403, Hôpital E. Herriot, and Claude Bernard University, Lyon 69003, France Marc K. Drezner (2:479) Department of Medicine, University of Wisconsin Medical School, Madison, Wisconsin 53792 Thomas A. Einhorn (1:3) Department of Orthopedic Surgery, Boston University School of Medicine, Boston, Massachusetts 02118
John A. Eisman (2:169) Bone and Mineral Research Program, Garvan Institute of Medical Research, St. Vincent’s Hospital, Sydney, New South Wales 2010, Australia Sol Epstein (2:327) Roche Laboratories, Nutley, New Jersey 07110 Kenneth G. Faulkner (2:433) GE Medical Systems, Madison, Wisconsin 53717 David Feldman (1:257) Department of Medicine, Division of Endocrinology, Stanford University School of Medicine, Stanford, California 94305 Lorraine A. Fitzpatrick (2:259) Division of Endocrinology and Metabolism and Internal Medicine, Mayo Clinic and Foundation, Rochester, Minnesota 55905 H. Fleisch (1:449) University of Berne, CH-3008 Berne, Switzerland Robert F. Gagel (1:247) Department of Endocrine Neoplasia and Hormonal Disorders, Division of Internal Medicine, University of Texas M.D. Anderson Cancer Center, Houston, Texas 77030 Patrick Garnero (2:459) INSERM Research Unit 403, Hôpital E. Herriot, and Synarc, Lyon 69003, France Harry K. Genant (2:411) Department of Radiology, University of California, San Francisco, San Francisco, California 94143 Jayashree A. Gokhale (1:107) Hospital for Special Surgery, Weill Medical College of Cornell University, New York, New York 10021 Deborah T. Gold (2:479) Departments of Psychiatry and Behavioral Sciences, Sociology, Psychology, and Aging Center, Duke University Medical Center, Durham, North Carolina 27710 Steven R. Goldring (2:351) Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, and New England Baptist Bone and Joint Institute, Harvard Institutes of Medicine, Boston, Massachusetts 02215 Gail A. Greendale (1:819) Department of Medicine, Division of Geriatrics, University of California, Los Angeles, School of Medicine, Los Angeles, California 90095 Jeane Ann Grisso (1:795) Center for Clinical Epidemiology and Biostatistics, Division of General Internal Medicine, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104 Coleman Gross (1:257) Department of Medicine, Division of Endocrinology, Stanford University School of Medicine, Stanford, California 94305
xviii
xix
CONTRIBUTORS
Gordon H. Guyatt (2:539) Department of Medicine and Clinical Epidemiology and Biostatistics, McMaster University, Hamilton, Ontario, Canada L8N 3Z5 Reinhard Gysin (1:405) Department of Medicine, Loma Linda University, and the Musculoskeletal Disease Center, Jerry L. Pettis Memorial Veterans Affairs Medical Center, Loma Linda, California 92357 Robert P. Heaney (1:669; 2:513) Creighton University, Omaha, Nebraska 68178 Hunter Heath III (2:259) U.S. Medical Division, Eli Lilly and Company, Indianapolis, Indiana 46285 Michael H. Heggeness (2:485) Department of Orthopaedic Surgery, Center for Spinal Disorders, Baylor College of Medicine, Houston, Texas 77030 N. Kathryn Henderson (2:169) Bone and Mineral Research Program, Garvan Institute of Medical Research, St. Vincent’s Hospital, Sydney, New South Wales 2010, Australia Ana O. Hoff (1:247) Department of Endocrine Neoplasia and Hormonal Disorders, Division of Internal Medicine, University of Texas M. D. Anderson Cancer Center, Houston, Texas 77030 Siu L. Hui (1:809) Department of Medicine, Indiana University School of Medicine, and the Regenstrief Institute for Health Care, Indianapolis, Indiana 46202 Marjorie K. Jeffcoat (2:363) Department of Periodontics, University of Alabama School of Dentistry, Birmingham, Alabama 35294 Michael Jergas (2:411) Department of Radiology, University of California, San Francisco, San Francisco, California 94143 C. Conrad Johnston (1:809) Department of Medicine, Indiana University School of Medicine, and the Regenstrief Institute for Health Care, Indianapolis, Indiana 46202 Stefan Judex (1:489) Musculo-Skeletal Research Laboratory, Department of Biomedical Engineering, State University of New York, Stony Brook, New York 11794 Gerard Karsenty (1:213) Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, Texas 77030 Jennifer L. Kelsey (1:535) Department of Health, Research, and Policy, Division of Epidemiology, Stanford University School of Medicine, Stanford, California 94305 Sundeep Khosla (2:49; 2:709) Department of Internal Medicine, Division of Endocrinology and Metabolism, Mayo Clinic and Foundation, Rochester, Minnesota 55905
Donald B. Kimmel (2:29) Osteoporosis Research Center, Creighton University, Omaha, Nebraska 68131 B. Jenny Kiratli (2:207) Spinal Cord Injury Center, Palo Alto Veterans Affairs Health Care System, Palo Alto, California 94304 Michael Kleerekoper (2:403) Department of Internal Medicine, Wayne State University, Detroit, Michigan 48201 Robert F. Klein (2:103) Bone and Mineral Research Unit, Oregon Health Sciences University, and the Portland Veterans Affairs Medical Center, Portland, Oregon 97201 Lynn Kohlmeier (2:341) Spokane Osteoporosis Center, Endocrine Associates of Spokane, Spokane, Washington 99204 Barry S. Komm (1:305) Women’s Health Research Institute, Wyeth-Ayerst Research, Radnor, Pennsylvania 19087 K.-H. William Lau (2:675) Departments of Medicine and Biochemistry, Jerry L. Pettis Veterans Affairs Medical Center, and Loma Linda University, Loma Linda, California 92357 Cassandra A. Lee (1:3) Department of Orthopedic Surgery, Boston University School of Medicine, Boston, Massachusetts 02118 Gary M. Leong (2:169) Bone and Mineral Research Program, Garvan Institute of Medical Research, St. Vincent’s Hospital, Sydney, New South Wales 2010, Australia Jane B. Lian (1:21) Department of Cell Biology, University of Massachusetts Medical School, Worcester, Massachusetts 01655 Robert Lindsay (2:577) Regional Bone Center, Helen Hayes Hospital, West Haverstraw, New York 10993 Kenneth W. Lyles (2:479) Departments of Medicine and Aging Center, Sarah W. Stedman Center for Nutritional Studies, GRECC, Veterans Affairs Medical Center, Duke University Medical Center, Durham, North Carolina 27710 Sharmilla Majumdar (2:3) Department of Radiology, University of California, San Francisco, San Francisco, California 94143 Peter J. Malloy (1:257) Department of Medicine, Division of Endocrinology, Stanford University School of Medicine, Stanford, California 94305 W. J. Maloney (2:385) Department of Orthopedic Surgery, Washington University School of Medicine, St. Louis, Missouri 63110 Robert Marcus (2:3; 2:341) Veterans Affairs Medical Center, Palo Alto, California 94304; and Department of Medicine, Stanford
xix
xx
CONTRIBUTORS
University School of Medicine, Stanford, California 94305 Thomas J. Martin (1:361) St. Vincent’s Institute of Medical Research, Melbourne 3065, Australia Kenneth B. Mathis (2:485) Department of Orthopaedic Surgery, Center for Spinal Disorders, Baylor College of Medicine, Houston, Texas 77030 Kenneth J. McLeod (1:489) Musculo-Skeletal Research Laboratory, Department of Biomedical Engineering, State University of New York, Stony Brook, New York 11794 L. Joseph Melton III (1:557; 2:49) Department of Internal Medicine, Division of Endocrinology and Metabolism, and Department of Health Sciences Research, Section of Clinical Epidemiology, Mayo Clinic and Foundation, Rochester, Minnesota 55905 P. J. Meunier (2:501) INSERM Unité 403, Faculté RTH Laennec, Lyon 69372, France Subburaman Mohan (1:405) Department of Medicine, Loma Linda University, and the Musculoskeletal Disease Center, Jerry L. Pettis Memorial Veterans Affairs Medical Center, Loma Linda, California 92357 Lis Mosekilde (2:725) Department of Cell Biology, Institute of Anatomy, University of Aarhus, DK-8000, Aarhus C, Denmark Douglas B. Muchmore (2:603) Lilly Research Laboratories, Eli Lilly and Company, Indianapolis, Indiana 46285 Gregory R. Mundy (1:373) Department of Medicine and Endocrinology, University of Texas Health Science Center, San Antonio, Texas 78284 Ran Namgung (1:599) Department of Pediatrics, Yonsei University College of Medicine, Seoul, Korea Dorothy Nelson (1:569) Department of Internal Medicine, Wayne State University, Detroit, Michigan 48201 Lorene Nelson (1:569) Department of Health Research and Policy, Division of Epidemiology, Stanford University School of Medicine, Stanford, California 94305 Robert A. Nissenson (1:221) Endocrine Unit, San Francisco Veterans Affairs Medical Center, and Departments of Medicine and Physiology, University of California, San Francisco, San Francisco, California 94121 Bjorn R. Olsen (1:189) Department of Cell Biology, Harvard Medical School, Boston, Massachusetts 02115
Eric S. Orwoll (1:339; 2:103) Bone and Mineral Unit, Oregon Health Sciences University, and the Portland Veterans Affairs Medical Center, Portland, Oregon 97201 Babatunde Oyajobi (1:373) Department of Medicine and Endocrinology, University of Texas Health Science Center, San Antonio, Texas 78284 Roberto Pacifici (2:85) Division of Bone and Mineral Diseases, Washington University, St. Louis, Missouri 63110 Charles Y. C. Pak (2:699) Center for Mineral Metabolism and Clinical Research, University of Texas Southwestern Medical Center, Dallas, Texas 75390 Socrates E. Papapoulos (2:631) Department of Endocrinology and Metabolic Diseases, University of Leiden Medical Center, 2333 ZA Leiden, The Netherlands A. Michael Parfitt (1:433) Division of Endocrinology, University of Arkansas for Medical Sciences, Little Rock, Arkansas 72205 Millan S. Patel (1:213) Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, Texas 77030 Huibert A. P. Pols (1:639) Department of Internal Medicine, Erasmus University Medical School, 3000 DR Rotterdam, The Netherlands Richard Prince (2:621) Department of Medicine, University of Western Australia, Sir Charles Gairdner Hospital, Perth 6009, Western Australia Xuezhong Qin (1:405) Department of Medicine, Loma Linda University, and the Musculoskeletal Disease Center, Jerry L. Pettis Memorial Veterans Affairs Medical Center, Loma Linda, California 92357 Yi-Xian Qin (1:489) Musculo-Skeletal Research Laboratory, Department of Biomedical Engineering, State University of New York, Stony Brook, New York 11794 Lawrence G. Raisz (2:19) Department of Endocrinology and Metabolism, University of Connecticut Health Center, Farmington, Connecticut 06032 Robert R. Recker (2:59) Osteoporosis Research Center, Creighton University, Omaha, Nebraska 68131 Michael S. Reddy (2:363) Department of Periodontics, University of Alabama School of Dentistry, Birmingham, Alabama 35294 Jonathan Reeve (1:585; 2:725) Department of Medicine, Addenbrooke’s Hospital, University of Cambridge, Cambridge CB2 2QQ, United Kingdom
xx
xxi
CONTRIBUTORS
Anthony M. Reginato (1:189) Department of Cell Biology, Harvard Medical School, Boston, Massachusetts 02115; and Arthritis Unit, Massachusetts General Hospital, Boston, Massachusetts 02114 Ian R. Reid (2:553) Department of Medicine, The University of Auckland, Auckland 1, New Zealand B. Lawrence Riggs (2:49) Department of Internal Medicine, Division of Endocrinology and Metabolism, Mayo Clinic and Foundation, Rochester, Minnesota 55905 Rene Rizzoli (1:621) Department of Internal Medicine, Division of Bone Diseases (World Health Organization Collaborating Center for Osteoporosis and Bone Diseases), University Hospital, Geneva CH-1211, Switzerland Pamela Gehron Robey (1:107) Bone Research Branch, National Institute of Dental Research, National Institutes of Health, Bethesda, Maryland 20892 Gideon A. Rodan (1:361) Department of Bone Biology and Osteoporosis Research, Merck Sharp & Dohme Research Laboratories, West Point, Pennsylvania 19486 Clifford Rosen (2:747) Maine Center for Osteoporosis Research and Education, St. Joseph Hospital, Bangor, Maine 04401 Michael Rosenblatt (2:769) Department of Medicine, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts 02215 F. Patrick Ross (1:73) Department of Pathology, Washington University School of Medicine, St. Louis, Missouri 63110 Clinton T. Rubin (1:489) Musculo-Skeletal Research Laboratory, Department of Biomedical Engineering, and the Center for Biotechnology, State University of New York, Stony Brook, New York 11794 Loran M. Salamone (1:741) University of Pittsburgh, Pittsburgh, Pennsylvania 15261 Adina Schneider (2:303) Mount Sinai Hospital, New York, New York 10029 D. J. Schurman (2:385) Division of Orthopedic Surgery, Stanford University School of Medicine, Stanford, California 94305 Ann V. Schwartz (1:795) Department of Epidemiology and Biostatistics, University of California, San Francisco, School of Medicine, San Francisco, California 94143 Ego Seeman (1:771) Austin and Repatriation Medical Centre, University of Melbourne, Heidelberg, Melbourne 3084, Australia
Elizabeth Shane (2:303; 2:327) College of Physicians and Surgeons of Columbia University, New York, New York 10032 Jay R. Shapiro (2:271) Kennedy Krieger Institute, Baltimore, Maryland 21224; and Uniformed Services University, Bethesda, Maryland 20814 Janet Shaw (1:701) Departments of Exercise and Sport Science, University of Utah, Salt Lake City, Utah 84112 Kathy M. Shipp (2:479) Departments of Physical Therapy and Aging Center, Duke University Medical Center, Durham, North Carolina 27710 Shonni J. Silverberg (2:71) Department of Medicine, College of Physicians and Surgeons, Columbia University, New York, New York 10032 Ethel S. Siris (2:603) Columbia University College of Physicians and Surgeons, and Toni Stabile Center for the Prevention and Treatment of Osteoporosis, Columbia – Presbyterian Medical Center, New York, New York 10032 Charles W. Slemenda† (1:809) Department of Medicine, Indiana University School of Medicine, and the Regenstrief Institute for Health Care, Indianapolis, Indiana 46202 Steven R. Smith (2:195) Pennington Biomedical Research Center, Louisiana State University, Baton Rouge, Louisiana 70808 R. L. Smith (2:385) Orthopedic Research Laboratory, Division of Orthopedic Surgery, Stanford University School of Medicine, Stanford, California 94305; and Veterans Affairs Medical Center, Palo Alto, California 94304 Christine M. Snow (1:701) Bone Research Laboratory, Oregon State University, Corvallis, Oregon 97331 MaryFran Sowers (1:535; 1:721) Department of Epidemiology, University of Michigan School of Public Health, Ann Arbor, Michigan 48109 Bonny L. Specker (1:599) Ethel Austin Martin Program in Human Nutrition, South Dakota State University, Brookings, South Dakota 57007 Gary S. Stein (1:21) Department of Cell Biology, University of Massachusetts Medical School, Worcester, Massachusetts 01655 S. Aubrey Stoch (2:769) Department of Medicine, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts 02215 †
Deceased.
xxi
xxii
CONTRIBUTORS
Steven L. Teitelbaum (1:73) Department of Pathology, Washington University School of Medicine, St. Louis, Missouri 63110 Kathy Traianedes (1:373) Department of Medicine and Endocrinology, University of Texas Health Science Center, San Antonio, Texas 78284 Reginald C. Tsang (1:599) Department of Pediatrics, University of Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio 45229 André G. Uitterlinden (1:639) Department of Internal Medicine, Erasmus University Medical School, 3000 DR Rotterdam, The Netherlands Marjolein C. H. van der Meulen (1:471) Sibley School of Mechanical and Aerospace Engineering, Cornell University, Ithaca, New York 14853; and Biomechanics and Biomaterials, Hospital for Special Surgery, New York, New York 10021
Johannes P. T. M. van Leeuwen (1:639) Department of Internal Medicine, Erasmus University Medical School, 3000 DR Rotterdam, The Netherlands Marie Luz Villa (1:569) Department of Medicine, Division of Gerontology and Geriatrics, University of Washington School of Medicine, Seattle, Washington 98104 WenFang Wang (1:189) Department of Cell Biology, Harvard Medical School, Boston, Massachusetts 02115 Kristine M. Wiren (1: 339) Bone and Mineral Unit, Oregon Health Sciences University, and the Portland Veterans Affairs Medical Center, Portland, Oregon 97201 Christian Wüster (2:747) Department of Medicine, Novo Nordisk Pharma, 55127 Mainz, Germany
xxii
xxiii
CHAPTER 1 Chapter Title
Preface
It will come as no surprise to the reader that this second edition of Osteoporosis substantially outweighs its predecessor. The reason for this increase is the astonishing progress in osteoporosis medicine that has taken place during the 5 years since publication of the first edition. Major advances have occurred at both the basic science and clinical levels, and we have sought to incorporate them into the second edition. Some illustrative areas where new insights have emerged are skeletal differentiation and regulation, particularly regarding complementary actions of Hedgehog and PTH-related proteins; the RANK/RANK-ligand osteoprotegerin system as a central feature of osteoclast biology; the skeletal phenotypes of mouse gene knockout models; the discovery of a second estradiol receptor which is relatively enriched in bone; and the molecular basis of bisphosphonate action. These developing areas represent new and important conceptual themes that were unknown in 1995, but which have been accorded detailed consideration in the second edition. Many of the chapters devoted to basic science emphasize new concepts that offer novel molecular targets for future therapeutics. Great progress has also been made in clinical practice. When the first edition was published, therapeutic choices were confined to several antiresorptive drugs. Few controlled clinical trials had been conducted for any agent, and those that had been published relied on the surrogate end point of bone mineral density, rather than on fracture incidence. Large randomized controlled trials using fracture incidence as the primary outcome have now become the industry standard, and results of such trials have led to the registration of new agents for the prevention and treatment of postmenopausal osteoporosis, glucocorticoid-associated osteoporosis, and osteoporosis in men. Perhaps most exciting, the recent demonstration that parathyroid hormone substantially increases bone mineral density and reduces fracture incidence validates the concept of skeletal anabolic
therapy and offers for the first time a potential approach to the eventual cure of osteoporosis. In response to these and other developments, new chapters have been introduced and others have been expanded, updated, or divided. New chapters include contributions on the developmental biology of bone, gene knockout models, major European epidemiological fracture studies, micro-CT assessment of bone architecture, evidence-based analysis of osteoporosis therapy, regulatory considerations for design of osteoporosis trials, skeletal effects of phytoestrogens and selective estradiol receptor modulators, and novel approaches to osteoporosis therapeutics. In the interest of maintaining freshness of ideas, we initiated our own remodeling process by creating a modest degree of authorship turnover. We thank previous contributors for their excellent work and forewarn them that we may call on them again for future editions. We note with sadness the passing of two prominent members of the osteoporosis community who were involved with the first edition of this book, Charles W. Slemenda and Louis V. Avioli. Despite his young age, Charlie provided many valuable insights into multiple aspects of osteoporosis. Lou Avioli elected not to write a chapter, but we benefited greatly from his advice regarding content and authors as well as from his enthusiasm for the project. We are pleased to acknowledge once again the efforts of Dr. Jasna Markovac of Academic Press for her unflagging interest and inspiration for this project. We also greatly appreciate the Academic Press editorial staff, particularly Mica Haley and Jenny Wrenn, for their cheerful hard work and enthusiasm no matter how short the deadline or late the chapter. ROBERT MARCUS DAVID FELDMAN JENNIFER KELSEY Stanford, California
xxiii
CHAPTER 1
The Bone Organ System Form and Function CASSANDRA A. LEE AND THOMAS A. EINHORN Department of Orthopedic Surgery, Boston University School of Medicine, Boston, Massachusetts 02118
I. II. III. IV.
Introduction Composition and Organization of Bone Cellular Control of Bone Homeostasis Bone Modeling and Remodeling
V. Bone Biomechanics VI. Summary References
I. INTRODUCTION
its simplest sense, Wolff’s law suggests that “form follows function’’ [3]. To fulfill these structure/function relationships adequately, bone is constantly being broken down and rebuilt in a process called remodeling. The cellular link between bone resorbing cells or osteoclasts, and bone forming cells or osteoblasts, is known as coupling. How bone resorption and bone formation are linked is not entirely understood (see also Chapter 12), but the consequences of accentuating one or the other preferentially leads to disease. Too much bone resorption at the expense of formation results in osteoporosis, a loss of bone strength and integrity, resulting in fractures after minimal trauma. However, under normal states of bone homeostasis, the remodeling activities in bone serve to remove bone mass where the mechanical demands of the skeleton are low and form bone at those sites where mechanical loads are transmitted repeatedly. Hence, bone is a well-designed organ system whose ability to maintain itself depends on the integrated processing of external mechanical input and physiological signals from the systemic environment and the transduction of these demands into cellular and chemical events.
Bone is a vital, dynamic connective tissue that has evolved to reflect a balance between its two major functions, provision of mechanical integrity for locomotion and protection and involvement in the metabolic pathways associated with mineral homeostasis. In addition, bone is the primary site of hemopoiesis, and recent findings support its important role as a component of the immune system [1,2]. Since the observations of Galileo, it has been assumed that the inherent architecture of bone is influenced by the mechanical stresses associated with normal function. A more formal definition of these structure/function relationships was provided by German anatomists and engineers during the late 19th century in what has since been known as Wolff’s law [3]. The tenets of Wolff’s law were based on a recognition of the correlation between the patterns of trabecular alignment in bone and the directions of the principal stresses, which were estimated to occur during normal skeletal function. Under these physiological conditions, the structure/function relationships observed in bone, coupled with its role in maintaining mineral homeostasis, strongly suggest that it is an organ of optimum structural design. In
OSTEOPOROSIS, SECOND EDITION VOLUME 1
3
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
4
LEE AND EINHORN
II. COMPOSITION AND ORGANIZATION OF BONE Bone is a composite material composed of an organic and an inorganic phase. By weight, approximately 70% of the tissue is mineral or inorganic matter, water comprises 5 to 8%, and the organic or extracellular matrix makes up the remainder. Approximately 95% of the mineral phase is composed of a specific crystalline hydroxyapatite, and this is impregnated with impurities, which make up the remaining 5% of the inorganic phase. Ninety-eight percent of the organic phase is composed of type I collagen and a variety of noncollagenous proteins; cells make up the remaining 2% of this phase [4].
A. Organic Phase The organic phase of bone plays a wide variety of roles, determining the structure and the mechanical and biochemical properties of the tissue. Growth factors and cytokines, and extracellular matrix proteins such as osteonectin, osteopontin, bone sialoprotein, osteocalcin, proteoglycans, and other phosphoproteins and proteolipids, make small contributions to the overall volume of bone and major contributions to its biologic function (see also Chapter 4). Collagen is a ubiquitous protein of extremely low solubility, which consists of three polypeptide chains composed of approximately 1000 amino acids each. It is the major structural component of the bone matrix; constructed in the form of a triple helix of two identical 1(I) chains and one unique 2(I) chain cross-linked by hydrogen bonding between hydroxyproline and other charged residues. This produces a very rigid linear molecule 300 nm in length. Each molecule is aligned with the next in a parallel fashion in a quarter-staggered array to produce a collagen fibril. The collagen fibrils are then grouped in bundles to form the collagen fiber. Within the collagen fibril, gaps, called “hole zones,” exist between the ends of the molecules. In addition, “pores’’ exist between the sides of parallel molecules (Fig. 1). Noncollagenous proteins or mineral deposits can be found within these spaces, and mineralization of the matrix is thought to be initiated in the hole zones. Several noncollagenous proteins have been described in bone. One of the more extensively studied of these in bone is osteocalcin or bone-carboxyglutamic acid-containing protein (bone Gla protein). This is a small (5.8 kDa) protein in which three glutamic acid residues are carboxylated as a result of a vitamin K-dependent post translational modification. The carboxylation of these residues confers on this protein calcium and mineral binding properties. Osteocalcin accounts for 10 to 20% of the noncollagenous protein present in bone and is closely associated with the mineral phase. While the function of this bone-specific protein is
FIGURE 1 Collagen fiber and fibril structure showing putative locations of pores and hole zones. Reprinted with permission from T. A. Einhorn, Bone metabolism and metabolic bone disease, In “Orthopaedic Knowledge Update 4” (J. W. Frymoyer, ed.), pp. 69 – 88. Amer. Acad. Orthop. Surg., Rosemont, IL (1993).
not known, it is thought to play some role in attracting osteoclasts to sites of bone resorption. It may also regulate the rate of mineralization or the final shape assumed by mineral crystals. Other noncollagenous proteins found in bone may also be important to their calcium and mineral-binding properties. In addition, several of the bone matrix proteins, such as osteopontin, bone sialoprotein, bone acidic glycoprotein, thrombospondin, and fibronectin, contain arginine – glycine – aspartic acid (RGD) sequences. Such amino acid sequences, characteristic of cell-binding proteins, are recognized by a family of cell membrane proteins known as integrins. The integrins span the cell membrane and provide a link between the extracellular matrix and the cytoskeleton of the cell. Integrins on osteoblasts, osteoclasts, and fibroblasts provide a means for anchoring these cells to the extracellular matrix. Once anchored, the cells are then enabled to express their phenotype and conduct the type of activities that characterize their funcions [5]. Growth factors and cytokines such as transforming growth factor- (TGF-), insulin-like growth factor (IGF), osteoprotegerin (OPG), the tumor necrosis factors (TNFs), the interleukins, and the bone morphogenetic proteins (BMPs 2 – 10) are present in very small quantities in bone matrix. Such proteins have important effects regulating bone cell differentiation, activation, growth, and turnover (see Chapter 14). It is also likely that these growth factors serve as coupling factors that link the processes of bone formation and bone remodeling (Table 1).
5
CHAPTER 1 The Bone Organ System
TABLE 1
Noncollagenous Proteins of the Extracellular Matrix
Structural matrix molecules Osteocalcin
Vitamin K dependent. Made by osteoblasts. May inhibit mineral deposition. Regulates activity of osteoclasts and their precursors. May mark turning point between bone formation and bone resorption. Restricted to the osteogenic lineage
Osteopontin
Contains RGD site. Anchors osteoclasts to bone. Supports cell attachments. Possibly inhibits mineralization. May regulate tissue repair and proliferation. Highly expressed in bone and inflammatory tissue
Bone Sialoprotein
Made by osteoblasts and hypertrophic chondrocytes. May initiate mineralization. Supports cell attachment. Binds Ca2 with high affinity. Restricted to skeletal lineage
Matrix Gla protein
Vitamin K dependent. May inhibit mineralization. May function in cartilage metabolism. Expressed in a variety of connective tissues
Chondroiton sulfate proteoglycan I (decorin)
Regulation of collagen fibrillogenesis. Found in extracellular matrix space
Chondroiton sulfate proteoglycan II (biglycan)
May bind to collagen. Regulates mineralization in vitro. Pericellular location
Fibronectin
Contains RGD site. Binds to numerous cell types, fibrin, heparin, gelatin, and collagen. Expressed in variety of connective tissues
Thrombospondin
Contains RGD site. Functions in cell attachment. Binds heparin, platelets, type I and V collagen, thrombin, fibrinogen, laminin, plasminogen, plasminogen activator inhibitor, histidine-rich glycoprotein. Expressed in variety of connective tissues
Osteonectin
Strong affinity for Ca2 and hydroxyapatite. Binds to growth factors. Potential association with osteoblast growth and/or proliferation. May play role in matrix mineralization. Expressed in a variety of connective tissues
Enzymatic matrix modifiers Matrix metalloproteinase (MMP) 9
Regulates growth plate angiogenesis and apoptosis. Degrades components of the ECM. Cleaves type IV, V, and XI collagens, elastin
Lysyl oxidase
Copper-dependent extracellular enzyme that catalyzes oxidative deamination of elastin and collagen precursors, which in turn spontaneously cross-link collagen and elastin, thereby forming a mature and functional ECM
Stromelysin
Is MMP3. Degrades most constituents of the ECM. Activates other MMPs
Latent morphogens and cytokines Bone morphogenetic protein family
Osteoinductive effects, promotes osteogenesis, chondrogenesis, and some induce other tissue types
Tissue growth factor family
Mitogen for periosteal osteoprogenitors and marrow stromal cells. Direct stimulatory effect on bone collagen synthesis. Decreases bone resorption by inducing apoptosis of osteoclasts. Inhibits osteoblast differentiation in vitro
Fibroblast growth factor
Angiogenic properties. Important with neovascularization and wound healing. Important in healing and bone repair. Promotes bone cell replication
B. Inorganic Phase
C. Organization of Bone
The inorganic component of bone is composed mainly of a calcium phosphate mineral analogous to crystalline calcium hydroxyapatite. This apatite is present as a platelike crystal, which is 20 to 80 nm long and 2 to 5 nm thick. The small amounts of impurities in hydroxyapatite, such as carbonate, which can replace the phosphate groups, or chloride and fluoride, which can replace the hydroxyl groups, may alter certain physical properties of the crystal, such as its solubility [6]. These altered properties may impart important biologic effects that are critical to normal function.
The skeleton is composed of two parts: an axial skeleton, which includes the vertebrae, pelvis, and other flat bones such as the skull and sternum, and an appendicular skeleton, which includes all of the long bones. The long bones are divided into three parts: the epiphysis, metaphysis, and diaphysis. The epiphysis is that portion of the long bone found at either end and develops from a center of ossification that is distinct from the rest of the long bone shaft. It is separated from the rest of the bone by a layer of growth cartilage. This growth cartilage is known as the physis. The metaphysis is the zone between the physis and the
6
LEE AND EINHORN
FIGURE 2
Woven bone. Note arrangement of this bone tissue in which there is active bone formation (top), no particular lamellar organization of the tissue, and a high degree of cellularity.
central portion of the long bone shaft (known as the diaphysis). The metaphysis is the region where remodeling of the bone takes place during growth and development. The diaphysis comprises the majority of the length of a long bone. At the microscopic level, bone appears as an extremely well-organized tissue whereby mineral impregnates the collagen fibril array in such a way as to provide mechanical properties, which, in tension, are nearly as strong as cast iron. There are two types of bone tissue: woven bone and lamellar bone. Woven bone is considered immature or primitive bone and is normally found in the embryo, the newborn, in fracture callus, and in certain metaphyseal regions of the growing skeleton. It is also found in certain bone tumors, in patients with osteogenesis imperfecta, and in patients with Paget disease. Lamellar bone, however, is a more mature bone that results from the remodeling of woven bone or preexisting bone tissue. Woven, or primary bone, is a coarse-fibered tissue that does not show any uniform orientation of the collagen fibers (Fig. 2). It has more cells per unit volume than lamellar bone, and its mineral content and cells are randomly arranged. Newly formed woven bone, which is not as well mineralized as mature lamellar bone, contains particles with a smaller average crystal size. Moreover, the relative disorientation of the collagen fibers gives it isotropic mechanical characteristics; i.e., when tested, the mechanical behavior of woven bone is similar regardless of the orientation of the applied forces. At the time of birth, all bone in the body is woven. However, beginning at approximately 1 month of age, lamellar bone begins to develop. By 1 year
of age, lamellar bone will have effectively replaced much of the woven bone, as the latter is resorbed. By the age of 4, most of the bone in the body will be lamellar. Lamellar bone is a highly organized material with respect to the stress orientation of the collagen fibers (Fig. 3). As a result of this structural organization, lamellar bone exhibits anisotropic properties; i.e., the mechanical behavior of lamellar bone differs depending on the orientation of the applied forces. Moreover, its ability to resist loads is greatest when the forces are directed in a parallel fashion to the longitudinal axis of the collagen fibers. Anatomically, woven and lamellar bone are organized into trabecular (spongy or cancellous) and cortical (dense or compact) bone compartments. Cortical bone has four times the mass of trabecular bone, although the metabolic turnover rate of trabecular bone is much higher than that of cortical bone (bone turnover is a surface event, and trabecular bone has a greater surface area than cortical bone). Trabecular bone is found principally at the ends of long bones and in cuboid bones, such as the vertebrae (Fig. 4). The internal beams or plates of trabecular bone form a three-dimensional branching lattice, which is oriented along lines of stress. Trabecular bone is subject to a complex set of stresses and strains, although it is best designed for resisting of compressive loads. Cortical bone appears as an entirely different structure. It is solid and arranged not as interconnecting plates, but rather as cylinders, which are ellipsoid in cross section (Fig. 5). Cortical bone is usually subject to bending and torsional forces, as well as compressive loads.
7
CHAPTER 1 The Bone Organ System
FIGURE 3
Lamellar bone. Note organization of this tissue in which there is a well-delineated orientation of the collagen fibers and a coordinated alignment of the cells.
Haversian bone is the most complex type of cortical bone. It is composed of vascular channels surrounded circumferentially by lamellae of bone (Fig. 6). This complex arrangement of bone around the vascular channel is called
FIGURE 4
the osteon. The osteon is an irregular, branching, and anastomosing cylinder composed of a neurovascular canal surrounded by cell-permeated layers of bone matrix. Osteons are usually oriented in the long axis of the bone and are the
Trabecular bone. Note the perpendicular orientation between the horizontal and the vertical trabeculae.
8
LEE AND EINHORN
FIGURE 5 Cross-section of cortical bone from the middiaphysis of the tibia. Note that cortical bone is a solid structure arranged not as interconnecting plates, but as ellipsoid cylinders. On the inner surface (endosteum), there is a structure that resembles that of trabecular bone.
FIGURE 6
Close-up view of a section of cortical bone. Note the distribution of vascular channels forming the osteons.
9
CHAPTER 1 The Bone Organ System
FIGURE 7 Scanning electron micrograph of cortical bone showing individual osteons, surrounded by lamellar bone, which is impregnated with well-aligned cellular lacunae.
major structural units of cortical bone. They are connected to one another by Volkmanns canals, which are oriented perpendicularly to the osteon. Thus, cortical bone is a complex structure composed of many adjacent osteons and their interstitial and circumferential lamellae (Fig. 7). Most vessels in haversian canals have the ultrastructural features of capillaries, although some smaller-sized vessels may resemble lymphatic vessels. When examined histologically, these small vessels contain only precipitated protein; their endothelial walls are not surrounded by a basement membrane. The basement membrane of capillary walls may function as a rate-limiting or selective ion-limiting transport barrier because all material traversing the vessel wall must go through the basement membrane. The presence of this barrier is particularly important in calcium and phosphorous ion transport to and from bone. It is also important in explaining the response of bone to mechanical loads. The capillaries in the central canals are derived from the principal nutrient arteries of the bone: the epiphyseal and metaphyseal arteries [6]. The periosteum lines the outer surface of bone. It is composed of two layers. The outer fibrous layer is in direct contact with muscle and other soft tissue elements and is populated by undifferentiated fibroblast-like cells. The inner layer is known as the cambium layer and it is populated by fibroblast-appearing cells, many of which are committed
progenitors of chondrocytes and osteoblasts (Fig. 8).This layer contributes to appositional bone growth during bone development and is responsible for the expansion of the diameters of the long bones with aging.
III. CELLULAR CONTROL OF BONE HOMEOSTASIS A. Bone Cells Bone metabolism is regulated by bone cells, which respond to various environmental signals, including chemical, mechanical, electrical, and magnetic stimuli. In general, specific responses are governed by cellular receptors found on the membrane of the cell or intracellularly. Cell membrane receptors bind the exogenous signal and transfer the information across the cell’s cytoplasm to the nucleus through a series of interactions that involve a complex set of transduction mechanisms. Intracellular receptors (cytoplasmic or nuclear) bind the stimulus (usually a steroid hormone, which has crossed the cell membrane and entered the cell) and then translocate that effector to the nucleus where the steroid – receptor complex binds to a specific DNA promotor sequence of a gene. Three cell types are found in bone, osteoblasts, osteoclasts, and osteocytes.
10
LEE AND EINHORN
FIGURE 8
Close-up view of the periosteum of a long bone. The darker staining material at the lower portion of the figure is mineralized cortical bone. Above this is the periosteum, which consists of two layers. The outer layer shows elongated fibroblast-like cells embedded in a fibrous-like tissue. The inner layer, known as the cambium layer, is composed of a higher density of cells, which are slightly more plump and are embedded in a loose connective tissue.
Osteoblasts, generally regarded as bone-forming cells, govern bone metabolism (see Chapter 2). Their most obvious function is to synthesize osteoid, the protein component of bone tissue, but they also initiate bone resorption by
FIGURE 9
elaborating various neutral proteases (Fig. 9) [8]. The proteases remove surface osteoid, after which other cells participate in bone resorption. Because osteoblasts contain the receptors for most chemical mediators of bone metabolism,
Low-power view of osteoblasts lining the bone surface.
11
CHAPTER 1 The Bone Organ System
FIGURE 10
including bone resorption, they play a critical role in the regulation of bone turnover. Osteocytes are abundant in mineralized bone matrix, but their function is poorly understood. Evidence suggests that they may receive mechanical input signals and transmit these stimuli to other cells in bone (Fig. 10). Osteoblasts and osteocytes are both derived from the same mesenchymal stem cell precursor (found in bone marrow stroma, periosteum, soft tissues, and possibly peripheral blood vessel endothelium). Once an osteoblast has synthesized osteoid and mineralized it, the osteoblast becomes an osteocyte. Evidence suggests that the lineage of the cell type can be identified by the expression of specific cell surface antigens, the expression of alkaline phosphatase when the cell is in its secretory osteoblastic phase, and the loss of alkaline phosphatase activity when the cell evolves to an osteocyte [8]. Osteoclasts, the active agents in bone resorption, are ultimately responsible for the remodeling of bone (see Chapter 3). In cortical bone, they are found at the apex of the classical “cutting cone’’ (Fig. 11, see also color plate). In trabecular bone they create the resorptive cavities, known as Howship’s lacunae (Fig. 12), seen on bone surfaces undergoing active remodeling. Osteoclasts are multinucleated, but their progenitors are hemopoietic mononuclear cells.
Differentiation toward an osteoclastic phenotype occurs early in the development of these cells.
B. Cellular Mechanisms in Bone All bone surfaces are continuous and lined by resting osteoblastic lining cells. On close inspection, it is possible to observe small intercellular gaps between the cells and their cytoplasmic processes. The lining osteoblasts are in communication with osteocytes through cell processes within the canaliculi that form gap junctions. (Fig. 13) Rapid fluxes of bone calcium across these junctions may be involved in the transmission of information between osteoblasts on the bone surface and osteocytes within the structure of bone itself. Although the cellular layer protects the bone from the extracellular fluid space, the osteoblasts on the bone surface are in direct chemical contact with the osteocytes within the mineralized bone by virtue of these cellular processes within the canaliculi. This organizational structure is consistent with the concept that bone cells are in intimate communication with each other and that osteoblasts receive the majority of local and systemic signals and then transmit them to other cells in bone. Conversely, strain-generated signals such as streaming potentials could
12
LEE AND EINHORN
FIGURE 11
A cutting cone in cortical bone. Note that it is composed of an osteoclast, which appears to be tunneling through bone, and this cell is followed closely by a group of endothelial cells that ultimately transition to osteoblasts. (See also color plate.)
be perceived by osteocytes, and their regulatory information can be passed on to the osteoblasts. Osteoclasts at specific bone sites are activated only after disruption of the osteoid layer that covers the bone sur-
FIGURE 12
faces, a bone lining osteoblast-mediated effect. This exposure of the underlying mineralized matrix may be caused by the degradation of surface osteoid by the neutral proteases elaborated by flat, elongated osteoblasts or by the
Low-power view of osteoclasts forming Howship’s lacunae.
13
CHAPTER 1 The Bone Organ System
This molecule, related to the TNF family, upregulates osteoclast development and increases the activity of mature osteoclasts [9]. It may also play a role in immunologic tissue by regulating the interaction between T-cells and dendritic cells in vitro [10]. In vivo, RANKL exists in the cell as a ligand, anchored to the surface of an osteoblast or a bone marrow stromal cell and interacts with RANK on a preosteoclast. RANKL also exists in a secreted form that binds directly to the preosteoclast [11]. It is here that it promotes osteoclast differentiation from hemopoietic precursors. Once this reaction occurs, the preosteoclast will differentiate to a mature osteoclast only if the macrophage colony-stimulating factor (M-CSF) is present. Various other factors, such as PTH, interleukin-11, prostaglandin E2 (PGE2), and 1,25(OH)2D, upregulate RANKL on the surface of osteoblasts, hence increasing the development and activity of osteoclasts. The interaction between RANKL and its preosteoclast receptor is controlled by osteoprotegerin (OPG), a secreted protein. OPG is a soluble decoy protein that may modulate communication between osteoblasts and osteoclasts, thus playing a major role in bone homeostasis. Its main function is to halt bone resorption by inhibiting osteoclast formation via interruption of RANKL. In vitro, studies have shown that OPG can induce apoptosis of osteoclast-like cells (see Chapter 3). The proposed mechanism for OPG function is via reduction/disruption of formation of the F-actin ring in isolated osteoclasts. The F-actin ring is a cytoskeletal structure correlated with bone resorption [12].
IV. BONE MODELING AND REMODELING A. Bone Modeling FIGURE 13
Osteocyte (OC) cytoplasmic process in contact with the secreting surface of an osteoblast (OB). TEM 16,125. Reprinted with permission from C. Palumbo, Morphological study of intercellular junctions during osteocyte differentiation, In “Bone” (R. Baron, ed.), pp. 401 – 409. Pergamon Press, Elmsford, NY.
contraction of osteoblasts in response to stimulation by parathyroid hormone (PTH), 1,25-dihydroxyvitamin D, or prostaglandins of the E series. This contraction allows osteoclasts to gain access to the mineralized bone. Three recently discovered molecules that possibly play a key role in the coordinated maintenance of bone homeostasis are osteoprotegerin (OPG), RANK, (receptor activator of NF-B), and its ligand, RANKL, (see Chapter 3). The ligand (RANKL) has proven to be an essential factor in osteoclast differentiation by promoting osteoclast formation.
Bone formation begins in utero and continues throughout adolescence until skeletal maturity (see Chapter 5). Long bones form by two mechanisms. Endochondral ossification occurs in the long bones or the appendicular skeleton. It involves the differentiation of mesenchymal lineage cells to chondroblasts and then chondrocytes with the synthesis of a proteoglycan-rich, type II collagen-based extracellular matrix. This matrix is then modified biochemically by enzymes elaborated by hypertrophic chondrocytes to produce an environment that will permit the deposition of calcium. Once the extracellular matrix is calcified, it becomes a target for blood vessel invasion, and with this angiogenic response comes osteoclasts (which degrade the calcified cartilage) and osteoblast precursors. The calcified cartilage first formed is known as the primary spongiosum, and the bone that is laid down upon this tissue is known as secondary spongiosum. This secondary spongiosum is in the form of woven bone.
14
LEE AND EINHORN
Intramembranous bone formation occurs in the flat bones of the skeleton such as the skull and pelvis. It involves the direct formation of bone tissue by cells of the mesenchymal lineage, which have already undergone a biological commitment to the formation of osteoblastic cells. Bones also grow in length by endochondral ossification and in width by intramembranous or subperiosteal new bone formation.
B. Bone Remodeling Following skeletal maturity, bone continues to remodel throughout life and adapt its material properties to the mechanical demands placed upon it (see Chapters 12 and 15). Remodeling also has the function of maintaining the biomechanical competence of the skeleton by preventing the accumulation of fatigue damage and maintaining a tissue whose components are available for mineral homeostasis. Remodeling is thus a continuous process whereby there is a constant removal and replacement of whole volumes of bone tissue, a process conducted by osteoclasts and osteoblasts on bone surfaces, which, together with their precursor cells, form the bone remodeling system. A knowledge of bone remodeling is essential to the understanding of the pathophysiology of osteoporosis. The bone remodeling process governs the way bone is replaced, gained, or lost at specific sites and the way cumulative effects determine the three-dimensional structure of bone. The rate of turnover also determines the age of the bone tissue, and various physical and chemical properties of the bone are dependent on age and function [13]. According to bone histologists (see Chapter 15), the skeleton is composed of individual structural units or bone metabolic units (BMU) [6]. Cortical bone constitutes approximately 80% of the skeletal mass and trabecular bone approximately 20%. Bone surfaces may be undergoing formation or resorption, or they may be inactive. These processes occur throughout life in both cortical and trabecular bone. Bone remodeling is a surface phenomenon and it occurs on periosteal, endosteal, haversian canal, and trabecular surfaces. The rate of cortical bone remodeling, which may be as high as 50% per year in the midshaft of the femur during the first 2 years of life, eventually declines to a rate of 2 to 5% per year in the elderly. Rates of remodeling in trabecular bone are proportionally higher throughout life and may normally be 5 to 10 times higher than cortical bone remodeling rates in the adult [13]. The BMU of cortical bone is the osteon or haversian system, a cylinder of about 200 to 250 m in diameter running parallel to the long axis of the bone. As described earlier, the canals are connected to each other by transverse Volkmann’s canals and periodically either divide or reunite to form a branching network. Osteons form approximately two-thirds of cortical bone volume, a
proportion that falls with age, with the remainder consisting of interstitial bone representing the previous generations of osteons. There are also subperiosteal and subendosteal circumferential lamellae. In trabecular bone, the BMU is constructed differently. In two-dimensional sections, the trabecular surfaces are shaped like thin crescents about 600 m long and about 60 m in depth. Three-dimensionally, these BMUs are actually larger than they appear in two-dimensional histological sections with prolongations in different directions that interlock with adjacent BMUs [14]. These BMUs follow the same shape as the trabecular surface, most of which are concave toward the marrow. Under normal conditions, the remodeling process of resorption followed by formation is closely coupled and results in no net change in bone mass. As such, the BMU consists of a group of cells that participate in remodeling in a concerted and coordinated fashion. Cortical bone remodeling proceeds via cutting cones (Fig. 11) and is similar to processes in other hard biological tissues. Cuttings cones, or sheets of osteoclasts, bore holes through the hard bone, leaving tunnels, which appear in cross section as cavities. The head of the cutting cone consists of osteoclasts that resorb the bone. Following closely behind the osteoclast is a capillary loop and a population of endothelial cells and perivascular mesenchymal cells that are progenitors for osteoblasts and soon begin to lay down osteoid and refill the resorption cavity. By the end of the process, a new osteon will have been formed. Trabecular bone remodeling proceeds on the surface of bone whereby osteoclasts resorb bone at specific sites (Fig. 14). These areas are then filled in with newly formed osteoid. The mechanisms that control the activity and site specificity of this process are unknown. Cortical bone remodeling occurs in discrete temporal foci that are active for about 4 to 8 months [13]. Mononuclear precursor cells proliferate into a team of new osteoclasts that initiate the cutting cone. The osteoclasts then are removed (or disappear), and there is a quiescent interval or “reversal phase’’ during which time the newly resorbed area of bone is smoothed and a layer of cement substance (osteoid) is deposited. The team of osteoblasts that follow the osteoclast attempts to replace exactly as much bone as has been removed. If it is successful in this venture, bone homeostasis will have been maintained. If not, osteoporosis may result. One of the conditions that affects the ability of the bone-resorbing and bone-forming cells to be coupled and result in no net change in bone mass is aging. During the process of remodeling, changes in the amount of bone are slow and changes in the shape of bone are barely perceptible. According to the model proposed by Parfitt, the normal remodeling sequence in bone follows a scheme of quiescence, activation, resorption, reversal, formation, and return
15
CHAPTER 1 The Bone Organ System
FIGURE 14
High-power scanning electron micrograph of an intersection of bone trabeculae showing multiple resorption sites. (Courtesy of S. Goldstein, University of Michigan.)
to quiescence (see Chapter 15). In the adult, approximately 80% of trabecular and approximately 95% of intracortical bone surfaces are inactive with respect to bone remodeling [15,16]. The surface of bone is covered by a layer of thin flattened lining cells approximately 15 m in diameter, which arise by terminal transformation of osteoblasts. Between these lining cells and bone is a layer of unmineralized osteoid. These lining cells have receptors for a variety of substances, which are important for initiating bone resorption (PTH, PGE2), and may respond to such substances by resorbing this surface osteoid, which is covering the bone. In doing so, mineralized bone will be exposed and the activation sequence of bone remodeling may be initiated. The conversion of a small area of bone surface from quiescence to activity is referred to as activation. The cycle of this response begins with the recruitment of osteoclasts followed by the initiation of mechanisms for their attraction (chemotaxis) and attachment to the bone surfaces. Several known growth factors may be active in promoting chemotaxis. In addition, several proteins are known to be attachment factors for osteoclasts, such as those that contain the RGD amino acid sequences as noted earlier. Osteopontin, osteocalcin, and osteonectin
may be important proteins in this process (see Chapters 3 and 4). In the adult skeleton, activation occurs about every 10 s. For intracortical remodeling, osteoclast precursors travel to the site of activation via the circulation, gaining access to the site by either a Volkmann or a haversian canal. In trabecular remodeling, activation occurs at sites that are apposed to bone marrow cells. The osteoclast is a very mobile cell that can resorb bone over an area approximately two to three times the area with which it is in direct contact. In cortical bone, the osteoclast and the cutting cone travel at a speed of about 20 or 40 m per day, roughly parallel to the long axis of the bone and about 5 to 10 m per day perpendicular to the main direction of advance [7]. In trabecular bone, osteoblasts erode to a depth of about two-thirds of the final cavity; the remainder of the cavity is eroded more slowly by mononuclear cells [17]. The reversal phase is a time interval between the completion of resorption and the initiation of bone formation at a particular skeletal site. Under normal conditions, it lasts about 1 to 2 weeks. The appearance of new osteoblasts at the base of the resorption cavity depends on chemotaxis for these osteoblasts and their progenitors, as
16
LEE AND EINHORN
well as conditions that stimulate proliferation. Hence chemotaxis, attachment, proliferation, and differentiation occur in a stepwise and concerted fashion in order for bone formation ultimately to take place. Electron microscope studies of the reversal phase have shown the release of osteocytes and the appearance of mononuclear phagocytes, some of which appear to be smoothing over the ragged surface left by the resorption process [18]. Because the coupling mechanism appears to be important to this bone remodeling process, the release of proteins from the bone matrix may be important in coordinating the activity between osteoclasts and osteoblasts. Unlike bone resorption, bone formation is a two-step process. First, the osteoid is synthesized and laid down at specific sites. Following this, the osteoblast must mineralize this newly formed protein matrix. The new matrix begins to mineralize after about 5 to 10 days from the time of deposition and, as a result, matrix apposition and mineralization are systematically out of step as the osteoid seam first increases and then decreases in thickness. The rate of mineral apposition can be measured directly in vivo after double tetracycline labeling; the mean distance between fluorescent bands divided by the time interval between the midpoints of the tetracycline labels can be used to calculate the mineral apposition rate [15].
V. BONE BIOMECHANICS A. Basic Concepts A basic knowledge of biomechanics must begin with an understanding of the terms stress and strain. These terms are used to describe the phenomenon whereby, when a force is applied to bone, the bone will not only be deformed from its original shape, but an internal resistance will be generated to counter the applied force. This internal resistance is known as stress and it is equal in magnitude but opposite in direction to the applied force. Because stress is distributed over the entire cross-sectional area of a section of bone, it is expressed as units of force per unit area. Strain is a term used to describe the changes in shape that bone experiences when it is subjected to an applied force. Strain is dimensionless and is therefore reported as a fraction or a percentage. It is equivalent to the change in length divided by the original length of the section of bone. Although an applied force can be directed at bone from any angle producing any set of complex stress patterns, all stresses can be resolved into three types: tension, compression, and shear Tension is produced in bone when two forces are directed away from each other along the same straight line. The resistance to a loading situation of this type is produced by intermolecular attractive forces, which
prevent the bone from being pulled apart. An example of a tensile force producing a failure in bone occurs when a tendon or ligament that is inserted into bone undergoes acute loading and, instead of failing within its own substance (tearing), it detaches itself from the bone by actually pulling a piece of bone off with it. Compression results from two forces that are directed toward each other along the same straight line. The common vertebral compression fracture sustained in osteoporotic patients is an example of the failure of bone as a result of this type of loading configuration (Fig. 15). Finally, when two forces are directed parallel to each other but not along the same line, shear stresses are produced. In nature, the three basic stress types can combine as a result of a variety of complex loading configurations and lead to different fracture patterns (Fig. 16). Bending results from a combination of tensile and compressive forces and leads to clinical fractures that show a predominantly transverse fracture pattern. Torsion or twisting produces shear stresses along the entire length of a bone and can result in spiral fractures. Comminuted fractures appear as shattered bones and this occurs because the amount of energy transmitted to the bone is so great that a variety of fracture lines have to be propagated in order for it to be dissipated. Because stress and strain are properties related to the quality of the tissue experiencing the load, the quality of bone can influence the magnitude of the stresses and strains generated. In normal, well-mineralized bone, physiological stresses will generally result in small strains. In poorly mineralized tissue, such as osteomalacic bone, bone will experience larger strains in response to the same stress. Moreover, because in nature, forces are applied to bone not only from perpendicular and horizontal directions but also from oblique angles, conditions will arise in which a variety of complex mechanical relationships will be generated. When bones show different mechanical properties if loads are transmitted from different directions, they express a property known as anisotropy [22]. In general, bone resists loads best when the loads are oriented in the direction of customary loading. For example, the femur is much better adapted to resisting compressive loads than bending loads [20]. Thus, the same stresses generated in the femur if one were to jump from a 4foot wall and land on his feet (compressive stresses) may fracture the femur if they were oriented from a transverse direction(bending stresses). In the proximal femur (hip joint area), loads are best resisted when they are transmitted along lines that are parallel to the trabecular systems [21].
B. Biomechanical Properties The biomechanical properties of bone can be described at two levels. First, the material properties of bone are
17
CHAPTER 1 The Bone Organ System
FIGURE 15
Sagittal section of a vertebral compression fracture. Note collapse of the vertebra through the central trabecular
section.
FIGURE 16
Fracture patterns in a cylindrical section of bone subjected to different complex loading configurations. (a) Pure tensile loading produces a purely transverse fracture. (b) Pure compressive loading produces an oblique fracture. (c) Torsional loading produces a spiral fracture. (d) Bending, a combination of compressive and tensile loading, produces an essentially transverse fracture with a small fragment on the concave side. (e) Bending in which compressive forces make a greater contribution to the complex loading configuration, leading to a transverse fracture with a larger fragment on the concave side.
defined by the tissue level qualities of the tissue, which are independent of its structure or geometry. Second, there are structural properties in bone, which are manifest when bone functions as a whole anatomical unit. Classically, the material properties of bone are defined by performing standardized mechanical tests on uniform, machined specimens of bone. Structural properties are determined on whole sections of bone whose normal geometry has been maintained. It is important to recognize that when a patient sustains a fracture, that event most likely represents the failure of bone at either the material or the structural level or both. When a uniform section of bone is tested under controlled laboratory conditions, and the applied forces and deformations are known, four basic mechanical properties of bone can be derived from a plot of the stress – strain relationship. These properties are stress, modulus, energy absorptive capacity, and deformation. By convention, stress is plotted on the ordinate (y axis) and strain on the abscissa (x axis). Figure 17 shows a stress – strain plot of an idealized material. Considering that combinations of the three types of stress can produce different stress patterns as a result of different types of externally applied loads (tension, compression, bending, and torsion), the terms used to define the parameters of the y axis can include any of these loading conditions. Under these circumstances, the stress – strain
18
FIGURE 17 A standard stress/strain curve of bone loaded in bending. The linear portion of the curve represents the elastic region and the slope of this part of the curve is used to derive the stiffness of the bone. Loading in this region will result in nonpermanent deformation, and the energy returned to the bone when the load is removed is known as resilience. The nonlinear portion of the curve represents the plastic region in which the bone will be permanently deformed by the load. The junction of these two regions defines the yield point and the stress here is known as the elastic limit. The maximum stress at the point of failure is known as the ultimate strength of the bone. The maximum strain at this point represents the bone’s ductility. The area under the curve is known as the strain energy, and the total energy stored at the point of fracture defines the toughness of the material. Reprinted with permission from T. A. Einhorn, Calcif. Tissue Int. 51, 333 – 339 (1992).
curve is actually a load versus deformation relationship whereby the y axis could be labeled as torque, compressive load, tensile force, or shear. At low levels of stress there is a linear relationship between the applied load and the resultant deformation. This proportionality is known as the modulus of elasticity or Young’s modulus. It is a measure of the rigidity of the bone tissue and is equivalent to the slope of the linear part of the curve. It is calculated by dividing the stress by the strain at any point along this straight line. This linear part of the curve is also known as the elastic region. The physiological significance of this property relates to the fact that forces applied to bone at any point along this line will only deform the bone temporarily. After the load is removed, it will return to its original shape. At the point where the curve becomes nonlinear, the elastic region ends and the stress at this point is known as the elastic limit. Further loading beyond this point will result in a permanent deformation in the material and this property is known as plasticity. This part of the curve is known as the plastic region. Not all materials (and, for that matter, not all bones) have significant plastic properties. Instead, a material may exhibit elastic deformation but, upon reaching its yield point, will fail. This type of material is considered to be brittle.
LEE AND EINHORN
The maximum height of the curve defines the maximum stress in the bone tissue and is the point at which the material fails. The strain at the point of failure is known as the ductility. The area under the curve is a measure of the energy absorptive capacity or strain energy. This energy is dissipated when the bone fractures and is lost at the point of failure. Note that energy stored by the bone up to the point where it reaches its elastic limit is known as resilience. This energy is recovered if the bone returns to its original shape after the load is removed. When whole bones are subjected to experimental or physiological loading conditions, their mechanical behavior is dependent not only on the mass of the tissue and its material properties, but also on its geometry and architecture. When whole bones are loaded to failure, they, too, produce curves that are configured like the curve shown in Fig. 21. However, under these conditions, the slope of the linear portion defines the stiffness of bone. Strength is determined by extrapolating the height of the curve to the y axis. Depending on the loading conditions, strength can be expressed in terms of compressive, bending, tensile, or torsional strength. The area under the curve now defines the toughness of the bone. Of particular interest in the study of metabolic bone diseases are fractures of the vertebral bodies. Here, the vertebrae fracture as a result of predominantly axially directed compressive loads. The reduced load-bearing capacity of each vertebra is related to the material properties of the bone as well as the way in which the vertebral trabeculation is altered through the processes of postmenopausal and agerelated bone loss. Studies have shown that it is the removal of the horizontal trabeculae or lateral support cross-ties that alter the architectural arrangement and lead to reduced load-bearing capacity [22]. As a result, the vertical trabeculae begin to behave as columns and, as such, are subjected to critical buckling loads [23]. A 50% reduction in the cross-sectional area contributed by these horizontal trabeculae will be associated with a 75% reduction in the loadbearing capacity of the vertebral body [24]. Most fractures occurring in nature result from a combination of axial compression, bending, and torsion. In bending or torsion, the cross-sectional area of a structure is more important in resisting loads than its mass or density [25]. Ideally, in bending or torsion, bone should be distributed as far away from the neutral axis of the load as possible. The geometric parameter used to describe this phenomenon in bending is the “areal moment of inertia’’ [25]. Similarly, in torsion, deformation would be resisted more efficiently if bone were distributed further away from the neutral axis of torsion. This property is known as the “polar moment of inertia’’ [25]. During aging, the outer cortical diameter of bone increases and the cortical wall diameter becomes thinner [26,27]. This results from the combined effects of increased endosteal resorption and periosteal bone formation. Although the net effect may be cortical thinning, the
CHAPTER 1 The Bone Organ System
increased diameter of the bone distributes the material further from the neutral axis and improves its resistance to bending and torsional loads.
VI. SUMMARY Bone is a mechanically optimized organ system whose composition and organization reflect the functional demands made upon it. Far from being an inert substance, it is also a living tissue that serves several important functions in the organism. As a biological entity, bone tissue is a composite material composed of a proteinaceous extracellular matrix or ground substance that has been impregnated by an inorganic calcium phosphate mineral phase. In this sense, it can be likened to a material such as fiberglass with flexibilities, rigidities, and other mechanical properties related to the composite nature of its components. However, what distinguishes a material like bone from other composite tissues and materials is the fact that it is constantly being broken down and rebuilt in the process known as remodeling. The cellular link between bone resorbing cells, osteoclasts, and bone-forming cells, osteoblasts, may be regulated by the release of small molecules from the extracellular matrix during bone resorption. The complexities of the remodeling process require further investigation, but current knowledge suggests that it is composed of several phases, including quiescence, activation, resorption, reversal, and formation. Although the transduction of mechanical signals through bone and the stimulation of bone cells by hormonal agents have been studied extensively, more information is needed before a truly comprehensive understanding of the bone organ system can be developed. Ultimately, it would be important to know how tissue development, cellular function, mechanical input, and the material properties of the tissue are organized in time, space, and function in order to maintain bone homeostasis. The information and concepts described in this chapter form the basis for understanding bone as a tissue as well as an organ system. The ability to maintain homeostasis, prevent bone resorption, and, perhaps more importantly, enhance bone formation will be essential to the management of osteoporosis now and in the future.
References 1. I. A. Bab and T. A. Einhorn, Polypeptide factors regulating osteogenesis and bone marrow repair. J. Cell Biochem. 55, 358 – 365 (1994). 2. M. Horowitz and R. L. Jilka, Colony stimulating factors in bone remodeling, In “Cytokines and Bone Metabolism” (M. Gowen, ed.), pp. 185 – 227. CRC Press, Boca Raton, FL, 1992. 3. J. Wolff, In “Das Gesetz der Transformation der Knochen” (A. Hirchwald, ed.). Berlin, 1892. 4. T. A. Einhorn, Bone metabolism and metabolic bone disease. In “Orthopaedic Knowledge Update 4 Home Study Syllabus” (J. W. Frymoyer, ed.), pp. 69 – 88. Am. Acad. Orthop. Surg., Rosemont, 1994.
19 5. E. Ruoslahti, Integrins. J. Clin. Invest. 87, 1 – 5 (1991). 6. F. S. Kaplan, W. C. Hayes, T. M. Keaveny, A. L. Boskey, T. A. Einhorn, and J. P. Iannotti, Form and function of bone, In “Orthopaedic Basic Science” (S. R. Simon, ed.), pp. 127 – 184. Am. Acad. Orthop. Surg., Rosemont, 1994. 7. T. A. Einhorn and R. J. Majeska, Neutral proteases in regenerating bone. Clin. Orthop. 262, 286 – 297 (1991). 8. S. P. Bruder and A. I. Caplan, Terminal differences of osteogenic cells in the embryonic chick tibia is revealed by a monoclonal antibody against osteocytes. Bone 11, 189 – 198 (1990). 9. T. J. Martin, E. Romas, and M. T. Gillespie, Interleukins in the control of osteoclast differentation. Crit Rev. Eukaryot. Gene Expr. 8, 107 – 23 (1998) 10. Y. Y. Kong, H. Yoshida, I. Sarosi, H. L. Tan, E. Timms, C. Capparelli, S. Morrony, A. J. Oliveirados-Santos, G. Van, A. Itie, W. Khoo, A. Wakeham, C. R. Dunstan, D. L. Lacey, T. W. Mak, W. J. Boyle, and J. M. Penninger, OPGL is a key regulator of osteoclastogenesis, lymphocyte development and lymph-node organogenesis. Nature (England) 397, 315 – 323, (1999). 11. N. Nakagawa, M. Kinosaki, K. Yamaguchi, N. Shimma, H. Yasuda, K. Yano, T. Morinaga, and K. Higasio, RANK is the essential signaling receptor for osteoclast differentiation factor in osteoclastogenesis. Biochem Biophys Res. Commun. 253, 395 – 400 (1998). 12. Y. Hakeda, Y. Kobayashi, K. Yamaguchi, H. Yasuda, E. Tsuda, K. Higashio, T. Miyata, and M. Kumegawa, Osteoclastogenesis inhibitory factor (OCIF) directly inhibits bone-rresorbing activity of isolated mature osteoclasts. Biochem Biophys Res Commun. 251, 796 – 801 (1998). 13. A. M. Parfitt, Bone remodeling: Relationship to the amount and structure of bone, and the pathogenesis and prevention of fractures, In “Osteoporosis: Etiology, Diagnosis, and Management” (B. L. Riggs and L. J. Melton, eds.), pp. 45 – 93. Raven Press, New York, 1988. 14. J. Kragstrup and F. Melsen, Three-dimensional morphology of trabecular bone osteons reconstructed from serial sections. Metab. Bone Dis. Relat. Res. 5, 127 – 130 (1983). 15. A. M. Parfitt, The psychological and clinical significance of bone histomorphometric data. In “Bone Histomorphometry. Techniques and Interpretations” (R. Recker, ed.) pp. 143 – 223. CRC Press, Boca Raton, FL, 1983 16. A. M. Parfitt, The cellular basis of bone remodeling: The quantum concept reexamined in light of recent advances in cell biology of bone. Calcif. Tissue Int. 36, S37 – S45 (1984). 17. E. F. Eriksen, F. Melsen, and L. Mosekilde, Reconstruction of the resorptive site in iliac trabecular bone: A kinetic model for bone resorption in 20 normal individuals. Metab. Bone Dis. Rel. Res. 5, 235 – 242 (1984). 18. A. M. Parfitt, A. R. Villaneuva, M. M. Crouch, C. H. E. Mathews, and H. Duncan, Classification of osteoid seams by combined use of cell morphology and tetracycline labeling: Evidence for intermittency of mineralization. In “Bone Histomorphometry: Second International Workshop” (P. J. Meunier, ed.), pp. 299 – 310. Armour Montagu, Paris, 1977. 19. L. J. Melton, E. Y. S. Chao, and J. M. Lane, Biomechanical aspects of fractures. In “Osteoporosis: Etiology, Diagnosis and Management” (B. L. Riggs and L. J. Melton, eds.), pp. 111 – 131. Raven Press, New York, 1988. 20. A. H. Burstein, D. T. Reilly, and M. J. Martens, Aging of bone tissue: Mechanical properties. J. Bone Jt. Surg. 58A, 82 – 86 (1976). 21. T. D. Brown and A. B. Ferguson, Jr., The development of a computational stress analysis of the femoral head. J. Bone Jt. Surg. 60A, 619 – 629 (1978). 22. Li, Mosekilde, A. Viidik, and L. E. Mosekilde, Correlation between the compressive strength of iliac and vertebral trabecular bone in normal individuals. Bone 6, 291 – 925 (1985).
20 23. P. R. Townsend, Buckling studies of single human trabeculae. J. Biomech. 8, 199 – 201 (1975). 24. H. Yamada, “Strength of Biological Materials” (F. G. Evans, ed.). Williams and Williams, Baltimore, MD, 1970. 25. T. A. Einhorn, Bone strength: The bottom line. Calcif. Tissue Int. 51, 333 – 339 (1992). 26. R. W. Smith and R. Walker, Femoral expansion in aging women: Implications for osteoporosis and fractures. Henry Ford Hosp. Med. J. 28, 168 – 170 (1980).
LEE AND EINHORN 27. C. B. Ruff and W. C. Hayes, Superiosteal expansion and cortical remodeling of the human femur and tibia with aging. Science 217, 945 – 948 (1982). 28. D. B. Phemister, The pathology of ununited fractures of the neck of the femur with special reference to the head. J. Bone Jt. Surg. 21, 681 – 693 (1939). 29. A. M. Pankovich, Primary internal fixation of femoral neck fractures. Arch. Surg. 110, 20 – 26 (1975).
CHAPTER 2
Osteoblast Biology JANE B. LIAN AND GARY S. STEIN Department of Cell Biology, University of Massachusetts Medical School, Worcester, Massachusetts 01655
I. Overview II. Embryonic Origins and Signaling Cascades for Osteogenesis III. In Vivo Tissue Level Organization of Osteoblasts: Phenotypic Features and Function IV. Cellular Cross-Talk of Osteoblast Lineage Cells: Functions in Calcium Homeostasis, Bone Turnover, and Hematopoiesis
V. Osteoblasts in Vitro: Stages in Development of the Osteoblast Phenotype VI. Molecular Mechanisms Mediating Progression of Osteoblast Differentiation VII. Concluding Remarks References
I. OVERVIEW
of genes associated with the biosynthesis, organization, and mineralization of the bone extracellular matrix. This chapter provides a perspective of the signaling pathways and molecular mechanisms mediating osteoblast growth, differentiation, and activity that serve as a basis for understanding the factors regulating bone development and growth, the continued remodeling of bone, and the regeneration of injured tissue. Recent advances in the identification of obligatory factors for development of the skeleton that also contribute to osteoblast growth and differentiation in the adult skeleton will be presented. During the past decade, numerous model systems have been developed to study the cell biology of bone and gain insight into various cell types important for bone formation and function. The advantages and limitations of recently described in vivo and cellular models to address physiological and molecular mechanisms regulating osteoblast growth and differentiation will be discussed. Knowledge of unique properties and definition of the mechanisms that control progression through the osteoblast cell lineage will allow a rational intervention for abnormalities in skeletal development, fracture repair, pathologies of metabolic bone diseases, and implant stability. These are basic biological questions and
Bone formation takes place not only during embryonic development and growth but throughout life to support normal bone remodeling and fracture repair. The requirement for continuous renewal of bone, through the remodeling process involving resorption and formation, necessitates recruitment, proliferation, and differentiation of osteoblast-lineage cells. Subpopulations of osteoblasts are recognizable in vivo morphologically in relation to tissue organization and exhibit subtle phenotypic differences with respect to the expression of genes and responses to physiologic mediators of bone formation. This chapter presents current understanding of the phenotypic definition of the spectrum of bone-forming cells with respect to their functional properties and responses. It is now apparent that growth factor, cytokine, and hormone responsive regulatory signals mediate competency for the expression of genes associated with metabolic responses as a function of the stages of osteoblast growth and differentiation. Osteoblast differentiation is a multistep series of events modulated by an integrated cascade of gene expression that initially supports proliferation and the sequential expression
OSTEOPOROSIS, SECOND EDITION VOLUME 1
21
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
22
LIAN AND STEIN
concerns of today’s clinician for the treatment of bonerelated disorders.
II. EMBRYONIC ORIGINS AND SIGNALING CASCADES FOR OSTEOGENESIS The complexities of bone formation are immediately apparent in the embryo where different regions of the skeleton arise from specific primordial structures and skeletogenesis involves two different processes. Intramembranous bone formation, as occurs in the development of the flat bones of the skull, results from the differentiation of mesenchymal cell condensations directly to osteoblasts. The endochondral sequence of bone formation, as occurs for all long bones, involves the differentiation of mesenchymal progenitors first to form a cartilage template of the bone, which is then replaced by bone. Progenitors of the boneforming cells for all osseous tissues derive from the mesodermal germ cell layer. The dorsal paraxial mesoderm gives rise to somites and the sclerotome, which is a source of cells for most of the axial skeleton (vertebrae, skull, ribs, sternum), the lateral plate mesoderm gives rise to the appendicular skeleton (limbs), and the cephalic mesoderm gives rise to the neural crest, which provides progenitor cells for facial skeletal structures. Thus, different regions of
FIGURE 1
the skeleton have distinct embryonic lineages reflecting origins from these specific primordial structures. Considering these different embryonic developmental programs of the mesoderm to form intramembranous bone and subtypes of endochondral bone (e.g., limbs and vertebrae), an early osteoprogenitor may divert from a stem cell at these specific skeletal sites. It has been shown that axial and appendicular-derived osteoblasts exhibit different responses to hormones [e.g., 1,2]. It remains to be determined whether this selective activity reflects the tissue environment or inherent properties of the cells selected at an early stage during osteoblast differentiation. Our understanding of skeletal patterning and limb development has been expanded significantly by characterization of the signaling factors and transcription factors that serve as morphogenic determinants of bone formation [3 – 8]. Well-documented interactions between epithelium and mesenchyme are first necessary for tissue differentiation [9] and are reviewed elsewhere [10]. Signals essential for limb patterning arise from the zone of polarizing activity, which resides in the posterior limb mesoderm, underlying the epithelial apical ectodermal ridge (AER) of the developing limb bud. Gradients of morphogenic factors and integration of the various regulatory cascades reflect the complexity of skeletal development (Fig. 1). Growth factors and their receptors in the transforming growth factor- (TGF-) superfamily, epidermal growth factor (EGF), and the fibroblast
Signaling molecules regulating position and pattern formation and osteoblast differentiation during development of the skeleton. Encircled regulators, the FGF family, BMPs, Hedgehogs, and HOX genes, can be considered major signaling “centers” in that they induce (arrows) the expression of several different classes of genes essential for skeletal development. Inhibitory controls are indicated ( ). Interrelationships are evident by feedback loops (e.g., FGFs and Hedgehogs), coordinate positive and negative regulation of common genes, and potential interaction of factors from one center with factors regulated by a different center. Relevant references include 21,22,27,28,47,67 – 70,75 – 77,698 – 705, as well as those described in the text.
23
CHAPTER 2 Osteoblast Biology
growth factor (FGF) family are implicated in mesenchymal condensations [11,12]. Basic fibroblast growth factor (FGF-2) and retinoic acid have been characterized as soluble mediators of communication between the AER and mesodermal progress zone [10,13 – 17]. FGF-2 activates several signaling pathways [18], including, for example, Wnt genes [19], Notch ligand expression [20,21], Hedgehog factors, and transcriptional regulators, such as helix – loop – helix proteins [22] (Fig. 1). Indian hedgehog (Ihh) is a key factor in the normal development of endochondral bone formation [23]. It is expressed abundantly in the developing growth plate in the mature and hypertrophic chondrocyte zones [24] is a key regulator of chondrocyte maturation. Ihh signaling is mediated through its receptor patch, and this signaling can be antagonized by a cell surface protein [25]. Several studies indicate that the rate of cartilage differentiation by Ihh is mediated by the parathyroid hormone-related protein (PTHrP) and its receptor [24,26 – 28]. PTHrP supports chondrocyte proliferation. Thus together Ihh and PTHrP regulate the proportions of proliferating and hypertrophic chondrocytes. Indeed, the PTHrP receptor will rescue the Ihh null mouse [27]. Knowledge of how mesenchymal condensations are initiated and grow and how their sizes and boundaries are regulated is being accrued through genetic studies in mice and the characterization of molecular defects in skeletal development [7] (Table 1). These studies have identified several specific players. Extracellular matrix molecules, cell surface receptors, and cell adhesion molecules, such as fibronectin, tenascin, syndecan, and N-CAM, initiate condensation and
TABLE 1
set boundaries for the forming mesenchyme, whereas Hox genes modulate the proliferation of cells within condensations [7]. Significant contributions to these earlier stages of skeletal development are provided by the TGF- superfamily of regulatory factors [reviewed in 29 – 32]. Bone morphogenetic proteins (BMPs) are actively involved in determining parameters of size and shape during mesenchymal cell condensation [33,34]. Selective expression of BMPs regulate mesenchymal condensations [35] and contribute to restriction of options for lineages. BMP-2, BMP-4, and BMP-7 (also designated OP-1) are potent inducers of osteogenesis in vivo and cell differentiation in vitro [36 – 41]. Specificity of the activities of TGF-s with target cells is regulated by their activation of distinct factors. TGF- and BMP bind to distinct receptors; each has associated kinase activity that phosphorylates Smad proteins. Smads 2 and 3 mediate TGF- responses, whereas Smads 1, 5, and 8 are activated by BMP receptors and transduce BMP signals. Interactions between receptor-activated Smads and Smad 4, a DNA-binding Smad, result in translocation of the complex to the nucleus for the transcription of target genes [29,42 – 44]. BMPs regulate several classes of genes that are key factors in normal bone development influencing spatial and temporal events (Fig. 1). These include fibroblast growth factors [45], homeobox containing genes [46 – 49], which contribute to position and pattern formation, cell adhesion proteins [50,51], and transcription factors such as helix – loop – helix [22,52], winged helix [53 – 55], SOX9, which is essential for chondrocyte differentiation [56,57], and RUNX2/core binding factor
Representative Mouse Models Involving Factors Related to Skeletal Development and Bone Formation
Homeobox proteins BAPX1 [648,649] (null)
Expressed in prechondrogenic cells, perinatal lethal skeletal dysplasia; malformations and absence of specific bones
Goosecoid [650] (null)
Craniofacial due to condensation defects
DLX-5 [651,652] (null)
Craniofacial and bone defects, die at birth
Msx-2 [653,654] (null)
Skull ossification defects
Msx-2 [655] (gain of function)
Craniosynostasis
TGF- superfamily TGF- receptor type II (transgenic dominant negative) a. Expressed in osteoblasts [656]
a. Decreased remodeling; increased trabecular bone
b. Expressed in skeleton [657]
b. Osteoarthritis
TGF-2 [658] (transgenic)
Development normal; osteoporosis postnatal
BMP-2 [659] (null)
Failure of mesoderm induction
BMP-4 [660,661] (null)
Die early in gastrulation
BMP-5 [34] (mutant)
Short-ear mouse gene (continues)
24
LIAN AND STEIN
TABLE 1 TGF- superfamily GDF-5 [662] (mutant)
(continued)
Brachypodism; bone and joint development problems
BMP-7 [663,664] (null)
Die after birth; polydactly; rib abnormalities
BMP receptor type I [665] (null)
Die 9.5 days post coital
BMP receptor type IB [666] (null)
Appendicular skeletal defects related to reduced chondrocyte proliferation and differentiation
Bone matrix proteins Osteopontin [667] (null)
Development is normal; increased osteoclastogenesis
Osteocalcin [277,278] (null)
Thickened bone increased mineral, but decreased crystal size
Osteonectin [668,669] (null)
Osteopenia; mesenchymal cell proliferation increased
Thrombospondin [670,671] (null)
Increased vascular density; increased marrow osteoprogenitor; increased cortical density
Biglycan [280] (null)
Thin bones, expression in preosteogenic cells
Decorin [672] (null)
Skin fragility
Alkaline phosphatase [673 – 675] (null)
Arrest of chondrocyte differentiation; defective matrix mineralization
Gelatinase [676] (null)
-3 Integrin [677] (null) BMP-5 [34] (mutant)
Transient aberrant growth plate vascularization and ossification Increased bone mass, but osteosclerotic; defective osteoclasts Short-ear mouse
Growth factors and receptors FGF-2 [261] (null)
Trabecular bone mass decreased; decreased mineralization of bone marrow stromal ex vivo cultures
FGF-2 [260] (transgene)
Increased apoptosis in calvaria
FGF-3 receptors [678,679] (null)
Skeletal overgrowth
(mutations) [680]
Achondroplasias dwarfism
FGF-1 receptor [681,682]
Embryonic growth and mesodermal patterning; axial organization and limb development
PTH/PTHrP receptor [80] (transgene)
Delays endochondral bone formation (EBF)
PTH/PTHrP receptor [28] (null)
Accelerates EBF; increase in osteoblast number and matrix; delay in vascular invasion; decrease in trabecular bone
PTHrP [74] (null)
Die at birth; premature chondrocyte differentiation and accelerated EBF
PTHrP x PTH/PTHrPR [78] (double knockout)
Accelerates differentiation of growth plate chondrocytes and EBFs; no delay in vascular invasion
PTHrP [683] (transgene)
In cartilage, delays chondrocyte maturation; mice are born with cartilaginous skeletons
IGF-1 [684] (transgene)
In bone, increases trabecular bone volume without proliferation
Other regulatory factors Indian Hedgehog [23] (null)
Severe dwarfism in the null mouse
Sonic Hedgehog [685]
A segment polarity gene; absence of distal limbs, most of the ribs and spinal column
Noggin [63] (null)
A BMP antagonist; cartilage hyperplasia
C-Abl [686] (null)
A nonreceptor tyrosine kinase; osteoporotic and osteoblast maturation defect
p27kip1 [253,687,688] (null)
A cyclin-dependent kinase inhibitor; larger size animals and bones; increased osteoprogenitors in bone marrow ex vivo cultures
TWRY [689] (mutant)
Nucleotide pyrophosphatase (NPPS) gene mutation; abnormal calcification of cartilage, spine, tendons, limbs
Transcriptional regulators RAR [690] (transgene)
Interferes with chondrogenesis, appendicular skeletal defect
SOX-9 (null)
An HMG domain TF expressed in cartilage lethal malformation syndrome XY reversal
c-Fos [691] (transgene)
Tumors in cartilage and bone
c-Fos [692] (null)
Osteopetrosis
Fra-1 [693] (transgene)
Progressive increase in bone mass; osteosclerotic
Fra-1 [693,694] (null)
Embryonic lethal from a placental defect
MFH-1 [695]
A forkhead or winged helix transcription factor; embryonic lethal and perinatal with skeletal defects
Plzf [696]
Promyelocytic leukemia zinc finger proteins; a growth inhibitory and proapoptotic factor essential for skeletal patterning
ATF-2 [697]
Chondrodysplasia
CBFA1 [58,59] (null)
A runt domain protein, absence of mineralized tissue essential for osteoblast differentiation
25
CHAPTER 2 Osteoblast Biology
A1 (CBFA1), which is essential for skeletal formation [58,59]. When growth of condensation ceases, the next stage of skeletal development, cellular differentiation, initiates. Here, Chordin [60], Noggin, and potentially Gremlin [61] play a critical role in skeletal development by inhibiting BMP2 signaling [62 – 65]. Feedback loops and complementary regulation of many of these factors (Fig. 1) ensure the progression of cellular differentiation, cartilage development, endochondral bone formation, and intramembranous osteogenesis [66 – 70]. Systemic factors are also critical for the early stages of limb formation and influence the renewal of osteogenesis throughout adult life. While parathyroid hormone (PTH) stimulates the growth of osteoprogenitor populations [71], PTHrP functions as a cellular cytokine regulating cell growth for differentiation in development [72,73] and is a key regulator of chondrocyte maturation [74]. PTHrP influences Hox gene expression and both peptide and its receptor are upregulated by TGF- [75 – 77]. Ablation of the PTH/PTHrP receptor in mice [28,74,78] and mutations in the receptor in human [79] reveal its central role in the regulation of endochondral bone formation (EBF). The disorganized growth plate of the receptor knockout mouse can be rescued by expression of the receptor; however, endochondral bone formation becomes delayed [80,81] (Table 1). Commitment of stem cells to specific mesenchymal lineages occurs early in development of the limb. Transcription
FIGURE 2
factors which function as “master switches” mediate cell differentiation by induction of a set of phenotypic genes that characterize the muscle, adipocyte, chondrocyte, or osteoblast cells (Fig. 2). Such regulatory proteins have also been implicated in key roles during the progressive differentiation of osteoblasts through specific stages of maturation (Fig. 2, lower panel). Among the most notable are cfos, a protooncogene, the helix – loop – helix proteins (Twist, Id, Scleraxis), leucine zipper proteins (hXBP-1), zinc finger proteins (zif268), homeodomain proteins (Msx-2, Dlx-5), steroid receptors, and runt domain transcription factors (RUNX2/Cbfa1/AML/PEBP2), which are obligatory for osteoblast differentiation. Modifications in the representation of classes of transcription factors at specific sites during embryogenesis and at different stages of osteoblast differentiation (Fig. 2, bottom) reflect linkage to the transcriptional control of osteoblast phenotype development. Helix – loop – helix factors, negative regulators of osteogenesis, illustrate this point. Id (inhibitor of differentiation), twist, and scleraxis are expressed in mesoderm of developing embryo [82,83]. Scleraxis is expressed in cells that form the skeleton and is not detected at the onset of ossification [83]. Id and Twist expression must be downregulated for osteoblast differentiation to proceed [84,85]; overexpression of these factors inhibits osteogenesis in vitro [86]. Msx-2 and Dlx-5 are members of the homeodomain gene family of transcription factors [87]. Both factors are expressed in mesenchymal cells at sites that will undergo skeletogenesis
Transcriptional control of phenotype lineages from mesenchymal stem cells. (Top) Activation of gene transcription for commitment to the muscle cell phenotype by MyoD to adipocytes by PPAR2 and C/EBP, chondrocytes by SOX9, and osteoblasts by BMP-2 and CBFA1 is illustrated. Several in vitro studies demonstrate that phenotype commitment can be blocked or switched by forced expression of the regulatory factor characteristic of a different lineage (designated ) [175,706 – 710]. (Bottom) Developmental expression of several transcription factors influencing osteoblast differentiation. Thick lines represent highest cellular levels.
26 [88 – 90]. Their importance for normal bone formation is realized by skeletal abnormalities that result from mutations or misexpression of these factors [91; and see Table 1]. Msx-2 must be down regulated for progression of osteoblast differentiation [92,57,8]. The runt homology domain-related core-binding factor RUNX2 has been shown to play an essential role in bone formation in the embryo, demonstrated by the inhibition of mineralized tissue formation in the CBFA1/ RUNX2 null mutation mouse model [58,59]. The family of RUNX/CBFA transcription factors comprises three related genes that each support tissue specification and organogenesis [93– 95]. RUNX1 (CBFA2/AML-1B/PEBP2B) and its partner proteins CBF are critical for hematopoietic cell differentiation [96 – 98]; RUNX3 (CBFA3/AML-2/PEBP2C) is required for gut development (Dr. Y. Ito, Kyoto University, Japan, personal communication); RUNX2 (CBFA1/ AML-3/ PEBP2A) is essential for the differentiation of osteoblastic cells for formation of the mineralized skeleton. In human, these factors were first designated as AML because they were identified as the protein encoded by a gene locus rearranged in acute myelogenous leukemia (AML). Other names include core-binding factor (CBFA) and polyoma enhancer-binding protein (PEBP2). The Human Genome Nomenclature Committee has now referred to this family as RUNX. CBFA1 was initially identified in bone as an osteoblastspecific DNA-binding protein that activated transcription of the tissue-specific osteocalcin gene [99 – 103] (also see Chapter 6). The obligatory role of CBFA1 for the formation of mature bone in the developing skeleton has been shown by the absence of a calcified skeleton and bone formation in CBFA1 null mutant mice [58,59]. The heterozygous mice exhibited phenotypic features akin to mouse models [104] and human cleidocranial dysplasia (CCD) abnormalities. Various mutations in CBFA1 were then identified in patients with CCD [105 – 108]. CBFA1 is expressed not only in abundance in osteoblasts and hypertrophic chondrocytes, but in cartilage [109 – 111], thymus [112], and testis [113], as well as early stage osseous and chondroprogenitor cells Marrow mesenchymal stromal cells lacking overt expression of chondrocyte or osteoblast markers and several nonosseous mesenchymal cell lines also have significant CBFA1 expression [114 – 116]. The transient expression of CBFA1 in early embryogenesis, followed by an upregulation in late stages of bone development [58,102,103], suggests that CBFA1 may be important in both early specification of the mesenchymal stromal phenotype and in supporting the final stages of osteoblast differentiation [103,117]. Evidence is also provided by the observations that RUNX antisense blocks in vitro differentiation of osteoblasts to the final mineralization stage [103]. Furthermore, CBFA1 can induce the expression of bone-related genes in nonosseous
LIAN AND STEIN
cells [102,103,118]. Finally, a dominant-negative CBFA1 mutant protein, expressed only in mature osteoblasts by the osteocalcin promoter, resulted in osteopenic bone due to decreased osteoblast activity [117]. The activities of growth regulators of skeletal development, FGFs, TGF-1, or BMP and CBFA1, appear to be closely linked. Activating mutations in FGF receptors cause premature fusion of the calvarial sutures in human disorders [119] and in a recently described mouse model [120]. FGF expression in this disorder is linked to expression of CBFA1. FGFR1 and ligands FGF2 and FGF8 induce CBFA1 in vivo and in C3H10T1/2 cells. Several reports have shown that CBFA1 is a downstream target of BMP-2 in osteoblastic and nonosseous cell lines [114,118,121,122] and BMP-7 [102]. Evidence has been presented demonstrating a protein – protein interaction of the CBFA2 factor with Smad1 or Smad 5 for functional cooperativity of TGF-mediated gene transcription [123]. Furthermore, Smad2 and CBFA1 cooperativity in osteoblasts has been shown [124]. These studies reflect cross-talk between two classes of signaling factors essential for osteogenesis, the Smad proteins, which mediate TGF-1 and BMP activities, and CBFA1 factors. Other studies suggest that CBFA1 induction of osteogenesis may require a BMP responsive factor. For example, while both TGF and BMP-2 induce CBFA1 expression [121] in the premyoblastic C2C12 cell line, only BMP-2 leads to expression of the osteoblast phenotype. Thus, BMP-2, but not TGF-1 signaling, leads to a factor that, together with CBFA1, may be required to mediate osteogenesis. This concept is further supported by the ability of BMP-2 to activate osteocalcin in calvarial cells from CBFA1 null mice [58]. In addition, subclones of the osteoblastic MC3T3-E1 cell line, which expressed CBFA1, were found incompetent for the synthesis of a mineralized bone ECM [125]. Thus, while CBFA1 gene ablation studies in mice revealed that CBFA1 is necessary for bone formation and can activate some osteogenic genes in nonosseous cells, CBFA1 may not be sufficient for de novo skeletal formation. We still do not understand if CBFA1 can provide a cell with the same cascade of signals induced by the bone morphogenetic proteins necessary for bone formation. In summary, our present knowledge of signaling factors and transcriptional regulators of bone development is growing. The significance of many of these growth factors and morphogens in regulating the progression of the pluripotent stem cell and multipotential mesenchymal cell to the committed osteoprogenitor and finally to recognizable osteoblasts is appreciated from mouse models in which these genes or receptors for these proteins have been expressed in transgenic mouse models or ablated (null mutations) with consequences on formation of the skeleton (summarized in Table 1). However, identification of the sequelae of events and integration of their activities can only be postulated at the present time. It is apparent that the convergence of multiple pathways and the
27
CHAPTER 2 Osteoblast Biology
coordination of activities are operative through complex feedback loops as illustrated above (Fig. 1), not only in development, but in renewal of osteogenesis in the adult [126,127].
III. IN VIVO TISSUE LEVEL ORGANIZATION OF OSTEOBLASTS: PHENOTYPIC FEATURES AND FUNCTION Based on morphological and histological studies, osteoblastic cells are categorized in a presumed linear sequence progressing from osteoprogenitor cells to preosteoblasts, which mature to osteoblasts and then to lining cells or osteocytes [128 – 134] (Fig. 3). There is a gradient of differentiation that can be observed morphologically either in the periosteum or in the marrow as the osteoprogenitor cell reaches the bone surface and the osteoblast phenotype becomes fully expressed. Osteoblasts that are derived from proliferating osteoprogenitors can be observed in clusters at the bone surface (Fig. 4, see also color plate). These cells synthesize the bone extracellular matrix, designated osteoid (Fig. 5, see also color plate). In metabolic bone disorders leading to decreased calcium or phosphate deposition in bone, as in vitamin D deficiency, wide osteoid seams are evident (Fig. 5B). Mineralization leads to the
FIGURE 3
final stage of osteoblast differentiation. When the boneforming osteoblast becomes encased in its own mineralized matrix, they are osteocytes. On a quiescent bone surface, the osteoblast flattens to a lining cell, forming an endosteum. Four forms of the osteoblast cell lineage are thus recognized in vivo. They are the committed progenitors (preosteoblasts), mature osteoblasts, osteocytes, and the bone-lining cell (Figs. 4 and 5). Particularly important in defining phenotypic differences is an understanding of gene expression and protein localization at the single cell level in relation to the development of bone tissue organization. During the past, high-resolution in situ hybridization methods have been developed for bone sections that effectively support the assessment of mRNA levels in specific cells and with respect to intracellular localization [135 – 139]. The application of immunological detection methods [140 – 142] to sections of intact bone has supported the establishment of linkage between patterns of gene expression observed in culture with those that occur developmentally in tissue. Distinct gradients of particular markers are evident, a phenomenon that may be related to a coupling of cell polarity with physiological function [143]. Polymerase chain reaction (PCR) methods have been applied successfully to determinations of cellular RNAs in single cells [144], and in situ PCR protocols offer the potential for extending this approach [145,146]. Variable levels of expressed genes are observed in neighboring cells.
Growth and differentiation of osteoblast lineage cells. Progression through the osteoblast lineage from a pluripotent mesenchymal stem cell to a mature osteocyte is regulated by numerous physiologic mediators, including, but not limited to, transforming growth factor- (TGF-), superfamily members, steroid hormones, growth regulators, and transcription factors. Each of these subpopulations of osteoblasts have many common features, but exhibit distinct different properties/functions and express osteoblast phenotypic genes to varying extents (see text for details).
28
FIGURE 4 Organization of osteoblasts in bone tissue. (A) Intramembranous bone formation (40). (B) Bone-forming surface (200), Goldner and von Kossa stain shows (A) osteoblast clusters and (B) surface osteoblasts, preosteocytes below the osteoid (blue), and mineralized tissue (black) with an osteocyte in its lacunae. Osteoprogenitor cells are evident at higher magnification behind the large surface cuboidal osteoblasts. Osteocyte cellular processes that extend through canaliculi cannot be seen with light level microscopy. (See also color plate.)
Viable approaches are thereby provided for investigating protein functions within multiple contexts that range from transactivation in the nucleus to support of extracellular matrix mineralization. The following summarizes the morphologic and phenotypic functional features of each osteoblast population. For a more detailed description of bone cell ultrastructure, the reader is referred to Refs. 130,132, 134, and 147.
A. Osteoprogenitors and Preosteoblasts: More Cells Build Bigger Bones Progression of the most primitive pluripotent cell to the undifferentiated multipotential mesenchymal cell and presumed osteoprogenitor is not understood. The expression and subsequent differentiation of the early osteoprogenitor
LIAN AND STEIN
FIGURE 5 Mineralization defects due to vitamin D deficiency and low serum calcium results in osteomalacia. Cortical bone sections from a 6week control rat (A) and a rat maintained on a vitamin D-deficient diet from weanling to 6 weeks. (B) Osteoblast differentiation from the periosteal and endosteal surfaces is seen, as well as a wide osteoid due to a much greater decrease in the rate of mineral deposition. Toluidine blue/von Kossa stain (200). (See also color plate.)
cells should be considered within the context of embryologic development, bone formation during growth, and bone tissue remodeling in the adult skeleton. Under each of these circumstances, progenitor cells must be responsive to a broad spectrum of regulatory signals that mediate their proliferation, commitment, and progression of phenotype development, as well as sustaining their structural and functional properties. In fully developed bone, there is a requirement for utilization of the same factors that can mediate the growth and differentiation of osteoprogenitor cells during skeletal development, as well as for osteoblast differentiation during bone remodeling and fracture healing in the adult [127,148]. From a developmental perspective, mesenchymalderived osteoprogenitor cells arise/reside in the periosteal tissue or the bone marrow stroma. The marrow and its stromal “bedding” give rise to multipotential cells of both
CHAPTER 2 Osteoblast Biology
hematopoietic lineage (origin of osteoclasts) and nonhematopoietic lineage cells from which many tissuespecific cells derive, such as chondrocytes, myoblasts, and adipocytes. When suspensions of marrow cells are plated in vitro, clonal colonies of adherent fibroblasts are formed; each derived from the single cell that has been designated as the colony-forming fibroblastic unit or CFU/F [149,150]. Formation of CFUs requires the presence of hematopoietic cells [151]. A proportion of these cells have high proliferative and differentiation capacity and exhibit characteristics of stem cells when transplanted in the closed environment of a diffusion chamber [152] or transplanted into the circulation [153]. Studies have demonstrated that marrow stromal cells can be transfected with reporter genes and then transplanted either into specific tissue sites (Subcutaneous, bone tumors, fracture sites) or systemically (e.g., by tail vein infusion) and they maintain competency for differentiation in vivo [154]. For clinical applications of genetically engineered cells for the local reconstruction of bone tissue [155 – 159] or to treat systemic bone diseases [153,160 – 163], in vitro expansion and modification of the cells forming adherent marrow colonies are important considerations. However, mechanisms related to the retention of in vivo properties, homing, engraftment, and differentiation after transplantation must be addressed [159,164 – 166]. Many groups have shown the adherent marrow population differentiates in vivo and in vitro to several mesenchymal lineage cells: adipocytes, chondrocytes, osteoblasts, and myoblasts [reviewed in 167,116,168 – 174]. The plasticity of these lineages is indicated by several lines of evidence. Forced expression of the transcription factors that function as “master switches” (Fig. 2) in phenotype commitment can transdifferentiate a cell to a different phenotype. The reciprocal relationship between adipocyte and osteoblast differentiation is suggested by such studies [115,175]. Forced expression of PPAR2 in marrow stromal cell lines results in the inhibition of terminal osteoblast differentiation with concomitant downregulation of CBFA1 [115,175]. The bipotential property of the late stage osteoprogenitor or preadipocyte [176 – 183] is markedly sensitive to biological regulatory signals influencing “master switch” transcription factor expression. Regulatory signals influencing osteogenesis in preference to adipogenesis can include BMP-2 and the BMP receptor [115,179,182, 184 – 187] and TGF-1 [188]. Retinoid signaling pathways [189] and leptin signaling (through the central nervous system) [181,190] favor adipogenesis. In contrast, 1,25(OH)2D3 inhibits adipogenesis [191]. One molecular mechanism for commitment to the osteogenic or adipogenic lineage which has been identified involves the role of mitogen-activated protein kinase family members (ERK, JNK, and p38). ERK functions as positive regulators of osteogenic differentiation. Inhibiting ERK activation blocked
29 the bone phenotype and resulted in the adipogenic differentiation of human mesenchymal stem cells [192]. Implications for the plasticity of osteoprogenitor and adipocyte cells in relation to osteoporosis and the aging skeleton have been reviewed in detail [180,193]. Presently, a key obstacle in understanding the origin of osteoblast lineage cells is the inability to identify an osteoprogenitor cell prior to the expression of bone phenotypic properties. Using characterization of hematopoietic stem cells as a paradigm, several groups have developed antibodies to cell surface proteins using presumptive marrow stromal cell populations. These reagents have the potential for both recognition and purification of skeletal stem cells [172,194 – 202]. STRO-1 positive cells [203] are well documented with respect to their pluripotential and osteoprogenitor properties [196,204 – 206]. Antigens for other antibodies have been characterized. Interestingly, an antigen to a cell surface marker antibody (SB-10) produced in response to mesenchymal stem cells is the activated leukocyte cell adhesion molecule ALCAM [195]. Expression of ALCAM becomes downregulated in concert with changes in morphology and detection of alkaline phosphatase activity of the periosteal osteoprogenitors as they migrate and develop into osteoblasts. There still remains considerable debate as to whether pluripotential osteoprogenitors (marked, for example, as STRO-1 or ALCAM positive cells) have the capacity to function as a “stem cell.” Proliferation and differentiation of the osteoprogenitor pool is influenced by several regulatory factors. Plateletderived growth factor and epidermal growth factor [207] have been identified as important in stimulating expansion of the CFU/F [151]. The leukemia inhibitory factor (LIF) maintains stem cell populations and osteoprogenitors and inhibits their differentiation in vitro [208,209], but has also been reported to have osteogenic activity in vivo [210]. The fibroblast growth factor-2 [211] and TGF-1 are potent mitogens for periosteal osteoprogenitors and marrow stromal cells [212 – 214]. The osteoinductive effects of bone morphogenetic proteins are complex, modulating growth, osteoinduction, and even apoptosis, depending on the specific BMP, concentration dependency, and the progenitor cell phenotype [36,185,215,216]. BMP-2 rapidly induces osteoblast differentiation in marrow stromal cells [37,183, 217,218], but the effect is slower in a number of pluripotent cell lines [219,220] and in the mouse myogenic C2C12 cell line [221]. These growth factors are expressed and produced by osteoblast lineage cells and are stored in the bone extracellular matrix. A local mechanism for stimulating the proliferation of progenitors in the bone microenvironment is thereby provided [222,223]. The osteoprogenitor appears to have limited selfrenewal capacity compared to the stem cell [224]. In contrast, a key feature of the osteoprogenitor/preosteoblast population is its capacity to divide and increase the size of
30
LIAN AND STEIN
bone. Labeling studies ([3H]thymidine and autoradiography) indicate that the proliferating cells are principally confined to progenitor cells and preosteoblasts with very few osteoblasts labeled [225 – 227]. The determined osteoprogenitor is recognizable in bone as a preosteoblast. Preosteoblasts are usually observed as one or two layers of cells behind the osteoblast near bone-forming surfaces [130 – 134,147,228]; i.e., they are usually present where active mature osteoblasts are laying down a bone matrix. These cells appear elongated, fibroblastic, or spindle shaped with an oval or elongated nucleus and with notable glycogen content. However, some of the cells in the layer found directly behind the active osteoblasts may appear morphologically closer to the cuboidal osteoblast and can clearly be defined as preosteoblasts (Fig. 4). Preosteoblasts may express a few phenotypic markers of the osteoblast, e.g., alkaline phosphatase activity, but less than mature osteoblasts [147,229]. The preosteoblast, however, has not yet acquired many of the differentiated characteristics of mature osteoblasts; for example, there is no evidence of a developed rough endoplasmic reticulum [130].
B. Cell Cycle Regulatory Parameters: Control Mechanisms Supporting Osteoblast Growth With recognition of decreased osteoblast surfaces in osteoporotic bone [230] and reports of decreased marrow
osteoprogenitors with age [231 – 235], defining mechanisms contributing to the regulation of proliferative activity in osteoblast lineage cells is increasing in importance. To understand regulatory parameters of proliferation, one must consider mechanisms that support the requisite responsiveness to growth factors through signaling pathways and the consequent induction of proliferation. To explain the induction, synthesis, activation, and suppression of the complex and interrelated regulatory factors associated with the growth control of osteoprogenitor cell proliferation in vivo, an understanding of mechanisms that control cell proliferation is required. Proliferation is controlled through the cell cycle by the activity of regulatory proteins which support progression of cells that have responded to a mitogenic stimulus through DNA replication and cell division. The cell cycle is a stringent growth-regulated series of sequential biochemical and molecular events that support genome replication and mitotic division [reviewed in 236]. The stages of the cell cycle and checkpoints that monitor competency of cells to progress through DNA replication and mitotic division are illustrated in Fig. 6. Suppression of certain cell cycle regulated genes is requisite for the cessation of proliferation and upregulation of phenotypic genes. When quiescent cells (G0) are stimulated to proliferate and divide, they enter G1, the first phase of the cell cycle where the enzymes required for DNA replication are synthesized. Before a cell can progress through G1 and begin DNA synthesis (S phase), it must pass through a checkpoint
FIGURE 6 Control of cell cycle progression in bone cells. The cell cycle is regulated by several critical cell cycle checkpoints (indicated by check marks), at which competency for cell cycle progression is monitored. Entry into an exit from the cell cycle is controlled by growth-regulatory factors (e.g., cytokines, growth factors, cell adhesion, and/or cell – cell contact) that determine the self-renewal of stem cells and expansion of precommitted progenitor cells. The biochemical parameters associated with each cell cycle checkpoint are indicated. Options for defaulting to apoptosis during G1 and G2 are evaluated by surveillance mechanisms that assess the fidelity of structural and regulatory parameters of cell cycle control. Apoptosis also occurs in mature differentiated bone cells.
CHAPTER 2 Osteoblast Biology
in late G1, which is known as the restriction point [237]. At this cell cycle restriction point, both positive and negative external growth signals are integrated. If conditions are appropriate, the cell proceeds through the remainder of G1 and enters the S phase. Once the cell passes the restriction point, it is refractory to withdrawal of mitogens or to growth inhibitory signals and is committed to progressing through the remainder of the cell cycle unless it is subjected to DNA damage or metabolic disturbance [237]. In mammalian cells, progression through the cell cycle is regulated by a cascade of cyclin-containing growth regulatory factors that transduce growth factor-mediated signals into discrete phosphorylation events, controlling the expression of genes responsible for both initiation of proliferation and competency for cell cycle progression. Cyclin activity is modulated by the formation of complexes with a family of threonine/serine kinases designated cyclin-dependent kinases (cdk) [238,239]. Cdks are regulated by both positive and negative phosphorylation, as well as by their reversible association with specific cyclins during defined phases of the cell cycle [240]. In general, the levels of cdk proteins remain relatively constant during the cell cycle, whereas the expression of specific cyclins is confined to distinct phases of the cell cycle where they are degraded quickly after having completed their function. An emerging concept is that the cyclins and cdks are responsive to regulation by the phosphorylation-dependent signaling pathways associated with activities of the early response genes, which are upregulated following the mitogen stimulation of proliferation [reviewed in 240 – 243]. Cyclindependent phosphorylation activity is functionally linked to activation and suppression of both p53 and RB-related tumor suppressor genes [244,245]. p53 accumulates in response to stress, inducing arrest at G1 or G2. The retinoblastoma protein (Rb), a tumor suppressor, is a member of a family of related proteins that include p105, p107, and p130. Rb has been shown to have a critical role in the regulation of cell proliferation, particularly in progression through G1 [reviewed in 244]. Rb functions as a signal transducer, receiving both growth-promoting and -inhibitory signals and linking them to the transcriptional machinery required for cell cycle progression or cell cycle arrest. In quiescent cells or cells reentering G1 from mitosis, Rb exists in an underphosphorylated or dephosphorylated state. Phosphorylation of Rb occurs late in G1 and modifies the activities of regulatory complexes that are required for gene expression linked to the onset of S phase [246]. The activities of the cdk are downregulated by a series of inhibitors (designated CDIs) and mediators of ubiquitination, which signal destabilization and/or destruction of these regulatory complexes in a cell cycle-dependent manner [247]. The cyclin inhibitory protein (CIP) class of CDI includes the proteins p21, p27, and p57. Growth arrest is, in part, due to induction of the cyclin-dependent
31 kinase inhibitor protein p21, which can interact with multiple cyclin – cdk complexes. The INK class is represented by proteins p15, p16, p18, and p19, which are linked to apoptosis control mechanisms. Expression of cell cycle regulatory proteins, cyclins, and cyclin-dependent kinases appears not to be solely confined to control of proliferation but, for example, associated differentiation in bone osteoblasts and nonosseous cells [248 – 250]. Cell cycle regulatory factors, particularly cyclin E, have been noted in several systems involved in the regulation of differentiation, in myoblasts [251], in osteoblasts [249], and in promyeloid cell differentiation into macrophages [252]. During osteoblast differentiation, cyclin-dependent kinase inhibitors (cdki) are also developmentally expressed. The cdki p21 (CIP/WAF1) is expressed in the growth period. In contrast, p27 (KIP-1) is expressed in the immediate postproliferative period and is upregulated again during differentiation [253]. Studies characterizing bone abnormalities associated with null mutations of cell cycle and cell growth regulatory factors have revealed their significance in providing signals for the control of both the number and the differentiation of bone-related cells. For example, marrow harvested from p27-/- mice shows a three- to fourfold increase in osteogenic nodule formation compared to wild type. Thus absence of this cdk inhibitor allows the marrow population to extend their growth phase, increasing the cell numbers. This expansion of the osteoprogenitor population is consistent with the larger size of the animals and the proportionally increased cortical width of the long bones [253]. Investigations of the effects of growth factors and osteogenic hormones on cell cycle target genes are increasing our understanding of their precise molecular mechanisms in the regulation of growth, differentiation, and apoptosis of osteoprogenitor cells and osteoblasts (Fig. 6). Several studies have reported BMP-2 and BMP-4 induction of cell cycle arrest in the G1 phase that is mediated by enhanced expression of the p21 cyclin inhibitor [254] and rapid induction of cyclin G, a cyclin that is increased after the induction of p53 by DNA damage [255]. Both of these events are linked to the induction of apoptosis, and in the developing tooth, p21 and BMP-4 are coexpressed in cells destined to undergo apoptosis in a transitional epithelial structure known as the enamel knot [256]. The apoptotic-promoting effects of BMP-2 have been reported to oppose the estradiol-induced growth of human breast cancer cells. Where estradiol stimulates cyclins and cyclin-dependent kinases, the BMP induction of the cyclin kinase inhibitor p21 leads to the inactivation of cyclin D1 [257]. The abundance of TGF- and BMPs in the early stages of osteoblast maturation and the targeting of BMP action to p21 may provide a mechanism not only for promoting osteogenic differentiation, but for apoptosis of proliferating cells that are recruited to the bone surface and may not progress to the mature osteocyte.
32
LIAN AND STEIN
Other cytokines and growth factors that target the proliferation phase have their effects coupled through p21. IL-6 promotes differentiation and exhibits antiapoptotic effects on human osteoblasts [258]. The effects of IL-6 on the p21 promoter are mediated by STAT-binding proteins and a STAT response element in the p21 promoter. Fibroblast growth factors are classic mitogens of the osteoprogenitor pool, as well as modulators of osteoblast differentiation [259 – 263]. FGF signaling also activates STAT1 and p21 [264], a mechanism that accounts for the ability of FGF-2 to induce both mitogenic responses and growth arrest in cancer cells [264,265]. TGF- also inhibits cell cycle progression in part through the upregulation of p21 gene expression [253,266]. Regulation of the p21 promoter is mediated by the TGF- induction of Smad3 and Smad4 [266,267]. The steroid hormone 1,25(OH)2D3 exerts antiproliferative effects in undifferentiated cells also mediated by the enhanced expression of p21 [268] and p27 [253]. This finding is consistent with the high levels of p27 in mature osteoblasts and 1,25(OH)2D3 induction of markers of the mature osteoblast phenotype. It is becoming increasingly evident that each step in the regulatory cycles (cell cycle, cyclin/cdk cycle, cdki cycle) governing proliferation is responsive to multiple signaling pathways and has multiple regulatory options. The diversity in cyclin – cyclin-dependent kinase complexes accommodates the control of proliferation under multiple biological circumstances and provides functional redundancy as a compensatory mechanism. Similarly, the inhibitors of cyclin – cdk complexes bind to and regulate multiple cyclin – cdk-containing complexes at several checkpoints [240,269,270]. The regulatory events associated with these proliferation-related cycles support control within the contexts of (a) responsiveness to a broad spectrum of positive and negative mitogenic factors, (b) cell – cell and cell – extracellular matrix interactions, (c) monitoring genome integrity and invoking DNA repair and/or apoptotic mechanisms if required, and (d) competency for differentiation. Perturbation of any of these cell cycle regulatory mechanisms can result in unregulated or neoplastic growth.
C. Osteoblasts: Producer of the Bone Extracellular Matrix When the preosteoblast ceases to proliferate, a key signaling event occurs for development of the mature osteoblast from the spindle-shaped osteoprogenitor. The osteoblast expresses all of the differentiated functions required to synthesize bone. Osteoblasts are defined in vivo by their appearance along the bone surface as large cuboidal cells actively producing matrix (Figs. 4 and 5), which is not yet calcified (osteoid tissue). Several structural features characterize this osteoblast, including its size and
cuboidal morphology, a round distinguishing nucleus at the base of the cell (opposite to the bone surface), a strongly basophilic cytoplasm, and a prominent Golgi complex located between the nucleus and the apex of the cell [271]. At the ultrastructural level, one observes an extremely welldeveloped rough endoplasmic reticulum with dilated cisternae and a dense granular content, and a large circular Golgi complex consisting of multiple Golgi stacks. These are typical characteristics of a secretory cell. The osteoblast synthesizes and vectorially secretes most of the bone ECM protein; others are accumulated. Fetal bone is enriched in type III collagen, whereas type I collagen accounts for 80 – 90% of the osteoid in the adult with minor collagens type V. Specialized noncollagenous proteins for cell adhesion and with calcium and phosphatebinding properties are found in variable amounts changing with age. The most abundant noncollagenous proteins include osteonectin, osteocalcin, bone sialoprotein, and osteopontin; and based on many in vitro studies, it is assumed that they participate in specialized functions necessary for both the structural integrity of bone tissue and bone turnover. The protein properties and gene regulation of bone extracellular matrix constituents are detailed elsewhere in this volume (see Chapter 4). The primary functional activity of the active surface osteoblast is production of an extracellular matrix with competency for mineralization. In this regard, the high level of the tissue-nonspecific alkaline phosphatase (TNAP) (bone, kidney, liver isoform) and the ability to synthesize a number of noncollagenous proteins that are in either representative or restricted abundance in mineralized tissues are important features. For example, hypertrophic chondrocytes and odontoblasts share some phenotypic features with osteoblasts, as osteocalcin expression [272,273] and high alkaline phosphatase activity, sufficient to allow for histochemical detection in cells associated with the active formation of mineralized matrix. Alkaline phosphatase activity, a hallmark of the osteoblast phenotype, is a widely accepted marker of new bone formation and early osteoblast activity. Gradations of enzyme intensity and mRNA expression are found in bone with lowest levels (or absence) in osteocytes and osteoprogenitors and maximal levels in surface osteoblasts and hypertrophic chondrocytes at the mineralization front [134,229]. Alkaline phosphatase activity is still considered critical to the initiation of mineralization, a concept supported by characterization of the genetic defect in hypophosphatasia [274]. Several of the noncollagenous bone-related proteins are implicated in initiating or regulating the mineral phase of bone from studies of the biochemical, molecular, and functional properties [275,276] (Table 1). However, ablation of the genes encoding some of the more abundant and bone restricted noncollagenous proteins (osteocalcin [277,278], osteopontin [279], biglycan [280]) have resulted in subtle changes in
CHAPTER 2 Osteoblast Biology
the mineral phase of bone. Future directions in which double mutants are derived may reveal severe mineralization defects. The anabolic activities of osteoblasts are regulated in large part by the insulin-like growth factor (IGF)/IGFbinding protein (IGFBP) system [281; reviewed in 282]. IGF-I and IGF-II are synthesized in may tissues and both are highly expressed in active osteoblasts [283]. IGFs stimulate cell proliferation and collagen synthesis [284 – 286] and, at the same time, inhibit matrix collagen degradation by decreasing collagenase 3 transcription [287]. The synthesis of IGF-I is regulated by physiologic mediators of bone formation, PTH stimulates [288], whereas glucocorticoids [289] and growth factors (e.g., FGF-2, PDGF-2) are inhibitory [290]. The activities of IGF-I and IGF-II are regulated by a family IGFBPs, designated IGFBP-1 through IGFBP-6 [291]. These binding proteins have either stimulatory effects (e.g. IGFBP-5) [286,292 – 294] or inhibitory activity (e.g., IGFBP4) [294] and appear to be expressed at different levels in subpopulations of osteoblasts [295]. Several clinical studies are revealing associations of IGF-1 and IGFBPs with metabolic bone diseases [296 – 298, and reviewed by 299]. Regulation of bone development and cellular differentiation by the extracellular matrix (ECM) is well established [300 – 302] (see Chapter 4). Osteoblast differentiation and functional activity are supported by cell – matrix interactions [303,304]. Functional studies establishing requirement of the type I collagen and other ECM components in promoting osteoblast/osteocyte differentiation have been carried out by modifying the production of osteoblastsecreted products or culturing osteoblasts on various matrices [305 – 308]. The molecular mechanisms mediating osteoprogenitor/osteoblast – matrix interactions are being identified. A spectrum of integrins have been shown to be expressed by osteoblasts [309]. Interactions between integrin receptors and fibronectin are required for both osteoblast differentiation [310,311] and cell survival [312]. Osteoblasts appear to use 1 integrins to adhere to the full range of RGD-containing bone matrix proteins. Disruption of the collagen – 21 interaction suppresses expression of the osteoblast phenotype [313,314]. The noncollagenous RGD containing bone matrix protein, bone sialoprotein (BSP), was also shown to mediate the collagen-induced osteoblast differentiation of bone marrow cells [315]. This activity of BSP is consistent with the reported coordinated expression of 3 integrin and BSP during osteoblast differentiation in vitro [316,317]. Based on early findings of Vukicevic et al. [306], who reported laminin-1 stimulates osteoblast differentiation, and recent observations that osteoprogenitors selectively attach to laminin-1 [318], it will be instructive to address the specificity of cell – matrix interactions with respect to subpopulations of osteoblast lineage cells. Studies are identifying integrin receptors in the
33 regulation of BMP-2-mediated differentiation of mesenchymal cells [319]. In addition to cell – matrix interactions, cell – cell communication is important for the differentiation and maturation of osteoblasts. Cytoplasmic processes on the secreting side of the surface osteoblast extend deep into the osteoid matrix and are in contact with the extended cellular processes of osteocytes. Junctional complexes (gap junctions) are often found between the osteoblasts on the surface as well as between cellular processes. In this manner, surface osteoblasts establish cell – cell communication with neighboring cells in the mineralized matrix. Gap junctions are a structure of six multiple protein units (connexins) that couple with an identical unit in a neighboring cell to form a channel connecting the two cytoplasms. Studies in osteoblasts [320,321] suggest that the selective utilization of connexin proteins contributes to the modulation of molecular permeability. Several members of the cadherin family of cell – cell adhesion proteins are expressed in osteoblasts, including cadherin-11, cadherin-4, N-cadherin, and OB-cadherin [50,51]. N-cadherin is present on proliferative preosteoblastic cells and may support osteoblast differentiation [322], but is lost as they become osteocytic [323]. In contrast, OB-cadherin is barely detected in osteoprogenitor cells and is upregulated in alkaline phosphatase-expressing cells [324]. Indeed, the relative abundance of cadherin defines the differentiation pathway of mesenchymal precursors to specific lineages [51,325]. Expression of R-cadherin, N-cadherin, and cadherin-11, present in progenitor cell lines, is modified in response to differentiation, e.g., R-cadherin is downregulated and cadherin-11 upregulated in response to BMP2-induced osteogenesis. In addition, ICAM-1 and VCAM-1 have been reported on the osteoblast surface, thereby providing a potential mechanism for T-cell interactions that contribute to the regulation of bone turnover [326]. Signaling pathways from the extracellular matrix through the cytoskeleton and finally to the nucleus, which allow expression and upregulation of bone-specific and bone-related genes, need to be investigated. For example, -catenin, which colocalizes and coprecipitates with cadherins [50], is a potential candidate. CD44, the hyaluronate receptor, is a nonintegrin adhesion receptor that is linked to the cytoskeleton. CD44 has been identified as a useful marker for osteocyte differentiation [200,201] and is also expressed in osteoclasts [327].
D. Osteocytes and Bone-Lining Cells: Gatekeepers of the Structural Integrity of Bone As the active matrix-forming osteoblast becomes encased in the mineralized matrix, the cell differentiates further into osteocytes. Labeling studies suggest that the
34 transition from an osteoblast to an osteocyte is approximately 3 – 5 days [169,328]. The osteocyte is considered the most mature or terminally differentiated cell of the osteoblast lineage. Osteocytes are embedded in bone matrix occupying spaces (lacunae) in the interior of bone and are connected to adjacent cells by long cytoplasmic projections, enriched in microfilaments, and lie within channels (canaliculi) through the mineralized matrix. These cell processes maintain contact with other osteocytes or with processes from the cells lining the bone surface [134,329]. It is estimated that 25,000/mm3 osteocytes are found embedded within the bone, contributing to the metabolic functions of bone [330]. Some, but not all, of the biochemical features of the osteoblast are expressed in the osteocyte. The morphologic properties of osteocytes change as the surrounding osteoid mineralizes and reflects their functional activity (Fig. 3). Being derived from osteoblasts, a young osteocyte has many of the ultrastructural characteristics of a cell involved in protein synthesis (rough endoplasmic reticulum, large Golgi), except that there is a decrease in the volume of the cell. An older osteocyte, located deeper within the calcified bone, shows less of these features; and in addition, glycogen stores become evident in its cytoplasm. Osteocytes have been shown to synthesize new bone matrix at the surface of the lacunae, and there is some evidence for their ability to resorb calcified bone from the same surface [331]. The osteocyte is a terminally differentiated cell, not capable of cell division even when separated from its matrix. In isolated cultures, they retain their cellular projections [332]. Osteocytes will be phagocytized and digested together with the other components of bone during osteoclastic bone resorption. On quiescent bone surfaces, the osteoblast develops into a flattened bone-lining cell of a single layer forming the endosteum against the marrow and underlying the periosteum directly on the mineralized surfaces. These osteoblasts are in direct communication with the osteocytes within the mineralized matrix through cellular processes that lie within the canaliculi. In the adult, the majority of bone surfaces are occupied by another osteoblast subtype with distinct phenotypic features. The bone-lining cell displays a flat and highly elongated cell shape with a spindle-shaped nucleus and fewer organelles than active osteoblasts [130, 271,329,333]. They are considered to provide a selective barrier between bone and other extracellular fluid compartments and contribute to mineral homeostasis by regulating the fluxes of calcium and phosphate in and out of bone fluids [333]. The organization of osteocytes in bone reflects their function and their capacity to respond, allowing these cells to communicate and transmit regulatory signals. The osteocytes and surface-lining cells form a continuum, or syncytium, by connection of their cytoplasmic projections through gap junctions that facilitate the exchange of both mechanical and metabolic signals [334,335] for
LIAN AND STEIN
responsiveness to physiologic demands on the skeleton. Osteoblasts and osteocytes are coupled metabolically and electrically through different gap junction proteins called connexins, described earlier [321,336,337]. Rapid fluxes of bone calcium across these junctions are thought to facilitate the transmission of information between osteoblasts on the bone surface and osteocytes within the structure of bone [338]. This structural organization and the direct contact of the active osteoblast or surface lining cells with the osteocytes is consistent with the concept that bone cells, responding to varying physiological signals, can communicate their responses. Bone-lining cells receive the majority of systemic and local signals and can transmit these to osteocytes. Reciprocally, mechanical forces on the bone produce stress-generated signals that are perceived by osteocytes, which then transmit the regulatory information to surface osteoblasts. Stress-generated electric potentials experienced by bone are either produced by strain in the organic components (piezo electric potential) or result from electrolyte fluid flow produced by deformation of the bone (streaming potential) [339 and reviewed in 340]. Thus, the ability of bone to act as a tissue responding to physiological homeostatic demands and functioning as a structural connective tissue organ to meet physical demands depends on communication among its resident cells. Studies in bone tissue and isolated cells following applied stress have advanced our understanding of osteocyte functions and responses [341]. Osteocytes produce IGF-1 and release prostaglandins in response to stress [342]. Direct evidence that osteocytes sense mechanical loading [343] has been demonstrated by rapid changes in metabolic activity by [3H]-uridine uptake [344], increased metabolic activity (e.g., glucose-6-phosphate dehydrogenase) [345], increased gene expression [346 – 348], and activation of a volume-sensitive Ca2 influx pathway potentiated by PTH [349]. Signals induced by fluid flow that have been reported include prostaglandin PGE-2, cAMP, and nitrous oxide [350 – 352]. In addition, extracellular matrix receptors, such as the integrins and CD44 receptors, are thought to mediate cellular sensing of mechanical loads [353]. The disruption of cell – matrix interactions by loading could induce a mechanical twisting of the integrins [354]. Integrins are tightly coupled to the cytoskeleton [355] and together the integrin – cytoskeleton complex facilitates the transduction of mechanical signals that may ultimately lead to modifications in gene expression [356]. Thickening of actin stress fibers and increased synthesis of cytoskeleton components in osteoblasts in response to mechanical strain have been documented [357]. The majority of the evidence to date suggests that mechanical tension can trigger bone remodeling and may favor bone formation. Increased osteopontin expression and synthesis may facilitate bone remodeling by osteoclasts [348,358,359]. However, it has been reported that mechanical strain inhibits expression of the osteoclast
CHAPTER 2 Osteoblast Biology
differentiation factor TRANCE [360]. Related to bone formation, several studies have demonstrated significant increases in PGE-2, which stimulates IRF-1 in osteocytes in response to mechanical strain [361 and reviewed in 282]. BMP-2 and BMP-4 (but not other BMPs) were induced in a model of distraction osteogenesis [362]. Thus, mechanical strain induces factors for the proliferation, differentiation, and anabolic activities of osteoblasts [363]. The life span of osteoblast lineage cells is dependent on several factors. Because more osteoblasts are recruited to bone remodeling sites than can be organized on the bone surface for further differentiation by mineralizing osteoid, a high percentage of surface preosteoblasts will die [364]. Apoptosis of preosteoblast clusters may be triggered by the lack of an adequate ECM and appropriate cell – matrix interactions for survival [312]. In vitro, osteoblast apoptosis is particularly evident in cells on the surface of multilayered bone nodules formed in primary calvarial cell cultures [365]. Apoptosis is a general mechanism for limiting organ size in embryonic development [366] and in the adult when there is a need to regenerate tissue. Here growth factor and cytokine effects on apoptosis of specific cell subpopulations in bone are likely to be contributing to tissue turnover [258,364,367]. In contrast to osteoblasts, osteocytes are very long lived in their lacunae, but will undergo apoptosis when their structural integrity is compromised. Microfracture in bones [368] and disruption of cell – cell contacts with the consequent inability to receive stimulatory signals and cell nutrients will lead to apoptosis. Increased empty lacunae and apoptotic cells (detected by DNA fragmentation using the TUNEL assay) are observed during bone turnover in aged human bone [369,370], in glucocorticoidtreated mouse models [371], and following estrogen withdrawal [372]. Parathyroid hormone [373], bisphosphonates [374], and estrogen [375,376] have been reported to prevent apoptosis. Of significance, calbindin-D (28k), which is expressed in osteoblasts, suppresses apoptosis by reducing caspase-3 activity through protein – protein interactions [377]. More studies are required to address intracellular apoptotic control mechanisms and pathways operative in the environment of the osteoblast [365].
IV. CELLULAR CROSS-TALK OF OSTEOBLAST LINEAGE CELLS: FUNCTIONS IN CALCIUM HOMEOSTASIS, BONE TURNOVER, AND HEMATOPOIESIS The biological significance and unanswered questions of the interrelationship of bone tissue cells with the hematopoietic and immune systems were highlighted at a National Institute of Health conference [378]. Osteoblasts
35 control osteoclast differentiation from hematopoietic precursors. They also support long-term bone marrow cultures and regulate hematopoiesis by the production of stimulatory factors (e.g., GM-CSF) [379,380], as well as by cell – cell interactions between early hematopoietic cells and osteoblasts via 1 integrins on CD34 cell and various cell adhesion on bone marrow stromal cells (e.g., VCAM1) [381] [reviewed in 382]. The immune and bone organ systems are linked by the production of multiple cytokines from T lymphocytes regulating bone turnover by the modulation of both osteoblast and osteoclast activities. Cross-talk between osteoblasts and other cellular systems is beginning to be investigated. Endothelial cell (EC) and osteoblast cross-talk is likely to be important for vascular invasion into the bone matrix. Osteoblasts secrete paracrine factors that regulate endothelial cell function [383], including vascular endothelial growth factor (VEGF) and its receptors [384]. VEGF secreted by ECs has been reported to enhance the anabolic effects of vitamin D3 on osteoblasts [385] and to be necessary for angiogenesis during endochondral bone formation in vivo [386]. Of note, bone sialoprotein, which is upregulated in osteogenic tumors and mediates cell attachment via V3 integrins, can promote adhesion of endothelial cells [387]. An important function of osteoblast lineage cells is their response to endocrine factors and the production of paracrine and autocrine factors for the sequelae of events mediating bone turnover. At all stages, from the initial activation of bone resorption to formation of new bone at the resorption site in the adult (bone remodeling unit), crosstalk between osteoblast lineage cells and other cell phenotypes is necessitated for the regulation of bone remodeling. The coupling of osteoblast and osteoclast activities through cross-talk is mediated by several mechanisms, which are detailed later. Direct interactions between a ligand on osteoblast lineage cells and its receptor on mononuclear preosteoclasts activate signaling cascades for osteoclast differentiation. Numerous indirect pathways are operative in which calciotrophic hormones and cytokines stimulate the secretion of factors from one cell type that activate or suppress activities of the other cell phenotype. Osteoprogenitors in the marrow, surface osteoblasts, and osteocytes have receptors for cytokines [reviewed in 388,389], parathyroid hormone, 1,25(OH)2D3 [390,391], and estrogen [392], which are key regulators of osteoclast activities. In addition to cytokine and hormone feedback loops, the transcriptional control of genes involved in the activation or suppression of osteoclast and osteoblast functions are coordinated. Two observations suggest a potential interrelationship between osteoblasts and osteoclasts by an unknown mechanism. Annexin II was identified as a vitamin D3 receptor [393] mediating the nongenomic rapid effects of 1,25(OH)2D3, which increase intracellular calcium in osteoblasts [394 – 396]. Annexin II was also reported to be
36
LIAN AND STEIN
secreted by osteoclasts and stimulate osteoclast differentiation and GM-CSF [397], an autocrine and paracrine product of osteoblasts [398]. These features provide mechanisms for regulating osteoblast and osteoclast activities to support calcium homeostasis and bone turnover. Osteoblasts and stromal osteoprogenitors primarily support osteoclastogenesis by the secretion of soluble cytokines and through cell – cell interactions with osteoclast precursors. Stromal osteoblastic cells are the major source of the macrophage colony-stimulating factor (M-CSF/CSF1) and the IL-6 family of cytokines, which are potent stimulators of bone resorption and participate in osteoclastogenesis at early and later stages. Other cytokines, such as TNF- and IL-1, which are predominantly derived from monocytes and have mitogenic effects on the mononuclear osteoclast precursor, feed back on stromal cells to regulate their production of M-CSF, as well as IL-11, another key regulator of osteoclast formation [reviewed in 399; 400,401]. IL-6, IL-11, and other bone-resorbing factors, such as LIF and oncostatin M, transduce their regulatory signals through the gp130 signal transduction pathway. Cytokine production by human bone marrow stromal cells can be effected by age and estrogen status [402]. PTH and vitamin D stimulate osteoblast production of IL-6, as do IL-1,
TNF- and TGF-. The mechanisms by which these factors affect the generation of early osteoclast precursors from CFU-GM colonies, osteoclast differentiation, and activity have been well documented (in several reviews [403 – 406]) (see Chapter 3). Two discoveries provide insight into mechanisms for the essential role of osteoblasts in mediating osteoclastogenesis directly through cell – cell interactions (Fig. 7). One is the characterization of osteoprotegerin (OPG), also designated osteoclastogenesis inhibitory factor (OCIF), a secreted protein with strong homology to the TNF receptor family. OPG is expressed in several tissues, including bone, cartilage, kidney, and blood vessels [407,408]. Several experimental approaches established OPG as a soluble factor competent to inhibit osteoclast differentiation [409 – 411]. Expression of the gene in osteoblast lineage cells is upregulated by calcium and is downregulated by the glucocorticoid, dexamethasone [411]. These findings are consistent with the conditions required for osteoclast differentiation in cocultures of bone marrow stromal cells and spleen cells [412]. Overexpression of OPG in transgenic mice resulted in severe osteopetrosis [407]. The presence of F4/80 positive osteoclast precursors in these mice suggested that OPG inhibits terminal stages of osteoclast differentiation.
FIGURE 7 Osteoblast – osteoclast cross-talk in the regulation of bone turnover. M-CSF and OPGL/RANKL are osteoblast factors that promote osteoclast differentiation: M-CSF binds to the c-fos receptor on the mononuclear preosteoclast, whereas RANKL mediates cell – cell interaction with the RANK receptor to promote fusion of the preosteoclast to the differentiated multinucleated resorbing cell. Other regulatory hormones, cytokines, and bone matrix proteins are shown. A soluble “decoy” OPG receptor is also produced by osteoblasts that can block osteoclast precursor interactions with stromal cells.
37
CHAPTER 2 Osteoblast Biology
In search of the specific ligand for OPG, several groups cloned a novel member of the transmembrane TNF ligand family, leading to the discovery of an osteoclast differentiation factor [411,415,416]. The ligand is identical to a TNF cytokine family member designated TRANCE (TNF-related activation induced cytokine) [417], a cytokine that regulates T-cell-dependent immune responses and to RANK (receptor activation of NF-,) a TNF receptor family member cloned from T cells [418]. A standard nomenclature has been proposed for the TNF factors related to bone resorption, designating the ligand which is expressed at high levels in osteoblasts as RANKL [419]. Importantly, RANKL was demonstrated to have competency for inducing osteoclast formation from hematopoietic cells in the absence of stromal cells [415,416]. The studies define RANKL as acting directly with RANK on osteoclast precursors. The soluble decoy receptor which blocks this interaction remains designated OPG (Fig. 7). Mice with a disrupted RANKL gene completely lack osteoclasts because of the inability of osteoblasts to support their differentiation [420]. Of interest, activating mutations in RANK have been identified as the cause of the bone disorder familial expansile osteolysis [421]. Thus, RANKL is essential for the differentiation of osteoclasts from the early stages of mononuclear cell fusion, activation, and survival. Together RANK, RANKL, and M-CSF represent essential factors required for coupling stromal/osteoblastic cells to the formation of osteoclasts and are approximately controlled by cytokine and hormonal mediators of bone resorption for regulated bone turnover.
It is instructive to consider regulated expression of RANK and RANKL in osteoblasts in an environment poised for bone resorption or bone formation. Expression of the RANKL ligand in osteoblasts is upregulated by stimulator of bone resorption while OPG is downregulated ([411]; and reviewed in Hofbauer et al. [422]). Notably, regulatory elements for the osteoblast differentiation transcription factor CBFA1 occur in both the OPF and OPG-L gene promoters [423,424]. CBFA1 upregulates OPG in osteoblasts [423], thereby suppressing osteoclastogenesis when bone formation is needed. Regulation of the RANKL promoter by CBFA1 was not examined; however, CBFA1 regulates RANKL mRNA, while vitamin D decreases CBFA1 cellular levels [425,426] and increases RANKL expression. These changes are consistent with vitamin D mediating osteoclast differentiation. Thus, through transcriptional control of the CBFA1 promoter by the same steroid hormone that regulates osteoclastogenesis, as well as transcriptional regulation of OPG and OPG-L by CBFA1, feedback loops can contribute to a balance between bone resorption and bone formation (illustrated in Fig. 8). The OPG system is discussed extensively in Chapters 3, 6, 12, and 14. Crosstalk between osteoclast activity and osteoblast recruitment maintains the fidelity of bone tissue organization. Following the activation and resorption phases of the bone remodeling sequence, the recruitment, proliferation, and differentiation of osteoprogenitors and osteoblasts on the resorbed surface is accomplished in part by the resorbed bone microenvironment. Stored growth factors in the bone matrix provide a local concentration to initiate the formation phase by recruitment of osteoprogenitors to the
FIGURE 8 CBFA and vitamin D3 regulate bone turnover by coordination of gene transcription. Vitamin D promotes osteoclast differentiation in part by increasing RANKL. Modest downregulation of the mouse CBFA1 promoter by 1,25(OH)2D3 [425] may ensure physiologic levels of this factor. CBFA1 is a positive regulator of OPG expressing bone resorption while promoting bone formation. Thus, OPG (RANK) /RANKL ratios may be regulated through the combined activities of vitamin D and CBFA.
38
LIAN AND STEIN
resorbed bone surface. Here the role of TGF-1 is central to increasing osteoblast differentiation at sites of bone resorption [427].
V. OSTEOBLASTS IN VITRO: STAGES IN DEVELOPMENT OF THE OSTEOBLAST PHENOTYPE A. Cell Culture Models 1. PRIMARY CELL CULTURES Valuable insight into the biology and pathology of bone has been obtained over the past several years from cultured mammalian and avian osteoblasts that undergo developmental expression of genes and establishment of bone tissue-like organization, analogous to osteoblast differentiation in vivo. The clinical relevance of culture models should not be overlooked. Primary cell cultures offer
TABLE 2 Cell Culture Models for Studying Commitment, Growth, Differentiation, and Physiological Responses of Osteoblastic Cells Primary cell cultures for osteoblast differentiation Periosteum [441] Calvariae [432] Trabeculae [434] Cortical bone [435] Marrow stroma [437,439] Tumor-derived cell lines [477] ROS 17/2.8; ROS 24/1 UMR106 SaOS-2 U2-OS MG-63 Clonal and transformed cell lines i. Pluripotential/bipotential C3H10T1/2 [462] ROB-C26 [459] 2T3 [182] MLB13 Myc [183] ii. Monopotential MC3T3-E1 (preosteoblast) [125] C2C12 (premyoblast) [461] MLO-Y4 (osteocytic) [473] iii. Marrow stromal origin ST-2 [463] W20-17 [37]
several advantages, particularly for studying cell growth control mechanisms and differentiation in the context of a mineralizing matrix. Primary cultures often reflect the in vivo status of the osteoblast and phenotypic properties of a mouse model [253,261] (Table 2). Cultured calvarial osteoblasts from osteopetrotic rats that exhibit precocious and intensified mineralization within the intact animal during development retain this defect in vitro [428,429]. The pathology is reflected by parallel in vitro and in vivo modification in the temporal sequence of cell growth and tissue-specific gene expression as well as by features of cell organization and extracellular matrix mineral deposition that are revealed at the histochemical and ultrastructural levels. In several mouse models, marrow cultures supporting osteoblast differentiation and osteoclastogenesis have revealed cell type-specific defects and provided insight into molecular mechanisms [430]. The method of releasing populations of cells from a bone extracellular matrix by outgrowth cultures of bone fragments or by enzymatic digestion with trypsin and collagenase was developed several decades ago [431 – 435]. Calvarial or trabeculae-derived osteoblasts support the exploration of mechanisms associated with the differentiation of cells committed to the osteoblast phenotype. Marrow cell cultures permit the evaluation of gene expression related to the commitment of stem cells to the osteoblast lineage, initial stages of differentiation, and the limits of plasticity that characterize pluripotent osteoprogenitor cells [133,171,202, 436 – 440]. Osteoprogenitor cells that differentiate have also been isolated from the periosteum [441,442]. Primary cultures of normal diploid fetal rat calvarialderived osteoblasts form multilayered cellular nodules having bone tissue-like properties [143,443 – 447], whereas adherent marrow-derived cells give rise to osteoblastic colonies. Osteoblasts isolated from more mature bone as described for chick [446], bovine, and human [447] form a uniform layer of mineralized matrix. The isolated cells from bone tissue are not a heterogeneous mixture, but will become heterogeneous in long-term culture as the bone matrix is synthesized and mineralized osteoblasts result from outgrowth or released cells, particularly from fetal tissues. Although primary cell cultures are a mixed population of cells, ultrastructural, histochemical, single cell gene analyses of the cell layers and bone nodule colonies revealed that preosteoblastic, osteoblast, and osteocyte-like cells can be identified in these different models. Primary cell cultures may not always be appropriate or practical for some lines of experimentation. They are limited in their ability to maintain phenotypic properties with passaging. The cumulative observations from primary cultures of fetal and adult bone, marrow, and periosteum reported over the last decade reveal considerations for the following in interpretation of these studies. The age of isolation influences the growth properties and representations of subpopulations of bone-forming cells. The expression of
CHAPTER 2 Osteoblast Biology
osteogenic and other phenotypic responses appears to also be related to bone sites and cell passages. Thus, studies of osteoblast activities must be controlled carefully. Cells closer to the progenitor/preosteoblast stage are differentiated more readily in vitro. It should additionally be noted that protocols have been developed for the culture of human bone cells from multiple sites offering viable options for the application of culture techniques to the evaluation of skeletal diseases or for the evaluation of selective responses of osteoblasts that have been observed in vivo [435,448,449]. Although limitations must be acknowledged, the effective use of cultured human osteoblasts in assessing functional activity related to disease has been validated, e.g., with cells from patients with Paget’s disease [450], osteogenesis imperfecta [451], and other skeletal disorders [452]. Several studies have indicated that osteoblasts from osteoporotic patients and control subjects exhibit few differences when compared in vitro [453,454]. Caution must be exercised in the interpretation of osteoblast responses to agents in vitro relative to in vivo effects. For example, osteoblasts harvested from bone biopsies of osteoporotic patients treated with fluoride exhibit increased proliferative activity [455]. In contrast, fluoride is not mitogenic to osteoblasts in culture [456]. 2. TUMOR-DERIVED, TRANSFORMED, AND IMMORTALIZED CELL LINES Several classes of nonmalignant, clonal murine and rat cell lines have facilitated the investigation of biochemical and molecular parameters of osteoblast differentiation (Table 2). Calvarial-derived cells that undergo spontaneous immortalization during passage in cultures [433,457] exhibit selective, and often unstable [458], expression of bone cell phenotypic properties and responses to steroid hormones and growth factors. Pluripotent clonal cell lines are being used frequently to examine molecular mechanisms of cell phenotype commitment. The ROB-C26 [459], isolated from neonatal rat calvaria, has tripotential (osteogenic, myogenic, and adipogenic) properties. A myogenic cell line (C2C12 cells) has been well characterized with respect to its differentiation potential in response to TGF-1 and BMP-2/4 [221]. TGF-1 blocks myotube formation only, whereas BMP-2/4 induces the osteoblast phenotype. The expression of BMPs and their receptors [460] and Smad signaling factors, as well as responses to different BMPs [461], have been studied [43,408]. The pluripotential C3H10T/2 cell line, which differentiates into adipocytes, chondrocytes, and myotubes when treated with azacytidine, will also differentiate to osteoblasts in response to BMP-2 [219,462]. The stromal cell line W20-17 [37], which was subcloned (by limiting dilution from bone marrow), and other similar mouse stromal lines (ST-2 cells) [463] have bipotential properties. These cells, which are widely used to support osteoclast formation, can
39 be differentiated to mature osteoblasts by BMP-2. The MC3T3-E1 monopotential preosteoblastic cell line [464] that was obtained by subjecting calvarial-derived osteoblasts (from C57B1/6 mice) to scheduled passaging retained the ability to undergo a development sequence of gene expression [465]. These cells establish a mineralized bone extracellular matrix [466] and, like primary rat calvarial cells, differentiation can be modulated by various stimuli. In recent years, subclones from these cell lines have been isolated with distinct properties, likely reflecting stable stages of osteoblast maturation [125]. Viral transformation of calvarial-derived rodent [467], marrow stromal cells, or human osteoblasts has also provided model systems for addressing regulatory mechanisms operative in bone cells [183,467 – 469]. The introduction of a temperature-sensitive viral gene, which at permissive temperatures selectively supports proliferation or postproliferative phenotypic gene expression, offers new options in the pursuit of skeletal regulatory mechanisms in both murine [182,183] and human [470,471] cells that are involved in development and bone disease. Characterized properties reveal that viral transformation appears to restrict their properties to mono- or bipotential lineages at an early stage of commitment, e.g., mouse limb bud-derived cell lines (MLB13 Myc clones 14 and 17) [183] and human marrow stromal cells [472]. The 2T3 cell line from mouse calvaria [182] and the MLO-Y4 osteocyte-like cell line from long bone [473] have been established from transgenic mice harboring the SV40 large T antigen. Stable cell lines permit conditional and reversible expression of osteoprogenitor or osteoblast phenotypic properties that are developmentally regulated. The limitation is that these cell lines may not reproducibly support formation of a mineralized matrix. Osteosarcoma cell lines, which typically express a limited gene characteristic of bone cells in vivo, have been utilized by many investigators to support studies directed to the control of genes expressed during osteoblast proliferation and differentiation. A series of rat osteosarcomaderived cell lines (ROS) exhibit a wide range of phenotypic responses. The ROS 17/2.8 [474] exhibits most properties of mature osteoblasts, including high levels of osteocalcin [475]; the ROS 24/1 cell line lacks the vitamin D receptors [396]. Several human osteosarcoma cell lines have been widely used for the characterization of hormonal responses. These include the rat UMR-106 and human MG-63, SAOS-2, U2-OS, and TE-85 cell lines. The large quantity of cells available permits isolation and characterization of gene regulatory molecules that control transcription at the level of the gene, as well as the signaling pathways that mediate responsiveness to growth and phenotypic cues [476]. However, caution must be exercised in assuming that identical regulatory parameters are operative in normal diploid osteoblasts and osteosarcoma cells
40 [477 – 482]. The abrogation of key components of growth control and proliferation – differentiation interrelationships in transformed and tumor cells results in a consequential coexpression of cell growth and tissue-specific genes that are, in normal diploid cells, sequentially expressed in response to stringently regulated physiological signaling mechanisms.
B. Developmental Sequence of Gene Expression Characterizes Stages in the Osteoblast Differentiation Pathway Primary cell cultures and cultures of established lines that produce an organized bone-like matrix provided a basis for studies that have mapped the temporal expression of cell growth and tissue-specific genes during the progressive establishment of the osteoblast phenotype [446,477, 483 – 486]. Profiles of gene expression defined developmental stages of osteoblast maturation and allowed for investigating regulatory mechanisms that support the progression of osteoblast growth and differentiation and developmental stage-specific responses to physiological mediators of bone formation and remodeling that include growth factors and steroid hormones [486 – 497]. These characterizations have facilitated the investigation of gene regulatory signaling pathways and control mechanisms associated with development and maintenance of bone tissuelike organization. The sequential expression of cell growth and tissue-specific genes presented in Fig. 9 (see also color plate) has been mapped during progressive development of the bone cell phenotype in cultured osteoblasts and marrow stromal cells from several species and several sites [128,167,320, 483,486,498,499] by the combined application of Northern blot analysis, in situ hybridization, nuclear run-on transcription and histochemistry. Four principal developmental periods can be defined [499] by expression of the major functional bone matrix proteins, often designated “phenotypic markers”. Initially, proliferation supports expansion of the osteoblast cell population to form a multilayered cellular nodule and biosynthesis of the type I collagen bone extracellular matrix (ECM). At this time, genes requisite for the activation of proliferation (e.g., c-myc, c-fos, c-jun) and cell cycle progression (e.g., histones, cyclins) are expressed together with the expression of genes encoding growth factors (e.g., FGF, IGF-I), cell adhesion proteins (e.g., fibronectin), and others associated with the regulation of ECM biosynthesis (e.g., TGF, type I collagen). However, several of the BMPs reach peak expression in immediate postproliferative osteoblasts [500] and may function in the regulation of related BMPs supporting osteoblast growth and differentiation [501,502]. For example, BMP-2 is not only autoregulated but additionally downregulated by
LIAN AND STEIN
BMP-4 and BMP-6 in osteoblasts [503]. However, BMP-2 enhances BMP-3 and BMP-4 expression during the mineralization stage [504]. Following the initial proliferation period, a second stage of gene expression is associated with the maturation and organization of the bone ECM. Genes that contribute to rendering the extracellular matrix competent for mineralization (e.g., alkaline phosphatase) are upregulated. Collagen synthesis continues and undergoes cross-link maturation [505]. Two principal transition points are key elements of this temporal expression of genes that support the progression of differentiation (Fig. 9B, bottom). These transitions have been established experimentally and defined functionally as restriction points during osteoblast differentiation to which developmental expression of genes can proceed but cannot pass without additional cellular signaling [483]. The first transition point is at the completion of the proliferation period when genes for cell cycle and cell growth control are downregulated, and expression of genes encoding proteins for extracellular matrix maturation and organization is initiated. The absence during the proliferation period of gene expression observed in postproliferative mature osteoblasts is called “phenotype suppression” [506]. The model is supported by the binding of regulatory factors abundant in proliferating osteoblasts (e.g., oncogene-encoded [507] or helix – loop – helix proteins [84,85,508,509]) to regulatory elements in postproliferative expressed genes, which results in the suppression of transcription of the genes until later in development. The second transition is at the onset of extracellular matrix mineralization. Signals for the third developmental period involve gene expression related to the accumulation of hydroxyapatite in the ECM. Genes encoding several proteins with mineral binding proteins (e.g. osteopontin, osteocalcin and bone sialoprotein) exhibit maximal expression at this time when mineralization of the bone tissue-like organized matrix is ongoing. This profile suggests functional roles for these proteins in regulation of the ordered deposition of hydroxyapatite. A fourth developmental period follows in mature mineralized cultures during which time collagenase is elevated, apoptotic activity occurs, and compensatory proliferative activity is evident [510,511]. Although the biological significance of gene expression during the fourth developmental stage remains to be formally established, it appears to serve an editing/remodeling function for modifications in the bone extracellular matrix that sustain the structural and functional properties of the tissue. This working model of osteoblast differentiation has been supported by studies that demonstrate the stage-specific expression of family members of numerous classes of proteins that regulate development of the bone cell phenotype. These include, for example, the IGF family of binding proteins [512], cyclin and cyclin-dependent kinases [249], bone morphogenic proteins [492,500], oncogene-encoded proteins
CHAPTER 2 Osteoblast Biology
41
FIGURE 9 Stages of maturation in primary cultures of bone-derived cells. (A) Morphology and histochemical staining of fetal rat calvaria-derived osteoblasts at three stages: proliferation (P) with toluidine blue, matrix maturation (MM) with alkaline phosphatase staining, and mineralization (M) detection by the von Kossa stain. Multilayers of osteoblasts form a typical bone nodule having a mineralized extracellular matrix. (See also color plate.) (B) Northern blot analyses of marker genes expressed maximally at each stage. (See also color plate.) (C) Schematic illustration of several genes temporally expressed during 35 days of culture. Increases in mRNA transcripts of OC and OP during the culture period parallel calcium (Ca2 ) deposition. The increase and peak levels of collagenase are related to remodeling or editing of the extracellular matrix. Induction of genes reflecting the apoptosis of cells associated with the mineralized nodules is shown [365,483,494]. (D) The reciprocal and functionally coupled relationship between cell growth and differentiation-related gene expression is illustrated by arrows. Broken vertical lines between the three principal periods indicate experimentally established transition or restriction points requiring the downregulation of cell growth and ECM signaling events for the progression of differentiation.
42 [507], heat shock proteins [513], homeodomain proteins [514], and TGF- receptors [515]. Notably, several genes associated with skeletal cells, osteonectin [305,485] and matrix Gla protein [516], are expressed constitutively. The sequential and stringently regulated expression of genes that defines periods of osteoblast phenotype development is illustrated schematically as a reciprocal and functionally coupled relationship between proliferation and differentiation (Fig. 9B, bottom). The extent to which developmental expression of genes is sequential and mutually exclusive rather than concomitant is in part dependent on the position in the osteoblast lineage. For example, while in calvarial-derived osteoblasts, alkaline phosphatase expression is only observed in postproliferative cells, and alkaline phosphatase is compatible with proliferation in earlier stage bone marrow-derived osteoprogenitor cells. The interrelationship between proliferation and differentiation often times provides explanations for discordance between in vivo and in vitro effects. FGF, for example, is a potent mitogen for mesenchymal cells, chondrocytes, and osteoblasts and stimulates endosteal bone formation [212]. FGF-2 is useful for fracture repair in vivo [517]. However, FGF effects in vitro are stage specific. If added to a proliferation stage, mineralization and subsequent maturation of osteoblasts are inhibited [260,263,518]. In part, the results are mediated by continued proliferation modifying the differentiated phenotype and induced levels of collagenase modifying the ECM in a manner incompatible with ECM mineralization. In postproliferative cultures, FGF-2 promotes differentiation [263].
C. Differential Responsiveness of Hormones and Growth Factors as a Function of Stage of Osteoblast Maturation Based on the in vitro models of osteoblast differentiation, we can better understand the properties and physiologic responses of the cells at their individual stages of differentiation. This is best exemplified by selective responses in bone marrow stromal cell cultures and calvarial-derived osteoblasts to growth factors [263,491,518] and hormones [487,519 – 525]. The parathyroid hormone will promote the differentiation of preosteoblasts but suppress late stages of maturation [526,527]. Although caution should be exercised in translation from in vitro to in vivo effects of PTH on bone formation, these studies indicate that even pulsed PTH administration may increase bone formation by stimulating the proliferation of progenitors, and not by anabolic effects of differentiated osteoblasts [528]. It is established that TGF- stimulates the replication of progenitor cells and directly stimulates collagen synthesis. When proliferating calvarial osteoblasts are exposed to TGF-, a block in differentiation is observed
LIAN AND STEIN
[490 – 492,529]. The mitogenic effects of TGF- are not apparent on mature postproliferative osteoblasts. The insulin-like growth factors [284] and fibroblast growth factors (FGF-1, FGF-2) [530], which have distinct roles in development [11], also have selective effects on osteoblast subpopulation in vitro. The various steroid hormones, including glucocorticoids [519,531], 1,25(OH)2D3 [487,489,520], and estrogen [521], also have selective effects, either promoting differentiation of the cells at early stages of maturation or inhibiting anabolic activities and promoting resorptive properties of the osteoblast at later stages. In general, growth factors and steroid hormones have the most robust responses in immature osteoblasts and can radically modify their program of differentiation when added to proliferating cells. Glucocorticoids regulate numerous osteoblast-expressed genes, which contribute to bone formation, including cytokines, growth factors, and bone matrix proteins [reviewed in 193,531,532] (see Chapter 44). The molecular mechanisms by which glucocorticoids exert selective effects on a particular gene are complex, but numerous examples have been documented. Increases in alkaline phosphatase, osteocalcin, and collagen are observed at early maturation stages, but inhibition of these genes occurs in differentiated osteoblasts [519,533,534]. Both transcriptional and posttranscriptional mRNA stability contribute to positive and negative regulation of a gene, as shown for osteocalcin [519] and collagen [535]. Glucocorticoids promote osteoblast colony formation in human and rat marrow-derived cells and accelerate osteoblast differentiation in proliferating calvarial derived cells, reflected by both increased numbers and size of the bone nodules and early mineralization [439,493,494]. Because postproliferative cultures cannot be stimulated to produce more mineralizing nodules, the mature osteoblast is refractory to growth stimulation by dexamethasone [129,493,494,536]. This may be a consequence of glucocorticoids pushing cells to terminal differentiation or apoptosis, thereby depleting the pool of preosteoblasts capable of nodule formation [129,371,493]. It is noteworthy that dexamethasone exerts an antiproliferative effect on mouse osteoblasts [525,537] and blocks their maturation. These findings, together with glucocorticoid effects on osteoclast activity (reviewed in [193], contribute to glucocorticoid-induced osteopenia observed in vivo when pharmacologic doses of glucocorticoids are required [290,531, 538 – 541] (see Chapter 44). The active metabolite of vitamin D, 1,25(OH)2D3, has complex effects on the skeletal system related to targeting of many cell types, dose, and timing [542 – 544] (see Chapter 9). Vitamin D is a biphasic regulator of osteoblast activity for bone formation and bone resorption. Vitamin D regulates the expression of genes in osteoblasts that form the bone ECM or provide signals for osteoclast differentiation. The up-
CHAPTER 2 Osteoblast Biology
regulation by 1,25(OH)2D3 of numerous osteoblast parameters related to bone matrix formation and mineralization (e.g., collagen, alkaline phosphatase, osteopontin, osteocalcin, matrix Gla protein), and bone resorption by cytokines (e.g., osteoprotegerin [545]), reflects influences of the hormone on osteoblast function and regulation of bone turnover. However, pharmacological doses and long-term exposure of this hormone to rats can result in abnormalities of bone formation [544,546,547]. In vitro analysis of 1,25(OH)2D3 in primary cultures of rat osteoblasts show stage-dependent effects. The steroid is antiproliferative in the growth period and can block formation of the mineralized nodule when introduced during the growth period [487 –489,524,548] or stimulate differentiation-related gene expression in mature osteoblasts. Because of these properties, acute versus continuous exposure of cells to 1,25(OH)2D3 can lead to opposing results [487,489]. The selective effects based on the stage of osteoblast maturation are evident by in situ hybridization studies, in vivo and in vitro, and accompanying changes in cell morphology in vitro [136,137,143,536,549]. These changes in morphology and gene expression may relate to bone-resorbing effects of 1,25(OH)2D3 on surface-lining osteoblasts, which must (a) retract to allow for osteoclast interaction with the bone mineral and (b) provide local factors for the induction of osteoclast activity. Studies from many groups using different osteoblast models have reinforced two important concepts: (1) that the stage of osteoblast maturation influences the selective responsiveness of specific genes to hormones or growth factors and (2) that there is a window of responsiveness of a cell during which the factor can alter development and maintenance of the bone cell phenotype. These analyses have provided clinically relevant information toward an understanding of the consequences incurred by the osteoblast when exposed to therapeutic agents that may stimulate or inhibit cell proliferation or differentiation.
VI. MOLECULAR MECHANISMS MEDIATING PROGRESSION OF OSTEOBLAST DIFFERENTIATION A. Stage-Specific Expression of Transcription Factors and Their Contribution to OsteoblastSpecific Gene Expression The progression of osteoblast differentiation requires the sequential activation and suppression of genes that encode phenotypic and regulatory proteins. Key molecular events initiate from the extracellular matrix and signaling molecules, such as BMPs and TGF-s, that can indirectly result in a cascade of gene expression (see earlier discussion). In addition, regulatory factors that directly engage in protein – DNA, as well as protein – protein interactions, are
43 important mechanisms for both the activation and the suppression of genes reflecting stages of osteoblast maturation [550]. Transcription factors described in Section I, which contribute to position and pattern formation in the developing embryo, provide mechanisms for regulating the progression of osteoblast differentiation in the adult. The selective representation of these factors during osteoblast differentiation and family members within a class of transcription factors (Fig. 2), as well as evidence for their functional consequences (e.g., by forced expression, antisense or antibody blocking studies) on osteoblasts, provides compelling evidence for their regulatory effects in driving osteoblast maturation. The focus on transcription factors that regulated the developmental expression of osteocalcin [reviewed in 551,552] has been further supported by the characterization of mice carrying null mutations of several transcription factors (Table 1). Homeodomain proteins, Fos/Jun family members, helix – loop – helix factors, and RUNX2/CBFA1 proteins are among the well-characterized transcription factors (see Fig. 2). The homeodomain protein-binding sites contribute to bone-specific expression of several genes, collagen type I [553,554], osteocalcin [514,555], and, most recently, regulated expression of bone sialoprotein by Dlx5 [556]. During osteoblast differentiation in vitro, Msx-2 is expressed maximally in the preosteoblast and is subsequently downregulated [514] with the onset of osteocalcin expression. In situ hybridization studies confirm the reciprocal expression of osteocalcin and Msx-2 in cells of developing bone [557]. In a reciprocal fashion, Dlx-5 appears in the postproliferative osteoblast and increases during mineralization [558]. It appears that there is selectivity for the expression of homeodomain factors during osteoblast differentiation as well as developmental variations in activities [557]. Consistent with the developmental expression of Msx-2 in early osteoprogenitors and Dlx-5 in mature osteoblasts, Msx-2 downregulates OC, whereas Dlx-5 can positively regulate OC and collagen as well. Protein – protein interactions between Msx-2 and Dlx-5 are also determinants for developmental activities of these factors, as demonstrated by Zhang et al. [559]. The molecular mechanisms of this heterodimer interaction in alleviating suppression of the osteocalcin gene were demonstrated [555]. These activities may in part reflect the sequence content of homeodomain responsive promoter elements in genes that are expressed developmentally in osteoblasts [514,553,558,560,561]. Helix – loop – helix transcription factors are expressed at high levels in proliferating osteoblasts, thereby facilitating downregulation of the gene in these immature cells [85,509]. This complexity ensures developmental, tissuespecific regulated expression of the postproliferative bonespecific genes in osteoblasts. The fos and jun family members also exhibit developmental stage-specific expression and activities during osteoblast
44
LIAN AND STEIN
differentiation in vivo [562] and in vitro [507]. Studies of c-fos expression in transgenic mouse models reveal the importance of c-fos in establishing the osteoblast phenotype [563]. In vivo immunohistochemical staining reveals that c-fos is expressed in osteoprogenitor cells, in the perichondrium and periosteal tissues, but not in mature osteoblasts [562]. During osteoblast differentiation in vitro, c-fos and jun expression is expressed maximally in proliferating preosteoblasts. However, retention and upregulation of fra2 and jun-D protein levels and mRNA expression were observed postproliferatively in differentiated osteoblasts [564]. Confirmation of the functional activities of these factors in regulating progressive maturation of the osteoblast phenotype has been demonstrated by several experimental approaches, including specific repressor and enhancer activities of c-fos/ c-jun and fra2/jun-D, respectively, on the osteocalcin promoter, as well as inhibition of osteoblast differentiation resulting from antisense inhibition of fra2 expression [564]. RUNX2/CBFA1, described in Section I as an essential transcription factor for osteogenesis, regulates osteoblast differentiation. It is noteworthy that CBFA1 is present in the osteoprogenitors where osteoblast-specific genes are not expressed. CBFA1 mRNA, cellular protein levels, and DNA-binding activity are increased during osteoblast differentiation [103] and may contribute to specific CBFA1 activities in cells at distinct stages of osteoblast differentiation. In consideration of the precise functional activities of CBFA1 in the regulation of osteoblastic genes, multiple molecular mechanisms are operative. Oftentimes expression of a gene can be regulated during development (embryonic versus postnatal) or in relation to cell-type specificity by the production of protein isoforms resulting from either alternative usage of ATG initiation sites or splicing variant. Several examples of alternatively spliced products imparting tissue specificity have been recorded in genes. The Fibronectin gene gives rise to multiple splice variants, and the splicing process is cell type, developmental stage, and age regulated [565,566]. In 2(XI) collagen, alternative
FIGURE 10
splicing gives rise to proteins that are expressed in a tissuespecific manner. Isoforms lacking exons 6 – 8 of the 2(XI) collagen gene code for a protein that is found in cartilaginous tissue of developing limbs and axial skeleton, whereas the transcript containing exons 6 – 8 is found in nonchondrogenic tissue such as calvaria and periosteum [567]. These mechanisms may contribute to the timing of expression or the protein isoforms may exhibit specialized functions. Several NH2 and COOH-terminal CBFA1 isoforms have been reported (Fig. 10) [93,568,569]. Both the aminoterminal isoforms, type I (MRIPV) and type II (MASNS) isoforms, have nearly equivalent transactivation activity on different promoters and in various cell lines [118, 570 – 572]. To date, no studies have systematically addressed the relative expression of these CBFA1 isoforms and other isoforms resulting from C-terminal splicing events. Noteworthy in the regulation of CBFA1 activities are posttranslational modifications, including phosphorylation by MAPK [573]. In addition, CBFA1 interacts with several coregulatory proteins, both activators and suppressors of CBFA1 activity [95,96,574,575,575,576], which also appear to be developmentally expressed during osteoblast differentiation [574]. CBFA1 also has properties that can mediate modifications in chromatin, an important component of tissue-specific transcriptional control that is described in the following section.
B. Contributions of Nuclear Architecture to Transcriptional Control during Osteoblast Differentiation It is becoming increasingly apparent that nuclear architecture provides a basis for support of the stringently regulated modulation of cell growth and tissue-specific transcription necessary for the onset and progression of osteoblast differentiation. Here, multiple lines of evidence point to contributions by three levels of nuclear organization to in
CBFA1 NH2 terminal isoforms present in osteoblast lineage cells. Schematic illustration of the origin of CBFA1 regulated by an upstream (PU) promoter is transcribed at MET-1 in exon 1 (MASNS isoform or type II designation by Harada et al. [118]). The type I isoform (MRIPV), regulated by the downstream promoter (PD), was the first CBFA1 protein to be cloned from T cells [94], which is also expressed in mesenchymal lineage cell lines, chondrocytes, and osteoblasts [114 – 118]. The MASNS isoform was initially identified by Stewart et al. [96a] and originates at MET 69 regulated by the same downstream promoter (PU) from which OSF-2 was first characterized [102]. OSF-2 is poorly transcribed from the MET-1, which does not reside in a Kozak sequence [96b].
CHAPTER 2 Osteoblast Biology
vivo transcriptional control where structural parameters of the genome are functionally coupled to regulatory events. The primary level of gene organization establishes a linear ordering of promoter regulatory elements. This representation of regulatory sequences reflects competency for the responsiveness to physiological regulatory signals as discussed earlier. The osteocalcin gene provides a paradigm for the involvement of nuclear organization in transcriptional control that is linked to bone formation, homeostatic regulation, and bone remodeling. The regulatory elements of the bone-specific osteocalcin gene are organized in a manner that supports responsiveness to physiologic mediators and developmental expression in relation to bone cell differentiation. Characterized regulatory elements and cognate transcription factors can support osteocalcin suppression in nonosseous cells, immature osteoblasts, and growthstimulated osteoprogenitors/osteoblasts, transcriptional activation in postproliferative cells, and steroid hormone enhancement (Fig. 11). For example, the OC box (99 to 76 bp) is a multipartite element that binds several classes of transcription factors. The OC box binds homeodomain proteins (e.g., Msx-2, Dlx-5 [514,577]) and a bone tissuespecific complex of unknown origin, designated the OC box-binding protein (OCBP), which contributes to the activation of gene transcription [578,579]. The minimal OC promoter containing the OC box is sufficient to support osteoblast-specific expression in vitro. A bipartite element in the proximal promoter confers FGF-2 and cAMP responsiveness [580]. This element regulates the suppression of osteocalcin synthesis in response to numerous modulators of cell growth, including IGF-1, bFGF, cAMP, and PTH [581]. A series of AP-1 sites occur in the OC gene promoter [506,582,583], including one that is the TGRE [584]. Fos (c-fos, fral, fra2) and jun (c-jun, jun-D, jun-B) oncogeneencoded families of transcription factors form homo or heterodimeric complexes that regulate gene expression at AP-1 motifs. These AP-1 sites provide another example of regulatory sequences that contribute to positive and negative regulation dependent on the biological circumstances. The c-fos/c-jun heterodimer represses, whereas the fra2/jun-D heterodimer activates transcription. The latter is more abundant in postproliferative osteoblasts. The vitamin D responsive element (VDRE) functions as an enhancer [585 – 588], binding the transcriptionally active VDR/RXR heterodimer complex (see Chapter 9). The core motif, two steroid half elements with a three nucleotide space, is highly conserved [589 – 591]. However, subtle variations, both within the core domain and in the flanking sequences, render VDRE promoter elements of various genes selectively ligand responsive in a developmental and tissue-specific manner. Specificity of VDRE utilization is further conferred by protein – DNA and/or protein – protein interactions in addition to the VDR/RXR complexes
45 [479,591 – 595]. The osteocalcin VDRE transcription factor complex appears to be a target for modifications in vitamin D-mediated transcription by other physiologic factors, including glucocorticoids [592], TNF-a [596], and retinoic acid [597 – 601]. It is evident from available findings that the linear organization of gene regulatory sequences is necessary but insufficient to accommodate the requirements for physiological responsiveness to homeostatic, developmental, and tissue-related regulatory signals. Parameters of chromatin structure and nucleosome organization are a second level of genome architecture. There is a requirement to render promoter regulatory elements competent for protein – DNA and protein – protein interactions that mediate positive and negative controls. Additionally, activities of regulatory complexes at the proximal and distal promoter must be integrated. Modifications in chromatin reduce the distance between promoter elements, thereby supporting interactions between the modular components of transcriptional control. Each nucleosome (approximately 140 nucleotide base pairs wound around a core complex of two each of H3, H4, H2, and H2B histone proteins) contracts linear spacing by sevenfold. Folding of nucleosome arrays into solenoid-type structures provides a potential for interactions that support synergism between promoter elements and responsiveness to multiple signaling pathways. The molecular mechanisms that mediate chromatin remodeling are being defined [reviewed in 602 – 605]. A family of proteins comprising multimeric protein complexes has been described in yeast (SWI/SNF complex) and in mammalian cells that promote transcription by altering chromatin structure [604 – 608]. These alterations render DNA sequences containing regulatory elements accessible for binding cognate transcription factors and mediate protein – protein interactions that influence the structural and functional properties of chromatin. Multiple lines of evidence suggest that the remodeling of nucleosomal structure involves alterations in histone – DNA and/or histone – histone interactions. Histone acetylation and phosphorylation posttranslational modifications have been functionally linked with changes in nucleosomal structure that alter the accessibility to specific regulatory elements [605]. Core histone hyperacetylation enhances the binding of most transcription factors to nucleosomes [609 – 611]. Alterations in the chromatin organization of the osteocalcin (OC) gene promoter during osteoblast differentiation provide a paradigm for remodeling chromatin structure and nucleosome organization that is linked to a long-term commitment to phenotype-specific gene expression [612 – 615]. In nonosseous cells, the packing of chromatin contributes to the extent that promoter elements are accessible to transcriptional activation complexes. An array of nucleosomes on the OC promoter in nonosseous cells contributes to maintaining the suppression of gene transcription. Figure 11 schematically depicts modifications in chromatin structure and
46
LIAN AND STEIN
FIGURE 11
Developmental remodeling of nucleosome organization in the osteocalcin gene promoter correlates with transcriptional activity during the differentiation of normal diploid osteoblasts. The positioning of nucleosomes in the osteocalcin gene promoter was determined by combining DNase I, micrococcal nuclease, and restriction endonuclease digestions with indirect end labeling [272]. Filled circles represent the placement of nucleosomes with the shadows indicating movement within nuclease-protected segments. Note the differences in nucleosome organization of the gene between nonosseous cells when the OC gene is not transcribed (line 1) and in osteoblasts expressing osteocalcin (lines 2 and 3). Vertical arrows correspond to the limits of the distal (600 to 400) and proximal (170 to 70) sites of DNase I hypersensitivity (DHS), which are observed in osteoblasts when the OC is expressed. DHS is increased in response to vitamin D (represented by larger arrows). Positions of CBFA1 elements (sites A, B, C), the VDRE (465 to 437), the osteocalcin box (OC box 99 to 77), and TATA element (31 to 28) are designated.
nucleosome organization that parallel competency for transcription and the extent to which the osteocalcin gene is activated and transcribed in bone cells. Basal expression and enhancement of osteocalcin gene transcription are accompanied by two changes in the structural properties of chromatin when the gene is activated in a bone cell. When the gene is activated in osteoblasts, there is a rearrangement in nucleosome placement [613]. A single nucleosome becomes positioned between proximal regulatory elements, and distal steroid hormone-dependent regulatory sequences provide a basis for accessibility of transactivation factors to cognate sequences (Fig. 11). This model is derived from experimen-
tal evidence. DNase I hypersensitivity is detected in two regulatory domains of the promoter encompassing the VDRE and CBFA sequences in the proximal and distal promoter [612 – 614]. Vitamin D treatment enhances DNase I hypersensitivity. These domains contribute to tissue-specific and vitamin D enhancer activity. A third level of nuclear architecture that contributes to transcriptional control is provided by the nuclear matrix [616]. The anastomosing network of fibers and filaments that constitute the nuclear matrix supports the structural properties of the nucleus as a cellular organelle and accommodates structural modifications associated with
47
CHAPTER 2 Osteoblast Biology
proliferation, differentiation, and changes necessary to sustain phenotypic requirements of specialized cells. As the intricacies of gene organization and regulation are elucidated, the implications of a fundamental biological paradox become strikingly evident. With a limited representation of gene-specific regulatory elements and a low abundance of cognate transactivation factors, how can sequence-specific interactions occur to support a threshold for the initiation of transcription within nuclei of intact cells? Viewed from a quantitative perspective, the in vivo regulatory challenge is to account for the formation of functional transcription initiation complexes with a nuclear concentration of regulatory sequences that is approximately 20 nucleotides per 2.5 yards of DNA and a similarly restricted level of DNAbinding proteins. Regulatory functions of the nuclear matrix include, but are by no means restricted to, DNA replication [617], gene location [618,619], imposition of physical constraints on the chromatin structure that support the formation of loop domains, concentration, and targeting of transcription factors [620 – 623], RNA processing and transport of gene transcripts [624 – 628], and posttranslational modifications of chromosomal proteins, as well as imprinting and modifications of chromatin structure [629]. Chromatin loop domains (10 – 100 kb) are tethered to components of the nuclear matrix through MAR (matrix attachment regions) sequences.
The initial indication of linkage between the nuclear matrix and control of osteoblast differentiation was first provided by the observation of a parallel representation of nuclear matrix proteins and developmental stage-specific gene expression [630]. Direct involvement of the nuclear matrix in the control of bone-specific gene transcription is provided by several lines of evidence. Two nuclear matrix DNA-binding proteins regulate the activation of osteocalcin gene transcription, identified as NMP1. YY1 modulates vitamin D enhancer activity and NMP2 characterized as RUNX/CBFA1 is a transcriptional activator protein and contributes to promoter organization [99]. The mechanism by which transcription factors contribute to overall promoter architecture is becoming clear. Transcription factors and chromatin-remodeling factors associated with the nuclear matrix, such as CBFA factors [631] and the YY1 transcription factor [632], which regulate osteocalcin gene activity, can contribute to conformational modifications in the promoter structure. Through protein – DNA interactions of the CBFA element with nuclear matrix-associated CBFA factors, local chromatin changes are induced by coactivator/corepressor proteins associated with the DNAbinding complex, and the OC promoter becomes poised for transcription. Figure 12 extends the model presented in Fig. 11 by illustrating the three-dimensional organization of the promoter that is permissive for the binding of
FIGURE 12 Role of the nuclear matrix-associated CBFA1 factor in organizing promoter architecture. A three-dimensional representation of the osteocalcin promoter is shown. The interaction of OC CBFA recognition motifs with CBFA1 provides a conformation stability mediated by the tight association of CBFA1 with the nuclear matrix scaffold. This conformation, together with the positioned nucleosome, facilitates the integration of regulatory signals between the proximal (TATA) and the distal (VDRE) elements necessary for basal and vitamin D enhancer activity. Mutations of three CBFA elements in the rat OC gene resulting in loss of DNase I hypersensitivity and vitamin D responsiveness support this model [647].
48
LIAN AND STEIN
FIGURE 13
CBFA-interacting proteins regulating CBFA transactivation functions. Coregulator proteins interact directly with the indicated domains of CBFA transcription factors. Structurally and functionally homologous segments include the conserved DNA-binding runt homology domain, transcriptional activation, and suppression domains [643,711,712], as well as subcellular targeting signals [631,713]. The promoter organizing functions of CBFA factors may involve interacting proteins, including CBF [714], necessary for DNA binding of CBFA to regulatory elements, coactivators ALY [633] and p300, the latter having histone acetylase (HAT) activity [634], and Groucho/TLE [574,575,639], a negative regulator of CBFA enhancer activity. In addition, other transcription factors [635], Smad signal transduction proteins [123], and steroid receptors [637] contribute to synergistic gene expression with CBFA1 (shown in Fig. 14) [reviewed in 94,95].
transactivation factors and protein – protein interactions between distal and proximal elements (e.g., TFIIB TATA box-binding protein and the VDR/RXR). We postulate that this promoter architecture is mediated and stabilized by transcription factors that are targeted to subnuclear domains, such as CBFA1. Among the numerous CBFA-interacting proteins that have been identified (Fig. 13), several have enzyme activities for chromatin remodeling (Fig. 14). Coactivators include ALY [633] and p300, which has histone acetylase (HAT) activity [634], as well as interacting transcriptional regulators [635], ear-2 [636], signal transduction proteins [123], and steroid receptors [637,638]. Corepressor modulators include HES [575] and the negative regulatory factor Groucho/TLE [576,639], which is also associated with the nuclear matrix [640]. This protein has histone deacetylase activity and a histone H3-interacting domain [641,642]. We have shown that the highest levels of Groucho at the mRNA and cellular protein levels are observed in early stage osteoblasts [574]. Groucho/TLE is also abundant in muscle tissue and may contribute to the suppression of osteocalcin in nonosseous cells and proliferating osteoblasts. As indicated in Fig. 13 (bottom), cooperative interactions between CBFA and other transcription factors, through as yet undefined mechanisms, have been reported. The synergistic transactivation of gene transcription between CBFA
and the Ets factors is well documented [95,96], as well as CBFA with C/EBP factors [643] and Smad factors. These interactions provide options for positive and negative regulation of a spectrum of CBFA-regulated genes within the cell or related to a cellular phenotype. Of interest, osteoblast-specific synergistic interactions between an estrogen response element and CBFA elements have been reported [638]. These respective activities contribute to the enhancer and suppressor transcriptional properties of CBFA factors as illustrated in Fig. 14 [574,575,644]. Several reports have indicted that not all osteoblast-expressed genes containing CBFA regulatory sequences are activated by Cbfa1, some are repressed; e.g., BSP [645] and the CBFA1 gene are autoregulated by CBFA1 [646]. The significance of the localization of these transcription factors in specific subnuclear domains, as well as their ability to interact with chromatin modifying proteins, is related to tissue-specific, as well as steroid hormone-dependent gene control of transcription. Colocalization of these regulatory components that are functionally linked to activation and suppression have been established in situ [574]. Validation of these models is further provided by studies in which the CBFA sites in the rat OC promoter were mutated. Three CBFA motifs are strategically positioned in the bone-specific rat osteocalcin promoter. Sites A and B flank the vitamin D response element in the distal
49
CHAPTER 2 Osteoblast Biology
FIGURE 14
Model of CBFA transcriptional regulation of gene activation and suppression mediated by modifications in chromatin architecture. CBFA factors interact with histone acetylases (HATs) (e.g., p300) to open chromatin domains for the binding of regulatory factors that lead to transcriptional activation. CBFA interactions with coregulators having associated histone deacetylase activity (HDAC) result in the suppression of transcription.
promoter, and sites B and C define the orders of a positioned nucleosome in the proximal promoter. The functional significance of multiple CBFA elements in contributing to promoter structure was addressed by mutating individual or multiple AML sites within the context of the native rat ( 1.1 kb) osteocalcin promoterCAT [647]. All three sites are required for maximal basal expression of the rOC promoter. Strikingly, mutation of the three CBFA1 sites leads to abrogation of responsiveness to vitamin D, glucocorticoids, and TGF-. The mechanism of this loss of promoter responsiveness was related to an absence of DNase I hypersensitive sites at the vitamin D response element and over the proximal tissue-specific basal promoter. These findings strongly support a multifunctional role for AML factors in regulating gene expression, not only as a simple transactivator, but also by facilitating modification in promoter architecture and chromatin organization. It is already apparent that local nuclear environments generated by the multiple aspects of nuclear structure are intimately tied to the developmental expression of cell growth and tissue-specific genes. Membrane-mediated initiation of signaling pathways that ultimately influence transcription has been recognized for some time. Here, the mechanisms that sense, amplify, dampen, and/or integrate regulatory signals involve structural as well as functional
components of cellular membranes. Extending the structure – regulation paradigm to nuclear architecture expands the cellular context in which cell structure – gene expression interrelationships are operative. The nuclear structure is a primary determinant of transcriptional control. Thus, the power of addressing gene expression within the threedimensional context of nuclear structure would be difficult to overestimate.
VII. CONCLUDING REMARKS This chapter has presented the cell biology of osteoblasts within the content of our current understanding of the regulatory controls operative in promoting osteoblast differentiation. We have attempted to address how physiologic parameters of gene expression are integrated to support the requirements of bone development and functional integrity of the tissue. During osteoblast phenotype development and bone formation, stages of maturation are defined by levels of expression of subsets of osteoblast genes. A cohort of tissue-specific, developmental, steroid hormone and growth factor-related transcription factor complexes impinge on gene transcription, providing a complex and integrated series of regulatory signals for the selective activation and repression of genes related to osteoblast activity. Additionally,
50
LIAN AND STEIN
the effects of a hormone or growth factor on expression of a specific gene are related to the phenotype as well as the stage of cellular maturation because of the different representation of proteins that contribute to gene regulation. Thus clinical consideration for treatment and therapeutic regimens can be approached with greater knowledge of the consequential effects at the level of gene-regulating responses. We have presented a growing body of evidence for cross-talk at all levels of bone biology, including among the subpopulations of bone cells, between the extracellular matrix and intracellular signaling factors, and finally at the DNA level between transcription factor complexes at multiple elements. Such interactions and complexities need to be considered in future applications of therapeutic strategies.
Acknowledgments The authors gratefully appreciate preparation of the manuscript by Judy Rask and thank colleagues Janet Stein and André van Wijnen for helpful discussions and members of our research group: Chaitali Banerjee, Amjad Javed, Hicham Drissi, Kaleem Zaidi, and Eva Balint. The National Institutes of Health grants supporting the research program related to this chapter include AR45688, AR45689, AR39588, and DE12528. The contents of this chapter are solely the responsibility of the authors and do not necessarily represent the official views of the National Institutes of Health.
References 1. S. Suwanwalaikorn, M. van Auken, M. I. Kang, S. Alex, L. E. Braverman, and D. T. Baran, Site selectivity of osteoblast gene expression response to thyroid hormone localized by in situ hybridization. Am. J. Physiol. 272, E212 – E217 (1997). 2. C. Kasperk, J. Wergedal, D. Strong, J. Farley, K. Wangerin, H. Gropp, R. Ziegler, and D. J. Baylink, Human bone cell phenotypes differ depending on their skeletal site of origin. J. Clin. Endocrinol. Metab. 80, 2511 – 2517 (1995). 3. A. P. Davis, D. P. Witte, H. M. Hsieh-Li, S. S. Potter, and M. R. Capecchi, Absence of radius and ulna in mice lacking hoxa-11 and hoxd-11. Nature 375, 791 – 795 (1995). 4. T. Lufkin, M. Mark, C. P. Hart, P. Dolle, M. LeMeur, and P. Chambon, Homeotic transformation of the occipital bones of the skull by ectopic expression of a homeobox gene. Nature 359, 835 – 841 (1992). 5. E. Laufer, C. E. Nelson, R. L. Johnson, B. A. Morgan, and C. Tabin, Sonic hedgehog and Fgf-4 act through a signaling cascade and feedback loop to integrate growth and patterning of the developing limb bud. Cell 79, 993 – 1003 (1994). 6. F. Rousseau, J. Bonaventure, L. Legeai-Mallet, A. Pelet, J. M. Rozet, P. Maroteaux, M. Le Merrer, and A. Munnich, Mutations in the gene encoding fibroblast growth factor receptor-3 in achondroplasia. Nature 371, 252 – 254 (1994). 7. B. K. Hall and T. Miyake, All for one and one for all: Condensations and the initiation of skeletal development. BioEssays 22, 138–147 (2000). 8. J. Jernvall and I. Thesleff, Reiterative signaling and patterning during mammalian tooth morphogenesis. Mech. Dev. 92, 19 – 29 (2000). 9. B. K. Hall, The induction of neural crest-derived cartilage and bone by embryonic epithelia: An analysis of the mode of action of an epithelial-mesenchymal interaction. J. Embryol. Exp. Morphol. 64, 305 – 320 (1981).
10. R. L. Johnson, R. D. Riddle, and C. J. Tabin, Mechanisms of limb patterning. Curr. Opin. Genet. Dev. 4, 535 – 542 (1994). 11. S. Iseki, A. O. Wilkie, and G. M. Morriss-Kay, Fgfr1 and Fgfr2 have distinct differentiation- and proliferation-related roles in the developing mouse skull vault. Development 126, 5611 – 5620 (1999). 12. B. L. Hogan, Morphogenesis. Cell 96, 225 – 233 (1999). 13. A. D. Weston, V. Rosen, R. A. Chandraratna, and T. M. Underhill, Regulation of skeletal progenitor differentiation by the BMP and retinoid signaling pathways. J. Cell Biol. 148, 679 – 690 (2000). 14. E. Koyama, E. B. Golden, T. Kirsch, S. L. Adams, R. A. Chandraratna, J. J. Michaille, and M. Pacifici, Retinoid signaling is required for chondrocyte maturation and endochondral bone formation during limb skeletogenesis. Dev. Biol. 208, 375 – 391 (1999). 15. L. Niswander, C. Tickle, A. Vogel, I. Booth, and G. R. Martin, FGF4 replaces the apical ectodermal ridge and directs outgrowth and patterning of the limb. Cell 75, 579 – 587 (1993). 16. N. Wanek, D. M. Gardiner, K. Muneoka, and S. V. Bryant, Conversion by retinoic acid of anterior cells into ZPA cells in the chick wing bud. Nature 350, 81 – 83 (1991). 17. C. Thaller and G. Eichele, Isolation of 3,4-didehydroretinoic acid, a novel morphogenetic signal in the chick wing bud. Nature 345, 815 – 819 (1990). 18. H. J. Kim, D. P. Rice, P. J. Kettunen, and I. Thesleff, FGF-, BMPand Shh-mediated signalling pathways in the regulation of cranial suture morphogenesis and calvarial bone development. Development 125, 1241 – 1251 (1998). 19. C. Hartmann and C. J. Tabin, Dual roles of wnt signaling during chondrogenesis in the chicken limb. Development 127, 3141 – 3159 (2000). 20. A. B. Zimrin, M. S. Pepper, G. A. McMahon, F. Nguyen, R. Montesano, and T. Maciag, An antisense oligonucleotide to the notch ligand jagged enhances fibroblast growth factor-induced angiogenesis in vitro. J. Biol. Chem. 271, 32499 – 32502 (1996). 21. S. Artavanis-Tsakonas, K. Matsuno, and M. E. Fortini, Notch signaling. Science 268, 225 – 232 (1995). 22. D. P. Rice, T. Aberg, Y. Chan, Z. Tang, P. J. Kettunen, L. Pakarinen, R. E. Maxson, and I. Thesleff, Integration of FGF and TWIST in calvarial bone and suture development. Development 127, 1845 – 1855 (2000). 23. B. St Jacques, M. Hammerschmidt, and A. P. McMahon, Indian hedgehog signaling regulates proliferation and differentiation of chondrocytes and is essential for bone formation. Genes Dev. 13, 2072 – 2086 (1999). 24. A. Vortkamp, K. Lee, B. Lanske, G. V. Segre, H. M. Kronenberg, and C. J. Tabin, Regulation of rate of cartilage differentiation by Indian hedgehog and PTH-related protein. Science 273, 613 – 622 (1996). 25. P. T. Chuang and A. P. McMahon, Vertebrate Hedgehog signalling modulated by induction of a Hedgehog- binding protein. Nature 397, 617 – 621 (1999). 26. U. I. Chung, B. Lanske, K. Lee, E. Li, and H. Kronenberg, The parathyroid hormone/parathyroid hormone-related peptide receptor coordinates endochondral bone development by directly controlling chondrocyte differentiation. Proc. Natl. Acad. Sci. USA 95, 13030 – 13035 (1998). 27. S. J. Karp, E. Schipani, B. St Jacques, J. Hunzelman, H. Kronenberg, and A. P. McMahon, Indian hedgehog coordinates endochondral bone growth and morphogenesis via parathyroid hormone relatedprotein-dependent and -independent pathways. Development 127, 543 – 548 (2000). 28. B. Lanske, A. C. Karaplis, K. Lee, A. Luz, A. Vortkamp, A. Pirro, M. Karperien, L. H. K. Defize, C. Ho, R. C. Mulligan, A. B. AbouSamra, H. Juppner, G. V. Segre, and H. M. Kronenberg, PTH/PTHrP receptor in early development and Indian hedgehog- regulated bone growth. Science 273, 663 – 666 (1996). 29. J. Massague, TGF-beta signal transduction. Annu. Rev. Biochem. 67, 753 – 791 (1998).
51
CHAPTER 2 Osteoblast Biology 30. A. Erlebacher, E. H. Filvaroff, S. E. Gitelman, and R. Derynck, Toward a molecular understanding of skeletal development. Cell 80, 371 – 378 (1995). 31. N. A. Wall and B. L. Hogan, TGF-beta related genes in development. Curr. Opin. Genet. Dev. 4, 517 – 522 (1994). 32. D. M. Kingsley, The TGF-beta superfamily: New members, new receptors, and new genetic tests of function in different organisms. Genes Dev. 8, 133 – 146 (1994). 33. K. M. Lyons, R. W. Pelton, and B. L. Hogan, Patterns of expression of murine Vgr-1 and BMP-2a RNA suggest that transforming growth factor-beta-like genes coordinately regulate aspects of embryonic development. Genes Dev. 3, 1657 – 1668 (1989). 34. D. M. Kingsley, A. E. Bland, J. M. Grubber, P. C. Marker, L. B. Russell, N. G. Copeland, and N. A. Jenkins, The mouse short ear skeletal morphogenesis locus is associated with defects in a bone morphogenetic member of the TGF superfamily. Cell 71, 399 – 410 (1992). 35. S. Vainio, I. Karavanova, A. Jowett, and I. Thesleff, Identification of BMP-4 as a signal mediating secondary induction between epithelial and mesenchymal tissues during early tooth development. Cell 75, 45 – 58 (1993). 36. A. Yamaguchi, T. Katagiri, T. Ikeda, J. Wozney, V. Rosen, E. Wang, A. Kahn, T. Suda, and S. Yoshiki, Recombinant human bone morphogenetic protein-2 stimulates osteoblastic maturation and inhibits myogenic differentiation in vitro. J. Cell Biol. 113, 681 – 687 (1991). 37. R. S. Thies, M. Bauduy, B. A. Ashton, L. Kurtzberg, J. M. Wozney, and V. Rosen, Recombinant human bone morphogenetic protein-2 induces osteoblastic differentiation in W-20-17 stromal cells. Endocrinology 130, 1318 – 1324 (1992). 38. E. A. Wang, V. Rosen, J. S. D’Alessandro, M. Bauduy, P. Cordes, T. Harada, D. I. Israel, R. M. Hewick, K. M. Kerns, P. LaPan, D. P. Luxenberg, D. McQuaid, I. K. Moutsatsos, J. Nove, and J. M. Wozney, Recombinant human bone morphogenetic protein induces bone formation. Proc. Natl. Acad. Sci. USA 87, 2220 – 2224 (1990). 39. T. K. Sampath, J. C. Maliakal, P. V. Hauschka, W. K. Jones, H. Sasak, R. F. Tucker, K. H. White, J. E. Coughlin, M. M. Tucker, and R. H. Pang, Recombinant human osteogenic protein-1 (hOP-1) induces new bone formation in vivo with a specific activity comparable with natural bovine osteogenic protein and stimulates osteoblast proliferation and differentiation in vitro. J. Biol. Chem. 267, 20352 – 20362 (1992). 40. V. Rosen and R. S. Thies, The BMP proteins in bone formation and repair. Trends. Genet. 8, 97 – 102 (1992). 41. I. Asahina, T. K. Sampath, I. Nishimura, and P. V. Hauschka, Human osteogenic protein-1 induces both chondroblastic and osteoblastic differentiation of osteoprogenitor cells derived from newborn rat calvaria. J. Cell Biol. 123, 921 – 933 (1993). 42. C. H. Heldin, K. Miyazono, and P. ten Dijke, TGF-beta signalling from cell membrane to nucleus through SMAD proteins. Nature 390, 465 – 471 (1997). 43. R. Nishimura, Y. Kato, D. Chen, S. E. Harris, G. R. Mundy, and T. Yoneda, Smad5 and DPC4 are key molecules in mediating BMP-2induced osteoblastic differentiation of the pluripotent mesenchymal precursor cell line C2C12. J. Biol. Chem. 273, 1872 – 1879 (1998). 44. Y. Zhang and R. Derynck, Regulation of Smad signalling by protein associations and signalling crosstalk. Trends Cell Biol. 9, 274 – 279 (1999). 45. G. R. Martin, The roles of FGFs in the early development of vertebrate limbs. Genes Dev. 12, 1571 – 1586 (1998). 46. K. Miyama, G. Yamada, T. S. Yamamoto, C. Takagi, K. Miyado, M. Sakai, N. Ueno, and H. Shibuya, A BMP-inducible gene, dlx5, regulates osteoblast differentiation and mesoderm induction. Dev. Biol. 208, 123 – 133 (1999). 47. Y. S. Hu, H. Zhou, V. Kartsogiannis, J. A. Eisman, T. J. Martin, and K. W. Ng, Expression of rat homeobox gene, rHOX, in developing
48.
49.
50.
51.
52.
53.
54.
55.
56. 57.
58.
59.
60.
61.
62.
63.
64.
and adult tissues in mice and regulation of its mRNA expression in osteoblasts by bone morphogenetic protein 2 and parathyroid hormone-related protein. Mol. Endocrinol. 12, 1721 – 1732 (1998). S. J. Tang, P. A. Hoodless, Z. Lu, M. L. Breitman, R. R. McInnes, J. L. Wrana, and M. Buchwald, The Tlx-2 homeobox gene is a downstream target of BMP signalling and is required for mouse mesoderm development. Development 125, 1877 – 1887 (1998). S. Rastegar, H. Friedle, G. Frommer, and W. Knochel, Transcriptional regulation of Xvent homeobox genes. Mech. Dev. 81, 139 – 149 (1999). S. L. Cheng, F. Lecanda, M. K. Davidson, P. M. Warlow, S. F. Zhang, L. Zhang, S. Suzuki, T. St. John, and R. Civitelli, Human osteoblasts express a repertoire of cadherins, which are critical for BMP-2-induced osteogenic differentiation. J. Bone Miner. Res. 13, 633 – 644 (1998). E. Hay, J. Lemonnier, D. Modrowski, A. Lomri, F. Lasmoles, and P. J. Marie, N- and E-cadherin mediate early human calvaria osteoblast differentiation promoted by bone morphogenetic protein-2. J. Cell. Physiol. 183, 117 – 128 (2000). X. L. Yang, H. Matsuura, Y. Fu, T. Sugiyama, and N. Miura, MFH1 is required for bone morphogenetic protein-2-induced osteoblastic differentiation of C2C12 myoblasts. FEBS Lett. 470, 29 – 34 (2000). M. Yamagata, and M. Noda, The winged-helix transcription factor CWH-3 is expressed in developing neural crest cells. Neurosci. Lett. 249, 33 – 36 (1998). A. Buchberger, M. Schwarzer, T. Brand, O. Pabst, K. Seidl, and H. H. Arnold, Chicken winged-helix transcription factor cFKH-1 prefigures axial and appendicular skeletal structures during chicken embryogenesis. Dev. Dyn. 212, 94 – 101 (1998). D. M. Dean, R. R. Berger, and M. M. Sanders, A winged-helix family member is involved in a steroid hormone-triggered regulatory circuit. Endocrinology 139, 4967 – 4975 (1998). C. Healy, D. Uwanogho, and P. T. Sharpe, Regulation and role of Sox9 in cartilage formation. Dev. Dyn. 215, 69 – 78 (1999). W. Bi, J. M. Deng, Z. Zhang, R. R. Behringer, and B. de Crombrugghe, Sox9 is required for cartilage formation. Nature Genet. 22, 85 – 89 (1999). T. Komori, H. Yagi, S. Nomura, A. Yamaguchi, K. Sasaki, K. Deguchi, Y. Shimizu, R. T. Bronson, Y.-H. Gao, M. Inada, M. Sato, R. Okamoto, Y. Kitamura, S. Yoshiki, and T. Kishimoto, Targeted disruption of Cbfa1 results in a complete lack of bone formation owing to maturational arrest of osteoblasts. Cell 89, 755 – 764 (1997). F. Otto, A. P. Thornell, T. Crompton, A. Denzel, K. C. Gilmour, I. R. Rosewell, G. W. H. Stamp, R. S. P. Beddington, S. Mundlos, B. R. Olsen, P. B. Selby, and M. J. Owen, Cbfa1, a candidate gene for cleidocranial dysplasia syndrome, is essential for osteoblast differentiation and bone development. Cell 89, 765 – 771 (1997). S. Piccolo, Y. Sasai, B. Lu, and E. M. De Robertis, Dorsoventral patterning in Xenopus: Inhibition of ventral signals by direct binding of chordin to BMP-4. Cell 86, 589 – 598 (1996). D. R. Hsu, A. N. Economides, X. Wang, P. M. Eimon, and R. M. Harland, The Xenopus dorsalizing factor Gremlin identifies a novel family of secreted proteins that antagonize BMP activities. Mol. Cell 1, 673 – 683 (1998). L. B. Zimmerman, J. M. Jesus-Escobar, and R. M. Harland, The Spemann organizer signal noggin binds and inactivates bone morphogenetic protein 4. Cell 86, 599 – 606 (1996). L. J. Brunet, J. A. McMahon, A. P. McMahon, and R. M. Harland, Noggin, cartilage morphogenesis, and joint formation in the mammalian skeleton. Science 280, 1455 – 1457 (1998). W. C. Smith and R. M. Harland, Expression cloning of noggin, a new dorsalizing factor localized to the Spemann organizer in Xenopus embryos. Cell 70, 829 – 840 (1992).
52 65. E. Gazzerro, V. Gangji, and E. Canalis, Bone morphogenetic proteins induce the expression of noggin, which limits their activity in cultured rat osteoblasts. J. Clin. Invest. 102, 2106 – 2114 (1998). 66. E. Abe, M. Yamamoto, Y. Taguchi, B. Lecka-Czernik, C. A. O’Brien, A. N. Economides, N. Stahl, R. L. Jilka, and S. C. Manolagas, Essential requirement of BMPs-2/4 for both osteoblast and osteoclast formation in murine bone marrow cultures from adult mice: antagonism by noggin. J. Bone Miner. Res. 15, 663 – 673 (2000). 67. T. Kameda, C. Koike, K. Saitoh, A. Kuroiwa, and H. Iba, Developmental patterning in chondrocytic cultures by morphogenic gradients: BMP induces expression of indian hedgehog and noggin. Genes Cells 4, 175 – 184 (1999). 68. C. Marcelle, M. R. Stark, and M. Bronner-Fraser, Coordinate actions of BMPs, Wnts, Shh and noggin mediate patterning of the dorsal somite. Development 124, 3955 – 3963 (1997). 69. E. Hirsinger, D. Duprez, C. Jouve, P. Malapert, J. Cooke, and O. Pourquie, Noggin acts downstream of Wnt and Sonic Hedgehog to antagonize BMP4 in avian somite patterning. Development 124, 4605 – 4614 (1997). 70. S. Pathi, J. B. Rutenberg, R. L. Johnson, and A. Vortkamp, Interaction of Ihh and BMP/Noggin signaling during cartilage differentiation. Dev. Biol. 209, 239 – 253 (1999). 71. J. M. Hock, J. Fonseca, M. Gunness-Hey, B. E. Kemp, and T. J. Martin, Comparison of the anabolic effects of synthetic parathyroid hormone- related protein (PTHrP) 1 – 34 and PTH 1 – 34 on bone in rats. Endocrinology 125, 2022 – 2027 (1989). 72. M. Mannstadt, H. Juppner, and T. J. Gardella, Receptors for PTH and PTHrP: Their biological importance and functional properties. Am. J. Physiol. 277, F665 – F675 (1999). 73. J. M. Moseley and T. J. Martin, Parathyroid hormone-related protein: Physiological actions. In “Principles of Bone Biology” (J. P. Bilezikian, L. G. Raisz, and G. A. Rodan, eds.), pp. 363 – 376. Academic Press, San Diego, 1996. 74. A. C. Karaplis, A. Luz, J. Glowacki, R. T. Bronson, V. L. Tybulewicz, H. M. Kronenberg, and R. C. Mulligan, Lethal skeletal dysplasia from targeted disruption of the parathyroid hormonerelated peptide gene. Genes Dev. 8, 277 – 289 (1994). 75. D. B. Pateder, R. N. Rosier, E. M. Schwarz, P. R. Reynolds, J. E. Puzas, M. D’Souza, and R. J. O’Keefe, PTHrP expression in chondrocytes, regulation by TGF-beta, and interactions between epiphyseal and growth plate chondrocytes. Exp. Cell Res. 256, 555 – 562 (2000). 76. R. Serra, A. Karaplis, and P. Sohn, Parathyroid hormone-related peptide (PTHrP)-dependent and -independent effects of transforming growth factor beta (TGF-beta) on endochondral bone formation. J. Cell Biol. 145, 783 – 794 (1999). 77. L. K. McCauley, A. J. Koh, C. A. Beecher, Y. Cui, J. D. Decker, and R. T. Franceschi, Effects of differentiation and transforming growth factor beta 1 on PTH/PTHrP receptor mRNA levels in MC3T3-E1 cells. J. Bone Miner. Res. 10, 1243 – 1255 (1995). 78. B. Lanske, M. Amling, L. Neff, J. Guiducci, R. Baron, and H. M. Kronenberg, Ablation of the PTHrP gene or the PTH/PTHrP receptor gene leads to distinct abnormalities in bone development. J. Clin. Invest. 104, 399 – 407 (1999). 79. A. S. Jobert, P. Zhang, A. Couvineau, J. Bonaventure, J. Roume, M. Le Merrer, and C. Silve, Absence of functional receptors for parathyroid hormone and parathyroid hormone-related peptide in Blomstrand chondrodysplasia. J. Clin. Invest. 102, 34 – 40 (1998). 80. E. Schipani, B. Lanske, J. Hunzelman, A. Luz, C. S. Kovacs, K. Lee, A. Pirro, H. M. Kronenberg, and H. Juppner, Targeted expression of constitutively active receptors for parathyroid hormone and parathyroid hormone-related peptide delays endochondral bone formation and rescues mice that lack parathyroid hormone-related peptide. Proc. Natl. Acad. Sci. USA 94, 13689 – 13694 (1997). 81. B. Zerega, S. Cermelli, P. Bianco, R. Cancedda, and F. D. Cancedda, Parathyroid hormone [PTH(1-34)] and parathyroid hormone-related
LIAN AND STEIN
82.
83.
84.
85.
86.
87. 88.
89.
90.
91.
92.
93.
94.
95.
96.
protein [PTHrP(1-34)] promote reversion of hypertrophic chondrocytes to a prehypertrophic proliferating phenotype and prevent terminal differentiation of osteoblast-like cells. J. Bone Miner. Res. 14, 1281 – 1289 (1999). Z. F. Chen and R. R. Behringer, Twist is required in head mesenchyme for cranial neural tube morphogenesis. Genes Dev. 9, 686 – 699 (1995). P. Cserjesi, D. Brown, K. L. Ligon, G. E. Lyons, N. G. Copeland, D. J. Gilbert, N. A. Jenkins, and E. N. Olson, Scleraxis: A basic helix-loop-helix protein that prefigures skeletal formation during mouse embryogenesis. Development 121, 1099 – 1110 (1995). T. Ogata and M. Noda, Expression of Id, a negative regulator of helix-loop-helix DNA binding proteins, is down-regulated at confluence and enhanced by dexamethasone in a mouse osteoblastic cell line, MC3T3E1. Biochem. Biophys. Res. Commun. 180, 1194 – 1199 (1991). S. S. Murray, C. A. Glackin, K. A. Winters, D. Gazit, A. J. Kahn, and E. J. Murray, Expression of helix-loop-helix regulatory genes during differentiation of mouse osteoblastic cells. J. Bone Miner. Res. 7, 1131 – 1138 (1992). C. A. Glackin, M. Lee, G. Lowe, S. Morales, J. Wergedal, and D. Strong, Overexpressing human TWIST in high SaOS (HSaOS) cells results in major differences in cellular morphology, proliferation rates, and ALP activities: Knocking out human Id-2 greatly reduces the expression of all TWIST family members, indicating that Id-2 may regulate TWIST during osteoblast differentiation. J. Bone Miner. Res. 12, S154 – S154 (1997). [Abstract] W. J. Gehring, M. Affolter, and T. Burglin, Homeodomain proteins. Annu. Rev. Biochem. 63, 487 – 526 (1994). M. Mina, J. Gluhak, W. B. Upholt, E. J. Kollar, and B. Rogers, Experimental analysis of Msx-1 and Msx-2 gene expression during chick mandibular morphogenesis. Dev. Dyn. 202, 195 – 214 (1995). G. Q. Zhao, S. Zhao, X. Zhou, H. Eberspaecher, M. Solursh, and B. de Crombrugghe, rDlx, a novel distal-less-like homeoprotein is expressed in developing cartilages and discrete neuronal tissues. Dev. Biol. 164, 37 – 51 (1994). D. Ferrari, L. Sumoy, J. Gannon, H. Sun, A. M. Brown, W. B. Upholt, and R. A. Kosher, The expression pattern of the Distal-less homeobox-containing gene Dlx-5 in the developing chick limb bud suggests its involvement in apical ectodermal ridge activity, pattern formation, and cartilage differentiation. Mech. Dev. 52, 257 – 264 (1995). E. W. Jabs, U. Muller, X. Li, L. Ma, W. Luo, I. S. Haworth, I. Klisak, R. Sparkes, M. L. Warman, J. B. Mulliken, M. L. Snead, and R. Maxson, A mutation in the homeodomain of the human MSX2 gene in a family affected with autosomal dominant craniosynostosis. Cell 75, 443 – 450 (1993). M. Dodig, T. Tadic, M. S. Kronenberg, S. Dacic, Y. H. Liu, R. Maxson, D. W. Rowe, and A. C. Lichtler, Ectopic Msx2 overexpression inhibits and Msx2 antisense stimulates calvarial osteoblast differentiation. Dev. Biol. 209, 298–307 (1999). N. A. Speck and T. Stacy, A new transcription factor family associated with human leukemias. Crit. Rev. Eukary. Gene Expr. 5, 337 – 364 (1995). J. J. Westendorf and S. W. Hiebert, Mammalian runt-domain proteins and their roles in hematopoiesis, osteogenesis and leukemia. J. Cell. Boichem. (32 – 33) Suppl., 51 – 58 (1999). Y. Ito, Molecular basis of tissue-specific gene expression mediated by the runt domain transcription factor PEBP2/CBF. Genes Cells 4, 685 – 696 (1999). E. Ogawa, M. Maruyama, H. Kagoshima, M. Inuzuka, J. Lu, M. Satake, K. Shigesada, and Y. Ito, PEBP2/PEA2 represents a family of transcription factors homologous to the products of the Drosophila runt gene and the human AML1 gene. Proc. Natl. Acad. Sci. USA 90, 6859 – 6863 (1993).
53
CHAPTER 2 Osteoblast Biology 96a. M. Stewart, A. Terry, M. Hu, M. O’Hara, K. Blyth, E. Baxter, E. Cameron, D. E. Onions, and J. C. Neil, Proviral insertions induce the expression of bone-specific isoforms of PEBP2alphaA (CBFA1): Evidence for a new myc collaborating oncogene. Proc. Natl. Acad. Sci. USA 94, 8646 – 8651 (1997). 96b. K. Thirunavukkarasu, M. Mahajan, K. W. McLarren, S. Stifani, and G. Karsenty, Two domains unique to osteoblast-specific transcription factor Osf2/Cbfa1 contribute to its transactivation function and its inability to heterodimerize with Cbf beta, Mol. Cell Biol. 18, 4197–4208 (1998). 97. M. Niki, H. Okada, H. Takano, J. Kuno, K. Tani, H. Hibino, S. Asano, Y. Ito, M. Satake, and T. Noda, Hematopoiesis in the fetal liver is impaired by targeted mutagenesis of a gene encoding a non-DNA binding subunit of the transcription factor, polyomavirus enhancer binding protein 2/core binding factor. Proc. Natl. Acad. Sci. USA 94, 5697 – 5702 (1997). 98. T. Okuda, J. van Deursen, S. W. Hiebert, G. Grosveld, and J. R. Downing, AML1, the target of multiple chromosomal translocations in human leukemia, is essential for normal fetal liver hematopoiesis. Cell 84, 321 – 330 (1996). 99. H. L. Merriman, A. J. van Wijnen, S. Hiebert, J. P. Bidwell, E. Fey, J. Lian, J. Stein, and G. S. Stein, The tissue-specific nuclear matrix protein, NMP-2, is a member of the AML/CBF/PEBP2/runt domain transcription factor family: Interactions with the osteocalcin gene promoter. Biochemistry 34, 13125 – 13132 (1995). 100. C. Banerjee, S. W. Hiebert, J. L. Stein, J. B. Lian, and G. S. Stein, An AML-1 consensus sequence binds an osteoblast-specific complex and transcriptionally activates the osteocalcin gene. Proc. Natl. Acad. Sci. USA 93, 4968 – 4973 (1996). 101. P. Ducy and G. Karsenty, Two distinct osteoblast-specific cis-acting elements control expression of a mouse osteocalcin gene. Mol. Cell. Biol. 15, 1858 – 1869 (1995). 102. P. Ducy, R. Zhang, V. Geoffroy, A. L. Ridall, and G. Karsenty, Osf2/Cbfa1: A transcriptional activator of osteoblast differentiation. Cell 89, 747 – 754 (1997). 103. C. Banerjee, L. R. McCabe, J.-Y. Choi, S. W. Hiebert, J. L. Stein, G. S. Stein, and J. B. Lian, Runt homology domain proteins in osteoblast differentiation: AML-3/CBFA1 is a major component of a bone specific complex. J. Cell. Biochem. 66, 1 – 8 (1997). 104. D. O. Sillence, H. E. Ritchie, and P. B. Selby, Animal model: Skeletal anomalies in mice with cleidocranial dysplasia. Am. J. Med. Genet. 27, 75 – 85 (1987). 105. S. Mundlos, F. Otto, C. Mundlos, J. B. Mulliken, A. S. Aylsworth, S. Albright, D. Lindhout, W. G. Cole, W. Henn, J. H. M. Knoll, M. J. Owen, R. Mertelsmann, B. U. Zabel, and B. R. Olsen, Mutations involving the transcription factor CBFA1 cause cleidocranial dysplasia. Cell 89, 773 – 779 (1997). 106. S. Mundlos, Cleidocranial dysplasia: Clinical and molecular genetics. J. Med. Genet. 36, 177 – 182 (1999). 107. B. Lee, K. Thirunavukkarasu, L. Zhou, L. Pastore, A. Baldini, J. Hecht, V. Geoffroy, P. Ducy, and G. Karsenty, Missense mutations abolishing DNA binding of the osteoblast- specific transcription factor OSF2/CBFA1 in cleidocranial dysplasia. Nature Genet. 16, 307 – 310 (1997). 108. Y. Zhang, N. Yasui, N. Kakazu, T. Abe, K. Takada, S. Imai, M. Sato, S. Nomura, T. Ochi, S. Okuzumi, H. Nogami, T. Nagai, H. Ohashi, and Y. Ito, PEBP2alphaA/CBFA1 mutations in Japanese cleidocranial dysplasia patients. Gene 244, 21 – 28 (2000). 109. H. Enomoto, M. Enomoto-Iwamoto, M. Iwamoto, S. Nomura, M. Himeno, Y. Kitamura, T. Kishimoto, and T. Komori, Cbfa1 is a positive regulatory factor in chondrocyte maturation. J. Biol. Chem. 275, 8695 – 8702 (2000). 110. M. Inada, T. Yasui, S. Nomura, S. Miyake, K. Deguchi, M. Himeno, M. Sato, H. Yamagiwa, T. Kimura, N. Yasui, T. Ochi, N. Endo, Y. Kitamura, T. Kishimoto, and T. Komori, Maturational disturbance of
111. 112.
113.
114.
115.
116.
117.
118.
119.
120.
121.
122.
123.
124.
chondrocytes in Cbfa1-deficient mice. Dev. Dyn. 214, 279 – 290 (1999). I. S. Kim, F. Otto, B. Zabel, and S. Mundlos, Regulation of chondrocyte differentiation by Cbfa1. Mech. Dev. 80, 159 – 170 (1999). M. Satake, S. Nomura, Y. Yamaguchi-Iwai, Y. Takahama, Y. Hashimoto, M. Niki, Y. Kitamura, and Y. Ito, Expression of the Runt domain-encoding PEBP2 alpha genes in T cells during thymic development. Mol. Cell. Biol. 15, 1662 – 1670 (1995). S. Ogawa, H. Harada, M. Fujiwara, S. Tagashira, T. Katsumata, and H. Takada, Cbfa1, an essential transcription factor for bone formation, is expressed in testis from the same promoter used in bone. DNA Res. 7, 181 – 185 (2000). K. Tsuji, Y. Ito, and M. Noda, Expression of the PEBP2A/ AML3/CBFA1 gene is regulated by BMP4/7 heterodimer and its overexpression suppresses type I collagen and osteocalcin gene expression in osteoblastic and nonosteoblastic mesenchymal cells. Bone 22, 87 – 92 (1998). F. Gori, T. Thomas, K. C. Hicok, T. C. Spelsberg, and B. L. Riggs, Differentiation of human marrow stromal precursor cells: Bone morphogenetic protein-2 increases OSF2/CBFA1, enhances osteoblast commitment, and inhibits late adipocyte maturation. J. Bone Miner. Res. 14, 1522 – 1535 (1999). K. Satomura, P. Krebsbach, P. Bianco, and R. P. Gehron, Osteogenic imprinting upstream of marrow stromal cell differentiation. J. Cell. Biochem. 78, 391 – 403 (2000). P. Ducy, M. Starbuck, M. Priemel, J. Shen, G. Pinero, V. Geoffroy, M. Amling, and G. Karsenty, A Cbfa1-dependent genetic pathway controls bone formation beyond embryonic development. Genes Dev. 13, 1025 – 1036 (1999). H. Harada, S. Tagashira, M. Fujiwara, S. Ogawa, T. Katsumata, A. Yamaguchi, T. Komori, and M. Nakatsuka, Cbfa1 isoforms exert functional differences in osteoblast differentiation. J. Biol. Chem. 274, 6972 – 6978 (1999). M. Muenke and A. Wilkie, Craniosynostosis syndromes. In “The Metabolic and Molecular Bases of Inherited Disease” (C. R. Scriver, A. L. Beaudet, W. S. Sly, D. Valle, B. Childs, and B. Vogelstein, eds.). McGraw-Hill, New York, 2000. Y. X. Zhou, X. Xu, L. Chen, C. Li, S. G. Brodie, and C. X. Deng, A Pro250Arg substitution in mouse Fgfr1 causes increased expression of Cbfa1 and premature fusion of calvarial sutures. Hum. Mol. Genet. 9, 2001 – 2008 (2000). M. H. Lee, A. Javed, H. J. Kim, H. I. Shin, S. Gutierrez, J. Y. Choi, V. Rosen, J. L. Stein, A. J. van Wijnen, G. S. Stein, J. B. Lian, and H. M. Ryoo, Transient upregulation of CBFA1 in response to bone morphogenetic protein-2 and transforming growth factor 1 in C2C12 myogenic cells coincides with suppression of the myogenic phenotype but is not sufficient for osteoblast differentiation. J. Cell. Biochem. 73, 114 – 125 (1999). Y. Takazawa, K. Tsuji, A. Nifuji, H. Kurosawa, Y. Ito, and M. Noda, An osteogenesis-related transcription factor, core-binding factor A1, is constitutively expressed in the chondrocytic cell line TC6, and its expression is upregulated by bone morphogenetic protein-2. J. Endocrinol. 165, 579 – 586 (2000). J. Hanai, L. F. Chen, T. Kanno, N. Ohtani-Fujita, W. Y. Kim, W. H. Guo, T. Imamura, Y. Ishidou, M. Fukuchi, M. J. Shi, J. Stavnezer, M. Kawabata, K. Miyazono, and Y. Ito, Interaction and functional cooperation of PEBP2/CBF with Smads: Synergistic induction of the immunoglobulin germline Calpha promoter. J. Biol. Chem. 274, 31577 – 31582 (1999). K. S. Lee, H. J. Kim, Q. L. Li, X. Z. Chi, C. Ueta, T. Komori, J. M. Wozney, E. G. Kim, J. Y. Choi, H. M. Ryoo, and S. C. Bae, Runx2 is a common target of transforming growth factor beta1 and bone morphogenetic protein 2, and cooperation between runx2 and smad5 induces osteoblast-specific gene expression in the pluripotent mesenchymal precursor cell line C2C12. Mol. Cell Biol. 20, 8783–8792 (2000).
54 125. D. Wang, K. Christensen, K. Chawla, G. Xiao, P. H. Krebsbach, and R. T. Franceschi, Isolation and characterization of MC3T3-E1 preosteoblast subclones with distinct in vitro and in vivo differentiation/mineralization potential. J. Bone Miner. Res. 14, 893 – 903 (1999). 126. A. Vortkamp, S. Pathi, G. M. Peretti, E. M. Caruso, D. J. Zaleske, and C. J. Tabin, Recapitulation of signals regulating embryonic bone formation during postnatal growth and in fracture repair. Mech. Dev. 71, 65 – 76 (1998). 127. C. Ferguson, E. Alpern, T. Miclau, and J. A. Helms, Does adult fracture repair recapitulate embryonic skeletal formation? Mech. Dev. 87, 57 – 66 (1999). 128. G. S. Stein and J. B. Lian, Molecular mechanisms mediating developmental and hormone regulated expression of genes in osteoblasts: an integrated relationship of cell growth and differentiation. In “Cellular and Molecular Biology of Bone” (A. Noda, ed.), pp. 47 – 95. Academic Press, San Diego, 1993. 129. J. E. Aubin, K. Turksen, and J. N. M. Heersche, Osteoblastic cell lineage. In “Cellular and Molecular Biology of Bone” (M. Noda, ed.), pp. 1 – 45. Academic Press, San Diego, 1993. 130. S. C. Marks, Jr. and S. N. Popoff, Bone cell biology: The regulation of development, structure, and function in the skeleton. Am. J. Anat. 183, 1 – 44 (1988). 131. P. J. Nijweide, A. van der Plas, and A. A. Othof, Osteoblastic differentiation. In “Cell and Molecular Biology of Vertebrate Hard Tissue” (D. Evered and S. Harnett, eds.), pp. 61 – 72. Wiley, New York, 1988. 132. E. Bonucci, The histology, histochemistry and ultrastructure of bone. In “Bone Regulatory Factors” (A. Pecile and B. de Bernard, eds.), pp. 15 – 39. Plenum Press, New York, 1990. 133. M. Owen and A. J. Friedenstein, Stromal stem cells: Marrow-derived osteogenic precursors. In “Cell and Molecular Biology of Vertebrate Hard Tissues” (D. Evered and S. Harnett, eds.), pp. 42 – 60. Wiley, Chichester, England, 1988. 134. M. E. Holtrop, The ultrastructure of bone. Ann. Clin. Lab. Sci. 5, 264 – 271 (1975). 135. S. Nomura, A. J. Wills, D. R. Edwards, J. K. Heath, and B. L. Hogan, Developmental expression of 2ar (osteopontin) and SPARC (osteonectin) RNA as revealed by in situ hybridization. J. Cell Biol. 106, 441 – 450 (1988). 136. M. Weinreb, D. Shinar, and G. A. Rodan, Different pattern of alkaline phosphatase, osteopontin, and osteocalcin expression in developing rat bone visualized by an in situ hybridization. J. Bone Miner. Res. 5, 831 – 842 (1990). 137. J. N. M. Heersche, S. M. Reimers, J. L. Wrana, M. M. Y. Waye, and A. K. Gupta, Changes in expression of alpha 1 type 1 collagen and osteocalcin mRNA in osteoblasts and adontoblasts at different stages of maturity as shown by in situ hybridization. Proc. Finn. Dent. Soc. 88, 173 – 182 (1992). 138. M. Sandberg, H. Autio-Harmainen, and E. Vuorio, Localization of the expression of types I, III, and IV collagen, TGF-beta 1 and c-fos genes in developing human calvarial bones. Dev. Biol. 130, 324 – 334 (1988). 139. K. Lee, J. D. Deeds, S. Chiba, M. Un-No, A. T. Bond, and G. V. Segre, Parathyroid hormone induces sequential c-fos expression in bone cells in vivo: In situ localization of its receptor and c-fos messenger ribonucleic acids. Endocrinology 134, 441 – 450 (1994). 140. M. P. Mark, C. W. Prince, S. Gay, R. L. Austin, M. Bhown, R. D. Finkelman, and W. T. Butler, A comparative immunocytochemical study on the subcellular distributions of 44 kDa bone phosphoprotein and bone gamma- carboxyglutamic acid (Gla)- containing protein in osteoblasts. J. Bone Miner. Res. 2, 337 – 346 (1987). 141. M. D. McKee and A. Nanci, Ultrastructural, cytochemical and immunocytochemical studies of bone and its interfaces. Cells Mater. 3, 219 – 243 (1993). 142. C. S. Carlson, H. M. Tulli, M. J. Jayo, R. F. Loeser, R. P. Tracy, K. G. Mann, and M. R. Adams, Immunolocalization of noncollage-
LIAN AND STEIN
143.
144.
145.
146.
147. 148.
149.
150.
151.
152.
153.
154.
155.
156.
157.
158.
nous bone matrix proteins in lumbar vertebrae from intact and surgically menopausal cynomolgus monkeys. J. Bone Miner. Res. 8, 71 – 81 (1993). S. M. Pockwinse, L. G. Wilming, D. M. Conlon, G. S. Stein, and J. B. Lian, Expression of cell growth and bone specific genes at single cell resolution during development of bone tissue-like organization in primary osteoblast cultures. J. Cell. Biochem. 49, 310 – 323 (1992). F. Liu, L. Malaval, A. K. Gupta, and J. E. Aubin, Simultaneous detection of multiple bone-related mRNAs and protein expression during osteoblast differentiation: Polymerase chain reaction and immunocytochemical studies at the single cell level. Dev. Biol. 166, 220 – 234 (1994). C. Zeng, S. McNeil, S. Pockwinse, J. A. Nickerson, L. Shopland, J. B. Lawrence, S. Penman, S. W. Hiebert, J. B. Lian, A. J. van Wijnen, J. L. Stein, and G. S. Stein, Intranuclear targeting of AML/CBF regulatory factors to nuclear matrix-associated transcriptional domains. Proc. Natl. Acad. Sci. USA 95, 1585 – 1589 (1998). G. S. Stein, A. J. van Wijnen, J. L. Stein, J. B. Lian, S. Pockwinse, and S. McNeil, Interrelationships of nuclear structure and transcriptional control: Functional consequences of being in the right place at the right time. J. Cell Biochem. 70, 200 – 212 (1998). S. B. Doty and B. H. Schofield, Enzyme histochemistry of bone and cartilage cells. Prog. Histochem. Cytochem. 8, 1 – 38 (1976). T. Nakase, S. Nomura, H. Yoshikawa, J. Hashimoto, S. Hirota, Y. Kitamura, S. Oikawa, K. Ono, and K. Takaoka, Transient and localized expression of bone morphogenetic protein 4 messenger RNA during fracture healing. J. Bone Miner. Res. 9, 651 – 659 (1994). J. E. Aubin, Osteoprogenitor cell frequency in rat bone marrow stromal populations: Role for heterotypic cell-cell interactions in osteoblast differentiation. J. Cell Biochem. 72, 396 – 410 (1999). A. J. Friedenstein, R. K. Chailakhjan, and K. S. Lalykina, The development of fibroblast colonies in monolayer cultures of guinea- pig bone marrow and spleen cells. Cell Tissue Kinet. 3, 393 – 403 (1970). A. J. Friedenstein, N. V. Latzinik, Gorskaya YuF, E. A. Luria, and I. L. Moskvina, Bone marrow stromal colony formation requires stimulation by haemopoietic cells. Bone Miner. 18, 199 – 213 (1992). A. J. Friedenstein, R. K. Chailakhyan, and U. V. Gerasimov, Bone marrow osteogenic stem cells: In vitro cultivation and transplantation in diffusion chambers. Cell Tissue Kinet. 20, 263 – 272 (1987). Z. Hou, Q. Nguyen, B. Frenkel, S. K. Nilsson, M. Milne, A. J. van Wijnen, J. L. Stein, P. Quesenberry, J. B. Lian, and G. S. Stein, Osteoblast-specific gene expression after transplantation of marrow cells: Implications for skeletal gene therapy. Proc. Natl. Acad. Sci. USA 96, 7294 – 7299 (1999). P. H. Krebsbach, S. A. Kuznetsov, K. Satomura, R. V. Emmons, D. W. Rowe, and P. G. Robey, Bone formation in vivo: Comparison of osteogenesis by transplanted mouse and human marrow stromal fibroblasts. Transplantation 63, 1059 – 1069 (1997). S. D. Boden, L. Titus, G. Hair, Y. Liu, M. Viggeswarapu, M. S. Nanes, and C. Baranowski, Lumbar spine fusion by local gene therapy with a cDNA encoding a novel osteoinductive protein (LMP-1). Spine 23, 2486 – 2492 (1998). D. Gazit, G. Turgeman, P. Kelley, E. Wang, M. Jalenak, Y. Zilberman, and I. Moutsatsos, Engineered pluripotent mesenchymal cells integrate and differentiate in regenerating bone: a novel cell-mediated gene therapy. J. Gene Med. 1, 121 – 133 (1999). T. D. Alden, D. D. Pittman, G. R. Hankins, E. J. Beres, J. A. Engh, S. Das, S. B. Hudson, K. M. Kerns, D. F. Kallmes, and G. A. Helm, In vivo endochondral bone formation using a bone morphogenetic protein 2 adenoviral vector. Hum. Gene Ther. 10, 2245 – 2253 (1999). R. T. Franceschi, D. Wang, P. H. Krebsbach, and R. B. Rutherford, Gene therapy for bone formation: In vitro and in vivo osteogenic activity of an adenovirus expressing BMP7. J. Cell Biochem. 78, 476 – 486 (2000).
55
CHAPTER 2 Osteoblast Biology 159. S. P. Bruder and B. S. Fox, Tissue engineering of bone: Cell based strategies. Clin. Orthop. S68 – S83 (1999). 160. J. C. Marx, J. A. Allay, D. A. Persons, S. A. Nooner, P. W. Hargrove, P. F. Kelly, E. F. Vanin, and E. M. Horwitz, High-efficiency transduction and long-term gene expression with a murine stem cell retroviral vector encoding the green fluorescent protein in human marrow stromal cells. Hum. Gene Ther. 10, 1163 – 1173 (1999). 161. A. I. Caplan, D. Reuben, and S. E. Haynesworth, Cell-based tissue engineering therapies: the influence of whole body physiology. Adv. Drug Deliv. Rev. 33, 3 – 14 (1998). 162. K. D. Riew, N. M. Wright, S. Cheng, L. V. Avioli, and J. Lou, Induction of bone formation using a recombinant adenoviral vector carrying the human BMP-2 gene in a rabbit spinal fusion model. Calcif. Tissue Int. 63, 357 – 360 (1998). 163. J. A. Allay, J. E. Dennis, S. E. Haynesworth, M. K. Majumdar, D. W. Clapp, L. D. Shultz, A. I. Caplan, and S. L. Gerson, LacZ and interleukin-3 expression in vivo after retroviral transduction of marrowderived human osteogenic mesenchymal progenitors. Hum. Gene Ther. 8, 1417 – 1427 (1997). 164. D. J. Prockop, Marrow stromal cells as stem cells for nonhematopoietic tissues. Science 276, 71 – 74 (1997). 165. J. B. Lian, G. S. Stein, J. L. Stein, and A. J. van Wijnen, Marrow transplantation and targeted gene therapy to the skeleton. Clin. Orthop. Rel. Res. 3795, 5146–5155 (2000). 166. J. B. Lian and C. Niyibizi, Breakout Session 4: Gene therapy for genetic diseases of bone. Clin. Orthop. Rel. Res. 3795, 5159–5163 (2000). 167. J. E. Aubin, Bone stem cells. J. Cell Biochem. Suppl. 30 – 31, 73 – 82 (1998). 168. P. Bianco, M. Riminucci, S. Kuznetsov, and P. G. Robey, Multipotential cells in the bone marrow stroma: Regulation in the context of organ physiology. Crit. Rev. Eukaryot. Gene Expr. 9, 159 – 173 (1999). 169. M. Owen, Marrow stromal stem cells. J. Cell Sci. 10, 63 – 76 (1988). 170. M. F. Pittenger, A. M. Mackay, S. C. Beck, R. K. Jaiswal, R. Douglas, J. D. Mosca, M. A. Moorman, D. W. Simonetti, S. Craig, and D. R. Marshak, Multilineage potential of adult human mesenchymal stem cells. Science 284, 143 – 147 (1999). 171. S. E. Haynesworth, J. Goshima, V. M. Goldberg, and A. I. Caplan, Characterization of cells with osteogenic potential from human marrow. Bone 13, 81 – 88 (1992). 172. S. E. Haynesworth, M. A. Baber, and A. I. Caplan, Cell surface antigens on human marrow-derived mesenchymal cells are detected by monoclonal antibodies. Bone 13, 69 – 80 (1992). 173. A. Muraglia, R. Cancedda, and R. Quarto, Clonal mesenchymal progenitors from human bone marrow differentiate in vitro according to a hierarchical model. J. Cell Sci. 113(Pt. 7), 1161 – 1166 (2000). 174. Y. Negishi, A. Kudo, A. Obinata, K. Kawashima, H. Hirano, N. Yanai, M. Obinata, and H. Endo, Multipotency of a bone marrow stromal cell line, TBR31-2, established from ts-SV40 T antigen gene transgenic mice. Biochem. Biophys. Res. Commun. 268, 450 – 455 (2000). 175. B. Lecka-Czernik, I. Gubrij, E. J. Moerman, O. Kajkenova, D. A. Lipschitz, S. C. Manolagas, and R. L. Jilka, Inhibition of Osf2/Cbfal expression and terminal osteoblast differentiation by PPARgamma2. J. Cell Biochem. 74, 357 – 371 (1999). 176. H. A. Kodama, Y. Amagai, H. Koyama, and S. Kasai, Hormonal responsiveness of a preadipose cell line derived from newborn mouse calvaria. J. Cell. Physiol. 112, 83 – 88 (1982). 177. J. N. Beresford, J. H. Bennett, C. Devlin, P. S. Leboy, and M. E. Owen, Evidence for an inverse relationship between the differentiation of adipocytic and osteogenic cells in rat marrow stromal cell cultures. J. Cell Sci. 102, 341 – 351 (1992). 178. J. M. Bennett, D. Catovsky, M. T. Daniel, G. Flandrin, D. A. Galton, H. R. Gralnick, and C. Sultan, Proposal for the recognition of
179.
180.
181.
182.
183.
184.
185.
186.
187.
188.
189.
190.
191.
192.
193.
194.
195.
minimally differentiated acute myeloid leukaemia (AML-MO). Br. J Haematol. 78, 325 – 329 (1991). X. Ji, D. Chen, C. Xu, S. E. Harris, G. R. Mundy, and T. Yoneda, Patterns of gene expression associated with BMP-2-induced osteoblast and adipocyte differentiation of mesenchymal progenitor cell 3T3-F442A. J. Bone Miner. Metab. 18, 132 – 139 (2000). M. E. Nuttall, A. J. Patton, D. L. Olivera, D. P. Nadeau, and M. Gowen, Human trabecular bone cells are able to express both osteoblastic and adipocytic phenotype: Implications for osteopenic disorders. J. Bone Miner. Res. 13, 371 – 382 (1998). T. Thomas, F. Gori, S. Khosla, M. D. Jensen, B. Burguera, and B. L. Riggs, Leptin acts on human marrow stromal cells to enhance differentiation to osteoblasts and to inhibit differentiation to adipocytes. Endocrinology 140, 1630 – 1638 (1999). D. Chen, X. Ji, M. A. Harris, J. Q. Feng, G. Karsenty, A. J. Celeste, V. Rosen, G. R. Mundy, and S. E. Harris, Differential roles for bone morphogenetic protein (BMP) receptor type IB and IA in differentiation and specification of mesenchymal precursor cells to osteoblast and adipocyte lineages. J. Cell Biol. 142, 295 – 305 (1998). V. Rosen, J. Nove, J. J. Song, R. S. Thies, K. Cox, and J. M. Wozney, Responsiveness of clonal limb bud cell lines to bone morphogenetic protein 2 reveals a sequential relationship between cartilage and bone cell phenotypes. J. Bone Miner. Res. 9, 1759 – 1768 (1994). S. Nishida, N. Endo, H. Yamagiwa, T. Tanizawa, and H. E. Takahashi, Number of osteoprogenitor cells in human bone marrow markedly decreases after skeletal maturation. J. Bone Miner. Metab. 17, 171 – 177 (1999). D. A. Puleo, Dependence of mesenchymal cell responses on duration of exposure to bone morphogenetic protein-2 in vitro. J. Cell. Physiol 173, 93 – 101 (1997). M. E. Nuttall and J. M. Gimble, Is there a therapeutic opportunity to either prevent or treat osteopenic disorders by inhibiting marrow adipogenesis? Bone 27, 177 – 184 (2000). J. M. Gimble, C. Morgan, K. Kelly, X. Wu, V. Dandapani, C. S. Wang, and V. Rosen, Bone morphogenetic proteins inhibit adipocyte differentiation by bone marrow stromal cells. J. Cell Biochem. 58, 393 – 402 (1995). R. A. Ignotz and J. Massague, Type beta transforming growth factor controls the adipogenic differentiation of 3T3 fibroblasts. Proc. Natl.Acad. Sci.USA 82, 8530 – 8534 (1985). A. Chawla and M. A. Lazar, Peroxisome proliferator and retinoid signaling pathways co-regulate preadipocyte phenotype and survival. Proc. Natl. Acad. Sci. USA 91, 1786 – 1790 (1994). P. Ducy, M. Amling, S. Takeda, M. Priemel, A. F. Schilling, F. T. Beil, J. Shen, C. Vinson, J. M. Rueger, and G. Karsenty, Leptin inhibits bone formation through a hypothalamic relay: A central control of bone mass. Cell 100, 197 – 207 (2000). K. A. Kelly and J. M. Gimble, 1,25-Dihydroxy vitamin D3 inhibits adipocyte differentiation and gene expression in murine bone marrow stromal cell clones and primary cultures. Endocrinolgy 139, 2622 – 2628 (1998). R. K. Jaiswal, N. Jaiswal, S. P. Bruder, G. Mbalaviele, D. R. Marshak, and M. F. Pittenger, Adult human mesenchymal stem cell differentiation to the osteogenic or adipogenic lineage is regulated by mitogenactivated protein kinase. J. Biol. Chem. 275, 9645 – 9652 (2000). S. C. Manolagas, Birth and death of bone cells: Basic regulatory mechanisms and implications for the pathogenesis and treatment of osteoporosis. Endocr. Rev. 21, 115 – 137 (2000). S. P. Bruder, D. J. Fink, and A. I. Caplan, Mesenchymal stem cells in bone development, bone repair, and skeletal regeneration therapy. J. Cell. Biochem. 56, 283 – 294 (1994). S. P. Bruder, N. S. Ricalton, R. E. Boynton, T. J. Connolly, N. Jaiswal, J. Zaia, and F. P. Barry, Mesenchymal stem cell surface antigen SB-10 corresponds to activated leukocyte cell adhesion
56
196.
197.
198.
199.
200.
201. 202.
203.
204.
205.
206.
207.
208.
209.
210.
211.
212.
213.
214.
LIAN AND STEIN molecule and is involved in osteogenic differentiation. J. Bone Miner. Res. 13, 655 – 663 (1998). S. Gronthos, S. E. Graves, S. Ohta, and P. J. Simmons, The STRO-1 fraction of adult human bone marrow contains the osteogenic precursors. Blood 84, 4164 – 4173 (1994). A. Wetterwald, W. Hoffstetter, M. G. Cecchini, B. Lanske, C. Wagner, H. Fleisch, and M. Atkinson, Characterization and cloning of the E11 antigen, a marker expressed by rat osteoblasts and osteocytes. Bone 18, 125 – 132 (1996). C. J. Joyner, A. Bennett, and J. T. Triffitt, Identification and enrichment of human osteoprogenitor cells by using differentiation stagespecific monoclonal antibodies. Bone 21, 1 – 6 (1997). P. J. Nijweide and R. J. Mulder, Identification of osteocytes in osteoblast-like cell cultures using a monoclonal antibody specifically directed against osteocytes. Histochemistry 84, 342 – 347 (1986). D. E. Hughes, D. M. Salter, and R. Simpson, CD44 expression in human bone: a novel marker of osteocytic differentiation. J. Bone Miner. Res. 9, 39 – 44 (1994). H. H. Jamal and J. E. Aubin, CD44 expression in fetal rat bone: In vivo and in vitro analysis. Exp. Cell Res. 223, 467 – 477 (1996). K. Turksen and J. E. Aubin, Positive and negative immunoselection for enrichment of two classes of osteoprogenitor cells. J. Cell Biol. 114, 373 – 384 (1991). P. J. Simmons and B. Torok-Storb, Identification of stromal cell precursors in human bone marrow by a novel monoclonal antibody, STRO-1. Blood 78, 55 – 62 (1991). R. J. Byers, J. Brown, C. Brandwood, P. Wood, W. Staley, L. Hainey, A. J. Freemont, and J. A. Hoyland, Osteoblastic differentiation and mRNA analysis of STRO-1-positive human bone marrow stromal cells using primary in vitro culture and poly (A) PCR. J. Pathol. 187, 374 – 381 (1999). K. Stewart, S. Walsh, J. Screen, C. M. Jefferiss, J. Chainey, G. R. Jordan, and J. N. Beresford, Further characterization of cells expressing STRO-1 in cultures of adult human bone marrow stromal cells. J. Bone Miner. Res. 14, 1345 – 1356 (1999). B. O. Oyajobi, A. Lomri, M. Hott, and P. J. Marie, Isolation and characterization of human clonogenic osteoblast progenitors immunoselected from fetal bone marrow stroma using STRO-1 monoclonal antibody. J. Bone Miner. Res. 14, 351 – 361 (1999). M. E. Owen, J. Cave, and C. J. Joyner, Clonal analysis in vitro of osteogenic differentiation of marrow CFU-F. J. Cell. Sci. 87, 731 – 738 (1987). L. Malaval, A. K. Gupta, and J. E. Aubin, Leukemia inhibitory factor inhibits osteogenic differentiation in rat calvaria cell cultures. Endocrinology 136, 1411 – 1418 (1995). L. Malaval, A. K. Gupta, F. Liu, P. D. Delmas, and J. E. Aubin, LIF, but not IL-6, regulates osteoprogenitor differentiation in rat calvaria cell cultures: Modulation by dexamethasone. J. Bone Miner. Res. 13, 175 – 184 (1998). J. Cornish, K. Callon, A. King, S. Edgar, and I. R. Reid, The effect of leukemia inhibitory factor on bone in vivo. Endocrinology 132, 1359 – 1366 (1993). I. Martin, A. Muraglia, G. Campanile, R. Cancedda, and R. Quarto, Fibroblast growth factor-2 supports ex vivo expansion and maintenance of osteogenic precursors from human bone marrow. Endocrinology 138, 4456 – 4462 (1997). T. Nakamura, K. Hanada, M. Tamura, T. Shibanushi, H. Nigi, M. Tagawa, S. Fukumoto, and T. Matsumoto, Stimulation of endosteal bone formation by systemic injections of recombinant basic fibroblast growth factor in rats. Endocrinology 136, 1276 – 1284 (1995). M. Noda and J. J. Camilliere, In vivo stimulation of bone formation by transforming growth factor-beta. Endocrinology 124, 2991 – 2994 (1989). J. M. Hock, E. Canalis, and M. Centrella, Transforming growth factor-beta stimulates bone matrix apposition and bone cell
215.
216.
217.
218.
219.
220.
221.
222.
223.
224.
225.
226. 227. 228.
229.
230.
231.
replication in cultured fetal rat calvariae. Endocrinology 126, 421 – 426 (1990). T. Onishi, Y. Ishidou, T. Nagamine, K. Yone, T. Imamura, M. Kato, T. K. Sampath, P. T. Dijke, and T. Sakou, Distinct and overlapping patterns of localization of bone morphogenetic protein (BMP) family members and a BMP type II receptor during fracture healing in rats. Bone 22, 605 – 612 (1998). M. Komaki, T. Katagiri, and T. Suda, Bone morphogenetic protein-2 does not alter the differentiation pathway of committed progenitors of osteoblasts and chondroblasts. Cell Tissue Res. 284, 9 – 17 (1996). D. J. Rickard, T. A. Sullivan, B. J. Shenker, P. S. Leboy, and I. Kazhdan, Induction of rapid osteoblast differentiation in rat bone marrow stromal cell cultures by dexamethasone and BMP-2. Dev. Biol. 161, 218 – 228 (1994). A. Yamaguchi, T. Ishizuya, N. Kintou, Y. Wada, T. Katagiri, J. M. Wozney, V. Rosen, and S. Yoshiki, Effects of BMP-2, BMP-4, and BMP-6 on osteoblastic differentiation of bone marrow-derived stromal cell lines, ST2 and MC3T3-G2/PA6. Biochem. Biophys. Res. Commun. 220, 366 – 371 (1996). E. A. Wang, D. I. Israel, S. Kelly, and D. P. Luxenberg, Bone morphogenetic protein-2 causes commitment and differentiation in C3H10T1/2 and 3T3 cells. Growth Factors 9, 57 – 71 (1993). T. Katagiri, A. Yamaguchi, T. Ikeda, S. Yoshiki, J. M. Wozney, V. Rosen, E. A. Wang, H. Tanaka, S. Omura, and T. Suda, The non-osteogenic mouse pluripotent cell line, C3H10T1/2, is induced to differentiate into osteoblastic cells by recombinant human bone morphogenetic protein-2. Biochem Biophys Res. Commun. 172, 295 – 299 (1990). T. Katagiri, A. Yamaguchi, M. Komaki, E. Abe, N. Takahashi, T. Ikeda, V. Rosen, J. M. Wozney, A. Fujisawa-Sehara, and T. Suda, Bone morphogenetic protein-2 converts the differentiation pathway of C2C12 myoblasts into the osteoblast lineage. J. Cell Biol. 127, 1755 – 1766 (1994). P. V. Hauschka, A. E. Mavrakos, M. D. Iafrati, S. E. Doleman, and M. Klagsbrun, Growth factors in bone matrix. J. Biol. Chem. 261, 12665 – 12674 (1986). M. W. Long, J. A. Robinson, E. A. Ashcraft, and K. G. Mann, Regulation of human bone marrow-derived osteoprogenitor cells by osteogenic growth factors. J. Clin. Invest. 95, 881 – 887 (1995). L. Malaval, F. Liu, P. Roche, and J. E. Aubin, Kinetics of osteoprogenitor proliferation and osteoblast differentiation in vitro. J. Cell Biochem. 74, 616 – 627 (1999). E. A. Tonna and E. P. Cronkite, An autoradiographic study of periosteal cell proliferation with tritiated thymidine. Lab. Invest. 11, 455 – 461 (1967). B. L. Scott, Thymidine-3H electron microscope radioautography of osteogenic cells in the fetal rat. J. Cell Biol. 35, 115 – 126 (1967). M. Owen, Cell population kinetics of an osteogenic tissue. J. Cell Biol. 19, 19 – 32 (1963). S. C. Luk, C. Nopajaroonsri, and G. T. Simon, The ultrastructure of endosteum: A topographic study in young adult rabbits. J. Ultrastruct. Res. 46, 165 – 183 (1974). S. P. Bruder and A. I. Caplan, Terminal differentiation of osteogenic cells in the embryonic chick tibia is revealed by a monoclonal antibody against osteocytes. Bone 11, 189 – 198 (1990). E. F. Eriksen, S. F. Hodgson, R. Eastell, S. L. Cedel, W. M. O’Fallon, and B. L. Riggs, Cancellous bone remodeling in type I (postmenopausal) osteoporosis: Quantitative assessment of rates of formation, resorption, and bone loss at tissue and cellular levels. J. Bone Miner. Res. 5, 311 – 319 (1990). A. M. Parfitt, C. H. E. Mathews, A. R. Villanueva, and M. Kleerekoper, Relationships between surface, volume, and thickness of iliac trabecular bone in aging and osteoporosis. J. Clin. Invest. 72, 1396 – 1409 (1983).
57
CHAPTER 2 Osteoblast Biology 232. J. Glowacki, Influence of age on human marrow. Calcif. Tissue Int. 56, S50 – S51 (1995). 233. R. J. Bergman, D. Gazit, A. J. Kahn, H. Gruber, S. McDougall, and T. J. Hahn, Age-related changes in osteogenic stem cells in mice. J. Bone Miner. Res. 11, 568 – 577 (1996). 234. P. J. Roholl, E. Blauw, C. Zurcher, J. A. Dormans, and H. M. Theuns, Evidence for a diminished maturation of preosteoblasts into osteoblasts during aging in rats; an ultrastructural analysis. J. Bone Miner. Res. 9, 355 – 366 (1994). 235. G. D’Ippolito, P. C. Schiller, C. Ricordi, B. A. Roos, and G. A. Howard, Age-related osteogenic potential of mesenchymal stromal stem cells from human vertebral bone marrow. J. Bone Miner. Res. 14, 1115 – 1122 (1999). 236. W. E. Mercer, Checking on the cell cycle. J. Cell. Biochem. Suppl. 30 – 31, 50 – 54 (1998). 237. A. B. Pardee, G1 events and regulation of cell proliferation. Science 246, 603 – 608 (1989). 238. D. O. Morgan, Principles of CDK regulation. Nature 374, 131 – 134 (1995). 239. S. S. Millard and A. Koff, Cyclin-dependent kinase inhibitors in restriction point control, genomic stability, and tumorigenesis. J. Cell. Biochem. Suppl 30 – 31, 37 – 42 (1998). 240. G. S. Stein, T. A. Owen, A. J. van Wijnen, B. Cho, J. L. Stein, and J. B. Lian, Mechanisms of action of skeletal growth factors in osteoblasts. In “Skeletal Growth Factors” (E. Canalis, ed.), pp. 51 – 65. Lippincott Williams & Wilkins, Philadelphia, PA, 2000. 241. T. Hunter and J. Pines, Cyclins and cancer II: Cyclin D and cdk inhibitors come of age. Cell 79, 573 – 582 (1994). 242. T. K. MacLachlan, N. Sang, and A. Giordano, Cyclins, cyclin-dependent kinases and cdk inhibitors: Implications in cell cycle control and cancer. Crit. Rev. Eukaryot. Gene Expr. 5, 127 – 156 (1995). 243. L. H. Hartwell and T. A. Weinert, Checkpoints: Controls that ensure the order of cell cycle events. Science 246, 629 – 634 (1989). 244. P. Stiegler, M. Kasten, and A. Giordano, The RB family of cell cycle regulatory factors. J. Cell. Biochem. Suppl 30 – 31, 30 – 36 (1998). 245. C. Cadwell and G. P. Zambetti, Regulators and mediators of the p53 tumor suppressor. J. Cell. Biochem. Suppl 30 – 31, 43 – 49 (1998). 246. R. A. Weinberg, The retinoblastoma protein and cell cycle control. Cell 81, 323 – 330 (1995). 247. D. M. Koepp, J. W. Harper, and S. J. Elledge, How the cyclin became a cyclin: regulated proteolysis in the cell cycle. Cell 97, 431 – 434 (1999). 248. H. Kiyokawa, V. M. Richon, R. A. Rifkind, and P. A. Marks, Suppression of cyclin-dependent kinase 4 during induced differentiation of erythroleukemia cells. Mol. Cell. Biol. 14, 7195 – 7203 (1994). 249. E. Smith, B. Frenkel, R. Schlegel, A. Giordano, J. B. Lian, J. L. Stein, and G. S. Stein, Expression of cell cycle regulatory factors in differentiating osteoblasts: Postproliferative up-regulation of cyclins B and E. Cancer Res. 55, 5019 – 5024 (1995). 250. R. F. Morrison and S. R. Farmer, Role of PPARgamma in regulating a cascade expression of cyclin- dependent kinase inhibitors, p18(INK4c) and p21(Waf1/Cip1), during adipogenesis. J Biol. Chem. 274, 17088 – 17097 (1999). 251. E. G. Reynaud, K. Pelpel, M. Guillier, M. P. Leibovitch, and S. A. Leibovitch, p57(Kip2) stabilizes the MyoD protein by inhibiting cyclin E-Cdk2 kinase activity in growing myoblasts. Mol. Cell. Biol. 19, 7621 – 7629 (1999). 252. Q. Liu, R. W. VanHoy, J. H. Zhou, R. Dantzer, G. G. Freund, and K. W. Kelley, Elevated cyclin E levels, inactive retinoblastoma protein, and suppression of the p27(KIP1) inhibitor characterize early development of promyeloid cells into macrophages. Mol. Cell. Biol. 19, 6229 – 6239 (1999). 253. H. Drissi, D. Hushka, F. Aslam, Q. Nguyen, E. Buffone, A. Koff, A. van Wijnen, J. B. Lian, J. L. Stein, and G. S. Stein, The cell cycle
254.
255. 256.
257.
258.
259.
260.
261.
262.
263.
264.
265.
266.
267.
268.
269.
regulator p27kip 1 contributes to growth and differentiation of osteoblasts. Cancer Res. 59, 3705 – 3711 (1999). K. Yamato, S. Hashimoto, N. Okahashi, A. Ishisaki, K. Nonaka, T. Koseki, M. Kizaki, Y. Ikeda, and T. Nishihara, Dissociation of bone morphogenetic protein-mediated growth arrest and apoptosis of mouse B cells by HPV-16 E6/E7. Exp. Cell Res. 257, 198 – 205 (2000). K. Okamoto and C. Prives, A role of cyclin G in the process of apoptosis. Oncogene 18, 4606 – 4615 (1999). J. Jernvall, T. Aberg, P. Kettunen, S. Keranen, and I. Thesleff, The life history of an embryonic signaling center: BMP-4 induces p21 and is associated with apoptosis in the mouse tooth enamel knot. Development 125, 161 – 169 (1998). N. Ghosh-Choudhury, G. Ghosh-Choudhury, A. Celeste, P. M. Ghosh, M. Moyer, S. L. Abboud, and J. Kreisberg, Bone morphogenetic protein-2 induces cyclin kinase inhibitor p21 and hypophosphory lation of retinoblastoma protein in estradiol-treated MCF-7 human breast cancer cells. Biochim. Biophys. Acta 1497, 186 – 196 (2000). T. Bellido, C. A. O’Brien, P. K. Roberson, and S. C. Manolagas, Transcriptional activation of the p21(WAF1, CIP1, SD11) gene by interleukin-6 type cytokines: A prerequisite for their pro- differentiating and ant-apoptotic effects on human osteoblastic cells. J Biol. Chem. 273, 21137 – 21144 (1998). S. Walsh, C. Jefferiss, K. Stewart, G. R. Jordan, J. Screen, and J. N. Beresford, Expression of the developmental markers STRO-1 and alkaline phosphatase in cultures of human marrow stromal cells: Regulation by fibroblast growth factor (FGF)-2 and relationship to the expression of FGF receptors 1 – 4. Bone 27, 185 – 195 (2000). A. Mansukhani, P. Bellosta, M. Sahni, and C. Basilico, Signaling by fibroblast growth factors (FGF) and fibroblast growth factro receptor 2 (FGFR2)- activating mutations blocks mineralization and induces apoptosis in osteoblasts. J. Cell Biol. 149, 1297 – 1308 (2000). A. Montero, Y. Okada, M. Tomita, M. Ito, H. Tsurukami, T. Nakamura, T. Doestschman, J. D. Coffin, and M. M. Hurley, Disruption of the fibroblast growth factor-2 gene results in decreased bone mass and bone formation. J. Clin. Invest. 105, 1085 – 1093 (2000). H. Tanaka, H. Ogasa, J. Barnes, and C. T. Liang, Actions of bFGF on mitogenic activity and lineage expression in rat osteoprogenitor cells: Effect of age. Mol. Cell. Endocrinol. 150, 1 – 10 (1999). F. Debiais, M. Hott, A. M. Graulet, and P. J. Marie, The effects of fibroblast growth factor-2 on human neonatal calvaria osteoblastic cells are differentiation stage specific. J. Bone Miner. Res. 13, 645 – 654 (1998). M. R. Johnson, C. Valentine, C. Basilico, and A. Mansukhani, FGF signaling activates STAT1 and p21 and inhibits the estrogen response and proliferation of MCF-7 cells. Oncogene 16, 2647 – 2656 (1998). H. Wang, M. Rubin, E. Fenig, A. DeBlasio, J. Mendelsohn, J. Yahalom, and R. Wieder, Basic fibroblast growth factor causes growth arrest in MCF-7 human breast cancer cells while inducing both mitogenic and inhibitory G1 events. Cancer Res. 57, 1750 – 1757 (1997). K. Pardali, A. Kurisaki, A. Moren, P. T. Dijke, D. Kardassis, and A. Moustakas, Role of Smad proteins and transcription factor Sp1 in p21WAF1/Cip1 regulation by transforming growth factor-beta. J. Biol. Chem. 275, 29244 – 29256 (2000). M. B. Datto, Y. Yu, and X. F. Wang, Functional analysis of the transforming growth factor beta responsive elements in the WAF1/ Cip1/p21 promoter. J. Biol. Chem. 270, 28623 – 28628 (1995). M. Liu, M. H. Lee, M. Cohen, M. Bommakanti, and L. P. Freedman, Transcriptional activation of the Cdk inhibitor p21 by vitamin D3 leads to the induced differentiation of the myelomonocytic cell line U937. Genes Dev. 10, 142 – 153 (1996). J. W. Harper, G. R. Adami, N. Wei, K. Keyomarsi, and S. J. Elledge, The p21 Cdk-interacting protein Cip1 is a potent inhibitor of G1 cyclin- dependent kinases. Cell 75, 805 – 816 (1993).
58 270. Y. Xiong, G. J. Hannon, H. Zhang, D. Casso, R. Kobayashi, and D. Beach, p21 is a universal inhibitor of cyclin kinases. Nature 366, 701 – 704 (1993). 271. D. A. Cameron, The Golgi apparatus in bone and cartilage cells. Clin. Orthop. 58, 191 – 211 (1968). 272. C. F. Descalzi, C. Gentili, P. Manduca, and R. Cancedda, Hypertrophic chondrocytes undergo further differentiation in culture. J. Cell Biol. 117, 427 – 435 (1992). 273. J. B. Lian, M. D. McKee, A. M. Todd, and L. C. Gerstenfeld, Induction of bone-related proteins, osteocalcin and osteopontin, and their matrix ultrastructural localization with development of chondrocyte hypertrophy in vitro. J. Cell. Biochem. 52, 206 – 219 (1993). 274. P. S. Henthorn, M. Raducha, K. N. Fedde, M. A. Lafferty, and M. P. Whyte, Different missense mutations at the tissue-nonspecific alkaline phosphatase gene locus in autosomal recessively inherited forms of mild and severe hypophosphatasia. Proc. Natl. Acad. Sci. USA 89, 9924 – 9928 (1992). 275. A. L. Boskey, Matrix proteins and mineralization: an overview. Connect. Tissue Res. 35, 357 – 363 (1996). 276. A. L. Boskey, Biomineralization: Conflicts, challenges, and opportunities. J. Cell Biochem. Suppl 30 – 31, 83 – 91 (1998). 277. A. L. Boskey, S. Gadaleta, C. Gundberg, S. B. Doty, P. Ducy, and G. Karsenty, Fourier transformed infrared microspectroscopic analysis of bones of osteocalcin deficient mice provides insight into the function of osteocalcin. Bone 23, 187 (1998). 278. P. Ducy, C. Desbois, B. Boyce, G. Pinero, B. Story, C. Dunstan, E. Smith, J. Bonadio, S. Goldstein, C. Gundberg, A. Bradley, and G. Karsenty, Increased bone formation in osteocalcin-deficient mice. Nature 382, 448 – 452 (1996). 279. H. Yoshitake, S. R. Rittling, D. T. Denhardt, and M. Noda, Osteopontin-deficient mice are resistant to ovariectomy-induced bone resorption [published erratum appears in Proc. Natl. Acad. Sci. USA 96(19), 1094 (1999)]. Proc. Natl. Acad. Sci. USA 96, 8156 – 8160 (1999). 280. T. Xu, P. Bianco, L. W. Fisher, G. Longenecker, E. Smith, S. Goldstein, J. Bonadio, A. Boskey, A. M. Heegaard, B. Sommer, K. Satomura, P. Dominguez, C. Zhao, A. B. Kulkarni, P. G. Robey, and M. F. Young, Targeted disruption of the biglycan gene leads to an osteoporosis-like phenotype in mice. Nature Genet. 20, 78 – 82 (1998). 281. J. M. Hayden, S. Mohan, and D. J. Baylink, The insulin-like growth factor system and the coupling of formation to resorption. Bone 17, 93S – 98S (1995). 282. C. A. Conover, Insulin-like growth factors and the skeleton. In “Skeletal Growth Factors” (E. Canalis, ed.), pp. 101 – 116. Lippincott Williams & Wilkins, Philadelphia, PA, 2000. 283. D. M. Shinar, N. Endo, D. Halperin, G. A. Rodan, and M. Weinreb, Differential expression of insulin-like growth factor-I (IGF-I) and IFG- II messenger ribonucleic acid in growing rat bone. Endocrinology 132, 1158 – 1167 (1993). 284. T. Thomas, F. Gori, T. C. Spelsberg, S. Khosla, B. L. Riggs, and C. A. Conover, Response of bipotential human marrow stromal cells to insulin-like growth factors: Effect on binding protein production, proliferation, and commitment to osteoblasts and adipocytes. Endocrinology 140, 5036 – 5044 (1999). 285. J. M. Hock, M. Centrella, and E. Canalis, Insulin-like growth factor I has independent effects on bone matrix formation and cell replication. Endocrinology 122, 254 – 260 (1988). 286. V. Gangji, S. Rydziel, B. Gabbitas, and E. Canalis, Insulin-like growth factor II promoter expression in cultured rodent osteoblasts and adult rat bone. Endocrinology 139, 2287 – 2292 (1998). 287. E. Canalis, S. Rydziel, A. M. Delany, S. Varghese, and J. J. Jeffrey, Insulin-like growth factors inhibit interstitial collagenase synthesis in bone cell cultures. Endocrinology 136, 1348 – 1354 (1995). 288. R. C. Pereira and E. Canalis, Parathyroid hormone increases mac25/insulin-like growth factor-binding protein-related protein-1 expression in cultured osteoblasts. Endocrinology 140, 1998 – 2003 (1999).
LIAN AND STEIN 289. R. C. Pereira, F. Blanquaert, and E. Canalis, Cortisol enhances the expression of mac25/insulin-like growth factor- binding protein-related protein-1 in cultured osteoblasts. Endocrinology 140, 228 – 232 (1999). 290. A. M. Delany, Y. Dong, and E. Canalis, Mechanisms of glucocorticoid action in bone cells. J. Cell. Biochem. 56, 295 – 302 (1994). 291. V. Hwa, Y. Oh, and R. G. Rosenfeld, The insulin-like growth factorbinding protein (IGFBP) superfamily. Endocr. Rev. 20, 761 – 787 (1999). 292. C. Richman, D. J. Baylink, K. Lang, C. Dony, and S. Mohan, Recombinant human insulin-like growth factor-binding protein-5 stimulates bone formation parameters in vitro and in vivo. Endocrinology 140, 4699 – 4705 (1999). 293. B. Gabbitas and E. Canalis, Insulin-like growth factors sustain insulin-like growth factor-binding protein-5 expression in osteoblasts. Am. J. Physiol. 275, E222 – E228 (1998). 294. X. Qin, D. Byun, K. W. Lau, D. J. Baylink, and S. Mohan, Evidence that the interaction between insulin-like growth factor (IGF)- II and IGF binding protein (IGFBP)-4 is essential for the action of the IGFII-dependent IGFBP-4 protease. Arch. Biochem. Biophys. 379, 209 – 216 (2000). 295. Y. Chen, H. Shu, C. Ji, S. Casinghino, K. Kim, C. M. Gundberg, M. Centrella, and T. L. McCarthy, Insulin-like growth factor binding proteins localize to discrete cell culture compartments in periosteal and osteoblast cultures from fetal rat bone. J. Cell. Biochem. 71, 351 – 362 (1998). 296. P. Gamero, E. Sornay-Rendu, and P. D. Delmas, Low serum IGF-1 and occurrence of osteoporotic fractures in postmenopausal women. Lancet 355, 898 – 899 (2000). 297. P. Lakatos, J. Foldes, Z. Nagy, I. Takacs, G. Speer, C. Horvath, S. Mohan, D. J. Baylink, and P. H. Stern, Serum insulin-like growth factor-I, insulin-like growth factor binding proteins, and bone mineral content in hyperthyroidism. Thyroid 10, 417 – 423 (2000). 298. J. G. Kim, C. S. Shin, Y. M. Choi, S. Y. Moon, S. Y. Kim, and J. Y. Lee, The relationship among circulating insulin-like growth factor components, biochemical markers of bone turnover and bone mineral density in postmenopausal women under the age of 60. Clin. Endocrinol. (Oxford) 51, 301 – 307 (1999). 299. C. J. Rosen, Clinical aspects of the insulin-like growth factors. In “Skeletal Growth Factors” (E. Canalis, ed.), pp. 117 – 128. Lippincott Williams & Wilkins, Philadelphia, PA, 2000. 300. J. C. Adams and F. M. Watt, Regulation of development and differentiation by the extracellular matrix. Development 117, 1183 –1198 (1993). 301. C. H. Damsky, Extracellular matrix-integrin interactions in osteoblast function and tissue remodeling. Bone 25, 95 – 96 (1999). 302. M. A. Horton Interactions of connective tissue cells with the extracellular matrix. Bone 17, 51S – 53S (1995). 303. W. J. Grzesik and P. G. Robey, Bone matrix RGD glycoproteins: Immunolocalization and interaction with human primary osteoblastic bone cells in vitro. J. Bone Miner. Res. 9, 487 – 496 (1994). 304. R. J. Majeska, M. Port, and T. A. Einhorn, Attachment to extracellular matrix molecules by cells differing in the expression of osteoblastic traits. J. Bone Miner. Res. 8, 277 – 289 (1993). 305. M. P. Lynch, J. L. Stein, G. S. Stein, and J. B. Lian, The influence of type I collagen on the development and maintenance of the osteoblast phenotype in primary and passaged rat calvarial osteoblasts: Modification of expression of genes supporting cell growth, adhesion, and extracellular matrix mineralization. Exp. Cell Res. 216, 35 – 45 (1995). 306. S. Vukicevic, F. P. Luyten, H. K. Kleinman, and A. H. Reddi, Differentiation of canalicular cell processes in bone cells by basement membrane matrix components: Regulation by discrete domains of laminin. Cell 63, 437 – 445 (1990). 307. A. G. Andrianarivo, J. A. Robinson, K. G. Mann, and R. P. Tracy, Growth on type I collagen promotes expression of the osteoblastic
CHAPTER 2 Osteoblast Biology
308.
309.
310.
311.
312.
313.
314.
315.
316.
317.
318. 319.
320.
321.
322.
323.
324.
325.
phenotype in human osteosarcoma MG-63 cells. J Cell Physiol. 153, 256 – 265 (1992). R. T. Franceschi and B. S. Iyer, Relationship between collagen synthesis and expression of the osteoblast phenotype in MC3T3-E1 cells. J. Bone Miner. Res. 7, 235 – 246 (1992). S. Gronthos K. Stewart, S. E. Graves, S. Hay, and P. J. Simmons, Integrin expression and function on human osteoblast-like cells. J. Bone Miner. Res. 12, 1189 – 1197 (1997). A. M. Moursi, R. K. Globus, and C. H. Damsky, Interactions between integrin receptors and fibronectin are required for calvarial osteoblast differentiation in vitro. J. Cell Sci. 110(Pt 18), 2187 – 2196 (1997). A. M. Moursi, C. H. Damsky, J. Lull, D. Zimmerman, S. B. Doty, S. Aota, and R. K. Globus, Fibronectin regulates calvarial osteoblast differentiation. J. Cell Sci. 109(Pt 6), 1369 – 1380 (1996). R. K. Globus, S. B. Doty, J. C. Lull, E. Holmuhamedov, M. J. Humphries, and C. H. Damsky, Fibronectin is a survival factor for differentiated osteoblasts. J. Cell Sci. 111, 1385 – 1393. G. A. Gronowicz and M. E. Derome, Synthetic peptide containing Arg-Gly-Asp inhibits bone formation and resorption in a mineralizing organ culture system of fetal rat parietal bones. J. Bone Miner. Res. 9, 193 – 201 (1994). M. Mizuno, R. Fujisawa, and Y. Kuboki, Type I collagen-induced osteoblastic differentiation of bone-marrow cells mediated by collagen-alpha2beta1 integrin interaction. J. Cell. Physiol. 184, 207 – 213 (2000). M. Mizuno, T. Imai, R. Fujisawa, H. Tani, and Y. Kuboki, Bone sialoprotein (BSP) is a crucial factor for the expression of osteoblastic phenotypes of bone marrow cells cultured on type I collagen matrix. Calcif. Tissue Int. 66, 388 – 396 (2000). G. B. Schneider, S. W. Whitson, and L. F. Cooper, Restricted and coordinated expression of beta3-integrin and bone sialoprotein during cultured osteoblast differentiation. Bone 24, 321 – 327 (1999). T. V. Byzova, W. Kim, R. J. Midura, and E. F. Plow, Activation of integrin alpha(V)beta(3) regulates cell adhesion and migration to bone sialoprotein. Exp. Cell Res. 254, 299 – 308 (2000). P. Roche, H. A. Goldberg, P. D. Delmas, and L. Malaval, Selective attachment of osteoprogenitors to laminin. Bone 24, 329 – 336 (1999). A. Jikko, S. E. Harris, D. Chen, D. L. Mendrick, and C. H. Damsky, Collagen integrin receptors regulate early osteoblast differentiation induced by BMP-2. J. Bone Miner. Res. 14, 1075 – 1083 (1999). T. H. Steinberg, R. Civitelli, S. T. Geist, A. J. Robertson, E. Hick, R. D. Veenstra, H. Z. Wang, P. M. Warlow, E. M. Westphale, and J. G. Laing, Connexin43 and connexin45 form gap junctions with different molecular permeabilities in osteoblastic cells. EMBO J. 13, 744 – 750 (1994). D. T. Yamaguchi, D. Ma, A. Lee, J. Huang, and H. E. Gruber, Isolation and characterization of gap junctions in the osteoblastic MC3T3-E1 cell line. J. Bone Miner. Res. 9, 791 – 803 (1994). S. L. Ferrari, K. Traianedes, M. Thorne, M. H. Lafage-Proust, P. Genever, M. G. Cecchini, V. Behar, A. Bisello, M. Chorev, M. Rosenblatt, and L. J. Suva, A role for N-cadherin in the development of the differentiated osteoblastic phenotype. J. Bone Miner. Res. 15, 198 – 208 (2000). Y. S. Lee and C. M. Chuong, Adhesion molecules in skeletogenesis. I. Transient expression of neural cell adhesion molecules (NCAM) in osteoblasts during endochondral and intramembranous ossification. J. Bone Miner. Res. 7, 1435 – 1446 (1992). M. Okazaki, S. Takeshita, S. Kawai, R. Kikuno, A. Tsujimura, A. Kudo, and E. Amann, Molecular cloning and characterization of OB-cadherin, a new member of cadherin family expressed in osteoblasts. J Biol. Chem. 269, 12092 – 12098 (1994). C. S. Shin, F. Lecanda, S. Sheikh, L. Weitzmann, S. L. Cheng, and R. Civitelli, Relative abundance of different cadherins defines differentiation of mesenchymal precursors into osteogenic, myogenic, or adipogenic pathways. J. Cell Biochem. 78, 566 – 577 (2000).
59 326. Y. Tanaka, I. Morimoto, Y. Nakano, Y. Okada, S. Hirota, S. Nomura, T. Nakamura, and S. Eto, Osteoblasts are regulated by the cellular adhesion through ICAM-1 and VCAM-1, J. Bone Miner. Res. 10, 1462 – 1469 (1995). 327. H. Nakamura, S. Kenmotsu, H. Sakai, and H. Ozawa, Localization of CD44, the hyaluronate receptor, on the plasma membrane of osteocytes and osteoclasts in rat tibiae. Cell Tissue Res. 280, 225 – 233 (1995). 328. R. W. Young, Cell proliferation and specialization during endochondral osteogenesis in young rats. J. Cell Biol. 14, 357 (1962). 329. D. N. Menton, D. J. Simmons, S. L. Chang, and B. Y. Orr, From bone lining cell to osteocyte—an SEM study. Anat. Rec. 209, 29 – 39 (1984). 330. J. E. Puzas, The osteoblast. In “Primer on the Metabolic Bone Diseases and Disorders of Mineral Metabolism” (M. J. Favus, ed.), pp. 15 – 21. Raven Press, New York, 1993. 331. E. M. Aarden, E. H. Burger, and P. J. Nijweide, Function of osteocytes in bone. J. Cell. Biochem. 55, 287 – 299 (1994). 332. A. van der Plas and P. J. Nijweide, Isolation and purification of osteocytes. J. Bone Miner. Res. 7, 389 – 396 (1992). 333. S. C. Miller and W. S. Jee, The bone lining cell: A distinct phenotype? Calcif. Tissue Int. 41, 1 – 5 (1987). 334. S. B. Doty, Morphological evidence of gap junctions between bone cells. Calcif. Tissue Int. 33, 509 – 512 (1981). 335. S. C. Rawlinson, J. R. Mosley, R. F. Suswillo, A. A. Pitsillides, and L. E. Lanyon, Calvarial and limb bone cells in organ and monolayer culture do not show the same early responses to dynamic mechanical strain. J. Bone Miner. Res. 10, 1225 – 1232 (1995). 336. R. Civitelli, E. C. Beyer, P. M. Warlow, A. J. Robertson, S. T. Geist, and T. H. Steinberg, Connexin43 mediates direct intercellular communication in human osteoblastic cell networks. J. Clin. Invest. 91, 1888 – 1896 (1993). 337. H. J. Donahue, K. J. McLeod, C. T. Rubin, J. Andersen, E. A. Grine, E. L. Hertzberg, and P. R. Brink, Cell-to-cell communication in osteoblastic networks: Cell line- dependent hormonal regulation of gap junction function. J. Bone Miner. Res. 10, 881 – 889 (1995). 338. C. T. Rubin and L. E. Lanyon, Osteoregulatory nature of mechanical stimuli: Function as a determinant for adaptive bone remodeling. J. Orthop. Res. 5, 300 – 310 (1987). 339. S. Weinbaum, S. C. Cowin, and Y. Zeng, A model for the excitation of osteocytes by mechanical loading- induced bone fluid shear stresses. J. Biomech. 27, 339 – 360 (1994). 340. T. A. Einhorn, Biomechanics of bone. In “Principles of Bone Biology” (J. P. Bilezikian, L. G. Raisz, and G. A. Rodan, eds.), pp. 25 – 38. Academic Press, San Diego, 1996. 341. K. J. McLeod, C. T. Rubin, M. W. Otter, and Y. X. Qin, Skeletal cell stresses and bone adaptation. Am. J. Med. Sci. 316, 176 – 183 (1998). 342. J. Klein-Nulend, A. van der Plas, C. M. Semeins, N. E. Ajubi, J. A. Frangos, P. J. Nijweide, and E. H. Burger, Sensitivity of osteocytes to biomechanical stress in vitro. FASEB J. 9, 441 – 445 (1995). 343. C. H. Turner, M. R. Forwood, and M. W. Otter, Mechanotransduction in bone: Do bone cells act as sensors of fluid flow? FASEB J. 8, 875 – 878 (1994). 344. A. J. el Haj, S. L. Minter, S. C. Rawlinson, R. Suswillo, and L. E. Lanyon, Cellular responses to mechanical loading in vitro. J. Bone Miner. Res. 5, 923 – 932 (1990). 345. T. M. Skerry, L. Bitensky, J. Chayen, and L. E. Lanyon, Early strainrelated changes in enzyme activity in osteocytes following bone loading in vivo. J. Bone Miner. Res. 4, 783 – 788 (1989). 346. S. Nomura and T. Takano-Yamamoto, Molecular events caused by mechanical stress in bone. Matrix Biol. 19, 91 – 96 (2000). 347. J. M. Lean, C. J. Jagger, T. J. Chambers, and J. W. Chow, Increased insulin-like growth factor I mRNA expression in rat osteocytes in response to mechanical stimulation. Am. J. Physiol. 268, E318 – E327 (1995).
60 348. C. D. Toma, S. Ashkar, M. L. Gray, J. L. Schaffer, and L. C. Gerstenfeld, Signal transduction of mechanical stimuli is dependent on microfilament integrity: Identification of osteopontin as a mechanically induced gene in osteoblasts. J. Bone Miner. Res. 12, 1626 – 1636 (1997). 349. A. Miyauchi, K. Notoya, Y. Mikuni-Takagaki, Y. Takagi, M. Goto, Y. Miki, T. Takano-Yamamoto, K. Jinnai, K. Takahashi, M. Kumegawa, K. Chihara, and T. Fujita, Parathyroid hormone-activated volumesensitive calcium influx pathways in mechanically loaded osteocytes. J Biol. Chem. 275, 3335 – 3342 (2000). 350. T. N. McAllister, T. Du, and J. A. Frangos, Fluid shear stress stimulates prostaglandin and nitric oxide release in bone marrow-derived preosteoclast-like cells. Biochem. Biophys. Res. Commun. 270, 643 – 648 (2000). 351. S. W. Fox, T. J. Chambers, and J. W. M. Chow, Nitric oxide is an early mediator of the induction of bone formation by mechanical stimulation. J. Bone Miner. Res. 10, S201 – S201 (1995). [Abstract] 352. E. H. Burger, J. Klein-Nulend, C. M. Semeins, N. E. Ajubi, and P. J. Nijweide, Osteocytes but not periosteal fibroblasts produce nitric oxide (NO) in response to pulsatile fluid flow. Trans. Orthop. Res. Soc. 21, 531 – 531 (1996). [Abstract] 353. N. Wang, J. P. Butler, and D. E. Ingber, Mechanotransduction across the cell surface and through the cytoskeleton. Science 260, 1124 – 1127 (1993). 354. D. M. Salter, J. E. Robb, and M. O. Wright, Electrophysiological responses of human bone cells to mechanical stimulation: Evidence for specific integrin function in mechanotransduction. J. Bone Miner. Res. 12, 1133 – 1141 (1997). 355. A. L. Hitt and E. J. Luna, Membrane interactions with the actin cytoskeleton. Curr. Opin. Cell Biol. 6, 120 – 130 (1994). 356. F. M. Pavalko, N. X. Chen, C. H. Turner, D. B. Burr, S. Atkinson, Y. F. Hsieh, J. Qiu, and R. L. Duncan, Fluid shear-induced mechanical signaling in MC3T3-E1 osteoblasts requires cytoskeletonintegrin interactions. Am. J Physiol 275, C1591 – C1601 (1998). 357. M. C. Meazzini, C. D. Toma, J. L. Schaffer, M. L. Gray, and L. C. Gerstenfeld, Osteoblast cytoskeletal modulation in response to mechanical strain in vitro. J. Orthop. Res. 16, 170 – 180 (1998). 358. K. Terai, T. Takano-Yamamoto, Y. Ohba, K. Hiura, M. Sugimoto, M. Sato, H. Kawahata, N. Inaguma, Y. Kitamura, and S. Nomura, Role of osteopontin in bone remodeling caused by mechanical stress. J. Bone Miner. Res. 14, 839 – 849 (1999). 359. R. R. Miles, C. H. Turner, R. Santerre, Y. Tu, P. McClelland, J. Argot, B. S. DeHoff, C. W. Mundy, P. R. Rosteck, Jr., J. Bidwell, J. P. Sluka, J. Hock, and J. E. Onyia, Analysis of differential gene expression in rat tibia after an osteogenic stimulus in vivo: Mechanical loading regulates osteopontin and myeloperoxidase. J. Cell Biochem. 68, 355 – 365 (1998). 360. J. Rubin, T. Murphy, M. S. Nanes, and X. Fan, Mechanical strain inhibits expression of osteoclast differentiation factor by murine stromal cells. Am. J. Physiol. Cell. Physiol. 278, C1126 – C1132 (2000). 361. H. Drissi, A. Lomri, F. Lasmoles, X. Holy, E. Zerath, and P. J. Marie, Skeletal unloading induces biphasic changes in insulin-like growth factor-I mRNA levels and osteoblast activity. Exp. Cell Res. 251, 275 – 284 (1999). 362. M. Sato, T. Ochi, T. Nakase, S. Hirota, Y. Kitamura, S. Nomura, and N. Yasui, Mechanical tension-stress induces expression of bone morphogenetic protein (BMP)-2 and BMP-4, but not BMP-6, BMP-7, and GDF-5 mRNA, during distraction osteogenesis. J. Bone Miner. Res. 14, 1084 – 1095 (1999). 363. T. S. Gross, J. L. Edwards, K. J. McLeod, and C. T. Rubin, Strain gradients correlate with sites of periosteal bone formation. J. Bone Miner. Res. 12, 982 – 988 (1997). 364. R. L. Jilka, R. S. Weinstein, T. Bellido, A. M. Parfitt, and S. C. Manolagas, Osteoblast programmed cell death (apoptosis):
LIAN AND STEIN
365.
366. 367. 368.
369.
370.
371.
372.
373.
374.
375.
376.
377.
378.
379.
380.
381.
382. 383.
modulation by growth factors and cytokines. J. Bone Miner. Res. 13, 793 – 802 (1998). M. Lynch, C. Capparelli, J. L. Stein, G. S. Stein, and J. B. Lian, Apoptosis during bone-like tissue development in vitro. J. Cell. Biochem. 68, 31 – 49 (1998). D. P. Rice, H. J. Kim, and I. Thesleff, Apoptosis in murine calvarial bone and suture development. Eur. J Oral Sci. 107, 265 – 275 (1999). D. E. Hughes and B. F. Boyce, Apoptosis in bone physiology and disease. Mol. Pathol. 50, 132 – 137 (1997). O. Verborgt, G. J. Gibson, and M. B. Schaffler, Loss of osteocyte integrity in association with microdamage and bone remodeling after fatigue in vivo. J. Bone Miner. Res. 15, 60 – 67 (2000). A. L. J. J. Bronckers, W. Goei, G. Luo, G. Karsenty, R. N. D’Souza, D. M. Lyaruu, and E. H. Burger, DNA fragmentation during bone formation in neonatal rodents assessed by transferase-mediated end labeling. J. Bone Miner. Res. 11, 1281 – 1291 (1996). B. S. Noble, H. Stevens, N. Loveridge, and J. Reeve, Identification of apoptotic changes in osteocytes in normal and pathological human bone. Bone 20, 273 – 282 (1997). R. S. Weinstein, R. L. Jilka, A. M. Parfitt, and S. C. Manolagas, Inhibition of osteoblastogenesis and promotion of apoptosis of osteoblasts and osteocytes by glucocorticoids: Potential mechanisms of their deleterious effects on bone. J. Clin. Invest. 102, 274 – 282 (1998). A. Tomkinson, J. Reeve, R. W. Shaw, and B. S. Noble, The death of osteocytes via apoptosis accompanies estrogen withdrawal in human bone. J. Clin. Endocrinol. Metab. 82, 3128 – 3135 (1997). R. L. Jilka, R. S. Weinstein, T. Bellido, P. Roberson, A. M. Parfitt, and S. C. Manolagas, Increased bone formation by prevention of osteoblast apoptosis with parathyroid hormone. J. Clin. Invest. 104, 439 – 446 (1999). L. I. Plotkin, R. S. Weinstein, A. M. Parfitt, P. K. Roberson, S. C. Manolagas, and T. Bellido, Prevention of osteocyte and osteoblast apoptosis by bisphosphonates and calcitonin. J. Clin. Invest. 104, 1363 – 1374 (1999). A. Tomkinson, E. F. Gevers, J. M. Wit, J. Reeve, and B. S. Noble, The role of estrogen in the control of rat osteocyte apoptosis. J. Bone Miner. Res. 13, 1243 – 1250 (1998). D. G. Monroe, D. F. Jin, and M. M. Sanders, Estrogen opposes the apoptotic effects of bone morphogenetic protein 7 on tissue remodeling. Mol. Cell. Biol. 20, 4626 – 4634 (2000). T. Bellido, M. Huening, M. Raval-Pandya, S. C. Manolagas, and S. Christakos, Calbindin-D28K is expressed in osteoblastic cells and suppresses their apoptosis by inhibiting caspase-3 activity. J Biol. Chem. 275, 26328 – 26332 (2000). W. J. Sharrock, Bone and the hematopoietic and immune systems: A report of the proceedings of a scientific workshop. J. Bone Miner. Res. 13, 537 – 543 (1998). R. S. Taichman and S. G. Emerson, Human osteoblasts support hematopoiesis through the production of granulocyte colonystimulating factor. J. Exp. Med. 179, 1677 – 1682 (1994). H. A. Kodama, Y. Amagai, H. Koyama, and S. Kasai, A new preadipose cell line derived from newborn mouse calvaria can promote the proliferation of pluripotent hemopoietic stem cells in vitro. J. Cell. Physiol. 112, 89 – 95 (1982). P. J. Simmons, B. Masinovsky, B. M. Longenecker, R. Berenson, B. Torok-Storb, and W. M. Gallatin, Vascular cell adhesion molecule-1 expressed by bone marrow stromal cells mediates the binding of hematopoietic progenitor cells. Blood 80, 388 – 395 (1992). R. S. Taichman and S. G. Emerson, The role of osteoblasts in the hematopoietic microenvironment. Stem Cells 16, 7 – 15 (1998). G. Fiorelli, C. Orlando, S. Benvenuti, F. Franceschelli, S. Bianchi, P. Pioli, A. Tanini, M. Serio, F. Bartucci, and M. L. Brandi, Characterization, regulation, and function of specific cell membrane receptors for insulin-like growth factor I on bone endothelial cells. J. Bone Miner. Res. 9, 329 – 337 (1994).
CHAPTER 2 Osteoblast Biology 384. M. M. Deckers, M. Karperien, B. C. van der, T. Yamashita, S. E. Papapoulos, and C. W. Lowik, Expression of vascular endothelial growth factors and their receptors during osteoblast differentiation. Endocrinology 141, 1667 – 1674 (2000). 385. D. S. Wang, M. Miura, H. Demura, and K. Sato, Anabolic effects of 1,25-dihydroxyvitamin D3 on osteoblasts are enhanced by vascular endothelial growth factor produced by osteoblasts and by growth factors produced by endothelial cells. Endocrinology 138, 2953 – 2962 (1997). 386. H. P. Gerber, T. H. Vu, A. M. Ryan, J. Kowalski, Z. Werb, and N. Ferrara, VEGF couples hypertrophic cartilage remodeling, ossification and angiogenesis during endochondral bone formation. Nature Med. 5, 623 – 628 (1999). 387. A. Bellahcene, K. Bonjean, B. Fohr, N. S. Fedarko, F. A. Robey, M. F. Young, L. W. Fisher, and V. Castronovo, Bone sialoprotein mediates human endothelial cell attachment and migration and promotes angiogenesis. Circ. Res. 86, 885 – 891 (2000). 388. T. Suda, I. Nakamura, E. Jimi, and N. Takahashi, Regulation of osteoclast function. J. Bone Miner. Res. 12, 869 – 879 (1997). 389. G. D. Roodman, Advances in bone biology: The osteoclast. Endocr. Rev. 17, 308 – 332 (1996). 390. A. van der Plas, E. M. Aarden, J. H. Feijen, A. H. de Boer, A. Wiltink, M. J. Alblas, L. de Leij, and P. J. Nijweide, Characteristics and properties of osteocytes in culture. J. Bone Miner. Res. 9, 1697 – 1704 (1994). 391. G. Boivin, P. Mesguish, J. W. Pike, R. Bouillon, P. J. Meunier, M. R. Haussler, P. M. Dubois, and G. Morel, Ultrastructural immunolocalization of endogenous 1,25-dihydroxyvitamin D3 and its receptor in osteoblasts and osteocytes from neonatal mouse and rat calvaria. Bone Miner. 3, 125 – 136 (1987). 392. E. F. Eriksen, D. S. Colvard, N. J. Berg, M. L. Graham, K. G. Mann, T. C. Spelsberg, and B. L. Riggs, Evidence of estrogen receptors in normal human osteoblast-like cells. Science 241, 84 – 86 (1988). 393. D. T. Baran, J. M. Quail, R. Ray, J. Leszyk, and T. Honeyman, Annexin II is the membrane receptor that mediates the rapid actions of 1 alpha, 25-dihydroxyvitamin D3. J. Cell. Biochem. 78, 34 – 46 (2000). 394. M. Lieberherr, Effects of vitamin D3 metabolites on cytosolic free calcium in confluent mouse osteoblasts. J. Biol. Chem. 262, 13168 – 13173 (1987). 395. R. Civitelli, Y. S. Kim, S. L. Gunsten, A. Fujimori, M. Huskey, L. V. Avioli, and K. A. Hruska, Nongenomic activation of the calcium message system by vitamin D metabolites in osteoblast-like cells. Endocrinology 127, 2253 – 2262 (1990). 396. D. T. Baran, A. M. Sorensen, V. Shalhoub, T. Owen, A. Oberdorf, G. Stein, and J. Lian, 1 Alpha, 25-dihydroxyvitamin D3 rapidly increases cytosolic calcium in clonal rat osteosarcoma cells lacking the vitamin D receptor. J. Bone Miner. Res. 6, 1269 – 1275 (1991). 397. C. Menaa, R. D. Devlin, S. V. Reddy, Y. Gazitt, S. J. Choi, and G. D. Roodman, Annexin II increases osteoclast formation by stimulating the proliferation of osteoclast precursors in human marrow cultures. J. Clin. Invest. 103, 1605 – 1613 (1999). 398. D. Modrowski, A. Lomri, and P. J. Marie, Endogenous GM-CSF is involved as an autocrine growth factor for human osteoblastic cells. J. Cell. Physiol. 170, 35 – 46 (1997). 399. M. Alsina and G. D. Roodman, Nonskeletal effects of cytokines secreted by bone cells. In “Skeletal Growth Factors” (E. Canalis, ed.), pp. 375 – 384. Lippincott Williams & Wilkins, Philadelphia, PA, 2000. 400. S. L. Teitelbaum, M. M. Tondravi, and F. P. Ross, Osteoclast biology. In “Osteoporosis” (R. Marcus, D. Feldman, and J. Kelsey, eds.), pp. 61 – 94. Academic Press, San Diego, 1996. 401. P. A. Hill, A. Tumber, S. Papaioannou, and M. C. Meikle, The cellular actions of interleukin-11 on bone resorption in vitro. Endocrinology 139, 1564 – 1572 (1998).
61 402. D. Cheleuitte, S. Mizuno, and J. Glowacki, In vitro secretion of cytokines by human bone marrow: Effects of age and estrogen status. J. Clin. Endocrinol. Metab. 83, 2043 – 2051 (1998). 403. T. Suda, N. Takahashi, N. Udagawa, E. Jimi, M. T. Gillespie, and T. J. Martin, Modulation of osteoclast differentiation and function by the new members of the tumor necrosis factor receptor and ligand families. Endocr. Rev. 20, 345 – 357 (1999). 404. S. V. Reddy and G. D. Roodman, Control of osteoclast differentiation. Crit. Rev. Eukaryot. Gene Expr. 8, 1 – 17 (1998). 405. S. L. Teitelbaum, M. M. Tondravi, and F. P. Ross, Osteoclasts, macrophages, and the molecular mechanisms of bone resorption. J. Leukocyle. Biol. 61, 381 – 388 (1997). 406. K. Insogna, S. Tanaka, L. Neff, W. Horne, J. Levy, and R. Baron, Role of c-Src in cellular events associated with colony-stimulating factor-1-induced spreading in osteoclasts. Mol. Reprod. Dev. 46, 104 – 108 (1997). 407 .W. S. Simonet, D. L. Lacey, C. R. Dunstan, M. Kelley, M. S. Chang, R. Luthy, H. Q. Nguyen, S. Wooden, L. Bennett, T. Boone, G. Shimamoto, M. DeRose, R. Elliott, A. Colombero, H. L. Tan, G. Trail, J. Sullivan, E. Davy, N. Bucay, L. Renshaw-Gegg, T. M. Hughes, D. Hill, W. Pattison, P. Campbell, and W. J. Boyle, Osteoprotegerin: A novel secreted protein involved in the regulation of bone density. Cell 89, 309 – 319 (1997). 408. N. Yamamoto, S. Akiyama, T. Katagiri, M. Namiki, T. Kurokawa, and T. Suda, Smad1 and smad5 act downstream of intracellular signalings of BMP-2 that inhibits myogenic differentiation and induces osteoblast differentiation in C2C12 myoblasts. Biochem. Biophys. Res. Commun. 238, 574 – 580 (1997). 409. B. S. Kwon, S. Wang, N. Udagawa, V. Haridas, Z. H. Lee, K. K. Kim, K. O. Oh, J. Greene, Y. Li, J. Su, R. Gentz, B. B. Aggarwal, and J. Ni, TR1, a new member of the tumor necrosis factor receptor superfamily, induces fibroblast proliferation and inhibits osteoclastogenesis and bone resorption. FASEB J. 12, 845 – 854 (1998). 410. E. Tsuda, M. Goto, S. Mochizuki, K. Yano, F. Kobayashi, T. Morinaga, and K. Higashio, Isolation of a novel cytokine from human fibroblasts that specifically inhibits osteoclastogenesis. Biochem. Biophys. Res. Commun. 234, 137 – 142 (1997). 411. H. Yasuda, N. Shima, N. Nakagawa, S. I. Mochizuki, K. Yano, N. Fujise, Y. Sato, M. Goto, K. Yamaguchi, M. Kuriyama, T. Kanno, A. Murakami, E. Tsuda, T. Morinaga, and K. Higashio, Identity of osteoclastogenesis inhibitory factor (OCIF) and osteoprotegerin (OPG): A mechanism by which OPG/OCIF inhibits osteoclastogenesis in vitro. Endocrinology 139, 1329 – 1337 (1998). 412. N. Udagawa, N. Takahashi, T. Akatsu, H. Tanaka, T. Sasaki, T. Nishihara, T. Koga, T. J. Martin, and T. Suda, Origin of osteoclasts: Mature monocytes and macrophages are capable of differentiating into osteoclasts under a suitable microenvironment prepared by bone marrow-derived stromal cells. Proc. Natl. Acad. Sci. USA 87, 7260 – 7264 (1990). 413. N. Bucay, I. Sarosi, C. R. Dunstan, S. Morony, J. Tarpley, C. Capparelli, S. Scully, H. L. Tan, W. Xu, D. L. Lacey, W. J. Boyle, and W. S. Simonet, Osteoprotegerin-deficient mice develop early onset osteoporosis and arterial calcification. Genes Dev. 12, 1260 – 1268 (1998). 414. A. Mizuno, N. Amizuka, K. Irie, A. Murakami, N. Fujise, T. Kanno, Y. Sato, N. Nakagawa, H. Yasuda, S. Mochizuki, T. Gomibuchi, K. Yano, N. Shima, N. Washida, E. Tsuda, T. Morinaga, K. Higashio, and H. Ozawa, Severe osteoporosis in mice lacking osteoclastogenesis inhibitory factor/osteoprotegerin. Biochem. Biophys. Res. Commun. 247, 610 – 615 (1998). 415. H. Yasuda, N. Shima, N. Nakagawa, K. Yamaguchi, M. Kinosaki, S. Mochizuki, A. Tomoyasu, K. Yano, M. Goto, A. Murakami, E. Tsuda, T. Morinaga, K. Higashio, N. Udagawa, N. Takahashi, and T. Suda, Osteoclast differentiation factor is a ligand for osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical to TRANCE/RANKL. Proc. Natl. Acad. Sci. USA 95, 3597 – 3602 (1998).
62 416. D. L. Lacey, E. Timms, H. L. Tan, M. J. Kelley, C. R. Dunstan, T. Burgess, R. Elliott, A. Colombero, G. Elliott, S. Scully, H. Hsu, J. Sullivan, N. Hawkins, E. Davy, C. Capparelli, A. Eli, Y. X. Qian, S. Kaufman, I. Sarosi, V. Shalhoub, G. Senaldi, J. Guo, J. Delaney, and W. J. Boyle, Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell 93, 165 – 176 (1998). 417. B. R. Wong, R. Josien, S. Y. Lee, B. Sauter, H. L. Li, R. M. Steinman, and Y. Choi, TRANCE (tumor necrosis factor [TNF]-related activation-induced cytokine), a new TNF family member predominantly expressed in T cells, is a dendritic cell-specific survival factor. J. Exp. Med. 186, 2075 – 2080 (1997). 418. D. M. Anderson, E. Maraskovsky, W. L. Billingsley, W. C. Dougall, M. E. Tometsko, E. R. Roux, M. C. Teepe, R. F. DuBose, D. Cosman, and L. Galibert, A homologue of the TNF receptor and its ligand enhance T-cell growth and dendritic-cell function. Nature 390, 175 – 179 (1997). 419. Proposed standard nomenclature for new tumor necrosis factor family members involved in the regulation of bone resorption. The American Society for Bone and Mineral Research President’s Committee on Nomenclature. J. Bone Miner. Res. 15, 2293 – 2296 (2000). 420. Y. Y. Kong, H. Yoshida, I. Sarosi, H. L. Tan, E. Timms, C. Capparelli, S. Morony, A. J. Oliveira-dos-Santos, G. Van, A. Itie, W. Khoo, A. Wakeham, C. R. Dunstan, D. L. Lacey, T. W. Mak, W. J. Boyle, and J. M. Penninger, OPGL is a key regulator of osteoclastogenesis, lymphocyte development and lymph-node organogenesis. Nature 397, 315 – 323 (1999). 421. A. E. Hughes, S. H. Ralston, J. Marken, C. Bell, H. MacPherson, R. G. Wallace, W. van Hul, M. P. Whyte, K. Nakatsuka, L. Hovy, and D. M. Anderson, Mutations in TNFRSF11A, affecting the signal peptide of RANK, cause familial expansile osteolysis. Nature Genet. 24, 45 – 48 (2000). 422. L. C. Hofbauer, S. Khosla, C. R. Dunstan, D. L. Lacey, W. J. Boyle, and B. L. Riggs, The roles of osteoprotegerin and osteoprotegerin ligand in the paracrine regulation of bone resorption. J. Bone Miner. Res. 15, 2 – 12 (2000). 423. K. Thirunavukkarasu, D. L. Halladay, R. R. Miles, X. Yang, R. J. Galvin, S. Chandrasekhar, T. J. Martin, and J. E. Onyia, The osteoblast-specific transcription factor Cbfa1 regulates the expression of osteoprotegerin (OPG), a potent inhibitor of osteoclast differentiation and function. J. Biol. Chem. 275, 25163 – 25172 (2000). 424. R. Kitazawa, S. Kitazawa, and S. Maeda, Promoter structure of mouse RANKL/TRANCE/OPGL/ODF gene. Biochim. Biophys. Acta 1445, 134 – 141 (1999). 425. H. Drissi, Q. Luc, R. Shakoori, S. Chuva de Sousa Lopes, J.-Y. Choi, A. Terry, M. Hu, S. Jones, J. C. Neil, J. B. Lian, J. L. Stein, A. J. Van Wijnen, and G. S. Stein, Transcriptional autoregulation of the bone related CBFA1/RUNX2 gene. J. Cell. Physiol. 184, 341 – 350 (2000). 426. Y. H. Gao, T. Shinki, T. Yuasa, H. Kataoka-Enomoto, T. Komori, T. Suda, and A. Yamaguchi, Potential role of Cbfa1, an essential transcriptional factor for osteoblast differentiation, in osteoclastogenesis: regulation of mRNA expression of osteoclast differentiation factor (ODF). Biochem. Biophys. Res. Commun. 252, 697 – 702 (1998). 427. A. Erlebacher, E. H. Filvaroff, J. Q. Ye, and R. Derynck, Osteoblastic responses to TGF-beta during bone remodeling. Mol. Biol. Cell 9, 1903 – 1918 (1998). 428. M. E. Jackson, V. Shalhoub, J. B. Lian, G. S. Stein, and S. C. Marks, Jr., Aberrant gene expression in cultured mammalian bone cells demonstrates an osteoblast defect in osteopetrosis. J. Cell. Biochem. 55, 366 – 372 (1994). 429. V. Shalhoub, M. E. Jackson, C. Paradise, G. S. Stein, J. B. Lian, and S. C. Marks, Jr., Heterogeneity of colony stimulating factor-1 gene expression in the skeleton of four osteopetrotic mutations in rats and mice. J. Cell. Physiol. 166, 340 – 350 (1996).
LIAN AND STEIN 430. K. Aoki, E. Didomenico, N. A. Sims, K. Mukhopadhyay, L. Neff, A. Houghton, M. Amling, J. B. Levy, W. C. Horne, and R. Baron, The tyrosine phosphatase SHP-1 is a negative regulator of osteoclastogenesis and osteoclast resorbing activity: Increased resorption and osteopenia in me(v)/me(v) mutant mice. Bone 25, 261 – 267 (1999). 431. W. A. Peck, S. J. Birge, and S. A. Fedak, Bone cells: Biochemical and biological studies after enzymatic isolation. Science 146, 1476 – 1477 (1964). 432. G. L. Wong and D. V. Cohn, Target cells in bone for parathormone and calcitonin are different: enrichment for each cell type by sequential digestion of mouse calvaria and selective adhesion to polymeric surfaces. Proc. Natl. Acad. Sci. USA 72, 3167 – 3171 (1975). 433. J. E. Aubin, J. N. M. Heersche, M. J. Merriless, and J. Sodek, Isolation of bone cell clones with differences in growth, hormone responses and extracellular matrix production. J. Cell Biol. 92, 452 – 461 (1982). 434. P. G. Robey and J. D. Termine, Human bone cells in vitro. Calcif. Tissue Int. 37, 453 – 460 (1985). 435. P. J. Marie, Human osteoblastic cells: A potential tool to assess the etiology of pathologic bone formation. J. Bone Miner. Res. 9, 1847 – 1850 (1994). 436. L. Malaval, D. Modrowski, A. K. Gupta, and J. E. Aubin, Cellular expression of bone-related proteins during in vitro osteogenesis in rat bone marrow stromal cell cultures. J. Cell Physiol. 158, 555 – 572 (1994). 437. S.-L. Cheng, J. W. Yang, L. Rifas, S.-F. Zhang, and L. V. Avioli, Differentiation of human bone marrow osteogenic stromal cell in vitro: Induction of the osteoblast phenotype by dexamethasone. Endocrinology 134, 277 – 286 (1994). 438. B. A. Ashton, C. C. Eaglesom, I. Bab, and M. E. Owen, Distribution of fibroblastic colony-forming cells in rabbit bone marrow and assay of their osteogenic potential by an in vivo diffusion chamber method. Calcif. Tissue Int. 36, 83 – 86 (1984). 439. P. S. Leboy, J. N. Beresford, C. Devlin, and M. E. Owen, Dexamethasone induction of osteoblast mRNAs in rat marrow stromal cell cultures. J. Cell. Physiol. 146, 370 – 378 (1991). 440. G. E. Schoeters, L. Saint-Georges, H. R. Van den, and O. Vanderborght, Mineralization of adult mouse bone marrow in vitro. Cell Tissue Kinet. 21, 363 – 374 (1988). 441. H. Nakahara, J. E. Dennis, S. P. Bruder, S. E. Haynesworth, D. P. Lennon, and A. I. Caplan, In vitro differentiation of bone and hypertrophic cartilage from periosteal-derived cells. Exp. Cell Res. 195, 492 – 503 (1991). 442. Y. Koshihara, M. Hirano, M. Kawamura, H. Oda, and S. Higaki, Mineralization ability of cultured human osteoblast-like periosteal cells does not decline with aging. J. Gerontol. 46, B201 – B206 (1991). 443. B. Ecarot-Charrier, F. H. Glorieux, M. van der Rest, and G. Pereira, Osteoblasts isolated from mouse calvaria initiate matrix mineralization in culture. J. Cell. Biol. 96, 639 – 643 (1983). 444. J. R. Nefussi, M. L. Boy-Lefevre, H. Boulekbache, and N. Forest, Mineralization in vitro of matrix formed by osteoblasts isolated by collagenase digestion. Differentiation 29, 160 – 168 (1985). 445. U. Bhargava, M. Bar-Lev, C. G. Bellows, and J. E. Aubin, Ultrastructural analysis of bone nodules formed in vitro by isolated fetal rat calvaria cells. Bone 9, 155 – 163 (1988). 446. L. C. Gerstenfeld, S. D. Chipman, J. Glowacki, and J. B. Lian, Expression of differentiated function by mineralizing cultures of chicken osteoblasts. Dev. Biol. 122, 49 – 60 (1987). 447. S. W. Whitson, W. Harrison, M. K. Dunlap, D. E. Bowers, Jr., L. W. Fisher, P. G. Robey, and J. D. Termine, Fetal bovine cells synthesize bone-specific matrix proteins. J. Cell Biol. 99, 607 – 614 (1984). 448. L. G. Jenis, B. Ongphiphadhanakul, L. E. Braverman, G. S. Stein, J. B. Lian, R. Lew, and D. T. Baran, Responsiveness of gene expression markers of osteoblastic and osteoclastic activity to calcitonin in
CHAPTER 2 Osteoblast Biology
449.
450.
451.
452.
453.
454.
455.
456.
457.
458.
459.
460.
461.
462.
463.
464.
the appendicular and axial skeleton of the rat in vivo. Calcif. Tissue Int. 54, 511 – 515 (1994). P. J. Marie, M. C. de Vernejoul, D. Connes, and M. Hott, Decreased DNA synthesis by cultured osteoblastic cells in eugonadal osteoporotic men with defective bone formation. J. Clin. Invest 88, 1167 – 1172 (1991). B. G. Mills, F. R. Singer, L. P. Weiner, and P. A. Holst, Long-term culture of cells from bone affected by Paget’s disease. Calcif. Tissue Int. 29, 79 – 87 (1979). J. E. Puzas, J. S. Brand, and K. V. Jackman, Isolation and characterization of bone cells from a patient with osteogenesis imperfecta type 1b. Bone Miner. 1, 373 – 382 (1986). C. Silve, B. Grosse, C. Tau, M. Garabedian, J. Fritsch, P. D. Delmas, G. Cournot-Witmer, and S. Balsan, Response to parathyroid hormone and 1,25-dihydroxyvitamin D3 of bone- derived cells isolated from normal children and children with abnormalities in skeletal development. J Clin. Endocrinol. Metab. 62, 583 – 590 (1986). M. Kassem, K. Brixen, L. Mosekilde, and E. F. Eriksen, Human marrow stromal osteoblast-like cells do not show reduced responsiveness to in vitro stimulation with growth hormone in patients with postmenopausal osteoporosis. Calcif. Tissue Int. 54, 1 – 6 (1994). M. M. Wong, L. G. Rao, H. Ly, L. Hamilton, S. Ish-Shalom, W. Sturtridge, J. Tong, R. McBroom, R. G. Josse, and T. M. Murray, In vitro study of osteoblastic cells from patients with idiopathic osteoporosis and comparison with cells from nonosteoporotic controls. Osteoporos. Int. 4, 21 – 31 (1994). P. J. Marie, M. C. de Vernejoul, and A. Lomri, Stimulation of bone formation in osteoporosis patients treated with fluoride associated with increased DNA synthesis by osteoblastic cells in vitro. J Bone Miner. Res. 7, 103 – 113 (1992). J. B. Kopp and P. G. Robey, Sodium fluoride does not increase human bone cell proliferation or protein synthesis in vitro. Calcif. Tissue Int. 47, 221 – 229 (1990). K. W. Ng, P. R. Gummer, V. P. Michelangeli, J. F. Bateman, T. Mascara, W. G. Cole, and T. J. Martin, Regulation of alkaline phosphatase expression in a neonatal rat clonal calvarial cell strain by retinoic acid. J. Bone Miner. Res. 3, 53 – 61 (1988). J. E. Aubin, I. Tertinegg, R. Ber, and J. N. Heersche, Consistent patterns of changing hormone responsiveness during continuous culture of cloned rat calvaria cells. J. Bone Miner. Res. 3, 333 – 339 (1988). A. Yamaguchi and A. J. Kahn, Clonal osteogenic cell lines express myogenic and adipocytic developmental potential. Calcif. Tissue Int. 49, 221 – 225 (1991). M. Namiki, S. Akiyama, T. Katagiri, A. Suzuki, N. Ueno, N. Yamaji, V. Rosen, J. M. Wozney, and T. Suda, A kinase domain-truncated type I receptor blocks bone morphogenetic protein-2-induced signal transduction in C2C12 myoblasts. J. Biol. Chem. 272, 22046 – 22052 (1997). M. Inada, T. Katagiri, S. Akiyama, M. Namika, M. Komaki, A. Yamaguchi, K. Kamoi, V. Rosen, and T. Suda, Bone morphogenetic protein-12 and -13 inhibit terminal differentiation of myoblasts, but do not induce their differentiation into osteoblasts. Biochem. Biophys. Res. Commun. 222, 317 – 322 (1996). D. Gazit, R. Ebner, A. J. Kahn, and R. Derynck, Modulation of expression and cell surface binding of members of the transforming growth factor-beta superfamily during retinoic acid-induced osteoblastic differentiation of multipotential mesenchymal cells. Mol. Endocrinol. 7, 189 – 198 (1993). N. Udagawa, N. Takahashi, T. Akatsu, T. Sasaki, A. Yamaguchi, H. Kodama, T. J. Martin, and T. Suda, The bone marrow-derived stromal cell lines MC3T3-G2/PA6 and ST2 support osteoclast-like cell differentiation in cocultures with mouse spleen cells. Endocrinology 125, 1805 – 1813 (1989). H. Kodama, Y. Amagai, H. Sudo, S. Kasai, and S. Yamamoto, Establishment of a clonal osteogenic cell line from newborn mouse calvaria. Jpn. J. Oral Biol. 23, 899 – 901 (1981).
63 465. J.-Y. Choi, B.-H. Lee, K.-B. Song, R.-W. Park, I.-S. Kim, K.-Y. Sohn, J.-S. Jo, and H.-M. Ryoo, Expression patterns of bone-related proteins during osteoblastic differentiation in MC3T3-E1 cells. J. Cell. Biochem. 61, 609 – 618 (1996). 466. H. Sudo, H.-A. Kodama, Y. Amagai, S. Yamamoto, and S. Kasai, In vitro differentiation and calcification in a new clonal osteogenic cell line derived from newborn mouse calvaria. J. Cell Biol. 96, 191 – 198 (1983). 467. J. K. Heath, S. B. Rodan, K. Yoon, and G. A. Rodan, Rat calvarial cell lines immortalized with SV-40 large T antigen: Constitutive and retinoic acid-inducible expression of osteoblastic features. Endocrinology 124, 3060 – 3068 (1989). 468. E. Mathieu, G. Schoeters, P. F. vander, and J. Merregaert, Establishment of an osteogenic cell line derived from adult mouse bone marrow stroma by use of a recombinant retrovirus. Calcif. Tissue Int. 50, 362 – 371 (1992). 469. A. Houghton, B. O. Oyajobi, G. A. Foster, R. G. Russell, and B. M. Stringer, Immortalization of human marrow stromal cells by retroviral transduction with a temperature sensitive oncogene: Identification of bipotential precursor cells capable of directed differentiation to either an osteoblast or adipocyte phenotype. Bone 22, 7 – 16 (1998). 470. B. S. Komm, S. Wang, J. G. Haddad, and C. R. Lyttle, Characterization of conditionally transformed osteoblast-like cells derived from rat calvaria. J. Bone Miner. Res. 8, S359 – S359 (1993). [Abstract] 471. P. E. Keeting, R. E. Scott, D. S. Colvard, M. A. Anderson, M. J. Oursler, T. C. Spelsberg, and B. L. Riggs, Development and characterization of a rapidly proliferating, well-differentiated cell line derived from normal adult human osteoblast-like cells transfected with SV40 large T antigen. J. Bone Miner. Res. 7, 127 – 136 (1992). 472. K. C. Hicok, T. Thomas, F. Gori, D. J. Rickard, T. C. Spelsberg, and B. L. Riggs, Development and characterization of conditionally immortalized osteoblast precursor cell lines from human bone marrow stroma. J. Bone Miner. Res. 13, 205 – 217 (1998). 473. Y. Kato, J. J. Windle, B. A. Koop, G. R. Mundy, and L. F. Bonewald, Establishment of an osteocyte-like cell line, MLO-Y4. J. Bone Miner. Res. 12, 2014 – 2023 (1997). 474. R. J. Majeska, S. B. Rodan, and G. A. Rodan, Maintenance of parathyroid hormone response in clonal rat osteosarcoma lines. Exp. Cell Res. 111, 465 – 468 (1978). 475. S. K. Nishimoto and P. A. Price, Secretion of the vitamin K-dependent protein of bone rat osteosarcoma cells: Evidence for an intracellular precursor. J. Biol. Chem. 255, 6579 – 6583 (1980). 476. J. M. Nelissen, R. Torensma, M. Pluyter, G. J. Adema, R. A. Raymarkers, Y. van Kooyk, and C. G. Figdor, Molecular analysis of the hematopoiesis supporting osteoblastic cell line U2-OS. Exp. Hematol. 28, 422 – 432 (2000). 477. G. A. Rodan and M. Noda, Gene expression in osteoblastic cells. Crit. Rev. Eukary. Gene Expr. 1, 85 – 98 (1990). 478. G. S. Stein, J. B. Lian, T. A. Owen-Hughes, J. Holthuis, R. Bortell, and A. J. van Wijnen, Molecular mechanisms that mediate a functional relationship between proliferation and differentiation. In “Molecular and Cellular Approaches to the Control of Proliferation and Differentiation” (G. S. Stein and J. B. Lian, eds.), pp. 299 – 341. Academic Press, Orlando, 1992. 479. R. Bortell, T. A. Owen, V. Shalhoub, A. Heinrichs, M. A. Aronow, C. Rochette-Egly, Y. Lutz, J. L. Stein, J. B. Lian, and G. S. Stein, Constitutive transcription of the osteocalcin gene in osteosarcoma cells is reflected by altered protein-DNA interactions at promoter regulatory elements. Proc. Natl. Acad. Sci. USA 90, 2300 – 2304 (1993). 480. J. Holthuis, T. A. Owen, A. J. van Wijnen, K. L. Wright, A. RamseyEwing, M. B. Kennedy, R. Carter, S. C. Cosenza, K. J. Soprano, J. B. Lian et al., Tumor cells exhibit deregulation of the cell cycle histone gene promoter factor HiNF-D. Science 247, 1454 – 1457 (1990).
64 481. F. M. van den Ent, A. J. van Wijnen, T. J. Last, R. Bortell, J. L. Stein, J. B. Lian, and G. S. Stein, Concerted control of multiple histone promoter factors during cell density inhibition of proliferation in osteosarcoma cells: Reciprocal regulation of cell cycle-controlled and bone-related genes. Cancer Res. 53, 2399 – 2409 (1993). 482. J. B. Lian, G. S. Stein, J. L. Stein, R. Bortell, T. A. Owen, V. Shalhoub, A. Heinrichs, C. W. van Oordt, and A. J. van Wijnen, Implications of abrogated growth control on expression of a cell cycleregulated histone and the osteoblast-specific osteocalcin genes in osteosarcoma cells. In “Frontiers of Osteosarcoma Research: Interdisciplinary Survey of Clinical and Research Advances” (J. F. Novak and J. H. McMaster, eds.), pp. 227 – 287. Hogrefe & Huber, Seattle, WA, 1993. 483. T. A. Owen, M. Aronow, V. Shalhoub, L. M. Barone, L. Wilming, M. S. Tassinari, M. B. Kennedy, S. Pockwinse, J. B. Lian, and G. S. Stein, Progressive development of the rat osteoblast phenotype in vitro: Reciprocal relationships in expression of genes associated with osteoblast proliferation and differentiation during formation of the bone extracellular matrix. J. Cell. Physiol. 143, 420 – 430 (1990). 484. L. D. Quarles, D. A. Yohay, L. W. Lever, R. Caton, and R. J. Wenstrup, Distinct proliferative and differentiated stages of murine MC3T3-E1 cells in culture: An in vitro model of osteoblast development. J. Bone Miner. Res. 7, 683 – 692 (1992). 485. K. Ibaraki, J. D. Termine, S. W. Whitson, and M. F. Young, Bone matrix mRNA expression in differentiating fetal bovine osteoblasts. J. Bone Miner. Res. 7, 743 – 754 (1992). 486. G. S. Stein and J. B. Lian, Molecular mechanisms mediating proliferation/differentiation interrelationships during progressive development of the osteoblast phenotype. Endocr. Rev. 14, 424 – 442 (1993). 487. T. A. Owen, M. S. Aronow, L. M. Barone, B. Bettencourt, G. S. Stein, and J. B. Lian, Pleiotropic effects of vitamin D on osteoblast gene expression are related to the proliferative and differentiated state of the bone cell phenotype: Dependency upon basal levels of gene expression, duration of exposure, and bone matrix competency in normal rat osteoblast cultures. Endocrinology 128, 1496 – 1504 (1991). 488. H. Ishida, C. G. Bellows, J. E. Aubin, and J. N. Heersche, Characterization of the 1,25-(OH)2D3-induced inhibition of bone nodule formation in long-term cultures of fetal rat calvaria cells. Endocrinology 132, 61 – 66 (1993). 489. M. Broess, A. Riva, and L. C. Gerstenfeld, Inhibitory effects of 1,25(OH)2 vitamin D3 on collagen type I, osteopontin, and osteocalcin gene expression in chicken osteoblasts. J Cell. Biochem. 57, 440 – 451 (1995). 490. M. E. Antosz, C. G. Bellows, and J. E. Aubin, Effects of transforming growth factor beta and epidermal growth factor on cell proliferation and the formation of bone nodules in isolated fetal rat calvaria cells. J. Cell. Physiol. 140, 386 – 395 (1989). 491. E. C. Breen, R. A. Ignotz, L. McCabe, J. L. Stein, G. S. Stein, and J. B. Lian, TGF beta alters growth and differentiation related gene expression in proliferating osteoblasts in vitro, preventing development of the mature bone phenotype. J. Cell. Physiol. 160, 323 – 335 (1994). 492. S. E. Harris, L. F. Bonewald, M. A. Harris, M. Sabatini, S. Dallas, J. Q. Feng, N. Ghosh-Choudhury, J. Wozney, and G. R. Mundy, Effects of transforming growth factor beta on bone nodule formation and expression of bone morphogenetic protein 2, osteocalcin, alkaline phosphatase, and type I collagen mRNA in long-term cultures of fetal rat calvarial osteoblasts. J. Bone Miner. Res. 9, 855 – 863 (1994). 493. C. G. Bellows, J. E. Aubin, and J. N. M. Heersche, Physiological concentrations of glucocorticoids stimulate formation of bone nodules from isolated rat calvaria cells in vitro. Endocrinology 121, 1985 – 1992 (1987).
LIAN AND STEIN 494. V. Shalhoub, D. Conlon, M. Tassinari, C. Quinn, N. Partridge, G. S. Stein, and J. B. Lian, Glucocorticoids promote development of the osteoblast phenotype by selectively modulating expression of cell growth and differentiation associated genes. J. Cell. Biochem. 50, 425 – 440 (1992). 495. E. Canalis and J. B. Lian, 1,25-Dihydroxyvitamin D3 effects on collagen and DNA synthesis in periosteum and periosteum-free calvariae. Bone 6, 457 – 460 (1985). 496. J. R. Harrison, D. N. Petersen, A. C. Lichtler, A. T. Mador, D. W. Rowe, and B. E. Kream, 1,25-Dihydroxyvitamin D3 inhibits transcription of type I collagen genes in the rat osteosarcoma cell line ROS 17/2.8. Endocrinology 125, 327 – 333 (1986). 497. M. A. Aronow, L. C. Gerstenfeld, T. A. Owen, M. S. Tassinari, G. S. Stein, and J. B. Lian, Factors that promote progressive development of the osteoblast phenotype in cultured fetal rat calvaria cells. J. Cell. Physiol. 143, 213 – 221 (1990). 498. G. S. Stein, J. B. Lian, and T. A. Owen, Relationship of cell growth to the regulation of tissue- specific gene expression during osteoblast differentiation. FASEB J. 4, 3111 – 3123 (1990). 499. G. S. Stein and J. B. Lian, Molecular mechanisms mediating proliferation-differentiation interrelationships during progressive development of the osteoblast phenotype: Update 1995. Endocr. Rev. 4, 290 – 297 (1995). 500. N. Ghosh-Choudhury, M. A. Harris, J. Q. Feng, G. R. Mundy, and S. E. Harris, Expression of the BMP 2 gene during bone cell differentiation. Crit. Rev. Eukaryot. Gene Expr. 4, 345 – 355 (1994). 501. F. J. Hughes, J. Collyer, M. Stanfield, and S. A. Goodman, The effects of bone morphogenetic protein-2, -4, and -6 on differentiation of rat osteoblast cells in vitro. Endocrinology 136, 2671 – 2677 (1995). 502. S. D. Boden, G. Hair, L. Titus, M. Racine, K. McCuaig, J. M. Wozney, and M. S. Nanes, Glucocorticoid-induced differentiation of fetal rat calvarial osteoblasts is mediated by bone morphogenetic protein-6. Endocrinology 138, 2820 – 2828 (1997). 503. R. C. Pereira, S. Rydziel, and E. Canalis, Bone morphogenetic protein-4 regulates its own expression in cultured osteoblasts. J. Cell Physiol. 182, 239 – 246 (2000). 504. D. Chen, M. A. Harris, G. Rossini, C. R. Dunstan, S. L. Dallas, J. Q. Feng, G. R. Mundy, and S. E. Harris, Bone morphogenetic protein 2 (BMP-2) enhances BMP-3, BMP-4, and bone cell differentiation marker gene expression during the induction of mineralized bone matrix formation in cultures of fetal rat calvarial osteoblasts. Calcif. Tissue Int. 60, 283 – 290 (1997). 505. L. C. Gerstenfeld, A. Riva, K. Hodgens, D. R. Eyre, and W. J. Landis, Posttranslational control of collagen fibrillogenesis in mineralizing cultures of chick osteoblasts. J. Bone Miner. Res. 8, 1031 – 1043 (1993). 506. J. B. Lian, G. S. Stein, R. Bortell, and T. A. Owen, Phenotype suppression: A postulated molecular mechanism for mediating the relationship of proliferation and differentiation by Fos/Jun interactions at AP-1 sites in steroid responsive promoter elements of tissuespecific genes. J. Cell. Biochem. 45, 9 – 14 (1991). 507. L. R. McCabe, M. Kockx, J. Lian, J. Stein, and G. Stein, Selective expression of fos- and jun-related genes during osteoblast proliferation and differentiation. Exp. Cell Res. 218, 255 – 262 (1995). 508. S. R. Siddhanti and L. D. Quarles, Molecular to pharmacologic control of osteoblast proliferation and differentiation. J. Cell. Biochem. 55, 310 – 320 (1994). 509. M. Tamura and M. Noda, Identification of a DNA sequence involved in osteoblast-specific gene expression via interaction with helixloop-helix (HLH)-type transcription factors. J. Cell Biol. 126, 773 – 782 (1994). 510. M. P. Lynch, J. L. Stein, G. S. Stein, and J. B. Lian, Apoptosis during in vitro bone formation. J. Bone Miner. Res. 9, S352 – S352 (1994). [Abstract]
65
CHAPTER 2 Osteoblast Biology 511. S. K. Winchester, S. R. Bloch, G. J. Fiacco, and N. C. Partridge, Regulation of expression of collagenase-3 in normal, differentiating rat osteoblasts. J. Cell Physiol. 181, 479 – 488 (1999). 512. R. S. Birnbaum and D. M. Wiren, Changes in insulin-like growth factor-binding protein expression and secretion during the proliferation, differentiation, and mineralization of primary cultures of rat osteoblasts. Endocrinology 135, 223 – 230 (1994). 513. A. R. Shakoori, A. M. Oberdorf, T. A. Owen, L. A. Weber, E. Hickey, J. L. Stein, J. B. Lian, and G. S. Stein, Expression of heat shock genes during differentiation of mammalian osteoblasts and promyelocytic leukemia cells. J. Cell. Biochem. 48, 277 – 287 (1992). 514. H. M. Hoffmann, K. M. Catron, A. J. van Wijnen, L. R. McCabe, J. B. Lian, G. S. Stein, and J. L. Stein, Transcriptional control of the tissue-specific, developmentally regulated osteocalcin gene requires a binding motif for the Msx family of homeodomain proteins. Proc. Natl. Acad. Sci. USA 91, 12887 – 12891 (1994). 515. Y. Takeuchi, S. Fukumoto, K. Nakayama, and T. Matsumoto, Decrease in transforming growth factor- receptors by differentiation of osteoblastic cells is dependent upon collagen production. J. Bone Miner. Res. 9, S126 – S126 (1994). [Abstract] 516. L. M. Barone, T. A. Owen, M. S. Tassinari, R. Bortell, G. S. Stein, and J. B. Lian, Developmental expression and hormonal regulation of the rat matrix Gla protein (MGP) gene in chondrogenesis and osteogenesis. J. Cell. Biochem. 46, 351 – 365 (1991). 517. N. Kawaguchi, H. F. DeLuca, and M. Noda, Id gene expression and its suppression by 1,25-dihydroxyvitamin D3 in rat osteoblastic osteosarcoma cells. Proc. Natl. Acad. Sci. USA 89, 4569 – 4572 (1992). 518. K.-T. Tang, C. Capparelli, J. L. Stein, G. S. Stein, J. B. Lian, A. C. Huber, L. E. Braverman, and W. J. DeVito, Acidic fibroblast growth factor inhibits osteoblast differentiation in vitro: Altered expression of collagenase, cell growth-related and mineralization-associated genes. J. Cell. Biochem. 61, 152 – 166 (1996). 519. V. Shalhoub, F. Aslam, E. Breen, A. van Wijnen, R. Bortell, G. S. Stein, J. L. Stein, and J. B. Lian, Multiple levels of steroid hormonedependent control of osteocalcin during osteoblast differentiation: Glucocorticoid regulation of basal and vitamin D stimulated gene expression. J. Cell. Biochem. 69, 154 – 168 (1998). 520. L. C. Gerstenfeld, D. Zurakowski, J. L. Schaffer, D. P. Nichols, C. D. Toma, M. Broess, S. P. Bruder, and A. I. Caplan, Variable hormone responsiveness of osteoblast populations isolated at different stages of embryogenesis and its relationship to the osteogenic lineage. Endocrinology 137, 3957 – 3968 (1996). 521. P. V. N. Bodine, R. A. Henderson, J. Green, M. Aronow, T. Owen, G. S. Stein, J. B. Lian, and B. S. Komm, Estrogen receptor- is developmentally regulated during osteoblast differentiation and contributes to selective responsiveness of gene expression. Endocrinology 139, 2048 – 2057 (1998). 522. T. Ishizuya, S. Yokose, M. Hori, T. Noda, T. Suda, S. Yoshiki, and A. Yamaguchi, Parathyroid hormone exerts disparate effects on osteoblast differentiation depending on exposure time in rat osteoblastic cells. J Clin. Invest. 99, 2961 – 2970 (1997). 523. Y. Ishida and J. N. Heersche, Progesterone- and dexamethasone-dependent osteoprogenitors in bone cell populations derived from rat vertebrae are different and distinct. Endocrinology 140, 3210 – 3218 (1999). 524. C. G. Bellows, S. M. Reimers, and J. N. Heersche, Expression of mRNAs for type-I collagen, bone sialoprotein, osteocalcin, and osteopontin at different stages of osteoblastic differentiation and their regulation by 1,25dihydroxyvitamin D3. Cell Tissue Res. 297, 249 – 259 (1999). 525. C. G. Bellows, A. Ciaccia, and J. N. Heersche, Osteoprogenitor cells in cell populations derived from mouse and rat calvaria differ in
526.
527. 528. 529.
530.
531.
532.
533.
534.
535.
536.
537.
538.
539. 540.
541.
542.
their response to corticosterone, cortisol, and cortisone. Bone 23, 119 – 125 (1998). A. Hollnagel, M. Ahrens, and G. Gross, Parathyroid hormone enhances early and suppresses late stages of osteogenic and chondrogenic development in a BMP-dependent mesenchymal differentiation system (C3H10T1/2). J. Bone Miner. Res. 12, 1993 – 2004 (1997). C. Gray and S. J. Jones, PTH(1-34) suppresses appositional bone formation by cultured rat cranial osteoblasts. Bone 23, 453 – 457 (1998). M. H. Kroll, Parathyroid hormone temporal effects on bone formation and resorption. Bull. Math. Biol. 62, 163 – 188 (2000). M. Iwasaki, K. Nakata, H. Nakahara, T. Nakase, T. Kimura, K. Kimata, A. I. Caplan, and K. Ono, Transforming growth factor-beta 1 stimulates chondrogenesis and inhibits osteogenesis in high density culture of periosteum- derived cells. Endocrinology 132, 1603 – 1608 (1993). P. J. Marie, A. Lomri, F. Debiais, and J. Lemonnier, Fibroblast growth factors and osteogenesis. In “Skeletal Growth Factors” (E. Canalis, ed.), pp. 179 – 196. Lippincott Williams & Wilkins, Philadelphia, 2000. Y. Ishida and J. N. Heersche, Glucocorticoid-induced osteoporosis: Both in vivo and in vitro concentrations of glucocorticoids higher than physiological levels attenuate osteoblast differentiation. J. Bone Miner. Res. 13, 1822 – 1826 (1998). E. Canalis, Clinical review 83: Mechanisms of glucocorticoid action in bone: implications to glucocorticoid-induced osteoporosis. J. Clin. Endocrinol. Metab. 81, 3441 – 3447 (1996). C. Kasperk, U. Schneider, U. Sommer, F. Niethard, and R. Ziegler, Differential effects of glucocorticoids on human osteoblastic cell metabolism in vitro. Calcif. Tissue Int. 57, 120 – 126 (1995). H. C. Tenenbaum and J. N. Heersche, Dexamethasone stimulates osteogenesis in chick periosteum in vitro. Endocrinology 117, 2211 – 2217 (1985). A. M. Delany, B. Y. Gabbitas, and E. Canalis, Cortisol downregulates osteoblast alpha 1 (I) procollagen mRNA by transcriptional and posttranscriptional mechanisms. J. Cell Biochem. 57, 488 – 494 (1995). S. M. Pockwinse, J. L. Stein, J. B. Lian, and G. S. Stein, Developmental stage-specific cellular responses to vitamin D and glucocorticoids during differentiation of the osteoblast phenotype: Interrelationship of morphology and gene expression by in situ hybridization. Exp. Cell Res. 216, 244 – 260 (1995). J. B. Lian, V. Shalhoub, F. Aslam, B. Frenkel, J. Green, M. Hamrah, G. S. Stein, and J. L. Stein, Species-specific glucocorticoid and 1,25dihydroxyvitamin D responsiveness in mouse MC3T3-E1 osteoblasts: Dexamethasone inhibits osteoblast differentiation and vitamin D downregulates osteocalcin gene expression. Endocrinology 138, 2117 – 2127 (1997). S. C. Manolagas and R. S. Weinstein, New developments in the pathogenesis and treatment of steroid-induced osteoporosis. J. Bone Miner. Res. 14, 1061 – 1066 (1999). S. F. Hodgson, Corticosteroid-induced osteoporosis. Endocrinol. Metab. Clin. NA 19, 95 – 111 (1990). D. J. Defranco, J. B. Lian, and J. Glowacki, Differential effects of glucocorticoid on recruitment and activity of osteoclasts induced by normal and osteocalcin-deficient bone implanted in rats. Endocrinology 131, 114 – 121 (1992). I. R. Reid, G. E. Chapman, T. R. C. Fraser, A. D. Davies, A. S. Surus, J. Meyer, N. L. Hyq, and H. K. Ibbertson, Low serum osteocalcin levels in glucocorticoid-treated asthmatics. J. Clin. Endocrinol. Metab. 62, 379 – 383 (1986). J. B. Lian and G. S. Stein, Vitamin D regulation of osteoblast growth and differentiation. In “Nutrition and Gene Expression” (C. D. Berdanier and J. L. Hargrove, eds.), pp. 391 – 429. CRC Press, Boca Raton, FL, 1992.
66 543. J. M. Hock, M. Gunness-Hey, J. Poser, H. Olson, N. H. Bell, and L. G. Raisz, Stimulation of undermineralized matrix formation by 1,25 dihydroxyvitamin D3 in long bones of rats. Calcif. Tissue Int. 38, 79 – 86 (1986). 544. P. J. Marie and R. Travers, Continuous infusion of 1,25-dihydroxyvitamin D3 stimulates bone turnover in the normal young mouse. Calcif. Tissue Int. 35, 418 – 425 (1983). 545. G. J. Atkins, D. R. Haynes, S. M. Geary, M. Loric, T. N. Crotti, and D. M. Findlay, Coordinated cytokine expression by stromal and hematopoietic cells during human osteoclast formation. Bone 26, 653 – 661 (2000). 546. J. K. Burmester, R. J. Wiese, N. Maeda, and H. DeLuca, Structure and regulation of the rat 1,25-dihydroxyvitamin D3 receptor. Proc. Natl. Acad. Sci. USA 85, 9499 – 9502 (1988). 547. J. B. Lian, D. L. Carnes, and M. J. Glimcher, Bone and serum concentrations of osteocalcin as a function of 1,25-dihydroxyvitamin D3 circulating levels in bone disorders in rats. Endocrinology 120, 2123 – 2130 (1987). 548. L. C. Gerstenfeld, M. Broess, S. Bruder, A. Caplan, and W. J. Landis, Regulation of osteoblast extracellular matrix formation: Posttranslational regulation of extracellular matrix deposition and relationship between embryonic development and hormonal response. In “Chemistry and Biology of Mineralized Tissues: Proceedings of the Fourth International Conference on the Chemistry and Biology of Mineralized Tissues” (H. Slavkin and P. Price, eds.), pp. 287 – 315. Elsevier, Amsterdam, 1992. 549. H. Zhou, P. Choong, R. McCarthy, S. T. Chou, T. J. Martin, and K. W. Ng, In situ hybridization to show sequential expression of osteoblast gene markers during bone formation in vivo. J. Bone Miner. Res. 9, 1489 – 1499 (1994). 550. G. S. Stein, J. B. Lian, J. L. Stein, A. J. van Wijnen, and M. Montecino, Transcriptional control of osteoblast growth and differentiation. Physiol. Rev. 76, 593 – 629 (1996). 551. J. B. Lian and G. S. Stein, Osteoblast biology. In “Osteoporosis” (R. Marcus, D. Feldman, and J. Kelsey, eds.), pp. 23 – 60. Academic Press, San Diego, 1996. 552. J. E. Aubin and F. Liu, The osteoblast lineage. In “Principles of Bone Biology” (J. P. Bilezikian, L. G. Raisz, and G. A. Rodan, eds.), pp. 51 – 68. Academic Press, San Diego, 1996. 553. M. Dodig, M. S. Kronenberg, A. Bedalov, B. E. Kream, G. Gronowicz, S. H. Clark, K. Mack, Y. H. Liu, R. Maxon, Z. Z. Pan, W. B. Upholt, D. W. Rowe, and A. C. Lichtler, Identification of a TAAT-containing motif required for high level expression of the COL1A1 promoter in differentiated osteoblasts of transgenic mice. J. Biol. Chem. 271, 16422 – 16429 (1996). 554. J. A. Rossert, S. S. Chen, H. Eberspaecher, C. N. Smith, and B. de Crombrugghe, Identification of a minimal sequence of the mouse pro-1(I) collagen promoter that confers high-level osteoblast expression in transgenic mice and that binds a protein selectively present in osteoblasts. Proc. Natl. Acad. Sci. USA 93, 1027 – 1031 (1996). 555. E. P. Newberry, T. Latifi, and D. A. Towler, Reciprocal regulation of osteocalcin transcription by the homeodomain proteins msx2 and dlx5. Biochemistry 37, 16360 – 16368 (1998). 556. M. D. Benson, J. L. Bargeon, G. Xiao, P. E. Thomas, A. Kim, Y. Cui, and R. T. Franceschi, Identification of a homeodomain binding element in the bone sialoprotein gene promoter that is required for its osteoblast- selective expression. J. Biol. Chem. 275, 13907 – 13917 (2000). 557. M. Bidder, T. Latifi, and D. A. Towler, Reciprocal temporospatial patterns of Msx2 and Osteocalcin gene expression during murine odontogenesis. J. Bone Miner. Res. 13, 609 – 619 (1998). 558. H.-M. Ryoo, H. M. Hoffmann, T. L. Beumer, B. Frenkel, D. A. Towler, G. S. Stein, J. L. Stein, A. J. van Wijnen, and J. B. Lian, Stage-specific expression of Dlx-5 during osteoblast differentiation:
LIAN AND STEIN
559.
560.
561.
562.
563.
564.
565.
566.
567.
568.
569.
570.
571.
572.
573.
574.
Involvement in regulation of osteocalcin gene expression. Mol. Endocrinol. 11, 1681 – 1694 (1997). H. Zhang, G. Hu, H. Wang, P. Sciavolino, N. Iler, M. M. Shen, and C. Abate-Shen, Heterodimerization of Msx and Dlx homeoproteins results in functional antagonism. Mol. Cell. Biol. 17, 2920 – 2932 (1997). J. Rossert, H. Eberspaecher, and B. de Crombrugghe, Separate cisacting DNA elements of the mouse pro-alpha 1 (I) collagen promoter direct expression of reporter genes to different type I collagen-producing cells in transgenic mice. J. Cell Biol. 129, 1421 – 1432 (1995). D. A. Towler, C. D. Bennett, and G. A. Rodan, Activity of the rat osteocalcin basal promoter in osteoblastic cells is dependent upon homeodomain and CP1 binding motifs. Mol. Endocrinol. 8, 614 – 624 (1994). M. Machwate, A. Jullienne, M. Moukhtar, and P. J. Marie, Temporal variation of c-fos proto-oncogene expression during osteoblast differentiation and osteogenesis in developing bone. J. Cell. Biochem. 57, 62 – 70 (1995). A. E. Grigoriadis, K. Schellander, Z.-Q. Wang, and E. F. Wagner, Osteoblasts are target cells for transformation in c-fos transgenic mice. J. Cell Biol. 122, 685 – 701 (1993). L. R. McCabe, C. Banerjee, R. Kundu, R. J. Harrison, P. R. Dobner, J. L. Stein, J. B. Lian, and G. S. Stein, Developmental expression and activities of specific fos and jun proteins are funtionally related to osteoblast maturation: Role of fra-2 and jun D during differentiation. Endocrinology 137, 4398 – 4408 (1996). A. R. Kornblihtt, C. G. Pesce, C. R. Alonso, P. Cramer, A. Srebrow, S. Werbajh, and A. F. Muro, The fibronectin gene as a model for splicing and transcription studies. FASEB J. 10, 248 – 257 (1996). A. Staffa, N. H. Acheson, and A. Cochrane, Novel exonic elements that modulate splicing of the human fibronectin EDA exon. J Biol. Chem. 272, 33394 – 33401 (1997). M. Sugimoto, T. Kimura, N. Tsumaki, Y. Matsui, K. Nakata, H. Kawahata, N. Yasui, Y. Kitamura, S. Nomura, and T. Ochi, Differential in situ expression of alpha2(XI) collagen mRNA isoforms in the developing mouse. Cell Tissue Res. 292, 325 – 332 (1998). V. Geoffroy, D. A. Corral, L. Zhou, B. Lee, and G. Karsenty, Genomic organization, expression of the human CBFA1 gene, and evidence for an alternative splicing event affecting protein function. Mamm. Genome 9, 54 – 57 (1998). Z. S. Xiao, R. Thomas, T. K. Hinson, and L. D. Quarles, Genomic structure and isoform expression of the mouse, rat and human Cbfa1/Osf2 transcription factor. Gene 214, 187 – 197 (1998). Z. S. Xiao, T. K. Hinson, and L. D. Quarles, Cbfa1 isoform overexpression upregulates osteocalcin gene expression in non-osteoblastic and pre-osteoblastic cells. J. Cell. Biochem. 74, 596 – 605 (1999). C. Banerjee, A. Javed, J. Choi, J. Green, A. van Wijnen, J. Stein, G. Stein, and J. Lian, Representation and activities of multiple Cbfa transcription factors in osteoblasts. Bone 23, S321 – S321 (1998). [Abstract] A. Javed, B. Guo, S. Hiebert, J. Choi, J. Green, S. Zhao, M. A. Osborne, S. Stifani, J. L. Stein, G. S. Stein, A. J. van Wijnen, and J. B. Lian, RUNX (CBF/AML/PEBP2) dependent repression of tissuespecific gene transcription involves in situ association with Groucho/TLE/R-esp proteins. J. Bone Miner. Res. 15, S498 – S498 (2000). [Abstract] G. Xiao, D. Jiang, P. Thomas, M. D. Benson, K. Guan, G. Karsenty, and R. T. Franceschi, MAPK pathways activate and phosphorylate the osteoblast-specific transcription factor, Cbfa1. J. Biol. Chem. 275, 4453 – 4459 (2000). A. Javed, B. Guo, S. Hiebert, J.-Y. Choi, J. Green, S.-C. Zhao, M. A. Osborne, S. Stifani, J. L. Stein, J. B. Lian, A. J. van Wijnen, and G. S. Stein, Groucho/TLE/R-Esp proteins associate with the nuclear matrix and repress RUNX (CBF/AML/PEBP2) dependent
67
CHAPTER 2 Osteoblast Biology
575.
576.
577.
578.
579.
580.
581.
582.
583.
584.
585.
586.
587.
588.
589.
activation of tissue-specific gene transcription. J. Cell Sci. 113, 2221 – 2231 (2000). K. W. McLarren, R. Lo, D. Grbavec, K. Thirunavukkarasu, G. Karsenty, and S. Stifani, The mammalian basic helix loop helix protein HES-1 binds to and modulates the transactivating function of the runt-related factor cbfa1. J. Biol. Chem. 275, 530 – 538 (2000). D. Levanon, R. E. Goldstein, Y. Bernstein, H. Tang, D. Goldenberg, S. Stifani, Z. Paroush, and Y. Groner, Transcriptional repression by AML1 and LEF-1 is mediated by the TLE/Groucho corepressors. Proc. Natl. Acad. Sci. USA 95, 11590 – 11595. D. A. Towler, S. J. Rutledge, and G. A. Rodan, Msx-2/Hox 8.1: A transcriptional regulator of the rat osteocalcin promoter. Mol. Endocrinol. 8, 1484 – 1493 (1994). H. M. Hoffmann, T. L. Beumer, S. Rahman, L. R. McCabe, C. Banerjee, F. Aslam, J. A. Tiro, A. J. Van Wijnen, J. L. Stein, G. S. Stein and J. B. Lian, Bone tissue-specific transcription of the osteocalcin gene: role of an activator osteoblast-specific complex and suppressor hox proteins that bind the OC box. J. Cell. Biochem. 61, 310 – 324 (1996). H. Hoffmann, J. Green, A. J. van Wijnen, J. L. Stein, G. S. Stein, and J. B. Lian, Expression screening of factors binding to the osteocalcin bone-specific promoter element OC Box I: Isolation of a novel osteoblast differentiation-specific factor. J. Cell. Biochem. 80, 156 – 168 (2000). J. M. Boudreaux and D. A. Towler, Synergistic induction of osteocalcin gene expression: Identification of a bipartite element conferring fibroblast growth factor 2 and cyclic AMP responsiveness in the rat osteocalcin promoter. J. Biol. Chem. 271, 7508 – 7515 (1996). G. Boguslawski, L. V. Hale, X. P. Yu, R. R. Miles, J. E. Onyia, R. F. Santerre, and S. Chandrasekhar, Activation of osteocalcin transcription involves interaction of protein kinase A- and protein kinase C-dependent pathways. J. Biol. Chem. 275, 999 – 1006 (2000). T. A. Owen, R. Bortell, S. A. Yocum, S. L. Smock, M. Zhang, C. Abate, V. Shalhoub, N. Aronin, K. L. Wright, A. J. van Wijnen, J. L. Stein, T. Curran, J. B. Lian, and G. S. Stein, Coordinate occupancy of AP-1 sites in the vitamin D responsive and CCAAT box elements by Fos-Jun in the osteocalcin gene: Model for phenotype suppression of transcription. Proc. Natl. Acad. Sci. USA 87, 9990 – 9994 (1990). K. Ozono, T. Sone, and J. W. Pike, The genomic mechanism of action of 1,25-dihydroxyvitamin D3. J. Bone Miner. Res. 6, 1021 – 1027 (1991). C. Banerjee, J. L. Stein, A. J. van Wijnen, B. Frenkel, J. B. Lian, and G. S. Stein, Transforming growth factor-beta 1 responsiveness of the rat osteocalcin gene is mediated by an activator protein-1 binding site. Endocrinology 137, 1991 – 2000 (1996). M. B. Demay, J. M. Gerardi, H. F. DeLuca, and H. M. Kronenberg, DNA sequences in the rat osteocalcin gene that bind the 1,25-dihydroxyvitamin D3 receptor and confer responsive to 1,25-dihydroxyvitamin D3. Proc. Natl. Acad. Sci. USA 87, 369 – 373 (1990). E. R. Markose, J. L. Stein, G. S. Stein, and J. B. Lian, Vitamin Dmediated modifications in protein-DNA interactions at two promoter elements of the osteocalcin gene. Proc. Natl. Acad. Sci. USA 87, 1701 – 1705 (1990). S. A. Kerner, R. A. Scott, and J. W. Pike, Sequence elements in the human osteocalcin gene confer basal activation and inducible response to hormonal vitamin D3. Proc. Natl. Acad. Sci. USA 86, 4455 – 4459 (1989). C. M. Terpening, C. A. Haussler, P. W. Jurutka, M. A. Galligan, B. S. Komm, and M. R. Haussler, The vitamin D-responsive element in the rat bone Gla protein gene is an imperfect direct repeat that cooperates with other cis-elements in 1,25-dihydroxyvitamin D3-mediated transcriptional activation. Mol. Endocrinol. 5, 373 – 385 (1991). M. B. Demay, M. S. Kiernan, H. F. DeLuca, and H. M. Kronenberg, Characterization of 1,25-dihydroxyvitamin D3 receptor interactions
590.
591.
592.
593.
594.
595.
596.
597.
598.
599.
600.
601.
602.
603.
604.
with target sequences in the rat osteocalcin gene. Mol. Endocrinol. 6, 557 – 562 (1992). M. Schrader, S. Nayeri, J. P. Kahlen, K. M. Muller, and C. Carlberg, Natural vitamin D3 response elements formed by inverted palindromes: Polarity-directed ligand sensitivity of vitamin D3 receptorretinoid X receptor heterodimer-mediated transactivation. Mol. Cell. Biol. 15, 1154 – 1161 (1995). P. N. MacDonald, D. R. Sherman, D. R. Dowd, S. C. Jefcoat, Jr., and K. DeLisle, The vitamin D receptor interacts with general transcription factor IIB. J. Biol. Chem. 270, 4748 – 4752 (1995). T. A. Owen, R. Bortell, V. Shalhoub, A. Heinrichs, J. L. Stein, G. S. Stein, and J. B. Lian, Postproliferative transcription of the rat osteocalcin gene is reflected by vitamin D-responsive developmental modifications in protein-DNA interactions at basal and enhancer promoter elements. Proc. Natl. Acad. Sci. USA 90, 1503 – 1507 (1993). J. C. G. Blanco, I.-M. Wang, S. Y. Tsai, M.-J. Tsai, B. W. O’Malley, P. W. Jurutka, M. R. Haussler, and K. Ozato, Transcription factor TFIIB and the vitamin D receptor cooperatively activate ligand-dependent transcription. Proc. Natl. Acad. Sci. USA 92, 1535 – 1539 (1995). A. Staal, A. J. van Wijnen, J. B. Lian, J. L. Stein, H. DeLuca, J. C. Birkenhager, H. A. P. Pols, J. P. T. M. van Leeuwen, and G. S. Stein, Different identities of nuclear protein complexes recognizing the vitamin D responsive elements of the osteocalcin and osteopontin genes. J. Bone Miner. Res. 10, S167 – S167 (1995). [Abstract] P. N. MacDonald, D. R. Dowd, S. Nakajima, M. A. Galligan, M. C. Reeder, C. A. Haussler, K. Ozato, and M. R. Haussler, Retinoid X receptors stimulate and 9-cis retinoic acid inhibits 1,25-dihydroxyvitamin D3-activated expression of the rat osteocalcin gene. Mol. Cell. Biol. 13, 5907 – 5917 (1993). H. Kuno, S. M. Kurian, G. N. Hendy, J. White, H. F. DeLuca, C. O. Evans, and M. S. Nanes, Inhibition of 1,25-dihydroxyvitamin D3 stimulated osteocalcin gene transcription by tumor necrosis factoralpha: Structural determinants within the vitamin D response element. Endocrinology 134, 2524 – 2531 (1994). R. Schule, K. Umesono, D. J. Mangelsdorf, J. Bolado, J. W. Pike, and R. M. Evans, Jun-Fos and receptors for vitamins A and D recognize a common response element in the human osteocalcin gene. Cell 61, 497 – 504 (1990). B. Cheskis and L. P. Freedman, Ligand modulates the conversion of DNA-bound vitamin D3 receptor (VDR) homodimers into VDRretinoid X receptor heterodimers. Mol. Cell. Biol. 14, 3329 – 3338 (1994). M. Schrader, I. Bendik, M. Becker-Andre, and C. Carlberg, Interaction between retinoic acid and vitamin D signaling pathways. J. Biol. Chem. 268, 17830 – 17836 (1993). S. A. Kliewer, K. Umesono, D. J. Mangelsdorf, and R. M. Evans, Retinoid X receptor interacts with nuclear receptors in retinoic acid, thyroid hormone and vitamin D3 signalling. Nature 355, 446 – 449 (1992). R. Bortell, T. A. Owen, J. P. Bidwell, P. Gavazzo, E. Breen, A. J. van Wijnen, H. F. DeLuca, J. L. Stein, J. B. Lian, and G. S. Stein, Vitamin D-responsive protein-DNA interactions at multiple promoter regulatory elements that contribute to the level of rat osteocalcin gene expression. Proc. Natl. Acad. Sci. USA 89, 6119 – 6123 (1992). G. S. Stein, A. J. van Wijnen, M. Montecino, J. L. Stein, and J. B. Lian, Nuclear structure/gene expression interrelationships. J. Cell. Physiol. 181, 240 – 250 (1999). N. J. McKenna, Z. Nawaz, S. Y. Tsai, M. J. Tsai, and B. W. O’Malley, Distinct steady-state nuclear receptor coregulator complexes exist in vivo. Proc. Natl. Acad. Sci. USA 95, 11697 – 11702 (1998). A. N. Imbalzano, Energy-dependent chromatin remodelers: Complex complexes and their components. Crit. Rev. Eukaryot. Gene Expr. 8, 225 – 255 (1998).
68 605. J. L. Workman and R. E. Kingston, Alteration of nucleosome structure as a mechanism of transcriptional regulation. Annu. Rev. Biochem. 67, 545 – 579 (1998). 606. H. Kwon, A. N. Imbalzano, P. A. Khavari, R. E. Kingston, and M. R. Green, Nucleosome disruption and enhancement of activator binding by a human SWI/SNF complex. Nature 370, 477 – 481 (1994). 607. J. Cote, J. Quinn, J. L. Workman, and C. L. Peterson, Stimulation of GAL4 derivative binding to nucleosomal DNA by the yeast SWI/SNF complex. Science 265, 53 – 60 (1994). 608. C. L. Peterson, Y. Zhao, and B. T. Chait, Subunits of the yeast SWI/SNF complex are members of the actin- related protein (ARP) family. J Biol. Chem. 273, 23641 – 23644 (1998). 609. M. Vettese-Dadey, P. Walter, H. Chen, L.-J. Juan, and J. L. Workman, Role of the histone amino termini in facilitated binding of a transcription factor. Mol. Cell. Biol. 14, 970 – 981 (1994). 610. M. Vettese-Dadey, P. A. Grant, T. R. Hebbes, C. Crane-Robinson, C. D. Allis, and J. L. Workman, Acetylation of histone H4 plays a primary role in enhancing transcription factor binding to nucleosomal DNA in vitro. EMBO J. 15, 2508 – 2518 (1996). 611. K. Ura, H. Kurumizaka, S. Dimitrov, G. Almouzni, and A. P. Wolffe, Histone acetylation: Influence on transcription, nucleosome mobility and positioning, and linker histone-dependent transcriptional repression. EMBO J. 16, 2096 – 2107 (1997). 612. M. Montecino, S. Pockwinse, J. Lian, G. Stein, and J. Stein, DNase I hypersensitive sites in promoter elements associated with basal and vitamin D dependent transcription of the bone- specific osteocalcin gene. Biochemistry 33, 348 – 353 (1994). 613. M. Montecino, J. Lian, G. Stein, and J. Stein, Changes in chromatin structure support constitutive and developmentally regulated transcription of the bone-specific osteocalcin gene in osteoblastic cells. Biochemistry 35, 5093 – 5102 (1996). 614. M. Montecino, B. Frenkel, J. Lian, J. Stein, and G. Stein, Requirement of distal and proximal promoter sequences for chromatin organization of the osteocalcin gene in bone-derived cells. J. Cell. Biochem. 63, 221 – 228 (1996). 615. E. C. Breen, A. J. van Wijnen, J. B. Lian, G. S. Stein, and J. L. Stein, In vivo occupancy of the vitamin D responsive element in the osteocalcin gene supports vitamin D-dependent transcriptional upregulation in intact cells. Proc. Natl. Acad. Sci. USA 91, 12902 – 12906 (1994). 616. G. S. Stein, J. B. Lian, S. I. Dworetzky, T. A. Owen, R. Bortell, J. P. Bidwell, and A. J. van Wijnen, Regulation of transcription-factor activity during growth and differentiation: Involvement of the nuclear matrix in concentration and localization of promoter binding proteins. J. Cell. Biochem. 47, 300 – 305 (1991). 617. R. Berezney and D. S. Coffey, Nuclear protein matrix: Association with newly synthesized DNA. Science 189, 291 – 292 (1975). 618. B. D. Nelkin, D. M. Pardoll, and B. Vogelstein, Localization of SV40 genes within supercoiled loop domains. Nucleic Acids Res. 8, 5623 – 5633 (1980). 619. S. I. Robinson, B. D. Nelkin, and B. Vogelstein, The ovalbumin gene is associated with the nuclear matrix of chicken oviduct cells. Cell 28, 99 – 106 (1982). 620. S. I. Dworetzky, K. L. Wright, E. G. Fey, S. Penman, J. B. Lian, J. L. Stein, and G. S. Stein, Sequence-specific DNA-binding proteins are components of a nuclear matrix-attachment site. Proc. Natl. Acad. Sci. USA 89, 4178 – 4182 (1992). 621. A. J. van Wijnen, J. P. Bidwell, E. G. Fey, S. Penman, J. B. Lian, J. L. Stein, and G. S. Stein, Nuclear matrix association of multiple sequence-specific DNA binding activities related to SP-1, ATF, CCAAT, C/EBP, OCT-1, and AP-1. Biochemistry 32, 8397 – 8402 (1993). 622. A. Stief, D. M. Winter, W. H. Stratling, and A. E. Sippel, A nuclear DNA attachment element mediates elevated and position- independent gene activity. Nature 341, 343 – 345 (1989). 623. J. Schaack, W. Y. Ho, P. Freimuth, and T. Shenk, Adenovirus terminal protein mediates both nuclear matrix association and
LIAN AND STEIN
624.
625. 626.
627.
628.
629.
630.
631.
632.
633.
634.
635.
636.
637.
638.
639.
640.
efficient transcription of adenovirus DNA. Genes Dev. 4, 1197 – 1208 (1990). J. B. Lawrence, R. H. Singer, and L. M. Marselle, Highly localized tracks of specific transcripts within interphase nuclei visualized by in situ hybridization. Cell 57, 493 – 502 (1989). S. Zeitlin, A. Parent, S. Silverstein, and A. Efstratiadis, Pre-mRNA splicing and the nuclear matrix. Mol. Cell Biol. 7, 111 – 120 (1987). K. C. Carter, D. Bowman, W. Carrington, K. Fogarty, J. A. McNeil, F. S. Fay, and J. B. Lawrence, A three-dimensional view of precursor messenger RNA metabolism within the mammalian nucleus. Science 259, 1330 – 1335 (1993). D. L. Spector, Higher order nuclear organization: Three-dimensional distribution of small nuclear ribonucleoprotein particles. Proc. Natl. Acad. Sci. USA 87, 147 – 151 (1990). B. J. Blencowe, J. A. Nickerson, R. Issner, S. Penman, and P. A. Sharp, Association of nuclear matrix antigens with exon-containing splicing complexes. J. Cell Biol. 127, 593 – 607 (1994). J. R. Davie, Nuclear matrix, dynamic histone acetylation and transcriptionally active chromatin. Mol. Biol. Rep. 24, 197 – 207 (1997). S. I. Dworetzky, E. G. Fey, S. Penman, J. B. Lian, J. L. Stein, and G. S. Stein, Progressive changes in the protein composition of the nuclear matrix during rat osteoblast differentiation. Proc. Natl. Acad. Sci. USA 87, 4605 – 4609 (1990). C. Zeng, A. J. van Wijnen, J. L. Stein, S. Meyers, W. Sun, L. Shopland, J. B. Lawrence, S. Penman, J. B. Lian, G. S. Stein, and S. W. Hiebert, Identification of a nuclear matrix targeting signal in the leukemia and bone-related AML/CBF transcription factors. Proc. Natl. Acad. Sci. USA 94, 6746 – 6751 (1997). S. McNeil, B. Guo, J. L. Stein, J. B. Lian, S. Bushmeyer, E. Seto, M. L. Atchison, S. Penman, A. J. van Wijnen, and G. S. Stein, Targeting of the YY1 transcription factor to the nucleolus and the nuclear matrix in situ: The C-terminus is a principal determinant for nuclear trafficking. J. Cell. Biochem. 68, 500 – 510 (1998). L. Bruhn, A. Munnerlyn, and R. Grosschedl, ALY, a contextdependent coactivator of LEF-1 and AML-1, is required for TCRalpha enhancer function. Genes Dev. 11, 640 – 653 (1997). I. Kitabayashi, A. Yokoyama, K. Shimizu, and M. Ohki, Interaction and functional cooperation of the leukemia-associated factors AML1 and p300 in myeloid cell differentiation. EMBO J. 17, 2994 – 3004 (1998). H. Ogihara, T. Kanno, E. Morii, D. K. Kim, Y. M. Lee, M. Sato, W. Y. Kim, S. Nomura, Y. Ito, and Y. Kitamura, Synergy of PEBP2/CBF with mitranscription factor (MITF) for transactivation of mouse mast cell protease 6 gene. Oncogene 18, 4632 – 4639 (1999). M. Y. Ahn, G. Huang, S. C. Bae, H. J. Wee, W. Y. Kim, and Y. Ito, Negative regulation of granulocytic differentiation in the myeloid precursor cell line 32Dcl3 by ear-2, a mammalian homolog of Drosophila seven-up, and a chimeric leukemogenic gene, AML1/ETO. Proc. Natl. Acad. Sci. USA 95, 1812 – 1817 (1998). Y. M. Ning and D. M. Robins, AML3/CBFalpha1 is required for androgen-specific activation of the enhancer of the mouse sex-limited protein (Slp) gene. J. Biol. Chem. 274, 30624 – 30630 (1999). H. Sasaki-Iwaoka, K. Maruyama, H. Endoh, T. Komori, S. Kato, and H. Kawashima, A trans-acting enhancer modulates estrogenmediated transcription of reporter genes in osteoblasts. J. Bone Miner. Res. 14, 248 – 255 (1999). B. D. Aronson, A. L. Fisher, K. Blechman, M. Caudy, and J. P. Gergen, Groucho-dependent and -independent repression activities of Runt domain proteins. Mol. Cell Biol. 17, 5581 – 5587 (1997). B. Guo, A. Javed, J. Green, J. L. Stein, J. B. Lian, A. J. van Wijnen, and G. S. Stein, Groucho/TLE proteins associate with the nuclear matrix and repress Cbfa/AML mediated transactivation on osteocalcin promoter. Bone 23, S184 – S184 (1998). [Abstract]
69
CHAPTER 2 Osteoblast Biology 641. G, Chen, J. Fernandez, S. Mische, and A. J. Courey, A functional interaction between the histone deacetylase Rpd3 and the corepressor groucho in Drosophila development. Genes Dev. 13, 2218 – 2230 (1999). 642. A. Palaparti, A. Baratz, and S. Stifani, The Groucho/transducin-like enhancer of split transcriptional repressors interact with the genetically defined amino-terminal silencing domain of histone H3. J. Biol. Chem. 272, 26604 – 26610 (1997). 643. S. W. Hiebert, W. Sun, J. N. Davis, T. Golub, S. Shurtleff, A. Buijs, J. R. Downing, G. Grosveld, M. F. Roussell, D. G. Gilliland, N. Lenny, and S. Meyers, The t(12;21) translocation converts AML-1B from an activator to a repressor of transcription. Mol. Cell. Biol. 16, 1349 – 1355 (1996). 644. B. Lutterbach, J. J. Westendorf, B. Linggi, S. Isaac, E. Seto, and S. W. Hiebert, A mechanism of repression by acute myeloid leukemia-1, the target of multiple chromosomal translocations in acute leukemia. J. Biol. Chem. 275, 651 – 656 (2000). 645. A. Javed, G. L. Barnes, A. Javed, B. O. Jassanya, J. L. Stein, L. Gerstenfeld, G. S. Stein, and J. B. Lian, Cbfa family members act as a negative regulator of the bone sialoprotein gene. J. Bone Miner. Res. 14, S165 – S165 (1999). [Abstract] 646. H. Drissi, Q. Luc, R. Shakoori, S. Chuva de Sousa Lopes, J.-Y. Choi, A. Terry, M. Hu, S. Jones, J. C. Neil, J. B. Lian, J. L. Stein, A. J. van Wijnen, and G. S. Stein, Transcriptional autoregulation of the bone related CBFA1/RUNX2 gene. J. Cell. Physiol. 184, 341 – 350 (2000). 647. A. Javed, S. Gutierrez, M. Montecino, A. J. van Wijnen, J. L. Stein, G. S. Stein, and J. B. Lian, Multiple Cbfa/AML sites in the rat osteocalcin promoter are required for basal and vitamin D responsive transcription and contribute to chromatin organization. Mol. Cell. Biol. 19, 7491 – 7500 (1999). 648. C. Tribioli and T. Lufkin, The murine Bapx1 homeobox gene plays a critical role in embryonic development of the axial skeleton and spleen. Development 126, 5699 – 5711 (1999). 649. L. A. Lettice, L. A. Purdie, G. J. Carlson, F. Kilanowski, J. Dorin, and R. E. Hill, The mouse bagpipe gene controls development of axial skeleton, skull, and spleen. Proc. Natl. Acad. Sci. USA 96, 9695 – 9700 (1999). 650. J. A. Rivera-Perez, M. Wakamiya, and R. R. Behringer, Goosecoid acts cell autonomously in mesenchyme-derived tissues during craniofacial development. Development 126, 3811 – 3821 (1999). 651. M. J. Depew, J. K. Liu, J. E. Long, R. Presley, J. J. Meneses, R. A. Pedersen, and J. L. Rubenstein, Dlx5 regulates regional development of the branchial arches and sensory capsules. Development 126, 3831 – 3846 (1999). 652. D. Acampora, G. R. Merlo, L. Paleari, B. Zerega, M. P. Postiglione, S. Mantero, E. Bober, O. Barbieri, A. Simeone, and G. Levi Craniofacial, vestibular and bone defects in mice lacking the distalless-related gene dlx5. Development 126, 3795 – 3809 (1999). 653. I. Satokata, L. Ma, H. Ohshima, M. Bei, I. Woo, K. Nishizawa, T. Maeda, Y. Takano, M. Uchiyama, S. Heaney, H. Peters, Z. Tang, R. Maxson, and R. Maas, Msx2 deficiency in mice causes pleiotropic defects in bone growth and ectodermal organ formation. Nature Genet. 24, 391 – 395 (2000). 654. A. O. Wilkie, Z. Tang, N. Elanko, S. Walsh, S. R. Twigg, J. A. Hurst, S. A. Wall, K. H. Chrzanowska, and R. E. Maxson, Jr., Functional haploinsufficiency of the human homeobox gene MSX2 causes defects in skull ossification. Nature Genet. 24, 387 – 390 (2000). 655. Y. H. Liu, Z. Tang, R. K. Kundu, L. Wu, W. Luo, D. Zhu, F. Sangiorgi, M. L. Snead, and R. E. Maxson, Msx2 gene dosage influences the number of proliferative osteogenic cells in growth centers of the developing murine skull: A possible mechanism for MSX2mediated craniosynostosis in humans. Dev. Biol. 205, 260 – 274 (1999). 656. E. Filvaroff, A. Erlebacher, J. Ye, S. E. Gitelman, J. Lotz, M. Heillman, and R. Derynck, Inhibition of TGF-beta receptor signaling in
657.
658.
659.
660.
661.
662.
663.
664.
665.
666.
667.
668.
669.
670.
671.
672.
osteoblasts leads to decreased bone remodeling and increased trabecular bone mass. Development 126, 4267 – 4279 (1999). R. Serra, M. Johnson, E. H. Filvaroff, J. LaBorde, D. M. Sheehan, R. Derynck, and H. L. Moses, Expression of a truncated, kinasedefective TGF-beta type II receptor in mouse skeletal tissue promotes terminal chondrocyte differentiation and osteoarthritis. J. Cell Biol. 139, 541 – 552 (1997). A. Erlebacher and R. Derynck, Increased expression of TGF-beta 2 in osteoblasts results in an osteoporosis-like phenotype. J. Cell Biol. 132, 195 – 210 (1996). H. Zhang and A. Bradley, Mice deficient for BMP2 are nonviable and have defects in amnion/chorion and cardiac development. Development 122, 2977 – 2986 (1996). K. A. Lawson, N. R. Dunn, B. A. Roelen, L. M. Zeinstra, A. M. Davis, C. V. Wright, J. P. Korving, and B. L. Hogan, BMP4 is required for the generation of primordial germ cells in the mouse embryo. Genes Dev. 13, 424 – 436 (1999). G. Winnier, M. Blessing, P. A. Labosky, and B. L. Hogan, Bone morphogenetic protein-4 is required for mesoderm formation and patterning in the mouse. Genes Dev. 9, 2105 – 2116 (1995). E. E. Storm, T. V. Huynh, N. G. Copeland, N. A. Jenkins, D. M. Kingsley, and S. J. Lee, Limb alterations in brachypodism mice due to mutations in a new member of the TGF beta-superfamily. Nature 368, 639 – 643 (1994). A. T. Dudley, K. M. Lyons, and E. J. Robertson, A requirement for bone morphogenetic protein-7 during development of the mammalian kidney and eye. Genes Dev. 9, 2795 – 2807 (1995). G. Luo, C. Hofmann, A. L. Bronckers, M. Sohocki, A. Bradley, and G. Karsenty, BMP-7 is an inducer of nephrogenesis, and is also required for eye development and skeletal patterning. Genes Dev. 9, 2808 – 2820 (1995). Y. Mishina, A. Suzuki, N. Ueno, and R. R. Behringer, Bmpr encodes a type I bone morphogenetic protein receptor that is essential for gastrulation during mouse embryogenesis. Genes Dev. 9, 3027 – 3037 (1995). S. E. Yi, A. Daluiski, R. Pederson, V. Rosen, and K. M. Lyons, The type I BMP receptor BMPRIB is required for chondrogenesis in the mouse limb. Development 127, 621 – 630 (2000). S. R. Rittling, H. N. Matsumoto, M. D. McKee, A. Nanci, X. R. An, K. E. Novick, A. J. Kowalski, M. Noda, and D. T. Denhardt, Mice lacking osteopontin show normal development and bone structure but display altered osteoclast formation in vitro. J. Bone Miner. Res. 13, 1101 – 1111 (1998). A. D. Bradshaw, A. Francki, K. Motamed, C. Howe, and E. H. Sage, Primary mesenchymal cells isolated from SPARC-null mice exhibit altered morphology and rates of proliferation. Mol. Biol. Cell 10, 1569 – 1579 (1999). A. M. Delany, M. Amling, M. Priemel, C. Howe, R. Baron, and E. Canalis, Osteopenia and decreased bone formation in osteonectindeficient mice. J. Clin. Invest. 105, 915 – 923 (2000). T. R. Kyriakides, Y. H. Zhu, L. T. Smith, S. D. Bain, Z. Yang, M. T. Lin, K. G. Danielson, R. V. Iozzo, M. LaMarca, C. E. McKinney, E. I. Ginns, and P. Bornstein, Mice that lack thrombospondin 2 display connective tissue abnormalities that are associated with disordered collagen fibrillogenesis, an increased vascular density, and a bleeding diathesis. J. Cell Biol. 140, 419 – 430 (1998). K. D. Hankenson, S. D. Bain, T. R. Kyriakides, E. A. Smith, S. A. Goldstein, and P. Bornstein, Increased marrow-derived osteoprogenitor cells and endosteal bone formation in mice lacking thrombospondin 2. J. Bone Miner. Res. 15, 851 – 862 (2000). K. G. Danielson, H. Baribault, D. F. Holmes, H. Graham, K. E. Kadler, and R. V. Iozzo, Targeted disruption of decorin leads to abnormal collagen fibril morphology and skin fragility. J. Cell Biol. 136, 729 – 743 (1997).
70 673. K. G. Waymire, J. D. Mahuren, J. M. Jaje, T. R. Guilarte, S. P. Coburn, and G. R. MacGregor, Mice lacking tissue non-specific alkaline phosphatase die from seizures due to defective metabolism of vitamin B-6. Nature Genet. 11, 45 – 51 (1995). 674. S. Narisawa, N. Frohlander, and J. L. Millan, Inactivation of two mouse alkaline phosphatase genes and establishment of a model of infantile hypophosphatasia. Dev. Dyn. 208, 432 – 446 (1997). 675. K. N. Fedde, L. Blair, J. Silverstein, S. P. Coburn, L. M. Ryan, R. S. Weinstein, K. Waymire, S. Narisawa, J. L. Millan, G. R. MacGregor, and M. P. Whyte, Alkaline phosphatase knock-out mice recapitulate the metabolic and skeletal defects of infantile hypophosphatasia. J. Bone Miner. Res. 14, 2015 – 2026 (1999). 676. T. H. Vu, J. M. Shipley, G. Bergers, J. E. Berger, J. A. Helms, D. Hanahan, S. D. Shapiro, R. M. Senior, and Z. Werb, MMP9/gelatinase B is a key regulator of growth plate angiogenesis and apoptosis of hypertrophic chondrocytes. Cell 93, 411 – 422 (1998). 677. K. P. McHugh, K. Hodivala-Dilke, M. H. Zheng, N. Namba, J. Lam, D. Novack, X. Feng, F. P. Ross, R. O. Hynes, and S. L. Teitelbaum, Mice lacking beta3 integrins are osteosclerotic because of dysfunctional osteoclasts. J. Clin. Invest 105, 433 – 440 (2000). 678. C. Deng, A. Wynshaw-Boris, F. Zhou, A. Kuo, and P. Leder, Fibroblast growth factor receptor 3 is a negative regulator of bone growth. Cell 84, 911 – 921 (1996). 679. J. S. Colvin, B. A. Bohne, G. W. Harding, D. G. McEwen, and D. M. Ornitz, Skeletal overgrowth and deafness in mice lacking fibroblast growth factor receptor 3. Nature Genet. 12, 390 – 397 (1996). 680. R. Shiang, L. M. Thompson, Y. Z. Zhu, D. M. Church, T. J. Fielder, M. Bocian, S. T. Winokur, and J. J. Wasmuth, Mutations in the transmembrane domain of FGFR3 cause the most common genetic form of dwarfism, achondroplasia. Cell 78, 335 – 342 (1994). 681. T. P. Yamaguchi, K. Harpal, M. Henkemeyer, and J. Rossant, fgfr-1 is required for embryonic growth and mesodermal patterning during mouse gastrulation. Genes Dev. 8, 3032 – 3044 (1994). 682. C. X. Deng, A. Wynshaw-Boris, M. M. Shen, C. Daugherty, D. M. Ornitz, and P. Leder, Murine FGFR-1 is required for early postimplantation growth and axial organization. Genes Dev. 8, 3045 – 3057 (1994). 683. E. C. Weir, W. M. Philbrick, M. Amling, L. A. Neff, R. Baron, and A. E. Broadus, Targeted overexpression of parathyroid hormone-related peptide in chondrocytes causes chondrodysplasia and delayed endochondral bone formation. Proc. Natl. Acad. Sci. USA 93, 10240 – 10245 (1996). 684. G. Zhao, M. C. Monier-Faugere, M. C. Langub, Z. Geng, T. Nakayama, J. W. Pike, S. D. Chernausek, C. J. Rosen, L. R. Donahue, H. H. Malluche, J. A. Fagin, and T. L. Clemens, Targeted overexpression of insulin-like growth factor I to osteoblasts of transgenic mice: Increased trabecular bone volume without increased osteoblast proliferation. Endocrinology 141, 2674 – 2682 (2000). 685. C. Chiang, Y. Litingtung, E. Lee, K. E. Young, J. L. Corden, H. Westphal, and P. A. Beachy, Cyclopia and defective axial patterning in mice lacking Sonic hedgehog gene function. Nature 383, 407 – 413 (1996). 686. B. Li, S. Boast, S. K. de los, I. Schieren, M. Quiroz, S. L. Teitelbaum, M. M. Tondravi, and S. P. Goff, Mice deficient in Abl are osteoporotic and have defects in osteoblast maturation. Nature Genet. 24, 304 – 308 (2000). 687. M. L. Fero, M. Rivkin, M. Tasch, P. Porter, C. E. Carow, E. Firpo, K. Polyak, L. H. Tsai, V. Broudy, R. M. Perlmutter, K. Kaushansky, and J. M. Roberts, A syndrome of multiorgan hyperplasia with features of gigantism, tumorigenesis, and female sterility in p27(Kip1)deficient mice. Cell 85, 733 – 744 (1996). 688. H. Kiyokawa, R. D. Kineman, K. O. Manova-Todorova, V. C. Soares, E. S. Hoffman, M. Ono, D. Khanam, A. C. Hayday, L. A. Frohman, and A. Koff, Enhanced growth of mice lacking the cyclindependent kinase inhibitor function of p27(Kip1). Cell 85, 721 – 732 (1996).
LIAN AND STEIN 689. A. Okawa, I. Nakamura, S. Goto, H. Moriya, Y. Nakamura, and S. Ikegawa, Mutation in Npps in a mouse model of ossification of the posterior longitudinal ligament of the spine. Nature Genet. 19, 271 – 273 (1998). 690. D. E. Cash, C. B. Bock, K. Schughart, E. Linney, and T. M. Underhill, Retinoic acid receptor alpha function in vertebrate limb skeletogenesis: a modulator of chondrogenesis. J Cell Biol. 136, 445 – 457. 691. A. E. Grigoriadis, Z. Q. Wang, and E. F. Wagner, Fos and bone cell development: Lessons from a nuclear oncogene. Trends Genet. 11, 436 – 441 (1995). 692. A. E. Grigoriadis, Z.-Q. Wang, M. G. Cecchini, W. Hofstetter, R. Felix, H. A. Fleisch, and E. F. Wagner, c-Fos: A key regulator of osteoclast-macrophage lineage determination and bone remodeling. Science 266, 443 – 448 (1994). 693. K. Matsuo, W. Jochum, J. M. Owens, T. J. Chambers, and E. F. Wagner, Function of Fos proteins in bone cell differentiation. Bone 25, 141 (1999). 694. M. Schreiber, C. Poirier, A. Franchi, R. Kurzbauer, J. L. Guenet, G. F. Carle, and E. F. Wagner, Structure and chromosomal assignment of the mouse fra-1 gene, and its exclusion as a candidate gene for oc (osteosclerosis). Oncogene 15, 1171 – 1178 (1997). 695. K. Iida, H. Koseki, H. Kakinuma, N. Kato, Y. Mizutani-Koseki, H. Ohuchi, H. Yoshioka, S. Noji, K. Kawamura, Y. Kataoka, F. Ueno, M. Taniguchi, N. Yoshida, T. Sugiyama, and N. Miura, Essential roles of the winged helix transcription factor MFH-1 in aortic arch patterning and skeletogenesis. Development 124, 4627 – 4638 (1997). 696. M. Barna, N. Hawe, L. Niswander, and P. P. Pandolfi, Plzf regulates limb and axial skeletal patterning. Nature Genet. 25, 166 – 172 (2000). 697. A. M. Reimold, M. J. Grusby, B. Kosaras, J. W. Fries, R. Mori, S. Maniwa, I. M. Clauss, T. Collins, R. L. Sidman, M. J. Glimcher, and L. H. Glimcher, Chondrodysplasia and neurological abnormalities in ATF-2-deficient mice. Nature 379, 262 – 265 (1996). 698. S. Murakami, M. Kan, W. L. McKeehan, and B. de Crombrugghe, Up-regulation of the chondrogenic Sox9 gene by fibroblast growth factors is mediated by the mitogen-activated protein kinase pathway. Proc. Natl. Acad. Sci. USA 97, 1113 – 1118 (2000). 699. B. K. Zehentner, C. Dony, and H. Burtscher, The transcription factor Sox9 is involved in BMP-2 signaling. J. Bone Miner. Res. 14, 1734 – 1741 (1999). 700. M. M. Hurley, S. Tetradis, Y. F. Huang, J. Hock, B. E. Kream, L. G. Raisz, and M. G. Sabbieti, Parathyroid hormone regulates the expression of fibroblast growth factor-2 mRNA and fibroblast growth factor receptor mRNA in osteoblastic cells. J. Bone Miner. Res. 14, 776 – 783 (1999). 701. M. J. Bitgood and A. P. McMahon, Hedgehog and BMP genes are coexpressed at many diverse sites of cell-cell interaction in the mouse embryo. Dev. Biol. 172, 126 – 138 (1995). 702. M. Iwasaki, A. X. Le, and J. A. Helms, Expression of indian hedgehog, bone morphogenetic protein 6 and gli during skeletal morphogenesis. Mech. Dev. 69, 197 – 202 (1997). 703. Y. Zhang, X. Zhao, Y. Hu, T. St. Amand, M. Zhang, R. Ramamurthy, M. Qiu, and Y. Chen, Msx1 is required for the induction of Patched by Sonic hedgehog in the mammalian tooth germ. Dev. Dyn. 215, 45 – 53 (1999). 704. K. Zhao, W. Wang, O. J. Rando, Y. Xue, K. Swiderek, A. Kuo, and G. R. Crabtree, Rapid and phosphoinositol-dependent binding of the SWI/SNF-like BAF complex to chromatin after T lymphocyte receptor signaling. Cell 95, 625 – 636 (1998). 705. X. Xu, M. Weinstein, C. Li, M. Naski, R. I. Cohen, D. M. Ornitz, P. Leder, and C. Deng, Fibroblast growth factor receptor 2 (FGFR2)mediated reciprocal regulation loop between FGF8 and FGF10 is essential for limb induction. Development 125, 753 – 765 (1998).
71
CHAPTER 2 Osteoblast Biology 706. W. C. Yeh, Z. Cao, M. Classon, and S. L. McKnight, Cascade regulation of terminal adipocyte differentiation by three members of the C/EBP family of leucin zipper proteins. Genes Dev. 9, 168 – 181 (1995). 707. S. O. Freytag, D. L. Paielli, and J. D. Gilbert, Ectopic expression of the CCAAT/enhancer-binding protein alpha promotes the adipogenic program in a variety of mouse fibroblastic cells. Genes Dev. 8, 1654 – 1663 (1994). 708. P. Tontonoz, E. Hu, and B. M. Spiegelman, Stimulation of adipogenesis in fibroblasts by PPAR gamma 2, a lipid-activated transcription factor. Cell 79, 1147 – 1156 (1994). 709. E. Hu, P. Tontonoz, and B. M. Spiegelman, Transdifferentiation of myoblasts by the adipogenic transcription factors PPAR gamma and C/EBP alpha. Proc. Natl. Acad. Sci. USA 92, 9856 – 9860 (1995). 710. D. J. Rickard, M. Kassem, T. E. Hefferan, G. Sarkar, T. C. Spelsberg, and B. L. Riggs, Isolation and characterization of osteoblast
711. 712.
713.
714.
precursor cells from human bone marrow. J. Bone Miner. Res. 11, 312 – 324 (1996). N. Lenny, S. Meyers, and S. Hiebert, Functional domains of the t(8;21) fusion protein, AML-1/ETO. Oncogene 11, 1761 – 1769 (1995). S. Meyers, N. Lenny, W.-H. Sun, and S. W. Hiebert, AML-2 is a potential target for transcriptional regulation by the t(8;21) and t(12;21) fusion proteins in acute leukemia. Oncogene 13, 303 – 312 (1996). Y. Ito, and S. C. Bae, The runt domain transcription factor, PEBP2/CBF, and its involvement in human leukemia. In “Oncogenes as Transcriptional Regulators” (M. Yaniv and J. Ghysdael, eds.), Vol. 2, pp. 107 – 132. Birkhauser Verlag, Basel, Switzerland. T. Kanno, Y. Kanno, L. F. Chen, E. Ogawa, W. Y. Kim, and Y. Ito, Intrinsic transcriptional activation-inhibition domains of the polyomavirus enhancer binding protein 2/core binding factor alpha subunit revealed in the presence of the beta subunit. Mol. Cell Biol. 18, 2444 – 2454 (1998).
CHAPTER 3
Osteoclast Biology F. PATRICK ROSS AND STEVEN L. TEITELBAUM Department of Pathology, Washington University School of Medicine, St. Louis, Missouri 63110
I. II. III. IV. V.
VI. Mechanisms of Bone Resorption VII. Humoral Regulation of Osteoclastic Bone Resorption VIII. Diseases of the Osteoclast References
Introduction Osteoclast Morphology Origin of Osteoclasts Models of Osteoclast Function Osteoclast Attachment and Polarization
I. INTRODUCTION
polykaryons generally juxtaposed to bony surfaces and often in resorption pits (Howship’s lacunae) (Fig. 1). These excavations are the products of the capacity of the osteoclast to degrade both organic and inorganic matrices of bone, an event in which mineral removal precedes collagenolysis [1]. Consequently, resorption pits are lined by demineralized collagen, which has not yet been degraded, a phenomenon best appreciated by scanning electron microscopy [2] (Fig. 2). All forms of adult osteoporosis reflect enhanced bone resorption relative to formation and should be viewed in the context of the remodeling cycle. Bone remodeling, a process characterized by the coupling of osteoclast and osteoblast recruitment, occurs throughout life in thousands of sites within the human skeleton. While the fundamental purpose of bone remodeling is unknown, it probably serves to replace effete, aged bone with that which is newly synthesized. Remodeling is initiated by osteoclasts, or their precursors, attaching to trabecular or endosteal bone surfaces. The mechanism by which the osteoclast binds to bone has been a focus of intense investigation and its recent unraveling promises to yield novel approaches to osteoporosis prevention. In this regard, it has been hypothesized that a thin layer of unmineralized type I collagen covering the surface of bone must be removed prior to osteoclast attachment [3]. Degradation of this surface collagen is proposed to be the
Osteoporosis, regardless of etiology, always represents enhanced bone resorption relative to formation. Thus, insights into the pathogenesis of this disease, and progress in its prevention and/or cure, depend on understanding the mechanisms by which bone is degraded. The osteoclast is the principal, if not exclusive, resorptive cell of bone. It is a member of the monocyte/ macrophage family, and most successful treatments of osteoporosis, to date, target osteoclastic bone resorption. Major advances have been witnessed in delineating the mechanisms by which osteoclasts are generated from their precursors and stimulated to degrade bone once they are fully differentiated. In fact, the essential osteoclastogenic molecules are now in hand and, most importantly, are the targets of rational antiosteoporosis drug design. These potential therapeutic agents reflect advances made in understanding osteoclast biology.
II. OSTEOCLAST MORPHOLOGY The osteoclast is a multinucleated cell whose capacity to degrade hard tissues depends on cell/matrix contact. Thus, when viewed in histological sections, osteoclasts appear as
OSTEOPOROSIS, SECOND EDITION VOLUME 1
73
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
74
ROSS AND TEITELBAUM
FIGURE 1 Osteoclasts (arrows) in resorption bay (Howship’s lacuna). Resorptive cells are juxtaposed to bone, and their nuclei are polarized toward the antiresorptive plasma membrane (nondecalcified, Goldner stain). consequence of secretion of neutral collagenase by osteoblasts [4]. While provocative, this hypothesis is unproven and, in fact, the putative layer of lining collagen may be an artifact of tissue fixation.
FIGURE 2
Upon attaching to a nascent remodeling site, osteoclasts polarize their resorptive machinery (infra vide) toward the cell/bone interface and generate a Howship’s lacuna approximately 50 m deep. Once the degradative phase of
Scanning electron micrograph of resorption pit formed on bone slice by isolated avian osteoclast. Bundles in the bottom of the pit represent bone collagen that the cell has demineralized in preparation for degradation.
CHAPTER 3 Osteoclast Biology
the remodeling cycle is complete, osteoclasts are replaced in the resorptive bay by mononuclear cells of unknown origin, which in turn are replaced by cells of osteoblastic lineage. Histological sections of normal human osteoclasts contain varying numbers of nuclei generally maximizing at 10. Nuclear number may mirror osteoclast activity. For example, osteoclasts in Paget’s disease, a disorder of accelerated resorption, are enormous and often contain as many as 100 nuclei [5]. Alternatively, hypernucleated osteoclasts are also encountered in circumstances of resorptive dysfunction, such as various forms of osteopetrosis. Whether the abundant osteoclast nuclei seen in some osteopetrotic patients reflect increased levels of osteoclastogenic agents, such as parathyroid hormone, or an abnormality intrinsic to the osteoclast per se is not known. Osteoclast nuclei are distinct from one another, a feature useful in distinguishing the cell from the megakaryocyte. Moreover, osteoclast nuclei are typically polarized away from the plasma membrane juxtaposed to bone, residing closest to the antiresorptive surface of the cell. Formation of the osteoclast polykaryon probably depends on matrix attachment. Furthermore, osteoclast multinucleation is not the product of endomitosis, as mitotic figures are rarely encountered. However, when placed in culture with mononuclear phagocytes, osteoclasts incorporate new nuclei and expel others [6]. Thus, the most compelling hypothesis holds that once an osteoclast, or its progenitor, attaches to a putative resorptive site, additional mononuclear precursors that fuse with the immobilized cell are recruited. While the life span of osteoclasts is unknown, they clearly undergo programmed cell death. Increased numbers of apoptotic osteoclasts appear when exposed to specific families of bisphosphonates, particularly those that modulate the mevalonate pathway [7]. Alternatively, agents such as interleukin-1 (IL-1), M-CSF, and RANK-ligand (RANKL) extend the cell’s survival (vide infra). Osteoclasts are rich in lysosomal enzymes, with the most studied being tartrate-resistant acid phosphatase (TRAP). While presence of this enzyme in histological sections serves to identify osteoclasts, the bone isoform of TRAP is also expressed by macrophages derived from other organs, most notably lung and spleen [8,9]. At first glance these data would suggest that TRAP is of little use in identifying osteoclasts formed in culture. Alternatively, Suda and colleagues generated bona fide osteoclasts in vitro from alveolar and splenic macrophages, and the same is true regarding circulating human monocytes [10], suggesting that TRAP-expressing cells resident in these tissues are, in fact, capable of osteoclastogenesis [11]. Similarly, the antiosteoclastogenic cytokine interleukin-4 (IL-4) prompts appearance of TRAP-negative polykaryons in culture, thereby associating resorptive activity with expression of
75 the enzyme [12]. Thus, while osteoclasts generated in vitro from macrophages invariably express TRAP, confirmation of their identity requires demonstration of other phenotypic features, preferably the capacity to resorb bone. Because osteoclasts are enormous, often larger than 100 m in diameter, histological sections, typically 5 to 10 m thick, accommodate only a small proportion of each cell. Reflecting the size and complexity of the osteoclast, a given section may contain apparently separate fragments of the same cell, generally considered distinct osteoclasts by bone histomorphometrists. Thus, so-called mononuclear osteoclasts may represent a portion of a multinucleated cell. The likelihood of multiple fragments of the same cell in a single slide underscores the limitations of attempting to distinguish altered numbers of osteoclasts from changes in size or plasma membrane complexity. Being a member of the monocyte/macrophage family, osteoclasts share a number of morphological features with foreign body giant cells, such as abundant lysosomes. However, the unique capacity to resorb bone endows the osteoclast with morphological features distinct from those of other related macrophage-like cells. For example, unlike foreign body giant cells, osteoclasts are rich in mitochondria and, in fact, probably represent the cell containing the greatest density of these organelles [13]. Clearly, however, the morphological feature of the osteoclast that most dramatically distinguishes it from related cells is its ruffled membrane (Fig. 3). This complex infolding of plasma membrane, polarized to the cell/bone interface, is best appreciated by transmission electron microscopy. The ruffled membrane is unique to the osteoclast and matrix specific, appearing only when the cell is in contact with bone. Because of the enormity of an osteoclast, its plasma membrane may touch bone in more than one location and thus separate sites of ruffling will develop within the same cell. Moreover, the extent of ruffled membrane formation appears to parallel the degradative activity of the cell and its exposure to resorptive agonists or antagonists [14]. In fact, as osteoclasts alternate between their resorptive/adherent and mobile/detached states, they respectively express and lose their ruffled membranes. While yet to be visualized in situ, ruffled membrane formation probably represents the insertion of proton pumpbearing vesicles into the plasma membrane, a process similar to exocytosis (Fig. 4). In fact, the small GTPase, Rab-3, which modulates the fusion of exocytic vesicles to the plasma membrane, may also regulate ruffled membrane formation [15]. The ruffled membrane is structurally distinct from the nonresorptive plasma membrane in that it contains abundant, “spike-like” structures shown by light [16] (Fig. 5, see also color plate) and ultrastructural immunocytochemistry to represent vesicular proton pumps. These same projections are present in acidifying vesicles within the cytoplasm. This
76
ROSS AND TEITELBAUM
FIGURE 3
Transmission electron micrograph of osteoclast attached to bone (B). The convoluted infolding of plasmalemma is the ruffled membrane (R). The sealing or “clear” zone (C) is rich in actin filaments oriented perpendicular to the bone surface. Note the abudance of mitochondria (arrows).
observation supports the hypothesis that these proton pumpbearing vacuoles are precursors of the ruffled membrane. While its role, if any, in the resorptive process is not clear, TRAP activity also polarizes to the ruffled membrane and, like other lysosomal enzymes, is secreted into the resorptive microenvironment [17], where it dephosphorylates osteopontin, thereby altering the adhesive properties of this important bone matrix protein [18]. Osteoclasts contain a prominent cytoskeleton, components of which are critical to resorptive function. Microtubules extend from organizing centers to the periphery of the cell. These structures polymerize on cell/substrate adherence and, in so doing, acquire the capacity to transport vesicles [19]. Thus, microtubules are likely to play a central role in osteoclast polarization, delivering proton pump-containing vacuoles for insertion into the resorptive membrane [20] (Fig. 6). The intimacy between osteoclasts and bone required for resorption is reflected, in the cell, by the matrix attachment or “sealing” zone (Fig. 3). This distinct morphological entity, organized as a ring, completely surrounds the ruffled membrane. While organelle free, and thus also referred to as the “clear zone,” this attachment area is, in actuality, rich in microfilaments polarized perpendicularly to the bone surface [21]. The sealing zone is a well-demarcated structure, not unique to the osteoclast, and is also formed
when other members of the macrophage/monocyte family attach to the matrix. Like microtubules, microfilaments organize upon adherence to bone, and the filamentous distribution of F-actin directly correlates with the resorptive activity of osteoclasts [22] (Fig. 7, see also color plate). In fact, when osteoclasts are in the process of degrading bone, Factin localizes in a ring-like structure of punctate plasma membrane protrusions known as focal adhesions or podosomes [23] (Fig. 8). In addition to F-actin, these structures contain matrix-recognizing integrins and proteins such as vinculin and talin, which link these attachment heterodimers to the cytoskeleton [24]. Thus, actin filaments in osteoclasts probably anchor the plasma membrane to the cytoskeleton and are central to the means by which osteoclasts recognize bone. Following formation of a resorptive lacuna, the cell becomes motile, an event attended by dissolution of the actin ring. Evidence in hand indicates that the cyclical mobilization of the actin cytoskeleton in osteoclasts is under the aegis of the Rho family of GTPases [25]. While the magnitude of podosome expression appears to mirror the degradative activity of the cell [26], the punctate appearance of these attachment structures suggests that they do not serve as the osteoclast’s “tight seal.” Unraveling the molecular mechanisms serving to isolate the resorptive microenvironment
77
CHAPTER 3 Osteoclast Biology
FIGURE 4
Model of ruffled membrane formation. In the nonattached state the osteoclast is unpolarized, with acidifying vesicles distributed throughout its cytoplasm. Once in contact with bone, matrix-derived signals, probably mediated via integrins such as v3, prompt targeting of acidifying vesicles to the apical (resorptive) surface of the cell. Insertion of these vesicles into the bone-adjacent plasma membrane yields the characteristic, highly ruffled resorptive surface of the cell. The “spikes” within the vesicles represent the H-ATPase (proton pump).
of the osteoclast is a challenge likely to yield important insights into regulating its activity.
III. ORIGIN OF OSTEOCLASTS Avian osteoclasts (OC) immunostained for vacuolar HATPase (proton pump). (A) Cells are juxtaposed to bone, and the pump (brown-staining reaction product, arrows) is polarized toward the cell – bone interface. (B) Osteoclasts are unattached to bone, and HATPase is distributed diffusely throughout the cytoplasm. Reprinted with permission from Blair et al., Science 245, 855 – 856. Copyright 1989 American Association for the Advancement of Science. (See also color plates).
FIGURE 5
The hematopoietic origin of osteoclasts is now in hand. Its recognition, however, follows a contentious history with initial debate focusing on whether osteoclasts and osteoblasts derive from a common precursor [27]. Early attempts to resolve this controversy involved experiments in which the circulations of two rats were joined. Using this parabiotic approach, Gothlin and Ericsson (reviewed in Ref. 28) established that osteoclasts migrating to a fracture are derived from the blood of its partner. In contrast, osteoblasts are not donor derived, suggesting that their ontogeny differs from that of bone-resorbing cells.
The hematopoietic origin of osteoclasts is also suggested by experiments in which quail cartilaginous limb buds were transplanted onto chick chorioallantoic membranes [29]. The host circulation vascularized the rudiments, which
78
ROSS AND TEITELBAUM
FIGURE 6 Model of osteoclast polarization. In the nonattached state the osteoclast is unpolarized, with acidifying vesicles distributed throughout its cytoplasm. Once in contact with bone, matrix-derived signals, probably mediated via integrins such as v3, prompt targeting by trafficking along microtubules of acidifying vesicles to the apical (resorptive) surface of the cell. Insertion of these vesicles into the bone-adjacent plasma membrane yields the characteristic, highly ruffled resorptive surface of the cell.
FIGURE 8 Isolated murine osteoclast stained with rhodamine phalloidin to delineate F-actin. The fluorescent rings are well-organized sealing zones containing several layers of punctate, focal adhesions, or podosomes.
FIGURE 7 Varying organization of the actin cytoskeleton during the different stages of osteoclast activity. Prior to resorptive phase (stages 1 – 3), actin (red) and vinculin (green) are found in punctate, podosomal structures. During resorption (stage 4), podosomes are lost and actin forms a dense continuous band surrounded by a double ring of vinculin and talin (not shown). During the postresorptive period (stage 5), the system reverts toward its original state. (See also color plate.)
79
CHAPTER 3 Osteoclast Biology
ultimately developed into bone. Most of the osteoclasts present in the neovascularized graft were of chick origin, indicating that they were derived hematopoietically. The hematopoietic origin of osteoclasts, and its distinction from that of osteoblasts, is also established by studies of murine and human osteopetrosis, a disease discussed in detail later (Section VIII). Given the hematopoietic derivation of the osteoclast, efforts turned to identifying its precursor. Because of phenotypic similarities, the cell was suspected of membership in the monocyte/macrophage family. Initial experiments aimed at determining if such is the case involved the injection of thorium dioxide-labeled macrophages into sibling rats who immediately thereafter had a femur fractured. Initially, only labeled monocytes and macrophages were present in the fracture but, ultimately, thorium dioxide-bearing osteoclasts appeared. These observations prompted the search for functional similarities between osteoclasts and macrophages [28]. Early in vitro studies suggested that peripheral blood monocytes and elicited rodent peritoneal macrophages are capable of degrading bone [30 – 32], but it is unlikely that the models used in these experiments are reflective of bona fide resorption. Moreover, despite formation of TRAP-expressing polykaryons by monocytic leukemic lines placed in osteoclastogenic conditions [33], these cells are not authentic osteoclasts. The first successful attempt at generating osteoclasts in vitro utilized long-bone primordia [34]. Culture of these rudiments with embryonic liver and marrow mononuclear cells, but not peripheral blood monocytes or peritoneal macrophages, propagates osteoclasts. Colony-forming unitgranulocyte – macrophage (CFU-GM) or more primitive stem cells are also capable of differentiating into bone resorptive polykaryons. The suggestion that osteoclasts are derived from pluripotent hematopoietic stem cells [35] is supported by experiments in which spleen cells taken from 5-fluorouracil-treated mice were cultured with IL-3 [36]. Blast cell colonies so formed mature into TRAP-expressing polykaryons responsive to calcitonin and are capable of resorbing bone. In contrast, mature macrophages are unable, in this system, to differentiate into osteoclasts. The experiments discussed thus far indicate that isolated, primitive hematopoietic precursors, but not more differentiated macrophages, are capable of osteoclast differentiation in vitro. The studies of Takahashi et al. [37] provided insight into the apparent osteoclastogenic capacity of immature and mature macrophages. In these experiments, mouse marrow cells in the presence of 1,25(OH)2D3 or parathyroid hormone (PTH) formed osteoclasts [38]. Interestingly, most TRAPexpressing polykaryons appeared near alkaline phosphataseexpressing cells, suggesting that osteoblasts play a role in osteoclast differentiation. This group next demonstrated that osteoclasts are induced by coculture of mouse spleen cells and osteoblasts [39]. Furthermore, marrow-derived stromal
cell lines may substitute for osteoblasts in inducing osteoclastogenesis [38,40]. Most importantly, in this coculture system, mature monocytes and macrophages differentiate into osteoclasts [11]. As will be discussed, the means by which stromal cells and osteoblasts promote osteoclastogenesis are in hand. Colony-forming assays, along with osteoclast and macrophage markers, serve to define further the lineage of human osteoclasts. CD34 hematopoietic stem cells, maintained with granulocyte – macrophage colony-stimulating factor (GM-CSF), form CFU-GM colonies [41]. Treatment of these colonies with 1,25(OH)2D3 generates granulocyte, macrophage, and mixed colonies, as well as those consisting of polygonal cells. In defined conditions, only CFUGM and polygonal cell colonies yield osteoclast-like cells, a few of which are capable of forming small resorptive pits. More recent studies, discussed in Section VII, provide further evidence that cells in the myeloid lineage contain osteoclast precursors. Thus, the hematopoietic origin of osteoclasts is established as is the cells’ membership in the monocyte/ macrophage family. Moreover, it is likely that macrophages at various maturational states may differentiate into osteoclasts.
IV. MODELS OF OSTEOCLAST FUNCTION The majority of resorptive experiments in whole animals involve the administration of drugs, systemic hormones, and cytokines (reviewed in Ref. 42). Additional studies include those exploring the effects of weightlessness [43,44], a relevant question in view of the possibility of long-term space travel. Interpretation of these experiments has depended on the quantitation of osteoclast number or net resorptive activity. The advent of densitometric techniques applicable to small animals [45] and of assays for products of rodent bone degradation [46] has increased the usefulness of these models. This approach has yielded important phamacological information, but the possibility that the target for the intervention is not the osteoclast itself, but rather an intermediary cell which modulates the bone resorptive polykaryon, is a major problem confounding whole animal experiments. However, gene deletion techniques have led to invaluable information regarding the molecular mechanisms of osteoclastogenesis in vivo. Fetal mouse [34] and rat bone cultures [47] utilizing intact explanted rudiments facilitate the exploration of paracrine, as opposed to endocrine factors in the resorptive process. The presence of an infinite variety of other cells confounds the direct assessment of osteoclast function. None the less, these systems continue to yield important data regarding specific aspects of osteoclast biology. Bone biopsy has been used as a means of estimating osteoclast activity in humans [48]. In addition to its invasive
80 nature, this approach is limited in that it does not actually measure resorptive activity but provides an estimate of osteoclast number. Noninvasive determinants of human osteoclast function include urinary markers such as hydroxyproline [49] and the dehydropyridinoline moiety [50]. This last approach utilizes a sensitive enzyme-linked immunosorbent assay (ELISA) to detect fragments arising directly from the degradation of mature bone collagen and is probably the current method of choice. Highly enriched or pure populations of osteoclasts would obviate many of the difficulties attending whole animal or organ culture systems. Early attempts utilized enzymatic digestion of bone and/or bone marrow suspensions, coupled with rapid sedimentation yielding small numbers of both mammalian osteoclasts [51 – 54]. The discovery that hens fed a low-calcium diet mobilize abundant osteoclasts [55] permits isolation of these cells in numbers sufficient for biochemical studies, including characterization of an acidic, cathepsin B-like collagenolytic enzyme [56] and a reconstitutible proton pump complex [57]. The same cells, following extensive purification, were used to confirm the presence of pp60c-src in osteoclasts [58], a protein essential to the cells’ polarization [59]. Because avian osteoclasts probably lack the calcitonin receptor, their relationship to mammalian osteoclasts is tenuous. Large numbers of rabbit osteoclasts have been obtained by simple adhesion techniques [60]. Sequencing of the cDNA library generated from these cells confirmed the presence of mRNA for osteopontin and cathepsin K [60,61]. Finally, freshly isolated rat osteoclasts lend themselves to the exploration of intracellular signaling [62,63]. A related strategy, applied to both human and avian systems, is based on immunoisolation of osteoclasts using monoclonal antibodies to suitable surface markers [64 – 66]. Such antibodies were raised to human osteoclastomas and screened for their ability to inhibit bone resorption in vitro. A single clone, 23C6, recognizes the integrin receptor, v3 [67]. This reagent has been used as a blocking antibody and to obtain highly purified polykaryons from which cDNA libraries have been generated [68]. Expression cloning of such a library led to the discovery that annexin II, a putative plasma membrane calcium channel, is secreted by osteoclasts and stimulates their activity [69]. In analogous studies, monoclonal antibodies were raised to purified avian osteoclasts [64]. Such an antibody, 121F, identifies a cell surface protein related structurally to superoxide dismutase [70]. The potential importance of this finding is underscored by the fact that oxygen-free radicals inhibit osteoclastic bone resorption [71]. While the ability to isolate and study mature osteoclasts is essential to understanding the mechanisms by which they degrade bone, the use of differentiated resorptive polykaryons precludes the identification of osteoclast precursors and the biochemical signals regulating osteoclastogenesis. Thus,
ROSS AND TEITELBAUM
efforts have focused on the generation of osteoclasts in vitro. The first successful attempts utilized mononuclear cells derived from calcium-deficient hens known to contain many mature osteoclasts [55]. When these TRAP-negative monocytic cells are cultured, at high density, they fuse within 5 – 6 days, yielding an almost homogeneous population of bone resorptive polykaryons with many, but not all, components of the osteoclast phenotype [2]. The large numbers of both precursors and fused cells allow for performance of a range of biochemical and cell biological studies. Thus, mature cells were characterized with respect to the integrins present and, most importantly, the ability of an antibody against the integrin v3 to block bone resorption [72]. Given that 1,25(OH)2D3 and retinoic acid stimulate osteoclastogenesis in vivo, the precursors served as a means to examine regulation, by these hormones, of this functionally important integrin [73,74]. In fact, the promoter of the 3 gene contains both vitamin D and retinoic acid responsive elements [75,76]. The same generated avian polykaryons have been used to demonstrate a decrease in intracellular calcium concentrations following liganding of the integrin v3 with soluble peptides [77]. Attempts to generate mammalian osteoclasts in vitro have also been rewarding. In early studies, unfractionated marrow cultures, treated with 1,25(OH)2D3, led to the production of small numbers of rabbit, primate, and human osteoclast-like cells [78 – 80]. The cells, most notably those of human origin, exhibit a limited ability to excavate characteristic resorption pits. In contrast, primary murine osteoblasts [39] and several clonal stromal cells lines [11,40], when cocultured with purified murine monocytic precursors, engender greater numbers of multinucleated cells, which resorb bone avidly and express the complete osteoclast phenotype. This system, which depends on contact between stromal cells/osteoblasts and osteoclast precursors, has been used to demonstrate that prostaglandin E2 and IL-4 have opposite effects on osteoclastogenesis [81]. Using the same assay and a function-blocking antibody to macrophage colony-stimulating factor (M-CSF), it was possible to reproduce, in vitro [82], in vivo data concerning the essential nature of this cytokine (M-CSF) for osteoclast production [83,84]. Studies have elucidated the minimal requirements for the generation of murine and human osteoclasts. In short, three proteins are sufficient to induce the differentiation of macrophages into osteoclasts. These osteoclastogenic proteins are M-CSF, receptor for activation of nuclear factor-kB (RANK), and RANK ligand (RANK-L).1 Interestingly, the soluble protein, osteoprotegerin (OPG), competes with RANK as a decoy receptor for RANK-L and thus attenuates osteoclast differentiation. Human bone marrow-derived CD34 cells [85,86] and murine 1 RANK-L is synonymous with osteoprotegerin ligand (OPGL), TRANCE, and osteoclast differentiation factor.
81
CHAPTER 3 Osteoclast Biology
spleen- or bone marrow-derived macrophages serve as osteoclast precursors, targeted by RANK-L and M-CSF [87,88]. RANK-L exists as soluble and membrane-bound forms and is produced by vascular endothelium [89], activated T cells [90], and, probably of most significance to osteoclast biology, those of the osteoblast/stromal lineage. This observation explains the mechanism by which osteoblasts and marrow stromal cells prompt macrophage differentiation into osteoclasts. OPG is also a product of osteoblastic/stromal cells [91], whereas RANK is found on cells of the osteoclast lineage [92,93]. With these proteins in hand, it is now feasible to generate large numbers of highly purified osteoclasts, using as targets, cells of the macrophage/monocyte lineage [88], now known to be osteoclast precursors. This approach has facilitated the execution of a wide range of biochemical and molecular experiments, adding substantially to our understanding of bone biology, a subject discussed further in Section VII.
V. OSTEOCLAST ATTACHMENT AND POLARIZATION As a prerequisite to initiating resorption, the osteoclast must attach to bone and create a functional, polarized plasma membrane. This binding event involves interactions between membrane-bound receptors on the cell surface and matrix proteins in bone. A number of surface markers have been identified on osteoclasts and/or their precursors [94], but, with the exception of integrins, few have been characterized functionally with respect to attachment. Antibody-blocking experiments suggest that members of the 1 family of integrins may participate in the resorptive process, presumably by binding to type I collagen [95]. However, the most compelling data indicate that v3 is the integrin most essential to osteoclast activity. Initial studies implicating this heterodimer in bone degradation were performed in vitro using antibodies to v3 [72] or peptides that block its function [96]. These experiments indicate that the integrin, on osteoclasts, binds one or more bone matrix proteins containing the motif Arg-Gly-Asp (RGD) [72,97 – 100]. The most convincing evidence that v3 is essential to osteoclast function comes, however, from the 3 knockout mouse [101]. This animal has dysfunctional osteoclasts as evidenced by their failure to form actin rings, a normal ruffled membrane or to resorb bone in vitro. Most importantly, these mice, which fail to express v3, are hypocalcemic and become progressively osteosclerotic with age. These studies suggest that v3 is a promising therapeutic target to inhibit accelerated bone resorption as occurs in postmenopausal osteoporosis. In fact, an organic molecule mimicking RGD prevents the profound bone loss rapidly following oophorectomy in the rat [102].
While the precise bone matrix protein(s) recognized by v3 the integrin in vivo is not known, osteopontin is a reasonable candidate. For example, immunoelectron microscopy suggests that v3 and OPN colocalize in bone [103], and an immunopurified antibody to OPN inhibits osteoclast/bone interaction [72]. Perhaps related more directly to polarization, the avian osteoclast cytoskeleton reorganizes upon v3 occupancy by soluble osteopontin [104]. In this instance, the mechanism involves activation of the important intracellular enzyme phosphoinositol-3 kinase, inhibition of which arrests bone resorption in vitro [105]. Activated phosphoinositol-3 kinase generates phosphotidylinositol-3,4 bisphosphate. This phospholipid, by binding to the actin-capping protein gelsolin, causes actin depolymerization and subsequent reelongation [106]. The most compelling evidence that osteopontin is pivotal to osteoclast function is the fact that the osteopontin knockout mouse is protected from oophorectomy-induced osteoporosis [107]. While the v3 integrin is pivotal to the resorptive process, its most important function is probably not formation of the tight seal, but transmission of bone-derived, intracellular signals. Thus, plating of mononuclear cells committed to osteoclast differentiation on a matrix recognized by v3 activates c-src [108], a plasma membrane protooncogene essential for osteoclast polarization and bone resorptive activity [109]. While it has been hypothesized that activated c-src associates with the tyrosine kinase pyk-2, which prompts a cascade of events probably responsible for organizing the osteoclast cytoskeleton during bone resorption [108], this remains to be proven. The dramatic polarization of the osteoclast, a phenomenon probably mediated by v3, distinguishes the cell from other members of the monocyte/macrophage family. It is clear that osteoclast polarization requires bone matrix recognition. The means by which the most dramatic manifestation of such polarization, namely the ruffled membrane, is generated are beginning to emerge. The information in hand suggests that the process involves migration, via microtubules, of acidifying vesicles from the trans-Golgi network to the bone-apposed plasma membrane into which they insert under the aegis of a Rab GTPase [15]. Thus, ruffled membrane formation is a process akin to regulated exocytosis, eventuating in focal complexity of the plasma membrane due to vesicular incorporation. A provocative alternative hypothesis holds that it is the endosomal, rather than exocytic, pathway that contributes to the ruffled membrane [110].
VI. MECHANISMS OF BONE RESORPTION The functional role of the osteoclast is to resorb bone, a composite matrix consisting of both inorganic and organic elements. The inorganic component is largely substituted
82
ROSS AND TEITELBAUM
FIGURE 9
Essential components of osteoclast-mediated bone degradation. See text for details.
hydroxyapatite, whereas the organic phase contains 20 or more proteins, with type 1 collagen the single major species (90% of total protein by weight) [111]. The initial step in bone resorption is attachment of the cell to the matrix, followed by the creation of an isolated, extracellular, resorptive microenvironment bordered by the ruffled membrane. Events leading to attachment and formation of the resorptive space have been discussed earlier. Studies with isolated osteoclasts reveal that dissolution of the inorganic phase of bone precedes that of protein [1]. Demineralization involves acidification of the extracellular microenvironment [112] (Fig. 9), a process mediated by a vacuolar H-ATPase in the ruffled membrane of the polarized cell [113] (Fig. 5). The structure and functional activity of this multienzyme complex are very similar, if not identical, to that of the analogous proton pump in the intercalated cell of the kidney [114,115]. This pump is a multimer, in which only some of the units are intrinsic membrane proteins. Others, including the 70-kDa protein containing ATPase activity, are attached, noncovalently, to subunits buried in the membrane. It is possible that one or more subunits may be present as an isotype, producing an osteoclast-specific form of the pump [116]. In support of this complex being the critical moiety in osteoclast acidification, the fungal metabolite bafilomycin A, a potent and specific inhibitor of all vacuolar proton
pumps, blocks bone resorption [117]. The intact proton pump has been isolated from avian osteoclasts, and the identity of several subunits to those present in other vacuolar pumps was established by Western analysis [57]. Importantly, the activity of the isolated complex is restored by the incorporation into lipid vesicles [57]. As with all members of the vacuolar pump family, extrusion of a proton through the plasma membrane is accompanied by the hydrolysis of one molecule of ATP. Given the fact the osteoclast transports protons extracellularly by an electrogenic mechanism raises the issue of maintenance of intracellular pH. Turning to the paradigm of the renal intercalated cell, Teti et al. [118] found that osteoclasts express, on their antiresorptive border, an energy-independent Cl/HCO3 exchanger similar to band 3 of the erythrocyte. Finally, electroneutrality is preserved by a plasma membrane Cl channel, charge coupled to the HATPase, resulting in the secretion of HCl into the resorptive microenvironment [119,120]. Because blockade of the Cl channel arrests H secretion, it is likely that impaired anion transport also impedes bone resorption. Acidification of the isolated resorptive environment prompts mineral mobilization as well as subsequent solubilization of the organic phase of bone [1], the products of which are endocytosed by the osteoclast, and transported to, and released at its antiresorptive surface [121,122].
83
CHAPTER 3 Osteoclast Biology
Given the low pH at the osteoclast/bone interface, the lysosomal family of acidic hydrolases, delivered to the resorptive microenvironment via the mannose-6-phosphate receptor [123,124], presented themselves as candidate molecules to degrade bone collagen and noncollagenous proteins. In fact, there is compelling evidence that, in the mammalian osteoclast, cathepsin K fulfills this role. Cathepsins have the capacity to fragment bone organic matrix [56] and, most importantly, mice deleted of the cathepsin K gene have increased skeletal mass [125]. Thus, cathepsin K is a potential antiosteoporosis target. Several studies reveal a possible role for neutral collagenases in the early phase of bone resorption. Bone is believed to be covered by a thin layer of osteoid (unmineralized matrix), which, some posit, needs be removed prior to osteoclastic degradation [4]. This hypothesis is, however, controversial, and the osteoblast is proposed to be the source of the putative neutral collagenase in this circumstance [126 – 129]. The most compelling evidence that these enzymes participate in the resorptive process comes from the demonstration that mutation in vivo of the site in type 1 collagen targeted by neutral collagenases attenuates bone resorption [130]. In addition to degrading bone, the osteoclast also resorbs mineralized cartilage in the hypertrophic zone of the growth plate. There is little evidence that the so-called “chondroclast” is distinct from the osteoclast; in fact, cartilage resorption by the cell is mediated via matrix metalloproteinase-9 [131]. While the means by which osteoclasts degrade bone are reasonably well defined, less is known regarding the mechanisms terminating their activity. The most provocative argument holds that sensing a high ambient Ca2 within the resorptive space, by a plasma membrane cation receptor, prompts withdrawal of the osteoclast from the bone surface and terminates resorption [132].
osteoclast recruitment. Injection of RANK-L into mice results in a rapid increase in serum calcium, indicating that the molecule also stimulates the resorptive activity of mature osteoclasts [134]. Targeted deletion of the RANK-L, OPG, and RANK genes prompts phenotypes consistent with the role of each protein as described in the current model. Thus, mice lacking either RANK-L [135] or RANK [136] are severely osteopetrotic secondary to lack osteoclasts, while the absence of OPG results in profound osteoporosis [137]. Binding of M-CSF and RANK-L to their respective receptors, c-fms and RANK, is the necessary and sufficient event to initiate osteoclastogenesis. Discovery of these molecules has yielded insights into the interplay between regulatory humoral and local factors and the intracellular signals controlling formation of the osteoclast (Fig. 10, see also color plate). In brief, a range of hormones, cytokines, and growth factors, targeting primarily to mesenchymederived cells in the bone microenvironment, control the expression of M-CSF, RANK-L, and OPG, with the overall impact on osteoclast recruitment dependent mainly on the ratio of OPG to RANK-L. Finally, the macrophage products, IL-1, TNF, and IL-6, regulate the capacity of stromal cells to promote osteoclast precursor differentiation (reviewed in Ref. 138). These proinflammatory cytokines also directly target myeloid precursors and mature osteoclasts, initiating signals prompting cell fusion and altered function and survival. Treatment of osteoclast precursors with IL-1 accelerates fusion and generates bone-resorbing polykaryons [139]. The cytokine also increases the life span of mature osteoclasts by a mechanism involving the activation of NF-B [140]. A more controversial
VII. HUMORAL REGULATION OF OSTEOCLASTIC BONE RESORPTION As discussed earlier, four molecules, M-CSF, RANK, RANK-L, and OPG, acting in concert, are major regulators of osteoclast formation and function. OPG, a member of the TNF receptor family, inhibits osteoclast formation. RANK-L, a novel member of the TNF superfamily, is the OPG ligand. Whereas RANK-L is primarily membrane bound, it also exists as a soluble form. Not surprisingly, RANK, the receptor for RANK-L, expressed on osteoclasts and their precursors, is a TNF receptor-related protein. The fact that purified macrophages give rise to numerous functional osteoclasts, when treated only with M-CSF and RANK-L [88] and that osteoclastogenesis is blocked by OPG [133], establishes these molecules as central to
FIGURE 10
Regulation of osteoclast formation. A range of hormones and cytokines, targeting to stromal cells, enhance expression of M-CSF and RANK-L, the pivotal osteoclastogenic molecules, and the proinflammatory cytokines IL-1, IL-6, and TNF. These proteins act to stimulate both formation and activity of mature polykaryons. HSC, hematopoietic stem cell. (See also color plate.)
84
FIGURE 11
Key elements of RANK signaling. Receptor ligation is followed by the recruitment of adaptor molecules, including TRAF6, which, interacting with c-src, stimulates the PI-3 K/Akt pathway. Additionally, the NF-B and AP-1 families of transcription factors, key elements in osteoclast formation and function, are activated. Akt plays a role in phosphorylation of the IKK complex [367]. (See also color plate.)
hypothesis holds that TNF induces osteoclast formation in a manner not requiring RANK-L [141]. While the extracellular domain of RANK identifies it as a member of the TNF receptor superfamily, the intracellular amino acid sequence is unlike that of all orthologs. Despite these differences, the known proximal and distal signaling events arising from ligation of RANK and other members of the family are similar (Fig. 11, see also color plate). Two critical distal events in RANK signaling are the activation of JNK and the NF-B transcription complex. Mice lacking both the p50 and p52 NF-B subunits fail to generate osteoclasts [142]. Similarly, the AP-1 family proteins, c-fos and c-jun, are targets of JNK and play a role in osteoclast formation. Thus, animals lacking c-fos generate no multinucleated, bone-resorbing cells [143], nor do precursors which overexpress a form of c-jun containing a mutation that renders the transcription factor nonactivatable by JNK [144]. The proximal RANK-initiated signals are more complex and may reflect those following ligation of other members of the TNF receptor superfamily. For example, activation of the type 1 TNF receptor, which promotes osteoclast formation [145], is followed by binding of TRAFs, adaptor proteins capable of driving a variety of downstream signals [146]. Similarly, overexpressed RANK in embryonic human kidney cells recruits TRAFs 1, 2, 3, and 5 [147]. Because mice lacking TRAF 2, 3, or 5 have normal bones [148 – 150], these molecules may not be necessary for RANK-induced osteoclastogenesis. However, these three TRAFs bind to overlapping sites in the RANK cytoplasmic tail [146] and thus it may be necessary to generate mice lacking combinations of TRAFs 2, 3, and 5 to establish the role of these adaptor proteins in osteoclast function.
ROSS AND TEITELBAUM
Absence of TRAF6 results in osteopetrosis. While this observation raises the possibility that TRAF6 links RANK-L to osteoclastogenesis, TRAF6-/- osteopetrotic mice have numerous osteoclasts, which, however, are unable to attach to bone, form a ruffled membrane, or degrade the matrix [151]. These animals bear a striking resemblance to those lacking c-src [59], a fact explained by the constitutive binding of the protooncogene to RANK and its interaction with TRAF6 following receptor occupancy [152]. The subsequent downstream events (Fig. 11) include stimulation of phosphatidylinositol-3 kinase (which by phosphorylating Akt, initiates cytoskeletal reorganization), antiapoptotic signals, and activation of NF-B [153], all key components of osteoclast function. In murine osteoclasts, RANK-derived signals stimulate JNK1 (but not JNK2) via the MEKMEKK axis (Srivastava et al., unpublished data). In short, the link between RANK activation and the AP-1 and NF-B families of transcription factors suggests these signals represent key components of RANK-induced osteoclast formation. However, the presence of numerous, albeit nonfunctional, osteoclasts in TRAF6 null mice indicates that generation of these cells does not require TRAF6-initiated signals. As will become evident, the majority of the osteoclasttargeting cytokines/hormones/steroids exert their influence, wholly or in part, by their ability to increase or decrease M-CSF, OPG, and RANK-L expression. Thus, if M-CSF production is enhanced, the number of precursors, and their ability to differentiate, is augmented. Similarly, a net excess of RANK-L activity (as a result of decreased OPG secretion, increased RANK-L synthesis, or both) stimulates osteoclast recruitment and function. A net deficiency of RANK-L activity (arising from the reciprocal set of circumstances) will arrests osteoclast differentiation and activity. Finally, a number of cytokines, in addition to controlling formation of osteoclasts, through their effects on OPG and RANK-L expression, also target the mature polykaryon, altering its resorptive capacity. Table 1, which summarizes the role of the various cytokines regulating production of M-CSF, OPG, and RANKL, reveals an apparent conundrum, namely that a given osteoclast agonist can increase both OPG and RANK-L. The net effect for any single molecule is explained by assuming greater potency for the induction of either the osteoclaststimulating or inhibiting moiety. It must be recognized that in vivo, the processes of osteoclast formation and subsequent function are controlled in a continuous and rapid manner, reflecting the net effect of the numerous regulators discussed next.
A. Parathyroid Hormone Parathyroid hormone is a major accelerator of bone remodeling. Because the remodeling process is initiated by
85
CHAPTER 3 Osteoclast Biology
TABLE 1 Summary of Factors that Modulate Expression of RANK-L, OPG, and M-CSFa RANK-L
OPG
M-CSF
q
q
q
q
p
Hormones 1,25(OH)2vitamin D3 Estrogen q
p
TNF
q
q
IL-1
q
q
IL-6
q
IL-11
q
PTH Cytokines
q
modulating cells and osteoclasts is necessary [11,82]. The capacity of PTH to increase OPGL and decrease OPG [167] explains the osteoclastogenic properties of the hormone. Intact PTH and parathyroid hormone-related protein (PTHrP) target the same receptor [168] and exert similar biological effects on bone cells [169]. In contrast, the carboxyl-terminal portion of PTHrP may uniquely dampen osteoclastic bone resorption [170]. It should be pointed out that the biological implications of this phenomenon are controversial, as it occurs only in selected systems [171]. Whether carboxyl-terminal PTHrP targets osteoclasts or exerts its inhibitory effects via intermediary cells is not clear [172].
Growth factors TGF-
p
q q
BMP-2 Others Prostaglandin E2
B. Cytokines and Chemokines
q
p
a Adapted, with additions, from Hofbauer et al. [138]. Data on the role of PTH are found in Lee and Lorenzo [167].
osteoclast recruitment, states of PTH excess are typically associated with morphological and biochemical evidence of increased resorptive activity. Importantly, PTH sensitizes bone to 1,25(OH)2D3 [154]. Thus, it is not surprising that 1,25(OH)2D3, in the parathyroprivic state, does not mobilize osteoclasts [155]. Osteoclasts, in vivo or in organ culture, undergo profound changes when exposed to PTH. The population of these cells grows proportional to PTH concentration and the hormone augments the ruffled membrane area and the number of cells exhibiting this resorptive organelle [21,156]. Clear zone and overall cell size are also enhanced [157]. The fact that PTH administration induces c-fos mRNA expression in osteoclasts [158] indicates that the hormone prompts the differentiation of precursors along an osteoclastic pathway [143]. Taken together, these observations suggest that PTH stimulates osteoclast recruitment and activates differentiated polykaryons. While it had been proposed that the hormone targets osteoclasts directly, high-affinity PTH receptors have not been demonstrated in these cells [159,160], nor is there is convincing evidence that the hormone exerts a direct biological effect on isolated osteoclasts [161,162]. However, osteoblasts [158] and marrow stromal cells [163] from which they are derived are sensitive to PTH. The fact that isolated osteoclasts, cocultured with these cells, respond to PTH indicates that osteoblasts and marrow stromal cells serve as intermediaries in PTH-stimulated, osteoclastic bone resorption [164]. PTH-treated osteoblasts or stromal cells secrete resorptive activity [165,166], but contact between
Cytokines regulating osteoclast biology can be divided broadly into three groups: those that facilitate differentiation, survival, and proliferation of precursors; those that promote osteoclast precursor differentiation, and those that alter the resorptive capacity of the mature polykaryon. As discussed later, a number of cytokines exert more than one function. Some act at both proliferative and differentiation stages, whereas others also target mature osteoclasts, modulating bone degradation. Osteoclast formation and activity are controlled by both systemic hormones and locally produced cytokines/chemokines. The latter molecules are themselves regulated in a complex manner in which paracrine and/or autocrine events play a central role. Proteins such as stem cell factor, IL-1, IL-3, IL-6, GMCSF, M-CSF, erythropoietin, epidermal growth factor, and basic fibroblast growth factor are required for the optimal differentiation of stem cells to CFU-GM [85,173], an early osteoclast precursor [41]. As discussed elsewhere in this chapter, GM-CSF and M-CSF play overlapping/redundant roles as regarding the formation of precursors capable of being induced to form bone-resorbing polykarons. Members of the hematopoietic family of cytokines target osteoclast precursors, stimulating both proliferation and differentiation. Other molecules, produced in the bone microenvironment by marrow stromal cells, immature monocytes, and osteoblasts and their precursors, as well as T lymphocytes [88,174,175], are also capable of modulating osteoclastogenesis by acting in a paracrine and/or autocrine manner to regulate the production of M-CSF, RANK-L, and OPG. Whereas most cytokines promote osteoclast formation by stimulating proliferation and/or differentiation of osteoclast precursors, several, including M-CSF, RANK-L, IL-1, IL-6, and TNF, also directly modulate mature osteoclasts. When binding to early osteoclast precursors, M-CSF provides a signal required for survival, proliferation, and maturation [176,177]. In contrast, the same protein
86 increases the motility of mature osteoclasts, with consequent diminution in bone resorption [178,179]. The cytokines IL-6, IL-11, oncostatin M, and leukocyte inhibitory factor share a common signaling pathway. The heterodimeric signaling complex comprises a common subunit, gp130, linked to a second protein, specific for each cytokine [180]. Whereas IL-6 and IL-11 [181,182] are important regulators of osteoclast function, the roles of oncostatin M and leukocyte inhibitory factor are less clear [183,184]. Mice unable to transduce signals from any member of the gp130 family as a result of targeted deletion of the common subunit have normal bones [185]. This result is not surprising given the multiplicity of pathways whereby control of the critical regulator of osteoclast formation and function, namely RANK-L, can be exercised (Fig. 10). Ligation of IL-6 to its receptor, in the mature osteoclast, promotes bone resorption [68], a process accentuated in Paget’s disease [186], in which the cytokine appears to play an important pathogenetic role [187]. The resorptive properties of IL-6 have been documented with osteoclastomaderived, osteoclast-like cells treated with antisense IL-6 oligonucleotides [188]. An antibody that blocks IL-6 function arrests bone loss in vivo [189,190] and blunts oophorectomy-induced osteoclastogenesis [190,191]. Mice in which the IL-6 gene has been removed by targeted deletion do not lose bone following oophorectomy [192]. Finally, IL-6 has been shown to play a role in PTH-induced bone resorption in vivo [193]. Taken together, these results provide strong support for the hypothesis that IL-6 appears to modulate, in part, the effects of estrogen deprivation on bone resorption (see Chapter 41). A molecular explanation for this is provided by the observation that IL-6, targeting to mature osteoclasts, blunts the inhibitory role of high extracellular calcium by a pathway that involves induction, on the surface of the cell, of ADP-ribosyl cyclase [194,195]. In contrast, Kitazawa et al. [189] found decreased numbers of osteoclasts in both oophorectomized and sham-operated mice administered the anti-IL-6 antibody. Furthermore, this antibody, administered in vivo, decreases the histological evidence of bone resorption but not urinary excretion of pyrodinoline cross-links. A naturally occurring inhibitor of IL-1 action, the IL-1 receptor antagonist is produced by monocytes [196] and dampens osteoclast formation and bone resorption in oophorectomized rats and mice [189,197]. Similarly, TNFbinding protein, a soluble form of the TNF receptor that blunts TNF activity, decreases the number of osteoclasts produced by murine cells in vivo and in vitro [189,198]. Furthermore, mice transgenic for the TNF-binding protein, a soluble TNF receptor, are resistant to oophorectomy-induced bone loss, indicating a central role for this cytokine in the most significant form of osteoporosis [199]. Regulation of cytokine expression as relating to osteoclastogenesis is incompletely understood. As regarding
ROSS AND TEITELBAUM
inductive proteins, IL-1 and TNF stimulate the synthesis of M-CSF and IL-11 by marrow stromal cells [200,201]. The same agonists, as well as IL-6, regulate their own production by monocytes via a complex set of paracrine and autocrine signals [202]. The release of IL-6 by stromal cells is stimulated by PTH [203]. As discussed later, estrogen controls the production of several cytokines central to different aspects of osteoclast biology. There are relatively few reports examining the role of chemokines in bone biology. Macrophage inflammatory protein-1, a product of osteoblasts, stimulates motility but decreases the resorptive activity of isolated osteoclasts [204]. Similarly, IL-8, secreted by osteoclasts [205], acts via an autocrine loop to inhibit bone resorption [206]. The latter finding is in conflict with an earlier report that osteoclast recruitment and matrix degradation increase in parallel with IL-8-induced neoangiogenesis [207].
C. Estrogen Estradiol is critical to skeletal preservation in both women [208] and men [209 – 211], as its denial prompts rapid loss, especially of cancellous bone. The estrogendeprived skeleton, particularly as seen shortly after menopause, exhibits histological features of accelerated remodeling, including abundant osteoclasts and resorption bays [3]. Biochemical evidence of enhanced osteoclastic activity in estrogen-deficient patients includes increased hydroxyproline excretion [212]. Interestingly, tamoxifen, which antagonizes estrogen in other organs, is an estrogen agonist in bone, wherein the effects of both steroids are additive [213,214]. Given the identification of molecular targets of estrogenic activity, the means by which the steroid dampens osteoclast function are now more clearly understood. While the role of estrogen is discussed in greater detail in Chapters 10 and 41, it is pertinent that the sex steroid not only modulates M-CSF, but, by controlling monocytic secretion of IL-1 and TNF, the ratio of OPG to RANK-L as well. Likewise, there are reports that the steroid either does [190] or does not [189,215,216] block the secretion of IL-6 from the same cells. There is, however, agreement that estrogen inhibits the production of IL-6 by monocytes [217 – 220]. Additionally, while the sex steroid dampens the secretion of GM-CSF and IL-1 receptor antagonist from cultured human monocytes [220,221], it does not alter the production of IL-11 by stromal cells [181]. Estrogen also diminishes expression of the type I IL-1 receptor on osteoclast and their precursors, thereby limiting the capacity of the cytokine to target these cells [222]. Finally, in addition to modulating the expression of factors that stimulate osteoclast formation and function, estrogen also induces osteoclast apoptosis, a subject discussed in Section VIII.
87
CHAPTER 3 Osteoclast Biology
D. Vitamin D 1,25(OH)2D3, the biologically active form of vitamin D, is a potent inducer of resorption whether administered in vivo [223,224] or added to bone organ culture [22,225]. The molecular basis of this activity is thought to lie largely in the ability of this secosteroid to stimulate, by osteoblastic/stromal cells, a net increase in RANK-L as opposed to OPG. The molecular basis for this result has been defined by the work of Kitazawa and colleagues [226], who demonstrated a functional vitamin D response element in the murine RANK-L promoter. A secondary component of the activity of 1,25(OH)2D3 is that it increases the production of M-CSF, the only cytokine, besides RANK-L, necessary to produce osteoclasts in vitro from both murine and human precursors [88]. A number of structural modifications of the vitamin D molecule have yielded analogues of 1,25(OH)2D3 with distinct biological properties of potential therapeutic importance. For example, 22-oxacalcitriol, while being an effective inducer of cell maturation, is only 1% as effective as 1,25(OH)2D3 in stimulating osteoclastic bone resorption [227]. The molecular basis of this observation is not clear at this time. Vitamin D analogues are discussed in detail in Chapter 9. The fact that vitamin D receptor-ablated mice in whom normal calcium and phosphate levels had been preserved by dietary means exhibited no differences from their wild-type littermates in bone histomorphometry suggests that the major role of the steroid in bone function is to maintain mineral homeostasis [228]. Finally, by promoting expression of type 1 IL-1 and E2 receptors in marrow-derived stromal cells [229,230], the secosteroid primes them to respond to these agonists.
E. Glucocorticoids Gluocorticoid-induced osteoporosis is among the most devastating forms of osteopenia. It predominantly affects trabecular bone [231,232], being manifest most profoundly in the axial skeleton. The subject is discussed extensively in Chapter 44. While glucocorticoids appear to increase osteoclastic activity in vivo, their effect on resorbing cells is complex. Patients chronically treated with glucocorticoids have increased numbers of osteoclasts [233] and resorption bays [234]. There is also biochemical evidence of accelerated skeletal degradation such as enhanced hydroxyproline excretion [235]. Given their osteoclastogenic properties, it is surprising that the resorptive activity and survival of isolated osteoclasts exposed to these agents are attenuated [236]. This conundrum prompted a hypothesis that glucocorticoids, in low dose, permit osteoclast precursor differentiation, whereas higher concentrations blunt resorption [237]. A more likely explanation
of the apparent discrepancy between in vivo and in vitro effects of glucocorticoids on osteoclasts relates to the steroid’s actions on mesenchymal marrow cells. By simultaneously increasing RANK-L, and decreasing OPG, the net effect is an increase in osteoclast number and resorptive capacity [138]. In addition to these direct effects, glucocorticoids suppress intestinal absorption of calcium [238]. This event, in conjunction with augmented renal calcium loss [239], provokes secondary hyperparathyroidism [233,240]. Furthermore, the resorptive impact of such relatively mild hyperparathyroidism is amplified as glucocorticoids sensitize bone cells to PTH [237] and 1,25(OH)2D3 [241], agents known to stimulate osteoblasts to accelerate osteoclast activity. Given these data, it is not surprising that osteoclasts do not proliferate in parathyroidectomized animals receiving glucocorticoids [242]. Finally, while their effects on PTH responsivity are in keeping with stimulated osteoclastic bone degradation, glucocorticoids also inhibit the production of other molecules known to enhance resorption. For example, the steroid suppresses the expression of IL-1 and IL-6 [243] and blunts the synthesis of prostaglandins [244].
F. Thyroid Hormone Thyroid hormone excess, whether endogenous [232,245] or iatrogenic [246-248], prompts bone loss in humans and experimental animals. The skeletal changes of hyperthyroidism are those of accelerated remodeling with an abundance of osteoclasts, osteoblasts, and osteoid [246,247]. In some instances, one encounters peritrabecular marrow fibrosis (osteitis fibrosa) [246,249]. The effects of thyroid on bone are discussed in Chapter 47. Thyroid hormone stimulates bone resorption in vivo [250,251], in organ culture [252,253], and in isolated cell systems consisting of osteoclasts and osteoblasts [254]. The urine of hyperthyroid patients contains increased amounts of the products of bone degradation, including hydroxyproline [250,255] and pyrodinoline cross-links [256,257]. However, the effect of thyroid hormone on osteoclasts is indirect and, like other stimulatory agents, is mediated by osteoblasts [254]. At this time there are no reports as to the impact of thyroid hormone treatment of osteoblastic/stromal cells on the OPG/RANK-L ratio.
G. Prostaglandins Prostaglandins exert dramatic and diverse effects on osteoclast function. While most is known about the resorptive impact of prostaglandins of the E series (PGE), thromboxane [258] and products of the lipooxygenase pathway [259,260] also modulate the osteoclast (see Chapter 13).
88
ROSS AND TEITELBAUM
The predominant effect of a relatively prolonged exposure of the osteoclast to PGE is enhanced resorption. Interestingly, the resorptive effects of intense mechanical force on osteoclast function may also be prostaglandin mediated [261]. While administration of prostaglandins in vivo has not been shown to accelerate resorption, cyclooxygenase inhibitors dampen bone remodeling [262]. Moreover, PGE potently induces osteoclast activity in organ culture [263]. Once again the ability to increase RANK-L and concomitantly decrease OPG explains the resorptive impact of PGE [264,265]. In contrast to the ability of PGE to stimulate osteoclast formation via upregulation of RANK-L on stromal cells, the eicosinoid inhibits the activity of preformed osteoclasts [81,266]. Isolated osteoclasts, when exposed to PGE, cease resorbing bone [266]. The cells transiently contract in a manner similar to that attending calcitonin exposure [267], a finding consistent with the fact that, like calcitonin, prostaglandins express their biological properties through cAMP generation. In fact, many of the resorptive effects of prostaglandins are mimicked by cell-permeant analogues of the nucleotide [81,268]. Interestingly, while prostaglandins and calcitonin both activate adenylate cyclase, the two agents appear to impact the enzyme differently [269].
H. Calcitonin Calcitonin, when administered to vertebrates, dampens bone resorption by directly targeting osteoclasts [161,270]. Thus, the hormone prompts a rapid decline in circulating Ca2 activity [271] and urinary hydroxyproline excretion [272], phenomena indicative of suppressed bone degradation. Furthermore, addition of calcitonin to any in vitro mammalian osteoclast model blunts the capacity of the cell to degrade bone. In actuality, the failure of calcitonin to suppress resorption in a mammalian system raises concern as to whether the model contains bona fide osteoclasts. However, there is no convincing evidence that endogenously secreted calcitonin directly modulates osteoclast activity (see Chapter 73). Reflecting the rapid biological effects of the hormone, a characteristic feature of mammalian osteoclasts is the expression of calcitonin receptors [273]. In fact, calcitoninbinding sites are the only peptide hormone receptors unequivocally resident in these cells. While avian osteoclasts probably lack calcitonin receptors [268], mammalian polykaryons failing to express them are not osteoclasts. However, calcitonin receptors are not unique to the osteoclast, as they also appear in macrophages derived from nonskeletal sources [274], such as lung [275]. This finding is in keeping with the capacity of various macrophage precursors, including those of alveolar origin, to differentiate
in vitro into polykaryons with the complete osteoclast phenotype [11]. When added to organ culture, calcitonin decreases the number of osteoclasts [276] and nuclei per cell [277]. The ruffled membrane disappears, a process reflecting its internalization and vacuolar conversion [123]. Proteins destined for the ruffled membrane, such as the mannose-6-phosphate receptor, are rerouted to intracellular vesicles [123]. TRAP secretion is altered by calcitonin, transiently increasing but ultimately declining [278]. Similarly, osteopontin mRNA expression by osteoclasts is blunted by the hormone [279]. Within seconds of calcitonin exposure, osteoclasts become immobile and contract [51,280,281], a phenomenon associated with changes in the microtubule [282] and microfilament [267] cytoskeleton (Fig. 12). The arrested motility of the cell has been attributed to two distinct events termed quiescence and retraction, each of which may be mediated by distinct signaling pathways [283]. Prolonged exposure to calcitonin results in a declining sensitivity of osteoclasts to the hormone [284,285]. The escape phenomenon, which reduces the therapeutic efficacy of the hormone, has been attributed to induced PTH secretion [286] and anticalcitonin antibody generation [287]. Takahashi and co-workers [288] have shown, however, that calcitonin downregulates its own receptor in osteoclasts. Calcitonin exposure is followed in the osteoclast by an immediate increase in intracellular Ca2 activity and generation of cyclic AMP [289]. The hormone also activates protein kinase C [287]. These accumulated data suggest that calcitonin mediates a number of intraosteoclastic events via distinct signaling pathways. The effects of PTH and calcitonin are discussed further in Chapter 12.
I. Superoxide and Nitric Oxide Oxygen-derived free radicals and nitric oxide (NO) alter osteoclastic bone resorption dramatically [290]. Superoxide is produced by osteoclasts resorbing bone [71,291,292] and therefore may function in an autocrine manner. These anions are located in the ruffled membrane [290] and are induced when osteoclasts contact skeletal matrix [71]. Osteoclasts generate superoxide via NADPH oxidase [292]. The cells also contain superoxide dismutase [70,293], which, when added to osteoclast cultures, decreases resorptive activity [294]. Superoxide anion generation may also play a role in regulated bone resorption. For example, IL-1 and PTH-induced osteoclast activity is blunted by superoxide dismutase [71], and calcitonin diminishes superoxide generation [71,291]. In contrast to superoxide, hitric oxide is antiresorptive [295,296], and the two molecules may act together to regulate osteoclast function [290]. NO targets directly to osteoclasts [295,296] and, like calcitonin, inhibits spreading
89
CHAPTER 3 Osteoclast Biology
FIGURE 12 Osteoclasts exposed to cyclic AMP-inducing agents, including calcitonin and prostaglandin E, undergo rapid contraction and motility arrest. Here, an untreated, isolated rat osteoclast (A) was exposed for 2 h to 108M dibutyryl cyclic AMP (B), inducing immediate contraction and immobility (phase contrast).
[295]. The molecular mechanisms responsible for the osteoclast-inhibiting effects of NO and CT differ [295]. NO is produced by osteoclasts via NO synthase activity [290]. The enzyme in this cell is both constitutively present and regulated [290]. Thus, like superoxide, NO may autoregulate osteoclast function. However, this short-lived reactive radical is also synthesized in abundance by endothelial cells that are in close proximity to, and may govern the activity of, osteoclasts [295]. Perhaps reflecting their known capacity to modulate osteoclasts, osteoblasts also secrete NO in response to inflammatory cytokines [297,298]. Interestingly, osteopontin dampens NO synthesis by renal cells, an event perhaps mediated through the integrin v3 [299]. Whether the bone matrix protein also has an impact on osteoclastic production of NO is unknown.
J. Osteoclast Apoptosis Net bone resorption represents the sum of osteoclast activity and number, with the latter parameter itself dependent on the rates of cell formation and death. Treatment of mature osteoclasts with estrogen results in the rapid initiation of apoptosis, an event mediated by transforming growth factor [300]. Nitrogen-containing bisphosphonates induce osteoclast death, an event resulting from the blunted synthesis
of farnesyl and geranylgeranyl pyrophosphate, two moieties central to the acylation, and thus activation, of members of the Ras superfamily [301 – 303]. Signaling through Ras proteins modulates cell survival. Separately, bisphosphonates lacking the nitrogen atom stimulate osteoclast death by an undetermined mechanism, but one that does not involve small GTPases [303]. Caspases, a family of cysteine proteases [304], are involved in both the initiation and the execution phases of apoptosis, and these molecules modulate the life span of the osteoclast. Thus, treatment of osteoclast cultures with a pan inhibitor of caspase action, Z-VAD-Fmk, decreases the number of apoptotic cells [305]. Preliminary data indicate the likely involvement of caspases 3 and/or 7, both executor, or downstream caspases [306]. The proximal signals leading to the activation of these distally acting proteases, in osteoclasts, are unknown.
VIII. DISEASES OF THE OSTEOCLAST A. Osteopetrosis The osteopetroses are a group of congenital diseases characterized by increased skeletal mass due to dysfunctional osteoclasts. Despite pathogenetic heterogeneity, all
90
FIGURE 13
Molecular mechanisms of osteopetrosis. See text for details. (See also color plate.)
forms of osteopetrosis are distinguished by radiopaque bones, with loss of differentiation between the cortex and the marrow. The molecular defect of a number of osteopetrotic mutants, particularly of murine origin, are now defined and serve as a rich source of information regarding the origin and the function of osteoclasts (Fig. 13 see also color plate). 1. OSTEOPETROSIS DUE TO DEFECTIVE OSTEOCLAST PROGENITORS Initial evidence regarding the origin of osteoclasts was provided by Walker [307,308] wherein the circulation of a normal mouse was joined to that of an osteopetrotic, microphthalmic (mi/mi), or gray-lethal (gl/gl) littermate. By 6 weeks of age all of the excess mutant skeleton was resorbed. Walker concluded that mature osteoclasts, or their progenitors, were introduced from the normal mouse into the mutant’s marrow and subsequent experiments support this hypothesis. For example, the skeletal lesion of irradiated gl/gl or mi/mi mice is cured by infused, normal marrow or spleen cells. Conversely, mi/mi or gl/gl spleen cells induce osteopetrosis in irradiated normal mice [309]. Together, these experiments provide direct evidence for the hematopoietic origin of osteoclasts. These studies also establish that ultimate proof of the mutated gene of interest is, in fact, essential for osteoclast differentiation and requires rescue of osteopetrotic mice with normal osteoclast precursors, or administration of the gene product. The technique of homologous recombination has yielded a plethora of osteopetrotic knockout mice and major insights into the genes regulating osteoclastogenesis. For example, deletion of the macrophage and B-cell-specific PU.1 gene generates mice completely devoid of osteoclasts and their precursors [310]. These animals represent the earliest known developmental defect in osteoclastogenesis. c-fos, deletion of which prompts osteopetrosis [143], is a widely expressed nuclear protooncogene belonging to a multigene family including fosB, fra-1, and fra-2 [311]. While the fos mutant lacks osteoclasts, marrow macrophages are actually increased relative to wild type. Thus, c-fos may exert its osteoclast differentiating effect distal to PU.1. Similarly, deletion of the p50 and p52
ROSS AND TEITELBAUM
subunits of NF-B prompts osteoclast-deficient osteopetrosis in which macrophage differentiation is extant [312]. NF-B, therfore, impacts osteoclastogenesis in a manner reminiscent of c-fos. The osteopetrotic phenotype of the NF-B knockout indicates cell surface receptors activating this transcription complex may also participate in the osteoclastogenic process. Thus, deletion of RANK-L [135] or its functional blockade by the overexpression of osteoprotegerin [91] each results in osteopetrotic mice devoid of osteoclasts. 2. OSTEOPETROSIS DUE TO DEFECTIVE OSTEOCLAST FUNCTION Among the most interesting of the osteopetrotic mutants is the c-src -/- mouse [59]. This animal, produced by homologous recombination, exhibits no detectable abnormalities in brain or platelets where pp60c-src is normally abundant, but is affected by a form of osteopetrosis in which osteoclasts are numerous. This latter finding suggests that the osteoclast lineage is intact but that the cell is functionally impaired [109]. In fact, electron microscopic examination reveals that src -/- osteoclasts contain no ruffled membranes (Fig. 14). Moreover, osteoclasts generated in vitro from src mutant marrow cells fail to pit bone. These observations signify that lack of bone resorption by src -/- osteoclasts is intrinsic to this cell and not reflective of the bone microenvironment [313]. Similar results were obtained with deletion of TRAF6 [151], a protein associating with the intracellular domain of RANK and which signals following association with c-src [152]. The absence of ruffling in src mutant osteoclasts suggests that pp60c-src is fundamental to osteoclast polarization and may participate in transporting vacuolar H -ATPase from internal pools of acidic vesicles to the plasma membrane. Although this hypothesis lacks direct proof, it is supported by several lines of circumstantial evidence. For example, pp60c-src is a specific marker of macrophages committed to the osteoclast as compared to the host defense phenotype [145]. Both pp60c-src and the vacuolar H -ATPase associate with microtubules in the putative pathway toward the formation of the ruffled membrane [20]. The osteoclast src protein is increased and decreased, respectively, in response to parathyroid hormone and calcitonin [314]. Moreover, bovine adrenal gland chromaffin granules are highly enriched in pp60c-src [315 – 317]. Because these structures are secretory organelles, they provide indirect evidence that the tyrosine kinase has a role in vesicle transport and/or secretion, events akin to ruffled membrane formation. Interestingly, the role of pp60c-src in osteoclast function appears to involve its capacity both as a tyrosine kinase and docking protein, as a kinase-inactive construct substantially, but not completely, rescues the c-src-/- osteoclast [318]. With the realization that the v3 integrin and cathepsin K may be critical to optimal osteoclast function, mice
CHAPTER 3 Osteoclast Biology
FIGURE 14 Osteoclasts from normal and c-src-/- mice. Wild-type cells (A) contain a well-developed ruffled membrane (r), which is lacking in the osteopetrotic mutant (B). Reproduced from J. Clin. Invest. 90, 1622 – 1627 (1992). Copyright permission of The Society for Clinical Investigation.
91
92 devoid of one or the other molecule have been generated [101,125]. While each lack the characteristic pathological features of osteopetrosis, they both contain dysfunctional osteoclasts eventuating in failure to normally resorb bone and, hence, enhanced bone mass. Two general phenotypes of osteopetrosis have been described in humans. One is relatively asymptomatic, whereas the other, usually present at birth, is fatal in infancy or early childhood. While the molecular mechanisms underlying the malignant forms of osteopetrosis are not yet resolved, there is information regarding the “benign’’ phenotype. One such example is characterized by renal tubular acidosis. Consistent with the capacity of the isoform’s inhibitors to also block parathyroid hormone-induced bone resorption, mutation of the carbonic anhydrase II gene is responsible for this form of benign osteopetrosis [319]. Carbonic anhydrase II, in normal osteoclasts, catalyzes the hydration of CO2 to carbonic acid. Dissociation of carbonic acid into bicarbonate ions and protons permits formation of the resorptive microenvironment, which is probably defective in carbonic anhydrase II-deficient patients. Considering the importance of proton secretion in bone degradation, one might expect inactivating mutations of the osteoclast H -ATPase to impair the resorptive process. Indeed, osteoclasts generated in vitro from marrow cells of a patient with craniometaphyseal dysplasia, a rare genetic sclerosing bone disease, fail to resorb bone [320]. These osteoclasts do not express the H -ATPase as determined by immunohistochemistry. 3. OSTEOPETROSIS DUE TO DEFECTIVE OSTEOCLASTOGENIC MICROENVIRONMENT The murine recessive op mutation prompts an osteopetrotic phenotype characterized by a failure to generate monocytes, macrophages, and osteoclasts [321,322]. These animals are not cured by marrow transplantation. Alternatively, op/op hematopoietic progenitors, administered to wild-type mice, differentiate into normal osteoclasts. These experiments prove that the op mutation does not target macrophage and osteoclast progenitors but rather that the microenvironment in which these cells develop is defective [323,324]. In fact, op/op bone cells and fibroblasts fail to produce competent M-CSF [84,325]. Furthermore, the op mutation maps to a single base pair insertion in the coding region of the M-CSF gene, resulting in a translation frame shift, and insertion of a stop codon 21 bp downstream, producing a truncated, nonfunctional M-CSF protein [326]. The observation that the op/op mouse is rescued by subcutaneous injections of recombinant M-CSF provides additional evidence that lack of this protein is the cause of its osteopetrosis [83,327,328]. The op/op mouse underscores the importance of the hematopoietic microenvironment in osteoclast development and points to the fact that osteoclasts and macrophages share a common lineage.
ROSS AND TEITELBAUM
Surprisingly, op/op mice progressively recover from lack of both macrophages and osteoclasts with age and, by 22 weeks, the marrow cavity size and cellularity appear unremarkable [329]. The numbers of mononuclear phagocytes, macrophages, and osteoclast precursors progressively increase to normal. Thus, the animal is able to compensate for a lack of M-CSF, presumably by other cytokines with whose function it overlaps. The spontaneous rescue of the op/op mouse has been shown to reflect expression of GM-CSF, a protein with many activities shared, in vitro, with M-CSF [330]. Supporting this contention, GM-CSF administered to op/op mice during their osteopetrotic phase cures the bone phenotype. This observation, taken with the normal skeletons of GM-CSF knockout mice [331], indicates that either M-CSF or GM-CSF, in the absence of the other, will promote osteoclastogenesis. 4. CURE OF HUMAN OSTEOPETROSIS Based on the animal models described earlier, a female patient with autosomal-recessive osteopetrosis was cured after receiving her brother’s marrow. In this instance, the donor origin of osteoclasts was established by following the Y chromosome [332]. The pretransplantation abundance of osteoclasts, albeit dysfunctional, established the patient’s capacity to provide an osteoclastogenic environment for normal marrow. Given an appropriate donor, marrow transplantation is the treatment of choice for malignant osteopetrosis.
B. Paget’s Disease Paget’s disease of bone is a paradigm of remodeling gone awry. As such, the initiator of bone remodeling, namely the osteoclast, is the pathogenetic cell. In fact, despite the array of abnormal histological features attending Paget’s disease, it is the appearance of the osteoclast that is pathognomonic of this condition [333]. The number of osteoclasts, and their nuclei, are increased greatly in Paget’s disease and the cells are enormous (Fig. 15). In contrast to normal osteoclasts in which nuclei polarize toward the nonresorptive surface, those in pagetic cells are distributed diffusely throughout the cytoplasm. Reflecting osteoclast size and number, bone resorption is accelerated greatly. Resorption bay volume and prevalence are increased manyfold [333,334]. Indeed, the initial cellular event in the pagetic skeleton is rapid resorption, often resulting in large, radiographically evident, lytic lesions [335]. Interestingly, one finds histological evidence of enhanced resorption in radiographically uninvolved bone. It is unresolved whether this phenomenon represents subclinical Paget’s disease or, as has been suggested [333,336], mild
93
CHAPTER 3 Osteoclast Biology
FIGURE 15
Osteoclasts (arrows) in bone biopsy taken from a patient with Paget’s disease. The cells are enormous and contain many nuclei, which, in contrast to normal osteoclasts (see Fig. 1), are nonpolarized (nondecalcified, toluidine blue stain).
hyperparathyroidism. In any event, given the central role that the osteoclast plays in Paget’s disease, it is not surprising that successful treatment entails arresting bone resorption [337,338]. While the precise etiology of Paget’s disease is not yet understood, the measles virus appears to be the most likely candidate. For example, osteoclasts in pagetic bone contain filamentous nuclear [339 – 341] and cytoplasmic inclusions [342] typical of the paramyxovirus family. Most importantly, experiments performed in Roodman’s laboratory demonstrate that targeting this virus to osteoclast precursors in vivo yields a mouse with osteoclasts mimicking those of pagetic patients [343]. This laboratory also found that similar to the in vivo situation, bone marrow derived from patients with Paget’s disease gives rise to numerous, large, hypernucleated osteoclast-like cells [344]. In contrast to polykaryons generated from normal marrow, these multinucleated cells exhibit increased sensitivity to 1,25-dihydroxyvitamin D and produce the osteoclastogenic cytokine IL-6, suggesting that an autocrine event mediates osteoclast proliferation in the disease [345]. Supporting this hypothesis is the fact that IL-6 circulates in increased amounts in pagetic patients [186]. The abundance of osteoclasts in Paget’s disease reflects the proliferative capacity of precursor cells. The number of CFU-GM colonies formed from pagetic marrow is increased, as is replication of CD34 cells, the earliest osteoclast progenitor yet identified in human [41]. It is of interest that pagetic marrow cells not in the osteoclast lineage (i.e., CD34 cells) may serve an accessory function in the
generation of resorptive polykaryons [41]. In this regard, pagetic stromal cells overexpress the RANK ligand, and marrow isolated from affected patients generates increased numbers of osteoclasts in response to the cytokine [346].
C. Cancer Osteolysis is the product of many malignant neoplasms either resident in or distant from bone. In most instances, it appears that tumor-induced osteolysis reflects the recruitment of osteoclasts, by-products of the neoplasm or bone matrix per se to the site of potential bone destruction. Such recruitment may be (i) the result of intimate interactions of a localized solid tumor, in bone, with osteoclast progenitors, (ii) humoral factors secreted by a tumor absent of skeletal metastases, or (iii) bidirectional paracrine stimulation of the tumor and bone marrow cells by cytokines. 1. SOLID TUMORS RESIDENT IN BONE Tumors commonly metastatic to the skeleton include those primary in lung, prostate, thyroid, and kidney [347]. Breast cancer is, however, by far the predominant source of osteolytic lesions. Reflecting excessive bone resorption, carcinoma of the breast is frequently complicated by hypercalcemia and fracture. The evidence that osteoclasts are pivotal to development of metastatic tumor-induced osteolysis includes: (i) in contrast to their ability to degrade soft tissues, cancer cells appear to have a limited, if any, capacity to resorb bone [334];
94
ROSS AND TEITELBAUM
FIGURE 16 Osteoclastic bone resorption in skeletal metastasis. Osteoclasts recruited by tumor resorb bone. Breast cancer cells (arrows) are in close proximity to osteoclasts (arrowheads) and follow the polykaryons into resorption bays (hematoxylin and eosin). (ii) breast cancer-induced osteolysis is blunted by osteoclast-inhibiting bisphosphonates [348]; (iii) osteoclasts are abundant in foci of metastatic breast cancer (Fig. 16); and (iv) breast cancer cells secrete potential osteoclastogenic agents [349 – 351]. In fact, evidence shows that parathyroid hormone-related protein is pivotal to the osteolytic properties of metastatic breast cancer. This protein, particularly produced by those breast cancers with a predisposition for skeletal metastasis, prompts marrow stromal cells, or osteoblasts, to synthesize osteoclast-recruiting cytokines [352]. The tumor-recruited osteoclasts resorb bone, releasing stored growth factors, notably transforming growth factor , which induces the neoplasm to proliferate and synthesize additional PTHrP [353]. This being the case, one would expect inhibition of tumor-induced osteoclastic bone resorption to decrease the tumor burden; in fact, bisphosphonates have such an effect [354]. 2. HUMORAL HYPERCALCEMIA OF MALIGNANCY PTHrP is also the active agent secreted by most tumors inducing humoral hypercalcemia of malignancy [355,356]. The amino terminus of the molecule is recognized by the PTH receptor. As this receptor has not been shown
convincingly to function in osteoclasts, the osteoclastogenic, and consequent, hypercalcemic effects of PTHrP are probably mediated by osteoblasts or marrow stromal cells in which the PTH/PTHrP receptor is abundant. 3. MULTIPLE MYELOMA In contrast to other B-cell malignancies, lytic bone lesions are common in patients with multiple myeloma. In fact, myeloma represents one of few malignancies almost always associated with osteolysis without an apparent osteoblastic component [357]. The central role osteoclasts play in the generation of myeloma bone lesions is underscored by the dramatic impact of bisphosphonates [358]. Bone biopsies from myeloma patients contain abundant osteoclasts in close proximity to the tumor [359,360]. Thus, myeloma cells alone, or in combination with surrounding stromal cells, likely produce factors capable of activating or recruiting osteoclasts. Medium conditioned by extirpated myeloma contains candidate molecules, including IL-6, IL-1, TNF, and macrophage inflammatory protein [361,362]. IL-6, which impacts osteoclasts profoundly, is also critical to myeloma proliferation [363,364]. Interestingly, and in keeping with the role of the
95
CHAPTER 3 Osteoclast Biology
cytokine in osteoclastogenesis attending postmenopausal osteoporosis, IL-6 mRNA is expressed by stromal and not myeloma cells. These observations raise the issue as to how myeloma cells induce cytokine production by accessory cells. Evidence indicates that the osteoclastogenic properties of myeloma cells depend on their contact, via the 41 integrin, with VCAM-1 on stromal cells [365]. 4. INFLAMMATORY OSTEOLYSIS Inflammatory osteolysis is a common form of clinically significant bone loss. This family of conditions includes alveolar bone loss attending periodontal disease, periprosthetic osteolysis, which frequently follows orthopedic implant surgery, and osteopenia accompanying rheumatoid arthritis. Each of these conditions is characterized by abundant osteoclast proliferation, eventuating in dramatic, localized bone destruction. In the case of rheumatoid arthritis, the osteopenia may also be systemic. TNF, interacting with its p55 (type 1) receptor, appears to be central to the osteoclastogenesis of inflammatory osteolysis [145,366] and, hence, newly available drugs inhibiting this cytokine hold therapeutic promise in these disorders. 5. POSTMENOPAUSAL OSTEOPOROSIS Postmenopausal (type 1) osteoporosis is due to an absolute increase bone resorption relative to the premenopausal state [190]. This enhancement of resorptive activity reflects the absence of estrogen’s antiosteoclastogenic effects. Thus, bone biopsies taken following estrogen withdrawal are rich in osteoclasts [190]. The means by which estrogen suppresses osteoclastogenesis was discussed earlier and in Chapter 41.
References 1. H. C. Blair, A. J. Kahn, E. C. Crouch, J. J. Jeffrey, and S. L. Teitelbaum, Isolated osteoclasts resorb the organic and inorganic components of bone. J. Cell Biol. 102, 1164 – 1172 (1986). 2. J. I. Alvarez, S. L. Teitelbaum, H. C. Blair, E. M. Greenfield, N. A. Athanasou, and F. P. Ross, Generation of avian cells resembling osteoclasts from mononuclear phagocytes. Endocrinology 128, 2324 – 2335 (1991). 3. H. M. Frost, Tetracycline-based histological analysis of bone remodeling. Calcif. Tissue Int. 3, 211 – 237 (1969). 4. T. J. Chambers and K. Fuller, Bone cells predispose bone surfaces to resorption by exposure of mineral to osteoclastic attack. J. Cell Sci. 76, 155 – 165 (1985). 5. M. D. Fallon, H. A. Schwamm, An update on the pathogenesis, pathophysiology, and treatment of osteitis deformans. Path. Annu. 24, 115 – 159 (1989). 6. A. Teti, G. Volleth, A. Carano, and A. Zambonin-Zallone, The effects of parathyroid hormone on 1,25-dihydroxyvitamin D3 on monocyteosteoclast fusion. Calcif. Tissue Int. 42, 302 – 308 (1988). 7. J. E. Fisher, M. J. Rogers, J. M. Halasy, S. P. Luckman, D. E. Hughes, P. J. Masarachia, G. Wesolowski, R. G. Russell, G. A. Rodan, and A. A. Reszka, Alendronate mechanism of action:
8. 9.
10. 11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
Geranylgeraniol, an intermediate in the mevalonate pathway, prevents inhibition of osteoclast formation, bone resorption, and kinase activation in vitro. Proc. Natl. Acad. Sci. USA 96, 133 – 138 (1999). T. Efstratiadis and D. W. Moss, Tartrate-resistant acid phosphatase in human alveolar macrophages. Enzyme 34, 140 – 143 (1985). J. Schindelmeiser, D. Munstermann, and H. Witzel, Histochemical investigations on the localization of the purple acid phosphatase in the bovine spleen. Histochemistry 87, 13 – 19 (1987). H. M. Massey and A. M. Flanagan, Human osteoclasts derive from CD14-positive monocytes. Br. J. Haematol. 106, 167 – 170 (1999). N. Udagawa, N. Takahashi, T. Akatsu, H. Tanaka, T. Sasaki, T. Nishihara, and T. Suda, Origin of osteoclasts: Mature monocytes and macrophages are capable of differentiating into osteoclasts under a suitable microenvironment prepared by bone marrow-derived stromal cells. Proc. Natl. Acad. Sci. USA 87, 7260 – 7264 (1990). A. McInnes and D. M. Rennick, Interleukin 4 induces cultured monocytes/macrophages to form giant multinucleated cells. J. Exp. Med. 167, 598 – 611 (1988). M. E. Holtrop, Light and electronmicroscopic structure of osteoclasts. In “Bone” (B. K. Hall, ed.), pp. 1 – 29. CRC Press, Boca Raton, FL, 1991. M. D. Fallon, S. L. Teitelbaum, and A. J. Kahn, Multinucleation enhances macrophage-mediated bone resorption. Lab. Invest. 49, 159 – 164 (1983). Y. Abu-Amer, S. L. Teitelbaum, P. Schlesinger, and F. P. Ross, Expression and regulation of RAB3 proteins in osteoclasts and their precursors. J. Bone Miner. Res. 14, 1855 – 1860 (1999). D. M. Kallio, P. R. Garant, and C. Minkin, Evidence of coated membranes in the ruffled border of osteoclasts. J. Ultrastruct. Res. 37, 169 – 177 (1971). G. Andersson, B. Ek-Rylander, and C. Minkin, Acid phosphatases. In “Biology and Physiology of the Osteoclast” (B. R. Rifkin and C. V. Gay, eds.), pp. 55 – 80. CRC Press, Boca Raton, FL, 1992. B. Ek-Rylander, M. Flores, M. Wendel, D. Heinegard, and G. Andersson, Dephosphorylation of osteopontin and bone sialoprotein by osteoclastic tartrate-resistant acid phosphatase: Modulation of osteoclast adhesion in vitro. J. Biol. Chem. 269, 14853 – 14856 (1994). O. Fukushima, P. J. Bekker, and C. V. Gay, Characterization of the functional stages of osteoclasts by enzyme histochemistry and electron microscopy. Anat. Rec. 231, 298 – 315 (1991). Y. Abu-Amer, F. P. Ross, P. Schlesinger, M. M. Tondravi, and S. L. Teitelbaum, Substrate recognition by osteoclast precursors induces s-crc/microtubule association. J. Cell Biol. 137, 247 – 258 (1997). M. E. Holtrop and L. G. Raisz, Comparison of the effects of 1,25dihydroxycholecalciferol, prostaglandin E2, and osteoclast-activating factor with parathyroid hormone on the ultrastructure of osteoclasts in cultured long bones of fetal rats. Calcif. Tissue Int. 29, 201 (1979). M. E. Holtrop, K. A. Cox, M. B. Clark, M. F. Holick, and C. S. Anast, 1,25-Dihydroxycholecalciferol stimulates osteoclasts in rat bones in the absence of parathyroid hormone. Endocrinology 108, 2293 – 2301 (1981). A. Zambonin-Zallone, A. Teti, A. Carano, and P. C. Marchisio, The distribution of podosomes in osteoclasts cultured on bone laminae: Effect of retinol. J. Bone Miner. Res. 3, 517 – 523 (1988). A. Zambonin-Zallone, A. Teti, M. Grano, A. Rubinacci, M. Abbadini, M. Gaboli, and P. C. Marchisio, Immunocytochemical distribution of extracellular matrix receptors in human osteoclasts: A 3 integrin is colocalized with vinculin and talin in the podosomes of osteoclastoma giant cells. Exp. Cell Res. 182, 645 – 652 (1989). D. Zhang, N. Udagawa, I. Nakamura, H. Murakami, S. Saito, K. Yamasaki, Y. Shibasaki, N. Morii, S. Narumiya, N. Takahashi, and T. Suda, The small GTP-binding protein, rho p21, is involved in bone resorption by regulating cytokskeletal organization in osteoclasts. J. Cell Sci. 108, 2285 – 2292 (1995).
96 26. A. Miyauchi, K. A. Hruska, E. M. Greenfield, R. Barattolo, S. Colucci, A. Zambonin-Zallone, S. L. Teitelbaum, and A. Teti, Osteoclast cytosolic calcium, regulated by voltage operated calcium channels and extracellular calcium, controls podosome assembly and bone resorption. J. Cell Biol. 111, 2543 – 2552 (1990). 27. H. Rasmussen and P. Bordier, The cellular basis of metabolic bone disease. N. Engl. J. Med. 289, 25 – 32 (1973). 28. G. Gothlin and J. L. E. Ericsson, The osteoclast: Review of ultrastructure, origin, and structure-function relationship. Clin. Orthop. Rel. Res. 120, 201 – 231 (1976). 29. A. J. Kahn and D. J. Simmons, Investigation of cell lineage in bone using a chimaera of chick and quail embryonic tissue. Nature 258, 325 – 327 (1975). 30. A. J. Kahn, C. C. Stewart, and S. L. Teitelbaum, Contact-mediated bone resorption by human monocytes in vitro. Science 199, 988 – 990 (1978). 31. S. L. Teitelbaum, C. C. Stewart, and A. J. Kahn, Rodent peritoneal macrophages as bone resorbing cells. Calcif. Tissue Int. 27, 255 – 261 (1979). 32. A. J. Kahn, J. D. Malone, and S. L. Teitelbaum, Osteoclast precursors, mononuclear phagocytes, and bone resorption. Trans. Assoc. Am. Phys. 94, 267 – 278 (1981). 33. M. Nagai and N. Sato, Reciprocal gene expression of osteoclastogenesis inhibitory factor and osteoclast differentiation factor regulates osteoclast formation. Biochem. Biophys. Res. Commun. 257, 719 – 723 (1999). 34. E. H. Burger, J. W. M. Van Der Meer, J. K. Van De Gevel, J. C. Gribnau, C. W. Thesingh, and R. Van Furth, In vitro formation of osteoclasts from long-term cultures of bone marrow mononuclear phagocytes. J. Exp. Med. 156, 1604 – 1614 (1982). 35. B. A. A. Scheven, J. W. M. Visser, and P. J. Nijweide, In vitro osteoclast generation from different bone marrow fractions, including a highly enriched haematopoietic stem cell population. Nature 321, 79 – 81 (1986). 36. N. Kurihara, T. Suda, Y. Miura, H. Nakauchi, H. Kodama, K. Hiura, Y. Hakeda, and M. Kumegawa, Generation of osteoclasts from isolated hematopoietic progenitor cells. Blood 74, 1295 – 1302 (1989). 37. N. Takahashi, H. Yamana, S. Yoshiki, G. D. Roodman, G. R. Mundy, S. J. Jones, A. Boyde, and T. Suda, Osteoclast-like cell formation and its regulation by osteotropic hormones in mouse bone marrow cultures. Endocrinology 122, 1373 – 1382 (1988). 38. N. Takahashi, T. Akatsu, T. Sasaki, G. C. Nicholson, J. M. Moseley, T. J. Martin, and T. Suda, Induction of calcitonin receptors by 1 alpha, 25-dihydroxyvitamin D3 in osteoclast-like multinucleated cells formed from mouse bone marrow cells. Endocrinology 123, 1504 – 1510 (1988). 39. N. Takahashi, T. Akatsu, N. Udagawa, T. Sasaki, A. Yamaguchi, J. M. Moseley, T. J. Martin, and T. Suda, Osteoblastic cells are involved in osteoclast formation. Endocrinology 123, 2600 – 2602 (1988). 40. N. Udagawa, N. Takahashi, T. Akatsu, T. Sasaki, A. Yamaguchi, H. Kodama, T. J. Martin, and T. Suda, The bone marrow-derived stromal cell lines MC3T3-G2/PA6 and ST2 support osteoclast-like cell differentiation in cocultures with mouse spleen cells. Endocrinology 125, 1805 – 1813 (1989). 41. N. Kurihara, C. Chenu, M. Miller, C. Civin, and G. D. Roodman, Identification of committed mononuclear precursors for osteoclastlike cells formed in long term human marrow cultures. Endocrinology 126, 2733 – 2741 (1991). 42. T. Suda, N. Takahashi, and T. J. Martin, Modulation of osteoclast differentiation. Endocr. Rev. 13, 66 – 80 (1992). 43. R. K. Globus, D. D. Bikle, B. Halloran, and E. Morey-Holton, Skeletal response to dietary calcium in a rat model simulating weightlessness. J. Bone Miner. Res. 1, 191 – 197 (1986).
ROSS AND TEITELBAUM 44. F. Tilton, J. Degioanni, and V. Schneider, Long-term follow-up of skylab bone demineralization. Aviat. Space Environ. Med. 51, 1209 – 1216 (1980). 45. M. G. Griffin, R. Kimble, W. Hopfer, and R. Pacifici, Dual-energy X-ray absorptiometry of the rat: Accuracy, precision, and measurement of bone loss. J. Bone Miner. Res. 8, 795 – 800 (1993). 46. S. P. Robins, H. Woitge, R. Hesley, J. Ju, S. Seyedin, and M. J. Seibel, Direct, enzyme-linked immunoassay for urinary deoxypyridinoline as a specific marker for measuring bone resorption. J. Bone Miner. Res. 9, 1643 – 1649 (1994). 47. L. G. Raisz, Stimulation of bone resorption by parathyroid hormone in tissue culture. Nature 197, 1015 – 1017 (1963). 48. S. L. Teitelbaum, J. E. Russell, J. M. Bone, J. J. Gilden, and L. V. Avioli, The relationship of biochemical and histometric determinants of uremic bone. Arch. Pathol. Lab. Med. 103, 228 – 230 (1979). 49. P. D. Delmas, Biochemical markers of bone turnover in osteoporosis. In “Osteoporosis” (B. L. Riggs and L. J. Melton, eds.), pp. 297 – 316. Raven Press, New York, 1988. 50. M. J. Seibel, S. P. Robins, and J. P. Bilezikian, Urinary pyridinium crosslinks of collagen: Specific markers of bone resorption in metabolic bone disease. Trends Endocrinol. Metab. 3, 263 – 270 (1992). 51. T. J. Chambers and C. J. Magnus, Calcitonin alters behavior of isolated osteoclasts. J. Pathol. 136, 27 – 39 (1982). 52. R. Nelson and G. Bauer, Isolation of osteoclasts by velocity sedimentation at unit gravity. Calcif. Tissue Res. 22, 303 – 309 (1977). 53. P. Osdoby, M. Martini, and A. Caplan, Isolated osteoclasts and their presumed progenitor cells, the monocyte, in culture. J. Exp. Zool. 244, 331 – 339 (1982). 54. T. Hefley and P. Stern, Isolation of osteoclasts from fetal rat long bones. Calcif. Tissue Int. 34, 480 – 487 (1982). 55. A. Zambonin-Zallone, A. Teti, and M. Primavera, Isolated osteoclasts in primary culture: First observation on structure and survival in culture media. Anat. Embryol. 165, 405 – 413 (1982). 56. H. C. Blair, S. L. Teitelbaum, L. E. Grosso, D. L. Lacey, H.-L. Tan, D. W. McCort and J. J. Jeffrey, Extracellular matrix degradation at acid pH: Avian osteoclast acid collagenase isolation and characterization. Biochem. J. 290, 873 – 884 (1993). 57. J. P. Mattsson, P. H. Schlesinger, D. J. Keeling, S. L. Teitelbaum, D. K. Stone, and X.-S. Xie, Isolation and reconstitution of a vacuolar-type proton pump of osteoclast membranes. J. Biol. Chem. 269, 24979 – 24982 (1994). 58. W. C. Horne, L. Neff, D. Chatterjee, A. Lomri, J. B. Levy, and R. Baron, Osteoclasts express high levels of pp60c-src in association with intracellular membranes. J. Cell Biol. 119, 1003 – 1013 (1992). 59. P. Soriano, C. Montgomery, R. Geske, and A. Bradley, Targeted disruption of the c-src proto-oncogene leads to osteopetrosis in mice. Cell 64, 693 – 702 (1991). 60. K. Tezuka, T. Sato, H. Kamioka, P. J. Nijweide, K. Tanaka, T. Matsuo, M. Ohta, N. Kurihara, Y. Hakeda, and M. Kumegawa, Identification of osteopontin in isolated rabbit osteoclasts. Biochem. Biophys. Res. Commun. 186, 911 – 917 (1992). 61. K. Tezuka, Y. Tezuka, A. Maejima, T. Sato, K. Nemoto, H. Kamioka, Y. Hakeda and M. Kumegawa, Molecular cloning of a possible cysteine proteinase predominantly expressed in osteoclasts. J. Biol. Chem. 269, 1106 – 1109 (1994). 62. Z. Zimolo, G. Wesolowski, H. Tanaka, J. L. Hyman, J. R. Hoyer, and G. A. Rodan, Soluble v3-integrin ligands raise [Ca2]i in rat osteoclasts and mouse-derived osteoclast-like cells. Am. J. Physiol. 266, C376 – C381 (1994). 63. G. Shankar, I. Davison, M. H. Helfrich, W. T. Mason, and M. A. Horton, Integrin receptor-mediated mobilisation of intranuclear calcium in rat osteoclasts. J. Cell Sci. 105, 61 – 68 (1993). 64. M. J. Oursler, L. V. Bell, B. Clevinger, and P. Osdoby, Identification of osteoclast-specific monoclonal antibodies. J. Cell Biol. 100, 1592 – 1600 (1985).
97
CHAPTER 3 Osteoclast Biology 65. T. J. Chambers, K. Fuller, J. A. Darby, J. A. S. Pringle, and M. A. Horton, Monoclonal antibodies against osteoclasts inhibit bone resorption in vitro. Bone Miner. 1, 127 – 135 (1986). 66. P. Collin-Osdoby, M. J. Oursler, D. Webber, and P. Osdoby, Osteoclast-specific monoclonal antibodies coupled to magnetic beads provide a rapid and efficient method of purifying avian osteoclasts. J. Bone Miner. Res. 6, 1353 – 1365 (1991). 67. J. Davies, J. Warwick, N. Totty, R. Philp, M. Helfrich, and M. Horton, The osteoclast functional antigen, implicated in the regulation of bone resorption, is biochemically related to the vitronectin receptor. J. Cell Biol. 109, 1817 – 1826 (1989). 68. Y. Ohsaki, S. Takahashi, T. Scarcez, A. Demulder, T. Nishihara, R. Williams, and G. D. Roodman, Evidence for an autocrine/paracrine role for interleukin-6 in bone resorption by giant cells from giant cell tumors of bone. Endocrinology 131, 2229 – 2234 (1992). 69. S. Takahashi, S. V. Reddy, J. M. Chirgwin, R. Devlin, C. Haipek, J. Anderson, and G. D. Roodman, Cloning and identification of annexin II as an autocrine/paracrine factor that increases osteoclast formation and bone resorption. J. Biol. Chem. 269, 28696 – 28701 (1994). 70. M. J. Oursler, P. Collin-Osdoby, L. Li, E. Schmitt, and P. Osdoby, Evidence for an immunological and functional relationship between superoxide dismutase and a high molecular weight osteoclast plasma membrane glycoprotein. J. Cell. Biochem. 46, 331 – 344 (1991). 71. I. R. Garrett, B. F. Boyce, R. O. Oreffo, L. Bonewald, J. Poser, and G. R. Mundy, Oxygen-derived free radicals stimulate osteoclastic bone resorption in rodent bone in vitro and in vivo. J. Clin. Invest. 85, 632 – 639 (1990). 72. F. P. Ross, J. I. Alvarez, J. Chappel, D. Sander, W. T. Butler, M. C. Farach-Carson, K. A. Mintz, P. G. Robey, S. L. Teitelbaum, and D. A. Cheresh, Interactions between the bone matrix proteins osteopontin and bone sialoprotein and the osetoclast integrin v3 potentiate bone resorption. J. Biol. Chem. 268, 9901 – 9907 (1993). 73. M. Chiba, S. L. Teitelbaum, and F. P. Ross, Treatment of avian osteoclast precursors with retinoic acid alters the rate of synthesis of 3 but not v mRNA and increases association of the heterodimer, by stimulating the synthesis of 3 protein. J. Bone Miner. Res. 8, S144 (1993). 74. H. Mimura, X. Cao, F. P. Ross, M. Chiba, and S. L. Teitelbaum, 1,25(OH)2D3 vitamin D3 transcriptionally activates the 3-integrin subunit gene in avian osteoclast precursors. Endocrinology 134, 1061 – 1066 (1994). 75. X. Cao, S. L. Teitelbaum, S. Dedhar, L. Zhang, and F. P. Ross, Retinoic acid and calreticulin, a novel ER-derived transcription factor, inhibit 1,25(OH)2D3-induced gene transcription. J. Bone Miner. Res. 9, S144 (1994). 76. X. Cao, F. P. Ross, L. Zhang, P. N. MacDonald, J. Chappel, and S. L. Teitelbaum, Cloning of the promoter for the avian integrin 3 subunit gene and its regulation by 1,25-dihydroxyvitamin D3. J. Biol. Chem. 268, 27371 – 27380 (1993). 77. A. Miyauchi, J. Alvarez, E. M. Greenfield, A. Teti, M. Grano, S. Colucci, A. Zambonin-Zallone, F. P. Ross, S. L. Teitelbaum, and D. Cheresh, Recognition of osteopontin and related peptides by an v3 integrin stimulates immediate cell signals in osteoclasts. J. Biol. Chem. 266, 20369 – 20374 (1991). 78. K. Fuller and T. J. Chambers, Generation of osteoclasts in cultures of rabbit bone marrow and spleen cells. J. Cell. Physiol. 132, 441 – 452 (1987). 79. B. R. MacDonald, G. R. Mundy, S. Clark, E. A. Wang, T. J. Kuehl, E. R. Stanley, and G. D. Roodman, Effects of human recombinant CSF-GM and highly purified CSF-1 on the formation of multinucleated cells with osteoclast characteristics in long-term bone marrow cultures. J. Bone Miner. Res. 1, 227 – 233 (1986). 80. G. D. Roodman, K. J. Ibbotson, B. R. MacDonald, T. J. Kuehl, and G. R. Mundy, 1,25-dihydroxyvitamin D3 causes formation of
81.
82. 83.
84.
85.
86.
87.
88.
89.
90.
91.
92.
93.
94.
95.
96.
multinucleated cells with several osteoclast characteristics in cultures of primate marrow. Proc. Natl. Acad. Sci. USA 82, 8213 – 8217 (1985). D. L. Lacey, J. M. Erdmann, S. L. Teitelbaum, H.-L. Tan, J. Ohara, and A. Shioi, Interleukin-4, interferon gamma and prostaglandin E impact the osteoclast forming potential of murine bone marrow macrophages in vitro. Endocrinology 136, 2367 – 2376 (1995). A. Shioi, F. P. Ross, and S. L. Teitelbaum, Enrichment of generated murine osteoclasts. Calcif. Tissue Int. 55, 387 – 394 (1994). H. Kodama, M. Nose, S. Niida, and A. Yamasaki, Essential role of macrophage colony-stimulating factor in the osteoclast differentiation supported by stromal cells. J. Exp. Med. 173, 1291 – 1294 (1991). W. Wiktor-Jedrzejczak, A. Bartocci, A. W. Jr. Ferrante, A. AhmedAnsari, K. W. Sell, J. W. Pollard, and E. R. Stanley, Total absence of colony-stimulating factor 1 in the macrophage-deficient osteopetrotic (op/op) mouse. Proc. Natl. Acad. Sci. USA 87, 4828 – 4832 (1990). D. N. Haylock, L. B. To, T. L. Dowse, C. A. Juttner, and P. J. Simmons, Ex vivo expansion and maturation of peripheral blood CD34 cells into the myeloid lineage. Blood 80, 1405 – 1412 (1992). S. Huang and W. M. M. Terstappen, Lymphoid and myeloid differentiation of single human CD34, HLA-DR, CD38 hematopoietic stem cells. Blood 83, 1515 – 1526 (1994). D. L. Lacey, E. Timms, H. L. Tan, M. J. Kelley, C. R. Dunstan, T. Burgess, R. Elliott, A. Colombero, G. Elliott, S. Scully, H. Hsu, J. Sullivan, N. Hawkins, E. Davy, C. Capparelli, A. Eli, Y. X. Qian, S. Kaufman, I. Sarosi, V. Shalhoub, G. Senaldi, J. Guo, J. Delaney, and W. J. Boyle, Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell 93, 165 – 176 (1998). J. M. Quinn, J. Elliott, M. T. Gillespie, and T. J. Martin, A combination of osteoclast differentiation factor and macrophage-colony stimulating factor is sufficient for both human and mouse osteoclast formation in vitro. Endocrinology 139, 4424 – 4427 (1998). P. Collin-Osdoby, L. Rothe, J. Weihe, T. Sunyer, and P. Osdoby, Osteoprotegerin ligand (OPGL) is expressed by human vascular endothelial cells and upregulated in response to the calciotropic agents TNF, PTH, or vitamin D3. J. Bone Miner. Res. 14, S232 (2000). R. Josien, B. R. Wong, H. L. Li, R. M. Steinman, and Y. Choi, TRANCE, a TNF family member, is differentially expressed on T cell subsets and induces cytokine production in dendritic cells. J. Immunol. 162, 2562 – 2568 (1999). W. S. Simonet, D. L. Lacey, C. R. Dunstan, M. Kelley, M. S. Chang, R. Luthy, H. Q. Nguyen, S. Wooden, L. Bennett, T. Boone, G. Shimamoto, M. DeRose, R. Elliott, A. Colombero, H. L. Tan, G. Trail, J. Sullivan, E. Davy, N. Bucay, L. Renshaw-Gegg, T. M. Hughes, D. Hill, W. Pattison, P. Campbell, and W. J. Boyle, Osteoprotegerin: A novel secreted protein involved in the regulation of bone density. Cell 89, 309 – 319 (1997). N. Nakagawa, M. Kinosaki, K. Yamaguchi, N. Shima, H. Yasuda, K. Yano, T. Morinaga, and K. Higashio, RANK is the essential signaling receptor for osteoclast differentiation factor in osteoclastogenesis. Biochem. Biophys. Res. Commun. 253, 395 – 400 (1998). H. Hsu, D. L. Lacey, C. R. Dunstan, I. Solovyev, A. Colombero, E. Timms, H.-L. Tan, G. Elliott, M. J. Kelley, I. Sarosi, L. Wang, X. Z. Xia, R. Elliott, L. Chiu, T. Black, S. Scully, C. Capparelli, S. Morony, G. Shimamoto, M. B. Bass, and W. J. Boyle, Tumor necrosis factor receptor family member RANK mediates osteoclast differentiation and activation induced by osteoprotegerin ligand. Proc. Natl. Acad. Sci. USA 96, 3540 – 3545 (1999). M. A. Horton and M. H. Helfrich, Antigenic markers of osteoclasts. In “Biology and Physiology of the Osteoclast.” (B. R. Rifkin and C. V. Gay, eds.), pp. 33 – 54. CRC Press, Boca Raton, FL, 1992. S. Nesbitt, A. Nesbit, M. Helfrich, and M. Horton, Biochemical characterization of human osteoclast integrins: Osteoclasts express v3, 21 and v1 integrins. J. Cell Biol. 268, 16737 – 16745 (1993). J. E. Fisher, M. P. Caulfield, M. Sato, H. A. Quartuccio, R. J. Gould, V. M. Garsky, G. A. Rodan, and M. Rosenblatt, Inhibition of osteoclastic
98
97.
98.
99.
100.
101.
102.
103.
104.
105.
106.
107.
108.
109.
110.
111. 112.
113.
114.
ROSS AND TEITELBAUM bone resorption in vivo by echistatin, an “arginyl-glycyl-aspartyl” (RGD)-containing protein. Endocrinology 132, 1411 – 1413 (1993). M. Grano, P. Zigrino, S. Colucci, G. Zambonin, L. Trusolino, M. Serra, N. Baldini, A. Teti, P. C. Machisio, and A. Z. Zallone, Adhesion properties and integrin expression of cultured human osteoclastlike cells. Exp. Cell Res. 212, 209 – 218 (1994). G. van der Pluijm, H. Mouthaan, C. Baas, H. de Groot, S. Papapoulos, and C. Lowik, Integrins and osteoclastic resorption in three bone organ cultures: Differential sensitivity to synthetic Arg-Gly-Asp peptides during osteoclast formation. J. Bone Miner. Res. 9, 1021 – 1028 (1994). M. Flores, M. Norgard, D. Heinegard, F. P. Reinholt, and G. Andersson, RGD-directed attachment of isolated rat osteoclasts to osteopontin, bone sialoprotein, and fibronectin. Exp. Cell Res. 201, 526 – 530 (1992). M. A. Horton, M. L. Taylor, T. R. Arnett, and M. H. Helfrich, Arggly-asp (RGD) peptides and the anti-vitronectin receptor antibody 23C6 inhibit dentine resorption and cell spreading by osteoclasts. Exp. Cell Res. 195, 368 – 375 (1991). K. P. McHugh, K. Hodivala-Dilke, M. H. Zheng, N. Namba, J. Lam, D. Novack, X. Feng, F. P. Ross, R. O. Hynes, and S. L. Teitelbaum, Mice lacking 3 integrins are osteosclerotic due to dysfunctional osteoclasts. J. Clin. Invest. 105, 433 – 440 (2000). V. W. Engleman, G. A. Nickols, F. P. Ross, M. A. Horton, S. L. Settle, P. G. Ruminski, and S. L. Teitelbaum, A peptidomimetic antagonist of the v3 integrin inhibits bone resorption in vitro and prevents osteoporosis in vivo. J. Clin. Invest. 99, 2284 – 2292 (1997). F. P. Reinholt, K. Hultenby, A. Oldberg, and D. Heinegard, Osteopontin: A possible anchor of osteoclasts to bone. Proc. Natl. Acad. Sci. USA 87, 4473 – 4475 (1990). K. A. Hruska, F. Rolnick, M. Huskey, U. Alvarez, and D. Cheresh, Engagement of the osteoclast integrin alpha v beta 3 by osteopontin stimulates phosphatidylinositol 3-hydroxyl kinase activity. Endocrinology 136, 2984 – 2992 (1995). I. Nakamura, T. Sasaki, S. Tanaka, N. Takahashi, E. Jimi, T. Kurokawa, Y. Kita, S. Ihara, T. Suda, and Y. Fukui, Phosphatidylinositol-3 kinase is involved in ruffled border formation in osteoclasts. J. Cell Physiol. 172, 230 – 239 (1997). M. Chellaiah and K. Hruska, Osteopontin stimulates gelsolin-associated phosphoinositide levels and phosphatidylinositol triphosphatehydroxyl kinase. Mol. Biol. Cell 7, 743 – 753 (1996). H. Yoshitake, S. R. Rittling, D. T. Denhardt, and M. Noda, Osteopontindeficient mice are resistant to ovariectomy-induced bone resorption. Proc. Natl. Acad. Sci. USA 96, 8156 – 8160 (1999). L. T. Duong, P. T. Lakkakorpi, I. Nakamura, M. Machwate, R. M. Nagy, and G. A. Rodan, PYK2 in osteoclasts is an adhesion kinase, localized in the sealing zone, activated by ligation of alpha(v)beta3 integrin, and phosphorylated by src kinase. J. Clin. Invest. 102, 881 – 892 (1998). B. F. Boyce, T. Yoneda, C. Lowe, P. Soriano, and G. R. Mundy, Requirement of pp60c-src expression for osteoclasts to form ruffled borders and resorb bone in mice. J. Clin. Invest. 90, 1622 – 1627 (1992). H. Palokangas, M. Mulari, and H. K. Vaananen, Endocytic pathway from the basal plasma membrane to the ruffled border membrane in bone-resorbing osteoclasts. J. Cell Sci. 110, 1767 – 1780 (1997). A. L. Boskey, Noncollagenous matrix proteins and their role in mineralization. Bone Miner. 6, 111 – 123 (1989). I. A. Silver, R. J. Murrills, and D. J. Etherington, Microelectrode studies on the acid microenvironment beneath adherent macrophages and osteoclasts. Exp. Cell Res. 175, 266 – 276 (1988). H. C. Blair, S. L. Teitelbaum, R. Ghiselli, and S. Gluck, Osteoclastic bone resorption by a polarized vacuolar proton pump. Science 245, 855 – 857 (1989). Z. Q. Wang and S. Gluck, Isolation and properties of bovine kidney brush border vacuolar H()-ATPase: A proton pump with enzymatic
115.
116.
117.
118.
119.
120.
121. 122.
123.
124.
125.
126.
127.
128.
129.
130.
and structural differences from kidney microsomal H()-ATPase. J. Biol. Chem. 265, 21957 – 21965 (1990). S. L. Alper, J. Natale, S. Gluck, H. F. Lodish, and D. Brown, Subtypes of intercalated cells in rat kidney collecting duct defined by antibodies against erythroid band 3 and renal vacuolar H-ATPase. Proc. Natl. Acad. Sci. USA 86, 5429 – 5433 (1989). D. Chatterjee, M. Chakraborty, M. Leit, L. Neff, S. JamsaKellokumpu, R. Fuchs, and R. Baron, Sensitivity to vanadate and isoforms of subunits A and B distinguish the osteoclast proton pump from other vacuolar H ATPases. Proc. Natl. Acad. Sci. USA 89, 6257 – 6261 (1992). J. P. Mattsson, K. Vaananen, B. Wallmark, and P. Lorentzon, Omeprazole and bafilomycin, two proton pump inhibitors: differentiation of their effects on gastric, kidney and bone H-translocating ATPases. Biochim. Biophys. Acta 1065, 261 – 268 (1991). A. Teti, H. C. Blair, P. Schlesinger, M. Grano, A. Zambonin-Zallone, A. J. Kahn, S. L. Teitelbaum, and K. A. Hruska, Extracellular protons acidify osteoclasts, reduce cytosolic calcium, and promote expression of cell-matrix attachment structures. J. Clin. Invest. 84, 773 – 780 (1989). H. C. Blair, S. L. Teitelbaum, H.-L. Tan, C. M. Koziol, and P. H. Schlesinger, Passive chloride permeability charge coupled to HATPase of avian osteoclast ruffled membrane. Am. J. Physiol. 260, C1315 – C1324 (1991). P. H. Schlesinger, H. C. Blair, S. L. Teitelbaum, and J. C. Edwards, Characterization of the osteoclast ruffled border chloride channel and its role in bone resorption. J. Biol. Chem. 272, 18636 – 18643 (1997). S. A. Nesbitt and M. A. Horton, Trafficking of matrix collagens through bone-resorbing osteoclasts. Science 276, 266 – 269 (1997). J. Salo, P. Lehenkari, M. Mulari, K. Metsikko, and H. K. Vaananen, Removal of osteoclast bone resorption products by transcytosis. Science 276, 270 – 273 (1997). R. Baron, L. Neff, W. Brown, D. Louvard, and P. J. Courtoy, Selective internalization of the apical plasma membrane and rapid redistribution of lysosomal enzymes and mannose-6-phosphate receptors during osteoclast inactivation by calcitonin. J. Cell Sci. 97, 439 – 447 (1990). H. C. Blair, S. L. Teitelbaum, P. A. Schimke, J. D. Konsek, C. M. Koziol, and P. H. Schlesinger, Receptor-mediated uptake of mannose-6-phosphate bearing glycoprotein by isolated chicken osteoclasts. J. Cell. Physiol. 137, 476 – 482 (1988). M. Gowen, F. Lazner, R. Dodds, R. Kapadia, J. Feild, M. Tavaria, I. Bertoncello, F. Drake, S. Zavarselk, I. Tellis, P. Hertzog, C. Debouck, and I. Kola, Cathepsin K knockout mice develop osteopetrosis due to a deficit in matrix degradation but not demineralization. J. Bone Miner. Res. 14, 1654 – 1663 (1999). J. A. Hamilton, S. Lingelbach, N. C. Partridge, and T. J. Martin, Regulation of plasminogen activator by bone-resorbing hormones in normal and malignant osteoblasts. Endocrinology 116, 2186 – 2191 (1985). M. C. Meikle, S. Bond, R. M. Hembry, J. Compston, P. I. Croucher, and J. J. Reynolds, Human osteoblasts in culture synthesize collagenase and other matrix metalloproteinases in response to osteotropic hormones and cytokines. J. Cell Sci. 103, 1093 – 1099 (1992). N. C. Partridge, B. E. Kemp, M. C. Veroni, and T. J. Martin, Activation of adenosine 3’,5’-monophosphate-dependent protein kinase in normal and malignant bone cells by parathyroid hormone, prostaglandin E2, and prostacyclin. Endocrinology 108, 220 – 225 (1981). J. Pfeilschifter, J. Erdmann, W. Schmidt, A. Naumann, H. W. Minne, and R. Ziegler, Differential regulation of plasminogen activator and plasminogen activator inhibitor by osteotropic factors in primary cultures of mature osteoblasts and osteoblast precursors. Endocrinology 126, 703 – 711 (1990). W. G. Zhao, M. H. Byrne, B. F. Boyce, and S. M. Krane, Bone resorption induced by parathyroid hormone is strikingly diminished in collagenase-resistant mutant mice. J. Clin. Invest. 103, 517 – 524 (1999).
99
CHAPTER 3 Osteoclast Biology 131. T. H. Vu, J. M. Shipley, G. Bergers, J. E. Berger, J. A. Helms, D. Hanahan, S. D. Shapiro, R. M. Senior, and Z. Werb, MMP-9/gelatinase B is a key regulator of growth plate angiogenesis and apoptosis of hypertrophic chondrocytes. Cell 93, 411 – 422 (1998). 132. M. Zaidi, B. S. Moonga, and O. A. Adebanjo, Novel mechanisms of calcium handling by the osteoclast: A review-hypothesis. Proc. Assoc. Am. Physicians 111, 319 – 327 (1999). 133. S. Morony, C. Capparelli, R. Lee, G. Shimamoto, T. Boone, D. L. Lacey, and C. R. Dunstan, A chimeric form of osteoprotegerin inhibits hypercalcemia and bone resorption induced by IL-1 beta, TNF-alpha, PTH, PTHrP, and 1,25(OH)(2)D-3. J. Bone Miner. Res. 14, 1478 – 1485 (1999). 134. T. L. Burgess, Y. Qian, S. Kaufman, B. D. Ring, G. Van, C. Capparelli, M. Kelley, H. Hsu, W. J. Boyle, C. R. Dunstan, S. Hu, and D. L. Lacey, The ligand for osteoprotegerin (OPGL) directly activates mature osteoclasts. J. Cell Biol. 145, 527 – 538 (1999). 135. Y. Y. Kong, H. Yoshida, I. Sarosi, H. L. Tan, E. Timms, C. Capparelli, S. Morony, D. S. A. Oliveira, G. Van, A. Itie, W. Khoo, A. Wakeham, C. R. Dunstan, D. L. Lacey, T. W. Mak, W. J. Boyle, and J. M. Penninger, OPGL is a key regulator of osteoclastogenesis, lymphocyte development and lymphnode organogenesis. Nature 397, 315 – 323 (1999). 136. J. Li, I. Sarosi, X.-Q. Yan, S. Morony, C. Capparelli, H.-L. Tan, S. McCabe, R. Elliott, S. Scully, G. Van, S. Kaufman, S.-C. Juan, Y. Sun, J. Tarpley, L. Martin, K. Christensen, J. McCabe, P. Kostenuik, H. Hsu, F. Fletcher, C. R. Dunstan, D. L. Lacey, and W. J. Boyle, RANK is the intrinsic hematopoietic cell surface receptor that controls osteoclastogenesis and regulation of bone mass and calcium metabolism. Proc. Natl. Acad. Sci. USA 97, 1566 – 1571 (2000). 137. N. Bucay, I. Sarosi, C. R. Dunstan, S. Morony, J. Tarpley, C. Capparelli, S. Scully, H. L. Tan, W. Xu, D. L. Lacey, W. J. Boyle, and W. S. Simonet, Osteoprotegerin-deficient mice develop early onset osteoporosis and arterial calcification. Genes Dev. 12, 1260 – 1268 (1998). 138. L. C. Hofbauer, S. Khosla, C. R. Dunstan, D. L. Lacey, W. J. Boyle, and B. L. Riggs, The roles of osteoprotegerin and osteoprotegerin ligand in the paracrine regulation of bone resorption. J. Bone Miner. Res. 15, 2 – 12 (2000). 139. E. Jimi, I. Nakamura, L. T. Duong, T. Ikebe, N. Takahashi, G. A. Rodan, and T. Suda, Interleukin 1 induces multinucleation and boneresorbing activity of osteoclasts in the absence of osteoblasts/stromal cells. Exp. Cell Res. 247, 84 – 93 (1999). 140. E. Jimi, I. Nakamura, T. Ikebe, S. Akiyama, N. Takahashi, and T. Suda, Activation of NF-kappaB is involved in the survival of osteoclasts promoted by interleukin-1. J. Biol. Chem. 273, 8799 – 8805 (1998). 141. K. Kobayashi, N. Takahashi, E. Jimi, N. Udagawa, M. Takami, S. Kotake, N. Nakagawa, M. Kinosaki, K. Yamaguchi, N. Shima, H. Yasuda, T. Morinaga, K. Higashio, T. J. Martin, and T. Suda, Tumor necrosis factor stimulates osteoclast differentiation by a mechanism independent of the ODF/RANKL-RANK interaction. J. Exp. Med. 191, 275 – 285 (2000). 142. G. Franzoso, L. Carlson, L. Xing, L. Poljak, E. W. Shores, K. D. Brown, A. Leonardi, T. Tran, B. F. Boyce, and U. Siebenlist, Requirement for NF-kappaB in osteoclast and B-cell development. Genes Dev. 11, 3482 – 3496 (1997). 143. A. E. Grigoriadis, Z.-Q. Wang, M. G. Cecchini, W. Hofstetter, R. Felix, H. A. Fleisch, and E. F. Wagner, c-Fos: A key regulator of osteoclast-macrophage lineage determination and bone remodeling. Science 266, 443 – 448 (1994). 144. J.-P. David, K. Sabapathy, A. Behrens, B. Berger, K. Hochedlinger, O. Hoffmann, and E. F. Wagner, c-Jun phosphorylation by JNK1 but not by JNK2 is necessary for efficient osteoclastogenesis in response to TRANCE stimulation. J. Bone Miner. Res. 14, S149 (1999). 145. Y. Abu-Amer, F. P. Ross, J. Edwards, and S. L. Teitelbaum, Lipopolysaccharide-stimulated osteoclastogenesis is mediated by
146.
147.
148.
149.
150.
151.
152.
153.
154.
155.
156.
157.
158.
159.
160.
tumor necrosis factor via its p55 receptor. J. Clin. Invest. 100, 1557 – 1565 (1997). R. H. Arch, R. W. Gedrich, and C. B. Thompson, Tumor necrosis factor receptor-associated factors (TRAFs): A family of adapter proteins that regulates life and death. Genes Dev. 12, 2821 – 2830 (1998). L. Galibert, M. E. Tometsko, D. M. Anderson, D. Cosman, and W. C. Dougall, The involvement of multiple tumor necrosis factor receptor (TNFR)-associated factors in the signaling mechanisms of receptor activator of NF-kappaB, a member of the TNFR superfamily. J. Biol. Chem. 273, 34120 – 34127 (1998). H. Nakano, S. Sakon, H. Koseki, T. Takemori, K. Tada, M. Matsumoto, E. Munechika, T. Sakai, T. Shirasawa, H. Akiba, T. Kobata, S. M. Santee, C. F. Ware, P. D. Rennert, M. Taniguchi, H. Yagita, and K. Okumura, Targeted disruption of Traf5 gene causes defects in CD40- and CD27-mediated lymphocyte activation. Proc. Natl. Acad. Sci. USA 96, 9803 – 9808 (1999). W. C. Yeh, A. Shahinian, D. Speiser, J. Kraunus, F. Billia, A. Wakeham, J. L. de la Pompa, D. Ferrick, B. Hum, N. Iscove, P. Ohashi, M. Rothe, P. V. Goeddel, and T. W. Mak, Early lethality, functional NF-kappaB activation, and increased sensitivity to TNF-induced cell death in TRAF2-deficient mice. Immunity 7, 715 – 725 (1997). Y. Xu, G. Cheng, and D. Baltimore, Targeted disruption of TRAF3 leads to postnatal lethality and defective T-dependent immune responses. Immunity 5, 407 – 415 (1996). M. A. Lomaga, W. C. Yeh, I. Sarosi, G. S. Duncan, C. Furlonger, A. Ho, S. Morony, C. Capparelli, G. Van, S. Kaufman, H. A. van der, A. Itie, A. Wakeham, W. Khoo, T. Sasaki, Z. D. Cao, J. M. Penninger, C. J. Paige, D. L. Lacey, C. R. Dunstan, W. J. Boyle, D. V. Goeddel, and T. W. Mak, TRAF6 deficiency results in osteopetrosis and defective interleukin-1, CD40, and LPS signaling. Genes Dev. 13, 1015 – 1024 (1999). B. R. Wong, D. Besser, N. Kim, J. R. Arron, M. Vologodskaia, H. Hanafusa, and Y. Choi, TRANCE, a TNF family member, activates Akt/PKB through a signaling complex involving TRAF6 and c-src. Mol. Cell 4, 1041 – 1049 (1999). P. J. Coffer, J. Jin, and J. R. Woodgett, Protein kinase B (c-Akt): A multifunctional mediator of phosphatidylinositol 3-kinase activation. Biochem. J. 335, 1 – 13 (1998). J. P. T. M. Van Leeuwen, J. C. Birkenhager, M. P. Bos, G. J. Van Der Bemd, M. P. Herrmann-Erlee, and H. A. P. Pols, Parathyroid hormone sensitizes long bones to the stimulation of bone resorption by 1,25-dihydroxyvitamin D3. J. Bone Miner. Res. 7, 303 – 309 (1992). J. A. Delmez, A. S. Dusso, E. Slatopolsky, and S. L. Teitelbaum, Modulation of renal osteodystrophy by extrarenal production of calcitriol. Am. J. Nephrol. 15, 85 – 89 (1995). G. J. King, M. E. Holtrop, and L. G. Raisz, The relation of ultrastructural changes in osteoclasts to resorption in bone cultures stimulated with parathyroid hormone. Metabol. Bone Dis. Rel. Res. 1, 67 (1978). E. Barengolts, R. Buschmann, D. H. Shevrin, E. C. Abramson, and S. C. Kukreja, Effects of hypercalcemia-producing tumor extract and parathyroid hormone on osteoclast ultrastructure. Acta Anat. 137, 160 – 164 (1991). K. Lee, J. D. Deeds, S. Chiba, M. Un-No, A. T. Bond, and G. V. Segre, Parathyroid hormone induces sequential c-fos expression in bone cells in vivo: In situ localization of its receptor and c-fos messenger ribonucleic acids. Endocrinology 134, 441 – 450 (1994). C. M. Silve, G. T. Hradek, A. L. Jones, and C. D. Arnaud, Parathyroid hormone receptor in intact embryonic chicken bone: Characterization and cellular localization. J. Cell Biol. 94, 379 – 386 (1982). M. F. Rouleau, H. Warshawsky, and D. Goltzman, Parathyroid hormone binding in vivo to renal, hepatic, and skeletal tissues of the rat using a radiographic approach. Endocrinology 118, 919 – 931 (1986).
100 161. T. J. Chambers, N. A. Athanasou, and K. Fuller, Effect of parathyroid hormone and calcitonin on the cytoplasmic spreading of isolated osteoclasts. J. Endocrinol. 102, 281 – 286 (1984). 162. T. J. Chambers, P. M. J. McSheehy, B. M. Thomson, and K. Fuller, The effect of calcium-regulating hormones and prostaglandins on bone resorption by osteoclasts disaggregated from neonatal rabbit bones. Endocrinology 116, 234 – 239 (1985). 163. J. Vilamitjana-Amedee, R. Bareille, F. Rouais, A. I. Caplan, and M. F. Harmand, Human bone marrow stromal cells express an osteoblastic phenotype in culture. In Vitro Cell Dev. Biol. Anim. 29A, 699 – 707 (1993). 164. N. Shevde, P. Anklesaria, J. S. Greenberger, I. Bleiberg, and J. Glowacki, Stromal cell-mediated stimulation of osteoclastogenesis. Proc. Natl. Acad. Sci. USA 205, 306 – 315 (1994). 165. H. M. I. Perry, W. Skogen, J. C. Chappel, G. D. Wilner, A. J. Kahn, and S. L. Teitelbaum, Conditioned medium from osteoblast-like cells mediate parathyroid hormone induced bone resorption. Calcif. Tissue Int. 40, 298 – 300 (1987). 166. H. M. I. Perry, W. Skogen, J. Chappel, A. J. Kahn, G. D. Wilner, and S. L. Teitelbaum, Partial characterization of a parathyroid hormonestimulated factor(s) from osteoblast-like cells. Endocrinology 125, 2075 – 2082 (1989). 167. S. K. Lee and J. A. Lorenzo, Parathyroid hormone stimulates TRANCE and inhibits osteoprotegerin messenger ribonucleic acid expression in murine bone marrow cultures: Correlation with osteoclast-like cell formation. Endocrinology 140, 3552 – 3561 (1999). 168. G. V. Segre, A. B. Abou-Samra, H. Juppner, E. Schipani, T. Force, P. Urena, M. Freeman, X. F. Kong, L. F. Jr. Kolakowski, and J. Hock, Characterization of cloned PTH/PTHrP receptors. J. Endocrinol. Invest. 15, 11 – 17 (1992). 169. E. Blind, F. Raue, V. Knappe, J. Schroth, and R. Siegler, Cyclic AMP formation in rat bone and kidney cells is stimulated equally by parathyroid hormone-related protein (PTHrP) 1 – 34 and PTH 1 – 34. Exp. Clin. Endocrinol. 101, 150 – 155 (1993). 170. A. J. Fenton, T. J. Martin, and G. C. Nicholson, Carboxyl-terminal parathyroid hormone-related protein inhibits bone resorption by isolated chicken osteoclasts. J. Bone Miner. Res. 9, 515 – 519 (1994). 171. T. Sone, H. Kohno, H. Kikuchi, T. Ikeda, R. Kasai, Y. Kikuchi, R. Takeuchi, J. Konishi, and C. Shigeno, Human parathyroid hormonerelated peptide(107 – 111) does not inhibit bone resorption in neonatal mouse calvariae. Endocrinology 131, 2742 – 2746 (1992). 172. H. Kaji, T. Sugimoto, M. Kanatani, A. Miyauchi, T. Kimura, S. Sakakibara, M. Fukase, and K. Chihara, Carboxyl-terminal parathyroid hormone fragments stimulate osteoclast-like cell formation and osteoclastic activity. Endocrinology 134, 1897 – 1904 (1994). 173. T. Tamura, N. Udagawa, N. Takahashi, C. Miyaura, S. Tanaka, Y. Yamada, Y. Koishihara, Y. Ohsugi, K. Kumaki, and T. Taga, Soluble interleukin-6 receptor triggers osteoclast formation by interleukin 6. Proc. Natl. Acad. Sci. USA 90, 11924 – 11928 (1993). 174. N. J. Horwood, V. Kartsogiannis, J. M. Quinn, E. Romas, T. J. Martin, and M. T. Gillespie, Activated T lymphocytes support osteoclast formation in vitro. Biochem. Biophys. Res. Commun. 265, 144 – 150 (1990). 175. Y. Y. Kong, U. Feige, I. Sarosi, B. Bolon, A. Tafuri, S. Morony, C. Capparelli, J. Li, R. Elliott, S. McCabe, T. Wong, G. Campagnuolo, E. Moran, E. R. Bogoch, G. Van, L. T. Nguyen, P. S. Ohashi, D. L. Lacey, E. Fish, W. J. Boyle, and J. M. Penninger, Activated T cells regulate bone loss and joint destruction in adjuvant arthritis through osteoprotegerin ligand. Nature 402, 304 – 309 (1999). 176. N. Takahashi, N. Udagawa, S. Tanaka, H. Murakami, I. Owan, T. Tamura, and T. Suda, Postmitotic osteoclast precursors are mononuclear cells which express macrophage-associated phenotypes. Dev. Biol. 163, 212 – 221 (1994).
ROSS AND TEITELBAUM 177. S. Tanaka, N. Takahashi, N. Udagawa, T. Tamura, T. Akatsu, E. R. Stanley, T. Kurokawa, and T. Suda, Macrophage colony-stimulating factor is indispensable for both proliferation and differentiation of osteoclast progenitors. J. Clin. Invest. 91, 257 – 263 (1993). 178. J. Owens and T. J. Chambers, Macrophage colony-stimulating factor (M-CSF) induces migration in osteoclasts in vitro. Biochem. Biophys. Res. Commun. 195, 1401 – 1407 (1993). 179. G. Hattersley, E. Dorey, M. A. Horton, and T. J. Chambers, Human macrophage colony-stimulating factor inhibits bone resorption by osteoclasts disaggregated from rat bone. J. Cell. Physiol. 137, 199 – 203 (1988). 180. P. C. Heinrich, I. Behrmann, G. Muller-Newen, F. Schaper, and L. Graeve, Interleukin-6-type cytokine signalling through the gp130/ Jak/STAT pathway. Biochem. J. 334, 297 – 314 (1998). 181. G. Girasole, G. Passeri, R. L. Jilka, and S. C. Manolagas, Interleukin-11: A new cytokine critical for osteoclast development. J. Clin. Invest. 93, 1516 – 1524 (1994). 182. E. Romas, N. Udagawa, H. Zhou, T. Tamura, M. Saito, T. Taga, D. J. Hilton, T. Suda, K. W. Ng, and T. J. Martin, The role of gp130-mediated signals in osteoclast development: regulation of interleukin 11 production by osteoblasts and distribution of its receptor in bone marrow cultures. J. Exp. Med. 183, 2581 – 2591 (1996). 183. C. B. Ware, M. C. Horowitz, B. R. Renshaw, J. S. Hunt, D. Liggitt, S. A. Koblar, B. C. Gliniak, H. J. McKenna, T. Papayannopoulou, B. Thoma, et al. Targeted disruption of the low-affinity leukemia inhibitory factor receptor gene causes placental, skeletal, neural and metabolic defects and results in perinatal death. Development 121, 1283 – 1299 (1995). 184. N. Malik, H. S. Haugen, B. Modrell, M. Shoyab, and C. H. Clegg, Developmental abnormalities in mice transgenic for bovine oncostatin M. Mol. Cell. Biol. 15, 2349 – 2358 (1995). 185. K. Kawasaki, Y. H. Gao, S. Yokose, Y. Kaji, T. Nakamura, T. Suda, K. Yoshida, T. Taga, T. Kishimoto, H. Kataoka, T. Yuasa, H. Norimatsu, and A. Yamaguchi, Osteoclasts are present in gp130-deficient mice. Endocrinology 138, 4959 – 4965 (1997). 186. G. D. Roodman, N. Kurihara, Y. Ohsaki, A. Kukita, D. Hosking, A. Demulder, J. F. Smith, and F. R. Singer, Interleukin 6. A potential autorcrine/paracrine factor in Paget’s disease of bone. J. Clin. Invest. 89, 46 – 52 (1992). 187. N. Kurihara, S. V. Reddy, C. Menaa, D. Anderson, and G. D. Roodman, Osteoclasts formed by normal human bone marrow cells transduced with the measles virus nucleocapsid gene express a pagetic phenotype. J. Clin. Invest. 105, 607– 614 (2000). 188. S. V. Reddy, S. Takahashi, M. Dallas, R. E. Williams, L. Neckers, and G. D. Roodman, Interleukin-6 antisense deoxyoligonucleotides inhibit bone resorption by giant cells from human giant cell tumors of bone. J. Bone Miner. Res. 9, 753 – 757 (1994). 189. R. Kitazawa, R. B. Kimble, J. L. Vannice, V. T. Kung, and R. Pacifici, Interleukin-1 receptor antagonist and tumor necrosis factor binding protein decrease osteoclast formation and bone resorption in ovariectomized mice. J. Clin. Invest. 94, 2397 – 2406 (1994). 190. R. L. Jilka, G. Hangoc, G. Girasole, G. Passeri, D. C. Williams, J. S. Abrams, B. Boyce, H. Broxmeyer, and S. C. Manolagas, Increased osteoclast development after estrogen loss: Mediation by interleukin6. Science 257, 88 – 91 (1992). 191. G. Girasole, R. L. Jilka, G. Passeri, S. Boswell, G. Boder, D. C. Williams, and S. C. Manolagas, 17 beta-estradiol inhibits interleukin-6 production by bone marrow-derived stromal cells and osteoblasts in vitro: A potential mechanism for the antiosteoporotic effect of estrogens. J. Clin. Invest. 89, 883 – 891 (1992). 192. V. Poli, R. Balena, E. Fattori, A. Markatos, M. Yamamoto, H. Tanaka, G. Ciliberto, G. A. Rodan, and F. Costantini, Interleukin-6 deficient mice are protected from bone loss caused by estrogen depletion. EMBO J. 13, 1189 – 1196 (1994).
CHAPTER 3 Osteoclast Biology 193. A. Grey, M. A. Mitnick, U. Masiukiewicz, B. H. Sun, S. Rudikoff, R. L. Jilka, S. C. Manolagas, and K. Insogna, A role for interleukin-6 in parathyroid hormone-induced bone resorption in vivo. Endocrinology 140, 4683 – 4690 (1999). 194. L. Sun, O. A. Adebanjo, B. S. Moonga, S. Corisdeo, H. K. Anandatheerthavarada, G. Biswas, T. Arakawa, Y. Hakeda, A. Koval, B. Sodam, P. J. Bevis, A. J. Moser, F. A. Lai, S. Epstein, B. R. Troen, M. Kumegawa, and M. Zaidi, CD38/ADP-ribosyl cyclase: A new role in the regulation of osteoclastic bone resorption. J. Cell Biol. 146, 1161 – 1172 (1999). 195. O. A. Adebanjo, B. S. Moonga, T. Yamate, L. Sun, C. Minkin, E. Abe, and M. Zaidi, Mode of action of interleukin-6 on mature osteoclasts: Novel interactions with extracellular Ca2 sensing in the regulation of osteoclastic bone resorption. J. Cell Biol. 142, 1347 – 1356 (1998). 196. W. P. Arend, H. G. Welgus, R. C. Thompson, and S. P. Eisenberg, Biological properties of recombinant human monocyte-derived interleukin 1 receptor antagonist. J. Clin. Invest. 85, 1694 – 1697 (1990). 197. R. B. Kimble, J. L. Vannice, D. C. Bloedow, R. C. Thompson, W. Hopfer, V. T. Kung, C. Brownfield, and R. Pacifici, Interleukin-1 receptor antagonist decreases bone loss and bone resorption in ovariectomized rats. J. Clin. Invest. 93, 1959 – 1967 (1994). 198. M. E. Choen-Solal, A. M. Graulet, M. A. Denne, J. Gueris, D. Baylink, and M. C. de Vernejoul Peripheral monocyte culture supernatants of menopausal women can induce bone resorption: Involvement of cytokines. J. Clin. Endocrinol. Metab. 77, 1648 – 1653 (1993). 199. P. Ammann, R. Rizzoli, J. P. Bonjour, S. Bourrin, J. M. Meyer, P. Vassalli, and I. Garcia, Transgenic mice expressing soluble tumor necrosis factor-receptor are protected against bone loss caused by estrogen deficiency. J. Clin. Invest. 99, 1699 – 1703 (1997). 200. R. Felix, H. Fleisch, and P. R. Elford, Bone-resorbing cytokines enhance release of macrophage colony-stimulating activity by osteoblastic cell MC3T3-E1. Calcif. Tissue Int. 44, 356 – 360 (1989). 201. J. Glowacki, N. Shevde, and J. S. Greenberger, Distinct roles of interleukins-1,-6, and-11 in osteoclastic differentiation of hemopoietic progenitors. J. Bone Miner. Res. 8, A190 (1993). 202. D. D. Chaplin and K. Hogqist, Interactions between TNF and interleukin-1. In “Tumor Necrosis Factors.” The Molecules and Their Emerging Role in Medicine” (B. Beutler, ed.), pp. 197 – 220. Raven Press, New York. 203. Y. Sakagami, G. Girasole, X. P. Yu, H. S. Boswell, and S. C. Manolagas, Stimulation of interleukin-6 production by either calcitonin gene-related peptide or parathyroid hormone in two phenotypically distinct bone marrow-derived murine stromal cell lines. J. Bone Miner. Res. 8, 811 – 816 (1993). 204. K. Fuller, J. M. Owens, and T. J. Chambers, Macrophage inflammatory protein-1 alpha and IL-8 stimulate the motility but suppress the resorption of isolated rat osteoclasts. J. Immunol. 154, 6065 – 6072 (1995). 205. L. Rothe, P. Collin-Osdoby, Y. Chen, T. Sunyer, L. Chaudhary, A. Tsay, S. Goldring, L. Avioli, and P. Osdoby, Human osteoclasts and osteoclast-like cells synthesize and release high basal and inflammatory stimulated levels of the potent chemokine interleukin-8. Endocrinology 139, 4353 – 4363 (1998). 206. P. Collin-Osdoby, D. Kirsch, F. Anderson, O. Joost, A. Dean and P. Osdoby, The chemokine IL-8 as an autocrine inhibitor of osteoclast bone resorptive activity via IL-8 receptors expressed by avian osteoclasts and human osteoclast-like cells. J. Bone Miner. Res. 11, S179 (1996). 207. P. A. Collin-Osdoby, L. Rothe, S. Bekker, and F. Anderson, Osteoclast recruitment, formation, and bone resorptive activity are increased in parallel with neoangiogenesis on the chick chorioallantoic membrane in response to the angiogenic peptides bFGF or IL-8. Bone 23, S550 (1998).
101 208. C. Christiansen, M. S. Christensen, and I. Transbol, Bone mass in postmenopausal women after withdrawal of oestrogen/gestagen replacement therapy. Lancet 1, 459 – 461 (1981). 209. E. P. Smith, J. Boyd, G. R. Frank, H. Takahashi, R. M. Cohen, B. Specker, T. C. Williams, D. B. Lubahn, and K. S. Korach, Estrogen resistance caused by a mutation in the estrogen-receptor gene in a man. N. Engl. J. Med. 331, 1056 – 1061 (1994). 210. J. P. Bilezikian, A. Morishima, J. Bell, and M. M. Grumbach, Increased bone mass as a result of estrogen therapy in a man with aromatase deficiency. N. Engl. J. Med. 339, 599 – 603 (1998). 211. E. R. Simpson, Genetic mutations resulting in estrogen insufficiency in the male. Mol. Cell. Endocrinol. 145, 55 – 59 (1998). 212. R. Civitelli, S. Gonnelli, F. Zacchei, S. Bigazzi, A. Vattimo, L. V. Avioli, and C. Gennari, Bone turnover in postmenopausal osteoporosis: Effect of calcitonin treatment. J. Clin. Invest. 82, 1268 – 1274 (1988). 213. R. T. Turner, G. K. Wakley, K. S. Hannon, and N. H. Bell, Tamoxifen prevents the skeletal effects of ovarian hormone deficiency in rats. J. Bone Miner. Res. 2, 449 – 456 (1987). 214. R. T. Turner, G. K. Wakley, K. S. Hannon, and N. H. Bell, Tamoxifen inhibits osteoclast-mediated resorption of trabecular bone in ovarian hormone-deficient rats. Endocrinology 122, 1146 – 1150 (1988). 215. D. Rickard, G. Russell, and M. Gowen, Oestradiol inhibits the release of tumour necrosis factor but not interleukin 6 from adult human osteoblasts in vitro. Osteoporos. Int. 2, 94 – 102 (1992). 216. L. R. Chaudhary, T. C. Spelsberg, and B. L. Riggs, Production of various cytokines by normal human osteoblast-like cells in response to interleukin-1 and tumor necrosis factor-: Lack of regulation by 17-estradiol. Endocrinology 130, 2528 – 2534 (1992). 217. G. Pioli, G. Basini, M. Pedrazzoni, G. Musetti, V. Ulietti, D. Bresciani, P. Villa, A. Bacci, G. Hughes, G. Russell, and M. Passeri, Spontaneous release of Interleukin-1 and interleukin-6 by peripheral blood monocytes after ovariectomy. Clin. Sci. 83, 503 – 507 (1992). 218. S. H. Ralston, R. G. G. Russell, and M. Gowen, Estrogen inhibits release of tumor necrosis factor from peripheral blood mononuclear cells in post-menopausal women. J. Bone Miner. Res. 5, 983 – 988 (1990). 219. R. Pacifici, L. Rifas, R. McCracken, I. Vered, C. McMurty, L. V. Avioli, and W. A. Peck, Ovarian steroid treatment blocks a postmenopausal increase in blood monocyte interleukin 1 release. Proc. Natl. Acad. Sci. USA 86, 2398 – 2402 (1989). 220. R. Pacifici, C. Brown, E. Puscheck, E. Friedrich, E. Slatopolsky, D. Maggio, R. McCracken, and L. V. Avioli, Effect of surgical menopause and estrogen replacement on cytokine release from human blood mononuclear cells. Proc. Natl. Acad. Sci. USA 88, 5134 – 5138 (1991). 221. R. Pacifici, J. L. Vannice, L. Rifas, and R. D. Kimble, Monocytic secretion of interleukin-1 receptor antagonist in normal and osteoporotic women: Effects of menopause and estrogen/progesterone therapy. J. Clin. Endocrinol. Metab. 77, 1135 – 1141 (1993). 222. T. Sunyer, J. Lewis, P. Collin-Osdoby, and P. Osdoby, Estrogen’s boneprotective effects may involve differential IL-1 receptor regulation in human osteoclast-like cells. J. Clin. Invest. 103, 1409 – 1418 (1999). 223. Y. Tanaka and H. F. DeLuca, Bone mineral mobilization of 1,25(OH)2D3, a metabolite of vitamin D. Arch. Biochem. Biophys. 146, 574 – 578 (1971). 224. J. J. Reynolds, H. Pavlovitch, and S. Falsan, 1,25-Dihydroxycholecalciferol increases bone resorption in thyroparathyroidectomized mice. Calcif. Tissue Res. 21, 207 – 212 (1976). 225. L. G. Raisz, C. L. Trummel, M. F. Holick, and H. F. DeLuca, 1,25Dihydroxycholecalciferol: A potent stimulator of bone resorption in tissue culture. Science 175, 768 – 769 (1972). 226. R. Kitazawa, S. Kitazawa, and S. Maeda, Promoter structure of mouse RANKL/TRANCE/OPGL/ODF gene. Biochim. Biophys. Acta 1445, 134 – 141 (1990).
102 227. J. Abe, M. Morikawa, and K. Miyamoto, Synthetic analogues of vitamin D3 with an oxygen atom in the side chain skeleton: A trial of the development of vitamin D compounds which exhibit potent differentiation-inducing activity without inducing hypercalcemia. FEBS Lett. 226, 58 – 62 (1987). 228. M. Amling, M. Priemel, T. Holzmann, K. Chapin, J. M. Rueger, R. Baron, and M. B. Demay, Rescue of the skeletal phenotype of vitamin D receptor-ablated mice in the setting of normal mineral ion homeostasis: Formal histomorphometric and biomechanical analyses. Endocrinology 140, 4982 – 4987 (1999). 229. D. L. Lacey, L. E. Grosso, S. A. Moser, J. Erdmann, H. L. Tan, R. Pacifici and D. T. Villareal, IL-1-induced murine osteoblast IL-6production is mediated by the type 1 IL-1 receptor and is increased by 1,25-dihydroxyvitamin D3. J. Clin. Invest. 91, 1731 – 1742 (1993). 230. T. Bellido, G. Girasole, G. Passeri, X. P. Yu, H. Mocharla, R. L. Jilka, A. Notides, and S. C. Manolagas, Demonstration of estrogen and vitamin D receptors in bone marrow-derived stromal cells: Upregulation of the estrogen receptor by 1,25-dihydroxyvitamin D3. Endocrinology 133, 553 – 562 (1993). 231. E. Seeman, H. W. Wagner, K. P. Offord, R. Kumar, W. J. Johnson, and B. L. Riggs, Differential effects of endocrine dysfunction on the axial and appendicular skeleton. J. Clin. Invest. 69, 1302 – 1309 (1982). 232. J. C. Gallagher, J. Aaron, A. Horsman, R. Wilkinson, and B. E. Nordin, Corticosteroid osteoporosis. Clin. Endocrinol. Metab. 2, 255 – 368 (1973). 233. T. J. Hahn, L. R. Halstead, S. L. Teitelbaum, and B. H. Hahn, Altered mineral metabolism in glucocorticoid-induced osteopenia. J. Clin. Invest. 64, 655 – 665 (1979). 234. D. W. Dempster, M. A. Arlot, and P. J. Meunier, Mean wall thickness and formation periods of trabecular bone packets in corticosteroidinduced osteoporosis. Calcif. Tissue Int. 35, 410 – 417 (1983). 235. A. G. Need, J. C. Philcox, J. C. Hartley, and B. E. Nordin, Calcium metabolism and osteoporosis in corticsteroid-treated postmenopausal women. Aust. N. Z. J. Med. 16, 341 – 346 (1986). 236. J. Tobias and T. J. Chambers, Glucocorticoids impair bone resorptive activity and viability of osteoclasts disaggregated from neonatal rat long bones. Endocrinology 125, 1290 – 1295 (1989). 237. G. L. Wong, Basal activities and hormone responsiveness of osteoclast-like and osteoblast-like bone cells are regulated by glucocorticoids. J. Biol. Chem. 254, 6337 – 6340 (1979). 238. C. Colette, L. Monnier, N. Pares Herbute, F. Blotman, and J. Mirouze, Calcium absorption in corticoid treated subjects: Effects of a single oral dose of calcitriol. Horm. Metab. Res. 19, 335 – 338 (1987). 239. H. K. Nielsen, K. Thomsen, E. F. Eriksen, P. Charles, T. Storm, and L. Mosekilde, The effect of high-dose glucocorticoid administration on serum bone gamma carboxyglutamic acid-containing protein, serum alkaline phosphatase and vitamin D metabolites in normal subjects. Bone Miner. 4, 105 – 113 (1988). 240. Y. Suzuki, Y. Ichikawa, E. Saito, and M. Homma, Importance of increased urinary calcium excretion in the development of secondary hyperparathyroidism of patients under glucocorticoid therapy. Metab. Clin. Exp. 32, 151 – 156 (1983). 241. G. L. Wong, B. P. Lukert, and J. S. Adams, Glucocorticoids increase osteoblast-like bone cell response to 1,25(OH)2D3. Nature 285, 254 – 257 (1980). 242. S. C. Kukreja, E. N. Bowser, G. K. Hargis, W. J. Henderson, and G. A. Williams, Mechanism of glucocorticoid-induced osteopenia: Role of parathyroid glands. Proc. Soc. Exp. Biol. Med. 152, 358 – 361 (1976). 243. H. Ishikawa, H. Tanaka, K. Iwato, O. Tanabe, H. Asaoku, M. Nobuyoshi, I. Yamamoto, M. Kawano, and A. Kuramoto, Effect of glucocorticoids on the biologic activities of myeloma cells: Inhibition of interleukin-1 osteoclast activating factor-induced bone resorption. Blood 75, 715 – 720 (1990).
ROSS AND TEITELBAUM 244. J. L. Masferrer, S. T. Reddy, B. S. Zweifel, K. Seibert, P. Needleman, R. S. Gilbert, and H. R. Herschman, In vivo glucocorticoids regulate cyclooxygenase-2 but not cyclooxygenase-1 in peritoneal macrophages. J. Pharmacol. Exp. Ther, 270, 1340 – 1344 (1994). 245. S. A. Fraser, J. B. Anderson, D. A. Smith, and G. M. Wilson, Osteoporosis and fractures following thyrotoxicosis. Lancet 1, 981 – 983 (1971). 246. M. F. Fallon, H. M. I. Perry, M. Bergfeld, D. Droke, L. V. Avioli, and S. L. Teitelbaum, Exogenous hyperthyroidism with osteoporosis. Arch. Intern. Med. 143, 442 – 444 (1983). 247. M. Yamamoto, A. Markatos, J. G. Seedor, P. Masarachia, M. Gentile, G. A. Rodan, and R. Balena, The effects of the amino bisphosphonate alendronate on thyroid hormone-induced osteopenia in rats. Calcif. Tissue Int. 53, 278 – 282 (1993). 248. J. J. Stepan and Z. Limanova, Biochemical assessment of bone loss in patients on long-term thyroid hormone treatment. Bone Miner. 17, 377 – 388 (1992). 249. R. H. Follis, Skeletal changes associated with hyperthyroidism. Bull. Johns Hopkins Hosp. 92, 405 (1953). 250. J. C. Krakauer and M. Kleerekoper, Borderline-low serum thyrotropin level is correlated with increased fasting urinary hydroxyproline excretion. Arch. Intern. Med. 152, 360 – 364 (1992). 251. E. DeMenis, G. Da Rin, I. Roiter, P. Legovini, G. Foscolo, and N. Conte, Bone turnover in overt and subclinical hyperthyroidism due to autonomous thyroid hormone. Horm. Res. 37, 217 – 220 (1992). 252. H. Kawaguchi, C. C. Pilbeam, F. N. Woodiel, and L. G. Raisz, Comparison of the effects of 3,5,3’-triiodothyroacetic acid and triiodothyronine on bone resorption in cultured fetal rat long bones and neonetal mouse calvariae. J. Bone Miner. Res. 9, 247 – 253 (1994). 253. G. R. Mundy, J. L. Shapiro, J. G. Bandelin, E. M. Canalis, and L. G. Raisz, Direct stimulation of bone resorption by thyroid hormones. J. Clin. Invest. 58, 529 – 534 (1976). 254. J. M. Britto, A. J. Fenton, W. R. Holloway, and G. C. Nicholson, Osteoblasts mediate thyroid hormone stimulation of osteoclastic bone resorption. Endocrinology 134, 169 – 176 (1994). 255. P. Charles, J. W. Poser, and L. Mosekilde, Estimation of bone turnover evaluated by 47Ca-kinetics: Efficiency of serum bone gamma-carboxyglutamic acid-containing protein, serum alkaline phosphatase, and urinary hydroxyproline excretion. J. Clin. Invest. 76, 2254 – 2258 (1985). 256. P. Garnero, V. Vassy, A. Bertholin, J. P. Riou, and P. D. Delmas, Markers of bone turnover in hyperthyroidism and the effects of treatment. J. Clin. Endocrinol. Metab. 78, 955 – 959 (1994). 257. R. D. Harvey, K. C. McHardy, I. W. Reid, F. Paterson, P. D. Bewsher, A. Duncan, and S. P. Robins, Measurement of bone collagen degradation in hyperthyroidism and during thyroxine replacement therapy using pyridinium crosslinks as specific urinary markers. J. Clin. Endocrinol. Metab. 72, 1189 – 1194 (1991). 258. S. Saito, K. Yamasaki, S. Yamada, A. Matsumoto, T. Akatsu, N. Takahashi, Y. Shibasaki, T. Suda, and T. Fukuhara, A stable analogue of thromboxane A2, 9,11-epithio-11,12-methanothromboxane A2, stimulates bone resorption in vitro and osteoclast-like cell formation in mouse marrow culture. Bone Miner. 12, 15 – 23 (1991). 259. W. E. Gallwitz, G. R. Mundy, C. H. Lee, M. Qiao, G. D. Roodman, M. Raftery, S. J. Gaskell, and L. F. Bonewald, 5-lipoxygenase metabolites of arachidonic acid stimulate isolated osteoclasts to resorb calcified matrices. J. Biol. Chem. 268, 10087 – 10094 (1993). 260. S. Meghji, J. R. Sandy, A. M. Scutt, W. Harvey, and M. Harris, Stimulation of bone resorption by lipoxygenase metabolites of arachidonic acid. Prostaglandins 36, 139 – 149 (1988). 261. K. Imamura, H. Ozawa, T. Hiraide, N. Takahashi, Y. Shibasaki, T. Fukuhara, and T. Suda, Continuously applied compressive pressure induces bone resorption by a mechanism involving prostaglandin E2 synthesis. J. Cell. Physiol. 144, 222 – 228 (1990).
CHAPTER 3 Osteoclast Biology 262. J. L. Saffar and P. Leroux, Role of prostaglandins in bone resorption in a synchronized remodeling in the rat. Bone 9, 141 – 145 (1988). 263. J. R. Nefussi and R. Baron, PGE2 stimulates both resorption and formation of bone in vitro: Differential responses of the periosteum and the endosteum in fetal rat long bone cultures. Anat. Rec. 211, 9 – 16 (1985). 264. M. R. Wani, K. Fuller, N. S. Kim, Y. Choi, and T. Chambers, Prostaglandin E2 cooperates with TRANCE in osteoclast induction from hemopoietic precursors: Synergistic activation of differentiation, cell spreading, and fusion. Endocrinology 140, 1927 – 1935 (1999). 265. C. S. Lader and A. M. Flanagan, Prostaglandin E2, interleukin 1 alpha, and tumor necrosis factor-alpha increase human osteoclast formation and bone resorption in vitro. Endocrinology 139, 3157 – 3164 (1998). 266. T. J. Chambers and N. N. Ali, Inhibition of osteoclastic motility by prostaglandins I2, E2 and 6-oxoE1/. J. Pathol. 139, 383 – 397 (1983). 267. P. T. Lakkakorpi and H. K. Vaananen, Calcitonin, prostaglandin E2 and dibutyrl cyclic adenosine 3’,5’-monophosphate disperse the specific microfilament structure in resorbing osteoclasts. J. Histochem. Cytochem. 38, 1487 – 1493 (1990). 268. G. C. Nicholson, S. A. Livesey, J. M. Moseley, and T. J. Martin, Actions of calcitonin, parathyroid hormone, and prostaglandin E2 on cyclic AMP formation in chicken and rat osteoclasts. J. Cell. Biochem. 31, 229 – 241 (1986). 269. G. C. Nicholson, J. M. Moseley, A. J. P. Yates, and T. J. Martin, Control of cyclic adenosine 3’,5’-monophosphate production in osteoclasts: Calcitonin-induced persistent activation and homologous densensitization of adenylate cyclase. Endocrinology 120, 1902 – 1908 (1987). 270. J. Kaneshisa, Time course of “escape” from calcitonin-induced inhibition of motility and resorption of disaggregated osteoclasts. Bone 10, 125 – 129 (1989). 271. D. H. Copp, A. G. F. Davidson, and B. Cheney, Evidence for a new parathyroid hormone which lowers blood calcium. Proc. Can. Fed. Biol. Sci. 4, 17 (1961). 272. D. Gonzalez, G. Ghiringhelli, and C. Mautalen, Acute antiosteoclastic effect of salmon calcitonin in osteoporotic women. Calcif. Tissue Int. 38, 71 – 75 (1986). 273. G. C. Nicholson, J. M. Moseley, P. M. Sexton, F. A. O. Mendelsohn, and T. J. Martin, Abundant calcitonin receptors in isolated rat osteoclasts. J. Clin. Invest. 78, 355 – 360 (1986). 274. S. R. Goldring, M. Roelke, and J. Glowacki, Multinucleated cells elicited in response to implants of devitalized bone particles possess receptors for calcitonin. J. Bone Miner. Res. 3, 117 – 120 (1988). 275. A. Vignery, M. J. Raymond, H. Y. Qian, F. Wang, and S. A. Rosenzweig, Multinucleated rat alveolar macrophages express functional receptors for calcitonin. Am. J. Physiol. 261, F1026 – F1032 (1991). 276. A. J. Fenton, T. J. Martin, and G. C. Nicholson, Long-term culture of disaggregated rat osteoclasts: Inhibition of bone resorption and reduction of osteoclast-like cell number by calcitonin and PTHrP[107 – 139]. J. Cell. Physiol. 155, 1 – 7 (1993). 277. D. M. Kallio, P. R. Garant, and C. Minkin, Ultrastructural effects of calcitonin on osteoclasts in tissue cultures. J. Ultrastruct. Res. 39, 205 – 216 (1972). 278. S. Yumita, G. C. Nicholson, D. J. Rowe, G. N. Kent, and T. J. Martin, Biphasic effect of calcitonin on tartrate-resistant acid phosphatase activity in isolated rat osteoclasts. J. Bone Miner. Res. 6, 591 – 597 (1991). 279. H. Kaji, T. Sugimoto, A. Miyauchi, M. Fukase, K. Tezuka, Y. Hakeda, M. Kumegawa, and K. Chihara, Calcitonin inhibits osteopontin mRNA expression in isolated rabbit osteoclasts. Endocrinology 135, 484 – 487 (1994). 280. S. Pandalai, and C. V. Gay, Effects of parathyroid hormone, calcitonin, and dibutyrl-cyclic AMP on osteoclast area in cultured chick tibia. J. Bone Miner. Res. 5, 701 – 705 (1990).
103 281. T. J. Chambers, and A. Moore, The sensitivity of isolated osteoclasts to morphological transformation by calcitonin. J. Clin. Endocrinol. Metab. 57, 819 – 824 (1983). 282. M. H. Zheng, J. M. Papadimitriou, and G. S. Nicholson, A quantitative cytochemical investigation of osteoclasts and multinucleated giant cells. Histochem. J. 23, 180 – 188 (1991). 283. M. Zaidi, H. K. Datta, B. S. Moonga, and I. Maclntyre, Evidence that the action of calcitonin on rat osteoclasts is mediated by two G proteins acting via separate post-receptor pathways. J. Endocrinol. 126, 473 – 481 (1990). 284. H. H. Messer and D. H. Copp, Changes in response to calcitonin following prolonged administration to intact rats. Proc. Soc. Exp. Biol. Med. 146, 643 – 647 (1974). 285. J. F. Obie and C. W. Cooper, Loss of calcemic effects of calcitonin and parathyroid hormone infused continuously into rats using the alzet osmotic minipumps. J. Pharmacol. Exp. Ther. 209, 422 – 428 (1979). 286. N. Glajchen, S. Thomas, P. Jowell, S. Epstein, F. Ismail, and M. Fallon, The effect of high dose salmon calcitonin on bone mineral metabolism in the normal rat. Calcif. Tissue Int. 46, 28 – 32 (1990). 287. Y. Su, M. Chakraborty, M. H. Nathanson, and R. Baron, Differential effects of the 3’,5’-cyclic adenosine monophosphate and protein kinase C pathways on the response of isolated rat osteoclasts to calcitonin. Endocrinology 131, 1497 – 1502 (1992). 288. S. Takahashi, S. Goldring, M. Katz, S. Hilsenbeck, R. Williams, and G. D. Roodman, Downregulation of calcitonin receptor mRNA expression by calcitonin during human osteoclast-like cell differentiation. J. Clin. Invest. 95, 167 – 171 (1995). 289. B. S. Moonga, A. S. Alam, P. J. Bevis, F. Avaldi, C. L. Soncini, C. L. Huang, and M. Zaidi, Regulation of cytosolic free calcium in isolated rat osteoclasts by calcitonin. J. Endocrinol. 132, 241 – 249 (1992). 290. P. Collin-Osdoby, G. A. Nickols, and P. Osdoby, Bone cell function, regulation, and communication: Role for nitric oxide. J. Cell. Biochem. 56, 1 – 10 (1994). 291. L. L. Key, Jr., W. L. Ries, R. G. Taylor, B. D. Hays, and B. L. Pitzer, Oxygen derived free radicals in osteoclasts: The specificity and location of the nitroblue tetrazolium reaction. Bone 11, 115 – 119 (1990). 292. M. J. Steinbeck, W. H. Appel, Jr., A. J. Verhoeven, and M. J. Karnovsky, NADPH-oxidase expression and in situ production of superoxide by osteoclasts actively resorbing bone. J. Cell Biol. 126, 765 – 772 (1994). 293. S. T. Deahl, L. W. Oberley, T. D. Oberley, and J. H. Elwell, Immunohistochemical identification of superoxide dismutases, catalase, and glutathione-S-transferases in rat femora. J. Bone Miner. Res. 7, 187 – 198 (1992). 294. W. L. Ries, L. L. Key, Jr., and R. M. Rodriguiz, Nitroblue tetrazolium reduction and bone resorption by osteoclasts in vitro inhibited by a manganese-based superoxide dismutase mimic. J. Bone Miner. Res. 7, 931 – 939 (1992). 295. I. MacIntyre, M. Zaidi, A. S. Alam, H. K. Datta, B. S. Moonga, P. S. Lidbury, M. Hecker, and J. R. Vane, Osteoclastic inhibition: An action of nitric oxide not mediated by cyclic GMP. Proc. Natl. Acad. Sci. USA 88, 2936 – 2940 (1991). 296. T. P. Kasten, P. Collin-Osdoby, N. Patel, P. Osdoby, M. Krukowski, T. P. Misko, S. L. Settle, M. G. Currie, and G. A. Nickols, Potentiation of osteoclast bone-resorption activity by inhibition of nitric oxide synthase. Proc. Natl. Acad. Sci. USA 91, 3569 – 3573 (1994). 297. C. W. Lowik, P. H. Nibbering, M. van de Ruit, and S. E. Papapoulos, Inducible production of nitric oxide in osteoblast-like cells and in fetal mouse bone explants is associated with suppression of osteoclastic bone resorption. J. Clin. Invest. 93, 1465 – 1472 (1994). 298. P. D. Damoulis and P. V. Hauschka, Cytokines induce nitric oxide production in mouse osteoblasts. Biochem. Biophys. Res. Commun. 201, 924 – 931 (1994).
104 299. S. M. Hwang, C. A. Lopez, D. E. Heck, C. R. Gardner, D. L. Laskin, J. D. Laskin, and D. T. Denhardt, Osteopontin inhibits induction of nitric oxide synthase gene expression by inflammatory mediators in mouse kidney epithelial cells. J. Biol. Chem. 269, 711 – 715 (1994). 300. D. E. Hughes, A. Dai, J. C. Tiffee, H. H. Li, G. R. Mundy, and B. F. Boyce, Estrogen promotes apoptosis of murine osteoclasts mediated by TGF-beta. Nature Med. 2, 1132 – 1136 (1996). 301. S. P. Luckman, D. E. Hughes, F. P. Coxon, R. Graham, G. Russell, and M. J. Rogers, Nitrogen-containing bisphosphonates inhibit the mevalonate pathway and prevent post-translational prenylation of GTP-binding proteins, including Ras. J. Bone Miner. Res. 13, 581 – 589 (1998). 302. S. P. Luckman, F. P. Coxon, F. H. Ebetino, R. G. Russell, and M. J. Rogers, Heterocycle-containing bisphosphonates cause apoptosis and inhibit bone resorption by preventing protein prenylation: Evidence from structure-activity relationships in J774 macrophages. J. Bone Miner. Res. 13, 1668 – 1678 (1998). 303. R. G. Russell, M. J. Rogers, J. C. Frith, S. P. Luckman, F. P. Coxon, H. L. Benford, P. I. Croucher, C. Shipman, and H. A. Fleisch, The pharmacology of bisphosphonates and new insights into their mechanisms of action. J. Bone Miner. Res. 14(Suppl. 2), 53 – 65 (1999). 304. D. W. Nicholson, Caspase structure, proteolytic substrates, and function during apoptotic cell death. Cell Death Differ. 6, 1028 – 1042 (1999). 305. N. Okahashi, M. Koide, E. Jimi, T. Suda, and T. Nishihara, Caspases (interleukin-1-beta-converting enzyme family proteases) are involved in the regulation of the survival of osteoclasts. Bone 23, 33 – 41 (1998). 306. N. Namba, X. Feng, J. C. Chappel, S. F. Dowdy, S. L. Teitelbaum, and F. P. Ross, A dominant negative form of caspase-7, the major downstream caspase in osteoclasts, inhibits endogenous apoptosis. J. Bone Miner. Res. 14, S148 (1999). 307. D. G. Walker, Congenital osteopetrosis in mice cured by parabiotic union with normal siblings. Endocrinology 91, 916 – 920 (1972). 308. D. G. Walker, Osteopetrosis cured by temporary parabiosis. Science 180, 275 (1973). 309. D. G. Walker, Spleen cells transmit osteopetrosis in mice. Science 190, 785 – 787 (1975). 310. M. M. Tondravi, S. McKercher, K. Anderson, J. M. Erdmann, M. Quiroz, R. Maki, and S. L. Teitelbaum, Targeted deletion of PU.1 induces osteopetrosis rescuable by bone marrow transplantation. Nature 386, 81 – 84 (1997). 311. R. Muller, Role of Fos proteins in the induction of transformation and DNA synthesis. In “The Fos and Jun Families of Transcription Factors” (P. Angel and P. Herrlich, eds.), pp. 189 – 197. CRC Press, Boca Raton, FL, 1994. 312. V. Iotsova, J. Caamano, J. Loy, Y. Young, A. Lewin, and R. Bravo, Osteopetrosis in mice lacking NFkB1 and NFkB2. Nature Med. 3, 1285 (1997). 313. C. Lowe, T. Yoneda, B. F. Boyce, H. Chen, G. R. Mundy, and P. Soriano, Osteopetrosis in src-deficient mice is due to an autonomous defect of osteoclasts. Proc. Natl. Acad. Sci. USA 90, 4485 – 4489 (1993). 314. T. Yoneda, M. Niewolna, C. Lowe, E. Izbicka, and G. R. Mundy, Hormonal regulation of pp60c-src expression during osteoclast formation in vitro. Mol. Endocrinol. 7, 1313 – 1318 (1993). 315. C. Grandori and H. Hanafusa, pp60c-src is complexed with a cellular protein in subcellular compartments involved in exocytosis. J. Cell Biol. 107, 2125 – 2135 (1988). 316. S. J. Parsons and C. E. Creutz, pp60c-src activity detected in the chromaffin granule membrane. Biochem. Biophys. Res. Commun. 34, 736 – 742 (1986). 317. K. M. Oddie, J. S. Litz, J. C. Balserak, D. M. Payne, C. E. Creutz, and S. J. Parsons, Modulation of pp60c-src tyrosine kinase activity during secretion in stimulated bovine adrenal chromaffin cells. J. Neurosci, Res. 24, 38 – 48 (1989).
ROSS AND TEITELBAUM 318. P. Schwartzberg, L. Xing, C. A. Lowell, E. Lee, L. Garrett, S. Reddy, G. D. Roodman, B. Boyce, and H. E. Varmus, Complementation of osteopetrosis in src-/- mice does not require src kinase activity. J. Bone Miner. Res. 11, S135 (1996). 319. W. S. Sly, D. Hewett-Emmett, M. P. Whyte, Y.-S. Yu, and R. E. Tashian, Carbonic anhydrase II deficiency identified as the primary defect in the autosomal recessive syndrome of osteopetrosis with renal tubular acidosis and cerebral calcification. Proc. Natl. Acad. Sci. USA 80, 2752 – 2756 (1983). 320. T. Yamamoto, N. Kurihara, K. Yamaoka, K. Ozono, M. Okada, K. Yamamoto, S. Matsumoto, T. Michigami, J. Ono, and S. Okada, Bone marrow-derived osteoclast-like cells from a patient with craniometaphyseal dysplasia lack expression of osteoclast-reactive vacuolar proton pump. J. Clin. Invest. 91, 362 – 367 (1993). 321. S. C. Marks, Jr. and P. W. Lane, Osteopetrosis, a new recessive skeletal mutation on chromosome 12 of the mouse. J. Hered. 67, 11 – 18 (1976). 322. W. W. Wiktor-Jedrzejczak, A. Ahmed, C. Szczylik, and R. R. Skelly, Hematological characterization of congenital osteopetrosis in op/op mouse: Possible mechanism for abnormal macrophage differentiation. J. Exp. Med. 156, 1516 – 1527 (1982). 323. N. Takahashi, N. Udagawa, T. Akatsu, H. Tanaka, Y. Isogai, and T. Suda, Deficiency of osteoclasts in osteopetrotic mice is due to a defect in the local microenvironment provided by osteoblastic cells. Endocrinology 128, 1792 – 1796 (1991). 324. S. C. Marks, Jr., M. F. Seifert, and J. L. McGuire, Congenitally osteopetrotic (op/op) mice are not cured by transplants of spleen or bone marrow cells from normal littermates. Metabol. Bone Dis. Rel. Res. 5, 183 – 186 (1984). 325. R. Felix, M. G. Cecchini, W. Hofstetter, P. R. Elford, A. Stutzer, and H. Fleisch, Impairment of macrophage colony-stimulating factor production and lack of resident bone marrow macrophages in the osteopetrotic op/op mouse. J. Bone Miner. Res. 5, 781 – 789 (1990). 326. H. Yoshida, S.-I. Hayashi, T. Kunisada, M. Ogawa, S. Nishikawa, H. Okamura, T. Sudo, L. D. Shultz, and S.-I. Nishikawa, The murine mutation osteopetrosis is in the coding region of the macrophage colony stimulating factor gene. Nature 345, 442 – 443 (1990). 327. R. Felix, M. G. Cecchini, and H. Fleisch, Macrophage colony stimulating factor restores in vivo bone resorption in the op/op mouse. Endocrinology 127, 2592 – 2594 (1990). 328. H. Kodama, A. Yamasaki, M. Nose, S. Niida, Y. Ohgama, M. Abe, M. Kumegawa, and T. Suda, Congenital osteoclast deficiency in osteopetrotic (op/op) mice is cured by injections of macrophage colony-stimulating factor. J. Exp. Med. 173, 269 – 272 (1991). 329. S. J. Begg, J. M. Radley, J. W. Pollard, O. T. Chisholm, E. R. Stanley, and I. Bertoncello, Delayed hematopoietic development in osteopetrotic (op/op) mice. J. Exp. Med. 177, 237 – 242 (1993). 330. Y. Y. Myint, K. Miyakawa, M. Naito, L. D. Shultz, Y. Oike, K. Yamamura, and K. Takahashi, Granulocyte/macrophage colonystimulating factor and interleukin-3 correct osteopetrosis in mice with osteopetrosis mutation. Am. J. Pathol. 154, 553 – 566 (1999). 331. E. Stanley, G. J. Lieschke, D. Grail, D. Metcalf, G. Hodgson, J. A. M. Gall, D. W. Maher, J. Cebon, V. Sinickas, and A. R. Dunn, Granulocyte/macrophage colony-stimulating factor-deficient mice show no major perturbation of hematopoiesis but develop a characteristic pulmonary pathology. Proc. Natl. Acad. Sci. USA 91, 5592 – 5596 (1994). 332. P. F. Coccia, W. Krivit, J. Cervenka, C. Clawson, J. H. Kersey, T. H. Kim, M. E. Nesbit, N. K. Ramsay, P. I. Warkentin, S. L. Teitelbaum, A. J. Kahn, and D. M. Brown, Successful bone-marrow transplantation for infantile malignant osteopetrosis. N. Engl. J. Med. 302, 701 – 708 (1980). 333. P. J. Meunier, J. M. Coindre, C. M. Edouard, and M. E. Arlot, Bone histomorphometry in Paget’s disease. Arthritis Rheum. 23, 1095 – 1103 (1980).
CHAPTER 3 Osteoclast Biology 334. A. Boyde, Scanning electron microscope studies of bone. In “The Biochemistry and Physiology of Bone” (G. H. Bourne, ed.), Vol. 2, p. 259. Academic Press, New York, 1972. 335. B. Maldague and J. Malghem, Dynamic radiologic patterns of Paget’s disease of bone. Clin. Orthop. Rel. Res. 217, 126 – 151 (1987). 336. E. S. Siris, T. P. Clemens, D. McMahon, A. Gordon, T. P. Jacobs, and R. E. Canfield, Parathyroid function in Paget’s disease of bone. J. Bone Miner. Res. 4, 75 – 79 (1989). 337. R. G. G. Russell, R. Smith, C. Preston, R. J. Walton, and C. G. Woods, Diphosphonates in Paget’s disease. Lancet 1, 894 (1974). 338. M. R. A. Khairi, P. Meunier, C. Edouard, P. Courpron, J. Bernard, G. P. DeRosa, and C. C. Johnston, Quantitative bone histology in Paget’s disease of bone: Influence of sodium etidronate therapy. Calcif. Tissue Res. 22, 355 (1977). 339. A. Rebel, K. Malkani, M. Basle, and C. H. Bregeon, Osteoclast ultrastructure in Paget’s disease. Calcif. Tissue Res. 20, 187 (1976). 340. B. G. Mills, and F. R. Singer, Nuclear inclusions in Paget’s disease of bone. Science 194, 201 – 202 (1976). 341. G. Gherardi, V. L. O. Cascio, and E. Bonucci, Fine structure of nuclei and cytoplasm of osteoclasts in Paget’s disease of bone. Histopathology 4, 63 – 74 (1980). 342. L. Harvey, T. Gray, M. N. C. Beneton, D. L. Douglas, J. A. Kanis, and R. G. G. Russell, Ultrastructural features of the osteoclasts from Paget’s disease of bone in relation to a viral aetiology. J. Clin. Pathol. 35, 771 – 779 (1982). 343. S. V. Reddy, C. Menaa, N. Kurihara, G. Landucci, B. A. Koop, J. J. Windle, D. N. Forthal, and D. G. Roodman, Toward an in vivo model of Paget’s disease. J. Bone Miner. Res. 14, S143 (1999). 344. A. Kukita, C. Chenu, L. M. McManus, G. R. Mundy, and G. D. Roodman, Atypical multinucleated cells form in long term marrow cultures from patients with Paget’s disease. J. Clin. Invest. 85, 1280 – 1286 (1990). 345. J. A. Hoyland, A. J. Freemont, and P. T. Sharpe, Interleukin-6, IL-6 receptor, and IL-6 nuclear factor gene expression in Paget’s disease. J. Bone Miner. Res. 9, 75 – 80 (1994). 346. C. Menaa, S. V. Reddy, J. Bruder, D. Anderson, J. Cornish, T. Cundy, N. Kurihara, and G. D. Roodman, Enhanced RANK ligand expression and responsivity in Paget’s disease. J. Bone Miner. Res. 14, S143 (1999). 347. M. Tubiana-Hulin, Incidence, prevalence, and distribution of bone metastases. Bone 12, S9 – S10 (1991). 348. S. D. Averbuch, New bisphosphonates in the treatment of bone metastases. Cancer 72, 3443 – 3452 (1993). 349. A. Valentin-Opran, G. Elison, S. Saez, and G. R. Mundy, Estrogens and antiestrogens stimulate release of bone resorbing activity by cultured human breast cancer cells. J. Clin. Invest. 75, 726 – 731 (1985). 350. W. J. Kort, A. M. Bijma, J. J. van Dam, A. C. van der Ham, J. M. Hekking, H. F. van der Ingh, W. S. Meijer, M. G. Wilgenburg, and F. J. Zijlstra, Eicosanoids in breast cancer patients before and after mastectomy. Prostagl. Leuko. Ess. Fatty Acids 45, 319 – 327 (1992). 351. S. M. Scholl, P. Crocker, R. Tang, P. Pouillart, and J. W. Pollard, Is colony-stimulating factor-1 a key media of breast cancer invasion and metastasis? Mol. Carcinogen. 7, 207 – 211 (1993).
105 352. T. A. Guise, and G. R. Mundy, Cancer and bone. Endocr. Rev. 19, 18 – 54 (1998). 353. J. J. Yin, K. Selander, J. M. Chirgwin, M. Dallas, B. G. Grubbs, R. Wieser, J. Massague, G. R. Mundy, and T. A. Guise, TGF-beta signaling blockade inhibits PTHrP secretion by breast cancer cells and bone metastases development. J. Clin. Invest. 103, 197 – 206 (1999). 354. I. J. Diel, E. F. Solomayer, S. D. Costa, C. Gollan, R. Goerner, D. Wallwiener, M. Kaufmann, and G. Bastert, Reduction in new metastases in breast cancer with adjuvant clodronate treatment. N. Engl. J. Med. 339, 357 – 363 (1998). 355. A. A. Budayr, R. A. Nissenson, R. F. Klein, K. K. Pun, O. H. Clark, D. Diep, C. D. Arnaud, and G. J. Strewler, Increased serum levels of a parathyroid hormone-like protein in malignancy-associated hypercalcemia. Ann. Intern. Med 111, 807 – 812 (1989). 356. N. Horiuchi, M. P. Caulfield, J. E. Fisher, M. E. Goldman, R. L. McKee, J. E. Reagan, J. J. Levy, R. F. Nutt, S. B. Rodan, T. L. Schofield, T. L. Clemens, and M. Rosenblatt, Similarity of synthetic peptide from human tumor to parathyroid hormone in vivo and in vitro. Science 238, 1566 – 1568 (1987). 357. J. F. Rossi, R. Bataille, D. Chappard, C. Alexandre, and C. Janbon, B cell malignancies presenting with unusual bone involvement and mimicking multiple myeloma: Study of nine cases. Am. J. Med. 83, 10 – 16 (1987). 358. R. A. Kyle, Maintenance therapy and supportive care for patients with multiple myeloma. Semin. Oncol. 26, 35 – 42 (1999). 359. G. R. Mundy, L. G. Raisz, R. A. Cooper, G. P. Schechter, and S. E. Salmon, Evidence for the secretion of an osteoclast stimulating factor in myeloma. N. Engl. J. Med. 291, 1041 – 1046 (1974). 360. A. Valentin-Opran, S. A. Charhon, P. J. Meunier, C. M. Edouard, and M. E. Arlot, Quantitative histology of myeloma-induced bone changes. Br. J. Haematol. 52, 601 – 610 (1982). 361. X. G. Zhang, R. Bataille, M. Jourdan, S. Saeland, J. Banchereau, P. Mannoni, and B. Klein, Granulocyte-macrophage colony-stimulating factor synergizes with interleukin-6 in supporting the proliferation of human myeloma cells. Blood 76, 2599 – 2605 (1990). 362. M. Kawano, I. Yamamoto, K. Iwato, H. Tanaka, H. Asaoku, O. Tanabe, H. Ishikawa, M. Nobuyoshi, Y. Ohmoto, and Y. Hirai, Interleukin-1 beta rather than lymphotoxin as the major bone resorbing activity in human multiple myeloma. Blood 73, 1646 – 1649 (1989). 363. M. Kawano, T. Hirano, T. Matsuda, T. Taga, Y. Horii, K. Iwato, H. Asaoku, B. Tang, O. Tanabe, and H. Tanaka, Autocrine generation and requirement of BSF-2/IL-6 for human multiple myeloma. Nature 332, 83 – 85 (1988). 364. X. G. Zhang, B. Klein, and R. Bataille, Interleukin-6 is a potent myeloma-cell growth factor in patients with aggressive multiple myeloma. Blood 74, 11 – 13 (1989). 365. Y. Mori, T. Michigami, M. Dallas, M. Niewolna, B. Story, R. Lobb, G. R. Mundy, and T. Yoneda, Anti-4 integrin antibody suppresses the bone disease of myeloma and disrupts myeloma-marrow stromal cell interactions. J. Bone Miner. Res. 14, S173 (1999). 366. K. D. Merkel, J. M. Erdmann, K. P. McHugh, Y. Abu-Amer, F. P. Ross, and S. L. Teitelbaum, Tumor necrosis factor- mediates orthopedic implant osteolysis. Am. J. Pathol. 154, 203 – 210 (1999).
CHAPTER 4
The Biochemistry of Bone JAYASHREE A. GOKHALE AND ADELE L. BOSKEY Hospital for Special Surgery, Weill Medical College of Cornell University, New York, New York 10021
PAMELA GEHRON ROBEY Bone Research Branch, National Institute of Dental Research, National Institutes of Health, Bethesda, Maryland 20892
I. II. III. IV. V.
VI. Other Proteins VII. Requirements for Matrix Mineralization VIII. Pathways of Matrix Mineralization References
Introduction Collagen Glycosaminoglycan-Containing Proteins Glycoproteins Gla-Containing Proteins
in the cells that destroy bone (osteoclasts) and, in addition, the osteoblastic precursors that will replace the calcified cartilage with bona fide bone. The initial bone formed, woven bone, is a rather unorganized conglomeration of collagenous and noncollagenous proteins that induce the precipitation of mineral. Through modeling by osteoclasts, this primordial bone is removed and replaced by the formation of lamellar bone, a more highly organized structure with alternating layers of mineralized extracellular matrix, whose plywood-like structure provides bone with its mechanical strength. This structure is further reinforced by the development of the Haversian canal system centered around a blood vessel and a nerve, providing nutrients and signals to the cells entombed in bone (osteocytes), while maintaining communication through osteocytic cell processes in canaliculae. Initially it was hypothesized that mineralized matrices were composed of a unique set of matrix proteins, the combination of which would initiate the precipitation of hydroxapatite. However, with the development of techniques for isolation of the components without degradation [2,3] and cell culture systems that faithfully recapitulate the osteoblastic phenotype [4 – 9], it became apparent that most,
I. INTRODUCTION A. Bone: The Tissue The skeleton is essentially responsible for providing not only structural support and protection to the body’s organs, but also for serving as a reservoir for calcium, magnesium, and phosphate, ions that are of critical importance in physiology. The fabric of bone is a unique composite of living cells embedded in a remarkable threedimensional structure of extracellular matrix that is stabilized by a mineral, a carbonate-rich analogue of the geologic mineral hydroxyapatite. During development, mesenchymal cells form the skeleton via two basic pathways [reviewed in 1]. Endochondral bone is formed by an initial condensation of mesenchymal cells that induces the expression of the chondrocytic phenotype. The cartilaginous structure that is formed through the sequential expression of a number of genes that regulate morphogenesis serves as a temporary model. As part of that developmental sequence, the cartilage becomes calcified. The provisional calcified cartilagenous precursor is subsequently replaced by bone. Invasion by blood vessels brings
OSTEOPOROSIS, SECOND EDITION VOLUME 1
107
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
108
GOKHALE, BOSKEY, AND ROBEY
if not all, of the proteins synthesized by bone-forming cells are also synthesized by nonskeletal cells. There are few, if any, truly bone-specific proteins. However, it is now clear that the composition of bone is quite different: 70 – 90% of bone is composed of mineral with only 10 – 30% represented by protein. The ratio of collagenous to noncollagenous proteins differs greatly from that of other tissues, with collagenous protein composing 90% of the organic matrix (compared to 10 – 20% in other tissues) and noncollagenous proteins accounting for 10% (compared to as much as 80– 90% in soft tissues with the exception of tendon). In addition, all of the known collagenous and noncollagenous proteins in bone studied to date differ from those in other tissues in their chemical nature. These diverse forms result from alternative splicing of mRNA and different posttranslational modifications such as glycosylation, phosphorylation, and sulfation. These chemical differences most likely influence the function of these proteins, and the appropriate mixture provides the extracellular matrix with the ability to calcify. While the mineralized matrix is viewed by many as mere cement, it is actually a fairly dynamic aggregate structure. Although the most abundant proteins in bone matrix play structural roles in providing the scaffolding and binding sites for the regulation of mineral deposition and turnover, these proteins also function in modulating cellular activity. Cell – matrix interactions have become increasingly recognized as important determinants during all stages of development and tissue homeostasis. In addition, extracellular matrix proteins are the secretory products of cells in the osteoblastic lineage and, as such, they represent biochemical markers of the formation process (either as complete forms or precursor molecules) or the resorption process (in their degraded form). Therefore, it is necessary to have an indepth understanding of the biosynthetic process by which these proteins are formed. In addition, in order to decipher intelligently the clues provided by their assay ofin body fluids for determination of the status of the skeleton, it is important to understand their function in bone homeostasis.
B. Bone Matrix Formation: The Role of Maturational Stage It is now well accepted that bone formation is accomplished by cells in the osteoblastic lineage that pass through a series of maturational stages [10 – 13; see also Chapter 2]. The lineage is composed of cells that start off as uncommitted precursors. These precursors may be highly proliferative during development, but most likely at maturity, these stem cells are quiescent, thus preserving a reservoir of cells that have gone through a limited number of mitoses. Upon command, by signals that have yet to be fully identified, they become committed to the osteoblastic lineage and are
recognizable as fibroblast-like, proliferative osteoprogenitors. At some point, their rate of proliferation slows, and they are more aptly termed preosteoblasts, mainly by virtue of their location immediately adjacent to the real workhorse of the lineage, the osteoblast, on the opposite side of where mineralization will occur or is occurring. The term osteoblast is best thought of as a histological definition that describes a particular cell that has a large nucleus with prominent nucleoli indicative of a high rate of genetic expression. Additionally, it has a greatly expanded rough endoplasmic reticulum that is somewhat polarized such that the cell secretes enormous amounts of matrix toward the mineralization front, creating a layer of unmineralized osteoid. For reasons that are not yet known, a limited number of cells disengage themselves from the osteoblastic layer and are left behind as apposition proceeds. As these cells become buried in matrix that becomes mineralized (through a somewhat nebulous stage termed an osteoid osteocyte or an osteoblastic osteocyte), the cell maintains its communications with cells above it in the osteoblastic layer. This occurs via retention of cell contacts as cell processes become surrounded by mineralized matrix. Consequently, the most differentiated member of the lineage, the osteocyte, is still in communication with the osteoblastic layer above it. It is very likely that these embedded osteocytes serve a mechanoreceptor function. Through the cell processes in the canaliculae, they constantly monitor the surrounding environment and signal to the osteoblastic layer or lining cell layer when and where resorption is needed to remove and refurbish fatigued matrix or to respond to alterations in mechanical load. The expression of bone matrix proteins is somewhat stage specific (Fig. 1). However, currently it is still not perfectly clear whether one cell at one particular stage of maturation (the osteoblast) is capable of making all the components of the mineralized matrix or whether the process requires cells at different stages of maturation simultaneously to be present. Considering the amount of coupling that occurs between cells at different stages of maturation through gap and tight junctions, a great deal of cooperativity in the bone-forming process is likely. In vitro model systems have not really sorted out this issue, as none of the cell culture systems currently available go through the different stages of maturation in a synchronous fashion. Uncommitted progenitors, preosteoblasts, osteoblasts, and sometimes even osteocytes can be identified in populations of cells that are in the process of forming bone nodules in vitro (the current state of the art in vitro) [14]. It has been demonstrated that proliferation of a single cell gives rise to the whole spectrum of maturational stages, and nature must have a reason for reestablishing this continuum. Although identification of the various maturational stages and the characterization of the biological properties of the cells within each stage have advanced our under-
109
CHAPTER 4 The Biochemistry of Bone
FIGURE 1
Maturational stage and bone matrix gene expression. Osteoblastic cells pass through a series of maturational stages, each of which can be partially characterized by the bone matrix proteins that they produce. In addition, osteoclasts also secrete proteins that become incorporated into mineralized matrix.
standing of bone formation greatly, the mechanism of how and when various factors influence bone matrix protein expression is not clear. The literature is cluttered with numerous reports describing the effects of hormones, growth factors, cytokines, chemokines, and so on osteoblastic metabolism, and the results of these studies are extremely variable. This variability may reflect the fact that the effects that these factors exert most likely depend on animal species, donor age and sex, stage of maturation of the cells, state of transformation, and certain cell culture parameters, such as length of time in culture, density, and nutrient conditions. This chapter focuses on the structural aspects of bone matrix proteins and their genes and only highlights what factors are known to influence gene expression. The topic of what factors influence expression at a particular stage of maturation is addressed in more detail in Chapter 2. Only a few years ago, relatively few human bone matrix genes had been isolated and characterized. The current list is quite lengthy, and most of the information provided in this chapter pertains to human genes and proteins unless information is available only from another animal species.
C. Mineral The mechanical strength of bone is attributable to the presence of mineral, which converts the pliable organic
matrix into a more rigid structure [15,16]. In the composite structure of cells, protein and mineral, the bone mineral (apatite) crystals, approximately 300 Å in their longest dimension, are aligned along the collagen fibril axis [17 – 19]. A variety of structural analyses, including X-ray and electron diffraction [20 – 22], infrared spectroscopy [23], high-voltage electron microscopy [24], nuclear magnetic resonance (NMR), and X-ray absorption fine structure analysis (EXAFS) [25 – 27] have shown that mineral crystals are analogous to the naturally occurring geologic mineral, hydroxyapatite (Ca10(PO4)6(OH)2) (Fig. 2). In bone, mineral includes numerous ions not found in pure hydroxyapatite. For example, HPO42, CO32, Mg2, Na, F, and citrate are adsorbed on the crystal surfaces and/or substituted in the lattice for the constituent Ca2, PO43, and OH ions [28 – 34]. This poorly crystalline apatite, because of its small size and large number of latticesubstituted and surface-adsorbed ion impurities, goes into solution more readily than the larger, more perfect crystals of geologic hydroxyapatite. For example, some ions (HPO42, CO32, Na2, and F) can be in the lattice and on the surface, whereas others (Mg2 and citrate) prefer surface locations. This altered solubility allows bone mineral to play an important role in Ca2, Mg2, and PO43 ion homeostasis [35]. Despite claims of the presence of other mineral phases in bone, e.g., brushite [36], octacalcium phosphate [37],
110
GOKHALE, BOSKEY, AND ROBEY
FIGURE 2 Crystal lattice structure. A portion of the apatite structure is depicated as it would be viewed along the length (c axis) of the hydroxyapatite crystal, showing the hexagonal arrangement of the Ca2 and PO43ions about the OH position.
amorphous calcium phosphate [38], and whitlockite [39], current evidence supports the viewpoint that bone mineral is predominantly apatitic, with numerous, perhaps unique, impurities [40]. This viewpoint is maintained for the rest of the chapter.
gene activation and regulation, (ii) RNA transcription and processing, (iii) translation into protein, posttranslational modification, and secretion from the cell, and finally (iv) deposition in the extracellular matrix. The following description pertains specifically to the individual chains of type I collagen. However, virtually all of these processes apply to the other proteins found in bone.
II. COLLAGEN In the vertebrate body, the major structural protein is collagenous in nature. Collagen is defined as a trimeric molecule composed of chain subunits that are wound together to form a triple helix [41]. A significant feature of component chains is that their primary sequence is composed almost entirely of a repeating triplet sequence, GlyX-Y, where X is often proline and Y is often hydroxyproline [42]. Collagenous proteins are either homotrimeric, composed of three identical chains, or heterotrimeric, with two or three different chains. Currently there are over 23 different chains that associate to yield 13 different types of collagens, with 6 more potential types that to date have only been identified at the cDNA level [43 – 45]. The various forms of collagen are as diverse as the tissues in which they are found. These variations on a theme convey distinct features such that the collagen type is uniquely suited to carry out particular functions in a given tissue. In each of these tissues, collagen most likely serves a mechanical function. For example, in mineralized tissues, collagen fibrils provide tensile strength [15,46]. Given the complex nature of collagen expression, it serves as a useful example for describing the pathway of (i)
A. Gene Structure The structure of mammalian genes is as varied as the proteins for which they code. An example of a prototype gene structure is shown in Fig. 3. In the 5 region of the gene, there is usually a sequence that is not transcribed from DNA into RNA. As described later, this untranslated region (UTR), the promoter, is the primary site for the regulation of gene activity. Structures can also form due to the presence of palindromic type sequences (mirror image and complimentary) that allow the DNA to take on conformations other than the typical Watson – Crick double helix. These unusual conformations (such as z DNA) may serve as recognition sites for certain factors and enzymes that regulate gene activity [47 – 50]. Following this 5 UTR is the portion of the gene that will be transcribed into an RNA sequence through the action of RNA polymerases. Genes contain sequences that are found in mature RNA and dictate the amino acid sequence of protein (exons), interspersed with sequences that are removed and do not end up in mature mRNA or in the translated protein (introns). The exon – intron structure of a gene
111
CHAPTER 4 The Biochemistry of Bone
FIGURE 3 Basic structure of prototype gene. In addition to exons (E), introns (I), and the transcription start site (toward the 3 end), the gene contains a promoter toward the 5 end, upstream of the transcription start site. It includes defined sequences or cis-acting elements such as CCAAT and TATA boxes. Trans-acting factors (proteins) can bind to cis elements to regulate transcription.
can be simple, with only a single exon, or highly complex, with multiple exons as exemplified by the genes for type I collagen. The gene for the constituent chains of type I collagen, COL1A1, located on chromosome 17q21.3-q22, is 18 kb in length and contains 51 exons [51,52], whereas COL1A2, on chromosome 7q21.3-q22, is 35 kb and has 52 exons [52 – 54]. It is often the case that exons code for functional regions of the resulting protein, and may thereby represent an ancestral gene. In many of the genes for collagen chains, the exons are often 54 bases, or multiples thereof, that code for six [GLY-X-Y]n triplets. Following the open reading frame that designates the protein sequence is another stretch of DNA sequence that codes for mRNA that again does not end up in protein. This 3-untranslated region (3 UTR) contains sequences that may also influence gene activity. Signals for polyadenylation, another posttranscriptional modification, are also located within this region [55,56].
B. Gene Activation and Transcription The process of gene activation is a complex series of events that are mediated by cis and trans-acting factors (see Fig. 3). Cis-acting elements are defined as sequences present in DNA that are required for the binding of factors resident within the nuclear environment that influence (either positively or negatively) gene transcription. They include the binding sites for enzymes that are required for the synthesis of RNA complementary to the DNA template, such as polymerases. These types of cis-acting elements are exemplified by the sequences TATA and CAAT, which have been postulated to serve as the binding/orientation sites for
polymerases. In addition, GC-rich regions may also regulate polymerase activity, perhaps through the formation of three-dimensional structures. Cis-acting elements are found most commonly upstream (5) from the RNA transcription start site (the promoter region), where polymerase initiates the synthesis of mRNA. However, cis elements can also be located in sites quite distant from the transcription start site and may also reside in intronic sequences, or in the 3 regions of the gene. Cis-acting elements are also the binding sites for transacting factors, which are defined as proteins that either singly or in combination have the ability to up- or downregulate gene activity. Trans-acting elements are extremely diverse, and many are the gene products of protooncogenes or receptors bound to their ligands. These DNA-binding proteins have a number of structural motifs that allow them to bind to the double helix [reviewed in 57], including helix – loop – helix [58,59], and leucine zipper proteins, which can also have important functions in the cytoplasm [60]. Other motifs include HMG-1 box-binding proteins [61] and zinc-binding proteins (fingers, twists, and clusters) [62]. By binding to specific sequences, these complexes can either enhance the activity of the polymerizing enzymes, leading to high levels of a particular mRNA species, or suppress their activity, such that there are very few or virtually no copies produced. The promoters for COL1A1 and COL1A2 have been characterized in detail and contain similar but not identical promoter elements [63 – 67]. At 29 bp from the transcription start site, the COL1A1 promoter contains a TATA box, whereas it is absent in the COL1A2 promoter. Further upstream, both contain a CCAAT sequence (100 bp in COL1A1 and 82 bp in COL1A2), as well as a long stretch of Cs and Ts, which confer S1 nuclease and DNAse
112
GOKHALE, BOSKEY, AND ROBEY
hypersensitivity, implying a relatively open structure. The COL1A1 sequence then becomes A-G rich, which could indicate another potential change in tertiary structure due to base pairing between these two regions. An SP1 element (the binding site of a constitutive transcription factor) is also present. Studies utilizing rather extended regions of this COL1A1 promoter have demonstrated that it is active. However, maximal activity may require sequences present in the first intron, which also contains SP1 and DNAse hypersensitive regions [68]. This concept remains controversial because both stimulatory [69,70] and inhibitory [70] elements have been described in this region. The 2(I) contains a positive enhancer in the first intron and a CAATbinding protein (CTF/NF1)-binding site [71]. Other elements include a vitamin D response element (VDRE) in COL1A1 [72] and a CAAT-like region that binds to nuclear factor 1 (NF1) in the COL1A2 promoter [73]. It is of interest that the amount of mRNA for COL1A1 is twice the amount of COL1A2, a ratio that is reflected in the final triple helical molecule. Consequently, there must be factors (e.g., ascorbic acid) that regulate the production of these two gene products such that there are never excessive amounts of COL1A2 mRNA [74 – 77].
C. Gene Regulation While there is only one copy of the genes that code for the COL1A1 and COL1A2 in mammalian genomes, the regulation of type I collagen production in bone is somewhat different from that in soft connective tissues. It has been recently found that different parts of the COL1A1 promoter are required to maintain production by bone-forming cells in vitro [78 – 79]. In bone cell and organ cultures, collagen synthesis increased by heparin [80], organic phosphate [81], interleukin (IL)-4 [82], and gallium [83]. In contrast, collagen synthesis is decreased by prostaglandin E2 [84] 1,25-dihydroxyvitamin D3 [85], cortisol [86], parathyroid hormone (PTH) [87], epidermal growth factor (EGF) [88], basic fibroblast growth factor (bFGF) [89], IL10 [90], and lead [91]. Although the COL1A1 promoter contains a VDRE-like element, binding of this element by the VDR that has bound to its ligand inhibits expression. In addition, removing this element from the promoter does not totally abolish the inhibitory effect of 1,25-dihydroxyvitamin D3, indicating that other cis- and/or trans-acting factors are involved [85]. Depending on the concentration and the stage of the cell culture, dexamethasone, can either increase or decrease collagen synthesis [92,93].
D. RNA Processing Once transcription is initiated, a precursor form of RNA (often termed Hn or heteronuclear RNA) is transcribed
from the DNA template. This precursor form contains the intronic sequences that must be removed prior to translation to yield mRNA with only exonic, 5 UTR, and 3 UTR sequences. This process is accomplished through the action of a number of enzymes that associate to form what is termed a spliceosome [94,95]. Several types of splicing reactions can occur based on the sequences that bridge the exon and intron. Splicing occurs by bringing together the junction consensus sequence (GT ........ AT) at both the 5 and the 3 end of the intron and cleavage by enzymes via formation of a lariat-like structure [96]. Differential splicing can also occur via exon skipping, or exon splitting due to the presence of splice junctions that are buried within the exonic sequences. The factors that regulate differential splicing are not well understood but are believed to provide a particular cell type with the ability to generate different mRNA species. The mRNA is further modified by the addition of a methyl cap at the 5 end. The addition of a polyadenyl tail in the noncoding region at the 3 end completes the formation of the mature mRNA. The regulation of polyadenylation (choice between different polyadenylation sites and length of polyadenylation) is not well known but is implicated in regulating the half-life of the mRNA. At this point, the fully processed mRNA species are transported to the cytoplasm (Fig. 4). In the case of COL1A1, the sizes are 7.2 and 5.9 kb, and for COL1A2, the sizes are 6.5 and 5.5 kb.
E. Translation and Secretion Once in the cytoplasm, the mRNA species associate with ribosomes, possibly through the formation of stem loops that have been described in some mRNA species, including the mRNA for COL1A1 [97]. The initiation codon in most proteins is AUG (or CUG) for methionine. The initial translation of secreted proteins produces a signal peptide that allows binding of the ribosome and nascent polypeptide chain to the cytoplasmic surface of the endoplasmic reticulum (ER). As translation proceeds, the protein is extruded into the lumen of the ER and the signal peptide is cleaved. However, it should be noted that, like many other proteins, collagen chains are synthesized in a precursor form. Noncollagenous sequences are found at both amino (pN) and the carboxyl (pC) termini, resulting in a chain termed a pro chain [reviewed in 98]. During passage of the nascent polypeptides through the ER, posttranslational processing begins by the action of certain enzymes that are resident along the secretory pathway. Prolyl hydroxylases will form hydroxyprolyl residues immediately preceding glycyl residues in the triplet sequence. In most cases, proline is hydroxylated in the 4 position; however, in certain tissues, another enzyme will hydroxylate in the 3 position. Prolyl hydroxylation is required
113
CHAPTER 4 The Biochemistry of Bone
FIGURE 4 Gene activation and RNA transcription. Gene activity is regulated by the interaction of cis-acting (DNA sequences) and trans-acting (nuclear factors) elements that either initiate or repress transcription. Once a gene is activated, polymerases translate the antisense genomic sequence, including the exons (which ultimately code for the protein) and the introns (intervening sequences). The initial transcription product, HnRNA, is processed through the action of spliceosomes to remove the intervening sequences. RNA is further modified by methylation of the 5 end and by the addition of poly(A) to the 3 end prior to transport to the cytoplasm.
for the formation of a stable helical structure because the hydroxyl groups participate in intrahelix hydrogen bonding. Certain lysyl residues will also be hydroxylated by lysyl hydroxylase in the ER. Hydroxylysyl residues can be further modified through the action of sugar transferases that add first galactosyl residues, and sometimes an additional glucosyl residue to form galactosyl-hydroxylysine and glucosyl-galactosyl-hydroxylysine, respectively. The function of these sugar modifications is unknown. Once translation of the pro chains has been completed, the noncollagenous precursor sequences at the carboxy termini of two pro1 and one pro2 chains associate and bond through disulfide bridges. This aggregation is then followed by triple helix formation progressing from the C terminus moving toward the N terminus as the protein passes through the ER. Posttranslational modifications continue until the target residues are no longer accessible due to triple helix formation [reviewed in 98]. The formation of this unique conformation is dependent on the GLY-X-Y sequence throughout the helical domain of the molecule. This sequence allows the formation of rodlike triple helices [41], as glycine is the only amino acid small enough to fit within the center of the triple helix. The X and Y amino acids occur on the surface of the triple helix and are arranged in charged clusters, which facilitate the interaction of the individual molecules in the formation of
fibrils [99]. It is currently believed that there are specific domains within the fibrils that interact with cells, fibronectin, decorin, and other matrix molecules, but these have not been well identified. Individual chains of the collagen molecule coil about one another in an extended rigid helix. The structure is stabilized by hydrogen bonding between OH groups on hydroxyproline and intrachain water [100] and by aldehyde derived cross-links [101]. More details of the collagen structure have bene reviewed by Kuhn [102]. In addition, the crystal structure of a small triple helical peptide [103] has provided confirmation of earlier structure predictions. With the passage of the hydroxylated and glycosylated precursor to the Golgi apparatus, the procollagen molecule is further modified by the addition of N-linked oligosaccharides to certain asparaginyl residues in the carboxy-terminal precursor region. Initially, a high mannose structure is transferred to the acceptor residue, and the oligosaccharide is trimmed and rebuilt by the addition of N-acetylgalactosamine and N-acetylglucosamine to form a complex type oligosaccharide. Again, the function of these sugar modifications is unknown, although they are thought to be important for transport of the molecule from the cell [104]. In the Golgi apparatus, O-linked glycosylation reactions also occur; however, collagen chains are not modified in this fashion.
114
GOKHALE, BOSKEY, AND ROBEY
The process by which collagen molecules are moved from one cellular compartment to another is not completely known, but it has been speculated to be mediated by proteins termed chaperonins, some of which are heat shock proteins [105]. Due to the major role that collagen plays in all tissues, and the complicated nature of its biosynthesis, its processing and secretion may be tightly regulated by such proteins. It is likely that a derangement in the secretory apparatus that affects collagen secretion would result in disease. The triple helical and posttranslationally modified procollagen molecules within secretory granules are then transported to the cell surface and extruded from the cell (Fig. 5).
F. Deposition and Fibril Formation Although the exact process of fibril formation is not fully characterized, it has been speculated that it can occur either prior to or during the process of secretion by the formation of deep invaginations of the cell membrane, which provide a secluded environment [106]. It is clear that the nontriple helical precursor extensions must be removed in
FIGURE 5
an orderly fashion in order for normal fibers to form. Specific peptidases remove the N-terminal extension (yielding pC collagen) or the C-terminal extension (yielding pN collagen). When both extensions are removed, the fully mature triple helical collagen molecule still contains short nontriple helical telopeptides at both termini. Interestingly, it has been suggested that both pN and pC propeptides participate in feedback inhibition of collagen synthesis [107 – 111]. In addition, the pN peptide remains, at least in part, within bone matrix and was identified as the 24-kDa phosphoprotein [112,113]. The pC propeptide escapes into the circulation and has been used as a measure of collagen biosynthesis [114]. However, because pC can be contributed from the synthesis by soft tissues as well, it is not totally reflective of bone formation. The mature collagen molecules then form head to tail and lateral associations, and it is thought that the lateral associations are inhibited by the presence of the pN collagen. The addition of pN collagen or pC collagen molecules to the outermost layer of fibers may in fact dictate when fibril growth stops. Small fibers tend to be coated with pN collagen, whereas larger fibers are associated with pC collagen [115,116]. Fiber diameter growth may also be regulated by
Protein translation, modification, and secretion. Once in the cytoplasm, mRNA is translated into protein by ribosomes. In the case of bone matrix proteins (and the majority of secreted proteins), mRNA codes for a signal peptide that allows the ribosome to attach to the endoplasmic reticulum, allowing the nascent peptide to be extruded. In the lumen, hydroxylases modify certain prolyl and lysyl residues. Once the C-terminal portion is completed, the C propeptides of three chains associate and triple helix formation proceeds to the amino terminus. The pathway to secretion proceeds through the Golgi apparatus where the molecule is further modified by the addition of oligosaccharides to the C propeptide, and galactosyl and glucosyl-galactosyl residues to lysine and hydroxylysine residues by sugar transferases. Upon secretion, propeptides at N and C termini are cleaved, and the molecules are deposited in a four-dimensional staggered array and are further stabilized by covalent cross-links.
115
CHAPTER 4 The Biochemistry of Bone
the binding of other collagenous or noncollagenous proteins to the outer fibril surface [116 – 119]. This is an important point because it may be that the surface of type I collagen fibrils is not directly accessible for binding, and that collagen may be functioning in a purely structural fashion when intact within bone collagen fibers. Other activities may occur once fibril-associated proteins are removed or when the collagen fibril is degraded. The structure of fibrillar type I collagen has been described in detail [120]. Katz’s structural models, based on low angle X-ray scattering, confirm the structural pattern predicted by Hodge and Petrushka [121]. According to this model, individual collagen fibrils are aligned in a quarterstaggered array, with a 280-nm periodicity. As a result of the quarter stagger, there are gaps (holes) within the fibrillar structures, and it is in these gaps and in the overlapping regions adjacent to them that bone mineral first appears [18,122 – 125]. Specifically, Traub and co-workers [125] have shown that the first mineral crystals appear in a specific region of the collagen fibril known as the e band. However, it should be noted that the majority of these studies have been performed in mineralizing turkey tendon, which may not be identical to other patterns of matrix mineralization. Once fibers have been formed in the extracellular environment, they are further stabilized by the formation of inter- and intramolecular cross-links. This process occurs through the action of lysyl oxidase, which converts lysyl and hydroxylysyl residues in the telopeptide regions to allysine and hydroxyallysine, which are aldehydic in nature. Immature cross-links that are reducible by sodium borohydride are formed by the condensation of the aldehydes with other lysyl and hydroxylysyl residues within the helical region of a neighboring molecule. With time, these crosslinks become more insoluble by condensation with histidinyl residues or its aldehydic derivatives. Through this time-dependent process, as reviewed (126), up to four
TABLE 1 Collagen
chains can ultimately become cross-linked, leading to great stability and insolubility of the collagenous scaffolding (Fig. 5).
G. Collagen Types A broad range in the collagen pattern and molar mixture is displayed from one tissue to another and certain types are concentrated in specific tissues [reviewed in 43,44]. In comparison, bone matrix proper contains a rather limited array of collagen types (Table 1), which will be discussed in detail in this chapter. Based on their structural features, collagens can be roughly divided into two groups: fibrillar and nonfibrillar. Fibrillar collagens (types I, II, III, and V) are by far the most abundant forms and are found in the interstitial spaces of connective tissues throughout the body. Type I collagen, the predominant collagen of skin, tendon, and bone, is composed of two 1(I) and one 2(1) chains and forms the major scaffolding of virtually all the connective tissues (with the exception of cartilage). Cartilage contains predominantly type II collagen ([1(II)]3) with limited amounts of other collagen types as described later. Type III, composed of three identical 1(III) chains, and type V, composed of a combination of 1(V), 2(V), and 3(V) chains, are often codistributed with type I. Fetal tissue contains proportionally more type III collagen, which has been reported to coat the surfaces of type I fibrils or to form thin reticulin-like fibrils. Type V collagen is often pericellular and is also enriched in particular tissues such as smooth muscle and blood vessels. The nonfibrillar collagens (types IV, VI, VII, VIII, IX, X, and XI) are characterized by triple helical domains that are either shorter or longer than those of the fibrillar types and may contain stretches of noncollagenous sequences. Type IV collagen, with the composition [1(IV)22(IV)], is
Collagen Types Found in Bone Matrix Location/function
Molecular structure
Type I:[1(1)2 (1)] and [1(1)3]
Constitutes 90% of matrix in the bone matrix. Acts as scaffolding and binds to other proteins that initiate hydroxyapatite deposition
67-nm banded fibrils
Type III:[1(III)3]
Present only in trace amounts and can regulate collagen fiber thickness
67-nm banded, coats type I fibrils
Type :[1(V)22(V)] and [1(V)2(V)
Their absence can result in collagen fibrils of larger diameter
67-nm banded, coats type I fibrils in some tissues
Type X:[1(X)3]
Present in hypertrophic cartilage and can be involved in matrix organization via formation of the template for type I collagen
Probably fishnet-like lattice
116
GOKHALE, BOSKEY, AND ROBEY
found in basement membranes, including those that surround vascular endothelial cells that invade bone during osteogenesis. Type VI collagen is significantly shorter than other collagen types and is composed of three distinct chains. These molecules form rope-like microfibrillar structures. Anchoring fibrils are composed of type VII collagen, which is 1.5 times longer than type I. Another short chain collagen, type VIII, is found in Descemet’s membrane of the eye, is synthesized by endothelial cells in culture, and may be related to type X collagen found in calcified cartilage. The localization of type X to hypertrophic chondrocytes is highly specific, but it does not appear to have a major role in cartilage calcification. Type IX and type XI are homologous to type V and are minor constituents in cartilage. Type IX belongs to the socalled FACIT class (fibril associated collagen with interrupted triplex). Structures of the isolated FACIT collagenous proteins, as distinct from those of the fibrillar collagens, include nontriple helical domains, as predicted from the non-GLY-X-Y repeats. Type IX is composed of three different types of chains, 1(IX), 2(IX), and 3(IX), that form a short and a long triple helix joined by a flexible hinge region. A glycosaminoglycan chain is also attached to one of the chains at the amino terminus, making this collagen a proteoglycan as well. Type IX has been found covalently attached to and as a coating to type II collagen fibrils and covalently attached to it. Type XII is similar to type IX but has three projections extending from the triple helix. This type may also be associated with type I fibrils in tendon. Type XIV (as well as XII) is structurally related to the type IX collagen fibrils, which trim type II collagen in cartilage.
H. Bone Collagen(s) Although bone matrix proper has been reported to contain only type I collagen, other types are certainly present but not at the levels found in soft connective tissues. Several FACITs have been detected in bone [127,128], and there are occasional reports of low levels of type III and type V [129,130] molecules as well. The FACIT collagens found associated with type I collagen in other tissues are types XII and XIV. Given the potential role of these low abundance collagens in regulating fibril diameter, it is possible that the collagen fibrils in bone grow to much larger diameters than in soft tissues due to the reduced proportion of these diameter-regulating types. However, it may be that these other collagen forms may originate from the vasculature that infiltrates bone to a great extent and may persist in the bone during chemical extraction prior to demineralization. Based on sequence analysis [131], type XII collagen is predicted to be a Mr 340,000 protein with repeating
domains consisting of regions homologous to the type III motif of fibronectin and a von Willebrand factor domain [132]. In addition, there is a type IX collagen-like domain, and several RGD cell-binding domains [132]. The presence of collagen, cell binding, and matrix-binding domains implies that these FACIT collagens may have novel functions. Type XIV collagen’s sequence indicates that it has a similar series of fibronectin and von Willebrand factor domains [132]. Type XII collagen is a homotrimer with two triple helical domains and a large (Mr 190,000) N-terminal globular domain. The helical domains appear to form arms stretching out from the globular domain [129], resulting in a cruciform appearance [133] with one thin and three thick arms. Like type IX collagen, type XII collagen made by fibroblasts, but not by all cells, contains CS chains [133]. In bone, type XII collagen is seen in the periosteum [134] and is made by periosteal cells in culture [135]. Type XII-like collagens also trim the surface of type I fibrils [136], as does a type XIV variant [137]. The FACIT collagens seem to have a fundamental role in determining matrix structure, as demonstrated by animals lacking or containing mutated forms of the FACIT collagens [138]. These animals exhibit a spectrum of bone and cartilage disorders, presumably due to abnormal fibril formation. In bone, the proteoglycan decorin, as well as types XII and XIV collagens, may be important for regulating in vitro and perhaps in situ type I collagen fibrillogenesis. The observation that a type XII molecule trims the fibrils at specific sites, while proteoglycans are found at other sites along the collagen fibrils, supports this view. This will be discussed in more detail later. Type I collagen is found not only in bone, tendon, and dentin, but also in the sclera of the eyes, the lungs, the cornea, and the skin. These type I collagens are not identical but differ in the extent of all the posttranslational modifications [102]. The predominant form of glycosylation in bone is galactosyl-hydroxylylysine, whereas in soft tissue it is glucosyl-galactosyl hydroxylysine [139,140]. In bone, the cross-linking pattern is also different and originates from the hydroxyallysine pathway, resulting in the formation of a lysyl-pyridinoline cross-link, as opposed to soft connective tissue where cross-links originate predominantly from the allysine pathway leading to hydroyxlysylpyridinoline [141]. This modification is now the basis of a clinical assay that can measure specifically the degradation of type I collagen from bone [126]. The altered crosslinking pattern in bone has been speculated to be due to the deposition of mineral within collagen fibrils [101]. These altered cross-links may be important for the mechanical properties of the tissue. In all connective tissues, the collagens serve mechanical functions, providing elasticity and structure for the component tissues. Although type I collagen is widely distributed,
117
CHAPTER 4 The Biochemistry of Bone
the tissue most affected when type I collagen metabolism is abnormal is bone. In bone, extensive data indicate a link between type 1 collagen and bone strength [142]. Insight into these functions comes from detailed analysis of human and other animal tissues in which collagen synthesis is altered. Classic examples are the various forms of osteogenesis imperfecta (brittle bone disease) in which bone fragility has been associated with spontaneous mutations in the type I collagen genes [99,143 – 147] and the Mov-13 mouse in which a viral insertion within the first intron totally silences the 1(I) gene [148]. The remaining 2(I) chains are unable to form a stable triple helical molecule [although 1(I) can form homotrimers]. These mice die during gestation, indicating a critical role for type I collagen during development [149,150]. Transgenic mice in which other mutations are introduced [151,152] or in which the entire 2(I) chain is ablated [153] provide animal models for various types of osteogenesis imperfecta (OI) [154]. The bone changes in such animal models of OI show features similar to those seen in humans. Specifically, when glycine is substituted by aspartate or cysteine at positions close to the C terminus, the disease mimics the perinatal lethal phenotype [155]. Similar substitutions midmolecule generally result in a less severe phenotype [145,146,154] characterized by the presence of thinner than normal collagen fibrils and active osteoblasts containing numerous collagen fibrils. When the entire 2(I) chain is absent, animals that make an 1(I) trimer show even fewer bone abnormalities [153], similar to reported cases in humans [156]. However, there are many exceptions to this general rule of mutation location within the molecule as being predictive of severity of the phenotype. The mineral crystals in the bones of patients and transgenic animals with OI tend to be smaller than those in agematched control bones [157,158]. In the OI mouse (oim) that lacks the 2(I) chain [153], tendon [159] and bone 160 mineralization is aberrant. In the oim tendon, the crystals occasionally appear outside the collagen matrix, a feature never noted when collagen production is normal [161]. Similarly, in oim bones, the pattern of initial mineral deposition and crystal growth along the collagen differs from normal: the crystals appear outside the collagen matrix and with regions of collagen that are less mineralized than those in normal controls [158]. In addition, the presence of thinner fibrils in OI patients may be insufficient to provide nucleation and scaffolding sites, which can potentially translate into fragile bones [162]. It is not known at this time whether mineral seen away from the collagen fibrils was formed in the absence of a collagen backbone or whether it “broke away’’ and was later seen out in the matrix because the collagen structure was not sufficient to support it. Collagen per se does not cause mineral deposition; i.e., it is not a mineral nucleator, as it lacks the appropriate conformation that matches the ion surface of the deposited mineral surface [163]. Nonetheless, data from OI tissues clearly
demonstrate the importance of collagen for providing a scaffold to organize the mineral. As discussed later, noncollagenous matrix proteins appear to initiate and regulate mineral deposition [reviewed in 164,165]. Thus, an additional function of type I collagen is to provide a site for the retention of noncollagenous proteins, some of which appear to be covalently bound [166] whereas others are more loosely associated through specific collagen-binding domains. In patients with OI, decreases in some matrix proteins [167,168] may be due to the deficit of collagen to which they bind or to absolute decreases in protein production [169,170] associated with “protein suicide” resulting from the destruction of abnormal collagen within the cell [99]. Decreased type I collagen production as seen in these OI models demonstrates the template-like and mechanical functions of collagen. It is likely that animals null for 1(I) are not compatible with life because of the role type I collagen plays in the development of lung, blood vessels, and mesenchyme. However, overexpression of type I collagen also provides insight into collagen function. While finding an increase in type I collagen production in bone is relatively rare, it has been reported that MAV.2-0 (myeloblastosis-associated retrovirus) causes an osteoblastic hyperplasia and a relative increase in cortical bone thickness [171], providing another animal model for the characterization of collagen function.
III. GLYCOSAMINOGLYCANCONTAINING PROTEINS Proteoglycans are a class of macromolecules characterized by the covalent attachment of long chains of repeating disaccharides that are often sulfated, termed glycosaminoglycans (GAGs). The different types of GAGs, chondroitin sulfate (CS), dermatan sulfate (DS), keratan sulfate (KS), heparan sulfate (HS), and hyaluronan (HA, which is unsulfated), are named based on the sugar composition of the repeating disaccharides (Fig. 6). GAGs, with the exception of HA, are synthesized in the Golgi apparatus by the sequential addition of sugar residues to a sugar transporter molecule, dolichol phosphate. Certain sugar residues become sulfated through the action of sulfotransferases; however, the level of sulfation can be extremely variable even within one GAG chain. Ultimately, the growing GAG is transferred from the dolichol phosphate to an acceptor site on a protein core. These sites are most frequently specific seryl or threonyl residues, but in the case of KS, the acceptor site is an asparaginyl residue. There are many different families of core proteins but the factors that dictate which residues become gagosylated and by what type of GAG are not well understood. In contrast to the other GAG chains, HA has not been found covalently attached to a core
118
GOKHALE, BOSKEY, AND ROBEY
FIGURE 6 Disaccharide composition of glycosaminoglycans (GAGs). The GAG side chains that are covalently attached to proteoglycan core proteins are composed of repeating disaccharide units. The composition of the disaccharides, along with modifications by acetylation, results in the formation of chondroitin sulfate, which is epimerized to form dermatan sulfate, heparan sulfate, and keratan sulfate. Hyaluronan is the sole GAG that remains unsulfated and is not covalently linked to core proteins.
protein. HA does associate with certain cartilage proteoglycans by a noncovalent linkage. Proteoglycans can be glycosylated by the addition of N- and O-linked oligosaccharides. In addition to sulfation of the GAG chains, the oligosaccharides can also be phosphorylated and/or sulfated, generating a class of macromolecules with multiple posttranslational modifications.
A. Aggrecan During endochondral bone formation, cartilage first hypertrophies and forms a temporary mineralized tissue, calcified cartilage, which is then invaded by blood vessels. The invading vasculature brings with it (1) cells that destroy the calcified matrix (osteoclasts or a related cell type) and (2) osteoprogenitor cells. Because cartilage macromolecules can be in close proximity to forming bone and may actually be incorporated into the initial bony tissue, it is im-
portant to describe its extracellular matrix constituents. It is also not clear whether all of the large CS-proteoglycans isolated from bones are remnants of those in calcified cartilage or specific bone products. The presence of elevated amounts of CS-proteoglycans in the bones of osteopetrotic animals with defective osteoclasts was linked to the inability of these animals to resorb calcified cartilage [172]. The basic scaffolding upon which cartilage matrix is built is type II collagen. In addition, a number of proteoglycans, including (i) a large CS molecule, termed aggrecan reviewed in 173, have the ability to form aggregates with HA and (ii) two small proteoglycans, decorin and biglycan [reviewed in 174 – 177]. Other proteins, including COMP, CD-RAP, chondroadherin, and matrilin-1, are present, but at lower levels [178,179]. Intact aggrecan has a Mr 2.5 million kDa, with a core protein ranging in apparent Mr between 180,000 and 370,000 with just over 100 GAG chains (mostly CS, but with some KS) of about Mr 25,000 (Fig. 7). Based on enzymatic cleavage and sequence homology, five
119
CHAPTER 4 The Biochemistry of Bone
FIGURE 7
A representation of the chemical features of the large hyaluronic acid-binding proteoglycan, aggrecan. GAG, glycosaminoglycan; CS, chondroitin sulfate; KS, keratan sulfate; G1, G2, G3, globular domains (see text for description); EGF, epidermal growth factor; CRP, C-reactive protein.
domains have been defined in the core protein: three globular domains, two of which bind to HA (G1 and G2), an interglobular domain, a domain rich in serine-glycine repeats to which the CS and KS glycosaminoglycan chains are attached, and the third globular domain, G3, at the C terminus. Each of the globular domains are cysteine rich and stabilized by disulfide bonds [180]. The G1 domain in the N terminus is structurally homologous [181] to “link protein” [reviewed in 182], which stabilizes the interaction between the proteoglycan and HA in cartilage [183]. The adjacent G2 domain provides a flexible hinge, whereas the more linear central domain is the site of KS and CS covalent attachment. The C-terminal G3 domain is also predicted from sequence analysis to be globular and contains a set of epidermal growth factor (EGF)-like and complement regulatory protein (CRP)-like sequences at the carboxy terminus [184,185]. The individual GAG chains form extended flexible structures, whereas the serines in the central domains have -D-xylose attachments with restricted orientation [186]. Whereas NMR analysis of isolated CS chains has provided insight into their conformation in different solutions [187], their actual conformation when attached to the core protein is not known. The human aggrecan gene is located on chromosome 15q26 [188]. However, the complete genomic sequence has been reported only in the rat and is 63 kb in length. There are 18 exons with 30 kb of intronic sequence separating the first and the second exon. The intermediate exons roughly encode for each of the structural domains of the molecule with the exception of G1-B (which contains the hyaluronicbinding domain) and G2-B (which contains link protein-
like sequences), which are split between two exons, and G3, the lectin-binding domain, which is coded for by three exons. The splice junctions are primarily symmetrical phase 1 type (in frame) [189,190]. The rat gene promoter lacks a TATA box, and the major transcription start site is located in close proximity with a number of SP1 sites. In addition, there are four AP2 sites located 120 kb upstream in a GC-rich region, and two of the SP1 sites overlap. A GC-rich region is also found in the first exon, which also has four AP2 sites. A potential NF-b (nuclear factor) site is located between 123 and 103 bp, and another AP2 site is located between 81 and 37 bp [190]. The resulting mRNA species of 8.2 and 8.9 kb predict a 2124 amino acid residue protein, including a 19 residue signal peptide. A stretch of 1164 residues contain Ser-Gly repeats, the CS attachment site [189,190]. Structures of the core proteins, individual domains, and segments of these domains have been determined based on NMR and molecular modeling [186,191] and neutron and X-ray scattering of molecules in solution [192]. These studies reveal that the core protein has a fairly linear structure. The protein core of the aggrecan-like proteoglycans (CS/KS-containing) is fairly homologous in a wide variety of tissues, ranging from tadpole tails to human articular cartilage [118]. The structures of isolated individual large aggregating proteoglycans (aggrecan) from hyaline cartilage and the aggregates which they form have been visualized at the electron microscopic (EM) level by rotary shadowing [193,194]. While EM studies of the cartilage molecule support a bottle brush-like conformation predicted earlier from physical and chemical analysis [183],
120 NMR and light-scattering studies of these molecules in solution indicate that they have a more compact form [191,192]. Whether the aggrecan found in bone is there as a bone product and not as a cartilage remnant has yet to be determined. The function of these large proteoglycans in bone is also unknown. In other tissues, the CS-proteoglycans serve a hydrodynamic function, aiding in the retention of both water and cations, and the exclusion of anions [195]. The large proteoglycans form gels, which can both swell and retain water, contributing to the osmotic pressure of the tissue [196]. These proteoglycans in cartilage are also responsible for matrix organization [185], in part through interactions with the glycosaminoglycan chain of type IX collagen that trims the type II collagen fibrils [197]. Each of these properties, as reviewed elsewhere [198], depend on the arrangement, size, and number of the constituent CS chains. In cartilage, proteoglycans are believed to play a major role in the regulation of calcification [for review see 199]. The large aggregating cartilage proteoglycans, in solution can inhibit hydroxyapatite formation and growth [200 – 203]. They can also chelate calcium [204] and serve as a source of calcium ions for mineralization if they are degraded into non-Ca2-binding fragments. Although there is some debate as to whether this chelation is involved in the inhibition of mineralization, it is clear that proteoglycans and their component GAGs stearically block hydroxyapatite formation and growth [201]. It has also been shown that when proteoglycans are degraded, their inhibitory ability is lost or diminished both in vivo [205] and in vitro [206]. However, because the relative amount of large CS-proteoglycans in bone is low, it seems unlikely that they have a critical role in preventing osteoid mineralization. Unfortunately, this cannot be proven based on existing data because none of the mutant animals in which the large CS-proteoglycan is altered in cartilage were reported to show bone abnormalities. This is the case for the cartilage matrix deficiency (cmd) mouse, which lacks a core protein [207], and the brachymorphic mouse (bm), which has a deficiency of the enzyme required for the sulfation of glycosaminoglycan side chains [208]. Only in the case of the bm was the mineral examined, and no abnormalities in the bone were detected by X-ray diffraction, infrared, or EM, although cartilage calcification was altered [208]. In the nanomelic chick, which contains a mutated core protein [209], studies show that although there is no overt bone defect, certain bones are abnormally shaped, perhaps due to the altered distribution of load applied to cartilage-deficient tissues during development. The likely functions for the large CS proteoglycans in cartilage are maintenance of tissue hydration and the regulation of mechanical [210] and hormonal signal transduction. However, it is not known if aggrecan or a related molecule provides these functions in bone.
GOKHALE, BOSKEY, AND ROBEY
B. Versican Another large CS proteoglycan related to aggrecan is found in cartilage at low levels. It has been termed versican, as it is found in variable forms in a large number of extracellular matrices. The protein core, with a Mr 360,000, has a structure similar to that of aggrecan with the exception that it lacks the G2 domain. In addition, versican contains only 12 – 15 CS side chains (Mr 45,000) in contrast to 100 in aggrecan [184] (Fig. 8). The distribution of versicans is somewhat ubiquitous as it is found in smooth muscle and aorta [211], glomeruli [212], brain [213,214], and epidermis [215]. A versican-like molecule is highly enriched in developing mesenchyme destined to become bone and may serve to capture space [216]. Versicans from soft connective tissues are capable of aggregating with HA [185]. In addition, it is now apparent that PG-M, identified as a constituent of the matrix surrounding condensing limb bud mesenchyme, is a splice variant of versican [217]. However, versican isolated from bone cell cultures apparently does not aggregate with HA (Fedarko and Gehron Robey, unpublished results). The versican gene localizes to human chromosome 5q12 – q14 [218,219]. The human gene has been isolated [220], is over 90 kb in length, and is composed of 15 exons with a splice variant that utilizes an additional exon [221]. The sequence predicts a 20 residue signal peptide and a 2389 residue mature protein [222]. Exon 1 codes for the signal peptide and a few amino acids found in the mature molecule, and the HA-binding region (G1) is in exons 3 – 6. These exons share homology with the other HA-binding protein, the link protein. This region also contains an Ig-like protein conformation whose function is unknown. Subsequently, exons 6 and 7 are differentially utilized and contain GAG attachment sites. The carboxy-terminal domain (G3), which contains homology to selectins, EGF, and CRP, is contained within exons 9 – 14. The 3-untranslated region contains three different potential polyadenylation sites [220]. The promoter region has a TATA box located 16 bp upstream from the transcription start site. Transfection analysis indicates that there is a positive enhancer between 209 and 445 bp and a negative element between 445 and 632. Other elements present include an XRE (xenobiotic responsive element that may downregulate P450 levels), SP1-binding sites, CRE (cyclic AMP responsive element), and a CCAAT transcription factor-binding site. Based on differential splicing and polyadenylation, three mRNA species of 10, 9, and 8 kb are produced [217]. Versican expression in other tissues can be modulated by a number of factors, including transforming growth factor (TGF-) [223,224], PDGF [223], and interleukin-1 [225]. There is little information on the modulation of its expression during bone development, but it appears to be
121
CHAPTER 4 The Biochemistry of Bone
FIGURE 8
A representation of the chemical features of the widely distributed proteoglycan that is related to, but not identical with, aggrecan. CS, chondroitin sulfate; G1 and G3, globular domains (see text for description); EGF, epidermal growth factor; CRP, C-reactive protein.
upregulated by TGF- in adult human bone cells and fetal bovine long bone cells (Heegaard and Gehron Robey, unpublished data). A potential function for this molecule is to serve as a bridge between the extracellular environment and the cell by binding to HA via the amino-terminal-binding region and to molecules that have yet to be identified on the cell via the carboxy-terminal domain [184]. Versican may also be involved in cell motility and growth [226]. Proteoglycans are known to interact with growth factors, providing a storage site, or facilitating interactions between the extracellular matrix and the cell. Some structural elements within the molecule may also function directly as signals for cellular growth or differentiation [227]. For example, versican contains EGF-like sequences such that upon its destruction, it may serve to stimulate proliferation of osteoprogenitors. EGF has been reported to stimulate proliferation of osteoblastic cells in vitro [228].
C. Small Leucine-Rich Repeat Progteoglycans Another family of proteoglycans is represented by a group whose protein core is characterized by a smaller size and a leucine-rich repeat sequence that is approximately 24 amino acids in length [reviewed in 176,229]. These small proteoglycans are broadly distributed and are found in the
extracellular matrices of many tissues [230]. In cartilage and bone, there are several members of this family, including decorin and biglycan. While they are highly homologous, they exhibit distinctly different patterns of expression and tissue localization, indicative of divergent functions within these tissues. It has been reported that proteoglycans must be removed prior to matrix mineralization and that they inhibit hydroxyapatite crystal growth in test tube experiments, although these two statements do not really comment on the role of a particular species of proteoglycan during matrix mineralization in vivo. It is apparent that there is a net loss of sulfate during matrix mineralization [231,232]. However, using animals and organ cultures, it has been found that a species of sulfate-labeled material accumulates rapidly at the mineralization front [233,234]. One possibility is that the versicanlike molecule is being removed and replaced by the smaller proteoglycans found in bone [235]. Considering the fact that versican has more GAGs per protein core than the small proteoglycans, this would account for the net decrease in sulfate content; however, these data do not rule out the presence of small proteoglycans during matrix mineralization. The leucine-rich core proteins of decorin and biglycan are quite similar in sequence to that of the ribonuclease inhibitor protein. Their structure (in the absence of Ca2 ions) has been determined using X-ray crystallography to 2.3 Å resolution. The ribonuclease inhibitor protein structure
122 consists of 15 leucine-rich repeats, alternatively 28 and 29 residues long, each forming right-handed -- structural motifs resulting in a horseshoe shape as opposed to a globular structure, with the parallel sheets in the horseshoe circumference [236]. The -- motif is frequently seen in proteins as a way of connecting two antiparallel strands; however, it can also be used to connect parallel strands, resulting in a more open structure [237]. This interesting sequence has also been identified in two morphogenetic proteins in Drosophila, chaoptin and toll, in von Willebrand factor-binding protein (GP 1) and in the leucine-rich protein in plasma [238], and in a number of other proteoglycans, which have formed a family, SLRPs, some of which are also found in bone. 1. DECORIN (PG-II, PG-40) Decorin, so named for its ability to bind to and decorate collagen fibrils [118,119,239,240], has also been termed PG-II and PG-40. In soft connective tissues and bone cell cultures, decorin (and biglycan) has DS side chains, whereas in bone, decorin (and biglycan) bears CS [reviewed in 176,241,242]. This difference is of interest because DS is formed from CS by the action of a specific epimerase, and this epimerization can occur uniformly or focally within the GAG chain. The factors that regulate whether a proteoglycan will bear CS as opposed to DS are not known. As one would expect from its proposed function of binding to collagen fibrils, decorin is fairly widely distributed and is found virtually coincident with type I collagen, although the timing of its appearance may be somewhat different. Histochemical studies show the presence of proteoglycans with low molecular weight in the d and e bands of type I collagen fibrils, which disappeared when mineralization occurred [116,117,239]. These histochemical data first suggested that a small proteoglycan might also play a role in mineralization. In the developing skeleton, it decorin is also distributed more uniformly than in the mature animal. In cartilage, decorin is present in very low levels and is restricted to the interterritorial matrix [243]. As bone is formed, it is produced by preosteoblasts and osteoblasts, but its synthesis is not maintained by osteocytes [243]. Nonglycanated forms of decorin (and biglycan) have been found in the intervertebral disk [244]. Decorin has a core protein of approximately Mr 38,000, which includes 10 of the leucine-rich repeat sequences. Although there are three potential GAG attachment sites, only one is utilized with the attachment of a single GAG chain of Mr 40,000, resulting in a molecule with an apparent Mr 130,000 as determined by sodium dodecyl sulfate –polyacrylamide gel electrophoresis (SDS-PAGE) [245] (Fig. 9). The decorin molecule was predicted to contain loops of strands [246]. It has been demonstrated by far-UV CD spectroscopy that both decorin and biglycan exist as predominantly
GOKHALE, BOSKEY, AND ROBEY
sheets, and biglycan has a significantly higher helical structure and assumes different structures in the solution 247. The human gene for decorin has been localized to 12q23 [248 – 250]. In mouse, it is located on chromosome 10, just proximal to the steel locus [251]. A restriction fragment length polymorphism (RFLP ) is also present in humans [252]. The gene is over 38 kb in length and contains 9 exons. The gene shares 55% homology to and is organized in a similar fashion to the biglycan gene (described later) except that the intronic sequences are much longer (two of which are 5.4 kb and greater than 13.2 kb) [249]. The leucine-rich repeat sequences are not organized specifically within the exons. Although the 5 end of the gene eluded characterization for some time, it is now evident that there are two alternatively spliced leader exons (exons 1a and 1b) [249,250]. Exon 1a contains two GC-rich sequences, whereas 1b contains two TATA boxes and one CAAT box that are in close proximity to the transcriptional start site [253]. In addition, AP1, AP5, and NF-b sites were also identified along with a homopurine/homopyrimidine mirror repeat sequence. This region has been postulated to take on a hairpin triplex structure that is unique to decorin compared to the promoters of the other SLRPs. Gene transcription results in a major mRNA species of 1.6 kb and a minor species of 1.9 kb [254,255]. The sequence predicts a 359 residue protein that includes a 30 residue prepro peptide. The synthesis of decorin can be modulated by TGF- [256 – 258]. Dexamethasone upregulates decorin, but downregulates biglycan [259]. The phytoestrogen ipriflavonemetabolite III has been shown to upregulate decorin [260]. Mechanical loading appears to stimulate the synthesis of decorin, but not of biglycan in cartilage [261]. While it would appear that the propeptide is cleaved from the mature decorin in bone, evidence shows that it is maintained in other tissues such as cartilage [262]. Decorin has been shown to bind to and regulate the fibrillogenesis of type I, type II, and type VI collagens [263,264]. It also has a high affinity for thrombospondin [265], TGF- [266], other growth factors, and fibronectin [267,268]. In bone, the proposed functions of decorin are the regulation of collagen fibril diameter and fibril orientation, and possibly the prevention of premature osteoid calcification. However, it is not clear if decorin within the tissue is actually inhibitory to matrix mineralization. Decorin isolated from skin was initially shown to regulate the rate of collagen fibrillogenesis in vitro [269]. Whether bone decorin has the same effect has not yet been demonstrated. Interestingly, bone decorin must be chemically different from tendon decorin, as a different peptide map can be generated by V8 protease [270]. Bone collagen fibrils are a composite of type I with trace amounts of types III, V, and XI and are also trimmed by decorin. Therefore, it is difficult to determine which of these are essential to the
123
CHAPTER 4 The Biochemistry of Bone
FIGURE 9 The two most abundant proteoglycans present in bone matrix are the small chondroitin sulfate/dermatan sulfate proteoglycans, decorin and biglycan. The core protein of each is highly homologous to a number of proteins due to the presence of a leucine-rich repeat sequence. CS, chondroitin sulfate; DS, dermatan sulfate; C – C, disulfide bonding.
regulation of collagen fibrillogenesis and maturation in this tissue. Decorin has been reported to bind with high avidity to type VI collagen [264] and hence may be more concentrated in areas where type VI collagen is more abundant, e.g., in the upper zones of epiphyseal growth plate [271]. Decorin, in solution, has a high affinity for type I collagen (Kd M10) [272] and, in contrast, has a low affinity for hydroxyapatite (Kd 13 g/ mol). When bound to a solid surface, decorin binds rather nonspecifically to apatite relative to other GAG-containing molecules (N, the number of binding sites 333) [273]. Further, it has no detectable direct effect on solution-mediated hydroxyapatite formation or growth [273]. With a low affinity for Ca2 (0.001 g/mg) [267] and a higher affinity for other divalent cations [275], the disappearance of decorin from the collagen fibrils in bone [117] indicates that it is unlikely that it has a direct role in mineralization. However, once removed, other possible nucleators may be exposed, which in turn would facilitate mineralization. Whether this removal also
affects collagen cross-linking and fibril spacing to facilitate mineralization, as well as unmasking nucleators localized under the decorin, remains to be determined, as does the mechanism responsible for the removal of decorin. It is interesting to note that collagen fibril diameter in bone is somewhat reduced in the decorin knockout mouse (M. Young, personal communication), and decorin expression is reduced in certain skin diseases characterized by excessive keratinization [276], stressing the physiologic importance of decorin in regulating fibril formation and collagen – matrix interactions. It is also possible that a key function of decorin is to serve as a binding site for growth factors in these tissues [277]. This may be relevant in the keratinization diseases, in the kidney [276], and in cartilage repair, as well as in the bone where there are other binding sites for these growth factors (see later). It is also of interest to note that there is a decreased expression of decorin in some patients, with OI [169,170,278]. Decorin has also been proposed to play a role in matrix organization by
124
GOKHALE, BOSKEY, AND ROBEY
FIGURE 10
Biglycan immunolocalization in newly formed trabecular bone. Biglycan was found predominantly in osteoid (Ost) and, to a lesser extent, in mineralized matrix (MM) in human bone. Osteocytic lacunae (OCy) also contain high levels of biglycan. Courtesy of Dr. Paolo Bianco.
binding to other matrix proteins, such as fibronectin [267,268]. Interestingly, it has been found that decorin (and biglycan) inhibits the attachment of bone cells to fibronectin, perhaps by blocking the fibronectin cell attachment site that mediates binding. Decorin also binds to thrombospondin, another matrix protein that is speculated to serve as a matrix organizer. It has been reported that decorin binds to TGF- and blocks its activity in fibroblasts [266]; however, it may also activate TGF- in bone cells [279]. 2. BIGLYCAN (PG-I, PG-SI) Biglycan, also known as PG-I and PG-S, is highly related to decorin and exhibits 55% homology at the protein level [238,280]. Despite its high homology to decorin, it is evident that biglycan is uniquely distributed and present in a temporal and spatial pattern not coincident with that of decorin, indicative of a distinctly different function. Biglycan has been localized to the pericellular environment of endothelial, epithelial, and muscle cells [243]. However, in skeletal tissues, biglycan does not appear to be associated with the collagen fibrils, although the biglycan mouse has highly abnormal collagen fibrils in boneskin (M. Young, personal communication). In contrast, it is more abundant in the growth plate [243] and is concentrated in the intraterritorial matrix and in preosteogenic cells, implying a role in early bone development. It is highly upregulated in differentiating bone cell cultures [281] and, interestingly, it is maintained (at the mRNA and protein levels) in osteocytic lacunae and in the canaliculae, which contain
the osteocytic cell processes (Fig. 10). This implies that biglycan may be important in osteocytic cell metabolism. In mineralizing osteoblast cultures, biglycan expression begins during the preosteoblastic phase of the culture and is commensurate with calcium uptake in contrast to decorin, which increases gradually and remains high as mineralization progressed [282]. Biglycan has a protein core of Mr 37,000 to which (in most forms) two GAG side chains are attached. The aminoterminal domain contains the GAG attachment sites, followed by 12 of the leucine-rich repeat sequences (Fig. 9). The first and the last repeat contain a characteristic pattern of cysteinyl residues that result in a particular pattern of intramolecular disulfide bonding [238,283]. The carboxy domain has a sequence that is unique to biglycan (and differs from decorin and other leucine-rich repeat sequence containing proteins). Biglycan has the propensity to self-aggregate in solution [284], although the physiological relevance of these aggregated structures is not known. It has also been determined that nonglycanated biglycan can be secreted into the matrix and that the level of the core protein alone increases with age in human articular cartilage [174]. Other than predicted structures from the amino acid sequences of the biglycan core protein [238], little is known about the tertiary and quaternary structure of biglycan. The gene for the biglycan core protein is localized to Xq27-ter in humans, the only matrix protein that is not on an autosomal chromosome [285]. The gene is 7 kb in size and codes for a 368 residue proform that is processed to form a
125
CHAPTER 4 The Biochemistry of Bone
mature core with 331 residues [238,285]. There are eight exons with relatively small introns compared to those in decorin. Exon 2 codes for the signal peptide, propeptide, and GAG attachment sites. The bulk of the core, the leucine-rich repeat sequences, are not obviously distributed between the remaining exons. The 3 portion of the gene is rich in CT and CA repeats [286] and may result in the Z DNA conformation. The promoter does not contain a TATA box or a CCAAT box, but has a number of cis-acting elements, including SP1, AP1, AP2, NF1, and NF-b-binding sites [285,287]. A number of in vitro transfection experiments have shown that the SP1 site is active, but that it may be binding to a factor other than those described previously [287]. The final mRNA species are 2.1 and 2.6 kb in size. A number of factors have been reported to regulate biglycan synthesis in bone cell cultures and often the pattern is distinctly different from those affecting decorin. TGF- increases biglycan (Gehron Robey, unpublished results) in normal human bone cells and in MC3T3 [258]. Other cell lines were not stimulated by TGF-, but IGF-I and IGF-II increased biglycan levels [288]. Retinoic acid suppresses biglycan in chondrocytes [289]. Dexamethasone and 1,25-dihydroxyvitamin D3 have been reported to decrease its expression in human bone and marrow cell cultures [290,291]. Fluoride, given to cultures of rat osteoblasts in clinically relevant concentrations, decreased GAG chain length and composition [292]. Biglycan (and decorin) binds to type V collagen [293], a collagen abundant in blood vessels. They both also inhibit the activity of antithrombin by interacting with heparin cofactor II. Thus, it has been suggested that one of the functions of biglycan is to create a thromboresistant vascular surface [293]. Whether this is the same for the chondroitin sulfate containing biglycan of bone remains to be determined. In solution, biglycan binds only low amounts of calcium (0.012 mM/ g) but appreciable amounts of small divalent cations such as zinc [294]. Its in vitro interaction with type I collagen can be blocked by increasing phosphate concentrations, implying that biglycan, as distinct from decorin, which is not affected by solution phosphate, interacts through its anionic residues [295]. Also, in solution, biglycan at low concentrations can promote apatite formation, whereas at higher concentrations it inhibits the growth and proliferation of mineral crystals [273]. These effects appear to be due to the highly specific high-affinity binding of biglycan for apatite (Kd 294 g/ mol). The relative extent of apatite formation induced by biglycan compared to other mineral nucleators and its absence from bone collagen fibrils suggest that its primary function may not be related to mineralization of bone. In calcifying cartilage, however, biglycan may play some role in the mineralization process. Biglycan appears to have a regulatory role in bone development. This suggestion is based on the observation that
patients with Turner’s syndrome (genotype 45, XO; i.e., females that are missing one or part of one X chromosome) have decreased biglycan levels, short stature, and other skeletal deformities, including early onset osteoporosis [296]. The biglycan knockout mouse also has a short stature, altered mechanical properties, and altered mineral distribution 297. In contrast, there is overexpression of biglycan in patients with Klinefelter’s disease (genotype 47, XXX) in which patients are excessively tall [296]. Biglycan, like decorin, also binds to a variety of growth factors in solution [266]; thus, it may have functions similar to those of decorin regarding the storage of these growth factors. Biglycan interacts with a number of extracellular components, including the cell-binding domain of fibronectin (thereby inhibiting attachment mediated by fibronectin) [268,298 – 300], TGF- [266], which may alter its activity, and heparin cofactor II [293]. Nonglycanated biglycan has been found to bind to colony-stimulating factor (CSF), which then stimulates the proliferation of nonadherent thymic and peritoneal exudate cells [301]. 3. FIBROMODULIN Fibromodulin is another SLRP that contains keratan sulfate found predominantly in articular cartilage, but also exists in bone [302,303]. The amount of fibromodulin correlates with the size of collagen fibrils in cartilage [304]. In developing bone induced by demineralized bone matrix, fibromodulin is heavily localized to fibrillar bundles [305]. Observations from the fibromodulin knockout mouse have indicated that in the absence of functional fibromodulin, collagen fibrils in tail tendon are abnormal. In these mice, fibrils are significantly thinner, indicating a role for fibromodulin in collagen fibrillogenesis 306. Similarly, a decrease in fibromodulin content and a alteration in structure have been shown to correlate with the aging-related degeneration of vertebral disks 307. The intact protein is approximately 59 kDa, and the protein core shares a great deal of homology with decorin and biglycan, but bears keratan sulfate GAG chains linked to asparaginyl residues rather than chondroitin or dermatan sulfate linked to seryl/threonyl residues [308]. There is also a considerable amount of tyrosine sulfation at the aminoterminal portion of the molecule. The human gene, which is at least 8.5 kb, has been isolated and partially characterized. It has an intron – exon organization, which differs markedly from that of decorin and biglycan. The first exon contains the untranslated region, and most of the coding sequence is located in the second exon, with the remainder residing in the third exon [308]. Fibromodulin interacts with triple helical types I and II collagen [309] and is capable of binding TGF- [310]. Decorin and fibromodulin are the most active collagenbinding proteins, which bind to distinctly different regions on collagen fibrils [309].
126
GOKHALE, BOSKEY, AND ROBEY
D. Heparan Sulfate Proteoglycans Although heparan sulfate proteoglycans are not usually found within the extracellular matrix of bone, they are cell surface associated, either by phosphoinositol linkages that are cleavable by phospholipase C or by transmembrane insertions. There are at least two heparan sulfate proteoglycans present on cells in the osteoblastic lineage, and most likely there are more. The most predominant type is similar to the syndecan family, first identified on mammary epithelial cells [311]. The intact syndecan molecule is Mr 400,000 and contains a core protein of Mr 80,000 and several heparan sulfate side chains of Mr 60,000 [312]. Glycan, a helper receptor for the type I and type II TGF- receptors, has also been identified in most connective tissue cells [313,314]. It has also been determined that bFGF binding to its receptor is facilitated by heparan sulfate on the cell surface [315], although the identity of this molecule has not been determined in bone cells.
E. Other SRLPs, Proteoglycans, and LeucineRich Repeat Proteins The structures and functions of the less abundant proteoglycans have been reviewed by Hardingham and Fosang [316]. In addition to decorin and biglycan, which are class 1 SRLPs [229], a third small proteoglycan whose core protein structure has not been described is the hydroxyapatite (HA)-binding proteoglycan, HA-PGIII, isolated from porcine bone by Nagata et al. [317]. An analogous hydroxyapatite binding protein from bovine bone has been isolated by Hashimoto et al. [318] and was shown to be covalently cross-linked to type I collagen. The localization of this protein in bone has not been described nor are there data on its function, its ability to regulate fibrillogenesis, or bind growth factors. A protein that contains the leucine-rich repeat sequence was originally identified as osteoinductive factor (OIF) [319,320]. However, the osteoinduction component of this preparation is in fact TGF-, and the other protein has now been renamed osteoglycin. It is not known whether this protein exists as a proteoglycan. However, it is of interest that all of the three proteins found in bone with the leucinerich repeat sequences have the ability to bind to TGF-. Another proteoglycan, PG-Lb, also known as epiphycan, has been isolated from calcifying cartilage and is very homologous to osteoglycin [321 – 324]. PG-100 was found in fibroblasts and preosteoblasts [325], and HA-PGII has been found to be associated with forming mineral crystals [317]. Given the low abundance and the variability of expression between one animal species and another, it is not clear what role they play in bone formation and/or mineralization. Until details of the structures and distributions of these
proteins in bone are determined, it is difficult to speculate on their functions. However, by analogy with the other small proteoglycans, it is likely that they will play some role in regulating matrix structure, collagen fibril diameter, and interaction with other matrix molecules.
F. Hyaluronan Hyaluronan is a ubiquitous component of connective tissue matrices [326] consisting of repeating sequences of glucuronic acid and N-acetylglucosamine linked by 1-3 and 1-4 glycosidic linkages. Chains may be several thousand residues in length, placing hyaluronan among the highest molecular weight glycosaminoglycans. Comper and Laurent [195] described the structure of hyaluronan based on models of fluid flow in the 1970s, showing that the hyaluronan formed space-filling gels. In mature bone, hyaluronan is only a small percentage ( 5% on a weight basis) of the total glycosaminoglycans present [327]. The properties of hyaluronan have been reviewed by Knudson and Knudson [328]. Unlike other GAGs, the synthesis of hyaluronan occurs outside of the cell by a large complex of sugar transferases that have yet to be fully characterized. Consequently, the synthesis of hyaluronan is most likely regulated by factors that are quite distinct from those that regulate other GAGs. Hyaluronan is produced at least during early phases of osteogenesis in vitro; however, very little is known about its metabolism in bone. Large amounts are produced by bone cells in culture [329]. To date, there have been no reports on the modulation of hyaluronan production by osteoblastic cells in response to growth factors and hormones that are known to affect bone. This paucity of information stems from the fact that there is no available antibody that binds to it specifically, and because it is synthesized by a multienzyme complex, there is no single probe for measuring mRNA. However, there is an indirect histochemical assay that relies on a peroxidase-conjugated link protein that will bind to hyaluronan in a specific fashion [330], and methods for blocking hyaluronan binding to the hyaluronan receptor (CD44) are available [331]. In order to understand the role of hyaluronan in bone formation, it would be of interest to apply these techniques to bone Hyaluronan binds weakly and nonspecifically to apatite, with affinity constants varying as a function of hyaluronan chain length [332,333]. There is some discrepancy in reports of the effects of hyaluronan on apatite formation and growth. However, these discrepancies may be related to the different systems used for the studies. Thus, where only pH was kept constant, hyaluronan of molecular weight 104 was shown to have no effect on apatite growth at low Ca P products [333]. With higher Ca P products maintained at constant composition, hyaluronan of molecular weight 108
127
CHAPTER 4 The Biochemistry of Bone
inhibited apatite growth slightly [332]. This difference is not attributable to hydrodynamic or size effects, as it has been shown that smaller molecular weight hyaluronans have a similar action [334]. Because the inhibition was small relative to other macromolecules studied, one may conclude that hyaluronan in bone is not apt to have a role in the control of mineralization. The structure and regulation of hyaluronan-binding proteins have been reviewed 335. Hyaluronan forms stable complexes with the large aggregating proteoglycans (aggrecan and versican) of cartilage [336] and with collagen-binding heparan sulfates [337]. The interaction of hyaluronan with the heparan sulfate of cell membranes is believed to be one of the ties that hold cells in place in a variety of tissues and this may also be the case in bone. Hyaluronan has been reported to play a role in cell migration and cell adhesion and invasion, as well as differentiation and proliferation of connective tissue cells [338,339]. An additional function of hyaluronan is the maintenance of tissue hydration, a function that it fulfills in many connective tissues. Removal of this molecule by the enzyme hyaluronidase results in tissue desiccation and facilitates tissue mineralization in cartilage [340]. Because of its abundance in early development, hyaluronan is thought to play a role in embryogenesis. Hyaluronan has been shown to facilitate mesenchymal cell movement [341] and cell adhesion [342] and it appears to bind via CD44 as the receptor [343]. It has been suggested that a major function of hyaluronan in mesenchyme-derived tissues is to act as a space filler, holding water and thereby increasing tissue volume [344]. In chondrocyte cultures, Knudson and Toole [345] have shown that hyaluronan promotes cell proliferation, probably because it provides a space/volume where cells can be anchored and provided with nutrition. Studies also show that hyaluronan can alter the extent of collagen production and proliferation of fibroblasts in culture [346]. Hyaluronan may also function as an organizer of the bone marrow [347] and may regulate hematopoiesis that is supported by steroid [348]. Serum hyaluronan concentrations are elevated in rheumatoid arthritis, osteoarthritis, liver cirrhosis, Werner syndrome, renal failure, psoriasis, and various malignancies; consequently, it may be a useful marker for measuring disease activity [349]. It is also of interest that intercellular adhesion molecule-1 (ICAM-1), a cell surface receptor for hyaluronan found on endothelial cells, is upregulated in inflamed tissue [350]. Consequently, hyaluronan may be involved in this tissue reaction.
IV. GLYCOPROTEINS This class of proteins is characterized by the covalent linkage of sugar moieties attached via asparaginyl or seryl residues as described earlier. Collagen also contains another
form of glycosylation (galactosyl and glucosyl-galactosylhydroxylysine), which is virtually specific to collagen. In bone, the most relevant and abundant glycoproteins are represented by alkaline phosphatase, osteonectin, and the cell attachment proteins, which include, but are not limited to, sialoproteins. These glycoproteins may also be further modified by posttranslational sulfation and phosphorylation.
A. Alkaline Phosphatase Alkaline phosphatase is not typically thought of as a matrix protein. However, studies indicate that under normal conditions, alkaline phosphatase is shed by cells of the osteoblastic lineage in culture as they pass through the G2/M phase of the cell cycle [351], and a Ca2-binding glycoprotein with alkaline phosphatase activity has been isolated from mineralized matrix extracts [352]. It is possible that this protein is indeed alkaline phosphatase that has been shed from the cell surface by cleavage of its phosphoinositol type linkage with phospholipase C or in a membranebound form (matrix vesicles). Although this enzymic activity is shared by many types of tissues, there is no doubt that the induction of alkaline phosphatase activity in uncommitted progenitors marks the entry of a cell into the osteoblastic lineage and is a hallmark in bone formation. However, it should be noted that many studies utilize the enzymatic activity alone to measure the presence or absence of alkaline phosphatase. This may be somewhat misleading, as it is possible that the protein can be present but inactive or, conversely, that the protein can be partially degraded and still maintain enzymatic activity. As in all studies where the most reliable information is obtained by measuring both mRNA levels and protein levels, in studies involving alkaline phosphatase, a clearer picture of the metabolism of this protein may be best obtained by measuring the mRNA and the enzymatic activity, as well as the status of the protein (intact versus degraded). Histological localization of alkaline phosphatase through utilization of a chromophore-generating enzyme substrate marks very specific sites within bone marrow and sites of new bone formation [353 – 356] (Fig. 11). Developmental studies in vivo and in vitro have shown that expression precedes mineralization and is maintained during early stages of hydroxyapatite deposition [10 – 12,357 – 359]. Within marrow, cells that are supportive of hematopoiesis are alkaline phosphatase positive (Westen – Bainton cells) and represent, at least in part, members of the bone marrow stromal stem cell population [360]. It has been reported that these cells are related to preadipocytes. In certain pathological conditions, when hematopoiesis is halted, these cells lose their alkaline phosphatase activity and proceed to form fat cells [361]. During osteogenesis, the alkaline phosphatase reaction product very
128
FIGURE 11
Alkaline phosphatase in developing bone. By histochemical staining for alkaline phosphatase activity during development, areas that are destined to become bone, as shown here in developing human subperiosteal bone, can be clearly illustrated. The fibrous layer (F) of the periosteum is negative, whereas preosteoblasts (POb) and osteoblasts (Ob) produce high levels of activity. Although a glycoprotein with alkaline phosphatase activity has been isolated from the bone matrix, it is not easily detected in mineralized matrix (MM) by this histochemical assay. Courtesy of Dr. Paolo Bianco.
clearly demarcates areas that will become bone from those that will not. Initially, the cells are all rather flat and spindle shaped, indicative of preosteoblasts, but as osteogenesis continues, some cells (usually those in close proximity to blood vessels) take on the morphology of osteoblasts (plump, highly biosynthetically active cells). As cells mature further to become osteocytes, alkaline phosphatase activity is lost, presumably due to the fact that its function is no longer required [355]. The enzyme exists as a dimer and the identical monomers have an Mr 50,000 – 85,000 depending on animal species and degree of posttranslational modification. Each monomer binds to two zinc and one magnesium ion in each active site. Although the bone/liver/kidney isozyme is the product of the same gene, there are tissue-specific posttranslational modifications that can be detected by monoclonal antibodies [362 – 364]. The enzyme is glycosylated
GOKHALE, BOSKEY, AND ROBEY
and attached to the cytoplasmic membrane on the external surface through a phosphatidyl-inositol-glycan group, which can be cleaved by phospholipase C, thereby releasing it from the cell surface [365,366]. Based on sequence and structural analyses, the E. Scherichia coli enzyme shows extensive homology with mammalian alkaline phosphatase, the major differences being on the enzyme surface [367]. The structure of the enzyme isolated from E. coli has been determined by X-ray crystallographic analyses of single crystals of the (a) Zn2-containing enzyme [368], (b) enzyme with inorganic phosphate bound to the active site [369], and (c) cadmium salt of the enzyme [369]. From E. coli single crystal data, the enzyme takes the form of a twofold symmetrical dimer (100 50 50 Å in dimensions), with the two active sites separated 30 Å from each other. Each subunit consists of a central 10-stranded sheet surrounded by 15 -helices of various lengths. The active site is in the carboxyl end of the central sheet, and all the ligands for the three metal ions come from this structure. One serine has been shown to be involved in phosphorylation and dephosphorylation reactions [369]. The phosphate is closely associated with all three metal ions, and the guanidinium group from an arginyl residue. The coordination of the metal ions in the active site is reported to be very similar to the active site of phospholipase C [370], which shows no other structural homology with alkaline phosphatase. The human gene for alkaline phosphatase is located on chromosome 1 [371]. It contains 12 exons, is over 50 kb in length [372], and has an RFLP [373]. The rat gene is at least 49 kb with 13 exons and has a similar gene organization [374]. The gene predicts a protein with 524 amino acids that includes a 17 amino acid signal peptide. The potential for heterogeneity between the bone/liver/kidney isozymes may reflect the fact that there are five potential glycosylation sites. The C-terminal region is hydrophobic, as would be expected for a protein that is linked to the cell membrane via phosphoinositol. It has been determined that regulation of the bone/liver/kidney isozyme is controlled by two leader exons, 1A and 1B [375], with alternative promoters separated by 25 kb. The upstream promoter is used preferentially by bone cells and facilitates its high level of expression in this cell type [376]. The transcription start site is preceded by a GC-rich region, a TATA box, and three SP1 sites [377,378]. The downstream promoter is constitutively active, produces low levels of activity, and is used in the kidney. Three mRNA species of 2.5, 4.1, and 4.7 kb are produced as the result of differential splicing [379]. The list of factors that regulate alkaline phosphatase is lengthy and the results are extremely variable. Expression can be affected by oxygen tension [380] and, in general, requires the presence of serum [381]. In human bone cells,
129
CHAPTER 4 The Biochemistry of Bone
1,25-dihydroxyvitamin D3 stimulates alkaline phosphatase [382]. In immortalized rat osteoblastic cells, retinoic acid also stimulated the enzyme [383], and the region 108 to 45 from the transcription start site was required for this stimulation [384]. Pretreatment of fibroblastic cells with substances that induce cAMP followed by retinoic acid treatment causes an elevation in the mRNA coded for by the exon 1B promoter [385]. IL-4 causes an increase in alkaline phosphatase mRNA in proliferating osteoblastic cells [82]. Dexamethasone increases the utilization of the upstream promoter, resulting in an increase in the tissue-specific expression of alkaline phosphatase [386]. Calcitonin increases alkaline phosphatase expression in osteosarcoma cells [387]. In deer antlers, an interesting model of bone formation, alkaline phosphatase was found to be increased by IGF-I [388]. Ascorbate has also been reported to upregulate alkaline phosphatase expression in chrondrocyte cultures [389]. Alkaline phosphatase activity has been found to be decreased by IL-10 [90] and by lead [91]. Reaction mechanisms for the E. coli enzyme have been predicted from kinetic labeling studies [390], as well as from the single crystal structure [369]. These studies show that the enzyme has phosphate transfer activity and hydrolase activity, transferring a phosphate to or from the serine in the active site from or to the substrate, respectively. The predicted structural identity between mammalian and E. coli enzymes [367] implies a conserved mechanism and suggests that the bone (and cartilage), liver, and kidney isozymes have similar nonspecific phosphomonoesterase activities. The specific function of alkaline phosphatase in bone has been debated for more than 60 years. In the 1920s, Robison, before the existence of the multiple isozymes was known, suggested that the function of this enzyme was to hydrolyze phosphate esters, providing a source of inorganic phosphate [391]. Human studies have shown that different isoform structures are likely to exist due to differences in glycosylation. These studies have also shown that higher alkaline phosphatase activity was present in trabecular bone than in cortical bone [392]. In diseases such as rickets, with impaired mineralization, activity of the bone isozyme is increased [393]. It has also been observed that the commencement of mineralization in culture coincides with increased alkaline phoasphatase activity, indicating that this enzyme was necessary for proper mineralization [see reviews in 394,395]. This hypothesis was confirmed by the discovery of mutations in the alkaline phosphatase gene in hypophosphatasia, a disease characterized by improper mineral deposition [396], and by the observation that cells that do not normally mineralize will form a mineralized matrix when transfected with the alkaline phosphatase gene [397]. Because mineral deposition in vitro will occur (i) when alkaline phosphatase is added to cell-free solutions of calcium ions and phosphate esters [398], (ii) in cell cultures
in which alkaline phosphatase activity is inhibited by levamisole [399,400], and (iii) in the absence of alkaline phosphatase substrates [401,402], the transfection studies do not necessarily identify a definitive function. Mice with null mutations in either the tissue-nonspecific alkaline phosphatase [403] or the bone-specific alkaline phosphatase also provide evidence of the importance of alkaline phosphatase for mineralization [404]. Even the tissue-nonspecific alkaline phosphatase knockout shows increased osteoid and defective growth plate development, leading one to question what the specific mechanism of action of alkaline phosphatase might be. It had been suggested that alkaline phosphatase could also function as a protein phosphate transferase in bone [405] and such transferase activity was noted at physiologic rather than basic pH (required for optimal hydrolytic activity). The transferase activity is in line with the mechanistic studies, and the abundance of phosphorylated proteins implies the need for such an activity. In this light, it should be noted that using non hydrolyzable adenosine triphosphate derivatives retards mineralization in culture [406], perhaps because matrix protein phosphorylation is impaired. A role for alkaline phosphatase in signal transduction may also be speculated based on its analogy to other glycan-linked proteins [365,407]. However, few experimental data support this function. Although a precise function for bone alkaline phosphatase is not known, it seems apparent from hypophosphatasia defects [396] that the enzyme plays a crucial role in mineralization. Because its phosphohydrolase activity is optimal at pH 10 [394], a less direct role than hydrolyzing phosphate esters to produce elevated phosphate concentrations is likely. Part of the function of alkaline phosphatase in mineralization may be associated with its abundance in the membrane-bound bodies, matrix vesicles, believed to be the foci of initial mineralization (vide infra). Another postulated function is the dissolution of calcium pyrophosphate dihydrate crystals in cartilage [408].
B. Osteonectin (SPARC, BM-40) As a result of the development of novel techniques for the extraction of bone matrix proteins in a nondegraded form [2,3], one of the first noncollagenous bone matrix proteins to be isolated and characterized was osteonectin [409]. Osteonectin can constitute up to 15% of the noncollagenous proteins in bone depending on the developmental age and the animal species [410]. Although it was initially thought to be bone specific, with the advent of sensitive antibodies and in situ hybridization, it became apparent that osteonectin is expressed in a number of tissues during development and by many cell types in vitro. In fact, osteonectin was independently identified as SPARC (secreted phosphoprotein acidic
130
GOKHALE, BOSKEY, AND ROBEY
and rich in cysteine) in parietal endoderm [411], culture shock protein in endothelial cells [412], and BM-40 from the EHS mouse basement membrane tumor [413,414]. By in situ hybridization in mouse and human tissue [415,416], osteonectin was found in skin, tendon, developing whiskers, certain nerve trunks, and parietal endoderm. Steroid-producing tissues such as the adrenals, testis, and ovary also produced osteonectin [417]. Immunohistochemical analysis localized transient osteonectin protein to decidua [418] and developing spermatozoa [419], whereas low, but constant, levels were found in renal distal tubule epithelial cells [420,421] and duct cells of salivary gland [420]. In the skeleton, osteonectin mRNA is found in odontoblasts, periosteal cells, osteoblasts, and, to a lesser extent, in osteocytes and hypertrophic chondrocytes [415,416]. By immunohistochemistry (Fig. 12), osteoblasts and osteoprogenitor cells contain osteonectin [422,423], as do odontoblasts [424], periodontal ligament, and gingival cells [425]. Again, low levels are found in chondrocytes and hypertrophic chondrocytes in the growth plate [422,423,426]. During subperiosteal bone formation, there is no detectable osteonectin antigen in the cells or matrix in the fibrous outer layer. Preosteoblasts have a small amount of cytoplasmic reactivity, but osteonectin cannot be detected by immunostaining. In the osteoblastic layer, there is a dramatic
FIGURE 12
upregulation of expression with intense staining of both the cytoplasm and the osteoid, but expression is much lower in osteocytic cells [12,422]. Woven bone contains less osteonectin than more mature, lamellar bone [410], which may explain the differences in concentration between bones of rats (less lamellar in character) and human (more lamellar and osteonal in nature). By radioimmunoassay (RIA), it is noted that bone contains 1000 – 10,000 times more osteonectin than any other connective tissue, such as skin and tendon [427]. Osteonectin is produced transiently in many tissues undergoing differentiation, and it is produced constitutively only by cell types that appear to have ion transport as a common feature. It is associated with development, remodeling, cell turnover, and tissue repair. In vitro studies have revealed that it is involved in counteradhesion and antiproliferation of cells 428. It accumulates in mineralized matrices (dentin, bone, calcified cartilage) and in the granules of platelets [429]. It has also been found to be associated with prostate tumors [430]. The identification of a bone matrix protein in platelet granules was the first in a series of discoveries that many bone matrix proteins are found in platelets and megakaryocytes. The reason for the localization of these proteins in cells of this distinctly different lineage is not known. Osteonectin has an apparent Mr 35,000 without reduction of disulfide bonds and appears to increase in size up to
Osteonectin immunolocalization in trabecular bone. (A) Osteonectin is highly enriched in the osteoid and in osteoblasts (Ob) and, to a lesser extent, in osteocytes (open arrow). In addition, osteonectin is also enriched between lamellae (small arrows), which is better demonstrated by immunofluorescence (B). Courtesy of Dr. Paolo Bianco.
CHAPTER 4 The Biochemistry of Bone
Mr 40,000 – 46,000 following reduction, indicative of intrachain disulfide bonds (Fig. 13). The estimated size is Mr 32,000 by molecular sieve chromatography and Mr 29,000 by equilibrium sedimentation, suggesting that it takes on a rather compact hydrodynamic conformation [429]. Osteonectin is also phosphorylated and glycosylated, two posttranslational modifications that may be regulated differentially depending on the tissue of origin. Due to the nature of the amino acid composition and the nature of the posttranslational modifications, osteonectin is acidic with a pI of 5 [431]. There is a single gene (20,000 kb) for osteonectin located on human chromosome 5 [432] at 5q31 – q33, and with one RFLP in the 5 region [433]. The gene contains 10 exons, several of which code for potentially functional regions. While the exons are relatively small (130 bases), the first intron is approximately 10 kb in length. The coding sequence predicts a 17 residue signal peptide and a 286 residue mature protein. The signal peptide is contained in exon 2, whereas exon 3 contains the amino terminus. This region varies from one animal species to another; however, it contains a preponderance of acidic residues. If this region takes on an -helical conformation, the orientation of the carboxy side chains away from the helix would result in the generation of 12 low-affinity calcium-binding sites. Subse-
FIGURE 13
The chemical characteristics of osteonectin indicate the presence of two -helical regions at the amino terminus, along with an ovomucoid-like sequence with extensive disulfide bonding and two EF hand structures.
131 quent exons code for a cysteine-rich region with homology to an ovomucoid-like (serine protease inhibitor) sequence. A structure that is highly homologous to the EF hand (highaffinity Ca2-binding structure) found in calmodulin is encoded by exon 10 and another one, although not as highly homologous, is found in exon 9 [434 – 436]. Sequence predictions indicate that the EF hand domain is in the C terminus. Several other domains, in addition to the EF hand region, have been defined, including a disulfide-rich domain and a pentapeptide KKGHK domain [437]. Three additional genes show high homology with osteonectin. One gives rise to a synaptic junction glycoprotein in rat, SCI [438], and hevin in human [439], a neural retina protein, QR1 [440], testican [441], and FRP [442]. These molecules contain the carboxy-terminal portion of osteonectin (three-fourths of the molecule) and substitute different amino-terminal ends. The role of these homologous forms of osteonectin in their respective tissues is not known. The promoter region of the gene has been isolated and shown to be active by transfection experiments [430,435, 443 – 445]. The promoter does not contain TATA or CCAAT sequences but contains a purine-rich region with GA repeats between 55 and 126. The bovine promoter has a complimentary CCCT-rich sequence further upstream that is believed to cooperate with the GGA-rich region to control transcription. This information, along with S1 nuclease sensitivity, suggests that this region may become hinged or form a triplex conformation. A novel nuclear factor from osteoblastic cells that binds to this region has been identified [445]. There are also numerous other sites, including multiple SP1 sites, one AP1, one CRE, a growth hormone element (GHE), a heat shock element (HSE), and a metal responsiveness element (MRE). The first exon also contains four CCTG repeats, sequences that have been implicated in transcriptional control. In connective tissues, the gene is transcribed to form mRNA species of 2.2 and 3.0 kb. The cDNA codes for a 17 residue signal peptide and a 303 residue protein with a 17 residue signal peptide [411,432,434,443], but with no evidence to suggest the existence of a propeptide [245]. Osteonectin may be differentially glycosylated and/or phosphorylated [446], a possible explanation for differences noted in molecular weights and reactivity with different antibody preparations. There are at least two potential Nglycosylation sites that bear diantenary oligosaccharides (an intermediate between high mannose and complex type oligosaccharides that contains variable amounts of sialic acid and fucose) [447]. It has been found that intracellular forms of osteonectin found in megakaryocytes and certain osteosarcomas have a different carbohydrate structure from secreted forms of osteonectin found in bone [448]. These structural differences may also be reflected in functional differences [449] and could explain differences in the reactivity of osteonectin with different monoclonal antibodies [450].
132
GOKHALE, BOSKEY, AND ROBEY
Factors that regulate the synthesis of osteonectin are not well understood. It appears that the control of osteonectin synthesis is deregulated rapidly, as cells of various tissues begin to produce osteonectin in vitro when they do not normally synthesize it in vivo. Biosynthesis has been studied in a number of cell culture systems, and the regulation and association of osteonectin with mineral formation appear to be dependent on the animal species and culture conditions. In cartilage, IL-1 decreases osteonectin synthesis, whereas TNF- and IL-6 have no effect. TGF-, PDGF, and IGF-I are stimulatory and able to reverse the effect of IL-1 [451]. In bovine bone cell cultures that exhibit extensive mineralization, osteonectin appeared at early stages and remained high thereafter [282]. The effect of TGF- is variable and a stimulation [452] as well as a lack of effect [453] have been reported. The expression of osteonectin by normal human bone cells is not altered dramatically by any treatment (Gehron Robey, unpublished results), although very modest increases with dexamethasone, retinoic acid, IGF-I and dibutyryl cAMP have been reported in other systems [288,411,454]. Interestingly, osteonectin is expressed at constant levels even after heat shock, whereas other proteins decrease after this form of stress [455]. This resistance to heat shock may be related to the presence of a potential heat shock element in the promoter. The only structural studies of osteonectin/SPARC/BM40 other than those predicted from cDNA sequences [e.g., 434,437,456] are NMR evaluations [457] and circular dichroism (CD) studies [71]. The CD measurements were interpreted as showing that the protein conformation had 6% helix, 26% sheet, with the remainder nonordered in the absence of Ca2, and less than 3% helix, 21% sheet, and 82% nonordered structure in its presence. Analysis of the crystal structure has shown that a follistatin-like
FIGURE 14
domain is related to the serine-protease domain of the Kazal family. There is an insertion into the inhibitory loop in BM-40 and a protruding N-terminal hairpin with striking similarities to epidermal growth factor 458. NMR evaluations showed the presence of a typical EF hand [457], which in other systems is involved in calcium chelation and calcium transport (Fig. 14). The initial investigations of the function of osteonectin demonstrated that when bound to denatured collagen, osteonectin bound calcium and phosphate ions, suggesting that it was promoting mineral deposition [409]. Osteonectin and its metalloprotease-cleaved fragments also bind to type I collagen [409,459], to types III and V collagens [448,460], and to thrombospondin [461]. Such interactions are likely to be important in determining the organization of the osteoid in bone. Later studies showed that osteonectin was an efficient inhibitor of hydroxyapatite formation in solution [462]. As discussed later, these concepts are not mutually exclusive. Because macromolecules that in low concentration act as nucleators by binding to matrix to provide suitable substrate surface for nucleation can in higher concentrations bind directly to one or more faces on the growing crystal. This can block further growth of the crystal. The tissue distribution of osteonectin within bone suggests, however, that it is not involved in the initiation of mineralization [463]. Expressed by cells in the unmineralized, mineralizing, and mineralized bone, osteonectin accumulates only within the mineralized matrix. Whether it has a specific function in further regulating growth and proliferation of mineral or simply accumulates within the mineralized tissue because of its affinity for hydroxyapatite (Kd 8 108, 11.3 mg osteonectin/gm apatite [460]) is not yet known. A role in the regulation of mineralization does not appear to be a principal function for osteonectin because it is
Structure of an EF hand high-affinity Ca2-binding site. Depiction of the theoretical structure and the amino acid sequence for the EF hand, which has an extremely high affinity for ionized calcium. Courtesy of Dr. Neal S. Fedarko.
133
CHAPTER 4 The Biochemistry of Bone
a widely expressed protein and, in other tissues, it appears to be more involved in regulating cell shape and cell matrix interactions [437] via calcium binding. For example, a fragment of the molecule has been shown to regulate the proliferation of endothelial cells [464]; the intact molecule acts to regulate cell shape in an ion-independent fashion [465,466]. Similar to thrombospondin, it modulates cell – matrix interactions [467]. The original description of the young osteonectin knockout mouse focused on the formation development of cataracts [468]. More recent studies of older mice indicate that the mice develop osteopenia with a significant loss of trabecular bone associated withdue to a decreased rate in bone formation [469]. A number of functions have been suggested by are derived from experiments utilizing synthetic peptides from different parts of the molecule and determining their effect on endothelial cell cultures. While this is an extremely useful approach that provides much intriguing information, it is not clear whether the intact molecule has the same functions or if the effects are cell specific. A synthetic peptide influenced endothelial cultures undergoing tube formation by decreasing the synthesis of fibronectin and thrombospondin-1 and by increasing the synthesis of plasminogen activator inhibitor-1 (PAI-1) [470]. The intact protein decreased bFGF-induced endothelial cell migration [471]. The binding of osteonectin to endothelial cells similarly may be through one of the EF hand structures at the carboxy terminus [472]. Osteonectin and a peptide derived from a noncalcium-binding region have the ability to stop endothelial cell cycle progression, although it does not have the same effect on fibroblasts [473]. In endothelial cells, it may play a role in barrier function by regulating endothelial cell shape [474]. In addition a peptide that becomes conformationally constrained by binding to Ca2 inhibits the proliferation of endothelial cells [475]. Overexpression of osteonectin causes attachment and spreading of endothelial cells in calcium-deficient medium [476,477]. The various activities of osteonectin in embryogenesis and during wound repair have been reviewed extensively by Sage and co-workers [428,478]. Osteonectin may also bind to copper and serve as a source of this ion during angiogenesis [479]. Osteonectin is produced during mid- and late stages of wound repair [480] and interacts with PDGF (specifically with the chain) and may modulate its activity by inhibiting binding to its receptor [481]. Treatment of fibroblasts with osteonectin induces metalloproteinase activity [482] and may anchor plasminogen and increase its activation, pointing to a function in tissue remodeling [483]. The disassembly of focal adhesion may be through a follistatin-like region that contains one of the EF hand structures [484]. Another interesting study utilized antiosteonectin in developing frog embryos and found that neurulation and myotome development were impaired [485]. Overexpression of osteonectin in C. elegans caused devel-
opmental abnormalities as well as paralysis [486]. Each of these studies demonstrate that osteonectin can influence development in a number of connective tissues. However, while it is known to affect bone development in the mouse, a its specific role in bone has yet to be established.
C. Tetranectin Tetranectin is a tetrameric protein with a Mr of 21,000 (subunits with Mr of 5800) that was first isolated from serum and found to bind to the kringle 4 domain of plasminogen [487]. It is immunolocalized in developing woven bone [488] and is expressed by osteoblastic cultures undergoing matrix mineralization [488]. The cDNA has been cloned [489] and the gene has been isolated [490]. The gene is 12 kb in length and contains three exons. The cDNA predicts for a 21 residue signal peptide and a 181 residue mature protein. It has sequence homology with asialoprotein receptor and the G3 domain of aggrecan and versican core proteins [491]. Overexpression of tetranectin by tumor cells caused an increase in matrix mineralization upon implantation into nude mice [488]. These data suggest that tetranectin may play a role in mineral deposition.
D. RGD-Containing Glycoproteins One of the major breakthroughs in the field of cell – matrix interactions was the discovery of a sequence contained within matrix proteins that would bind to cell surface receptors, thereby mediating attachment [492]. These cell surface receptors belong, for the most part, to the family of integrins. These receptors are formed by one subunit and one subunit, each of which have a cytoplasmic extension that may associate with intracellular signaling pathways, a transmembrane domain and an extracellular domain. The extracellular domains of the and subunits form a binding pocket that recognizes the cell attachment sequence in the extracellular matrix protein [reviewed in 493,494]. While there are a number of amino acid sequences that bind to integrins, the most frequently utilized sequence is ArgGly-Asp (RGD) [492]. In bone there are at least seven matrix proteins that contain RGD: thrombospondin, fibronectin, vitronectin, collagen (described earlier), osteopontin, bone sialoprotein, and fibrillin. The reason for this redundancy is not entirely clear, although a clue may lie in the fact that these proteins exhibit different patterns of expression during bone formation. Consequently, the interaction of cells with an ever-changing matrix may mediate, in part, changes in cellular metabolism. What follows is a description of the RGD-containing proteins identified in bone, presented roughly in the order of their appearance during developmental bone formation.
134
GOKHALE, BOSKEY, AND ROBEY
1. THROMBOSPONDIN(S) This complex, modular glycoprotein [495 – 498] was first isolated and characterized from the granules of platelets [499]. Since its initial discovery, it has been found that there are at least five members of this family. The various forms are found in a large variety of connective tissues, in particular, in areas of demarcation such as at the dermal – epidermal junction in the skin, surrounding muscle fibers, separating glandular epithelium, and in the lung in peritubular spaces [500,501]. Thrombospondin is relatively less abundant in mineralized matrix relative to other glycoproteins; however, due to its complex chemical nature, it is most likely active in modulating cell metabolism. The expression of thrombospondin during development is sporadic and usually coincides with morphogenetic events such as cell proliferation, migration, and commitment, followed by its disappearance as differentiation continues [502]. In bone, it is found during early stages of osteogenesis. Immunohistochemical localization indicated low levels of expression in the periosteum, with primary localization in developing osteoid [503] (Fig. 15). Osteoblastic cells are stained intensely. There is a moderate accumulation of thrombospondin in mineralized matrix [504] and, by Western blotting, the protein can be detected in bone matrix extracts [503]. Thrombospondin is a highly complex molecule with a Mr 450,000 [reviewed in 505,506] (Fig. 16), composed of three identical subunits ranging in Mr from 150,000 to
180,000 that are held together by disulfide bonds. Each monomer has a number of intramolecular disulfide bonds that give rise to a molecule with a roughly dumbbell shape with distinct functional domains. The small amino-terminal globular domain contains a fibrinogen-like sequence along with a region that may have cell-binding activities [507] and heparin- and platelet-binding sites. In addition to homologies to the propeptide of the (1)I chains of types I and III collagen, von Willebrand factor, and the circumsporozoite protein from Plasmodium falciparum, this small globular domain is attached to an extended stalk region that contains three properdin-like (type I) and three EGF-like (type II) repeat sequences. There is a cluster of cysteine residues in the stalk region that participate in the cross-linking of the monomers and binding sites for types I and V collagens, thrombin, fibrinogen, laminin, plasminogen, and plasminogen activator. A large disulfide-bonded domain makes up the carboxy-terminal region of the molecule and contains sequence homologies to parvalbumin and fibrinogen, with seven calmodulin-like (type III) repeat sequences, although this sequence does not take on the EF hand structure. This region binds to the histidine-rich glycoprotein of serum, activates platelet aggregation, and has multiple Ca2-binding sites. Binding of Ca2 participates in the conformation of the globular domain. The RGD sequence is within the Ca2-binding region; consequently, it is not clear whether the RGD actually is active in mediating cell attachment under normal physiological conditions [508].
FIGURE 15 Thrombospondin immunolocalization in developing bone. Thrombospondin is first expressed in the preosteoblastic layer (POb) in the periosteum of human developing long bone and is highly concentrated in osteoid and osteoblasts (Ob). Somewhat lower levels are maintained in the mineralized matrix (MM). Courtesy of Dr. Wojciech J. Grzesik.
135
CHAPTER 4 The Biochemistry of Bone
FIGURE 16 Thrombospondin is a disulfide-linked trimer that has globular domains at the amino and carboxy terminus interconnected by a stalk region. Each of these domains has a number of binding sites for other proteins, suggesting numerous potential functions in cell – matrix interactions. The cell attachment consensus sequence, RGD, is in the carboxy-terminal domain; however, its availability depends on the calcium ion concentration, which is known to affect the conformation of this region.
Initially it was thought that there was only one gene for thrombospondin. However, it is now apparent that in humans there are at least five (TSP-5 being equivalent to the cartilage molecule COMP), located on chromosomes 1 (TSP-3), 5 (TSP-4), 6 (TSP-2), 15 (TSP-1), and 19 (COMP) [509 – 514], all at least 16 kb in length. While the coding sequences are all highly homologous and differ only in the number of times that the type I, II, and III sequences are repeated, they utilize different promoters [515]. The complete pattern of expression of the different thrombospondin genes is not complete to date [516], although it is known that TSP-1, TSP-2, and TSP-3 are expressed in bone [517,518]. A promoter from the thrombospondin 1 gene has been isolated and characterized [519,520]. It contains a TATA box and an Egr1 site that is flanked by overlapping GC boxes, followed by a GC-rich region. Binding sites for NFY, AP2, SP1, and an SRE have also been identified. Based on the inhibition of TSP-1 transcription by c-jun, an AP1 site may also be present [521]. The resulting mRNA is 6.1 kb [495]. The organization of the TSP-2 and TSP-3 promoters is similar [522 – 524]. Thrombospondin synthesis has been demonstrated in several cell culture systems, including those of adult human
trabecular bone cells [503], in cultures of rat marrow stromal fibroblasts undergoing nodule formation [525], MC3T3 [526], and MG-63 cells [527]. Synthesis appears to be inhibited by dexamethasone [527]. While reducing the overall net synthesis of thrombospondin, TGF-, increases the amount of thrombospondin retained by the cell layer/matrix fraction [503]. Mutant (Tsp-2 null) mice have abnormalities in connective tissue structure and function, including fragile skin [528] and bone (Bornstein, personal communication). Mutant mice have increased cortical bone thickness and density, which has been explained in terms of effects on cell adhesion and differentiation, and collagen fibrillogenesis [529]. While the precise functions of the thrombospondins are not known, they have been postulated to play a role in development, angiogenesis, tumorogenesis, and wound healing [516,530 – 536], and bone remodeling 537. Thrombospondins bind to the small proteoglycan, decorin [265], and along with this molecule may bind to growth factors such as TGF-, which later serve as cell signals. Thrombospondins also bind to osteonectin [461], and considering their colocalization within the granules of platelets, this complex, along with PDGF and TGF-, may have a
136
GOKHALE, BOSKEY, AND ROBEY
function in repair processes in soft connective tissues as well as in fracture healing. The functions of the various isoforms of thrombospondin have been indicated by several in vitro studies. In an osteoblast-like cell culture (MC3T3 cells), mRNA for one form of thrombospondin increased during differentiation [526], suggesting that thrombospondin has a role in this process. Thrombospondin has been shown to increase osteoclast resorption of dentin slices independent of increases in osteoclast attachment [537]. Thus it might similarly affect differentiation in bone, either by regulating the latter processes or by directly causing the activation of cell signals. Thrombospondin has been shown to bind to TGF- [525]. In vitro, thrombospondin has been found to be active in the attachment of osteoblastic cells. However, cell spreading required the synthesis of other molecules [503]. Thrombospondin has been reported to bind the vitronectin receptor, v3, in an RGD-dependent fashion. However, it was found that attachment and spreading of osteoblastic cells were not inhibited by GRGDS, a competitive inhibitor used routinely to demonstrate the RGD dependency of cell attachment [504,538]. Thrombospondin may also bind to decorin, which is known to interfere with the cell attachment to fibronectin [539]. The precise cell surface receptor responsible for mediating cell adhesion to thrombospondin is not known, although a cell surface receptor that recognizes the sequence VVM in the carboxy terminus of the molecule has been identified [540]. In properdin-like repeats, the VTCG sequence has been found to mediate the
FIGURE 17
attachment of platelets, monocytes, endothelial cells, and certain tumor cells via cell surface CD36 [541 – 543]. Thrombospondin has been shown to interact with fibrinogen/fibrin but not with heparin [544]. Mice that lack thrombospondin have disordered collagen in their fragile soft tissues, increased bone density, and altered fibroblast cell attachment [528]. Bone mineral properties have not yet been determined. However, the properties of these mutant animals confirm the importance of thrombospondin in bone development. Interestingly, unlike other RGD proteins in bone, thrombospondin does not mediate osteoclast cell attachment [545,546]. 2. FIBRONECTIN Fibronectin, one of the most abundant extracellular matrix proteins, is also a major constituent of serum [reviewed in 547,548]. It is produced by virtually all connective tissue cells at some stage of development and accumulates in extracellular matrices throughout the body. Fibronectin appears at an early stage of bone formation during development [504] or during induction by demineralized bone matrix [549,550]. Osteoblasts and osteocytes stain intensely for fibronectin and it is also accumulated in mineralized matrix [504] (Fig. 17). Western blotting analysis of bone extracts indicates that it is relatively abundant. Although fibronectin is a product of osteoblastic cells, it could also be adsorbed from the circulation; consequently, there may be multiple forms from different sources entombed within mineralized matrix.
Fibronectin immunolocalization in developing bone. There is virtually no expression of fibronectin in the periosteum and preosteoblasts (POb) of developing long bone, but there are high levels of expression in the osteoblastic layer (Ob). In addition, high levels of fibronectin are maintained in the mineralized matrix (MM). Courtesy of Dr. Wojciech J. Grzesik.
137
CHAPTER 4 The Biochemistry of Bone
Fibronectin is a dimeric protein with a Mr 400,000, composed of two subunits of Mr 250,000 that are highly homologous, but variable depending on the cell source, held together by two disulfide bonds near the carboxy termini (Fig. 18). Each of the subunits has multiple domains that bind to fibrin (domains I and VIII), heparin, and certain bacteria (domain I), gelatin and collagen (domain II), DNA (domain IV), cell surfaces, including the RGD site (domain VI), and another heparin-binding site (domain VII). Each of these functional domains is composed of different combinations of homologous repeats (types I, II, and III). Type I and II repeats are formed of sequences 45 – 50 amino acids in length that contain a disulfide loop. The type III repeat is twice as long but does not contain a loop structure. The RGD sequence is located in a type III sequence in domain VI approximately one-third of the way from the carboxy terminus. Details of the fibronectin structure have been evaluated from combinations of predictions from amino acid sequence, single crystal X-ray [551], and NMR studies of isolated domains [552,553]. Fibronectin is considered to have three types of domains, two unique sheets containing type I moieties at the N terminus, 12 type II domains each with hydrophobic pockets, and 17 – 19 type III domains. The overall structure consists of 35% anti parallel sheets and no helices. There is a uniform distribution of these pleated sheets along a long chain, resulting in independent
FIGURE 18
structural domains [554]. Most of the type III domains, one of which contains the cell-binding RGD sequence, are in a flexible region of the molecule. The NMR structure [552] suggests that the abundant sheets form stacks separated by linker regions, which stabilize the molecule. Each of these domains has a distinct function [555]. The N-terminal domain seems to be required for extracellular matrix deposition [556], another domain is required for binding chondroitin sulfate [557], and the C terminus appears to be needed to stimulate its own synthesis [558]. Interactions with cells are modulated by a type III domain containing the RGD [559,560]. The linking region between these domains allows the molecule to stretch to up to seven times its native length [554]. The detailed functions of each of these domains, and the way that they interact with the cytoskeletons of connective tissue cells, have been reviewed [548]. What is apparent from each of these structure – function studies is that fibronectin is essential for matrix deposition. The fibronectin gene is located on chromosome 7 and is very complex, with up to 50 exons [561]. The chicken gene is 50 kb [562] and six RFLPs have been identified [563]. The functional domains, composed of type I, II, and III repeat sequences, are each coded for by an exon. Consequently, it would appear that fibronectin arose from the duplication of multiple genes. Nucleic acid sequence analysis indicates that approximately 90% of the entire coding
Fibronectin is composed of nonidentical subunits that are disulfide bonded at their carboxy termini. The molecule is composed of a series of repeating units (types I, II, and III) that give rise to domains with affinities for other proteins. There are several known splice variants (with or without EIIIB, EIIIA, and V; see text for description). The splice variant present in bone is not known. The cell attachment consensus sequence in a type III unit is RGD; however, other sequences that participate in cell attachment have been identified.
138 sequence is composed of these three repeats. The fibrinbinding domains are composed mainly of type I repeats, whereas the collagen domain is mainly type III repeats. The human gene promoter has been identified and it contains TATA and CCAAT boxes, is GC rich, and has an SP1 and a CRE-binding site [564,565]. Promoter analysis indicates that the CCAAT and the CRE located between164 and - 90 are essential for gene activity. However, gelshift analysis indicates that there are different complements of proteins that bind to this region depending on the tissue source [566,567]. Other studies have shown that the factor (a heterodimer of 43 kDa and a 73-kDa protein) termed ATF2, which binds to the CRE, facilitates the recruitment of factors that bind to the CCAAT element [568]. The gene is transcribed to form mRNA of 7.5 kb, but as might be anticipated, there is a great deal of heterogeneity based on differential splicing and up to 20 different mRNA species have been identified [569,570]. Within the mRNA sequences there are three regions, EIIIA, EIIIB, and V, that can be inserted or deleted depending on the tissue. An example is seen in the differences between plasma (void of E, but containing V regions) and tissue fibronectins (which contain various combinations of Es), which are the result of exon skipping. Differences between fibronectin produced by different cell types have also been found to be the result of exon subdivision (splicing within an exon). Factors that regulate differential splicing are not known, nor is the nature of the splice variant produced by bone cells. Fibronectin synthesis has been demonstrated in many osteoblastic cell culture systems. However, there is not much information on the nature of factors that regulate its production by bone cells. In adult human trabecular bone cell cultures, TGF- increases fibronectin synthesis [453]. In rat and human cells, PTH and TGF- increased fibronectin up to 11-fold. Estrogen caused a decrease in PTH-stimulated levels but had no effect on TGF--stimulated levels [571]. Gallium nitrate, currently under investigation as a therapeutic compound for increasing bone mass, also stimulates fibronectin synthesis in rat calvarial cells and ROS 17/2.8 cells [83]. The elimination of the fibronectin gene in transgenic animals (and all its multisphere variants) was lethal in utero; connective tissues did not form, indicating that fibronectin is a component essential for development of these tissues [572]. Fibronectin is capable of mediating cell attachment and spreading of rat osteosarcoma cells and normal human trabecular bone cells in vitro [504,573]. Numerous integrins mediate cell attachment to fibronectin, including the v3 vitronectin receptor. Interestingly, the attachment of normal cells to fibronectin was RGD independent, indicating that another receptor may be responsible. In fact, it was found that the osteoblastic layer in developing bone, which is
GOKHALE, BOSKEY, AND ROBEY
actively producing fibronectin, is also positive for the 4 integrin subunit [504]. This subunit can combine with 1 to form a fibronectin receptor that is RGD independent. The sequence that mediates attachment to 41 is not yet known. These results are somewhat different from what has been reported for rat osteoblastic cells, which were reported to attach to fibronectin in an RGD-dependent fashion. However, examination of data indicates that incubation with extremely high concentrations of GRGDS (200 M) decreased attachment by only 45% [574]. Fibronectin also mediates attachment of osteoclasts; however, it has been shown that osteoclasts utilize the v3 integrin [545,546, 575]. High-resolution electron microscopy studies have demonstrated that fibronectin can play a role in early biological crystal nucleation, which may be of significance in ectopic calcification, primary nucleation in calcified tissue, and bone ingrowth on ceramic implants [576]. Fibronectin has also been shown to cause apatite formation in solution [577]. 3. VITRONECTIN Vitronectin, also termed the S-protein of the complement system, is produced predominantly by the liver. It is found in serum at concentrations of 200 – 400 g/ml and in extracellular matrices [578]. In fact, it was first identified as the “serum-spreading” factor [579]. It is also found in basement membranes, but generally appears in most matrices containing the fibrillar collagens. It is detectable in developing bone by immunohistochemistry and is found in a very limited number of cells lying on the surface of newly formed bone. However, it is not clear that these cells are in fact osteoblasts. In addition, bone matrix is only faintly stained by immunological techniques, indicating accumulation in bone matrix at very low levels [504]. However, prior to mineral deposition, vitronectin is increased in concentration in the unmineralized osteoid 580, implying that it may be involved in preparing the matrix for mineralization. Although vitronectin is synthesized primarily in the liver, it is also produced by mesenchymal cells at low levels [581,582]. It may also be a biosynthetic product of osteoblastic cells [580]. The protein has a Mr 70,000, and the primary structure of human vitronectin was predicted from cDNA analysis by Oldberg et al. [583] and Jenne and Stanley [584]. An RGD sequence close to the amino terminus and a heparin-binding site in the carboxyl terminus were predicted. Ehrlich’s study [585] has defined several additional homologous domains in the mammalian vitronectin sequences obtained from different sources. From the amino to the carboxy terminus there is a “somatomedin B” domain that is rich in cysteines, followed by an RGD cell attachment site, a collagen-binding domain, a cross-linking site for transglutaminase, a plasminogen-binding site, a heparin-binding site, a PAI-binding site, and an endogenous cleavage site. Sites for sulfation and
CHAPTER 4 The Biochemistry of Bone
cAMP-dependent phosphorylation are also present. The human gene is located on chromosome 17q [586]. In vitro, vitronectin is very active in mediating attachment of all cell types. This in vitro finding may be somewhat misleading, as of all the serum proteins, vitronectin is the most active in binding to standard tissue culture plastic. Consequently, when cells are cultured in vitro, there may be a selection process whereby cells that bear the vitronectin receptor v3 have a distinct advantage over cells that do not have this receptor. Integrins other than the v3 may also bind to vitronectin [587]. Bone cells attach very strongly to vitronectin [504,588]. However, it is not clear what role this interaction plays in bone formation. Because osteoclasts have integrins that are vitronectin receptors [589], it would seem likely that the vitronectin in bone is needed for osteoclast adherence. Further, antibodies to the vitronectin receptor inhibit in vitro osteoclast action [589]. However, the vitronectin receptor distribution on the osteoclasts changes depending on whether the cells are attached or motile [590]. In addition, other RGD-containing proteins such as osteopontin or bone sialoprotein may be utilized by the osteoclast’s v3 receptor. Vitronectin has been shown to inhibit secondary nucleation of apatite crystals in vitro [591]; however, a direct effect on mineral deposition has not been established. 4. OSTEOPONTIN (BSP-1, SPP, pp66, Eta-1) This acidic glycoprotein was identified independently in several cell systems. In bone, it was termed BSP-1, now known as osteopontin [592 – 594]; however, it was also described as a secreted phosphoprotein, SPP, and pp66, a protein that is dramatically upregulated by cell transformation [595 – 597] and in association with tumor progression [598 – 599]. In addition to bone where osteopontin is largely accumulated, it is found in the kidney tubule epithelium, especially in the loop of Henle and the distal convoluted tubule [600], and it is secreted into forming urine (perhaps to inhibit crystal formation) [601 – 603]. It is also produced by mammary epithelium, as it is found at high concentrations in milk [604]. Immunolocalization and in situ hybridization have identified osteopontin in the uterus and placental membranes and in metrial gland cells [440,605]. In adult human tissue, its expression is found in epithelial cells of the gastrointestinal tract, gall bladder, pancreas, urinary and reproductive tracts, lung bronchi, mammary and salivary glands, sweat ducts, and smooth muscle [606], as well as in a variety of tumors [607]. Osteopontin is also found within neuronal cells in the brain and in the inner ear [440,608 – 610]. Platelets and megakaryocytes also contain mRNA for osteopontin, although the level of actual protein appears to be very low [611]. In general, data that have emerged in recent years indicate that osteopontin mediates autocrine – paracrine func-
139 tions in the regulation of tissue formation and plays a role in tissue homeostasis [reviewed in 612]. Inspection of osteopontin production at the cellular level during subperiosteal bone formation indicates that it is produced by osteoblasts and, to a lesser extent, by osteocytes, making it a late marker of osteoblastic differentiation and an early marker of matrix mineralization [11,12,613 – 616]. Marrow also contains cells that contain both osteopontin message and protein. In bone marrow ablation studies in which gene expression was measured as a function of new bone formation, osteopontin exhibited a biphasic pattern of expression. Levels were found to be high during the early phase of the process, most likely corresponding to the initial phase of cell proliferation [617,618]. Subsequently, levels fell to baseline, but began to climb to maximal levels just prior to or coincident with matrix mineralization [463,619]. In addition to being produced by osteoblasts, osteopontin is produced by hypertrophic chondrocytes [620 – 622] and osteoclasts [601,609,610,623,624]. Osteopontin has been localized in bone matrix at the EM level and is highly enriched at cement lines [613,625,626]. Osteopontin is also found in dentin [609]. Some studies have reported that osteopontin is not prominent in osteoid and is most prominent in cells close to the metaphyseal/diaphyseal border where there is active resorption, although the cells surrounding the osteoclasts contained the mRNA for osteopontin whereas the osteoclasts did not [627]. The molecular weight of osteopontin is highly variable depending on the method of analysis. By SDS-PAGE, the Mr varies from 44,000 to 75,000 depending on the percentage of acrylamide in the gel [245,582]. By equilibrium sedimentation analysis, it has a Mr of 44,000 (Fig. 19). Due to the nature of posttranslational modifications, it does not stain well with Coomassie brilliant blue, but becomes blue with Stains All [239,628] in agreement with its acidic pI of 5.0. The structure of osteopontin was predicted by Prince from the primary sequence of bovine osteopontin [629,630]. There is an RGD cell-binding domain, and a single poly-aspartyl repeat sequence. This poly-aspartyl sequence is highly conserved in all species, implying a functional importance for this domain. Direct analysis of the protein indicates that the bone form has an N-linked oligosaccharide, 5 – 6 O-linked chains, 12 phosphoserine residues, and 1 phosphothreonine [592]. Chick, rat, mouse, and human proteins show considerable homology, although potential phosphorylation sites vary [631]. In a posttranslational modification, osteopontin becomes cross-linked to fibronectin through the action of transglutaminase [632], which may further stabilize its deposition in bone matrix. The structure of a segment of the osteopontin molecule has been determined by NMR methods [633]. This structural study examined small peptides in the cell-binding (RGD) domain and defined the stereospecificity of the
140
GOKHALE, BOSKEY, AND ROBEY
FIGURE 19 The osteopontin molecule is composed of numerous stretches of helix (depicted as cylinders) interconnected in several cases by -pleated sheets, one of which contains the cell attachment consensus sequence (RGD). A stretch of polyaspartic acid (Poly Asp), along with phosphorylated residues (PO4), makes osteopontin a highly acidic molecule. Adapted from Denhardt and Guo, FASEB J. 7, 1475 – 1482 (1993).
serine peptide, GRGDSL. Both the RGD cell-binding domain and a RGD-free cell binding domain in the N terminus have the structures required for integrin interactions needed for cell attachment [634]. Although to date there have been no conformational studies on intact osteopontin, elegant NMR studies of the dentin phosphoprotein’s (phosphophoryn) conformation at different pHs reveal some general features of phosphorylated protein structures that might appear in osteopontin [635,636]. These NMR studies showed that dentin phosphophoryn consisted of structurally organized repeating charged clusters separated by neutral amino acids that were postulated to serve as flexible hinges. The charge clusters are those involved in calcium chelation, and most likely in interactions with the mineral. The availability of expression systems for the native protein [e.g., 637] can facilitate determination of function based on extents of phosphorylation and varying sialic acid contents. Sequence analysis demonstrates that osteopontin and the other phosphorylated sialoproteins have structural features (-pleated sheets containing anionic and phosphorylated residues) that make them well suited for interactions with hydroxyapatite [638]. The osteopontin gene has been isolated in many animal species [639 – 643] and is localized to 4q13 – 21 in humans. The gene contains seven exons and one RFLP [644], and several different alleles [645] have been reported. While the amino acid sequence is highly conserved, there are signifi-
cant differences that appear to be the result of differential splicing of certain exons in different tissues [640,644,646]. In bone, the mRNA predicts a 301 residue protein that includes a 16 residue signal peptide [630,644], whereas osteopontin from osteosarcoma appears to have an insertion due to alternative splicing [646]. The osteopontin promoter is highly complex as would be expected, given the ranges of tissues in which it is synthesized at very precise times and locations. The first kilobase of the mouse osteopontin promoter has been studied intensely. It contains a TATA box, an inverted CCAAT, and a GC box going from 3 to 5 upstream from the transcription start site. There is a positive enhancer between 543 and 253 bp and a negative element between 777 and 543 [642]. There are five PEA-3 (polyoma enhancer activator) sites, multiple TPA sites, SP1, thyroid hormone response (THR), growth hormone factor (GHF), AP4, AP5, AP1, ras activation element (RAE) sites, and a vitamin D response element (VDRE) site [647]. Transcription in bone gives rise to a 1.6-kb mRNA. Due to the correlation of osteopontin production with initial matrix mineralization [648], there have been many studies on the effect of growth factors and hormones on osteopontin synthesis [649]. In ROS 17/2.8 cells, osteopontin is stimulated by 1,25-dihydroxyvitamin D3 [650,651] and TGF- [652]. Long-term treatment with TGF- caused a decrease in expression of osteopontin, indicating a decrease
CHAPTER 4 The Biochemistry of Bone
in osteoblastic phenotype [653]. Osteopontin synthesis is also decreased by dexamethasone [654] and PTH, perhaps due to the induction of cAMP [655]. Endothelin-1, a product of endothelial cells, stimulates osteopontin in osteosarcoma cells [656]. Intermittent compressive forces also stimulate osteopontin [657]. In osteoclast cultures, it has been found that calcitonin inhibits osteopontin mRNA [658]. Phosphoproteins in general have long been linked to the mineralization process based on their accumulation at the mineralization front [106,659] and on the inability of dephosphorylated bone matrices to support mineralization in metastable calcium phosphate solutions [660,661]. The two major non collagenous matrix proteins osteopontin and bone sialoprotein, which accumulate in cement lines and in spaces among mineralized fibrils, show different densities and distribution throughout the tissue. Furthermore, the amount of these proteins generally correlates with the type of bone, speed of tissue formation, and packing density of collagen fibrils [662]. At the molecular level there is indirect evidence from cell cultures deprived of ATP that matrix protein phosphorylation is an essential step in the formation of a mineralizable matrix [406]. Osteopontin is an inhibitor of hydroxyapatite formation and growth in a variety of in vitro systems [663 – 665]. Dephosphorylated osteopontin lacks the ability to inhibit hydroxyapatite formation or growth [663,664], indicating the importance of the phosphate residues and explaining, in part, why osteopontin from different tissues with varying degrees of phosphorylation [e.g., 666] may have diverse effects on mineral formation and growth. Identification of the phosphorylated residues and protein domains required for this inhibition remain to be determined, both in solution and when the protein is bound to a matrix, thereby retaining the conformation it would have in situ. Based on the EM appearance of apatite crystals grown in the presence of 0 – 100 g/ml osteopontin, it appears that this protein blocks crystal elongation [663] rather than secondary nucleation, as is the case for dentin phosphophoryn [667]. This implied that osteopontin binds with high affinity to one or more apatite crystal faces. In fact, it has been shown [613] that osteopontin binds to hydroxyapatite with both high specificity (N 0.026 mol/m2) and high affinity (Kd 1087 g/ mol). It is of interest to note that osteopontin inhibits elastin calcification [668] and in vitro calcification mediated by vascular smooth muscle cells [669]. Thus osteopontin may be important in the prevention of atherosclerotic plaque development. In urine, an osteopontin homologue, uropontin, is similarly believed to prevent the formation and growth of kidney stones based on the in vitro effects of uropontin on oxalate crystal formation and agglomeration [601,602]. Antibodies to this protein localize osteopontin in layers on the kidney stone, suggesting binding to non calcium phosphate minerals. Although it was initially thought that osteopontin might be the phosphorylated
141 protein that was responsible for the nucleation of bone mineral, in solution, osteopontin does not appear to act as a nucleator in any of the in vitro systems studied to date [663,665]. These differences could be due to whether osteopontin is free (in vitro) or bound to the matrix (in vivo). In the latter case, it could be acting as a nucleator. In a number of connective tissue cells in culture, osteopontin appears to be an early marker of cell adherence, being visible immunohistochemically as soon as cells attach to the substratum. Osteopontin promotes osteoblastic cell attachment in vitro [504,573] and therefore may be important in determining the arrangement of cells in the matrix. It has been reported that osteopontin utilizes the vitronectin receptor, v3; however, the evidence is indirect [670]. A fragment of osteopontin also promotes osteoblastic attachment and, interestingly, it does not contain RGD [634]. The evidence for the utilization of the v3 by osteoclasts is much more direct, as attachment could be blocked with a monoclonal specific for the vitronectin receptor [575]. It is possible that this is the initial function of the phosphorylated bone sialoproteins (BSPs) and that later they play different roles by regulating (by osteopontin) and initiating (by BSP) mineralization. Likewise, osteopontin may also play a part later in recruiting osteoclasts to the mineral surface [671] and mediating their activity, as it has been reported that binding of osteopontin by osteoclasts can trigger intracellular signaling [672]. It also appears that osteoclasts dephosphorylate osteopontin, which may in turn alter its activity following bone resorption [673]. It has been shown that osteopontin inhibits the production of nitric oxide synthase (NOS) stimulated by cytokines in kidney cells [674]. This may be of interest in light of the potential role for NOS in osteoblast activity [675]. Another function is in association with osteopontin’s identity to an early T lymphocyte activation-1 (Eta-1) gene. This gene regulates resistance to different strains of viruses, and this resistance has been shown to correlate to different osteopontin alleles [645]. With respect to bone, studies from osteopontin knockout mice indicate that it is not necessary for normal bone development [676]. However, these animals have larger crystals in their bones than age-matched controls [677] and defective osteoclast function [676]. Other functions of osteopontin include enhancement of cell survival by inhibiting apoptosis [676]. This may explain why increased osteopontin expression is associated with metastatic tumor cells 678. In general, whenever ectopic mineral is formed in disease states, many of the bone matrix proteins are found because their expression has been induced by factors that are not yet well understood, or alternatively, due to their affinity for hydroxyapatite. For example, osteopontin has been found in diffuse calcification sites in association with atherosclerosis [679], in mineralized foci within certain mammary tumors [680,681], granulomas
142 of various etiologies [682], and pristane-induced calcified deposits in mice (Gokhale, unpublished data). 5. BONE SIALOPROTEIN (FORMERLY BSP-II) Another glycoprotein, somewhat more bonespecific than osteopontin, is the heavily sialylated glycoprotein, bone sialoprotein (formerly known as BSP-II). A fragment of the protein was first isolated by Andrews and co-workers [683], who initially termed it bone sialoprotein. Subsequently Fisher and co-workers isolated the intact molecule [684] and Oldberg and co-workers later determined the sequence [685]. It can comprise up to 10% of the noncollagenous protein of bone, depending on the animal species. Outside of the skeleton under normal circumstances, examples of BSP expression are very limited. BSP is present in the circulation and may derive in part from platelets [686]. It is possible that BSP is a product of megakaryocytes [687]; however, studies utilizing gray platelets (which lack granules) indicate that BSP may be adsorbed from the serum [686]. BSP is very specific to mineralized tissues and is found in dentin, cementum, and certain regions of hypertrophic chondrocytes [245,628,684, 688 – 691]. In normal healthy tissue, there may be low levels of BSP expression in mammary epithelial cells [692; Van der Pluijm and Gehron Robey, unpublished data] and it is associated with microcalcifications in breast carcinoma [693]. Within the skeleton, BSP expression is also quite limited. During subperiosteal bone formation, the fibrous periosteum and preosteoblastic layers are devoid of expression. Cells in the osteoblastic layer contain BSP, which appears just before or coincident with mineralization. By immunohistochemical staining, BSP localization is not uniform, with only groups of well-defined osteoblasts staining intensely for BSP. The pattern of localization is somewhat atypical in that the cytoplasm is not uniformly stained as seen with other bone matrix proteins. Localization is enriched in the Golgi apparatus. This unique localization was verified at the EM level by the immunogold technique in calcified and undecalcified sections. In tracing the biosynthesis through the secretory pathway, it was noted that BSP, translated and transported through the RER, is concentrated in the Golgi apparatus [353]. Subsequently, it is packaged into secretory vesicles that contain a discrete packet of material that appears to have a honeycomblike substructure. These secretory packets move to the cell surface where their contents are deposited in toto and mineralize immediately in a pericellular area that is relatively devoid of collagen fibrils. These packets (which do not appear to be enclosed within a membrane and are smaller than what has been called an extracellular matrix vesicle) appear to be the first detectable mineralized structures in new bone [353]. Despite this precise localization of BSP with initial matrix mineralization, it is not produced continuously in subse-
GOKHALE, BOSKEY, AND ROBEY
quent stages. After the initial deposition of mineral, the same cells that were previously BSP positive become devoid of BSP, despite the fact that they are identical positionally and morphologically to their previous state of development. These data suggest that the secretion of BSP is not constitutive, but regulated [353]. As shown by in situ hybridization and immunohistochemistry, BSP is also found as chondrocytes begin to hypertrophy during endochondral bone formation, suggesting another example of the osteoblast-like character of hypertrophic chondrocytes [622,689]. The region that separates the developing bone from the cartilage, the lamina limitans, and developing osteoid in this area are also intensely stained [627,689]. In adult trabecular bone, BSP is distributed in a lamellar pattern, and osteoid osteocytes and osteocytes contain the protein (Fig. 20). BSP mRNA and protein were also localized to morphologically well-defined osteoclasts sitting within Howship’s lacunae (Fig. 21). This was the first demonstration of a bone matrix protein being synthesized by both osteoblasts and osteoclasts [689]. BSP has an apparent Mr of approximately 75,000 as judged by SDS-PAGE and is composed of 50% carbohydrate (12% sialic acid, 7% glucosamine, 6% galactosamine) (Fig. 22). It is also rich in aspartic acid, glutamic acid, and glycine, and due to this unique composition, it does not stain well with Coomassie brilliant blue, but is stained by Stains All [684]. BSP, distinct from osteopontin, has two or three sets of polyglutamic acid stretches, each starting with a serine/phosphoserine, and tends to be more highly glycosylated and less phosphorylated [648]. Structure prediction [694] places the polyglutamate stretches in an helical domain, whereas the proline rich cell-binding RGD containing domain would occur at a V-shaped segment, with the arms of the V highly anionic. The low content of hydrophobic amino acids predicts an open, extended structure [694] analogous to that observed for the bovine BSP in rotary shadowed micrographs [628]. In addition to glycosylation, BSP can also be phosphorylated and sulfated [695]. The sulfate can be localized to either carbohydrate side chains or tyrosine residues [696]. From sequence homologies, a region for such tyrosine sulfation was noted to be between the postulated apatite and the RGD cell-binding sites [694]. Interestingly, in rabbits, BSP is also a proteoglycan. Rabbit bone extracts revealed a protein with keratan sulfate side chains that was identified as BSP after removal of the side chains by keratanase [697]. This keratan sulfate-BSP may also be related to a similar molecule that was identified as a constituent of medullary bone in chickens, which is elevated and decreased during the egg-laying cycle [698]. The functional significance of this posttranslational modification is not known; however, it is an example of how a bone matrix protein can take on different forms, and presumably different functional properties, depending on the animal species. There is also an RGD cell attachment domain in
143
CHAPTER 4 The Biochemistry of Bone
FIGURE 20
Bone sialoprotein immunolocalization in trabecular bone. While highly expressed within the osteoblasts in developing bone, BSP in adult bone is found primarily in the matrix between lamellae (small arrows) and is often enriched at cement or reversal lines (not shown). In addition, some, but not all, osteoid osteocytes (Ost Ocy) and osteocytes (Ocy) contain BSP. Courtesy of Dr. Paolo Bianco.
BSP, located near the carboxy terminus, which recognizes the vitronectin receptor [670] and facilitates the in vitro attachment of fibroblasts [699], osteoblastic cells [504,588], and osteoclasts [575]. A S. aureus-binding site is located
near the amino terminus [700], a factor that implicates BSP in the etiology of osteomyelitis [701]. Although NMR data on RGD peptides exist [553], detailed conformational studies of intact BSP are not available.
FIGURE 21 Bone sialoprotein immunolocalization and in situ hybridization in osteoclasts. Osteoclasts contain both BSP (A) and its mRNA (B and C, Nomarski optics), the first demonstration of the dual source of BSP in bone. It is speculated that osteoclasts utilize endogenous BSP to attach to the bone matrix in preparation for forming the sealing zone. It is not yet known if BSP mRNA is transcribed within osteoclastic nuclei or if mRNA is translated. It is possible that osteoclasts gain these moieties by fusing with osteoblastic cells. Courtesy of Dr. Paolo Bianco.
144
GOKHALE, BOSKEY, AND ROBEY
FIGURE 22 Sequence analysis predicts the presence of multiple stretches of polyglutamic acid (Poly-Glu) in the first half of the molecule and tyrosine-rich regions in the amino- and carboxy-terminal domains. In the carboxy-terminal region, many of these tyrosines are sulfated. The cell attachment consensus sequence (RGD) is flanked by such regions at the carboxy terminus of the molecule. The molecule is composed of 50% carbohydrate, including a high concentration of sialic acid residues. Glycosylation is somewhat restricted to the amino-terminal 50% of the molecule. Adapted from Fisher, McBride, Termine, and Young, J. Biol. Chem. 265, 2347 – 2351 (1990).
However, NMR analysis of a recombinant C-terminal region that contains the RGD indicated the presence of a random coil. It is likely, however, that similar to phosphophoryn, BSP chelation of calcium will involve phosphorylated clusters, and perhaps the sulfated tyrosines are unique to this form of BSP, as contrasted with the other two phosphorylated sialoproteins [317,695,696]. The human gene for bone sialoprotein was initially localized to 4q28-q32 [702 – 704], but now appears to be linked to the osteopontin gene and dentin matrix protein-1 at 4q1321 which have a similar organization and may have arisen by gene duplication (Fisher and Young, personal communication). It is approximately 15 kb in length, containing seven exons, the first six of which are small, with most of the coding sequence located in the last exon [703,704]. The signal peptide and first two amino acids of the mature protein are coded for in exon 2, and exons 3 and 4 contain regions that are rich in tyrosine and phenylalanine. The polyglutamyl acid stretches are contained in exons 5 and 6, and the RGD region is in exon 7. The splice junctions are all type 0, which means that differential splicing would leave the codon intact. Consequently, any splice variant results in an mRNA that will remain within the reading frame and not change the coding sequence. The cDNA codes for a 320 residue protein that includes a 16 residue prepeptide such that the mature protein (unglycosylated) has a predicted molecular weight of 33,600 [702].
The promoter region contains some unusual characteristics [704,705]. There is an inverted TATA and CCAAT box in close proximity to an AP1 site (148 to 142 bp), a CRE (122 to 116 bp), and a homeobox-binding site (200 to 191 bp). A retinoic acid response element (RARE) is present and overlaps with a glucocorticoid response element (GRE, 1038 to 1022 bp). A VDRE overlapping the inverted TATA has also been identified [703]. There is a polypurine (CT rich) stretch that is also found in the osteopontin promoter [643], which can possibly take on DNA triplex conformation [706]. An AC-rich region is also present, which may take on a left-handed helical configuration. This type of structure can either stimulate or inhibit transcription of the gene [reviewed in 707]. A functional YY-1 site has been identified in intron 1 [704,707]. However, the elements that convey tissue specificity to the expression of this gene have not yet been determined. Transcription of the gene results in a mRNA of 2.0 kb, although higher molecular forms have been described [702, Beresford, personal communication]. Studies on the regulation of BSP in vitro have been hampered somewhat by the fact that it is only reproducibly produced in cultures that are actively mineralizing. Because most in vitro model systems are not mineralizing, expression levels are usually low, but detectable. An exception is the cell line UMR-106-BSP, a variant that constitutively produces high levels of BSP [696]. Biosynthesis of BSP is
CHAPTER 4 The Biochemistry of Bone
tightly coordinated with the maturational stage of osteoblastic cells. Using fetal bovine bone cell cultures, BSP mRNA was increased rapidly at the stage of matrix mineralization. These data suggest that the production of BSP is regulated by autocrine factors that are produced by the cells as they go through different stages of maturation. Studies utilizing 1,25-dihydroxyvitamin D3 have shown that unlike osteopontin, BSP synthesis is decreased. A derivative of ipriflavone (metabolite III) has been reported to increase the synthesis of bone sialoprotein [260]. BSP is relatively unique to mineralizing tissues [702,709], expressed only by those cells forming a mineralized matrix [648]. These observations, combined with solution studies showing that BSP acts as a hydroxyapatite nucleator [665] and immunohistochemical data showing that BSP expression in culture is maximal during early matrix mineralization [282,648], suggest that BSP is involved in bone mineralization [317,710,711]. The solution studies of the effect of BSP on apatite formation suggest which structural features are important in this process. When the effect of BSP on mineralization is monitored in an agar gel or at constant pH in solution, it facilitates hydroxyapatite deposition [665,712], although BSP can also block seed growth [713]. Blocking the carboxylic groups, presumably those in the polyglutamyl domains, destroys the nucleation abilities of BSP, while dephosphorylating the molecule has less of an effect. This suggests that apatite – BSP interactions occur predominantly through the polyglutamyl repeats; however, other portions of the molecule are also involved [713]. Although solution data do not prove that BSP has this same function in situ, it does demonstrate the nature of the interaction between BSP and hydroxyapatite. Effects of the variable phosphorylation of BSP remain to be determined. Numerous studies have shown that BSP is active in mediating cell attachment of a number of different cell types that is dependent on RGD [504,573,575,588,589,714]. In studies where it was noted that endogenously produced enzymes could cleave BSP into more or less defined fragments, it was noted that there is a cleavage site just upstream from the RGD that results in a large fragment that lacks RGD and a small fragment that contains RGD at the amino terminus. Surprisingly, the large fragment, which lacks RGD, was able to support cell attachment, whereas the small fragment, which contains the RGD, was relatively inactive [588]. These data suggest that there is a sequence upstream from the RGD that is capable of supporting the attachment and spreading of bone cells, whereas RGD without this upstream region is not active. Although it was initially thought that these data suggested binding of BSP to an integrin other than the v3, current data suggest that the fragments associate with different regions within the same binding site on the integrin, albeit with different affinity and efficacy 713. Because the RGD region is clustered with tyrosine phosphorylation consensus sequences (tyrosyl
145 residues preceded by an acidic residue and a turn), it was thought that the degree of sulfation may also influence cell attachment. However, unsulfated BSP (prepared from UMR-106-BSP cells treated with the sulfation inhibitor chlorate) was able to support attachment and spreading [715]. It should be noted, however, that cell attachment is an in vitro assay that may not totally reflect the subtleties of a process that occurs in vivo. Although data are not definitive, it appears that osteoblastic cells attach to BSP via the v3 integrin (vitronectin) receptor. Data that osteoclasts use this receptor are much more definitive and several studies have demonstrated that the attachment of osteoclasts can be blocked with antibodies against it [575,589]. In addition, changes in intracellular Ca2 and pH levels in osteoclasts have been shown to be altered, indicating that the binding of BSP (and osteopontin as well) may initiate a signaling pathway and ultimately cause a change in cellular metabolism. It is not yet known what stage of osteoclast formation and/or activation these two molecules are influencing, as there is some controversy as to whether they mediate formation of the sealing zone, which is required for the formation of the extracellular lysosome-like compartment. Other potential functions include mediating interactions between cells and collagen fibrils [716]. Early studies that utilized the injection of isotopic sulfate into animals along with autoradiography and biochemical isolation techniques described a population of sulfated molecules that moved rapidly from the osteoblastic cells to the mineralization front [233,234]. In addition, microprobe analysis indicated that newly formed hydroxyapatite was in close proximity to sulfate [717]. Because proteoglycans account for virtually all of the free sulfate that is incorporated into macromolecules, it was thought that this population of sulfated macromolecules was proteoglycan(s) [233,234]. However, new data that BSP can also be sulfated may provide strong evidence for the role of BSP in matrix mineralization. There are a few suggestions that abnormal BSP metabolism may result in abnormal bone formation and/or disease. In one case, it was demonstrated that BSP was lacking in a cow with an osteopetrosis-like syndrome (Weintroub, personal communication). Given the potential role of BSP in osteoclastic resorption, this may be an indication that in some cases of osteopetrosis, the matrix required for optimal osteoclastic activity is lacking. It has been demonstrated that the production of BSP by certain primary cancers (breast, prostate, and thryoid, to name a few) coincides with the invasion of the tumor into the surrounding tissue and the appearance of mineralized foci within the primary tumor. Furthermore, the levels of BSP in the primary tumors are prognostic of long-term outcome [718,719]. Studies with carcinoma cell lines have shown that metastatic cells do in fact preferentially attach to BSP [720], and it has been
146 demonstrated that cell-bound BSP binds to factor H, a mediator in the alternate complement pathway, and may be involved in the escape of tumor cells from immunosurveillance [721] BSP knockout mice [722], which have a totally nonfunctional BSP gene, were reported to be indistinguishable from wild-type mice at birth, 8.5 days, and 1 month. However, at 1 year they were 25% smaller than the wild type. Histologically, the predominant observation in long bones and in the calvaria was a decrease in marrow space. Incisors were also elongated. X-ray diffraction of the homogenized bones of the knockout animals revealed no differences in mineral crystal relative to controls (Aubin, personal communication). Detailed analyses of spatial changes in mineral properties have not yet been reported. 6. BONE ACIDIC GLYCOPROTEIN-75 (BAG-75) Another sialoprotein originally isolated from rat bone has an apparent Mr of 75,000 and hence is called bone acidic glycoprotein-75 (BAG-75) [723 – 725]. This protein is heavily glycosylated and contains 7% sialic acid and 8% phosphate. Thirty percent of the residues are acidic in nature. It is found only in bone, dentin, and growth plate cartilage, but in culture, cells from soft connective tissues have also been found to synthesize low levels of this protein [724]. It is not known what percentage of this protein exists in bone. BAG-75 shares sequence homology with the dentin phosphoprotein, phosphophoryn, osteopontin, BSP, and dentin matrix protein and therefore may have similar structural features [723,724,726]. The structures of phosphophoryn fragments, predicted from NMR data, were discussed earlier with osteopontin. The cDNA and the gene have not yet been cloned for this molecule. However, some data are available from direct amino acid sequencing. The amino terminus is about 30% homologous with osteopontin. In addition, it does contain polyacid stretches as do osteopontin and bone sialoprotein [723,727]. However, BAG-75 contains both polyaspartate and polyglutamate domains, as well as several phosphorylation sites and an RGD cell binding site [638]. The protein binds with high affinity to both hydroxyapatite and Ca2 ions [723], as well as to collagen [727]. Immunolocalized next to cells in bone and concentrated in newly formed osteoid, this protein may combine the properties of osteopontin (a mineralization inhibitor) and bone sialoprotein (a nucleator). Preliminary studies with its homologue dentin matrix protein demonstrate that it may have both functions (Boskey, unpublished data). BAG-75 also inhibits the resorptive activity of osteoclasts, presumably by blocking access to bone mineral [728]. Related to BAG-75 is dentin matrix protein-1 [729,730] originally isolated from teeth, but now shown to be expressed by bone marrow stromal cells (Fisher, personal
GOKHALE, BOSKEY, AND ROBEY
communication). The nonphosphorylated recombinant DMP-1 [731] is a potent apatite nucleator in solution (Boskey, unpublished data). 7. MICROFIBRILLAR PROTEINS As in many connective tissues, bone matrix also contains microfibrillar structures [732]. Although the protein constituents of the microfibrills have not been completely identified, by analogy to other tissues they are likely composed of a family of proteins that include type VI collagen, fibrillins, fibulins, latent TGF-binding proteins (LTBPs), the MAGP proteins, and Big-h3 [733]. Fibrillins and LTGP are cysteine-rich glycoproteins that are related by their composition of repeating EGF motifs. Fibrillin-1 and 2 are similar but not identical components of the microfibrillar aggregates in the skeleton, as well as in eyes and vascular tissues [734,735]. The structure of fibrillin-1 was predicted by Sakai’s group [736,737] to consist of eight repeated EGF-like motifs, an RGD cellbinding region, several cysteine motifs, and a cysteine-poor COOH terminus. The structure as deduced from primary sequence is thought to be stabilized by calcium-binding EGF domains [736,738]. The proteins associate with glycoproteins, lysyloxidase, a 58-kDa microfibril associated protein, and other protein species forming beaded domains [739]. Fibrillin mutations have been found to cause Marfan’s syndrome, a disease characterized by skeletal, ocular and cardiovascular abnormalities [740]. The association of fibrillins with microfilaments and elastin implies an anchoring function. Fibrillins only form microfibrils in the presence of Ca2; the recently reported mutations in Marfan’s patients who lacked this Ca2-binding domain [738] speak to the importance of these microfibrillar structures. How the microfibrils regulate the growth of long bones is not known; however, immunolabeling in the chick embryo demonstrated labeling in the primary axis and the ventral surface of the notochord [741], suggesting a role in early development. Related to the fibrillins are the latent TGF--binding proteins (LTBPs) and, as their name implies, their proposed function includes binding to latent TGF-. However, they also appear to be microfibrillar proteins based on immunohistochemical analysis colocalizing LTPB-1 with fibrillin at light microscopic and EM levels. These studies show that in addition to regulating TGF-1 activity, LTBP1 may function as a structural component of connective tissue microfibrils. LTBP1 may therefore be a candidate gene for Marfan-related connective tissue disorders in which linkage to fibrillins has been excluded [742]. 8. OTHER CELL-BINDING PROTEINS As microanalytical techniques become more refined, both at the protein level by immunohistochemistry, radioimmunoassays, and enzyme-linked immunosorbent
147
CHAPTER 4 The Biochemistry of Bone
assays (ELISA) and at the nucleic acid level by reverse transcriptase polymerase chain reaction (RT-PCR), it is certain that the list of glycoproteins, including RGD-containing proteins, will continue to grow. These proteins, while not as abundant as those described earlier, most likely make significant contributions to the development of bone. As transgenic animals are generated that either lack or contain a mutated form of these molecules, and/or as chemical amounts of purified proteins are generated through the use of recombinant techniques for use in cell culture and other model systems, more information will become available on the precise roles of all the glycoproteins in bone development. Tenascin (neuronectin), originally identified in muscle, is also made in bone cells [743,744], bone marrow cells [745], and cartilage cells [746]. Tenascin has a hexameric quaternary structure, with six arms attached to a central globular domain [747]. It has been found to be associated with mesenchyme, which is differentiating into bone or cartilage [748,749], but was absent from mature tissue. This indicates that tenascin may be important for early matrix development, where it acts as a modulator of cell – extracellular matrix interactions [747]. A 36-kDa cartilage matrix protein with an RGD sequence [750], this protein has the highest affinity for fibroblasts of all the bone matrix proteins tested to date [751]. Based on protein and DNA sequencing, the protein consists of a 337 amino acid peptide chain, with 10 leucine-rich repeats surrounded by disulfide-bonded loops at each end. Based on its strong cell-binding ability and its distribution, it is most likely that this protein is involved in regulating cell adherence in a variety of connective tissues. Another proteoglycan, osteoadherin, a member of the leucin-rich repeat family, has been purified from bovine bone [752]. Biochemical analysis has shown that it has a molecular mass of 85 da and contains RGD sequences and keratan sulfate chains. It is synthesized primarily by osteoblasts and can bind to hydroxyapatite with efficiency similar to that of fibronectin. This binding appears to be mediated by integrin V3. It has been demonstrated that depending on the species, osteoadherin and osteomodulin are the same protein/gene [753]. A 90 Da protein, periostin (previously called OSF-2), is expressed preferentially in periosteum and periodontal ligaments 754, suggesting a potential role in bone and tooth formation as well as in maintenance of structure. It is secreted by osteoblasts and osteoblast-like cell lines and exists as several isoforms. Distinct from the other cell-binding proteins discussed here, it does not contain a RGD sequence (L. Bonewald, personal communication). Biochemical analysis has shown that glycosylation is not a major component of this protein. In vitro functional studies have shown that the antiperiostin antibody inhibits attachment and spreading of MC3T3-E1 cells, indicating that periostin
is involved in cell adhesion. In addition, TGF- increased the expression of periostin in primary osteoblastic cells significantly. These results indicated that periostin may play a role in recruitment and attachment of osteoblast precursors in the periosteum.
V. GLA-CONTAINING PROTEINS Bone contains a number of proteins that are modified posttranslationally by vitamin K-dependent enzymes to form the amino acid -carboxy glutamic acid (gla). Due to the sequence requirements of the carboxylating enzymes, the gla proteins of bone share some sequence homology with certain blood coagulation factors that require -carboxylation to maintain their activity.
A. Osteocalcin Osteocalcin was the first bone matrix protein to be isolated by the use of nondegradative techniques from acid demineralized bone [755,756]. It comprises up to 15% of the noncollagenous protein, although the level varies depending on the animal species [410] and accounts for up to 80% of the total gla content of mature bone [757]. It was initially reported to be virtually exclusive to bone and was considered the only bone-specific protein. Consequently, a great deal of effort has gone into trying to determine when and where osteocalcin is synthesized and how it functions in bone metabolism. Extensive screening of protein and RNA extracts [758,759] and tissue sections by immunohistochemistry [613,760 – 762] from virtually all tissues has failed to detect osteocalcin in any tissue other than dentin and bone. The one exception was marrow, which led to the discovery that megakaryocytes and platelets contain mRNA for osteocalcin (along with the mRNAs for many of the other bone matrix proteins). It is still not clear how much osteocalcin protein is actually synthesized by platelets and megakaryocytes or what its function is in these cells [611,763]. During bone development, osteocalcin production is very low and does not reach maximal levels until late stages [764 – 767]. By immunohistochemistry, mineralization fronts stain intensely for osteocalcin, but it has been difficult to demonstrate osteocalcin in osteoid and in cells. However, using an antibody against the precursor form of osteocalcin, the primary cell type that stains in developing human subperiosteal bone is the osteocyte. This antibody intensely stained the cell processes in canaliculae and suggests that perhaps osteocalcin bypasses the osteoid layer by being secreted directly at the mineralization front through the osteocytic cell processes [768] (Fig. 23). However, it should be noted that in other animal species, particularly in
148
GOKHALE, BOSKEY, AND ROBEY
FIGURE 23
Osteocalcin immunolocalization in developing bone. Localization of osteocalcin using an antibody against the mature secreted form of the protein sharply demarcates the mineralization front (MF) in developing bone (A). Note, however, the lack of localization within cells that should be synthesizing this molecule. However, when utilizing an antibody raised against the precursor peptide (which is not maintained within mineralized matrix), it can be seen that osteoid osteocytes and osteocytes contain high levels of the proform of the molecule (B and C). Courtesy of Dr. Paolo Bianco.
rat, endosteal osteoblasts are positive for osteocalcin [758,760,761]. The protein has a molecular weight of 5300 but migrates with an apparent Mr of 14,000 Da on SDS-PAGE. The entire sequence has been determined by direct protein sequencing [769,770]. Depending on the animal species, there is one intramolecular disulfide bond and three to five residues of -carboxy glutamic acid [756] (Fig. 24). The original structural predictions by Hauschka and Carr [771] based on circular dichroism suggested that osteocalcin had a structure with an extensive (40%) helix in the presence of calcium ions. As detailed elsewhere [772], the predicted structure of osteocalcin in the presence of Ca2 consists of two antiparallel -helical domains, one containing -carboxyglutamic acid residues and one rich in acidic amino acids. Both these domains were proposed as sites for calcium chelation. The -carboxyglutamic acids were calculated to be 0.5 nm apart, corresponding to the 0.55-nm interatomic spacing of Ca2 ions in the 001 plane of the apatite lattice, suggesting that this domain might be involved in binding to the mineral. A -pleated sheet in the C terminus was suggested as a cell-binding site. More recent insight into the osteocalcin structure comes from comparisons of NMR data for Ca2 and Pb2 salts [773] and Ca2 and Lu3 (lutecium) salts [774]. These NMR studies show that Pb and Lu3 compete for Ca2 binding sites. Because Pb2 blocks the binding of osteocalcin to hydroxyapatite in
FIGURE 24 This small molecule contains two stretches of helix (depicted as cylinders) and two regions of -pleated sheet (arrows). The carboxylated residues of glutamic acid in the amino-terminal helix orient the carboxy groups to the exterior, thereby conferring calcium ion binding with relatively high affinity. There is one intramolecular disulfide bridge (C – C) in the middle region of the molecule. Adapted from Hauschka and Carr, Biochemistry 21, 258 – 272 (1985).
CHAPTER 4 The Biochemistry of Bone
solution, such data imply that the osteocalcin – apatite interaction occurs through the same domain as Ca2 chelation in solution. Comparison of Lu3 and Ca2 data for the dog apoprotein demonstrate, the presence of two high-affinitybinding sites for Ca2 and the conformational changes that occur when Ca2 is present. The human gene is localized on chromosome 1 [775] and the structure has been determined [776]. The gene is 1.2 kb with four exons that predict a protein of 125 amino acids. The signal peptide contains 26 amino acids in exon 1, a propeptide of 49 amino acids in exon 2 along with the -carboxylation recognition sequence, two stretches that become -carboxylated in exon 3, and the remainder of the molecule and untranslated region in exon 4 [777,778]. Interestingly, the mouse genome contains three osteocalcin genes, two of which are activated in bone and one activated in the kidney [754]. Although some of the basic elements have been determined in the human promoter, most of the extensive characterization has been done primarily in rodent promoters. It contains a TATA box and a CCAAT box. In addition there is one NF1-binding site and one AP2-binding site, a viral core enhancer, and a CRE. There is a VDRE at 463 to 437 bp [779] that is flanked by other nuclear-binding sites [780 – 783]. A GRE is found in close proximity to the TATA (16 to 1 bp) [784]. A vitamin A response element has also been identified [785]. Apparent AP1 binding sites are also located in close proximity to, or overlapping with, the VDRE and the CAAT box [786]. Because of the highly specific nature of osteocalcin expression, the promoter has been scrutinized intensely to determine what properties convey tissue specificity. This has led to the characterization of the “osteocalcin box” [787,788], located between 99 and 76 bp, which is functionally active [789,790] and contains a binding site for Msx-1 or Msx-2 (homeodomain proteins). Msx-related factors have been found to be synthesized by osteoblastic cells [791]. The first intron has also been characterized and found to have a potential silencer [792,793]. Further characterization of this promoter led to the identification of a binding site, OSE2, located between bp 146 and 132, which binds the transcription factor, cbfa1, the so-called osteogenic “master gene”[794]. The effects of growth factors and hormones on osteocalcin vary somewhat, which most likely reflects differences in the culture systems under investigation (stage of maturity, length of time in culture, etc.). Osteocalcin synthesis is upregulated by 1,25-dihydroxyvitamin D3 [781 – 783,788, 795 – 797] and by 22-oxacalcitriol [798]. In general, most factors decrease osteocalcin expression, such as PTH [799], glucocorticoids [800,801], TGF- [802,803], PGE2 [799], IL-1 [804,805], TNF- [804], IL-10 [90], and lead [91]. Mechanical loading has also been reported to have a negative effect [806].
149 The proposed functions for osteocalcin have been reviewed extensively [807,808]. Because osteocalcin has a high and relatively specific affinity for apatite, probably due to the binding of the gla domain to the 100 (a axis) face of the apatite crystal, the protein has been proposed as a specific regulator of the length of the mineral crystals in bone. Osteocalcin is not expressed in culture until mineralization starts [808,809], which fits the model that it is a regulator of the size and habit of the mineral crystals rather than a promoter of mineral crystal formation. In solution, osteocalcin is an effective crystal growth inhibitor [429,810], whereas -carboxyglutamate-free osteocalcin has no such effect, which is in good agreement with the proposed function. Similarly, during new bone formation, osteocalcin staining and expression occur after mineralization starts [811,812], demonstrating its function in later stages of bone formation and remodeling. Osteocalcin is made by osteoblasts/osteocytes and is considered to be a marker of osteoblastic function [813,814] as well as a coupler of osteoblast – osteoclast action. It also appears to be important for induction of the osteoclast phenotype [815]. Studies probing the function of osteocalcin have shown that osteocalcin-deficient bone particles are not resorbed readily [816], in good agreement with the report that osteocalcin is involved in osteoclast recruitment [817,818]. This concept is also supported by the defective osteocalcin production noted in some humans and animals with osteopetrosis, a severely deforming disease characterized by the failure to remodel bone and calcified cartilage [819 – 821]. Osteocalcin may have other functions. The osteoclastrecruiting function has been questioned based on earlier observations that animals with osteocalcin-deficient bones due to warfarin treatment fail to show significant abnormalities in bone growth and fracture healing [755]. These animals did show excessive growth plate mineralization [822], which could be interpreted as a failure to remodel calcified cartilage. Whether this remodeling effect was due to the failure of osteocalcin to bind to the mineral in calcified cartilage or to a defect in the actions of matrix-gla protein (see earlier discussion), which is produced in the cartilage, is not known. As reviewed here, sufficient data link osteocalcin to a regulatory role in bone mineral maturation. It is likely, however, that several proteins have a similar function, thus it will be difficult to prove this function in vivo. Lambs that lack -carboxylated osteocalcin show decrease in both new bone formation and bone resorption [823], supporting a role for osteocalcin in remodeling. Mice with a null allele (both bone genes deleted) appear normal at birth and at 4 weeks, but by 24 weeks, they have increased bone mass. Preliminary analysis suggested a role for osteocalcin in regulating bone growth in vivo 794. Mineral crystals in the bones of osteocalcin-null animals fail to mature [824], supporting its role in bone remodeling.
150
GOKHALE, BOSKEY, AND ROBEY
B. Matrix Gla Protein Studies that measured the content of gla residues in bone during development indicated that gla-containing proteins were present at earlier stages of development than could be accounted for by the expression of osteocalcin, the major gla-containing protein in bone. Consequently, the presence of matrix gla protein, MGP, was predicted prior to its actual isolation and purification. It was first isolated from bone due to its copurification with BMP [825,826]. While there is little information on the developmental expression of MGP throughout the body, it is known that it is expressed in lung, kidney, heart, cartilage [827], and smooth muscle cells [828]. MGP is more abundant in cartilage than in bone [829]. In the skeleton, MGP expression appears early and remains at the same level at all subsequent stages of development [765]. However, close examination of early developmental stages indicates that MGP expression is much more limited, found only in lung bud and at the limb bud mesenchyme – epithelial interface [830]. MGP has a Mr 15,000 although it migrates as a substantially larger molecule on SDS-PAGE. The secreted form contains five residues of gla and one disulfide bridge in a 77 to 79 amino acid residue protein. A distinct physical property of matrix gla protein is its insolubility in physiologic solutions (10 g/ml) and its tendency to selfassociate via hydrophobic interactions. It also appears that a propeptide present at the carboxy terminus is removed to form the mature protein [831]. Matrix gla proteins from five different species have been shown to have phosphorylated serine residues [832,833]. Thus the protein is a phosphorylated gla protein. Due to its insolubility, along with difficulties in isolating a purified protein, it has been difficult to predict the potential structure of MGP, and detailed conformational studies are not yet available. However, cDNA sequences for several species allowed the primary structure to be predicted [832,833]. The MGP gene has been localized to 12p [834]. The gene is approximately 3.9 kb long and contains four exons. The signal peptide is coded for by exon 1 and an helical region by exon 2. The recognition sequence for the carboxylating enzymes is found in exon 3, and a sequence that actually becomes -carboxylated is in exon 4. There are a series of AluI repeats in the 3 -untranslated region of the gene. The promoter has been characterized and found to contain a TATA box and a CAAT box, along with a perfect palindromic sequence that is similar to a RARE [835]. Little is known about the factors that regulate the synthesis of MGP. In ROS 17/2.8 cells, 1,25-dihydroxyvitamin D3 was found to stimulate its synthesis, although higher levels were needed than for the induction of osteocalcin. It is also apparent that retinoic acid induces MGP expression greatly, which may be highly significant for cartilage development [835].
The matrix gla protein (along with osteocalcin) was initially suggested to be important for the process of endochondral ossification because warfarin-treated rats showed premature epiphyseal closure [822], indicative of impaired remodeling of calcified cartilage. At the time of the report, the matrix gla protein had not been identified, but because cartilage chondrocytes were not thought to produce osteocalcin, a second gla protein was postulated. When the matrix gla protein was identified as a chondrocyte product [827], warfarin data were reinterpreted to consider a potential role for the matrix gla protein. The findings that retinoic acid, a known teratogen, stimulated matrix gla protein gene expression, coupled with the observation that matrix gla protein copurifies with the bone morphogenetic proteins, led Price to suggest that the matrix gla protein may mediate effects of retinoic acid during cellular differentiation, perhaps by serving as a natural carrier for BMP. A recent discovery indicates that the matrix gla protein can be found in a variety of soft tissues and that its mRNA is abundant in lung tissue, although the gla protein content was low [836]. This suggests that the protein has a more general secretory function. Mice that have their MGP gene deleted die prematurely because of massive tracheal cartilage and blood vessels calcification [828]. The endochondral cartilage in these animals is also excessively mineralized, but trabecular and cortical bones appear comparable in mineral properties to age-matched controls [837]. This is convincing evidence that MGP is an in vivo inhibitor of mineralization. This has further been shown in cell culture studies in which ablation of the MGP to sternal chondrocyte cultures resulted in dystrophic mineralization, whereas the addition of exogenous MGP prevented calcification under conditions in which mineralization is normally observed [838]. Additional insights into MGP function may come from reports that in breast cancer cells there is an increase in the expression of MGP [839]. Interestingly, in prostate cells undergoing apoptosis, there is also an increase in MGP [840]. In light of data in knockout animals and in cell culture, it seems likely that expression of this protein may be a protective action by the cell against unwanted calcification. Expression of MGP in cells from non-onion fractures, but not healing fractures, indicates its importance in bone healing [841].
C. Protein S This gla-containing protein is synthesized primarily in the liver, but has been isolated from bone matrix [842]. Although synthesis was demonstrated in osteosarcoma cells [768,842], it has not been reported to be synthesized endogenously in bone. However, it should be noted that patients with protein S deficiency suffer from osteopenia, and
151
CHAPTER 4 The Biochemistry of Bone
consequently this protein, irrespective of its origin, most likely plays a role in bone metabolism.
VI. OTHER PROTEINS A. Proteolipids Proteolipids, as a general class of macromolecules, are membrane components, consisting of a hydrophobic protein component and covalently bound lipid [843,844]. Proteolipids have been isolated from a variety of connective tissues, including bone [845 – 848] and calcified cartilage [849,850], where there are cell and matrix vesicle membrane components [851]. Based on analyses of the apoprotein amino acid compositions, it is clear that there may be more than one type of proteolipid component in bone. Although the primary structures of several of the bone and/or cartilage proteolipid apoproteins have been determined [852 – 854], their detailed structure is not known. Although an oral bacterial proteolipid associated with calcification has been sequenced [855], structures of the bone and cartilage proteolipids have not yet been described. However, the structures of several other proteolipids that appear to have common features have been described in detail. The protein conformations of a variety of proteolipid proteins have been determined using NMR [856], fluorescent labeling [857,858], electron microscopy and quasi-elastic light scattering [859], and Fourier transform infrared spectroscopy [860]. In general, these transmembrane proteins have hydrophobic domains throughout the molecule, including N and C termini. In many cases, the transmembrane domain helices are highly ordered, although they have abundant hydrophobic residues (e.g., polyvaline). These hydrophobic domains facilitate interactions with the lipids in the membranes in which the proteolipids are contained. In contrast, the N- and C-termini are generally flexibly disordered and contain covalently linked lipids, such as palmitoyl-cysteine or acidic phospholipids. Bone and related cartilage proteolipids have several functions. Based on studies with the bacterial analogue of the chondrocyte proteolipid [848,861], a role in proton transport has been postulated [862,863]. This activity is quite distinct from that of another chondrocyte proteolipid, which is a phosphate ion transporter, taking phosphate ions into and out of the cell [864]. They have also been demonstrated to act as hydroxyapatite nucleators both in vitro [845,847,850], in a gelatin gel [865], and when implanted in a millipore chamber in vivo [866]. The calcifiable proteolipids are associated [867,868] with a complex consisting of decreasing molar amounts of calcium, the acidic phospholipids, and inorganic phosphate [869]. These complexes are capable of acting as hydroxyapatite nucleators in the presence, as well as in the absence, of the proteolipid
apoprotein [867,870,871]. Such complexes were also shown to be components of the membranes of extracellular matrix vesicles [872] and to be able to act as ion transporters [873]. Proteolipids isolated from cartilage matrix vesicles have been shown to include the annexins [874]. Annexins are Ca2 and phospholipid-binding proteins [875] previously known as lipocortin, calpactin, endonexin, chromobindin, anchorin, and so on [876]. Annexins have also been shown to have ion transporter roles [877], but other functions are also known. Lipocortin I is a phospholipase A2 inhibitor [878], and anchorin is a collagen-and cytoskeletal-binding protein [879]. The annexins are synthesized by osteoblasts [880] as well as by chondrocytes [855,881] and are abundant in matrix vesicle membranes where they are involved in the initiation of calcification [882]. Annexins share a 17 amino acid residue homology, which is probably important for the Ca2-dependent phospholipid binding [879]. Wu et al. [851] have reported that the “nucleational core complex,” i.e., the essential structure needed to cause hydroxyapatite formation in matrix vesicles, consists of annexins (i.e., proteolipid), Ca2, Pi2, and phosphatidylserine. This is in good agreement with earlier reports indicating that cartilage proteolipids and their associated complexed acidic phospholipids were hydroxyapatite nucleators. Because proteolipids are concentrated in matrix vesicle membranes [872] and because most are involved in ion transport, these proteolipids seem to be of general importance in accumulating ions within the cell and/or extracellular matrix vesicles. As ions accumulate within vesicles, in the presence of proteolipids and phosphatidylserine, mineral crystals form associated with the membranes. Thus, the two functions of the proteolipids may be closely related.
B. Enzymes and Inhibitors It is apparent from the slow destruction of bone matrix proteins that occurs with increasing age [883,884] that enzymes are present within the mineralized matrix. The origin of these enzymes can be from the circulation due to their affinity for hydroxyapatite or from neighboring cells that are highly associated in bone such as endothelial cells, hematopoietic cells, and cartilage cells. In addition, cells in the osteoblastic lineage also synthesize a number of enzymes and their inhibitors. A source of enzymatic activity that has often been overlooked is the osteoclasts themselves, which secrete lysosomal types of enzymes into the extracellular compartment formed by the sealing zone. 1. METALLOPROTEINASES The matrix metalloproteinases (MMPs) [reviewed in 885] are a family of enzymes that can be roughly divided into collagenases, which produce the initial cleavage of
152 native, triple helical collagen; gelatinases, which further degrade the structurally compromised collagen, and stromelysins, which degrade proteoglycans. However, there is some degree of overlap among these groups. It has long been known that connective tissue cells that are synthesizing extracellular matrix proteins also have the capacity to synthesize proteins that destroy the very same proteins. Osteoblastic cells are no exception, and it has been reported that they synthesize collagenase-1 (MMP-1) [886], collagenase-3 (MMP-13) [887], gelatinase A (MMP-2) [888] and gelatinase B (MMP-9), [889 – 892], and MT1-MMP (vide infra). Some osteoblastic cells have been found to bear a cell surface receptor for collagenase. In addition, matrix vesicles contain MMPs [893] along with a stromelysin-like activity that has the ability to degrade proteoglycans [894,895]. Osteoclasts also have been reported to contain collagenases [896,897]. To date, a number of transgenic animals have been generated that are deficient in an MMP, but generally did not display a skeletal defect. The one exception is the recently reported MT1-MMP knockout mouse, which while normal at birth quickly develops a severe skeletal phenotype characterized by dwarfism, osteopenia, and arthritis [898]. Tissue inhibitors (TIMPs, forms 1 and 2) inhibit the activity of metalloproteinases [899]. However, it is not clear to date at which point this activity is produced. Depending on the cell culture system, it is apparent that there is a great deal of variability in the ability to produce these enzymes and inhibitors. In those cell culture systems producing MMPs, it has been found that MMP activity is stimulated by PTH [891,900], TNF- [901], and retinoic acid [902]. The crystal structures of two of the matrix metalloproteases [903] and one form of collagenase [904] have been resolved. These are both members of the stromelysin III family (to date, 11 such enzymes have been identified). Structural features include a histidine-rich -helical catalytic domain that chelates the metal ion required for activity [905], a C-terminal domain that is involved in linking the enzyme to the membrane, and pre- and pro-N-terminal domains that are thought to be responsible for maintaining the enzyme in inactive form via interaction with TIMP [906]. The catalytic domain is a spherical region with a methionine-based turn containing Zn2 and Ca2-binding sites, with proline-86, phenylalanine-79, and aspartate-232 forming the base of the active site residues [904]. The crystal structure of elastase [906], the enzyme responsible for the degradation of elastin, has similar features. The crystal structure of the active site of collagenase [905,907] shows the presence of a five-stranded -pleated sheet, three helices, and binding sites for Ca2 and Zn2, but here the zinc is chelated only by histidines. Collagenase cleaves at a unique site in the collagen triple helix [908] and at a minor site in the nonhelical N-terminal
GOKHALE, BOSKEY, AND ROBEY
region. Activated by tyrosine kinase-dependent phosphorylation, collagenase-mediated turnover of the bone matrix seems essential during growth and repair, but not during early development. Thus, homozygous transgenic mice whose type I collagen does not contain the unique cleavage site appear normal at birth, but develop thickened skin, uteri, and bone during growth and have impaired fracture healing [909]. In bone, osteoclasts produce several cysteine proteases as well as metalloproteases. It is believed that during osteoclastic resorption the cysteine proteases that have acidic optimal pHs act first [910,911]. Then, as the mineral is dissolved in the acidic environment and the acidity is neutralized, the matrix metalloproteases function. Specific inhibitors of the cysteine proteases have been used effectively to inhibit osteoclastic resorption [912]. Such inhibitors, while blocking the actions of the cysteine proteases, increase the activities of some lysosomal enzymes [913]. As reviewed in detail elsewhere [911], metalloproteases degrade proteoglycans, collagens, and matrix proteins [914 – 916]. One of the most important of the cysteine protease degradative enzymes in bone may be cathepsin K, as this enzyme is expressed mainly by osteoclasts and appears to initiate the bone degradation process. In this light, cathepsin K knockout animals have osteopetrosis associated with abnormal matrix degradation but normal mineral resorption [917,918]. 2. PLASMINOGEN ACTIVATOR (PA) AND PLASMINOGEN ACTIVATOR INHIBITOR (PAI) Another enzyme system that is represented in bone matrix is plasminogen activator (both urinary and tissue types, uPA and tPA) and PAI-1 [919, reviewed in 920]. PA and PAI have been identified in both transformed and nontransformed cell culture systems [921], (Kopp and Gehron Robey, unpublished results). However, it is also possible that enzymic activity detected in bone matrix can also originate from other neighboring cell types. The proposed functions of these proteins are varied. They range from the controlled activation of other enzymes, such as matrix metalloproteinases, to the activation of growth factors, such as TGF- and the IGFs that are bound and modulated by the IGF-binding proteins. The synthesis of enzymes and inhibitors is regulated by a number of growth factors and hormones, and in general they are modulated in different directions; i.e., if the enzyme goes up the inhibitor goes down, and vice versa. PA is upregulated by PTH [921 – 924], 1,25-dihydroxyvitamin D3 [925], and acidic and basic FGF, EGF, and PDGF [926]. PAI is concurrently decreased by these factors. Activity is either upregulated or downregulated by TGF-, depending on the cell type [927], although there is generally an increase in PAI-1 [926,928]. PA activity is decreased and PAI activity is increased by glucocorticoids [929].
153
CHAPTER 4 The Biochemistry of Bone
3. MATRIX PHOSPHOPROTEIN KINASES A third category of enzymes that appear to be critical for the formation of mineralized connective tissues are the extracellular matrix phosphoprotein kinases. Matrix protein kinases isolated from bone and dentin [930 – 934] are casein II kinases, whose activities can be inhibited by heparin and 2,3-diphosphoglycerate. Analogous to some tyrosine kinases found in the extracellular matrix [935], these phosphoprotein kinases are responsible for the extracellular addition of phosphate to noncollagenous matrix proteins [936]. Deficient phosphorylation due to altered casein kinase II activity has been reported in the hypophosphatemic mouse, an animal model of human hypophosphatemic rickets, that is resistant to phosphate and vitamin D treatment [937]. Further, because there are protein kinases in the extracellular matrix, it is likely that phosphoprotein phosphatases are also present.
C. Bone Morphogenetic Proteins (BMPs) The existence of BMPs (as a concept rather than as isolated proteins) was first suggested by Huggins [938] and later by Urist [939] who observed that the demineralized bone matrix, when implanted in an ectopic site, would cause cartilage and bone formation. Their observations led to an intense search to purify and characterize the factor present in bone. What followed was the identification of a class of proteins that is now recognized as the TGF- superfamily, which include the BMPs, and are capable of inducing bone when implanted into various sites within test animals. The BMPs, with or without carriers, alone or in combination, all seem to have the ability to cause mesenchymal cells to differentiate, and hence to induce bone formation (see Chapter 5 by Olsen). Effects of BMPs on cell morphology and function, proliferation, and differentiation have been reviewed in detail elsewhere [940 – 942]. However, it is now clear that there is a great deal of redundancy and that these proteins are not specific to bone. It is also not yet known how the temporal and spatial pattern of expression of these proteins controls developmental processes. The recognition that the response to these proteins is mediated by cell surface receptors [943] has opened the door to a rapidly expanding field, which will provide a great deal of insight into how this class of proteins functions. Six of the seven well-characterized BMPs are members of the TGF- superfamily whose primary structures have been elucidated by molecular cloning [941]. Of these, only BMP-1 does not have the TGF- structure. The carboxy-terminal regions of BMP-2 – BMP-7 have high structural homology with seven cysteine residues at conserved locations. The mature protein (lacking the propeptides) forms dimers. As reviewed by Wozney and co-workers [941,944,945],
these may exist as hetero-or homodimers, each with different activities. All of the mature BMPs have basic amino termini, which persist following the removal of long propeptides. The mature chains are also glycosylated. BMP-1, in contrast, has domains resembling those in metallo endoproteases and the C1 and C1q of components of complement. The crystal structure of one of the TGF-s (TGF-2) has been determined [946]. It has been shown by NMR [947] that it has an extended rather than a globular structure without a hydrophobic core due to the arrangement of the four disulfide bonds, which form the so-called “TGF- knot’’ in one section of the molecule. Such extended structures common to the TGF- superfamily may be important for interaction with the receptors of the TGF s, which are transmembrane serine-threonine kinases [948]. Based on sequence analyses, it has been found that all of the BMPs have hydrophobic leader sequences, suggesting that they are secreted proteins. This is in line with their proposed functions as growth factor bone-inducing agents. The identification of the defect in the short-ear mutant mouse [949] as an abnormality in the BMP5 gene led to the suggestion that BMPs are involved in the formation, patterning, and repair of specific morphological features. BMP has also been linked to the patterning in the developing chick limb [950]. Another illustration of effects of BMPs is seen in mouse model of fibrodysplasia ossificans progressive, which lacks this protein [951]. Studies have identified BMP receptors on mesenchymal cells, providing a clue as to how these soluble bone derived proteins may function as growth factors in situ. These types of studies have shown that the BMP family is functionally redundant in that phenotypic changes often are not apparent unless more than one of these factors is knocked out [952] (see Chapter 5).
D. Growth Factors The skeletal matrix and mineral are repositories for growth factors [953 – 956], such as the transforming growth factor- family [941], insulin-like growth factors [957], platelet-derived growth factor [958], basic and acidic fibroblast growth factors [953], interleukins [959,960], granulocyte – macrophage colony-stimulating factor [961], tumor necrosis factor [962], and the bone morphogenetic proteins (see earlier discussion). Many of these growth factors are products of bone cells as well as other cells and have specific receptors on osteoblasts, osteoclasts, and osteocytes [963]. These receptors include cell surface receptors that are linked to G-proteins, have directly linked kinase activities [964], and/or that are channel linked [965]. Thus they serve autocrine (self-regulating) and paracrine (regulating other cells) functions [966]. Their actions are regulated by factors such as systemic hormones and by mechanical stress.
154
GOKHALE, BOSKEY, AND ROBEY
The actions of these growth factors in bone are being reported at an exponential rate and have been reviewed by Gowen et al. [962], Rosen et al. [957], and Yoneda [967] (see also Chapter 14). In general these growth factors have mitogenic and proliferative activities in bone, and many influence the expression of specific phenotypic proteins, thereby affecting both bone formation and resorption. The ability to affect these processes depends on the presence of receptors for the particular growth factor or cytokine. The mechanism of action of cytokine binding to such receptors generally involves action of a transmembrane protein kinase, which in turn results in phosphorylation or dephosphorylation of multiple proteins in the signal transduction cascade [948].
E. Serum Proteins (2-HS-Glycoprotein, Fetuin, Albumin, etc.) The list of nonstructural proteins that have been identified in bone that originate from serum and become entrapped in bone is quite lengthy [968]. This ability of bone to adsorb such a broad spectrum of proteins makes it uniquely suited to be the reservoir of multiple growth factors or proteins that may be required and liberated at least in part by bone resorption. Albumin, 2-HS-glycoprotein, fetuin, transferrin, 1-antitrypsin, 1-antichymotrypsin, IgG, haptoglobulin, hemopexin, serum cholinesterase, and soluble fibronectin are among the plasma proteins that accumulate in bone in detectable amounts [969 – 971]. Their accumulation is most likely due to their binding to hydroxyapatite. This affinity for hydroxyapatite is frequently the basis of the separation of different classes of the immunoglobulins [972], removal of interfering plasma constituents [973] and other purification schemes. Human 2-HS-glycoprotein is produced in the liver and circulates in the bloodstream [974]. There is some debate as to whether connective tissue cells can produce the protein, but it is apparent that it specifically accumulates in mineralized tissues. The high concentration of 2-HS-glycoprotein in bone cannot be due entirely to the presence of blood in the tissue, as the relative amount is too high and the ratio of the protein to albumin is enhanced 100-fold in bone relative to blood [970]. Chondrocytes have been shown to express 2-HSglycoprotein where the protein appears to enhance endochondral ossification. The message is absent from normal human bone cell lines [975], indicating that it probably accumulates in human bone by adsorption from the circulation. In rodents a bone sialoprotein has been shown to have high homology with the 2-HS-glycoprotein [976,977]. However, while Ohnishi et al. [977] found that the protein was expressed by osteoblasts, immunolocalization of the analogous protein by Mizuno et al. [978] failed to show its
presence within the cell per se, implying that like the human homologue it is deposited in bone from other sources. 2-HS-Glycoprotein consists of two nonidentical glycosylated peptide chains (chains A and B) that are held together by disulfide bonds [979,980]. These subunits are characterized by repeating Ala-Ala and Pro-Pro sequences. The initial biosynthetic product is a single polypeptide that is cleaved to form the A and B chains of the mature molecule [981]. The major glycoside is sialic acid [982]. Prediction of the primary structure and chemical digestion studies have led to descriptions of the domain structure of the protein [983,984]. In addition to the single disulfide bond linking the two individual chains by their extreme NH2 and COOH ends, there are five intradisulfide bonds in the A (heavy) chain. The light B chain has no intrachain S – S bonds. The A chain is composed of three domains, consisting of S – S linked loops. Of the five loops that span 4- to 19-amino-acid residues, two highly homologous loops form one domain, flanked on either side by the other tandem repeats. This structure is typical of that of bovine fetuin [985], a member of the cystatin superfamily, and of the histidine-rich glycoprotein family [984]. The cystatin family proteins all have “cystatin-like” domains of linearly arranged disulfide linked loops [984,986] and a variety of N-glycosides [987] that vary within the family, and for the same protein they vary with species. Members of the cystatin family are inhibitors of cysteine proteases. The 2HS-glycoprotein is also homologous to a nonphosphorylated sialoprotein found in rodent bone [976,977]; however, the rodent counterpart of 2-HS-glycoprotein consists of one rather than two chains. 2-HS-Glycoprotein can bind TGF-/BMP cytokines and block their osteogenic activity in culture. Studies with dexamethasone-treated rat bone marrow cell cultures have shown that recombinant fetuin as well as native serum protein can inhibit osteogenesis by similar efficiency. Northern blots showed that both 2-HS-glycoprotein and high doses of TGF- suppressed transcripts of alkaline phosphatase, osteopontin, collagen type I, and bone sialoprotein. These data suggest that 2-HS-glycoprotein (together with TGF) is involved in regulation of osteogenesis in remodeling bone [988]. The human gene sequence for 2-HS-glycoprotein is known [989]. The gene is on chromosome 3, and two RFLPs have been identified. The single mRNA species predicts an 18 residue signal peptide, followed by a sequence that codes for the A and B peptides with an intervening sequence between them. This sequence is presumably lost during cleavage of the precursor to form the mature molecule. In tissues other than bone, 2-HS-glycoprotein has several functions. It has been shown to have opsonic properties [971], to promote endocytosis, and to bind DNA [990,991].
155
CHAPTER 4 The Biochemistry of Bone
In liver, 2-HS-glycoprotein (pp63) plays an essential role in insulin signaling, inhibiting the insulin-dependent tyrosine kinase activity of the insulin receptor [992] without competing with insulin binding. It is also believed to be involved in the immune response [993,994]. 2-HS-Glycoprotein has been suggested to play a variety of roles in bone, but none of these has been proven conclusively. It appears at selected times during development and may be involved in differentiation [995]. This idea is supported by earlier studies demonstrating that 2-HS-glycoprotein is found in higher concentrations in young bone as opposed to mature bone [996,997]. The protein appears to have a higher affinity for calcium phosphates than other serum proteins, as the addition of calcium and phosphate to serum led to the removal of all the 2-HS-glycoprotein but removed less than 1% of the albumin [970]. 2-HS-Glycoprotein is also believed to be involved in bone mineralization [998], as decreased levels of this protein are seen in bones with abnormal mineralization pattern due to malnourishment, certain malignancies [993], and Paget’s disease [970]. In contrast, 2-HS-glycoprotein levels are high in some patients with OI [996]. The suggestion that this glycoprotein might affect mineralization is also supported by solution studies. In fact, the ability of serum to inhibit the solution-mediated conversion of amorphous calcium phosphate to hydroxyapatite [999] was attributed to the presence of this high-affinity glycoprotein. This was later confirmed in unpublished in vitro studies where purified 2-HS-glycoprotein was seen to be a much better inhibitor than other serum proteins. The 2-HS-glycoprotein-deficient mouse did not show skeletal abnormalities [1000]; however, the serum from these animals did not inhibit apatite formation as efficiently as that from wildtype animals. The mutant animals also developed ectopic calcifications, confirming the role of this protein as a serum inhibitor of calcification. Other studies suggest that the protein is involved in the recruitment of osteoclast precursors to resorptive sites [817,1001]. Because of its homology to cysteine protease inhibitors, a role in regulating bone resorption seems likely, and it may be that the high concentrations in very young bone may be related to the need to limit remodeling during development. Whole serum [1002] and albumin have been shown to inhibit hydroxyapatite growth in solution [1002 – 1004]. The ability of albumin to inhibit apatite growth is attributed to the affinity of albumin for apatite [973,1005 – 1007]. Specifically, albumin at 50 – 250 g/ml alters the linear rate of growth of apatite seed crystals by binding to the mineral on several faces [1003] and blocking the growth of crystal agglomerates [1004]. Although the primary function of albumin in bone is not apt to be one of regulation of mineralization, the extent of inhibition of hydroxyapatite growth in solution indicated that phosphoproteins were more effective
inhibitors than albumin. Ions such as citrate and magnesium were also less effective in retarding apatite growth in solution [1003]. Transferrin [1008], IgG, IgE, and the other serum proteins also bind to apatite. Studies from the Laboratories of Borkey and Heywood indicated that IgG had no effect on hydroxyapatite formation, morphology, or growth. None of the other serum proteins has been reported to have any effect on either inhibition or promotion of mineralization. Effects on other events in bone matrix deposition are not known; however, because these serum proteins increase in concentration as mineralization progresses [970], it is unlikely that they play significant roles in the bone deposition process.
VII. REQUIREMENTS FOR MATRIX MINERALIZATION Analyses of diseased tissues and tissues from transgenic animals indicate that there are a number of cellular and extracellular factors essential for physiologic mineral deposition. Physiologic mineral deposition is distinguished here from dystrophic or pathologic mineral deposition by its ordered, characteristic appearance on an oriented collagenous matrix. In contrast, in dystrophic calcium phosphate deposition in general, and dystrophic apatite deposition in particular, the mineral crystals are often not collagen associated. They surround and engulf cells and may be longer in size and different in included ion composition from physiologic mineral [1009,1010]. From the just described definition, it is apparent that for physiologic mineral deposition, there must be an appropriate collagen-based matrix. This is emphasized by (a) the smaller size of hydroxyapatite crystals in OI bone [157] and (b) the relative abundance of mineral that is not associated with collagen in these bones with deficient and/or impaired collagen production [158]. Although in some cases the defective mineralization in the OI bones may also be attributed to altered matrix protein production [167] or retention, collagen is clearly an absolute requirement for physiologic bone mineralization. Similarly, because fibronectin forms the basis on which collagen is deposited, it must be also a requirement. Equally apparent from this definition is the essential presence of Ca2+ and Pi ions for mineralization. Calcium ions may be supplied from the cells or from circulating or localized calcium-binding proteins. Phosphate ions may be derived from the breakdown of pyrophosphate, an abundant metabolic product, from hydrolysis of phosphoesters or phosphoproteins, or from circulating Pi ions. The exact Ca2 and Pi content of the extracellular fluid of bone is not known, but in cartilage, micropuncture studies showed the pH to be 7.58, and total Ca2 and total Pi to be 1 – 12 and 3 – 12 mg/dl, respectively [231]. For the formation of
156
GOKHALE, BOSKEY, AND ROBEY
apatite, a basic environment is also required, and many of the highly anionic matrix proteins probably contribute to creating this environment. Which of these matrix proteins is truly essential for mineralization of bone can only be guessed until the sequence of protein expression is determined precisely and appropriate knockout and transgenic models are developed. Even in these cases it may be difficult to prove an essential role for mineralization, as it is already apparent that there are redundant controls of this critical process. It is certain that the cells are required for the production of a physiologic matrix, synthesizing and exporting necessary enzymes, growth factors, and matrix molecules. As discussed later, the formation of extracellular matrix vesicles is also apt to prove critical for the initiation of mineralization in some cases.
VIII. PATHWAYS OF MATRIX MINERALIZATION Apatite crystals found in bone are distinct from most naturally occurring geological crystals, being smaller and containing more impurities. In addition, bone apatite crystals have a plate-like habit [1011], are arranged in an oriented fashion on a collagen-based matrix, and have a very limited size distribution [19]. In general, the mineral crystals in bone (and dentin) are smaller than those in enamel [1012] and in dystrophic deposits in severely atherosclerotic plaques [1013] or other soft tissue calcifications [202]. The bone mineral crystals do vary in size with tissue site, age, and disease [157,1014], but the range in the lengths of the smallest bone mineral crystals and their orientation implies that their growth must be regulated. Bone mineralization is thus distinct from solution-mediated Ca2 phosphate precipitation where similarly sized, nonoriented small crystals that are formed can ripen to appreciably larger sizes [1015]. This is also distinct from geologic apatite formation where high temperatures and pressures yield extremely large single crystals.
A. Physical Chemistry of Mineralization Calcium phosphate precipitation from solution can yield a variety of phases, depending on the pH, Ca to Pi ratio, and solution supersaturation [21,1015,1016]. When the pH is in the physiologic range (7.4 – 7.8), apatite formation occurs with solution Ca:P molar ratios as high as 2 to 1 and as low as 1 to 1 as long as the solution is supersaturated with respect to apatite (has a Ca P product that exceeds the solubility product for apatite). Depending on the supersaturation, intermediate phases such as amorphous calcium phosphate [1017], octacalcium phosphate [1018 – 1020], or
other intermediates may form [21,1021], but in all these cases, the final product is apatitic. Apatite crystals develop in solution when individual ions or ion clusters associate in the same orientation as in the crystal lattice that they are trying to form. When sufficient ion clusters are oriented correctly, they can persist in solution and can serve as a “critical nucleus” for further crystal growth. Homogeneous nucleation, in which crystals form de novo, is a rare process [1015]. Thus it is likely that in most instances of solution-mediated apatite deposition, nucleation occurs on foreign materials such as dust, scratches on the container, and buret tips. Such heterogeneous nucleation yields the initial crystals, which then facilitate additional growth by the process of secondary nucleation. In secondary nucleation, growth sites on the preformed apatite crystals serve as branch points for the formation of new crystals, analogous in many ways to the branching of the growing glycogen molecule during glycogenesis. Proliferation by secondary nucleation results in numerous small crystals. Crystal growth, in the absence of secondary nucleation, would result in fewer, but larger crystals. This suggests that most of the crystals in bone form by a secondary nucleation-like process or by growth from individual nuclei. Unfortunately, what regulates crystal size in bone cannot be determined from studies of proteinfree solutions. The mineral in bone, as in other physiologically calcified tissues, is associated with an organic matrix (Fig. 25). The protein(s) within such matrices can alter the nucleation and growth of mineral crystals in several ways. When isolated in an environment relatively free of body fluids, the protein(s) can chelate Ca2 or Pi ions, reducing the fluid supersaturations, which in turn would prevent crystal nucleation and/or growth. The protein(s) can form a protected environment around the crystal nucleus, sequestering it and thus preventing crystal growth, or stabilizing the nucleus, protecting it from the external environment. The protein(s) may bind Ca2 and/or Pi ions, forming a surface that resembles the apatite surface. In this manner, the protein serves as an epitaxial (similar surface) nucleator, thereby providing a surface for the start of nucleation. The protein(s) may also bind to one or more faces of the growing crystal because its side chains match positions in the lattice, thereby blocking growth in one or more directions or even blocking growth beyond a specific size. The protein(s) might also bind to other proteins, changing their conformation and their ability to affect crystal nucleation and growth according to the pathways described earlier, or might associate with cells resulting in a change in the extracellular Ca P concentration, or pH. The elegant ultrastructural studies of Addadi and colleagues that combined X-ray crystallographic and electron microscopic techniques provide illustrations for each of these mechanisms for the formation of larger crystals of
157
CHAPTER 4 The Biochemistry of Bone
FIGURE 25
Cell-mediated matrix mineralization in developing bone. Early mineralization in chick bone. Electron micrograph showing a 17-day-old embryonic tibia, stained with uranyl acetate and lead citrate. Mineral clusters (C) outside the osteoblast (OB) are associated with collagen (thin arrows) and extracellular matrix vesicles (inset). Empty vesicles (thick arrows) as well as vesicles with mineral are seen. Courtesy of Dr. Steven B. Doty.
calcium carbonates, calcium sulfates, brushite, and octacalcium phosphate [163,1022,1023]. For example, fibronectin has been shown to bind to the ionic surfaces of calcite that did not include water molecules, but does not bind at all to brushite whose surfaces all have bound water [1024]. The acidic macromolecules from sea animals have been shown to determine the shape of calcite crystals [1025]. Cells have been shown to interact with specific faces on such crystals in the presence and absence of RGD-containing macromolecules [1026]. Scanning electron micrographs have similarly been used to identify the binding sites for polyaspartic acid, mollusk shell proteins, and rat dentin phosphoprotein on the surface of octacalcium phosphate [1024]. There are also examples of each of these mechanisms from solution studies of apatite formation. For these, to date, there is no direct evidence of the exact nature of the protein – mineral interaction. Studies of the effects of bone matrix proteins on apatite formation include studies in which preformed seed crystals are added to Ca P solutions, and the rate of crystal growth is determined at fixed Ca P and fixed pH [21] or variable Ca P OH [1021].
Other studies have looked at the formation (nucleation and growth) of apatite from solutions in the presence of insoluble proteins, proteins immobilized on polyanionic beads [1027], or proteins in solution [462]. Diffusion studies in which the protein is held within an agarose [664], silicate [712], or denatured collagen [202] gel have also provided insight into apatite nucleation and growth. From such studies, one can find examples of the mechanisms listed earlier. It should be emphasized, however, that a protein, because of its affinity for apatite, may, in low concentrations, act as a nucleator and in higher concentrations serve to regulate crystal growth. The multifunctional roles of each of the bone matrix proteins should be apparent from the previous sections of this review. The extracellular matrix vesicles (Fig. 25, inset, discussed later) and their component lipids may facilitate Ca and P accumulation, while shielding the apatite nucleus. As reviewed in detail elsewhere [1028], illustration of this behavior in vesicles is seen in the iron oxide forming bacteria and in model liposomes. In model liposomes, where Ca2 accumulation is facilitated by an ionophore, initial mineral
158
GOKHALE, BOSKEY, AND ROBEY
crystals form inside the liposomes in association with the liposome membrane where they eventually grow and puncture the liposome membrane and become exposed to the external solution [1029 – 1032]. Proteins that can prevent apatite nucleation by chelating Ca2 include aggrecan and its associated proteoglycan aggregates [200]. This may be more complicated in vivo where free Ca2 levels may be affected by Donnan equilibrium effects. However, it has been shown that phosphate can then cause the release of this Ca2, facilitating the mineralization process [664,1028]. An example of a protein that appears to block crystal growth by binding to a specific surface or surfaces is osteopontin, which binds with high specificity and decreases the length of the crystals formed in the collagen-gel system [663]. This is in contrast to the dentin phosphoprotein phosphophoryn, which blocks secondary nucleation by binding to growth sites [667]. Immobilized albumin, osteocalcin, and osteonectin appear to act as nucleators [1027], although evidence shows that the beads on which these proteins are immobilized are themselves nucleators. Fibronectin has also been shown to be capable of nucleating apatite in solution [577]. The only bone/dentin specific matrix proteins shown to act as nucleators are bone sialoprotein [665,712,1033], biglycan [1034], and dentin matrix protein 1 (Boskey, unpublished data). When stripped of its associated matrix proteins, collagen itself, does not nucleate apatite, but when its associated phosphoproteins are present, apatite nucleation on collagen occurs in a dose-dependent manner directly related to the extent of protein phosphorylation [124,660]. Collagen peptides containing the carboxylate region, but not the Nterminal domain can also nucleate apatite in solution [1035].The complexed acidic phospholipids and their associated proteolipids [848] and lipoproteins [851] also nucleate apatite formation in solution [851,1036] and when implanted in a cell-free environment in vivo [866].
B. Cell-Regulated Mineralization In each of the just mentioned solution-mediated studies, hydroxyapatite formation was performed in the absence of cells. However, it is the cells that regulate bone mineralization by producing and secreting the appropriate extracellular matrix components, regulating their interactions, and controlling the flux of Ca, P, and OH ions in their extracellular environments. One must look to cell-mediated studies to understand the pathways of matrix mineralization in vitro [reviewed in 1037]. Unfortunately, most of these cellmediated studies have been concerned with understanding the origin and differentiation of the bone cells rather than with the mineralization process per se. However, from analyses of the cell culture systems, one can learn much about mineralization. Another complicating issue with the
cell culture studies is that most rely on 10 mM -glycerolphosphate, an excellent substrate for alkaline phosphatase, as a source of inorganic phosphate. The hydrolysis of 10 mM -glycerophosphate results in inorganic phosphate concentrations two to five times higher than physiologic levels, in turn producing a medium Ca P product much higher (10 mM2) than physiologic (2 – 5 mM2). In such high Ca P medium, larger, nonoriented mineral crystals distinct from those that exist in situ may be formed [1010,1038]. Organ culture for the study of bone formation was first described by Dame Honor Fell in 1925 [1039]. Similar studies of cultured limb rudiments have been used more recently to show stages of differentiation and the effects of a variety of hormones and growth factors on bone formation [1010,1040]. Tenenbaum et al. used an inverted folded chick periosteum model as their organ culture system [1041,1042] with similar results. The limitation of the organ culture systems for the study of the mineralization process is the variety of cell types present. For that reason, isolated osteoblast cultures or cultures of cells that differentiate into osteoblasts and form a mineralized matrix may be preferable. Based on the system described by Friedenstein et al. [1043], Owen [1044], Ashton et al. [1045], and Benayahu et al. [1046], many investigators have stimulated marrow stem cell cultures to differentiate into osteoblasts [reviewed in 1047]. In these, and in isolated osteoblast systems, steroids such as dexamethasone [1048 – 1050], retinoic acid [383], BMP [1051], and -glycerophosphate [1052] are thought to enhance this differentiation. Osteoblasts isolated from fetal tissues, e.g., fetal rat calvariae, are commonly used to study osteoblasts, as are tumor-derived osteoblast-like cells. Examples of these include ROS17/2.8 [6], MG-63 [1053], SaOS [1054], and MC3T3-E1 [1055] cell lines. Methods have also been developed for isolating and culturing cells from more mature human (and animal) bones [9] and immortalized cell lines [1056 – 1058]. As techniques emerge, there will undoubtedly be an increase in the number of cell culture model systems that will exhibit several phenotypic traits. However, it will have to be determined by stringent criteria (physiological matrix mineralization verified at the EM level) whether these model systems will be useful in studying late stages of bone formation (i.e., matrix mineralization). Rat calvarial cells plated at low density yield isolated colonies, some of which are “osteoblast like” based on morphology, matrix composition, alkaline phosphatase expression, and mineralization [525,1059,1060]. In these osteoblast nodules/colonies, active type I collagen expression follows cell proliferation and then is reduced to baseline levels, whereas alkaline phosphatase activity and expression, osteopontin, collagenase, and osteocalcin expression commence after this. In fact, in such cultures, osteocalcin expression is highly correlated with mineral accretion
CHAPTER 4 The Biochemistry of Bone
[814]. BSP, osteopontin, and osteocalcin expression continue to increase as mineralization proceeds [10]. Mineral deposition in such cultures has been characterized by electron microscopy and wide angle X-ray diffraction [1061 – 1063] and shown to contain hydroxyapatite crystals, although not always oriented in the same direction as the collagen fibrils. These types of cell and organ culture systems have allowed the definition of the proposed sequential events involved in the recruitment and proliferation of osteoprogenitor cells and their differentiation into osteoblasts, and the temporal changes in the expression of bone matrix proteins [10]. Consistently, in each of the different model systems [9,11,608,814,1064,1065], osteopontin, collagen, osteonectin, and BSP expression have been shown to precede mineral deposition, whereas osteocalcin expression occurs after mineral deposition has commenced [10]. This pattern is consistent with the functions suggested for these proteins in this chapter, and although such cultures have not yet defined which of these proteins are necessary for mineralization, they did shed some light on the debate concerning the site of initial calcification in bone. In the 1960s the identification of extracellular matrix vesicles (MV) [1066,1067] as the location of the first mineral deposits in a variety of mineralizing tissues, including membranous bone [1068] and calcifying cartilage [1066,1067,1069], led to the suggestion that these membrane-bound bodies were the site of initial calcification in these tissues. Later studies, reviewed by Anderson [1070,1071] and Wuthier et al. [1072] revealed that these bodies were enriched in enzymes associated with mineralization, among them alkaline phosphatase, ATPase, and proteoglycan-degrading enzymes. The MV membranes contained proteolipids capable of facilitating Ca ion transport into the vesicles, and also had matrix collagens (type II, IX, and X) associated with them [1073]. Tissue slices, suspensions of isolated vesicles, and liposome models of matrix vesicles, in solutions of calcium and phosphate, each led to the deposition of apatite. The mechanism seemed to involve the transport of Ca ions into a vesicle that already contained a high concentration of phosphate ions [1073], formation and growth of vesicle associated mineral crystals, and rupture of the vesicle and the spread of larger mineral crystals into the extravesicular environment. As discussed earlier, the vesicle in solution served the purpose of providing a protected/stabilizing environment for the formation of the first crystals. When inhibitors of mineralization were included with the vesicles, mineralization was facilitated over vesicle-free controls [1074]. Vesicles also serve to provide enzymes that can modify the extracellular matrix, facilitating the proliferation of crystals. Extracellular matrix vesicles have been reported to be the site of initial mineralization in bone and cartilage and organ culture systems as described earlier. In such systems,
159 the mineral-associated vesicles are seen prior to the appearance of bulk mineral. One of the major issues in the debate between those that believe that collagen-based matrix is the initial site of mineralization and those who believe mineralization starts in vesicles concerns how the mineral would travel from the vesicles to its highly aligned/oriented sites on the collagen. Culture data support the matrix vesicle theory but do not prove a link between vesicle and collagenmediated mineralization. An association between vesicles and collagen suggests such a link may exist [1075]. The application of computerized axial tomographs to high-resolution electron micrographs by Landis has provided some explanations. Looking at the highly oriented calcified turkey tendon [159], mineral was seen in vesicles at sites different from mineral associated with the collagen. However, at certain sites, the mineral could be seen branching (showing the link between vesicle and collagen associated mineral). Thus when viewed in three dimensions, it appeared that the mineral does not “swim” but rather grows out radially, extending to the collagen fibril. The question remains whether this mineral associates with mineral already in the collagen or whether mineral from the vesicle manages to grow out to specific sites along the collagen fibril. From the simplistic point of view, the former seems more likely, as annealing of crystals does take place. To date, there is no evidence that collagen-based mineralization cannot occur in the absence of vesicles. Further, because mineralization starts at numerous discrete sites within the collagen fibrils, it has been argued that there would not be sufficient space for matrix vesicles to serve as so many distinct nuclei [19]. In contrast, from solution-based data, it is clear that there are matrix proteins associated with the collagen that can act as nucleators, implying that the latter mechanism may predominate. Collagen fibrils depleted of matrix proteins are not mineralized readily in solution. Collagen mineralization as seen at the EM level starts at specific sites, the so-called e bands, where several matrix proteins are known to localize [125]. Some of the matrix proteins in these sites may shield others, preventing nucleation. Others associated with the collagen may serve as nucleators, as described earlier, allowing the oriented deposition of mineral crystals. The growth of these crystals may also become limited by the space available in the collagen fibrils and by the adsorption of other matrix proteins. From the material presented in this review, we can suggest that (a) decorin, which trims the collagen fibrils and disappears from the e bands in mineralized fibrils [117], matrix gla-protein, and 2-HS glycoprotein are among the shielding agents, (b) bone sialoprotein and other phosphorylated proteins, which are closely associated with the collagen hole zones, serve as nucleators, and (c) several of the other matrix proteins produced by osteoblasts may play a role in regulating crystal size.
160
GOKHALE, BOSKEY, AND ROBEY
Acknowledgments The authors thank Drs. Paolo Bianco, Wojciech J. Grzesik, Neal S. Fedarko, and Steven B. Doty for providing photographic materials and Luz Ingles for secretarial assistance. Dr. Boskey’s research as discussed in this review was supported by NIH Grants DE04141, AR037661, and AR41325.
References 1. P. Gehron Robey, Normal bone formation – structure. In “Bone Formation and Repair” (C. T. Brighton, G. Friedlander, and J. M. Lan, eds.), pp. 3 – 12, AAOS, Rosemont, IL, 1994. 2. J. D. Termine, A. B. Belcourt, P. J. Christner, K. M. Conn, and M. U. Nylen, Properties of dissociatively extracted fetal tooth matrix proteins. I. Principal molecular species in developing bovine enamel. J. Biol. Chem. 255, 9760 (1980). 3. J. D. Termine, A. B. Belcourt, K. M. Conn, and H. K. Kleinman, Mineral and collagen-binding proteins of fetal calf bone. J. Biol. Chem. 256, 10403 (1981). 4. W. Peck, S. Birge, and C. Fedak, Bone cells: Biochemical and biological studies after enzymatic isolation. Science 146, 1476 (1964). 5. G. L. Wong and P. V. Cohn, Target cells in bone for parahormone and calcitonin are different: Enrichment for each cell type by sequential digestion of mouse calvaria and selective adhesion to polymeric surfaces. Proc. Natl. Acad. Sci. USA. 72,(8); 3167 – 3171 (1975). 6. R. J. Majeska, S. B. Rodan, and G. A. Rodan, Parathyroid hormone-responsive clonal cell lines from rat osteosarcoma. Endocrinology 107, 1494 (1980). 7. N. C. Partridge, J. J. Jeffrey, L. S. Ehlich, S. L. Teitelbaum, C. Fliszar, H. G. Welgus, and A. J. Kahn, Hormonal regulation of the production of collagenase and a collagenase inhibitor activity by rat osteogenic sarcoma cells. Endocrinology 120, 1956 (1987). 8. J. N. Beresford, J. A. Gallagher, J. W. Poser, and R. G. G. Russell, Production of osteocalcin by human bone cells in vitro. Effects of 1,25(OH)2D3, 24,25(OH)2D3, parathyroid hormone, and glucocorticoids. Metab. Bone Dis. Relat. Res. 5, 229 (1984). 9. P. Gehron Robey and J. D. Termine, Human bone cells in vitro. Calcif. Tissue Int. 37, 453 – 460 (1985). 10. G. S. Stein and J. B. Lian, Molecular mechanisms mediating developmental and hormone-regulated expression of genes in osteoblasts: An integrated relationship of cell growth and development, In “Cellular and Molecular Biology of Bone” (M. Noda, ed.), pp. 47 – 95, Academic Press, New York, 1993. 11. P. G. Strauss, E. I. Closs, J. Schmidt, and V. Erfle, Gene expression during osteogenic differentiation in mandibular condyles in vitro. J. Cell Biol. 110, 1369 – 1378 (1990). 12. P. Gehron Robey, P. Bianco, and J. D. Termine, The cellular biology and molecular biochemistry of bone formation. In “Disorders of Mineral Metabolism” (F. L. Coe and M. J. Favus, eds.), pp. 241 – 263. Raven Press, New York, 1992. 13. A. Caplan, The mesengenic process. Clin. Plast. Surg. 21, 429 – 435 (1994). 14. F. Liu, L. Malaval, A. K. Gupta, and J. E. Aubin, Simultaneous detection of multiple bone-related mRNAs and protein expression during osteoblast differentiation: Polymerase chain reaction and immunocytochemical studies at the single cell level. Dev. Biol. 166, 220 – 234 (1994). 15. S. Lee and C. L. Davidson, The role of collagen in the elastic properties of calcified tissues. J. Biomech. 10, 473 – 486 (1977). 16. A. L. Boskey, Mineralization, structure, and function of bone. In “Dynamics of Bone and Cartilage Metabolism” (M. J. Seibel, S. P.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28. 29. 30.
31.
32.
33.
Robins, and J. P. Bilzekian, eds.), pp.153 – 162. Academic Press, 1999. R. A. Robinson and M. L.Watson, Crystal-collagen relationships in the bone as observed in the electron microscope. III. Crystal and collagen morphology as a function of age. Ann. NY. Acad. Sci. 60, 596 – 628 (1955). A. L. Arsenault and M. D. Grynpas, Crystals in calcified epiphyseal cartilage and cortical bone of the rat. Calcif. Tissue Int. 43, 219 – 225 (1988). J. Christoffersen and W. J. Landis, A contribution with review to the description of mineralization of bone and other calcified tissues in vivo. Anat. Rec. 230, 435 – 450 (1991). A. S. Posner and F. Betts, Molecular control of tissue mineralization. In “The Chemistry and Biology of Mineralized Connective Tissues” (A. Veis, ed.), pp. 257 – 271, Elsevier, Amsterdam, 1981. E. D. Eanes, Dynamic aspects of apatite phases of mineralized tissues, model studies. In “The Chemistry and Biology of Mineralized Tissues” (W. T. Butler, ed.), pp. 213 – 221, Ebsco Media, Birmingham, AL, 1985. J. Moradian-Oldak, F. Frolow, L. Addadi, and S. Weiner, Interactions between acidic matrix macromolecules and calcium phosphate ester crystals: Relevance to carbonate apatite formation in biomineralization, Proc. Roy. Soc. Lond. B. 247, 47 – 55 (1992). R. T. Bailey and C. Holt, Fourier transform infrared spectroscopy and characterisation of biological calcium phosphates. In “Calcified Tissue,” (D. W. L. Hukins, ed.), chapt. 5. Macmillan Press, Houndsmills, England, 1989. D. D. Lee, W. J. Landis, and M. J. Glimcher, The solid, calciumphopshate mineral phases in embryonic chick bone characterized by high-voltage electron microscopy. J Bone Mineral Res. 1, 425 – 432 (1986). E. D. Eanes, L. Powers, and J. L. Costa, Extended X-ray absorption fine structure (EXAFS) studies on calcium in crystalline and amorphous solids of biological interest. Cell Calcium 2, 251 – 262 (1981). W. P. Aue, A. H. Roufosse, M. J. Glimcher, and R. G. Griffin, Solid state phosphorus nuclear magnetic resonance study of synthetic solid phase of calcium phosphate; potential models of bone mineral. Biochemistry 23, 6110 – 6114 (1984). D. W. L. Hukins, A. J. Cox, and J. E. Harries, EXAFS characterisation of poorly crystalline deposits from biological systems in the presence of highly crystalline material. J. Physique 48, 1181 – 1184 (1986). A. S. Posner, Crystal chemistry of bone mineral. Physiol. Rev. 40, 760 – 792 (1969). R. A. Young, Biological apatite vs hydroxyapatite at the atomic level. Clin. Orthop. Rel. Res. 113, 249 – 262 (1975). L. Cohen and R. Kitzes, Infrared spectroscopy and magnesium content of bone mineral in osteoporotic women. Israel J. Med. Sci. 17, 1123 – 1125 (1981). C. Rey, M. Shimizu, B. Collins, and M. J. Glimcher, Resolution-enhanced fourier transform infrared spectroscopy study of the environment of phosphate ions in early deposits of a solid phase of calcium-phosphate in bone and enamel, and their evolution with age. I. Investigations in the nu 4 PO4 domain. Calcif. Tissue Int. 46, 384 – 394 (1990). C. Rey, M. Shimizu, B. Collins, and M. J. Glimcher, Resolution-enhanced Fourier transform infrared spectroscopy study of the environment of phosphate ion in the early deposits of a solid phase of calcium phosphate in bone and enamel and their evolution with age. 2. Investigations in the nu3PO4 domain. Calcif. Tissue Int. 49, 383 – 388 (1991). C. Rey, V. Renugopalakrishnan, M. Shimizu, B. Collins, and M. J. Glimcher, A resolution-enhanced Fourier transform infrared
161
CHAPTER 4 The Biochemistry of Bone
34.
35.
36.
37. 38.
39.
40.
41.
42. 43. 44. 45. 46.
47.
48.
49.
50.
51.
52.
53.
54.
spectroscopic study of the environment of the CO32 ion in the mineral phase of enamel during its formation and maturation. Calcif. Tissue Int. 49, 2259 – 2268 (1991). J. D. Termine and D. R. Lundy, Hydroxide and carbonate in rat bone mineral and its synthetic analogues. Calcif. Tissue Res. 13, 73 – 82 (1973). A. J. Kahn and N. C. Partridge, Bone resorption in vivo. In “Bone” (B. K. Hall, ed.), Vol. 2, pp. 199 – 140. CRC Press, Boca Raton, FL, 1990. A. H. Roufosse, W. J. Landis, W. K. Sabine, and M. G. Glimcher, Identification of brushite in newly deposited bone mineral from embryonic chiecks. J. Ultrastruct. Res. 68, 235 – 255 (1979). W. E. Brown, Crystal growth of bone mineral. Clin. Orthop. Rel. Res. 44, 205 – 220 (1966). A. S. Posner and F. Betts, Synthetic amorphous calcium phosphate and its relation to bone mineral structure. Accts. Chem. Res. 8, 273 – 281 (1975). F. C. M. Driessens, J. W. E. Van Dijk, and J. M. P. M. Borggreven, Biological calcium phosphates and their role in the physiology of bone and dental tissues. Calcif. Tissue Res. 26, 127 – 136 (1978). Y. Wu, M. J. Glimcher, C. Rey, and J. L. Ackerman, A unique protomate phosphate group in bone mineral not present in synthetic calcium phosphates: Identification by phosphorus-31 solid state NMR spectroscopy. J. Mol. Biol. 244, 423 – 435 (1994). K. A. Piez, Molecular and aggregate structures of collagens. In “Extracellular Matrix Biochemistry” (K. A. Piez and A. H. Reddi, eds.), pp. 1 – 40. Elsevier, New York, 1984. P. P. Fietzek and K. Kuhn, The primary structure of collagen, Int. Rev. Connect. Tissue Res. 7, 1 – 60 (1976). G. R. Martin, R. Timpl, P. K. Muller, and K. Kuhn, The genetically distinct collagens. Trends Biochem. Sci. 251, 285 – 287 (1985). K. S. E. Cheah, Collagen genes and inherited connective tissue disease. Biochem. J. 229, 287 – 303 (1985). J. C. Brown and R. Timpl, The collagen superfamily. Int. Arch. Allergy. Immunol. 107(4), 484 – 490 (1995). D. A. D. Parry, The molecular and fibrillar structure of collagen and its relationship to the mechanical properties of connective tissue. Biophys. Chem. 29, 195 – 209 (1988). D. B. Haniford and D. E. Pulleybank, Transition of a cloned d(AT)n-d(AT)n tract to a cruciform in vivo. Nucleic Acid Res. 13, 4343 – 4363 (1985). M. J. McLean and R. D. Wells, The role of DNA sequence in the formation of Z-DNA versus cruciforms in plasmids. J. Biol. Chem. 263, 7370 – 7377 (1988). C. V. Heath, R. M. Denome, and C. N. Cole, Spatial constraints on polyadenylation signal function. J. Biol. Chem. 265, 9098 – 9104 (1990). L. H. Naylor and E. M. Clark, d(TG)n-d(CA)n sequences upstream of the rat prolactin gene form Z-DNA and inhibit gene transcription. Nucleic Acids Res. 18, 1595 – 1601 (1990). M.-L. Chu, W. De Wet, M. Bernard, J. F. Ding, M. Morabito, J. Myers, C. Williams, and F. Ramirez, Human pro allpha 1(I) collagen gene structure reveals evolutionary conservation of a pattern of introns and exons. Nature 310, 337 – 340 (1984). F. Ramirez, M. Bernard, M.-L. Chu, L. Dickson, F. Sangiorgi, D. Weil, W. de Wet, C. Junien, and M. Sobel, Isolation and characterization of the human fibrillar collagen genes. Ann. N.Y. Acad. Sci. 460, 117 – 129 (1985). W. deWet, M. Bernard, V. Benson-Chanda, M.-L. Chu, L. Dickson, D. Weil, and F. Ramirez, Organization of the human pro-alpha 2(I) collagen gene. J. Biol. Chem. 262, 16032 – 16036 (1987). H. Kuivaniemi, G. Tromp, M. L. Chu, and D. J. Prockop, Structure of a full-length cDNA clone for the prepro2(I) chain of human type I procollagen. Comparison with the chicken gene confirms
55.
56. 57. 58.
59. 60.
61. 62.
63.
64.
65.
66.
67.
68.
69.
70.
71. 72.
73.
74.
unusual patterns of gene conservation. Biochem. J. 252, 633 – 640 (1988). S. Strickland, J. Huarte, D. Belin, A. Vassalli, R. J. Rickles, and J. D. Vassalli, Antisense RNA directed against the 3 noncoding region prevents dormant mRNA activation in mouse oocytes. Science 241, 680 – 684 (1988). J. L. Manley, Polyadenylation of mRNA precursors. Biochim. Biophys. Acta 6950(1), 1 – 12 (1988). W. M. Krojewska, Regulation of transcription in eukaryotes by DNA-binding proteins. Int. J. Biochem. 24, 1885 – 1899 (1992). C. Murre, G. Bain, M. A. van Dijk , I. Engel, B. A. Furnari, M. E. Massari, J. R. Matthews, M. W. Quong, R. R. Rivera, and M. H. Stuiver, Structure and function of helix-loop-helix proteins. Biochim. Biophys. Acta. 1218, 129 – 135 (1994). S. E. Phillips, Built by association: Structue and funtion of helixloop-helix DNA-binding proteins. Structure 2, 1 – 4 (1994). V. J. Dwarki, M. Montminy, and I. M. Verma, Both the basic region and the “leucine ziper” domain of the cyclic AMP response element binding (CREB) protein are essential for transcriptional activation. EMBO J. 9, 225 – 232 (1990). D. Landsman and M. Bustin, A signature for the HMG-1 box DNAbinding proteins. Bioessays 15, 539 – 546 (1993). B. L. Vallee and D. S. Auld, Functional zinc-binding motifs in enzymes and DNA-binding proteins. Faraday Discuss 93, 47 – 65 (1992). M. L. Chu, W. deWet, M. Bernard, and F. Ramirez, Fine structural analysis of the human pro collagen gene: Promoter structure, AluI repeats and polymorphism transcripts. J. Biol. Chem. 260, 2315 – 2320 (1985). M. H. Finer, S. Aho, and L. Gerstenfeld, Unusual DNA sequences located within the promoter region and the first intron of the chicken proalpha1(I) collagen gene. J. Biol. Chem. 262, 13323 (1987). A. Lichtler, M. L. Stover, and J. Angilly, Isolation and characterization of the rat 1(I) collagen promoter: Regulation by 1,25-dihydroxyvitamin D3. J. Biol. Chem. 264, 3072 (1989). A. Schmidt, P. Rossi, and B. de Crombrugghe, Transcriptional control of the mouse 2(I) collagen gene: Functional deletion analysis of the promoter and evidence for cell-specific expression. Mol. Cell. Biol. 6, 347 – 354 (1986). G. Vogeli, H. Ohkubo, M. E. Sobel, Y. Yamada, I. Pastan, and B. de Crombrugghe, Structure of the promoter for chicken 2 type I collagen gene, Proc. Natl. Acad. Sci. USA 78, 5534 – 5538 (1981). G. T. Merlino, C. McKeon, B. de Crombrugghe et al., Regulation of the expression of genes encoding types I, II and III collagen during chick embryonic development. J. Biol. Chem. 258, 10041 (1983). C. M. S. Rossouw, W. P. Vergeer, S. J. du Plooy, M. P. Bernard, F. Ramirez, and W. J. de Wet, DNA sequences in the first intron of the human pro-1(I) collagen gene enhance transcription. J. Biol. Chem. 262, 15151 – 15157 (1987). P. Bornstein, J. McKay, and J. K. Morishima, Regulatory elements in the first intron contribute to transcriptional control of the human 1(I) collagen gene. Proc. Natl. Acad. Sci. USA 82, 1321 (1987). E. Vuorio and B. de Crombrugghe, The family of collagen genes. Annu. Rev. Biochm. 59, 837 – 872 (1990). A. Lichtler, M. L. Stover, J. Angilly, B. Kream, and D. W. Rowe, Isolation and characterization of the rat alpha1(I) collagen promoter. J. Biol. Chem. 269, 16518 (1994). R. D’Souza, K. Niederreither, and B. de Crombrugghe, Osteoblastspecific expression of the alpha 2(I) collagen promoter in transgenic mice: correlation with the distribution of TGF-1. J. Bone Miner. Res. 8, 1127 – 1136 (1993). W. deWet, M. L. Chu, and D. J. Prockop, The mRNAs for the pro(I) and pro2(I) chains of type I procollagen are translated at the same
162
75.
76. 77. 78.
79.
80.
81.
82.
83.
84.
85.
86.
87.
88.
89.
90.
GOKHALE, BOSKEY, AND ROBEY rate in normal human fibroblasts and fibroblasts from two variants of osteogenesis imperfecta with altered steady state rations of the two mRNAs. J. Biol. Chem. 258, 14385 – 14389 (1983). J. Vuust, M. E. Sobel, and G. R. Martin, Regulation of type I collagen synthesis: total pro1(I) and pro2(I) mRNAs are maintained in a 2:1 ratio under varying rates of collagen synthesis. Eur. J. Biochem. 151, 449 – 453 (1985). P. Bornstein, D. Horlein, and J. McPherson, “Myelofibrosis and the Biology of Connective Tissue,” p. 61 A. R. Liss, New York, 1984. P. Bornstein and H. Sage, Structurally distinct collagen types. Annu. Rev. Biochem. 49, 957 – 1003 (1980). P. H. Krebsbach, J. R. Harrison, A. C. Lichtler, C. O. Woody, D. W. Rowe, and B. E. Kream, Transgenic expression of COL1A1-chloramphenicol acetyltransferase fusion genes in bone: Differential utilization of promoter elements in vivo and in cultured cells. Mol. Cell. Biol. 13, 5168 – 5174 (1993). D. Pavlin, A. C. Lichtler, A. Bedalov, B. E. Kream, J. R. Harrison, H. F. Thomas, G. A. Gronowicz, S. H. Clark, C. O. Woody, and D. W. Rowe, Differential utilization of regulatory domains within the 1(I) collagen promoter in osseous and fibroblastic cells. J. Cell Biol. 116, 227 – 236 (1992). Y. Shibata, Y. Abiko, K. Goto, Y. Moriya, and H. Takiguchi, Heparin stimulates the collagen synthesis in mineralized cultures of the osteoblast-like cell line, MC3T3-E1. Biochem. Int. 28, 335 – 344 (1992). H. C. Tenenbaum, H. Limeback, C. A. McCulloch, H. Mamujee, B. Sukhu, and M. Torontali, Osteogenic phase-specific co-regulation of collagen synthesis and mineralization by beta-glycerophosphate in chick periosteal cultures. Bone 13, 129 – 138 (1992). K. Nohtomi, K. Sato, K. Shizume, K. Yamazaki, H. Demura, K. Hosada, Y. Murata, and H. Seo, Stimulation of interleukin-4 of cell proliferation and mRNA expression of alkaline phosphatase and collagen type I in human osteoblast-like cells of trabecular bone. Bone Miner. 27, 69 – 79 (1994). R. S. Bockman, P. T. Guidon, Jr., L. C. Pan, R. Salvatori, and A. Kawaguchi, Gallium nitrate increases type I collagen and fibronectin mRNA and collagen protein levels in bone and fibroblast cells. J. Cell Biochem. 52, 396 – 403 (1993). L. G. Raisz, P. M. Fall, D. N. Petersen, A. Lichtler, and B. E. Kream, Prostaglandin E2 inhibits alpha 1(I)procollagen gene transcription and promoter activity in the immortalized rat osteoblastic clonal cell line. Mol. Endocrinol. 7, 17 – 22 (1993). D. Pavlin, A. Bedalov, M. S. Kronenberg, B. E. Kream, D. W. Rowe, C. L. Smith, J. W. Pike, and A. C. Lichtler, Analysis of regulatory regions in the COL1A1 gene responsible for 1,25-dihydroxyvitamin D3-mediated transcriptional repression in osteoblastic cells. J. Cell Biochem. 56, 490 – 501 (1994). L. G. Raisz, P. M. Fall, B. Y. Gabbitas, T. L. McCarthy, B. E. Cream, and E. Canalis, Effects of prostaglandin E2 on bone formation in cultured fetal rat calvariae: Role of insulin-like growth factor-I. Endocrinology 133, 1504 – 1510 (1993). B. E. Kream, D. LaFrancis, D. N. Peterson, C. Woody, S. Clark, D. W. Rowe, and A. Lichtler, Parathyroid hormone represses ? 1(I) collagen promoter activity in cultured calvariae from neonatal transgenic mice. Mol. Endocrinol. 7, 399 – 408 (1993). R. Hata, H. Hori, Y. Nagai, S. Tanaka, M. Kondo, M. Hiramatsu, N. Utsumi, and M. Kumegawa, Selective inhibition of type I collagen synthesis in osteoblastic cells by epidermal growth factor. Endocrinology 115, 867 – 876 (1984). M. M. Hurley, C. Abreu, J. R. Harrison, A. C. Lichtler, L. G. Raisz, and B. E. Kream, Basic fibroblast growth factor inhibits type I collagen gene expressionin osteblastic MC3T3-E1 cells. J. Biol. Chem. 268, 5588 – 5593 (1993). P. Van Vlasselaer, B. Borremans, U. van Gorp, J. R. Dasch, and R. De Waal-Malefyt, Interleukin 10 inhibits transforming growth
91. 92.
93.
94. 95. 96.
97.
98. 99. 100.
101.
102.
103.
104. 105.
106.
107.
108.
109. 110.
111.
112.
factor-beta (TGF-beta) synthesis required for osteogenic commitment of mouse bone marrow cells. J. Cell Biol. 124, 569 – 577 (1994). R. F. Klein and K. M. Wiren, Regulation of osteoblastic gene expression by lead. Endocrinology 132, 2531 – 2537 (1993). J. W. Dietrich, E. M. Canalis, D. M. Maina, and L. G. Raisz, Effects of glucocorticoids on fetal rat bone collagen synthesis in vitro. Endocrinology 104, 715 – 721 (1979). E. Canalis, Effect of glucocorticoids on type I collagen synthesis, alkaline phosphatase activity, and deoxyribonucleic acid content in cultured rat calvariae, Endocrinology 112, 931 – 939 (1983). A. Newman, RNA splicing: Activity in the spliceosome. Curr. Biol. 4, 462 – 464 (1994). S. M. Berget, Exon recognition in vertebrate splicing. J. Biol. Chem. 270, 2411 – 2414 (1995). M. J. Moore and P. A. Sharp, Evidence for two active sites in the spliceosome provided by stereochemistry of pre-mRNA splicing. Nature 365, 364 – 368 (1993). Y. Yamada, M. Mudryj, and B. de Crombrugghe, A uniquely conserved regulatory signal is found around the translation initiation site in three different collagen genes. J. Biol. Chem. 258, 14914 (1983). J. H. Fessler and L. I. Fessler, Biosynthesis of procollagen. Annu. Rev. Biochem. 47, 129 – 162 (1978). D. J. Prockop, Mutations in collagen genes as a cause of connective-tissue diseases. N. Engl. J. Med. 326, 540 – 546 (1992). G. N. Ramachandran and C. Ramakrishnan, Molecular structure. In “Biochemistry of Collagen” (G. N. Ramachandran and A. H. Reddi, eds.), pp. 45 – 84. Plenum Press, New York, 1976. M. Yamauchi, E. P. Katz, K. Otsubo, K. Teraoka, and G. L. Mechanic, Cross-linking and stereospecific structure of collagen in mineralized skeletal tissues. Connect. Tissue Res. 21, 159 – 169 (1989). K. Kuhn, The classical collagens: Type I, II and III. In “Structure and Function of Collagen Types” (R. Mayne and R. E. Burgeson, eds.), pp. 1 – 42, Academic Press, Orlando, 1987. J. Bella, M. Eaton, B. Brodsky, and H. M. Berman, Crystal and molecular structure of a collagen-like peptide at 1.9 angstroms. Science 266, 75 – 81 (1994). R. Kornfeld and S. Kornfeld, Comparative aspects of glycoprotein structure. Annu. Rev. Biochem. 45, 217 – 237 (1976). J. J. Sauk, K. Norris, J. M. Kerr, M. J. Sommerman, and M. F. Young, Diverse forms of stress result in changes in celular levels of osteonectin/SPARC witout altering mRNA levels in osteoligament cells. Calicif. Tissue Int. 49, 58 – 62 (1991). M. Weinstock and C. P. Leblond, Radioautographic visualization of the deposition of a phosphoprotein at the mineralization front in the dentin of the rat incisor. J. Cell Biol. 56, 838 – 845 (1973). L. Paglia, J. Wilczek, L. Diaz de Leon, G. R. Martin, D. Horlein, and P. Muller, Inhibition of procollagen cell-free synthesis by amino-terminal extension peptides. Biochemistry 18, 5030 – 5034 (1979). W. Schlumberger, M. Thie, H. Volmer et al., Binding and uptake of Col 1(I), a peptide capable of inhibiting collagen synthesis in fibroblasts. J. Cell Biol. 46, 244 (1988). M. Wiestner, S. Fischer, W. Dessau et al., Collagen types synthesized by isolated calvarium cells. Exp. Cell Res. 133, 115 (1981). C. H. Wu, C. B. Donovan, and G. Y. Wu, Evidence for pretranslational regulation of collagen synthesis by procollagen propeptides. J. Biol. Chem. 261, 10482 – 10484 (1989). R. S. Ayacock, R. Raghow, G. P. Striklin, J. M. Seyer, and A. H. Kang, Post transcriptional inhibition of collagen and fibronectin synthesis by a synthetic homolog of a portion of the carboxy-terminal propeptide of human type I collagen. J. Biol. Chem. 261, 14355 – 14360 (1986). J. D. Termine, A. B. Belcourt, K. M. Conn et al., Mineral and collagen-binding proteins of fetal calf bone. J. Biol. Chem. 256, 10403 (1981).
CHAPTER 4 The Biochemistry of Bone 113. L. W. Fisher, P. Gehron Robey, N. Tuross, A. S. Otsuka, D. A. Tepen, F. S. Esch, S. Shimasaki, and J. D. Termine, The Mr 24,000 phosphoprotein from developing bone is the N-terminal propeptide of the 1 chain of type I collagen. J. Biol. Chem. 262, 13457 – 13462 (1987). 114. D. R. Eyre, New biomarkers of bone resorption. J. Clin. Endocrinol. Metab. 74, 470A – 470C (1992). 115. R. A. Fleischmajer and J. S. Perlish, The role of the amino propeptide in collagen fibrillogenesis. Rheumatology 10, 91 (1986). 116. R. A. Fleischmajer, L. W. Fisher, E. D. MacDonald, L. Jacobs Jr., J. S. Perlish, and J. D. Termine, Decorin interacts with fibrillar collagen of embryonic and adult human skin. J. Struct. Biol. 106, 82 – 90 (1991). 117. J. E. Scott and M. Haigh, Proteoglycan type I collagen fibril interactions in bone and non-calcifying connective tissues. Biosci. Rep. 5, 71 – 81 (1985). 118. J. E. Scott, Proteoglycan:collagen interactions in connective tissues: Ultrastructural, biochemical, functional and evolutionary aspects. Int. J. Biol. Macromol. 13, 157 – 161 (1991). 119. J. E. Scott, Proteoglycan-fibrillar collagen interactions. Biochem. J. 252, 313 – 323 (1988). 120. E. P. Katz and S. T. Li, The intermolecular space of reconstituted collagen fibrils. J. Mol. Biol. 73, 351 – 369 (1973). 121. A. J. Hodge and J. A. Petruska, Recent studies with electron microscope on ordered aggregates of the tropocollagen macromolecule. In “Aspects of Protein Structure” (G. N. Ramachandran, ed.), pp. 289 – 300. Academic Press, New York, 1963. 122. A. L. Arsenault, Crystal-collagen relationships in calcified turkey leg tendons visualized by selected-area dark field electron microscopy. Calcif. Tissue Int. 43, 202 – 212 (1988). 123. M. E. Maitland and A. L. Arsenault, A correlation between the distribution of biological apatite and amino acid sequence of type I collagen. Calcif. Tissue Int. 48, 341 – 352 (1991). 124. M. Glimcher, Recent studies of the mineral phase in bone and its possible linkage to the organic matrix by protein-bound phosphate bonds. Phil. Trans. R. Soc. Lond. B. 304, 479 – 508 (1984). 125. W. Traub, A. Jodaikin, T. Arad, A. Veis, and B. Sabsay, Dentin phosphophoryn binding to collagen fibrils. Matrix 12, 197 – 201 (1992). 126. D. R. Eyre, M. A. Paz, and P. M. Gallop, Cross-linking in collagen and elastin. Annu. Rev. Biochem. 53, 717 – 748 (1984). 127. C. Niyibizi and D. R. Eyre, Identification of the cartilage alpha1(IX) chain in type V collagen from bovine bone. FEBS Lett. 242, 314 – 318 (1989). 128. T. Pihlajaniemi and M. Tammnen, The alpha1 chain of type XIII collagen: Polypeptide structure, alternative splicing and tissue distribution. In Structure, “Molecular Biology and Pathology of Collagen.” (R. Fleishmajer, B. R. Olsen, and K. Kuhn, eds.). New York Academy of Science. 129. M. van der Rest, Collagen of bone. In “Bone Matrix and Bone Specific Products” (B. K. Hall, ed.), pp. 187 – 239. CRC Press, Boca Raton, FL, 1991. 130. C. Niyibizi and D. R. Eyre, Bone type V collagen: Chain composition and location of a trypsin cleavage site. Connect. Tissue Res. 20, 247 – 250 (1989). 131. M. Yamagata, K. Yamada, S. S. Yamada, and T. Shinomura, The complete primary structure of type XII collagen shows a chimeric molecule with reiterated fibronectin type III motifs, von Willebrand factor A motifs, a domain homologous to a non-collagenous region of type IX collagen, and a short collagenous domain. J. Cell Biol. 115, 209 – 221 (1991). 132. S. J. Perkins, K. F. Smith, S. C. Williams, and P. I. Haris, The secondary structure of von Willebrand factor type A domains in factor B of human complement by Fourier transform infrared spectroscopy: Is occurrence in collagen types VI, VII, XII and XIV, the integrins and other proteins? J. Mol. Biol. 238, 104 – 119 (1994).
163 133. M. Koch, C. Bernasconi, and M. Chiquet, A major oligomeric fibroblast proteoglycan identified as a novel large form of type-XII collagen. Eur. J. Biochem. 207, 847 – 856 (1992). 134. S. P. Oh, C. M. Griffith, E. D. Hay, and B. R. Olsen, Tissue-specific expression of type XII collagen during mouse embryonic development. Dev. Dyn. 196, 37 – 46 (1993). 135. H. Nakahara, K. Watanabe, S. P. Sugrue, B. R. Olsen, and A. I. Caplan, Temporal and spatial distribution of type XII collagen in high cell density culture of periosteal-derived cells. Dev. Biol. 142, 481 – 485 (1991). 136. D. R. Keene, G. P. Lundstrum, M. P. Morris, D. W. Stoddard, and R. E. Burgeson, Two type XII-like collagens localize in the surface of banded collagen fibrils. J. Cell Biol. 113, 971 – 978 (1991). 137. J. Trueb and B. Trueb, Type XIV collagen is a variant of indulin. Eur. J. Biochem. 207, 549 – 557 (1993). 138. B. R. Olsen, The roles of collagen genes in skeletal development and morphogenesis. Experientia 51, 194 – 195 (1995). 139. S. A. Pinnell, R. Fox, and S. M. Krane, Human collagens, differences in glycosylated hydroxylysines in skin and bone. Biochim. Biophys. Acta 229, 119 (1971). 140. J. R. Segrest and L. W. Cunningham, Variations in human urinary O-hydroxylysyl glucoside levels and their relationship to collagen metabolism. J. Clin. Invest. 49, 1497 (1970). 141. L. Knott and A. J. Bailey, Collagen cross-links in mineralizing tissues: A review of their chemistry, function, and clinical relevance. Bone 22(3), 181 – 187 (1998). 142. A. L. Boskey, T. L. Wright, and R. D. Blank, Collagen and bone strength, J. Bone Miner. Res. 3, 330 – 335 (1999). 143. D. J. Prockop, Mutations that alter the primary structure of type I collagen: The periols of a system for generation large structures by the principle of nucleated growth. J. Biol. Chem. 265, 15349 – 15352 (1990). 144. R. J. Wenstrup, M. C. Willing, B. J. Starman, and P. H. Byers, Distinct biochemical phenotypes predict clinical severity in nonlethal variants of osteogenesis imperfecta. Am. J. Hum. Genet. 46, 975 – 982 (1990). 145. P. H. Byers, Inherited disorders of collagen gene structure and expression. Am. J. Hum. Genet. 34, 72 – 80 (1989). 146. P. H. Byers, Brittle bones – fragile molecules: Disorders of collagen gene structure and expression. Trends Genet. 6, 293 – 300 (1990). 147. H. Kuivaniemi, G. Tromp, and D. J. Prockop, Mutations in collagen genes: Causes of rare and some common diseases in humans. FASEB J. 5, 2052 – 2060 (1991). 148. J. Lohler, R. Timpl, and R. Jaenisch, Embryonic lethal mutation in mouse collagen I gene causes rupture of blood vessels and is associated with erythropoietic and mesenchymal cell death. Cell 38(2), 597 – 607 (1984). 149. K. Harbers, M. Juehan, H. Delius, and R. Jaenisch, Insertion of retrovirus into the first intron of 1(I) collagen gene leads to embryonic fetal mutation in mice. Proc. Natl. Acad. Sci. USA 81, 1504 – 1508 (1984). 150. A. Schniecke, K. Harbers, and R. Jaenisch, Embryonic lethal mutation in mice induced by retrovirus insertion into the alpha1(I) collagen gene. Nature 304, 315 – 320 (1983). 151. A. Stacey, J. Bateman, T. Choi, T. Mascara, W. Cole, and R. Jaenisch, Perinatal lethal osteogenesis imperfecta in transgenic mice bearing an engineered mutant pro-1(I) collagen gene. Nature 332, 131 – 136 (1988). 152. J. Bonadio, T. L. Saunders, E. Tsai, S. A. Goldstein, J. MorrisWiman, L. Brinkley, D. F. Dolan, R. A. Altschuler, J. E. Hawkins, J. F. Bateman, T. Maresca, and R. Jaenisch, Transgenic mouse model of the mild dominant form of osteogenesis imperfecta. Proc. Natl. Acad. Sci. USA 87, 7145 – 7149 (1990). 153. S. D. Chipman, H. O. Sweet, D. J. McBride, M. T. Davidson, S. C. Marks Jr, A. R. Schuldiner, D. W. Rowe, R. J. Wenstrup, and J. R.
164
154.
155.
156.
157.
158.
159.
160.
161.
162.
163.
164.
165. 166. 167.
168.
169.
170.
171.
172.
GOKHALE, BOSKEY, AND ROBEY Shapiro, Defective pro a 2 (I) collagen synthesis in a new recessive mutation in mice: A potential model for human osteogenesis imperfecta, Proc. Natl. Acad. Sci. USA 90, 1701 – 1705 (1993). J. Bonadio and S. A. Goldstein, Our understanding of inherited skeletal fragility and what this has taught us about bone structure and function. In “Cellular and Molecular Biology of Bone” (M Noda, ed.), pp. 169 – 189. Academic Press, San Diego, 1993. J. S. Khillan, A. E. Olsen, S. Kontusaari, B. Solov, and D. J. Prockop, Trangenic mice that express a mini-gene version of the human gene for type I procollagen (COLA1) developed a phenotype resembling a lethal form of osteogenesis imperfecta. J. Biol. Chem. 266(34), 23373 – 23379 (1991). W. de Wet, T. Pihlajaniemi, J. C. Myers, T. E. Kelly, and D. J. Prockop, Synthesis of a shortened pro-2 chain, and decreased synthesis of pro-a2 (I) chains in a proband with osteogenesis imperfecta. J. Biol. Chem. 258, 7721 – 7728 (1983). U. Vetter, E. D. Eanes, J. B. Kopp, J. D. Termine, and P. Gehron Robey, Changes in apatite crystal size in bones of patients with osteogenesis imperfecta. Calcif. Tissue Int. 49, 248 – 250 (1991). W. Traub, T. Arad, U. Vetter, and S. Weiner, Ultrastructural studies of bones from patients with osteogenesis imperfecta. Matrix Biol. 14, 337 – 345 (1994). P. Fratzl, O. Paris, K. Klaushofer, and W. J. Landis, Bone mineralization in an osteogenesis imperfecta mouse model studied by small-angle x-ray scattering. J. Clin. Invest. 97(2) 396 – 402 (1996). N. P. Camacho, L. Hou, T. R. Toledano, W. A. Ilg, C. F. Brayton, C. L. Raggio, L. Root, and A. L. Boskey, The material basis for reduced mechanical properties in oim mice bones. J. Bone. Miner. Res. 14(2), 264 – 272 (1999). W. Traub, T. Arad, and S. Weiner, Three-dimensional ordered distribution of crystals in turkey tendon collagen fibers. Proc. Natl. Acad. Sci. USA 86, 9822 – 9826 (1989). P. Sarathchandra, F. M. Pope, and S. Y. Ali, Morphometric analysis of type I collagen fibrils in the osteoid of osteogenesis imperfecta. Calcif. Tissue Int. 65(5), 390 – 395 (1999). L. Addadi and S. Weiner, Interaction between acidic proteins and crystals: Stereochemical requirements in biomineralization. Proc. Natl. Acad. Sci. USA 82, 4110 – 4114 (1985). L. W. Fisher and J. D. Termine, Noncollagenous proteins influencing the local mechansims of calcification. Clin. Orthop. 200, 362 – 365 (1985). A. L. Boskey, Mineral-matrix interactions in bone and cartilage. Clin. Orthopaed. 281, 244 – 274 (1992). W. Traub, T. Arad, and S. Weiner, Origin of mineral crystal growth in collagen fibrils. Matrix 12, 251 – 255 (1992). U. Vetter, L. W. Fisher, K. P. Mintz, J. B. Kopp, N. Tuross, J. D. Termine, and P. Gehron Robey, Osteogenesis imperfecta: Changes in noncollagenous proteins in bone. J. Bone Miner. Res. 6, 501 – 505 (1991). L. W. Fisher, M. A. Drum, P. Gehron Robey, K. M. Conn, and J. D. Termine, Osteonectin content in human osteogenesis imperfecta bone shows a range similar to that of two bovine models of OI. Calcif. Tissue Int. 40, 260 – 264 (1987). N. S. Fedarko, M. Moerike, R. Brenner, P. Gehron Robey, and U. Vetter, Extracellular matrix formation by osteoblasts from patients with osteogenesis imperfecta. J. Bone Min. Res. 7, 921 – 930 (1992). N. S. Fedarko, P. Gehron Robey, and U. K. Vetter, Extracellular matrix stoichiometry in osteoblasts from patients with osteogenesis imperfecta. J. Bone Miner. Res. 10, 1122 – 1129 (1995). A. Boyde, A. J. Banes, and R. M. Dullamans, A morphological study of an avian bone disorder caused by myeloblastosis-associated virus. Metab. Bone Dis. Rel. Res. 1, 235 (1978). A. L. Boskey and S. C. Marks, Mineral and matrix alterations in the bones of incisors-absent (ia/ia) osteopetrotic rat. Calcif. Tissue Int. 37, 287 – 292 (1985).
173. N. B. Schwartz, E. W. Pirok, III, J. R. Mensch, Jr., and M. S. Domowicz, Domain organization, genomic structure, evolution, and regulation of expression of the aggrecan gene family. Prog. Nucleic Acid Res. Mol. Biol. 62, 177 – 225 (1999). 174. K. M. Dyne, M. Valli, A. Forlino, M. Mottes, H. Kress, and G. Cetta, Deficient expression of the small proteoglycan decorin in a case of severe/lethal osteogenesis imperfecta. Am. J. Med. Genet. 53, 161 – 166 (1996). 175. N. S. Fedarko, Isolation and purification of proteoglycans. EXS 70, 9 – 35 (1994). 176. H. Kresse, H. Hausser, and E. Schinherr, Small proteoglycans. In “Proteoglycans” (P. Jolläs, ed.), pp. 73 – 100. Birkhauser Verlag, Basel, Switzerland, 1994. 177. M. Yanagishita, Function of proteoglycans in the extracellular matrix. Acta Pathol. Japonica 43, 283 – 293 (1993). 178. P. J. Neame, H. Tapp, and A. Azizan, Noncollagenous, nonproteoglycan macromolecules of cartilage. Cell. Mol. Life Sci. 55, 1327 – 1340 (1999). 179. F. Deak, D. Piecha, C. Bachrati, M. Paulsson, and I. Kiss, Primary structure and expression of matrilin-2, the closest relative of cartilage matrix protein within the von Willebrand factor type A-like module superfamily. J. Biol. Chem. 272, 9268 – 9274 (1997). 180. J. D. Sandy, C. R. Flannery, R. E. Boynton, and P. J. Neame, Isolation and characterization of disulfide-bonded peptides from the three globular domains of aggregating cartilage proteoglycan. J. Biol. Chem. 265, 21108 – 21113 (1990). 181. M. Morgelin, M. Paulsson, T. E. Hardingham, D. Heinegard, and J. Engel, Cartilage proteoglycans: Assembly with hyaluronate and link protein as studied by electron microscopy. Biochem J. 253, 175 – 185 (1988). 182. P. J. Neame and F. P. Barry, The link proteins. Experientia 49, 393 – 402 (1993). 183. T. E. Hardingham, The role of link protein in the structure of cartilage proteoglycan aggregates. Biochem. J. 177, 237 – 247 (1979). 184. L. Kjellän and U. Lindahl, Proteoglycans: Structures and interactions. Annu. Rev. Biochem. 60, 443 – 475 (1991). 185. D. Heinegard and A. Oldberg, Structure and biology of cartilage and bone matrix noncollagenous macromolecules. FASEB J. 3, 2042 – 3053 (1989). 186. H. Paulssen, R. Busch, V. Sinnwell, and A. Poller-Kruger, Conformation analysis of D-xylose-containing O-glycopeptide sequences. Carbohydr. Res. 214, 227 – 234 (1991). 187. J. E. Scott, Supramolecular organization of extracellular matrix glycosaminoglycans in vitro and in the tissues. FASEB J. 6, 2639 – 2645 (1992). 188. J. R. Korenberg, X. N. Chen, K. Doege, J. Grover, and P. J. Roughley, Assignment of the human aggrecan gene (AGC1) to 15q26 using fluorescence in situ hybridization analysis. Genomics 16(2), 546 – 548 (1993). 189. K. J. Doege, M. Sasaki, E. Horigan, J. R. Hassell, and Y. Yamada, Complete primary structure of the rat cartilage proteoglycan core protein deduced from cDNA clones. J. Biol. Chem. 262, 17757 – 17767 (1987). 190. K. J. Doege, K. Garrison, S. N. Coulter, and Y. Yamada, The structure of the rat aggrecan gene and preliminary characterization of its promoter. J. Biol. Chem. 269, 29232 – 29240 (1994). 191. D. A. Horita, P. J. Hajduk, P. F. Goetinck, and L. E. Lerner, NMR studies of peptides derived from the putative binding regions of cartilage proteins: No evidence for binding to hyaluronan. J. Biol. Chem. 269, 1699 – 1704 (1994). 192. S. J. Perkins, A. S. Nealis, D. G. Dunham, T. E. Hardingham, and H. Muir, Neutron and X-ray solution-scattering studies of the ternary complex between proteoglycan-binding region, link protein and hyaluronan. Biochem. J. 285, 263 – 268 (1992).
CHAPTER 4 The Biochemistry of Bone 193. J. A. Buckwalter and L. C. Rosenberg, Structural changes during development in bovine fetal epiphyseal cartilage. Coll. Rel. Res. 3, 489 – 504 (1983). 194. J. A. Buckwalter, P. J. Roughley, and L. C. Rosenberg, Age-related changes in cartilage proteoglycans: Quantitative electron miscroscopic studies. Microsc. Res. Tech. 28, 398 – 408 (1994). 195. W. D. Comper, and T. C. Laurent, Physiological function of connective tissue polysaccharides. Physiol. Rev. 58, 255 – 315 (1978). 196. W. D. Comper and R. P. W. Williams, Hydrodynamics of concentration proteoglycan solutions. J. Biol. Chem. 262, 13464 – 13471 (1987). 197. Nishimura, Y. Muragaki, and B. R. Olsen, Tissue-specific forms of type IX collagen-proteoglycan arise from the use of two widely separated promoters. J. Biol. Chem. 264, 20033 – 20041 (1989). 198. W. D. Comper, Physicochemical aspects of cartilage extracellular matrix. In “Cartilage: Molecular Aspects” (B. Hall and S. Newman, eds.), pp. 59 – 96. CRC Press, Boca Raton, FL, 1991. 199. G. H. Hunter, Role of proteoglycan in the provisional calcification of cartilage. Clin. Orthop. Rel. Res. 262, 256 – 263 (1991). 200. C. C. Chen, A. L. Boskey, and L. C. Rosenberg, The inhibitory effect of cartilage proteoglycans on hydroxyapatite growth. Calcif. Tissue Int. 36, 285 – 290 (1984). 201. C. C. Chen, and A. L. Boskey, Mechanisms of proteoglycan inhibition of hydroxyapatite growth. Calcif. Tissue Int. 37, 395 – 400 (1985). 202. A. L. Boskey, Hydroxyapatite formation in a dynamic gel system: Effects of type I collagen, lipids, and proteoglycans. J. Phys. Chem. 93, 1628 – 1633 (1989). 203. D. E. Eanes, Mixed phospholipid liposome calcification. Bone Miner. 17, 269 – 272 (1992). 204. G. K. Hunter , An ion-exchange mechanism of cartilage calcification. Connect. Tissue Res. 16, 111 – 120 (1987). 205. A. L. Boskey, M. Maresca, A. L. Armstrong, and M. G. Ehrlich, Treatment of proteoglycan aggregates with growth plate neutral metalloprotease reduces their ability to inhibit hydroxyapatite proliferation in a gelatin gel. J. Orthopaed. Res. 10, 313 – 319 (1992). 206. D. E. Eanes and A. W. Hailer, Effect of ultrafilterable fragments from chondroitinase and protease-treated aggrecan on calcium phosphate precipitation in liposomal suspensions. Calcif. Tissue Int. 55, 176 – 179 (1994). 207. K. Kimata, H. Barrach, K. Brown, and J. P. Pennypacker, Absence of proteoglycan core protein in cartilage from the cmd/cmd (cartilage matrix deficiency) mouse, J. Biol. Chem. 256, 6961 – 6968 (1981). 208. A. L. Boskey, M. Maresca, and A. Hjerpe, Hydroxyapatite formation in the presence of proteoglycans of reduced sulfate content: Studies in the brachymorphic mouse. Calcif. Tissue Int. 49, 389 – 393 (1991). 209. W. S. Argraves, P. J. McKeown-Longo, and P. F. Goetinck, Absence of proteoglycan core protein in the cartilage mutant nanomelia. FEBS Lett. 131, 265 – 268 (1981). 210. S. Weinbaum, S. C. Cowin, and Y. Zeng, A model for the excitation of osteocytes by mechanical loading-induced bone fluid shear stresses. J. Biomech. 27, 339 – 360 (1994). 211. L. Y. Yao, C. Moody, E. Schonherr, T. N. Wight, and L. J. Sandell, Identification of the proteoglycan versican in aorta and smooth muscle cells by DNA sequence analysis, in situ hybridization and immunohistochemistry. Matrix Biol. 14, 213 – 225 (1994). 212. G. J. Thomas, M. T. Bayliss, K. Harper, R. M. Mason, and M. Davies, Glomerular mesangial cells in vitro synthesize an aggregating proteoglycan immunologically related to versican, Biochem. J. 302, 49 – 56 (1994). 213. T. J. Crawford, I. G. Melhado, and F. R. Jirik, Expression of versican mRNA is developmentally regulated in the brain of the embryonic chick and the developing rat. Brain Res. Dev. Brain Res. 76, 264 – 267 (1993).
165 214. A. Bignami, G. Perides, and F. Rahemtulla, Versican, a hyaluronate-binding proteoglycan of embryonal precartilaginous mesenchyma, is mainly expressed postnatally in rat brain. J. Neurosci. Res. 34, 97 – 106 (1993). 215. D. R. Zimmerman, M. T. Dours-Zimmerman, M. Schubert, and L. Bruckner-Tuderman, Versican is expressed in the proliferating zone in the epidermis and in association with the elastic network of the dermis. J. Cell Biol. 124, 817 – 825 (1994). 216. L. W. Fisher, The nature of the proteoglycans of bone. In “The Chemistry and Biology of Mineralized Tissues” (W. T. Butler, ed.), pp. 188 – 196. EBSCO Media, Birmingham, AL, 1985. 217. K. Ito, T. Shinomura, M. Zako, M. Ujita, and K. Kimata, Multiple forms of mouse PG-M, a large chondroitin sulfate proteoglycan generated by alternative splicing. J. Biol. Chem. 270, 958 – 965 (1995). 218. M. Yamagata, T. Shinomura, and K. Kimata, Tissue variation of two large chondroitin sulfate proteoglycans (PG-M/versican and PG-H/aggrecan) in chick embryos. Anat. Embryol. 187, 433 – 444 (1993). 219. R. V. Iozzo, M. F. Naso, L. A. Cannizzaro, J. J. Wasmuth, and J. D. McPherson, Mapping of the versican proteoglycan gene (CSPG2) to the long arm of human chromosome 5 (5q12 – 5q14). Genomics 14, 845 – 851 (1992). 220. M. F. Naso, D. R. Zimmermann, and R. V. Iozzo, Characterization of the complete genomic structure of the human versican gene and functional analysis of its promoter. J. Biol. Chem. 269, 32999 – 33008 (1994). 221. M. T. Dours-Zimmermann and D. R. Zimmermann, A novel glycosaminoglycan attachment domain identified in two alternative splice variants of human versican. J. Biol. Chem. 269, 32992 – 32998 (1994). 222. D. R. Zimmermann and E. Ruoslahti, Multiple domains of the large fibroblast proteoglycan, versican, EMBO J. 8; 2575 – 2588 (1990). 223. E. Schonherr, H. T. Jarvelainen, L. J. Sandell, and T. Wight, Effects of platelet-derived growth factor and transforming growth factorbeta 1 on the synthesis of a large versican-like chondroitin sulfate proteoglycan by arterial smooth muscle cells. J. Biol. Chem. 266, 17640 – 17647 (1991). 224. V. M. Kahari, H. Larjava, and J. Uitto, Differential regulation of extracelular matrix proteoglycan (PG) gene expression: Transforming growth factor-beta 1 up-regulates biglycan (PGI), and versican (large fibroblast PG) but down-regulates decorin (PGII) mRNA levels in human fibroblasts in culture. J. Biol. Chem. 266, 10608 – 10615 (1991). 225. E. E. Qwarnstrom, H. T. Jarvelainen, M. G. Kinsella, C. O. Ostberg, L. J. Sandell, R. C. Page, and T. N. Wight, Interleukin-1 beta regulation of fibroblast proteoglycan synthesis involves a decrease in versican steady-state mRNA levels. Biochem. J. 294, 613 – 620 (1993). 226. G. Perides, F. Rahemtulla, W. S. Lane, R. A. Asher, and A. Bignami, Isolation of a large aggregating proteoglycan from human brain. J. Biol. Chem. 267, 23883 – 23887 (1992). 227. J. Engel, Domains in proteins and proteoglycans of the extracellular matrix with functions in assembly and cellular activities. Int. J. Biol. Macromol. 13, 147 – 151 (1991). 228. M. E. Antosz, C. G. Bellows, and J. E. Aubin, Biphasic effects of epidermal growth factor on bone nodule formation by isolated rat calvaria cells in vitro. J. Bone Miner. Res. 2, 385 – 393 (1987). 229. N. Matsushima, T. Ohyanagi, T. Tanaka, and R. H. Kretsinger, Super-motifs and evolution of tandem leucine-rich repeats within the small proteoglycans: Biglycan, decorin, lumican, fibromodulin, PRELP, keratocan, osteoadherin, epiphycan, and osteoglycin. Proteins 38, 210 – 225 (2000). 230. H. Kresse, H. Hausser, E. Schonherr, and K. Bittner, Biosynthesis and interactions of small chondroitin/dermatan sulphate proteoglycans. Eur. J. Clin. Chem. Clin. Biochem. 32, 259 – 254 (1994).
166 231. D. S. Howell, J. C. Pita, J. F. Marquez, and J. E. Madruga, Partition of calcium, phosphate and protein in the fluid phase aspirated at calcifying sites in epiphyseal cartilage. J. Clin. Invest. 47, 1121 – 1132 (1968). 232. J. Klein-Nulend, P. Veldhuijzen, R. J. van de Stadt et al., Influence of intermittent compressive force on proteoglycan content in calcifying growth plate cartilage in vitro. J. Biol. Chem. 262, 15490 (1987). 233. C. W. Prince, F. Rahemtulla, and W. T. Butler, Metabolism of rat bone proteoglycans in vivo. Biochem. J. 216, 589 – 596 (1983). 234. C. W. Prince, F. Rahemtulla, and W. T. Butler, Incorporation of [35S]sulphate into glycosaminoglycans by mineralized tissues in vivo. Biochem. J. 224, 941 – 945 (1984). 235. L. W. Fisher, The nature of proteoglycans of bone. In “The Chemistry and Biology of Mineralized Tissues” (W. T. Butler, ed.), pp. 188 – 196, EBSCO Media, Birmingham, AL, 1985. 236. B. Kobe and J. Deisnhofer, Crystal structure of porcine ribonuclease inhibitor, a protein with leucine-rich repeats. Nature 366, 751 – 756 (1993). 237. C. Branden and J. Tooze , Motifs of protein structure. In “Introduction to Protein Structure,” Chap. 2, pp. 11 – 32. Garland Publishing, New York, 1991. 238. L. W. Fisher, J. D. Termine, and M. F. Young, Deduced protein sequence of bone small proteoglcyan I (biglycan) shows homology with proteoglcyan II (decorin) and several non-connective tissue proteins in a variety of species. J. Biol. Chem. 264, 4571 – 4576 (1989). 239. G. A. Pringle and C. M. Dodd, Immunoelectron microscopy localization of the core protein of decorin near the d and e bands of tendon collagen fibrils by use of monoclonal antibodies. J. Histochem. Cytochem. 38, 1405 – 1411 (1990). 240. D. Simonescu, R. V. Iozzo, and N. A. Kefalides, Bovine pericardial proteoglycan: Biochemical, immunochemical and ultrastructural studies. Matrix 9, 301 – 310 (1989). 241. L. W. Fisher, J. D. Termine, S. W. Dejter, S. W. Whitson, M. Yanagishita, J. H. Kimura, V. C. Hascall, H. K. Kleinman, J. R. Hassell, and B. Nilson, Proteoglycans of developing bone. J. Biol. Chem. 258, 6588 – 6594 (1983). 242. H. U. Choi, T. L. Johnson, P. Subhash, L.-H. Tang, and L. Rosenberg, Characterization of the dermatan sulfate proteoglycans, DSPGI and DS-PGII, from bovine articular cartilage and skin isolated by octyl-spharose chromatography. J. Biol. Chem. 264, 2876 – 2884 (1989). 243. P. Bianco , L. W. Fisher, M. F. Young, J. D. Termine, and P. Gehron Robey, Expression and localization of the two small proteoglycans biglycan and decorin in developing human skeletal and non-skeletal tissues. J. Histochem. Cytochem. 38, 1549 – 1563 (1990). 244. B. Johnstone, M. Markopoulos, P. Neame, and B. Caterson, Identification and characterization of glycanated and non-glycanated forms of biglycan and decorin in the human intervertebral disc. Biochem. J. 292, 661 – 666 (1993). 245. L. W. Fisher, G. R. Hawkins, N. Tuross, and J. D. Termine, Purification and partial characterization of small proteoglycans I and II, bone sialoproteins I and II, and osteonectin from the mineral compartment of developing human bone. J. Biol. Chem. 262, 9702 – 9708 (1987). 246. D. D. Krantz, R. Ziovetski, B. L. Kagan, and S. L. Ziporsky, Ampipathic -structure of a leucine-rich repeat peptide. J. Biol. Chem. 266, 16901 – 16907 (1991). 247. P. Krishnan, A. M. Hocking, J. M. Scholtz, C. N. Pace, K. K. Holik, and D. J. McQuillan, Distinct secondary structures of the leucinerich repeat proteoglycans decorin and biglycan: Glycosylation-dependent conformational stability. J. Biol. Chem. 274(16), 10945 – 10950 (1999). 248. O. W. McBride, L. W. Fisher, and M. F. Young, Localization of PGI (biglycan, BGN) and PGII (decorin, DCN, PG-40) genes on human
GOKHALE, BOSKEY, AND ROBEY
249.
250.
251.
252.
253.
254.
255.
256.
257.
258.
259.
260.
261.
262.
263.
264.
265.
chromosomes Xq13-qter and 12q, respectively. Genomics 6(2); 219 – 225 (1990). K. G. Danielson, A. Fazzio, I. Cohen, L. A. Cannizzaro, I. Eichstetter, and R. V. Iozzo, The human decorin gene: Intron-exon organization, discovery of two alternatively spliced exons in the 5 untranslated region, and mapping of the gene to chromosome 12q23. Genomics 15, 146 – 160 (1993). U. Vetter, W. Vogel, W. Just, M. F. Young, and L. W. Fisher, Human decorin gene: Intron-exon junctions and chromosomal localization, Genomics 15, 161 – 168 (1993). T. Scholzen, M. Solursh, S. Suzuki, R. Reiter, J. L. Morgan, A. M. Buchberg, L. D. Siracusa, and R. V. Iozzo, The murine decorin, complete cDNA cloning genomic organization, chromosomal assignment, and expression during organogenesis and tissue differentiation. J. Biol. Chem. 269, 28270 – 28281 (1994). M. D. Briggs and D. H. Cohn, Length polymorphism within a complex dinucleotide repeat in the human decorin (DCN) gene. Hum. Mol. Genet. 2, 1087 (1993). M. Santra, K. G. Danielson, and R. V. Iozzo, Structural and functional characterization of the human decorin gene promoter: A homopurine-homopyrimidine S1 nuclease-sensitive region is involved in transcriptional control. J. Biol. Chem. 269, 579 – 587 (1994). T. Krusius, K. R. Gehlsen, and E. A. Ruoslahti, fibroblast chondroitin sulfate proteoglycan core proteincontains lectin-like and growth factor-like sequences. J. Biol. Chem. 262; 13120 – 13125 (1987). A. A. Day, C. I. McQuillan, J. D. Termine, and M. R. Young, Molecular cloning and sequence analysis of the cDNA for small proteoglycan II of bovine bone. Biochem. J. 248, 801 – 805 (1987). A. Bassols and J. Massague, Transforming growth factor- regulates the expression and structure of extracellular matrix chondroitin/dermatan sulfate proteoglycans. J. Biol. Chem. 263, 3039 – 3045 (1988). D. M. Rosen, S. A. Stempien, A. Y. Thompson, and S. M. Seyedin, Transforming growth factor-beta modulates the expression of osteoblast and chondroblast phenotypes in vitro. J. Cell Physiol. 134, 337 – 346 (1988). Y. Takeuchi, T. Matsumoto, E. Ogata, and Y. Shishiba, Effects of transforming growth factor 1 and L-ascorbate on synthesis and distribution of proteoglycans in murine osteoblast-like cells, J. Bone Miner. Res. 8, 823 – 829 (1993). S. Kimoto, S. L. Cheng, S. F. Zhang, and L. V. Avioli, The effect of glucocorticoid on the synthesis of biglycan and decorin in human osteoblasts and bone marrow stromal cells. Endocrinology 135(6); 2423 – 2431 (1994). S. L. Cheng, S. F. Zhang, T. L. Nelson, P. M. Warlow, and R. Civitelli, Stimulation of human osteoblast differentiation and function by ipriflavone and its metabolites. Calcif. Tissue Int. 55, 356 – 362 (1994). N. A. Visser, G. P. Vankampen, M. H. Dekoning, and J. K. Vanderkorst, The effects of loading on the synthesis of biglycan and decorin in intact mature articular cartilage in vitro. Connect. Tissue Res. 30, 241 – 250 (1994). R. S. Sawhney, T. M. Hering, and L. J. Sandell, Biosynthesis of small proteoglycan II (decorin) by chondrocytes and evidence for a procore protein. J. Biol. Chem. 266, 9231 – 9240 (1991). K. G. Vogel and J. A. Trotter, The effect of proteoglycans on the morphology of collagen fibrils formed in vitro. Collagen Rel. Res. 7, 105 – 114 (1987). D. J. Bidanset, C. Guidry, L. C. Rosenberg, H. Choi, R. Timpl, and M. Hook, Binding of the proteoglycan decorin to collagen type VI. J. Biol. Chem. 267, 5250 – 5256 (1992). B. Merle, H. V. Choi, L. Rosenberg, J. Lawler, P. D. Delmas, and P. Clezardin, In vitro interaction between decorin and thrombospondin. In “FECTS XIV Meeting,” p. F13. Lyon, France, 1994.
CHAPTER 4 The Biochemistry of Bone 266. Y. Yamaguchi, D. M. Mann, and E. Rouslahti, Negative regulation of transforming growth factor- by the proteoglycan decorin. Nature 346, 281 – 284 (1990). 267. G. Schmidt, H. Hausser, and H. Kresse, Interaction of the small proteoglycan decorin with fibronectin. Biochem. J. 266, 591 – 595 (1990). 268. M. Winnermoller, G. Schmidt, and H. Kresse, Influence of decorin on fibroblast adhesion to fibronectin. Eur J Cell Biol. 54, 10 – 17 (1991). 269. K. G. Vogel, M. Paulsson, and D. Heinegard, Specific inhibition of type I and type II collagen fibrillogenesis by the small proteoglycan of tendon. Biochem. J. 223, 587 – 597 (1984). 270. K. G. Vogel and L. W. Fisher, Comparisons of antibody reactivity and enzyme sensitivity between small proteoglycans from bovine tendon, bone, and cartilage. J. Biol. Chem. 261, 11334 – 11340 (1986). 271. D. E. Ashurst and Y. S. Bland, The collagens of the soft tissues during development in the knee and hip joints of the rabbit: Their synthesis and distribution. In “FECTS XIV Meeting,” p. 4H. Lyon, France, 1994. 272. D. V. Brown and K. C. Vogel, Characteristics of the in vitro interaction of a small proteoglycan (PGII) of bovine tendon with type I collagen. Matrix 9, 468 – 478 (1989). 273. A. L. Boskey, L. Spevak, and L. C. Rosenberg, Effects of bone CSproteoglycans, DS-decorin, and DS-biglycan on hydroxyapatite formation in a gelatin gel, Calci. Tissue Int. 61(4); 298 – 305 (1997). 274. A. L. Boskey and L. Spevak, The effects of decorin and biglycan on in vitro apatite formation in a gelatin gel. In “FECTS XIV Meeting,” p. N6. Lyon, France (1994). 275. L. Rosenberg, H. U. Choi, L. Tang, T. Johnson, J. Liu, and T. M. Lave, Metal-binding properties of biglycan. Trans 40th ORS 19, 5 (1994). 276. P. M. Steinlen, E. Maessen, J. Kresse, I. M. Van Vlijmen et al., Expression of tenascin, biglycan and decorin in disorders of keratinization. Br. J. Dermatol. 130, 564 – 568 (1994). 277. E. Rouslahti and Y. Yamaguchi, Proteoglycans as modulators of growth factor activities. Cell 64, 867 – 869 (1991). 278. K. M. Dyne, M. Valli, M. Forlino, M. Mottes, and G. Cetta, Decreased decorin expression in patients with osteogenesis imperfecta. In “FECTS Meeting,” p. E6. Lyon, France, 1994. 279. Y. Takeuchi, Y. Kodama, and T. Matsumoto, Bone matrix decorin binds transforming growth factor- and enhances its bioactivity. J. Biol. Chem. 51, 32634 – 32638 (1994). 280. T. Krusius and E. Ruoslahti, Primary structure of an extracellular matrix proteoglycan coreprotein deduced from cloned cDNA. Proc. Natl. Acad. Sci. USA. 83, 7683 – 7687 (1986). 281. N. S. Fedarko, J. D. Termine, and P. Gehron Robey, Temporal sequence of proteoglycan biosynthesis by normal human bone cell in vitro. J. Biol. Chem. 265, 12200 – 12209 (1990). 282. K. Ibaraki, S. W. Whitson, J. D. Termine, and M. F. Young, Bone matrix mRNA expression in differentiating fetal bovine osteoblasts. J. Bone Miner. Res. 7, 743 – 754 (1992). 283. P. G. Neame, H. C. Choi, and L. C. Rosenberg, The primary structure of the core protein of the small leucine-rich proteoglycan (PGI) from bovine articular cartilage. J. Biol. Chem. 264, 8653 – 8661 (1989). 284. J. Liu, T. M. Laue, H. U. Choi, L.-H. Tang, and L. Rosenberg, The self-association of biglycan from bovine articular cartilage. J. Biol. Chem. 269, 28366 – 28373 (1994). 285. L. W. Fisher, A. M. Heegaard, U. Vetter, W. Vogel, W. Just, J. D. Termine, and M. F. Young, Human biglycan gene. J. Biol. Chem. 266, 14371 – 14377 (1991). 286. V. Asundi, K. Cowan, D. Matzura, W. Wagner, and K. L. Dreher, Characterization of extracellular matrix proteoglycan transcripts expressed by vascular smooth muscle cells. Eur. J. Cell Biol. 52, 98 – 104 (1990).
167 287. A. M. Heegaard, P. Gehron Robey, W. Vogel, W. Just, R. L. Widom, J. Scholler, J. Fisher, and M. F. Young, Functional characterization of the human biglycan 5-flanking DNA and binding of the transcription factor c-Krox. J. Bone Miner. Res. 12, 2050 – 2060 (1997). 288. D. Thiebaud, H. L. Guenther, A. Porret, P. Burckhardt, H. Fleisch, and W. Hofstetter, Regulation of collagen type I and biglycan mRNA levels by hormones and growth factors in normal and immortalized osteoblastic cell lines. J. Bone Miner. Res. 9, 1347 – 1354 (1994). 289. D. Pearson and J. Sasse, Differential regulation of biglycan and decorin by retinoic acidin bovine chondrocytes. J. Biol. Chem. 267, 25364 – 25370 (1992). 290. P. Gehron Robey, N. S. Fedarko, W. Lindner et al., Biosynthesis of a small protoeglycan (PG-I) by human bone cells. J. Cell Biol. 105, 299a (1987). 291. S. Kimoto, S.-L. Cheng, S.-F. Zhang, and L. V. Avioli, The effect of glucocorticoid on the synthesis of biglycan and decorin in human osteoblasts and bone marrow stromal cells. Endocrinology 135, 2423 – 2431 (1994). 292. R. J. Waddington, and M. S. Langley, Structural analysis of proteoglycans synthesized by mineraling bone cells in vitro in the presence of fluoride. Matrix Biol. 17, 255 – 268 (1998). 293. H. C. Whinna, H. U. Choi, L. C. Rosenberg, and F. C. Church, Interaction of heparin cofactor II with biglycan and decorin. J. Biol. Chem. 268, 3920 – 3924 (1993). 294. L. Rosenberg and E. B. Hunziker, Induction of repair in partial thickness articular cartilage lesions by timed release of TGF. In “ORS 40th Annual Mtg,” pp. 236 – 241, 1994. 295. G. Pogany, D. J. Hernandez, and K. G. Vogel, The in vitro interaction of proteoglycans with type I collagen is modulated by phosphate. Arch. Biochem. Biophys. 313, 102 – 111 (1994). 296. U. Vetter, N. S. Fedarko, M. F. Young, J. D. Termine, W. Just, W. Vogel, and P. Gehron Robey, Biglycan synthesis in fibroblasts of patients with Turner’s syndrome and other X-chromosome abnormalities. J. Bone Min. Res. 6, S133 (1991). 297. T. Xu, P. Bianco, L. W. Fisher, G. Longenecker, E. Smith, S. Goldstein, J. Bonadio, A. Boskey, A. M. Heegaard, B. Sommer, K. Satomura, P. Dominguez, C. Zhao, A. B. Kulkarni, P. G. Robey, and M. F. Young, Targeted disruption of the biglycan gene leads to an osteoporosis-like phenotype in mice. Nature Genet. 20, 78 – 82 (1998). 298. G. Schmidt, H. Hausser, and H. Kresse, Interaction of the small proteoglycan decorin with fibronectin: Involvement of the sequence NKISK of the core protein. Biochem. J. 280, 411 – 414 (1991). 299. D. J. Bidanset, R. LeBaron, L. Rosenberg, J. E. Murphy-Ullrich, and M. Hook, Regulation of cell substrate adhesion: Effects of small galactosaminoglycan-containing proteoglycans. J. Cell Biol. 118, 1523 – 1531 (1992). 300. P. Gehron Robey, W. J. Grzesik, The biochemistry of bone forming cells: Cell matrix interactions. In “Proceedings of the International Conference on Tooth Eruption, Resorption, and Replacement by Implants” (Z. Davidovitch, ed.), pp. 167 – 172. EBSCO Press, Birmingham, AL, 1994. 301. I. Kamo, A. Kikuchi, I. Nonaka, E. Yamada, and J. Kondo, Haemopoietic activity associated with biglycan like proteoglycan. Biochem. Biophys. Res. Commun. 195, 1119 – 1126 (1993). 302. H. Greiling, Structure and biological functions of keratan sulfate proteoglycans. EXS 70, 101 – 102 (1994). 303. S. Ekman, and D. Heinegard, Immunohistochemical localization of matrix proteins in the femoral joint cartilage of growing commercial pigs. Vet. Pathol. 29(6), 514 – 520 (1992). 304. H. Hedlund, S. Mengarelli-Widholm, D. Heinegard, F. P. Reinhold, and O. Svensson, Fibromodulin distribution and association with collagen. Matrix Biol. 14, 227 – 232 (1994).
168 305. A. Hulth, O. Johnell, L. Lindberg, and D. Heinegard, Sequential appearance of macromolecules in bone induction in the rat. J. Orthop. Res. 11, 367 – 378 (1993). 306. L. Svensson, A. Aszodi, F. P. Reinholt, R. Fassler, D. Heinegard, and A. Oldberg, Fibromodulin-null mice have abnormal collagen fibrils, tissue organization, and altered lumican deposition in tendon. J. Biol. Chem. 274, 9636 – 9647 (1999). 307. R. Sztrolovics, M. Alini, J. S. Mort, and P. J. Roughley, Age-related changes in fibromodulin and lumican in human intervertebral discs. Spine 24, 1765 – 1771 (1999). 308. P. Antonsson, D. Heinegard, and A. Oldberg, Structure and deduced amino acid sequence of the human fibromodulin gene. Biochim. Biophys. Acta. 1174, 204 – 206 (1993). 309. E. Hedbom and D. Heinegård, Binding of fibromodulin and decorin to separate sites on fibrillar collagens. J. Biol. Chem. 268, 27307 – 27312 (1993). 310. D. Fukushima, R. Batzow, A. Hildebrand, and E. Ruoslahti, Localization of transforming growth factor beta binding site in betaglycan: Comparison with small extracellular matrix proteoglycans. J. Biol. Chem. 28, 22710 – 22715 (1993). 311. M. Bernfield, R. Kokenyesi, M. Kato, M. T. Hinkes, J. Spring, R. L. Gallo, and E. J. Lose, Biology of the syndecans: A family of transmembrane heparan sulfate proteoglycans. Annu. Rev. Cell Biol. 8, 365 – 393 (1992). 312. J. N. Beresford, N. S. Fedarko, L. W. Fisher, R. J. Midura, M. Yanagishita, J. D. Termaine, and P. Gehron Robey, Analysis of the proteoglycans synthesized by human bone cells in vitro. J. Biol. Chem. 262, 17164 – 17172 (1987). 313. J. L. Andres, K. Stanley, S. Cheifetz, and J. Massague, Membraneanchored and soluble forms of betaglycan, a polymorphic proteoglycan that binds transforming growth factor-beta. J. Cell Biol. 109(6 Pt 1), 3137 – 3145 (1989). 314. P. R. Segarini and S. M. Seyedin, The high molecular weight receptor to transforming growth factor-beta contains glycosaminoglycan chains. J. Biol. Chem. 263(17), 8366 – 8370 (1988). 315. M. Klagsbrun, The affinity of fibroblast growth factors (FGFs) for heparin; FGF-heparan sulfate interactions in cells and extracellular matrix. Curr. Opin. Cell Biol. 2(5), 857 – 863 (1990). 316. T. E. Hardingham and A. J. Fosang, Proteoglycans: Many forms and many functions. FASEB J. 6, 861 – 870 (1992). 317. T. Nagata, H. A. Goldberg, Q. Zhang, C. Domenicucci, and J. Sodek, Biosynthesis of bone proteins by fetal porcine calvariae in vitro: Rapid association of sulfated sialoproteins (secreted phosphoprotein-1 and bone sialoprotein) and chondroitin sulfate proteoglycan (CS-PGIII) with bone mineral. Matrix 11, 86 – 100 (1991). 318. L. G. Hashimoto, B. Caterson, and M. Yamauchi, EDTA-insoluble, calcium-binding proteoglycans in bone. In “FECTS XIV Meeting,” p. N10. Lyon, France, 1994. 319. H. Bentz, A. Y. Thompson, R. Armstrong, R. J. Chang, K. A. Piez, and D. M. Rosen, Transforming growth factor-beta 2 enhances the osteoinductive activity of a bovine bone-derived fraction containing bone matrix morphogenetic protein 2 and 3. Matrix 11, 269 – 275 (1991). 320. M. Ujita, T. Shinomura, and K. Kimata, Molecular cloning of the mouse osteoglycin-encoding gene. Gene 158, 237 – 240 (1995). 321. T. Shinomura, and K. Kimata, Proteoglycan-Lb, a small dermatan sulfate proteoglycan expressed in embryonic chick epiphyseal cartilage, is structurally related to osteoinductive factor. J. Biol. Chem. 267, 1265 – 1270 (1992). 322. K. Kurita, T. Shinomura, M. Ujita, M. Zako, D. Kida, H. Iwata, and K. Kimata, Occurrence of PG-Lb, a leucine-rich small chondroitin/dermatan sulphate proteoglycan in mammalian epiphyseal cartilage: molecular cloning and sequence analysis of the mouse cDNA. Biochem. J. 318, 909 – 914 (1996). 323. H. J. Johnson, L. Rosenberg, H. U. Choi, S. Garza, M. Hook, and P. J. Neame, Characterization of epiphycan, a small proteoglycan with
GOKHALE, BOSKEY, AND ROBEY
324.
325.
326. 327.
328.
329.
330.
331. 332.
333.
334.
335. 336. 337.
338. 339.
340.
341. 342.
343.
344.
a leucine-rich repeat core protein. J. Biol. Chem. 272, 18709 – 18717 (1997). Y. Iwata, T. Shinomura, K. Kurita, M. Zako, and K. Kimata, The gene structure and organization of mouse PG-Lb, a small chondroitin/dermatan sulphate proteoglycan. Biochem. J. 331, 959 – 964 (1998). A. Bosse, H. Kresse, K. Schwarz, and K. M. Muller, Immunohistochemical characterization of the small proteoglycans decorin and proteoglycan-100 in heterotopic ossification. Calcif. Tissue Int. 54, 119 – 124 (1994). B. P. Toole, “Cell Biology of Extracellular Matrix,” (E. D. Hay, ed.), New York, 1991. S. O. Hjertquist and L. Veljens, The glycosaminoglycans of dog compact bone and epiphyseal cartilage in the normal state and in hyper parathyroidism. Calcif. Tissue Res. 2, 314 – 333 (1968). C. B. Knudson and W. Knudson, Hyaluronan-binding proteins in development, tissue homeostasis, and disease. FASEB J. 7, 1233 – 1241 (1993). N. S. Fedarko, J. D. Termine, and P. Gehron Robey, High-performance liquid chromatographic separation of hyaluronan and four proteoglycans produced by human bone cell cultures. Anal Biochem. 188, 398 – 407 (1990). S. J. Green, G. Tarone, and C. B. Underhill, Distribution of hyaluronate and hyaluronate receptors in the adult lung. J. Cell Sci. 89, 145 – 156 (1988). C. B. Knudson, Hyaluronan receptor-directed assembly of chondrocyte pericellular matrix. J. Cell Biol. 120, 825 – 834 (1993). P. Paskalakis, D. H. Vynios, C. P. Tsiganos, E. Dalas, C. Maniatis, and P. G. Koutsoukos, Effect of proteoglycans on hydroxyapatite growth in vitro: The role of hyaluronan. Biochim. Biophys. Acta. 1158, 129 – 136 (1993). A. L. Boskey and B. L. Dick. Hyaluronan interactions with hydroxyapatite do not alter in vitro hydroxyapatite crystal proliferation and growth, Matrix 11, 442 – 446 (1991). P. Paschalakis, D. H. Vynios, C. P. Tsiganos, E. Dalas, C. Maniatis, and P. G. Koutsoukos, Effect of proteoglycans on hydroxyapatite growth in vitro: The role of hyaluronan. Biochim. Biophys. Acta. 1158, 129 – 136 (1993). A. J. Day, The structure and regulation of hyaluronan-binding proteins. Biochem. Soc. Trans. 27, 115 – 121 (1999). V. C. Hascall, Interaction of cartilage proteoglycans with hyaluronic acid. J. Supramol. Struct. 7, 101 – 120 (1977). J. Koda, A. Rapraeger, and M. Bernfield, Heparan sulfate proteoglycans from mouse mammary epithelial cells: Cell surface proteoglycan as a receptor for interstitial collagens. J. Biol. Chem. 260, 8157 – 8162 (1985). B. Toole, Developmental role of hyaluronate. Connect. Tissue Res. 10, 83 – 100 (1982). G. Oster, J. Murray, and P. Maini, A model for chondrogenic condensations in the developing limb: The role of extracellular matrix and cell tractions. J. Embryol. Exp. Morphol. 89, 93 – 112 (1985). I. Binderman, R. M. Green, and J. P. Pennypacker, Calcification of differentiating skeletal mesenchyme in vitro. Science 206, 222 – 225 (1979). B. P. Toole, G. Jackson, and J. Gross, Hyaluronate in morphogenesis in vitro. Proc. Natl. Acad. Sci. USA 69, 1384 – 1389 (1972). L. A. Goldstein, D. F. H. Zhou, L. J. Picker, C. N. Minty, R. F. Bargatze, J. F. Ding, and E. L. Butcher, A human lymphocyte homing receptor, the hermes antigen, is related to cartilage proteoglycan and link proteins. Cell 56, 1063 – 1072 (1984). J. Lesley and R. Hyman, CD44 can be activated to function as an hyaluronic acid receptor in normal murine T cells. Eur. J. Immunol. 22, 2719 – 2723 (1992). P. H. Weigel, G. M. Fuller, and R. D. Le Beof, A model for the role of the hyaluronic acid and fibrin in the early events during the
CHAPTER 4 The Biochemistry of Bone
345.
346.
347. 348.
349.
350.
351.
352.
353.
354.
355.
356.
357.
358.
359.
360.
361.
inflammatory response and wound healing. J. Theor. Biol. 119, 219 – 228 (1986). C. Knudson and B. Toole, Changes in the pericellular matrix during differentiation of limb bud mesoderm. Dev. Biol. 112, 308 – 318 (1985). B. A. Mast, R. F. Diegelmann, T. M. Krummel, and K. Cohen, Hyaluronic acid modulates proliferation, collagen and protein synthesis of cultured fetal fibroblasts. Matrix 13, 441 – 446 (1993). J. J. Minguell, Is hyaluronic acid the organizer of the extracellular matrix in marrow stroma? Exp. Hematol. 21, 7 – 8 (1993). M. Siczkowski, T. Andrew, S. Amos, and M. Y. Gordon, Hyaluronic acid regulates the function and distribution of sulfated glycosaminoglycans in bone marrow stromal cultures. Exp. Hematol. 21, 126 – 130 (1993). A. Engstrom-Laurent, Changes in hyaluronan concentrationin tissues and body fluids in disease states. Ciba Found. Symp. 143, 233 – 247 (1989). P. A. G. McCourt, B. Ek, N. Forsberg, and S. Gustafson, Intercellular adhesion molecule-1 is a cell surface receptor for hyaluronan. J. Biol. Chem. 269, 30081 – 30084 (1994). N. S. Fedarko, P. Bianco, U. K. Vetter, and P. Gehron Robey, Human bone cell enzyme expression and cellular heterogeneity: Correlation of alkaline phosphatase enzyme activity with cell cycle. J. Cell Physiol. 144, 115 – 121 (1990). B. de Bernard, P. Bianco, E. Bonucci, M. Costantini, G. C. Lunazzi, P. Martinuzzi, C. Modricky, L. Moro, E. Panfili, and P. Pollesello, Biochemical and immunohistochemical evidence that in cartilage an alkaline phosphatase is a Ca-binding glycoprotein, J. Cell Biol. 103, 1615 – 1623 (1986). P. Bianco, M. Riminucci, G. Silvestrini, E. Bonucci, J. D. Termine, L. W. Fisher, and P. Gehron Robey, Localization of bone sialoprotein (BSP) to golgi and post-golgi secretory structures in osteoblasts and to discrete sites in early bone matrix. J. Histochem. Cytochem. 41, 193 – 203 (1993). P. Bianco and A. Boyde, Alkaline phosphatase cytochemistry in confocal scanning light microcopy for imagining the bone marrow stroma. Basic Appl. Histochem. 33, 17 – 23 (1989). P. Bianco, N. S. Fedarko, E. Bonucci, J. D. Termine, and P. Gehron Robey, Acid and alkaline phosphatase activities of bone cells revisited. In “Proceedings of the Int. Soc. Bioanaloging Skeletal Implants,” pp. 17 – 27, 1991. P. Bianco, P. Ballanti, S. Mazzaferro, G. Coen, and E. Bonucci, Combined use of tetracycline fluorescence and alkaline phosphatase cytochemisry in the evaluation of osteoblastic activity in bone biopsies. Basic Appl. Histochem. 31(Suppl.), 29 (1987). J. Zernik, K. Twarog, and W. B. Upholt, Regulation of alkaline phosphatase and ? 2(I) procollagen synthesis during early intramembranous bone formation in the rat mandible. Differentiation 44, 207 – 215 (1990). S. Sakano, Y. Murata, T. Miura, H. Iwata, K. Sato, N. Matsui, and H. Seo, Collagen and alkaline phosphatase gene expression during bone morphogenetic protein (BMP-induced cartilage and bone differentiation. Clin. Orthop. 292, 337 – 344 (1993). P. Collin, J. R. Nefussi, A. Wetterwald, V. Nicolas, M. L. BoyLefever, H. Fleisch, and N. Forest, Expression of collagen, osteocalcin and bone alkaline phosphatase in a mineralizing rat osteoblastic cell culture. Calcif. Tissue Int. 50, 175 – 183 (1992). P. Bianco, and M. Riminucci, The bone marrow stroma in vivo: Ontogeny, structure, cellular composition and changes in disease. In “Marrow Stromal Cell Culture” (J. N. Beresford and M. E. Owen, eds.), pp. 10 – 25. Cambridge Univ. Press, Cambridge, UK. P. Bianco, M. Costantini, and L. C. Dearden, Alkaline phosphatase positive precursors of adipocytes in the human bone marrow. Br. J. Haematol. 88, 401 – 403 (1988).
169 362. B. C. Nair, R. J. Majeska, and G. A. Rodan, Rat alkaline phosphatase. I. Purification and characterization of the enzyme from osteosarcoma: Generation of monoclonal and polyclonal antibodies. Arch. Biochem. Biophys. 254, 18 – 27 (1987). 363. B. C. Nair, D. E. Johnson, R. J. Majeska et al., Rat alkaline phosphatase. II. Structural similarities between the osteosarcoma, bone, kidney, and placental isoenzymes. Arch. Biochem. Biophys. 254, 28 – 34 (1987). 364. S. Nagoya, T. Uede, T. Wada, S. Ishii, S. Yamawaki, and K. Kikuchi, Detection of bone-type alkaline phsophatase by monoclonal antibodies reacting with human osteosarcoma-associated antigen. Jpn. J. Cancer Res. 82, 862 – 870 (1991). 365. M. G. Low, The glycosyl-phosphatidylinositol anchor of membrane proteins. Biochim. Biophys. Acta. 988, 427 – 454 (1989). 366. M. Noda, K. Yoon, G. A. Rodan, and D. E. Koppel, High lateral mobility of endogenous and transfected alkaline phosphatase: A phosphitdylinositol anchored membrane protein. J. Cell Biol. 105, 1671 – 1677 (1987). 367. E. E. Kim and H. W. Wyckoff, Structure of alkaline phosphatases. Clin. Chem. Acta 186, 175 – 187 (1989). 368. H. W. Wyckoff, M. D. Handschumacher, H. M. K. Murthy, and J. M. Sowadski, The three-dimensional structure of alkaline phosphatase from E coli. Adv. Enzymol. Rel. Areas Mol. Biol. 55, 453 – 480 (1983). 369. E. E. Kim and H. W. Wyckoff, Reaction mechanisms of alkaline phosphatase based on crystal structures: Two-metal ion catalysis. J. Mol. Bio. (England) 218, 449 – 464 (1991). 370. S. Hansen, L. K. Hansen, and E. Hough, The crystal structure of tris-inhibited phospholipase C from Bacillus cereus at 1.9 A resolution: The nature of the metal ion in site 2. J. Mol. Biol. 231(3), 870 – 876 (1993). 371. D. M. Swallow, Mapping of the gene coding for the human liver/bone/kidney isozyme of alkaline phosphatase to chromosome 1. Ann. Hum. Genet. 60, 229 – 235 (1986). 372. H. H. Hsu, The partial sequencing of human nonspecific (bone/liver/kidney) alkaline phosphatase gene. Int. J. Biochem. 23, 391 – 393 (1991). 373. J. E. Beever and H. A. Lewin, RFLP at the bovine liver, bone and kidney alkaline phosphatase (ALPL) locus. Anim. Genet. 23(6), 577 (1992). 374. M. Terao, M. Studer, M. Gianni, and E. Garattini, Isolation and characterization of the mouse liver/bone/kidney-type alkaline phosphatase gene. Biochem. J. 268(3), 641 – 648 (1990). 375. M. Studer, M. Terao, M. Gianni, and E. Garattini, Characterization of a second promoter for the mouse liver/bone/kidney-type alkaline phosphatase gene: Cell and tissue specific expression. Biochem. Biophys. Res. Commun. 179(3), 1352 – 1360 (1991). 376. J. Zernik, B. Kream, and K. Twarog, Tissue-specific and dexamethasone-inducible expression of an alkaline phosphatase from alternative promoters of the rat bone/liver/kidney/placenta gene. Biochem. Biophys. Res. Commun. 176, 1149 – 1156 (1991). 377. M. Kiledjian and T. Kadesch, Analysis of the human liver/bone/kidney alkaline phosphatase promoter in vivo and in vitro. Nuclei Acids Res. 18, 957 – 961 (1990). 378. J. Zernik, M. A. Thiede, K. Twarog, M. L. Stover, G. A. Rodan, W. B. Upholt, and D. W. Rowe, Cloning and analysis of the 5 region of the rat bone/liver/kidney/placenta alkaline phosphatase gene: A dual function promter. Matrix 10, 38 – 47 (1990). 379. S. Matsuura, F. Kishi, and T. Kajii, Characterization of a 5-flanking region of the human liver/bone/kidney alkaline phosphatase gene: Two kinds of mRNA from a single gene. Biochem. Biophys. Res. Commun. 168(3), 993 – 1000 (1990). 380. J. A. Deren, F. S. Kaplan, and C. T. Brighton, Alkaline phosphatase production by periosteal cells at various oxygen tensions in vitro. Clin. Orthop. 252, 307 – 312 (1990).
170 381. L. Y. Yohay, J. Zhang, K. M. Thraikill, J. M. Arthur, and L. D. Quarles, Role of serum in the developmental expression of alkaline phosphatase in MC3T3-E1 osteoblasts. J. Cell Physiol. 158, 467 – 475 (1994). 382. E. Kyeyune-Nyombi, K. H. Lau, D. J. Baylink, and D. D. Strong, 1,25-Dihydroxyvitamin D3 stimulates both alkaline phosphatase gene transcription and mRNA stability in human bone cells. Arch. Biochem. Biophys. 291, 316 – 325 (1991). 383. K. W. Ng, P. R. Gummer, V. P. Michelangeli, J. E. Baterman, T. Mascara, W. G. Cole, and T. J. Martin, Regulation of alkaline phosphatase expression in a neonatal rat calvarial cell strain by retinoic acid. J. Bone Miner. Res. 3, 43 – 61 (1988). 384. J. K. Heath, L. J. Suva, K. Yoon, M. Kiledjian, T. J. Martin, and G. A. Rodan, Retinoic acid stimulates transcriptional activity from the alkaline phosphatase promoter in the immortalized rat calvarial cell line, RCT-1. Mol. Endocrinol. 6, 636 – 646 (1992). 385. M. Gianni, M. Terao, S. Sozzani, and E. Garattini, Retinoic acid and cyclic AMP synergistically induce the expression of liver/bone/kidney-type alkaline phosphatase gene in L929 fibroblastic cells. Biochem. J. 296, 67 – 77 (1993). 386. M. Subramaniam, D. Colvard, P. E. Keeting, K. Rasmussen, B. L. Riggs, and T. C. Spelsberg, Glucocorticoid regulation of alkaline phosphatase, osteocalcin, and proto-oncogenes in normal human osteoblast-like cells. J. Cell. Biochem. 50, 411 – 424 (1992). 387. J. R. Farley, S. L. Hall, S. Herring, Calcitonin acutely increases net 45Ca uptake and alters alkline phosphatase specific activity in human osteosarcoma cells. Metabolism 42(1); 97 – 104 (1993). 388. P. J. Price, B. O. Oyajobi, R. O. Oreffo, and R. G. G. Russell, Cells cultured from the growing tip of red dear antler express alkaline phosphatase and proliferate in response to insulin-like growth factor-I. J. Endocrionol. 143, R9 – R16 (1994). 389. P. S. Leboy, L. Vaias, B. Uschman, E. Golub, and S. L. Adams, Pacific M’ ascorbic acid induces alkaline phosphatase, type X collgen, and calcium deposition in cultured chick chondrocytes. J. Biol. Chem. 264, 17281 – 17286 (1989). 390. J. F. Chlebowski and J. E. Coleman, Pre-steady state kinetics of phosphorothioate hydrolysis by alkaline phosphatase. J. Biol. Chem. 247, 6007 – 6010 (1989). 391. R. Robison, The possible significance of hexose phosphoric esters in ossification. Biochem. J. 17, 286 – 293 (1923). 392. P. Magnusson, L. Larsson, M. Magnusson, M. W. Davie, and C. A. Sharp, Isoforms of bone alkaline phosphatase: Characterization and origin in human trabecular and cortical bone. J. Bone Miner. Res. 14, 1926 – 1933 (1999). 393. G. I. Baroncelli, S. Bertelloni, C. Ceccarelli, V. Amato, and G. Saggese, Bone turnover in children with vitamin D deficiency rickets before and during treatment. Acta. Paediatr. 89, 513 – 518 (2000). 394. M. P. Whyte, Alkaline phosphatase: Physiological role explored in hypophosphatasia. In “Bone and Mineral Research” (W. A. Peck, ed.), Vol. 6, pp. 175 – 218. Elsevier, Amsterdam, 1989. 395. H. Harris, The human alkaline phosphatases: What we know and what we don’t know. Clin. Chim. Acta 186, 133 – 150 (1989). 396. M. J. Weiss, D. E. C. Cole, K. Ray, M. P. Whyte, M. A. Lafferty, R. A. Mulivor, and H. Harris, A misuse mutation in the human liver/bone/kidney alkaline phosphatase causing a lethal form of hyphophosphatasia. Proc. Natl. Acad. Sci. USA 85, 7666 – 7669 (1989). 397. K. Yoon, E. Golub, and G. A. Rodan, Alkaline phosphatase transfected cells promote calcium and phosphate deposition. Connect. Tissue Res. 22, 17 – 25 (1989). 398. H. I. Khouja, A. Bevington, G. J. Kemp, and R. G. Russell, Calcium and orthophosphate deposits in vitro do not imply osteoblastmediated mineralization: Mineralization by betaglycerophosphate in the absence of osteoblasts. Bone 11, 385 – 391 (1990).
GOKHALE, BOSKEY, AND ROBEY 399. M. D. Fallon, M. P. Whyte, and S. L. Teitelbaum, Stereospecific inhibition of alkaline phosphatase by L-tetramisole prevents in vitro cartilage calcification. Lab. Invest. 43, 489 – 494 (1980). 400. B. Zimmerman, H. C. Wachtel, and J. Vormann, Kinetics of ßglycerophosphate-induced endochondral mineralization in vitro: Calcium accumulation, alkaline phosphatase activity, and effects of levamisole. Calcif. Tissue Int. 51, 54 – 61 (1992). 401. A. L. Boskey, W. Ziecheck, P. Guidon, and S. B. Doty, Gallium nitrate inhibits alkaline phosphatase activity in differentiating mesenchymal cell culture. Bone Min. 20, 179 – 192 (1993). 402. Y. Ishikawa and R. E. Wuthier, Development of an in vitro mineralization model with growth plate chondrocytes that does not require ß-glycerophosphate. Bone Miner. 17, 152 – 157 (1992). 403. K. N. Fedde, L. Blair, J. Silverstein, S. P. Coburn, L. M. Ryan, R. S. Weinstein, K. Waymire, S. Narisawa, J. L. Millan , G. R. MacGregor, and M. P. Whyte, Alkaline phosphatase knock-out mice recapitulate the metabolic and skeletal defects of infantile hypophopshatasia. J. Bone Miner. Res. 14, 2015 – 2026 (1999). 404. S. Narisawa, N. Frohlander, and J. L. Millan, Inactivation of two mouse alkaline phosphatase genes and establishment of a model of infantile hypophosphatasia. Dev. Dyn. 208; 432 – 446 (1997). 405. W. M. Burch, G. Hammer, and R. E. Wuthier, Phosphotyrosine and phosphoprotein phosphatase activity of alkaline phosphatase in mineralizing cartilage. Metabolism 34, 169 – 175 (1985). 406. A. L. Boskey, S. B. Doty, R. Mendelsohn, and I. Binderman, ATP promotes mineralization in differentiating chick limb-bud mesenchymal cell cultures. Micros. Res. Techniq. 28, 492 – 504 (1994). 407. L. N. Wu, W. B. Valhmu, G. C. Lloyd, B. R. Genge, and R. E. Wuthier, Isolation of two glycosylated forms of membrane-bound alkaline phosphatase from avian growth plate cartilage matrix vesicle-enriched microsomes. Bone Miner. 7, 113 – 125 (1989). 408. Y. Xu, K. P. Pritzer, and T. F. Cruz, Characteriztion of chondrocyte alkaline phosphatase as a potential mediator in the dissolution of calcium pyrophosphate dihydrate crystals. J. Rheumatol. 21, 912 – 919 (1994). 409. J. D. Termine, H. K. Kleinman, S. W. Whitson, K. M. Conn, M. L. McGarvey, and G. R. Martin, Osteonectin, a bone-specific protein linking mineral to collagen. Cell 26, 99 – 105 (1981). 410. K. M. Conn and J. D. Termine, Matrix protein profiles in calf bone development. Bone 6, 33 – 36 (1985). 411. I. J. Mason, A. Taylor, J. G. Williams, H. Sage, and B. L. M. Hogan, Evidence from molecular cloning that SPARC, a major product of mouse embryo parietal endoderm, is related to endothelial cell “culture shock” glycoprotein of MR-43000. EMBO J. 5, 1465 – 1472 (1986). 412. H. Sage, C. Johnson, and P. Bornstein, Characterization of a novel serum albumin binding glycoprotein secreted by endothelial cells in culture. J. Biol. Chem. 259, 3993 – 4007 (1984). 413. M. Dziadek, M. Paulsson, M. Aumailley and R. Timpl, Purification and tissue distribution of a small protein (BM-40) extracted from a basement membrane tumor. Eur. J. Biochem. 161, 455 – 464 (1986). 414. K. Mann, R. Deutzmann, M. Paulsson, and R. Timpl, Solubilization of protein BM-40 from a basement membrane tumor with chelating agents and evidence for its identity with osteonectin and SPARC. FEBS Lett. 218, 167 – 172 (1987). 415. P. W. H. Holland, S. J. Harper, and J. H. McVey, In vivo expression of mRNA for the Ca-binding protein SPARC (osteonectin) revealed by in situ hybridization. J. Cell Biol. 105, 473 – 482 (1987). 416. M. Metsaranta, M. F. Young, M. Sandberg, J. Termine, and E. Vuorio, Localization of osteonectin expression in human fetal skeletal tissues by in situ hybridization. Calcif. Tissue Int. 45, 146 – 152 (1989). 417. P. L. Porter, E. H. Sage, T. F. Lane, S. E. Funk, and A. M. Gown, Distribbution of SPARC in normal and neoplastic human tissue. J. Histochem. Cytochem. 43, 791 – 800 (1995).
171
CHAPTER 4 The Biochemistry of Bone 418. W. M. Wewer, R. Albrechtsen, L. W. Fisher et al., Osteonectin/ SPARC/BM-40 in human decidua and carcinoma, tissues characterized by de novo formation of basement membrane. Am. J. Pathol. 132, 345 (1988). 419. H. Sage, A secreted protein associated with cellular proliferation inhibits cell spreading in vitro. J. Cell Biol. 107, 587a (1989). 420. P. Bianco, J. B. Kopp, and L. W. Fisher, New data on the distribution of osteonectin/SPARC in human and mouse tissues and their potential implications. J. Bone Miner. Res. 4(Suppl.), S245 (1989). 421. J. B. Kopp, P. Bianco, M. F. Young, J. D. Termine, and P. Gehron Robey, Differential expression of osteopontin and osteonectin by cultured renal tubular epithelial cells. Kidney Int. 41, 56 – 64 (1992). 422. P. Bianco, G. Silverstrini, J. D. Termine, and E. Bonucci, Immunohistochemical localization of osteonectin in developing human bone using monoclonal antibodies. Calcif. Tissue Int. 43, 155 – 161 (1988). 423. G. Jundt, L. H. Berghauser, J. D. Termine et al., Osteonectin: A differentiation marker of bone cells. Cell Tissue Res. 248, 409 (1987). 424. P. S. Tung, C. Domenicucci, S. Wasi et al., Specific immunohistochemical localization of osteonectin and collagen types I and III in fetal and adult porcine dental tissues. J. Histochem. Cytochem. 33, 531 (1985). 425. S. Kasugai, T. Nagata, and J. Sodek, Temporal studies on the tissue compartmentalization of bone sialoprotein (BSP), osteopontin (OPN), and SPARC protein during bone formation in vitro. J. Cell Physiol. 152, 467 – 77 (1992). 426. P. Gehron Robey, L. W. Fisher, J. T. Stubbs, and J. D. Termine, Biosynthesis of osteonectin and a small proteoglycan (PG-II) by connective tissue cells in vitro. In “Development and Diseases of Cartilage and Bone Matrix” (A. Sen and T. Thornhill, eds.), pp. 115 – 125. A. R. Liss, New York, 1986. 427. D. D. Stenner, R. P. Tracy, B. L. Riggs, and K. G. Mann, Human platelets contain and secrete osteonectin, a major protein of mineralized bone. Proc. Natl. Acad, Sci. USA 83, 6892 – 6896 (1986). 428. Q. Yan, and E. H. Sage, SPARC, a matricellular glycoprotein with important biological functions. J. Histochem. Cytochem. 47, 1495 – 1506 (1999). 429. R. W. Romberg, P. G. Werness, P. Lollar, B. L. Riggs, and K. G. Mann, Isolation and characterization of native adult osteonectin. J. Biol. Chem. 260, 2728 – 2736 (1985). 430. R. Thomas, L. D. True, J. A. Bassuk, P. H. Lange, and R. L. Vessella, Differential expression of osteonectin/SPARC during human prostate cancer progression. Clin. Cancer Res. 6, 1140 – 1149 (2000). 431. R. W. Romberg, P. G. Werness, P. Lollar, B. L. Riggs, and K. G. Mann, Isolation and characterization of native adult osteonection. J. Biol. Chem. 260, 2728 – 2736 (1985). 432. A. Swaroop, B. L. M. Hogan, and U. Francke, Molecular analysis of the cDNA for human SPARC/Osteonectin/BM-40: Sequence, expression and localization of the gene to chromosome 5q31 – q33. Genomics 2, 37 (1988). 433. R. C. Schwartz, M. F. Young, and P. Tsipouras, Two RFLPs in the 5 end of the human osteonectin (ON) gene. Nucleic Acids Res. 16(18), 9076 (1988). 434. M. E. Bolander, M. F. Young, L. W. Fisher, Y. Yamada, and J. D. Termine, Osteonectin cDNA sequence reveals potential binding regions for calcium and hydroxyapatite and shows homologies with both a basement membrane protein (SPARC) and a serine-protease inhibitor (ovomucoid). Proc. Natl. Acad. Sci. USA 85, 2919 – 2923 (1988). 435. D. M. Findlay, L. W. Fisher, C. I. McQuillan, J. D. Termine, and M. F. Young, Isolation of the osteonectin gene: Evidence that a variable region of the osteonectin molecule is encoded within one exon. Biochemistry 27, 1483 – 1489 (1988). 436. H. H. McVey, S. Nomura, P. Kelly, I. J. Mason, and B. L. Hogan,
437.
438.
439.
440.
441.
442.
443.
444.
445.
446.
447.
448.
449.
450.
451.
452.
453.
Characterization of the mouse SPARC/Osteonectin gene: Intron/exon organization and an unusual promoter region. J. Biol. Chem. 263, 11111 – 11116 (1988). J. A. Bassuk, M. L. Iruela-Arispa, T. F. Lane, J. M. Benson, R. A. Berg, and E. H. Sage, Molecular analyses of chicken embryo sparc (osteonectin). Eur. J. Biochem. 218, 117 – 127 (1993). I. G. Johnston, T. Paladin, J. W. Curd, and I. R. Brown, Molecular cloning of SC1: A putative brain extracellular matrix glyprotein showing partial similarity to osteonectin/BM40/SPARC. Neuron 2, 165 – 176 (1990). P. Maurer, C. Hohenadl, E. Hohenester, W. Gohring, R. Timpl, and J. Engel, The C-terminal portion of BM-40 (SPARC/osteonectin) is an autonomously folding and crystallisable domain that binds calcium and collagen IV. J. Mol. Biol. 253, 347 – 357 (1995). M. Guermah, P. Crisanti, D. Laugier, L. Bidou, B. Pessac, and G. Calothy, Transcription of quail gene expressed in embryonic retinal cells is shut off sharply at hatching. Proc. Natl. Acad. Sci. USA 88, 4503 – 4507 (1991). P. M. Alliel, J. P. Perin, P. Jolles, and F. J. Bonnet, Testican, a multidomain testicular proteoglycan resembling modulators of cell social behaviour. Eur. J. Biochem. 214, 347 – 350 (1993). A. Zwijsen, H. Blockx, W. Van Arnhem, J. Willems, L. Fransen, K. Devos, J. Raymackers, A. Van de Voorde, and H. Slegers, Characterization of a rat C6 glioma-secreted follistatin-related protein (FRP): Cloning and sequence of the human homologue. Eur. J. Biochem. 225, 937 – 946 (1994). M. F. Young, D. M. Findlay, P. Dominguez et al., Osteonectin promoter: DNA sequence analysis and S1 endonuclease site, potentially associated with transcriptional control in bone cells. J. Biol. Chem. 264, 450 (1989). P. Dominguez, K. Ibaraki, P. Gehron Robey, T. Hefferan, J. D. Termine, and M. F. Young, The expression of the osteonectin gene is potentially controlled by multiple cis and trans acting elements. J. Bone Mines. Res. 6, 1127 – 1136 (1991). K. Ibaraki, P. Gehron Robey, and M. F. Young, The transcription of the osteonectin gene is potentially controlled by AP1, AP2, AP3 and a novel GAGA binding factor. Connect Tissue Res. 27, 130 (1992). L. Malaval, B. Darbouret, C. Preaudat, J. P. Jolu, and P. D. Delmas, Intertissular variations in osteonectin: A monoclonal antibody directed to bone osteonectin shows reduced affinity for platelet osteonectin. J. Bone Miner. Res. 6(4), 315 – 323 (1991). R. C. Hughes, A. Taylor, H. Sage, and B. L. M. Hogan, Distinct patterns of glycosylation of colligin, a collagen-binding glycoprotein, and SPARC (osteonectin), a secreted Ca-binding glycoprotein: Evidence for the localization of colligin in the endoplasmic reticulum. Eur. J. Biochem. 163, 57 – 65 (1987). R. J. Kelm, Jr., and K. G. Mann, The collagen binding specificity of bone and platelet osteonectin is related to differences in glycosylation. J. Biol. Chem. 266, 9632 – 9639 (1991). R. J. Kelm, Jr., G. A. Hair, K. G. Mann, and B. W. Grant, Characterization of human osteoblast and megakaryocyte-derived osteonectin (SPARC). Blood 80, 3112 – 3119 (1992). X. C. Villarreal, L. Malaval, K. G. Mann, P. Delmas, and G. L. Long, Epitope mapping of two monoclonal antibodies to the central portion of human osteonectin. Calcif. Tissue Int. 48, 138 – 141 (1991). S. Chandrasekhar, A. K. Harvey, M. G. Johnson, and G. W. Becker, Osteonectin/SPARC is a product of articular chondrocytes/cartilage and is regulated by cytokines and growth factors. Biochim. Biophys. Acta. 1221(1), 7 – 14 (1994). M. Noda, and G. A. Rodan, Type beta transforming growth factor regulates expression of genes encoding bone matrix proteins. Connect. Tissue Res. 21(1 – 4), 71 – 75 (1989). J. L. Wrana, M. Maeno, B. Hawrylyshyn, K. L. Yao, C. Domenicucci, and J. Sodek, Differential effects of transforming
172
454.
455.
456.
457.
458.
459.
460.
461.
462.
463.
464.
465.
466.
467.
468.
469.
469a.
470.
GOKHALE, BOSKEY, AND ROBEY growth factor-? on the synthesis of extracellular matrix proteins by normal fetal rat calvarial bone cell populations. J. Cell Biol. 106, 915 – 924 (1988). H. Zhou, R. G. Hammonds, Jr., D. M. Findlay, P. J. Fuller, T. J. Martin, and K. W. Ng, Retinoic acid modulation of mRNA levels in malignant, nontransformed, and immortalized osteoblasts. J. Bone Miner. Res. 6(7), 767 – 777 (1991). M. Neri, F. Descalzi-Cancedda, and R. Cancedda, Heat-shock response in cultured chick embryo chondrocytes: Osteonectin is a secreted heat-shock protein. Eur. J. Biochem. 205, 569 – 574 (1992). M. F. Young, A. A. Day, P. Dominguez, C. I. McQuillan, L. W. Fisher, and J. D. Termine, Structure and expression of osteonectin mRNA in human tissue. Connect. Tissue Res. 24, 17 – 28 (1990). J. Engel, W. Taylor, M. Paulsson, H. Sage, and B. Hogan, Calcium binding domains and calcium-induced conformational transition of SPARC/BM-40/osteonectin, an extracellular glycoprotein expressed in mineralized and nonmineralized tissues. Biochemistry 26, 6958 – 6965 (1987). E. Hohenester, P. Maurer, and R. Timpl, Crystal structure of a pair of follistatin-like and EF-hand calcium-binding domains in BM-40. EMBO J. 16(13), 3778 – 3786 (1997). B. Tyree, The partial degradation of osteonectin by a bone-derived metalloprotease enhances binding to type I collagen. J. Bone Miner. Res. 4, 877 – 883 (1989). H. Sage, R. B. Vernon, S. E. Funk, E. A. Everitt, and J. Angello, SPARC, a secreted protein associated with cellular proliferation, inhibits cell spreading in vitro and exhibits Ca2 dependent binding to the extracellular matrix. J. Cell Biol. 109, 341 – 356 (1989). P. Clezardin, L. Malavel, A.-S. Ehrensperger, P. Delmas, M. Dechavanne, and J. L. McGregor, Complex formation of human thrombospondin with osteonectin. Eur. J. Biochem. 175, 275 – 284 (1988). R. W. Romberg, P. G. Werness, B. L. Riggs, and K. G. Mann, Inhibition of hydroxyapatite crystal growth by bone-specific and other calcium-binding proteins. Biochemistry 25, 1176 – 1180 (1986). H. J. Roach, Why does bone matrix contain non-collagenous proteins? The possible roles of osteocalcin, osteonectin, osteopontin, and bone sialoprotein in mineralisation and resorption. Cell. Biol. Int. 18, 617 – 628 (1994). S. E. Funk and E. H. Sage, The Ca2 binding glycoprotein SPARC modulates cell cycle progression in bovine aortic endothelial cells. Proc. Natl. Acad. Sci. USA 88, 2648 – 2652 (1991). E. H. Sage, Modulation of endothelial cell shape by SPARC does not involve chelation of extracellular Ca2 and Mg2. Biochem. Cell. Biol. 70, 56 – 62 (1992). T. F. Lane and H. Sage, Functional mapping of SPARC: Peptides from two distinct Ca(2) binding sites modulates cell shape. J. Cell Biol. 111, 3065 – 3076 (1990). H. Sage and P. Bornstein, Extracellular proteins that modulate cellmatrix interations: SPARC, tenascin, and thrombospondin. J. Biol. Chem. 266, 14831 – 14834. D. T. Gilmour, G. J. Lyon, M. B. Carlton, J. R. Sanes, J. M. Cunningham, J. R. Anderson, B. L. Hogan, M. J. Evans, and W. H. Colledge, Mice deficient for the secreted glycoprotein SPARC/osteonectin/BM40 develop normally but show severe age-onset cataract formation and disruption of the lens. EMBO J. 17, 1860 – 1870 (1998). A. M. Delany, M. Amling, M. Priemel, C. Howe, R. Baron, and E. Canalis, Osteopenia and decreased bone formation in osteonectin-deficient mice [published erratum appears in J. Clin. Invest. 105(9), 1325 (2000)]. J. Clin. Invest. 105, 915 – 923 (2000). A. M. Delany, M. Amling, M. Priemel, C. Howe, R. Baron, and E. Canalis, Osteopenia and decreased bone formation in osteonectin-deficient mice. J. Clin. Invest. 105, 1325 (2000). T. F. Lane, M. L. Iruela-Arispe, and E. H. Sage, Regulation of gene expression by SPARC during angiogenesis in vitro: Changes in
471.
472.
473.
474.
475.
476.
477.
478. 479.
480.
481.
482.
483.
484.
485.
486.
487.
488.
fibronectin, thrombospondin-1, and plasminogen activator inhibitor-1. J. Biol. Chem. 267, 16736 – 16745 (1992). P. Hasselaar and E. H. Sage, SPARC antagonizes the effect of basic fibroblast growth factor on the migration of bovine aortic endothelial cells. J. Cell. Biochem. 49, 272 – 283 (1992). J. C. Yost and E. H. Sage, Specific interaction of SPARC with endothelial cells is mediated through a carboxyl-terminal sequence containing a calcium-binding EF hand. J. Biol. Chem. 268, 25790 – 25796 (1993). S. E. Funk and E. H. Sage, Differential effects of SPARC and cationic SPARC peptides on DNA synthesis by endothelial cells and fibroblasts. J. Cell. Physiol. 154, 53 – 63 (1993). S. E. Goldblum, X. Ding, S. E. Funk, and E. H. Sage, SPARC (secreted protein acidic and rich in cysteine) regulates endothelial cell shape and barrier function. Proc. Natl. Acad. Sci. USA 91, 3448 –3452 (1994). E. H. Sage, J. A. Bassuk, J. C. Yost, M. J. Folkman, and T. F. Lane, Inhibition of endothelial cell proliferation by SPARC is mediated through a Ca2-binding EF-hand sequence. J. Cell. Biochem. 57, 127 – 140 (1995). E. A. Everitt and E. H. Sage, Expression of SPARC is correlated with altered morphologies in transfected F9 embryonal carcinoma cells. Exp. Cell Res. 199, 134 – 146 (1992). E. A. Everitt and E. H. Sage, Overexpression of SPARC in stably transfected F9 cells mediates attachment and spreading in Ca(2)deficient medium. Biochem. Cell. Biol. 70, 1368 – 1379 (1992). T. F. Lane and E. H. Sage, The biology of SPARC, a protein that modulates cell-matrix interactions. FASEB J. 8, 163 – 173 (1994). T. F. Lane, M. L. Iruela-Arispe, R. S. Johnson, and E. H. Sage, SPARC is a source of copper-binding peptides that stimulate angiogenesis. J. Cell Biol. 125, 929 – 943 (1994). M. J. Reed, P. Puolakkainen, T. F. Lane, D. Dickerson, P. Bornstein, and E. H. Sage, Differential expression of SPARC and thrombospondin 1 in wound repair: Immunolocalization and in situ hybridization. J. Histochem. Cytochem. 41, 1467 – 1477 (1993). E. W. Raines, T. F. Lane, M. L. Ireula-Arispe, R. Ross, and E. H. Sage, The extracellular glycoprotein SPARC interacts with plateletderived growth factor (PDGF-AB and -BB and inhibits the binding of PDGF to its receptors. Proc. Natl. Acad. Sci. USA 89, 1281 – 1285 (1992). P. M. Tremble, T. F. Lane, E. H. Sage, and Z. Werb, SPARC, a secreted protein associated with morphogenesis and tissue remodeling, induces expression of metalloproteinases in fibroblasts through a novel extracellular matrix-dependent pathway. J. Cell Biol. 121, 1433 – 1444 (1993). R. J. Kelm, Jr., N. A. Swords, T. Orfeo, and K.G. Mann, Osteonectin in matrix remodeling, a plasminogen-osteonectin-collagen complex. J. Biol. Chem. 269, 30147 – 30153 (1994). J. E. Murphy-Ullrich, T. F. Lane, M. A. Pallero, and E. H. Sage, SPARC mediates focal adhesion disassembly in endothelial cells through a follistatin-like region and the Ca2-binding EF-hand. J. Cell Biochem. 57, 341 – 350 (1995). L. Purcell, J. Guia-Gray, S. Scanga, and M. Ringuette, Developmental anomalies of xenopus embroys following microinjection of SPARC antibodies. J. Exp. Zool. 265, 153 – 164 (1993). J. E. Schwarzbauer and C. S. Spencer, The Caenorhabditis elegans homologue of the extracellular calcium binding protein SPARC/osteonectin affects nematode body morphology and mobility. Mol. Biol. Cell 4, 941 – 952 (1993). I. Clemmensen, L. C. Petersen, and C. Kluft, Purification and characterization of a novel, oligomeric, plasminogen kringle 4 binding protein from human plasma: Tetranectin. Eur. J. Biochem. 156, 327 – 333 (1986). U. M. Wewer, K. Ibaraki, P. Schjorring, M. E. Durkin, M. F. Young, and R. Albrechtsen, A potential role for tetranectin in mineralization during osteogenesis. J. Cell Biol. 127, 1767 – 1775 (1994).
CHAPTER 4 The Biochemistry of Bone 489. U. M. Wewer and R. Albrechtsen, Tetranectin, a plasminogen kringle 4-binding protein. Cloning and gene expression pattern in human colon cancer. Lab. Invest. 67, 253 – 262 (1992). 490. L. Berglund and T. E. Petersen, The gene structure of tetranectin, a plasminogen binding protein. FEBS Lett. 309, 15 – 19 (1992). 491. Fuhlendorff, I. Clemmensen, and S. Magnusson, Primary structure of tetranectin, a plasminogen kringle 4 binding plasma protein: Homology with a sialoglycoprotein receptors and cartilage proteoglycan core protein. Biochemistry 26, 6757 – 6764 (1987). 492. E. Ruoslahti and M. D. Pierschbacher, New perspectives in cell adhesion: rgd and integrins. Science 238, 491 (1987). 493. R. O. Hynes, Integrins: A family of cell surface receptors. Cell 48, 548 – 554 (1987). 494. R. O. Hynes, Integrins: Versatility modulation and signalling in cells adhesion. Cell 69, 11 – 25 (1992). 495. D. B. Donoviel, P. Framson, C. F. Eldridge, M. Cooke, S. Kobayashi, and P. Bornstein, Structural analysis and expression of the human thrombospondin gene promoter. J. Biol. Chem. 263, 18590 – 18593 (1988). 496. J. W. Lawler, L. H. Derick, J. E. Connolly, J. H. Chen, and F. C. Chao, The structure of human platelet thrombospondin. J. Biol. Chem. 260, 3762 – 3772 (1985). 497. S. S. Margossian, J. W. Lawler, and H. S. Slayter, Physical characterization of platelet thrombospondin. J. Biol. Chem. 256(14), 7495 – 7500 (1981). 498. J. Lawler, F. C. Chao, and C. M. Cohen, Evidence for calciumsensitive structure in platelet thrombospondin: Isolation and partial characterization of thrombospondin in the presence of calcium. J. Biol. Chem. 257(20), 12257 – 12265 (1982). 499. N. L. Baenziger, G. N. Brodie, and P. W. Majerus, Isolation and properties of a thrombin-sensitive protein of human platelets. J. Biol. Chem. 248, 2723 – 2731 (1972). 500. T. N. Wight, G. J. Raugi, S. M. Mumby et al., Light microscopic immunolocalization of thrombospondin in human tissues. J. Histochem. Cytochem. 33, 295 (1985). 501. K. S. O’Shea, L. H. Liu, L. H. Kinnunen, and Y. M. Dixit, Role of the extracellular matrix protein thrombospondin in the early development of the mouse embryo. J. Cell Biol. 111(6 Pt 1), 2713 – 2723 (1990). 502. K. S. O’Shea, L. H. J. Liu, and V. M. Dixit, Thrombospondin and a 140 kd fragment promote adhesion and neurite outgrowth from embryonic central and peripheral neurons and from PC12 cells. Neuron 7, 231 – 237 (1991). 503. P. Gehron Robey, M. F. Young, L. W. Fisher, and T. D. McClain, Thrombospondin is an osteoblast-derived component of mineralized extracellular matrix. J. Cell Biol. 108, 719 – 727 (1989). 504. W. J. Grzesik and P. Gehron Robey, Bone matrix RGD glycoproteins: Immunolocalization and interaction with human primary osteoblastic bone cells in vitro. J. Bone Miner. Res. 9, 487 – 496 (1994). 505. W. A. Frazier, Thrombospondin: A modular adhesive glycoprotein of platelets and nucleated cells. J. Cell Biol. 105, 625 – 632 (1987). 506. P. Bornstein, Thrombospondins: Structure and regulation of expression. FASEB J. 6, 3290 – 3299 (1992). 507. D. D. Roberts, D. M. Haverstick. V. M. Dixit, W. A. Frazier, S. A. Santoro, and V. Ginsburg, The platelet glycoprotein thrombospondin binds specifically to sulfated glycolipids. J. Biol. Chem. 260, 9405 – 9411 (1985). 508. X. Sun, D. F. Mosher, and A. Rapraeger, Heparan sulfate-mediated binding of epithelial cell surface proteoglycan to thrombospondin. J. Biol. Chem. 264, 2885 – 2889 (1989). 509. F. W. Wolf, R. L. Eddy, T. B. Shows, and V. M. Dixit, Structure and chromosomal localization of the human thrombospondin gene. Genomics 6(4), 685 – 691 (1990). 510. E. Jaffe, P. Bornstein, and C. M. Disteche, Mapping of the thrombospondin gene to human chromosome 15 and mouse chromosome 2 by in situ hybridization. Genomics 7(1), 123 – 126 (1990).
173 511. T. L. LaBell and P. H. Byers, Sequence and characterization of the complete human thrombospondin 2 cDNA: Potential regulatory role for the 3 untranslated region. Genomics 17(1), 225 – 229 (1993). 512. H. L. Vos, S. Devarayalu, Y. de Vries, and P. Bornstein, Thrombospondin 3 (Thbs3), a new member of the thrombospondin gene family. J. Biol. Chem. 267(17), 12192 – 12196 (1992). 513. G. Newton, S. Weremowicz, C. C. Morton, N. A. Jenkins, D. J. Gilbert, N. G. Copeland, and J. Lawler, The thrombospondin-4 gene. Mamm. Genome 10, 1010 – 1016 (1999). 514. J. T. Hecht, M. Deere, E. Putnam, W. Cole, B. Vertel, H. Chen, and J. Lawler, Characterization of cartilage oligomeric matrix protein (COMP) in human normal and pseudoachondroplasia musculoskeletal tissues. Matrix Biol. 17, 269 – 278 (1998). 515. T. Shingu and P. Bornstein, Characterization of the mouse thrombospondin 2 gene. Genomics 16, 78 – 84 (1993). 516. J. Adams and J. Lawler, Curr. Biol. 3, 188 – 190 (1993). 517. J. A. Carron, W. B. Bowler, S. C. Wagstaff, and J. A. Gallagher, Expression of members of the thrombospondin family by human skeletal tissues and cultured cells. Biochem. Biophys. Res. Commun. 263, 389 – 391 (1999). 518. M. Melnick, H. Chen, Y. Zhou, and T. Jaskoll, Thrombospondin-2 gene expression and protein localization during embryonic mouse palate development. Arch. Oral Biol. 45, 19 – 25 (2000). 519. P. Framson and P. Bornstein, A serum response element and a binding site for NF-Y mediate the serum response of the human thrombospondin 1 gene. J. Biol. Chem. 268, 4989 – 4996 (1993). 520. T. Shingu and P. Bornstein, Overlapping Egr-1 and Sp1 sites function in the regulation of transcription of the mouse thrombospondin 1 gene. J. Biol. Chem. 269, 32551 – 32557 (1994). 521. A. Mettouchi, F. Cabon, N. Montreau, P. Vernier, G. Mercier, D. Blangy, H. Tricoire, P. Vigier, and B. Binetruy, SPARC and thrombospondin genes are repressed by the c-jun oncogene in rat embryo fibroblasts. EMBO J. 13, 5668 – 5678 (1994). 522. P. Bornstein, S. Devarayalu, P. Li, C. M. Disteche, and P. Framson, A second thrombospondin gene in the mouse is similar in organization to thrombospondin 1 but does not respond to serum. Proc. Natl. Acad. Sci. USA 88, 8636 – 8640 (1991). 523. K. W. Adolph, G. L. Long, S. Winfield, E. I. Ginns, and P. Bornstein, Structure and organization of the human thrombospondin 3 gene (THBS3). Genomics 27, 329 – 336 (1995). 524. K. W. Adolph, D. J. Liska, and P. Bornstein, Analysis of the promoter and transcription start sites of the human thrombospondin 2 gene (THBS2). Gene 193, 5 – 11 (1997). 525. L. Malaval, D. Modrowski, A. K. Gupta, and J. E. Aubin, Cellular expression of bone related proteins during in vitro osteogenesis in rat bone marrow stromal cell cultures. J. Cell Physiol. 158, 555 – 572 (1994). 526. N. V. Sherbina and P. Bornstein, Modulation of thrombospondin gene expression during osteoblast differentiation in MC3T3-E1 cells. Bone 13, 197 – 201 (1992). 527. Z. Abbadia, J. Amiral, M. C. Trzeciak, P. D. Delmas, and P. Clezardin, The growth-supportive effect of thromobospondin (TSP1) and the expression of TSP1 by human MG-63 osteoblastic cells are both inhibited by dexamethasone. FEBS Lett. 335, 161 – 166 (1993). 528. T. R. Kyriakides, Y. H. Zhu, L. T. Smith, S. D. Bain, Z. Yang, M. T. Lin, K. G. Danielson, R. V. Iozzo, M. LaMarca , C. E. McKinney, E. I. Ginns, and P. Bornstein, Mice that lack thrombospondin 2 display connective tissue abnormalities that are associated with disordered collagen fibrillogenesis, an increased vascular density, and a bleeding diathesis. J. Cell Biol. 140, 419 – 430 (1998). 529. T. R. Kyriakides, Y. H. Zhu, Z. Yang, and P. Bornstein, The distribution of the matricellular protein thrombospondin 2 in tissues of embryonic and adult mice. J. Histochem. Cytochem. 46, 1007 – 1015 (1998).
174 530. M. L. Iruela-Arispe, P. Bornstein, and H. Sage,Thrombospondin exerts an antiangiogenic effect on cord formation by endothelial cells in vitro. Proc. Natl. Acad. Sci. USA 88, 5026 – 5030 (1991). 531. M. L. Iruela-Arispe, D. J. Liska, E. H. Sage, and P. Bornstein, Differential expression of thrombospondin 1, 2, and 3 during murine development. Dev. Dynam. 197, 40 – 56 (1993). 532. G. J. Raugi, J. E. Oleud, and A. M. Gown,Thrombospondin in early human wound tissue. J. Invest. Dermatol. 89, 551 – 554 (1987). 533. M. J. Reed, P. Puolakkainen, T. F. Lane, D. Dickerson et al., Differential expression of SPARC and thrombospondin 1 in wound repair: Immunolocalization and in situ hybridization. J. Histochem. Cytochem. 41, 1467 – 1477 (1993). 534. V. Castle, J. Varani, S. Fligiel, E. V. Prochownik, and V. Dixit, J. Clin Invest. 87, 1883 – 1888 (1991). 535. D. J. Good, P. J. Poiverini, F. Rastinejad, M. M. Le Beau, R. S. Lemons, W. Frazier, and N. P. Bouck, A tumor suppression dependent-inhibitors of angiogeneses in immunologically and functionally indistinguishable from a fragment of thrombospondin. Proc. Natl. Acad. Sci. USA 87, 6624 – 6628 (1990). 536. S. S. Tolsma, O. V. Volpert, D. J. Good, W. A. Frazier, P. J. Polverini, and N. Bouck, Peptides derived from two spearate domains of the matrix protein thrombospondin-1 have anti-antiogenic activity. J. Cell Biol. 122, 497 – 511 (1993). 537. J. A. Carron, W. D. Fraser, and J. A. Gallagher, Thrombospondin promotes resorption by osteoclasts in vitro. Biochem. Biophys. Res. Commun. 213, 1017 – 1025 (1995). 538. S. Schultz-Cherry, H. Chen, D. F. Mosher, T. M. Misenheimer, H. C. Krutzsch, D. D. Roberts, and J. E. Murphy-Ullrich, Regulation of transforming growth factor- activation by discrete sequences of thrombospondin 1. J. Biol. Chem. 270, 7304 – 7310 (1995). 539. M. Winnemoller, P. Schîn, P. Vischer, and H. Kresse, Interactions between thrombospondin and the small proteoglycan decorin: Interference with cell attachment. Eur. J. Biol. 59, 47 – 55 (1992). 540. G. Gao and W. A. Frazier, Identification of a receptor candidate for the carboxyl-terminal cell binding domain of thrombospondins. J. Biol. Chem. 269, 29650 – 29657 (1994). 541. C. A. Prater, J. Plotkin, D. Jaye, and W. A. Frazier, The properdinlike type I repeats of human thrombospondin contain a cell attachment site. J. Cell Biol. 112, 1031 – 1040 (1991). 542. W. X. Li, R. J. Howard, and L. L. K. Leung, Identification of SVTCG in thrombospondin as the conformation-dependent, high affinity binding site for its receptor, CD36. J. Biol. Chem. 268, 16179 – 16184 (1993). 543. A. S. Asch, I. Liu, F. M. Briccetti, J. W. Barnwell, F. KwakyeBerko, A. Dokun, J. Goldberger, and M. Pernambuco, Analysis of CD36 binding domains: Ligand specificity controlled by dephosphorylation of an ectodomain. Science 262, 1436 – 1440 (1993). 544. T. S. Panetti, B. J. Kudryk, and D. F. Mosher, Interaction of recombinant procollagen and properdin modules of thrombospondin-1 with heparin and fibrinogen/fibrin. J. Biol. Chem. 274, 430 – 437 (1999). 545. M. Grano, P. Zigrino, S. Colucci, G. Zambonin et al., Adhesion properties and integrin expression of cultured human osteoclast-like cells. Exp. Cell Res. 212, 209 – 218 (1994). 546. M. A. Horton, E. L. Dorey, S. A. Nesbitt, J. Samanen, F. E. Ali, J. M. Stadel, A. Nichols, R. Greig, and M. H. Helfrich, Modulation of vitronectin receptor-mediated osteoclast adhesion by Arg-GlyAsp peptide analogs: A structure-function analysis. J. Bone Miner. Res. 8(2), 239 – 247 (1993). 547. R. Hynes, Molecular biology of fibronectin, Annu. Rev. Cell Biol. 1, 67 – 90 (1985). 548. D. F. Mosher (ed.), “Fibronectin,” Academic Press, New York, 1989. 549. R. E. Weiss and A. H. Reddi, Appearance of fibronectin during the differentiation of cartilage, bone and bone marrow. J. Cell Biol. 88, 630 (1981).
GOKHALE, BOSKEY, AND ROBEY 550. M. J. Weiss, P. S. Henthorn, M. A. Lafferty, C. Slaughter, M. Raducha, and H. Harris, Isolation and characterization of a cDNA encoding a human liver/bone/kidney-type alkaline phosphatase. Proc. Natl. Acad. Sci. USA 83, 7182 – 7186 (1986). 551. D. J. Leahy, H. P. Erickson, I. Aukhil, P. Joshi, and W. A. Hendricson, Crystallization of a fragment of human fibronectin: Introduction of methionine by site-directed mutagenesis to allow phasing via selenomethionine. Proteins 19, 48 – 54 (1994). 552. A. M. Krezel, G. Wagner, J. Seymour- Ulmer, and R. A. Lazarus, Structure of the RGD protein decorsin: Conserved motif and distinct function in leech proteins that affect blood clotting. Science 264, 1944 – 1947 (1994). 553. R. F. Cachau, E. H. Serpersu, A. S. Mild van, T. August, and L. M. Amzel, Recognition in cell adhesion: A compulsive study of RGDcontaining peptides by Monte Carlo and NMR methods. J. Mol. Recog. 2, 179 – 186 (1989). 554. H. P. Erickson, Reversible unfolding of fibronectin type III and immunoglobulin domains provides the structural basis for stretch and elasticity of titin and fibronectin. Proc. Natl. Acad. Sci. USA 21, 10114 – 10118 (1994). 555. J. E. Schwarzbauer, Alternatives splicing of fibronectin: Three variants, three functions. Bioessays 13, 527 – 533 (1991). 556. J. E. Schwarzbauer, Identification of the fibronectin sequences required for assembly of a fibrillar matrix. J. Cell Biol. 113, 463 – 473 (1991). 557. F. J. Barkalow and J. E. Schwarzbauer, Interactions between fibronectin and chondroitin sulfate are modulated by molecular context. J. Biol. Chem. 269, 3957 – 3962 (1994). 558. K. Ichihara-Tanaka, T. Maeda, K. Titani, and K. Sekiguchi, Matrix assembly of recombinant fibronectin polypeptide consisting of amino-terminal 70 kDa and carboxyl-terminal 37 kDa regions. FEBS Lett. 299(2), 155 – 158 (1992). 559. J. Sottile, J. Schwarzbauer, J. Seleque, and D. F. Mosher, Five type I modules of fibronectin form a functional unit that binds to fibroblasts and Staphylococcus aureus. J. Biol. Chem. 266, 12840 – 12843 (1991). 560. T. Darribere, V. E. Koteliansky, M. A. Chernousov, and S. K. Akiyama, Distinct regions of human fibronectin are essential for fibril assembly in an in vivo developing system. Dev. Dyn. 194, 63 – 70 (1992). 561. K. Vibe-Pedersen, S. Magnusson, and F. E. Baralle, Donor and acceptor splice signals within an exon of the human fibronectin gene: A new type of differential splicing. FEBS Lett. 207, 287 – 291 (1986). 562. H. Hirano, Y. Yamada, M. Sullivan, B. de Crombrugghe, I. Pastan, and K. M. Yamada, Isolation of genomic DNA clones spanning the entire fibronectin gene. Proc. Natl. Acad. Sci. USA 80, 46 – 50 (1983). 563. R. Gardella, M. Colombi, and S. Barlati, Human fibronectin gene (FN1) RFLPs: Mapping and linkage disequilibrium analysis. Hum. Genet. 92, 639 – 641 (1993). 564. D. C. Dean, C. L. Bowlus, and S. Bourgeois, Cloning and analysis of the promoter region of the human fibronectin gene. Proc. Natl. Acad. Sci. USA 84, 1876 – 1880 (1987). 565. D. D. Dean, Z. Schwartz, L. Bonewald, O. E. Muniz, S. Morales, R. Gomez, B. P. Brooks, M. Qiao, D. S. Howell, and B. D. Boyan, Matrix vesicles produced by osteoblast-like cells in culture become significantly enriched in proteoglycan-degrading metalloproteinases after addition of beta-glycerophosphate and ascorbic acid. Calcif. Tissue Int. 54, 399 – 408 (1994). 566. S. Miao, P. K. Suri, L. Shu-Ling, A. Abraham, N. Cook, P. Milos, and M. A. Zern, Role of the cyclic AMP response element in rat fibronectin gene expression. Hepatology 17, 882 – 890 (1993). 567. A. F. Muro, V. A. Bernath, and A. R. Kornblihtt, Interaction of the 170 cyclic AmP response element with the adjacent CCAAT box
CHAPTER 4 The Biochemistry of Bone
568.
569.
570.
571.
572.
573.
574.
575.
576.
577.
578.
579.
580.
581. 582. 583.
584.
585.
in the human fibronectin gene promoter. J. Biol. Chem. 267, 12767 – 12774 (1992). A. Srebrow, A. F. Muro, S. Wergajh, P. A. Sharp, and A. R. Hornblihtt, The CRE-binding factor ATF-2 facilitates the occupation of the CAAT box in the fibronectin gene promoter. FEBS Lett. 327, 25 – 28 (1993). A. R. Kornblihtt, K. Umezawa, K. Vibe-Pedersen, and F. E. Baralle, Primary structure of human fibronectin: Differential splicing may generate at lease 10 polypeptides from a single gene. EMBO J. 4, 1755 – 1759 (1985). R. S. Patel, E. Odermatt, J. E. Schwarzbauer, and R. O. Hynes, Organization of the fibronectin gene provides evidence for exon shuffling during evolution. EMBO J. 6(9), 2565 – 2572 (1987). C. Eielson, D. Kaplan, M. A. Mitnick, I. Paliwal, and K. Insogna, Estrogen modulates parathyroid hormone-induced fibronectin productcion in human and rat osteoblast-like cells. Endocrinology 135, 1639 – 1644 (1994). E. L. George, E. N. Georges-Labouesse, R. S. Patel-King, H. Rayburn, and R. O. Hynes, Defects in mesoderm, neural tube and vascular development in mouse embryos lacking fibronectin. Development 119, 1079 – 1091 (1993). R. J. Majeska, M. Port, and T. A. Einhorn, Attachment to extracellular matrix molecules by cells differing in the expression of osteoblastic traits. J. Bone Miner. Res. 8(3), 277 – 289 (1993). D. A. Puleo and R. Bizios, Mechanisms of fibronectin-mediated attachment of osteoblasts to substrates in vitro. Bone Miner. 18, 215 – 226 (1992). F. P. Ross, J. Chappel, J. I. Alvarez, D. Sander, W. T. Butler, M. C. Farach-Carson, K. A. Mintz, P. G. Robey, S. L. Teitelbaum, and D. A. Cheresh, Interactions between the bone matrix proteins osteopontin and bone sialoprotein and the osteoclast integrin alpha v beta 3 potentiate bone resorption. J. Biol. Chem. 268, 9901 – 9907 (1993). G. Daculsi, P. Pilet, M. Cottrel, and G. Guicheux, Role of fibronectin during biological apatite crystal nucleation: Ultrastructural characterization. J. Biomed. Mater. Res. 47(2), 228 – 233 (1999) G. Daculsi, P. Pilet, M. Cottrel, and G. Guicheux, Role of fibronectin during biological apatite crystal nucleation: Ultrastructural characterization. J. Biomed. Mater. Res. 47, 228 – 233 (1997). D. Seiffert, T. J. Podor, and D. J. Loskutoff, Distribution of Vitronectin. In “Biology of Vitronectins and Their Receptors” (K. T. Preissner, S. Rosenblatt, C. Kost, J. Wegerhoff, and D. F. Mosher, eds.), pp. 75 – 81. Excerpta Medica, Amsterdam, 1993. E. G. Hayman, M. D. Pierschbacher, S. Suzuki, and E. Ruoslahti, Serum spreading factor (vitronectin) is present at the cell surface and in tissues. Proc. Natl. Acad. Sci. USA 80, 4003 – 4007 (1983). T. Kumagai, I. Lee, Y. Ono, M. Maeno, and M. Takagi, Ultrastructural localization and biochemical characterization of vitronectin in developing rat bone. Histochem. J. 30, 111 – 119 (1998). K. T. Preissner, Structure and biological role of vitronectin. Annu. Rev. Cell. Biol. 7, 275 – 310 (1991). K. T. Preissner and D. Jenne, Structure of vitronectin and its biological role in haemostasis. Throm. Haemost. 66, 123 – 132 (1991). A. Oldberg, E. G. Hayman, M. D. Pierschbacher, and E. Ruoslahti, Complete amino acid sequence of human vitronectin deduced from cDNA: Similarity of cell attachment sites in vitronectin and fibronectin. EMBO J. 4, 2519 – 2524 (1985). D. Jenne and K. K. Stanley, Molecular cloning of S-protein, a link between complement coagulation and cell-substrate adhesion. EMBO J. 4, 3153 – 3157 (1985). H. J. Ehrlich, B. Richter, D. von der Ahe, and K. T. Preissner, Primary structure of vitronectins and homology with other proteins. In “Biology of Vitronectins and their Receptors” (K. T. Preissner, S. Rosenblatt, C. Kost, J. Wegerhoff, and D. F. Mosher, eds.), pp. 59 – 66. Excerpta Medica, Amsterdam, 1993.
175 586. K. T. Preissner and D. Jenne, Vitronectin: A new molecular connection in haemostasis. Thromb. Haemost. 66(2), 189 – 194 (1991). 587. J. W. Smith, D. J. Vestal, S. V. Irwin, T. A. Burke, and D. A. Cheresh, Purification and functional characterization of integrin alpha v beta 5. An adhesion receptor for vitronectin. J. Biol. Chem. 265(19), 11008 – 11013 (1990). 588. K. P. Mintz, W. J. Grzesik, R. J. Midura, P. G. Robey, J. D. Termine, and L. W. Fisher, Purification and fragmentation of nondenatured bone sialoprotein: Evidence for a cryptic, RGD-resistant cell attachment domain. J. Bone Miner. Res. 8, 985 – 995 (1993). 589. J. Davies, J. Warwick, N. Totty, R. Philip, M. Helfrich, and M. Horton, The osteoclast functional antigen, implicated in the regulation of bone resorption, is biochemically related to the vitronectin receptor. J. Cell Biol. 109, 1817 – 1826 (1980). 590. R. Baron, J.-H. Ravesloot, L. Neff, M. Chakraborty, D. Chatterjee, A. Lomri, and W. Horne, Cellular and molecular biology of the osteoclast. In “Cellular and Molecular Biology of Bone” (M. Noda, ed.), p. 445 – 495. Academic Press, San Diego, 1993. 591. R. P. M. Rohanizadeh, J. M. Bouler, D. Couchourel, Y. Fortun, and G. Daculsi, Apatite precipitation after incubation of biphasic calcium-phosphate ceramic in various solutions: Influence of seed species and proteins. J. Biomed. Mater. Res. 42, 530 – 539 (1998). 592. C. W. Prince, T. Oosawa, W. T. Butler, M. Tomana, A. S. Bhon, and R. E. Schrohenloher, Isolation, characterization and biosynthesis of a phosphorylated glycoprotein formed at bone. J. Biol. Chem. 262, 2900 – 2907 (1987). 593. A. Franzen and D. Heinegard, Isolation of new phosphorylated glycoprotein from mineralized phase of bone that exhibits limited homology to adhesive protein osteopontin. J. Biol. Chem. 252, 15938 – 15945 (1988). 594. A. Oldberg, A. Franzen, and D. Heinegard, The primary structure of a cell-binding bone sialoprotein. J. Biol. Chem. 263, 19430 – 19432 (1988). 595. J. H. Smith and D. T. Denhardt, Molecular cloning of a tumor promoter-inducible mRNA found in JB6 mouse epidermal cells. J. Cell Biochem. 34, 13 – 22 (1987). 596. G. R. Laverdure, D. Banerjee, I. Chackalaparampil, and B. B. Mukerjee, FEBS Lett. 222, 261 – 265 (1987). 597. D. R. Senger, C. A. Perruzzi, and A. Papadopoulos, Elevated expression of secreted phosphoprotein 1 (osteopontin, 2ar) as a consequence of neoplastic transformation. Anticancer Res. 9, 1291 – 1300 (1989). 598. A. M. Craig, M. Nemir, B. B. Mukerjee, A. F. Chambers, and D. T. Denhardt, Identification of the major phosphoprotein secreted by many rodent cell lines as 2ar/osteopontin. Biochem. Biophys. Res. Commun. 157, 166 – 173 (1988). 599. A. F. Chambers, E. I. Behrend, S. M. Wilson, and D. T. Denhardt, Induction of expression of osteopontin (OPN; secreted phosphoprotein) in metastatic, ras-transformed NIH 3T3 cells. Anticancer Res. 12(1), 43 – 47 (1992). 600. C. A. Lopez, J. R. Hoyer, P. D. Wilson, P. Waterhouse, and D. T. Denhardt, Heterogeneity of osteopontin expression among nephrons in mouse kidneys and enhanced expressionin sclerotic glomeruli. Lab. Invest. 69, 355 – 363 (1993). 601. H. Shiraga, W. Min, W. J. van Dusen, M. D. Clayman, D. Miner, C. L. T. Terrell, J. R. Sherbotie, J. W. Foreman, C. Pryzysiecki, E. G. Nelson, and J. R. Hoyer, Inhibition of calcium oxalate crystal growth in vitro by uropontin: Another member of the aspartic acid-rich protein superfamily, Proc. Natl. Acad. Sci. USA 89, 426 – 430 (1992). 602. E. M. Worcester, S. S. Blumenthal, A. M. Beshensky, and D. L. Lewand, The calcium oxalate crystal growth inhibitor protein produced by mouse kidney cortical cells in culture is osteopontin. JBMR 7, 1029 – 1036 (1992). 603. K. Kohr, Y. Suzuki, K. Yoshida, and K. Yamamoto, Molecular cloning and sequencing of cDNA encoding urinary stone protein
176
604.
605.
606.
607.
608.
609.
610.
611.
612.
613. 614.
615.
616.
617.
618.
619.
620.
GOKHALE, BOSKEY, AND ROBEY which is identical to osteopontin. Biochem. Biophys. Res. Commun. 184, 859 – 864 (1992). D. R. Senger, C. A. Perruzzi, A. Papadopoulos, and D. G. Tenen, Purification of a human milk protein closely similar to tumor-secreted phosphoproteins and osteopontin. Biochim. Biophys. Acta 996(1 – 2), 43 – 48 (1989). P. Waterhouse, R. S. Parhar, X. Guo, R. K. Lala, and D. T. Denhardt, Regulated temporal and spatial expression of the calciumbinding proteins calcyclin and OPN (osteopontin) in mouse tissues during pregnancy. Mol. Reprod. Dev. 32, 315 – 323 (1992). L. F. Brown, B. Berse, L. Vande Water, A. Papadopoulos-Sergion, C. A. Perruzzi, E. J. Manseau, H. F. Dvorak, and D. R. Senger, Expression and distribution of osteopontin in human tissues; widespread association with luminal epithelial surfaces. Mol. Biol. Cell. 3, 1169 – 1180 (1992). X. Guo, Y. P. Zhang, D. A. Mitchell, D. T. Denhardt, and A. F. Chambers, Identification of a ras-activated enhancer in the mouse osteopontin promoter and its interaction with a putative ETS-related transcription factor whose activity correlates with the metastatic potential of the cell. Mol. Cell Biol. 15(1), 476 – 487 (1995). K. Yoon, R. Buenaga, and G. A. Rodan, Tissue specificity and developmental expression of rat osteopontin. Biochem. Biophys. Res. Commun. 148, 1129 – 1136 (1987). M. P. Mark, W. T. Butler, C. W. Prince et al., Developmental expression of 44-kDa bone phosphoprotein (osteopontin) and bone gamma-carboxyglutamic acid (Gla)-containing protein (osteocalcin) in calcifying tissues fo rat. Differentiation 37, 123 (1988). M. F. P. Mark, C. W. Prince, S. Gay et al., 44-kDal bone phosphoprotein (osteopontin) antigenicity at ectopic sites in newborn rats: Kidney and nervous tissues. Cell Tissue Res. 251, 23 (1988). S. L. Smock, D. N. Petersen, W. A. Grasser, D. T. Denhardt, and M. A. Thiede, The osteoblastic phenotype of rat bone marrow megakaryocytes and peripheral blood platelets. J. Bone Miner. Res. 9(Suppl. 1), S126 (1994). M. P. Mark, C. W. Prince, S. Gay, R. L. Austin, M. Bnown, R. D. Finkelman, and W. T. Butler, A comparative immunocytochemical study on the subcellular distributions of 44 kDa bone phosphoprotein and bone gamma-carboxy-glutamic acid (Gla)-containing protein in osteoblasts. J. Bone Miner. Res. 2(4), 337 – 346 (1987). Deleted at proof. M. P. Mark, C. W. Prince, T. Oosawa, S. Gay, A. L. Bronkers, and W. T. Butler, Immunohistochemical demonstration of a 44-kD phosphoprotein in developing rat bones. J. Histochem. Cytochem. 35, 707 – 715 (1987). J. Chen, K. Singh, B. B. Mukherjee, and J. Sodek, Developmental expression of osteopontin (OPN) mRNA in rat tissues: Evidence for a role for OPN in bone formation and resorption. Matrix 13, 113 – 123 (1993). J. Chen, Q. Zhang, C. A. G. McCulloch, and J. Sodek, Immunohistochemical localization of bone sialoprotein in fetal porcine bone tissues; comparison with secreted phosphoprotein 1 (SPP-1, osteopontin) and SPARC (osteonectin). Histochem. J. 23, 281 – 289 (1991). L. J. Suva, J. B. Seedor, N. Endo, H. A. Quartuccio, D. D. Thompson, I. Bab, and G. A. Rodan, Pattern of gene expression following rat tibial marrow ablation. J. Bone Miner. Res. 8, 379 – 388 (1993). R. Quarto, D. Thomas, and C. T. Liang, Bone progenitor cell deficits and the age-associated decline in bone repair capacity. Calcif. Tissue Int. 56, 123 – 129 (1995). M. D. McKee, M. C. Farach-Carson, W. T. Butler, P. V. Hauschka, and A. Nanci, Ultrastructural immunolocalization of noncollagenous (osteopontin and osteocalcin) and plasma (albumin and alpha2 HS-glycoprotein) proteins in rat bone. J. Bone Miner. Res. 8, 485 – 496 (1993). J. B. Lian, M. D. McKee, A. M. Todd, and L. C. Gerstenfeld, Induction of bone-related proteins, osteocalcin and osteopontin, and
621.
622.
623.
624.
625.
626.
627.
628.
629. 630.
631. 632.
633.
634.
635.
636.
637.
638.
their matrix ultrastructural localization with development of chondrocyte hypertrophy in vitro. J. Cell Biochem. 52, 206 – 219 (1993). C. S. Carlson, H. M. Tulli, M. J. Jayo, R. F. Loeser, R. P. Tracy, K. G. Mann, and M. R. Adams, Immunolocalization of noncollagenous bone matrix proteins in lumbar vertebrae from intact and surgically menopausal cynomogous monkeys. J. Bone Miner. Res. 8, 71 – 81 (1993). M. Galotto, G. Campanile, G. Robino, F. D. Cancedda, P. Bianco, and R. Cancedda, Hypertrophic chondrocytes undergo further differentiation to osteoblast-like cells and participate in the initial bone formation in developing chick embryo. J. Bone Miner. Res. 9(8), 1239 – 1249 (1994). K. Merry, R. Dodds, A. Littlewood, and M. Gowen, Expression of osteopontin mRNA by osteoclasts and osteoblasts in modelling adult human bone. J. Cell Sci. 104, 1013 – 1020 (1993). K. Tezuka, T. Sato, H. Kamioka, P. J. Nijweide et al., Identification of osteopontin in isolated rabbit osteoclasts. Biochem. Biophys. Res. Commun. 186, 911 – 917 (1992). M. D. McKee and A. Nanci, Osteopontin and the bone remodeling sequence: Colloidal-gold immunocytochemistry of an interfacial extracellular matrix protein. Ann. N. Y. Acad. Sci. 760, 177 – 189 (1995). M. P. Mark, C. W. Prince, T. Oosawa, S. Gay, A. L. Bronckers, and W. T. Butler, Immunohistochemical demonstration of a 44-KD phosphoprotein in developing rat bones. J. Histochem. Cytochem. 35, 707 – 715 (1987). K. Hultenby, F. P. Reinholt, M. Norgard, A. Oldberg, M. Wendel, and D. Heinegard, Distribution and synthesis of bone sialoprotein in metaphyseal bone of young rats show a distinctly different pattern from that of osteopontin. Eur. J. Cell. Biol. 63, 230 – 239 (1994). A. Franzen and D. Heinegard, Isolation and characterization of two sialoproteins present only in bone calcified matrix. Biochem. J. 232, 715 – 724 (1985). C. W. Prince, Secondary structure predictions for rat osteopontin. Connect. Tissue Res. 21, 15 – 20 (1989). J. M. Kerr, L. W. Fisher, J. D. Termine, and M. F. Young, The cDNA cloning and RNA distribution of bovine osteopontin. Gene 108(2), 237 – 243 (1991). W. T. Butler, The nature and significance of osteopontin. Connect. Tissue Res. 23, 123 – 126 (1989). S. Beninati, D. R. Senger, E. Cordella-Miele, A. B. Mukherjee, I. Chackalaparampil, V. Shanmugam, K. Singh, and B. B. Mukherjee, Osteopontin: Its transglutaminase-catalyzed posttranslational modifications and cross-linking to fibronectin. J. Biochem. (Tokyo) 115, 675 – 682 (1994). H. Mickos, K. Sunderg, and B. Luning, Synthesis, NMR spectra and function of peptides with alpha-methylserine attached to the RGD sequence of osteopontin. Acta Chem. Scand. 46, 989 – 993 (1992). S. Van Dijk, J. A. D’Errico, M. J. Somerman, M. C. Farach-Carson, and W. T. Butler, Evidence that a non-RGD domain in rat osteopontin is involved in cell attachment. J. Bone Miner. Res. 8, 1499 – 1506 (1993). J. A. Evans and S. I. Chan, Phosphophoryn, a biomineralization template protein: pH-dependent protein folding experiments. Biopolymers 34, 507 – 527 (1994). J. S. Evans, T. Chio, and S. I. Chan, Phosphophoryn, an “acidic” biomineralization regulatory protein: Conformational folding in the presence of Cd (II). Biopolymers 34, 1359 – 1375 (1994). S. Ashkar, D. B. Teplow, M. J. Glimcher, and R. A. Saavedra, In vitro phosphorylation of mouse osteopontin expressed in E coli. Biochem. Biophys. Res. Commun. 191, 126 – 133 (1993). J. P. Gorski, Acidic phosphoproteins from bone matrix: A structural rationalization of their role in biomineralization. Calcif. Tissue Int. 50, 391 – 396 (1992).
177
CHAPTER 4 The Biochemistry of Bone 639. N. Hijiya, M. Setoguchi, K. Matsuura, Y. Higuchi, S. Akizuki, and S. Yamamoto, Cloning and characterization of the human osteopontin gene and its promoter. Biochem. J. 303, 255 – 262 (1994). 640. K. Rafidi, I. Simkina, E. Johnson, M. A. Moore, and L. C. Gerstenfeld, Characterization of the chicken osteopontin-encoding gene. Gene 140, 163 – 169 (1994). 641. E. I. Behrend, A. F. Chambers, S. M. Wilson, and D. T. Denhardt, Comparative analysis of two alternative first exons reported for the mouse osteopontin gene. J. Biol. Chem. 268, 11172 – 11175 (1993). 642. A. M. and Craig, D. T. Denhardt, The murine gene encoding secreted phosphoprotein 1 (osteopontin): Promoter structure, activity, and induction in vivo by estrogen and progesterone. Gene 100, 163–171 (1991). 643. Y. Miyazaki, M. Setoguchi, S. Yoshida, Y. Higuchi, S. Akizuki, and S. Yamamoto, The mouse osteopontin gene: Expression in monocytic lineages and complete nucleotide sequence. J. Biol. Chem. 265, 14432 – 14438 (1990). 644. M. F. Young, J. M. Kerr, J. D. Termine, U. M. Wewer, M. G. Wang, O. W. McBride, and L. W. Fisher, cDNA cloning, mRNA distribution and heterogeneity, chromosomal location, and RFLP analysis of human osteopontin (OPN). Genomics 7, 491 – 502 (1990). 645. R. Patarca, R. A. Saavedra, and H. Cantor, Molecular and cellular basis of genetic resistance to bacterial infection: The role of the early T-lymphocyte activation-1/osteopontin gene. Crit. Rev. Immunol. 13, 225 – 246 (1993). 646. M. C. Kieffer, D. M. Aauer, and P. J. Barr, The cDNA and derived amino acid sequence for human osteopontin. Nucleic Acids Res. 17, 3306 (1989). 647. D. T. Denhardt and X. Guo, Osteopontin: A protein with diverse functions. FASEB J. 7, 1475 – 1482 (1993). 648. J. Sodek, J. Chen, S. Kasugai, T. Nagata, Q. Zhang, M. D. McKee, and A. Nanci, Elucidating the functions of bone sialoprotein and osteopontin in bone formation. In “The Chemistry and Biology of Mineralized Tissues” (H. Slavkin and P. Price, eds.), pp. 297 – 306. Elsevier, Amsterdam, 1992. 649. M. Noda and D. T. Denhardt, Regulation of osteopontin gene expression in osteoblasts. Ann. N. Y. Acad. Sci. 760, 242 – 248 (1995). 650. C. W. Prince and W. T. Butler, 1,25-Dihydroxyvitamin D3 regulates the biosynthesis of osteopontin, a bone-derived cell attachment protein, in clonal osteoblast-like osteosarcoma cells. Coll. Relat. Res. 7, 305 (1987). 651. P. L. Chang and C. W. Prince, 1 alpha, 25-Dihydroxyvitamin D3 enhances 12-O-tetradecanoylphorbol-13 acetate-induced tumorigenic transformation and osteopontin expression in mouse JB6 epidermal cells. Cancer Res. 53, 2217 – 2220 (1993). 652. M. Noda, K. Yoon, C. W. Prince, W. T. Butler, andd G. A. Rodan, Transcriptional regulation of osteopontin production in rat osteosarcoma cells by type transforming growth factor. J. Biol. Chem. 263, 13916 – 13921 (1988). 653. S. E. Harris, L. F. Bonewald, M. A. Harris, M. Sabatini et al., Effects of transforming growth factor beta on bone nodule formation and expression of bone morphogenetic protein 2, osteocalcin, osteopontin, alkaline phosphatase, and type I collagen mRNA in longterm cultures of fetal rat calvarial osteoblasts. J. Bone Miner. Res. 9, 855 – 863 (1994). 654. K. Yoon, R. Buenaga, and G. A. Rodan, Tissue specificity and developmental expression of rat osteopontin. Biochem. Biophys. Res. Commun. 148(3), 1129 – 1136 (1987). 655. M. Noda and G. A. Rodan, Transcriptional regulation of osteopontin production in rat osteoblast-like cells by parathyroid hormone. J. Cell. Biol. 108, 713 (1989). 656. M. Shioide and M. Noda, Endothelin modulates osteopontin and osteocalcin messenger ribonucleic acid expression in rat osteoblastic osteosarcoma cells. J. Cell. Biochem. 53, 176 – 180 (1993). 657. T. Kubota, M. Yamauchi, J. Onozaki, S. Sato et al., Influence of an intermittent compressive force on matrix protein expression by
658.
659. 660.
661.
662.
663.
664.
665.
666.
667.
668.
669.
670.
671.
672.
673.
674.
675.
ROS 17/2.8 cells, with selective stimulation of osteopontin. Arch. Oral Biol. 38, 23 – 30 (1993). Kaji, T. Sugimoto, A. Miyauchi, M. Fukase et al., Calcitonin inhibits osteopontin mRNA expression in isolated rabbit osteoclasts. Endocrinology 135, 484 – 487 (1994). S. T. Li and E. Katz, On the state of demineralized matrices of bone and dentin. Calcif. Tissue Res. 22, 275 – 284 (1977). A. Endo, Potential role of phosphoprotein in collagen mineralization: An experimental study in vitro. J. Orthop. Assoc. 61, 563 – 569 (1987). M. J. Glimcher, Mechanisms of calcification in bone: Role of collagen fibrils and collagen-phosphoprotein complexes in vitro and in vivo. Anat. Rec. 224, 139 – 153 (1989). A. Nanci, Content and distribution of noncollagenous matrix proteins in bone and cementum: Relationship to speed of formation and collagen packing density. J. Struct. Biol. 126, 256 – 269 (1999). A. L. Boskey, M. Maresca, W. Ullrich, S. B. Doty, W. T. Butler, and C. W. Prince, Osteopontin-hydroxyapatite interactions in vitro: Inhibition of hydroxyapatite formation and growth in a gelatin-gel. Bone Miner. 22, 147 – 159 (1993). G. K. Hunter, C. L. Kyle, and H. A. Goldberg, Modulation of crystal formation by bone phosphoproteins: Structural specificity of the osteopontin-mediated inhibition of hydroxyapatite formation. Biochem. J. 300, 723 – 728 (1994). G. K. Hunter and H. A. Goldberg, Nucleation of hydroxyapatite by bone sialoprotein. Proc. Natl. Acad. Sci. USA 90, 8562 – 8565 (1993). K. Singh, A. B. Mukherjee, and M. W. DeVouge, Physiological properties and differential glycosylation of phosphorylated and non-phosphorylated forms of osteopontin secreted by normal rat kidney cells. J. Biol. Chem. 265, 18696 – 18701 (1992). A. L. Boskey, M. Maresca, S. Doty, B. Sabsay, and A. Veis, Concentration dependent effect of dentin-phosphophoryn in the regulation of in vitro hydroxyapatite formation and growth. Bone Miner. 11, 55 – 65 (1990). M. Bacarani-Contri, F. Taparelli, and I. Pasquali-Ronchetti, Osteopontin is a constitutive component of normal elastin fibers in human skin and aorta. Matrix Biol., in press. T. Wada, M. D. McKee, S. Steitz, and C. M. Giachelli, Calcification of vascular smooth muscle cell cultures: Inhibition by osteopontin. Circ. Res. 84(2), 166 – 178 (1999). A. Oldberg, A. Franzen, D. Heinegard, M. Pierschbacher, and E. Ruoslahti, Identification of a bone sialoprotein receptor in ostesarcoma cells. J. Biol. Chem. 263, 19433 – 19436 (1988). F. P. Reinholt, K. Hultenby, A. Oldberg, and D. Heinegard, Osteopontin: A possible anchor of osteoclasts to bone surface. Proc. Natl. Acad. Sci. USA 87, 4473 – 4475 (1990). A. Miyauchi, J. Alvarez, E. M. Greenfield, A. Teti, M. Grano, S. Colucci, A. Zambonin-Zallone, F. P. Ross, S. L. Teitelbaum, and D. Cheresh, Binding of osteopontin to the osteoclast integrin alpha v beta 3. Osteoporos Int. 3(Suppl. 1,) 132 – 135 (1993). B. Ek-Rylander, M. Flores, W. Wendel, D. Heinegard, and G. Andersson, Dephosphorylation of osteopontin and bone sialoprotein by osteoclastic tartrate-resistant acid phosphatase: Modulation of osteoclast adhesion in vitro. J. Biol. Chem. 269, 14853 – 14856 (1994). S. Hwang, C. A. Lopez, D. E. Heck, C. R. Gardner, D. L. Laskin, J. D. Laskin, and D. T. Denhardt, Osteopontin inhibits induction of nitric oxide synthase gene expression by inflammatory mediators in mouse kidney epithelial cells. J. Biol. Chem. 269, 711 – 715 (1994). S. H. Ralston, D. Todd, M. Helfrich, N. Benjamin, and P. S. Grabowski, Human osteoblast-like cells produce nitric oxide and express inducible nitric oxide synthase. Endocrinology 135(1), 330 – 336 (1994).
178 676. S. R. Rittling, H. N. Matsumoto, M. D. McKee, A. Nanci, X. R. An, K. E. Novick, A. J. Kowalski, M. Noda, and D. T. Denhardt, Mice lacking osteopontin show normal development and bone structure but display altered osteoclast formation in vitro. J. Bone Miner. Res. 13, 1101 – 1111 (1998). 677. A. L. Boskey and E. P. Paschalis, Matrix proteins and biomineralization. In “Engineering Bone” (J. E. Davies, ed.), in press. 678. D. T. Denhardt and M. Noda, Osteopontin expression and function: Role in bone remodeling. J. Cell. Biochem. Suppl. 30-31, 92 – 102 (1998). 679. L. A. Fitzpatrick, A. Severson, W. D. Edwards, and R. T. Ingram, Diffuse calcification in human coronary arteries: Association of osteopontin with atherosclerosis. J. Clin. Invest. 94, 1597 – 1604 (1994). 680. A. Bellahcene and V. Castronovo, Increased expression of osteonectin and osteopontin, two bone matrix proteins, in human breast cancer. Am. J. Pathol. 146, 95 – 100 (1995). 681. S. Hirota, A. Ito, J. Nagoshi, M. Takeda, A. Kurata, Y. Takatsuka, K. Kohri, S. Nomura, and Y. Kitamura, Expression of bone matrix protein messenger ribonucleic acids in human breast cancers: Possible involvement of osteopontin in development of calcifying foci. Lab. Invest. 72, 64 – 69 (1995). 682. I. Carlson, K. Tognazzi, E. J. Manseau, H. F. Dvorak, and L. F. Brown, Osteopontin is strongly expressed by histiocytes in granulomas of diverse etiology. Lab. Invest. 77,103 – 108 (1997). 683. A. T. Andrews, G. M. Herring, and P. W. Kent, Some studies on the composition of bovine cortical-bone sialoprotein. Biochem J. 104(3), 705 – 715 (1967). 684. L. W. Fisher, S. W. Whitson, L. V. Avioli, and J. D. Termine, Matrix sialoprotein of developing bone. J. Biol. Chem. 258, 12723 – 12727 (1983). 685. A. Oldberg, A. Franzen, and D. Heinegard, Primary structure of bone sialoprotein. J. Biol. Chem. 263, 19430 – 19432 (1988). 686. C. Chenu and P. D. Delmas, Platelets contribute to circulating levels of bone sialoprotein in humans. J. Bone Miner. Res. 7, 47 – 54 (1992). 687. M. A. Thiede, S. L. Smock, D. N. Petersen, W. A. Grasser, D. D. Thompson, and S. K. Nishimoto, Presence of messenger ribonucleic acid encoding osteocalcin, a marker of bone turnover, in bone marrow megakaryocytes and peripheral blood platelets. Endocrinology 135, 929 – 937 (1994). 688. Q. Zhang, C. Domenicucci, H. A. Goldberg, J. L. Wrana, and J. Sodek, J. Biol. Chem. 262, 9702 – 9708 (1990). 689. P. Bianco, L. W. Fisher, M. F. Young, J. D. Termine, and P. Gehron Robey, Expression of bone sialoprotein (BSP) in developing human tissues. Calcif. Tissue Int. 49, 421 – 426 (1991). 690. J. Chen, H. S. Shapiro, and J. Sodek, Developmental expression of bone sialoprotein mRNA in rat mineralized connective tissues. J. Bone Miner. Res. 7, 987 – 997 (1992). 691. J. Chen, M. D. McKee, A. Nanci, and J. Sodek, Bone sialoprotein mRNA expression and ultrastructural localization in fetal porcine calvarial bone: Comparisons with osteopontin. Histochem. J. 26, 67 – 78 (1994). 692. A. Bellahcene, N. Antoine, N. Clausse, E. Tagliabue, L. W. Fisher, J. M. Kerr, P. Jares, and V. Castronovo, Detection of bone sialoprotein in human breast cancer tissue and cell lines at both protein and messenger ribonucleic acid levels. Lab. Invest. 75, 203 – 210 (1996). 693. V. Castronovo and A. Bellahcene, Evidence that breast cancer associated microcalcifications are mineralized malignant cells. Int. J. Oncol. 12, 305 – 308 (1998). 694. B. Ecarot-Charrier, F. Bouchard, and C. Delloye, Bone sialoprotein II synthesized by cultured osteoblasts contains tyrosine sulfate. J. Biol. Chem. 264, 20049 – 20053 (1989). 695. R. J. Midura, D. J. McQuillan, K. J. Benham, L. W. Fisher, and V. C. Hascall, A rat osteogenic cell line (UMR106-01) synthesizes a
GOKHALE, BOSKEY, AND ROBEY
696.
697.
698.
699.
700.
701.
702.
703. 704.
705.
706.
707.
708.
709.
710.
711.
712.
713.
highly sulfated form of bone sialoprotein. J. Biol. Chem. 265, 5285 – 5291 (1990). H. S. Shapiro, J. Chen, J. L. Wrana, Q. Zhang, M. Blum, and J. Sodek, Characterization of porcine bone sialoprotein: Primary structure and cellular expression. Matrix 13, 431 – 440 (1993) R. W. Kinne and L. W. Fisher, Keratan sulfate proteoglycan in rabbit compact bone is bone sialoprotein II. J. Biol. Chem. 262, 10206 – 10211 (1987). L. W. Fisher and H. Schraer, The glycosaminoglycans of estrogeninduced medullary bone in Japanese quail. Arch. Biochem. Biophys. 205, 396 – 403 (1980). M. J. Somerman, L. W. Fisher, R. A. Foster, and J. J. Sauk, Human bone silaoproteins I and II enhance fibroblast attachment in vitro. Calcif. Tissue Int. 43, 50 – 53 (1988). L. W. Fisher, Structure/function studies of the sialoproteins and proteoglycans of bone: It is still the early days. In “The Chemistry and Biology of Mineralized Tissues” (H. Slavkin and P. Price, eds.), pp. 177 – 186. Elsevier, Amsterdam, 1992. A. Oldberg, C. Ryden, A. I. Yacoub, I. Maxe, and D. Heinegard, Specific binding of bone sialoprotein to Staphylococcus aureus isolated from patients with osteomyelitis. Eur. J. Biochem. 184(2), 331 – 336. L. W. Fisher, O. W. McBride, J. D. Termine, and M. F. Young, Human bone sialoprotein: Deduced protein sequence and chromosomal localization. J. Biol. Chem. 265, 2347–2351 (1990). J. J. Li and J. Sodek, Cloning and characterization of the rat bone sialoprotein gene promoter. Biochem. J. 289(Pt 3), 625 – 629 (1993). J. M. Kerr, L. W. Fisher, J. D. Termine, M. G. Wang, O. W. McBride, and M. F. Young, The human bone sialoprotein gene (IBSP): Genomic localization and characterization. Genomics 17, 408 – 415 (1993). R. H. Kim, H. S. Shapiro, J. J. Li, J. L. Wrana, and J. Sodek, Characterization of the human bone sialoprotein (BSP) gene and its promoter sequence. Matrix Biol. 14, 31 – 40 (1994). A. Bianchi, R. D. Wells, N. H. Heintz, and M. S. Caddle, Sequences near the origin of replication of the DHFR locus of Chinese hamster ovary cells adopt left-handed Z-DNA and triplex structures. J. Biol. Chem. 265, 21789 – 21796 (1990). M. F. Young, K. Ibaraki, J. M. Kerr, and A. M. Heegaard, Molecular and cellular biology of the major noncollagenous proteins in bone. In (M. Noda, ed.), pp. 191 – 234. J. M. Kerr, D. R. Hiscock, W. Grzesik, P. G. Robey, and M. F. Young, The human bone sialoprotein gene contains an NF-E1/YY1 cis-acting sequence with putative regulatory activity. Calcif. Tissue Int. 60, 276 – 282 (1997). A. Oldberg, B. Jirskog-Hed, S. Axelsson, and D. Heinegard, Regulation of bone sialoprotein mRNA by steroid hormones. J. Cell Biol. 109, 3183 – 3186 (1989). J. Chen, H. S. Shapiro, J. L. Wrana, S. Reimers, J. N. M. Heersche, and J. Sodek, Localization of bone sialoprotein (BSP) expression to sites of mineralized tissue formation in fetal rat tissues by in situ hybridization. Matrix 11, 133 – 143 (1991). J. Chen, C. A. G. McCulloch, and J. Sodek, Bone sialoprotein in developing porcine dental tissues: Cellular expression and comparison of tissue localization with osteopontin and osteonectin. Arch. Oral Biol. 38, 241 – 249 (1993). G. K. Hunter and H. A. Goldberg, Modulation of crystal formation by bone phosphoproteins: Role of glutamic acid-rich sequences in the nucleation of hydroxyapatite by bone sialoprotein. Biochem. J. 302, 175 – 179 (1994). J. T. Stubbs, Jr., K. P. Mintz, E. D. Eanes, D. A. Torchia, and L. W. Fisher, Characterization of native and recombinant bone sialoprotein: Delineation of the mineral-binding and cell adhesion domains and structural analysis of the RGD domain. J. Bone Miner. Res. 12, 1210 – 1222 (1997).
CHAPTER 4 The Biochemistry of Bone 714. M. E. Flores, M. Norgard, D. Heinegard, F. P. Reinholt, and G. Andersson, RGD-directed attachment of isolated rat osteoclasts to osteopontin, bone sialoprotein, and fibronectin. Exp. Cell Res. 201, 526 – 530 (1992). 715. K. P. Mintz, L. W. Fisher, W. J. Grzesik, V. C. Hascall, and R. J. Midura, Chlorate-induced inhibition of tyrosine sulfation on bone sialoprotein synthesized by a rat osteoblast-like cell line (UMR 106-01 BSP). J. Biol. Chem. 269, 4845 – 4852 (1994). 716. R. Fujisawa and Y. Kuboki, Affinity of bone sialoprotein and several other bone and dentin acidic proteins to collagen fibrils. Calcif. Tissue Int. 51, 438 – 442 (1992). 717. A. L. Arsenault and F. P. Ottensmeyer, Quantitative spatal destributions of calcium, phosphorus and sulfur in calcifying epiphysis by high resolution electron spectroscopic imaging. Proc. Natl. Acad. Sci. USA 80, 1322 – 1326 (1983). 718. A. Bellahcene, M. P. Merville, and V. Castronovo, Expression of bone sialoprotein, a bone matrix protein, in human breast cancer. Anticancer Res. 54, 2823 – 2826 (1994). 719. D. Waltregny, A. Bellahcene, X. de Leval, B. Florkin, U. Weidle, and V. Castronovo, Increased expression of bone sialoprotein in bone metastases compared with visceral metastases in human breast and prostate cancers. J. Bone Miner. Res. 15, 834 – 843 (2000). 720. G. van der Pluijm, H. J. M. Vloedgraven, S. Papapoulos, C. W. G. M. Lowik, W. J. Grzesik, J. M. Kerr, P. Gehron Robey, Attachment characteristics and involvement of integrins in adhesion of breast cancer cell lines to extracellular bone matrix constituents. Lab. Invest. 77(6), 665 – 675 (1997). 721. N. S. Fedarko, B. Fohr, P. G. Robey, M. F. Young, and L. W. Fisher, Factor H binding to bone sialoprotein and osteopontin enables tumor cell evasion of complement-mediated attack. J. Biol. Chem. 275, 16666 – 16672 (2000). 722. J. E. Aubin, A. K. Gupta, R. Zirngbl, and J. Rossant, Knockout mice lacking bone sialoprotein expression have bone abnormalities. J. Bone Miner. Res. 11S, S29 (1996). 723. J. F. Gorski and K. Shimizu, Isolation of new phosphorylated glycoprotein from mineralized phase of bone that exhibits limited homology to adhesive protein osteopontin. J. Biol. Chem. 263, 15938 – 15945 (1988). 724. J. P. Gorski, D. Griffin, G. Dudley, C. Stanford, R. Thomas, C. Huang, E. Lai, B. Karr, and M. Solursh, Bone acidic glycoprotein75 is a major synthetic product of osteoblastic cells and localized as 75- and/or 50-kDa forms in mineralized phases of bone and growth plate and in serum. J. Biol. Chem. 265, 14956 – 14963 (1990). 725. J. P. Gorski, Is all bone the same? Distinctive distributions and properties of non-collagenous matrix proteins in lamellar vs. woven bone imply the existence of different underlying osteogenic mechanisms. Crit. Rev. Oral. Biol. Med. 9, 201 – 223 (1998). 726. A. George, J. Gui , N. A. Jenkins, D. J. Gilbert, N. G. Copeland, and A. Veis , In situ localization and chromosomal mapping of the AG1 (Dmp1) gene. J. Histochem. Cytochem. 42, 1527 – 1531 (1994). 727. Y. Chen, B. S. Bal, and J. P. Gorski, Calcium and collagen binding properties of osteopontin, bone sialoprotein, and bone acidic glycoprotein-75 from bone. J. Biol. Chem. 267, 24871 – 24878 (1992). 728. M. Sato, W. Grasser, S. Harm, C. Fullenkamp, and J. P. Gorski, Bone acidic glycoprotein 75 inhibits resorption activity of isolated rat and chicken osteoclasts. FASEB J. 6, 2966 – 2976 (1992). 729. A. George, B. Sabsay, P. A. L. Simonian, and A. Veis, Characterization of a novel dentin matrix acidic phosphoprotein. J. Biol. Chem. 268, 12624 – 12630 (1993). 730. A. George, R. Silbertsein, and A. Veis, In situ hybridization shows Dmp1 (AG1) to be a developmentally regulated dentin-specific protein produced by mature odontoblasts. Connect. Tissue 43s 67 – 92 (1995).
179 731. R. Srinivasan, B. Chen, J. P. Gorski, and A. George, Recombinant expression and characterization of dentin matrix protein-1. Connect. Tissue Res. 40, 251 – 258 (1999). 732. D. R. Keene, L. Y. Sakai, and R. E. Burgeson, Human bone contains type III collagen, type VI collagen, and fibrillin: Type III collagen is present on specific fibers that may mediate attachment of tendons, ligaments, and periosteum to calcified bone cortex. J. Histochem. Cytochem. 39, 59 – 69 (1991). 733. S. Kitahama, M. A. Gibson, G. Hatzinikolas, S. Hay, J. L. Kuliwaba, A. Evdokiou, G. J. Atkins, and D. M. Findlay, Expression of fibrillins and other microfibril-associated proteins in human bone and osteoblast-like cells. Bone 27, 61 – 67 (2000). 734. K. Kainulainen, L. Karttunen, L. Puhakka, L. Sakai, and L. Peltonen, Mutations in the fibrillin gene responsible for dominant ectopia lentis and neonatal Marfan syndrome. Nature Genet. 6, 64 – 69 (1994). 735. F. Ramirez and L. Pereira, The fibrillins. Int. J. Biochem. Cell. Biol. 31, 255 – 259 (1999). 736. G. M. Corson, S. C. Chalberg, H. C. Dietz, N. L. Charbonneau, and L. Y. Sakai, Fibrillin binds calcium and is coded by cDNAS that reveal a multidomain structure and alternatively spliced exons at the 5 end. Genomics 17, 476 – 484 (1993). 737. D. P. Reinhardt, G. M. Corson, and L. Y. Sakai, Expression and characterization of recombinant fibrillin peptides. In “FECTS Lyon, p. F17, 1994. 738. C. M. Kielty and C. A. Shuttleworth, The role of calcium in the organization of fibrillin microfibrils. FEBS Lett. 336, 323 – 326 (1993). 739. M. Raghunath, C. M. Kielty, K. Kainulainen, A. Child, L. Peltonen, and B. Steinmann, Analyses of truncated fibrillin caused by a 366 bp deletion in the FBN1 gene resulting in Marfan syndrome. Biochem. J. 302 889 – 896 (1994). 740. L. Peltonen and K. Kainulainen, Elucidation of the gene defect in Marfan syndrome: Success by two complementary research strategies. FEBS Lett. 307, 116 – 121 (1992). 741. B. C. Gallagher, L. Y. Sakai, and C. D. Little, Fibrillin delineates the primary axis of the early avian embryo. Dev. Dyn. 196, 70 – 78 (1993). 742. S. L. Dallas, D. R. Keene, S. P. Bruder, J. Saharinen, L. Y. Sakai, G. R. Mundy, and L. F. Bonewald, Role of the latent transforming growth factor beta binding protein 1 in fibrillin-containing microfibrils in bone cells in vitro and in vivo. J. Bone Miner. Res. 15, 68 – 81 (2000). 743. E. J. Mackie, I. Thesleff, and R. Chiquet-Ehrismann, Tenascin is associated with chondrogenic and osteogenic differentiation in vivo and promotes chondrogenesis in vitro. J. Cell Biol. 105, 2569 – 2579 (1987). 744. M. Ekblom, R. Fassler, B. Tomasini-Johansson, K. Nilsson, and P. Ekblom, Downregulation of tenascin expression by glucocorticoids in bone marrow stromal cells and in fibroblasts. J. Cell Biol. 123, 1037 – 1045 (1993). 745. Y. Soini, D. Kamel, M. Apaja-Sakkinen, I. Virtanen, and V. P. Lehto, Tenascin immunoreactivity in normal and pathological bone marrow. J. Clin. Pathol. 46, 218 – 221 (1993). 746. X. Chevalier, N. Groult, B. Larget-Piet, L. Zardi, and W. Hornebeck, Tenascin distribution in articular cartilage from normal subjects and from patients with osteoarthritis and rheumatoid arthritis. Arthritis Rheum. 7, 1013 – 1022 (1994). 747. J. F. Riou, M. Umbhauer, D. L. Shi, and J. C. Boucaut, Tenascin: A potential modulator of cell-extracellular matrix interactions during vertebrate embryogenesis. Biol. Cell 75, 1 – 9 (1992). 748. E. J. Mackie and R. P. Tucker, Tenascin in bone morphogenesis: Expression by osteoblasts and cell type-specific expression of splice variants. J. Cell Sci. 3, 765 – 771 (1992). 749. L. Vakeva, E. Mackie, T. Kantomaa, and I. Thesleff, Comparison of the distribution patterns of tenascin and alkaline phosphatase in
180
750.
751.
752.
753.
754.
755.
756.
757.
758.
759.
760.
761.
762.
763.
764.
765.
766.
GOKHALE, BOSKEY, AND ROBEY developing teeth, cartilage, and bone of rats and mice. Anat. Rec. 228, 69 – 76 (1990). T. Larsson, Y. Sommarin, M. Paulsson, and P. Antonsson, Cartilage matrix proteins: A basic 36-kDa protein with a restricted distribution to cartilage and bone. J. Biol. Chem. 266, 20428 – 20433 (1991). P. J. Neame, Y. Sommarin, R. E. Boynton, and D. Heinegard, The structure of a 38-kDa leucine-rich protein (chondroadherin) isolated from bovine cartilage. J. Biol. Chem. 269(34), 21547 – 21554 (1994). M. Wendel, Y. Sommarin, and D. Heinegard, Bone matrix proteins: Isolation and characterization of a novel cell-binding keratan sulfate proteoglycan (osteoadherin) from bovine bone. J. Cell Biol. 141, 839 – 847 (1998). R. Buchaille, M. L. Couble, H. Macgloire, and F. Bleicher, Expression of the small leucine-rich proteoglycan osteoadherin/osteomodulin in human dental pulp and developing rat teeth. Bone 27, 265 – 276 (2000). K. Horiuchi, N. Amizuka, S. Takeshita, H. Takamatsu, M. Katsuura, H. Ozawa, Y. Toyama, L. F. Bonewald, and A. Kudo, A Identification and characterization of a novel protein, periostin, with restricted expression to periosteum and periodontal ligament and increased expression by transforming growth factor beta. J. Bone Miner. Res. 14, 1239 – 1249 (1999). P. A. Price, A. S. Otsuka, J. W. Poser et al., Characterization of gamma-carboxyglutamic acid containing protein from bone. Proc. Natl. Acad. Sci. USA 73, 1447 (1976). P. V. Hauschka, J. B. Lian, and P. M. Gallop, Direct identification of the calcium-binding amino acid, gamma-carboxyglutamic acid in mineralized tissue. Proc. Natl. Acad. Sci. USA 72(10), 3925 – 3929. P. A. Price, A. S. Otsuka, J. W. Poser, J. Kristaponis, and N. Raman, Characterization of gamma-carboxyglutamic-acid containing protein from bone. Proc. Natl. Acad. Sci. USA 73, 1447 – 1451 (1976). J. D. Fraser and P. A. Price, Lung, heart and kidney express high levels of mRNA for the vitamin K-dependent matrix Gla protein; implications for the possible functions of matrix gla protein and for the tisssue distribution of the -carboxylase, J. Biol. Chem. 263, 11033 – 11036 (1988). P. A. Price, Gla-containing proteins of mineralized tissues. In “Chemistry and Biology of Mineralized Tissues” (H. Slavkin and P. Price, eds.), pp. 169 – 176. Elsevier, Amsterdam, 1992. A. L. J. J. Bronckers, S. Gay, M. T. DiMuzio, and W. T. Butler, Immunolocalization of carboxyglutamic acid containing proteins in developing rat bones. Coll. Relat. Res. 5, 273 – 281 (1985). A. L. J. J. Bronckers, S. Gay, R. D. Finkelman, and W. T. Butler, Developmental appearance of Gla proteins (osteocalcin) and alkaline phosphatase in tooth germs and bone of the rat. Bone Miner. 2, 361 – 373 (1987). A. J. Camarda, W. T. Butler, R. D. Finkelman, and A. Nanci, Immunocytochemical localization of carboxyglutamic acid-containing proteins and osteocalcin in rat bone and dentin. Calcif. Tissue Int. 40, 349 – 355 (1987). J. C. Fleet and J. M. Hock, Identification of osteocalcin mRNA in nonosteoid tissue of rats and humans by reverse transcription-polymerase chain reaction. J. Bone Miner. Res. 9, 1565 – 1573 (1994). P. V. Hauschka and P. V. Reid, Timed appearance of a calciumbinding protein containing carboxyyglutamic acid in developing chick bone. Dev. Biol. 65, 426. Y. Otawara and P. A. Price, Developmental appearance of matrix Gla protein during calcification in the rat. J. Biol. Chem. 261, 10828 (1986). P. A. Price, J. W. Lothringer, and S. K. Nishimoto, Absence of the vitamin K-dependent bone protein in fetal rat mineral: Evidence for another carboxyglutamic acid-containing component in bone. J. Biol. Chem. 255, 2938 – 2942 (1980).
767. P. A. Price, J. W. Lothringer, S. A. Baukol, and A. H. Reddi, Developmental appearance of the vitamin K-dependent proteins in bone metabolism. Annu. Rev. Nutr. 8, 865 (1988). 768. R. Kasai, P. Bianco, P. Gehron Robey, and A. J. Kahn, Production and characterization of an antibody against the human bone GLA protein (BGP/osteocalcin) propeptide and its use in immunocytochemistry of bone cells. Bone Miner. 25, 167 – 182 (1994). 769. L. C. Pan and P. A. Price, The propeptide of rat bone gamma-carboxyglutamic acid protein share homology with other vitamin Kdependent protein precursors. Proc. Natl. Acad. Sci. USA 82(18), 6109 – 6113. 770. J. W. Poser, F. S. Esch, N. C. Ling et al., Isolation and sequence of the vitamin K-dependent protein from human bone: Undercarboxylation of the first glutamic acid residue. J. Biol. Chem. 255, 8685 (1980). 771. P. V. Hauschka and S. A. Carr, Calcium-dependent alpha-helical structure in osteocalcin. Biochemistry 21, 2538 – 2547 (1982). 772. P. V. Hauschka and F. H. Wians, Jr., Osteocalcin-hydroxyapatite interactions in the extracellular organic matrix of bone. Anat. Rec. 224, 180 – 188 (1988). 773. T. L. Dowd, J. F. Rosen, C. M. Gundberg, and R. K. Gupta, The displacement of calcium from osteocalcin at submicrolar concentration of free lead. Biochim. Biophys. Acta 1226, 131 – 137 (1994). 774. D. T. Isbell, S. Du, A. G. Schroering, G. Colombo, and J. G. Shelling, Metal ion binding to dog osteocalcin studied by 1H NMR spectroscopy. Biochemistry 32, 11352 – 11362 (1993). 775. E. Puchacz, J. G. Lian, G. S. Stein, J. Wozney, K. Heubner, and C. Croce, Chromosomal localization of the human osteocalcin gene. Endocrinology 124, 2648 (1989). 776. A. J. Celeste, V. Rosen, and J. L. Buecker, Isolation of the human gene for bone Gla protein. EMBO J. 5, 1885 – 1889 (1986). 777. L. C. Pan and P. A. Price, The propeptide of rat bone gammacarboxyglutamic acid protein shares homology with other vitamin K-dependent protein precursors. Proc. Natl. Acad. Sci. USA 82, 6109 – 6113 (1985). 778. L. C. Pan, M. K. Williamson, and P. A. Price, Sequence of the precursor to rat bone gamma-carboxyglutamic acid protein that assimulates in warfarin-treated osteosarcoma cells. J. Biol. Chem. 260, 13398 – 13401 (1985). 779. S. A. Kerner, R. A. Scott, and J. W. Pike, Sequence elements in the human osteocalcin gene confer basalactivation and inducible response to hormonal vitamin D3. Proc. Natl. Acad. Sci. USA 84, 4455 – 4459 (1989). 780. J. P. Bidwell, A. J. van Wijnen, E. G. Fey, S. Dworetzky, Penman, J. L. Stein, J. B. Lian, and G. S. Stein, Osteocalcin gene promoter binding factors are tissue-specific nuclear matrix components. Proc. Natl. Acad. Sci. USA 90, 3162 – 3166 (1993). 781. M. D. Demay, D. A. Roth, and H. M. Kronenberg, Regions of the rat osteocalcin gene which mediate the effect of 1,25-dihydroxyvitamin D3 on gene transcription. J. Biol. Chem. 264, 2279 – 2282 (1989). 782. M. D. Demay, J. M. Gerardi, H. F. DeLuca, and H. M. Kronenberg, DNA sequences in the rat osteocalcin gene that bind the 1,25-dihydroxyvitamin D3 receptor and confer responsive to 1,25-dihydroxy vitamin D3. Proc. Natl. Acad. Sci. USA 87, 369 – 373 (1990). 783. K. Yoon, S. J. C. Rutledge, R. F. Buenaga, and G. A. Rodan, Characterization of the rat osteocalcin gene: Stimulation of promoter activity by 1,25-dihydroxy vitamn D3. Biochemistry 27, 8521 – 8526 (1988). 784. A. A. J. Heinrichs, R. Bortell, S. Rahman, J. L. Stein, E. S. Alnemri, G. Litwack, J. B. Lian, and SG. S. Stein, Identification of multiple glucocorticoid receptor binding sites in the rat osteocalcin gene promoter. Biochemistry 32, 11436 – 11444 (1993). 785. R. Schule, K. Umesono, D. J. Magelsdorf, J. Bolado, J. W. Pike, and R. M. Evans, Jun-Fos and receptors for vitamins A and D
CHAPTER 4 The Biochemistry of Bone
786.
787.
788.
789.
790.
791.
792.
793.
794.
795.
796.
797.
798.
799.
800.
recognize a common response element in the human osteocalcin gene. Cell 61, 497 – 504 (1990). T. A. Owen, M. A. Aronow, L. M. Baron, B. Bettencourt, G. S. Stein, and J. B. Lian, Pleiotropic effecs of vitamin D on osteoblast gene expresion are related to the prolifrative and differentiatied state of the bone cell phenotype: Dependency upon basal levels of gene expression, duration of exposure and bone matrix competency in noral rat osteoblast cultures. Endocrinology 128, 1496 – 1504 (1991). J. Lian, C. Stewart, E. Puchacz, S. Mackowiak, V. Shalhoub, D. Collart, G. Zambetti, and G. S. Stein, Structure of the rat osteocalcin gene and regulation of vitamin D-dependent expression. Proc. Natl. Acad. Sci. USA 86, 1143 – 1147 (1989). E. R. Markose, J. L. Stein, G. S. Stein, and J. B. Lian, Vitamin Dmediated modifications in protein-DNA interactions a two promoter elements of the osteocalcin gen. Proc. Natl. Acad. Sci. USA 87, 1701 – 1705 (1990). A. A. J. Heinrichs, C. Banerjee, R. Bortell, T. A. Owen, J. L. Stein, G. S. Stein, and J. B. Lian, Identification and characterization of two proximal elements in the rat osteocacin gene promoter that may confer species-specific regulation. J. Cell Biochem. 53, 240 – 250 (1993). A. J. J. Heinrichs, R. Bortell, M. Bourke, J. B. Lian, G. S. Stein, and J. L. Stein, Proximal promoter binding protein contributes to developmental, tissue-restricted expression of the rat osteocalcin gene. J. Cell Biochem. 57, 90 – 100 (1995). H. M. Hoffmann, K. M. Catron, A. J. van Wijnen, L. R. McCabe, J. B. Lian, G. S. Stein, and J. L. Stein, Transcriptional control of the tissue-speciic, developmentally regulated osteocalcin gene requires a binding motif for the Msx family of homeodomain proteins. Proc. Natl. Acad. Sci. USA 91, 12887 – 12891 (1994). R. Frenkel, J. Mijnes, M. A. Aronow, G. Zambetti, C. Banerjee, J. L. Stein, J. B. Lian, and G. S. Stein, Position and orientation-selective silencer in protein-coding sequences of the rat osteocalcin gene. Biochemistry 32, 13636 – 13643 (1993). B. Frenkel, M. Menetcino, J. L. Stein, J. B. Lian, and G. S. Stein, A composite intragenic silencer domain exhibits negative and positive transcriptional control of the bone-specific osteocalcin gene: Promoter and cell type requirements. Proc. Natl. Acad. Sci. USA 91, 10923 – 10927 (1994). P. Ducy, R. Zhang, V. Geoffroy, A. L. Ridall, and G. Karsenty, Osf2/Cbfa1: A transcriptional activator of osteoblast differentiation. Cell 89, 747 – 754 (1997). J. Fraser, Y. Otawara, and P.A. Price, 1,25-dihydroxyvitaine D3 stimulates the synthesis of matrix -carboxyglutaic acid protein by osteosarcoma cells: Mutually exclusive expression of vitamin K-dependent bone proteins by clonal osteoblastic cell lines. J. Biol. Chem. 263, 911 – 916 (1988). C. M. Terpening, C. A. Haussler, P. W. Jurutka, M. A. Galliggan, B. S. Komm, and M. R. Haussler, The vitamin D-responsive element in the rat bone gla protein gene is an imperfect direct repeat that cooperates with other cis-elements in 1,25-dihydroxyvitamin D3-mediated transcriptional activation. Mol. Endocrinol. 5, 373 – 385 (1991). P. Maenpaa, A. Mahonen, and A. Pirskanen, Hormonal regulation of vitamin D receptor levels and osteocalcin synthesis in human osteosarcoma cells. Calcif. Tissue Int. 49, S85 – S86 (1991). N. Pernalete, T. Mori, Y. Nishii, E. Slatopolsky, and A. J. Brown, The activity of 22-oxacalcitriol in osteoblast-like (ROS 17/2.8) cells. Endocrinology 129, 778 – 784 (1991). D. Lajeunesse, G. M. Kiebzak, C. Frondoza, and B. Sacktor, Regulation of osteocalcin secretion by human primary bone cells and by the human osteosarcoma cell line MG-63. Bone Miner. 14, 237 – 250 (1991). A. M. Delany, Y. Dong, and E. Canalis, Mechanisms of glucocorticoid action in bone cells. J. Cell Biochem. 56, 295 – 302 (1994).
181 801. G. Schepmoes, E. Breen, T. A. Owen, M. A. Aronow, G. S. Stein, and J. B. Lian, Infuence of dexamethasone on the vitamin D mediated regulation of the osteocalcin gene expression. J. Cell Biochem. 47, 184 – 196 (1991). 802. A. Pirskanen, T. Jaaskelainen, and P. H. Maenpaa, Effects of transforming growth factor beta 1 on the regulation of osteocalcin synthesis in human MG-63 osteosarcoma cells. J. Bone Miner. Res. 9, 1635 – 1642 (1994). 803. R. T. Ingram, S. K. Bonde, B. L. Riggs, and L. A. Fitzpatrick, Effects of transforming growth factor beta (TGF beta) and 1,25 dihydroxyvitamin D3 on the function, cytochemistry and morphology of normal human osteoblast-like cells. Differentiation 55, 153 – 163 (1994). 804. R. S. Taichman and P. V. Hauschka, Effects of interleukin-1 beta and tumor necrosis factor-alpha on osteoblastic expression of osteocalcin and mineralized extracellular matrix in vitro. Inflammation 16, 587 – 601 (1992). 805. D. B. Evans, R. A. Bunning, and R. G. Russell, The effects of recombinant human interleukin-1 beta on cellular proliferation and the production of prostaglandin E2, plasminogen activator, osteocalcin and alkaline phosphatase by osteoblast-like cells derived from human bone. Biochem. Biophys. Res. Commun. 166, 208 – 216 (1990). 806. D. M. Raab-Cullen, M. A. Thiede, D. N. Petersen, D. B. Kimmel, and R. R. Recker, Mechanical loading stimulates rapid changes in periosteal gene expression. Calcif. Tissue Int. 55, 473 – 478 (1994). 807. J. B. Lian and C. M. Gundberg, Osteocalcin-biochemical considerations and clinical applications. Clin. Orthop. 226, 267 – 291 (1988). 808. P. V. Hauschka, J. B. Lian, D. E. C. Cole, and C. Gundberg, Osteocalin and matrix gla protein: Vitamin K-dependent proteins in bone. Physiol. Rev. 69, 990 – 1047 (1989). 809. L. C. Gerstenfeld, S. D. Chipman, J. Glowacki, and J. B. Lian, Expression to differentiate function by mineralizing cultures of chicken osteoblasts. Dev. Biol. 122, 49 – 60 (1986). 810. A. L. Boskey, F. H. Wians., Jr., and P. V. Hauschka, The effect of osteocalcin on in vitro lipid-induced hydroxyapatite formation and seeded hydroxyapatite growth. Calcif. Tissue Int. 37, 75 (1985). 811. C. G. Groot, J. K. Danes, J. Blok, A. Hoogendijk, and P. V. Hauschka, Light and electron microscopic demonstration of osteocalcin antigenicity in embryonic and adult rat bone. Bone 7, 379 – 385 (1986). 812. J. N. M. Heersche, S. Reimers, J. L. Wrana, M. M. Y. Waye, and A. K. Gupta, Changes in expression of alpha I collagen and osteocalcin mRNA in osteoblasts and odontoblasts at different stages of maturity as shown by in situ hybridization. Proc. Finn. Dent. Soc. 88, 173 – 182 (1992). 813. G. A. Rodan and M. Noda, Gene expression in osteoblastic cells. Crit. Rev. Eukaryot. Gene Expr. 1(2), 85 – 98 (1991). 814. M. A. Aronow, L. C. Gerstenfeld, T. A. Owen, M. Tassinari, G. S. Stein, and J. B. Lian, Factors that promote progressive development of the osteoclast phenotype in cultured rat calvarial cells. J. Cell Physiol. 143, 213 – 221 (1990). 815. W. H. Liggett, J. B. Lian, J. S. Greenberger, and J. Glowacki, Osteocalcin promotes differentiation of osteoclast progenitors from murine long-term bone marrow cultures. J. Cell. Biochem. 55, 190 – 199 (1994). 816. D. J. DeFranco, J. Glowaki, K. A. Cox, and J. B. Lian, Normal bone particles are preferentially resorbed in the presence of osteocalcindeficient bone particles in vitro. Calcif. Tissue Int. 49, 43 (1991). 817. J. D. Malone, S. L. Teitelbaum, G. L. Griffin, R. M. Senior, and A. J. Kahn, Recruitment of osteoclast precursors by purified bone matrix constituents. J. Cell Biol. 92, 227 – 230 (1982). 818. J. Glowacki and J. B. Lian, Impaired recruitment and differentiation of osteoclast progenitors by osteocalcin-depleted bone implants. Cell Differ. 21, 247 (1987).
182 819. J. B. Lian and S. C. Marks, Jr., Osteopetrosis in the rat: Coexistence of reductions in osteocalcin and bone resorption. Endocrinology 126, 955 – 962 (1990). 820. V. Shalhoub, B. Betterncourt, M. E. Jackson, C. A. Mackay, J. Lian, and G. S. Stein, Abnormalities of phosphoprotein gene expression in three osteopetrotic rat mutations: Elevated mRNA transcripts, protein synthesis, and accumulation in bone of mutant animals. J. Cell Physiol. 158, 110 – 120 (1994). 821. J. Bollerslev, S. C. Marks, Jr, L. Mosekilde, and J. B. Lian, Cortical bone osteocalcin content and matrix composition in autosomal dominant osteopetrosis type I. Eur. J. Endocrinol. 130, 592 – 594 (1994). 822. P. A. Price, M. K. Williamson, T. Haba, R. B. Dell, and W. S. S. Jee, Excessive mineralization with growth plate closure in rats on chronic warfarin treatment. Proc. Natl. Acad. Sci. USA 79, 7734 – 7738 (1982). 823. P. Pastoureau, P. Vergnaud, P. J. Meunier, and P. D. Delmas, Osteopenia and bone-remodeling abnormalitities in warfarin-treated lambs. J. Bone Miner. Res. 8, 1417 – 1426 (1992). 824. A. L. Boskey, S. Gadaleta, C. Gundberg, S. B. Doty, P. Ducy, and G. Karsenty, Fourier transform infrared microspectroscopic analysis of bones of osteocalcin-deficient mice provides insight into the function of osteocalcin. Bone 23, 187 – 196 (1998). 825. P. A. Price, M. R. Urist, and Y. Otawara, Matrix gla protein, a new g-carboxyglutamic acid-containing protein which is associated with the organic matrix of bone. Biochem. Biophys. Res. Commun. 117, 765 – 771 (1983). 826. P. A. Price and M. K. Williamson, Primary structure of bovine matrix gla protein, a new vitamin K-dependent bone protein. J. Biol. Chem. 260, 14791 – 14975 (1985). 827. J. E. Hale, J. D. Fraser, and P. A. Price, The identification of matrix gla protein in cartilage. J. Biol. Chem. 263, 5820 – 5824 (1988). 828. G. Luo, P. Ducy, M. D. McKee, G. J. Pinero, E., Loyer, R. R. Behringer, and G. Karsenty, Spontaneous calcification of arteries and cartilage in mice lacking matrix GLA protein. Nature 386, 78 – 81 (1997). 829. J. S. Rice, M. K. Williamson, and P. A. Price. Isolation and sequence of the vitamin K-dependent matrix Gla protein from the calcified cartilage of the soupfin shark. J. Bone Miner. Res. 9, 567 – 576 (1994). 830. G. Luo, R. D’Souza, D. Hogue, and G. Karsenty, The matrix glaprotein gene is a marker of the chondrogenesis cell lineage during mouse development. J. Bone Miner. Res. 10, 325 – 334 (1995). 831. J. E. Hale, M. K. Williamson, and P. A. Price, Carboxyl-terminal proteolytic processing of matrix gla protein. J. Biol. Chem. 266, 21145 – 21149 (1991). 832. P. A. Price, J. S. Rice, and M. K. Williamson, Conserved phosphorylation of serines in the Ser-x-glu/Ser (P) sequences of the vitamin K-dependent matrix gla protein from shark, lamb, rat, cow, and human. Protein Sci. 5, 822 – 830 (1994). 833. M. T. B. Wiedemann and D. Belluoccio, Molecular cloning of avian matrix Gla protein. Biochim. Biophys. Acta 1395, 47 – 49 (1998). 834. M. L. Cancela, C.-L. Hsieh, U. Francke, and P. A. Price, Molecular structure, chromosome assignment and promoter organization of the human matrix Gla protein gene. J. Biol. Chem. 265, 15040 – 15048 (1990). 835. M. L. Cancela and P. A. Price, Retinoic acid induces matrix gla protein gene expression in human cells. Endocrinology 130, 102–108 (1992). 836. S. R. Rannels, M. L. Cancela, E. B. Wolpert, and P. A. Price, Matrix gla protein mRNA expression in cultured type II pneuomocytes. Am. J. Physiol. 265, L270 – L278 (1993). 837. A. L. Boskey, M. D. McKee, and G. P. Karsenty, Mineral characterization of bones and soft tissues in matrix gla protein-deficient mice. In “Chemistry and Biology of Mineralized Tissues” (M. Goldberg, A. Boskey, and C. Robinson, eds.), pp. 63 – 67 (2000).
GOKHALE, BOSKEY, AND ROBEY 838. K. Yagami, J. Y. Suh, M. Enomoto-Iwamoto, E. Koyama, W. R. Abrams, I. M. Shapiro, M. Pacifici, and M. Iwamoto, Matrix GLA protein is a developmental regulator of chondrocyte mineralization and, when constitutively expressed, blocks endochondral and intramembranous ossification in the limb. J. Cell. Biol. 147, 1097 – 1108 (1999). 839. L. Chen, J. P. O’Bryan, H. S. Smith, and E. Liu, Overexpression of matrix Gla protein mRNA in malignant human breast cells: Isolation by differential cDNA hybridization. Oncogene 5, 1391 – 1395 (1990). 840. M. M. Briehl and R. L. Misefeld, Isolation and characterization of transcripts induced by androgen withdrawal and apoptotic cell death in the rat ventral prostate. Mol. Endocrinol. 5, 1381 – 1388 (1991). 841. D. M. Lawton, J. G. Andrew, D. R. Marsh, J. A. Hoyland, and A. J. Freemont, Expression of the gene encoding the matrix gla protein by mature osteoblasts in human fracture non-unions. Mol. Pathol. 52, 92 – 96 (1999). 842. C. Maillard, M. Berruyer, C. M. Serre, M. Dechavane, and P. D. Delmas, Protein S, a vitamin K-dependent protein, is a bone matrix component synthesized and secreted by osteoblasts. Endocrinology 130, 1599 – 1604 (1992). 843. J. Folch-Pi and P. Stoffyn, Proteolipids from membrane systems. Ann. N. Y. Acad. Sci. 195, 86 – 107 (1972). 844. J. Folch-Pi and J. Sakura, Preparation of the proteolipid apo-protein from bovine heart, liver and kidney. Biochim. Biophys. Acta 427, 401 – 427 (1977). 845. J. Ennever, B. Boyan-Salyers, and L. J. Riggan, Proteolipid and bone matrix calcification in vitro. J. Dent. Res. 56, 967 – 970 (1977). 846. J. Ennever, J. J. Vogel, and L. J. Riggan, Phospholipids of a bone matrix calcification nucleator. J. Dent. Res. 57, 731 – 734 (1978). 847. A. L. Boskey, Phospholipids and calcification: An overview. In “Cell-Mediated Calcification and Matrix Vesicles” (S. Yousuf Ali, ed.), pp. 175 – 179. Elsevier, Amsterdam, 1986. 848. B. D. Boyan, Proteolipid-dependent calcification. In “The Chemistry and Biology of Mineralized Tissues” (W. F. Butler, ed.), pp. 125 – 131. EBSCO Media, Birmingham, 1985. 849. B. D. Boyan and N. M. Ritter, Proteolipid-lipid relationships in normal and vitamin D-deficient chick cartilage. Calcif. Tissue Int. 36, 332 – 337 (1984). 850. B. D. Boyan, Z. Schwartz, L. D. Swain, and A. Khare, Role of lipids in calcification of cartilage. Anat. Rec. 224, 211 – 219 (1989). 851. L. N. Wu, T. Yoshimori, B. R. Genge, G. R. Sauer, and R. E. Wuthier, Characterization of the nucleational core complex responsible for mineral induction by growth plate cartilage matrix vesicles. J. Biol. Chem. 268, 25084 – 25094 (1993). 852. R. E. Wuthier, Mechanism of de novo mineral formation by matrix vesicles. Connect. Tissue Res. 22, 27 – 32 (1989). 853. G. J. Barton, R. H. Newman, P. S. Freemont, and M. J. Crumpton, Amino acid sequence analysis of the annexin super-gene family. Eur. J. Biochem. 198, 749 – 760 (1991). 854. J. Johansson, T. Szyperski, T. Curstedt, and K. Wuthrich, The NMR structure of the pulmonary surfactant-associated polypeptide SP-C in an apolar solvent contains a valyl-rich alpha-helix. Biochemistry 33, 6015 – 6023 (1994). 855. S. van Dijk, D. D. Dean, Y. Liu, Y. Zhao, J. M. Chirgwin, Z. Schwartz, and B. D. Boyan, Purification, amino acid sequence, and cDNA sequence of a novel calcium-precipitating proteolipid involved in calcification of corynebacterium matruchotii. Calc. Tissue Int. 62, 350 – 358 (1998). 856. K. Jung, H. Jung, and H. R. Kaback, Dynamics of lactose permease of Escherichia coli determined by site-directed fluorescence labeling. Biochemistry 33, 3980 – 3985 (1994).
CHAPTER 4 The Biochemistry of Bone 857. J. Stim, A. A. Bernardo, F. T. Kear, Y. Y. Qiu, and J. A. Arruda, Renal cortical basolateral Na/HCO3 co-transporter. II. Detection of conformational changes with fluorescein isothiocyanate labeling. J. Memb. Biol. 140, 39 – 46 (1994). 858. M. B. Beest, K. Hoekstra, A. Sein, and D. Hoekstra, Reconstruction of proteolipid protein: Some properties and its role in interlamellar attachment. Biochem. J. 300, 545 – 552 (1994). 859. G. Fernandez-Ballester, J. Castesana, A. M. Fernandez, and J. L. Arrondo, A role for cholesterol as a structural effector of the nicotinic acetylcholine receptor. Biochemistry 33, 4065 – 4071 (1994). 860. B. D. Boyan, W. J. Landis, J. Knight, G. Dereszewski, and J. Zeagler, Microbial hydroxyapatite formation as a model of proteolipiddependent membrane-mediated calcification. Scan. Electr. Microsc. 4, 1793 – 1800 (1984). 861. L. D. Swain, R. D. Renthal, and B. D. Boyan, Resolution of ion translocating proteolipid subclasses active in bacterial calcification. J. Dent. Res. 68, 1094 – 1097 (1989). 862. J. D. Swain and B. D. Boyan, Ion-translocating properties of calcifiable proteolipids. J. Dent. Res. 67, 526 – 530 (1988). 863. C. Montessuit, J. Caverzassio, and J. P. Bonjour, Mechanism of Pi accumulation in maxrip vesecles and its potential role in the calcification process. Bone Miner. 17, 386 (1992). 864. A. L. Boskey, Hydroxyapatite formation in a dynamic gelatin gel system: Effects of type I collagen, lipids, and proteoglycans. J. Phys. Chem. 93, 1628 – 1633 (1989). 865. C. L. Raggio, B. D. Boyan, and A. L. Boskey, In vivo hydroxyapatite formation induced by lipids. J. Bone Miner. Res. 1, 409 – 415 (1986). 866. B. D. Boyan-Salyers and A. L. Boskey, Relationship between protolipids and calcium-phospholipid-phosphate complexes in Bacterionema matruchotii calcification. Calcif. Tissue Int. 30, 167 – 174 (1980). 867. A. L. Boskey and B. D. Boyan-Salyers, Relationship between proteolipids and calcium-phospholipid-phosphate complexes in bacterionema matruchotti calcification. Calcif. Tissue Int. 30, 167 – 174 (1978). 868. A. L. Boskey and A. S. Posner, The role of synthetic and bone extracted calcium-phospholipid-phosphate complexes in hydroxyapatite formation. Calcif. Tissue Res. 23, 251 – 258 (1977). 869. B. D. Boyan and A. L. Boskey, Co-isolation of proteoplipids and calcium-phospholipid-phosphate complexes. Calcif. Tissue Int. 36, 214 – 218 (1984). 870. A. L. Boskey, The role of Ca-phospholipid-PO4 complexes, membrane lipids and matrix vesicles in calcification. In “Matrix Vesicles” (A. Ascenzi, E. Bonnucci, and B. de Bernard, eds.), pp. 161 – 167. Wichtig Editore, Milan Italy, 1981. 871. A. L. Boskey, Phospholipids and biological calcification. In “Proc. First Int. Conf. Chem. Biol. Mineralized Connective Tissues” (A. Veis, ed.), pp. 531 – 537. Elsevier/North Holland, New York, 1981. 872. R. E. Wuthier and S. Gore, Partition of inorganic ion and phospholipids in isolated cell, membrane and matrix vesicle fractions: Evidence for Ca:Pi acidic phospholipid complexes. Calcif. Tissue Res. 24, 163 – 171 (1977). 873. A. Yaari, I. Shapiro, and A. L. Boskey, Ionophoretic activity of a Ca-PL-PO4 complex. Calcif. Tissues Int. 36, 317 – 319 (1984). 874. B. R. Genge, L. N. Wu, H. D. Adkinsson, IV, and R. E. Wuthier, Matrix vesicle annexins exhibit proteolipid-like properties: Seclective partitioning into lipophilic solvents under acidic conditions. J. Biol. Chem. 226, 10678 – 10685 (1991). 875. R. B. Pepinsky, R. Tizard, R. J. Mattaliano, K. L. Sinclair, G. T. Miller, J. L. Browning, E. P. Chow, C. Burne, K. S. Huang, D. Pratt, L. Wachter, C. Hession, A. Z. Frey, and B. P. Wallner, Five distinct calcium and phospholipid binding proteins share homology with lipocortin I. J. Biol. Chem. 263, 10799 – 10811 (1988). 876. M. J. Crumpton and J. R. Dedman, Protein terminology tangle. Nature 345, 212 (1990).
183 877. E. Rojas, H. B. Pollard, H. T. Haigler, C. Parra, and A. L. Burns, Calcium-activated endonexin II forms calcium channels across acidic phospholipids bilayer membranes. J. Biol. Chem. 265, 21207 – 21215 (1990). 878. G. Weber and E. Ferber, Selective inhibition of phospholipase A2 by different lipocortins. Biol. Chem. Hoppe Seyler 371, 725 – 731 (1990). 879. M. J. Geisow, U. Fritsche, J. M. Hexham, B. Dash, and T. Johnson, A consensus amino-acid sequence repeat in Torpedo and mammalian Ca-dependent membrane-binding proteins. Nature 320, 636 – 638 (1986). 880. F. Suarez, B. Rothhut, C. Comera, L. Touqui et al., Expression of annexin I, II, V and VI by rat osteoblasts in primary culture: Stimulation of annexin I expression by dexamethasone. J. Bone Miner. Res. 8, 1201 – 1210 (1993). 881. B. R. Genge, L. N. Wu, and R. E. Wuthier, Differential fractionation of matrix vesicle proteins: Further characterization of the acidic phospholipid-dependent Ca-binding proteins. J. Biol. Chem. 265, 10678 – 10685 (1990). 882. R. E. Wuthier, Mechanism of de novo mineral formation by matrix vesicles. Connect. Tissue Res. 22, 27 – 33 (1989). 883. L. W. Fisher and J. D. Termine, Purification of the noncollagenous proteins from bone: Technical pitfalls and how to avoid them. In “Current Advances in Skeletogenesis” (A. Ornoy, A. Harell, and J. Sela, eds.), pp. 467 – 472. Elsevier, Amsterdam, 1985. 884. I. R. Dickson and M. K. Bagga, Changes with age in the noncollagenous proteins of human bone. Connect. Tissue Res. 14, 77 – 85 (1985). 885. H. Nagase and J. F. J. Woessner, Matrix metalloproteinases. J. Biol. Chem. 274, 21491 – 21494 (1999). 886. S. Gack, R. Vallon, J. Schmidt, A. Grigoriadis, J. Tuckermann, J. Schenkel, H. Weiher, E. F. Wagner, and P. Angel, Expression of interstitial collagenase during skeletal development of the mouse is restricted to osteoblast-like cells and hypertrophic chondrocytes. Cell Growth Differ. 6, 759 – 767 (1995). 887. N. Johansson, U. Saarialho-Kere, K. Airola, R. Herva, L. Nissinen, J. Westermarck, E. Vuorio, J. Heino, and V. M. Kahari, Collagenase-3 (MMP-13) is expressed by hypertrophic chondrocytes, periosteal cells, and osteoblasts during human fetal bone development. Dev. Dyn. 208, 387 – 397 (1997). 888. M. C. Meikle, S. Bord, R. M. Hembry, and J. J. Reynolds, The synthesis of collagenase, gelatinase-A (72 kDa) and -B (95 kDa), and TIMP-1 and -2 by human osteoblasts from normal and arthritic bone. Bone 17, 255 – 260 (1995). 889. J. K. Heath, S. J. Atkinson, M. C. Meikle et al., Mouse osteoblasts synthesize collagenase in response to bone resorbing agents. Biochim. Biophys. Acta 802, 151 (1984). 890. K. Otsuka, J. Sodek, and H. Limeback, Synthesis of collagenase and collagenase inhibitors by osteoblast-like cells in culture. Eur. J. Biochem. 145, 123 (198). 891. D. K. Scott, K. D. Brinckerhoff, J. C. Clohisy, C. O. Quinn, and N. C. Partridge, Parathyroid hormone induces transcription of collagenase in rat osteoblastic cells by a mechanism using cyclic adenosine 3,5-monophosphate and requiring protein synthesis. Mol. Endocrinol. 6, 2153 – 2159 (1992). 892. J. A. Lorenzo, C. C. Pilbeam, J. F. Kalinowski, and M. S. Hibbs, Production of both 92- and 72-kDa gelatinases by bone cells. Matrix 12, 282 – 290 (1992). 893. T. H. Omura, A. Noguchi, C. A. Johanns, J. J. Jeffrey, and N. C. Partridge, Identification of a specific receptor for interstitial collagenase on osteoblastic cells. J. Biol. Chem. 269, 24994 – 24998 (1994). 894. N. Katsura and K. Yamada, Isolation and characterization of a metalloprotease associated with chicken epiphyseal cartilage matrix vesicles. Bone 7, 137 – 143 (1986).
184 895. S. Bord, A. Horner, C. A. Beeton, R. M. Hembry, and J. E. Compston, Tissue inhibitor of matrix metalloproteinase-1 (TIMP-1) distribution in normal and pathological human bone. Bone 24, 229 – 235 (1999). 896. J. M. Dellaise, Y. Eeckout, L. Neff, C. Francois-Gillet, A. Herriet, Sutt, G. Vaes, and R. Baron, (Pro)collagenase (matrix metalloproteinases-1) is present in rodent osteoclasts and in the underlying bone-resorbing compartment. J. Cell Sci. 106, 1071 – 1082 (1993). 897. G. Leloup, J. M. Delaisse, and G. Vaes, Relationship of the plasminogen activator/plasmin cascade to osteoclast invasion and mineral resorption in explanted fetal metatarsal bones. J. Bone Miner. Res. 891 – 902 (1994). 898. K. Holmbeck, P. Bianco, J. Caterina, S. Yamada, M. Kromer, S. A. Kuznetsov, M. Mankani, P. G. Robey, A. R. Poole, I. Pidoux, J. M. Ward, and H. Birkedal-Hansen, MT1-MMP-deficient mice develop dwarfism, osteopenia, arthritis, and connective tissue disease due to inadequate collagen turnover. Cell 99, 81 – 92 (1999). 899. M. C. Meikle, S. Bord, R. M. Hembry, and J. J. Reynolds, Rabbit calvarial osteoblasts in culture constitutively synthesize progelatinase-A, and TIMP1 and TIMP-2. Biochim. Biophys. Acta 1224, 99 – 102 (1994). 900. A. T. Vernillo, N. S. Ramamurthy, H. M. Lee, and B. R. Rifkin, Parathyroid hormone regulation of matrix degra 75, 3 ding enzymes in rat osteoblastic osteosarcoma 17/2.8 cells. Res. Commun. Chem. Pathol. Pharmacol. 323 – 339 (1992). 901. F. S. Panagakos and S. Kumar, Modulation of proteases and their inhibitors in immortal human osteoblast-like cells by tumor necrosis factor-alpha in vitro. Inflammation 18, 243 – 265 (1994). 902. S. Varghese, S. Rydziel, J. J. Jeffrey, and E. Canalis, Regulation of interstitital collagenase expression and collagen degradation by retinoic acid in bone cells. Endocrinology 134, 2438 – 2444 (1994). 903. W. Bode, F. X. Gomis-Ruth, and W. Stockler, Astacins serralysins, snake venom and matrix metalloproteinases exhibit identical zincbinding environments (HEXXHSSGXXH and Met-turn) and topologies and should be grouped into a common family, the “metzincins,” FEBS Lett. 331, 134 – 140 (1993). 904. P. Reinemer, F. Grams, R. Huber, and T. Kleine, Structural implications for the role of the N-terminus in the “Superactivatin” of collagenases. A crystallographic study. FEBS Lett. 338, 227 – 233 (1994). 905. W. Bode, P. Reinemer, R. Huber, and T. Kleine, The X-ray crystal structure of the catalytic domain of human neutrophil collagenase inhibited by a substrate analogue reveals the essentials for catalysis and specificity. EMBO J. 13, 1263 – 1269 (1994). 906. D. B. McKay, M. M. Thayer, K. M. Flaherty, H. Pley, and D. Benvegnu, Crystallograpic structures of the elastase of Pseudomonas aeruginosa. Matrix Suppl. 1, 112 – 115 (1992). 907. J. C. Spurlino, A. M. Smallwood, D. D. Carlton, T. M. Banks, K. J. Vavra, J. S. Johnson, E. R. Cook, J. Falvo, R. C. Wahl, and T. A. Pulvino, 1.56 A structure of mature truncated human fibroblast collagenase. Proteins 19(2), 98 – 109 (1994). 908. M. F. French, A. Bhown, and H. E. Van Wart, Identification of Clostridium collagenase hyperactive sites in type I, II, and III collagens: Lack of correlation with local triple helical stability. J. Protein Chem. 11, 83 – 97 (1992). 909. H. Wu, X. Liu, M. Bynne, Y. Harada, S. M. Krane, and R. Jaenisch, Collagenases are critical for skeletal and soft connective tissue remodeling : Abnormalities in mice that express, in the germline, an engineered mutation in the Col 1a1 gene that encodes collagenase resistance. J. Bone Miner Res. 9, S141(1994). 910. V. Everts, J. M. Delaisse, W. Korper, A. Niehof, G. Vaes, and W. Beertsen, Degradation of collagen in the bone-resorbing compartment underlying the osteoclast involves both cysteineproteinases and matrix metalloproteases. J. Cell. Physiol. 150, 221 – 231 (1992).
GOKHALE, BOSKEY, AND ROBEY 911. G. Vaes, J. M. Delaisse, and Y. Eeckhout, Relative roles of collagenase and lysosomal cysteine-proteinases in bone resorption. Matrix Suppl. 1, 383 – 388 (1992). 912. P. A. Hill, D. J. Buttle, S. J. Jones, A. Boyde, M. Murata, J. J. Reynolds, and M. C. Meikle, Inhibition of bone resorption by selective inactivators of cysteine proteinases. J. Cell Biochem. 45, 118–130 (1994). 913. J. P. Montenez, J. M. Delaisse, P. M. Tulkens, and B. K. Kishore, Increased activities of cathepsin B and other lysosomal hydrolases in fibroblasts and bone tissue cultured in the presence of cysteine proteinases inhibitors. Life Sci. 55, 1199 – 1208 (1994). 914. J. J. Wu, M. W. Lark, L. E. Chun, and D. R. Eyre, Sites of stromelysin cleavage in collagen types II, IX, X, and XI of cartilage. J. Biol. Chem. 266, 5625 – 5628 (1991). 915. C. Niyibizi, R. Chan, J. J. Wu, and D. Eyre, A 92 kDA gelatinase (MMP-9) cleavage site in native type V collagen. Biochem. Biophys. Res. Commun. 202, 328 – 333 (1994). 916. A. M. Bollen and D. R. Eyre, Direct extraction of gelatinases from rat bone. Connect. Tissue Res. 29, 223 – 230 (1993). 917. M. Gowen, F. Lazner, R. Dodds, R. Kapadia, J. Feild, M. Tavaria, I. Bertoncello, F. Drake, S. Zavarselk, I. Tellis, P. Hertzogm, C. Debouck, and I. Kola, Cathepsin K knockout mice develop osteopetrosis due to a deficit in matrix degradation but not demineralization. J. Bone Miner. Res. 14, 1654 – 1663 (1999). 918. P. Saftig, E. Hunziker, V. Everts, S. Jones, A. Boyde, O. Wehmeyer, A. Suter, and K. von Figura, Functions of cathepsin K in bone resorption: Lessons from cathepsin K deficient mice. Adv. Exp. Med. Biol. 477, 293 – 303 (2000). 919. K. Hoekman, C. W. Lîwik, M. van de Ruit, O. L. Bijvoet, J. H. Verheijen, and S. E. Papapoulos, The effect of tissue type plasminogen activator (tPA) on osteoclastic resorption in embryonic mouse long bone explants: A possible role for the growth factor domain of tPA. Bone Miner. 17, 1 – 13 (1992). 920. T. J. Martin, E. H. Allan, and S. Fukumoto, The plasminogen activator and inhibitor system in bone remodelling. Growth Regul. 3, 209 – 214 (1993). 921. J. A. Hamilton, S. R. Lingelbach, N. C. Partridge et al., Stimulation of plasminogen activator in osteoblast like cells by bone-resorbing hormones. Biochem. Biophys. Res. Commun. 122, 230 (1984). 922. B. D. Catherwood, L. Titus, C. O. Evans, J. Rubin, S. D. Boden, and M. S. Nanes, Increased expression of tissue plasminogen activator messenger ribonucleic acid is an immediate esponse to parathyroid hormone in neonatal rat osteoblasts. Endocrinology 134, 1429 – 1436 (1994). 923. S. Fukumoto, E. H. Allan, J. A. Yee, T. D. Gelehrter, and T. J. Martin, Plasminogen activator regulation in osteblasts: Parathyroid hormone inhibition of type-1 plasminogen activator inhibitor and its mRNA. J. Cell. Physiol. 152, 346 – 355 (1992). 924. J. Pfeilschifter, J. Erdmann, W. Schmidt, A. Naumann, H. W. Minne, and R. Ziegler, Differential regulation of plasminogen activator and plasminogen activator inhibitor by osteotropic factors in primary cultures of mature osteoblasts and osteoblast precursors. Endocrinology 126, 703 – 711 (1990). 925. S. Fukumoto, E. H. Allan, and T. J. Martin, Regulation of plasminogen activator inhibitor-1 (PAI-1) expression by 1,25-dihydroxyvitamin D-3 in normal and malignant rat osteoblasts. Biochim. Biophys. Acta 1201, 223 – 228 (1994). 926. S. L. Cheng, V. Shen, and W. A. Peck, Regulation of plasminogen activator and plasminogen activator inhibitor production by growth factors and cytokines in rat calvarial cells. Calcif. Tissue Int. 49, 321 – 327 (1991). 927. F. W. Fawthrop, B. O. Oyajobi, R. A. Bunning, and R. G. Russell, The effect of transforming growth factor beta on the plasminogen activator activity of normal human osteoblast-like cells and a human osteosarcoma cell line MG-63. J. Bone Miner. Res. 7, 1363 – 1371 (1992).
CHAPTER 4 The Biochemistry of Bone 928. E. H. Allan, R. Zeheb, T. D. Gelehrter, J. H. Heaton, S. Fukumoto, J. A. Yee, and T. J. Martin, Transforming growth factor beta inhibits plasminogen activator (PA) activity and stimulates production of urokinase-type PA, PA inhibitor-1 mRNA, and protein in rat osteoblast-like cells. J. Cell. Physiol. 149, 34 – 43 (1991). 929. S. Fukumoto, E. H. Allan, R. Zeheb, T. D. Gelehrter, and T. J. Martin, Glucocorticoid regulation of plasminogen activator inhibitor-1 messenger ribonucleic acid and protein in normal and malignant rat osteoblasts. Endocrinology 130, 797 – 804 (1992). 930. Y. Mikuni-Takagaki and M. J. Glimcher, Post-translational processing of chicken bone phosphoproteins: Identification of the bone phosphoproteins of embryonic tibia. Biochem. J. 268, 585 – 591 (1990b). 931. Y. Mikuni-Takagaki and M. J. Glimcher, Post-translational processing of chicken bone phosphoproteins. Identification of bone (phospho)protein kinase. Biochem. J. 268, 593 – 597 (1990a). 932. C. Sfeir and A. Veis, Casein kinase localization in the enodplasmic reticulum of the RO”S 17/2.8 cell line. J. Bone Miner. Res. 10, 607 – 615 (1995). 933. Y. Suzuki, A. Yamaguchi, T. Ikeda, T. Kawase, S. Saito, and Y. Mikuni-Takagaki, In situ phosphorylation of bone and dentin proteins by the casein kinase II-like enzyme. J. Dent. Res. 77, 1799 – 1806 (1998). 934. A. Veis, C. Sfeir, and C. B. Wu, Phosphorylation of the proteins of the extracellular matrix of mineralized tissues by casein kinase-like activity. Crit. Rev. Oral. Biol. Med. 8, 360 – 379 (1997). 935. E. H. Fischer, H. Charbonneau, and N. K. Tonks, Protein tyrosine phosphatases: A diverse family of intracellular and transmembrane enzymes. Science 253, 401 – 420 (1991). 936. E. Salih, S. Ashkar, L. C. Gerstenfeld, and M. J. Glimcher, Identification of the phosphorylated sites of metabolically 32P-labeled osteopontin from cultured chicken osteoblasts. J. Biol. Chem. 272, 13966 – 13973 (1997). 937. L. Rifas, S. Cheng, L. R. Halstead, A. Gupta, K. A. Hruska, and L. V. Avioli, Skeletal casein kinase activity defect in the HYP mouse. Calcif. Tissue Int. 61, 256 – 259 (1997). 938. C. B. Huggins, The formation of bone under the influence of epithelium of the urinary tract. Arch. Surg. 22, 377 – 408 (1931). 939. M. R. Urist, Bone: Formation by autoinduction, Science 150, 893 – 899 (1965). 940. M. R. Urist, R. J. De Lange, and A. M. Finerman, Bone cell differentiation and growth factors. Science 220, 680 – 686 (1983). 941. J. M. Wozney, Bone morphogenetic proteins and their gene expression. In “Cellular Molecular Biology of Bone,” pp. 131 – 168. Academic Press, New York, 1993. 942. A. H. Reddi and N. S. Cunningham, Initiation and promotion of bone differentiation by bone morphogenetic proteins. J. Bone Miner. Res. 8, S499 – S502 (1993). 943. S. Cheifetz, BMP receptors in limb and tooth formation. Crit. Rev. Oral. Biol. Med. 10, 182 – 198 (1999). 944. J. M. Wozney, The bone morphogenetic protein family: Multifunctional cellular regulators in the embryo and adult. Eur. J. Oral. Sci. 106, 160 – 166 (1998). 945. J. M. Wozney and V. Rosen, Bone morphogenetic protein and bone morphogenetic protein gene family in bone formation and repair. Clin. Orthop. 346, 26 – 37 (1998). 946. S. Daopin, M. Li, and D. R. Davies, Crystal structure of TGF-beta 2 refined at 1.8 A resolution. Proteins 17, 176 – 192 (1993). 947. S. J. Archer, A. Bax, A. B. Roberts, M. B. Sporn, Y. Ogawa, K. A. Piez, J. A. Weatherbee, M. L. Tsang, R. Lucas, and B. L. Zheng, Transforming growth factor beta 1: Secondary structure as determined by heteronuclear magnetic resonance spectroscopy. Biochemistry 32(4), 1164 – 1171 (1993). 948. J. L. Wrana, L. Attisano, R. Wieser, F. Ventura, and J. Massague, Mechanism of activation of the TGF-ß receptor. Nature 370, 341 – 346 (1994).
185 949. D. M. Kingsley, What do BMPs do in mammals? Clues from the mouse short-ear mutation. Trends Genet. 10, 16 – 21 (1994). 950. P. H. Frances, M. K. Richardson, P. M. Brickell, and C. Ticle, Bone morphogenetic proteins and a signalling pathway that controls patterning in the developing chick limb. Development 120, 209 – 218 (1994). 951. F. S. Kaplan, J. A. Tabas, and M. A. Zasloff, Fibrodysplasia ossificans proressiva: A clue from the fly? Calcif. Tissue Int. 47, 117 – 125 (1989). 952. J. L. Christian, BMP, Wnt and Hedgehog signals: How far can they go? Curr. Opin. Cell. Biol. 12, 244 – 249 (2000). 953. P. V. Hauschka, A. E. Mavrakos, M. D. Iafrati, S. E. Doleman, and M. Klagsbrun, Growth factors in bone matrix: Isolation of multiple types by affinity chromatography on heparin-sepharose. J. Biol. Chem. 261, 12665 – 12674 (1986). 954. E. Canalis, T. McCarthy, and M. Centrella, Isolation of growth factors from adult bovine bone. Calcif. Tissue Int. 43, 346 – 351 (1988). 955. D. J. Baylink, R. D. Finkelman, and S. Mohan, Growth factors to stimulate bone formation. J. Bone Miner. Res. 8, S565 – S572 (1993). 956. E. Solheim, Growth factors in bone. Int. Orthop. 22, 410 – 416 (1998). 957. C. J. Rosen, L. R. Donahue, and S. J. Hunter, Insulin-like growth factors and bone: The osteoporosis connection. Proc. Soc. Exp. Biol. Med. 206, 83 – 102 (1994). 958. D. T. Graves, A. J. Owen, R. K. Barth, P. Tempst, A. Winoto, L. Fors, and L. E. Hood, Detection of c-sis transcripts and synthesis of PDGF-like proteins by human osteosarcoma cells. Science 265, 972 (1984). 959. Y. Ishimi, C. Miyaura, C. H. Jin, T. Akatsu, T. Abe, Y. Nakamura, A. Yamaguichi, S. Yoshiki, T. Matsuda, T. Hirano, T. Kishimoto, and T. Suda, IL-6 is produced by osteoblasts and induces bone resorption. J. Immunol. 145, 3297 – 3303 (1990). 960. P. E. Keeting, L. Rifas, S. A. Harris, D. S. Colvard, T. C. Spelsberg, W. A. Peck, and B. L. Riggs, Evidence for interleukin-1ß production by cultured normal human osteoblast-like cells. J. Bone Miner. Res. 6, 827 – 833 (1991). 961. E. C. Weir, K. L. Insogna, and M. C. Horowitz, Osteoblast-like cells secrete granulocyte-macrophage colony-stimulating factor in response to parathyroid hormone and lipopolysaccaride. Endocrinology 124, 899 – 904 (1989). 962. M. Gowen, K. Chapman, A. Littlewood, D. Hughes, D. Evans, and G. Russell, Production of tumor necrosis factor by human osteoblasts is modulated by other cytokines, but not by osteotropic hormones. Endocrinology 126, 1250 – 1255 (1990). 963. G. R. Mundy, Immune system and bone remodeling. Trends Endocrin. Metab. 1, 307 – 311 (1990). 964. J. D. Bjorge, T. O. Chan, M. Antcza, H. J. Kung, and D. J. Fujita, Activated type I phosphoinostol kinase is associated with the epidermal growth factor (EGF) receptor following EGF stimulation. Proc. Natl. Acad. Sci. USA 87, 3816 – 3820 (1990). 965. M. Berridge, The molecular basis of communication within the cell. Sci. AM 253, 142 – 152 (1985). 966. M. Centrella, T. L. McCarthy, and E. Canalis, Growth factors and cytokines. In “Bone” (B. Hall, ed.), Vol. 4, pp. 47 – 73. CRC Press, Boca Raton, FL, 1992. 967. T. Yoneda, Cytokines in bone: Local translators in cell-to-cell communications. In “Cellular and Molecular Biology of Bone” (M. Noda, ed.), pp. 375 – 412. Academic Press, New York, 1993. 968. P. D. Delmas, R. P. Tracy, B. L. Riggs et al., Identification of the noncollagenous proteins of bovine bone by two-dimensional gel electrophoresis. Calcif. Tissue Int. 36, 308 (1984). 969. I. R. Dickson, A. R. Poole, and A. Veis, Localization of plasma 2HS-glycoprotein in mineralizing human bone. Nature 256, 430 – 432 (1975).
186 970. B. A. Ashton, H.-J. Hohling, and J. T. Triffit, Plasma proteins present in human cortical bone: Enrichment of the 2HS-glycoprotein. Calcif. Tissue Res. 22, 27 – 33 (1976). 971. R. A. Terkeltaub, D. A. Santoro, G. Mandel, and N. Mandel, Serum and plasma inhibit neutrophil stimulation by hydroxyapatite crystals: Evidence that serum alpha 2-HS glycoprotein is a potent and specific crystal-bound inhibitor. Arthritis Rheum. 31, 1081 – 1089 (1988). 972. K. Aoyama and J. Chiba, Separation of different molecular forms of mouse IgA and IgM monoclonal antibodies by high-performance liquid chromatography on spherical hydroxyapatite beads. J. Immunol. Methods 162, 201 – 210 (1993). 973. H. Iwase, K. Gondo, T. Koike, and I. Ono, Novel precolumn deproteinization method using a hydroxyapatite cartridge for the determination of theophylline and diazepam in human plasma by highperformance liquid chromatography with ultraviolet detection. J. Chromatogr. B. Biomed. Appl. 655, 73 – 81 (1994). 974. H. E. Schultze, K. Heide, and H. Haupt, Charterisierung eines niedermolekularen a2 mukoids aus human serum. Natur-wissenschaften. 49, 440–453 (1962). 975. F. Yang, Z. Schwartz, L. Swain, C.-C. Lee, B. Bowman, and B. Boyan, 2-HS-glycoprotein: Expression in chondrocytes and augmentation of alkaline phosphatase and phospholipase A2 activity. Bone 12, 7 – 15 (1991). 976. T. Ohnishi, N. Arakaki, O. Nakamura, S. Hirono, and Y. Daikuhara, Purification, characterization, and studies on biosynthesis of a 59-kDA bone sialic acid-containing protein (BSP) from rat mandible using a monoclonal antibody: Evidence from rat mandible using a monoclonal antibody that 59-kDa BSP may be the rat counterpart of human 2-HS glycoprotein and is synthesized by both hepatocytes and osteoblasts. J. Biol. Chem. 266, 14636 – 14645 (1991). 977. T. Ohnishi, O. Nakamura, M. Ozawa, N. Arakaki, T. Muramatsu, and Y. Daikuhara, Molecular cloning and sequence analysis of cDNA for a 59kD bone sialoprotein of the rat: Demonstration that it is a counterpart of human a2-HS glycoprotein and bovine fetuin. J. Bone Miner Res. 8, 367 – 377 (1993). 978. M. Mizuno, M. C. Farach-Carson, G. J. Pinero, R. Fujisawa, J. C. Brunn, J. M. Seyer, G. R. Bousfield, M. P. Mark, and W. T. Butler, Identification of the rat bone 60K acidic glycoprotein as 2HSglycoprotein. Bone Miner. 13, 1 – 12 (1991). 979. F. Gejyo, J. L. Chang, W. Burgi, K. Schmid, G. D. Offner, and R. H. Troxler, Characterization of the ß-chain of human plasma a2HS-glycoprotein. J. Biol. Chem. 258, 4966 – 4971 (1983). 980. R. H. Troxler and K. Schmid, The complete aminoacid sequence of the A-chain of human plasma 2-HS glycoprotoein. J. Biol. Chem. 261, 1665 – 1676 (1986). 981. C. C. Lee, B. H. Bowman, and F. Yang, Human 2-HS-glycoprotein: The A and B chains with a connecting sequence are encoded by a single mRNA transcript. Proc. Natl. Acad. Sci. USA 84, 4403 – 4407 (1987). 982. K. Matsushima, M. Cheng, and S. Migita, Purification and physicochemical characterization of human alpha 2-HS-glycoprotein. Biochim. Biophys. Acta 701, 200 – 205 (1982). 983. T. Araki, Y. Yoshioka, and K. Schmid , The position of the disulfide bonds in human plasma ? 2 HS-glycoprotein and the repeating double disulfide bonds in the domain structure. Biochim. Biophys. Acta 994, 195 – 199 (1989). 984. J. Kellerman, H. Haupt, E.-A. Auerswald, and W. Muller-Esterl, The arrangement of disulfide loops in human 2-HS glycoprotein: Similarity to the disulfide bridge structures of cystatins and kininogens. J. Biol. Chem. 264, 14121 – 14128 (1989). 985. D. L. Christie, K. M. Dziegielewska, R. M. Hill, and N. R. Sanders, Fetuin: The bovine homologue of human 2HS glycoprotein. FEBS Lett. 214, 226 – 227 (1987).
GOKHALE, BOSKEY, AND ROBEY 986. M. Haasemann, P. Nawratil, and W. Muller-Esterl, Rat tyrosine kinase inhibitor shows sequence similarity to human alpha 2-HS glycoprotein and bovine fetuin. Biochem. J. 274, 899 – 902 (1991). 987. T. Hayase, K. G. Rice, K. M. Dziegielewska, M. Kuhlenschmidt et al., Comparison of N-glycosides of fetuins from different species and human alpha 2-HS-glycoprotein. Biochemistry 31, 4915 – 4921 (1992). 988. C. Binkert, M. Demetriou, B. Sukhu, M. Szweras, H. C. Tenenbaum, and J. W. Dennis, Regulation of osteogenesis by fetuin. J. Biol. Chem. 274, 28514 – 28520 (1999). 989. C. C. Lee, B. H. Bowman, and F. M. Yang, Human alpha 2-HS-glycoprotein: The A and B chains with a connecting sequence are encoded by a single mRNA transcript. Proc. Natl. Acad. Sci. USA 84(13), 4403 – 4407 (1987). 990. J. G. Lewis and C. M. Andre, A serum DNA-binding protein in malignant disease. FEBS Lett. 92, 211 (1978). 991. J. G. Lewis and C. M. Andre, Effect of human a2HS-glycoprotein on mouse macrophage function. Immunology 3, 329 (1980). 992. P. R. Srinivas, A. S. Wagner, L. V. Reddy, D. D. Deutsch et al., Serum alpha 2-HS-glycoprotein is an inhibitor of the human insulin receptor at the tyrosine kinase level. Mol. Endocrinol. 7, 1145 – 1155 (1993). 993. W. P. Bradley, A. P. Blasco, J. F. Weiss, J. C. Alexander, N. A. Silverman, and P. B. Chretien, Correlations among serum proteinbound carbohydrates, serum glycoprotein, lymphocyte reactivity and tumor burden in cancer patients. Cancer 40, 2264 – 2275 (1977). 994. C. J. van Oss, C. F. Gillman, P. M. Bronson, and J. R. Border, Opsonic-properties of human serum 2-HS-glycoprotein. Immun. Commun. 3, 329 – 335 (1974). 995. F. Yang, Z. L. Chen, J. M. Bergeron, R. L. Cupples, and W. E. Friedrichs, Human 2-HS-glycoprotein/bovine fetuin homoluge in mice: Identification and developmental retulation of the gene. Biochim. Biophys. Acta 1130, 149 – 156 (1992). 996. I. R. Dickson, M. K. Bagga, and C. R. Paterson, Variations in the serum concentration and urine excretion of 2-HS-glycoprotein, a bone-related protein, in normal individuals and in patients with osteogenesis imperfecta. Calcif. Tissue Int. 35, 16 – 20 (1983). 997. Quelch, W. G. Cole, and R. A. Melick, Noncollagenous proteins in normal and pathological human bone. Calcif. Tissue Int. 36, 545 – 549 (1985). 998. J. T. Triffit, U. Gehauer, B. A. Ashton, M. H. Owen, and J. J. Reynolds, Origin of plasma 2-HS-glycoprotein and its accumulation in bone. Nature 262, 226 – 227 (1976). 999. N. C. Blumenthal, F. Betts, and A. S. Posner, Effect of carbonate and biological macromolecules on formation and properties of hydroxyapatite. Calcif. Tissue Res. 18, 81 – 90 (1975). 1000. W. Jahnen-Dechent, T. Schinke, A. Trindl, W. Muller-Esterl, F. Sablitzky, S. Kaiser, and M. Blessing, Cloning and targeted deletion of the mouse fetuin gene. J. Biol. Chem. 272, 31496 – 31503 (1997). 1001. G. C. Colclasure, W. S. Lloyd, M. Lamkin, W. Gonnerman, R. F. Troxler, G. D. Offner, W. Burgi, K. Schmid, and R. B. Nimberg, Human serum 2HS-glycoprotein modulates in vitro bone resorption. J. Clin. Endocrinol. Metab. 66, 187 – 192 (1988). 1002. J. Garnett and P. Dieppe, The effects of serum and human albumin on calcium hydroxyapatite crystal growth. Biochem. J. 266, 863 – 868 (1990). 1003. H. Gilman and D. W. Hukins, Seeded growth of hydroxyapatite in the presence of dissolved albumin. J. Inorg. Biochem. 55, 21 – 30 (1994). 1004. H. Gilman and D. W. Hukins, Seeded growth of hydroxyapatite in the presence of dissolved albumin at constant composition. J. Inorg. Biochem. 55, 31 – 39 (1994). 1005. H. Furedi-Milhofer, Adsorption of human serum albumin on precipitated hydroxyapatite. J. Colloid. Interface Sci. 69, 460 – 480 (1979).
CHAPTER 4 The Biochemistry of Bone 1006. M. J. Mura-Galelli, J. C. Voegel, S. Behr, E. F. Bres, and P. Schaaf, Adsorption/desorption of human serum albumin on hydroxyapatite: A critical analysis of the Langmuir model. Proc. Natl. Acad. Sci. USA 88, 5557 – 5561 (1991). 1007. Y. Yoshiuka, F. Gejyo, T. Marti, E. E. Ricki, W. Burgi, G. D. Offner, S. Shimabayashi, Y. Tanizawa, and K. Ishida, Effect of phosphorylated organic compound on the adsorption of bovine serum albumin by hydroxyapatite. Chem. Pharm. Bull. 39, 2183 – 2188 (1991). 1008. H. Iwase, K. Gondo, T. Koike, and I. Ono, Novel precolumn deproteinization method using a hydroxyapatite cartridge for the determination of theophylline and diazepam in human plasma by high-performance liquid hromotography with ultraviolet detection. J. Chromotogr. B Biomed. Appl. 665, 73 – 81 (1994). 1009. A. L. Boskey, V. Vigorita, and P. G. Bullough, Calcium-acidic phospholipid-phosphate complexes: Promoter of mineralization common to pathologic hydroxyapatite-containing calcifications. Am. J. Pathol. 133, 22 – 29 (1988). 1010. G. Gronowicz, F. N. Woodiel, M. B. McCarthy, and L. G. Raisz, In vitro mineralization of fetal rat parietal bones in defined serum-free medium: Effect of ß-glyceral phosphate. J. Bone Miner. Res. 4, 313 – 324 (1989). 1011. D. D. Lee and M. J. Glimcher, Three-dimensional spatial relationship between the collagen fibrils and the inorganic calcium phosphate crystals of pickeral (Americanus americanus) and herring (Clupea harengus) bone. J. Mol. Biol. 217, 487 – 501 (1991). 1012. S. Weiner, Organization of extracellularly mineralized tissues: A comparative study of biological crystal growth. CRC Crit. Rev. Biochem. 20, 365 – 408 (1986). 1013. E. Dmitrovsky and A. L. Boskey, Calcium acidic phospholipid phosphate complexes in human atherosclerotic aorta. Calcif. Tissue Int. 37, 121 – 125 (1985). 1014. J. M. Burnell, D. J. Baylink, C. H. Chestnut, III, and E. J. Teubner, The role of skeletal calcium deficiency in postmenopausal osteoporosis. Calcif. Tissue Int. 38, 187 – 192 (1986). 1015. A. E. N. Nielsen, “Kinetics of Precipitation.” Pergammon Press, Oxford, 1964. 1016. G. H. Nancollas and B. Tomazic, Growth of calcium phosphate on hydroxyapatite crystals: Effect of supersaturation and ionic medium. J. Phys. Chem. 78, 2218 – 2225 (1974). 1017. A. L. Boskey and A. S. Posner, Conversion of amorphous calcium phosphate to microcrystalline hydroxyapatite: A pH-dependent, solution-mediated, solid-solid conversion. J. Phys. Chem. 77, 2313 – 2317 (1973). 1018. W. T. Brown and L. C. Chow, Chemical properties of bone mineral. Annu. Rev. Mater. Sci. 6, 213 – 262 (1976). 1019. E. D. Eanes and J. L. Meyer, The maturation of crystalline calcium phosphates in adqueous suspensions at physiologic pH. Calcif. Tissue Res. 23, 259 – 269 (1977). 1020. W. E. Brown, N. Eidelman, and B. Tomazic, Octacalcium phosphate as a precursor in biomineral formation. Adv. Dent. Res. 1, 306 – 313 (1987). 1021. L. J. Bréceviç and H. Füredi-Milhofer, Precipitation of calcium phosphates from electrolyte solutions. II. The formation and transformation of precipitates. Calcif. Tissue Res. 10, 82 – 90 (1972). 1022. J. Moradian-Oldak, S. Weiner, L. Addadi, W. J. Landis, and W. Traub, Electron imaging and diffraction study of individual crystals of bone, mineralized tendon and synthetic carbonate apatite. Connect. Tissue Res. 25, 219 – 228 (1991). 1023. H. Füredi-Milhofer, J. Moradian-Oldak, S. Weiner, A. Veis, K. P. Mintz, and L. Addadi, Interactions of matrix proteins from mineralized tissues with octacalcium phosphate. Connect. Tissue Res. 30, 251 – 264 (1994). 1024. D. Hanein, B. Geiger, and L. Addadi, Fibronectin adsorption to surfaces of hydrated crystals: An analysis of the importance of bound water in protein– substrate interactions. Langmuir 9, 1058 – 1065 (1993a).
187 1025. J. Aizenberg, S. Albeck, S. Weiner, and L. Addadi, Crystal – protein interactions studied by overgrowth of calcite on biogenic skeletal elements. J. Crystal Growth 142, 156 – 164 (1994). 1026. D. Hanein, H. Sabanay, L. Addadi, and B. Geiger, Selective interactions of cells with crystal surfaces: Implications for the mechanism of cell adhesion. J. Cell Sci. 104, 275 – 288 (1993). 1027. A. Linde, A. Lussi, and M. A. Crenshaw, A mineral induction by immobilized polyanionic proteins. Calcif. Tissue Int. 44, 286 – 295 (1989). 1028. G. K. Huneter, In vitro studies on matrix-mediated mineralization. In “Bone Metabolism and Mineralization” (B. K. Hall, ed.), pp. 225 – 247. CRC Press, Boca Raton, FL, 1992. 1029. B. R. Heywood and E. D. Eanes, An ultrastructural study of the effects of acidic phospholipid substitutions on calcium phosphate precipitation in anionic liposomes. Calcif. Tissue Int. 50, 149 – 156 (1992). 1030. E. D. Eanes, A. W. Hailer, R. J. Midura, and V. C. Hascall, Proteoglycan inhibition of calcium phosphate precipitation in liposomal suspensions. Glycobiology 2, 571 – 578 (1992). 1031. D. Skrtic and E. D. Eanes, Effect of different phospholipid-cholesterol membrane compositions on liposome-mediated formation of calcium phosphates. Calcif. Tissue Int. 50, 253 – 260 (1992a). 1032. D. Skrtic and E. D. Eanes, Membrane-mediated precipitation of calcium phosphate in model liposomes with matrix vesicle-like lipid composition. Bone Miner. 16, 109 – 119 (1992). 1033. B. Ganss, R. H. Kim, and J. Sodek, Bone sialoprotein. J. Crit. Rev. Oral Biol. Med. 10, 79 – 98 (1999). 1034. A. L. Boskey, L. Spevak, S. B. Doty, and L. Rosenberg, Effects of bone CS-proteoglycans, DS-decorin, and DS-biglycan on hydroxyapatite formation in a gelatin gel. Calc. Tissue Int. 61, 298 – 305 (1997). 1035. K. Sato, Y. Kumagai, and J. Tanaka, Apatite formation on organic monolayers in simulated body environment. J. Biomed. Mater. Res. 50, 16 – 20 (2000). 1036. A. L. Boskey, B. D. Boyan, S. B. Doty, A. Feliciano, K. Greer, D. Weiland, L. D. Swain, and Z. Schwartz, Studies of matrix vesicleinduced mineraliztion in a gelatin gel. Bone Miner. 17, 257 – 263 (1992). 1037. P. Gehron Robey, Cell-mediated mineralization in vitro. In “Mineralization in Biological Systems” (E. Bonucci, ed.), pp. 107 – 127. CRC Press, Boca Raton, FL, 1992. 1038. H. I. Roach, Induction of normal and dystrophic mineralization by glycero-phosphates in long-term bone organ culture. Calcif. Tissue Int. 50, 553 – 562 (1992). 1039. H. Fell, The histogenesis of cartilage and bone in the long bones of the embryonic fowl. J. Morpol. 40, 417 – 459 (1925). 1040. H. I. Roach, Long-term organ culture of embryonic chick femurs: A system for investigating bone and cartilage formation at an intermediate level of organization. J. Bone Miner. Res. 5, 85 – 100 (1990). 1041. H. C. Tenenbaum and J. N. M. Heersche, Differentiation of osteoblasts and formation of mineralized bone in vitro. Calcif. Tissue Int. 34, 76 – 79 (1982). 1042. H. C. Tenenbaum and J. N. M. Heersche, Dexamethasone stimulates osteogenesis in chick periosteum in vitro. Endocrinology 117, 2211 – 2217 (1985). 1043. A. J. Friedenstein, K. V. Petrakova, A. I. Kurolesova, and G. P. Frolova, Heterotopic transplants of bone marrow: Analysis of precursor cells for osteogenic and hematopoietic tissues. Transplant 6, 230 – 247 (1968). 1044. M. Owen, Marrow stromal stem cells. J. Cell. Sci. S10, 63 – 76 (1988). 1045. B. A. Ashton, T. D. Allen, C. R. Howlett, C. C. Eaglesom, A. Hattori, and M. Owen, Formation of bone and cartilage by marrow stromal cells in diffusion chambers in vivo. Clin. Orthop. Relat. Res. 151, 294 – 307 (1980).
188 1046. D. Benayahu, A. Fried, D. Zipori, and S. Weintroub, Subpopulations of marrow stromal cells share a variety of osteoblastic markers. Calcif. Tissue Int. 49, 202 – 207 (1991). 1047. J. N. Beresford, Osteogenic stem cells and the stromal system of bone and marrow. Clin. Orthop. Relat. Res. 240, 270 – 280 (1989). 1048. A. E. Grigoriadis, J. N. M. Heersche, and J. E. Aubin, Differentation of muscle fat cartilage and bone from progenitor cells present in a bone derived clonal cell population: Effects of dexamethasone. J. Cell. Biol. 106, 2139 – 2151 (1988). 1049. C. G. Bellows and J. E. Aubin, Determination of numbers of osteoprogenitors present in isolated fetal rat calvaria cells in vitro. Dev. Biol. 133, 8 – 13 (1989). 1050. C. G. Bellows, J. E. Aubin, and J. N. M. Heersche, Physiological concentrations of glucocorticoids timulate formation of bone nodules from isolated rat calvaria cells in vitro. Endocrinology 121, 1985 – 1992 (1989). 1051. A. Yamaguichi, T. Katagiri, T. Ikeda, J. M. Wozney, V. Rosen, E. A. Want, A. J. Kahn, T. Suda, and S. Yoshiki, Recombinant human bone morphogenetic protein-2 stimulates osteoblastic maturation and inhibits myogenic differentiation in vitro. J. Cell Biol. 113, 681 – 687 (1991). 1052. S. Sakamoto, M. Sakamoto, L. Goldberg, L. Colarusso, and Y. Gotoh, Mineralization induced by beta-glycerophosphate in cultures leads to a marked incrased in collagenase synthesis by mouse osteogenic MC3T3-E1 cells under subsequent stimulation with heparin. Biochem. Biophys. Res. Commun. 162, 773 – 780 (1989). 1053. R. T. Franceschi, C. J. Linson, T. C. Peter, and P. R. Romano, Regulation of cellular adhesion and fibronectin synthesis by 1 alpha,25dihydroxyvitamin D3. J. Biol. Chem. 262(9), 4165 – 4171 (1987). 1054. S. B. Rodan, Y. Imai, M. A. Thiede, G. Wesolowski, D. Thompson, Z. Bar-Shavit, S. Shull, K. Mann, and G. A. Rodan, Characterization of a human osteosarcoma cell line (Saos-2) with osteoblastic properties. Cancer Res. 47(18), 4961 – 4966 (1987). 1055. H. Sudo, H. A. Kodama, Y. Amagai, S. Yamamoto, and S. Kasai, In vitro differentiation and calcification in a new clonal osteogenic cell line derived from newborn mouse calvaria. J. Cell Biol. 96(1), 191 – 198 (1983). 1056. J. K. Heath, S. B. Rodan, K. G. Yoon, and G. A. Rodan, Rat calvarial cell lines immortalized with SV-40 large T-antigen: Constitutive and retinoic acid-inducible expression of osteoblastic features. Endocrinology 124, 3060 – 3068 (1989). 1057. P. E. Keeting, R. E. Scott, D. S. Colvard, M. A. Anderson, M. J. Oursler, T. C. Spelsberg, and B. L. Riggs, Development and characterization of a rapidly proliferating, well-differentiated cell line derived from normal adult human osteoblast-like cells transfected with SV40 large T antigen. J. Bone Miner. Res. 7(2), 127 – 136 (1992). 1058. S. A. Harris, R. J. Enger, B. L. Riggs, and T. C. Spelsberg, Development and characterization of a conditionally immortalized human fetal osteoblastic cell line. J. Bone Miner. Res. 10(2), 178 – 186 (1995). 1059. T. S. Owen, M. S. Aronow, V. Shaloub, D. M. Barone, L. Wiling, M. Tassinari, M. B. Kennedy, S. Pockwinse, J. B. Lian, and G. S.
GOKHALE, BOSKEY, AND ROBEY
1060.
1061.
1062.
1063.
1064.
1065.
1066. 1067. 1068.
1069. 1070. 1071. 1072.
1073.
1074.
1075.
Stein, Progressive development of the rat osteoblast phenotype in vitro: Reciprocal relationships in expression of genes association with osteoblast proliferation and differentiation during formation of the bone extraellular matris. J. Cell. Physiol. 143, 420 – 430 (1990). K. Turksen and J. E. Aubin, Positive and negative immunoselection for enrichment of two classes of osteoprogenitor cells. J. Cell Biol. 114(2), 373 – 384 (1991). B. Ecarot-Charrier, N. Shepard, G. Charette, M. Grynpas, and F. H. Glorieux, Mineralization in osteoblast cultures: A light and electron microscopic study. Bone 9, 147 – 154 (1988). L. C. Gerstenfeld, S. D. Chipman, C. M. Kelly, K. J. Hodgens, and D. D. Lee, Collagen expression, ultrastructural assembly, and mineralization in cultures of chick embryo osteoblasts. J. Cell Biol. 106, 979 – 989 (1988). J. R. Nefussi, M. L. Boy-Lefebre, H. Boulekbache, and N. Forest, Mineralization in vitro of matrix formed by osteoblasts isolated by collagenase digestion. Differentiation 29, 160 – 168 (1985). M. Weinreb, D. Shinar, and G. A. Rodan, Different pattern of alkaline phosphatase, osteopontin, and osteocalcin expression in developing rat bone visualized by in situ hybridization. J. Bone Miner. Res. 5, 831 – 842 (1990). S. W. Whitson, M. A. Whitson, D. E. Bowers, Jr., and M. C. Falk, Factors influencing synthesis and mineralization of bone matrix from fetal bovine bone cells grown in vitro. J. Bone Miner. Res. 7(7), 727 – 741 (1992). H. C. Anderson, Electron microscopic studies of induced cartilage development and calcification. J. Cell Biol. 35, 81 (1967). E. Bonnuci, Fine structure of early cartilage calcification. J. Ultrastruct. Res. 201, 33 (1967). G. W. Bernard and D. C. Pease, An electron microscopic study of initial intramembranous osteogenesis. Am. J. Anat. 125, 271 – 290 (1969). S. Y. Ali, Analysis of matrix vesicles and their role in the calcification of epiphyseal cartilage. Fed. Proc. 35, 135 – 142 (1976). H. C. Anderson, Matrix vesicle calcification. Metab. Bone Dis. Rel. Res. 1, 83 (1978). H. C. Anderson, Biology of disease: Mechanism of mineral formation in bone. Lab. Invest. 60, 320 – 331 (1989). R. E. Wuthier, L. N. Wu, G. R. Sauer, and B. R. Genge, Mechanism of matrix vesicle calcification: Characterization of ion channels and the nucleational core of growth plate vesicles. Bone Miner. 17, 290 – 295 (1992). R. E. Wuthier, Electrolytes of isolated epiphyseal chondrocytes, matrix vesicles, and extracellular fluid. Calcif. Tissue Res. 23, 125 – 133 (1977). T. Kirsch and R. E. Wuthier, Stimulation of calcification of growth plate cartilage matrix vesicles by binding to type II and X collagens. J. Biol. Chem. 269, 11462 – 11469 (1994). L. N. Wu, B. R. Genge, G. C. Lloyd, and R. E. Wuthier, Collagenbinding proteins in collagenase-released matrix proteins and different types of collagen. J. Biol. Chem. 266, 195 – 203 (1991).
CHAPTER 5
Developmental Biology of Bone ANTHONY M. REGINATO,*† WENFANG WANG,* AND BJORN R. OLSEN* *Department of Cell Biology, Harvard Medical School, Boston, Massachusetts 02115, and † Arthritis Unit, Massachusetts General Hospital, Boston, Massachusetts 02114
I. II. III. IV. V.
Introduction Craniofacial Bone Development Axial Bone Development Limb Bone Development Sonic Hedgehog and Indian Hedgehog in Skeletal Development VI. Bone Morphogenetic Proteins, Homologues, and Their Antagonists in Skeletal Development
VII. Fibroblast Growth Factors and Their Receptors in Skeletal Development VIII. Chondrocyte Differentiation and Endochondral Bone Formation IX. Osteoblast Differentiation and Function X. Cell – Matrix Interactions in Skeletal Development XI. Summary and Perspectives References
I. INTRODUCTION
initial German descriptions) of the future bones (Fig. 1, see also color plate). The cartilage anlagen is replaced by bone during endochondral ossification. During this process, chondrocytes in the center of cartilage undergo a program of hypertrophy and apoptosis and the matrix calcifies and is invaded by blood vessels, osteoblasts, and osteoclasts. Gradually, the cartilage is replaced by bone and bone marrow. At the same time, differentiation of cells to osteoblasts in the surrounding perichondrium leads to the formation of a sleeve of bone that surrounds the developing bone marrow space. As the bone marrow space expands toward the ends of the anlagen, the process of chondrocyte hypertrophy, apoptosis, and vessel ingrowth continues at the ends, defining what becomes the epiphyseal growth plate (Fig. 2, see also color plate). Growth plates are the centers for longitudinal bone growth and their cartilage is composed of at least three types of cells: (1) resting chondrocytes, (2) proliferating
The development of the vertebrate skeleton depends on the regulated differentiation, function, and interactions of its cellular components. The cells are derived from three sources: cranial neural crest, paraxial mesoderm, and lateral mesoderm. The cranial neural crest gives rise to most of the craniofacial skeleton, the paraxial mesoderm forms the axial skeleton, and the lateral plate mesoderm generates the appendicular skeleton. Mesenchymal cells from these three compartments undergo condensation at the future sites of cartilage or bone and differentiate into chondrocytes or osteoblasts depending on the location within the developing skeleton. In membranous bones, such as the calvaria and the mandible, mesenchymal cells differentiate directly into osteoblasts, whereas in endochondral bones, mesenchymal cells differentiate into chondrocytes that produce cartilage models (frequently called anlagen, based on
OSTEOPOROSIS, SECOND EDITION VOLUME 1
189
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
190
REGINATO, WANG, AND OLSEN
FIGURE 1 Diagram illustrating how mesenchymal cell condensations give rise to membranous bones (e.g., in calvaria) by direct differentiation of osteoblasts, bone matrix production, and bone growth and remodeling, or to endochondral bone (e.g., in long, tubular bones) by differentiation of chondrocytes, production of cartilage models (anlagen), followed by replacement of the cartilage by bone, and bone growth and remodeling. In some cases (e.g., the mandible) membranous bone is formed around a cartilage model (in the case of the mandible, this cartilage is the Meckel’s cartilage), but does not replace the cartilage as in endochondral ossification. (See also color plate.)
FIGURE 2
Proximal end (left) and proximal growth plate region (right) of tibia from a 17-day old mouse. The growth plate, with layers of proliferating and hypertrophic cartilage, separates the bone and marrow of the secondary ossification center in the epiphysis from the trabecular bone and marrow in the diaphysis of the tibia. Courtesy of Dr. D. Glotzer. (See also color plate.)
191
CHAPTER 5 Developmental Biology of Bone
FIGURE 3 In situ hybridization showing expression of Col10a1 in hypertrophic chondrocytes of growth plates of tibia and metatarsal bones from a mouse embryo at day 17.5 of development. Courtesy of Dr. E. Zelzer. (See also color plate.)
chondrocytes, and (3) hypertrophic chondrocytes, which are larger and surrounded by a mineralized matrix containing type X collagen as a specific marker (Fig. 3, see also color plate). Hypertrophic chondrocytes die through a process of apoptosis, and osteoblasts, which are brought in with invading blood vessels, form trabecular bone under the growth plate. Ultimately, all cartilage in anlagen is replaced by bone with the exception of the articular surfaces of joints and costal cartilage. Current understanding of the molecular pathways that control patterning, growth, and differentiation of the various skeletal elements has been derived from a combination of genetics, experimental developmental biology, and in vitro cellular and biochemical studies. Knowledge of embryonic skeletal development will provide a better basis for generating strategies to repair cartilage and bone in patients with skeletal diseases such as osteoporosis.
II. CRANIOFACIAL BONE DEVELOPMENT Craniofacial bones arise from two distinct lineages of skeletogenic mesenchyme: neural crest and paraxial mesoderm. Neural crest cells from the caudal midbrain and rhombomeres 1, 2, and 4 produce cartilage and bones of the face, jaws, and the front and sides of the brain case (frontal, temporal and parietal bones) [1 – 3]. The paraxial mesoderm gives rise to the posterior parts of the head skeleton and the skull base. Most of the craniofacial bones, such as the calvaria and jaws, are formed by membranous bone formation. During this process, osteoblasts are formed
by direct differentiation from mesenchymal cells (Fig. 1). Condensed mesenchymal cells form a multilayered membrane and osteogenesis begins within the core of the membrane and radiates outward. In some bones, such as the mandible, membranous bone formation occurs around a core of cartilage (Meckel’s cartilage) (Fig. 1). The transcription factor Cbfa1 is required for osteoblast differentiation and is thus essential for both membranous and endochondral bone formation [4 – 6]. Disruption of Cbfa1 in mice leads to a lack of osteoblast differentiation and complete absence of bone [5,6]. Craniofacial bone formation is dynamic and complex and no simple molecular pattern has emerged from studies of gene knockouts in mice. Inactivation of genes for transcription factors such as gsc (murine homologue of Drosophila goosecoid [7]), Otx2 (homologue of Drosophila orthodenticle – otd [8,9]), Mf1 (forkhead winged-helix transcription factor [10]), Dlx1, 2, and 5 [11,12], and Pax3 [13] affects proliferation, differentiation, and/or migration of neural crest cells and cause malformation or absence of certain craniofacial bones. In humans, abnormalities that affect neural crest cells, and consequently patterning of craniofacial bones, include several types of Waardenburg syndrome (Table 1) [14]. Types 1 and 3 of the syndrome, caused by mutations in the transcription factor PAX3, are characterized by a combination of craniofacial bone and soft tissue abnormalities, partial albinism, hearing loss, spina bifida, cleft lip/palate, and scapular anomalies [15]. In the calvarium, osteogenic fronts from neighboring skeletal elements meet at the cranial sutures. During early postnatal life these sutures remain open to allow the cranial vault to grow and expand to accommodate the enlarging
192
REGINATO, WANG, AND OLSEN
TABLE 1
Human Genetic Disorders and Their Associated Genes
Human syndrome Waardenburg syndrome type 1 (WS1)
Gene(s) affecteda PAX3
Waardenburg syndrome type 3 (WS3)
PAX3
Spondylocostal dysostosis (SCDO1)
DLL3
Synpolydactyly (SPD)
HOXD13
Pallister – Hall syndrome (PHS)
GLI3
Postaxial polydactyly type A (PAP-A)
GLI3
Greig cephalopolysyndactyly syndrome (GCPS)
GLI3
Brachydactyly (BD)
CDMP1/GDF5
Acromesomelic dysplasia Hunter – Thompson type (AMDH)
CDMP1/GDF5
Acromesomelic dysplasia Grebe type (AMDG)
CDMP1/GDF5
Proximal symphalangism (SYM1)
NOG
Multiple synostosis syndrome 1 (SYNS1)
NOG
Pfeiffer syndrome
FGFR1, FGFR2
Crouzon syndrome (CS)
FGFR2
Apert syndrome
FGFR2
Jackson – Weiss syndrome (JWS)
FGFR2
Beare – Stevenson syndrome
FGFR2
Saethre – Chotzen syndrome (SCS)
TWIST
Muenke syndrome
FGFR3
Crouzon syndrome with acanthosis nigricans
FGFR3
Achondroplasia (ACH)
FGFR3
Hypochondroplasia (HCH)
FGFR3
Thanatophoric dysplasia (TD)
FGFR3
Campomelic dysplasia (CMD1)
SOX9
Jansen metaphyseal chondrodysplasia
PTHrPR
Blomstrand chondrodysplasia
PTHrPR
Schmid metaphyseal chondrodysplasia (MCDS)
COL10A1
Cleidocranial dysplasia (CCD)
CBFA1
Osteogenesis imperfecta (OI)
COL1A1, COL1A2
Achondrogenesis type II (ACGII)
COL2A1
Hypochondrogenesis
COL2A1
Spondyloepiphyseal dysplasia congenita (SEDC)
COL2A1
Kniest dysplasia
COL2A1
brain. Premature fusion of the sutures results in craniosynostosis. Mutations in the transcription factors TWIST [16,17] and MSX2 [18] and in the fibroblast growth factor receptors FGFR1, 2, and 3 [19] cause various types of human craniosynostoses, including Saethre – Chotzen, Pfeiffer, Apert, and Crouzon syndromes (see later).
III. AXIAL BONE DEVELOPMENT The axial skeleton of vertebrates is composed of segmental units of vertebrae, intervertebral discs, and ribs. This metameric pattern can be traced back to the somites of early embryos. Somites, paired and morphologically distinct segmental units, are formed on either side of the neural tube and notochord [20]. The process of somitogenesis begins at the head level and continues in a craniocaudal sequence. The segmentation is accompanied by regionalization along the craniocaudal axis, leading to specific segmental identities (e.g., a thoracic vertebra can be distinguished easily from a lumbar vertebra). The dorsolateral part of somites gives rise to the epithelial dermomyotome, whereas the ventromedial part forms the sclerotome [21]. Sclerotomal cells differentiate into chondrocytes that make the vertebral bodies (Fig. 4, see also color plate). A vertebra is formed by the combination of the caudal half of one sclerotome and the rostral half of the next sclerotome. This process is called resegmentation [21]. In N-cadherin-null mutant mice, each epithelial somite is split into rostral and caudal halves, whereas in N-cadherin/cad11 double mutants, more fragmented somites are observed [22]. This suggests that cells in the rostral and caudal compartments of a somite have distinct adhesive affinities and that both N-cadherin and cad11 play roles in holding cells in these two compartments together before resegmentation.
Stickler syndrome Type 1 (STL1)
COL2A1
Type 2 (STL2)
COL11A1
Type 3 (STL3)
COL11A2
Marshall syndrome
COL11A1
Multiple epiphyseal dysplasia (MED)
COL9A2, COL9A3, COMP, DTDST
a For gene notations we have followed the standard of using capital letters for human genes and a capital letter followed by lowercase letters for mouse genes.
FIGURE 4
Diagram illustrating migration of sclerotomal cells from their somitic origin toward the notochord. Modified from Mundlos and Olsen [34]. (See also color plate.)
193
CHAPTER 5 Developmental Biology of Bone
The segmentation of paraxial mesoderm into somites is regulated by Notch signaling pathways. The transmembrane receptor Notch1 is strongly expressed in presomitic mesoderm, and mouse embryos carrying null alleles of Notch1 die during gestation with a delay in the transition from presomitic mesoderm to epithelial somites [23,24]. Expression of the Notch ligand Delta-like (Dll1) starts in the paraxial mesoderm, in presomitic mesoderm and posterior halves of somites [25]. In Dll1-deficient mouse embryos, the primary metameric presomites are established, but the segments have no craniocaudal polarity and no epithelial somites form [26]. Mutations in a downstream effector of Notch signaling, lunatic fringe, also affect somite formation. In lunatic fringe-deficient mice, the boundaries between individual somites fail to form, resulting in a severely disorganized axial skeleton [27,28]. The Mesp2 gene [29] and presenilin genes (PS1 and PS2) have been shown to regulate Notch signaling [30]. Disruption of Mesp2 alters the expression of Dll1 and abolishes the expression of Notch1 and Notch2 [30]. Presenilin genes are required for the spatiotemporal expression of Notch1 and Dll1 [31]. Mice with disrupted Dll3 alleles have vertebral and rib defects that are similar to many human abnormalities of the axial skeleton [32], and mutations in DLL3 have been described in patients with spondylocostal dysostosis [33]. The signaling molecule sonic hedgehog (Shh) controls proliferation and differentiation of cells in the sclerotomes. Shh is expressed in the notochord and the floor plate of the neural tube (Fig. 5A, see also color plate). It promotes sclerotome formation and represses the formation of the dermomyotome, as shown by induction of the sclerotomal marker Pax1 and repression of the dermomyotomal marker Pax3 in vitro and in vivo [35]. Shh-null mice lack most of the sclerotomal derivatives: the vertebral column and the posterior portion of the ribs [36]. Therefore, Shh is critical for development of the axial skeleton. More detailed studies indicate that Shh is not required for the initial induction of the sclerotome, as Shh-null mice display close to normal expression of molecular markers (Pax1, Myf5, and Pax3) for sclerotome, myotome, and dermomyotome [36]. Instead, Shh signaling appears to be required for maintenance of these markers [35,37]. One of the genes that Shh induces and maintains the expression of is the paired-box gene Pax1 [35]. In mice, expression of Pax1 can be detected in sclerotomal cells from embryonic day 8.5 onward [38] (Fig. 5B, see also color plate). These Pax1-positive cells are committed to chondrogenesis. Mice that are homozygous for either mutated (undulated) or inactivated Pax1 are defective in sclerotome differentiation and formation of the vertebral column [38,39]. In the absence of Pax1, the gradual loss of Sox9 and Col2a1 expression in the segmented mesenchyme prevents the sclerotomes from undergoing chondrogenesis [40].
FIGURE 5
(A) Whole mount in situ hybridization showing expression of Sonic hedgehog in the notochord of a mouse at day 9.5 of development. (B) Whole mount in situ hybridization showing expression of the transcription factor Pax1 in the sclerotomes of a mouse embryo at day 9.5 of development. Courtesy of Dr. S. Mundlos. (See also color plates.)
Early craniocaudal identities of the axial skeleton are determined by the expression of different Hox genes [41 – 43]. Almost all segments, such as vertebrae and their attached muscles, differ in size, shape, and structure. This regionalization can be visualized by specific patterns of Hox gene expression or “Hox codes.” Changes in the Hox code lead to a shifting of the regional borders and axial identities called homeotic transformations [44,45].
IV. LIMB BONE DEVELOPMENT Forelimbs and hindlimbs have distinct morphology; however, they share a basic skeletal pattern. A single, proximal long bone (humerus in forelimb and femur in hindlimb) articulates distally with a pair of long bones (ulna and radius in the forelimb and tibia and fibula in the hindlimb). The long bones are followed by a series of
194
FIGURE 6
Diagram illustrating migration of cells into limb bud mesenchyme from somites and lateral plate mesoderm. Modified from Mundlos and Olsen [34]. (See also color plate.)
carpals (forelimb) or tarsals (hindlimb) and digits. This complex three-dimensional structure is patterned along three main axes: proximal – distal (shoulder to digits), anterior – posterior (thumb to small finger), and dorsal – ventral (back of hand to palm). During embryonic development, paired limb buds appear along the lateral body wall, covered at the distal end by a layer of tall epithelial cells called the apical ectodermal ridge (AER). Mesenchymal cells underneath the AER have two separate origins: somites and lateral plate mesoderm (Fig. 6, see also color plate). Lateral plate mesodermal cells give rise to cartilage, tendons, and other connective tissues, whereas cells that originate from the somites migrate into limb buds and form myogenic cells of the muscles [46,47]. The limb axes are established and maintained by different sets of signaling molecules. Factors involved along one axis may sometimes regulate genes responsible for a different axis, thereby making processes along the three axes intimately linked. The proximal – distal axis is primarily controlled by fibroblast growth factors (FGFs) (such as Fgf4 and Fgf8). These factors, released by cells of the AER, stimulate the proliferation of cells in the underlying mesenchyme, the progress zone [48]. The ventral – dorsal axis is controlled by Wnt7a [49] and En-1 [50], and the anterior – posterior axis is regulated by Shh [51]. Shh is expressed by mesenchymal cells in the posterior region of the limb bud and defines the zone of polarizing activity (ZPA) [51]. A positive feedback loop is established between the AER and the ZPA. Fgf activity is required to maintain Shh expression by the ZPA, and Shh induces Fgf expression in the AER [52]. The cross-talk between Shh and Fgfs may be mediated by the Bmp antagonist gremlin (see later) [53]. An upstream regulator of Shh, the basic helix – loop – helix transcription factor dHand, may control Shh expression and establishment of the ZPA [54,55].
REGINATO, WANG, AND OLSEN
Mice lacking the Shh gene have no distal skeletal elements in the limbs; nevertheless, the more proximal skeletal elements (humerus in the forelimbs and femur in the hindlimbs) do develop [36]. The homeobox genes Meis1 and Meis2 have been shown to be expressed in proximal regions of the limb bud and both play an important role in specifying cell fates and differentiation patterns along the proximal – distal axis of the limbs [56,57]. The Brachyury-related (Tbx) genes, Tbx4 and Tbx5, are important for establishing the identities of the limbs [58]. Tbx5 expression is restricted to forelimb mesenchyme buds and Tbx4 to the hindlimbs [59]. Misexpression of Tbx4 in chick forelimb buds leads to the transformation of forelimb structures to those of the hindlimb [60,61]. Misexpression of Pitx1, a homeobox-containing transcription factor that regulates Tbx4 expression in the wing buds of chick embryos, induces a leg-like phenotype [62]. Pitx1-null mice have abnormalities in tibia, fibula, and tarsal bones and they resemble the bones of the forelimb [63,64]. All the major downstream mediators of limb bud patterning genes are not known, but homeobox genes (members of the Hoxa and Hoxd clusters), zinc finger transcription factors (Gli genes), and members of the transforming growth factor (TGF-) superfamily play essential roles. Genes of the Hoxd cluster are expressed in overlapping regions in the posterior and distal zones of the limb bud [65,66]. In humans, expansions of a polyalanine stretch in the N-terminal region of Hoxd13 cause synpolydactyly (SPD), a dominantly inherited human malformation of hands and feet. In heterozygous patients, digits are shorter than normal and central syndactyly is seen [67,68]. A phenocopy of SPD arises when Hoxd11, 12, and 13 gene functions are all eliminated in mice, indicating that the SPD protein interferes with the function of the three Hox proteins by a dominant-negative mechanism [69]. Homozygous mice carrying the same mutation as that found in the human SPD syndrome have the same phenotype as human patients [70]. A positive feedback exists between Hoxd11 or 12 and Shh. Hoxd12 misexpression in transgenic mice produces the transformation of anterior digits to digits of posterior morphology and digit duplications [71]. Hoxd11 or 12 genes can directly amplify the posterior Shh polarizing signal to reinforce the positive feedback loop during limb bud outgrowth [71].
V. SONIC HEDGEHOG AND INDIAN HEDGEHOG IN SKELETAL DEVELOPMENT The Hedgehog (Hh) gene was first identified in Drosophila where it is involved in the patterning of larval segments and adult appendages [72]. Sonic hedgehog (Shh)
195
CHAPTER 5 Developmental Biology of Bone
and Indian hedgehog (Ihh) are vertebrate homologues of Hh [73]. They are secreted molecules that act as extracellular signals to regulate cell proliferation and differentiation. First synthesized as large precursors, they are processed by an autoproteolytic activity of the C-terminal domain [74]. The resulting N-terminal product (the signaling part of the molecule) is further modified by the addition of a cholesterol moiety, which plays a role in facilitating the appropriate spatial distribution [75]. Therefore, these autocatalytic reactions not only release the active form but also affect the distribution of the patterning signal in tissues. The current understanding of the mechanism of Hh signaling in Drosophila is as follows [76]: Upon Hh binding to its transmembrane receptor Patched (Ptc) [77,78], another transmembrane protein, Smoothened (Smo) [79,80], is released from the Ptc – Smo complex and triggers downstream signals via Cubitus interruptus (Ci), a five zinc finger-containing transcription factor [81]. In the presence of Hh signaling, the full-length Ci protein (Ci155) is stabilized and translocates into cell nuclei where it mediates the transcriptional activation of Hh target genes [82]. In the absence of Hh binding, Ptc forms a complex with Smo and inhibits Smo action. This results in proteolysis of Ci155 by a cAMP-activated protein kinase (PKA)-dependent mechanism to generate a 75-kDa N-terminal fragment (Ci75), which enters the nucleus and acts as a repressor of Hh targets [83]. In vertebrates, a similar mechanism has been demonstrated, although it is more complex. Instead of one Ci gene, three Ci homologues have been identified, Gli1, Gli2, and Gli3 [84]. Mammalian Gli proteins display extensive homology to each other but only limited homology to the Drosophila Ci protein outside the zinc finger regions. Gli3 is processed proteolytically in a PKA-dependent manner similar to Ci in flies in the absence of Shh signaling (Gli3 – 190 to Gli3 – 83) [85]. Overexpression of Gli3 represses Gli1 expression in cell culture [86,87]. Expression of Gli proteins is associated with activation of the Shh pathway [88]. Despite the finding that Gli1 is dispensable for normal mouse development [89], the loss of Gli2 or Gli3 function affects many processes of Shh-dependent organogenesis [90 – 94]. Gli3 is particularly important in vertebrate limb patterning. Mutations in GLI3 have been identified in Pallister – Hall syndrome (PHS), postaxial polydactyly type A (PAP-A), and Greig cephalopolysyndactyly syndrome (GCPS). PHS and PAP-A are dominant human syndromes associated with postaxial polydactyly [95]. Several genetic lesions identified in PHS [96,97] result in truncations that cause an increased level of GLI3 fragments with strong repressor activities [85]. Unlike PHS, GCPS in humans and extra-toes (Xt) and polydactyly Nagoya (Pdm) (Table 2) in mice are caused by mutations that abolish Gli3 function [98 – 100]; in these disorders the levels of Gli3 repressor activity are therefore decreased. In mice, this derepression leads to ectopic Shh expression in
TABLE 2
Mouse Genetic Disorders and Their Associated Genes
Mouse syndrome
Gene affecteda
Extra toes (xt)
Gli3
Polydactyly Nagoya (Pdm)
Gli3
Short ear (se)
Bmp5
Brachypodism (bp)
Gdf5
Osteogenesis imperfecta (oim)
Col1a2
Disproportionate micromelia (dmm)
Col2a1
Chondrodysplasia (cho)
Col11a1
Cartilage matrix deficiency (cmd)
Agc
a For gene notations we have followed the standard of using capital letters for human genes and a capital letter followed by lowercase letters for mouse genes
the anterior mesenchyme of the developing limb bud and, therefore, extra digits. The role of Ihh in growth plate function is discussed in the next section.
VI. BONE MORPHOGENETIC PROTEINS, HOMOLOGUES, AND THEIR ANTAGONISTS IN SKELETAL DEVELOPMENT Bone morphogenetic proteins (BMPs) are members of the transforming growth factor- (TGF-) superfamily, which regulate a number of embryonic events [101 – 103]. Major subdivisions within the superfamily include the TGF-s, BMPs (excluding BMP1), growth/differentiation factors 1 – 10 (called GDFs or CDMPs, a subclass of BMPs), inhibins, activins, Vg-related genes, nodal-related genes, and glial-derived neurotropic factor. BMP genes are vertebrate homologues of the Drosophila decapentaplegic, a target of the Hedgehog pathway. A large number of processes are known to require BMPs, including the proliferation of mesodermal cells [104]; growth and regression of the AER at the distal tip of limb buds [105,106]; formation of the antero – posterior limb axis and Hox gene expression [107,108]; initiation of chondrogenesis and cartilage differentiation [109 – 113]; myogenesis [110,114]; and interdigital apoptosis during limb development [115 – 117]. BMPs owe their name to the early discovery that demineralized bone fragments implanted subcutaneously or intramuscularly in animals induce bone formation [118]. A search for responsible factors resulted in the identification of the family of bone morphogenetic proteins [119]. To date, 15 BMPs have been identified and are further divided into groups according to their amino acid sequence
196
REGINATO, WANG, AND OLSEN
similarities [120]. Structural studies of BMPs reveal that they contain a mature domain that is cleaved proteolytically, allowing monomeric units to become dimers that are stabilized by disulfide bridges. Expression of different BMP genes in a cell can produce homodimers and heterodimers with variable degrees of glycosylation; these variations can influence the activities and effects of BMPs [121]. BMPs elicit their effects on target cells through binding to specific cell surface type I (BMPR-IA or BMPR-IB) and type II receptors (BMPR-II) [122]. The activated receptors (serine-threonine kinases) in turn activate SMAD proteins (homologues of mothers against decapentaplegic in Drosophila), which relay the signal from the cell cytoplasm to the nucleus [123]. Individual receptor molecules have low affinity for BMPs; however, with the formation of heterotetrameric complexes (two type I receptor molecules and two type II receptor molecules held together by the BMP ligand), high affinity is achieved [124 – 126]. Type II receptors are active kinases that function upstream of the type I receptors but can independently initiate cell signals [127]. On binding to BMP2, 4, or 7, the type II receptor kinase transphosphorylates the type I receptor [127,128].
FIGURE 7
Specific signals appear to be determined primarily by the type I receptor [129]. The type I receptor phosphorylates a serine residue in SMAD1, 5, and possibly 8. After phosphorylation, these SMADs associate with SMAD4 as heterooligomers and translocate to the nucleus where they accumulate rapidly. The SMAD1 signaling pathway appears to be regulated negatively by SMAD6, which inhibits SMAD1 signaling through binding to the type I receptor and competing with SMAD4 for binding to the activated SMAD1. This produces an inactive complex of SMAD1 and SMAD6 [130,131]. The C-terminal domain of SMAD1 is required for DNA binding and subsequent transcriptional activation [124,132]. In a similar manner, SMAD7 acts as an inhibitor of SMAD2 – SMAD4 complexes downstream of TGF- signaling [133]. Regulation of BMP effects depends on the distribution of specific signaling receptors, their functional states, and secreted protein antagonists (Fig. 7, see also color plate). BMP antagonists share the functional property of binding specifically to BMPs, thus preventing their interaction with the receptors. Antagonists include molecules such as noggin [134], DAN [135], Drm [136], chordin [137], and
Diagram illustrating how different BMP antagonists (chordin, noggin, follistatin, and members of the DAN family) can bind BMP dimers and thereby control the amount of BMPs available for binding to the signaling receptor complex (consisting of type I and type II receptors) on the cell surface. The antagonists are thought to be inactivated by proteolytic fragmentation. Following ligand binding, the type II receptor transphosphorylates (orange arrow) and activates the type I receptor; this initiates signaling to the nucleus. Modified from Cho and Blitz [123]. (See also color plate.)
197
CHAPTER 5 Developmental Biology of Bone
follistatin [138]. Noggin binds several BMPs with high affinities, with marked preference for BMP2 and BMP4 over BMP7 [139]. Noggin is expressed in chondrogenic condensations and appears to regulate the shape and size of the cartilaginous anlagen by controlling the effects of BMPs [140]. Gremlin, a member of the DAN family, appears to be important for establishing prechondrogenic condensations. It is highly conserved through evolution and can bind to and block BMP2, 4, and 7 [141]. At early stages of limb development, gremlin is expressed under the control of the AER and ZPA in a pattern that is complementary to these BMPs. The function of gremlin may be to confine chondrogenesis to the central core of the limb. Analysis of different mouse mutations suggests that there is considerable overlap between the functions of various BMPs. Targeted inactivation of the genes for Bmp2, 4, and 7 in mice leads to embryonic or perinatal lethality; these Bmps must therefore have important roles in early developmental processes. Bmp2 mutations result in malformations of the amnion, chorion, and the heart [142]. Bmp4 and BmprIA null mice have early lethal defects in mesoderm formation [143,144]. Alterations in kidney, ears, eyes, ribs, sternum, and skull and polydactyly are seen in Bmp7 knockout mice [145,146]. Homozygous mutant mice carrying a targeted deletion of Gdf11/Bmp11 exhibit anterior directed homeotic transformations throughout the axial skeleton and posterior displacement of the hindlimbs, suggesting that Gdf11 acts globally to specify positional identity along the anterior-posterior axis [147]. In addition to developmental abnormalities seen in knock – out mice, several skeletal disorders have been demonstrated to be caused by mutations in BMP genes. The short ear (se) and the brachypodism (bp) mutations in mice are caused by mutations in Bmp5 [148] and Gdf5 [149], respectively. Brachydactyly type C is the result of haploinsufficiency mutations in CDMP1 (cartilage-derived morphogenetic protein 1; also called GDF5) [150]. Homozygosity for CDMP1 mutations in humans causes acromesomelic dysplasia Hunter – Thompson type [151,152]. This disorder is characterized by short stature and shortening of forearms and lower legs, as well as the long bones of hands and feet. In the more severe Grebe type, affected individuals have a null allele of CDMP1 on one chromosome and an allele on the other chromosome with a dominant-negative mutation. The abnormal protein probably forms nonfunctional heterodimers with other BMPs, preventing their secretion [152]. Interference with BMP signaling causes the absence of proximal interphalangeal joints, fusion of wrist and ankle joints, and conductive deafness. This disorder, called proximal symphalangism (SYM1), has been shown to be caused by mutations in the gene for the BMP-binding molecule noggin [153]. Noggin mutations are also associated with multiple synostosis (SYNS1) syndromes, characterized by fusion in several joints (elbows, hips, intervertebral joints)
in addition to joints in the hands and feet [153]. Not surprisingly, in mice with inactivated noggin alleles, limb cartilage becomes hypoplastic and joints fail to form [140]. Based on studies of mice with inactivated BmprIB or Gdf5 alleles, it has been suggested that BmprIB regulates distal limb chondrogenesis through both Gdf5-dependent and -independent processes, and that, reciprocally, Gdf5 acts through both IB and other type I receptors [154,155].
VII. FIBROBLAST GROWTH FACTORS AND THEIR RECEPTORS IN SKELETAL DEVELOPMENT Fibroblast growth factors are essential for several aspects of bone development, not only in the ossification of cranial sutures and limb bud outgrowth, but also in growth plate function, i.e., longitudinal bone growth. Since their first discovery as a mitogenic activity in pituitary extracts [156,157], FGFs have been shown to support the proliferation of a variety of mesenchymal and epithelial cells, regulate cell migration, differentiation, and chemotaxis [158], and be involved in a variety of nonskeletal and skeletal developmental processes [159,160]. Proteins encoded by different FGF genes (Fgf1 – Fgf19) range in size from 155 to 268 amino acid residues. Each polypeptide contains a conserved region of approximately 120 amino acids that confers a common tertiary structure and the ability to bind heparin or heparan sulfate proteoglycans (HSPGs) [161]. FGFs are released into the extracellular matrix and bind to two classes of cell surface receptors: low-affinity and high-affinity receptors. The low-affinity FGF receptors are heparan sulfate-containing proteoglycans such as syndecan and glypican [162]. These proteoglycans may restrict the ability of FGFs to diffuse far from the cells that release them [163] and may facilitate signal transduction by oligomerizing and presenting the FGF ligands to high-affinity receptors (FGFRs) [164]. Thus, treating cells with heparin-degrading enzymes or with inhibitors of glycosaminoglycan sulfation inhibits the binding of and response to FGFs [165]. FGF receptors constitute a family of four transmembrane receptor tyrosine kinases encoded by four distinct genes (Fgfr1 – Fgfr4) [166]. These proteins consist of an extracellular ligand binding domain, a single transmembrane domain, and an intracellular tyrosine kinase domain (Fig. 8, see also color plate). The extracellular domain contains two or three immunoglobulin-like (Ig-like) domains, depending on alternate RNA splicing. One splicing event involves the N-terminal domain (domain I), leading to a form of the receptor with only two Ig-like domains. The ligand-binding properties of the alternately spliced receptor are similar to those of the full-length receptor, suggesting
198
FIGURE 8
Diagram showing the various domains in an FGFR3 receptor molecule and different mutations associated with four osteochondrodysplasias. The amino acid residues are numbered from the N terminus (left) of the molecule, with glycine at position 380 being localized in the transmembrane-spanning region of the molecule. Standard single letter names are used for amino acid residues; TER indicates termination codon. (See also color plate.)
that domain I is not critical for ligand binding. In contrast, the alternative splicing of exons encompassing domain III results in either IIIb or IIIc isoforms of the FGFR1, 2, and 3 receptors and affects their ligand specificity dramatically [167,168]. Interestingly, the IIIb isoform appears to be expressed in epithelial lineages, whereas the IIIc isoform is expressed primarily in mesenchymal cells [169 – 172]. After ligand binding, the receptor molecules dimerize, followed by autophosphorylation of tyrosine residues in the intracellular domain. Both homodimer and heterodimer interactions can occur between different FGF receptors within a cell [173]. The activated receptor in turn activates downstream signaling targets, including the RAS/MAP kinase and phosphatidylinositol pathways, ultimately influencing mitogenesis and differentiation [174 – 176]. Insights into the role of FGF receptor signaling in skeletal development have been gained from studies of human genetic diseases and mice that carry mutations in the receptors. Mutations in different domains of FGFR1, 2, and 3 cause a number of human craniosynostosis and dwarfism syndromes [177,178]. All the mutations are dominantly inherited, with the vast majority representing missense mutations that result in “gain of function.” The mutations cause excessive activation of the receptor, sometimes in a ligandindependent fashion, resulting in altered cell proliferation and/or differentiation [179].
REGINATO, WANG, AND OLSEN
Mutations in FGFR1 cause Pfeiffer syndrome, whereas mutations in FGFR2 can result in Apert syndrome, Crouzon’s syndrome, Jackson – Weiss syndrome, or Beare – Stevenson syndrome, depending on the location and nature of the mutation and genetic background effects. Mutations in FGFR3 are also associated with several distinct clinical syndromes [178]. Muenke syndrome (MS), or Muenke nonsyndromic coronal craniosynostosis, is characterized by incomplete penetrance and a wide phenotypic spectrum. MS patients may present with uni- or bilateral coronal craniosynostosis and midface hypoplasia, a downslanted palpebral fissure, or ptosis [180]. Crouzon syndrome with acanthosis nigricans is a rare condition seen in a subgroup of Crouzon patients with cutaneous manifestations of hyperpigmentation, hyperkeratosis, melanocytic nevi, and verrous hyperplasia. The craniofacial abnormalities are similar to those seen in patients with Crouzon syndrome [181,182]. Finally, different mutations in FGFR3 cause the dwarfism conditions hypochondroplasia (HCH), achondroplasia ACH), and thanatophoric dysplasia (TD) (Fig. 8) [183, 184]. Achondroplasia is among the most common and well known of the skeletal dysplasias and is characterized by short stature with rhizomelic shortening of the limbs, reduced elbow extension, genu varum, trident hand, exaggerated lumbar lordosis, frontal bossing, and midface deficiencies. Hypochondroplasia has similar limb and spinal features, but they are less severe. The face is normal, but the head circumference is larger than normal. Thanatophoric dysplasia, a neonatal lethal dwarfism, is clinically divided into two types: type 1 and type 2 thanatophoric dysplasia. In type 1, the long tubular bones are curved and the vertebral bodies are flat. In type 2, the long tubular bones are straight and the vertebral bodies are not as flat as in type 1, and craniosynostosis (cloverleaf skull) is present in most cases. Type 1 is the result of one of several mutations, both in extracellular and intracellular domains of FGFR3. Type 2 is caused by a mutation in the second kinase domain of FGFR3. The underlying defect in HCH, ACH, and TD is the disruption of normal, regulated proliferation and differentiation of chondrocytes in the epiphyseal growth plates of long bones, caused by gain-of-function mutations in FGFR3 [185 – 188]. A hypothesis for the phenotypic differences between the two types of thanatophoric dysplasia and achondroplasia is based on the dual developmental origin of the skull, in that it is formed by both endochondral and membranous ossification. Mutations that affect endochondral bone would cause deficiencies primarily of the cranial base and the nasal capsule, with a secondary effect on membranous bone, resulting in a shortened anterior cranial fossa and midface hypoplasia. This would lead to an abnormal fusion of sutures between membranous bones of the skull (frontal and parietal) during intrauterine development
CHAPTER 5 Developmental Biology of Bone
and sagittal synostosis. In this view, the severity of the cranial phenotype would depend on the degree of hypoplasia of the nasal capsule and the cranial base, and a spectrum of craniofacial deformities ranging from achondroplasia to thanatophoric dysplasia would be the consequence [184]. Although attractive, the hypothesis does not explain why the limbs and vertebrae are affected more severely in type 1 than type 2 thanatophoric dysplasia and why the skull is involved more severely in type 2 than in type 1. Part of the explanation may be that the common mutation in type 1, R248C, is known to completely activate the receptor in a ligand-independent manner, whereas the type 2 mutation, K650E, activates the receptor to a lesser degree and receptor activation is still partially ligand dependent [see 184]. In addition, the type 2 mutant FGFR3 receptor may dimerize with FGFR2 in the skull, and give rise to a more severe skull phenotype than the mutant homodimer of FGFR3 in type 1. Alternatively, craniosynostosis may be the result of heterodimerization of the mutant FGFR3 receptor with FGFR2 at the edges of membranous bones in the calvaria rather than cartilage hypoplasia [184]. Confirmation that constitutive receptor activation is involved in FGFR-based skeletal dysplasias has come from studies of knockout mice. Fgfr3 knockout mice display a phenotype with overgrowth of the vertebral column, long bones, and deafness due to an expansion of the zone of proliferating and hypertrophic chondrocytes in growth plate cartilages. In many ways, this phenotype is the opposite of the phenotypes of human achondroplasia caused by activating mutations in FGFR3 [189,190] and suggests a role for FGFR3 in restraining chondrocyte proliferation and maturation during endochondral bone formation. Transgenic mice that overexpress the Fgfr3 ligand Fgf2 exhibit skeletal malformations that include long bone shortening, and microcephaly, again suggesting that enhanced signaling through FGFR3 underlies FGFR3-based dysplasias [191]. Targeted disruption of the mouse Fgf2 gene leads to a phenotype with multiple abnormalities, including decreased bone formation [192].
VIII. CHONDROCYTE DIFFERENTIATION AND ENDOCHONDRAL BONE FORMATION The process of mesenchymal cell condensation, proliferation, and differentiation of chondrocytes is called chondrogenesis. Mesenchymal cells present in the anlagen initially express both type I collagen and a splice variant of type II collagen, type IIa, which is not chondrocyte specific [193]. Chondrocytes later switch on the specific isoform of type IIb collagen. Sox9, a member of the HMG box family of
199 transcription factor genes, is expressed in mesenchymal condensations before and during chondrogenesis [194,195]. The expression of Col2A1 is dependent on Sox9. Sox9 binds to the Col2A1 enhancer located in the first intron of the gene and activates transcription of the gene in vivo and in vitro [196 – 199]. Embryonic stem cells carrying inactivated Sox9 alleles fail to undergo chondrogenesis within mesenchymal condensations when mixed with wild-type cells [200], and chondrogenic markers such as Col2a1 and aggrecan are not expressed. Heterozygous loss-of-function mutations in SOX9 lead to campomelic dysplasia, a skeletal dysplasia characterized by abnormalities in all skeletal elements that are formed by endochondral ossification [201,202]. Sox9 also plays a role in the differentiation of Sertoli cells in male gonads; this provides an explanation for the high frequency of autosomal sex reversal in campomelic dysplasia. The products of other Sox genes, L-Sox5 and Sox6, form a complex with Sox9 during the activation of Col2a1 in vitro [203,204]. Other known molecules that affect chondrogenesis are Gdf5 and BmprIB. Gdf5 is expressed in the end regions of cartilage anlagen [149,205]. The disruption of the Gdf5 gene in mice leads to abnomalities in mesenchymal condensations, and mutations in Gdf5 result in brachypodism (see earlier discussion) [149]. BmprIB expression in chick limb buds precedes the future cartilage anlagen and its activity is involved in the initial steps of chondrogenesis (see earlier discussion) [206]. The replacement of cartilage by bone and bone marrow during endochondral bone formation starts by the hypertrophy of chondrocytes in the center of cartilage anlagen. The hypertrophic extracellular matrix mineralizes, blood vessels invade the cartilage, and a primary ossification center is formed (Fig. 9, see also color plate). This results in the deposition of trabecular bone within the cartilage, as the extracellular matrix of hypertrophic cartilage is degraded and hypertrophic chondrocytes undergo apoptosis. At the same time, bone is deposited in a sleeve around the primary ossification center. As primary ossification centers expand and secondary centers are established in the end regions of cartilage anlagen, cartilaginous growth plates remain as the centers of longitudinal growth (Fig. 2). Activities of the growth plate are regulated by several known cytokines. The parathyroid hormone-related peptide (PTHrP) is expressed in the periarticular region, whereas Ihh is expressed in prehypertrophic chondrocytes [207,208]. Chondrocyte hypertrophy is accelerated in PTHrP-deficient mice [209], indicating that PTHrP promotes proliferation and inhibits hypertrophy. Ihh has also been shown to stimulate chondrocyte proliferation, and Ihh signaling is required for PTHrP function [210]. The inactivation of Ihh genes in mice confirms that PTHrP signaling requires Ihh; however, Ihh plays multiple roles in PTHrPdependent and -independent pathways [211,212]. In
200
REGINATO, WANG, AND OLSEN
protein vascular endothelial growth factor (VEGF). This is crucial for blood vessel invasion into and resorption of hypertrophic cartilage. Anti-VEGF treatment using a soluble receptor in mice prevents capillary invasion into hypertrophic cartilage and impairs trabecular bone formation [219].
IX. OSTEOBLAST DIFFERENTIATION AND FUNCTION FIGURE 9
Immunohistochemical staining of blood vessels in the leg of a mouse embryo at day 17.5 of development. Staining with anti-CD31 antibodies shows vessels in the diaphyseal region of the developing tibia, where trabecular bone and bone marrow replace cartilage. Note that no vessels penetrate into the cartilage regions of the proximal and distal ends of the tibia. ep, epidermis. Courtesy of Dr. N. Fukai. (See also color plate.)
Ihh-null mice, PTHrP is undetectable and chondrocyte differentiation is abnormal. Furthermore, Cbfa1 is not expressed in the perichondrium and no endochondral bone formation occurs. This indicates that Ihh signaling is required for the normal differentiation of chondrocytes and osteoblasts during endochondral bone formation. Signaling by PTHrP is mediated by the PTH/PTHrP receptor [208]. The receptor is a G-protein-coupled receptor with seven membrane-spanning domains and is expressed in proliferating and prehypertrophic chondrocytes, as well as in osteoblasts. Targeted disruption of the receptor gene in mice leads to acceleration of the transition from proliferative to hypertrophic chondrocytes, resulting in premature ossification and short-limbed dwarfism [208]. An activating mutation in the receptor causes a decrease in hypertrophy and results in an autosomal dominant form of dwarfism, Jansen metaphyseal chondrodysplasia [213]. A loss-offunction mutation also causes dwarfism, Blomstrand chondrodysplasia [214]. Fgfr3 is expressed in nonproliferating chondrocytes, represses Ihh signaling, and inhibits Bmp4 expression in growth plates [215]. As described earlier, activating mutations in FGFR3 cause hypochondroplasia, achondroplasia, and thanatophoric dysplasia. Consistent with this, inactivation of Fgfr3 genes results in bone overgrowth in mice. Hypertrophic chondrocytes synthesize extracellular matrix proteins that are significantly different from those that are synthesized by small chondrocytes in the growth plates. Collagen X, encoded by Col10A1, is unique to the hypertrophic matrix and serves as an excellent marker for hypertrophic chondrocytes [216]. A mutation in Col10A1 causes Schmid metaphyseal chondrodysplasia in humans [217]; transgenic mice expressing a dominant-negative form of collagen X develop spondylometaphyseal dysplasia [218]. Hypertrophic chondrocytes also synthesize the angiogenic
Osteoblasts, originating from mesenchymal cells, are responsible for bone matrix deposition in both membranous and endochondral bone formation. A key regulator of osteoblast differentiation and function is Cbfa1 [4 – 6]. Cbfa1 is a member of the Runt domain-containing transcription factor family. Members of this family share a DNA-binding domain that is homologous to Drosophila Runt protein. Cbfa1 was initially identified as an enhancer binding protein for polyoma virus and murine leukemia virus and was given various names, such as PEA2, PEBP2a, CBF (core binding factor), and AML3 (acute myeloid leukemia 3 gene) [220]. It was described independently as Osf2 for its ability to bind the cis-acting elements of osteocalcin genes [221]. Cbfa1 is highly expressed in preosteoblasts and mature osteoblasts, and many bone matrix proteins, including bone sialoprotein and collagen type I, require Cbfa1 for their expression [4]. No bone is formed in Cbfa1-deficient mice, although the cartilage skeleton is patterned normally. In null mice, osteoblasts do not differentiate [5,6]. The activity of Cbfa1 is required not only for embryonic bone development, but also for postnatal bone growth. Bone formation ceases when Cbfa1 activity is blocked by expression of a dominant-negative mutant containing only the DNA-binding Runt domain [222]. Haploinsufficiency of Cbfa1 causes Cleidocranial dysplasia (CCD) in both mice and humans. The disorder is characterized by delayed closure of sutures, hypoplastic clavicles, hypoplastic pelvis, and short stature [223,224]. Cbfa1 is also expressed in chondrocytes in that it is first turned on in prechondrocytes, turned off in proliferating chondrocytes, and then turned on again in hypertrophic chondrocytes [225,226]. Blood vessel invasion into hypertrophic cartilage does not occur in Cbfa1-null mice.
X. CELL – MATRIX INTERACTIONS IN SKELETAL DEVELOPMENT Skeletal development depends critically on the synthesis of extracellular matrix components of cartilage and bone. This is illustrated by the large number of skeletal dysplasias that are secondary to mutations in matrix molecules. In
CHAPTER 5 Developmental Biology of Bone
bone, type I collagen, a heterotrimer of two 1(I) and one 2(I) chains, represents 90% of the total protein mass. Mutations in COL1A1, the gene coding for 1(I), as well as mutations in COL1A2, the gene coding for the 2(I) chain, cause osteogenesis imperfecta (OI) [227]. Osteogenesis imperfecta comprises a heterogeneous group of inherited disorders characterized by brittle bone disease leading to increased bone fracture, hearing loss, blue sclerae, and dentinogenesis imperfecta. Patients with OI type I, a mild phenotype, fracture their bones easily, but have normal bone healing and little or no deformity. These patients have a nonfunctioning or “null” collagen gene allele, resulting in a reduced collagen content in bone [228,229]. Patients with more severe phenotypes (lethal, type II; severe deformity, type III; and moderate deformity, type IV) generally have point mutations in the triple-helical domain of either the 1(I) or the 2(I) chains, with glycine replaced by another, bulkier amino acid residue [230,231]. Large multiexon deletions or duplications are not common, but have been reported to give rise to phenotypes that range from mild to lethal [see 232]. The exact mechanism by which these different mutations cause the phenotypes is largely unknown, but general principles are emerging. Mutations that affect the C-terminal propeptide prevent the incorporation of the mutant chain into trimeric molecules and result in the mildest phenotypes. Mutations that affect the Gly-X-Y repeat domain, and allow the mutant chain to participate in triple helix formation, act in a dominant-negative manner and result in a more severe phenotype. The phenotype is largely determined by the degree of secretion of mutant and normal collagen molecules into the extracellular space and their subsequent incorporation into collagen fibrils. If a mutant collagen molecule is incorporated into the matrix, it will affect the overall stability of the fibrils. In contrast, if the mutant collagen molecule is not incorporated into the matrix, the mature matrix will be affected less severely [see 232]. Several animal models for osteogenesis imperfecta exist. One such model is the Mov13 mouse, in which a retroviral insertion results in the transcriptional block of Col1a1 [233,234]. Transgenic mice have been generated with mild to lethal phenotypes [235 – 237]. Another model is a nonlethal recessive mutation (oim) in mice with features of osteopenia, fractures, and progressive skeletal deformities. In oim mice, a point mutation leads to alteration of the last 48 amino acid residues of the pro2(I) collagen chain; this prevents the association with pro1(I) collagen chains during assembly of the heterotrimeric molecule [238]. As a result, 1(I) homotrimers are formed in both skin and bone. This closely resembles the situation in patients with type III osteogenesis imperfecta [239]. Transgenic mice expressing a partially deleted Col1a1 gene show phenotypic variability and incomplete penetrance of spontaneous fractures even
201 on an inbred background [237]. This suggests that the phenotypic variability is an inherent stochastic feature of the expression of mutated collagens. Type II collagen, a homotrimer encoded by COL2A1, is the major structural component of cartilage. In addition, it is expressed in other structures, such as the vitreous of the eye, the nucleus pulposus of intervertebral discs, and the tectorial membrane of the inner ear. Mutations in type II collagen cause a spectrum of diseases known as “type II collagenopathies.” The severity ranges from developmental lethality (achondrogenesis type II, hypochondrogenesis), to moderately severe dwarfism (spondyloepiphyseal dysplasia congenita, Kniest dysplasia), and to normal stature with premature osteoarthritis (Stickler and Marshall syndromes) [240]. In the case of lethal mutations that lead to an absence of type II collagen in embryonic cartilage, some type of cartilage develops with collagen I substituting for collagen II, chondrocytes differentiate, and bones are formed [241]. Mutations that cause a moderately severe phenotype (spondylopepiphyseal dysplasia, Kniest dysplasia) generally lead to a reduced secretion and content of type II collagen in cartilage [242 – 244]. Patients with these mutations have disproportionate dwarfism with rhizomelic shortening of the limbs, while the hands and feet are minimally affected. The mildest phenotypes (Stickler syndrome with premature osteoarthritis) are caused by premature stop codons resulting in a null allele [244,245]. Several mouse models of the human type II collagenopathies have been generated [246,247]. A naturally occurring mouse mutant, disproportionate micromelia (dmm), has been found to be caused by a three nucleotide deletion in Col2a1, leading to the substitution of Lys-Thr with Asn in the C-propeptide of the type II procollagen molecule [248]. In addition to type II collagen, a number of quantitatively minor collagens exist in cartilage. They include types IX, X, XI, and XII collagen. Type IX collagen is a member of the FACIT group of extracellular matrix proteins, which also includes types XII, XIV, XVI, and XIX. Type IX collagen is a heterotrimeric nonfibrillar collagen composed of three different chains, 1(IX), 2(IX), and 3(IX), that are coexpressed with type II collagen in cartilage and other cartilage-like tissues. Type IX molecules are localized on the surface of the fibrils where they get cross-linked to residues within the N- and C-telopeptides of type II collagen; their function is probably to stabilize the fibril network [see 249]. Mutations in collagen IX cause autosomal-dominant multiple epiphyseal dysplasia (MED). This disorder includes the mild Ribbing type, the more severe Fairbank type, and some unclassified MED types. Splice-site mutations in COL9A2, causing skipping of exon 3 in 2(IX) transcripts, result in a mild phenotype with short stature and early-onset osteoarthritis [250,251]. Further analysis of MED has shown similar skipping of exon 3 in the a3(IX) gene transcript, caused by mutations in the intron 3
202 splice – donor site of COL9A3 [252 – 254]. Transgenic mice overexpressing a dominant-negative truncated form of the 1(IX) chain show mild chondrodysplasia and progressive osteoarthritis [255]. Mice homozygous for a null mutation in Col9a1 exhibit normal skeletal development, but develop progressive osteoarthritis-like changes in articular cartilage after birth [256]. Type X collagen, a member of the short chain collagen family, is a homotrimeric molecule that is expressed by hypertrophic chondrocytes during endochondral bone formation (Fig. 3). Transgenic mice, expressing a collagen X gene with a large in-frame deletion in the central triple helical domain, as well as mice with Col10a1-null alleles, develop variable spondylometaepiphyseal chondrodysplasia after birth [218,257]. Histological analysis shows a reduction in the height of the hypertrophic zone, severe reduction in the size and number of bone trabeculae below the growth plate cartilage, and lymphopenia in the bone marrow, the thymus, and the circulation. In humans, mutations in COL10A1 have been shown to cause the autosomaldominant disorder Schmid metaphyseal chondrodysplasia, characterized by bowing of the legs, growth retardation of the extremities, and coxa vara [217]. All the mutations identified are clustered in the C-terminal nontriple helical NC1 domain of the molecule. They are likely to prevent the formation of trimeric molecules, and the mutant chains are consequently not secreted by chondrocytes. This suggests that the phenotype seen in Schmid metaphyseal chondrodysplasia is caused by haploinsufficiency, although dominant-negative mechanisms cannot be ruled out for at least some of the mutations [258 – 260]. Type XI copolymerizes with type II collagen and appears to regulate the diameter of cartilage fibrils. It is a heterotrimer composed of products of three different genes: COL11A1, COL11A2, and COL2A1. Mutations in COL11A1 have been shown to result in Marshall or Stickler syndrome, characterized by high myopia, vitreoretinal degeneration, cleft palate, midfacial hypoplasia, premature osteoarthritis, and hearing defects [261,262]. Extensive genotype – phenotype comparisons of patients with Stickler, Stickler-like, or Marshall syndromes have suggested that null-allele mutations in COL2A1 cause the typical Stickler phenotype (vitreoretinal degeneration common and hearing loss less common), whereas splicing mutations in the COL11A1 gene are responsible for the Marshall syndrome (hearing loss common and vitreoretinal degeneration less common). Patients with glycine substitutions or small in-frame deletions in the COL11A1 gene (dominantnegative mutations) may have a mixed phenotype characteristic of both syndromes [263]. In contrast to mutations in COL2A1, mutations in COL11A2 are associated with a Stickler-like syndrome without eye involvement [264]. The explanation for this is the presence of a unique form of type XI collagen in the vitreous, which contains an 2(V)
REGINATO, WANG, AND OLSEN
chain instead of the 2(XI) chain of type XI collagen in cartilage [265]. Defects of mice with a recessively inherited chondrodysplasia (cho) provide further insights into the role of collagen type XI in skeletal development. In cho mice, a single nucleotide deletion creates a premature stop codon in the N-terminal region of 1(XI) collagen [266]. Heterozygous animals are relatively unaffected; however, with age they develop osteoarthritis. Homozygous animals die at birth with short limbs, a short snout, and a cleft palate. The vertebral column is short and the thoracic cage is small. Growth plate cartilage is disorganized with soluble proteoglycan aggregates and thicker than normal collagen fibrils. The presence of thick fibrils provides further evidence of the role of type XI collagen in regulating fibril diameters. The proteolytic processing of type XI collagen differs from that of type II collagen in that the N-propeptide domains are not cleaved after secretion. During embryogenesis, these peptide domains control fibril diameter by localizing on the fibril surface and limiting the addition of type II collagen molecules to the surface by steric hindrance [267]. cho mice have thick fibrils leading to the formation of a large-pore network of fewer and thicker fibrils instead of a meshwork of thin fibrils found in wild-type cartilage. Changes in the fibril meshwork cause the proteoglycan aggregates to be loosely entrapped within the mutant matrix and they are more soluble than in wild-type cartilage. Cartilage and bone also contain numerous noncollagenous components, of which the large cartilage proteoglycan or aggrecan is the most studied. Aggrecan contains a core protein that is modified by substitution with chondroitin and keratan sulfate chains and oligosacharides. Aggrecan binds to hyaluronic acid and forms large complexes, which are stabilized by link proteins. The proteoglycan aggregates bind water and give cartilage its ability to withstand compressive forces [268]. Within the growth plate, aggrecan and link proteins are coexpressed with type II collagen. Although aggrecan would appear to be a prime candidate for chondrodysplasias, to date no mutations in the aggrecan gene have been identified in humans. However, mutations have been identified in mice and chicken. Cartilage matrixdeficient mice (cmd) have an autosomal recessive mutation causing short limbs, tail, and snout and a cleft palate. A 7 bp deletion in the aggrecan gene has been demonstrated to result in a truncated molecule and a major reduction in the amount of aggrecan in cartilage matrix [269,270]. In chicken, the nanomelia phenotype is similar to that of the murine cmd and results from a premature stop codon in the avian aggrecan gene [271]. Link protein (LP) stabilizes the aggregates of aggrecan and hyaluronan and is important for the normal organization of hypertrophic chondrocytes in growth plate cartilage [272]. Targeted mutations in the LP gene give rise to defects in cartilage in homozygous mice and delayed bone formation with short limbs and abnormal
CHAPTER 5 Developmental Biology of Bone
craniofacial development. The animals have small epiphyses, flared metaphyses of the long bones, and flattened vertebrae, characteristic of spondyloepiphyseal dysplasia. There is a reduction in aggrecan deposition in the hypertrophic zone in growth plates and a decrease in the number of prehypertrophic and hypertrophic chondrocytes [272]. Biglycan (Bgn) is an extracellular matrix proteoglycan that is enriched in bone and other skeletal connective tissues. In vitro studies indicate that Bgn may function in connective tissue metabolism by binding to collagen fibrils and TGF-. Bgn-deficient mice show a reduced rate of growth and decreased bone mass and display an osteoporotic-like phenotype [273]. Cartilage oligomeric matrix protein (COMP) is another noncollagenous component of cartilage, composed of a pentamer of five identical subunits. COMP belongs to the thrombospondin family of matrix molecules and is localized in the pericellular, territorial matrix in cartilage. The protein contains several repeat domains, including eight calcium-binding calmodulin-like repeats. Mutations in COMP have been described in various types of multiple epiphyseal dysplasia [274 – 277]. MED is thus a heterogeneous disorder, caused by mutations in either collagen IX or COMP. Perlecan is a large heparan sulfate proteoglycan with a wide tissue distribution and multiple potential functions [278]. The perlecan core protein consists of several distinct protein modules organized in five domains, resembling pearls on a string. Domain I contains a glycosaminoglycan side chain that binds basic fibroblast growth factor FGF2 and has been shown to promote mitogenic and angiogenic activities. Other domains are capable of binding several other small and large molecules, including FGF7, fibronectin, heparin, laminin, PDGF-BB, and cell surface integrins [see 279]. Although perlecan plays a major role as a component of basement membranes, it also has an important function in cartilage. French et al. [280] demonstrated that 10T1/2 multipotential mouse embryonic fibroblasts aggregate into a dense cartilagenous nodule when cultured on perlecan, suggesting that perlecan promotes chondrogenic differentiation. Perlecan homozygous knockout mice develop severe osteochondral defects characterized by dwarfism, cleft palate, short limbs, and a short and abnormally curved vertebral column [281]. The phenotype resembles that of Col2a1-deficient [282] and dmm mice [248]. The perlecan-null bones show minor changes in epiphyseal cartilage but severe abnormalities in the growth plates. The proliferating zone is disorganized and the hypertrophic zone shows signs of increased extracellular matrix synthesis and is frequently separated from the proliferating zone [281]. It is possible that perlecan protects the extracellular matrix of cartilage by inactivating matrix proteases or masking/protecting proteins against proteolytic degradation. Also, through its ability to bind and sequester
203 Fgfs, perlecan may modulate Fgfr3 signaling pathways and thus influence both chondrocyte proliferation and differentiation. Finally, given recent evidence for the importance of heparan sulfate proteoglycans in transducing Hedgehog signals, perlecan may even be important for Ihh signaling [283]. Matrilins or cartilage matrix (Crtm) proteins are members of a novel family of extracellular matrix proteins consisting of von Willebrand factor A-like (vWFA-like) domains, epidermal growth factor (EGF)-like domains, and a coiled -helical motif. Four members of the family have been identified. Matrilin1 and matrilin3 are expressed mainly in hyaline cartilage, whereas matrilin2 and matrilin4 are expressed in a wide variety of extracellular matrices [284,285]. The roles of matrilins in cartilage and skeletal development are largely unknown. The cross-linking of matrilin1 to aggrecan core protein [286], as well as its association with type II collagen-containing fibrils [287], suggests a possible role as a structural connector. Mice carrying a null mutation in the Crtm gene coding for matrilin1 demonstrate a normal phenotype with no detectable abnormalities of matrix organization [288]. These findings suggest that matrilin1 is structurally not a critical, limiting component in cartilage structure and that other matrilins may have functionally redundant roles.
XI. SUMMARY AND PERSPECTIVES Recent advances in developmental biology, skeletal cell biology, and molecular genetics have increased our understanding of bone development significantly. We are now beginning to understand the various genetic pathways that control the patterning of mesenchymal condensations. Critical genes and regulatory cytokines involved in both chondrocyte and osteoblast differentiation have been identified. Genes and mechanisms that regulate joint formation and bone growth are becoming known. The roles of extracellular matrix proteins, not only as structural components, but also as modulators of signal transduction, are coming into sharper focus. However, much needs to be accomplished, including a better understanding of endochondral ossification and the function of cartilage in bone remodeling and the factors that control cortical and trabecular bone thickness. What mechanisms trigger the formation of primary versus secondary ossification centers in the cartilage anlagen of long bones? What distinguishes articular cartilage from epiphyseal growth plate cartilage? How is cartilage maintained throughout development and postnatal life? What determines the specific sizes and shapes of bones? These are challenging questions, but given the current interest in skeletal biology, the existence of a large number of inherited bone diseases in humans, new methodologies for genetic manipulations in various animal models, and the
204
REGINATO, WANG, AND OLSEN
rapid advances of the human and mouse genome projects, it seems likely that several of these questions will be answered relatively soon. It is hoped that the topics highlighted in this chapter will stimulate research aimed at providing answers that will benefit patients with developmental and metabolic bone diseases, including osteoporosis.
References 1. D. M. Noden, The role of the neural crest in patterning of avian cranial skeletal, connective, and muscle tissues. Dev. Biol. 96, 144 – 165 (1983). 2. D. M. Noden, Cell movements and control of patterned tissue assembly during craniofacial development. J. Craniofac. Genet. Dev. Biol. 11, 192 – 213 (1991). 3. M. Bronner-Fraser, Neural crest cell formation and migration in the developing embryo. FASEB J. 8, 699 – 706 (1994). 4. P. Ducy, R. Zhang, V. Geoffroy, A. L. Ridall, and G. Karsenty, Osf2/Cbfa1: A transcriptional activator of osteoblast differentiation. Cell 89, 747 – 754 (1997). 5. T. Komori, H. Yagi, S. Nomura, A. Yamaguchi, K. Sasaki, K. Deguchi, Y. Shimizu, R. T. Bronson, Y. H. Gao, M. Inada, M. Sato, R. Okamoto, Y. Kitamura, S. Yoshiki, and T. Kishimoto, Targeted disruption of Cbfa1 results in a complete lack of bone formation owing to maturational arrest of osteoblasts. Cell 89, 755 – 764 (1997). 6. F. Otto, A. P. Thornell, T. Crompton, A. Denzel, K. C. Gilmour, I. R. Rosewell, G. W. H. Stamp, R. S. P. Beddington, S. Mundlos, B. R. Olsen, P. B. Selby, and M. J. Owen, CBFA1, a candidate gene for the cleidocranial dysplasia syndrome, is essential for osteoblast formation and bone development. Cell 89, 765 – 771 (1997). 7. J. A. Rivera-Perez, M. Wakamiya, and R. R. Behringer, Goosecoid acts cell autonomously in mesenchyme-derived tissues during craniofacial development. Development 126, 3811 – 2821 (1999). 8. D. Acampora, S. Mazan, Y. Lallemand, V. Avantaggiato, M. Maury, A. Simeone, and P. Brulet, Forebrain and midbrain regions are deleted in Otx2-/- mutants due to a defective anterior neuroectoderm specification during gastrulation. Development 121, 3279 – 3290 (1995). 9. I. Matsuo, S. Kuratani, C. Kimura, N. Takeda, and S. Aizawa, Mouse Otx2 functions in the formation and patterning of rostral head. Genes Dev. 9, 2646 – 2658 (1995). 10. T. Kume, K. Y. Deng, V. Winfrey, D. B. Gould, M. A. Walter, and B. L. Hogan, The forkhead/winged helix gene Mf1 is disrupted in the pleiotropic mouse mutation congenital hydrocephalus. Cell 93, 985 – 996 (1998). 11. M. Qiu, A. Bulfone, I. Ghattas, J. J. Meneses, L. Christensen, P. T. Sharpe, R. Presley, R. A. Pedersen, and J. L. Rubenstein, Role of the Dlx homeobox genes in proximodistal patterning of the branchial arches: Mutations of Dlx-1, Dlx-2, and Dlx-1 and -2 alter morphogenesis of proximal skeletal and soft tissue structures derived from the first and second arches. Dev. Biol. 185, 165 – 184 (1997). 12. D. Acampora, G. R. Merlo, L. Paleari, B. Zerega, M. P. Postiglione, S. Mantero, E. Bober, O. Barbieri, A. Simeone, and G. Levi, Craniofacial, vestibular and bone defects in mice lacking the Distal-lessrelated gene Dlx5. Development 126, 3795 – 3809 (1999). 13. E. Dahl, H. Koseki, and R. Balling, Pax genes and organogenesis. Bioessays 19, 755 – 765 (1997). 14. A. P. Read and V. E. Newton, Waardenburg syndrome. J. Med. Genet. 34, 656 – 665 (1997). 15. M. Tassabehji, A. P. Read, V. E. Newton, R. Harris, R. Balling, P. Gruss, and T. Strachan, Waardenburg’s syndrome patients have
16.
17.
18.
19.
20. 21.
22.
23.
24.
25.
26.
27. 28.
29.
30.
31.
32.
33.
mutations in the human homologue of the Pax-3 paired box gene. Nature 355, 635 – 636 (1992). V. El, Ghouzzi, M. Le Merrer, F. Perrin-Schmitt, E. Lajeunie, P. Benit, D. Renier, P. Bourgeois, A. L. Bolcato-Bellemin, A. Munnich, and J. Bonaventure, Mutations of the TWIST gene in the Saethre-Chotzen syndrome. Nature Genet. 15, 42 – 46 (1997). T. D. Howard, W. A. Paznekas, E. D. Green, L. C. Chiang, N. Ma, R. I. Ortiz de, Luna, C. Garcia Delgado, M. Gonzalez-Ramos, A. D. Kline, and E. W. Jabs, Mutations in TWIST, a basic helix-loop-helix transcription factor, in Saethre-Chotzen syndrome. Nature Genet. 15, 36 – 41 (1997). E. W. Jabs, U. Muller, X. Li, L. Ma, W. Luo, I. S. Haworth, I. Klisak, R. Sparkes, M. L. Warman, J. B. Mulliken, et al., A mutation in the homeodomain of the human MSX2 gene in a family affected with autosomal dominant craniosynostosis. Cell 75, 443 – 450 (1993). U. Hehr and M. Muenke, Craniosynostosis syndromes: From genes to premature fusion of skull bones. Mol. Genet. Metab. 68, 139 – 151 (1999). P. P. Tam and P. A. Trainor, Specification and segmentation of the paraxial mesoderm. Anat. Embryol. (Berlin) 189, 275 – 305. B. Christ, C. Schmidt, R. Huang, J. Wilting, and B. Brand-Saberi, Segmentation of the vertebrate body. Anat. Embryol. 197, 1 – 8 (1998). K. Horikawa, G. Radice, M. Takeichi, and O. Chisaka, Adhesive subdivisions intrinsic to the epithelial somites. Dev. Biol. 215, 182 – 189 (1999). P. J. Swiatek, C. E. Lindsell, F. F. del Amo, G. Weinmaster, and T. Gridley Notch1 is essential for postimplantation development in mice. Genes Dev. 8, 707 – 719 (1994). R. A. Conlon, A. G. Reaume, and J. Rossant, Notch1 is required for the coordinate segmentation of somites. Development 121, 1533 – 1545 (1995). B. Bettenhausen, M. Hrabe de Angelis, D. Simon, J. L. Guenet, and A. Gossler, Transient and restricted expression during mouse embryogenesis of Dll1, a murine gene closely related to Drosophila Delta. Development 121, 2407 – 2418 (1995). M. Hrabe de Angelis, J. McIntyre, II, and A. Gossler, Maintenance of somite borders in mice requires the Delta homologue DII1. Nature 386, 717 – 721 (1997). N. Zhang and T. Gridley, Defects in somite formation in lunatic fringe-deficient mice. Nature 394, 374 – 377 (1998). Y. A. Evrard, Y. Lun, A. Aulehla, L. Gan, and R. L. Johnson, Lunatic fringe is an essential mediator of somite segmentation and patterning. Nature 394, 377 – 381 (1998). Y. Saga, N. Hata, H. Koseki, and M. M. Taketo, Mesp2: A novel mouse gene expressed in the presegmented mesoderm and essential for segmentation initiation. Genes Dev. 11, 1827 – 1839 (1997). D. B. Donoviel, A. K. Hadjantonakis, M. Ikeda, H. Zheng, P. S. Hyslop, and A. Bernstein, Mice lacking both presenilin genes exhibit early embryonic patterning defects. Genes Dev. 13, 2801 – 2810 (1999). P. C. Wong, H. Zheng, H. Chen, M. W. Becher, D. J. Sirinathsinghji, M. E. Trumbauer, H. Y. Chen, D. L. Price, L. H. Van der Ploeg, and S. S. Sisodia, Presenilin 1 is required for Notch1 and DII1 expression in the paraxial mesoderm. Nature 387, 288 – 292 (1997). K. Kusumi, E. S. Sun, A. W. Kerrebrock, R. T. Bronson, D. C. Chi, M. S. Bulotsky, J. B. Spencer, B. W. Birren, W. N. Frankel, and E. S. Lander, The mouse pudgy mutation disrupts Delta homologue Dll3 and initiation of early somite boundaries. Nature Genet. 19, 274 – 278 (1998). M. P. Bulman, K. Kusumi, T. M. Frayling, C. McKeown, C. Garrett, E. S. Lander, R. Krumlauf, A. T. Hattersley, S. Ellard, and P. D. Turnpenny, Mutations in the human Delta homologue, DLL3, cause axial skeletal defects in spondylocostal dysostosis. Nature Genet. 24, 438 – 441 (2000).
205
CHAPTER 5 Developmental Biology of Bone 34. S. Mundlos and B. R. Olsen, Defects in skeletal morphogenesis. In “Defects in Skeletal Morphogenesis” (P. Royce and B. Steinmann, eds.). Wiley, New York, in press. 35. C. Marcelle, S. Ahlgren, and M. Bronner-Fraser, In vivo regulation of somite differentiation and proliferation by Sonic Hedgehog. Dev. Biol. 214, 277 – 287 (1999). 36. C. Chiang, Y. Litingtung, E. Lee, K. E. Young, J. L. Corden, H. Westphal, and P. A. Beachy, Cyclopia and defective axial patterning in mice lacking Sonic hedgehog gene function. Nature 383, 407 – 413 (1996). 37. L. C. Murtaugh, J. H. Chyung, and A. B. Lassar, Sonic hedgehog promotes somitic chondrogenesis by altering the cellular response to BMP signaling. Genes Dev. 13, 225 – 237 (1999). 38. J. Wallin, J. Wilting, H. Koseki, R. Fritsch, B. Christ, and R. Balling, The role of Pax-1 in axial skeleton development. Development 120, 1109 – 1121 (1994). 39. B. Wilm, E. Dahl, H. Peters, R. Balling, and K. Imai, Targeted disruption of Pax1 defines its null phenotype and proves haploinsufficiency. Proc. Natl. Acad. Sci. USA 95, 8692 – 8697 (1998). 40. H. Peters, B. Wilm, N. Sakai, K. Imai, R. Maas, and R. Balling, Pax1 and Pax9 synergistically regulate vertebral column development. Development 126, 5399 – 5408 (1999). 41. R. Krumlauf, Hox genes in vertebrate development. Cell 78, 191 – 201 (1994). 42. A. C. Burke, C. E. Nelson, B. A. Morgan, and C. Tabin, Hox genes and the evolution of vertebrate axial morphology. Development 121, 333 – 346 (1995). 43. B. Favier, and P. Dolle, Developmental functions of mammalian Hox genes. Mol. Hum. Reprod. 3, 115 – 131 (1997). 44. F. Chen, J. Greer, and M. R. Capecchi, Analysis of Hoxa7/Hoxb7 mutants suggests periodicity in the generation of the different sets of vertebrae. Mech. Dev. 77, 49 – 57 (1998). 45. J. Zakany, M. Gerard, B. Favier, and D. Duboule, Deletion of a HoxD enhancer induces transcriptional heterochrony leading to transposition of the sacrum. EMBO J. 16, 4393 – 4402 (1997). 46. R. L. Johnson and C. J. Tabin, Molecular models for vertebrate limb development. Cell 90, 979 – 990 (1997). 47. C. Tickle and G. Eichele, Vertebrate limb development. Annu. Rev. Cell Biol. 10, 121 – 152 (1994). 48. J. K. Ng, K. Tamura, D. Buscher, and J. C. Izpisua-Belmonte, Molecular and cellular basis of pattern formation during vertebrate limb development. Curr. Top. Dev. Biol. 41, 37 – 66 (1999). 49. B. A. Parr and A. P. McMahon, Dorsalizing signal Wnt-7a required for normal polarity of D-V and A-P axes of mouse limb. Nature 374, 350 – 353 (1995). 50. C. A. Loomis, R. A. Kimmel, C. X. Tong, J. Michaud, and A. L. Joyner, Analysis of the genetic pathway leading to formation of ectopic apical ectodermal ridges in mouse Engrailed-1 mutant limbs. Development 125, 1137 – 1148 (1998). 51. R. D. Riddle, R. L. Johnson, E. Laufer, and C. Tabin, Sonic hedgehog mediates the polarizing activity of the ZPA. Cell 75, 1401 – 1416 (1993). 52. X. Sun, M. Lewandoski, E. N. Meyers, H. Y. Liu, R. E. Maxson, and G. R. Martin, Conditinal inactivation of Fgf4 reveals complexity of signalling during limb bud development. Nature Genet. 25, 83 – 86 (2000). 53. A. Zuniga, A. P. Haramis, A. P. McMahon, and R. Zeller, Signal relay by BMP antagonism controls the SHH/FGF4 feedback loop in vertebrate limb buds. Nature 401, 598 – 602 (1999). 54. J. Charite, D. G. McFadden, and E. N. Olson, The bHLH transcription factor dHAND controls Sonic hedgehog expression and establishment of the zone of polarizing activity during limb development. Development 127, 2461 – 2470 (2000). 55. M. Fernandez, Teran, M. E. Piedra, I. S. Kathiriya, D. Srivastava, J. C. Rodriguez Rey, and M. A. Ros, Role of dHAND in the anterior-posterior
56.
57.
58. 59.
60.
61.
62. 63.
64.
65.
66.
67.
68.
69. 70.
71.
72.
73.
polarization of the limb bud: Implications for the Sonic hedgehog pathway. Development 127, 2133 –2142 (2000). J. Capdevila, T. Tsukui, C. Rodriquez Esteban, V. Zappavigna, and J. C. Izpisua Belmonte, Control of vertebrate limb outgrowth by the proximal factor Meis2 and distal antagonism of BMPs by Gremlin. Mol. Cell 4, 839 – 849 (1999). N. Mercader, E. Leonardo, N. Azpiazu, A. Serrano, G. Morata, C. Martinez, and M. Torres, Conserved regulation of proximodistal limb axis development by Meis1/Hth. Nature 402, 425 – 429 (1999). H. Simon, T-box genes and the formation of vertebrate forelimb- and hindlimb specific pattern. Cell Tissue Res. 296, 57 – 66 (1999). D. L. Chapman, N. Garvey, S. Hancock, M. Alexiou, S. I. Agulnik, J. J. Gibson Brown, J. Cebra Thomas, R. J. Bollag, L. M. Silver, and V. E. Papaioannou, Expression of the T-box family genes, Tbx1-Tbx5, during early mouse development. Dev. Dyn. 206, 379 – 390 (1996). J. K. Takeuchi, K. Koshiba Takeuchi, K. Matsumoto, A. VogelHopker, M. Naitoh-Matsuo, K. Ogura, N. Takahashi, K. Yasuda, and T. Ogura, Tbx5 and Tbx4 genes determine the wing/leg identity of limb buds. Nature 398, 810 – 814 (1999). C. Rodriguez-Esteban, T. Tsukui, S. Yonei, J. Magallon, K. Tamura, and J. C. Izpisua Belmonte, The T-box genes Tbx4 and Tbx5 regulate limb outgrowth and identity. Nature 398, 814 – 818 (1999). M. Logan and C. J. Tabin, Role of Pitx1 upstream of Tbx4 in specification of hindlimb identity. Science 283, 1736 – 1739 (1999). C. Lanctot, A. Moreau, M. Chamberland, M. L. Tremblay, and J. Drouin, Hindlimb patterning and mandible development require the Ptx1 gene. Development 126, 1805 – 1810 (1999). D. P. Szeto, C. Rodriguez-Esteban, A. K. Ryan, S. M. O’Connell, F. Liu, C. Kioussi, A. S. Gleiberman, Izpisua-J. C. Belmonte, and M. G. Rosenfeld, Role of the Bicoid-related homeodomain factor Pitx1 in specifying hindlimb morphogenesis and pituitary development. Genes Dev. 13, 484 – 494 (1999). P. Dolle, J. C. Izpisua-Belmonte, E. Boncinelli, and D. Duboule, The Hox-4.8 gene is localized at the 5 extremity of the Hox-4. Mech. Dev. 36, 3 – 13 (1991). C. E. Nelson, B. A. Morgan, A. C. Burke, E. Laufer, E. DiMambro, L. C. Murtaugh, E. Gonzales, L. Tessarollo, L. F. Parada, and C. Tabin, Analysis of Hox gene expression in the chick limb bud. Development 122, 1449 – 1466 (1996). Y. Muragaki, S. Mundlos, J. Upton, and B. R. Olsen, Altered growth and branching patterns in synpolydactyly caused by mutations in HOXD13. Science 272, 548 – 551 (1996). F. R. Goodman, S. Mundlos, Y. Muragaki, D. Donnai, M. L. Giovannucci-Uzielli, E. Lapi, F. Majewski, J. McGaughran, C. McKeown, W. Reardon, J. Upton, R. M. Winter, B. R. Olsen, and P. J. Scambler, Synpolydactyly phenotypes correlate with size of expansions in HOXD13 polyalanine tract. Proc. Natl. Acad. Sci. USA 94, 7458 – 7463 (1997). J. Zakany and D. Duboule, Hox genes in digit development and evolution. Cell Tissue Res. 296, 19 – 25 (1999). K. R. Johnson, H. O. Sweet, L. R. Donahue, P. Ward-Bailey, R. T. Bronson, and M. T. Davisson, A new spontaneous mouse mutation of Hoxd13 with a polyalanine expansion and phenotype similar to human synpolydactyly. Hum. Mol. Genet. 7, 1033 – 1038 (1998). V. Knezevic, R. De Santo, K. Schughart, U. Huffstadt, C. Chiang, K. A. Mahon, and S. Mackem, Hoxd-12 differentially affects preaxial and postaxial chondrogenic branches in the limb and regulates Sonic hedgehog in a positive feedback loop. Development 124, 4523 – 4536 (1997). J. J. Lee, D. P. von Kessler, S. Parks, and P. A. Beachy, Secretion and localized transcription suggest a role in positional signaling for products of the segmentation gene hedgehog. Cell 71, 33 – 50 (1992). Y. Echelard, D. J. Epstein, B. St-Jacques, L. Shen, J. Mohler, J. A. McMahon, and A. P. McMahon, Sonic hedgehog, a member of a
206
74.
75.
76. 77. 78.
79.
80.
81. 82.
83.
84.
85.
86.
87.
88.
89.
90.
91.
92.
REGINATO, WANG, AND OLSEN family of putative signaling molecules, is implicated in the regulation of CNS polarity. Cell 75, 1417 – 1430 (1993). J. J. Lee, S. C. Ekker, D. P. von Kessler, J. A. Porter, B. I. Sun, and P. A. Beachy, Autoproteolysis in hedgehog protein biogenesis. Science 266, 1528 – 1537 (1994). J. A. Porter, K. E. Young, and P. A. Beachy, Cholesterol modification of hedgehog signaling proteins in animal development. Science 274, 255 – 259 (1996). P. W. Ingham, Transducing Hedgehog: The story so far. EMBO J. 17, 3505 – 3511 (1998). Y. Chen and G. Struhl, Dual roles for patched in sequestering and transducing Hedgehog. Cell 87, 553 – 563 (1996). D. M. Stone, M. Hynes, M. Armanini, T. A. Swanson, Q. Gu, R. L. Johnson, M. P. Scott, D. Pennica, A. Goddard, H. Phillips, M. Noll, J. E. Hooper, F. de Sauvage, and A. Rosenthal, The tumoursuppressor gene patched encodes a candidate receptor for Sonic hedgehog. Nature 384, 129 – 134 (1996). M. van den Heuvel and P. W. Ingham, Smoothened encodes a receptor-like serpentine protein required for hedgehog signalling. Nature 382, 547 – 551 (1996). J. Alcedo, M. Ayzenzon, T. Von Ohlen, M. Noll, and J. E. Hooper, The Drosophila smoothened gene encodes a seven-pass membrane protein, a putative receptor for the hedgehog signal. Cell 86, 221 – 232 (1996). P. Aza-Blanc and T. B. Kornberg, Ci: A complex transducer of the hedgehog signal. Trends Genet. 15, 458 – 462 (1999). C. H. Chen, D. P. von Kessler, W. Park, B. Wang, Y. Ma, and P. A. Beachy (1999). Nuclear trafficking of Cubitus interruptus in the transcriptional regulation of Hedgehog target gene expression. Cell 98, 305 – 316. (1999) Y. Chen, N. Gallaher, R. H. Goodman, and S. M. Smolik, Protein kinase A directly regulates the activity and proteolysis of cubitus interruptus. Proc. Natl. Acad. Sci. USA 95, 2349 – 2354 (1998). J. M. Ruppert, K. W. Kinzler, A. J. Wong, S. H. Bigner, F. T. Kao, M. L. Law, H. N. Seuanez, S. J. O’Brien, and B. Vogelstein, The GLIKrüppel family of human genes. Mol. Cell Biol. 8, 3104 – 3113 (1988). B. Wang, J. F. Fallon, and P. A. Beachy, Hedgehog-regulated processing of Gli3 produces an anterior/posterior repressor gradient in the developing vertebrate limb. Cell 100, 423 – 434 (2000). P. Dai, H. Akimaru, Y. Tanaka, T. Maekawa, M. Nakafuku, and S. Ishii, Sonic Hedgehog-induced activation of the Gli1 promoter is mediated by GLI3. J. Biol. Chem. 274, 8143 – 8152 (1999). A. Ruiz i Altaba, Combinatorial Gli gene function in floor plate and neuronal inductions by Sonic hedgehog. Development 125, 2203 – 2212 (1998). C. C. Hui, D. Slusarski, K. A. Platt, R. Holmgren, and A. L. Joyner, Expression of three mouse homologs of the Drosophila segment polarity gene cubitus interruptus, Gli, Gli-2, and Gli-3, in ectodermand mesoderm-derived tissues suggests multiple roles during postimplantation development. Dev. Biol. 162, 402 – 413 (1994). H. L. Park, C. Bai, K. A. Platt, M. P. Matise, A. Beeghly, C. Hui, M. Nakashima, and A. L. Joyner, Mouse gli1 mutants are viable but have defects in SHH signaling in combination with a gli2 mutation. Development 127, 1593 – 1605 (2000). J. Motoyama, J. Liu, R. Mo, Q. Ding, M. Post, and C. C. Hui, Essential function of Gli2 and Gli3 in the formation of lung, trachea and oesophagus Nature Genet. 20, 54 – 57 (1998). R. Mo, A. M. Freer, D. L. Zinyk, M. A. Crackower, J. Michaud, H. H. Heng, K. W. Chik, X. M. Shi, L. C. Tsui, S. H. Cheng, A. L. Joyner, and C. Hui, Specific and redundant functions of Gli2 and Gli3 zinc finger genes in skeletal patterning and development. Development 124, 113 – 123 (1997). Z. Hardcastle, R. Mo, C. C. Hui, and P. T. Sharpe, The Shh signalling pathway in tooth development: Defects in Gli2 and Gli3 mutants. Development 125, 2803 – 2811 (1998).
93. Q. Ding, J. Motoyama, S. Gasca, R. Mo, H. Sasaki, J. Rossant, and C. C. Hui, Diminished Sonic hedgehog signaling and lack of floor plate differentiation in Gli2 mutant mice. Development 125, 2533 – 2543 (1998). 94. M. P. Matise, D. J. Epstein, H. L. Park, K. A. Platt, and A. L. Joyner, Gli2 is required for induction of floor plate and adjacent cells, but not most ventral neurons in the mouse central nervous system. Development 125, 2759 – 2770 (1998). 95. L. G. Biesecker, Strike three for GLI3 [published erratum appears in Nature Genet. 18, 88] (1998). Nature Genet. 17, 259 – 260 (1997). 96. S. Kang, J. M. Graham, Jr., A. H. Olney, and L. G. Biesecker, GLI3 frameshift mutations cause autosomal dominant Pallister-Hall syndrome. Nature Genet. 15, 266 – 268 (1997). 97. U. Radhakrishna, A. Wild, K. H. Grzeschik, and S. E. Antonarakis, Mutation in GLI3 in postaxial polydactyly type A. Nature Genet. 17, 269 – 271 (1997). 98. A. Vortkamp, M. Gessler, and K. H. Grzeschik, GLI3 zinc-finger gene interrupted by translocations in Greig syndrome families. Nature 352, 539 – 540 (1991). 99. C. C. Hui and A. L. Joyner, A mouse model of Greig cephalopolysyndactyly syndrome: The extra-toes J mutation contains an intragenic deletion of the Gli3 gene. Nature Genet. 3, 241 – 246. 100. H. Thien and U. Ruther, The mouse mutation Pdn (Polydactyly Nagoya) is caused by the integration of a retrotransposon into the Gli3 gene. Mamm. Genome 10, 205 – 209 (1999). 101. B. L. Hogan, Bone morphogenetic proteins: Multifunctional regulators of vertebrate development. Genes Dev. 10, 1580 – 1594 (1996). 102. D. M. Kingsley, The TGF-beta superfamily: new members, new receptors, and new genetic tests of function in different organisms. Genes Dev. 8, 133 – 146 (1994). 103. D. M. Kingsley, What do BMPs do in mammals? Clues from the mouse short-ear mutation. Trends Genet. 10, 16 – 21 (1994). 104. L. Niswander and G. R. Martin, FGF-4 and BMP-2 have opposite effects on limb growth. Nature 361, 68 – 71 (1993). 105. Y. Ganan, D. Macias, R. D. Basco, R. Merino, and J. M. Hurle, Morphological diversity of the avian foot is related with the pattern of msx gene expression in the developing autopod. Dev. Biol. 196, 33 – 41 (1998). 106. S. Pizette and L. Niswander, BMPs negatively regulate structure and function of the limb apical ectodermal ridge. Development 126, 883 – 894 (1999). 107. Y. Kawakami, T. Ishikawa, M. Shimabara, N. Tanda, M. EnomotoIwamoto, M. Iwamoto, T. Kuwana, A. Ueki, S. Noji, and T. Nohno, BMP signaling during bone pattern determination in the developing limb. Development 122, 3557 – 3566 (1996). 108. D. M. Duprez, K. Kostakopoulou, P. H. Francis-West, C. Tickle, and P. M. Brickell, Activation of Fgf-4 and HoxD gene expression by BMP-2 expressing cells in the developing chick limb. Development 122, 1821 – 1828 (1996). 109. D. Duprez, E. J. Bell, M. K. Richardson, C. W. Archer, L. Wolpert, P. M. Brickell, and P. H. Francis-West, Overexpression of BMP-2 and BMP-4 alters the size and shape of developing skeletal elements in the chick limb. Mech. Dev. 57, 145 – 157 (1996). 110. D. M. Duprez, M. Coltey, H. Amthor, P. M. Brickell, and C. Tickle, Bone morphogenetic protein-2 (BMP-2) inhibits muscle development and promotes cartilage formation in chick limb bud cultures. Dev. Biol. 174, 448 – 452 (1996). 111. D. Macias, Y. Ganan, T. K. Sampath, M. E. Piedra, M. A. Ros, and J. M. Hurle, Role of BMP-2 and OP-1 (BMP-7) in programmed cell death and skeletogenesis during chick limb development. Development 124, 1109 – 1117 (1997). 112. R. Merino, Y. Ganan, D. Macias, A. N. Economides, K. T. Sampath, and J. M. Hurle, Morphogenesis of digits in the avian limb is controlled by FGFs, TGFs, and noggin through BMP signaling. Dev. Biol. 200, 35 – 45 (1998).
CHAPTER 5 Developmental Biology of Bone 113. M. Enomoto-Iwamoto, M. Iwamoto, Y. Mukudai, Y. Kawakami, T. Nohno, Y. Higuchi, S. Takemoto, H. Ohuchi, S. Noji, and K. Kurisu (1998). Bone morphogenetic protein signaling is required for maintenance of differentiated phenotype, control of proliferation, and hypertrophy in chondrocytes. J. Cell Biol. 140, 409 – 418 (1998). 114. H. Amthor, B. Christ, M. Weil, and K. Patel, The importance of timing differentiation during limb muscle development. Curr. Biol. 8, 642 – 652 (1998). 115. H. Zou and L. Niswander, Requirement for BMP signaling in interdigital apoptosis and scale formation. Science 272, 738 – 741 (1996). 116. Y. Yokouchi, J. Sakiyama, T. Kameda, H. Iba, A. Suzuki, N. Ueno, and A. Kuroiwa, BMP-2/-4 mediate programmed cell death in chicken limb buds. Development 122, 3725 – 3734 (1996). 117. Y. Ganan, D. Macias, M. Duterque-Coquillaud, M. A. Ros, and J. M. Hurle, Role of TGFs and BMPs as signals controlling the position of the digits and the areas of interdigital cell death in the developing chick limb autopod. Development 122, 2349 – 2357 (1996). 118. M. R. Urist, Bone: Formation by autoinduction. Science 150, 893 – 899 (1965). 119. J. M. Wozney, V. Rosen, A. J. Celeste, L. M. Mitsock, M. J. Whitters, R. W. Kriz, R. M. Hewick, and E. A. Wang, Novel regulators of bone formation: Molecular clones and activities. Science 242, 1528 – 1534 (1988). 120. A. H. Reddi, Role of morphogenetic proteins in skeletal tissue engineering and regeneration. Nature Biotechnol. 16, 247 – 252 (1998). 121. J. M. Wozney, Bone morphogenetic proteins and their expression. In “Bone Morphogenetic Proteins and Their expression” (M. Noda, ed.), pp. 131 – 165 Academic Press, San Diego, (1993). 122. S. Cheifetz, BMP receptors in limb and tooth formation. Crit. Rev. Oral Biol. Med. 10, 182 – 198 (1999). 123. K. W. Cho and I. L. Blitz, BMPs, Smads and metalloproteases: Extracellular and intracellular modes of negative regulation. Curr. Opin. Genet. Dev. 8, 443 – 449 (1998). 124. F. Liu, A. Hata, J. C. Baker, J. Doody, J. Carcamo, R. M. Harland, and J. Massague, A human Mad protein acting as a BMP-regulated transcriptional activator. Nature 381, 620 – 623 (1996). 125. T. Nohno, T. Ishikawa, T. Saito, K. Hosokawa, S. Noji, D. H. Wolsing, and J. S. Rosenbaum, Identification of a human type II receptor for bone morphogenetic protein-4 that forms differential heteromeric complexes with bone morphogenetic protein type I receptors. J. Biol. Chem. 270, 22522 – 22526 (1995). 126. B. L. Rosenzweig, T. Imamura, T. Okadome, G. N. Cox, H. Yamashita, P. ten Dijke, C. H. Heldin, and K. Miyazono, Cloning and characterization of a human type II receptor for bone morphogenetic proteins. Proc. Natl. Acad. Sci. USA 92, 7632 – 7636 (1995). 127. J. L. Wrana, L. Attisano, R. Wieser, F. Ventura, and J. Massague, Mechanism of activation of the TGF- receptor. Nature 370, 341 – 347 (1994). 128. J. Massague, TGF signaling: Receptors, transducers, and Mad proteins. Cell 85, 947 – 950 (1996). 129. J. Carcamo, F. M. Weis, F. Ventura, R. Wieser, J. L. Wrana, L. Attisano, and J. Massague, Type I receptors specify growth-inhibitory and transcriptional responses to transforming growth factor and activin. Mol. Cell Biol. 14, 3810 – 3821 (1994). 130. A. Hata, G. Lagna, J. Massague, and A. Hemmati-Brivanlou, Smad6 inhibits BMP/Smad1 signaling by specifically competing with the Smad4 tumor suppressor. Genes Dev. 12, 186 – 197 (1998). 131. T. Imamura, M. Takase, A. Nishihara, E. Oeda, J. Hanai, M. Kawabata, and K. Miyazono, Smad6 inhibits signalling by the TGF- superfamily Nature 389, 622 – 626 (1997). 132. J. Kim, K. Johnson, H. J. Chen, S. Carroll, and A. Laughon, Drosophila Mad binds to DNA and directly mediates activation of vestigial by Decapentaplegic. Nature 388, 304 – 308 (1997).
207 133. H. Hayashi, S. Abdollah, Y. Qiu, J. Cai, Y. Y. Xu, B. W. Grinnell, M. A. Richardson, J. N. Topper, M. A. Gimbrone, Jr., J. L. Wrana, and D. Falb, The MAD-related protein Smad7 associates with the TGF receptor and functions as an antagonist of TGF signaling. Cell 89, 1165 – 1173 (1997). 134. W. C. Smith and R. M. Harland, Expression cloning of noggin, a new dorsalizing factor localized to the Spemann organizer in Xenopus embryos. Cell 70, 829 – 840 (1992). 135. E. Stanley, C. Biben, S. Kotecha, L. Fabri, S. Tajbakhsh, C. C. Wang, T. Hatzistavrou, B. Roberts, C. Drinkwater, M. Lah, M. Buckingham, D. Hilton, A. Nash, T. Mohun, and R. P. Harvey, DAN is a secreted glycoprotein related to Xenopus cerberus. Mech. Dev. 77, 173 – 184 (1998). 136. J. J. Pearce, G. Penny, and J. Rossant, A mouse cerberus/Dan-related gene family. Dev. Biol. 209, 98 – 110 (1999). 137. Y. Sasai, B. Lu, H. Steinbeisser, D. Geissert, L. K. Gont, and E. M. De Robertis, Xenopus chordin: A novel dorsalizing factor activated by organizer-specific homeobox genes. Cell 79, 779 – 790 (1994). 138. F. S. Esch, S. Shimasaki, M. Mercado, K. Cooksey, N. Ling, S. Ying, N. Ueno, and R. Guillemin. Structural characterization of follistatin: A novel follicle-stimulating hormone release-inhibiting polypeptide from the gonad. Mol. Endocrinol. 1, 849 – 855 (1987). 139. L. B. Zimmerman, J. M. De Jesus-Escobar, and R. M. Harland, The Spemann organizer signal noggin binds and inactivates bone morphogenetic protein 4. Cell 86, 599 – 606 (1996). 140. L. J. Brunet, J. A. McMahon, A. P. McMahon, and R. M. Harland, Noggin, cartilage morphogenesis, and joint formation in the mammalian skeleton. Science 280, 1455 – 1457 (1998). 141. D. R. Hsu, A. N. Economides, X. Wang, P. M. Eimon, and R. M. Harland, The Xenopus dorsalizing factor Gremlin identifies a novel family of secreted proteins that antagonize BMP activities. Mol. Cell 1, 673 – 683 (1998). 142. H. Zhang and A. Bradley, Mice deficient for BMP2 are nonviable and have defects in amnion/chorion and cardiac development. Development 122, 2977 – 2986 (1996). 143. G. Winnier, M. Blessing, P. A. Labosky, and B. L. Hogan, Bone morphogenetic protein-4 is required for mesoderm formation and patterning in the mouse. Genes Dev. 9, 2105 – 2116 (1995). 144. Y. Mishina, A. Suzuki, N. Ueno, and R. R. Behringer, Bmpr encodes a type I bone morphogenetic protein receptor that is essential for gastrulation during mouse embryogenesis. Genes Dev. 9, 3027 – 3037 (1995). 145. A. T. Dudley, K. M. Lyons, and E. J. Robertson, A requirement for bone morphogenetic protein-7 during development of the mammalian kidney and eye. Genes Dev. 9, 2795 – 2807 (1995). 146. G. Luo, C. Hofmann, A. L. Bronckers, M. Sohocki, A. Bradley, and G. Karsenty, BMP-7 is an inducer of nephrogenesis, and is also required for eye development and skeletal patterning. Genes Dev. 9, 2808 – 2820 (1995). 147. A. C. McPherron, A. M. Lawler, and S. J. Lee, Regulation of anterior/posterior patterning of the axial skeleton by growth/differentiation factor 11. Nature Genet. 22, 260 – 264 (1999). 148. D. M. Kingsley, A. E. Bland, J. M. Grubber, P. C. Marker, L. B. Russell, N. G. Copeland, and N. A. Jenkins, The mouse short ear skeletal morphogenesis locus is associated with defects in a bone morphogenetic member of the TGF superfamily. Cell 71, 399 – 410 (1992). 149. E. E. Storm, T. V. Huynh, N. G. Copeland, N. A. Jenkins, D. M. Kingsley, and S. J. Lee, Limb alterations in brachypodism mice due to mutations in a new member of the TGF -superfamily. Nature 368, 639 – 643 (1994). 150. A. Polinkovsky, N. H. Robin, J. T. Thomas, M. Irons, A. Lynn, F. R. Goodman, W. Reardon, S. G. Kant, H. G. Brunner, I. van der Burgt, D. Chitayat, J. McGaughran, D. Donnai, F. P. Luyten, and M. L. Warman, Mutations in CDMP1 cause autosomal dominant brachydactyly type C. Nature Genet. 17, 18 – 19 (1997).
208 151. J. T. Thomas, K. Lin, M. Nandedkar, M. Camargo, J. Cervenka, and F. P. Luyten, A human chondrodysplasia due to a mutation in a TGFbeta superfamily member. Nature Genet. 12, 315 – 317 (1996). 152. J. T. Thomas, M. W. Kilpatrick, K. Lin, L. Erlacher, P. Lembessis, T. Costa, P. Tsipouras, and F. P. Luyten, Disruption of human limb morphogenesis by a dominant negative mutation in CDMP1. Nature Genet. 17, 58 – 64 (1997). 153. Y. Gong, D. Krakow, J. Marcelino, D. Wilkin, D. Chitayat, R. BabulHirji, L. Hudgins, C. W. Cremers, F. P. Cremers, H. G. Brunner, K. Reinker, D. L. Rimoin, D. H. Cohn, F. R. Goodman, W. Reardon, M. Patton, C. A. Francomano, and M. L. Warman, Heterozygous mutations in the gene encoding noggin affect human joint morphogenesis. Nature Genet. 21, 302 – 304 (1999). 154. S. T. Baur, J. J. Mai, and S. M. Dymecki, Combinatorial signaling through BMP receptor IB and GDF5: Shaping of the distal mouse limb and the genetics of distal limb diversity. Development 127, 605 – 619 (2000). 155. S. E. Yi, A. Daluiski, R. Pederson, V. Rosen, and K. M. Lyons, The type I BMP receptor BMPRIB is required for chondrogenesis in the mouse limb. Development 127, 621 – 630 (2000). 156. H. A. Armelin, Pituitary extracts and steroid hormones in the control of 3T3 cell growth. Proc. Natl. Acad. Sci. USA 70, 2702 – 2706 (1973). 157. D. Gospodarowicz, Localisation of a fibroblast growth factor and its effect alone and with hydrocortisone on 3T3 cell growth. Nature 249, 123 – 127 (1974). 158. M. Klagsbrun, The fibroblast growth factor family: Structural and biological properties. Prog Growth Factor Res 1, 207 – 235 (1989). 159. G. R. Martin, The roles of FGFs in the early development of vertebrate limbs. Genes Dev. 12, 1571 – 1586 (1998). 160. M. C. Naski and D. M. Ornitz, FGF signaling in skeletal development. Front. Biosci. 3, 781 – 794 (1998). 161. S. Faham, R. E. Hileman, J. R. Fromm, R. J. Linhardt, and D. C. Rees, Heparin structure and interactions with basic fibroblast growth factor. Science 271, 1116 – 1120 (1996). 162. N. Perrimon and M. Bernfield, Specificities of heparan sulphate proteoglycans in developmental processes. Nature 404, 725 – 728 (2000). 163. C. Basilico and D. Moscatelli, The FGF family of growth factors and oncogenes. Adv. Cancer Res. 59, 115 – 165 (1992). 164. I. J. Mason, The ins and outs of fibroblast growth factors. Cell 78, 547 – 552 (1994). 165. B. B. Olwin and A. Rapraeger, Repression of myogenic differentiation by aFGF, bFGF, and K-FGF is dependent on cellular heparan sulfate. J. Cell Biol. 118, 631 – 639 (1992). 166. D. E. Johnson and L. T. Williams, Structural and functional diversity in the FGF receptor multigene family. Adv. Cancer Res. 60, 1 – 41 (1993). 167. T. Miki, D. P. Bottaro, T. P. Fleming, C. L. Smith, W. H. Burgess, A. M. Chan, and S. A. Aaronson, Determination of ligand-binding specificity by alternative splicing: Two distinct growth factor receptors encoded by a single gene. Proc. Natl. Acad. Sci. USA 89, 246 – 250 (1992). 168. A. T. Chellaiah, D. G. McEwen, S. Werner, J. Xu, and D. M. Ornitz, Fibroblast growth factor receptor (FGFR) 3: Alternative splicing in immunoglobulin-like domain III creates a receptor highly specific for acidic FGF/FGF-1. J. Biol. Chem. 269, 11620 – 11627 (1994). 169. A. Avivi, A. Yayon, and D. Givol, A novel form of FGF receptor-3 using an alternative exon in the immunoglobulin domain III. FEBS Lett. 330, 249 – 252 (1993). 170. G. Yan, Y. Fukabori, G. McBride, S. Nikolaropolous, and W. L. McKeehan, Exon switching and activation of stromal and embryonic fibroblast growth factor (FGF)-FGF receptor genes in prostate epithelial cells accompany stromal independence and malignancy. Mol. Cell Biol. 13, 4513 – 4522 (1993).
REGINATO, WANG, AND OLSEN 171. E. Gilbert, F. Del Gatto, P. Champion-Arnaud, M. C. Gesnel, and R. Breathnach, Control of BEK and K-SAM splice sites in alternative splicing of the fibroblast growth factor receptor 2 pre-mRNA. Mol. Cell Biol. 13, 5461 – 5468 (1993). 172. E. Scotet and E. Houssaint, The choice between alternative IIIb and IIIc exons of the FGFR-3 gene is not strictly tissue-specific. Biochim. Biophys. Acta 1264, 238 – 242 (1995). 173. F. Bellot, G. Crumley, J. M. Kaplow, J. Schlessinger, M. Jaye, and C. A. Dionne, Ligand-induced transphosphorylation between different FGF receptors. EMBO J. 10, 2849 – 2854 (1991). 174. G. Szebenyi, and J. F. Fallon, Fibroblast growth factors as multifunctional signaling factors. Int. Rev. Cytol. 185, 45 – 106 (1999). 175. H. Kouhara, Y. R. Hadari, T. Spivak-Kroizman, J. Schilling, D. Bar-Sagi, I. Lax, and J. Schlessinger, A lipid-anchored Grb2-binding protein that links FGF-receptor activation to the Ras/MAPK signaling pathway. Cell 89, 693 – 702 (1997). 176. G. Sa and T. Das, Basic fibroblast growth factor stimulates cytosolic phospholipase A2, phospholipase C-gamma1 and phospholipase D through distinguishable signaling mechanisms. Mol. Cell. Biochem. 198, 19 – 30 (1999). 177. D. Burke, D. Wilkes, T. L. Blundell, and S. Malcolm, Fibroblast growth factor receptors: lessons from the genes. Trends Biochem. Sci. 23, 59 – 62 (1998). 178. M. R. Passos-Bueno, W. R. Wilcox, E. W. Jabs, A. L. Sertie, L. G. Alonso, and H. Kitoh, Clinical spectrum of fibroblast growth factor receptor mutations. Hum. Mutat. 14, 115 – 125 (1999). 179. M. K. Webster and D. J. Donoghue, FGFR activation in skeletal disorders: Too much of a good thing. Trends Genet. 13, 178 – 182 (1997). 180. M. Muenke, K. W. Gripp, D. M. McDonald-McGinn, K. Gaudenz, L. A. Whitaker, S. P. Bartlett, R. I. Markowitz, N. H. Robin, N. Nwokoro, J. J. Mulvihill, H. W. Losken, J. B. Mulliken, A. E. Guttmacher, R. S. Wilroy, L. A. Clarke, G. Hollway, L. C. Ades, E. A. Haan, J. C. Mulley, M. M. Cohen, Jr., G. A. Bellus, C. A. Francomano, D. M. Moloney, S. A. Wall, and A. O. Wilkie, A unique point mutation in the fibroblast growth factor receptor 3 gene (FGFR3) defines a new craniosynostosis syndrome. Am. J. Hum. Genet. 60, 555 – 564 (1997). 181. D. Wilkes, P. Rutland, L. J. Pulleyn, W. Reardon, C. Moss, J. P. Ellis, R. M. Winter, and S. Malcolm, A recurrent mutation, ala391glu, in the transmembrane region of FGFR3 causes Crouzon syndrome and acanthosis nigricans. J. Med. Genet. 33, 744 – 748 (1996). 182. G. A. Meyers, S. J. Orlow, I. R. Munro, K. A. Przylepa, and E. W. Jabs, Fibroblast growth factor receptor 3 (FGFR3) transmembrane mutation in Crouzon syndrome with acanthosis nigricans. Nature Genet. 11, 462 – 464 (1995). 183. J. Bonaventure, F. Rousseau, L. Legeai-Mallet, M. Le Merrer, A. Munnich, and P. Maroteaux, Common mutations in the fibroblast growth factor receptor 3 (FGFR 3) gene account for achondroplasia, hypochondroplasia, and thanatophoric dwarfism. Am. J. Med. Genet. 63, 148 – 154 (1996). 184. M. M. Cohen, Jr., Achondroplasia, hypochondroplasia and thanatophoric dysplasia: Clinically related skeletal dysplasias that are also related at the molecular level. Int. J. Oral Maxillofac. Surg. 27, 451 – 455 (1998). 185. R. Shiang, L. M. Thompson, Y. Z. Zhu, D. M. Church, T. J. Fielder, M. Bocian, S. T. Winokur, and J. J. Wasmuth, Mutations in the transmembrane domain of FGFR3 cause the most common genetic form of dwarfism, achondroplasia. Cell 78, 335 – 342 (1994). 186. P. L. Tavormina, R. Shiang, L. M. Thompson, Y. Z. Zhu, D. J. Wilkin, R. S. Lachman, W. R. Wilcox, D. L. Rimoin, D. H. Cohn, and J. J. Wasmuth, Thanatophoric dysplasia (types I and II) caused by distinct mutations in fibroblast growth factor receptor 3. Nature Genet. 9, 321 – 328 (1995).
CHAPTER 5 Developmental Biology of Bone 187. F. Rousseau, P. Saugier, M. Le Merrer, A. Munnich, A. L. Delezoide, P. Maroteaux, J. Bonaventure, F. Narcy, and M. Sanak, Stop codon FGFR3 mutations in thanatophoric dwarfism type 1. Nature Genet. 10, 11 – 12 (1995). 188. G. A. Bellus, I. McIntosh, E. A. Smith, A. S. Aylsworth, I. Kaitila, W. A. Horton, G. A. Greenhaw, J. T. Hecht, and C. A. Francomano, A recurrent mutation in the tyrosine kinase domain of fibroblast growth factor receptor 3 causes hypochondroplasia. Nature Genet. 10, 357 – 359 (1995). 189. C. Deng, A. Wynshaw-Boris, F. Zhou, A. Kuo, and P. Leder, Fibroblast growth factor receptor 3 is a negative regulator of bone growth. Cell 84, 911 – 921 (1996). 190. J. S. Colvin, B. A. Bohne, G. W. Harding, D. G. McEwen, and D. M. Ornitz, Skeletal overgrowth and deafness in mice lacking fibroblast growth factor receptor 3. Nature Genet. 12, 390 – 397 (1996). 191. J. D. Coffin, R. Z. Florkiewicz, J. Neumann, T. Mort-Hopkins, G. W. Dorn, II, P. Lightfoot, R. German, P. N. Howles, A. Kier, and B. A. O’Toole, Abnormal bone growth and selective translational regulation in basic fibroblast growth factor (FGF-2) transgenic mice. Mol. Biol. Cell 6, 1861 – 1873 (1995). 192. A. Montero, Y. Okada, M. Tomita, M. Ito, H. Tsurukami, T. Nakamura, T. Doetschman, J. D. Coffin, and M. M. Hurley, Disruption of the fibroblast growth factor-2 gene results in decreased bone mass and bone formation. J. Clin. Invest. 105, 1085 – 1093 (2000). 193. V. C. Lui, L. J. Ng, J. Nicholls, P. P. Tam, and K. S. Cheah, Tissuespecific and differential expression of alternatively spliced alpha 1(II) collagen mRNAs in early human embryos. Dev. Dyn. 203, 198 – 211 (1995). 194. L. J. Ng, S. Wheatley, G. E. Muscat, J. Conway-Campbell, J. Bowles, E. Wright, D. M. Bell, P. P. Tam, K. S. Cheah, and P. Koopman, SOX9 binds DNA, activates transcription, and coexpresses with type II collagen during chondrogenesis in the mouse. Dev. Biol. 183, 108 – 121 (1997). 195. E. Wright, M. R. Hargrave, J. Christiansen, L. Cooper, J. Kun, T. Evans, U. Gangadharan, A. Greenfield, and P. Koopman, The Sryrelated gene Sox9 is expressed during chondrogenesis in mouse embryos. Nature Genet. 9, 15 – 20 (1995). 196. V. Lefebvre and B. de Crombrugghe, Toward understanding Sox9 function in chondrocyte differentiation. Matrix Biol. 16, 529 – 540 (1998). 197. D. M. Bell, K. K. Leung, S. C. Wheatley, L. J. Ng, S. Zhou, K. W. Ling, M. H. Sham, P. Koopman, P. P. Tam, and K. S. Cheah, Sox9 directly regulates the type-II collagen gene. Nature Genet. 16, 174 – 178 (1997). 198. K. K. Leung, L. J. Ng, K. K. Ho, P. P. Tam, and K. S. Cheah, Different cis-regulatory DNA elements mediate developmental stage- and tissue-specific expression of the human COL2A1 gene in transgenic mice. J. Cell Biol. 141, 1291 – 1300 (1998). 199. V. Lefebvre, W. Huang, V. R. Harley, P. N. Goodfellow, and B. de Crombrugghe, Sox9 is a potent activator of the chondrocyte-specific enhancer of the pro alpha1(II) collagen gene. Mol. Cell Biol. 17, 2336 – 2346 (1997). 200. W. Bi, J. M. Deng, Z. Zhang, R. R. Behringer, and B. de Crombrugghe, Sox9 is required for cartilage formation. Nat. Genet. 22, 85 – 89 (1999). 201. T. Wagner, J. Wirth, J. Meyer, B. Zabel, M. Held, J. Zimmer, J. Pasantes, F. Bricarelli, J. Keutel, E. Hustert, U. Wolf, N. Tommerup, W. Schempp, and G. Scherer, Autosomal sex reversal and campomelic dysplasia are caused by mutations in and around the SRY-related gene SOX9. Cell 79, 1111 – 1120 (1994). 202. J. W. Foster, M. A. Dominguez-Steglich, S. Guioli, G. Kowk, P. A. Weller, M. Stevanovic, J. Weissenbach, S. Mansour, I. D. Young, and P. N. Goodfellow, Campomelic dysplasia and autosomal sex reversal caused by mutations in an SRY-related gene. Nature 372, 525 – 530 (1994).
209 203. G. Zhou, V. Lefebvre, Z. Zhang, H. Eberspaecher, and B. de Crombrugghe, Three high mobility group-like sequences within a 48-base pair enhancer of the Col2a1 gene are required for cartilage-specific expression in vivo. J. Biol. Chem. 273, 14989 – 14997 (1998) 204. V. Lefebvre, P. Li, and B. de Crombrugghe, A new long form of Sox5 (L-Sox5), Sox6 and Sox9 are coexpressed in chondrogenesis and cooperatively activate the type II collagen gene. EMBO J. 17, 5718 – 5733 (1998). 205. E. E. Storm and D. M. Kingsley, GDF5 coordinates bone and joint formation during digit development. Dev. Biol. 209, 11 – 27 (1999). 206. H. Zou, K. M. Choe, Y. Lu, J. Massague, and L. Niswander, BMP signaling and vertebrate limb development. Cold Spring Harb. Symp. Quant. Biol. 62, 269 – 272 (1997). 207. M. J. Bitgood and A. P. McMahon, Hedgehog and Bmp genes are coexpressed at many diverse sites of cell-cell interaction in the mouse embryo. Dev. Biol. 172, 126 – 138 (1995). 208. B. Lanske, A. C. Karaplis, K. Lee, A. Luz, A. Vortkamp, A. Pirro, M. Karperien, L. H. K. Defize, C. Ho, R. C. Mulligan, A. B. AbouSamra, H. Juppner, G. V. Segre, and H. M. Kronenberg, PTH/PTHrP receptor in early development and Indian hedgehog-regulated bone growth. Science 273, 663 – 666 (1996). 209. A. C. Karaplis, A. Luz, J. Glowacki, R. T. Bronson, V. L. Tybulewicz, H. M. Kronenberg, and R. C. Mulligan, Lethal skeletal dysplasia from targeted disruption of the parathyroid hormone-related peptide gene. Genes Dev. 8, 277 – 289 (1994). 210. A. Vortkamp, K. Lee, B. Lanske, G. V. Segre, H. M. Kronenberg, and C. J. Tabin, Regulation of rate of cartilage differentiation by Indian hedgehog and PTH-related protein. Science 273, 613 – 622 (1996). 211. St-B. Jacques, M. Hammerschmidt, and A. P. McMahon, Indian hedgehog signaling regulates proliferation and differentiation of chondrocytes and is essential for bone formation. Genes Dev. 13, 2072 – 2086 (1999). 212. S. J. Karp, E. Schipani, B. St-Jacques, J. Hunzelman, H. Kronenberg, and A. P. McMahon, Indian Hedgehog coordinates endochondral bone gorwth and morphogenesis via parathyroid hormone relatedprotein-dependent and -independent pathways. Development 127, 543 – 548 (2000). 213. E. Schipani, K. Kruse, and H. Jüppner, A constitutively active mutant PTH-PTHrP receptor in Jansen-type metaphyseal chondrodysplasia. Science 268, 98 – 100 (1995). 214. A. S. Jobert, P. Zhang, A. Couvineau, J. Bonaventure, J. Roume, M. Le Merrer, and C. Silve, Absence of functional receptors for parathyroid hormone and parathyroid hormone-related peptide in Blomstrand chondrodysplasia. J. Clin. Invest. 102, 34 – 40 (1998). 215. M. C. Naski, J. S. Colvin, J. D. Coffin, and D. M. Ornitz, Repression of hedgehog signaling and BMP4 expression in growth plate cartilage by fibroblast growth factor receptor 3. Development 125, 4977 – 4988 (1998). 216. K. Iyama, Y. Ninomiya, B. R. Olsen, T. F. Linsenmayer, R. L. Trelstad, and M. Hayashi, Spatiotemporal pattern of type X collagen gene expression and collagen deposition in embryonic chick vertebrae undergoing endochondral ossification. Anat. Rec. 229, 462 – 472 (1991). 217. M. L. Warman, M. H. Abbott, S. S. Apte, T. Hefferon, I. McIntosh, D. H. Cohn, J. T. Hecht , B. R. Olsen, and C. A. Francomano, A type X collagen mutation causes Schmid metaphyseal chondrodysplasia. Nature Genet. 5, 79 – 82 (1993). 218. O. Jacenko, P. A. LuValle, and B. R. Olsen, Spondylometaphyseal dysplasia in mice carrying a dominant negative mutation in a matrix protein specific for cartilage-to-bone transition. Nature 365, 56 – 61 (1993). 219. H. P. Gerber, T. H. Vu, A. M. Ryan, J. Kowalski, Z. Werb, and N. Ferrara, VEGF couples hypertrophic cartilage remodeling, ossification and angiogenesis during endochondral bone formation. Nature Med. 5, 623 – 628 (1999).
210 220. N. A. Speck and T. Stacy, A new transcription factor family associated with human leukemias. Crit. Rev. Eukaryot. Gene Expr. 5, 337 – 364 (1995). 221. P. Ducy and G. Karsenty, Two distinct osteoblast-specific cis-acting elements control expression of a mouse osteocalcin gene. Mol. Cell. Biol. 15, 1858 – 1869 (1995). 222. P. Ducy, M. Starbuck, M. Priemel, J. Shen, G. Pinero, V. Geoffroy, M. Amling, and G. Karsenty, A Cbfa1-dependent genetic pathway controls bone formation beyond embryonic development. Genes Dev. 13, 1025 – 1036 (1999). 223. S. Mundlos, F. Otto, C. Mundlos, J. B. Mulliken, A. S. Aylsworth, S. Albright, D. Lindhout, W. G. Cole, W. Henn, J. H. M. Knoll, M. J. Owen, B. U. Zabel, R. Mertelsmann, and B. R. Olsen, Mutations involving the transcription factor CBFA1 cause cleidocranial dysplasia. Cell 89, 773 – 779 (1997). 224. L. F. Huang, N. Fukai, P. B. Selby, B. R. Olsen, and S. Mundlos, Mouse clavicular development: Analysis of wild-type and cleidocranial dysplasia mutant mice. Dev. Dyn. 210, 33 – 40 (1997). 225. I. S. Kim, F. Otto, B. Zabel, and S. Mundlos, Regulation of chondrocyte differentiation by Cbfa1. Mech. Dev. 80, 159 – 170 (1999). 226. M. Inada, T. Yasui, S. Nomura, S. Miyake, K. Deguchi, M. Himeno, M. Sato, H. Yamagiwa, T. Kimura, N. Yasui, T. Ochi, N. Endo, Y. Kitamura, T. Kishimoto, and T. Komori, Maturational disturbance of chondrocytes in Cbfa1-deficient mice. Dev. Dyn. 214, 279 – 290 (1999). 227. M. S. Kocher and F. Shapiro, Osteogenesis imperfecta J. Am. Acad. Orthop. Surg. 6, 225 – 236 (1998). 228. M. C. Willing, S. P. Deschenes, R. L. Slayton, and E. J. Roberts, Premature chain termination is a unifying mechanism for COL1A1 null alleles in osteogenesis imperfecta type I cell strains. Am. J. Hum. Genet. 59, 799 – 809 (1996). 229. S. B. M. Deak, A. F. Nicholls, F. M. Pope, and D. J. Prockop, The molecular defect in a nonlethal variant of osteogenesis imperfecta. J. Biol. Chem. 258, 15192 – 15197 (1983). 230. J. F. Bateman, I. Moeller, M. Hannagan, D. Chan, and W. G. Cole, Characterization of three osteogenesis imperfecta collagen a1(I) glycine to serine mutations demonstrating a position-dependent gradient of phenotypic severity. J. Biochem. 288, 131 – 135 (1992). 231. Q. Wang, B. M. Orrison, and J. C. Marini, Two additional cases of osteogenesis imperfecta with substitutions for glycine in the alpha 2(I) collagen chain: A regional model relating mutation location with phenotype. J. Biol. Chem. 268, 25162 – 25167 (1993). 232. S. Mundlos and B. R. Olsen, Heritable diseases of the skeleton. II. Molecular insights into skeletal development-matrix components and their homeostasis. FASEB J. 11, 227 – 233 (1997). 233. A. Schnieke, K. Harbers, and R. Jaenisch, Embryonic lethal mutation in mice induced by retrovirus insertion into the alpha 1(I) collagen gene. Nature 304, 315 – 320 (1983). 234. J. Bonadio, T. L. Saunders, E. Tsai, S. A. Goldstein, J. MorrisWiman, L. Brinkley, D. F. Dolan, R. A. Altschuler, J. E. Hawkins, Jr., J. F. Bateman, et al., Transgenic mouse model of the mild dominant form of osteogenesis imperfecta. Proc. Natl. Acad. Sci. USA 87, 7145 – 7149 (1990). 235. A. Stacey, J. Bateman, T. Choi, T. Mascara, W. Cole, and R. Jaenisch, Perinatal lethal osteogenesis imperfecta in transgenic mice bearing an engineered mutant pro-1(I) collagen gene. Nature 332, 131 – 135 (1988). 236. J. S. Khillan, A. S. Olsen, S. Kontusaari, B. Sokolov, and D. J. Prockop, Transgenic mice that express a mini-gene version of the human gene for type I procollagen (COL1A1) develop a phenotype resembling a lethal form of osteogenesis imperfecta. J. Biol. Chem. 266, 23373 – 23379 (1991). 237. R. Pereira, K. Halford, B. P. Sokolov, J. S. Khillan, and D. J. Prockop, Phenotypic variability and incomplete penetrance of spontaneous
REGINATO, WANG, AND OLSEN
238.
239.
240.
241.
242.
243.
244.
245.
246.
247.
248.
249.
250.
251.
252.
fractures in an inbred strain of transgenic mice expressing a mutated collagen gene (COL1A1). J. Clin. Invest. 93, 1765 – 1769 (1994). S. D. Chipman, H. O. Sweet, D. J. J. McBride, M. T. Davisson, S. C. J. Marks, A. R. Shuldiner, R. J. Wenstrup, D. W. Rowe, and J. R. Shapiro, Defective pro alpha 2(I) collagen synthesis in a recessive mutation in mice: A model of human osteogenesis imperfecta. Proc. Natl. Acad. Sci. USA 90, 1701 – 1705 (1993). T. Pihlajaniemi, L. A. Dickson, F. M. Pope, V. R. Korhonen, A. Nicholls, D. J. Prockop, and J. C. Myers, Osteogenesis imperfecta: Cloning of a pro2(I) Collagen gene with a frameshift mutations. J. Biol. Chem. 259, 12941 – 13944 (1984). J. Spranger, A. Winterpacht, and B. Zabel, The type II collagenopathies: A spectrum of chondrodysplasias. Eur. J. Pediatr 153, 56 – 65 (1994). D. Chan, W. G. Cole, C. W. Chow, S. Mundlos, and J. F. Bateman, A COL2A1 mutation in achondrogenesis type II results in the replacement of type II collagen by type I and III collagens in cartilage. J. Biol. Chem. 270, 1747 – 1753 (1995). D. Chan and W. G. Cole, Low basal transcription of genes for tissuespecific collagens by fibroblasts and lymphoblastoid cells: Application to the characterization of a glycine 997 to serine substitution in alpha 1(II) collagen chains of a patient with spondyloepiphyseal dysplasia. J. Biol. Chem. 266, 12487 – 12494 (1991). D. Chan, T. K. Taylor, and W. G. Cole, Characterization of an arginine 789 to cysteine substitution in alpha 1 (II) collagen chains of a patient with spondyloepiphyseal dysplasia. J. Biol. Chem. 268, 15238 – 15245 (1993). A. Winterpacht, M. Hilbert, U. Schwarze, S. Mundlos, J. Spranger, and U. Zabel, Kniest and Stickler dysplasia phenotypes caused by collagen type II gene (COL2A1) defect. Nature Genet. 3, 323 – 326 (1993). N. N. Ahmad, L. Ala-Kokko, R. G. Knowlton, S. A. Jimenez, E. J. Weaver, J. I. Maguire, W. Tasman, and D. J. Prockop, Stop codon in the procollagen II gene (COL2A1) in a family with the Stickler syndrome (arthro-ophthalmopathy). Proc. Natl. Acad. Sci. USA 88, 6624 – 6627 (1991). P. Vandenberg, J. S. Khillan, D. J. Prockop, H. Helminen, S. Kontusaari, and L. Ala-Kokko, Expression of a partially deleted gene of human type II procollagen (COL2A1) in transgenic mice produces a chondrodysplasia. Proc. Natl. Acad. Sci. USA 88, 7640 – 7644 (1991). M. Rintala, M. Metsaranta, S. Garofalo, B. de Crombrugghe, E. Vuorio, and O. Ronning, Abnormal craniofacial morphology and cartilage structure in transgenic mice harboring a Gly : Cys mutation in the cartilage-specific type II collagen gene. J. Craniofac. Genet. Dev. Biol. 13, 137 – 146 (1993). J. M. Pace, Y. Li, R. E. Seegmiller, C. Teuscher, B. A. Taylor, and B. R. Olsen, Disproportionate micromelia (Dmm) in mice caused by a mutation in the C-propeptide coding region of Col2a1. Dev. Dyn. 208, 25 – 33 (1997). Y. Ninomiya and B. R. Olsen, Collagens. In “Collagens” (T. Kreis and R. Vale, eds.), pp. 380 – 404. Oxford Univ. Press, Oxford, (1999). Y. Muragaki, E. C. Mariman, S. E. van Beersum, M. Perala, J. B. van Mourik, M. L. Warman, B. R. Olsen, and B. C. Hamel, A mutation in the gene encoding the alpha 2 chain of the fibril-associated collagen IX, COL9A2, causes multiple epiphyseal dysplasia (EDM2). Nature Genet. 12, 103 – 105 (1996). E. C. Spayde, A. P. Joshi, W. R. Wilcox, M. Briggs, D. H. Cohn, and B. R. Olsen, Exon skipping mutation in the COL9A2 gene in a family with multiple epiphyseal dysplasia. Matrix Biol. 19, 121 – 128 (2000). C. G. Bonnemann, G. F. Cox, F. Shapiro, J.-J. Wu, C. A. Feener, T. G. Thompson, D. C. Anthony, D. R. Eyre, B. T. Darras, and L. M. Kunkel, A mutation in the alpha 3 chain of type IX collagen causes
CHAPTER 5 Developmental Biology of Bone
253.
254.
255.
256.
257.
258.
259.
260.
261.
262.
263.
264.
265.
266.
autosomal dominant multiple epiphyseal dysplasia with mild myopathy. Med. Sci. 97, 1212 – 1217 (2000). J. Lohiniva, P. Paassilta, U. Seppanen, O. Vierimaa, S. Kivirikko, and L. Ala-Kokko, Splicing mutations in the COL3 domain of collagen IX cause multiple epiphyseal dysplaisa. Am. J. Med. Genet. 90, 216 – 222 (2000). P. Paassilta, J. Lohiniva, S. Annunen, J. Bonaventure, M. Le Merrer, L. Pai, and L. Ala-Kokko, COL9A3: A third locus for multiple epiphyseal dysplasia. Am. J. Hum. Genet. 64, 1036 – 1044 (1999). K. Nakata, K. Ono, J. Miyazaki, B. R. Olsen, Y. Muragaki, E. Adachi, K. Yamamura, and T. Kimura, Osteoarthritis associated with mild chondrodysplasia in transgenic mice expressing alpha 1(IX) collagen chains with a central deletion. Proc. Natl. Acad. Sci. USA 90, 2870 – 2874 (1993). R. Faessler, P. N. J. Schnegelsberg, J. Dausman, Y. Muragaki, T. Shinya, M. T. McCarthy, B. R. Olsen, and R. Jaenisch, Mice lacking a1(IX) collagen develop noninflammatory degenerative joint disease. Proc. Natl. Acad. Sci. USA 91, 5070 – 5074 (1994). C. J. Gress and O. Jacenko, Growth plate compressions and altered hematopoiesis in collagen X null mice. J. Cell Biol. 149, 983 – 993 (2000). D. Chan, Y. M. Weng, A. M. Hocking, S. Golub, D. J. McQuillan, and J. F. Bateman, Site-directed mutagenesis of human type X collagen: Expression of alpha1(X) NC1, NC2, and helical mutations in vitro and in transfected cells. J. Biol. Chem. 271, 13566 – 13572 (1996). D. Chan, Y. M. Weng, H. K. Graham, D. O. Sillence, and J. F. Bateman, A nonsense mutation in the carboxyl-terminal domain of type X collagen causes haploinsufficiency in schmid metaphyseal chondrodysplasia. J. Clin. Invest. 101, 1490 – 1499 (1998). D. Chan, S. Freddi, Y. M. Weng, and J. F. Bateman, Interaction of collagen alpha1(X) containing engineered NC1 mutations with normal alpha1(X) in vitro: Implications for the molecular basis of schmid metaphyseal chondrodysplasia. J. Biol. Chem. 274, 13091 – 13097 (1999). A. J. Richards, J. R. Yates, R. Williams, S. J. Payne, F. M. Pope, J. D. Scott, and M. P. Snead, A family with Stickler syndrome type 2 has a mutation in the COL11A1 gene resulting in the substitution of glycine 97 by valine in alpha 1(XI) collagen. Hum. Mol. Genet. 5, 1339 – 1343 (1996). A. J. Griffith, L. K. Sprunger, D. A. Sirko-Osadsa, G. E. Tiller, M. H. Meisler, and M. L. Warman, Marshall syndrome associated with a splicing defect at the COL11A1 locus. Am. J. Hum. Genet. 62, 816 – 823 (1998). S. Annunen, J. Körkkö, M. Czarny, M. L. Warman, H. G. Brunner, H. Kääriäinen, J. B. Mulliken, L. Tranebjaerg, D. G. Brooks, G. Cox, M. A. Curtis, S. Davenport, C. Friedrich, I. Kaitila, M. Krawczynski, A. Latos-Bielenska, S. Mukai, B. R. Olsen, N. Shinno, M. Somer, M. Vikkula, J. Zlotogora, D. J. Prockop, and L. Ala-Kokko, Splicing mutations of 54-bp exons in the COL11A1 gene cause Marshall syndrome, but other mutations cause overlapping Marshall/Stickler phenotypes. Am. J. Hum. Genet. 65, 974 – 983 (1999). M. Vikkula, E. C. M. Mariman, C. H. Lui, N. Zhidkova, G. E. Tiller, M. B. Goldbring, S. E. C. van Beersum, M. Malefijt, F. H. J. van den Hoogen, H.-H. Ropers, R. Mayne, K. Cheah, B. R. Olsen, M. L. Warman, and H. G. A. Brunner, Autosomal dominant and recessive osteochondrodysplasias associated with the COL11A2 locus. Cell 80, 431 – 437 (1995). R. Mayne, R. G. Brewton, P. M. Mayne, and J. R. Baker, Isolation and characterization of the chains of type V/type XI collagen present in bovine vitreous. J. Biol. Chem. 268, 9381 – 9386 (1993). Y. Li, D. A. Lacerda, M. L. Warman, D. R. Beier, H. Yoshioka, Y. Ninomiya, J. T. Oxford, N. P. Morris, K. Andrikopoulos, F. Ramirez, et al., A fibrillar collagen gene, Col11a1, is essential for skeletal morphogenesis. Cell 80, 423 – 430 (1995).
211 267. B. R. Olsen, New insights into the function of collagens from genetic analysis. Curr. Opin. Cell Biol. 7, 720 – 727 (1995). 268. H. Watanabe, Y. Yamada, and K. Kimata, Roles of aggrecan, a large chondroitin sulfate proteoglycan, in cartilage structure and function. J. Biochem. 124, 687 – 693 (1998). 269. H. Watanabe, K. Kimata, S. Line, D. Strong, L. Y. Gao, C. A. Kozak, and Y. Yamada, Mouse cartilage matrix deficiency (cmd) caused by a 7 bp deletion in the aggrecan gene. Nature Genet. 7, 154 – 157 (1994). 270. H. Watanabe, K. Nakata, K. Kimata, I. Nakanishi, and Y. Yamada, Dwarfism and age-associated spinal degeneration of heterozygote cmd mice defective in aggrecan. Proc. Natl. Acad. Sci. USA 94, 6943 – 6947 (1997). 271. H. Li, N. B. Schwartz, and B. M. Vertel, cDNA cloning of chick cartilage chondroitin sulfate (aggrecan) core protein and identification of a stop codon in the aggrecan gene associated with the chondrodystrophy, nanomelia. J. Biol. Chem. 268, 23504 – 23511 (1993). 272. H. Watanabe and Y. Yamada, Mice lacking link protein develop dwarfism and craniofacial abnormalities. Nature Genet. 21, 225 – 229 (1999). 273. T. Xu, P. Bianco, L. W. Fisher, G. Longenecker, E. Smith, S. Goldstein, J. Bonadio, A. Boskey, A. M. Heegaard, B. Sommer, K. Satomura, P. Dominguez, C. Zhao, A. B. Kulkarni, P. G. Robey, and M. F. Young, Targeted disruption of the biglycan gene leads to an osteoporosis-like phenotype in mice. Nature Genet. 20, 78 – 82 (1998). 274. M. D. Briggs, S. M. G. Hoffman, L. M. King, A. S. Olsen, H. Mohrenweiser, J. G. Leroy, G. R. Mortier, D. L. Rimoin, R. S. Lachman, E. S. Gaines, J. A. Cekleniak, R. G. Knowlton, and D. H. Cohn, Pseudoachondroplasia and multiple epiphyseal dysplasia due to mutations in the cartilage oligomeric matrix protein gene. Nature Genet. 10, 330 – 336 (1995). 275. J. Hecht, L. D. Nelson, E. Crowder, Y. Wang, F. F. B. Elder, W. R. Harrison, C. A. Francomano, C. K. Prange, G. G. Lennon, M. Deere, and J. Lawler, Mutations in exon 17B of cartilage oligomeric matrix protein (COMP) cause pseudoachondroplasia. Nature Genet. 10, 325 – 329 (1995). 276. R. Ballo, M. D. Briggs, D. H. Cohn, R. G. Knowlton, P. H. Beighton, and R. S. Ramesar, Multiple epiphyseal dysplasia, ribbing type: A novel point mutation in the COMP gene in a South African family [published erratum appears in Am. J. Med. Genet. 71, 494 (1997)]. Am. J. Med. Genet. 68, 396 – 400 (1997). 277. E. Delot, L. M. King, M. D. Briggs, W. R. Wilcox, and D. H. Cohn, Trinucleotide expansion mutations in the cartilage oligomeric matrix protein (COMP) gene. Hum. Mol. Genet. 8, 123 – 128 (1999). 278. R. V. Iozzo, Matrix proteoglycans: From molecular design to cellular function. Annu. Rev. Biochem. 67, 609 – 652 (1998). 279. B. R. Olsen, Life without perlecan has its problems. J. Cell Biol. 147, 909 – 911 (1999). 280. M. M. French, S. E. Smith, K. Akanbi, T. Sanford, J. Hecht, M. C. Farach-Carson, and D. D. Carson, Expression of the heparan sulfate proteoglycan, perlecan, during mouse embryogenesis and perlecan chondrogenic activity in vitro. J. Cell Biol. 145, 1103 – 1115 (1999). 281. M. Costell, E. Gustafsson, A. Aszodi, M. Morgelin, W. Bloch, E. Hunziker, K. Addicks, R. Timpl, and R. Fassler, Perlecan maintains the integrity of cartilage and some basement membranes. J. Cell Biol. 147, 1109 – 1122 (1999). 282. A. Aszodi, D. Chan, E. Hunziker, J. F. Bateman, and R. Fassler, Collagen II is essential for the removal of the notochord and the formation of intervertebral discs. J. Cell Biol. 143, 1399 – 1412 (1998). 283. I. The, Y. Bellaiche, and N. Perrimon, Hedgehog movement is regulated through tout velu-dependent synthesis of a heparan sulfate proteoglycan. Mol. Cell 4, 633 – 639 (1999). 284. F. Deak, D. Piecha, C. Bachrati, M. Paulsson, and I. Kiss, Primary structure and expression of matrilin-2, the closest relative of cartilage
212 matrix protein within the von Willebrand factor type A-like module superfamily. J. Biol. Chem. 272, 9268 – 9274 (1997). 285. F. Deak, R. Wagener, I. Kiss, and M. Paulsson, The matrilins: A novel family of oligomeric extracellular matrix proteins. Matrix Biol. 18, 55 – 64 (1999). 286. N. Hauser, M. Paulsson, D. Heinegard, and M. Morgelin, Interaction of cartilage matrix protein with aggrecan: Increased covalent cross-linking with tissue maturation. J. Biol. Chem. 271, 32247 – 32252 (1996).
REGINATO, WANG, AND OLSEN 287. N. Winterbottom, M. M. Tondravi, T. L. Harrington, F. G. Klier, B. M. Vertel, and P. F. Goetinck, Cartilage matrix protein is a component of the collagen fibril of cartilage. Dev. Dyn. 193, 266 – 276 (1992). 288. A. Aszodi, J. F. Bateman, E. Hirsch, M. Baranyi, E. B. Hunziker, N. Hauser, Z. Bosze, and R. Fassler, Normal skeletal development of mice lacking matrilin 1: Redundant function of matrilins in cartilage? Mol. Cell Biol. 19, 7841 – 7845 (1999).
CHAPTER 6
Mouse Genetics as a Tool to Study Bone Development and Physiology MILLAN S. PATEL AND GERARD KARSENTY Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, Texas 77030
I. II. III. IV.
V. Independence of Bone Resorption from Bone Formation during Bone Remodeling VI. Bone Formation Is Centrally Regulated in Vivo References
Introduction The Osteoprotegerin Pathway Vitamin D Receptor Cbfa1
I. INTRODUCTION
peutic perspective. Moreover, the ability to delete several contiguous genes or an entire cell population expands considerably the scope of the information that can be obtained and analyzed. The second reason stems from the issues that are at stake in bone biology. As with every organ, little is known about the mechanisms controlling skeletal formation during development. Mouse genetics has shown that it is, along with chick embryology, the most powerful and elegant approach to address these questions for any organ, including the skeleton. However, what is specific to bone and only very few other organs is that we still know little about the molecular mechanisms controlling the function of its constituent cells: the osteoblasts and, to a lesser degree, the osteoclasts. That such questions of physiology and pathophysiology can be addressed successfully by mouse genetics has become increasingly evident over time. Although there are physiologic differences between mice and humans, the similarities for every organ far outnumber these differences. This has now been amply demonstrated for most organ physiology
Since the early 1990s, mouse genetics, defined here as a group of techniques aimed at altering gene expression and/or function in vivo, has invaded every field of biology, including bone biology. In doing so it has revolutionized thinking in each field it has entered. There are several reasons that fully justify such prominence, however, and this point deserves emphasis: mouse genetics is like every other field of biology, by itself it does not have all the answers to the questions confronting bone biology. One reason to explain the popularity of mouse genetics comes from the versatility of the technology. Indeed, it is now possible to delete a gene, decrease or increase its expression, or express it ectopically. The fact that these different manipulations can be done in vivo, during development or after birth, and in the cell type of choice, allows two equally important questions regarding any gene product to be addressed: what it normally does and what it can do. This latter aspect may be of greater interest from a thera-
OSTEOPOROSIS, SECOND EDITION VOLUME 1
213
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
214 including bone. The demonstration is so eloquent that the argument that mouse findings cannot be used to understand human physiology has lost most of its credibility. A final reason explaining the importance of mouse genetics as a tool comes from its track record. The functions of most cloned genes have now been studied in mice. In a few instances these studies have generated such clear-cut and unexpected results that they have substantially altered the way we think. There is one of the best examples of the revolutionary power of mouse genetics. The c-src gene codes for a tyrosine kinase membrane protein and is expressed ubiquitously. Surprisingly, c-src-deficient mice have only one phenotype: osteopetrosis due primarily to a functional defect of the osteoclast [1]. Further domain function analyses of c-src in vivo showed that this phenotype was tyrosine kinase domain independent. The function of c-src and the relative importance of each of its domains could not have been obtained without in vivo studies. This example best illustrates the power of mouse genetics in skeletal biology. Recent developments that illustrate the fundamental changes in conception that mouse genetics has brought to bone biology include the discovery of osteoprotegerin (OPG) and its ligand (RANKL, receptor activator of NFB ligand/ODF, osteoclast differentiation factor), the phenotype of vitamin D receptor (VDR)-deficient mice, the dual functions of Cbfa1 during development and postnatally, the independence of bone resorption from bone formation during bone remodeling, and the regulation of osteoblast function by the hypothalamus.
II. THE OSTEOPROTEGERIN PATHWAY Osteoclast biology has been profoundly revolutionized by the identification of a group of secreted molecules that positively or negatively regulate osteoclast differentiation and function. By screening an intestinal cDNA library for novel secreted molecules, a group at Amgen, Inc. identified a new member of the TNF receptor superfamily [2]. Named osteoprotegerin, this molecule contains no hydrophobic transmembrane-spanning sequence, suggesting that it is a soluble receptor. This molecule is identical to the osteoclastogenesis inhibitory factor (OCIF) purified and subsequently cloned by a group at Snow Brand Milk Products Co., Ltd. using a biochemical approach [3]. A daily intraperitoneal injection of recombinant OPG/OCIF in large amounts, as well as a high level of overexpression in transgenic mice, resulted in osteopetrosis due to arrested osteoclast differentiation. The identification and functional study of OPG/OCIF was historically of critical importance as it demonstrated that, in addition to steroid hormones and known peptide hormones, there are novel secreted molecules able to control osteoclast differentiation. The
PATEL AND KARSENTY
specificity of OPG/OCIF function for inhibiting osteoclast differentiation was further illustrated by the phenotype of OPG/OCIF-deficient mice. These mice develop severe osteoporosis due to increased numbers of osteoclasts [4,5]. The identification of a soluble receptor with such a powerful inhibitory effect on osteoclast differentiation suggested that it may act by binding a factor with osteoclast differentiation activity. This factor was cloned by both groups at nearly the same time. The Amgen group used recombinant OPG/OCIF to screen for OPG/OCIF ligand on the surface of various cell lines in a ligand-panning assay [6]. For this screening they rightly did not limit their search solely to novel molecules. The protein they isolated and called osteoprotegerin ligand (OPGL) had in fact been cloned previously and called TRANCE or RANK ligand (RANKL). This illustrates the limits of large genomic screens, as selection of only novel genes would have missed RANKL. At the same time, Yasuda et al. [7] from Snow Brand purified to homogeneity the osteoclast differentiation factor (ODF) and showed that it was the RANK ligand. For the sake of clarity it will be referred to here as RANKL or ODF, although this nomenclature controversy has not been settled. In vitro, RANKL/ODF has all the attributes of a genuine osteoclast differentiation factor: it favors osteoclast differentiation; in conjunction with M-CSF (macrophage colony stimulating factor), it bypasses the need for stromal cells and 1,25(OH)2 vitamin D3 to induce osteoclast differentiation; and it activates mature osteoclasts to resorb mineralized bone [8]. RANKL/ODF exists either in bound form on the membrane of osteoblast progenitors and other cells or as a soluble molecule in the bone microenvironment and in blood. Systemic administration of RANKL/ODF leads to increased bone resorption in vivo, and deletion of the RANKL gene leads to mice that lack osteoclasts and develop severe osteopetrosis in addition to immunological defects [9]. Two observations are of potential interest. First, because RANKL/ODF is secreted by osteoblasts and other cells, the secreted form could conceivably be the active form while the membrane-bound RANKL/ODF could be acting as a reservoir. Second, RANKL/ODF is also expressed by T cells, cells that can in vitro induce osteoclastogenesis [10]. The production and secretion of RANKL/ODF systemically or locally by T cells may explain some of the bone abnormalities observed in inflammatory disorders [9]. As mentioned previously, ODF is also called RANK ligand because it binds to a receptor, present on T cells and bone marrow stromal cells, called RANK [11]. Transgenic mice expressing a soluble form of RANK develop an osteopetrosis similar to that observed in RANKL/ODFdeficient mice. A polyclonal antibody against the RANK extracellular domain promotes osteoclastogenesis in bone marrow cultures, suggesting that RANK activation mediates the effect of RANKL/ODF [12]. The signal transduction
215
CHAPTER 6 Mouse Genetics as a Tool
pathway initiated following binding of RANKL/ODF to RANK has also been partly elucidated, thus establishing a cascade from extracellular signals to nuclear effectors. The intracellular domain of RANK contains two binding sites for members of a family of proteins called TNF receptor-associated factors (TRAFs) [13]. TRAFs have been implicated in mediating signals induced by a subset of TNF receptor family members. RANK contains a binding site for TRAF6 in its intracellular domain [14]. Importantly, TRAF6-deficient mice exhibit an osteopetrotic phenotype due to defective osteoclast function, thus providing the beginning of a signal transduction cascade leading to osteoclast terminal differentiation [15]. This observation is even more important as TRAFs appear to enhance c-src function [16] and to control the activation of NF-B, a transcription factor required for osteoclast differentiation [17,18]. Given the information available it is now possible to illustrate this signal transduction pathway along with those transcription factors known to affect osteoclast differentiation (see Fig. 1). What has been learned from the OPG-RANK-RANKL pathway could not have been performed as quickly and as thoroughly without using a genetic approach.
III. VITAMIN D RECEPTOR Vitamin D is a pleiotropic hormone whose active form is thought to play key roles in intestinal calcium absorption, renal calcium and phosphate conservation, and osteoclastic bone resorption. Vitamin D was hypothesized to induce monocytic stem cells in the bone marrow to differentiate into osteoclasts and to indirectly regulate osteoclastic activity through its effects on osteoblasts [19]. However, when mice were generated that lacked the vitamin D receptor (VDR), normal numbers of osteoclasts were seen [20]. Because VDR-deficient mice had normal bone at 3 weeks of age but showed a 40% reduction by 7 weeks, this further suggests that osteoclast function was not severely impaired. Most importantly, the florid rickets demonstrated by these mice was rescued by calcium supplementation, as had been shown more than a decade earlier for a human patient with vitamin D-resistant rickets [21]. In fact, VDR-deficient mice confirmed that in the entire animal, even though VDR appears to have a minor direct role in bone cell biology, its major role is indirect through effects on calcium regulation.
IV. Cbfa1
FIGURE 1
(A) Genes involved in osteoclast differentiation and maturation in vivo. mi, microphthalmia. (B) Signal transduction pathway leading to a functional osteoclast.
Transcriptional hierarchies are a well-established mechanism for cell-type specification. Extensive analysis of the promoter of the most osteoblast-specific gene, osteocalcin, was used to identify the first osteoblast-specific cis-acting elements (OSEs), OSE1 and OSE2. OSE2 is required for high-level gene expression and binds to an osteoblastspecific nuclear factor [22]. A biochemical and molecular approach revealed that this factor was identified as corebinding factor A1 (Cbfa1). Cbfa proteins are mammalian homologues of the Drosophila runt protein [23]. These proteins, along with their Drosophila and Caenorhabditis elegans counterparts, form a new family of transcription factors whose hallmark is a highly conserved 128 amino acid DNA-binding domain called the runt domain [24]. To date, three distinct Cbfa proteins encoded by different genes have been identified [24]. Several lines of evidence support the role of Cbfa1 as a transcriptional activator of osteoblast differentiation. Its pattern of expression during early development is confined to the mesenchymal condensations, which model the future skeleton, and its expression is detectable as early as 10.5 days postcoitum (dpc) [25]. At this early stage of skeletogenesis, Cbfa1 expression identifies a common progenitor for the chondrocytic and osteoblastic lineages. Beginning at 14.5 dpc and thereafter throughout life, Cbfa1 expression is restricted to cells of the osteoblastic lineage and, apart from decreasing expression in prehypertrophic and hypertrophic chondrocytes as chondrogenesis proceeds is absent from any other cell type [25 – 29]. To date, Cbfa1 remains the most specific and
216 earliest marker of osteogenesis known. Moreover, its function is consistent with its osteoblast-specific expression. Cbfa1 binds to the promoters of genes expressed predominatly in osteoblasts, such as (I) collagen, bone sialoprotein, osteopontin, and osteocalcin, as well as being able to positively regulate their expression in tissue culture and in vivo. Further experiments showed that ectopic expression of Cbfa1 in fibroblast cell lines or in primary skin fibroblasts leads to the acquisition of an osteoblastic genotype by these cells, suggesting it may be sufficient for osteoblastogenesis [26]. However, a more convincing demonstration that Cbfa1 was necessary for osteoblast formation came from mouse genetics. Cbfa1-deficient mice die immediately after birth and skeletal studies showed a complete absence of osteogenesis with a consequent lack of both endochondral and intramembranous bone formation. Cartilage formation was grossly unaffected in these mutant mice [25,27]. Detailed histological analysis of the skeleton of these mice revealed an arrest of osteoblast differentiation before 14.5 dpc and a lack of expression of the osteoblast proteins, osteopontin and osteocalcin. Taken together, these data demonstrate that the Cbfa1 gene is essential for the differentiation of osteoblasts and thus for bone formation during development of the skeleton. Analysis of mice heterozygous for the Cbfa1 deletion demonstrated specific skeletal defects that were confined to those bones that form directly from mesenchymal precursors by intramembranous ossification. The most obvious of these defects, clavicular hypoplasia and delayed cranial bone ossification resulting in widely patent fontanelles, strongly resembled the human autosomal dominant disease cleidocranial dysplasia (CCD). A mouse mutation with similar features (also named Ccd) mapped to the same location as Cbfa1, and the Cbfa1 gene was found to be partially deleted in these mice [25]. Analysis of human patients with CCD also showed mutations of Cbfa1, some in the highly conserved runt domain, demonstrating that haploinsufficiency for this gene has functional consequences [30,31]. The similarity of the mouse and human phenotypes is important as it exemplifies how useful the mouse is to uncover the pathways controlling skeletal development in humans. As seen later the same holds true when it comes to skeletal physiology. The finding that Cbfa1 is expressed postnatally at high levels in osteoblasts, and the delay in skeletal maturation shown by CCD patients, led Ducy et al. [32] to examine, through mouse genetics, the consequences of selective postnatal inhibition of Cbfa1 function in osteoblasts using a dominant-negative form of the protein. This dominantnegative form consisted of the DNA-binding domain of Cbfa1 ( Cbfa1), which had a greater binding affinity for DNA than the native protein but lacked transactivation capabilities. The osteocalcin promoter was used to direct expression of this dominant-negative protein exclusively to differentiated osteoblasts. Because osteocalcin is virtually not expressed during development, this dominant-negative form of Cbfa1 was also not expressed during development.
PATEL AND KARSENTY
As expected, mice overexpressing Cbfa1 had no features of Ccd at birth and were indistinguishable from their wildtype littermates. Over time, however, the transgenic mice developed marked osteopenia despite normal numbers of osteoblasts. The osteopenia affected both cortical and trabecular bone, and a 70% decrease in the bone formation rate was documented, suggesting that functional inhibition of Cbfa1 postnatally markedly inhibits differentiated osteoblast function. This second function of Cbfa1, the regulation of osteoblast physiology, was accompanied by a profound decrease in transcription of the collagen genes and that of all noncollagenous protein genes studied in the transgenic mice. Endogenous Cbfa1 expression was also decreased in the transgenic mice. Investigation of this phenomenon [32] revealed that Cbfa1 is a potent stimulator of its own expression and thus identified an autoregulatory loop at the top of a gene cascade controlling postnatal osteoblast function. These results illustrate how developmentally important genes can also play later roles in physiology. As mentioned above, Cbfa1 is also expressed in prehypertrophic chondrocytes and, to a lesser extent, in hypertrophic chondrocytes at early stages of chondrogenesis. Consistent with this pattern of expression, some, but not all, skeletal elements in Cbfa1-deficient mice lack hypertrophic chondrocytes [25,27,28,29]. To address the role of Cbfa1 during chondrogenesis transgenic mice were generated in which Cbfa1 was expressed in nonhypertrophic chondrocytes [30]. Continuous expression of Cbfa1 in nonhypertrophic chondrocytes induced chondrocyte hypertrophy and endochondral ossification in locations where it normally never occurs. To determine whether this was due to transdifferentiation of chondrocytes into osteoblasts or to a specific hypertrophic chondrocyte differentiation ability of Cbfa1, the transgene was introduced into Cbfa1-deficient mice. The transgene restored chondrocyte hypertrophy and vascular invasion but did not induce osteoblast differentiation. The rescue was cell-autonomous as skeletal elements not expressing the transgene were unaffected. Despite the lack of osteoblasts in the rescued mice, there were multinucleated, TRAP-positive cells reabsorbing the hypertrophic cartilage matrix. These results identify Cbfa1 as a hypertrophic chondrocyte differentiation factor and provide a genetic argument for a common regulation of osteoblast and chondrocyte differentiation mediated by Cbfa1.
V. INDEPENDENCE OF BONE RESORPTION FROM BONE FORMATION DURING BONE REMODELING Coupling between the functions of bone resorption and bone formation was originally proposed as the mechanism by which bone mass is maintained constant throughout the
CHAPTER 6 Mouse Genetics as a Tool
reproductive years [33]. Uncoupling of necessity occurs during longitudinal bone growth, fracture repair, and secondary ossification. These processes are collectively referred to as bone modeling to distinguish them from the proposed coupled state, referred to as bone remodeling. A key prediction of the coupling hypothesis is that absence or functional deficit of one cell type should greatly inhibit the function of the other cell type. In several mouse models with genetic defects of osteoclasts, however, osteopetrosis ensues, suggesting that osteoblast activity is not constrained by a lack of osteoclast function. Mouse genetics was used by Corral et al. [34] to construct an in vivo model of inducible osteoblast ablation in order to test the reverse: whether osteoclast activity would be inhibited in the complete absence of bone formation. Transgenic mice harboring the herpes simplex virus thymidine kinase gene under control of the osteocalcin promoter were generated and, upon administration of ganciclovir, longitudinal growth as well as bone formation activity ceased. Despite this, bone was progressively lost over time. This loss was due to osteoclast activity because it could be inhibited by the classical antiresorptive agent alendronate. These results showed that, in functional terms, bone resorption is independent of bone formation in vivo. RANKL/ODF expressed by osteoblast progenitors likely plays a role in cross-regulation; however, the transgenic results described earlier indicate that this occurs at the level of osteoblast differentiation, not at the level of osteoblast function. Another mouse genetic experiment demonstrates that the converse observation is also true. Mice deficient in 3 integrin that have a defect in bone resorption develop osteosclerosis, as bone formation is not affected [35].
VI. BONE FORMATION IS CENTRALLY REGULATED IN VIVO Treatment of osteoporosis requires modulation of bone cell function per se, and to this end an unexpected but critically important finding in osteocalcin-thymidine kinase transgenic mice was that after withdrawal of ganciclovir, there was a complete reversal of the morphologic, histologic, and histomorphometric findings within 4 weeks. The striking precision of this normalization process in such a short time suggests that osteoblasts had two speeds to deposit bone matrix in this model. Upon withdrawal of ganciclovir they could quickly deposit a huge amount of bone matrix to fill up the empty bone. Once that was achieved they could “slow down” their rate of extracellular matrix production and reenter the normal cycle of resorption/formation to maintain bone mass. This versatility was then interpreted as an indication that bone formation was an endocrine function [34]. This led to the discovery of leptin as the most potent inhibitor of bone formation identified to date.
217 Two clinical observations have been made repeatedly over the past several decades regarding the pathophysiology of osteoporosis. The first is that bone loss invariably follows the cessation of gonadal function and the second is that obesity protects from osteoporosis. In endocrine terms, these observations suggest that bone mass, fat mass, and gonadal function may be regulated by common molecules. This hypothesis, involving whole organism physiology, could only be tested efficiently using mouse genetics. Given its known functions, leptin was a good candidate molecule with which to test the hypothesis. Leptin-deficient (ob/ob) mice show striking obesity and hypogonadism. These phenotypes are recessive. The latter feature should significantly predispose to low bone mass; however, these mice have an increased bone formation rate that leads to a high bone mass phenotype [36]. This high bone mass phenotype is also seen in leptin receptor-deficient mice and in leptin receptor-deficient fa/fa rats (M. Amling, personal communication). Importantly, the high bone mass phenotype precedes the appearance of obesity in ob/ob mice. In contrast, fat-free mice [37], which have very low leptin levels, do have high bone mass, demonstrating that fat is not required as an intermediary. Expression of neither leptin nor its receptor could be detected in primary osteoblasts or in bone, and leptin signaling could not be detected in cultured, primary osteoblasts, suggesting that its mechanism of action is not autocrine, paracrine, or endocrine. In contrast, intracerebroventricular (icv) infusion of leptin resulted in correction of the high bone mass phenotype in ob/ob mice as well as bone loss in wild-type mice. These effects were shown to be specifically due to a decrease in bone formation with normal numbers of osteoblasts and no alteration of osteoclast function, indicating that leptin is a specific inhibitor of osteoblast function. Because the effects of leptin on body weight are mediated by binding of the hormone to receptors in the hypothalamus and icv infusion of leptin into the third ventricle inhibits bone formation, it is likely that the hypothalamus is the major site regulating osteoblast function and thereby bone mass homeostasis (Fig. 2). This central regulation of bone remodeling is the only form of regulation yet identified that can overcome the deleterious effect of hypogonadism on bone mass. This point best underscores its physiologic relevance.
FIGURE 2 Central regulation of body weight, osteoblast function, and gonadal function by leptin and possibly other secreted molecules.
218
PATEL AND KARSENTY
How does this study relate to human physiology and the initial hypothesis that bone mass, fat mass, and gonadal function share common regulation? Both humans and mice with diet-induced obesity show increased leptin levels, proportional to their fat mass. The lack of functional inhibition by leptin in these situations has been termed “leptin resistance”; thus, the protective nature of obesity for osteoporosis can be thought of as a state of resistance to the osteopenic effects of leptin. The findings described earlier, which relied entirely on mouse and rat genetics, demonstrate unambiguously that bone formation is a centrally regulated function. The functional importance of this pathway is reinforced by the absence of low bone mass in human patients with leptin deficiency, despite their hypogonadism. This in turn exemplifies the usefulness of the mouse as a tool to study human physiology. This new view of bone remodeling, when reinforced by the central action of other hormones affecting bone metabolism, may have profound implications for our understanding and management of osteoporosis.
References 1. P. Soriano, C. Montgomery, R. Geske, and A. Bradley, Targeted disruption of the c-src proto-oncogene leads to osteopetrosis in mice. Cell 64, 693 – 702 (1991). 2. W. S. Simonet, D. L. Lacey, C. R. Dunstan, M. Kelley, M. S. Chang, R. Luthy, H. Q. Nguyen, S. Wooden, L. Bennett, T. Boone, G. Shimamoto, M. DeRose, R. Elliott, A. Colombero, H. L. Tan, G. Trail, J. Sullivan, E. Davy, N. Bucay, G. L. Renshaw, T. M. Hughes, D. Hill, W. Pattison, P. Campbell, and W. J. Boyle, Osteoprotegerin: A novel secreted protein involved in the regulation of bone density. Cell 86, 309 – 319 (1997). 3. E. Tsuda, M. Goto, S. Mochizuki, K. Yano, F. Kobayashi, T. Morinaga, and K. Higashio, Isolation of a novel cytokine from human fibroblasts that specifically inhibits osteoclastogenesis. Biochem. Biophys. Res. Commun. 234, 137 – 142 (1997). 4. N. Bucay, I. Sarosi, C. R. Dunstan, S. Morony, J. Tarpley, C. Capparelli, S. Scully, H. L. Tan, W. Xu, D. L. Lacey, W. J. Boyle, and W. S. Simonet, Osteoprotegerin-deficient mice develop early onset osteoporosis and arterial calcification. Genes Dev. 12, 1260 – 1268 (1998). 5. A. Mizuno, N. Amizuka, K. Irie, A. Murakami, N. Fujise, T. Kanno, Y. Sato, N. Nakagawa, H. Yasuda, S. Mochizuki, T. Gomibuchi, K. Yano, N. Shima, N. Washida, E. Tsuda, T. Morinaga, K. Higashio, and H. Ozawa, Severe osteoporosis in mice lacking osteoclastogenesis inhibitory factor/osteoprotegerin. Biochem. Biophys. Res. Commun. 247, 610 – 615 (1998). 6. D. L. Lacey, E. Timms, H.-L. Tan, M. J. Kelley, C. R. Dunstan, T. Burgess, R. Elliot, A. Colombero, G. Elliot, S. Scully, H. Hsu, J. Sullivan, N. Hawkins, E. Davy, C. Capparelli, A. Eli, Y.-X. Qian, S. Kaufman, I. Sarosi, V. Shalhoub, G. Senaldi, J. Guo, J. Delaney, and W. J. Boyle, Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell 93, 165 – 176 (1998). 7. H. Yasuda, N. Shima, N. Nakagawa, K. Yamaguchi, M. Kinosaki, S. Mochizuki, A. Tomoyasu, K. Yano, M. Goto, A. Murakami, E. Tsuda, T. Morinaga, K. Higashio, N. Udagawa, N. Takahashi, and T. Suda, Osteoclast differentiation factor is a ligand for osteoprotegerin/ osteoclastogenesis-inhibitory factor and is identical to TRANCE/ RANKL. Proc. Natl. Acad. Sci. USA. 95, 3597 – 3602 (1998).
8. T. L. Burgess, Y. Qian, S. Kaufman, B. D. Ring, G. Van, C. Capparelli, M. Kelley, H. Hsu, W. J. Boyle, C. R. Dunstan, S. Hu, and D. L. Lacey, The ligand for osteoprotegerin (OPGL) directly activates mature osteoclasts. J. Cell. Biol. 145, 527 – 538 (1999). 9. Y. Y. Kong, U. Feige, I. Sarosi, B. Bolon, A. Tafuri, S. Morony, C. Capparelli, J. Li, R. Elliott, S. McCabe, T. Wong, G. Campagnuolo, E. Moran, E. R. Bogoch, G. Van, L. T. Nguyen, P. S. Ohashi, D. L. Lacey, E. Fish, W. J. Boyle, and J. M. Penninger, Activated T cells regulate bone loss and joint destruction in adjuvant arthritis through osteoprotegerin ligand. Nature 402, 304 – 309 (1999). 10. N. J. Horwood, N. Udagawa, J. Elliott, D. Grail, H. Okamura, M. Kurimoto, A. R. Dunn, T. Martin, and M. T. Gillespie, Interleukin 18 inhibits osteoclast formation via T cell production of granulocyte macrophage colony-stimulating factor. J. Clin. Invest. 101, 595 – 603 (1998). 11. D. M. Anderson, E. Maraskovsky, W. L. Billingsley, W. C. Dougall, M. E. Tometsko, E. R. Roux, M. C. Teepe, R. F. DuBose, D. Cosman, and L. Galibert, A homologue of the TNF receptor and its ligend enhance T-cell growth and dendritic-cell function. Nature 390, 175 – 179 (1997). 12. H. Hsu, D. L. Lacey, C. R. Dunstan, I. Solovyev, A. Colombero, E. Timms, H. L. Tan, G. Elliott, M. J. Kelley, I. Sarosi, L. Wang, X. Z. Xia, R. Elliott, L. Chiu, T. Black, S. Scully, C. Capparelli, S. Morony, G. Shimamoto, M. B. Bass, and W. J. Boyle, Tumor necrosis factor receptor family member RANK mediates osteoclast differentiation and activation induced by osteoprotegerin ligand. Proc. Natl. Acad. Sci. USA 96, 3540 – 3545 (1999). 13. B. R. Wong, R. Josien, S. Y. Lee, M. Vologodskaia, R. M. Steinman, and Y. Choi, The TRAF family of signal transducers mediates NFkappaB activation by the TRANCE receptor. J. Biol. Chem. 273, 28355 – 28359 (1998). 14. B. G. Darnay, J. Ni, P. A. Moore, and B. B. Aggarwal, Activation of NF-kappaB by RANK requires tumor necrosis factor receptor-associated factor (TRAF) 6 and NF-kappaB-inducing kinase: Identification of a novel TRAF6 interaction motif. J. Biol. Chem. 274, 7724 – 7731 (1999). 15. M. A. Lomaga, W. C. Yeh, I. Sarosi, G. S. Duncan, C. Furlonger, A. Ho, S. Morony, C. Capparelli, G. Van, S. Kaufman, A. Van der Heiden, A. Itie, A. Wakeham, W. Khoo, T. Sasaki, Z. Cao, J. M. Penninger, C. J. Paige, D. L. Lacey, C. R. Dunstan, W. J. Boyle, D. V. Goeddel, and T. W. Mak, TRAF6 deficiency results in osteopetrosis and defective interleukin-1, CD40, and LPS signaling. Genes Dev. 13, 1015 – 1024 (1999). 16. B. R. Wong, D. Besser, N. Kim, J. R. Arron, M. Vologodskaia, H. Hanafusa, and Y. Choi, TRANCE, a TNF family member, activates Akt/PKB through a signaling complex involving TRAF6 and c-Src. Mol. Cell. 4, 1041 – 1049 (2000). 17. G. Franzoso, L. Carlson, L. Xing, L. Poljak, E. W. Shores, K. D. Brown, A. Leonardi, T. Tran, B. F. Boyce, and U. Siebenlist, Requirement for NF-kappaB in osteoclast and B-cell development. Genes Dev. 11, 3482 – 3496 (1997). 18. V. Iotsova, J. Caamano, J. Loy, Y. Yang, A. Lewin, and R. Bravo, Osteopetrosis in mice lacking NF-kappaB1 and NF-kappaB2. Nature Med. 3, 1285 – 1289 (1997). 19. M. F. Holick, Vitamin D: Photobiology, metabolism, mechanism of action, and clinical applications. In “Primer on the Metabolic Bone Diseases and Disorders of Mineral Metabolism” (M. J. Favus et al. eds.), 4th Ed. pp. 92 – 98. Lippincott, Williams & Wilkins, Philadelphia, 1999. 20. T. Yoshizawa, Y. Handa, Y. Uematsu, S. Takeda, K. Sekine, Y. Yoshihara, T. Kawakami, K. Arioka, H. Sato, Y. Uchiyama, S. Masushige, A. Fukamizu, T. Matsumoto, and S. Kato, Mice lacking the vitamin D receptor exhibit impaired bone formation, uterine hypoplasia and growth retardation after weaning. Nature. Genet. 16, 391 – 396 (1997). 21. S. Balsan, M. Garabedian, M. Larchet, A. M. Gorski, G. Cournot, C. Tau, A. Bourdeau, C. Silve, and C. Ricour, Long-term nocturnal calcium infusions can cure rickets and promote normal mineralization in
CHAPTER 6 Mouse Genetics as a Tool
22.
23.
24.
25.
26.
27.
28.
29. 30.
hereditary resistance to 1,25-dihydroxyvitamin D. J. Clin. Invest. 77, 1661 – 1667 (1986). P. Ducy and G. Karsenty, Two distinct osteoblast-specific cis-acting elements control expression of a mouse osteocalcin gene. Mol. Cell. Biol. 15, 1858 – 1869 (1995). J. B. Duffy, M. A. Kania, and J. P. Gergen, Expression and function of the Drosophila gene runt in early stages of neural development. Development 113, 1223 – 1230 (1991). J. J. Westendorf and S. W. Hiebert, Mammalian runt-domain proteins and their roles in hematopoiesis, osteogenesis, and leukemia. J. Cell. Biochem. Suppl. 32 – 33, 51 – 58 (1999). F. Otto, A. P. thornell, T. Crompton, A. Denzel, K. C. Gilmour, I. R. Rosewell, G. W. H. Stamp, R. S. P. Beddington, S. Mundlos, B. R. Olsen, P. B. Selby, and M. J. Owen, Cbfa1, a candidate gene for cleidocranial dysplasia syndrome, is essential for osteoblast differentiation and bone development. Cell 89, 765 – 771 (1997). P. Ducy, R. Zhang, V. Geoffroy, A. Ridall, and G. Karsenty, Osf2/Cbf1: A transcriptional activator of osteoblast differentiation. Cell 89, 747 – 754 (1997). T. Komori, H. Yahi, S. Nomura, A. Yamaguchi, K. Sasaki, K. Deguchi, Y. Shimizu, R. T. Bronson, Y. H. Gao, M. Inada, M. Sato, R. Okamoto, Kitamura, S. Yoskiki, and T. Kishimoto, Targeted disruption of Cbfa1 results in a complete lack of bone formation owing to maturational arrest of osteoblasts. Cell 89, 755 – 764 (1997). M. Inada, T. Yasui, S. Nomura, S. Miyake, K. Deguchi, M. Himeno, M. Sato, H. Yamagiwa, T. Kimura, N. Yasui, T. Ochi, N. Endo, Y. Kitamura, T. Kishimoto, and T. Komori, Maturational disturbance of chondrocytes in Cbfa1-deficient mice. Dev. Dyn. 214, 279 – 290 (1999). I. S. Kim, F. Otto, B. Zabel, and S. Mundlos, Regulation of chondrocyte differentiation by Cbfa1. Mech. Dev. 80, 159 – 170 (1999). Deleted at proof.
219 31. B. Lee, K. Thirunabukkarasu, L. Zhou, L. Pastore, A. Baldini, J. Hecht, V. Geoffroy, P. Ducy, and G. Karsenty, Missense mutations abolishing DNA binding of the osteoblast-specific transcription factor OSF2/CBFA1 in cleidocranial dysplasia. Nature. Genet. 16, 307 – 310 (1997). 32. S. Mundlos, F. Otto, C. Mundlos, J. B. Mulliken, A. S. Aylsworth, S. Albright, D. Lindhout, W. G. Cole, W. Henn, J. H. M. Knoll, M. J. Owen, R. Mertelsmann, B. U. Zabel, and B. R. Olsen, Mutations involving the transcription factor CBFA1 cause cleidocranial dysplasia. Cell 89, 773 – 779 (1997). 33. P. Ducy, M. Starbuck, M. Priemel, J. Shen, G. Pinero, V. Geoffroy, M. Amling, and G. Karsenty, A Cbfa1-dependent genetic pathway controls bone formation beyond embryonic development. Genes Dev. 13, 1025 – 1036 (1999). 34. H. M. Frost, “Bone Biodynamics,” p. 315. Little, Brown, Boston, 1964. 35. D. Corral, M. Amling, M. Prienel, E. Loyer, S. Fuchs, P. Ducy, R. Baron, and G. Karsenty, Dissociation between bone resorption and bone formation in osteopenic transgenic mice. Proc. Natl. Acad. Sci. USA 95, 13835 – 13840 (1998). 36. K. P. McHugh, K. Hodivala-Dilke, M. H. Zheng, N. Namba, J. Lam, D. Novack, X. Feng, F. P. Ross, R. O. Hynes, and S. L. Teitelbaum, Mice lacking beta3 integrins are osteosclerotic because of dysfunctional osteoclasts. J. Clin. Invest. 105, 433 – 440 (2000). 37. P. Ducy, M. Amling, S. Takeda, M. Priemel, A. F. Schilling, F. T. Beil, J. Shen, C. Vinson, J. M. Rueger, and G. Karsenty, Leptin inhibits bone formation through a hypothalamic relay: A central control of bone mass. Cell 100, 197 – 207 (2000). 38. J. Moitra, M. M. Mason, M. Olive, D. Krylov, O. Gavrilova, B. Marcus-Samuels, L. Feigenbaum, E. Lee, T. Aoyama, M. Eckhaus, M. L. Reitman, and C. Vinson, Life without white fat: A transgenic mouse. Genes Dev. 12, 3168 – 3181.
CHAPTER 7
Parathyroid Hormone and Parathyroid HormoneRelated Protein ROBERT A. NISSENSON
Endocrine Unit, San Francisco Veterans Affairs Medical Center, and Departments of Medicine and Physiology, University of California, San Francisco, San Francisco, California 94121
I. Introduction II. Parathyroid Hormone (PTH) III. Parathyroid Hormone-Related Protein (PTHrP)
IV. Mechanism of Action of PTH and PTHrP References
I. INTRODUCTION
(paracrine) factor that controls the development, morphogenesis, and function of a variety of tissues including (but not limited to) those involved in skeletal and mineral homeostasis. PTH and PTHrP are tied together historically in that PTHrP was discovered as a result of the quest to understand the pathogenesis of malignancy-associated hypercalcemia. However, they are also related structurally and produce their major physiological effects by activating a common receptor, the PTH/PTHrP (or PTH1) receptor. This chapter focuses on our current understanding of the physiology and mechanism of action of these polypeptides.
Parathyroid hormone (PTH) and PTH-related protein (PTHrP) are major factors that regulate skeletal physiology and mineral homeostasis. The appearance of parathyroid glands during the evolution of terrestrial vertebrates underscores the primary functional role of parathyroid hormone (PTH): the maintenance of adequate levels of plasma-ionized calcium in the face of a calcium-deficient terrestrial environment. The secretion of PTH by parathyroid glands is stimulated when plasma ionized calcium levels fall. Once secreted, PTH acts to restore normal levels of ionized calcium through an integrated series of actions on bone, kidney, and (indirectly) the intestine. When present as a circulating factor, PTHrP, produces target cell effects that resemble those of PTH. This is most evident in malignancy-associated hypercalcemia where tumors elaborate sufficient quantities of PTHrP to produce biochemical abnormalities overlapping those seen in primary hyperparathyroidism. However, the major physiological function of PTHrP is to act as a local
OSTEOPOROSIS, SECOND EDITION VOLUME 1
II. PARATHYROID HORMONE (PTH) A. Secretion The parathyroid glands first appear during evolution with the movement of animals from an aquatic environment to a terrestrial environment deficient in calcium. Maintenance of
221
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
222
FIGURE 1
Relationship between plasma levels of ionized calcium and the release of PTH(1 – 84) in normal humans. Variations in plasma ionized calcium were achieved by the infusion of calcium or EDTA. Note the sigmoidal relationship, ensuring significant changes in PTH secretion with small variations in ionized calcium. Reproduced from Brown [1], with permission.
adequate levels of plasma ionized calcium (1.0 – 1.3 mM) is required for normal neuromuscular function, bone mineralization, and many other physiological processes. The parathyroid gland secretes PTH in response to very small decrements in blood-ionized calcium in order to maintain the normocalcemic state. As discussed later, PTH accomplishes this task by promoting bone resorption and releasing calcium from the skeletal reservoir; by inducing renal conservation of calcium and excretion of phosphate; and by indirectly enhancing intestinal calcium absorption by increasing the renal production of the active vitamin D metabolite [1,25(OH)2D] vitamin D. The parathyroid gland functions in essence as a “calciostat,” sensing the prevailing bloodionized calcium level and adjusting the secretion of PTH accordingly (Fig. 1) [1]. The relationship between ionized calcium and PTH secretion is a sigmoidal one, allowing significant changes in PTH secretion in response to very small changes in plasma-ionized calcium. In addition to providing acute regulation of PTH secretion, ionized calcium is also a primary factor controlling chronic secretion of the hormone. Thus, sustained hypocalcemia promotes increased expression of the PTH gene [2,3] and results in parathyroid hyperplasia. A common example of the latter is the marked parathyroid hyperplasia (secondary hyperparathyroidism) that frequently accompanies chronic renal failure. 1,25(OH)2D also serves as a negative regulator of PTH gene expression and parathyroid cell hyperplasia. In chronic renal failure, both hypocalcemia and reduced circulating levels of 1,25(OH)2D presumably contribute to the progression of secondary hyperparathyroidism. There has been great progress in our understanding of how extracellular calcium controls PTH secretion [4,5]. The
ROBERT A. NISSENSON
plasma membrane of parathyroid cells contains high levels of a calcium-sensing receptor (CaR). Unlike intracellular calcium-binding proteins, which have an affinity for free calcium in the nanomolar range (consistent with intracellular levels of free calcium), the CaR binds calcium in the millimolar range. The receptor is a member of the G-protein-coupled receptor (GPCR) superfamily. It contains calcium-binding elements in its extracellular domain and signaling determinants in its cytoplasmic regions. Calcium binding to the receptor triggers activation of the G-proteins Gq and (to a lesser extent) Gi, resulting in the stimulation of phospholipase C and inhibition of adenylyl cyclase, respectively. This results in an increase in intracellular calcium and a decrease in cyclic AMP levels in parathyroid cells. By mechanisms that are not yet clear, these signaling pathways serve to suppress the synthesis and secretion of PTH. When blood ionized calcium falls, there is less signaling by the CaRs on the parathyroid cell and PTH secretion consequently increases. The essential role of the CaR can best be seen in individuals harbouring loss-of-function mutations in the CaR gene. In the heterozygous state, such mutations result in familial hypocalciuric hypercalcemia, characterized by inappropriately high levels of PTH secretion in the face of hypercalcemia [6]. These individuals are quantitatively resistant to the suppressive effect of calcium on PTH secretion due to the reduced number of parathyroid CaRs. In the homozygous state, patients display a severe increase in PTH secretion with life-threatening hypercalcemia (neonatal severe primary hyperparathyroidism).
B. Metabolism Early studies demonstrated that PTH circulates in multiple forms that can be distinguished by radioimmunoassays specific for different regions of the PTH molecule [7 – 9]. This heterogeneity has two origins (Fig. 2). PTH(1 – 84) is subject to metabolism within the parathyroid gland, resulting in the secretion of PTH fragments as well as the intact molecule. In addition, PTH(1 – 84) is metabolized in peripheral tissues. Midregion and carboxyl-terminal fragments of PTH have a much longer half-life in the circulation than PTH(1 – 84) [10 – 13]. As a result, midregion and carboxyl-terminal fragments of PTH circulate at much higher concentrations than intact PTH(1 – 84). Rapid plasma clearance of PTH is due primarily to hepatic metabolism, with a lesser contribution by the kidneys [14 – 16]. Peripheral metabolism generates mid- and carboxylterminal fragments of PTH that resemble those secreted by the parathyroid gland. Mid- and carboxyl-terminal PTH fragments are cleared by renal excretion, and thus circulating levels of these fragments are highly dependent on renal function. Extremely high levels of PTH detected with antibodies against the mid- and carboxyl-regions of the hormone in many patients with end-stage renal disease thus
223
CHAPTER 7 PTH and PTH-Related Protein
FIGURE 2
Metabolism and clearance of PTH. PTH is subject to proteolytic cleavage in the parathyroid gland, as well as in liver and kidney, resulting in the presence of inactive midregion and carboxyl-terminal PTH fragments in the circulation. Amino-terminal PTH fragments are apparently degraded rapidly and do not accumulate in the circulation. Intact PTH has a short half-life in the circulation (2 – 4 min) due to hepatic and renal metabolism. Midregion and carboxy-terminal PTH fragments are cleared by glomerular filtration. They have a much longer half-life that is dependent on the level of renal function. Reproduced from Endres et al. [341], with permission.
reflect a combination of secondary hyperparathyroidism and reduce renal clearance of PTH fragments. Mid- and carboxyl-region PTH fragments lack the amino-terminal 1 – 34 sequence of the hormone required for binding to PTH/PTHrP receptors and producing the classical effects of PTH on kidney and bone. Metabolism of PTH could produce biologically active, amino-terminal fragments of PTH, but there is little evidence for the presence of significant levels of amino-terminal PTH fragments in the circulation [17] or for significant secretion of such fragments by the parathyroid gland [18]. Presumably, both the parathyroid gland and peripheral organs contain enzymes that degrade amino-terminal fragments of PTH. This ensures that circulating levels of biologically active PTH are derived exclusively from glandular secretion of PTH(1 – 84). A few studies have demonstrated potential biological effects of mid- or carboxyl-region fragments of PTH [19 – 21], and there is also evidence for the existence of membrane receptors for these fragments [22,23]. However, the biological role of PTH fragments remains unclear. Calcium-sensitive cathepsins are responsible for cleaving PTH (1 – 84) within the parathyroid gland. Intraglandular cleavage occurs between residues 34 and 35 or between
residues 36 and 37 [24,25], and a greater proportion of PTH is cleaved under conditions of hypercalcemia [26]. The amino-terminal fragments so produced are degraded rapidly within the parathyroid gland, and thus calcium-sensitive cleavage constitutes a mechanism for the inactivation of PTH. Therefore, the level of plasma calcium determines not only the rate of synthesis and secretion of PTH, but also the extent to which secreted PTH is biologically active.
C. Physiological Actions 1. BONE a. Bone Resorption The major physiological role of PTH is to mobilize calcium from bone in order to maintain an adequate level of plasma-ionized calcium. This is accomplished by a direct action of PTH on bone that results in increased osteoclastic bone resorption and increased flux of calcium from bone into blood. Administration of PTH produces rapid movement of calcium out of bone, an effect that is associated with structural changes in cells lining the endosteal surface [27]. It has been suggested that these lining cells form an epithelial-like barrier between the
224
ROBERT A. NISSENSON
circulation and the bone extracellular fluid [28,29], and PTH may act on these cells to promote calcium transport. PTH enhances osteoclastic bone resorption within 15 min of its administration [30] and produces a sustained increase in bone resorption that appears to require the recruitment and differentiation of new osteoclasts. PTH-induced bone resorption involves the dissolution of hydroxyapatite bone mineral in the acidic microenvironment created by the osteoclast, as well as the degradation of collagen and other matrix proteins by proteolytic enzymes. Over the years, the cellular and molecular basis of the ability of PTH to promote osteoclastic bone resorption has been a source of considerable puzzlement. PTH receptors have been localized to bone-forming osteoblasts and their precursors [31,32], but it is not clear that mature osteoclasts possess PTH receptors [31 – 35]. Indeed, PTH is not able to activate isolated osteoclasts in vitro unless osteoblast-like cells are also present [36,37]. These findings suggest that PTH may produce its actions on osteoclasts indirectly, perhaps through direct interaction with cells of the osteoblast lineage. It is possible that PTH action on osteoblast lining cells alters their attachment to the surface of bone or reduces cell – cell interactions, allowing osteoclasts to gain access to the mineralized bone surface. Indeed, PTH has dramatic effects on the morphology of isolated osteoblasts [38] and alters osteoblast expression of connexin 43, a protein involved in cell – cell communication [39 – 41]. In addition, osteoblasts are known to respond to PTH by secreting proteins such as collagenase [42 – 46] and plasminogen activators [47 – 49], which may facilitate osteoclastic bone resorption. Indeed, PTH-stimulated bone resorption is blunted in mice expressing a mutated form of type I collagen that is resistant to digestion by collagenase [50]. As mentioned earlier, studies on isolated osteoclasts indicate that these cells do not display increased bone resorption
FIGURE 3
in the presence of PTH unless accessory cells such as osteoblasts are also present [36,37,51]. Osteoblasts secrete several cytokines that could potentially influence osteoclast activity through a paracrine mechanism [52 – 54]. However, it appears that direct contact between accessory cells and osteoclasts is required for PTH-induced osteoclast activation [55]. An explanation for this derives from the discovery of the role of osteoprotegerin ligand (RANKL) and its receptor (RANK) in the regulation of osteoclast differentiation and function [56,57] (See Chapters 2, 3, and 12). RANK is a tumor necrosis factor (TNF)- receptor-related protein that is expressed in osteoclast precursors as well as in differentiated osteoclasts. RANK signaling in osteoclast precursors promotes differentiation to functional osteoclasts, and RANK signaling in differentiated osteoclasts enhances bone resorption and inhibits apoptosis [58 – 61]. In either case, RANKL binding to RANK is required for signaling. RANKL is not a secreted protein but rather is an intrinsic membrane protein expressed on the surface of cells of the osteoblast lineage. Thus, direct contact between cells of the osteoblast lineage and osteoclasts or their precursors is required for the engagement of RANKL with RANK leading to osteoclast differentiation and activation. RANKL is required for normal osteoclast development and function, and mice lacking RANKL show a loss of functional osteoclasts and osteopetrosis [62]. Cells in the microenvironment of bone also secrete a truncated TNF- receptor-like molecule termed osteoprotegerin (OPG) that binds to RANKL and prevents RANK signaling [63 – 65]. The importance of OPG as a tonic suppressor of bone turnover is evident from findings with mice lacking functional expression of OPG. These animals display increased bone resorption and osteoporosis [66,67]. Current evidence indicates that the RANKL/RANK system plays a major role in PTH-induced bone resorption and calcium mobilization (Fig. 3). Administration of exogenous,
Regulation of osteoclast differentiation and activation by PTH. Binding of PTH by receptors on osteoblasts results in increased expression of osteoprotegerin ligand (RANKL) on the cell surface. Activation of the PTH receptor can also reduce the secretion of the RANKL inhibitor osteoprotegerin (OPG), which is produced by cells in the bone microenvironment. These effects of PTH promote the action of RANKL on its receptor (RANK) on the surface of osteoclast precursors and mature osteoclasts. RANK signaling, together with the action of macrophage colony-stimulating factor, stimulates the differentiation of osteoclast precursors and promotes the activation of mature osteoclasts.
225
CHAPTER 7 PTH and PTH-Related Protein
soluble RANKL to mice elicits severe hypercalcemia within 1 day of administration, and increased osteoclast activity and bone loss are evident within 3 days [58]. Administration of OPG (RANKL antagonist) blocks the calcemic action of exogenous PTH in vivo [63]. Addition of OPG also inhibits PTH-induced osteoclast activation and bone resorption in vitro [68,69]. PTH is reported to increase the ratio of RANKL:OPG expressed by osteoblastic cells, an effect that is largely due to the ability of PTH to increase the expression of RANKL [60,70,71]. PTH also inhibits the expression of OPG [70,72]. Factors such as interleukin (IL)-6 [73] and insulin-like growth factor (IGF)-1 [74] have been suggested to play a role in mediating the actions of PTH on osteoclast formation and bone resorption. However, induction of RANKL in cells of the osteoblast lineage may well be the major mechanism underlying these effects of PTH under physiological conditions. b. Bone Formation PTH acts directly on cells of the osteoblast lineage, thereby influencing osteoblast differentiation and function, and consequently bone formation. Administration of PTH intermittently to animals or humans results in a marked anabolic response of the skeleton [75 – 82] (See also Chapter 77). PTH promotes bone formation in both trabecular and cortical bone, and these actions are associated with increased trabecular thickness and increased bone strength [83 – 89]. High levels of PTH are known to produce an increase in the number of osteoblasts, which results in part from the coupling between increased osteoclastic resorption and new bone formation. However, intermittent treatment with low doses of PTH produces a direct positive effect on bone formation that is independent of preceding bone resorption. The cellular basis for this action of PTH is not fully understood, but there are a number of potential targets for PTH (Fig. 4). PTH could promote the differentiation of stromal cell osteoblast precursors to matrix-synthesizing mature osteoblasts. The hormone could also function as an osteoblast mitogen or increase the life span of mature osteoblasts, thereby increasing their pool size. Finally, PTH could directly enhance the ability of mature osteoblasts to synthesize and secrete matrix proteins and to promote matrix mineralization. Data support a number of these possibilities. PTH receptors are present on osteoblast precursors, including bone marrow stromal cells [90 – 92]. Available evidence indicates that PTH increases the number of active, replicating osteoblasts, but its direct effect on the replication of osteoblastic cells is variable [93 – 96]. Model systems for osteoblast differentiation in vitro reveal a positive effect of PTH on differentiation, depending on the dose and mode of exposure, with intermittent treatment with low doses being most consistently effective [97 – 99]. Active osteoblasts are subject to death by apoptosis and that PTH treatment reduces the number of apoptotic osteoblasts in vivo [100].
FIGURE 4
Possible mechanisms contributing to the anabolic effect of PTH in bone. PTH may act on bone marrow stromal cell precursors to promote their differentiation to functional osteoblasts. PTH could also act directly on osteoblasts to increase their number or their functional activity. Finally, PTH could increase the life span of mature osteoblasts by inhibiting their death via apoptosis.
Taken together, available data support the notion that PTH elicits an increase in osteoblast number via actions to promote osteoblast differentiation and to inhibit osteoblast apoptosis. Direct effects of PTH on osteoblasts in vitro to promote the synthesis of matrix proteins have also been reported [101 – 103]. Some of these effects may be secondary to the release of osteoblast growth factors such as IGF-1 [104], and their relevance to the anabolic action of PTH in vivo remains to be established. Intermittent (e.g., once daily) treatment with PTH elicits skeletal effects in which increased bone formation predominates, whereas continuous treatment with high doses of PTH results in a major increase in bone resorption. Continuous treatment of target cells with high doses of PTH results in a loss of responsiveness (desensitization), and it is possible that the anabolic effects of PTH are particularly sensitive to hormone-induced desensitization. Intermittent administration of PTH could allow for resensitization of the anabolic response prior to administration of a subsequent dose of hormone. However, continuous administration of lower doses of PTH also elicits an anabolic skeletal response, suggesting that the balance between bone resorption and anabolism
226 may be related to the dose of PTH rather than to its intermittent administration. The effects of PTH also differ depending on the nature of the skeletal site, with trabecular bone displaying the greatest increase in mass in response to PTH. At doses of PTH that are anabolic in trabecular bone, cortical bone displays increased bone resorption as well as increased bone formation. The net effect of PTH treatment on cortical bone mass in thus variable. 2. KIDNEY PTH produces a series of renal actions that help ensure that calcium mobilized from bone contributes optimally for the maintenance of plasma-ionized calcium levels. The reanl actions of PTH include inhibition of renal phosphate reabsorption, stimulation of renal calcium reabsorption, and increased production of 1,25(OH)2D. The ability of PTH to inhibit renal phosphate reabsorption has been known for many years, providing the basis for the clinical Ellsworth – Howard test of renal responsiveness to the hormone [105]. Patients with primary hyperparathyroidism display hypophosphatemia and decreased renal tubular reabsorption of phosphate, whereas hypoparathyroid patients are hyperphosphatemic and have increased phosphate reabsorption. Phosphate forms a complex with free calcium in blood. Thus, for a given level of serum calcium, ionized calcium will be reduced as serum phosphate increases. Under conditions of relative hypocalcemia (e.g., during chronic dietary calcium deficiency), PTH secretion is increased, resulting in increased bone resorption. Both calcium and phosphate are released from hydroxyapitite during the process of bone resorption. By promoting renal excretion of phosphate, PTH facilitates a rise in ionized as well as total plasma calcium. Phosphate reabsorption in the proximal renal tubule is dependent in part on the activity of the type IIa sodium – phosphate cotransporter. The phosphaturic action of PTH derives from the action of the hormone to inhibit the function of this transporter [106]. The type IIa transporter is located in the apical plasma membrane and permits the coupled transport of sodium and phosphate from the tubule into the renal cell. Exposure of proximal tubular cells to PTH results in a reduced Vmax of the transporter [107, 108], and this is associated with a decrease in the amount of the transporter in the apical plasma membrane [109]. Acute exposure of the proximal tubular cells to PTH enhances the endocytosis and subsequent lysosomal degradation of the transporter, and this may be the major mechanism responsible for rapid PTH-induced inhibition of renal phosphate reabsorption [110,111]. PTH appears to regulate the type II transporter by enhancing its rate of turnover rather than by suppressing its synthesis [112]. PTH also acts to increase renal calcium reabsorption, thus ensuring that only small amounts of calcium released during PTH-induced bone resorption are lost via renal
ROBERT A. NISSENSON
excretion. The major sites for this effect of PTH are in the distal convoluted tubule and the thick ascending limb of Henle’s loop [113,114]. Evidence indicates that distal renal tubular calcium reabsorption is an active process that requires calcium influx through dihydropyridine-sensitive calcium channels located in the apical plasma membrane [115]. Drugs that inhibit these channels are effective in blocking PTH-induced renal calcium reabsorption. Unlike voltage-sensitive calcium channels in excitable tissues, PTH-responsive calcium channels in the distal nephron are activated by membrane hyperpolarization [116]. PTH appears to open calcium channels by inducing hyperpolarization of the apical plasma membrane. Calcium entering the distal renal tubular cell in this manner is transported into the extracellular compartment via a sodium – calcium exchanger present on the basolateral plasma membrane [117]. PTH promotes intestinal calcium absorption indirectly through an action to increase circulating levels of 1,25(OH)2D. This vitamin D metabolite acts directly on intestinal epithelial cells to increase the efficiency of calcium (and phosphate) absorption. Primary hyperparathyroidism is commonly associated with increased circulating levels of 1,25(OH)2D, whereas reduced levels of this metabolite are present in hypoparathyroidism [118]. PTH produces this effect by increasing the rate of production of 1,25(OH)2D through activation of the 25(OH)D-1-hydroxylase enzyme located in the proximal renal tubule [119 – 122]. The gene encoding this enzyme has been cloned in several laboratories [123 – 125]. Studies in vivo as well as in cultured renal cell lines indicate that PTH increases the expression of the 25(OH)D-1-hydroxylase gene through a transcriptional mechanism [126,127].
III. PARATHYROID HORMONERELATED PROTEIN (PTHrP) A. Role in Malignancy-Associated Hypercalcemia The frequent occurrence of hypercalcemia in individuals with a variety of malignancies has been recognized for many years. An important clue to the pathogenesis of malignancy-associated hypercalcemia (MAH) came with the recognition 20 years ago that many such individuals display an increased excretion of renal-derived (“nephrogenous”) cyclic AMP [128]. Activation of the renal PTH receptor by elevated circulating levels of PTH in hyperparathyroidism was the only known cause of increased nephrogenous cyclic AMP, and thus it was suggested that malignant tumors are capable of producing a factor that activates PTH receptors. Plasma levels of immunoreactive PTH were
227
CHAPTER 7 PTH and PTH-Related Protein
found to be low in patients with MAH, indicating that the relevant circulating factor was not PTH itself. Using the activation of PTH receptors as an assay, several groups succeeded in isolating and ultimately identifying the PTH-like etiologic factor in MAH [129 – 132]. This factor was termed PTH-related protein (1) because of its ability to bind to and activate the PTH receptor and (2) because of its limited sequence similarity to PTH [133 – 135]. The PTHrP gene is subject to alternative splicing, resulting in the production of three protein products ranging from 139 to 173 amino acids differing only in their carboxyl-terminal sequence [136,137]. PTHrP is capable of reproducing the major target cell actions of PTH and (like PTH) does so via the amino-terminal 34 amino acids or so of the protein. A comparison of the 1 – 34 sequences of PTH and PTHrP reveals significant amino acid homology, with identity in 8 of the 13 amino terminal residues. Two of the known contact sites between PTH and the PTH/PTHrP receptor are within this 13 amino-acid homologous region [138], indicating that these ligands use very similar mechanisms to activate their common receptor. The molecular mechanisms underlying the over expression of PTHrP by malignant tumors remain unclear. As the mass of PTHrP-expressing tumor cells expands, systemic levels of PTHrP eventually increase sufficiently to allow the peptide to elicit endocrine effects on PTH/PTHrP receptors in bone and kidney, resulting in MAH.
B. Physiological Roles Although PTHrP produces PTH-like target cell effects in patients with MAH, circulating levels of PTHrP are very low to undetectable in normal individuals. This, coupled with the widespread expression of the PTHrP gene in normal tissues, suggested that PTHrP was likely to have physiological functions as a local, paracrine factor rather than as a systemic hormone. Subsequent studies have confirmed that PTHrP indeed plays an important role as a paracrine factor in a wide variety of tissues (Table 1) [139 – 143], as summarized next. 1. ENDOCHONDRAL BONE DEVELOPMENT The first direct evidence concerning a physiological role for PTHrP appeared in 1994 with the report of the phenotype of mice lacking the expression of PTHrP due to targeted gene ablation [144]. These animals died shortly after birth and were found to display a form of short-limbed dwarfism with generalized chondrodysplasia. The most striking feature of mice lacking the expression of PTHrP is the disruption of normal endochondral ossification. Although the most obvious gross phenotypic abnormality is short-limbed dwarfism, the defect in endochondral bone formation is generalized. The role of PTHrP is best under-
TABLE 1 Target tissue
Target Tissue Actions of PTHrP Actions
Cartilage
Inhibits terminal chondrocyte differentiation; increases chondrocyte proliferation
Bone
Regulates bone resorption
Mammary gland
Facilitates branching morphogenesis of mammary epithelium; may play an endocrine or paracrine role in lactation
Skin
Inhibits terminal differentiation of keratinocytes; promotes normal hair follicle development
Teeth
Promotes normal tooth eruption
Extraembryonic endoderm Enhances the differentiation of primitive endoderm to parietal endoderm Smooth muscle
Serves as a general smooth muscle relaxant
Central nervous system
Inhibits neuronal L-type calcium channel activity; protects neurons from excitotoxicity
Placenta
Maintains the positive maternal – fetal transplacental calcium gradient
stood in the context of the homeostatic mechanisms regulating the differentiation of cartilage and bone during endochondral bone formation. In the long bones, chondrogenesis is initiated by the differentiation of mesenchymal cell precursors that form nodules and begin to express characteristic genes, including those encoding type II collagen and other cartilage matrix proteins [145]. These early chondrocytes are mitotically active, but the cells in the center of the nodule become hypertrophic, cease dividing, and express gene products characteristic of mature chondrocytes (such as type X collagen). Hypertrophic chondrocytes undergo programmed cell death (apoptosis), which is accompanied by vascular invasion. Subsequently, the cartilage scaffold is replaced by bone. In the growing animal, this process is continued in the growth plate where the differentiation process is subject to tight temporal and spatial control. Mesenchymal cell differentiation and early chondrocyte proliferation occur in a columnar array inward from the articular surface. This spatial profile is extended as the chondrocytes become prehypertrophic, then hypertrophic. After the hypertrophic cells undergo apoptosis, the cartilaginous scaffold is remodeled and subsequently replaced by bone. The control of endochondral bone formation is maintained by a complex series of extracellular cues and intracellular signaling pathways. One of these factors is Indian hedgehog (Ihh), a member of the ancient hedgehog family of secreted patterning molecules. Ihh functions to promote chondrocyte proliferation and to maintain the pool of proliferating chondrocytes, thus extending the length of the differentiating cartilaginous growth plate prior to terminal
228
FIGURE 5
Regulation of chondrocyte differentiation by PTHrP and Indian hedgehog (Ihh). Ihh is secreted by postmitotic prehypertrophic chondrocytes and acts as a negative feedback regulator of the differentiation of proliferating chondrocytes. This occurs, in part, via the effect of Ihh to induce the expression of PTHrP in perichondrial cells. This occurs either directly or through an unknown mediator(s). PTHrP then acts directly on PTH/PTHrP receptors in proliferating chondrocytes to inhibit their differentiation. There is also evidence that Ihh inhibits chondrocyte proliferation through a PTHrP-independent mechanism.
differentiation and ossification [146]. Ihh is produced by postmitotic prehypertrophic chondrocytes, suggesting that the factor may serve as a negative feedback signal that slows the rate of transition of chondrocytes from the proliferative to the prehypertrophic pool. Ihh also appears to directly act on cells of the osteoblast lineage to promote their differentiation to mature bone-forming cells [146]. PTHrP appears to mediate some, but not all, of the actions of Ihh on endochondral bone formation (Fig. 5). PTHrP directly inhibits the differentiation of proliferating chondrocytes to postmitotic prehypertrophic cells. Lack of PTHrP results in accelerated chondrocyte differentiation with shortened growth plates and premature ossification. The cellular composition of the growth plates of PTHrP / animals is abnormal, with a marked reduction in the number of proliferating chondrocytes. Conversely, overexpression of PTHrP in chondrocytes of mice bearing a collagen II promoter-PTHrP transgene resulted in a distinct form of chondrodysplasia, which is characterized by shortlimbed dwarfism and delayed ossification [147]. At birth, these animals displayed a cartilaginous endochondral skeleton, and histological evaluation revealed a marked suppression of the chondrocyte differentiation program. By 7 weeks of age ossification was evident, but the long bones remained foreshortened and misshapen. Similar abnormalities are seen in humans with hereditary Jansen’s metaphyseal chondrodysplasia. The latter disorder has been associated with mutations in the PTH/PTHrP receptor that result in constitutive receptor activation [148]. Ihh acts directly or indirectly on cells in the periarticular perichondrium to increase expression of the PTHrP gene [149]. The effect of Ihh (or the related protein Sonic hedgehog) to delay terminal differentiation of chondrocytes in the long bones was not seen in PTHrP / or in PTH/PTHrP
ROBERT A. NISSENSON
receptor / mice, indicating an intermediary role of PTHrP in Ihh action in endochondral bone formation [149,150]. Consistent with this conclusion, a type II collagen promoter-driven constitutively active PTH/PTHrP receptor transgene has been reported to rescue the abnormally accelerated chondrocyte differentiation program in Ihh / mice [151]. These animals nonetheless displayed short-limbed dwarfism and decreased chondrocyte proliferation, demonstrating that PTHrP is not the only mediator of the multiple actions of Ihh on endochondral ossification. This conclusion is further supported by the observation that the severity of short-limbed dwarfism is much more severe in Ihh /, PTHrP / mice than in Ihh /, PTHrP / mice [151]. Solid evidence shows that the PTH/PTHrP receptor is responsible for initiating the actions of PTHrP on the differentiation of growth plate chondrocytes. The PTH/PTHrP receptor is expressed in proliferating chondrocytes as well as in cells in the transitional zone between proliferating and hypertrophic chondrocytes, where the regulation of terminal differentiation occurs [152]. PTH/PTHrP receptor / mice display growth plate abnormalities similar to those seen in PTHrP / mice [150]. Patients with inherited mutations in the PTH/PTHrP receptor that cause constitutive (i.e., ligand-independent) signaling (Jansen’s metaphyseal chondrodysplasia) display growth plate abnormalities similar to those seen in mice overexpressing a collagen II promoter-PTHrP transgene [153]. Lack of expression of functional PTH/PTHrP receptors in humans is associated with Blomstrand chondrodysplasia [154,155], a lethal disorder characterized by premature endochondral ossification [156]. Precisely how signaling by the PTH/PTHrP receptor results in the delay of chondrocyte differentiation in the transitional zone is unclear. It is known that programmed cell death (apoptosis) occurs during the late terminal differentiation of chondrocytes. This process has been shown to be inhibited by PTHrP, which upregulates the antiapoptotic protein bcl-2 through a cyclic AMP-dependent mechanism [157]. Mice lacking expression of a functional bcl-2 gene are known to display accelerated differentiation of growth plate chondrocytes, although the severity of the phenotype is much less than that seen in PTHrP / mice. Thus, inhibition of apoptosis may be one of several pathways involved in PTHrP-induced suppression of chondrocyte differentiation. 2. MAMMARY GLAND DEVELOPMENT Targeted overexpression of PTHrP in mammary myoepithelial cells of transgenic mice provided direct evidence of a possible role for PTHrP in mammary gland development [158]. The mammary ducts of 18- to 21-day-old transgenic mice were normal in terms both in the size of the ducts and in the branching morphogenesis of the developing gland. However, by 6 weeks of age, the transgenic animal displayed a delay in the development of the mammary duct
229
CHAPTER 7 PTH and PTH-Related Protein
system and a reduction in the degree of ductal branching. Pregnant transgenic animal displayed similar defects, as well as a diminished formation of terminal ductules. Overexpression of PTH in mammary myoepithilial cells of transgenic mice produced identical morphogenetic defects, indicating that this action of PTHrP is mediated by the PTH/PTHrP receptor. The postnatal role of PTHrP in mammary gland development was studied in PTHrP / mice expressing a PTHrP transgene targeted to cartilage [159], allowing postnatal survival. At 4 months of age, female transgenic mice lack mammary glands. The mammary fat pads appear normal, but mammary epithelial ducts are missing. PTHrP / mice display arrest of mammary duct development beginning between days 15 and 18 of embryogenesis. At this time, there is a degeneration of epithelial elements within the ducts, and the initiation of normal branching morphogenesis of the mammary glands does not occur. In normal animals, PTHrP is expressed in mammary epithelial cells [159,160], whereas functional PTH/PTHrP receptors are expressed in the underlying mesenchyme [159,161]. This pattern of expression suggests that PTHrP is an epithelial signal that acts on PTH/PTHrP receptors in mesenchymal cells to promote mammary epithelial morphogenesis. Consistent with this notion, PTH/PTHrP receptor / mice display the same defects in embryonic mammary development seen in PTHrP / mice. Moreover, normal morphogenesis requires PTH/PTHrP receptor expression specifically in mammary mesenchymal cells [161]. The factors that regulate epithelial production of PTHrP, and the nature of the mesenchymal targets of PTH/PTHrP receptor signaling, are unknown. The mesenchymal genes encoding tenascin C and the androgen receptor are induced by PTHrP [162]. PTHrP / or PTH/PTHrP receptor / male mice fail to display the normal androgen-dependent apoptotic destruction of the mammary bud, suggesting that induction of the androgen receptor by PTHrP is essential for sexual dimorphism during mammary development. 3. SKIN AND TOOTH DEVELOPMENT Keratinocytes were the first normal cells shown to express PTH-like bioactivity [163] and subsequently the PTHrP gene [163]. PTHrP is expressed in the basal layer through the granulosa layer of the skin, with epidermal expression detectable as early as day 14 of embryogenesis in the rat [164,165]. PTH/PTHrP receptors are present in dermal fibroblasts [166], and novel binding sites for PTHrP are detected in keratinocytes [167]. In cultured human keratinocytes, suppression of PTHrP production resulted in increased cell proliferation [168] and decreased differentiation [169]. Thus, PTHrP may have a role in the local regulation of epidermal cell proliferation and differentiation. Targeted overexpression of PTHrP in basal keratinocytes and outerroot sheath cells of hair follicles in transgenic
mice resulted in a failure of ventral hair eruption that was evident within 6 days after birth [170]. Dorsal hair was evident, but its eruption was delayed and the hairs were shorter and thinner compared to those of normal littermates. Histological evaluation of the transgenic mice revealed thickening of the ventral epidermis and expansion and increased cellularity of the dermis. Hair follicle development was delayed substantially in both ventral and dorsal skin of transgenic mice. These effects are probably due to disruption of the normal epithelial – mesenchymal interactions required for proper hair follicle development and epidermal differentiation. PTHrP / mice that have been rescued by expression of a type II collagen-PTHrP transgene display thinning of the epidermis with hypoplastic sebaceous glands and thinning of hair [171]. These abnormalities could be reversed by targeted expression of PTHrP in skin, indicating that PTHrP expression in basal keratinocytes is necessary for maintaining normal epithelial – mesenchymal interactions during epidermal differentiation. Inhibition of PTHrP action in skin was found to produce an increase in the number of follicles involved in active hair growth, further supporting a role of PTHrP in promoting hair follicle development [172]. PTHrP apparently maintains the pool of proliferating keratinocytes by suppressing their terminal differentiation. However, the underlying mechanisms responsible for regulating expression of PTHrP and the presumed mesenchymal responses remain obscure and are not understood at the molecular level. The fibroblast growth factor (FGF)-like factor keratinocyte growth factor (KGF) has been reported to be induced by PTHrP in dermal fibroblasts [173]. Because KGF is a critical regulator of keratinocyte growth and differentiation, it could be a mediator of PTHrP action in skin. PTHrP / mice also show cranial chondrodystrophy with a failure in normal tooth eruption [174]. In normal animals, PTHrP is expressed in the enamel epithelium, whereas the PTH/PTHrP receptor is expressed in the adjacent dental mesenchyme and in alveolar bone. These findings suggest that PTHrP is a regulator of epithelial – mesenchymal interactions during tooth development, as well as promoting the resorption of alveolar bone that is required for normal tooth eruption. 5. OTHER ACTIONS OF PTHRP PTHrP is expressed in a variety of smooth muscles where it functions as a local muscle-relaxing agent. Increased intraluminal pressure (either from muscle contraction or from expanding intraluminal contents) is a known stimulus for PTHrP gene expression. Myometrial expression of PTHrP peaks just before the end of pregnancy, and this effect is specific for the pregnant uterine horn in unilaterally pregnant animals [175]. Mechanotransduction is likely to be the primary stimulus, as physical stretch induces PTHrP expression in the nonpregnant rat uterus [176].
230 Human amniotic fluid contains high levels of PTHrP [177,178], and it is possible that PTHrP produced in the amnion plays a role in suppressing myometrial contractions and/or in regulating chorionic blood flow. PTHrP is also expressed in gastric and bladder smooth muscle and promotes muscle relaxation in these tissues in response to distension [179,180]. Pharmacological doses of PTH can reproduce the relaxing effects of PTHrP, strongly indicating the involvement of the PTH/PTHrP receptor. PTHrP has effects on both the contractility and the proliferation of vascular smooth muscle. PTHrP is widely expressed in vascular smooth muscle, and administration of PTHrP in vivo and in vitro elicits a vasodilitory response [181,182]. Expression of PTHrP in vascular smooth muscle is increased in experimental models of hypertension and in response to vasoconstrictors such as angiotensin II [183,184]. Targeted overexpression of PTHrP in vascular smooth muscle of transgenic mice results in decreased baseline blood pressure as well as in a diminished hypotensive response to exogenous PTHrP, with the latter possibly due to desensitization [185]. The role of endogenous PTHrP is seen in transgenic mice overexpressing the PTH/PTHrP receptor in vascular smooth muscle [186]. These animals are hypotensive and (as expected) are hyperresponsive to exogenous PTHrP with respect to vasodilitation. PTHrP is also induced in the blood vessels bathing skeletal muscle after muscle stimulation, perhaps promoting new capillary formation in response to increased muscle contraction [187]. Taken together, these results implicate PTHrP as an important physiological regulator of static blood pressure and as a counterregulatory factor secreted in response to vasoconstriction. The genes encoding PTHrP and the PTH/PTHrP receptor are widely expressed in the central nervous system, with particularly high levels seen in cerebellar granule cells [188,189]. These cells also express high levels of L-type calcium channels, and expression of PTHrP appears to be induced by depolarization-induced calcium influx through these channels [190]. Cerebellar granule cells are subject to excitatory cell death in response to agents such as kainic acid, which trigger calcium entry through L-type calcium channels. PTHrP blocks this excitatory cell death by inhibiting L-type calcium channel activity through a mechanism that probably involves cyclic AMP signaling via the PTH/PTHrP receptor [191]. This is consistent with previous reports that exogenous PTH inhibits L-type calcium channel activity [192]. These findings suggest that PTHrP functions as a neuronal survival factor produced in response to neuroexcitatory stimuli. Addition of a blocking antibody to PTHrP prevents cerebellar granule cell survival under depolarizing conditions, strongly suggesting that PTHrP is the endogenous factor responsible for neuroprotection [193]. As discussed earlier, PTHrP is expressed in the myometrium during pregnancy in response to distension
ROBERT A. NISSENSON
produced by the growing fetus. By inducing relaxation of uterine smooth muscle, locally produced PTHrP permits progressive intrauterine growth of the fetus and may also assist in maintaining the uterus in a quiescent state until the onset of parturition. PTHrP also plays an important role in the fetal-placental unit during pregnancy. The protein is expressed in human amniotic tissue and may serve to increase chorionic blood flow [177,178]. A role for PTHrP in placental calcium transport is suggested by studies demonstrating that the loss of the positive maternal – fetal placental calcium gradient produced by parathyroidectomy of fetal sheep could be restored by perfusion of the placenta with PTHrP [194]. The fetal parathyroid gland is a site of expression of PTHrP [195], suggesting that this might be the source of PTHrP responsible for maintaining the positive maternal – fetal calcium gradient. Moreover, relative to wild type littermates, PTHrP / fetuses are hypocalcemic and have a reduced ability to accumulate calcium from the mother’s circulation [196]. Thus, fetal PTHrP appears to have an endocrine action to maintain sufficient fetal calcium levels for skeletal growth and mineralization. A role for PTHrP during lactation was first suggested by the observation that suckling is a powerful stimulator of mammary PTHrP gene expression [197]. However, the precise physiological role of PTHrP during the period of lactation remains controversial. There is evidence for the existence of a factor, distinct from PTH and vitamin D metabolites, that is capable of maintaining plasma calcium homeostasis in lactating women. Indeed, systemic maternal PTHrP levels have been reported to increase during suckling [198] and to be elevated during lactation [199]. However, others have reported that circulating levels of PTHrP are unchanged during lactation [200,201]. A further complication is the observation that extremely large quantities of PTHrP are secreted into milk rather than into the bloodstream during lactation [200]. Suckling animals and humans thus ingest large amounts of PTHrP over an extended time period. However, evidence that milk-derived PTHrP is absorbed in an active form and/or is physiologically important in suckling infants or animals is lacking. It may be that suckling-induced expression of PTHrP has a local effect (e.g., increased blood flow) that facilitates mammary function during lactation.
IV. MECHANISM OF ACTION OF PTH AND PTHrP A. Signal Transduction Many of the actions of PTH and PTHrP are initiated by binding of these proteins to the PTH/PTHrP receptor, a Gprotein-coupled receptor that activates two G-proteins and
CHAPTER 7 PTH and PTH-Related Protein
FIGURE 6
Signal transduction by the PTH/PTHrP receptor. PTH and PTHrP bind to determinants in the extracellular domain and in the body of the receptor. This leads to conformational changes in the transmembrane helices and consequent structural changes in the cytoplasmic domain. The latter permit productive interaction between the receptor and the G-proteins Gs and Gq, activating the adenylyl cyclase (AC) and phospholipase C (PL-C) signaling pathways, respectively. These pathways are thought to cooperate in determining the cellular response to the receptor activation. Most available evidence supports a primary role of the cyclic AMP/protein kinase A (PK-A) pathway in mediating the biological effects of PTH/PTHrP receptor activation, with the PL-C pathway playing a modulatory role.
thereby two major signal transduction pathways (Fig. 6). Shortly after the discovery of the cyclic AMP signaling pathway, it was found that PTH is capable of increasing levels of cyclic AMP in target cells through activation of the enzyme adenylyl cyclase [202 – 205]. Cyclic AMP is a second messenger in the cellular action of a wide variety of hormones and other extracellular regulatory molecules. It activates cyclic AMP-dependent protein kinase (PK-A), which in turns phosphorylates and thereby regulates key proteins that participate in physiological responses. Very little is known about the identity of substrates of PK-A that are phosphorylated in response to PTH/PTHrP receptor activation. These presumably include transcription factors, ion channels, transporters, and enzymes involved in cellular metabolism. PTH/PTHrP receptors also activate phospholipase C, an enzyme that hydrolyzes the plasma membrane phospholipid phosphatidylinositol-4,5-bisphosphate to produce diacylglycerol (DG) and soluble 1,4,5-inositol trisphosphate (IP3). DG and IP3 function as second messengers, the former by activating protein kinase-C (PK-C) and the
231 latter by binding to and opening calcium channels on the membrane of the endoplasmic reticulum, thereby increasing cytosolic-free calcium. The PTH/PTHrP receptor is clearly required for PTHstimulated bone resorption [206], and a number of studies have been carried out to identify the nature of the relevant signaling pathway(s). Agents that raise cellular cyclic AMP levels (e.g., analogues of cyclic AMP, forskolin) are capable of eliciting bone resorption in organ culture [207 – 211]. In addition, inhibition of cyclic AMP phosphodiesterase (thus augmenting the cellular cyclic AMP response to PTH) potentiates PTH-induced bone resorption [212]. Activation of phospholipase C-related pathways with calcium ionophores and phorbol esters also promotes bone resorption in organ culture [213 – 215], and inhibition of protein kinase C is reported to block PTH-stimulated bone resorption [216,217]. However, at least in mouse calvarial cultures, the effects of calcium ionophores and phorbol esters require the intermediary synthesis of prostaglandins, whereas PTH-induced bone resorption does not [218]. Moreover, under some circumstances, these agents can inhibit bone resorption [219-221]. Thus, available evidence indicates that the cyclic AMP pathway plays a primary second messenger role in the stimulation of bone resorption by PTH. PTH-induced differentiation of hematopoietic precursors to osteoclast-like cells also involves the cyclic AMP pathway [222 – 224], although the phospholipase C pathway may also contribute [225]. In primary cultures of human bone marrow stromal cells, the cyclic AMP pathway has been shown to downregulate expression of the OPG gene [226], which could allow for greater bone resorption in response to PTH-induced production of RANKL. From earlier studies of PTH-induced bone resorption, it is expected that the cyclic AMP pathway would promote expression of RANKL in osteoclast precursors. However, studies of the RANKL promoter failed to demonstrate a stimulatory effect of cyclic AMP [227], indicating that effects of this pathway on RANKL expression are likely to be indirect. There has been great interest in defining the signaling events that are responsible for the anabolic response of the skeleton to intermittent administration of PTH. Progress in this area has been hampered by the lack of a useful in vitro model system for the investigation of the anabolic response to PTH and the uncertainty about the cellular basis of this effect. PTH generally has been reported to have an antiproliferative effect on cultured osteoblasts, although it is reported to promote proliferation in an osteoblast precursor model [228]. PTH can also promote osteoblast differentiation in vitro, depending on the time and duration of treatment [98,99,229]. In vivo studies have demonstrated that amino-terminal fragments of both PTH and PTHrP are anabolic, implicating the PTH/PTHrP receptor as the likely initiator of this skeletal response. Interestingly, PTH(1 – 30)
232 and PTH(1 – 31), which activate adenylyl cyclase but have a greatly reduced ability to activate phospholipase C, are effective as anabolic agents in bone [230,231]. This result suggests that the cyclic AMP pathway is the major mediator of the anabolic actions of PTH. However, it should be noted that cyclic AMP signaling has generally been linked to the inhibition of osteoblast proliferation and differentiation in vitro [232 – 235]. This finding suggests that these in vitro model systems of osteoblast proliferation and differentiation may not be relevant to the mechanism of PTH-induced anabolism. Alternatively, signaling pathways other then the cyclic AMP system may participate in the anabolic response to PTH. Microdissection studies revealed the presence of PTHstimulated cyclic AMP generation in the proximal convoluted tubule where sodium-dependent phosphate cotransport occurs [236,237]. Analogues of cyclic AMP were found to be effective in reproducing the phosphaturic effect of PTH [238 – 241]. In pseudohypoparathyroidism Ia, genetic deficiency of Gs- (a protein that links receptors to the activation of adenylyl cyclase) is associated with resistance to the phosphaturic action of PTH [242 – 245]. With the discovery that an opossum kidney cell line (OK) retains PTH receptors [246] and PTH-inhibited sodium – phosphate cotransport [247], it became possible to carry out studies on the mechanisms of PTH inhibition of phosphate transport. Cyclic AMP clearly has a primary, although not exclusive, role in the negative regulation of sodium – phosphate cotransport [247 – 250] [251]. Cyclic AMP (like PTH) promotes rapid downregulation of the type II sodium – phosphate cotransporter in OK cells via enhanced transporter endocytosis and lysosomal degradation [106,109,111,252]. Activation of PK-C by the PTH/PTHrP receptor may also contribute to the inhibition of phosphate transport, as treatment of OK cells with PMA or other phorbol esters substantially inhibits sodium – phosphate cotransport and reduces the expression of the type II cotransporter in some [249,253 – 256] but not all [252] studies. The cyclic AMP pathway is known to be important in mediating the effect of PTH to increase the activity of the 25(OH)D-1-hydroxylase in the proximal renal tubule [122,257,258]. PTH has a positive effect on the renal expression of the 1-hydroxylase mRNA in vivo [126,127]. This appears to occur at the level of gene transcription, as upstream elements in the 5-region of the 1-hydroxylase gene confer transcriptional responses to PTH and forskolin in cultured kidney cells [259 – 261]. The precise elements in the promoter responsible for these effects have not been identified, but putative binding sites for the transcription factors CREB and AP-1 are present and represent possible targets. PTH-stimulated phospholipase C activation might also contribute to the 1-hydroxylase response, as the combination of a calcium ionophore and PMA was shown to promote a sustained increase in 1,25(OH)2D production in
ROBERT A. NISSENSON
perifused rat proximal tubule cells [262]. Under some circumstances, inhibitors of PK-C have been shown to suppress PTH-induced renal production of 1,25(OH)2D [263]. In light of these findings, it is possible that phospholipase C has a role in the transcriptional response of the 1-hydroxylase gene to PTH. The PTH-induced stimulation of renal calcium transport in the distal convoluted tubule appears to require activation of both PK-A and PK-C pathways [113]. Inhibition of either of these kinases suppresses PTH-induced calcium uptake by distal tubular cells [264]. Moreover, simultaneous activation of both kinases was shown to be necessary and sufficient to reproduce the effect of PTH on calcium uptake [265]. PTH does not appear to increase the activity of phospholipase C in the distal renal tubule [266], suggesting that an alternative mechanism exists for the PTH-induced generation of diacylglycerol. In this regard, PTH is capable of increasing the activity of phospholipase D, an enzyme that hydrolyzes phosphatidyl choline to produce phosphatidic acid and, indirectly, diacylglycerol [266,267]. It is possible that the activation of phospholipase D participates in the activation of PK-C that is reported to occur in response to PTH as well as amino-terminally truncated PTH fragments [268]. There is as yet very little information concerning the signaling mechanisms responsible for mediating the developmental and morphogenetic actions of PTHrP. It is likely that the cyclic AMP signaling pathway is of primary importance, and genetic deficiency of the -subunit of Gs produces a constellation of developmental abnormalities (Abright’s hereditary osteodystrophy) that overlap those seen in animals lacking PTHrP or the PTH/PTHrP receptor. However, the precise role of adenylyl cyclase and phospholipase C signaling pathways in mediating specific paracrine actions of PTHrP remains to be defined.
B. PTH/PTHrP Receptors 1. ACTIVATION OF G-PROTEINS Early studies on the PTH/PTHrP receptor demonstrated a prominent role for GTP and its analogues in regulating ligand – receptor affinity and signaling, suggesting that this receptor couples to GTP-binding (G) proteins [269 – 274]. The cloning of the cDNA encoding the PTH/PTHrP receptor [275] revealed a predicted protein sequence containing seven putative membrane-spanning domains (Fig. 7), a topology characteristic of members of the GPCR superfamily [276,277]. In the case of the PTH/PTHrP receptor, the major G-proteins that can be activated are Gs and Gq. Activation of Gs leads to increased adenylyl cyclase activity, resulting in increased cellular levels of cyclic AMP and activation of PK-A. Activation of Gq results in the stimulation of phospholipase
233
CHAPTER 7 PTH and PTH-Related Protein
FIGURE 7 Structural model of the PTH/PTHrP receptor indicating the presence of seven membrane-spanning helices, which surround a central polar cavity. The receptor contains a large, glycosylated N-terminal extracellular domain and a long C-terminal cytoplasmic tail. Agonist binding to the receptor alters the relative orientation of the transmembrane helices, promoting the activation of specific G-proteins. See text for details.
C, resulting in the mobilization of intracellular calcium and activation of PK-C. Preference of the PTH/PTHrP receptor for the cyclic AMP signaling pathway is suggested by studies on PTH target cells in vitro, where the activation of adenylyl cyclase generally occurs at lower concentrations of added PTH than activation of phospholipase C [278]. These findings are consistent with the observations that the cyclic AMP pathway is most closely associated with most of the physiological effects of PTH on bone and kidney, with activation of phospholipase C playing a modulatory role. 2. RECEPTOR ACTIVATION MECHANISMS When the cDNA sequence of the PTH/PTHrP receptor was first delineated [275], it was apparent that it encoded a protein with a predicted overall structure consistent with those of other known GPCRs. In particular, the receptor was modeled as containing seven membrane-spanning helices, with a large amino-terminal extracellular domain, three extracellular loops, three intracellular loops, and a large carboxy-terminal cytoplasmic tail (Fig. 7). However, the PTH/PTHrP receptor does not share a number of the specific sequence motifs present in the largest subfamily of GPCRs (the so-called class I family that includes receptors
for a diverse group of ligands ranging from photons to polypeptide hormones). Rather, the PTH/PTHrP receptor is a member of a second GPCR subfamily (class II) that includes receptors for calcitonin, glucagon, and a number of other polypeptide ligands. Members of the class II GPCR subfamily are presumed to share a common basic mechanism of G-protein activation, but have evolved determinants of specificity that permit binding and activation by only the appropriate peptide ligand. Mutagenesis studies have been carried out to investigate the structural features in the PTH/PTHrP receptor that are important for agonist binding and for maintaining receptor specificity. These studies have demonstrated that the large amino-terminal extracellular domain of the receptor contains critical determinants of agonist binding affinity [279 – 281]. However, the body of the receptor, which includes the extracellular loops and the transmembrane domains, also plays a role in ligand binding as well as in maintaining ligand specificity [280,282 – 284]. In a recent series of elegant biochemical studies, sites of interaction between amino-terminal PTH fragments and the PTH/PTHrP receptor have been mapped. These studies have demonstrated multiple points of contact between the 1 – 34 region of PTH/PTHrP and the receptor. Specifically, there is interaction between position 23 in the ligand and the extreme N terminus of the extracellular domain of the receptor [285]; between amino 13 of the ligand and the membrane-proximal portion of the N terminal extracellular domain of the receptor [286]; and between the N-terminus of the ligand and the extracellular end of the sixth transmembrane domain of the receptor [287]. This latter interaction is presumably required to initiate the conformational shift in the transmembrane domain of the receptor that is required for signal transduction [288]. This involves the exposure of key amino acids in the second and third cytoplasmic loops of the PTH/PTHrP receptor that are required for activation of Gs and Gq [289,290]. 3. RECEPTOR REGULATION Signal transduction by GPCRs is generally subject to strict regulatory control. This control can occur in response to agonist binding (homologous regulation) or in response to factors acting though separate pathways (heterologous regulation). Acute control of signaling is accomplished by blocking the ability of agonist-occupied receptors to sustain the activation of G-proteins (desensitization) and by physically moving the receptors into an intracellular compartment effectively separating them from G-proteins (sequestration). Chronic regulation of receptor signaling is accomplished by agonist-induced changes in steady-state levels of expression of receptors due to increased receptor catabolism following receptor internalization (downregulation) and to changes in de novo receptor synthesis. Homologous regulation commonly involves all of these mechanisms, whereas
234 heterologous regulation most often occurs through changes in steady-state levels of receptor expression. Many studies have documented homologous regulation of PTH/PTHrP receptor signaling. Treatment of cultured bone and kidney cells with PTH generally dampens the adenylyl cyclase and phospholipase C responses to a second addition of the hormone [291 – 300]. In most studies, desensitization of the PTH response occurs rapidly, within minutes of initial exposure to PTH, suggesting that the PTH/PTHrP receptor has become acutely uncoupled from its cognate G-proteins. The mechanisms underlying acute desensitization have been well studied for GPCRs such as rhodopsin and -adrenergic receptors [301 – 303]. The major mechanism underlying acute desensitization of these receptors is phosphorylation of the cytoplasmic domain of the receptor by a GPCR kinase (GRK). GRKs are serine/threonine kinases that phosphorylate only the agonist-occupied receptor, and phosphorylation facilitates the interaction of the receptor with a member of the arrestin protein family. Arrestin binding to the receptor sterically interferes with the interaction between the receptor and G-proteins, thus preventing signal transmission. There is mounting evidence that a similar mechanism applies to desensitization of PTH/PTHrP receptor signaling. The PTH/PTHrP receptor is subject to phosphorylation in response to agonist binding [304,305], and this appears to occur largely if not exclusively on serine residues in the cytoplasmic tail [305 – 307]. The kinase involved appears to be a member of the GRK family, possibly GRK2 [306,308], and a dominant-inhibitor of GRK function can suppress PTH/PTHrP receptor desensitization in human osteoblast-like cells [309]. Long-term treatment with PTH results in a loss of cellular PTH/PTHrP receptors (downregulation) and a corresponding reduction in the maximal signaling response to the hormone [299,310 – 313]. Evidence shows that this process may have pathophysiological relevance. For example, vitamin D deficiency can be associated with target cell resistance to PTH [314 – 316]. In animal studies, this resistance can be reversed by parathyroidectomy, suggesting that it is the secondary hyperparathyroidism that is responsible for target cell resistance [317]. Infusion of PTH to levels seen in severe secondary hyperparathyoidism produces downregulation of PTH/PTHrP receptors and a reduction in the adenylyl cyclase response to PTH [310]. In chronic renal failure, factors other than hyperparathyroidism may also contribute to reduced target cell expression of PTH/PTHrP receptors [318]. The initial step in downregulation of PTH/PTHrP receptors appears to be agonist-induced accumulation of the receptor in plasma membrane clathrincoated pits [31,319]. These pits are endocytic organelles that pinch off from the plasma membrane, thus becoming endocytic vesicles. Once internalized, PTH/PTHrP receptors can be recycled to the plasma membrane or can presumably progress further down the endocytic pathway to
ROBERT A. NISSENSON
the lysosomes for degradation. The molecular mechanisms underlying the agonist-induced internalization of the PTH/PTHrP receptor are not entirely clear. Agonist-stimulated receptor phosphorylation may play a role in osteoblastic cells [309], although receptor phosphorylation is not required for endocytosis in some cellular settings [307]. Arrestins have been implicated as mediators of GPCR endocytosis, and studies indicate that arrestins can become associated with the PTH/PTHrP receptor following agonist binding [320]. In addition, the cytoplasmic tail of the PTH/PTHrP receptor contains a tyrosine-based sequence that has been implicated in promoting the internalization of other membrane receptors. Mutation of this sequence markedly inhibits agonist-induced PTH/PTHrP receptor endocytosis [319]. Further studies are needed to define the relative contribution of these mechanisms to PTH/PTHrP receptor internalization and downregulation. Another mechanism for the regulation of PTH/PTHrP receptor levels is through changes in expression of the receptor gene. In osteoblastic cells, PTH is reported to decrease levels of PTH/PTHrP receptor mRNA by a mechanism involving the cyclic AMP pathway [321,322]. This may be due to direct transcriptional activation of the PTH/PTHrP receptor gene by PK-A-activated transcription factors, but the details of this pathway have yet to be elucidated. Homologous control of PTH/PTHrP receptor expression appears to be target cell specific in that PTH reportedly does not reduce expression of the PTH/PTHrP receptor gene in the kidneys of rats with secondary hyperparathyroidism [318,323]. Heterologous factors are also reported to regulate levels of PTH/PTHrP receptor expression in bone and kidney. The cytokine TGF upregulates the expression of the PTH/PTHrP receptor in osteoblastic osteosarcoma cells [324], although the opposite effect is reported in primary cultures of fetal rat osteoblasts [325] and in OK cells [326]. Dexamethasone treatment produces an increase in expression of the PTH/PTHrP receptor in osteoblastic cells, but not in kidney cells [327,328], whereas 1,25(OH)2D downregulates expression of the PTH/PTHrP receptor gene [329]. It should be pointed out that most of these studies have been carried out in cultured bone and kidney cells in vitro, and more needs to be done to establish the physiological relevance of the changes in PTH/PTHrP receptor gene expression.
C. Nontraditional Mechanisms of Action of PTHrP The discovery of PTHrP was based on the PTH-like endocrine actions of this peptide in patients with malignancyassociated hypercalcemia. The classical mechanism of action of PTHrP is thus to bind to and activate the widely expressed PTH/PTHrP receptor. The amino-terminal 1 – 34 domain of
235
CHAPTER 7 PTH and PTH-Related Protein
FIGURE 8
Potential mechanisms of action of PTHrP. The majority of the actions of PTHrP result from the binding of the amino-terminal portion of the protein to the PTH/PTHrP receptor, leading to the activation of adenylyl cyclase (AC) and phospholipase C (PLC). Activation of these effector enzymes results in increased cellular levels of cyclic AMP (cAMP), intracellular calcium, and protein kinase C (PKC). PTHrP is also processed posttranslationally, producing midregion and carboxyl-terminal fragments of the protein. These fragments have cellular effects that are presumably mediated by novel membrane receptors (Rm and Rc), acting through unknown signaling pathways. PTHrP has also been localized to the nucleus of cells (intracrine action) where it may regulate nuclear functions such as mitosis, apoptosis, and RNA processing.
PTHrP is responsible for binding to the PTH/PTHrP receptor, thus initiating signal transduction. However, it appears that the PTH/PTHrP receptor does not mediate all of the physiological actions of PTHrP. Two additional mechanisms have been identified by which PTHrP can potentially influence cellular function (Fig. 8). One involves the notion of PTHrP as a polyhormone that yields mid- and carboxyl-region fragments with distinct biological activities that are presumably mediated by novel cell surface receptors. The second mechanism relates to the ability of PTHrP to translocate to the nucleus of cells in which it is expressed, thereby altering cell proliferation and/or gene expression. 1. PTHRP AS A POLYHORMONE The PTHrP gene is subject to alternative splicing, resulting in multiple protein products (ranging from 139 to 173
amino acids) that differ only in the extent of their C termini [142]. Only the amino-terminal 34 amino acids are needed to produce all of the PTH-like actions of PTHrP on the PTH/PTHrP receptor, and several groups have been interested in assessing a possible biological role for the remainder of the molecule. Indeed, PTHrP is subject to posttranslational proteolytic processing to produce a midregion fragment (amino acids 38 – 94) and a C-terminal fragment (amino acids 107 – 139) as well as PTHrP(1 – 36) [140]. Fragments of PTHrP are secreted by some cells, at least in vitro, and thus have the potential to elicit biological responses in a paracrine or endocrine fashion. Synthetic PTHrP(107 – 139) has been reported to elicit biological effects such as inhibition of bone resorption [330], stimulation of osteoblast proliferation [331], and stimulation of interleukin-6 expression in osteoblasts [332]. The nature of the receptor and signaling pathway responsible for these actions of PTHrP is unclear, although the latter effect appeared to involve activation of protein kinase C. A physiological role for PTHrP fragments is suggested by studies of placental calcium transport. As mentioned earlier, the fetal parathyroid gland is required to maintain the normal positive maternal – fetal calcium gradient, at least in sheep. This gradient can be restored in parathyroidectomized fetuses by the administration of midregion fragments of PTHrP, but not by PTH or by amino-terminal PTHrP fragments [196,333]. This effect must therefore be initiated by a receptor distinct from the classical PTH/ PTHrP receptor. 2. INTRACRINE ACTIONS OF PTHRP Several studies have demonstrated that, once synthesized, PTHrP can localize to the nucleolus as well as being secreted. Nucleolar localization requires the presence of a targeting signal in the carboxyl region of the molecule [334] and occurs through an interaction with the targeting protein importin [335]. Secreted PTHrP can also be taken up by cells and translocated to the nucleus, and this appears to involve a receptor distinct from the PTH/PTHrP receptor [336]. Although the functional significance of nuclear PTHrP has yet to be definitely established, a number of intriguing findings have been reported. Intracellular expression of PTHrP has been shown to protect chondrocytes from apoptosis induced by serum deprivation, and this effect was dependent on the presence of an intact nucleolar localization signal [337]. Nuclear localization of PTHrP is associated with mitogenesis in cultured vascular smooth muscle cells [338]. The mechanisms underlying this mitogenic effect are not entirely clear. In cultured keratinocytes, PTHrP is present in the nucleolus during the G1 phase of the cell cycle, but redistributes to the cytoplasm during cell division [339]. Interestingly, PTHrP is phosphorylated by the cell cycle regulatory kinase CDC2-CDK2, and this appears to promote translocation of the PTHrP from the
236
ROBERT A. NISSENSON
nucleus to the cytoplasm [340]. It is possible that PTHrP acts, at least in part, through direct interaction with ribonucleoprotein complexes, as PTHrP is capable of binding directly to RNA via a polybasic region within the nuclear localization signal [336]. Further work is needed to more clearly define the significance of this unusual mode of action of PTHrP.
Acknowledgments I am grateful to Margaret Bencsik for her skillful assistance in the preparation of this manuscript. Portions of the work discussed here were supported by NIH Grant DK35323 and by the Medical Research Service of the Department of Veterans’ Affairs. Dr. Nissenson is a Senior Research Career Scientist of the Department of Veterans’ Affairs.
References 1. E. M. Brown, Extracellular Ca2 sensing, regulation of parathyroid cell function, and role of Ca2 and other ions as extracellular (first) messengers. Physiol. Rev. 71, 371 – 411 (1991). 2. J. Russell, D. Lettieri, and L. M. Sherwood, Direct regulation by calcium of cytoplasmic messenger ribonucleic acid coding for preproparathyroid hormone in isolated bovine parathyroid cells. J. Clin. Invest. 72, 1851 – 1855 (1983). 3. T. Naveh-Many and J. Silver, Regulation of parathyroid hormone gene expression by hypocalcemia, hypercalcemia, and vitamin D in the rat. J. Clin. Invest. 86, 1313 – 1319 (1990). 4. T. Yamaguchi, N. Chattopadhyay, and E. M. Brown, G protein-coupled extracellular Ca2 (Ca2o)-sensing receptor (CaR): Roles in cell signaling and control of diverse cellular functions. Adv. Pharmacol. 47, 209 – 253 (2000). 5. E. M. Brown, Physiology and pathophysiology of the extracellular calcium-sensing receptor. Am. J. Med. 106, 238 – 253 (1999). 6. M. R. Pollak, C. E. Seidman, and E. M. Brown, Three inherited disorders of calcium sensing. Medicine 75, 115 – 123 (1996). 7. S. A. Berson and R. S. Yalow, Immunochemical heterogeneity of parathyroid hormone in plasma. J. Clin. Endocrino. Metab. 28, 1037 – 1047 (1968). 8. J. M. Canterbury and E. Reiss, Fractionation of circulating parathyroid hormone (PTH) in man. J. Lab. Clin. Med. 78, 814 (1971). 9. C. D. Arnaud, Immunochemical heterogeneity of circulating parathyroid hormone in man: Sequel to an original observation by Berson and Yalow. Mount Sinai J. Med. 40, 422 – 432 (1973). 10. S. B. Oldham, E. J. Finck, and F. R. Singer, Parathyroid hormone clearance in man. Metab. Clin. Exp. 27, 993 – 1001 (1978). 11. G. V. Segre, H. D. Niall, J. F. Habener, and J. T. Potts, Jr., Metabolism of parathyroid hormone: Physiologic and clinical significance. Am. J. Med. 56, 774 – 784 (1974). 12. J. Fox, M. Scott, R. A. Nissenson, and H. D. Heath, Effect of plasma calcium concentration on the metabolic clearance rate of parathyroid hormone in the dog. J. Lab. Clin. Med. 102, 70 – 77 (1983). 13. S. K. Libutti, H. R. Alexander, D. L. Bartlett, M. L. Sampson, M. E. Ruddel, M. Skarulis, S. J. Marx, A. M. Spiegel, W. Simmonds, and A. T. Remaley, Kinetic analysis of the rapid intraoperative parathyroid hormone assay in patients during operation for hyperparathyroidism. Surgery 126, 1145 – 1150 (1999).
14. K. J. Martin, K. A. Hruska, J. J. Freitag, S. Klahr, and E. Slatopolsky, The peripheral metabolism of parathyroid hormone. N. Engl. J. Med. 301, 1092 – 1098 (1979). 15. K. Martin, K. Hruska, A. Greenwalt, S. Klahr, and E. Slatopolsky, Selective uptake of intact parathyroid hormone by the liver: Differences between hepatic and renal uptake. J. Clin. Invest. 58, 781 – 788 (1976). 16. K. A. Hruska, A. Korkor, K. Martin, and E. Slatopolsky, Peripheral metabolism of intact parathyroid hormone: Role of liver and kidney and the effect of chronic renal failure. J. Clin. Invest. 67, 885 – 892 (1981). 17. F. R. Bringhurst, A. M. Stern, M. Yotts, N. Mizrahi, G. V. Segre, and J. T. Potts, Jr., Peripheral metabolism of PTH: Fate of biologically active amino terminus in vivo. Am. J. Physiol. 255, E886 – E893 (1988). 18. J. A. Flueck, F. P. Di Bella, A. J. Edis, J. M. Kehrwald, and C. D. Arnaud, Immunoheterogeneity of parathyroid hormone in venous effluent serum from hyperfunctioning parathyroid glands. J. Clin. Invest. 60, 1367 – 1375 (1977). 19. S. Erdmann, W. Müller, S. Bahrami, S. I. Vornehm, H. Mayer, P. Bruckner, K. von der Mark, and H. Burkhardt, Differential effects of parathyroid hormone fragments on collagen gene expression in chondrocytes. J. Cell Biol. 135, 1179 – 1191 (1996). 20. M. Nasu, T. Sugimoto, H. Kaji, J. Kano, and K. Chihara, Carboxylterminal parathyroid hormone fragments stimulate type – 1 procollagen and insulin-like growth factor-binding protein – 5 mRNA expression in osteoblastic UMR-106 cells. Endoc. J. 45, 229 – 234 (1998). 21. T. Tsuboi and A. Togari, Comparison of the effects of carboxylterminal parathyroid hormone peptide[53 – 84] and aminoterminal peptide[1-34] on mouse tooth germ in vitro. Arch. Oral Bio. 43, 335 – 339 (1998). 22. N. Inomata, M. Akiyama, N. Kubota, and H. Jüppner, Characterization of a novel parathyroid hormone (PTH) receptor with specificity for the carboxyl-terminal region of PTH-(1-84). Endocrinology 136, 4732 – 4740 (1995). 23. H. Takasu, H. Baba, N. Inomata, Y. Uchiyama, N. Kubota, K. Kumaki, A. Matsumoto, K. Nakajima, T. Kimura, S. Sakakibara, T. Fujita, K. Chihara, and I. Nagai, The 69 – 84 amino acid region of the parathyroid hormone molecule is essential for the interaction of the hormone with the binding sites with carboxyl-terminal specificity. Endocrinology 137, 5537 – 5543 (1996). 24. R. R. MacGregor, J. W. Hamilton, G. N. Kent, R. E. Shofstall, and D. V. Cohn, The degradation of proparathormone and parathormone by parathyroid and liver cathepsin B. J. Biol. Chem. 254, 4428 – 4433 (1979). 25. J. W. Hamilton, R. L. Jilka, and R. R. MacGregor, Cleavage of parathyroid hormone to the 1-34 and 35-84 fragments by cathepsin D-like activity in bovine parathyroid gland extracts. Endocrinology 113, 285 – 292 (1983). 26. G. P. Mayer, J. A. Keaton, J. G. Hurst, and J. F. Habener, Effects of plasma calcium concentration on the relative proportion of hormone and carboxyl fragments in parathyroid venous blood. Endocrinology 104, 1778 – 1784 (1979). 27. R. Talmage, The demand for bone calcium in mainenance of plasma calcium concentration. In “Mechanisms of Localized Bone Loss” (R. Horton, T. M. Tarplay, W. F. Davis, eds.), pp. 73 – 92. Info Retrieval, Washington, DC, 1978. 28. R. A. M. Talmage, Physiological role of parathyroid hormone. In “Handbook of Physiology” (R. A. A. Grup, ed.), pp. 343 – 351. American Physiological Society, Washington, DC, 1967. 29. W. F. Neuman, M. W. Neuman, and C. R. Myers, Blood:bone disequilibrium. III. Linkage between cell energetics and Ca fluxes. Am. J. Physiol. 236, C244 – C248 (1979). 30. M. E. Holtrop and G. J. King, The ultrastructure of the osteoclast and its functional implications. Clin. Orthop. Rela. Res. 475, 177 – 196 (1977). 31. C. M. Silve, G. T. Hradek, A. L. Jones, and C. D. Arnaud, Parathyroid hormone receptor in intact embryonic chicken bone: Characterization and cellular localization. J. Cell Biol. 94, 379 – 386 (1982).
CHAPTER 7 PTH and PTH-Related Protein 32. M. F. Rouleau, H. Warshawsky, and D. Goltzman, Parathyroid hormone binding in vivo to renal, hepatic, and skeletal tissues of the rat using a radioautographic approach. Endocrinology 118, 919 – 931 (1986). 33. A. Teti, R. Rizzoli, and A. Zambonin Zallone, Parathyroid hormone binding to cultured avian osteoclasts. Biochem. Biophys. Res. Commun. 174, 1217 – 1222 (1991). 34. N. Agarwala and C. V. Gay, Specific binding of parathyroid hormone to living osteoclasts. J. Bone Miner. Res. 7, 531 – 539 (1992). 35. L. G. Rao, T. M. Murray, and J. N. Heersche, Immunohistochemical demonstration of parathyroid hormone binding to specific cell types in fixed rat bone tissue. Endocrinology 113, 805 – 10 (1983). 36. T. J. Chambers, P. M. McSheehy, B. M. Thomson, and K. Fuller, The effect of calcium-regulating hormones and prostaglandins on bone resorption by osteoclasts disaggregated from neonatal rabbit bones. Endocrinology 116, 234 – 239 (1985). 37. R. L. Jilka, Are osteoblastic cells required for the control of osteoclast activity by parathyroid hormone? Bone Miner. 1, 261 – 266 (1986). 38. S. S. Miller, A. M. Wolf, and C. D. Arnaud, Bone cells in culture: Morphologic transformation by hormones. Science 192, 1340 – 1343 (1976). 39. R. Civitelli, K. Ziambaras, P. M. Warlow, F. Lecanda, T. Nelson, J. Harley, N. Atal, E. C. Beyer, and T. H. Steinberg, Regulation of connexin43 expression and function by prostaglandin E2 (PGE2) and parathyroid hormone (PTH) in osteoblastic cells. J. Cell. Biochem. 68, 8 – 21 (1998). 40. P. C. Schiller, B. A. Roos, and G. A. Howard, Parathyroid hormone up-regulation of connexin 43 gene expression in osteoblasts depends on cell phenotype. J. Bone Miner. Res. 12, 2005 – 2013 (1997). 41. H. J. Donahue, K. J. McLeod, C. T. Rubin, J. Andersen, E. A. Grine, E. L. Hertzberg, and P. R. Brink, Cell-to-cell communication in osteoblastic networks: Cell line-dependent hormonal regulation of gap junction function. J. Bone Miner. Res. 10, 881 – 889 (1995). 42. J. E. Puzas and J. S. Brand, Parathyroid hormone stimulation of collagenase secretion by isolated bone cells. Endocrinology 104, 559 – 562 (1979). 43. J. K. Heath, S. J. Atkinson, M. C. Meikle, and J. J. Reynolds, Mouse osteoblasts synthesize collagenase in response to bone resorbing agents. Biochim. Biophys. Acta 802, 151 – 154 (1984). 44. Y. Eeckhout, J. M. Delaissé, and G. Vaes, Direct extraction and assay of bone tissue collagenase and its relation to parathyroid-hormoneinduced bone resorption. Biochem. J. 239, 793 – 796, (1986). 45. N. C. Partridge, J. J. Jeffrey, L. S. Ehlich, S. L. Teitelbaum, C. Fliszar, H. G. Welgus, and A. J. Kahn, Hormonal regulation of the production of collagenase and a collagenase inhibitor activity by rat osteogenic sarcoma cells. Endocrinology 120, 1956 – 1962 (1987). 46. R. L. Jilka, Procollagenase associated with the noncalcified matrix of bone and its regulation by parathyroid hormone. Bone 10, 353 – 358 (1989). 47. J. A. Hamilton, S. Lingelbach, N. C. Partridge, and T. J. Martin, Regulation of plasminogen activator production by bone-resorbing hormones in normal and malignant osteoblasts. Endocrinology 116, 2186 – 2191 (1985). 48. G. Leloup, C. Peeters-Joris, J. M. Delaisse, G. Opdenakker, and G. Vaes, Tissue and urokinase plasminogen activators in bone tissue and their regulation by parathyroid hormone. J. Bone Miner. Res. 6, 1081 – 1090 (1991). 49. T. Nonaka, H. Matsumoto, W. Shimada, K. Okada, H. Fukao, S. Ueshima, H. Kikuchi, S. Tanaka, and O. Matsuo, Effect of bone resorbing factors on u-PA and its specific receptor in osteosarcoma cell line. Clin. Chim. Acta 223, 129 – 142 (1993). 50. W. Zhao, M. H. Byrne, B. F. Boyce, and S. M. Krane, Bone resorption induced by parathyroid hormone is strikingly diminished in collagenase-resistant mutant mice. J. Clin. Invest. 103, 517 – 524 (1999).
237 51. M. C. de Vernejoul, M. Horowitz, J. Demignon, L. Neff, and R. Baron, Bone resorption by isolated chick osteoclasts in culture is stimulated by murine spleen cell supernatant fluids (osteoclastactivating factor) and inhibited by calcitonin and prostaglandin E2. J. Bone Miner. Res. 3, 69 – 80 (1988). 52. S. Amano, S. Hanazawa, K. Hirose, Y. Ohmori, and S. Kitano, Stimulatory effect on bone resorption of interleukin-1-like cytokine produced by an osteoblast-rich population of mouse calvarial cells. Calcif. Tissue Int. 43, 88 – 91 (1988). 53. M. C. Horowitz, D. L. Coleman, P. M. Flood, T. S. Kupper, and R. L. Jilka, Parathyroid hormone and lipopolysaccharide induce murine osteoblast-like cells to secrete a cytokine indistinguishable from granulocyte-macrophage colony-stimulating factor. J. Clin. Invest. 83, 149 – 157 (1989). 54. D. B. Evans, R. A. Bunning, J. Van Damme, and R. G. Russell, Natural human IL-1 beta exhibits regulatory actions on human bonederived cells in vitro. Biochem. Biophys. Res. Commun. 159, 1242 – 1248 (1989). 55. E. Jimi, I. Nakamura, H. Amano, Y. Taguchi, T. Tsurukai, M. Tamura, N. Takahashi, and T. Suda, Osteoclast function is activated by osteoblastic cells through a mechanism involving cellto-cell contact [corrected and republished with original paging, article originally printed in Endocrinology 137(5), 2187 – 2190 (1996)]. Endocrinology 137, 2187 – 2190 (1996). 56. T. Suda, N. Takahashi, N. Udagawa, E. Jimi, M. T. Gillespie, and T. J. Martin, Modulation of osteoclast differentiation and function by the new members of the tumor necrosis factor receptor and ligand families. Endocr. Rev. 20, 345 – 357 (1999). 57. L. C. Hofbauer, S. Khosla, C. R. Dunstan, D. L. Lacey, W. J. Boyle, and B. L. Riggs, The roles of osteoprotegerin and osteoprotegerin ligand in the paracrine regulation of bone resorption. J. Bone Miner. Res. 15, 2 – 12 (2000). 58. D. L. Lacey, E. Timms, H. L. Tan, M. J. Kelley, C. R. Dunstan, T. Burgess, R. Elliott, A. Colombero, G. Elliott, S. Scully, H. Hsu, J. Sullivan, N. Hawkins, E. Davy, C. Capparelli, A. Eli, Y. X. Qian, S. Kaufman, I. Sarosi, V. Shalhoub, G. Senaldi, J. Guo, J. Delaney, and W. J. Boyle, Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell 93, 165 – 176 (1998). 59. T. L. Burgess, Y. Qian, S. Kaufman, B. D. Ring, G. Van, C. Capparelli, M. Kelley, H. Hsu, W. J. Boyle, C. R. Dunstan, S. Hu, and D. L. Lacey, The ligand for osteoprotegerin (OPGL) directly activates mature osteoclasts. J. Cell Biol. 145, 527 – 538 (1999). 60. H. Yasuda, N. Shima, N. Nakagawa, K. Yamaguchi, M. Kinosaki, S. Mochizuki, A. Tomoyasu, K. Yano, M. Goto, A. Murakami, E. Tsuda, T. Morinaga, K. Higashio, N. Udagawa, N. Takahashi, and T. Suda, Osteoclast differentiation factor is a ligand for osteoprotegerin/ osteoclastogenesis-inhibitory factor and is identical to TRANCE/ RANKL. Proc. Nat. Acad. Sci. USA 95, 3597 – 3602 (1998). 61. J. M. Quinn, J. Elliott, M. T. Gillespie, and T. J. Martin, A combination of osteoclast differentiation factor and macrophage-colony stimulating factor is sufficient for both human and mouse osteoclast formation in vitro. Endocrinology 139, 4424 – 4427 (1998). 62. Y. Y. Kong, H. Yoshida, I. Sarosi, H. L. Tan, E. Timms, C. Capparelli, S. Morony, A. J. Oliveira-dos-Santos, G. Van, A. Itie, W. Khoo, A. Wakeham, C. R. Dunstan, D. L. Lacey, T. W. Mak, W. J. Boyle, and J. M. Penninger, OPGL is a key regulator of osteoclastogenesis, lymphocyte development and lymph-node organogenesis. Nature 397, 315 – 323 (1999). 63. W. S. Simonet, D. L. Lacey, C. R. Dunstan, M. Kelley, M. S. Chang, R. Lüthy, H. Q. Nguyen, S. Wooden, L. Bennett, T. Boone, G. Shimamoto, M. DeRose, R. Elliott, A. Colombero, H. L. Tan, G. Trail, J. Sullivan, E. Davy, N. Bucay, L. Renshaw-Gegg, T. M. Hughes, D. Hill, W. Pattison, P. Campbell, W. J. Boyle et al., Osteoprotegerin: A novel secreted protein involved in the regulation of bone density. Cell 89, 309 – 319 (1997).
238 64. H. Yasuda, N. Shima, N. Nakagawa, S. I. Mochizuki, K. Yano, N. Fujise, Y. Sato, M. Goto, K. Yamaguchi, M. Kuriyama, T. Kanno, A. Murakami, E. Tsuda, T. Morinaga, and K. Higashio, Identity of osteoclastogenesis inhibitory factor (OCIF) and osteoprotegerin (OPG): A mechanism by which OPG/OCIF inhibits osteoclastogenesis in vitro. Endocrinology 139, 1329 – 1337 (1998). 65. E. Tsuda, M. Goto, S. Mochizuki, K. Yano, F. Kobayashi, T. Morinaga, and K. Higashio, Isolation of a novel cytokine from human fibroblasts that specifically inhibits osteoclastogenesis. Biochem. Biophys. Res. Commun. 234, 137 – 142 (1997). 66. N. Bucay, I. Sarosi, C. R. Dunstan, S. Morony, J. Tarpley, C. Capparelli, S. Scully, H. L. Tan, W. Xu, D. L. Lacey, W. J. Boyle, and W. S. Simonet, Osteoprotegerin-deficient mice develop early onset osteoporosis and arterial calcification. Genes Dev. 12, 1260 – 1268 (1998). 67. A. Mizuno, N. Amizuka, K. Irie, A. Murakami, N. Fujise, T. Kanno, Y. Sato, N. Nakagawa, H. Yasuda, S. Mochizuki, T. Gomibuchi, K. Yano, N. Shima, N. Washida, E. Tsuda, T. Morinaga, K. Higashio, and H. Ozawa, Severe osteoporosis in mice lacking osteoclastogenesis inhibitory factor/osteoprotegerin. Biochem. Biophys. Res. Commun. 247, 610 – 615 (1998). 68. K. Tsukii, N. Shima, S. Mochizuki, K. Yamaguchi, M. Kinosaki, K. Yano, O. Shibata, N. Udagawa, H. Yasuda, T. Suda, and K. Higashio, Osteoclast differentiation factor mediates an essential signal for bone resorption induced by 1 alpha,25-dihydroxyvitamin D3, prostaglandin E2, or parathyroid hormone in the microenvironment of bone. Biochem. Biophys. Res. Commun. 246, 337 – 341 (1998). 69. B. S. Kwon, S. Wang, N. Udagawa, V. Haridas, Z. H. Lee, K. K. Kim, K. O. Oh, J. Greene, Y. Li, J. Su, R. Gentz, B. B. Aggarwal, and J. Ni, TR1, a new member of the tumor necrosis factor receptor superfamily, induces fibroblast proliferation and inhibits osteoclastogenesis and bone resorption. FASEB J. 12, 845 – 854 (1998). 70. S. K. Lee and J. A. Lorenzo, Parathyroid hormone stimulates TRANCE and inhibits osteoprotegerin messenger ribonucleic acid expression in murine bone marrow cultures: Correlation with osteoclast-like cell formation. Endocrinology 140, 3552 – 3561 (1999). 71. N. J. Horwood, J. Elliott, T. J. Martin, and M. T. Gillespie, Osteotropic agents regulate the expression of osteoclast differentiation factor and osteoprotegerin in osteoblastic stromal cells. Endocrinology 139, 4743 – 4746 (1998). 72. T. Murakami, M. Yamamoto, K. Ono, M. Nishikawa, N. Nagata, K. Motoyoshi, and T. Akatsu, Transforming growth factor-beta1 increases mRNA levels of osteoclastogenesis inhibitory factor in osteoblastic/stromal cells and inhibits the survival of murine osteoclastlike cells. Biochem. Biophys. Res. Commun. 252, 747 – 752 (1998). 73. A. Grey, M. A. Mitnick, U. Masiukiewicz, B. H. Sun, S. Rudikoff, R. L. Jilka, S. C. Manolagas, and K. Insogna, A role for interleukin-6 in parathyroid hormone-induced bone resorption in vivo. Endocrinology 140, 4683 – 4690 (1999). 74. H. Kaji, T. Sugimoto, M. Kanatani, K. Nishiyama, M. Nasu, and K. Chihara, Insulin-like growth factor-I mediates osteoclast-like cell formation stimulated by parathyroid hormone. J. Cell. Physiol. 172, 55 – 62 (1997). 75. T. Yonaga, Action of parathyroid hormone, with special reference to its anabolic effect on different kinds of tissues in rats (I). Bull. Tokyo Med. Dent. Univ. 25, 237 – 248 (1978). 76. J. Reeve, P. J. Meunier, J. A. Parsons, M. Bernat, O. L. Bijvoet, P. Courpron, C. Edouard, L. Klenerman, R. M. Neer, J. C. Renier, D. Slovik, F. J. Vismans, and J. T. Potts, Jr, Anabolic effect of human parathyroid hormone fragment on trabecular bone in involutional osteoporosis: A multicentre trial. Br. Med. J. 280, 1340 – 1344 (1980). 77. W. B. High, H. E. Black, and C. C. Capen, Histomorphometric evaluation of the effects of low dose parathyroid hormone administration on cortical bone remodeling in adult dogs. Lab. Invest. 44, 449 – 454 (1981).
ROBERT A. NISSENSON 78. M. Gunness-Hey and J. M. Hock, Increased trabecular bone mass in rats treated with human synthetic parathyroid hormone. Metab. Bone Dis. Relat. Res. 5, 177 – 181 (1984). 79. T. J. Wronski, C. F. Yen, H. Qi, and L. M. Dann, Parathyroid hormone is more effective than estrogen or bisphosphonates for restoration of lost bone mass in ovariectomized rats. Endocrinology 132, 823 – 831 (1993). 80. C. C. Liu and D. N. Kalu, Human parathyroid hormone-(1 – 34) prevents bone loss and augments bone formation in sexually mature ovariectomized rats. J. Bone Miner. Res. 5, 973 – 982 (1990). 81. J. F. Whitfield, P. Morley, and G. E. Willick, The bone-building action of the parathyroid hormone: Implications for the treatment of osteoporosis. Drugs Aging 15, 117 – 129 (1999). 82. D. W. Dempster, F. Cosman, M. Parisien, V. Shen, and R. Lindsay, Anabolic actions of parathyroid hormone on bone [published erratum appears in Endocr. Rev. 15(2), 261 (1994)]. Endocr. Rev. 14, 690 – 709 (1993). 83. T. Hirano, D. B. Burr, C. H. Turner, M. Sato, R. L. Cain, and J. M. Hock, Anabolic effects of human biosynthetic parathyroid hormone fragment (1 – 34), LY333334, on remodeling and mechanical properties of cortical bone in rabbits. J. Bone Miner. Res. 14, 536 – 545 (1999). 84. L. Zhang, N. Endo, N. Yamamoto, T. Tanizawa, and H. E. Takahashi, Effects of single and concurrent intermittent administration of human PTH (1 – 34) and incadronate on cancellous and cortical bone of femoral neck in ovariectomized rats. Tohoku J. Exp. Med. 186, 131 – 141 (1998). 85. M. Sato, G. Q. Zeng, and C. H. Turner, Biosynthetic human parathyroid hormone (1 – 34) effects on bone quality in aged ovariectomized rats. Endocrinology 138, 4330 – 4337 (1997). 86. B. D. Baumann and T. J. Wronski, Response of cortical bone to antiresorptive agents and parathyroid hormone in aged ovariectomized rats. Bone 16, 247 – 253 (1995). 87. N. E. Lane, D. B. Kimmel, M. H. Nilsson, F. E. Cohen, S. Newton, R. A. Nissenson, and G. J. Strewler, Bone-selective analogs of human PTH(1 – 34) increase bone formation in an ovariectomized rat model. J. Bone Miner. Res. 11, 614 – 625 (1996). 88. N. E. Lane, J. M. Thompson, G. J. Strewler, and J. H. Kinney, Intermittent treatment with human parathyroid hormone (hPTH[1 – 34]) increased trabecular bone volume but not connectivity in osteopenic rats. J. Bone Miner. Res. 10, 1470 – 1477 (1995). 89. L. Mosekilde, C. H. Søgaard, C. C. Danielsen and O. Tørring, The anabolic effects of human parathyroid hormone (hPTH) on rat vertebral body mass are also reflected in the quality of bone, assessed by biomechanical testing: a comparison study between hPTH-(1 – 34) and hPTH-(1 – 84). Endocrinology 129, 421 – 428 (1991). 90. K. C. Hicok, T. Thomas, F. Gori, D. J. Rickard, T. C. Spelsberg, and B. L. Riggs, Development and characterization of conditionally immortalized osteoblast precursor cell lines from human bone marrow stroma. J. Bone Miner. Res. 13, 205 – 217, (1998). 91. R. W. Zhang, S. C. Supowit, X. Xu, H. Li, M. D. Christensen, R. Lozano, and D. J. Simmons, Expression of selected osteogenic markers in the fibroblast-like cells of rat marrow stroma. Calcif. Tissue Int. 56, 283 – 291, (1995). 92. M. F. Rouleau, J. Mitchell, and D. Goltzman, Characterization of the major parathyroid hormone target cell in the endosteal metaphysis of rat long bones. J. Bone Miner. Res. 5, 1043 – 1053 (1990). 93. X. Y. Zang, Y. B. Tan, Z. L. Pang, W. Z. Zhang, and J. Zhao, Effects of parathyroid hormone and estradiol on proliferation and function of human osteoblasts from fetal long bone: An in vitro study. Chin. Med. J. 107, 600 – 603 (1994). 94. M. Sabatini, C. Lesur, M. Pacherie, P. Pastoureau, N. Kucharczyk, J. L. Fauchère, and J. Bonnet, Effects of parathyroid hormone and agonists of the adenylyl cyclase and protein kinase C pathways on bone cell proliferation. Bone 18, 59 – 65 (1996).
CHAPTER 7 PTH and PTH-Related Protein 95. B. R. MacDonald, J. A. Gallagher, and R. G. Russell, Parathyroid hormone stimulates the proliferation of cells derived from human bone. Endocrinology 118, 2445 – 2449 (1986). 96. N. C. Partridge, A. L. Opie, R. T. Opie, and T. J. Martin, Inhibitory effects of parathyroid hormone on growth of osteogenic sarcoma cells. Calcif. Tissue Int. 37, 519 – 525 (1985). 97. S. Nishida, A. Yamaguchi, T. Tanizawa, N. Endo, T. Mashiba, Y. Uchiyama, T. Suda, S. Yoshiki, and H. E. Takahashi, Increased bone formation by intermittent parathyroid hormone administration is due to the stimulation of proliferation and differentiation of osteoprogenitor cells in bone marrow. Bone 15, 717 – 723 (1994). 98. P. C. Schiller, G. D’Ippolito, B. A. Roos, and G. A. Howard, Anabolic or catabolic responses of MC3T3-E1 osteoblastic cells to parathyroid hormone depend on time and duration of treatment. J. Bone Miner. Res. 14, 1504 – 1512 (1999). 99. T. Ishizuya, S. Yokose, M. Hori, T. Noda, T. Suda, S. Yoshiki, and A. Yamaguchi, Parathyroid hormone exerts disparate effects on osteoblast differentiation depending on exposure time in rat osteoblastic cells. J. Clin. Invest. 99, 2961 – 2970 (1997). 100. R. L. Jilka, R. S. Weinstein, T. Bellido, P. Roberson, A. M. Parfitt, and S. C. Manolagas, Increased bone formation by prevention of osteoblast apoptosis with parathyroid hormone. J. Clin. Invest. 104, 439 – 446 (1999). 101. D. Thiébaud, H. L. Guenther, A. Porret, P. Burckhardt, H. Fleisch, and W. Hofstetter, Regulation of collagen type I and biglycan mRNA levels by hormones and growth factors in normal and immortalized osteoblastic cell lines. J. Bone Miner. Res. 9, 1347 – 1354 (1994). 102. R. Yang and L. C. Gerstenfeld, Signal transduction pathways mediating parathyroid hormone stimulation of bone sialoprotein gene expression in osteoblasts. J. Biol. Chem. 271, 29839 – 29846 (1996). 103. G. Boguslawski, L. V. Hale, X. P. Yu, R. R. Miles, J. E. Onyia, R. F. Santerre, and S. Chandrasekhar, Activation of osteocalcin transcription involves interaction of protein kinase A- and protein kinase C-dependent pathways. J. Biol. Chem. 275, 999 – 1006 (2000). 104. E. Canalis, M. Centrella, W. Burch, and T. L. McCarthy, Insulin-like growth factor I mediates selective anabolic effects of parathyroid hormone in bone cultures. J. Clin. Invest. 83, 60 – 65 (1989). 105. J. H. R Ellsworth, Studies on physiology of parathyroid glands: Some responses of normal human kidneys and blood to intravenous parathyroid extract. Bull. Johns Hopkins Hosp. 55, 296 (1934). 106. H. Murer, I. Forster, H. Hilfiker, M. Pfister, B. Kaissling, M. Lötscher, and J. Biber, Cellular/molecular control of renal Na/Picotransport. Kidney Int. Suppl. 65, S2 – S10 (1998). 107. T. A. K. Berndt, Renal regulation of phosphate excretion. In “The Kidney: Physiology and Pathophysiology” (D. A. G. Seldin, ed.), pp. 2511 – 2532. Raven Press, New York, 1992. 108. H. Murer, Homer Smith Award: Cellular mechanisms in proximal tubular Pi reabsorption: Some answers and more questions. J. Am. Soc. Nephrol. 2, 1649 – 1665 (1992). 109. H. Murer, I. Forster, N. Hernando, G. Lambert, M. Traebert, and J. Biber, Posttranscriptional regulation of the proximal tubule NaPi-II transporter in response to PTH and dietary P(i). Am. J. Physiol. 277, F676 – F684 (1999). 110. I. Keusch, M. Traebert, M. Lötscher, B. Kaissling, H. Murer, and J. Biber, Parathyroid hormone and dietary phosphate provoke a lysosomal routing of the proximal tubular Na/Pi-cotransporter type II. Kidney Int. 54, 1224 – 1232 (1998). 111. Y. Zhang, J. M. Norian, C. E. Magyar, N. H. Holstein-Rathlou, A. K. Mircheff, and A. A. McDonough, In vivo PTH provokes apical NHE3 and NaPi2 redistribution and Na-K-ATPase inhibition. Am. J. Physiol. 276, F711 – F719 (1999). 112. S. A. Kempson, M. Lötscher, B. Kaissling, J. Biber, H. Murer, and M. Levi, Parathyroid hormone action on phosphate transporter mRNA and protein in rat renal proximal tubules. Am. J. Physiol. 268, F784 – F791 (1995).
239 113. P. A. Friedman and F. A. Gesek, Calcium transport in renal epithelial cells. Am. J. Physiol. 264, F181 – F198 (1993). 114. D. W. Seldin, Renal handling of calcium. Nephron 81, Suppl. 1, 2 – 7 (1999). 115. B. J. Bacskai and P. A. Friedman, Activation of latent Ca2 channels in renal epithelial cells by parathyroid hormone. Nature 347, 388 – 391 (1990). 116. F. A. Gesek and P. A. Friedman, On the mechanism of parathyroid hormone stimulation of calcium uptake by mouse distal convoluted tubule cells. J. Clin. Invest. 90, 749 – 758 (1992). 117. P. A. Friedman, Codependence of renal calcium and sodium transport. Annu. Rev. Physiol. 60, 179 – 197 (1998). 118. N. A. Breslau, Normal and abnormal regulation of 1,25-(OH)2D synthesis. Am. J. Med. Sci. 296, 417 – 425 (1988). 119. D. R. Fraser and E. Kodicek, Regulation of 25-hydroxycholecalciferol-1-hydroxylase activity in kidney by parathyroid hormone. Nature New Biol. 241, 163 – 166 (1973). 120. M. J. Favus, Factors affecting calcium metabolism in disorders of the kidney. Ann. Clin. Lab. Sci. 11, 327 – 332 (1981). 121. R. Kremer and D. Goltzman, Parathyroid hormone stimulates mammalian renal 25-hydroxyvitamin D3-1 alpha-hydroxylase in vitro. Endocrinology 110, 294 – 296 (1982). 122. C. R. Rost, D. D. Bikle, and R. A. Kaplan, In vitro stimulation of 25hydroxycholecalciferol 1 alpha-hydroxylation by parathyroid hormone in chick kidney slices: Evidence for a role for adenosine 3,5monophosphate. Endocrinology 108, 1002 – 1006 (1981). 123. T. Monkawa, T. Yoshida, S. Wakino, T. Shinki, H. Anazawa, H. F. Deluca, T. Suda, M. Hayashi, and T. Saruta, Molecular cloning of cDNA and genomic DNA for human 25-hydroxyvitamin D3 1 alphahydroxylase. Biochem. Biophys. Res. Commun. 239, 527 – 533 (1997). 124. G. K. Fu, D. Lin, M. Y. Zhang, D. D. Bikle, C. H. Shackleton, W. L. Miller, and A. A. Portale, Cloning of human 25-hydroxyvitamin D-1 alpha-hydroxylase and mutations causing vitamin D-dependent rickets type 1. Mol. Endocrinol. 11, 1961 – 1970 (1997). 125. F. H. Glorieux and R. St-Arnaud, Molecular cloning of (25-OH D)-1 alpha-hydroxylase: An approach to the understanding of vitamin D pseudo-deficiency. Recent Progr. Horm. Res. 53, 341 – 349 (1998). 126. A. Murayama, K. Takeyama, S. Kitanaka, Y. Kodera, Y. Kawaguchi, T. Hosoya, and S. Kato, Positive and negative regulations of the renal 25-hydroxyvitamin D3 1alpha-hydroxylase gene by parathyroid hormone, calcitonin, and 1alpha,25(OH)2D3 in intact animals. Endocrinology 140, 2224 – 2231 (1999). 127. T. Shinki, Y. Ueno, H. F. DeLuca, and T. Suda, Calcitonin is a major regulator for the expression of renal 25-hydroxyvitamin D3-1alphahydroxylase gene in normocalcemic rats. Proc. Natl. Acad. Sci. USA 96, 8253 – 8258 (1999). 128. A. F. Stewart, R. Horst, L. J. Deftos, E. C. Cadman, R. Lang, and A. E. Broadus, Biochemical evaluation of patients with cancerassociated hypercalcemia: Evidence for humoral and nonhumoral groups. N. Engl. J. Med. 303, 1377 – 1383 (1980). 129. S. B. Rodan, K. L. Insogna, A. M. Vignery, A. F. Stewart, A. E. Broadus, S. M. D’Souza, D. R. Bertolini, G. R. Mundy, and G. A. Rodan, Factors associated with humoral hypercalcemia of malignancy stimulate adenylate cyclase in osteoblastic cells. J. Clin. Invest. 72, 1511 – 1515 (1983). 130. G. J. Strewler, R. D. Williams, and R. A. Nissenson, Human renal carcinoma cells produce hypercalcemia in the nude mouse and a novel protein recognized by parathyroid hormone receptors. J. Clin. Invest. 71, 769 – 774 (1983). 131. S. A. Rabbani, J. Mitchell, D. R. Roy, R. Kremer, H. P. Bennett, and D. Goltzman, Purification of peptides with parathyroid hormone-like bioactivity from human and rat malignancies associated with hypercalcemia. Endocrinology 118, 1200 – 1210 (1986). 132. J. M. Moseley, M. Kubota, H. Diefenbach-Jagger, R. E. Wettenhall, B. E. Kemp, L. J. Suva, C. P. Rodda, P. R. Ebeling, P. J. Hudson,
240
133.
134.
135.
136.
137.
138.
139. 140.
141.
142.
143.
144.
145.
146.
147.
148.
ROBERT A. NISSENSON J. D. Zajac et al., Parathyroid hormone-related protein purified from a human lung cancer cell line. Proc. Natl. Acad. Sci. USA 84, 5048 – 5052 (1987). L. J. Suva, G. A. Winslow, R. E. Wettenhall, R. G. Hammonds, J. M. Moseley, H. Diefenbach-Jagger, C. P. Rodda, B. E. Kemp, H. Rodriguez, E. Y. Chen et al., A parathyroid hormone-related protein implicated in malignant hypercalcemia: Cloning and expression. Science 237, 893 – 896 (1987). G. J. Strewler, P. H. Stern, J. W. Jacobs, J. Eveloff, R. F. Klein, S. C. Leung, M. Rosenblatt, and R. A. Nissenson, Parathyroid hormonelike protein from human renal carcinoma cells: Structural and functional homology with parathyroid hormone. J. Clin. Invest. 80, 1803 – 1807 (1987). M. Mangin, A. C. Webb, B. E. Dreyer, J. T. Posillico, K. Ikeda, E. C. Weir, A. F. Stewart, N. H. Bander, L. Milstone, D. E. Barton et al., Identification of a cDNA encoding a parathyroid hormone-like peptide from a human tumor associated with humoral hypercalcemia of malignancy. Proc. Natl. Acad. Sci. USA 85, 597 – 601 (1988). M. Mangin, K. Ikeda, B. E. Dreyer, L. Milstone, and A. E. Broadus, Two distinct tumor-derived, parathyroid hormone-like peptides result from alternative ribonucleic acid splicing. Mol. Endocrinol. 2, 1049 – 1055 (1988). M. A. Thiede, G. J. Strewler, R. A. Nissenson, M. Rosenblatt, and G. A. Rodan, Human renal carcinoma expresses two messages encoding a parathyroid hormone-like peptide: Evidence for the alternative splicing of a single-copy gene. Proc. Natl. Acad. Sci. USA 85, 4605 – 4609 (1988). M. Mannstadt, H. Jüppner, and T. J. Gardella, Receptors for PTH and PTHrP: Their biological importance and functional properties. Am. J. Physiol. 277, F665 – F675 (1999). G. J. Strewler, The physiology of parathyroid hormone-related protein. N. Engl. J. Med. 342, 177 – 185 (2000). J. J. Wysolmerski and A. F. Stewart, The physiology of parathyroid hormone-related protein: An emerging role as a developmental factor. Annu. Rev. Physiol. 60, 431 – 460 (1998). B. Lanske and H. M. Kronenberg, Parathyroid hormone-related peptide (PTHrP) and parathyroid hormone (PTH)/PTHrP receptor. Crit. Rev. Eukary. Gene Express. 8, 297 – 320 (1998). W. M. Philbrick, J. J. Wysolmerski, S. Galbraith, E. Holt, J. J. Orloff, K. H. Yang, R. C. Vasavada, E. C. Weir, A. E. Broadus, and A. F. Stewart, Defining the roles of parathyroid hormone-related protein in normal physiology. Physiol. Rev. 76, 127 – 173 (1996). M. E. Dunbar, J. J. Wysolmerski, and A. E. Broadus, Parathyroid hormone-related protein: From hypercalcemia of malignancy to developmental regulatory molecule. Am. J. Med. Sci. 312, 287 – 294 (1996). A. C. Karaplis, A. Luz, J. Glowacki, R. T. Bronson, V. L. Tybulewicz, H. M. Kronenberg, and R. C. Mulligan, Lethal skeletal dysplasia from targeted disruption of the parathyroid hormone-related peptide gene. Genes Dev. 8, 277 – 289 (1994). A. Erlebacher, E. H. Filvaroff, S. E. Gitelman, and R. Derynck, Toward a molecular understanding of skeletal development. Cell 80, 371 – 378 (1995). B. St-Jacques, M. Hammerschmidt, and A. P. McMahon, Indian hedgehog signaling regulates proliferation and differentiation of chondrocytes and is essential for bone formation. Genes Dev. 13, 2072 – 2086 (1999). E. C. Weir, W. M. Philbrick, M. Amling, L. A. Neff, R. Baron, and A. E. Broadus, Targeted overexpression of parathyroid hormonerelated peptide in chondrocytes causes chondrodysplasia and delayed endochondral bone formation. Proc. Natl. Acad. Sci. USA 93, 10240 – 10245 (1996). E. Schipani, K. Kruse, and H. Jüppner, A constitutively active mutant PTH-PTHrP receptor in Jansen-type metaphyseal chondrodysplasia. Science 268, 98 – 100 (1995).
149. A. Vortkamp, K. Lee, B. Lanske, G. V. Segre, H. M. Kronenberg, and C. J. Tabin, Regulation of rate of cartilage differentiation by Indian hedgehog and PTH-related protein. Science 273, 613 – 622 (1996). 150. B. Lanske, A. C. Karaplis, K. Lee, A. Luz, A. Vortkamp, A. Pirro, M. Karperien, L. H. K. Defize, C. Ho, R. C. Mulligan, A. B. AbouSamra, H. Jüppner, G. V. Segre, and H. M. Kronenberg, PTH/PTHrP receptor in early development and Indian hedgehog-regulated bone growth. Science 273, 663 – 666 (1996). 151. S. J. Karp, E. Schipani, B. St-Jacques, J. Hunzelman, H. Kronenberg, and A. P. McMahon, Indian hedgehog coordinates endochondral bone growth and morphogenesis via parathyroid hormone relatedprotein-dependent and -independent pathways. Development 127, 543 – 548 (2000). 152. K. Lee, J. D. Deeds, and G. V. Segre, Expression of parathyroid hormone-related peptide and its receptor messenger ribonucleic acids during fetal development of rats. Endocrinology 136, 453 – 463 (1995). 153. M. Jansen, Uber atypische chondrodystrophie (achondroplasie) und uber eine noch nicht beschriebene angeborene wachstumsstarung des knochensystems: Metaphysare dysostosis. Z. Orthop. Chir. 61, 253 – 286 (1934). 154. A. S. Jobert, P. Zhang, A. Couvineau, J. Bonaventure, J. Roume, M. Le Merrer, and C. Silve, Absence of functional receptors for parathyroid hormone and parathyroid hormone-related peptide in Blomstrand chondrodysplasia. J. Clin. Invest. 102, 34 – 40 (1998). 155. A. C. Karaplis, B. He, M. T. Nguyen, I. D. Young, D. Semeraro, H. Ozawa, and N. Amizuka, Inactivating mutation in the human parathyroid hormone receptor type 1 gene in Blomstrand chondrodysplasia. Endocrinology 139, 5255 – 5258 (1998). 156. S. Blomstrand, I. Claësson and J. Säve-Söderbergh, A case of lethal congenital dwarfism with accelerated skeletal maturation. Pediatr. Radiol. 15, 141 – 143 (1985). 157. M. Amling, L. Neff, S. Tanaka, D. Inoue, K. Kuida, E. Weir, W. M. Philbrick, A. E. Broadus, and R. Baron, Bcl-2 lies downstream of parathyroid hormone-related peptide in a signaling pathway that regulates chondrocyte maturation during skeletal development. J. Cell Biol. 136, 205 – 213 (1997). 158. J. J. Wysolmerski, J. F. McCaughern-Carucci, A. G. Daifotis, A. E. Broadus, and W. M. Philbrick, Overexpression of parathyroid hormone-related protein or parathyroid hormone in transgenic mice impairs branching morphogenesis during mammary gland development. Development 121, 3539 – 3547 (1995). 159. J. J. Wysolmerski, W. M. Philbrick, M. E. Dunbar, B. Lanske, H. Kronenberg, and A. E. Broadus, Rescue of the parathyroid hormone-related protein knockout mouse demonstrates that parathyroid hormone-related protein is essential for mammary gland development. Development 125, 1285 – 1294 (1998). 160. S. L. Ferrari, R. Rizzoli, and J. P. Bonjour, Parathyroid hormonerelated protein production by primary cultures of mammary epithelial cells. J. Cell. Physiol. 150, 304 – 311 (1992). 161. M. E. Dunbar, P. Young, J. P. Zhang, J. McCaughern-Carucci, B. Lanske, J. J. Orloff, A. Karaplis, G. Cunha, and J. J. Wysolmerski, Stromal cells are critical targets in the regulation of mammary ductal morphogenesis by parathyroid hormone-related protein. Dev. Biol. 203, 75 – 89 (1998). 162. M. E. Dunbar, P. R. Dann, G. W. Robinson, L. Hennighausen, J. P. Zhang, and J. J. Wysolmerski, Parathyroid hormone-related protein signaling is necessary for sexual dimorphism during embryonic mammary development. Development 126, 3485 – 3493 (1999). 163. J. J. Merendino, Jr., K. L. Insogna, L. M. Milstone, A. E. Broadus, and A. F. Stewart, A parathyroid hormone-like protein from cultured human keratinocytes. Science 231, 388 – 390 (1986). 164. R. V. Campos, S. L. Asa, and D. J. Drucker, Immunocytochemical localization of parathyroid hormone-like peptide in the rat fetus. Cancer Res. 51, 6351 – 6357 (1991).
241
CHAPTER 7 PTH and PTH-Related Protein 165. E. J. Atillasoy, W. J. Burtis, and L. M. Milstone, Immunohistochemical localization of parathyroid hormone-related protein (PTHRP) in normal human skin. J. Invest. Dermatol. 96, 277 – 280 (1991). 166. K. K. Pun, C. D. Arnaud, and R. A. Nissenson, Parathyroid hormone receptors in human dermal fibroblasts: Structural and functional characterization. J. Bone Miner. Res. 3, 453 – 460 (1988). 167. J. J. Orloff, M. B. Ganz, A. E. Ribaudo, W. J. Burtis, M. Reiss, L. M. Milstone, and A. F. Stewart, Analysis of PTHRP binding and signal transduction mechanisms in benign and malignant squamous cells. Am. J. Physiol. 262, E599 – E607 (1992). 168. S. M. Kaiser, P. Laneuville, S. M. Bernier, J. S. Rhim, R. Kremer, and D. Goltzman, Enhanced growth of a human keratinocyte cell line induced by antisense RNA for parathyroid hormone-related peptide. J. Biol. Chem. 267, 13623 – 13628 (1992). 169. S. M. Kaiser, M. Sebag, J. S. Rhim, R. Kremer, and D. Goltzman, Antisense-mediated inhibition of parathyroid hormone-related peptide production in a keratinocyte cell line impedes differentiation. Mol. Endocrinol. 8, 139 – 147 (1994). 170. J. J. Wysolmerski, A. E. Broadus, J. Zhou, E. Fuchs, L. M. Milstone, and W. M. Philbrick, Overexpression of parathyroid hormone-related protein in the skin of transgenic mice interferes with hair follicle development. Proc. Natl. Acad. Sci. USA 91, 1133 – 1137 (1994). 171. J. Foley, B. J. Longely, J. J. Wysolmerski, B. E. Dreyer, A. E. Broadus, and W. M. Philbrick, PTHrP regulates epidermal differentiation in adult mice. J. Invest. Dermatol. 111, 1122 – 1128 (1998). 172. M. F. Holick, S. Ray, T. C. Chen, X. Tian, and K. S. Persons, A parathyroid hormone antagonist stimulates epidermal proliferation and hair growth in mice. Proc. Natl. Acad. Sci. USA 91, 8014 – 8016 (1994). 173. E. A. Blomme, Y. Sugimoto, Y. C. Lin, C. C. Capen, and T. J. Rosol, Parathyroid hormone-related protein is a positive regulator of keratinocyte growth factor expression by normal dermal fibroblasts. Mol. Cell. Endocrinol. 152, 189 – 197 (1999). 174. W. M. Philbrick, B. E. Dreyer, I. A. Nakchbandi, and A. C. Karaplis, Parathyroid hormone-related protein is required for tooth eruption. Proc. Natl. Acad. Sci. USA 95, 11846 – 11851 (1998). 175. M. A. Thiede, A. G. Daifotis, E. C. Weir, M. L. Brines, W. J. Burtis, K. Ikeda, B. E. Dreyer, R. E. Garfield, and A. E. Broadus, Intrauterine occupancy controls expression of the parathyroid hormone-related peptide gene in preterm rat myometrium. Proc. Natl. Acad. Sci. USA 87, 6969 – 6973 (1990). 176. A. G. Daifotis, E. C. Weir, B. E. Dreyer, and A. E. Broadus, Stretchinduced parathyroid hormone-related peptide gene expression in the rat uterus. J. Biol. Chem. 267, 23455 – 23458 (1992). 177. J. E. d. Ferguson, J. V. Gorman, D. E. Bruns, E. C. Weir, W. J. Burtis, T. J. Martin, and M. E. Bruns, Abundant expression of parathyroid hormone-related protein in human amnion and its association with labor. Proc. Natl. Acad. Sci. USA 89, 8384 – 8388 (1992). 178. A. M. Germain, H. Attaroglu, P. C. MacDonald, and M. L. Casey, Parathyroid hormone-related protein mRNA in avascular human amnion. J. Clin. Endocrinol. Metab. 75, 1173 – 1175 (1992). 179. M. Yamamoto, S. C. Harm, W. A. Grasser, and M. A. Thiede, Parathyroid hormone-related protein in the rat urinary bladder: A smooth muscle relaxant produced locally in response to mechanical stretch. Proc. Natl. Acad. Sci. USA 89, 5326 – 5330 (1992). 180. M. Ito, A. Ohtsuru, H. Enomoto, S. Ozeki, M. Nakashima, T. Nakayama, K. Shichijo, I. Sekine, and S. Yamashita, Expression of parathyroid hormone-related peptide in relation to perturbations of gastric motility in the rat. Endocrinology 134, 1936 – 1942 (1994). 181. G. A. Nickols, A. D. Nana, M. A. Nickols, D. J. DiPette, and G. K. Asimakis, Hypotension and cardiac stimulation due to the parathyroid hormone-related protein, humoral hypercalcemia of malignancy factor. Endocrinology 125, 834 – 841 (1989). 182. T. Massfelder, N. Fiaschi-Taesch, A. F. Stewart, and J. J. Helwig, Parathyroid hormone-related peptide: A smooth muscle tone and pro-
183.
184.
185.
186.
187.
188.
189.
190.
191.
192.
193.
194.
195.
196.
liferation regulatory protein. Curr. Opin. Nephrol. Hypertension 7, 27 – 32 (1998). K. Takahashi, D. Inoue, K. Ando, T. Matsumoto, K. Ikeda, and T. Fujita, Parathyroid hormone-related peptide as a locally produced vasorelaxant: Regulation of its mRNA by hypertension in rats. Biochem. Biophys. Res. Commun. 208, 447 – 455 (1995). C. J. Pirola, H. M. Wang, A. Kamyar, S. Wu, H. Enomoto, B. Sharifi, J. S. Forrester, T. L. Clemens, and J. A. Fagin, Angiotensin II regulates parathyroid hormone-related protein expression in cultured rat aortic smooth muscle cells through transcriptional and post-transcriptional mechanisms. J. Biol. Chem. 268, 1987 – 1994 (1993). S. Maeda, R. L. Sutliff, J. Qian, J. N. Lorenz, J. Wang, H. Tang, T. Nakayama, C. Weber, D. Witte, A. R. Strauch, R. J. Paul, J. A. Fagin, and T. L. Clemens, Targeted overexpression of parathyroid hormone-related protein (PTHrP) to vascular smooth muscle in transgenic mice lowers blood pressure and alters vascular contractility. Endocrinology 140, 1815 – 1825 (1999). J. Qian, J. N. Lorenz, S. Maeda, R. L. Sutliff, C. Weber, T. Nakayama, M. C. Colbert, R. J. Paul, J. A. Fagin, and T. L. Clemens, Reduced blood pressure and increased sensitivity of the vasculature to parathyroid hormone-related protein (PTHrP) in transgenic mice overexpressing the PTH/PTHrP receptor in vascular smooth muscle. Endocrinology 140, 1826 – 1833 (1999). A. G. Schneider, K. Leuthäuser, and D. Pette, Parathyroid hormonerelated protein is rapidly up-regulated in blood vessels of rat skeletal muscle by low-frequency stimulation. Pflug. Arch. Eur. J. Physiol. 439, 167 – 173 (1999). E. C. Weir, M. L. Brines, K. Ikeda, W. J. Burtis, A. E. Broadus, and R. J. Robbins, Parathyroid hormone-related peptide gene is expressed in the mammalian central nervous system. Proc. Natl. Acad. Sci. USA 87, 108 – 112 (1990). D. R. Weaver, J. D. Deeds, K. Lee, and G. V. Segre, Localization of parathyroid hormone-related peptide (PTHrP) and PTH/PTHrP receptor mRNAs in rat brain. Brain Res. Mol. Brain Res. 28, 296 – 310 (1995). E. H. Holt, A. E. Broadus, and M. L. Brines, Parathyroid hormonerelated peptide is produced by cultured cerebellar granule cells in response to L-type voltage-sensitive Ca2 channel flux via a Ca2/calmodulin-dependent kinase pathway. J. Biol. Chem. 271, 28105 – 28111 (1996). M. L. Brines, Z. Ling, and A. E. Broadus, Parathyroid hormone-related protein protects against kainic acid excitotoxicity in rat cerebellar granule cells by regulating L-type channel calcium flux. Neurosci. Lett. 274, 13 – 16 (1999). P. K. Pang, R. Wang, J. Shan, E. Karpinski, and C. G. Benishin, Specific inhibition of long-lasting, L-type calcium channels by synthetic parathyroid hormone. Proc. Natl. Acad. Sci. USA 87, 623 – 627 (1990). T. Ono, K. Inokuchi, A. Ogura, Y. Ikawa, Y. Kudo, and S. Kawashima, Activity-dependent expression of parathyroid hormone-related protein (PTHrP) in rat cerebellar granule neurons: Requirement of PTHrP for the activity-dependent survival of granule neurons. J. Biol. Chem. 272, 14404 – 14411 (1997). C. P. Rodda, M. Kubota, J. A. Heath, P. R. Ebeling, J. M. Moseley, A. D. Care, I. W. Caple, and T. J. Martin, Evidence for a novel parathyroid hormone-related protein in fetal lamb parathyroid glands and sheep placenta: Comparisons with a similar protein implicated in humoral hypercalcaemia of malignancy. J. Endocrinol. 117, 261 – 271 (1988). R. J. MacIsaac, I. W. Caple, J. A. Danks, H. Diefenbach-Jagger, V. Grill, J. M. Moseley, J. Southby, and T. J. Martin, Ontogeny of parathyroid hormone-related protein in the ovine parathyroid gland. Endocrinology 129, 757 – 764 (1991). C. S. Kovacs, B. Lanske, J. L. Hunzelman, J. Guo, A. C. Karaplis, and H. M. Kronenberg, Parathyroid hormone-related peptide
242
197.
198.
199.
200.
201.
202.
203.
204.
205.
206.
207.
208. 209.
210. 211.
212.
213.
214.
215.
216.
ROBERT A. NISSENSON (PTHrP) regulates fetal – placental calcium transport through a receptor distinct from the PTH/PTHrP receptor. Proc. Natl. Acad. Sci. USA 93, 15233 – 15238 (1996). M. A. Thiede and G. A. Rodan, Expression of a calcium-mobilizing parathyroid hormone-like peptide in lactating mammary tissue. Science 242, 278 – 280 (1988). M. Yamamoto, L. T. Duong, J. E. Fisher, M. A. Thiede, M. P. Caulfield, and M. Rosenblatt, Suckling-mediated increases in urinary phosphate and 3,5-cyclic adenosine monophosphate excretion in lactating rats: Possible systemic effects of parathyroid hormone-related protein. Endocrinology 129, 2614 – 2622 (1991). K. J. Mather, C. L. Chik, and B. Corenblum, Maintenance of serum calcium by parathyroid hormone-related peptide during lactation in a hypoparathyroid patient. J. Clin. Endocrinol. Metabol. 84, 424 – 427 (1999). A. A. Budayr, B. P. Halloran, J. C. King, D. Diep, R. A. Nissenson, and G. J. Strewler, High levels of a parathyroid hormone-like protein in milk. Proc. Natl. Acad. Sci. USA 86, 7183 – 7185 (1989). M. Sowers, D. Zhang, B. W. Hollis, B. Shapiro, C. A. Janney, M. Crutchfield, M. A. Schork, F. Stanczyk, and J. Randolph, Role of calciotrophic hormones in calcium mobilization of lactation. Am. J. Clin. Nutr. 67, 284 – 291 (1998). L. R. Chase and G. D. Aurbach, Renal adenyl cyclase: Anatomically separate sites for parathyroid hormone and vasopressin. Science 159, 545 – 547 (1968). L. R. Chase, S. A. Fedak, and G. D. Aurbach, Activation of skeletal adenyl cyclase by parathyroid hormone in vitro. Endocrinology 84, 761 – 768 (1969). R. Marcus and G. D. Aurbach, Bioassay of parathyroid hormone in vitro with a stable preparation of adenyl cyclase from rat kidney. Endocrinology 85, 801 – 810 (1969). T. Dousa and I. Rychlík, The effect of parathyroid hormone on adenyl cyclase in rat kidney. Biochim. Biophys. Acta 158, 484 – 486 (1968). B. Lanske, P. Divieti, C. S. Kovacs, A. Pirro, W. J. Landis, S. M. Krane, F. R. Bringhurst, and H. M. Kronenberg, The parathyroid hormone (PTH)/PTH-related peptide receptor mediates actions of both ligands in murine bone. Endocrinology 139, 5194 – 5204 (1998). G. Vaes, Parathyroid hormone-like action of N6-2-O-dibutyryladenosine-35 (cyclic)-monophosphate on bone explants in tissue culture. Nature 219, 939 – 940 (1968). L. G. Raisz, Physiologic and pharmacologic regulation of bone resorption. N. Engl. J. Med. 282, 909 – 916 (1970). M. P. Herrmann-Erlee, Studies on the role of cyclic AMP in parathyroid hormone-induced embryonic bone resorption. J. Endocrinol. 48, lix-lxi (1970). W. A. Peck and S. Klahr, Cyclic nucleotides in bone and mineral metabolism. Adv. Cyclic Nucleotide Res. 11, 89 – 130 (1979). J. A. Lorenzo, S. Sousa, and, J. Quinton, Forskolin has both stimulatory and inhibitory effects on bone resorption in fetal rat long bone cultures. J. Bone Miner. Res. 1, 313 – 317 (1986). M. P. Herrmann-Erlee and J. M. V. D. Meer, The effects of dibutyryl cyclic AMP, aminophylline and propranolol on PTE-induced bone resorption in vitro. Endocrinology 94, 424 – 434 (1974). R. Dziak and P. Stern, Parathyromimetic effects of the ionophore, A23187, on bone cells and organ cultures. Biochem. Biophys. Res. Commun. 65, 1343 – 1349 (1975). J. A. Lorenzo and S. Sousa, Phorbol esters stimulate bone resorption in fetal rat long-bone cultures by mechanisms independent of prostaglandin synthesis. J. Bone Miner. Res. 3, 63 – 67 (1988). D. C. Abraham, C. L. Wadkins, and H. H. Conaway, Enhancement of fetal rat limb bone resorption by phorbol ester (PMA) and ionophore A-23187. Calcif. Tissue Int. 42, 191 – 195 (1988). M. P. Bos, W. Most, J. P. van Leeuwen, and M. P. Herrmann-Erlee, Role of protein kinase C (PKC) in bone resorption: Effect of the
217.
218.
219.
220.
221.
222.
223.
224.
225.
226.
227.
228.
229.
230.
231.
232.
specific PKC inhibitor 1-alkyl-2-methylglycerol. Biochem. Biophys. Res. Commun. 184, 1317 – 1323 (1992). S. M. Sprague, M. M. Popovtzer, M. Dranitzki-Elhalel, and H. Wald, Parathyroid hormone-induced calcium efflux from cultured bone is mediated by protein kinase C translocation. Am. J. Physiol. 271, F1139 – F1146 (1996). E. Bornefalk, S. Ljunghall, and O. Ljunggren, Bone resorption induced by A23187 is abolished by indomethacin: implications for second messenger utilised by parathyroid hormone. Eur. J. Pharmacol. 345, 333 – 338 (1998). J. L. Ivey, D. R. Wright, and A. H. Tashjian, Jr., Bone resorption in organ culture: Inhibition by the divalent cation ionophores A23187 and X-537A. J. Clin. Invest. 58, 1327 – 1338 (1976). R. Dziak and P. H. Stern, Responses of fetal rat bone cells and bone organ cultures to the ionophore, A23187. Calcif. Tissue Res. 22, 137 – 147 (1976). M. Ransjö and U. H. Lerner, 12-O-tetradecanoylphorbol-13-acetate, a phorbol ester stimulating protein kinase C, inhibits bone resorption in vitro induced by parathyroid hormone and parathyroid hormonerelated peptide of malignancy. Acta Physiol. Scand. 139, 249 – 250 (1990). H. Kaji, T. Sugimoto, M. Kanatani, M. Nasu, and K. Chihara, Estrogen blocks parathyroid hormone (PTH)-stimulated osteoclast-like cell formation by selectively affecting PTH-responsive cyclic adenosine monophosphate pathway. Endocrinology 137, 2217 – 2224 (1996). M. Ransjö, A. Lie, and E. J. Mackie, Cholera toxin and forskolin stimulate formation of osteoclast-like cells in mouse marrow cultures and cultured mouse calvarial bones. Eur. J. Oral Sci. 107, 45 – 54 (1999). Y. H. Gao and M. Yamaguchi, Inhibitory effect of genistein on osteoclast-like cell formation in mouse marrow cultures. Biochem. Pharmacol. 58, 767 – 772 (1999). T. Sugimoto, M. Kanatani, H. Kaji, T. Yamaguchi, M. Fukase, and K. Chihara, Second messenger signaling of PTH- and PTHRPstimulated osteoclast-like cell formation from hemopoietic blast cells. Am. J. Physiol. 265, E367 – E373 (1993). H. Brändström, K. B. Jonsson, C. Ohlsson, O. Vidal, S. Ljunghall, and O. Ljunggren, Regulation of osteoprotegerin mRNA levels by prostaglandin E2 in human bone marrow stroma cells. Biochem. Biophys. Res. Commun. 247, 338 – 341 (1998). R. Kitazawa, S. Kitazawa, and S. Maeda, Promoter structure of mouse RANKL/TRANCE/OPGL/ODF gene. Biochim. Biophys. Acta 1445, 134 – 141 (1999). T. Onishi, W. Zhang, X. Cao, and K. Hruska, The mitogenic effect of parathyroid hormone is associated with E2F-dependent activation of cyclin-dependent kinase 1 (cdc2) in osteoblast precursors. J. Bone Miner. Res. 12, 1596 – 1605 (1997). Y. Isogai, T. Akatsu, T. Ishizuya, A. Yamaguchi, M. Hori, N. Takahashi, and T. Suda, Parathyroid hormone regulates osteoblast differentiation positively or negatively depending on the differentiation stages. J. Bone Miner. Res. 11, 1384 – 1393 (1996). R. H. Rixon, J. F. Whitfield, L. Gagnon, R. J. Isaacs, S. Maclean, B. Chakravarthy, J. P. Durkin, W. Neugebauer, V. Ross, W. Sung et al., Parathyroid hormone fragments may stimulate bone growth in ovariectomized rats by activating adenylyl cyclase. J. Bone Miner. Res. 9, 1179 – 1189 (1994). J. F. Whitfield, P. Morley, G. E. Willick, S. MacLean, V. Ross, J. Barbier, and R. J. Isaacs, Stimulation of femoral trabecular bone growth in ovariectomized rats by human parathyroid hormone (hPTH)(1 – 30)NH(2). Calcif. Tissue Int. 65, 143 – 147 (1999). A. Inoue, Y. Hiruma, S. Hirose, A. Yamaguchi, and H. Hagiwara, Reciprocal regulation by cyclic nucleotides of the differentiation of rat osteoblast-like cells and mineralization of nodules. Biochem. Biophys. Res. Commun. 215, 1104 – 1110 (1995).
243
CHAPTER 7 PTH and PTH-Related Protein 233. H. Kaneki, I. Takasugi, M. Fujieda, M. Kiriu, S. Mizuochi, and H. Ide, Prostaglandin E2 stimulates the formation of mineralized bone nodules by a cAMP-independent mechanism in the culture of adult rat calvarial osteoblasts. J. Cell. Biochem. 73, 36 – 48 (1999). 234. A. J. Koh, C. A. Beecher, T. J. Rosol, and L. K. McCauley, 3,5Cyclic adenosine monophosphate activation in osteoblastic cells: Effects on parathyroid hormone – 1 receptors and osteoblastic differentiation in vitro. Endocrinology 140, 3154 – 3162 (1999). 235. J. Kano, T. Sugimoto, M. Fukase, and T. Fujita, The activation of cAMP-dependent protein kinase is directly linked to the regulation of osteoblast proliferation (UMR-106) by parathyroid hormone. Biochem. Biophys. Res. Commun. 177, 365 – 369 (1991). 236. D. Chabardès, M. Imbert, A. Clique, M. Montégut, and F. Morel, PTH sensitive adenyl cyclase activity in different segments of the rabbit nephron. Pflug. Arch. Eur. J. Physiol. 354, 229 – 239 (1975). 237. S. Torikai and M. Imai, A simple method to determine adenylate cyclase activity in isolated single nephron segments by radioimmunoassay for succinyl adenosine 3,5-cyclic monophosphate. Tohoku J. Exp. Med. 129, 91 – 99 (1979). 238. C. A. Harris and J. F. Seely, Resistance to the phosphaturic effect of parathyroid hormone in the hamster. Am. J. Physiol. 237, F175 – F181 (1979). 239. T. J. Berndt, M. J. Onsgard, and F. G. Knox, Effect of cAMP analogue infusion on phosphate reabsorption in phosphate-deprived rats. Am. J. Physiol. 255, F96 – F99 (1988). 240. N. Beck and B. B. Davis, Impaired renal response to parathyroid hormone in potassium depletion. Am. J. Physiol. 228, 179 – 183 (1975). 241. T. Mano, K. Uchimura, R. Hayashi, T. Kobahashi, K. Fujiwara, M. Makino, H. Kakizawa, M. Nagata, A. Nakai, M. Wada, A. Nagasaka, and M. Itoh, Increased urinary phosphate excretion in pseudohypoparathyroidism type II with long-term treatment with phosphodiesterase inhibitor. Horm. Metab. Res. 31, 602 – 605 (1999). 242. A. M. Moses, N. Breslau, and R. Coulson, Renal responses to PTH in patients with hormone-resistant (pseudo) hypoparathyroidism. Am. J. Med. 61, 184 – 189 (1976). 243. R. M. Neer, G. W. Tregear, and J. T. Potts, Jr., Renal effects of native parathyroid hormone and synthetic biologically active fragments in pseudohypoparathyroidism and hypoparathyroidism. J. Clin. Endocrinol. Metab. 44, 420 – 423 (1977). 244. F. Albright, CH, Burnett, and P. H. Smith, Pseudohypoparathyroidism: An example of “Seabright-Bantam syndrome.” Endocrinology 30, 922 – 932 (1942). 245. M. A. Levine, Pseudohypoparathyroidism: From bedside to bench and back. J. Bone Miner. Res. 14, 1255 – 1260 (1999). 246. A. P. Teitelbaum and G. J. Strewler, Parathyroid hormone receptors coupled to cyclic adenosine monophosphate formation in an established renal cell line. Endocrinology 114, 980 – 985 (1984). 247. K. Malmström and H. Murer, Parathyroid hormone inhibits phosphate transport in OK cells but not in LLC-PK1 and JTC-12.P3 cells. Am. J. Physiol. 251, C23 – C31 (1986). 248. J. A. Cole, L. R. Forte, S. Eber, P. K. Thorne, and R. E. Poelling, Regulation of sodium-dependent phosphate transport by parathyroid hormone in opossum kidney cells: Adenosine 3,5-monophosphatedependent and -independent mechanisms. Endocrinology 122, 2981 – 2989 (1988). 249. G. Quamme, J. Pfeilschifter, and H. Murer, Parathyroid hormone inhibition of Na/phosphate cotransport in OK cells: Intracellular [Ca2] as a second messenger. Biochim. Biophys. Acta 1013, 166 – 172 (1989). 250. K. J. Martin, C. L. McConkey, A. K. Jacob, E. A. Gonzalez, M. Khan, and J. J. Baldassare, Effect of U-73,122, an inhibitor of phospholipase C, on actions of parathyroid hormone in opossum kidney cells. Am. J. Physiol. 266, F254 – F258 (1994). 251. J. H. Segal and A. S. Pollock, Transfection-mediated expression of a dominant cAMP-resistant phenotype in the opossum kidney (OK)
252.
253.
254.
255.
256.
257.
258.
259.
260.
261.
262.
263.
264.
265.
266.
267.
268.
cell line prevents parathyroid hormone-induced inhibition of Naphosphate cotransport: A protein kinase-A-mediated event. J. Clin. Invest. 86, 1442 – 1450 (1990). E. D. Lederer, S. S. Sohi, J. M. Mathiesen, and J. B. Klein, Regulation of expression of type II sodium-phosphate cotransporters by protein kinases A and C. Am. J. Physiol. 275, F270 – F277 (1998). M. F. Pfister, J. Forgo, U. Ziegler, J. Biber, and H. Murer, cAMPdependent and -independent downregulation of type II Na-Pi cotransporters by PTH. Am. J. Physiol. 276, F720 – F725 (1999). M. Nakai, Y. Kinoshita, M. Fukase, and T. Fujita, Phorbol esters inhibit phosphate uptake in opossum kidney cells: A model of proximal renal tubular cells. Biochem. Biophys. Res. Commun. 145, 303 – 308 (1987). J. A. Cole, S. L. Eber, R. E. Poelling, P. K. Thorne, and L. R. Forte, A dual mechanism for regulation of kidney phosphate transport by parathyroid hormone. Am. J. Physiol. 253, E221 – E227 (1987). K. Malmström, G. Stange, and H. Murer, Intracellular cascades in the parathyroid-hormone-dependent regulation of Na/phosphate cotransport in OK cells. Biochem. J. 251, 207 – 213 (1988). J. Welsh, V. Weaver, and M. Simboli-Campbell, Regulation of renal 25(OH)D3 1 alpha-hydroxylase: Signal transduction pathways. Biochem. Cell Biol. 69, 768 – 770 (1991). H. L. Henry, Parathyroid hormone modulation of 25-hydroxyvitamin D3 metabolism by cultured chick kidney cells is mimicked and enhanced by forskolin. Endocrinology 116, 503 – 510 (1985). H. L. Brenza, C. Kimmel-Jehan, F. Jehan, T. Shinki, S. Wakino, H. Anazawa, T. Suda, and H. F. DeLuca, Parathyroid hormone activation of the 25-hydroxyvitamin D3 – 1alpha-hydroxylase gene promoter. Proc. Natl. Acad. Sci. USA 95, 1387 – 1391 (1998). A. Murayama, K. Takeyama, S. Kitanaka, Y. Kodera, T. Hosoya, and S. Kato, The promoter of the human 25-hydroxyvitamin D3 1 alphahydroxylase gene confers positive and negative responsiveness to PTH, calcitonin, and 1 alpha,25(OH)2D3. Biochem. Biophys. Res. Commun. 249, 11 – 16 (1998). X. F. Kong, X. H. Zhu, Y. L. Pei, D. M. Jackson, and M. F. Holick, Molecular cloning, characterization, and promoter analysis of the human 25-hydroxyvitamin D3-1alpha-hydroxylase gene. Proc. Natl. Acad. Sci. USA 96, 6988 – 6993 (1999). H. K. Ro, V. Tembe, and M. J. Favus, Evidence that activation of protein kinase-C can stimulate 1,25-dihydroxyvitamin D3 secretion by rat proximal tubules. Endocrinology 131, 1424 – 1428 (1992). M. Janulis, V. Tembe, and M. J. Favus, Role of protein kinase C in parathyroid hormone stimulation of renal 1,25-dihydroxyvitamin D3 secretion. J. Clin. Invest. 90, 2278 – 2283 (1992). J. G. Hoenderop, J. J. De Pont, R. J. Bindels, and P. H. Willems, Hormone-stimulated Ca2 reabsorption in rabbit kidney cortical collecting system is cAMP-independent and involves a phorbol ester-insensitive PKC isotype. Kidney Int. 55, 225 – 233 (1999). G. Hilal, D. Claveau, M. Leclerc, and M. G. Brunette, Ca2 transport by the luminal membrane of the distal nephron: Action and interaction of protein kinases A and C. Biochem. J. 328, 371 – 375 (1997). P. A. Friedman, F. A. Gesek, P. Morley, J. F. Whitfield, and G. E. Willick, Cell-specific signaling and structure-activity relations of parathyroid hormone analogs in mouse kidney cells. Endocrinology 140, 301 – 309 (1999). A. T. Singh, J. G. Kunnel, P. J. Strieleman, and P. H. Stern, Parathyroid hormone (PTH)-(1 – 34), [Nle(8,18),Tyr34]PTH-(3 – 34) amide, PTH-(1 – 31) amide, and PTH-related peptide-(1 – 34) stimulate phosphatidylcholine hydrolysis in UMR106 osteoblastic cells: Comparison with effects of phorbol 12,13-dibutyrate. Endocrinology 140, 131 – 137 (1999). P. Morley, J. F. Whitfield, and G. E. Willick, Design and applications of parathyroid hormone analogues. Curr. Med. Chem. 6, 1095 – 1106 (1999).
244 269. H. Kather, W. Tschöpe, and B. Simon, Human fat cell adenylate cyclase: Modulation of parathyroid hormone action by guanine nucleotides. Res. Exp. Med. 171, 201 – 204 (1977). 270. V. P. Michalangeli, N. H. Hunt, and T. J. Martin, States of activation of chick kidney adenylate cyclase induced by parathyroid hormone and guanyl nucleotides. J. Endocrinol. 72, 69 – 79 (1977). 271. E. Bellorin-Font and K. J. Martin, Regulation of the PTH-receptorcyclase system of canine kidney: Effects of calcium, magnesium, and guanine nucleotides. Am. J. Physiol. 241, F364 – F373 (1981). 272. R. E. Rizzoli, T. M. Murray, S. J. Marx, and G. D. Aurbach, Binding of radioiodinated bovine parathyroid hormone-(1 – 84) to canine renal cortical membranes. Endocrinology 112, 1303 – 1312 (1983). 273. A. P. Teitelbaum, R. A. Nissenson, and C. D. Arnaud, Coupling of the canine renal parathyroid hormone receptor to adenylate cyclase: Modulation by guanyl nucleotides and N-ethylmaleimide. Endocrinology 111, 1524 – 1533 (1982). 274. R. A. Nissenson, S. R. Abbott, A. P. Teitelbaum, O. H. Clark, and C. D. Arnaud, Endogenous biologically active human parathyroid hormone: Measurement by a guanyl nucleotide-amplified renal adenylate cyclase assay. J. Clin. Endocrinol Metab. 52, 840 – 846 (1981). 275. H. Jüppner, A. B. Abou-Samra, M. Freeman, X. F. Kong, E. Schipani, J. Richards, L. F. Kolakowski, Jr., J. Hock, J. T. Potts, Jr., H. M. Kronenberg, and G. V. Segre, A G protein-linked receptor for parathyroid hormone and parathyroid hormone-related peptide. Science 254, 1024 – 1026 (1991). 276. T. Schöneberg, G. Schultz, and T. Gudermann, Structural basis of G protein-coupled receptor function. Mol. Cell. Endocrinol. 151, 181 – 193 (1999). 277. C. D. Strader, T. M. Fong, M. P. Graziano, and M. R. Tota, The family of G-protein-coupled receptors. FASEB J. 9, 745 – 754 (1995). 278. M. Babich, H. Choi, R. M. Johnson, K. L. King, G. E. Alford, and R. A. Nissenson, Thrombin and parathyroid hormone mobilize intracellular calcium in rat osteosarcoma cells by distinct pathways. Endocrinology 129, 1463 – 1470 (1991). 279. H. Jüppner, E. Schipani, F. R. Bringhurst, I. McClure, H. T. Keutmann, J. T. Potts, Jr., H. M. Kronenberg, A. B. Abou-Samra, G. V. Segre, and T. J. Gardella, The extracellular amino-terminal region of the parathyroid hormone (PTH)/PTH-related peptide receptor determines the binding affinity for carboxyl-terminal fragments of PTH(1 – 34). Endocrinology 134, 879 – 884 (1994). 280. P. R. Turner, T. Bambino, and R. A. Nissenson, A putative selectivity filter in the G-protein-coupled receptors for parathyroid hormone and secretion. J. Biol. Chem. 271, 9205 – 9208 (1996). 281. C. Bergwitz, T. J. Gardella, M. R. Flannery, J. T. Potts, Jr., H. M. Kronenberg, S. R. Goldring, and H. Jüppner, Full activation of chimeric receptors by hybrids between parathyroid hormone and calcitonin: Evidence for a common pattern of ligand-receptor interaction. J. Biol. Chem. 271, 26469 – 26472 (1996). 282. T. J. Gardella, H. Jüppner, A. K. Wilson, H. T. Keutmann, A. B. Abou-Samra, G. V. Segre, F. R. Bringhurst, J. T. Potts, Jr., S. R. Nussbaum, and H. M. Kronenberg, Determinants of [Arg2]PTH-(1 – 34) binding and signaling in the transmembrane region of the parathyroid hormone receptor. Endocrinology 135, 1186 – 1194 (1994). 283. P. R. Turner, T. Bambino, and R. A. Nissenson, Mutations of neighboring polar residues on the second transmembrane helix disrupt signaling by the parathyroid hormone receptor. Mol. Endocrinol. 10, 132 – 139 (1996). 284. T. J. Gardella, M. D. Luck, M. H. Fan, and C. Lee, Transmembrane residues of the parathyroid hormone (PTH)/PTH-related peptide receptor that specifically affect binding and signaling by agonist ligands. J. Biol. Chem. 271, 12820 – 12825 (1996). 285. M. Mannstadt, M. D. Luck, T. J. Gardella, and H. Juppner, Evidence for a ligand interaction site at the amino-terminus of the parathyroid hormone (PTH)/PTH-related protein receptor from cross-linking and mutational studies. J. Biol. Chem. 273, 16890 – 16896 (1998).
ROBERT A. NISSENSON 286. A. T. Zhou, R. Bessalle, A. Bisello, C. Nakamoto, M. Rosenblatt, L. J. Suva, and M. Chorev, Direct mapping of an agonist-binding domain within the parathyroid hormone/parathyroid hormone-related protein receptor by photoaffinity crosslinking. Proc. Natl. Acad. Sci. USA 94, 3644 – 3649. 287. A. Bisello, A. E. Adams, D. F. Mierke, M. Pellegrini, M. Rosenblatt, L. J. Suva, and M. Chorev, Parathyroid hormone-receptor interactions identified directly by photocross-linking and molecular modeling studies. J. Biol. Chem. 273, 22498 – 22505 (1998). 288. S. P. Sheikh, J. P. Vilardarga, T. J. Baranski, O. Lichtarge, T. Iiri, E. C. Meng, R. A. Nissenson, and H. R. Bourne, Similar structures and shared switch mechanisms of the beta2-adrenoceptor and the parathyroid hormone receptor: Zn(II) bridges between helices III and VI block activation. J. Biol. Chem. 274, 17033 – 17041 (1999). 289. Z. Huang, Y. Chen, S. Pratt, T. H. Chen, T. Bambino, R. A. Nissenson, and D. M. Shoback, The N-terminal region of the third intracellular loop of the parathyroid hormone (PTH)/PTH-related peptide receptor is critical for coupling to cAMP and inositol phosphate/Ca2 signal transduction pathways. J. Biol. Chem. 271, 33382 – 33389 (1996). 290. A. Iida-Klein, J. Guo, M. Takemura, M. T. Drake, J. T. Potts, Jr., A. Abou-Samra, F. R. Bringhurst, and G. V. Segre, Mutations in the second cytoplasmic loop of the rat parathyroid hormone (PTH)/PTHrelated protein receptor result in selective loss of PTH-stimulated phospholipase C activity. J. Biol. Chem. 272, 6882 – 6889 (1997). 291. G. L. Wong, Induction of metabolic changes and down regulation of bovine parathyroid hormone-responsive adenylate cyclase are dissociable in isolated osteoclastic and osteoblastic bone cells. J. Biol. Chem. 254, 34 – 37 (1979). 292. W. A. Peck and G. Kohler, Hormonal and nonhormonal desensitization in isolated bone cells. Calcif. Tissue Int. 32, 95 – 103 (1980). 293. S. R. Goldring, J. M. Dayer, and M. Rosenblatt, Factors regulating the response of cells cultured from human giant cell tumors of bone to parathyroid hormone. J. Clin. Endocrinol. Metab. 53, 295 – 300 (1981). 294. W. I. Chao and L. R. Forte, Rat kidney cells in primary culture: Hormone-mediated desensitization of the adenosine 3,5-monophosphate response to parathyroid hormone and calcitonin. Endocrinology 111, 252 – 259 (1982). 295. H. L. Henry, N. S. Cunningham, and T. A. Noland, Jr., Homologous desensitization of cultured chick kidney cells to parathyroid hormone. Endocrinology 113, 1942 – 1949 (1983). 296. C. Bergwitz, A. B. Abou-Samra, R. D. Hesch, and H. Jüppner, Rapid desensitization of parathyroid hormone dependent adenylate cyclase in perifused human osteosarcoma cells (SaOS-2). Biochim. Biophys. Acta 1222, 447 – 456 (1994). 297. S. K. Lee and P. H. Stern, Studies on the mechanism of desensitization of the parathyroid hormone-stimulated calcium signal in UMR106 cells: Reversal of desensitization by alkaline phosphatase but not by protein kinase C downregulation. J. Bone Miner. Res. 9, 781 – 789 (1994). 298. A. Fujimori, A. Miyauchi, K. A. Hruska, K. J. Martin, L. V. Avioli, and R. Civitelli, Desensitization of calcium messenger system in parathyroid hormone-stimulated opossum kidney cells. Am. J. Physiol. 264, E918 – E924 (1993). 299. S. Fukayama, A. H. Tashjian, Jr., and F. R. Bringhurst, Mechanisms of desensitization to parathyroid hormone in human osteoblast-like SaOS-2 cells. Endocrinology 131, 1757 – 1769 (1992). 300. K. K. Pun, P. W. Ho, R. A. Nissenson, and C. D. Arnaud, Desensitization of parathyroid hormone receptors on cultured bone cells. J. Bone Miner. Res. 5, 1193 – 1200 (1990). 301. M. Bünemann and M. M. Hosey, G-protein coupled receptor kinases as modulators of G-protein signalling. J. Physiol. 517, 5 – 23 (1999). 302. R. J. Lefkowitz, G protein-coupled receptors. III. New roles for receptor kinases and beta-arrestins in receptor signaling and desensitization. J. Biol. Chem. 273, 18677 – 18680 (1998).
245
CHAPTER 7 PTH and PTH-Related Protein 303. J. G. Krupnick and J. L. Benovic, The role of receptor kinases and arrestins in G protein-coupled receptor regulation. Annu. Rev. Pharmacol. Toxicol. 38, 289 – 319 (1998). 304. E. Blind, T. Bambino, and R. A. Nissenson, Agonist-stimulated phosphorylation of the G protein-coupled receptor for parathyroid hormone (PTH) and PTH-related protein. Endocrinology 136, 4271 – 4277 (1995). 305. F. Qian, A. Leung, and A. Abou-Samra, Agonist-dependent phosphorylation of the parathyroid hormone/parathyroid hormone-related peptide receptor. Biochemistry 37, 6240 – 6246 (1998). 306. E. Blind, T. Bambino, Z. Huang, M. Bliziotes, and R. A. Nissenson, Phosphorylation of the cytoplasmic tail of the PTH/PTHrP receptor. J. Bone Miner. Res. 11, 578 – 586 (1996). 307. N. Malecz, T. Bambino, M. Bencsik, R. A. Nissenson, Identification of phosphorylation sites in the G-protein coupled receptor for parathyroid hormone: Receptor phosphorylation is not required for agonist-induced internalization. Mol. Endocrinol. 12, 1846 – 1856 (1998). 308. F. Dicker, U. Quitterer, R. Winstel, K. Honold, and M. J. Lohse, Phosphorylation-independent inhibition of parathyroid hormone receptor signaling by G protein-coupled receptor kinases. Proc. Natl. Acad. Sci. USA 96, 5476 – 5481 (1999). 309. S. Fukayama, G. Kong, J. L. Benovic, E. Meurer, and A. H. Tashjian, Jr., Beta-adrenergic receptor kinase-1 acutely regulates PTH/PTHrP receptor signalling in human osteoblastlike cells. Cell. Signal. 9, 469 – 474 (1997). 310. C. A. Mahoney and R. A. Nissenson, Canine renal receptors for parathyroid hormone. Down-regulation in vivo by exogenous parathyroid hormone. J. Clin. Invest. 72, 411 – 421 (1983). 311. A. P. Teitelbaum, C. M. Silve, K. O. Nyiredy, and C. D. Arnaud, Down-regulation of parathyroid hormone (PTH) receptors in cultured bone cells is associated with agonist-specific intracellular processing of PTH-receptor complexes Endocrinology 118, 595 – 602 (1986). 312. I. Yamamoto, C. Shigeno, J. T. Potts, Jr., and G. V. Segre, Characterization and agonist-induced down-regulation of parathyroid hormone receptors in clonal rat osteosarcoma cells. Endocrinology 122, 1208 – 1217 (1988). 313. J. Mitchell and D. Goltzman, Mechanisms of homologous and heterologous regulation of parathyroid hormone receptors in the rat osteosarcoma cell line UMR-106. Endocrinology 126, 2650 – 2660 (1990). 314. L. R. Forte, G. A. Nickols, and C. S. Anast, Renal adenylate cyclase and the interrelationship between parathyroid hormone and vitamin D in the regulation of urinary phosphate and adenosine cyclic 3,5monophosphate excretion. J. Clin. Invest. 57, 559 – 568 (1976). 315. S. G. Massry, R. Stein, J. Garty, A. I. Arieff, J. W. Coburn, A. W. Norman, and R. M. Friedler, Skeletal resistance to the calcemic action of parathyroid hormone in uremia: Role of 1,25 (OH)2 D3. Kidney Int. 9, 467 – 474 (1976). 316. I. G. Lewin, S. E. Papapoulos, G. N. Hendy, S. Tomlinson, and J. L. O’Riordan, Reversible resistance to the renal action of parathyroid hormone in human vitamin D deficiency. Clin. Sci. 62, 381 – 387 (1982). 317. G. A. Nickols, D. L. Carnes, C. S. Anast, and L. R. Forte, Parathyroid hormone-mediated refractoriness of rat kidney cyclic AMP system. Am. J. Physiol. 236, E401 – E409 (1979). 318. P. Ureña, M. Mannstadt, M. Hruby, A. Ferreira, F. Schmitt, C. Silve, R. Ardaillou, B. Lacour, A. B. Abou-Samra, G. V. Segre et al., Parathyroidectomy does not prevent the renal PTH/PTHrP receptor down-regulation in uremic rats. Kidney Int. 47, 1797 – 1805 (1995). 319. Z. Huang, Y. Chen, and R. A. Nissenson, The cytoplasmic tail of the G-protein-coupled receptor for parathyroid hormone and parathyroid hormone-related protein contains positive and negative signals for endocytosis. J. Biol. Chem. 270, 151 – 156 (1995). 320. S. L. Ferrari, V. Behar, M. Chorev, M. Rosenblatt, and A. Bisello, Endocytosis of ligand-human parathyroid hormone receptor 1
321.
322.
323.
324.
325.
326.
327.
328.
329.
330.
331.
332.
333.
334.
335.
complexes is protein kinase C-dependent and involves betaarrestin2. Real-time monitoring by fluorescence microscopy. J. Biol. Chem. 274, 29968 – 29975 (1999). E. A. González and K. J. Martin, Coordinate regulation of PTH/PTHrP receptors by PTH and calcitriol in UMR 106-01 osteoblast-like cells. Kidney Inte. 50, 63 – 70 (1996). J. W. Jongen, E. C. Willemstein-van Hove, J. M. van der Meer, M. P. Bos, H. Jüppner, G. V. Segre, A. B. Abou-Samra, J. H. Feyen, and M. P. Herrmann-Erlee, Down-regulation of the receptor for parathyroid hormone (PTH) and PTH-related peptide by PTH in primary fetal rat osteoblasts. J. Bone Miner. Res. 11, 1218 – 1225 (1996). G. Turner, C. Coureau, M. R. Rabin, B. Escoubet, M. Hruby, O. Walrant, and C. Silve, Parathyroid hormone (PTH)/PTH-related protein receptor messenger ribonucleic acid expression and PTH response in a rat model of secondary hyperparathyroidism associated with vitamin D deficiency. Endocrinology 136, 3751 – 3758 (1995). L. K. McCauley, C. A. Beecher, M. E. Melton, J. R. Werkmeister, H. Jüppner, A. B. Abou-Samra, G. V. Segre, and T. J. Rosol, Transforming growth factor-beta1 regulates steady-state PTH/PTHrP receptor mRNA levels and PTHrP binding in ROS 17/2.8 osteosarcoma cells. Mol. Cell. Endocrinol. 101, 331 – 336 (1994). J. W. Jongen, E. C. Willemstein-Van Hove, J. M. Van der Meer, M. P. Bos, H. Jüppner, G. V. Segre, A. B. Abou-Samra, J. H. Feyen, and M. P. Herrmann-Erlee, Down-regulation of the receptor for parathyroid hormone (PTH) and PTH-related peptide by transforming growth factor-beta in primary fetal rat osteoblasts. Endocrinology 136, 3260 – 3266 (1995). F. Law, J. P. Bonjour, and R. Rizzoli, Transforming growth factorbeta: A down-regulator of the parathyroid hormone-related protein receptor in renal epithelial cells. Endocrinology 134, 2037 – 2043 (1994). P. Ureña, A. Iida-Klein, X. F. Kong, H. Jüppner, H. M. Kronenberg, A. B. Abou-Samra, and G. V. Segre, Regulation of parathyroid hormone (PTH)/PTH-related peptide receptor messenger ribonucleic acid by glucocorticoids and PTH in ROS 17/2.8 and OK cells. Endocrinology 134, 451 – 456 (1994). J. Yaghoobian and T. B. Drüeke, Regulation of the transcription of parathyroid-hormone/parathyroid-hormone-related peptide receptor mRNA by dexamethasone in ROS 17/2.8 osteosarcoma cells. Nephrol. Dialysis Transplant. 13, 580 – 586 (1998). H. Wald, M. Dranitzki-Elhalel, R. Backenroth, and M. M. Popovtzer, Evidence for interference of vitamin D with PTH/PTHrP receptor expression in opossum kidney cells. Pflug. Arch. Eur. J. Physiol. 436, 289 – 294 (1998). J. Cornish, K. E. Callon, G. C. Nicholson, and I. R. Reid, Parathyroid hormone-related protein-(107 – 139) inhibits bone resorption in vivo. Endocrinology 138, 1299 – 1304 (1997). J. Cornish, K. E. Callon, C. Lin, C. Xiao, J. M. Moseley, and I. R. Reid, Stimulation of osteoblast proliferation by C-terminal fragments of parathyroid hormone-related protein. J. Bone Miner. Res. 14, 915 – 922 (1999). F. De Miguel, P. Martinez-Fernandez, C. Guillen, A. Valin, A. Rodrigo, M. E. Martinez, and P. Esbrit, Parathyroid hormone-related protein (107 – 139) stimulates interleukin-6 expression in human osteoblastic cells. J. Am. Soc. Nephrol. 10, 796 – 803 (1999). A. D. Care, S. K. Abbas, D. W. Pickard, M. Barri, M. Drinkhill, J. B. Findlay, I. R. White, and I. W. Caple, Stimulation of ovine placental transport of calcium and magnesium by mid-molecule fragments of human parathyroid hormone-related protein. Exp. Physiol. 75, 605 – 608 (1990). M. T. Nguyen and A. C. Karaplis, The nucleus: A target site for parathyroid hormone-related peptide (PTHrP) action. J. Cell. Biochem. 70, 193 – 199 (1998). M. H. Lam, L. J. Briggs, W. Hu, T. J. Martin, M. T. Gillespie, and D. A. Jans, Importin beta recognizes parathyroid hormone-related
246 protein with high affinity and mediates its nuclear import in the absence of importin alpha. J. Biol. Chem. 274, 7391 – 7398 (1999). 336. M. M. Aarts, D. Levy, B. He, S. Stregger, T. Chen, S. Richard, and J. E. Henderson, Parathyroid hormone-related protein interacts with RNA. J. Biol. Chem. 274, 4832 – 4838 (1999). 337. J. E. Henderson, N. Amizuka, H. Warshawsky, D. Biasotto, B. M. Lanske, D. Goltzman, and A. C. Karaplis, Nucleolar localization of parathyroid hormone-related peptide enhances survival of chondrocytes under conditions that promote apoptotic cell death. Mol. Cell. Biol. 15, 4064 – 4075 (1995). 338. T. Massfelder, P. Dann, T. L. Wu, R. Vasavada, J. J. Helwig, and A. F. Stewart, Opposing mitogenic and anti-mitogenic actions of parathyroid hormone-related protein in vascular smooth muscle
ROBERT A. NISSENSON cells: A critical role for nuclear targeting. Proc. Natl. Acad. Sci. USA 94, 13630 – 13635 (1997). 339. M. H. Lam, S. L. Olsen, W. A. Rankin, P. W. Ho, T. J. Martin, M. T. Gillespie, and J. M. Moseley, PTHrP and cell division: Expression and localization of PTHrP in a keratinocyte cell line (HaCaT) during the cell cycle. J. Cell. Physiol. 173, 433 – 446 (1997). 340. M. H. Lam, C. M. House, T. Tiganis, K. I. Mitchelhill, B. Sarcevic, A. Cures, R. Ramsay, B. E. Kemp, T. J. Martin, and M. T. Gillespie, Phosphorylation at the cyclin-dependent kinases site (Thr85) of parathyroid hormone-related protein negatively regulates its nuclear localization. J. Biol. Chem. 274, 18559 – 18566 (1999). 341. D. Endres, R. Villanueva, C. F. Sharp, Jr., and F. R. Singer, Measurement of parathyroid hormone. Endocrinol. Metab. Clin. North Am. 18, 611 – 629 (1989).
CHAPTER 8
Calcitonin ANA O. HOFF, GILBERT J. COTE, AND ROBERT F. GAGEL Department of Endocrine Neoplasia and Hormonal Disorders, Divison of Internal Medicine, University of Texas M.D. Anderson Cancer Center, Houston, Texas 77030
I. II. III. IV.
V. Physiology and Mechanism of Action VI. The Calcitonin Receptor Superfamily VII. Summary References
Introduction Calcitonin Gene Products Calcitonin Structure/Function Relationships Secretion and Metabolism
processed to produce only CGRP. The third is a pseudogene and is not expressed. The fourth encodes the peptide amylin. In addition, a fifth gene encoding a vasodilator named adrenomedullin will be discussed because of its homology to CGRP and interactions with a family of CT-like receptors. This chapter focuses on calcitonin and therefore only the transcriptional and posttranscriptional regulation of the CALC I or CGRP gene will be discussed. The CALC I or CGRP gene is expressed in a broad spectrum of neuroendocrine cells, but is predominantly produced by the C or calcitonin-producing cell of the thyroid gland. The C cell migrates from the neural crest during embryonic life to a position in the thyroid gland. In other species (fish and birds), C cells are located in a discrete organ in the neck region called the ultimobranchial body [5]. More than 90% of the calcitonin circulating in plasma is normally produced by the thyroidal C cell; the remaining 10% or less is produced in a variety of neuroendocrine cells in the pituitary, lung, adrenal glands, prostate, and other sites. A cDNA for CT was identified in 1981 [6]; subsequent investigators identified a second mRNA with partial sequence homology to the CT cDNA and named it calcitonin gene-related peptide [7]. The 5 portions of this cDNA are identical to similar portions of the CT cDNA, whereas the 3 portion encodes CGRP. Further elucidation of the
I. INTRODUCTION In 1961 Copp made the observation that a protein extract derived from the ultimobranchial body of salmon caused a lowering of the serum calcium when injected into a rodent [1, 2]. He named this substance calcitonin (CT). In mammalian species this activity was subsequently localized to the thyroid gland [3] and shown to be produced by the C or calcitonin-producing cells of the thyroid gland [4]. The active principle was subsequently shown to be a 32 amino acid peptide, released in response to an increase in the plasma calcium concentration. Calcitonin interacts with a specific G-protein-coupled receptor on the osteoclast surface and causes inhibition of bone resorption.
II. CALCITONIN GENE PRODUCTS The calcitonin gene family includes at least four separate genes. The first (CALC I or CGRP) encodes calcitonin and a second peptide, calcitonin gene-related peptide (CGRP) (Fig. 1). The second gene (CALC II or CGRP), undoubtedly a duplication of the first, has a similar structure to the first, but the primary RNA transcript of this gene is
OSTEOPOROSIS, SECOND EDITION VOLUME 1
247
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
248
HOFF, COTE, AND GAGEL
FIGURE 1 Calcitonin gene expression. The CT/CGRP gene is transcribed in a variety of cell types, including thyroidal C cells and neurons. The primary transcript of CT/CGRP is processed to include (CT-specific pathway) or exclude (CGRP-specific pathway) exon 4 encoding CT. Following translation of either CT or CGRP mRNA, the propeptide (proCT or proCGR) undergoes additional proteolytic cleavage. CT or CGRP is cleaved from the central portion of each propeptide and generates amino-and carboxy-terminal peptides (N- and C-proCT or N- and C-proCGRP).
genomic structure of the CT/CGRP gene led to the recognition that CT or CGRP is produced as a result of alternative RNA processing (Fig. 1). Following transcription there is constitutive RNA processing that combines exons 1 – 4 of the CT transcript to produce a mRNA encoding proCT (with polyadenylation resulting in truncation of the mRNA immediately following exon 4), a processing pathway that is used predominantly in the thyroidal C cell. A second processing choice results in the combining of exons 1 – 3 and 5 and 6 to produce a mRNA encoding CGRP. This processing pathway results in the exclusion of exon 4 (containing the sequences encoding CT) from the final mRNA, a cell-specific processing choice that occurs mainly in neural types of cells in the brain, spinal cord, and various sympathetic ganglia. In the
normal thyroidal C cell, 99% of the primary RNA transcript is processed down a CT-specific pathway; in neuronal tissue, 95% of the primary transcript is processed in a CGRPspecific manner [8 – 11]. The regulatory pathways that control this alternative splice in the C cell are complex and likely involve multiple factors that enhance (CT-specific pattern) or suppress (CGRP-specific pattern) recognition of exon 4 containing the coding sequence for calcitonin. Only one set of regulatory factors has been identified, those involved in recognition and polyadenylation of exon 4 (Fig. 1) [12]. Both CT and CGRP mRNAs encode propeptides that are subsequently processed further [13]. The CT monomer (the hypocalcemic factor identified by Copp) is located centrally
249
CHAPTER 8 Calcitonin
in the proCT molecule. Cleavage of the CT monomer results in the production of two additional peptides: amino (N)-and carboxy (C)-terminal proCT (Fig. 1). Although CproCT has been shown to have some bone effects, it is unclear at present whether these effects are significant. CGRP is similarly cleaved from a central location within the proCGRP molecule, producing an amino- and carboxy-terminal fragment. There is considerable similarity between the structure of monomeric CT or CGRP. Both contain an amino-terminal ring structure defined by disulfide bonds, considerable amino-terminal homology, and both are amidated. These findings suggest that either CT or CGRP originated as a result of an intragenic duplication. The only product derived from the second CT/CGRP gene is CGRP, a peptide with a high degree of homology to CGRP. Although there is a CT-like exon (exon 4) in the second gene, it is not included in the final mRNA. Therefore, the only product from this gene is CGRP. Amylin is the secretory product of the fourth calcitonin gene. This gene is expressed predominantly in islet cells of the pancreas where amyloid composed of amylin sheets has been implicated in the pathogenesis of diabetes mellitus. There is no known function of amylin in calcium metabolism, although its interaction with the CT and CGRP receptors is discussed later. Adrenomedullin is a 52 amino acid peptide with an amino-terminal ring structure and approximately 25% homology to CGRP. It is a potent vadodilator [14]. Although it has no known relationship to calcium metabolism and is expressed from a separate gene, it is considered a member
of the CT/CGRP superfamily because of its biological similarity to CGRP and its interaction with the CT/CGRP family of receptors [15,16].
III. CALCITONIN STRUCTURE/ FUNCTION RELATIONSHIPS Based on primary structure, the CTs can be divided into three groups: artiodactyl, including porcine, bovine, and ovine, which differ by four amino acids; primate/rodent, including human and rat calcitonins, which differ by two amino acids; and teleost/avian, including salmon, eel, goldfish, and chicken calcitonins, which differ by three amino acids (Fig. 2). In several different biological assays, the order of potency of the CTs is teleost artiodactyl human, although the absolute potency of each is dependent on the species studied. Fish and avian CTs are products of the ultimobranchial glands, which remain in those species as discrete organs, whereas in mammals, C cells migrate from the neural crest to be dispersed within the developing thyroid gland. Studies of substituted, deleted, and otherwise modified CTs have provided considerable information regarding structure/activity relationships of the CT molecule. For example, although the ring structure appears to protect and stabilize the molecule, linear analogues of salmon CT retain full hypocalcemic activity and adenylate cyclase activation [17,18]. Consistent with this, increased stability of the ring (as in aminosuberic 1 – 7 eel calcitonin) increases biological stability and potency. The carboxy-terminal
FIGURE 2 Amino acid alignment of calcitonins from various species. All calcitonins have a cysteine (C) at positions 1 and 7 linked by a disulfide bridge, a glycine (G) at position 28, and a proline amide (P) at position 32. Residues 4, 5, and 6 are conserved in all species. Residues identical to salmon calcitonin sCT) are boxed. eCT, eel calcitonin; gCT, goldfish calcitonin; stCT, stingray calcitonin; cCT, chicken calcitonin; pCT, porcine calcitonin; oCT, ovine calcitonin; bCT, bovine calcitonin; dCT, dog calcitonin; hCT, human calcitonin; rCT, rat calcitonin; and rbCT, rabbit calcitonin.
250 TABLE 1
HOFF, COTE, AND GAGEL
General Structure-Function Relationships for Monmeric Calcitonin
Integrity of the amino-terminal ring structure is important for the stability in circulation in some species Greater stability of the ring (as in aminosuberic 1–7 eel calcitonin) enhances potency Full biological activity can be preserved despite substitutions or deletions of one residue within the ring The sequence Leu-Ser-Thr–Cys at positions 4–7 in the ring structure is invariant Deletions in other parts of the molecule may impair biological activity, depending on spatial arrangements of hydrophobic–hydrophilic residues Single or multiple substitutions in the human CT molecule can enhance biological activity Carboxyl-terminal proline amide is essential for activity
proline amide is essential, but numerous changes are tolerated in the 8 – 22 domain. General observations regarding the biological activity of the various CT species are outlined in Table 1.
IV. SECRETION AND METABOLISM Calcitonin secretion is regulated by the extracellular calcium concentration. Increasing the extracellular calcium concentration results in increased secretion of CT [19]. This effect is mediated by the interaction of calcium with the calcium-sensing receptor (CaR), a G-protein-coupled receptor [20] expressed in high levels on the thyroidal C cell [21,22]. In contrast to the parathyroid cell, where increasing calcium concentration activates the CaR and links to pathways that inhibit parathyroid hormone secretion, activation of this same receptor in the C cell causes increased secretion of CT [23,24]. Thus the same receptor system is linked differently in two cell types, a useful physiologic adaptation that results in reduced secretion of the parathyroid hormone (a hormone that raises the serum calcium concentration) and enhanced secretion of CT (a hormone that lowers serum calcium concentration) in the presence of hypercalcemia. Other peptides, including glucagon, gastrin, cholecystokinin, secretin, and vasoactive intestinal peptide, also stimulate CT release from the C cell, although the physiologic relevance of these interactions is unclear [25]. Glucagon appears to act through a cAMP-dependent pathway; the mechanism by which gastrin stimulates CT release is less clear. Therapy with omeprazole and other proton pump inhibitors that inhibit gastric acid production increases CT secretion through increased gastrin secretion. A broad spectrum of evidence indicates that voltagegated calcium channels and the release of calcium from
intracellular pools are important in both calcium and cAMP-mediated hormone release [23,26]. Alcohol and exercise stimulate CT release [27], although the mechanisms are not understood. Pentagastrin is used routinely to detect CT abnormalities in patients with medullary thyroid carcinoma [28,29]. Calcitonin gene expression is also affected by a variety of factors. Activation of cAMP and protein kinase C pathways [30,31] and glucocorticoid treatment [32,33] increases transcription of the CT gene, whereas 1,25dihydroxy vitamin D3 inhibits transcription [34]. During the past decade a body of evidence has developed that serum CT concentrations are increased in sepsis. Although there was initial skepticism about whether the immunoreactive material was CT, studies using more specifi assays indicate that the circulating peptide is proCT rather than CT 1 – 32 (monomeric CT) and that the CT/CGRP gene is transcribed in infected tissues [35]. Even more provocative are results that show infected animals have greater survival if the proCT is neutralized by passive immunization [36]. Circulating plasma or serum CT values are higher in men than in women [37,38] and decrease with age [37,38], although these differences disappear when more specifi CT assays are utilized to measure calcitonin [37]. There are large increases in plasma immunoreactive CT during pregnancy and lactation, and it has been proposed that protection of the maternal skeleton from excessive calcium depletion may be the major physiological role of CT [39]. Estrogen therapy has been reported to increase basal CT concentrations in postmenopausal women [40], although studies utilizing a sensitive extraction assay for CT have not confirmed these observations [41].
V. PHYSIOLOGY AND MECHANISM OF ACTION Calcitonin and its related peptides (CGRP, CGRP amylin, and adrenomedullin) are small peptides that exert their biological effect by binding to and activating a family of G-protein-coupled receptors (Fig. 3). The first identified member of this family was the CT receptor (CTR) [42,43]. This receptor is expressed on osteoclasts, renal tubular cells, neurons in the central and peripheral nervous system, and other cell types [44]. Several different isoforms of the CT receptor are produced by alternative RNA processing [45,46], which couple differently to downstream transduction pathways [47,48]. The CTR is expressed on the osteoclast during the final stages of its differentiation. Binding of CT to the osteoclast CTR results in the activation of several downstream signaling pathways, leading to a decrease in bone resorption. This has been demonstrated in several ways. Addition of
251
CHAPTER 8 Calcitonin
FIGURE 3
The relationship between the calcitonin (CTR) or CT-like receptor (hCTR2 or rCTLR) and receptor activity modifying proteins (RAMP). Current information indicates that different combinations of CTR2 or CTLR and RAMPs 1 – 3 create receptors for CGRP, adrenomedullin, and amylin.
CT to in vitro bone culture assays (consisting of calvarial or long bones from newborn mice) inhibits bone resorption caused by a variety of resorptive agents, including parathyroid hormone, parathyroid hormone-related protein, interleukins, 1,25-dihydroxy vitamin D, and prostaglandins [49]. The effect of CT is of limited duration in these systems. Within 48 h following repetitive administration of CT to the bone culture system, there is a loss of the inhibitory effect of CT to prevent resorptive agent-mediated bone resorption termed “escape” [50]. This effect appears to be mediated by a decrease in CTR caused by some
combination of receptor recycling [51,52] and reduced synthesis of CTR [53 – 55]. Studies of isolated osteoclasts have also demonstrated rapid effects of CT. One model system utilizes isolated osteoclasts placed on coverslips or bone slices [56]. Within minutes after exposure to CT at concentrations as low as 3 pg/ml there is a retraction of the osteoclast ruffled border from its attachment points and decreased bone resorption [56]. Other studies have demonstrated reductions in the production of proteolytic enzymes, an increase in the pH of the contents of the resorption pit induced by CT, and
252
HOFF, COTE, AND GAGEL
alterations in the integrins, which mediate attachment of the osteoclast to the bone surface. Studies in a transgenic animal model in which the CT/CGRP gene has been deleted by homologous recombination provide additional support for the acute effect of CT to inhibit bone resorption. In mice treated with 1,25dihydroxy vitamin D there is greater hypercalcemia in CT/CGRP -/- mice than in wild-type control animals. Further evidence that the hypercalcemia is mediated by increased bone resorption is provided by the demonstration that urine deoxypyridinoline crosslinks are severalfold greater in CT/CGRP -/- mice than in wild-type control animals [57]. Collectively, data accumulated since the discovery in the 1960s argue for a role of CT to protect against hypercalcemia [58]. Acute increases in the serum calcium concentration cause release of CT from C cells. Binding of CT to the CTR on the osteoclast causes rapid inhibition of bone resorption. The effect of CT on the osteoclast is transient
with reversal of the effect within 48 h, exactly the type of response that is optimal for maintenance of the serum calcium concentration within a narrow range. Studies in CT/CGRP -/- mice hint at other effects of the CT/CGRP gene. CT/CGRP -/- mice have a greater bone mass at 1 and 3 months. Furthermore, in oophorectomized CT/CGRP -/- mice, there is preservation of bone mass, whereas wild-type control mice lose the expected amount of bone mass under the same conditions. The preservation of bone mass appears to be mediated by increased bone formation in the oophorectomized CT/ CGRP -/- mice [57]. Interestingly, there is no evidence of increased bone resorption in CT/CGRP -/- mice. These studies suggest a more complex effect of the CT/CGRP gene that includes effects on bone formation. These studies are currently in their infancy, but provide provocative new information about a broader role for this gene. CT also has effects on renal tubular function, causing natriuresis through an effect on the Na/H exchanger and
FIGURE 4 Calcitonin receptor isoforms of rat (top) and human (bottom) with extracellular domains (e1 to e4) and intracellular domains (i1 to i4) indicated. The C1b isoform contains a 37 amino acid insert in e2, compared to C1a. The human CTR cloned from BIN67 cells contains a 16 amino acid insert in i1 compared with the receptor cloned from T47D cells.
253
CHAPTER 8 Calcitonin
Na,K-ATPase [59]. Binding of CT to the renal CTR results in the activation of protein kinase A and C pathways in a cell cycle-dependent manner [60]. In addition, there is also activation of the ERK 1/2 pathways [61]. The exact physiologic role of CT in Na transport has been more difficult to establish. CT/CGRP -/- mice have no identifiable electrolyte abnormalities, although detailed studies of tubular function have not been performed in these animals.
VI. THE CALCITONIN RECEPTOR SUPERFAMILY The CTR superfamily includes receptors for CT, CGRP, amylin, and adrenomedullin (Fig. 3). The CTR was the first identified member of this receptor family [43, 62] and is unique in that it functions without an accessory protein (Fig. 3). Efforts to identify a receptor for CGRP, amylin, and adrenomedullin were unsuccessful until it was recognized that a family of receptor activity-modifying proteins (RAMPS) join with CT-like receptors (CTLR) to form receptors for these proteins. Coexpression of RAMP 1, 2, or 3 with a CT-like receptor (hCTR2 or rCTLR) creates highaffinity binding sites for CGRP [63], amylin [63], and adrenomedullin [64 – 66]. A current view of how these receptor proteins interact with each other and the four ligands is shown in Fig. 3. The CTR RNA is further modified, presumably by alternative RNA processing, to produce several unique forms of the receptor that differ by insertion or deletion of small components of the receptor (Fig. 4). The alternative RNA processing results in two major receptor modifications. The first is the insertion of a 37 amino acid in the second extracellular domain (designated C1b in Fig. 4). Cells expressing the C1b variant have a lower adenylate cyclase response [67]. In the human there is a second receptor variant with a 16 amino acid insert in the first intracellular domain (Fig. 4) [48]. The most commonly expressed form of the human CTR appears to be the variant without the 16 amino acid insert (Fig. 4, T47D). Analysis suggests that the presence of the 16 amino acid insert in the first intracellular loop of the hCTR results in greater affinity for CT binding, but lowers phospholipase D and protein kinase C activation. This effect is thought to be related to a modification of a G-protein-binding sequence by the 16 amino acid insert [68].
VII. SUMMARY In summary, the CTR is a member of the class of seven trans-membrane G-protein-coupled receptors. CTRs in osteoclasts and models of renal tubular epithelium link to both protein kinase A- and C-dependent pathways. In the
osteoclast the CTR links to cellular processes, leading to osteoclast detachment from the matrix surface and cessation of bone resorption; in the renal tubular cell, activation of the CTR affects sodium transport and cell growth.
References 1. D. H. Copp, A. G. F. Davidson, and B. A. Cheney, Evidence for a new parathyroid hormone which lowers blood calcium. Proc. Can. Fed. Biol. Soc. 4, 17 (1961). 2. D. H. Copp, E. C. Cameron, B. A. Cheney, A. G. F. Davidson, and K. G. Henze, Evidence for calcitonin-a new hormone from the parathyroid that lowers blood calcium. Endocrinology 70, 638 – 649 (1962). 3. P. F. Hirsch, G. F. Gauthier, and P. L. Munson, Thyroid hypocalcemic principle and recurrent laryngeal nerve injury as factors affecting the response to parathyroidectomy in rats. Endocrinology 73, 244 – 252 (1963). 4. G. V. Foster, A. Baghdiantz, M. A. Kumar, E. Slack, H. A. Soliman, and I. MacIntyre, Thyroid origin of calcitonin. Nature (London) 202, 1303 – 1305 (1964). 5. C. S. Le-Lievre and N. M. Le-Douarin, Mesenchymal derivatives of the neural crest: Analysis of chimaeric quail and chick embryos. J Embryol. Exp. Morphol. 34, 125 – 54 (1975). 6. J. W. Jacobs, R. H. Goodman, W. W. Chin, P. C. Dee, J. F. Habener, N. H. Bell, and J. T. Potts, Jr., Calcitonin messenger RNA encodes multiple polypeptides in a single precursor. Science 213, 457 – 459 (1981). 7. S. G. Amara, V. Jonas, M. G. Rosenfeld, E. S. Ong, and R. M. Evans, Alternative RNA processing in calcitonin gene expression generates mRNAs encoding different polypeptide products. Nature 298, 240 – 244 (1982). 8. H. Lou, K. M. Neugebauer, R. F. Gagel, and S. M. Berget, Regulation of alternative polyadenylation by U1 snRNPs and SRp20. Mol. Cell. Biol. 18, 4977 – 4985 (1998). 9. H. Lou and R. F. Gagel, Alternative RNA processing: Its role in regulating expression of calcitonin/calcitonin gene-related peptide. J. Endocrinol. 156, 401 – 405 (1998). 10. H. Lou, D. M. Helfman, R. F. Gagel, and S. M. Berget, Polypyrimidine tract-binding protein positively regulates inclusion of an alternative 3-terminal exon. Mol. Cell. Biol. 19, 78 – 85 (1999). 11. J. M. Yeakley, F. Hedjran, J. P. Morfin, N. Merillat, M. G. Rosenfeld, and R. B. Emeson, Control of calcitonin/calcitonin gene-related peptide pre-mRNA processing by constitutive intron and exon elements. Mol. Cell. Biol. 13, 5999 – 6011 (1993). 12. H. Lou, R. F. Gagel, and S. M. Berget, An intron enhancer recognized by splicing factors activates polyadenylation. Genes Dev. 10, 208 – 219 (1996). 13. M. Zaidi, B. S. Moonga, P. J. Bevis, Z. A. Bascal, and L. H. Breimer, The calcitonin gene peptides: Biology and clinical relevance. Crit. Rev. Clin. Lab. Sci. 28, 109 – 174 (1990). 14. K. Kitamura, J. Sakata, K. Kangawa, M. Kojima, H. Matsuo, and T. Eto, Cloning and characterization of cDNA encoding a precursor for human adrenomedullin. Biochem. Biophys. Res. Commun. 194, 720 – 725 (1993). 15. S. J. Wimalawansa, Amylin, calcitonin gene-related peptide, calcitonin, and adrenomedullin: A peptide superfamily. Crit. Rev. Neurobiol. 11, 167 – 239 (1997). 16. R. Muff, W. Born, and J. A. Fischer, Calcitonin, calcitonin generelated peptide, adrenomedullin and amylin: Homologous peptides, separate receptors and overlapping biological actions. Eur. J. Endocrinol. 133, 17 – 20 (1995).
254 17. P. M. Sexton, H. G. Schneider, C. S. DSantos, F. A. Mendelsohn, B. E. Kemp, J. M. Moseley, T. J. Martin, and D. M. Findlay, Reversible calcitonin binding to solubilized sheep brain binding sites. Biochem. J. 273, 179 – 184 (1991). 18. R. C. Orlowski, R. M. Epand, and A. R. Stafford, Biologically potent analogues of salmon calcitonin which do not contain an N-terminal disulfide-bridged ring structure. Eur. J. Biochem. 162, 399 – 402 (1987). 19. C. W. Cooper, L. J. Deftos, and J. T. Potts, Jr., Direct measurement of in vivo secretion of pig thyrocalcitonin by radioimmunoassay. Endocrinology 88, 747 – 754 (1971). 20. E. M. Brown, G. Gamba, D. Riccardi, M. Lombardi, R. Butters, O. Kifor, A. Sun, M. A. Hediger, J. Lytton, and S. C. Hebert, Cloning and characterization of an extracellular Ca(2)-sensing receptor from bovine parathyroid. Nature 366, 575 – 580 (1993). 21. J. E. Garrett, H. Tamir, O. Kifor, R. T. Simin, K. V. Rogers, A. Mithal, R. F. Gagel, and E. M. Brown, Calcitonin-secreting cells of the thyroid express an extracellular calcium receptor gene. Endocrinology 136, 5202 – 5211 (1995). 22. M. Freichel, A. Zink-Lorenz, A. Holloschi, M. Hafner, V. Flockerzi, and F. Raue, Expression of a calcium-sensing receptor in a human medullary thyroid carcinoma cell line and its contribution to calcitonin secretion. Endocrinology 137, 3842 – 3848 (1996). 23. F. Raue and H. Scherubl, Extracellular calcium sensitivity and voltage-dependent calcium channels in C cells. Endocr. Rev. 16, 752 – 764 (1995). 24. D. S. McGehee, M. Aldersberg, K. P. Liu, S. Hsuing, M. J. Heath, and H. Tamir, Mechanism of extracellular Ca2 receptor-stimulated hormone release from sheep thyroid parafollicular cells. J. Physiol. 502, 31 – 44 (1997). 25. B. A. Roos, L. J. Deftos, and P. Rorabaugh, Calcitonin secretion in vitro. II. Regulatory effects of enteric mammalian polypeptide hormones on trout C-cell cultures. Endocrinology 98, 1284 – 1288 (1976). 26. R. M. Fried and A. H. J. Tashjian, Unusual sensitivity of cytosolic free Ca2 to changes in extracellular Ca2 in rat C-cells. J. Biol. Chem. 261, 7669 – 7674 (1986). 27. J. F. Aloia, P. Rasulo, L. J. Deftos, A. Vaswani, and J. K. Yeh, Exercise-induced hypercalcemia and the calciotropic hormones. J. Lab. Clin. Med. 106, 229 – 232 (1985). 28. J. F. Hennessey, S. A. Wells, D. A. Ontjes, and C. W. Cooper, A comparison of pentagastrin injections and calcium infusion as provocative agents for the detection of medullary carcinoma of the thyroid. J. Clin. Endocr. Metab. 39, 487 – 495 (1974). 29. R. F. Gagel, A. H. Tashjian, Jr., T. Cummings, N. Papathanasopoulos, M. M. Kaplan, R. A. DeLellis, H. J. Wolfe, and S. Reichlin, The clinical outcome of prospective screening for multiple endocrine neoplasia type 2a: An 18-year experience. N. Engl. J. Med. 318, 478 – 484 (1988). 30. Y. T. Monla, S. Peleg, and R. F. Gagel, Cell type-specific regulation of transcription by cyclic adenosine 3,5-monophosphateresponsive elements within the calcitonin promoter. Mol. Endocrinol. 9, 784 – 793 (1995). 31. A. de Bustros, S. B. Baylin, C. L. Berger, B. A. Roos, S. S. Leong, and B. D. Nelkin, Phorbol esters increase calcitonin gene transcription and decrease c-myc mRNA levels in cultured human medullary thyroid carcinoma. J. Biol. Chem. 260, 98–104 (1985). 32. M. Muszynski, R. S. Birnbaum, and B. A. Roos, Glucocorticoids stimulate the production of preprocalcitonin-derived secretory peptides by a rat medullary thyroid carcinoma cell line. J. Biol. Chem. 258, 11678 – 11683 (1983). 33. G. J. Cote and R. F. Gagel, Dexamethasone differentially affects the levesl of calcitonin and calcitonin gene-related peptide mRNAs expressed in a human medullary thyroid carcinoma cell line. J. Biol. Chem. 261, 15524–15528 (1986).
HOFF, COTE, AND GAGEL 34. G. J. Cote, D. G. Rogers, E. S. Huang, and R. F. Gagel, The effect of 1,25-dihydroxyvitamin D3 treatment on calcitonin and calcitonin gene-related peptide mRNA levels in cultured human thyroid C-cells. Biochem. Biophys. Res. Commun. 149, 239 – 243 (1987). 35. R. H. Snider, Jr., E. S. Nylen, and K. L. Becker, Procalcitonin and its component peptides in systemic inflammation: Immunochemical characterization. J. Invest. Med. 45, 552 – 560 (1997). 36. E. S. Nylen, K. T. Whang, R. H. Snider, Jr., P. M. Steinwald, J. C. White, and K. L. Becker, Mortality is increased by procalcitonin and decreased by an antiserum reactive to procalcitonin in experimental sepsis. Crit. Care Med. 26, 1001 – 1006 (1998). 37. R. D. Tiegs, J. J. Body, J. M. Barta, and H. D. Heath, Secretion and metabolism of monomeric human calcitonin: Effects of age, sex, and thyroid damage. J. Bone Miner. Res. 1, 339 – 349 (1986). 38. L. J. Deftos, M. H. Weisman, G. W. Williams, D. B. Karpf, A. M. Frumar, B. J. Davidson, J. G. Parthemore, and H. L. Judd, Influence of age and sex on plasma calcitonin in human beings. N. Engl. J. Med. 302, 1351 – 1353 (1980). 39. J. C. Stevenson, C. J. Hillyard, I. MacIntyre, H. Cooper, and M. I. Whitehead, A physiological role for calcitonin: Protection of the maternal skeleton. Lancet 2, 769 – 770 (1979). 40. J. C. Stevenson, C. J. Hillyard, G. Abeyasekera, K. G. Phang, and I. MacIntyre, Calcitonin and the calcium-regulating hormones in postmenopausal women: Effect of oestrogens. Lancet 1, 693 – 695 (1981). 41. D. L. Hurley, R. D. Tiegs, J. Barta, K. Laakso, and H. Heath III, Effects of oral contraceptive and estrogen administration on plasma calcitonin in pre- and postmenopausal women. J. Bone Miner. Res. 4, 89 – 95 (1989). 42. H. Y. Lin, T. L. Harris, M. S. Flannery, A. Aruffo, E. H. Kaji, A. Gorn, L. F. Kolakowski, Jr., M. Yamin, H. F. Lodish, and S. R. Goldring, Expression cloning and characterization of a porcine renal calcitonin receptor. Trans. Assoc. Am. Phys. 104, 265 – 272 (1991). 43. H. Y. Lin, T. L. Harris, M. S. Flannery, A. Aruffo, E. H. Kaji, A. Gorn, L. F. Kolakowski, Jr., H. F. Lodish, and S. R. Goldring, Expression cloning of an adenylate cyclase-coupled calcitonin receptor. Science 254, 1022 – 1024 (1991). 44. S. R. Goldring, A. H. Gorn, M. Yamin, S. M. Krane, and J. T. Wang, Characterization of the structural and functional properties of cloned calcitonin receptor cDNAs. Horm. Metab. Res. 25, 477 – 480 (1993). 45. M. Ikegame, M. Rakopoulos, H. Zhou, S. Houssami, T. J. Martin, J. M. Moseley, and D. M. Findlay, Calcitonin receptor isoforms in mouse and rat osteoclasts. J. Bone Miner. Res. 10, 59 – 65 (1995). 46. S. K. Lee, S. R. Goldring, and J. A. Lorenzo, Expression of the calcitonin receptor in bone marrow cell cultures and in bone: A specifi marker of the differentiated osteoclast that is regulated by calcitonin. Endocrinology 136, 4572 – 4581 (1995). 47. P. Orcel, H. Tajima, Y. Murayama, T. Fujita, S. M. Krane, E. Ogata, S. R. Goldring, and I. Nishimoto, Multiple domains interacting with Gs in the porcine calcitonin receptor. Mol. Endocrinol. 14, 170 – 182 (2000). 48. E. E. Moore, R. E. Kuestner, S. D. Stroop, F. J. Grant, S. L. Matthewes, C. L. Brady, P. M. Sexton, and D. M. Findlay, Functionally different isoforms of the human calcitonin receptor result from alternative splicing of the gene transcript. Mol. Endocrinol. 9, 959 – 968 (1995). 49. J. Friedman and L. G. Raisz, Thyrocalcitonin: Inhibitor of bone resorption in tissue culture. Science 150, 1465 – 1467 (1965). 50. J. A. Wener, S. J. Gorton, and L. G. Raisz, Escape from inhibition or resorption in cultures of fetal bone treated with calcitoninand parathyroid hromone. Endocrinology 90, 752 – 759 (1972). 51. A. H. J. Tashjian, D. R. Wright, J. L. Ivey, and A. Pont, Calcitonin binding sites in bone: Relationships to biological response and “escape.” Recent Prog. Horm. Res. 34, 285 – 334 (1978).
255
CHAPTER 8 Calcitonin 52. N. S. Krieger, R. S. Feldman, and A. H. Tashjian, Jr., Parathyroid hormone and calcitonin interactions in bone: Irradiation-induced inhibition of escape in vitro. Calcif. Tissue Int. 34, 197 – 203 (1982). 53. D. Inoue, C. Shih, D. L. Galson, S. R. Goldring, W. C. Horne, and R. Baron, Calcitonin-dependent down-regulation of the mouse C1a calcitonin receptor in cells of the osteoclast lineage involves a transcriptional mechanism. Endocrinology 140, 1060 – 1068 (1999). 54. S. Wada, N. Udagawa, N. Nagata, T. J. Martin, and D. M. Findlay, Calcitonin receptor down-regulation relates to calcitonin resistance in mature mouse osteoclasts. Endocrinology 137, 1042 – 1048 (1996). 55. S. Wada, T. J. Martin, and D. M. Findlay, Homologous regulation of the calcitonin receptor in mouse osteoclast-like cells and human breast cancer T47D cells. Endocrinology 136, 2611 – 2621 (1995). 56. T. J. Chambers and A. Moore, The sensitivity of isolated osteoclasts to morphological transformation by calcitonin. J. Clin. Endocrinol. Metab. 57, 819 – 824 (1983). 57. A. O. Hoff, M. Amling, and R. F. Gagel, CT/CGRP knockout mice. Personal communication (2000). 58. B. A. Roos, M. Yoon, S. V. Cutshaw, and D. N. Kalu, Calcium regulatory action of endogenous rat calcitonin demonstrated by passive immunization with calcitonin antibodies. Endocrinology 107, 1320 – 1326 (1980). 59. M. Chakraborty, D. Chatterjee, F. S. Gorelick, and R. Baron, Cell cycle-dependent and kinase-specific regulation of the apical Na/H exchanger and the Na,K-ATPase in the kidney cell line LLC-PK1 by calcitonin. Proc. Natl. Acad. Sci. USA 91, 2115 – 2119 (1994). 60. M. Chakraborty, D. Chatterjee, S. Kellokumpu, H. Rasmussen, and R. Baron, Cell cycle-dependent coupling of the calcitonin receptor to different G proteins. Science 251, 1078 – 1082 (1991). 61. Y. Chen, J. F. Shyu, A. Santhanagopal, D. Inoue, J. P. David, S. J. Dixon, W. C. Horne, and R. Baron, The calcitonin receptor stimulates Shc tyrosine phosphorylation and Erk1/2 activation: Involvement of
62.
63.
64.
65.
66.
67.
68.
Gi, protein kinase C, and calcium. J. Biol. Chem. 273, 19809 – 19816 (1998). A. H. Gorn, H. Y. Lin, M. Yamin, P. E. Auron, M. R. Flannery, D. R. Tapp, C. A. Manning, H. F. Lodish, S. M. Krane, and S. R. Goldring, Cloning, characterization, and expression of a human calcitonin receptor from an ovarian carcinoma cell line. J. Clin. Invest. 90, 1726 – 1735 (1992). L. M. McLatchie, N. J. Fraser, M. J. Main, A. Wise, J. Brown, N. Thompson, R. Solari, M. G. Lee, and S. M. Foord, RAMPs regulate the transport and ligand specificity of the calcitonin-receptor-like receptor. Nature 393, 333 – 339 (1998). N. Buhlmann, K. Leuthauser, R. Muff, J. A. Fischer, and W. Born, A receptor activity modifying protein (RAMP)2-dependent adrenomedullin receptor is a calcitonin gene-related peptide receptor when coexpressed with human RAMP1. Endocrinology 140, 2883 – 2890 (1999). K. Leuthauser, R. Gujer, A. Aldecoa, R. A. McKinney, R. Muff, J. A. Fischer, and W. Born, Receptor-activity-modifying protein 1 forms heterodimers with two G-protein-coupled receptors to define ligand recognition. Biochem. J. 351, 347 – 351 (2000). R. Muff, N. Buhlmann, J. A. Fischer, and W. Born, An amylin receptor is revealed following co-transfection of a calcitonin receptor with receptor activity modifying proteins-1 or -3. Endocrinology 140, 2924 – 2927 (1999). S. Houssami, D. M. Findlay, C. L. Brady, T. J. Martin, R. M. Epand, E. E. Moore, E. Murayama, T. Tamura, R. C. Orlowski, and P. M. Sexton, Divergent structural requirements exist for calcitonin receptor binding specificity and adenylate cyclase activation. Mol. Pharmacol. 47, 798 – 809 (1995). F. Naro, M. Perez, S. Migliaccio, D. L. Galson, P. Orcel, A. Teti, and S. R. Goldring, Phospholipase D-and protein kinase C isoenzymedependent signal transduction pathways activated by the calcitonin receptor. Endocrinology 139, 3241 – 3248 (1998).
CHAPTER 9
Vitamin D Biology, Action, and Clinical Implications DAVID FELDMAN, PETER J. MALLOY, AND COLEMAN GROSS Department of Medicine, Division of Endocrinology, Stanford University School of Medicine, Stanford, California 94305
I. II. III. IV. V. VI. VII.
VIII. 1,25(OH)2D3 Analogues with Decreased Calcemic Activity IX. Actions of Vitamin D in Classical Target Organs to Regulate Mineral Homeostasis X. Actions of 1,25(OH)2D in Nonclassical Target Organs XI. Vitamin D and Osteoporosis References
Introduction Vitamin D Metabolism Pathways of Activation and Inactivation of Vitamin D Mechanism of 1,25(OH)2D Action Nongenomic Effects of Vitamin D Physiology: Regulation of Serum Calcium Genetic Disorders and Vitamin D Receptor Polymorphisms
I. INTRODUCTION
bone or mineral metabolism, including antiproliferative, prodifferentiating, and immunosuppressive activities. This chapter describes the basic biology of vitamin D, including its metabolism, physiology, mechanism of action, and its diverse functions in the body, including those actions that relate to mineral metabolism as well as the newer actions. Several reviews of vitamin D mechanism of action and function have been published [1 – 9] as well as a comprehensive book addressing all areas of vitamin D [10]. Specific issues relating to vitamin D and osteoporosis are discussed in Chapter 68.
Vitamin D is one of the major regulators of calcium homeostasis in the body and is critically important for normal mineralization of bone. The active hormone, 1,25-dihydroxyvitamin D [1,25(OH)2D], is produced by sequential hydroxylations of vitamin D in the liver (25hydroxylation) and the kidney (1-hydroxylation). 1,25(OH)2D, acting through the vitamin D receptor (VDR), acts by a genomic mechanism identical to the classical steroid hormones to regulate target gene transcription. The traditional actions of 1,25(OH)2D are to enhance calcium and phosphate absorption from the intestine in order to maintain normal concentrations in the circulation and to provide adequate amounts of these minerals to the boneforming site for the mineralization of bone to proceed normally. However, in the past decade, it has become increasing clear that vitamin D has many additional functions that implicate the hormone in a wide array of actions relating to bone formation as well as to other areas unrelated to
OSTEOPOROSIS, SECOND EDITION VOLUME 1
A. Chemistry, Structure, and Terminology Vitamin D exists in two forms: vitamin D3 (cholecalciferol) and vitamin D2 (ergocalciferol). When written without a subscript the designation vitamin D denotes either D2 or D3. Sunlight, in the form of UV B rays, cleaves the B ring between carbon 9 and 10 to open the ring and create a
257
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
258
FELDMAN, MALLOY, AND GROSS
B. History The unfolding of the story of vitamin D from its discovery as an antirachitic factor and designation as a vitamin to its transition from being considered a vitamin to its recognition as a hormone has all occurred within the past 75 years. However, the substance appears to be evolutionarily very ancient, produced by phytoplankton exposed to sunlight approximately 750 million years ago [12]. The history of the identification of vitamin D, the beneficial effects of sunlight on rickets, the elucidation of the pathway of conversion of vitamin D to 1,25(OH)2D, and the realization that vitamin D is a steroid hormone have been detailed in multiple reviews [12 – 15].
II. VITAMIN D METABOLISM A. Dietary Sources
FIGURE 1
1,25(OH)2D metabolic pathways. UV B indicates ultraviolet radiation (wavelength 290 – 320 nm) emitted from the sun. Liver 25 refers to hepatic 25-hydroxylase, and kidney 24R and 1 are renal 24-hydroxylase and 1-hydroxylase, respectively. Reproduced with permission from M. F. Holick, In “Endocrinology” (L. J. DeGroot et al., eds.). Saunders, Philadelphia, 1995.
secosteroid structure (Fig. 1). By this process the precursor (provitamin) molecules, 7-dehydrocholesterol in animals and ergosterol in plants, are converted to the secosteroids, vitamin D3 and vitamin D2, respectively [11]. The two secosteroids differ only in the presence of a methyl group at carbon 28 and a double bond between carbon 22 and 23 on the side chain of vitamin D2. Vitamin D2 and vitamin D3 are handled identically in the body and converted, via two hydroxylation steps, to the active hormones, 1,25(OH)2D2 or 1,25(OH)2D3 (calcitriol).
There are two sources of vitamin D: dietary intake and endogenous production (Fig. 1). Endogenous vitamin D production occurs in the skin as a result of UV light exposure, and this synthetic process, which distinguishes vitamin D from the true vitamins, will be discussed later (Section IIB). Whereas only vitamin D3 is produced in the skin, both vitamin D2 from plant sources and vitamin D3 from animal sources are available in the diet. Vitamins D2 and D3 are biologically inactive and, as discussed in detail later, must be converted to hydroxylated metabolities to exhibit hormonal activity. Foods naturally containing substantial amounts of vitamin D are relatively few: egg yolks, liver, and fish liver oils (cod liver oil). In the United States the primary dietary source of vitamin D is fortified milk, which nominally contains 400 IU/quart, although this has been found to vary considerably [16]. Other supplemented sources may include cereals, breads, and fortified margarine. The recommended daily allowance for adults is 200 IU/day, for pregnant and lactating women it is 400 IU/day, for infants less than 6 months it is 300 IU/day, and for children over 6 months it is 400 IU/day. However, on average, adult intake is estimated to be less than 100 IU/day, suggesting that dietary sources of vitamin D play a minor role in vitamin D homeostasis (see Section XI,D for consequences on bone). Findings indicate that the recommended daily intake of vitamin D may be insufficient [17], especially in the elderly [18], and when sunlight exposure is limited [19,20]. The problem is magnified in the homebound elderly [21,22]. Seasonal sunlight deficiency contributes to vitamin D insufficiency [23,24]. Many studies suggest that fortification is necessary to augment daily intake and maintain baseline stores of vitamin D [25]. Even in those taking supplements, especially the elderly or individuals who are ill and
259
CHAPTER 9 Vitamin D
hospitalized, hypovitaminosis D may be common [26] and may contribute to osteoporotic fracture [27]. Evidence that vitamin D supplementation reduces fractures has been accumulating [22,28]. It is unwise to assume that vitamin D status is normal, even if patients are taking 400 IU supplementation. Many authors have concluded that 800 IU/day would be an effective intake yet still safe. This subject is further discussed in Chapter 68. Vitamin D is fat soluble and dietary sources are absorbed via the lymphatics in the proximal small bowel. Important factors for absorption include (i) gastric, pancreatic, and biliary secretions, (ii) formation of micelles, (iii) diffusion through the unstirred layer adjacent to the intestinal mucosa, (iv) brush border membrane uptake, (v) incorporation into chylomicrons, and (vi) absorption into the lymphatics. The mechanism of intestinal calcium absorption and its regulation by vitamin D has been reviewed by Wasserman [29]. Disorders that interfere with the processes just described or that disrupt the small bowel mucosa can interfere with vitamin D absorption include cystic fibrosis, chronic pancreatitis with pancreatic insufficiency, biliary obstruction, sprue (gluten enteropathy), inflammatory bowel disease involving the small bowel, and gastrointestinal surgery [30]. Assessing the absorption of vitamin D may be clinically important in patients with these or related conditions. After an oral dose of vitamin D, blood levels begin to rise at 4 h peak by 12 h, and return close to baseline by 72 h. This pharmacokinetic profile provides a useful clinical test for assessing vitamin D absorption. The serum vitamin D concentration can be measured 12 h after an oral dose of 50,000 IU of vitamin D; a value of 50 ng/ml is indicative of normal vitamin D absorption, whereas malabsorption is indicated when values are 10 ng/ml [31]. The subject of disordered vitamin D absorption is discussed more fully in Chapter 48. Although most cases of rickets are due to vitamin D deficiency, a study of rachitic children in Nigeria suggests that calcium deficiency may also contribute to this condition. The children responded better to treatment with calcium alone or calcium and vitamin D than treatment with vitamin D alone [32].
B. Photobiology of Vitamin D: Endogenous Production The photobiology of vitamin D3 has been reviewed [33,34]. Ultraviolet (UV) radiation emitted from the sun and transmitted to the earth’s surface can be broadly divided into two spectra: UV A (wavelength 320 – 400 nm) and UV B (wavelength 290 – 320 nm). Light energy is transmitted to the epidermis and dermis where stores of 7-dehydrocholesterol (provitamin D3) are located (Fig. 1). UV B radiation causes scission of the C9 – C10 bond in the steroid, yielding
the “split” or secosteroid previtamin D3. Thermal equilibration within the skin occurs over a day, converting previtamin D3 to Vitamin D3. Vitamin D3 binds to the circulating vitamin D-binding protein (DBP) and thus leaves the skin and enters the circulation. During prolonged exposure to UV B radiation, previtamin D3 synthesis plateaus at about 15% of the 7-dehydrocholesterol skin content and leads to the increasing production of the biologically inert compounds lumisterol and a small amount of tachysterol from previtamin D3. This restriction on previtamin D3 formation may serve as a mechanism to prevent the overproduction of vitamin D3. Several factors have been found to affect the cutaneous synthesis of vitamin D3, including latitude and seasonal variation, skin pigmentation, the use of topical sunscreens, and age. In addition, 1,25(OH)2D may feed back on the skin to add to the regulation since it acts on epidermal constituents [35]. In addition, UV B radiation inhibits levels of VDR, suggesting the existence of a feedback mechanism in that UV B initiates vitamin D synthesis in keratinocytes and, at the same time, limits VDR abundance [36]. 1. LATITUDE AND SEASON Because the conversion of 7-dehydrocholesterol to previtamin D3 in the skin requires UV B radiation, the amount of previtamin D3 synthesized is related to the amount of UV B radiation absorbed by the skin. The amount of solar radiation reaching the surface of the earth is limited by the changing zenith angle of the sun and decreases with increasing global latitude. Similarly, the incident radiation on the surface of the earth is diminished during the fall and winter months when the sun is lower in the sky. Therefore, the variation in cutaneous UV B radiation exposure due to seasonal variation or geographical location can influence the amount of vitamin D3 synthesized in the skin. As a result, no previtamin D3 is synthesized in Boston (42° N latitude) from November to February, and 10° farther north, in Edmonton, this period is extended from October to March. In more southerly locations, such as Los Angeles and Puerto Rico, previtamin D3 synthesis occurs year round [12]. An interesting commentary on the relative importance of sunlight was described by Holick [12] in a study of naval personnel onboard submarines. Submariners who were not exposed to sunlight for 3 months failed to maintain adequate vitamin D levels even while ingesting 600 IU/day of vitamin D, supporting the concept that 800 IU/day may be necessary to maintain normal vitamin D levels in the absence of adequate sunlight. 2. SKIN PIGMENTATION The degree of skin pigmentation (i.e., melanin content) also affects cutaneous vitamin D3 production. Melanin protects the body from excess sunlight by acting as a sink to absorb UV B rays and acts as a competitor of 7dehydrocholesterol for UV B radiation. Therefore, the more
260
FELDMAN, MALLOY, AND GROSS
melanin that is present in the skin, the less UV B radiation is available for previtamin D3 synthesis. However, individuals with high melanin levels compensate by increasing the conversion of 25OHD to 1,25(OH)2D [37]. Loomis [38] raised the hypothesis that melanin pigmentation evolved in people living near the equator to prevent the excessive production of vitamin D due to constant exposure to sunlight. As people migrated away from the equatorial regions, their sunlight exposure was shortened and, in order to allow adequate production of vitamin D and prevent rickets, the melanin levels in their skin diminished. When individuals of different skin pigmentation were exposed to the same suberythemic dose of UV radiation (27 mJ/cm2), whites showed the largest incremental rise in serum vitamin D concentrations, whereas Asians showed an intermediate increase and black individuals the smallest rise [39]. Basal concentrations of 25(OH)D are lower in young healthy blacks than in young healthy whites; however, their 1,25(OH)2D concentrations are higher than in whites, possibly due to relative secondary hyperparathyroidism [40]. Increased skin pigmentation does not limit the absolute amount of previtamin D3 made, but rather it extends the period of sunlight exposure necessary to reach the maximum production of previtamin D3 [41]. This time interval for maximum previtamin D3 production ranges from 0.5 h in lightly pigmented individuals to 3 h or more in darker pigmented subjects. 3. SUNSCREENS, SUN EXPOSURE, AND AGE Interestingly, similar to melanin, topical sunscreens act as a competitor of the photochemical production of vitamin D3 by absorbing UV radiation. Para-amino benzoic acidbased preparations with an SPF 8 rating can significantly block the cutaneous production of vitamin D3. Age is also a variable that can influence the production of vitamin D3, as the amount of 7-dehydrocholesterol in the skin and the efficiency of previtamin D3 photoproduction decreases as a consequence of advancing age [33].
C. Transport in Circulation: Vitamin D-Binding Protein (DBP) Group-specific component (Gc), a 58-kDa plasma globulin, was originally described immunologically in 1959 [42], and approximately 16 years later Gc was identified as a vitamin D-binding protein [43]. There appears to be a single binding site for calciferols on each DBP molecule; however, only about 5% of the binding sites are normally occupied, probably due to the high concentration of DBP in the circulation [44,45]. The binding affinity of DBP for the vitamin D metabolites is as follows: 25(OH)D3 24,25(OH)2D3 vitamin D3 1,25(OH)2D3
1,24,25(OH)3D3. The affinity of DBP for vitamin D2 and vitamin D3 is similar. DBP is primarily synthesized in the liver [45], and serum concentrations of DBP increase in pregnancy and with estrogen treatment, whereas it is decreased in liver disease, malnutrition, and nephrotic syndrome. Calcitropic hormones, however, do not appear to regulate the synthesis of DBP. In addition to liver, DBP mRNA has been detected in several rat tissues, including testis, kidney, yolk sac, placenta, and adipose tissue [45]. Using the sensitive method of reverse transcriptasepolymerase chain reaction (RT-PCR), DBP transcripts have been found additionally in lung, heart, gut, spleen, uterus, and brain [45], as well as in activated monocytes [46]. Its role when produced locally is still unclear. DBP is very polymorphic, with over 120 variants being described [47], making it useful as a population marker [48]. The gene encoding the human DBP has been cloned, its genomic structure established [44,45], and homology with both -fetoprotein and albumin recognized [45]. Interestingly, albumin binds 10% of the vitamin D sterols in the circulation. The genes encoding DBP, -fetoprotein, and albumin are all located on chromosome 4 and are considered part of the same gene family based on a conserved triple domain structure [45]. Vitamin D3 synthesized in the skin travels in plasma almost entirely bound to DBP, whereas vitamin D2 obtained in the diet is associated with both lipoproteins (chylomicrons) and DBP [49]. Like other steroid hormones in the circulation, the free or unbound 1,25(OH)2D is in equilibrium with the bound form. It is the free fraction of the 1,25(OH)2D that is hormonally active and binding to DBP inhibits accessibility of the steroid to the cell and prolongs 1,25(OH)2D halflife [50]. In serum, approximately 0.04% of 25(OH)D and 0.4% of 1,25(OH)2D are found in the free form. DBP functions as a reservoir of 25(OH)D and serves as a buffer to prevent the too rapid tissue delivery of the steroids to target cells. DBP thereby prevents vitamin D deficiency and presents 25OHD for renal activation to 1,25(OH)2D [51]. Several findings suggest that DBP may have other critical roles in the body in addition to being the vitamin D transport protein. As alluded to earlier, it circulates at micromolar concentrations, 100-fold in excess of its main ligand 25(OH)D, and is only 5% occupied with calciferols [44,45]. DBP binds monomeric G-actin molecules and is part of the extracellular actin scavenger system that protects the organism from the effects of filamentous F-actin formation when actin is released following cell lysis [52]. Additionally, DBP has been shown to be membrane associated on a number of cell types, either acquired from serum or synthesized by the cell [53]. The function of membraneassociated DBP is unclear at present but it may aid in sterol transport into the cell or it may play a role in modulating the function of 1,25(OH)2D by limiting its interaction with the cell and the VDR [45].
261
CHAPTER 9 Vitamin D
Although there are no reports of patients with DBP deficiency, a DBP knockout mouse has been described [54]. DBP null (-/-) mice are phenotypically normal and fertile. However, they have lower circulating concentrations of 25(OH)D and 1,25(OH)2D when fed a normal diet and exhibit secondary hyperparathyroidism and bone changes when fed a vitamin D-deficient diet. These findings were not seen in control normal mice and support the concept that DBP acts as a storehouse for vitamin D metabolites, thus protecting the animal in times of vitamin D deficiency. DBP markedly prolonged the serum half-life of 25(OH)D and less dramatically prolonged the half-life of vitamin D by slowing its hepatic uptake and increasing the efficiency of its conversion to 25(OH)D in the liver. After an overload of vitamin D, DBP -/- mice were unexpectedly less susceptible to hypercalcemia and its toxic effects. DBP knockout mice also show an increase in hepatic uptake and clearance of vitamin D. Thus, the role of DBP is to maintain stable serum stores of vitamin D metabolites and modulate the rates of its bioavailability, activation, and end-organ responsiveness. These properties may have evolved to stabilize and maintain serum levels of vitamin D in environments with variable vitamin D availability [54].
D. Megalin Megalin is a large multifunctional endocytic clearance receptor for circulating proteins that has been implicated in vitamin D uptake and delivery to the kidney for activation to 1,25(OH)2D [55]. Knockout of the megalin gene in mice usually is lethal, but the few survivors were characterized as having severe rickets [55]. The findings suggested that DBP may be a ligand for megalin and that megalin is critical for 25OHD uptake by the kidney. However, because megalin is located on the luminal membrane of the proximal tubule, this theory suggests that DBP carrying vitamin D is filtered by the glomerulus and reabsorbed by the tubular cells. Thus, knockout mice with the null (-/-) megalin genotype develop proteinuria [56] and lose their vitamin D – DBP complex into the urine, leading to vitamin D deficiency and rickets [55]. In the presence of megalin, the vitamin D – DBP complex would bind to megalin and be internalized by endocytic vesicle formation. Megalin may also affect vitamin D function by binding PTH and PTHrP and directing these proteins toward endocytic lysosomal catabolism. Megalin may also be involved in vitamin D uptake into other cells, facilitating its metabolism and action. Leheste et al. [56] have speculated that the Fanconi syndrome, characterized by proteinuria and osteomalacia, might be related to defects in megalin as represented by the megalin null (-/-) mouse. The role of megalin in the entry of vitamin D metabolites into metabolic or target tissues remains to be clarified.
E. Intracellular Binding Proteins Adams and colleagues [57,58] have described intracellular vitamin D-binding proteins (IDBPs), which they speculate play a role in the intracellular movement of vitamin D metabolites. The IDBPs are related to heat shock 70 (HSP 70) proteins and, as chaperones, contain intracellular organelle-targeting sequences to direct bound molecules to various intracellular destinations. Differences in IDBPs may explain the relative resistance of New World primates to vitamin D action. Gacad and Adams [58] demonstrated that this form of resistance is associated with the overexpression of an IDBP. Their data suggest that this IDBP is relatively specific for 25OHD3 and that additional HSP 70like binding proteins are present in New World primates that are specific for 1-hydroxylated-vitamin D metabolites as well as other steroid hormones. These authors hypothesize that the movement of vitamin D metabolites within cells is controlled by IDBPs, directing substrate toward metabolic pathways for activation or inactivation by enzymes or to the VDR to activate target genes [57,58].
F. Assays of Vitamin D Metabolites Assays of 25(OH)D and 1,25(OH)2D provide valuable tools to assess the vitamin D status of patients [59]. The best indicator of the overall vitamin D status of an individual, 25(OH)D, was originally measured by competitive binding assays, but is now measured by radioimmunoassay. The sensitivity of the assay was improved by the use of 125I-labeled 25(OH)D. Although measurement of 1,25(OH)2D is more difficult because it circulates at approximately 1000-fold lower concentrations than 25(OH)D, i.e., picograms per milliliter instead of nanograms per milliliter, an 125I-based radioimmunoassay is now available for determining 1,25(OH)2D concentrations. In the clinical setting, measurement of 25OHD is the most useful and measurement of 1,25(OH)2D is confirmatory. However, in cases of genetic disease, such as 1-hydroxylase deficiency (see Section VII,A) or hereditary vitamin D-resistant rickets (HVDRR) (see Section VII,B), or in cases of hypercalcemia, measurement of 1,25(OH)2D is critical to understand the pathophysiology.
III. PATHWAYS OF ACTIVATION AND INACTIVATION OF VITAMIN D A. 25-Hydroxylation The first step in activation of vitamin D to the biologically active hormone, 1,25(OH)2D, is hydroxylation at the carbon 25 position in the liver [60]. Although liver
262 parenchymal cells are the primary site for 25-hydroxylation, enzyme activity may be expressed in other tissues in other species (e.g.) in birds, kidney and intestine have 25hydroxylase activity. The 25-hydroxylase enzyme is a cytochrome P450, present in both microsomes and mitochondria in rats, but in humans, significant amounts of the 25-hydroxylase are only found in mitochondria [60, 61]. The gene encoding the human 25-hydroxylase (CYP27) has been cloned [62 – 64] and has been localized to chromosome 2q33-qter [63]. The gene product of CYP27 encodes a protein with sterol 27-hydroxylase as well as 25hydroxylase activities, the former step being important in the biosynthetic pathway of bile acids [61]. The rare genetic disease cerebrotendinous xanthomatosis is due to a deficiency of 27-hydroxylase activity [63]. This defect results in the accumulation of bile acid precursors and cholestanol, which deposit in the brain and peripheral nerves, and forms tuberous xanthomata [65]. These patients also have been reported to have low bone mineral density associated with low 25(OH)D levels and increased fracture risk [66]. A deficiency in the enzymatic activity is not clinically apparent unless severe hepatic failure develops [30]. The 25-hydroxylase is not a tightly controlled enzyme in contrast to the 1-hydroxylase (see Section IIIB2). In patients with hypervitaminosis D, 25(OH)D concentrations increase markedly (as much as 15-fold), whereas those of 1,25(OH)2D are relatively normal [67]. Production of 25(OH)D depends primarily on the concentration of vitamin D; however, higher basal vitamin D and 25(OH)D levels may diminish the production of 25(OH)D in vivo. However, 1,25(OH)2D has been shown to limit the production of 25(OH)D. Treatment with 1,25(OH)2D prevented the increase seen in circulating 25(OH)D after oral vitamin D given to volunteers [68]. This effect may be explained by the increased metabolism of 25(OH)D to 24R,25-dihydroxyvitamin D [24,25(OH)2D] due to the induction of 24hydroxylase by 1,25(OH)2D (see Section IIIC2) and therefore increased the metabolic clearance rate of 25(OH)D [69,70]. Calcium may also have a direct modulatory role on 25-hydroxylase activity [60]. However, in vivo, the role of calcium to modulate 25-hydroxylase activity is likely mediated via changes in PTH, which influences the production of 1,25(OH)2D, which in turn increases the metabolism of 25(OH)D through 24-hydroxylation. Although 25-hydroxylase activity may be decreased in cerebrotendinous xanthomatosis, abnormalities in the enzyme are otherwise very rare [60]. Severe liver disease may lead to metabolic bone disease, but etiologies are complex and include malabsorption, dietary changes, and effects of medication and alcohol, as well as defective 25hydroxylation. This subject is discussed more fully in Chapter 48.
FELDMAN, MALLOY, AND GROSS
B. 25-Hydroxyvitamin D-1-Hydroxylase 1. THE 25-HYDROXYVITAMIN D-1-Hydroxylase Enzyme Following hydroxylation in the liver, 25(OH)D is transported in the circulation bound to DBP and the kidney accomplishes the final step of vitamin D activation, namely 1-hydroxylation. This step is apparently megalin dependent. The 25-hydroxyvitamin D-1-hydroxylase (1-hydroxylase) is a mitochondrial P450 enzyme present in low abundance and localized to the proximal tubule of the nephron [71]. As a mixed function oxidase the enzyme requires NADPH, molecular oxygen, ferredoxin, and ferredoxin reductase for activity. In 1997, cDNAs for the 1-hydroxylase from the mouse, rat, and human were cloned [72 – 77]. The predicted amino acid sequence confirms that the 1-hydroxylase gene (CYP1 or CYP27B1) is a member of the cytochrome P450 enzyme superfamily. The 1-hydroxylase exhibits significant homologies to the vitamin D-25-hydroxylase (CYP27) and the 25-hydroxyvitamin D-24-hydroxylase (CYP24) enzymes. The human 1-hydroxylase gene is approximately 5 kb in length and is composed of nine exons. Fluorescent in situ hybridization analysis localized the gene to chromosome 12q13.3, confirming earlier reports that the gene defect causing 1-hydroxylase deficiency was linked to chromosome 12q14, close to the gene coding for the vitamin D receptor [78,79]. The gene is expressed in kidney epithelial cells in both proximal and distal tubules as well as selected other sites. 2. REGULATION OF 1-Hydroxylase In contradistinction to 25-hydroxylase, renal 1-hydroxylase is a tightly regulated enzyme and the critical determinant of 1,25(OH)2D synthesis (Fig. 2). The overall regulation of 1-hydroxylase is determined by the calcium and phosphorus requirements of the organism and is mediated by several bioactive substances. The principal regulator of 1-hydroxylase is parathyroid hormone (PTH) [40,80]; however, other important regulators include phosphate, 1,25(OH)2D itself, calcium, and calcitonin. The production of 1,25(OH)2D may also be modulated by other hormones, such as estrogen, prolactin, and growth hormone, but these effects in mammalian systems appear to be small. Analysis of the human 1-hydroxylase promoter has identified positive response elements for PTH and calcitonin and a negative response element for 1,25(OH)2D [81,82]. In normal calcemic states, the expression of 1hydroxylase is determined by the concentrations of calcitonin and 1,25(OH)2D [83]. In hypocalcemic states, the expression of 1-hydroxylase is determined by the levels of
CHAPTER 9 Vitamin D
263 inhibitor of PKA abrogated the PTH-mediated upregulation of 1-hydroxylase gene expression [81]. It is now well established that PTH secretion is regulated by the concentrations of both Ca2 and 1,25(OH)2D in the blood [low Ca2 increases PTH; high 1,25(OH)2D decreases PTH]. In the parathyroid gland, the circulating Ca2 concentrations are monitored by the Ca2-sensing receptor (CaR) described by Chattopadhyay and colleagues [88], whereas the VDR downregulates PTH synthesis and secretion [86].
FIGURE 2
Regulation of 1-hydroxylase and 24-hydroxylase activities
in kidney.
PTH and 1,25(OH)2D [81,82]. Positive and negative regulation of the 1-hydroxylase gene by PTH, calcitonin, or 1,25(OH)2D3 has been demonstrated at the transcriptional level in kidneys of animals and in a mouse proximal tubule cell line. Although data strongly support the role of PTH and 1,25(OH)2D in regulating renal 1-hydroxylase, the regulation of 1-hydroxylase expression by these hormones in nonrenal tissues remains to be determined. However, it is clear that the 1-hydroxylase enzyme expressed in renal and nonrenal tissues is encoded by the same gene, as mutations causing 1-hydroxylase deficiency have been found in both renal [77] and nonrenal tissues, including keratinocytes [76] and blood cells [84]. a. PTH Evidence that PTH is the primary regulator of 1-hydroxylase is substantial [85 – 87]. 1,25(OH)2D levels are increased in hyperparathyroidism and reduced in hypoparathyroidism. After parathyroidectomy, 1,25(OH)2D levels fall and are increased after administration of PTH to normal subjects and to patients with hypoparathyroidism. Moreover, substantial in vitro data indicate that PTH markedly stimulates 1-hydroxylase activity in mammalian renal slices, isolated renal tubules, and cultured renal cells. PTH has been shown to stimulate 1-hydroxylase gene promoter activity, most likely by increasing cAMP. cAMP stimulates protein kinase A (PKA) activity, which phosphorylates CRE-binding protein (CREB) and modulates transcription through cAMP-responsive elements (CRE) present in the 5’ sequence flanking the 1-hydroxylase gene. Alternatively, cAMP may modulate promoter activity by transactivation through AP-1 and AP-2 sites. In addition, an
b. Phosphate Phosphate is the second most important physiological regulator of the 1-hydroxylase with high phosphate levels suppressing and low levels stimulating enzyme activity [86,87,89]. Indeed, thyroparathyroidectomized rats maintained on high calcium, low phosphorus diets metabolize 25(OH)D to 1,25(OH)2D despite the absence of PTH, suggesting that PTH may, in part, indirectly influence 1-hydroxylase activity by way of its regulation of phosphate. In humans, phosphorus restriction increases 1,25(OH)2D concentrations to 180% of control, and phosphorus supplementation decreases circulating 1,25(OH)2D by 29% [90]. These changes reflect alterations in the synthetic rate rather than changes in the half-life of the enzyme, demonstrating the important role played by phosphate on 1-hydroxylase. The effect of elevated phosphate to inhibit 1-hydroxylation contributes to the development of renal osteodystrophy during chronic renal failure and is part of the rationale for using phosphate binder therapy to delay the onset of bone disease in these patients [86,87,89]. c. 1,25(OH)2D Interestingly, 1,25(OH)2D regulates its own production. This activity is mediated directly at the level of the 1-hydroxylase in the kidney and indirectly by inhibition of PTH (as described earlier). Low 1,25(OH)2D concentrations promote 1-hydroxylase activity and 1,25(OH)2D synthesis, whereas high 1,25(OH)2D levels inhibit enzyme activity [71,86]. The ability of 1,25(OH)2D to inhibit 1-hydroxylase activity has been demonstrated in vitro as well as in vivo [71]. This effect involves both PTHdependent and PTH-independent mechanisms; 1,25(OH)2D directly (PTH-independent) decreases 1-hydroxylase activity as well as decreasing PTH secretion (PTH-dependent). In vivo, however, it is difficult to separate the contribution of changes in calcium or PTH from direct 1,25(OH)2D actions because of the tight linkage of these systems. In VDR null (-/-) mice, 1-hydroxylase gene expression is increased (a phenomenon used to help in the cloning of this elusive gene [72]) and the upregulation of 1-hydroxylase by PTH was evident. However, a downregulation of 1-hydroxylase gene expression by 1,25(OH)2D3 was not observed, implying that the VDR is essential for the negative regulation of this gene by
264 1,25(OH)2D3, probably via an effect on PTH transcription [72,81]. Another complexity in vivo is the finding that the administration of 1,25(OH)2D chronically can regulate its serum concentration by increasing its metabolic clearance rate by induction of the 24-hydroxylase enzyme (see Section III,C) [91]. d. Calcium Although regulation of 1-hydroxylase in response to changes in serum calcium levels is mainly due to changes in PTH, calcium may act independently as well. Calcium restriction increases 1,25(OH)2D synthesis in thyroparathyroidectomized rats [92]. Additionally, an increase in calcium concentration in the media of cultured chick renal tubule cells and rat renal tissue slices leads to a decreased production of 1,25(OH)2D. The effect of calcium in the regulation of 1-hydroxylase may explain why some patients with severe hyperparathyroidism and very high serum calcium levels exhibit low 1,25(OH)2D values [93]. Although the underlying mechanism for this finding is obscure, one might speculate that the calcium-sensing receptor (CaR) originally described in parathyroid glands [94] and also found in the kidney [95] may mediate this effect [88]. However, studies in VDR null (-/-) mice indicate that calcium is likely an indirect modulator of 1-hydroxylase, as in the absence of 1,25(OH)2D action, changes in calcium did not alter the levels of 1-hydroxylase activity [81]. e. Calcitonin Calcitonin can also stimulate 1,25(OH)2D synthesis in thyroparathyroidectomized rats [96]. Similarly, 1,25(OH)2D levels increase after calcitonin administration to patients with X-linked hypophosphatemic rickets [97] as well in the HYP mouse [98] where the 1-hydroxylase response to PTH is abnormal. In normocalcemic rats where PTH concentrations are relatively low, calcitonin has been shown to be a major regulator of the renal enzyme [83]. Analysis of the human 1-hydroxylase gene promoter has demonstrated a positive regulatory region for calcitonin [81]. f. Chronic Renal Failure In a model of chronic renal failure, in 5/6ths nephectomized rats the renal 1-hydroxylase gene expression decreased and the positive effects of PTH and calcitonin were diminished [81]. This study, and others like it, also showed that PTH and calcitonin positively regulate renal 1-hydroxylase gene expression via protein kinase A (PKA)-dependent and-independent pathways, respectively, and that 1,25(OH)2D3 is a negative regulator. Furthermore, in a moderate state of chronic renal failure, renal cells expressing the 1-hydroxylase gene appear to have a diminished potential to respond to the positive regulators, PTH and calcitonin [86,87,89,99]. g. Local Production of 1,25(OH)2D The kidney is the major source of circulating detectable 1,25(OH)2D. How-
FELDMAN, MALLOY, AND GROSS
ever, humans and animals devoid of functioning renal tissue exhibit low but detectable 1,25(OH)2D concentrations in the circulation [99]. Several extrarenal tissues, including skin [76], bone [100], macrophages [84,101], colon [102], placenta [103], and prostate [104], have been shown to exhibit 1-hydroxylase activity. Macrophages can convert 25OHD to 1,25(OH)2D and, in cases of an increased macrophage pool in the body, 1,25(OH)2D production by these cells can lead to hypercalcemia [99, 105]. The usual regulators of renal 1-hydroxylase, PTH, calcitonin, and 1,25(OH)2D, apparently do not play a primary role in controlling extrarenal 1-hydroxylase activity. For example, when overexpression of macrophage 1-hydroxylase activity and hypercalcemia occurs as in sarcoid, PTH is suppressed [106,107]. Potential regulators of 1-hydroxylase in macrophages include cytokines and the nitric oxide system [99,108].
C. 25-Hydroxyvitamin D-24-Hydroxylation in Kidney and Other Sites 1. THE 25-HYDROXYVITAMIN D-24-HYDROXYLASE ENZYME 25-Hydroxyvitamin D-24-hydroxylase (24-hydroxylase) is a mitochondrial P450 enzyme, which, in general, is expressed in cells that are responsive to 1,25(OH)2D [109]. 24-Hydroxylase can convert 25(OH)D to 24,25(OH)2D, which may have some biological activity (see Section III,C,3). However, the formation of 24,25(OH)2D is generally considered to represent the first step in the degradative and excretory pathway of vitamin D (Figs. 1 and 2). The enzyme can also hydroxylate 1,25(OH)2D to 1,24,25(OH)3D, initiating the inactivation pathway of the active hormone. Thus this enzyme acts to protect the body from the overproduction of 1,25(OH)2D [110]. The intestine is a major site of hormonal inactivation by virtue of its abundant 24-hydroxylase activity [111]. In the nephron, the enzyme is distributed in the proximal and distal tubules, the glomerulus, and the mesangium [112]. Cloning of the gene encoding the rat kidney 24-hydroxylase (CYP24) [113] led to the cloning of the human gene [64,114] and characterization of its promoter region [115]. Earlier there had been some question about whether the 24-hydroxylase and 1-hydroxylase activities resided in the same protein. Data now firmly indicate that these are in fact distinct enzymes encoded by separate genes. The human 24-hydroxylase is present on chromosome 20q13 [114]. 2. REGULATION OF 24-HYDROXYLASE ACTIVITY The regulation of the 24-hydroxylase activity (see Fig. 2) has been reviewed [109]. 1,25(OH)2D is the primary regulator of 24-hydroxylase, causing a marked induction of enzymatic activity and mRNA levels via a VDR-mediated genomic pathway (see Section IV,H). Two vitamin D response
265
CHAPTER 9 Vitamin D
elements (VDREs, see Section IV,F) have been identified in the promoter of the 24-hydroxylase gene [116,117]. Because 24-hydroxylase can be induced by 1,25(OH)2D in many VDR-containing cells, regulation of this gene product has proven to be an excellent marker of 1,25(OH)2D activity. Measurement of 24-hydroxylase enzyme activity and induction of mRNA by 1,25(OH)2D have been employed extensively in studies of cultured dermal fibroblasts from HVDRR patients [7,118] (see Section VII,B) and in studies of kidney, bone, and intestinal cells, as well as in studies of new target organs and malignant cells. In the kidney, PTH stimulates 1-hydroxylase and inhibits 24-hydroxylase [119], effects that are opposite to those of 1,25(OH)2D. However, because the intestine does not respond to PTH, and 24-hydroxylase is not downregulated by PTH in cultured bone cells [120], the action of PTH on renal 24-hydroxylase appears to be most important. Calcitonin has been shown to downregulate 24-hydroxylase mRNA and enzyme activity in rat intestine in vivo [121], suggesting the presence of an intestinal calcitonin receptor and a heretofore unanticipated function for this hormone. 3. CONTROVERSY OVER WHETHER 24,25(OH)2D EXHIBITS DISTINCT BIOLOGIC ACTIVITY Controversy over the biological activity of 24,25(OH)2D has existed for a number of years and the issue still causes lively debate [122]. It is clear that 24,25(OH)2D can bind to the VDR and, at a high concentration, induce 1,25(OH)2Dlike activity [123]. Findings have shown that 24,25(OH)2D can transactivate gene expression via an osteocalcin-VDRE mediated by a VDR mechanism identical to the pathway of 1,25(OH)2D action [124]. Whether 24,25(OH)2D has a unique receptor [125,126] and unique actions independent of 1,25(OH)2D, particularly during development or in selected tissues, has been speculated on for a number of years [123,127]. Evidence supporting a role for 24,25(OH)2D has been presented [126,127] and refuted [123]. In HYP mice, administration of 24,25(OH)2D3 did not merely mimic 1,25(OH)2D3 but caused unique actions, including increasing bone size, dry weight, and bone mineral content, without causing hypercalcemia or activating bone resorption as did 1,25(OH)2D [128]. In growth plate chondrocytes where 24-hydroxylase is expressed, high dose treatment with 24,25(OH)2D has been shown to be involved in the process of regulating bone growth, development, and repair [126]. A 24-hydroxylase knockout mouse model has been generated to address the physiological role of 24,25(OH)2D [110]. 24-hydroxylase null (-/-) mice showed a reduced clearance of 1,25(OH)2D3 from the circulation and have a high degree of perinatal lethality most likely due to hypercalcemia. Histological examination of bone from these mice showed an accumulation of unmineralized osteoid matrix in calvaria, mandible, clavicle, and cortical surfaces of long bones, sites of intramembraneous ossification.
Treatment with 24,25(OH)2D rescued the bone phenotype normalizing the development of the calvaria and the accumulation of osteoid at the periosteal surface of long bones, all suggesting a role for 24,25(OH)2D [110]. However, because 24-hydroxylase initiates 1,25(OH)2D3 inactivation, these mice have high circulating 1,25(OH)2D3. To rule out the contribution of elevated 1,25(OH)2D3 concentrations acting via the VDR in this study, a subsequent study examined a double knockout mouse generated by crossing 24hydroxylase (-/-) mice with VDR (-/-) mice. The animals were fed a high calcium diet to maintain normal calcium concentrations in the serum [129]. Whereas 24-hydroxylase (-/-), VDR (/) mice showed reduced amounts of mineralized tissue in the mandible and cranial bones, 24-hydroxylase (-/-), VDR (-/-) double knockout mice showed normal bone formation at all sites. Data indicate that the impaired mineralization phenotype seen in 24-hydroxylase (-/-) mice was due to the increase in 1,25(OH)2D3 action on the bone because of loss of the 24-hydroxylase inactivation pathway. The authors conclude that 24,25(OH)2D3 is not an essential hormone for bone formation [129]. The slower rate of 1,25(OH)2D3 metabolism in VDR-ablated mice was also seen in other studies [130]. 4. OTHER METABOLITES Less research has centered around the role of 1,24,25(OH)3D as an active hormone; however, in early studies, 1,24,25(OH)3D was shown to stimulate intestinal calcium absorption, mobilize calcium from bone, and exhibit antirachitic activity in rats [131,132]. The steroid binds to the VDR, but with lower affinity than 1,25(OH)2D [123] and, as a result, 1,24,25(OH)3D has diminished potency when compared to 1,25(OH)2D. The initial step in the catabolic pathway of 1,25(OH)2D is 24-hydroxylation, which leads to the inactivation of the active hormone and the production of more polar metabolites, eventually leading to calcitroic acid [11,132,133]. The affinity of the 24-hydroxylase enzyme is 5 to 10 times greater for 1,25(OH)2D than 25(OH)D, making 1,25(OH)2D the preferred substrate. Subsequent hydroxylations and oxidations lead to the production of the C-23 carboxylic acid derivative, calcitroic acid, the major final excretory product of vitamin D metabolism (Fig. 1). The details of these complex catabolic steps have been reviewed [11,133]. Vitamin D can be metabolized to more than 35 metabolites, some of which have been shown to exhibit biological activity [11]. Using the Caco-2 human colon adenocarcinoma cell line, Bischof et al. [134] showed that in proliferating undifferentiated cells, 1,25(OH)2D3 is converted into several side chain metabolites through the C-24 oxidation pathway. In contrast, in differentiated cells where the C-24 oxidation pathway is inactive, a more polar compound produced in significant amounts was found and identified as
266
FELDMAN, MALLOY, AND GROSS
1,25-dihydroxy-3-cholecalcifierol [1,25(OH)2-3-epi-D3]. This study showed that the state of cell differentiation influenced the metabolism of 1,25(OH)2D3 and leads us to speculate about possible autocrine or intracrine effects of vitamin D metabolites to maintain differentiation and limit cell proliferation. The activity of 1,25(OH)2-3-epi-D3 is only slightly, but not significantly, less active than the native 1,25(OH)2D3 in suppressing bovine PTH gene transcription and secretion, despite having a 30-fold lower affinity for the VDR [135 – 137]. Both 1,25(OH)2D3 and 1,25(OH)2-3-epi-D3 suppress PTH secretion by 50%. In metabolism studies using bovine parathyroid cells, the concentration of 1,25(OH)2-3-epi-D3 was even higher than that of the parent substrate 1,25(OH)2D3, suggesting that this diastereomer has a slower rate of metabolism. Thus, production and accumulation of 1,25(OH)2-3-epi-D3 as a major stable metabolite of 1,25(OH)2D3 in parathyroid and other tissues may contribute to the prolonged effects of 1,25(OH)2D3 on gene transcription.
IV. MECHANISM OF 1,25(OH)2D ACTION 1,25(OH)2D regulates calcium metabolism and promotes other physiologic actions by a VDR-mediated mechanism analogous to the classical steroid hormones. The VDR, a member of the steroid – thyroid – retinoid receptor gene superfamily, acts as a regulator of target gene transcription. Several reviews of the 1,25(OH)2D-VDR system have been published [1,2,4,5,7,9,138,139], and the subject is covered extensively elsewhere [10].
A. The Vitamin D Receptor The identification of 1,25(OH)2D as the active metabolite of vitamin D led to a search for the specific protein that functioned as a receptor for this hormone. Early studies demonstrated the existence of a target tissue protein that interacted with vitamin D metabolites [140]. The protein was subsequently very well characterized in a number of laboratories and was shown to have an approximate molecular mass of 50 kDa, to exhibit saturable and specific high-affinity binding of 1,25(OH)2D, to be located in the nucleus (see later), and to bind to DNA. The binding affinity for the vitamin D metabolites roughly correlated with their potency [141], although many exceptions have been noted. Analysis of the structure of the VDR by limited proteolysis showed that the DNA-binding region was distinct from the 1,25(OH)2D-binding portion of the protein, thus dividing the protein into two functional domains [142]. 1,25(OH)2Dbinding and DNA-binding activities were shown to be sen-
sitive to sulfhydryl and metal-chelating reagents, indicating that the protein contained essential cysteine residues and required heavy metals for these activities [143]. The VDR was eventually purified to near homogeneity from chick intestinal tissue [144,145], and monoclonal antibodies to the chick VDR [146] and later to the porcine VDR [147] were developed. From biochemical and immunocytochemical studies, the VDR has been found in the nuclei of many normal human tissues, including intestine, kidney, bone, parathyroids, thyroid, skin, adrenal, liver, breast, pancreas, muscle, prostate, and lymphocytes, demonstrating that the protein is widely distributed in normal human tissues [5,148 – 152]. It has also been identified in malignant cells from most of these sites. VDR is a phosphoprotein that undergoes a hormone-dependent phosphorylation in intact cells [153]. In embryonic chick duodenal organ culture, phosphorylation of the receptor is strongly induced within 1 h by 1,25(OH)2D and this occurs prior to the initiation of calcium uptake or the induction of calcium-binding protein [154]. The VDR has been shown to be phosphorylated on serine residues in vitro by protein kinase C [155], casein kinase II [156,157], and cAMP-dependent protein kinase [158], which strongly suggests that phosphorylation of the VDR may be an essential event required for 1,25(OH)2D-induced gene activation [155 – 158]. Phosphorylation of VDR by protein kinase A (PKA) results in a reduction in the transactivation capacity of the receptor in response to 1,25(OH)2D3 [158]. In vivo, Ser208 has been shown to be the major site of phosphorylation by casein kinase II during 1,25(OH)2D3 treatment [156, 157]. In vivo, overexpression of casein kinase II causes an increase in 1,25(OH)2D-induced transcriptional activity [159]. Phosphorylation of Ser51 by protein kinase C diminishes VDR binding and nuclear localization of the VDR [155,160] so that differential phosphorylation may play a role in determining VDR activity.
B. The Gene Encoding the VDR In 1986, the chick VDR cDNA was cloned by McDonnell et al. [161], which subsequently led to the cloning of the human VDR cDNA by Baker et al. [162]. The original VDR cDNA described by Baker et al. [162] contains 4800 nucleotides and encodes a protein of 427 amino acids with a predicted molecular mass of 48,000 Da. The cDNA sequence encodes a protein containing a cysteinerich region corresponding to the DNA-binding domain (DBD) that is highly conserved among othe steroid receptors, thus confirming that the VDR is a member of the steroid hormone receptor superfamily [163]. Structure – function analyses of the VDR showed that the DBD is localized to the amino-terminal region of the polypeptide [164]. This region contains nine cysteine residues, eight of
267
CHAPTER 9 Vitamin D
FIGURE 3 Domains of the VDR. The DNA-binding domain consists of two zinc finger modules located at the amino-terminal portion of the receptor. The ligand-binding domain contains 12 -helices shown as open boxes and 3 turns shown as a filled box. E1 and AF-2 subregions of the receptor are important in transactivation. which coordinate zinc binding to form a two zinc finger structure. The ligand-binding domain (LBD) is contained in the region C-terminal to the DBD (Fig. 3). In humans, the VDR gene has been localized to chromosome 12q13-14 [79,165,166], in close proximity to the 1hydroxylase gene [78]. The structural organization of the VDR gene has been delineated by two groups [167, 168]. Miyamoto et al. [167] initially characterized the VDR gene and its promoter and showed that the gene is composed of at least 11 exons that span 75 kb of DNA (Fig. 4). The VDR protein is encoded by 8 exons. Exon 2 and 3 encode the DBD and exons 4 – 9 encode the LBD. Exon 2 contains the most proximal 3 bp of the 5-noncoding sequence, the translation initiation site, and nucleotide sequence that encodes the first zinc finger module. Exon 3 lies approximately 15 kb downstream and encodes the second zinc finger module. The 12 -helices of the ligand-binding domain are encoded by exons 4 and 6 – 9. Exons 4, 5, and 6 encode the D region or “hinge.” Exon 5 encodes a less well-conserved region of the receptor and may represent an insertion. Exons 6, 7, 8, and 9 encode a portion of the hinge and the carboxy-terminal “E” region, as well as approximately 3200 nucleotides of the 3-noncoding sequence [167]. The original human VDR cDNA sequence, which contained two potential methionine start sites, was proposed
FIGURE 4
to encode a 427 amino acid polypeptide [162]. However, subsequent DNA sequence analyses of the VDR gene from normal individuals or patients with HVDRR (see Section VII,B) identified a polymorphism in the ATG codon encoding the first methionine [169]. By not initiating at the first ATG (M1), but at an ATG three codons downstream (M4), the polymorphism lead to a VDR shortened by three amino acids (424 amino acids) (see Fig. 11). Individuals may be homozygotic or heterozygotic at the site, leading to the detection of two closely sized VDR bands on Western blots [170,171]. The numbering system employed in this chapter uses the first methionine (M1) as the starting point, as has been the convention in most papers published to date.
C. Alternate Splicing and VDR Promoters The original VDR cDNA sequence reported by Baker et al. [162] contained approximately 115 bp of the 5-noncoding sequence. Miyamoto et al. [167] showed that portions of this noncoding sequence were contained on separate exons. They subsequently identified three noncoding exons upstream of exon 2. These exons, 1A, 1B, and 1C, were shown to be spliced differentially to generate
Organization of the VDR chromosomal gene. The human VDR gene is located on chromosome 12q13-14 and spans approximately 75 k of DNA. The gene is composed of at least six 5 noncoding exons and 8 coding exons. Alternative splicing results in at least 14 types of transcripts. Locations of the start (ATG) and termination (TGA) codons are indicated.
268
FELDMAN, MALLOY, AND GROSS
three or four mRNA species each with the potential to encode a 424 or 427 amino acid protein, depending on the polymorphism site at M1. Crofts et al. [168] expanded these findings further using RT-PCR. They showed that the 5-noncoding region of the VDR gene was distributed among at least six exons (1A – 1F), which could be spliced differentially to generate 14 mRNA transcripts [168]. Most intriguingly, these authors showed that two of the transcripts have the potential to encode N-terminal variants 23 to 50 amino acids longer than the classical 427 amino acid VDR. These two alternatively spliced transcripts encode VDRs of 450 and 477 amino acids. The polymorphic N terminus of the VDR (M1 vs M4) has been shown to influence transactivation [170], possibly by modulating TFIIB interactions [172]. Whether the presence of additional amino acids upstream of the polymorphic site modulates activity of the VDR is currently under investigation. Miyamoto et al. [167] identified a putative promoter sequence upstream of exon 1A. This GC-rich sequence contains bindings sites for transcription factor SP-1, but did not contain a TATA box. Interestingly, the intron sequence between exon 1C and exon 2 was capable of responding to retinoic acid. Crofts et al. [168] subsequently showed that the expression of the VDR gene was directed by multiple promoters upstream of exons 1A, 1D, and 1F. The two alternatively spliced transcripts encoding VDRs of 450 and 477 amino acids originated from a promoter upstream of exon 1D. Especially intriguing was the finding that one subset of transcripts originating from exon 1F, the most distal exon, was expressed only in the parathyroids, kidney, and intestine, tissues involved in calcium regulation. This finding raises the possibility of differential tissue expression of variant forms of the VDR [168].
D. Three-Dimensional Structure of the VDR LBD Models of the three-dimensional structure of the LBD of the VDR have been developed from computer-modeling studies [173,174]. More recently, the crystal structure of the VDR LBD bound to 1,25(OH)2D3 was solved and reported by Rochel et al. [175]. To obtain crystals of the LBD the authors deleted amino acids 165 – 215. Although this region of the receptor is referred to as an “insertion,” it contains the site of serine phosphorylation, Ser208. However, this region is poorly conserved among nuclear receptors and has not been shown to have a direct effect on VDR transactivation [175]. The deleted VDR LBD, amino acids 118 – 425 with the insertion deleted ( 165 – 215), exhibited 1,25(OH)2D3 and analogue binding affinities similar to the wild-type VDR and was functionally active in transactivation assays when fused to the Gal-4 DBD.
Three-dimensional structure of the holo-VDR LBD. helices are shown as cylinders and three sheets located between helix 5 and helix 6 as arrows. Helix 12 is shown in black and the ligand 1,25(OH)2D3 is docked. The location of the insertion domain deleted from the LBD is shown. Reproduced with permission from D. Moras [Mol. Cell 5, 173 – 179 (2000)]. (See also color plate.)
FIGURE 5
As shown in Fig. 5 (see also color plate), the VDR LBD based on the crystal structure is composed of 12 -helices and 3 sheets [175]. The ligand-binding pocket forms a large cavity of 693 Å and is lined with hydrophobic amino acid residues. When bound to the VDR, the A ring of 1,25(OH)2D3 embraces helix H3 and orients toward the C terminus of helix H5. The 1-hydroxyl group forms hydrogen bonds with Ser237 (H3) and Arg274 (H5), and the 3-hydroxyl group forms bonds with Ser278 (H5) and Tyr143. The conjugated triene connecting the A and C rings fits into a hydrophobic channel formed between Ser275 (loop H5- ) and Trp268 ( 1) on one side and Leu233 (H3) on the other side. The C ring contacts Trp286, and the C18 methyl group is aimed at Val234 in helix H3. The 25-hydroxyl group forms hydrogen bonds with His305 (loop H6H7) and His397 (H11). The AF-2 (activation function-2) domain is contained within helix H12. From crystallographic studies of other receptors [176,177], the H12 helix is repositioned following ligand binding such that repositioning locks the ligand in the cavity of the ligandbinding pocket. The repositioning of H12 also leads to the formation of protein surfaces that allow interaction with specific comodulators and transcriptional activation.
269
CHAPTER 9 Vitamin D
Upon ligand binding, the position of helix H12 is stabilized by hydrophobic interactions involving Thr415, Leu417, Val418, Leu419, Val421, and Phe422 from helix H12 with residues Asp232, Val234, Ser235, Ile238, and Gln239 from helix H3, Ala267 and Ile268 from helix H5, and His397 and Try401 from helix H11. In addition, a salt bridge is formed between Lys264 (H4) and Gln420 (H12), and hydrogen bonding occurs between Ser235 (H3) and Thr415 (H12). Val234, Ile268, His397, and Tyr401 also interact with the ligand, indicating that the repositioning of helix H12 is controlled by 1,25(OH)2D3 [175]. Mutations have been created in several amino acids predicted to be important in ligand binding. Mutagenesis of His397 to alanine completely abolished ligand binding, whereas mutations Ser237Ala and Cys288Ala significantly reduced the binding affinity of the receptor [174]. In addition, two natural mutations found in patients with HVDRR — Arg274Leu, which also abolished ligand binding [178], and His305Gln, which reduced the binding affinity about 10-fold [179]—also showed the importance of these amino acid residues in ligand binding. Ligand-binding modeling has also been extended to docking vitamin D analogues. In some cases, such as MC903 and EB1089, which have side chain modifications, only minor adjustments to the C and D rings are required for docking. Howevers in 20-epi analogues like KH1080 and 20-epi 1,25(OH)2D3, only the low energy conformers could be docked [175]. The large volume of the binding pocket accommodates structural differences in ligand but does not as yet explain the differential activity of various vitamin D analogues (see Section VIII).
E. Heterodimerization Transcriptional activation of target genes by 1,25(OH)2D is complex and involves a sequence of events centered around the VDR. The VDR acts as a trans-acting factor that interacts with specific VDREs located in the promoter regions of 1,25(OH)2D responsive genes. The first target gene shown to have a VDRE was the osteocalcin gene [180 – 182]. In early studies examining the binding of the VDR to the osteocalcin VDRE in yeast, it was demonstrated that a protein from a nuclear extract from mammalian cells was required for the binding to occur. This factor was originally termed a nuclear accessory factor (NAF) [183 – 185] or receptor auxiliary factor (RAF) [186]. NAF/RAF was later determined to be the retinoid X receptor (RXR), a member of the steroid – thyroid – retinoid gene superfamily [187]. RXR is a 55-kDa protein that binds 9cis-retinoic acid as its ligand [188,189] and is found widely distributed in cells and tissues, including those that do not express the VDR [185]. RXR has now been shown to be the heterodimerization partner of a number of other recep-
tors in the steroid – thyroid – retinoid gene superfamily, including thyroid receptor, retinoic acid receptor, and peroxisome proliferator activating receptor [190]. Mutagenesis of the VDR has demonstrated that the E1 region (overlapping H3 – H4) and helix H10 are required for high-affinity binding to RXR. Other regions of the receptor probably contribute to the RXR interface [2,175]. Phosphorylation of RXR by mitogen-activated protein kinase (MAPK) has been shown to inhibit 1,25(OH)2D signaling [191].
F. VDREs and Target Genes 1,25(OH)2D induces a wide array of biological responses, some resulting in an upregulation of specific mRNAs and others that downregulate protein expression. Stimulatory or inhibitory actions may depend on tissue specificity or on the state of cellular differentiation. A number of proteins have been shown to be stimulated by 1,25(OH)2D, including osteocalcin, Eta-1 (for early T lymphocyte activation-1 also known as osteopontin), fibronectin, alkaline phosphatase, carbonic anhydrase, 24-hydroxylase, prolactin, c-fos, PSA, calbindin-D9K, calbindin-D28K, p21, p27, 3 integrin, sodium-dependent phosphate cotransporter 2 (NPT2), nerve growth factor (NGF), transforming growth factor 2 (TGF 2), estrogen receptor (ER), lipoprotein lipase, and aromatase (CYP19). Proteins downregulated by 1,25(OH)2D include collagen, PTH, PTHrP, calcitonin, IL-2, atrial naturetic peptide, Bcl-2, 1hydroxylase, and c-myc. A number of these genes have been shown to have VDREs in their promoters. The human osteocalcin VDRE (GGGTGA ACG GGGGCA) is an imperfect hexanucleotide direct repeat separated by a three nucleotide spacer sequence, a so-called DR3 structure [180 – 182]. Unlike the osteocalcin VDRE, the mouse osteopontin VDRE is a perfect direct repeat of the hexanucleotide GGTTCA separated by a three nucleotide spacer. It is clear that additional aspects of the VDRE region determine whether the 1,25(OH)2D action on a target gene is stimulatory or inhibitory. The complexity in understanding how 1,25(OH)2D responsive genes are regulated in vivo is highlighted by the different VDRE structure that exist as well as the multiple possibilities for comodulator interactions.
G. Comodulator Interactions A large number of proteins, known as comodulators or coregulators, have been found to interact with the nuclear receptors and are required for gene transcription [2,8,9,139]. The particular coregulatory proteins recruited to nuclear receptors, influenced by conformational changes in the receptor – ligand complex, may contribute to the
270
FELDMAN, MALLOY, AND GROSS
specificity of transcriptional regulation [192]. Specific interactions with the VDR have been shown for a number of proteins. In addition to heterodimerizing with RXR, the VDR has been shown to bind SRC-1, a member of the p160 class of comodulators. SRC-1 has been shown to have intrinsic histone acetyltransferase (HAT) activity [193]. Interaction of the VDR with SRC-1 is ligand dependent and involves the AF-2 domain (helix H12) and helix H3 [194,195]. Mutagenesis of Tyr236 to alanine in H3 blocks SRC-1 interaction and transactivation but does not interfere with RXR or ligand binding [195]. The VDR has also been shown to interact with basal transcription factor TFIIB, a protein associated with the basal transcriptional machinery [172,196 – 199]. TFIIB interaction does not involve the AF-2 domain [198], instead it has been shown to bind to the N-terminal portion of the receptor [172, 199]. Using the VDR as bait in the yeast two-hybrid system, Baudino et al. [200] isolated a protein they termed NCoA62 (nuclear receptor coactivator; 62,000 Da). Addition of NCoA-62 to transactivation assays augmented 1,25(OH)2D3 responsiveness; however, the role of this protein in transactivation has not been defined as yet. A large class of proteins collectively called vitamin D – receptor-interacting proteins or DRIPs was isolated by affinity chromatography using glutathione S-transferase (GST)-VDR LBD immobilized to glutathione agarose [201,202]. The DRIP complex only bound to the immobilized VDR when 1,25(OH)2D3 was present. In cell-free transcription assays, DRIPs mediate ligand-dependent gene transcription by the VDR. At least 13 proteins consititute the DRIP complex, although only DRIP205 binds directly to the VDR. The other DRIPs must be recruited to the growing complex of proteins subsequent to DRIP205 binding. Smad3, a member of the Smad protein family of intracellular signal transducers of the TGF- -BMP superfamily, has also been shown to act as a coactivator of the VDR [203, 204]. Smad3 binds to the VDR LBD and enhances the transactivation function of the receptor. The Smad3– VDR complex is stabilized by SRC-1 [204].
H. Transactivation of Target Genes In the bloodstream, 1,25(OH)2D circulates mostly bound to DBP in equilibrium with a small amount of unbound or free hormone. It was generally believed that the unoccupied VDR was localized to the nucleus and that no special mechanism was required for the 1,25(OH)2D to enter the cell and make its way to the nucleus. However, 1,25(OH)2D may bind to a membrane receptor or interact with tubulin during its delivery to the VDR and translocation to the nucleus [205,206]. The role of 1,25(OH)2D in the transport of VDR from cytoplasm to nucleus has been examined
using green fluorescent protein (GFP)-tagged chimeras of full-length or truncated constructs of the VDR [207]. 1,25(OH)2D treatment showed translocation of cytoplasmic VDR to the nucleus and colocalization with RXR [208]. Truncation of either the LBD or the AF-2 region of VDR abolished hormone-dependent translocation and transactivation. The findings support the model of hormone-dependent VDR translocation and indicate that translocation from the cytoplasm to the nucleus is part of the receptor activation process. An intact AF-2 domain is required for this translocation [207]. Photobleaching data suggest that the VDR shuttles back and forth between the cytoplasm and the nucleus and that ligand increases the nuclear accumulation of VDR [208]. A simplified model of 1,25(OH)2D-regulated gene transactivation is shown in Fig. 6 (see also color plate). In the absence of 1,25(OH)2D, the VDR exhibits a low affinity for RXR or is in the cytoplasm. Upon 1,25(OH)2D binding, the VDR translocates to the nucleus and RXR proteins heterodimerize and form a high-affinity complex that acquires the ability to recognize and bind with high affinity to VDREs through their cognate DBDs. As determined from the crystal structure of the VDR LBD and models based on other steroid receptors, the VDR undergoes a major conformational change upon 1,25(OH)2D binding. The repositioning of helix H12 upon the upstream -helices in the E1 domain forms high-affinity protein surfaces capable of interacting with specific comodulator proteins required for transactivation. The VDR – RXR heterodimer attracts coactivator proteins such as SRC-1 and NCoA-62. SRC-1, through its intrinsic HAT activity, derepresses chromatin so that nucleosomes are rearranged and naked DNA is accessible, allowing for the assembly of the transcriptional apparatus. TATA-binding protein-associated factors (TAFs) are also recruited to target TATA/TBP-binding sites. Other proteins, including DRIPs and TFIIB, serve to stabilize the preiniation complex. Transcription is then initiated by RNA polymerase II. For stimulatory actions, mRNAs transcription is increased and the specifically induced mRNAs are translated into proteins that carry out the biological effects of the hormone. For inhibitory actions, mRNA transcription is suppressed, thereby reducing the expression level of selected proteins.
I. Regulation of VDR Abundance Within each target tissue, the level of VDR is not fixed but rather is regulated dynamically by a variety of physiological and developmental signals. The concentration of VDR expressed in a target cell determines the amplitude of the response evoked by 1,25(OH)2D treatment. Upregulation of VDR enhances the response to 1,25(OH)2D, and downregulation diminishes the response [209 – 213]. Of
271
CHAPTER 9 Vitamin D
FIGURE 6
Model of 1,25(OH)2D gene transactivation. Upon entering the cell, 1,25(OH)2D3 binds to the VDR, leading to the formation of a VDR:RXR heterodimer (1). The heterodimeric complex subsequently binds to vitamin D response elements (VDREs) in promoter regions of target genes through their cognate DNA-binding domains (2). Conformational changes in the VDR:RXR heterodimer initiate the recruitment of coactivating proteins, including SRC-1 and NCoA-62, to the oligomeric complex. The histone deacetylase activity of SRC-1 modifies the chromatin structure and facilitates essential contact with the general transcription apparatus (3). Additional proteins are recruited to the complex such as TBP and TAFs for targeting promoter elements (4). Binding of TFIIB and DRIPs to the complex stabilizes the preinitiation complex (5). Once the proteins have been assembled, transcription is initiated by RNA polymerase II (6). Reproduced with permission from M. R. Haussler [J. Cell. Biochem. Suppl. 32/33, 110 – 122 (1999)]. (See also color plate.)
the many factors that regulate VDR abundance, the hormone 1,25(OH)2D itself is an important modulator that increases the concentration of the receptor (homologous upregulation). Other regulators include steroid and peptide hormones, growth factors, activators of specific second messenger pathways, and intracellular calcium, which may up- or downregulate the concentration of VDR (heterologous regulation) [213]. VDR abundance has been shown to depend on the proliferation/differentiation status of the target cells, and VDR changes are also detectable during neonatal development in different tissues [213]. 1. HOMOLOGOUS REGULATION The VDR is upregulated by 1,25(OH)2D and other vitamin D metabolites that bind to the VDR itself (homologous regulation), and this has been observed both in vitro [214] and in vivo [215 – 217]. The magnitude of homologous upregulation varies from 2 to 10-fold depending on the target cell. In pig kidney cells, human skin fibroblasts, and human mammary cancer cells (MCF-7), the VDR content increases when the cells are treated with 1,25(OH)2D3, 1,24,25(OH)3D3, 24,25(OH)2D3, and 25(OH)D3, and the concentrations required for maximal upregulation closely reflect the affinities of the various metabolites for the VDR [214]. Several studies have shown that the upregulation of
the VDR is due to an increase in the transcription of the VDR gene [161,217,218]; however, other studies have found that the upregulation is mainly due to the stabilization of the ligand-occupied VDR [219 – 221]. Either one or both of these phenomena may be operative depending on the target cells under study [214,220]. When examined carefully in pig kidney cells, about two-thirds of the upregulation appeared to be due to the stabilization of the VDR and one-third due to the increased synthesis of the VDR protein [214]. Homologous upregulation of VDR appears to play a role in the treatment of psoriasis, a hyperproliferative skin disorder. Chen et al. [222] have shown that the response to 1,25(OH)2D treatment in patients with psoriasis correlated with the upregulation of VDR in psoriatic skin. In patients who showed clinical improvement with treatment, significant upregulation of VDR mRNA was observed in psoriatic lesions, whereas there was no upregulation in patients who did not respond to 1,25(OH)2D. 2. HETEROLOGOUS REGULATION Various hormones, including steroid and peptide hormones, and growth factors regulate VDR abundance (heterologous regulation) in a cell and tissue-specific manner. In cultured cells, the abundance of VDR has been shown to be closely related to the rate of cell proliferation, with VDR
272 levels being higher in proliferating cells than in quiescent cells [223,224]. Also, in some cell systems, the induction of differentiation leads to decreased VDR levels [218,225, 226]. However, activation of the protein kinase C pathway by mitogens such as basic fibroblast growth factor and phorbol esters and/or elevation of intracellular Ca2 led to a significant decrease in VDR abundance, despite stimulating cell proliferation [227]. The downregulation appears to result from a destabilization of the VDR mRNA or a decrease in VDR gene expression. In contrast, elevation of intracellular cAMP has been shown to increase VDR abundance by increasing VDR mRNA expression [211,228]. The mechanism of the cAMP-mediated upregulation may be due to an increase in the transcriptional rate of the VDR gene, as shown by a study characterizing a 1.5-kb promoter region of the mouse VDR gene [229]. Miyamoto et al. [167] cloned and characterized approximately 1.5 kb of the promoter region of the human VDR gene 5 of the transcriptional start site and demonstrated the presence of several potential binding sites for the SP-1 transcription factor and putative sites for other transcriptional activators, including cAMP response elements. Prostaglandins have also been shown to regulate VDR levels in human leukemia cells, probably through the elevation of cAMP levels in these cells [230]. While studying the regulation of the VDR gene in the intestine, Yamamoto et al. [231] demonstrated the presence of an intestine-specific cis element in the human VDR gene promoter (at 3731 to 3720), which interacts with a caudal-related homeodomain transcription factor, Cdx-2. Cdx-2 is able to activate VDR gene transcription in the intestine by binding to this element. Glucocorticoids [232 – 235], estrogens [236,237], retinoic acid [238,239], and PTH [212,228,240] have been shown to regulate VDR abundance. Changes in VDR abundance elicited by these hormones are reflected in the magnitude of 1,25(OH)2D responsiveness. However, species differences among various rodent models prevent extrapolation to humans. Even within a species, there may be tissue-specific differences. An intron fragment 3 of exon 1C of the human VDR gene appears to confer retinoic acid sensitivity, suggesting a molecular mechanism for the regulation of VDR by retinoic acid [167]. In the case of excess glucocorticoids, there is a resistance to 1,25(OH)2D, whereas in the case of excess PTH there is enhanced 1,25(OH)2D responsiveness [212,228]. Thus these hormones modulate target cell sensitivity to 1,25(OH)2D in part through the regulation of VDR levels. The abundance of VDR in parathyroid glands has been studied extensively [86]. It has been postulated that reduced VDR concentration in parathyroid glands may be related to a lack of vitamin D suppression of the gland [241]. This may contribute to the pathogenesis of secondary hyperparathyroidism in chronic renal failure by reducing the inhibition by 1,25(OH)2D of parathyroid hormone secretion
FELDMAN, MALLOY, AND GROSS
[6,242]. The low serum levels of 1,25(OH)2D in chronic renal failure may accentuate this effect [242]. Similarly, vitamin D status may alter the pattern of signs and symptoms in primary hyperparathyroidism [243 – 245].
V. NONGENOMIC EFFECTS OF VITAMIN D A. Rapid Responses to 1,25(OH)2D In addition to the classical VDR-mediated genomic pathway, 1,25(OH)2D has also been shown to elicit rapid responses [126,246,247]. The term “rapid response” is used to describe the biological effects of 1,25(OH)2D that occur within a few minutes after hormone treatment and are considered too rapid to be explained by a VDR-mediated genomic pathway. Rather, they are thought to be mediated by a direct action of 1,25(OH)2D on the plasma membrane of target cells stimulating a signal transduction pathway involving the rapid opening of voltage-sensitive Ca2 channels and activation of protein kinases. Some of the 1,25(OH)2D-induced rapid responses include changes in intracellular calcium flux, alteration in phospholipid metabolism and phosphate transport, and changes in alkaline phosphatase and adenylate cyclase activities. Also, “transcaltachia,” a process of transluminal transport of Ca2 across the intestine, has been shown to occur rapidly when vitamin D-repleted animals are treated with 1,25(OH)2D3. The rapid Ca2 transport has been proposed to be facilitated by endocytic and lysosomal vesicles, which deliver the Ca2 to the basal – lateral membrane where it is released by exocytosis into the lamina propria. However, because the transcaltachia response requires vitamin D-replete animals, a preexisting condition induced by 1,25(OH)2D may be operative, and thus transcaltachia may ultimately depend on a 1,25(OH)2D – VDR-mediated genomic pathway. One rapid effect observed in 1,25(OH)2D3-treated cultured human fibroblasts was the rapid accumulation of cGMP, which colocalizes with reorganizing VDRs within 15 s after calcitriol addition [248]. However, this rapid effect was dependent on having a functional VDR, as no increase in cGMP was seen in fibroblasts from patients with HVDRR containing defective receptors. In contrast, in rat osteosarcoma cells (ROS 24/1) that are totally devoid of detectable VDRs, 1,25(OH)2D3 induced the rapid flux of Ca2 into the cells, suggesting that a 1,25(OH)2D3 signaling system independent of the VDR exits in at least in some cells [249]. Several lines of evidence support the existence of a nongenomic 1,25(OH)2D-mediated signal transduction pathway. For instance, the antagonist 1,25(OH)2D3, which has a minimal effect on 1,25(OH)2D-induced genomic actions, blocks the effect of 1,25(OH)2D3 on transcaltachia
273
CHAPTER 9 Vitamin D
[250]. Similarly, some vitamin D analogues, such as the 6s-cis-blocked conformer that binds poorly to the VDR, are able to generate the transcaltachia response in perfused chick intestine and Ca2 influx in ROS 17/2.8 cells [247, 251]. In NB4 cells, an acute promyelocytic leukemia cell line, the 6-s-cis-blocked conformer was 20 times more effective at priming the cells for monocytic differentiation than the natural hormone. This response was attenuated by the 1,25(OH)2D3, a specific antagonist of the nongenomic response [252]. The 6-s-cis analogue, 125(OH)2lumisterol3, also induces transcaltachia and stimulates Ca2 uptake in the ROS 17/2.8 osteosarcoma cell line [251]. 125(OH)2lumisterol3 was also shown to augment glucoseinduced insulin secretion in rat pancreatic islet cells while also increasing intracellular Ca2 concentrations [253].
B. Membrane VDR One possible mechanism for these rapid, nongenomic effects is that there is a specific membrane receptor for 1,25(OH)2D that mediates the rapid responses. The presence of a specific binding protein for 1,25(OH)2D3 has been identified in plasma membrane preparations, and a protein that binds [3H]1,25(OH)2D3 was partially purified from a basal – lateral membrane preparation from chick intestine [254]. The putative membrane VDR exhibits a different ligand specificity than the classical nuclear VDR. Using antiserum to this protein, Nemere et al. [254] demonstrated that this protein was present in rat costochondral cartilage cells and that the antibody could block the 1,25(OH)2D3 increase in protein kinase C activity. Immunohistochemistry demonstrated that both resting and growth zone chondrocytes express the protein, but levels are highest in the growth zone. The binding protein is present in both plasma membranes and matrix vesicles and has a molecular mass of 66,000 Da. The putative membrane VDR was also shown to mediate the antiproliferative effects of 1,25(OH)2D3 on chondrocytes [126]. However, at the time of writing, the membrane receptor has not as yet been cloned or characterized at the molecular level.
bined excretion in feces (720 mg) and urine (280 mg) [257]. The coordinated interaction of 1,25(OH)2D and PTH to regulate 1-hydroxylase activity plays a major role in the maintenance of calcium balance (Fig. 7). Small decreases in serum calcium result in increases in PTH secretion, which stimulates the upregulation of 1-hydroxylase activity, and increased renal phosphate excretion. The combination of increased PTH and decreased phosphate leads to enhanced 1-hydroxylase activity. The augmented synthesis of 1,25(OH)2D enhances intestinal calcium absorption to restore the calcium concentration toward normal levels, which in turn feeds back to diminish PTH secretion, thereby limiting the further production of 1,25(OH)2D. In addition, 1,25(OH)2D feeds back on the kidney to inhibit the further production of 1,25(OH)2D by downregulating 1-hydroxylase gene expression while stimulating 24hydroxylase gene expression. Furthermore, serum calcium is maintained by the combined actions of PTH and 1,25(OH)2D on the bone to increase bone resorption and by the action of PTH on the kidney to increase calcium reabsorption. In hypercalcemic states, PTH is suppressed by a signal transmitted via the parathyroid CaR [94], and the entire process is reversed. In rat parathyroid glands and kidney, the expression of the CaR gene is increased by 1,25(OH)2D but not by Ca2 [258]. Upregulation of the CaR is thought to be involved in the suppressive effects of vitamin D compounds on PTH secretion. The selective action of noncalcemic vitamin D analogues that have a greater suppressive effect on PTH expression may allow for their potential use in therapeutic situations with elevated PTH concentrations [259] (see Section VIII).
VI. PHYSIOLOGY: REGULATION OF SERUM CALCIUM A. Interaction of PTH and Vitamin D to Regulate Serum Calcium The concentration of Ca2 in plasma and extracellular fluid is maintained within a narrow range, with variations up or down being associated with untoward effects [6,87,255,256]. In the balanced state, the dietary intake of approximately 1000 mg of calcium is equal to the com-
FIGURE 7 PTH.
Regulation of Ca2 levels in the blood by 1,25(OH)2D and
274
FELDMAN, MALLOY, AND GROSS
B. Extrarenal 1,25(OH)2D Synthesis and Hypercalcemic States Under normal physiological conditions the kidney is the primary site of 1,25(OH)2D formation. However, small amounts of 1,25(OH)2D are produced in various other tissues, and in selected pathological conditions the extrarenal production of 1,25(OH)2D may significantly contribute to alterations in calcium homeostasis [260]. Tissues shown to synthesize 1,25(OH)2D from 25(OH)D include human decidua and placenta, bone cells, keratinocytes, spleen, melanoma cells, hepatoma cells, and peritoneal, synovial, and pulmonary monocytes and macrophages [261]. Hypercalcemia can be expected to occur in 7 to 24% of patients with sarcoidosis [262]. Proof of the clinical significance of extrarenal production of 1,25(OH)2D was provided from studies on an anephric patient with sarcoidosis who developed hypercalcemia [263]. Cultured pulmonary alveolar macrophages from patients with sarcoidosis [101,107] and homogenized sarcoid lymph node tissue [264] have been shown to be capable of producing 1,25(OH)2D. In addition to sarcoidosis, other granulomatous disorders have been associated with hypercalcemia and elevated 1,25(OH)2D levels, including tuberculosis, leprosy, silicone-induced granulomatosis, and disseminated candidiasis [260,261]. Hypercalcemia in lymphoma patients due to elevations in 1,25(OH)2D is a well-known phenomenon and has been reviewed [265]. Both Hodgkin’s and non-Hodgkin’s lymphoma have been associated with elevated circulating 1,25(OH)2D [261,265]. Hypercalcemia in these disorders is estimated to occur in 5% of patients with Hodgkin’s disease and in 15% of patients with non-Hodgkin’s lymphoma. In one report, 1,25(OH)2D levels were elevated in 55% of a group of 22 hypercalcemic patients with non-Hodgkin’s lymphoma, and many of the normocalcemic patients with non-Hodgkin’s lymphoma had evidence of dysregulated 1,25(OH)2D synthesis [266]. Lymphocytes transformed with HTLV-1 have been shown to convert 25(OH)D to 1,25(OH)2D in vitro indicating that these lymhoma-like cells have 1-hydroxylase activity, and evidence shows that lymphomatous tissue in vitro can convert 25(OH)D to 1,25(OH)2D. However, whether the lymphoma cell itself or associated macrophages are responsible for the 1-hydroxylase activity found in lymphoma patients is still unclear at present [266]. Elevated 1,25(OH)2D concentrations are observed in pregnancy and appear to increase as gestation progresses [267]. DBP is stimulated by estrogens, and both total and free 1,25(OH)2D moieties are elevated during pregnancy and estrogen therapy [268, 269]. Only the free hormone is thought to be active [270]. The increased 1,25(OH)2D may augment the intestinal absorption of calcium that occurs during pregnancy, which is necessary to supply calcium to the developing fetal skeleton [267]. Extrarenal production of 1,25(OH)2D has been shown to occur in anephric
pregnant rats [271], in human placental tissue in vitro [103, 272], and in two pregnant women with pseudohypoparathyroidism, in whom renal production of 1,25(OH)2D was low [273]. The metabolism of vitamin D during pregnancy has been reviewed [274]. The use of the antifungal drug ketoconazole as a diagnostic test or as therapy for hypercalcemic states has been suggested [275 – 278]. Ketoconazole inhibits fungal growth by blocking the P450 enzyme 14-demethylase in the pathway to ergosterol synthesis [279]. The drug has been shown to inhibit mammalian P450 enzymes, including 24-hydroxylase [280] and 1-hydroxylase [281].
C. Local 1-Hydroxylases, Possible Autocrine/Intracrine Activity As a form of therapy in the preantibiotic era, patients with tuberculosis were often sent to sanitoria to increase their exposure to sunlight. It now seems possible that the benefits of sunlight may in part have been due to increased vitamin D synthesis. Data suggest that local 1-hydroxylase expression by macrophages leads to functionally important local production of 1,25(OH)2D, which acts in an intracrine, autocrine, and/or paracrine fashion. There is developing evidence that vitamin D deficiency is associated with an increased risk of contracting tuberculosis [282,283]. Furthermore, the synthesis of 1,25(OH)2D by macrophages may suppress Mycobacterium tuberculosis growth, perhaps by altering the local release of cytokines or nitric oxide [284]. Alternatively, 1,25(OH)2D may enhance macrophage function, perhaps by stimulating the local production of Eta-1 to enhance cell-mediated immunity [285]. In addition, analysis of polymorphisms in the VDR, especially when coupled with vitamin D deficiency, suggest that there is an increased risk of tuberculosis in certain populations [283]. On the other hand, in inflammatory arthritis, cytokines may elicit the local production of 1,25(OH)2D in the joints where 1,25(OH)2D may contribute to periarticular bone loss [286,287]. 1-Hydroxylase activity has been found in human colon [102] and prostate [104] tissue and cancer cell lines. Although the reason for its expression in these organs is unknown, it is possible that the local production of 1,25(OH)2D serves as an antiproliferative [288 – 291] and/or prodifferentiating [102,218,288,292] agent in colon and/or prostate development. Further studies in this area will help clarify the importance of 1-hydroxylase expression in these organs. Keratinocytes express 1-hydroxylase activity, and cultured keratinocytes from skin biopsies have been used to examine mutations that are associated with 1-hydroxylase deficiency [76,293]. 1,25(OH)2D acts as as antiproliferative and prodifferentiating agent in keratinocytes [35,294] and
275
CHAPTER 9 Vitamin D
is useful in treating psoriasis [295]. Recent data suggest that the beneficial effects of 1,25(OH)2D on psoriasis are due to its antiproliferative/prodifferentiating actions, as well as its ability to inhibit inflammation in dermal cells [296,297].
VII. GENETIC DISORDERS AND VITAMIN D RECEPTOR POLYMORPHISMS Examples of both over- and underproduction of the 1hydroxylated vitamin D sterols are not uncommon. Disorders associated with increased renal production of 1,25(OH)2D include hyperparathyroidism and tumoral calcinosis. Conditions that have decreased production of 1,25(OH)2D as part of their clinical picture include hypoparathyroidism and pseudohypoparathyroidism, renal failure, X-linked hypophosphatemic rickets, oncogenic osteomalacia, and hereditary 1-hydroxylase deficiency [298 – 303].
A. 1-Hydroxylase Deficiency (VDDR-I, PDDR) The clinical findings of hereditary complete deficiency of renal 1-hydroxylase were first described in 1961 by Prader et al. [304]. This disease was known as vitamin D-dependent rickets type I (VDDR-I) or pseudo vitamin D deficiency type I and, more recently, as pseudo vitamin D deficiency rickets (PDDR). This chapter refers to this genetic disorder as 1-hydroxylase deficiency now that the disease is proven to be caused by mutations in the cytochrome P450 1-hydroxylase gene (refered to as either CYP27B1 or CYP1). 1-Hydroxylase deficiency is a rare autosomal recessive disease that is manifested at an early age [76,84,302,305,306]. Hypocalcemia, elevated PTH levels, increased alkaline phosphatase, and low urine calcium are found. Affected children present with hypotonia, muscle
weakness, growth failure, and rickets. Tetany and convulsions may occur with severe hypocalcemia. As expected, patients with 1-hydroxylase deficiency have normal 25(OH)D concentrations and low levels of 1,25(OH)2D. Circulating 1,25(OH)2D does not increase after PTH infusion, consistent with defective 1-hydroxylase activity. Very large doses of vitamin D or 25(OH)D are required for adequate treatment of 1-hydroxylase deficiency; often 20,000 to over 100,000 IU of vitamin D daily is needed. However, modest doses of 1,25(OH)2D (0.25 – 2 g/day), which bypass the deficient enzyme, tend to be sufficient to restore calcium to normal and heal the rickets [307]. 1-Hydroxylase deficiency was presumed to be due to mutations in the gene encoding the 1-hydroxylase enzyme, and family linkage studies localized the defect to chromosome 12q14, close to the locus for the VDR [78, 79]. Since the cloning of the 1-hydroxylase gene, several groups have demonstrated that this disease is caused by mutations in the 1-hydroxylase gene. A number of missense mutations scattered throughout the entire region of the 1-hydroxylase gene have been identified that disrupt the enzyme activity. In one particularly interesting case, a patient with 1-hydroxylase deficiency was shown to have two mutations: a frameshift and a deletion on separate alleles, which together resulted in two null mutations [76]. Figure 8 illustrates the location of 1-hydroxylase mutations thus far reported [72,76,77,293,302,305,308].
B. Hereditary 1,25-Dihydroxyvitamin D-Resistant Rickets (HVDRR) Hereditary 1,25-dihydroxyvitamin D-resistant rickets (HVDRR), also known as vitamin D-dependent rickets type II (VDDR-II) or pseudo-vitamin D deficiency type II, is a rare genetic disease that arises as a result of mutations in the gene encoding the VDR [7,118,309 – 312]. This autosomal recessive disease is characterized by early onset rickets, hypocalcemia, secondary hyperparathyroidism, and, in
FIGURE 8 Mutations in the CYP27B1 gene causing 1-hydroxylase deficiency. Boxes represent exon sequences with intron sequences illustrated by connecting lines. Shaded boxes include untranslated sequences. Reproduced with permission from S. Kato [Mol. Cell. Endocrinol. 156, 7 – 12 (1999)].
276 contrast to 1-hydroxylase deficiency, normal or elevated serum 1,25(OH)2D levels. The heterozygotic parents have no evidence of bone disease, but a history of consanguinity is usually present. In many cases, total body alopecia accompanies the disease and provides initial evidence of the HVDRR syndrome [7]. The first description of a mutation in any member of the steroid receptor gene superfamily causing a hormone resistant phenotype was presented by Hughes et al. [313], who identified missense mutations in two unrelated families whose VDRs exhibited abnormal DNA binding [314,315]. In one family, a Gly33Asp mutation was identified in the first zinc finger module of the DBD, whereas in the second family, a Arg73Gly mutation was identified in the second zinc finger of the DBD [313]. The mutations were recreated by site-directed mutagenesis of the wild-type VDR cDNA and were shown to exhibit the properties of the native
FELDMAN, MALLOY, AND GROSS
mutant VDR, as well as to be defective in 1,25(OH)2D3induced gene transactivation [316]. These results clearly established that defects in the VDR gene were the etiology of 1,25(OH)2D resistance and the origin of the HVDRR phenotype. Several other mutations in the DBD have since been described [169,317 – 319] and all of the mutations reported are summarized elsewhere [7] (see Fig. 9). The first LBD mutation to be described was a nonsense mutation (TAC to TAA) that resulted in the premature termination of the VDR at amino acid 295 (Tyr295stop) [320]. The same mutation was subsequently found in multiple interrelated families comprising a large kindred [321]. Several other mutations that result in premature termination of the VDR have now been described [178,322]. Two premature termination signals occurred as a result of frameshifts. In one case, a 5 donor splice site was mutated, which caused exon 4 to be skipped in the processing of the VDR
FIGURE 9 Mutations in the VDR causing hereditary vitamin D-resistant rickets (HVDRR). (A) The two zinc finger modules and the amino acid composition of the DBD. Conserved amino acids are depicted as shaded circles. Mutations are indicated by large arrows and circles. The location of the intron separating exon 2 and exon 3, which encode the separate zinc finger modules, is indicated by an arrow. Numbers specify amino acid number. (B) Location of -helices (H1-H12) of the VDR LBD. -helices are depicted as open boxes, and the region containing turns is drawn as a shaded box. E1 and AF-2 regions are shown above the -helices. The location of the mutations is indicated by arrows. fs, frameshift mutation.
CHAPTER 9 Vitamin D
transcript. In the other case, a cryptic 5 splice site was generated by a mutation in exon 6 (Fig. 9). The first missense mutation identified in the VDR LBD was an Arg274Leu. The mutant receptor exhibited a decreased binding affinity for [3H]1,25(OH)2D3 of about 1000-fold [178]. Crystallographic studies of the VDR have shown that Arg274 is a contact point for the 1-hydroxyl group of 1,25(OH)2D3 [175]. Several other missense mutations have now been identified in the VDR LBD (Fig. 9). A His305Gln mutation caused a 5- to 10-fold reduction in [3H]1,25(OH)2D3 binding and a similar reduction in gene transactivation [179]. Crystallographic studies of the VDR have shown that His305 makes contact with the 25hydroxyl group of 1,25(OH)2D3 [175]. An Arg391Cys mutation was shown not to affect ligand binding but interaction with RXR [323]. In one special case, a patient who exhibited all the signs of HVDRR, including alopecia, no mutation was found in the VDR [324]. This case illustrates the fact that additional proteins are involved in 1,25(OH)2D3 signaling and that disruption of these proteins may cause the HVDRR syndrome. Indeed, a recently published report suggests that genetic defects in coactivator molecules are associated with steroid hormone-resistant syndromes [325]. A prenatal test for HVDRR has been described in a study using amniotic fluid cells or chorionic villus samples [326]. HVDRR was confirmed using assays to determine the level of [3H]1,25(OH)2D3 binding and 1,25(OH)2D3induced 24-hydroxylase activity, as well as by restriction fragment length polymorphism (RFLP) analysis [327]. The successful treatment of children with HVDRR who are unresponsive to large doses of vitamin D derivatives or oral calcium supplements has been achieved by the chronic intravenous administration of calcium [328 – 330]. The intravenous calcium infusions were given nightly over a period of many months and, by bypassing the intestinal defect in calcium absorption, over time were able to correct the hypocalcemia. The treatment eventually resulted in the normalization of serum calcium levels, correction of secondary hyperparathyroidism, and normal mineralization of bone and healing of rickets on X-ray. The clinical improvement can be sustained if adequate serum calcium and phosphorus concentrations are maintained. Despite healing of the rickets, the alopecia does not improve as a consequence of the treatment. In 1997, VDR knockout mouse models were generated by two groups. Yoshizawa et al. [331] disrupted exon 2 to generate the VDR null (-/-)genotype, whereas Demay and colleagues [332, 333] deleted exon 3. In both models, VDR (-/-) mice were phenotypically normal at birth, suggesting that 1,25(OH)2D3 actions are not necessary for normal embryogenesis. After weaning, the mice became hypocalcemic and developed rickets similar to patients with HVDRR. Alopecia also appeared progressively as
277 the mice aged. Most of the VDR (-/-) mice generated by Yoshizawa et al. [331] were infertile and died by 15 weeks after birth. Uterine hypoplasia and impaired folliculogenesis were also noted. Mice generated by Li et al. [332] survived at least 6 months. In both mouse models, the survival of the knockout mice was enhanced by a high calcium diet supplemented with lactose [333]. Many, but not all, of the abnormalities in the reproductive organs were eliminated by the maintenance of normal calcium levels [334]. A role for 1,25(OH)2D3 in regulating estrogen levels via its regulation of aromatase gene expression [335] was not corrected by calcium repletion [334]. As in the human disease HVDRR, normalization of calcium eliminated many of the abnormalities, but not alopecia.
C. X-Linked Hypophosphatemic Rickets (XLH) X-linked hypophosphatemia (XLH) is an X-linked dominant disorder causing phosphate wasting and hypophosphatemia [299,303]. The primary defect is expressed as an abnormality of the renal proximal tubule that impairs phosphate reabsorption. Patients with XLH develop progressively severe skeletal abnormalities and growth retardation. The clinical presentation ranges from mild abnormalities to severe rickets and osteomalacia. Children with the disease exhibit rachitic bone deformities, including enlargement of the wrists and knees and bowing of the lower extremities secondary to rickets. The clinical features of XLH are not apparent until 6 – 12 months of age and may also include defects in tooth development and premature cranial synostosis. The disorder is also associated with low or inappropriately normal circulating levels of 1,25(OH)2D in the presence of low serum phosphate, which would normally increase 1-hydroxylase activity, suggesting the defective regulation of 1-hydroxylase. The gene causing XLH has been cloned and identified as PHEX, a PHosphate-regulating gene with homologies to Endopeptidases located on the X chromosome [336,337]. The PHEX gene is homologous to a family of endopeptidase genes, which includes endothelin-converting enzyme1 and neutral endopeptidase. The PHEX gene encodes a 749 amino acid membrane-bound protein that is expressed in bone, adult ovary, lung, and fetal liver. A number of genetic defects in the PHEX gene have been identified in XLH patients, including deletions, insertions, duplications, splice site, and missense and nonsense mutations [337]. The X-linked dominant expression of the disorder is likely due to haploinsufficiency rather than dominant-negative effects, as many of the mutations are inactivating mutations [303]. A possible role for PHEX in the pathophysiology of XLH is illustrated in Fig. 10 [303]. As seen in the model,
278
FELDMAN, MALLOY, AND GROSS
FIGURE 10 Model for X-linked hypophosphatemia. In the osteoblast, both the PHEX protein and phosphatonin are made. Phosphatonin is secreted from the cell in an active state (PTNa). Under normal conditions, the PHEX protein degrades some of the phosphatonin to an inactive polypeptide (PTNi). The remaining active phosphatonin is then able to bind to a membrane receptor on the renal tubule cell surface. A signal is then transmitted to the cell to downregulate NPT2 activity and regulate phosphate reabsorption. In XLH, defects in the PHEX protein prevent the degradation of phosphatonin, allowing for greater circulating levels of the hormone. The signal transmitted by the higher phosphatonin levels is amplified, resulting in a greater inhibition of NPT2 activity, leading to phosphate wasting. Reproduced with permission from M. K. Drezner [Kidney Int. 57, 9 – 18 (2000)].
normal osteoblasts express both PHEX and “phosphatonin,” a putative phosphate-lowering peptide hormone. The membrane-bound PHEX degrades some of the active phosphatonin to an inactive metabolite. The remaining phosphatonin in the circulation then interacts with a renal tubule cell receptor. Through an unknown mechanism, a signal is sent to downregulate to a small degree the sodium-dependent phosphate cotransporter (NPT2) in the kidney. In XLH, defective PHEX proteins are unable to degrade phosphatonin, leading to excessive amounts of this protein in the circulation. Consequently, the signal to downregulate the NPT2 is magnified, leading to phosphate wasting [303]. The pathophysiology of XLH is similar to oncogenic osteomalacia or tumor-induced osteomalacia (TIO) in which phosphate depletion predominates [338]. However, in TIO, the acquired hypophosphatemic state is secondary to a tumor, which is generally a small benign lesion of mesenchymal origin. Removal of the tumor leads to the remission of the clinical abnormalities, suggesting that a putative circulating factor, “phosphatonin,” is produced by the tumor and causes phosphate wasting by the kidney [303]. As in XLH, phosphatonin in TIO has been postulated to act on signaling pathways that affect the NPT2 and clearance of phosphate [303].
D. VDR Polymorphisms The role of VDR polymorphisms in osteoporosis is covered extensively in several reviews [339, 340] and in Chapter 26. Here we briefly discuss the VDR polymorphisms and their potential role in predicting the risk of developing osteoporosis. Whereas mutations in the VDR gene can cause dramatic changes in gene function [7], polymorphisms may have no effect on function, or they may be functional, but the functional difference between the variants is usually subtle [171]. Osteoporosis has strong polygenic influences, and variance in bone mineral density (BMD) is estimated to be 50 – 80% heritable [339 – 341]. The potential for polymorphic variants affecting BMD and contributing substantially to this heritable risk was originally raised by Morrison et al. [342]. They identified several polymorphisms in the VDR gene and focused on three sites in the 3 region of the gene. The sites, denoted by the restriction enzyme that identified the variants by RFLP, were BsmI and ApaI located in the intron between exon, 8 and 9, and TaqI, located in exon 9 (see Fig. 11). A capital letter (e.g., B) denotes the absence of the restriction site, whereas a lowercase letter (e.g., b) denotes the presence of the restriction site in RFLP analysis. Morrison et al. [342] hypothesized that these VDR polymorphisms were associated with approximately a full standard deviation difference in BMD and
CHAPTER 9 Vitamin D
279
FIGURE 11 Polymorphisms in the human VDR chromosomal gene. The six 5 noncoding exons and 8 coding exons are depicted as shaded boxes. The location of the start codon polymorphism (SCP) is shown above exon 2 and the FokI polymorphism it generates is shown below the exon. The lowercase f is used to indicate the absence of the FokI restriction site and a 427 amino acid protein. The uppercase F is used to indicate the presence of the FokI site and a 424 amino acid protein. The location of the B, A, and T polymorphisms comprising the BsmI, ApaI, and TaqI restriction sites are shown. A variable length poly(A) microsatellite in exon 9 is also shown.
approximately 70% of the genetic variation in BMD. The possibility that a single gene, even one as central to bone metabolism as the VDR, contributed a major portion of the genetic basis of osteoporosis was an exciting but controversial hypothesis. However, multiple studies that followed in the wake of this paper either could not confirm the effect or found smaller differences in BMD [343 – 345]. A meta-analysis of 16 studies concluded that these polymorphisms contributed only a small effect on BMD, in the range of 1 – 2% [344]. Cooper and Umbach [344] summarized these findings by saying that the VDR polymorphisms represented one genetic factor affecting BMD, but further research into the mechanisms, clinical significance, and its relation between other genetic and environmental factors was needed. Although Howard et al. [346] eventually retracted some of their findings, their paper ignited great interest in the genetic basis of osteoporosis. Whereas the 3 polymorphisms did not change the structure or function of the VDR [343], interest soon developed in a polymorphism at the translation start site (ATG → ACG), which changed the amino-terminal end of the VDR [347]. The polymorphic variants detected by FokI RFLP cause translation to start at either the first methionine (denoted f or M1), generating a 427 amino acid protein, or a second methionine (denoted F or M4) three amino acids downstream of the first methionine generating a 424 amino acid protein. The F/M4 short VDR was found to have significantly more functional activity than the f/M1 long VDR [170]. Jurutka et al. [172] showed that the decrease in f/M1 activity was due to a decrease in binding of the transcription factor TFIIB. Several population studies have shown that f/M1 homozygotic individuals have decreased BMD [170, 347 – 350], but not all populations showed this effect [351]. Importantly, Ames et al. [352] showed that f/M1
polymorphism was associated with a decrease in BMD in healthy children aged 7 – 12. Moreover, ff homozygotes show a decrease in intestinal calcium absorption. FF homozygotes had a mean calcium absorption that was 41.5% greater than ff homozygotes and 17% greater than Ff heterozygotes. These results suggest a substantial relationship between FokI VDR polymorphism and bone metabolism, both at the level of intestinal calcium absorption and BMD. Other studies also emphasize the importance of the interaction of environmental factors, such as dietary calcium intake, with genetic factors, such as VDR polymorphisms [350]. It seems clear that osteoporosis is a polygenic disease and that polymorphisms in the VDR do not fully explain all of the heritable aspects. As discussed in Chapter 26, full genome scans and other approaches are actively being pursued to discover additional loci that contribute to the inheritance of traits that are associated with low BMD and increased fracture risk. It seems likely that an interaction of multiple polymorphisms and environmental factors will cumulatively determine osteoporotic risk. Interestingly, polymorphisms in the VDR considered to be associated with osteoporosis are also associated with the risk of developing prostate cancer [353]. Ingles et al. [354] and Taylor et al. [355] showed that polymorphisms at the 3 end of the gene, and a poly (A) track, 16 – 22 nucleotides in length, in the 3-untranslated region of the VDR mRNA, are associated with an approximate four fold increase in prostate cancer risk. However, as in the studies of osteoporotic risk, apparently not all populations show this association [356]. A connection between VDR polymorphisms and the risk of other malignancies, including breast and colon cancers, is underway. Initial studies in breast cancer suggest that VDR polymorphisms are associated with
280
FELDMAN, MALLOY, AND GROSS
breast cancer risk and progression [357 – 360]. These VDR polymorphisms have also been associated with a number of other conditions, including osteoarthritis and hyperparathyroidism [343], which are discussed further in Chapter 26.
VIII. 1,25(OH)2D3 ANALOGUES WITH DECREASED CALCEMIC ACTIVITY A. Agonists In addition to being a major regulator of calcium metabolism, 1,25(OH)2D exhibits many nonclassical actions in the body, including affecting cell growth, promoting cell differentiation, and suppressing the immune response (see Section IX). These properties make 1,25(OH)2D3 an attractive candidate for treating a number of serious diseases. However, to treat these diseases effectively, the dose of 1,25(OH)2D3 might well be in the range that would induce hypercalciuria, hypercalcemia, and renal stones; therefore, these unfavorable side effects limit its clinical utility. However, structural analogues of 1,25(OH)2D3 have been developed that exhibit a reduced calcemic response compared to 1,25(OH)2D3, yet retain many of the other therapeutically useful properties of the hormone, thus increasing their therapeutic potential [10,361 – 367]. Multiple analogues have been developed by the Roche company [368], the Leo Company [369], and the Chugai Co. [370], as well as by various investigators [135,371,372]. Although scores of analogues have been studied, the more effective analogues thus far have been the ones with sidechain modifications [371]. Changes in the 1,25(OH)2D3 molecule include the insertion of extra carbons, oxygen or
FIGURE 12
unsaturation in the carbon side chain, 16-ene derivatives, 19-nor derivatives, 20-epi derivatives, and 3-epi derivatives. The structures of a few of the more prominently studied analogues are depicted in Fig. 12. Many analogues have been shown to have a reduced calcemic response and a greater growth inhibitory potency compared to 1,25(OH)2D3. The mechanism for the differential activity displayed by the analogues is not totally clear but may be related to a number of properties: (a) decreased binding to DBP [373,374], (b) altered metabolic clearance and/or production of metabolites that retain significant biological activity [133,137,375], (c) increased ability to induce dimerization with RXR [376] or recruit coregulatory proteins [377], (d) increased ability to act preferentially to maintain an active conformation of the VDR within selected target tissues or on a limited number of target genes [378,379] and (e) ability to prevent degradation of the VDR [380]. There has also been progress in developing nonsteroidal molecules that mimic 1,25(OH)2D binding to the VDR [381,382]. This approach may make it possible to employ a much larger array of analogues and possibly lead to the concept of designer vitamin D drugs that exhibit specific target gene activation and thus have minimal side effects. The vitamin D analogues MC903, EB1089, and KH1060 have been developed by the Leo Pharmaceutical Company [369]. MC903, also known as calcipotriol, is less potent than 1,25(OH)2D3 in causing hypercalcemia, yet it is equivalent to 1,25(OH)2D3 in inhibiting the growth and in inducing differentiation of the human histiocytic lymphoma cell line U937. Calcipotriol has a low affinity for DBP and is metabolized rapidly to inactive metabolites when administered systemically. The drug is currently marketed for the topical treatment of psoriasis [295]. EB1089 exhibits approximately the same
Structure of 1,25(OH)2D3 and side chain modifications in four analogues with reduced calcemic activity.
281
CHAPTER 9 Vitamin D
affinity for the VDR as 1,25(OH)2D3, but is more potent than 1,25(OH)2D3 in inhibiting the growth of multiple cancer cell lines [291,383 – 386]. This drug is in clinical trials for the treatment of breast, pancreatic, and other cancers [387,388]. KH1060, also less calcemic than 1,25(OH)2D3, affects the immune function, including inhibiting IL-2-induced T lymphocyte differentiation [389]. This property may make KH1060 useful in treating autoimmune diseases, organ transplantation, and other conditions requiring immunosuppression. The vitamin D analogue 22-oxa-1,25(OH)2D3 (OCT), developed by Chugai Pharmaceuticals [370], has a lower affinity for VDR than 1,25(OH)2D3 but is 10 times more potent than 1,25(OH)2D3 in differentiating the myeloid leukemia cell line HL-60 and 100-fold less active in bone mobilization [390]. OCT, like 1,25(OH)2D3, also suppresses PTH production and is a potent inhibitor of renal 1-hydroxylase activity. OCT is being studied for use in chronic renal failure patients to inhibit excessive PTH secretion [370]. A number of analogues synthesized by the Roche company have shown to have increased antiproliferative activity and decreased calcemic activity [368]. A number of these analogues have been studied in various in vitro and in vivo models of cancer with promising results (see Section X) [391 – 397].
B. Antagonists Among the various naturally occurring metabolites of 1,25(OH)2D is (23S, 25R)-1-hydroxyvitamin D3-26,23-lactone. This molecule has some 1,25(OH)2D-like activities in dogs and humans, but when given to vitamin D-deficient rats in pharmacological doses, it led to hypocalcemia [398]. This finding suggested possible antagonist activities, and a series of analogues were studied to determine whether a vitamin D antagonist could be developed. The most promising candidate thus far is the analogue (23S)-25-dehydro-1-hydroxyvitamin D3-26,23-lactone (TEI-9647), which has been shown to have antagonist activity in several systems, including HL-60, SaOS-2, and MG-63 cells [398, 399]. The analogue binds to the VDR and appears to prevent heterodimer formation with RXR and subsequent recruitment of the coactivator SRC-1 [399]. TEI-9647 has a small amount of agonist activity, suggesting that it is a partial agonist/antagonist. The major action is as an antagonist, which may be clinically useful in selected hypercalcemic states.
IX. ACTIONS OF VITAMIN D IN CLASSICAL TARGET ORGANS TO REGULATE MINERAL HOMEOSTASIS The classical actions of 1,25(OH)2D on intestine, bone and kidney (see Fig. 7) include improved efficiency of in-
testinal calcium absorption, increased calcium mobilization from bone, and maintenance of adequate concentrations of calcium and phosphate in the extracellular fluid to promote the normal mineralization of bone [14,33,127]. In recent years, additional mechanisms by which 1,25(OH)2D modulates calcium homeostasis have been demonstrated, including autoregulation of 1,25(OH)2D synthesis as well as regulation of the calciotropic peptides PTH and calcitonin. These classical actions of 1,25(OH)2D to regulate calcium homeostasis are discussed in this section. The nonclassical, newly recognized actions of 1,25(OH)2D on many additional target cells, apparently unrelated to maintenance of systemic mineral homeostasis, are discussed in Section X.
A. 1,25(OH)2D Actions in Intestine 1. OVERVIEW OF CALCIUM ABSORPTION Calcium and phosphate are absorbed along the length of the small intestine. Calcium is mostly absorbed proximally in the duodenum, whereas vitamin D-dependent phosphate absorption occurs more distally in the jejunum and ileum [29]. VDR are present along the entire course of the small intestine, with the highest concentration proximally and the levels decreasing distally [141]. The abundance of VDR in the duodenum is the highest of all organs reported, and at any cross-sectional level along the intestine, VDR content is highest in crypts and decreases as the cells progress up the villus [400]. VDR are also present throughout the colon [401] and are expressed in colon cancer cell lines as well as in cancer specimens removed at surgery [218, 402 – 404]. Three mechanisms for intestinal calcium absorption have been described [29]. (i) In the absence of vitamin D, calcium is absorbed by a paracellular passive route; the rate of absorption is driven by mass action and is a function of the calcium concentration. (ii) In the presence of 1,25(OH)2D, a transcellular, saturable process is stimulated. This 1,25(OH)2D-dependent process takes some hours to develop and enhances calcium absorption greatly. (iii) Transcaltachia is the process of very rapid change in calcium flux that occurs within minutes in isolated perfused duodenum and is believed to be stimulated by vitamin D but via a nongenomic pathway [246]. Transcaltachia is described further in Section IV,B on nongenomic actions of vitamin D. Details of the process of calcium absorption are still not completely delineated at the molecular level [29]. In pioneering studies, Wasserman et al. [405] found that vitamin D induced calcium-binding proteins, now known as calbindins, in chick intestinal mucosa. 1,25(OH)2D induces the 9000 Mr vitamin D-dependent calbindin-D9k in mammalian intestine and also calbindin-D28k and calbindin-D9k in mammalian kidney [406,407]. In intestine, these
282 high-affinity calcium-binding proteins are located in the cytosol of columnar epithelial cells of the intestine and are involved in the translocation of calcium. However, their role is not completely defined and their induction alone cannot explain the process of calcium absorption, as calbindin-D levels are not correlated directly with the time course of 1,25(OH)2D-mediated calcium transport. Nevertheless, the calbindins have been an important tool for studying 1,25(OH)2D action on the intestine [29,406,407]. Two other proteins that are also regulated by vitamin D and probably play a role in calcium absorption are the integral membrane calcium-binding protein (IMCAL) [408] and Ca,Mg-dependent ATPase (Ca2 pump) located in the basolateral membrane of intestinal and renal cells, which uses ATP as an energy source to pump Ca2 out of the cell across a concentration gradient [409]. In addition to regulating these gene products, 1,25(OH)2D also plays a role in intestinal cell differentiation, elongating the villi, and inducing the polyamine pathway [410]. A combination of these and other effects coordinately expressed in the intestine mediate 1,25(OH)2D action to enhance calcium absorption. A model of transcellular intestinal calcium transport includes three sequential steps with calcium moving from protein to protein along an uphill gradient of calcium-binding affinities from apical to basolateral membrane [29]: (i) rapid entry of calcium at the brush border with transient sequestration subjacent to the microvillar membrane; (ii) transfer of calcium from the brush border complex to calbindin-D, which has a higher affinity for Ca2; and (iii) eventual transfer to the Ca/Mg-dependent ATPase or Ca2 pump, which has the highest affinity for Ca2 and extrusion against a concentration gradient at the basolateral membrane. This process may be similar in kidney where equivalent molecular species exist [257]. The overall contribution to calcium transport of the transcaltachia process [246], an endocytic pathway proposed to rapidly transport calcium via exocytosis of Ca2 containing vesicles by a nongenomic pathway, remains to be determined [29]. 2. ACTION OF VITAMIN D METABOLITES ON CALCIUM ABSORPTION Heaney and colleagues [255, 411, 412] have investigated the calcium absorptive response to graded doses of vitamin D3, 25(OH)D, and 1,25(OH)2D in healthy adult men. While no relationship was found between baseline absorption and serum vitamin D metabolite levels, all three vitamin D compounds significantly elevated 45Ca absorption from a 300-mg calcium load given as part of a standard test meal. 1,25(OH)2D was active even at the lowest dose (0.5 g/day), and the slope was such that doubling of absorption would occur at an oral dose of approximately 3 g/day. 25(OH)D was also active in elevating absorption and did so without raising total 1,25(OH)2D levels. On the
FELDMAN, MALLOY, AND GROSS
basis of the dose – response curves for 1,25(OH)2D and 25(OH)D, the two compounds exhibited a molar ratio for physiological potency of approximately 100:1. The absorptive effect of vitamin D3 was seen only at the highest dose level (1250 g, or 50,000 IU/day) and was apparently mediated by conversion to 25(OH)D. Analysis of pooled 25(OH)D data from both 25(OH)D- and vitamin D3-treated groups suggests that approximately one-eighth of circulating vitamin D-like absorptive activity under untreated conditions in winter may reside in 25(OH)D. This is a substantially larger share than has been predicted from studies of in vitro receptor binding [255,411,412]. 3. CHANGES IN CALCIUM ABSORPTION WITH AGE Osteoporosis is often associated with decreased intestinal calcium absorption with increasing age, and this phenomena is speculated to contribute to its pathogenesis [413, 414]. In rats there is an age-related decrease in the induction of calbindin protein in response to 1,25(OH)2D in the duodenum, but not in the ileum or kidney [415]. This decline in protein expression may be due to decreased translation of calbindin D9k mRNA in the duodenum with age. Several studies have suggested that intestinal VDR declines with age in the rat [213]. Duodenal biopsies of human subjects showed a slight trend toward a decrease of VDR abundance in the intestine with age [416]. However, the change in VDR abundance did not correlate with calcium absorption efficiency [417]. Estradiol may be an additional regulator of calcium absorption, as a direct effect of estradiol on intestinal calcium absorption independent of 1,25(OH)2D has been demonstrated [418]. 4. HYPERCALCIURIA Idiopathic hypercalciuria, the commonest form of renal stone disease, is characterized by the hyperabsorption of calcium, hypercalciuria, and normal or elevated 1,25(OH)2D concentrations [419]. Hypercalciuria in genetic hypercalciuric stone-forming (GHS) rats has been studied as a model for human intestinal calcium hyperabsorptive conditions [419,420]. GHS rats with normal serum 1,25(OH)2D levels are hyperabsorptive and have a greater number of VDRs than normal in the intestine, kidney, and bone. Administration of 1,25(OH)2D3 increases VDR gene expression significantly in GHS but not normocalciuric animals. Results suggest that GHS rats hyperrespond to minimal doses of 1,25(OH)2D3 by upregulating VDR gene expression. This unique characteristic suggests that GHS rats may be susceptible to minimal fluctuations in serum 1,25(OH)2D3, which may pathologically amplify the actions of 1,25(OH)2D3 on Ca2 metabolism that thus contribute to the hypercalciuria and stone formation [420]. Whether this mechanism also causes some form of human hypercalciuria and renal stones remains to be proven.
283
CHAPTER 9 Vitamin D
B. 1,25(OH)2D Actions in Bone Bone undergoes constant remodeling involving osteoclast-mediated bone resorption and osteoblast-mediated bone formation (see Chapter 12). 1,25(OH)2D is a major regulator of both formation and resorption. The detailed actions of 1,25(OH)2D on bone are discussed more completely in Chapter 2 on osteoblasts, Chapter 3 on osteoclasts, and Chapter 4 on matrix proteins and mineralization. Vitamin D is necessary for normal mineralization of the skeleton, and when it is deficient, a mineralization defect develops, causing rickets in growing children and osteomalacia in adults [421]. 1,25(OH)2D actions on bone are complex, and both direct and indirect effects have been described. Direct actions on the bone are complicated further because 1,25(OH)2D actions appear to be induced in several bone cell types, including osteoblasts, bone stromal cells, and osteoclasts. In addition, the nature of the response to 1,25(OH)2D is dependent on the differentiation state of the bone cell. VDRs are expressed in osteoblasts [422,423] and direct actions of 1,25(OH)2D3 on these cells include the modulation of cell growth [210] and stimulation of differentiation [424]. 1,25(OH)2D3 induces osteoblasts to progress from immature, proliferating cells to differentiated, nondividing cells that synthesize matrix proteins and mineralize bone. Several 1,25(OH)2D3 upregulated gene products have been identified, including osteocalcin, Eta-1 (osteopontin), alkaline phosphatase, matrix gla protein (osteocalcin), and IMCAL (Chapter 2). Collagen synthesis, however, is downregulated by 1,25(OH)2D3 in osteoblasts [425]. Although 1,25(OH)2D has been well known to promote bone mineralization since its discovery as an antirachitic agent many years ago, there is no definitive evidence that direct actions of 1,25(OH)2D on bone are required for normal bone mineralization. Effects to promote mineralization appear to be due mainly to 1,25(OH)2D actions on the intestine to enhance calcium and phosphate absorption to ensure optimal delivery of these ions to bone-forming cells. This concept of permissive action is supported by studies showing a restoration of normal bone mineralization in the absence of vitamin D action when adequate calcium and phosphorous are provided by intravenous infusion to vitamin D-deficient rats, VDR knockout mice, and children with HVDRR [7]. In the latter situation, chronically administered i.v. calcium infusions, which bypass the intestinal site of 1,25(OH)2D action, can achieve normalization of serum calcium levels, reversal of secondary hyperparathyroidism, and promote healing of the mineralization defect of rickets, despite the fact that 1,25(OH)2D action at the bone is prevented because of defective VDR [7]. These studies highlight the essential role of 1,25(OH)2D action on the intestine and indicate that the actions of the hormone on bone are indirect in regard to the process of mineralization.
There are nontheless many consequential effects of 1,25(OH)2D on bone, often in conjunction with PTH [424]. It has been known for many years that 1,25(OH)2D stimulates bone resorption [426]. This effect appears to be due to 1,25(OH)2D actions to directly stimulate the differentiation of precursor cells, mononuclear phagocytes of the macrophage lineage, to fuse into mature multinucleated osteoclasts [427]. This process, osteoclastogenesis, involves a complex interaction of osteoclast precursor cells, osteoblasts, and bone stromal cells. Together with other factors, 1,25(OH)2D promotes the early stages of osteoclastogenesis by direct actions on the osteoclast precursor cells. During the later stages of this differentiation process, the developing osteoclasts seem to lose their VDR, and 1,25(OH)2D stimulation of differentiation becomes indirect by acting on cells in the osteoblast lineage, possibly osteoblast stromal cells, to induce osteoclast differentiating factor(s) (see Chapter 3). Studies reveal that osteoclastogenesis is regulated by osteoclast differentiation factor (ODF) or RANKL (RANK ligand), an osteoclastogenic factor of osteoblastic origin, and osteoprotegerin (OPG), a potent inhibitor of osteoclastogenic activity. These factors are discussed extensively in Chapters 3, 6, and 12–14. RANKL is a member of the tumor necrosis factor (TNF) ligand family and OPG is a member of the tumor necrosis factor receptor family. Overexpression of OPG in transgenic mice leads to osteopetrosis, whereas OPG knockout mice develop severe osteoporosis. The ratio of RANKL to OPG determines the level of osteoclastogenic activity, and 1,25(OH)2D3 has been shown to regulate both of these factors. However, like PTH and IL-2, 1,25(OH)2D3 increases the ratio of RANKL:OPG, leading to increased osteoclastogenesis and bone resorption [428].
C. 1,25(OH)2D Actions in Kidney VDR are present in kidney [429], and renal calbindinD28k [406] and calmodulin [430], among other proteins, are regulated by 1,25(OH)2D [431]. However, the most important renal actions of 1,25(OH)2D are probably the regulation of the 1- and 24-hydroxylases (see Fig. 2). 1,25(OH)2D has a short and a long loop feedback to regulate its own production (see Fig. 7). In the presence of adequate 1,25(OH)2D levels, the short loop feedback is a direct renal action of 1,25(OH)2D to inhibit 1-hydroxylase and to induce 24-hydroxylase gene expression. The two actions coordinately drive 25(OH)D into 24,25(OH)2D, an inactivation pathway, and inhibit further 1,25(OH)2D synthesis. The long loop feedback is via 1,25(OH)2D inhibition of PTH gene expression, as PTH is the major stimulator of 1-hydroxylase activity. The 1,25(OH)2D action on PTH is also mediated indirectly via 1,25(OH)2D regulation of
284
FELDMAN, MALLOY, AND GROSS
the serum Ca2 concentration, which will rise subsequent to the calcemic actions of 1,25(OH)2D on intestine and bone (see Fig. 7). 1,25(OH)2D has been implicated in regulating renal calcium and phosphate excretion; however, its role here is not as well defined. The components of the intestinal calcium transport system are also present in the kidney, including VDR and 1,25(OH)2D-dependent calbindin-D28k, as well as calbindin-D9k and a membrane calcium pump, all located within the same cells of the distal convoluted tubule. It is likely that 1,25(OH)2D stimulates calcium transport in the distal tubule via a calbindin D-dependent mechanism similar to the intestine [431]. In chronic renal failure, as the mass of functional renal tissue declines, the production of 1,25(OH)2D diminishes with resultant vitamin D insufficiency and secondary hyperparathyroidism. Coexisting hyperphosphatemia leads to the development of renal osteodystrophy [87,256]. In addition to preventing hyperphosphatemia, replacement of 1,25(OH)2D has become a cornerstone of managing patients with this syndrome. Initially, oral 1,25(OH)2D3 and then intravenous 1,25(OH)2D3 were used. Currently, newer less calcemic analogues are being studied for possible improved results, including 22-Oxa-1,25(OH)2D3 (22-oxacalcitriol or OCT), 19-nor-1,25(OH)2D2 (19-norD2), and 1(OH)D2 [363]. These analogues have been tested in animal models of uremia and in clinical trials [80]. Intravenous 19-norD2 and oral 1(OH)D2 have been approved for use in the United States; OCT is currently under review. The mechanisms by which these analogues exert their selective actions on the parathyroid glands, to suppress secondary hyperparathyroidism without causing hypercalcemia, are still under investigation.
D. 1,25(OH)2D Action on the Parathyroid Glands and Regulation of PTH The parathyroid glands possess VDR and are an important component of the systemic regulation of calcium homeostasis by 1,25(OH)2D [80,432]. The major effect of 1,25(OH)2D in this site is to suppress PTH secretion by inhibiting mRNA and protein synthesis. The other major regulator of PTH secretion is serum Ca2, which acts via the calcium sensing receptor in the parathyroid glands [94]. It has been suggested that the weight of parathyroid adenomas is related to vitamin D nutrition, indicating the importance of the feedback of vitamin D to inhibit parathyoid growth [245]. Patients with chronic renal failure develop secondary hyperparathyroidism, partly due to the decreased renal production of 1,25(OH)2D by the diseased kidneys. In addition, inappropriately elevated PTH secretion may result from decreased levels of VDR in the parathyroid gland of
uremic patients, resulting in the less efficient suppression of PTH synthesis by 1,25(OH)2D [242]. Studies indicate that the decrease in VDR is not distributed uniformly in parathyroid glands from chronic renal failure patients and that selected areas of low VDR content exhibit the most severe hyperplasia [433]. Suppression of elevated PTH in secondary hyperparathyroidism of chronic renal failure may be accomplished by the administration of 1,25(OH)2D3 or its analogues as described earlier. Better PTH suppression with less hypercalcemia is achieved with intermittent i.v. administration of 1,25(OH)2D3, which results from higher peak serum levels that are achieved with this regimen. The use of vitamin D analogues that elicit a reduced calcemic response, especially i.v. 19-norD2, oral 1(OH)D2 and oral OCT, may in the future yield improved PTH suppression without hypercalcemia and provide a more effective treatment for secondary or tertiary hyperparathyroidism in chronic renal failure [363].
E. Regulation of PTHrP and Calcitonin 1,25(OH)2D3 inhibits PTHrP expression in many normal tissues as well as malignant cells [434] but not all tissues (e.g., prostate), [435]. This may add to the beneficial effects of 1,25(OH)2D3 in the treatment of cancer with metastases to bone and/or in humeral hypercalcemia of malignancy. The less calcemic analogue of 1,25(OH)2D3, EB1089, was shown to adequately suppress PTHrP production by a squamous cell cancer xenografted into mice and reverse the hypercalcemic state caused by excess PTHrP [436]. Although use of a vitamin D preparation in a hypercalcemic state might at first appear counterintuitive, the less calcemic analogues may have a role in suppressing pathologic levels of PTHrP in humoral hypercalcemia of malignancy. Calcitonin is another calciotropic peptide hormone regulated by 1,25(OH)2D3 [437]. Inhibition of mRNA and protein expression has been demonstrated in vivo in rat and in vitro in medullary thyroid cancer cells. These issues are discussed in more detail in Chapters 7 and 8.
X. ACTIONS OF 1,25(OH)2D IN NONCLASSICAL TARGET ORGANS In recent years a number of additional actions of 1,25(OH)2D beyond merely regulating mineral homeostasis have been discovered in numerous nonclassical target organs. Many of these actions involve the promotion of cellular differentiation and inhibition of cell growth and appear to be unrelated to the regulation of total body calcium metabolism. VDR expression and 1,25(OH)2D effects have been demonstrated in a variety of cells and tissues, including hematopoietic, immunologic, epidermal, and cancer
285
CHAPTER 9 Vitamin D
cell systems. These diverse actions of 1,25(OH)2D and its analogues have been the subject of several reviews [152,365,434,438 – 440].
A. 1,25(OH)2D Effects on Cell Growth and Differentiation in Normal and Malignant Tissues VDR are expressed in many normal and malignant cell types, indicating a wide array of previously unrecognized potential targets for 1,25(OH)2D action [152]. In many of these nonclassical tissues, 1,25(OH)2D acts to inhibit cell proliferation [152,439], although in selected settings, 1,25(OH)2D may stimulate cell proliferation [441,442]. Also, in a number of systems, 1,25(OH)2D affects cells by promoting cellular differentiation [152,434,438,439]. The hormonal modulation of intracellular calcium and the regulation of expression of nuclear oncogenes [439,443] have been raised as possible mechanisms for these antiproliferative and differentiating effects of 1,25(OH)2D. Additional mechanisms are postulated to include inhibition of the passage of cells through the cell cycle, regulation of paracrine growth factors, stimulation of terminal differentiation, and induction of apoptosis [397,444 – 447]. The induction of p21 and p27, cell cycle-dependent kinase (Cdk) inhibitors, has emerged as a major mechanism for cell cycle arrest [386,448], whereas the downregulation of Bcl-2 and perhaps other antiapoptotic factors play a major role in the induction of apoptosis [444,447].
B. Possible Role of Vitamin D in Cancer Prevention or Therapy Because of its actions to inhibit cellular proliferation and stimulate cellular differentiation, 1,25(OH)2D has been considered for possible “chemoprevention” or “differentiation” therapy in a number of malignant cell types that possess VDR [434]. 1. COLON CANCER VDR are present in the colon [401] and in colon cancer cell lines, as well as in surgically removed colon cancers [449]. The possibility that calcium and/or vitamin D may be active in decreasing colon cancer has been examined by several groups [446,450]. Garland et al. [451] have suggested that higher levels of 25(OH)D protect against colon cancer. Eisman and coworkers showed that 1,25(OH)2D3 administration could inhibit the growth of colon cancer xenografts in nude mice [452]. A number of colon cancer models have been shown to be inhibited and/or differentiated by 1,25(OH)2D or its analogues, both in vitro and in vivo [218, 404, 446, 453].
2. BREAST CANCER VDR are present in normal breast and breast cancer cell lines and in many human cancer specimens [391,444]. 1,25(OH)2D3 was shown to suppress the growth of breast cancer cell lines, xenografts in nude mice, and carcinogen (nitrosomethyl urea, NMU)-induced breast cancer in rats [391]. Also, EB1089 (see Fig. 12), an analogue with reduced calcemic potency, has been shown to inhibit the proliferation of MCF-7 human breast cancer cells in vitro. In in vivo studies, EB1089 was found to be more potent than 1,25(OH)2D3 at inhibiting tumor growth induced by the carcinogen NMU and less likely to induce hypercalcemia. It therefore has been considered to have therapeutic potential as an antitumor agent. Clinical trials are currently underway to determine whether this antiproliferative action of vitamin D analogues on breast cancer will be effective in humans [387]. In a separate breast cancer prevention trial in rats, also using the NMU-induced breast cancer model, Anzano et al. [454] tested the effectiveness of the vitamin D analogue 1,25(OH)2D-16-ene-23-yne-26,27-hexafluoro. After 5 – 7 months of therapy they found that the compound extended tumor latency, lessened tumor incidence, enhanced the tamoxifen effect to reduce tumor burden, and increased the number of tumor-free rats. Also, the rats did not develop hypercalcemia as a result of treatment with this analogue. A number of investigators have shown that 1,25(OH)2D or its analogues are antiproliferative in breast cancer cell models and involve a number of different pathways [444,455 – 459]. In addition to its antiproliferative effects, evidence shows that 1,25(OH)2D stimulates apoptosis in some breast cancer cells [394]. 3. PROSTATE CANCER On the basis of geographic patterns of ultraviolet radiation throughout the contiguous United States and epidemiological data on prostate cancer incidence, a hypothesis was raised by Schwartz and colleagues that vitamin D deficiency may be a risk factor [460] and that increased sunlight exposure may protect against clinical prostate cancer [461]. In a prediagnostic study with stored sera, 1,25(OH)2D blood levels were found to be an important predictor for palpable and anaplastic tumors in men over 57 years of age but not for incidentally discovered or well-differentiated tumors [462]. VDR are present in prostate cancer cell lines [288,463] and in normal prostate [289], and 1,25(OH)2D3 has been shown to inhibit the growth of all these cell types in culture [288,289,392]. Furthermore, the prostate growth-inhibiting activity of selected vitamin D analogues with reduced calcemic potency (e.g., EB1089 and OCT) was even greater than 1,25(OH)2D3 [291], raising the possibility of the therapeutic potential of these drugs in the treatment of prostate cancer. A number of studies have demonstrated an antiproliferative activity of
286 1,25(OH)2D3 and vitamin D analogues in multiple prostate cancer models [292,393,395,397,447]. The induction of apoptosis may play some role in the antiproliferative activity in some prostate cancer cells [464]. Clinical trials have begun to address the utility of 1,25(OH)2D3 in treating prostate cancer patients [465,466]. 4. HEMATOPOIETIC CELLS: MYELOID CELLS AND LEUKEMIA In addition to promoting osteoclastogenesis from macrophage precursors described earlier in the section on bone (Section IXB), 1,25(OH)2D3 has been shown to stimulate a variety of immature hematopoietic myeloid cells to differentiate into mature cells, including M-1 mouse myeloid leukemic cells, HL-60 human promyelocytic leukemia cells, U-937 human monocytic cells, and peripheral human monocytes [445]. In these cells, VDR were present and the differentiation process was accompanied by the inhibition of cell proliferation. In HL-60 cells, the 1,25(OH)2D3-induced response appears to be due to the induction of terminal differentiation mediated by the inhibition of expression of the c-myc oncogene [467]. Liu et al. [448] have shown that 1,25(OH)2D stimulates myeloid leukemic cells lines to terminally differentiate into monocytes/macrophages. Using the myelomonocytic U937 cell line, they showed that 1,25(OH)2D induces the expression of the Cdk inhibitor p21 (WAF1/CIP1), which caused the cells to terminally differentiate [448]. These effects on leukemic cells in vitro as well as prolongation of survival time of mice inoculated with myeloid leukemia cells [468] have led to the consideration of using 1,25(OH)2D3 therapeutically in human leukemia as “differentiation” therapy [445]. Munker et al. [469] reviewed the potential use of 1,25(OH)2D and analogues to treat leukemia. Their work showed that 1,25(OH)2D differentiated both normal and leukemic cells. They suggested that 1,25(OH)2D alone, or in combination with retinoids or chemotherapeutic agents, would be useful in the treatment of patients and warrants clinical trials in patients with leukemia.
C. Immune System: 1,25(OH)2D Actions on Immunosuppression and Cytokines In addition to 1,25(OH)2D3 effects on myeloid cells described earlier and on monocytic/macrophage precursors that are differentiated into osteoclasts (described in Section IXB), 1,25(OH)2D3 has many important immunomodulatory effects [365,389,440,470]. Circulating resting T and B cells do not express VDR but when blast transformed or mitogen activated these cells do express VDR and respond to 1,25(OH)2D3 [439]. 1,25(OH)2D3 treatment of
FELDMAN, MALLOY, AND GROSS
VDR-positive cells results in growth inhibition and decreased IL-2 production. Monocytes and macrophages also express VDR. 1,25(OH)2D3 modulates the production of numerous interleukins, cytokines, and various oncogenes and transcription factors, including interleukins 1, 2, 3, and 6; interferon-; leukemia inhibitory factor (LIF); tumor necrosis factor (TNF); and granulocyte/macrophage colony-stimulating factor (GM-CSF), as well as several oncogenes, including c-myc, fos, myb, fms, and fgr [152, 438,439]. 1,25(OH)2D3 also suppresses the production of helper T cells. Although the actions of 1,25(OH)2D3 are predominantly inhibitory, some genes may be stimulated, depending on the activation state of the immune cells. In various animal models, 1,25(OH)2D3 reduces immune responses when administered prior to induction or early in the disease process [389,470]. 1,25(OH)2D3 inhibits the induction of several autoimmune diseases in animals, such as experimental autoimmune encephalomyelitis, lupus, thyroiditis, and type I diabetes. 1,25(OH)2-24-oxo-16ene-D3, a natural metabolite of the vitamin D analogue 1,25(OH)216ene-D3, can initiate immunosuppressive effects equal to the parent compound without causing hypercalcemia in vivo [471,472]. The significance of the immunomodulating properties of 1,25(OH)2D3 remains poorly understood; however, the possible applications to the clinical setting of leukemia, autoimmune disease, and transplantation are currently being explored. As described in Section VI,C, activated macrophages can synthesize 1,25(OH)2D from circulating 25(OH)D, as has been shown in sarcoid, tuberculosis, and other granulomatous diseases. Interferon- has been shown to stimulate the local production of 1,25(OH)2D in these cells [261]. 1,25(OH)2D inhibits lymphocyte proliferation, interleukin 2, and interferon- production, as well as other cytokines (IL-12) and immunoglobulins [438]. The local production of 1,25(OH)2D may have an autocrine/paracrine function, which may act locally to suppress the inflammatory response, particularly leading to the inhibition of helper T cell subset type 1 (Th1) [389]. The various findings suggest the possibility of therapeutic application of 1,25(OH)2D3 and analogues in autoimmunity and transplantation [473]. 1,25(OH)2D immunosuppressive activity has been well studied in the autoimmune model of diabetes that develops spontaneously in nonobese diabetic (NOD) mice [362]. Type I diabetes can be prevented without generalized immunosuppression by nonhypercalcemic analogues of 1,25(OH)2D when treatment is started early, i.e., before the autoimmune attack, reflected by insulitis. Even if the autoimmune disease is already active, treatment with 1,25(OH)2D analogues can prevent clinical diabetes when this therapy is combined with a short induction course of an immunosuppressant such as cyclosporin A.
287
CHAPTER 9 Vitamin D
D. 1,25(OH)2D Effects on Skin: Use in the Treatment of Psoriasis In addition to being the site of initiation of vitamin D synthesis, skin is also a 1,25(OH)2D target organ [35]. Human dermal fibroblasts and keratinocytes possess VDR and are 1,25(OH)2D3 responsive [474]. For this reason, cultured dermal fibroblasts are used frequently to study HVDRR [7,309, 310]. 1,25(OH)2D3 inhibits proliferation and stimulates terminal differentiation of keratinocytes, including stimulation of involucrin, cornified envelop development, and transglutaminase activity [35]. Other cells within the skin also contain VDR and appear to be 1,25(OH)2D3 targets as well. Melanoma cells express VDR, and 1,25(OH)2D3 induces differentiation and inhibits cell proliferation [475]. VDR are also present in the hair follicle within the skin [151]. It is of interest that children with HVDRR [476], as well as VDRablated mice [331,332], have alopecia; however, the mechanism is unknown. When HVDRR patients respond to treatment with calcium infusion and the bone mineralization defect of rickets is reversed, no improvement in alopecia is noted either in HVDRR patients [7] or in VDR-ablated mice [333]. Evidence from studies in VDR-ablated mice suggests that skin differentiation is not dependent on 1,25(OH)2D3 acting through its nuclear receptor, and the authors speculate that the associated alopecia in these animals may be due to nongenomically mediated vitamin D toxicity from elevated 1,25(OH)2D3 concentrations [477]. Psoriasis is a hyperproliferative disorder of the epidermis, which responds to treatment with vitamin D preparations applied topically or administered systemically [295,478]. The antipsoriatic effect may be due to the antiproliferative action of the hormone, but may also involve immunosuppressive properties induced by 1,25(OH)2D3 [296]. Newer vitamin D analogues with reduced calcemic activity are being developed to improve the therapeutic potential of treating this disease. Interestingly, in keratinocytes, the VDR levels are downregulated within a few hours after UV B irradiation [36]. These results strongly suggest the existence of a feedback mechanisms in that UV B initiates vitamin D synthesis in keratinocytes and at the same time limits VDR abundance. These findings provide an explanation for the reported lack of any additive effect between 1,25(OH)2D and UV B phototherapy in the treatment of psoriasis.
E. 1,25(OH)2D Action in the Nervous System: NGF, Alzheimer’s Disease, and Aging The first evidence for the presence of VDR in brain came from autoradiographic studies using [3H]1,25(OH)2D3
to localize the receptor [479]. In rodents, [3H]1,25(OH)2D3 binding sites were located throughout the brain from the basal forebrain to the midbrain and hindbrain [479, 480]. VDR were detected biochemically in the hippocampus by showing VDR mRNA expression using in situ hybridization [481]. Very little is known about 1,25(OH)2D actions in the central nervous system [482]. Calbindin-D28k in the brain is not vitamin D dependent; however, 1,25(OH)2D3 was found to stimulate choline acetyl transferase activity in the bed nucleus of the stria terminalis [483]. Furthermore, nerve growth factor (NGF) mRNA levels were stimulated by 1,25(OH)2D3 in mouse L929 fibroblasts, an in vitro model of nerve cell function [484,485], and studies demonstrated that 1,25(OH)2D3 induced NGF mRNA levels in the hippocampus and cortex [486]. In the intact organism, 1,25(OH)2D3 treatment results in improved memory performance of young adult rats in the Morris water maze test [487]. Interestingly, VDR mRNA expression was found to be decreased in the hippocampus of patients with Alzheimer’s disease [481]. A possible role of decreased 1,25(OH)2D or VDR with aging, leading to decreased NGF production in the brain, has raised conjecture about a possible role of decreased vitamin D action in the neurodegeneration found with aging or Alzheimer’s disease [482]. VDR levels have been shown to decrease with aging in the intestine [416], and although a connection to the brain is highly speculative at this time, some role for 1,25(OH)2D in the central nervous system seems clear. Alzheimer’s disease (AD) patients are susceptible to hypovitaminosis D due to their being elderly and confined to a hospital. A study of 46 ambulatory elderly women with AD showed that 26% had decreased 25(OH)D (5 – 10 ng/ml) and 54% had osteomalacic levels ( 5 ng/ml) [488]. Those with decreased vitamin D had increased PTH and decreased BMD. Many AD patients were sunlight deprived and consumed less than 100 IU of vitamin D per day. Vitamin D deficiency due to sunlight deprivation and malnutrition, together with compensatory hyperparathyroidism, contributes significantly to reduced BMD and increased risk of hip fractures in patients with AD [488].
F. 1,25(OH)2D Action on the Reproductive System The role of 1,25(OH)2D in reproduction has been examined in chickens and rats where 1,25(OH)2D appears to play a role in normal ovulation, fetal and neonatal bone development, milk production, and maintenance of normocalcemia and mineral homeostasis in the neonate [489]. Extrarenal synthesis of 1,25(OH)2D takes place in the placenta, which also expresses VDR. In addition, 1,25(OH)2D stimulates human placental lactogen (hPL) expression from
288 trophoblast cells, and a VDRE has been demonstrated in the 5 upstream region of the hPL gene, supporting a role for 1,25(OH)D in placental function [490]. VDR and vitamin D-dependent Ca2-binding protein are found in a number of additional tissues, including testis, uterus, pancreas, pituitary, thyroid, gonads, and muscle, including the heart [152,438,491], but the functional role of 1,25(OH)2D in these sites is unclear and will require further investigation. In VDR-ablated mice, uterine hypoplasia and ovarian abnormalities were detected in females and testicular defects and sperm abnormalities in males [331]. However, many of these defects were improved after calcium nutrition was normalized [334]. However, in females the estradiol levels were still somewhat reduced and gonadotropin levels somewhat elevated, suggesting a residual defect unrelated to calcium. These parameters normalized with administered estradiol. Because the aromatase gene is regulated by vitamin D [335], an effect on estradiol synthesis may affect fertility in VDR-ablated mice and HVDRR subjects [334].
XI. VITAMIN D AND OSTEOPOROSIS The importance of vitamin D in the etiology and treatment of osteoporosis will be discussed in detail in a number of subsequent chapters in this book, especially Chapters 40 and 68. The use of vitamin D and its analogues to prevent and treat ostoporosis has been reviewed [366,492,493]. In brief, several potential mechanisms have been put forward to implicate vitamin D in the development of osteoporosis. (i) The possibility that polymorphisms within the gene encoding the VDR contribute substantially to genetic differences in osteoporosis risk has been raised by Morrison et al. [342]. The basis for this genetic effect on osteoporosis risk is presumably as a hereditary factor affecting “peak bone mass,” but the mechanism is unknown. At this time the VDR genotype hypothesis remains controversial, as other groups, using subjects of different ethnic background, have not found a similar correlation between the different VDR alleles and bone density [344,345] (see Chapter 26). (ii) An age-related decline in renal 1,25(OH)2D production, due in part to a diminished renal response to PTH and reduced intestinal calcium absorption [414]. There appears to be a defect in the renal response to PTH so that older women with osteoporosis require greater amounts of PTH to stimulate 1,25(OH)2D production. (iii) A relative decrease in circulating 1,25(OH)2D has been considered a contributing factor in the development of senile osteoporosis [494]. A low vitamin D state from an inadequate diet and decreased exposure to sunlight as people age, especially in the house-bound elderly, contribute to malabsorption of calcium and vitamin D “insufficiency” in the elderly [17]. Other studies concur that there is a high prevalence of vitamin D insufficiency in the elderly, even in
FELDMAN, MALLOY, AND GROSS
an active community. These individuals may have established vertebral osteoporosis with increased bone turnover, decreased BMD at the hip, and thus an enhanced risk of further osteoporotic fractures in comparison with vitamin D-sufficient subjects [495]. (iv) An age-related decline in intestinal VDR creating a relative 1,25(OH)2D-resistant state and impairing intestinal calcium absorption [416]. All of these factors coordinately contribute to age-related bone loss, which, according to some studies, can be ameliorated by vitamin D and calcium supplements, [496 – 498]. It is interesting to note that there is a rapid recovery of BMD following the resolution of vitamin D insufficiency [499]. In a paper discussing vitamin D and osteoporosis, Lau and Baylink [500] argue that the type of vitamin D deficiency that exists determines the response to vitamin D or 1,25(OH)2D3 and its analogues. Vitamin D deficiency may be (1) primary vitamin D deficiency, which is due to a deficiency of vitamin D, the parent compound; (2) a deficiency of 1,25(OH)2D3 resulting from decreased renal production; and (3) resistance to 1,25(OH)2D3 action at the target tissues, which could be related to decreased VDR levels in the intestine with age. Each type of deficiency has been implicated as a potential cause of intestinal calcium malabsorption, secondary hyperparathyroidism, and senile osteoporosis. Primary vitamin D deficiency can be corrected by vitamin supplements, whereas 1,25(OH)2D3 deficiency or resistance may require 1,25(OH)2D3 or an analogue to correct the high serum PTH and the calcium malabsorption. In addition, some elderly patients have decreased intestinal Ca absorption, but with apparently normal vitamin D metabolism. Although the cause is unclear, these patients, as well as other patients with secondary hyperparathyroidism (not due to decreased renal function), show a decrease in serum PTH and an increase in calcium absorption in response to therapy with 1,25(OH)2D3 or an analogue. Some form of vitamin D therapy—vitamin D, 1,25(OH)2D3, or an analogue—can be used to correct all types of age-dependent impairments in intestinal calcium absorption and secondary hyperparathyroidism during aging. With respect to postmenopausal osteoporosis, there is strong evidence that 1,25(OH)2D3 or its analogues may have bone-sparing actions. However, these effects appear to be the result of pharmacologic actions on bone formation and resorption rather than through replenishing a deficiency [500]. Vitamin D has direct actions to affect estrogen synthesis by regulating the activity of aromatase in osteoblasts [335] and estrogen half-life by regulaing 17-hydroxysteroid dehydrogenase in keratinocytes [501]. The impact of these effects on multiple organs and their potential role in modulating vitamin D actions remain to be fully clarified. Vitamin D has effects on muscle, and findings suggest that vitamin D insufficiency may be associated with decreased muscle strength [502] and therefore increased rates of falling [503]. In ambulatory nursing home and hostel
289
CHAPTER 9 Vitamin D
residents, individuals who fall have lower serum 25hydroxyvitamin D and higher serum parathyroid hormone concentrations than other residents. The association between falling and serum PTH persists after adjustment for other variables [503]. An increased rate of falling can contribute to increased fractures in vitamin D-insufficient individuals. In a special case of glucocorticoid-induced osteoporosis, vitamin D plus calcium is superior to no therapy or calcium alone in the management of patients according to a recent meta-analysis. However, vitamin D is less effective than some osteoporosis therapies, such as bisphosphonates (see Chapter 44). Therefore, treatment with vitamin D plus calcium, as a minimum, should be recommended to patients receiving long-term corticosteroids [504]. Although most studies address the common problem of insufficient vitamin D, there are also difficulties with overdoses of vitamin D. Hypervitaminosis D may result from drinking milk that is incorrectly fortified with vitamin D [505] or from excessive dietary supplements [506]. After resolution of occult vitamin D intoxication in patients using dietary supplements that contained unadvertised high levels of vitamin D, resolution of vitamin D intoxication was associated with a rebound in bone mineral density [506]. The relationship of vitamin D to osteoporosis is discussed completely in Chapter 68.
References 1. S. Christakos, M. Raval-Pandya, R. P. Wernyj, and W. Yang, Genomic mechanisms involved in the pleiotropic actions of 1,25-dihydroxyvitamin D3. Biochem. J. 316, 361 – 371 (1996). 2. M. R. Haussler, G. K. Whitfield, C. A. Haussler, J. Hsieh, P. D. Thompson, S. H. Selznick, C. E. Dominguez, and P. W. Jurutka, The nuclear vitamin D receptor: Biological and molecular regulatory properties revealed. J. Bone Min. Res. 13, 325 – 349 (1998). 3. L. L. Issa, G. M. Leong, and J. A. Eisman, Molecular mechanism of vitamin D receptor action. Inflamm. Res. 47, 451 – 475. 4. G. Jones, S. A. Strugnell, and H. F. DeLuca, Current understanding of the molecular actions of vitamin D. Physiol. Rev. 78, 1193 – 1231 (1998). 5. A. W. Norman, Receptors for 1,25(OH)2D3: Past, present, and future. J. Bone Miner. Res. 13, 1360 – 1369 (1998). 6. A. J. Brown, A. Dusso, and E. Slatopolsky, Vitamin D. Am. J. Physiol. 277, F157 – F175 (1999). 7. P. J. Malloy, J. W. Pike, and D. Feldman, The vitamin D receptor and the syndrome of hereditary 1,25-dihydroxyvitamin D-resistant rickets. Endocr. Rev. 20, 156 – 188 (1999). 8. N. J. McKenna, R. B. Lanz, and B. W. O’Malley, Nuclear receptor coregulators: Cellular and molecular biology. Endocr. Rev. 20, 321 – 344 (1999). 9. C. Rachez and L. P. Freedman, Mechanisms of gene regulation by vitamin D3 receptor: A network of coactivator interactions. Gene 246, 9 – 21 (2000). 10. D. Feldman, F. H. Glorieux, and J. W. Pike, (eds.) (1997). “Vitamin D.” Academic Press, San Diego, 1997. 11. R. L. Horst and T. A. Reinhardt, Vitamin D metabolism. In “Vitamin D” (D. Feldman, F. Glorieus, and J. W. Pike, eds.), pp. 13 – 31. Academic Press, San Diego 1997.
12. M. F. Holick, McCollum Award Lecture, 1994: Vitamin D — new horizons for the 21st century. Am. J. Clin. Nutr. 60, 619 – 630 (1994). 13. E. Kodicek, The story of vitamin D, from vitamin to hormone. Lancet 1, 325 – 329 (1974). 14. H. F. DeLuca, The vitamin D story: A collaborative effort of basic science and clinical medicine. FASEB J. 2, 224 – 236 (1988). 15. H. F. DeLuca, Historical overview. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 3 – 11. Academic Press, San Diego, 1997. 16. M. F. Holick, Q. Shao, W. W. Liu, and T. C. Chen, The vitamin D content of fortified milk and infant formula. N. Engl. J. Med. 326, 1178 – 1181 (1992). 17. M. C. Chapuy, P. Preziosi, M. Maamer, S. Arnaud, P. Galan, S. Hercberg, and P. J. Meunier, Prevalence of vitamin D insufficiency in an adult normal population. Osteoporos. Int. 7, 439 – 443 (1997). 18. M. F. Holick, Vitamin D requirements for humans of all ages: New increased requirements for women and men 50 years and older. Osteoporos. Int. 8 (Suppl. 2), S24 – S29 (1998). 19. H. Glerup, K. Mikkelsen, L. Poulsen, E. Hass, S. Overbeck, J. Thomsen, P. Charles, and E. F. Eriksen, Commonly recommended daily intake of vitamin D is not sufficient if sunlight exposure is limited. J. Intern. Med. 247, 260 – 268 (2000). 20. M. C. Chapuy, A. M. Schott, P. Garnero, D. Hans, P. D. Delmas, and P. J. Meunier, Healthy elderly French women living at home have secondary hyperparathyroidism and high bone turnover in winter: EPIDOS Study Group. J. Clin. Endocrinol. Metab. 81, 1129 – 1133 (1996). 21. F. M. Gloth, III, C. M. Gundberg, B. W. Hollis, J. G. Haddad, Jr., and J. D. Tobin, Vitamin D deficiency in homebound elderly persons. JAMA 274, 1683 – 1686 (1995). 22. M. C. Chapuy, M. E. Arlot, P. D. Delmas, and P. J. Meunier, Effect of calcium and cholecalciferol treatment for three years on hip fractures in elderly women. Br. Med. J. 308, 1081 – 1082 (1994). 23. R. A. Bouillon, J. H. Auwerx, W. D. Lissens, and W. K. Pelemans, Vitamin D status in the elderly: Seasonal substrate deficiency causes 1,25-dihydroxycholecalciferol deficiency. Am. J. Clin. Nutr. 45, 755 – 763 (1987). 24. B. Dawson-Hughes, G. E. Dallal, E. A. Krall, S. Harris, L. J. Sokoll, and G. Falconer, Effect of vitamin D supplementation on wintertime and overall bone loss in healthy postmenopausal women. Ann. Intern. Med. 115, 505 – 512 (1991). 25. M. J. McKenna, Differences in vitamin D status between countries in young adults and the elderly. Am. J. Med. 93, 69 – 77 (1992). 26. M. K. Thomas, D. M. Lloyd-Jones, R. I. Thadhani, A. C. Shaw, D. J. Deraska, B. T. Kitch, E. C. Vamvakas, I. M. Dick, R. L. Prince, and J. S. Finkelstein, Hypovitaminosis D in medical inpatients. N. Engl. J. Med. 338, 777 – 783 (1998). 27. M. S. LeBoff, L. Kohlmeier, S. Hurwitz, J. Franklin, J. Wright, and J. Glowacki, Occult vitamin D deficiency in postmenopausal US women with acute hip fracture. JAMA 281, 1505 – 1511 (1999). 28. M. C. Chapuy and P. J. Meunier, Vitamin D insufficiency in adults and the elderly. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 679 – 693. Academic Press, San Diego, 1997. 29. R. H. Wasserman, Vitamin D and the intestinal absorption of calcium and phosphorus. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 259 – 273. Academic Press, San Diego, 1997. 30. E. B. Mawer and M. Davies, Bone disorders associated with gastrointestinal and hepatobiliary disease. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 831 – 847. Academic Press, San Diego, 1997. 31. C. W. Lo, P. W. Paris, T. L. Clemens, J. Nolan, and M. F. Holick, Vitamin D absorption in healthy subjects and in patients with intestinal malabsorption syndromes. Am. J. Clin. Nutr. 42, 644 – 649 (1985). 32. T. D. Thacher, P. R. Fischer, J. M. Pettifor, J. O. Lawson, C. O. Isichei, J. C. Reading, and G. M. Chan, A comparison of calcium,
290
33.
34.
35.
36.
37.
38. 39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52. 53.
FELDMAN, MALLOY, AND GROSS vitamin D, or both for nutritional rickets in Nigerian children. N. Engl. J. Med. 341, 563 – 568 (1999). M. F. Holick, Vitamin D: Photobiology, metabolism, and clinical applications. In “Endocrinology” (L. J. DeGroot, M. Besser, H. G. Burger, D. L. Loriaux, J. C. Marshall, W. D. Odell, J. T. Potts, and A. H. Rubentstein, eds.), 3rd Ed., pp. 990 – 1014. Saunders, Philadelphia, 1995. M. F. Holick, Photobiology of vitamin D. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 33 – 39. Academic Press, San Diego, 1997. D. D. Bikle, Vitamin D and skin. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 379 – 394. Academic Press, San Diego, 1997. S. J. Courtois, S. Segaert, H. Degreef, R. Bouillon, and M. Garmyn, Ultraviolet B suppresses vitamin D receptor gene expression in keratinocytes. Biochem. Biophys. Res. Commun. 246, 64 – 69 (1998). L. Y. Matsuoka, J. Wortsman, T. C. Chen, and M. F. Holick, Compensation for the interracial variance in the cutaneous synthesis of vitamin D. J. Lab. Clin. Med. 126, 452 – 457 (1995). W. F. Loomis, Skin-pigment regulation of vitamin-D biosynthesis in man. Science 157, 501 – 506 (1967). L. Y. Matsuoka, J. Wortsman, J. G. Haddad, P. Kolm, and B. W. Hollis, Racial pigmentation and the cutaneous synthesis of vitamin D. Arch. Dermatol. 127, 536 – 538 (1991). N. H. Bell, A. Greene, S. Epstein, M. J. Oexmann, S. Shaw, and J. Shary, Evidence for alteration of the vitamin D-endocrine system in blacks. J. Clin. Invest. 76, 470 – 473 (1985). M. F. Holick, J. A. MacLaughlin, and S. H. Doppelt, Regulation of cutaneous previtamin D3 photosynthesis in man: Skin pigment is not an essential regulator. Science 211, 590 – 593 (1981). J. Hirschfeld, Immune-electrophoretic demonstration of qualitative differences in human sera and their relation to the haptoglobins. Acta Pathol. Microbiol. Scand. 47, 160 – 168 (1959). S. P. Daiger, M. S. Schanfield, and L. L. Cavalli-Sforza, Groupspecific component (Gc) proteins bind vitamin D and 25-hydroxyvitamin D. Proc. Natl. Acad. Sci. USA 72, 2076 – 2080 (1975). N. E. Cooke, J. F. McLeod, X. K. Wang, and K. Ray, Vitamin D binding protein: Genomic structure, functional domains, and mRNA expression in tissues. J. Steroid Biochem. Mol. Biol. 40, 787 – 793 (1991). N. E Cooke and J. G. Haddad, Vitamin D binding protein. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 87 – 101. Academic Press, San Diego 1997. A. Sabbatini, M. Petrini, L. Mattii, P. Arnaud, and R. M. Galbraith, Vitamin D binding protein is produced by human monocytes. FEBS Lett. 323, 89 – 92 (1993). H. Cleve and J. Constans, The mutants of the vitamin-D-binding protein: More than 120 variants of the GC/DBP system. Vox. Sang. 54, 215 – 225 (1988). L. L. Cavalli-Sforza, P. Menozzi, and A. Piazza, “The History and Geography of Human Genes.” Princeton Univ. Press, Princeton, NJ, 1994. J. G. Haddad, L. Y. Matsuoka, B. W. Hollis, Y. Z. Hu, and J. Wortsman, Human plasma transport of vitamin D after its endogenous synthesis. J. Clin. Invest. 91, 2552 – 2555, (1993). R. Bouillon, K. Allewaert, D. Z. Xiang, B. K. Tan, and H. van Baelen, Vitamin D analogs with low affinity for the vitamin D binding protein: Enhanced in vitro and decreased in vivo activity. J. Bone Miner. Res. 6, 1051 – 1057 (1991). J. P. Berg, Vitamin D-binding protein prevents vitamin D deficiency and presents vitamin D for its renal activation. Eur. J. Endocrinol. 141, 321 – 322 (1999). W. M. Lee and R. M. Galbraith, The extracellular actin-scavenger system and actin toxicity. N. Engl. J. Med. 326, 1335 – 1341 (1992). M. Guoth, A. Murgia, R. M. Smith, M. B. Prystowsky, N. E. Cooke, and J. G. Haddad, Cell surface vitamin D-binding protein
54.
55.
56.
57.
58.
59.
60.
61. 62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
(GC-globulin) is acquired from plasma. Endocrinology 127, 2313 – 2321 (1990). F. F. Safadi, P. Thornton, H. Magiera, B. W. Hollis, M. Gentile, J. G. Haddad, S. A. Liebhaber, and N. E. Cooke, Osteopathy and resistance to vitamin D toxicity in mice null for vitamin D binding protein. J. Clin. Invest. 103, 239 – 251 (1999). A. Nykjaer, D. Dragun, D. Walther, H. Vorum, C. Jacobsen, J. Herz, F. Melsen, E. I. Christensen, and T. E. Willnow, An endocytic pathway essential for renal uptake and activation of the steroid 25-(OH) vitamin D3 Cell 96, 507 – 515 (1999). J. R. Leheste, B. Rolinski, H. Vorum, J. Hilpert, A. Nykjaer, C. Jacobsen, P. Aucouturier, J. O. Moskaug, A. Otto, E. I. Christensen, and T. E. Willnow, Megalin knockout mice as an animal model of low molecular weight proteinuria. Am. J. Pathol. 155, 1361 – 1370 (1999). M. A. Gacad, H. Chen, J. E. Arbelle, T. LeBon, and J. S. Adams, Functional characterization and purification of an intracellular vitamin D-binding protein in vitamin D-resistant new world primate cells: Amino acid sequence homology with proteins in the hsp-70 family. J. Biol. Chem. 272, 8433 – 8440 (1997). M. A. Gacad and J. S. Adams, Proteins in the heat shock-70 family specifically bind 25-hydroxyvitamin D3 and 17-estradiol. J. Clin. Endocrinol. Metab. 83, 1264 – 1267 (1998). B. W. Hollis, Detection of vitamin D and its major metabolites. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 587 – 606. Academic Press, San Diego 1997. M. Gascon-Barre, The vitamin D 25-hydroxylase. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 41 – 55. Academic Press, San Diego 1997. K. I. Okuda, Liver mitochondrial P450 involved in cholesterol catabolism and vitamin D activation. J. Lipid Res. 35, 361 – 372 (1994). E. Usui, M. Noshiro, and K. Okuda, Molecular cloning of cDNA for vitamin D3 25-hydroxylase from rat liver mitochondria. FEBS Lett. 262, 135 – 138 (1990). J. J. Cali, C. L. Hsieh, U. Francke, and D. W. Russell, Mutations in the bile acid biosynthetic enzyme sterol 27-hydroxylase underlie cerebrotendinous xanthomatosis. J. Biol. Chem. 266, 7779 – 7783 (1991). Y. D. Guo, S. Strugnell, D. W. Back, and G. Jones, Transfected human liver cytochrome P-450 hydroxylates vitamin D analogs at different side-chain positions. Proc. Natl. Acad. Sci. USA 90, 8668 – 8672 (1993). I. Bjorkhem and S. Skrede, Familial diseases with storage of sterols other than cholesterol: Cerebrotendinous xanthomatosis and phytosterolemia. In “The Metabolic Basis of Inherited Disease.” (C. R. Scriver, A. L. Beaudet, W. S. Sly, and D. Valle, eds.), pp. 1283 – 1302. McGraw Hill, New York, 1999. V. M. Berginer, S. Shany, D. Alkalay, J. Berginer, S. Dekel, G. Salen, G. S. Tint, and D. Gazit, Osteoporosis and increased bone fractures in cerebrotendinous xanthomatosis. Metabolism 42, 69 – 74 (1993). M. R. Hughes, D. J. Baylink, P. G. Jones, and M. R. Haussler, Radioligand receptor assay for 25-hydroxyvitamin D2/D3 and 1, 25dihydroxyvitamin D2/D3. J. Clin. Invest. 58, 61 – 70 (1976). N. H. Bell, S. Shaw, and R. T. Turner, Evidence that 1,25-dihydroxyvitamin D3 inhibits the hepatic production of 25-hydroxyvitamin D in man. J. Clin. Invest. 74, 1540 – 1544 (1984). B. P. Halloran, D. D. Bikle, M. J. Levens, M. E. Castro, R. K. Globus, and E. Holton, Chronic 1,25-dihydroxyvitamin D3 administration in the rat reduces the serum concentration of 25-hydroxyvitamin D by increasing metabolic clearance rate. J. Clin. Invest. 78, 622 – 628 (1986). M. R. Clements, L. Johnson, and D. R. Fraser, A new mechanism for induced vitamin D deficiency in calcium deprivation. Nature 325, 62 – 65 (1987). H. L. Henry, The 25-hydroxyvitamin D 1-hydroxylase. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 57 – 68. Academic Press, San Diego 1997.
CHAPTER 9 Vitamin D 72. K. Takeyama, S. Kitanaka, T. Sato, M. Kobori, J. Yanagisawa, and S. Kato, 25-Hydroxyvitamin D3 1-hydroxylase and vitamin D synthesis. Science 277, 1827 – 1830 (1997). 73. R. St-Arnaud, S. Messerlian, J. M. Moir, J. L. Omdahl, and F. H. Glorieux, The 25-hydroxyvitamin D 1--hydroxylase gene maps to the pseudovitamin D-deficiency rickets (PDDR) disease locus. J. Bone Miner. Res. 12, 1552 – 1559 (1997). 74. T. Monkawa, T. Yoshida, S. Wakino, T. Shinki, H. Anazawa, H. F. Deluca, T. Suda, M. Hayashi, and T. Saruta, Molecular cloning of cDNA and genomic DNA for human 25-hydroxyvitamin D3 1hydroxylase. Biochem. Biophys. Res. Commun. 239, 527 – 533 (1997). 75. T. Shinki, H. Shimada, S. Wakino, H. Anazawa, M. Hayashi, T. Saruta, H. F. DeLuca, and T. Suda, Cloning and expression of rat 25-hydroxyvitamin D3-1-hydroxylase cDNA. Proc. Natl. Acad. Sci. USA 94, 12920 – 12925 (1997). 76. G. K. Fu, D. Lin, M. Y. Zhang, D. D. Bikle, C. H. Shackleton, W. L. Miller, and A. A. Portale, Cloning of human 25-hydroxyvitamin D-1 -hydroxylase and mutations causing vitamin D-dependent rickets type 1. Mol. Endocrinol. 11, 1961 – 1970 (1997). 77. S. Kitanaka, K. Takeyama, A. Murayama, T. Sato, K. Okumura, M. Nogami, Y. Hasegawa, H. Niimi, J. Yanagisawa, T. Tanaka, and S. Kato, Inactivating mutations in the 25-hydroxyvitamin D3 1hydroxylase gene in patients with pseudovitamin D-deficiency rickets. N. Engl. J. Med. 338, 653 – 661 (1998). 78. M. Labuda, K. Morgan, and F. H. Glorieux, Mapping autosomal recessive vitamin D dependency type I to chromosome 12q14 by linkage analysis. Am. J. Hum. Genet. 47, 28 – 36 (1990). 79. M. Labuda, T. M. Fujiwara, M. V. Ross, K. Morgan, J. Garcia-Heras, D. H. Ledbetter, M. R. Hughes, and F. H. Glorieux, Two hereditary defects related to vitamin D metabolism map to the same region of human chromosome 12q13-14. J. Bone Miner. Res. 7, 1447 – 1453 (1992). 80. E. S. Slatopolsky and A. J. Brown, Vitamin D and renal failure. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 849 – 865. Academic Press, San Diego 1997. 81. A. Murayama, K. Takeyama, S. Kitanaka, Y. Kodera, Y. Kawaguchi, T. Hosoya, and S. Kato, Positive and negative regulations of the renal 25-hydroxyvitamin D3 1-hydroxylase gene by parathyroid hormone, calcitonin, and 1,25(OH)2D3 in intact animals. Endocrinology 140, 2224 – 2231 (1999). 82. X. F. Kong, X. H. Zhu, Y. L. Pei, D. M. Jackson, and M. F. Holick, Molecular cloning, characterization, and promoter analysis of the human 25-hydroxyvitamin D3-1-hydroxylase gene. Proc. Natl. Acad. Sci. USA 96, 6988 – 6993 (1999). 83. T. Shinki, Y. Ueno, H. F. DeLuca, and T. Suda, Calcitonin is a major regulator for the expression of renal 25-hydroxyvitamin D3-1hydroxylase gene in normocalcemic rats. Proc. Natl. Acad. Sci. USA 96, 8253 – 8258 (1999). 84. S. J. Smith, A. K. Rucka, J. L. Berry, M. Davies, S. Mylchreest, C. R. Paterson, D. A. Heath, M. Tassabehji, A. P. Read, A. P. Mee, and E. B. Mawer, Novel mutations in the 1-hydroxylase (P450c1) gene in three families with pseudovitamin D-deficiency rickets resulting in loss of functional enzyme activity in blood-derived macrophages. J. Bone Miner. Res. 14, 730 – 739 (1999). 85. N. H. Bell, Vitamin D-endocrine system. J. Clin. Invest. 76, 1 – 6 (1985). 86. P. Beckerman and J. Silver, Vitamin D and the parathyroid. Am. J. Med. Sci. 317, 363 – 369 (1999). 87. J. Silver, C. Yalcindag, A. Sela-Brown, R. Kilav, and T. Naveh-Many, Regulation of the parathyroid hormone gene by vitamin D, calcium and phosphate. Kidney Int. Suppl. 73, S2 – S7 (1999). 88. N. Chattopadhyay, A. Mithal, and E. M. Brown, The calciumsensing receptor: A window into the physiology and pathophysiology of mineral ion metabolism. Endocr. Rev. 17, 289 (1996).
291 89. E. Slatopolsky, A. Dusso, and A. J. Brown, The role of phosphorus in the development of secondary hyperparathyroidism and parathyroid cell proliferation in chronic renal failure. Am. J. Med. Sci. 317, 370 – 376 (1999). 90. A. A. Portale, B. P. Halloran, M. M. Murphy, and R. Morris, Jr., Oral intake of phosphorus can determine the serum concentration of 1,25dihydroxyvitamin D by determining its production rate in humans. J. Clin. Invest. 77, 7 – 12 (1986). 91. B. P. Halloran and M. E. Castro, Vitamin D kinetics in vivo: Effect of 1,25-dihydroxyvitamin D administration. Am. J. Physiol. 256, E686 – E691 (1989). 92. N. A. Breslau, Normal and abnormal regulation of 1,25-(OH)2D synthesis. Am. J. Med. Sci. 296, 417 – 425 (1988). 93. J. Wortsman, J. G. Haddad, J. T. Posillico, and E. M. Brown, Primary hyperparathyroidism with low serum 1,25-dihydroxyvitamin D levels. J. Clin. Endocrinol. Metab. 62, 1305 – 1308 (1986). 94. E. M. Brown, G. Gamba, D. Riccardi, M. Lombardi, R. Butters, O. Kifor, A. Sun, M. Hediger, J. Lytton, and S. C. Hebert, Cloning, expression, and characterization of an extracellular Ca2 sensing receptor from bovine parathyroid. Nature 366, 575 – 580 (1993). 95. D. Riccardi, J. Park, W. Lee, G. Gamba, E. M. Brown, and S. Hebert, Cloning and functional expression of a rat kidney extracellular calcium/polyvalent cation-sensing receptor. Proc. Natl. Acad. Sci. USA 92, 131 – 135 (1995). 96. P. Jaeger, W. Jones, T. L. Clemens, and J. P. Hayslett, Evidence that calcitonin stimulates 1,25-dihydroxyvitamin D production and intestinal absorption of calcium in vivo. J. Clin. Invest. 78, 456 – 461 (1986). 97. M. J. Econs, B. Lobaugh, and M. K. Drezner, Normal calcitonin stimulation of serum calcitriol in patients with X-linked hypophosphatemic rickets. J. Clin. Endocrinol. Metab. 75, 408 – 411 (1992). 98. T. Nesbitt, B. Lobaugh, and M. K. Drezner, Calcitonin stimulation of renal 25-hydroxyvitamin D-1-hydroxylase activity in hypophosphatemic mice: Evidence that the regulation of calcitriol production is not universally abnormal in X-linked hypophosphatemia. J. Clin. Invest. 79, 15 – 19 (1987). 99. N. H. Bell, Renal and nonrenal 25-hydroxyvitamin D-1hydroxylases and their clinical significance. J. Bone Miner. Res. 13, 350 – 353 (1998). 100. G. A. Howard, R. T. Turner, D. J. Sherrard, and D. J. Baylink, Human bone cells in culture metabolize 25-hydroxyvitamin D3 to 1,25-dihydroxyvitamin D3 and 24,25-dihydroxyvitamin D3. J. Biol. Chem. 256, 7738 – 7740 (1981). 101. J. S. Adams, O. P. Sharma, M. A. Gacad, and F. R. Singer, Metabolism of 25-hydroxyvitamin D3 by cultured pulmonary alveolar macrophages in sarcoidosis. J. Clin. Invest. 72, 1856 – 1860 (1983). 102. H. S. Cross, M. Peterlik, G. S. Reddy, and I. Schuster, Vitamin D metabolism in human colon adenocarcinoma-derived Caco-2 cells: Expression of 25-hydroxyvitamin D3-1-hydroxylase activity and regulation of side-chain metabolism. J. Steroid Biochem. Mol. Biol. 62, 21 – 28 (1997). 103. Y. Weisman, A. Harell, S. Edelstein, M. David, Z. Spirer, and A. Golander, 1,25-dihydroxyvitamin D3 and 24,25-dihydroxyvitamin D3 in vitro synthesis by human decidua and placenta. Nature 281, 317 – 319 (1979). 104. G. G. Schwartz, L. W. Whitlatch, T. C. Chen, B. L. Lokeshwar, and M. F. Holick, Human prostate cells synthesize 1,25-dihydroxyvitamin D3 from 25-hydroxyvitamin D3. Cancer Epidemiol. Biomarkers Prev. 7, 391 – 395 (1998). 105. M. Davies, M. E. Hayes, J. A. Yin, J. L. Berry, and E. B. Mawer, Abnormal synthesis of 1,25-dihydroxyvitamin D in patients with malignant lymphoma. J. Clin. Endocrinol. Metab. 78, 1202 – 1207 (1994). 106. N. H. Bell, J. Shary, S. Shaw, and R. T. Turner, Hypercalcemia associated with increased circulating 1,25 dihydroxyvitamin D in a patient with pulmonary tuberculosis. Calcif. Tissue Int. 37, 588 – 591 (1985).
292 107. J. S. Adams, F. R. Singer, M. A. Gacad, O. P. Sharma, M. J. Hayes, P. Vouros, and M. F. Holick, Isolation and structural identification of 1,25-dihydroxyvitamin D3 produced by cultured alveolar macrophages in sarcoidosis. J. Clin. Endocrinol. Metab. 60, 960 – 966 (1985). 108. J. S. Adams and S. Y. Ren, Autoregulation of 1,25-dihydroxyvitamin D synthesis in macrophage mitochondria by nitric oxide. Endocrinology 137, 4514 – 4517 (1996). 109. J. Omdahl and B. May, The 25-hydroxyvitamin D 24-hydroxylase. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 69 – 85. Academic Press, San Diego, 1997. 110. R. St-Arnaud, Targeted inactivation of vitamin D hydroxylases in mice. Bone 25, 127 – 129 (1999). 111. J. P. Goff, T. A. Reinhardt, G. W. Engstrom, and R. L. Horst, Effect of dietary calcium or phosphorus restriction and 1,25-dihydroxyvitamin D administration on rat intestinal 24-hydroxylase. Endocrinology 131, 101 – 104 (1992). 112. R. Kumar, J. Schaefer, J. P. Grande, and P. C. Roche, Immunolocalization of calcitriol receptor, 24-hydroxylase cytochrome P-450, and calbindin D28k in human kidney. Am. J. Physiol. 266, F477 – F485 (1994). 113. Y. Ohyama, M. Noshiro, and K. Okuda, Cloning and expression of cDNA encoding 25-hydroxyvitamin D3 24-hydroxylase. FEBS Lett. 278, 195 – 198 (1991). 114. C. N. Hahn, E. Baker, P. Laslo, B. K. May, J. L. Omdahl, and G. R. Sutherland, Localization of the human vitamin D 24-hydroxylase gene (CYP24) to chromosome 20q13.2 : q13.3. Cytogenet. Cell. Genet. 62, 192 – 193 (1993). 115. K. S. Chen and H. F. DeLuca, Cloning of the human 1 alpha,25-dihydroxyvitamin D3 24-hydroxylase gene promoter and identification of two vitamin D-responsive elements. Biochem. Biophys. Acta. 1263, 1 – 9 (1995). 116. Y. Ohyama, K. Ozono, M. Uchida, T. Shinki, S. Kato, T. Suda, O. Yamamoto, M. Noshiro, and Y. Kato, Identification of a vitamin D-responsive element in the 5-flanking region of the rat 25-hydroxyvitamin D3 24-hydroxylase gene. J. Biol. Chem. 269, 10545 –10550 (1994). 117. C. Zierold, H. M. Darwish, and H. F. DeLuca, Identification of a vitamin D-response element in the rat calcidiol (25-hydroxyvitamin D3) 24-hydroxylase gene. Proc. Natl. Acad. Sci. USA 91, 900 – 902 (1994). 118. P. J. Malloy, J. W. Pike, and D. Feldman, Hereditary 1,25-dihydroxyvitamin D resistant rickets. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 756 – 788. Academic Press, San Diego 1997. 119. T. Shigematsu, N. Horiuchi, Y. Ogura, T. Miyahara, and T. Suda, Human parathyroid hormone inhibits renal 24-hydroxylase activity of 25-hydroxyvitamin D3 by a mechanism involving adenosine 3,5monophosphate in rats. Endocrinology 118, 1583 – 1589 (1986). 120. A. Nishimura, T. Shinki, C. H. Jin, Y. Ohyama, M. Noshiro, K. Okuda, and T. Suda, Regulation of messenger ribonucleic acid expression of 1,25-dihydroxyvitamin D3-24-hydroxylase in rat osteoblasts. Endocrinology 134, 1794 – 1799 (1994). 121. M. J. Beckman, J. P. Goff, T. A. Reinhardt, D. C. Beitz, and R. L. Horst, In vivo regulation of rat intestinal 24-hydroxylase: Potential new role of calcitonin. J. Clin. Endocrinol. Metab. 135, 1951 – 19551 (1994). 122. R. St-Arnaud and F. H. Glorieux, 24,25-Dihydroxyvitamin D: Active metabolite or inactive catabolite? Endocrinology 139, 3371 – 3374 (1998). 123. R. Brommage and H. F. DeLuca, Evidence that 1,25-dihydroxyvitamin D3 is the physiologically active metabolite of vitamin D3. Endocr. Rev. 6, 491 – 511 (1985). 124. M. Uchida, K. Ozono, and J. W. Pike, Activation of the human osteocalcin gene by 24R,25-dihydroxyvitamin D3 occurs through the vitamin D receptor and the vitamin D-responsive element. J. Bone Miner. Res. 9, 1981 – 1987 (1994).
FELDMAN, MALLOY, AND GROSS 125. D. Somjen, G. J. Somjen, Y. Weisman, and I. Binderman, Evidence for 24,25-dihydroxycholecalciferol receptors in long bones of newborn rats. Biochem. J. 204, 31 – 36 (1982). 126. H. A. Pedrozo, Z. Schwartz, S. Rimes, V. L. Sylvia, I. Nemere, G. H. Posner, D. D. Dean, and B. D. Boyan, Physiological importance of the 1,25(OH)2D3. J. Bone Miner. Res. 14, 856 – 867 (1999). 127. A. W. Norman, J. Roth, and L. Orci, The vitamin D endocrine system: Steroid metabolism, hormone receptors, and biological response (calcium binding proteins). Endocr. Rev. 3, 331 – 366 (1982). 128. T. Yamate, T. Hiroyuki, Y. Nagai, H. Yamato, N. Taniguchi, T. Nakamura, and Y. Seino, Bone-forming ability of 24R,25-dihydroxyvitamin D3 in the hypophosphatemic mouse. J. Bone Miner. Res. 9, 1967 – 1974 (1994). 129. R. St. Arnaud, A. Arabian, R. Travers, F. Barletta, M. Raval-Pandya, K. Chapin, J. Depovere, C. Mathieu, S. Christakos, M. B. Demay, and F. H. Glorieux, Deficient mineralization of intra membranous bone in vitamin D-24-hydroxylase ablated mice is due to elevated 1,25-dihydroxy vitamin D and not to the absence of 24,25-dihydroxy vitamin D. Endocrinology 141, 2658–2666 (2000). 130. B. Endres, S. Kato, and H. F. DeLuca, Metabolism of 1,25dihydroxyvitamin D3 in vitamin D receptor-ablated mice in vivo. Biochemistry 39, 2123 – 2129 (2000). 131. M. F. Holick, A. Kleiner-Bossaller, H. K. Schnoes, P. M. Kasten, I. T. Boyle, and H. F. DeLuca, 1,24,25-Trihydroxyvitamin D3: A metabolite of vitamin D3 effective on intestine. J. Biol. Chem. 248, 6691 – 6696 (1973). 132. L. Castillo, Y. Tanaka, H. F. DeLuca, and N. Ikekawa, On the physiological role of 1,24,25-trihydroxyvitamin D3. Miner. Electrolyte Metab. 1, 198 – 207 (1978). 133. G. Jones, Analog metabolism. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 973 – 994. Academic Press, San Diego 1997. 134. M. G. Bischof, M. L. Siu-Caldera, A. Weiskopf, P. Vouros, H. S. Cross, M. Peterlik, and G. S. Reddy, Differentiation-related pathways of 1,25-dihydroxycholecalciferol metabolism in human colon adenocarcinoma-derived Caco-2 cells: Production of 1,25dihydroxy-3epi-cholecalciferol. Exp. Cell. Res. 241, 194 – 201 (1998). 135. M. L. Siu-Caldera, H. Sekimoto, A. Weiskopf, P. Vouros, K. R. Muralidharan, W. H. Okamura, J. Bishop, A. W. Norman, M. R. Uskokovic, I. Schuster, and G. S. Reddy, Production of 1,25dihydroxy-3-epi-vitamin D3 in two rat osteosarcoma cell lines (UMR 106 and ROS 17/2.8): Existence of the C-3 epimerization pathway in ROS 17/2.8 cells in which the C-24 oxidation pathway is not expressed. Bone 24, 457 – 463 (1999). 136. H. Sekimoto, M. L. Siu-Caldera, A. Weiskopf, P. Vouros, K. R. Muralidharan, W. H. Okamura, M. R. Uskokovic, and G. S. Reddy, 1,25-dihydroxy-3-epi-vitamin D3: In vivo metabolite of 1,25dihydroxyvitamin D3 in rats. FEBS Lett. 448, 278 – 282 (1999). 137. A. J. Brown, C. Ritter, E. Slatopolsky, K. R. Muralidharan, W. H. Okamura, and G. S. Reddy, 1,25-dihydroxy-3-epi-vitamin D3, a natural metabolite of 1,25-dihydroxyvitamin D3, is a potent suppressor of parathyroid hormone secretion. J. Cell. Biochem. 73, 106 – 113 (1999). 138. M. R. Haussler, C. A. Haussler, P. W. Jurutka, P. D. Thompson, J. C. Hsieh, L. S. Remus, S. H. Selznick, and G. K. Whitfield, The vitamin D hormone and its nuclear receptor: Molecular actions and disease states. J. Endocrinol. 154(Suppl.), S57 – S73 (1997). 139. G. K. Whitfield, P. W. Jurutka, C. A. Haussler, and M. R. Haussler, Steroid hormone receptors: Evolution, ligands, and molecular basis of biologic function. J. Cell. Biochem. 32 – 33(Suppl.), 110 – 122 (1999). 140. M. R. Haussler and A. W. Norman, Chromosomal receptor for a vitamin D metabolite. Proc. Natl. Acad. Sci. USA 62, 155 – 162 (1969).
CHAPTER 9 Vitamin D 141. D. Feldman, T. A. McCain, M. A. Hirst, T. L. Chen, and K. W. Colston, Characterization of a cytoplasmic receptor-like binder for 1 alpha, 25-dihydroxycholecalciferol in rat intestinal mucosa. J. Biol. Chem. 254, 10378 – 10384 (1979). 142. E. A. Allegretto and J. W. Pike, Trypsin cleavage of chick 1,25-dihydroxyvitamin D3 receptors: Generation of discrete polypeptides which retain hormone but are unreactive to DNA and monoclonal antibody. J. Biol. Chem. 260, 10139 – 10145 (1985). 143. J. W. Pike, Evidence for a reactive sulfhydryl in the DNA binding domain of the 1,25-dihydroxyvitamin D3 receptor. Biochem. Biophys. Res. Commun. 100, 1713 – 1719 (1981). 144. J. W. Pike and M. R. Haussler, Purification of chicken intestinal receptor for 1,25-dihydroxyvitamin D. Proc. Natl. Acad. Sci. USA 76, 5785 – 5789 (1979). 145. R. U. Simpson, A. Hamstra, N. C. Kendrick, and H. F. DeLuca, Purification of the receptor for 1,25-dihydroxyvitamin D3 from chicken intestine. Biochemistry 22, 2586 – 2594 (1983). 146. J. W. Pike, Monoclonal antibodies to chick intestinal receptors for 1,25-dihydroxyvitamin D3: Interaction and effects of binding on receptor function. J. Biol. Chem. 259, 1167 – 1173 (1984). 147. M. C. Dame, E. A. Pierce, J. M. Prahl, C. E. Hayes, and H. F. DeLuca, Monoclonal antibodies to the porcine intestinal receptor for 1,25-dihydroxyvitamin D3: Interaction with distinct receptor domains. Biochemistry 25, 4523 – 4534 (1986). 148. W. E. Stumpf, M. Sar, F. A. Reid, Y. Tanaka, and H. F. DeLuca, Target cells for 1,25-dihydroxyvitamin D3 in intestinal tract, stomach, kidney, skin, pituitary, and parathyroid. Science 206, 1188 – 1190 (1979). 149. K. Colston, M. Hirst, and D. Feldman, Organ distribution of the cytoplasmic 1,25-dihydroxycholecalciferol receptor in various mouse tissues. Endocrinology 107, 1916 – 1922 (1980). 150. T. L. Clemens, K. P. Garrett, X. Y. Zhou, J. W. Pike, M. R. Haussler, and D. W. Dempster, Immunocytochemical localization of the 1,25dihydroxyvitamin D3 receptor in target cells. Endocrinology 122, 1224 – 1230 (1988). 151. U. Berger, P. Wilson, R. A. McClelland, K. Colston, M. R. Haussler, J. W. Pike, and R. C. Coombes, Immunocytochemical detection of 1,25-dihydroxyvitamin D receptors in normal human tissues. J. Clin. Endocrinol. Metab. 67, 607 – 613 (1988). 152. M. R. Walters, Newly identified actions of the vitamin D endocrine system. Endocr. Rev. 13, 719 – 764 (1992). 153. J. W. Pike and N. M. Sleator, Hormone-dependent phosphorylation of the 1,25-dihydroxyvitamin D3 receptor in mouse fibroblasts. Biochem. Biophys. Res. Commun. 131, 378 – 385 (1985). 154. T. A. Brown and H. F. DeLuca, Phosphorylation of the 1,25dihydroxyvitamin D3 receptor: A primary event in 1,25-dihydroxyvitamin D3 action. J. Biol. Chem. 265, 10025 – 10029 (1990). 155. J. C. Hsieh, P. W. Jurutka, M. A. Galligan, C. M. Terpening, C. A. Haussler, D. S. Samuels, Y. Shimizu, N. Shimizu, and M. R. Haussler, Human vitamin D receptor is selectively phosphorylated by protein kinase C on serine 51, a residue crucial to its trans-activation function. Proc. Natl. Acad. Sci. USA 88, 9315 – 9319 (1991). 156. P. W. Jurutka, J. C. Hsieh, P. N. MacDonald, C. M. Terpening, C. A. Haussler, M. R. Haussler, and G. K. Whitfield, Phosphorylation of serine 208 in the human vitamin D receptor: The predominant amino acid phosphorylated by casein kinase II, in vitro, and identification as a significant phosphorylation site in intact cells. J. Biol. Chem. 268, 6791 – 6799 (1993). 157. G. M. Hilliard, R. G. Cook, N. L. Weigel, and J. W. Pike, 1,25-dihydroxyvitamin D3 modulates phosphorylation of serine 205 in the human vitamin D receptor: Site-directed mutagenesis of this residue promotes alternative phosphorylation. Biochemistry 33, 4300 – 4311 (1994). 158. P. W. Jurutka, J. C. Hsieh, and M. R. Haussler, Phosphorylation of the human 1,25-dihydroxyvitamin D3 receptor by cAMP-dependent protein kinase, in vitro, and in transfected COS-7 cells. Biochem. Biophys. Res. Commun. 191, 1089 – 1096 (1993).
293 159. P. W. Jurutka, J. C. Hsieh, S. Nakajima, C. A. Haussler, G. K. Whitfield, and M. R. Haussler, Human vitamin D receptor phosphorylation by casein kinase II at Ser-208 potentiates transcriptional activation. Proc. Natl. Acad. Sci. USA 93, 3519 – 3524 (1990). 160. J. C. Hsieh, P. W. Jurutka, S. Nakajima, M. A. Galligan, C. A. Haussler, Y. Shimizu, N. Shimizu, G. K. Whitfield, and M. R. Haussler, Phosphorylation of the human vitamin D receptor by protein kinase C: Biochemical and functional evaluation of the serine 51 recognition site. J. Biol. Chem. 268, 15118 – 15126 (1993). 161. D. P. McDonnell, D. J. Mangelsdorf, J. W. Pike, M. R. Haussler, and B. W. OMalley, Molecular cloning of complementary DNA encoding the avian receptor for vitamin D. Science 235, 1214 – 1217 (1987). 162. A. R. Baker, D. P. McDonnell, M. Hughes, T. M. Crisp, D. J. Mangelsdorf, M. R. Haussler, J. W. Pike, J. Shine, and B. W. OMalley, Cloning and expression of full-length cDNA encoding human vitamin D receptor. Proc. Natl. Acad. Sci. USA 85, 3294 – 3298 (1988). 163. D. J. Mangelsdorf, C. Thummel, M. Beato, P. Herrlich, G. Schutz, K. Umesono, B. Blumberg, P. Kastner, M. Mark, P. Chambon, et al. (1995). The nuclear receptor superfamily: The second decade. Cell 83, 835 – 839 (1995). 164. D. P. McDonnell, R. A. Scott, S. A. Kerner, B. W. O’Malley, and J. W. Pike, Functional domains of the human vitamin D3 receptor regulate osteocalcin gene expression. Mol. Endocrinol. 3, 635 – 644 (1989). 165. J. H. Faraco, N. A. Morrison, A. Baker, J. Shine, and P. M. Frossard, ApaI dimorphism at the human vitamin D receptor gene locus. Nucleic Acids Res. 17, 2150 (1989). 166. J. Szpirer, C. Szpirer, M. Riviere, G. Levan, P. Marynen, J. J. Cassiman, R. Wiese, and H. F. DeLuca, The Sp1 transcription factor gene (SP1) and the 1,25-dihydroxyvitamin D3 receptor gene (VDR) are colocalized on human chromosome arm 12q and rat chromosome 7. Genomics 11, 168 – 173 (1991). 167. K. Miyamoto, R. A. Kesterson, H. Yamamoto, Y. Taketani, E. Nishiwaki, S. Tatsumi, Y. Inoue, K. Morita, E. Takeda, and J. W. Pike, Structural organization of the human vitamin D receptor chromosomal gene and its promoter. Mol. Endocrinol. 11, 1165 – 1179 (1997). 168. L. A. Crofts, M. S. Hancock, N. A. Morrison, and J. A. Eisman, Multiple promoters direct the tissue-specific expression of novel N-terminal variant human vitamin D receptor gene transcripts. Proc. Natl. Acad. Sci. USA 95, 10529 – 10534 (1998). 169. T. Saijo, M. Ito, E. Takeda, A. H. Huq, E. Naito, I. Yokota, T. Sone, J. W. Pike, and Y. Kuroda, A unique mutation in the vitamin D receptor gene in three Japanese patients with vitamin D-dependent rickets type II: Utility of single-strand conformation polymorphism analysis for heterozygous carrier detection. Am. J. Hum. Genet. 49, 668 – 673 (1991). 170. H. Arai, K. I. Miyamoto, Y. Taketani, H. Yamamoto, Y. Iemori, K. Morita, T. Tonai, T. Nishisho, S. Mori, and E. Takeda, A vitamin D receptor gene polymorphism in the translation initiation codon: Effect on protein activity and relation to bone mineral density in Japanese women. J. Bone Miner. Res. 12, 915 – 921 (1997). 171. C. Gross, A. V. Krishnan, P. J. Malloy, T. R. Eccleshall, X. Y. Zhao, and D. Feldman, The vitamin D receptor gene start codon polymorphism: A functional analysis of FokI variants. J. Bone Miner. Res. 13, 1691 – 1699 (1998). 172. P. W. Jurutka, L. S. Remus, G. K. Whitfield, P. D. Thompson, J. C. Hsieh, H. Zitzer, P. Tavakkoli, M. A. Galligan, H. T. Dang, C. A. Haussler, and M. R. Haussler, The polymorphic N terminus in human vitamin D receptor isoforms influences transcriptional activity by modulating interaction with transcription factor IIB. Mol. Endocrinol. 14, 401 – 420 (2000). 173. A. W. Norman, D. Adams, E. D. Collins, W. H. Okamura, and R. J. Fletterick, Three-dimensional model of the ligand binding domain of the nuclear receptor for 1,25-dihydroxyvitamin D3. J. Cell. Biochem. 74, 323 – 333 (1999).
294 174. K. Yamamoto, H. Masuno, M. Choi, K. Nakashima, T. Taga, H. Ooizumi, K. Umesono, W. Sicinska, J. VanHooke, H. F. DeLuca, and S. Yamada, Three-dimensional modeling of and ligand docking to vitamin D receptor ligand binding domain. Proc. Natl. Acad. Sci. USA 97, 1467 – 1472 (2000). 175. N. Rochel, J. M. Wurtz, A. Mitschler, B. Klaholz, and D. Moras, The crystal structure of the nuclear receptor for vitamin D bound to its natural ligand. Mol. Cell. 5, 173 – 179 (2000). 176. K. Paech, P. Webb, G. G. Kuiper, S. Nilsson, J. Gustafsson, P. J. Kushner, and T. S. Scanlan, Differential ligand activation of receptors ER and ER at AP1 sites. Science 277, 1508 – 1510 (1997). 177. A. K. Shiau, D. Barstad, P. M. Loria, L. Cheng, P. J. Kushner, D. A. Agard, and G. L. Greene, The structural basis of estrogen receptor/coactivator recognition and the antagonism of this interaction by tamoxifen. Cell 95, 927 – 937 (1998). 178. K. Kristjansson, A. R. Rut, M. Hewison, J. L. ORiordan, and M. R. Hughes, Two mutations in the hormone binding domain of the vitamin D receptor cause tissue resistance to 1,25 dihydroxyvitamin D3. J. Clin. Invest. 92, 12 – 16 (1993). 179. P. J. Malloy, T. R. Eccleshall, C. Gross, L. Van Maldergem, R. Bouillon, and D. Feldman, Hereditary vitamin D resistant rickets caused by a novel mutation in the vitamin D receptor that results in decreased affinity for hormone and cellular hyporesponsiveness. J. Clin. Invest. 99, 297 – 304 (1997). 180. S. A. Kerner, R. A. Scott, and J. W. Pike, Sequence elements in the human osteocalcin gene confer basal activation and inducible response to hormonal vitamin D3. Proc Natl. Acad. Sci. USA 86, 4455 – 4459 (1989). 181. J. Lian, C. Stewart, E. Puchacz, S. Mackowiak, V. Shalhoub, D. Collart, G. Zambetti, and G. Stein, Structure of the rat osteocalcin gene and regulation of vitamin D-dependent expression. Proc. Natl. Acad. Sci. USA 86, 1143 – 1147 (1989). 182. N. A. Morrison, J. Shine, J. C. Fragonas, V. Verkest, M. L. Mc Menemy, and J. A. Eisman, 1,25-dihydroxyvitamin D-responsive element and glucocorticoid repression in the osteocalcin gene. Science 246, 1158 – 1161 (1989). 183. J. Liao, K. Ozono, T. Sone, D. P. McDonnell, and J. W. Pike, Vitamin D receptor interaction with specific DNA requires a nuclear protein and 1,25-dihydroxyvitamin D3. Proc. Natl. Acad. Sci. USA 87, 9751 – 9755 (1990). 184. T. Sone, S. Kerner, and J. W. Pike, Vitamin D receptor interaction with specific DNA. Association as a 1,25-dihydroxyvitamin D3modulated heterodimer. J. Biol. Chem. 266, 23296 – 23305 (1991). 185. T. Sone, K. Ozono, and J. W. Pike, A 55-kilodalton accessory factor facilitates vitamin D receptor DNA binding. Mol. Endocrinol. 5, 1578 – 1586 (1991). 186. P. N. MacDonald, D. R. Dowd, S. Nakajima, M. A. Galligan, M. C. Reeder, C. A. Haussler, K. Ozato, and M. R. Haussler, Retinoid X receptors stimulate and 9-cis retinoic acid inhibits 1,25-dihydroxyvitamin D3-activated expression of the rat osteocalcin gene. Mol. Cell. Biol. 13, 5907 – 5917 (1993). 187. D. J. Mangelsdorf, E. S. Ong, J. A. Dyck, and R. M. Evans, Nuclear receptor that identifies a novel retinoic acid response pathway. Nature 345, 224 – 229 (1990). 188. A. A. Levin, L. J. Sturzenbecker, S. Kazmer, T. Bosakowski, C. Huselton, G. Allenby, J. Speck, C. Kratzeisen, M. Rosenberger, A. Lovey et al., 9-cis retinoic acid stereoisomer binds and activated the nuclear receptor RXR alpha. Nature 355, 359 – 361 (1992). 189. R. A. Heyman, D. J. Mangelsdorf, J. A. Dyck, R. B. Stein, G. Eichele, R. M. Evans, and C. Thaller, 9-cis retinoic acid is a high affinity ligand for the retinoid X receptor. Cell 68, 397 – 406 (1992). 190. D. J. Mangelsdorf and R. M. Evans, The RXR heterodimers and orphan receptors. Cell 83, 841 – 850 (1995).
FELDMAN, MALLOY, AND GROSS 191. C. Solomon, J. H. White, and R. Kremer, Mitogen-activated protein kinase inhibits 1,25-dihydroxyvitamin D3-dependent signal transduction by phosphorylating human retinoid X receptor alpha. J. Clin. Invest. 103, 1729 – 1735 (1999). 192. D. Robyr, A. P. Wolffe, and W. Wahli, Nuclear hormone receptor coregulators in action: Diversity for shared tasks. Mol. Endocrinol. 14, 329 – 347 (2000). 193. T. E. Spencer, G. Jenster, M. M. Burcin, C. D. Allis, J. Zhou, C. A. Mizzen, N. J. McKenna, S. A. Onate, S. Y. Tsai, M. J. Tsai, and B. W. OMalley, Steroid receptor coactivator-1 is a histone acetyltransferase. Nature 389, 194 – 198 (1997). 194. R. K. Gill, L. M. Atkins, B. W. Hollis, and N. H. Bell, Mapping the domains of the interaction of the vitamin D receptor and steroid receptor coactivator-1. Mol. Endocrinol. 12, 57 – 65 (1998). 195. D. M. Kraichely, J. J. R. Collins, R. K. DeLisle, and P. N. Mac Donald, The autonomous transactivation domain in helix H3 of the vitamin D receptor is required for transactivation and coactivator interaction. J. Biol. Chem. 274, 14352 – 14358 (1999). 196. J. C. Blanco, I. M. Wang, S. Y. Tsai, M. J. Tsai, B. W. OMalley, P. W. Jurutka, M. R. Haussler, and K. Ozato, Transcription factor TFIIB and the vitamin D receptor cooperatively activate ligand-dependent transcription. Proc. Natl. Acad. Sci. USA 92, 1535 – 1539 (1995). 197. P. N. MacDonald, D. R. Sherman, D. R. Dowd, S. C. Jefcoat, Jr., and R. K. DeLisle, The vitamin D receptor interacts with general transcription factor IIB. J. Biol. Chem. 270, 4748 – 4752 (1995). 198. P. W. Jurutka, J. C. Hsieh, L. S. Remus, G. K. Whitfield, P. D. Thompson, C. A. Haussler, J. C. Blanco, K. Ozato, and M. R. Haussler, Mutations in the 1,25-dihydroxyvitamin D3 receptor identifying C-terminal amino acids required for transcriptional activation that are functionally dissociated from hormone binding, heterodimeric DNA binding, and interaction with basal transcription factor IIB, in vitro. J. Biol. Chem. 272, 14592 – 14599 (1997). 199. H. Masuyama, S. C. Jefcoat, Jr., and P. N. MacDonald, The N-terminal domain of transcription factor IIB is required for direct interaction with the vitamin D receptor and participates in vitamin D-mediated transcription. Mol. Endocrinol. 11, 218 – 228 (1997). 200. T. A. Baudino, D. M. Kraichely, S. C. Jefcoat, Jr., S. K. Winchester, N. C. Partridge, and P. N. MacDonald, Isolation and characterization of a novel coactivator protein, NCoA-62, involved in vitamin D-mediated transcription. J. Biol. Chem. 273, 16434 – 16441 (1998). 201. C. Rachez, Z. Suldan, J. Ward, C. P. Chang, D. Burakov, H. Erdjument-Bromage, P. Tempst, and L. P. Freedman, A novel protein complex that interacts with the vitamin D3 receptor in a liganddependent manner and enhances VDR transactivation in a cell-free system. Genes Dev. 12, 1787 – 1800 (1998). 202. C. Rachez, B. D. Lemon, Z. Suldan, V. Bromleigh, M. Gamble, A. M. Naar, H. Erdjument-Bromage, P. Tempst, and L. P. Freedman, Ligand-dependent transcription activation by nuclear receptors requires the DRIP complex. Nature 398, 824 – 828 (1999). 203. Y. Yanagi, M. Suzawa, M. Kawabata, K. Miyazono, J. Yanagisawa, and S. Kato, Positive and negative modulation of vitamin D receptor function by transforming growth factor- signaling through smad proteins. J. Biol. Chem. 274, 12971 – 12974 (1999). 204. J. Yanagisawa, Y. Yanagi, Y. Masuhiro, M. Suzawa, M. Watanabe, K. Kashiwagi, T. Toriyabe, M. Kawabata, K. Miyazono, and S. Kato, Convergence of transforming growth factor- and vitamin D signaling pathways on SMAD transcriptional coactivators. Science 283, 1317 – 1321 (1999). 205. J. Barsony, J. W. Pike, H. F. DeLuca, and S. J. Marx, Immunocytology with microwave-fixed fibroblasts shows 1, 25-dihydroxyvitamin D3-dependent rapid and estrogen-dependent slow reorganization of vitamin D receptors. J. Cell Biol. 111, 2385 – 2395 (1990). 206. J. Barsony and W. McKoy, Molybdate increases intracellular 3,5guanosine cyclic monophosphate and stabilizes vitamin D receptor
CHAPTER 9 Vitamin D
207.
208.
209.
210.
211.
212.
213.
214.
215.
216.
217.
218.
219.
220.
221.
222.
223.
224.
association with tubulin-containing filaments. J. Biol. Chem. 267, 24457 – 24465 (1992). A. Racz and J. Barsony, Hormone-dependent translocation of vitamin D receptors is linked to transactivation. J. Biol. Chem. 274, 19352 – 19360 (1999). K. Prüfer, C. Schröder, A. Racz, and J. Barsony. Cell cycle dependence of vitamin D receptor expression. In “vitamin D endocrine system: Structural, biological, genetic and clinical aspects” (A. W. Norman, R. Bouillon, and M. Thomasset, eds.) pp. 379 – 382. University of California, Printing and Reprographics, Riverside, 2000. M. Hirst and D. Feldman, Regulation of 1,25(OH)2 vitamin D3 receptor content in cultured LLC-PK1 kidney cells limits hormonal responsiveness. Biochem. Biophys. Res. Commun. 116, 121 – 127 (1983). T. L. Chen, J. M. Li, T. Van Ye, C. M. Cone and D. Feldman, Hormonal responses to 1,25-dihydroxyvitamin D3 in cultured mouse osteoblast-like cells-modulation by changes in receptor level. J. Cell. Physiol. 126, 21 – 28 (1986). A. V. Krishnan and D. Feldman, Cyclic adenosine 3,5-monophosphate upregulates 1,25-dihydroxyvitamin D3 receptor gene expression and enhances hormone action. Mol. Endocrinol. 6, 198 – 206 (1992). A. V. Krishnan, S. D. Cramer, F. R. Bringhurst, and D. Feldman, Regulation of 1,25-dihydroxyvitamin D3 receptors by parathyroid hormone in osteoblastic cells: Role of second messinger pathways. Endocrinology 136, 705 – 712 (1995). A. V. Krishnan and D. Feldman, Regulation of vitamin D receptor abundance. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 179 – 200. Academic Press, San Diego 1997. E. M. Costa, M. A. Hirst, and D. Feldman, Regulation of 1,25dihydroxyvitamin D3 receptors by vitamin D analogs in cultured mammalian cells. Endocrinology 117, 2203 – 2210 (1985). E. M. Costa and D. Feldman, Homologous up-regulation of the 1,25 (OH)2 vitamin D3 receptor in rats. Biochem. Biophys. Res. Commun. 137, 742 – 747 (1986). M. J. Favus, D. J. Mangelsdorf, V. Tembe, B. J. Coe, and M. R. Haussler, Evidence for in vivo upregulation of the intestinal vitamin D receptor during dietary calcium restriction in the rat. J. Clin. Invest. 82, 218 – 224 (1988). M. Strom, M. E. Sandgren, T. A. Brown, and H. F. DeLuca, 1,25-Dihydroxyvitamin D3 up-regulates the 1,25-dihydroxyvitamin D3 receptor in vivo. Proc. Natl. Acad. Sci. USA 86, 9770 – 9773 (1993). X. Zhao and D. Feldman, Regulation of vitamin D receptor abundance and responsiveness during differentiation of HT-29 human colon cancer cells. Endocrinology 132, 1808 – 1814 (1993). R. J. Wiese, A. Uhland-Smith, T. K. Ross, J. M. Prahl, and H. F. DeLuca, Up-regulation of the vitamin D receptor in response to 1,25-dihydroxyvitamin D3 results from ligand-induced stabilization. J. Biol. Chem. 267, 20082 – 20086 (1992). D. Santiso-Mere, T. Sone, G. T. Hilliard, J. W. Pike, and D. P. McDonnell, Positive regulation of the vitamin D receptor by its cognate ligand in heterologous expression systems. Mol. Endocrinol. 7, 833 – 839 (1993). N. C. Arbour, J. M. Prahl, and H. F. DeLuca, Stabilization of the vitamin D receptor in rat osteosarcoma cells through the action of 1,25-dihydroxyvitamin D3. Mol. Endocrinol. 7, 1307 – 1312 (1993). M. L. Chen, A. Perez, D. K. Sanan, G. Heinrich, T. C. Chen, and M. F. Holick, Induction of vitamin D receptor mRNA expression in psoriatic plaques correlates with clinical response to 1,25-dihydroxyvitamin D3. J. Invest. Dermatol. 106, 637 – 641 (1996). T. L. Chen and D. Feldman, Regulation of 1,25-dihydroxyvitamin D3 receptors in cultured mouse bone cells: Correlation of receptor concentration with the rate of cell division. J. Biol. Chem. 256, 5561 – 5566 (1981). A. Krishnan and D. Feldman Stimulation of 1,25-dihydroxyvitamin D3 receptor gene expression in cultured cells by serum and growth factors. J. Bone Miner. Res. 6, 1099 – 1107 (1991).
295 225. S. Pillai D. D. Bikle, and P. M. Elias, 1,25-Dihydroxyvitamin D production and receptor binding in human keratinocytes varies with differentiation. J. Biol. Chem. 263, 5390 – 5395 (1988). 226. X. P. Yu, H. Mocharla, F. G. Hustmyer, and S. C. Manolagas, Vitamin D receptor expression in human lymphocytes: Signal requirements and characterization by western blots and DNA sequencing. J. Biol. Chem. 266, 7588 – 7595 (1991). 227. A. V. Krishnan and D. Feldman, Activation of protein kinase-C inhibits vitamin D receptor gene expression. Mol. Endocrinol. 5, 605 – 612 (1991). 228. J. P. van Leeuwen, J. C. Birkenhager, T. Vink-van Wijngaarden, G. J. van den Bemd, and H. A. Pols, Regulation of 1,25-dihydroxyvitamin D3 receptor gene expression by parathyroid hormone and cAMPagonists. Biochem. Biophys. Res. Commun. 185, 881 – 886 (1992). 229. F. Jehan and H. F. DeLuca, Cloning and characterization of the mouse vitamin D receptor promoter. Proc. Natl. Acad. Sci. USA 94, 10138 – 10143 (1997). 230. S. J. Smith, L. M. Green, M. E. Hayes, and E. B. Mawer, Prostaglandin E2 regulates vitamin D receptor expression, vitamin D-24-hydroxylase activity and cell proliferation in an adherent human myeloid leukemia cell line (Ad-HL60). Prostaglandins Other Lipid Mediat. 57, 73 – 85 (1999). 231. H. Yamamoto, K. Miyamoto, B. Li, Y. Taketani, M. Kitano, Y. Inoue, K. Morita, J. W. Pike, and E. Takeda, The caudal-related homeodomain protein Cdx-2 regulates vitamin D receptor gene expression in the small intestine. J. Bone Miner. Res. 14, 240 – 247 (1999). 232. T. L. Chen, C. M. Cone, E. Morey-Holton, and D. Feldman, Glucocorticoid regulation of 1,25(OH)2-vitamin D3 receptors in cultured mouse bone cells. J. Biol. Chem. 257, 13564 – 13569 (1982). 233. M. Hirst and D. Feldman, Glucocorticoid regulation of 1,25(OH)2 vitamin D3 receptors: Divergent effects on mouse and rat intestine. Endocrinology 111, 1400 – 1402 (1982). 234. M. Hirst and D. Feldman, Glucocorticoids down-regulate the number of 1,25-dihydroxyvitamin D3 receptors in mouse intestine. Biochem. Biophys. Res. Commun. 105, 1590 – 1596 (1982). 235. H. K. Nielsen, E. F. Eriksen, T. Storm, and L. Mosekiilde, The effects of short-term, high-dose treatment with prednisone on the nuclear uptake of 1,25-dihydroxyvitamin D3 in monocytes from normal human subjects. Metabolism 37, 109 – 114 (1988). 236. M. R. Walters, An estrogen-stimulated 1,25-dihydroxyvitamin D3 receptor in rat uterus. Biochem. Biophys. Res. Commun. 103, 721 – 726 (1981). 237. Y. Liel, S. Kraus, J. Levy, and S. Shany, Evidence that estrogens modulate activity and increase the number of 1,25-dihydroxyvitamin D receptors in osteoblast-like cells (ROS 17/2.8). Endocrinology 130, 2597 – 2601 (1992). 238. P. M. Petkovich, J. N. Heersche, D. O. Tinker, and G. Jones, Retinoic acid stimulates 1,25-dihydroxyvitamin D3 binding in rat osteosarcoma cells. J. Biol. Chem. 259, 8274 – 8280 (1984). 239. T. L. Chen and D. Feldman, Retinoic acid modulation of 1,25(OH)2 vitamin D3 receptors and bioresponse in bone cells: Species differences between rat and mouse. Biochem. Biophys. Res. Commun. 132, 74 – 80 (1985). 240. T. A. Reinhardt and R. L. Horst, Parathyroid hormone downregulates 1,25-dihydroxyvitamin D receptors (VDR) and VDR messenger ribonucleic acid in vitro and blocks homologous upregulation of VDR in vivo. Endocrinology 127, 942 – 948 (1990). 241. S. B. Borwn, T. T. Brierley, N. Palanisamy, I. B. Salusky, W. Goodman, M. Brandi, T. B. Drueke, E. Sarfati, P. Urena, R. S. Chaganti, J. W. Pike, and A. Arnold, Vitamin D receptor as a candidate tumor-suppressor gene in severe hyperparathyroidism of uremia. J. Clin. Endocrinol. Metab. 85, 868 – 872 (2000). 242. A. B. Korkor, Reduced binding of [3H]1,25-dihydroxyvitamin D3 in the parathyroid glands of patients with renal failure. N. Engl. J. Med. 316, 1573 – 1577 (1987).
296 243. S. J. Silverberg, E. Shane, D. W. Depster, and J. P. Bilezikian, The effects of vitamin D insufficiency in patients with primary hyperparathyroidism. Am. J. Med. 107, 561 – 567 (1999). 244. D. Feldman, Vitamin D, parathyroid hormone, and calcium: A complex regulatory network. Am. J. Med. 107, 637 – 639 (1999). 245. D. S. Rao, M. Honasoge, G. W. Divine, E. R. Phillips, M. W. Lee, M. R. Ansari, G. B. Talpos, and A. M. Parfitt, Effect of vitamin D nutrition on parathyroid adenoma weight: Pathogenetic and clinical implications. J. Clin. Endocrinol. Metab. 85, 1054 – 1058 (2000). 246. A. W. Norman, Rapid biological responses mediated by 1,25dihydroxyvitamin D3: A case study of transcaltachia (rapid hormonal stimulation of intestinal calcium transport). In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 233 – 256. Academic Press, San Diego, 1997. 247. A. W. Norman, X. Song, L. Zanello, C. Bula, and W. H. Okamura, Rapid and genomic biological responses are mediated by different shapes of the agonist steroid hormone, 1,25(OH)2 vitamin D3. Steroids 64, 120 – 128 (1999). 248. J. Barsony and S. J. Marx, Rapid accumulation of cyclic GMP near activated vitamin D receptors. Proc. Natl. Acad. Sci. USA 88, 1436 – 1440 (1991). 249. D. T. Baran, A. M. Sorensen, V. Shalhoub, T. Owen, A. Oberdorf, G. Stein and J. Lian, 1-,25-dihydroxyvitamin D3 rapidly increases cytosolic calcium in clonal rat osteosarcoma cells lacking the vitamin D receptor. J. Bone Miner. Res. 6, 1269–1275 (1991). 250. A. W. Norman, R. Bouillon, M. C. Farach-Carson, J. E. Bishop, L. X. Zhou, I. Nemere, J. Zhao, K. R. Muralidharan, and W. H. Okamura, Demonstration that 1,25-dihydroxyvitamin D3 is an antagonist of the nongenomic but not genomic biological responses and biological profile of the three A-ring diastereomers of 1,25-dihydroxyvitamin D3. J. Biol. Chem. 268, 20022 – 20030 (1993). 251. A. W. Norman, W. H. Okamura, M. W. Hammond, J. E. Bishop, M. C. Dormanen, R. Bouillon, H. van Baelen, A. L. Ridall, E. Daane, R. Khoury, and M. C. Farach-Carson, Comparison of 6-s-cis- and 6-s-trans-locked analogs of 1,25-dihydroxyvitamin D3 indicates that the 6-s-cis conformation is preferred for rapid nongenomic biological responses and that neither 6-s-cis- nor 6-s-trans-locked analogs are preferred for genomic biological responses. Mol. Endocrinol. 11, 1518 – 1531 (1997). 252. M. Bhatia, J. B. Kirkland, and K. A. Meckling-Gill, Monocytic differentiation of acute promyelocytic leukemia cells in response to 1,25-dihydroxyvitamin D3 is independent of nuclear receptor binding. J. Biol. Chem. 270, 15962 – 15965 (1995). 253. M. Kajikawa, H. Ishida, S. Fujimoto, E. Mukai, M. Nishimura, J. Fujita, Y. Tsuura, Y. Okamoto, A. W. Norman, and Y. Seino, An insulinotropic effect of vitamin D analog with increasing intracellular Ca2 concentration in pancreatic -cells through nongenomic signal transduction. Endocrinology 140, 4706 – 4712 (1999). 254. I. Nemere, Z. Schwartz, H. Pedrozo, V. L. Sylvia, D. D. Dean, and B. D. Boyan, Identification of a membrane receptor for 1,25-dihydroxyvitamin D3 which mediates rapid activation of protein kinase C. J. Bone Miner. Res. 13, 1353 – 1359 (1998). 255. R. P. Heaney, Vitamin D: Role in the calcium economy. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 485 – 497. Academic Press, San Diego, 1997. 256. E. Slatopolsky, A. Brown, and A. Dusso, Pathogenesis of secondary hyperparathyroidism. Kidney Int. Suppl. 73, S14 – 19 (1999). 257. J. A. Johnson and R. Kumar, Renal and intestinal calcium transport: Roles of vitamin D and vitamin D-dependent calcium binding proteins. Semin. Nephrol. 14, 119 – 128 (1994). 258. A. J. Brown, M. Zhong, J. Finch, C. Ritter, R. McCracken, J. Morrissey, and E. Slatopolsky, Rat calcium-sensing receptor is regulated by vitamin D but not by calcium. Am. J. Physiol. 270, F454 – F460 (1996).
FELDMAN, MALLOY, AND GROSS 259. J. L. Finch, A. J. Brown, and E. Slatopolsky, Differential effects of 1,25-dihydroxyvitamin D3 and 19-nor-1,25-dihydroxyvitamin D2 on calcium and phosphorus resorption in bone. J. Am. Soc. Nephrol. 10, 980 – 985 (1999). 260. J. S. Adams, Extrarenal production and action of active vitamin D metabolites in human lymphoproliferative diseases. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 903–921. Academic Press, San Diego, 1997. 261. A. Dusso, A. Brown, and E. Slatopolsky, Extrarenal production of calcitriol. Semin. Nephrol. 14, 144 – 155 (1994). 262. L. E. Siltzbach, D. G. James, E. Neville, J. Turiaf, J. P. Battesti, O. P. Sharma, Y. Hosoda, R. Mikami, and M. Odaka, Course and prognosis of sarcoidosis around the world. Am. J. Med. 57, 847 – 852 (1974). 263. G. L. Barbour, J. W. Coburn, E. Slatopolsky, A. W. Norman, and R. L. Horst, Hypercalcemia in an anephric patient with sarcoidosis: evidence for extrarenal generation of 1,25-dihydroxyvitamin D. N. Engl. J. Med. 305, 440 – 443 (1981). 264. R. S. Mason, T. Frankel, Y. L. Chan, D. Lissner, and S. Posen, Vitamin D conversion by sarcoid lymph node homogenate. Ann. Intern. Med. 100, 59 – 61 (1984). 265. J. F. Seymour and R. F. Gagel, Calcitriol: The major humoral mediator of hypercalcemia in Hodgkins disease and non-Hodgkin’s lymphomas. Blood 82, 1383 – 1394 (1993). 266. J. F. Seymour R. F. Gagel, F. B. Hagemeister, M. A. Dimopoulos, and F. Cabanillas, Calcitriol production in hypercalcemic and normocalcemic patients with non-Hodgkin lymphoma. Ann. Intern. Med. 121, 633 – 640 (1994). 267. R. Kumar, W. R. Cohen, P. Silva, and F. H. Epstein, Elevated 1,25dihydroxyvitamin D plasma levels in normal human pregnancy and lactation. J. Clin. Invest. 63, 342 – 344 (1979). 268. D. D. Bikle, E. Gee, B. Halloran, and J. G. Haddad, Free 1,25-dihydroxyvitamin D levels in serum from normal subjects, pregnant subjects, and subjects with liver disease. J. Clin. Invest. 74, 1966 – 1971 (1984). 269. C. Cheema, B. F. Grant, and R. Marcus, Effects of estrogen on circulating “free” and total 1,25-dihydroxyvitamin D and on the parathyroid-vitamin D axis in postmenopausal women. J. Clin. Invest. 83, 537 – 542 (1989). 270. S. Vargas, R. Bouillon, H. Van Baelen, and L. G. Raisz, Effects of vitamin D-binding protein on bone resorption stimulated by 1,25 dihydroxyvitamin D3. Calcif. Tissue Int. 47, 164 – 168 (1990). 271. T. K. Gray, G. E. Lester, and R. S. Lorenc, Evidence for extra-renal 1-hydroxylation of 25-hydroxyvitamin D3 in pregnancy. Science 204, 1311 – 1313 (1979). 272. J. A. Whitsett, M. Ho, R. C. Tsang, E. J. Norman, and K. G. Adams, Synthesis of 1,25-dihydroxyvitamin D3 by human placenta in vitro. J. Clin. Endocrinol. Metab. 53, 484 – 488 (1981). 273. N. A. Breslau and J. E. Zerwekh, Relationship of estrogen and pregnancy to calcium homeostasis in pseudohypoparathyroidism. J. Clin. Endocrinol. Metab. 62, 45 – 51 (1986). 274. A. D. Care, Vitamin D in pregnancy, the fetoplacental unit, and lactation. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 437 – 446. Academic Press, San Diego, 1997. 275. A. R. Glass and C. Eil, Ketoconazole-induced reduction in serum 1,25-dihydroxyvitamin D and total serum calcium in hypercalcemic patients. J. Clin. Endocrinol. Metab. 66, 934 – 938 (1988). 276. J. S. Adams, O. P. Sharma, M. M. Diz, and D. B. Endres, Ketoconazole decreases the serum 1,25-dihydroxyvitamin D and calcium concentration in sarcoidosis-associated hypercalcemia. J. Clin. Endocrinol. Metab. 70, 1090 – 1095 (1990). 277. N. A. Breslau, G. M. Perminger, B. V. Adams, J. Otey, and C. Y. C. Pak, Use of ketoconazole to probe the pathogenetic importance of 1,25-dihydroxyvitamin D in absorptive hypercalciuria. J. Clin. Endocrinol. Metab. 75, 1446 – 1452 (1992).
CHAPTER 9 Vitamin D 278. O. P. Sharma, Vitamin D, calcium, and sarcoidosis. Chest 109, 535 – 539 (1996). 279. D. Feldman, Ketoconazole and other imidazole derivatives as inhibitors of steroidogenesis. Endocr. Rev. 7, 409 – 420 (1986). 280. D. S. Loose, P. B. Kan, M. A. Hirst, R. A. Marcus, and D. F. Feldman, Ketoconazole blocks adrenal steroidogenesis by inhibiting cytochrome P450-dependent enzymes. J. Clin. Invest. 71, 1495 – 1499 (1983). 281. H. L. Henry, Effect of ketoconazole and miconazole on 25-hydroxyvitamin D3 metabolism by cultured chick kidney cells. J. Steroid Biochem. 23, 991 – 994 (1985). 282. A. S. Douglas, S. Ali, and S. S. Bakhshi, Does vitamin D deficiency account for ethnic differences in tuberculosis seasonality in the UK? Ethn. Health 3, 247 – 253 (1998). 283. R. J. Wilkinson, M. Llewelyn, Z. Toossi, P. Patel, G. Pasvol, A. Lalvani, D. Wright, M. Latif, and R. N. Davidson, Influence of vitamin D deficiency and vitamin D receptor polymorphisms on tuberculosis among Gujarati Asians in west London: A case-control study. Lancet 355, 618 – 621 (2000). 284. K. A. Rockett, R. Brookes, I. Udalova, V. Vidal, A. V. Hill, and D. Kwiatkowski, 1,25-Dihydroxyvitamin D3 induces nitric oxide synthase and suppresses growth of Mycobacterium tuberculosis in a human macrophage-like cell line. Infect. Immun. 66, 5314 – 5321 (1998). 285. S. Ashkar, G. F. Weber, V. Panoutsakopoulou, M. E. Sanchirico, M. Jansson, S. Zawaideh, S. R. Rittling, D. T. Denhardt, M. J. Glimcher, and H. Cantor, Eta-1 (osteopontin): An early component of type-1 (cell-mediated) immunity. Science 287, 860 – 864 (2000). 286. E. B. Mawer, M. E. Hayes, P. E. Still, M. Davies, G. A. Lumb, J. Palit, and P. J. Holt, Evidence for nonrenal synthesis of 1,25-dihydroxyvitamin D in patients with inflammatory arthritis. J. Bone Miner. Res. 6, 733 – 739 (1991). 287. S. J. Smith, M. E. Hayes, P. L. Selby, and E. B. Mawer, Autocrine control of vitamin D metabolism in synovial cells from arthritic patients. Ann. Rheum. Dis. 58, 372 – 378 (1999). 288. R. Skowronski, D. Peehl, and D. Feldman, Vitamin D and prostate cancer: 1,25-Dihydroxyvitamin D3 receptors and actions in prostate cancer cell lines. Endocrinology 132, 1952 – 1960 (1993). 289. D. M. Peehl, R. J. Skowronski, G. K. Leung, S. T. Wong, T. A. Stamey, and D. Feldman, Antiproliferative effects of 1,25-dihydroxy-vitamin D3 on primary cultures of human prostatic cells. Cancer Res. 54, 805 – 810 (1994). 290. G. G. Schwartz, T. A. Oeler, M. R. Uskokovic, and R. R. Bahnson, Human prostate cancer cells: Inhibition of proliferation by vitamin D analogs. Anticancer Res. 14, 1077 – 1081 (1994). 291. R. J. Skowronski, D. M. Peehl, and D. Feldman, Actions of vitamin D3 analogs on human prostate cancer cell lines: Comparison with 1,25-dihydroxyvitamin D3. Endocrinology 136, 20 – 26 (1995). 292. B. R. Konety, G. G. Schwartz, J. S. Acierno, Jr., M. J. Becich, and R. H. Getzenberg, The role of vitamin D in normal prostate growth and differentiation. Cell Growth Differ. 7, 1563 – 1570 (1996). 293. G. K. Fu, A. A. Portale, and W. L. Miller, Complete structure of the human gene for the vitamin D 1-hydroxylase, P450c1. DNA Cell. Biol. 16, 1499 – 1507 (1997). 294. D. D. Bikle, E. Gee, and S. Pillai, Regulation of keratinocyte growth, differentiation, and vitamin D metabolism by analogs of 1,25-dihydroxyvitamin D. J. Invest. Dermatol. 101, 713 – 718 (1993). 295. K. Kragballe, Psoriasis and other skin diseases. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 1213 – 1225. Academic Press, San Diego, 1997. 296. J. Reichrath, A. Perez, S. M. Muller, T. C. Chen, A. Kerber, F. A. Bahmer, and M. F. Holick, Topical calcitriol (1,25-dihydroxyvitamin D3) treatment of psoriasis: An immunohistological evaluation. Acta Derm. Venereol. 77, 268 – 272 (1997).
297 297. J. Reichrath, S. M. Muller, A. Kerber, H. P. Baum, and F. A. Bahmer, Biologic effects of topical calcipotriol (MC 903) treatment in psoriatic skin. J. Am. Acad. Dermatol. 36, 19 – 28 (1997). 298. R. St-Arnaud and F. H. Glorieux, Vitamin D and bone development. In “Vitamin D” (D. Feldman, F. H. Gloriux, and J. W. Pike, eds.), pp. 293 – 304. Academic Press, San Diego, 1997. 299. M. K. Drezner, Clinical disorders of phosphate homeostasis. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 733 – 753. Academic Press, San Diego, 1997. 300. T. O. Carpenter and K. L. Insogna, The hypocalcemic disorders: Differential diagnosis and therapeutic use of vitamin D. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 923 – 936. Academic Press, San Diego, 1997. 301. R. Bouillon, The many faces of rickets. N. Engl. J. Med. 338, 681 – 682 (1998). 302. W. L. Miller and A. A. Portale, Genetic disorders of vitamin D biosynthesis. Endocrinol. Metab. Clin. North Am. 28, 825 – 840 (1999). 303. M. K. Drezner, PHEX gene and hypophosphatemia. Kidney Int. 57, 9 – 18 (2000). 304. V. A. Prader, R. Illig, and E. Heierli, Eine besondere form der primaren vitamin-D-resistenten rachitis mit hypocalcamie und autosomal-dominantem erbgang: Die hereditare pseudo-mangelrachitis. Helv. Paediatr. Acta. 16, 452 – 468 (1961). 305. F. H. Glorieux and R. St-Arnaud, Molecular cloning of (25-OH D)1 -hydroxylase: An approach to the understanding of vitamin D pseudo-deficiency. Recent Prog. Horm. Res. 53, 341 – 349; discussion 350 (1998). 306. S. Kato, Genetic mutation in the human 25-hydroxyvitamin D3 1-hydroxylase gene causes vitamin D-dependent rickets type I. Mol. Cell. Endocrinol. 156, 7 – 12 (1999). 307. F. H. Glorieux and R. St-Arnaud, Vitamin D pseudodeficiency. In “Vitamin D” (D. Feldman, F. H. Glorieux, and J. W. Pike, eds.), pp. 755 – 764. Academic Press, San Diego, 1997. 308. S. Kato, J. Yanagisawa, A. Murayama, S. Kitanaka, and K. Takeyama, The importance of 25-hydroxyvitamin D3 1hydroxylase gene in vitamin D-dependent rickets. Curr. Opin. Nephrol. Hypertens. 7, 377 – 383 (1998). 309. D. Feldman and P. J. Malloy, Hereditary 1,25-dihydroxyvitamin D resistant rickets: Molecular basis and implications for the role of 1,25(OH)2D3 in normal physiology. Mol. Cell. Endocrinol. 72, C57 – C62 (1990). 310. M. R. Hughes, P. J. Malloy, B. W. OMalley, J. W. Pike, and D. Feldman, Genetic defects of the 1,25-dihydroxyvitamin D3 receptor. J. Recept. Res. 11, 699 – 716 (1991). 311. P. J. Malloy and D. Feldman, “Molecular Defects in the Vitamin D Receptor Associated with Hereditary 1,25-Dihydroxyvitamin D Resistant Rickets” (M. Holick, ed.). Humana Press, 1998. 312. P. J. Malloy and D. Feldman, Vitamin D resistance. Am. J. Med. 106, 355 – 370 (1999). 313. M. R. Hughes, P. J. Malloy, D. G. Kieback, R. A. Kesterson, J. W. Pike, D. Feldman, and B. W. OMalley, Point mutations in the human vitamin D receptor gene associated with hypocalcemic rickets. Science 242, 1702 – 1705 (1988). 314. M. A. Hirst, H. I. Hochman, and D. Feldman, Vitamin D resistance and alopecia: A kindred with normal 1,25-dihydroxyvitamin D binding, but decreased receptor affinity for deoxyribonucleic acid. J. Clin. Endocrinol. Metab. 60, 490 – 495 (1985). 315. P. J. Malloy, Z. Hochberg, J. W. Pike, and D. Feldman, Abnormal binding of vitamin D receptors to deoxyribonucleic acid in a kindred with vitamin D-dependent rickets, type II. J. Clin. Endocrinol. Metab. 68, 263 – 269 (1989). 316. T. Sone, R. A. Scott, M. R. Hughes, P. J. Malloy, D. Feldman, B. W. O’Malley, and J. W. Pike, Mutant vitamin D receptors which confer hereditary resistance to 1,25-dihydroxyvitamin D3 in humans are transcriptionally inactive in vitro. J. Biol. Chem. 264, 20230 –20234 (1989).
298 317. T. Sone, S. J. Marx, U. A. Liberman, and J. W. Pike, A unique point mutation in the human vitamin D receptor chromosomal gene confers hereditary resistance to 1,25-dihydroxyvitamin D3. Mol. Endocrinol. 4, 623 – 631 (1990). 318. H. Yagi, K. Ozono, H. Miyake, K. Nagashima, T. Kuroume, and J. W. Pike, A new point mutation in the deoxyribonucleic acidbinding domain of the vitamin D receptor in a kindred with hereditary 1,25-dihydroxyvitamin D-resistant rickets. J. Clin. Endocrinol. Metab. 76, 509 – 512 (1993). 319. P. J. Malloy, Y. Weisman, and D. Feldman, Hereditary 1,25dihydroxyvitamin D-resistant rickets resulting from a mutation in the vitamin D receptor deoxyribonucleic acid-binding domain. J. Clin. Endocrinol. Metab. 78, 313 – 316 (1994). 320. H. H. Ritchie, M. R. Hughes, E. T. Thompson, P. J. Malloy, Z. Hochberg, D. Feldman, J. W. Pike, and B. W. OMalley, An ochre mutation in the vitamin D receptor gene causes hereditary 1,25-dihydroxyvitamin D3-resistant rickets in three families. Proc. Natl. Acad. Sci. USA 86, 9783 – 9787 (1989). 321. P. J. Malloy, Z. Hochberg, D. Tiosano, J. W. Pike, M. R. Hughes, and D. Feldman, The molecular basis of hereditary 1,25-dihydroxyvitamin D3 resistant rickets in seven related families. J. Clin. Invest. 86, 2071 – 2079 (1990). 322. R. J. Wiese, H. Goto, J. M. Prahl, S. J. Marx, M. Thomas, A. al-Aqeel, and H. F. DeLuca, Vitamin D-dependency rickets type II: Truncated vitamin D receptor in three kindreds. Mol. Cell. Endocrinol. 90, 197 – 201 (1993). 323. G. K. Whitfield, S. H. Selznick, C. A. Haussler, J. C. Hsieh, M. A. Galligan, P. W. Jurutka, P. D. Thompson, S. M. Lee, J. E. Zerwekh, and M. R. Haussler, Vitamin D receptors from patients with resistance to 1,25-dihydroxyvitamin D3: Point mutations confer reduced transactivation in response to ligand and impaired interaction with the retinoid X receptor heterodimeric partner. Mol. Endocrinol. 10, 1617 – 1631 (1996). 324. M. Hewison, A. R. Rut, K. Kristjansson, R. E. Walker, M. J. Dillon, M. R. Hughes, and J. L. ORiordan, Tissue resistance to 1,25-dihydroxyvitamin D without a mutation of the vitamin D receptor gene. Clin. Endocrinol. 39, 663 – 670 (1993). 325. M. I. New, S. Nimkarn, D. D. Brandon, S. Cunningham-Rundles, R. C. Wilson, R. S. Newfield, J. Vandermeulen, N. Barron, C. Russo, D. L. Loriaux, and B. OMalley, Resistance to several steroids in two sisters. J. Clin. Endocrinol. Metab. 84, 4454 (1999). 326. Y. Weisman, N. Jaccard, C. Legum, Z. Spirer, G. Yedwab, L. Even, S. Edelstein, A. M. Kaye, and Z. Hochberg, Prenatal diagnosis of vitamin D-dependent rickets, type II: Response to 1,25-dihydroxyvitamin D in amniotic fluid cells and fetal tissues. J. Clin. Endocrinol. Metab 71, 937 – 943 (1990). 327. Y. Weisman, P. J. Malloy, A. V. Krishnan, N. Jaccard, D. Feldman, and Z. Hochberg, In “Ninth workshop on Vitamin D,” p. 106 (1994). 328. S. Balsan, M. Garabedian, M. Larchet, A. M. Gorski, G. Cournot, C. Tau, A. Bourdeau, C. Silve, and C. Ricour, Long-term nocturnal calcium infusions can cure rickets and promote normal mineralization in hereditary resistance to 1,25-dihydroxyvitamin D. J. Clin. Invest. 77, 1661 – 1667 (1986). 329. Z. Hochberg, D. Tiosano, and L. Even, Calcium therapy for calcitriol-resistant rickets. J. Pediatr. 121, 803 – 808 (1992). 330. A. al-Aqeel, P. Ozand, S. Sobki, W. Sewairi, and S. Marx, The combined use of intravenous and oral calcium for the treatment of vitamin D dependent rickets type II (VDDRII). Clin. Endocrinol. 39, 229 – 237 (1993). 331. T. Yoshizawa, Y. Handa, Y. Uematsu, S. Takeda, K. Sekine, Y. Yoshihara, T. Kawakami, K. Arioka, H. Sato, Y. Uchiyama, S. Masushige, A. Fukamizu, T. Matsumoto, and S. Kato, Mice lacking the vitamin D receptor exhibit impaired bone formation, uterine hypoplasia and growth retardation after weaning. Nature Genet. 16, 391 – 396 (1997).
FELDMAN, MALLOY, AND GROSS 332. Y. C. Li, A. E. Pirro, M. Amling, G. Delling, R. Baron, R. Bronson, and M. B. Demay, Targeted ablation of the vitamin D receptor: An animal model of vitamin D-dependent rickets type II with alopecia. Proc. Natl. Acad. Sci. USA 94, 9831 – 9835, (1997). 333. M. Amling, M. Priemel, T. Holzmann, K. Chapin, J. M. Rueger, R. Baron, and M. B. Demay, Rescue of the skeletal phenotype of vitamin D receptor-ablated mice in the setting of normal mineral ion homeostasis: Formal histomorphometric and biomechanical analyses. Endocrinology 140, 4982 – 4987 (1999). 334. K. Kinuta, H. Tanaka, T. Moriwake, K. Aya, S. Kato, and Y. Seino, Vitamin D is an important factor in estrogen biosynthesis of both female and male gonads. Endocrinology 141, 1317 – 1324 (2000). 335. S. Tanaka, M. Haji, R. Takayanagi, Y. Sugioka, and H. Nawata, 1,25Dihydroxyvitamin D3 enhances the enzymatic activity and expression of the messenger ribonucleic acid for aromatase cytochrome P450 synergistically with dexamethasone depending on the vitamin D receptor level in cultured human osteoblasts. Endocrinology 137, 1860 – 1869 (1996). 336. F. Francis, T. M. Strom, S. Hennig, A. Boddrich, B. Lorenz, O. Brandau, K. L. Mohnike, M. Cagnoli, C. Steffens, S. Klages, K. Borzym, T. Pohl, C. Oudet, M. J. Econs, P. S. Rowe, R. Reinhardt, T. Meitinger, and H. Lehrach, Genomic organization of the human PEX gene mutated in X-linked dominant hypophosphatemic rickets. Genome Res. 7, 573 – 585 (1997). 337. P. S. Rowe, C. L. Oudet, F. Francis, C. Sinding, S. Pannetier, M. J. Econs, T. M. Strom, T. Meitinger, M. Garabedian. A. David, M. A. Macher, E. Questiaux, E. Popowska, E. Pronicka, A. P. Read, A. Mokrzycki, F. H.Glorieux, M. K. Drezner, A. Hanauer, H. Lehrach, J. N. Goulding, and J. L. ORiordan, Distribution of mutations in the PEX gene in families with X-linked hypophosphataemic rickets (HYP). Hum. Mol. Genet. 6, 539 – 549 (1997). 338. Q. Cai, S. F. Hodgson, P. C. Kao, V. A. Lennon, G. G. Klee, A. R. Zinsmiester, and R. Kumar, Brief report: Inhibition of renal phosphate transport by a tumor product in a patient with oncogenic osteomalacia. N. Engl. J. Med. 330, 1645 – 1649 (1994). 339. J. A. Eisman, Genetics of osteoporosis. Endocr. Rev. 20, 788 – 804 (1999). 340. J. M. Zmuda, J. A. Cauley, and R. E. Ferrell, Recent progress in understanding the genetic susceptibility to osteoporosis. Genet. Epidemiol. 16, 356 – 367 (1999). 341. A. G. Uitterlinden, H. Burger, Q. Huang, E. Odding, C. M. van Duijn, A. Hofman, J. C. Birkenhager, J. P. T. M. van Leeuwen, and H. A. P. Pols, Vitamin D receptor genotype is associated with radiographic osteoarthritis at the knee. J. Clin. Invest. 100, 259 – 263 (1997). 342. N. A. Morrison, J. C. Qi, A. Tokita, P. J. Kelly, L. Crofts, T. V. Nguyen, P. N. Sambrook, and J. A. Eisman, Prediction of bone density from vitamin D receptor alleles. Nature 367, 284 – 287 (1994). 343. C. Gross, T. R. Eccleshall, and D. Feldman, Vitamin D receptor gene alleles and osteoporosis. In “Principles of Bone Biology” (J. P. Bilezikian, L. G. Raisz, and G. A. Rodan, eds.), pp. 917 – 933. Academic Press, San Diego, 1996. 344. G. S. Cooper and D. M. Umbach, Are vitamin D receptor polymorphisms associated with bone mineral density? A meta-analysis. J. Bone Miner. Res. 11, 1841 – 1849 (1996). 345. G. Gong, H. S. Stern, S. C. Cheng, N. Fong, J. Mordeson, H. W. Deng, and R. R. Recker, The association of bone mineral density with vitamin D receptor gene polymorphisms. Osteoporos. Int. 9, 55 – 64 (1999). 346. G. Howard, T. Nguyen, N. Morrison, T. Watanabe, P. Sambrook, J. Eisman, and P. Kelly, Genetic influences on bone density: Physiological correlates of vitamin D receptor gene alleles in premenopausal women. Notification of genotype corrections. J. Clin. Endocrinol. Metab. 83, 1043 (1998).
CHAPTER 9 Vitamin D 347. C. Gross, T. R. Eccleshall, P. J. Malloy, M. L. Villa, R. Marcus, and D. Feldman, The presence of a polymorphism at the translation initiation site of the vitamin D receptor gene is associated with low bone mineral density in postmenopausal Mexican-American women. J. Bone Miner. Res. 11, 1850 – 1855 (1996). 348. S. S. Harris, T. R. Eccleshall, C. Gross, B. Dawson-Hughes, and D. Feldman, The vitamin D receptor start codon polymorphism (FokI) and bone mineral density in premenopausal American black and white women. J. Bone Miner. Res. 12, 1043 – 1048 (1997). 349. L. Gennari, L. Becherini, R. Mansani, L. Masi, A. Falchetti, A. Morelli, E. Colli, S. Gonnelli, C. Cepollaro, and M. L. Brandi, FokI polymorphism at translation initiation site of the vitamin D receptor gene predicts bone mineral density and vertebral fractures in postmenopausal Italian women. J. Bone Miner. Res. 14, 1379 – 1386 (1999). 350. S. Ferrari, D. Manen, J. P. Bonjour, D. Slosman, and R. Rizzoli, Bone mineral mass and calcium and phosphate metabolism in young men: Relationships with vitamin D receptor allelic polymorphisms. J. Clin. Endocrinol. Metab. 84, 2043 – 2048 (1990). 351. T. R. Eccleshall, P. Garnero, C. Gross, P. D. Delmas, and D. Feldman, Lack of correlation between the start codon polymorphism of the vitamin D receptor and bone mineral density in premenopausal French women. J. Bone Miner. Res. 13, 31 – 35 (1998). 352. S. K. Ames, K. J. Ellis, S. K. Gunn, K. C. Copeland, and S. A. Abrams, Vitamin D receptor gene Fokl polymorphism predicts calcium absorption and bone mineral density in children. J. Bone Miner. Res. 14, 740 – 746 (1999). 353. D. Feldman, Androgen and vitamin D receptor gene polymorphisms: The long and short of prostate cancer risk. J. Natl. Cancer Inst. 89, 109 – 111 (1997). 354. S. A. Ingles, R. K. Ross, M. C. Yu, R. A. Irvine, G. La Pera, R. W. Haile, and G. A. Coetzee, Association of prostate cancer with genetic polymorphisms in vitamin D receptor and androgen receptor. Cancer Inst. 89, 166 – 170 (1997). 355. J. A. Taylor, A. Hirvonen, M. Watson, G. Pittman, J. L. Mohler, and D. A. Bell, Association of prostate cancer with vitamin D receptor gene polymorphism. Cancer Res. 56, 4108 – 4110 (1990). 356. A. S. Kibel, S. D. Isaacs, W. B. Isaacs, and G. S. Bova, Vitamin D receptor polymorphisms and lethal prostate cancer. J. Urol. 160, 1405 – 1409 (1998). 357. J. E. Curran, T. Vaughan, R. A. Lea, S. R. Weinstein, N. A. Morrison, and L. R. Griffiths, Association of A vitamin D receptor polymorphism with sporadic breast cancer development. Int. J. Cancer 83, 723 – 726 (1999). 358. A. C. Lundin, P. Soderkvist, B. Eriksson, M. Bergman-Jungestrom, and S. Wingren, Association of breast cancer progression with a vitamin D receptor gene polymorphism, South-East Sweden Breast Cancer Group. Cancer Res. 59, 2332 – 2334 (1999). 359. M. Ruggiero, S. Pacini, M. Amato, S. Aterini, and V. Chiarugi, Association between vitamin D receptor gene polymorphism and nephrolithiasis. Miner. Electrolyte Metab. 25, 185 – 190 (1999). 360. S. A. Ingles, D. G. Garcia, W. Wang, A. Nieters, B. E. Henderson, L. N. Kolonel, R. W. Haile, and G. A. Coetzee, Vitamin D receptor genotype and breast cancer in Latinas (United States). Cancer Causes Control 11, 25 – 30 (2000). 361. R. Bouillon, W. H. Okamura, and A. W. Norman, Structure-function relationships in the vitamin D endocrine system. Endocr. Rev. 16, 200 – 257 (1995). 362. K. M. Casteels, C. Mathieu, M. Waer, D. Valckx, L. Overbergh, J. M. Laureys, and R. Bouillon, Prevention of type I diabetes in nonobese diabetic mice by late intervention with nonhypercalcemic analogs of 1,25-dihydroxyvitamin D3 in combination with a short induction course of cyclosporin A. Endocrinology 139, 95 – 102 (1998).
299 363. A. J. Brown and E. Slatopolsky, Vitamin D analogs: Perspectives for treatment. Miner. Electrolyte Metab. 25, 337 – 341 (1999). 364. E. Slatopolsky, A. Dusso, and A. Brown, New analogs of vitamin D3. Kidney Int. Suppl. 73, S46 – 51 (1999). 365. A. Verstuyf, L. Verlinden, S. Segaert, E. Van Etten, C. Mathieu, and R. Bouillon, Nonclassical effects of l, 25-dihydroxyvitamin D3 and its analogs. Miner. Electrolyte Metab. 25, 345 – 348 (1999). 366. J. A. Kanis, Vitamin D analogs: From renal bone disease to osteoporosis. Kidney Int. Suppl. 73, S77 – 81 (1999). 367. M. Quack and C. Carlberg, Selective recognition of vitamin D receptor conformations mediates promoter selectivity of vitamin D analogs. Mol. Pharmacol. 55, 1077 – 1087 (1999). 368. M. R. Uskokovic, G. P. Studzinski, and S. G. Reddy, The 16-ene vitamin D analogs. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 1045 – 1070. Academic Press, San Diego, 1997. 369. L. Binderup, E. Binderup, and W. O. Godtfredsen, Development of new vitamin D analogs. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 1027 – 1043. Academic Press, San Diego 1997. 370. N. Kubodera, K. Sato, and Y. Nishii, Characteristics of 22-oxacalcitriol (OCT) and 2-(3-hydroxypropoxy)-calcitriol (ED-71). In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 1071 – 1086. Academic Press, San Diego 1997. 371. W. H. Okamura and G. D. Zhu, Chemistry and design: Structural biology of vitamin D action. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 939 – 971. Academic Press, San Diego 1997. 372. S. Peleg, C. Nguyen, B. T. Woodard, J. K. Lee, and G. H. Posner, Differential use of transcription activation function 2 domain of the vitamin D receptor by 1,25-dihydroxyvitamin D3 and its A ring-modified analogs. Mol. Endocrinol. 12, 525 – 535 (1998). 373. R. Bouillon, A. Verstuyf, J. Zhao, B. K. Tan, and H. Van Baelen, Nonhypercalcemic vitamin D analogs: Interactions with the vitamin D-binding protein. Horm. Res. 45, 117 – 121 (1996). 374. A. J. Brown and E. Slatopolsky, Mechanisms for the selective actions of vitamin D analogs. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 995 – 1009. Academic Press, San Diego, 1997. 375. F. J. Dilworth, G. R. Williams, A. M. Kissmeyer, J. L. Nielsen, E. Binderup, M. J. Calverley, H. L. Makin, and G. Jones, The vitamin D analog, KH1060, is rapidly degraded both in vivo and in vitro via several pathways: Principal metabolites generated retain significant biological activity. Endocrinology 138, 5485 – 5496 (1997). 376. X. Y. Zhao, T. R. Eccleshall, A. V. Krishnan, C. Gross, and D. Feldman, Analysis of vitamin D analog-induced heterodimerization of vitamin D receptor with retinoid X receptor using the yeast twohybrid system. Mol. Endocrinol. 11, 366 – 378 (1997). 377. W. Yang and L. P. Freeman, 20-Epi analogues of 1,25-dihydroxyvitamin D3 are highly potent inducers of DRIP coactivator complex binding to the vitamin D3 receptor. J. Biol. Chem. 274, 16838 – 16845 (1999). 378. S. Peleg, M. Sastry, E. D. Collins, J. E. Bishop, and A. W. Norman, Distinct conformational changes induced by 20-epi analogues of 1,25-dihydroxyvitamin D3 are associated with enhanced activation of the vitamin D receptor. J. Biol. Chem. 270, 10551 – 10558 (1995). 379. S. Peleg, Molecular basis for differential action of vitamin D analogs. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 1011 – 1025. Academic Press, San Diego 1997. 380. T. Jaaskelainen, S. Ryhanen, A. Mahonen, H. F. DeLuca, and P. H. Maenpaa, Mechanism of action of superactive vitamin D analogs through regulated receptor degradation. J. Cell. Biochem. 76, 548 – 558 (2000). 381. A. Verstuyf, L. Verlinden, H. Van Baelen, K. Sabbe, C. D’Hallewyn, P. De Clercq, M. Vandewalle, and R. Bouillon, The biological activity
300
382.
383.
384.
385.
386.
387.
388.
389.
390.
391.
392.
393. 394.
395.
396.
FELDMAN, MALLOY, AND GROSS of nonsteroidal vitamin D hormone analogs lacking both the C- and D-rings. J. Bone Miner. Res. 13, 549 – 558 (1998). T. C. Polek, S. Murthy, S. E. Blutt, E. A. Allegretto, and N. Wrigel, Vitamin D receptor (VDR) agonists inhibit prostate cancer cell growth in vitro and in a nude mouse tumor model. American Assoc. Cancer Research Annual Meeting, San Francisco, CA, April 1 – 5, 2000, Abstract 27/2, p. 426. K. W. Colston, A. G. Mackay, S. Y. James, L. Binderup, S. Chander, and R. C. Coombes, EB-1089: A new vitamin D analogue that inhibits the growth of breast cancer cells in vivo and in vitro. Biochem. Pharmacol. 44, 2273 – 2280 (1992). M. Haq, R. Kremer, D. Goltzman, and S. A. Rabbani, A vitamin D analogue (EB1089) inhibits parathyroid hormone-related peptide production and prevents the development of malignancy-associated hypercalcemia in vivo. J. Clin. Invest. 91, 2416 – 2422 (1993). B. L. Lokeshwar, G. G. Schwartz, M. G. Selzer, K. L. Burnstein, S. H. Zhuang, N. L. Block, and L. Binderup, Inhibition of prostate cancer metastasis in vivo: A comparison of 1,25-dihydroxyvitamin D (calcitriol) and EB1089. Cancer Epidemiol. Biomarkers Prev. 8, 241 – 248 (1999). W. H. Park, J. G. Seol, E. S. Kim, C. W. Jung, C. C. Lee, L. Binderup, H. P. Koeffler, B. K. Kim, and Y. Y. Lee, Cell cycle arrest induced by the vitamin D3 analog EB1089 in NCI-H929 myeloma cells is associated with induction of the cyclin-dependent kinase inhibitor p27. Exp. Cell Res. 254, 279 – 286 (2000). T. Gulliford, J. English, K. W. Colston, P. Menday, S. Moller, and R. C. Coombes, A phase I study of the vitamin D analogue EB 1089 in patients with advanced breast and colorectal cancer. Br. J. Cancer 78, 6 – 13 (1996). T. R. J. Evans, K. J. Hamberg, T. Skov, H. L. O. Haahr, P. Menday, C. Bay, and L. Binderup. Seocalcitol (EB 1089)-clinical experience to date. In “Vitamin D endocrine system: Structural, biological, genetic and clinical aspects” (A. W. Norman, R. Bouillon, and M. Thomasset, eds.) pp. 485 – 488, University of California, Printing and Reprographics, Riverside, 2000. J. Lemire, The role of vitamin D3 in immunosuppression: Lessons from autoimmunity and transplantation. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 1167 – 1181. Academic Press, San Diego, 1997. J. Abe, M. Morikawa, K. Miyamoto, S. Kaiho, M. Fukushima, C. Miyaura, E. Abe, T. Suda, and Y. Nishii, Synthetic analogues of vitamin D3 with an oxygen atom in the side chain skeleton: A trial of the development of vitamin D compounds which exhibit potent differentiation-inducing activity without inducing hypercalcemia. FEBS Lett. 226, 58 – 62 (1987). K. Colston, Vitamin D and brest cancer: Therapeutic potential of new vitamin D analogs. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 1107 – 1123. Academic Press, San Diego, 1997. C. Gross, D. M. Peehl, and D. Feldman, Vitamin D and prostate cancer. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 1125 – 1139. Academic Press, San Diego, 1997. G. J. Miller, Vitamin D and prostate cancer: Biologic interactions and clinical potentials. Cancer Metastasis Rev. 17, 353 – 360 (1998). J. Welsh, K. Van Weelden, L. Flanagan, I. Byrne, E. Nolan, and C. J. Narvaez (1998). The role of vitamin D3 and antiestrogens in modulating apoptosis of breast cancer cells and tumors. Subcell. Biochem. 30, 245 – 270. B. R. Konety, C. S. Johnson, D. L. Trump, and R. H. Getzenberg, Vitamin D in the prevention and treatment of prostate cancer. Semin. Urol. Oncol. 17, 77 – 84 (1999). H. Hofer, G. Ho, M. Peterlik, M. R. Uskokovic, J. K. Lee, M. C. White, G. H. Posner, and H. S. Cross, Biological effects of 1hydroxy- and 1-(hydroxymethyl)-vitamin D compounds relevant for potential colorectal cancer therapy. J. Pharmacol. Exp. Ther. 291, 450 – 455 (1999).
397. D. Feldman, X. Y. Zhao, and A. V. Krishnan, Vitamin D and prostate cancer. Endocrinology 141, 5 – 9 (2000). 398. D. Miura, K. Manabe, K. Ozono, M. Saito, Q. Gao, A. W. Norman, and S. Ishizuka, Antagonistic action of novel 1,25-dihydroxyvitamin D3-26, 23- lactone analogs on differentition of human leukemia cells (HL-60) induced by 1,25-dihydroxyvitamin D3. J. Biol. Chem. 274, 16392 – 16399 (1999). 399. K. Ozono, M. Saito, D. Miura, T. Michigami, S. Nakajima, and S. Ishizuka, Analysis of the molecular mechanism for the antagonistic action of a novel 1,25-dihydroxyvitamin D3 analogue toward vitamin D receptor function. J. Biol. Chem. 274, 32376 – 32381 (1999). 400. S. D. Chan and D. Atkins, The temporal distribution of the 1,25dihydroxycholecalciferol receptor in the rat jejunal villus. Clin. Sci. 67, 285 – 290 (1984). 401. M. A. Hirst and D. Feldman, 1,25-Dihydroxyvitamin D3 receptors in mouse colon. J. Steroid Biochem. 14, 315 – 319 (1981). 402. P. Lointier, F. Meggouh, D. Pezet, M. Dapoigny, P. N. Dieng, S. Saez, and J. Chipponi, Specific receptors for 1,25-dihydroxyvitamin D3 (1,25-DR) and human colorectal carcinogenesis. Anticancer Res. 9, 1921 – 1924 (1989). 403. P. Lointier, F. Meggouh, P. Dechelotte, D. Pezet, C. Ferrier, J. Chipponi, and S. Saez, 1,25-Dihydroxyvitamin D3 receptors and human colon adenocarcinoma. Br. J. Surg. 78, 435 – 439 (1991). 404. W. M. Tong, G. Bises, Y. Sheinin, A. Ellinger, D. Genser, R. Potzi, F. Wrba, E. Wenzl, R. Roka, N. Neuhold, M. Peterlik, and H. S. Cross, Establishment of primary cultures from human colonic tissue during tumor progression: Vitamin D responses and vitamin-Dreceptor expression. Int. J. Cancer 75, 467 – 472 (1998). 405. R. H. Wasserman, R. A. Corradino, C. S. Fullmer, and A. N. Taylor, Some aspects of vitamin D action; calcium absorption and the vitamin D-dependent calcium-binding protein. Vitam. Horm. 32, 299 – 324 (1974). 406. S. Christakos, J. D. Beck, and S. J. Hyllner, Calbindin-D28K. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 209–221. Academic Press, San Diego 1997. 407. M. Thomasset, Calbindin-D9K. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 223 – 232. Academic Press, San Diego, 1997. 408. S. Kowarski, L. A. Cowen, M. T. Takahashi, and D. Schachter, Tissue distribution and vitamin D dependence of IMCAL in the rat. Am. J. Physiol. 253, G411 – 419 (1987). 409. R. H. Wasserman, J. S. Chandler, S. A. Meyer, C. A. Smith, M. E. Brinkak, C. S. Fullmer, J. T. Penniston, and R. Kumar, Intestinal calcium transport and calcium extrusion processes at the basolateral membrane. J. Nutr. 122, 662 – 671 (1992). 410. T. Suda, T. Shinki, and N. Takahashi, The role of vitamin D in bone and intestinal cell differentiation. Annu. Rev. Nutr. 10, 195 – 211 (1990). 411. R. P. Heaney, M. J. Barger-Lux, M. S. Dowell, T. C. Chen, and M. F. Holik, Calcium absorptive effects of vitamin D and its major metabolites. J. Clin. Endocrinol. Metab. 82, 4111 – 4116 (1997). 412. R. P. Heaney, Lessons for nutritional science from vitamin D. Am. J. Clin. Nutr. 69, 825 – 826 (1999). 413. B. L. Riggs, H. W. Wahner, L. D. Melton, L. S. Richelson, H. L. Judd, and W. M. O’Fallon, Dietary calcium intake and rates of bone loss in women. J. Clin. Invest. 80, 979 – 982 (1987). 414. B. P. Halloran and A. A. Portale, Vitamin D metabolism: The effects of aging. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 541 – 554. Academic Press, San Diego, 1997. 415. H. J. Armbrecht, M. A. Boltz, S. Christakos, and M. E. Bruns, Capacity of 1,25-dihydroxyvitamin D to stimulate expression of calbindin D changes with age in the rat. Arch. Biochem. Biophys. 352, 159 – 164 (1998). 416. P. R. Ebeling, M. E. Sandgren, E. P. DiMagno, A. W. Lane, H. F. DeLuca, and B. L. Riggs, Evidence of an age-related decrease in
CHAPTER 9 Vitamin D
417.
418.
419.
420.
421.
422.
423.
424.
425.
426.
427.
428.
429.
430.
431.
432.
433.
intestinal responsiveness to vitamin D: relationship between serum 1,25-dihydroxyvitamin D3 and intestinal vitamin D receptor concentrations in normal women. J. Clin. Endocrinol. Metab. 75, 176 – 182 (1992). M. J. Barger-Lux, R. P. Heaney, S. J. Lanspa, J. C. Healy, and H. F. DeLuca, An investigation of sources of variation in calcium absorption efficiency [published erratum appears in J. Clin. Endocrinol. Metab. 80(7), 2068 (1995)]. J. Clin. Endocrinol. Metab. 80, 406 – 411 (1995). E. M. Colin, G. J. Van Den Bemd, M. Van Aken, S. Christakos, H. R. De Jonge, H. F. Deluca, J. M. Prahl, J. C. Birkenhager, C. J. Buurman, H. A. Pols, and J. P. Van Leeuwen, Evidence for involvement of 17beta-estradiol in intestinal calcium absorption independent of 1,25-dihydroxyvitamin D3 level in the Rat. J. Bone Miner. Res. 14, 57 – 64 (1999). M. J. Favus and F. L. Coe, Idiopathic hypercalciuria and nephrolithiasis. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 867 – 881. Academic Press, San Diego 1997. J. Yao, P. Kathpalia, D. A. Bushinsky, and M. J. Favus, Hyperresponsiveness of vitamin D receptor gene expression to 1,25-dihydroxyvitamin D3: A new characteristic of genetic hypercalciuric stoneforming rats. J. Clin. Invest. 101, 2223 – 2232 (1995). A. M. Parfitt, Vitamin D and the pathogenesis of rickets and osteomalacia. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 645 – 662. Academic Press, San Diego, 1997. B. E. Kream, M. Jose, S. Yamada, and H. F. DeLuca, A specific high-affinity binding macromolecule for 1,25-dihydroxyvitamin D3 in fetal bone. Science 197, 1086 – 1088 (1977). T. L. Chen, M. A. Hirst, and D. Feldman, A receptor-like binding macromolecule for 1,25-dihydroxycholecalciferol in cultured mouse bone cells. J. Biol. Chem. 254, 7491 – 7494 (1979). J. E., Aubin and J. N. M. Heersche, Vitamin D and osteoblasts. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 313 – 328. Academic Press, San Diego, 1997. B. E. Kream and A. C. Lichtler, Vitamin D regulation of type I collagen synthesis and gene expression in bone. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 201 – 207. Academic Press, San Diego, 1997. L. G., Raisz, C. L. Trummel, M. F. Holick, and H. F. DeLuca, 1,25dihydroxycholecalciferol: A potent stimulator of bone resorption in tissue culture. Science 175, 768 – 769 (1972). T. Suda and N. Takahashi, Vitamin D and osteoclastogenesis. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 329 – 340. Academic Press, San Diego, 1997. N. J. Horwood, J. Elliott, T. J. Martin, and M. T. Gillespie, Osteotropic agents regulate the expression of osteoclast differentiation factor and osteoprotegerin in osteoblastic stromal cells. Endocrinology 139, 4743 – 4746 (1998). K. W. Colston and D. Feldman, Demonstration of 1, 25-dihydroxycholecalciterol cytoplasmic receptor-like binder in mouse kidney cytosol. J. Clin. Endocrinol. 49, 798 – 800 (1979). S. D. Antrobus and M. R. Walters, 1,25-Dihydroxyvitamin D effects in the kidney: Induction of calmodulin binding proteins. Biochem. Biophys. Res. Commun. 185, 636 – 640 (1992). R. Kumar, Vitamin D and the kidney. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 275 – 292. Academic Press, San Diego, 1997. J. Silver and T. Naveh-Many, Vitamin D and the parathyroid glands. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 353 – 367. Academic Press, San Diego, 1997. N. Fokuda, H. Tanaka, Y. Tominaga, M. Fukagawa, K. Kurokawa, and Y. Seino, Decreased 1,25-dihydroxyvitamin D3 receptor density is associated with a more severe form of parathyroid hyperplasia in chronic uremic patients. J. Clin. Invest. 92, 1436 – 1443 (1993).
301 434. J. P. T. M. Van Leeuwen and H. A. P. Pols, Vitamin D: Anticancer and differentiation. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 1089 – 1105. Academic Press, San Diego, 1997. 435. S. D. Cramer, D. M. Peehl, M. G. Edgar, S. T. Wong, L. J. Deftos, and D. Feldman, Parathyroid hormone-related protein (PTH-P) is an epidermal growth factor-regulated secretory product of human prostatic epithelial cell. Prostate 29, 20 – 29 (1996). 436. K. El Abdaimi, V. Papavasiliou, S. A. Rabbani, J. S. Rhim, D. Goltzman, and R. Kremer, Reversal of hypercalcemia with the vitamin D analogue EB1089 in a human model of squamous cancer. Cancer Res. 59, 3325 – 3328 (1999). 437. R. F. Gagel and S. Peleg, Thyroidal C cells and medullary thyroid carcinoma as target tissues for vitamin D action. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 369 – 377. Academic Press, San Diego, 1997. 438. D. D. Bikle, Clinical counterpoint: Vitamin D: new actions, new analogs, new therapeutic potential. Endocr. Rev. 13, 765 – 784 (1992). 439. S. C. Manolagas, X. P. Yu, G. Girasole, and T. Bellido, Vitamin D and the hematolymphopoietic tissue: A 1994 update. Semin. Nephrol. 14, 129 – 143 (1994). 440. M. Hewison and J. L. H. O’Riordan, Immunomodulatory and cell differentiation, effects of vitamin D. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 447 – 462. Academic Press, San Diego, 1997. 441. R. Baier, A. Grauer, M. Lazaretti-Castro, R. Ziegler, and F. Raue, Differential effects of 1,25-dihydroxyvitamin D3 on cell proliferation and calcitonin gene expression. Endocrinology 135, 2006 – 2011 (1994). 442. P. H. Itin, M. R. Pittelkow, and R. Kumar, Effects of vitamin D metabolites on proliferation and differentiation of cultured human epidermal keratinocytes grown in serum-free or defined culture medium. Endocrinology 135, 1793 – 1798 (1994). 443. G. P. Studzinski, J. A. McLane, and M. R. Uskokovic, Signaling pathways for vitamin D-induced differentiation: Implications for therapy of proliferative and neoplastic diseases. Crit. Rev. Eukaryot. Gene Expr. 3, 279 – 312 (1993). 444. J. Welsh, Vitamin D compounds as potential therapeutics for estrogen-independent breast cancer. Nutrition 13, 915 – 917 (1997). 445. R. Morosetti and H. P. Koeffler, Vitamin D compounds and analogs: Effects on normal and abnormal hematopoiesis. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 1155 – 1166. Academic Press, San Diego, 1997. 446. W. M. Tong, H. Hofer, A. Ellinger, M. Peterlik, and H. S. Cross, Mechanism of antimitogenic action of vitamin D in human colon carcinoma cells: Relevance for suppression of epidermal growth factor-stimulated cell growth. Oncol. Res. 11, 77 – 84 (1999). 447. S. E. Blutt and N. L. Weigel, Vitamin D and prostate cancer. Proc. Soc. Exp. Biol. Med. 221, 89 – 98 (1999). 448. M. Liu M. H. Lee, M. Cohen, M. Bommakanti, and L. P. Freedman, Transcriptional activation of the Cdk inhibitor p21 by vitamin D3 leads to the induced differentiation of the myelomonocytic cell line U937. Genes Dev. 10, 142 – 153 (1996). 449. T. A. Brasitus and M. D. Sitrin, Chemoprevention of colon cancer by vitamin D3 and its metabolites/analogs. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 1141 – 1154. Academic Press, San Diego 1997. 450. C. F. Garland, F. C. Garland, and E. D. Gorham, Can colon cancer incidence and death rates be reduced with calcium and vitamin D? Am. Clin. Nutr. 54 (Suppl. 1), 193S – 201S (1991). 451. C. F. Garland, G. W. Comstock, F. C. Garland, K. J. Helsing, E. K. Shaw, and E. D. Gorham, Serum 25-Hydroxyvitamin D and colon cancer: Eight-year prospective study. Lancet 2, 1176 – 1178 (1989). 452. J. A. Eisman, D. H. Barkla, and P. J. Tutton, Suppression of in vivo growth of human cancer solid tumor xenografts by 1,25-dihydroxyvitamin D3. Cancer Res. 47, 21 – 25 (1987).
302 453. S. R. Evans, A. M. Schwartz, E. I. Shchepotin, M. Uskokovic, and I. B. Shchepotin, Growth inhibitory effects of 1,25-dihydroxyvitamin D3 and its synthetic analogue, 1,25-dihydroxy-16-ene-23yne-26,27hexafluoro-19-nor-cholecalcifero1(Ro 25-6760), on a human colon cancer xenograft. Clin. Cancer Res. 4, 2869 – 2876 (1998). 454. M. A. Anzano, J. M. Smith, M. R. Uskokovic, C. W. Peer, L. T. Mullen, J. J. Letterio, M. C. Welsh, M. W. Shrader, D. L. Logsdon, C. L. Driver, C. C. Brown, A. B. Roberts, and M. B. Sporn, (1994). 1-,25-Dihydroxy-16-ene-23-yne-26,27-hexafluorocholecalciferol (Ro24-5531), a new deltanoid (vitamin D analogue) for prevention of breast cancer in the rat. Cancer Res. 54, 1653 – 1656 (1994). 455. E. Elstner, D. Heber, and H. P. Koeffler, 20-Epi-vitamin D3 analogs: Potent modulators of proliferation and differentiation of breast cancer cell lines in vitro. Adv. Exp. Med. Biol. 399, 53 – 70. 456. T. Vink-van Wijngaarden, H. A. Pols, C. J. Buurman, J. C. Birkenhager, and J. P. van Leeuwen, Inhibition of insulin- and insulin-like growth factor-I-stimulated growth of human breast cancer cells by 1,25-dihydroxyvitamin D3 and the vitamin D3 analogue EB1089. Eur. J. Cancer 32A, 842 – 848 (1996). 457. M. Koike, E. Elstner, M. J. Campbell, H. Asou, M. Uskokovic, N. Tsuruoka, and H. P. Koeffler, 19-nor-hexafluoride analogue of vitamin D3: A novel class of potent inhibitors of proliferation of human breast cell lines. Cancer Res. 57, 4545 – 4550 (1997). 458. G. Wu, R. S. Fan, W. Li, V. Srinivas, and M. G. Brattain, Regulation of transforming growth factor- type II receptor expression in human breast cancer MCF-7 cells by vitamin D3 and its analogues. J. Biol. Chem. 273, 7749 – 7756 (1998). 459. K. W. Colston, C. M. Perks, S. P. Xie, and J. M. Holly, Growth inhibition of both MCF-7 and Hs578T human breast cancer cell lines by vitamin D analogues is associated with increased expression of insulin-like growth factor binding protein-3. J. Mol. Endocrinol. 20, 157 – 162 (1998). 460. G. G. Schwartz and B. S. Hulka, Is vitamin D deficiency a risk factor for prostate cancer? Anticancer Res. 10, 1307 – 1311 (1990). 461. C. L. Hanchette and G. G. Schwartz, Geographic patterns of prostate cancer mortality. Evidence for a protective effect of ultraviolet radiation. Cancer 70, 2861 – 2869 (1992). 462. E. H. Corder, H. A. Guess, B. S. Hulka, G. D. Friedman, M. Sadler, R. T. Vollmer, B. Lobaugh, M. K. Drezner, J. H. Vogelman, and N. Orentreich, Vitamin D and prostate cancer: A prediagnostic study with stored sera. Cancer Epidemiol. Biomarkers Prev. 2, 467 – 447 (1993). 463. G. J. Miller, G. E. Stapleton, J. A. Ferrara, M. S. Lucia, S. Pfister, T. E. Hedlund, and P. Upadhya, The human prostatic carcinoma cell line LNCaP expresses biologically active, specific receptors for 1,25-dihydroxyvitamin D3. Cancer Res. 52, 515 – 520 (1992). 464. S. E. Blutt, T. J. McDonnell, T. C. Polek, and N. L. Weigel, Calcitriol-induced apoptosis in LNCaP cells is blocked by overexpression of Bcl-2. Endocrinology 141, 10 – 17 (2000). 465. C. Gross, T. Stamey, S. Hancock, and D. Feldman, Treatment of early recurrent prostate cancer with 1,25-dihydroxyvitamin D3 (calcitriol) [published erratum appears in J. Urol. 160(3 Pt 1), 840]. J. Urol. 159, 2035 – 2039; discussion 2039 – 2040 (1998). 466. D. Trump, P. A. Hershberger R. A. Modzelewski, R. J. Bernardi, J. Muindi, R. M. Rueger, W. Yu, M. J. Egorin, K. E. Blum, and C. S. Johnson. Clinical development of calcitriol and calcitriol analogues as anticancer agents. In “Vitamin D endocrine system: Structural, biological, genetic and clinical aspects” (A. W. Norman, R. Bouillon, and M. Thomasset, eds.) pp. 469 – 475. University of California, Printing and Reprographics, Riverside, 2000. 467. P. H. Reitsma, P. G. Rothberg, S. M. Astrin, J. Trial, Z. Bar-Shavit, A. Hall, S. L. Teitelbaum, and A. J. Kahn, Regulation of myc gene expression in HL-60 leukaemia cells by a vitamin D metabolite. Nature 306, 492 – 494 (1983). 468. Y. Honma, M. Hozumi, E. Abe, K. Konno, M. Fukushima, S. Hata, Y. Nishii, H. F. DeLuca, and T. Suda, 1,25-Dihydroxyvitamin D3
FELDMAN, MALLOY, AND GROSS
469.
470.
471.
472.
473.
474.
475.
476.
477.
478.
479.
480.
481.
482.
483.
484.
485.
and 1-hydroxyvitamin D3 prolong survival time of mice inoculated with myeloid leukemia cells. Proc. Natl. Acad. Sci. USA 80, 201 – 204 (1983). R. Munker, W. Zhang, E. Elstner, and H. P. Koeffler, Vitamin D analogs, leukemia and WAF1. Leuk. Lymphoma 31, 279 – 284 (1998). C. Mathieu, K. Casteels, and R. Bouillon, Vitamin D and diabetes. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 1183 – 1196. Academic Press, San Diego, 1997. J. M. Lemire, D. C. Archer, and G. S. Reddy, (1994). 1,25Dihydroxy-24-OXO-16ene-vitamin D3, exerts immunosuppressive activity equal to its parent without causing hypercalcemia in vivo. Endocrinology 135, 2818 – 2821 (1994). C. Mathieu, K. Casteels, M. Waer, J. Laureys, D. Valckx, and R. Bouillon, Prevention of diabetes recurrence after syngeneic islet transplantation in NOD mice by analogues of 1,25(OH)2D3 in combination with cyclosporin A: mechanism of action involves an immune shift from Th1 to Th2. Transplant Proc. 30, 541 (1998). A. Verstuyf, S. Segaert, L. Verlinden, K. Casteels, R. Bouillon, and C. Mathieu, Recent developments in the use of vitamin D analogues. Curr. Opin. Nephrol. Hypertens. 7, 397 – 403 (1998). D. Feldman, T. Chen, M. Hirst, K. Colston, M. Karasek, and C. Cone, Demonstration of 1,25-dihydroxyvitamin D3 receptors in human skin biopsies. J. Clin. Endocrinol. Metab. 51, 1463 – 1465 (1980). K. Colston, J. M. Colston, and D. Feldman, 1,25-Dihydroxyvitamin D3 and malignant melanoma: The presence of receptors and inhibition of cell growth in culture. Endocrinology 108, 1083 – 1086 (1981). S. J. Marx, M. M. Bliziotes, and M. Nanes, Analysis of the relation between alopecia and resistance to 1,25-dihydroxyvitamin D. Clin. Endocrinol. 25, 373 – 381 (1986). Y. Sakai and M. B. Demay, Evaluation of keratinocyte proliferation in vitamin D receptor knockout mice. Endocrinology 141, 2043 – 2049 (2000). E. L. Smith, S. H. Pincus, L. Donovan, and M. F. Holick, A novel approach for the evaluation and treatment of psoriasis. Oral or topical use of 1,25-dihydroxyvitamin D3 can be a safe and effective therapy for psoriasis. J. Am. Acad. Dermatol. 19, 516 – 528 (1988). W. E. Stumpf, H. J. Bidmon, L. Li, C. Pilgrim, A. Bartke, A. Mayerhofer, and C. Heiss, Nuclear receptor sites for vitamin D-soltriol in midbrain and hindbrain of Siberian hamster (Phodopus sungorus) assessed by autoradiography. Histochemistry 98, 155 – 164 (1992). I. M. Musiol, W. E. Stumpf, H. J. Bidmon, C. Heiss, A. Mayerhofer, and A. Bartke, Vitamin D nuclear binding to neurons of the septal, substriatal and amygdaloid area in the Siberian hamster (Phodopus sungorus) brain. Neuroscience 48, 841 – 848 (1992). M. K. Sutherland, M. J. Somerville, L. K. Yoong, C. Bergeron, M. R. Haussler, and D. R. McLachlan, Reduction of vitamin D hormone receptor mRNA levels in Alzheimer as compared to Huntington hippocampus: Correlation with calbindin-28k mRNA levels. Brain Res. Mol. Brain Res. 13, 239 – 250 (1992). S. Carswell, Vitamin D in the nervous system: Actions and therapeutic potential. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 1197 – 1211. Academic Press, San Diego 1997. J. Sonnenberg, V. N. Luine, L. C. Krey, and S. Christakos, 1,25Dihydroxyvitamin D3 treatment results in increased choline acetyltransferase activity in specific brain nuclei. Endocrinology 118, 1433 – 1439 (1986). D. Wion, D. MacGrogan, I. Neveu, F. Jehan, R. Houlgatte, and P. Brachet, 1,25-Dihydroxyvitamin D3 is a potent inducer of nerve growth factor synthesis. J. Neurosci. Res. 28, 110 – 114 (1991). I. M. Musiol and D. Feldman, 1,25-dihydroxyvitamin D3 induction of nerve growth factor in L929 mouse fibroblasts: Effect of vitamin D receptor regulation and potency of vitamin D3 analogs. Endocrinology 138, 12 – 18 (1997).
CHAPTER 9 Vitamin D 486. M. S. Saporito, E. R. Brown, K. C. Hartpence, H. M. Wilcox, J. L. Vaught, and S. Carswell, Chronic 1,25-dihydroxyvitamin D3-mediated induction of nerve growth factor mRNA and protein in L929 fibroblasts and in adult rat brain. Brain Res. 633, 189 – 196 (1994). 487. P. A. De Viragh, D. Wolfer, H. A. Lipp, and M. R. Celio, (eds.), “Behavioral Changes in Chronically D-Hyper-vitaminotic Animals.” Walter de Gruyter, New York, 1988. 488. Y. Sato, T. Asoh, and K. Oizumi, High prevalence of vitamin D deficiency and reduced bone mass in elderly women with Alzheimer’s disease. Bone 23, 555 – 557 (1998). 489. B. P. Halloran, Is 1,25-dihydroxyvitamin D required for reporduction? Proc. Soc. Exp. Biol. Med. 191, 227 – 232 (1989). 490. A. Stephanou, R. Ross, and S. Handwerger, Regulation of human placental lactogen expression by 1,25-dihydroxyvitamin D3. Endocrinology 135, 2651 – 2656 (1994). 491. H. Reichel, H. P. Koeffler, and A. W. Norman, The role of the vitamin D endocrine system in health and disease. N. Engl. J. Med. 320, 980 – 991 (1989). 492. R. Eastell and B. L. Riggs, Vitamin D and osteoporosis. In “Vitamin D” (D. Feldman, F. Glorieux, and J. W. Pike, eds.), pp. 695–711. Academic Press, San Diego, 1997. 493. R. Nuti, E. Bonucci, D. Brancaccio, J. C. Gallagher, C. Gennari, G. Mazzuoli, M. Passeri, and P. Sambrook, The role of calcitriol in the treatment of osteoporosis. Calcif. Tissue Int. 66, 239 – 240 (2000). 494. B. L. Riggs and L. J. Melton, Involutional osteoporosis. N. Engl. J. Med. 314, 1676 – 1686 (1986). 495. O. Sahota, T. Masud, P. San, and D. J. Hosking, Vitamin D insufficiency increases bone turnover markers and enhances bone loss at the hip in patients with established vertebral osteoporosis. Clin. Endocrinol. (Oxf.) 51, 217 – 221 (1999). 496. M. C. Chapuy, M. E. Arlot, F. Duboeuf, J. Brun, B. Crouzet, S. Arnaud, P. D. Delmas, and P. J. Meunier, Vitamin D3 and calcium to prevent hip fractures in the elderly women. N. Engl. J. Med. 327, 1637 – 1642 (1992).
303 497. M. W. Tilyard, G. F. Spears, J. Thomson, and S. Dovey, Treatment of postmenopausal osteoporosis with calcitriol or calcium. N. Engl. J. Med. 326, 357 – 362 (1992). 498. J. C. Gallagher, Prevention of bone loss in postmenopausal and senile osteoporosis with vitamin D analogues. Osteoporos. Int. 1, 172 – 175 (1993). 499. J. S. Adams, V. Kantorovich, C. Wu, M. Javanbakht, and B. W. Hollis, Resolution of vitamin D insufficiency in osteopenic patients results in rapid recovery of bone mineral density. J. Clin. Endocrinol. Metab. 84, 2729 – 2730 (1999). 500. K. H. Lau and D. J. Baylink, Vitamin D therapy of osteoporosis: Plain vitamin D therapy versus active vitamin D analog (D-hormone) therapy. Calcif. Tissue Int. 65, 295 – 306 (1999). 501. S. V. Hughes, E. Robinson, R. Bland, H. M. Lewis, P. M. Stewart, and M. Hewison, 1,25-dihydroxyvitamin D3 regulates estrogen metabolism in cultured keratinocytes. Endocrinology 138, 3711 – 3718 (1997). 502. H. A. Bischoff, H. B. Stahelin, N. Urscheler, R. Ehrsam, R. Vonthein, P. Perrig-Chiello, A. Tyndall, and R. Theiler, Muscle strength in the elderly: Its relation to vitamin D metabolites. Arch. Phys. Med. Rehabil. 80, 54 – 58 (1999). 503. M. S. Stein, J. D. Wark, S. C. Scherer, S. L. Walton, P. Chick, M. Di Carlantonio, J. D. Zajac, and L. Flicker, Falls relate to vitamin D and parathyroid hormone in an Australian nursing home and hostel. J. Am. Geriatr. Soc. 47, 1195 – 1201 (1999). 504. S. Amin, M. P. LaValley, R. W. Simms, and D. T. Felson, The role of vitamin D in corticosteroid-induced osteoporosis: A meta-analytic approach. Arthritis Rheum. 42, 1740 – 1751 (1999). 505. C. H. Jacobus, M. F. Holick, Q. Shao, T. C. Chen, I. A. Holm, J. M. Kolodny, G. E. Fuleihan, and E. W. Seely, Hypervitaminosis D associated with drinking milk. N. Engl. J. Med. 326, 1173 – 1177 (1992). 506. J. S. Adams and G. Lee, Gains in bone mineral density with resolution of vitamin D intoxication. Ann. Intern. Med. 127, 203 – 206 (1997).
CHAPTER 10
Regulation of Bone Cell Function by Estrogens BARRY S. KOMM AND PETER V. N. BODINE Women’s Health Research Institute, Wyeth-Ayerst Research, Radnor, Pennsylvania 19087
I. II. III. IV. V. VI.
VII. Estrogenic Responses in Bone Cells VIII. Estrogen-Related Receptor- and Osteopontin Gene Expression IX. Nongenomic Actions of Estrogens in Bone Cells X. Summary References
Introduction What Is an Estrogen? Estrogen Receptors ER and ER Knockout Mice Estrogens and Bone: Overview Estrogen Receptors in Bone Cells
I. INTRODUCTION
number of molecules, both steroidal and nonsteroidal in nature. Endogenous vertebrate estrogens are 18 carbon, four-ringed structures [7] (Fig. 1, see also color plate) derived from cholesterol. The most abundant estrogenic steroids in humans include estrone (E1), 17-estradiol (E2), and estriol (E3). In addition, there is an array of estrogenic metabolites that display variable estrogenic activity as well as several well-characterized B-ring-saturated estrogens [8]. Beyond these classic estrogens, several estrogenic substances obtained from plant sources (phytoestrogens), synthetic estrogens (i.e., diethylstibestrol), and a relatively large group of xenobiotics (e.g., DDT, biphenols) have been classified as estrogens. Finally, there is a growing number of molecules, originally classified as antiestrogens, but currently undergoing reclassification (based on their biological activity) that are represented by a diverse set of chemical structures (Fig. 1) and are collectively referred to as selective estrogen receptor modulators (SERMs) [9,10].
Estrogens and their diverse effects on bone remodeling are perhaps less well characterized than one would predict. The positive impact of estrogens on the skeleton has been well known and documented since the early 1940s, and estrogen remains the primary form of osteoporosis treatment in the world [1 – 6]. However, the mechanisms by which estrogens regulate bone remodeling and therby protect the skeleton continue to undergo intense evaluation.
II. WHAT IS AN ESTROGEN? Before discussing the role estrogens play in bone, it is important to define what an estrogen is and the abundance of basic science that describes the multiple facets of estrogenic activity. Estrogens are represented by a large
OSTEOPOROSIS, SECOND EDITION VOLUME 1
305
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
306
KOMM AND BODINE
FIGURE 1
Structures of a variety of compounds that can be classified as members of the family of estrogens. In red are classical steroidal estrogens represented by the three predominant circulating estrogens detected in mammals. In pink is the nonsteroidal and potent estrogen diethylstilbestrol. In green are two phytoestrogens, both nonsteroidal but functionally characterized as estrogens. In black is the potent steroidal antiestrogen ICI182780. This compound has been described as a pure estrogen receptor antagonist; however, its characterization is still under examination. At the bottom of the figure in blue are three generations of selective estrogen receptor modulators (SERMs). Originally referred to, like ICI182780, as antiestrogens, this group of compounds exhibits mixed functional activity all seemingly transduced by estrogen receptors. What all these compounds (and there are hundreds more) have in common is that they bind to the estrogen receptors and functionally affect estrogen receptor activity. In some cases the effects are only as agonists, or as relatively potent antagonists, but most commonly as mixed function ligands with their effects related to the cellular target and the specific genes that are being monitored. (See also color plate.)
III. ESTROGEN RECEPTORS A. Members of the Nuclear Receptor Superfamily What this assortment of compounds has in common is that they exert their function via a single class of nuclearlocalized proteins: estrogen receptors. The two currently recognized members of the estrogen receptor family are referred to as estrogen receptor (ER) [11 – 13] and estrogen receptor (ER) [14,15]. Estrogen receptors belong to a large superfamily (Table 1) of nuclear acting receptors represented by members that bind the classical group of steroid hormones, including glucocorticoids, progestins, androgens and mineralocorticoids. In addition to these, other members include the receptors for Vitamin D, retinoids, thyroid hormones, oxysterols, farnesol, prostanoids, and ecdysone. Well over 50 members of this superfamily remain for which a ligand has not been identified, and they are referred to as orphan nuclear receptors [16 – 19].
Steroid receptors share many common features. Structurally, this group of proteins can be dissected into discrete regions with different functions [11,20]. The regions are designated simply as A,B,C,D,E, and F (Fig. 2, see also color plate). The unifying feature characteristic of each nuclear receptor family member is a two zinc finger domain (region C) associated with DNA binding (DNA-binding TABLE 1
Members of the Steroid/Thyroid/Retinoid Nuclear Receptor Superfamily
Androgen
Estrogen (,)
Glucocorticoid
Mineralocorticoid
Progesterone (A,B)
Thyroid hormone (,)
Vitamin D
Retinoic acid (,,)
Retinoid X receptor (,,)
Peroxisome proliferator activating receptor (,,)
Pregnane receptor
Ecdysone
Orphan receptors (50)
307
CHAPTER 10 Regulation of Bone Cell Function by Estrogens
B. Coactivators and Corepressors
FIGURE 2
Schematic structure of nuclear hormone receptors. This family of receptors, which includes the estrogen receptors, can be represented as cassettes with interchangeable units. The A/B domain at the N terminus contains at least one transactivation domain (AF-1) that is ligand independent. The A/B domain is adjacent to the C domain, which represents the DNA-binding domain containing two cysteine loops that each intercalate one zinc molecule to form DNA-binding fingers. This domain is highly conserved among the family members. The D domain is much less well defined but has been described as the hinge domain and contains a nuclear localization signal; however, other sites in the ER have been linked to nuclear localization outside of the D domain. The E domain represents the ligand-binding domain and is not as conseved as the C domain. Additionally, embedded within the ligand binding domain is another transactivation domain: AF-2, which is ligand dependent, unlike AF-1. The F domain of the receptor does not display any clear function; however, removal of small parts of this domain can affect receptor function (both ligand binding and transactivation). (See also color plate.)
domain=DBD). The receptors are DNA-binding proteins that interact with specific DNA sequences [21,22] [e.g., estrogen receptor element (ERE) and androgen receptor element (ARE)] via two cysteine-rich domains that intercalate zinc to form binding “fingers.” Sequence homology in this domain among members of this family is relatively high, and while there is amino acid disparity in the DBD, the cysteine residues can be aligned for all of the receptors, supporting their derivation from a common ancestral protein. The other domains are a ligand-binding domain (region D,E, and F), nuclear localization domain (D), and a hinge domain (D). In addition, two transactivation domains, AF-1, and AF-2, reside in the amino (A/B)and carboxy (e)- terminal portions of the protein, respectively [23]. The mechanism through which information is transduced from the ligand by the receptor has been the subject of intense research since the 1960s. It has become clear that ligand binding to the estrogen receptor initiates a number of processes. Ligand binding produces a change in conformtion, which for several members of the family, including the ER, appears to begin with the displacement of heat shock proteins [24,25]. Subsequently, two liganded estrogen receptors dimerize [26], are biochemically modified (e.g., acetylation and phosphorylation) [27], and then bind to specific DNA sequences. In this simple model, the “activated” ER complex can act as an enhancer or repressor of gene transcriptional activity [28,29].
The model for ER regulation of gene transcription has gained complexity with the discovery of several proteins that interact with the estrogen receptor, as well as other members of the steroid hormone receptor superfamily. These proteins, referred to as coregulators, or comodulators are represented by both coactivators [30,31] and corepressors [32,33]. Several coregulators have been identified and are represented by a diverse group of proteins and RNA [34]. Not unlike the nuclear receptors, several of these proteins contain specific regions associated with independent function [35], including histone acetylation, CREB binding protein interaction domains, and a nuclear receptor interaction domain (NRID) [36,37]. The corepressors contain histone deacetylase domains [33]. Within the NRID, one or more LXXLL motifs interact with the estrogen receptor and other members of the superfamily [38,39]. This binding has been verified by co-crystalization of the ER ligand-binding domain with a small peptide containing an LXXLL domain from the coactivator protein, GRIP 1 (SRC-2) [40], and has been shown to interact specifically with a region of the receptor represented by helices 3,4,5, and 12 [40,41]. The interaction of these coactivators via the NRID has also been demonstrated to be associated with increased transcriptional activity of the ER [42]. The transcriptional complex is composed of an array of proteins, which would include several coactivators whose roles may vary; however, some definitely serve to bridge the enhancer region of ER binding on DNA with the basal transcriptional machinery. The vitamin D receptor interacting protein/thyroid receptor activator protein (DRIP/TRAP) complex of proteins (10 proteins) plays the dual role of transcriptional activation and bridging the transcriptional enhancer complex with the basal transcriptional complex. Not all proteins in the DRIP complex have been shown to interact with the ER, and this complex does not play only a functional role in transcriptional enhancement with nuclear steroid hormone receptors [43 – 45]. Additionally, this is not to say that the estrogen receptor cannot interact directly with proteins associated with the basal transcriptional machinery, as has been suggested for the vitamin D receptor.
C. Alternate Pathways for Estrogenic Activity The model just described for ER activity is characterized as a multiple series of steps initiated by ligand binding (Fig. 3, see also color plate). However, it has become clear that ERs function through other mechanistic pathways to affect various physiologic functions in both a ligand dependent and independent fashion. Data demonstrate that the ER can be activated by growth factors working through protein kinase A and C or in concert with these kinases [46,47].
308
KOMM AND BODINE
FIGURE 3
Estrogens can affect cell function through several pathways. Classically, an estrogen diffuses through the plasma membrane to interact with a nuclear localized receptor (ER or ER or both). The binding of ligand results in a rapid conformational change in the receptor and other biochemical modifications, such as phosphorylation and acetylation. Associated with the changes in conformation are interactions and displacements of proteins from the class of coactivators and corepressors. These proteins form a transcriptional complex linking the receptor DNA complex to the basal transcriptional machinery, resulting in changes in gene transcriptional activity. The liganded receptor can also interact with the AP-1 and SF-1 complex to regulate gene transcription indirectly (i.e., not through estrogen receptors directly binding to DNA). The liganded receptor also interacts with the NF-B protein complex to inhibit gene activity regulated by NF-B. The unliganded ER can be activated by growth factors such as EGF and IGF-1. The stimulated phosphorylation of the ER is sufficient to generate genetic responses regulated by ER without any ligand-induced change in conformation. Other data also support the contention that a plasma membrane-associated ER exists and can transduce information via second messengers. (See also color plate.)
These kinases phosphorylate the receptor (also a result of ligand binding) predominantly at serine 118 [48]. This phosphorylation appears to be sufficient to activate the receptors to then recruit the appropriate coactivators, bind to DNA, and enhance transcriptional activity. This has been shown to occur in a human breast tumor cell model (MCF7 cells) [48,49], a rat osteosarcoma cell line overexpressing ER [50], and a human ovarian adenocarcinoma cell line, BG-1 [51]. In these cell models, treatment with IGF-1, EGF, and other activators of the A and C kinases activates the ER in these cells, as shown by transactivation of an ERE-driven promoter construct. These promoters are minimally active and require functional ER to detect transcriptional activity. It is important to note that in these models, ER antagonists such as ICI-164384 block the stimulation
whether activation is attributed to an estrogen such as 17 estradiol or to kinase activators, supporting the contention that the effects are ER dependent. Another classic model to evaluate estrogen action is the rodent uterine response to estrogens. Here, relatively low doses of an estrogen, typically 17- estradiol, stimulate an increase in uterine wet weight coupled with various uterine histological changes and the expression of several marker genes. If the animals are estrogen deficient (i.e., ovaries removed) and treated with epidermal growth factor, the uterine response in many aspects is quite similar to that seen with 17- estradiol treatment alone [52]. Data utilizing the ER knockout mouse has verified the estrogen receptor requirement for this response [46]. Additionally, it has also been shown that 17-estradiol can activate the MAPK
309
CHAPTER 10 Regulation of Bone Cell Function by Estrogens
pathway, resulting in a rapid alteration in intracellular calcium [53]. Because the estrogen receptor is required for these effects to be manifested, these data strongly support the conclusion that the ER can function in ways other than its role as a transcription factor in the nucleus. While growth factors like IGF-1 and EGF have been demonstrated to activate the ER, it has also been shown that estrogens can inhibit the response associated with various interleukins, especially IL-1, TNF-, and IL-6. This has specific relevance to the role estrogens play in the skeleton. IL-6 and TNF- function via (NF-B) activation and the subsequent interaction of this complex with specific DNA sequences. The skeletal effects of these cytokines are negative (i.e., they increase resorption resulting in a loss of bone mass). ERs interact with NF-B in a ligand-dependent manner to inhibit its activity [54 – 56]. The exact mechanism for suppression is not known, but may be through interference of ER with NF-B binding to DNA, or ER inhibition of appropriate NF-B activation. ER:NF-B interactions and involvement in bone will be discussed in more detail later in this chapter. In addition to the inhibitory effect that the liganded ER has on NF-B, ERs also affect signaling through the AP-1 (JUN-FOS) [57], Sp1 [58], and SF-1 [59,60] transcriptional complexes. Like NF-B, this involves protein:protein interaction. For the AP-1 complex, the effect is compound specific. A compound such as tamoxifen, known as an ER antagonist, functions as a relatively potent agonist through AP-1 and the ER. Activation by tamoxifen is through the AF-1 domain in the N-terminal (A/B) region of the ER [57]. This type of response cannot be evoked with a more classical ERE-driven promoter by tamoxifen. These alternate pathways may partially explain the mixed functional activity that has been reported for SERMs (see Section IV). The final and, at this time, most controversial mechanism through which estrogens can elicit a response appears to involve a process that does not utilize nuclear-localized estrogen receptors or cytosolic ER. Instead, the response to compounds that would be classified as estrogenic appears to function via plasma membrane-associated estrogen receptors [61 – 64]. This would account for physiological responses to estrogens that occur rapidly and would be considered unlikely to be mediated through a nuclear receptor-mediated transcriptional process. Various examples representing different mechanistic explanations include plasma membrane residing ER that bind ligand and transmit information via G-coupled protein receptors [64]; for example, an associated increase in cAMP in response to 17-estradiol has been demonstrated. A twist on this paradigm involves sex hormone-binding globulin (SHBG), a serum protein that transports sex steroid hormones and is capable of binding to plasma membrane receptors. Cells with attached SHBG treated with dihydrotestosterone or 17-estradiol exhibit changes in intracellular cAMP levels,
suggesting an alternative mechanism for 17-estradiol signaling, but only through unliganded SHBG [65]. Additionally, calcium transport has been shown to be directly influenced by estrogens at the level of the calcium channel [66]. Quite recently, the MaxiK channel associated with ion flux via a voltage-gated channel in bladder smooth muscle has been shown to be directly regulated by estrogens [67].
D. ER Estrogens can elicit a variety of physiological responses and, until 1996, it was believed that transduction of information occurred through one nuclear receptor protein (ER). However, as mentioned earlier, a second protein has been identified that also exhibits high-affinity binding for estrogens and dubbed estrogen receptor (ER) [14,15,68]. Its chromosome location differs from that of human ER (14 vs 6, respectively) [69]. The two transcripts differ in length with ER coding for a protein of 530 amino acids [70] and ER coding for a protein of 595 amino acids [71]. Additionally, their tissue distribution varies, especially in the brain, ovary, uterus, and prostate [72]. At this point in time, the functional role of ER remains to be proven. In vitro transcription assays have shown that ER, like ER, dimerizes and binds to DNA (specifically EREs). However, it has been shown that under appropriate conditions, ER heterodimerizes with ER, and the resulting complex binds to DNA more avidly than the ER homodimer [73]. However, the transcriptional activity of the heterodimer is similar to that of the ER homodimer, but differs from the ER homodimer. The affinity of 17-estradiol for the two receptors is essentially identical, but clearly under in vitro conditions, ER is a more effective activator of transcription than ER [70]. Another characteristic difference between these two receptors is their apparent variation in ligand affinity. Whereas 17-estradiol binding affinity is the same, another estrogen, the phytoestrogen genestein, shows a remarkable preference for ER (30-fold) [74]. The interaction of coactivators with these two proteins also differs. This information, coupled with the distinct tissue distribution and apparent differences in ligand preference, suggests that specific ligands may exist that activate one receptor preferentially over the other [75]. If this is the case, then it also seems quite possible that these compounds could be synthesized and specifically activate only one of the receptors. The pharmaceutical implications of this possibility are obvious.
E. Structure of ER and ER by X-Ray Crystallography Both ER and ER ligand-binding domains (LBDs) have been crystallized [40,76,77]. ER cocrystallized with DES,
310
KOMM AND BODINE
17-estradiol, and 4-OH tamoxifen demonstrates that these ligands generate two different conformations of the ER LBD. With a natural agonist, 17-estradiol, or a synthetic agonist, DES, the ligand fits snuggly into a pocket, and helix 12 (12 of 12 helices in LDB crystallized) appears to fold over the binding pocket [40]. With the SERM 4-OH tamoxifen, helix 12 no longer covers the binding pocket, and now shifts in position to a region that masks amino acids in helices 3,4, and 5. As it turns out, the hydrophobic surface created by those amino acids is critical for the interaction of members of the p160 coactivator family (SRC1,2, and 3) [40]. Indeed, transcriptional activation studies performed with these coactivators in the presence of various anti-estrogens reveal little to no activity, thereby supporting structural data and the importance of the AF-2 domain in estrogen receptor transactivation.
F. Tissue Selective Estrogens (SERMs) What has become clear recently is the fact that ERs are rather willing partners for a wide variety of ligands. This is unlike the other members of the steroid receptor superfamily, which at least, to date, demonstrate more stringent ligand-binding parameters. Compounds with rather diverse structures have been demonstrated to bind with high affinity to the estrogen receptor and exhibit various potencies depending on the end points evaluated. Classically, the targets of estrogen action were the uterus, breast, and liver. In the past two decades, it has also been shown that estrogens directly affect the skeleton, central nervous system (CNS), immune system, cardiovascular system, and the gastrointestinal tract. The discovery of ER has led to the inclusion of the prostate as an estrogen target tissue in males, along with some tissues common to both sexes (i.e., bone, cardiovascular, immune). Obviously, depending on the tissue, the genetic response to estrogens varies. There may be a group of genes that respond similarly in all tissues to a particular agonist, but the key end responses are most likely tissue selective as a result of the responsiveness of a specific set of genes. Thus, in the uterus, a collection of genetic end points can be quantitated that are distinct from the those of breast. This is a critical premise defining the role of tissue selective estrogens (or SERMs) and their clinical applications [8,10]. Perhaps all estrogens are selective and a change in nomenclature is in order. Nevertheless, one example of a tissue selective estrogen would be a compound that behaves as an ER agonist in the skeleton, but as an antagonist (no intrinsic activity but would antagonize estrogens) in the uterus. Tamoxifen, which was originally targeted for contraception, turned out to be a better antiestrogen on breast tissue and was developed as a treatment for hormone (estrogen) responsive breast cancer. As more data were generated using tamoxifen, it was seen to affect several other tissues besides the breast [78]. Some of the effects were positive or
desirable (estrogen agonist activity), such as on the skeleton and lipid profiles, whereas others were considered negative or undesirable such as the antagonist effect in the CNS and the agonist effect on the uterus [79 – 82]. How could this be? Clearly, all “tissue selective estrogens” do not behave identically. Because of structural diversity, their impact on ER function due to different receptor conformation varies [83] and, conceptually, this together with differences in the various target tissues must account for the differences in responses that are seen when comparing these compounds.
IV. ER AND ER KNOCKOUT MICE In an effort to define more clearly the physiologic role(s) of both ER and ER, knockout (KO) mice have been generated [84,85]. Neither KO is lethal and the phenotype exhibited by mice was not as predictable as anticipated. ERKO and ERKO (ER knockout) animals do not demonstrate a striking skeletal phenotype, suggesting that presence of either receptor suffices to maintain skeletal estrogen responsiveness. There is a small, but significant decrease in bone length in both sexes of ERKO animals. This is not seen in ERKOs. Bone mineral density is affected minimally in both KO strains [86]. Ovariectomy of either knockout results in osteopenia typical of wild-type mice and rats, supporting the fact that either receptor is capable of maintaining “normal” modeling in the mouse. One dramatic example of a human man who suffers from ER inactivation (a point mutation resulting in a premature stop condon) has been reported [87]. This man exhibits an overt phenotype where longitudinal bone growth has not terminated (no epiphyseal closure) and bone mineral density has been compromised. Although not published, it appears that this man expresses normal ER and normal androgen receptors. The patients skeletal phenotype is opposite that seen in mice lacking ER, which should warn us (once again) about extrapolating results from rodents to humans. Human data also, at least in this man, suggest that ER and androgen receptors are not sufficient to overcome the inactivation of ER in all aspects of skeletal function where estrogens are required. ERKO mice are characterized by atrophic uteri, ovarian malfunction, and tremendously increased circulating concentrations of estrogens. The testes are abnormal in appearance, wet weight, and function. Successful production of ERKO animals requires heterozygote crossing due to the reproductive impairment in both sexes when both ER alleles are inactivated. ERKO animals, like their ERKO counterparts, exhibit ovarian changes; however, unlike ERKOs, which have hemorrhagic ovaries, ERKOs demonstrate some mature follicles, but reduced numbers compared to normal wild-type mice, resulting in reduced fecundity. Uteri of these mice are normal, and circulating estrogens
311
CHAPTER 10 Regulation of Bone Cell Function by Estrogens
are normal. Testicular histology and function are normal as is male reproductive behavior; however, with age, prostate and bladder hyperplasia has been reported. ER receptor distribution is clearly distinct from ER; there is some overlap, but there is absolutely no ER in specific CNS regions, in ovarian granulosa cells, or, in males, in prostate. Animal data indicate that ER plays a dominant role in the uterus and the ovary, which raises questions as to the absolute necessity of ER in the granulosa cell. One hopes that double knockout animals, which are becoming available, will aid in elucidating ER function more clearly than the individual gene knockout examples. Early data on males reveal that the bone phenotype resembles ERKO animals, again bringing into question the role of ER in the normal developing and remodeling skeleton [88].
V. ESTROGENS AND BONE: OVERVIEW Estrogens are important regulators of skeletal development and homeostasis [89]. This is demonstrated by the dramatic loss of bone that occurs after menopause [90,91] (see Chapter 41). Moreover, estrogens are considered to be firstline therapy for the treatment of postmenopausal osteoporosis [5,92] (see Chapter 69). The reason is that these steroid hormones not only suppress bone resorption and turnover, but also relieve additional menopausal symptoms, such as hot flashes [5,92]. However, the impact of estrogens on bone goes beyond the female skeleton. It is becoming increasingly recognized that these hormones not only play a major role in the cause and prevention of postmenopausal osteoporosis, but also contribute to the development of age-related bone loss and so-called type II osteoporosis, which affects both older women and men [93] (see Chapter 38). Estrogens have both direct and indirect effects on the skeleton [89,94,93]. The extraskeletal actions relevant to calcium homeostasis include the regulation of intestinal calcium absorption [95,96] or secretion [97] the modulation of serum 1,25-dihydroxyvitamin D (vitamin D) concentrations renal calcium excretion, and the secretion of parathyroid hormone (PTH) [93,94]. The direct action of estrogens on bone cells is the subject of the reminder of this chapter. Although some of this work has been reviewed previously [e.g., 89,98 – 101], our goal is to provide a comprehensive upto date review of the literature and some insights into the complexities and mechanisms of estrogen action in the skeleton.
VI. ESTROGEN RECEPTORS IN BONE CELLS Many cell types in the skeleton express ERs. These include cells of both osteoblast and osteoclast lineages, as
well as chondrocytes and endothelial cells. For historical reasons, our discussion of this work will begin with cells of the osteoblast lineage, as these were the first bone-derived cells reported to express the ER.
A. Estrogen Receptors in Osteoblasts Prior to 1987, bone cells were not generally considered to be direct targets for estrogens [102]. However, this view began to change in 1987 when Gray et al. [103] reported that 17-estradiol decreased proliferation and increased alkaline phosphatase activity in rat UMR-106 osteosarcoma cells, which are an in vitro model for the osteoblast or bone-forming cell [104] (see Chapter 2). This report was followed the subsequent year by four publications demonstrating that rat and human osteoblastic cells expressed ERs and/or exhibited estrogenic responses. Komm et al. [105] showed specific binding sites for 125I labeled 17-estradiol in nuclear extracts from rat ROS 17/2.8 and human HOSTE85 osteosarcoma cells, as well as ER mRNA expression by these cells. These authors also reported that 17-estradiol upregulated type I procollagen and transforming growth factor (TGF)-1 mRNA levels in HOS-TE85 cells. Eriksen et al. [106] described specific nuclear-binding sites for [3H]17-estradiol in explant cultures of normal human osteoblasts (hOBs), in addition to ER mRNA expression by these cells. This group also demonstrated that 17-estradiol upregulated the nuclear progesterone receptor (PR) content of hOB cells. Kaplan et al. [107] showed by both immunocytochemistry and ligand-binding assays that osteoblasts in cystic bone lesions from a female patient with McCune – Albright syndrome (fibrous dysplasia) expressed ER. Finally, Ernst et al. [108] reported that 17-estradiol increased the proliferation of primary rat osteoblasts (ROBs) and upregulated 1 type I procollagen mRNA levels in these cells. Since these initial observations over a decade ago, ER expression has been reported to occur in a dozen different in vitro osteoblast models as well as in osteoblasts from in situ studies of bone (Table 2). These models represent a variety of mammalian and avian species. Moreover, ER expression has been determined using either Northern blot or reverse transcriptase-polymerase chain reaction (RTPCR) analysis for mRNA and Western blot or immunocytochemistry for protein. In addition, ER function has been determined by ligand-binding, DNA-binding, and estrogen response element (ERE) reporter gene assays as well as endogenous responses (which will be discussed later). Analysis of ligand-binding data indicates that osteoblasts express relatively low numbers of ERs (60-4500/cell) of high-affinity (Kd 0.05 – 1.1 nM for 17-estradiol) [105 – 107, 109 – 115]. Although these levels are much lower than is observed in uterine and breast cells, which express high amounts of ER, they are consistent with the degree of
312
KOMM AND BODINE
expression seen in other “nonclassical” estrogen responsive tissues [116]. Altogether, these results provide unequivocal evidence that osteoblasts express functional ERs and are one of the direct targets of estrogen action in the skeleton. In 1996, the discovery of the second ER, ER, was reported [117]. This discovery resulted in renaming the original ER as ER. Because each of these had a distinct — albeit overlapping — tissue distribution, investigators began to reexamine ER expression in osteoblasts in the light of these new findings. As outlined in Table 2, in situ studies of rat and human bone have demonstrated that osteoblasts express both ER isoforms [118 – 124]. Moreover, several in vitro osteoblast models, including primary rat and human osteoblasts, have been shown to express both ER and ER [115,122,126 – 129]. However, after reexamining the earlier literature, it is unclear in some instances if a specific osteoblastic cell line expresses either one or the other — or both — ER isoforms. This is particularly true for the human osteosarcoma cell lines HOS-TE85, SaOS-2, and MG-63 (Table 2). Unpublished results from our laboratory using RTPCR analysis indicate that these human osteosarcoma cell lines express only ER mRNA. Although osteoblasts appear to express both ER and ER, it is not known if the isoforms heterodimerize in these cells and what impact this may have on estrogenic responses. Moreover, the ER isoforms appear to be independently regulated during osteoblast development, which may account for the differential effects of estrogens on these cells. In primary rat osteoblasts (ROBs) [125,128] and in SV-HFO transformed human fetal osteoblastic cells [126], ER mRNA expression increases with the increasing stage of differentiation. However, ER mRNA levels either remain constant [125] or also increase [126] with advancing cellular development. Thus, the ratio of ER to ER in osteoblasts may vary as the cells progress from the preosteoblast to the mature osteocyte. Such variation might contribute to the differential estrogenic responses that have been observed in these cells [128] (this will be discussed further in a later section). Support for this idea comes from work by Hall and McDonnell [130]. Using transient transfection assays, these authors showed the following: (1) ER functions as a transdominant inhibitor of ER transcriptional activity at subsaturating steroid levels, (2) ER and ER can heterodimerize in cells, and (3) ER can interact with target gene promoters in the absence of ligand. Thus, Hall and McDonnell concluded that the relative levels of expression of these two receptor isoforms would determine how a cell responds to either estrogens or antiestrogens.
B. Estrogen Receptors in Osteocytes and Lining Cells Osteoblasts, which arise from mesenchymal stem cells in the bone marrow, undergo further differentiation to either
lining cells or osteocytes [104] (see Chapter 2). Lining cells are thought to be quiescent osteoblasts that line the mineralized bone matrix and regulate access of the osteoclasts to this tissue [131]. In contrast, osteocytes are osteoblasts that become embedded within the mineralized matrix and assume a stellate or dendritic morphology [132,133]. The primary function of osteocytes, which are the most abundant cell type in mature bone, appears to be mechanosensory [132,133]. As such, they are involved in strain perception and the adaptive mediation of physical forces on bone modeling and remodeling [134,133]. Osteocytes and lining cells may also be targets for estrogens [133]. As outlined in Table 3, evidence has been obtained from in situ studies that mammalian and avian osteocytes express ERs. Receptor expression in these cells has been shown to occur using in situ hybridization for mRNA and immunocytochemistry for protein. Moreover, as with osteoblasts, human osteocytes have been reported to express both ER and ER [118,122,124,135,136]. Unpublished observations from our laboratory indicate that a conditionally immortalized human osteocyte cell line (HOB-05-T1) expresses both ER and ER mRNAs (as measured by RT-PCR), and that these receptors are functional based on the transactivation of an ERE reporter gene by 17-estradiol. Estrogenic responses in osteocytes will be discussed in a later section. At least two publications document ER expression in bone-lining cells. Ohashi et al. [137] reported that lining cells in Japanese quail bone contained ERs, where Kusec et al. [119] showed ER mRNA and protein expression in human-lining cells. Although these studies suggest that estrogens may play a role in the physiology of these cells, there are as yet no identified estrogenic responses in lining cells. One limitation to these types of investigations is the absence of an in vitro model to study lining cell biology.
C. Estrogen Receptors in Bone Marrow Stromal Cells Pluripotent mesenchymal stem cells of bone marrow have the capacity to become osteoblasts, as well as chondrocytes, adipocytes, myoblasts, and fibroblasts [138,139]. Like other cells of the osteoblast lineage, these bone marrow stromal cells (BMSCs) express ERs and are estrogen responsive. As summarized in Table 4, primary BMSCs from rodents and humans, as well as some immortalized bone marrow stromal cell lines, have been reported to express ER and ER. In these studies, ER expression was demonstrated using RT-PCR and Northern hybridization for mRNA, immunocytochemistry for protein, and cytosolic ligand-binding assays for receptor function. Oreffo et al. [140] reported that human BMSCs express ER mRNA based on Northern blot analysis and that its expression increases as the cells undergo differentiation to osteoblasts.
313
CHAPTER 10 Regulation of Bone Cell Function by Estrogens
TABLE 2 Isoform
Estrogen Receptors in Osteoblasts System
Observation
Ref.
ER and ER
Rat ROS 17/2.8 osteosarcoma cells
mRNA Ligand binding Protein ERE-tk-CAT
[105] [113] [114] [125]
ER (?)
Human HOS-TE85 osteosarcoma cells
mRNA Ligand binding Protein
[105] [109] [294]
ER and ER
Primary human OB (hOB) cells
mRNA Ligand binding Protein ERE-tk-Luc/Cat
[106] [110] [295] [224] [127] [296] [122]
ER and ER
Human bone
Protein mRNA
[107] [118] [119] [122] [124]
ER (?) and ER
Human SaOS-2 osteosarcoma cells
Ligand binding mRNA Protein
[109] [122]
ER and ER
Rat bone
mRNA
[297] [121] [123]
ER and ER
Primary rat OB (ROB) cells
mRNA ERE-tk-CAT
[237] [125] [128] [137]
ER (?)
Japanese quail bone
Protein
ER (?)
Immortalized human HOBIT cells
mRNA
[111]
ER and ER
Immortalized mouse MC-3T3-E1 cells
mRNA Protein
[112] [294] [298]
ER (?)
Primary mouse OB cells
mRNA Protein
[294]
ER (?) and ER
Human MG-63 osteosarcoma cells
mRNA Protein
[299] [122]
ER and ER
Rat UMR-106 osteosarcoma cells
Ligand binding Protein mRNA ERE-tk-CAT
[114] [298]
ER and ER
Immortalized human HOB-03-CE6 cells
mRNA Ligand binding DNA binding ERE-tk-Luc
[115],[129]
ER
Rabbit bone
mRNA Protein
[119]
ER and ER
Transformed human SV-HFO cells
mRNA
[126]
ER
Mouse bone
mRNA Protein
[122]
314
KOMM AND BODINE
TABLE 3
Estrogen Receptors in Osteocytes
TABLE 5
Estrogen Receptors in Cells of the Osteoclast Lineage
Isoform
System
Observation
Ref.
ER (?)
Japanese quail bone
Protein
[137]
ER and ER
Human bone
Protein mRNA
[135] [118] [119] [122] [136] [124]
ER and ER (?)
Human bone
Protein mRNA
[144] [118] [122]
ER (?)
Chicken osteoclasts
Ligand binding mRNA Protein
[145]
ER
Human giant cell tumors
mRNA Protein
[146]
ER
Human FLG-29.1 preosteoclastic cells
Ligand binding mRNA Protein ERE-tk-Cat
[151]
Isoform
ER (?)
Pig bone
Protein
[135]
ER (?)
Guinea pig bone
Protein
[119]
ER
Rabbit bone
mRNA Protein
[119]
ER
Mouse bone
mRNA Protein
[122]
ER and ER
Immortalized human HOB-05-T1 cells
mRNA EREtk-Luc
Bodine and Komm, unpublished
Likewise, Dieudonne et al. [141] stated that immortalized human bone marrow stromal fibroblasts (BMSFs) isolated from a patient with a mutated ER gene, as well as nonimmortalized control BMSFs from normal patients, expressed ER mRNA as determined by RT-PCR. Moreover, the nonimmortalized control BMSFs were acknowledged to express the wild-type ER message. Estrogenic responses in BMSCs will be discussed in a later section.
D. Estrogen Receptors in Cells of the Osteoclast Lineage Osteoclasts are multinucleated giant cells responsible for bone resorption [142,143] (see Chapter 3). They arise from hemopoietic stem cells of the monocyte/macrophage linTABLE 4 Isoform
Estrogen Receptors in Bone Marrow Stromal Cells System
Observation
Ref.
ER
Mouse / LDA11 cells
Ligand binding mRNA
[300]
ER
Mouse MBA 13.2 cells
Ligand binding mRNA
[300]
ER
Mouse BMSCs
mRNA
[300] [301] [298]
ER and ER
Rat BMSCs
mRNA
[121]
ER and ER
Mouse ST2 cells
mRNA Protein
[298]
ER and ER
Human BMSCs
mRNA
[141] [140]
System
Observation
Ref.
ER (?)
Rabbit osteoclasts
mRNA
[148]
ER
Mouse hemopoietic blast cells
mRNA
[152]
ER
Rat preosteoclasts
mRNA
[302]
ER
Primary human osteoclasts
mRNA
[147]
ER
Human TCG 51 preosteoclastic cells
Protein
[303]
ER (?)
Mouse bone
Protein
[122]
eages, which, like BMSCs, are found in the bone marrow [143]. Because the primary therapeutic effect of estrogens on the postmenopausal skeleton is to suppress bone resorption [90,91], it seems logical that cells of the osteoclastic lineage would express ERs. However, the direct action of estrogens on osteoclasts is less generally accepted by workers in the field than is an indirect effect mediated by cells of the osteoblast lineage. Table 5 summarizes the evidence for ER expression by osteoclastic cells. In 1990, Pensler et al. [144] reported that human osteoclasts isolated from membranous bone (pediatric craniotomies) expressed ERs based on immunocytochemistry of fixed cells and radioimmunoassay (RIA) of cell lysates. Subsequently, Oursler and colleagues described the presence of ERs in osteoclasts purified from either chicken long bones [145] or human giant cell tumors (hGCTs) of bone (i.e., osteoclastomas) [146]. For these studies, the authors used a monoclonal antibody (121F) generated to chicken osteoclasts to purify mature osteoclasts ( 90% pure) from these tissues. ER expression was then demonstrated using either Northern blot analysis [145] or RT-PCR [146] for ER mRNA, Western blot analysis for receptor protein [145], and a nuclear ligand-binding assay, which indicated that the chicken osteoclasts contained 5000 – 6000 ERs/nucleus [145]. Two groups confirmed that human osteoclasts express ER mRNA. Hoyland et al. [118] used in situ RT-PCR to demonstrate the presence of ER message in normal human bone samples, whereas Sunyer et al. [147] used RT-PCR to reveal the expression of this mRNA in
315
CHAPTER 10 Regulation of Bone Cell Function by Estrogens
purified normal human osteoclasts (hOCLs). ER mRNA has also been reported to be expressed by isolated mature rabbit osteoclasts [148]. Thus, at least five separate laboratories have found evidence for ER expression in osteoclasts. In contrast, Collier et al. [149] failed to detect either ER or ER mRNAs in pure preparations of microisolated osteoclasts from hGCTs. Moreover, the authors confirmed their results using fluorescence in situ hybridization (FISH), which showed that the tumor mononuclear cells expressed ER mRNA, whereas the multinuclear osteoclasts did not express this gene. The reason for this discrepancy is not clear. However, Oursler [142] has postulated that prior in vivo exposure to estrogens may have downregulated ER levels in the osteoclasts examined by Collier and co-workers [149]. This conclusion is based on the work of Pedersen et al. [150], who reported that in vivo treatment of 5-week-old chickens with 17-estradiol dramatically suppressed ER protein levels in the purified osteoclasts. Preosteoclasts also appear to express ERs (Table 5). For example, Fiorelli et al. [151] used RT-PCR (for ER), Western blot analysis, a nuclear extract ligand-binding assay, and an ERE reporter gene assay to demonstrate the presence of functional ERs in human leukemic FLG 29.1 cells. The ligand-binding assay showed that this cell line, which can be induced to express an osteoclast-like phenotype, contained approximately 400 ERs/nucleus. Moreover, Kanatani et al. [152] demonstrated that mouse hemopoietic blast cells, which contain osteoclast progenitors, express ER mRNA based on RT-PCR. Estrogenic responses in osteoclastic cells will be discussed in a later section.
E. Estrogen Receptors in Chondrocytes and Other Bone-Associated Cells Estrogens play an important role in the regulation of human longitudinal bone growth and skeletal maturation [89] (see Chapter 25). These steroid hormones accelerate endochondral bone formation in early adolescence, but also initiate epiphyseal growth plate fusion in late adolescence. Consistent with these observations, chondrocytes express both ER and ER. As outlined in Table 6, rabbit, mouse, rat, human, and pig chondrocytes are all reported to possess ERs. These observations are based on in situ hybridization for ER mRNA [119], immunocytochemistry for ER and ER proteins [119,153 – 156], and cytosolic ligand-binding assays [157 – 160]. Scatchard analysis of ligand-binding data indicates that chondrocytes express relatively low amounts of ER (3.9 – 11.2 fmol/mg protein) [160] of high affinity (Kd 0.12 – 0.87 nM for 17-estradiol) [157,160]. Thus, these ER parameters are comparable to those found in osteoblasts [115]. In human growth plate chondrocytes, ER was reported to be expressed by resting, proliferative, and hypertrophic cells [119] whereas ER expression was shown to be restricted to hypertrophic cells [156]. Thus, these ER iso-
TABLE 6 Isoform
Estrogen Receptors in Chondrocytes System
Observation
Ref.
ER
Rabbit chondrocytes
Ligand binding mRNA Protein
[157] [119] [155]
ER and ER
Human chondrocytes
Protein
[153] [154] [119] [156] [303]
mRNA
ER (?)
Rat chondrocytes
Protein
[160] [155]
ER (?)
Pig bone
Protein
[135]
ER (?)
Guinea pig
Protein
[135]
forms may have distinct roles in the regulation of endochondral bone growth and maturation. Estrogenic responses in chondrocytes will be discussed in a later section. At least one report describes the expression of ERs in bone-derived endothelial cells [161]. Using bovine bone endothelial (BBE) cells, the authors showed that these cells expressed ER mRNA by Northern hybridization and contained specific binding sites for [3H]17-estradiol (Kd 17.2 nM, Bmax 32,000 sites/cell). Treatment of the cells with 17-estradiol enhanced proliferation and suppressed PTH-stimulated cyclic adenosine monophosphate (cAMP) accumulation. As described in more detail later, both of these estrogenic responses have also been observed in osteoblasts. Thus, this study suggests that estrogens may regulate bone angiogenesis as well as bone formation and resorption.
F. Summary It is clear from the numerous studies reviewed in this section that many cell types in the skeleton express ERs. These estrogen responsive cell types include bone marrow progenitor cells as well as mature osteoblasts, osteoclasts, and chondrocytes. In the osteoblast lineage, each cell type — from the BMSC to the osteocyte or lining cell — has been shown to be a potential estrogen target. Thus, the totality of the effects of estrogen on the skeleton may, to a large extent, be equivalent to the sum of its action on all of these cell types. The following section reviews the estrogenic responses of skeletal cells and places them in the context of in vivo knowledge of estrogen action.
VII. ESTROGENIC RESPONSES IN BONE CELLS Consistent with the expression of ERs by many bone cell types, there are also many estrogenic responses in these
316
KOMM AND BODINE
cells. Our review of the literature on estrogenic responses indicates that findings are sometimes contradictory. We will attempt to place the estrogenic response in the context of estrogen’s known physiologic and therapeutic function in the skeleton.
A. Estrogenic Responses in Cells of the Osteoblast Lineage Due to the profusion of in vitro models, much of what we know about estrogen action on bone cells is in relationship to the osteoblast. As summarized in Table 7, 43 estrogenic responses have been identified in 15 different in vitro osteoblast models. To make sense of these observations we have separated them into six different categories: regulation of osteoblast number; regulation of matrix production and mineralization; regulation of growth factor expression and responsiveness; regulation of factors that modulate bone resorption; regulation of receptor expression and signal transduction; and miscellaneous responses. Moreover, we have indicated which in vitro model(s) was reported to exhibit each of the responses. The reason for doing this is to determine if a given response represents a generalized estrogenic effect in an osteoblast or whether it might be specific to a particular cell line (e.g., immortalized MC-3T3-E1 mouse cells) or cell type (e.g., osteosarcoma-derived cells). From our viewpoint, primary cultures are the most pertinent osteoblast models for attempting to extrapolate in vitro estrogen affects to in vivo relevance. On the other hand, caution should be applied to observations that are only made in osteosarcoma cells, as these are generally considered to be unreliable models of osteoblasts biology [162,163]. When available, we have also noted when an in vitro estrogenic response has been observed in vivo and therefore may be relevant physiologically or pharmacologically (see Chapter 37). 1. REGULATION OF OSTEOBLAST NUMBER Using UMR-106 rat osteosarcoma cells, Gray et al. [103] reported that 17-estradiol decreases osteoblastic cell proliferation. In the same study, 17-estradiol also increased alkaline phosphatase activity. Given the limitations of osteosarcoma cells as models of osteoblast biology [162,163] these results suggested that estrogens might potentiate cellular differentiation, as the mature rat osteoblast expresses high levels of alkaline phosphatase and no longer divides [162,104]. Subsequent to this publication, other researches described similar results using four additional in vitro osteoblast models (Table 7). These include primary osteoblasts isolated from the tibias of 17-estradiol-treated ovariectomized (OVX) rats [164]. Moreover, Westerlind et al. [165] confirmed these observations in vivo by showing that the potent nonsteroidal estrogen, diethylstilbestrol (DES), reduces the [3H]thymidine-labeling index of tibial
osteoblasts in OVX rats. Thus, a suppressive effect of estrogens on osteoblast proliferation is consistent with an inhibitory action of the steroid on bone turnover [89 – 91]. In contrast, other laboratories using additional in vitro models, as well as ROBs, have reported that estrogens increase osteoblast proliferation and DNA synthesis (Table 7). There are several possible explanations for these discrepancies. First, with the exception of studies using UMR106 and ROBs, other publications that showed that 17 estradiol suppresses proliferation used cell lines that overexpressed ER. Thus, as was concluded by Watts and King [166], overexpression of the ER may inhibit cell proliferation by interfering with transcription artifactually. If this is true, then a transfected ER may not necessarily function the same as the endogenous ER. However, studies reporting that 17-estradiol stimulated osteoblast proliferation all used in vitro models that naturally expressed ERs. At least two groups have reported that in vitro treatment of ROBs with 17-estradiol enhances cell proliferation or DNA synthesis [108,167], whereas Modrowski et al. [164] used isolated osteoblasts from in vivo-treated OVX rats to show that the steroid inhibits proliferation. Consequently, these two experimental paradigms may generate cells that are in different stages of differentiation (e.g., preosteoblastic versus mature osteoblasts), and these stages may respond differently to estrogens [168,128]. While a suppressive effect of estrogens on osteoblast proliferation is consistent with a potentiation of differentiation or a suppression of bone turnover, a stimulatory effect might reflect an expansion of the preosteoblast pool [169]. For example, Qu et al. [170] presented evidence that treatment of primary mouse BMSC cultures with 17-estradiol stimulates cellular proliferation and differentiation into osteoblastic cells. This in vitro observation is consistent with an in vivo study of Somjen et al. [171], which reported that 17-estradiol stimulates DNA synthesis in rat bone. Moreover, in OVX Swiss – Webster mice, high doses of 17 estradiol (50 – 100 g/mouse/week, s.c., for 4 weeks) increased both endosteal and cancellous bone formation, as well as inhibited bone resorption [172]. Thus, under some circumstances, estrogens may stimulate bone formation [173] as well as inhibit resorption and turnover. However, the stimulatory action of the steroid may represent a pharmacological or toxic effect rather than a physiologic or therapeutic response [174]. In addition to regulating cell division, estrogens have now been shown to control osteoblast apoptosis. Gohel et al. [175] reported that 17-estradiol blocks the induction of apoptosis by cortisol in primary rat and mouse osteoblasts. These in vitro observations were confirmed by an in vivo experiment showing that 17-estradiol decreased the number of apoptotic osteoblasts in the calvaria of dexamethasone-treated mice. Consequently, estrogens may modulate osteoblast number by regulating both proliferation and
317
CHAPTER 10 Regulation of Bone Cell Function by Estrogens
TABLE 7
Estrogenic Responses in Cells of the Osteoblast Lineage
Response
System
Ref.
Regulation of cell number Decreases proliferation and decreases DNA synthesis
Rat UMR-106 osteosarcoma cells Human HTB-96 cells overexpressing ER Primary rat OB (ROB) cells Rat ROS.SMER-14 cells overexpressing ER Human hFOB/ER9 cells overexpressing ER
Rat ROS 17/2.8 osteosarcoma cells Rat bone Increases proliferation and increases DNA synthesis
Primary ROB cells
[103] [254] [304] [166] [164] [167] [113] [212] [168] [178] [203] [305] [165]
Primary human OB (hOB) cells Human HOS-TE85 osteosarcoma cells Primary mouse bone marrow stromal cells Rat bone Mouse bone
[108] [181] [167] [181] [112] [177] [176] [306] [170] [171] [169]
Primary ROB cells Primary mouse OB cells Mouse bone
[175] [175] [175]
Rat UMR-106 osteosarcoma cells
Immortalized human HOB-03-CE6 cells Primary ROB cells
[103] [254] [113] [176] [177] [168] [178] [115] [128]
Decreases alkaline phosphatase
Primary ROB cells Rat bone
[128] [180]
Increases osteocalcin
Primary ROB cells
[128]
Decreases osteocalcin
Rat ROS 17/2.8 osteosarcoma cells Human hFOB/ER9 cells overexpressing ER
[179] [168] [178] [128] [180] [179] [182]
Transformed rat RCT-1 and -3 cells Immortalized mouse MC-3T3-E1 cells
Inhibits glucocorticoid-induced apoptosis
Regulation of matrix production and mineralization Increases alkaline phosphatase
Rat ROS.SMER-14 cells overexpressing ER Primary hOB cells Immortalized mouse MC-3T3-E1 cells Human hFOB/ER9 cells overexpressing ER
Primary ROB cells Rat bone
Increases osteonectin
Primary ROB cells
[128]
Decreases osteonectin
Primary ROB cells Rat bone
[128] [180]
Increases type I collagen
Human HOS-TE85 osteosarcoma cells Primary ROB cells
[105] [108] [181] [128] [181]
Transformed rat RCT-1 and -3 cells
(continues)
318
KOMM AND BODINE
TABLE 7
(continued)
Response
System Primary hOB cells
Immortalized mouse MC-3T3-E1 cells
Ref. [110] [127] [296] [177]
Decreases type I collagen
Primary hOB Cells Rat bone
[296] [180] [182]
Increases mineralization
Human HOS-TE85 osteosarcoma cells Primary human OB (SaM-1) cells
[185] [185]
Regulation of growth factor expression and responsiveness Increases TGF-1
Human HOS-TE85 osteosarcoma cells Rat UMR-106 osteosarcoma cells Primary hOB Cells Primary mouse OB cells Primary ROB cells Rat ROS 17/2.8 osteosarcoma cells Rat bone
[105] [197] [198] [199] [128] [179] [199] [179]
Increases TGF-3
Human MG-63 osteosarcoma cells Rat bone
[202] [200]
Increases TIEG
Human hFOB/ER9 cells overexpressing ER
[203]
Increases BMP-6
Human hFOB/ER9 cells overexpressing ER
[205]
Increases IGF-I
Rat UMR-106 osteosarcoma cells Primary ROB cells
[207] [181] [307] [181] [308]
Transformed Rat RCT-1 and -3 cells Human hFOB/ER9 cells overexpressing ER Increases growth hormone Receptor
Rat UMR-106 osteosarcoma cells Primary hOB cells
[208] [208]
Increases IGF-BPs
Primary ROB cells Human hFOB/ER9 cells overexpressing ER Human SaOS-2 osteosarcoma cells
[211] [212]
Decreases IGF-BP3
Primary human bone marrow stromal cells
[215]
Blocks PGE2-induced IGF-1
ROB cells overexpressing ER
[216]
Mouse / LDA11 marrow stromal cells Primary hOB cells Primary ROB cells Primary mouse OB cells
[217] [217] [217] [217] [309] [217] [310] [225]
[213]
Regulation of factors that modulate bone resorption Decreases IL-6
Immortalized mouse MC-3T3-E1 cells Human SaOS-2 cells overexpressing ER Human hFOB/ER9 cells overexpressing ER Immortalized human HOB-03-CE6 cells Human MG-63 osteosarcoma cells Primary human bone marrow stromal cells In vivo (mice)
[115] [311] [219] [218] [312]
Decreases TNF-
Primary hOB cells
[223]
Decreases gp80 and gp130
Mouse / LDA11 marrow stromal cells Immortalized mouse MC-3T3-E1 cells
[227] [227]
319
CHAPTER 10 Regulation of Bone Cell Function by Estrogens
TABLE 7
(continued)
Response
System
Ref.
Increases OPG
Human hFOB/ER9 cells overexpressing ER Primary hOB Cells
[229]
Suppresses PTH action
Human SaOS-2 osteosarcoma cells
[232] [233] [235] [213] [181] [237] [181] [234] [198] [177] [115] [236]
Transformed rat RCT-1 and -3 Cells Primary ROB cells Primary mouse OB cells Primary hOB cells Immortalized mouse MC-3T3-E1 cells Immortalized human HOB-03-CE6 cells In vivo (humans) Enhances PTH action
Increases IL-1
Human SaOS-2 osteosarcoma cells
[229]
Primary ROB cells Primary hOB cells
[238] [239] [239] [239]
Immortalized human HOBIT cells
[241]
Regulation of receptor expression and signal transduction Increases PR
Primary hOB cells Human hFOB/ER9 cells overexpressing ER
[106] [242] [225]
Antagonizes VD3 responsiveness
Rat UMR-106 osteosarcoma cells
[207]
Increases VDR and VD3 responsiveness
Rat ROS 17/2.8 osteosarcoma cells Human OGA osteosarcoma cells
[243] [243]
Increases ER
Primary hOB cells
[127] [296] [128] [136] [124]
Primary ROB cells In vivo (human bone) Decreases ER
Primary ROB cells In vivo (human bone)
[128] [136]
Decreases IP3 receptor
Rat UMR-106 osteosarcoma cells
[245] [247] [245] [247] [245] [247] [245] [247] [245] [247]
Human SaOS-2 osteosarcoma cells Primary ROB cells Immortalized mouse MC-3T3-E1 cells G-292 human osteosarcoma cells Increase basal NOS
Human HOS-TE85 osteosarcoma cells In vivo (rats)
[248] [250]
Decrease cytokine-induced NO
Immortalized mouse MC-3T3-E1 cells
[251]
Enhances bradykinin action
Primary hOB cells
[252]
Increases CK
Primary ROB cells Immortalized mouse MC-3T3-E1 cells Rat ROS 17/2.8 osteosarcoma cells Rat bone
[171] [171] [171] [171]
Increases HSP-27
Immortalized mouse MC-3T3-E1 cells
[253]
Increases AST, GGT, LDH, and transferrin
Rat UMR-106 osteosarcoma cells
[254]
Miscellaneous response
320 viability. As will be reviewed in later sections, estrogens may also suppress osteocyte apopotosis but induce the programmed cell death of osteoclasts. 2. REGULATION OF MATRIX PRODUCTION MINERALIZATION
AND
One of the most commonly observed estrogenic responses in osteoblasts is the upregulation of alkaline phosphatase expression, which is an important phenotypic marker of the osteoblast lineage [104] (see Chapter 2). Estrogens have been reported to increase either alkaline phosphatase mRNA levels and/or activity in seven different in vitro osteoblast models (Table 7). These models include rat osteosarcoma cell lines [103,113], primary cultures of ROB or hOB cells [128,176], immortalized mouse MC-3T3-E1 cells [177], and the conditionally immortalized human osteoblast cell lines hFOB/ER9 and HOB-03-CE6 [115,168,178]. However, in the case of ROB cells, 17estradiol has also been reported to downregulate alkaline phosphatase expression [128]. The explanation for this discrepancy is that 17-estradiol regulates the steady-state mRNA levels of this enzyme in a differentiation selective manner [128]. In post-proliferative/nodule-forming stage ROB cells (i.e., mature osteoblasts), 17-estradiol estradiol suppresses alkaline phosphatase expression, whereas in postmineralization stage cells (i.e., osteocytes), the steroid hormone increases enzyme message levels. This same pattern of regulation also holds true for the noncollagenous bone matrix proteins osteocalcin and osteonectin [128]. Estrogens also regulate the expression of osteocalcin (Table 7), which is the most selective phenotypic marker of the osteoblast lineage [104]. As noted previously, 17estradiol downregulates steady-state osteocalcin mRNA levels in postproliferative/nodule-forming stage ROB cells, but upregulates it in postmineralization stage cells [128]. Moreover, estrogens decrease osteocalcin expression in ROS 17/2.8 osteosarcoma cells [179] and in hFOB/ER9 cells, which over express human ER [168,178]. Confirmation that estrogens downregulates alkaline phosphatase, osteocalcin and osteonectin mRNA levels in vivo comes from Turner et al. [180], who reported that DES treatment of OVX rats decreased the expression of these messages in periosteal osteoblasts isolated from long bones. Again, suppression of osteoblastic activity as measured by the expression of bone matrix proteins would be consistent with a reduction in bone turnover. The most abundant bone matrix protein is of course type I collagen [104] (see Chapter 4), and it is perhaps not surprising that estrogens have been shown to regulate its expression (Table 7). Komm et al. [105] and Ernst et al. [108] were the first to report that 17-estradiol upregulated 1 type 1 procollagen mRNA levels in HOS-TE85 human osteosarcoma cells and ROB cells, respectively. Subsequent studies confirmed these observations in hOBs [110], MC-
KOMM AND BODINE
3T3-E1 [177], and transformed rat RCT-1 and RCT-3 cell lines [181]. In contrast to these in vitro studies, type I collagen expression does not appear to be upregulated by estrogens in vivo. In fact, mRNA levels for this bone matrix protein have been reported to increase in OVX rat bones [182,183], and estrogens have been observed to either suppress this increase [180,182], or have no effect [184]. Once again, the in vivo observations are consistent with the concept that estrogen deficiency increases bone resorption and bone turnover, and that estrogens reduce these effects [89 – 91]. Finally, at least one report describes the effects of estrogens on mineralization. Takeuchi et al. [185] showed that 17-estradiol at concentrations of 1 – 100 nM increased the calcium content of extracellular matrix laid down in vitro by either HOS-TE85 human osteosarcoma cells or primary human osteoblasts (referred to as SaM-1 cells). 3. REGULATION OF GROWTH FACTOR EXPRESSION RESPONSIVENESS
AND
Another aspect of osteoblast biology that estrogens have been shown to regulate is growth factor expression and responsiveness. Bone is an abundant reservoir for several growth factors, including isoforms of transforming growth factor (TGF)-, bone morphogenetic proteins (BMPs), and insulin-like growth factors (IGFs) [186 – 191] (see Chapters 13 and 14). These peptides are synthesized and secreted by cells of the osteoblast and/or osteoclast lineages, and they regulate the proliferation, differntiation, and activities of these cell types [186,187,189 – 193]. In fact, growth factors, together with other cytokines, provide the elaborate communication network that couples osteoclastic bone resorption to osteoblastic bone formation [89,99] (see Chapter 12). Moreover, it is the disruption of this network that, to a large extent, leads to accelerated bone resorption and increased bone turnover after menopause [99,194 – 196]. The first bone cell-derived growth factor whose expression was shown to be regulated by estrogens was TGF-1. Komm et al. [105] reported in 1988 that 17-estradiol treatment of HOS-TE85 human osteosarcoma cells upregulated the steady-state levels of TGF-1 mRNA. As outlined in Table 7 estrogens have also been shown to increase TGF1 mRNA expression and/or TGF- protein secretion in rodent osteosarcoma cell lines [179,197], as well as primary cultures of human, mouse, and rat osteoblasts [128,198,199]. Moreover, estrogens have been observed to increase TGF- expression in bone in vivo. Finkelman et al. [199] reported that treatment of OVX rat with 17 estradiol upregulated TGF- protein levels in long bones, and lkeda et al. [179] demonstrated that TGF-1 mRNA levels decreased in the tibia of OVX rats. However, neither Westerlind et al. [183] nor Yang et al. [200] were able to confirm these findings. Although TGF- regulates osteoblast proliferation, differentiation, and activity in vitro
CHAPTER 10 Regulation of Bone Cell Function by Estrogens
and promotes bone formation in vivo [186,187,190], it has also been reported to inhibit osteoclast differentiation and activity in vitro [89,99]. Thus, an increase in osteoblastic TGF- production would be consistent with a therapeutic antiresorptive effect of estrogens [90,91]. Estrogens, as well as tissue-selective estrogens (TSEs) [92] or SERMs [201], have also been reported by at least one group to increase TGF-3 expression by osteoblastic cells (Table 7). Yang et al. [200] observed an increase in TGF-3 mRNA levels in the femurs of OVX rats that were treated with either 17-estradiol or the SERM raloxifene; in contrast, the message levels for either TGF-1 or TGF2 were unaffected by these treatments. Although in situ studies to identify the celltype(s) responsible for this expression were not reported, this same group subsequently demonstrated that 17-estradiol or raloxifene upregulated TGF-3 mRNA levels in MG-63 human osteosarcoma cells [202]. These observations were extended by contransfection studies in MG-63 cells using human TGF-3 promoter – reporter gene constructs and human ER expression vectors [200,202]. These experiments indicated that a variety of estrogens and TSEs/SERMs upregulated TGF-3 promoter activity in an ER-dependent manner. Although these results were intriguing, an apparent disconnection occurred between in vitro and in vivo pharmacology, as the potency and efficacy of compounds in this in vitro assay did not correlate with their bone-sparing activities in vivo. Moreover, 17-estradiol antagonized raloxifene in this in vitro system [202]. In any event, as with TGF-1, upregulation of TGF-3 expression in bone by either estrogens or a TSE/SERM would be consistent with an antiresorptive effect, as this isoform also inhibits in vitro osteoclastic differentiation and activity [200]. In addition to upregulating TGF- expression in osteoblasts, estrogens may act like these peptides in terms of their downstream effects. For instance, 17-estradiol has been reported by Tau et al. [203] to increase the expression of TIEG (TGF- inducible early gene) in conditionally immortalized hFOB/ER9 human fetal osteoblasts. The expression of this gene is also increased by TGF- in human osteoblastic cells [204]. Treatment of this cell line with 17-estradiol, or overexpression of TIEG, causes a reduction in DNA synthesis. These results suggest that at least part of the mechanism by which estrogens inhibit osteoblast proliferation may involve upregulation of TIEG. Estrogens appear to regulate the expression of additional members of the TGF- superfamily. In 1998, Rickard et al. [205] reported that treatment of hFOB/ER9 cells with 17estradiol increased both steady-state mRNA levels and protein levels of BMP-6 (Table 7). In contrast, the steroid hormone had no effect on TGF-1, TGF-2, BMP-2, BMP-4, or BMP-5 expression. Like TGF-s, BMPs also have autocrine and paracrine effects on a variety of skeletal cells [186,189]. More recently, van den Wijngaard et al. [206]
321 reported that antiestrogens or TSEs/SERMs such as tamoxifen, raloxifene, and ICI-164,384 upregulated human BMP-4 promoter-luciferase expression in U2-OS human osteosarcoma cells that were cotransfected with hER but not hER. However, this response required expression of relatively high receptor levels and was blocked by cotreatment with 17-estradiol. As there is no evidence to date that endogenous BMP-4 expression is increased in osteoblasts without ER overexpression, it is unclear whether this observation has any bearing on the pharmacological actions of TSEs/SERMs in the skeleton. In addition to members of the TGF-/BMP family, estrogens have been observed to regulate the expression of components of the osteoblastic IGF/growth hormone (GH) system as well. Gray et al. [207] were the first to report that 17-estradiol treatment upregulated the secretion of IGF-I and IGF-II from UMR-106 rat osteosarcoma cells. These results were confirmed, at least for IGF-I, in three additional osteoblast models, including ROBs (Table 7). Likewise, 17-estradiol increased GH receptor expression and GH action in UMR-106 cells and normal human osteoblast cultures [208]. In contrast, in vivo studies by Turner and coworkers [184,209] in OVX rats failed to verify these in vitro observations. In fact, these authors demonstrated that estrogen loss resulted in an increase in IGF-I mRNA expression in calvarial periosteum and that DES treatment suppressed this increase. Because IGFs increase bone formation, resorption, and turnover [187,188], upregulation of osteoblastic IGF expression following 17-estradiol treatment in vitro is inconsistent with a suppressive effect of the steroid hormone on resorption and turnover in vivo [89 – 91]. However, the in vitro studies were confirmed by Erdmann et al. [210], who showed that supraphysiological doses of 17-estradiol increased IGF-I protein content in femoral shaft bone matrix of OVX rats. However, these authors cautioned that this stimulatory effect of estrogens occurred only at relatively high concentrations of steroid and that this may not be relevant to the normal physiological actions of the hormone. Because high estrogen doses stimulate bone formation in OVX mice [169,172], upregulation of IGF-I levels in bone may be part of the mechanism by which this pharmacological effects occurs. Estrogens also increase IGF-binding protein (IGF-BP) secretion and expression by ROBs [211], hFOB/ER9 [212], and SaOS-2 human osteosarcoma cells [213] (Table 7). IGF-BPs are secreted proteins that bind IGF-I and IGF-II and regulate their bioavailability and activity [192,214]. Consequently, the IGF-BPs can either enhance or inhibit IGF action. Moreover, in some instances, these BPs may also act independently of IGFs. Of the six IGF-BPs, all of which are expressed by human osteoblasts [214], IGF-BP4 is considered to the most inhibitory to IGF activity [192]. In 1996, Kassem et al. [212] demonstrated that 17-estradiol increased IGF-BP4 mRNA expression and secretion in
322 hFOB/ER9 conditionally immortalized fetal human osteoblasts that overexpress hER. In contrast, the steroid had no effect on either IGF-II or IGF-BP3 expression. In addition, 17-estradiol decreased IGF-BP4 proteolysis. Because 17-estradiol also inhibited DNA synthesis by these cells, the authors proposed that upregulation of IGF-BP4 levels in the bone microenvironment might contribute to the suppressive action of estrogens on bone formation observed in vivo [89]. However, Rosen et al. [215] reported that 17-estradiol suppressed IGF-BP3 secretion from a primary culture of human BMSCs. Another potential mechanism by which estrogens may suppress IGF-dependent bone turnover is through antagonism of induced IGF-1 expression. Using ROBs that were contransfected with a human ER expression vector, McCarthy et al. [216] reported that 17-estradiol suppressed PGE2-induced rat IGF-I promoter-luciferase activity. However, basal promoter function was unaffected by the hormone. 4. REGULATION OF FACTORS THAT MODULATE BONE RESORPTION As noted in preceding sections, the therapeutic actions of estrogens in vivo primarily involve the suppression of bone resorption and turnover [5,89]. One of the chief estrogenic targets for these antiresorptive effects are cells of the osteoblast lineage [99,194 – 196]. As outline, in Table 7, at least five different effects of estrogens on osteoblasts and their progenitors involve the suppression of cytokine production, cytokine action, or bone-resorbing hormone activity (see Chapter 13). One of the most commonly reported estrogenic effects in cells of the osteoblast lineage is the downregulation of synthesis of interleukin (IL)-6, a cytokine that promotes differentiation of osteoclast progenitors to mature bone-resorbing cells [138,194 – 196] (see Chapter 41). In 1992, Girasole et al. [217] reported that 17-estradiol suppressed the induction of IL-6 secretion by tumor necrosis factor (TNF)- or IL-1 in mouse / LDA11 stromal cells, MC-3T3-E1 immortalized mouse osteoblastic cells, or primary cultures of rat and human osteoblasts. Moreover, in neonatal mouse calvarial-derived bone cell cultures that contain osteoblasts as well as osteoclast progenitors, 17estradiol inhibited both TNF--stimulated IL-6 production and osteoclast development. In addition, equivalent suppression was also observed with an anti-IL-6 antibody, indicating that IL-6 was involved in this process. These in vitro observations were confirmed by the same group later that year in an in vivo study in mice [218]. These findings were also corroborated by Cheleuitte et al. [219], who used cultured BMSCs isolated from postmenopausal women. These authors showed that basal and IL-1-stimulted IL-6 secretion from incubated BMSCs was reduced significantly relative to age-matched control when cells were isolated
KOMM AND BODINE
from women using estrogen replacement therapy (ERT). The mechanism for the inhibition of IL-6 expression by 17-estradiol was determined by Pottratz et al. [220], who showed that it was through an ER mediated indirect effect on IL-6 promoter activity. Subsequent studies have demonstrated that the ER interferes with NF-B activity, although the precise molecular events involved in this suppression await eluctation [reviewed in 221]. Although several other research groups using a variety of in vitro osteoblast models have corroborated these findings (Table 7), others have been unable to verify IL-6 as a target for estrogen action [222 – 225]. These reports used primary cultures of hOBs or human BMSCs, which are known to express relatively low and variable amounts of ER [106,224]. Our laboratory offered a possible explanation for this discrepancy. Using conditionally immortalized human HOB-03-CE6 cells that naturally express functional ERs [115], we showed that the bone-resorbing cytokines TNF- and IL-1/ are potent suppressors of ligand-dependent receptor activity [129]. In this cell line, 17-estradiol downregulates basal IL-6 mRNA levels [115], but does not block the induction of IL-6 secretion by either TNF- or IL-1 [129]. Thus, we postulated that in osteoblasts that normally express low ER levels, TNF- and IL-1/ may inactivate the receptor before it can blunt IL-6 production. Although Rickard et al. [223] were unable to demonstrate that 17-estradiol suppressed IL-1-induced IL-6 secretion from hOB cells, they did show that the steroid downregulated the release of TNF- from these cells in response to IL-1 stimulation. Estrogens have also been shown to blunt IL-6 responsiveness in osteoblastic and BMSCs cells. The IL-6 receptor is a bipartite complex composed of two transmembrane glycoproteins. One is an 80,000-Da protein (gp80) that binds the cytokine, whereas the other is a dimer of 130,000 Da proteins (gp130) that is involved in signal transduction to the JAK/STAT (Janus kinase/signal transducer and activator of transcription) pathway [226]. Lin et al. [227] reported that 17-estradiol downregulated gp80 and gp130 mRNA levels, as well as gp130 protein levels, in /LDA11 stromal cells. Likewise, the steroid hormone also suppressed the induction of gp130 mRNA by PTH, IL11, or leukemia inhibitory factor (LIF) in MC-3T3-E1 osteoblastic cells. Although cells of the osteoblast lineage produce many proteins that potentiate osteoclastogenesis and osteoclastic activity, one termed RANKL (receptor activator of NF-B ligand) appears to be critical for this process [reviewed in [143,228] (see also Chapters 3, 12, and 13) RANKL is a membrane protein found on the surface of osteoblasts and BMSCs. Moreover, it is the ligand for RANK (receptor activator of NF-B), a transmembrane protein that is expressed by osteoclast progenitors and mature bone-resorbing cells. The binding of RANKL to RANK stimulates the
323
CHAPTER 10 Regulation of Bone Cell Function by Estrogens
differentiation of osteoclast progenitors to mature osteoclasts. Additionally, it activates mature cells. However, RANKL is also a ligand for a secreted decoy receptor called osteoprotegerin (OPG). Osteoblasts and BMSCs synthesize OPG as well as RANKL [99,228], and OPG suppresses bone resorption by sequestering RANKL [99,143,228]. Consequently, given the antiresorptive nature of estrogens, it is not surprising that these hormones have been observed to increase OPG expression by osteoblasts. Using both conditionally immortalized hFOB/ER9 fetal human osteoblastic cells as well as hOBs, Hofbauer et al. [229] demonstrated that 17-estradiol up regulated OPG mRNA levels and increased OPG secretion. One potential mechanism by which estrogens suppress cytokine expression in BMSCs has been elucidated by Srivastava et al. [230]. Using primary cultures of BMSCs isolated from mice, these authors showed that ovariectomy results in increased nuclear levels of phosphorylated Egr-1, a transcription factor that modulates expression of the cytokine macrophage colony-stimulating factor (M-CSF). M-CSF, in turn, is an important inducer (together with RANKL) of osteoclast differentiation [143]. When compared to non phosphorylated Egr-1, the phosphorylated protein binds less well to another transcription factor, Sp-1; this results in increased nuclear levels of free Sp-1, which leads to increased transctivation of the M-CSF gene in BMSCs. Conversely, treatment of wild-type OVX mice with 17-estradiol decreases the levels of phosphorylated Egr-1 in the nucleus of BMSCs and therefore down regulates M-CSF expression. Protein antagonists of IL-1 and TNF- mimic this down regulation. In contrast, 17-estradiol has no effect on M-CSF expression in OVX mice that lack Egr-1. Another commonly reported osteoblastic response to estrogens is the suppression of PTH action. Like estrogens, PTH is an important hormonal regulator of bone metabolism [231] (see Chapter 7). Osteoblasts are the primary targets for PTH action in bone and mediate both anabolic and catabolic activities of this hormone. In fact, one of the bone-resorbing effects of PTH on osteoblastic cells is the upregulation of RANKL expression [143]. As summarized in Table 7, treatment of seven different in vitro osteoblast models with 17estradiol has been shown to block the ability of those cells to respond to PTH. Typically, 17-estradiol has been observed to inhibit the PTH-stimulated increase in intracellular cAMP content [115,177, 181,232,233]. However, the steroid has also been reported to interfere with some of the downstream effects of the peptide as well [198,213,227,234]. In at least one instance, PTH has also been shown to block an estrogenic effect in an osteoblast [203]. Furthermore, the suppressive effect of estrogens on PTH activity has also been observed clinically. Using urinary biochemical markers of bone resorption (see Chapter 60), Cosman et al. [236] reported that postmenopausal women treated with estrogens
exhibited a markedly blunted response to a continuous intravenous infusion of h PTH (1 – 34). The mechanism by which estrogens interfere with PTH signaling is not clear. Using SaOS-2 human osteosarcoma cells, Monroe and Tashjian [233] proposed that suppression was due to a decrease in membrane-associated adenylyl cyclase activity. However, this mechanism does not appear to be applicable to HOB-03-CE6 conditionally immortalized human osteoblasts, as the inhibitory actions of 17-estradiol are selective for PTH over PGE2 and forskolin-stimulated cAMP production [115]. Ernst et al. [237] suggested that the ability of 17-estradiol to reduce PTH-stimulated cAMP production in RCT-3 transformed rat osteoblasts was due to a nongenomic action of the steroid, because it was observed within 4 h of treatment and was not enhanced by overexpression of ER. Although these data are suggestive of a nongenomic effect, they are by no means conclusive. While most studies have demonstrated an antagonistic effect of estrogens on PTH activity or cytokine expression, a few reports have shown the opposite to occur (Table 7). For example, 17-estradiol has been observed to enhance PTH responsiveness. In dexamethasone-conditioned SaOS2 cells, 17-estradiol and PTH potentiate each others stimulatory effect on alkaline phosphatase activity [238]. Whereas in SaOS-2 cells, as well as in primary rat and human osteoblasts, the steroid enhances the ability of PTH to stimulate fibronectin production [239]. Although these reports appear to contradict the antagonistic effects of estrogens on PTH activity in osteoblasts, PTH receptors are coupled to at least two signal transduction pathways [240], and estrogens may act differentially on these second messenger systems. Likewise, using a T-antigen-transformed human osteoblast cell line (HOBIT), Pivirotto et al. [241] presented evidence that 17-estradiol upregulates IL-1 mRNA levels. However, because this effect has only been reported to occur in HOBIT cells, its biological significance is questionable. 5. REGULATION OF RECEPTOR EXPRESSION SIGNAL TRANSDUCTION
AND
Estrogens modulate the expression of several receptors in osteoblasts. At least three members of the nuclear receptor superfamily are known to be regulated by these steroids. As occurs in uterine and breast cells, treatment of either hOBs or conditionally immortalized hFOB/ER9 cells with 17-estradiol upregulates progesterone receptor (PR) expression [106,225,242]. The steroid has also been observed to increase vitamin D receptor (VDR) levels and vitamin D responsiveness in two osteosarcoma cell lines [243,244]. In addition, it either increases [127,128] or decreases [128] ER mRNA levels in primary cultures of human and rat osteoblasts, respectively. In the case of ROB cells, our laboratory demonstrated that 17-estradiol downregulates ER expression in day 14 nodule-forming cultures (osteoblastic
324 cells), while it upregulates receptor expression in day 30 late mineralization stage cultures (osteocytic cells) [128]. Consistent with these observations, Hoyland et al. [136] have reported that ERT or hormone replacement therapy (HRT) decreases the number of ER mRNA-positive osteoblasts in human bone biopsies. However, ERT/HRT increases the number of ER protein-positive osteocytes in these biopsies. Thus, estrogens play a role in both directly regulating osteoblastic activity and modulating the hormonal responsiveness of the cells. Estrogens have also been reported to regulate additional signal transduction pathways in osteoblasts (Table 7). One interesting finding is that 17-estradiol downregulates mRNA expression of the type I inositol trisphosphate (IP3) receptor in several in vitro osteoblast models [245]. This receptor is a transmembrane calcium channel found on the “calciosome,” a specialized component of endoplasmic reticulum that is involved in the storage and release of IP3 sensitive intracellular calcium [246]. This receptor is therefore essential for the phosphoinositide-signaling pathway. Because bone-resorbing agents such as PTH, prostaglandins, and bradykinin utilize this pathway, suppression of type I IP3 receptor expression by estrogens in osteoblasts may lead to decreased bone resorption and turnover. Although the human type I IP3 receptor promoter does not contain a consensus ERE, 17-estradiol nevertheless downregulates promoter – reporter gene constructs when transfected transiently into G292 human osteosarcoma cells [247]. Another interesting observation is the upregulation of endothelial nitric oxide synthase (ecNOS or NOS-1) mRNA expression and enzyme activity in estrogen-treated HOS TE-85 human osteosarcoma cells [248]. Because a high nitric oxide (NO) content has been reported to inhibit in vitro osteoclastic bone resorption [249], this estrogenic effect is also consistent with an antiresorptive role for the steroid. An in vivo study with OVX rats confirms these results. Wimalawansa et al. [250] reported that treatment of OVX rats with either 17-estradiol or nitroglycerine (an NO donor) reversed lumbar spine bone loss as measured by dual-energy X-ray absorptiometry (DXA). In contrast, cotreatment with 17-estradiol and NG-nitro-L-arginine methyl ester (L-NAME, a NOS inhibitor) blocked the bone-sparing effects of the steroid hormone. In contrast to these observations regarding basal NO production, Van Bezooijen et al. [251] reported that 17-estradiol treatment of mouse-immortalized MC-3T3-E1 osteoblasts suppressed cytokine-induced (NOS-2 mediated) NO synthesis. This finding may reflect the generally antagonistic nature of estrogens toward cytokine action (i.e., IL-1 and TNF-) in the skeleton. Finally, pretreatment of hOBs with 17-estradiol has been reported to increase bradykinin responsiveness as measured by the release of arachindonic acid from the cells [252]. However, because bradykinin stimulates bone re-
KOMM AND BODINE
sorption, the physiological significance of this observation is unclear. 6. MISCELLANEOUS RESPONSES As outlined at the end of Table 7, treatment of several rodent osteoblastic cell models with 17-estradiol has been reported to increase creatine kinase (CK) [171], heat shock protein (HSP)-27 [253] and aspartate aminotransferase (AST), -glutamyl transferase (GGT), lactate dehydrogenase (LDH), and transferrin [254]. However, the physiological or therapeutic significance of these responses is unclear. The upregulation of CK activity by 17-estradiol was also observed in rat bone in vivo, and this may represent another anabolic effect of the steroid [171]. 7. SUMMARY As described in the preceding sections, about a third (15/43) of the estrogenic responses observed in a broad range of in vitro osteoblast and BMSC models are consistent with the suppressive effects of estrogens on bone resorption and bone turnover in vivo. In other instances, such as anabolic effects, a disconnection occurs between the in vitro responses and the in vivo physiology of these steroids. In vivo, increased bone turnover on estrogen depletion is driven primarily by increased osteoclastic bone resorption and the subsequent inadequate ability of osteoblastic bone formation to keep pace with this accelerated bone loss [89 – 91]. However, in vitro studies with osteoblasts are almost always performed with pure cultures of cells (i.e., cloned osteoblastic cell lines) and in the absence of osteoclasts. Consequently, the opportunity for coupling between the two cell types is lost [255]. Thus, in isolation, estrogens appear to exert both stimulatory and inhibitory effects on osteoblastic function. In some in vitro models, such as hFOB/ER9 cells [168] or ROBs [128], these differential effects seem to occur as a result of changes that arise during cellular differentiation. However, it is not known if estrogens have divergent actions on osteoblasts as they undergo maturation in vivo. Another possible explanation for the apparent anabolic effects of estrogens on osteoblasts in vitro is that these may represent a pharmacological response to the steroid and not a physiological one [174].
B. Estrogenic Responses in Osteocytes Only a few of estrogenic responses have been observed in osteocytes, and all of these reports come from in situ studies. In what may well be the first publication on this subject, Whitson [256] described the results of an electron microscopic analysis of metatarsal bones isolated from vehicle and 17-estradiol-n-valerate-treated female rabbits. Although the results were not quantitative, the author noted that the number of tight junctions (possibly gap junctions)
325
CHAPTER 10 Regulation of Bone Cell Function by Estrogens
formed between osteocytes was greater in bones from estrogen-treated animals. Moreover, he suggested that this increased tight junction formation might be related to an accelerated osteogenesis. Twenty-five years later, Tomkinson et al. [257] reported the findings of a clinical study of premenopausal women who were treated with a gonadotropin-releasing hormone (GnRH) analogue for endometriosis. Transiliac biopsies were taken before and after GnRH analogue therapy, which resulted in a dramatic decrease in serum 17-estradiol concentrations. Although osteocyte lacuna density was not affected by the treatment, the percentage of lacunae containing viable osteocytes (as determined by cell-associated lactate dehydrogenase activity) was reduced in all but one of the six patients, suggesting that estrogen deficiency is associated with increased osteocyte apoptosis [133]. Because one of the functions of osteocytes is to serve as mechanosensors [132,134,258], these observations also implied that estrogen deficiency could lead to increased bone fragility (and therefore increased fracture) at weight-bearing skeletal sites with or without an accompanying net bone loss. The same group confirmed this clinical study the following year using OVX rats [259]. In this preclinical model of estrogen deficiency, OVX increased the number of apoptotic osteocytes (as determined by DNA strand fragmentation) in both trabecular and cortical bone of the tibia. In addition, repletion with 17-estradiol reversed this increase and returned the apoptotic index to sham values. In another in situ study of OVX rats, lkeda et al. [260] observed that osteopontin mRNA expression increased after OVX in osteocytes that were located in metaphyseal trabecular bone of the femur, but not in those found in the epiphysis. Because osteopontin is one of the bone matrix proteins to which osteoclasts are known to bind [104], these data suggested a possible role for osteocytes in regulating bone resorption. Our laboratory has also presented evidence that osteocytic cells may play a role in modulating osteoclastic activity [261]. Using a conditionally immortalized human preosteocytic (i.e., osteoid osteocyte) cell line (HOB-01C1), we showed that these cells secrete high amounts of IL-6 and monocyte chemoattractant protein (MCP)-1 in response to treatment with the bone-resorbing cytokines IL-1 and TNF-. Together, IL-6 and MCP-1, in addition to other factors, might stimulate osteoclast differentiation and recruitment to a specific bone-remodeling site. Another potential regulatory target for estrogens in osteocytes is ER. Using immunofluorescence to study ER protein expression in human bone biopsies, Braidman and colleagues [124,136] reported that ERT/HRT increases the number of ER protein-positive osteocytes and osteoblasts. Curiously, the number of ER mRNA-positive osteoblasts was observed to decline with ERT/HRT [136]. As noted previously, osteocytes are postulated to serve as mechansenors [132,134,258]. As such, they are thought
to translate the effects of weight bearing or weightlessness into either increases or decreases in bone mineral density, respectively. Several studies suggest that estrogens regulate the process of mechanosensory stimulation, and that mechanical strain and estrogen action may share common signaling pathways. Using organ cultures of rat ulnae isolated from female rats, Cheng et al. [262,263] reported that both 17-estradiol and mechanical loading stimulated [3H]thymidine and [3H]proline incorporation into the bones. Moreover, when the treatments were combined, a synergistic effect was observed. Thus, estrogens appeared to enhance the osteogenic response of the bones to mechanical strain. A subsequent study by the same group using primary cultures of rat long bone-derived osteoblasts demonstrated that both 17-estradiol and mechanical strain increase cellular DNA synthesis [167]. Furthermore, these increases were suppressed by cotreatment with the antiestrogen ICI-182,780. Although osteoblasts are probably not the targets for mechanical loading in vivo [258], these results nevertheless suggest that mechanical strain can activate the ER. The observation that mechanical strain and estrogens appear to share common signal transduction pathways is supported by an in vivo study of Westerlind et al. [264]. Using OVX rats, these authors showed that estrogen deficiency resulted in a preferential lose of cancellous bone from a site that experiences low mechanical strain (distal femur metaphysis), whereas one that experiences high-strain energies (distal femur epiphysis) does not lose bone (even though bone turnover was increased at both sites). In addition, increased mechanical loading (treadmill exercise) suppressed OVX-induced cancellous bone loss from the proximal tibial metaphysis. Conversely, treatment of OVX animals with 17-estradiol suppressed tibial cancellous bone loss that resulted from decreased mechanical loading (unilateral sciatic neurotomy). Finally, there is also evidence that these preclincal findings may translate to humans. For example, in a small clinical study of postmenopausal women, Kohrt et al. [265] reported that HRT and weight-bearing exercise had an additive effect on total body bone mineral accretion. Thus, the efficacies of HRT and weight-bearing exercise on the skeleton seem to be enhanced by concurrent use. Although the just-mentioned studies do not specifically address the role of estrogens in osteocyte biology per se, the implications of this work is that osteocytes — as the major mechanosensory cell in bone — are at least one of the targets for these effects.
C. Estrogenic Responses in Cells of the Osteoclast Lineage In addition to indirectly inhibiting bone resorption through cells of the osteoblast lineage, estrogens have also
326 been reported to have direct suppressive effects on cells of the osteoclast lineage [142]. The most extensive evidence for a direct inhibitory effect of estrogens on mature osteoclasts comes from the work of Oursler [142]. Using both avian and hGCT-derived osteoclasts that were highly purified (90% homogeneous) with an osteoclast-specific monoclonal antibody (121F), Oursler reported that 17-estradiol inhibits in vitro bone resorption by these preparations [145,146,150,266 – 268]. Estrogenic responses in these studies include the upregulation of c-fos, c-jun, TGF-2, TGF3, and TGF-4 mRNA levels; the downregulation of tartrate-resistant acid phosphatase (TRAP), cathepsin B, cathepsin D, LEP-100, and lysozyme message levels; the induction of total TGF- protein secretion (due mostly to an increase in TGF-3); and the suppression of TRAP, cathepsin B, cathepsin L, and -glucuronidase activity, as well as lysozyme protein production. The majority of these effects are consistent with an estrogen-mediated decrease in osteoclast activity and subsequent bone resorption. For example, TGF- is an inhibitor of bone resorption, whereas lysozomal proteases like the cathepsins are involved in digesting the bone matrix [142]. Confirmation that estrogens suppress osteoclastic gene expression in vivo comes from the studies of Zheng et al. [269], who demonstrated that treatment of OVX rats with 17-estradiol decreased the expression of TRAP mRNA in bone. Additional support for a direct effect of estrogens on osteoclasts comes from the work of Sunyer et al. [147]. Employing normal human osteoclasts (hOCLs) that were also purified to 90% homogeneity with the 121F monoclonal antibody, these authors reported that 17-estradiol decreased the mRNA levels of the signaling receptor for IL-1 (IL-1RI) and increased the message levels of the IL-1 decoy receptor (IL-1RII). This change in receptor expression correlated with a suppression of IL-1-mediated IL-8 expression by the steroid hormone. Moreover, 17estradiol pretreatment abrogated the reduction of hOCL apoptosis by IL-1. Finally, Mano et al. [148] have demonstrated that 17-estradiol also inhibits the in vitro bone resorption of purified rabbits osteoclasts and reduces the expression of cathepsin K mRNA by these cells. However, some studies have failed to detect a direct inhibitory effect of estrogens on mature osteoclasts. For example, Williams et al. [270] were unable to suppress bone resorption of purified avian osteoclasts with either 17estradiol or DES. However, high ( M) levels of the TSE/ SERM tamoxifen decreased osteoclast activity. Likewise, calmodulin antagonists had a similar effect. Additional experiments led the authors to conclude that tamoxifen acted through a membrane-associated target to suppress osteoclastic bone resorption independently of the ER. This target appeared to be similar or related to the target for the calmodulin inhibitors. As indicated earlier and in preceding sections, estrogens have been observed to increase the expression of TGF- by
KOMM AND BODINE
both osteoblasts and osteoclasts. In addition, these steroids suppress osteoblast apoptosis, but enhance programmed osteoclast cell death [271,272]. Hughes et al. [273] elegantly demonstrated a connection among estrogens, TGF-, and osteoclast apoptosis. These authors showed that treatment of marrow culture-derived murine osteoclasts with 17 estradiol increased the percentage of cells undergoing apoptosis. Likewise, treatment of the cultures with TGF-1 also increased osteoclast apoptosis. Moreover, the induction of osteoclast programmed cell death by 17-estradiol was blocked by coincubation with a pan-specific TGF- antibody. Consistent with its bone-sparing effects [201], treatment of the osteoclast-containing cultures with tamoxifen also increased apoptosis of these cells. These in vitro observations were confirmed by an in vivo study in which OVX mice were treated with 17-estradiol. Because the marrow culture system used by Hughes et al. [273] was a heterogeneous cell population, the promotion of osteoclast apoptosis by 17-estradiol could have resulted from either a direct action of the steroid on osteoclasts or an indirect effect on another cell type like osteoblasts or BMSCs. In addition to inducing apoptosis of mature osteoclasts, estrogens may also have similar effects on osteoclast progenitors. Zecchi-Orlandini et al. [274] reported that 17 estradiol induced apoptosis of the human monoblastic leukemia cell line FLG 29.1, which has characteristics resembling preosteoclasts. Moreover, treatment of this cell line with the TSE/SERM raloxifene [201] also induced apoptosis [275]. FLG 29.1 cells can be stimulated to form osteoclast-like cells in vitro by treatment with phorbol ester, 1,25(OH)2vitamin D, or osteoblast-derived factors [276]. These agents also induce the expression of a novel superoxide dismutase-related membrane glycoprotein, which is the osteoclast-specific antigen that is recognized by the 121F monoclonal antibody. Incubation of cells with 17-estradiol suppresses the induction of this antigen by phorbol ester [276]. Thus, these results suggest that estrogens may also suppress osteoclast differentiation by acting directly on their progenitors. Additional reports also indicate that estrogens can suppress osteoclast differentiation. Schiller et al. [277] have demonstrated that 17-estradiol antagonizes the induction of osteoclast-like cell formation by 1,25(OH)2 vitamin D in primary cultures of mouse bone marrow cells. In addition, these authors showed that the ability of 1,25(OH)2 vitamin D to stimulate osteoclast differentiation is at least partially mediated by an upregulation of IL-6 secretion and that 17-estradiol blocks this effect as well. Estrogens also suppresses PTH-stimulated osteoclast formation. Using primary mouse hemopoietic blast cell cultures, which were reportedly free of stromal cells and osteoblasts, Kanatani et al. [152] presented evidence that these osteoclast precursors contain PTH receptor mRNA based on RT-PCR. These cells also express ER message. Treatment of mouse
327
CHAPTER 10 Regulation of Bone Cell Function by Estrogens
hemopoietic blast cell cultures with either 1,25(OH)2 vitamin D or PTH (1 – 34) induces the formation of osteoclastlike cells (i.e., TRAP-positive multinucleated cells). However, cotreatment of the cultures with 17-estradiol blunts the stimulation of osteoclast differentiation by PTH, but not by 1,25(OH)2 vitamin D. These authors also demonstrated that 17-estradiol blocks osteoclast-like cell formation induced by agents that activate adenylyl cyclase or mimic cAMP, but not ones that activate protein kinase C or increase intracellular calcium. Although an earlier report from the same group suggested that estrogens suppress PTH-induced osteoclast differentiation indirectly through an effect on osteoblasts [235], the study by Kanatani et al. [152] concluded that this inhibitory effect may also be due to a direct action on osteoclast progenitor cells. In summary, there is substantial evidence to conclude that estrogens inhibit osteoclast differentiation and activity in two ways: one that occurs indirectly via the osteoblast and stromal cell and a second that occurs directly through interaction with the ER in osteoclast progenitors and mature osteoclasts. However, as with other aspects of estrogen action on bone cells, this area of research is controversial.
D. Estrogenic Responses in Chondrocytes Another important target cell in the skeleton for estrogens is the chondrocyte. As noted in an earlier section, these cells have been shown to express both ER and ER. Moreover, chondrocytes have also been reported to exhibit estrogenic responses. In vivo, it is well known that estrogens accelerate endochondral growth during puberty and potentiate epiphyseal closure at the end of the growth spurt [89] (see Chapter 25). Consistent with these physiological responses, 17-estradiol has been observed to decrease the in vitro proliferation and/or DNA synthesis of embryonic duck [278] and rat chondrocytes [279]. In duck chondrocytes, 17-estradiol also suppressed sulfated proteoglycan synthesis [278], whereas in fetal rabbit [280] and human chondrocytes [281], the steroid had the opposite effect. Additional in vitro estrogenic effects in rat chondrocytes include the upregulation of alkaline phosphatase activity and collagen production, which are consistent with a potentiation of cellular differentiation by the steroid [279].
VIII. ESTROGEN-RELATED RECEPTOR- AND OSTEOPONTIN GENE EXPRESSION In addition to expressing ER and ER, osteoblasts also express a related member of the nuclear receptor superfamily known as estrogen-related receptor (ERR)-1 or [282 – 284]. ERR- is an orphan receptor that shares
68% amino acid identity with ER and ER in the DNAbinding domain, but only 36% identity in the ligand-binding domain [284]. Consequently, it does not bind 17estradiol but instead is constitutively active in serum-containing medium [284]. However, this constitutive activity is diminished upon charcoal treatment of the serum [284]. ERR-, as well as the related ERR-, transactivates promoters containing either an ERE or a SF-1-response element (SFRE) [284]. ER also binds to both of these DNA response elements, whereas ER does not bind to the SFRE [284]. ERR- mRNA is highly expressed in the ossification zones of the developing mouse skeleton (long bones, vertebrae, ribs, and skull), as well as in some human osteosarcoma cell lines (HOS-TE85 and SaOS-2) and hOBs [282]. Given this expression pattern, as well as the knowledge that the osteopontin promoter contains an SFRE, it is perhaps not surprising that cotransfection of rat ROS 17/2.8 osteosarcoma cells with ERR and an osteopontin promoter – reporter gene construct resulted in the transactivation of this promoter [282 – 284]. Moreover, transient transfection of ROS 17/2.8 cells and immortalized mouse MC-3T3-E1 cells with ERR produced an upregulation of endogenous osteopontin mRNA levels [283]. Taken together, these data demonstrate that osteopontin gene expression in the osteoblast is regulated not only by ER in an estrogen-dependent manner, but also by ERR in an estrogen-independent manner [284]. In contrast, ER does not appear to regulate this gene [284]. Thus, these observations also point to a potential functional difference between the biological roles of ER and ER in the osteoblast. However, because osteopontin is an apparent binding site for osteoclasts to the bone matrix [104], the physiological significance of its upregulation by estrogens via either ER or ERR- in a ligand-independent manner is unclear.
IX. NONGENOMIC ACTIONS OF ESTROGENS IN BONE CELLS Although the majority of estrogenic effects are believed to be mediated by one of the nuclear ERs, some responses may also originate at the plasma membrane [285,286]. Estrogens have been reported to produce rapid effects (within seconds or minutes) on a variety of cell types, including bone cells [285,286]. These nongenomic actions are thought to be mediated via a membrane receptor. However, it is unclear whether this receptor is a membrane-localized form of a nuclear ER or if it is a distinct transmembrane protein, such as a G-protein-coupled receptor (GPCR) [285,286]. In a series of papers using primary female rat osteoblasts, Lieberherr and co-workers presented convincing evidence for rapid, membrane-derived effects of 17estradiol [287 – 289]. Treatment of ROB cells with low con-
328 centrations (1 pM – 1 nM) of 17-estradiol increased intracellular calcium concentrations within 10 to 30 [287]. Through the use of various inhibitors, the source of this calcium was shown to be both extracellular, via plasma membrane channels, and intracellular from the endoplasmic reticulum or calciosome. Cells within the same time frame also produced IP3 and diacylglycerol (DAG) after treatment with the steroid. Because inhibitors of both phospholipase C (PLC) and Gi proteins blocked the release of IP3 and DAG, the authors concluded that 17-estradiol acted through a GPCR [287]. Consistent with estrogens working through a distinct membrane receptor and not simply a membrane-localized ER, tamoxifen was neither an agonist nor an antagonist of 17-estradiol. Subsequent studies by this group refined the model to include activation of PLC2 by subunits [288,289]. In contrast, 1,25(OH)2 vitamin D, which also has rapid effects on female ROB cells, was shown to act via modulation of PLC-1 by G (q/11) [288,289]. A potential downstream target for the rapid generation of a membrane-derived signal by 17-estradiol was reported by Endoh et al. [290]. These authors showed that treatment of ROS 17/2.8 cells with 17-estradiol activated the mitogen-activated protein kinase (MAPK) within 5 min. Estrogens may also produce rapid nongenomic effects in cells of the osteoclast lineage [41,291 – 293]. For example, using the human preosteoclastic cell line FLG 29.1, Fiorelli et al. [61] demonstrated that 17-estradiol stimulated an increase in intracellular pH within 50, as well as an increase in intracellular cAMP and cGMP after 30 min. In addition, Brubaker and Gay [293] reported that treatment of isolated avian osteoclasts with 17-estradiol caused a depolarization of the plasma membrane potential within seconds of adding the steroid to the cells. The mechanism for the depolarization appeared to be due to the regulation of potassium channel activity. The net effect of this rapid, nongenomic estrogenic response could be an inhibition of osteoclastic acidification.
X. SUMMARY Estrogens clearly play a critical role in bone biology. The burst in research aimed at elucidating the functional role of estrogens in bone remodeling that has occurred since the mid-1970s has led to the discovery of a multitude of potential pathways that are affected by estrogens in the skeleton. The sheer abundance of estrogenic-related regulated events in bone cells supports the contention that estrogens, working through their receptors, play key roles in the development and maintenance of a normal skeleton. Questions that remain to be answered relate to the differences in the skeletal response to the various types of estrogens (estradiol vs phytoestrogens vs SERMS). All estrogens do not evoke the same response in bone whether looking at a
KOMM AND BODINE
specific gene’s regulation in isolated osteoblasts or a global skeletal response in vivo. Why this occurs is not known. The complexity of the bone-remodeling process coupled with multiple sites where an estrogen could elicit an effect will make it difficult to fully answer the question, but as technology advances, so will the possibility of answering our tough questions.
References 1. F. Albright, P. H. Smith, and A. M. Richardson, Postmenopausal osteoporosis: Its clinical features. JAMA 116, 2465 – 2474 (1941). 2. A. W. Diddle and I. Q. Smith, Postmenopausal osteoporsis: The role of estrogens. South. Med. J. 77, 868 – 874 (1984). 3. M. C. Chapuy and P. J. Meunier, Prevention and treatment of osteoporosis. Aging Clin. Exp. Res. 7, 164 – 173 (1995). 4. J. P. Bilezikian, Estrogens and postmenopausal osteoporosis: Was Albright right after all? J. Bone Miner. Res. 13, 774 – 776 (1998). 5. R. Eastell, Treatment of postmenopausal osteoporosis. Engl. J. Med. 338, 736 – 746 (1998). 6. A. Horsman, M. Jones, R. Francis, and C. Nordin, The effect of estrogen dose on postmenopausa bone loss. N. Engl. J. Med. 309, 1404 – 1407 (1983). 7. T. M. Willson, Chemistry: Structure and function relationships. Estrogen Antiestrogens Basic Clin. Aspects 21 – 22 (1997). 8. M. O Connell, Pharmacokinetic and pharmacologic variation between different estrogen products. J. Clin. Pharmacol. 35, 18s – 24s (1995). 9. H. U. Bryant and W. H. Dere, Selective estrogen receptor modulators: An alternative to hormone replacement therapy (44204). P.S.E.B.M. 217, 45 – 52 (1998). 10. T. A. Grese and J. A. Dodge, Selective estrogen receptor modulators (SERMs). Curr. Pharm. Design 4, 71 – 92 (1998). 11. V. Kumar, S. Green, G. Stack, M. Berry, J.-R. Jin, and P. Chambon, Functional domains for the human estrogen receptor. Cell. 51, 941 – 951 (1987). 12. S. Green, P. Walter, G. Greene, A. Krust, C. Goffin, E. Jensen, G. Scrace, M. Waterfield, and P. Chambon, Cloning of the human oestrogen receptor cDNA. J. Steroid Biochem. 24, 77 – 83 (1986). 13. P. Walter, S. Green, G. Greeene, A. Krust, J.-M. Bornert, J.-M. Jeltsch, A. Staub, E. Jensen, G. Scrace, M. Waterfield, and P. Chambon, Cloning of the human estrogen receptor cDNA. Proc. Natl. Acad. Sci. USA 82, 7889 – 7893 (1985). 14. G. J. M. Kuiper, E. Enmark, M. Pelto-Huikko, S. Nilsson, and J. Gustafsson, Cloning of a novel estrogen receptor expressed in rat prostate and ovary. Proc. Natl. Acad. USA 93, 5925 – 5930 (1996). 15. S. Mosselman, J. Polman, and R. Dijkema, ER-beta: Identification and characterization of a novel human estrogen receptor. FEBS Lett. 392, 49 – 53 (1996). 16. B. O Malley, The steroid receptor superfamily: More excitement predicted for the future. Mol. Endocrinol. 4, 363 – 369 (1990). 17. C. Weinberger, V. Giguere, S. Hollenberg, M. G. Rosenfeld, and R. M. Evans, Human steroid receptors and erbA proto-oncogene products: Members of a new superfamily of enhancer binding proteins. Cold Spring Harbor Symp. Quant. Biol. 51, 759 – 772 (1986). 18. Q. Zhao, S. Khorasanizadeh, Y. Miyoshi, M. A. Lazar, and F. Rastinejad, Structural elements of an orphan nuclear receptorDNA complex. Mol. Cell. 1, 849 – 861 (1998). 19. B. Blumberg and R. M. Evans, Orphan nuclear receptors: New ligands and new possibilities. Genes Dev. 12, 3149 – 3155 (1998). 20. R. Kumar and E. Thompson, The structure of the nuclear hormone receptors. Steroids 64, 310 – 319 (1999).
CHAPTER 10 Regulation of Bone Cell Function by Estrogens 21. L. Klein-Hitpass, S. Y. Tsai, G. L. Greene, J. H. Clark, M.-J. Tsai, and B. W. O’Malley, Specific binding of estrogen receptor to the estrogen response element. Mole. Cell. Biol. 9, 43 – 49 (1989). 22. G. Klock, U. Strahle, and G. Schutz, Oestrogen and glucocorticoid responsive elements are closely related by distinct. Nature 329, 734 – 736 (1987). 23. M. T. Tzukerman, A. Esty, D. Santiso-Mere, P. Danielian, M. G. Parker, R. B. Stein, J. W. Pike, and D. P. McDonnell, Human estrogen receptor transactivational capacity is determined by both cellular and promoter context and mediated by two fuctionally distinct intramolecular regions. Mol. Endocrinol. 8, 21 – 30 (1994). 24. K. Inno, M. Haino, M. Iwasaki, N. Ono, T. Horigome, and H. Sugano, Reconstruction of the 9 S estrogen receptor with heat shock protein 90. FEBS. Lett. 267, 157 – 159 (1990). 25. J. Devin-Leclerc, X. Meng, F. Delahaye, P. Leclerc, E.-E. Baulieu, and M.-G. Catelli, Interaction and dissociation by ligands of estrogen receptor and Hsp90: The antiestrogen RU 58668 induces a protein synthesis-dependent clustering of the receptor in the cytoplasm. Mol. Endocrinol. 12, 842 – 854 (1998). 26. V. Kumar and P. Chambon, The estrogen receptor binds tightly to its responsive element as a ligand-induced homodimer. Cell 55, 145 – 156 (1988). 27. F. Auricchio, Phosphorylation of steroid receptors. J. Steroid Biochem. 32, 613 – 622 (1989). 28. W. W. Grody, W. T. Schrader, and B. W. O Malley, Activation transformation, and subunit structure of steroid hormone receptors. Endocr. Rev. 3, 141 – 163 (1982). 29. B. Katzenellenbogen, B. Bhardwaj, H. Fang, B. Ince, F. Pakdel, J. Reese, D. Schodin, and C. Wrenn, Hormone binding and transcription activation by estrogen receptors: Analyses using mammalian and yeast systems. J. Steroid Biochem. Mol. Biol. 47, 39 – 48 (1993). 30. N. J. McKenna, R. B. Lanz, and B. W. O Malley, Nuclear receptor coregulators: Cellular and molecular biology. Endocr. Rev. 20, 321 – 344 (1999). 31. J. D. Chen and H. Li, Coactivation and corepression in transcriptional regulation by steroid/nuclear hormone receptors. Crit. Rev. Eukary. Gene Express. 8, 169 – 190 (1998). 32. P. Clement-Lacroix, C. Ormandy, L. Lepescheux, P. Ammann, D. Damotte, V. Goffin, B. Bouchard, M. Amling, M. Gaillard-Kelly, N. Binart, R. Baron, and P. A. Kelly, Osteoblasts are a new target for prolactin: Analysis of bone formation in prolactin receptor knockout mice. Endocrinology 140, 96 – 105 (1999). 33. N. Huang, E. vom Baur, J.-M. Garnier, T. Lerouge, J.-L. Vonesch, Y. Lutz, P. Chambon, and R. Losson, Two distinct nuclear receptor interaction domains in NSD1, a novel SET protein that exhibits characteristics of both copressors and coactivators. EMBO J. 17, 3398 – 3412 (1998). 34. R. M. Lavinski, K. Jepsen, T. Heinzel, J. Torchia, T.-M. Mullen, R. Schiff, A. L. Del-Rio, M. Ricote, S. Ngo, J. Gemsch, S. G. Hilsenbeck, C. K. Osborne, C. K. Glass, M. G. Rosenfeld, and D. W. Rose, Diverse signaling pathways modulate nuclear receptor recruitment of N-CoR and SMRT complexes. Proc. Natl. Acad. Sci. USA 95, 2920 – 2925 (1998). 35. C. M. Klinge, Estrogen receptor interaction with co-activators and co-repressors. Steroids 65, 227 – 251 (2000). 36. J. T. Kadonaga, Eukaryotic transcription: An interlaced network of transcription factors and chromatin-modifying machines. Cell 92, 307 – 313 (1998). 37. E. Korzus, J. Torchia, D. W. Rose, L. Xu, R. Kurokawa, E. M. Mclnerney, T. Mullen, C. K. Glass, and M. G. Rosenfeld, Transcription factor-specific requirements for coactivators and their acetyltransferase functions. Science 279, 703 – 707 (1998). 38. H. Y. Mak, S. Hoare, P. M. A. Henttu, and M. G. Parker, Molecular determinants of the estrogen receptor-coactivator interface. Mol. Cell. Biol. 19, 3895 – 3903 (1999).
329 39. P. M. A. Henttu, E. Kalkhoven, and M. G. Parker, AF-2 activity and recruitment of steroid receptor coactivator 1 to the estrogen receptor depend on a lysine residue conserved in nuclear receptors. Mol. Cell. Biol. 17, 1832 – 1839 (1997). 40. A. K. Shiau, D. Barstad, P. M. Loria, L. Cheng, P. J. Kushner, D. A. Agard, and G. L. Greene, The structural basis of estrogen receptor/coactivator recognition and the antagonism of this interaction by tamoxifen. Cell. 95, 927 – 937 (1998). 41. W. Feng, R. C. J. Ribeiro, R. L. Wagner, H. Nguyen, J. W. Apriletti, R. J. Fletterick, J. D. Baxter, P. J. Kushner, and B. L. West, Hormone-dependent coactivator binding to a hydrophobic, cleft on nuclear receptors. Science 280, 1747 – 1749 (1998). 42. W. L. Kraus and J. T. Kadonaga, p300 and estrogen receptor cooperatively activate transcription via differential enhancement of initiation and reinitiation. Genes Dev. 12, 331 – 342 (1998). 43. C. Rachez, Z. Suldan, J. Ward, C.-P. B. Chang, D. Burakov, H. Erdjument-Bromage, P. Tempst, and L. P. Freedman, A novel protein complex that interacts with the vitamin D3 receptor in a liganddependent manner and enhances VDR transactivation in a cell-free system. Genes. Dev. 12, 1787 – 1800 (1998). 44. L. P. Freedman, Increasing the complexity of coactivation in nuclear receptor signaling. Cell 97, 5 – 8 (1999). 45. M. Ito, C.-X. Yaun, S. Malik, W. Gu, J. D. Fondell, S. Yamamura, Z.-Y. Fu, X. Zhang, J. Qin, and R. G. Roeder, Identity between TRAP and SMCC complexes indicates novel pathways for the function of nuclear receptors and diverse mammalian activators. Mol. Cell 3, 361 – 370 (1999). 46. S. W. Curtis, T. Washburn, C. Sewall, R. DiAugustine, J. F. Couse, and K. S. Korach, Cross-talk between growth factors and estrogen receptor. Estrogen Vessel Wall 31 – 41. 47. B. S. Katzenellenbogen and M. J. Norman, Multihormonal regulation of the progesterone receptor in MCF-7 human breast cancer cells: Interrelationships among insulin/insulin-like growth factor-I, serum, and estrogen. Endocrinology 126, 891 – 898 (1990). 48. P. Joel, A. Traish and D. Lannigan, Estradiol-induced phosphorylation of serine 118 in the estrogen receptor is independent of p42/p44 mitogen-activated protein kinase. J. Biol. Chem. 273, (1998). 49. H. Cho and B. Katzenellenbogen, Synergistic activation of estrogen receptor-mediated transcription by estradiol and protein kinase activators. Mol. Endocrinol. 7, 441 – 452 (1993). 50. S. Migliccio, S. Bernardini, W. Westsel, K. Korach, T. Faraggiana, and A. Teti, Protein kinase C modulates estrogen receptors in differentiated osteoblastic cells in vitro. Steroids 63, 352 – 354 (1998). 51. D. Ignar-Trowbridge, M. Pimentel, M. Parker, J. McLachlan, and K. Korach, Peptide growth factor cross-talk with the estrogen receptor requires the A/B domain and occurs independently of protein kinase C or estradiol. Endocrinology 137, 1735 – 1744 (1996). 52. D. Ignar-Trowbridge, M. Pimentel, C. Teng, K. Korach, and J. McLachlan, Cross talk between peptide growth factor and estrogen receptor signaling systems. Environ Health Perspect. 103, 35 – 38 (1995). 53. T. Improta-Brears, A. Whorton, F. Codazzi, J. York, T. Meyer, and D. McDonnell, Estrogen-induced activation of mitogen-activated protein kinase requires mobilization of intracellular calcium. Proc. Natl. Acad. Sci. USA 96, 4686 – 4691 (1999). 54. S. Manolagas, The role of IL-6 type cytokines and their receptors in bone. Ann. N. Y. Acad. Sci. 98, 194 – 204 (1998). 55. T. Bellido, C. O’Brien, P. Roberson, and S. Manolagas, Transcriptional activation of the p21WAF1, C1P1, SD11 gene by interluekin-6 type cytokines. J. Biol. Chem. 273, 21137 – 21144 (1998). 56. G. Cerillo, A. Rees, N. Manchanda, C. Reilly, I. Brogan, A. White, and M. Needham, The oestrogen receptor regulates NFB and AP-1 activity in a cell-specific manner. J. Steroid. Biochem. Mol. Bio. 67, 79 – 88 (1998).
330 57. P. Webb, G. Lopez, R. Uht, and P. Kushner, Tamoxifen activation of the estrogen receptor/AP-1 pathway: Potential origin for the cellspecific estrogen-like effects of antiestrogens. Mol. Endocrinol. 9, 443 – 456 (1995). 58. C. Qin, P. Singh, and S. Safe, Transcriptional activation of insulinlike growth factor-binding protein-4 by 17-estradiol in MCF-7 cells: Role of estrogen receptor-Sp1 complexes. Endocrinology 140, 2501 – 2508 (1999). 59. Y. Sadovsky and P. Crawford, Developmental and physiological roles of the nuclear receptor steroidogenic factor-1 in the reproductive system. J. Soc. Gynecol. Invest. 5, 6 – 12 (1998). 60. J. Bertherat, The nuclear receptor SF-1 (steroidogenic factor-1) is no longer an orphan. Eur. J. Endocrinol. 138, 32 – 33 (1998). 61. G. Fiorelli, F. Gori, U. Frediani, F. Franceschelli, A. Tanini, C. TostiGuerra, S. Benvenuti, L. Gennari, L. Becherini, and M. L. Brandi, Membrane binding sites and non-genomic effects of estrogen in cultured human pre-osteoclastic cells. J. Steroid Biochem. Mol. Biol. 59, 233 – 240 (1996). 62. P. S. Cooke, D. L. Buchanan, D. B. Lubahn, and G. R. Cunha, Mechanism of estrogen action: Lessons from the estrogen receptoralpha knockout mouse. Biol. Reprod. 59, 470 – 475 (1998). 63. P. Monje and R. Boland, Characterization of membrane estrogen binding proteins from rabbit uterus. Mol. Cell. Endocrinol. 147, 75 – 84 (1999). 64. M. Razandi, A. Pedram, G. Greeene, and E. Levin, Cell membrane and nuclear estrogen receptors (ERs) originate from a single transcript: Studies of ER and ER expressed in Chinese hamster ovary cells. Mol. Endocrinol. 13, 307 – 319 (1999). 65. W. Rosner, D. Hryb, M. Khan, A. Nakhla, and N. Romas, Sex hormone-binding globulin binding to cell membranes and generation of a second messenger. J. Androl. 13, 101 – 106 (1992). 66. E. Colin, G. Van Den Bemd, M. Van Aken, S. Christakos, H. De Jonge, H. DeLeuca, J. Prahl, J. Birkenhager, C. Buurman, H. Pols, and J. Van Leeuwen, Evidence for involvement of 17-estradiol in intestinal calcium absorption independent of 1,25-dihydroxyvitamin D3 level in the rat. J. Bone. Miner. Res. 14, 57 – 64 (1999). 67. M. Valverde, P. Rojas, J. Amigo, D. Cosmelli, P. Orio, M. Bahamonde, G. Mann, C. Vergara, and R. Latorre, Acute activation of Maxi-K channels (hSlo) by estradiol binding to the subunit. Science 285, 1929 – 1931 (1999). 68. G. Tremblay, A. Tremblay, N. Copeland, D. Gilbert, N. Jenkins, F. Labrie, and V. Giguere, Cloning, chromosomal localization, and functional anaysis of the murine estrogen receptor . Mol. Endocrinol 11, 353 – 365 (1997). 69. E. Enmark, P. Pelto-Huikko, K. Grandien, S. Lagercrantz, J. Lagercrantz, G. Fried, M. Nordenskjold, and J.-A. Gustafsson, Human estrogen receptor -Gene structure, chromosomal localization, and expression pattern. J. Clin. Endocrinol. Metab. 82, 4258 – 4265 (1997). 70. R. Bhat, D. Harnish, P. Stevis, C. Lyttle, and B. Komm, A novel human estrogen receptor : Identification and functional analysis of aditional N-terminal amino acids. J. Steroid Biochem. Mol. Biol. 67, 233 – 240 (1998). 71. D. Chen, P. Pace, R. Coombes, and S. Ali, Phosphorylation of human estrogen receptor by protein kinase A regulates dimerization. Mol. Cell. Bio. 19, 1002 – 1015 (1999). 72. P. Shughrue, M. Lane, P. Scrimo, and I. Mechenthaler, Comparative distribution of estrogen receptor- (ER-) and (ER-) mRNA in the rat pituitary, gonad, and reproductive tract. Steroids 63, 498 – 504 (1998). 73. S. Cowley, S. Hoare, S. Mosselman, and M. Parker, Estrogen receptors and form heterodimers on DNA. J. Biol. Chem. 272, 19858 – 19862 (1997). 74. G. Kuiper, S. Nilsson, and J.-A. Gustafsson, Characteristics and function of the novel estrogen receptor . Horm. Signal. 1, 89 – 112 (1998).
KOMM AND BODINE 75. J. Sun, M. Meyers, B. Fink, R. Rajendran, J. Katzenbellenbogen, and B. Katzenbellenbogen, Novel ligands that function as selective estrogens or antiestrogens for estrogen receptor- or estrogen receptor-. Endocrinology 140, 800 – 804 (1999). 76. A. Brzozowski, A. Pike, Z. Dauter, R. Hubbard, T. Bonn, O. Engstrom, L. Ohman, G. Greene, J.-A. Gustafsson, and M. Carlquist, Molecular basis of agonism and antagonism in the oestrogen receptor. Nature. 389, 753 – 758 (1997). 77. A. Pike, A. Brzozowski, R. Hubbard, T. Bonn, A.-G. Thorsell, O. Engstrom, J. Ljunggren, J.-A. Gustafsson, and M. Carlquist, Structure of the ligand-binding domain of oestrogen receptor beta in the presence of a partial agonist and a full antagonist. EMBO J. 18, 4608 – 4618 (1999). 78. W. Gradishar and V. Jordan, Clinical potential of new antiestrogens. J. Clin. Oncol. 15, 840 – 852 (1997). 79. R. Love, H. Barden, R. Mazess, S. Epstein, and R. Chappell, Effect of tamoxifen on lumbar spine bone mineral density in postmenopausal women after 5 years. Arch. Intern. Med. 154, 2585 – 2588 (1994). 80. C. A. Frolik, H. U. Bryant, E. C. Black, D. E. Magee, and S. Chandrasekhar, Time-dependent changes in biochemical bone markers and serum cholesterol in ovariectomized rats: Effects of raloxifene HC1, tamoxifen, estrogen, and alendronate. Bone 18, 621 – 627 (1996). 81. P. Carthew, R. Edwards, and B. Nolan, Uterotrophic effects of tamoxifen, toremifene, and raloxifene do not predict endometrial cell proliferation in the ovariectomized CD1 mouse. Toxicol. Appl. Pharmacol. 158, 24 – 32 (1999). 82. T. Nickelsen, E. G. Lufkin, B. L. Riggs, D. A. Cox, and T. H. Crook, Raloxifene hydrochloride, a selective estrogen receptor modulator: Safety assessment of effects on cognitive function and mood in postmenopausal women. Psychoneuroendocrinology 24, 115 – 128 (1999). 83. J. Norris, L. Paige, D. Christensen, C.-Y. Chang, M. Huacani, D. Fan, P. Hamilton, D. Fowlkes, and D. Mcdonnell, Peptide antagonists of the human estrogen receptor. Science 285, 744 – 746 (1999). 84. D. Lubahn, J. Moyer, T. Golding, J. Couse, K. Korach, and O. Smithies, Alteration of reproductive function but not prenatal sexual development after insertional disruption of the mouse estrogen receptor gene. Proc. Natl. Acad. Sci. USA 90, 11162 – 11166 (1993). 85. J. Krege, J. Hodgin, J. Couse, E. Enmark, M. Warner, J. Mahler, M. Sar, K. Korach, J.-A. Gustatsson, and O. Smithies, Generation and reproductive phenotypes of mice lacking estrogen receptor Proc. Natl. Acad. Sci. USA 95, 15677 – 15682 (1998). 86. J. Couse and K. Korach, Estrogen receptor null mice: What have we learned and where will they lead us? Endocrinol. Rev. 20, 358 – 417 (1999). 87. E. Smith, J. Boyd, G. Frank, H. Takahashi, R. Cohen, B. Specker, T. Williams, D. Lubahn, and K. Korach, Estrogen resistance caused by a mutation in the estrogen-receptor gene in a man. Engl. J. Med. 331, 1056 – 1061 (1994). 88. O. Vidal, M. K. Lindberg, K. Hollberg, D. J. Baylink, G. Andersson, D. B. Lubahn, S. Mohan, J.-A. Gustafsson, and C. Ohlsson, Estrogen receptor specificity in the regulation of skeletal growth and maturation in male mice. Proc. Natl. Acad. Sci. USA 97, 5474 – 5479 (2000). 89. R. T. Turner, B. L. Riggs, and T. C. Spelsberg, Skeletal effects of estrogen. Endocr. Rev. 15, 275 – 300 (1994). 90. B. L. Riggs and L. J. I. Melton, The prevention and treatment of osteoporosis. Engl. Med. 327, 620 – 627 (1992). 91. D. W. Dempster and R. Lindsay, Pathogenesis of osteoporosis. Lancet 341, 797 – 805 (1993). 92. B. S. Komm and P. V. Bodine, The ongoing saga of osteoporosis treatment. J. Cell. Biochem. Suppl. 30/31, 277 – 283 (1998).
CHAPTER 10 Regulation of Bone Cell Function by Estrogens 93. B. L. Riggs, S. Khosla, and L. J. Melton, A unitary model for involutional osteoporosis: Estrogen deficiency causes both type I and tyep II osteoporosis in postmenopausal women and contributes to bone loss in aging men. J. Bone Miner. Res. 13, 763 – 773 (1998). 94. R. L. Prince, Counterpoint: Estrogen effects on calciotropic hormones and calcium homeostasis. Endoc. Rev. 15, 301 – 309 (1994). 95. Y. Liel, S. Shany, P. Smirnoff, and B. Schwartz, Estrogen increases 1,25-dihydroxyvitamin D receptors expression and bioresponse in the rat duodenal mucosa. Endocrinology 140, 280 – 285 (1999). 96. M. ten Bolscher, J. C. Netelenbos, R. Barto, L. M. Van Buuren, and W. J. F. Van der Vijgh, Estrogen regulation of intestinal calcium absorption in the intact and ovariectomized adult rat. J. Bone Miner. Res. 14, 1197 – 1202 (1999). 97. C. R. Draper, I. M. Dick, and R. L. Prince, The effect of estrogen deficiency on calcium balance in mature rats. Calcif. Tissue Int. 64, 325 – 328 (1999). 98. M. J. Oursler, M. Kassem, R. Turner, B. L. Riggs, and T. C. Spelsberg, Regulation of bone cell function by gonadal steroids. In “Osteoporosis” (R. Marcus, D. Feldman, and J. Kelsey), pp. 237 – 260. Academic Press, San Diego, 1996. 99. T. C. Spelsberg, M. Subramaniam, B. L. Riggs, and S. Khosla, The actions and interactions of sex steroids and growth factors cytokines on the skeleton. Mol. Endocrinol. 13, 819 – 828 (1999). 100. D. J. Rickard, M. Subramaniam, and T. C. Spelsberg, Molecular and cellular mechanisms of estrogen action on the skeleton. J. Cell. Biochem. Suppl. 32/33, 123 – 132 (1999). 101. R. Bland, Steroid hormone receptor expression and action in bone. Clin. Sci. 98, 217 – 240 (2000). 102. B. L. Riggs and L. J. I. Melton, Involutional osteoporosis. Engl. J. Med. 314, 1676 – 1686 (1986). 103. T. K. Gray, T. C. Flynn, K. M. Gray, and L. M. Nabell, 17beta-estradiol acts directly on the clonal osteoblastic cell line UMR106. Proc. Natl. Acad. Sci. USA 84, 6267 – 6271 (1987). 104. J. B. Lian, G. S. Stein, E. Canalis, P. Gehron Robey, and A. L. Boskey, Bone formation: Osteoblast lineage cells, growth factors, matrix proteins, and the mineralization process. In “Primer on the Metabolic Bone Diseases and Disorders of Mineral Metabolism” (M. J. Favus), pp. 14 – 29. Lippincott Williams & Wilkins, Philadelphia, 1999. 105. B. S. Komm, C. M. Terpening, D. J. Benz, K. A. Graeme, A. Gallegos, M. Korc, G. L. Greene, B. W. OMalley, and M. R. Haussler, Estrogen binding, receptor mRNA, and biologic response in osteoblast-like osteosarcoma cells. Science 241, 81 – 84 (1988). 106. E. F. Eriksen, D. S. Colvard, N. J. Berg, M. L. Graham, K. G. Mann, T. C. Spelsberg, and B. L. Riggs, Evidence of estrogen receptors in normal human osteoblast-like cells. Science 241, 84 – 86 (1988). 107. F. S. Kaplan, M. D. Fallon, S. D. Boden, R. Schmidt, M. Senior, and J. G. Haddad, Estrogen receptors in bone in a patient with polyostotic fibrous dysplasia (McCune-Albright Syndrome). N. Engl. J. Med. 319, 421 – 425 (1988). 108. M. Ernst, C. Schmid, and E. R. Froesch, Enhanced osteoblast proliferation and collagen gene expression by estradiol. Proc. Natl. Acad. Sci. USA 85, 2307 – 2310 (1988). 109. M. C. Etienne, J. L. Fischel, G. Milano, P. Formento, J. L. Formento, M. Francoual, M. Frenay, and M. Namer, Steroid receptors in human osteoblast-like cells. Eur. J. Cancer 26, 807 – 810 (1990). 110. D. J. Benz, M. R. Haussler, and B. S. Komm, Estrogen binding and estrogenic responses in normal human osteoblast-like cells. J. Bone Miner. Res. 6, 531 – 541 (1991). 111. P. E. Keeting, R. E. Scott, D. S. Colvard, M. A. Anderson, M. J. Oursler, T. C. Spelsberg, and B. L. Riggs, Development and characterization of a rapidly proliferating, well-differentiated cell line derived from normal adult human osteoblast-like cells transfected with SV40 large T antigen. J. Bone Miner. Res. 7, 127 – 136 (1992).
331 112. A. Masuyama, Y. Ouchi, T. Sato, T. Hosoi, T. Nakamura, and H. Orimo, Characteristics of steroid receptors in cultured MC3T3-E1 osteoblastic cells and effect of steroid hormones on cell proliferation. Calcif. Tissue Int. 51, 376 – 381 (1992). 113. S. Migliaccio, V. L. Davis, M. K. Gibson, T. K. Gray, and K. S. Korach, Estrogens modulate the responsiveness of osteoblast-like cells (ROS 17/2.8) stably transfected with estrogen receptor. Endocrinology 130, 2617 – 2624 (1992). 114. V. L. Davis, J. F. Couse, T. K. Gray, and K. S. Korach, Correlation between low levels of estrogen receptors and estrogen responsiveness in two rat osteoblast-like cell lines. J. Bone Miner. Res. 9, 983 – 991 (1994). 115. P. V. Bodine, J. Green, H. A. Harris, R. A. Bhat, G. S. Stein, J. B. Lian, and B. S. Komm, functional properties of a conditionally phenotypic, estrogen-responsive, human osteoblast cell line. J. Cell. Biochem. 65, 368 – 387 (1997). 116. D. R. Ciocca and L. M. Vargas Roig, Estrogen receptors in human nontarget tissues: Biological and clinical implications. Endocr. Rev. 16, 35 – 62 (1995). 117. S. Nilsson, G. Kuiper, and J.-A. Gustafsson, ER-beta: A novel estrogen receptor offers the potential for new drug development. Trends Endrocrinol. Metab. 9, 387 – 395 (1998). 118. J. A. Hoyland, A. P. Mee, P. Baird, I. P. Braidman, E. B. Mawer, and A. J. Freemont, Demonstration of estrogen receptor mRNA in bone using in situ reverse-transcriptase polymerase chain reaction. Bone 20, 87 – 92 (1997). 119. V. Kusec, A. S. Virdi, R. Prince, and J. T. Triffitt, Localization of estrogen receptor-alpha in human and rabbit skeletal tissues. J. Clin. Endocrinol. Metab. 83, 2421 – 2428 (1998). 120. D. N. Petersen, G. T. Tkalcevic, P. H. Koza-Taylor, T. G. Turi, and T. A. Brown, Identification of estrogen receptor beta2, a functional variant of estrogen receptor beta expressed in normal rat tissues. Endocrinology 139, 1082 – 1092 (1998). 121. S. K. Lim, Y. J. Won, H. C. Lee, K. B. Huh, and Y. S. Park, A PCR analysis of ER alpha and ER beta mRNA abundance in rats and the effect of ovariectomy. J. Bone Miner. Res. 14, 1189 – 1196 (1999). 122. O. Vidal, L. G. Kindblom, and C. Ohlsson, Expression and localization of estrogen receptor-beta in murine and human bone. J. Bone Miner. Res. 14, 923 – 929 (1999). 123. S. H. Windahl, M. Norgard, G. Kuiper, J. A. Gustafsson, and G. Andersson, Cellular distribution of estrogen receptor beta in neonatal rat bone. Bone 26, 117 – 121 (2000). 124. I. Braidman, C. Baris, L. Wood, P. Selby, J. Adams, A. Freemont, and J. Hoyland, Preliminary evidence for impaired estrogen receptor-alpha protein expression in osteoblasts and osteocytes from men with idiopathic osteoporosis. Bone 26, 423 – 427 (2000). 125. Y. Onoe, C. Miyaura, H. Ohta, S. Nozawa, and T. Suda, Expression of estrogen receptor beta in rat bone. Endocrinology 138, 4509 – 4512 (1997). 126. J. Arts, G. G. Kuiper, J. M. Janssen, J. A. Gustafsson, C. W. Lowik, H. A. Pols, and J. P. van Leeuwen, Differential expression of estrogen receptors alpha and beta mRNA during differentiation of human osteoblast SV-HFO cells. Endocrinology 138, 5067 – 5070 (1997). 127. M. A. Ankrom, J. A. Patterson, P. Y. dAvis, U. K. Vetter, M. R. Blackman, P. D. Sponseller, M. Tayback, P. G. Robey, J. R. Shapiro, and N. S. Fedarko, Age-related changes in human oestrogen receptor alpha function and levels in osteoblasts. Biochem. J. 333, 787 – 794 (1998). 128. P. V. Bodine, R. A. Henderson, J. Green, M. Aronow, T. Owen, G. S. Stein, J. B. Lian, and B. S. Komm, Estrogen receptor-alpha is developmentally regulated during osteoblast differentiation and contributes to selective responsiveness of gene expression. Endocrinology 139, 2048 – 2057 (1998). 129. P. V. N. Bodine, H. A. Harris, and B. S. Komm, Suppression of ligand-dependent estrogen receptor activity by bone-resorbing
332
130.
131.
132. 133. 134. 135.
136.
137.
138.
139. 140.
141.
142.
143.
144.
145.
146.
147.
148.
KOMM AND BODINE cytokines in human osteoblasts. Endocrinology 140, 2439 – 2451 (1999). J. M. Hall and D. P. McDonnell, The estrogen receptor beta-isoform (ERbeta) of the human estrogen receptor modulates ERalpha transcriptional activity and is a key regulator of the cellular response to estrogens and antiestrogens. Endocrinology 140, 5566 – 5578 (1999). R. R. Recker, Emryology, anatomy, and microstructure of bone. In “Disorders of Bone and Mineral Metabolism” (F. L. Coe and M. J. Favus eds.), pp. 219 – 240. Raven Press, New York, 1992. E. M. Aarden, E. H. Burger, and P. J. Nijweide, Functions of osteocytes in bone. J. Cell. Biochem. 55, 287 – 299 (1994). B. S. Noble and J. Reeve, Osteocyte function, osteocyte death and bone fracture resistance. Mol. Cell. Endocrinol. 159, 7 – 13 (2000). L. E. Lanyon, Osteocytes, strain detection, bone modeling and remodeling. Calcif. Tissue Int. 53 (Suppl. 1), S102 – S107 (1993). I. P. Braidman, L. K. Davenport, D. H. Carter, P. L. Selby, E. B. Mawer, and A. J. Freemont, Preliminary in situ identification of estrogen target cells in bone. J. Bone Miner. Res. 10, 74 – 80 (1995). J. A. Hoyland, C. Baris, L. Wood, P. Baird, P. L. Selby, A. J. Freemont, and I. P. Braidman, Effect of ovarian steroid deficiency on oestrogen receptor alpha expression in bone. J. Pathol. 188, 294 – 303 (1999). T. Ohashi, S. Kusuhara, and K. Ishida, Immunoelectron microscopic demonstration of estrogen receptors in osteogenic cells of Japanese quail. Histochemistry 96, 41 – 44 (1991). S. C. Manolagas and R. L. Jilka, Bone marrow, cytokines, and bone remodeling: Emerging insights into the pathophysiology of osteoporosis. N. Engl. J. Med. 332, 305 – 311 (1995). J. E. Aubin, Bone stem cells. J. Cell. Biochem. Suppl. 30/31, 73 – 82 (1998). R. O. C. Oreffo, V. Kusec, S. Romberg, and J. T. Triffitt, Human bone marrow osteoprogenitors express estrogen receptor-alpha and bone morphogenetic proteins 2 and 4 mRNA during osteoblastic differentiation. J. Cell. Biochem. 75, 382 – 392 (1999). S. C. Dieudonne, T. Xu, J. Y. Chou, S. A. Kuznetsov, K. Satomura, M. Mankani, N. S. Fedarko, E. P. Smith, P. G. Robey, and M. F. Young, Immortalization and characterization of bone marrow stromal fibroblasts from a patient with a loss of function mutation in the estrogen receptor-alpha gene. J. Bone Miner. Res. 13, 598 – 608 (1998). M. J. Oursler, Estrogen regulation of gene expression in osteoblasts and osteoclasts. Crit. Rev. Eukary. Gene Express. 8, 125 – 140 (1998). T. Suda, N. Takahashi, N. Udagawa, E. Jimi, M. T. Gillespie, and T. J. Martin, Modulation of osteclast differentiation and function by new members of the tumor necrosis factor receptor and ligand families. Endocr. Rev. 20, 345 – 357 (1999). J. M. Pensler, C. B. Langman, J. A. Radosevich, M. L. Maminta, M. Mangkornkanok, R. Higbee, and A. Molteni, Sex steroid hormone receptors in normal and dysplastic bone disorders in children. Bone Miner. Res. 5, 493 – 498 (1990). M. J. Oursler, P. Osdoby, J. Pyfferoen, B. L. Riggs, and T. C. Spelsberg, Avian osteoclasts as estrogen target cells. Proc. Nat. Acad. Sci. USA 88, 6613 – 6617 (1991). M. J. Oursler, L. Pederson, L. Fitzpatrick, B. L. Riggs, and T. Spelsberg, Human giant cell tumors of the bone (osteoclastomas) are estrogen target cells. Proc. Nal. Acad. Sci. USA 91, 5227 – 5231 (1994). T. Sunyer, J. Lewis, P. Collin-Osdoby, and P. Osdoby, Estrogen’s bone-protective effects may involve differential IL-1 receptor regulation in human osteoclast-like cells. J. Clin. Invest. 103, 1409 – 1418 (1999). H. Mano, T. Yuasa, T. Kameda, K. Miyazawa, Y. Nakamaru, M. Shiokawa, Y. Mori, T. Yamada, K. Miyata, H. Shindo, H. Azuma, Y. Hakeda, and M. Kumegawa, Mammalian mature osteoclasts as
149.
150.
151.
152.
153.
154.
155.
156.
157.
158.
159.
160.
161.
162.
163.
164.
165.
estrogen target cells. Biochem. Biophys. Res. Commun. 223, 637 – 642 (1996). F. M. Collier, W. H. Huang, W. R. Holloway, J. M. Hodge, M. T. Gillespie, L. L. Daniels, M. H. Zheng, and G. C. Nicholson, Osteoclasts from human giant cell tumors of bone lack estrogen receptors. Endocrinology 139, 1258 – 1267 (1998). L. Pederson, M. Kremer, N. T. Foged, B. Winding, C. Ritchie, L. A. Fitzpatrick, and M. J. Oursler, Evidence of a correlation of estrogen receptor level and avian osteoclast estrogen responsiveness. J. Bone Miner. Res. 12, 742 – 752 (1997). G. Fiorelli, F. Gori, M. Petilli, A. Tanini, S. Benvenuti, M. Serio, P. Bernabel and M. L. Brandi, Functional estrogen receptors in a human preosteoclastic cell line. Proc. Natl. Acad. Sci. USA 92, 2672 – 2676 (1995). M. Kanatani, T. Sugimoto, Y. Takahashi, H. Kaji, R. Kitazawa, and K. Chihara, Estrogen via the estrogen receptor blocks cAMPmediated parathyroid hormone (PTH)-stimulated osteoclast formation. J. Bone Miner. Res. 13, 854 – 862 (1998). H. Ben-Hur, G. Mor, I. Blickstein, I. Likhman, F. Kohen, R. Dgani, V. Insler, P. Yaffe, and A. Ornoy, Localization of estrogen receptors in long bones and vertebrae of human fetuses. Calci. Tissue Int. 53, 91 – 96 (1993). H. Ben-Hur, H. H. Thole, A. Mashiah, V. Insler, V. Berman, E. Shezen, D. Elias, A. Zuckerman, and A. Ornoy, Estrogen, progesterone and testosterone receptors in human fetal cartilaginous tissue: Immunohistochemical studies. Calci. Tissue Int. 60, 520 – 526 (1997). J. Kennedy, C. Baris, J. A. Hoyland, P. L. Selby, A. J. Freemont, and I. P. Braidman, Immunofluorescent localization of estrogen receptoralpha in growth plates of rabbits, but not in rats, at sexual maturity. Bone 24, 9 – 16 (1999). L. O. Nilsson, A. Boman, L. Savendahl, G. Grigelioniene, C. Ohlsson, E. M. Ritzen, and J. Wroblewski, Demonstration of estrogen receptor-beta immunoreactivity in human growth plate cartilage. J. Clin. Endocrinol. Metab. 84, 370 – 373 (1999). N. Dayani, M. T. Corvol, P. Robel, B. Eychenne, B. Moncharmont, L. Tsagris, and R. Rappaprot, Estrogen receptors in cultured rabbit articular chondrocytes: Influence of age. J. Steroid Biochem. 31, 351 – 356 (1988). B. A. Monaghan, F. S. Kaplan, C. R. Lyttle, M. D. Fallon, S. D. Boden, and J. G. Haddad, Estrogen receptors in fracture healing. Clin. Orthop. Rel. Res. 280, 277 – 280 (1992). H. Pinus, A. Ornoy, N. Patlas, P. Yaffe, and Z. Schwartz, Specifi beta-estradiol binding in cartilage and serum from young mice and rats is age dependent. Connect. Tissue Res. 30, 85 – 98 (1993). E. Nasatzky, Z. Schwartz, W. A. Soskolne, B. P. Brooks, D. D. Dean, B. D. Boyan, and A. Ornoy, Evidence for receptors specific for 17beta-estradiol and testosterone in chondrocyte cultures. Connect. Tissue Res. 30, 277 – 294 (1994). M. L. Brandi, C. Crescioli, A. Tanini, U. Frediani, D. Agnusdei, and C. Gennari, Bone endothelial cells as estrogen targets. Calci. Tissue Int. 53, 312 – 317 (1993). G. S. Stein and J. B. Lian, Molecular mechanisms mediating proliferation/differentiation interrelationships during progressive development of the osteoblastic phenotype. Endocr. Rev. 14, 424 – 442 (1993). J. Clover and M. Gowen, Are MG-63 and HOS TE85 human osteosarcoma cell lines representative models of the osteoblastic phenotype? Bone 15, 585 – 591 (1994). D. Modrowski, L. Miravet, M. Feuga, and P. J. Marie, Increased proliferation of osteoblast precursor cells in estrogen-deficient rats. Am. J. Physiol. 264, E190 – E196 (1993). K. C. Westerlind, G. K. Wakley, G. L. Evans, and R. T. Turner, Estrogen does not increase bone formation in growing rats. Endocrinology 133, 2924 – 2934 (1993).
333
CHAPTER 10 Regulation of Bone Cell Function by Estrogens 166. C. K. Watts and R. J. King, Overexpression of estrogen receptor in HTB 96 human osteosarcoma cells results in estrogen-induced growth inhibition and receptor cross talk. J. Bone Miner. Res. 9, 1251 – 1258 (1994). 167. E. Damien, J. S. Price, and L. E. Lanyon, The estrogen receptor’s involvement in osteoblasts adaptive response to mechanical strain. J. Bone Miner Res. 13, 1275 – 1282 (1998). 168. J. A. Robinson, S. A. Harris, B. L. Riggs, and T. C. Spelsberg, Estrogen regulation of human osteoblastic cell proliferation and differentiation. Endocrinology 138, 2919 – 2927 (1997). 169. A. Samuels, M. J. Perry, and J. H. Tobias, High-dose estrogeninduced osteogenesis in the mouse is partially suppressed by indomethacin. Bone 25, 675 – 680 (1999). 170. Q. Qu, P. L. Harkonen, and H. K. Vaananen, Comparative effects of estrogen and antiestrogens on differentiation of osteoblasts in mouse bone marrow culture. J. Cell. Biochem. 73, 500 – 507 (1999). 171. D. Somjen, Y. Weisman, A. Harell, E. Berger, and A. M. Kaye, Direct and sex-specific stimulation by sex steroids of creatine kinase activity and DNA synthesis in rat bone. Proc. Natl. Acad. Sci. USA 86, 3361 – 3365 (1989). 172. S. D. Bain, M. C. Bailey, D. L. Celino, M. M. Lantry, and M. W. Edwards, High-dose estrogen inhibits bone resorption and stimulates bone formation in the ovariectomized mouse. J. Bone Miner. Res. 8, 435 – 442 (1993). 173. C. H. Turner, Editorial: Do estrogens increase bone formation? Bone 12, 305 – 306 (1991). 174. R. T. Turner, Mice, estrogen, and postmenopausal osteoporosis. Bone Miner. Res. 14, 187 – 191 (1999). 175. A. Gohel, M. B. McCarthy, and G. Gronowicz, Estrogen prevents glucocorticoid-induced apoptosis in osteoblasts in vivo and in vitro. Endocrinology 140, 5339 – 5347 (1999). 176. H. J. Verhaar, C. A. Damen, S. A. Duursma, and B. A. Scheven, A comparison of the action of progestins and estrogen on the growth and differentiation of normal adult human osteoblast-like cells in vitro. Bone 15, 307 – 311 (1994). 177. R. J. Majeska, J. T. Ryaby, and T. A. Einhorn, Direct modulation of osteoblastic activity with estrogen. J. Bone Joint Surg. 76, 713 – 721 (1994). 178. J. A. Robinson, K. M. Waters, R. T. Turner, and T. C. Spelsberg, Direct action of naturally occurring estrogen metabolites on human osteoblastic cells. J. Bone Miner. Res. 15, 499 – 506 (2000). 179. T. Ikeda, C. Shigeno, R. Kasai, H. Kohno, S. Ohta, H. Okumura, J. Konishi, and T. Yamamuro, Ovariectomy decreases the mRNA levels of transforming growth factor-beta1 and increases the mRNA levels of osteocalcin in rat bone in vivo. Biochem. Biophy. Commun. 194, 1228 – 1233 (1993). 180. R. T. Turner, D. S. Colvard, and T. C. Spelsberg, Estrogen inhibition of periosteal bone formation in rat long bones: Down-regulation of gene expression for bone matrix proteins. Endocrinology 127, 1346 – 1351 (1990). 181. M. Ernst, J. K. Heath, and G. A. Rodan, Estradiol effects on proliferation, messenger ribonucleic acid for collagen and insulin-like growth factor-I, and parathyroid hormone-stimulated adenylate cyclase activity in osteoblastic cells from calvariae and long bones. Endocrinology 125, 825 – 833 (1989). 182. M. A. Salih, C. C. Liu, B. H. Arjmandi, and D. N. Kalu, Estrogen modulates the mRNA levels for cancellous bone protein of ovariectomized rats. Bone Miner. 23, 285 – 299 (1993). 183. K. C. Westerlind, T. J. Wronski, G. L. Evans, and R. T. Turner, The effect of long-term ovarian hormone deficiency on transforming growth factor-beta and bone matrix protein mRNA expression in rat femora. Biochem. Biophys. Res. Commun 200, 283 – 289 (1994). 184. R. T. Turner, P. Backup, P. J. Sherman, E. Hill, G. L. Evans, and T. C. Spelsberg, Mechanism of action of estrogen on intramembra-
185.
186.
187.
188.
189.
190.
191.
192.
193.
194. 195. 196. 197.
198.
199.
200.
201. 202.
203.
nous bone formation: Regulation of osteoblast differentiation and activity. Endocrinology 131, 883 – 889 (1992). M. Takeuchi, M. Tokin, and K. Nagata, Tamoxifen directly stimulates the mineralization of human osteoblast-like osteosarcoma cells through a pathway independent of estrogen response element. Biochem. Biophys. Res. Commun. 210, 295 – 301 (1995). M. Centrella, M. C. Horowitz, J. M. Wozney, and T. L. McCarthy, Transforming growth factor-B gene family members and bone. Endocr. Rev. 15, 27 – 39 (1994). T. A. Linkhart, S. Mohan, and D. J. Baylink, Growth factors for bone growth and repair: IGF, TGF-beta and BMP. Bone 19, 1S – 12S (1996). R. N. Margolis, E. Canalis, and N. C. Partridge, Invited review of a workshop: anabolic hormones in bone: basic research and therapeutic potential. J. Clin. Endocrinol. Metab. 81, 872 – 877 (1996). J. M. Wozney and V. Rosen, Bone morphogenetic protein and bone morphogeneitic protein gene family in bone formation and repair. Clin. Orthop. Rel. Res. 346, 26 – 37 (1998). L. F. Bonewald, Regulation and regulatory activities of transforming growth factor beta. Crit. Rev. Eukary. Gene Express. 9, 33 – 44 (1999). C. J. Rosen, Serum insulin-like growth factors and insulin-like growth factor-binding proteins: Clinical implications. Clin. Chem. 45, 1384 – 1390 (1999). J. L. Jones and D. R. Clemmons, Insulin-like growth factors and their binding proteins: Biological actions. Endocr. Rev. 16, 3 – 34 (1995). P. P.-C. Hu, M. B. Datto, and X.-F. Wang, Molecular mechanisms of transforming growth factor-beta signaling. Endocr. Rev. 19, 349 – 363 (1998). M. C. Horowitz, Cytokines and estrogen in bone: Anti-osteoporotic effects. Science 260, 626 – 627 (1993). R. Pacifici, Estrogen, cytokines, and pathogenesis of postmenopausal osteoporosis. J. Bone Miner. Res. 11, 1043 – 1051 (1996). R. L. Jilka, Cytokines, bone remodeling, and estrogen deficiency: A 1998 update. Bone 23, 75 – 81 (1998). T. K. Gray, B. Lipes, T. Linkhart, S. Mohan, and D. J. Baylink, Transforming growth factor beta mediates the estrogen induced inhibition of UMR106 cell growth. Connect. Tissue Res. 20, 23 – 32 (1989). M. J. Oursler, C. Cortese, P. Keeting, M. A. Anderson, S. K. Bonde, B. L. Riggs, and T. C. Spelsberg, Modulation of transforming growth factor-beta production in normal human osteoblast-like cells by 17 beta-estradiol and parathyroid hormone. Endocrinology 129, 3313 – 3320 (1991). R. D. Finkelman, N. H. Bell, D. D. Strong, L. M. Demers, and D. J. Baylink, Ovariectomy selectively reduces the concentration of transfoming growth factor b in rat bone: Implications for estrogen deficiency-associated bone loss. Proc. Natl. Acad. Sci. USA 89, 12190 – 12193 (1992). N. N. Yang, H. U. Bryant, S. Hardikar, M. Sato, R. J. Galvin, A. L. Glasebrook, and J. D. Termine, Estrogen and raloxifene stimulate transforming growth factor-beta 3 gene expression in rat bone: A potential mechanism for estrogen- or raloxifene-mediated bone maintenance. Endocrinology 137, 2075 – 2084 (1996). D. P. McDonnell, The molecular pharmacology of SERMs. Trends Endocrinol. Metab. 10, 301 – 311 (1999). N. N. Yang, M. Venugopalan, S. Hardikar, and A. Glasebrook, Identification of an estrogen response element activated by metabolites of 17beta-estradiol and raloxifene. Science 273, 1222 – 1225 (1996). K. R. Tau, T. E. Hefferan, K. M. Waters, J. A. Robinson, M. Subramaniam, B. L. Riggs, and T. C. Spelsberg, Estrogen regulation of a transforming growth factor-beta inducible early gene that inhibits deoxyribonucleic acid synthesis in human osteoblasts. Endocrinology 139, 1346 – 1353 (1998).
334 204. M. Subramaniam, S. A. Harris, M. J. Oursler, K. Rasmussen, B. L. Riggs, and T. C. Spelsberg, Identification of a novel TGF-betaregulated gene encoding a putative zinc finger protein in human osteoblasts. Nucleic Acids Res. 23, 4907 – 4912 (1995). 205. D. J. Rickard, L. C. Hofbauer, S. K. Bonde, F. Gori, T. C. Spelsberg, and B. L. Riggs, Bone morphogenetic protein-6 production in human osteoblastic cell lines. Selective regulation by estrogen. J. Clin. Invest. 101, 413 – 422 (1998). 206. A. van den Wijngaard, W. R. Mulder, R. Dijkema, C. J. C. Boersma, S. Mosselman, E. J. J. van Zoelen, and W. Olijve, Antiestrogens specifically up-regulate bone morphogenic protein-4 promoter activity in human osteoblastic cells. Mol. Endocrinol. 14, 623 – 633 (2000). 207. T. K. Gray, S. Mohan, T. A. Linkhart, and D. J. Baylink, Estradiol stimulates in vitro the secretion of insulin-like growth factors by the clonal osteoblastic cell line, UMR106. Biochem. Biophys. Res. Commun. 158, 407 – 412 (1989). 208. M. C. Slootweg, D. Swolin, J. C. Netelenbos, O. G. Isaksson, and C. Ohlsson, Estrogen enhances growth hormone receptor expression and growth hormone action in rat osteosarcoma cells and human osteoblast-like cells. J. Endocrinol. 155, 159 – 164 (1997). 209. R. T. Turner, L. S. Kidder, M. Zhang, S. A. Harris, K. C. Westerlind, A. Maran, and T. J. Wronski, Estrogen has rapid tissue-specific effects on rat bone. J. Appl. Physiol. 86, 1950 – 1958 (1999). 210. J. Erdmann, S. Storch, J. Pfeilschifter, P. Ochlich, R. Ziegler, and F. Bauss, Effects of estrogen on the concentration of insulin-like growth factor-1 in rat bone. Bone 22, 503 – 507 (1998). 211. C. Schmid, M. Ernst, J. Zapf, and E. R. Froesch, Release of insulinlike growth factor carrier proteins by osteoblasts: Stimulation by estradiol and growth hormone. Biochem. Biophys. Res. Commun. 160, 788 – 794 (1989). 212. M. Kassem, R. Okazaki, D. De Leon, S. A. Harris, J. A. Robinson, T. C. Spelsberg, C. A. Conover, and B. L. Riggs, Potential mechanism of estrogen-mediated decrease in bone formation: Estrogen increases production of inhibitory insulin-like growth factor-binding protein-4. Proc. Assoc. Am. Phys. 108, 155 – 164 (1996). 213. Y. Kudo, M. Iwashita, T. Iguchi, Y. Takeda, N. Hizuka, K. Takano, and T. Muraki, Estrogen and parathyroid hormone regulate insulinlike growth factor binding protein-4 in SaOS-2 cells. Life Sci. 61, 165 – 170 (1997). 214. V. Hwa, Y. Oh, and R. G. Rosenfeld, The insulin-like growth factor binding protein (IGFBP) superfamily. Endocr. Rev. 20, 761 – 787 (1999). 215. C. J. Rosen, D. Verault, C. Steffens, D. Cheleuitte, and J. Glowacki, Effects of age and estrogen status on the skeletal IGF regulatory system. Endocrin. 7, 77 – 80 (1997). 216. T. L. McCarthy, C. Ji, H. Shu, S. Casinghino, K. Crothers, P. Rotwein, and M. Centrella, 17beta-estradiol potently suppresses cAMP-induced insulin-like growth factor-1 gene activation in primary rat osteoblast cultures. J. Biol. Chem. 272, 18132 – 18139 (1997). 217. G. Girasole, R. L. Jilka, G. Passeri, S. Boswell, G. Boder, D. C. Williams, and S. C. Manolagas, 17b-estradiol inhibits interleukin-6 production by bone marrow-derived stromal cells and osteoblastic cells in vitro: A potential mechanism for the antiosteoporotic effect of estrogens. J. Clin. Invest. 89, 883 – 891 (1992). 218. R. L. Jilka, G. Hangoc, G. Girasole, G. Passeri, D. C. Williams, J. S. Abrams, B. Boyce, H. Broxmeyer, and S. C. Manolagas, Increased osteoclast development after estrogen loss: Mediation by interleukin-6. Science 257, 88 – 91 (1992). 219. D. Cheleuitte, S. Mizuno, and J. Glowacki, In vitro secretion of cytokines by human bone marrow: Effects of age and estrogen status. J. Clin. Endocrinol. Metab. 83, 2043 – 2051 (1998). 220. S. T. Pottratz, T. Bellido, H. Mocharla, D. Crabb, and S. C. Manolagas, 17beta-estradiol inhibits expression of human interleukin-6 pro-
KOMM AND BODINE
221.
222.
223.
224.
225.
226. 227.
228.
229.
230.
231. 232.
233.
234.
235.
236.
237.
moter-reporter constructs by a receptor-dependent mechanism. J. Clin. Invest. 93, 944 – 950 (1994). L. I. McKay and J. A. Cidlowski, Molecular control of immune/ inflammatory responses: Interactions between nuclear factor-kappaB and steroid receptor-signaling pathways. Endocr. Rev. 20, 435 – 459 (1999). L. R. Chaudhary, T. C. Spelsberg, and B. L. Riggs, Production of various cytokines by normal human osteoblast-like cells in response to interleukin-1 beta and tumor necrosis factor-alpha: Lack of regulation by 17 beta-estradiol. Endocrinology 130, 2528 – 2534 (1992). D. Rickard, G. Russell, and M. Gowen, Oestradiol inhibits the release of tumor necrosis factor but not interleukin-6 from adult human osteoblasts in vitro. Osteoporosis Int. 2, 94 – 102 (1992). L. Rifas, J. S. Kenney, M. Marcelli, R. Pacifici, S.-L. Cheng, L. L. Dawson, and L. V. Avioli, Production of interleukin-6 in human osteoblasts and human bone marrow stromal cells: Evidence that induction by interleukin-1b and tumor necrosis factor-a is not regulated by ovarian steroids. Endocrinology 136, 4056 – 4067 (1995). M. Kassem, S. A. Harris, T. C. Spelsberg, and B. L. Riggs, Estrogen inhibits interleukin-6 production and gene expression in a human osteoblstic cell line with high levels of estrogen receptors. J. Bone Miner. Res. 11, 193 – 199 (1996). T. Taga and T. Kishimoto, GP130 and the interkeukin-6 family of cytokines. Annu. Rev. Immunol. 15, 797 – 819 (1997). S. C. Lin, T. Yamate, Y. Taguchi, V. Z. Borba, G. Girasole, C. A. OBrien, T. Bellido, E. Abe, and S. C. Manolagas, Regulation of the gp80 and gp130 subunits of the IL-6 receptor by sex steroids in the murine bone marrow. J. Clin. Invest. 100, 1980 – 1990 (1997). L. C. Hofbauer, S. Khosla, C. R. Dunstan, D. L. Lacey, W. J. Boyle, and B. L. Riggs, The roles of osteoprotegerin and osteoprotegerin ligand in the paracrine regulation of bone resorption. J. Bone Miner. Res. 15, 2 – 12 (2000). L. C. Hofbauer, S. Khosla, C. R. Dunstan, D. L. Lacey, T. C. Spelsberg, and B. L. Riggs, Estrogen stimulates gene expression and protein production of osteoprotegerin in human osteoblastic cells. Endocrinology 140, 4367 – 4370 (1999). S. Srivastava, M. N. Weitzmann, R. B. Kimble, M. Rizzo, M. Zahner, J. Milbrandt, F. P. Ross, and R. Pacifici, Estrogen blocks M-CSF gene expression and osteoclast formation by regulating phosphorylation of Egr-1 and its interaction with Sp-1. J. Clin. Invest. 102, 1850 – 1859 (1998). T. Chevalley and R. Rizzoli, Influence de I’hormone parathyroidienne sur Ios. Press Med. 28, 547 – 553 (1999). S. Fukayama and A. H. Tashjian, Direct modulation by estradiol of the response of human bone cells (SaOS-2) to human parathyroid hormone (PTH) and PTH-related protein. Endocrinology 124, 397 – 401 (1989). J. J. Monroe and A. H. J. Tashjian, Pretreatment with 17betaestradiol attenuates basal- and PTH-stimulated membrane adenylyl cyclase activity in human osteoblast-like SAOS-2 cells. Bichem. Biophys. Res. Commun. 225, 320 – 325 (1996). C. C. Pilbeam, J. Klein-Nulend, and L. G. Raisz, Inhibition by 17bestradiol of PTH-stimulated resorption and prostaglandin production in cultured neonatal mouse calvariae. Biochem. Biophys. Res. Commun. 163, 1319 – 1324 (1989). H. Kaji, T. Sugimoto, M. Kanatani, M. Nasu, and K. Chihara, Estrogen blocks parathyroid hormone (PTH)-stimulated osteoclast-like cell formation by selectively affecting PTH-responsive cyclic adenosine monophosphate pathway. Endocrinology 137, 2217 – 2224 (1996). F. Cosman, V. Shen, F. Xie, M. Seibel, A. Ratcliffe, and R. Lindsay, Estrogen protects against bone resorbing effects of parathyroid hormone infusion: Assesment by use of biochemical markers. Ann. Intern. Med. 118, 337 – 343 (1993). M. Ernst, M. G. Parker, and G. A. Rodan, Functional estrogen receptors in osteoblastic cells demonstrated by transfection with a reporter
CHAPTER 10 Regulation of Bone Cell Function by Estrogens
238.
239
240.
241.
242.
243.
244.
245.
246.
247.
248.
249.
250.
251.
252.
253.
254.
gene containing an estrogen response element. Mol. Endocrinol. 5, 1597 – 1606 (1991). L. G. Rao, J. N. Wylie, M. S. K. Sutherland, and T. M. Murray, 17bestradiol and parathyroid hormone potentiate each others stimulatory effects on alkaline phosphatase activity in SaOS-2 cells in a differeniation- dependent manner. Endocrinology 134, 614 – 620 (1994). C. Eielson, D. Kaplan, M. A. Mitnick, I. Paliwal, and K. Insogna, Estrogen modulates parathyroid hormone-induced fibronectin production in human and rat osteoblast-like cells. Endocrinology 135, 1639 – 1644 (1994). V. Behar, C. Nakamoto, Z. Greenberg, A. Bisello, L. J. Suva, M. Rosenblatt, and M. Chorev, Histidine at position 5 is the specificity “switch” between two parathyroid hormone receptor subtypes. Endocrinology 137, 4217 – 4224 (1996). L. A. Pivirotto, D. S. Cissel, and P. E. Keeting, Sex hormones mediate interleukin-1b production by human osteoblastic HOBIT cells. Mol. Cell Endocrinol. 111, 67 – 74 (1995). S. A. Harris, K. R. Tau, R. J. Enger, D. O. Toft, B. L. Riggs, and T. C. Spelsberg, Estrogen response in the hFOB 1.19 human fetal osteoblastic cell line stably transfected with the human estrogen receptor gene. J. Cell. Biochem. 59, 193 – 201 (1995). Y. Liel, S. Kraus, J. Levy, and S. Shany, Evidence that estrogens modulate activity and increase the number of 1,25-dihydroxyvitamin D receptors in osteoblast-like cells (ROS 17/2.8). Endocrinology 130, 2597 – 2601 (1992). M. Ishibe, T. Nojima, T. Ishibashi, T. Koda, K. Kaneda, R. N. Rosier, and J. E. Puzas, 17b-Estradiol increases the receptor number and modulates the action of 1,25-dihydroxyvitamin D3 in human osteosarcoma-derived osteoblast-like cells. Calcif. Tissue Int. 57, 430 – 435 (1995). K. L. Kirkwood, R. Dziak, and P. G. Bradford, Inositol trisphosphate receptor gene expression and hormonal regulation in osteoblast-like cell lines and primary osteoblast cell cultures. J. Bone Miner. Res. 11, 1889 – 1896 (1996). J. W. J. Putney and G. S. J. Bird, The inositol phosphate-calcium signaling system in nonexcitable cells. Endocr. Rev. Rev. 14, 610 – 631 (1993). K. L. Kirkwood, K. Homick, M. B. Dragon, and P. G. Bradford, Cloning and characterization of the type I inositol 1,4,5-trisphosphate receptor gene promoter: Regulation by 17beta-estradiol in osteoblasts. J. Biol. Chem. 272, 22425 – 22431 (1997). K. E. Armour and S. H. Ralston, Estrogen upregulates endothelial constitutive nitric oxide synthase expression in human osteoblastlike cells. Endocrinology 139, 799 – 802 (1998). S. H. Ralston, L.-P. Ho, M. H. Helfrich, P. S. Grabowski, P. W. Johnston, and N. Benjamin, Nitric oxide: a cytokine-induced regulator of bone resorption. J. Bone Miner. Res. 10, 1040 – 1049 (1995). S. J. Wimalawansa, G. De Marco, P. Gangula, and C. Yallampalli, Nitric oxide donor alleviates ovariectomy-induced bone loss. Bone 18, 301 – 304 (1996). R. L. Van Bezooijen, C. Van der Bent, S. E. Papapoulos, and C. W. G. M. Lowik, Oestrogenic compounds modulate cytokine-induced nitric oxide production in mouse osteoblast-like cells. J. Pharm. Pharmacol. 51, 1409 – 1414 (1999). D. S. Cissel, M. Murty, D. L. Whipkey, J. D. Blaha, G. M. Graeber, and P. E. Keeting, Estrogen pretreatment increases arachidonic acid release by bradykinin stimulated normal human osteoblast-like cells. J. Cell. Biochem. 60, 260 – 270 (1996). L. F. Cooper and K. Uoshima, Differential estrogenic regulation of small Mr heat shock protein expression in osteoblasts. J. Biol. Chem. 269, 7869 – 7873 (1994). D. D. Bankson, N. Rifai, M. E. Williams, L. M. Silverman, and T. K. Gray, Biochemical effects of 17 beta-estradiol on UMR106 cells. Bone Miner. 6, 55 – 63 (1989).
335 255. E. F. Eriksen, Normal and pathological remodeling of human trabecular bone: Three dimensional reconstruction of the remodeling sequence in normals and in metabolic bone disease. Endocr. Rev. 7, 379 – 408 (1984). 256. S. W. Whitson, Tight junction formation in the osteon. Clin. Orthop. Relat. Res. 86, 206 – 213 (1972). 257. A. Tomkinson, J. Reeve, R. W. Shaw, and B. S. Noble, The death of osteocytes via apoptosis accompanies estrogen withdrawal in human bone. J. Clin. Endocrinol. Metab. 82, 3128 – 3135 (1997). 258. R. L. Duncan and C. H. Turner, Mechanotransduction and the functional response of bone to mechanical strain. Calcif. Tissue Int. 57, 344 – 358 (1995). 259. A. Tomkinson, E. F. Gevers, J. M. Wit, J. Reeve, and B. S. Noble, The role of estrogen in the control of rat osteocyte apoptosis. J. Bone Miner. Res. 13, 1243 – 1250 (1998). 260. T. Keda, A. Yamaguchi, S. Yokose, Y. Nagal, H. Yamato, T. Nakamura, H. Tsurukami, T. Tanizawa, and S. Yoshiki, Changes in biological activity of bone cells in ovariectomized rats revealed by in situ hybrdization. J. Bone Miner. Res. 11, 780 – 788 (1996). 261. P. V. Bodine, S. K. Vernon, and B. S. Komm, Establishment and hormonal regulation of a conditionally transformed preosteocytic cell line from adult human bone. Endocrinology 137, 4592 – 4604 (1996). 262. M. Z. Cheng, G. Zaman, S. C. F. Rawlinson, R. F. L. Suswillo, and L. E. Lanyon, Mechanical loading and sex hormone interactions in organ cultures of rat ulna. J. Bone Miner. Res. 11, 502 – 511 (1996). 263. M. Z. Cheng, G. Zaman, S. C. F. Rawlinson, A. A. Pitsillides, R. F. L. Suswillo, and L. E. Lanyon, Enhancement by sex hormones of the osteoregulatory effects of mechanical loading and prostaglandins in explants of rat ulna. J. Bone Miner. Res. 12, 1424 – 1430 (1997). 264. K. C. Westerlind, T. J. Wronski, E. L. Ritman, Z.-P. Luo, K.-N. An, N. H. Bell, and R. T. Turner, Estrogen regulates the rate of bone turnover but bone balance in ovariectomized rats is modulated by prevailing mechanical strain. Proc. Natl. Acad. Sci. USA 94, 4199 – 4204 (1997). 265. W. M. Kohrt, D. B. Snead, E. Slatopolsky, and S. J. J. Birge, Additive effects of weight-bearing exercise and estrogen on bone mineral density in older women. J. Bone Miner. Res. 10, 1303 – 1311 (1995). 266. M. J. Oursler, L. Pederson, J. Pyfferoen, P. Osdoby, L. Fitzpatrick, and T. C. Spelsberg, Estrogen modulation of avian osteoclast lysosomal gene expression. Endocrinology 132, 1373 – 1380 (1993). 267. M. Kremer, J. Judd, B. Ritkin, J. Auszmann, and M. J. Oursler, Estrogen modulation of osteoclast lysosomal enzyme secretion. J. Cell. Biochem. 57, 271 – 279 (1995). 268. J. A. Robinson, B. L. Riggs, T. C. Spelsberg, and M. J. Oursler, Osteoclasts and transforming growth factor-beta: Estrogen-mediated isoform-specific regulation of production. Endocrinology 137, 615 – 621 (1996). 269. M. H. Zheng, T.-T. A. Lau, R. Prince, A. Criddle, S. Wysocki, M. Beilharz, J. M. Papadimitriou, and D. J. Wood, 17beta-estradiol suppresses gene expression of tartrate-resistant acid phosphatase and carbonic anhydrase II in ovariectomized rats. Calcif. Tissue Int. 56, 166 – 169 (1995). 270. J. P. Williams, H. C. Blair, M. A. McKenna, S. E. Jordan, and J. M. McDonald, Regulation of avian osteoclastic H -ATPase and bone resorption by tamoxifen and calmodulin antagonists. Effects independent of steroid receptors. J. Biol. Chem. 271, 12488 – 12495 (1996). 271. B. F. Boyce, D. E. Hughes, K. R. Wright, L. Xing, and A. Dai, Recent advances in bone biology provide insight into the pathogenesis of bone diseases. Lab. Invest. 79, 83 – 94 (1999). 272. S. C. Manolagas, Birth and death of bone cells: Basic regulatory mechanisms for the pathogenesis and treatment of osteoporosis. Endocr. Rev. 21, 115 – 137 (2000). 273. D. E. Hughes, A. Dai, J. C. Tiffee, H. H. Li, G. R. Mundy, and B. F. Boyce, Estrogen promotes apoptosis of murine osteoclasts mediated by TGF-beta. Nature Med. 2, 1132 – 1136 (1996).
336 274. S. Zecchi-Orlandini, L. Formigli, A. Tani, S. Benvenuti, G. Fiorelli, L. Papucci, S. Capaccioli, G. E. Orlandini, and M. L. Brandi, 17 beta-estradiol induces apoptosis in the preosteoclastic FLG 29.1 cell line. Biochem. Biophys. Res. Commun. 255, 680 – 685 (1999). 275. G. Fiorelli, V. Martineti, F. Gori, S. Benvenuti, U. Frediani, L. Formigli, S. Zecchi, and M. L. Brandi, Heterogeneity of binding sites and bioeffects of raloxifene on the human leukemic cell line FLG 29.1. Biochem. Biophy. Res. Commun. 240, 573 – 579 (1997). 276. Z. Khalkhali-Ellis, P. Collin-Osdoby, L. Li, M. L. Brandi, and P. Osdoby, A human homolog of the 150 kD avian osteoclast membrane antigen related to superoxide dismutase and essential for bone resorption is induced by developmental agents and opposed by estrogen in FLG 29.1 cells. Calcif. Tissue Int. 60, 187 – 193 (1997). 277. C. Schiller, R. Gruber, K. Redlich, G. M. Ho, F. Katzgraber, M. Willheim, P. Pietschmann, and M. Peterlik, 17Beta-estradiol antagonizes effects of 1 alpha,25-dihydroxyvitamin D3 on interleukin-6 production and osteoclast-like cell formation in mouse bone marrow primary cultures. Endocrinology 138, 4567 – 4571 (1997). 278. M. M. Takahashi and T. Noumura, Sexually dimorphic and laterally asymmetric development of the embryonic duck syrinx: Effect of estrogen on in vitro cell proliferation and chondrogenesis. Dev. Biol. (Orlando) 121, 417 – 422 (1987). 279. E. Nasatzky, Z. Schwartz, B. D. Boyan, W. A. Soskolne, and A. Ornoy, Sex-dependent effects of 17beta-estradiol on chondrocyte differentiation in culture. J. Cell. Physiol. 154, 359 – 367 (1993). 280. M.-T. Corvol, A. Carrascosa, L. Tsagris, O. Blanchard, and R. Rappaport, Evidence for a direct in vitro action of sex steroids on rabbit cartilage cells during skeletal growth: Influence of age and sex. Endocrinology 120, 1422 – 1429 (1987). 281. O. Blanchard, L. Tsagris, R. Rappaport, G. Duval-Beaupere, and M.-T. Corvol, Age-dependent responsivness of rabbit and human cartilage cells to sex steroids in vitro. J. Steroid Biochem. Mol. Biol. 40, 711 – 716 (1991). 282. E. Bonnelye, J. M. Vanacker, T. Dittmar, A. Begue, X. Desbiens, D. T. Denhardt, J. E. Aubin, V. Laudet, and B. Fournier, The ERR-1 orphan receptor is a transcriptional activator expressed during bone development. Mol. Endocrinol. 11, 905 – 916 (1997). 283. J. M. Vanacker, C. Delmarre, X. J. Guo, and V. Laudet, Activation of the osteopontin promoter by the orphan nuclear receptor estrogen receptor related alpha. Cell Growth Differ. 9, 1007 – 1014 (1998). 284. J. M. Vanacker, K. Pettersson, J. A. Gustafsson, and V. Laudet, Transcriptional targets shared by estrogen receptor- related receptors (ERRs) and estrogen receptor (ER) alpha, but not by ERbeta. EMBO J. 18, 4270 – 4279 (1999). 285. E. R. Levin, Cellular functions of the plasma membrane estrogen receptor. Trends Endocrinol. Metab. 10, 374 – 377 (1999). 286. M. J. Kelly and E. J. Wagner, Estrogen modulation of G-protein coupled receptors. Trends Endocrinol. Metab. 10, 369 – 374 (1999). 287. M. Lieberherr, B. Grosse, M. Kachkache, and S. Balsan, Cell signaling and estrogens in female rat osteoblasts: A possible involvement of unconventional nonnuclear receptors. J. Bone Miner. Res. 8, 1365 – 1376 (1993). 288. V. Le Mellay, B. Grosse, and M. Lieberherr, Phospholipase C beta and membrane action of calcitriol and estradiol. J. Biol. Chem. 272, 11902 – 11907 (1997). 289. V. Le Mellay, F. Lasmoles, and M. Lieberherr, Galpha q/11 and Gbeta gamma proteins and membrane signaling of calcitriol and estradiol. J. Cell. Biochem. 75, 138 – 146 (1999). 290. H. Endoh, H. Sasaki, K. Maruyama, K.-I. Takeyama, I. Waga, T. Shimizu, S. Kato, and H. Kawashima, Rapid activation of MAP kinase by estrogen in the bone cell line. Biochem. Biophys. Res. Commun. 235, 99 – 102 (1997). 291. C. V. Gay, N. L. Kief, and P. J. Bekker, Effect of estrogen on acidification in osteoclasts. Biochem. Biophys. Res. Commun. 192, 1251 – 1259 (1993).
KOMM AND BODINE 292. K. D. Brubaker and C. V. Gay, Specific binding of estrogen to osteoclast surfaces. Biochem. Biophys. Res. Commun. 200, 899 – 907 (1994). 293. K. D. Brubaker and C. V. Gay, Depolarization of osteoclast plasma membrane potential by 17 beta-estradiol. J. Bone Miner. Res. 14, 1861 – 1866 (1999). 294. A. Ikegami, S. Inoue, T. Hosoi, Y. Mizuno, T. Nakamura, Y. Ouchi, and H. Orimo, Immunohistochemical detection and northen blot analysis of estrogen receptors in osteoblastic cells. J. Bone Miner. Res. 8, 1103 – 1109 (1993). 295. K. Grandien, M. Backdahl, O. Ljunggren, J.-A. Gustafsson, and A. Berkenstam, Estrogen target tissue determines alternative promoter utilization of the human estrogen receptor gene in osteoblasts and tumor cell lines. Endocrinology 136, 2223 – 2229 (1995). 296. R. Delaveyne-Bitbol and M. Garabedian, In vitro responses to 17 beta-estradiol throughout pubertal maturation in female human bone cells. J. Bone Miner. Res. 14, 376 – 385 (1999). 297. M. E. Bolander, M. E. Joyce, S. E. Boden, B. Oliver, and A. Heydemann, Estrogen receptor mRNA expression during fracture healing in the rat detected by polymerase chain reaction amplification. In “Calcium Regulation and Bone Metabolism” (D. V. Cohn, F. H. Glorieux, and T. J. Martin eds.). Elsevier, New York, 1990. 298. R. Gruber, K. Czerwenka, F. Wolf, G. M. Ho, M. Willheim, and M. Peterlik, Expression of the vitamin D receptor, of estrogen and thyroid hormone receptor alpha- and beta-isoforms, and of the androgen receptor in cultures of native mouse bone marrow and of stromal/osteoblastic cells. Bone 24, 465 – 473 (1999). 299. A. Mahonen and P. H. Maenpaa, Steroid hormone modulation of vitamin D receptor levels in human MG-63 osteosarcoma cells. Biochem. Biophys. Res. Commun. 205, 1179 – 1186 (1994). 300. T. Bellido, G. Girasole, G. Passeri, X. P. Yu, H. Mocharla, R. L. Jilka, A. Notides, and S. C. Manolagas, Demonstration of estrogen and vitamin D receptors in bone marrow-derived stromal cells: Upregulation of the estrogen receptor by 1,25-dihydroxyvitamin-D3. Endocrinology 133, 553 – 562 (1993). 301. Q. Qu, M. Perala-Heape, A. Kapanen, J. Dahllund, J. Salo, H. K. Vaananen, and P. Harkonen, Estrogen enhances differentiation of osteoblasts in mouse bone marrow culture. Bone 22, 201 – 209 (1998). 302. W. H. Huang, A. T. Lau, L. L. Daniels, H. Fujii, U. Seydel, D. J. Wood, J. M. Papadimitriou, and M. H. Zheng, Detection of estrogen receptor alpha, carbonic anhydrase II and tartrate-resistant acid phosphatase mRNAs in putative mononuclear osteoclast precursor cells of neonatal rats by fluorescence in situ hybridization. J. Mol. Endocrinol. 20, 211 – 219 (1998). 303. R. O. C. Oreffo, V. Kusec, A. S. Virdi, A. M. Flanagan, M. Grano, A. Zambonin-Zallone, and J. T. Triffitt, Expression of estrogen receptor-alpha in cells of the osteoclastic lineage. Histochem. Cell. Biol. 111, 125 – 133 (1999). 304. C. K. Watts, M. G. Parker, and R. J. King, Stable transfection of the oestrogen receptor gene into a human osteosarcoma cell line. J. Steroid Biochem. 34, 483 – 490 (1989). 305. M. Z. Cheng, G. Zaman, S. C. F. Rawlinson, S. Mohan, D. J. Baylink, and L. E. Lanyon, Mechanical strain stimulates ROS cell proliferation through IGF-II and estrogen through IGF-I. J. Bone. Miner. Res. 14, 1742 – 1750 (1999). 306. A. Ikegami, S. Inoue, T. Hosoi, M. Kaneki, Y. Mizuno, Y. Akedo, Y. Ouchi, and H. Orimo, Cell cycle-dependent expression of estrogen receptor and effect of estrogen on proliferation of synchronized human osteoblast-like osteosarcoma cells. Endocrinology 782 – 789 (1994). 307. M. Ernst and G. A. Rodan, Estradiol regulation of insulin-like growth factor-I expression in osteoblastic cells: Evidence for transcriptional control. Mol. Endocrinol. 5, 1081 – 1089 (1991). 308. M. Kassem, R. Okazaki, S. A. Harris, T. C. Spelsberg, C. A. Conover, and B. L. Riggs, Estrogen effects on insulin-like
CHAPTER 10 Regulation of Bone Cell Function by Estrogens growth factor gene expression in a human osteoblastic cell line with high levels of estrogen receptor. Calcif. Tissue Int. 62, 60 – 66 (1998). 309. Q. Qu, P. L. Harkonen, J. Monkkonen, and H. K. Vaananen, Conditioned medium of estrogen-treated osteoblasts inhibits osteoclast maturation and function in vitro. Bone 25, 211 – 215 (1999). 310. B. Huo, D. A. Dossing, and M. T. Dimuzio, Generation and characterization of a human osteosarcoma cell line stably transfected with
337 the human estrogen receptor gene. J. Bone Miner. Res. 10, 769 – 781 (1995). 311. S. Koka, T. M. Petro, and R. A. Reinhardt, Estrogen inhibits interleukin-1beta-induced interleukin-6 production by human osteoblastlike cells. J. Interf. Cytokine Res. 18, 479 – 483 (1998). 312. G. Passeri, G. Girasole, R. L. Jilka, and S. C. Manolagas, Increased interleukin-6 production by murine bone marrow and bone cells after estrogen withdrawal. Endocrinology 133, 822 – 828 (1993).
CHAPTER 11
Skeletal Biology of Androgens KRISTINE M. WIREN AND ERIC S. ORWOLL Bone and Mineral Unit, Oregon Health Sciences University, and the Portland Veterans Affairs Medical Center, Portland, Oregon 97201
VI. The Role of Aromatization: Effects of Replacement Sex Steroids after Castration VII. Gender Specificity in the Actions of Sex Steroids VIII. The Animal Model of Androgen Resistance IX. Summary References
I. Introduction II. Cellular Biology of the Skeletal Effects of Androgens III. Androgen Actions in Bone IV. The Role of Androgen Metabolism V. Animal Studies of the Skeletal Effects of Androgen
I. INTRODUCTION
estrogen replacement alone [8 – 11]. At the same time, treatment with nonaromatizable androgen alone and in combination with estrogen also result in distinct changes in bone mineral density in females [12]. These reports illustrate the independent actions of androgens and estrogens on the skeleton. Thus, in both men and women it is probable that androgens and estrogens each have important, yet distinct, functions during bone development and in the subsequent maintenance of skeletal homeostasis. With the awakening awareness of the importance of the effects of androgen on skeletal homeostasis, and the potential to make use of this information for the treatment of bone disorders, much is to be learned.
The obvious impact of menopause on skeletal health has focused much of the research on the general action of gonadal steroids on the specific effects of estrogen. However, androgens, independently, have important beneficial effects on skeletal development and on the maintenance of bone mass in both men and women. Thus, androgens (1) influence growth plate maturation and closure, helping to determine longitudinal bone growth during development, (2) mediate dichotomous regulation of cancellous and cortical bone mass, leading to a sexually dimorphic skeleton, (3) modulate peak bone mass acquisition, and (4) inhibit bone loss [1 – 4]. In castrate animals, replacement with nonaromatizable androgens (e.g., dihydrotestosterone) yields beneficial effects that are clearly distinct from those observed with estrogen replacement [5, 6]. In intact females, blockade of the androgen receptor with the specific androgen receptor antagonist hydroxyflutamide results in osteopenia [7]. Data suggest that combination therapy with both estrogen and androgenic steroids is more effective than
OSTEOPOROSIS, SECOND EDITION VOLUME 1
II. CELLULAR BIOLOGY OF THE SKELETAL EFFECTS OF ANDROGENS The mechanisms by which androgens affect skeletal homeostasis have been the focus of intensified research [13]. Androgen receptors are present in a variety of cells found in bone [14], clearly identifying bone as a target
339
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
340
WIREN AND ORWOLL
tissue for androgen. The direct actions of androgen that influence the complex processes of proliferation, differentiation, mineralization, and gene expression in the osteoblast have also been documented. Androgen effects on bone may also be indirectly modulated and/or mediated by other autocrine and paracrine factors in the bone microenvironment.
A. Molecular Mechanisms of Androgen Action in Bone Cells: The Androgen Receptor Direct characterization of AR expression in a variety of tissues, including bone, was made possible by the cloning of the AR cDNA [15,16]. The AR is a member of the class I (so-called classical or steroid) nuclear receptor superfamily, as are estrogen, progesterone, mineralocorticoid, and glucocorticoid receptors [17]. These steroid receptors are ligandinducible transcription factors with a highly conserved modular design. In the absence of ligand, AR is found in the nucleus [18] in a large complex of molecular chaperonins, consisting of loosely bound heat shock and other accessory proteins. As lipids, androgens can diffuse freely through the plasma membrane into the nucleus to bind the AR. Once bound by ligand, the AR is activated and released from this protein complex, allowing the formation of homodimers (or potentially heterodimers) that activate a cascade of nuclear events [19]. Bound to DNA, the AR influences transcription and translation of a specific network of genes, leading to the cellular response to the steroid. A steroid hormone target tissue is frequently defined as one that possesses both functional levels of the steroid receptor and a measurable response in the presence of hormone. Bone tissue clearly meets this standard with respect to androgen. Colvard et al. [20] first reported the presence of AR mRNA and specific androgen-binding sites in normal human osteoblastic cells. The abundance of both androgen and estrogen receptor (ER) proteins was similar, suggesting that androgens and estrogens each play important roles in skeletal physiology (Fig. 1). Subsequent reports have confirmed AR mRNA expression and/or the presence of androgen-binding sites in both normal and clonal, transformed osteoblastic cells derived from a variety of species [18, 21 – 25]. The size of the AR mRNA transcript in osteoblasts (about 10 kb) is similar to that described in prostate and other tissues [15], as is the size of the AR protein analyzed by Western blotting ( 110 kDa) [24]. There is a report of two isoforms of AR protein in human osteoblast-like cells ( 110 and 97 kDa) [26] similar to that observed in human fibroblasts [27]. Whether these isoforms possess similar functional activities in bone when expressed at similar levels as described in other tissue [28] is yet to be determined. The number of specific androgen-binding sites in osteoblasts varies, depending on methodology and the cell source, from 1000 to 14000 sites/cell [23,24,29,30], but is
FIGURE 1 Nuclear AR and ER binding in normal human osteoblast-like cells. Dots represent the mean calculated number of molecules per cell nucleus for each cell strain. (Left) Specific nuclear binding of [3H]R1881 (methyltrienolone, an androgen analogue) in 12 strains from normal men and 13 strains from normal women. (Right) Specific nuclear [3H]estradiol binding in 15 strains from men and 15 strains from women. Horizontal lines indicate mean receptor concentrations [20]. in a range seen in other androgen target tissues. Furthermore, the binding affinity of the AR found in osteoblastic cells (Kd 0.5 – 2 109) is typical of that found in other tissues. Androgen binding is specific, without significant competition by estrogen, progesterone, or dexamethasone [20,24,30]. Finally, testosterone and dihydrotestosterone (DHT) appear to have similar binding affinities [21,24]. All these data are consistent with the notion that the direct biologic effects of androgenic steroids in osteoblasts are mediated at least in part via classic mechanisms associated with the AR as a member of the steroid hormone receptor superfamily as described earlier. In addition to the classical AR present in bone cells, several other androgen-dependent signaling pathways may be present. Specific binding sites for weaker adrenal androgens [dehydroepiandrosterone (DHEA)] have been described [31], raising the possibility that DHEA or similar androgenic compounds can also have direct effects in bone. In fact, Bodine et al. [32] showed that DHEA caused a rapid inhibition of c-fos expression in human osteoblastic cells that was more robust than seen with the classical androgens (DHT, testosterone, androstenedione). Nevertheless, all androgenic compounds significantly increased transforming growth factor- (TGF-) activity in osteoblastic cells. Androgens may also be specifically bound in osteoblastic cells by a 63-kDa cytosolic protein [33]. There are reports of distinct AR polymorphisms identified in different races that may have a biological impact on androgen responses [34], but this has not been explored with respect to bone tissue. These different isoforms have the potential to interact in distinct fashions with other signaling molecules, such as cJun [35]. Finally, androgens may regulate osteoblast activity
CHAPTER 11 Skeletal Biology of Androgens
341
via rapid nongenomic mechanisms [36] through receptors at the bone cell surface [37], as has also been shown for estrogen [38]. The role and biologic significance of these nonclassical signaling pathways in androgen-mediated responses in bone are still relatively unexplored.
osteoclasts in human bone slices [14]. Because the major effects of androgens on skeletal remodeling and maintenance of bone mineral density seem to be mediated by cells of osteoblast lineage [43], the biologic relevance of potential AR expression osteoclasts is unclear.
B. Localization of Androgen Receptor Expression
C. Regulation of Androgen Receptor Expression
Clues about the potential sequella of AR signaling might be derived from a better understanding of the cell types in which expression is documented. In the bone microenvironment, the localization of AR expression has been described in intact human bone by Abu et al. [14] using immunocytochemical methods. In developing bone from young adults, ARs were predominantly expressed in active osteoblasts at sites of bone formation (Fig. 2). ARs were also observed in osteocytes embedded in the bone matrix. Importantly, the pattern of both AR distribution and the level of expression was similar in males and in females. Furthermore, AR was also observed within the bone marrow in mononuclear cells and endothelial cells of blood vessels. Expression of the AR has also been characterized in cultured osteoblastic cell populations isolated from bone biopsy specimens, determined at both the mRNA level and by binding analysis [30]. Expression varied according to the skeletal site of origin and age of the donor of the cultured osteoblastic cells: AR expression was higher at cortical and intramembranous sites and lower in cancellous bone. This distribution pattern correlates with androgen responsiveness. AR expression was highest in osteoblastic cultures generated from young adults and somewhat lower in samples from either prepubertal or senescent bone. Again, no differences were found between male and female samples, suggesting that differences in receptor number per se do not underlie development of a sexually dimorphic skeleton. AR and ER have also been shown in bone marrow-derived stromal cells [39], which are responsive to sex steroids during the regulation of osteoclastogenesis. Because androgens are so important in bone development at the time of puberty, it is not surprising that ARs are also present in epiphyseal chondrocytes [14,40]. The expression of ARs in such a wide variety of cell types known to be important for bone modeling during development, and remodeling in the adult, provides evidence for direct actions of androgens in bone and cartilage tissue, and these results illustrate the complexity of androgen effects on bone. Osteoclasts may be a target for sex steroid regulation, as they have been shown to possess ERs [41], but a direct effect of androgens on osteoclast function has not been demonstrated. Mizuno et al. [42] described the presence of AR immunoreactivity in mouse osteoclast-like multinuclear cells, but expression was not detected in bona fide
The regulation of AR expression in osteoblasts is incompletely characterized. Homologous regulation of AR by androgen has been described that is tissue specific. Upregulation by androgen exposure is seen in a variety of osteoblastic cells [18,25,44,45], whereas in prostatic tissue, downregulation of AR after androgen exposure is observed. The androgen-mediated upregulation observed in osteoblasts, at least in part, occurs through changes in AR gene transcription (Fig. 3). As in other tissues, increased AR protein stability may also play a part. No effect, or even inhibition, of AR mRNA by androgen exposure in other osteoblastic models has also been described [30,46]. The mechanism(s) that underlies tissue specificity in autologous AR regulation, and the possible biological significance of distinct autologous regulation of AR, is not yet understood. It is possible that AR upregulation by androgen in bone may result in an enhancement of androgen responsiveness at times when androgen concentrations are rising or elevated. In addition, AR expression in osteoblasts may be upregulated by exposure to glucocorticoids, estrogen, or 1, 25-dihydroxyvitamin D3 [26]. Except for the immunocytochemical detection of AR expression in bone slices described previously, regulation during osteoblast differentiation has not been described. Whether other hormones, growth factors, or agents influence AR expression in bone is also unknown. Finally, whether the AR in osteoblasts undergoes posttranslational processing that might influence receptor signaling (stabilization, phosphorylation, and so on) as described in other tissues [47,48], and the potential functional implications [49], is also unknown. Ligandindependent activation of AR has been described in other tissues [50], but has not yet been explored in bone. AR activity may also be influenced by receptor modulators, such as nuclear receptor coactivators or corepressors [51,52]. These coactivators/corepressors can influence the downstream signaling of nuclear receptors to reflect both the cellular context and the particular promoter. AR-specific coactivators have been identified [53], many of which interact with the ligand-binding domain of the receptor [54]. Expression and regulation of these modulators may thus influence the ability of steroid receptors to regulate gene expression in bone [55], but this has been underexplored with respect to androgen action. A preliminary report suggested the presence of androgen-specific coactivators in osteoblastic cells [56].
342
WIREN AND ORWOLL
FIGURE 2
Localization of AR in normal tibial growth plate and adult osteophytic human bone. (a) Morphologically, sections of the growth plate consist of areas of endochondral ossification with undifferentiated (small arrowhead), proliferating (large arrowheads), mature (small arrow), and hypertrophic (large arrow) chondrocytes. Bar: 80 m. An inset of an area of the primary spongiosa is shown in b. (b) Numerous osteoblasts (small arrowheads) and multinucleated osteoclasts (large arrowheads) on the bone surface. Mononuclear cells within the bone marrow are also present (arrows). Bar: 60 m. (c) In the growth plate, AR is predominantly expressed by hypertrophic chondrocytes (large arrowheads). Minimal expression is observed in mature chondrocytes (small arrowheads). Receptors are rarely observed in the proliferating chondrocytes (arrow). (d) In the primary spongiosa, the AR is predominantly and highly expressed by osteoblasts at modeling sites (arrowheads). Bar: 20 m. (e) In osteophytes, AR is also observed at sites of endochondral ossification in undifferentiated (small arrowheads), proliferating (large arrowheads), mature (small arrows), and hypertrophic-like (large arrow) chondrocytes. Bar: 80 m. (f) A higher magnification of e showing proliferating, mature, and hypertrophic-like chondrocytes (large arrows, small arrows, and very large arrows, respectively). Bar: 40 m. (g) At sites of bone remodeling, the receptors are highly expressed in osteoblasts (small arrowheads) and also in mononuclear cells in the bone marrow (large arrowheads). Bar: 40 m. (h) AR is not detected in osteoclasts (small arrowheads). Bar: 40 m. B, bone: C, cartilage; BM, bone marrow [14].
III. ANDROGEN ACTIONS IN BONE A. Effects of Androgens on Proliferation and Apoptosis in Osteoblastic Cells Androgens have direct effects on osteoblast proliferation and expression in vitro, but the nature of these effects remains controversial; both stimulation and inhibition of
osteoblast proliferation have been reported. Benz and coworkers. showed that prolonged androgen exposure in the presence of serum inhibited proliferation (cell counts) by 15 – 25% in a transformed human osteoblastic line (TE-85), with testosterone and DHT being nearly equally effective. Hofbauer et al. [57] examined the effect of DHT in hFOB/AR-6, an immortalized human osteoblastic cell line stably transfected with an AR expression construct (with
343
CHAPTER 11 Skeletal Biology of Androgens
FIGURE 3 Dichotomous regulation of AR mRNA levels in osteoblast-like and prostatic carcinoma cell lines after exposure to androgen. (A) Time course of changes in AR mRNA abundance after DHT exposure in human SaOS-2 osteoblastic cells and human LNCaP prostatic carcinoma cells. To determine the effect of androgen exposure on hAR mRNA abundance, confluent cultures of either osteoblast-like cells (SaOS-2) or prostatic carcinoma cells (LNCaP) were treated with 10 8 M DHT for 0, 24, 48, or 72 h. Total RNA was then isolated and subjected to RNase protection analysis with 50 g total cellular RNA from SaOS-2 osteoblastic cells and 10 g total RNA from LNCaP cultures. (B) Densitometric analysis of AR mRNA steady-state levels. The AR mRNA to -actin ratio is expressed as the mean SEM compared to the control value from three to five independent assessments [44].
4000 receptors/cell). In this line, DHT treatment inhibited cell proliferation by 20 – 35%. Finally, Kasperk et al. [26] reported that prolonged DHT treatment inhibited normal human osteoblastic cell proliferation (cell counts) in cultures pretreated with DHT. In contrast, Kasperk et al. [58,59] also demonstrated in serum free primary cultures of murine and passaged human osteoblast-like cells that a variety of androgens increase DNA synthesis (assessed by [3H]thymidine incorporation) up to nearly 300% (Table 1) and cell counts by 200%. Again, testosterone and nonaromatizable androgens (DHT and fluoxymesterone) were nearly equally effective regulators. Consistent with increased proliferation, testosterone and DHT have also been reported to cause an increase in creatine kinase activity and [3H]thymidine incorporation into DNA in rat diaphyseal bone [60]. The differences observed with androgen-mediated changes in osteoblastic cell proliferation may be due to the different model systems employed (transformed osteoblastic cells vs second to fourth
passage normal human cells) and/or may reflect differences in the culture conditions (e.g., state of differentiation, receptor number, times of treatment, phenol red-containing vs phenol red-free, or serum containing vs serumfree). These differences also suggest an underlying complexity and subtlety for androgen regulation of osteoblast proliferation. TABLE 1
Effect of Androgens (1 nM) on [3H]Thymidine Uptake into DNA of Mouse Bone Cellsa Counts per minute
% control
Control
235 14
100 6
DHT
479 23
204 9
0.001
Testosterone
429 47
182 20
0.01
Fluoxymesterone
423 44
180 18
0.01
Methenolone
633 55
269 23
0.001
a
Data from Kasperk et al. [58].
P
344
WIREN AND ORWOLL
As a component of the control of osteoblast survival, it is also important to consider programmed cell death, or apoptosis [61]. A variety of skeletal cell types have been shown to undergo apoptosis [62,63]. In particular, as the osteoblast population differentiates in vitro, the mature bone cell phenotype undergoes apoptosis [64]. Modulation of bone cell apoptosis by steroid hormones has been shown: glucocorticoids enhance and estrogen treatment prevents apoptosis of osteoclasts [65] and osteoblasts/osteocytes [66 – 68]. Furthermore, evidence shows that the osteocytic population is particularly sensitive to the effects of estrogen withdrawal, which induces apoptosis [69,70]. Androgen exposure has been shown to influence apoptosis in other tissues [71,72], but the effects of either androgen exposure or androgen withdrawal in bone have not been described.
variety of model systems that androgens either inhibit [57] or have no effect on alkaline phosphatase activity [25,74], which may reflect both the complexity and the dynamics of osteoblastic differentiation. There are also reports of androgen-mediated increases in type I -1 collagen protein and mRNA levels [21,73,74] and increased osteocalcin secretion [26]. Consistent with increased collagen production, androgen treatment has also been shown to stimulate mineral accumulation in a time-and dose-dependent manner [25,26,75]. These results suggest that under certain conditions androgens enhance osteoblast differentiation and may thus play an important role in the regulation of bone matrix production and/or organization. This effect is also consistent with an overall stimulation of bone formation, as is observed clinically after androgen treatment.
B. Effects of Androgens on Differentiation of Osteoblastic Cells
C. Interaction with Other Factors to Modulate Bone Formation and Resorption
Osteoblast differentiation can be characterized by changes in alkaline phosphatase activity and/or alterations in the expression of important extracellular matrix proteins, such as type I collagen, osteocalcin, and osteonectin. Enhanced osteoblast differentiation, as measured by increased matrix production, has been shown to result from androgen exposure. Androgen treatment in both normal osteoblasts and transformed clonal human osteoblastic cells (TE-89) appears to increase the proportion of cells expressing alkaline phosphatase activity, thus representing a shift toward a more differentiated phenotype (Fig. 4) [58]. Kasperk and colleogues. reported dose-dependent increases in alkaline phosphatase activity in both high and low alkaline phosphatase subclones of SaOS2 cells [73] and human osteoblastic cells [26]. However, there are also reports in a
The effects of androgens on osteoblast activity must certainly also be considered in the context of the very complex endocrine, paracrine, and autocrine milieu in the bone microenvironment. Systemic and/or local factors presumably act in concert to influence bone cell function. This has been well described with regard to modulation of the effects of estrogen on bone [see, for example, 76 – 78]. Androgens have also been shown to regulate well-known modulators of osteoblast proliferation or function. The most extensively characterized growth factor influenced by androgen exposure is TGF-. TGF- is stored in bone (the largest reservoir for TGF-) in a latent form and has been show to be a mitogen for osteoblasts [79,80]. Androgen treatment increases TGF- activity in human osteoblast primary cultures (Fig. 5). The expression of some TGF- mRNA transcripts (apparently TGF-2) was increased, but no effect on TGF-1 mRNA abundance was observed [32,59]. At the protein level, specific immunoprecipitation analysis reveals DHTmediated increases in TGF- activity to be predominantly in the form of TGF-2 [26,32]. DHT has also been shown to inhibit both TGF- and TGF--induced early gene (TIEG) expression that correlates with growth inhibition in this cell line [57]. TIEG has been shown to be a transcription factor that may mediate some TGF- effects [81]. These results are consistent with the notion that TGF- may mediate androgen effects on osteoblast proliferation. However, TGF1 mRNA levels are increased by androgen treatment in human clonal osteoblastic cells (TE-89), under conditions where osteoblast proliferation is slowed [21]. Thus, the specific TGF- isoform may determine osteoblast responses. It is interesting to note that at the level of bone, orchiectomy drastically reduces bone content of TGF- levels, and testosterone replacement prevents its occurrence
FIGURE 4
Effect of DHT on ALP-positive and ALP-negative cells in normal mouse, human osteosarcoma (TE89) monolayer cell culture, and normal human osteoblast line. (*** p0.001; ** p0.01; * p0.1). Control values in cells per mm2 for mouse bone cells were: 90 5; TE89 cells: 75 7; and human bone cells: 83 14 [58].
CHAPTER 11 Skeletal Biology of Androgens
Induction of total TGF- activity by gonadal and adrenal androgens in human osteoblast (hOB) cell-conditioned media. Cells were treated for 24 – 48 h with vehicle or steroids. After treatment, conditioned media were saved and processed for the TGF- bioassay. Results are presented as the mean SEM of three to four experiments; *P 0.05; **P 0.02, ***P 0.0005 (Behren’s – Fisher t test) compared to the 48-h ethanol control. ETOH, ethanol; TEST, testosterone; DHT, DHT; ASD, androstenedione; DHEA, dehydroepiandrosterone; DHEA-S, DHEA-sulfate [32].
FIGURE 5
345 Other growth factor systems may also be influenced by androgens. Conditioned media from DHT-treated normal osteoblast cultures are mitogenic, and DHT pretreatment increases the mitogenic response to fibroblast growth factor and to insulin-like growth factor II (IGF-II) [59]. In part, this may be due to slight increases in IGF-II binding in DHT-treated cells [59], as the IGF-I and IGF-II content of osteoblast-conditioned media is not affected by androgen [59,83]. Although most studies have not found regulation of IGF-I or IGF-II abundance by androgen exposure [24,59,83], it has been shown that IGF-I mRNA levels are significantly upregulated by DHT [84]. Androgens also modulate the expression of components of the AP-1 transcription factor, as has been shown with the inhibition of cfos expression in proliferating normal osteoblast cultures [32]. Thus, androgens may accelerate osteoblast differentiation via a mechanism whereby growth factors or other mediators of differentiation are regulated by androgen exposure. Finally, androgens modulate responses to other important osteotropic hormones/regulators. Testosterone and DHT specifically inhibit the cAMP response elicited by parathyroid hormone or parathyroid hormone-related protein in the human clonal osteoblast-like cell line SaOS-2, whereas the inactive or weakly active androgen 17 epitestosterone had no effect (Fig. 7). This inhibition may
[82] (Fig. 6). These data support the findings that androgens influence cellular expression of TGF- and suggest that the bone loss associated with castration is related to a reduction in growth factor abundance induced by androgen deficiency.
FIGURE 6
Effects of orchiectomy and T replacement on isoforms of TGF- in long bones. Results are mean SE of four to six animals. Rats underwent sham operation or orchiectomy (ORX), and 1 week later were given either placebo or 100 mg of testosterone in 60-day slow-release pellets. Specimens were obtained 6 weeks after surgery. All forms of TGF- were reduced by orchiectomy (at least p0.0002), whereas there was no change in those with testosterone replacement [82].
FIGURE 7 Actions of testosterone and 17 -epitestosterone on cAMP accumulation stimulated by hPTH1-34 (5.0 nM) or hPTHrP1-34 (5.0 nM) in human SaOS-2 cells. Cells were pretreated without or with the steroid hormones (10 9M) for 24 h. Each bar gives the mean value, and brackets give the SE for four to five dishes [85].
346 be mediated via an effect on the parathyroid hormone receptor-Gs-adenylyl cyclase [85 – 87]. The production of prostaglandin E2 (PGE2), another important regulator of bone metabolism, is also affected by androgens. Pilbeam and Raisz [88] showed that both DHT and testosterone potently inhibited both parathyroid hormone (Fig. 8) and interleukin-1-stimulated PGE2 production in cultured neonatal mouse calvaria. The effects of androgens on parathyroid hormone action and PGE2 production suggest that androgens could act to modulate (reduce) bone turnover in response to these agents. Finally, both androgen [13] and estrogen [77, 89] inhibit the production of interleukin-6 by osteoblastic cells. In stromal cells of the bone marrow, androgens have been shown to have potent inhibitory effects on the production of interleukin-6 (Table 2) and the subsequent stimulation of osteoclastogenesis by marrow osteoclast precursors [39]. Interestingly, adrenal androgens (androstenediol, androstenedione, dehydroepiandrosterone) have similar inhibitory activities on interleukin-6 gene expression and protein production by stroma [39]. The loss of inhibition of interleukin-6 production by androgen may contribute to the marked increase in bone remodeling and resorption that follows orchidectomy. Moreover, androgens inhibit the expression of the genes encoding the two subunits of the IL-6 receptor (gp80 and gp130) in murine bone marrow, another mechanism that may blunt the effects of this osteoclastogenic cytokine in intact animals [90]. In these aspects, the effects of androgens seem to be very similar to those of estrogen, which may also inhibit osteoclastogenesis via mechanisms that involve interleukin-6 inhibition.
WIREN AND ORWOLL
TABLE 2 Effect of Androgens on CytokineInduced IL-6 Production by Murine Bone Marrow Stromal Cells a Treatment
IL-6
IL-1 TNF
4.27 1.43
IL-1 TNF testosterone (10 12 M)
3.87 0.33
IL-1 TNF testosterone (10 11 M)
2.90 0.42*
IL-1 TNF testosterone (10 10 M)
2.09 0.33*
IL-1 TNF testosterone (10 9 M)
1.12 0.49*
IL-1 TNF testosterone (10 8 M)
1.03 0.04*
7
IL-1 TNF testosterone (10
M)
1.01 0.48*
IL-1 TNF DHT (10 12 M)
4.05 0.19
IL-1 TNF DHT (10 11 M)
2.97 0.48*
IL-1 TNF DHT (10 10 M)
2.31 0.86*
9
IL-1 TNF DHT (10
M)
1.72 0.43*
IL-1 TNF DHT (10 8 M)
0.65 0.21*
IL-1 TNF DHT (10 7 M)
1.41 0.82*
a
Murine stromal cells ( / LDA11 cells) were cultured for 20 h in the absence or the presence of different concentrations of either testosterone or DHT. Then IL-1 (500 U/ml) and TNF (500 U/ml) were added and cells were maintained for another 24 h in culture. Values indicate means ( SD) of triplicate cultures from one experiment. Data were analyzed by one-way ANOVA. IL-1, interleukin-1; TNF, tumor necrosis factor. *P 0.05, versus cells not treated with steroids as determined by Dunnet’s test. Neither testosterone nor DHT had an affect on cell number [39].
IV. THE ROLE OF ANDROGEN METABOLISM A. Metabolism of Androgens in Bone: Aromatase and 5 -Reductase Activities
FIGURE 8 Effect of testosterone on PTH-stimulated PGE2 production in cultured neonatal calvariae as a function of time. Each bone was precultured for 24 h in 1 ml medium with or without 10 9 M T and then transferred to similar medium with 2.5 nM PTH. Media were sampled (0.1 ml) at the indicated times. Data were corrected for the volume of medium removed. Each point represents the mean SEM for six bones in one experiment. The effect of T on PTH-stimulated PGE2 production was significant (P 0.05) at 6, 12, and 24 h [88].
There is abundant evidence in a variety of tissues that the ultimate cellular effects of androgens may be the result not only of direct action of androgen, but also of the effects of sex steroid metabolites formed as the result of local enzyme activities. The most important of these androgen metabolites are estradiol (formed by the aromatization of testosterone) and 5 -DHT (the result of 5 reduction of testosterone). Evidence shows that both aromatase and 5 -reductase are present in bone tissue, at least to some measurable extent, but their biologic relevance is still controversial. 5 -reductase activity was first described in crushed rat mandibular bone by Vittek et al. [91]; and Schweikert et al. [92] reported similar findings in crushed human spongiosa. Two different 5 -reductase genes encode type 1 and type 2 isozymes in many mammalian species [93], but the
CHAPTER 11 Skeletal Biology of Androgens
isozyme present in human bone has not been characterized. In osteoblast-like cultures derived from orthopedic surgical waste, androstenedione (the major circulating androgen in women) can be reversibly converted to testosterone via 17-hydroxysteroid dehydrogenase activity and to 5 -androstanedione via 5 -reductase activity, whereas testosterone is converted to DHT via 5 -reductase activity [94]. The principal metabolites of androstenedione are 5 -androstanedione in the 5 -reductase pathway and testosterone in the 17-hydroxysteroid dehydrogenase pathway. Essentially the same results were reported in experiments with human epiphyseal cartilage and chondrocytes [95]. In general, the Km values for bone 5 -reductase activity are similar to those in other androgen responsive tissues [24,92]. The cellular populations in these studies were mixed and hence the specific cell type responsible for the activity is unknown. Turner and co-workers found that periosteal cells do not have detectable 5 -reductase activity [96], raising the possibility that the enzyme may be functional in only selected skeletal compartments, and that testosterone may be the active metabolite at this clinically important site. From a clinical perspective, the general importance of this enzymatic pathway is suggested by the presence of skeletal abnormalities in patients with 5 -reductase type 2 deficiency [97]. However, Bruch et al. [94] found no significant correlation between enzyme activities and bone volume. In mutant null mice lacking 5 -reductase type 1 (mice express very little type 2 isozyme), the effect on the skeleton cannot be analyzed due to midgestational fetal death [98]. Treatment of male animals with finasteride (an inhibitor of type 1 5 -reductase activity) does not recapitulate the effects of castration [99], indicating that reduction of testosterone to DHT by the type 1 isozyme is not a major determinant in the effects of gonadal hormones on bone. While available data point to a possible role for 5 reduction in the mechanism of action for androgen in bone, the clinical impact of this enzyme, which isozyme may be involved, whether it is uniformly present in all cells involved in bone modeling/remodeling, or whether local activity is important at all remains uncertain. The biosynthesis of estrogens from androgen precursors is catalyzed by the microsomal enzyme aromatase cytochrome P450 (P450arom, the product of the CYP19 gene). It is an enzyme well known to be both expressed and regulated in a very pronounced tissue-specific manner [100]. Aromatase activity has been reported in bone from mixed cell populations derived from both sexes [101 – 103] and from osteoblastic cell lines [24,104,105]. Aromatase expression in intact bone has also been documented by in situ hybridization and immunohistochemical analysis [103]. Aromatase mRNA is expressed predominantly in lining cells, chondrocytes, and some adipocytes, but there is no detectable expression in osteoclasts. At least in vertebral bone, the aromatase fibroblast (1b type) promoter is
347 predominantly utilized [103]. The enzyme kinetics in bone cells seem to be similar to those in other tissues, although the Vmax may be increased by glucocorticoids [105]. Aromatase can produce the potent estrogen estradiol, but can also form the weaker estrogen estrone from its adrenal precursors androstenedione and dehydroepiandrosterone [101]. In addition to aromatase itself, osteoblasts contain enzymes that are able to interconvert estradiol and estrone (estradiol-17B hydroxysteroid dehydrogenase) and to hydrolyze estrone sulfate to estrone (estrone sulfatase) [104]. Nawata et al. [101] have reported that dexamethasone and 1,25-dihydroxyvitamin D3 synergistically enhance aromatase activity and aromatase mRNA (P450arom) expression in human osteoblast-like cells. There is no other information concerning the regulation of aromatase in bone, although this is an area of obvious interest given the potential importance of the enzyme and its regulation by a variety of mechanisms (including androgens and estrogens) in other tissues [100,106]. The clinical impact of aromatase activity has been suggested by the reports of women [107] and men [108,109] with aromatase deficiencies who presented with a skeletal phenotype. The presentation of men with aromatase deficiency is very similar to that of a man with ER- deficiency, namely an obvious delay in bone age, lack of epiphyseal closure, and tall stature [110], suggesting that aromatase (and thus estrogen action) has a substantial role to play during skeletal development in the male (see Chapter 10 on estrogen action). In one case, estrogen therapy for a man with an aromatase deficiency was associated with an increase in bone mass [111]. Inhibition of aromatization in young growing orchidectomized male rats, with a nonsteroidal inhibitor, vorozole, results in decreases in bone mineral density and changes in skeletal modeling, as does castration with a resulting reduction in both androgens and estrogens. However, vorozole therapy induces less dramatic effects on bone turnover [112]. Inhibition of aromatization in older orchidectomized males resembles castration with similar increases in bone resorption and bone loss, suggesting that aromatase activity may also play a role in skeletal maintenance in males [113]. These studies herald the importance of aromatase activity (and estrogen) in the mediation of androgen action in bone. The finding of these enzymes in bone clearly raises the difficult issue of the origin of androgenic effects. Do they arise from direct androgen effects (as is suggested by the actions of nonaromatizable androgens) or to some extent from the local production of estrogenic intermediates? Nevertheless, there is substantial evidence that some, if in fact not most, of the biologic actions of androgens in the skeleton are direct. As noted previously, both in vivo and in vitro systems reveal the effects of the nonaromatizable androgen DHT to be essentially the same as those of testosterone (vida infra). In addition, blockade of the AR with the
348 receptor antagonist flutamide results in osteopenia as a result of reduced bone formation [7]. These reports clearly indicate that androgens, independent of estrogenic metabolites, have primary effects on osteoblast function. However, the clinical reports of subjects with aromatase deficiency also highlight the relevance of androgen metabolism to biopotent estrogens in bone. Elucidation of the regulation of steroid metabolism, and the potential mechanisms by which androgenic and estrogenic effects are coordinated, may have physiological, pathophysiological, and therapeutic implications.
B. Direct Effects of Androgens on Other Cell Types in Bone in Vitro Similar to the effects noted in osteoblastic populations, androgens regulate chondrocyte proliferation and expression. Androgen exposure promotes chondrogenesis as shown with increased creatine kinase and DNA synthesis after androgen exposure in cultured epiphyseal chondrocytes [40,114]. Increased [35S]sulfate incorporation into newly synthesized proteoglycan [115] and increased alkaline phosphatase activity [116] are androgen mediated. Regulation of these effects are obviously complex, as they were dependent on the age of the animals and the site from which chondrocytes were derived. Thus, in addition to effects on osteoblasts, multiple cell types in the skeletal milieu are regulated by androgen exposure.
V. ANIMAL STUDIES OF THE SKELETAL EFFECTS OF ANDROGEN The effects of androgens on bone remodeling have been examined fairly extensively in animal models. Much of this work has been in species not perfectly suited to reflect human bone metabolism (rodents), and certainly the field remains incompletely explored. Nevertheless, animal models do provide valuable insights into the effects of androgens at organ and cellular levels. Most studies of androgen action have been performed in male rats, in which rapid skeletal growth occurs until about 4 months of age, at which time epiphyseal growth slows markedly (although never completely ceases at some sites). Because the effects of androgen deficiency may be different in growing and more mature animals [1], it is appropriate to consider the two situations independently.
A. Effects on Epiphyseal Function and Bone Growth during Skeletal Development In most mammals there is a marked gender difference in bone morphology. The mechanisms responsible for these differences are complex and presumably involve both androgenic and estrogenic actions. Estrogens are particu-
WIREN AND ORWOLL
larly important for the regulation of epiphyseal function and act to reduce the rate of longitudinal growth via influences on chondrocyte proliferation and action, as well as on the timing of epiphyseal closure [117]. Androgens appear to have somewhat opposite effects and tend to promote long bone growth, chondrocyte maturation, and metaphyseal ossification. Androgen deficiency retards those processes [118]. Nevertheless, excess concentrations of androgen will accelerate aging of the growth plate and reduce growth potential [119], possibly via conversion to estrogens. Although the specific roles of sex steroids in the regulation of epiphyseal growth and maturation remain somewhat unresolved, evidence shows that androgens do have direct effects independent of those of estrogen. For instance, testosterone injected directly into the growth plates of rats increases plate width [120]. In a model of endochondral bone development based on the subcutaneous implantation of demineralized bone matrix in castrate rats, both testosterone and DHT increase the incorporation of calcium during osteoid formation [75]. Interestingly, in this model androgens reduced the incorporation of [35S]sulfate into glycosaminoglycans early in the developing cartilage. Certainly, androgens are known to interact considerably with the growth hormone – IGF system in the coordination of skeletal growth. Whereas androgens can induce a clear stimulation of growth in intact animals, in growth hormone deficiency that effect is essentially eliminated [121], underscoring the codependence of these two hormonal systems in the control of pubertal skeletal change. In sum, these data support the contention that androgens play a direct role in chondrocyte physiology, but how these actions are integrated with those of other regulators is unclear.
B. Effects on Bone Mass in Growing Male Animals The most dramatic effect of androgens during growth is on bone size. Male animals have larger bones, and particularly thicker cortices than females [117,122]. The effects of androgens on bone mass maturation can to some extent be assessed by observing the results of androgen withdrawal. In most studies, orchiectomy in young rats results in deficits in cortical bone mass within 2 – 4 weeks. The calcium content of the femur or tibia [123 – 127], whole femoral, tibial or body bone mineral density [99,127,128], and tibial diaphyseal cortical area [6] have been shown to be lower in castrated than in sham-operated controls. Similar trends have been reported in young, castrate male mice [129]. In animals followed for longer periods after castration (90 days), the density of cortical bone was reduced slightly (but not significantly), but bone area was clearly lessened in the diaphysis of the femur [126]. At least in part, the reduction in cortical bone mass appears to result
349
CHAPTER 11 Skeletal Biology of Androgens
from a reduction in the periosteal bone formation rate induced by gonadectomy in males [5,6]. This response differs from that induced by oophorectomy, which results in an increase in periosteal apposition in the period immediately after surgery (Fig. 9). This divergent trend in the periosteal response to castration in male and female animals abolishes the sexual dimorphism usually present in radial bone growth. Endosteal bone formation does not seem to be affected by orchiectomy [5]. As another indication that the cortical skeleton is affected by androgens, the characteristic acute increase in creatine kinase activity induced from diaphyseal bone by androgen treatment is abolished by orchiectomy [130]. For unclear reasons, it remains intact in epiphyseal specimens. Although castration in the male tends to slow growth and weight gain, the effects on cortical bone histomorphometry are present in pair-fed rats and in groups in which there was no difference in growth rates [5,6], indicating that the skeletal effects are not merely the indirect result of changes in body size or composition. The effects of androgens on cortical bone architecture have biomechanical implications. For instance, in studies of long-term androgen administration to female primates, Kasra and Grynpas [122] showed that cortical bone dimensions were increased and that tibias in treated animals were stronger, tougher, and stiffer. Thus, androgen deficiency during growth reduces cortical bone mass and strength primarily by blunting periosteal bone apposition. The lack of significant change in bone density following castration suggests that there is not a major impact of androgen deficiency on cortical porosity. Whereas estrogens appear to increase endosteal bone apposition, androgens probably have little effect at that site [117,122].
Cancellous bone mass is also reduced in castrate young male rats. Tibial metaphyseal bone volume and vertebral bone mineral density are clearly reduced [5,99,126], an effect that is seen rapidly following castration [5]. The reduction in bone volume is dramatic, with differences between control and castrate of 40 – 50% appearing in 4 – 10 weeks [99,131]. Rosen et al. [99] showed that measures of trabecular bone volume and mineral density diverged much more than areal measures of the proximal tibia or distal femur (by dual-energy X-ray absorptiometry) and speculated that this difference reflected a more intense bone deficit from trabecular than from cortical compartments. An important issue that remains unresolved is whether the bone deficit is a result of actual loss of bone mass following castration or whether the differences between castrate and control animals result from a failure of castrate animals to accrue bone normally. Nevertheless, bone changes following orchiectomy occur in the presence of an increase in skeletal blood flow [124,127], osteoclast numbers and surface [5], serum and urine calcium levels [5], and increased serum tartrateresistant acid phosphatase activity [128]. All these findings strongly suggest an increase in bone remodeling and bone resorption. However, in one report, distal femoral bone loss following castration was accompanied by a reduction in bone remodeling [132]. Parathyroid hormone concentrations have not been measured in these experiments, but vitamin D concentrations do not appear to be altered by orchiectomy [5]. In sum, trabecular bone mass, as well as cortical mass, appears to be dependent on adequate androgen action in the growing male animal, but the specific mechanisms that mediate that effect are not well delineated. For instance, the role of aromatization (see after) and the interaction of androgens and growth factors (e.g., IGF-1) are not clear.
C. Mature Male Animals
FIGURE 9
(A) The effect of ovariectomy (OVX) on periosteal bone formation rate. The mean SE (vertical bar) and tetracycline-labeling period (horizontal line) for intact controls () and OVX () rats are shown as a function of time after OVX. p 0.01 for all OVX time points compared to intact controls. (B) The effect of orchiectomy (ORX) on periosteal bone formation rate. The mean SE and tetracycline-labeling period for intact controls () and ORX () are shown as a function of time after ORX. p 0.01 for all ORX time points compared to same labeling period in intact controls [6].
In mature rats, castration also results in osteopenia. At a time when longitudinal growth has slowed markedly, pronounced differences between intact and castrate animals appear in cortical bone ash weight per unit length, crosssectional area, thickness, and bone mineral density [133 – 138] (Fig. 10 and 11). Periosteal bone accretion is reduced [138]. Because periosteal bone formation is known to be increased after oophorectomy, reduced by estrogens, and increased by androgens [6], the reduction in periosteal bone formation after orchidectomy is presumably mediated by a reduction in androgen action. Endocortical bone loss is accelerated in orchiectomized animals [135,139], and because endocortical remodeling is apparently not affected strongly by androgens, this may be a result of estrogen withdrawal by castration [5]. As might be expected in light of these changes, the maximum compressive load is
350
FIGURE 10
WIREN AND ORWOLL
Bone mass (unit ash: milligram of ash per millimeter of height), cross-sectional area, and wall thickness of the proximal femur cylinder from intact (I) and orchiectomized (O) rats (vertical bar SE). Comparison among mean values at different ages for intact (I**) and castrated (O*) rats. *P 0.05; *** P 0.001 [135].
FIGURE 11 Microphotographs of 200-m-thick mid diaphyseal cross sections from 24-month-old intact (a) and orchidectomized (b) rats taken with a polarization microscope. Magnification 14 [135].
decreased in cortical bone, although when corrected for cortical mass, bone strength is normal [135]. In addition to changes in bone size, increased intracortical resorption cavities are reported to result from orchiectomy [133,140]. Cancellous bone volume is reduced rapidly after castration as well (Fig. 12) [133,138], and osteopenia becomes pronounced with time [137]. This bone loss appears to result in part from increased bone resorption, as it is associated with increased resorption cavities, osteoclasts, and blood flow [133,134,138]. Dynamic histomorphometric and biochemical measures of bone remodeling increase quickly [126,137], with evidence of increased osteoclast numbers only 1 week after castration [138] (Table 3). These changes include an increase in osteoblastic activity, as well as increased bone resorption [138,140]. In the SAMP6 mouse, a model of accelerated senescence in which osteoblastic function is impaired, the rise in remodeling following orchidec-
tomy is blunted, which has been interpreted as evidence that the early changes after gonadectomy are dependent on osteoblast-derived signals [43]. Although estrogen has been shown to increase osteoprotegerin (OPG) production (and thus to reduce osteoclastognesis), the effects of androgens on OPG are unknown [141]. This information will be key to the understanding of the effects of androgen on remodeling, as androgens reduce osteoclast formation [142]. This initial phase of increased bone remodeling activity appears to subside somewhat with time [126,134] and by 4 months there is evidence of a depression in bone turnover rates in some skeletal areas (Fig. 13) [134]. As in younger animals, indices of mineral metabolism are not altered by these changes in skeletal metabolism [137]. As a potential model for the effects of hypogonadism in humans, animal models therefore suggest an early phase of high bone turnover and bone loss after orchidectomy,
351
CHAPTER 11 Skeletal Biology of Androgens
FIGURE 12 (2) Microradiograph of a midsagittal longitudinal section from the proximal end of a femur in a control rat illustrating the normal appearance of distribution of compact and spongy bone ( 7). (3) Microradiograph of a midsagittal longitudinal section from the proximal end of a femur in a 4-month postcastrate rat. Much of the spongiosa has been lost. ( 7). (4) Microradiograph of a midsagittal longitudinal section from the distal end of a femur in a control rat ( 14). (5) Microradiograph of a midsagittal longitudinal section from the distal end of a femur in a 4-month postcastrate rat. Metaphaseal spongiosa has almost disappeared and compacta is thin ( 14) [133].
followed by a later reduction in remodeling rates. How long bone loss continues, and at what rate, is unclear. Both cortical and trabecular compartments are affected. The remodeling imbalance responsible for loss of bone mass appears complex, as there are changes in rates of both bone formation and resorption, and patterns that vary from one skeletal compartment to another. These changes are very similar to those noted in female animals after castration, in which a loss of estrogen has been associated with a stimulation of osteoblast progenitor differentiation, in association with an
increase in osteoclast numbers, bone resorption, and bone loss [143]. The mechanisms by which androgens mediate their skeletal effects in mature animals are not known. The direct effects of androgen on bone cells suggest that androgens act in part without intermediates. However estrogens derived from aromatization (either in bone or other tissues) play a role. Because sex steroids influence growth hormone and growth factor regulation, they may also be involved. In fact, growth hormone treatment of orchidectomized animals
352
WIREN AND ORWOLL
TABLE 3
Static Histomorphometry of Proximal Tibial Metaphysis Cancellous Bone in Sham-Operated or Orchiectomized (ORX) 4-Month-Old Male Rats after 1, 2, or 4 Weeksa
N
Bone volume a (%)
Trabecular number (n/mm)
Trabecular thickness (m)
Osteoblast surface b (%)
Osteoclast surface c (%)
Number of osteoclasts d (n/mm)
Sham
10
10 1
2.6 0.3
41.0 2.6
4.9 0.7
2.1 0.3
0.67 0.07
ORX
9
10 1
2.4 0.3
41.6 2.7
3.8 0.8
3.2 0.7
1.03 0.22
Sham
9
91
2.4 0.3
39.8 2.6
4.4 1.0
1.7 0.3
0.50 0.1
ORX
9
71
2.0 0.4
38.6 2.2
8.3 2.8
2.8 0.5
0.81 0.1
1 week
2 weeks
4 weeks Sham
8
91
2.2 0.4
41.9 4.0
1.9 0.g
2.0 0.5
0.57 0.2
ORX
12
61
1.4 0.2
41.8 2.5
12.8 2.2*
3.8 0.5
1.8 0.1
a Data expressed as mean SEM. Significant treatment effect by two-way ANOVA: ap 0.02; cp 0.001, dp 0.002. Significant interaction effect by two-way ANOVA: bp 0.003. n, number. * Significantly different from sham , p 0.01. Data from Gunness and Orwoll [138].
recapitulates many (although not all) of the effects of testosterone replacement [140].
D. Androgens in the Female Animal
FIGURE 13 Evolution of the bone calcium turnover rate after castration (ratio of castrated/sham-operated animals). * P 0.05 [134].
Of course androgens are present in females as well as males and may affect bone metabolism. In castrate female rats, DHT administration suppresses elevated concentrations of bone resorption markers, as well as those of increased osteocalcin [144]. However, alkaline phosphatase activity increases further. Additional evidence to support the contention that androgens play a role in females includes the fact that antiandrogens are capable of evoking osteopenia in intact (i.e., fully estrogenized) female rats [7, 145] (Fig. 14). This obviously suggests that androgens provide crucial support to bone mass independent of estrogens. Of interest, the character of the bone loss induced by flutamide suggested that estrogen prevents bone resorption whereas androgens stimulate bone formation. In periosteal bone, DHT and testosterone appear to stimulate bone formation in orchidectomized young male rats, whereas in castrate females they suppress bone formation [6], perhaps reflecting an interaction or synergism between sex steroids and their effects on bone. There is also some information concerning androgens in other animal models, including primates. For instance, in adult female cynomolgus monkeys, testosterone treatment increased cortical and trabecular bone density as well as biomechanical strength [122].
353
CHAPTER 11 Skeletal Biology of Androgens
FIGURE 14
FIGURE 16
VI. THE ROLE OF AROMATIZATION: EFFECTS OF REPLACEMENT SEX STEROIDS AFTER CASTRATION
cancellous bone loss. In fact, estradiol resulted in an absolute increase in trabecular bone volume not achieved with androgen replacement. Similarly, estrogen was reported to antagonize the increase in blood flow resulting from castration and to increase bone ash weight more consistently than testosterone. Although data thus far available are incomplete, these studies raise obvious questions of the overlap between the actions of androgens and estrogens in bone. The gender reversal of estrogen replacement in male animals is also instructive. Nonaromatizable androgens are capable of preventing or reversing osteopenia and abnormalities in bone remodeling in oophorectomized females [6,146]. These actions apparently result from the suppression of cancellous bone resorption as well as stimulation of periosteal and endosteal bone formation [146]. Very similar results have been reported following the treatment of oophorectomized animals with dehydroepiandrosterone [6]. Moreover, blockage of androgen action with an AR antagonist in female rats already treated with an estrogen antagonist increases bone loss and indices of osteoclast activity more than treatment with an estrogen antagonist alone [147], indicating that ovarian androgens (apart from estrogens) exert a protective effect on bone. Analogously, androstenedione reduces (although does not abrogate) cancellous bone loss (and remodeling alterations) in oophorectomized animals treated with an aromatase inhibitor [148,149]. This protective effect was blocked by the addition of an AR antagonist [149]. Finally, whereas aromatase inhibition in male rats reduces bone mass, the large increase in remodeling induced by orchidectomy does not occur in these animals [112]. Also, orchidectomy in ERKO mice further reduces bone mass [150]. The latter observation implicates a role for androgens in the maintenance of bone mass in ERKO mice, but ER may also be playing a
Effects of buserelin alone, flutamide alone, and buserelin and flutamide in combination on total body calcium values after 4 weeks. Results are mean SD, n 7 – 8 [7].
Essentially all of the alterations induced by orchiectomy (in both growing and mature animals) can be prevented at least in part by replacement with either testosterone or nonaromatizable androgens [6,124,127,130,131,139,140] (Figs. 15 and 16). These results strongly suggest that aromatization of androgens to estrogens cannot fully explain the actions of androgens on bone metabolism. However estrogens also seem to prevent bone loss following castration in male animals. Vanderschueren et al. [137] reported that estradiol (and nandrolone) was capable of not only preventing the increase in biochemical indices stimulated by orchiectomy, but also preventing cortical and
FIGURE 15
Cancellous bone volume (BV/TV) in control and androgen-treated orchiectomized rats (vertical) compared to a reference group consisting of age-matched untreated intact rats of group 1 (horizontal lines). *p 0.05 [131].
Osteoblast-lined cancellous bone surface (percentage of total surface) in control and androgen-treated orchiectomized rats (vertical bars) compared to a reference group consisting of age-matched untreated intact rats (horizontal lines). *p 0.05 [131].
354 role. Newly available ER and KO mice have not yet been completely characterized. Aromatase-deficient mice have been developed [151], and the study of their skeletal phenotype should be very revealing. This should be a particularly useful model in view of human males with aromatase deficiency and impaired skeletal development [111]. In sum, these studies strongly support an independent role for androgens in skeletal homeostasis. Clearly estrogens play an important role in both genders. The interaction of the two is not yet defined but is of major interest.
VII. GENDER SPECIFICITY IN THE ACTIONS OF SEX STEROIDS Although still controversial, there may be gender specific responses in osteoblastic cells to sex steroids. Somjen and colleagues have shown that the increase in creatine kinase that occurs from bone cells in vivo and in vitro is gender specific (i.e., male animals or cells derived from male bones respond only to androgens, whereas females or female-derived cells respond only to estrogens) [130,152]. This gender specificity appears to depend on the previous history of exposure of animals to androgens (or estrogens). How much gender-specific effects might affect bone metabolism in the intact animal is completely unknown. In addition, in most mammals, there is a marked gender difference in morphology that results in a sexually dimorphic skeleton. The mechanisms responsible for these differences are obviously complex and presumably involve both androgenic and estrogenic actions on the skeleton. It is becoming increasingly clear that estrogens are particularly important for the regulation of epiphyseal function and act to reduce the rate of longitudinal growth via influences on chondrocyte proliferation and function, as well as on the timing of epiphyseal closure [117]. Androgens, however appear to have opposite effects on the skeleton, tending to promote long bone growth, chondrocyte maturation, and metaphyseal ossification. Furthermore, the most dramatic effect of androgens is on bone size, particularly cortical thickness [117,122], as androgens appear to have genderspecific effects on periosteal bone formation [6]. This difference has biomechanical implications, with thicker bones having greater resistance to fracture. At the same time, the response of the adult skeleton to the same intervention is distinct in males and females. For example, in a model of disuse osteopenia, antiorthostatic suspension significantly reduces the bone formation rate at the endocortical perimeter in males, but in females it decreases the bone formation rate along the periosteal perimeter [153]. Gender-specific responses in vivo and in vitro, and the mechanism(s) that underlies such responses in bone cells, may thus have significant implications in treatment options for metabolic bone disease.
WIREN AND ORWOLL
VIII. THE ANIMAL MODEL OF ANDROGEN RESISTANCE The testicular feminized (Tfm, AR deficient) male rat provides an interesting model for the study of the unique effects of androgens in bone. In these rats, androgens are presumed to be incapable of action, but estrogen and androstenedione concentrations are considerably higher than in normal males [154,155]. Clear increases also exist in Tfm male rats in serum concentrations of calcium, phosphorus, and osteocalcin, whereas IGF-1 concentrations are decreased. Estimates of bone mass suggest that Tfm rats have reduced longitudinal and radial growth rates, but that cancellous volume and density are similar to those of normal rats. In selected sites, measures of bone mass and remodeling were intermediate between normal male and female values. However, castration reduced bone volume markedly in Tfm male rats, suggesting a major role for estrogens in skeletal homeostasis (Fig. 17). This model indicates that androgens have an independent role to play in normal bone growth and metabolism, but the model is complex and not easily dissected. This is very similar to the study of humans with the androgen insensitivity syndrome. Marcus et al. [156] reported that there is a deficit in bone mineral density in women with androgen insensitivity even when compliance with estrogen replacement is excellent [156]. However, inadequate estrogen replacement appeared to worsen the deficit, and other environmental factors are difficult to quantitate.
IX. SUMMARY The influences of androgens on bone are obviously pervasive and complex. Androgens have a multiplicity of effects on skeletal cells in vitro. In vivo, they are particularly
FIGURE 17
Cancellous bone volume of the proximal metaphysis of the tibia in male, female, testicular feminized (Tfm), and orchiectomized (orch) male rats [155].
CHAPTER 11 Skeletal Biology of Androgens
dramatic during growth in males, but almost certainly play an important role during this period in females as well. Throughout the rest of life, androgens affect skeletal function in both sexes. Nevertheless, relatively little has been done to unravel the mechanisms by which androgens contribute to the physiology and pathophysiology of bone, and there is still much to be learned about the roles of androgens at all levels. The interaction of androgens and estrogens and how their respective actions can be utilized for specific diagnostic and therapeutic benefit are important but unanswered issues. With an increase in the understanding of the nature of androgen effects will come greater opportunities to use their positive actions in the prevention and treatment of a wide variety of bone disorders.
References 1. D. Vanderschueren and R. Bouillon, Androgens and bone. Calcif Tissue Int. 56, 341 – 346 (1995). 2. E. S. Orwoll, Androgens. In “Principles of Bone Biololy” (J. P. Bilezikian, L. G. Raisz, and G. A. Rodan, eds.), pp. 563 – 580. Academic Press, San Diego, 1996. 3. D. Vanderschueren, Androgens and their role in skeletal homeostasis. Horm. Res. 46, 95 – 98 (1996). 4. E. S. Orwoll, (ed.), Androgen and bone: Clinical aspects. In “Osteoporosis in Men,” pp. 247 – 274. Academic Press, San Diego, 1999. 5. R. T. Turner, K. S. Hannon, L. N. Demers, J. Buchanan, and N. Bell, Differential effects of gonadal function in bone histomorphometry in male and female rats. J. Bone Miner. Res. 4, 557 – 563 (1989). 6. R. T. Turner, G. K. Wakley, and K. S. Hannon, Differential effects of androgens on cortical bone histomorphometry in gonadectomized male and female rats. J. Orthop. Res. 8, 612 – 617 (1990). 7. A. Goulding and E. Gold, Flutamide-mediated androgen blockade evokes osteopenia in the female rat. J. Bone Miner. Res. 8, 763 – 769 (1993). 8. N. B. Watts, M. Notelovitz, M. C. Timmons, W. A. Addison, B. Wiita, and L. J. Downey, Comparison of oral estrogens and estrogens plus androgen on bone mineral density, menopausal symptoms, and lipid-lipoprotein profiles in surgical menopause. Obstet. Gynecol. 85, 529 – 537 (1995). 9. L. G. Raisz, B. Wiita, A. Artis, A. Bowen, S. Schwartz, M. Trahiotis, K. Shoukril, and J. Smith, Comparison of the effects of estrogen alone and estrogen plus androgen on biochemical markers of bone formation and resorption in postmenopausal women. J. Clin. Endocrinol. Metab. 81, 37 – 43 (1996). 10. M. J. Rosenberg, T. D. N. King, and M. C. Timmons, Estrogen-androgen for hormone replacement. J. Reprod. Med. 42, 394 – 404 (1997). 11. E. Barrett-Connor, Efficacy and safety of estrogen/androgen therapy. J. Reprod. Med. 43, 746 – 752 (1998). 12. V. Coxam, B. M. Bowman, M. Mecham, C. M. Roth, M. A. Miller, and S. A. Miller, Effects of dihydrotestosterone alone and combined with estrogen on bone mineral density, bone growth, and formation rates in ovariectomized rats. Bone 19, 107 – 114 (1996). 13. L. C. Hofbauer and S. Khosla, Androgen effects on bone metabolism: Recent progress and controversies. Eur. J. Endocrinol. 140, 271 – 286 (1999). 14. E. O. Abu, A. Horner, V. Kusec, J. T. Triffitt, and J. E. Compston, The localization of androgen receptors in human bone. J. Clin. Endocrinol. Metab. 82, 3493 – 3497 (1997).
355 15. C. Chang, J. Kokontis, and S. Liao, Structural analysis of complementary DNA and amino acid sequences of human and rat androgen receptors. Proc. Natl. Acad. Sci. USA 85, 7211 – 7215 (1988). 16. D. B. Lubahn, D. R. Joseph, M. Sar, J. Tan, H. N. Higgs, R. E. Larson, F. S. French, and E. M. Wilson, The human androgen receptor: Complementary deoxyribonucleic acid cloning, sequence analysis, and gene expression in prostate. Mol. Endocrinol. 2, 1265 – 1275 (1988). 17. D. J. Mangelsdorf, C. Thummel, M. Beato, P. Herrlich, G. Schutz, K. Umesono, B. Blumberg, P. Kastner, M. Mark, P. Chambon, and R. M. Evans, The nuclear receptor superfamily: The second decade. Cell 83, 835 – 839 (1995). 18. Y. H. Zhuang, M. Blauer, A. Pekki, and P. Tuohimaa, Subcellular location of androgen receptor in rat prostate, seminal vesicle and human osteosarcoma MG-63 cells. J. Steroid Biochem. Mol. Biol. 41, 693 – 696 (1992). 19. C. Chang, N. Saltzman, S. Yeh, W. Young, E. Keller, H.-J. Lee, C. Wang, and A. Mizokami, Androgen receptor: An overview. Crit. Rev. Eukaryot. Gene Expr. 5, 97 – 125 (1995). 20. D. Colvard, T. Spelsberg, E. Eriksen, P. Keeting, and L. B. Riggs, Evidence of steroid receptors in human osteoblast-like cells. Connect. Tissue Res. 20, 33 – 40 (1989). 21. D. J. Benz, M. R. Haussler, M. A. Thomas, B. Speelman, and B. S. Komm, High-affinity androgen binding and androgenic regulation of a1(I)-procollagen and transforming growth factor-B steady state messenger ribonucleic acid levels in human osteoblast-like osteosarcoma cells. Endocrinology 128, 2723 – 2730 (1991). 22. E. S. Orwoll, L. Stribrska, E. E. Ramsay, and E. Keenan, Androgen receptors in osteoblast-like cell lines. Calcif. Tissue Int. 49, 182 – 187 (1991). 23. P. Liesegang, G. Romalo, M. Sudmann, L. Wolf, and H.-U. Schweikert, Human osteoblast-like cells contain specific saturable, high-affinity glucocorticoid, androgen, estrogen and 1a,25-dihydroxycholecalciferol receptors. J. Andro. 15, 194 – 199 (1994). 24. Y. Nakano, I. Morimoto, O. Ishida, T. Fujihira, A. Mizokami, A. Tanimoto, N. Yanagihara, F. Izumi, and S. Eto, The receptor, metabolism and effects of androgen in osteoblastic MC3T3-E1 cells. Bone Miner. 26, 245 – 259 (1994). 25. M. Takeuchi, H. Kakushi, and M. Tohkin, Androgens directly stimulate mineralization and increase androgen receptors in human osteoblast-like osteosarcoma cells. Biochem. Biophys. Res. Commun. 204, 905 – 911 (1994). 26. C. H. Kasperk, G. K. Wakley, T. Hierl, and R. Ziegler, Gonadal and adrenal androgens are potent regulators of human bone cell metabolism in vitro. J. Bone Miner. Res. 12, 464 – 471 (1997). 27. C. M. Wilson and M. J. McPhaul, A and B forms of the androgen receptor are present in human genital skin fibroblasts. Proc. Natl. Acad. Sci. USA 91, 1234 – 1238 (1994). 28. T. Gao and M. J. McPhaul, Functional activities of the A and B forms of the human androgen receptor in response to androgen receptor agonists and antagonists. Mol. Endocrinol. 12, 654 – 663 (1998). 29. A. Masuyama, Y. Ouchi, F. Sato, T. Hosoi, T. Nakamura, and H. Orimo, Characteristics of steroid hormone receptors in cultured MC3T3-E1 osteoblastic cells and effect of steroid hormones on cell proliferation. Calcif. Tissue Int. 51, 376 – 381 (1992). 30. C. Kasperk, A. Helmboldt, I. Borcsok , S. Heuthe, O. Cloos, F. Niethard, and R. Ziegler, Skeletal site-dependent expression of the androgen receptor in human osteoblastic cell populations. Calcif. Tissue Int. 61, 464 – 473 (1997). 31. A. W. Meikle, R. W. Dorchuck, B. A. Araneo, J. D. Stringham, T. G. Evans, S. L. Spruance, and R. A. Daynes, The presence of a dehydroepiandrosterone-specific receptor binding complex in murine T cells. J. Steroid Biochem. Mol. Biol. 42, 293 – 304 (1992). 32. P. V. N. Bodine, B. L. Riggs, and T. C. Spelsberg, Regulation of c-fos expression and TGF-B production by gonadal and adrenal
356
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49. 50.
WIREN AND ORWOLL androgens in normal human osteoblastic cells. J. Steroid Biochem. Mol. Biol. 52, 149 – 158 (1995). K. Wrogemann, G. Podolsky, J. Gu, and E. Rosenmann, A 63 -kDa protein with androgen-binding activity is not from the androgen receptor. Biochem. Cell. Biol. 69, 695 – 701 (1991). C. A. Pettaway, Racial differences in the androgen/androgen receptor pathway in prostate cancer. J. Natl. Med. Assoc. 91, 653 – 660 (1999). A. J. Grierson, R. C. Mootoosamy, and C. C. J. Miller, Polyglutamine repeat length influences human androgen receptor/c-Jun mediated transcription. Neurosci. Lett. 277, 9 – 12 (1999). H. Peterziel, S. Mink, A. Schonert, M. Becker, H. Klocker, and A. C. B. Cato, Rapid signalling by androgen receptor in prostate cancer cells. Oncogene 18, 6322 – 6329 (1999). M. Lieberherr and B. Grosse, Androgens increase intracellular calcium concentration and inositol 1,4,5-triphosphate and diacylglycerol formation via a pertussis toxin-sensitive G-protein. J. Biol. Chem. 269, 7217 – 7223 (1994). M. Lieberherr, B. Grosse, M. Kachkache, and S. Balsan, Cell signaling and estrogens in female rat osteoblasts: A possible involvement of unconventional nonnuclear receptors. J. Bone Miner. Res. 8, 1365 – 1376 (1993). T. Bellido, R. J. Jilka, B. F. Boyce, G. Girasole, H. Broxmeyer, A. Dalrymple, R. Murry, and S. C. Manolagas, Regulation of interleukin-6, osteoclastogenesis and bone mass by androgens: The role of the androgen receptor. J. Clin. Invest. 95, 2886 – 2895 (1995). A. Carracosa, L. Audi, M. A. Ferrandez, and A. Ballabriga, Biological effects of androgens and identification of specific dihydrotestosterone-binding sites in cultured human fetal epiphyseal chondrocytes. J. Clin. Endocrinol. Metab. 70, 134 – 140 (1990). M. J. Oursler, P. Osdoby, J. Pyfferoen, B. L. Riggs, and T. C. Spelsberg, Avian osteoclasts as estrogen target cells. Proc. Natl. Acad. Sci. USA 88, 6613 – 6617 (1991). Y. Mizuno, T. Hosoi, S. Inoue, A. Ikegami, M. Kaneki, Y. Akedo, T. Nakamura, Y. Ouchi, C. Chang, and H. Orimo, Immunocytochemical identification of androgen receptor in mouse osteoclast-like multinucleated cells. Calcif. Tissue Int. 54, 325 – 326 (1994). R. S. Weinstein, R. L. Jilka, A. M. Parfitt, and S. C. Manolagas, The effects of androgen deficiency on murine bone remodeling and bone mineral density are mediated via cells of the osteoblastic lineage. Endocrinology 138, 4013 – 4021 (1997). K. M. Wiren, X. Zhang, C. Chang, E. Keenan, and E. Orwoll, Transcriptional up-regulation of the human androgen receptor by androgen in bone cells. Endocrinology 138, 2291 – 2300 (1997). K. Wiren, E. Keenan, X. Zhang, B. Ramsey, and E. Orwoll, Homologous androgen receptor up-regulation in osteoblastic cells may be associated with enhanced functional androgen responsiveness. Endocrinology 140, 3114 – 3124 (1999). L. C. Hofbauer, K. C. Hicok, M. J. Schroeder, S. A. Harris, J. A. Robinson, and S. Khosla, Development and characterization of a conditionally immortalized human osteoblastic cell line stably transfected with the human androgen receptor gene. J. Cell. Biochem. 66, 542 – 551 (1997). J. A. Kemppainen, M. V. Lane, M. Sar, and E. M. Wilson, Androgen receptor phosphorylation, turnover, nuclear transport, and transcriptional activation. J. Biol. Chem. 267, 968 – 974 (1992). T. Ikonen, J. J. Palvimo, P. J. Kallio, P. Reinikainen, and O. A. Janne, Stimulation of androgen-regulated transactivation by modulators of protein phosphorylation. Endocrinology 135, 1359 – 1366 (1994). L. J. Blok, P. E. de Ruiter, and A. O. Brinkmann, Androgen receptor phosphorylation. Endocrin. Res. 22, 197 – 219 (1996). A. Culig, A. Hobisch, M. V. Cronauer, C. Radmayr, J. Trapman, A. Hittmair, G. Bartsch, and H. Klocker, Androgen receptor activation in prostatic tumor cell lines by insulin-like growth factor-I, keratinocyte growth factor, and epidermal growth factor. Cancer Res. 54, 5474 – 5478 (1994).
51. K. B. Horwitz, T. A. Jackson, D. L. Bain, J. K. Richer, G. S. Takimoto, and L. Tung, Nuclear receptor coactivators and corepressors. Mol. Endocrinol. 10, 1167 – 1177 (1996). 52. N. J. McKenna, R. B. Lanz, and B. W. O’Malley, Nuclear receptor coregulators: cellular and molecular biology. Endocr. Rev. 20, 321 – 344 (1999). 53. H. E. MacLean, G. L. Warne, and J. D. Zajac, Localization of functional domains in the androgen receptor. J. Steroid Biochem. Mol. Biol. 62, 233 – 242 (1997). 54. S. Yeh and C. Chang, Cloning and characterization of a specific coactivator, ARA70, for the androgen receptor in human prostate cells. Proc. Natl. Acad. Sci. USA 93, 5517 – 5521 (1996). 55. M. R. Haussler, C. A. Haussler, P. W. Jurutka, P. D. Thompson, J. C. Hsieh, L. S. Remus, S. H. Selznick, and G. K. Whitfield, The vitamin D hormone and its nuclear receptor: Molecular actions and disease states. J. Endocrinol. 154 (Suppl), S57 – S73 (1997). 56. K. M. Wiren, J. R. Lundblad, X.-W. Zhang, and E. S. Orwoll, Androgen receptor function in osteoblasts: Expression of coactivators and cointegrators. Endocr. Soc. [Abstract P2-466] 57. L. C. Hofbauer, K. C. Hicok, and S. Khosla, Effects of gonadal and adrenal androgens in a novel androgen-responsive human osteoblastic cell line. J. Cell Biochem. 71, 96 – 108 (1998). 58. C. H. Kasperk, J. E. Wergedal, J. R. Farley, T. A. Linkhart, R. T. Turner, and D. J. Baylink, Androgens directly stimulate proliferation of bone cells in vitro. Endocrinology 124(3), 1576 – 1578 (1989). 59. C. Kasperk, R. Fitzsimmons, D. Strong, S. Mohan, J. Jennings, J. Wergedal, and D. Baylink, Studies of the mechanism by which androgens enhance mitogenesis and differentiation in bone cells. J. Clin. Endocrinol. Metab. 71, 1322 – 1329 (1990). 60. D. Somjen, Y. Weisman, A. Harell, E. Berger, and A. M. Kaye, Direct and sex-specific stimulation by sex steroids of creatine kinase activity and DNA synthesis in rat bone. Proc. Natl. Acad. Sci. USA 86, 3361 – 3365 (1989). 61. A. H. Wyllie, J. F. R. Kerr, and A. R. Currie, Cell death: The significance of apoptosis. Int. Rev. Cytol. 68, 251 – 307 (1980). 62. D. E. Hughes and B. F. Boyce, Apoptosis in bone physiology and disease. J. Clini. Pathol. Mol. Pathol. 50, 132 – 137 (1997). 63. S. C. Manolagas and R. S. Weinstein, New developments in the pathogenesis and treatment of steroid-induced osteoporosis. J. Bone. Miner. Res. 14, 1061 – 1066 (1999). 64. M. P. Lynch, C. Capparelli, J. L. Stein, G. S. Stein, and J. B. Lian, Apoptosis during bone-like tissue development in vitro. J. Cell. Biochem. 68, 31 – 49 (1998). 65. D. W. Dempster, B. S. Moonga, L. S. Stein, W. R. Horbert, and T. Antakly, Glucocorticoids inhibit bone resorption by isolated rat osteoclasts by enhancing apoptosis. J. Endocrinol. 154, 397 – 406 (1997). 66. S. C. Manolagas, Cellular and molecular mechanisms of osteoporosis. Aging (Milano) 10, 182 – 190 (1998). 67. R. S. Weinstein, R. L. Jilka, A. M. Parfitt, and S. C. Manolagas, Inihibition of osteoblastogenesis and promotion of apoptosis of osteoblasts and osteocytes by glucocorticoid: Potential mechanisms of their deleterious effects on bone. J. Clin. Invest. 104, 1363 – 1374 (1999). 68. A. Gohel, M. B. McCarthy, and G. Gronowicz, Estrogen prevents glucocorticoid-induced apoptosis in osteoblasts in vivo and in vitro. Endocrinology 140, 5339 – 5347 (1999). 69. A. Tomkinson, J. Reeve, R. W. Shaw, and B. S. Noble, The death of osteocytes via apoptosis accompanies estrogen withdrawal in human bone. J. Clin. Endocrinol. Metab. 82, 3128 – 3135 (1997). 70. A. Tomkinson, E. F. Gevers, J. M. Wit, J. Reeve, and B. S. Noble, The role of estrogen in the control of rat osteocyte apoptosis. J. Bone Miner. Res. 18, 1243 – 1250 (1998). 71. M. T. Abreu-Martin, A. Chari, A. A. Palladino, N. A. Craft, and C. L. Sawyers, Mitogen-activated protein kinase kinase kinase 1 activates androgen receptor-dependent transcription and apoptosis in prostate cancer. Mol. Cell. Biol. 19, 5143 – 5154 (1999).
CHAPTER 11 Skeletal Biology of Androgens 72. I. K. Lim, H. J. Joo, K. S. Choi, E. Sueoka, M. S. Lee, M. S. Ryu, and H. Fuijiki, Protection of 5-dihydrotestosterone against TGF-Binduced apoptosis in FaO cells and induction of mitosis in HepG2 cells. Int. J. Cancer 72, 351 – 355 (1997). 73. C. H. Kasperk, K. Faehling, I. Borcsok, and R. Ziegler, Effects of androgens on subpopulations of the human osteosarcoma cell line SaOS2. Calcif. Tissue Int. 58, 376 – 382 (1996). 74. C. Gray, K. W. Colston, A. G. Mackay, M. L. Taylor, and T. R. Arnett, Interaction of androgen and 1,25-dihydroxyvitamin D3: Effects on normal rat bone cells. J. Bone Miner. Res. 7, 41 – 46 (1992). 75. S. P. Kapur and A. H. Reddi, Influence of testosterone and dihydrotestosterone on bond-matrix induced endochondral bone formation. Calcif. Tissue Int. 44, 108 – 113 (1989). 76. M. C. Horowitz, Cytokines and estrogen in bone: Anti-osteoporotic effects. Science 260, 626 – 627 (1993). 77. M. Kassem, S. A. Harris, T. C. Spelsberg, and B. L. Riggs, Estrogen inhibits interleukin-6 production and gene expression in a human osteoblastic cell line with high levels of estrogen receptors. J. Bone. Miner. Res. 11, 193 – 199 (1996). 78. H. Kawaguchi, C. C. Pilbeam, S. J. Vargas, E. E. Morse, J. C. Lorenzo, and L. G. Raisz, Ovariectomy enhances and estrogen replacement inhibits the activity of bone marrow factors that stimulate prostaglandin production in cultured mouse calvariae. J. Clin. Invest. 96, 539 – 548 (1995). 79. M. Centrella, M. C. Horowitz, J. M. Wozney, and T. L. McCarthy, Transforming growth factor- gene family members and bone. Endocr. Rev. 15, 27 – 39 (1994). 80. S. E. Harris, L. F. Bonewald, M. A. Harris, M. Sabatini, S. Dallas, J. Q. Feng, N. Ghosh-Choudhury, J. Wozney, and G. R. Mundy, Effects of transforming growth factor on bone nodule formation and expression of bone morphogenetic protein 2, osteocalcin, osteopontin, alkaline phosphatase, and type I collagen mRNa in long-term cultures of fetal rat calvarial osteoblasts. J. Bone. Miner. Res. 9, 855 – 863 (1994). 81. M. Subramaniam, S. A. Harris, M. J. Oursler, K. Rasmussen, B. L. Riggs, and T. C. Spelsberg, Identification of a novel TGF-betaregulated gene encoding a putative zinc finger protein in human osteoblasts. Nucleic Acids Res. 23, 4907 – 4912 (1995). 82. R. K. Gill, R. T. Turner, T. J. Wronski, and N. H. Bell, Orchiectomy markedly reduces the concentration of the three isoforms of transforming growth factor in rat bone, and reduction is prevented by testosterone. Endocrinology 139, 546 – 550 (1998). 83. E. Canalis, M. Centrella, and T. L. McCarthy, Regulation of insulinlike growth factor-II production in bone cultures. Endocrinology 129, 2457 – 2462. 84. F. Gori, L. C. Haufbauer, C. A. Conover, and S. Khosla, Effects of androgens on the insulin-like growth factor system in an androgenresponsive human osteoblastic cell line. Endocrinology 140, 5579 – 5586 (1999). 85. S. Fukayama and H. J. Tashjian, Direct modulation by androgens of the response of human bone cells (SaOS-2) to human parathyroid hormone (PTH) and PTH-related protein. Endocrinology 125, 1789 – 1794 (1989). 86. A. Gray, H. A. Feldman, J. B. McKinlay, and C. Longcope, Age, disease and changing sex hormone levels in middle-aged men: Results of the Massachusetts male aging study. J. Clin. Endocrinol. Metab. 73, 1016 – 1025 (1991). 87. A. Vermeulen, Clinical review 24: Androgens in the aging male. J. Clin. Endocrinol. Metab. 73, 221 – 224 (1991). 88. C. C. Pilbeam and L. G. Raisz, Effects of androgens on parathyroid hormone and interleukin-1-stimulated prostaglandin production in cultured neonatal mouse calvariae. J. Bone Miner. Res. 5, 1183 – 1188 (1990). 89. G. Passeri, G. Girasole, R. L. Jilka, and S. C. Manolagas, Increased interleukin-6 production by murine bone marrow and bone cells after estrogen withdrawal. Endocrinology 133, 822 – 828 (1993).
357 90. S.-C. Lin, T. Yamate, Y. Taguchi, V. Z. C. Borba, G. Girasole, C. A. O’Brian, T. Bellido, E. Abe, and S. C. Manolagos, Regulation of the gp80 and gp130 subunits of the IL-6 receptor by sex steroids in the murine bone marrow. J. Clin. Invest. 101, 1942 – 1950 (1997). 91. J. Vittek, K. Altman, G. G. Gordon, and A. L. Southren, The metabolism of 7a-3H-testosterone by rat mandibular bone. Endocrinology 94, 325 – 329 (1974). 92. H. U. Schweikert, W. Rulf, N. Niederle, H. E. Schafer, E. Kreck, and F. Kruck, Testosterone metabolism in human bone. Acta Endocrinol. 95, 258 – 264 (1980). 93. D. W. Russell and J. D. Wilson, Steroid 5-reductase: Two genes/two enzymes. Annu. Rev. Biochem. 63 (1994). 94. H.-R. Bruch, L. Wolf, R. Budde, G. Romalo, and H.-U. Schweikert, Androstenedione metabolism in cultured human osteoblast-like cells. J. Clin. Endocrinol. Metab. 75, 101 – 105 (1992). 95. L. Audi, A. Carrascosa, and A. Ballabriga, Androgen metabolism by human fetal epiphyseal cartilage and its condrocytes in primary culture. J. Clin. Endocrinol. Metab. 58, 819 – 825 (1984). 96. R. T. Turner, B. Bleiberg, D. S. Colvard, P. E. Keeting, G. Evans, and T. C. Spelsberg, Failure of isolated rat tibial periosteal cells to 5a reduce testosterone to 5a-dihydrotestosterone. J. Bone Miner. Res. 5, 775 – 779 (1990). 97. L. Fisher, M. D. Kogut, R. J. Moore, U. Goebelsmann, J. J. Weitzmann, H. Isaacs, J. E. Griffin, and J. D. Wilson, Clinical, endocrinological and enzymatic characterization of two patients with 5areductase deficiency: Evidence that a single enzyme is responsible for the 5a-reduction of cortisol and testosterone. J. Clin. Endocrinol. Metab. 47, 653 – 664 (1978). 98. M. S. Mahendroo, K. M. Cala, C. P. Landrum, and D. W. Russell, Fetal death in mice lacking 5-reductase type 1 caused by estrogen excess. Mol. Endocrinol. 11, 917 – 927 (1997). 99. H. N. Rosen, S. Tollin, R. Balena, V. L. Middlebrooks, W. G. Beamer, L. R. Donohue, C. Rosen, A. Turner, M. Holick, and S. L. Greenspan, Differentiating between orchiectomized rats and controls using measurements of trabecular bone density: A comparison among DXA, histomorphometry, and peripheral quantitative computerized tomography. Calcif. Tissue Int. 57, 35 – 39 (1995). 100. E. R. Simpson, M. S. Mahendroo, G. D. Means, M. W. Kilgore, M. M. Hinshelwood, S. Graham-Lorence, B. Amarneh, Y. Ito, C. R. Fisher, M. D. Michael, C. R. Mendelson, and S. E. Bulun, Aromatase cytochrome P450, the enzyme responsible for estrogen biosynthesis. Endocr. Rev 15, 342 – 355 (1994). 101. H. Nawata, S. Tanaka, R. Takayanagi, Y. Sakai, T. Yanase, S. Ikuyama, and M. Haji, Aromatase in bone cell: Association with osteoporosis in postmenopausal women. J. Steroid Biochem. Mol. Biol. 53, 165 – 174 (1995). 102. H.-U. Schweikert, L. Wolf, and G. Romalo, Estrogen formation from androstenedione in human bone. Clin. Endocrinol. 43, 37 – 42 (1995). 103. H. Sasano, M. Uzuki, T. Sawai, H. Nagura, G. Matsunaga, O. Kashimoto, and N. Harada, Aromatase in human bone tissue. J. Bone Miner. Res. 12, 1416 – 1423 (1997). 104. A. Purohit, A. M. Flanagan, and M. J. Reed, Estrogen synthesis by osteoblast cell lines. Endocrinology 131, 2027 – 2029 (1992). 105. S. Tanaka, M. Haji, Y. Nishi, T. Yanase, R. Takayanagi, and H. Nawata, Aromatase activity in human osteoblast-like osteosarcoma cell. Calcif. Tissue Int. 52, 107 – 109 (1993). 106. S. E. Abdelgadir, J. A. Resko, S. R. Ojeda, E. D. Lephart, M. J. McPhaul, and C. E. Roselli, Androgens regulate aromatase cytochrome P450 messenger ribonucleic acid in rat brain. Endocrinology 135, 395 – 401 (1994). 107. F. A. Conte, M. M. Grumbach, Y. Ito, C. R. Fisher, and E. R. Simpson, A syndrome of female pseudohermaphrodism, hypergonadotropic hypogonadism, and multicystic ovaries associated with missense mutations in the gene encoding aromatase (P450arom). J. Clin. Endocrinol. Metab. 78, 1287 – 1292 (1994).
358 108. C. Carani, K. Qin, M. Simoni, M. Faustini-Fustini, S. Serpente, J. Boyd, K. S. Korach, and E. R. Simpson, Effect of testosterone and estradiol in a man with aromatase deficiency. N. Engl. J. Med. 337, 91 – 95 (1997). 109. A. Morishima, M. M. Grumbach, E. R. Simpson, C. Fisher, and K. Qin, Aromatase deficiency in male and female siblings caused by a novel mutation and the physiological role of estrogens. J. Clin. Endocrinol. Metab. 80, 3689 – 3698 (1995). 110. E. P. Smith, J. Boyd, G. R. Frank, H. Takahashi, R. M. Cohen, B. Specker, T. C. Williams, D. B. Lubahn, and K. S. Korach, Estrogen resistance caused by a mutation in the estrogen-receptor gene in a man. N. Engl. J. Med. 331, 1056 – 1061 (1994). 111. J. P. Bilezikian, A. Morishima, J. Bell, and M. M. Grumbach, Increased bone mass as a result of estrogen therapy in a man with aromatase deficiency. N. Engl. J. Med. 339, 599 – 603 (1998). 112. D. Vanderschueren, E. Van Herck, J. Nijs, and A. G. H. Ederveen, Aromatase inhibition impairs skeletal modeling and decreases bone mineral density in growing male rats. Endocrinology 138, 2301 – 2307 (1997). 113. D. Vanderschueren, E. Van Herck, R. De Coster, and R. Bouillon, Aromatization of androgens is important for skeletal maintenance of aged male rats. Calcif. Tissue Int. 59, 179 – 183 (1996). 114. D. Somjen, Y. Weisman, Z. Mor, A. Harell, and A. M. Kaye, Regulation of proliferation of rat cartilage and bone by sex steroid hormones. J. Steroid Biochem. Mol. Biol. 40, 717 – 723 (1991). 115. M. T. Corvol, A. Carracosa, L. Tsagris, O. Blanchard, and R. Rappaport, Evidence for a direct in vitro action of sex steroids on rabbit cartilage cells during skeletal growth: Influence of age and sex. Endocrinology 120, 1422 – 1429 (1987). 116. Z. Schwartz, E. Nasatzky, A. Ornoy, B. P. Brooks, W. A. Soskolne, and B. D. Boyan, Gender-specific, maturation-dependent effects of testosterone on chondrocytes in culture. Endocrinology 134, 1640 – 1647 (1994). 117. R. T. Turner, B. L. Riggs, and T. C. Spelsberg, Skeletal effects of estrogen. Endocr. Rev. 15, 275 – 300 (1994). 118. H. E. Lebovitz and G. S. Eisenbarth, Hormonal regulation of cartilage growth and metabolism. Vitam. Horm. (NY) 33, 575 – 648 (1975). 119. J. P. Iannotti, Growth plate physiology and pathology. Orthop. Clin. North Am. 21, 1 – 17 (1990). 120. S. G. Ren, S. Malozowski, P. Sanchez, D. E. Sweet, D. L. Loriaux, and F. Cassorla, Direct administration of testosterone increases rat tibial epiphyseal growth plate width. Acta Endocr. (Copenhagen) 21, 401 – 405 (1989). 121. L. E. Underwood and J. J. Van Wyk, Normal and aberrant growth. In “Williams Textbook of Endocrinology” (J. D. Wilson and D. W. Foster, eds.), 8th Ed., pp. 1079 – 1138. Saunders, Philadelphia, PA, 1992. 122. M. Kasra and M. D. Grynpas, The effects of androgens on the mechanical properties of primate bone. Bone 17, 265 – 270 (1995). 123. P. D. Saville and C. S. Lieber, Increases in skeletal calcium and femur cortex thickness in undernutrition. J. Nutr. 99, 141 – 144 (1969). 124. A. Schoutens, M. Verhas, M. L’Hermite-Baleriaux, M. L’Hermite, A. Verschaeren, N. Dourov, M. Mone, A. Heilporn, and A. Tricot, Growth and bone haemodynamic responses to castration in male rats: Reversibility by testosterone. Acta Endocrinol. 107, 428 – 432 (1984). 125. J. M. Hock, I. Gera, J. Fonseca, and L. G. Raisz, Human parathyroid hormone (1 – 34) increases bone mass in ovariectomized and orchidectomized rats. Endocrinology 122, 2899 – 2904 (1988). 126. D. Vanderschueren, I. Jans, E. Van Herck, K. Moermans, J. Verhaeghe, and R. Bouillon, Time-related increase of biochemical markers of bone turnover in androgen-deficient male rats. Bone Miner. 26, 123 – 131 (1994). 127. J. Kapitola, J. Kubickova, and J. Andrle, Blood flow and mineral content of the tibia of female and male rats: Changes following cas-
WIREN AND ORWOLL
128.
129.
130.
131.
132.
133.
134.
135.
136.
137.
138.
139.
140.
141.
142.
143.
144.
145.
tration and/or administration of estradiol or testosterone. Bone 16, 69 – 72 (1995). B. Ongphiphadhanakul, S. Alex, L. F. Braverman, and D. T. Baran, Excessive L-thyroxine therapy decreases femoral bone densities in the male rat: Effect of hypogonadism and calcitonin. J. Bone Miner. Res. 7, 1227 – 1231 (1992). A. Ornoy, S. Giron, R. Aner, M. Goldstein, B. D. Boyan, and Z. Schwartz, Gender dependent effects of testosterone and 17bestradiol on bone growth and modelling in young mice. Bone Miner. 24, 43 – 58 (1994). D. Somjen, Z. Mor, and A. M. Kaye, Age dependence and modulation by gonadectomy of the sex-specific response of rat diaphyseal bone to gonadal steroids. Endocrinology 134, 809 – 814 (1994). G. K. Wakley, H. D. J. Schutte, K. S. Hannon, and R. T. Turner, Androgen treatment prevents loss of cancellous bone in the orchidectomized rat. J. Bone Miner. Res. 6, 325 – 330 (1991). J. Tuukkanen, Z. Peng, and H. K. Vaananen, Effect of running exercise on the bone loss induced by orchidectomy in the rat. Calcif. Tissue Int. 55, 33 – 37 (1994). C. S. Wink and W. J. L. Felts, Effects of castration on the bone structure of male rats: a model of osteoporosis. Calcif. Tissue Int. 32, 77 – 82 (1980). M. Verhas, A. Schoutens, M. L’hermite-Baleriaux, N. Dourov, A. Verschaeren, M. Mone, and A. Heilporn, The effect of orchidectomy on bone metabolism in aging rats. Calcif. Tissue Int. 39, 74 – 77 (1986). C. C. Danielsen, L. Mosekilde, and T. T. Andreassen, Long-term effect of orchidectomy on cortical bone from rat femur: Bone mass and mechanical properties. Calcif. Tissue Int. 50, 169 – 174 (1992). D. Vanderschueren and E. Van Herck, Androgen deficiency effects in the rat skeleton. Second Workshop on Using the Live Rate in Skeletal Studies, Chicago. D. Vanderschueren, E. Van Herck, A. M. H. Suiker, W. J. Visser, L. P. C. Schot, and R. Bouillon, Bone and mineral metabolism in aged male rats: Short and long term effects of androgen deficiency. Endocrinology 130, 2906 – 2916 (1992). M. Gunness and E. S. Orwoll, Early induction of alterations in cancellous and cortical bone histology after orchiectomy in mature rats. J. Bone. Miner. Res. 10, 1735 – 1744 (1995). D. Vanderschueren, E. Van Herck, P. Schot, E. Rush, T. Einhorn, P. Geusens, and R. Bouillon, The aged male rat as a model for human osteoporosis: Evaluation by nondestructive measurements and biomechanical testing. Calcif. Tissue Int. 53, 342 – 347 (1993). G. Prakasam, J. K. Yeh, M.-M. Chen, M. Castro-Magana, C. T. Liang, and J. F. Aloia, Effects of growth hormone and testosterone on cortical bone formation and bone density in aged orchiectomized rats. Bone 24, 491 – 497 (1999). N. O. Vidal, H. Brandstrom, K. B. Jonsson, and C. Ohlsson, Osteoprotegerin mRNA is expressed in primary human osteoblast-like cells: Downregulation by gluccocorticoids. J. Endocrinol. 159, 191 – 195 (1998). S. Manolagas, T. Bellido, and R. Jilka, Sex steroids, cytokines and the bone marrow: New concepts on the pathogenesis of osteoporosis. In “Non-reproductive actions of Sex Steroids,” pp. 187 – 202. Wiley, Chichester, 1995. R. L. Jilka, K. Takahashi, M. Munshi, D. C. Williams, P. K. Roberson, and S. C. Manolagas, Loss of estrogen upregulates osteoblastogenesis in the murine bone marrow. J. Clin. Invest. 101, 1942 – 1950 (1998). R. A. Mason and H. A. Morris, Effects of dihydrotestosterone on bone biochemical markers in sham and oophorectomized rats. J. Bone Miner. Res. 12, 1431 – 1437 (1997). C. Lea, N. Kendall, and A. M. Flanagan, Casodex (a nonsteroidal antiandrogen) reduces cancellous, endosteal, and periosteal bone formation in estrogen-replete female rats. Calcif. Tissue Int. 58, 268 – 272 (1996).
CHAPTER 11 Skeletal Biology of Androgens 146. J. H. Tobias, A. Gallagher, and T. J. Chambers, 5a-dihydrotestosterone partially restores cancellous bone volume in osteopenic ovariectomized rats. Am J. Physiol. 267, E853 – E859 (1994). 147. C. K. Lea and A. M. Flanagan, Ovarian androgens protect against bone loss in rats made oestrogen deficient by treatment with ICI 182,780. J. Endocrinol. 160, 111 – 117 (1999). 148. C. K. Lea, V. Moxham, M. J. Reed, and A. M. Flanagan, Androstenedione treatment reduces loss of cancellous bone volume in ovariectomized rats in a dose-responsive manner and the effect is not mediated by oestrogen. J. Endocrinol. 156, 331 – 339 (1998). 149. C. K. Lea and A. M. Flanagan, Physiological plasma levels of androgens reduce bone loss in the ovariectomized rat. Am. J. Physiol. 274, E328 – 335 (1998). 150. M. Taki, K. S. Kimbro, T. F. Washburn, A. Kiyokawa, F. Okada, C. Fukuda, M. Miyamoto, and K. S. Korach, Effects of estrogen receptor destruction on femurs from male mice. J. Bone Miner. Res. 12 (Suppl. 1), S430 (1997). [Abstract] 151. C. R. Fisher, K. H. Graves, A. F. Parlow, and E. R. Simpson, Characterization of mice deficient in aromatase (arKO) because of targeted disruption of the cyp19 gene. Proc. Natl. Acad. Sci. USA 95, 6965 – 6970 (1998).
359 152. Y. Weisman, F. Cassoria, S. Malozowski, R. J. J. Krieg, D. Goldray, A. M. Kaye, and D. Somjen, Sex-specific response of bone cells to gonadal steroids: Modulation in perinatally androgenized females and in testicular feminized male rats. Steroids 58, 126 – 133 (1993). 153. T. A. Bateman, J. J. Broz, M. L. Fleet, and S. J. Simske, Differing effects of two-week suspension on male and female mouse bone metabolism. Biomed. Sci. Instrum. 34, 374 – 379 (1997). 154. D. Vanderschueren, E. Van Herck, A. M. H. Suiker, W. J. Visser, L. P. C. Scot, K. Chung, R. S. Lucas, T. A. Einhorn, and R. Bouillon, Bone and mineral metabolism in the androgen-resistant (testicular feminized) male rat. J. Bone Miner. Res. 8, 801 – 809 (1993). 155. D. Vanderschueren, E. Van Herck, P. Geusens, A. Suiker, W. Visser, K. Chung, and R. Bouillon, Androgen resistance and deficiency have difference effects on the growing skeleton of the rat. Calcif. Tissue Int. 55, 198 – 203 (1994). 156. R. Marcus, D. Leary, D. L. Schneider, E. Shane, M. Favus, and C. A. Quigley, The contribution of testosterone to skeletal development and maintenance: Lessons from the androgen insensitivity syndrome. J. Clin. Endocrinol. Metab. 85, 1032 – 1037. 72 (2000).
CHAPTER 12
Coupling of Bone Resorption and Formation during Bone Remodeling THOMAS J. MARTIN GIDEON A. RODAN
I. II. III. IV. V.
St. Vincent’s Institute of Medical Research, Melbourne 3065, Australia Department of Bone Biology and Osteoporosis Research, Merck Sharp & Dohme Research Laboratories, West Point, Pennsylvania 19486
VI. Bone Mass Homeostasis VII. Role of Mechanical Function (Strain) in the Coupling of Bone Resorption to Bone Formation VIII. Integrated View of the Coupling of Bone Resorption and Bone Formation References
Introduction Sequence of Cellular Events in Bone Remodeling Interaction of Osteoblast Lineage Cells with Osteoclasts Similarities between Bone Remodeling and Inflammation Factors Proposed to Mediate the Coupling of Bone Formation to Bone Resorption
I. INTRODUCTION
In order to maintain skeletal balance, which one could name skeletal homeostasis, bone resorption intiates bone formation, which restores the amount of bone removed by resorption. Two main sets of observations pointed to the concept of coupling. Kinetic studies using radiotracers of calcium or strontium to estimate the rates of bone formation and resorption in animals or humans have shown that when bone resorption increases under physiological or pathological conditions, bone formation increases as well [1]. Hyperparathyroidism and estrogen deficiency are examples of conditions in which resorption and formation are increased. Similarly, when bone resorption decreases, e.g., during estrogen replacement therapy, bone formation does too [2]. The second type of evidence is histological. Examination of bone sections showed that osteoclastic bone resorption and osteoblastic bone formation are contiguous during bone remodeling and can be logically conceived to
Bone remodelling refers to the renewal process whereby small pockets of old bone, disposed throughout the skeleton and separated from others anatomically as well as chronologically, are replaced by new bone throughout adult life. Remodeling is essential for the maintenance of normal bone structure and for calcium homeostasis. The process is such that the entire adult human skeleton is replaced on the average in 10 years. The concept that bone formation and resorption are coupled during the bone remodeling process was developed in the 1970s. It is based on the principle that bone resorption occurs in order to release calcium for physiological needs and to reshape the bone structure to equip it better for its mechanical function, and formation restores the bone that is lost.
OSTEOPOROSIS, SECOND EDITION VOLUME 1
361
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
362 follow each other in the “basic multicellular unit” (BMU), which describes a packet of bone being resorbed and rebuilt [3]. The histomorphometric estimation of remodeling (or turnover) rates is based on the assumption that resorption precedes formation [3]. This “coupling” has been amply confirmed and, with some exceptions discussed later, is a general characteristic of bone remodeling. A number of discoveries in the late 1990s have revealed much about the molecular signaling pathways that influence local processes in bone, including osteoblast differentiation and function and the control of osteoclast formation and activity. As yet it is unknown what controls the extent of bone resorption and the extent of bone formation that replaces it and how in particular the two are contrived to be equal. This chapter outlines the cellular aspects of the bone remodeling process, how cells of the osteoblast lineage influence the resorption process, as well as bone formation, and considers current views of possible cellular and molecular mechanisms by which bone formation is coupled to resorption.
II. SEQUENCE OF CELLULAR EVENTS IN BONE REMODELING Cancellous bone remodeling starts on the bone surface, usually covered by lining cells, a single layer of flat cells derived from osteoblasts that have ceased to deposit matrix. These cells are proposed to respond to stimuli that initiate the remodeling cycle, of which there are many candidates both among circulating hormones and among cytokines generated locally, either by stromal/osteoblastic cells or cells of the immune system. The very nature of the remodeling process, occurring as it does in different parts of the skeleton at different times, highlights the importance of locally generated, regulatory factors in the process. When the cycle is initiated, say by parathyroid hormone (PTH) or by mechanical strain, which would generate cytokines or prostanoids [4], it has been proposed that the thin layer of nonmineralized matrix under these cells is initially digested by collagenase to expose the mineralized matrix that osteoclasts can resorb [5 – 7]. PTH is known to stimulate collagenase production and secretion in osteoblastic cells [8], and evidence in strong support of the role of collagenase is provided by a study in genetically mutated mice whose collagenase is unable to digest type I collagen [9]. In these mice the bone-resorbing action of PTH is severely attenuated. According to this model of initiation of resorption, osteoclasts (or their late precursors) near the stage of final maturation and activation must be available close to the required sites. Although this model is likely, there is no direct in vivo evidence for it. Certainly, once the process starts, then new osteoclasts must be generated, and all the bone-resorbing
MARTIN AND RODAN
hormones, cytokines, and prostanoids can promote this process through mechanisms discussed later in this chapter. Through the interaction of osteoblast lineage cells with osteoclasts, the latter start resorbing bone, a process lasting 2 to 4 weeks and carried out by groups of osteoclasts to a depth of about 30 m [10]. Toward the end of resorption, mononuclear cells, proposed to complete the resorption process, are seen at the bottom of resorption pits [11]. These are followed by cells, which may be related to macrophages, which appear along with preosteoblasts on the resorption surface. The transition from resorption to formation is called reversal, and the reversal plane can be identified microscopically with certain stains [12] or polarized light. The reversal line (cement line) contains a large abundance of osteopontin [13], which is produced both by osteoclasts and by osteoblasts. This is an arginine – glycine – aspartic acid (RGD) containing extracellular matrix protein [14], which interacts with vitronectin receptors v3 in osteoclasts and primarily v5 in osteoblasts. These integrin receptors were shown not only to mediate cell attachment to the extracellular matrix, but also to act as signal-transducing receptors [15]. Studies on mice that lack expression of the osteopontin gene indicate that osteopontin plays a role in bone resorption [16]. During maturation, osteoblasts become cuboidal, polarized cells, which are rich in endoplasmic reticulum and contain a large oval nucleus. Osteoblasts are connected to each other and form a contiguous layer. They seem to cooperate in the production of the extracellular bone matrix, as the dimensions of the fibrillar organization of collagen exceed the size of single cells. Moreover, because the organization of collagen is so well suited to withstand the tensile mechanical forces exerted on bone, osteoblasts probably sense and respond to mechanical strain. Osteocytes, which are embedded in bone and connected with each other and with surface cells by canaliculae, are particularly well situated to carry out this function. Another indication that this process may be important is the reduction in the rate of bone formation caused by immobilization [17] and weightlessness [18]. To replace the bone resorbed during a couple of weeks, bone formation continues for several months. Osteoblasts then become gradually flatter and the osteoid surface thins until a very flat layer of lining cells covers a very thin layer of nonmineralized matrix on quiescent bone surfaces. The remodeling of cortical bone follows similar stages, triggered by cues that may start in cells lining the Haversian canals or in osteocytes [19]. Osteoclasts excavate a “cutting cone,” which is refilled by osteoblast activity. Open questions in this sequence of events relate primarily to the signals that govern osteoclast and osteoblast recruitment and termination of osteoclast and osteoblast activity, the identity of cells at the reversal
363
CHAPTER 12 Coupling of Bone Resorption and Formation
phase, and the precise composition of the matrix on the reversal line.
III. INTERACTION OF OSTEOBLAST LINEAGE CELLS WITH OSTEOCLASTS Chambers [20] and Rodan and Martin [21] proposed that the stimulation of osteoclast activity by most agents is mediated by osteoblast lineage cells. This hypothesis was based on the fact that PTH receptors and PTH responses were much better documented for osteoblasts than osteoclasts, prostanoid stimulation of bone resorption correlated with effects on cyclic AMP accumulation in osteoblasts, and PTH and 1,25(OH)2D3 induced pronounced shape changes in osteoblastic cells. Similar changes in lining cells could expose the bone matrix, making it accessible to osteoclasts. Bone resorption studies with isolated osteoclasts supported the indirect modes of action of PTH, interleukin (IL)-1, tumor (TNF)-necrosis factor , 1,25-dihydroxyvitamin D, and the cytokines that use gp 130 as a signal transducer [22 – 25]. In vitro studies of osteoclast formation from bone marrow cells have convicingly demonstrated the requirement for osteoblasts or stromal cells [26]. This, together with the fact that actual contact between these cells and osteoclast precursors is necessary [26], strongly indicated that a molecule expressed on the cell membrane of osteoblast/stromal cells is important in promoting osteoclast formation. This prediction was fulfilled with the discovery of RANK ligand (RANKL), also known as “osteoclast differentiation factor” (ODF), a 316 amino acid, type II transmembrane protein, which is a member of the TNF ligand family [27,28] (see Chapter 3). Produced by osteoblastic stromal cells and activated T cells, RANKL, in the presence of colony stimulating factor1 (M-CSF), but without any accompanying stromal/osteoblastic cells, promotes the formation of osteoclasts from hemopoietic cells. When RANKL (-/-) mice were generated, they were found to be osteopetrotic because of failed osteoclast formation [29]. The action of RANKL is antagonized by osteoprotegerin (OPG), a soluble member of the TNF receptor family, produced by osteoblastic stromal cells as a decoy receptor, which inhibits RANKL action. Overexpression of OPG in transgenic mice results in osteopetrosis [30], and OPG (-/-) mice exhibit severe bone loss through excessive osteoclast formation and bone resorption [31]. The receptor for RANKL on hemopoietic cells is RANK (receptor activator of NF-KappaB), and RANK (-/-) mice are also osteopetrotic [32]. In addition to these effects on osteoclast formation, RANKL is able to activate mature osteoclasts [33] and OPG to inhibit their activity [34]. The formation of RANKL and of OPG in osteoblastic stromal cells is regulated by the hormones and cytokines that influence bone
resorption [35]. The identification in the promoter region of OPG of functional binding sites for the osteoblast master switch transcription factor, Cbfal [36] provides a molecular mechanism for the control of bone homeostasis by the osteoblast lineage. In addition to promoting and maintaining the osteoblast phenotype and thereby favoring bone formation, Cbfal can drive the stromal lineage toward the capacity to inhibit bone resorption by promoting OPG formation (see Chapter 6). These discoveries of the late 1990s have revealed much more of the cellular and molecular processes involved in the generation of resorption sites, and therefore in the bone remodeling process. There is little doubt of the importance in remodeling of the TNF and TNF receptor ligand family members. Their roles, to be considered in the context of hormone and drug actions upon bone, will undoubtedly have applications for new therapeutic approaches.
IV. SIMILARITIES BETWEEN BONE REMODELING AND INFLAMMATION In searching for molecules that link bone resorption to formation, it may be useful to point to striking similarities between bone remodeling and inflammation (Table 1). Inflammation starts with trauma produced by injury or by a foreign body. Bone remodeling starts with a stimulus, sometimes mechanical, which exposes the mineralized bone surface. In inflammation, the foreign body is recognized by white blood cells, e.g., macrophages, which start secreting cytokines and growth factors. The cytokines stimulate the production and migration of other white blood cells to the site of inflammation. Bone exposed to mechanical strain, which probably initiates remodeling, attracts mononuclear cells, which stain positively for nonspecific esterases [37]. Many cytokines involved in inflammation are potent stimulators of osteoclastic bone resorption and osteoclast differentiation in vitro [26] (see Chapter 13). IL-1 and TNF- are among the most powerful stimulators of bone resorption yet identified, inducing TABLE 1 Comparison of Sequence of Events and Cellular Interactions in Bone Remodeling and Inflammation Stage
Inflammation
Bone remodeling
Injury
Tissue damage foreign body
Mechanical?
Reaction
White blood cells, macrophage (local and hematogenous)
Osteoclasts
Repair
Mesenchymal cells (perivascular, fibroblasts)
Osteoblast lineage
Fibrosis, scar formation
Bone formation
364 osteoclast formation by effects on osteoblast/stromal cells to produce RANKL and M-CSF [38] and reduce OPG production. Most importantly a RANKL/RANK-independent pathway of the osteoclastogenic effect of TNF- has been demonstrated [39], through which TNF directly programs bone marrow macrophage precursors to osteoclasts, with their activation dependent on IL-1. The potential importance of this for inflammatory bone disease is evident. Indeed the impact of the immune system on bone cell function has become increasingly apparent. T cells produce many cytokines that have an impact on osteoblast or osteoclast differentiation. Although T cells represent about 2 to 3% of bone marrow cells, they become an abundant population in inflammatory states, e.g., periodontal disease and rheumatoid arthritis. IL-1 and TNF- are predominantly derived from monocytes. Among T-cell-derived cytokines, interferon (IFN)-, granulocyte macrophage colony stimulating factor (GM-CSF), IL-4, and IL-13 function as negative regulators of osteoclastogenesis [40]. IL-17 is a T-cell cytokine that promotes osteoclast formation and bone resorption through a prostaglandin dependent mechanism, similar to that of IL-1 [41]. IL-18, a stromal/osteoblastic product, inhibits osteoclast formation by acting on T cells to promote GM-CSF production [42]. It may be that local imbalances of pro and antiosteoclastogenic cytokines determine whether there is a net loss of bone in inflammatory conditions affecting bone directly. These discoveries of the late 1990s have revealed further cellular and molecular processes involved in the generation of resorption sites, and therefore in the bone remodeling process. One of the main antiosteoporotic effects of estrogen is to inhibit proliferation and differentiation of osteoclast precursors. The precise mechanism of these effects and the cellular targets of estrogen have yet to be fully elucidated. Estrogen receptors are expressed by monocytes, osteoblasts, and osteoclast precursors, as well as osteoclasts. Thus, estrogen could suppress osteoclastogenesis by regulating any one or more of these cell types. Current evidence indicates that production of at least five factors — IL-1, TNF, IL-6, and IL-6 receptor complex, M-CSF and GMCSF — is enhanced in conditions of estrogen deficiency [40]. In view of their pro-osteoclastogenic effects, all of these cytokines are considered potential mediators of the effects of estrogen on bone. A late phase in inflammation is the recruitment of fibroblasts, which produce matrix and encapsulate the foreign body. Fibroblast growth factor (FGF) and other growth factors are involved in this process [43]. The analogous phase in bone remodeling is the recruitment of osteoblasts, which cover the resorption surface with mineralized matrix. PGE, IL-1, transforming growth factor (TGF-), and FGF were all shown to stimulate bone formation in vivo [44 – 47]. An important part of inflammation is neovascularization, probably stimulated by FGF, vascular endothelial growth factor
MARTIN AND RODAN
(VEGF), and other cytokines [48]. The importance of angiogenesis in osteogenesis has long been recognized, and bone-derived cells were shown to produce VEGF in response to PGE [49], and VEGF was shown to promote osteoclast formation in vitro [50]. This analogy suggests that in inflammation, T cells can substitute for osteoblastic stromal cells in promoting osteoclast formation, and evidence has been produced for that [51,52]. It suggests further that during the resorption process or at its termination, factors released by osteoclasts or cells present on the reversal surface, e.g., macrophages, attract the preosteoblasts to that surface. Interestingly, osteopontin, an extracellular molecule made by macrophages and osteoclasts, is found in inflammatory and atherosclerotic lesions [53 – 55], is present on the reversal surface [13], and is chemotactic [14]. Osteopontin is also produced abundantly by macrophages found in tumors [56] and tumors are often encapsulated by fibrous tissue. Osteopontin seems to be one of the molecules that plays a role in bone remodeling [16].
V. FACTORS PROPOSED TO MEDIATE THE COUPLING OF BONE FORMATION TO BONE RESORPTION Baylink and colleagues [57,58] suggested that “coupling” is due to bone formation factors released from the bone matrix during bone resorption. Indeed, a large number of substances that are mitogenic to osteoblasts or stimulate bone formation in vivo could be extracted from bone matrix [59]. These include insulin-like growth factor, (IGF) I and II [60], acidic and basic fibroblast growth factor (FGF) [61], transforming growth factor 1 and 2 [62] and TGF- heterodimers [63], bone morphogenetic proteins (BMPs) 2, 3, 4, 6, and 7 [64 – 66], platelet-derived growth factor (PDGF) [67], and probably others. Several questions should be considered regarding the role of these substances in the coupling of bone formation to bone resorption: (i) which cells produce them and under what circumstances, (ii) do they stimulate bone formation in vivo, (iii) can they be released from the matrix in active from and in controlled amounts during bone resorption, (iv) is there evidence for an increase in the abundance of these substances at sites of bone remodeling, and (v) are there regulatory mechanisms by which they are activated? IGF-I, IGF-II, bFGF, TGF-, and PDGF are produced by rat osteoblastic cells. IGF-I and IGF-II production is enhanced by stimulators of bone formation, such as PGE and PTH [68,69]. Elevated levels of IGF-I mRNA were found in bone from estrogen-deficient rats, where bone turnover is increased [70]. During bone growth in rats, there is a close association between osteogenesis and IGF-I expression [71]. However, following marrow ablation, which causes a
365
CHAPTER 12 Coupling of Bone Resorption and Formation
substantial increase in bone formation, the rise in IGF-I mRNA was seen after the histological appearance of differentiated osteoblasts, suggesting that it did not initiate bone formation in that system [72]. In human bone, the major form of IGF is IGF-II, which was also shown to be produced by human bone cells in culture [57,73]. Bone is one of the most abundant sources of TGF [74]. This growth factor is produced by all osteoblastic cells examined and its production is increased by estrogen and FGF (in osteosarcoma cells) [75,76]. BMPs are members of the TGF- superfamily. BMP-2 and BMP-4 are produced in adult bovine preodontoblasts [77], as well as in human fetal teeth [78]. BMP-7 (OP-1) was localized in human embryos in hypertrophic chondrocytes, osteoblasts, and periosteum, as well as other tissues [65], whereas BMP-3 was found in human embryonic lung and kidney, in addition to perichondrium, periosteum, and osteoblasts [79]. Both bFGF [80] and PDGF [67] were shown to be produced by bone cells or bone explants in culture. These factors could thus be involved in bone remodeling, but the time and site for their synthesis and secretion in vivo have not yet been determined. Prostaglandin E, primarily E2, is another bone cellproduced cytokine, which in vitro is upregulated by mechanical strain [81] and stimulates both bone resorption and formation [44,82]. Further light on mechanisms by which new bone is formed comes from the discovery of Cbfal, an essential transcription factor required for osteoblast differentiation [83,84]. Cbfal not only programs the primitive mesenchymal cell to express osteoblast-specific genes, but has also been shown to be important for maintaining the osteoblast phenotype in mature bone [85]. In remodeling, osteoblasts are recruited from a pool of committed cells and need to sustain the osteoblast phenotype. The regulated expression of Cbfal may be an important aspect of this. Study of its regulation is at an early stage, but it will be important to know whether growth factor effects proceed through the Cbfal pathway in the remodeling process. Several growth factors stimulate bone formation in vivo. IGF-I, injected into humans or rats, increases both bone resorption and bone formation [86], and reports on its effect on the bone balance are inconclusive [87]. When injected together with the IGF-binding protein IGF BP-3 into rats, it was reported to increase bone volume [88]. BMPs injected into bone stimulate bone formation locally and produce a positive bone balance. TGF-, from the same family of proteins, has a similar effect. When injected next to the periosteum or endosteum, there is substantial augmentation in local bone formation in rats and other species [46,89,90]. At the same time, there is an increase in endocortical bone resorption; thus, like IGF-I, TGF- seems to stimulate both resorption and formation; however, the local balance is clearly positive. bFGF, injected both locally and systemically, was also reported to increase bone formation [47,91]. Prostaglandin (PG)E1 and E2 have long been known to be
potent stimulators of bone formation when given either locally [92] or systemically, both to humans or to experimental animals [93,94]. These substances could thus contribute to the bone formation observed in remodeling if secreted or released in active form at the appropriate site and time. It was proposed that TGF-, which is produced as an inactive precursor in bone and bone cells [95,96], is present in the matrix and can be activated by acidification or proteolytic cleavage and is activated by resorbing osteoclasts [97]; TGF- activity was recovered from conditioned media of in vitro-resorbing osteoclasts [98]. It remains to be shown if the other growth factors also survive the proteolytic cleavage of the acidic hydrolases present in the resorption lacunae. Other questions raised by this model of coupling, via growth factor release from the matrix, relate to the time course and the distance between resorption and formation processes and whether activation can be controlled with sufficient precision in this way. Osteoclastic bone resorption proceeds for about 2 – 3 weeks before formation follows and continues for 3 – 4 months. The osteoblast precursors, which should respond to the “coupling factors”, could be many micrometers away from where active osteoclast resorption is in progress. Osteoblastic lineage cells produce TGF- in latent form, and IGFs are complexes bound to a family of specific, high-affinity binding proteins (IGFBPs), which regulate their bioavailability [99]. TGF- may be released from latent complexes at appropriate sites in bone by plasmin generated locally through the action of plasminogen activators in a manner that is controlled temporally and spatially by hormones and cytokines [100]. A similar local control could free IGF-I from association with its inhibitory binding protein [101]. Although there is no obvious skeletal phenotype in mice with inactivated genes for plasminogen activators, in vivo investigation of such possibilities would require treatment of such animals with anabolic agents such as PTH. Another way to explain coupling using an embryological paradigm is by implicating the surface left by osteoclasts, the so-called reversal surface, as an initiating influence. If active growth and/or differentiation factors are contained in this surface, they clearly could play a role by acting on osteoblasts or intermediary cells, which recruit the osteoblasts. Local matrix molecules, such as osteopontin, could also play such a role. Most of all, in vivo evidence is needed to show the presence by immunochemistry and the activity by bioassays, as illustrated for TGF- in vitro, of specific growth factors at bone remodeling sites. The technology for such investigations may become available soon.
VI. BONE MASS HOMEOSTASIS The putative biochemical mediators of bone resorption and bone formation described earlier do not explain a major aspect of “coupling,” namely, what determines the extent of
366 bone resorption and bone formation in each remodeling cycle. There clearly is bone mass homeostasis. All healthy individuals have a bone mineral density or bone mineral content that distributes normally around a mean with a standard deviation of about 10%. Bone mass is clearly genetically controlled [102,103] and there is much interest in possible genes that may be involved [103,104] (see chapter 26). Bone mass or bone mineral content, as measured, for example, in the lumbar spine noninvasively by dual energy X-ray absorptiometry (DXA), is determined by the amount of both cortical and cancellous bone. The amount of cortical bone is determined by periosteal bone formation, which continues throughout life, as well as endosteal and Haversian bone remodeling. Cancellous bone volume is determined by the relative extent of bone resorption and bone formation on the cancellous bone surface. The genetic determinants of bone mass thus should control these processes. Steroid hormones and sex hormones in particular are likely to participate in the genetic determination of bone mass. Men clearly have larger and thicker bones than women. The reduction in bone mass due to estrogen or androgen deficiency is well documented. Moreover, an estrogen receptor-deficient man [105], as well as mice in which the estrogen receptor was “knocked out” [106,107], was reported to have bone defects. In addition, the epiphyses closed very late in that man, suggesting an estrogen role in that function in males, as well as in females. It is not known exactly how steroids control bone formation or bone resorption. Receptors for sex steroids have been detected in osteoblastic cells from various species, including humans, and estrogens were shown to inhibit osteoclast activity in vitro [108]. The effect of estrogens on bone is discussed in detail in chapter 41. The sex steroids could have both direct and indirect effects, acting both on bone resorption and bone formation. Because these are systemic hormones and their concentration is most likly not determined by skeletal function, they do not generate the signals that terminate resorption or formation. They can provide a general background for the cellular responses to such signals. Frost [109] has indeed proposed that estrogen concentrations determine the “set point” for the response of the skeleton to mechanical signals. Mechanical stimuli may be the most direct input in the maintenance of bone mass and thus play a central role in bone mass homeostasis and by extension in the coupling of bone resorption and bone formation.
VII. ROLE OF MECHANICAL FUNCTION (STRAIN) IN THE COUPLING OF BONE RESORPTION TO BONE FORMATION The effect of mechanical forces on bone formation and resorption, mediated by the strain in the matrix, has long
MARTIN AND RODAN
been known and is very well documented. A decrease in mechanical load produced by immobilization or weightlessness causes a reduction in bone mass, which is due both to increased bone resorption, which occurs initially, and to decreased bone formation, which is sustained for a longer duration [17,110]. Eventually the system reaches a new steady state where the available bone mass is probably adequate for the prevailing mechanical load. The effects of weightlessness are a clear illustration of “uncoupling”, bone resorption being increased and bone formation decreased, implicating mechanical load in the “coupling” phenomenon. Examination of trabeculae from human vertebrae by scanning electron microscopy provided a visual illustration of this phenomenon [111]. Trabeculae, which were not mechanically loaded because one of their edges was loose and disconnected, showed very extensive resorption without evidence of bone formation. However, trabeculae, which were connected at both ends and thus mechanically loaded, had shallower resorption lacunae and evidence of bone formation. The increase in bone turnover produced by a reduction in mechanical load, the lower bone formation rate produced by immobilization, and the stabilization of bone mass at a new lower steady state pointed to mechanical strain as a factor that couples bone resorption to bone formation. In trabeculae mechanically weakened by resorption, and possibly cortical bone as well, bone formation would be stimulated until the strain is dissipated. The resulting structure would thus be ideally suited to sustain the prevailing strain. This would explain trabecular architecture, which matches the strain distribution in the bone and would explain the increase in the diameter of long bones to compensate for decreased bone mass observed in mice with osteogenesis imperfecta [112]. It could explain why the gain produced by an inhibitor of bone resorption, such as estrogen or bisphosphonates, eventually levels off, possibly when the existing bone mass has maximized its resistance to the prevailing loads. Consistent with this model is the fact that the potent bone resorption inhibitor alendronate did not cause any changes in bone mass in nonosteopenic minipigs [113]. Furthermore, in osteoporotic patients treated with inhibitors of bone resorption, bone mass continues to increase for some time after the filling of the remodeling spaces and the increment in mechanically loaded cortical bones at the hip, for example, is larger than in less loaded ones, such as the wrist. Experimental studies suggest that relatively limited mechanical input is probably sufficient to maintain the “genetically programmed” skeletal mass [114], that short-term bone loss can be caused by total rest or weightlessness (hypogravity) [115], and that very strenuous exercise, such as professional tennis playing, is necessary to produce exercise dependent significant increases in bone mass [116]. Thus, if we accept the fact that bone mass and bone structure are controlled by mechanical strain and that bone formation is
367
CHAPTER 12 Coupling of Bone Resorption and Formation
proportional to mechanical strain, we have to conclude that mechanical strain is at least one of the factors that couples bone resorption to formation. How this is brought about at biochemical and molecular levels has not been satisfactorily elucidated and remains one of the current challenges of skeletal reseach.
TABLE 2
Bone Remodeling/Skeletal Homeostasis
Skeletal functions Homeostasis of calcium and other ions Mechanical support and levers for muscle action Support of hemopoiesis Participating cells Osteoblasts (mesenchyme-derived cells)
VIII. INTEGRATED VIEW OF THE COUPLING OF BONE RESORPTION AND BONE FORMATION Bone has three major functions: mechanical support, homeostasis of calcium and other ions, and housing of hemopoiesis (see Table 2). Bone remodeling is initiated by stimuli generated to fulfill one of these functions. Mechanical stimuli are clearly local and were shown to be able to initiate remodeling [117]. The change in extracellular matrix strain, perceived by lining cells or osteocytes, likely primes a specific site in bone for remodeling. Local events could include the release of arachidonic acid metabolites, probably prostaglandin E, and other cytokines, plus possibly direct activation of osteoclasts or osteoclast precursors, leading to a local round of bone remodeling. It is not known what determines the extent of resorption and the depth or the size of the resorption lacunae. Bone formation is then initiated through (i) direct stimulation of osteoblast precursors by the initial factors, such as PGE; (ii) release of growth factors from the matrix or from other cells at the resorption site, such as vascular cells or macrophages; (iii) interaction with matrix molecules at the resorption surfaces, such as osteopontin; or (iv) all of the above. Once in progress, bone formation probably continues as long as the bone-forming cells perceive the osteogenic stimulus of mechanical strain. Likely transducers of that strain are integrins, through which cells are anchored in the matrix. Integrins were shown to act as signal transducing receptors [15] and to affect the phosphorylation of intracellular molecules in ways similar to those produced by growth factor receptors [118,119], possibly leading to similar outcomes of gene expression and protein synthesis. Both bone resorption and bone formation, presumbly controlled by the homeostatic inputs of mechanical forces, occur in an endocrine “field.” Thus, factors that suppress osteoclast activity, such as estrogens, would modulate the rate and possibly the extent of the resorptive phase, which would increase in the absence of estrogen or in the presence of stimulators of osteoclast activity, such as interleukins or parathyroid hormone. The same may hold true for bone formation where factors reported to enhance osteoblast activity, such as IGF, androgens, TGF- and BMPs, and others, may augment the rate and possibly the extent of the bone-forming phase. If the kinetic constraints, determined primarily by the rate of
Osteocytes (osteoblast lineage cells) Lining cells (osteoblast lineage cells) Marrow stromal cells (mesenchyme-derived cells) Osteoclasts (hemopoietically derived) B lymphocytes (hemopoietically derived) T lymphocytes (hemopoietically derived) Molecular mediators Major endocrine factors Parathyroid hormone Sex steroids (estrogens and androgens) Calcitonin Glucocorticoids Calcitriol [1,25(OH)2D] Thyroid hormones Paracrine/autocrine factors Insulin-like growth factors (IGFs) and IGF-binding proteins Transforming growth factor family, including bone morphogenetic proteins (BMPs, 2, 4, 6, and others) Fibroblast growth factor family Prostanoids (PGE2 and others) Interleukins (IL-1, -6, -11, -17, and others) Colony-stimulating factors (M-CSF and GM-CSF) Tumor necrosis factors (RANK Ligand, TNF- and others) TNF receptors (osteoprotegerin) Parathyroid hormone-related peptide Matricrine factors Collagen (type I) Osteopontin Fibronectin Vitronectin Thrombospondin Mechanical stimuli
bone resorption, are not rate limiting, the steady-state bone density is most likely determined by the mechanical load. However, if bone resorption proceeds at an excessive pace that becomes rate limiting, such as in estrogen deficiency, bone formation, albeit increased, will not keep pace and bone loss will occur. Once bone resorption is slowed down by estrogen or other therapy, the bone mass can again reach its homeostatic level, determined by mechanical loads and possibly systemic regulators [120]. Thus, the relative
368 effects of various hormones and other factors would be to modulate the resorption or formation arm of the equation, permitting or preventing the maintenance of the homeostatic bone mass and the rate at which it is reached. This is another way of expressing the “set point hypothesis” for the mechanical control of bone mass [109]. For example, not all estrogen-deficient women or hyperthyroid patients with increased bone turnover lose bone to the same extent. This could also explain why exercise may be more effective in maintaining or gaining bone mass in estrogen-replete postmenopausal women. The second stimulus for bone turnover is calcium recruitment from the skeleton, initiated by PTH. Cortical bone seems to be a preferential target for PTH-stimulated bone resorption, possibly a reflection of the distribution of PTH receptors among osteoblast lineage cells [121,122]. However, elevated PTH concentrations may increase the general level of bone resorption wherever it occurs, augmenting bone loss produced by a lack of mechanical function. The beneficial effects of calcium supplements and vitamin D on hip fractures are consistent with this effect, as well as the bone gain observed after parathyroidectomy in vertebral BMD, which also contains a considerable amount of cancellous bone [123]. The third function of the skeleton, housing of the hemopoietic system, probably does not affect bone mass significantly under usual circumstances, but may lie at the basis of the response of the skeleton to lymphokines and other cytokines and explain the bone loss associated with inflammation in periarticular regions and the periodontium. It has been reported that increased red blood cell formation enlarges the marrow cavity [for review, see 124], and malignancies of the bone marrow, such as multiple myeloma, are clearly associated with extensive bone resorption. The feedback mechanisms, which come into play for enlarging the marrow cavity when increased hemopoiesis is needed, are probably mediated by the interleukins, which increase osteoclastogenesis, such as IL-1, IL-6, IL-11, and TNF-. Production of these interleukins during inflammation or in response to local tumors would lead to similar bone destruction. The similarity between the phases of inflammation and bone remodeling was pointed out earlier, but it is not yet known if factors involved in the later steps of inflammation, probably FGF and TGF-, which were shown to stimulate osteoblast proliferation, play a role in bone formation during normal bone remodeling. In addition to the local control described earlier, an intriguing possibility of central control of bone remodeling and homeostasis comes from the discovery that both ob/ob mice (leptin gene mutated to inactivity) and db/db mice (leptin receptor inactive) have greatly increased bone mass despite their hypogonadism and increased circulating glucocorticoid. Strikingly, this phenotype is corrected by intracerebroventricular injection of leptin [120], suggesting that the hypothalamus releases a bone mass regulatory substance.
MARTIN AND RODAN
In conclusion, an integrated view of the bone remodeling process should take into account that bone mass is controlled homeostatically by mechanical function in a hormonal environment (or by hormones in a mechanical field) and that there is a close relationship between bone and hemopoiesis and a similarity between bone remodeling and the cycle of inflammation and tissue repair. Rapidly accumulating new information should test and undoubtedly modify these hypotheses.
References 1. W. H. Harris and R. P. Heaney, Skeletal renewal and metabolic bone disease. N. Engl. J. Med. 280, 193 – 202 (1969). 2. K. M. Prestwood, C. C. Pilbeam, J. A. Burleson, F. N. Woodiel, P. D. Delmas, L. J. Deftos, and L. G. Raisz, The short term effects of conjugated estrogen on bone turnover in older women. J. Clin. Endocrinol. Metab. 79, 366 – 371 (1994). 3. H. M. Frost, Dynamics of bone remodeling. In “Bone Biodynamics,” pp. 315 – 333. Littel, Brown, Boston, 1964. 4. G. A. Rodan, Perspectives: Mechanical loading, estrogen deficiency, and the coupling of bone formation to bone resorption. J. Bone Miner. Res. 6, 527 – 530 (1991). 5. S. Sakamoto and M. Sakamoto, Bone collagenase, osteoblasts and cell-mediated bone resorption. In “Bone and Mineral Research,” pp. 49 – 102. Elsevier, Amsterdam, 1986. 6. T. J. Chambers, J. A. Darby, and K. Fuller, Mammalian collagenase predisposes bone surfaces to osteoclastic resorption. Cell Tissue Res. 241, 671 – 675 (1985). 7. T. J. Chambers and K. Fuller, Bone cells predispose bone surfaces to resorption by exposure of mineral to osteoclastic contact. J. Cell Sci. 6, 155 – 165 (1985). 8. N. C. Partridge, J. J. Jeffrey, L. S. Ehlich, S. L. Teitelbaum, C. Fliszar, H. G. Welgus, and A. J. Kahn, Hormonal regulation of the production of collagenase and a collagenase inhibitor activity by rat osteogenic sarcoma cells. Endocrinology 120, 1956 – 1962 (1987). 9. W. Zhao, M. H. Byrne, B. F. Boyce, and S. M. Krane, Bone resorption induced by parathyroid hormone is strikingly diminished in collagenaseresistant mutant mice. J. Clin. Invest. 103, 517 – 524 (1999). 10. N. J. Garrahan, P. I. Croucher, and J. E. Compston, A computerized technique for the quantitative assessment of resorption cavities in trabecular bone. Bone 11, 241 – 246 (1990). 11. E. F. Eriksen, Normal and pathological remodeling of human trabecular bone: Three dimensional reconstruction of the remodeling sequence in normals and in metabolic disease. Endocr. Rev. 4, 379 – 408 (1986). 12. A. R. Villanueva, C. Sypitkowski, and A. M. Parfitt, A new method for identification of cement lines in undecalcified, plastic embedded sections of bone. Stain Technol. 61, 83 – 88 (1986). 13. J. Chen, M. D. McKee, A. Nanci, and J. Sodek, Bone sialoprotein mRNA expression and ultrastructural localization in fetal porcine calvarial bone: Comparisons with osteopontin. Histochem. J. 26, 67 – 78 (1994). 14. D. T. Denhardt and X. Guo, Osteopontin: A protein with diverse functions. FASEB J. 7, 1475 – 1482 (1993). 15. R. Hynes, Integrins: Versatility, modulation, and signaling in cell adhesion. Cell 69, 11 – 25 (1992). 16. H. Yoshitake, S. R. Rittling, D. T. Denhardt, and M. Noda, Osteopontin-deficient mice are resistant to ovariectomy-induced bone resorption. Proc. Natl. Acad. Sci. USA 96, 10944 (1999).
CHAPTER 12 Coupling of Bone Resorption and Formation 17. M. Weinreb, G. A. Rodan, and D. D. Thompson, Osteopenia in the immobilized rat hind limb is associated with increased bone resorption and decreased bone formation. Bone 10, 187 – 194 (1989). 18. P. Backup, K. Westerlind, S. Harris, T. Spelsberg, B. Kline, and R. Turner, Spaceflight results in reduced mRNA levels for tissuespecific proteins in the musculoskeletal system. Am. J. Physiol. 266, E567 – E573 (1994). 19. T. M. Skerry, L. Bitensky, J. Chayen, and L. E. Lanyon, Early strainrelated changes in enzyme activity in osteocytes following bone loading in vivo. J. Bone Miner. Res. 4, 783 – 788 (1989). 20. T. J. Chambers, The cellular basis for bone resorption. Clin. Orthop. Relat. Res. 151, 283 – 293 (1980). 21. G. A. Rodan and T. J. Martin, Role of osteoblasts in hormonal control of bone resorption: A hypothesis. Calcif. Tissue Int. 33, 349 – 351 (1981). 22. P. M. J. McSheehy and T. J. Chambers, Osteoblastic cells mediate osteoclastic responsiveness to parathyroid hormone. Endocrinology 118, 824 – 828 (1986). 23. B. M. Thomson, J. Saklatvala and T. J. Chambers, Osteoblasts mediate interleukin-1 stimulation of bone resorption by rat osteoclasts. J. Exp. Med. 164, 104 – 112 (1986). 24. H. M. Perry, W. Skogen, J. Chappel, A. J. Khan, G. Wilner and S. L. Teitelbaum, Partial characterization of a parathyroid hormone stimulated resorption factor(s) from osteoblast-like cells. Endocrinology 125, 2075 – 2082 (1989). 25. P. M. J. McSheehy and T. J. Chambers, 1,25-Dihydroxyvitamin D3 stimulates rat osteoblastic cells to release a soluble factor that increases osteoclastic bone resorption. J. Clin. Invest. 80, 425 – 429 (1987). 26. T. Suda, N. Takahashi, and T. J. Martin, Modulation of osteoclast differentiation. Endoc. Rev. 13, 66 – 80 (1992). 27. D. L. Lacey, E. Timms, H. L. Tan, M. J. Kelley, C. R. Dunstan, T. Burgess, R. Elliott, A. Colombero, G. Elliott, S. Scully, et al., Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell 93, 165 – 176 (1998). 28. H. Yasuda, N. Shima, N. Nakagawa, S. I. Mochizuki, K. Yano, N. Fujise, Y. Sato, M. Goto, K. Yamaguchi, M. Kuriyama, et al., Identity of osteoclastogenesis inhibitory factor (OCIF) and osteoprotegerin (OPG): A mechanism by which OPG/OCIF inhibits osteoclastogenesis in vitro. Endocrinology 139, 1329 – 1337 (1998). 29. Y. Y. Kong, H. Yoshida, I. Sarosi, H. L. Tan, E. Timms, C. Capparelli, S. Morony, A. J. Oliveira-dos-Santos, G. Van, A. Itie, et al., OPGL is a key regulator of osteoclastogenesis, lymphocyte development and lymph-node organogenesis. Nature 397, 315 – 323 (1999). 30. W. S. Simonet, D. L. Lacey, C. R. Dunstan, M. Kelley, M. S. Chang, R. Luthy, H. Q. Nguyen, S. Wooden, L. Bennett, T. Boone, et al., Osteoprotegerin: A novel secreted protein involved in the regulation of bone density. Cell 89, 309 – 319 (1997). 31. N. Bucay, I. Sarosi, C. R. Dunstan, S. Morony, I. Tarpley, C. Capparelli, S. Scully, H. L. Tan, W. Xu, D. L. Lacey, W. I. Boyle, and W. S. Simonet, Osteoprotegerin-deficient mice develop early onset osteoporosis and arterial calcification. Genes Dev. 12, 1260 – 1268 (1998). 32. W. C. Dougall, M. Glaccum, K. Charrier, K. Rohirbach, K. Brasel, et al., RANK is essential for osteoclast and lymph node development. Genes Dev. 13, 2412 – 2424 (1999). 33. T. L. Burgess, Y. Qian, S. Kautman, B. D. Ring, G. Van, C. Capparelli, M. Kelley, H. Hsu, W. I. Boyle, C. R. Dunstan, S. Hu, and D. L. Lacey, The ligand for osteoprotegerin OPGL directly activates mature osteoclasts. J. Cell. Biol. 145, 527 – 538 (1999). 34. B. S. Kwon, S. Wang, N. Udagawa, V. Haridas, Z. H. Lee, K. K. Kim, K. O. Oh, J. Greene, Y. Li, J. Su, R. Gentsz, B. B. Aggarwal, and J. Ni, TR1, a new member of the tumor necrosis factor receptor superfamily, induces fibroblast proliferation and inhibits osteoclastogenesis and bone resorption. FASEB J. 12, 845 – 854 (1998).
369 35. N. J. Horwood, J. Elliott, T. J. Martin, and M. T. Gillespie, Osteotropic agents regulate the expression of osteoclast differentiation factor and osteoprotegerin in osteoblastic stromal cells. Endocrinology 139, 4743 – 4746 (1998). 36. K. Thirunavukkarasu, D. L. Halladay, R. R. Miles, Yang, R. I. S. Galvin, S. Chandrasekhar, T. S. Martin, and J. E. Onyia, The osteoblast-specific transcription factor Cbfal regulates the expression of osteoprotegerin (OPG), a potent inhibitor of osteoclast differentiation and function. J. Biol. Chem. in press. 37. R. L. Baron, L. Neff, P. Tran Van, J. R. Nefussi, and A. Vignery, Kinetic and cytochemical identification of osteoclast precursors and their differentiation into multinucleated osteoclasts. Am. J. Pathol. 122, 363 – 378 (1986). 38. T. Suda, N. Takahashi, N. Udagawa, E. Jimi, M. T. Gillespie, and T. J. Martin, Modulation of osteoclast differentiation and function by the new members of the human necrosis factor receptor and ligand families. Endocrin. Rev. 20, 345 – 357 (1999). 39. K. Kobayashi, N. Takahashi, E. Jimi, N. Udagawa, M. Takami, S. Kotake, N. Nakagawa, M. Kinosaki, K. Yamaguchi, N. Shima, H. Yasuda, T. Morinaga, K. Higashio, T. J. Martin, and T. Suda, Tumor necrosis factor alpha stimulates osteoclast differentiation by a mechanism independent of ODF/RANKL-RANK intereaction. J. Exp. Med. 191, 275 – 286. 40. T. J. Martin, E. Romas, and M. T. Gillespie, Interleukins in the control of osteoclast differentiation. Crit. Rev. Euk. Gene Expr. 8, 107 – 123 (1998). 41. S. Kotake, N. Udagawa, N. Takahashi, K. Matsuzaki, K. Itoh, S. Ishiyama, S. Saito, K. Inoue, N. Kamateni, M. T. Gillespie, T. J. Martin, and T. Suda, IL-17 in synovial fluids from patients with rheumatoid arthritis is a potent stimulator of osteoclastogenesis. J. Clin. Invest. 103, 1345 – 1352 (1999). 42. N. Horwood, N. Udagawa, J. Elliott, D. Grail, H. Okamura, M. Kurimoto, A. Dunn, T. J. Martin, and M. T. Gillespie, IL-18 inhibits osteoclast formation via T-cell production of GM-CSF. J. Clin. Invest. 103, 1345 – 1352 (1999). 43. D. Gospodarowicz, Fibroblast growth factor. Clin. Orthop. Relat. Res. 257, 231 – 248 (1990). 44. W. S. S. Jee, H. Z. Ke, and X. J. Li, Long-term anabolic effects of prostaglandin-E2 on tibial diaphyseal bone in male rats. Bone Miner. 15, 33 – 55 (1991). 45. B. F. Boyce, T. B. Aufdemorte, I. R. Garrett, A. J. P. Yates, and G. R. Mundy, Effects of interluekin-1 on bone turnover in normal mice. Endocrinology 125, 1142 – 1150 (1989). 46. M. Noda and J. J. Camilliere, In vivo stimulation of bone formation by transforming growth factor. Endocrinology 124, 2991 – 2994 (1989). 47. T. Nakamura, K. Hanada, M. Tamura, T. Shibanushi, H. Nigi, M. Tagawa, S. Fukumoto, and T. Matsumoto, Stimulation of endosteal bone formation by systemic injections of recombinant basic fibroblast growth factor in rats. Endocrinology 136, 1276 – 1284 (1995). 48. J. Folkman, Angiogenesis in cancer, vascular rheumatoid and other disease. Nature Med. 1, 27 – 31 (1995). 49. S. Harada, J. A. Nagy, K. A. Sullivan, K. A. Thomas, N. Endo, G. A. Rodan, and S. B. Rodan, Induction of vascular endothelial growth factor expression by prostaglandin E2 and E1 in osteoblasts. J. Clin. Invest. 93, 2490 – 2496 (1994). 50. S. Niida, M. Kaku, H. Amano, H. Yoshida, H. Kataoka, S. Nishikawa, K. Tanne, N. Maeda, S. I. Nishikawa, and H. Kodama, Vascular endothelial growth factor can substitute for macrophage colony-stimulating factor in the support of osteoclastic bone resorption. J. Exp. Med. 190, 293 – 298 (1999). 51. N. J. Horwood, V. Kartsogiannis, J. M. W. Quinn, E. Romas, T. J. Martin, and M. T. Gillespie, Activated T lymphocytes support osteoclast formation in vitro. Biochem. Biophys. Res. Commun. 265, 144 – 150 (1999).
370 52. Y. Y. Kong, U. Feige, U. Sarosi, B. Bolon, A. Tafuri, S. Morony, et al., Activated T cells regulate bone loss and joint destruction in adjuvant arthritis through osteoprotegerin ligand. Nature 402, 304 – 309 (1999). 53. C. M. Giachelli, N. Bae, M. Almeida, D. T. Denhardt, C. E. Alpers, and S. M. Schwartz, Osteopontin is elevated during neointima formation in rat arteries and is a novel component of human atherosclerotic plaques. J. Clin. Invest. 92, 1686 – 1696 (1993). 54. E. R. OBrien, M. R. Garvin, D. K. Stewart, T. Hinohara, J. B. Simpson, S. M. Schwatz, and C. M. Giachelli, Osteopontin is synthesized by macrophage, smooth muscle, and endothelial cells in primary and restenotic human coronary atherosclerotic plaques. Arterioscler. Thromb. 14, 1648 – 1656 (1994). 55. T. Ikeda, T. Shirasawa, Y. Esaki, S. Yoshiki, and K. Hirokawa, Osteopontin mRNA is expressed by smooth muscle-derived foam cells in human atherosclerotic lesions of the aorta. J. Clin. Invest. 92, 2814 – 2820 (1993). 56. D. R. Senger, C. A. Peruzzi, C. F. Gracey, A. Papadopoulos, and D. G. Tenen, Secreted phosphoproteins associated with neoplastic transformation: Close homology with plasma proteins cleaved during blood coagulation. Cancer Res. 48, 5770 – 5774 (1988). 57. S. Mohan, J. C. Jennings, T. A. Linkhart, and D. J. Baylink, Primary structure of human skeletal growth factor: Homology with human insulin-like growth factor-II. Biochim. Biophys. Acta 966, 44 – 55 (1988). 58. S. Mohan and D. J. Baylink, The role of IGF-II in the coupling of bone formation to resorption. In “Modern Concepts of Insulin-like Growth Factors,” pp. 169 – 184. Elsevier, New York, 1991. 59. P. V. Hauschka, A. E. Mavrakos, M. D. Iafrati, S. E. Doleman, and M. Klagsburn, Growth factors in bone matrix: Isolation of multiple types by affinity chromatography on heparin sepharose. J. Biol. Chem. 261, 12665 – 12674 (1986). 60. C. Schmid and M. Ernst, Insulin-like growth factors. In “Cytokines and Bone Metabolism,” pp. 229 – 265. CRC Press, Boca Raton, FL, 1992. 61. S. B. Rodan and G. A. Rodan, Fibroblast growth factor and plateletdrived growth factor. In “Cytokines and Bone Metabolism,” pp. 115 – 145. CRC Press, Boca Raton, FL, 1992. 62. M. Centrella, T. L. McCarthy, and E. Canalis, Transforming growth factor-beta and remodeling in bone. J. Bone Joint Surg. 73-A, 1418 – 1428 (1991). 63. Y. Ogawa, D. K. Schmidt, J. R. Daschy, R. J. Chang, and C. B. Glaser, Purification and characterization of transforming growth factor beta2.3 and beta1.2 heterodimers from bovine bone. J. Biol. Chem. 267, 2325 – 2328 (1992). 64. S. E. Harris, M. Sabatini, M. A. Harris, J. Q. Feng, J. Wozney, and G. R. Mundy, Expression of bone morphogenetic protein messenger RNA in prolonged cultures of fetal rat calvarial cells. J. Bone Miner. Res. 9, 389 – 394 (1994). 65. S. Vukicevic, V. Latin, P. Chen, R. Batorsky, A. H. Reddi, and K. Sampath, Localization of osteogenic protein-1 (bone morphogenetic protein-7) during human embryonic development: High affinity to basement membranes. Biochem. Biophys. Res. Commun. 198, 693 – 700 (1994). 66. S. Vukicevic, V. M. Paralkar, N. S. Cunningham, J. S. Gutkind, and A. H. Reddi, Autoradiographic localization of osteogenin binding sites in cartilage and bone during rat embryonic development. Dev. Biol. 140, 209 – 214 (1990). 67. S. Rydziel, C. Ladd, T. L. McCarthy, M. Centrella, and E. Canalis, Determination and expression of platelet-derived growth factorAA in bone cell cultures. Endocrinology 130, 1916 – 1922 (1992). 68. T. L. McCarthy, M. Centrella, and E. Canalis, Parathyroid hormone enhances the transcript and polypeptide levels of insulin-like growth factor I in osteoblast-enriched cultures from fetal rat bone. Endocrinology 124, 1247 – 1253 (1989).
MARTIN AND RODAN 69. T. L. McCarthy, M. Centrella, L. G. Raisz, and E. Canalis, Prostaglandin E2 stimulates insulin-like growth factor I synthesis in osteoblast-enriched cultures from fetal rat bone. Endocrinology 128, 2895 – 2900 (1991). 70. R. T. Turner, P. H. Backup, P. H. Sherman, E. Hill, G. L. Evans, and T. C. Spelsberg, Mechanism of action of estrogen on intramembranous bone formation and regulation of osteoblast differentiation and activity. Endocrinology 131, 883 – 889 (1992). 71. D. M. Shinar, N. Endo, D. Halperin, G. A. Rodan, and M. Weinreb, Differential expression of insulin-like growth factor-I (IGF-I) and IGF-II messenger ribonucleic acid in growing rat bone. Endocrinology 132, 1158 – 1167 (1993). 72. L. J. Suva, J. G. Seedor, N. Endo, H. A. Quartuccio, D. D. Thompson, I. Bab, and G. A. Rodan, Pattern of gene expression following rat tibial marrow ablation. J. Bone Miner. Res. 8, 379 – 388 (1993). 73. S. Mohan and D. J. Baylink, Evidence that the inhibition of TE85 human bone cell proliferation by agents which stimulate cAMP production may in pat be mediated by changes in the IGF-II regulatory system. Growth Regul. 1, 110 – 118 (1991). 74. M. Centrella, M. C. Horowitz, J. M. Wozney, and T. L. McCarthy, Transforming growth factor-beta gene family members and bone. Endocr. Rev. 15, 27 – 39 (1994). 75. B. S. Komm, C. M. Terpening, D. J. Benz, K. A. Graeme, A. Gallegos, M. Korc, G. L. Greene, B. W. OMalley, and M. R. Haussler, Estrogen binding, receptor mRNA, and biologic response in osteoblast-like osteosarcoma cells. Science 241, 81 – 84 (1988). 76. M. Noda and R. Vogel, Fibroblast growth factor enhances type 1 transforming growth factor gene expression in osteoblast-like cells. J. Cell Biol. 109, 2529 – 2535 (1989). 77. M. Nakashima, H. Nagasawa, Y. Yamada, and A. H. Reddi, Regulatory role of transforming growth factor-beta, bone morphogenetic protein-2, and protein-4 on gene expression of extracellular matrix proteins and differentiation of dental pulp cells. Dev. Biol. 162, 18 – 28 (1994). 78. K. Heikinheimo, Stage-specific expression of decapentaplegic-Vgrelated genes 2,4 and 6 (bone morphogenetic proteins 2,4 and 6) during human tooth morphogenesis. J. Dent. Res. 73, 590 – 597 (1994). 79. S. Vukicevic, M. N. Helder, and F. P. Luyten, Developing human lung and kidney are major sites for synthesis of bone morphogenetic protein-3 (osteogenin). J. Histochem. Cytochem. 42, 869 – 875 (1994). 80. R. K. Globus, J. Plouet, and D. Gospodarowicz, Cultured bovine bone cells synthesize basic fibroblast growth factor and store it in their extracellular matrix. Endocrinology 124, 1539 – 1547 (1989). 81. D. Somjen, I. Binderman, E. Berger, and A. Harell, Bone remodelling induced by physical stress is prostaglandin E2 mediated. Biochim. Biophys. Acta 627, 91 – 100 (1980). 82. L. G. Raisz and T. J. Martin, Prostaglandins in bone and mineral metabolism. In “Bone and Mineral Research” (W. A. Pecked), pp. 286 – 310. Elsevier, Amsterdam, 1984. 83. P. Ducy, R. Zhan, V. Geoffroy, A. L. Ridall, and G. Karsenty, Osf2/Cbfal: A transcriptional activator of osteoblast differentiation. Cell 89, 747 – 754 (1997). 84. T. Komori, H. Yagi, S. Nomura, A. Yamaguchi, K. Sasaki, K. Deguchi, Y. Shimizu, R. T. Bronson, Y. H. Gao, M. Inada, M. Sato, R. Okamoto, Y. Kitamura, S. Yoshiki, and T. Kishimoto, Targeted disrutpion of Cbfal results in a complete lack of bone formation owing to maturational arrest of osteoblasts. Cell 89, 755 – 764 (1997). 85. P. Ducy, M. Starbuck, M. Priemel, J. Shen, G. Pinero, V. Geoffroy, M. Amling, and G. Karsenty, A Cbfal-dependent genetic pathway controls bone formation beyond embryonic development. Genes Dev. 13, 1025 – 1036 (1994). 86. E. M. Spencer, C. C. Liu, E. C. C. Si, and G. A. Howard, In vivo actions of insulin-like growth factor (IGF-I) on bone formation and resorption in rats. Bone 12, 21 – 26 (1991).
371
CHAPTER 12 Coupling of Bone Resorption and Formation 87. J. H. Tobias, J. W. M. Chow, and T. J. Chambers, Opposite effects of insulin-like growth factor-I on the formation of trabecular and cortical bone in adult female rats. Endocrinology 131, 2387–2392 (1992). 88. C. M. Bagi, R. Brommage, L. Deleon, S. Adams, D. Rosen, and A. Sommer, Benefit of systemically administered rhIGF-I/IGFBP-3 on cancellous bone in ovariectomized rats. J. Bone Miner. Res. 9, 1301 – 1312 (1994). 89. M. A. Critchlow, Y. S. Bland, and D. E. Ashhurst, The effects of age on the response of rabbit periosteal osteoprogenitor cells to exogeneous transforming growth factor-beta2. J. Cell Sci. 107, 499 – 516 (1994). 90. M. E. Joyce, A. B. Roberts, M. B. Sporn, and M. E. Bolander, Transforming growth factor and the initiation of chondrogenesis and osteogenesis in the rat femur. J. Cell Biol. 110, 2195 – 2207 (1990). 91. H. Kawaguchi, T. Kurokawa, K. Hanada, Y. Hiyama, M. Tamura, E. Ogata, and T. Matsumoto, Stimulation of fracture repair by recombinant human basic fibroblast growth factor in normal and streptozotocin-diabetic rats. Endocrinology 135, 774 – 781 (1994). 92. S. C. Marks, Jr. and S. C. Miller Local infusion of prostaglandin- E1 stimulates mandibular bone formation in vivo. J. Oral Pathol. 17, 500 – 505 (1988). 93. J. E. Lund, The toxicology of PGE1 and PGI2. In “Prostaglandins in Clinical Medicine: Cardiovascular and Thrombotic Disorders,” pp. 93 – 109. Yearbook Medical Publishers, New York, 1982. 94. W. S. S. Jee, S. Mori, X. J. Li, and S. Chan, Prostaglandin E2 enhances cortical bone mass and activates intracortical bone remodeling in intact and ovariectomized female rats. Bone 11, 253 – 266 (1990). 95. S. L. Dallas, S. Park-Snyder, K. Miyazono, D. Twardwik, G. R. Mundy, and L. F. Bonewald, Characterization and autoregulation of laten transforming growth factor beta (TGFbeta) complexes in osteoblast-like cell lines. J. Biol. Chem. 269, 6815 – 6822 (1994). 96. L. F. Bonewald, L. Wakefield, R. O. C. Oreffo, A. Escobedo, D. R. Twardzik, and G. R. Mundy, Latent forms of transforming growth factor beta (TGFbeta) derived from bone cultures: Identification of a naturally occurring 100-kDa complex with similarity to recombinant TGFbeta. Mol. Endocrinol. 5, 741 – 745 (1991). 97. R. O. C. Oreffo, G. R. Mundy, S. M. Seyedin, and L. F. Bonewald, Activation of the bone-derived latent TGF beta complex by isolated osteoclasts. Biochem. Biophys. Res. Commun. 158, 817 – 823 (1989). 98. J. Pfeilschifter and G. R. Mundy, Modulation of type beta transforming growth factor activity in bone cultures by osteotropic hormones. Proc. Natl. Acad. Sci. USA 84, 2024 – 2028 (1987). 99. L. E. Underwood, J. P. Thissen, S. Lemozy, J. M. Ketelslegers, and D. R. Clemmons, Hormonal and nutritional regulation if IGF-I and its binding proteins. Horm. Res. 42, 145 – 151 (1994). 100. T. J. Martin, E. H. Allan, and S. Fukumoto, The plasminogen activator and inhibitor system in bone remodelling. Growth Regul. 3, 209 – 214 (1993). 101. P. G. Campbell, J. F. Novak, T. B. Yanosick, and J. H. McMaster, Involvement of the plasmin system in dissociation of the insulin-like growth factor-binding protein complex. Endocrinology 130, 1401 – 1412 (1992). 102. J. C. Christian, C. Slemenda, and C. C. Johnston, (1989). Heritability of bone mass: A longitudinal study in aging male twins. Am. J. Hum. Genet. 44, 429 – 433 (1989). 103. N. A. Morrison, J. Cheng Qi, A. Tokita, P. J. Kelly, L. Crofts, T. V. Nguyen, P. N. Sambrook, and J. A. Eisman, Prediction of bone density from vitamin D receptor alleles. Nature 367, 284 – 287 (1994). 104. L. D. Spotila, C. D. Constantinou, L. Sereda, A. Ganguly, B. L. Riggs, and D. J. Prockop, Mutation in a gene for type I procollagen (COL1A2) in a woman with post-menopausal osteoporosis: Evidence for phenotypic and genotypic overlap with mild osteogenesis imperfecta. Proc. Natl. Acad. Sci. USA 88, 5423 – 5427 (1991). 105. E. P. Smith, J. Boyd, G. R. Frank, H. Takahashi, R. M. Cohen, B. Specker, T. C. Williams, D. B. Lubahn, and K. S. Korach, Estrogen
106.
107. 108.
109.
110.
111.
112.
113.
114. 115.
116.
117. 118.
119.
120.
121.
122.
123.
124.
resistance caused by a mutation in the estrogen-receptor gene in a man. N. Eng. J. Med. 331, 1056 – 1061 (1994). D. B. Lubahn, J. S. Moyer, T. S. Golding, J. F. Couse, K. S. Korach, and O. Smithies, Alteration of reproductive function but not prenatal sequal development after insertional disruption of the mouse estrogen receptor gene. Proc. Natl. Acad. Sci. USA 90, 11162 – 11166 (1993). K. S. Korach, Insights from the study of animals lacking functional estrogen receptor. Science 266, 1524 – 1527 (1994). M. J. Oursler, L. Pederson, L. Fitzpatrick, B. L. Riggs, and T. Spelsberg, Human giant cell tumors of the bone (osteoclastomas) are estrogen target cells. Proc. Natl. Acad. Sci. USA 91, 5227 – 5231 (1994). H. M. Frost, The mechanostat: A proposed pathogenic mechanism of osteoporosis and the bone mass effects of mechanical and nonmechanical agents. Bone 2, 73 – 85 (1987). D. T. Thompson and G. A. Rodan, Indomethacin inhibition of tenotomy-induced bone resorption in rats. J. Bone Miner Res. 3, 409 – 414 (1988). L. Mosekilde, Consequences of the remodeling process for vertebral trabecular bone structure: A scanning electron microscopy study (uncoupling of unloaded structures). Bone Miner. 10, 13 – 35 (1990). J. Bonadio, K. J. Jepsen, M. K. Mansoura, R. Jaenisch, J. L. Kuhn, and S. A. Goldstein, A murine skeletal adaptation that significantly increases cortical bone mechanical properties. J. Clin. Invest. 92, 1697 – 1705 (1993). M.-H. Lafage, R. Balena, M. A. Battle, M. Shea, J. G. Seedor, H. Klein, W. C. Hayes, and G. A. Rodan, Comparison of alendronate and sodium fluoride effects on cancellous and cortical bone in minipigs. J. Clin. Invest. 95, 2127 – 2133 (1995). C. T. Rubin and L. E. Lanyon, Regulation of bone mass by mechanical strain magnitude. Calcif. Tissue Int. 37, 411 – 417 (1985). V. Schneider, V. Ogancy, A. LeBlanc, A. Rakhmanov, A. Bakulin, A. Grigoriev, and L. VaRonin, Spaceflight bone loss and changes in fat and lean body mass. J. Bone Miner. Res. 7, S122 (1992). A. L. Huddleston, D. Rockwell, D. N. Kulund, and R. B. Harrison, Bone mass in lifetime tennis athletes. J. Am. Med. Assoc. 244, 1107 – 1109 (1980). S. Morii and D. B. Burr, Increased intracortical remodeling following fatigue damage. Bone 14, 103 – 109 (1993). L. J. Kornberg, S. H. Earp, C. E. Turner, C. Prockop, and R. L. Juliano, Signal transduction by integrins: Increased protein tyrosine phosphorylation caused by clustering betal integrins. Proc. Natl. Acad. Sci. USA 88, 8392 – 8396 (1991). J. L. Guan, J. E. Trevithick, and R. O. Fibronectin/integrin interaction induces tyrosine phosphorylation of a 120-kDa protein. Cell Regul. 2, 951 – 954 (1991). P. Ducy, M. Amling, S. Takeda, M. Priemel, A. F. Schilling, F. T. Beil, J. H. Shen, C. Vinson, J. M. Rueger, and G. Karsenty, Leptin inhibits bone formation through a hypothalamic relay: A central control of bone mass. Cell 100, 197 – 207 (2000). M. F. Rouleau, J. Mitchell, and D. Goltzman, In vivo distribution of parathyroid hormone receptors in bone: Evidence that a predominant osseous target cell is not the mature osteoblast. Endocrinology 123, 187 – 191 (1988). M. F. Rouleau, J. Mitchell, and D. Goltzman, Characterization of the major parathyroid hormone target cell in the endosteal metaphysis of rat long bones. J. Bone Miner. Res. 5, 1043 – 1053 (1990). S. J. Silverberg, F. Gartenberg, D. McMahon, and J. P. Bilezikian, Parathyroidectomy improves bone density at cancellous sites in asymptomatic primary hyperparathyroidism. J. Bone Miner. Res. 8, S169 (1993). D. Shinar and G. A. Rodan, Relationships and interactions between bone and bone marrow. In “The Hematopoietic Microenvironment: The Functional and Structural Basis of Blood Cell Development” (M. W. Long and M. S. Wicha, (eds.), pp. 79 – 109. Johns Hopkins Univ. Press, Baltimore, MD, 1993.
CHAPTER 13
Cytokines and Bone Remodeling GREGORY R. MUNDY, BABATUNDE OYAJOBI, KATHY TRAIANEDES, SARAH DALLAS, AND DI CHEN Department of Medicine and Endocrinology, University of Texas Health Science Center, San Antonio, Texas 78284
VIII. Interleukin (IL)-1 and IL-1 Signaling in Osteoclast-like Cells IX. Tumor Necrosis Factor X. Interleukin-6 XI. Vascular Endothelial Growth Factor XII. IL-15, IL-17, and IL-18 XIII. Archidonic Acid Metabolites: Prostaglandins and Leukotrienes XIV. Transforming Growth Factor- XV. Bone Morphogenetic Proteins XVI. Conclusion References
I. Introduction II. Evidence for a Role of Cytokines in Osteoclastic Bone Resorption III. The Osteoclast as a Cell Source of Cytokines Involved in Osteoclastic Resorption IV. The Osteoblast as a Cell Source of Cytokines Involved in Osteoclastic Resorption V. RANK Ligand and Its Signaling Receptor, RANK VI. Osteoprotegerin VII. Macrophage – Colony-Stimulating Factor and Its Signaling Receptor, c-fms
I. INTRODUCTION
recent advances. A list of the cytokines reviewed in this chapter is provided in Table 1. Recent advances in molecular biological techniques have meant that most of the biological activities ascribed to cytokines have now been associated with specific molecules, and their receptors identified and molecularly cloned. Several of these cytokines and their cognate receptors have been shown to be expressed by bone cells, marrow cells, or accessory cells in the bone microenvironment. Moreover, studies using knockout and transgenic mice have increased our understanding of the complex signal transduction mechanisms utilized by cytokines and are opening up new and exciting areas of study. Cytokines tend to be pleiotropic and multifactorial and may have overlapping and seemingly redundant biological effects. Some of this redundancy is
In normal individuals, bone is continuously being remodeled and this is achieved via a finely regulated balance between the processes of bone formation and resorption mediated by osteoblasts and osteoclasts, respectively. This bone remodeling is regulated, in part, by local factors including cytokines generated in the bone microenvironment. The purpose of this chapter is to summarize what is currently known about the role of cytokines and their receptors in bone remodeling. In the past few years, there has been an explosion of information on multiple aspects of the effects of cytokines on bone. This has become an enormous topic, and it will not be possible to cover all aspects in this chapter. Rather, it will focus on important
OSTEOPOROSIS, SECOND EDITION VOLUME 1
373
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
374 TABLE 1 Cytokines Produced in Bone Microenvironment with Major Effects on Osteoclasts and Osteoblasts Osteoclastogenic cytokines RANK ligand Osteoprotegrin Macrophage – colony stimulating factor Interleukin-1 Tumor necrosis factor Interleukin-6 Vascular endothelial growth factor Interleukin-15, -16, and -17 Prostaglandins and leukotrienes Osteoblastogenic cytokines Transforming growth factor- Bone morphogenetic proteins
apparent in the receptor mechanisms and signal transduction pathways used by groups of cytokines. Classic examples that illustrate this vividly are the various cytokines belonging to the interleukin (IL)-6 family, such as IL-6, leukemia inhibitory factor (LIF), oncostatin-M, and IL-11, which utilize a common signal transduction protein known as gp130. These cytokines bind to distinct membrane-associated receptors, which form hetero- or homodimers upon binding to the ligand. These dimers then complex with gp 130, leading to its activation by the phosphorylation of tyrosine residues. This subsequently activates several tyrosine kinase cascades within the cells by a common tyrosine kinase known as JAK2. One of these cascades involves phosphorylation of a transcription factor known as STAT-2. Another involves ras and MAP-2 kinase and leads to phosphorylation of the transcription factor, nuclear factor (NF) – IL-6 [1]. These signal transduction pathways and those used by other cytokines are only now being studied in bone cells, but observations already made in other cells and tissues are holding true for bone with just a few exceptions. The reasons that individual members of cytokine families have seemingly distinct effects on cells involved in bone remodeling remain unclear.
II. EVIDENCE FOR A ROLE OF CYTOKINES IN OSTEOCLASTIC BONE RESORPTION A considerable amount of data has been accumulate since the mild-1970s that indicate that cytokines play a role in both physiological bone remodeling. As mentioned previously, osteoclast formation and activity are regulated by factors that are generated in the bone microenvironment acting in an autocrine, paracrine, or juxtacrine fashion.
GREGORY R. MUNDY ET AL.
These include macrophage – colony-stimulating factor (M-CSF), also known as colony-stimulating factor-1 (CSF1), IL-6, IL-1, IL-11, tumor necrosis factor (TNF)-, TNF, granulocyte – macrophage colony-stimulating factor (GM-CSF), transforming growth factor (TGF)-, TGF-, leukemia inhibitory factor (LIF), and bone morphogenetic proteins (BMPs). For some of these cytokines, the precise cellular source within the bone microenvironment has not been defined, although possibilities include immune cells and bone cells of either osteoblastic or osteoclastic lineages. In addition, biologically active forms of some of these cytokines may be derived from sequestered stores within bone matrix. However, the relative importance of cytokines in the formation of new osteoclasts and the activation of mature osteoclasts in vivo is still unclear. The availability of transgenic animal technology has meant that the role of cytokines in bone metabolism is now increasingly being examined in an in vivo context. However, the functional redundancies among related cytokines or groups of cytokines that share common signaling pathways mean that the direct ablation of individual genes for these factors does not always impact bone remodeling adversely. With the exception of M-CSF, as exemplified by the op/op variant of osteopetrosis, naturally occurring models of total deficiency of any of the factors mentioned previously are rare. In the op/op mouse model of osteopetrosis, a frameshift mutation in the coding region of the csf-1 gene leads to the failure of secretion of biologically active M-CSF by stromal cells, osteoblasts, or other accessory cells. Consequently, mature macrophages do not survive for long and osteoclasts fail to form during the neonatal period, resulting in inadequately remodeled bone. However, osteoclasts do form beyond the neonatal period with sufficient function to reverse the osteopetrosis by 22 weeks, indicating that MCSF is not required for osteoclast formation beyond the first few weeks of life. These data show that in the mouse, secretion of biologically active M-CSF is an absolute requirement for normal osteoclast formation during this early period of life. The impairment in osteoclastic resorption and osteopetrosis can be rescued by the exogenous administration of M-CSF during the neonatal period [2 – 4]. Moreover, studies have also shown that the exogenous administration of vascular endothelial growth factor (VEGF) to neonatal op/op mutant also reverse osteopetrosis, suggesting that this factor may be responsible, in part, for the spontaneous improvement observed in affected op/op mice as they mature [5]. Interestingly, GM-CSF and IL-3, which are other major growth factors for cells of the monocyte/ macrophage lineage, can also partially reverse the osteopetrosis in these mutant mice [6], implying that other factors are essential for normal bone remodeling in this form of osteopetrosis. Consistent with this hypothesis, enforced expression of bcl-2 in cells of the monocyte/macrophage lineage resulting in their prolonged survival partially rescues
375
CHAPTER 13 Cytokines and Bone Remodeling
the osteoclast defect in csf-1 mutant op/op mice [7]. In summary, studies with the mutant M-CSF op/op mouse provide clear evidence that cytokines are involved in physiological bone resorption. There are other lines of evidence from studies using transgenic mice and inhibitors of cytokine activites that indicate that cytokines such as IL-1, IL-6, and TNF are important in disorders of bone remodeling in vivo. However, available data are complex to interpret and there are often conflicting reports. IL-1 and TNF have consistently been shown to play a major role in the rapid bone loss associated with estrogen-depleted states such as postmenopause and after ovariectomy [8], and there is also evidence from studies with mice lacking the IL-1 type 1 receptor that imply that IL-1 may be an important mediator of the effects of ovariectomy on bone mass. Mice deficient in the type I IL-1 receptor (IL-1R1), which is the signaling receptor for both IL-1 and IL-1, do not lose bone after ovariectomy [9]. The soluble p75 TNF receptor blocks the osteoclastogenic effect of TNF [10], and mice engineered to overexpress this soluble TNF receptor do not lose bone after ovariectomy [11]. Regarding IL-6, the increase in bone resorption observed following ovariectomy in mice can be corrected by the administration of neutralizing antibodies to IL-6, as shown by the experiments of Jilka and colleagues [12]. The same abnormality can be reversed by treatment of mice with estrogen. Also, IL-6 knockout mice are protected against bone loss induced by ovariectomy, further implicating IL-6 in bone remodeling in vivo [13]. Surprisingly, transgenic mice overexpressing IL-6 do not have osteopenia as would be predicted [14]. These seemingly discrepant observations are probably related to the fact that in the presence of estrogen deficiency, there is likely increased production of several cytokines by cells in the bone marrow microenvironment leading to increased osteoclastic activity. Whether production of each cytokine is largely by a particular cell type has yet to be defined. There is considerable evidence that IL-6 and/or TNF may synergize with IL-1 to enhance osteoclastic bone resorption. For example, Pacifici and colleagues have suggested that the simultaneous block of IL-1 and TNF may be necessary to completely abrogate the rapid bone loss seen in the early postovariectomy period [15,16]. However, neutralization of either of these factors may nevertheless lead to some decrease in bone resorption in certain situations. There is also evidence that other osteotropic cytokines may be involved in other disease states. Much of this evidence comes from “gain of function” rather than “loss of function” experiments, with evidence that there is increased production of certain cytokines associated with increased bone loss. This is certainly true in myeloma, in some solid tumors, and in chronic inflammatory diseases associated with a local increase in bone loss such as rheumatoid arthritis and periodontal diseases [17]. It is also true in Paget’s
disease where there is increased production of proresorptive cytokines by multinucleated osteoclasts, especially IL-6 [18 – 20]. Because overproduction of these cytokines in these conditions may enhance bone resorption through the stimulation of osteoclast formation and differentiation, pathologic bone lesions associated with a large increase in osteoclasts may be self-perpetuating.
III. THE OSTEOCLAST AS A CELL SOURCE OF CYTOKINES INVOLVED IN OSTEOCLASTIC RESORPTION Abundant evidence indicates that osteoclast formation and activity are regulated by factors generated in the bone cell microenvironment. As mentioned in the preceding section, these factors may be produced by immune cells or cells in the osteoblast lineage or be derived from the bone matrix itself. However, convincing data support the notion that the osteoclast itself may also be a source of autocrine or paracrine factors, which can modulate bone remodeling. The subject of osteoclast as a secretory cell has been reviewed comprehensively elsewhere [21 – 24] (see also Chapter 3). The osteoclast expresses IL-6 in prodigious amounts. Moreover, IL-6, at least in human systems, can stimulate the formation of cells with osteoclast characteristics [18]. Antibodies to IL-6 inhibit bone resorption by isolated human giant cells on calcified matrices, and, similarly, antisense oligonucleotides to IL-6 inhibit the capacity of human giant cells to form resorption pits on sperm whale dentine [25]. Furthermore, it appears that IL-6 may mediate some of the effects of IL-1 and TNF on bone resorption, as an anti-IL-6 neutralizing antibody and a potent IL-6 antagonist that binds to IL6 receptor but does no dimerize with gp130 both blocked IL-1 and TNF-induced osteoclast formation in human marrow cultures [26]. However, IL-6 is not the only cytokine that is produced by isolated osteoclasts. TGF-, interleukin1, annexin-II (lipocortin-II), and human stem cell antigen I [24,27,28] have all been shown to be expressed by osteoclasts, and each of these factors may regulate osteoclasts function. TGF- inhibits osteoclast formation [29,30] and is a powerful stimulator of osteoclast apoptosis [31]. These effects are probably mediated, in part, via paracrine mechanisms involving alterations in the stromal/osteoblastic cell expression of the receptor activator of NF-B ligand (RANKL) and osteoprotegerin (OPG) [32,33]. TGF- may also generate prostaglandins in the microenvironment of osteoclasts [30], which can exert independent effects on osteoclast formation and activity, most probably via modulating RANKL expression [34,35]. IL-1 and annexin-II both stimulate osteoclast formation, whereas human stem cell antigen I inhibits osteoclast formation. The relative importance of all these osteoclast products in the formation of new osteoclasts is not clear. However, one possibility is that
376 as osteoclasts undergo apoptosis within the bone remodeling unit, at the conclusion of the remodeling sequence, some of these cytokines may be released by the dying osteoclast to produce a new generation of osteoclasts derived from their marrow precursors.
IV. THE OSTEOBLAST AS A CELL SOURCE OF CYTOKINES INVOLVED IN OSTEOCLASTIC RESORPTION There is compelling evidence from ex vivo studies that the commitment of osteoclast progenitors (spleen, bone marrow, or peripheral blood derived) to differentiate to multinucleated cells with characteristics of mature osteoclasts requires direct cell – cell contact with osteoblastic or related marrow stromal cells [36]. Furthermore, it has been known for some time that almost all of the known bone-resorbing cytokines, such as IL-1, IL-6 and IL-11, as well as the systemic bone-resorbing hormones, such as parathyroid hormone (PTH), PTHrelated protein (PTHrP), 1,25-dihydroxyvitamin D3, and PGE2, appear to exert their effect only in the presence of stromal/osteoblastic cells [36]. Because these agents activate different signal transduction pathways on osteogenic cells, it was recognized that there is a convergence in their downstream response, and the existence of a membrane-associated factor on the surface of cells of the osteoblastic lineage essential for osteoclast progenitors to proliferate and differentiate was therefore proposed. This factor, which was variously termed “osteoclast differentiation factor” (ODF) and stromal cell-derived osteoclast formation activity (SOFA), was postulated to be inducible by cytokines and hormones known to regulate osteoclast differentiation. Although M-CSF was known to be membrane associated and to be important for osteoclastic bone resorption, recombinant M-CSF alone could not induce osteoclast formation in the absence of stromal/osteoblastic cells. Anderson and colleagues [37] reported the molecular cloning of a novel membrane-bound member of the TNF receptor (TNFR) family from a cDNA library established from human bone marrow-derived myeloid dendritic cells. Simultaneously, they reported the cloning of the mouse orthologue of the receptor from a fetal mouse liver cDNA library. This receptor, which activated (NF-B) activity, was designated receptor activator of NF-B (RANK) and a search for its cognate ligand led to the cloning of RANK ligand (RANKL). RANKL was shown to be identical to TNF--related activation-induced cytokine (TRANCE), a TNF ligand family member cloned from murine thymoma EL40.5 cells and shown to activate c jun-Nterminal kinase [38]. Subsequently, using a novel secreted TNFR homologue known as osteoprotegerin (OPG)/osteoclastogenesis-inhibitory factor (OCIF) as a probe, two groups independently reported the cloning of the same molecule, which they designated OPG ligand (OPGL) and osteoclast
GREGORY R. MUNDY ET AL.
differentiation factor (ODF), respectively [39,40]. As will be discussed later, RANKL/TRANCE/OPGL/ODF have now been shown to be the same molecule whose expression is obligatory for osteoclastic resorption and normal bone modeling and remodeling. As proposed by Suda and others, we will refer to this cytokine hereafter as RANKL [41]. This subject is also discussed in Chapters 3 and 12.
V. RANK LIGAND AND ITS SIGNALING RECEPTOR, RANK RANKL is synthesized as a type II integral membrane protein with its N terminus in the cytoplasm and a C terminus extending extracellularly. Expression of RANKL in a human embryonic kidney fibroblast (293) cell line was reported to generate a membrane-bound as well as a secreted form of RANKL representing the extracellular C-terminal domain [40]. It has also been reported that the extracellular domain of RANKL can be cleaved by a TNF-convertase (TACE)-like enzyme in vitro and that recombinant RANKL can be cleaved by purified TACE [42]. This cleaved form retains some biological activity in osteoclast assay systems [42]. Although it has been reported that T cells shed RANKL on activation in vitro [43], there is as yet no evidence that a soluble form of RANKL exists in vivo or is generated by proteolytic cleavage in the bone microenvironment. Nevertheless, it remains a possibility that the extracellular domain of RANKL is shed by tumor-associated metalloproteinases as has been described for other members of the TNF ligand family, such as TNF and Fas. In this regard, our group identified a factor from a human tumor associated with osteoclastosis and hypercalcemia that appears to be a novel cytokine that stimulates osteoclast formation in the presence of M-CSF [44], and another group also identified another factor from a mouse tumor with similar biological activity [45]. RANKL, in the presence of M-CSF, induces osteoclast formation in all model systems presently available to study osteoclast development. For example, RANKL stimulated the formation of osteoclasts from spleen-derived osteoclast progenitors in the absence of osteoblasts/stromal cells and this was abolished completely by simultaneously adding OPG [46] or a recombinant soluble form of the extracellular domain of RANK fused to the Fc region of human immunoglobulin (RANK.Fc) [47]. In the presence of M-CSF, RANKL also stimulated osteoclast formation in human and murine bone marrow cultures and also in human peripheral blood monocyte cultures [48 – 50], and it induced the formation of TRAP-positive colonies in an agar culture of bone marrow cells [40]. Treatment of stromal/osteoblastic cells of human and murine origins with known stimulators of osteoclast formation, 1,25(OH)2D3, PTH, PGE2, IL-11,
CHAPTER 13 Cytokines and Bone Remodeling
IL-6, IL-1, and TNF, induces or enhances RANKL messenger RNA levels [34,35,40,46,51 – 54]. Treatment of 45Ca-labeled fetal mouse or rat long bones with a recombinant soluble form of RANKL also stimulated the release of 45Ca from bone, which was completely inhibited by simultaneously adding OPG or RANK.Fc [46,47]. Like OPG, polyclonal antibodies against RANKL inhibited bone resorption in organ cultures induced not only by soluble RANKL but also by 1,25(OH)2D3, PTH, PGE2, and IL-1 [51,55] These results clearly indicate that bone resorption induced by these osteotropic factors is mediated by RANKL. In culture systems where there is essentially no continuing osteoclast formation, such as isolated rat osteoclasts, recombinant RANKL has been shown to induce actin rings in the cells rapidly and to increase their bone-resorbing activity markedly, independent of stromal/osteoblastic cells [55,56]. Administration of recombinant OPG to mice injected with saline or with RANKL daily for 7 days led to a rapid disappearance of osteoclasts (as early as 6 h) from bone surfaces with evidence of osteoclast apoptosis [57]. A single parenteral administration of recombinant RANKL increased blood ionized calcium within 1 h in mice [56]. Also, systemic injection of RANKL twice daily for 3 days led to a sustained hypercalcemia, althouth the number of osteoclasts was almost identical to those of untreated mice [40]. Taken together, these data suggest that RANKL stimulates not only osteoclast differentiation but also activates mature osteoclasts as well as prolongs their survival, therby having a direct impact on their function. RANKL knockout mice have been generated that exhibit typical osteopetrosis with total occlusion of bone marrow space within endosteal bone in the early neonatal period. The bones of these RANKL null mutant mice lack osteoclasts, although osteoclast progenitors were present that were shown to differentiate into functionally active osteoclasts when cocultured with normal osteoblasts/stromal cells from wild-type litter mates [58]. Although these results suggest that RANKL is an absolute requirement for osteoclast development, it remains unknown whether there is a spontaneous reversal of the osteopetrotic phenotype with age in RANKL (-/-) mice as seen in adult csf-1 null mutant mice as there are presently no data on older RANKL null mutant mice. It is likely that the increased expression of RANKL may play a role in pathological situations associated with bone destruction such as malignancies. We have reported that direct cell – cell contact between myeloma cells and marrow stroma-derived ST2 cells enhances RANKL expression in the myeloma and also stimulates production of a soluble factor(s) capable of enhancing RANKL expression in marrow stromal cells [59]. We postulated that increased RANKL expression within the bone microenvironment cells may explain the increased osteoclastic activity and destructive bone lesions that are characteristic of multiple myeloma. This is likely to be true for other tumors metastatic to bone such as
377 breast cancer. In this regard, it has also been demonstrated that although RANKL expression was undetectable in either mouse breast cancer cells or bone marrow stromal cells, its expression was elevated markedly in cocultures of both cell types [60]. It has also been reported that substantial numbers of multinucleated cells with osteoclastic characteristic form in coculture of activated human CD4 T helper cells and adherent murine splenic osteoclast precursors in the presence of M-CSF, independent of stromal/osteoblastic cells [61]. This suggests that in chronic inflammatory tissues characterized by CD4 T-cell infiltration such as rheumatoid arthrititic synovium, this mechanism may be responsible for the extensive localized bone destruction. As mentioned earlier, RANKL was originally cloned as a ligand for the receptor, RANK. RANK is a type I transmembrane protein with a C-terminal cytoplasmic tail much longer than that of all known members of the TNFR superfamily. Like other members of the family, RANK has four extracellular cysteine-rich domains. However, unlike most other TNFR family members, RANK messenger RNA is expressed ubiquitously with highest levels in skeletal muscle and thymus and in spleen- and bone-marrow-derived osteoclast precusors [37,63]. To date, RANK has been shown to bind only to RANKL; it does not bind other members of the TNF ligand family, such as lymphotoxin, TNF, Fas ligand, CD27 ligand, CD30 ligand, CD40 ligand, 4-1BB ligand, or TRAIL. It has also been demonstrated conclusively that the formation of mature osteoclasts from osteoclasts precursors as well as activation of mature osteoclasts can only be induced via RANK signaling [55,62 – 64]. As with RANKL knockout mice, RANK null mutant mice also exhibit severe osteopetrotic phenotype with a complete absence of osteoclasts [63]. RANK also activates c-jun N-terminal kinase (JNK) in immune cells such as T cells, dendritic cells, and spleen-derived hematopoietic osteoclast progenitors [55,62]. RANK also activates NF-B and JNK in a macrophage cell line RAW 264.7, which has been shown to differentiate into osteoclast-like cells when treated with RANKL and M-CSF [62]. However, it remains unclear whether RANK also activates the JNK pathway in marrow-derived osteoclast precursors. However, there is no unequivocal evidence to date that the JNK pathway is at all involved in osteoclast formation and activation. Interestingly, overexpression of RANK in human embryonic kidney fibroblast 293 cells induces ligand-independent NF-B and JNK activation, suggesting that pathological conditions associated with RANK expression may result in increased osteoclast formation independent of RANKL. As RANK has no intrinsic kinase activity, it activates NF-B via interactions with the TNF receptor-associated family (TRAFs) of adaptor molecules [65]. Several members of the TRAF family have been implicated in regulating signals from various TNF/TNFR family members. Evidence shows that TRAF2, TRAF5,
378 and TRAF6 interact with the C-terminal 85 amino acid cytoplasmic tail of RANK, and it is likely that the signals through RANK are mediated primarily through these TRAFs [63,66 – 70]. Of these three TRAFs, TRAF6 appears unique in several respects. First it interacts with a novel Cterminal domain of the cytoplasmic tail of RANK distinct from the known binding motifs for TRAF1, TRAF2, TRAF3, and TRAF5, although TRAF6 also associates with a short N-terminal sequence within the cytoplasmic domain [67,68]. Second, overexpression of an N-terminal truncated TRAF6, acting as a dominant negative, inhibited RANKLinduced NF-B activation in the human embryonic kidney 293 cell line. Third, unlike other TRAFs, TRAF6 has also been implicated in IL-1-induced NF-B activation [71]. Finally, whereas other TRAF null mutant mice currently available, such as TRAF2 null mutant mice, have a normal skeletal phenotype, TRAF6 knockout mice exhibit severe osteopetrosis with defective bone remodeling and delayed tooth eruptions [69]. However, unlike in RANKL (-/-) mice, the bones of TRAF6 (-/-) mice had a few osteoclasts, suggesting that there might be some redundancy in TRAF usage in osteoclast development. A few years ago, two groups independently generated mice that were lacking both p50 and p52 subunits of NFb [72,73] and reported that these double knockout (nf-b-1 and nf-B-2) mice developed severe osteopetrosis because of a defect in osteoclast differentiation. There was a complete absence of osteoclasts, although there were osteoclast progenitors and the number of macrophages was increased. Surprisingly, the osteopetrotic phenotype could be rescued by bone marrow transplant from wild-type littermates. Recent developments in our understanding of the molecular mechanisms of the intracellular signal transduction pathways induced when RANKL binds to its cognate receptor RANK have now clarified observations made with double NF-B null mutant mice. Although overexpression of RANK in human embryonic kidney 293 cells stimulated JNK and NF-B, when the C-terminal cytoplasmic tail necessary for TRAF binding was deleted, the truncated RANK receptor was still capable of stimulating JNK activity but not NF-B. This suggests that interaction with TRAFs is critical for NF-B activation but not for the activation of the JNK pathway. Taken together with the complete absence of osteoclasts in NF-B double knockout mice, the fact that the JNK pathway remains intact even in the absence of NF-B activation indicates that it is most unlikely that the JNK pathway is involved in osteoclast development. We have reported that a genetically engineered form of RANK generated by fusing the entire extracellular domain to the Fc region of human lgG1 (RANK.Fc) blocks hypercalcemia induced by PTHrP-secreting human tumor xenografts in nude mice. This further confirms the critical role of RANKL/RANK interaction in osteoclastic bone
GREGORY R. MUNDY ET AL.
resorption [47]. Transgenic mice overexpressing a soluble RANK.Fc fusion protein have severe osteopetrosis because of a marked reduction in osteoclast numbers and a decrease in bone resorption indices [62]. Figure 1 represents a diagram showing the interactions among RANK, RANK ligand, and OPG.
VI. OSTEOPROTEGERIN Our understanding of the biology of bone modeling and remodeling was given an impetus with the discovery of a novel secreted member of the TNFR superfamily. One group isolated a heparin-binding protein from conditioned media of human fibroblast cultures, which profoundly inhibited osteoclast formation. This protein was thus designated “osteoclastogenesis inhibitory factor” (OCIF) [74]. Independently, another group cloned a novel TNFR family member that constitutively lacked a transmembrane domain and was thus secreted. When expressed, the recombinant protein was shown to inhibit both physiological and pathological bone resorption, and hepatic overexpression of the gene in transgenic mice resulted in severe osteopetrosis. The receptor was therefore termed osteoprotegerin [75]. Subsequent molecular cloning of the cDNA coding for OCIF revealed that it was identical to OPG [39]. Other groups also independently cloned the same receptor molecule and the TNF receptorlike molecule 1 (TR1) and follicular dendritic cell-derived receptor I (FDCR-1) [76 – 78] have each been shown to have complete sequence identity to OPG/OCIF. As proposed by Suda et al. [41], we will hereafter refer to the protein (including OCIF, FDCR-1, and TR1) as OPG. Like other members of the TNF receptor family, OPG has four cysteine-rich domains (DI – D4). In addition, there are two homologous death domain regions (D5 and D6) in OPG. Both D5 and D6 share structural features with other death domains previously described in other members of the TNFR family, including the TNF receptor p55, Fas, DR3, and TRAIL receptor. These death domains have been shown to mediate apoptotic signals. Although the precise role of D5 and D6 of OPG is still not known, the death domain-homologous regions are active in mediating apoptotic signals [79]. OPG has only two known ligands, RANKL and TRAIL, both of which are type II membrane-bound TNF homologues [40,80]. In contrast, OPG circulates in vivo in measurable quantities in both human and rodent sera [81,82], and there is now incontrovertible evidence that it acts as a nonsignaling decoy receptor for RANKL and thereby as a regulator of bone turnover [41] (also see Chapter 12). Serum concentrations of OPG increase with age in both men and women and were significantly higher in postmenopausal osteoporotic women compared to age-matched
379
CHAPTER 13 Cytokines and Bone Remodeling
FIGURE 1
Interactions among RANK ligand, its receptor RANK, and the decoy receptor/in-
hibitor OPG.
controls. It was suggested that the increased levels of OPG in the former group reflect a compensatory response to the enhanced bone resorption in the postmenopausal period rather than a cause of the osteoporosis [82]. This is the only study to determine serum OPG levels in humans and it remains to be seen whether these early data are reproduced in different cohorts of patients. The role of OPG in normal bone remodeling has been highlighted in more detail in studies with OPG-deficient mice produced by targeted disruption of the gene [83,84]. OPG (-/-) mice are viable and fertile, but they exhibit profound osteoporosis from birth caused by enhanced osteoclast formation and formation and function as well as prolonged osteoclast survival. Histological analyses showed a destruction of growth plates and a lack of trabeculae, and histomorphometrical analyses revealed an increase in bone resorption indices in long bones of adult OPG null mutant mice. This is accompanied by a marked decrease in the strength and mineral density of their bones. Interestinly, the osteoblast surface area was also increased in OPG-deficient mice. OPG (-/-) mice also develop calcification of the aorta and renal arteries. These results indicate that OPG is a physiological regulator of osteoclast-mediated bone resorption during postnatal bone growth. It also suggests that OPG may play a role in preventing the calcification of larger arteries.
In the presence of M-CSF, RANKL induced osteoclast formation from spleen cells in the absence of osteoblasts/ stromal cells, and this was abolished completely by simultaneously adding OPG. OPG also strongly inhibits osteoclast formation induced by a range of osteotropic agents, including 1,25(OH)2D3, PTH, PGE2, IL-1, and IL-11, in cocultures of osteoblasts/stromal cells and hemapoietic osteoclast progenitors. Interestingly, in contrast to their stimulatory effects on RANKL mRNA expression, PGE2, 1,25(OH)2D3, and dexamethasone strongly inhibit OPG mRNA expression, suggesting that the regulation of OPG levels is also critical for osteoclastogenesis induced by known osteotropic factors [39,40,85–88]. This has led some workers to postulate that downregulation of OPG may be one of the mechanisms involved in glucocorticoid-induced osteoporosis [86]. OPG also directly inhibits the bone-resorbing activity of isolated mature osteoclasts [89]. As mentioned earlier, treatment of 45Ca-radiolabeled fetal mouse long bones with a soluble form of RANKL also stimulated the release of 45Ca from the bone tissues, which was completely inhibited by the simultaneous addition of OPG [46]. This effect of OPG to inhibit bone resorption is due, in part, to its ability to suppress osteoclast survival [90]. In contrast, OPG gene expression and production in marrow stromal/osteoblastic cells are markedly upregulated by TGF-, [32,33], which likely explains the powerful effect of TGF- to inhibit osteoclast formation [30] and enhance
380 osteoclast apoptosis [31]. In vivo, parenteral administration of OPG results in a marked increase in bone mineral density and bone volume associated with a decrease of active osteoclast number in normal and ovariectomized rats [91]. Serum calcium concentration was also decreased rapidly by the parenteral administration of OPG, independent of any changes in urinary calcium excretion, in thyroparathyroidectomized rats whose serum calcium levels were raised acutely by the administration of PTH [92]. This suggests that OPG, in addition to its effect on osteoclastogenesis, also affects the function and/or survival of mature osteoclasts. OPG also decreased serum calcium levels in tumor-bearing nude mice [91,93], suggesting that it has therapeutic potential for the treatment of hypercalcemic conditions, such as those associated with malignancy.
VII. MACROPHAGE – COLONYSTIMULATING FACTOR AND ITS SIGNALING RECEPTOR, c-fms One of the cytokines that has been clearly shown to play an important role in bone resoption is M-CSF. M-CSF on its own does not stimulate osteoclastic bone resorption in organ culture assays. However, it has been known since the mid-1980s that it is capable of stimulating the formation of cells with osteoclast characteristics in long-term human marrow cultures, as well as in murine marrow cultures [94 – 97]. Like IL-1, M-CSF induces the fusion of preosteoclasts [98,99] and prolongs survival of the multinucleated osteoclast-like cells [100]. However, unlike IL-1, M-CSF does not augment their pit-forming capacity when seeded on calcified matrices [55,99 – 101]. M-CSF has also been implicated in the bone disease osteopetrosis [2,3,102]. As discussed earlier, studies on the murine op/op model of osteopetrosis have clearly shown that M-CSF is required for normal osteoclastogenesis and bone remodeling in the mouse, at least up until the late neonatal period. Studies have also suggested a role for M-CSF in adult bone remodeling. Pacifici and colleagues have provided evidence that IL-1 and TNF concentrations are increased in vivo in estrogen-deficiency states. There is also a substantial body of evidence to indicate that the production of both secreted and cell surface forms of M-CSF by bone marrow stromal cells (BMSC) and osteoblastic cells is regulated by osteotropic cytokines, including IL-1 and TNF [103 –106]. Furthermore, the increased ability of BMSC to support osteoclast formation in the estrogen-deficient state is via IL-1 and TNF-mediated stimulation of M-CSF production [107]. Finally, it has been demonstrated that estrogen blocks M-CSF production by BMSC by directly inhibiting its gene expression [108]. M-CSF mediates its effects on osteoclastic bone resorption through a receptor tyrosine kinase, the protooncogene known as c-fms. Presumably, the presence of this receptor
GREGORY R. MUNDY ET AL.
tyrosine kinase on osteoclast precursors is responsible for M-CSF mediating its effects on osteoclast formation. There may be a hierarchy of receptor tyrosine kinases involved in normal and pathological bone resorption. Other stimulators of bone resorption that mediate their effects on osteoclasts and receptor tyrosine kinases include epidermal growth factor (EGF), TGF-, and platelet-derived growth factor (PDGF). The EGF receptor itself is a receptor tyrosine kinase. PDGF mediates its effects on osteoclast formation presumably through a receptor tyrosine kinase. Because activation of these different receptor tyrosine kinases leads to osteoclast formation and osteoclastic bone resorption, they likely play an important role in bone resorption. However, these receptor tyrosine kinases may not be the only tyrosine kinases involved in bone resorption. Mice deficient in expression of the nonreceptor tyrosine kinase c-src also develop osteopetrosis with failure of osteoclastic bone resorption [109]. However, in these mice, the defect differs from that which occurs in mice with op/op osteopetrosis. In c-src-deficient osteopetrotic mice, there is a failure of ruffled border formation and polarization of the osteoclasts [110]. Nevertheless, osteoclastic form normally. It appears that this receptor tyrosine kinase may, among other things, be involved in osteoclast polarization, which is required for normal osteoclastic bone resorption. Thus, there may be a hierarchy of tyrosine kinases involved in normal osteoclastic bone resorption. Interestingly, c-src has been implicated in signaling by M-CSF. Treatment of normal isolated osteoclasts with M-CSF results in increased osteoclast size and cytoplasmic spreading [111 – 113], which is associated with increased src kinase activity [111].
VIII. INTERLEUKIN (IL)-1 AND IL-1 SIGNALING IN OSTEOCLAST-LIKE CELLS IL-1 is the osteotropic cytokine about which the most is known, at least as far as its effects on bone are concerned. Both IL-1 and IL-1 are powerful stimulators of osteoclastic bone resorption in vitro [114 – 116] and in vivo [116 – 118]. The effects of both Il-1 and IL-1 on bone appear to be identical. IL-1 affects cells at all stages in the osteoclast lineage. This has been demonstrated both in vitro and in vivo [119 – 120]. In vitro studies using cultures of human and murine marrow mononuclear cells that contain committed osteoclasts show that IL-1 increases the formation of multinucleated osteoclasts. When IL-1 is infused or injected in vivo, there is increased appearance not only of mature multinucleated cells, but also of marrow mononuclear cells. This result is indicative of not only an increase in more mature cells in the osteoclast lineage, but also in
CHAPTER 13 Cytokines and Bone Remodeling
granulocyte – macrophage colony-forming units representative of early osteoclast precursors (CFU-GM) [121], an effect probably mediated by IL-1-induced IL-6 [122]. The major effects of IL-1 are probably on the earlier steps in the osteoclast lineage. IL-1 exerts, at least part of, its in vitro and in vivo effects by stimulating prostaglandin synthesis [117]. When IL-1 is injected into the subcutaneous tissue overlying the calvariae, there is accumulation of chronic inflammatory cells associated with an increase in osteoclast activity. This increase can be partially inhibited by the administration of indomethacin, a prostaglandin synthesis inhibitor [123]. IL-1 exerts systemic as well as local actions. Infusions of IL-1 by an osmotic minipump show evidence of increased osteoclastic bone resorption at distant sites [116]. This is associated with a progressive increase in blood-ionized calcium concentrations. IL-1 also has complex effects on osteoblasts [124,125]. These probably depend on whether the interleukin is administered intermittently or continuously. If administered continuously, IL-1 inhibits bone formation. In contrast, when administered intermittently, interleukin-1 induces the proliferation of bone cells, which stimulate osteoblast differentiation into mature bone-forming osteoblasts upon withdrawal [117]. This has been demonstrated clearly in vivo, where intermittent injection of IL-1 leads to a transient increase in bone resorption followed by prolonged osteoblastic bone formation and repair of the defect. Continuous or local administration of IL-1 in sufficient doses leads to a progressive increase in extracellular fluid calcium concentrations [116,117]. However, following injections of IL-1, there is an initial transient decrease in whole blood-ionized calcium levels [118]. This decrease is relatively small, but nevertheless extremely reproducible. It seems to be mediated via prostaglandin synthesis, as it can be abrogated by the concomitant administration of indomethacin. IL-1 has been implicated in a number of disease states associated with increased bone destruction, including all diseases where proinflammatory cells infiltrate and accumulate in tissues adjacent to bone. Thus, it has been implicated in the localized osteolysis that occurs in rheumatoid arthritis and in periodontal disease. It has also been shown that IL-1 is produced by solid tumors associated with the hypercalcemia of malignancy [126]. The form of IL-1 produced in these tumors is IL-1, and it is usually produced in conjunction with the parathyroid hormonerelated protein (PTHrP). The effects of IL-1 and PTHrP are synergistic on bone resorption [126]. This likely occurs because they act at different stages in the osteoclast lineage. IL-1 increases the pool of uncommitted osteoclast progenitors or CFU-GM. In contrast, PTHrP has no effect on CFUGM, but expands markedly the pools of cells at later stages in the osteoclast lineage. IL-1 has also been implicated in the localized bone destruction associated with myeloma [127]. Freshly isolated marrow cells derived from patients
381 with myeloma release IL-1 into the media, and the boneresorbing activity present in the conditioned media can be inhibited by neutralizing antibodies to IL-1 or by IL-1 receptor antagonists (IL-1RA). Possibly the most controversial suggestion for a role of IL-1 in disease states is in postmenopausal osteoporosis. Pacifici and co-workers [128] have suggested that mononuclear cells in the estrogen-deficient state release excessive amounts of IL-1. They postulate that this action is responsible for the increase in bone turnover seen following ovariectomy or at menopause [8]. As support for their concepts, they have shown that peripheral blood monocytes in patients with postmenopausal osteoporosis produce excessive amounts of IL-1 and that this increased production can be inhibited by treatment of the patients with estrogen. They have also shown that bone loss associated with ovariectomy in the rat can be reduced by treatment with IL1RA [15,129]. Other workers have also provided data that support the notion that elevated IL-1 levels may play a role in the bone loss associated with estrogen withdrawal. For example, it has been reported that IL-1R1-deficient mice do not lose bone after ovariectomy [9]. As with most of the other osteotropic cytokines, IL-1 probably mediate its effects on 4 differentiation of osteoclast progenitors indirectly, as it does not induce osteoclast formation in the absence of stromal/osteoblastic cells. Indeed, IL-1 has been reported to induce RANKL expression in human and murine osteoblastic cells [130,131]. However, IL-1 can also promote osteoclastic bone resorption by mechanisms independent of stromal/osteoblastic cells. Moreover, it appears that of the known osteotropic cytokines and hormones, only IL-1 and RANKL are capable of acting directly on mature osteoclasts, which is consistent with the demonstration that osteoclasts express the IL-1 receptor type 1 (IL1R1) [132]. In studies using near homogeneous populations of postmitotic osteoclast precursors (prefusion osteoclasts), it has been demonstrated that like RANKL, IL-1 can prolong the survival of osteoclasts in vitro, independent of stromal/osteoblastic cells. Furthermore, IL-1 can also enhance the resorptive capacity of these osteoclasts on calcified matrices in the absence of stromal/osteoblastic cells and independent of RANKL. It is now known that there are similarities in the signaling pathways utilized by RANKL and IL-1, and the specific receptors for both cytokines, RANK and IL1 receptor, share certain intracellular signal transducers. The effect of IL-1 to induce the fusion of preosteoclasts as well as promote the survival of the multinucleated osteoclasts formed is mediated via IL-1 binding to IL-1 type 1 receptors (IL-1R) on preosteoclasts and direct activation of NF-B, independent of RANKL [99,100,132,133]. As already mentioned, binding of either cytokine to its cognate receptor activates NF-B, and in either case, NF-B activation is preceded by the recruitment of TRAF6 [55,69,99]. Indeed the inflammatory response of TRAF6-deficient mice to IL-1
382 challenge is defective [69]. Interestingly, OPG inhibits the activation of NF-B and JNK induced by RANKL, but not by IL-1 [55]. Because no osteoclasts formed at all in bones of RANKL (-/-) mice, it has been proposed that RANKL is obligatory for osteoclast formation and that it is also responsible for maintaining osteoclast activity under physiological conditions. In contrast, although its presence is not obligatory for osteoclastogenesis, by prolonging the life span of osteoclasts and enhancing their resorptive activity, IL-1 may play a major contributory role in the osteoclastic bone destruction associated with chronic inflammatory diseases such as periodontitis.
IX. TUMOR NECROSIS FACTOR TNF and lymphotoxin are classic type II membrane cytokines that stimulate osteoclastic bone resorption both in vitro and in vivo [134,135]. Their effects in vivo have been demonstrated using Chinese hamster ovarian (CHO) cells transfected with the human TNF- gene. Nude mice bearing tumors that express TNF in large amounts develop hypercalcemia and demonstrate increased osteoclastic bone resorption [135]. TNF stimulates cells at all stages in the osteoclast lineage, in much the same way as does IL-1. The effects of TNF on bone-forming cells have been less well studied but are also probably similar to those of IL-1 [125]. TNF has been implicated in hypercalcemia in several human and animal tumors associated with the humoral hypercalcemia of malignancy [136 – 138]. Antibodies to TNF reduce the blood-ionized calcium in these models as well as some of the other paraneoplastic syndromes associated with maligancy, including leukocytosis and cachexia. In these models, TNF is not produced by the tumor cells but rather by the host immune cells, possibly as part of the immune defense mechanism generated by the presence of the tumor [136]. Lymphotoxin (tumor necrosis factor-) has been implicated in the same sorts of diseases of bone as for tumor necrosis factor. Lymphotoxin is produced by stable myeloma cell lines and may be responsible for the localized osteolysis associated with myeloma [139]. When given by injection or infusion, lymphotoxin causes hypercalcemia and increased bone resorption in rodents [139].
X. INTERLEUKIN-6 Interleukin-6 is a multifunctional cytokine that appears to have a number of unique effects in bone. These appear to differ from those of other osteotropic cytokines and include the following.
GREGORY R. MUNDY ET AL.
1. IL-6 is generated in bone in response to osteotropic hormones such as PTH and cytokines such as IL-1 and TNF- [140,141]. It appears that the production of IL-6 in bone is much greater in response to cytokines than to systemic hormones [142]. For example, IL-1 induces approximately 10 times more IL-6 in bone than PTH [142]. There is not a perfect relationship between IL-6 production in bone and bone resorption. Although IL-6 mediates the effects of PTHrP, IL-1, and TNF on osteoclastic bone resorption in murine organ cultures, it remains unclear whether IL-6 mediates the effects of PTH and PTHrP on osteoclast formation in human bone marrow cultures [26]. 2. IL-6 appears to enhance the effects of other cytokines and systemic hormones on bone resorption both in vitro and in vivo [143]. We have found that not only does IL-6 have synergistic effects with interleukin-1 and PTH in organ culture and cell culture systems for assessing bone resorption, it also has synergistic effects on the boneresorbing capacity of PTH in vivo. This has been shown using CHO cells transfected with PTH and with CHO cells transfected with IL-6 [143]. The effects of both agents together are much greater than either agent alone. 3. The effects of IL-6 on bone resorption in vivo alone are modest. We have found that when IL-6 is expressed by CHO cells, there are only modest effects on serum calcium, and bone resorption is not observed unless enormous amounts of circulating IL-6 are present [144]. This is in contrast to other cytokines, such as IL-1, TNF-, and lymphotoxin. 4. It has been suggested that the effects of IL-6 on bone resorption in vivo may be enhanced by the presence of the soluble IL-6 receptor (sIL-6R). Simultaneous treatment with IL-6 and sIL-6R induced the formation of multinucleated cells with features of authentic osteoclasts in cocultures of hematopoietic osteoclast precursors and osteoblastic cells [145]. This finding may be very important, as sIL-6R is present in the circulation of patients with multiple myeloma in increased amounts [146], and it may explain enhanced effects of IL-6 on bone in this condition. 5. The source of IL-6 in bone has not been definitively clarified. IL-6 may be present in bone matrix, but it is produced in prodigious amounts by osteoclasts as well as stromal cells and osteoblasts. Whether the most important source of IL-6 is osteoclasts or cells in the osteoblast lineage remains to be determined, as many more osteoblasts than osteoclasts are present in bone preparations. 6. IL-6 receptors are present on CFU-GM colonies, which are presumed to be the earliest stage in the macrophage/osteoclast lineage, and IL-6 stimulates the formation of early osteoclast precursors from these CFUGM colonies. However, it is now clear that IL-6 induction of osteoclast differentiation is dependent solely on IL-6 receptors expressed on cells of the osteogenic lineage and
383
CHAPTER 13 Cytokines and Bone Remodeling
not on osteoclast progenitors [147]. This effect of IL-6 is probably due to its ability to enhance RANKL expression by stromal cells [130]. IL-6 has been implicated in a number of disease states. Jilka and colleagues [12] have suggested that excess production of IL-6 may account in large part for the bone loss associated with ovariectomy and estrogen withdrawal. These workers have shown that neutralizing antibodies to IL-6 reduces osteoclastic resorption associated with ovariectomy in mice. They propose that IL-6 production by stromal cells and cells in the osteoblast lineage is enhanced in the presence of estrogen deficiency. There were no significant differences in osteoclast numbers between IL-6-deficient and wild-type mice. Furthermore, ovariectomy did not induce any change in osteoclast number in IL-6-deficient mice compared to wild-type mice [13]. These data suggest that IL-6 may play a more important role in osteoclast development in pathological conditions such as estrogen-depleted states rather than in normal development. This is further emphasized by observations of morphologically normal osteoclasts in bones of gp 130-deficient mice, at least in the early neonatal period [148]. IL-6 has been implicated in the bone loss associated with myeloma. It is clear that IL-6 is expressed by some malignant plasma cells in multiple myeloma, although the actual amounts of IL-6 expressed in myeloma have been conflicting and controversial [149]. Some groups believe that the major source of IL-6 expression is not myeloma cells but rather stromal cells in the bone marrow [150]. Whether IL-6 is an autocrine or paracrine growth factor in myeloma remains controversial. Because IL-6 has such powerful effects to potentiate the bone-resorbing actions of factors such as PTH and PTHrP, its excess production may be important in some disease states where there is overproduction of factors such as PTH or PTHrP, which could lead to the local production of IL-6 in bone. This may be true in some patients with severe primary hyperparathyroidism and secondary hyperparathyroidism and in some malignancies. In each of these conditions, there is excess production of peptides, which induce IL-6 production in osteoblasts, and experimentally there is good evidence to believe that IL-6 production in bone may enhance the bone-resorbing effects of other factors such as PTH or PTHrP on murine osteoclasts. In a search for naturally occurring inhibitors of IL-6, conditioned media harvested from human and murine immune cells were examined. It was found that the monocyte – macrophage cell lines U937 and P388DI produce a biological activity, which impaired the proliferative effects of IL-6 on bone [151]. These factors were purified to homogeneity and it was found that they could be ascribed to a factor in the TGF- superfamily, activin A. Activin A has
previously been shown to be present in considerable amounts in the bone matrix and therefore may act as a stored, endogenous inhibitor of IL-6.
XI. VASCULAR ENDOTHELIAL GROWTH FACTOR VEGF has been implicated in hypertrophic cartilage remodeling, endochondral ossification, and angiogenesis [152]. It appears that VEGF-mediated capillary invasion is an essential signal regulating growth plate morphogenesis and triggering cartilage remodeling. Interestingly, it has been shown that, as with M-CSF, a single injection of recombinant VEGF can induce osteoclast recruitment and survival in the neonatal period in osteopetrotic (op/op) mice [5]. Also, recombinant VEGF can substitute for M-CSF in the formation of osteoclast-like cells in vitro in the presence of RANKL [153]. Although both cytokines are not related, these data suggest that M-CSF and VEGF have overlapping functions with regards to osteoclastic bone resorption. However, a clear role for VEGF in adult bone remodeling remains to be demonstrated.
XII. IL-15, IL-17, AND IL-18 IL-15 is an IL-2-like cytokine produced almost exclusively by T cells and which binds the same receptor as IL-2. IL-15 has been reported to stimulate the formation of TRAP-positive, calcitonin receptor-positive multinucleated osteoclast-like cells in rat bone marrow cultures that resorb calcified matrices [154]. Although IL-15 is a potent inducer of TNF, this effect to stimulate the formation of osteoclastlike cells is not blocked by a specific anti-TNF-neutralizing antibody. Although IL-15 and IL-2 also share some receptor components, IL-2 does not stimulate the formation of osteoclast-like cells. IL-15 levels in synovial fluids of rheumatoid arthritis patients are markedly elevated [155], raising the possibility that this cytokine may play a role in the local destruction of bone associated with chronic inflammatory disease. IL-17 is also a product of activated T cells that induces the production of prostaglandin E2 and IL-6 by bone marrow stromal cells and has been demonstrated to stimulate osteoclast-like cells in vitro via a PGE2-dependent mechanism [156]. Furthermore, although it had no effect on either basal or IL-1-induced bone resorption in bone organ cultures, IL-17 markedly enhanced TNF--induced osteoclastic bone resorption in fetal mouse long bones [157]. It was proposed that this cytokine also plays a role in the bone destruction associated with rheumatoid arthritis. The levels of IL-17 are also markedly elevated in rheumatoid arthritis synovial fluids compared to osteoarthritis synovial fluids of
384 from normal controls. More recent studies have shown that the effect of IL-17 to stimulate osteoclast formation is via cyclooxygenase-2-dependent PGE2 production, which in turn stimulates RANKL expression in stromal/osteoblastic cells [158]. IL-18 is a proinflammatory cytokine originally described as a product of activated macrophages but since demonstrated to be produced by marrow stromal/osteoblastic cells [157,159]. Unlike IL-15 and IL-17, IL-18 inhibits osteoclast formation in cocultures of murine spleen cells and osteoblasts, an effect likely mediated via T-cell-produced GM-CSF, as neutralizing antibodies to GM-CSF abolished osteoclast formation [159]. IL-18, which is homologous to IL-1, signals by binding to IL-1R-related protein I (IL1RrP-1), which is in turn highly homologous to IL-1R. Both IL-1R and IL-1RrP-1 associate with IL-1R-associated kinase (IRAK) and both recruit TRAF6 and both activate NF-B [120]. It remains to be seen how the two seemingly divergent downstream responses can be generated by a near identical signaling cascade.
XIII. ARACHIDONIC ACID METABOLITES: PROSTAGLANDINS AND LEUKOTRIENES Cellular activation by a variety of different stimuli results in the remodeling of membrane phospholipids to generate biologically active lipid mediators that can function intracellularly and extracellularly. A number of distinct classes of lipid mediators, or eicosanoids, are derived from a common precursor, arachidonic acid (AA). Eicosanoids, which include the prostaglandins (PGs), leukotrienes (LTs), lipoxins, and epoxides [160,161] (Fig. 2), are implicated in a variety of pathological and physiological processes, and individual members can exert opposing responses either to stimulate or to inhibit inflammation [161]. For example, thromboxane and prostacyclin have opposing functions in hemodynamics; and leukotrienes are proinflammatory, whereas the lipoxins act as endogenous antiinflammatory eicosanoids. In general, activation of phospholipase A2 results in the hydrolysis of membrane phospholipids and the subsequent release of AA. AA metabolites are then synthesized by the action of specific enzymes on AA. PGs and LTs are synthesized locally and function as autocrine, paracrine, and perhaps intracrine mediators to elicit signals in response to ligand binding and initiate responses in the immediate environment. Both cyclooxygenases and lipoxygenases translocate from the cytosolic compartment to the nuclear membrane or to the nucleus [162,163]. Translocation to the nucleus or nuclear membrane suggests that these enzymes produce eicosanoids that are active within the nucleus, as well as producing secreted eicosanoids capable of paracrine
GREGORY R. MUNDY ET AL.
FIGURE 2 Arachidonic acid metabolic pathway: 5-LO, 5-lipoxygenase; FLAP, 5-lipoxygenase-activating protein; leukotriene (LT) A4, B4, C4, D4, E4; 5-HPETE, 5-hydroperoxytetraenoic acid; 5-HETE, 5-hydroxytetraenoic acid; and COX-1 and -2, cyclooxygenase 1 and 2.
stimulation. This suggestion has merit, as metabolites of both cyclooxygenase and 5-lipoxygenase pathways show binding and activation of nuclear receptors. PG-J2, a metabolite of PG-D2, activates the peroxisome proliferatoractivated receptor- (PPAR-), a transcription factor that regulates adipocyte formation [164]. PPAR- is also a target for thiazolidinediones [165]. Leukotriene B4 (LTB4) binds PPAR-, a transcription factor that regulates the expression of enzymes involved in the oxidation of fatty acids [164 – 166]. PPAR- has been identified in osteoblasts and is associated with a potential switch to the adipocyte phenotype in these cells [167,168]. A possible role for PPAR- in osteoblasts has not been clearly established. Both types of metabolites also bind to classic seven-transmembrane Gcoupled cell surface receptors. For example, PGE2 binds to four isoforms of the PGE receptor, EP1 – EP4 [169], and their presence in osteoblasts has been demonstrated previously [170]. In bone, EP4 and EP2 appear to be critical for the anabolic effects of PGE2 [171], as well as indirect stimulation of osteoclastogenesis [172] and direct inhibition of osteoclast function [173]. Cell surface receptors for leukotriene LTB4 [174] and cysteinyl leukotrienes LTC4, D4, and E4 have also been identified [175]. The pathway most studied with respect to bone is the cyclooxygenase (COX) pathway. Two isoforms of cyclooxygenase have been identified: prostaglandin synthase1 (PGS-1, COX-1) and prostaglandin synthase-2 (PGS-2, COX-2) [176]. COX-1 is expressed constitutively and is primarily responsible for maintaining prostaglandin-mediated physiological functions. COX-2 responds to various stimuli producing prostaglandins involved in inflammation and growth regulation [177,178]. Prostaglandins have many diverse functions in humans, including blood clotting,
CHAPTER 13 Cytokines and Bone Remodeling
nerve growth, would healing, kidney function, and blood vessel tone [179]. Also, the stimulation of COX-2 can be inhibited by glucocorticoids, whereas COX-1 cannot be inhibited by glucocorticoids [180,181]. This differential response to glucocorticoids has been the impetus to develop nonsteroidal anti-inflammatory drugs that selectively inhibit the proinflammatory COX-2 enzyme [182]. Severe bone loss is a major adverse effect of glucocorticoid therapy (see Chapter 44) [183] and may be due, in part, to the combined inhibition of PG synthesis as well as stimulation of LT synthesis (discussed later). The role of the PGs in bone metabolism have been studied extensively for two decades [179,184,185]. Prostaglandins of the E series stimulate bone resorption in vitro [186] but they also stimulate bone formation in vitro and in vivo [187 – 189]. in vivo studies using rodents [190 – 192], dogs [193], and humans [194] have demonstrated the anabolic effects of PGE2. PGE2 has been shown to stimulate bone formation, improve bone structure, and increase bone mass and strength [191,195]. The bone resorptive effects of PGE2 may be critical for the initial phase of bone remodeling. A significant increase in total and osteoid-covered eroded surfaces was observed in cancellous bone sites after 5 days of PGE2 treatment in rats [196]. Although the eroded surface was not elevated in the tibial shaft at the fifth day of treatment, the eroded surface covered with osteoid was increased on the endocortical surface, which the authors concluded was due to PGE2-stimulated bone resorption on this surface prior to day 5. The osteoid perimeter was increased with PGE2 treatment as early as 5 days, and bone formation indices were increased after 10 days of treatment. This study indicates that resorption and formation were rapid with a tapering off of the resorptive phase of the response. There appears to be a dose-dependent response to PGE2 by osteoclasts; at lower doses PGE2 stimulates bone formation, whereas at higher doses PGE2 leads to bone resorption, possibly without diminishing bone formation. Because prostaglandins and leukotrienes are known inflammatory mediators within the body, the bone remodeling unit could be considered an organ-specific inflammatory response to certain proinflammatory stimuli (cytokines, bone defects, micro cracks, stress, exposure of matrix components). Inflammation is a normal response during healing, but the inflammatory effect and responses are cell-type specific. The response is extremely localized, in which the osteoclast is responsible for removing defective bone and the osteoblast is necessary for the healing phase of the inflammatory response. During the resorptive phase, the production of PGE2 is elevated. This is followed by the bone formation, or healing phase, which most likely correlates with a reduction in PGE2 levels. This would suggest that bone formation proceeds without further effects on resorption. Under pathological conditions, constant stimulus and production
385 of inflammatory mediators may lead to progressive bone loss. To this end, it is not surprising to find that cytokines can stimulate bone cells to proliferate, differentiate, and undergo apoptosis in a similar manner to other cell types. Although the in vivo anabolic effects of PGE2 have been well established, the tissue nonspecificity of the metabolite makes it an unlikely therapeutic agent for bone loss. Side effects such as severe diarrhea, hair loss, and decreased physical activity accompanied systemic or oral treatment in animals. The leukotrienes and peptidoleukotrienes are 5-lipoxygenase (5-LO) metabolites of arachidonic acid that appear to have unique effects on bone. Previous work examining the role of leukotrienes in bone metabolism has mainly focused on their effects on osteoclasts. In vitro and in vivo evidence shows that leukotriene 5-LO metabolites (namely 5-HETE and the cysteinyl-leukotrienes LTC4, LTD4, and LTE4) stimulate the formation and activity of avian osteoclasts in vitro [197]. Studies show that leukotriene B4 (LTB4) stimulates bone resorption both in vitro and in vivo [198,199]. When LTB4 was injected over the calvaria of mice, there was a significant increase in osteoclast numbers per unit surface area of bone [198]. Meghji and co-workers [200] had also found LTB4 to be a more potent activator of bone resorption in the mouse calvarial assay. These studies indicated that 5-LO metabolites stimulate the recruitment, formation, and activation of osteoclasts. However, these effects may be indirect, as mature mouse osteoclasts and human giant cells do not express mRNA for either the 5-LO enzyme or the LTB4 receptor. The response of avian osteoclasts may be species specific or the effects of the leukotrienes occur in a precursor population. Leukotrienes and lipoxins may be important in hematopoiesis by regulating the production of committed progenitors: CFU-GM and BFU-E [201 – 204]. It may be that fully differentiated, mature osteoclasts lose the ability to synthesize leukotrienes or do not express cell surface receptors. Few studies have directly addressed the effects of 5-LO metabolites on osteoblast function. Previous reports have shown that LTB4 inhibited the proliferation of normal osteoblastic rat calvarial cells, as well as the osteoblastic cell lines SaOS-2 and G292, and increased intracellular calcium release in osteoblasts derived from neonatal mice calvaria [205]. Data suggest that mice lacking the functional gene for 5-LO have increased cortical bone thickness compared to wild-type mice [206] and have significantly different mechanical properties of bone compared to wild-type animals [207]. These observations suggest that increased bone formation may occur in the absence of leukotriene synthesis. More recently, the capacity of osteoblasts to differentiate and to form bone was inhibited in both the bone nodule formation assay and the calvarial organ culture assay in the presence of 5-LO metabolites [208]. The exogenous addition of 5-hydroxyeicosatetraenoic acid (5-HETE) and LTB4
386 showed a dose-dependent decrease in alkaline phosphatase activity consistent with the inhibition of osteoblast differentiation and inhibition of bone nodule formation in fetal rat calvarial cultures [208]. These studies all support the hypothesis that 5-LO metabolites are negative regulators of bone formation. Leukotrienes and cysteinyl-leukotrienes have been implicated in a number of chronic inflammatory conditions, such as rheumatoid arthritis (RA), asthma, psoriasis, periodontal disease, and inflammatory bowel disease [209]. 5-LO metabolites may be responsible for decreased osteoblast function or decreased bone formation in conditions of elevated 5-LO metabolite production, such as the acutephase inflammatory response and rheumatoid arthritis. Osteoblasts express mRNA for all enzymes in the leukotriene pathway necessary for leukotriene synthesis. LTB4 receptor mRNA was also expressed by these cells, potentially indicating an autocrine/paracrine role for these metabolites in osteoblasts. In contrast to this, because mammalian osteoclasts (mouse and human giant cells) [210] do not express LTB4 receptor mRNA, the regulation by leukotrienes may be either at the precursor level or indirectly through the osteoblast. Leukotriene synthesis inhibitors or receptor antagonists have been used successfully to treat asthma [211,212]. These compounds may have therapeutic potential with regards to bone formation. Inhibition of leukotriene synthesis with the use of 5-LO enzyme inhibitors or leukotriene receptor antagonists resulted in increased bone-like nodule formation, which was negated by the coaddition of indomethacin, indicating cross talk between the 5-LO and COX pathways. The possible interaction between cyclooxygenase and lipoxygenase pathways was demonstrated in the HT29 cl.19A enterocyte cell line, which has 5-LO metabolism without 5-LO activating protein (FLAP) [213]. FLAP cDNA-transfected clones resulted in increased COX-2 expression and PGE2 synthesis. If 5-LO metabolites regulate bone formation negatively, and in the complete absence of 5-LO expression and hence leukotriene synthesis, the basal prostaglandin levels (COX-1 and 2) may be sufficient to maintain bone formation, which eventually leads to the accumulation of bone in these animals (Fig. 3). Glucocorticoids are the mainstay treatment of many inflammatory conditions, including arthritis and asthma. The net in vivo effect of glucocorticoids on bone is to inhibit osteoblast differentiation and function [214]. They activate osteoclasts, inhibit intestinal calcium absorption, and suppress the gonadal axis. At supraphysiological concentrations in vivo, as observed in patients with hypercortisolism (Cushing’s syndrome) [215] and steroid-induced osteoporosis [216], there is an increased bone turnover rate, resulting from increased bone resorption and decreased bone formation. A fine balance exists between soluble mediators released by activated cells of the immune system (cytokines) and
GREGORY R. MUNDY ET AL.
FIGURE 3
Potential shunting between 5-lipoxygenase and cyclooxygenase pathways. Leukotrienes negatively regulate bone formation mediated by prostaglandins. Physiological levels of leukotrienes and prostaglandins work in concert to regulate bone formation. Inhibition of LT synthesis or blockade of their receptors may lead to increased bone formation or a reduction in bone loss associated with inflammatory conditions. Basal cyclooxygenase expression is controlled by COX1, which is not affected by glucocorticoids. COX2, the inducible enzyme, is inhibited by glucocorticoids. Low levels of PGE2 are necessary and stimulate bone formation in vivo. 5-HETE and the leukotrienes stimulate bone resorption and inhibit formation.
products released by the neuroendocrine system in response to these inflammatory mediators. For the most part, homeostasis is maintained under conditions of “stress”, however, in conditions of chronic inflammation, such as rheumatoid arthritis, there remains an imbalance [217]. In general, inflammatory cytokines, such as IL-1, IL-6, and TNF-, stimulate the production of corticotrophin-releasing hormone and arginine vasopressin from the hypothalamus. This in turn stimulates the release of ACTH from the pituitary, followed by glucocorticoid secretion by the adrenal cortex and indirect effects on gonadal function. The hormone products of the hypothalamic – pituitary – adrenal and the hypothalamic – pituitary – gonadal axes are capable of modulating cytokine production [217]. Glucocorticoids are the most potent endogenous inhibitors of immune and inflammatory processes, including the production of inflammatory cytokines [217]. In newly diagnosed, untreated RA patients, elevated levels of IL-6 stimulate the production of CRH, ACTH, and cortisol [218]. However, although overall hypothalamic – pituitary – adrenal axis activity appears to be normal, it may be insufficient to inhibit the chronic inflammatory condition. IL-6 levels were increased significantly in RA patients compared to controls in the early morning hours. In the face of elevated cytokines, perhaps due to glucocorticoid insufficiency or resistance of target tissues [219], these cytokines are able to stimulate osteoclast formation and activation. This effect appears to be mediated through the IL-6 receptor present in osteoblasts [147]. In coculture experiments, dexamethasone treatment increased IL-6 receptor mRNA in osteoblasts. Although the effects of cytokines on bone have been well documented [119 – 222], only a few groups have demonstrated their effects on the 5-LO pathway in osteoblasts. IL-1 has been shown to stimulate 5-hydroxyeicosatetraenoic
387
CHAPTER 13 Cytokines and Bone Remodeling
acid (5-HETE) and PGE2 production in G292 osteoblasticlike cells [223]. The same investigators later demonstrated dose-dependent effects on the cyclooxygenase and lipoxygenase pathways by TNF- [224]. High-dose TNF- (108M) stimulated PGE2 production, whereas low-dose TNF- (1010M) stimulated 15-HETE production. An intermediate dose resulted in 5, 12, and 15-HETE and leukotriene B4 (LTB4) production. In other cell types, IL-1 and IL-6 stimulate the production of 5-HETE [225]. Although glucocorticoids have been used effectively for the symptomatic treatment of RA, their long-term consequence is osteopenia. The anti-inflammatory use of glucocorticoid has been mainly as a replacement therapy in regimens that vary from low-dose oral administration to high-dose pulse therapy and intra-articular injection, which can modulate the activity of activated T lymphocytes and macrophages [226]. The most noted effect of glucocorticoid administration is related to the inhibitory effect on the inducible cyclooxygenase enzyme, COX2. O’Banion et al. [227] demonstrated inhibition of COX-2 by glucocorticoids. Messenger RNA for COX-2 was completely blocked by dexamethasone treatment in murine fibroblasts and IL-1-stimulated human monocytes. Similarly, dexamethasone inhibited COX-2 stimulation and PGE2 production by IL-1 and TNF- [228] and PTH [229] in human osteoblasts. However, differential regulation of COX-2 and 5-LO by dexamethasone has been reported. Goppelt-Struebe et al. [230] showed increased 5-LO activating protein (FLAP) and decreased COX-2 expression in the human monocytic leukemia cell line THP-1. In a longer, conditioning experiment, Riddick et al. [231] showed that 5-LO and FLAP mRNA were increased by dexamethasone in these cells. These experiments suggest a shunting of the arachidonic acid cascade from the cyclooxygenase pathway to the 5-LO pathway (Fig. 3). These data indicate that metabolites of the 5-LO pathway are negative regulators of bone formation. The coordinate regulation of COX and LOX pathways may be required to maintain bone metabolism, and it is the imbalance between the pathways that may account for bone loss. The continued presence of these metabolites in the bone environment might account, in part, for the bone loss associated with chronic inflammatory conditions.
XIV. TRANSFORMING GROWTH FACTOR- Transforming growth factor- has potent effects on the activity of both osteoblasts and osteoclasts and is therefore another important modulator of bone remodeling (see also Chapter 14). Large amounts of TGF- are stored in bone matrix in a latent form, making bone the most abundant source of TGF- in the body [232,233]. TGF-1 is the predominant isoform, making up 80 – 90% of matrix-stored
TGF-, with smaller amounts of TGF-2 and -3 [234 – 237]. Bone matrix-bound TGF- can be released and activated by resorbing osteoclasts [238 – 239]. Results from in vitro and in vivo studies on the effects of TGF- in bone have been somewhat controversial. However, a general consensus is that active TGF- has stimulatory effects on bone formation and inhibitory effects on bone resorption. Thus TGF- has been viewed as a coupling factor, which links bone resorption to subsequent bone formation.
A. Effects of TGF- in Bone In vivo studies with TGF- have shown that, unlike bone morphogenetic proteins, TGF- is not able to stimulate new bone formation in ectopic sites. However, TGF-1 does stimulate new bone formation when injected in close proximity to bone. TGF-1 injected over the calvaria of mice induces new bone formation [240 – 243], as does TGF-1 injected into skull defects [244] and TGF-1 and -2 injected systemically [245 – 247]. The overall effect of TGF- to enhance bone formation appears to be due to TGF- exerting effects at multiple stages in the osteoblast life cycle. Thus, TGF- is a potent chemotactant, which attracts osteoblast precursors to the resorption defect [248]. TGF- also appears to be mitogenic for these osteoblast precursors [249], stimulating them to proliferate at the site of new bone formation. The effect of TGF- on the mature osteoblast may then be to stimulate the production of matrix proteins, such as type I collagen, leading to the production of osteoid. However, in order for mineralization to proceed, it appears that TGF- must be withdrawn. Thus, when primary cultures of fetal rat calvarial osteoblasts or mineralizing bone organ cultures are treated continuously with TGF-, mineralization is inhibited [250 – 252]. Similarly, in animal models of TGF--induced bone formation, TGF- injections must be given and then bone formation allowed to continue in the absence of further treatment. If TGF- injections continue, bone formation will be inhibited. Interestingly, in transgenic mice, which chronically overexpress TGF-2 driven by the osteocalcin promoter, a similar inhibition of mineralization is seen, producing an osteopenic phenotype [253]. The effects of TGF- on bone resorption remain controversial. Following local injection over the calvaria in mice, TGF- does not stimulate bone resorption locally on the periosteal side of the bone, as seen with interleukin-1 or PTHrP. However, it does cause an increase in osteoclast number in the marrow spaces [241]. Its effects in in vitro models of bone resportion appear to be system specific. In cultured fetal rat long bone and in human and murine bone marrow cultures, TGF- inhibits osteoclast formation and osteoclastic bone resorption [29,30]. In contrast, in neonatal mouse calvariae, it stimulates osteoclastic bone resorption, probably through the stimulation of prostaglandin production [254].
388
GREGORY R. MUNDY ET AL.
An exciting recent finding is that in addition to inhibiting osteoclast formation and activity, TGF- may inhibit bone resorption by stimulating osteoclast apoptosis. Furthermore, TGF- appears to mediate the stimulation of osteoclast apoptosis by estrogen, suggesting that it may play a role in postmenopausal osteoporosis [31].
B. Targeted Disruption of TGF- Genes Reveals Isoform-Specific Roles in the Skeleton The generation of mice lacking the genes for TGF-1, 2, and 3 has provided valuable insights into the role of TGFs in bone [255]. Each of these gene knockouts has specific, but nonoverlapping, bone phenotypes, suggesting a specific role for each isoform. Mice lacking the TGF-1 gene, which are born to heterozygous mothers, die shortly after weaning due to inflammatory disease [256 – 257]. Although no gross abnormalities are observed in the bones of these animals, they do show decreased bone mass, reduced bone length, decreased bone elasticity, and are smaller than their normal siblings [258]. The potential contribution of TGF-1 to bone loss in postmenopausal or aging animals has not yet been assessed using this model due to the difficulty in maintaining mice to adulthood. Additional roles for TGF-1 may yet be revealed following various pathological challenges in these animal models. Mice that lack the gene for the TGF-2 isoform show a wide range of developmental defects, including cranial defects, cardiac, lung, and urogenital defects [259]. Mice lacking the gene for TGF-3 show failure of the palatial shelves to fuse, resulting in cleft palate. These mice also show abnormal lung development and die soon after birth [260]. The TGF-1 knockout model has been complicated by the observation that TGF- can pass from mother to fetus both across the placenta and in the mother’s milk [261]. Thus, in TGF-1 knockouts born to heterozygous mothers, this “maternal rescue” phenomenon may mean that the mice are not totally lacking TGF-1, but should be viewed more correctly as TGF-1 deficient. To eliminate the effects of multifocal inflammation, TGF- null deletions were created in severe combined immunodeficient (SCID) mice [262]. This enabled some of the knockout mice to survive to adulthood. These mice are 50 – 80% the size of their normal littermates, show a lack of vigor, and do not thrive, however, these investigators were able to breed a TGF-1 null female with a heterozygous male. The two TGF-1 null pups born to this TGF-1 null female were viable, suggesting that lack of TGF-1 is not lethal in the embryo.
C. TGF- Gene Polymorphisms Various polymorphisms of the TGF-1 gene have been associated with low bone mass and fracture risk in osteoporotic
women [263 – 265]. Some of these polymorphisms have been shown to result in reduced circulating TGF-1 concentrations. This suggests that TGF-1 allelic variants may be important determinants of bone mass and that analysis of the TGF-1 genotype may prove useful in the prevention and management of osteoporosis. In support of these genetic findings, animal studies have shown that osteopenia in old male mice was due to reduced TGF- content of the bone matrix and reduced TGF- responsiveness of marrow osteoprogenitors [266].
D. TGF- Receptors TGF-s signal through serine/threonine kinase receptors, which are composed of type I and type II subunits [267 – 269]. TGF- binds to the type II receptor, which then recruits the type I receptor to the complex. The type II receptor then phosphorylates the type I receptor, which initiates the signaling cascade through phosphorylation of a family of downstream signaling molecules known as Smads. TGF- signals through Smads 2 and 3, which then associate with Smad 4. Smad 4 translocates the complex to the nucleus, where the complex, in conjunction with other DNA-binding proteins, initiates the transcription of TGF-regulated genes. In addition, these activated Smads can be inhibited by the negative regulators, Smad 6 and Smad 7. Bone cells express both type I and type II receptors and are capable of transducing TGF- signals via the Smad pathway [235,270,271]. Experiments using transgenic mice have emphasized the importance of TGF- receptor signaling in skeletal tissue. Expression of a dominant-negative type II TGF- receptor in osteoblasts under control of the osteocalcin promoter resulted in increased trabecular bone in transgenic mice [272]. Expression of the same dominant-negative TGF- type II receptor in mouse perichondrium/periosteum, synovium, and articular cartilage, under control of a metallothionine-like promoter, resulted in progressive skeletal degeneration resembling osteoarthritis, suggesting that TGF- may be important for the normal maintenance of synovial joints [273]. Bone cells also express betaglycan, which is known as the type III TGF- receptor. This membrane-associated proteoglycan binds TGF- with high affinity but does not appear to signal directly. The type III receptor is thought to function as a cell surface reservoir for TGF-, which presents TGF- to the type II receptor [274].
E. TGF- Latency TGF- is produced by most cells, including bone cells, as one or more latent complexes that must be activated in order for TGF- to exert its biological effects (Fig. 4).
389
CHAPTER 13 Cytokines and Bone Remodeling
The forms of TGF- found in the bone microenvironment and the potential mechanisms for the regulation of TGF- activity by bone cells. TGF- is made by osteoblasts in two latent forms: a small latent complex (100 kDa) that lacks the 190-kDa latent TGF--binding protein (LTBP-1) and a large latent complex that contains LTBP-1. LTBP-1 appears to target the latent complex to the bone matrix for storage in a network of fibrillar structures. Distinct from this function associated with latent TGF-, LTBP-1 may also function as an extracellular matrix protein that influences the structure of the bone matrix. Latent TGF- can be activated by numerous mechanisms. One well-known cellular mechanism is the activation by resorbing osteoclasts. Another mechanism is through the action of proteases. Proteases, such as plasmin, can easily activate the small latent complex through proteolytic digestion of the precursor portion. However, in the large latent complex, the hinge region of LTBP-1 is highly susceptible to cleavage by proteases, thereby releasing the large latent complex containing a truncated form of LTBP-1. The target for this form of latent TGF- is being investigated.
FIGURE 4
Because virtually all cells express surface receptors for TGF-, activation of latent TGF- may be key point for the regulation of TGF- activity. Bone cells produce at least three forms of latent TGF-, including small (100 Da) and large (290 Da) latent TGF-1 and TGF-2 complexes [275,276]. The 100-Da small latent TGF- complex consists of the mature 25-kDa TGF- homodimer, which is cleaved from but remains noncovalently associated with a 75-kDa portion of the propeptide homodimer. The large (290-kDa) complex is identical, except that one of the propeptide chains is disulfide linked to a third protein called the latent transforming growth factor -binding protein-1 (LTBP1). This protein, originally identified as a component of the large latent TGF- complex [277,278], is clearly a matrix protein, which colocalizes with fibrillin-1 in microfibrillar structures in bone matrix [279,280]. LTBP1 is thought to play an important role in the regulation of TGF- at multiple levels. Thus, LTBP1 has been shown to facilitate the secretion of latent TGF- from the cell [281] and is also important in targeting TGF- for storage in the extracellular matrix [279,282,283]. By undergoing proteolytic cleavage, LTBP1 may also provide a mechanism for the release of latent TGF- from the extracellular matrix [279,282,284]. LTBP1 belongs to a family of recently identified matrix proteins, which share homology with fibrillins 1 and 2 and appear to be important components of connective tissue microfibrils [285]. Four LTBPs
have been identified, which range in size from approximately 180 to 310 kDa. LTBPs 1,3, and 4 all appear to bind small latent TGF-, but there are conflicting reports as to whether LTBP2 binds to TGF-. Although it is known that bone cells express LTBP2 and LTBP3 [286; Dallas, unpublished observations], future studies are clearly warranted to determine their specific roles in bone cell function and in regulation of TGF-. Activation of latent TGF- complexes involves the dissociation of mature TGF- from the propeptide. This can be achieved in vitro by extremes of pH, by chaotropic agents, by heat or via the action of proteases. Several studies have documented factors that stimulate or inhibit the activation of latent TGF- in various cell systems. However, very little is known about the actual mechanisms for activation of TGF- by cells [287]. A protease-mediated mechanism appears the most likely mechanism for the majority of cell types. In support of this, evidence for a plasmin-mediated activation mechanism has been reported in UMR-106 osteosarcoma cells [288]. Osteoclasts are interesting in that they are one of the few cell types for which an acid-mediated mechanism of activation for TGF- seems likely. The pH in the sealed zone, under the ruffled border of the osteoclast, has been reported to be as low as 5 [289]. Significant amounts of TGF- would be expected to be activated at this pH. Thus, osteoclasts may be unique in utilizing an acid-mediated mechanism of activation of latent TGF-.
390 In addition to its roles in matrix storage and release of latent TGF-, LTBP1 may play a role in the activation of latent TGF-. Flaumenhaft and co-workers [290] showed that antibodies to LTBP1 inhibited the activation of latent TGF- by cocultures of smooth muscle and endothelial cells. LTBP1 may also modulate activation by protecting latent TGF- from activation by proteases until the complex is bound to the surface of an appropriate cell that expresses cell surface-bound protease systems for activation. An important role for thrombospondin in the activation of TGF- has been suggested by studies showing that thrombospondin can activate latent TGF- [291] and, more recently, by the observation that thrombospondin null mice show a phenotype resembling the TGF- knockout mouse [292]. However, the role of thrombospondin in the activation of TGF- in bone cells remains undetermined. Future studies are clearly required to unravel the mechanisms of activation of TGF- in bone and will be the key to understanding the regulation of this important factor in bone. Figure 4 shows the relationship between TGF- and its associated binding proteins.
XV. BONE MORPHOGENETIC PROTEINS Studies of stromal cell transplantation in vivo [293] and analysis of different mouse or rat clonal cell populations in vitro [294 – 296] provide evidence that multipotential mesenchymal cells can differentiate into different cell types, including osteoblasts and chondroblasts. These mesenchymal cells have the capacity to undergo the commitment process to give rise to progeny with more limited or monopotential differentiation capacity. The mechanism of commitment and specification of uncommitted mesenchymal precursor cells to the osteoblast lineage is not fully understood. BMPs appear to play regulatory roles in the commitment of mesenchymal precursor cells to the osteoblast lineage [297]. BMPs are originally identified from bone matrix using an ectopic bone formation assay [298,299]. When BMPs are implanted subcutaneously or intramuscularly in mice or rats, they induce massive amounts of new cartilage and bone at implantation sites. BMP-2 has been shown to induce embryonic limb cells to differentiate into mature chondroblasts and osteoblasts [300]. BMP-2, BMP-4, and BMP-7 induce mesenchymal precursor C3H10T1/2 cells to differentiate into mature chondroblasts and osteoblasts [301,302]. These results indicate the regulatory roles of BMPs in the commitment of mesenchymal cells to the osteoblast and chondroblast lineages (also see Chapters 5 and 14). BMPs mediate their functions through type I and type II serine/threonine kinase receptors. To determine which type I BMP receptor is responsible for the commitment of mesenchymal precursor cells to the osteoblast lineage and for
GREGORY R. MUNDY ET AL.
BMP-induced osteoblast differentiation, we have stably transfected dominant-negative truncated and constitutively active type IA and IB BMP receptors (BMPR-IA and BMPR-IB) into a mesenchymal precursor cell line, 2T3. Overexpression of the truncated type IB BMP receptor, but not the truncated type IA BMP receptor, blocks the commitment of 2T3 cells to the osteoblast lineage and BMP-2induced osteoblast differentiation [167]. The differentiation pathway of 2T3 cells is respecified to the adipocyte lineage after type IB BMP receptor signaling is blocked [167]. These findings suggest that the temporal acquisition of the type IB BMP receptor in mesenchymal precursor cells may be a key step for the commitment of mesenchymal cells to the osteoblast lineage.
A. Role of BMPs and BMP Receptors in Osteoblast and Chondroblast Differentiation BMPs play important roles in osteoblast differentiation. BMP-2 stimulates osteoblast differentiation in primary osteoblastic cells [303,304] and in cell lines derived from osteogenic tissues [305 – 307]. BMPs also induce nonosteogenic precursor cells to differentiate into cells with osteoblast phenotypes. For example, BMP-2 induces myoblast C2C12 cells to differentiate into osteoblasts [308]. BMP-2, BMP-6, and BMP-7 have all been shown to induce both hypertrophy in cultured chondrocytes and osteogenesis from mesenchymal stem cells [309]. BMP-6 may play a central role in chondrocyte differentiation and endochondral bone formation, as glucocorticoid and estrogen selectively upregulate BMP-6 expression [310,311], and PTHrP, a key factor in chondrocyte differentiation, inhibits BMP-6 expression in chondrocytes [312]. BMP-2 upregulates BMP-2 and BMP-4 mRNA expression in primary osteoblastic cells [303,304], suggesting that BMP-2 acts as an autocrine and paracrine factor during osteoblast differentiation. The autocrine effect of BMP-2 may be mediated in part through transcription factors, Dlx2 and Dlx5, as BMP2 stimulates Dlx2 and Dlx5 mRNA expression and Dlx2 can bind and activate BMP-2 gene transcription [313,314]. BMPs may stimulate osteoblast differentiation in part through activating the osteoblast-specific transcription factor, Osf2/Cbfa1. In Osf2/Cbfa1 knockout mice, both membranous bones of the skull and endochondral bones in the rest of the skeleton are absent [315]. Heterozygous mutations of the Osf2/Cbfa1 gene are found in humans with cleidocranial dysplasia (CCD) [316,317]. BMPs may be important regulators of Osf2/Cbfa1, as BMP-7 and BMP-2 stimulate Osf2/Cbfa1 expression in pluripotent mouse fibroblast C3H10T1/2 cells and in mesenchymal precursor 2T3 cells [167,318]. In vitro studies using dominant-negative truncated and constitutively active type IA and IB BMP receptors reveal
391
CHAPTER 13 Cytokines and Bone Remodeling
the essential roles of the type IB BMP receptor in chondroblast and osteoblast differentiation. Expression of a dominant-negative type II and type IB BMP receptors in immature chondrocytes leads to a loss of differentiated function and inhibits the expression of chondrocyte marker genes such as type II collagen and aggrecan [319]. When the truncated type IB BMP receptor is overexpressed in mesenchymal precursor 2T3 cells, the osteoblast differentiation properties of 2T3 cells are blocked. Both BMP-2-induced mineralized bone matrix formation and osteoblast-specific gene expression such as Osf2/Cbfa1 and osteocalcin are inhibited [167].
in the shapes of the epiphyses of the long bones, suggesting that the type IB BMP receptor is essential for the formation of joints [329]. This phenotype shows close similarities to that of the bp/bp (brachypodism) mouse, which contains the mutated GDF-5 gene [326]. GDF-5 has been shown to bind specifically to the type IB BMP receptor but not to the other type I receptors [330]. In BMPR-IA homozygous null mutant mice, morphological defects are first detected at 7.0 days postcoitum (dpc) and no mesoderm is formed in the mutant embryos, suggesting that BMP signaling through the type IA BMP receptor is essential for mesoderm formation during gastrulation [331].
B. Role of BMPs and BMP Receptors in Bone Development and Bone Formation
C. Signaling Mechanism of BMP Receptors (Fig. 5)
The function of BMP-2, BMP-6, and BMP-7 in ectopic bone formation and in fracture repair have been well characterized [309,320,321]. These peptides all induce ectopic bone formation and accelerate the healing processes during fracture repair. Generation of null mutant mice lacking BMP-2 or BMP-4 genes results in early embryonic lethality, before any skeletal formation has been initiated [322,323]. Tissue-specific knockout of these genes will clarify the specific roles of these genes in bone development and bone formation. In BMP-7-deficient mice, skeletal abnormalities are identified in discrete areas: the rib cage, the skull, and the hindlimbs [324], suggesting that BMP-7 plays a role in bone development and patterning. Consistent with the roles for BMPs in bone development and bone formation, studies of naturally occurring mutations have shown that different members of the BMP family may control the formation of different morphological features in the mammalian skeleton. Mutation in the BMP5 gene is associated with a wide range of skeletal defects, including reductions in long bone width and the size of several vertebral processes and an overall lower body mass [325]. Mutations in the GDF5 (CDMP-1/BMP-11) gene result in brachypodism in mice [326] and Hunter – Thompson-type chondrodysplasia in humans [327]. These findings demonstrate the critical roles of BMPs in bone development and bone formation. In vivo studies using virus-mediated kinase defective and constitutively active type IA and IB BMP receptors demonstrate that the type IB BMP receptor is required for cartilage formation. In chicken embryos, chondrogenesis of limb mesenchymal cells and bone formation in the distal limb bud are markedly inhibited by dominant-negative type IB BMP receptors, but not by dominant-negative type IA BMP receptors [328]. Mice lacking the type IB BMP receptor gene are viable and display brachydactyly due to fusion and severe reduction in length of the first and second phalages. X-ray analysis of BMPR-IB-/- mice reveals defects
BMPs signal through serine/threonine kinase receptors, composed of type I and type II components. Three type I receptors have been shown to bind to BMP ligands: BMPRIA, BMPR-IB, and ActR-IA (also termed ALK-3, ALK-6, and ALK-2, respectively) [332,333]. Three type II receptors for BMPs are also identified. BMPR-II, ActR-IIA, and ActR-IIB [334 – 337]. Whereas BMPR-IA, BMPR-IB, and BMPR-II are specific to BMPs, ActR-IIA and ActR-IIB are also signaling receptors for activins. ActR-IA has been shown to bind both BMP-7 and activin when expressed in COS cells. In embryonic P19 cells and MC3T3-E1 cells, endogenous ActR-IA mediates only BMP signaling, but not activin signaling [338]. No cross interaction between BMP and TGF- receptors has yet been demonstrated. Signal transduction through serine/threonine kinase receptors has been best characterized in the TGF- receptor system. It is likely that BMPs transduce signals in a similar
FIGURE 5
The BMP ligand – receptor and signal transduction pathway.
392
GREGORY R. MUNDY ET AL.
fashion. Type I and type II BMP receptors are both indispensable for signal transduction; after ligand binding they form a heteromeric-activated receptor complex. Both type I and type II BMP receptors are composed of three parts: short extracellular domains with 10 – 12 cysteine residues; single membrane-spanning domains; and intracellular domains with a serine/threonine kinase region [332 – 336]. Preceding the serine/threonine kinase domain, type I BMP receptors, but not type II BMP receptors, have a domain that contains a characteristic SGSGS motif (GS domain) [332,333]. The GS domain, which plays an important role in signal transduction, distinguishes type I serine/threonine kinase receptors from type II receptors. In the TGF- and activin receptor systems, ligands bind to type II receptors in the absence of type I receptors. Type I receptors can bind ligands only in the presence of type II receptors. However, BMPs bind weakly to type I receptors in the absence of type II receptors. In the presence of type II receptors, the binding of BMPs to type I receptors is accelerated [332,333]. In the TGF- receptor system, the type II receptor kinase transphosphorylates the GS domain in the type I receptor, which leads to activation of the type I receptor kinase [337]. Phosphorylation of the type I receptor is required for signal transduction, as TGF--mediated responses are impaired after mutation of serine and threonine residues in the GS domain of type I receptor or after mutation in the type II receptor, rendering it incapable to phosphorylate the type I receptor [339,340]. These observations indicate that the type II receptor is a primary binding protein for ligand and that the type I receptor acts as an effector in the signal transduction. This notion is supported by the observation that the mutation of glutamine in the GS domain of the type I BMP receptor to aspartic acid results in a receptor with a constitutively activated kinase. In these mutants, signals are transduced from the type I BMP receptor in the absence of ligand and the type II BMP receptor [341].
D. Downstream Molecules of BMP Receptor Signaling Type I BMP receptor substrates include a recently identified protein family, the Smad proteins, that play a central role in the relay of BMP signals from the receptor to target genes in the nucleus. Smad 1 [341], Smad 5 [342], and Smad 8 [343] are phosphorylated by BMP receptors in a ligand-dependent manner. These receptor-regulated Smads physically associate with the ligand-activated receptor complex and undergo phosphorylation at the C terminus [342]. After release from the receptor, Smad proteins associate with the related protein Smad 4, which acts as a shared partner. This complex translocates into the nucleus and
participates in gene transcription. In vertebrates, Smad proteins consist of three regions: a conserved N-terminal domain (MH1 domain), a conserved C-terminal domain (MH2 domain), and a more divergent linker region. In the transcriptional complex, the Smads contact DNA via their N-terminal domains [344]. The C-terminal domain of Smads mediates Smad – receptor interaction [345]. Activation of specific genes by Smads is conducted by interaction with specific DNA-binding proteins. The Xenopus protein Fast1 is the prototypic Smad-recruiting, DNAbinding factor [346]. Fast1, which contains a “winged helix” DNA-binding domain, binds to the activin response element (ARE) and is absolutely required for activation of the Mix.2 gene in response to activin or TGF-. Fast1 bound to DNA alone does not activate transcription. However, recruitment of an activated Smad2 – Smad 4 complex to the ARE by Fast1 results in activation of Mix.2 expression [347]. One of the transcription factors, which interacts with Smad 1, has been identified as a homeodomain DNAbinding protein, Hoxc-8. Hoxc-8 can serve as a transcriptional repressor for osteopontin gene transcription. Interaction of Smad 1 with Hoxc-8 relieves the repressive activity of Hoxc-8 and activates target gene transcription. The interactive regions of Smad 1 with Hoxc-8 are in the MH1 domain and linker region [348]. Consistent with these findings, transgenic mice overexpressing Hoxc-8 show abnormal cartilage, which is characterized by an accumulation of proliferating chondrocytes and reduced maturation [349]. Interestingly, Smad 2, a downstream molecule for TGF- signaling, interacts with a homeodomain DNAbinding protein, TGIF, a repressor of transcription, when it moves into the nucleus with Smad 4. The Smad 2 – Smad 4 complex can recruit TGIF and histone deacetylases (HDACs) to a Smad target promoter, repressing transcription [350]. These findings provide molecular evidence for the functional connection between growth factors activin/BMP/TGF- and homeodomain DNA-binding transcription factors. The immediate downstream target genes for BMP signaling during osteoblast and chondroblast differentiation have yet to be identified. Dlx2 and Dlx5 are candidate target genes as immediate downstream molecules for BMP signaling, as BMP-2 induces Dlx2 and Dlx5 mRNA expression within 1 h and without the synthesis of new proteins [313,314].
XVI. CONCLUSION Rapidly accumulating evidence shows cytokines generated in the bone microenvironment control bone remodeling. There is now a body of data derived from in vivo studies in animals which show that over- or underproduction of certain cytokines cause profound effects on bone. These
393
CHAPTER 13 Cytokines and Bone Remodeling
fundamental observations have the potential of not only increasing our understanding of the pathophysiology of osteoporosis, but also leading to new and better forms of therapy using these molecules as targets for drug discovery programs.
13.
14.
Acknowledgments
15.
We are grateful to Nancy Garrett for her assistance in the preparation of the manuscript. This work was supported by Grants P01 CA40035, AR28149, AR07464, RR01346, DK45229, and AR42372 from the National Institutes of Health.
16.
17.
References 1. S. Akira and T. Kishimoto, NF-IL-6 and NF-B in cytokine gene regulation. Adv. Immunol. 65, 1 – 46 (1997). 2. R. Felix, M. G. Cecchini, and H. Fleisch, Macrophage-colony stimulating factor restores in vivo bone resorption in the op/op osteopetrotic mouse. Endocrinology 127, 2592 – 2594 (1990). 3. H. Kodama, A. Yamasaki, M. Nose, S. Niida, Y. Ohgame, U. Abe, M. Kumegawa, and T. Suda, Congenital osteoclast deficiency in osteopetrotic (op/op) mice is cured by injections of macrophage colony stimulating factor. J. Exp. Med. 173, 269 – 272 (1991). 4. H. Kodama, A. Yamasaki, M. Abe, S. Niida, Y. Yakeda, and H. Kawashima, Transient recruitment of osteoclasts and expression of their function in osteopetrotic (op/op) mice by a single injection of macrophage colony-stimulating factor. J. Bone Miner. Res. 8, 45 – 50 (1993). 5. S. Niida, M. Kaku, H. Amano, H. Yoshida, H. Kataoka, S. Nishikawa, K. Tanne, N. Maeda, S. I. Nishikawa, and H. Kodama, Vascular endothelial growth factor can substitute for macrophage colony stimulating factor in the support of osteoclastic bone resorption. J. Exp. Med. 190, 293 – 298 (1999). 6. Y. Y. Myint, K. Miyakawa, M. Naito, L. D. Shultz, Y. Oike, K. I. Yamamura, and K. Takahashi, Granulocyte/macrophage colonystimulating factor and interleukin-3 correct osteopetrosis in mice with osteopetrosis mutation. Am. J. Pathol. 154, 553 – 566 (1999). 7. E. Lagasse and I. L. Weissman, Enforced expression of Bcl-2 in monocytes rescues macrophages and partially reverses osteopetrosis in op/op mice. Cell 89, 102 – 131 (1997). 8. R. Pacifici, Cytokines, estrogen and postmenopausal osteoporosis. Endocrinology 139, 2659 – 2661 (1998). 9. J. A. Lorenzo, A. Naprta, Y. Rao, C. Alander, M. Glaccum, M. Widmer, G. Gronowicz, J. Kalinowski, and C. C. Pilbeam, Mice lacking the type I interleukin-1 receptor do not lose bone mass after ovariectomy. Endocrinology 139, 3022 – 3025 (1998). 10. Y. Abu-Amer, F. P. Ross, J. Edwards, and S. L. Teitelbaum, Lipopolysaccharide-stimulated osteoclastogenesis is mediated by tumor necrosis factor via its P55 receptor. J. Clin. Invest. 100, 1557 – 1565 (1997). 11. P. Ammann, R. Rizzoli, J. P. Bonjour, S. Bourrin, J. M. Meyer, P. Vasalli, and I. Garcia, Transgenic mice expressing soluble tumor necrosis factor-receptor are protected against bone loss caused by estrogen deficiency. J. Clin. Invest. 99, 1699 – 1702 (1997). 12. R. L. Jilka, G. Hagoc, G. Girasole, G. Passeri, D. C. Williams, J. S. Abrams, B. Boyce, H. Broxmeyer, and S. C. Manolagas, Increased
18.
19.
20.
21. 22.
23.
24. 25.
26.
27.
28.
29.
osteoclast development after estrogen loss: Mediation by interleukin6. Science 257, 88 – 91 (1992). V. Poli, R. Balena, E. Fattori, A. Markatos, M. Yamamoto, H. Tanaka, G. Cilberto, G. A. Rodan, and F. Constantini, Interleukin6 deficient mice are protected from bone loss caused by estrogen depletion. EMBO J. 13, 1189 – 1196 (1994). H. Kitamura, H. Kawata, F. Takahashi, Y. Higuchi, T. Furuichi, H. Ohkawa, Bone marrow neutrophilia and suppressed bone turnover in human interleukin-6 transgenic mice. Am. J. Pathol. 147, 1682 – 1692 (1995). R. Kitazawa, R. B. Kimble, J. L. Vannice, V. T. Kung, and R. Pacifici, Interleukin-1 receptor antagonist and tumor necrosis factor binding protein decrease osteoclast formation and bone resorption in ovariectomized mice. J. Clin. Invest. 94, 2397 – 2406 (1994). R. B. Kimble, A. B. Matayoshi, J. L. Vannice, V. T. Kung, C. Williams, and R. Pacifici, Simultaneous block of interleukin-1 and tumor necrosis factor is required to completely prevent bone loss in the early post-ovariectomy period. Endocrinology 136, 3054 – 3061 (1995). G. R. Mundy and T. J. Martin, Pathophysiology of skeletal complications of cancer. In “Physiology and Pharmacology of Bone: Handbook of Experimental Pharmacology” (G. R. Mundy and T. J. Martin, eds.), Chap. 18, pp. 641 – 671. Springer, Berlin. G. D. Roodman, N. Kurihara, Y. Ohsaki, A. Kukita, D. Hosking, A. Demulder, J. F. Smith, and F. R. Singer, Interleukin 6: A potential autocrine/paracrine factor in Paget’s disease of bone. J. Clin. Invest. 89, 46 – 52 (1992). R. J. O’Keefe, L. A. Teot, D. Singh, J. E. Puzas, R. N. Rosier, and D. G. Hicks, Osteoclasts constitutively express regulators of bone resorption: An immunohistochemical and in situ hybridization study. Lab. Invest. 76, 457 – 465 (1997). B. G. Mills and A. Fausta, Cytokines expressed in multinucleated cells: Paget’s disease and giant cell tumors versus normal bone. Calcif. Tissue Int. 61, 16 – 21 (1997). S. Reddy and G. D. Roodman, Control of osteoclast differentiation. Crit. Rev. Eukaryot. Gene Expr. 8, 1 – 17 (1998). T. J. Martin, E. Romas, and M. T. Gillespie, Interleukins in the control of osteoclast differentiation. Crit. Rev. Eukaryot. Gene Expr. 8, 107 – 123 (1998). D. Heyman, J. Guicheux, F. Gouin, N. Passuti, and G. Dalculsi, Cytokines, growth factors and osteoclasts. Cytokine 10, 155 – 168 (1998). G. D. Roodman, Cell biology of the osteoclast. Exp. Hematol. 27, 1229 – 1241 (1999). S. V. Reddy, S. Takahashi, M. Dallas, R. E. Williams, L. Neckers, and G. D. Roodman, IL-6 antisense deoxyoligonucleotides inhibit bone resorption by giant cells from human giant cell tumors of bone. J. Bone Miner. Res. 9, 753 – 757 (1994). R. D. Devlin, S. V. Reddy, sR. Savino, G. Ciliberto, and G. D. Roodman, IL-6 mediates the effects of IL-1 or TNF, but not PTHrP or 1,25(OH)2D3, on osteoclast-like cell formation in normal human bone marrow cultures. J. Bone Miner. Res. 13, 393 – 399 (1998). S. J. Choi, R. D. Devlin, C. Menaa, H. Chung, G. D. Roodman, and S. V. Reddy, Cloning and identification of human Sca as a novel inhibitor of osteoclast formation and bone resorption. J. Clin. Invest. 102, 1360 – 1368 (1998). C. Menaa, R. D. Devlin, S. V. Reddy, Y. Gazitt, S. J. Choi, and G. D. Roodman, Annexin II increases osteoclast formation by stimulating the proliferation of osteoclast precursors in human marrow cultures. J. Clin. Invest. 103, 1605 – 1613 (1999). C. Chenu, J. Pfeilschifter, G. R. Mundy, and G. D. Roodman, Transforming growth factor inhibits formation of osteoclast-like cells in long term human marrow cultures. Proc. Natl. Acad. Sci. USA 85, 5683 – 5687 (1988).
394 30. J. Pfeilschifter, S. Seyedin, and G. R. Mundy, Transforming growth factor inhibits bone resorption in fetal rat long bone cultures. J. Clin. Invest. 82, 380 – 385 (1988). 31. D. E. Hughes, A. Dai, J. C. Tiffee, H. H. Li, G. R. Mundy, and B. F. Boyce, Estrogen promotes apoptosis of murine osteoclasts mediated by TGF-. Nature Med. 2, 1132 – 1136 (1996). 32. H. Takai, M. Kanematsu, K. Yano, E. Tsuda, K. Higashio, K. Ikeda, K. Watanabe, and Y. Yamada, Transforming growth factor- stimulates the production of osteoprotegerin/osteoclastogenesis inhibitory factor by bone marrow stromal cells. J. Biol. Chem. 278, 27091 – 27096 (1998). 33. T. Murakami, M. Yamamoto, M. Yamamoto, K. Ono, M. Nishikawa, N. Nagata, K. Motoyoshi, and T. Akatsu, Transforming growth factor-1 increases mRNA levels of osteoclastogenesis inhibitory factor in osteoblastic/stromal cells and inhibits the survival of murine osteoclast-like cells. Biochem. Biophys. Res. Commun. 252, 747 – 752 (1998). 34. K. Fuller, B. Wong, S. Fox, Y. Choi, and T. J. Chambers, TRANCE is necessary and sufficient for osteoblast-mediated activation of bone resorption in osteoclasts. J. Exp. Med. 188, 997 – 1001 (1998). 35. M. R. Wani, K. Fuller, N. S. Kim, Y. Choi, and T. J. Chambers, Prostaglandin E2 cooperates with TRANCE in osteoclast induction from hemopoietic precursors: Synergistic activation of differentiation, cell spreading and fusion. Endocrinology 140, 1927 – 1935 (1999). 36. T. Suda, I. Nakamura, E. Jimi, and N. Takahashi, Regulation of osteoclast function. J. Bone Miner. Res. 12, 869 – 879 (1997). 37. D. M. Anderson, E. Maraskovsky, W. L. Billingsley, W. C. Dougall, M. E. Tomesko, E. R. Roux, M. C. Teepe, R. F. DuBose, D. Cosman, and L. Galibert, A homologue of the TNF receptor and its ligand enhance T-cell growth and dendritic cell function. Nature 390, 175 – 179 (1997). 38. B. R. Wong, J. Rho, J. Arron, E. Robinson, J. Orlinick, M. Chao, S. Kalachikov, E. Cayani, F. S. Bartlett, W. N. Frankel, S. Y. Lee, and Y. Choi, TRANCE is a novel ligand of the tumor necrosis factor receptor family that activates c-jun N-terminal kinase in T cells. J. Biol. Chem. 272, 25190 – 25194 (1997). 39. H. Yasuda, N. Shima, N. Nakagawa, S. I. Mochizuki, K. Yano, N. Fujise, Y. Sato, M. Goto, K. Yamaguchi, M. Kuriyama, T. Kanno, A. Murakami, E. Tsuda, T. Morinaga, and K. Higashio, Identity of osteoclastogenesis inhibitory factor (OCIF) and osteoprotegerin (OPG): A mechanism by which OPG/OCIF inhibits osteoclastogenesis in vitro. Endocrinology 139, 1329 – 1337 (1998). 40. W. S. Simonet, D. L. Lacey, E. Timms, H. L. Tan, M. J. Kelley, C. R. Dunstan, T. Burgess, R. Elliott, A. Colombero, G. Elliott, S. Scully, H. Hsu, J. Sullivan, N. Hawkins, E. Davy, C. Capparelli, A. Eli, Y. X. Qian, S. Kaufman, I. Sarosi, V. Shalhoub, G. Senaldi, J. Guo, J. Delaney, and W. J. Boyle, Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell 93, 165 – 176 (1998). 41. T. Suda, N. Takahashi, N. Udagawa, E. Jimi, M. T. Gillespie, and T. J. Martin, Modulation of osteoclast differentiation and function by the new members of the tumor necrosis factor receptor and ligand families. Endocr. Rev. 20, 345 – 357 (1999). 42. L. Lum, B. R. Wong, R. Josien, J. D. Becherer, H. ErdjumentBromage, J. Schonlondorff, P. Tempsta, Y. Choi, and C. P. Blobel, Evidence for a role of a tumor necrosis factor (TNF-) converting enzyme-like protease in shedding of TRANCE, a TNF family member involved in osteoclastogenesis and dendritic cell survival. J. Biol. Chem. 274, 13613 – 13618 (1999). 43. B. R. Wong, R. Josien, and Y. Choi, TRANCE is a TNF family member that regulates dendritic cell and osteoclast function. J. Leukocyte Biol. 65, 715 – 724 (1999). 44. T. Yoneda, I. Kato, L. F. Bonewald, W. H. Burgess, and G. R. Mundy, A novel cytokine with osteoclastopoietic activity. J. Periodontal Res. 28, 521 – 522 (1993).
GREGORY R. MUNDY ET AL. 45. M. Y. Lee, D. R. Eyre, and W. R. A. Osborne, Isolation of a murine osteoclast colony stimulating factor. Proc. Natl. Acad. Sci. USA 88, 8500 – 8504 (1991). 46. K. Tsukii, N. Shima, S. I. Mochizuki, K. Yamaguchi, M. Kinosaki, K. Yano, O. Shibata, N. Udagawa, H. Yasuda, T. Suda, and K. Higashio, Osteoclast differentiation factor mediates an essential signal for bone resorption induced by 1,25-dihydroxyvitamin D3, prostaglandin E2, or parathyroid hormone in the microenvironment of bone. Biochem. Biophys. Res. Commun. 246, 337 – 341 (1998). 47. B. O. Oyajobi, P. J. Williams, K. Traianedes, T. Yoneda, D. M. Anderson, and G. R. Mundy, Efficacy of a genetically-engineered soluble receptor activator of NF-B (RANK) fusion protein on bone resorption in vitro and in vivo. J. Bone Miner. Res. 14, 1124 (1999). 48. V. Shalhoub, J. Faust, W. J. Boyle, C. R. Dunstan, M. Kelley, S. Kaufman, S. Scully, G. Van, and D. L. Lacey, Osteoprotegerin and osteoprotegerin ligand effects on osteoclast formation from human peripheral blood mononuclear cell precursors. J. Cell. Biochem. 72, 251 – 261 (1999). 49. J. M. W. Quinn, J. Elliott, M. T. Gillespie, and T. J. Martin, A combination of osteoclast differentiation factor and macrophage-colony stimulating factor is sufficient for both human and mouse osteoclast formation in vitro. Endocrinology 139, 4424 – 4427 (1998). 50. K. Matsuzaki, N. Udagawa, N. Takahashi, K. Yamaguchi, H. Yasuda, N. Shima, T. Morinaga, Y. Toyama, Y. Yabe, K. Higashio, and T. Suda, Osteoclast differentiation factor (ODF) induces osteoclastlike cell formation in human peripheral blood mononuclear cell cultures. Biochem. Biophys. Res. Commun. 246, 199 – 204 (1998). 51. H. Yasuda, N. Shima, N. Nakagawa, K. Yamaguchi, M. Kinosaki, S. I. Mochizuki, A. Tomoyasu, K. Yano, M. Goto, A. Murakami, E. Tsuda, T. Morinaga, K. Higashio, N. Udagawa, N. Takahashi, and T. Suda, Osteoclast differentiation factor is a ligand for osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical to TRANCE/RANKL. Proc. Natl. Acad. Sci. USA 95, 3597 – 3602 (1998). 52. N. J. Horwood, J. Elliott, T. J. Martin, and M. T. Gillespie, Osteotropic agents regulate the expression of osteoclast differentiation factor and osteoprotegerin in osteoblastic stromal cells. Endocrinology, 139, 4743 – 4746 (1998). 53. S. K. Lee and J. A. Lorenzo, Parathyroid hormone stimulates TRANCE and inhibits osteoprotegerin messenger ribonucleic acid expression in murine bone marrow cultures: Correlation with osteoclast-like cell formation. Endocrinology 140, 3552 – 3561 (1999). 54. L. C. Hofbauer, D. L. Lacey, C. R. Dunstan, T. C. Spelsberg, B. L. Riggs, and S. Khosla, Interleukin-1 and tumor necrosis factor-, but not interleukin-6, stimulate osteoprotegerin ligand gene expression in human osteoblastic cells. Bone 25, 255 – 259 (1999). 55. E. Jimi, N. Akiyama, T. Tsurukai, N. Okahashi, K. Kobayashi, N. Udagawa, T. Nishihara, N. Takahashi, and T. Suda, Osteoclast differentiation factor (ODF) acts as a multifunctional regulator in murine osteoclast differentiation and function. J. Immunol. 162, 434 – 442 (1999). 56. T. L. Burgess, Y. X. Qian, S. Kaufman, B. D. Ring, G. Van, C. Capparelli, M. Kelley, H. Hsu, W. J. Boyle, C. R. Dunstan, S. Hu, and D. L. Lacey, The ligand for osteoprotegerin (OPGL) directly activates mature osteoclasts. J. Cell Biol., 145, 527 – 538 (1999). 57. D. L. Lacey, H. L. Tan, J. Lu, A. Eli, G. Van, F. Fletcher, W. J. Boyle, T. Juan, and A. Polverino, Osteoprotegerin ligand is essential for osteoclast survival in vitro and in vivo. J. Bone Miner. Res. 14(suppl), S179 (1999) [Abstract 1182]. 58. Y. Y. Kong, H. Yoshida, I. Sarosi, H. L. Tan, E. Timms, C. Cappparelli, S. Morony, A. J. Olivera-dos Santos, G. Van, A. Itie, W. Khoo, A. Wakeham, C. R. Dunstan, D. L. Lacey, T. W. Mak, W. J. Boyle, and J. M. Penniger, OPGL is a key regulator of osteoclastogenesis, lymphocyte development and lymph node organogenesis. Nature 397, 315 – 323 (1999).
CHAPTER 13 Cytokines and Bone Remodeling 59. B. O. Oyajobi, K. Traianedes, T. Yoneda, and G. R. Mundy, Expression of RANK Ligand (RANKL) by myeloma cells requires binding to bone marrow stromal cell via an 41-VCAM-1 interaction. Bone 23(Suppl.), S180 (1998). 60. N. Chikatsu, Y. Takeuchi, T. Kikuchi, E. Ogata, and T. Fujita, Adhesion of cancer cells to bone marrow stromal cells induces ODF/OPGL expression and osteoclast-like cell formation in vitro. 2nd International Conference on Cancer-Induced Bone Diseases, Davos, Switzerland. 61. N. J. Horwood, V. Kartsogiannis, M. H. C. Iam, E. Romas, J. M. W. Quinn, T. J. Martin, and M. T. Gillespie, Activated T cells are capable of inducing osteoclast formation: A mechanism for rheumatoid arthritis. Bone 23 (Suppl.), S214 (1998). 62. H. Hsu, D. L. Lacey, C. R. Dunstan, I. Solovyev, A. Colombero, E. Timms, H. L. Tan, G. Elliott, M. J. Kelley, I. Sarosi, L. Wang, X. Z. Xia, R. Elliott, L. Chiu, T. Black, S. Scully, C. Capparelli, S. Morony, G. Shimamoto, M. B. Bass, and W. J. Boyle, Tumor necrosis factor receptor family member RANK mediates osteoclast differentiation and activation induced by osteoprotegerin ligand. Proc. Natl. Acad. Sci. USA 96, 3540 – 3545 (1999). 63. B. Dougall, M. Glaccum, K. Charrier, K. Rohrbach, K. Brasel, T. De Smedt, E. Daro, J. Smith, M. E. Tomesko, C. R. Maliszewski, A. Armstrong, V. Shen, S. Bain, D. Cosman, D. Anderson, P. J. Morrisey, J. J. Peschon, and J. Schuh, RANK is essential for osteoclast and lymph node development. Genes Dev. 13, 2412 – 2424 (1999). 64. N. Nakagawa, M. Kinosaki, K. Yamaguchi, N. Shima, H. Yasuda, K. Yano, T. Morinaga, and K. Higashio, RANK is the essential signaling receptor for osteoclast differentiation factor in osteoclastogenesis. Biochem. Biophys. Res. Commun. 253, 395 – 400 (1998). 65. R. H. Arch, R. W. Gedrich, and C. B. Thompson, Tumor necrosis factor receptor-associated factors (TRAFs): A family of adapter proteins that regulate life and death. Genes Dev. 12, 2821 – 2830 (1998). 66. B. G. Darnay, J. Ni, P. A. Moore, and B. B. Aggarwal, Activation of NF-B by RANK requires tumor necrosis factor-receptor associated factor (TRAF) 6 and -inducing kinase: Identification of a novel TRAF6 interaction motif. J. Biol. Chem. 274, 7724 – 7731 (1999). 67. B. R. Wong, R. Josien, S. Y. Lee, M. Vologodskaia, R. M. Steinman, and Y. Choi, The TRAF family of signal transducers mediates the NF-B activation by the TRANCE receptor. J. Biol. Chem. 273, 28355 – 28359 (1998). 68. L. Galibert, M. E. Tomesko, D. M. Anderson, D. Cosman, and W. C. Dougall, The involvement of multiple tumor necrosis factor receptor (TNFR)-associated factors in signaling mechanisms of receptor activator of NF-B, a member of the TNFR superfamily. J. Biol. Chem. 273, 34120 – 34127 (1998). 69. M. A. Lomaga, W. C. Yeh, I. Sarosi, G. S. Duncan, C. Furlonger, A. Ho, S. Morony, C. Capparelli, G. Van, S. Kaufman, A. van der Heiden, A. Itie, A. Wakeham, W. Khoo, T. Sasaki, Z. Cao, J. M. Penninger, C. J. Page, D. L. Lacey, C. R. Dunstan, W. J. Boyle, D. V. Goeddel, and T. W. Mak, TRAF-6 deficiency results in osteopetrosis and defective interleukin-1, CD40, and LPS signaling. Genes Dev. 13, 1015 – 1024 (1999). 70. H. H. Kim, D. E. Lee, J. N. Shin, Y. S. Lee, Y. M. Jeon, C. H. Chung, J. Ni, B. S. Kwon, and Z. H. Lee, Receptor activator of NF-B recruits multiple TRAF family adaptors and activates c-jun N-terminal kinase. FEBS Lett. 443, 297 – 302 (1999). 71. Z. Cao, J. Xiong, M. Takeuchi, T. Kurma, and D. V. Goeddel, TRAF6 is a signal transducer for interleukin-1. Nature 383, 443 – 446 (1996). 72. V. Iotsova, J. Caamano, J. Loy, Y. Yang, A. Lewin, and R. Bravo, Osteopetrosis in mice lacking NF-B1 and NF-B2. Nature Med. 3, 1285 – 1289 (1997). 73. G. Franzoso, L. Carlson, L. Xing, L. Pljak, E. W. Shores, K. D. Brown, A. Leonardi, T. Tran, B. F. Boyce, and U. Siebenlist, Requirement for NF-B in osteoclast and B-cell development. Genes Dev. 11, 3482 – 3496 (1997).
395 74. E. Tsuda, M. Goto, S. I. Mochizuki, K. Yano, F. Kobayashi, T. Morinaga, and K. Higashio, Isolation of a novel cytokine from human fibroblasts that specifically inhibits osteoclastogenesis. Biochem. Biophys. Res. Commun. 234, 137 – 142 (1997). 75. W. S. Simonet, D. L. Lacey, C. R. Dunstan, M. Kelley, M. S. Chang, R. Luthy, H. Q. Nguyen, S. Wooden, L. Bennett, T. Boone, G. Shimamoto, M. DeRose, R. Elliott, A. Colombero, H. L. Tan, G. Trail, J. Sullivan, E. Davy, N. Bucay, L. Renshaw-Gegg, T. M. Hughes, D. Hill, W. Pattison, P. Campbell, S. Sander, G. Van, J. Tarpley, P. Derby, R. Lee, and W. J. Boyle, Osteoprotegerin: A novel secreted protein involved in the regulation of bone density. Cell 89, 309 – 319 (1997). 76. K. B. Tan, J. Harrop, M. Reddy, P. Young, J. Terrett, J. Emery, G. Moore, and A. Truneh, Characterization of a novel TNF-like ligand and recently described TNF ligand and TNF receptor superfamily genes and their constitutive and inducible expression in hematopoietic and non-hematopoietic cells. Gene 204, 35 – 46 (1997). 77. B. S. Kwon, S. Wang, N. Udagawa, V. Haridas, Z. H. Lee, K. K. Kim, K. O. Oh, J. Greene, Y. Li, J. Su, R. Gentsz, B. B. Aggarwal, and J. Ni, TR1, a new member of the tumor necrosis superfamily, induces fibroblast proliferation and inhibits osteoclastogenesis and bone resorption. FASEB J. 12, 845 – 854 (1998). 78. T. J. Yun, P. M. Chaudhary, G. L. Shu, J. K. Frazer, M. K. Ewings, S. M. Schwartz, V. Pascual, L. E. Hood, and E. A. Clark, OPG/ FDCR-1, a TNF receptor family member, is expressed in lymphoid cells and is up-regulated by ligating CD40. J. Immunol. 161, 6113 – 6121 (1998). 79. T. Morinaga, N. Nakagawa, H. Yasuda, E. Tsuda, and K. Higashio, Cloning and characterization of the gene encoding human osteoprotegerin/osteoclastogenesis-inhibitory factor. Eur. J. Biochem. 254, 685 – 691 (1998). 80. J. G. Emery, P. McDonnell, M. B. Burke, K. C. Deen, S. Lyn, C. Silverman, E. Dul, E. R. Appelbaum, C. Eichman, R. DiPrinizio, I. E. Dodds, James, M. Rosenberg, J. C. Lee, and P. R. Young, Osteoprotegerin is a receptor for the cytotoxic ligand TRAIL. J. Biol. Chem. 273, 14363 – 14367 (1998). 81. H. Kawaguchi, C. Manabe, C. Miyaura, H. Chikuda, K. Nakamura, and M. Kuro-o, Independent impairment of osteoblast and osteoclast differentiation in klotho mouse exhibiting low-turnover osteopenia. J. Clin. Invest. 104, 229 – 237 (1999). 82. K. Yano, E. Tsuda, N. Washida, F. Kobayashi, M. Goto, A. Harada, K. Ikeda, K. Higashio, and Y. Yamada, Immunological characterization of circulating osteoprotegerin/osteoclastogenesis inhibitory factor: Increased serum concentrations in post-menopausal osteoporosis. J. Bone Miner. Res. 14, 518 – 527 (1999). 83. N. Bucay, I. Sarosi, C. R. Dunstan, S. Morony, J. Tarpley, C. Capparelli, S. Scully, H-L. Tan, W. Xu, D. L. Lacey, W. J. Boyle, and W. S. Simonet, Osteoprotegerin-deficient mice develop early onset osteoporosis and arterial calcification. Genes Dev. 12, 1260 – 1268 (1998). 84. A. Mizuno, N. Amizuka, K. Irie, A. Murakami, N. Fujise, T. Kanno, Y. Sato, N. Nakagawa, H. Yasuda, S. A. Mochizuki, A. Tomoyasu, K. Yano, N. Shima, N. Washida, E. Tsuda, T. Morinaga, K. Higashio, and H. Ozawa, Severe osteoporosis in mice lacking osteoclastogenesis inhibitory factor/osteoprotegerin. Biochem. Biophys. Res. Commun. 247, 610 – 615 (1998). 85. H. Brandstrom, K. B. Jonsson, C. Ohlsson, O. Vidal, S. Ljunghall, and O. Ljunggren, Regulation of osteoprotegerin mRNA levels by prostaglandin E2 in human bone marrow stroma cells. Biochem. Biophys. Res. Commun. 248, 338 – 341 (1998). 86. L. C. Hofbauer, F. Gori, B. L. Riggs, D. L. Lacey, C. R. Dunstan, T. C. Spelsberg, and S. Khosla, Stimulation of osteoprotegerin ligand and inhibition of osteoprotegerin production by glucocorticoids in human osteoblastic lineage cells: Potential paracrine mechanisms of glucocorticoid-induced osteoporosis. Endocrinology 140, 4382 – 4389 (1999).
396 87. O. N. A. Vidal, H. Brandstrom, K. B. Jonsson, and C. Ohlsson, Osteoprotegerin mRNA is expressed in primary human osteoblast-like cells: Downregulation by glucocorticoids. J. Endocrinol. 159, 191 – 195 (1998). 88. O. N. A. Vidal, K. Sjogren, B. I. Eriksson, O. Ljunggren, and C. Ohlsson, Osteoprotegerin mRNA is increased by interleukin-1 in the human osteosarcoma cell line MG-63 and in human osteoblastlike cells. Biochem. Biophys. Res. Commun. 248, 696 – 700 (1998). 89. Y. Hakeda, Y. Kobayashi, K. Yamaguchi, H. Yasuda, E. Tsuda, K. Higashio, T. Miyata, and M. Kumegawa, Osteoclastogenesisinhibitory factor (OCIF) directly inhibits bone-resorbing activity of isolated mature osteoclasts. Biochem. Biophys. Res. Commun. 251, 796 – 801 (1998). 90. T. Akatsu, T. Murakami, M. Nishikawa, K. Ono, N. Shinomiya, E. Tsuda, S.-I. Mochizuki, K. Yamaguchi, M. Kinosaki, K. Higashio, M. Yamamoto, and N. Nagata, Osteoclastogenesis-inhibitory factor suppresses osteoclast survival by interfering in the interaction of stromal cells with osteoclasts. Biochem. Biophys. Res. Commun. 250, 229 – 234 (1998). 91. T. Akatsu, T. Murakami, K. Ono, M. Nishikawa, E. Tsuda, S-I. Mochizuki, N. Fujise, K. Higashio, K. Motoyoshi, M. Yamamoto, and N. Nagata, Osteoclastogenesis-inhibitory factor exhibits hypocalcemic effects in normal mice and in hypercalcemic nude mice carrying tumors associated with humoral hypercalcemia of malignancy. Bone 23, 495 – 498 (1998). 92. M. Yamamoto, T. Murakami, M. Nishikawa, E. Tsuda, S. Mochizuki, K. Higashio, T. Akatsu, K. Motoyoshi, and N. Nagata, Hypocalcemic effect of osteoclastogenesis inhibitory factor/osteoprotegerin in the thyroparathyroidectomized rat. Endocrinology 139, 4012 – 4015 (1998). 93. S. Morony, C. Caparelli, R. Lee, G. Shimamoto, T. Boone, D. L. Lacey, and C. R. Dunstan, A chimeric form of osteoprotegerin inhibits hypercalcemia and bone resorption induced by IL-1, TNF-, PTH, PTHrP and 1,25(OH)2D3. J. Bone Miner. Res. 14, 1478 – 1485 (1999). 94. B. R. MacDonald, G. R. Mundy, S. Clark, E. A. Wang, T. J. Kuehl, E. R. Stanley, and G. D. Roodman, Effects of human recombinant CSF-GM and highly purified CSF-1 on the formation of multinucleated cells with osteoclast characteristics in long term bone marrow cultures. J. Bone Miner. Res. 1, 227 – 233 (1986). 95. S. Tanaka, N. Takahashi, N. Udagawa, T. Tamura, T. Akatsu, E. R. Stanley, T. Kurokawa, and T. Suda, Macrophage colony stimulating factor is indispensable for both proliferation and differentiation of osteoclast progenitors. J. Clin. Invest. 91, 257 – 263 (1993). 96. U. Sarma and A. M. Flanagan, Macrophage colony stimulating factor induces substantial osteoclast generation and bone resorption in human bone marrow cultures. Blood 88, 2531 – 2540 (1996). 97. C. S. Lader and A. M. Flanagan, Prostaglandin E2, interleukin 1 and tumor necrosis factor- increase human osteoclast formation and bone resorption in vitro. Endocrinology 139, 3157 – 3164 (1998). 98. H. Amano, S. Yamada, and R. Felix, Colony-stimulating factor-1 stimulates the fusion process in osteoclasts. J. Bone Miner. Res. 13, 846 – 843 (1998). 99. E. Jimi, I. Nakamura, L. T. Duong, T. Ikebe, N. Takahashi, G. A. Rodan, and T. Suda, Interleukin 1 induces multinucleation and boneresorbing activity of osteoclasts in the absence of osteoblast/stromal cells. Exp. Cell Res. 247, 84 – 93 (1999). 100. E. Jimi, T. Shuto, and T. Koga, Macrophage colony stimulating factor and interleukin 1 maintain the survival of osteoclast-like cells. Endocrinology 136, 808 – 811 (1995). 101. K. Fuller, J. M. Owens, C. J. Jagger, A. Wilson, and R. Moss, Macrophage colony-stimulating factor stimulates survival and chemotactic behavior in isolated osteoclasts. J. Exp. Med. 178, 1733 – 1744 (1993).
GREGORY R. MUNDY ET AL. 102. H. Yoshida, S. I. Hayashi, T. Kunisada, M. Ogawa, S. Nishikawa, H. Okamura, T. Sudo, L. D. Shultz, and S. I. Nishikawa, The murine mutation osteopetrosis is in the coding region of the macrophage colony stimulating factor gene. Nature 345, 442 – 443 (1990). 103. M. L. Sherman, B. L. Weber, R. Datta, and D. W. Kufe, Transcriptional and posttranscriptional regulation of macrophage-specific colony stimulating factor gene expression by tumor necrosis factor. J. Clin. Invest. 85, 442 – 447 (1990). 104. J. H. Falkenburg, M. A. Harrington, R. A. de Paus, W. K. Walsh, R. Daub, J. E. Landegent, and H. E. Broxmeyer, Differential transcriptional and posttranscriptional regulation of gene expression of the colony-stimulating factors by interleukin-1 and fetal bovine serum in murine fibroblasts. Blood 78, 658 – 665 (1991). 105. S. D. Isaacs, X. Fan, D. Fan, H. Gewant, T. C. Murphy, P. Farmer, W. R. Taylor, M. S. Nanes, and J. Rubin, Role of NF-B in the regulation of macrophage colony stimulating factor by tumor necrosis factor- in 5T2 bone stromal cells. J. Cell. Physiol. 179, 193 – 200 (1999). 106. G. Q. Yao, B. H. Sun, E. E. Hammond, E. N. Spencer, M. C. Horowitz, K. L. Insogna, and E. C. Weir, The cell-surface form of colony-stimulating factor-1 is regulated by osteotropic agents and supports formation of multinucleated osteoclast-like cells. J. Biol. Chem. 273, 4119 – 4128 (1998). 107. R. B. Kimble, S. Srivastava, F. P. Ross, A. Matayoshi, and R. Pacifici, Estrogen deficiency increases the ability of stromal cells to support osteoclastogenesis via an IL-1 and TNF mediated stimulation of M-CSF production. J. Biol. Chem. 271, 28890 – 28897 (1996). 108. S. Srivastava, M. N. Weitzmann, R. B. Kimble, M. Rizzo, M. Zahner, J. Milbrandt, F. P. Ross, and R. Pacifici, Estrogen blocks M-CSF gene expression and osteoclast formation by regulating phosphorylation of Egr-1 and its interaction with Sp-1. J. Clin. Invest. 102, 1850 – 1859 (1998). 109. P. Soriano, C. Montgomery, R. Geske, and A. Bradley, Targeted disruption of the c-src proto-oncogene leads to osteopetrosis in mice. Cell 64, 693 – 702 (1991). 110. B. F. Boyce, T. Yoneda, C. Lowe, P. Soriano, and G. R. Mundy, Requirement of pp60c-src expression of osteoclasts to form ruffled borders and resorb bone. J. Clin. Invest. 90, 1622 – 1627 (1992). 111. K. L. Insogna, M. Sahni, A. B. Grey, S. Tanaka, W. C. Horne, L. Neff, M. Mitnick, J. B. Levy, and R. Baron, Colony stimulating factor-1 induces cytoskeletal re-organization and c-src dependent tyrosine phosphorylation of selected cellular proteins in rodent osteoclasts. J. Clin. Invest. 100, 2476 – 2485 (1997). 112. A. Teti, A. Taranta, S. Migliaccio, A. Degiorgi, E. Santandrea, I. Villanova, T. Faraggiana, M. Chellaiah, and K. A. Hruska, Colony stimulating factor-1-induced osteoclast spreading depends on substrate and requires vitronectin receptor and c-src proto-oncogene. J. Bone Miner. Res. 13, 50 – 58 (1998). 113. R. L. Lees, and J. H. M. Heersche, Macrophage colony stimulating factor increases bone resorption in dispersed osteoclast cultures by increasing osteoclast size. J. Bone Miner. Res. 14, 937 – 945 (1999). 114. M. Gowen, M. C. Meikle, and J. J. Reynolds, Stimulation of bone resorption in vitro by a non-prostanoid factor released by human monocytes in culture. Biochem. Biophys. Acta 762, 471 – 474 (1983). 115. M. Gowen, and G. R. Mundy, Actions of recombinant interleukin-1, interleukin-2 and interferon gamma on bone resorption in vitro. J. Immunol. 136, 2478 – 2482 (1986). 116. M. Gowen, G. Nedwin, and G. R. Mundy, Preferential inhibition of cytokine stimulated bone resorption by recombinant interferon gamma. J. Bone Miner. Res. 1, 469 – 474 (1986). 117. B. F. Boyce, T. B. Aufdemorte, I. R. Garrett, A. J. P. Yates, and G. R. Mundy, Effects of interleukin-1 on bone turnover in normal mice. Endocrinology 125, 1142 – 1150 (1989).
CHAPTER 13 Cytokines and Bone Remodeling 118. B. F. Boyce, A. J. P. Yates, and G. R. Mundy, Bolus injections of recombinant human interleukin-1 cause transient hypocalcemia in normal mice. Endocrinology 125, 2780 – 2783 (1989). 119. J. Pfeilschifter, C. Chenu, A. Bird, G. R. Mundy, and G. D. Roodman, Interleukin-1 and tumor necrosis factor stimulate the formation of human osteoclastlike cells in vitro. J. Bone Miner. Res. 4, 113 – 118 (1989). 120. E. Thomassen, T. A. Bird, B. R. Renshaw, M. K. Kennedy, and J. E. Sims, Binding of interleukin-18 to the interleukin-1 receptor homologous receptor IL-1RP1 leads to activation of signaling pathways similar to those used by interleukin-1. J. Interferon Cytokine Res. 18, 1077 – 1088 (1998). 121. H. L. Uy, T. A. Guise, J. DeLaMata, S. D. Taylor, B. M. Story, M. R. Dallas, B. F. Boyce, G. R. Mundy, and G. D. Roodman, Effects of PTH-rP and PTH on osteoclasts and osteoclast precursors in vivo. Endocrinology 136, 3207 – 3212 (1995). 122. C. A. Dinarello, Biologic basis for interleukin-1 in disease. Blood 77, 1627 – 1652 (1996). 123. R. M. Hinson, J. A. Williams, and E. Shacter, Elevated interleukin-6 is induced by prostaglandin E2 in a murine model of inflammation: Possible role of cyclooxygenase-2. Proc. Natl. Acad. Sci. USA 93, 4885 – 4890 (1996). 124. M. Gowen, D. D. Wood, G. R. Mundy, and R. G. G. Russell, Studies on the actions of interleukin-1 on bone metabolism: IL-1 stimulation of bone cell proliferation, and inhibition of IL-1-induced bone resorption by interferon gamma. Brit. J. Rheum. 24, 147 – 149 (1985). 125. D. Smith, M. Gowen, and G. R. Mundy, Effects of interferon gamma and other cytokines on collagen synthesis in fetal rat bone cultures. Endocrinology 120, 2494 – 2499 (1987). 126. K. Sato, Y. Fujii, K. Kasono, M. Ozawa, H. Imamura, Y. Kanaji, H. Kurosawa, T. Tsushima, and K. Shizume, Parathyroid hormonerelated protein and interleukin-1 synergistically stimulate bone resorption in vitro and increase the serum calcium concentration in mice in vivo. Endocrinology 124, 2172 – 2178 (1989). 127. M. Kawano, H. Tanaka, H. Ishikawa, M. Nobuyoshi, K. Iwato, H. Asaoku, O. Tanabe, and A. Kuramoto, Interleukin-1 accelerates autocrine growth of myeloma cells through interleukin-6 in human myeloma. Blood 73, 2145 – 2148 (1989). 128. R. Pacifici, J. L. Vannice, L. Rifas, and R. B. Kimble, Monocytic secretion of interleukin-1 receptor antagonist in normal and osteoporotic women: Effects of menopause and estrogen/progesterone therapy. J. Clin. Endocrin. Metab. 77, 1135 – 1141 (1993). 129. R. B. Kimble, J. L. Vannice, D. C. Bloedow, R. C. Thompson, W. Hopfer, V. Kung, C. Brownfield, and R. Pacifici, Interleukin-1 receptor antagonist decreases bone loss and bone resorption in ovariectomized rats. J. Clin. Invest. 93, 1959 – 1967 (1994). 130. C. A. O’Brien, I. Gubrij, S. C. Lin, R. L. Saylors, and S. C. Manolagas, STAT3 activation in stromal/osteoblastic cells is required for induction of the receptor activator of NF-B ligand and stimulation of osteoclastogenesis by gp 130-utilizing cytokines or interleukin-1 but not 1,25-dihydroxyvitamin D3 or parathyroid hormone. J. Biol. Chem. 274, 19310 – 19308 (1999). 131. L. C. Hofbauer, B. L. Riggs, D. L. Lacey, C. R. Dunstan, T. C. Spelsberg, and S. Khosla, Glucocorticoids stimulate osteoprotegerin ligand and inhibit osteoprotegerin production in human osteoblastic lineage cells: Potential paracrine mechanisms of glucocorticoidinduced osteoporosis. 18th Endocrine Society Meeting, San Diego, CA. 132. E. Jimi, I. Nakamura, T. Ikebe, N. Akiyama, N. Takahashi, and T. Suda, Activation of NF-B is involved in the survival of osteoclasts promoted by interleukin 1. J. Biol. Chem. 273, 8799 – 8805 (1998). 133. E. Jimi, T. Ikebe, N. Takahashi, N. Hirata, T. Suda, and T. Koga, Interleukin 1 activates an NF-B-like factor in osteoclast-like cells. J. Biol. Chem. 271, 4605 – 4608 (1996).
397 134. D. R. Bertolini, G. E. Nedwin, T. S. Bringman, D. D. Smith, and G. R. Mundy, Stimulation of bone resorption and inhibition of bone formation in vitro by human tumour necrosis factors. Nature 319, 516 – 518 (1986). 135. R. A. Johnson, B. F. Boyce, G. R. Mundy, and G. D. Roodman, Tumors producing human TNF induce hypercalcemia and osteoclastic bone resorption in nude mice. Endocrinology 124, 1424 – 1427 (1989). 136. T. Yoneda, M. M. Alsina, J. B. Chavez, L. Bonewald, R. Nishimura, and G. R. Mundy, Evidence that tumor necrosis factor plays a pathogenetic role in the paraneoplastic syndromes of cachexia, hypercalcemia, and leukocytosis in a human tumor in nude mice. J. Clin. Invest. 87, 977 – 985 (1991). 137. M. Sabatini, A. J. Yates, R. Garrett, J. Chavez, J. Dunn, L. Bonewald, and G. R. Mundy, Increased production of tumor necrosis factor by normal immune cells in a model of the humoral hypercalcemia of malignancy. Lab. Invest. 63, 676 – 682 (1990). 138. M. Sabatini, J. Chavez, G. R. Mundy, and L. F. Bonewald, Stimulation of tumor necrosis factor release from monocytic cells by the A375 human melanoma via granulocyte-macrophage colony stimulating factor. Cancer Res. 50, 2673 – 2678 (1990). 139. I. R. Garrett, B. G. M. Durie, G. E. Nedwin, A. Gillespie, T. Bringman, M. Sabatini, D. R. Bertolini, and G. R. Mundy, Producion of lymphotoxin, a bone-resorbing cytokine, by cultured human myeloma cells. N. Engl. J. Med. 317, 526 – 532 (1987). 140. J. H. M. Feyen, P. Elford, F. E. Dipadova, and U. Trechsel, Interleukin-6 is produced by bone and modulated by parathyroid hormone. J. Bone Miner. Res. 4, 633 – 638 (1989). 141. Y. Ishimi, C. Miyaura, C. H. Jin, T. Akatsu, E. Abe, Y. Nakamura, A. Yamaguchi, S. Woshiki, T. Matsuda, and T. Hirano, IL-6 is produced by osteoblasts and induces bone resorption. J. Immunol. 145, 3297 – 3303 (1990). 142 J. Gilles, M. Qiao, G. R. Mundy, and L. F. Bonewald, Preferential stimulation of interleukin-6 production in bone by cytokines compared with systemic factors. J. Bone Miner. Res. 9, B433 (1994) [Abstract]. 143. J. De La Mata, H. Uy, T. A. Guise, B. Story, B. F. Boyce, G. R. Mundy, and G. D. Roodman, IL-6 enhances hypercalcemia and bone resorption mediated by PTH-rP in vivo. J. Clin. Invest. 95, 2846 – 2852 (1995). 144. K. Black, I. R. Garrett, and G. R. Mundy, Chinese hamster ovarian cells transfected with the murine interleukin-6 gene cause hypercalcemia as well as cachexia, leukocytosis and thrombocytosis in tumor-bearing nude mice. Endocrinology 128, 2657 – 2659 (1991). 145. T. Tamura, N. Udagawa, N. Takahashi C. Miyaura, S. Tanaka, Y. Yamada, Y. Koishihara, Y. Ohsugi, K. Kumaki, and T. Taga, Soluble interleukin-6 receptor triggers osteoclast formation by interleukin-6 Proc. Natl. Acad. Sci. USA 90, 11924 – 11928, (1993). 146. J. P. Gaillard, R. Bataille, H. Brailly, C. Zuber, K. Yasukawa, M. Attal, N. Maruo, T. Tage T. Kishimoto, and B. Klein, Increased and highly stable levels of functional soluble interleukin-6 receptor in sera of patients with monoclonal gammopathy. Eur. J. Immunol. 23, 820 – 824, (1993). 147. N. Udagawa, N. Takahashi, T. Katagiri, T. Tamura S. Wada, D. M. Findlay, T. J. Martin, H. Hirota, T. Taga, T. Kishimoto, and T. Suda, Interleukin (IL)-6 induction of osteoclast differentiation depends on IL-6 receptors expressed on osteoblastic cells but not on osteoclast progenitors. J. Exp. Med. 182, 1461 – 1468, (1995). 148. K. Kawasaki, Y. H. Gao, S. Yokose, Y. Kaji, T. Nakamura, T. Suda, K. Yoshida, T. Taga, T. Kishimoto, H. Kataoka, T. Yuasa, H. Norimatsu, and A. Yamaguchi, Osteoclasts are present in gp 130-deficient mice. Endocrinology 138, 4959 – 4965 (1997). 149. N. Nishimoto, Y. Shima, K. Yoshizaki, and T. Kishimoto, Myeloma biology and therapy: Present status and future developments. Hematol. Oncol. Clinics North Am. 11, 159 – 172 (1997).
398 150. B. Klein, X. G. Zhang, Z. Y. Lu, and R. Bataille, Interleukin-6 in human multiple myeloma. Blood 74, 863 – 872 (1995). 151. G. Gutierrez, G. Rossini, J. Chavez, T. Nishihara, T. Yoneda, and G. R. Mundy, Identification, purification and expression of a naturally-occurring inhibitor of interleukin-6 (IL-6). J. Bone Miner. Res. 9, 79 (1994) [Abstract]. 152. H. P. Gerber, T. H. Vu, A. M. Ryan, J. Kowalski, Z. Werb, and N. Ferrara, VEGF couples hypertrophic cartillage remodeling, ossification and angiogenesis during endochondral bone formation. Nature Med. 5, 623 – 628 (1999). 153. S. Niida, M. Kaku, H. Amano, H. Yoshida, H. Kataoka, S. Nishikawa, K. Tanne, N. Maeda, S. I. Nishikawa and H. Kodama, Vascular endothelial growth factor can substitute for macrophage colony stimulating factor in the support of osteoclastic bone resorption. J. Exp. Med. 190, 293 – 298 (1999). 154. Y. Ogata, A. Kukita, T. Kukita M. Komine, A. Miyahara, S. Miyazaki, and O. Kohashi, A novel role of IL-15 in the development of osteoclasts: Inability to replace its activity with IL-2. J. Immunol. 162, 2754 – 2760 (1999). 155. I. B. Mclnnes, J. AI-Mughales, M. Field, B. P. Leung, F. P. Huang R. Dixon, R. D. Sturrock, P. C. Wilkinson, and F. Y. Lew, The role of interleukin-15 in T-cell migration and activation in rheumatoid arthritis. Nature Med. 2, 175 – 178 (1996). 156. N. Udagawa, N. J. Horwood, J. Elliot, A. Mackay, J. Owens, H. Okamura, M. Kurimoto, T. J. Chambers, T. J. Martin, and M. T. Gillespie, Interleukin 18 (interferon- inducing factor) is produced by osteoblasts and acts via granulocyte-macrophage colony stimulating factor and not via interferon- to inhibit osteoclast formation. J. Exp. Med. 185, 1005 – 1017 (1997). 157. R. L. Van Bezooijen, H. C. M. Farih-Sips, S. E. Papapoulos, and C. W. G. M. Lowik, Interleukin 17: A new bone acting cytokine in vitro. J. Bone Miner. Res. 14, 1513 – 1521 (1999). 158. S. Kotake, N. Udagawa, N. Takahashi, K. Matzuzaki, K. Itoh, S. Lyama, S. Saito K. Inoue, N. Kamatani, M. T. Gillespie, T. J. Martin, and T. Suda, IL-17 in synovial fluids from patients with rheumatoid arthritis is a potent stimulator of osteoclastogenesis. J. Clin. Invest. 103, 1345 – 1352 (1999). 159. N. J. Horwood, N. Udagawa, J. Elliott, D. Grail, H. Okamura, M. Kurimoto, A. R. Dunn, T. J. Martin and M. T. Gillespie, Interleukin 18 inhibits osteoclast formation via T cell production of granulocyte macrophage colony stimulating factor. J. Clin. Invest. 101, 595 – 603 (1998). 160. E. Sigal, The molecular biology of mammalian arachidonic acid metabolism. Am. J. Physiol. 260, L13 – L28 (1991). 161. C. N. Serhan, J. Z. Haeggstrom, and C. C. Leslie, Lipid mediator networks in cell signaling: Update and impact of cytokines. FASEB J. 10, 1147 – 1158 (1996). 162. J. C. Otto and W. L. Smith, Prostaglandin endoperoxide synthases-1 and -2. J. Lipid Mediat. Cell Signal. 12, 139 – 156 (1995). 163. M. Peters-Golden, Cell biology of the 5-lipoxygenase pathway. Am. J. Respir. Crit. Care. Med. 157, S227 – S232 (1998). 164. K. Schoonjans, B. Staels, and J. Auwerx, The peroxisome proliferator activated receptors (PPARS) and their effects on lipid metabolism and adipocyte differentiation. Biochim. Biophys. Acta 1302, 93 – 109 (1996). 165. B. M. Forman, J. Chen, and R. M. Evans, Hypolipidemic drugs, polyunsaturated fatty acids, and eicosanoids are ligands for peroxisome proliferator-activated receptors alpha and delta. Proc. Natl. Acad. Sci. USA 94, 4312 – 4317 (1997). 166. P. R. Devchand, H. Keller, J. M. Peters, M. Vazquez, F. J. Gonzalez, and W. Wahli, The PPARalpha-leukotriene B4 pathway to inflammation control. Nature 384, 39 – 43 (1996). 167. D. Chen, X. Ji, M. A. Harris, J. Q. Feng, G. Karsenty, A. J. Celeste, V. Rosen, G. R. Mundy, and S. E. Harris, Differential roles for bone morphogenetic protein (BMP) receptor type IB and IA in dif-
GREGORY R. MUNDY ET AL.
168.
169.
170.
171.
172.
173.
174.
175.
176.
177.
178.
179.
180.
181.
182.
183.
184.
ferentiation and specification of mesenchymal precursor cells to osteoblast and adipocyte lineages. J. Cell Biol. 142, 295 – 305 (1998). D. D. Diascro, Jr. R. L. Vogel, T. E. Johnson, K. M. Witherup, S. M. Pitsenberger, S. J. Rutledge D. J. Prescott, G. A. Rodan, and A. Schmidt, High fatty acid content in rabbit serum is responsible for the differentiation of osteoblasts into adipocyte-like cells. J. Bone Miner. Res. 13, 96 – 106 (1998). A. Lchikawa, Molecular biology of prostaglandin E receptors – Expression of multi-function by PGE receptor subtypes and isoforms Nippon Rinsho 56, 1813 – 1818 (1998). M. Suda, K. Tanaka, K. Natsui, T. Usui, I. Tanaka, M. Fukushima, C. Shigeno, J. Konishi, S. Narumiya, A. Ichikawa, and N. Nakao, Prostaglandin E receptor subtypes in mouse osteoblastic cell line. Endocrinology 137, 1698 – 1705 (1996). M. Weinreb, A. Grosskopf, and N. Shir, The anabolic effect of PGE2 in rat bone marrow cultures is mediated via the EP4 receptor subtype. Am. J. Physiol. 276, E376 – E383 (1999). K. Ono, T. Akatsu, T. Murakami, M. Nishikawa, M. Yamamoto, N. Kugai, K. Motoyoshi, and N. Nagata, Important role of EP4, a subtype of prostaglandin (PG) E receptor, in osteoclast-like cell formation form mouse bone marrow cells induced by PGE2. J. Endocr. 158, R1 – R5 (1998). M. Mano, Y. Hakeda, H. Mano, H. Kiyomura, and M. Kumegawa, Prostaglandin E2 directly inhibits osteoclastic bone resorption through EP2 receptor. Bone 23, (Suppl.) W066 (1998). T. Yokomizo, T. Izumi, K. Chang, Y. Takuwa, and A. Shimizu, G-protein-coupled receptor for leukotriene B4 that mediates chemotaxis. Nature 387, 620 – 624 (1997). K. R. Lynch, G. P. O’Neill, Q. Liu, D. S. Im, N. Sawyer, K. M. Metters, N. Coulombe M. Abramovitz, D. J. Figueroa, Z. Zeng, B. M. Connolly, C. Bai, C. P. Austin, A. Chateauneuf, R. Stocco, G. M. Greig, S. Kargman, S. B. Hooks, E. Hosfield, D. L. Williams, Jr. A. W. Ford-Hutchinson, C. T. Caskey, and J. F. Evans, Characterization of the human cysteinyl leukotriene CysL T1 receptor. Nature 399, 789 – 793 (1999). H. R. Herschman, Regulation of prostaglandin synthase-1 and prostaglandin synthase-2. Cancer Metastasis Rev. 13, 241 – 256 (1994). B. P. Kennedy, C. C. Chan, S. A. Culp, and W. A. Cromlish, Cloning and expression of rat prostaglandin endoperoxide synthase (cyclooxygenase)-2 cDNA. Biochem. Biophys. Res. Commun. 197, 494 – 500 (1993). Y. Harada, K. Hatanaka, M. Kawamura, M. Saito, M. Ogino, M. Majima, T. Ohno, K. Ogino, K. Yamamoto, Y. Taketani, S. Yamamoto, and M. Katori Role of prostaglandin H synthase-2 in prostaglandin E2 formation in rat carrageenin-induced pleurisy. Prostaglandins 51, 19 – 33 (1996). R. N. Dubois, S. B. Abramson, L. Crofford, R. A. Gupta, L. S. Simon, L. B. Van De Putte, and P. E. Lipsky, Cyclooxygenase in biology and disease. FASEB J. 12, 1063 – 1067 (1998). D. A. Kujubu and H. R. Herschman, Dexamethasone inhibits mitogen induction of the TIS 10 prostaglandin synthase/cyclooxygenase gene. J. Biol. Chem. 267, 7991 – 7994 (1992). J. L. Masferrer, K. Seibert, B. Zweifel, and P. Needleman, Endogenous glucocorticoids regulate an inducible cyclooxygenase enzyme. J. Biol. Chem. 89, 3917 – 3921 (1992). H. H. Versteeg, P. M. van Bergen en Henegouwen, S. J. van Deventer, and M. P. Peppelenbosch, Cyclooxygenase-dependent signalling: Molecular events and consequences. FEBS Lett. 445, 1 – 5 (1999). E. J. Canalis, Clinical review 83: Mechanisms of glucocorticoid action in bone: Implications to glucocorticoid-induced osteoporosis. Clin. Endocrinol. Metab. 81, 3441 – 3447 (1996). S. C. Miller and S. C. Marks, Effects of prostaglandins on the skeleton. Clin. Plast. Surg. 21, 393 – 400 (1994).
CHAPTER 13 Cytokines and Bone Remodeling 185. H. Kawaguchi, C. C. Pilbeam, J. R. Harrison, and L. G. Raisz, The role of prostaglandins in the regulation of bone metabolism. Clin. Orthop. 313, 36 – 46 (1995). 186. D. C. Klein and L. G. Raisz, Prostaglandins: Stimulation of bone resorption in tissue culture. Endocrinology 86, 1436 – 1440 (1970). 187. K. Ueda, A. Saito, H. Nakano, M. Aoshima, M. Yokata, R. Muraoka, and T. Iwaya, Cortical hypersteosis following long-term administration of prostaglandin E2 in infants with cyanotic congenital heart disease. J. Pediatr. 97, 834 – 836 (1980). 188. R. W. Norrdin, W. S. Jee, and W. B. High, The role of prostaglandins in bone in vivo. Prostaglandins Leukot. Essent. Fatty Acids 41, 139 – 149 (1990). 189. T. Nagata, K. Kaho, S. Nishikawa, H. Shinohara, Y. Wakano, and H. Ishida, Effect of prostaglandin E2 on mineralization of bone nodules formed by fetal rat calvarial cells. Calcif. Tissue Int. 55, 451 – 457 (1994). 190. W. S. Jee, B. Y. Lin, Y. F. Ma, and H. Z. Ke, Extra cancellous bone induced by combined prostaglandin E2 and risedronate administration is maintained after their withdrawal in older female rats. J. Bone Miner. Res. 10, 963 – 970 (1995). 191. W. S. Jee and Y. F. Ma, The in vivo anabolic actions of prostaglandins in bone. Bone 21, 297 – 304 (1997). 192. Y. F. Ma, Z. Pan, W. S. Jee, C. H. Lin, H. H. Liang, H. Chen, S. Pun, and X. J. Li, Intermittent on/off prostaglandin E2 and risedronate are equally anabolic as daily PGE2 alone treatment in cortical bone of ovariectomized rats. J. Bone Miner. Res. 12, 2108 – 2112 (1997). 193. X. J. Li, W. S. Jee, Y. L. Li, and P. Patterson-Buckendahl, Transient effects of subcutaneously administered prostaglandin E2 on cancellous and cortical bone in young adult dogs. Bone 11, 353 – 364 (1990). 194. H. R. Jorgensen, H. Svanholm, and A. Host, Bone formation induced in an infant by systemic prostaglandin-E2 administration. Acta. Orthop. Scand. 59, 464 – 466 (1988). 195. H. Z. Ke, V. W. Shen, H. Qi, D. T. Crawford, D. D. Wu, X. G. Liang, K. L. Chidsey-Frink, C. M. Pirie, H. A. Simmons, and D. D. Thompson, Prostaglandin E2 increases bone strength in intact rats and in ovariectomized rats with established osteopenia. Bone 23, 249 – 255 (1998). 196. C. H. Lin, W. S. Jee, Y. F. Ma, and R. B. Setterberg, Early effects of prostaglandin E2 on bone formation and resorption in different bone sites of rats. Bone 17, 255S – 259S (1995). 197. W. E. Gallwitz, G. R. Mundy, C. H. Lee, M. Qiao, G. D. Roodman, M. Raftery, S. J. Gaskell, and L. F. Bonewald, 5-lipoxygenase metabolites of arachidonic acid stimulate isolated osteoclasts to resorb calcified matrices. J. Biol. Chem. 268, 10087 – 10094 (1993). 198. C. Garcia, B. F. Boyce, J. Gilles, M. Dallas, M. Qiao, G. R. Mundy, and L. F. Bonewald, Leukotriene B4 stimulates osteoclastic bone resorption both in vitro and in vivo. J. Bone Miner. Res. 11, 1619 – 1627 (1996). 199. C. Garcia, M. Qiao, D. Chen, M. Kirchen, W. Gallwitz, G. R. Mundy, and L. F. Bonewald, Effects of synthetic peptidoleukotrienes on bone resorption in vitro. J. Bone Miner. Res. 11, 521 – 529 (1996). 200. S. Meghji, J. R. Sandy, A. M. Scutt, W. Harvey, and M. Harris, Stimulation of bone resorption by lipoxygenase metabolites of arachidonic acid. Prostaglandins 36, 139 – 149 (1988). 201. Z. Estrov, D. S. Halperin, F. Coceani, and M. H. Freedman, Modulation of human marrow haematopoiesis by leucotrienes in vitro. Br. J. Haematol. 69, 321 – 327 (1988). 202. D. Pasquale and G. Chikkappa, Lipoxygenase products regulate proliferation of granulocyte-macrophage progenitors. Exp. Hematol. 21, 1361 – 1365 (1993). 203. L. Stenke, P. Reizenstein, and J. A. Lindgren, Leukotrienes and lipoxins: New potential performers in the regulation of human myelopoiesis. Leukemia Res. 18, 727 – 732 (1994).
399 204. F. Dupuis, V. Desplat, V. Praloran, and Y. E. Denizot, Effects of lipidic mediators on the growth of human myeloid and erythroid marrow progenitors. J. Lipid Mediat. Cell Signal. 16, 117 – 125 (1997). 205. W. Ren and R. Dziak, Effects of leukotrienes on osteoblastic cell proliferation. Calcif. Tissue Int. 49, 197 – 201 (1993). 206. L. F. Bonewald, M. Flynn, M. Qiao, M. R. Dallas, G. R. Mundy, and B. F. Boyce, Mice lacking 5-lipoxygenase have increased cortical bone thickness. Adv. Exp. Med. Biol. 433, 299 – 302 (1997). 207. D. P. Nicolella, A. E. Nicholls, L. F. Bonewald, and J. Lankford, Mechanical properties of bone from mice lacking 5-lipoxygenase differ significantly from normal mice. J. Bone Miner. Res. 12 (Suppl) 223 (1997). 208. K. Traianedes, M. R. Dallas, I. R. Garrett, G. R. Mundy, and L. F. Bonewald, 5-Lipoxygenase metabolites inhibit bone formation in vitro. Endocrinology 139, 3178 – 3184 (1998). 209. W. R. Henderson, The role of leukotrienes in inflammation. Ann. Intern. Med. 121, 684 – 697 (1994). 210. F. M. Collier, W. H. Huang, W. R. Holloway, J. M. Hodge, M. T. Gillespie, L. L. Daniels, M. H. Zheng, and G. C. Nicholson, Osteoclasts from human giant cell tumors of bone lack estrogen receptors. Endocrinology 139, 1258 – 1267 (1998). 211. P. M. Renzi, Antileukotriene agents and asthma: The dart that kills the elephant? CMAJ 160, 217 – 223 (1999). 212. S. E. Wenzel, Leukotriene receptor antagonists and related compounds. Can. Respir. J. 6, 189 – 193 (1999). 213. S. Battu, J. L. Beneytout, M. Pairet, and M. Rigaud, Cyclooxygenase-2 up-regulation after FLAP transfection in human adenocarcinoma cell line HT29 cl. 19A. FEBS Lett. 437, 49 – 55 (1998). 214. R. S. Weinstein, J. A. Jilka, A. M. Parfitt, and S. C. Manolagos, Inhibition of osteoblastogenesis and promotion of apoptosis of osteoblasts and osteocytes by glucocorticoids: Potential mechanisms of their deleterious effects on bone. J. Clin. Invest. 102, 274 – 282 (1998). 215. A. Sartorio, A. Conti, and B. Ambrosi, Bone and collagen turnover in patients with active and preclinical Cushings syndrome and in subjects with adrenal incidentaloma. J. Clin. Endocrinol. Metab. 83, 2605 – 2606 (1998). 216. D. J. Baylink, Glucocorticoid-induced osteoporosis. N. Engl. J. Med. 309, 306 – 308 (1983). 217. R. L. Wilder, Neuroendocrine-immune system interactions and autoimmunity. Annu. Rev. Immunol. 13, 307 – 338 (1995). 218. L. J. Crofford, T. K. Konstantine, G. Mastorakos, M. A. Magiakou, J. Wells, K. S. Kanik, P. W. Gold, G. P. Chrousos, and R. L. Wilder, Circadian relationships between interleukin (IL)-6 and hypothalamic-pituitary-adrenal axis hormones: Failure of IL-6 to cause sustained hypercortisolism in patients with early untreated rheumatoid arthritis. J. Clin. Endocrinol. Metab. 82, 1279 – 1283 (1997). 219. M. R. Podgorski, N. J. Goulding, N. K. Hall, R. J. Flower, and P. J. Maddison, Autoantibodies to lipocortin-1 are associated with impaired glucocorticoid responsiveness in rheumatoid arthritis. J. Rheumatol. 19, 1668 – 1671 (1992). 220. G. R. Mundy, G. D. Roodman, L. F. Bonewald, T. Yoneda, and M. Sabatini, Effects of TNF and lymphotoxin on bone cells. Immunol. Ser. 56, 483 – 498 (1992). 221. G. R. Mundy, Myeloma bone disease. Eur. J. Cancer 34, 246 – 51 (1998). 222. S. C. Manolagas, The role of IL-6 type cytokines and their receptors in bone. Ann. N.Y. Acad. Sci. 840, 194 – 204 (1998). 223. W. Zhang and R. Dziak, Effects of interleukin-1 alpha on arachidonic acid metabolism in human osteosarcoma osteoblastic cells. Prostagland. Leukot. Essent. Fatty Acids 49, 521 – 526 (1993). 224. W. Zhang and R. Dziak, Tumor necrosis factor alpha stimulates arachidonic acid metabolism in human osteoblastic osteosarcomal cells. Prostagland. Leukot. Essent. Fatty Acids 54, 427 – 431 (1996).
400 225. S. S. Edwin, S. L. LaMarche, D. Thai, D. W. Branch, and M. D. Mitchell, 5-Hydroxyeicosatetraenoic acid biosynthesis by gestational tissues: Effects of inflammatory cytokines. Am. J. Obstet. Gynecol. 169, 1467 – 1471 (1993). 226. J. W. Bijlsma, Prevention of glucocorticoid induced osteoporosis. Ann. Rheum. Dis. 56, 507 – 509 (1997). 227. M. K. OBanion, V. D. Winn, and D. A. Young, cDNA cloning and functional activity of a glucocorticoid-regulated inflammatory cyclooxygenase. Proc. Natl. Acad. Sci. USA 89, 4888 – 4892 (1992). 228. A. J. de Brum-Fernandes, S. Laporte, M. Heroux, M. Lora, C. Patry, H. A. Menard, R. Dumais, and R. Leduc, Expression of prostaglandin endoperoxide synthase-1 and prostaglandin endoperoxide synthase-2 in human osteoblasts. Biochem. Biophys. Res. Commun. 198, 955 – 960 (1994). 229. F. M. Maciel, P. Sarrazin, S. Morisset, M. Lora, C. Patry, R. Dumais, and A. J. de Brum-Fernandes, Induction of cyclooxygenase-2 by parathyroid hormone in human osteoblasts in culture. J. Rheumatol. 24, 2429 – 2435 (1997). 230. M. Goppelt-Struebe, D. Schaefer, and A. J. Habenicht, Differential regulation of cyclo-oxygenase-2 and 5-lipoxygenase-activating protein (FLAP) expression by glucocorticoids in monocytic cells. Br. J. Pharmacol. 122, 619 – 624 (1997). 231. C. A. Riddick, W. L. Ring, J. R. Baker, C. R. Hodulik, and T. D. Bigby, Dexamethasone increases expression of 5-lipoxygenase and its activating protein in human monocytes and THP-1 cells. Eur. J. Biochem. 246, 112 – 118 (1997). 232. P. V. Hauschka, A. E. Mavrakos, S. E. Iafrati, S. E. Doleman, and M. Klagsbrun, Growth factors in bone matrix. J. Biol. Chem. 261, 12665 – 12674 (1986). 233. S. M. Seyedin, T. C. Thomas, A. Y. Thompson, D. M. Rosen, and K. A. Piez, Purification and characterization of two cartilage-inducing factors from bovine demineralized bone. Proc. Natl. Acad. Sci. USA 82, 2267 – 2271 (1985). 234. L. F. Bonewald and S. L. Dallas, The role of active and latent TGF in bone formation. J. Cell Biochem. 55, 350 – 357 (1994). 235. M. Centrella, M. C. Horowitz, J. M. Wozney, and T. L. McCarthy, Transforming growth factor-beta gene family members and bone. Endocr. Rev. 15, 27 – 39 (1994). 236. A. B. Roberts, Molecular and cell biology of TGF-beta. Miner. Electrolyte Metab. 24, 111 – 119 (1998). 237. L. F. Bonewald, Regulation and regulatory activities of transforming growth factor . Crit. Rev. Eukary. Gene Express. 9, 33 – 44 (1999). 238. J. Pfeilschifter, and G. R. Mundy, Modulation of transforming growth factor beta activity in bone cultures by osteotropic hormones. Proc. Natl. Acad. Sci. USA 84, 2024 – 2028 (1987). 239. S. L. Dallas, G. R. Mundy, and L. F. Bonewald, Proteolytic cleavage of the latent TGF binding protein-1 by avian osteoclasts: A possible mechanism for release of latent TGF from bone matrix. J. Bone Miner. Res. 12 (Suppl. 1), S418 (1997). 240. E. J. Mackie and U. Trechsel, Stimulation of new bone formation in vivo by transforming growth factor : remodeling of woven bone and lack of inhibition by indomethacin. Bone 11, 295 – 300 (1990). 241. C. Marcelli, A. J. Yates, and G. R. Mundy, In vivo effects of human recombinant transforming growth factor beta on bone turnover in normal mice. J. Bone Miner. Res. 5, 1087 – 1096 (1990). 242. M. Noda and J. J. Camilliere, In vivo stimulation of bone formation by transforming growth factor-beta. Endocrinology 124, 2991 – 2994 (1989). 243. R. Fujimoto, T. Tanizawa, S. Nishida, N. Yamamoto, S. Soshi, N. Endo, and H. E. Takahashi, Local effects of transforming growth factor-beta 1 on rat calvaria: Changes depending on the dose and the injection site. J. Bone Miner. Metab. 17, 11 – 17 (1999). 244. L. S. Beck, L. DeGuzman, W. P. Lee, Y. Xu, L. A. McFatridge, N. A. Gillett, and E. P. Amento, TGF1 induces bone closure of skull defects. J. Bone Miner. Res. 6, 1257 – 1265 (1991).
GREGORY R. MUNDY ET AL. 245. D. Rosen, S. C. Miller, E. DeLeon, A. Y. Thompson, H. Bentz, M. Matthews, and S. Adams, Systemic administration of recombinant transforming growth factor beta 2 (rTGF-2) stimulates parameters of cancellous bone formation in juvenile and adult rats. Bone 15, 355 – 359 (1994). 246. M. Machwate, E. Zerath, X. Holy, M. Hott, D. Godet, A. Lomri, and P. J. Marie, Systemic administration of transforming growth factor2 prevents the impaired bone formation and osteopenia induced by unloading in rats. J. Clin. Invest. 96, 1245 – 1253 (1995). 247. D. Gazit, Y. Zilberman, L. Passi-Even, R. Ebner, and A. Kahn, In vivo administration of TGF beta 1 increases marrow osteoprogenitor number and selectively reverses the bone loss seen in old “osteoporotic” mice. J. Bone Miner. Res. 10 (Suppl. 1), 25 (1995). 248. J. Pfeilschifter, O. Wolf, A. Nautmann, H. W. Minne, G. R. Mundy, and R. Ziegler, Chemotactic response of osteoblast-like cells to TGF. J. Bone Miner. Res. 5, 825 – 830 (1990). 249. J. M. Hock, E. Canalis, and M. Centrella, Transforming growth factor stimulates bone matrix apposition and bone cell replication in cultured fetal rat calvariae. Endocrinology 126, 421 – 426 (1990). 250. M. E. Antosz, C. G. Bellows, and J. E. Aubin, Effects of transforming growth factor and epidermal growth factor on cell proliferation and the formation of bone nodules in isolated fetal rat calvaria cells. J. Cell. Physiol. 140, 386 – 395 (1989). 251. S. E. Harris, L. F. Bonewald, M. A. Harris, M. Sabatini, S. L. Dallas, J. Feng, N. Ghosh-Choudhury, J. Wozney, and G. R. Mundy, Effects of TGF on bone nodule formation and expression of bone morphogenetic protein-2, osteocalcin, osteopontin, alkaline phosphatase and type I collagen mRNA in prolonged cultures of fetal rat calvarial osteoblasts. J. Bone Miner. Res. 9, 855 – 863 (1994). 252. T. L. Chen and R. L. Bates, Recombinant human transforming growth factor beta 1 modulates bone remodeling in a mineralizing bone organ culture. J. Bone Miner. Res. 8, 423 – 434 (1993). 253. A. Erlebacher and R. Derynck, Increased expression of TGF2 in osteoblasts results in an osteoporosis-like phenotype. J. Cell Biol. 132, 195 – 210 (1996). 254. A. H. Tashjian, E. F. Voelkel, M. Lazzaro, F. R. Singer, A. B. Roberts, R. Derynck, M. E. Winkler, and L. Levine, Alpha and beta human transforming growth factors stimulate prostaglandin production and bone resorption in cultured mouse calvaria. Proc. Natl. Acad. Sci. USA 82, 4535 – 4538 (1985). 255. J. J. Letterio and E. P. Bottinger, TGF-beta knockout and dominantnegative receptor transgenic mice. Miner. Electrolyte Metab. 24, 161 – 167 (1998). 256. M. M. Shull, I. Ormsby, A. B. Kier, S. Pawlowski, R. J. Diebold, M. Yin, R. Allen, C. Sidman, G. Proetzel, D. Calvin, N. Annunziata, and T. Doetschman, Targeted disruption of the mouse transforming growth factor 1 gene results in multifocal inflammatory disease. Nature 359, 693 – 699 (1992). 257. A. B. Kulkarni, C. G. Huh, D. Becker, A. Geiser, M. Lyght, K. C. Fkander, A. B. Roberts, M. B. Sporn, J. M. Ward, and S. Karlsson, Transforming growth factor 1 null mutation in mice causes excessive inflammatory response and early death. Proc. Natl. Acad. Sci. USA 90, 770 – 774 (1993). 258. A. G. Geiser, Q. Q. Zeng, M. Sato, L. M. Helverding, T. Hirano, and C. H. Turner, Decreased bone mass and bone elasticity in mice lacking the transforming growth factor beta-1 gene. Bone 23, 87 – 93 (1998). 259. L. P. Sanford, I. Ormsby, A. C. Gittenberger-de Groot, H. Sariola, R. Friedman, G. P. Boivin, E. I. Cardell, and T. Doetschman, TGF beta 2 knockout mice have multiple developmental defects that are non-overlapping with other TGF beta knockout phenotypes. Development 124, 2659 – 2670 (1997). 260. G. Proetzel, S. A. Pawlowski, M. V. Wiles, M. Yin, G. P. Boivin, P. N. Howles, J. Ding, M. W. Ferguson, and T. Doetschman, Transforming growth factor beta 3 is required for secondary palate fusion. Nature Genet. 11, 409 – 414 (1995).
CHAPTER 13 Cytokines and Bone Remodeling 261. J. L. Letterio, A. G. Geiser, A. B. Kulkarni, N. S. Roche, M. B. Sporn, and A. B. Roberts, Maternal rescue of transforming growth factor-1 null mice. Science 264, 1936 – 1938 (1994). 262. R. J. Diebold, M. J. Eis, M. Yin, I. Ormsby, G. P. Boivin, B. J. Darrow, J. E. Saffitz, and T. Doetschman, Early onset multifocal inflammation in the TGF1-null mouse is lymphocyte-mediated. Proc. Natl. Acad. Sci. USA 92, 12215 – 12219 (1995). 263. B. L. Langdahl, J. Y. Knudsen, H. K. Jensen, N. Gregersen, and E. F. Eriksen, A sequence variation: 713-8delC in the transforming growth factor-beta 1 gene has higher prevalence in osteoporotic women than in normal women and is associated with very low bone mass in osteoporotic women and increased bone turnover in both osteoporotic and normal women. Bone 20, 289 – 294 (1997). 264. Y. Yamada, A. Miyauchi, J. Goto, Y. Takagi, H. Okuizumi, M. Kanematsu, H. Takai, A. Harada, and K. Ikeda, Association of a polymorphism of the transforming growth factor-beta 1 gene with genetic susceptibility to osteoporosis in postmenopausal japanese women. J. Bone Miner. Res. 13, 1569 – 1576 (1998). 265. Y. Yamada, T. Hosoi, F. Makimoto, H. Tanaka, Y. Seino, and K. Ikeda, Transforming growth factor beta-1 gene polymorphisms and bone mineral density in Japanese adolescents. Am. J. Med. 106, 477 – 479 (1999). 266. D. Gazit, Y. Zilberman, R. Ebner, and A. Kahn, Bone loss (osteopenia) in old male mice results from diminished activity and availability of TGF. J. Cell. Biochem. 70, 478 – 488 (1998). 267. J. L. Wrana, TGF-beta receptors and signaling mechanisms. Miner. Electrolyte Metab. 24, 120 – 130 (1998). 268. J. Massague, TGF signal transduction. Annu. Rev. Biochem. 67, 753 – 791 (1998). 269. R. Derynck and X. H. Feng, TGF-beta receptor signalling. Biochim. Biophys. Acta 1333, F105 – F150 (1997). 270. M. Centrella, C. Ji, and T. L. McCarthy, Control of TGF-beta receptor expression in bone. Front. Biosci. 3, D113 – 124 (1998). 271. A. Horner, P. Kemp, C. Summers, S. Bord, N. J. Bishop, A. W. Kelsall, N. Coleman, and J. E. Compston, Expression and distribution of transforming growth factor-beta isoforms and their signalling receptors in growing human bone. Bone 23, 95 – 102 (1998). 272. E. H. Filvaroff, A. Erlebacher, J. O. Ye, S. E. Gitelman, J. C. Lotz, M. R. Heilmann, and R. Derynck, Increased trabecular bone in transgenic mice expressing a dominant-negative, type II TGF receptor in osteoblasts. J. Bone Miner. Res. 12(Suppl. 1), S118 (1997). 273. R. Serra, M. Johnson, E. H. Filvaroff, J. LaBorde, D. M. Sheehan, R. Derynck, and H. L. Moses, Expression of a truncated, kinase-defective TGF type II receptor in mouse skeletal tissue promotes terminal chondrocyte differentiation and osteoarthritis. J. Cell Biol. 139, 541 – 552 (1997). 274. F. Lopez-Casillas, J. L. Wrana, and J. Massague, Betaglycan presents ligand to the TGF signaling receptor. Cell 73, 1435 – 1444 (1993). 275. L. F. Bonewald, L. Wakefield, R. O. C. Oreffo, A. Escobedo, D. R. Twardzik, and G. R. Mundy, Latent forms of transforming growth factor-beta (TGF) derived from bone cultures: Identification of a naturally occurring 100-kDa complex with similarity to recombinant latent TGF. Mol. Endocrinol. 5, 741 – 751 (1991). 276. S. L. Dallas, S. Park-Snyder, K. Miyazono, D. Twardzik, G. R. Mundy, and L. F. Bonewald, Characterization and autoregulation of latent transforming growth factor (TGF) complexes in osteoblastlike cell lines. J. Biol. Chem. 269, 6815 – 6822 (1994). 277. T. Kanzaki, A. Olofsson, A. Moren, C. Wernstedt, U. Hellman, K. Miyazono, L. Claesson-Welsh, and C. H. Heldin, TGF1 binding protein: A component of the large latent complex of TGF1, with multiple repeat sequences. Cell 61, 1051 – 1061 (1990). 278. T. Tsuji, F. Okada, K. Yamaguchi, and T. Nakamura, Molecular cloning of the large subunit of transforming growth factor type masking protein and expression of the mRNA in various rat tissues. Proc. Natl. Acad. Sci. USA 87, 8835 – 8839 (1990).
401 279. S. L. Dallas, K. Miyazono, T. M. Skerry, G. R. Mundy, and L. F. Bonewald, Dual role for the latent TGF binding protein (LTBP) in storage of latent TGF in the extracellular matrix and as a structural matrix protein. J. Cell Biol. 131, 539 – 549 (1995). 280. S. L. Dallas, D. R. Keene, S. P. Bruder, J. Saharinen, L. Y. Sakai, G. R. Mundy, and L. F. Bonewald, Role of the latent-TGF binding protein-1 in fibrillin-containing microfibrils in bone cells in vitro and in vivo. J. Bone Miner. Res. 15, 68 – 81 (2000). 281. K. Miyazono, A. Olofsson, P. Colosetti, and C. H. Heldin, A role of the latent TGF1-binding protein in the assembly and secretion of TGF1. EMBO J. 10, 1091 – 1101 (1991). 282. J. Taipale, K. Miyazono, C. H. Heldin, and J. Keski-Oja, Latent transforming growth factor-1 associates to fibroblast extracellular matrix via latent TGF- binding protein. J. Cell Biol. 124, 171 – 181 (1994). 283. J. Taipale, J. Saharinen, K. Hedman, and J. Keski-Oja, Latent transforming growth factor-1 and its binding protein are components of extracellular matrix microfibrils. J. Histochem. Cytochem. 44, 875 – 889 (1996). 284. H. A. Pedrozo, Z. Schwartz, R. Gomez, A. Ornoy, W. Xin-Sheng, S. L. Dallas, L. F. Bonewald, D. D. Dean, and B. D. Boyan, Growth plate chondrocytes store latent TGF in their matrix through latent TGF binding protein. J. Cell. Physiol. 177, 343 – 354 (1998). 285. S. Sinha, C. Nevett, C. A. Shuttleworth, and C. M. Kielty, Cellular and extracellular biology of the latent transforming growth factorbeta binding proteins. Matrix Biol. 17, 529 – 545 (1999). 286. W. Yin, E. Smiley, J. Germiller, R. P. Mecham, J. V. Florer, R. J. Wenstrup, and J. Bonadio, Isolation of a novel latent transforming growth factor- binding protein gene (LTBP-3). J. Biol. Chem. 270, 10147 – 10160 (1995). 287. R. Flaumenhaft, S. Kojima, M. Abe, and D. B. Rifkin, Activation of latent transforming growth factor beta. Adv. Pharmacol. 24, 51 – 76 (1993). 288. J. A. Yee, L. Yan, J. C. Dominguez, E. H. Allan, and T. J. Martin, Plasminogen-dependent activation of latent transforming growth factor beta (TGF) by growing cultures of osteoblast-like cells. J. Cell. Physiol. 157, 528 – 534 (1993). 289. I. A. Silver, R. J. Murrills, and O. J. Etherington, Microelectrode studies on the acid microenvironment beneath adherent macrophages and osteoclasts. Exp. Cell Res. 175, 266 – 276 (1998). 290. R. Flaumenhaft, M. Abe, Y. Sato, K. Miyazono, C. H. Heldin, and D. B. Rifkin, Role of the latent TGF- binding protein in the activation of latent TGF by co-cultures of endothelial and smooth muscle cells. J. Cell Biol. 120, 995 – 1002 (1993). 291. S. Schultz-Cherry, H. Chen, D. F. Mosher, T. M. Misenheimer, H. C. Krutzsch, D. D. Roberts, and J. E. Murphy-Ullrich, Regulation of transforming growth factor- activation by discrete sequences of thrombospondin 1. J. Biol. Chem. 269, 26783 – 26788 (1995). 292. S. E. Crawford, V. Stellmach, J. E. Murphy-Ullrich, S. M. F. Ribeiro, J. Lawler, R. O. Hynes, G. P. Boivin, and N. Bouck, Thrombospondin1 is a major activator of TGF in vivo. Cell 93, 1159 – 1170 (1998). 293. A. J. Friedenstein, P. K. Chailakhyan, and U. V. Gerasimov, Bone marrow osteogenic stem cells: In vitro cultivation and transplantation in diffusion chambers. Cell Tissue Kinet. 20, 263 – 272 (1987). 294. S. M. Taylor and P. A. Jones, Multiple new phenotypes induced in 10T1/2 and 3T3 cells treated with 5-azacytidine. Cell 17, 771 – 779 (1979). 295. J. H. Bennett C. J. Joyner, J. T. Triffitt, and M. E. Owen, Adipocytic cells cultured from marrow have osteogenic potential. J. Cell Sci. 99, 131 – 139 (1991). 296. J. N. Beresford, J. H. Bennett, C. Delving, P. S. Leboy, and M. E. Owen, Evidence for an inverse relationship between the differentiation of adipocytic and osteogenic cells in rat marrow stromal cell cultures. J. Cell Sci. 102, 341 – 351 (1992). 297. J. E. Aubin, F. Liu, L. Malaval, and A. K. Gupta, Osteoblast and chondroblast differentiation. Bone 17, 77S – 83S (1995).
402 298. M. R. Urist, Bone formation by autoinduction. Science 150, 893 – 899 (1965). 299. J. M. Wozney, V. Rosen, A. J. Celeste, L. M. Mitsock, M. J. Whitters, R. Kriz, R. Hewick, and E. A. Wang, Novel regulators of bone formation: Molecular clones and activities. Science 242, 1528 – 1534 (1988). 300. V. Rosen, J. Nove, J. J. Song, R. S. Thies, K. Cox, and J. M. Wozney, Responsiveness of clonal limb bud cell lines to bone morphogenetic protein 2 reveals a sequential relationship between cartilage and bone cell phenotypes. J. Bone Miner. Res. 9, 1759 – 1768 (1994). 301. M. Ahrens, T. Ankenbauer, D. Schroder, A. Hollnagel, H. Mayer, and G. Gross, Expression of human bone morphogenetic protein-2 or -4 in murine mesenchymal progenitor C3H10T1/2 cells induces differentiation into distinct mesenchymal cell lineages. DNA Cell Biol. 12, 871 – 880 (1993). 302. I. Asahina, T. K. Sampath, and P. V. Hauschka, Human osteogenic protein-1 induces chondroblastics, osteoblastics and/or adipocyte differentiation of clonal murine target cells. Exp. Cell Res. 222, 38 – 47 (1996). 303. S. E. Harris, J. Q. Feng, M. A. Harris, N. Ghosh-Choudhury, M. R. Dallas, J. M. Wozney, and G. R. Mundy, Recombinant bone morphogenetic protein 2 accelerates bone cell differentiation and stimulates BMP-2 mRNA expression and BMP-2 promoter activity in primary fetal rat calvarial osteoblast cultures. Mol. Cell. Differ. 3, 137 – 155 (1995). 304. D. Chen, M. A. Harris, G. Rossini, C. R. Dunstan, S. L. Dallas, J. Q. Feng, G. R. Mundy, and S. E. Harris, Bone morphogenetic protein 2 (BMP-2) enhances BMP-3, BMP-4 and bone cell differentiation marker gene expression during the induction of mineralized bone matrix formation in cultures of fetal rat calvarial osteoblasts. Calcif. Tissue Int. 60, 283 – 290 (1997). 305. A. Yamaguchi, T. Katagiri, T. Ikeda, J. M. Wozney, V. Rosen, E. A. Wang, A. Kahn, T. Suda, and S. Yoshiki, Recombinant human bone morphogenetic protein-2 stimulates osteoblastic maturation and inhibits myogenic differentiation in vitro. J. Cell. Biol. 113, 681 – 687 (1991). 306. Y. Takuwa, C. Ohse, E. A. Wang, J. M. Wozney, and K. Yamashita, Bone morphogenetic protein-2 stimulates alkaline phosphatase activity and collagen synthesis in cultured osteoblastic cells. Biochem. Biophys. Res. Commun. 174, 96 – 101 (1991). 307. R. S. Thies, M. Bauduy, B. A. Ashton, L. Kurtzberg, J. M. Wozney, and V. Rosen, Recombinant human bone morphogenetic protein-2 induces osteoblastic differentiation in W-20-17 stromal cells. Endocrinology 130, 1318 – 1324 (1992). 308. T. Katagiri, A. Yamaguchi, M. Komaki, E. Abe, N. Takahashi, T. Ikeda, V. Rosen, J. M. Wozney, A. Fujisawa-Sehara, and T. Suda, Bone morphogenetic protein-2 converts the differentiation pathway of C2C12 myoblasts into the osteoblast lineage. J. Cell Biol. 127, 1755 – 1766 (1994). 309. J. M. Wozney and V. Rosen, Bone morphogenetic protein and bone bone morphogenetic protein gene family in bone formation and repair. Clin. Orthop. 346, 26 – 37 (1998). 310. S. D. Boden, G. Hair, L. Titus, M. Racine, K. Mccuaig, J. M. Wozney, and M. S. Nanes, Glucocorticoid-induced differentiation of fetal rat calvarial osteoblasts is mediated by bone morphogenetic protein-6. Endocrinology 138, 2820 – 2828 (1997). 311. D. Rickard, L. C. Hofbauer, S. Bonde, F. Gori, T. C. Spelsberg, and B. L. Riggs, Bone morphogenetic protein-6 production in human osteoblastic cell lines: selective regulation by estrogen. J. Clin. Invest. 101, 413 – 422 (1998). 312. C. D. Grimsbud, P. R. Romano, M. DSouza, J. D. Puzas, P. R. Reynolds, R. N. Rosier, and R. J. OKeefe, BMP-6 is an autocrine stimulator of chondrocyte differentiation. J. Bone Miner. Res. 14, 475 – 482 (1999).
GREGORY R. MUNDY ET AL. 313. K. Miyama, G. Yamada, T. S. Yamamoto, C. Takagi, K. Miyado, M. Sakai, N. Ueno, and H. Shibuya, A BMP-inducible gene, Dlx5, regulates osteoblast differentiation and mesoderm induction. Dev. Biol. 208, 123 – 133 (1999). 314. S. E. Harris, M. A. Harris, N. Ghosh-Choudhury, C. Xu, D. Chen, L. Wang, J. Rubenstein, and G. R. Mundy, Dlx2 homeobox gene is rapidly induced by BMP-2 and directly regulates BMP-2 transcription through a cell-specific enhancer region. Bone 23, S197 (1998). 315. T. Komori, H. Yagi, S. Nomura, A. Yamaguchi, K. Sasaki, K. Deguchi, Y. Shimizu, R. T. Bronson, Y. Gao, M. Inada, M. Sato, R. Okamoto, R. Kitamura, S. Yoshiki, and T. Kishimoto, Targeted disruption of Cbfa1 results in a complete lack of bone formation owing to maturational arrest of osteoblasts. Cell 89, 755 – 764 (1997). 316. S. Mundlos, F. Otto, C. Mundlos, J. B. Mulliken, A. S. Aylsworth, S. Albright, D. Lindhout, W. G. Cole, W. Henn, J. H. M. Knoll, M. J. Owen, R. Mertelsmann, B. U. Zabel, and B. R. Olsen, Mutations involving the transcription factor Cbfa1 cause cleidocranial dysplasia. Cell 89, 773 – 779 (1997). 317. F. Otto, A. P. Thornell, T. Crompton, A. Denze, K. C. Gilmour, I. R. Rosewell, G. W. H. Stamp, R. S. P. Beddington, S. Mundlos, B. R. Olsen, P. Selby, and M. Owen, Cbfa 1, a candidate gene for cleidocranial dysplasia syndrome, is essential for osteoblast differentiation and bone development. Cell 89, 765 – 771 (1997). 318. P. Ducy, R. Zhang, V. Geoffroy, A. L. Ridall, and G. Karsenty, Osf2/Cbfa 1: A transcriptional activator of osteoblast differentiation. Cell 89, 747 – 754 (1997). 319. M. Enomoto-Iwamoto, M. Iwamoto, Y. Mukudai, Y. Kawakami, T. Nohno, Y. Higuchi, S. Takemoto, H. Ohuchi, S. Noji, and K. Kurisu, Bone morphogenetic protein signaling is required for maintenance of differentiated phenotype, control of proliferation and hypertrophy in chondrocytes. J. Cell Biol. 140, 409 – 418 (1998). 320. S. E. Gitelman, M. S. Kobrin, J. Q. Ye, A. R. Lopez, A. Lee, and R. Derynck, Recombinant Vgr-1/BMP-6-expressing tumors induce fibrosis and endochondral bone formation in vitro. J. Cell Biol. 126, 1595 – 1609 (1994). 321. T. K. Sampath, J. C. Maliakal, P. V. Hauschka, W. K. Jones, H. Sasak, R. F. Tucker, K. H. White, J. E. Coughlin, M. M. Tucker, and R. H. Pang, Recombinant human osteogenic protein-1 (hOP-1) induces bone formation in vivo with a specific activity comparable with natural bovine osteogenic protein and stimulates osteoblast proliferation and differentiation in vitro. J. Biol. Chem. 267, 20352 – 20362 (1992). 322. H. Zhang and A. Bradley, Mice deficient for BMP-2 are nonviable and have defects in amnion/chorion and cardiac development. Development 122, 2977 – 2986 (1996). 323. G. Winnier, M. Blessing, P. A. Labosky, and B. L. M. Hogan, Bone morphogenetic protein-4 is required for mesoderm formation and patterning in the mouse. Genes Dev. 9, 2105 – 2116 (1995). 324. G. Luo, C. Hofmann, A. L. J. J. Bronckers, M. Sohocki, A. Bradley, and G. Karsenty, BMP-7 is an inducer of nephrogenesis, and is also required for eye development and skeletal patterning. Genes Dev. 9, 2808 – 2820 (1995). 325. D. M. Kingsley, A. E. Bland, J. M. Grubber, P. C. Marker, L. B. Russell, N. G. Copeland, and N. A. Jenkins, The mouse short ear skeletal morphogenesis is associated with defects in a bone morphogenetic member of the TGF superfamily. Cell 71, 399 – 410 (1992). 326. E. E. Storm, T. V. Huynh, N. G. Copeland, N. A. Jenkins, D. M. Kingsley, and S. J. Lee, Limb alterations in brachypodism mice due to mutations in a new member of the TGF-superfamily. Nature 368, 639 – 643 (1994). 327. J. T. Thomas, K. Lin, M. Nandedkar, M. Camargo, J. Cervenka, and F. P. Luyten, A human chondrodysplasia due to a mutation in a TGF superfamily member. Nature Gene. 12, 315 – 317 (1996). 328. Y. Kawakami, T. Ishikawa, M. Shimabara, N. Tanda, M. EnomotoIwamoto, M. Iwamoto, T. Kuwana, A. Ueki, S. Noji, and T. Nohno,
CHAPTER 13 Cytokines and Bone Remodeling
329.
330.
331.
332.
333.
334.
335.
336.
337.
BMP signaling during bone pattern determination in the developing limb. Development 122, 3557 – 3566 (1996). S. Yi, A. Daluiski, M. Zhao, V. Rosen, and K. Lyons, The bone morphogenetic protein receptor type IB is required for the development of digits and the maintenance of articular cartilage. 45th Annual Meeting of the Orthopaedic Research Society, Anaheim, CA (1999). [Abstract 61] H. Nishitoh, H. Ichijo, M. Kimura, T. Matsumoto, F. Makishima, A. Yamaguchi, H. Yamashita, S. Enomoto, and K. Miyazono, Identification of type I and type II serine/threonine kinase receptors for growth/differentiation factor-5. J. Biol. Chem. 271, 21345 – 21352 (1996). Y. Mishina, A. Suzuki, N. Ueno, and R. B. Behringer, Bmpr encodes a type I bone morphogenetic protein receptor that is essential for gastrulation during mouse embryogenesis. Genes Dev. 9, 3027 – 3037 (1995). B. B. Koenig, J. S. Cook, D. H. Wolsing, J. Ting, J. P. Tiesman, P. E. Correa, C. A. Olson, A. L. Pecquet, F. Ventura, R. A. Grant, G. Chen, J. L. Wrana, J. Massague, and J. S. Rosenbaum, Characterization and cloning of a receptor for BMP-2 and BMP-4 from NIH 3T3 cells. Mol. Cell. Biol. 14, 5961 – 5974 (1994). P. ten Dijke, H. Yamashita, T. K. Sampath, A. H. Reddi, M. Estevez, D. L. Riddle, H. Ichijo, C. H. Heldin, and K. Miyazono, Identification of type I receptors for osteogenic protein-1 and bone morphogenetic protein-4. J. Biol. Chem. 269, 16985 – 16988 (1994). M. Kawabata, A. Chytil, and H. L. Moses, Cloning of a novel type II serine/threonine kinase receptor through interaction with the type I transforming growth factor- receptor. J. Biol. Chem. 270, 5625 – 5630 (1995). T. Nohno, T. Ishikawa, T. Saito, K. Hosokawa, S. Noji, D. H. Wolsing, and J. S. Rosenbaum, Identification of a human type II receptor for bone morphogenetic protein-4 that forms differential heteromeric complexes with bone morphogenetic protein type I receptors. J. Biol. Chem. 270, 22522 – 22526 (1995). B. L. Rosenzweig, T. Imamura, T. Okadome, G. N. Cox, H. Yamashita, P. ten Dijke, C. Heldin, and K. Miyazono, Cloning and characterization of a human type II receptor for bone morphogenetic proteins. Proc. Natl. Acad. Sci. USA 92, 7632 – 7636 (1995). H. Yamashita, P. ten Dijke, D. Huylebroeck, T. K. Sampath, M. Andries, J. S. Smith, C. H. Heldin, and K. Miyazono, Osteogenic protein-1 binds to activin type II receptors and induces certain activinlike effects. J. Cell. Biol. 130, 217 – 226 (1995).
403 338. M. Macias-Silva, P. A. Hoodless, S. J. Tang, M. Buchwald, and J. L. Wrana, Specific activation of Smad 1 signaling pathways by the BMP7 type I receptor, ALK2. J. Biol. Chem. 273, 25628 – 25636 (1998). 339. J. Carcamo, A. Zentella, and J. Massague, Disruption of transforming growth factor b signaling by a mutation that prevents transphosphorylation within the receptor complex. Mol. Cell. Biol. 15, 1573 – 1581 (1995). 340. R. Wieser, J. L. Wrana, and J. Massague, GS domain mutations that constitutively activeate TbR-I, the downstream signaling component in the TGF- receptor complex. EMBO J. 14, 2199 – 2208 (1995). 341. P. A. Hoodless, T. Haerry, S. Abdollah, M. Stapleton, M. B. OConnor, L. Attisano, and J. L. Wrana, MADR1, a MAD-related protein that functions in BMP2 signaling pathways. Cell 85, 489 – 500 (1996). 342. R. Nishimura, Y. Kato, D. Chen, S. E. Harris, G. R. Mundy and T. Yoneda, Smad5 and DPC4 are key molecules in mediating BMP2-induced osteoblastic differentiation of the pluripotent mesenchymal precursor cell line C2C12. J. Biol. Chem. 273, 1872 – 1879 (1998). 343. Y. Chen, A. Bhushan, and W. Vale, Smad 8 mediates the signaling of the receptor serine kinase. Proc. Natl. Acad. Sci. USA 94, 12938 – 12943 (1997). 344. Y. Shi, Y. F. Wang, L. Jayaraman, H. Yang, J. Massague, and N. P. Pavletich, Crystal structure of Smad MH1 domain bound to DNA: insights on DNA binding in TGF signaling. Cell 94, 585 – 594 (1998). 345. R. S. Lo, Y. G. Chen, Y. Shi, N. Pavletich, and J. Massague, The L3 loop: A structural motif determining specific interactions between SMAD proteins and TGF receptors. EMBO J. 17, 996 – 1005 (1998). 346. X. Chen, M. J. Rubock, and M. Whitman, A transcriptional partner of MAD proteins in TGF- signaling. Nature 383, 691 – 696 (1996). 347. X. Chen, E. Weisberg, V. Fridmacher, M. Watanabe, G. Naco, and M. Whitman, Smad 4 and FAST-1 in the assembly of activin-responsive factor. Nature 389, 85 – 89 (1997). 348. X. Shi, X. Yang, D. Chen, Z. Chang, and X. Cao, Smad 1 interacts with homeobox DNA-binding proteins in bone morphogenetic protein signaling. J. Biol. Chem. 274, 13711 – 13717 (1999). 349. Y. G. Tueh, D. P. Gardner, and C. Kappen, Evidence for regulation of cartilage differentiation homeobox gene Hoxc-8. Proc. Natl. Acad. Sci. USA 95, 9956 – 9961 (1998). 350. D. Wotlon, R. S. Lo, S. Lee, and J. Massague, A Smad transcriptional corepressor. Cell 97, 29 – 39 (1999).
CHAPTER 14
Bone Growth Factors XUEZHONG QIN, REINHARD GYSIN, SUBBURAMAN MOHAN, AND DAVID J. BAYLINK Department of Medicine, Loma Linda University, and the Musculoskeletal Disease Center, Jerry L. Pettis Memorial Veterans Affairs Medical Center, Loma Linda, California 92357
I. Introduction II. Bone Growth Factors
III. Conclusions References
factor availability to cells of osteoblastic lineage is through the release of growth factors from bone during bone resorption [11]. Although this latter growth factor-releasing process is part of the mechanism of the coupling of bone formation to resorption, which is itself a very important topic, we will not deal with this process, as it is covered in detail in Chapter 12. Therefore, this chapter focuses on the molecular mechanisms of action, the in vitro and in vivo functions, and the potential therapeutic relevance of each growth factor. We will also provide the latest information available from knockout, transgenic, and natural mutant studies on these growth factors, as this information is important in bringing to our attention potential adverse effects of a growth factor on other tissues, in addition to its beneficial effect on bone formation. Finally, we will briefly comment on the potential role of a given growth factor in the pathogenesis of osteoporosis.
I. INTRODUCTION Since the early discovery by Marshal Urist [1] that demineralized bone extract could induce bone formation ectopically, scores of papers have been published demonstrating that bone is a storehouse for growth factors that are capable of stimulating both osteoblast cell proliferation and osteoblast differentiation [2 – 10]. Moreover, the finding that osteoblast line cells secrete factors that are mitogenic to bone cells in vitro led to studies on the identification of the active principle from culture media conditioned by a variety of untransformed and transformed bone cell lines in vitro [2,3,5 – 7,9]. These findings showing mitogenic activity in bone or in conditioned medium (CM) from osteoblastic cultures have been extended to demonstrate specific classes of growth factors produced by osteoblasts, including transforming growth factor- (TGF-), bone morphogenetic proteins (BMPs), insulin-like growth factors (IGFs), and fibroblast growth factors (FGFs). A comprehensive treatise on bone growth factors would include regulation of growth factor secretion, together with molecular mechanisms of action and in vitro and in vivo functions. Regarding growth factor secretion, osteoblasts along the trabecular bone and the endosteal surfaces are exposed to growth factors secreted by cells of osteoblast lineage and osteoclast lineage, as well as bone marrow cells. In addition, a unique aspect of growth
OSTEOPOROSIS, SECOND EDITION VOLUME 1
II. BONE GROWTH FACTORS A. The TGF- System To date, five TGF- isoforms have been identified [12] (also see Chapter 13). TGF-1, 2, and 3 are present in mammalian species and TGF-4 and TGF-5 are present only in chicken and Xenopus, respectively. Two heterodimers
405
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
406 termed TGF-1,2 and TGF-2,3 have also been identified in bovine bone extracts [4]. Of the three members of the TGF- family known to be produced by various mammalian tissues, TGF-1 appears to be the major member produced by human osteoblasts and stored in bone. TGF-2 is also known to be produced by human bone cells and stored in bone but to a lesser extent [8,13,14]. Active TGF- exists as disulfidelinked dimers comprised of identical 12,500-Dalton subunits. Studies from TGF- transgenic and knockout experiments indicate that TGF-s are multifunctional genes, although their effects on skeletal tissues appear to be similar. This section focuses on studies related to the in vitro and in vivo effects of TGF- on bone metabolism, as well as the mechanism of TGF- action. 1. IN VITRO STUDIES TGF- produces diverse effects on skeletal tissues and cells in vitro. Studies on the effects of TGF- on cultured osteoblasts and bones in organ cultures have led to a large number of conflicting reports. TGF- stimulates prostaglandin-mediated bone resorption in mouse bone organ culture [15] and inhibits proliferation of isolated fetal rat calvaria cells at low cell density [16], as well as clonal mouse calvaria-derived cells (MC3T3E1) [17]. TGF- also inhibits osteoblast differentiation in MC3T3E1 cells [17] and normal rat osteoblasts [18]. In contrast, TGF- stimulates cell growth in fetal rat calvaria [19]. TGF- increases protein synthesis [19] and collagen expression in fetal rat calvaria cells [16] and stimulates alkaline phosphatase (ALP) in rat osteosarcoma cells [20]. Treatment of rat osteoblastic ROS 17/2.8 cells with a TGF- increases ALP activity in a dose-dependent fashion with an ED50 of approximately 0.2 ng/ml. This effect is more profound during the logarithmic growth than at confluence [17]. Chondrocytes also respond to TGF- in vitro by increasing ALP activity in a biphasic manner [21 – 23] with a maximal response observed at a dose of 0.22 ng/ml [22]. However, TGF- inhibits glycosaminoglycan and collagen production in fully committed mature articular chondrocytes [24]. The discrepancy among studies regarding the effects of TGF- on bone cells appears to be related to the stages of cell differentiation, cell culture conditions, and TGF- doses, as well as the variables that are measured. Differences in TGF- receptor signaling could also contribute to the observed differences in target cell response to TGF-. In general, both chondrocytes and osteoblasts appear to be more sensitive to TGF- at their early developmental stages. Although the effects of TGF- in vitro on bone cells are controversial, TGF- has been shown to be a strong and consistent bone-inductive molecule in vivo (see Section II,A,4). The effect of TGF- on bone cells can be modulated by other factors. For example, TGF- and 1,25(OH)2D3 synergistically increase ALP activity in normal human
XUEZHONG QIN ET AL.
osteoblasts [8,14] and ALP activity and collagen synthesis in rat chondrocytes [22]. Yaeger et al. [25] found that IGFI, insulin, or TGF-s alone did not stimulate induction of aggrecan and type II collagen expression by adult human articular chondrocytes. However, TGF-1 or TGF-2, together with IGF-I or insulin, strongly induces the expression of these proteins. Similarly, acidic FGF (aFGF) or basic FGF (bFGF) alone is unable to induce odontoblast differentiation, whereas both aFGF and bFGF act synergistically with TGF-1 or IGF-I to induce odotoblast differentiation [26]. These data suggest that responses of cells to exogenously added TGF- may vary depending on the type and amount of growth factors produced endogenously by cells in vitro. 2. MECHANISM OF ACTION TGF- has been known to regulate the expression of a number of genes associated with osteoblast activity [8,27 – 29]. TGF- action is mediated through its binding to a distinct heteromeric receptor complex, including type I and type II serine/threonine kinase type receptors. Activation of the receptor complex is initiated upon the binding of the ligand to type II receptor, which then recruits, phosphorylates, and activates the type I receptor [12,30,31]. The activated type I receptor propagates the signal to downstream targets, including the Smad proteins [12,30,31]. To date, at least seven Smads have been identified; their potential role in mediating TGF- action is illustrated in Fig. 1. Understanding the mechanism of TGF- action has been enhanced by studies on TGF- regulation of transcription factors and the interaction of these transcription factors with Smads. Over the last decade, several TGF- and activin-responsive elements, including AP-1, Sp-1, and CTF/NF-1-binding sites, have been identified in the promoters of various genes [30,31]. Among the various TGF-regulated transcription factors, AP-1, a heterodimer of Fos and Jun proteins, has been the most extensively studied, as it clearly plays an important role in the regulation of proliferation and differentiation of osteoblast cell lines [28,32,33]. AP-1-regulated promoters are transcriptionally activated by TGF- [34 – 36]. This activation appears to be mediated through the TGF--stimulated c-Jun/c-Fos expression [37]. Studies have provided evidence for an interaction between AP-1 and Smads in mediating TGF- action [36,38]. Moreover, the osteoblast-specific transcription factor, core-binding factor (Cbfa1), has been suggested to be involved in Smad-mediated TGF- signaling in bone cells [39]. Thus, TGF- effects on target cells may, in part, be mediated via the activation of several known transcription factors. In addition to the AP-1-mediated PKC pathway, previous studies also suggest that TGF--induced gene expression may involve the MAP kinase pathway [30]. It is anticipated that identification of additional TGF--regulated
407
CHAPTER 14 Bone Growth Factors
Proposed model for the role of Smads in TGF- signaling pathway. Binding of a TGF family member to its type II receptor in concert with a type I receptor leads to formation of a receptor complex and phosphorylation of the type I receptor. The activated type I receptor subsequently phosphorylates Smad-2 or Smad-3 (this step can be inhibited by Smad 6/7). Upon phosphorylation, the Smad-2 or Smad-3 homodimer associates with the Smad-4 homodimer to form a heterohexamer that translocates into the nucleus. The Smad complex then activates transcription of target genes through an intermediary transcription factor or by binding to DNA directly [30] (reproduced with permission).
FIGURE 1
transcription factors in these signaling pathways will further advance our understanding of the molecular mechanism of TGF- action in bone. 3. TRANSGENIC/KNOCKOUT STUDIES To date, null mice have been generated for each of the three TGF- isoforms [40 – 43]. TGF-1 knockout mice develop a rapid wasting syndrome with excessive inflammatory response and die by 3 – 4 weeks of age [40,41]. The longitudinal growth and total mineral content are decreased in these knockout mice [44]. In TGF-1 transgenic mice, overexpression of TGF-1 increases plasma TGF-1 concentration and causes glomerulosclerosis [45]. Effects on skeletal tissues in these mice have not been examined. TGF-2 knockout mice exhibit perinatal mortality and a wide range of developmental defects, including some skeletal tissues, such as the spinal column [43]. Because there is little phenotypic overlap between TGF-1 and TGF-2 null mice, these two isoforms appear to have different functions. In contrast to the positive effect of TGF-2 on bone formation via local administration (see Section II,A,4), overexpression of TGF-2 in osteoblasts using the osteocalcin
promoter in mice resulted in progressive bone loss associated with increases in osteoblastic matrix deposition and osteoclastic bone resorption [46]. TGF-3 knockout mice exhibit an incomplete penetrant failure of the palatal shelves to fuse, leading to cleft palate [42,47]. Results from these studies demonstrate that TGF-s are multifunctional genes and that different forms of TGF- may have distinct biological functions. In TGF- knockout mice, one would expect to see a reduction in bone formation based on the in vivo studies described later. However, such effects have not been obvious. Furthermore, instead of producing an increase in bone formation, the overexpression of TGF-2 in osteoblasts led to an increase in bone resorption (see earlier discussion). This latter observation may not be relevant to the effects of TGF- in humans for the following reason. TGF- in mice is known to stimulate prostaglandin synthesis, which in turn is a strong osteolytic agent in mice. In contrast, TGF- does not stimulate prostaglandin synthesis in rats, and the effect of TGF- in rats is to decrease rather than increase bone resorption. Therefore, whereas the genetic studies using these mice provide important information on the in vivo function of TGF-s, these data need to
408 be interpreted with caution when they are applied to other species. 4. IN VIVO TREATMENT STUDIES The findings that TGF- family members are expressed during normal fracture healing and are actively produced by bone cells suggest that TGF- plays an important role in fracture healing [48,49]. Therefore, a number of studies using various animal model systems have been conducted to evaluate the effect of TGF- on bone formation. The results from representative studies are briefly discussed. a. Local Effects The calvaria model has been used frequently to evaluate the effect of TGF- on bone formation in either intact animals [50 – 52] or animals with critical sized defects [53 – 56]. Direct injection of TGF- onto the periostea of parietal bones of neonatal rats [50] or to the subcutaneous tissue overlying the calvaria of mice [51] increases the thickness of parietal bone. Because the stimulatory effect is partially inhibited by concomitant treatment with indomethacin, an inhibitor of prostaglandin synthesis, it is speculated that TGF-1 effects on bone formation may, in part, be mediated by prostaglandin synthesis in mice [51]. However, this contention is not supported by studies using rats [57]. More recently, Fujimoto, et al. [52] reported that subcutaneous administration of TGF-1 onto the parietal surface in rats had a negative effect on bone formation at the injection site, whereas direct injection of TGF-1 in the subperiosteal layer induced new bone formation. These data suggest that the TGF- effect can vary depending on the site. Studies using calvarial critical sized defect models have generally shown that TGF-, via appropriate carriers, can heal the defects in various animal models [53,56], although a lack of response has also been described [54]. The effect of local TGF- administration on bone regeneration in long bones has also been evaluated in intact animals [48,49] or animals with critical sized bone defects [58 – 62]. Subperiosteal injection of TGF- into the nonfractured rat femur results in mesenchymal cell proliferation and the subsequent initiation of chondrogenesis and intramembranous bone formation [48,49]. Short-term administration of TGF-1 directly into the bone marrow space attenuates the stimulation of osteoclastic resorption induced by ovariectomy in rats [60]. Repeated applications of TGF- to rabbit and rat tibial fractures for 6 weeks resulted in calluses that are larger and mechanically stronger than those in controls [58,59]. In contrast, a single injection of TGF-2 [60 or 600 ng) into the developing callus of rabbit tibial fractures 4 days after fracture had no significant effect on fracture healing within 5 to 14 days [61]. These studies suggest that multiple injections of TGF into the callus of the fractured bone are needed to enhance the healing process. Beck et al. [62] reported that
XUEZHONG QIN ET AL.
implantation of TGF-1 (1 g), together with autologous bone marrow, is as effective as the implantation of autogenous cortical bone graft in healing segmental defects in the radius of rabbits, suggesting that bone marrow may serve as an ideal carrier for TGF-. The effect of TGF- on bone formation in other skeletal sites has also been evaluated [56,63,64]. For example, implantation of TGF-1 in a 5-mm critical size defect in the rat mandible results in a dose-dependent (0.1 to 20 g) bone bridging at both 12 and 24 days [56]. In contrast, TGF-1 has a negative effect on bone regeneration when the access of cells from the periosteum to the defect is prevented by microporous membrane. Wikesjo et al [63] reported that implantation of 20 g rhTGF-1 in CaCO3/hydroxyethyl starch carrier into periodontal defects in dogs is not effective in regenerating bone. The authors concluded that rhTGF-1 has a limited potential in periodontal repair. In contrast, local TGF-1 administration has been shown to be effective in the stimulation of articular cartilage repair in mice with arthritis induced by zymosan injection [64]. Overall, the studies just described demonstrate that TGF-, via local administration, is effective in bone regeneration at various skeletal sites in small experimental animals. Whether TGF- is effective in fracture healing in large animals, such as nonhuman primates, needs to be evaluated in the future before its application to patients with large bone fractures. b. Systemic Effects Previous studies have shown that systemic TGF- administration also enhances bone formation [65 – 68]. In ovariectomized rats, daily administration of TGF-2 for 5 weeks by subcutaneous injections stimulates bone formation and prevents ovariectomy-induced bone loss [65]. Short-term systemic administration of rhTGF-2 (5 or 14 days) also stimulates cancellous bone formation in both juvenile and adult rats [66]. Histomorphometric analysis revealed that TGF-2 treatment did not affect the number of osteoclasts or the number of osteoclast nuclei per cell. Consistent with these anabolic effects, Zerath et al. [67] found that treatment of rats with TGF-2 at a dose of 2 g/kg/day prevented the decrease in chondrocyte number and proliferative zone growth in the proliferative zone induced by unloading. Gazit et al. [68] reported that the effect of systemic injection of TGF-1 on bone formation in mice is affected by the age of the animals. Daily administration of TGF-1 at 0.5 or 5 g/day for 20 days into 24-month-old mice caused a significant increase in trabecular bone volume, bone formation, and mineral apposition rate and enhanced fracture healing. However, these stimulatory effects are either not seen or are less significant in younger mice (4 weeks old). The studies just summarized suggest that TGF- is unique in that it increases bone formation and, simultaneously, decreases bone resorption. However, the potential
409
CHAPTER 14 Bone Growth Factors
adverse effects associated with the systemic administration of TGF- have not been fully examined in these studies. It has been shown previously that systemic TGF-1 administration in mice causes severe anemia [69] and that overexpression of TGF-1 in transgenic mice leads to progressive glomerulosclerosis [45,70]. Thus, it will be essential to determine if the systemic administration of TGF- at an effective dose for promoting fracture healing has any significant systemic deleterious effects before the therapeutic potential of TGF- can be fully evaluated. 5. IMPAIRMENT IN TGF- SYSTEM AND PATHOGENESIS OF BONE LOSS Regarding the functional significance of growth factors, such as TGF-, stored in bone, it has been proposed that growth factors are released in a bioactive form during osteoclastic bone resorption to act in a paracrine manner on osteoblast precursors and mature osteoblasts to ensure sitespecific bone replacement [5,7]. Thus, if the extent of the refill of the resorption cavity depends partly on the amount of TGF- stored in bone, then the level of the TGF- sequestered in bone should vary in response to physiological and pathological conditions. Consistent with this argument, we have found an age-related decline in the amount of TGF- stored in the cortical bone of both men and women [71]. More recently, Gazit et al. [72] reported that the matrix of long bones of old mice contains significantly less TGF- than that of young mice. In addition, bone marrow osteoblast progenitor cells isolated from old mice produce less TGF- in vitro than those isolated from young mice. The molecular mechanisms responsible for this age-related decrease in TGF- synthesis can only be speculated upon at this time. In this regard, studies have shown that estrogen deficiency and vitamin D deficiency both resulted in reduced cortical TGF- content in rats [73,74]. In addition, treatment of bone cells with estradiol, testosterone, or 1,25(OH)2D3 has also been reported to increase TGF- production [73 – 76]. Taken together these findings suggest that the decreased serum concentrations of sex steroid hormones with advancing age could lead to a decrease in the expression of TGF- by osteoblasts and thus could contribute to the age-related decline in the skeletal concentration of TGF-. Because physical activity also seems to regulate TGF- production, the decrease in physical activity with aging could also contribute to the observed decrease in skeletal TGF- content with age [77]. In view of the findings that the amount of TGF- in human bone decreases with age and that the osteoblast production of TGF- is regulated by sex steroid hormones and physical activity, it is tempting to speculate that the uncoupling between bone formation and resorption in the elderly could, in part, be contributed by a decreased amount of growth factors, such as TGF-, secreted by osteoblasts for contemporary use and storage in bone for later use (Fig. 2).
B. The BMP System The discovery of BMPs stems from the findings that implantation of demineralized bone matrix at ectopic sites causes an induction of heterotopic (extraskeletal) bone development [1]. Subsequently, investigation on the identification of osteoinductive molecules [78] has led to the discovery of several BMPs, including BMP-2, BMP-3, and BMP-4 [79,80], as well as BMP-7 and BMP-8 [81,82]. To date, more than 15 BMPs have been identified [10,83 – 86]. More recently, Paralkar et al [87] cloned a novel member of the BMP family from prostate, designated prostate-derived factor (PDF), which shows a structural and functional relationship to the BMPs. All these BMPs, except BMP-1, belong to the TGF- superfamily and share significant sequence homology in the carboxy-terminal region with a conserved pattern of seven cysteine residues. BMPs are synthesized as precursor forms and are cleaved at the C-terminal region to yield mature proteins. Active BMPs exist as dimers formed by disulfide bond bridging [88]. Genetic studies suggest that BMPs are produced by multifunctional genes that control not only the growth and development of skeletal tissue, but also the morphogenesis and function of other tissues and organs [10,83 – 86]. This section focuses on current understanding of the role of BMPs in regulating bone cell functions and in vivo studies on BMPs in bone regeneration (also see Chapter 13). 1. IN VITRO STUDIES Recruitment of active osteoblasts is essential for the regeneration of bone at the site of osteoclastic bone resorption. This process is dependent on the continuous supply of mature osteoblasts differentiated from hematopoietic stem cells. In this regard, treatment with BMP-2 increases ALP activity and osteocalcin production in human bone marrow stromal cells [89,90] and stimulates ALP activity and collagen synthesis in primary cultures of fetal calvarial osteoblasts [91]. The stimulatory effect of BMP-2 on osteoblast differentiation can be enhanced by ascorbic acid [91,92]. Treatment of MC3T3-E1 cells with BMP-4 also increases ALP activity [93]. This effect is enhanced by IL-1 (IL-1 alone has no effect on ALP activity) but is suppressed by TNF-. Treatment of secondary rat calvarial cultures with antisense oligodeoxynucleotides to BMP-6 suppresses glucocorticoidinduced osteoblast differentiation [94]. This inhibition can be rescued by treatment with rhBMP-6. In fetal rat calvarial cell cultures, treatment with BMP-7 dose dependently increases bone nodule formation and osteoblast marker gene expression [95]. Erlacher et al. [96] compared the effect of BMP-7 with other BMP subfamily members, i.e., the cartilagederived morphogenetic protein (CDMP)-1 (BMP-15) and CDMP-2 (BMP-13) in stimulating osteogenic differentiation and chondrogenic differentiation. It was found that these three BMP members are equally potent in stimulating
410
XUEZHONG QIN ET AL.
Model of potential role of a deficiency of TGF- in the pathogenesis of bone loss. Based on previous findings that the production of TGF- by osteoblasts is increased by treatment with sex steroid hormones [74 – 76] and that the serum levels of sex steroid hormones decrease with age, it is postulated that the osteoblast cell production of TGF- decreases with advances in age. The decrease in osteoblast cell production of TGF- is a likely explanation for why the skeletal content of TGF- is decreases with age [71]. Because TGF- treatment increases bone formation parameters in vitro and in vivo and decreases osteoclast cell survival in vitro, the underproduction of TGF- could, in part, lead to the age-related uncoupling of bone formation to resorption.
FIGURE 2
chondrogenic differentiation, whereas BMP-7 is more potent than CDMPs in promoting osteogenic differentiation in both primary fetal chondrocytes and established osteoblast cell lines. Thus, while many of the BMPs examined thus far induce differentiated functions of osteoblasts, BMP members belonging to the CDMP subfamily are more potent in stimulating chondrogenic differentiation compared with other BMP subfamily members. Studies suggest that heterodimeric BMPs are more potent than homodimers in stimulating osteoblast differentiation [97 – 99]. Compared with the BMP-4 or BMP-7 homodimer, the BMP-4/7 heterodimer more potently stimulates osteocalcin production in MC3T3-E1 cells [97] and ALP activity in rat bone marrow stromal cells [98]. Coexpression of BMP-2 with BMP-7 in Chinese hamster ovary cells has yielded a BMP-2/7 heterodimer with a specific activity of 20-fold higher than BMP homodimers in stimulating ALP production [99]. Similar results have also been obtained with the BMP-2/6 heterodimer. In addition to these in vitro data, it has also been demonstrated that these BMP heterodimers are more potent in stimulating bone formation in vivo (see
Section II,B,5). Although these research findings are intriguing, the mechanism by which these BMP heterodimers are more potent than homodimers has not yet been determined. Taken together, these findings demonstrate that BMPs consistently stimulate osteoblast differentiation, as well as chondrocyte differentiation, from stem cells and early differentiated bone cells, and that these effects can be modulated by other factors, as well as BMP heterodimer formation. 2. MECHANISM OF ACTION The actions of BMPs are mediated through the BMP receptors (BMPR), which are structurally and functionally similar to the receptors for activin and TGF-. Two types of BMP serine/threonine kinase receptors, BMPR-I and BMPR-II, have been identified. Type I BMP receptors (BMPR-I) exist as two forms: BMPR-IA and BMPR-IB (BMPR-Is). In vitro studies suggest that optimal binding of BMPs requires the coordination of both BMPR-Is and BMPR-II [85,100]. It is generally accepted that BMPR-II activates BMPR-Is and that signals are mediated through the BMPR-Is [10,85].
411
CHAPTER 14 Bone Growth Factors
Understanding of BMP signaling pathways has been enhanced greatly by the discovery of a group of molecules named “Smads,” which mediate BMP/BMPR signaling in either a positive or a negative manner. Models of BMP intracellular signal transduction have been proposed by various investigators [10,84 – 86]. The Smad-mediated signaling pathway of BMP and TGF- (Fig. 1) is very similar, except that the activated BMPR-Is specifically phosphorylate Smad 1 or Smad 5 and, possibly, Smad 8 (refer to Section II,A). Similarly, phosphorylation of these Smads is inhibited and modulated by Smads 6 and 7. Phosphorylated Smad 1 or 5 interacts with Smad 4 and enters the nucleus to activate transcription of the early BMP response genes. Overexpression of mutant Smad 5 or DPC-4/Smad 4 has been shown to block BMP-2 induction of ALP [101]. The role of Smad 5 in BMP signaling is supported by the findings that Smad 5 knockout mice exhibit phenotypes similar to those of BMP-2 and BMP-4 knockout mice [102]. Although the nuclear targets of BMP signaling are largely unknown, studies suggest that homeodomain transcription factors such as T-cell leukemia homeobox genes (Tlx), muscle segment homeobox homologue genes (Msx), and distal-less homeobox genes (Dlx) are induced by BMPs. BMP-2 has been shown to rapidly activate Tlx-2 expression in mouse embryos [103]. BMP-4 can increase the expression of Dlx-1 and -2 in dental mesenchyme [104]. In fetal mouse calvaria organ culture, BMP-4 increases the expression of both Msx-1 and Msx-2, whereas FGF-4 increases only the expression of Msx-1 [105]. In addition to these homeodomain transcription factors, the osteoblastspecific transcription factor Cbfa1 is also suggested to be an important downstream mediator of BMPs, based on the finding that BMP-7 increases Cbfa1 expression in bone cells [106]. Further studies on the relationship between these BMP-regulated transcription factors and BMP-specific Smads will be essential to enhance our understanding of the mechanisms involved in BMP signaling. 3. BMP ANTAGONISTS Studies demonstrate that the biological action of BMPs can be modulated by BMP antagonists, such as noggin and chordin [107 – 109]. Noggin, originally identified in Xenopus embryos [110], blocks BMP-4-induced ALP activity in murine bone marrow stromal cells [107] and decreases the stimulatory effects of BMP-2, -4, and -6 on DNA synthesis, collagen synthesis, and ALP activity in rat osteoblasts [111]. Studies on the mechanism by which noggin inhibits BMP activity revealed that noggin binds to BMP-2 and BMP-4 with high affinity, but not to TGF- [107]. In vitro, noggin prevents BMP-4 from binding to its receptor [107]. Noggin knockout mice are lethal and reveal severe developmental skeletal abnormalities [112]. The excess BMP activity in noggin knockout mice increases the recruitment of cells to cartilage, expanding the cartilage at the expense
of other tissues and causing larger growth plates [112]. Consistent with the idea that noggin may act to prevent excess BMP activity, Gazzerro et al. [111] found that treatment of rat osteoblasts with BMP-2 dramatically increased noggin expression as early as 2 h. Chordin is another BMP antagonist [109,113]. Similar to noggin, chordin binds to BMP-2 and -4, but not to activin or TGF- [109]. Chordin prevents the binding of BMP-4 to its receptors and, consequently, inhibits BMP-4-stimulated ALP activity in the multipotential mesodermal cell line 10T1/2 . The first mammalian form of chordin cDNA has been cloned from mice [114]. It has been speculated that chordin may play a major role during gastrulation of the mammalian embryo. The role of chordin in bone metabolism has not yet been established. Based on in vitro and in vivo findings that noggin and chordin are potent inhibitors of BMP actions, it can be speculated that the balance between the production of BMP and the production of these BMP antagonists may be crucial for the regulation of new bone formation. Thus, a better understanding of the mechanisms controlling the synthesis of these antagonists will be important in terms of improving BMP therapy or new drug discovery using inhibitors of BMP antagonists in the treatment of bone loss. 4. TRANSGENIC/KNOCKOUT/MUTANT STUDIES Characterization of the abnormalities associated with BMP gene mutations has provided important information on the in vivo role of BMPs (Table 1). BMP-2 and BMP-4 knockout mice have defects in mesoderm formation and die at an early stage of embryonic development [115,116]. BMP-4 heterozygous mice show abnormalities in various organs/tissues, including hindlimb [117]. Natural mutations in the BMP-5 gene in mice have short ears and exhibit abnormalities in the skull and axial parts of the skeleton [118]. BMP-6 knockout mice are viable and fertile with normal skeletal development at birth [119]. However, in late gestation embryos, a delay in ossification of the developing sternum is revealed, and these defects are slightly exacerbated by the disruption of both BMP-5 and BMP-6 genes. BMP-7 knockout mice die shortly after birth with developmental defects in the kidney, eyes, and skeleton [120]. BMP-2/7 and BMP-5/7 double heterozygous mice do not reveal any abnormalities, whereas BMP-4/7 double heterozygotes develop minor skeletal defects in the rib cage and the distal part of the limbs [121]. BMP-8B (not present in humans) knockout mice exhibit germ cell deficiency and infertility [122]. Natural mutations in growth and differentiation factor (GDF)-5/CDMP-1 gene cause brachypodism in mice [123] and chondrodysplasia in humans [124]. GDF-8 knockout mice exhibit increased skeletal muscle mass [125]. GDF-9 knockout mice have defects in ovarian folliculogenesis [126]. GDF-10 (BMP-3B) knockout mice do not exhibit apparent abnormalities [127]. It is suggested that GDF-10 may be functionally redundant with other
412
XUEZHONG QIN ET AL.
TABLE 1
Abnormality Associated with BMP Gene Mutations
Genetic pertubation
Phenotype
Ref.
Single BMP knockout BMP-2
Nonviable; defects in mesoderm formation and cardiac development
[115]
BMP-4
Death at early embryo, defect in mesodermal development
[116]
BMP-6
Viable, fertile, but delay in sternum ossification
[119]
BMP-7
Die shortly after birth, defect in development of multiple tissues, including bone
[120]
BMP-8B
Germ cell deficiency and infertility
[122]
GDF-8
Excessive muscle development
[125]
GDF-10
No apparent abnormalities
[127]
Similar to BMP-6 knockout
[119]
BMP-2 and BMP-7
No apparent abnormalities
[121]
BMP-5 and BMP-7
No apparent abnormalities
[121]
BMP-4 and BMP-7
Minor defect in rib cage and limbs
[121]
Double BMP knockout Homozygous BMP-5 and BMP-6 Heterozygotes
Natural mutants BMP-5
Short ear mice, abnormalities in skull and axial parts of skeleton
[118]
GDF-5 (CDMP-1)
Brachypodism in mice and chondrodysplasia in human
[123,124]
growth factor-like molecules. These genetic studies suggest that some BMPs may have distinct functions, whereas others may functionally overlap each other or have redundant functions with other growth factors. Finally, these studies emphasize that BMPs control the development of not only skeleton but also other tissues. 5. IN VIVO TREATMENT STUDIES a. Local Effects Studies of BMP actions in vivo have concentrated on the development of local delivery systems for the BMPs. Materials that have frequently served as BMP carriers include hydroxyapatite, biodegradable polymers, titanium, and fibrous glass membrane [128 – 133]. By using appropriate delivery systems, BMPs have been shown to be effective in bone regeneration in segmental fracture [134 – 136] and spinal fusion models in experimental animals [137,138]. The following briefly summarizes the findings of several representative studies. Cook et al. [135] evaluated the effect of rhBMP-7 (OP-1) using bovine collagen as a carrier, on bone formation in segmental defects in monkeys. The majority of the ulnae treated with 0.25 to 2 mg BMP-7/400 mg bovine collagen exhibited complete healing at 6 to 8 weeks. Boden et al. [138] evaluated the efficacy of using a ceramic material (hydroxyapatite-tri-calcium phosphate) loaded with rhBMP-2 as a bone graft substitute in a primate model of intertransverse spinal fusion. All monkeys treated with ceramic
blocks loaded with BMP-2 (6 to 12 mg) achieved solid fusion, whereas no fusion occurred in monkeys treated with an autogenous iliac crest bone graft. In a rabbit critical-sized defect model, insertion of a porous poly DL-lactic acid implant soaked with rhBMP-2 dose and time dependently induced new bone formation in the 20-mm-long defect [136]. By 8 weeks postimplantation, 35 or 70 g BMP-2 restored the cortices and marrow elements. When adsorbed onto porous hydroxyapatite, BMP-3 has been shown to induce rapid bone formation in calvarial defects in baboons [128]. BMP-2 and BMP-4 have also been shown to induce dentin formation by stimulating the differentiation of dental pulp cells into odontoblasts [130,139]. These studies demonstrate that BMP, when administered with a suitable carrier, effectively regenerates bone at various skeletal sites. It has been reported that a combination of BMP-2 and TGF- increases bone formation more potently than BMP-2 alone in ceramic bovine bone implanted into the thigh muscle of mice [140]. This synergistic effect is suggested to be due to their different preferential effects on cells at different stages. Ono et al. [141] reported that PGE1 promoted the osteogenic effect of BMP-2 in a rabbit model using porous hydroxyapatite ceramic pellets as the carrier. This effect was suggested to be mediated indirectly through PGE1-induced IL-6 and TGF- production. Thus, it appears that the in vivo bone-forming activity of BMPs can be mediated by other locally produced osteoregulatory factors.
413
CHAPTER 14 Bone Growth Factors
Interestingly, it has been shown that the osteoinductive effect of BMPs in vivo can be enhanced by heterodimer formation. Implantation of Xenopus BMP-4 or -7, together with a pure collagen carrier in rats, dose dependently induces new bone formation, but administration of the BMP4/7 heterodimer is more effective [98]. In addition, Israel et al. [99] reported that coexpression of various BMPs in mammalian cells led to BMP heterodimer formation. The purified BMP-2/7 or BMP-2/6 heterodimer is 5- to 10-fold more potent than BMP homodimers in inducing cartilage and bone formation [99]. These in vivo data are, thus, consistent with in vitro data showing that BMP heterodimers are more potent in stimulating osteoblast differentiation (see Section II,B,1). Future studies should emphasize the mechanism by which these BMP heterodimers exhibit more potent osteoinductive activity. Efforts to identify the molecules that stimulate BMP-2 promoter activity have led to the discovery that “statins” (which are the commonly prescribed HMG-lA reductase inhibition drugs for reducing serum cholesterol concentrations and lowering the risk of heart attack) are stimulators of BMP-2 production [142]. Statins, such as lovostatin, increased bone formation when injected subcutaneously over the calvaria of mice and increased cancellous bone volume when administered orally to rats. Whether the stimulation of bone formation by statins is solely due to their stimulatory effect on BMP-2 production remains to be studied. b. Gene Therapy Studies described earlier have demonstrated the clinical utility of BMPs in bone regeneration. However, these studies do not completely mimic the clinical situation in which defects are often much larger. Therefore, a single application of BMP may not be very effective in these situations. Moreover, this approach has been limited by the lack of availability of an ideal carrier that can provide a sustained release of BMP. Thus, BMP gene therapy offers an alternative strategy. In this regard, Fang et al. [143] reported that the implantation of degradable matrices (mainly bovine collagen) soaked with BMP4 expression plasmid led to functional union of the segmental defect in rats 9 weeks postimplantation. The fracture healing is facilitated by codelivery of the PTH(1 – 84) gene. However, further studies are warranted to evaluate the reproducibility of this strategy as the use of plasmid, in general, cannot achieve high transfection efficiency and only provides a transient expression of the therapeutic genes. Subsequently, it has been reported that the delivery of the BMP-2 gene using either the ex vivo [144 – 146] or the in vivo [147,148] approach is able to induce bone formation in muscle or heal a bone fracture. However, a significant osteoinductive effect is only observed in immunodeficient animals. The lack of significant effect in immunocompetent animals is due to the production of the adenoviral proteins encoded by the vectors,
which causes a severe immune response that ultimately inactivates the expression system. Therefore, the use of first-generation adenoviral vector (E1 and E3 deleted) to deliver BMP genes in clinical bone repair may have very limited potential. Despite the various problems associated with the current strategies, it is conceivable that BMP gene therapy may become a powerful therapeutic tool in bone repair with the development of new vector systems, such as the “gutless” adenoviral vector that does not cause a severe immune response. In addition, application of BMP gene therapy can be improved further if long-term, regulatable gene expression and systemic delivery with tissue-specific expression can be achieved in the future.
C. The IGF System IGFs represent the most abundant growth factors produced by osteoblasts and stored in bone. Previous studies have provided strong support for the concept that IGFs function in an autocrine/paracrine manner to regulate bone metabolism [149 – 152]. It has been demonstrated that the actions of IGFs on bone metabolism are modulated by multiple regulatory components of the IGF system, which includes IGF-I and IGF-II, type I and type II IGF receptors, at least six high-affinity IGF-binding proteins (IGFBP-1 to -6), and IGFBP proteases. In this regard, understanding the role of each component of the IGF system in bone metabolism is critical to understanding the pathogenic role of the IGF system in the development of osteoporosis and, also, in improving the utility of IGFs as a future therapy in the treatment of osteoporosis. Therefore, our primary focus in this section is on the scientific advancement related to studying the action of each IGF system component in bone. 1. IGFS a. In Vitro Studies IGFs have been shown to stimulate the proliferation of serum-free cultures of osteoblasts derived from the bone of various species [149 – 152]. Approximately 40 – 50% of total basal cell proliferation in serumfree bone cell culture has been shown to be attributable to IGFs [153]. IGFs also promote osteoblast differentiation based on the findings that treatment with IGFs in bone cell cultures increases the activity of ALP [153 – 156]. Hill et al. [157] demonstrated that IGF-I and IGF-II are among a few growth factors that strongly inhibit cell apoptosis in primary mouse osteoblast cultures. The antiapoptotic effect of IGFs is enhanced by PDGF, which by itself has no effect on osteoblast survival. However, Kawakami et al. [158] reported that IGF-I potentiated the Fas-mediated apoptosis of human primary osteoblasts and MG63 osteosarcoma cells. These data contrast to findings in our laboratory that IGF-II
414 dose and time dependently inhibited both basal and okadaic acid- or TNF-induced apoptosis in primary human osteoblasts and human osteosarcoma cells [159]. IGFs also stimulate collagen synthesis and decrease collagen degradation [153,155,160,161], which is consistent with the inhibitory effect of IGF-I on interstitial collagenase expression in rat osteoblast cultures [161]. These data suggest that IGFs play an important role in osteoblast proliferation, differentiation, and apoptosis. New evidence suggests that IGFs may increase osteoclastogenesis [162]. PTH-induced osteoclastogenesis and GH-induced osteoclastic resorption in vitro can be blocked significantly by IGF-I neutralization antibody [163,164]. In addition, Hou et al. [165] reported that the induction of nuclear fragmentation in serum-depleted cultures of purified mature osteoclasts was inhibited dose dependently by IGF-I in the picomolar range but not by 1 nM insulin. The stimulatory effect of IGFs on the formation and activity of both osteoblasts and osteoclasts may, in part, contribute to IGFinduced bone formation, as well as bone resorption in vivo under certain conditions. b. Mechanism of Action Although IGFs clearly play an important role in bone metabolism, the molecular mechanism by which they exert their biological actions in bone is just beginning to be understood. Accordingly, studies using cell model systems have begun to elucidate the IGF signaling pathways [166,167], and may provide important information for future studies in bone. Because it is well accepted that the major biological actions of both IGF-I and IGF-II are mediated through the type I IGF receptor, studies on the IGF signaling mechanism have predominantly been focused to that receptor and its downstream mediators. The type I IGF receptor controls cell proliferation by at least three different means: it is mitogenic, it causes transformation, and it protects cells from apoptosis [168,169]. Studies suggest that the functional domain in the type I receptor that controls each of these cellular responses is different. The domain conferring the transforming signal is located in the C terminus of the receptor, which is separated from the mitogenic signal located within the tyrosine kinase domain [168,170,171]. Type I IGF receptor mutants with disrupted kinase domain are capable of suppressing apoptosis, suggesting that the antiapoptotic domain is different from the mitogenic domain [171]. Although the postreceptor signaling mechanism for these responses is not well defined, previous studies have clearly demonstrated that type I receptor, upon binding to IGFs, and after autophosphorylation directly phosphorylates at least two classes of substrate proteins: insulin receptor substrates (IRSs) and SH-2-containing proteins (Shc). Phosphorylation of these molecules leads to activation of the MAP kinase cascade and PI-3 kinase pathways [166,167]. The role of IRS-1 in IGF receptor signaling is strongly supported by
XUEZHONG QIN ET AL.
the findings that targeted IRS-I gene disruption in mice led to a reduction in muscle PI-3 and MAP kinase activity, severe growth retardation, and resistance to both IGFs and insulin [167]. However, further downstream mediators that control various IGF biological actions remain to be identified. Because both insulin and type I IGF receptors are able to phosphorylate IRS and Shc, future identification of the downstream mediators diverging the actions of insulin and IGFs could provide important information with regard to reducing the hypoglycemic effect of IGFs. c. Transgenic/Knockout Studies IGF transgenic and knockout studies in mice have provided animal models to study the functions of IGFs in vivo. IGF-II knockout mice are apparently normal and fertile, but their body weights are only 60% of the wild type [172,173]. While survival is normal in mice lacking IGF-II, the majority of IGF-I knockout mice die shortly after birth, and surviving mice are infertile and exhibit delayed bone development [174]. In IGF-I and IGF-II double knockout mice, birth weight is further reduced, and all the mice die shortly after birth [174]. In contrast, transgenic mice overexpressing hIGF-I exhibit 30% higher body weight 4 weeks after birth than the wild type [175]. Moreover, transgenic mice overexpressing IGF-I in osteoblasts using an osteocalcin promoter to target expression exhibit increased cancellous bone volume [176]. These studies provide strong evidence that IGFs are crucial for both somatic growth and skeletal development. d. In Vivo Animal Treatment Studies To evaluate the effect of systemic IGF administration on bone formation, studies have been conducted using various animal model systems. Spencer et al. [177] showed that continuous infusion of IGF-I for 14 days in normal adult female rats increases the formation of both cortical and trabecular bone. Administration of IGF-I to 10 – 12-week-old growing rats for 4 weeks increases bone size and mineral content, although no effect on bone density is observed [178]. We have found that the systemic administration of IGF-I to 6month-old mice increases serum concentrations of bone formation markers [179]. In hypophysectomized rats, continuous IGF-I infusion for 18 days (0.3 mg/day) increases tibia epiphyseal width, longitudinal bone growth, and trabecular bone formation [180]. More relevant to osteoporosis, a number of studies have been performed to determine the effect of IGF-I treatment on osteopenia induced by ovariectomy [152]. Delivery of IGF-I via an osmotic minipump implanted subcutaneously for 6 weeks in ovariectomized rats results in a dose-dependent increase in bone mineral density in the lumbar spine and proximal and midshaft tibia [181]. Administration of human GH and IGF-I to aged ovariectomized rats prevents further loss of bone strength at sites containing trabecular bone [182]. In
415
CHAPTER 14 Bone Growth Factors
addition, IGF-I administration also increased bone formation in rats with osteopenia induced by liver cirrhosis [183]. Because hypoglycemia is a major adverse effect associated with systemic IGF administration, several studies have been performed to evaluate the efficacy of local IGF delivery on bone formation. Isgaard et al. [184] reported that local injection of IGF-I into the epiphyseal growth plate increased unilateral bone growth significantly. Local infusion of IGF-I (50 ng/day for 2 weeks) into the epiphyseal and metaphyseal junction at the distal femur of old rats increased trabecular bone volume and bone formation rate significantly without increasing bone resorption [185]. In another study [186], 12 daily injections of IGF-II (10 to 500 ng/day) into the outer periostea of parietal bones of neonatal rats dose dependently increased parietal bone mineral density and thickness. Consistent with these data, we found that a single local injection of IGF-I (0.2 or 1 g) over the parietal bone of old mice is able to increase ALP activity significantly in parietal bone extracts [179]. Taken together, these studies demonstrate that systemic or local IGF-I delivery to various skeletal sites is effective in promoting bone formation in experimental animals under both normal and pathological conditions, although the degree of effect varies among studies. e. In Vivo Human Treatment Studies A number of human clinical trials using IGF-I aimed at osteoporosis treatment have been conducted. Johansson et al. [187] first reported that bone formation and resorption markers in serum and urine are increased substantially during the treatment of osteoporotic men with rhIGF-I at a dose of 160 g/kg/day divided into two subcutaneous injections for 7 days. Subsequently, Ebeling et al. [188] showed that daily administration of IGF-I by a single subcutaneous injection for 6 days dose dependently increases the concentrations of markers for collagen synthesis and breakdown in healthy postmenopausal women. These authors demonstrated that the low dose of IGF-I (30 g/day) increases the level of bone formation but not the bone resorption markers. Johansson et al. [189,190] studied the effect of three doses of rhIGF-I (20, 40, or 80 g/kg/day) on bone turnover markers in 24 men with idiopathic osteoporosis. At a 20-g/kg dose, the increase in the serum PCP concentration remains elevated during the entire 6 weeks of injections. In contrast, urine deoxypyridinoline excretion is not altered by the same dose. Ghiron et al. [191] also found that 24 daily injections of IGF-I at 60 g/kg increases the levels of both resorption and formation markers, whereas treatment with IGF-I at a lower dose (15 g/day) only increased the levels of bone formation markers. Based on these preliminary data, a major problem with IGF treatment is that it usually increases not only bone formation but also bone resorption. Because IGFs are known to be the major mediator of GH actions, this may explain
why GH therapy also leads to an increase in both bone resorption and formation. Therefore, minimization of the stimulatory effect of IGFs on bone resorption represents an important issue in terms of improving IGF therapy. This goal may be achieved by the administration of IGF in combination with antiresorptive agents. 2. IGFBPS a. In Vitro Studies In addition to functioning as classic binding proteins (i.e., acting as transport proteins and prolonging the half-life of IGFs), recent studies demonstrate that different IGFBPs can exhibit independent actions which may either enhance or inhibit IGF actions. The following briefly describes our understanding of the biological effects of various IGFBPs in osteoblasts. Campbell and Novak [192] reported that purified IGFBP-1 inhibits IGF-I-induced cell proliferation in human MG63 osteosarcoma cells. In contrast to these results, we found that IGFBP-1, up to 30 ng/ml, had no effect on chick osteoblast cell proliferation [193]. rhIGFBP-2 inhibits IGF-I-induced bone cell proliferation at an apparent dose ratio of 10:1 [194]. The effect of IGFBP-3 seems to vary depending on the dose and culture conditions used. Ernst and Rodan [195] showed that IGFBP-3 augments the effects of IGF-I on rat osteoblast cell proliferation. We found that IGFBP-3 potentiates the IGF-I effect in osteoblasts at low concentrations, whereas it inhibits the IGF-I effect at high concentrations [196]. Among the six IGFBPs, IGFBP-4 is the only abundant osteoblast-produced IGFBP that consistently inhibits the mitogenic activity of IGFs in a variety of cell types, including osteoblasts [193,197 – 199]. In contrast to IGFBP-4, IGFBP-5 enhances IGF-induced cell proliferation in osteoblasts [199 – 201]. IGFBP-6 inhibits the mitogenic action of IGFII more potently than that of IGF-I in osteoblasts [202]. This differential effect is apparently due to the lower affinity of IGFBP-6 for IGF-I than for IGF-II. The distinct effect of each individual IGFBP, as well as the differential regulation of IGFBP production [152], may be critical in determining the degree of IGF-induced cellular responses in target tissues such as bone. b. Mechanism of Action The mechanisms by which IGFBPs modulate IGF actions in bone have not been clearly defined and may vary among IGFBPs. In this regard, evidence suggests that some of the IGFBPs may exert IGF-independent actions [199,203 – 205], in addition to modulating IGF actions [192,199,206 – 209]. Previous studies demonstrate that IGFBP – 4 acts primarily to inhibit human osteoblast proliferation by an IGF-dependent mechanism (Fig. 3), based on the following observations: (1) IGFBP-4 competes with IGF receptors for IGF binding in human osteoblasts in the monolayer culture and in purified type I IGF receptor preparations [199]; (2) IGFBP-4 has no
416
XUEZHONG QIN ET AL.
FIGURE 3
Proposed models for the actions of IGFBPs in bone cells. IGFBP-4 inhibits IGF binding to IGF receptors by binding IGFs near or at the receptor-binding site. This mechanism may also be applicable to the actions of IGFBP-1 and IGFBP-6 in osteoblasts, although further confirming studies need to be performed. With regard to IGFBP-5, three alternate models are proposed. In model 1, the complex of IGFBP5IGF binds to IGF receptors. In model 2, IGFBP-5 may bind to its own cell surface receptors and stimulate cell proliferation via IGF-independent mechanisms. In model 3, IGFBP-5-binding sites in the bone cell surface may recruit IGFs at the surface of cells, enabling the ligand to be captured easily by IGF receptors (reproduced with permission).
effect on human osteoblast proliferation induced by IGF-I or -II analogues with reduced affinity for IGFBP-4 [199]; and (3) IGFBP-4 proteolytic fragments or rhIGFBP-4 analogues with little IGF-binding activity are much less effective at inhibiting IGF-induced human osteoblast proliferation [208,209]. In addition to IGFBP-4, IGFBP-1 and IGFBP-6 may also act to inhibit osteoblast proliferation by an IGF-dependent mechanism based on the observations that: (1) IGFBP1 is less effective in inhibiting cell proliferation of MG63 cells induced by IGF analogues with low affinity for IGFBP-1 and (2) IGFBP-6 is less potent in inhibiting the mitogenic action of IGF-I than that of IGF-II (IGFBP-6 exhibits 100-fold higher affinity for IGF-II than for IGF-I) in oseoblasts [202]. IGFBP-5, which by itself associates with human osteoblast surfaces, can enhance the binding of IGFs to human osteoblasts in serum-free culture [199]. In addition, IGFBP-5 treatment further increases human osteoblast proliferation in the presence of IGF-I or IGF-II analogues with little IGFBP-5 binding affinity [199]. These findings suggest that IGFBP-5 may, in part, stimulate human osteoblast proliferation by an IGF-independent mechanism involving IGFBP-5-specific cell surface-binding sites (Fig. 3). In
addition, Jones and Clemmons [166] have proposed that association of IGFBP-5 with proteins on the cell surface or in the extracellular matrix (ECM) results in an increase in the local concentration of IGFs in the vicinity of the IGF receptor. Similarly, IGF-independent effects of IGFBP-3 in fibroblast cell cultures have also been described [203 – 205]. It has been reported that IGFBP-3 may be localized to the nuclei, although the physiological significance of this finding has not been determined [210]. It is anticipated that future studies on the role of IGFBP nuclear localization and cell surface IGFBP receptor/acceptor molecules will provide experimental data needed to determine the mechanism whereby some IGFBPs exert IGF-independent effects. c. Transgenic Studies To date, transgenic mice have been generated for IGFBP-1 [211], IGFBP-2 [212], IGFBP3 [213], and IGFBP-4 [214]. Transgenic mice overexpressing high levels of rat IGFBP-1 exhibit reduced birth weight (80 to 90% of the wild type), hyperglycemia, and impaired brain development [211,215]. IGFBP-2 transgenic mice also show reduced postnatal growth [212]. In IGFBP-3 transgenic mice, the rate of growth is not affected, although the weights of several tissues are higher than these of wild
CHAPTER 14 Bone Growth Factors
type mice [213]. Overexpression of rat IGFBP-4 in smooth muscle cells of transgenic mice using a myoblast-specific promoter causes muscle hypoplasia, a reciprocal phenotype to that of IGF-I transgenic mice [214]. Although the effects of overexpression of these IGFBPs on skeleton have not been examined, these transgenic studies suggest that IGFBP-1, -2, and -4 may play a negative role in growth regulation in vivo. d. In Vivo Treatment Studies Studies on the effect of IGFBPs on bone formation in vivo have been limited. Bagi et al. [216] first evaluated the effect of systemic administration of IGF-I alone or in combination with IGFBP-3 on bone formation in ovariectomized rats. At the highest dose used in this study, IGF-I (7.5 mg/kg), with an equimolar amount of IGFBP-3, significantly promotes bone formation. However, it is not known whether this effect is indeed due to the presence of IGFBP-3, as a treatment group receiving IGF-I only at this dose (7.5 mg/kg) was not included in this study. Subsequently, Narusawa et al. [217] compared the effect of administration of rhIGF-I alone or rhIGF-I/IGFBP-3 complex on bone formation in rats with combined ovariectomy and bilateral sciatic neurectomy. Injection of IGF-I alone three times a week for 4 weeks (0.3 or 3 mg/kg) did not significantly increase the bone formation rate. However, injection of the same dose of IGF-I along with an equimolar amount of IGFBP-3 significantly increased bone formation rate and cancellous bone volume of the lumbar vertebrae. These observations suggest that IGFBP-3 can potentiate the effect of IGF-I on bone formation (although the issue of whether IGFBP-3 alone affects bone formation was not addressed in this study). Regarding IGFBP-5, it has been demonstrated that nine daily subcutaneous injections of rhIGFBP-5 (50 g/day) in mice increase serum osteocalcin values as effectively as administration of IGF-I alone (13 g/day) or IGFIrhIGFBP-5 complex [218]. In addition, a single systemic administration of IGFBP-5 alone increased bone formation markers. Future studies using various doses of IGFBP-5 with or without IGF are needed to determine if the systemic administration of IGFBP-5 and IGF-I can cause an additive effect on bone formation. It is anticipated that further detailed studies on the effect of IGFBP, alone or in combination with IGFs, on bone formation could lead to increased understanding of the physiological role of each IGFBP in vivo and could provide important information for future optimization of IGF therapy in the treatment of osteoporosis and other diseases. 3. IGFBP PROTEASES In essentially all body fluids, IGFs form complexes with IGFBPs. The release of IGFs from these inactive IGF/IGFBP complexes is a prerequisite for IGFs to bind to their receptors and elicit a biological response. One mech-
417 anism would be through the degradation of IGFBPs by IGFBP proteases (Fig. 4). Although proteolysis of various IGFBPs by human osteoblast conditioned medium has been reported [152], the IGF-dependent IGFBP-4 protease produced by human osteoblasts is studied most extensively. This protease is unique in that the proteolysis of IGFBP-4 by cell-free human osteoblast CM can be enhanced by the addition of exogenous IGFs [207,219 – 221]. Based on the following, our characterization of human osteoblast-produced IGFBP-4 protease suggests that binding of IGF-II to IGFBP-4 may alter the IGFBP-4 conformation such that the cleavage site is more accessible to the protease (1) des [1– 6] IGF-II, with at least 200-fold reduced affinity for IGFBP-4, is much less effective than intact IGF-II in promoting IGFBP-4 proteolysis by human osteoblast CM [220] and (2) the IGFBP-4 analogue with deletion of Leu73, Met73, and His74 from the IGF- binding domain exhibits undetectable IGF-binding affinity and is not cleaved by the partially purified IGFBP-4 protease from human osteoblast CM [209]. Because IGFBP-4 is one of the most abundant IGFBPs produced by hOBs and its molar concentration in human osteoblast CM is at least 10-fold higher than that of IGF-II, we postulate that the IGFBP-4 protease may play an important role in the regulation of human osteoblast proliferation. This notion is supported by findings that protease-resistant IGFBP-4 analogues, which exhibit normal IGF-II and IGF-I binding affinity, are much more potent than the-wild-type IGFBP-4 in blocking IGF-II mitogenic actions in normal human osteoblast cultures, but not in the culture of MG63 cells, which do not produce IGFBP-4 protease [221]. In addition, IGFBP-4 protease activity has been known to be regulated by other osteoregulatory agents in addition to IGFs, such as TGF- and estrogen [219]. Lawrence et al. [222] reported that the IGF-II-dependent IGFBP-4 protease produced by human fibroblasts is identical to the previously identified pregnancy-associated plasma protein (PAPP)-A. Consistent with this new finding, we found that proteolysis of IGFBP-4 by human pregnancy serum is IGF-II dependent [223]. Moreover, pregnancy serum and human osteoblast-produced IGFBP-4 protease recognized the same cleavage site (Met135/Lys136). These data suggest that PAPP-A represents the major IGF-II dependent IGFBP-4 protease produced by human osteoblasts. Since the PAPPA cDNA has been cloned [224], the role of PAPP-A in osteoblast functions can be further confirmed by blocking its production using the antisense approach. In addition to the IGFBP-4 protease, human osteoblasts also produce a metal-dependent serine protease capable of degrading IGFBP-5 but not other IGFBPs [220,225]. However, the identity of the human osteoblast-produced IGFBP5 protease has not been determined. Based on findings that the activity of the IGFBP-5 protease changes in response to treatment with growth factors and osteoregulatory agents in
418
XUEZHONG QIN ET AL.
FIGURE 4 Potential role of IGFBP proteases in modulating IGF bioavailability to local tissues. Proteolysis of IGFBP-3 by serum protease leads to the disruption of IGF/IGFBP-3/ALS (available substance) complex and the release of IGFs. The released IGFs are either directly transported to the target tissues or bind to other IGFBPs to form smaller IGF/IGFBP complexes, which can cross the endothelium into the extracellular space. Upon transport into the extracellular space, IGFs are released from inactive IGFBP/IGF complexes (e.g., IGF/IGFBP-4] through action of the locally produced IGFBP proteases (e.g., the IGF-dependent IGFBP-4 protease/PAPP-A). However degradation of stimulatory IGFBPs, such as IGFBP-5, may attenuate the action of IGFs. Thus, the regulation of IGFBP protease production/activity may play a key role in controlling the availability of IGFs to the target tissues.
osteoblasts [220,225 – 228] and that IGFBP-5 proteolytic fragments are much less active compared to the intact IGFBP-5 in stimulating osteoblast proliferation [199], it can be proposed that the IGFBP-5-specific protease is also an important regulatory component of the IGF system in human osteoblasts. The regulation of IGFBP degradation and synthesis may be equally important in controlling the bioavailability of IGFBPs and, consequently, IGF actions in the local bone environment. Studies on the physiological role of IGFBP proteases have just begun and new understanding will emerge as the identity of various IGFBP proteases become established. 4. ROLE OF IGF SYSTEM IN THE PATHOGENESIS OF OSTEOPOROSIS The development of osteoporosis is determined by at least two major parameters: (1) peak bone density achieved during or shortly after puberty and (2) rate of bone loss with aging, particularly after menopause. These two parameters appear
to be affected by the IGF system. Studies demonstrate that the serum concentrations of IGF-I, IGFBP-3, and IGFBP-5, but not IGF-II, showed a significant positive correlation to bone mineral content, bone density, and metacarpal indexes during puberty in girls [229]. In addition, serum concentrations of IGFBP-3 and IGFBP-5 showed a significant positive correlation to the level of bone formation marker, osteocalcin. These data suggest that the IGF system may play an important role in skeletal development during sexual development, a stage of life that determines peak bone density. Although it has been well established that bone formation decreases with age, the mechanism which underlies this process awaits full elucdation. In this regard, serum concentrations of IGF-I are lower in elderly individuals [230,231] and postmenopausal women [232]. Serum IGF-I values are reduced in men with idopathic osteoporosis and correlate with lumbar spine bone mineral density [231]. Serum levels of IGF-I and other stimulatory IGF system components are significantly lower in hip fracture patients than in agematched controls [233]. The IGF-I content of cortical bone
419
CHAPTER 14 Bone Growth Factors
also decreases with age [71]. In addition, concentration of IGFBP-5 (a stimulator of IGF actions) in serum and bone decrease with aging [234]. In contrast, serum levels of IGFBP4 (an inhibitor of IGF actions) are elevated in elderly subjects with low bone density and correlate with serum concentrations of PTH [235,236]. Consistent with the hypothesis that the increase in serum PTH in osteoporotic patients may lead to overproduction of IGFBP-4, we found that treatment of human osteoblasts with PTH significantly increases IGFBP-4 concentration in the conditioned medium [236]. In addition to these age-related changes in the production of IGF system components, bone cells derived from aged rats [237] or humans [238,239] are much less responsive to IGFs than bone cells derived from their younger counterparts. Based on these findings, it is conceivable that the age-related decline in bone formation may be due, at least in part, to the overproduction of an inhibitory IGF component (i.e., IGFBP-4], underproduction of stimulatory IGF system components (IGF-I and IGFBP-5], and the decreased osteoblast responsiveness to IGFs in aged individuals. In addition to age-related osteoporosis, glucocorticoid induced osteoporosis represents a significant medical problem (see also Chapter 44). Although the mechanism for the glucocorticoid-induced bone loss is not clear, the impairment of the GH – IGF axis appears to play a role. Although results are not entirely consistent among studies, treatment of bone cells with glucocorticoid generally decreases the expression/production of stimulatory components of the IGF system, including IGF receptors, IGF-I, IGF-II, IGFBP-3, and IGFBP-5, but increases the inhibitory components of this system, such as IGFBP-6 [233,240 – 245]. Dexamethasone dose dependently suppressed basal, as well as GH or IGF-I stimulated rat chondrocyte proliferation and the GH stimulated IGF-I production in these cells [245]. In vivo, methylprednisolone treatment dose dependently decreased the concentrations of serum-free IGF-I in rats [246]. In a human longitudinal prospective study, glucocorticoid therapy decreased serum bone formation parameters and serum concentrations of IGF-I, IGF-II, and IGFBP-3 markedly [247]. In aggregate, these in vitro and in vivo studies provide evidence that the IGF system may be involved in the pathogenesis of osteoporosis induced by glucocorticoid treatment.
D. The FGF System Fibroblast growth factors belong to a large family of heparin-binding proteins that regulate cell proliferation, differentiation, and migration in many different tissues and play key roles during physiological and pathological conditions, such as wound healing, skeletal repair, neovascularization, and tumor growth [248]. FGFs were first isolated in the 1970s from bovine brain extracts based on their mitogenic and angiogenic activities [249], and currently 19
unique family members have been identified on the basis of their strong sequence homology in a core region of 120 amino acids. FGF1 (acidic FGF) and FGF2 (basic FGF) were the first two members of this family to be characterized and are the most widely studied in many tissues, including bone. In contrast to the almost ubiquitous tissue distribution and action of FGF2 and the wide distribution of FGF1, many other FGF family members have more restricted spatial and temporal expression patterns. Originally, FGF3 – 6 were isolated as protooncogenes from tumor cells [250 – 253]. FGF7 was originally known as keratinocyte growth factor, based on its identification as a selective mitogen for epithelial cells but not fibroblasts [254]. FGF8, or androgen-induced growth factor (AIGF), was isolated from a murine androgen-dependent carcinoma [255]. FGF9, originally termed GAF or glial-activating factor, was isolated from a human glioma cell line [256]. A group of 4 fibroblast homologous growth factors (FHFs) implicated in nervous system development were designated FGF11 – 14 [257]. The number of newly discovered FGFs has been growing rapidly, with FGF19 being the most recently added member [258]. Undoubtedly, more members of the FGF family will be discovered in the future and, with these discoveries, new functions are also likely to emerge. Work performed in many laboratories has established that FGFs play many important physiological roles in bone growth, remodeling, and repair. Both FGF1 and FGF2 stimulate osteoblast proliferation and promote bone growth. They are also synthesized by osteoblasts and are stored in bone. Because FGF2 is more potent than FGF1, most of the in vitro and in vivo studies have been performed with FGF2. In vitro, FGF2 has been shown to stimulate the proliferation of osteoblasts, chondrocytes, and periosteal cells, and to stimulate the formation of mineralized bone-like nodules in cultures of bone marrow stromal cells [259 – 262]. Based on these data and the following studies, it is suggested that FGF2 may play a role in normal bone remodeling and, in addition, may have therapeutic use in fracture repair and in the treatment of other local skeletal defects. 1. IN VITRO STUDIES FGF1 and FGF2 strongly stimulate the proliferation of bone cells under most culture conditions, but reduce markers of the differentiated phenotype, such as alkaline phosphatase and PTH-responsive adenylate cyclase activity [263,264]. In bovine, rodent, and human calvaria-derived cells, the effects on bone cell proliferation and differentiation appear to be uncoupled, indicating that FGFs have independent effects on cell replication and differentiation. The effects of FGF on osteoblastic cell differentiation and bone matrix formation are, however, conflicting. FGF2 has been reported to either stimulate or inhibit the production of type I collagen, the major differentiated product of
420 the osteoblast [265,266]. Two factors that influence the type of response are duration of FGF exposure and maturational stage of the cells being treated. Thus, short-term (24 h) treatment with FGF2 results in the stimulation of collagen I, whereas long-term exposure of osteoblast cultures to FGF2 inhibits type I collagen synthesis. Such inhibition by FGF appears to be transcriptional in nature and is mediated by DNA sequences in the promoter of the alpha1(I) collagen gene [267]. Further, the effect of FGF2 on human calvarial osteoblast cells is influenced by the level of cell maturation. FGF2 reduces the expression of osteoblast markers in less mature cells, while increasing osteocalcin production and matrix mineralization in more mature cells [268]. These observations indicate that FGF does not have a consistent effect on osteoblastic cell differentiation in vitro. Studies of the mechanism of action of FGF on osteoblasts should help clarify the function of FGF on osteoblastic differentiation. The observation of a dose-dependent biphasic effect of FGF2 on long bone growth has prompted several in vitro studies with isolated chondrocytes in order to elucidate this mechanism. In vitro, FGF2 inhibits terminal differentiation and hypertrophy of chondrocytes. Cultured chondrocytes that undergo terminal differentiation have fewer high-affinity binding sites for FGF [269]. This suggests that the action of FGF may be partially regulated at the level of receptor number. Similar to the situation with osteoblasts, there are discrepancies between different studies regarding the effect of FGF on thymidine incorporation and collagen synthesis in chondrocytes [270]. A possible explanation for the failure of these in vitro experiments to consistently represent the in vivo situation is that a cell culture system does not maintain the three-dimensional cellular interaction of the highly organized growth plate. The use of an organ system circumvents many of the limitations of cell culture. The addition of FGF2 to organ cultures from rat metatarsals inhibits longitudinal growth, which is consistent with the effect in vivo [271]. 2. MECHANISM OF ACTION The biological effects of FGFs are mediated through four high affinity transmembrane kinase receptors, known as FGFR1 – FGFR4. FGFRs structurally resemble other transmembrane protein kinase receptors and have been found in all contemporary vertebrates [272]. Four human genes encoding FGFR1 – FGFR4 have amino acid identities of 55 – 72% [273]. The receptors are monomeric in their native state and dimerize after binding with an FGF ligand. Dimerization activates tyrosine kinase and triggers downstream effects through multiple signaling pathways. The details of these pathways have not been fully established for all the FGFRs. Studies with FGFR1 and FGFR3 suggest that activation of Ras-dependent pathways results in the stimulation of mitogenesis, whereas STAT signaling
XUEZHONG QIN ET AL.
pathways are inhibitory [274]. Another signaling pathway may involve internalization and nuclear translocation of the ligand – FGFR complex [275]. An interesting aspect of the FGFR gene family is the existence of multiple variants generated by alternative splicing of mRNA, which can result in over 100 different protein sequences. Some of the variants have distinct affinities for individual FGF ligands and use different signaling pathways [276]. Taken together, the complex genetic organization of both FGF ligands and receptors presents a great number of regulatory mechanisms to mediate the biological effects of this extensive family of growth factors. This modulation can occur at transcriptional, translational, posttranslational, or signal pathway levels, thereby providing the means for fine-tuning the many biological processes that utilize FGF. In addition to high-affinity FGFRs, FGF also binds to heparin sulfate proteoglycans (HSPGs) [277] and to a lowaffinity, cysteine-rich transmembrane FGF-binding protein [278]. HSPGs are sulfated glycosaminoglycans bound to a core protein localized in the extracellular matrix. At this location, HSPGs may provide an extracellular storage compartment protecting FGFs from degradation. In response to external or internal stimuli, such as phosphorylation or proteolytic processing of the extracellular matrix, HSPGs may regulate the bioavailability of FGFs. Many studies have indicated an enhancement by HSPGs for FGF-induced biological activities. The most likely mechanism for this effect is by increasing the affinity of FGFs for FGFRs [279,280]. Heparin contains the same sugar residues as HSPGs and is often used in experiments to enhance FGF stability and potency. The role of the low-affinity, cysteine-rich binding protein has not been well studied. It binds FGF at a site that overlaps the high-affinity receptor binding site, as binding to both receptor types is mutually exclusive. A possible role for the low-affinity receptor may be in regulating intracellular FGF levels [281]. 3. TRANSGENIC/KNOCKOUT STUDIES Much can be learned about the role of FGF and FGFRs from transgenic studies or from the analysis of natural mutations [282]. Transgenic mice overexpressing FGF2 exhibit chondrodysplasia, which is characterized by long bones that are considerably shorter than those of wild-type animals [283]. Chondrodysplasia in animals overexpressing FGF2 is similar to skeletal malformations observed in human syndromes associated with mutations in FGFRs. Therefore, a possible explanation for this phenomenon is that high concentrations of FGF2 selectively activate certain subsets of FGFRs that exert a negative effect on bone growth. Surprisingly, FGF2 knockout mice were indistinguishable in appearance from their normal littermates [284], indicating that other members of the FGF family may be able
CHAPTER 14 Bone Growth Factors
to substitute for some FGF2 functions during early bone formation. However, significant neuronal defects in FGF2 knockout mice were observed in the neocortex, most notably a reduction in the neuronal density and disruption of the cytoarchitecture. In addition, the healing of skin wounds was delayed significantly in mice lacking FGF2. These results indicate that FGF2, while not essential for embryonic development, plays specific roles in cortical neurogenesis and wound healing that cannot be carried out by other FGF family members. Because of the redundancy of FGF signaling, it is difficult to assess the role of any particular member of the FGF family by elimination or overexpression of a single gene. Only specific functions that cannot be substituted by other FGFs will be manifest in obvious defects. FGF4, FGF8, and FGF10 appear to be the most crucial regulating factors during embryonic development. Mice lacking FGF4 or FGF8 are both embryonic lethal, whereas a targeted disruption in FGF10 resulted in mutant embryos without limbs [285 – 287]. The action of FGF8 and FGF10, both of which are expressed during limb development, is most likely mediated through FGFR2, as shown by expression of a soluble dominant-negative FGFR2 gene [288]. Knockout studies of all four high-affinity FGFRs have shown individually that FGFR1 and FGFR2 are the only receptors that are involved in limb development and are embryonic lethal [282,289]. FGFR1 mutants died early in gastrulation, displaying growth defects and axial mesodermal disorganization. In contrast to knockouts for FGFR1 and FGFR2, null mutations for FGFR3 and FGFR4, or both, have normal limbs. Of great interest regarding the regulation of bone growth was the finding that FGFR3 knockouts displayed excessive growth of long bones and vertebrae, indicating enhanced endochondral bone formation [290,291]. These findings have led to the suggestion that the FGFR3 kinase is a negative regulator of bone growth that acts by limiting endochondral ossification. This explanation is supported by observations of certain naturally occurring mutations of FGFR3 that cause short stature (achondroplasia and thanatophoric displasia). Because these syndromes exhibit increased ligand-independent activity of FGFR3, it is likely that the mutations act by enhancing the negative control normally exerted by FGFR3. Therefore, the identification of distinct FGF receptor mutations in individuals with human skeletal disorders provides supporting evidence that FGFs play an important role in growth and development of the skeleton. 4. IN VIVO TREATMENT STUDIES a. Local Applications During fracture repair, FGF1 and FGF2 are expressed at the earliest stage in macrophages and periosteal cells, and in osteoblasts and chondrocytes in later stages [49]. These observations provided the rationale for applications of FGF directly to the
421 fracture site in order to improve the healing process. In a diabetic rat model, where fracture healing is impaired due to reduced expression of FGF2 during early stages, a single application of recombinant FGF2 increased the volume and mineral content of the callus in both normal and diabetic rats [292]. In addition, FGF2 also provided a marked improvement in the mechanical stability of the healing fibula in all animals. While one study found no effect on fracture healing in the rabbit tibia with a single injection of FGF1 or FGF2 [293], more recent investigations have reported improved fracture healing by FGF in several species. In rabbits, single injections of FGF2 at 100 g or above in a 3-mm tibial bone defect resulted in increased volume and mineral content of newly made bone after 5 weeks [294]. A similar model in dogs used 200g of FGF2 and showed increased membraneous ossification after 2 weeks. After 4 weeks, the callus in the FGF-treated group was larger than in the vehicle-treated control group. Additionally, the fracture strength by week 16 was greater in FGF-treated animals than in controls [295]. Interestingly, the FGF group also had a markedly increased osteoclast number in the periosteal callus. This suggests that FGF2 stimulates both callus formation and osteoclastic resorption, thereby promoting fracture healing by stimulating bone remodeling. A single local application of FGF was also found to promote healing in a primate fracture model. As measured by radiography and mechanical and histological critera, a single injection of FGF2 in a stabilizing gel of hyaluronic acid was found to be more effective at promoting fracture healing in baboon fibulae than FGF2 alone or injected in fibronectin gel [296]. Based on these findings, it can be postulated that local delivery of FGF2 or FGF1 directly into the fracture site immediately after the injury may substantially enhance fracture healing in patients with both normal or impaired ability for fracture repair. In addition to promoting fracture healing, local application of FGFs has also been shown to promote regeneration at other sites. Intraosseous injection into the femur of ovariectomized rabbits to a depth of 3 – 5 mm from the cortex stimulated intraosseous bone formation after 4 weeks in a dose-dependent manner [297]. Slow release delivery methods for FGF are currently being developed aimed at increasing the efficacy of this growth factor. Specifically, FGF2 embedded in a gelatin-based hydrogel promoted complete closure within 21 days of a skull defect in primates [298]. In contrast, no closure occurred when the same dose of FGF2 was applied in solution. This type of experiment suggests that local delivery of FGF to sites of degeneration, such as the hip in generalized osteoporosis, may be used to promote new bone formation at this site and reduce the risk for hip fractures. b. Systemic Applications Similar to local application of FGFs, systemic application also promotes bone growth
422 and restoration in many systems. FGF1 prevented bone loss and increased new bone formation when injected at 0.2 mg/kg daily for 28 days into the tail vein of adult ovariectomized female rats [299]. FGF1 had substantial anabolic effects in sham-operated animals, where bone density increased two-fold. Without treatment, severe bone loss and trabecular disruption occurred in ovariectomized animals 6 months after the operation, similar to that seen in patients with osteoporosis. In these rats, FGF1 treatment induced extensive new woven bone formation with new trabecular-like structures, and bone density in the tibial metaphysis increased three-fold. More complex effects of systemically applied FGF on bone growth are suggested by the observation that systemic administration of FGF2 to growing rats events a dose-dependent biphasic effect on longitudinal bone growth [300]. In this study, daily intravenous injections of FGF2 at 0.1mg/kg in growing rats increased longitudinal bone growth and cartilage cell proliferation. In contrast, a higher dose (0.3 mg/kg) reduced the rate of bone growth and cartilage proliferation. As the rate of longitudinal bone growth depends on the rate of growth plate chondrogenesis, FGF2, at higher doses, may inhibit one of the cellular processes underlying chondrogenesis [271]. Pleiotropic effects of FGF on its target tissues throughout the body may preclude the clinical application of this very exciting therapeutic agent to systemically regenerate the skeleton. Such effects on nonskeletal tissues include hypotension [301] and hypertrophy of epithelial tissues, kidney, and lung, some of which would be considered adverse effects in patients who require only skeletal regeneration. Systemic application of FGF with gene therapy methods designed specificially to increase bone formation may circumvent such undesirable side effects of FGF action by targeting the vector delivery, or FGF expression, exclusively to bone tissue. New therapies using the growth factor potential of FGFs are currently being actively pursued, not only to stimulate bone formation, but also to stimulate new blood vessel formation in peripheral vascular disease and myocardial ischemia [302,303]. It seems likely that, in the future, FGF gene targeting to skeletal tissue will be accomplished so that we can realize the remarkable skeletal regenerative capacity of this growth factor.
III. CONCLUSIONS Growth factors, such as TGF-, BMP, IGF, and FGF, play a unique role in bone in that they are not only actively produced by bone cells for contemporary use, but are also stored in bone as an immediately available source of growth factors for future use when the old bone is destroyed by resorption. One or both of these mechanisms for growth factor elaboration in bone may be involved in the
XUEZHONG QIN ET AL.
several key physiological processes in bone metabolism: namely, development of peak bone density, bone remodeling (in which the coupling of bone formation to bone resorption plays a pivotal role), fracture repair, and the adaptive response in terms of density and architecture that bone exhibits in response to mechanical strains. Based on these above observations, it is not surprising that bone growth factor deficiencies have also been implicated in the pathogenesis of disease processes, such as osteoporosis. In this regard, both deficiencies of TGF- and the IGFs have been associated with osteoporosis. If a deficiency of growth factor can lead to bone loss, it follows that treatment with a growth factor might be an effective therapy to promote bone gain in osteoporosis. With regard to growth factor therapy, it is clear from animal studies that a number of growth factors, including TGF-, BMP, IGF, and FGF, can locally stimulate bone formation in vivo in a number of animal models, and it seems likely that these agents could be used in the near future in human conditions requiring stimulation of local bone growth, such as fracture healing and the treatment of local lesions in bone. However, the potential usefulness of skeletal growth factors to treat bone loss in generalized skeletal disorders, such as osteoporosis, requires more knowledge, as these multifunctional factors affect diverse tissues besides bone. Therefore, from the standpoint of safety, TGF-, BMP, and FGF can now be administered only locally. In the future it may be possible to develop the necessary methodologies to target these growth factors for the skeleton or specific sites in the skeleton. In this regard, the development of suitable delivery systems to target growth factors to areas of interest would be very critical. In contrast to TGF-s, BMPs, and FGFs, IGFs are produced locally as growth factors and also circulate in body fluids as hormones. Therefore, the IGFs may represent a growth factor group that can possibly be used to treat bone loss via systemic administration. Because IGF action can be enhanced by certain IGFBPs, the combination IGF/IGFBPs may prove more useful than IGF alone in the treatment of bone loss. In addition, simultaneous administration of antiresorptive agents may help reduce the bone-resorbing activity of IGF. This important area should demand attention in the future.
References 1. M. R. Urist, Bone: Formation by autoinduction. Science 150, 893 – 899 (1965). 2. P. V. Hauschka, A. E. Mavrakos, M. D. Iafrati, S. E. Doleman, and M. Klagsbrun, Growth factors in bone matrix: Isolation of multiple types by affinity chromatography on heparin-sepharose. J. Biol. Chem. 261, 12665 – 12674 (1986). 3. S. Mohan and D. J. Baylink, Bone growth factors. Clin. Orthop. 263, 30 – 48 (1991).
CHAPTER 14 Bone Growth Factors 4. Y. Ogawa, D. K. Schmidt, J. R. Dasch, R. J. Chang, and C. B. Glaser, Purification and characterization of transforming growth factor-beta 2.3 and -beta 1.2 heterodimers from bovine bone. J. Biol. Chem. 267, 2325 – 2328 (1992). 5. D. J. Baylink, R. D. Finkelman, and S. Mohan, Growth factors to stimulate bone formation. J. Bone Miner. Res. 8 (Suppl. 2), S565 – S572 (1993). 6. E. Canalis, J. Pash, and S. Varghese, Skeletal growth factors. Crit. Rev. Eukaryot. Gene Expr. 3, 155 – 166 (1993). 7. G. R. Mundy, Factors which stimulate bone growth in vivo. Growth Regul. 3, 124 – 128 (1993). 8. M. Centrella, M. C. Horowitz, J. M. Wozney, and T. L. McCarthy, Transforming growth factor-beta gene family members and bone. Endocr. Rev. 15, 27 – 39 (1994). 9. A. H. Reddi, Bone morphogenetic proteins, bone marrow stromal cells, and mesenchymal stem cells: Maureen Owen revisited. Clin. Orthop. 313, 115 – 119 (1995). 10. T. Sakou, Bone morphogenetic proteins: From basic studies to clinical approaches. Bone 22, 591 – 603 (1998). 11. D. J. Baylink, J. C. Jennings, and S. Mohan, Calcium and bone homeostasis and changes with aging In “Principles of Geriatric Medicine” (W. R. Hazzard, J. P. Blass, W. H. Ettinger, J. B. Halter, and J. G. Ouslander, eds.), pp. 1041 – 1056 (1999). 12. P. P. Hu, M. B. Datto, and X. F. Wang, Molecular mechanisms of transforming growth factor-beta signaling. Endocr. Rev. 19, 349 – 363 (1998). 13. J. C. Jennings and S. Mohan, Heterogeneity of latent transforming growth factor-beta isolated from bone matrix proteins. Endocrinology 126, 1014 – 1021 (1990). 14. L. F. Bonewald and S. L. Dallas, Role of active and latent transforming growth factor beta in bone formation. J. Cell. Biochem. 55, 350 – 357 (1994). 15. A. H. Tashjian, Jr., E. F. Voelkel, M. Lazzaro, F. R. Singer, A. B. Roberts, R. Derynck, M. E. Winkler, and L. Levine, Alpha and beta human transforming growth factors stimulate prostaglandin production and bone resorption in cultured mouse calvaria. Proc. Natl. Acad. Sci. USA 82, 4535 – 4538 (1985). 16. M. Centrella, T. L. McCarthy, and E. Canalis, Transforming growth factor beta is a bifunctional regulator of replication and collagen synthesis in osteoblast-enriched cell cultures from fetal rat bone. J. Biol. Chem. 262, 2869 – 2874 (1987). 17. M. Noda and G. A. Rodan, Type-beta transforming growth factor inhibits proliferation and expression of alkaline phosphatase in murine osteoblastlike cells. Biochem. Biophys. Res. Commun. 140, 56–65 (1986). 18. E. C. Breen, R. A. Ignotz, L. McCabe, J. L. Stein, G. S. Stein, and J. B. Lian, TGF beta alters growth and differentiation related gene expression in proliferating osteoblasts in vitro, preventing development of the mature bone phenotype. J. Cell. Physiol. 160, 323 – 335 (1994). 19. M. Centrella, J. Massague, and E. Canalis, Human platelet-derived transforming growth factor-beta stimulates parameters of bone growth in fetal rat calvariae. Endocrinology 119, 2306 – 2312 (1986). 20. J. Pfeilschifter and G. R. Mundy, Modulation of type beta transforming growth factor activity in bone cultures by osteotropic hormones. Proc. Natl. Acad. Sci. USA 84, 2024 – 2028 (1987). 21. Z. Schwartz, L. F. Bonewald, K. Caulfield, B. Brooks, and B. D. Boyan, Direct effects of transforming growth factor-beta on chondrocytes are modulated by vitamin D metabolites in a cell maturation-specific manner. Endocrinology 132, 1544 – 1552 (1993). 22. Z. Schwartz, V. L. Sylvia, Y. Liu, D. D. Dean, and B. D. Boyan, Treatment of resting zone chondrocytes with transforming growth factor- beta 1 induces differentiation into a phenotype characteristic of growth zone chondrocytes by downregulating responsiveness to 24,25-(OH)2D3 and upregulating responsiveness to 1,25-(OH)2D3. Bone 23, 465 – 470 (1998).
423 23. V. L. Sylvia, S. Mackey, Z. Schwartz, L. Schuman, R. Gomez, and B. D. Boyan, Regulation of protein kinase C by transforming growth factor beta 1 in rat costochondral chondrocyte cultures. J. Bone Miner. Res. 9, 1477 – 1487 (1994). 24. K. A. Skantze, C. E. Brinckerhoff, and J. P. Collier, Use of agarose culture to measure the effect of transforming growth factor beta and epidermal growth factor on rabbit articular chondrocytes. Cancer Res. 45, 4416 – 4421 (1985). 25. P. C. Yaeger, T. L. Masi, J. L. de Ortiz, F. Binette, R. Tubo, and J. M. McPherson, Synergistic action of transforming growth factor-beta and insulin-like growth factor-I induces expression of type II collagen and aggrecan genes in adult human articular chondrocytes. Exp. Cell Res. 237, 318 – 325 (1997). 26. A. Martin, F. J. Unda, C. Begue-Kirn, J. V. Ruch, and J. Arechaga, Effects of aFGF, bFGF, TGFbeta1 and IGF-I on odontoblast differentiation in vitro. Eur. J. Oral Sci. 106 (Suppl. 1), 117 – 121 (1998). 27. J. E. Wergedal, T. Matsuyama, and D. D. Strong, Differentiation of normal human bone cells by transforming growth factor-beta and 1,25(OH)2 vitamin D3. Metabolism 41, 42 – 48 (1992). 28. Y. Chen, A. Takeshita, K. Ozaki, S. Kitano, and S. Hanazawa, Transcriptional regulation by transforming growth factor beta of the expression of retinoic acid and retinoid X receptor genes in osteoblastic cells is mediated through AP – 1. J. Biol. Chem. 271, 31602 – 31606 (1996). 29. C. Pilbeam, Y. Rao, O. Voznesensky, H. Kawaguchi, C. Alander, L. Raisz, and H. Herschman, Transforming growth factor-beta1 regulation of prostaglandin G/H synthase-2 expression in osteoblastic MC3T3-E1 cells. Endocrinology 138, 4672 – 4682 (1997). 30. J. Massague, TGF-beta signal transduction. Annu. Rev. Biochem. 67, 753 – 791 (1998). 31. R. Derynck, Y. Zhang, and X. H. Feng, Smads: Transcriptional activators of TGF-beta responses. Cell 95, 737 – 740 (1998). 32. Z. Q. Wang, C. Ovitt, A. E. Grigoriadis, U. Mohle-Steinlein, U. Ruther, and E. F. Wagner, Bone and haematopoietic defects in mice lacking c-fos. Nature 360, 741 – 745 (1992). 33. N. Udagawa, J. Chan, S. Wada, D. M. Findlay, J. A. Hamilton, and T. J. Martin, c-Fos antisense DNA inhibits proliferation of osteoclast progenitors in osteoclast development but not macrophage differentiation in vitro. Bone 18, 511 – 516 (1996). 34. S. J. Felts, E. S. Stoflet, C. T. Eggers, and M. J. Getz, Tissue factor gene transcription in serum-stimulated fibroblasts is mediated by recruitment of c-Fos into specific AP-1 DNA-binding complexes. Biochemistry 34, 12355 – 12362 (1995). 35. G. Jin and P. H. Howe, Regulation of clusterin gene expression by transforming growth factor beta. J. Biol. Chem. 272, 26620 – 26626 (1997). 36. Y. Zhang, X. H. Feng, and R. Derynck, Smad3 and Smad4 cooperate with c-Jun/c-Fos to mediate TGF-beta-induced transcription. Nature 394, 909 – 913 (1998). 37. A. Takeshita, K. Imai, S. Kato, S. Kitano, and S. Hanazawa, 1alpha,25-dehydroxyvitamin D3 synergism toward transforming growth factor-beta1-induced AP – 1 transcriptional activity in mouse osteoblastic cells via its nuclear receptor. J. Biol. Chem. 273, 14738 – 14744 (1998). 38. L. J. Jonk, S. Itoh, C. H. Heldin, P. ten Dijke, and W. Kruijer, Identification and functional characterization of a Smad binding element (SBE) in the JunB promoter that acts as a transforming growth factor-beta, activin, and bone morphogenetic protein-inducible enhancer. J. Biol. Chem. 273, 21145 – 21152 (1998). 39. J. Li, K. Tsuji, T. Komori, K. Miyazono, J. L. Wrana, Y. Ito, A. Nifuji, and M. Noda, Smad2 overexpression enhances Smad4 gene expression and suppresses CBFA1 gene expression in osteoblastic osteosarcoma ROS17/2.8 cells and primary rat calvaria cells. J. Biol. Chem. 273, 31009 – 31015 (1998). 40. M. M. Shull, I. Ormsby, A. B. Kier, S. Pawlowski, R. J. Diebold, M. Yin, R. Allen, C. Sidman, G. Proetzel, D. Calvin, et al., Targeted
424
41.
42.
43.
44.
45.
46.
47.
48.
49. 50.
51.
52.
53.
54.
55.
56.
57.
XUEZHONG QIN ET AL. disruption of the mouse transforming growth factor-beta 1 gene results in multifocal inflammatory disease. Nature 359, 693 – 699 (1992). A. B. Kulkarni, C. G. Huh, D. Becker, A. Geiser, M. Lyght, K. C. Flanders, A. B. Roberts, M. B. Sporn, J. M. Ward, and S. Karlsson, Transforming growth factor beta 1 null mutation in mice causes excessive inflammatory response and early death. Proc. Natl. Acad. Sci. USA 90, 770 – 774 (1993). G. Proetzel, S. A. Pawlowski, M. V. Wiles, M. Yin, G. P. Boivin, P. N. Howles, J. Ding, M. W. Ferguson, and T. Doetschman, Transforming growth factor-beta 3 is required for secondary palate fusion. Nature Genet. 11, 409 – 414 (1995). L. P. Sanford, I. Ormsby, A. C. Gittenberger-de Groot, H. Sariola, R. Friedman, G. P. Boivin, E. L. Cardell, and T. Doetschman, TGFbeta2 knockout mice have multiple developmental defects that are non-overlapping with other TGFbeta knockout phenotypes. Development 124, 2659 – 2670 (1997). A. G. Geiser, Q. Q. Zeng, M. Sato, L. M. Helvering, T. Hirano, and C. H. Turner, Decreased bone mass and bone elasticity in mice lacking the transforming growth factor-beta1 gene. Bone 23, 87 – 93 (1998). M. M. Mozes, E. P. Bottinger, T. A. Jacot, and J. B. Kopp, Renal expression of fibrotic matrix proteins and of transforming growth factor-beta (TGF-beta) isoforms in TGF-beta transgenic mice. J. Am. Soc. Nephrol. 10, 271 – 280 (1999). A. Erlebacher and R. Derynck, Increased expression of TGF-beta 2 in osteoblasts results in an osteoporosis-like phenotype. J. Cell Biol. 132, 195 – 210 (1996). Y. Chai, Y. Sasano, P. Bringas, Jr., M. Mayo, V. Kaartinen, N. Heisterkamp, J. Groffen, H. Slavkin, and C. Shuler, Characterization of the fate of midline epithelial cells during the fusion of mandibular prominences in vivo. Dev. Dyn. 208, 526 – 535 (1997). M. E. Joyce, A. B. Roberts, M. B. Sporn, and M. E. Bolander, Transforming growth factor-beta and the initiation of chondrogenesis and osteogenesis in the rat femur. J. Cell Biol. 110, 2195 – 2207 (1990). M. E. Bolander, Regulation of fracture repair by growth factors. Proc. Soc. Exp. Biol. Med. 200, 165 – 170 (1992). M. Noda and J. J. Camilliere, In vivo stimulation of bone formation by transforming growth factor-beta. Endocrinology 124, 2991 – 2994 (1989). C. Marcelli, A. J. Yates, and G. R. Mundy, In vivo effects of human recombinant transforming growth factor beta on bone turnover in normal mice. J. Bone Miner. Res. 5, 1087 – 1096 (1990). R. Fujimoto, T. Tanizawa, S. Nishida, N. Yamamoto, S. Soshi, N. Endo, and H. E. Takahashi, Local effects of transforming growth factor-beta1 on rat calvaria: changes depending on the dose and the injection site. J. Bone Miner. Metab. 17, 11 – 17 (1999). L. S. Beck, E. P. Amento, Y. Xu, L. Deguzman, W. P. Lee, T. Nguyen, and N. A. Gillett, TGF-beta 1 induces bone closure of skull defects: temporal dynamics of bone formation in defects exposed to rhTGF-beta 1. J. Bone Miner. Res. 8, 753 – 761 (1993). C. Bosch, B. Melsen, R. Gibbons, and K. Vargervik, Human recombinant transforming growth factor-beta 1 in healing of calvarial bone defects. J. Craniofac. Surg. 7, 300 – 310 (1996). L. McKinney and J. O. Hollinger, A bone regeneration study: Transforming growth factor-beta 1 and its delivery. J. Craniofac. Surg. 7, 36 – 45 (1996). G. Zellin, S. Beck, R. Hardwick, and A. Linde, Opposite effects of recombinant human transforming growth factor-beta 1 on bone regeneration in vivo: Effects of exclusion of periosteal cells by microporous membrane. Bone 22, 613 – 620 (1998). E. J. Mackie and U. Trechsel, Stimulation of bone formation in vivo by transforming growth factor-beta: Remodeling of woven bone and lack of inhibition by indomethacin. Bone 11, 295 – 300 (1990).
58. M. Lind, B. Schumacker, K. Soballe, J. Keller, F. Melsen, and C. Bunger, Transforming growth factor-beta enhances fracture healing in rabbit tibiae. Acta Orthop. Scand. 64, 553 – 556 (1993). 59. H. M. Nielsen, T. T. Andreassen, T. Ledet, and H. Oxlund, Local injection of TGF-beta increases the strength of tibial fractures in the rat. Acta Orthop. Scand. 65, 37 – 41 (1994). 60. J. Beaudreuil, G. Mbalaviele, M. Cohen-Solal, C. Morieux, M. C. De Vernejoul, and P. Orcel, Short-term local injections of transforming growth factor-beta 1 decrease ovariectomy-stimulated osteoclastic resorption in vivo in rats. J. Bone Miner. Res. 10, 971 – 977 (1995). 61. M. A. Critchlow, Y. S. Bland, and D. E. Ashhurst, The effect of exogenous transforming growth factor-beta 2 on healing fractures in the rabbit. Bone 16, 521 – 527 (1995). 62. L. S. Beck, R. L. Wong, L. DeGuzman, W. P. Lee, B. Ongpipattanakul, and T. H. Nguyen, Combination of bone marrow and TGFbeta1 augment the healing of critical-sized bone defects. J. Pharm. Sci. 87, 1379 – 1386 (1998). 63. U. M. Wikesjo, S. S. Razi, T. J. Sigurdsson, D. N. Tatakis, M. B. Lee, B. Ongpipattanakul, T. Nguyen, and R. Hardwick, Periodontal repair in dogs: Effect of recombinant human transforming growth factor-beta1 on guided tissue regeneration. J. Clin. Periodontol. 25, 475 – 481 (1998). 64. H. L. Glansbeek, H. M. van Beuningen, E. L. Vitters, P. M. van der Kraan, and W. B. van den Berg, Stimulation of articular cartilage repair in established arthritis by local administration of transforming growth factor-beta into murine knee joints. Lab. Invest. 78, 133 – 142 (1998). 65. D. N. Kalu, E. Salerno, Y. Higami, C. C. Liu, F. Ferraro, M. A. Salih, and B. H. Arjmandi, In vivo effects of transforming growth factorbeta 2 in ovariectomized rats. Bone Miner. 22, 209 – 220 (1993). 66. D. Rosen, S. C. Miller, E. DeLeon, A. Y. Thompson, H. Bentz, M. Mathews, and S. Adams, Systemic administration of recombinant transforming growth factor beta 2 (rTGF-beta 2] stimulates parameters of cancellous bone formation in juvenile and adult rats. Bone 15, 355 – 359 (1994). 67. E. Zerath, X. Holy, J. M. Mouillon, B. Farbos, M. Machwate, C. Andre, S. Renault, and P. J. Marie, TGF-beta2 prevents the impaired chondrocyte proliferation induced by unloading in growth plates of young rats. Life Sci. 61, 2397 – 2406 (1997). 68. D. Gazit, Y. Zilberman, G. Turgeman, S. Zhou, and A. Kahn, Recombinant TGF-beta1 stimulates bone marrow osteoprogenitor cell activity and bone matrix synthesis in osteopenic, old male mice. J. Cell. Biochem. 73, 379 – 389 (1999). 69. J. A. Carlino, H. R. Higley, J. R. Creson, P. D. Avis, Y. Ogawa, and L. R. Ellingsworth, Transforming growth factor beta 1 systemically modulates granuloid, erythroid, lymphoid, and thrombocytic cells in mice. Exp. Hematol. 20, 943 – 950 (1992). 70. J. B. Kopp, V. M. Factor, M. Mozes, P. Nagy, N. Sanderson, E. P. Bottinger, P. E. Klotman, and S. S. Thorgeirsson, Transgenic mice with increased plasma levels of TGF-beta 1 develop progressive renal disease. Lab. Invest. 74, 991 – 1003 (1996). 71. V. Nicolas, A. Prewett, P. Bettica, S. Mohan, R. D. Finkelman, D. J. Baylink, and J. R. Farley, Age-related decreases in insulin-like growth factor-I and transforming growth factor-beta in femoral cortical bone from both men and women: Implications for bone loss with aging. J. Clin. Endocrinol. Metab. 78, 1011 – 1016 (1994). 72. D. Gazit, Y. Zilberman, R. Ebner, and A. Kahn, Bone loss (osteopenia) in old male mice results from diminished activity and availability of TGF-beta. J. Cell. Biochem. 70, 478 – 488 (1998). 73. R. D. Finkelman, T. A. Linkhart, S. Mohan, K. H. Lau, D. J. Baylink, and N. H. Bell, Vitamin D deficiency causes a selective reduction in deposition of transforming growth factor beta in rat bone: Possible mechanism for impaired osteoinduction. Proc. Natl. Acad. Sci. USA 88, 3657 – 3660 (1991). 74. R. D. Finkelman, N. H. Bell, D. D. Strong, L. M. Demers, and D. J. Baylink, Ovariectomy selectively reduces the concentration of
CHAPTER 14 Bone Growth Factors
75.
76.
77.
78.
79.
80.
81.
82.
83.
84. 85.
86.
87.
88.
89.
90.
91.
transforming growth factor beta in rat bone: Implications for estrogen deficiency-associated bone loss. Proc. Natl. Acad. Sci. USA 89, 12190 – 12193 (1992). C. Kasperk, R. Fitzsimmons, D. Strong, S. Mohan, J. Jennings, J. Wergedal, and D. Baylink, Studies of the mechanism by which androgens enhance mitogenesis and differentiation in bone cells. J. Clin. Endocrinol. Metab. 71, 1322 – 1329 (1990). M. J. Oursler, C. Cortese, P. Keeting, M. A. Anderson, S. K. Bonde, B. L. Riggs, and T. C. Spelsberg, Modulation of transforming growth factor-beta production in normal human osteoblast-like cells by 17 beta-estradiol and parathyroid hormone. Endocrinology 129, 3313 – 3320 (1991). K. C. Westerlind and R. T. Turner, The skeletal effects of spaceflight in growing rats: Tissue-specific alterations in mRNA levels for TGFbeta. J. Bone Miner. Res. 10, 843 – 848 (1995). T. K. Sampath and A. H. Reddi, Dissociative extraction and reconstitution of extracellular matrix components involved in local bone differentiation. Proc. Natl. Acad. Sci. USA 78, 7599 – 7603 (1981). J. M. Wozney, V. Rosen, A. J. Celeste, L. M. Mitsock, M. J. Whitters, R. W. Kriz, R. M. Hewick, and E. A. Wang, Novel regulators of bone formation: Molecular clones and activities. Science 242, 1528–1534 (1988). V. Rosen, J. M. Wozney, E. A. Wang, P. Cordes, A. Celeste, D. McQuaid, and L. Kurtzberg, Purification and molecular cloning of a novel group of BMPs and localization of BMP mRNA in developing bone. Connect. Tissue Res. 20, 313 – 319 (1989). E. Ozkaynak, D. C. Rueger, E. A. Drier, C. Corbett, R. J. Ridge, T. K. Sampath, and H. Oppermann, OP-1 cDNA encodes an osteogenic protein in the TGF-beta family. EMBO J. 9, 2085 – 2093 (1990). E. Ozkaynak, P. N. Schnegelsberg, D. F. Jin, G. M. Clifford, F. D. Warren, E. A. Drier, and H. Oppermann, Osteogenic protein-2: A new member of the transforming growth factor-beta superfamily expressed early in embryogenesis. J. Biol. Chem. 267, 25220 – 25227 (1992). A. H. Reddi, Cartilage morphogenesis: Role of bone and cartilage morphogenetic proteins, homeobox genes and extracellular matrix. Matrix Biol. 14, 599 – 606 (1995). A. H. Reddi, Role of morphogenetic proteins in skeletal tissue engineering and regeneration. Nature Biotechnol. 16, 247 – 252 (1998). M. Kawabata, T. Imamura, and K. Miyazono, Signal transduction by bone morphogenetic proteins. Cytokine Growth Factor Rev. 9, 49 – 61 (1998). J. M. Schmitt, K. Hwang, S. R. Winn, and J. O. Hollinger, Bone morphogenetic proteins: An update on basic biology and clinical relevance. J. Orthop. Res. 17, 269 – 278 (1999). V. M. Paralkar, A. L. Vail, W. A. Grasser, T. A. Brown, H. Xu, S. Vukicevic, H. Z. Ke, H. Qi, T. A. Owen, and D. D. Thompson, Cloning and characterization of a novel member of the transforming growth factor-beta/bone morphogenetic protein family. J. Biol. Chem. 273, 13760 – 13767 (1998). T. K. Sampath, J. E. Coughlin, R. M. Whetstone, D. Banach, C. Corbett, R. J. Ridge, E. Ozkaynak, H. Oppermann, and D. C. Rueger, Bovine osteogenic protein is composed of dimers of OP-1 and BMP-2A, two members of the transforming growth factor-beta superfamily. J. Biol. Chem. 265, 13198 – 13205 (1990). F. Lecanda, L. V. Avioli, and S. L. Cheng, Regulation of bone matrix protein expression and induction of differentiation of human osteoblasts and human bone marrow stromal cells by bone morphogenetic protein-2. J. Cell. Biochem. 67, 386 – 396 (1997). O. Fromigue, P. J. Marie, and A. Lomri, Bone morphogenetic protein – 2 and transforming growth factor-beta2 interact to modulate human bone marrow stromal cell proliferation and differentiation. J. Cell. Biochem. 68, 411 – 426 (1998). A. Chaudhari, E. Ron, and M. P. Rethman, Recombinant human bone morphogenetic protein-2 stimulates differentiation in primary cultures of fetal rat calvarial osteoblasts. Mol. Cell. Biochem. 167, 31 – 39 (1997).
425 92. Y. Torii, K. Hitomi, and N. Tsukagoshi, Synergistic effect of BMP-2 and ascorbate on the phenotypic expression of osteoblastic MC3T3E1 cells. Mol. Cell. Biochem. 165, 25 – 29 (1996). 93. T. Nakase, K. Takaoka, K. Masuhara, K. Shimizu, H. Yoshikawa, and T. Ochi, Interleukin-1 beta enhances and tumor necrosis factor-alpha inhibits bone morphogenetic protein-2-induced alkaline phosphatase activity in MC3T3-E1 osteoblastic cells. Bone 21, 17 – 21 (1997). 94. S. D. Boden, G. Hair, L. Titus, M. Racine, K. McCuaig, J. M. Wozney, and M. S. Nanes, Glucocorticoid-induced differentiation of fetal rat calvarial osteoblasts is mediated by bone morphogenetic protein-6. Endocrinology 138, 2820 – 2828 (1997). 95. I. W. Li, S. Cheifetz, C. A. McCulloch, K. T. Sampath, and J. Sodek, Effects of osteogenic protein-1 (OP-1, BMP-7) on bone matrix protein expression by fetal rat calvarial cells are differentiation stage specific. J. Cell. Physiol. 169, 115 – 125 (1996). 96. L. Erlacher, J. McCartney, E. Piek, P. ten Dijke, M. Yanagishita, H. Oppermann, and F. P. Luyten, Cartilage-derived morphogenetic proteins and osteogenic protein-1 differentially regulate osteogenesis. J. Bone Miner. Res. 13, 383 – 392 (1998). 97. M. Hazama, A. Aono, N. Ueno, and Y. Fujisawa, Efficient expression of a heterodimer of bone morphogenetic protein subunits using a baculovirus expression system. Biochem. Biophys. Res. Commun. 209, 859 – 866 (1995). 98. A. Aono, M. Hazama, K. Notoya, S. Taketomi, H. Yamasaki, R. Tsukuda, S. Sasaki, and Y. Fujisawa, Potent ectopic boneinducing activity of bone morphogenetic protein-4/7 heterodimer. Biochem. Biophys. Res. Commun. 210, 670 – 677 (1995). 99. D. I. Israel, J. Nove, K. M. Kerns, R. J. Kaufman, V. Rosen, K. A. Cox, and J. M. Wozney, Heterodimeric bone morphogenetic proteins show enhanced activity in vitro and in vivo. Growth Factors 13, 291 – 300 (1996). 100. H. Yamashita, P. Ten Dijke, C. H. Heldin, and K. Miyazono, Bone morphogenetic protein receptors. Bone 19, 569 – 574 (1996). 101. R. Nishimura, Y. Kato, D. Chen, S. E. Harris, G. R. Mundy, and T. Yoneda, Smad5 and DPC4 are key molecules in mediating BMP-2induced osteoblastic differentiation of the pluripotent mesenchymal precursor cell line C2C12. J. Biol. Chem. 273, 1872 – 1879 (1998). 102. H. Chang, D. Huylebroeck, K. Verschueren, Q. Guo, M. M. Matzuk, and A. Zwijsen, Smad5 knockout mice die at mid-gestation due to multiple embryonic and extraembryonic defects. Development 126, 1631 – 1642 (1999). 103. S. J. Tang, P. A. Hoodless, Z. Lu, M. L. Breitman, R. R. McInnes, J. L. Wrana, and M. Buchwald, The Tlx-2 homeobox gene is a downstream target of BMP signalling and is required for mouse mesoderm development. Development 125, 1877 – 1887 (1998). 104. M. Bei and R. Maas, FGFs and BMP4 induce both Msx1-independent and Msx1-dependent signaling pathways in early tooth development. Development 125, 4325 – 4333 (1998). 105. H. J. Kim, D. P. Rice, P. J. Kettunen, and I. Thesleff, FGF-, BMPand Shh-mediated signalling pathways in the regulation of cranial suture morphogenesis and calvarial bone development. Development 125, 1241 – 1251 (1998). 106. P. Ducy, R. Zhang, V. Geoffroy, A. L. Ridall, and G. Karsenty, Osf2/Cbfa1: A transcriptional activator of osteoblast differentiation. Cell 89, 747 – 754 (1997). 107. L. B. Zimmerman, J. M. De Jesus-Escobar, and R. M. Harland, The Spemann organizer signal noggin binds and inactivates bone morphogenetic protein 4. Cell 86, 599 – 606 (1996). 108. S. A. Holley, J. L. Neul, L. Attisano, J. L. Wrana, Y. Sasai, M. B. O’Connor, E. M. De Robertis, and E. L. Ferguson, The Xenopus dorsalizing factor noggin ventralizes Drosophila embryos by preventing DPP from activating its receptor. Cell 86, 607 – 617 (1996). 109. S. Piccolo, Y. Sasai, B. Lu, and E. M. De Robertis, Dorsoventral patterning in Xenopus: Inhibition of ventral signals by direct binding of chordin to BMP-4. Cell 86, 589 – 598 (1996).
426 110. W. C. Smith and R. M. Harland, Expression cloning of noggin, a new dorsalizing factor localized to the Spemann organizer in Xenopus embryos. Cell 70, 829 – 840 (1992). 111. E. Gazzerro, V. Gangji, and E. Canalis, Bone morphogenetic proteins induce the expression of noggin, which limits their activity in cultured rat osteoblasts. J. Clin. Invest. 102, 2106 – 2114 (1998). 112. L. J. Brunet, J. A. McMahon, A. P. McMahon, and R. M. Harland, Noggin, cartilage morphogenesis, and joint formation in the mammalian skeleton. Science 280, 1455 – 1457 (1998). 113. Y. Sasai, B. Lu, H. Steinbeisser, D. Geissert, L. K. Gont, and E. M. De Robertis, Xenopus chordin: A novel dorsalizing factor activated by organizer-specific homeobox genes. Cell 79, 779 – 790 (1994). 114. W. N. Pappano, I. C. Scott, T. G. Clark, R. L. Eddy, T. B. Shows, and D. S. Greenspan, Coding sequence and expression patterns of mouse chordin and mapping of the cognate mouse chrd and human CHRD genes. Genomics 52, 236 – 239 (1998). 115. H. Zhang and A. Bradley, Mice deficient for BMP2 are nonviable and have defects in amnion/chorion and cardiac development. Development 122, 2977 – 2986 (1996). 116. G. Winnier, M. Blessing, P. A. Labosky, and B. L. Hogan, Bone morphogenetic protein-4 is required for mesoderm formation and patterning in the mouse. Genes Dev. 9, 2105 – 2116 (1995). 117. N. R. Dunn, G. E. Winnier, L. K. Hargett, J. J. Schrick, A. B. Fogo, and B. L. Hogan, Haploinsufficient phenotypes in Bmp4 heterozygous null mice and modification by mutations in Gli3 and Alx4. Dev. Biol. 188, 235 – 247 (1997). 118. D. M. Kingsley, A. E. Bland, J. M. Grubber, P. C. Marker, L. B. Russell, N. G. Copeland, and N. A. Jenkins, The mouse short ear skeletal morphogenesis locus is associated with defects in a bone morphogenetic member of the TGF beta superfamily. Cell 71, 399 – 410 (1992). 119. M. J. Solloway, A. T. Dudley, E. K. Bikoff, K. M. Lyons, B. L. Hogan, and E. J. Robertson, Mice lacking Bmp6 function. Dev. Genet. 22, 321 – 339 (1998). 120. N. Gena, C. Martin-Seisdedos, P. McCue, and C. M. Groce, BMP-7 null mutation in mice: Developmental defects in skeleton, kidney and eye. Exp. Cell Res. 230, 28 – 37 (1997). 121. T. Katagiri, S. Boorla, J. L. Frendo, B. L. Hogan, and G. Karsenty, Skeletal abnormalities in doubly heterozygous Bmp4 and Bmp7 mice. Dev. Genet. 22, 340 – 348 (1998). 122. G. Q. Zhao, K. Deng, P. A. Labosky, L. Liaw, and B. L. Hogan, The gene encoding bone morphogenetic protein 8B is required for the initiation and maintenance of spermatogenesis in the mouse. Genes Dev. 10, 1657 – 1669 (1996). 123. E. E. Storm, T. V. Huynh, N. G. Copeland, N. A. Jenkins, D. M. Kingsley, and S. J. Lee, Limb alterations in brachypodism mice due to mutations in a new member of the TGF beta-superfamily. Nature 368, 639 – 643 (1994). 124. J. T. Thomas, K. Lin, M. Nandedkar, M. Camargo, J. Cervenka, and F. P. Luyten, A human chondrodysplasia due to a mutation in a TGFbeta superfamily member. Nature Genet. 12, 315 – 317 (1996). 125. A. C. McPherron, A. M. Lawler, and S. J. Lee, Regulation of skeletal muscle mass in mice by a new TGF-beta superfamily member. Nature 387, 83 – 90 (1997). 126. J. Dong, D. F. Albertini, K. Nishimori, T. R. Kumar, N. Lu, and M. M. Matzuk, Growth differentiation factor – 9 is required during early ovarian folliculogenesis. Nature 383, 531 – 535 (1996). 127. R. Zhao, A. M. Lawler, and S. J. Lee, Characterization of GDF-10 expression patterns and null mice. Dev. Biol. 212, 68 – 79 (1999). 128. U. Ripamonti, S. S. Ma, B. van den Heever, and A. H. Reddi, Osteogenin, a bone morphogenetic protein, adsorbed on porous hydroxyapatite substrata, induces rapid bone differentiation in calvarial defects of adult primates. Plast. Reconstr. Surg. 90, 382 – 393 (1992). 129. R. Langer and J. P. Vacanti, Tissue engineering. Science 260, 920 – 926 (1993).
XUEZHONG QIN ET AL. 130. E. Hedner and A. Linde, Efficacy of bone morphogenetic protein (BMP) with osteopromotive membranes: An experimental study in rat mandibular defects. Eur. J. Oral Sci. 103, 236 – 241 (1995). 131. M. Niederwanger and M. R. Urist, Demineralized bone matrix supplied by bone banks for a carrier of recombinant human bone morphogenetic protein (rhBMP-2): A substitute for autogeneic bone grafts. J. Oral Implantol. 22, 210 – 215 (1996). 132. Y. Yamazaki, S. Oida, K. Ishihara, and N. Nakabayashi, Ectopic induction of cartilage and bone by bovine bone morphogenetic protein using a biodegradable polymeric reservoir. J. Biomed. Mater. Res. 30, 1 – 4 (1996). 133. H. Takita, E. Tsuruga, I. Ono, and Y. Kuboki, Enhancement by bFGF of osteogenesis induced by rhBMP-2 in rats. Eur. J. Oral Sci. 105, 588 – 592 (1997). 134. T. N. Gerhart, C. A. Kirker-Head, M. J. Kriz, M. E. Holtrop, G. E. Hennig, J. Hipp, S. H. Schelling, and E. Wang, Healing segmental femoral defects in sheep using recombinant human bone morphogenetic protein. Clin. Orthop. 293, 317 – 326 (1993). 135. S. D. Cook, M. W. Wolfe, S. L. Salkeld, and D. C. Rueger, Effect of recombinant human osteogenic protein-1 on healing of segmental defects in non-human primates. J. Bone Joint. Surg. Am. 77, 734 – 750 (1995). 136. H. D. Zegzula, D. C. Buck, J. Brekke, J. M. Wozney, and J. O. Hollinger, Bone formation with use of rhBMP-2 (recombinant human bone morphogenetic protein-2). J. Bone Joint Surg. Am. 79, 1778 – 1790 (1997). 137. J. P. Sheehan, D. F. Kallmes, J. M. Sheehan, J. A. Jane, Jr., A. H. Fergus, C. G. diPierro, N. E. Simmons, D. D. Makel, and G. A. Helm, Molecular methods of enhancing lumbar spine fusion. Neurosurgery 39, 548 – 554 (1996). 138. S. D. Boden, G. J. Martin, Jr., M. A. Morone, J. L. Ugbo, and P. A. Moskovitz, Posterolateral lumbar intertransverse process spine arthrodesis with recombinant human bone morphogenetic protein 2/hydroxyapatite-tricalcium phosphate after laminectomy in the nonhuman primate. Spine 24, 1179 – 1185 (1999). 139. M. Nakashima, Induction of dentin formation on canine amputated pulp by recombinant human bone morphogenetic proteins (BMP)-2 and -4. J. Dent. Res. 73, 1515 – 1522 (1994). 140. X. Si, Y. Jin and L. Yang, Induction of new bone by ceramic bovine bone with recombinant human bone morphogenetic protein 2 and transforming growth factor beta. Int. J. Oral Maxillofac. Surg. 27, 310 – 314 (1998). 141. I. Ono, M. Inoue, and Y. Kuboki, Promotion of the osteogenetic activity of recombinant human bone morphogenetic protein by prostaglandin E1. Bone 19, 581 – 588 (1996). 142. G. Mundy, R. Garrett, S. Harris, J. Chan, D. Chen, G. Rossini, B. Boyce, M. Zhao, and G. Gutierrez, Stimulation of bone formation in vitro and in rodents by statins. Science 286, 1946 – 1949 (1999). 143. J. Fang, Y. Y. Zhu, E. Smiley, J. Bonadio, J. P. Rouleau, S. A. Goldstein, L. K. McCauley, B. L. Davidson, and B. J. Roessler, Stimulation of new bone formation by direct transfer of osteogenic plasmid genes. Proc. Natl. Acad. Sci. USA 93, 5753 – 5758 (1996). 144. J. R. Lieberman, L. Q. Le, L. Wu, G. A. Finerman, A. Berk, O. N. Witte, and S. Stevenson, Regional gene therapy with a BMP-2-producing murine stromal cell line induces heterotopic and orthotopic bone formation in rodents. J. Orthop. Res. 16, 330 – 339 (1998). 145. K. D. Riew, N. M. Wright, S. Cheng, L. V. Avioli, and J. Lou, Induction of bone formation using a recombinant adenoviral vector carrying the human BMP-2 gene in a rabbit spinal fusion model. Calcif. Tissue Int. 63, 357 – 360 (1998). 146. J. Lou, F. Xu, K. Merkel, and P. Manske, Gene therapy: Adenovirus-mediated human bone morphogenetic protein-2 gene transfer induces mesenchymal progenitor cell proliferation and differentiation in vitro and bone formation in vivo. J. Orthop. Res. 17, 43 – 50 (1999).
CHAPTER 14 Bone Growth Factors 147. D. S. Musgrave, P. Bosch, S. Ghivizzani, P. D. Robbins, C. H. Evans, and J. Huard, Adenovirus-mediated direct gene therapy with bone morphogenetic protein-2 produces bone. Bone 24, 541 – 547 (1999). 148. T. D. Alden, D. D. Pittman, G. R. Hankins, E. J. Beres, J. A. Engh, S. Das, S. B. Hudson, K. M. Kerns, D. F. Kallmes, and G. A. Helm, In vivo endochondral bone formation using a bone morphogenetic protein 2 adenoviral vector. Hum. Gene Ther. 10, 2245 – 2253 (1999). 149. E. Canalis, Insulin like growth factors and the local regulation of bone formation. Bone 14, 273 – 276 (1993). 150. C. J. Rosen, L. R. Donahue, and S. J. Hunter, Insulin-like growth factors and bone: the osteoporosis connection. Proc. Soc. Exp. Biol. Med. 206, 83 – 102 (1994). 151. C. J. Rosen and L. R. Donahue, Insulin-like growth factors and bone: the osteoporosis connection revisited. Proc. Soc. Exp. Biol. Med. 219, 1 – 7 (1998). 152. S. Mohan and D. J. Baylink, IGF system components and their role in bone metabolism In “Contemporary Endocrinology: The IGF System” (R. Rosenfeld and C. Roberts, eds.), pp. 457 – 496, Humana Press, Inc., Totowa, NJ, (1999). 153. S. Mohan and D. J. Baylink, The role of insulin-like growth factor-II in the coupling of bone formation and resorption In “Modern Concepts of Insulin-like Growth Factors” (E. M. Spencer, ed.), pp. 169 – 184, Elsevier Science, New York, 1991. 154. C. Schmid, T. Steiner and E. R. Froesch, (1984). Insulin-like growth factor I supports differentiation of cultured osteoblast-like cells. FEBS Lett 173, 48 – 52 (1984). 155. D. D. Strong, A. L. Beachler, J. E. Wergedal, and T. A. Linkhart, Insulin-like growth factor II and transforming growth factor beta regulate collagen expression in human osteoblast-like cells in vitro. J. Bone Miner. Res. 6, 15 – 23 (1991). 156. T. Onishi, S. Kinoshita, S. Shintani, S. Sobue, and T. Ooshima, Stimulation of proliferation and differentiation of dog dental pulp cells in serum-free culture medium by insulin-like growth factor. Arch. Oral Biol. 44, 361 – 371 (1999). 157. P. A. Hill, A. Tumber, and M. C. Meikle, Multiple extracellular signals promote osteoblast survival and apoptosis. Endocrinology 138, 3849 – 3858 (1997). 158. A. Kawakami, T. Nakashima, M. Tsuboi, S. Urayama, N. Matsuoka, H. Ida, Y. Kawabe, H. Sakai, K. Migita, T. Aoyagi, M. Nakashima, K. Maeda, and K. Eguchi, Insulin-like growth factor I stimulates proliferation and Fas-mediated apoptosis of human osteoblasts. Biochem. Biophys. Res. Commun. 247, 46 – 51 (1998). 159. J. Farley and B. Coffing, Apoptosis may determine the rate of appearance of skeletal alkaline phosphatase in serum. J. Bone Miner. Res. 14 (Suppl 1), (1999), [Abstract SA148]. 160. J. M. Hock, M. Centrella, and E. Canalis, Insulin-like growth factor I has independent effects on bone matrix formation and cell replication. Endocrinology 122, 254 – 260 (1988). 161. E. Canalis, S. Rydziel, A. M. Delany, S. Varghese, and J. J. Jeffrey, Insulin-like growth factors inhibit interstitial collagenase synthesis in bone cell cultures. Endocrinology 136, 1348 – 1354 (1995). 162. P. A. Hill, J. J. Reynolds, and M. C. Meikle, Osteoblasts mediate insulin-like growth factor-I and -II stimulation of osteoclast formation and function. Endocrinology 136, 124 – 131 (1995). 163. H. Kaji, T. Sugimoto, M. Kanatani, K. Nishiyama, M. Nasu, and K. Chihara, Insulin-like growth factor-I mediates osteoclast-like cell formation stimulated by parathyroid hormone. J. Cell. Physiol. 172, 55 – 62 (1997). 164. J. Guicheux, D. Heymann, A. V. Rousselle, F. Gouin, P. Pilet, S. Yamada, and G. Daculsi, Growth hormone stimulatory effects on osteoclastic resorption are partly mediated by insulin-like growth factor I: An in vitro study. Bone 22, 25 – 31 (1998). 165. P. Hou, T. Sato, W. Hofstetter, and N. T. Foged, Identification and characterization of the insulin-like growth factor I receptor in mature rabbit osteoclasts. J. Bone Miner. Res. 12, 534 – 540 (1997).
427 166. J. I. Jones and D. R. Clemmons, Insulin-like growth factors and their binding proteins: Biological actions. Endocr. Rev. 16, 3 – 34 (1995). 167. T. Kadowaki, K. Tobe, R. Honda-Yamamoto, H. Tamemoto, Y. Kaburagi, K. Momomura, K. Ueki, Y. Takahashi, T. Yamauchi, Y. Akanuma, and Y. Yazaki, Signal transduction mechanism of insulin and insulin-like growth factor-1. Endocr. J. 43 (Suppl), S33 – 41 (1996). 168. S. Li, M. Resnicoff, and R. Baserga, Effect of mutations at serines 1280 – 1283 on the mitogenic and transforming activities of the insulin-like growth factor I receptor. J. Biol. Chem. 271, 12254 – 12260 (1996). 169. R. Baserga, M. Resnicoff, C. D’Ambrosio, and B. Valentinis, The role of the IGF-I receptor in apoptosis. Vitam. Horm. 53, 65 – 98 (1997). 170. E. Surmacz, C. Sell, J. Swantek, H. Kato, C. T. Roberts, Jr., D. LeRoith, and R. Baserga, Dissociation of mitogenesis and transforming activity by C-terminal truncation of the insulin-like growth factor-I receptor. Exp. Cell Res. 218, 370 – 380 (1995). 171. R. O’Connor, A. Kauffmann-Zeh, Y. Liu, S. Lehar, G. I. Evan, R. Baserga, and W. A. Blattler, Identification of domains of the insulin-like growth factor I receptor that are required for protection from apoptosis. Mol. Cell. Biol. 17, 427 – 435 (1997). 172. T. M. DeChiara, E. J. Robertson, and A. Efstratiadis, Parental imprinting of the mouse insulin-like growth factor II gene. Cell 64, 849 – 859 (1991). 173. J. Wang, J. Zhou, L. Powell-Braxton, C. Bondy, Effects of IGF – 1 gene deletion on postnatal growth patterns. Endocrinology 140, 3391 – 3394 (1999). 174. J. Baker, J. P. Liu, E. J. Robertson, and A. Efstratiadis, Role of insulin-like growth factors in embryonic and postnatal growth. Cell 75, 73 – 82 (1993). 175. L. S. Mathews, R. E. Hammer, R. R. Behringer, A. J. D’Ercole, G. I. Bell, R. L. Brinster, and R. D. Palmiter, Growth enhancement of transgenic mice expressing human insulin-like growth factor I. Endocrinology 123, 2827 – 2833 (1988). 176. G. Zhao, M. C. Faugere, H. Malluche, J. Wang, T. Nakayama, J. Pike, J. Fagin, and T. Clemens, Osteoblast-specific overexpression of IGF-I in transgenic mice increases cancellous bone volume. J. Bone Miner. Res. 23 (Suppl 1), (1988) [Abstract 1099]. 177. E. M. Spencer, C. C. Liu, E. C. Si, and G. A. Howard, In vivo actions of insulin-like growth factor-I (IGF-I) on bone formation and resorption in rats. Bone 12, 21 – 26 (1991). 178. H. N. Rosen, V. Chen, A. Cittadini, S. L. Greenspan, P. S. Douglas, A. C. Moses, and W. G. Beamer, Treatment with growth hormone and IGF-I in growing rats increases bone mineral content but not bone mineral density. J. Bone Miner. Res. 10, 1352 – 1358 (1995). 179. N. Miyakoshi, C. Richman, X. Qin, D. J. Baylink, and S. Mohan, Effects of recombinant insulin-like growth factor-binding protein-4 on bone formation parameters in mice. Endocrinology 140, 5719 – 5728 (1999). 180. H. P. Guler, J. Zapf, E. Scheiwiller, and E. R. Froesch, Recombinant human insulin-like growth factor I stimulates growth and has distinct effects on organ size in hypophysectomized rats. Proc. Natl. Acad. Sci. USA 85, 4889 – 4893 (1988). 181. P. Ammann, R. Rizzoli, K. Muller, D. Slosman, and J. P. Bonjour, IGF-I and pamidronate increase bone mineral density in ovariectomized adult rats. Am. J. Physiol. 265, E770 – E776 (1993). 182. J. Verhaeghe, R. van Bree, E. Van Herck, H. Thomas, A. Skottner, J. Dequeker, L. Mosekilde, T. A. Einhorn, and R. Bouillon, Effects of recombinant human growth hormone and insulin-like growth factor-I, with or without 17 beta-estradiol, on bone and mineral homeostasis of aged ovariectomized rats. J. Bone Miner. Res. 11, 1723 – 1735 (1996). 183. A. Cemborain, I. Castilla-Cortazar, M. Garcia, J. Quiroga, B. Muguerza, A. Picardi, S. Santidrian, and J. Prieto, Osteopenia in
428
184.
185.
186.
187.
188.
189.
190.
191.
192.
193.
194.
195.
196. 197.
198.
199.
200.
XUEZHONG QIN ET AL. rats with liver cirrhosis: beneficial effects of IGF-I treatment. J. Hepatol. 28, 122 – 131 (1998). J. Isgaard, A. Nilsson, A. Lindahl, J. O. Jansson, and O. G. Isaksson, Effects of local administration of GH and IGF-1 on longitudinal bone growth in rats. Am. J. Physiol. 250, E367 – E372 (1986). A. Wakisaka, H. Tanaka, J. Barnes, and C. T. Liang, Effect of locally infused IGF-I on femoral gene expression and bone turnover activity in old rats. J. Bone Miner. Res. 13, 13 – 19 (1998). M. Ishibe, T. Ishibashi, K. Kaneda, T. Koda, R. N. Rosier, and J. E. Puzas, Stimulation of bone formation in vivo by insulin-like growth factor-II in rats. Calcif. Tissue Int. 63, 36 – 38 (1998). A. G. Johansson, E. Lindh, and S. Ljunghall, Insulin-like growth factor I stimulates bone turnover in osteoporosis. Lancet 339, 1619 (1992). P. R. Ebeling, J. D. Jones, W. M. O’Fallon, C. H. Janes, and B. L. Riggs, Short-term effects of recombinant human insulin-like growth factor I on bone turnover in normal women. J. Clin. Endocrinol. Metab. 77, 1384 – 1387 (1993). A. G. Johansson, E. Lindh, G. Kollerup, and S. Ljunghall, Dosedependent effects of IGF-I on the metabolism of collagen type I in male osteoporosis. J. Bone Miner. Res. 10 (Suppl 1), S249 (1995) [Abstract S393]. A. G. Johansson, E. Lindh, W. F. Blum, G. Kollerup, O. H. Sorensen, and S. Ljunghall, Effects of growth hormone and insulin-like growth factor I in men with idiopathic osteoporosis. J. Clin. Endocrinol. Metab. 81, 44 – 48 (1996). L. J. Ghiron, J. L. Thompson, L. Holloway, R. L. Hintz, G. E. Butterfield, A. R. Hoffman, and R. Marcus, Effects of recombinant insulinlike growth factor-I and growth hormone on bone turnover in elderly women. J. Bone Miner. Res. 10, 1844 – 1852 (1995). P. G. Campbell and J. F. Novak, Insulin-like growth factor binding protein (IGFBP) inhibits IGF action on human osteosarcoma cells. J. Cell. Physiol. 149, 293 – 300 (1991). S. Mohan, C. M. Bautista, J. Wergedal, and D. J. Baylink, Isolation of an inhibitory insulin-like growth factor (IGF) binding protein from bone cell-conditioned medium: A potential local regulator of IGF action. Proc. Natl. Acad. Sci. USA 86, 8338 – 8342 (1989). J. H. Feyen, D. B. Evans, C. Binkert, G. F. Heinrich, S. Geisse, and H. P. Kocher, Recombinant human [Cys281]insulin-like growth factor-binding protein 2 inhibits both basal and insulin-like growth factor I-stimulated proliferation and collagen synthesis in fetal rat calvariae. J. Biol. Chem. 266, 19469 – 19474 (1991). M. Ernst and G. A. Rodan, Increased activity of insulin-like growth factor (IGF) in osteoblastic cells in the presence of growth hormone (GH): Positive correlation with the presence of the GH-induced IGFbinding protein BP – 3. Endocrinology 127, 807 – 814 (1990). S. Mohan, Insulin-like growth factor binding proteins in bone cell regulation. Growth Regul. 3 67 – 70 (1993). S. Amarnani, H. L. Merriman, T. A. Linkhart, D. J. Baylink, and S. Mohan, Autocrine regulators of MC3T3-E1 cell proliferation. J. Bone Miner. Res. 8, 157 – 165 (1993). S. K. Durham, M. C. Kiefer, B. L. Riggs, and C. A. Conover, Regulation of insulin-like growth factor binding protein 4 by a specific insulin-like growth factor binding protein 4 proteinase in normal human osteoblast-like cells: Implications in bone cell physiology. J. Bone Miner. Res. 9, 111 – 117 (1994). S. Mohan, Y. Nakao, Y. Honda, E. Landale, U. Leser, C. Dony, K. Lang, and D. J. Baylink, Studies on the mechanisms by which insulin-like growth factor (IGF) binding protein-4 (IGFBP-4) and IGFBP-5 modulate IGF actions in bone cells. J. Biol. Chem. 270, 20424 – 20431 (1995). C. M. Bautista, D. J. Baylink, and S. Mohan, Isolation of a novel insulin-like growth factor (IGF) binding protein from human bone: A potential candidate for fixing IGF-II in human bone. Biochem. Biophys. Res. Commun. 176, 756 – 763 (1991).
201. D. L. Andress and R. S. Birnbaum, Human osteoblast-derived insulin-like growth factor (IGF) binding protein-5 stimulates osteoblast mitogenesis and potentiates IGF action. J. Biol. Chem. 267, 22467 – 22472 (1992). 202. N. Srinivasan, D. Edwall, T. A. Linkhart, D. J. Baylink, and S. Mohan, Insulin-like growth factor-binding protein-6 produced by human PC-3 prostate cancer cells: isolation, characterization and its biological action. J. Endocrinol. 149, 297 – 303 (1996). 203. B. Valentinis, A. Bhala, T. DeAngelis, R. Baserga, and P. Cohen, The human insulin-like growth factor (IGF) binding protein-3 inhibits the growth of fibroblasts with a targeted disruption of the IGF-I receptor gene. Mol. Endocrinol. 9, 361 – 367 (1995). 204. C. Lalou, C. Lassarre, and M. Binoux, A proteolytic fragment of insulin-like growth factor (IGF) binding protein-3 that fails to bind IGFs inhibits the mitogenic effects of IGF-I and insulin. Endocrinology 137, 3206 – 3212 (1996). 205. S. Mohseni-Zadeh and M. Binoux, Insulin-like growth factor (IGF) binding protein – 3 interacts with the type 1 IGF receptor, reducing the affinity of the receptor for its ligand: An alternative mechanism in the regulation of IGF action. Endocrinology 138, 5645 – 5648 (1997). 206. S. D. Chernausek, C. E. Smith, K. L. Duffin, W. H. Busby, G. Wright, and D. R. Clemmons, Proteolytic cleavage of insulin-like growth factor binding protein 4 (IGFBP-4). Localization of cleavage site to non-homologous region of native IGFBP-4. J. Biol. Chem. 270, 11377 – 11382 (1995). 207. C. A. Conover, S. K. Durham, J. Zapf, F. R. Masiarz, and M. C. Kiefer, Cleavage analysis of insulin-like growth factor (IGF)-dependent IGF-binding protein-4 proteolysis and expression of proteaseresistant IGF-binding protein-4 mutants. J. Biol. Chem. 270, 4395 – 4400 (1995). 208. X. Qin, D. D. Strong, D. J. Baylink, and S. Mohan, Structurefunction analysis of the human insulin-like growth factor binding protein-4. J. Biol. Chem. 273, 23509 – 23516 (1998). 209. X. Qin, D. Byun, D. D. Strong, D. J. Baylink, and S. Mohan, Studies on the role of human insulin-like growth factor-II (IGF-II)-dependent IGF binding protein (hIGFBP)-4 protease in human osteoblasts using protease-resistant IGFBP-4 analogs. J. Bone Miner. Res. 14, 2079 – 2088 (1999). 210. W. Li, J. Fawcett, H. R. Widmer, P. J. Fielder, R. Rabkin, and G. A. Keller, Nuclear transport of insulin-like growth factor-I and insulinlike growth factor binding protein-3 in opossum kidney cells. Endocrinology 138, 1763 – 1766 (1997). 211. K. Rajkumar, D. Barron, M. S. Lewitt, and L. J. Murphy, Growth retardation and hyperglycemia in insulin-like growth factor binding protein-1 transgenic mice. Endocrinology 136, 4029 – 4034 (1995). 212. A. Hoeflich, M. Wu, S. Mohan, J. Foll, R. Wanke, T. Froehlich, G. J. Arnold, H. Lahm, H. J. Kolb, and E. Wolf, Overexpression of insulin-like growth factor-binding protein-2 in transgenic mice reduces postnatal body weight gain. Endocrinology 140, 5488 – 5496 (1999). 213. L. J. Murphy, P. Molnar, X. Lu, and H. Huang, Expression of human insulin-like growth factor-binding protein-3 in transgenic mice. J. Mol. Endocrinol. 15, 293 – 303 (1995). 214. J. Wang, W. Niu, D. P. Witte, S. D. Chernausek, Y. E. Nikiforov, T. L. Clemens, B. Sharifi, A. R. Strauch, and J. A. Fagin, Overexpression of insulin-like growth factor-binding protein-4 (IGFBP-4) in smooth muscle cells of transgenic mice through a smooth muscle alpha-actin-IGFBP-4 fusion gene induces smooth muscle hypoplasia. Endocrinology 139, 2605 – 2614 (1998). 215. W. Ni, K. Rajkumar, J. I. Nagy, and L. J. Murphy, Impaired brain development and reduced astrocyte response to injury in transgenic mice expressing IGF binding protein-1. Brain Res. 769, 97 – 107 (1997). 216. C. M. Bagi, R. Brommage, L. Deleon, S. Adams, D. Rosen, and A. Sommer, Benefit of systemically administered rhIGF-I and
CHAPTER 14 Bone Growth Factors
217.
218.
219.
220.
221.
222.
223.
224.
225.
226.
227.
228.
229.
230. 231.
rhIGF-I/IGFBP-3 on cancellous bone in ovariectomized rats. J. Bone Miner. Res. 9, 1301 – 1312 (1994). K. Narusawa, T. Nakamura, K. Suzuki, Y. Matsuoka, L. J. Lee, H. Tanaka, Y. Seino, The effects of recombinant human insulin-like growth factor (rhIGF-1 and rhIGF-1/IGF binding protein-3 administration on rat osteopenia induced by ovariectomy with concomitant bilateral sciatic neurectomy. J. Bone Miner. Res. 10, 1853 – 1864 (1995). C. Richman, D. J. Baylink, K. Lang, C. Dony, and S. Mohan, Recombinant human insulin-like growth factor-binding protein-5 stimulates bone formation parameters in vitro and in vivo. Endocrinology 140, 4699 – 4705 (1999). S. K. Durham, B. L. Riggs, and C. A. Conover, The insulin-like growth factor-binding protein-4 (IGFBP-4)-IGFBP-4 protease system in normal human osteoblast-like cells: regulation by transforming growth factor-beta. J. Clin. Endocrinol. Metab. 79, 1752 – 1758 (1994). S. Kanzaki, S. Hilliker, D. J. Baylink, and S. Mohan, Evidence that human bone cells in culture produce insulin-like growth factor-binding protein-4 and -5 proteases. Endocrinology 134, 383 – 392 (1994). X. Qin, D. Byun, D. J. Baylink, and S. Mohan, Evidence that IGF binding domain in IGF binding protein (IGFBP)-4 is essential for the IGF-II enhancement of IGFBP-4 proteolysis by osteoblast-produced IGFBP-4 protease. J. Bone Miner. Res. 14 (Suppl 1), S462 (1999) [Abstract SU123]. J. B. Lawrence, C. Oxvig, M. T. Overgaard, L. Sottrup-Jensen, G. J. Gleich, L. G. Hays, J. R. Yates, III, and C. A. Conover, The insulinlike growth factor (IGF)-dependent IGF binding protein-4 protease secreted by human fibroblasts is pregnancy-associated plasma protein-A. Proc. Natl. Acad. Sci. USA 96, 3149 – 3153 (1999). D. Byun, S. Mohan, C. Kim, K. Suh, M. Yoo, H. Lee, D. J. Baylink, and X. Qin, Studies on the human pregnancy-induced insulin-like growth factor binding protein (IGFBP)-4 poetesses in serum: Determination of IGF-II dependency and localization of cleavage site. J. Clin. Endocrinol. Metab. 85, 373 – 381 (2000). T. Kristensen, C. Oxvig, O. Sand, N. P. Moller, and L. SottrupJensen, Amino acid sequence of human pregnancy-associated plasma protein-A derived from cloned cDNA. Biochemistry 33, 1592 – 1598 (1994). C. A. Conover and M. C. Kiefer, Regulation and biological effect of endogenous insulin-like growth factor binding protein-5 in human osteoblastic cells. J. Clin. Endocrinol. Metab. 76, 1153 – 1159 (1993). C. A. Conover, L. K. Bale, J. T. Clarkson, and O. Torring, Regulation of insulin-like growth factor binding protein-5 messenger ribonucleic acid expression and protein availability in rat osteoblast-like cells. Endocrinology 132, 2525 – 2530 (1993). R. Knutsen, Y. Honda, D. D. Strong, T. K. Sampath, D. J. Baylink, and S. Mohan, Regulation of insulin-like growth factor system components by osteogenic protein-1 in human bone cells. Endocrinology 136, 857 – 865 (1995). K. M. Thrailkill, L. D. Quarles, H. Nagase, K. Suzuki, D. M. Serra, and J. L. Fowlkes, Characterization of insulin-like growth factorbinding protein 5-degrading proteases produced throughout murine osteoblast differentiation. Endocrinology 136, 3527 – 3533 (1995). C. Libanati, D. J. Baylink, E. Lois-Wenzel, N. Srinvasan, and S. Mohan, Studies on the potential mediators of skeletal changes occurring during puberty in girls. J. Clin. Endocrinol. Metab. 84, 2807 – 2814 (1999). M. R. Hammerman, Insulin-like growth factors and aging. Endocrinol. Metab. Clin. North Am. 16, 995 – 1011 (1987). E. S. Kurland, C. J. Rosen, F. Cosman, D. McMahon, F. Chan, E. Shane, R. Lindsay, D. Dempster, and J. P. Bilezikian, Insulin-like growth factor-I in men with idiopathic osteoporosis. J. Clin. Endocrinol. Metab. 82, 2799 – 2805 (1997).
429 232. E. Romagnoli, S. Minisola, V. Carnevale, R. Rosso, M. T. Pacitti, A. Scarda, L. Scarnecchia, and G. Mazzuoli, Circulating levels of insulin-like growth factor binding protein 3 (IGFBP-3) and insulinlike growth factor I (IGF-I) in perimenopausal women. Osteoporos. Int. 4, 305 – 308 (1994). 233. S. Boonen, S. Mohan, J. Dequeker, J. Aerssens, D. Vanderschueren, G. Verbeke, P. Broos, R. Bouillon, and D. J. Baylink, Down-regulation of the serum stimulatory components of the insulin-like growth factor (IGF) system (IGF-I, IGF-II, IGF binding protein [BP]-3, and IGFBP-5) in age-related (Type II) femoral neck osteoporosis. J. Bone Miner. Res. 14, 2150 – 2158 (1999). 234. S. Mohan, J. R. Farley, and D. J. Baylink, Age-related changes in IGFBP-4 and IGFBP-5 levels in human serum and bone: Implications for bone loss with aging. Prog. Growth Factor Res. 6, 465 – 473 (1995). 235. C. Rosen, L. R. Donahue, S. Hunter, M. Holick, H. Kavookjian, A. Kirschenbaum, S. Mohan, and D. J. Baylink, The 24/25-kDa serum insulin-like growth factor-binding protein is increased in elderly women with hip and spine fractures. J. Clin. Endocrinol. Metab. 74, 24 – 27 (1992). 236. Y. Honda, E. C. Landale, D. D. Strong, D. J. Baylink, and S. Mohan, S. Recombinant synthesis of insulin-like growth factor-binding protein-4 (IGFBP-4): Development, validation, and application of a radioimmunoassay for IGFBP-4 in human serum and other biological fluids. J. Clin. Endocrinol. Metab. 81, 1389 – 1396 (1996). 237. H. Tanaka, Liang, C. T. Mitogenic activity but not phenotype expression of rat osteoprogenitor cells in response to IGF-I is impaired in aged rats. Mech. Aging Dev. 92, 1 – 10 (1996). 238. J. Pfeilschifter, I. Diel, U. Pilz, K. Brunotte, A. Naumann, and R. Ziegler, Mitogenic responsiveness of human bone cells in vitro to hormones and growth factors decreases with age. J. Bone Miner. Res. 8, 707 – 717 (1993). 239. Y. D’ Avis P, C. R. Frazier, J. R. Shapiro, and N. S. Fedarko, Agerelated changes in effects of insulin-like growth factor I on human osteoblast-like cells. Biochem. J. 324, 753 – 760 (1997). 240. A. M. Delany and E. Canalis, Transcriptional repression of insulinlike growth factor I by glucocorticoids in rat bone cells. Endocrinology 136, 4776 – 4781 (1995). 241. S. Rydziel and E. Canalis, Cortisol represses insulin-like growth factor II receptor transcription in skeletal cell cultures. Endocrinology 136, 4254 – 4260 (1995). 242. B. Gabbitas and E. Canalis, Cortisol enhances the transcription of insulin-like growth factor-binding protein-6 in cultured osteoblasts. Endocrinology 137, 1687 – 1692 (1996). 243. B. Gabbitas J. M. Pash, A. M. Delany, and E. Canalis, Cortisol inhibits the synthesis of insulin-like growth factor-binding protein-5 in bone cell cultures by transcriptional mechanisms. J. Biol. Chem. 271, 9033 – 9038 (1996). 244. T. Chevalley, D. D. Strong, S. Mohan, D. Baylink, and T. A. Linkhart, Evidence for a role for insulin-like growth factor binding proteins in glucocorticoid inhibition of normal human osteoblast-like cell proliferation. Eur. J. Endocrinol. 134, 591 – 601 (1996). 245. C. Jux, K. Leiber, U. Hugel, W. Blum, C. Ohlsson, G. Klaus, and O. Mehls, Dexamethasone impairs growth hormone (GH)-stimulated growth by suppression of local insulin-like growth factor (IGF)-I production and expression of GH- and IGF-I-receptor in cultured rat chondrocytes. Endocrinology 139, 3296 – 3305 (1998). 246. C. Skjaerbaek, J. Frystyk, T. Grofte, A. Flyvbjerg, M. S. Lewitt, R. C. Baxter, and H. Orskov, Serum free insulin-like growth factor-I is dose-dependently decreased by methylprednisolone and related to body weight changes in rats. Growth Horm. IGF Res. 9, 74 – 80 (1999). 247. C. Libanati, S. Rajaram, D. J. Baylink, and S. Mohan, In a longitudinal prospective study, glucocorticoid therapy causes a marked reduction in serum bone formation parameters and stimulatory insulin-like
430
248. 249.
250. 251.
252.
253.
254.
255.
256.
257.
258.
259.
260.
261.
262.
263.
264.
XUEZHONG QIN ET AL. growth factor (IGF) system components. J. Bone Miner. Res. 11 (Suppl 1), S150 (1966) [Abstract P239]. G. Szebenyi and J. F. Fallon, Fibroblast growth factors as multifunctional signaling factors. Int. Rev. Cytol. 185, 45 – 106 (1999). H. A. Armelin, Pituitary extracts and steroid hormones in the control of 3T3 cell growth. Proc. Natl. Acad. Sci. USA 70, 2702 – 2706 (1973). C. Dickson and G. Peters, Potential oncogene product related to growth factors (letter). Nature 326, 833 (1987). M. Taira, T. Yoshida, K. Miyagawa, H. Sakamoto, M. Terada, and T. Sugimura, cDNA sequence of human transforming gene hst and identification of the coding sequence required for transforming activity. Proc. Natl. Acad. Sci. USA 84, 2980 – 2984 (1987). X. Zhan, B. Bates, X. G. Hu, and M. Goldfarb, The human FGF – 5 oncogene encodes a novel protein related to fibroblast growth factors. Molec. Cell. Biol. 8, 3487 – 3495 (1988). I. Marics, J. Adelaide, F. Raybaud, M. G. Mattei, F. Coulier, J. Planche, O. de Lapeyriere, and D. Birnbaum, Characterization of the HST-related FGF6 gene, a new member of the fibroblast growth factor gene family. Oncogene 4, 335 – 340 (1989). P. W. Finch, J. S. Rubin, T. Miki, D. Ron, and S. A. Aaronson, Human KGF is FGF-related with properties of a paracrine effector of epithelial cell growth. Science 245, 752 – 755 (1989). A. Tanaka, K. Miyamoto, N. Minamino, M. Takeda, B. Sato, H. Matsuo, and K. Matsumoto, Cloning and characterization of an androgen-induced growth factor essential for the androgen-dependent growth of mouse mammary carcinoma cells. Proc. Natl. Acad. Sci. USA 89, 8928 – 8932 (1992). M. Miyamoto, K. Naruo, C. Seko, S. Matsumoto, T. Kondo, and T. Kurokawa, Molecular cloning of a novel cytokine cDNA encoding the ninth member of the fibroblast growth factor family, which has a unique secretion property. Mol. Cell. Biol. 13, 4251 – 4259 (1993). P. M. Smallwood, I. Munoz-Sanjuan, P. Tong, J. P. Macke, S. H. Hendry, D. J. Gilbert, N. G. Copeland, N. A. Jenkins, and J. Nathans, Fibroblast growth factor (FGF) homologous factors: New members of the FGF family implicated in nervous system development. Proc. Natl. Acad. Sci. USA 93, 9850 – 9857 (1996). T. Nishimura, Y. Utsunomiya, M. Hoshikawa, H. Ohuchi, and N. Itoh, Structure and expression of a novel human FGF, FGF – 19, expressed in the fetal brain. Biochim. Biophys. Acta 1444, 148 – 151 (1999). E. Canalis, M. Centrella, and T. McCarthy, Effects of basic fibroblast growth factor on bone formation in vitro. J. Clin. Invest. 81, 1572 – 1577 (1988). R. K. Globus, P. Patterson-Buckendahl, and D. Gospodarowicz, Regulation of bovine bone cell proliferation by fibroblast growth factor and transforming growth factor beta. Endocrinology 123, 98 – 105 (1988). C. H. Kasperk, J. E. Wergedal, S. Mohan, D. L. Long, K. H. Lau, and D. J. Baylink, Interactions of growth factors present in bone matrix with bone cells: effects on DNA synthesis and alkaline phosphatase. Growth Factors 3, 147 – 158 (1990). D. Noff, S. Pitaru, and N. Savion, Basic fibroblast growth factor enhances the capacity of bone marrow cells to form bone-like nodules in vitro. FEBS Lett. 250, 619 – 621 (1989). S. B. Rodan, G. Wesolowski, K. Yoon, and G. A. Rodan, Opposing effects of fibroblast growth factor and pertussis toxin on alkaline phosphatase, osteopontin, osteocalcin, and type I collagen mRNA levels in ROS 17/2.8 cells. J. Biol. Chem. 264, 19934 – 19941 (1989). V. Shen, G. Kohler, J. Huang, S. S. Huang, and W. A. Peck, An acidic fibroblast growth factor stimulates DNA synthesis, inhibits collagen and alkaline phosphatase synthesis and induces resorption in bone. Bone Miner. 7, 205 – 219 (1989).
265. E. Canalis and J. B. Lian, Effects of bone associated growth factors on DNA, collagen and osteocalcin synthesis in cultured fetal rat calvariae. Bone 9, 243 – 246 (1988). 266. T. L. McCarthy, M. Centrella, and E. Canalis, Effects of fibroblast growth factors on deoxyribonucleic acid and collagen synthesis in rat parietal bone cells. Endocrinology 125, 2118 – 2126 (1989). 267. M. M. Hurley, C. Abreu, J. R. Harrison, A. C. Lichtler, L. G. Raisz, and B. E. Kream, Basic fibroblast growth factor inhibits type I collagen gene expression in osteoblastic MC3T3-E1 cells. J. Biol. Chem. 268, 5588 – 5593 (1993). 268. F. Debiais, M. Hott, A. M. Graulet, and P. J. Marie, The effects of fibroblast growth factor – 2 on human neonatal calvaria osteoblastic cells are differentiation stage specific. J. Bone Miner. Res. 13, 645 – 654 (1998). 269. M. Iwamoto, A. Shimazu, K. Nakashima, F. Suzuki, and Y. Kato, Reduction of basic fibroblasts growth factor receptor is coupled with terminal differentiation of chondrocytes. J. Biol. Chem. 266, 461–467 (1991). 270. F. De Luca and J. Baron, Control of bone growth by fibroblast growth factors. Trends in Endocrinol. Metab. 10, 61 – 65 (1999). 271. E. E. Mancilla, F. De Luca, J. A. Uyeda, F. S. Czerwiec, and J. Baron, Effects of fibroblast growth factor-2 on longitudinal bone growth. Endocrinology 139, 2900 – 2904 (1998). 272. F. Coulier, P. Pontarotti, R. Roubin, H. Hartung, M. Goldfarb, and D. Birnbaum, Of worms and men: An evolutionary perspective on the fibroblast growth factor (FGF) and FGF receptor families. J. Mol. Evol. 44, 43 – 56 (1997). 273. K. Keegan, D. E. Johnson, L. T. Williams, and M. J. Hayman, Isolation of an additional member of the fibroblast growth factor receptor family, FGFR-3. Proc. Natl. Acad. Sci. USA 88, 1095 – 1099 (1991). 274. W. C. Su, M. Kitagawa, N. Xue, B. Xie, S. Garofalo, J. Cho, C. Deng, W. A. Horton, and X. Y. Fu, Activation of Stat1 by mutant fibroblast growth-factor receptor in thanatophoric dysplasia type II dwarfism. Nature 386, 288 – 292 (1997). 275. D. M. Kilkenny and D. J. Hill, Perinuclear localization of an intracellular binding protein related to the fibroblast growth factor (FGF) receptor 1 is temporally associated with the nuclear trafficking of FGF-2 in proliferating epiphyseal growth plate chondrocytes. Endocrinology 137, 5078 – 5089 (1996). 276. R. E. Friesel and T. Maciag, Molecular mechanisms of angiogenesis: fibroblast growth factor signal transduction. FASEB J. 9, 919 – 925 (1995). 277. C. Basilico and D. Moscatelli, The FGF family of growth factors and oncogenes In “Advances in Cancer Research.” pp. 115 – 165. Department of Microbiology, New York University School of Medicine, New York, (1992). 278. L. W. Burrus, M. E. Zuber, B. A. Lueddecke, and B. B. Olwin, Identification of a cysteine-rich receptor for fibroblast growth factors. Mol. Cell Biol. 12, 5600 – 5609 (1992). 279. M. Fannon and M. A. Nugent, Basic fibroblast growth factor binds its receptors, is internalized, and stimulates DNA synthesis in Balb/c3T3 cells in the absence of heparan sulfate. J. Biol. Chem. 271, 17949 – 17956 (1996). 280. M. Roghani, A. Mansukhani, P. Dell’Era, P. Bellosta, C. Basilico, D. B. Rifkin, and D. Moscatelli, Heparin increases the affinity of basic fibroblast growth factor for its receptor but is not required for binding. J. Biol. Chem. 269, 3976 – 3984 (1994). 281. Z. Zhou, M. E. Zuber, L. W. Burrus, and B. B. Olwin, Identification and characterization of a fibroblast growth factor (FGF) binding domain in the cysteine-rich FGF receptor. J. Biol. Chem. 272, 5167 – 5174 (1997). 282. X. Xu, M. Weinstein, C. Li, and C. Deng, Fibroblast growth factor receptors (FGFRs) and their roles in limb development. Cell Tissue Res. 296, 33 – 43 (1999). 283. J. D. Coffin, R. Z. Florkiewicz, J. Neumann, T. Mort-Hopkins, G. W. Dorn, P. Lightfoot, R. German, P. N. Howles, A. Kier, and B. A.
CHAPTER 14 Bone Growth Factors
284.
285.
286.
287.
288.
289.
290.
291.
292.
293.
O’-Toole, Abnormal bone growth and selective translational regulation in basic fibroblast growth factor (FGF-2) transgenic mice. Mol. Biol. Cell 6, 1861 – 1873 (1995). S. Ortega, M. Ittmann, S. H. Tsang, M. Ehrlich, and C. Basilico, Neuronal defects and delayed wound healing in mice lacking fibroblast growth factor 2. Proc. Natl. Acad. Sci. USA 95, 5672 – 5677 (1998). B. Feldman, W. Poueymirou, V. E. Papaioannou, T. M. DeChiara, and M. Goldfarb, Requirement of FGF-4 for postimplantation mouse development. Science 267, 246 – 249 (1995). E. N. Meyers, M. Lewandoski, and G. R. Martin, An Fgf8 mutant allelic series generated by Cre- and Flp-mediated recombination. Nature. Genet. 18, 136 – 141 (1998). H. Min, D. M. Danilenko, S. A. Scully, B. Bolon, B. D. Ring, J. E. Tarpley, M. DeRose, and W. S. Simonet, Fgf-10 is required for both limb and lung development and exhibits striking functional similarity to Drosophila branchless. Genes Dev. 12, 3156 – 3161 (1998). G. Celli, W. J. LaRochelle, S. Mackem, R. Sharp, and G. Merlino, Soluble dominant-negative receptor uncovers essential roles for fibroblast growth factors in multi-organ induction and patterning. EMBO J. 17, 1642 – 1655 (1998). E. Arman, R. Haffner-Krausz, Y. Chen, J. K. Heath, and P. Lonai, Targeted disruption of fibroblast growth factor (FGF) receptor 2 suggests a role for FGF signaling in pregastrulation mammalian development. Proc. Natl. Acad. Sci. USA 95, 5082 – 5087 (1998). J. S. Colvin, B. A. Bohne, G. W. Harding, D. G. McEwen, and D. M Ornitz, Skeletal overgrowth and deafness in mice lacking fibroblast growth factor receptor 3. Nature Genet. 12, 390 – 397 (1996). C. Deng, A. Wynshaw-Boris, F. Zhou, A. Kuo, and P. Leder, Fibroblast growth factor receptor 3 is a negative regulator of bone growth. Cell 84, 911 – 921 (1996). H. Kawaguchi, T. Kurokawa, K. Hanada, Y. Hiyama, M. Tamura, E. Ogata, T. Matsumoto, Stimulation of fracture repair by recombinant human basic fibroblast growth factor in normal and streptozotocin-diabetic rats. Endocrinology 135, 774 – 781 (1994). Y. S. Bland, M. A. Critchlow, and D. E. Ashhurst, D. E. Exogenous fibroblast growth factors-1 and -2 do not accelerate fracture healing in the rabbit. Acta Orthop. Scand. 66, 543 – 548 (1995).
431 294. T. Kato, H. Kawaguchi, K. Hanada, I. Aoyama, Y. Hiyama, T. Nakamura, K. Kuzutani, M. Tamura, T. Kurokawa, and K. Nakamura, Single local injection of recombinant fibroblast growth factor-2 stimulates healing of segmental bone defects in rabbits. J. Orthop. Res. 16, 654 – 659 (1998). 295. T. Nakamura, Y. Hara, M. Tagawa, M. Tamura, T. Yuge, H. Fukuda, and H. Nigi, Recombinant human basic fibroblast growth factor accelerates fracture healing by enhancing callus remodeling in experimental dog tibial fracture. J. Bone Miner. Res. 13, 942 – 949 (1998). 296. M. L. Radomsky, T. B. Aufdemorte, L. D. Swain, W. C. Fox, R. C. Spiro, and J. W. Poser, Novel formulation of fibroblast growth factor – 2 in a hyaluronan gel accelerates fracture healing in nonhuman primates. J. Orthop. Res. 17, 607 – 614 (1999). 297. K. Nakamura, H. Kawaguchi, I. Aoyama, K. Hanada, Y. Hiyama, T. Awa, M. Tamura, and T. Kurokawa, Stimulation of bone formation by intraosseous application of recombinant basic fibroblast growth factor in normal and ovariectomized rabbits. J. Orthop. Res. 15, 307 – 313 (1997). 298. Y. Tabata, K. Yamada, L. Hong, S. Miyamoto, N. Hashimoto, and Y. Ikada, Skull bone regeneration in primates in response to basic fibroblast growth factor. J. Neurosurg. 91, 851 – 856 (1999). 299. C. R. Dunstan, R. Boyce, B. F. Boyce, I. R. Garrett, E. Izbicka, W. H. Burgess, and G. R. Mundy, Systemic administration of acidic fibroblast growth factor (FGF-1) prevents bone loss and increases new bone formation in ovariectomized rats. J. Bone Miner. Res. 14, 953 – 959 (1999). 300. H. Nagai, R. Tsukuda, and H. Mayahara, Effects of basic fibroblast growth factor (bFGF) on bone formation in growing rats. Bone 16, 367 – 373 (1995). 301. P. Cuevas, F. Carceller, S. Ortega, M. Zazo, I. Nieto, G. GimenezGallego, Hypotensive activity of fibroblast growth factor. Science 254, 1208 – 1210 (1991). 302. T. K. Rosengart, S. R. Patel, and R. G. Crystal, Therapeutic angiogenesis: protein and gene therapy delivery strategies. J. Cardiovasc. Risk 6, 29 – 40 (1999). 303. J. M. Isner, The role of angiogenic cytokines in cardiovascular disease. Clin. Immunol. Immunopathol. 80, S82 – S91 (1996).
CHAPTER 15
Skeletal Heterogeneity and the Purposes of Bone Remodeling Implications for the Understanding of Osteoporosis A. M. PARFITT
Division of Endocrinology, University of Arkansas for Medical Sciences, Little Rock, Arkansas 72205
I. Introduction II. Skeletal Heterogeneity III. Purposes of Bone Remodeling
IV. Implications for Understanding Osteoporosis References
I. INTRODUCTION
sense, implies a target. The target value of any regulatory process in biology has been optimized by natural selection. Mechanisms have evolved that ensure that deviations from the target are detected and that corrective measures to restore the target value are carried out. In this sense, body temperature, extracellular fluid osmolality, tissue oxygen tension, and countless other physiologic quantities are regulated, but the mechanisms of regulation could not be determined until the existence of the target had been recognized and its precise nature defined. Is there a target for bone remodeling or for some characteristic of bone that is influenced by remodeling? The piecemeal, quantal nature of bone remodeling is well known. The process is carried out by temporary anatomic structures known as basic multicellular units (BMUs) [2 – 4],
The cells of bone influence its structure by means of four processes: growth, repair, modeling, and remodeling, the last being the basis of bone tissue turnover in the adult skeleton. The purposes of growth and repair are obvious. Modeling serves to adapt bones to changes in mechanical loading, a process that is most effective during growth [1]. But why does a tissue that can survive for thousands of years after death need to be maintained by periodic replacement during life? Most of those interested in bone, whether as physicians, as clinical investigators, or as basic scientists, show remarkably little interest in this fundamental question. Many articles and book chapters discuss the regulation of bone remodeling, but regulation, at least in the physiologic
OSTEOPOROSIS, SECOND EDITION VOLUME 1
433
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
434
A. M. PARFITT
which excavate and replace tunnels through cortical bone (osteonal remodeling) or trenches across the surface of cancellous bone (hemiosteonal remodeling). Each BMU includes two teams of executive cells (osteoclasts and osteoblasts), supported by blood vessels, nerves, and loose connective tissue. The life span of the BMU, which comprises separate stages of origination, progression, and termination, is measured in months, but the life span of osteoblasts while they are making bone is measured in weeks, and the life span of osteoclast nuclei is measured in days. During progression of the BMU through or across the surface of bone the spatial and temporal relationships between its components are maintained by the continued growth of the central capillary [3] and recruitment of new cells [3 – 5]. These cells, like the formed elements of the blood, originate from stem cells in the bone marrow [6], except that in the peripheral skeleton, osteoblasts are derived from local precursors [5]. For blood cells, as for other short-lived cells, control of cell production and survival is more important than control of differentiated cell function; although the details are less clear, the same also applies to bone cells [6]. Each type of blood cell is normally produced at a basal rate that is sufficient for ordinary purposes but which can be increased when needed [7]. For each cell type, the circumstances under which demand is increased are well known and are related to the function of the particular cell, although the cell types differ with respect to the time scale of this response, its specificity, the relative importance of reactive and anticipatory homeostasis [8], and the extent to which the control mechanisms have been elucidated. The importance of these relationships between supply and demand, and between demand and function, also applies to bone cells. For osteoblasts in the adult nongrowing skeleton, the demand is created by bone resorption, as the function of osteoblasts is to replace the bone removed by osteoclasts. However, the circumstances that create a demand for osteoclasts are much less well defined, as these circumstances are dictated by the purposes of bone remodeling. Indeed, the questions “What are the purposes of bone remodeling and how are they achieved?” are essentially equivalent to the questions “Where and when are osteoclasts needed, and how is this need recognized and satisfied?” The answers to these questions are different in different types of bone and in different regions of the skeleton.
II. SKELETAL HETEROGENEITY A. Structure and Function The structural differences between cortical bone, in which porosity and surface-to-volume ratio are low, and cancellous bone, in which these geometric quantities are high [9], are now widely recognized. All intermediate values for these quantities can occur, but they are infrequent,
TABLE 1
Subdivisions of the Skeleton
Feature
Central
Peripheral
Main bone tissue
Cancellous
Cortical
Main soft tissue
Viscera
Muscle
Main joint type
Various
Synovial
Cortices
Thin
Thick
Marrow
Hematopoietic
Fatty
Turnover
High
Low
implying that transitional structures tend to be temporary and short-lived [10]. Less often noted are the differences between axial and appendicular subdivisions of the skeleton (Table 1); the pelvis, defined anatomically as appendicular, behaves functionally as part of the axial skeleton, so that it is more accurate to contrast central with peripheral regions. This distinction is important because the different functions of the skeleton are divided differently between central and peripheral components. The primary function is load bearing: to support posture, permit movement (including locomotion), and provide protection for the soft tissues. Subsidiary functions participate in mineral homeostasis and provide a favorable microenvironment for hematopoiesis. For convenience the former functions will be referred to as “mechanical” and the latter as “metabolic” [7]. It is commonly believed that the mechanical functions are carried out mainly by cortical bone and the metabolic functions mainly by cancellous bone, regardless of their central or peripheral locations. In fact, the functions of the peripheral skeleton, cancellous as well as cortical, are mainly mechanical, whereas the central skeleton, cortical as well as cancellous, in addition to its mechanical function, participates to a much greater extent in the metabolic functions of bone. This revision in functional attribution is most striking for peripheral cancellous bone, such as in the metaphyses of the long bones [11]. As is evident from the orientation of the trabeculae (Fig. 1a), metaphyseal cancellous bone transmits loads from the joint surfaces to diaphyseal cortical bone. Indeed, the metaphyses are flared in shape precisely to make such load transmission possible. Similar functional and architectural considerations apply to the cancellous bone in the small bones of the hands and feet (Fig. 1b). As will subsequently be discussed in detail, there is no evidence that such peripheral cancellous bone participates to a significant extent in the metabolic functions of the skeleton, whether related to mineral homeostasis or to hematopoiesis.
B. Remodeling and Turnover The frequent assertion that cancellous bone has higher turnover than cortical bone is usually supported by comparing
435
CHAPTER 15 Skeletal Heterogeneity and the Purposes of Bone Remodeling
the vertebral bodies in particular [15], is quite unrepresentative of the peripheral skeleton [16,17]. Based on a variety of indirect methods, turnover in peripheral cortical bone is lower by about half than in the ribs (around 2%/year vs 4%/year) [18]. For peripheral cancellous bone, estimates of turnover are based on fewer data, but they suggest that the central – peripheral difference is greater than for cortical bone. During the treatment of osteomalacia, the increase in cancellous bone mineral was about 35% in the ilium, measured histologically, but only 1 – 2% in the distal radius, measured by single photon absorptiometry [19]. On the reasonable assumption that unmineralized osteoid tissue accumulates during the evolution of osteomalacia in proportion to the initial rate of turnover, this rate in the cancellous bone of the distal radius is normally only about 2%/year. This is similar to the estimate for peripheral cortical bone; because the surfaceto-volume ratio would be higher in cancellous bone, the activation frequency would be even lower than on the intracortical surfaces. Direct measurements of turnover in peripheral and central cancellous bone in the beagle confirm a much lower value for the former, even though the absolute values for both were higher than in human subjects [20]. FIGURE 1
(a) Examples of trabecular orientation in metaphyseal cancellous bone in the appendicular skeleton. Alignment with stress trajectories facilitates the transmission of loads from the joints to diaphyseal cortical bone. (b) Examples of trabecular orientation in the small bones of the feet. Alignment with stress trajectories facilitates the transmission of loads during locomotion to the ankle joint and hence to diaphyseal cortical bone in the tibia. Modified from Meyer [11].
central cancellous with peripheral cortical bone, but this is to confuse the geometrical and biological factors that influence turnover. The remodeling process occurs only on bone surfaces, and the intensity of remodeling is expressed by the activation frequency, which is the reciprocal of the average time interval between the initiation of consecutive cycles of remodeling at the same surface location, referred to as the regeneration period [12]. Turnover refers to volume replacement, which depends not only on the surface-defined activation frequency but on the surface-to-volume ratio. This geometrical property is about four to five times higher in typical cancellous bone than in typical cortical bone [9,13]. Consequently, the former could have a higher turnover despite a lower intensity of remodeling. Systematic site-specific measurements of turnover in the human skeleton are available only for the rib [2] and for the ilium. In the latter, because activation frequency is similar on cancellous, endocortical, and intracortical subdivisions of the endosteal envelope [14,15], the difference in turnover between cortical and cancellous bone at this site depends entirely on the difference in the surface-to-volume ratio. Unfortunately, the ilium, although probably representative of the central skeleton in general and of
C. Relationship to Marrow Composition In the embryo, hematopoietic marrow appears first in the yolk sac and subsequently migrates to the liver and spleen and then to the marrow cavities. At birth, hematopoiesis is active in cancellous bone throughout the skeleton but has virtually ceased at extramedullary sites [21,22]. During growth, there is gradual conversion of red to yellow marrow, a process that begins in the distal extremities and proceeds centripetally. By age 25, hematopoiesis has disappeared from the peripheral skeleton, except to a limited extent in the upper femora [22]. Macroscopically visible hematopoiesis continues in the central skeleton throughout life, although there is a gradual increase in the number of fat cells at the expense of hematopoietic cells. At any age, a sustained increase in demand can lead to the reappearance of hematopoiesis in the extremities [22,23]. Whether this results from reactivation of dormant local stem cells or from recolonization of fatty marrow by circulating stem cells is not known, although there is strong evidence that hematopoietic stem cells do circulate [24]. Data presented, although incomplete, indicate that in the adult human skeleton central cancellous bone has persistent hematopoiesis and high bone turnover, whereas peripheral cancellous bone has absent hematopoiesis and low bone turnover (Table 2). Furthermore, based on external radionuclide counting, there is a close correlation between the extent of hematopoiesis and bone blood flow [25]. When different bones sampled at autopsy were compared, there
436
A. M. PARFITT
TABLE 2
Cancellous Bone and Its Marrow
Feature
Red marrow
Yellow marrow
Bone type
Metabolic
Mechanical
Location
Central
Peripheral
Main function
Calcium homeostasis
Transmit loads
Support hematopoiesis
Absorb energy
Cellularity
High
Low
Blood flow
High
Low
Turnover
High
Low
III. PURPOSES OF BONE REMODELING
was a good relationship between the proportion of cancellous bone surface in contact with hematopoietic cells and the proportion engaged in bone remodeling [26]. In adult beagles, there is an even more striking correspondence between marrow composition and bone remodeling (Table 3). Adjacent to red marrow there is a 15% higher mineral apposition rate and an almost 10-fold higher bone formation rate than adjacent to yellow marrow, with corresponding differences in the uptake of plutonium [27,28]. If, as seems likely, there are no hematopoietic stem cells in yellow marrow, all osteoclasts in the peripheral skeleton, cancellous as well as cortical, must be derived from circulating mononuclear precursor cells [5,7]. In the axial skeleton, osteoclast precursors might be able to migrate directly to the bone surface, but participation of the local microcirculation has now been established [5,29]. The relationship between marrow composition and remodeling can be disturbed in pathologic conditions. For example, after ovariectomy, both bone turnover and amount of fat in the marrow increase [30]. No relationship between marrow composition and bone remodeling was found in a single patient with osteoporosis who died from an unrelated cause after the administration of tetracycline labels in preparation for bone biopsy [31]. The relationship is also disturbed by proximity to synovial joints; turnover is higher within 1 mm of the articular surface than at more distant locations [32]. Nevertheless, the spatial association between TABLE 3
Cancellous Bone Turnover in Normal Beaglesa
Site
Marrow
MARb(m/day)
BFRc(%/year)
Lumbar vertebra
Red
1.29 0.10
Proximal humerus
Red
1.23 0.10
89 18
Pelvis
Red
1.26 0.10
83 25
Proximal ulna
Yellow
0.90 0.06
13 6
Distal ulna
Yellow
0.97 0.07
73
106 9
Data expressed as mean SE. From Wronski et al. [27,28]. Mineral apposition rate; n 8. c Bone formation rate; n 4. a b
hematopoiesis and active remodeling appears to be characteristic of the healthy skeleton. To most observers, this is simply the expected consequence of the presence or absence of precursor cells in close proximity to the bone surface, but this is a superficial view. Why does cancellous bone need to turn over so much faster in some locations than in others?
There is probably no physiologic function other than bone remodeling that has attracted so much study in the face of so much uncertainty about why it occurs. Many in the field act as if they believed that the only purpose of remodeling was to cause osteoporosis and thus provide employment for scientists and business opportunities for the pharmaceutical industry! In the analysis of this problem, it seems reasonable to make two assumptions. First, periodic replacement of bone serves to maintain its ability to carry out its functions, as summarized previously. Second, because the most obvious difference between the old bone removed and the new bone put in its place is in their ages, excessive age of bone in some way compromises its functional capacity; in general, there is a reciprocal relationship between the rate of turnover and the mean age of the bone. Because the primary function of bone is mechanical, the primary purpose of remodeling of bone is to maintain its load-bearing capacity. This is accomplished partly by preventing structural damage at the microscopic and submicroscopic levels and partly by repairing such damage after it occurs.
A. Fatigue Damage and Mechanical Competence All structural materials that undergo repetitive cyclical loading are subject to fatigue, a phenomenon that has been studied most extensively in fabricated materials such as steel [33]. After a certain number of load cycles, tiny cracks appear, which are detectable at first at the ultramicroscopic level, but were probably preceded by damage at submicroscopic and molecular levels. If cyclical loading continues, the cracks extend and accumulate into macroscopic damage and eventually into overt structural failure. In bone, subtle molecular changes in matrix constituents may also appear with increasing material age [34]. The essence of fatigue is that in each cycle, the load is far below the instantaneous breaking strength of the intact material. Biological materials such as bone also undergo fatigue damage, but differ from man-made materials in their capacity for self-repair. The occurrence of fatigue damage has been demonstrated unequivocally in cortical bone [35,36], and there is compelling evidence that experimentally induced fatigue damage in
437
CHAPTER 15 Skeletal Heterogeneity and the Purposes of Bone Remodeling
cortical bone induces repair by remodeling so that the damaged bone is removed and replaced by new undamaged bone [36,37]. It is reasonable to assume that the same applies to load-bearing cancellous bone, although this has not yet been demonstrated. Microcracks occur in human cancellous bone [38,39], and various degrees of microdamage can be identified [40]; because they do not accumulate with age, unlike those in femoral cortical bone [41], they must be repaired by remodeling. However, there is no evidence that these lesions are due to fatigue; indeed, identical lesions can be produced experimentally by compression [42]. Microfractures in cancellous bone heal by callus formation rather than by remodeling [43], and most of them (at least in the vertebrae) can also be explained, not by fatigue, but by instantaneous overload, leading to failure by buckling [44]. Evidently, one function of remodeling in load-bearing bone is to provide a means for replacing bone that has undergone fatigue microdamage. However, this cannot be the only way in which remodeling maintains the mechanical competence of bone. The similarity between different members of the same species in the spatial distribution of remodeling activity at different skeletal sites [45] cannot be explained by a mechanism that is purely reparative. One of the most striking aspects of such remodeling maps is their bilateral symmetry, such that cross sections at the same level of bones on opposite sides of the body are virtually mirror images of one another [33,46]. It is exceedingly unlikely that such consistent symmetry could be the expression of fatigue damage repair, but it could well be an expression of fatigue damage prevention. Increased mechanical usage might stimulate remodeling before the occurrence of damage [47], but there are several difficulties with this proposal [2,33,46,48]. For material of the same mechanical properties, the major determinant of fatigue damage is the number of load cycles, and for the same level of physical activity, the major determinant of the number of load cycles is the age of the structure. The customary pattern and intensity of physical activity are species specific and so are genetically programmed [1]. Consequently, it seems possible that the remodeling map is the expression of a genetic program to prevent bone age from exceeding some critical level, a level that is different in different regions of the skeleton [45]. The contrast between the prevention of fatigue and other forms of damage by keeping bone age below some critical value and the repair of such damage by removal of the bone involved is analogous to the contrast between anticipatory and reactive homeostasis [8], except that the basis of the anticipation is genetic rather than physiologic. More specifically, it exemplifies the distinction between stochastic1 remodeling and targeted remodeling, a distinction that establishes an order of priority for different remodeling projects. There is a wide range of turnover rates consistent with skeletal health [2,18], and the low rates that occur in hypothyroidism [50] and hypoparathyroidism [51] do not
appear to increase fracture risk. Presumably this is because stochastic remodeling to prevent excessive bone age provides a substantial margin of safety. Consequently, curtailing the progression of a particular BMU, which is a likely basis of stochastic remodeling [52], is unlikely to have any harmful effects. However, targeted remodeling, to remove fatigued bone before the damage escalates from microscopic to macroscopic, must be carried out promptly or else it will fail in its purpose. The existence of such a temporal hierarchy has an important impact on the therapeutic reduction of bone turnover, a point that will subsequently be discussed in more detail. The mechanism of targeted remodeling is becoming a little clearer. The only cell that is in the right location to detect microscopic damage is the osteocyte. This cell can be activated by mechanically induced strain to increase protein synthesis [53], but the relationship of this phenomenon to damage detection is unknown. The osteocytes must transmit a signal to the cells lining the nearest free bone surface, instructing them to originate a new BMU, but whether the signal is biochemical, electrical, hydraulic, or neural is unknown [7]. In low turnover load-bearing bone, whether cortical or cancellous, there will be no osteoclast precursor cells in the vicinity of the surface so that the lining cells must transmit a second signal to the nearest capillary, which induces circulating mononuclear osteoclast precursors to leave the circulation by an area code mechanism, analogous to that which attracts circulating neutrophils to sites of inflammation [4,5,7]. Once the new BMU is in place, it must find its way to the site of damage. Matrix microdamage induces local osteocyte apoptosis in the bone that will be removed, and molecules released by dying cells could serve as a homing signal for the approaching BMU [54]. Many other factors can influence one or more steps in this complex process, but their role is more likely to be permissive than regulatory [3].
B. Metabolic Functions of Remodeling The foregoing argument has established three interconnected facts. First, the primary function of metaphyseal cancellous bone in the extremities is mechanical load bering. Second, the reason why load-bearing bone must be remodeled is to maintain its mechanical competence. Third, the rate of turnover of load-bearing bone adjacent to fatty marrow, whether cortical or cancellous, is low. Clearly, a low rate of turnover, of the order of 2 – 5%/year, is sufficient to maintain the mechanical competence of bone, regardless of its location in the skeleton or its geometric features. Consequently,
1 By stochastic is meant a succession of events that are individually random but collectively constitute a process that is amenable to study [49].
438 the rate of turnover of axial cancellous bone adjacent to hematopoietic marrow (15 – 35%/year) is much higher (by a factor of at least five) than is necessary to maintain mechanical competence. Unless this mechanically surplus or spare remodeling is simply a form of occupational therapy for cells with nothing better to do, it must subserve an entirely different purpose. This conclusion will not surprise the many endocrinologists who have always believed that the main purpose of bone remodeling was to support calcium homeostasis, but the restriction of this function mainly to cancellous bone adjacent to red marrow has not previously been emphasized. The relative importance of the mechanical and metabolic aspects of remodeling, debated inconclusively for many years [46], is evidently different in different regions of the skeleton, although both are essential to the organism as a whole. The most important nonmechanical function of bone remodeling concerns the regulation of calcium homeostasis. Bone is involved in both determining the steady-state target value for plasma-free calcium and correcting deviations from the target value [55]. Both of these processes depend on a relatively high rate of bone remodeling, but in quite different ways. Bone mineral also functions as a reservoir for sodium and as a buffer for hydrogen ion regulation. Bone remodeling may also provide biochemical support for hematopoiesis as well as the mechanical support provided by the bone itself. Both stem cell number and proliferative activity are greatest adjacent to the endosteal surface [56], and for this reason, bone-lining cells may need timely replacement. Bone matrix contains growth factors and other regulatory molecules, some of which may act on blood-forming cells rather than on bone cells. For several reasons, it could be advantageous for such molecules to be released into the bone marrow during bone resorption rather than directly from the cells involved in their biosynthesis. Possible reasons include cell polarization, with osteoblasts transporting substances away from, and osteoclasts toward, the marrow, the high proton concentration within the ruffled border of osteoclasts, and a need for intermittent rapid release rather than more continuous slow release. However, this is speculative, and the remainder of the discussion focuses on the relationship between bone remodeling and calcium homeostasis. Except under conditions of extreme calcium deprivation, the calcium homeostatic function of remodeling is not antagonistic to the mechanical function, as normally calcium homeostasis does not depend on continued net loss of calcium from bone. Steady-state levels of plasma-free calcium can be high, normal, or low, regardless of the directional changes in osteoclastic bone resorption or in calcium balance [57]. Plasma-free calcium is regulated by the joint effects of parathyroid hormone (PTH) on the renal tubular reabsorption of calcium and on the blood – bone equilibrium. This equilibrium is achieved when the inward and outward fluxes of calcium at quiescent bone surfaces are equal, and the calcium level at which this occurs is determined by
A. M. PARFITT
some effect of PTH on bone-lining cells [55]. For this mechanism to be effective, several conditions must be met. First, there must be a high blood flow, which is ensured by the proximity of hematopoietic marrow. Second, the bone at the surface must retain enough water to permit rapid diffusion of minerals, which is ensured by a high rate of remodeling. As bone ages, secondary mineralization proceeds slowly to completion by crystal enlargement and displacement of water, with a progressive decline in its ability to support the rapid mineral exchanges on which plasmacalcium homeostasis depends [55]. Stochastic remodeling could prevent excessive aging of surface bone, but as for fatigue damage, from time to time targeted remodeling will be needed to remove bone that has become hypermineralized. The mechanism of targeting is even less understood than for fatigue damage but should be simpler, as the bone to be removed is on rather than beneath the surface. As well as determining the steady-state target level of plasma-free calcium, the bone also participates in the correction of deviations from the target value. A fall in plasmafree calcium stimulates PTH secretion, which increases the outflow of calcium from bone, not only by shifting the balance of exchange at quiescent bone surfaces but also by increasing the resorptive activity of existing osteoclasts. This acute effect is quite separate from the long-term effect of PTH to increase activation frequency, osteoclast recruitment, and bone turnover in primary and secondary hyperparathyroidism. Obviously, the rapidity of the correction depends on the number of osteoclasts available, which is determined by the number of BMUs present and by the efficiency of the local circulation. The most important use for this mechanism is to accommodate the circadian changes in the supply of calcium from intestinal absorption, with an approximately 12- to 16-h period of eating, followed by an 8- to 12-h period of fasting, during which both PTH secretion and bone resorption increase [58]. In each BMU, the cutting cone (in osteonal remodeling) or hemicone (in hemiosteonal remodeling) advances more rapidly at night and slows down to allow the closing cone (or hemicone) to catch up during the day. This concertina-like action (Fig. 2) allows the skeleton to supply calcium at night when it is needed, without affecting the terminal balance of the BMUs and so without causing an irreversible loss of bone. A final aspect of remodeling and homeostasis is that the remodeling apparatus can supply a temporary but sustained demand for calcium lasting for many months by a temporary increase in BMU origination and a corresponding increase in the remodeling-dependent reversible mineral deficit [59]. The best known example is cyclic physiologic osteoporosis in deer, in which a seasonal increase in cortical porosity is entrained to the antler growth cycle [60]. The phenomenon has been demonstrated only in ribs; whether it is confined to the central skeleton or affects the peripheral skeleton as well is not known. The
CHAPTER 15 Skeletal Heterogeneity and the Purposes of Bone Remodeling
439
1. MECHANISMS OF BONE LOSS
FIGURE 2
Contribution of BMU-based remodeling to short-term demands for calcium. During the night, osteoclasts of the cutting cone (Rs) advance more quickly than osteoblasts of the closing cone (F), increasing the reversal zone (Rv). During the day, the cutting cone slows down and the closing cone catches up. The same concertina-like action can occur with cancellous BMUs (hemiosteonal remodeling). Reprinted with permission from Parfitt [55].
same phenomenon can partly satisfy the increased demand for calcium that occurs during growth, pregnancy, and lactation; based on densitometric data, in these circumstances the peripheral skeleton is also involved. During the adolescent growth spurt, some of the calcium needed for endochondral ossification and subperiosteal apposition is provided by a further increase in the already high cortical porosity, which subsides after cessation of longitudinal growth [1].
IV. IMPLICATIONS FOR UNDERSTANDING OSTEOPOROSIS “Osteoporosis” is a convenient term with which to cover the health implications of two related phenomena. First, bone mass in individuals falls with age. Second, partly as a result, the incidence of fractures in the population rises with age. Regrettably, for a variety of nonmedical and nonscientific reasons, it has become fashionable to define “osteoporosis” as a disease that is either present or absent, but in this text the term is used only in the former sense.
A. Pathogenesis of Fractures The relationship of bone remodeling to bone loss and to bone fragility is considered separately, as bone loss is not the only cause of increased bone fragility.
The most remarkable feature of age-related bone loss is its universality. There are useful analogies between osteoporosis and hypertension [61,62], but there are also differences. In some communities remote from western civilization, mean blood pressure does not rise with age. However, there is no subset of the human species in which mean bone mass does not fall with age, although the rate and magnitude of loss may differ between individuals and between groups [63]. Bone loss not only affects almost all persons but almost every bone, and it is of interest to compare the observed rates of loss at different skeletal sites with those predicted from remodeling theory. There are many problems in comparing rates of loss between different sites [64], including differences in methodology, instrumentation, and units. Rates of bone loss are usually expressed as a percentage of the initial value per year. This is not the best way of comparing measurements at the same site between individuals or groups [65], but in the absence of a better mathematical model, it is the most practical way of comparing different sites. For a few years after menopause, the rate of loss is substantially faster for vertebral cancellous bone than for cancellous bone at other sites and for cortical bone [66,67], but the wider the age range over which data are collected, the more similar the rates become. For example, 25 years after menopause the average amount of bone that has been lost in healthy women is about 35% of the initial value (or about 1.4%/year) in both the vertebral bodies and the distal forearm [66]. In the ilium, loss of cancellous bone, measured histologically in autopsy specimens, is about 1%/year in women between the ages of 25 and 75 [68,69]. About the same rate of loss is found in biopsy specimens, and the proportional loss of cancellous and cortical bone is very similar [9,13]. Likewise, in healthy women studied between the ages of 55 and 75 years, the average rates of loss (%/year) were 1.0 in the distal radius, 1.2 in the calcaneum, and 1.4 in the proximal radius [70]. Thus, at both central and peripheral sites, comprising various proportions of cortical and cancellous bone, the long-term rates of bone loss measured cross-sectionally are in the range of 1 – 1.5%/year. In cross-sectional studies the subjects differ not only in age but in year of birth and so may have been subject to different environmental influences [71]. This generational or cohort effect could increase the apparent rate of loss compared to longitudinal studies, but the latter can be continued only for relatively short periods. Furthermore, because the cohort effect would apply to every site, the real differences between sites are even smaller than they appear. All bone loss occurs from one of the internal surfaces of bone, and the rate of loss from any surface location depends on the average bone deficit at the end of each cycle of remodeling and the frequency with which cycles occur on that surface [12,72]. Thus for the same focal imbalance,
440 the rate of bone loss from a surface is proportional to the rate of remodeling on that surface [15,72]. It is impossible to measure remodeling rates at individual surface locations noninvasively, but biochemical indices of bone turnover reflect the aggregate of the separate contributions of each BMU currently present in the skeleton, although each index is also influenced by several other factors [73]. In accordance with remodeling theory, differences in these indices between persons are significantly correlated with differences in the rate of bone loss [74]. However, when different sites are compared, a serious paradox emerges. Remodeling theory predicts that for the same focal imbalance, the average rate of loss will be about five times higher from cancellous bone adjacent to red marrow than from cancellous bone adjacent to yellow marrow because of their difference in turnover, but sustained differences of even half this magnitude have never been demonstrated. The inescapable conclusion is that the degree of focal remodeling imbalance in, for example, the calcaneum is much greater than in the ilium, the only site where such imbalance has so far been measured [14,15]. For the same absolute rates of bone loss from a surface, the fractional loss depends on the thickness of bone beneath the surface, and hence is proportional to the surfaceto-volume ratio [65]. Accordingly, it would be expected that for the same degree of remodeling imbalance and the same frequency of remodeling activation, the average fractional rate of bone loss would be about five times higher in cancellous than in cortical bone because of their difference in surface-to-volume ratio. However, again, sustained differences in rates of bone loss of even half this magnitude have never been demonstrated. Only in the ilium have rates of both remodeling and bone loss been measured at both cortical and cancellous sites in the same bone. As mentioned previously, the results indicated similar rates of surface remodeling, similar fractional rates of bone loss, much larger absolute rates of loss from the endocortical surface, and, by inference, much greater remodeling imbalance on this surface [9,14,15]. In primary hyperparathyroidism, in normal age and menopause-related bone loss, and in patients with vertebral fracture, cortical thinning is mainly the result of increased resorption depth [14,15], which is the two-dimensional reflection of deeper penetration by endocortical BMUs. The same phenomenon has been demonstrated in the rib [75] and inferred for the metacarpal [72] and is presumably a universal feature of cortical bone loss throughout the skeleton. Furthermore, the similarity in fractional rates of bone loss indicates that the increase in resorption depth at different sites is inversely related to the customary rate of turnover and is positively related to the usual thickness of cortical bone, at each site. This is a remarkable and unexpected conclusion. When bone loss is both generalized and sustained, as in normal
A. M. PARFITT
aging, it appears that resorption depth at different sites increases to the extent necessary to bring about roughly the same rates of fractional bone loss and, as it were, “compensates” for differences in bone turnover contingent on differences in marrow composition and for differences in local bone structure and geometry. The only conceivable kind of explanation for such a phenomenon is biomechanical [2]. All mechanical influences on bone remodeling are mediated by strain, the technical term for relative deformation of a structural material as the result of load bearing. Similar fractional rates of bone loss throughout the skeleton will produce similar proportional changes in the strains that occur in different bones as a result of the same pattern and intensity of physical activity. Frost [76], building on earlier work by others [33,77], has proposed the existence of the “mechanostat,” which orchestrates the recruitment and activity of osteoclasts and osteoblasts in such a way that strain is maintained within an acceptable range [78]. The primary function of the mechanostat is to ensure that each bone during growth acquires the strength it needs to support the species-specific pattern and intensity of physical activity customary during adult life [1]. After growth has ceased, the mechanostat is much less effective in adapting the bones to an increase in mechanical demand, but is highly effective in adapting them to a decrease, accounting for the rapidity, severity, and usual irreversibility of bone loss consequent on disuse [79]. As a result of the sedentary lifestyle made possible by economic development, aging in most persons is accompanied by a progressive reduction in physical activity and muscle strength, of earlier onset and greater severity than is biologically mandated [80]. According to biomechanical theory, this should not increase the risk of fracture, as the reduced bone mass would remain appropriate to the reduced level of activity, but this does not take into account the age-related increase in the liability to fall, to which the mechanostat is blind. Frost has postulated that as a result of estrogen deficiency, the mechanostat is reset so that the skeleton responds not so much to actual but to erroneously perceived disuse [76]. A universal resetting of the mechanostat would not account for the disproportionately rapid loss of central cancellous bone in the first few years after menopause, so it is more likely not the result of estrogen deficiency, but of the aging process. 2. MECHANISMS OF BONE FRAGILITY Bone mass is related inversely to fracture risk, both current and future, but there are also qualitative abnormalities in bone that contribute to its fragility [81]. The best known and most well established of these relates to cancellous bone architecture. When cancellous bone is lost as a result of estrogen deficiency, whole structural elements are removed, leaving those that remain more widely separated and less well connected [68]. As a result, vertebral fracture
CHAPTER 15 Skeletal Heterogeneity and the Purposes of Bone Remodeling
risk is increased to a greater extent than would be expected for the reduction in bone mass [82]. This is probably why the presence of at least one vertebral fracture is an independent risk factor for further fractures [83]. The structural changes are the result of perforation of trabecular plates because the cutting hemicones of individual BMUs penetrate more deeply into the bone away from the surface [12,84]. This qualitative abnormality in osteoclast function appears now to be more likely due to delayed osteoclast apoptosis [85] than to excessively rapid resorption by individual osteoclasts [72]. However, a more fundamental problem may be loss of BMU directional control [86]. These various changes were attributed by Frost to resetting of the mechanostat [76], but estrogen deficiency leads for a few years to a disproportionately rapid loss of central cancellous bone adjacent to hematopoietic marrow. Furthermore, the occurrence of severe vertebral osteopenia in elite athletes with exercise-associated amenorrhea [87] indicates that the effects of estrogen deficiency are not prevented by increased physical activity. The second, more controversial, qualitative factor in bone fragility is the accumulation of fatigue microdamage. Frost [88] has proposed that normally there is such a wide margin of safety that the adverse effect of bone loss on bone fragility is mediated, not by a reduction in instantaneous breaking strength, but by fatigue damage accumulation due to increased strain in the bone that remains. However, most investigators believe that the margin of safety is not as great as Frost has claimed [89]. Frost has further proposed that a defective damage repair mechanism could be overwhelmed by even normal damage production. Increased bone age would increase susceptibility to fatigue damage, both directly (by exceeding the fatigue life) and indirectly (Fig. 3). Osteocyte death, which can occur spontaneously when bone age exceeds about 20 years [90], would impair the detection of fatigue damage, and the consequent perilacunar hypermineralization (or micropetrosis) would make the bone more brittle and more susceptible to fatigue damage [91]. The repair of microdamage by a new BMU could be delayed by an age-related decline in any of the intervening steps outlined previously or by a loss of the directional control needed for the new BMU to find its target [86], another possible consequence of osteocyte death [54]. Excessive bone age and its adverse consequences are most likely to occur in peripheral cortical bone because of its low rate of turnover, but even in central cancellous bone, turnover may be sufficiently low in some subjects that a significant fraction of interstitial bone may be older than 20 years, and this fraction is much higher in patients with vertebral fracture [92]. On the basis of this reasoning, it has been proposed that the subnormal turnover of central cancellous bone, a common finding in patients with vertebral fracture [72], was of pathogenetic significance, via increases in bone age and
441
FIGURE 3
Mechanisms whereby increased bone age could lead to accumulation of fatigue damage. Some effects would increase fatigue damage production, and some effects would decrease its detection and repair. Reproduced with permission from Parfitt [100].
susceptibility to fatigue damage [92]. At the time the previous version of this chapter was written [93], no evidence in support of this proposal had appeared and at an international meeting on osteoporosis held in 1993, it was withdrawn [44]. However, recent findings, although by no means establishing the hypothesis, suggest that this act of apostasy was premature. Osteocyte death, which increases in prevalence with subject age in the upper femur [94], appears not to change with age in the vertebral body when autopsy specimens are studied by respiratory enzyme histochemistry [95], but does increase with age in biopsy specimens of iliac cancellous bone studied by fluorescence and confocal microscopy [96]. The microcracks that occur in peripheral cortical bone have now also been observed in central cancellous bone, as mentioned previously [38], although it is not known whether they are more common in patients with vertebral fracture. It remains true that vertebral cancellous microfractures have not been shown to result from fatigue, even though they are frequently referred to as “fatigue fractures.” In the vertebral body, the perforations and loss of structural elements mentioned previously occur preferentially in horizontal rather than in vertical trabeculae. The compressive strength of a vertical trabecula will decline in proportion to the square of the unsupported length so that a 50% reduction in the number of horizontal trabeculae will lead to a fourfold increase in the susceptibility to buckling [44]. Based on estimates of in vivo stresses during normal activity [97] and on the production of microcracks by experimental compression [42], vertebral microfractures can be explained by instantaneous overload as a result of the architectural changes mentioned previously without the need to invoke a fatigue-based mechanism [44, 97,98]. Lower than normal turnover in central cancellous bone may not have the adverse effects on bone fragility that were
442 predicted because the turnover of cancellous bone adjacent to hematopoietic marrow is much higher than is needed to maintain its mechanical competence, which had not yet been deduced when the hypothesis was first proposed. In patients with vertebral fracture, central cancellous bone turnover may be much lower than in healthy subjects, but still high enough to prevent fatigue damage accumulation. In patients with osteoporotic vertebral fracture, the mineral density of iliac bone is decreased rather than increased [99,100]. This suggests that in some patients with osteoporotic vertebral fracture there is a substantial delay in secondary mineralization, the process whereby mineral crystals enlarge at the expense of water [55], which would remove much of the need for bone remodeling to prevent hypermineralization. Hypomineralization would be expected to reduce the strength of bone as a material [33] and to be an independent risk factor for bone fragility so that it is important to discover its pathogenesis. However, even if lower turnover of central cancellous bone is a consequence of hypomineralization rather than a cause of hypermineralization, there would still be accumulation of very old bone [92], which would be expected to exaggerate the age-related increase in osteocyte death and to compromise microdamage repair. Furthermore, the ability to direct a new BMU to regions of microdamage in interstitial bone could be impaired by poorly understood physical and cellular changes. If so, then the relationship between bone fragility and bone turnover is U shaped, as high bone turnover is a risk factor, because each remodeling site constitutes both a region of focal weakness and a stress concentrator [44,101]. Hip fractures share with vertebral fractures the inverse relationship of risk to bone mass, but differ from vertebral fractures with respect to the qualitative contribution to bone fragility. Loss of cancellous bone connectivity due to estrogen deficiency is less important, whereas fatigue damage accumulation is more important. Although small islands of hematopoietic tissue can persist in the upper femur much longer than at more distal sites, particularly in the femoral head, the proportion of red marrow is much lower than in the ilium [17]. There are no tetracycline-based measurements of bone remodeling in the upper femur, but other indices of bone remodeling are lower than in the ilium or vertebral body [26], and this difference is exaggerated in patients with hip fracture [17]. The proportion of osteocytes that are viable declines progressively with increasing subject age in the femoral neck [94]. True bone mineral density increases with age in the femoral shaft cortex [102], but not in the spine. Fatigue microdamage occurs in the cortical bone of both the femoral neck [103] and the femoral shaft, and in the latter, crack density increases exponentially with age, more so in women than in men [104]. Cancellous microfractures in the femoral head increase in number with age and with reduction in mineral density [105,106] and are significantly more frequent in hip fracture patients than in
A. M. PARFITT
TABLE 4
Fracture Pathogenesis at Different Sites Vertebra
Femoral neck
Function of cancellous bone
Metabolic
Mechanical
Marrow/turnover
Red/high
Yellow/low
Osteocyte death
Yes
Yes
Increase with age
Small
Large
Fatigue damage
?a
Yesb
Hypermineralization
No
Yesb
Main qualitative factor
Architecture
Bone age
a
Microdamage, not shown to be due to fatigue. In femoral cortical bone, not necessarily at fracture site.
b
controls, despite a statement by the authors to the contrary [107]; because of the lower bone turnover and differences in architecture, there is greater reason to invoke a fatiguebased mechanism than in the spine [100,105]. All these data indicate that increased bone age and its adverse effects on bone fragility (Fig. 3) are likely to be of major importance in the pathogenesis of hip fracture [100] (Table 4).
B. Prevention of Fractures It is customary to discuss the “prevention” and “treatment” of osteoporosis separately, but this is a misleading distinction, as the only therapeutic goal is to prevent fractures; whether one’s aim is to prevent the first fracture or a subsequent fracture does not alter this principle. Of the several aspects of fracture prevention, the theme of this chapter relates most clearly to the prevention and restoration of bone loss. Agents that accomplish these aims are usually referred to respectively as “inhibitors of bone resorption” and “stimulators of bone formation,” but these vague terms betray a serious lack of comprehension of bone remodeling.They ignore the indivisible unity of the BMU as a structural and functional entity, obscure the crucial distinction between effects on cell recruitment and effects on differentiated cell function, and engender the absurd notions that all bone resorption is bad and all bone formation is good. The former error is potentially more dangerous than the latter, so this aspect of therapy will be the focus of subsequent discussion. A long-term reduction in the rate of bone loss can be accomplished by a long-term reduction in activation frequency and a consequent reduction in bone turnover. How is this possible without frustrating the purposes of bone remodeling? Activation frequency is the best histologic index of the intensity of bone remodeling on a surface and is the main determinant of the rate of bone turnover, but it is not a measure of the frequency of BMU origination because it
CHAPTER 15 Skeletal Heterogeneity and the Purposes of Bone Remodeling
FIGURE 4 Relationship between BMU origination and remodeling activation. Activation frequency represents the product of frequency of BMU origination and the average distance of BMU progression. In this example, activation frequency would be the same with one BMU that progresses for 9 units distance, two BMUs that each progress for 4.5 units of distance, or three BMUs that each progress for 3 units of distance. However, the biological significance would be different because each BMU represents a separate remodeling project. Copyright 1995, A. M. Parfitt, used with permission.
also depends on the mean distance of BMU progression [4,108 – 110]. The effects on all histologic, biochemical, and radiokinetic indices of bone turnover would be the same whether, for example, one BMU traveled for nine units of distance through or across the surface of bone or each of three BMUs traveled for three units of distance (Fig. 4). The biological significance would be different, however, because each new BMU represents a separate remodeling project. Approximately 90% of new mononuclear osteoclast precursor cells are used to sustain the progression of existing BMUs, and only 10% are used to originate a new BMU [4]. Consequently, substantial changes in activation frequency and bone turnover can be brought about by manipulating the distance and duration of BMU progression without changing the frequency of BMU origination. Each episode of targeted remodeling requires a new BMU, but stochastic remodeling could be accomplished if each BMU progressed for a variable distance beyond its target [4]. This arrangement makes it possible for therapeutic agents to reduce activation frequency and bone turnover by curtailing BMU progression, without inhibiting BMU origination, and so to reduce stochastic remodeling without interfering with targeted remodeling. Obviously, the ability to prioritize different remodeling tasks is a feature of the remodeling system itself, not of the individual therapeutic agents. It must be assumed that the signals for osteoclast precursors to arrive at a particular location are more compelling for BMU origination than for BMU progression, more compelling for BMU progression toward its target than beyond its target, and more compelling for
443
the peripheral than for the central skeleton because of the difference in margin of safety. These hierarchies could reflect differences in the types as well as the amounts of signal molecules. However, therapeutic agents may differ in their ability to exploit these differences in signal strength. Agents that act directly on osteoclasts to reduce their resorptive activity are more likely to act indiscriminately on all osteoclasts throughout the skeleton, and in some locations this is likely to negate their purpose; consequently, the net outcome of the intervention could be harmful rather than beneficial. However, agents that reduce the supply of osteoclast precursor cells leave the remodeling system able to deploy its more limited resources to the best advantage. Not surprisingly, hormone replacement therapy (HRT) is the most effective means of preventing the adverse effects on bone of the hormone deficiency that results from menopausal ovarian failure. Estrogen deficiency increases the availability of osteoclast precursor cells [6] and so increases the stochastic component of bone remodeling by removing a constraint on BMU progression, particularly in the central skeleton. However, the most destructive consequence of estrogen deficiency is delayed osteoclast apoptosis [4,85], leading to deeper BMU penetration (reflected in two-dimensional histologic sections as increased resorption depth), trabecular plate perforation, and loss of connectivity. Both of these effects — increased osteoclast recruitment and delayed osteoclast apoptosis — are prevented by HRT, and ideally both of them should be prevented by any agent that is used as a substitute for HRT. Until recently, the most widely used substitute was calcitonin, but it has not been shown to promote earlier osteoclast apoptosis [111], and its effects on resorption depth are uncertain. Furthermore, its long-term use can be complicated by secondary hyperparathyroidism of sufficient severity that bone turnover may be increased rather than decreased [112]. The newer bisphosphonates appear to be more complete substitutes for HRT. Although their best known effect is to acutely inhibit the function of existing osteoclasts, in the long term, they reduce osteoclast recruitment by mechanisms that remain uncertain [4,113], promote earlier osteoclast apoptosis [114], and reduce resorption depth [115]. The safety of reducing bone turnover depends on the ability to limit stochastic remodeling preferentially in the central skeleton without interfering with targeted remodeling at any skeletal site. Obviously, there is a lower limit to osteoclast precursor cell recruitment below which the purposes of remodeling will be frustrated. As would be predicted from the earlier discussion, complete suppression of remodeling in beagles leads to the occurrence of spontaneous fractures after a few months [116]; this occurred with etidronate, which causes osteomalacia, but also with clodronate, which does not. A dangerous reduction in bone turnover could never occur with physiological agents such
444 as estrogen or calcitonin, but can be produced readily by bisphosphonates if given in excessive dose. Regrettably, there is very little information on what lower limit is safe. The safe level will be different in different regions of the skeleton, which reduces the value of biochemical indices of turnover to determine safety, as these are necessarily blind to regional differences. Quite low levels of whole body bone turnover are consistent with skeletal health when they occur naturally, but might conceal regional ill health when induced by therapeutic intervention. Reducing osteoclast recruitment to a level just sufficient to allow the completion of targeted remodeling but leaving no room for stochastic remodeling would also lead eventually to spontaneous fractures, but the time required would probably be measured in years rather than in months, which limits the use of animal models to determine long-term safety. As explained earlier, when vertical trabeculae have lost their horizontal supports, even normal remodeling may constitute a mechanical threat. In this situation, reducing turnover even within the normal range (defined by biochemical indices) may be useful in the prevention of vertebral fractures [44]. In a study of transdermal estrogen therapy, reduced turnover contributed independently to reduced vertebral fracture occurrence, in addition to the effect of increased bone mass [117], but the long-term effects of reducing turnover on hip fracture risk are less easily predictable. For reasons given previously, the adverse effects of prolonged bone age on bone fragility (Fig. 3) are likely to be more serious in the upper femur than in the spine (Table 4). The large increase in the use of bisphosphonates that followed the approval of alendronate by the FDA should reduce the incidence of vertebral fractures [118] and, for a few years, the incidence of other fractures, but data available today do not exclude the possibility that 10 – 20 years later there will be an epidemic of hip fractures. By then, a large proportion of the elderly population will have levels of bisphosphonate of one kind or another in the femoral heads and necks that are possibly dangerous, and there will be nothing that can be done about it. Two years of treatment with risedronate did not increase microdamage in the canine femoral neck [119], but a higher effective dose of alendronate did increase microdamage in the canine rib [120]. It is not too late to find out what is really going on in the bones of hip fracture patients, but only if we abandon the exclusive reliance on biochemical and densitometric methods and on histologic examination at a site chosen for its convenience rather than its relevance to the problem of greatest importance. An important recent discovery from direct examination is that clustered remodeling and giant resorption cavities due to confluence of clusters are more common in hip fracture patients than in age-matched control subjects [121]. A reduction in the frequency of such clusters could account for the early reduction in fracture risk by correction of vitamin D deficiency
A. M. PARFITT
[122] and also for the beneficial effect of bisphosphonate administration. However, if the clusters are an effect of defective directional control of BMUs, then reducing their number could compromise microdamage repair. Obviously a great deal more research will be needed to resolve these uncertainties.
References 1. A. M. Parfitt, The two faces of growth-Benefits and risks to bone integrity. Osteopor. Int. 4, 382 – 398 (1994). 2. H. M. Frost, “Intermediary Organization of the Skeleton.” CRC Press, Boca Raton, FL, 1986. 3. A. M. Parfitt, Osteonal and hemi-osteonal remodeling: The spatial and temporal framework for signal traffic in adult human bone. J. Cell. Biochem. 55, 273 – 286 (1994). 4. A. M. Parfitt, G. R. Mundy, G. D. Roodman, D. E. Hughes, and B. Boyce, A new model for the regulation of bone resorption. J. Bone Miner. Res. 11, 150 – 159 (1996). 5. A. M. Parfitt. Min: Review — Osteoclast precursors as leucocytes: Importance of the area code. Bone 23, 491 – 494 (1998). 6. S. C. Manolagas, Cell number versus cell vigor: What really matters to a regenerating skeleton? Endocrinology 140, 4377 – 4381. (1999). 7. A. M. Parfitt, Problems in the application of in vitro systems to the study of human bone remodeling. Proceedings from the Workshop on Human Models of Skeletal Aging, NIH. Calcif. Tissue Int. 56 (Suppl. 1), S5 – S7 (1995). 8. M. C. Moore-Ede, Physiology of the circadian timing system: Predictive versus reactive homeostasis. Am. J. Physiol. 250, R735 – R752 (1986). 9. J. Foldes, A. M. Parfitt, M.-S. Shih, D. S. Rao, and M. Kleerekoper, Structural and geometric changes in iliac bone: Relationship to normal aging and osteoporosis. J. Bone Miner. Res. 6, 759 – 766 (1991). 10. D. P. Fyhrie, S. M. Lang, and A. M. Parfitt, “Cortical and Cancellous Bone Structure in the Ilia of Normals and Osteoporotics,” p. 443. Orthopaedic Research Society, 40th Annual Meeting, 1994. [Abstract] 11. G. H. Meyer, Die Architectur der spongiosa. Arch. Anat. Physiol. Wissensch. 34, 615 – 628 (1867). 12. A. M. Parfitt, The physiologic and pathogenetic significance of bone histomorphometric data. In “Disorders of Bone and Mineral Metabolism” (F. L. Coe and M. J. Favus, eds.), 2nd Ed. Lippincott, Williams, and Wilkins, Philadelphia, in press. 13. Z.-H. Han, S. Palnitkar, D. S. Rao, D. Nelson, and A. M. Parfitt, Effect of ethnicity and age or menopause on the structure and geometry of iliac bone. J. Bone Miner. Res. 11, 1967 – 1975 (1996). 14. A. M. Parfitt, Surface specific bone remodeling in health and disease. In “Clinical Disorders of Bone and Mineral Metabolism” (M. Kleerekoper and S. Krane, eds.), pp. 7 – 14. Mary Ann Liebert, New York, 1989. 15. Z.-H. Han, S. Palnitkar, D. Rao, Sudhaker, D. Nelson, and A. M. Parfitt, Effects of ethnicity and age or menopause on the remodeling and turnover of iliac bone: Implications for mechanisms of bone loss. J. Bone Miner. Res. 12, 498 – 508 (1997). 16. D. W. Dempster, Relationship between the iliac crest bone biopsy and other skeletal sites. In “Clinical Disorders of Bone and Mineral Metabolism” (M. N. Kleerekoper and S. M. Krane, eds.), pp. 247 – 252. Mary Ann Liebert, New York, 1989. 17. I. Eventov, B. Frisch, Z. Cohen, and I. Hammel, Osteopenia, hematopoiesis, and bone remodelling in iliac crest and femoral biopsies: A prospective study of 102 cases of femoral neck fractures. Bone 12, 1 – 6 (1991).
CHAPTER 15 Skeletal Heterogeneity and the Purposes of Bone Remodeling 18. A. M. Parfitt, The physiologic and clinical significance of bone histomorphometric data. In “Bone Histomorphometry: Techniques and Interpretations” (R. Recker, ed.), pp. 142 – 223. CRC Press, Boca Raton, FL, 1983. 19. A. M. Parfitt, D. S. Rao, J. Stanciu, A. R. Villenueva, M. Kleerekoper, and B. Frame, Irreversible bone loss in osteomalacia: Comparison of radial photon absorptiometry with iliac bone histomorphometry during treatment. J. Clin. Invest. 76, 2403 – 2412 (1985). 20. D. B. Kimmel and W. S. S. Jee, A quantitative histologic study of bone turnover in young adult beagles. Anat. Rec. 203, 31 – 45 (1982). 21. A. J. Erslev, Medullary and extramedullary blood formation. Clin. Orthop. 52, 25 – 36 (1967). 22. S. N. Wickramasinghe, “Human Bone Marrow.” Blackwell Scientific, Oxford, 1975. 23. R. P. Custer, “An Atlas of the Blood and Bone Marrow.” Saunders, Philadelphia, 1949. 24. A. Kessinger, Circulating stem cells: Waxing hematopoietic. N. Engl. J. Med. 333(5), 315 – 316 (1995). 25. D. Van Dyke, Similarity in distribution of skeletal blood flow and erythropoietic marrow. Clin. Orthop. 52, 37 – 51 (1967). 26. B. Krempien, F. M. Lemminger, E. Ritz, and E. Weber, The reaction of different skeletal sites to metabolic bone disease: A micromorphometric study. Klin. Wochenschr. 56, 755 – 759 (1978). 27. T. J. Wronski, J. M. Smith, and W. S. S. Jee, The microdistribution and retention of injected 239Pu on trabecular bone surfaces of the beagle: Implications for the induction of osteosarcoma. Radiat. Res. 83, 74 – 89 (1980). 28. T. J. Wronski, J. M. Smith, and W. S. S. Jee, Variations in mineral apposition rate of trabecular bone within the beagle skeleton. Calcif. Tissue Int. 33, 583 – 586 (1981). 29. F. Melsen, L. Mosekilde, and E. F. Eriksen, Spatial distribution of sinusoids in relation to remodeling sites: Evidence for specialized sinusoidal structures associated with formative sites. J. Bone Miner. Res. 10(Suppl.), S209 (1995). 30. R. B. Martin, B. D. Chow, and P. A. Lucas, Bone marrow fat content in relation to bone remodeling and serum chemistry in intact and ovariectomized dogs. Calcif. Tissue Int. 46, 189 – 194 (1990). 31. J. Podenphant and U. Engel, Regional variations in histomorphometric bone dynamics from the skeleton of an osteoporotic woman. Calcif. Tissue Int. 40, 184 – 188 (1987). 32. H. Duncan, A. R. Villanueva, C. H. Mathews, J. C. Leisen, and A. M. Parfitt, Dynamics of bone destruction and repair in rheumatoid joints. Arthritis Rheum. 24, S97 (1981). 33. J. Currey, “The Mechanical Adaptations of Bones.” Princeton Univ. Press, Princeton, NJ, 1984. 34. J. J. Tomasek, S. W. Meyers, J. B. Basinger, D. T. Green, and R. L. Shew, Diabetic and age-related enhancement of collagen-linked fluorescence in cortical bones of rats. Life Sci. 55(11), 855 – 861 (1994). 35. D. B. Burr, M. R. Forwood, D. P. Fyhrie, R. B. Martin, M. B. Schaffler, and C. H. Turner, Bone microdamage and skeletal fragility in osteoporotic and stress fractures. J. Bone Miner. Res. 12, 6 – 15 (1997). 36. V. Bentolila, T. M. Boyce, D. P. Fyhrie, R. Drumb, T. M. Skerry, and M. B. Schaffler, Intracortical remodeling in adult rat long bones after fatigue loading. Bone 23, 275 – 281 (1998). 37. S. Mori and D. B. Burr, Increased intracortical remodeling following fatigue damage. Bone 14, 103 – 109 (1993). 38. T. E. Wenzel, M. B. Schaffler, and D. P. Fyhrie, In vivo trabecular microcracks in human vertebral bone. Bone 19(2), 89 – 95 (1996). 39. S. Mori, R. Harruff, W. Ambrosius, and D. B. Burr, Trabecular bone volume and microdamage accumulation in the femoral heads of women with and without femoral neck fractures. Bone 2 1, 521 – 526 (1997).
445
40. N. H. Fazzalari, M. R. Forwood, B. A. Manthey, K. Smith, and P. Kolesik, Three-dimensional confocal images of microdamage in cancellous bone. Bone 23, 373 – 378 (1998). 41. M. B. Schaffler, K. Choi, and C. Milgrom, Aging and matrix microdamage accumulation in human compact bone. Bone 17, 521 – 525 (1995). 42. D. P. Fyhrie and M. B. Schaffler, Failure mechanisms in human vertebral cancellous bone. Bone 15, 105 – 109 (1994). 43. E. P. M. Urovitz, V. L. Fornasier, M. I. Risen, and I. MacNab, Etiological factors in the pathogenesis of femoral trabecular fatigue fractures. Clin. Orthop. 127, 275 – 280 (1977). 44. A. M. Parfitt, Pathophysiology of bone fragility. In “Proceedings of the Fourth International Symposium on Osteoporosis” (C. Christiansen and B. Riis, eds.), pp. 164 – 166. Hong Kong, Handelstrykkeriet Aalborg ApS, Aalborg, Denmark, 1993. 45. G. Marotti, Map of bone formation rate values recorded throughout the skeleton of the dog. In “Bone Morphometry: Proceedings of the First Workshop” (Z. F. G. Jaworski, ed.), pp. 202 – 207. Univ. of Ottawa Press, 1976. 46. D. Dempster, Bone remodeling. In “Disorders of Bone and Mineral Metabolism” (F. Coe and M. Z. Favus, eds.), pp. 355 – 380. Raven Press, New York, 1992. 47. Z. F. G. Jaworski, Haversian systems and Haversian bone. In “Bone — A Treatise: Bone Metabolism and Mineralization” (B. K. Hall, ed.), Vol. 4, pp. 21 – 45. CRC Press, Boca Raton, FL, 1992. 48. R. B. Martin and D. B. Burr, “Structure, Function, and Adaptation of Compact Bone.” Raven Press, New York, 1989. 49. C. Morris, “Academic Press Dictionary of Science and Terminology.” Academic Press, San Diego, 1992. 50. E. F. Eriksen, T. Steiniche, L. Mosekilde, and F. Melsen, Histomorphometric analysis of bone in metabolic bone disease. Endocrinol. Metabol. Clin. North Am. 18, 919 – 954 (1989). 51. A. M. Parfitt, Idiopathic, surgical and other varieties of parathyroid hormone deficient hypoparathyroidism. In “Endocrinology” (L. DeGroot, ed.), 2nd Ed., pp. 1049 – 1064. Saunders, Philadelphia, 1989. 52. A. M. Parfitt, BMU Origination and Progression: Relationship to Targeted and Nontargeted Remodeling. Abstract for VIIIth Congress of the International Society of Bone Morphometry, 1999. 53. C. H. Turner and M. R. Forwood, What role does the osteocyte network play in bone adaptation? Bone 16, 283 – 285 (1995). 54. O. Verborgt, G. J. Gibson, and M. B. Schaffler, Loss of osteocyte integrity in association with microdamage and bone remodeling after fatigue in vivo. J. Bone Min. Res. 15, 60 – 67 (2000). 55. A. M. Parfitt, Calcium homeostasis. In “Handbook of Experimental Pharmacology: Physiology and Pharmacology of Bone” (G. R. Mundy and T. J. Martin, eds.), Vol. 107, pp. 1 – 65. Springer-Verlag, Heidelberg, 1993. 56. A. Deldar, H. Lewis, and L. Weiss, Bone lining cells and hematopoiesis: An electron microscopic study of canine bone marrow. Anat. Rec. 213, 187 – 201 (1985). 57. A. M. Parfitt, Equilibrium and disequilibrium hypercalcemia: New light on an old concept. Metab. Bone Dis. Rel. Res. 1, 279 – 293 (1979). 58. A. Blumsohn, K. Herrington, R. A. Hannon, P. Shao, D. R. Eyre, and R. Eastell, The effect of calcium supplementation on the circadian rhythm of bone resorption. J. Clin. Endocrinol. Metab. 79, 730 – 735 (1994). 59. A. M. Parfitt, Morphologic basis of bone mineral measurements: Transient and steady state effects of treatment in osteoporosis. Miner. Electrolyte Metab. 4, 273 – 287 (1980). 60. A. M. Parfitt, Integration of skeletal and mineral homeostasis. In “Osteoporosis: Recent Advances in Pathogenesis and Treatment” (H. F. DeLuca, H. Frost, W. Jee, C. Johnston, and A. M. Parfitt, eds.), pp. 115 – 126. University Park Press, Baltimore, MD, 1981.
446 61. A. M. Parfitt, Definition of osteoporosis: Age-related loss of bone and its relationship to increased fracture risk. In “Osteoporosis.” NIH Consensus Development Conference, 1984. 62. M. Kleerekoper and D. Nelson, Osteoporosis as a community health problem: Lessons learned from studying hypertension. HFH Med. J. 36, 113 – 116 (1988). 63. S. M. Garn, “The Earlier Gain and the Later Loss of Cortical Bone.” Thomas, Springfield, IL, 1970. 64. A. M. Parfitt, Interpretation of bone densitometry measurements: The disadvantages of a percentage scale and a discussion of available alternatives. J. Bone Miner. Res. 5, 537 – 540 (1990). 65. A. M. Parfitt, Is the rate of bone loss influenced by the initial value? Biological and statistical issues. Abstract, Perth International Bone Meeting, Fremantle, Australia. Osteopor. Int. 5, 309 – 310 (1995). 66. B. E. C. Nordin, A. G. Need, R. L. Prince, M. Horowitz, D. H. Gutteridge, and S. E. Papapoulos, Osteoporosis. In “Metabolic Bone and Stone Disease” (B. E. C. Nordin, A. G. Need, and H. A. Morris, eds.), Churchill Livingstone, London, 1993. 67. C. W. Slemenda, Adult bone loss. In “Osteoporosis” (R. Marcus, ed.), Blackwell Scientific, London, 1994. 68. A. M. Parfitt, C. H. E. Mathews, A. R. Villanueva, M. Kleerekoper, B. Frame, and D. S. Rao, Relationship between surface, volume and thickness of iliac trabecular bone in aging and in osteoporosis: Implications for the microanatomic and cellular mechanism of bone loss. J. Clin. Invest. 72, 1396 – 1409 (1983). 69. P. J. Meunier, Assessment of bone turnover by histomorphometry in osteoporosis. In “Osteoporosis: Etiology, Diagnosis, and Management” (B. L. Riggs and L. J. Melton, eds.), Raven Press, New York, 1988. 70. J. W. Davis, J. S. Grove, P. D. Ross, J. M. Vogel, and R. D. Wasnich, Relationship between bone mass and rates of bone change at appendicular measurement sites. J. Bone Miner. Res. 7, 719 – 725 (1992). 71. K. M. Davies, R. R. Recker, M. R. Stegman, and R. P. Heaney, Tallness versus shrinkage: Do women shrink with age or grow taller with recent birth date? J. Bone Miner. Res. 6, 1115, 1120 (1991). 72. A. M. Parfitt, Bone remodeling: Relationship to the amount and structure of bone, and the pathogenesis and prevention of fractures. In “Osteoporosis: Etiology, Diagnosis, and Management” (B. L. Riggs and L. J. Melton, eds.), Raven Press, London, 1988. 73. B. J. Pedersen, A. Schlemmer, C. Hassager, and C. Christiansen, Changes in the carboxyl-terminal propeptide of type I procollagen and other markers of bone formation upon five days of bed rest. Bone 17, 91 – 95 (1995). 74. C. Christiansen, Postmenopausal bone loss and the risk of osteoporosis. Osteopor. Int.(Suppl. 1), S47 – S51 (1994). 75. K. Wu, S. Jett, and H. M. Frost, Bone resorption rates in rib in physiological, senile, and postmenopausal osteoporoses. J. Lab. Clin. Med. 69, 810 – 818 (1967). 76. H. M. Frost, The mechanostat: A proposed pathogenetic mechanism of osteoporoses and the bone mass effects of mechanical and nonmechanical agents. Bone Miner. 2, 73 – 85 (1987). 77. L. E. Lanyon, A. E. Goodship, C. J. Pye, and J. H. MacFie, Mechanically adaptive bone remodeling. J. Biomech. 15, 141, 154 (1982). 78. H. M. Frost, Structural adaptations to mechanical usage (SATMU). Anat. Rec. 226, 414 – 422 (1990). 79. Z. F. G. Jaworski and H. K. Uhthoff, Disuse osteoporosis: Current status and problems. In “Current Concepts of Bone Fragility” (H. K. Uhthoff, ed.), p. 181. Springer-Verlag, Berlin, 1986. 80. H. M. Frost, On our age-related bone loss: Insights from a new paradigm. J. Bone Miner. Res. 12, 1539 – 1546 (1997). 81. R. P. Heaney, Is there a role for bone quality in fragility fractures? Calcif. Tissue Int. 53 (Suppl. 1), S3 – S6 (1993). 82. M. Kleerekoper, A. R. Villanueva, J. Stanciu, D. S. Rao, and A. M. Parfitt, The role of three dimensional trabecular microstructure in the pathogenesis of vertebral compression fractures. Calcif. Tissue Int. 37, 594 – 597 (1985).
A. M. PARFITT 83. P. D. Ross, J. W. Davis, R. S. Epstein, and R. D. Wasnich, Preexisting fractures and bone mass predict vertebral fracture incidence in women. Ann. Intern. Med. 114, 919 – 923 (1991). 84. EF Eriksen, B. Langdahl, A. Vesterby, J. Rungby, and M. Kassem, Hormone replacement therapy prevents osteoclastic hyperactivity: A histomorphometric study in early postmenopausal women. J. Bone Miner. Res. 14, 1217 – 1221 (1999). 85. D. E. Hughes, A. Dai, J. C. Tiffee, H. H. Li, G.R. Mundy, and B. F. Boyce, Estrogen promotes apoptosis of murine osteoclasts mediated by TGF. Nature Med. 2, 1132 – 1135 (1996). 86. A. M. Parfitt, Abnormal structure and fragility of bone as expressions of disordered remodeling. Abstract XXIV, European Symposium on Calcified Tissues, Aarhus, Calcif. Tissue Int. 56, 423 (1995). 87. B. L. Drinkwater, K. Nilson, C. H. Chesnut, W. J. Bremner, S. Shainholtz, and M. B. Southworth, Bone mineral content of amenorrheic and eumenorrheic athletes. N. Engl. J. Med. 311, 277 – 281 (1984). 88. H. M. Frost, The pathomechanics of osteoporosis. Clin. Orthop. 200, 198 – 225 (1985). 89. R. McN. Alexander, Optimum strengths for bones liable to fatigue and accidental fracture. J. Theor. Biol. 109, 621 – 636 (1984). 90. H. M. Frost, In vivo osteocyte death. J. Bone Jt. Surg. 42A, 138 – 143 (1960). 91. H. M. Frost, Micropetrosis. J. Bone Jt. Surg. 42A, 144 – 150 (1960). 92. A. M. Parfitt, M. Kleerekoper, and A. R. Villanueva, Increased bone age: Mechanisms and consequences. In “Osteoporosis” (C. Christiansen, C. Johansen, and B. J. Riis, eds.), pp. 301 – 308. Osteopress ApS, Copenhagen, 1987. 93. A. M. Parfitt, Skeletal heterogeneity and the purposes of bone remodeling: Implications for the understanding of osteoporosis. In “Osteoporosis” (R. Marcus, D. Feldman and J. Kelsey, eds.), pp. 315 – 329. Academic Press, San Diego, 1996. 94. S. Y. P. Wong, J. Kariks, R. A. Evans, C. R. Dunstan, and E. Hills, The effect of age on bone composition and viability in the femoral head. J. Bone Jt. Surg. 67A, 274 – 283 (1985). 95. C. R. Dunstan, N. M. Somers, and R. A. Evans, Osteocyte death and hip fracture. Calcif. Tissue Int. 53(Suppl. 1), S113 – S117 (1993). 96. S. J. Qui, S. Palnitkar, and D. S. Rao, Age-related changes in osteocyte density and distribution in human cancellous bone. J. Bone Miner. Res. 14(Suppl. 1), S308 (1999). 97. B. D. Snyder, S. Piazza, W. T. Edwards, and W. C. Hayes, Role of trabecular morphology in the etiology of age-related vertebral fractures. Calcif. Tissue Int. 53 (Suppl. 1), S14 – S22 (1993). 98. B. Vernon-Roberts and C. J. Pirie, Healing trabecular microfractures in the bodies of lumbar vertebrae. Ann. Rheum. Dis. 32, 406 – 412 (1973). 99. J. M. Burnell, D. J. Baylink, C. H. Chesnut III, M. W. Mathews, and E. Teubner, Bone matrix and mineral abnormalities in postmenopausal osteoporosis. Metabolism 31, 1113 – 1120 (1982). 100. A. M. Parfitt, Bone age, mineral density and fatigue damage. Calcif. Tissue Int. 53 (Suppl. 1), S82 – S86 (1993). 101. T. A. Einhorn, Bone strength: The bottom line. Calcif. Tissue Int. 51, 333 – 339 (1992). 102. M. Grynpas, Age and disease-related changes in the mineral of bone. Calcif. Tissue Int. 53 (Suppl. 1), S57 – S64 (1993). 103. M. B. Schaffler, T. M. Boyce, K. D. Lundin-Cannon, C. Milgrom, and D. P. Fyhrie, “Age-Related Architectural Changes and Microdamage Accumulation in the Human Femoral Neck Cortex,” p. 549. 41st Annual Meeting, Orthopaedic Research Society, 1995. [Abstract] 104. M. B. Schaffler, K. Choi, and C. Milgrom, Microcracks and aging in human femoral compact bone. Bone 17 (1995). 105. M. A. R. Freeman, R. C. Todd, and C. J. Pirie, The role of fatigue in the pathogenesis of senile femoral neck fractures. J. Bone Jt. Surg. 56B, 698 – 702 (1974).
CHAPTER 15 Skeletal Heterogeneity and the Purposes of Bone Remodeling 106. N. L. Fazzalari, Trabecular microfracture. Calcif. Tissue Int. 53(Suppl. 1), S143 – S147 (1993). 107. M. Wicks, R. Garrett, B. Vernon-Roberts, and N. Fazzalari, Absence of metabolic bone disease in the proximal femur in patients with fracture of the femoral neck. J. Bone Jt. Surg. 64B, 319 – 322 (1982). 108. A. M. Parfitt, The actions of parathyroid hormone on bone: Relation to bone remodeling and turnover, calcium homeostasis and metabolic bone disease. I. Mechanisms of calcium transfer between blood and bone and their cellular basis. Morphologic and kinetic approaches to bone turnover. Metabolism 25, 809 – 844 (1976). 109. R. B. Martin, On the histologic measurement of osteonal BMU activation frequency. Bone 15, 547 – 549 (1994). 110. C. J. Hernandez, S. J. Hazelwood, and R. B. Martin, The relationship between basic multicellular unit activation and origination in cancellous bone. Bone 25, 583 – 587 (1999). 111. D. E. Hughes, K. R. Wright, A. Sasaki, T. Yoneda, H. Uy, G. D. Roodman, G. R. Mundy, and B. F. Boyce, Bisphosphonates induce osteoclast apoptosis in vivo and in vitro, but calcitonin does not. J. Bone Miner. Res. 9 (Suppl. 1), S347 (1994). 112. R. Boyce, personal communication. 113. H. Fleisch, “Bisphosphonates in Bone Disease: From the Laboratory to the Patient,” 3rd Ed. Patterson Publishing Group/New York, 1997. 114. D. E. Hughes, K. R. Wright, H. L. Uy, S. Saski, T. Yoned, G. D. Roodman, G. R. Mundy, and B. F. Boyce, Bisphosphates promote apoptosis in murine osteoclasts in vitro and in vivo. J. Bone Miner. Res. 10, 1478 – 1487 (1995). 115. R. W. Boyce, C. L. Paddock, J. R. Gleason, W. K. Sletsema, and E. F. Eriksen, The effects of risedronate on canine cancellous bone remodeling: Three-dimensional kinetic reconstruction of the remodeling site. J. Bone Miner. Res. 10, 211 – 221 (1995).
447
116. L. Flora, G. S. Hassing, A. M. Parfitt, and A. R. Villanueva, Comparative skeletal effects of two diphosphonates in dogs. In “Bone Histomorphometry: Third International Workshop” (W. S. S. Jee and A. M. Parfitt, eds.), pp. 389 – 407. ArmourMontagu, Paris, 1981. 117. B. L. Riggs, L. J. Melton III, and W. M. O’Fallon, Postmenopausal osteoporosis: Evidence that antiresorptive and formation-stimulating regimens decrease vertebral fracture rate by independent mechanisms. In “Proceedings of the Fourth International Symposium on Osteoporosis” (C. Christiansen and B. Riis, eds.), Vol. 102, pp. 13 – 15. Hong Kong, Handelstrykkeriet Aalborg ApS, Aalborg, Denmark, 1993. 118. U. A. Liberman, S. R. Weiss, J. Broll et al., Effect of oral alendronate on bone mineral density and the incidence of fractures in postmenopausal osteoporosis. N. Engl. J. Med. 333, 1437 – 1443 (1995). 119. M. R. Forwood, D. B. Burr, Y. Takano, D. F. Eastman, P. N. Smith, and J. D. Schwardt, Risedronate treatment does not increase microdamage in the canine femoral neck. Bone 16, 643 – 650 (1995). 120. T. Mashiba, T. Hirano, C. H. Turner, M. R. Forwood, C. C. Johnston, and D. B. Burr, Suppressed bone turnover by bisphophonates increases microdamage accumulation and reduces some biomechanical properties in dog rib. J. Bone Miner. Res. 15, 613 – 620 (2000). 121. K. L. Bell, N. Loveridge, G. Jorda, J. Power, and J. Reeve, Merging of haversian canals within remodeling clusters; a cause of focal weakness in the femoral neck cortex? J. Bone Miner. Res. 14 (Suppl. 1), S265 (1999). 122. M. C. Chapuy, M. E. Arlot, P. D. Delmas, and P. J. Meunier, Effect of calcium and cholecalciferol treatment for three years on hip fractures in elderly women. Br. Med. J. 308, 1081 – 1082 (1994).
CHAPTER 16
Basic Biology of Bisphosphonates H. FLEISCH
University of Berne, CH-3008 Berne, Switzerland
I. Introduction II. Chemistry III. Actions
IV. Pharmacokinetics V. Animal Toxicology References
I. INTRODUCTION
P – C – P structure, the compounds are called geminal bisphosphonates. They are therefore analogues of pyrophosphate that contain a carbon instead of an oxygen atom. For the sake of simplicity, and because only P – C – P bisphosphonates have been found to exert strong activity on the skeleton, the geminal bisphosphonates will simply be called bisphosphonates in this review. This simplification is usually also made in the literature (Fig. 1). The geminal bisphosphonates have been known for a long time, the first synthesis by German chemists dating back to 1865 [7]. Etidronate was synthesized as early as 1897 [8]. They were used for a variety of industrial applications, among them as antiscaling agents [8]. The P – C – P structure allows a great number of possible variations, either by changing the two lateral chains on the carbon atom or by esterifying the phosphate groups. The first report on the biological action of bisphosphonates dates back to 1968/1999 [9– 11] Since then, many bisphosphonates have been investigated in animals and humans with respect to their effect on bone. Today, seven — alendronate, clodronate, etidronate, ibandronate, pamidronate, risedronate, and tiludronate — are available commercially in some countries for use in human bone disease (Fig. 2).
Bisphosphonates are a class of drugs developed in the past three decades for use in various diseases of bone and calcium metabolism. This chapter deals with the preclinical aspects of these compounds, with emphasis on those related to osteoporosis. The topics discussed are divided into the following sections: chemistry, effects on bone resorption and their mechanisms, effects on mineralization and their mechanisms, other effects, pharmacokinetics, and animal toxicology. Because the literature in this field is very large, references are generally restricted in the specific topic to the original ones and to those bringing new knowledge, as well as to newer reviews. For a more complete general update of the preclinical aspects of bisphosphonates, several newer reviews are available [1 – 6]. The clinical aspects of bisphosphonates will be covered in Chapter 72.
II. CHEMISTRY Bisphosphonates, formerly called diphosphonates, are compounds characterized by two C – P bonds. If the two bonds are located on the same carbon atom, resulting in a
OSTEOPOROSIS, SECOND EDITION VOLUME 1
449
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
450
H. FLEISCH
Each bisphosphonate has its own physicochemical and biological characteristics. This variability in effect makes it impossible to extrapolate with certainty from data for one compound to others so that each compound has to be considered on its own, with respect to both its use and its toxicology. The P – C – P bonds of bisphosphonates are stable to heat and most chemical reagents, and completely resistant to enzymatic hydrolysis, but can be hydrolyzed in solution by FIGURE 1
Chemical structure of pyrophosphate and bisphosphonates.
FIGURE 2 Chemical structure of bisphosphonates investigated for their effect on bone in humans. *Available commercially. From Fleisch [5].
451
CHAPTER 16 Basic Biology of Bisphosphonates
FIGURE 2
ultraviolet light. These compounds have a strong affinity for metal ions, among them calcium, with which they can form both soluble and insoluble complexes and aggregates, depending on the pH of the solution and the specific metal [12]. This can occur in vivo when large amounts are infused rapidly so great care must be taken when these compounds are given intravenously. Some uncertainty still exists as to the state of bisphosphonates when in solution.
III. ACTIONS A. Inhibition of Bone Resorption The main effect of pharmacologically active bisphosphonates is to inhibit bone resorption. Indeed, these compounds proved to be extremely powerful inhibitors of resorption when tested in a variety of conditions, both in vitro and in vivo.
(continued)
1. EFFECTS IN VITRO Bisphosphonates block bone resorption induced by various means in cell and organ culture. In the former, they inhibit the formation of pits by isolated osteoclasts cultured on mineralized substrata [13,14]. In organ culture they decrease the destruction of bone in embryonic long bones and in neonatal calvaria [15,16]. This inhibition is present whether resorption is stimulated or not. Up until now, the effect of all the stimulators of bone resorption, such as parathyroid hormone, 1,25(OH)2D, prostaglandins, and products of tumor cells, as well as others, has been inhibited by biphosphonates. In the past, the correlation between results obtained in calvaria in vitro and those obtained in vivo was rather poor. However, a more recent study performed with nine compounds varying in their activity by five to six orders of magnitude showed a satisfactory correlation using the 4- to 7-day-old mouse calvaria assay [17].
452
H. FLEISCH
2. EFFECTS IN VIVO a. Intact Animals In growing rats, bisphosphonates can block the degradation of both primary and secondary trabeculae, thus arresting the modeling and remodeling of metaphyses [18]. The latter therefore become club-shaped and radiologically more dense than normal, leading to a picture similar to that seen in congenital osteopetrotic animals. This effect, called the “Schenk assay,” is often used as an experimental assay to estimate the potency of new compounds [19] (Fig. 3). The inibition of bone resorption by bisphosphonates has also been documented using 45Ca kinetic studies and hydroxyproline excretion [20], as well as by other means. The effect occurs within 24 – 48 h [21] and is therefore slower than that of calcitonin. The decrease in resorption is accompanied, at least in the growing animal, by a positive calcium balance and an increase in the mineral content of bone and in bone mass
FIGURE 3
[20]. This is possible because of an increase in intestinal calcium absorption, consequent to an elevation of 1,25(OH)2D. The increase is, however, smaller than expected considering the dramatic decrease in bone resorption. This is due to the fact that, after a certain time, bone formation also decreases [20] because of the so-called coupling between formation and resorption characteristic of a decrease in remodeling. The main effect of bisphosphonates is therefore a reduction in bone turnover. It is not known how long the increase in balance lasts in the rat after discontinuation of the bisphosphonate. This increase is the basis for the administration of these compounds to prevent and treat osteoporosis in humans. Less is known about the effect in the normal adult animal. In dogs and minipigs, the long-term administration of alendronate did not lead to an increase in bone mass [22]. This might be explained by the physiological biomechanical homeostasis of bone structure, which would eliminate a
Inhibition of metaphyseal modeling and remodeling by a bisphosphonate in the growing rat. (Top) Locations of bone resorption in the rat tibia during longitudinal growth (left): osteoclasts resorb calcified cartilage (1), subperiosteal bone (2), and primary spongiosa (3), therefore enlarging the marrow cavity. Effect of clodronate (right). (Bottom) Microradiograph of a normal tibia (left) and of a bone from an animal treated with clodronate (right). Adapted from Schenk et al. (1973). Reproduced from Calcif. Tissue Res. 11, 196 – 214, with copyright permission from the author and Springer-Verlag, Heidelberg.
CHAPTER 16 Basic Biology of Bisphosphonates
453
biomechanically unnecessary excess of bone. This fact suggests that the fear of the dangers of long-term use of therapeutic doses may not be warranted. b. Animals with Experimentally Increased Bone Resorption Bisphosphonates can also prevent an experimentally induced increase in bone resorption. Thus, they impair resorption induced by many bone-resorbing agents such as, among others, parathyroid hormone [10,15], 1,25(OH)2D, and retinoids [23], the latter effect having been used to develop a powerful and rapid screening assay for new compounds. c. Animals with Experimentally Induced Osteoporosis Many osteoporosis models have been investigated for their response to the administration of bisphosphonates. All showed that bone loss could be prevented by bisphosphonates (Fig. 4). The first used was immobilization by a sciatic nerve section in the rat. The administration of etidronate and especially clodronate not only prevented bone loss, but actually led to an increase of radiological bone density and calcium content of the immobilized bone [24,25]. Similar results were obtained with spinal cord section [26] and other means of immobilization (Fig. 5). Because ovariectomy produces significant bone loss in animals it has been used frequently as a model for human postmenopausal osteoporosis. Bisphosphonates have been found to be very active in preventing the induced bone loss in rats [24 – 33], monkeys [34 – 36], dogs [37,38], and other animals. Also the loss induced by castration in males is attenuated by bisphosphonates [39,40]. Another model of clinical interest is corticosteroid induced bone loss. This loss is inhibited by bisphosphonates in the rabbit [41] and other animals. The same is true for the loss induced by other hormones, such as thyroid hormones [42,43]. Effects on the loss induced by low calcium diet are ambiguous. An interesting new model has been developed to mimic osteolysis and aseptic loosening around total hip arthroplasty. In this model, alendronate also inhibited bone destruction [44]. Of clinical importance is the finding that treatments that increase bone formation, such as prostaglandins, IGF-I and
FIGURE 4 Fleisch [5].
Osteoporosis models improved by bisphosphonates. From
FIGURE 5 Effect of 10 mg phosphorus/kg of clodronate subcutaneously on the bone loss induced by a sciatic nerve section in the rat. From Fleisch [5]. PTH, are still effective or maintained after their discontinuation in bisphosphonates-treated animals [32,45], resulting sometimes even in an additive effect of the two treatments. This additive effect of stimulators of bone formation and inhibitors of bone destruction opens an interesting possibility for future therapy. Another question was whether bisphosphonates could display an additive effect together with another inhibitor of bone resorption. This has been shown to be the case with estrogen in humans [46,47]. Practically all of the bisphosphonates tested have been effective. Listed in order of increasing potency in animals, these include etidronate [24,25,27,29,33], tiludronate [31,37,40], clodronate [24,25,39,41], pamidronate [42], olpadronate [26], incadronate [45], alendronate [30,34,35,43], risedronate [28,29,33], ibandronate [38], minodronate, and zoledronate [36]. In the case of etidronate, the effect was blurred at higher doses because they also inhibited mineralization. In the ovariectomized rat the preventive effect was maintained for a long period after discontinuation of the drug. However, many of these studies were performed in growing animals, in which the bisphosphonates increased bone mass conspicuously by inhibiting the resorption of the metaphyseal bone. In such animals, it is most often not possible to know whether an inhibition of the bone loss induced by the various procedures is due only to the effect of the compound on the induced bone loss or to a general effect on endogenous bone resorption. This is especially the case because the proper controls, namely nonosteoporotic animals, are often not available. d. Mechanisms of Action in Bone Loss The mechanism of action of the bisphosphonates in osteoporosis is still not completely understood. The prevention of bone loss is probably explained to a large extent by the decrease
454 in bone turnover, which by itself slows bone loss. This decrease also diminishes the fracture rate, as fewer trabeculae are destroyed. One explanation of the initially occurring increase in bone is that the decrease in bone resorption is followed only later by the “coupling-induced” diminution in formation, which brings an initial gain in calcium and bone balance through the reduction of the so-called remodeling space. In addition, bisphosphonates may also act at the individual bone mineral density (BMU). Indeed, it has been shown that they decrease the depth of the resorption site [35,48]. Because the amount of bone formed at each individual BMU is not decreased, but may possibly even be somewhat increased in animals [35,48] and humans [49], bone balance may be increased at this site. Whether this occurs is under discussion. Actually no increase of the trabecular bone has been reported [49] (Fig. 6). It must be remembered that in nearly all studies BMD and not real bone mass has been measured. A lower turnover will lengthen the life span of the BMU, thus permitting it to mineralize more completely, which will increase BMD without increasing bone mass. This has been described in alendronate-treated baboons [50]. Finally, it has been suggested that bisphosphonates may, to some extent, increase bone formation. Thus very low concentrations of bisphosphonates were found to increase cell multiplication, colony formation, nodule formation, mineralization, and osteocalcin synthesis in bone cell cultures (51– 53). As mentioned previously, some results suggest that the bone formed in individual BMU is possibly somewhat increased under bisphosphonate treatment [48]. Finally, since statins, which are inhibitors of
H. FLEISCH
the mevalonate pathway, increase bone formation at least in vitro, probably through an elevation of BMP-2, it could be conceivable that the bisphosphonates, which also inhibit the mevalonate pathway (see later), have a similar action. Thus, it could be that bisphosphonates might, under certain conditions, increase bone formation in vivo. However, this still needs to be verified.
e. Effect on Mechanical Properties of Bone The effects of bisphosphonates on the mechanical properties of the skeleton have been addressed only recently. This issue is of importance, as it is known that a long-lasting, extreme inhibition of bone resorption can lead to increased bone fragility, with the human osteopetrosis described by Albers – Schönberg being a good illustration. It appears that when not given in excess, bisphosphonates have a positive effect on mechanical characteristics, such as torsional torque, ultimate bending strength, stiffness, maximum elastic strength, Young’s modulus of elasticity, and others, both in normal animals and in various experimental osteoporosis models. The bisphosphonates that proved to be active include alendronate [22,54], clodronate, etidronate [27, 55,57], ibandronate, incadronate [56], minodronate, neridronate, olpadronate [57], pamidronate [55,58], risedronate, and tiludronate [59]. In contrast, when given in excess, bisphosphonates can induce bone to become more prone to fractures both because of an inhibition of mineralization, mainly with etidronate [60], or because of a excessive inhibition of bone resorption [60]. For review, see Ferretti [61] (Fig. 7).
FIGURE 7
FIGURE 6 Possible effect of bisphosphonates at the level of the individual BMU. From Fleisch [5].
Effect of alendronate given intravenously every 2 weeks for a period of 2 years to ovariectomized baboons on bone mineral density and mechanical strength. Squares, not ovariectomized; triangles, ovariectomized; circles, ovariectomized 0.05 mg/kg; diamonds, ovariectomized 0.25 mg/kg. MPa, megapascal. Adapted from Balena et al. [35]. Reproduced from J. Clin. Invest. 92, 2577 – 2586, with copyright permission from the author and the American Society of Clinical Investigation.
455
CHAPTER 16 Basic Biology of Bisphosphonates
The mechanisms leading to the improved mechanical strength are still poorly understood. They may not be caused uniquely, as previously thought, by a higher bone mass, but also by an improvement in architecture and probably a reduction in bone remodeling. Indeed, a higher number of remodeling sites, in which there is excessive osteoclastic destruction of bone, leads to the development of areas of stress concentration, and hence to increased fracture risk. Bisphosphonates may serve as a means of reducing these effects, hence reducing the incidence of new fractures. Another mechanism may be related to the increase in the mineral density following the lower bone turnover [50]. f. Animals with Experimentally Induced Osteolytic Tumors Bisphosphonates also inhibit bone resorption induced experimentally in vitro [62] or in vivo [63,64] by implantation of various tumor cells. They reduce the local destruction of bone near the invading tumor cells, as well as the resorption induced by systemically circulating factors. These effects lead to a partial or total prevention of hypercalcemia and hypercalciuria [63] and to a decrease of metastases and tumor burden [64,65]. This effect is the basis for their use in tumor-related bone disease. For review, see Fleisch [66] and Russell and Croucher [67]. g. Others Of interest in the dental field is the fact that they also slow down periodontal bone destruction in rats susceptible to periodontal disease and in experimental periodontitis in monkeys. Furthermore, they inhibit tooth movement induced by orthodontic procedures [68], and these effects can be achieved even when the compounds are administered topically [69]. Finally, several bisphosphonates inhibit local bone and cartilage resorption, preserve the joint architecture, and decrease the inflammatory reaction in various types of experimental arthritis [70]. 3. RELATIVE ACTIVITY OF BISPHOSPHONATES ON BONE RESORPTION The potency of bisphosphonates on bone resorption varies greatly from compound to compound. For etidronate, the dose required to inhibit resorption is relatively high: in the rat above 1 mg/kg parenterally per day. This dose is
FIGURE 8
very near that which impairs normal mineralization. One of the aims of bisphosphonate research has therefore been to develop compounds with a more powerful antiresorptive activity, without a stronger inhibition of mineralization. This has proven to be possible. Clodronate was already more potent than etidronate [10,15], and pamidronate was found to be even more active [71]. In more recent years, compounds have been developed that are up to 10,000 times more powerful than etidronate in the inhibition of bone resorption in experimental animals without being more active in inhibiting mineralization (Fig. 8). For the development of future compounds, it is of relevance that, so far, the potency evaluated in the rat corresponds quite well to that found in humans, at least with respect to their relative place in the scale of potency. However, the difference of activity between the least and the most potent compound is less in humans and depends on the disease for which the compounds are used. It is much smaller for osteoporosis, less so for Paget’s disease and tumor-induced hypercalcemia. At present, the structural requirements for activity are only partially defined. The length of the aliphatic carbon chain is important, with the effect on bone resorption increasing and then decreasing again with increasing chain length [72]. Adding a hydroxyl group to the carbon atom at position 1 increases activity, and compounds with a nitrogen atom in the side chains are more active. The first compound of the latter kind to be described, pamidronate, has an amino group at the end of the alkyl chain [71]. When the chain is altered in its length, the highest activity is present with a backbone of four carbons, as seen in alendronate [19]. A primary amine is not necessary for this activity, as dimethylation of the amino nitrogen of pamidronate, as seen in olpadronate, increases potency [73]. The latter can still be increased further when other groups are added to the nitrogen, as is the case for ibandronate, [1-hydroxy3[methylpentylamino)propylidene]bisphosphnate, which is extremely potent [74]. Geminal bisphosphonates containing cyclic substituents are also very potent, especially those containing a nitrogen atom in the ring, such as risedronate [75]. The most active compounds described so far, zoledronate [17] and minodronate, belong to this class and contain an imidazole ring.
Potency of some bisphosphonates to inhibit bone resorption in the rat. Compounds in each column are listed in alphabetical order. From Fleisch [5].
456 It must be noted that, at present, all effective compounds have a P – C – P structure, which appears to be a prerequisite for activity. The intensity of the effect is, however, also dependent on the side chain. A three-dimensional structural requirement appears to be involved. Indeed, stereoisomers of the same chemical structure have shown 10-fold differences in activity [76]. This opens the possibility of binding onto some kind of receptor. 4. MECHANISMS OF ACTION ON BONE RESORPTION a. General Concepts Our understanding of the mode of action of the bisphosphonates has made great progress. There is no doubt that the action in vivo is mediated mostly, if not completely, through mechanisms other than the physicochemical inhibition of crystal dissolution, as was initially postulated. However, the exact nature of these mechanisms is still not entirely unraveled. It may well be that several mechanisms operate simultaneously. b. Levels of Action The mechanism of action of the bisphosphonates can be considered at three levels, which are, however, tightly linked one to another. At the tissue level their main effect is a decrease in bone turnover, which is secondary to the inhibition of bone resorption. This effect is due to a decrease in the number and the activity of osteoclasts destroying bone [49,77], which leads to a decrease in the number of new BMUs. Because bone loss is intimately linked to turnover in diseases such as tumor-induced bone disease and osteoporosis, this loss will be reduced by the bisphosphonates. Furthermore, the bisphosphonates act to a certain extent at the individual BMU level by decreasing the depth of the resorption site [35,48,77]. Because the amount of new bone formed in the BMU is not decreased, but possibly even increased [35,48,50,77], local and consequently the whole body bone balance will be less negative or possibly sometimes even positive. Four mechanisms possibly can be involved at the cellular level. 1. Inhibition of osteoclast recruitment. Several bisphosphonates inhibit osteoclast differentiation in various culture systems of both cells [78] and bones [79,80]. Some experiments suggest that the effect occurs at the terminal step of the differentiation process [81]. 2. Inhibition of osteoclastic adhesion to the mineralized matrix. One study in vitro reports such an effect [82]. Whether this takes place in vivo is not yet established. However, there is now excellent evidence that bisphosphonates can inhibit the adhesion of tumor cells in vitro [83,84]. 3. Shortening of the life span of osteoclasts due to earlier apoptosis. It has been reported that bisphosphonates induce
H. FLEISCH
osteoclast programmed cell death (apoptosis], both in vitro and in vivo, and both in normal mice and in mice with increased bone resorption [85]. The ranking of effectiveness of clodronate, pamidronate, and risedronate was the same as seen in vivo. The effect was not due to toxic cell death. 4. Inhibition of osteoclast activity. While the former three mechanisms will lead to a decrease in the number of osteoclasts, which is usually seen after longer treatment, the fourth will lead to inactive osteoclasts, often seen at the beginning of treatment. Several facts suggest that this effect must be present. Thus, following bisphosphonate administration, the number of multinucleated osteoclasts on the bone surface often increases initially, despite a reduced bone resorption [18], however, the cells show changes in morphology and look inactive [18]. The changes are numerous and include alterations in the cytoskeleton, among others, in actin [14,86,87] and vinculin [87], as well as disruption of the ruffled border [18,88 – 90]. It is only later, after chronic administration, that the osteoclast number decreases. The cause for the initial increase is unknown. One possibility is that it could reflect a stimulation of osteoclast formation to compensate for the decrease in osteoclast activity. At the molecular level, the low concentration necessary for activity suggests some sort of “pocket” that induces a cellular transduction mechanism. This site could be either on the cell membrane or within the cell and might be an enzyme, a pump, or some other intracellular protein involved in the signaling cascade. One of these could be farnesyl pyrophosphate synthase, which has been found to be inhibited by nitrogen-containing bisphosphonates and responsible for at least some of their effects (see later) [91 – 93]. A great number of different biochemical effects on various cell types have been described in vitro, but fewer data exist on the osteoclasts themselves. Some of the changes that possibly relate to bone resorption, include reduction in lactic acid production by calvaria [94], various cells [95], and osteoclasts [96,97]. In the latter, an inhibition of the acid extrusion performed by a sodium-independent mechanism and of the vacuolar-type proton ATPase present in the ruffled border has been shown [96,97]. Bisphosphonates also inhibit lysosomal enzyme activity in vitro [98], and decrease prostaglandin synthesis in bone when added both in vitro and in vivo [99,100]. Furthermore, they inhibit matrix metalloproteinases [101], which might be also of interest in view of a possible use in arthritis [102]. An inhibition of certain protein tyrosine phosphatases (PTPase), namely PTPase and , has also been described [103]. Unfortunately, in many cases there is no structure – effect correlation between these effects in vitro and those on bone resorption in vivo when bisphosphonates of various antiresorbing activity are compared. It has been found that nitrogen-containing bisphosphonates can inhibit the mevalonate pathway [104,105] by in-
457
CHAPTER 16 Basic Biology of Bisphosphonates
It was also shown that some nonnitrogen-containing bisphosphonates that closely resemble pyrophosphate, such as etidronate, tiludronate, and clodronate, can be incorporated into the phosphate chain of ATP-containing compounds so that they become nonhydrolyzable. The new P – C – P containing ATP analogues inhibit cell function and may lead to apoptosis and cell death [107 – 109]. Thus, the bisphosphonates can be classified into two major groups with different modes of action. The latter may explain the various cellular changes described earlier (Fig. 10).
FIGURE 9
Effect of bisphosphonates on the mevalonate pathway. From
Fleisch [5].
hibiting farnesyl pyrophosphate synthase [91 – 93]. This leads to a decrease of the formation of isoprenoid lipids such as farnesyl- and geranylgeranylpyrophosphates. These are required for the posttranslational prenylation (transfer of fatty acid chains) of proteins, including the GTP-binding proteins Ras, Rho, Rac, and Rab. These proteins are important for many cell functions, including cytoskeletal assembly and intracellular signaling. Therefore, disruption of their activity will induce a series of changes leading to decreased activity, probably the main effect, and to earlier apoptosis in several cell types, including osteoclasts [104,105]. In osteoclasts the lack of geranylgeranylpyrophosphate is probably responsible for the effects [106] (Fig. 9).
FIGURE 10
c. Direct vs Indirect Effect through Other Cells It is probable that bisphosphonates influence osteoclasts either directly as a result of their cellular binding or intracellular uptake or indirectly via other cells. The direct effect is supported by many facts. As mentioned earlier, bisphosphonates inhibit the formation of resorption cavities by isolated osteoclasts deposited on calcified matrices in vitro [13,14]. One study [87] showed that morphological changes occurred only when the cells were actively resorbing the calcified matrix or if the bisphosphonate was injected into the cells. No changes occurred when the osteoclasts were not active, showing that the bisphosphonates has to be taken up with the resorbed mineral. A direct action on osteoclasts is also supported by the fact that, under certain conditions, bisphosphonates can enter cells [90,95,110], particularly those of the macrophage lineage [111]. They are taken up by the osteoclasts during the resorption process, a process favored by the fact that bisphosphonates also deposit preferentially under osteoclasts where they can attain very high concentrations, in the range
The two classes of bisphosphonates. Courtesy of Dr. M. Rogers.
458
H. FLEISCH
of 104 M or higher [90], and are then released from the mineral at the acid pH prevailing at this location. It is likely that bisphosphonates also act, at least in part, through other cells. One candidate is the osteoblast. It is now generally accepted that cells of osteoblastic lineage control the recruitment and activity of osteoclasts. One of the modulators involved in this mechanism appears to be bisphosphonates. Indeed, these compounds induce the osteoblasts to synthesize an inhibitor(s) for osteoclast recruitment and therefore of bone resorption [112 – 114]. In this regard it might be relevant that bisphosphonates prevent the apoptotic effect of various bone resorbers on osteoblasts and osteocytes [115]. Another candidate target cell population are the macrophages [111], which release many cytokines, which are able to modulate osteoclasts and are influenced by bisphosphonates. Thus, under certain conditions, bisphosphonates inhibit macrophage release of interleukin (IL)-1, IL-6, and TNF in vitro. Alternatively, at high concentrations, such as after intravenous administration, the release of these cytokines can be stimulated, producing an acute phase reaction in humans [116]. It is not known at present to which extent these mechanisms — the direct effect on the osteoclast or indirect action through osteoblastic or other cells — are operating in vivo and, if both do, which of the two is more important (Fig. 11).
B. Inhibition of Mineralization 1. EFFECTS IN VITRO The physicochemical effects of many of the bisphosphonates are very similar to those of pyrophosphate. Thus, they inhibit the formation [11,117,118], the aggregation, and
FIGURE 11 Summary of the effects of bisphosphonates on the osteoclast. From Fleisch [5].
also slow down the dissolution of calcium phosphate crystals [10]. All these effects are related to the marked affinity of these compounds for solid-phase calcium phosphate, on the surface of which they bind strongly [119]. This property is of great importance because it is the basis for the use of these compounds as skeletal markers in nuclear medicine and the basis for their selective localization in bone when used as drugs (Fig. 12). The inhibition of calcium phosphate formation is closely related to the affinity of the bisphosphonate to the solidphase calcium phosphate. The binding can be bidentate through the two phosphates, as is the case for clodronate, or it can be tridentate [120] through a third moiety, such as a hydroxyl or a nitrogen attached to the carbon atom. This is the case for most bisphosphonates used clinically today. The third binding site increases the affinity and hence the inhibitory effect on calcification. Bisphosphonates also inhibit the formation and the aggregation of calcium oxalate crystals [121]. 2. EFFECTS IN VIVO a. Ectopic Mineralization Like pyrophosphate, bisphosphonates inhibit calcification in vivo very efficiently. Thus, they prevent experimentally induced calcification of many soft tissues such as arteries and kidneys [11,117] and skin. They are active when administered either orally or parenterally. In arteries, they decrease not only mineral deposition, but also the accumulation of cholesterol, elastin, and collagen. Bisphosphonates, such as etidronate, can also inhibit the calcification of bioprosthetic heart valves, either when administered subcutaneously or when released locally from various matrices. Etidronate also inhibits ectopic ossification when given either systemically [122] or locally. This effect has led to the clinical use of etidronate in ectopic ossification, but normal mineralization is unfortunately inhibited as well. Similarly, certain bisphosphonates such as etidronate decrease the formation of experimental urinary stones [121]. However, the active dose also leads to the inhibition of normal mineralization of bone so that these compounds cannot be administered in this condition. Finally, topical administration of etidronate induces a decrease in the formation of dental calculus [123], a property that is exploited in some toothpastes.
FIGURE 12 Physicochemical effects of bisphosphonates on calcium phosphate. From Fleisch [5].
459
CHAPTER 16 Basic Biology of Bisphosphonates
b. Normal Mineralization The dose, at least of etidronate, that inhibits experimental ectopic mineralization also impairs the mineralization of normal calcified tissues such as bone and cartilage (Fig. 3) [18], dentine [124,125], and enamel [125]. The amount required to have this effect varies according to animal species and length of treatment. In contrast to bone resorption, where the different compounds vary greatly in their activity, this does not seem to be so much the case for the inhibition of mineralization. For most species and compounds, the effective daily parenteral dose is in the order of 1 – 10 mg of compound phosphorus per kilogram. Interestingly, clodronate inhibits normal mineralization somewhat less than etidronate, despite the fact that it is more active on bone resorption. This may be due to the fact that clodronate has no hydroxyl side chain and is therefore less bound to the mineral. The inhibition of calcification with high doses can lead to fractures [60] and to an impairment of fracture healing. The mineralization defect is eventually reversed after discontinuation of the drug. Bisphosphonates also inhibit calcification of bone in humans when given in larger amounts. This propensity to inhibit the mineralization of normal bone has hampered the therapeutic use of bisphosphonates in ectopic calcification. This is not the case for their use in bone resorption, as compounds have been developed that inhibit this process at doses at least 1000 times lower than those that inhibit mineralization. OF
3. MECHANISMS OF ACTION IN THE INHIBITION CALCIFICATION
There is a close relationship between the ability of an individual bisphosphonate to inhibit the formation of calcium phosphate in vitro and its effectiveness on calcification in vivo [117,126], strongly suggesting that the latter can be explained in terms of a physicochemical mechanism. However, additional effects on matrix formation, involving changes in glycosaminoglycan and collagen synthesis, may occur. These may be direct via cellular effects or mediated indirectly by effects on crystals. In contrast, there is no relation between crystal binding and bone resorption [127].
C. Other Actions In view of the large array of their effects on cells, it is surprising that bisphosphonates act almost exclusively on calcified tissues. This selectivity is explained by the strong affinity of these compounds for calcium phosphate, which allows them to be cleared very rapidly from blood and to be incorporated into calcified tissues, especially bone (see pharmacokinetics). However, some effects exist that are not, or not entirely, explained by the effects on bone. Thus several bisphosphonates inhibit local bone and cartilage resorption, preserve
the joint architecture, and decrease the inflammatory reaction in various types of experimental arthritis, such as that induced by Freund’s adjuvant, carrageenin, or collagen [128,129]. This effect is especially pronounced when bisphosphonates are encapsulated in liposomes [130]. The fact that not only is bone resorption decreased, but also the inflammatory reaction in the joint and in the paw itself is diminished [131] suggest that mechanisms other than those in bone, possibly involving the mononuclear phagocyte system, are operating. These results open the exciting possibility of using bisphosphonates in inflammatory arthritis, given either systemically or locally, possibly encapsulated in liposomes. Of interest, bisphosphonates or phosphonosulfonates linked to an isoprene chain are potent inhibitors of squalene synthase and hence are cholesterol-lowering agents in animals [132], which may open some interesting new therapeutic possibilities for these drugs.
IV. PHARMACOKINETICS Bisphosphonates are synthetic compounds that have not yet been found to occur naturally in animals or humans. No enzymes able to cleave the P – C – P bonds have been described. The bisphosphonates on which data have been published so far appear to be absorbed, stored, and excreted unaltered from the body. Therefore, these bisphosphonates seem to be nonbiodegradable in solution and in animals. However, it cannot be excluded that some bisphosphonates may be metabolized, especially in their side chains. Data from relatively few pharmacokinetic studies are available. Most of the published data have been obtained with alendronate, clodronate, etidronate, pamidronate, and tiludronate [For reviews, see 133 – 135].
A. Intestinal Absorption The bioavailability of an oral dose of a bisphosphonate in animals as well as in humans is low, lying between 1 and 10% [136], probably because of their low lipophilicity, which prevents transcellular transport, and their high negative charge, which prevents paracellular transport. Absorption is proportionally greater when large doses are given, such as with etidronate and clodronate. This is possibly the reason why it is generally lower for the more potent bisphosphonates, which are administered in lower amounts. Absorption is in general higher in the young and shows great inter- and intraspecies variation. This variability can present a problem in humans, especially for compounds such as etidronate, where the dose that shows an adverse event, such as an inhibition of mineralization, is close to that which inhibits bone resorption. The location of the
460
H. FLEISCH
absorption in the gastrointestinal tract is not yet elucidated, although it can occur in the stomach as well as in the small intestine. It appears to occur by passive diffusion, probably through a paracellular pathway [137]. Absorption is diminished substantially when the drug is given with meals, especially in the presence of calcium and iron. The mechanism of this reduction may be due to the conversion of the bisphosphonate into a nonabsorbable form or to a decrease of the absorption process itself. Therefore, bisphosphonates should never be given at mealtimes and never together with milk or dairy products or with iron supplements. For some unknown reasons, orange juice and coffee also decrease absorption [138].
B. Distribution In the blood, only part of the bisphosphonates are ultrafilterable [139,140]. The values vary between about twothirds to only a few percent and are strongly species dependent, being low for rats and higher for larger animals and humans. The nonfilterable fraction is either bound to proteins, especially albumin [140], or present in very small calcium-containing aggregates. Some 20 – 80% of the absorbed bisphosphonate is then taken up by bone, with the remainder being rapidly excreted in the urine. Skeletal uptake varies with species, sex, and age and with the dose and nature of the compound. In humans receiving clinical doses, values are about 20% for clodronate, 50% for etidronate, and more for alendronate and pamidronate. Sometimes bisphosphonates, especially pamidronate, deposit in other organs, mostly the liver and the spleen [141]. Such deposition is proportionally greater when large amounts of the compounds are given. At least part of this extraosseous deposition appears to be due to the formation of complexes with metals or to aggregates after too high or too rapid intravenous injection. These complexes are then phagocytosed by the macrophages of the reticuloendothelial system. Therefore, data obtained from studies using large amounts of labeled bisphosphonate given rapidly intravenously should be interpreted with caution. The formation of aggregates in the blood is thought to occur in humans following rapid intravenous injections of large quantities, possibly explaining the renal failure that can ensue. The circulating half-life of bisphosphonates is short, in the order of minutes in the rat [142]. In humans it is somewhat longer, 0.5 – 2 h. The rate of entry into bone is very fast, similar to that of calcium and phosphate. Bone clearance is compatible with a complete extraction by the skeleton after the first passage [142] so that skeletal uptake might be determined to a large extent by skeletal vascularization and blood flow. The areas of deposition were generally thought to be mostly those of bone formation. This property is used to measure areas of high bone turnover in nuclear medicine by
means of 99mTc-linked bisphosphonates. However, alendronate, when given in therapeutic doses, has been found to accumulate preferentially under osteoclasts [143]. This is also the case, although to a lesser extent, for etidronate when given in the same amount. When given at a therapeutic dose, the latter, however, accumulates equally under both cells. This suggests that when a bisphosphonate is given in small doses, which is the case for all newer compounds, it is likely to deposit preferentially in locations of bone resorption. The rapid uptake by bone means that the soft tissues are exposed to bisphosphonates for only short periods, explaining why practically only bone is affected in vivo. When bisphosphonates are given to humans in clinically effective doses, there seems to be no saturation in their total skeletal uptake, at least within periods as long as years or decades. In contrast, with continuous administration, the antiresorbing effect reaches a maximum relatively rapidly [71], both in animals and in human. The level of this maximal effect depends on the dose administered, as does the duration of the effect after discontinuation of the drug. The fact that a plateau of activity is reached, despite the fact that the bisphosphonate continues to be incorporated, suggests that the compounds are buried in the bone and become inactive (Fig. 13).
FIGURE 13 Effect of various doses of pamidronate administered daily sc on urinary hydroxyproline excretion in the rat. The maximal effect is obtained rapidly and depends on the dose given. Adapted and reproduced from Reitsma et al. [143]. Calcif. Tissue Int. 32, 145 – 157, with copyright permission from the author and Springer-Verlag, Heidelberg.
461
CHAPTER 16 Basic Biology of Bisphosphonates
Once deposited in the skeleton, part of the bisphosphonate is liberated again by physicochemical mechanisms. Once buried under new layers of bone, they will be released to a large extent only when the bone in which they were deposited is resorbed. Thus the half-life in the body depends on the rate of bone turnover itself. As the bisphosphonates slow down the resorption of the bone in which they are deposited, their half-life may be even longer than the normal half-life of the skeleton. The half-life of various bisphosphonates is between 3 months and up to a year in mice or rats [144], with clodronate being cleared somewhat faster than etidronate and pamidronate. For humans it is much longer, for some bisphosphonates over 10 years [145], and it is possible that a portion of the administered compounds remains in the body for life. However, this is also true for other bone seekers such as tetracyclines, heavy metals, and fluoride. There is no indication that the bisphosphonate buried in the skeleton has any pharmacological activity. On the contrary, in the rat, bone formed under administration of even high doses of alendronate can be resorbed normally. However, at sites where the bisphosphonate is deposited in large amounts, such as in high turnover locations of patients with bone metastases or with Paget’s disease, the long skeletal retention may explain why one single administration of a bisphosphonate can be active for long periods of time, both in animals and in humans.
FIGURE 14
Pharmacokinetics of bisphosphonates. From Fleisch [5].
Acute, subacute, and chronic administration in several animal species have in general revealed little toxicity. Teratogenicity, mitogenicity, and carcinogenicity tests have been negative. When bisphosphonates are administered subcutaneously, local toxicity can occur, with local necrosis. This is especially the case for the nitrogen-containing derivatives such as pamidronate.
A. Acute Toxicity
The renal clearance of bisphosphonates is high. When taking into account their only partial ultrafilterability, it can, at least in animals, exceed that of inulin, indicating active secretion [146,147]. The secretory pathway involved is not yet characterized. Urinary excretion is decreased in renal failure and the removal by peritoneal dialysis is poor, which has to be accounted for when the compounds are administered to patients with kidney disease (Fig. 14).
Acute toxicity can include induction of hypocalcemia and appears to be due mainly to the formation of complexes or aggregates with calcium, which lead to a decrease in ionized calcium. Toxicity therefore varies with the speed of infusion when the compounds are administered intravenously, so that the rate of infusion in humans must be controlled carefully when given in high doses. In contrast, it appears that very potent bisphosphonates, such as ibandronate, may be given safely given in doses up to 2 – 3 mg as an intravenous bolus injection. In the event of hypocalcemia, calcium infusion can correct the signs and symptoms rapidly, Some acute toxicity can also be due to renal tubular effects.
D. Other Modes of Application
B. Nonacute Toxicity
Bisphosphonates are also bioavailable to some extent when given intranasally and through the skin. This may open new modes of administration in clinical practice.
In view of the large array of cellular effects obtained in vitro with the bisphosphonates, one would have expected a large number of toxic effects. This is not the case, and when administered in pharmacological doses, bisphosphonates seem to act almost exclusively on calcified tissues and secondarily on plasma calcium, but are otherwise very well tolerated. This selectivity is explained by the strong affinity of these compounds for calcium phosphate, which allows them to be incorporated rapidly into calcified tissues, especially bone, and therefore to be cleared quickly from the blood.
C. Renal Clearance
V. ANIMAL TOXICOLOGY Published animal toxicological data are scanty and deal mostly with alendronate, clodronate, etidronate, incadronate, pamidronate, and tiludronate. Unfortunately, little is published about other compounds.
462 The nonskeletal toxicity associated with compounds used clinically occurs only when doses substantially larger than those which inhibit bone resorption are used. In general, the first organ to show cellular alterations with all bisphosphonates, as well as with polyphosphates and phosphate itself, is the kidney [148,149]. The liver, as well as the testis, the epididymis, the prostate, and possibly the lung, can in some cases also show alterations. Some inflammatory gastrointestinal changes have been described, with parenteral administration at high doses. After the appearance in humans of gastrointestinal adverse events after oral administration of nitrogen-containing bisphosphonates [150], special attention has been given to the effects of oral bisphosphonates in animals. Thus alendronate, when given orally to rats at suprapharmacological doses, has been reported to occasionally induce gastric and esophageal erosions and ulcerations and delay healing of indomethacin-induced gastric erosions. These effects are not attributable to changes in gastric acid secretion or prostaglandin synthesis, but are thought to be due to a topical irritant effect. No esophageal irritation occurred when the pH was above 3.5 in a dog model. Similar effects were reported with etidronate, risedronate, and tiludronate when given at pharmacologically equivalent doses. All these effects were obtained at doses much larger than those given to humans [151,152]. The most relevant toxicity associated with bisphosphonates is the inhibition of bone and cartilage calcification [17,22,153]. This starts to occur at parenteral doses of approximately 5 – 10 mg/kg daily. The radiological appearance resembles rickets or osteomalacia, although there are some histological differences. Fractures can occur after the long-term administration of high doses and are probably the result of defective mineralization. Developmental disturbances of enamel can also appear at high systemic doses. Fractures can also be induced occasionally by very large amounts inducing a pathological decrease in bone resorption without signs of osteomalacia or rickets [22]. The fractures may be caused by the extreme long-term decrease in bone turnover, which can itself lead to an increased fragility, as is well known in human congenital osteopetrosis, or by bone cell toxicity. Finally, certain bisphosphonates, such as etidronate and pamidronate, cross the placenta and can affect the fetus. Very large doses can lead to a decrease in the number of live pups, to fetal abnormalities of the skeleton and the skin, and to malformations and hemorrhages [154 – 158). In view of these results, bisphosphonates should, as a general rule, not be administered to pregnant women. It must be stressed that the results with one bisphosphonate cannot necessarily be extrapolated to other bisphosphonates. Indeed, toxicity, both in cell and organ culture and in vivo, varies greatly from one compound to another.
H. FLEISCH
References 1. G. A. Rodan and H. A. Fleisch, Bisphosphonates: Mechanisms of action. J. Clin. Invest. 97, 2692 – 2699 (1996). 2. F. H. Ebetino, M. D. Francis, M. J. Rogers, and R. G. G. Russell, Mechanisms of action of etidronate and other bisphosphonates. Rev. Contemp. Pharmacother. 9, 233 – 243 (1998). 3. H. Fleisch, Bisphosphonates: Mechanisms of action. Endocr. Rev. 19, 80 – 100 (1998). 4. G. A. Rodan, Mechanisms of action of bisphosphonates. Annu. Rev. Pharmacol. Toxicol. 38, 375 – 388 (1998). 5. H. Fleisch, “Bisphosphonates in Bone Disease: From the Laboratory to the Patient.” Academic Press San Diego, 2000. 6. M. J. Rogers, J. C. Frith, S. P. Luckman, F. P. Coxon, H. L. Benford, J. Monkkonen, S. Auriola, K. M. Chilton, and R. G. G. Russell, Molecular mechanisms of action of bisphosphonates. Bone 24, 73S – 79S (1999). 7. N. Menschutkin, Ueber die Einwirkung des Chloracetyls auf phosphorige Säure. Ann. Chem. Pharm. 133, 317 – 320 (1865). 8. L. J. M. J. Blomen, History of bisphosphonates: Discovery and history of the non-medical uses of bisphosphonates. In “Bisphosphonate on Bones” (O. L. M. Bijvoet, H. A., Fleisch, R. E. Canfield, and R. G. G. Russell, eds.), pp. 111 – 124. Elsevier, Amsterdam, 1995. 9. H. Fleisch, R. G. G. Russell, S. Bisaz, P. A. Casey, and R. C. Mühlbauer, The influence of pyrophosphate analogues (diphosphonates) on the precipitation and dissolution of calcium phosphate in vitro and in vivo. Calcif. Tissue Res. 2(Suppl.), 10 – 10A (1968). 10. H. Fleisch, R. G. G. Russell, and Francis, M. D. Diphosphonates inhibit hydroxyapatite dissolution in vitro and bone resorption in tissue culture and in vivo. Science 165, 1262 – 1264 (1969). 11. M. D. Francis, R. G. G. Russell, and H. Fleisch, Diphosphonates inhibit formation of calcium phosphate crystals in vitro and pathological calcification in vivo. Science, 165, 1264 – 1266 (1969). 12. M. L. Lamson, J. L. Fox, and W. I. Huguchi, Calcium and 1-hydroxyethylidene-1,1-bisphosphonic acid: Polynuclear complex formation in the physiological range of pH. Int. J. Pharm. 21, 143 – 154 (1984). 13. A. M. Flanagan and T. J. Chambers, Dichloromethylene-bisphosphonate (Cl2MBP) inhibits bone resorption through injury to osteoclasts that resorb Cl2MBP-coated bone. Bone Miner. 6, 33 – 43 (1989). 14. M. Sato and W. Grasser, Effects of bisphosphonates on isolated rat osteoclasts as examined by reflected light microscopy. J. Bone Miner. Res. 5, 31 – 40 (1990). 15. R. G. G. Russell, R. C. Mühlbauer, S. Bisaz, D. A. Williams, and H. Fleisch, The influence of pyrophosphate, condensed phosphates, phosphonates and other phosphate compounds on the dissolution of hydroxyapatite in vitro and on bone resorption induced by parathyroid hormone in tissue culture and in thyroparathyroidectomised rats. Calcif. Tissue Res. 6, 183 – 196 (1970). 16. J. J. Reynolds, C. Minkin, D. B. Morgan, D. Spycher, and H. Fleisch, The effect of two diphosphonates on the resorption of mouse calvaria in vitro. Calcif. Tissue Res. 10, 302 – 313 (1972). 17. J. R. Green, K. Müller, and K. A. Jaeggi, Preclinical pharmacology of CGP 42’446, a new, potent, heterocyclic bisphosphonate compound. J. Bone Miner. Res. 9, 745 – 751 (1994). 18. R. Schenk, W. A. Merz, R. Mühlbauer, R. G. G. Russell, and H. Fleisch, Effect of ethane-1-hydroxy-1,1-diphosphonate (EHDP) and dichloromethylene diphosphonate (Cl2MDP) on the calcification and resorption of cartilage and bone in the tibial epiphysis and metaphysis of rats. Calcif. Tissue Res. 11, 196 – 214 (1973). 19. R. Schenk, P. Eggli, H. Fleisch, and S. Rosini, Quantitative morphometric evaluation of the inhibitory activity of new aminobisphosphonates on bone resorption in the rat. Calcif. Tissue Int. 38, 342 – 349 (1986).
CHAPTER 16 Basic Biology of Bisphosphonates 20. A. B. Gasser, D. B. Morgan, H. A. Fleisch, and L. J. Richelle, The influence of two diphosphonates on calcium metabolism in the rat. Clin. Sci. 43, 31 – 45 (1972). 21. R. C. Mühlbauer and H. Fleisch, A method for continual monitoring of bone resorption in rats: Evidence for a diurnal rhythm. Am. J. Physiol. 259, R679 – R689 (1990). 22. C. P. Peter, J. Guy, M. Shea, W. Bagdon, W. F. Kline, and W. C. Hayes, Long-term safety of the aminobisphosphonate alendronate in adult dogs. I. General safety and biomechanical properties of bone. J. Pharmacol. Exp. Ther. 276, 271 – 276 (1996). 23. U. Trechsel, A. Stutzer, and H. Fleisch, Hypercalcemia induced with an arotinoid in thyroparathyroidectomized rats: New model to study bone resorption in vivo. J. Clin. Invest. 80, 1679 – 1686 (1987). 24. W. R. Michael, W. R. King, and M. D. Francis, Effectiveness of diphosphonates in preventing “osteoporosis” of disuse in the rat. Clin. Orthopaed. 78, 271 – 276 (1971). 25. R. C. Mühlbauer, R. G. G. Russell, D. A. Williams, and H. Fleisch, The effects of diphosphonates, polyphosphates, and calcitonin on “immobilisation” osteoporosis in rats. Eur. J. Clin. Invest. 1, 336 – 344 (1971). 26. A. Schoutens, M. Verhas, N. Dourov, P. Bergmann, F. Caulin, A. Verschaeren, M. Mone, and A. Heilporn, Bone loss and bone blood flow in paraplegic rats treated with calcitonin, diphosphonate, and indomethacin. Calcif. Tissue Int. 42, 136 – 142 (1988). 27. E. Shiota, Effects of diphosphonate on osteoporosis induced in rats: Roentgenological, histological and biomechanical studies. Fukuoka Acta Med. 76(6), 317 – 342 (1985). 28. J. E. McOsker and W. K. Sietsema, Preclinical pharmacology of risedronate, a novel bisphosphonate. In “Osteoporosis 1990, Third International Symposium on Osteoporosis, Copenhagen, Denmark” (C. Christiansen and K. Overgaard eds.), Osteopress ApS, Copenhagen, 1990. 29. T. J. Wronski, C-F. Yen, and K. S. Scott, Estrogen and diphosphonate treatment provide long-term protection against osteopenia in ovariectomized rats. J. Bone Miner. Res. 6, 387 – 394 (1991). 30. J. G. Seedor, H. A. Quartuccio, and D. D. Thompson, The bisphosphonate alendronate (MK-217) inhibits bone loss due to ovariectomy in rats. J. Bone Miner. Res. 6, 339 – 346 (1991). 31. P. Ammann, R. Rizzoli, J. Caverzasio, T. Shigematsu, D. Slosman, and J. P. Bonjour, Effects of the bisphosphonate tiludronate on bone resorption, calcium balance, and bone mineral density. J. Bone Miner. Res. 8, 1491 – 1498 (1993). 32. W. S. S. Jee, L. Tang, H. Z. Ke, R. B. Setterberg, and D. B. Kimmel, Maintaining restored bone with bisphosphonate in the ovariectomized rat skeleton: Dynamic histomorphometry of changes in bone mass. Bone 14, 493 – 498 (1993). 33. T. J. Wronski, L. M. Dann, H. Qi, and C. F. Yen, Skeletal effects of withdrawal of estrogen and diphosphonate treatment in ovariectomized rats. Calcif. Tissue Int. 53, 210 – 216 (1993). 34. D. D. Thompson, J. G. Seedor, H. Quartuccio, H. Solomon, C. Fioravanti, J. Davidson, H. Klein, R. Jackson, J. Clair, D. Frankenfield, E. Brown, H. A. Simmons, and G. A. Rodan, The bisphosphonate, alendronate, prevents bone loss in ovariectomized baboons. J. Bone Miner. Res. 7, 951 – 960 (1992). 35. R. Balena, B. C. Toolan, M. Shea, A. Markatos, E. R. Myers, S. C. Lee, E. E. Opas, J. G. Seedor, H. Klein, D. Frankenfield, H. Quartuccio, C. Fioravanti, J. Clair, E. Brown, W. C. Hayes, and G. A. Rodan, The effects of 2-year treatment with the aminobisphosphonate alendronate on bone metabolism, bone histomorphometry, and bone strength in ovariectomized non human primates. J. Clin. Invest. 92, 2577 – 2586 (1993). 36. N. Binkley, D. Kimmel, J. Bruner, A. Haffa, B. Davidowitz, C. Meng, V. Schaffer, and J. Green, Zoledronate prevents the development of absolute osteopenia following ovariectomy in adult Rhesus monkeys. J. Bone Miner. Res. 13, 1775 – 1782 (1998).
463 37. M. C. Vernejoul, de, Y. Jiang, F. Lacheretz, A. Barbier, P. Geusens, C. Morieux, and C. Pfersdorff, Prevention of bone loss following tiludronate administration to ovariectomized beagle dogs. In “Osteoporosis 1990, Third International Symposium on Osteoporosis, Copenhagen, Denmark” (C. Christiansen and K. Overgaard eds.), pp. 1119 – 1122. Osteopress ApS, Copenhagen, 1990. 38. M.-C. Monier-Faugère, R. M. Friedler, F. Bauss, and H. H. Malluche, A new bisphosphonate, BM 21.0955, prevents bone loss associated with cessation of ovarian function in experimental dogs. J. Bone Miner. Res. 8, 1345 – 1355 (1993). 39. C. S. Wink, M. S. Onge, and B. Parker, The effects of dichloromethylene bisphosphonate on osteoporotic femora of adult castrated male rats. Acta Anat. 124, 117 – 121 (1985). 40. A. Barbier, X. Edmonds-Alt, J. C. Brelière, and D. Ethgen, In vitro and in vivo osseous pharmacological profile of tiludronate: Implication for osteoporosis treatment. In “Third International Symposium on Osteoporosis, Copenhagen, Denmark” (C. Christiansen and K. Overgaard eds.), pp.1127 – 1130 Osteopress ApS, Copenhagen, 1990. 41. W. S. S. Jee, H. E. Black, and J. E. Gotcher, Effect of dichloromethane diphosphonate on cortisol-induced bone loss in young adult rabbits. Clin. Orthop. Res. 158, 39 – 51 (1981). 42. H. N. Rosen, E. K. Sullivan, V. L. Middlebrooks, A. J. Zeind, C. Gundberg, R. Dresner-Pollak, L. A. Maitland, J. M. Hock, A. C. Moses, and S. L. Greenspan, Parenteral pamidronate prevents thyroid hormone-induced bone loss in rats. J. Bone Miner. Res. 8(10), 1255 – 1261 (1993). 43. M. Yamamoto, A. Markatos, J. G. Seedor, P. Masarachia, M. Gentile, G. A. Rodan, and R. Balena, The effects of the aminobisphosphonate alendronate on thyroid hormone-induced osteopenia in rats. Calcif. Tissue Int. 53, 278 – 282 (1993). 44. A. S. Shanbhag, C. T. Hasselman, and H. E. Rubash, Inhibition of wear debris mediated osteolysis in a canine total hip arthroplasty model. Clin. Orthop. 344, 33 – 43 (1997). 45. Y. Takano, T. Tanizawa, T. Mashiba, N. Endo, S. Nishida, and H. E. Takahashi, Maintaining bone mass by bisphosphonate incadronate disodium (YM175) sequential treatment after discontinuation of intermittent human parathyroid hormone (1 – 34) administration in ovariectomized rats. J. Bone Miner. Res. 11, 169 – 177 (1996). 46. S. J. Wimalawansa, A four-year randomized controlled trial of hormone replacement and bisphosphonate, alone or in combination, in women with postmenopausal osteoporosis. Am. J. Med. 104, 219 – 226 (1998). 47. R. Lindsay, F. Cosman, R. O. Lobo, B. W. Walsh, S. T. Harris, J. T. Reagan, C. L. Liss, M. E. Melton, and C. A. Byrnes, Addition of alendronate to ongoing hormone replacement therapy in the treatment of osteoporosis: A randomized, controlled clinical trial. J. Clin. Endocrinol. Metab. 84, 3076 – 3081 (1999). 48. R. W. Boyce, C. L. Paddock, J. R. Gleason, W. K. Sietsema, and E. F. Eriksen, The effects of risedronate on canine cancellous bone remodeling: Three-dimensional kinetic reconstruction of the remodeling site. J. Bone Miner. Res. 10, 211 – 215 (1995). 49. P. M. Chavassieux, M. E. Arlot, C. Reda, L. Wei, A. J. Yates, and P. M. Meunier, Histomorphometric assessment of the long-term effects of alendronate on bone quality and remodeling in patients with osteoporosis. J. Clin. Invest. 100, 1475 – 1480 (1997). 50. P. J. Meunier and G. Boivin, Bone mineral density reflects bone mass but also the degree of mineralization of bone: Therapeutic implications. Bone 5, 373 – 377 (1997). 51. H. L. Guenther, H. E. Guenther, and H. Fleisch, The effects of 1-hydroxyethane-1, 1-diphosphonate and dichloromethanedisphosphonate on collagen synthesis by rabbit articular chondrocytes and rat bone cells. Biochem. J. 196, 293 – 301 (1981). 52. M. Tsuchimoto, Y. Azuma, O. Higuchi, I. Sugimoto, N. Hirata, M. Kiyoki, and I. Yamamoto, Alendronate modulates osteogenesis
464
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
H. FLEISCH of human osteoblastic cells in vitro. Jpn. J. Pharmacol. 66, 25 – 33 (1994). N. Giuliani, M. Pedrazzoni, G. Negri, G. Passeri, M. Impicciatore, and G. Girasole, Bisphosphonates stimulate formation of osteoblast precursors and mineralized nodules in murine and human bone marrow cultures in vitro and promote early osteoblastogenesis in young and aged mice in vivo. Bone 22, 455 – 461 (1998). J. A. Guy, M. Shea, C. P. Peter, R. Morrissey, and W. C. Hayes, Continuous alendronate treatment throughout growth, maturation, and aging in the rat results in increases in bone mass and mechanical properties. Calcif. Tissue Int. 53, 283 – 288 (1993). J. L. Ferretti, C. J. Delgado, R. F. Capozza, G. Cointry, E. Montuori, E. Roldán, A. Pérez-Lloret, and J. R. Zanchetta, Protective effects of disodium etidronate and pamidronate against the biomechanical repercussion of betamethasone-induced osteopenia in growing rat femurs. Bone Miner. 20, 265 – 276 (1993). H. Motoie, T. Nakamura, N. O’Uchi, H. Nishikawa, H. Kanoh, T. Abe, and H. Kawashima, Effects of the bisphosphonate YM175 on bone mineral density, strength, structure, and turnover in ovariectomized beagles on concomitant dietary calcium restriction. J. Bone Miner. Res. 10, 910 – 920 (1995). J. L. Ferretti, N. Mondelo, R. F. Capozza, G. R. Cointry, J. R. Zanchetta, and E. Montuori, Effects of large doses of olpadronate (dimethyl-pamidronate) on mineral density, cross-sectional architecture, and mechanical properties of rat femurs. Bone 16 (Suppl. 4), 285S – 293S 57 (1995). J. L. Ferretti, G. Cointry, R. Capozza, E. Montuori, E. Roldan, and A. Perez Lloret, Biomechanical effects of the full range of useful doses of (3-amino-1-hydroxypropylidene)-1,1-bisphosphonate (APD) on femur diaphyses and cortical bone tissue in rats. Bone Miner. 11, 111 – 122 (1990). P. Geusens, G. Nijs, G. Van der Perre, R. Van Audekercke, G. Lowet, S. Goovaerts, A. Barbier, F. Lacheretz, B. Remandet, Y. Jiang, and J. Dequeker, Longitudinal effect of tiludronate on bone mineral density, resonant frequency, and strength in monkeys. J. Bone Miner. Res. 7(6), 599 – 608 (1992). L. Flora, G. S. Hassing, A. M. Parfitt, and A. R. Villanueva, Comparative skeletal effects of two diphosphonates in dogs. Metab. Bone Dis. Relat. Res. 2, 389 – 407 (1980). J. L. Ferretti, Effects of bisphosphonates on bone biomechanics. In “Bisphosphonate on Bones” (O. L. M. Bijvoet, H. A. Fleisch, R. E. Canfield, and R. G. G. Russell, eds.), pp. 211 – 229. Elsevier, Amsterdam, 1995. A. Jung, B. Mermillod, C. Barras, M. Baud, and B. Courvoisier, Inhibition by two diphosphonates of bone lysis in tumor-conditioned media. Cancer Res. 41, 3233 – 3237 (1981). R. R. Martodam, K. S. Thornton, D. A. Sica, S. M. D’Souza, L. Flora, and G. R. Mundy, The effects of dichloromethylene diphosphonate on hypercalcemia and other parameters of the humoral hypercalcemia of malignancy in the rat Leydig cell tumor. Calcif. Tissue Int. 35, 512 – 519 (1983). A. Jung, J. Bornand, B. Mermillod, C. Edouard, and P. J. Meunier, Inhibition by diphosphonates of bone resorption induced by the Walker tumor of the rat. Cancer Res. 44, 3007 – 3011 (1984). A. Sasaki, B. F. Boyce, B. Story, K. R. Wright, M. Chapman, R. Boyce, G. R. Mundy, and T. Yoneda, Bisphosphonate risedronate reduces metastatic human breast cancer burden in bone in nude mice. Cancer Res. 55, 3551 – 3557 (1995). H. Fleisch, Bisphosphonates: Introduction and mechanisms of action in tumor prevention. In “Tumor Bone Diseases and Osteoporosis in Cancer Patients. Pathophysiology, Diagnosis, and Therapy” (J.-J. Body, ed.), pp. 357 – 376. Dekker, New York, 2000. R. G. Russell and P. J. Croucher, Clinical pharmacology of bisphosphonates. In “Tumor Bone Diseases and Osteoporosis in Cancer
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
79.
80.
81.
82.
83.
Patients: Pathophysiology, Diagnosis, and Therapy” (J.-J. Body, ed.), pp. 377 – 392. Dekker, New York, 2000. K. Shoji, H. Horiuchi, and H. Shinoda, Inhibitory effects of a bisphosphonate (risedronate) on experimental periodontitis in rats. J. Periodont. Res. 30, 27 – 84 (1995). K. Igarashi, H. Adachi, H. Mitani, and H. Shinoda, Inhibitory effect of the topical administration of a bisphosphonate (risedronate) on root resorption incident to orthodontic tooth movements in rats. J. Dent. Res. 75, 1644 – 1659 (1996). M. D. Francis, K. Hovancik, and R. W. Boyce, NE-58095: A diphosphonate which prevents bone erosion and preserves joint architecture in experimental arthritis. Int. J. Tissue React. 11, 239 – 252 (1989). P. H. Reitsma, O. L. M. Bijvoet, H. Verlinden-Ooms, and L. J. A. van der Wee-Pals, Kinetic studies of bone and mineral metabolism during treatment with (3-amino-1-hydroxypropylidene)-1,1-bisphosphonate (ADP) in rats. Calcif. Tissue Int. 32, 145 – 157 (1980). H. Shinoda, G. Adamek, R. Felix, H. Fleisch, R. Schenk, and P. Hagan, Structure-activity relationships of various bisphosphonates. Calcif. Tissue Int. 35, 87 – 99 (1983). P. M. Boonekamp, C. W. G. M. Löwik, L. J. A. van der Wee-Pals, M. L. L.van Wijk-van Lennep, and O. L. M. Bijvoet, Enhancement of the inhibitory action of APD on the transformation of osteoclast precursors into resorbing cells after dimethylation of the amino group. Bone Miner. 2, 29 – 42 (1987). R. C. Mühlbauer, F. Bauss, R. Schenk, M. Janner, E. Bosies, K. Strein, and H. Fleisch, BM 21.0955, a potent new bisphosphonate to inhibit bone resorption. J. Bone Miner. Res. 6, 1003 – 1011 (1991). W. K. Sietsema, F. H. Ebetino, A. M. Salvagno, and J. A. Bevan, Antiresorptive dose-response relationships across three generations of bisphosphonates. Drugs Exp. Clin. Res. XV, 389 – 396 (1989). M. Takeuchi, S. Sakamoto, M. Yoshida, T. Abe, and Y. Isomura, Studies on novel bone resorption inhibitors. I. Synthesis and pharmacological activities of aminomethylenebisphosphonate derivatives. Chem. Pharm. Bull. 41, 688 – 693 (1993). T. Storm, T. Steiniche, G. Thamsborg, and F. Melsen, Changes in bone histomorphometry after long-term treatment with intermittent, cyclic etidronate for postmenopausal osteoporosis. J. Bone Miner. Res. 8, 199 – 208 (1993). D. E. Hughes, B. R. MacDonald, R. G. G. Russell, and M. Gowen, Inhibition of osteoclast-like cell formation by bisphosphonates in long-term cultures of human bone marrow. J. Clin. Invest. 83, 1930 – 1935 (1989). P. M. Boonekamp, L. J. A. van der Wee-Pals, M. M. L. van Wijk-van Lennep, C. W. Thesing, and O. L. M. Bijvoet, Two modes of action of bisphosphonates on osteoclastic resorption of mineralized matrix. Bone Miner. 1, 27 – 39 (1986). P. M. Boonekamp, C. W. G. M. Löwik, L. J. A. van der Wee-Pals, M. L. L. van Wijk-van Lennep, and O. L. M. Bijvoet, Enhancement of the inhibitory action of APD on the transformation of osteoclast precursors into resorbing cells after dimethylation of the amino group. Bone Miner. 2, 29 – 42 (1987). C. W. G. M. Löwik, G. van der Pluijm, L. J. A. van der Wee-Pals, H. Bloys van Treslong-de Groot, and O. L. M. Bijvoet, Migration and phenotypic transformation of osteoclast precursors into mature osteoclasts: The effect of a bisphosphonate. J. Bone Miner. Res. 3, 185–192 (1988). S. Colucci, V. Minielli, G. Zambonin, N. Cirulli, G. Mori, M. Serra, V. Patella, A. Zambonin Zallone, and M. Grano, Alendronate reduces adhesion of human osteoclast-like cells to bone and bone proteincoated surfaces. Calcif. Tissue Int. 63, 230 – 235 (1998). G. van der Pluijm, H. Vloedraven, E. van Beek, L. van der Wee-Pals, C. Löwik, and S. Papapoulos, Bisphosphonates inhibit the adhesion of breast cancer cells to bone matrices in vitro. J. Clin. Invest. 98, 698 – 705 (1996).
465
CHAPTER 16 Basic Biology of Bisphosphonates 84. S. Boissier, S. Magnetto, L. Frappart, B. Cuzin, F. H. Ebetino, P. D. Delmas, and P. Clezardin, Bisphosphonates inhibit prostate and breast carcinoma cell adhesion to unmineralized and mineralized bone extracellular matrices. Cancer Res. 57, 3890 – 3894 (1997). 85. D. E. Hughes, K. R. Wright, H. L. Uy, A. Sasaki, T. Yoneda, G. D. Roodman, G. R. Mundy, and B. F. Boyce, Bisphosphonates promote apoptosis in murine osteoclasts in vitro and in vivo. J. Bone Miner. Res. 10, 1478 – 1487 (1995). 86. K. Selander, P. Lehenkari, and H. K. Väänänen, The effects of bisphosphonates on the resorption cycle of isolated osteoclasts. Calcif. Tissue Int. 55, 368 – 375 (1994). 87. H. Murakami, N. Takahashi, T. Sasaki, N. Udagawa, S. Tanaka, I. Nakamura, D. Zhang, A. Barbier, and T. Suda, A possible mechanism of the specific action of bisphosphonates on osteoclasts: Tiludronate preferentially affects polarized osteoclasts having ruffled borders. Bone 17, 137 – 144 (1995). 88. S. C. Miller and W. S. S. Jee, The effect of dichloromethylenediphosphonate, a pyrophosphate analog, on bone and bone cell structure in the growing rat. Anat. Rec. 193, 439 – 462 (1979). 89. C. M. T. Plasmans, P. H. K. Jap, W. Kuijpers, and T. J. J. H. Slooff, Influence of a diphosphonate on the cellular aspect of young bone tissue. Calcif. Tissue Int. 32, 247 – 256 (1980). 90. M. Sato, W. Grasser, N. Endo, R. Akins, H. Simmons, D. D. Thompson, E. Golub, and G. A. Rodan, Bisphosponate action. Alendronate localization in rat bone and effects on osteoclast ultrastructure. J. Clin. Invest. 88, 2095 – 2105 (1991). 91. E. van Beek, E. Pieterman, L. Cohen, C. Löwik, and S. Papapoulos, Nitrogen-containing bisphosphonates inhibit isopentenyl pyrophosphate isomerase/farnesyl pyrophosphate synthase activity with relative potencies corresponding to their antiresorptive potencies in vitro and in vivo. Biochem. Biophys. Res. Commun. 255, 491 – 494 (1999). 92. E. van Beek, E. Pieterman, L. Cohen, C. Löwik, and S. Papapoulos, Farnesyl pyrophosphate synthase is the molecular target of nitrogencontaining bisphosphonates. Biochem. Biophys. Res. Commun. 264, 108 – 111 (1999). 93. J. D. Bergstrom, R. G. Bostedor, P. J. Masarachia, A. A. Reszka, and G. Rodan, Alendronate is a specific, nanomolar inhibitor of farnesyl diphosphate synthase. Arch. Biochem. Biophys. 373, 231 – 241 (2000). 94. D. B. Morgan, A. Monod, R. G. G. Russell, and H. Fleisch, Influence of dichloromethylene diphosphonate (Cl2MDP) and calcitonin on bone resorption, lactate production and phosphatase and pyrophosphatase content of mouse calvaria treated with parathyroid hormone in vitro. Calcif. Tissue Res. 13, 287 – 294 (1973). 95. D. K. Fast, R. Felix, C. Dowse, W. F. Neuman, and H. Fleisch, The effects of diphosphonates on the growth and glycolysis of connective-tissue cells in culture. Biochem. J. 172, 97 – 107 (1978). 96. Z. Zimolo, G. Wesolowski, and G. A. Rodan, Acid extrusion is induced by osteoclast attachment to bone: Inhibition by alendronate and calcitonin. J. Clin. Invest. 96, 2277 – 2283 (1995). 97. P. David, H. Nguyen, A. Barbier, and R. Baron, The bisphosphonate tiludronate is a potent inhibitor of the osteoclast vacuolar H-ATPase. J. Bone Miner. Res. 11, 1498 – 1507 (1996). 98. R. Felix, R. G. G. Russell, and H. Fleisch, The effect of several diphosphonates on acid phosphohydrolases and other lysosomal enzymes. Biochim. Biophys. Acta 429, 429 – 438 (1976). 99. R. Felix, J. D. Bettex, and H. Fleisch, Effect of diphosphonates on the synthesis of prostaglandins in cultured calvaria cells. Calcif. Tissue Int. 33, 549 – 552 (1981). 100. K. Ohya, S. Yamada, R. Felix, and H. Fleisch, Effect of bisphosphonates on prostaglandin synthesis by rat bone cells and mouse calvaria in culture. Clin. Sci. 69, 403 – 411 (1985). 101. O. Teronen, Y. T. Konttinen, C. Lindqvist, T. Salo, T. Ingman,
102.
103.
104.
105.
106.
107.
108.
109.
110.
111.
112.
113.
114.
115.
H. Lauhio, S. Santavirta, H. Valleala, and T. Sorsa, Inhibition of matrix matalloproteinase-1 by dichloromethylene bisphosphonate (clodronate). Calcif. Tissue Int. 61, 59 – 61 (1997). O. Teronen, P. Heikkila, Y. T. Konttinen, M. Laitinen, T. Salo, R. Haanemaaijer, A. Teronen, P. Maisi, and T. Sorsa, MMP inhibition and downregulation by bisphosphonates. Ann. N. Y. Acad. Sci. 878, 453 – 465 (1999). A. Schmidt, S. J. Rutledge, N. Endo, E. E. Opas, H. Tanaka, G. Wesolowski, C. T. Leu, Z. Huang, C. Ramachandaran, S. B. Rodan, and G. A. Rodan, Protein-tyrosine phosphatase activity regulates osteoclast formation and function: Inhibition by alendronate. Proc. Natl. Acad. Sci. USA 93, 3068 – 3073 (1996). S. P. Luckman, F. P. Coxon, F. H. Ebetino, R. G. G. Russell, and M. J. Rogers, Heterocycle-containing bisphosphonates cause apoptosis and inhibit bone resorption by preventing protein prenylation: Evidence from structure-activity relationships in J774 macrophages. J. Bone Miner. Res. 13, 1668 – 1678 (1998). S. P. Luckman, D. E. Hughes, F. P. Coxon, R. G. G. Russell, and M. J. Rogers, Nitrogen-containing bisphosphonates inhibit the mevalonate pathway and prevent post-translational prenylation of GTP-binding proteins, including Ras. J. Bone Miner. Res. 13, 581 – 589 (1998). J. E. Fisher, M. J. Rogers, J. M. Halasy, S. P. Luckman, D. E. Hughes, P. J. Masarachia, G. Wesolowski, R. G. G. Russell, G. A. Rodan, and A. A. Reszka, Alendronate mechanism of action: Gernylgeraniol, an intermediate in the mevalonate pathway, prevents inhibition of osteoclast formation, bone resorption, and kinase activation in vitro. Proc. Natl. Acad. Sci. USA 96, 133 – 138 (1999). M. J. Rogers, R. G. Brown, V. Hodkin, R. G. G. Russell, and D. J. Watts, Bisphosphonates are incorporated into adenine nucleotides by human aminoacyl-tRNA synthetase enzymes. Biochem. Biophys. Res. Commun. 224, 863 – 869 (1996). S. Auriola, J. Frith, M. J. Rogers, A. Koivuniemi, and J. Mönkkönen, Identification of adenine nucleotide-containing metabolites of bisphosphonate drugs using ion-pair liquid chromatographyelectrospray mass spectrometry. J. Chromatogr. B 704, 187 – 195 (1997). J. C. Frith, J. Mönkkönen, G. M. Blackburn, R. G. G. Russell, and M. J. Rogers, Clodronate and liposome-encapsulated clodronate are metabolised to a toxic ATP analog, adenosine5 ,-dichloromethylene) triphosphate, by mammalian cells in vitro. J. Bone Miner. Res. 12, 1358 – 1367 (1997). R. Felix, H. L. Guenther, and H. Fleisch, The subcellular distribution of [14C] dichloromethylenebisphosphonate and [14C]1-hydroxyethylidene-1,1-bisphosphonate in cultured calvaria cells. Calcif. Tissue Int. 36, 108 – 113 (1984). M. G. Cecchini, R. Felix, H. Fleisch, and P. H. Cooper, Effects of bisphosphonates on proliferation and viability of mouse bone marrow-derived macrophages. J. Bone Miner. Res. 2, 135 – 142 (1987). M. Sahni, H. L. Guenther, H. Fleisch, P. Collin, and T. J. Martin, Bisphosphonates act on rat bone resorption through the mediation of osteoblasts. J. Clin. Invest. 91, 2004 – 2011 (1993). C. Vitté, H. Fleisch, and H. L. Guenther, Bisphosphonates induce osteoblasts to secrete an inhibitor of osteoclast-mediated resorption. Endocrinology 137, 2324 – 2333 (1996). M. Nishikawa, T. Akatsu, Y. Katayama, Y. Yasutomo, S. Kado, N. Kugai, M. Yamamoto, and N. Nagata, Bisphosphonates act on osteoblastic cells and inhibit osteoclast formation in mouse marrow cultures. Bone 18, 9 – 14 (1996). L. I. Plotkin, R. S. Weinstein, A. M. Parfitt, P. K. Roberson, S. C. Manolagas, and T. Bellido, Prevention of osteocyte and osteoblast apoptosis by bisphosphonates and calcitonin. J. Clin. Invest. 104, 1363 – 1374 (1999).
466 116. S. Adami, A. K. Bhalla, R. Dorizzi, F. Montesanti, S. Rosini, G. Salvagno, and V. Lo Cascio, The acute-phase response after bisphosphonate administration. Calcif. Tissue Int. 41, 326 – 331 (1987). 117. H. Fleisch, R. G. G. Russell, S. Bisaz, R. C. Mühlbauer, and D. A. Williams, The inhibitory effect of phosphonates on the formation of calcium phosphate crystals in vitro and on aortic and kidney calcification in vivo. Eur. J. Clin. Invest. 1, 12 – 18 (1970). 118. M. D. Francis, The inhibition of calcium hydroxyapatite crystal growth by polyphosphonates and polyphosphates. Calcif. Tissue Res. 3, 151 – 162 (1969). 119. A. Jung, S. Bisaz, and H. Fleisch, The binding of pyrophosphate and two diphosphonates by hydroxyapatite crystals. Calcif. Tissue Res. 11, 269 – 280 (1973). 120. B. L. Barnett and L. C. Strickland, Structure of disodium dihydrogen 1-hydroxyethylidene-diphosphonate tetrahydrate: A bone growth regulator. Acta Cryst. B35, 1212 – 1214 (1979). 121. D. Fraser, R. G. G. Russell, O. Pohler, W. G. Robertson, and H. Fleisch, The influence of disodium ethane-1-hydroxy-1,1-diphosphonate (EHDP) on the development of experimentally induced urinary stones in rats. Clin. Sci. 42, 197 – 207 (1972). 122. C. M. T. Plasmans, W. Kuypers, and T. J. J. H. Slooff, The effect of ethane-1-hydroxy-1,1-diphosphonic acid (EHDP) on matrix induced ectopic bone formation. Clin. Orthop. Rel. Res. 132, 233 – 243 (1978). 123. H. R. Mühlemann, D. Bowles, A. Schatt, and J. P. Bernimoulin, Effect of diphosphonate on human supragingival calculus. Helv. Odont. Acta 14, 31 – 33 (1970). 124. A. Larsson, The short-term effects of high doses of ethylene-1hydroxy-1,1-diphosphonates upon early dentin formation. Calcif. Tissue Res. 16, 109 – 127 (1974). 125. Y. Ogawa, Disturbances in enamel and dentin formation of rat incisor following injection with sodium fluoride, strontium(II) chloride and EHDP. Jpn. J. Oral Biol. 22, 199 – 226 (1980). 126. E. van Beek, M. Hoekstra, M. van de Ruit, C. Löwik, and S. Papapoulos, Structural requirements for bisphosphonate actions in vitro. J. Bone Miner. Res. 9, 1875 – 1882 (1994). 127. E. van Beek, C. Löwik, I. Oue, and S. Papapoulos, Dissociation of binding and antiresorptive properties of hydroxybisphosphonates by substitution of the hydroxyl with an amino group. J. Bone Miner. Res. 11, 1492 – 1499 (1996). 128. M. D. Francis, L. Flora, and W. R. King, The effects of disodium ethane-1-hydroxy-1,1-diphosphonate on adjuvant induced arthritis in rats. Calcif. Tissue Res. 9, 109 – 121 (1972). 129. C. J. Dunn, L. A. Galinet, H. Wu, R. A. Nugent, S. T. Schlachter, N. D. Staite, D. G. Aspar, G. A. Elliott, N. A. Essani, N. A. Rohloff, and R. J. Smith, Demonstration of novel anti-arthritic and antiinflammatory effects of diphosphonates. J. Pharmacol. Exp. Ther. 266, 1691 – 1698 (1993). 130. P. L. E. M. van Lent, L. A. M. van den Bersselaar, A. E. M. Holthuyzen, N. van, Rooijen, L. B. A. van de Putte, and W. B. van den Berg, Phagocytic synovial lining cells in experimentally induced chronic arthritis: Down-regulation of synovitis by CL2MDPliposomes. Rheumatol. Int. 13, 221 – 228 (1994). 131. T. Österman, K. Kippo, L. Laurén, R. Hannuniemi, and R. Sellman, Effect of clodronate on established adjuvant arthritis. Rheumatol. Int. 14, 139 – 147 (1994). 132. C. P. Ciosek, D. R. Magnin, T. W. Harrity, J. V. H. Logan, J. K. Dickson, E. M. Gordon, K. A. Hamilton, K. G. Jolibois, L. K. Kunselman, R. M. Lawrence, K. A. Mookhtiar, L. C. Rich, D. A. Slusarchyk, R. B. Sulsky, and S. A. Biller, Lipophilic 1,1-bisphosphonates are potent squalene synthase inhibitors and orally active cholesterol lowering agents in vivo. J. Biol. Chem. 268, 24832 – 24837 (1993). 133. J. H. Lin, Bisphosphonates: A review of their pharmacokinetic properties. Bone 18, 75 – 85 (1996).
H. FLEISCH 134. S. E. Papapoulos, Pharmacodynamics of bisphosphonates in man; implications for treatment. In “Bisphosphonate on Bones” (O. L. M. Bijvoet, H. A. Fleisch, R. E. Canfield, and R. G. G. Russell, eds.), pp. 231 – 263. Elsevier, Amsterdam, 1995. 135. A. G. Porras, S. D. Holland, and B. J. Gertz, Pharmacokinetics of alendronate. Clin. Pharmacokinet. 36, 315 – 328 (1999). 136. R. R. Recker and P. D. Saville, Intestinal absorption of disodium ethane-1-hydroxy-1,1-diphosphonate (disodium etidronate) using a deconvolution technique. Toxicol. Appl. Pharmacol. 24, 580 – 589 (1973). 137. X. Boulenc, E. Marti, H. Joyeux, C. Roques, Y. Berger, and G. Fabre, Importance of the paracellular pathway for the transport of new bisphosphonate using the human CACO-2 monolayers model. Biochem. Pharmacol. 46(9), 1591 – 1600 (1993). 138. B. J. Gertz, S. D. Holland, W. F. Kline, B. K. Matuszewski, A. Freeman, H. Quan, K. C. Lasseter, J. C. Mucklow, and A. G. Porras, Studies of the oral bioavailability of alendronate. Clin. Pharmacol. Ther. 58, 288 – 298 (1995). 139. W. H. Wiedmer, A. M. Zbinden, U. Trechsel, and H. Fleisch, Ultrafiltrability and chromatographic properties of pyrophosphate, 1-hydroxyethylidene-1,1 bisphosphonate, and dichloromethylenebisphosphonate in aqueous buffers and in human plasma. Calcif. Tissue Int. 35, 397 – 400 (1983). 140. J. H. Lin, I.-W. Chen, and F. A. De Luna, Nonlinear kinetics of alendronate: Plasma protein binding and bone uptake. Drug Metab. Dispos. 22, 400 – 405 (1994). 141. F. Wingen, and D. Schmähl, Distribution of 3-amino-1-hydroxypropane-1,1-diphosphonic acid in rats and effects on rat osteosarcoma. Arzneimittelforschung 35, 1565 – 1571 (1985). 142. S. Bisaz, A. Jung, and H. Fleisch, Uptake by bone of pyrophosphate, diphosphonates and their technetium derivatives. Clin. Sci. Mol. Med. 54, 265 – 272 (1978). 143. P. Masarachia, M. Weinreb, and G. A. Rodan, Comparison of the distribution of 3H-alendronate and 3H-etidronate in rat and mouse bones. Bone 19, 281 – 290 (1996). 144. J. Mönkkönen, H. M. Koponen, and P. Ylitalo, Comparison of the distribution of three bisphosphonates in mice. Pharmacol. Toxicol. 66, 294 – 298 (1990). 145. G. B. Kasting and M. D. Francis, Retention of etidronate in human, dog and rat. J. Bone Miner. Res. 7, 513 – 522 (1992). 146. U. Troehler, J. P. Bonjour, and H. Fleisch, Renal secretion of diphosphonates in rats. Kidney Int. 8, 6 – 13 (1975). 147. J. H. Lin, I.-W. Chen, A. Florencia, A. Deluna, and M. Hichens, Renal handling of alendronate in rats: An uncharacterized renal transport system. Drug Metab. Dispos. 20, 608 – 613 (1992). 148. C. L. Alden, R. D. Parker, and D. F. Eastman, Development of an acute model for the study of chloromethanediphosphonate nephrotoxicity. Toxicol. Pathol. 17, 27 – 32 (1989). 149. J. C. Cal and P. T. Daley-Yates, Disposition and nephrotoxicity of 3-amino-1-hydroxypropylidene-1,1-bisphosphonate (APD) in rats and mice. Toxicology 65, 179 – 197 (1990). 150. P. C. De Groen, D. F. Lubbe, L. J. Hirsch, A. Daifotis, W. Stephenson, D. Freedholm, S. Pryor-Tillotson, M. J. Seleznick, H. Pinkas, and K. K. Wang, Esophagitis associated with the use of alendronate. N. Engl. J. Med. 355, 1016 – 1021 (1996). 151. C. P. Peter, K. K. Handt, and S. M. Smith, Esophageal irritation due to alendronate sodium tablets. Digest. Dis. Sci. 43, 1998 – 2002 (1998). 152. C. P. Peter, M. V. Kindt, and J. A. Majka, Comparative study of potential for bisphosphonates to damage gastric mucosa of rats. Digest. Dis. Sci. 43, 1009 – 1015 (1998). 153. W. R. King, M. D. Francis, and W. R. Michael, Effect of disodium ethane-1-hydroxy-1,1-diphosphonate on bone formation. Clin. Orthop. 78, 251 – 270 (1971).
CHAPTER 16 Basic Biology of Bisphosphonates 154. M. Eguchi, T. Yamaguchi, E. Shiota, and S. Handa, Fault of ossification and calcification and angular deformities of long bones in the mouse fetuses caused by high doses of ethane-1-hydroxy-1, 1-diphosphonate (EHDP) during pregnancy. Cong. Anom. 22, 47 – 52 (1982). 155. P. Graepel, P. Bentley, H. Fritz, M. Miyamoto, and S. R. Slater, Reproduction toxicity studies with pamidronate. Arzneim.-Forsch. Drug Res. 42, 654 – 667 (1992). 156. A. Okazaki, T. Matsuzawa, M. Takeda, R. G. York, P. C. Barrow, V. C. King, and G. P. Bailey, Intravenous reproductive and develop-
467 mental toxicity studies of cimadronate (YM175), a novel bisphosphonate, in rats and rabbits. J. Toxicol. Sci. 20, 1 – 13 (1995). 157. A. Okazaki, H. Sakai, T. Matsuzawa, C. J. Perkin, and P. W. East, Intravenous single and repeated dose toxicity studies of cimadronate (YM175), a novel bisphosphonate, in rats. J. Toxicol. Sci. 20, 15 – 26 (1995). 158. A. Okazaki, T. Matsuzawa, C. J. Perkin, and M. H. Barker, Intravenous single and repeated dose toxicity studies of cimadronate (YM175), a novel bisphosphonate, in beagle dogs. J. Toxicol. Sci. 20, 27 – 36 (1995).
CHAPTER 17
Skeletal Development Mechanical Consequences of Growth, Aging, and Disease MARJOLEIN C. H. VAN DER MEULEN
DENNIS R. CARTER AND GARY S. BEAUPRÉ
Sibley School of Mechanical and Aerospace Engineering, Cornell University, Ithaca, New York 14853; and Biomechanics and Biomaterials, Hospital for Special Surgery, New York, New York 10021 Department of Mechanical Engineering, Biomechanical Engineering Program, Stanford University, Stanford, California 94305; and Rehabilitation Research and Development Center, Veterans Affairs Palo Alto, Palo Alto, California 94304
I. Developmental Mechanics in Skeletogenesis II. Mechanical Regulation of Bone Biology III. Mechanobiologic Self-Design of Bones
IV. Adaptational Mechanics in Aging and Disease References
I. DEVELOPMENTAL MECHANICS IN SKELETOGENESIS
the normal adult skeleton is “inherently” well designed from a mechanical point of view. Changes in bone tissue quality and/or quantity as a result of aging or disease then diminish the normal mechanical integrity of the skeleton, thereby increasing the risk of fracture. Mechanical deficits in the osteoporotic skeleton are thus often assessed with respect to a “normal” young or age-matched control group. The approach taken in this chapter is a bit different. We believe that the mechanics of the osteoporotic skeleton can best be understood when one appreciates the role of mechanics in skeletal development. We consider the mechanical integrity of the skeleton at any age to be a reflection of intrinsic genetic factors and the entire prior life history of mechanical and chemical epigenetic events. These events include numerous factors that are related to hormones, diet, and physical activity. Biomechanical
The bones of the adult skeleton are well designed for supporting the forces that are created during normal physical activities. The tubular shape of the diaphyses of long bones is ideal for withstanding the bending and torsional loads imposed on the bone shaft. Bone tissue at the ends of long bones and in short bones serves to support and distribute joint contact forces. The intricate architecture of the cancellous bone in these regions is well suited for this task. Indeed, the form and internal architecture of the entire skeleton are exquisitely matched to its mechanical function. Biomechanical considerations of the osteoporotic skeleton usually concentrate on the material and structural changes that compromise its mechanical integrity and thus can lead to bone fracture. Such an approach assumes that
OSTEOPOROSIS, SECOND EDITION VOLUME 1
471
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
472 changes in osteoporosis can then be directly linked to the physicochemical events that precede the disease and continue to affect the skeleton during the progression and treatment of the disease. Mechanical regulation of bone biology begins very early. At approximately 5 to 7 weeks of prenatal life, most of the skeletal elements, muscles, tendons, and ligaments characteristic of the adult have formed. Involuntary contractions of the newly formed muscle fibers commence and ossification is initiated in the cartilaginous endoskeleton. The intermittently imposed skeletal tissue stresses, deformations, and motions caused by muscular contractions then play an increasingly important role in modulating cartilage growth, ossification, and bone modeling and remodeling throughout the skeleton except for the cranium. By 15 weeks, all of the basic movements characteristic of full-term newborn infants can be observed [1]. The inhibition of muscular contractions and movements in utero results in abnormally low skeletal mass and strength (Fig. 1) [2 – 4]. After birth, further growth and ossification of the skeleton continues to be strongly influenced by physical activity and externally applied forces [5]. Long bones such as the femur begin to ossify when the primary bone collar appears at the midshaft of the cartilage anlage where the chondrocytes have reached a hypertrophic
VAN DER
state. Some have speculated that the hypertrophy of the chondrocytes at the midshaft may result in the release of chemical factors that induce osteogenesis in the perichondrium. Vascular invasion of the hypertrophic cartilage inside the primary bone collar results in a transient stage of endochondral bone formation followed by osteoclastic resorption. The medullary canal and endosteal surface are established. The entire anlage continues to grow in length by the proliferation and ossification of cartilage. Further growth and development of the diaphyseal cross section are achieved by direct bone apposition and resorption on the periosteal and endosteal surfaces, respectively. When one examines the structure of a typical long bone, the bone in different regions is associated with different ontogenetic processes (Fig. 2). A significant portion of the compact bone in the diaphysis has a developmental history that includes initial appositional bone formation. Extending from the center of the bone toward both of the bone ends, however, are conical regions that include the cancellous bone of the metaphyses and epiphyses. The bone in these areas was initially formed by endochondral ossification. With increasing age, secondary bone remodeling throughout the skeleton will progressively diminish the distinctions associated with the primary bone formed in the different regions. In considering the influence of mechanical factors on bone development, appositional bone
FIGURE 2 FIGURE 1 Radiographs of tibiae from (a) normal infant and (b) newborn with spinal muscular atrophy. Used with permission from Rodriguez et al., Calcif. Tissue Int. 43, 335 – 339 (1988).
MEULEN, CARTER, AND BEAUPRÉ
Schematic representation of long bone growth. The shaded regions on the right-hand figure are areas of appositional growth. Metaphyseal and epiphyseal cancellous bone is formed by endochondral ossification, shown in the white conical regions and bone ends. Used with permission from Carter et al., Bone 18 (Suppl. 1), 5S – 10S (1996).
473
CHAPTER 17 Skeletal Development
formation must be considered as a different process from endochondral bone formation. The subsequent bone modeling and remodeling in these two regions, however, are similar.
II. MECHANICAL REGULATION OF BONE BIOLOGY Wolff [7] wrote extensively on the relationship between physical forces and bone structure. He was influenced by the work of Wilhelm Roux [8], who was interested not only in the morphology of tissues and organ systems but also the mechanisms responsible for the development of specific morphological features. Roux was convinced that physical forces play a major role in development. He referred to the processes by which physicochemical factors regulate development as “Entwicklungsmechanik” or “developmental mechanics” [8]. Wolff stated, “Roux, as I do myself, distinguishes two periods in the life of every organism. One is embryonic. During this period the ‘organs expand, differentiate and grow.’ The other period is adulthood. During this period, growth and replacement of what is worn out takes place ‘only when stimulated.’ . . . ” (translation by Maquet and Furlong, 1986). In the writings of Roux and Wolff we see the seeds of a fundamental question regarding the relative importance of biological and mechanical regulation of skeletogenesis. Building on these concepts first introduced a century ago, we have developed a theory for bone adaptation in which biologic factors play an important role only in the initial phases of skeletal development and their influence diminishes over time [9,10]. Mechanobiologic influences, however, remain a fundamental influence on bone apposition and resorption throughout life. In this theory we describe the intensity of bone tissue mechanical loading in terms of a daily stress (or strain)
FIGURE 3
stimulus, b, that takes into account both the magnitude and the number of cycles of loading applied during daily activities [11]. For example, we might consider the daily stress stimulus for a nonathletic individual to consist of contributions from walking, stair climbing, and rising from a chair. For an individual who is athletic, we might include additional contributions from jogging, bicycling, running, and so on. In mathematical terms we define the daily stress stimulus as
b
n i b
m
1/m
,
(1)
day
where ni is the number of cycles of each load type i, b is a measure of stress intensity within the bone tissue, and the stress exponent, m, is an empirical constant. The stress exponent can be thought of as a weighting factor for the relative importance of the stress magnitude and the number of load cycles. For m 1, the stress magnitude and the number of load cycles are equally important. For m 1, those activities having high stress magnitudes would contribute more to the total stimulus. Alternatively, for m 1, those activities that are repeated many times each day would be relatively more important. Experimental data suggest that m lies in the range of 3 to 8, indicating that the magnitude of the cyclic stress is more important than the number of loading cycles [12]. We assume that if the imposed daily stress stimulus is greater than some target level of stimulus, the attractor stress stimulus (AS), then bone apposition will occur. If the imposed daily stress stimulus is less than the attractor stimulus, bone will be resorbed. The stress stimulus error, e, is the difference between the daily stress stimulus and the attractor stress stimulus, and is the driving force for bone adaptation. The block diagram shown in Fig. 3 is a schematic representation of our bone development and adaptation theory with the linear apposition/resorption rate, r. used as the feedback
Block diagram representation of bone remodeling having multiple feedback loops. Adapted from Beaupré et al., J. Orthop. Res. 8, 651 – 661 (1990).
474
VAN DER
MEULEN, CARTER, AND BEAUPRÉ
FIGURE 4
Simplified block diagram representation assuming that local nonstress effects (see Fig. 3) do not occur and that the bone attractor stress stimulus does not change appreciably due to formation or resorption. Adapted from Beaupré et al., J. Orthop. Res. 8, 651 – 661 (1990).
parameter. In this particular block diagram two feedback loops are shown. The upper feedback loop implies that bone adaptation and factors related to metabolic status, genotype, and local tissue interactions may influence the attractor stress stimulus, as well as the local tissue response. The lower feedback loop illustrates the interaction between bone adaptation and purely mechanical factors. By assuming that neither the attractor state stimulus nor the local tissue response changes during the course of bone adaptation, the upper loop in Fig. 3 can be eliminated (Fig. 4). In this representation we have divided the lower, mechanical feedback loop into two parallel paths: one path corresponding to changes in geometry (typically cortical changes) and the other path corresponding to changes in apparent bone density (typically cancellous changes). To represent the time-dependent nature of bone adaptation, we must establish a quantitative relationship between the stress stimulus error and the rates of bone apposition and resorption. We believe, as others do, that the rate relationship that describes the bone response to a given remodeling error is nonlinear [13 – 15]. Specifically, we think that when the stress stimulus error is within a range near zero (i.e., within the normal activity range) the rate of net bone apposition or resorption will be small. When the remodeling error is outside this range, however, the rates of bone apposition and resorption can increase dramatically. Three hypothetical rate relationships are shown in Fig. 5. The curve labeled “1” might represent the skull, which is shown having a lower attractor state stress and a relative insensitivity to unloading and bone resorption. The curve
labeled “2” might represent the periosteum of the femur, with a low sensitivity for bone resorption with unloading and a high sensitivity for bone apposition with increased loading. Finally, the curve labeled “3” might represent the endosteum of the femur, with a higher sensitivity for resorption than for apposition. These differences among the three curves could be related to local tissue interactions and cell populations associated with the different bone surfaces in question. In the following sections, this approach for describing the regulation of bone biology is presented in more detail and implemented for particular skeletal elements.
FIGURE 5 Hypothetical curves for three bone regions showing the rate of surface response as a function of the tissue level stress stimulus. See text for further discussion. Adapted from Beaupré et al., J. Orthop. Res. 8, 651 – 661 (1990).
475
CHAPTER 17 Skeletal Development
III. MECHANOBIOLOGIC SELF-DESIGN OF BONES Development and subsequent growth involve the coordinated change of shape, size, and material. The conceptual approach presented in Section II directly relates the in vivo mechanical loading environment to these changes during the development, growth, and adaptation of the skeleton. Using mathematical implementations on computers, we can model the various skeletal elements that are influenced by mechanics to simulate bone appositional and endochondral growth. The following sections first show how mechanics guides modeling changes in the long bone diaphysis during growth and development. Thereafter, a parallel development is demonstrated for the density and morphology changes in trabecular bone sites. The same fundamental relationships are used in both cases. Osteoporotic changes caused by alterations in bone loading are also simulated with the same models used for development.
A. Diaphyseal Compact Bone 1. DEVELOPMENT AND ADAPTATION In mammals the ossification of the long bones initiates with endochondral ossification and the formation of the primary bone collar at the middiaphysis of the cartilage anlage. At this time during development, fetal muscle contractions also commence [1]. Thereafter, the primary ossification center forms and the processes of endochondral ossification and direct bone apposition begin. Whereas the initial bone collar appears without the stimulation of fetal muscle contractions, further normal development and growth of the diaphysis are dependent on the mechanical loading environment created by these forces [2 – 4,16 – 18]. Therefore, the radial growth of the diaphysis may be considered as a combination of intrinsic biologic growth and mechanically regulated biologic (“mechanobiologic”) processes. Using this fundamental concept, we have developed an analytical model to simulate the roles of biological and mechanobiological factors in the development of the crosssectional geometry of the human femur [10]. We modeled the long bone diaphysis as a circular cross section defined by a periosteal (outer) and an endosteal (inner) radius. These sections were “grown” using bone surface apposition rates determined from underlying biologic growth and mechanically regulated biologic stimuli. Both growth processes were assumed to be functions of time and the particular bone surface. This model was used to simulate development under normal and altered loading conditions as well as adaptation to increased and decreased loading in the adult. To model intrinsic biological processes, we assumed that purely biological factors play a significant role in early
cross-sectional development and that their contribution gradually diminishes with time, becoming negligible in the later half of maturation. Biological factors were modeled as a periosteal surface apposition rate, which was a decaying exponential function of age. In our model of human femoral development, the biological growth rate decayed to zero by 6 years of age (Fig. 6). Mechanically regulated surface bone growth rates were calculated for the periosteum and endosteum based on the daily stress stimulus, [Section II, Eq. (1)], and added directly to the biological rate. Whereas the contribution of purely biological growth processes diminishes with time, extrinsic mechanical influences on long bone cross-sectional growth are fundamental processes that remain active throughout an individual’s lifetime. In this model, the mechanobiologic stimulus is the sole regulator of long bone cross-sectional geometry once the biologic contribution has disappeared. Using the daily stress stimulus to describe the cyclic in vivo load history, the stress stimulus attractor was chosen consistent with experimental data from the literature [19]. In vivo experimental studies using strain gages bonded to adult bone surfaces have shown that the magnitude of bone strains created during physical activity is similar across different bones in a variety of animals over several orders of body mass [20]. Similar constant peak strain levels have also been measured in growing animals [21 – 23]. In our model, therefore, the stress stimulus attractor was assumed to be identical at all diaphyseal locations and constant throughout life. The surface modeling rate – stress stimulus relationship (see Fig. 5) used to model the mechanobiological responses included a “lazy zone” in the region near the stress stimulus attractor. The relationship between the bone stress stimulus and the surface apposition rates was modeled differently on the periosteum and endosteum, reflecting different tissue
FIGURE 6
Periosteal apposition rate used to simulate intrinsic biologic growth. Used with permission from van der Meulen et al., Bone 14, 635 – 642 (1993).
476 properties and interactions on the two surfaces [24 – 27]. The bone apposition rates were implemented identically for the periosteum and endosteum; there was no resorptive response allowed on the periosteum. A symmetric stimulus – rate relationship was modeled on the endosteum. The width of the lazy zone was chosen to be 20% of the attractor stimulus value. This value was based on the bone’s mature thicknessto-radius ratio. Values of the apposition rates were based on experimental measurements of diaphyseal changes with aging [28]. To implement this model, one must apply an assumed mechanical loading that changes as a function of age. In vivo strain gage studies have shown that the surface stresses in the weight-bearing long bones are primarily longitudinal normal and shear stresses from combined bending and torsional moments 1 [22,29,30]. These in vivo moments are produced by actions of the muscles on the skeleton and were assumed to scale in proportion to muscle mass [10]. Muscle mass is approximately proportional to body mass in adult mammals [31]. Assuming that torsional moments are directly proportional to body mass, we obtained human body mass data during growth and aging and used this to simulate the age-dependent in vivo loading history (Fig. 7). For modeling simplicity, we applied an axisymmetric torsional moment as a representative loading history. The stress distribution determines the cross-sectional morphology; torsional and bending moments both produce stresses that increase linearly from the inner surface to the outer surface in a cylindrical structure with constant material properties. The magnitude of the moments at maturity (age 20 years) was calculated as that which produced a bone stress stimulus at the corner of the lazy zone. A strength of materials analysis for a hollow circular cylinder was used to determine the stress stimulus on the periosteum and endosteum once the load was applied to the bone cross section. A normal developmental loading history was applied to the model, and cross-sectional morphologies were developed with time. For each simulation we calculated the following parameters: the periosteal and endosteal radii, surface apposition/resorption rates and stress stimuli, cortical area, polar moment of inertia, section modulus, and ratio of bone thickness to periosteal radius. The simulation values were compared with experimental measurements of human femoral cross-sectional morphology. Starting from an initial bone collar, the biological growth rate alone (without any mechanical sensitivity) produced a cross section completely dependent on the magnitude of the intrinsic growth rate. These results are shown in Fig. 8a. As implemented, only periosteal apposition was present and periosteal expansion ceased after 6 years. When both biologic
VAN DER
FIGURE 7
Torsional moment applied to simulate mechanobiological growth. The corresponding body mass at each time point is indicated on the right-hand scale. Used with permission from van der Meulen et al., Bone 14, 635 – 642 (1993).
and mechanobiologic responses were implemented (Fig. 8b), a rapid expansion of both the periosteum and the endosteum occurred during development and subsequent growth. The dimensional increases stabilized at maturity, and thereafter a gradual age-related expansion and thinning of the cortex occurred throughout the remainder of life. Comparison of the simulation radius values with those measured by other researchers shows a very good correspondence (Fig. 9) [32 – 34]. Although the loading is constant throughout maturity, gradual, age-related periosteal expansion occurs and results in a diminution of the cortical area and an increase in the polar moment of inertia. Numerous studies have measured this continuing subperiosteal expansion with age [28,34 – 37] and showed similar area and moment of inertia results. When a moment is applied to a beam, the resulting surface stresses are proportional to the applied moment divided by the section modulus, a cross-sectional shape parameter. For a given cross-sectional geometry, the section modulus is defined as the polar moment of inertia divided
FIGURE 8 1 A “moment” results from a force applied at a distance from the point of interest.
MEULEN, CARTER, AND BEAUPRÉ
Simulation results for normal cross-sectional growth of the human femur. (a) Biologic growth only and (b) biologic and mechanobiologic growth. Used with permission from van der Meulen et al., Bone 14, 635 – 642 (1993).
477
CHAPTER 17 Skeletal Development
FIGURE 9 Simulation results for periosteal and endosteal radii over time. Experimental data shown from McCammon ([33] mean data), Smith and Walker ([34] mean data), and Martin and Atkinson ([32] individual data points). Used with permission from van der Meulen et al., Bone 14, 635 – 642 (1993). by the periosteal diameter. If the moments increase with increasing body mass, as we have assumed, then the section modulus must also increase in the same fashion if the stresses are to remain constant. Our mechanobiological model reflects this adaptation, and the section modulus increases nearly linearly with body mass during development while the relationship with age is more complex (Fig. 10). Clinical measurements made in our laboratory [38 – 41] reveal a poor linear relationship between section modulus and age in adolescents and young adults (Fig. 11). When the same data are plotted against body mass, however, a very strong linear relationship is evident (Fig. 12). Further implications of the section modulus are discussed in Section III,A,2. Once the validity of this model was established for examining normal growth and development, we used the
FIGURE 10
model to examine skeletal ontogeny with mechanical loading reduced to 40% of normal [42]. Functional adaptation in the normal adult was simulated by altering the loading at maturity (20 years of age) in two ways: (i) a 60% decrease in load magnitudes and (ii) a 25% increase over normal load levels. In these analyses, only the load levels were altered; all other models parameters were maintained at their normal values. For all cases, the periosteal and endosteal radius values were calculated and compared to the results for normal development (Fig. 13). The trends represented in the results were also compared to experimental results by others under qualitatively similar conditions. Reduction of the normal loading history by 60% throughout the lifetime of the individual produces an overall diminished cross section. The adult periosteal radius was reduced 25% compared to normal, and the endosteal radius was approximately 80% of the normal value. The thickness-to-radius ratio of the sections was only slightly reduced; however, the section modulus was reduced approximately 40% as a result of the periosteal radius reduction (Fig. 14). These results cannot be fully experimentally validated, but are similar to those observed in growing animals with reduced skeletal loading [43 – 48] and are consistent with clinical observations of children born with neuromuscular defects [3,16,49]. The rapid reduction of loading after reaching maturity results in extreme cortical thinning through arrested periosteal growth and increased endosteal expansion. Although the overall cross-sectional dimensions were much greater than when the loading was reduced throughout development, the resulting cross-sectional strength changes were similar for the two cases; the section moduli were nearly equally reduced (Fig. 14). These simulations may be compared to clinical results of cortical bone adaptation following spinal cord injury. Experimental measurements of
Normal increase in femoral diaphysis section modulus with age and body mass predicted by our theoretical model [10]. Used with permission from Carter et al., Bone 18 (Suppl. 1), 5S – 10S (1996).
478
VAN DER
MEULEN, CARTER, AND BEAUPRÉ
FIGURE 11
Section modulus plotted against age for Caucasian adolescents (r2 0.51). Male and female regressions are significantly different. Used with permission from van der Meulen et al., J. Orthop. Res. 14, 22 – 29 (1996).
diaphyseal bone mineral content (BMC) have been somewhat mixed, but several studies have shown decreased BMC [50 – 52] and increased fracture rates at femoral and tibial cortical sites [53,54]. An abrupt increase of applied loading at maturity produced approximately equivalent increases (6 – 7%) in endosteal and periosteal diameters and a 20% increase in the section modulus. These changes are consistent with increased bone strength observed in animal studies [55,56]. Direct comparison to clinical data is difficult for various reasons. Human exercise studies have been unable to produce bone hypertrophy with consistency and have reported increased, decreased, and unaltered bone mass. When increased bone density is
FIGURE 13
Effects of changes in bone loading during life on cross-sectional dimensions of the femoral diaphysis predicted by our theoretical models. Adapted from van der Meulen, Ph.D. thesis, Stanford University.
present, the changes are very modest, generally ranging from 0.5 to 3% [57]. Because the loading magnitudes are difficult to quantify, the true experimental levels are unknown and may be less than those that have been theoretically assumed. Finally, our simulations represent prolonged, sustained increased loading, and most in vivo studies follow the subjects for shorter periods. 2. MATERIAL AND STRUCTURAL STRENGTH
FIGURE 12 Regression of section modulus on body mass for Caucasian adolescents (r2 0.86). No significant effect of gender. Used with permission from van der Meulen et al., J. Orthop. Res. 14, 22 – 29 (1996).
In addition to mineral metabolism functions, the long bones of the skeleton primarily perform a structural role supporting our body mass and enabling locomotion. The structural response of a long bone to an applied force is a function of its material (or tissue) properties and geometry, and thus we need to examine these properties individually. The cortices of the long bones consist of dense cortical bone. Cortical bone is a transversely isotropic material with
479
CHAPTER 17 Skeletal Development
diaphyseal structural behavior. For example, the governing equation for the torsion of a hollow circular cylinder results in the following expression for the applied torsional moment, T, T
FIGURE 14 Effects of changes in bone loading on the section modulus of the femoral diaphysis predicted by our theoretical models.
a longitudinal elastic modulus that is nearly 50% greater than the transverse modulus [58]. Cortical bone is stronger in compression than in tension for both longitudinally and transversely applied loads and is weakest in shear loading (induced by torsion about the longitudinal axis) [58]. The material properties of bone tissue are affected by a variety of intrinsic and extrinsic factors. Contributing intrinsic properties include the bone microstructure, porosity, degree of mineralization, and density. In addition, age, race, hormones, and diet are determining factors. In the normal adult, the degree of mineralization, porosity, and density of cortical bone change relatively little. From childhood to maturity (age 8 to 26 years), the ash content of human femoral cortical bone increases only 6%, while its material strength increases 12.5% [59]. Femoral midshaft density increases 4.5% between adolescence and young adulthood [60]. In addition, no difference in bone material properties has been observed between males and females during the same period [61]. Because the material properties of cortical bone change little during postnatal growth, changes in femoral diaphyseal structural behavior are dominated by geometric changes. In the adult, when both material properties and geometry are relatively stable, differences in cortical structural strength and stiffness are also generally attributable to subtle geometric variations. Long bone cross-sectional geometry is fairly complex and varies along the bone length. Closed-form solutions to mechanical analyses, in contrast, are limited to known geometries and require many simplifying assumptions when applied to skeletal structures. In general, long bones are modeled as hollow, prismatic tubes of either a circular or a similar elliptical cross section [62 – 64]. Simple geometric relationships indicative of bone structural behavior can be derived from the governing equations for different loading conditions and used to understand
J , r
(2)
where J is the polar moment of inertia of the circular cross section, r is the outer (periosteal) radius of the tube, and is the maximum shear stress of bone tissue. Assuming that bone material properties are relatively constant, we expect that the torsional moment that can be withstood by a bone with cross-sectional properties r and J is proportional to the geometric term J/r, which is directly proportional to the section modulus, Z. The section modulus is defined as Z
J , D
(3)
where D is the periosteal diameter of the circular cross section (twice the radius). An analogous analysis for bending demonstrates that the applied force during bending is also directly proportional to the section modulus. Therefore, the cross-sectional morphology of a long bone is critical in determining its structural behavior. This result is confirmed by experimental results, which show a very strong correlation between bending strength and section modulus (Fig. 15). Taking these concepts one step further, Selker and Carter [65] defined the “whole bone strength index,” SB, for torsion as SB
J , DL
(4)
where L is the bone length. This index is proportional to the ultimate force required to fracture a long bone when it is
FIGURE 15
Linear regression of whole bone bending strength on section modulus for the human radius. Adapted from Martin and Burr, J. Biomech. 17, 195 – 201 (1984). The sample indicated by the square was omitted from linear regression because it was from the youngest individual and failed in a different mode than the others.
480 held by its ends and a torsional or transverse force is applied to the midshaft. This expression highlights the two important geometric aspects of whole bone strength: the cross-sectional resistance (indicated by J/D) and the bone length. To increase bone strength, there is a direct correspondence with the section modulus and an inverse relationship with bone length. Clinically, bone cross-sectional morphology is difficult to measure noninvasively and can be obtained only by tomographic techniques, which have several drawbacks and are not commonly used. Absorptiometric methods for measuring bone mass [e.g., single or dual photon absorptiometry, dual-energy X-ray absorptiometry (DXA)], however, can measure bone width and total bone mineral content only in the scan plane. The cross-sectional bending moment of inertia of the bone mineral can be directly determined for a plane perpendicular to the scan direction by integrating the absorption curve. With some assumptions similar to those presented earlier for determining whole bone structural behavior, it is possible to estimate the cross-sectional area, area moments of inertia, and section modulus of a cortical cross section [66,67]. The long bone diaphysis is modeled as a hollow circular tube with an outer diameter equal to the bone width measured in the scan plane, and the mineral density, porosity, and degree of mineralization are assumed [68]. Studies examining the relationships between cortical bone strength and the linear bone mineral density (BMC in g/cm) measured by projected radiography demonstrate a good correlation between whole bone strength and in vivo BMC at cortical sites (Fig. 16) [66,69]. These noninvasive imaging techniques have primarily been applied to cancellous bone sites in an attempt to predict fracture risk, which is discussed in the following sections (see Section II,B,2) and in other chapters of this book.
FIGURE 16
Ultimate bending moment regressed on bone mineral content at the fracture site for canine radii, ulnae, and tibiae. Adapted from Borders et al., Biochem. Eng. 99, 40 – 44 (1977).
VAN DER
MEULEN, CARTER, AND BEAUPRÉ
B. Cancellous Bone 1. DEVELOPMENT AND ADAPTATION As noted earlier, the development of cancellous bone in the metaphysis and epiphysis is inextricably tied to the process of endochondral ossification. It is useful, therefore, to understand how endochondral ossification proceeds in the developing skeleton. In mammals, endochondral ossification commences in the central area of the cartilage anlage shortly after the primary bone collar forms. This region of primary endochondral ossification then expands and progresses toward the bones ends, establishing primary ossification fronts in both directions. The cartilage directly ahead of each ossification front exhibits the characteristic feature of interstitial cartilage growth in which the chondrocytes undergo proliferation, maturation, and hypertrophy prior to ossification. Endochondral growth and ossification can proceed without local tissue mechanical loading, provided that the biological environment is appropriate to support bone formation. However, cyclic mechanical stresses (strains) caused by physical activity provide a complex history of physical stimuli throughout the cartilage tissue of the anlage. As a result of the stress distributions created, the cartilage growth and ossification process will be accelerated in some areas and retarded in others [70 – 72]. Mechanical influences on the rate of endochondral ossification are responsible for establishing the geometry of the ossification fronts, the appearance of secondary ossification centers, and the geometry of the growth plates. The cartilage loading histories at the bone ends are also responsible for the stabilization of the subchondral growth front at skeletal maturity and, therefore, are a key factor in establishing the thickness of the articular cartilage covering the joint surfaces [73]. Later in life, articular cartilage stresses and strains play a critical role in the pathogenesis of osteoarthrosis, which can be viewed as the final stage of endochondral ossification in the cartilage anlage [70 – 72]. Cyclic tissue stresses not only regulate the process of endochondral ossification, but also directly influence the organization and remodeling of the cancellous bone initially formed at ossification fronts. The bone formed is immediately exposed to cyclic stresses that are a result of the physical activity of the fetus and, later, of the child. Cancellous bone tissue remodels toward the attractor stress stimulus by increasing or decreasing the local bone apparent density (which is inversely related to porosity) while at the same time adjusting the local trabecular orientation. In effect, bone apposition and resorption take place on the surfaces of trabeculae rather than on the periosteal and endosteal surfaces as in the appositional bone modeling of the diaphysis. The bony architecture of the proximal femur and the changes that occur to that architecture due to altered loads
481
CHAPTER 17 Skeletal Development
received considerable attention in the mid- and late 19th century [7,8,74]. The proximal femur continues to attract the interest of both clinicians and researchers for a number of reasons, including osteoporosis and the risk of hip fracture, as well as the challenges for improving the longevity of hip arthroplasty. Figure 17a shows an anteroposterior radiograph of the proximal femur. Key architectural features include the primary load-bearing trabecular system and the secondary or arcuate trabecular system within the femoral head and the region of low density bone near the center of the femoral neck referred to as Ward’s triangle. The bone slice in Fig. 17b shows additional features, including the detailed arrangement of trabecular struts in the metaphyseal and more proximal regions and the dense cortices and hollow medullary canal at the more distal regions. Our research group has conducted computer simulations of cancellous bone remodeling in the proximal femur using finite element models that represent the geometry and typical daily loading conditions. In these models, the bone apparent density is incrementally adjusted based on the error between the attractor state stimulus and the imposed stress stimulus value at each location throughout the bony model. When this process is implemented on the
computer, the entire distribution of normal bone density and architecture can be developed. Similar simulations have been conducted by others [15,75]. These results strongly suggest that the development of normal cancellous bone architecture in endochondrally derived bone is achieved primarily by epigenetic mechanobiologic processes. In addition, when these methods are used to simulate bone remodeling in response to altered stress states caused by prosthesis implantations or changes in physical activity, these computer models predict changes in bone density distributions that are consistent with the changes observed clinically and experimentally [15,19,76]. The process of cancellous bone adaptation in response to a daily loading history was illustrated by Beaupré et al. [19]. The geometry of the proximal femur was represented using a two-dimensional finite element model (Fig. 18a). To provide a rough approximation of the daily loading history of the bone, three separated loading conditions were considered that represented different activities and joint orientations encountered in a typical day. Initially it was assumed that the cancellous bone density was constant throughout the entire bone. The cumulative stress stimulus in each element was then calculated, assuming that each
FIGURE 17 Morphology of the proximal human femur by (a) radiograph illustrating the bone tissue density distribution and (b) histological section showing cancellous bone architecture.
482 loading case was imposed for many loading cycles over the course of the day. Based on the magnitude of the stress stimulus calculated relative to the attractor state stimulus, the apparent density of each element was adjusted incrementally according to a time-dependent, bone remodeling rate law similar to that used in our appositional modeling simulations. The distribution of bone apparent density after 1 and 30 remodeling increments is shown in Figs. 18b and 18c. Note the development distally of dense cortices, and proximally of a compressive trabecular column through the femoral head, the trabecular band corresponding to the arcuate system in the lateral superior neck and the lowdensity region corresponding to Ward’s triangle. In addition to the distribution of bone density, our bone remodeling simulations can also provide an indication of the directionality of the trabecular bone [77,78]. The polar plots show the equivalent normal stresses at selected locations within the remodeling femur (Fig. 19). The major and minor axes of these polar plots are an indication of trabecular orientation. One interesting note is that the trabecular orientations at a given location need not be perpendicular. Compressive equivalent normal stresses predominate in the femoral head (Fig. 19b), whereas tensile equivalent normal stresses are seen near the greater trochanter and, to a lesser extent, in the region of the superior neck and within the arcuate system (Fig. 19c).
VAN DER
MEULEN, CARTER, AND BEAUPRÉ
Once the normal architecture of the proximal femur was created, three additional simulations were performed to predict bone changes at subsequent times. In the first simulation, the load magnitudes and the number of loading cycles were kept the same as during normal development. In the second simulation, the magnitudes of the loads and the number of cycles were reduced by 20% and in the third simulation the load magnitudes and number of cycles were increased by 20%. Continued normal loading caused little change in the bone density distributions (Fig. 20a). Reduced loading led to a general decrease in the bone density throughout the entire proximal femur (Fig. 20b). The general pattern of density distribution, however, remained similar to that of the normal femur. With increased bone loading, there was a general increase in the bone density everywhere but, again, the general pattern of density distribution remained unchanged (Fig. 20c). These simulation results are consistent with experimental and clinical studies of cancellous bone remodeling in response to changes in physical activity [79 – 81] and can be used to simulate architecture changes in the osteoporotic femur. 2. MATERIAL AND STRUCTURAL STRENGTH The local stress/strain history of the tissue strongly influences the cancellous bone microstructural characteristics that are established during morphogenesis and altered during
FIGURE 18 (a) Finite element mesh and loading conditions. Distribution of bone apparent density is shown after 1 (b) and 30 (c) remodeling increments. Adapted from Beaupré et al., J. Orthop. Res. 8, 662 – 670 (1990).
483
CHAPTER 17 Skeletal Development
FIGURE 19
Polar plots of bone equivalent normal stresses at selected locations in the proximal femur. Used with permission from Carter et al., J. Biomech. 22, 231 – 244 (1989).
functional adaptation. The tissue that is formed under this stress history exhibits material properties that can, in turn, be directly related to its chemical and microstructural character. Under normal circumstances, the two parameters that have been used most successfully to characterize cancellous bone tissue mechanical behavior are apparent density and trabecular orientation. In viewing sections of whole bones, one is often impressed by the strong variations of apparent density and trabecular orientation in the section. These distributions are directly associated with rather dramatic variations in tissue mechanical behavior throughout the bone. Furthermore, the
structural mechanical characteristics of whole bone strength and stiffness are determined by these variations. It is therefore important to appreciate the effect of apparent density and trabecular orientation on tissue mechanics. Clinical measures that reflect these characteristics in whole bones can then be better related to whole bone fracture risk. The apparent density of cancellous bone can be determined by cutting out a small (e.g., 5 5 5 mm) region of bone and washing the fat, marrow, and blood from the trabecular pores. The remaining tissue is then weighed to determine the mass of the mineralized tissue. The apparent density is the mass divided by the bulk volume (including
FIGURE 20 Bone density distributions for three different loading histories, starting with the distribution shown in Fig. 19c as the initial conditions: (a) normal loading, (b) load magnitude and number of cycles reduced by 20%, and (c) load magnitude and number of cycles increased by 20%. Adapted from Beaupré et al., J. Orthop. Res. 8, 662 – 670 (1990).
484
FIGURE 21 Relationship between trabecular bone ultimate compressive strength and apparent density. Adapted from Carter and Hayes, J. Bone Jt. Surg. 59A, 954 – 962 (1977).
pores). Although the extent of mineralization in cancellous bone is, on average, slightly less than that of cortical bone, the true density of bone tissue in cancellous bone is very close to that of cortical bone. The apparent density is, therefore, approximately proportional to the bone volume fraction, which is inversely proportional to the porosity. Bone strength is approximately proportional to the square of the apparent density (Fig. 21) [82,83]. This relationship can be used to describe the general strength characteristics of bone tissue from the most porous trabecular bone to fully compact bone. In compression, the range of strength that this represents is from less than 1 to more than 200 MPa. However, two bone specimens of the same apparent density may differ substantially in strength, depending on the trabecular microstructural characteristics. The most noticeable microstructural characteristic of cancellous bone of a specific apparent density is the organization of the trabeculae. As discussed in the previous section, trabeculae in a particular region tend to be preferentially oriented in the direction of the principal stresses imposed during daily activities. This organization imposes anisotropic characteristics on the tissue so that it is both stronger and stiffer in directions of most pronounced orientation. The anisotropic nature of cancellous bone can be documented using stereological methods that generally document the intersection of trabecular struts with a theoretical grid of parallel lines oriented in different directions [84]. Other stereological measures of secondary importance are mean trabecular width and the extent to which trabeculae are interconnected (trabecular connectivity).
VAN DER
MEULEN, CARTER, AND BEAUPRÉ
The importance of trabecular orientation on bone strength was shown in the early study of Galante et al. [85], who examined the influence of apparent density and trabecular orientation on the compressive strength of vertebral bone specimens. In these specimens there is a pronounced orientation bias in the superior – inferior direction. They found, as have others, a positive relationship between strength and apparent density. Specimens tested in a superior – inferior direction were found to be more than twice as strong as specimens of comparable apparent density that were tested in the medial – lateral direction (Fig. 22). In considering the risk of fracture in whole bones, one should be aware that developmental and adaptational mechanics act to design whole bones for the loads experienced during normal activities and not necessarily for the loads imposed during a traumatic episode. The density distributions and trabecular orientations established in the skeleton may therefore not be well suited for specific traumatic loads that are likely to cause fracture. Nevertheless, some general statements can be made concerning the fracture resistance of cancellous bone regions that are of concern clinically. If we ignore microstructural characteristics and load direction (as a first approximation), the most important parameters to consider are bone size and bone density. Big, dense bones are stronger than small, osteoporotic bones. The DXA techniques that are widely used to measure both density and size can serve to generate predictors of whole bone strength. It is important to recognize that the projected areal bone density measures of BMD (g/cm2) do not provide a true volumetric measure of bone density (g/cm3) [86]. Instead the areal bone density BMD is positively biased by bone size. Large bones with the same apparent density as small bones will have greater DXA estimates of BMD. This
FIGURE 22
Regression of trabecular bone compressive strength on apparent density of human vertebral bone demonstrating dependence of strength on trabecular orientation. The primary trabecular orientation in these specimens was in the superior–inferior direction, and tests were performed in the superior–inferior (SI), anterior–posterior (AP), and lateral (Lat) directions. Data from Galante et al., Calcif. Tissue Res. 5, 236–246 (1970).
CHAPTER 17 Skeletal Development
inherent bias may be fortuitous, as by inherently containing a component of bone size as well as density, BMD values turn out to be good predictors of whole bone strength. To understand how bone density and size contribute to the strength of a whole bone, consider hypothetical cubes of bone tissue with apparent density and cross-sectional area A. The force, F, required to fracture these cubes of bone would be proportional to 2A because the tissue fracture stress would be proportional to 2 and the tissue stress is equal to the applied force divided by A. The parameter 2A can be considered a strength index. If we take a DXA scan of these same cubes we would find that the areal density BMD (g/cm2) would be equal to t, where t is the thickness of the cube. The thickness of the cube is equal to the square root of A. Therefore, the areal density BMD is directly proportional to the square root of the strength index. One might, therefore, expect a correlation between areal density BMD and whole bone fracture strength. Although there is a good deal of scatter in whole bone testing, some investigators are finding reasonable correlations between areal density BMD values and strengths of cadaveric specimens tested in the laboratory. A relationship was found between failure load and femoral neck BMD by Courtney et al. [87], who tested young adult and elderly human femurs to failure in a fall-loading configuration. The ultimate load for both groups combined was found to be strongly positively correlated with the BMD measured at the femoral neck by DXA (Fig. 23).
IV. ADAPTATIONAL MECHANICS IN AGING AND DISEASE This chapter concentrated on the role of mechanical factors in the development and adaptation of the skeleton. This was done with the belief that the skeleton is a self-designing
FIGURE 23
Regression of failure load on BMD for the femoral neck of human femora. Adapted from Courtney et al., Calcif. Tissue Res. 55, 53 – 58 (1994).
485 structure and that the tissue mechanical characteristics and whole bone strength are primarily a consequence of the loading histories that are imposed during ontogeny. With this simple view one can demonstrate that increases or decreases in bone mass may appear as a direct result of changes in the intensity of daily physical activities. Whereas osteoporosis in many individuals may be partly due to decreases in skeletal loading, a purely mechanical view of the pathogenesis of osteoporosis is clearly an oversimplification. Skeletal developmental and adaptational mechanics must be evaluated in light of the many genetic, metabolic, and dietary factors that have been shown to influence bone density and strength in important ways. In general, nonmechanical factors can influence bone by either influencing the basic quality of bone (e.g., mineralization, chemical composition, ultrastructure) or simply increasing or decreasing the amount of bone that is present. In some instances, such as fluoride treatment, both bone quality and bone quantity are affected. Changes in bone quality are the result of changes in the basic biophysics of bone formation and mineralization. Changes in bone quantity alone can be realized by simply changing the balance of osteoblastic and osteoclastic activity as reflected by the number and activities of various populations of bone cells. The number of identifiable agents that influence bone cell biology are staggering and include estrogen, vitamin D, calcium, parathyroid hormone, calcitonin, bisphosphonates, and fluoride. The specific influences of these and other nonmechanical factors are described in other chapters of this book. The mechanisms of action of these factors may all differ but can be of two basic types. These chemical agents could either influence the mechanical regulation of bone cells or have a direct influence on bone cells and their precursors that is independent of the mechanical stimuli. In either case there would be apparent interactions between mechanical and nonmechanical factors. Such synergistic influences have been shown in both preclinical and clinical studies, indicating that mechanical stimuli for bone hypertrophy or atrophy can be altered by endocrine status, diet, or drugs [e.g., 88 – 90]. If one wishes to maintain a view of bone regulation that is dominated by mechanical factors, one can conceptually model the nonmechanical factors as agents that effectively alter the level of mechanical stimulus that is required to maintain bone. In the model of Fig. 3, we find that this viewpoint will cause variations in the attractor stress stimulus, AS, which regulates the mechanically related bone remodeling stimulus. This approach was used by Carter and Beaupré [91] to represent the influence of fluoride treatment on changes in bone volume fraction and was also used by Carter et al. [71] to represent genetic differences in bone mass among different individuals. Frost [92] has employed a similar perspective on factors that may alter the mechanical “set point” of bone. He proposed that such a perspective may be useful for viewing the many bone changes that are observed in osteoporosis.
486
VAN DER
We are only beginning to be able to understand and effectively model the many intrinsic and extrinsic (both mechanical and chemical) factors that control the development and maintenance of bone. It is clear, however, that the biomechanical characteristics of bone and its fracture risk are tied to its ontogenetic history. The structure and mechanical properties of the bones are, in fact, a direct reflection of prior mechanical loading, metabolic status, and diet. Prior mechanical function is perhaps the most dominant factor in determining the form and strength of bones.
References 1. J. I. P. de Vries, G. H. A. Visser, and H. F. R. Prechtl, The emergence of fetal behavior. I. Qualitative aspects. Early Hum. Dev. 7, 301 – 322 (1982). 2. B. K. Hall and S. W. Herring, Paralysis and growth of the musculoskeletal system in the embryonic chick. J. Morphol. 206, 45 – 56 (1990). 3. J. I. Rodriguez, A. Garcia-Alix, J. Palacios, and R. Paniagua, Changes in long bones due to fetal immobility caused by neuromuscular disease. J. Bone Jt. Surg. 70A, 1052 – 1060 (1988). 4. J. I. Rodriguez, J. Palacios, A. Garcia-Alix, I. Pastor, and R. Paniagua, Effects of immobilization on fetal bone development: A morphometric study in newborns with congenital neuromuscular diseases with intrauterine onset. Calcif. Tissue Int. 43, 335 – 339 (1988). 5. J. A. Ogden, Chondro-osseous development and growth. In “Fundamental and Clinical Bone Physiology” (M. R. Urist, ed.), pp. 108 – 171. Lippincott, Philadelphia, 1980. 6. D. R. Carter, M. C. H. van der Meulen, and G. S. Beaupré, Mechanical factors in bone growth and development. Bone 18(Suppl. 1), 5S – 10S (1996). 7. J. Wolff, “The Law of Bone Remodeling.” Springer-Verlag, Berlin, 1892. 8. W. Roux, “Gesammelte Abhandulgen über Entwicklungsmechanic der Organismen,” Vols. I and II. Wilhelm Engelmann, Leipzig, 1895. 9. D. R. Carter, Mechanical loading history and skeletal biology. J. Biomech. 20, 1095 – 1109 (1987). 10. M. C. H. van der Meulen, G. S. Beaupré, and D. R. Carter, Mechanobiologic influences in long bone cross-sectional growth. Bone 14, 635 – 642 (1993). 11. G. S. Beaupré, T. E. Orr, and D. R. Carter, An approach for timedependent bone modeling and remodeling: Theoretical development. J. Orthop. Res. 8, 651 – 661 (1990). 12. R. T. Whalen, D. R. Carter, and C. R. Steele, Influence of physical activity on the regulation of bone density. J. Biomech. 21, 825 – 837 (1988). 13. D. R. Carter, The relationship between in vivo strains and cortical bone remodeling. In “CRC Critical Reviews in Biomedical Engineering” (J. R. Bourne, ed.), pp. 1 – 28. CRC Press, Boca Raton, FL, 1982. 14. H. Frost, “Intermediary Organization of the Skeleton,” Vol. 1. CRC Press, Boca Raton, FL, 1986. 15. R. Huiskes, H. Weinans, H. J. Grootenboer, M. Dalstra, B. Fudala, and T. J. Sloof, Adaptive bone-remodeling theory applied to prosthetic-design analysis. J. Biomech. 20, 1135 – 1150 (1987). 16. Z. A. Ralis, H. M. Ralis, M. Randall, G. Watkins, and P. D. Blake, Changes in shape, ossification and quality of bone in children with spina bifida. Dev. Med. Child. Neurol. 18 (Suppl 37), 29 – 41 (1976). 17. M. Wong, J. Germiller, J. Bonadio, and S. A. Goldstein, Neuromuscular atrophy alters collagen gene expression, pattern formation, and mechanical integrity of the chick embryo long bone. Prog. Clin. Biol. Res. 383B, 587 – 597 (1993).
MEULEN, CARTER, AND BEAUPRÉ
18. J. E. Bertram, L. S. Greenberg, T. Miyake, and B. K. Hall, Paralysis and long bone growth in the chick: Growth shape trajectories of the pelvic limb. Growth Dev. Aging 61, 51 – 60 (1997). 19. G. S. Beaupré, T. E. Orr, and D. R. Carter, An approach for timedependent bone modeling and remodeling — Application: A preliminary remodeling situation. J. Orthop. Res. 8, 662 – 670 (1990). 20. C. T. Rubin and L. E. Lanyon, Dynamic strain similarity in vertebrates; an alternative to allometric limb bone scaling. J. Theor. Biol. 107, 321 – 327 (1984). 21. A. A. Biewener, S. M. Swartz, and J. E. A. Bertam, Bone modeling during growth: Dynamic strain equilibrium in the chick tibiotarsus. Calcif. Tissue Int. 39, 390 – 395 (1986). 22. K. Indrekvam, O. Husby, N. Gjerdet, L. Engester, and N. Langeland, Age-dependent mechanical properties of rat femur: Measured in vivo and in vitro. Acta Orthop. Scand. 62, 248 – 252 (1991). 23. T. S. Keller and D. M. Spengler, Regulation of bone stress and strain in the immature and mature rat femur. J. Biomech. 22, 1115 – 1127 (1989). 24. J. L. Vaughan, “The Physiology of Bone,” 3rd Ed. Clarendon, Oxford, England, 1981. 25. C. B. Ruff, A. Walker, and E. Trinkaus, Postcranial robusticity in Homo. III. Ontogeny. Am. J. Phys. Anthrop. 93, 35 – 54 (1994). 26. J. D. Currey, “The Mechanical Adaptations of Bones.” Princeton Univ. Press, Princeton, 1984. 27. A. M. Parfitt, Bone-forming cells in clinical conditions. In “Bone” (B. K. Hall, ed.), pp. 351 – 429. Telford Press, Caldwell, NH, 1992. 28. C. B. Ruff and W. C. Hayes, Subperiosteal expansion and cortical remodeling of the human femur and tibia with aging. Science 217, 945 – 947 (1982). 29. A. A. Biewener, Musculoskeletal design in relation to body size. J. Biomech. 24 (Suppl. 1), 19 – 29 (1991). 30. T. S. Gross, K. J. McLeod, and C. T. Rubin, Characterizing bone strain distributions in vivo using three triple rosette strain gages. J. Biomech. 25, 1081 – 1087 (1992). 31. H. N. Munro, Evolution of protein metabolism in mammals. In “Mammalian Protein Metabolism” (H. N. Munro, ed.), pp. 133 – 182. Academic Press, New York, 1969. 32. R. B. Martin and P. J. Atkinson, Age and sex-related changes in the structure and strength of the human femoral shaft. J. Biomech. 10, 223 – 231 (1977). 33. R. W. McCammon, “Human Growth and Development.” Thomas, Springfield, IL, 1970. 34. R. W. Smith and R. R. Walker, Femoral expansion in aging women: Implications for osteoporosis and fractures. Science 145, 156 – 157 (1964). 35. S. M. Garn, C. G. Rohmann, B. Wagner, and W. Ascoli, Continuing bone growth throughout life: A general phenomenon. Am. J. Phys. Anthrop. 26, 313 – 318 (1967). 36. C. B. Ruff and W. C. Hayes, Cross-sectional geometry of Pecos Pueblo femora and tibiae: A biomechanical investigation. II. Sex, age, and side differences. Am. J. Phys. Anthropol. 60, 383 – 400 (1983). 37. C. B. Ruff, Allometry between length and cross-sectional dimensions of the femur and tibia in Homo sapiens sapiens. Am. J. Phys. Anthrop. 65, 347 – 358 (1984). 38. M. C. H. van der Meulen, M. W. Ashford, Jr., B. J. Kiratli, L. K. Bachrach, and D. R. Carter, Determinants of femoral geometry and structure during adolescent growth. J. Orthop. Res. 14, 22 – 29 (1996). 39. M. Moro, M. C. H. van der Meulen, B. J. Kiratli, R. Marcus, L. K. Bachrach, and D. R. Carter, Body mass is the primary determinant of mid-femoral bone acquisition during adolescence. Bone 19, 519 – 526 (1996). 40. M. C. H. van der Meulen, R. Marcus, L. K. Bachrach, and D. R. Carter, Correspondence between theoretical models and DXA measurements of cross-sectional growth in adolescence. J. Orthop Res. 15, 473 – 476 (1997).
CHAPTER 17 Skeletal Development 41. M. C. H. van der Meulen, M. Moro, B. J. Kiratli, R. Marcus, and L. K. Bachrach, Mechanobiology of femoral neck structure during adolescence. J. Rehab. Res. Dev. 37, 201 – 209 (2000). 42. M. C. H. van der Meulen, “The Influence of Mechanics on Long Bone Development and Adaptation,” Ph.D. thesis, Stanford University, 1993. 43. M. Geiser and J. Trueta, Muscle action, bone rarefaction and bone formation. J. Bone Jt. Surg. 40B, 282 – 311 (1958). 44. L. E. Lanyon, The influence of function on the development of bone curvature. J. Zool. Lond. 192, 457 – 466 (1980). 45. S. R. Shaw, A. C. Vailas, R. E. Grindeland, and R. F. Zernicke, Effects of a one-week spaceflight on the morphological and mechanical properties of growing bone. Am. J. Physiol. Reg. Int. Comp. Physiol. 254, R78 — R83 (1988). 46. D. M. Spengler, E. R. Morey, D. R. Carter, R. T. Turner, and D. J. Baylink, Effects of spaceflight on structural and material strength of growing bone. Proc. Soc. Exp. Biol. Med. 174, 224 – 228 (1983). 47. H. K. Uhthoff and Z. F. G. Jaworski, Bone loss in response to longterm immobilisation. J. Bone Jt. Surg. 60B, 420 – 429 (1978). 48. M. C. H. van der Meulen, E. R. Morey-Holton, and D. R. Carter, Hindlimb suspension diminishes femoral cross-sectional growth in the rat. J. Orthop. Res. 13, 700 – 707 (1995). 49. S. W. Burke, V. P. Jameson, J. M. Roberts, C. E. Johnston, and J. Willis, Birth fractures in spinal muscular atrophy. J. Pediatr. Orthop. 6, 34 – 36 (1986). 50. F. Biering-Sørenson, H. H. Bohr, and O. P. Schaadt, Longitudinal study of bone mineral content in the lumbar spine, the forearm and the lower extremities after spinal cord injury. Eur. J. Clin. Invest. 20, 330 – 335 (1990). 51. A. Chantraine, B. Nugens, and C. M. Lapiere, Bone remodeling during the development of osteoporosis in paraplegia. Calcif. Tissue Int. 38, 323 – 327 (1986). 52. R. L. Prince, R. I. Price, and S. Ho, Forearm bone loss in hemiplegia: A model for the study of immobilization osteoporosis. J. Bone Miner. Res. 3, 305 – 310 (1988). 53. A. E. Comarr, R. H. Hutchinson, and E. Bors, Extremity fractures of patients with spinal cord injuries. Am. J. Surg. 103, 732 – 739 (1962). 54. K. T. Ragnarsson and G. H. Sell, Lower extremity fractures after spinal cord injury: A retrospective study. Arch. Phys. Med. Rehabil. 62, 418 – 423 (1981). 55. R. K. Martin, J. P. Albright, W. R. Clarke, and J. A. Niffenegger, Load-carrying effects on the adult beagle tibia. Med. Sci. Sports Exercise 13, 343 – 349 (1981). 56. S. L.-Y. Woo, S. C. Kuei, D. Amiel, M. A. Gomez, W. C. Hayes, F. C. White, and W. H. Akeson, The effect of prolonged physical training on the properties of long bone: A study of Wolff’s law. J. Bone Jt. Surg. 63A, 780 – 787 (1981). 57. R. Marcus and D. R. Carter, The role of physical activity in bone mass regulation. Adv. Sports Med. Fitness 1, 63 – 82 (1988). 58. D. T. Reilly and A. H. Burstein, The elastic and ultimate properties of compact bone tissue. J. Biomech. 8, 393 – 405 (1975). 59. J. D. Currey and G. Butler, The mechanical properties of bone tissue in children. J. Bone Jt. Surg. 57A, 810 – 814 (1975). 60. P. Atkinson and J. A. Weatherell, Variation in the density of the femoral diaphysis with age. J. Bone Jt. Surg. 49B, 781 – 788 (1967). 61. A. H. Burstein, D. T. Reilly, and M. Martens, Aging of bone tissue: Mechanical properties. J. Bone Jt. Surg. 58A, 82 – 86 (1976). 62. J. G. Kennedy and D. R. Carter, Long bone torsion. I. Effects of heterogeneity, anisotropy and geometric irregularity. J. Biomech. Eng. 107, 183 – 188 (1985). 63. M. E. Levenston, G. S. Beaupré, and M. C. H. van der Meulen, Improved method for analysis of whole bone torsion tests. J. Bone Miner. Res. 9, 1459 – 1465 (1994).
487 64. R. B. Martin, Determinants of the mechanical properties of bones. J. Biomech. 24 (Suppl. 1), 79 – 88 (1991). 65. F. Selker and D. R. Carter, Scaling of long bone fracture strength with animal mass. J. Biomech. 22, 1175 – 1183 (1989). 66. R. B. Martin and D. B. Burr, Non-invasive measurement of long bone cross-sectional moment of inertia by photon absorptiometry. J. Biomech. 17, 195 – 201 (1984). 67. T. M. Cleek and R. T. Whalen, Bone structural properties in the tibia and fibula using DXA. Int Workshop Bone Densitometry (2000). 68. V. K. Sarin, E. G. Laboa, G. S. Beaupré, B. J. Kiratli, D. R. Carter, and M. C. H. van der Meulen, DXA-derived section modulus and bone mineral, content predict long bone torsional strength. Acta Orthop. Scand. 70, 71 – 76 (1999). 69. S. Borders, K. R. Petersen, and D. Orne, Prediction of bending strength of long bones from measurements of bending stiffness and bone mineral content. J. Biomech. Eng. 99, 40 – 44 (1977). 70. D. R. Carter, T. E. Orr, D. P. Fyhrie, and D. J. Schurman, Influences of mechanical stress on prenatal and postnatal skeletal development. Clin. Orthop. 219, 237 – 250 (1987). 71. D. R. Carter, M. Wong, and T. E. Orr, Musculoskeletal ontogeny, phylogeny, and functional adaptation. J. Biomech. 24 (Suppl. 1), 3 – 16 (1991). 72. M. Wong and D. R. Carter, A theoretical model of endochondral ossification and bone architectural construction in long bone ontogeny. Anat. Embryol. 181, 523 – 532 (1990). 73. G. S. Beaupré, S. S. Stevens, and D. R. Carter, Mechanobiology in the development, maintenance, and degeneration of articular cartilage. J. Rehab. Res. Dev. 37, 145 – 151 (2000). 74. G. H. von Meyer, Die Architektur der Spongiosa. Arch. Anat. Physiol. Wiss. Med. 34, 615 – 628 (1867). 75. H. Weinans, R. Huiskes, and H. J. Grootenboer, The behavior of adaptive bone-remodeling simulation models. J. Biomech. 25, 1425 – 1441 (1992). 76. T. E. Orr, G. S. Beaupré, D. R. Carter, and D. J. Schurman, Computer predictions of bone remodeling around porous-coated implants. J. Arthrop. 5, 191 – 200 (1990). 77. D. R. Carter, T. E. Orr, and D. P. Fyhrie, Relationships between loading history and femoral cancellous bone architecture. J. Biomech. 22, 231 – 244 (1989). 78. C. R. Jacobs, J. C. Simo, G. S. Beaupré, and D. R. Carter, Adaptive bone remodeling incorporating simultaneous density and anisotropy considerations. J. Biomech. 30, 603 – 613 (1997). 79. G. P. Dalsky, K. S. Stocke, A. A. Ehsain, E. Slatopolsky, W. C. Lee, and S. J. Birge, Weight-bearing exercise training and lumbar bone mineral content in postmenopausal women. Ann. Int. Med. 108, 824 – 828 (1988). 80. C. L. Donaldson, S. B. Hulley, J. M. Vogel, R. S. Hattner, J. H. Bayers, and D. E. McMillan, Effect of prolonged bed rest on bone mineral. Metabolism 19, 1071 – 1084 (1970). 81. W. S. S. Jee and X. J. Li, Adaptation of cancellous bone to overloading in the adult rat: A single photon absorptiometry and histomorphometry study. Anat. Rec. 227, 418 – 426 (1990). 82. D. R. Carter, W. C. Hayes, and D. J. Schurman, Fatigue life of compact bone. II. Effects of microstructure and density. J. Biomech. 9, 211 – 218 (1976). 83. D. R. Carter and W. C. Hayes, The compressive behavior of bone as a two-phase porous structure. J. Bone Jt. Surg. 59A, 954 – 962 (1977). 84. A. D. Kuo and D. R. Carter, Computational methods for analyzing the structure of cancellous bone in planar sections. J. Orthop. Res. 9, 918 – 931 (1991). 85. J. Galante, W. Rostoker, and R. D. Ray, Physical properties of trabecular bone. Calcif. Tissue Res. 5, 236 – 246 (1970). 86. D. R. Carter, M. L. Bouxsein, and R. Marcus, New approaches for interpreting projected bone densitometry data. J. Bone Miner. Res. 7, 137 – 145 (1992).
488 87. A. C. Courtney, E. F. Wachtel, E. R. Myers, and W. C. Hayes, Effects of loading rate on strength of the proximal femur. Calcif. Tissue Int. 55, 53 – 58 (1994). 88. Y. Kodama, Y. Umemura, S. Nagasawa, W. G. Beamer, L. R. Donahue, C. R. Rosen, D. J. Baylink, and J. R. Farley, Exercise and mechanical loading increase periosteal bone formation and whole bone strength in C57BL/6J mice but not in C3H/Hej mice. Calcif. Tissue Int. 66, 298 – 306 (2000). 89. B. F. Halloran, D. D. Bikle, J. Harris, S. Tanner, T. Curren, and E. Morey-Holton, Regional responsiveness of the tibia to intermittent
VAN DER
MEULEN, CARTER, AND BEAUPRÉ
administration of parathyroid hormone as affected by skeletal unloading. J. Bone Miner. Res. 12, 1068 – 1074 (1997). 90. T. L. Robinson, C. Snow-Harter, D. R. Taaffe, D. Gillis, J. Shaw, and R. Marcus, Gymnasts exhibit higher bone mass than runners despite similar prevalence of amenorrhea and oligomenorrhea. J. Bone Miner. Res. 10, 26 – 35 (1995). 91. D. R. Carter and G. S. Beaupré, Effects of fluoride treatment on bone strength. J. Bone Miner. Res. 5 (Suppl. 1), S177 — S184 (1990). 92. H. M. Frost, The pathomechanics of osteoporoses. Clin. Orthop. 200, 198 – 225 (1985).
489
CHAPTER 18 Inhibition of Osteopenia by Biophysical Intervention
CHAPTER 18
Inhibition of Osteopenia by Biophysical Intervention CLINTON T. RUBIN,*,† STEFAN JUDEX,* KENNETH J. MCLEOD,* AND YI-XIAN QIN* *
Musculo-Skeletal Research Laboratory, Department of Biomedical Engineering, and the Center for Biotechnology, State University of New York, Stony Brook, New York 11794
†
I. II. III. IV.
V. Attenuation of Wolff’s Law by Systemic Disorders VI. Clinical Application of Biophysical Stimuli VII. Summary References
Introduction Wolff’s Law of Bone Adaptation Structural Demands on the Vertebrate Skeleton Modulation of Bone Tissue by Biophysical Stimuli
I. INTRODUCTION
mechanical environment of bone and demonstrating that physically based signals within these physiologic constraints can be osteogenic. Further, evidence will be provided that demonstrates that these regulatory signals diminish as we age because of sarcopenia (a state of diminished muscle mass and function), suggesting that a key etiologic factor in osteoporosis results from degeneration of the muscular system, rather than a primary dysfunction of the skeletal system per se. Finally, preliminary evidence is provided that extremely low-level mechanical signals are capable of inhibiting the rapid bone loss that typically follows menopause. While a biophysical approach contrasts sharply with pharmaceutical strategies for the treatment of osteoporosis, in essence, the structural success of the skeleton is a product of the ability of bone tissue to adapt to this constant barrage of mechanically based signals, and herein lies the basis for a unique treatment regimen for this debilitating skeletal disease. The elaborate cortical and trabecular morphology of the skeleton is a dynamic product of three competing and disparate goals: pressure to establish (and maintain) mechanical strength, the metabolic advantages inherent in a
Osteopenia, a condition of diminished bone mass, will become symptomatic only when mechanical demands exceed the structural viability of the skeleton. In other words, when function is no longer accommodated by form. While treatment of osteoporosis is principally oriented toward pharmaceutical prophylaxis, this chapter proposes a case for considering biophysical intervention in general (e.g., electrical, acoustic, thermal), and biomechanical treatment in particular. Much of the frustration and ambivalence toward biophysical treatment for osteoporosis may, in reality, be a result of a poorly defined osteogenic signal, exacerbated by decreasing sensitivity/responsiveness of the aging skeleton. However, there is accumulating basic science and clinical evidence that biophysical intervention may prove a safe and efficacious means of inhibiting and reversing osteoporosis, a goal that can be achieved by augmentation, rather than disruption, of bone remodeling processes. This chapter will try to build support for the clinical potential of biophysical stimuli by defining the functional
OSTEOPOROSIS, SECOND EDITION VOLUME 1
489
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
490 minimal mass, and tissue serving as the body’s principal mineral reservoir. This balance is achieved via site-specific formation and resorption of bone, reflecting a combative struggle between systemically based catabolic factors (e.g., parathyroid hormone, vitamin D) aimed at releasing calcium from the skeleton, and site-specific biophysical signals, which arise from function (e.g., mechanical strain, pressure generated fluid flow, stress-generated electrical potentials), which serve as local anabolic factors. When metabolic demands swamp the structural role, the skeleton can deteriorate to the point that it can no longer withstand normal functional load bearing. While osteopenia may result from an increased level of resorptive factors, it is as likely due to an age- or menopause-related attenuation of the physically based signals that serve to maintain bone structure. The premise of the chapter is that osteoporosis is an end product of a dysfunctional form/function relationship. This perspective is supported by the fact that this disease is most symptomatic at specific load bearing sites of the skeleton (e.g., femoral neck, lumbar spine). Paradoxically, even though the disease is focal in nature, the most accepted treatment protocols are administered systemically. Intervention that inhibits, disrupts, or even promotes the bone cell kinetics of the entire skeleton belies the need for a focal treatment strategy for a focal etiology. Further, an effective treatment for osteoporosis cannot realistically arise simply by statically retaining the bone mass or density that is present at any given point in time (e.g., antiresorptives). Bone quality is as important as bone quantity, emphasizing that the ideal therapy will be one that incorporates all aspects of normal bone turnover, not one that annihilates any given part of it. An optimal treatment would target a site-specific regimen for the inhibition and/or reversal of bone loss and achieve this without interrupting the delicate interplay between the cells responsible for bone remodeling. Ideally, this therapy would reestablish the osteogenic components of this balance to a level that would revitalize a normal remodeling equilibrium. The anabolic potential of load bearing, as well as the site-specific loss of bone density in osteoporosis, implies that an improved understanding of how biophysical signals mediate remodeling will facilitate their effective application to the treatment of these diseases.
II. WOLFF’S LAW OF BONE ADAPTATION The principal responsibility of the skeleton is to support the loads and bending 1moments that arise during activity, a responsibility that results in mechanical strain in the bone tissue. The ability of the skeleton to adapt to these functional demands was recognized over a century ago and is now referred to as Wolff’s law [1]. The basic premise of
CLINTON T. RUBIN
ET AL.
this “law” is that bone strives toward an optimized structure, which caters to an individual’s level of activity. Thus, each individual would tune the mass and morphology of the skeleton such that it was sufficient to withstand the extremes of functional loading, but not so massive as to make transportation a metabolic liability. While this Goldilocks paradigm of “just right” emphasizes the role of anabolic functional stimuli in defining skeletal morphology, identifying those components within the milieu that achieve this balance has proven difficult. Nonetheless, it is a venture worth pursuing. If the osteogenic constituents of the functional regime could be identified, focal modalities could be developed to treat skeletal disorders that exploit the extreme sensitivity of the musculoskeletal system to the mechanical domain. The search for the osteogenic components that control Wolff’s law has benefited from both qualitative and quantitative observations of the skeleton’s response to changes in its functional environment. The sensitivity of bone to physical and environmental stimuli is readily evident in a large body of clinically based studies that show the skeleton’s graded response to levels of exercise [2,3], as well as the bone lost due to reductions in gravitational force [4,5] and bed rest [6]. Importantly, evidence of local hypertrophy (e.g., humerus in tennis players [7,8]) or resorption (e.g., femoral neck after total arthroplasty [9,10]) following site-specific activities emphasizes that focally mediated adaptation is caused by changes in the local functional environment. While these studies portray the skeleton’s sensitivity to function, the difficulty in defining the complex loading history of the bone under study has precluded identification of the osteoregulatory component(s) embedded within the physical milieu. To address these limitations, analytic and empiric models have been developed to study physical influences on bone formation. Through the past two decades, specific components of the mechanical milieu have been proposed as the dominant stimulus for bone adaptation. These include strain magnitude [11], strain rate [12], electrokinetic currents [13], piezoelectric currents [14], fluid shear flow [15], and strain energy density [16]. While these parameters demonstrate strong correlations to specific skeletal morphologies, few have validated their prescience by predicting morphologic changes that would be stimulated by distinct loading condition [17]. The inability to identify a unifying principle for the mechanical control of bone adaptation may be aggravated by the underlying assumption that a structural efficiency paradigm (minimal skeletal strain/minimal skeletal mass) is itself the driving stimulus that regulates the remodeling process. Alternatively, bone cells may know little of the principles of structural mechanics and 1 Moment is an engineering term that describes the turning, twisting, or rotational effect of a force; M Nm.
CHAPTER 18 Inhibition of Osteopenia by Biophysical Intervention
491
instead are responding to “biologically relevant” parameters of functional milieu that are not necessarily linked to peak structural challenges. This “other than peak” perspective is employed in several biologic systems, which perceive and respond to exogenous stimuli, such as vision, hearing, and touch. In considering the mechanically mediated control of bone remodeling, there is little argument that biophysical stimuli are potent determinants of skeletal morphology; but too much may only damage the system. Indeed, we shut our eyes when it is too bright, cover our ears when it is too loud, and shed tears when the pressure is too great. To identify criteria by which these processes are controlled, it is necessary to look beyond the material consequences of a structure subject to load and consider the biologic benefit of a viable tissue subject to functional levels of strain.
significant safety factor reigns and that the skeleton can survive an errant step into a plot hole or the odd trip over a curb. However, imagining a mechanism whereby the skeleton “knows” it is loaded to half its yield strength seems unlikely, somehow altering its morphology based on a paranoia of strains that are much higher. Instead, in vivo and in vitro data suggest that the functional criteria that regulate adaptation, and the means by which bone cells perceive and respond to their functional milieu, are far more sophisticated than a magnitude “brute force” or “accumulation of microdamage” perspective. Indeed, these data suggest that adaptation to biophysical stimuli occurs to encourage specific components of the strain milieu, rather than to necessarily annihilate them. Two thousand microstrain may sound like an ominously large number, but in reality it represents an exceedingly small change in length from a material’s original length. Cartilage is subject to 25% compressive deformations, tendons experience functional tensile strain upward of 20%, and ligaments may well stretch 4 – 5% during the extremes of functional loading. The 20% strain of these connective tissues is two orders of magnitude greater than the 0.2% (2000 ) peak strains experienced by bone. Admittedly, even a 0.2% strain seems unwieldy for a bridge support or skyscraper, yet by the time a 10 , bone-lining cell is subject to 2000 , such deformation is on the order of angstroms. Clearly, if deformations of this order are to affect cell metabolism, the bone cell mechanosensory system must be exceedingly sensitive (Fig. 1).
III. STRUCTURAL DEMANDS ON THE VERTEBRATE SKELETON A. Cross-Species Similarity of Peak Bone Strain Magnitudes Regardless of the design or function of a vertebrate, strain is a ubiquitous product of a functionally loaded skeleton. Mechanical strain, therefore, is commonly considered a reasonable and efficient means of translating the intensity, duration, and manner of functional loading into a site-specific, generic signal relevant to the cells responsible for osteoregulation. One obvious goal of this strain-mediated form/function formula is to avoid fracture. Thus, bone loading and architecture must be coordinated to avoid the tissue’s yield strain of 0.7% (7000 microstrain () [18]). To establish the role of functional strain in defining skeletal morphology, as well as to determine how closely bone approaches its point of failure, the mechanical signals to which the skeleton is exposed must be determined. This goal is achieved by attaching strain gauges directly to a bone in vivo [19,20], which permits a number of critical observations to be made regarding the structural demands made on the skeleton. Peak strain magnitudes measured in diverse vertebrates are remarkably similar, ranging in amplitude from 2000 to 3500 [21,22]. Whether measured in metacarpal bone of a galloping horse, the tibia of a running human, the humerus of a flying goose, the femur of a trotting sheep, or the mandible of a chewing macaque, this “dynamic strain similarity” suggests that skeletal morphology is adjusted in such a way that functional activity elicits a very specific (and perhaps beneficial) level of strain to the bone tissue [23]. That strains of this magnitude are a factor of two below the yield point of bone material emphasizes that a
B. Absence of a Uniform Peak Strain Stimulus Models aimed at defining the osteogenic components of the overall strain history of alone (accumulated strain information over time) have focused on correlating bone morphology to the predominant characteristics of the mechanical environment of the bone, including peak strain magnitude, peak strain rate, peak strain energy density, or number of peak loading cycles. Certainly, that peak strain magnitudes among vertebrates are all very similar would support a hypothesis that achieving a specific level of peak strain is the Holy Grail toward which the bone tissue strives. However, models based on this hypothesis also commonly assume that a homogeneous state of strain persists across the cortex [24 – 27]. In other words, at either that point in the stride where the strains are greatest, or when strains are integrated over some time-averaging period, it is assumed that each area of the cortex is subject to the identical strain information and therefore the same stimulus for remodeling. In vivo strain gauge data, however, demonstrate the spatial distribution of peak normal and shear strains, as well as strain energy density (an aggregate of the stress/strain state), to be extremely non uniform [20,28]. For example, for a horse
492
FIGURE 1
Strain is defined as a (load-induced) change in length relative to the structure’s original length. One thousand microstrain, or 0.1% strain, reflects the amount of strain experienced by bone tissue during an activity such as walking. For a structure such as the 170-m Washington monument, 1000 would represent a 17-cm change in length over the entire structure. In a giraffe tibia, 1000 would reflect a 1-mm change over the bone’s original 1000-mm length. At the level of a 10-m bone lining cell sitting on the periosteum of that giraffe tibia, its dimensional change when subject to 1000 would be 100 A. The mechanisms responsible for perceiving and responding to such small biophysical signals, whatever they may be, must be extremely sensitive. Reproduced with permission from Rubin et al. [90].
walking at 2 ms1, at the point in the stride in which peak strain is achieved, the spatial distribution of normal strain in the metacarpal ranges from plus 13 in tension to minus 1048 in compression. Further, shear strain ranges from 54 to 360 (indicating opposite directions of shear), and peak strain energy density, which accounts for all components of the strain tensor, spans two orders of magnitude, from 117 to 10,602 Pa. A uniform strain stimulus is certainly not readily apparent in the functionally loaded skeleton. Bone structure would realize an important benefit if it were to adapt to elicit a homogeneous strain distribution; peak strain could be minimized while simultaneously minimizing total bone mass. Imagine trying to break a pencil by loading it only in an axial fashion, along its longitudinal shaft. Short of enlisting a materials test machine, such an endeavor would prove difficult. As any frustrated writer knows, the effortless way to snap a pencil is to bend it. Taking this design hazard into account, structural models of bone adaptation insist that bone strives to minimize risk of failure by ensuring that the bone is loaded axially, hence the uniform distribution of strain. It certainly seems aesthetically reasonable that architectural embellishments of the skeleton such as cross-sectional morphology and longitudinal curvature, together with antagonistic and synergistic
CLINTON T. RUBIN
ET AL.
muscle activity, all conspire to minimize any bending in the bone. Empirical evidence, however, measured from the appendicular skeleton during functional activity, demonstrates that the predominant (85%) component of strain is generated by bending, even though far less bone mass would be required to support the same loads if the bone were loaded axially [21,29,30]. The bending is sufficient to subject a significant portion of the cortex to longitudinal tension, thus mimicking the perilous state of the doomed pencil. Further, committing one surface to tension and another to compression means that the transition between these two areas creates a region of the cortex which experiences very low peak strain magnitudes. Even though this “neutral axis” is far removed from the area of the cortex subject to the peak strains, somehow tissue is retained in this strain demilitarized zone (DMZ) [31–33]. Clearly, bone cannot be presumed to be solely a compressive element, and strain cannot be presumed to be uniform across the cortex. It could be argued that a more uniform strain stimulus could be achieved via integration of strain information over time [26,27]. Such a time-dependent “strain memory” in bone cell networks has already been demonstrated, with noncollagenous matrix proteins “responding” to load by changing their morphology [34]. Over the course of time, a homogeneous strain signal could be achieved by ensuring all areas of the cortex are subject — at one point or another within the tissue’s memory span — to the peak strain milieu. However, this “equilibration presumption” is not consistent with in vivo data that show that the inhomogeneity of the instantaneous strain distribution becomes even more disparate as the strain energy is summed over the course of a stride (Fig. 2). Summing the functional strain milieu over an entire 24-h period demonstrates the range of total strain experienced between areas of the cortex to be huge [35], approaching three orders of magnitude. While high degrees of bending may provoke distinctly non uniform strain distribution in cortical bone, it does not necessarily preclude the possibility of an adaptive mechanism mediated by some aspect of strain. One solution to this dilemma is to suggest that bone cells in different regions of the cortex may be differentially sensitive to strain (some cells strive to 3000 in compression, some to 1500 in tension; others — near the neutral axis — are content with strains of 50 or 100 ). While this is appealing in its simplicity, the genetic logistics of a spatially specific strain sensitivity would be astronomical. Alternatively, it is possible that strain information is spatially integrated in three dimensions via a cell network facilitated by gap junction intercellular communication [36], such that the are of the cortex subject only to 100 resists resorption due to sufficient homeostatic signals received from adjacent areas subject to much higher strain. This “information integration” perspective is supported by the observation that the bone loss that parallels disuse occurs
493
CHAPTER 18 Inhibition of Osteopenia by Biophysical Intervention
FIGURE 2 The distribution of strain energy about the midshaft of the horse cannon bone (MCIII) during a gallop. Shown is strain energy density (SED) during that point in the stride in which peak strain is achieved (left), as well as the SED averaged over the entire stance phase of the stride (right). In the first case, SED ranges from a minimum of 600 Pa to a maximum of 56,000 Pa and, when averaged over the stride, from 461 to 51,375 Pa. Importantly, while the distribution of the peak and time averaged SED is very nonuniform, the manner in which the bone is loaded remains constant (i.e., the site of peak and minimal SED varies very little). Adapted from Gross et al. [20] uniformly about the cortex and through the diaphysis, even though the net change in bone strain caused by the absence of function varies widely [37]. The degeneration of cell:cell communication, which occurs with age, may contribute to the etiology of osteopenia [38]; even though the physical signals persist, the information super highway needs repaving. Evidence suggests that individual osteocytes also contain the ability to modulate the architecture of their individual lacunae, thereby “autoregulating” the mechanical environment to which they are exposed [71]. The bone cell would modulate its perception of strain by regulating the level of its attachment to the matrix, as well as the size of the periosteocytic space, thereby controlling the amount of direct deformation and/or shear to which it would be subjected. It is entirely feasible, and also biologically straightforward, to deem the level of strain in the matrix as irrelevant and instead focus solely on the amount and type of mechanical stimulus imposed on the bone cells. Accept for a moment that the osteocyte has a generic strain goal, similar at all sites through the skeleton. It does not really matter what that amount is, assume it is 100 . The osteocyte, regardless of whether the matrix is exposed to strains of 3000 in the tibia cortex or 300 in the mandibular trabeculae, can achieve the goal of 100 by increasing or decreasing its lacunar attachments rather than altering the mass and morphology of the bone. Near the neutral axis, the cell is highly coupled to the matrix such that very little strain influences cell metabolism in the same manner as a poorly coupled cell in a highly strained region of the cortex. With the osteocyte also capable of detaching from the matrix, through expression of collagenase, it is able to modulate its own mechanical microenvironment through the active coupling and uncoupling of its
membrane to the matrix [39]. In this manner, the osteocyte, identical in the mandible or tibia, would achieve its goal of 100 simply by modulating its own private space. This may also, in part, explain the “lazy zone” in bone (that strain region where an alteration in a mechanical signals fail to stimulate changes in bone architecture); the cell, not the tissue, is accommodating the new mechanical stimulus. In reality, this indicates that the bone can adapt without necessarily requiring the formation or resorption of tissue. Indeed, some exercise regimes that do not increase bone density per se may nevertheless improve the viability of the tissue, and thus improve the quality of bone, without influencing quantity.
C. Contribution of Muscle Dynamics to the Strain Environment That bone morphology ignores so many structural paradigms suggests that models of bone adaptation may rely too heavily on teleologic-based engineering concepts of structural optimization and safety, and should consider principles of evolution and physiology, which emphasize undirected development. Recognizing that bone is first a tissue and second (and quite conveniently) a structure, it is important to consider the biologic implications associated with biophysical stimuli. Indeed, tissue viability may depend on aspects of the mechanical environment that may not be at all rooted in minimal strain/minimal mass criteria, such as strain-dependent perfusion or strain-induced electrical currents. Alternatively, bone adaptation may depend on some camouflaged subset of the mechanical milieu, e.g., the mechanical strains induced by muscle.
494 While the symbiotic relationship between muscle and bone is obvious, only seldom is it explicitly considered in the context of one defining the other [40]. As muscle contraction imposes far smaller strains on the skeleton than those caused by ground reaction loads (e.g., impact), their role in defining bone morphology has not received much consideration. Although muscle-induced strains may indeed be relatively small, they are sustained for extended periods of time (e.g., in postural muscle activity), and thus , over time , may dominate a bone’s characteristic “strain history.” Examining this hypothesis, strain data from a variety of animals reveal the existence of a broad frequency range of strains in the appendicular skeleton, even during activity such as quiet standing [35]. While reaction forces due to locomotion give rise to large distinct strain components, spectral analysis of standing strain recordings shows significant strain information extending out to the frequency range of 50 cycles per second (Hz). While the magnitudes and frequency content of gait-related strains change transiently as a function of speed and gait, time-averaged strains (strain history) are dominated by standing strain spectra and are therefore quite stable over time and more uniform (Fig. 3). The spectral content of time averaged strain history, therefore, may better portray the wide range of strain information present in the functional milieu of the skeleton. From a stimulus standpoint, these persistent, low-amplitude signals may, when summed, be at least as important as the seldom occurring, and somewhat unpredictable, peak strain events [41]. Whether the skeleton is preferentially sensitive to a few, large strain events or a continual barrage of low magnitude events, must be evaluated at the tissue level, where specific mechanical signals can be introduced, and the resultant remodeling evaluated.
IV. MODULATION OF BONE TISSUE BY BIOPHYSICAL STIMULI It is clear that the skeletal organ is subject to a wide range of mechanical signals, including low to highfrequency strains, normal and shear strains, and compressive and tensile strains. It is also clear that the cells on and within the mineralized matrix are subject not only to mechanical parameters such as strain, but derivatives of tissue deformation such as fluid flow and electrokinetic currents, parameters that may represent an important physiologic pathway in mediating an adaptive response. Unfortunately, the study of bone at this organ level makes it difficult to categorically identify the osteogenic parameters of the biophysical milieu. Addressing the “mechanism” of signal transduction necessitates a move to the level of bone tissue.
CLINTON T. RUBIN
ET AL.
FIGURE 3
(a) A 2-min strain recording from the caudal longitudinal gage of the sheep tibia while the animal took a few steps with peak strains on the order of 200 . (b) A 20-s portion of that strain record shows peak strain events as large as 40 . (c) Further scaling down to a 3-s stretch of the strain recording illustrates events on the order of 5 . Adapted from Fritton et al. [35].
A. Identifying Osteogenic Parameters of the Strain Milieu The most efficient means of studying adaptive tissue responses to biophysical stimuli is through animal models in which the mechanical environment can be controlled accurately. Investigations designed to identify those components of the mechanical milieu that regulate skeletal adaptation have used a number of experimental approaches, including stress protection adjacent to implants of varied stiffness [42,43], overload caused by osteomy [18,44], and externally applied loading [45 – 47]. Although these applied loading experiments have contributed to our understanding of adaptation, they too have distinct limitations, the greatest of which is that the loads are applied for a limited, arbitrarily chosen period of time, yet for the remainder of the day the animals are able to apply uncontrolled, unmonitored loading to the bone under investigation.
CHAPTER 18 Inhibition of Osteopenia by Biophysical Intervention
495
To ensure that the adaptive modeling and remodeling that one observes is uniquely a product of those mechanical parameters that are applied, it becomes essential that the bone be exposed to minimal spurious loading events. These requirements have been met through several different animal models, ranging from the loading of cancellous bone in the distal femur of dogs [48] to the loading of tail vertebrae in rats [49]. In our laboratory, a primary means of studying the adaptation of cortical bone to biophysical stimuli has been the functionally isolated turkey ulna [31]. The advantage of this model is that the bone tissue is subject only to the mechanical [32] or electrical [50] regimen prescribed by the investigators, with no aberrant biophysical signals entering the preparation. In short, the ulna of adult male turkeys is functionally isolated by proximal and distal epiphyseal osteotomies, leaving the entire diaphyseal shaft undisturbed. Caps filled with methylmethacrylate are placed over the ends of the bones and percutaneous transfixing pins are placed through the caps. This model has demonstrated that 8 weeks of functional isolation alone will consistently result in a 10 – 15% loss of bone in 8 weeks, where an externally applied mechanical strain regimen, physiological in strain magnitude, can lead to significant increases in cross-sectional bone area depending on the signal parameters. Specifically, alterations in bone mass, turnover, and internal replacement are sensitive to changes in the magnitude [11] and distribution [51] generated within the bone tissue. Further, a loading regimen must be dynamic (time-varying) in nature; static loads do not influence bone morphology [52]. Moreover, the full osteogenic potential of a large amplitude (2000 ), low frequency (1 Hz) regimen is realized following only an extremely short ( 1 min) exposure to this stimulus [31].
response to be an appropriate interim strategy in the adaptive response to intense new mechanical challenges, we undertook a series of longer duration adaptation studies [54]. Following 16 weeks of a load regimen of only 100 cycles per day, inducing a peak compressive strain of 2000 , new bone stimulated in the turkey ulna model was lamellar, composed of primary and secondary osteons toward the original cortex and circumferential lamellae at the periphery, (Fig. 4, see also color plate). Remnants of the initial woven bone response seen at 4 weeks remained clearly visible at both 8 and 16 weeks as diffusely labeled interstitial elements within the newly formed lamellar construct. The presence of secondary osteons, circumferential lamellae, and an osteocyte density and organization similar to that seen in controls suggests that the presence of woven bone in the initial stages of the adaptive process is not
B. Long-Term Modeling Response to Mechanical Stimuli In the tissue level experiments discussed earlier the efficacy of any given mechanical stimulus to affect bone remodeling has typically been evaluated following 8 weeks of either disuse or externally applied loading. When a strongly osteogenic load regimen is applied, the magnitude and location of the adaptive response the consistent, yet the character of the periosteal response is often woven in nature and is not considered an ideal adaptive response under any conditions. If mechanically mediated modeling stimulated only woven bone, its potential use to combat osteoporosis would be diminished. To determine if this non-optimal tissue type persists over time (or disappears altogether), suggesting its appearance to be an aberrant reaction to surgery [53], or if lamellar bone replaces the woven response, emphasizing the woven
FIGURE 4
A fluorescent photomicrograph of the periosteal surface of a turkey ulna diaphysis following 8 (top) and 16 (bottom) weeks of a mechanical regimen sufficient to cause a peak of 2000 . The 8-week response shows consolidating primary bone. By 16 weeks, remnants of the original woven response can be seen serving as interstitial elements of primary and secondarily remodeled bone. In essence, the woven bone response has served as a strategic stage in the achievement of a structurally appropriate increase in bone mass. Reproduced with permission from Rubin et al. [54]. (See also color plate.)
496 necessarily a pathologic or transient reaction to injury, but instead may represent a normal – and strategic – stage in response to a potent mechanical stimulus. It also demonstrates that the response of bone to mechanical loading is not only anabolic, but that it produces lamellar bone, the “gold standard” of bone formation.
C. Differential Modeling/Remodeling to Distinct Components of the Strain Tensor Mechanical factors such as magnitude and duration are essentially “organ” level stimuli. Out of the widely diverse range of mechanical signals to which the tissue is exposed, it is essential to determine which components of this strain tensor (i.e., the complete strain state of the bone tissue) actually influence the metabolism of the osteocyte, osteoblast, or osteoclast to retain the status quo, initiate modeling, or turn on remodeling. While the strain tensor of the functional regimen is very complex, it can be described in general terms by two predominant components: dilatation (i.e., dilate; volume changes caused by hydrostatic stress) and deviatoric (i.e., deviate; shape changes caused by shear stress) parameters. If the control of bone adaptation demonstrates a differential response to discrete parameters of the mechanical milieu, the mechanisms that control bone morphology can be elucidated. This goal has been approached using the turkey ulna model of disuse osteopenia, in which the modeling and remodeling response was quantified following 4 weeks of either axial or torsional loading or disuse [55]. Each of the two load groups were subject to peak principal strains of 1000 (predominately normal strain in the axial case, and shear strain when subject to torsion). Of the three distinct groups, only disuse caused a significant change in gross areal properties as compared to controls (13% loss of bone). This suggests that both axial and torsional loading conditions are reasonable substitutes for the functional signals normally responsible for retention of bone mass, leaving the periosteal and endosteal envelopes unphased by disparate components of the strain tensor. The intracortical response, however, was found to depend strongly on the manner in which the bone was loaded. Disuse failed to increase the number of sites within the cortex actively involved in bone turnover (intracortical events), yet significant area was lost within the cortex due to a threefold increase in the mean size of each porotic site. Axial loading increased the degree of intracortical turnover as compared to intact controls, yet the average size of each porotic event remained identical to that of the control. Conversely, compared to the control, torsion elevated neither the number of porotic events, the area of bone lost from within the cortex, nor the size of the porotic event. It appears that bone tissue can readily differentiate between
CLINTON T. RUBIN
ET AL.
distinct components of the strain tensor, with strain per se necessary to retain coupled formation and resorption, deviatoric strain achieving this goal by maintaining the status quo, whereas dilatation elevates intracortical turnover, but retains coupling. These experiments suggest two critically important characteristics of bone cellular activity: modeling (surface) and remodeling (intracortical) activity are not necessarily coupled and the osteocyte population can differentiate between dilatational and deviatoric strains in the tissue. This suggests that the ability of the cell to distinguish between volumetric and shape changes is achieved through several distinct mechanisms; perhaps dilatation is sensed directly by the degree of coupling between cell and matrix, whereas deviatoric stresses, which cause fluid flow, influence cell activity via second messenger gradients. This hypothesis is supported by in vitro findings that confirm that bone cells perceive and respond, albeit differently, to both hydrostatic and shear stresses [56,57]. Most importantly, these data suggest that different components of the strain tensor have distinct regulatory roles. . . it is not the aggregate of strain per se that defines remodeling, but that independent components of the strain tensor have differential responsibilities in achieving and maintaining bone mass. Other biophysical factors derived from the strain environment may also be responsible for cellular signaling. Dynamic loading of bone tissue not only results in a dynamic stress – strain environment, but is also associated with other matrix-related events related to the fluid content within the porous space of bone tissue [58 – 60]. These interstitial flow phenomena are driven by gradients in tissue dilatation [17,61], which will be significantly more pronounced under conditions of axial loading than with torsional loads. Thus, observed differences between axial and torsional loading may be due to differences in fluid pressurization [62], flow-induced shear stresses [15], or strain-generated electric potentials [13,14] in the cellular microenvironment.
D. Osteogenic Potential of Low-Level Electric Fields Mechanical data described earlier demonstrate a distinct relationship of function to form. They even suggest a means by which the mechanical stimulus can be perceived and regulated, via the cell – matrix tethers facilitated by integrins and matrix proteins. Just the same, for an osteocyte entombed within the tissue matrix, these deformations are on the order of a few angstroms, suggesting that alternatives to the strain of the cell per se should be considered. One such mechanism for the coupling of mechanical deformation to cellular activity is the strain-induced movement of interstitial fluid in the bone, similar to the water flow through a sponge caused either by stretching or squeezing.
CHAPTER 18 Inhibition of Osteopenia by Biophysical Intervention
497
Whereas strain-induced fluid flow will certainly contribute to increased nutrient and metabolite transport, this fluid movement will also cause an electrokinetic interaction with the bone tissue. The potential role of electricity in the regulation of bone tissue was first considered over 40 years ago with the report of electrical currents being generated with the loading of dry bone [63]. The discovery of load-induced piezoelectric potentials in bone provided a means by which stress or strain could intrinsically alter the biophysical environment of the bone cell, and thus influence proliferation and differentiation [14]. This hypothesis became even more attractive when it was demonstrated that, in wet bone, two sources of electrical current existed in parallel; piezoelectric currents induced by the deformation of collagen and the relatively large electrokinetic currents (streaming potentials) produced by the strain-induced flow of charged constituents of extracellular fluids flowing through the charged matrix [58,60]. Measurement of the electric potentials generated by functional levels of strain shows that the average field intensities in bone are quite small, on the order of 1 V/ (one microvolt per microstrain [59]). Considering that the adult skeleton is seldom subject to strains exceeding 2000 [22], if endogenous fields are to influence bone morphology they must do so at field intensities below 2 mV/cm (2000 V/cm). This field intensity, in and of itself, is certainly sufficient to perturb the membrane potential of an osteoblast to the point of having some biologic effect [64]. However, as described previously, a large percentage of bone tissue is rarely subject to strains greater than 500 [22], yet bone mass is retained in these areas [17]. Therefore, if each cell at each site of the bone is responsible for its own assessment of the biophysical milieu in its adjacent matrix, field intensities well below the “threshold” of 1 mV/cm would somehow have to regulate bone cell metabolism. Because of their suspected role in regulating bone morphology, electric fields have been used in the clinic for the treatment of delayed fracture unions [65]. The majority of the signals used for pseudoarthroses, delayed union, and even avascular necrosis are complex pulsed electromagnetic fields (PEMFs), which utilize time-varying magnetic fields exogenously to induce electrical currents into the local tissue. These devices induce relatively high electric field intensities (10 mV/cm), with the energy distributed over a broad frequency range, from one to more than 1,000,000 Hz. To identify the osteogenic components of the PEMF, we investigated how changes in PEMF characteristics would affect the remodeling response. In the first series of experiments, electromagnetic fields were induced in the turkey ulna model of disuse osteopenia using Helmholtz coil pairs strapped to the wing of the animal [66]. Keeping the peak
of the magnetic field levels constant yet varying their rise time, both the PEMF waveshape and, correspondingly, the power of the induced electric field were changed. Exposing the turkey ulna preparation to 1 h per day, the maximum osteogenic effect was seen with rise times of 3.8 to 5.5 ms. In contrast to the 11% loss of bone caused by 8 weeks of disuse, these waveforms stimulated a 12% increase in bone area, a net benefit over disuse exceeding 20%. Rise times above or below this range were less effective in generating bone formation and, in some cases, were incapable of inhibiting bone loss. Because changes in magnetic field rise time alter both pulse duration and pulse width in a PEMF, we considered the possibility that the osteoregulatory efficacy of the PEMF (all for a given magnetic field amplitude) was related to the spectral distribution of energy in specific PEMF signals [67]. This correlation analysis indicated that even though the frequency content of the PEMFs spanned the 1 to 250-Hz range,the component of the field energy that correlated most strongly with the ability of the PEMF’s to stimulate new bone formation was that induced at frequencies below 75 Hz. The osteogenic capability of this low frequency bandwidth was evident even though less than 0.1% of the total PEMF energy was contained in this range. Subsequent studies have validated this suspected lowfrequency affinity, with sinusoidal fields induced in the range of 15 – 35 Hz appearing the most potently osteogenic [68]. For example, a 15-Hz magnetic field signal inducing electric field intensities on the order of only 1 to 10 V/cm is more effective in stimulating new bone formation than a PEMF signal inducing fields on the order of 1 – 10 mV/cm (Fig. 5). However, at frequencies below 15 Hz (and thus more aptly associated with locomotion) the osteogenic potential decreased dramatically, such that at 5 Hz and below, induced electric fields were incapable even of preventing disuse bone loss [28]. That electric fields in the 10- 100-Hz range can affect bone remodeling activity at intensities below 10 V/cm suggests that strains below 10 — if induced within this frequency range — can play an important role in mediating bone remodeling even though these strains would be 300-fold less than the peak strains induced during maximal activity (based on 1 V/ [59]). In essence, very low-magnitude, high-frequency physical signals persist in functionally loaded bone, and very low-magnitude, high-frequency electric fields are strongly osteogenic.
E. Osteogenic Potential of Frequency The bone tissue modeling/remodeling response is sensitive only to dynamic (time-varying) strains; static strains are ignored as a source of osteogenic stimuli [52]. This observation indicates that — rather than strain magnitude per
498
CLINTON T. RUBIN
ET AL.
FIGURE 5
Microradiographs of transverse sections of the ulna midshaft following 8 weeks of 1 h per day of various electric field regimens. An ulna subject only to a “dummy” coil, resulting in a 12% bone loss via intracortical and endosteal resorption (top left). An ulna isolated from function but subject to a 75-Hz simusoidal electric field inducing 10 V/cm (bottom left) showed little modeling or remodeling activity, with a net increase in bone mass of 3%. A signal of the same magnitude but induced at 15 Hz resulted in substantial new bone formation on both endosteal and periosteal surfaces, with little evidence of intracortical porosis, resulting in a 14% increase in bone area (top right). Reproduced with permission from McLeod and Rubin [68].
se — the total number of strain events, the number of strain events per unit time, or the strain rates involved in the loading regimen may be critical to bone mass and morphology. In cortical bone, 2000 induced at 0.5 Hz (cycles/second) maintains bone mass and achieves this with just four cycles of loading encompassing 8 per day [31]. Reducing this strain to 1000 at 1 Hz requires 100 cycles, and 100, to maintain bone mass [11]. Raising the loading frequency to 3 Hz, bone mass can be retained with 1800 cycles (600 of load) with peak-induced strains of only 800 [61]. With the same 600 per day loading regimen, only 200 is necessary to maintain cortical bone mass if the strain is applied at 30 Hz, a protocol employing 18,000 cycles of loading. When these 30-Hz mechanical signals are induced for 1 h per day (108,000 cycles), only 70 is necessary to inhibit bone loss. Plotted together, these data demonstrate that the sensitivity of cortical bone to mechanical loading goes up quickly with frequency (Fig. 6), and thus much lower strains are necessary to of maintain bone mass.
These data indicate that bone is preferentially sensitive to high-frequency mechanical stimuli. However, all the experiments discussed to this point were invasive in nature, using the functionally isolated turkey ulna preparation. Unless we are able to induce these osteogenic stimuli noninvasively, application to humans in the clinic will be greatly limited. Using whole body vibration as a means of inducing mechanical signals into the skeleton, our first studies again used turkeys, but rather than invasive studies on the ulna, these animals simply stood on a small oscillating plate [69]. Over a 2-month period, each animal was subjected to a 30-Hz sinusoidal vibration for 10 min each day, five days per week. Five animals were in each of four groups to test acceleration intensity of 0.1, 0.2, 0.3, and 0.9g (where 1.0g earth’s gravitational field, or 9.8 ms2). Eight control animals remained unstimulated. When not being loaded, animals roamed their pens freely. At 0.3g, this stimulation induced approximately 5 on the cortical surface of the tibia. Dynamic indices of bone formation (labeled surface, LS; mineral apposition rate; MAR) were determined for all animals via pulsed, double labels of tetracycline.
499
CHAPTER 18 Inhibition of Osteopenia by Biophysical Intervention
FIGURE 6
The sensitivity of cortical bone tissue to mechanical strain increases with loading frequency. The plot indicates the area increase (mm2) in cortical bone measured for each additional 1 imposed on the turkey ulna, at each of five loading frequencies spanning the 1- to 60-Hz range. Another way of interpreting these data is to consider that 1/10th of the strain is necessary to maintain cortical bone mass if the strain is induced at 60 Hz, rather than 1 Hz. Trabecular bone is even more responsive to frequency. Adapted from Qin et al. [61]
FIGURE 7
Trabeculae within the proximal tibial metaphysis of the eight control animals showed 1.9% labeled surface, with no measurable mineral apposition rate, whereas trabeculae in the femoral head of controls showed 1.8% LS, with no detectable MAR (Fig. 7, see also color plate). These data demonstrate that the skeletons of these adult animals were in a state of low bone turnover. The LS in trabeculae following 8 weeks of stimulation demonstrated a linear dose response with an increase in vibration intensity, extending to 50.7% in the tibia and 44.2% in the femur at 0.9g. In contrast, the MAR, once turned on at 0.1g (1.47 m/day in the tibia, 1.36 in the femur), failed to increase further by increasing intensity. These results suggest that brief exposure to extremely small strains, two orders of magnitude below the peak strains experienced by the skeleton, if induced at sufficiently high frequency, can be as important to skeletal design as large magnitude strains seen only occasionally. Further, that the strains induced noninvasively were far less than those in the invasive protocols (e.g., the isolated turkey ulna) suggests that trabeculae may be even more sensitive to strain than cortical bone. Regardless, the influence of the mechanical milieu on skeletal morphology as it
Fluorescent photomicrographs of trabeculae within the proximal tibial metaphysis of two 1.5-year-old turkeys. (Left) Trabeculae were harvested from a control (nonoscillated) animal. (Right) Trabeculae harvested from a turkey subject to 30 days of 5 min per day of plate vibrations at 30 Hz, inducing accelerations on the order of 0.3g. The fluorescent labeling schedule is identical in each animal. While mineral apposition rate and labeled surface were essentially zero in the skeletally mature controls, these dynamic indices of formation rose to 41 3.4% of labeled surface in the loaded group, with MAR rising to 1.9 0.22 m per day (N 4). By small fluctuations in gravity, induced at the proper frequency, trabecular bone formation is stimulated noninvasively in the weight bearing skeleton. Magnification 40. Adapted from Rubin et al. [93]. (See also color plate.)
500 relates to posture, speaking, and chewing, while generating relatively small signals, may be quite dramatic, depending on how long you stand, how loudly you speak, or how much you eat. If the small strains induced by the musculature during activities such as posture, speaking, or chewing are critically important to the establishment and maintenance of the skeleton, if those signals change with age, a key regulatory signal may be lost. The short-term (8 weeks) studies on turkeys show that noninvasive, very low intensity mechanical stimuli are osteogenic if applied above 10 Hz. It is essential, however, to demonstrate that this strongly anabolic stimulus can influence not only the quantity, but the quality of the bone. To determine if long-term (12 month) stimuli will ultimately improve the structural status of the bone, we have used DXA, peripheral quantitative computed tomography (pQCT), and histology (static and dynamic histomorphometry, CT) to evaluate the skeletal effects of a low-magnitude mechanical stimulus [70]. Eighteen adult female sheep, 5–7 years of age, were randomized into two groups; experimental and untreated controls. For 20 min/day 5 day/week, experimental sheep stood constrained in a chute such that only the hind limbs were subject to a vertical ground-based vibration, oscillating at 30 Hz, to create peak–peak accelerations of 0.3g. When the animals were not being treated, they joined the controls in a pasture area. Following 1 year of stimulation, the animals were euthanized and the femora and ulnae removed. Compared to untreated controls, the bone mineral density (BMD) of the proximal femur in stimulated animals was 5.4% greater, but this difference was not significant (p 0.1). Although pQCT also failed to demonstrate a significant difference in the total density of the proximal femur ( 6.5%; p 0.1) when this assay was used to selectively evaluate cortical and cancellous bone at the lesser trochanter, a 34.2% increase in trabecular density was observed in mechanically stimulated sheep (p , 0.01). This effect was substantiated by undecalcified bone histology, which demonstrated substantial increases in trabecular bone volume and trabecular number, and sharp decreases in trabecular spacing (Fig. 8, see also color plate). This effect appeared highly selective for cancellous bone, as there were no significant changes in any cortical bone parameters. To determine if the stimulus had any benefit to bone “quality,” high-resolution three-dimensional models were made of 1-cm cubes of trabecular bone harvested from the medical condyle of the femur, composed of 300 microtomographic slices for each cube [71]. The trabecular bone pattern factor, an index of connectivity, was decreased by 24.2% in animals subject to the noninvasive stimulus (p 0.03), reflecting a significant increase in connectivity and thus an improvement in the quality of bone. True physical property measurements of the bone samples, as performed using ultrasound and material testing, substantiated these
CLINTON T. RUBIN
ET AL.
findings, showing that the elastic modulus and stiffness of the bone subject to the low-level mechanical stimulus increased by 12% ( p 0.05), and strength by 27% ( p 0.05) [72]. Indeed, both the quantity and the quality of bone were enhanced by an extremely low-level mechanical stimulus.
V. ATTENUATION OF WOLFF’S LAW BY SYSTEMIC DISORDERS As sensitive as the skeletal system appears to be to mechanical stimuli, it becomes somewhat distressing that osteopenia should become symptomatic at skeletal sites subject to the greatest mechanical demand; as bone mass diminishes, a given load will engender a greater strain. However, if the strain increases, why then does the bone disappear? Something must happen to the level of the signal (e.g., suppressed streaming potentials caused by ageinduced change in interstitial fluid viscosity), the perception of that signal by the cells within bone (compromised matrix/cell interaction), or the means of responding to those signals (attenuated cell metabolism). Evidence is building that systemic distress such as age [73], calcium deficiency [74], or endocrine imbalance [75] will dramatically affect the interaction of biophysical stimuli with the modeling/remodeling response. For example, mechanical signals that are osteogenic in the young skeleton fail to stimulate bone formation in the old skeleton. Further, the osteopenia caused by disuse is markedly pronounced when combined with calcium deficiency (Fig. 9). Finally, even the modus operandi of disuse in normal bone is affected in conditions such as endocrinopathy: instead of a few, large porotic events occurring in the cortex, a huge number of smaller events arise. Perhaps it is this attenuation of the response of bone to mechanical signals that could explain why exercise is not considered ideal osteoporosis therapy for the postmenopausal or aging population [76]. Indeed, the risk of increasing signal strength to induce a response in the frail skeleton may induce the very failure that one is trying to prevent. To examine the inability of older bone to perceive mechanical stimuli, a series of experiments were devised to quantify the osteocyte population’s perception of exogenous signals. Obviously, the amplitude of the strain itself does not diminish in the older skeleton, if anything, it gets larger as bone mass disappears. Perhaps instead it is the means by which the cells perceive this deformation as a regulatory signal that has been affected. This hypothesis has been addressed in the turkey ulna model using the in situ reverse transcript polymerase chain reaction (RTPCR). These experiments show that the mRNA expression of many matrix proteins (e.g., type I collagen, osteopontin) and cell adhesion molecules (e.g., integrins) is reduced in the osteocytes of aging cortical bone [77]. For example,
501
CHAPTER 18 Inhibition of Osteopenia by Biophysical Intervention
FIGURE 8 Following 1 year of extremely low-level mechanical stimulation, parameters of both static and dynamic histomorphometry demonstrated a significant benefit to both the quantity and the quality of bone from exposure to the biophysical stimulus, as compared to control animals (top). Fluorescent photomicrographs of a transverse section at the lesser trochanter of the femur showing more trabeculae, which are thicker, than control. (See also color plate.)
approximately 3.0% of the osteocytes were positively labeled for collagen mRNA in the 1-year-old bone, with only 1.2% labeled in the old bone. Such a small, but significant, change will hardly convince anyone of a new etiologic factor in osteoporosis. However, while 12.8% of the osteocytes in the 1-year-old bone showed positive labeling for osteopontin, this labeling dropped 84% to 2.1% of osteocytes in the old bone. Finally, positive labeling for integrin 3 dropped from 5.3% in young bone to 0.8% in old bone, an 85% decrease in message for this protein [78]. These data reflect a diminished cell – matrix interaction; while the matrix still experiences strain, the regulatory information never reaches the osteocyte. Measuring the mRNA abundance in osteocytes following the load of old bone supports this mechanism of signal perception. Subjecting the older bone to a load regimen of 2000 at 1 Hz,
which is strongly osteogenic in young adult bone, was insufficient to stimulate any new bone formation, and protein expression remained essentially unchanged from the suppressed levels observed in the intact old bone. However, changing the mechanical signal to a low-magnitude, highfrequency signal increased the levels of expression of these critical proteins and stimulated substantial new bone formation (14% over the animals’ intact control). A 500- regimen at 30 Hz increased the number of osteocytes labeled for mRNA message of integrins to 7.4%, 2% above the basal level of young bone. Further, the osteopontin message increased to 13.9%, a level equivalent to that of young bone. Finally, the collagen type I message increased to 92.4% of osteocytes with label, a 300 increase in active osteocytes. This upregulation of mRNA activity is observed as early as 3 days into the loading regimen, long before the surfaces
502
CLINTON T. RUBIN
ET AL.
modulate its coupling to the matrix and morphology of its lacunae [54], it is further evidence of the ability of the cell to autoregulate its own mechanical environment (i.e., striving for the “optimal” mechanical milieu of 100). Age or menopause-induced reduction in the presence of these proteins would diminish the ability of the cell to sense some aspects of the mechanical signals within the bone tissue. However, aged bone tissue clearly retains the ability to respond to its mechanical environment, as reflected by the robust response to a high-frequency, low-magnitude mechanical stimulus.
VI. CLINICAL APPLICATION OF BIOPHYSICAL STIMULI The observations that relatively high-frequency, lowmagnitude mechanical stimuli can influence bone formation and resorption suggest that this signal can be used clinically to inhibit or reverse osteopenia. Certainly, lowlevel biophysical signals are simpler and safer to impose into the skeletal system than large stimuli. Recalling Goldilocks, it is important to remember that, before implementing these stimuli for a disease that may require three decades of treatment, the signal should not be “too big or too small, but just right.”
A. Deterioration of Muscle Dynamics as an Etiologic Factor in Osteopenia
Microradiographs of a 100-m midshaft section of an intact turkey hen ulna fed a normal diet (top). Following a 6-week period of a low calcium diet, the intact ulna shows a substantial amount of cortical thinning, endosteal resorption, subendosteal cavitations, and intracortical resorption (middle). From the functionally isolated ulna of the same bird, a more pronounced degree of tissue loss occurs in a bone subjected to the same calcium insufficiency, but not protected by functional loading (bottom). While the biophysical regulation of bone mass and morphology occurs locally, the systemic state of the animal can have an overwhelming effect. Reproduced with permission from Lanyon et al. [74].
FIGURE 9
have begun to produce osteoid [78]. These data suggest that extracellular matrix proteins provide a means of translating the mechanical loading dynamics of the skeleton to the osteocyte population. Recalling the ability of the osteocyte to
The suppression of the cell – matrix interaction in aged bone has already suggested that strain signals will not be transduced efficiently in older bone tissue [77]. Before completely blaming the bone cells’ inability to perceive or respond to the strain environment, it is worth considering whether the aged skeleton is instead lacking a critical component of the functionally induced endogenous signal. As suggested from strain gage recordings from the appendicular skeleton, low-level, high-frequency strains arise directly from muscle dynamics [35]. Further, these persistent, low-magnitude strains have been shown to be strongly osteogenic and may represent a strong stimulus in defining the morphology of the skeleton [69 – 72]. Therefore, if there is an age (or menopause)-induced change in the dynamics of these muscle oscillations, it could be argued that bone mass may deteriorate because these muscle-based signals also attenuate. To determine the role of muscle dynamics in the etiology of osteopenia, the spectral characteristics of muscle activity as a function of age were obtained through measurements of muscle surface vibration [79]. During the contraction of a muscle, radial expansion of the individual fibers results in fiber collisions and the
503
CHAPTER 18 Inhibition of Osteopenia by Biophysical Intervention
production of muscle sound or acoustic vibrations of the muscle body. The frequency of these vibrations reflects the firing rate of the motor units and, correspondingly, the force output of the motor unit, at least up to 80% maximum voluntary contraction during isometric contraction. The acoustic vibrations normal to the surface of the soleus muscle were recorded in 40 volunteers (20 – 83 years) using a low mass accelerometer. Recordings were made from the left and right soleus, a principal leg muscle associated with posture, while the volunteer sat (passive) and stood (active). Recordings from each leg were averaged and integrated to obtain spectral power in the boundaries of 1 – 50, 1 – 25 and 25 – 50 Hz. Spectra obtained from these recordings show muscle activity in the frequency range above 20 Hz decreases by a factor of three in the elderly as compared to that seen in young adults (Fig. 10), a sarcopenia consistent with loss of fast oxidative type fibers. As the high-frequency components seen in bone strain almost certainly arise through muscle activity, the deterioration of the postural muscle contraction spectra with age would contribute to a decrease in the spectral content of strain above 20 Hz. From this perspective, it can be argued that the sarcopenia that parallels aging, may well prove to be a principal etiologic factor in osteoporosis, as this portion of the strain spectra is demonstrably osteogenic. While the association between loss of muscle dynamics and bone mass does not demonstrate a causal relationship, it provides support for the concept that the chronic activity of postural muscles may be a dominant force in controlling bone mass. Further, if aging leads to the loss of specific muscle fibers critical to the maintenance of bone mass, osteoporosis could presumably be inhibited by providing a “surrogate” for the lost spectral strain history.
B. Transmissibility of Ground Reaction Forces to the Appendicular and Axial Skeleton The nature of the weight-supporting skeleton facilitates the transmission of mechanical energy into bone tissue in a relatively direct manner. That the weight-bearing skeleton certainly subjects the skeleton to strain, a dynamic strain on the skeleton can presumably be induced by perturbing its effective gravity. In other words, the static strain in your weight-bearing skeleton would rise if you suddenly found yourself on Jupiter, with the percentage increase in strain rising proportionally to the present increase in g force. If this change in g force, and therefore the change in strain that it caused, was achieved at a frequency that was demonstrated to be osteogenic, a unique means of influencing bone mass becomes possible. Fortunately, moving 20 million people to Jupiter may not be necessary, as the strains necessary to influence bone mass appear to be so small.
FIGURE 10
Age-related changes in soleus muscle dynamics during postural activity. Whereas low-frequency (1 – 25 Hz) spectra are only slightly affected by age, high-frequency muscle dynamics (25 – 50 Hz) are reduced markedly in the elderly. If these higher frequency vibrations are the dominant source of the high-frequency, low-magnitude strains in bone, it could be argued that the pathogenesis of osteopenia is rooted in degenerative changes in the neuromuscular system rather than bone tissue per se. Adapted from Huang et al. [80].
The modulation of g force can be accomplished by placing the standing human on a platform which is made to oscillate at a specific frequency and acceleration [80]. The strains arising from dynamic alterations in g force would be transfered into the skeleton along a normal trajectory, ensuring that the stimulus is concentrated at those sites with greatest weight-bearing responsibility (e.g., femoral neck), yet weak at sites not subject to resisting gravity (e.g., cranium). While conceptually simple, it must be demonstrated that groundbased accelerations are indeed transmitted through the bones and joints of the lower appendicular skeleton; little is known of the transmissibility of ground-based vibration at frequencies above 12 Hz [81]. To rigorously establish the relationship between acceleration at the plate surface and transmission of acceleration through the appendicular and axial skeleton, accelerations were measured from the femur and spine of the human standing on a vibrating platform [82]. Force transmission to these bones was determined using accelerometers mounted on Steinman pins imbedded transcutaneously in the spinous process of L3 and the greater trochanter of the right femur of six volunteers. To determine damping as a function of posture, data were also collected from subjects while standing with bent knees. For a constant force input (18N), plate accelerations increased with frequency at both the femur and spinous process of L4 (Fig. 11). When the subject stood erect, a negligible loss of acceleration was observed in the femur and spine in the lower frequency bands, yet transmissibility fell off by as much as 40% when the frequency approached 40 Hz. When the subject was asked to stand with bent knees, transmissibility fell to below 20% at the femur and spine. Presumably, this is due at least in part to the uncoupling of the body segments, such that they are no longer working efficiently as a fixed,
504
CLINTON T. RUBIN
ET AL.
FIGURE 11
Accelerometer readings as a function of time as measured from the center of the resonant plate, as well as the vertical component of L4 and femur of a 52-kg female oscillating at 29 Hz (left). Plotting site-specific accelerations as a function of frequency (right), it is clear that as frequency increases, the transmissibility of the ground-based vibrations fall in both the hip and the spine. Around 30 Hz, about 85% reaches the femur. In both figures, the solid line is the plate, the dashed line is the spine, and the dotted line is the hip. Adapted from Rubin et al. [82].
stiff system. More importantly, these measurements confirm the ability of the standing adult skeleton to transmit a substantial fraction of ground accelerations to regions of the weightbearing skeleton most susceptible to osteopenic bone loss.
C. Inhibition of Post Menopausal Bone Loss by Extremely Low-Level Mechanical Stimlui In vivo animal work, as well as the design and development of a prototype device suitable for humans, led to a 1-year feasibility trial on a small cohort of women [84]. The ability of a low-magnitude (0.2g), high-frequency (30 Hz) mechanical stimulation to inhibit postmenopausal osteopenia was evaluated in a prospective, randomized, double blind, placebo-controlled clinical trial. Sixty-two women, 3 – 8 years postmenopausal, enrolled in the pilot study. Thirty-one women underwent mechanical loading of the lower appendicular and axial skeleton for two 10 min periods per day, induced via floor-mounted devices that produced the mechanical stimulus. Accelerations of 0.2g are just perceptible and no adverse reactions were reported. Thirty-one women received placebo devices and underwent daily treatment for the same period of time. DXA was performed on the spine (L1-4), right and left proximal femurs, and nondominant radius at baseline and at approximately 3,6, and 12 months. A full complement of DXA data was obtained for 56 of the patients (28 treatment, 28 placebo; six subjects dropped from the study for reasons not related to the device). In a post hoc analysis, a linear re-
gression of the means was used to show that lumbar spine bone mineral density (BMD) declined by 3.3% ( 0.83, n 28) in the placebo group compared to only 0.8% (
0.82, n 28) in the treated group (p 0.03), reflecting a 2.5% benefit of the biomechanical intervention. A 3.3% treatment benefit was observed in the trochanter region of the hip, with a 2.9% ( 1.2) loss observed in the placebo group, yet with a 0.4% ( 1.2) gain in the treated group (p 0.03). At the distal radius, no significant differences were observed as a function of time or between groups, emphasizing the influence to be local. Stratifying the results based on patient body mass index (BMI) (kg/m 2), end point analysis confirmed the relationship between svelte stature and a greater degree of osteoporosis [84,85]; subjects with BMI 24 lost 2.5% ( 0.6) BMD over the course of the year (Fig. 12), where those with a BMI 24 did not show any change over the 12 month period. This stratification also demonstrates the ability of mechanical stimulation to inhibit this bone loss in the group at greatest risk; in subjects with BMI 24 who were exposed to the mechanical stimulus, bone loss in the spine was not significantly different than zero ( 0.2 0.7). The 2.7% difference between placebo and treatment groups was significant at p 0.01. Treated subjects with BMI 25 showed no apparent affect of treatment, perhaps because there was no bone loss to inhibit. Overall, these results indicate the potential of a unique, non-invasive biomechanical therapy for osteoporosis, representing a nonpharmacologic means of inhibiting the rapid decline of bone mineral density that follows menopause.
505
CHAPTER 18 Inhibition of Osteopenia by Biophysical Intervention
absence of a key regulatory stimulus to the bone tissue. To a certain extent, osteopenia may arise not through the inability of bone cells to respond to mechanical stimuli, but rather through a deficiency of a regulatory signal caused by muscle wasting. Regardless, the osteogenic potential of biophysical stimuli clearly points to their potential as a unique intervention for disorders and injuries of the musculoskeletal system. Whether such biophysical intervention will supersede pharmaceutical prophylaxes is doubtful, but not impossible.
Acknowledgments FIGURE 12
Stratification based on body mass index shows that the lighter women (BMI 24) lost on the order of 2.5% bone from the spine over the course of the year. Those thinner women, when exposed to lowlevel mechanical stimulation, inhibited this loss (p 0.005). As importantly, women with a BMI greater than 24 lost no bone over the course of the year, and thus it was not possible to demonstrate the efficacy of treatment to inhibit a loss that was not occurring (p 0.36). Adapted from Rubin et al. [83].
This work has been kindly supported by grants from the National Institutes of Health (AR39278, AR41011, AR41040, and AR43498), The Whitaker Foundation, Exogen, Inc., and The National Science Foundation (PYI 865105). The authors are grateful for the contributions made by our colleagues, in particular Jack Ryaby, Susannah and Chris Fritton, Yan-Qun Sun, Yi-Xian Qin, Steven Bain, Ted Gross, Simon Turner, and Robert Recker.
VII. SUMMARY
References
The role of biophysical stimuli in the achievement and maintenance of a structurally appropriate bone mass is clear. Indeed, these factors critical to retaining an effective skeletal structure have great potential for direct clinical applications, such as in fracture healing [86,87] or osseointegration [88]. In contrast to systemic, pharmaceutical intervention such as estrogens, bisphosphonates, or calcitonin, the attributes of such biophysical prophylaxes are that they are native to the bone tissue, safe at low intensities, incorporate all aspects of the remodeling cycle, will ultimately induce lamellar bone, and the relative amplitude of the signal will subside as formation persists (self-regulating and self-targeting). However, the widespread use of biophysical stimuli in the treatment of skeletal disorders will undoubtedly be delayed until we achieve a better understanding of the mechanisms by which they act [89]. At the organ level, biophysical signals exist as a normal, physiologic component of the functional milieu; strain energy appears on the occipital ridge of the macaque, the femur of the lizard, and the metacarpal of the horse. In addition to the large amplitude strains typically associated with functional activity, a strain signal, far less than 10 in amplitude, arises through muscular activity in the frequency band of 10 to 50 Hz. This signal is present in the cranial, axial, and appendicular skeleton and persists essentially at all times including passive actions such as standing and speaking. Indeed, the sarcopenia that parallels the aging process, and more specifically, the attenuation of the 20- to 50-Hz spectral content of muscle contraction, suggests that the absence of these signals may also indicate the
1. J. Wolff, “The Law of Bone Remodeling” (Das Gesetz der Transformation der Knochen). Originally published by Verlag von August Hirshwald, Berlin, 1892. (English translation by P. Maquet and R. Furlong, published by Springer Verlag, Berlin, 1986]. 2. T. Robinson, C. Snow-Harter, D. Taaffe, D. Gillis, J. Shaw, and R. Marcus, Gymnasts exhibit higher bone mass than runners despite similar prevalence of amenorrhea and oligomenorrhea. J. Bone Miner. Res. 10(1), 26 – 35 (1995). 3. C. Snow-Harter, M. Bouxsein, B. Lewis, D. Carter, and R. Marcus, Effects of resistance and endurance exercise on bone mineral status of young women: A randomized exercise intervention trail. J. Bone Miner. Res. 7(7), 761 – 769 (1992). 4. Hargens, Recent bed rest results and countermeasure development at NASA. Acta Physiol. Scand. 150(S616), 103 – 114 (1994). 5. D. Simmons, J. Russel, F. Winter, R. Baron, A. Vignery, T. Van Thuc, G. Rosenberg, and W. Walke, Space flight and the non-weight bearing bones of the rat skeleton (Cosmos 1129). Trans. 27th Orthop. Res. Soc. 6, 65 (1981). 6. Y. Nishimura, H. Fukuoka, M. Kiriyama, Y. Suzuki, S. Oyama, S. Ikawa, M. Higurashi, and A. Gunji, Bone turnover and calcium metabolism during 20 days bed rest in young healthy males and females. Acta Physiol. Scand. 150(S616), 27 – 35 (1994). 7. H. Jones, J. Priest, W. Hayes, C. Ticheno, and D. Nagel, Humeral hypertrophy in response to exercise. J. Bone J. Surg. 59A, 204 – 208 (1977). 8. Huddleston, D. Rockwel, D. Kurlund, and J. Harrison, Bone mass in lifetime tennis athletes. J. Am. Med. Assoc. 244, 1107 (1980). 9. P. Engh, Massin, and K. Suthers, Roentgenographic assessment of the biologic fixation of poroussurfaced femoral components. Clin. Orthop. Rel. Res. 257, 107 – 128 (1990). 10. J. Callaghan, The clinical results and basic science of total hip arthroplasty with porous coated prostheses. J. Bone J. Surg. 75A(2), 299 – 310 (1993). 11. L. Rubin, Lanyon, Regulation of bone mass by mechanical strain magnitude. Calcif. Tissue Int. 37, 411 – 417 (1985). 12. J. O’Conner, L. Lanyon, and H. MacFie, The influence of strain rate on adaptive bone remodelling. J. Biochem. 15, 767 – 781 (1982).
506 13. S. Pollack, D. Berretta, H. Steinberg, D. Steinberg, and C. Brighton, Osteoporosis and the electrical double layer. 6th Trans. Biol. Reprod. Growth Soc. (1986). 14. Bassett, Biologic significance of piezoelectricity. Calcif. Tissue Int. 1, 252 – 272 (1971). 15. S. Weinbaum, S. Cowin, and Y. Zeng, A model for the excitation of osteocytes by mechanical loading-induced bone fluid shear stresses. J. Biomech. 27(3), 339-360 (1994). 16. Fyhrie and D. Carter, A unifying principle relating stress to trabecular bone morphology. J. Orthop. Res. 4, 304 – 331 (1986). 17. T. S. Gross, J. L. Edwards, K. J. McLeod, and C. T. Rubin, Strain gradients correlate with sites of periosteal bone formation. J. Bone Miner. Res. 12(6), 982 – 988 (1997). 18. D. Carter, W. Harris, R. Vasu, and W. Caler, The mechanical and biological response of cortical bone to in vivo strain histories, In Mechanical Properties of Bone (S. Cowin, ed.) Vol. 45, p. 81. Am. Soc. Mech. Eng. 1981. 19. L. Lanyon and R. Smith, Bone strain in the tibia during normal quadrupedal locomotion. Acta Orthop. Scand. 41, 238 – 248 (1970). 20. T. Gross, K. McLeod, and C. Rubin, Characterizing the bone strain distributions in vivo using three triple rosette strain gages. J. Biomech. 25, 1081 – 1087 (1992). 21. C. Rubin, Skeletal strain and the functional significance of bone architecture. Calcif. Tissue Int. 36, 11 – 18 (1984). 22. C. Rubin and L. Lanyon, Dynamic strain similarity in vertebrates: An alternative to allometric limb bone scaling. J. Theor. Biol. 107, 321 – 327 (1984). 23. C. Rubin, K. McLeod, and S. Bain, Functional strains and cortical bone adaptation: Epigenetic assurance of skeletal integrity. J. Biomech. 23, 43 – 54 (1990). 24. S. Cowin, Bone remodeling of diaphyseal surfaces by torsional loads: Theoretical predictions. J. Biomech. 20, 1111 – 1120 (1987). 25. R. Huiskes, H. Weinans, H. Grootenboer, M. Dalstra, B. Fudala, and T. Slooff, Adaptive bone remodeling theory applied toprosthetic:Design analysis. J. Biomech. 20, 1135 – 1150 (1987). 26. Beaupre, T. Orr, and D. Carter, An approach for timedependent bone modeling and remodeling: Theoretical development. J. Orthop. Res. 8, 651 – 661 (1990). 27. M. Van Der Meulen, G. Beaupre, and D. Carter, Mechanobiologic influences in long bone cross-sectional growth. Bone 14, 635 – 642 (1993). 28. C. Rubin and K. McLeod, Endogenous control of bone morphology via frequency specific, low magnitude functional strain, In “Human Biology” (H. Weinans, ed.), pp. 79 – 90. World Scientific, Singapore, 1995. 29. C. Rubin and L. Lanyon, Limb mechanics as a function of speed and gait: Functional strains in the radius and tibia of horse and dog. J. Exp. Biol. 101, 187 – 211 (1982). 30. Biewener, Biomechanics of mammalian terrestrial locomotion. Science 250, 1097 – 1103 (1990). 31. C. Rubin and L. Lanyon, Regulation of bone formation by applied dynamic loads. J. Bone J. Surg. 66A, 397 – 402 (1984). 32. C. Rubin and L. Lanyon, Osteoregulatory nature of mechanical stimuli: Function as a determinant for adaptive remodeling in bone. J. Orthop. Res. 5, 300 – 310 (1987). 33. C. Rubin and K. McLeod, Biologic modulation of mechanical influences in bone remodeling, In “Biomechanics of Diarthroidal Joints” (V. Mow, A. Ratcliffe, and S. Woo, eds.), Chap. 21, pp. 97 – 118. Springer-Verlag, New York, 1990. 34. T. Skerry, R. Suswillo, A. El Haj, N. Ali, R. Dodds, and L. Lanyon, Load-induced proteoglycan orientation in bone tissue in vivo and in vitro. Calcif. Tissue Int. 46, 318 – 326 (1990). 35. S. P. Fritton, K. J. McLeod, and C. T. Rubin, Self-similarity in longterm strain recordings from the skeleton of disparate species: Evidence for the self-organization in bone. J. Biomech. 33, 317 – 326 (2000). 36. H. Donahue, K. McLeod, C. Rubin, J. Andersen, E. Grine, E. Hertzberg, and P. Brink, Cell to cell communication in osteoblastic
CLINTON T. RUBIN
37. 38.
39.
40. 41.
42.
43.
44. 45. 46.
47.
48.
49.
50.
51.
52. 53. 54.
55.
56.
57.
ET AL
networks: Cell line dependent hormonal regulation of gap junctions composed of connexin43. J. Bone Miner. Res. 10, 881 – 889 (1995). T. Gross, and C. Rubin, Uniformity of resorptive bone loss induced by disuse. J. Orthop. Res. 13(5), 708 – 714 (1995) H. Donahue, K. McLeod, C. Rubin, J. Andersen, E. Grine, E. Hertzberg, and P. Brink, Cell to cell communication in osteoblast networks: Cell line dependent hormonal regulation of gap junctions composed of connexin 43. J. Bone Mineral Res. 10(6), 881 – 889 (1995). C. T. Rubin, Y-Q. Sun, M. Hadjiargyrou, and K. McLeod, Increased expression of matrix metalloproteinase-1 mRNA in osteocytes precedes bone resorption as stimulated by disuse: Evidence for autoregulation of the cell’s mechanical environment? J. Orth op. Res. 17, 354 – 361 (1999). D. Carter and M. Wong, Mechanical stresses and endochondral ossification in the chondroepiphysis. J. Orthop. Res. 6, 148 – 154 (1988). D. Adams, A. Spirt, T. Brown, S. Fritton, C. Rubin, and R. Brand, Testing the “daily stress stimulus” hypothesis of bone remodeling with natural and experimentally controlled strain histories. J. Biomech. 30, 671 – 678 (1997). S. Woo, W. Akeson, R. Coutts, L. Rutherford, D. Doty, G. Jemmott, and D. Amiel, A comparison of cortical bone atrophy secondary to fixation with plates with large differences in bending stiffness. J. Bone J. Surg. 58A, 190 – 195 (1976). D. Carter, W. Harris, R. Vasu, and W. Caler, Stress fields in the implated and plated canine femur calculated from in vivo strain measurements. J. Biomech. 14, 63 – 70 (1981). L. Lanyon, A. Goodship, C. Pye, and J. MacFie, Mechanically adaptive bone remodeling. J. Biomech. 15(3), 141 – 154 (1982). J. Hert, M. Liskova, and J. Landa, Continuous and intermittent loading of tibia in rabbit, Folia Morphol. (Prague). 19, 290 – 317 (1971). A. Churches, C. Howlett, K. Waldron, and G. Ward, The response of living bone to controlled time varying loading. J. Biomech. 12, 35 – 45 (1979). C. Turner, M. Akhter, D. Raab, D. Kimmel, and R. Recker, A noninvasive, in vivo model for studying strain adaptive bone remodeling. Bone 12, 73 – 79 (1991). S. Goldstein, L. Matthews, J. Kuhn, and S. Hollister, Trabecular bone remodeling: An experimental model. J. Biomech. 24, 135 – 150 (1991). T. Chambers, M. Evans., T. Gardner, A. Turner-Smith, and J. Cho, Induction of bone formation in rat tail vertebrae by mechanical loading. Bone Miner. 20, 167 – 178 (1993). C. Rubin, H. Donahue, J. Rubin, and K. McLeod, Optimization of electric field parameters for the control of bone remodeling: Exploitation of an indigenous mechanism for the prevention of osteopenia. J. Bone Miner. Res. 8, S573 – S581 (1993). T. Brown, D. Pedersen, M. Gray, R. Brand, and C. Rubin, Identification of mechanical parameters initiating periosteal remodeling. J. Biochem. 23, 893 – 905 (1990). L. Lanyon and C. Rubin, Static versus dynamic loads as an influence on bone remodeling. J. Biomech. 17, 897 – 906 (1985). H. Frost, Perspectives: Bone’s mechanical usage windows. Bone Miner. 19, 257 – 271 (1992). C. Rubin, T. Gross, K. McLeod, and S. Bain, Stimulation of lamellar bone formation by potent mechanical stimuli: Morphologic stages in the achievement of lamellar bone. J. Bone Miner. Res. 10(3), 488 – 495 (1995). C. Rubin, T. Gross, Y. Qin, S. Fritton, F. Guilak, and K.J. McLeod, Differentiation of bone-tissue remodeling response to axial and torsional loading in the turkey ulna. J. Bone Jt. Surg. 78, 1523 – 1533 (1996). E. Burger, J. Klein-Nulend, and J. Veldhuuzen, Modulation of osteogenesis in fetal bone rudiments by mechanical stress in vitro. J. Biomech. 24 (Suppl. 1), 101 – 109 (1991). P. Davies and S. Trepathi, Mechanical stress mechanisms and cell: An endothelial paradigm. Circ. Res. 27, 239 – 245 (1993).
CHAPTER 18 Inhibition of Osteopenia by Biophysical Intervention
507
58. Y. Mak, Streaming potential in bone. Exercise Sports Sci. 17, 175 – 194 (1989). 59. M. Otter, V. Palmieri, D. Wu, K. K. Siez, L. MacGinitie, and G. Cochran, A comparative analysis of streaming potentials in vivo and in vitro. J. Orthop. Res. 10, 710 – 719 (1992). 60. R. Salzstein, S. Pollack, A. Mak, and N. Petrov, Electromechanical potentials in cortical bone. 1. A continuum approach. J. Biochem. 20(3), 261 – 270 (1987). 61. Y.-X. Qin, C. Rubin, and K. McLeod, Nonlinear dependence of loading intensity and cycle number in the maintenance of bone mass and morphology. J. Orthop Res. 16, 482 – 489 (1998). 62. Y. Zeng, S. Cowin, and S. Weinbaum, A fiber matrix model for fluid flow and streaming potentials in the canaliculi of an osteon. Ann. Biomed. Eng. 22, 280 – 292 (1994). 63. A. Yasuda, Piezoelectric activity of bone. J. Jpn. Orthop. Surg. Soc. 28, 267 (1954). 64. K. McLeod, H. Donahue, P. Levin, M. Fontaine, and C. Rubin, Electric fields modulate bone cell function in a density dependent manner. J. Bone Miner. Res. 8(8), 977 – 984 (1993). 65. A. Bassett, Beneficial effects of electromagnetic fields. J. Cell. Biochem. 51, 387 – 393 (1993). 66. C. Rubin, K. McLeod, and L. Lanyon, Prevention of osteoporosis by pulsed electromagnetic fields. J. Bone J. Surg. 71A(3), 411 – 418 (1989). 67. K. McLeod and C. Rubin, Frequency specific modulation of bone adaptation by induced electric fields. J. Theor. Biol. 145, 385 – 396 (1990) 68. K. McLeod and C. Rubin, The effect of low-frequency electrical fields on osteogenesis. J. Bone J. Surg. 74A, 920 – 929 (1992). 69. C. Rubin, C. Li, Y. Sun, C. Fritton, and K. McLeod, Non-invasive stimulation of trabecular bone formation via low magnitude, high frequency strain. 41st Orthop. Res. Soc. 20(2), 548 (1995). 70. C. Rubin, S. Turner, S. Bain, C. Mallinekrodt, and K. McLeod, Extremely low-level mechanical signals are anabolic to trabecular bone. Nature (in press) (2001). 71. C. Rubin, A. S. Turner, R. Muller, Y-X. Qin, and K. McLeod, Femoral bone density and trabecular number are increased by non-invasive low level mechanical stimuli, as quantified by micro-computed tomography. Trans. Orthop. Res. Soc. 25, 715 (2000). 72. Y. X. Qin, R. Mauser, C. Berman, W. Lin, and C. Rubin. The relationship between bone mineral density and ultrasonic velocity. Trans. Orthop. Res. Soc. 25, 568 (2000). 73. C. Rubin, S. Bain, and K. McLeod, Suppression of the osteogenic response in the aging skeleton. Calcif. Tissue Int. 50, 306 – 313 (1992). 74. L. Lanyon, C. Rubin, and G. Baust, Modulation of bone loss during calcium insufficiently by controlled dynamic loading. Calcif. Tissue Int. 38, 209 – 216 (1986). 75. S. Bain and C. Rubin, Metabolic modulation of disuse osteopenia: Endocrine-dependent site specificity of bone remodeling. J. Bone Miner. Res. 5(10), 1069 – 1075 (1990).
76. G. Rodan, Mechanical loading, estrogen deficiency, and the coupling of bone formation to bone resorption. J. Bone Miner. Res. 6(6), 527 – 530 (1991). 77. Y-Q. Sun, K. McLeod, and C. Rubin, Mechanically induced periosteal bone formation is paralleled by the upregulation of collagen type one mRNA in osteocytes as measured by in situ reverse transcript-polymerase chain reaction. Calcif. Tissue. Int. 57, 456 – 462 (1995). 78. C.T. Rubin, J. Zhi, and M. Hadjiargyrou, Expression of a novel and a known gene, upregulated by disuse, is downregulated by anabolic mechanical stimulation. J. Bone Miner. Res. 14(S1), S522 (1999). 79. Y-Q. Sun, K. J. McLeod, and C. Rubin, Rapid enhancement of coupling of osteocyte to matrix. Trans. 42nd. Ann. Mtg. Ortho. Res. Soc. 21(2), 267 (1996). 80. R. Huang, K. McLeod, and C. Rubin, Changes in the dynamics of muscle contraction as a function of age. A contributing factor to the etiology of osteoporosis? J. Geronto 54A(8), B352 – B357 (1999). 81. J. Fritton, C. Rubin, Y-X. Qin, and K. McLeod, Whole body vibration in the skeleton: Development of a resonance-based testing device. Ann. Biomed Eng. 25(5), 831 – 839 (1997). 82. M. Griffin, “Handbook of Human Vibration.” Academic Press, San Diego, (1990). 83. C. Rubin, K. McLeod, M. Pope, M. Magnysson, M. Rostedt, C. Fritton, and T. Hansson, Transmissibility of ground vibration to the axial and appendicular skeleton: An alternative strategy for the treatment of osteoporosis. 18th Am. Soc. Biomech. 79 – 80 (1994). 84. C. Rubin, R. Recker, D. Cullen, J. Ryaby, and K. McLeod, Prevention of bone loss in a postmenopausal population by low-level biomechanical intervention. Am. Soc. Bone Miner. Res. 23, 1106 (1998). 85. J. Aloia, and E. Flaster, Estimating the risk of fracture in osteopenic patients. Endocrinologist 5, 297 – 402 (1995). 86. P. Martin, C. Verhas, C. Als, I. Geerts, J. Paternot, and P. Bergmann, Influence of patient’s weight on measurements of bone mineral density. Osteopor. Int. 3, 199 – 203 (1993). 87. J. Heckman, J. Ryaby, J. McCabe, J. Frey, and R. Kilcoyne, Acceleration of tibial fracture healing by noninvasive low intensity ultrasound. J. Bone J. Surg. 76A(1), 26 – 34 (1994). 88. A.E. Goodship, T. Lawes, and C.T. Rubin, Low magnitude high frequency mechanical stimulation of endochondral bone repair. Trans. 43rd Orthop. Res. Soc. 22(1), 234 (1997). 89. C. Rubin and K. McLeod, Promotion of bony ingrowth by frequency specific, low amplitude mechanical strain. Clin. Orthop. Rel. Res. 298, 165 – 174 (1993). 90. J. Rubin, C. Rubin, and K. McLeod, Biophysical modulation of cell and tissue structure and function. Crit. Rev. Eukar. Gene Expr. 5, 177 – 191 (1995). 91. C. Rubin, T. Gross, H. Donahue, H. Guilak, and K. McLeod, Physical and environmental influences on bone formation. In “Bone Formation and Repair”, pp. 61 – 78. Amer. Acad. Orth. Surg. (1994).
CHAPTER 19
Biomechanics of Age-Related Fractures MARY L. BOUXSEIN
Orthopedic Biomechanics Laboratory, Department of Orthopedic Surgery, Beth Israel Deaconess Medical Center, and Harvard Medical School, Boston, Massachusetts 02215
I. Introduction II. Biomechanics of Bone: Basic Concepts and Age-Related Changes III. Biomechanics of Hip Fractures
IV. Biomechanics of Vertebral Fractures V. Summary and Clinical Implications References
I. INTRODUCTION
to the forces applied to the bone, as well as to its loadbearing capacity, are important determinants of fracture risk. In support of this concept, clinical studies have repeatedly shown that factors related to skeletal fragility, as well as to the loads applied to the skeleton, are important determinants of fracture risk [4 – 9]. Insight into the relative contributions of skeletal fragility versus skeletal loading may be gained by using a standard engineering approach for evaluating the risk of structural failure. To design a structure, engineers must consider the size and geometry of the structure, the materials from which it is to be made, and the types of loads to which it will be subjected. Using this information, the loads applied to the structure during its normal usage can be compared to the loads known to cause failure. This comparison of applied load versus failure load gives an estimate of how “safely” the structure is designed. If a structure’s design appears “unsafe,” it may be necessary to change the geometry of the structure (e.g., increase its size), use stronger materials, or reduce the applied loads. In practice, it is often difficult to estimate precisely the strength of a structure and the loads applied to it. Therefore, to reduce the likelihood of
Age-related fractures represent an immense and increasing public health issue. In the United States alone, there are an estimated 1.5 million fractures annually, with associated medical expenditures of nearly $14 billion [1,2]. Based on current demographic trends, the number of fractures and their associated costs are projected to double or triple in the near future [3]. Most importantly, the consequences of these fractures are enormous, as those who suffer fractures experience increased mortality rates, chronic pain and disability, and a decreased quality of life. Strategies designed to prevent fractures must be based on a sound understanding of their etiology. From an engineering viewpoint, fractures of any type are due to a structural failure of the bone. This failure occurs when the forces applied to the bone exceed its load-bearing capacity. The load-bearing capacity of a bone depends primarily on the material that comprise the bone (and its corresponding mechanical behavior), the geometry of the bone (its size, shape, and distribution of bone mass), and the specific loading conditions (Fig. 1). Thus it is clear that factors related
OSTEOPOROSIS, SECOND EDITION VOLUME 1
509
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
510
MARY L. BOUXSEIN
FIGURE 1 Characteristics of the spine that determine the capacity to carry load: The trabecular bone (left); the design and organization of the vertebral body (middle); and the loading conditions (right), which are illustrated as lifting in this figure, but could be any loading action. From Myers and Wilson [159].
unexpected failure, structures are often designed with very high safety factors. To apply these concepts in the study of the etiology of fractures, Hayes and colleagues introduced a parameter called the “factor of risk” [10]. The factor of risk, , is defined as the ratio of the load delivered to a bone (applied load) to the load-bearing capacity of that bone (failure load): applied load / failure load. Thus, when the factor of risk is low (1), the forces applied to the bone are much lower than those required to fracture it, and the bone is at low risk for fracture. However, when the factor of risk is high (1), fracture of the bone is predicted to occur. A high factor of risk can occur either when the bone is very weak and its load-bearing capacity is compromised or when very high loads, such as those resulting from trauma, are applied to the bone. In elderly individuals, it is likely that the coupling of a weak bone with an increased incidence of traumatic loading leads to the dramatic rise in fracture incidence with age [11,12]. To apply the factor of risk concept in studies of hip and vertebral fracture, the loads applied to the bone of interest and the corresponding load required to fracture the bone must be identified. For example, the majority of hip fractures are associated with a fall. Therefore, to compute the factor of risk for hip fracture due to a fall, information about the loads applied to the femur during a fall and about the load-bearing capacity of the femur in a fall configuration is required. While this approach is relatively easy to
conceive, in practice it is difficult to apply. There are surprisingly few data describing the magnitude and direction of loads applied to the skeleton during the activities of daily living and even fewer data describing the loads engendered during traumatic events, such as a trip, slip, or fall. Moreover, due to the complex morphology of the skeleton and associated muscle and tendon attachments, it is difficult to design a laboratory study that mimics the loading environment encountered by the bone in vivo. Therefore, it is challenging to determine the load-bearing capacity of skeletal elements under realistic loading conditions. Moreover, because these are “biologic structures,” both applied loads and structural capacity can change with aging, pharmacologic intervention, and disease. Nevertheless, despite these uncertainties and limitations, rough estimates of the factor of risk for hip and vertebral fracture can be derived to provide insights into the complex roles of loading severity and skeletal fragility in the etiology of age-related fractures. This chapter reviews clinical and laboratory studies related to the biomechanics of age-related fractures. It first presents basic concepts related to the biomechanics of bone, including a summary of the factors that determine the material and structural behavior of bone. It then evaluates the roles of skeletal loading and bone fragility as they relate to hip and spine fractures. These sections discuss the factors that are related to the loads applied to the skeleton, either through traumatic events or everyday activities; the factors that are related to the structural capacity of skeletal elements; and how these factors interact to influence fracture risk.
CHAPTER 19 Biomechanics of Age-Related Fractures
511
II. BIOMECHANICS OF BONE: BASIC CONCEPTS AND AGE-RELATED CHANGES
return to its original shape with no residual deformation. The slope of the load – deformation curve in this elastic region defines the structural stiffness (or rigidity) of the bone. In contrast to a bone’s behavior in the elastic region, beyond the yield point, the bone undergoes permanent deformation and will not return to its original shape even when the load is removed completely. At this point, the bone is said to be in the plastic region. If the load continues to increase, the ultimate or failure load is reached, after which the structure often fails catastrophically. The energy required to produce structural failure (“yield”) is computed as the area under the load – displacement curve and is sometimes referred to as the work to fracture. To determine the mechanical behavior of bone material itself, the geometry of the specimen must be accounted for. Thus, mechanical tests are conducted on specimens of a standardized geometry under controlled conditions. As the load is applied, the specimen deforms and internal forces are generated within the specimen. The resulting relative deformation at any point is called the strain at that point. The “intensity” of the internal forces is referred to as the stress at that point. The material properties are analogous to the structural properties discussed earlier except that the properties are determined from a plot of stress versus strain instead of load vs deformation. In practice, the load – deformation curve can be converted to a stress – strain curve by correcting for specimen geometry by applying appropriate formulas for converting load to stress and deformation to strain. For example, for a specimen loaded in compression, stress is equal to the applied load divided by the crosssectional area of the specimen and strain is equal to the deformation divided by the original length of the specimen (Fig. 3). The resistance of the material to deformation is described by the elastic (or Young’s) modulus, defined as the slope of the stress – strain curve in the elastic region. As the load is increased, the specimen undergoes permanent deformation and begins to yield. If the load is increased beyond the yield point, the specimen will eventually fail, at which point the strength or ultimate stress and ultimate (or failure) strain can be determined. The biomechanical property termed toughness reflects the amount of work per unit volume of material required to yield or fracture the specimen and can be computed as the area under the stress – strain curve. Tough bone will be more resistant to fracture, although it may yield at a lower stress and, according to that measure, be considered weaker. In addition, examining the pre- and postyield regions of the stress – strain curve may provide information regarding the tendency of the bone material to accumulate damage and the mechanisms underlying its failure. A material that fractures soon after yielding, and therefore sustains little postyield strain before fracture, is termed brittle. In contrast, a material that sustains relatively large postyield strains before fracturing is considered ductile.
A. Structural vs Material Behavior To understand the nature of skeletal fragility, it is important to distinguish between factors that affect the mechanical behavior of a whole bone as a structure (structural behavior) and those that affect the mechanical behavior of the bony tissue itself (material behavior) [13]. In general, structural properties are determined by both the size and the shape of the bone, along with the mechanical properties of the tissue that comprise the bone. As such, a bone’s structural properties are in large part determined by the amount of bone present. In comparison, the material properties of bone tissue are independent of specimen size and shape, thereby reflecting the intrinsic characteristics of the bony tissue itself. During any physical activity, a complex distribution of forces, or loads, is applied to the skeleton, and with the imposition of these forces, the skeleton undergoes deformations. It is this relationship between the forces applied to a bone and the resulting deformations — characterized by a load – deformation curve — that define the structural behavior of the whole bone (Fig. 2). The load – deformation curve reflects the amount of load needed to produce a unit deformation. As mentioned earlier, the shape of this curve depends on the size/shape of the bone, as well as the properties of the tissue that comprise it. Generally, load and deformation are linearly related until the yield point is reached, at which time the slope of the load – deformation curve is reduced. Before the yield point, the bone is considered to be in the elastic region and, if unloaded, would
FIGURE 2
The load vs deformation plot is used to describe the structural behavior of a specimen. The elastic region is distinguished from the plastic region by the yield region. In the elastic region, when the load is removed there will be no residual deformation and the bone will return to its original shape. In contrast, in the plastic region, the bone will undergo permanent deformations that will remain even if the load is removed.
512
MARY L. BOUXSEIN
FIGURE 3 Typical stress versus strain diagram for longitudinally (L) and transversely (T) oriented specimens of cortical bone from the diaphysis. For specimens tested in compression, load and displacement are converted to stress and strain by dividing by the cross-sectional area and original length of the specimen, respectively. The figure shows the inherent anisotropy in bone, as specimens testing in the longitudinal direction are significantly stronger than those tested in the transverse direction.
The elastic properties of isotropic materials, such as steel or rubber, are the same in all directions. The elastic properties of bone, however, depend on the orientation of the material with respect to the loading direction. Materials whose elastic properties are sensitive to loading direction are referred to as anisotropic materials. For example, cortical bone from the femoral diaphysis has a higher modulus and is stronger when loaded in the longitudinal direction than when loaded in the transverse direction [14 – 16] (Fig. 3). The anisotropic nature of bone reflects its function as a load-bearing structure, as it is generally strongest in the primary loading direction. Hence, the degree of anisotropy in bone varies with anatomical site and functional loading [17 – 19]. For instance, human trabecular bone from the vertebral body is much stronger in the vertical direction than in the transverse direction [20 – 22], yet trabecular bone from the iliac crest and central femoral head are nearly isotropic [18,23]. In a heterogeneous material such as bone, the definition of material properties is not straightforward. In describing the properties of bone as a tissue, one could consider the mechanical properties of single trabeculae, the calcified bone matrix, or small specimens of cortical or trabecular bone. For purposes of this review, we consider bone “material” to include the calcified bone matrix, the marrow spaces in trabecular bone, and Haversian and Volkmann’s canals in cortical bone [24]. With this approach, we take a continuum mechanics view of bone in that the specimen is small enough to be homogeneous (uniform), but large enough to include a sufficient number of trabeculae (for trabecular bone) or osteons (for cortical bone) to characterize the overall material behavior.
B. Age-Related Changes in the Material Properties of Bone The elastic modulus and ultimate strength of cortical [25 – 32] and cancellous [21,22,33 – 37] bone decrease with increasing age in both men and women. In human cortical bone from the femoral middiaphysis, the tensile and compressive strengths (Fig. 4) and elastic moduli decrease approximately 2% per decade after age 20 [25]. In addition, the incurred deformation and energy absorbed before
FIGURE 4
Age-related changes in the ultimate stress of human femoral cortical bone in tension and compression (error bars represent 1 SD). The mean change per decade is -2.1% for tension and -2.5% for compression. These data indicate that femoral cortical bone becomes weaker with age. Data from Burstein et al. [25].
513
CHAPTER 19 Biomechanics of Age-Related Fractures
is characterized by a decline in the apparent density of cancellous bone. It appears that the amount of bone is reduced and therefore the integrity of the trabecular network is compromised; however, the remaining bone is histologically normal. Relevant to this is the observation, first reported by Carter and Hayes [43,44] and later confirmed by others [36,45 – 47], of a nonlinear relationship between bone density and strength (Fig. 5), whereby a given change in bone density leads to a relatively greater change in bone strength. In support of this, Mosekilde and colleagues conducted a series of experiments demonstrating an age-related decline in the mechanical properties of vertebral trabecular bone [21,22,34,36,40,41]. In one study [36], they found that the ash density declines approximately 50% from ages 20 to 80, whereas the material properties (compressive elastic modulus, ultimate stress, and energy to failure) decrease approximately 75 – 90% (Table 1). In trabecular bone of the proximal tibia, an age-related decline in apparent density of 25% is accompanied by a 30 – 40% reduction in compressive strength and energy absorption properties [37]. In addition, the strength anisotropy of trabecular bone from human lumbar vertebrae increases with age, as the ratio of compressive strengths of vertically and horizontally loaded specimens increases from about 2 at age 20 to 3.5 at age 80 [20,21]. This observation may reflect age-related changes in the trabecular architecture of vertebral bodies, whereby horizontally oriented trabeculae thin and disappear to a greater extent than vertically oriented trabeculae [36,40,48,49]. Finally, it is important to note that in trabecular bone specimens from the iliac crest that were matched pairwise for density, yield stress was approximately 40% lower in specimens from older donors (60 years) compared with younger donors (40 years) [39]. These data suggest that factors other than decreased bone density may
FIGURE 5 Compressive modulus as a function of apparent density for trabecular bone specimens from a wide variety of species and anatomic sites. In general, the modulus varies as a power-law function of density, with an exponent of approximately two. From Keaveny and Hayes [203], with permission.
fracture decrease approximately 5 – 12% per decade, suggesting the bone becomes more brittle and less tough with increasing age [25,31,32]. Moreover, the energy required to fracture a cortical bone specimen under impact loading decreases three-fold between the ages of 3 and 90 [38]. These changes in the elastic and ultimate properties of cortical bone are likely the result of porosity increases with age. McCalden and colleagues [31] found that age was strongly correlated with porosity (r 0.73) and that porosity explained over 75% of the variability in cortical bone strength. In summary, age-related changes in cortical bone lead to a weaker, more brittle material. Human cancellous bone exhibits a similar age-related decline in material properties [21,22,34 – 37,39 – 42]. Aging TABLE 1
Age-Related Changes in Vertically Oriented Trabecular Bone Specimens Compressed in Either the Vertical or Horizontal Directiona Specimens compressed in vertical direction % per decade
Specimens compressed in horizontal direction
Correlation with age (r)
% per decade
Correlation with age (r)
8.7
0.85b
8.7
Not reported
Ultimate stress
12.8
0.79
15.5
0.87c
Elastic modulus
13.5
0.83
15.9
0.83c
Energy to failure
14
0.75
c
15.2
0.88c
0.45b
3.1
0.30d
Ash density
Ultimate strain
4
c c
a The mean percentage change per decade and the linear correlation with age are presented. Specimens were taken from 42 persons, aged 15 to 87. Data from Mosekilde et al. [36]. b p0.01. c p0.001. d 0.05p0.06.
514
MARY L. BOUXSEIN
contribute to the age-related decline in material properties of trabecular bone [46,47,50,51].
C. Factors That Influence the Mechanical Behavior of Bone as a Material Whereas the age-related changes in the material properties of cortical and trabecular bone are influenced by many factors, the major determinant of the mechanical properties of bone is its porosity or apparent density. The mechanical properties of cortical bone are strongly related to porosity and the degree of matrix mineralization [52 – 56]. Over 80% of the variation in the elastic modulus of cortical bone can be explained by a power – law relationship with mineralization (defined as calcium content) and porosity as explanatory variables [52,53]. In general, with increasing age, the degree of mineralization of the matrix increases, leading to stiffer, but more brittle material behavior [26,57]. The material properties of cancellous bone are also determined to a great extent by bone density. As mentioned previously, power – law relationships with bone density as the explanatory variable explain 60 – 90% of the variation the modulus and strength of cancellous bone [36,43 – 46]. These power – law relationships indicate that small changes in apparent density can lead to dramatic changes in mechanical behavior. For instance, a 25% decrease in apparent density, approximately equivalent to 15 – 20 years of agerelated bone loss [58,59], would lead to a 44% decrease in the strength of cancellous bone. Given the anisotropic nature of trabecular bone and the variation in predicted modulus for a given density [39,60], it is clear that density alone cannot explain all of the variability in the mechanical behavior of trabecular bone. Empirical observations and theoretical analyses indicate that trabecular architecture plays an important role in determining the mechanical properties of trabecular bone [46,47,50]. Trabecular architecture can be characterized by the thickness, number, and separation of the individual trabecular elements, as well as the extent to which these elements are interconnected. Advances in nondestructive, high-resolution imaging techniques have provided new insights into the relative influence of architecture and density on age-related changes in the mechanical behavior of cancellous bone [19,61 – 64]. However, defining the precise role of microarchitecture in prediction of the mechanical behavior of bone and its influence on fracture risk is complicated by the fact that microarchitecture characteristics are strongly correlated to each other and to bone density. As such, changes in trabecular architecture accompany the age-related declines in bone density. Trabecular number, trabecular thickness, and connectivity all decline with decreasing density, whereas trabecular separation and anisotropy increase [40,41,51,60,65 – 69]. Previous studies using architectural
features derived from a model that assumes that cancellous bone architecture is “plate like” suggested that architectural features provided only modest improvements in the prediction of mechanical properties over those provided by bone density alone [67,68]. However, these previous findings should be interpreted with caution, as recent data indicate significant differences in structural indices derived from the traditional plate model compared to those computed directly from high-resolution three-dimensional images [70]. For instance, Ulrich et al. [19] reported that indices of trabecular structure determined directly from three-dimensional micro-computed tomography data significantly improve the prediction of the mechanical behavior of cancellous bone specimens from several skeletal sites. Several studies have indicated that trabecular architecture differs in fracture subjects compared to those who have not suffered a fracture [71 – 73]. However, few of these studies have controlled for the confounding influence of differences in bone density between the two groups and few have investigated microarchitecture at the sites of fracture. Ciarelli and colleagues [74] measured microarchitecture of cancellous bone specimens from the femoral neck in subjects with hip fracture compared to unfractured autopsy patients. Whereas there were no differences in trabecular thickness, number, separation, or connectivity among samples matched for equal bone density, the degree of anisotropy differed between the two groups even after controlling for density differences. These data suggest a role for trabecular architecture in the etiology of fractures that may be independent of changes in bone density. Clearly, this is an area of great interest, and additional studies are required to define the role of in vivo assessments of trabecular architecture in the prediction of fracture risk [71,75 – 78]. As mentioned previously, because changes in trabecular architecture are strongly intercorrelated, it is difficult to discern the relative effect on bone strength of reductions in trabecular number versus trabecular thickness for both vertically and horizontally oriented trabecular struts. To address this issue, Silva and Gibson [79] developed a two-dimensional model of vertebral trabecular bone to simulate the effects of age-related changes in trabecular microstructure. They found that reductions in the number of trabeculae decreased bone strength two to five times more than reductions in trabecular thickness, which resulted in an identical decrease in bone density (Fig. 6). For instance, removing longitudinally oriented trabecular elements to create a 10% reduction in density resulted in a 70% reduction in bone strength. In contrast, reducing trabecular thickness to achieve a 10% reduction in density resulted in only a 20% reduction in strength. This study implies that it is important to maintain trabecular number in order to preserve bone strength with aging. Consequently, therapies designed to counter age-related declines in bone strength should strive to maintain or restore the number of trabeculae rather
515
CHAPTER 19 Biomechanics of Age-Related Fractures
FIGURE 6
A plot of the predicted effect of bone density reductions, either by a reduction in trabecular thickness or loss of trabecular elements, on the strength of vertebral cancellous bone. Strength reductions were at least twice as sensitive to changes in the number of trabeculae as to changes in the thickness of trabeculae. Findings were similar for loading in the transverse direction. From Silva and Gibson [79].
than just increasing the thickness of existing trabecular struts. A final aspect of trabecular architecture that may have been underappreciated until recently is the potential detrimental effect of increased variability in trabecular thickness and number within a given cancellous bone specimen [80,81]. Other factors may influence age-related changes in the mechanical behavior of bone, including the histologic structure (primary vs osteonal bone); the collagen content and orientation of collagen fibers; the number and composition of cement lines; and the presence of fatigue microdamage and microfractures [55,82 – 89]. For example, an increase in osteonal remodeling (and the subsequent increase in the number of cement lines) reduces the strength of the bone for single load applications. However, the cement lines act as deterrents to crack proliferation, possibly improving the mechanical behavior of bone under repetitive loading conditions [24]. Burr and colleagues [89] reviewed the potential role of skeletal microdamage in age-related fractures. They suggest that microdamage due to repetitive loading of bone likely initiates at the level of the collagen fiber or below and may include collagen fiber – matrix debonding, disruption of the mineral – collagen aggregate, and failure of the collagen fiber itself. They hypothesize that the accumulation and coalescence of these small defects eventually lead to microcracks that are visible under light microscopy. Although the relationship between existing microcracks and bone mechanical properties has not been established in vivo, investigators have shown that damage accumulation in devitalized bone leads to a decrease in bone strength [85,90]. Thus, the accumulation of microdamage in vivo may contribute to the increased fragility of the aging skeleton.
Microcracks occur naturally in human specimens from several anatomic locations, including trabecular bone from the femoral head and vertebral body, as well as cortical bone from the femoral and tibial diaphyses [88,91 – 95]. It appears that the incidence of microcracks increases with age, probably in an exponential fashion, and that after age 40, microdamage accumulates faster in women than in men [92,94] (Fig. 7). For instance, Mori and colleagues [91] reported that the density of microcracks in the femoral head of older women is more than double the density seen in younger women. In addition, they observed an inverse, nonlinear relationship between microcrack density and trabecular bone area, indicating that microcracks accumulate more rapidly as bone mass decreases. Similar evidence for a nonlinear relationship between microcrack density and trabecular bone area has been reported for vertebral trabecular bone specimens [93]. Thus, the accumulation of microdamage in vivo may contribute to the increased fragility of the aging skeleton.
D. Age-Related Changes in Bone Geometry Age-related changes in the material properties of bone tissue are frequently accompanied by a redistribution of the cortical and trabecular bone material. It is likely that the structural rearrangement of bone tissue is driven by “preprogrammed” behavior of the endosteal and periosteal bone
FIGURE 7
Bone microcrack density vs age. There is an exponential increase in microdamage accumulation in the femoral cortex in both men and women after the age of 40 years. Damage accumulation occurs about twice as rapidly in women as in men. From Burr et al. [89].
516
MARY L. BOUXSEIN
FIGURE 8 Age-related changes in the femoral midshaft demonstrating periosteal expansion and endosteal resorption. Data represent the mean 2 standard errors. From Ruff and Hayes [99], with permission. cells, as well as the local mechanical loading environment and biochemical signals [24,96]. Hence, the adaptation pattern depends on age, gender, skeletal site, physical activity patterns, and expression (local and systemic) of cytokines and growth factors. The general pattern of adaptation in the appendicular skeleton includes endosteal resorption and periosteal apposition of bone tissue (Fig. 8). Thus, the diameter of the bone increases, but the thickness of the cortex decreases. This redistribution of bone tissue away from the center of the bone allows the bone to better resist bending and torsional loads. Resistance to bending and torsional loading is particularly important, as the highest stresses in the appendicular skeleton are due to these loading modes [24]. The most efficient design for resisting bending and torsional loads involves distributing the material far from the neutral axis of bending
or torsion (generally the center of the bone). The distribution of mass about the center of a structural element is quantitatively described by the area moment of inertia. For example, consider three circular bars, each composed of the same material (Fig. 9). The resistance of each bar to tensile and compressive loads is directly proportional to the cross-sectional area. However, the resistance to bending and torsional loads is influenced not only by how much bone (i.e., the crosssectional area), but also by how it is distributed. Therefore, the structural capacity of bar C in bending or torsion is twice that of bar A due to its greater moment of inertia. Some studies indicate that both men and women exhibit endosteal resorption accompanied by periosteal expansion [97 – 101], whereas others report that women undergo geometric changes that lead to decreased bone strength [102 – 106]. Smith and Walker [101] studied femoral
FIGURE 9 Illustration of the influence of cross-sectional geometry on the structural strength of circular structures.
517
CHAPTER 19 Biomechanics of Age-Related Fractures
radiographs of 2030 women aged 45 to 90 and reported that periosteal diameter and cortical cross-sectional area (assuming a circular cross-section) both increased approximately 11% in 35 years. Furthermore, the section modulus (an indicator of the resistance to bending loads) increased 32% in the same time period. In contrast, Ruff and Hayes [106], using direct assessment of cadaveric femurs and tibiae from 75 Caucasian adults, reported that although both men and women undergo endosteal resorption and medullary expansion with age, only men show subperiosteal expansion and bone apposition at the femoral diaphysis. They reported that, in men, cortical area is nearly constant and moments of inertia increase slightly with age. In women, however, both cortical area and moments of inertia decrease with age. The authors concluded therefore, that in this sample from modern humans, only men exhibit bone remodeling patterns that would compensate for the agerelated decline in bone material properties. Age-related changes in bone shape and size have been observed not only at diaphyseal sites, but also at the vertebral body and femoral neck. However, age-related changes at these latter sites are much smaller than those seen at diaphyseal sites. In a study of 338 skeletons from the Smithsonian Institution, Ericksen [107] found that the transverse breadths of the L3 and L4 vertebrae increase slightly with increasing age in both men and women. Using indirect measurements, Mosekilde and Mosekilde [34] also found that, in men, the cross-sectional area of lumbar vertebrae is only very weakly associated with age (r 2 0.11), increasing approximately 25 – 30% from age 20 to 90. In women, however, there was no age-related change in vertebral cross-sectional area. Based on methodology originally presented by Martin and Burr [108], techniques have been developed to assess femoral geometry from X-ray absorptiometry exams, thereby allowing in vivo assessment of bone structure [102,109 – 112]. Investigators have used these methods to investigate race [111]- and sex [102,110,112]-based differences in femoral geometry. For instance, Beck and colleagues [102] studied the cross-sectional relationship between femoral neck geometry and age in 1044 women aged 18 – 89. In women under age 50, femoral neck bone mineral density (BMD) decreased on average 4% per decade with no observable changes in femoral neck geometry. In women over age 50, femoral neck BMD declined 2.5 times faster than in the younger group (on average 7% per decade). However, in contrast to the younger group, the older women also showed changes in femoral geometry wherein the cross-sectional area and cross-sectional moment of inertia of the femoral neck declined approximately 7 and 5% per decade, respectively. These data suggest that after menopause women lose bone mass at an accelerated rate, but that unlike men [110], they do not exhibit changes in femoral geometry that would
compensate for this loss. As a result, in women, the stress in the femoral neck during walking is predicted to increase from 25 to 40% from age 50 to 80. In another study, however, there were no age-related changes in the moment of inertia of the femoral neck in either men or women [112]. As can be seen from the results of the previous studies, the sex-specific nature of age-related changes in skeletal structure remains controversial. The discrepancies in findings related to sex-specific bone adaptation patterns may be attributed to several factors. Most importantly, most of these studies use a cross-sectional design, thereby possibly introducing secular changes that confound data and eliminating the possibility of a causal relationship with age. In addition, differences in methodology (direct vs in vivo measurements), subject populations (archaeological vs modern human specimens), and measurement site (femoral shaft vs femoral neck) likely contribute to the conflicting findings. Thus, it appears that women may have a reduced capacity to alter their bone geometry in order to preserve bone strength with aging compared to men. However, the extent to which this reduced capacity contributes to the increased fracture risk observed in women is unknown.
III. BIOMECHANICS OF HIP FRACTURES Recall that the “biomechanics” view of fractures states that a fracture occurs when the loads applied to the bone exceed its load-bearing capacity. Therefore, to study the etiology of hip fractures it is important first to identify what event(s) is associated with hip fractures, and determine the loads that are applied to the bone during that event, and what the load-bearing capacity of the femur is during that loading situation. It is estimated that over 90% of hip fractures in the elderly are associated with a fall [8,113]. Thus, studies of the etiology of hip fractures are complicated by the need to examine risk factors for falls, as well as risk factors for fracture. In addition, given that fewer than 2% of falls in the elderly result in a hip fracture [114 – 116], investigations of hip fracture etiology must also distinguish factors related to “high-risk” falls that result in fracture. Therefore, this section reviews clinical and laboratory studies related to factors influencing the loads applied to the femur during a fall and the load-bearing capacity of the femur in a fall configuration.
A. Factors That Influence Fall Severity: Loads Applied to the Femur during a Fall A fall can be defined as a sudden, unexpected event that results in a person coming to rest on a horizontal surface [4,117]. A fall can be further characterized by several
518 phases: (1) an instability phase resulting in fall initiation, (2) a descent phase, (3) an impact phase, and (4) a postimpact phase during which the faller comes to rest [4]. The definition of “fall severity” is more difficult. From a biomechanical perspective, fall severity can be described by the magnitude and direction of the load applied to the hip and the impact site. From a clinical perspective, Cummings and Nevitt [118] suggest that a high-risk fall includes (1) impact on or near the hip, (2) lack of active protective mechanisms such as an outstretched arm to break the fall, and (3) insufficient energy absorption by local soft tissues. Thus, by these criteria a high-risk fall could transmit a force to the proximal femur that exceeds the force required to fracture the hip. A few surveillance studies have been conducted to more fully characterize falls as they relate to hip fracture [4 – 7,9,115]. Among nursing home residents, falling to the side and impacting the hip or side of the leg increased the risk of hip fracture approximately 20-fold relative to falling in any other direction [4]. An increase in the potential energy content of the fall, computed from fall height and body mass, was also associated with an increased risk for fracture. Similar results were reported in a nested case-control analysis of the Study of Osteoporotic Fractures cohort, a large, prospective study in community-dwelling women [7]. Women who suffered a hip fracture were more likely to have fallen sideways or straight down and to have landed on or near the hip than women who fell and did not suffer a fracture [7]. Thus, these surveillance studies have identified several factors that are related to the “severity” of a fall in terms of hip fracture risk. From these data it is clear that a fall to the side represents a particularly risky event. Several laboratory investigations have been conducted to further study the characteristics of sideways falls. In a study of the descent phase of sideways falls, van den Kroonenberg and colleagues [119] estimated the impact velocities and energies that may occur during falls from standing height, the effect of muscle activity on these impact velocities, and insights into the high-risk nature of sideways falls. Six young, healthy adults (age 19 – 30) were asked to fall sideways, as naturally as possible, onto a thick gymnastics mattress. To investigate the effect of muscle activity on fall dynamics, subjects were instructed to fall either as relaxed as they could or to fall naturally, using the musculature of the trunk and upper extremity as they would in a reflex-mediated fall. To investigate potential protective mechanisms, during some falls subjects were instructed to try to break the fall with their arm. The vertical velocity at impact with the floor ranged between 2.1 and 4.8 m/sec. The impact velocity was 7% lower in relaxed than in muscle-active falls, a finding attributed to the observation that hip impact occurs closer to the feet in the muscle-relaxed case. Despite instructions to break the fall with an outstretched arm, only two of six subjects were able to do so (Fig. 10). In the
MARY L. BOUXSEIN
FIGURE 10 Example of a sideways fall onto a thick gymnastics mattress. Despite instructions to break the fall with the hand, only two of six subjects were able to do so. In the other subjects, hip impact occurred first, thus providing insight into the high-risk nature of sideways falls From van den Kroonenberg et al. [119], with permission.
remaining subjects, hip impact occurred first, followed by impact of the arm or hand. Finally, the authors found that, in these young adults, approximately 70% of the total energy available is dissipated during the descent phase of a sideways fall from standing height. This energy dissipation is likely due to muscle activity and the stiffness and damping characteristics of the hip and knee joints. Sabick and colleagues [120] reported that “active responses,” such as using the arm to break a fall, reduce the impact forces experienced at the hip during falls to the side. However, despite the potential for reducing fall severity via active responses, it is likely that with age, the ability to dissipate energy during a fall or to activate protective responses will decrease, and therefore it is quite likely that elderly individuals “fall harder” than young adults. The forces applied to the proximal femur during a sideways fall depend not only on the dynamics of the descent phase of the fall, but also on characteristics of the impact phase of the fall. Robinovitch and colleagues have conducted a series of experiments to study the potential roles of trochanteric soft tissues, muscle contraction, and body configuration in determining the load applied to the femur during a sideways fall with impact to the greater trochanter [121 – 125]. In these experiments, they used a “pelvis release” system (Fig. 11), in which a small force is applied to the lateral aspect of the hip and the dynamic response of the body is measured [121]. This system allows impact forces from falls to be predicted with reasonable accuracy from the body’s response to safe, simulated collisions [122]. They found that during a sideways fall with impact to the greater trochanter, only about 15% of the total impact force is distributed to structures peripheral to the hip, whereas the remainder of the force is delivered along a load path directly in line with the hip [123]. In addition, for the same body mass and height, sideways falls with the trunk in a more upright position are predicted to result in greater
CHAPTER 19 Biomechanics of Age-Related Fractures
519
FIGURE 11 Schematic diagram of the setup used for “pelvis release” experiments. The subject’s pelvis was supported by the sling, raised a small amount, and then released onto the force platform, which recorded the body’s dynamic response. Experiments were conducted in two body configurations — in the trunk-straight and trunk-flexed positions — to determine the effect of trunk position on fall impact dynamics. From Robinovitch et al. [123], with permission.
impact forces on the proximal femur than falls where the trunk is more horizontal at impact. To study the force attenuation and energy absorption properties of the soft tissues overlying the greater trochanter, tissue samples were obtained from nine cadavers, positioned over a surrogate proximal femur and pelvis, and subjected to a typical impact load associated with a sideways fall [125]. For a constant impact energy, trochanteric soft tissue thickness was strongly negatively correlated with the peak femoral impact force (r 2 0.91), such that the force applied to the femur decreased approximately 70 N per 1-mm increase in tissue thickness (Fig. 12). However, the force attenuation due to trochanteric soft tissues alone is likely insufficient to prevent hip fracture in falls where an elderly person lands directly on the hip [125]. These findings suggest that trochanteric padding systems may be effective means of reducing the load applied to the femur during a fall [126]. Finally, van den Kroonenberg et al. [127] developed a series of biomechanical models to estimate peak impact forces delivered to the proximal femur during a sideways fall from standing height. The models incorporated stiffness and damping parameters from the “pelvis-release” experiments [121 – 123], and the models’ behavior was compared with previous observations of the dynamics of voluntary sideways falls [119]. Using the most accurate model, peak impact forces applied to the greater trochanter ranged from 2900 to 4260 N ( 650 – 960 lbs) for 5th to 95th percentile woman based on weight and height. Thus, these findings support the idea that “the bigger they are, the harder they fall” [128]. Given an individual’s height and weight, these models can be used to estimate femoral impact forces associated with a sideways fall.
FIGURE 12 Effect of trochanteric soft tissue thickness on (top) the force delivered to the femur and (bottom) the energy absorbed by soft tissue for a constant energy impact directed laterally on the hip. From Robinovitch et al. [125].
B. Factors That Influence the Strength of the Proximal Femur As mentioned previously, several factors contribute to the load-bearing capacity of the proximal femur, including its intrinsic material properties, as well as the total amount (size) and spatial distribution (shape) of the bone tissue. Because the mechanical properties of both cortical and trabecular bone are strongly related to bone density, many have hypothesized that age-related bone loss is a primary contributor to the steep increase in hip fracture incidence with age. In support of this hypothesis, there is strong evidence from prospective clinical studies that low BMD, measured both at the hip and at other sites, is a risk factor for hip fracture [129 – 132]. Furthermore, case-control studies of elderly fallers have reported that low BMD of the hip is a risk factor for hip fracture that is independent of fall characteristics [5 – 7]. Several laboratory studies have evaluated the loadbearing capacity of the proximal femur using a configuration designed to simulate the single-leg stance phase of gait [109,133 – 139]. The loads required to fracture the femur in the stance phase of gait ranged from approximately 1000 to 13,000 N (225 – 3000 lbs). These studies demonstrated a
520 strong relationship between the load required to fracture the femur in this stance configuration and noninvasive measurements of bone geometry and bone mineral density or content. Other studies have evaluated the load-bearing capacity of the proximal femur in a configuration designed to simulate a sideways fall with impact to the greater trochanter [139 – 147]. Courtney and colleagues [141,142] studied the effect of age and loading rate on the failure load of the proximal femur in the fall configuration (Fig. 13). They found that at a slow loading rate (2 mm/sec), femurs from young individuals (age 17 – 51) were more than twice as strong as femurs from older individuals (age 59 – 83). At high loading rates (100 mm/sec), such as might be expected during a fall, femurs from both young and older individuals were approximately 20% stronger than at the slower loading rate [142]. However, femurs from the younger group were still approximately 80% stronger than those from the older group. Loading direction may also dramatically influence femoral failure loads. Greater loads are required to fracture femurs testing in a single-leg stance configuration than in a sideways fall configuration, further supporting the high risk of a sideways fall in terms of hip fracture risk [139]. Moreover, subtle differences in the direction of a sideways fall can influence femoral strength as much as 25 years of age-related bone loss [148,149]. In addition to age, loading rate, and loading direction, femoral geometry also influence the load-bearing capacity of the proximal femur. The relationship between femoral
FIGURE 13 Mean failure loads for cadaveric proximal femurs from young and elderly donors tested in a sideways fall configuration at slow and fast loading rates. For each loading rate, femurs from the younger individuals were 80 – 100% stronger than femurs from the older individuals. Data from Courtney et al. [141,142].
MARY L. BOUXSEIN
geometry and load-bearing capacity is not unexpected. Because the load-bearing capacity is a structural property, it is influenced by the size of the specimen. Therefore, larger femurs have a greater load-bearing capacity. Therefore, as expected, femoral neck area, neck width, and neck axis length are all positively correlated (r 2 0.21 – 0.79) with femoral failure loads [141,143,145]. It is interesting to note that the positive correlation between femoral neck length and femoral strength appears to contradict findings from clinical studies, where a longer hip axis length (HAL) is associated with a greater risk of hip fracture [150]. This discrepancy may be attributed to the differences in the portion of hip anatomy that is included in the in vitro measurements (neck axis length only) versus in vivo measurements (neck axis length plus acetabular thickness). Some evidence suggests that it is the “acetabular thickness” portion of the measurement that is associated with fracture risk, and not the “femoral neck length” portion [151]. Additional laboratory studies are required to understand the complex relationship between hip geometry and fracture risk. While it is important to understand what factors influence the load-bearing capacity of the femur in the laboratory environment, it is also critical to develop techniques that can be used clinically to predict femoral strength. Several studies have confirmed that noninvasive assessments of bone mineral density and geometry using dual-energy Xray absorptiometry (DXA) or quantitative computed tomography (QCT) are strongly correlated to the load-bearing capacity of human cadaveric femurs. Femoral bone mineral content and density explain over 80% of the variation in the load-bearing capacity of the proximal femur [138,141,143,145 – 147] (Fig. 14). In summary, the load-bearing capacity of cadaveric proximal femurs ranges from approximately 800 to 10,000 N (180 – 2250 lbs) and is influenced, at least in part, by femoral bone mineral density, femoral geometry, loading direction, and loading rate. At a given moment, an individual’s
FIGURE 14
Bone mineral density of the femoral neck versus femoral failure load of cadaveric proximal femurs. The femurs were tested to failure in a configuration designed to simulate a sideways fall with impact to the greater trochanter. From Bouxsein et al. [147], with permission.
521
CHAPTER 19 Biomechanics of Age-Related Fractures
bone density and geometry are constant, although they can readily change with age or therapeutic intervention. However, other factors, such as loading direction and loading rate that are influenced by the characteristics of the fall, may significantly influence fracture risk.
C. Interactions between Fall Severity and Femoral Strength: The Factor of Risk for Hip Fracture The concept of a factor of risk for fracture (introduced in Section I) suggests that low BMD is not the only indicator of risk, but rather that the loads applied to the bone must also be considered. Case-control studies have demonstrated the importance of both fall severity and bone mineral density as risk factors for hip fracture [5 – 7]. In a nested casecontrol analysis of the Study of Osteoporotic Fractures cohort, Nevitt and Cummings [7], studied 130 women who fell and suffered a hip fracture and a consecutive sample of 467 women who fell and did not fracture. They reported that among those who fell on or near their hip, those who fell sideways or straight down were at increased risk for hip fracture (odds ratio 4.3), whereas those who fell backward were less likely to suffer a hip fracture (odds ratio 0.2). Furthermore, low BMD at the femoral neck (odds ratio 2.6 for a 1 SD decrease) or calcaneus (odds ratio 2.4 for a 1 SD decrease) strongly increased the risk of fracture among those who fell on or near the hip. Greenspan and co-workers [5] reported similar findings in a study of 149 community-dwelling men and women, including 72 cases who fell and suffered a hip fracture and 77 control subjects who fell and did not fracture. They showed that in these elderly fallers, independent risk factors for hip fracture included characteristics related to fall severity, low bone mineral density at the hip, and body habitus (Table 2). The success of hip protectors in preventing hip fracture provides additional evidence of the strong relationship between falls and hip fracture risk [126,152,153].
Clinical studies provide valuable information about the independent contributions of fall severity and skeletal fragility to hip fracture risk. However, further insight may be achieved by considering a “factor of risk” for hip fracture. The previous two sections have described how laboratory techniques can be used to develop and validate methods for estimating the loads applied to the femur and the load-bearing capacity of the femur from data that can be acquired in a clinical setting. Thus, these findings can be used to estimate the factor of risk for hip fracture due to a sideways fall from standing height. Myers and co-workers [154] applied the factor of risk concept in a case-control study of elderly fallers. The numerator of the factor of risk, the applied load, was estimated from previous studies of the descent and impact phases of a sideways fall with impact to the lateral aspect of the hip [119,121 – 123,127]. Each individual’s body height and weight were used as input parameters for the model to estimate the impact force delivered to the proximal femur during a sideways fall from standing height. The denominator of the factor of risk, or load-bearing capacity of the proximal femur, was determined from linear regressions between noninvasive bone densitometry and femoral failure loads in a fall configuration [143]. For each subject, femoral bone mineral density was assessed by DXA and then used to estimate the femoral failure load. There was a strong association between the factor of risk and hip fracture in these elderly fallers, with the odds of hip fracture increasing by 5.1 for a 1 SD increase in the factor of risk (95% confidence interval: 2.9, 9.2) (Fig. 15). In comparison, the odds ratio for a 1 SD decrease in femoral BMD was 2.0 (95% confidence interval: 1.4, 2.6).
IV. BIOMECHANICS OF VERTEBRAL FRACTURES Investigations of the etiology and biomechanics of vertebral fractures are particularly difficult, as the precise definition of a vertebral fracture remains controversial [155,156].
TABLE 2 Multiple Logistic Regression Analysis of Factors Associated with Hip Fracture in Community-Dwelling Men and Women Who Fella Factor
Adjusted odds
95% Confidence interval
P
Fall to the side
5.7
2.3 – 14
0.001
Femoral neck BMD (g/cm2)b
2.7
1.6 – 4.6
0.001
Potential energy of fall (Joules)c
2.8
1.5 – 5.2
0.001
Body mass index (kg/m2)b
2.2
1.2 – 3.8
0.003
a
Data from Greenspan et al. [5]. Calculated for a decrease of 1 SD. c Calculated for an increase of 1 SD. b
522
MARY L. BOUXSEIN
FIGURE 15
Proportion of subjects with hip fracture in each quartile of the factor of risk for hip fracture (top) and femoral neck BMD (bottom). Data from Myers et al. [154].
In addition, many fractures that are identified by radiographic review are asymptomatic [157], further complicating the interpretation of many studies. In contrast to the growing recognition of the importance of bone fragility and fall severity in the etiology of hip fractures, the role of spinal loading in the etiology of age-related vertebral fractures has received relatively little attention [158,159]. As loads are applied to the spine during nearly every activity of daily living, it is crucial to distinguish which of these activities (and the resulting loads on the spine) are associated with vertebral fractures to try to understand the loading environment that leads to vertebral fractures.
A. Factors That Influence Loads Applied to the Spine Although no clinical study has yet examined the relative roles of bone fragility and load severity as risk factors for
TABLE 3
vertebral fracture, several investigators have reviewed medical records or interviewed patients to assess the “degree of trauma” associated with vertebral fractures [157,160 – 162]. A few observational studies have attempted to identify the activities surrounding the onset of a vertebral fracture. These investigators have reviewed medical records or interviewed patients to assess the “degree of trauma” associated with vertebral fractures [157,160 – 162]. Cooper et al. [157] reviewed medical records from a 5-year period to determine the circumstances associated with “clinically diagnosed” vertebral fractures in a populationbased sample of 341 Rochester, Minnesota, residents. In their study, a specific loading event was reported for approximately 50% of the total fractures (Table 3). In contrast to the commonly held belief that lifting plays a major role in the development of vertebral fractures, relatively few of the fractures were associated with lifting. Excluding fractures that were diagnosed incidentally, only 10% of fractures were associated with “lifting a heavy object,” whereas nearly 40% were associated with falling. In a hospitalbased study, Myers et al. [163] interviewed patients after diagnosis of vertebral fracture with respect to their activity at the time of fracture. Their results indicate that nearly 50% of acute, symptomatic vertebral fractures in individuals over age 60 are associated with a fall, whereas 20% are associated with “controlled” activities, such as bending, lifting, and reaching. Most of the remainder of the patients could not identify a specific activity at the time of fracture. Therefore determining the forces on the spine during controlled activities and falls may improve our understanding of the biomechanics of vertebral fractures. Although it is impossible to measure the loads on the vertebral bodies in vivo, investigators have used kinematic analysis, electromyographic measurements, and biomechanical modeling to estimate the loads on the lumbar spine during various activities [164 – 167]. The models use optimization techniques to estimate the trunk muscle forces and compressive forces on the spine during various tasks and
Circumstances Associated with Clinically Diagnosed Vertebral Fractures a
Reported activity/circumstance
Number of persons
% of symptomatic fractures
% of total fractures 3.5
Pathologic fracture
12
4
Traffic accident
20
7
6
Fall from greater than standing height
27
9
8
Fall from standing height or less
86
30
25
Lifting a heavy object
29
10
8.5
“Spontaneous”
113
39
33
Diagnosed incidentally (asymptomatic)
54
NA
16
a
Data from Cooper et al. [157].
523
CHAPTER 19 Biomechanics of Age-Related Fractures
have been verified by comparing predicted compressive spine loads and muscle activity with direct measurements of intradiscal pressure [167 – 170] and myoelectric trunk muscle activity [164,165,167,171 – 173]. These models were originally developed to study the potential origins and mechanisms of low back pain and injury in working adults. Therefore, they are generally based on anthropometric data from young, healthy adults and are limited to estimating the vertebral forces in the lumbar region. However, Wilson [174] extended these models to include the mid- and lower thoracic spine and incorporated geometric properties of the trunk using CT scans of older individuals. Using this model, the compressive forces applied to T8, T11, and L2 vertebrae during various activities for a woman who weighed 58 kg and was 1.6 m tall (mean values from a cohort of 120 women aged 65 yrs or older [175] were computed (Table 4). The estimated forces applied to the spine ranged from approximately 400 to 2100 N for typical activities. For example, rising from a chair without the use of one’s hands results in compressive forces equal to 60 and 173% of body weight on T11 and L2 vertebrae, respectively. These estimates reinforce the concept that subtle changes in body position can dramatically alter spinal loading. Standing straight and holding an 8-kg weight with the arms slightly extended creates a compressive load on L2 equal to 230% of body weight, whereas flexing the trunk forward 30° and holding the same weight generates a compressive force on L2 of over 320% of body weight. From these estimates, it is clear that everyday activities, such as rising from a chair or bending over and picking up a full grocery bag, can generate high forces on the spine. There are currently few biomechanicals models designed to estimate the load on the spine during falls. However, TABLE 4
based on data from a study of the dynamics of backwards falls [176] and from the previously described “pelvis-release” experiments [121], Myers and Wilson [159] estimated that the impact force on the pelvis due to a backward fall would be approximately 2000 – 2500 N. This impact force would be expected to be dissipated somewhat before reaching the thoracolumbar spine. Research is currently underway to further characterize the loads applied to the spine during falls in order to better our understanding of the circumstances surrounding acute vertebral fractures.
B. Factors That Influence Vertebral Strength The use of noninvasive assessments of skeletal status to predict vertebral strength in vivo is based on the assumption that much of the variability in the strength of whole vertebrae can be explained by variations in bone mineral density and/or geometry. As in other skeletal structures, the loadbearing capacity of a whole vertebra is determined by its intrinsic material properties, as well as its overall geometry and shape. The vertebral body is characterized by a central core of cancellous bone surrounded by a thin covering of condensed trabecular bone (often referred to as a “cortical shell“) [177,178]. In the spine, compressive loads are transferred from the intervertebral discs to adjacent vertebral bodies. Therefore, age-related changes in the properties of the intervertebral disc, the vertebral centrum, and the vertebral shell can each influence the load-bearing capacity of the vertebrae. For instance, the thickness of the outer shell decreases from approximately 400 – 500 m at age 20 – 40 to 200 – 300 m at age 70 – 80 and to 120 – 150 m in osteoporotic individuals [179]. This change in vertebral morphology likely influences the way that loads are transmitted
Predicted Compressive Loads on L2 and T11 Vertebrae during Various Activitiesa Predicted load on T11
Activity Relaxed standing
N 240
% of body weight 41
Predicted load on L2 N
% of body weight
290
51 173
Rising from a chair, without hands
340
60
980
Standing, holding 8 kg weight close to body
320
57
420
74
Standing, holding 8-kg weight arms extended
660
117
1302
230
Standing, trunk flexed 30°, extended
370
65
830
146
Standing, trunk flexed 30°, 18 kg with arms extended
760
135
1830
323
Lift 15 kg from floor, knees arms straight down
593
104
1810
319
a The loads were computed from the model developed by Wilson [174] for a woman who weighs 58 kg and is 162 cm tall, and are expressed as the absolute value (in Newtons, N) and as a percent of total body weight.
524 throughout the spine. For instance, whereas the relative contributions of the vertebral centrum and shell to overall vertebral strength remain controversial, it is suggested that the vertebral shell may account for 10 – 30% of vertebral strength in healthy individuals and, due to decreased bone mass in the trabecular centrum, from 50 to 90% in osteoporotic persons [178 – 183]. Understanding the relative contributions of the cortical shell and trabecular centrum throughout aging and disease may afford the development of therapeutic agents specifically designed to strengthen one of these bone compartments. A number of laboratory studies have investigated the relationships among the strength of human lumbar and thoracic vertebrae and age, bone density, and vertebral geometry [21,33,182,184 – 196]. These studies indicate that the strength of thoracolumbar vertebrae is reduced from a value of 8000 – 10,000 N at age 20 – 30 to 1000 – 2000 N by age 70 – 80 [179,197]. In severely osteoporotic individuals, the load-bearing capacity may be even less [191]. The strength of human vertebrae is strongly correlated with noninvasive estimates of vertebral bone density and geometry, with approximately 50 – 80% of the variance in load-bearing capacity explained by parameters measured noninvasively [21,182,185 – 194,196]. For example, strong correlations have been reported between bone density and vertebral cross-sectional area assessed by quantitative computed tomography (QCT) and vertebral failure loads [187,188]. In addition, several investigators have reported strong correlations between bone mineral density, assessed by dual-energy X-ray absorptiometry, and vertebral strength [191,192,194 – 196]. For example, Moro et al. [191] found that lumbar BMD assessed by DXA correlates strongly with the compressive failure load and energy to failure of both L2 and T11 vertebrae (Fig. 16). The standard error of the estimate for predicting vertebral failure load from lumbar BMD was 527 N (25% of the mean failure load) for T11 and 733 N (28% of the mean failure load) for L2.
FIGURE 16 Linear relationship between lumbar BMD and compressive failure loads of T11 and L2 vertebrae as reported by Moro et al. [191]. Correlation coefficients between lumbar BMD and failure loads of T11 and L2 were r 0.94 (p0.001) and r 0.89 (p0.001), respectively.
MARY L. BOUXSEIN
Thus, it appears that noninvasive assessments of bone mass and bone mineral density provide reasonable estimates of the failure loads of cadaveric vertebrae subjected to controlled compression tests in the laboratory. It remains to be seen whether BMD or other bone density parameters can predict the strength of vertebrae subjected to loading conditions that may more closely resemble the mechanical environment in vivo, such as falling or compression combined with forward flexion or compression combined with lateral bending.
C. Interactions between Spinal Loads and Vertebral Strength: The Factor of Risk for Vertebral Fracture Although it has not been clearly demonstrated by clinical surveillance studies, it seems reasonable to suggest that, similar to hip fractures, both bone fragility and skeletal loading are important factors in the etiology of vertebral fractures [198]. To investigate the potential roles of bone fragility and spinal loading, Myers and Wilson [159] estimated a factor of risk for vertebral fractures for various activities of daily living (Fig. 17). As before, the factor of risk was defined as the ratio between load applied to the bone and its load-bearing capacity for a given loading event. They estimated the numerator of the factor of risk (i.e., the applied load) using predictions of compressive loading in the spine from the model developed by Wilson [174]. The denominator of the factor of risk (i.e., the failure load) was estimated from linear regressions between lumbar BMD and the compressive failure load of cadaveric vertebrae [191]. Their predictions of the factor of risk indicate that osteopenic individuals may perform many activities wherein their factor of risk for vertebral fracture is close to or greater than one, suggesting that they are at high risk for fracture. These estimates show that a woman who bends over to pick up a 15-kg object is predicted to be at great risk for vertebral fracture (i.e., 1) when her lateral L2 BMD is less than 0.55 g/cm2. To put this in context, the mean lateral L2 BMD for a 65-year-old woman is 0.58 0.10 g/cm2 (199). Hence, for the same lifting activity (ie., picking up a 15-kg object), a 65-year-old woman whose spine BMD is 1 SD below the mean for her age would have a factor of risk equal to 1.4 and would be at high risk for fracture. To reduce her factor of risk below one without altering the applied load due to lifting, the osteopenic woman would have to increase her spine BMD by 20%, an increase much greater than is currently achieved through the use of pharmocologic agents [200]. Thus, individuals with extremely low bone mineral density may be at risk for vertebral fracture during simple activities such as tying one’s shoes or opening a window. Individuals with low bone mineral density (still in the osteopenic range) may be at risk for vertebral fracture when lifting groceries out of the car or
525
CHAPTER 19 Biomechanics of Age-Related Fractures
FIGURE 17 Factor of risk for vertebral fracture for eight common activities as a function of lumbar bone mineral density. The numerator of the factor of risk was determined from models of spine loading at L2 for an elderly woman of average height and weight. The demonimator was determined on the basis of regression analysis between lateral lumbar BMD and the load-bearing capacity of the L2 vertebrae. Values for lateral BMD cover a wide range. The t score (number of standard deviations from the mean value for BMD in young women) is approximately 1 for a BMD 0.9 g/cm2 and is 5 for BMD 0.4 g/cm2. The factor of risk is predicted to be greater than or close to 1 for low BMD values (shaded area). From Myers and Wilson [159], with permission.
picking up a toddler. These examples illustrate the need for strategies to prevent vertebral fractures, such as reducing spinal loading by avoiding certain “high risk” activities.
V. SUMMARY AND CLINICAL IMPLICATIONS This chapter emphasized the concept that age-related fractures represent a structural failure whereby the forces applied to the bone exceed its load-bearing capacity. Viewing fractures in this manner, it is clear that studies of their etiology must include factors that influence skeletal fragility or its load-bearing capacity, as well as those that influence the forces that are applied to the skeleton. The load-bearing capacity of a skeletal structure is determined by its intrinsic material properties as well as the total amount (size) and spatial distribution (shape) of the bone tissue. Considerable evidence indicates that the material properties, in particular the elastic modulus and ultimate strength, of both cortical and trabecular bone decrease with increasing age in both men and women. This decrease in material properties is likely due, in part, to age-related
reductions in bone mass, as the elastic modulus and strength of trabecular bone are related to density by a nonlinear relationship. Therefore, small changes in bone density can dramatically influence bone material properties. These decrements in bone density and material properties may be partially offset by geometric rearrangement of the bone tissue, particularly in the long bones, that helps preserve the bone’s ability to resist bending and torsional loads. Clinical investigations have confirmed that skeletal status and fall severity are both significant and independent risk factors for hip fracture [5 – 9]. Estimates of the forces applied to the proximal femur during a sideways fall range from 2900 to 4260 N for the 5th to 95th percentile woman based on height and weight. Factors that influence the load applied to the femur include, but are not limited to, fall height, fall direction, body habitus, muscle activity, trochanteric soft tissue thickness, and the intrinsic stiffness of the hip and knee joints. In comparison, estimates of the load required to fracture the elderly cadaveric femur in a configuration simulating a sideways fall range from 800 to 10,000 N. This femoral failure load is influenced by femoral bone mineral content and density; femoral geometry; and the direction and rate of
526 the applied load. In particular, it appears that subtle changes in the direction of the load applied to the femur during a fall can influence femoral failure loads as much as nearly 25 year’s worth of age-related bone loss. Many of these factors that influence fall severity and femoral strength are independent of femoral bone mineral density and thus may prove useful in improving current estimates of fracture risk that are based on bone densitometry alone. In contrast to hip fractures, relatively little is known about the combined roles of spinal loading and skeletal fragility in the etiology of vertebral fractures. Contrary to previously held beliefs that vertebral fractures are caused primarily by bending and lifting activities, evidence shows that falls may play a significant role in the etiology of vertebral fractures. In one study, nearly 40% of clinically diagnosed, symptomatic fractures were associated with falls, whereas 10% were attributed to lifting a heavy object [157]. Moreover, preliminary findings from a surveillance study of acute vertebral fractures indicate that approximately one-half of these fractures are associated with falls. Thus, future studies should incorporate assessments of fall severity in order to determine the characteristics of falls associated with vertebral fracture. In addition, models to estimate the loads applied to the spine during a fall should be developed. Mathematical models used to estimate the forces generated in the spine during bending and lifting activities indicate that compressive forces generated in the lower thoracic and upper lumbar spine range from approximately 400 to 2100 N. A comparison of these loads with predicted vertebral strengths suggests that activities of daily living may place the elderly, osteopenic person at high risk for vertebral fracture. To date, investigators have focused primarily on methods to prevent bone loss and to restore bone to the osteopenic skeleton. However, alternative approaches for fracture prevention that are directed at reducing the loads applied to the skeleton may prove to be both effective and cost-efficient. For example, trochanteric padding systems designed to reduce the load applied to the hip during a fall have shown great potential for reducing fracture risk [126,152,153]. In one study, analyzed on an intention to treat basis, hip fracture incidence was reduced 53% in the group assigned to wear the hip pads [152]. In addition, energy-absorbing floors, particularly in institutional environments, may help lower the risk of fractures due to falls [201,202]. Vertebral fracture incidence may be reduced by teaching high-risk patients to avoid activities that generate high loads on the spine and thereby put them at increased risk for fracture. Clearly, identification of these high-risk activities is critical to the success of this approach for preventing fractures. Ultimately, fracture prevention may be best achieved by an educational program designed to limit high-risk activities in conjunction with interventions targeted at increasing bone mass and reducing loads applied to the skeleton during traumatic events.
MARY L. BOUXSEIN
References 1. N. F. Ray, J. Chan, M. Thamer, and L. Melton, Medical expenditures for the treatment of osteoporotic fracture in the United States in 1995: Report from the National Osteoporosis Foundation. J. Bone Miner. Res. 12(1), 24 – 35 (1997). 2. A. Praemer, S. Furner, and D. Rice, “Musculoskeletal Conditions in the United States.” American Academy of Orthopaedic Surgeons, Rosemont, IL, (1999). 3. C. Cooper, G. Campion, and L. J. Melton, Hip fractures in the elderly: A world-wide projection. Osteopor Int. 2, 285 – 289 (1992). 4. W. Hayes, E. Myers, J. N. Morris, H. S. Yett, and L. A. Lipsitz, Impact near the hip dominates fracture risk in elderly nursing home residents who fall. Calcif. Tissue Int. 52, 192 – 198 (1993). 5. S. L. Greenspan, E. R. Myers, L. A. Maitland, N. M. Resnick, and W. C. Hayes, Fall severity and bone mineral density as risk factors for hip fracture in ambulatory elderly. JAMA 217(2), 128 – 133 (1994). 6. S. L. Greenspan, E. R. Myers, D. P. Kiel, R. A. Parker, W. C. Hayes, and N. M. Resnick, Fall direction, bone mineral density, and function: Risk factors for hip fracture in frail nursing home elderly. Am. J. Med. 104(6), 539 – 545 (1998). 7. M. C. Nevitt and S. R. Cummings, Type of fall and risk of hip and wrist fractures: The study of osteoporotic fractures. J. Am. Geriatr. Soc. 41, 1226 – 1234 (1993). 8. J. A. Grisso, J. L. Kelsey, B. L. Strom, G. Y. Chiu, G. Maislin, L. A. O’Brien, S. Hoffman, and F. Kaplan, Risk factors for falls as a cause of hip fracture in women. N. Engl. J. Med. 324, 1326 – 1331 (1991). 9. R. G. Cumming, and R. J. Klineberg, Fall frequency and characteristics and the risk of hip fractures. J. Am. Geriatr. Soc. 42(7), 774 – 778 (1994). 10. W. Hayes, S. Piazza, and P. Zysset, Biomechanics of fracture risk prediction of the hip and spine by quantitative computed tomography. In: “Radiologic Clinics of North America D. I. Rosenthal (ed.), Vol. 29, pp. 1 – 18 Saunders, Philadelphia. 11. S. L. Hui, C. W. Slemenda, and C. C. Johnston, Jr., Age and bone mass as predictors of fracture in a prospective study. J. Clin. Invest. 81(6), 1804 – 1809 (1988). 12. L. J. Melton III, Epidemiology of hip fractures: Implications of the exponential increase with age. Bone 18, 121S – 126S (1996). 13. C. H. Turner and D. B. Burr, Basic biomechanical measurements of bone: A tutorial. Bone 14, 595 – 608 (1993). 14. D. Reilly, A. Burstein, and V. Frankel, The elastic modulus for bone. J. Biomech. 7, 271 – 275 (1974). 15. D. Reilly and A. Burstein, The elastic and ultimate properties of compact bone tissue. J. Biomech. 8, 393 – 405 (1975). 16. R. B. Ashman, S. C. Cowin, W. C. Van Buskirk, and J. C. Rice, A continuous wave technique for the measurement of the elastic properties of cortical bone. J. Biomech. 17, 349 – 361 (1984). 17. T. Brown and A. Ferguson, Mechanical property distributions in the cancellous bone of the human proximal femur. Acta. Orthop. Scand. 51, 429 – 437 (1980). 18. M. J. Ciarelli, S. Goldstein, J. Kuhn, D. Cody, and M. Brown, Evaluation of orthogonal mechanical properties and density of human trabecular bone from the major metaphyseal regions with materials testing and computed tomography. J. Orthop. Res. 9, 674 – 682 (1991). 19. D. Ulrich, B. van Rietbergen, A. Laib, and P. Ruegsegger, The ability of three-dimensional structural indices to reflect mechanical aspects of trabecular bone. Bone 25(1), 55 – 60 (1999). 20. J. S. Arnold, M. H. Bartley, D. S. Bartley, S. A. Tont, and D. P. Henkins, Skeletal changes in aging and disease. Clin. Orthop. Rel. Res. 49, 37 (1966). 21. L. Mosekilde, A. Viidik, and L. Mosekilde, Correlation between the compressive strength of iliac and vertebral trabecular bone in normal individuals. Bone 6, 291 (1985).
CHAPTER 19 Biomechanics of Age-Related Fractures 22. L. Mosekilde and L. Mosekilde, Normal vertebral body size and compressive strength: Relations to age and to vertebral and iliac trabecular bone compressive strength. Bone 7, 207 – 212 (1986). 23. L. Mosekilde and L. Mosekilde, Iliac crest trabecular bone volume as predictor for vertebral compressive strength, ash density and trabecular bone volume in normal individuals. Bone 9, 195 – 199 (1988). 24. R. Martin, Aging and strength of bone as a structural material. Calcif. Tissue. Int. 53 (Suppl 1), S34 – S40 (1993). 25. A. Burstein, D. Reilly, and M. Martens, Aging of bone tissue: Mechanical properties. J. Bone Jt Surg. 58-A, 82 – 86 (1976). 26. J. Currey, The mechanical consequences of variation in the mineral content of bone. J. Biomech. 2, 1 – 11 (1969). 27. O. Lindahl and A. Lindgren, Cortical bone in man. II. Variation in tensile strength with age and sex. Acta. Orthop. Scand. 38, 141 – 147 (1967). 28. R. Melick and D. Miller, Variations of tensile strength of human cortical bone with age. Clin. Sci. 30, 243 – 248 (1966). 29. C. Smith and D. Smith, Relations between age, mineral density and mechanical properties of human femoral compacta. Acta. Orthop. Scand. 47, 496 – 502 (1976). 30. J. C. Wall, S. K. Chatterji, and J. W. Jeffrey, Age-related changes in the density and tensile strength of human femoral cortical bone. Calcif. Tissue. Int. 27, 10 – 108 (1979). 31. R. McCalden, J. McGeough, M. Barker, and C. Court-Brown, Agerelated changes in the tensile properties of cortical bone. J. Bone. Jt. Surg. 75-A, 1193 – 1205 (1993). 32. P. Zioupos and J. Currey, Changes in the stiffness, strength, and toughness of human cortical bone with age. Bone 22, 57 – 66 (1998). 33. G. H. Bell, O. Dunbar, and J. S. Beck, Variations in strength of vertebrae with age and their relation to osteoporosis. Calcif. Tissue. Res. 1, 75 – 86 (1967). 34. L. Mosekilde and L. Mosekilde, Sex differences in age-related changes in vertebral body size, density and biomechanical competence in normal individuals. Bone 11, 67 – 73 (1990). 35. J. Weaver and J. Chalmers, Cancellous bone: Its strength and changes with ageing and an evaluation of some methods for measuring its mineral content. J. Bone. Jt Surg. 48-A, 289 – 299 (1966). 36. L. Mosekilde, L. Mosekilde, and C. C. Danielson, Biomechanical competance of vertebral trabecular bone in relation of ash density and age in normal individuals. Bone 8, 79 – 85 (1987). 37. M. Ding, M. Dalstra, C. Danielsen, J. Kabel, I. Hvid, and F. Linde, Age variations in the properties of human tibial trabecular bone. J. Bone. Jt Surg. 79-B, 995 – 1002 (1997). 38. J. D. Currey, Changes in the impact energy absorption of bone with age. J. Biomech. 12, 459 – 469 (1979). 39. J. M. Britton and M. W. J. Davie, Mechanical properties of bone from iliac crest and relationship to L5 vertebral bone. Bone 11, 21 – 28 (1990). 40. L. Mosekilde, Age-related changes in vertebral trabecular bone architecture: Assessed by a new method. Bone 9, 247 – 250 (1988). 41. L. Mosekilde, Sex differences in age-related loss of vertebral trabecular bone mass and structure: Biomechanical consequences. Bone 10, 425 – 432 (1989). 42. O. Lindahl, Mechanical properties of dried defatted spongy bone. Acta. Orthop. Scand. 47, 11 – 19 (1976). 43. D. R. Carter and W. C. Hayes, Bone compressive strength: The influence of density and strain rate. Science 194, 1174 – 1176 (1976). 44. D. R. Carter and W. C. Hayes, The compressive behavior of bone as a two-phase porous structure. J. Bone. Jt. Surg. 59-A(7), 954 –962 (1977). 45. L. Gibson and M. Ashby, “Cellular Solids: Structure and Properties.” Pergamon Press, New York, (1988). 46. J. C. Rice, S. C. Cowin, and J. A. Bowman, On the dependence of the elasticity and strength of cancellous bone on apparent density. J. Biomech. 21(2), 155 – 168 (1988).
527 47. T. M. Keaveny and W. C. Hayes, A 20-year perspective on the mechanical properties of trabecular bone. J. Biomech. Eng. 115, 534–542 (1993). 48. P. Atkinson, Variation in trabecular structure of vertebrae with age. Calcif. Tissue. Res. 1, 24 – 32 (1967). 49. L. Twomey, J. Taylor, and B. Furniss, Age changes in the bone density and structure of the lumbar vertebral column. J. Anat. 136(1), 15 – 25 (1983). 50. L. Gibson, The mechanical behaviour of cancellous bone. J. Biomech. 18, 317 – 328 (1985). 51. J. Compston, Connectivity of cancellous bone: Assessment and mechanical implications. Bone 15, 463 – 466 (1994). 52. J. D. Currey, The effect of porosity and mineral content on the Young’s modulus of elasticity of compact bone. J. Biomech. 21(2), 131 – 139 (1988). 53. J. Currey, Physical characteristics affecting the tensile failure properties of compact bone. J. Biomech. 23, 837 – 844 (1990). 54. R. Martin and D. Burr, “The Structure, Function, and Adaptation of Compact Bone.” Raven Press, New York, (1989). 55. R. Martin and J. Ishida, The relative effects of collagen fiber orientation, porosity, density, and mineralization on bone strength. J. Biomech 22, 419 – 426 (1989). 56. M. Schaffler and D. Burr, Stiffness of compact bone: Effects of porosity and density. J. Biomech. 21, 13 – 16 (1988). 57. M. Grynpas, Age and disease-related changes in the mineral of bone. Calcif. Tissue. Int. 53 (Suppl. 1), S57 – S64 (1993). 58. P. Steiger, S. Cummings, D. Black, N. Spencer, and H. Genant, Agerelated decrements in bone mineral density in women over 65. J. Bone. Miner. Res. 7, 625 – 632 (1992). 59. S. Greenspan, L. Maitland, E. Myers, M. Krasnow, and T. Kido, Femoral bone loss progresses with age: A longitudinal study in women over age 65. J. Bone. Miner. Res. 9, 1959 – 1965 (1994). 60. D. Ulrich, B. Rietbergen, A. Laib, and P. Ruegsegger, Mechanical analysis of bone and its microarchitecture based on in vivo voxel images. Technol. Health. Care. 6(5-6), 421 – 427 (1998). 61. R. Muller and P. Ruegsegger, Micro-tomographic imaging for the nondestructive evaluation of trabecular bone architecture. Stud. Health. Technol. Inform. 40, 61 – 79 (1997). 62. P. Ruegsegger, B. Koller, and R. Muller, A microtomographic system for the nondestructive evaluation of bone architecture. Calcif. Tissue. Int. 58(1), 24 – 29 (1996). 63. S. Majumdar, D. Newitt, A. Mathur, D. Osman, A. Gies, E. Chiu, J. Lotz, J. Kinney, and H. Genant, Magnetic resonance imaging of trabecular bone structure in the distal radius: Relationship with Xray tomographic microscopy and biomechanics. Osteopor. Int. 6(5), 376 – 385 (1996). 64. S. Majumdar, M. Kothari, P. Augat, D. C. Newitt, T. M. Link, J. C. Lin, T. Lang, Y. Lu, and H. K. Genant, High-resolution magnetic resonance imaging: Three-dimensional trabecular bone architecture and biomechanical properties. Bone 22(5), 445 – 454 (1998). 65. B. D. Snyder, S. Piazza, W. T. Edwards, and W. C. Hayes, Role of trabecular morphology in the etiology of age-related vertebral fractures. Calcif. Tissue. Int. 53, S14 – S22 (1993). 66. C. Bergot, A. Laval-Jeantet, F Prêteux, and A. Meunier, Measurement of anisotropic vertebral trabecular bone loss during aging by quantitative image analysis. Calcif. Tissue. Int. 43, 143 – 149 (1988). 67. R. Goulet, S. Goldstein, M. Ciarelli, J. Kuhn, M. Brown, and L. Feldkamp, The relationship between the structural and orthogonal compressive properties of trabecular bone. J. Biomech 27, 375 – 389 (1994). 68. B. D. Snyder and W. C. Hayes, Multiaxial structure-property relations in trabecular bone. In “Biomechanics of Diarthrodial Joints.” pp. 31 – 59. V. C. Mow, A. Ratcliffe, and S. L.-Y. Woo, (eds.), Springer-Verlag, New York, (1990). 69. C. Turner, S. Cowin, J. Rho, R. Ashman, and J. Rice, The fabric dependence of the orthotropic elastic constants of cancellous bone. J. Biomech. 23, 549 – 561 (1990).
528 70. T. Hildebrand, A. Laib, R. Muller, J. Dequeker, and P. Ruegsegger, Direct three-dimensional morphometric analysis of human cancellous bone: Microstructural data from spine, femur, iliac crest, and calcaneus. J. Bone. Miner. Res. 14(7), 1167 – 1174 (1999). 71. T. M. Link, S. Majumdar, S. Grampp, G. Guglielmi, C. van Kuijk, H. Imhof, C. Glueer, and J. E. Adams, Imaging of trabecular bone structure in osteoporosis. Eur. Radiol. 9(9), 1781 – 1788 (1999). 72. E. Legrand, D. Chappard, C. Pascaretti, M. Duquenne, S. Krebs, V. Rohmer, M. F. Basle, and M. Audran, Trabecular bone microarchitecture, bone mineral density, and vertebral fractures in male osteoporosis. J. Bone. Miner. Res. 15(1), 13 – 19 (2000). 73. J. E. Aaron, P. A. Shore, R. C. Shore, M. Beneton, and J. A. Kanis, Trabecular architecture in women and men of similar bone mass with and without vertebral fracture. II. Three-dimensional histology. Bone 27(2), 277 – 282 (2000). 74. T. E. Ciarelli, D. P. Fyhrie, M. B. Schaffler, and S. A. Goldstein, Variations in three-dimensional cancellous bone architecture of the proximal femur in female hip fractures and in controls. J. Bone. Miner. Res. 15(1), 32 – 40 (2000). 75. A. Laib, T. Hildebrand, H. J. Hauselmann, and P. Ruegsegger, Ridge number density: A new parameter for in vivo bone structure analysis. Bone 21(6), 541 – 546 (1997). 76. A. Laib, H. J. Hauselmann, and P. Ruegsegger, In vivo high resolution 3D-QCT of the human forearm. Technol. Health. Care 6(5-6), 329 – 337 (1998). 77. S. Majumdar, T. M. Link, P. Augat, J. C. Lin, D. Newitt, N. E. Lane, and H. K. Genant, Trabecular bone architecture in the distal radius using magnetic resonance imaging in subjects with fractures of the proximal femur. Osteoporos. Int. 10(3), 231 – 239 (1999). 78. T. Lang, P. Augat, S. Majumdar, X. Ouyang, and H. K. Genant, Noninvasive assessment of bone density and structure using computed tomography and magnetic resonance. Bone 22(5 Suppl.), 149S – 153S (1998). 79. M. Silva and L. Gibson, Modeling the mechanical behavior of vertebral bone: Effects of age-related changes in microstructure. Bone 21, 191 – 199 (1997). 80. O. C. Yeh and T. M. Keaveny, Biomechanical effects of intraspecimen variations in trabecular architecture: A three-dimensional finite element study. Bone 25(2), 223 – 228 (1999). 81. M. Kothari, T. M. Keaveny, J. C. Lin, D. C. Newitt, and S. Majumdar, Measurement of intraspecimen variations in vertebral cancellous bone architecture. Bone 25(2), 245 – 250 (1999). 82. D. B. Burr, M. B. Schaffler, and R. G. Frederickson, Composition of the cement line and its possible mechanical role as a local interface in human compact bone. J. Biomech. 21(11), 939 – 945 (1988). 83. A. Burstein, J. Zika, K. Heiple, and L. Klein, Contribution of collagen and mineral to the elastic-plastic properties of bone. J. Bone Jt. Surg. 57, 956 – 961 (1975). 84. S. Carando, M. Portigliatti Barbos, A. Ascenzi, and A. Boyde, Orientation of collagen in human tibial and fibular shafts and possible correlation with mechanical properties. Bone 10, 139 – 142 (1989). 85. D. R. Carter and W. C. Hayes, Compact bone fatigue damage. I. Residual strength and stiffness. J. Biomech. 10, 325 – 337 (1977). 86. F. Evans and R. Vincintelli, Relations of the compressive properties of human cortical bone to histological structure and calcification. J. Biomech. 7, 1 – 10 (1974). 87. J. Katz and A. Meunier, The elastic anisotropy of bone. J. Biomech. 20, 1063 – 1070 (1987). 88. M. B. Schaffler, K. Choi, and C. Milgrom, Microcracks and aging in human femoral compact bone. Trans. Orthop. Res. Soc. 19, 190 (1994). 89. D. Burr, M. Forwood, D. Fyhrie, R. Martin, M. Schaffler, and C. Turner, Bone microdamage and skeletal fragility in osteoporotic and stress fractures. J. Bone Miner. Res. 12, 6 – 15 (1997).
MARY L. BOUXSEIN 90. S. Hoshaw, D. Cody, A. Saad, and D. Fhyrie, Decrease in canine proximal femoral ultimate strength and stiffnss due to fatigue damaage. J. Biomech. 30, 323 – 329 (1997). 91. S. Mori, R. Harruf, W. Ambrosius, and D. Burr, Trabecular bone volume and microdamage accumulation in the femoral heads of women with and without femoral neck fractures. Bone 21, 521 – 526 (1997). 92. M. Schaffler, K. Choi, and C. Milgrom, Aging and matrix microdamage accumulation in human compact bone. Bone 17, 521 – 525 (1995). 93. T. Wenzel, M. Schaffler, and D. Fyhrie, In vivo trabecular microcracks in human vertebral bone. Bone 19, 89 – 95 (1996). 94. T. Norman and Z. Wang, Microdamage of human cortical bone: Incidence and morphology in long bones. Bone 20, 375 – 379 (1997). 95. A. Courtney, W. Hayes, and L. Gibson, Age-related differences in post-yield damage in human cortical bone: Experiment and model. J. Biomech 29, 1463 – 1471 (1996). 96. M. C. H. van der Meulen, G. S. Beaupré, and D. R. Carter, Mechanobiologic influences in long bone cross-sectional growth. Bone 14, 635 – 642 (1993). 97. M. Bouxsein, K. Myburgh, M. van der Meulen, E. Lindenberger, and R. Marcus, Age-related differences in cross-sectional geometry of the forearm bones in healthy women. Calcif. Tissue Int. 54, 113 – 118 (1994). 98. S. Garn, C. Rohmann, B. Wagner, and W. Ascoli, Continuing bone growth throughout life: A general phenomenon. Am. J. Phys. Anthrop. 26, 313 – 318 (1967). 99. C. Ruff and W. Hayes, Subperiosteal expansion and cortical remodeling of the human femur and tibia with aging. Science 217, 945 – 947 (1982). 100. C. Ruff and W. Hayes, Cross-sectional geometry of Pecos Pueblo femora and tibiae: A biomechanical investigation. I. Method and general patterns of variation. Am. J. Phys. Anthropol. 60, 359 – 381 (1983). 101. R. Smith and R. Walker, Femoral expansion in aging women: Implications for osteoporosis and fractures. Science 145, 156 – 157 (1964). 102. T. Beck, C. Ruff, and K. Bissessur, Age-related changes in female femoral neck geometry: Implications for bone strength. Calcif. Tissue Int. 53 (Suppl. 1), S41 – S46 (1993). 103. D. Burr and R. Martin, The effects of composition, structure and age on the torsional properties of the human radius. J. Biomech. 16(8), 603 – 608 (1983). 104. R. Martin and P. Atkinson, Age and sex-related changes in the structure and strength of the human femoral shaft. J. Biomech. 10, 223 – 231 (1977). 105. H. Meema, Cortical bone atrophy and osteoporosis as a manifestation of aging. Am. J. Roentgenol. Radium Ther, Nucl. Med. 89, 1287 – 1295 (1963). 106. C. Ruff and W. Hayes, Sex differences in age-related remodeling of the femur and tibia. J. Orthop. Res. 6, 886 – 896 (1988). 107. M. F. Ericksen, Some aspects of aging in the lumbar spine. Am. J. Phys. Anthrop. 45, 575 – 580 (1976). 108. R. B. Martin and D. B. Burr, Non-invasive measurement of long bone cross-sectional moment of inertia by photon absorptiometry. J. Biomech. 17(3), 195 – 201 (1984). 109. T. J. Beck, C. B. Ruff, K. E. Warden, W. W. Scott, and G. U. Rao, Predicting femoral neck strength from bone mineral data. Invest. Radiol. 25, 6 – 18 (1990). 110. T. Beck, C. Ruff, W. Scott. Jr. C. Plato, J. Tobin, and C. Quan, Sex differences in geometry of the femoral neck with aging: A structural analysis of bone mineral data. Calcif. Tissue Int. 50, 24 – 29 (1992). 111. T. Nakamura, C. Turner, T. Yoshikawa, C. Slemenda, M. Peacock, D. Burr, Y. Mizuno, H. Orimo, Y. Ouchi, and C. Johnston, Do variations in hip geometry explain differences in hip fracture risk between Japanese and white Americans? J. Bone Minre. Res. 9, 1071 – 1076 (1994). 112. T. Yoshikawa, C. Turner, M. Peacock, C. Slemenda, C. Weaver, D. Teegarden, P. Markwardt, and D. Burr, Geometric structure of the
CHAPTER 19 Biomechanics of Age-Related Fractures
113.
114.
115.
116.
117. 118. 119.
120.
121.
122.
123.
124.
125.
126.
127.
128.
129.
130.
131.
132.
femoral neck measured using dual-energy x-ray absorptiometry. J. Bone Miner. Res. 9, 1053 – 1064 (1994). S. Cummings, D. Black, M. Nevitt, W. Browner, J. Cauley, H. Genant, S. Mascioli, and J. Scott, Appendicular bone density and age predict hip fractures in women. JAMA 263, 665 – 668 (1990). J. Michelson, A. Myers, R. Jinnah, Q. Cox, and M. Van Natta, Epidemiology of hip fractures among the elderly: Risk factors for fracture type. Clin. Orthop. Rel. Res. 311, 129 – 135 (1995). M. C. Nevitt, S. R. Cummings, and E. S. Hudes, Risk factors for injurious falls: A prospective study. J. Gerontol. 46, M164 – M170 (1991). M. Tinetti, Factors associated with serious injury during falls among elderly persons living in the community. J. Am. Geriatr. Soc. 35, 644 – 648 (1987). R. Sattin, Falls among older persons. Annu. Rev. Publ. Health. 13, 489 – 508 (1992). S. Cummings and M. Nevitt, A hypothesis: The causes of hip fracture. J. Gerontol. 44, M107-M111 (1989). A. van den Kroonenberg, W. Hayes, and T. McMahon, Hip impact velocities and body configurations for experimental falls from standing height. J. Biomech. 29, 807 – 811 (1996). M. Sabick, J. Hay, V. Goel, and S. Banks, Active responses decrease impact forces at the hip and shoulder in falls to the side J. Biomech. 32, 993 – 998 (1999). S. Robinovitch, W. Hayes, and T. McMahon, Prediction of femoral impact forces in falls on the hip. J. Biomech. Eng. 113, 366 – 374 (1991). S. Robinovitch, W. Hayes, and T. McMahon, Predicting the impact reponse of a nonlinear, single-degree-of-freedom shock-absorbing system from the measured step response. J. Biomech. Eng. 119, 221 – 227 (1997). S. Robinovitch, W. Hayes, and T. McMahon, Distribution of contact force during impact to the hip. Ann. Biomed. Eng. 25(3), 499 – 508 (1997). S. Robinovitch, T. McMahon, and W. Hayes, Energy-shunting hip padding system attenuates femoral impact force in a simulated fall. J. Biomech. Eng. 177, 409 – 413(1995). S. Robinovitch, T. McMahon, and W. Hayes, Force attenuation in trochanteric soft tissues during impact from a fall. J. Orthop. Res. 13, 956 – 962 (1995). J. Lauritzen, Hip fractures: Epidemiology, risk factors, falls, energy absorption, hip protectors, and prevention. Dan. Med. Bull. 44(2), 155 – 168 (1997). A. van den Kroonenberg, W. Hayes, and T. McMahon, Dynamic models for sideways falls from standing height. J. Biomech. Eng. 117, 309 – 318 (1995). D. Hemenway, D. Azrael, E. Rimm, D. Feskanich, and W. Willett, Risk factors for hip fracture in US men aged 40 through 75 years. Am. J. Public Health 84, 1843 – 1845 (1994). S. R. Cummings, D. M. Black, M. C. Nevitt, W. Browner, J. Cauley, C. Ensrud, H. K. Genant, L. Palermo, J. Scott, and T. M. Vogt, Bone density at various sites for prediction of hip fractures. Lancet 341, 72 – 75 (1993). M. Nevitt, O. Johnell, D. Black, K. Ensrud, H. Genant, and S. Cummings, Bone mineral density predicts non-spine fractures in very elderly women. Osteopor. Int. 4, 325 – 331 (1994). P. D. Ross, J. W. Davis, J. M. Vogel, and R. D. Wasnich, A critical review of bone mass and the risk of fractures in osteoporosis. Calcif. Tissue. Int. 46, 149 – 161 (1990). D. Hans, P. Dargent-Molina, A. Schott, J. Sebert, C. Cormier, P. Kotski, P. Delmas, J. Pouilles, G. Breart, and P. Meunier, Ultrasonographic heel measurements to predict hip fracture in elderly women: The EPIDOS prospective study. Lancet 348, 511 – 514 (1996).
529 133. N. Dalén, L. Hellström, and B. Jacobson, Bone mineral content and mechanical strength of the femoral neck. Acta Orthop. Scand. 47, 503 – 508 (1976). 134. I. Leichter, J. Y. Margulies, A. Weinreb, J. Mizrahi, G. C. Robin, B. Conforty, M. Makin, and B. Bloch, The relation between bone density, mineral content, and mechanical strength in the femoral neck. Clin. Orthop. Rel. Res. 163, 272 – 281 (1982). 135. J. Mizrahi, J. Y. Margulies, I. Leichter, and D. Deutsch, Fracture of the human femoral neck: Effect of density of the cancellous core. J. Biomed. Eng. 6, 56 – 62 (1984). 136. D. J. Sartoris, F. G. Sonner, J. Kosek, A. Gies, and D. R. Carter, Dualenergy projection radiography in the evaluation of femoral neck strength, density, and mineralization. Invest. Radiol. 20, 476 – 485 (1985). 137. A. Alho, T. Husby, and A. H iseth, Bone mineral content and mechanical strength: An ex vivo study on human femora at autopsy. Clin. Orthop. Rel. Res. 227, 292 – 297 (1988). 138. D. D. Cody, G. J. Gross, F. J. Hou, H. J. Spencer, S. A. Goldstein, and D. P. Fyhrie, Femoral strength is better predicted by finite element models than QCT and DXA. J. Biomech. 32(10), 1013 – 1020 (1999). 139. J. H. Keyak, S. A. Rossi, K. A. Jones, and H. B. Skinner, Prediction of femoral fracture load using automated finite element modeling. J. Biomech. 31(2), 125 – 133 (1998). 140. J. Lotz and W. Hayes, The use of quantitative computed tomography to estimate risk of fracture of the hip from falls. J. Bone Jt. Surg. 72-A(5), 689 – 700 (1990). 141. A. Courtney, E. F. Wachtel, E. R. Myers, and W. C. Hayes, Age-related reductions in the strength of the femur tested in a fall loading configuration. J. Bone Jt Surg. 77, 387 – 395 (1995). 142. A. Courtney, E. F. Wachtel, E. R. Myers, and W. C. Hayes, Effects of loading rate on the strength of the proximal femur. Calcif Tissue Int. 55, 53 – 58 (1994). 143. M. Bouxsein, A. Courtney, and W. Hayes, Ultrasound and densitometry of the calcaneus correlate with the failure loads of cadaveric femurs. Calcif. Tissue Int. 56, 99 – 103 (1995). 144. Deleted at proof. 145. X. Cheng, G. Lowet, S. Boonen, P. Nicholson, P. Brys, P. Nijs, and J. Dequeker, Assessment of the strength of proximal femur in vitro: Relationship to femoral bone mineral density and femoral geometry. Bone 20(3), 213 – 218 (1997). 146. E. M. Lochmuller, J. B. Zeller, D. Kaiser, F. Eckstein, J. Landgraf, R. Putz, and R. Steldinger, Correlation of femoral and lumbar DXA and calcaneal ultrasound, measured in situ with intact soft tissues, with the in vitro failure loads of the proximal femur. Osteopor. Int. 8(6), 591 – 598 (1998). 147. M. L. Bouxsein, B. S. Coan, and S. C. Lee, Prediction of the strength of the elderly proximal femur by bone mineral density and quantitative ultrasound measurements of the heel and tibia. Bone 25(1), 49 – 54 (1999). 148. T. P. Pinilla, K. C. Boardman, M. L. Bouxsein, E. R. Myers, and W. C. Hayes, Impact direction from a fall influences the failure load of the proximal femur as much as age-related bone loss. Calcif. Tissue Int. 58(4), 231 – 235 (1996). 149. C. Ford, T. Keaveny, and W. Hayes, The effect of impact direction on the structural capacity of the proximal femur during falls. J. Bone Miner. Res. 11, 377 – 383 (1996). 150. K. G. Faulkner, S. R. Cummings, D. Black, L. Palermo, C. C. Glüer, and H. K. Genant, Simple measurement of femoral geometry predicts hip fracture: The study of osteoporotic fracture. J. Bone Miner. Res. 8, 1211 – 1217 (1993). 151. C. C. Glüer, S. R. Cummings, A. Pressman, J. Li, K. Glüer, K. G. Faulkner, S. Grampp, and H. K. Genant, Prediction of hip fractures from pelvic radiographs: The study of osteoporotic fractures. J. Bone Miner. Res. 9, 671 – 677 (1994).
530 152. J. Lauritzen, M. Petersen, and B. Lund, Effect of external hip protectors on hip fractures. Lancet 341, 11 – 13 (1993). 153. M. Parker, L. Gillespie, and W. Gillespie, Hip protectors for preventing hip fractures in the elderly. Cochrane Database Syst. Rev. 2, CD001255 (2000). 154. E. R. Myers, S. N. Robinovitch, S. L. Greenspan, and W. C. Hayes, Factor of risk is associated with frequency of hip fracture in a casecontrol study. Trans. Orthop. Res. Soc. 19, 526 (1994) 155. T. O’Neill, J. Varlow, D. Felsenberg, O. Johnell, K. Weber, F. Marchant, P. Delmas, C. Cooper, J. Kanis, and A. Silman, Variation in vertebral height ratios in population studies. J. Bone Miner. Res. 9, 1895 – 1907 (1994). 156. M. Kleerekoper and D. Nelson, Vertebral fracture or vertebral deformity? Calcif. Tissue Int. 50, 5 – 6 (1992). 157. C. Cooper, E. Atkinson, W. Ó. Fallon, and L. Melton, Incidence of clinically diagnosed vertebral fractures: A population-based study in Rochester, Minnesota, 1985 – 1989. J. Bone Miner. Res. 7(2), 221 – 227 (1992). 158. M. C. Nevitt, Epidemiology of osteoporosis. In “Rheumatic Disease Clinics of North America: Osteoporosis” N. E. Lane, (ed.), Vol. 20, pp. 535 – 560. Saunders, Philadelphia, (1994). 159. E. Myers and S. Wilson, Biomechanics of osteoporosis and vertebral fractures. Spine 22, 25S – 31S (1997). 160. U. Bengnér, O. Johnell, and I. Redlund-Johnell, Changes in incidence and prevalence of vertebral fractures during 30 years. Calcif. Tissue Int. 42, 293 – 296 (1988). 161. U. Patel, S. Skingle, G. Campbell, A. Crisp, and I. Boyle, Clinical profile of acute vertebral compression fractures in osteoporosis. Br. J. Rheumato. 30, 418 – 421 (1991). 162. S. Santavirta, Y. Konttinen, M. Heliövaara, P. Knekt, P. Lüthje, and A. Aromaa, Determinants of osteoporotic thoracic vertebral fracture. Acta Orthop. Scand. 63(2), 198 – 202 (1992). 163. E. Myers, S. Wilson, and S. Greenspan, Vertebral fractures in the elderly occur with falling and bending. J. Bone Miner. Res. 11, S355 (1996). 164. G. B. J. Andersson, R. Örtengren, and A. Schultz, Analysis and measurement of the loads on the lumbar spine during work at a table. J. Biomech. 13, 513 – 520 (1980). 165. A. B. Schultz, G. B. J. Andersson, K. Haderspeck, R. Örtengren, M. Nordin, and R. Björk, Analysis and measurement of lumbar trunk loads in tasks involving bends and twists. J. Biomech. 15(9), 669 – 675 (1982). 166. A. B. Schultz and G. B. J. Andersson, Analysis of loads on the lumbar spine. Spine 6(1), 76 – 82 (1981). 167. A. Schultz, G. Andersson, R. Ortengren, K. Haderspeck, and A. Nachemson, Loads on the lumbar spine: Validation of a biomechanical analysis by measurements of intradiscal pressures and myoelectric signals. J. Bone Jt. Surg. 64A, 713 – 720 (1982). 168. M. Jäger and A. Luttmann, Biomechanical analysis and assessment of lumbar stress during load lifting using a dynamic 19-segment human model. Ergonomics 32, 93 – 112 (1989). 169. A. Nachemson and G. Elfström, Intravital dynamic pressure measurements in lumbar discs. Scand. J. Rehab. Med. Suppl. 1, 3 – 40 (1970). 170. A. L. Nachemson, The lumbar spine: An orthopaedic challenge. Spine 1(1), 59 – 69 (1976). 171. J. Cholewicki, S. McGill, and R. Norman, Comparison of muscle forces and joint load from an optimization and EMG assisted lumbar spine model: Towards development of a hybrid approach. J. Biomech. 28, 321 – 331 (1995). 172. A. Schultz, G. B. J. Andersson, R. Örtengren, R. Björk, and M. Nordin, Analysis and quantitative myoelectric measurements of loads on the lumbar spine when holding weights in standing postures. Spine 7, 390 – 397 (1982).
MARY L. BOUXSEIN 173. A. Yettram and M. Jackman, Equilibrium analysis for the forces in the human spinal column and its musculature. Spine 5(5), 402 – 411 (1980). 174. S. Wilson, Development of a model to predict the compressive forces on the spine associated with age-related vertebral fractures. Massachusetts Institute of Technology (1994). 175. S. Greenspan, L. Maitland-Ramsey, and E. Myers, Classification of osteoporosis in the elderly is dependent on site-specific analysis. Calcif. Tissue Int. 58, 409 – 414 (1996). 176. A. van den Kroonenberg, S. Wilson, E. Myers, W. Hayes, and T. McMahon Impact velocities and body configurations for backward falls from standing height. Submitted for publication. 177. L. Mosekilde and L. Mosekilde, Vertebral structure and strength in vivo and in vitro. Calcif. Tissue Int. 53, S121 – S126 (1993). 178. M. Silva, T. Keaveny, and W. Hayes, Load sharing between the shell and centrum in the lumbar vertebral body. Spine 22(2), 140–150 (1997). 179. L. Mosekilde, Osteoporosis: Mechanisms and models. In “Anabolic Treatments for Osteoporosis.” J. Whitfield, and P. Morley, (eds.), pp. 31 – 58. CRC Press, Boca Raton, FL, (1998). 180. K. Faulkner, C. Cann, and B. Hasedawa, Effect of bone distribution on vertebral strength: Assessment with a patient-specific nonlinear finite element analysis. Radiology 179, 669 – 674 (1991). 181. S. D. Rockoff, E. Sweet, and J. Bleustein, The relative contribution of trabecular and cortical bone to the strength of human lumbar vertebrae. Calcif. Tissue. Res. 3, 163 – 175 (1969). 182. R. J. McBroom, W. C. Hayes, W. T. Edwards, R. P. Goldberg, and A. A. White, Prediction of vertebral body compressive fracture using quantitative computed tomography. J. Bone Jt. Surg. 67-A, 1206 – 1214 (1985). 183. N. Yoganandan, J. Myklebust, J. Cusick, C. Wilson, and A. Sances, Functional biomechanics of the thoracolumbar vertebral cortex. Clin. Biomech. 3, 11 – 18 (1988). 184. M. H. Bartley, J. S. Arnold, R. K. Haslam, and W. S. S. Jee, The relationship of bone strength and bone quantity in health, disease, and aging. J. Gerontol. 21, 517 – 521 (1966). 185. M. Biggemann, D. Hilweg, S. Seidel, M. Horst, and P. Brinckmann, Risk of vertebral insufficiency fractures in relation to compressive strength predicted by quantitative computed tomography. Eur. J. Radiol. 13, 6 – 10 (1991). 186. M. Biggemann, D. Hilweg, and P. Brinckmann, Prediction of the compressive strength of vertebral bodies of the lumbar spine by quantitative computed tomography. Skel. Radiol. 17, 264 – 269 (1988). 187. P. Brinckmann, M. Biggeman, and D. Hilweg, Prediction of the compressive strength of human lumbar vertebrae. Clin. Biomech. 4, S1 – S27. 188. D. Cody, S. Goldstein, M. Flynn, and E. Brown, Correlations between vertebral regional bone mineral density (rBMD) and whole bone fracture load. Spine 16, 146 – 154 (1991). 189. S. A. Eriksson, B. O. Isberg, and J. U. Lindgren, Prediction of vertebral strength by dual photon absorptiometry and quantitative computed tomography. Calcif. Tissue Int. 44, 243 – 250 (1989). 190. T. Hansson, B. Roos, and A. Nachemson, The bone mineral content and ultimate compressive strength of lumbar vertebrae. Spine 5(1), 46 – 55 (1980). 191. M. Moro, A. T. Hecker, M. L. Bouxsein, and E. R. Myers, Failure load of thoracic vertebrae correlates with lumbar bone mineral density measured by DXA. Calcif. Tissue Int. 56, 206 – 209 (1995). 192. B. Myers, K. Arbogast, B. Lobaugh, K. Harper, W. Richardson, and M. Drezner, Improved assessment of lumbar vertebral body strength using supine lateral dual-energy x-ray absorptiometry. J. Bone Miner. Res. 9, 687 – 693 (1994). 193. A. Vesterby, L. Mosekilde, H. Gundersen, F. Melsen, L. Mosekilde, K. Holme, and S. S rensen, Biologically meaningful determinants of the in vitro strength of lumbar vertebrae. Bone 12, 219 – 224 (1991).
531
CHAPTER 19 Biomechanics of Age-Related Fractures 194. A. Tabensky, J. Williams, V. DeLuca, E. Briganti, and E. Seeman, Bone mass, areal, and volumetric bone density are equally accurate, sensitive, and specific surrogates of the breaking strength of the vertebral body: An in vitro study. J. Bone Miner. Res. 11, 1981 – 1988 (1996). 195. X. G. Cheng, P. H. Nicholson, S. Boonen, G. Lowet, P. Brys, J. Aerssens, G. Van der Perre, and J. Dequeker, Prediction of vertebral strength in vitro by spinal bone densitometry and calcaneal ultrasound. J. Bone Miner. Res. 12(10), 1721 – 1728 (1997). 196. E. M. Lochmuller, F. Eckstein, D. Kaiser, J. B. Zeller, J. Landgraf, R. Putz, and R. Steldinger, Prediction of vertebral failure loads from spinal and femoral dual- energy X-ray absorptiometry, and calcaneal ultrasound: An in situ analysis with intact soft tissues. Bone 23(5), 417 – 424 (1998). 197. M. Biggeman and P. Brinckman, Biomechanics of osteoporotic vertebral fractures. In “Vertebral Fracture in Osteoporosis.” H. Genant,
198.
199. 200. 201.
202. 203.
M. Jergas, C. van Kuijk, (eds.), pp. 21 – 34. Osteoporosis Research Group, University of California, San Francisco, (1995). W. C. Hayes, Biomechanics of cortical and trabecular bone: Implications for assessment of fracture risk. In “Basic Orthopaedic Biomechanics.” V. C. Mow, and W. C. Hayes (eds.), pp. 93 – 142. Raven Press, New York, (1991). Hologic, Hologic Reference Database, (1994). R. Eastell, Treatment of postmenopausal osteoporosis. N. Engl. J. Med. 338(11), 736 – 746 (1998). J. Casalena, D. Streit, P. Cavanaugh, Preliminary design of a dual stiffness floor for injury prevention. In “Injury Prevention through Biomechanics” K. Yang (ed.), p. 155. B. Maki, G. Fernie, Impact attenuation of floor coverings in simulated falling accidents. Appl. Ergon. 21, 107 – 114 (1990). T. M. Keaveny and W. Hayes, Mechanical properties of cortical and trabecular bone. In “Bone” (B. Hall, ed.), Vol. 7. CRC Press, Boca Raton, FL, (1993).
CHAPTER 20
Introduction to Epidemiologic Methods
I. II. III. IV. V. VI.
JENNIFER L. KELSEY
Department of Health Research and Policy, Division of Epidemiology, Stanford University School of Medicine, Stanford, California 94305
MARYFRAN SOWERS
Department of Epidemiology, University of Michigan School of Public Health, Ann Arbor, Michigan 48109
Introduction Descriptive and Analytic Studies Study Designs Some Useful Epidemiologic Concepts Some Frequently Used Statistics Criteria for Deciding Whether an Association Is Causal
VII. Sample Size Considerations VIII. Measurement Error IX. Measuring Diet and Bone Turnover Status as Examples of Measurement Issues X. Conclusions References
I. INTRODUCTION
magnitude and impact of diseases or other conditions in populations or in selected subgroups of populations. This information can be used in setting priorities for investigation and control, in deciding where preventive efforts should be focused, in evaluating the efficacy of therapeutic procedures, and in determining what type of treatment facilities are needed. For instance, knowledge of current hip fracture incidence rates in various parts of the world and projected large increases in the numbers of elderly in developing countries indicate that hip fractures will become major problems in all parts of the world in the future [3]. Accordingly, identifying risk factors in these regions and planning for more services become high priority. Epidemiologic studies may also be used to learn about the natural history, clinical course, and pathogenesis of diseases. Currently, several studies are being undertaken of the clinical course of vertebral osteoporosis. One study [4], for instance, reported that
Many of the chapters in this section focus on epidemiologic aspects of osteoporosis. Since many readers will have only a passing acquaintance with the terms, methods, and concepts used in epidemiology, we start this section with an introduction to epidemiologic methods. It is hoped that this introduction will be of help in reading subsequent chapters in this section, in reading and evaluating studies directly from the published literature, and in understanding some of the reasons that different results sometimes are reported from various studies of the same issue. Much of this material is adapted from a textbook of epidemiology [1] and from a paper on epidemiologic methods [2]. Epidemiology is the study of the occurrence and distribution of diseases and other health-related conditions in populations. It is used for many purposes. One is to determine the
OSTEOPOROSIS, SECOND EDITION VOLUME 1
535
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
536
KELSEY AND SOWERS
minor fractures, unless multiple, are not associated with back pain or loss of height. Another [5] found that most back pain cannot be attributed to vertebral deformities and that vertebral deformities cause significant pain, disability, or height loss only if vertebral height ratios fall four standard deviations below the normal mean. Most commonly, epidemiologic studies are undertaken to identify causes of disease, and it is this application of epidemiology that is the focus of this chapter.
II. DESCRIPTIVE AND ANALYTIC STUDIES A. Descriptive Epidemiology Descriptive studies provide information on patterns of disease occurrence in populations according to such attributes as age, gender, race, ethnicity, marital status, social class, occupation, geographic area, and time of occurrence. Usually, routinely collected data such as from hospital discharge records, death certificates, and general health surveys are used for descriptive studies. This information can be used to indicate the magnitude of a problem or to provide preliminary ideas about etiology. For instance, several decades ago, the marked loss of bone mass and increasing hip fracture incidence rates in women after age 50 or so suggested that menopause and its accompanying decreased estrogen levels might be involved in the etiology of osteoporosis [6]. This hypothesis has been borne out by many subsequent studies [7,8]. The variation in hip fracture incidence rates around the world has led to several hypotheses about reasons for the differences, including differences in diet, physical activity, frequency of falling [9], and, most recently, hip axis length [10]. The increasing hip fracture incidence rates in northern European countries have led to a great deal of speculation about reasons for the increases [9]. Correlations between a putative risk factor and a disease according to geographic region or over time, however, at most provide weak evidence that the factor causes the disease. There are so many differences between lifestyles and other characteristics of people living in different geographic areas and at different periods of time that singling out one factor as being the reason for the difference in incidence rates is usually impossible. Countries with high incidence rates of hip fracture compared to those with low incidence rates have many differences in their diets, as well as different levels of physical activity, neuromuscular functioning, medication use, and perhaps hip axis length and propensity to fall. Accordingly, analytic epidemiologic studies designed specifically to test hypotheses are used to provide more definitive information.
B. Analytic Epidemiology Analytic studies are designed to test causal hypotheses that have been generated from descriptive epidemiology, clinical observations, laboratory studies, and other sources, including analytic studies undertaken for other purposes. Whereas descriptive epidemiology describes how a disease is distributed in a population, analytic epidemiology tries to explain why. Because analytic studies often require the collection of new data, they tend to be more expensive than descriptive studies, but, if designed and executed properly, generally allow more definitive conclusions to be reached about causation. Most epidemiologic studies are observational; i.e., the investigator observes what is occurring in the study populations of interest and does not interfere with what he or she observes. For instance, an investigator could observe existing physical activity levels among individuals and relate those activity levels to bone mass or fracture occurrence. In contrast, in an experimental study, the investigator intervenes and assigns members of the study population to one exposure or treatment category or another, as in a randomized clinical trial. In such a trial, an investigator would randomly assign individuals (or communities) to programs with varying levels of physical activity and note changes in bone mass or the occurrence of fractures following implementation of the programs. Observational epidemiologic studies, to be described first, include case-control, cohort, and cross-sectional studies, as well as some hybrid designs. Then, experimental studies will be discussed briefly.
III. STUDY DESIGNS A. Case-Control Studies Case-control studies are those in which the investigator selects persons with a given disease (the cases) and persons without the given disease (the controls) for study. Usually the cases enter the study as they are diagnosed over time, and controls enter the study as they are identified over the same time period. The proportion of cases and controls with certain characteristics or past exposure to possible risk factors (e.g., ever used estrogen) are then determined and compared. For instance, Table 1 shows that in a case-control study of hip fractures in women [11], 16.6% of hip fracture cases and 23.5% of controls had ever used estrogen, suggesting some protection from use of estrogen. For a numerical measurement such as weight, the mean level of the characteristic of interest in the cases is compared to the mean level of the characteristic in the controls. In that same case-control study [11], the mean Quetelet index [weight (kg)/height2 (m2)] was 22.3 in cases and 25.6 in controls, suggesting an increased risk among thin women.
537
CHAPTER 20 Introduction to Epidemiologic Methods
TABLE 1
Case-Control Study Relating Hip Fracture in Women to Previous Estrogen Use Hip fracture Cases
Ever used estrogen Yes No
Controls
Total
26
39
65
131
127
258 323
Total
157
166
Percent having used estrogen
16.6
23.5
Note. Source: Nieves et al. [11].
Cases are generally persons seeking medical care for the disease. Only newly diagnosed cases are usually included in order to be more certain that the risk factor preceded the disease rather than being a consequence of the disease and so that rapidly fatal cases or cases of short duration are appropriately represented. For instance, fracture cases tend to change their physical activity patterns after the fracture, and if cases who had incurred their fracture in the past are included, it would be more difficult to differentiate the physical activity pattern that preceded the fracture from the physical activity pattern that resulted from the fracture. Many case-control studies have been undertaken to identify risk factors for hip fracture. Because hip fracture is almost always seen in a hospital, virtually all cases can be identified from hospital sources. For fractures of many other common sites, such as distal forearm fractures, only a select portion of all cases would be seen in hospital, so the representativeness of cases in hospital settings would be highly questionable. Efforts would have to be made to identify cases seen on an outpatient basis as well, a considerably more extensive undertaking in most settings. Choice of an appropriate control group is often one of the most difficult and controversial aspects of designing a case-control study. A useful working concept of what a control group should be has been provided by Miettinen [12]: the controls should be selected in an unbiased manner from those individuals who would have been included in the case series if they had developed the disease under study. The choice of which control group to use generally depends on the source of the cases, the relative costs of obtaining the various types of controls, and the facilities available to the investigator. If cases consist of all individuals developing the disease of interest in a defined population, then the single best control group would generally be a random sample of individuals (in the same age range and of the same sex, for instance) from the same source population who have not developed the disease. If cases are identified at certain hospitals that do not cover a defined geographic area, it is usually impossible to specify the source population from
which the cases arose. In this situation, controls are often chosen from among other patients admitted to similar services of the same hospitals as the cases, as one wants to obtain a source of controls subject to the same selective factors as the cases. It is usually desirable to include as controls people with a variety of other conditions so that no one disease is unduly represented among the control group. Generally, it is important to exclude potential controls who have had their disease for a long time because, like the cases, the presence of their disease may have influenced their exposure to possible risk factors. Such characteristics as physical activity, diet, weight, and medication use may change as a result of many diseases. Controls in case-control studies of hip fracture have included people from the same retirement community as the cases [13], people from the same prepaid health care plan as the cases [14], people selected at random from lists of Medicare recipients of the Health Care Financing Administration (HCFA) files [15], people sampled from the general population of the same geographic areas as the cases [16], and patients seen at the same hospitals as the cases for other conditions [17,18]. Obtaining controls through random digit dialing, a procedure frequently used in case-control studies in younger populations, often is not practical for case-control studies that include large numbers of elderly individuals, as an enormous number of telephone calls must be made to find an appropriate elderly person. In countries or other geographic units with population registries, controls might consist of a random sample of persons from the same geographic area as the cases in the appropriate age groups, as listed in the register. The various types of control groups have their own strengths and weaknesses. If hospital controls are used, the controls by definition are different from the cases in that they generally have another disease for which they have sought medical care. If smoking is the putative risk factor, for instance, there may be concern that hospital controls include more than their fair share of smokers, as smoking is associated with many diseases that require hospitalization. A major concern in using controls from HCFA files or obtained though random digit dialing in younger age groups is that a substantial proportion of potential controls (typically 30 – 40% in otherwise well-executed studies) may decline to participate, and it is possible that participants and nonparticipants differ in ways that affect study results. Cases and controls from prepaid health care plans or from retirement communities are generally more likely to participate in studies, thus giving higher response rates. In some situations when no single control group is obviously best, it may be helpful to have more than one control group with which to compare the cases. Information on exposure to putative risk factors may be obtained in several ways, depending on the nature of the exposure. Risk factor data are obtained most commonly by
538 means of questionnaires administered by trained interviewers to cases and controls. For instance, the only practical way to find out about a person’s smoking habits is to ask the person. Existing records may sometimes be used to find out about exposures such as medication use. Physical measurements or laboratory tests on sera or other tissue drawn from cases and controls may also be used, but it must be kept in mind that measurements of such attributes as bone density or markers of bone turnover made after the fracture has occurred may differ from the values of these attributes before the fracture occurred. Whichever methods are used, ensuring that ascertainment of exposure status is comparable in cases and controls is of the utmost importance. Certain cases and controls may be excluded from a study, such as those with other disorders that affect calcium metabolism and that are not of interest to the study being conducted. Although excluding cases and controls limits generalizability, the validity of the comparison between cases and controls must take highest priority. The general principle that the same exclusion criteria should be applied to cases and controls should be maintained whenever possible. If cases are restricted to a certain sex or age range, controls should be similarly restricted. If cases with certain medical conditions are excluded, then controls with those conditions should also be excluded. While equal application of exclusion criteria may sound reasonable and easy, in practice this may be more difficult. Undiagnosed disease such as Paget’s disease may exist among controls in a casecontrol study of hip fracture, as the controls may not have had as thorough a diagnostic workup as the cases. Inequitable access to health care between cases and controls can exacerbate this problem. In summary, case-control studies can provide much useful information about risk factors for diseases, including hip fracture and other fractures, in settings where fractures can be ascertained readily. Case-control studies are by far the most frequently undertaken type of analytic epidemiologic study. They can generally be carried out in a much shorter period of time than cohort studies (to be discussed later), do not require nearly so large a sample size, and consequently are less expensive. For a rare disease, case-control studies are usually the only practical approach to identifying risk factors. Certain potential problems and limitations that may arise in case-control studies need to be carefully considered before deciding whether a case-control study is appropriate in a given situation. Sackett [19] and Austin et al. [20] have listed and discussed a large number of possible sources of bias and error in case-control studies. Among the most common concerns are that (i) information on potential risk factors may not be available either from records or the participants’ memories, (ii) information on other relevant variables may not be available either from records or from the participants’ memories, (iii) cases may search for a cause for their
KELSEY AND SOWERS
disease and thereby be more likely to report an exposure than controls, (iv) the investigator may be unable to determine with certainty whether the agent was likely to have caused the disease or whether the occurrence of the disease was likely to have caused the person to be exposed to the agent, (v) identifying and assembling a case group representative of all cases may be unduly difficult, (vi) identifying and assembling an appropriate control group may be unduly difficult, and (vii) participation rates may be low. Because of these potential weaknesses, the case-control study is considered by some to be a type of study that merely provides leads to be followed up by more definitive cohort studies. However, decisions as to whether preventive actions should be taken often must be reached on the basis of information obtained from case-control studies. Each casecontrol study should be evaluated individually, as some studies are affected by error and bias very little and others a great deal.
B. Cohort Studies In a typical prospective cohort study, persons free of the disease of interest at the time of entry into the study are classified according to whether they are exposed to the risk factors of interest. The cohort is then followed for a period of time (which may be many years), and the incidence rates (number of new cases of disease per population at risk per unit time) or mortality rates (number of deaths per population at risk per unit time) in those exposed or not exposed are compared. A prospective cohort study may also involve measuring exposure status at the beginning of a study and determining how this relates to changes in an attribute (such as bone mass) over time. Cohort studies have a major advantage over case-control studies in that exposures or characteristics of interest are measured before the disease has developed (or before changes in an attribute take place). However, prospective cohort studies generally require large sample size, long-term follow-up of study subjects, large monetary expense, and complex administrative and organizational arrangements. The outcome of interest must be relatively common, or prohibitively large numbers of cohort members will be required in order to ensure adequate numbers experiencing that outcome. Therefore, prospective cohort studies are usually initiated under two circumstances: (1) when sufficient (but not definitive) evidence has been obtained from less expensive studies to warrant a more expensive cohort study and (2) when a new agent (e.g., a new widely used medication) is introduced that may alter the risk for several diseases. Results from a prospective cohort study of the association between bone mass measured at the mid-radius and fractures at all sites among a cohort of 521 Caucasian women of ages 22 – 95 years followed for an average of 6.5
539
CHAPTER 20 Introduction to Epidemiologic Methods
years are shown in Table 2 [21]. It can be seen that as bone mass increases, incidence rates for fracture decrease. Before this study was undertaken, it was considered possible on the basis of results from case-control studies that low bone mass could be a consequence rather than a cause of fractures. This prospective cohort study clearly showed that low bone mass preceded the fractures. In Table 2 the column headed “Person-years of followup” should be noted. The figures in this column indicate the total number of years that women in this cohort were actually under observation for the occurrence of a fracture. Because women entered and left the cohort at different times, the total number of years that each woman was under surveillance by the investigators for fracture occurrence had to be taken into account. The sum of the number of years each woman in a given bone mass category was at risk and under observation is the number of person-years for that category. In most studies, cohort members are removed from follow-up when they experience a first event because they are no longer at risk for being an incident case. In this study, fractures of different sites were each counted so that women were still considered at risk for a fracture of other sites even though they experienced a fracture of one site. Studies of bone mass are increasingly being designed as longitudinal studies in which individuals are measured repeatedly over time. This enables one to track an individual’s changes in bone mass or in a marker of bone turnover over time, or to track an entire cohort’s average changes in bone mass or in a marker of bone turnover over time. In contrast, in a cross-sectional study (see later), one could examine the relationship between age and bone mass in a group of people at one point in time, but could not determine how bone mass changes in individuals as they become older. When the same individuals are measured repeatedly over time, one can also more readily determine
TABLE 2 Prospective Cohort Study Relating Bone Mass (g/cm) at Midradius to Number of Fractures of All Sites in Caucasian Women
Bone mass
No. of fractures
Personyears of follow-up
Incidence rate per person-year
0.60
46
415.5
0.111
0.60 – 0.69
25
554.2
0.045
0.70 – 0.79
46
861.1
0.053
0.80 – 0.89
15
776.8
0.019
0.90 – 0.99
5
521.1
0.010
1.00
0
260.2
0
Note. Source: Hui et al. [21].
whether a change is attributable to age or to a specific event, such as menopause, that is correlated with age. Sowers et al. [22] provided an example of a prospective cohort study to identify attributes associated with changes in radial bone mineral density. They found that lower weight, smaller triceps, and less arm muscle mass were predictive of increased 5-year loss of bone mineral density in postmenopausal women, whereas use of estrogen for 5 years or longer and current use of thiazide diuretics were predictive of less loss of bone mineral density. Another example of repeated measurement of one variable over time would be a study to assess drift in a densitometer over a period of several months by measuring a phantom weekly over this period. Statistical analysis of repeated measurements requires specific methods that take into account that the measurements over time on each individual tend to be correlated with each other. (See Diggle et al. [23] for a description of methods of analysis of longitudinal data.) Cohorts are sometimes chosen because they are representative of the general population, such as in the Framingham Heart Study [24]. Studies to identify risk factors for hip fracture and loss of bone mass have been able to take advantage of data collected on these cohorts, even though they were not originally designed for this purpose. Although the ability to generalize from such studies makes them highly desirable, they are usually very expensive and tend to be associated with relatively large numbers of people lost to follow-up. Also, an exposure of interest may be uncommon in the general population so it sometimes may be more efficient to select a cohort with a higher proportion exposed or to select a cohort at higher risk of the disease, so that the sample size does not have to be so large. Other cohorts used in studies of fractures include the residents of retirement communities [25], members of prepaid health care plans [26], the Nurses’ Health Study cohort [27], and women recruited from available listings in four areas of the United States [28]. A related type of study is the retrospective cohort study (also called a historical cohort study). In this design, investigators assemble a cohort by reviewing records to identify exposures in the past (often decades ago). Based on recorded exposure histories, cohort members are divided into exposed and nonexposed groups or according to level of exposure. The investigator then reconstructs their subsequent disease experience up to some defined point in the more recent past or up to the present time. For instance, 35,767 people in a county in Norway participated in a health screening program in 1984 – 1986. Each participant had filled out a questionnaire on some background factors and personal health characteristics, including smoking habits, health status, weight, height, and physical activity. By linking this information to hospital admissions for hip fracture over the 3-year period 1986 – 1989, Forsen et al. [29] were able to determine that cigarette smoking was
540
KELSEY AND SOWERS
a risk factor for hip fracture independent of thinness, physical inactivity, and self-reported ill health. In the United States, the first National Health and National Examination Survey (NHANES I) has also provided information on potential risk factors for retrospective cohort studies. People examined in NHANES I during 1971 – 1975 were followed through 1987 in the NHANES I Epidemiologic Follow-Up Study. Looker et al. [30] related data obtained during 1971 – 1975 on calcium consumption of participants in NHANES I to subsequent occurrence of hip fracture. These data suggested that calcium may lower the risk for hip fracture in women who were at least 6 years past menopause and who did not use hormone replacement therapy. Retrospective cohort studies have many of the advantages of prospective cohort studies, but can be completed in a much more timely fashion and are therefore much less expensive. However, only when the necessary information on past exposure and other characteristics of interest has been reliably recorded can a retrospective cohort study be reasonably undertaken. In addition, it must be possible to trace a large proportion of the cohort members in order to determine whether they in fact developed the disease of interest. As with prospective cohort studies, it is usually feasible to carry out a retrospective cohort study only when the outcome of interest is relatively common. It is also frequently important to obtain information on characteristics of the cohort members other than the exposure history and the outcome of primary interest so as to make sure that those with and those without the exposure of interest are comparable in other relevant respects. If such information is not available, then the interpretation of the study results may be ambiguous. Thus, retrospective cohort studies are economical and useful when the necessary information on the past history of study subjects is available and when the outcome of interest is relatively common. In many situations, however, these conditions are not met.
C. Cross-Sectional Studies In a cross-sectional, or prevalence, study, exposure to a hypothesized risk factor or other characteristic of interest and the occurrence of a disease are measured at one time (or over a relatively short period of time) in a study population. Prevalence rates of disease (number of cases of existing disease per population at risk at a given point in time or time period) among those with and without the exposure or characteristic of interest are then compared. For a quantitative variable such as bone mineral density, the mean value of the exposed and nonexposed groups may be compared. For instance, in a cross-sectional study of the association between back muscle strength and spinal osteoporosis, it was found that osteoporotic women had lower muscle
strength than women without osteoporosis [31]. However, in cross-sectional studies, it is often difficult to differentiate cause and effect. As the authors themselves pointed out, a longitudinal study would be needed to determine whether weak back muscles contribute to the development of osteoporosis or are a consequence of osteoporosis. Cross-sectional studies of the association between calcium supplementation and bone mineral density would be difficult to interpret because people with low bone mineral density might take calcium once they were told about their low bone mineral density. Interpretation of findings from crosssectional studies is generally clear only for potential risk factors that will not change as a result of the disease, such as genotype. Prevalence studies include all cases of disease, new and old. Therefore, a second limitation of cross-sectional studies is that the case group tends to be weighted toward individuals with disease of long duration, as the chances for cases of long duration to be included are greater than those for cases who recover or die quickly. Thus, any associations found between an exposure and a disease may be more applicable to survivorship with disease rather than development of disease. Another use of prevalence studies is simply to describe the prevalence of a disease or condition in the population. For such studies to be useful, it is important that the individuals studied be representative of the population to which the results are to be generalized. Patients seen in tertiary care centers or in the practice of any one physician are seldom representative of all persons in the community with a disease, many of whom may not have even sought medical care. Accordingly, generalizations from such select groups of patients should be avoided.
D. Hybrid Study Designs It is sometimes possible to design a case-control study within either a retrospective or prospective cohort study. Consider a traditional cohort study in which an investigator wishes to find out whether a positive test result from a certain expensive serologic test is associated with an increased risk of hip fracture. In such a traditional cohort study, the investigator might start with blood samples drawn from 10,000 people free of hip fracture. The cohort might then be followed for 10 years to determine the incidence rate of hip fracture in those positive and in those negative on the serologic test. A modification of this traditional cohort design, called a nested case-control study, is illustrated in Fig. 1. The blood samples from the 10,000 people could be frozen and stored. Suppose that after 10 years had elasped, 200 people had incurred a hip fracture and 9800 had not. The stored serum samples from the 200 cases and a sample of, say, 400 of the
541
CHAPTER 20 Introduction to Epidemiologic Methods
FIGURE 1
Hypothetical nested case-control study of association between a serologic marker and hip
fracture.
9800 without the disease could then be tested. This sampling of nondiseased people would greatly reduce the cost from what it would be if the sera from all 10,000 cohort members had to be tested, yet the serologic status before disease occurrence would be measured. The proportion testing positive among the cases could then be compared to the proportion testing positive among the controls, as in a usual case-control study. Controls are selected from unaffected cohort members who are still alive and under surveillance at the time the cases developed the disease. Typically, the controls are matched to cases according to age, sex, and time of entry into the cohort. The availability of a variety of banks of stored serum around the world and the current interest in serologic predictors of disease make nested case-control studies an attractive and economical approach, as long as the serologic marker of interest does not undergo degradation over time. A case-cohort study is another method of increasing efficiency compared to a traditional retrospective or prospective cohort study. Like the nested case-control study, all cases and a sample of controls are selected. However, controls are sampled from the entire cohort, not just those free of disease, and are not matched to the cases. Rather, other relevant variables are taken into account in the statistical analysis. A case-cohort design is particularly useful when the associations between a serologic marker or other variable and several diseases are of interest.
E. Experimental Studies In general, the strongest evidence that a given exposure is a cause of a disease is produced from experimental
studies. In experimental studies, the investigator randomly assigns study subjects either to be exposed or not exposed to an agent, and then follows them through time to see what proportion of the exposed and unexposed develop certain diseases. Thus, the possibility that people choosing to be exposed to a certain factor are systematically different from those who do not is eliminated. Randomized clinical trials, which are one type of experimental study, have provided convincing evidence that estrogen replacement therapy protects against loss of bone mass, at least as long as it is used [7,8]. However, the protective effect of estrogen replacement therapy against coronary heart disease that has been reported from many observational studies was not uniformly accepted as causal because the association had not been tested in randomized trials, even though randomized trials have shown favorable effects on high-density lipoprotein cholesterol and low-density lipoprotein cholesterol levels [32]. Some people believed that women who use estrogen replacement therapy would be at lower risk for coronary heart disease even if they did not use estrogen. Users of replacement estrogen tend to be more physically active, healthier, and younger than women who do not use estrogen. Such characteristics would be associated with less coronary heart disease regardless of estrogen use [32]. In fact, the Heart Estrogen/Progestin Study (HERS), a randomized trial in women with previously diagnosed coronary heart disease (32a), found an initial increase in risk after initiation of use of an estrogen/progestin compound, followed by an apparent protective effect with increasing length of use. A more detailed description of issues that arise in randomized trials is included in Chapter 64.
542
KELSEY AND SOWERS
IV. SOME USEFUL EPIDEMIOLOGIC CONCEPTS A. Confounding Confounding is an issue that is especially likely to arise in observational studies. A statistical association between an exposure or other characteristic and a disease does not necessarily mean that one is a cause of the other. In ruling out other explanations for an association, confounding variables need to be considered. For instance, an investigator wishing to determine whether coffee drinking increases the risk for osteoporosis would have to be concerned with whether any observed statistical association between coffee consumption and osteoporosis was actually attributable to the tendency of coffee drinkers to smoke more cigarettes, not to use estrogen, to be thinner, or to have some other characteristic that puts them at an elevated risk for osteoporosis. Such variables are considered potential confounding variables and need to be considered in virtually all observational epidemiologic studies. A confounding variable is defined as a variable (e.g., cigarette smoking) that (i) is causally related to the disease under study independently of the exposure or characteristic of primary interest (e.g., coffee drinking) and (ii) is associated with the exposure or characteristic of primary interest in the study population, but (iii) is not a consequence of this exposure. Confounding variables may be taken into account in the study design by matching on the confounding variables, such as when cases are matched to controls on age and sex in a case-control study. For instance, Sowers et al. [33] described different hormone levels in premenopausal women with low bone mass compared to controls without low bone mass matched to cases by age, weight, and parity. Alternatively, confounding variables may be taken into account in the statistical analysis by using multivariate statistical methods, such as the Mantel – Haenzel procedure, logistic regression, Cox proportional hazard model, Poisson regression, or multiple linear regression. In the study in Table 2 relating bone mass to fracture risk, age was subsequently controlled in the analysis by a modification of Poisson regression [21]. Both matching in the study design and controlling in the analysis are valid ways of adjusting for confounding variables, and in fact it is possible to match roughly on certain variables in the study design and then control more finely in the analysis. Matching is used most frequently in case-control studies. The main considerations in deciding whether to match in the study design or control in the analysis in a case-control study are whether a given variable really is likely to be a confounder, the cost of obtaining information on the confounding variable so that it can be matched on in the study design, and whether the confounder is strongly related to the
disease and the exposure. The reader is referred elsewhere [1] for a discussion of which method of controlling confounding is more efficient under which circumstances. The same procedures that are used to determine exposure to putative risk factors are used to ascertain exposure to confounding variables, including questionnaires, medical records, laboratory tests, physical assessment, and special procedures. Measurement of potential confounding variables is highly important, as otherwise they cannot be controlled adequately in the analysis.
B. Effect Modification Effect modification, sometimes referred to as statistical interaction, also needs to be considered when studies are designed, analyzed, and interpreted. It occurs when the magnitude of the association between one variable and another differs according to the level of a third variable. For instance, it has been hypothesized [34], with some supporting empirical evidence [35], that any beneficial effect of supplemental calcium will be seen mainly in persons with low dietary calcium intake. Thus, the effect of supplemental calcium is modified by dietary calcium intake. It has been noted in at least one study [24] that use of estrogen replacement therapy in the immediate postmenopausal period protects against low bone mineral density only in women younger than about age 75 years. In other words, the effect of estrogen is modified by a person’s age (or years since menopause). Forsen et al. [29] found that the effect of cigarette smoking on hip fracture risk was greater in thin women than in heavy women. Detecting effect modification is an important component of the analysis of epidemiologic data.
V. SOME FREQUENTLY USED STATISTICS A. Relative Risk In cohort studies, the strength of the association between a putative risk factor and a disease is often measured by what is called a relative risk (or, more technically, a rate ratio or risk ratio; a discussion of the difference between a rate ratio and a risk ratio is beyond the scope of this chapter). A relative risk is simply the risk (or incidence rate) of disease in one group divided by the risk (or incidence rate) of disease in another group. For instance, a relative risk of 0.111/0.045 2.47 for fracture among women with bone mass of less than 0.60 g/cm compared to women with bone mass of 0.60 – 0.69 g/cm in Table 2 indicates that the former group has about two and one-half times the risk of
543
CHAPTER 20 Introduction to Epidemiologic Methods
fractures as women in the latter group. A relative risk of 0.68 for hip fracture among persons using thiazide diuretics compared to those not using them [36] indicates that the risk of hip fracture is reduced by almost one-third among users of these drugs. Relative risks give a meaningful idea of the extent to which an exposure elevates or decreases risk for disease and are important in assessing whether a causal relationship between an exposure and a disease exists.
B. Odds Ratio In case-control studies, risks and incidence rates are generally not available, so relative risks cannot be calculated. Instead, the odds ratio (ratio of exposed to nonexposed among cases divided by the ratio of exposed to nonexposed among controls) is calculated. It can be shown [37] that for all but the most common diseases ( 10% of the population affected, for instance), the odds ratio is a good approximation to the relative risk and can be interpreted similarly. In Table 1, an odds ratio of (26 127)/(39 131) 0.65 can be calculated, indicating that women who have ever used estrogen have about a 35% reduction in their risk of experiencing a hip fracture compared to women who have never used estrogen. In this same case-control study [18], an odds ratio of 1.9 was calculated for the association between lower extremity dysfunction and hip fracture, meaning that the risk of hip fracture is almost twice as great among those with lower extremity dysfunction as among those without.
C. Confidence Interval A confidence interval is often presented along with the estimate of the relative risk or odds ratio (or other parameter) in order to give a range of plausible values for the parameter being estimated. Confidence intervals provide more information than can be obtained simply by testing for statistical significance. A 95% confidence interval of 1.46 – 2.75 around a point estimate of relative risk of 2.00, for instance, gives the likely range of values for the true relative risk and indicates that a relative risk of less than 1.46 or greater than 2.75 can be ruled out with 95% confidence. The 95% confidence interval around the odds ratio of 1.9 for the association between lower extremity dysfunction and hip fracture mentioned earlier was 0.9 – 3.8 [18]. That 1.0 is included in the interval indicates that this association is not statistically significant at the P 0.05 level, although an odds ratio of greater than 1.0 is still quite likely and is certainly biologically plausible.
D. Statistically Adjusted Relative Risk or Odds Ratio When interpreting relative risks or odds ratios, the effects of confounding variables need to be taken into account. Statistical procedures for making adjustments for confounding variables are available and are described in textbooks of biostatistics [38] and epidemiology [1]. Briefly, a commonly used procedure for making statistical adjustments for confounding variables in case-control studies and cross-sectional studies when a disease is either present or absent (e.g., fracture vs no fracture) and when there are small numbers of confounding variables is the Mantel – Haenszel procedure. For instance, data in Table 1 might be subdivided by age group (which may be considered a potential confounding variable) and then summarized by the Mantel – Haenszel procedure to obtain an odds ratio adjusted for age. Logistic regression is used frequently when a disease is either present or absent and when there are several potential confounding variables or when a potential confounding variable is continuously distributed (e.g., weight). For instance, if an investigator wanted to examine the relationship between estrogen and hip fracture adjusting for age (measured as actual years of age or in broad age groups), weight, and several other variables simultaneously, he/she would use logistic regression to obtain an estimate of the odds ratio for the association between estrogen use and hip fracture, adjusted for any differences in the distributions of age, weight, and other variables between users and nonusers of estrogen. If cases and controls have been matched in the study design, then statistical methods that take the matching into account need to be employed. Procedures for making adjustments in cohort studies are based on similar principles, but must take into account the varying periods of time that different cohort members usually are followed and under observation. A form of the Mantel – Haenszel procedure that takes into account the length of follow-up of each cohort member may be used for small numbers of categorized potential confounding variables. Methods of analysis called Poisson regression and the Cox proportional hazard model can be used to calculate rate ratios adjusted for multiple potential confounding variables, taking into account the length of time that each cohort member has been followed. When the outcome of interest is a continuously distributed variable, such as bone mineral density or changes in bone mineral density over time, multiple regression can be used to estimate the relationship between an independent variable (e.g., age) and a dependent variable (e.g., bone mineral density), adjusting for potential confounding variables. If effect modification is present, statistics should be presented separately for groups in which the effects differ from each other.
544
VI. CRITERIA FOR DECIDING WHETHER AN ASSOCIATION IS CAUSAL Because most epidemiologic studies are observational rather than experimental, participants will usually differ in respect to other characteristics than just the exposure and disease of primary interest to the investigator. Sometimes confounding variables can be recognized, measured, and accounted for, but often they are unknown or only vaguely hypothesized. Also, any one study may have certain methodological deficiencies or may produce certain results by chance. Therefore, seldom will a single epidemiologic study provide definitive evidence for or against a hypothesis. Even results from several studies may not be convincing. With these considerations in mind, epidemiologists have developed criteria to be used as tests of whether a causal association exists. Not all criteria need to be fulfilled in all instances, nor are all equally important, but taken together they provide useful guidelines for determining whether an association between a given exposure and disease is causal. 1. Strength of association. For a positive association, the measure of association (relative risk or odds ratio) should be elevated, indicating that the exposed are at increased risk of disease over the unexposed or that those with disease are more likely to have histories of exposure than those without the disease. The greater the magnitude of these measures, the more likely the association is to be causal. As a rough rule of thumb, a relative risk or odds ratio of 2 indicates a moderate elevation in risk and a relative risk or odds ratio of 3 or more is considered strong. If no association between exposure and disease exists, the issue of causality does not arise, so establishing an association is an essential first step; the stronger the association, the more convincing is this aspect of the argument. Women with stroke, for instance, are 4 – 5 times more likely to fracture their hip than women not having had a stroke, making it more likely that this association is causal than if the relative risk were 1.5. Women in the highest quintile of body mass index (weight/height2) have only one-fifth the risk of experiencing a hip fracture compared to those in the lowest quintile [18]. 2. Statistical significance. A finding of statistical significance means that the result is unlikely to be a consequence of chance. Statistical significance depends on both the strength of the association and the number of people included in a study. If the sample size is inadequate, even relatively strong associations may not demonstrate statistical significance. Conversely, a tiny, biologically meaningless elevated risk can become “significant” with a very large sample size. For instance, among the 9704 women included in the baseline survey of the Study of Osteoporotic Fractures, lifetime caffeine consumption was inversely associated with bone mass, such that the
KELSEY AND SOWERS
equivalent of 10 cups of coffee per day over a period of 30 years was associated with a 1.1% decrease in radial bone mineral density [28]. Although statistically significant, the clinical significance of this finding is probably slight. 3. Ruling out alternative explanations. Once a significant association has been established (i.e., the exposure and disease are related, and the relationship is unlikely to be attributable to chance), other explanations for the observed association, such as methodologic deficiencies and confounding, should be carefully considered and ruled out. As mentioned earlier, if an association is found between coffee drinking and osteoporosis, it must be determined whether the association exists only because people who drink coffee also tend to smoke, be thin, not use estrogen, or have some other characteristic that influences their risk for osteoporosis. 4. Dose – response relationship. If increasing dose or length of exposure is associated with increasing risk, then the case for causality is considerably enhanced, as the likelihood is reduced that such a pattern could arise by chance or be attributable to confounding. Increasing length of use of hormone replacement therapy, for instance, is associated with a decreasing risk of hip fracture. Increasing weight is associated with a decreased risk [18]. The absence of a dose – response relationship does not disprove causality, however, as other patterns of association, such as a threshold effect, could also occur. 5. Removal of exposure. If the presence of an exposure increases risk of disease and removing the exposure reduces risk, belief in a causal association is strengthened. When estrogen replacement therapy is stopped, loss of bone mass resumes [8], thus strengthening the belief that the association is indeed causal. 6. Time order. As mentioned earlier in the discussion of the association between low back muscle strength and spinal osteoporosis, sometimes it is not clear whether an exposure caused a disease or the disease caused the exposure. This problem is particularly notable in crosssectional studies, where prevalent disease and exposure are determined simultaneously. Time order is unique among the causal criteria in that if disease can be shown to precede exposure, causality is definitely ruled out. 7. Predictive power. Hypotheses regarding presumed causal associations that can in turn be shown to predict future occurrences lend strong support to the belief in the causality. 8. Consistency. If associations of similar magnitude are found in different populations by different study methods, the likelihood of causality is increased substantially, as all studies are unlikely to have the same methodologic limitations or study population idiosyncrasies. Virtually all studies, for instance, show that estrogen replacement therapy protects against loss of bone mass, at least in the early postmenopausal period.
545
CHAPTER 20 Introduction to Epidemiologic Methods
9. Coherence with experimental data. When available, the results of well-designed experiments in which exposures are assigned at random are very convincing because the only factor on which groups differ, except by chance, is the exposure of interest. Randomized trials [7,8], for instance, have clearly shown that women in their early postmenopausal years who are randomly assigned to use estrogen replacement therapy have significantly less loss of bone mass than women randomized to placebo. However, many exposures cannot be ethically or practically assigned at random. In addition, some well-controlled experiments on a few carefully selected people may have little relevance to the general, free-living population. 10. Biologic plausibility. When a new finding fits well with the currently known biology of a disease, it is more plausible than if a whole new theory must be developed to explain the new finding. Protection against hip fracture from obesity, for instance, is biologically plausible because of the increased estrogen production in obese women [39] and the protection against hip fracture afforded during a fall from fatty tissue around the hips [40]. Another possible means of enhancing biologic plausibility is through laboratory experiments. However, what occurs in a laboratory setting or in experimental animals may have limited relevance to humans. It should be obvious from the discussion just given that decisions on the likelihood of causality are of necessity partly judgmental. What one person may believe is a causal association, another person may not. Lilienfeld [41] divided the degree of evidence for causation into three levels. At the first level, the evidence is considered sufficient for further study. For instance, studies suggesting a protective effect of vitamin C against osteoporosis [11] should be followed up with further study. At the second level, the evidence is considered sufficient to warrant public health action, even if the causal association has not been definitively established. Many people believe that the possible protection against osteoporosis and associated fractures from calcium supplementation fits into this category [35]. At the third level, the evidence is so strong that the causal association is considered part of the body of scientific knowledge. There is general agreement that the protective effect of estrogen replacement therapy against loss of bone mass in early postmenopausal women is established with this degree of certainty [42].
VII. SAMPLE SIZE CONSIDERATIONS There is little point in undertaking a study to determine whether an exposure is associated with a disease unless the number of study subjects is large enough that the association is in fact likely to be detected. Similarly, under most circumstances it would be wasteful to go to the expense of
including far more study subjects than are actually needed. Thus, determining the optimal sample size is an important component of planning a study. Many statistical and epidemiologic textbooks provide formulae for sample size estimation [1,37,43]. The sample size required depends on several conditions, all of which enter into the equations given later. First, what risk is one willing to take that the null hypothesis (of no difference) is rejected when it is in fact true? This is the value, usually taken to be 0.05, meaning that the investigator is willing to reject a null hypothesis incorrectly 1 out of 20 times. Smaller values of will require larger sample sizes. Second, what risk is the investigator willing to take that the null hypothesis is not rejected when it should be? This is the value, usually taken to be 0.10 or 0.20, meaning that the power to reject the null hypothesis of no association is 0.90 or 0.80, respectively. The greater the power (and thus the smaller the value of ), the larger the sample size that is needed. Third, how large a difference does one want to be able to detect? The smaller the difference, the larger the required sample size. Fourth, if the outcome of interest is a yes/no variable, what proportion of the population develops the disease in a cohort study or is exposed in a case-control study? Outcomes that affect roughly half the population will require smaller sample sizes than outcomes that are either very rare or very common. Thus, studying rare diseases with a cohort design or rare exposures with a case-control design requires enormous sample sizes. Fifth, what is the variance of what is being measured in the population? The greater the variance, the larger the sample size that will be needed. Two formulae that may be used in estimating required sample size are given here when the objective is to detect difference between two groups. To detect differences between means, the appropriate formula is n
(Z /2 Z )2 2 (d*) 2 r
To detect differences between proportions, the appropriate formula is n
(Z /2 Z )2 p(1 p) (r 1) (d*) 2 r
where d* is the value of the difference in means or proportions that one wishes to be able to detect; n is the number of exposed individuals in a cohort (or cross-sectional) study or the number of cases in a casecontrol study; r is the ratio of the number of unexposed individuals to the number of exposed individuals in a cohort (or crosssectional) study or the ratio of the number of controls to the number of cases in a case-control study; r 1 if the numbers in the two groups being compared are equal;
546
KELSEY AND SOWERS
is the standard deviation in the population for a continuously distributed variable; p1 is the proportion of exposed individuals who develop (or have) the disease in a cohort (or cross-sectional) study or the proportion of cases who are exposed in a case-control study; p2 is the proportion of unexposed individuals who develop (or have) the disease in a cohort (or cross-sectional) study or the proportion of controls who are exposed in a case-control study; p rp2 p is the weighted average of p1 and p2 , or p 1 1 r Z is called a standard normal deviate because it rescales distributions of measurements to have a mean of 0 and a standard deviation of 1, thus enabling one to use standard tables of the normal distribution. Most textbooks of basic statistics have tables that enable one to determine the level of significance (i.e., p value) that corresponds to a given value of Z. The symbol Z/2 refers to the standard normal deviate for a two-tailed test of statistical significance. A two-tailed test provides for the possibility that a difference between two groups might be either positive or negative. For 0.05, Z/2 1.96. Z is related to the probability that one fails to reject a null hypothesis that should in fact be rejected. For 0.20, Z 0.84; for 0.10, Z 1.28; for 0.05, Z 1.64. As an example, suppose that an investigator wants to undertake a case-control study to determine whether there is an association between alcohol use (categorized as yes versus no) and distal forearm fracture. Suppose the investigator knows that about 20% (i.e., p2 0.20) of the population drinks alcoholic beverages and that it is desired to detect a 10% difference between cases and controls (i.e., d* 0.10, p1 0.20 0.10 0.30). Further, suppose that the investigator wants to be 90% certain of detecting a difference of this magnitude (i.e., 1 0.90 0.10) and wants to find a difference when there really is none only 5% of the time (i.e., 0.05). Suppose that the desired ratio of controls to cases is 2:1 (i.e., r 2). Note that 0.30 2(0.20) p 0.23 1 2 Then, substituting into the second formula just given, n
(1.96 1.28)2 0.23 (1 0.23) (3) 279 cases (0.10)2 2
Number of controls 2 279 558. In trying to keep the sample size as small as possible, it is important that measurements be as precise as possible. With poor measurement, not only will a larger sample size be needed, but the estimated magnitude of an association will be a poor approximation to the true association. These issues are discussed next.
VIII. MEASUREMENT ERROR A. Nature of the Problem and Definitions A certain amount of measurement error is almost inevitable, whether in measurement of potential risk factors, disease status, or potential confounding variables. This discussion will be referring both to the validity or accuracy of a measurement, or the closeness with which the measurement approaches the true value, and to the reliability or reproducibility of a measurement, or the extent to which the same measurement is obtained on the same occasion by the same observer, on multiple occasions by the same observer, or by different observers on the same occasion. It is well known by investigators in the field of osteoporosis that it is difficult to measure some of the major putative environmental risk factors, such as diet, physical activity, coffee consumption, and alcohol consumption. Because much of this information is obtained from questionnaires, the quality of data obtained is often no better than the imperfect memory of individuals about such factors as their dietary habits or physical activity. Also, questionnaires would become much too tedious if too much detail were required of study subjects. Thus, in obtaining summary indicators of such variables as calcium consumption or physical activity, accuracy of measurement will, to some extent, be compromised. Data obtained from questionnaires present particular problems in case-control studies in which information is sought about exposures that took place decades before the disease manifests itself. If physical activity or diet during adolescence affects risk for fractures in older individuals, obtaining accurate information in a case-control study would in all likelihood be impossible. Even if information on an exposure is obtained from biologic assays, there is often no assurance that a single measurement is indicative of the cumulative exposure or the exposure at the time the disease was developing. Errors in classification of disease status are probably as small for fractures (except vertebral fractures) as for any other disease. However, if measures of bone mass are of interest, several measurement issues must be considered. First, can bone mass be measured accurately considering that marrow fat may distort the true measure of mineral? Second, the measure of bone density is areal, not volumetric. Use of conventional units (g/cm2) may cause selective misrepresentation of density. Third, risk factors such as body size influence the precision of the bone mass measure. Finally, while measures of bone density have been used as an indicator of fracture risk, this measure does not include all dimensions of bone structure that affect fracture risk, such as bone structure and microarchitecture. Errors in the measurement of confounding variables are also of concern. Often the critical confounding variables are just as difficult to measure as the exposure of primary interest.
547
CHAPTER 20 Introduction to Epidemiologic Methods
TABLE 3
Definitions of Sensitivity and Specificity True classification Present
Absent
TABLE 4 Hypothetical Example of Effect of Differential Misclassification of Exposure in Case-Control Study Relating Hip Fracture to Medication Use in Past 24 h Hip fracture cases
Imperfect classification Present
a
b
Absent
c
d
a c
b d
Total
Note. Sensitivity a/(a c); false negative rate c/(a c); specificity d/(b d); false positive rate b/(b d).
Controls
True status a Medication use
20
80
80
Total
100
100
20
Observed status with recall bias b Medication use
Some potential confounding variables, such as socioeconomic status, are difficult to conceptualize let alone measure. As will be discussed later, inadequate measurement of important confounding variables can lead to biases just as serious as those arising from errors in the measurement of exposure or disease. For ease of presentation, this discussion focuses mainly on measurement of binary variables (i.e., variables that take on only one of two values, such as disease present or absent or exposure present or absent). Toward the end of this section, we will briefly extend the discussion to quantitative variables such as bone mineral density. Sensitivity is defined as the proportion of those who truly have the characteristic that are correctly classified as having it by the measurement technique. Specificity is the proportion of those who truly do not have the characteristic that are correctly classified as not having it by the measurement technique. Table 3 shows how sensitivity and specificity may be calculated from a 2 2 table. The proportion false positive is 1 specificity and the proportion false negative is 1 sensitivity. Measurement of a binary characteristic is perfect only when sensitivity and specificity are both 100%. Unfortunately, sensitivity and specificity close to perfect are seldom achieved in practice. When sensitivity equals 1.00 minus specificity, the measurement method is no better than entirely random classification of study subjects. Measurement error is said to be differential if the magnitude of the error for one variable differs according to the actual value of another variable. Table 4 shows a hypothetical example of differential misclassification. Suppose a case-control study of hip fracture is being undertaken, and the exposure of interest is whether a certain medication was taken in the 24 h before the fracture occurred. Cases might want to blame the fracture on some external agent such as a medication and therefore might report use in the last 24 h when such use did not occur. Controls, however, might forget that they even had taken the medication in the past 24 h and therefore might underreport its use. Thus, an association between the medication and hip fracture such as that
25
75
85
Total
100
100
a b
15
True odds ratio (20 80)/(20 80) 1.00. Observed odds ratio (25 85)/(15 75) 1.89.
shown in the lower table might be observed when in fact no such association exists. Measurement error is said to be nondifferential when the magnitude of error for one variable does not vary according to the actual value of the other variable of interest. In other words, both sensitivity and specificity remain constant irrespective of the value of the other variables.
B. Effects of Nondifferential Misclassification of Discrete Variables Nondifferential misclassification in a 2 2 table always causes the measure of association (i.e., relative risk or odds ratio) to become closer to the null value. Table 5 shows what happens in a hypothetical case-control study in which an exposure is measured with sensitivity of 0.60 and specificity of 0.50, and misclassification is nondifferential. The true odds ratio of 2.22 would be observed to be only 1.06. With some variables, such as diet and physical activity, for which measurement error is undoubtedly substantial, it is difficult to know whether results showing no association between these variables and fractures occur because there really is no association or because measurement is poor. The extent of attenuation of measures of association depends in part on how common the exposure is. Table 6 shows a hypothetical case-control study in which the true odds ratio is 2.11, the exposure is fairly rare, the disease is measured without error, and the sensitivity of exposure measurement is 100% but the specificity is only 60%. (That is, 40% of those without the exposure are classified incorrectly as having the exposure.) It may be seen that the
548
KELSEY AND SOWERS
TABLE 5 Hypothetical Example of Nondifferential Misclassification of Exposure in Case-Control Study Cases
Controls
Total
True status in case-control study a Exposure
85
50
135
115
150
265
Total
200
200
400
Observed status in case-control study b Exposure
51 57 108
30 75 105
34 58 92
20 75 95
187
200
200
400
Total
213
True odds ratio (85 150)/(50 115) 2.22. Assume misclassification in measurement of exposure, but not disease, and sensitivity 0.60 and specificity 0.50. Observed odds ratio (108 95)/(105 92) 1.06. a b
observed odds ratio is 1.13 instead of 2.11. However, if the sensitivity is 60% (i.e., 40% of those with the exposure are classified incorrectly as not having the exposure) and the specificity 100%, then the odds ratio is reduced only to 2.06. Thus, for an uncommon exposure, it is important to have a
TABLE 6 Hypothetical Example of Effects of Different Values of Specificity and Sensitivity of Exposure Measurement on Attenuation of Odds Ratio in Case-Control Study with Uncommon Exposurea True status b Cases
Controls
Exposure
10
Total
5
90
95
100
100
Observed status c
Observed status d
Cases
Controls
Cases
46
43
6
Controls
Exposure
Total a
3
54
57
94
97
100
100
100
100
Assume misclassification in measurement of exposure, but not disease. True odds ratio (10 95)/(5 90) 2.11. c If specificity 0.60 and sensitivity 1.00, observed odds ratio 1.13. d If specificity 1.00 and sensitivity 0.60, observed odds ratio 2.06. b
highly specific measure in order to obtain a good estimate of the odds ratio. With a highly prevalent exposure, the situation is different. Here, with high sensitivity but only fair specificity, the odds ratio is attenuated only slightly, whereas with high specificity but low sensitivity, the odds ratio is reduced almost to 1.00. In this situation, a highly sensitive measure is desirable. When measurement error occurs when controlling for potential confounding variables, additional problems occur. If a confounding variable is measured imperfectly, then controlling for the confounding variable in the analysis will not entirely remove its effect because it was not measured with sufficient accuracy. Accordingly, if an investigator controls for the effect of physical activity when considering the possible effect of alcohol consumption on risk for hip fracture, the relative risk could change from 2.0 without adjustment for physical activity to 1.5 with adjustment. The investigator would not know whether there still is an independent effect of alcohol consumption or whether physical activity had been measured accurately, there would be no residual association between alcohol consumption and hip fracture. Furthermore, if the confounding variable is measured perfectly but the exposure variable is not (or vice versa), then the effect of the measurement error is typically to induce apparent effect modification when none exists [44]. Table 7, for instance, presents a hypothetical case-control study of the association between previous use of replacement estrogen and hip fracture. Suppose that the true situation (data on left) is that in both younger and older women, use of estrogen is associated with an odds ratio of 0.47. However, note that fewer women in the older strata have used estrogen. Suppose that sensitivity is 80% but that specificity is only 50%. (In other words, 50% of women who have not used estrogen say that they have used it.) Because the effect of misclassification on the odds ratio will be greater when prevalence of exposure is lower, the odds ratio becomes much closer to 1.0 in the stratum of older women than in the stratum of younger women (data on right). Thus, because of the poor specificity, it appears that there is effect modification by age when none exists. If measurement of exposure were perfect but measurement of the confounder less than ideal, then similar apparent effect modification could be induced. When both the exposure and the confounder are subject to nondifferential measurement error, the effects are less predictable. The adjusted estimate of the odds ratio may be even more biased than the unadjusted estimate that ignores confounding entirely [44]. Considering that epidemiologic studies are frequently trying to measure and control for several variables with substantial measurement error (e.g., food consumption, caffeine consumption, physical activity, alcohol consumption, cigarette smoking), it is no wonder that results from different studies are inconsistent, as measurement techniques and
549
CHAPTER 20 Introduction to Epidemiologic Methods
TABLE 7 Hypothetical Example in Which Confounding Variable (Age) and Disease (Hip Fracture) Are Measured without Error and Exposure (Estrogen Use) Is Imperfectly Measured (but Nondifferentially) Observed statusc
True status Hip fracture cases
Controls
Ages 50 – 74a
Hip fracture cases
Controls
Ages 50 – 74 d
Estrogen
320
500
Estrogen
596
Estrogen
680
500
Estrogen
404
350
1000
1000
1000
1000 515
Total Age 75b
Total
650
Age 75 e
Estrogen
24
50
Estrogen
507
Estrogen
976
950
Estrogen
493
485
1000
1000
1000
1000
Total
Total
True odds ratio 0.47. True odds ratio 0.47. c Sensitivity for measurement of estrogen use 0.80. Specificity for measurement of estrogen use 0.50. d Observed odds ratio 0.79. e Observed odds ratio 0.97. a b
the prevalence of exposures and confounders differ from study to study. It should also be mentioned that in analyzing data from tables other than simple 2 2 tables, there are circumstances under which nondifferential measurement error can make an association appear larger than it really is. The reader is referred to an article by Weinberg et al. [45] for a discussion of such situations.
C. Error Correction Methods for Discrete Variables Sometimes the accuracy of a measurement is known from previous studies or can be determined in a small substudy undertaken as part of an ongoing study, but it is impractical or too costly to use the accurate measurement on all study subjects. In such situations, error correction methods may be employed to correct for the effects of measurement error on the magnitude of the observed association. That is, the known values of specificity and sensitivity can be used to estimate the true proportion exposed from the observed proportion. If p denotes the observed proportion exposed, then the following formula may be used to estimate P, the true proportion exposed: p specificity 1 . P sensitivity specificity 1 Other methods of improving estimates even if sensitivity and specificity are not known are described elsewhere [1]. Also discussed [1] is the use of multiple imperfect measurements
of a given variable to improve accuracy rather than relying on a single imperfect measurement.
D. Quantification of the Reproducibility of Discrete Variables For quantifying the reproducibility of a discrete variable, the kappa statistic is used most frequently. Consider data in Table 8 [46]. Suppose it is known from medical records that 39 of 217 people were prescribed a certain medication. When administered a questionnaire, 14 of the 39 people who were prescribed a medication say that they were prescribed it, while 171 of 178 people who were not prescribed a medication say that they were not. The agreement between the two methods of ascertaining information (85%) is actually quite good. However, relatively few people in fact were prescribed the medication, so that even if all the study subjects said they had not had the medication prescribed, regardless of whether or not they had, agreement would still be high. Thus, a statistic is needed that takes into account the agreement that would be expected by chance. The kappa statistic, which takes into account chance agreement, is defined as observed agreement expected agreement . 1 expected agreement When two measurements agree only at the chance level, the value of kappa is zero. When the two measurements agree perfectly, the value of kappa is 1.0. In Table 8, the value of kappa is 0.39, indicating that the observed agreement is
550
KELSEY AND SOWERS
TABLE 8 Example of Calculation of kappa: Agreement between Personal Interview and Medical Chart for Use of Reserpine among Controls in a Case-Control Study in Two Retirement Communities History of use of reserpine according to medical chart Yes
No
Total
History of use of reserpine according to patient’s report Yes
14
7
21
No
25
171
196
Total
39
178
217
Note. Source: Paganini-Hill and Ross [46]. Chance-expected agreement [(21)(39) (196)(178)]/(217)2 0.7583. Observed agreement (14 171)/217 0.8525. kappa (0.8525 0.7583)/(1 0.7583) 0.39.
only 39% of the way between chance agreement and perfect agreement. O’Neill et al. [47] assessed the reproducibility of answers obtained by the questionnaire used in the European Vertebral Osteoporosis Study of persons ages 50 – 85 years by having a different interviewer readminister the same questionnaire within a 28-day period at four of the study sites. The kappa coefficient was 1.00 at all four study sites for the variable of ever having been pregnant, but ranged between only 0.25 and 0.63 at the four sites for activity level when ages 15 – 25 and from 0.17 to 0.62 for milk consumed at ages 15 – 25. It should be noted that for measurements of conditions that are uncommon, the value of kappa will be lower than for common conditions, even though the values of specificity and sensitivity remain the same. This property needs to be taken into account when interpreting values of kappa.
E. Continuously Distributed Variables For continuously distributed variables such as bone mineral density, measurement error is again a concern. A number of indices can be calculated that reflect the accuracy of the measure of interest [1]. One frequently used measure to reflect the accuracy or lack of accuracy is the standardized bias, which is defined as mean of measurements true mean . standard deviation of the true values Thus, for example, if a given technique for measuring bone mineral density systematically overestimates the true value by 0.02 g/cm2 and the standard deviation for true bone mineral density in the population is 0.10 g/cm2, then the technique has a standardized bias of 0.02/0.10. That is, the
imperfect technique tends to give values that are 0.20 of a standard deviation higher than the true values. A measure of the extent to which imperfectly measured values tend to fall in the same position relative to their mean as do the corresponding true values relative to the true mean is the correlation coefficient, which can range from 1.0 to 1.0. The correlation coefficient of reproducibility (which also can range from 1.0 to 1.0) is often used to assess the extent to which two imperfect sets of measurements agree. This coefficient indexes the extent to which the measurement tends to fall in the same position relative to the mean for the first set as it does relative to the mean for the second set. The square of this correlation coefficient indicates that proportion of the variance in one set of measurements that is captured by the other set of measurements. A correlation coefficient of 0.60 between two types of measures of bone mass, for instance, would indicate that (0.60)2 0.36 of the variance in one type of measure was captured by the other type of measure. It is important to note that both the correlation coefficient of reproducibility and kappa may give misleading indications of the extent of reproducibility if the errors in measurement are not independent of each other. For instance, information recorded in medical records may have been obtained from the patient herself so that data subsequently elicited from the patient by an interviewer may not be independent of what is found in the medical record. The coefficient of variation, or the standard deviation divided by the mean, is sometimes used as an indicator of the precision of a measure and is best interpreted taking into account the numeric values of the mean and standard deviation. It is particularly useful for assessing the relative amount of variation in situations in which as the mean increases, so does the standard deviation or when investigators want to compare their precision to that reported by others. For instance, Schott et al. [48] reported a coefficient of variation of 0.84% for the broadband ultrasonic attenuation of the calcaneous and 0.15% for speed of sound by measuring a phantom daily for 45 days using a Lunar Achilles ultrasonic instrument. When 20 volunteers were measured three times each, these coefficients of variation were 0.93 and 0.15%, respectively. The authors reported that these results were comparable to other reports in the literature. As is the case for binary variables, nondifferential error in measurement generally results in attenuation of associations between continuously distributed variables, such as the association between dietary calcium intake and bone mineral density. If the accuracy or reproducibility of a continuously distributed variable is available from previous studies, then methods are available to correct for this attenuation. Approaches to correction in correlation analysis with one or two variables measured with known error are described by Liu et al. [49] and Rosner and Willett [50]. Liu et al. [49] also presented an approach for correction when regression analysis is used.
CHAPTER 20 Introduction to Epidemiologic Methods
Multiple measurements can be used to increase the accuracy of certain continuously distributed variables, such as levels of hormones or bone turnover markers. For example, in a longitudinal study, Sowers et al. [51] observed that the between-person variability for bone-specific alkaline phosphatase was two times greater than the within-person variation. In contrast, when measuring bone mineral density with dual energy X-ray absorptiometry (DXA), the between-person variability was 11 to 29 times greater than the within-person variability. To reduce the relatively high within-person variability for the bone-specific alkaline phosphatase, the investigators sampled individuals at five different times and used the mean of the five samples as their measure of bone-specific alkaline phosphatase.
IX. MEASURING DIET AND BONE TURNOVER STATUS AS EXAMPLES OF MEASUREMENT ISSUES This section addresses in greater detail issues related to the measurement of dietary status and of bone turnover status. The measurement of these characteristics illustrates the pitfalls and problems in epidemiologic studies in which the data sources are interview and biologic specimens, respectively.
A. Dietary Intake The study of calcium intake and other dietary characteristics has long been a focus of those interested in risk factors for osteoporosis. However, the results of studies concerned with dietary intake of calcium have been inconsistent. These discrepant results may have their origin in either (i) the nature of and relatively weak role of dietary intake in the development of osteoporosis or (ii) the manner in which data were collected. Four general categories of assessment methods have been used to obtain information about diet in epidemiologic studies [52]. In a dietary record the individual records the amounts of each food and beverage consumed over a specified number of days (usually 4 days). The respondent must be able to record food weights, volumetric measures, and the content of mixed dishes. Accordingly, the participants in a study using dietary records tend to be highly select individuals who have the literacy, endurance, and motivation to participate in a demanding assessment. This selectivity ultimately limits the generalizability of findings. In addition, during the period that food records are being used, respondents may alter their behavior to ease the burden of completing the assessment or to avoid the perceived positive or negative social value placed on certain eating patterns.
551 In 24-h recalls, the respondent is asked to describe all food and beverages consumed in the previous 24 h, typically with a standard set of question probes and food models for assistance in identifying serving sizes. Because most individuals’ diets vary greatly from day to day, it is not appropriate to use data from a single 24-h recall to characterize an individual’s usual intake. The principal use of 24-h recalls should be to describe the average dietary intake of a group. Measurements on at least 50 people should be made to categorize a group. The food frequency method asks the respondents to report their usual frequency of foods consumed from a list of foods (ranging in length from about 10 to more than 100) for a specific period of time, most often in the past year. Responses to food frequency questionnaires can be used to rank individuals within a population according to their consumption of foods specified on the frequency list. These responses can also be linked by computer to food composition data bases; the one generally used in the United States is that of the U.S. Department of Agriculture [53]. A food composition data base provides information on the nutrient content of food groupings. Because food groupings used in food frequency questionnaires are generally quite broad (e.g., fruit juices), an average nutrient content is estimated for each grouping. The averages are obtained by weighting the contributions of the different constituents based on consumption patterns in the population. Thus, in the United States, the nutrient contribution of orange juice might be, say, 50% of the nutrient estimate for fruit juices, whereas apple and grape juice would contribute smaller proportions. The exact proportions to be attributed to different foods should be determined by the investigator based on the food groupings used in his or her particular questionnaire and the consumption patterns in the population being studied. Whether it is appropriate to use the food frequency methodology to quantify nutrient intake is controversial. Although many investigators use quantitative estimates derived from food frequency questionnaires as approximations of the usual intake, food frequency questionnaires have been observed to perform inconsistently in different populations, and the same questionnaire performs differently when administered under different conditions to the same population. Table 9 shows considerable differences in estimated calcium intake when the same questionnaire was administered by an interviewer (row 1) or filled out by the respondent (rows 2 and 3). When a food frequency questionnaire is administered by an interviewer, typically more than 90% of the frequencies are considered usable; however, when the food frequency questionnaire is self-administered, between 20 and 40% of frequencies may not be usable. In a food frequency questionnaire, the lack of details about diet makes it likely that quantification of nutrient intake will not be as accurate as dietary records or 24-h recalls. For instance, portion size may not be estimated
552
KELSEY AND SOWERS
TABLE 9 Mean (and Standard Deviation) of Dietary Calcium Intake (mg/day) Estimated from the National Cancer Institute Food Frequency Questionnaire According to Whether the Questionnaire Was Interviewer-Administered or Self-Administered, Women Categorized by Reproductive Status Mode of administration Interviewer-administered
Menstruating women
Women with hysterectomy
Women with oophorectomy 1161 (702)
1012 (577)
995 (559)
Self-administered (baseline)
714 (457)
697 (434)
853 (602)
Self-administered (year 1)
754 (440)
749 (441)
800 (434)
Note. Source: Sowers et al., unpublished data.
adequately and certain unusual foods eaten in large quantities by only a small proportion of the population may be difficult to take into account. Diet histories are typically a combination of a food frequency questionnaire, an interview about usual patterns of eating, and a 3-day diet record. Each method uses a data base with the nutrient content of food and either a reported or an imputed measure of serving size to estimate nutrient intake. While the diet history is noted for the volume of data gathered, it is used infrequently because of the cost of administration and the burden to the respondent. Because the diet history includes a 3-day food record, it is likely that only select groups of people will participate in assessment. Also, people may modify their behavior during the assessment period. Because each of the parts of the diet history provides a unique piece of information, data reduction and reconciliation of the findings of the 3-day food record and the food frequency present major problems. Establishing the role of dietary calcium intake in osteoporosis and fractures also requires various other pieces of information, including (i) knowledge of the intake of other nutrients, such as vitamin D as well as sunlight exposure; (ii) knowledge of intake of other nutrients that might be deleterious in high concentrations, such as fluoride or aluminum; (iii) information about nutrient intake and its adequacy in critical periods during growth, development, reproduction, and the menopausal transition; (d) an assessment of factors such as dietary sodium that may influence the excretion of calcium in the urine; and (e) information about factors that may facilitate or impede the absorption of dietary calcium, including intestinal disorders such as sprue. Additional problems are that the nutrient data banks do not have consistent documentation about important nutrients, such as vitamin D, fluoride, or aluminum content of foods. For example, the U.S. Department of Agriculture has only provisional tables for vitamin D. Other nutrients, such as fluoride, may be in part determined by the fluoride content of the water in which the food is prepared. The responsibility then rests with the individual investigator to
ascertain how much total fluoride is ingested. There are no validated instruments to assess retrospectively dietary calcium intake during critical periods such as adolescence and young adulthood. Furthermore, those assessments must take into account the change in the food supply over time. For instance, were school lunch programs available to provide a daily source of calcium during the Depression years?
B. Bone Turnover Markers Bone turnover markers may help define which component of bone remodeling is responsible for reduction in bone density or alteration in bone microarchitecture. Additionally, bone turnover markers can help define the responsiveness of bone tissue to preventive or therapeutic interventions. While these markers hold greater potential for understanding bone biology at a cellular level, they have limitations in epidemiologic studies. For instance, bone turnover markers have a greater variability than the measurement of bone mineral density [51]. Also, these turnover markers reflect the short-term events of bone changes required for homeostasis. As such, their measurement may reflect immediate events, but their extrapolation to events over a broader time frame is questionable. Multiple measurements of bone turnover markers are required to characterize the individual’s current status in regard to osteocalcin, bone-specific alkaline phosphatase, or N-telopeptides. The use of urine-based turnover markers is particularly useful in epidemiologic studies because it is easier to collect multiple urine samples than multiple blood samples. However, these urinary markers reflect not only the bone turnover processes, but also the integrity of the individual’s renal function and any changes in fluid status. Thus, in epidemiologic studies, problems and pitfalls occur with biologic data as well as with interview data. As with interview data, multiple sampling of biological material can improve the precision of measurement of an individual’s status. There is also a need to collect specimens in
553
CHAPTER 20 Introduction to Epidemiologic Methods
ways that are acceptable to a broad range of individuals so that studies are not based on a highly select group whose findings cannot be extrapolated to a larger population.
X. CONCLUSIONS Epidemiologic studies have added considerably to our understanding of the etiology, course, and consequences of osteoporosis. As will be described in subsequent chapters in this section, a variety of risk factors have been established and intriguing leads for further study have been suggested. However, for some potential risk factors, results of different studies have been rather inconsistent, such as those pertaining to some aspects of diet, physical activity, alcohol, certain medications, and reproductive variables. When trying to interpret these discrepant results, it is important to keep in mind common reasons for inconsistent results. Sometimes discrepant results may be explained by major flaws in study designs, such as failure to realize that the time sequence is unclear in cross-sectional studies, poor choice of controls in case-control studies, large numbers of people lost to follow-up in cohort studies, and, in any type of study, sample sizes that are too small and the use of specialized populations that are not representative of any larger population. However, for the most part, the variables for which results have tended to be inconsistent are those that are difficult to measure. It is important to keep in mind that if a characteristic cannot be measured well or for some reason is not measured well in an otherwise methodologically sound study, associations between that risk factor and a disease will be difficult to detect. If measurements of varying quality are used in different studies, then discrepant results may be expected. Different results may also be obtained when the prevalence of a potential risk factor varies from one study to another if there is some error in the measurement of that risk factor. Another reason for discrepant results is the extent to which confounding variables have been accurately measured and taken into account. Because many established and potential risk factors for osteoporosis, such as body build, diet, physical activity, alcohol consumption, caffeine consumption, and cigarette smoking, are correlated with each other, separating out the effect of any one of these variables is difficult. Errors in measuring these characteristics exacerbate this problem. Another reason for inconsistent results is that a risk factor may have a different effect in one subgroup of the population than in another. If one study includes mostly members of one subgroup and another study has mostly people from another subgroup, then results could be different in the two studies. Sometimes results from different studies are said to be discrepant when they really are not. One common reason
for this is varying sample size. If sample sizes differ from one study to another, then statistical significance is also likely to vary. If a result is statistically significant in one study and not in another, but the magnitudes of the association between the risk factor and the disease are similar in the studies, then this is evidence for consistency between the studies, not inconsistency. For instance, Cumming [35] found that although an apparently protective effect of supplemental calcium on loss of bone mass in adult women was statistically significant in some studies and not others, the magnitude of the slight protective effect in almost all studies was actually quite consistent. It is thus very important to consider magnitudes of associations and their confidence limits, not just statistical significance. Finally, if an effect is relatively small and there is even a modest amount of measurement error, it will be difficult for epidemiologic studies to detect the effect. The association between caffeine intake and osteoporosis may fall into this category. As knowledge of good epidemiologic methods and awareness of potential pitfalls in epidemiologic studies have become more widespread, the quality of studies has improved considerably. Also, results have tended to be interpreted more cautiously when potential problems in studies have been recognized. One key to continued improvement will be better methods of measurement of exposures, confounding variables, and, to some extent, outcome variables, whether by questionnaire, laboratory assay, densitometry, or other approach. Such improvements should help advance knowledge of the epidemiology of osteoporosis and associated fractures and should also enable epidemiologists to continue to provide further ideas for investigators in other disciplines, including endocrinology, biomechanics, and other areas.
References 1. J. L. Kelsey, A. S. Whittemore, A. S. Evans, and W. D. Thompson, “Observational Epidemiology,” 2nd Ed. Oxford Univ. Press, New York, 1996. 2. J. L. Kelsey and S. Parker, Epidemiology as an alternative to animal research. In “Non-Animal Techniques in Biomedical and Behavioral Research and Testing” (M. B. Kapis and S. C. Gad, eds.). Lewis Publishers, Boca Raton, FL, 1993. 3. C. Cooper, G. Campion, and L. J. Melton III, Hip fractures in the elderly: A world-wide projection. Osteopor. Int. 2, 285 – 289 (1992). 4. T. D. Spector, E. V. McCloskey, D. V. Doyle, and J. A. Kanis, Prevalence of vertebral fracture in women and the relationship with bone density and symptoms: The Chingford Study. J. Bone Miner. Res. 8, 817 – 822 (1993). 5. B. Ettinger, D. M. Black, M. E. Nevitt, A. C. Rundle, J. C. Cauley, S. R. Cummings, H. K. Genant, and the Study of Osteoporotic Fractures Research Group, Contribution of vertebral deformities to chronic back pain and disability. J. Bone Miner. Res. 7, 449 – 455 (1992). 6. F. Albright, E. Bloomberg, and P. H. Smith, Postmenopausal osteoporosis. Trans. Assoc. Am. Phys. 55, 298 – 305 (1940).
554 7. R. Lindsay, D. M. Hart, A. MacLean, A. C. Clark, A. Kraszewski, and J. Garwood, Bone response to termination of estrogen treatment. Lancet 1, 1325 – 1327 (1978). 8. C. Christiansen, M. S. Christiansen, and I. Transbol, Bone mass in postmenopausal women after withdrawal of estrogen/gestagen therapy. Lancet 1, 459 – 461 (1981). 9. S. Maggi, J. L. Kelsey, J. Litvak, and S. P. Heyse, Incidence of hip fractures in the elderly: A cross-national analysis. Osteopor. Int. 1, 232 – 241 (1991). 10. S. R. Cummings, J. A. Cauley, L. Palermo, P. D. Ross, R. P. Wasnich, D. Black, and K. G. Faulkner for the Study of Osteoporotic Fractures Research Group, Racial differences in hip axis lengths might explain racial differences in rates of hip fracture. Osteopor. Int. 4, 226 – 229 (1994). 11. J. W. Nieves, J. A. Grisso, and J. L. Kelsey, A case-control study of hip fracture: Evaluation of selected dietary variables and teenage physical activity. Osteopor. Int. 2, 122 – 127 (1992). 12. O. S. Miettinen, The “case-control” study: Valid selection of subjects. J. Chron. Dis. 38, 543 – 548 (1985). 13. A. Paganini-Hill, R. K. Ross, V. R. Gerkins, B. E. Henderson, M. Arthur, and T. M. Mack, Menopausal estrogen therapy and hip fractures. Ann. Intern. Med. 95, 28 – 31 (1981). 14. J. A. Grisso, J. L. Kelsey, L. A. O’Brien, C. G. Miles, S. Sidney, G. Maislin, K. LaPann, D. Moritz, B. Peters, and the Hip Fracture Study Group, Risk factors for hip fracture in men. Am. J. Epidemiology. 145, 786 – 793 (1997). 15. J. A. Grisso, J. L. Kelsey, B. L. Strom, L. A. O’Brien, G. Maislin, K. LaPann, L. Samelson, S. Hoffman, and the Northeast Hip Fracture Study Group, Risk factors for hip fracture in black women. N. Engl. J. Med. 330, 1555 – 1559 (1994). 16. N. S. Weiss, C. L. Ure, J. H. Ballard, A. R. Williams, and J. R. Daling, Decreased risk of fractures of the hip and lower forearm with postmenopausal use of estrogen. N. Engl. J. Med. 303, 1195 – 1198 (1980). 17. N. Kreiger, A. Gross, and G. Hunter, Dietary factors and fracture in postmenopausal women: A case-control study. Int. J. Epidemiol. 21, 953 – 958 (1992). 18. J. A. Grisso, J. L. Kelsey, B. L. Strom, G. Y. Chiu, G. Maislin, L. A. O’Brien, S. Hoffman, F. Kaplan, and the Northeast Hip Fracture Study Group, Risk factors for falls as a cause of hip fracture in women. N. Engl. J. Med. 324, 1326 – 1331 (1991). 19. D. L. Sackett, Bias in analytic research. J. Chron. Dis. 32, 51– 68 (1979). 20. H. Austin, H. A. Hill, W. D. Flanders, and R. S. Greenberg, Limitations in the application of case-control methodology. Epidemiol. Rev. 16, 65 – 76 (1994). 21. S. L. Hui, C. W. Slemenda, and C. C. Johnston, Jr., Age and bone mass as predictors of fracture in a prospective study. J. Clin. Invest. 81, 1804 – 1809 (1988). 22. M. Sowers, M. K. Clark, M. L. Jannausch, and R. B. Wallace, Body size, estrogen use and thiazide diuretic use affect 5-year radial bone loss in postmenopausal women. Osteopor. Int. 3, 314 – 321 (1993). 23. P. J. Diggle, K. Liang, and S. L. Zeger (eds.), “Analysis of Longitudinal Data.” Oxford Univ. Press, New York, 1998. 24. D. T. Felson, Y. Zhang, M. T. Hannon, D. P. Kiel, P. W. Wilson, and J. J. Anderson, The effect of postmenopausal estrogen therapy on bone density in elderly women. N. Engl. J. Med. 329, 1141 – 1146 (1993). 25. E. Barrett-Connor, J. C. Chang, and S. L. Edelstein, Coffee-associated osteoporosis offset by daily milk consumption, The Rancho Bernardo Study. JAMA 271, 280 – 283 (1994). 26. D. B. Petitti and S. Sidney, Hip fracture in women. Clin. Orthop. 246, 150 – 155 (1989). 27. M. Hernandez-Avila, G. A. Colditz, M. J. Stampfer, B. Rosner, F. E. Speizer, and W. C. Willett, Caffeine, moderate alcohol intake, and risk of fractures of the hip and forearm in middle-aged women. Am. J. Clin. Nutr. 54, 157 – 163 (1991).
KELSEY AND SOWERS 28. D. C. Bauer, W. S. Browner, J. A. Cauley, E. S. Orwell, J. C. Scott, D. M. Black, J. L. Tao, and S. R. Cummings for the Study of Osteoporotic Fractures Group, Factors associated with appendicular bone mass in older women. Ann. Intern. Med. 118, 657 – 665 (1993). 29. L. Forsen, A. Bjorndal, K. Bjartveit, T. H. Edna, J. Holman, V. Jessen, and G. Westberg, Interaction between current smoking, leanness, and physical inactivity in the prediction of hip fracture. J. Bone Miner. Res. 9, 1671 – 1678 (1994). 30. A. C. Looker, T. B. Harris, J. H. Madans, and C. T. Sempos, Dietary calcium and hip fracture risk: The NHANES I Epidemiologic Follow-Up Study. Osteopor. Int. 3, 177 – 184 (1993). 31. M. Sinaki, S. Khosla, P. J. Limburg, J. W. Rogers, and P. A. Murtaugh, Muscle strength in osteoporotic versus normal women. Osteopor. Int. 3, 8 – 12 (1993). 32. Institute of Medicine Committee to Review the NIH Women’s Health Initiative, “An Assessment of the NIH Women’s Health Initiative” (S. Thaul and D. Hotra, eds.). National Academy Press, Washington, DC, 1993. 32a. S. Hulley, D. Grady, T. Bush, C. Furberg, D. Herrington, B. Riggs, and E. Vittinghoff, Randomized trial of estrogen plus progestin for secondary prevention of coronary heart disease in postmenopausal women: Heart and Estrogen/Progestin Study (HERS) Research Group. JAMA 280, 605 – 613 (1998). 33. M. Sowers, B. Shapiro, M. A. Gilbraith, and M. Jannausch, Health and hormonal characteristics of premenopausal women with lower bone mass. Calcif. Tissue Int. 47, 130 – 135 (1990). 34. R. P. Heaney, Calcium, bone health, and osteoporosis. Bone Miner. Res. 4, 255 – 301 (1986). 35. R. G. Cumming, Calcium intake and bone mass: A quantitative review of the evidence. Calcif. Tissue Int. 47, 194 – 201 (1990). 36. A. Z. LaCroix, J. Wienpahl, L. R. White, R. B. Wallace, P. A. Scherr, L. K. George, J. Coroni-Huntley, and A. M. Ostfeld, Thiazide diuretic agents and the incidence of hip fracture. N. Engl. J. Med. 322, 286 – 290 (1990). 37. J. J. Schlesselman, “Case-Control Studies.” Oxford Univ. Press, New York, 1982. 38. S. Selvin, “Statistical Analysis of Epidemiologic Data.” Oxford Univ. Press, New York, 1991. 39. A. E. Schindler, A. Ebert, and E. Friedrich, Conversion of androstenedione to estrogen by human fat tissue. J. Endocrinol. Metab. 35, 627 – 630 (1972). 40. S. R. Cummings and M. C. Nevitt, A hypothesis: The causes of hip fractures. J. Gerontol. 44, M107 – M111 (1989). 41. A. M. Lilienfeld, Epidemiologic methods and inferences instudies of noninfectious diseases. Public Health Rep. 72, 51 – 60(1957). 42. R. Lindsay, Hormone replacement for prevention and treatment of osteoporosis. Am. J. Med. 95, 37S – 39S (1993). 43. J. F. Fleiss, “The Design and Analysis of Clinical Experiments.” Wiley, New York, 1986. 44. S. Greenland, The effect of misclassification in the presence of covariates. Am. J. Epidemiol. 112, 564 – 569 (1980). 45. C. R. Weinberg, D. M. Umbach, and S. Greenland, When will nondifferential misclassification of an exposure preserve the direction of a trend? Am. J. Epidemiol. 140, 565 – 571 (1994). 46. A. Paganini-Hill and R. K. Ross, Reliability of recall of drug usage and other health-related information. Am. J. Epidemiol. 116, 114 – 122 (1982). 47. T. W. O’Neill, C. Cooper, J. B. Cannata, J. B. Diaz Lopez, K. Hoszowski, O. Johnell, R. S. Lorene, B. Nilsson, H. Raspe, O. Stewart, and A. J. Silman on Behalf of the European Vertebral Osteoporosis (EVOS) Group, Reproducibility of a questionnaire on risk factors for osteoporosis in a multicentre prevalence survey: The European Vertebral Osteoporosis Study. Int. J. Epidemiol. 23, 559 – 565 (1994). 48. A. M. Schott, D. Hans, E. Somay-Rendu, P. D. Delmas, and P. J. Meunier, Ultrasound measurements on os calcis: Precision and
CHAPTER 20 Introduction to Epidemiologic Methods age-related changes in a normal female population. Osteopor. Int. 3, 249 – 254 (1993). 49. K. Liu, J. Stamler, A. Dyer, J. McKeever, and P. McKeever, Statistical methods to assess and minimize the role of intraindividual variability in obscuring the relationship between dietary lipids and serum cholesterol. J. Chron. Dis. 31, 399 – 418 (1978). 50. B. A. Rosner and W. C. Willett, Interval estimates for correlation coefficients corrected for within-person variation: Implications for
555 study design and hypothesis testing. Am. J. Epidemiol. 127, 337 – 386 (1988). 51. M. Sowers, Clinical epidemiology and ostoporosis: Measures and their interpretation. Clin. North Am. Endocrinol. 26, 219 – 231 (1997). 52. F. E. Thompson and T. Byers, Dietary Assessment Resource Manual. J. Nutr. 124, 11S, 2245S – 2317S (1994). 53. U. S. Department of Agriculture, “The Nutrient Composition of Foods,” Handbook 8, Vols. 1 – 23, Hyattsville, 1989.
CHAPTER 21
Magnitude and Impact of Osteoporosis and Fractures L. JOSEPH MELTON III CYRUS COOPER
I. II. III. IV.
Department of Health Sciences Research, Section of Clinical Epidemiology, Mayo Clinic and Foundation, Rochester, Minnesota 55905 MRC Environmental Epidemiology Unit, University of Southampton, Southampton General Hospital, Southampton SO16 6YD, England
V. Future Projections VI. Conclusions References
Introduction Magnitude of the Problem Fracture Epidemiology Impact of Osteoporotic Fractures
I. INTRODUCTION
osteoporosis and osteoporosis-related fractures and to determine the impact that the condition has on society.
When the term “osteoporosis” entered medical parlance in France and Germany during the past century [1], it implied a histological diagnosis (“porous bone”) that was subsequently refined to mean that bone tissue, while normally mineralized, was present in reduced quantity. This definitional approach culminates today in attempts to define osteoporosis on the basis of low bone mass [2], which can be assessed in vivo by a variety of noninvasive densitometric techniques (see Chapter 59). Low bone mineral density (BMD), in combination with impaired bone “quality” [3], leads to skeletal fragility and an increased risk of fracture, the clinical manifestation of osteoporosis [4]. Indeed, the realization that these fractures might result from an age-related reduction in bone strength antedated the histological observations [5]. This line of thought suggests that any assessment of osteoporosis must also include the associated fractures. The purposes of this review are to summarize epidemiologic data concerning the frequency of
OSTEOPOROSIS, SECOND EDITION VOLUME 1
II. MAGNITUDE OF THE PROBLEM How many people suffer from osteoporosis? It is clear from the preceding section that the answer to this question will depend on whether osteoporosis is defined on the basis of low bone mass or whether the emphasis, instead, is on osteoporosis-related fractures. Implicit in the definition of osteoporosis by bone mass alone is the relationship between this parameter and fracture risk. Low bone density is therefore analogous to high blood pressure, and the risk of fracture increases when bone density declines, just as the risk of stroke increases when blood pressure rises. A working group of the World Health Organization operationalized this concept by considering osteoporosis to be present when BMD levels in
557
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
558
MELTON AND COOPER
white women are more than 2.5 SD below the young normal mean [6]. To provide some comparability with earlier definitions that incorporated fracture, the subset of women with presumptive osteoporosis who also have a history of one or more fragility fractures are deemed to have severe (“established”) osteoporosis. Low bone mass (“osteopenia”) is defined by bone density levels more than 1 SD below the young normal mean, but less than 2.5 SD below. Some of the best data on the prevalence of osteoporosis come from the Third National Health and Nutrition Examination Survey (NHANES III), a large probability sample of the United States population. As assessed at the femoral neck, 20% of postmenopausal white women have osteoporosis [7]. No criterion for osteoporosis has been established for nonwhite women, but 10% of Hispanic women and 5% of African-American women have femoral neck BMD values more than 2.5 SD below the young normal mean for white women. Using the same cut-off value, osteoporosis prevalence rates for white, Hispanic, and African-American men age 50 years and over are 4, 2, and 3% respectively [7]. However, these race and genderspecific differences partly relate to overestimation of areal BMD (g/cm2) in individuals with larger skeletons [8]. When bone size is taken into account with bone mineral apparent density (BMAD, g/cm3), differences in bone density between men and women [9,10] and between women of different races [11 – 13] are reduced. In addition, osteoporosis is a systemic disease, and a greater proportion of the population is seen to be affected when more skeletal sites are assessed (Table 1). Most white women under age 50 have normal bone density, but with advancing age, the proportion with osteoporosis increases dramatically. Among women age 80 years and over, for example, only 3% have normal bone density at the hip, spine,
and forearm, while 27% have osteopenia at one skeletal site or another and 70% have osteoporosis. As judged from these data, an estimated 16.8 million (54%) postmenopausal white women in the United States have osteopenia and another 9.4 million (30%) have osteoporosis [14]. However, results vary from one geographic region to another [15]. The alternative approach is to assess fracture frequency. There are no data regarding the prevalence of distal forearm fracture, while the prevalence of hip fracture has been estimated at 51 per 1000 women age 65 and over [16] or 6 per 1000 men and women of all ages [17]. Most available information relates to vertebral fractures. In the European Vertebral Osteoporosis Study, for example, 15,570 men and women aged 50 – 79 years were selected from population registers in 36 European centers and evaluated according to a standardized protocol [18]. The overall prevalence of morphometrically defined vertebral deformity was 12% in both men and women, but the increase in prevalence with age was steeper among the women (Fig.1). Thus, the frequency of deformities in men aged 50 – 54 years was around 10%, rising to 18% at age 75 – 79 years, while the prevalence rose from 5% among women aged 50 – 54 years to over 24% at age 75 – 79 years. Using a different approach to morphometry, other studies have estimated the overall prevalence of vertebral deformities among postmenopausal women at 20 to 24% [19 – 22]. A frequency measure of interest to patients and clinicians is the probability of fracture over an average lifetime. Using fracture incidence rates from the United States, the estimated lifetime risk of a hip fracture is 17% in white women and 6% in white men [23]. This compares with risks of 16 and 5% for clinically diagnosed vertebral fractures and 16 and 2% for distal forearm fractures in white women and men, respectively. The lifetime risk of any of
TABLE 1 Proportion (%) of Rochester, Minnesota, Women with Bone Density Measurements More Than 2.5 SD below the Mean for Young Normal Womena Lumbar spine (%)
Either hip site (%)
50 – 59
7.6
3.9
3.7
14.8
60 – 69
11.8
8.0
11.8
21.6
70 – 79
25.0
24.5
23.1
38.5
80
32.0
47.5
50.0
70.0
Totalb
16.5
16.2
17.4
30.3
Age group
Midradius (%)
Spine, hip, or midradius(%)
a Mean is from 48 subjects under age 40 who were randomly sampled from the Rochester, Minnesota, population. None of them was known to have any disorder that might influence bone metabolism. From L. J. Melton III, How many women have osteoporosis now? J. Bone Miner. Res. 10, 175 – 177 (1995). b Age-adjusted to the population structure of 1990 United States white women 50 years of age and older.
CHAPTER
559
21 Magnitude and Impact of Osteoporosis and Fractures
FIGURE 1
Prevalence of vertebral deformities among European men and women by age. Data derived from the European Vertebral Osteoporosis Study. Reproduced from O’Neill et al. [18], with permission.
the three fractures is 40% for women and 13% for men from age 50 years onward [23]. In Great Britain, the lifetime risk of a hip fracture among 50-year-old women is 14%, compared to 3% for British men of comparable age [24]. This contrasts with lifetime risks of 11 and 2% for clinical diagnosed vertebral deformities and 13 and 2% for wrist fractures in white women and men, respectively. Because lifetime risk depends on life expectancy as well as fracture incidence, the lifetime risk of hip fracture in British women could rise to 24% by 2050 if life expectancy continues to increase [24].
III. FRACTURE EPIDEMIOLOGY Fracture incidence in the community is bimodal, with peaks in youth and advanced age [25,26]. In young people, fractures of the long bones predominate, often following substantial trauma, and the incidence is greater in young men than in young women. Above the age of 35 years, overall fracture incidence in women climbs steeply so that female rates become twice those in men [27]. At least 1.3 million fractures in the United States each year have been attributed to osteoporosis, presuming that 70% of all fractures in persons aged 45 years or over are due to the condition [28]. The three sites most closely associated with osteoporosis are fractures of the hip, spine, and distal forearm. However, the epidemiologic characteristics of these three fractures differ, suggesting the influence of different etiologic factors.
A. Hip Fracture In most populations, hip fracture incidence rates increase exponentially with age (Fig. 2). In Rochester, Minnesota, rates reach about 3.0% per year among women age 85 years and over and 1.9% among men in this age group
FIGURE 2
Age-specific incidence rates for hip, vertebral, and distal forearm fractures in men and women. Data derived from the population of Rochester, Minnesota. From C. Cooper and L. J. Melton III, Epidemiology of osteoporosis. Trends Endocrinol. Metab. 3, 224 – 229 (1992).
[29]. At all ages beyond 50 years, the incidence in women is about twice that in men. Because there are more elderly women than men, however, about 80% of all hip fractures occur in women. Worldwide, there were an estimated 1.7 million hip fractures in 1990, about 1,197,000 in women, and another 463,000 or so in men [30]. A minority of such fractures are due to overwhelming trauma or to specific lesions in the proximal femur, although severe trauma accounts for a greater proportion of the total in countries where hip fractures are uncommon [29]. The vast majority of hip fractures follow a fall from standing height or less in individuals with reduced bone strength [31]. Over a lifetime, bone density of the femoral neck declines an estimated 58 and 39% in women and men, respectively, while bone density of the intertrochanteric region of the proximal femur falls about 53 and 35% [32]. Each 1 SD decline in BMD is associated with a 2.0- to 3.6-fold increase in the age-adjusted risk of hip fracture, depending on the exact site in the proximal femur that is measured [33]. Simultaneously, there is a dramatic increase in the likelihood of falling each year, from about one in five women age 40 – 49 years to nearly half of women 85 years old and over, along with a third of men in the oldest age group [34]. The pathophysiology of falling is complex (see Chapter 32), but only about 1% of falls lead to a hip fracture [35]. This is because the amount of force delivered to the proximal femur depends on various protective responses and on the orientation of the faller (see Chapter 19). The seemingly inexorable bone loss and increased risk of falling that accompany the aging process suggest that hip fractures are inevitable. In fact, incidence rates vary substantially from one population to another. Thus, hip fractures are much less frequent among nonwhites than whites (see Chapter 22), but there is substantial variation even within populations of a given race and gender (Fig. 3). Thus, incidence rates are higher among white residents
560
MELTON AND COOPER
FIGURE 3 Hip fracture incidence around the world as a ratio of the rates observed in various populations to those expected for U.S. white women. From L. J. Melton III, Differing patterns of osteoporosis across the world, In “New Dimensions in Osteoporosis in the 1990s” (C. H. Chesnut, ed.), pp. 13 – 18. Excerpta Medica Asian, Hong Kong, 1991. of Scandinavia than comparable people in North America or Oceania [36]; even within Europe, hip fracture rates vary more than sevenfold from one country to another [37]. The higher incidence in urban than rural districts has been explained on the basis of lower bone mass among urban residents [38], but data from the United States show that the pattern is much more complex (Fig. 4). In over 2000 counties nationwide, the incidence of hip fractures in elderly white women was negatively associated with latitude (higher in the South), water hardness, and mean hours of January sunlight and positively associated with poverty levels, proportion of the land in farms, and proportion of the population with fluoridated water supplies [39]. Regional differences did not seem to be accounted for by variation in activity levels, obesity, cigarette smoking, alcohol consumption, or Scandinavian heritage. Additional studies are needed to identify the environmental factors associated with such regional differences.
B. Vertebral Fracture Vertebral fractures have been synonymous with osteoporosis since its earliest description as a metabolic bone disorder [40]. However, epidemiologic data remain scant because there is no universally accepted definition of vertebral fracture and because a substantial proportion are
asymptomatic (see Chapter 34). Although it has long been clear that some vertebral fractures do not reach clinical attention, the size of this fraction was unknown. The age-adjusted incidence of clinically diagnosed vertebral fractures has been estimated at 5.3 per 1000 person-years among white women aged 50 years and over [41]. This represents about 30% of the total incidence of vertebral fractures among American postmenopausal white women of 18 per 1000 person-years [21]. The incidence of clinically diagnosed vertebral fractures was 4.3 times greater than rates derived from United States hospital discharge data for spine fractures [42], suggesting that a third of all vertebral deformities come to medical attention with about 8% necessitating admission to hospital. Data for England and Wales imply that as few as 2% might be hospitalized [43], but this is almost certainly an underestimate due to incomplete diagnostic coding. Figure 2 illustrates the incidence of clinically diagnosed vertebral fractures. In men, incidence rates climb exponentially with age, adopting a pattern similar to that observed for hip fractures in the same population [29]. In women, there is a more linear increase such that vertebral fracture rates are higher than those for hip fracture before the age of 70 years, but not thereafter. Figure 2 also illustrates that vertebral fractures are a greater problem in men than previously recognized, with an overall age-adjusted incidence in women only 1.9 times greater than that in men [41], although sex ratios vary by geographic region [18]. Other
CHAPTER
561
21 Magnitude and Impact of Osteoporosis and Fractures
Age-adjusted incidence of hip fractures among white women 65 years of age by county of residence in the United States,(1984 – 1987). From S. J. Jacobsen, J. Goldberg, T. P. Miles, J. A. Brody, W. Stiers, and A. A. Rimm, Regional variation in the incidence of hip fracture. JAMA 264, 500 – 502 (1990).
FIGURE 4
than the expected relationship with secondary osteoporosis (see Chapter 51), strong underlying risk factors have been difficult to define [44 – 50]. Falls accounted for only a third of new vertebral fractures among Rochester men and women; the majority were due instead to the compressive loading associated with lifting, changing positions, and so on or were discovered only incidentally [41]. The most frequent vertebral levels involved were T8 and T12/L1. These correspond with biomechanically compromised regions of the spine: the midthoracic region, where the dorsal kyphosis is most pronounced, and the thoracolumbar junction, where the relatively rigid thoracic spine meets the freely moving lumbar segment [31].
C. Distal Forearm Fracture Distal forearm fractures almost always follow a fall on the outstretched arm [51]. They display a different pattern of incidence compared to hip or vertebral fractures (Fig. 2). In white women, incidence rates increase linearly from age 40 to age 65 years and then stabilize [29] for reasons that remain obscure but may relate to a change in the pattern of falling with advancing age so that elderly women with slower gait and impaired neuromuscular coordination are more likely to fall on their hip than on their wrist [51,52]. In addition, compared to hip fractures, a greater proportion
of distal forearm fractures occur outdoors, and a winter peak in incidence has been associated with periods of icy weather [53,54]. In men, the incidence of distal forearm fractures remains relatively constant and low between ages 20 and 80 years. Consequently, the majority of such fractures occur in women, and the age-adjusted female to male ratio of 4:1 is more marked than for hip or vertebral fractures. Nonetheless, the incidence of distal forearm fracture varies from one geographic area to another generally in parallel with hip fracture incidence rates [29].
D. Other Fractures Incidence rates for fractures of the proximal humerus, pelvis, proximal tibia and distal femur also increase with aging in elderly women but to a lesser extent in men. Along with most other fractures among elderly women, these have been directly associated with low bone density [55]. About 80% of proximal humerus fractures occur in individuals 35 years old and over, and three-fourths are in women [56]. The same general picture is seen in other populations [29]. Three-fourths or more of all proximal humerus fractures are due to moderate trauma, typically a fall from standing height or less [56,57]. Like the falls related to hip fracture, these seem to be more frequent in frail women with poor neuromuscular function [58].
562
MELTON AND COOPER
Dramatic age-related increases are also seen for pelvic fractures [29]. Overall, about two-thirds of all pelvic fractures occur among persons age 35 years or older, and nearly 70% are in women [59]. While multiple pelvic fractures and acetabular fractures are associated with severe trauma, moderate trauma accounts for nearly two-thirds of fractures of isolated pelvic bones and single breaks in the pelvic ring, which constitute about 80% of all pelvic fractures in the community [59]. Proximal tibia fractures have been classified as “composite” fractures, with peaks in incidence among the young and the old [60]. Incidence rates are highest in adolescent boys, but age-adjusted incidence rates are 30% greater among women [29]. Proximal tibia fractures in women outnumber those in men by about 2:1 after age 50 years [27,61]. Over 40% of proximal tibia fractures are due to severe trauma, while only a fourth are due to falls. Nonetheless, low bone mass increases the risk of leg fractures in elderly women [55], and the majority of proximal tibia fractures among older individuals are related to osteoporosis [62]. Distal femur fractures also have some of the characteristics of an age-related fracture [29]. The incidence of fractures of the femur shaft in Stockholm is greater in women than in men after age 65 years; the majority of the fractures in this age group are due to falls, and the incidence of femur shaft fractures due to moderate trauma increases exponentially with age in both sexes [63]. However, most fractures of the femur shaft are due to severe trauma. Moderate trauma, however, accounts for half of all distal femur fractures in Rochester, and this subgroup of cases exhibits age-related increases in incidence among both men and women like those seen for hip fractures [64].
IV. IMPACT OF OSTEOPOROTIC FRACTURES The adverse outcomes of osteoporotic fractures fall into three broad categories: mortality, morbidity, and cost.
A. Mortality The influence that hip, spine, and distal forearm fractures have on survival appears to differ with the type of fracture. Hip fractures are the most serious, leading to an overall reduction in survival of 10 – 20% [29]. Excess deaths occur mainly within the first 6 months and diminish with time, although the death rate may remain elevated for a number of years [65]. Mortality differs, however, by age and sex. In one population-based study, a relative survival of 92% was found for white hip fracture victims under 75 years of age, compared to only 83% for those aged 75 years
and over [29]. Despite their greater average age at the time of fracture, survival is better among women. This sex difference appears to arise from the greater frequency of other chronic diseases among men who sustain hip fractures [66]. Thus, while some deaths are attributable to acute complications of the fracture or of its surgical management [67], the majority appear to be due to serious coexisting illnesses [68]. There does not appear to be any excess mortality among patients who sustain distal forearm fractures [69] and, until recently, it was assumed that osteoporotic vertebral fractures were not attended by significant mortality either. However, 5-year survival among patients with clinically diagnosed vertebral fractures is only 61% compared to 76% expected for those of like age and sex (relative survival, 0.82) [69]. The proportionate excess of deaths is thus comparable to that following hip fracture (Table 2). Except for an excess of pulmonary deaths in women with severe vertebral deformities and kyphosis [70], strong associations with specific causes of death have not been identified [69,70], suggesting that impaired survival may be due to an indirect association with comorbid conditions that lead to an increased risk of osteoporosis. This explanation would accord with the observation that low bone density is itself associated with excess mortality from various causes [71 – 73].
B. Morbidity After allowing for the functional impairment expected in aged people, fractures of the hip, spine, and distal forearm result in an estimated 7% of women becoming dependent in the basic activities of daily living and cause nursing home care in a further 8% [74]. As reviewed elsewhere (see Chapter 34), hip fractures contribute most to this burden. They almost invariably necessitate hospitalization, and in 1985 the average length of hospital stay in England and Wales was 30 days so that 3500 National Health Service hospital beds were occupied daily. While these patients are at high risk of acute complications such as pressure sores, pneumonia, and urinary tract infections, the most important long-term outcome is impairment in the ability to walk. Around 20% of patients are nonambulatory even before fracture, but of those able to walk, half cannot walk independently afterward [75]. Ultimately, up to a third of hip fracture victims may become totally dependent [76], and the risk of institutionalization is great [77]. Nearly 140,000 nursing home admissions are attributable to hip fractures each year in the United States [78], and as many as 8% of all nursing home residents have had a hip fracture [79]. The health impact of vertebral fractures has proved considerably more difficult to quantify. As noted earlier, only a minority of incident vertebral deformities come to clinical attention. Nonetheless, vertebral fractures in patients aged
CHAPTER
563
21 Magnitude and Impact of Osteoporosis and Fractures
TABLE 2 Relative Survival Following Vertebral, Hip and Distal Forearm Fractures among Residents of Rochester, Minnesota, According to Duration of Follow-Up from Diagnosis Relative survival (95% CI) Time from diagnosis (year)
Vertebral
Hip
Forearm
1
0.96 (0.92 – 0.99)
0.88 (0.85 – 0.91)
1.00 (0.98 – 1.02)
2
0.93 (0.87 – 0.99)
0.87 (0.83 – 0.90)
1.00 (0.97 – 1.03)
3
0.92 (0.86 – 0.98)
0.86 (0.82 – 0.90)
1.01 (0.98 – 1.04)
4
0.84 (0.75 – 0.92)
0.83 (0.78 – 0.88)
0.99 (0.95 – 1.04)
5
0.82 (0.71 – 0.93)
0.83 (0.77 – 0.89)
1.00 (0.95 – 1.05)
45 years and older account for about 52,000 hospital admissions in the United States [79] and 2188 in England and Wales each year. The major consequences of vertebral fracture are back pain, kyphosis, and height loss. New compression fractures may give rise to severe back pain, which typically resolves over weeks or months [80]. A more protracted clinical course affects a proportion of patients who experience chronic pain while standing and during physical stress, particularly bending [81]. Not only physical function but self-esteem, body image, and mood also appear to be adversely affected in patients with vertebral fractures (see Chapter 61). Despite the fact that only about one-fifth of all patients with distal forearm fractures are hospitalized [82], they account for some 50,000 hospital admissions and over 400,000 physician visits in the United States each year [79]. Admission rates vary markedly with age, such that only 16% of forearm fractures occurring in women ages 45 – 54 years required inpatient care compared to 76% of those in women age 85 years and over [83]. There is a 30% risk of algodystrophy after these fractures [84], as well as an increased likelihood of neuropathies and posttraumatic arthritis [85]. Nearly half of all patients report only fair or poor functional outcomes at 6 months following a distal forearm fracture [86].
C. Economic Costs Fractures in the United States may cost as much as $20 billion per year, with hip fractures accounting for over a third of the total [87]. Because most hip fractures are in elderly individuals, wages foregone or years of life lost are not the primary determinants of cost. Eiskjaer et al. [88] found that the 9.2 years of potential life per 1000 women lost due to hip fracture was much lower than comparable figures for heart disease, stroke, and breast cancer (73, 29, and 20 per 1000 women, respectively). Instead, the greatest
expense is for inpatient medical services and nursing home care (Table 3). The direct costs in Table 3 include an estimated 547,000 hospitalizations and 4.6 million hospital bed days for the care of osteoporotic fractures in the United States in 1995 [78]. In Switzerland, osteoporotic fractures account for more hospital bed days than myocardial infarction and stroke at a cost of CHF 600 million [89]. In England, hip fractures alone consume one-fifth of all orthopedic beds, at a direct cost of £850 million per year in 1999 [90]. In France, an estimated 56,000 hip fractures annually cost about FF 3.5 billion [91]. Such figures are a source of concern to governmental leaders in almost every country.
V. FUTURE PROJECTIONS The cost of osteoporosis-related fractures can only rise in the future. Life expectancy is increasing around the globe and the number of elderly individuals is rising in every region. The estimated 323 million individuals in the world age 65 years or over at present is expected to rise to 1555 million by the year 2050 [30]. These demographic changes alone can be expected to cause the number of hip fractures throughout the world to increase from about 1.7 million in 1990 to 6.3 million in 2050 (Fig. 5). Around half of all hip fractures occurred in Europe and North America in 1990, but rapid aging of the Asian and Latin American populations will reduce the European and North American contribution to only 25% by 2050, with over half of all hip fractures occurring in Asia [30]. It is clear, therefore, that osteoporosis will truly become a global problem over the next half century and that measures are urgently required to avert this trend. Such projections would be worsened by any increase in hip fracture incidence. For example, an increase of only 1% in the age-adjusted incidence rate each year might cause the estimated number of hip fractures in 2050 to reach 8.1 million [92]. Fortunately, hip fracture rates appear to have
564
MELTON AND COOPER
TABLE 3
Health Care Expenditures Attributable to Osteoporotic Fractures in the United States by Type of Service and Type of Fracture (1995)a Type of service (millions of dollars)
Type of fracture Hip
Inpatient hospital
Emergency room
Outpatient physician
Outpatient hospital
Other outpatientb
Nursing home
Total
5576
130
67
9
90
2811
8682
Forearm
183
55
93
8
4
41
385
Spine
575
20
13
3
10
126
746
All other sites
2259
3632
297
45
91
899
3953
Total
8594
567
470
65
194
3875
13,764
a From N. F. Ray, J. K. Chan, M. Thamer, and L. J. Melton III, Medical expenditures for the treatment of osteoporotic fractures in the United States in 1995: Report from the National Osteoporosis Foundation. J. Bone Miner. Res. 12, 24 – 35 (1997). b Includes home health care, ambulance services, and medical equipment.
leveled off in the northern United States [93], in parts of Scandinavia [94 – 97] and in Great Britain [98], although rates in Asia have risen substantially in recent years [99]. On the basis of current trends, hip fractures might increase in the United Kingdom from 46,000 in 1985 to 117,000 in 2016 [100]. In Australia, they could rise from 10,150 in 1986 to 18,550 in 2011, with a doubling in the cost of care in constant dollars from $38 to $69 million annually [101]. Health authorities in Finland expect to see a three-fold increase in the number of hip fractures between 1997 and 2030 [102], while in Canada, there could be a four-fold increase by 2041 [103]. Incidence rates for fractures at other skeletal sites have also risen during the last half century [104]. There are three broad explanations for this trend. First, it might reflect the influence of some increasingly prevalent risk factor for
osteoporosis or for falling. Time trends for a number of osteoporosis risk factors, including oophorectomy, estrogen replacement therapy, cigarette smoking, alcohol consumption, and dietary calcium intake, do not match those observed for hip fractures [105], but physical activity could be a candidate. Ample evidence links inactivity to hip fracture risk [106 – 109], and the steepest increases in incidence have been observed in Asian countries, which have witnessed dramatic reductions in customary activity levels [99]. A second possible explanation is increasing frailty among the elderly population, especially since many of the disorders contributing to frailty are also associated with osteoporosis and the risk of falling [110]. Finally, the trends could arise from a cohort phenomenon, i.e., some adverse influence that acted at an earlier time is now manifesting as a rising fracture incidence in successive generations. For example, it has been speculated that the increase in adult height during this century led to a secular trend toward longer hip axis length, which may increase the risk of hip fracture [111]. However, analysis of data from the Oxford Record Linkage Study revealed a declining incidence of hip fracture in more recent birth cohorts [112]. This is in accord with a recent survey in southern England, where fracture rates increased in the elderly between 1978 and 1995 but were tending to decrease among people under 70 years of age [113].
VI. CONCLUSIONS
FIGURE 5
Estimated number of fractures (in thousands) for men and women in different regions of the world in 1990, 2025, and 2050. Modified from C. Cooper, G. Campion, and L. J. Melton III, Hip fractures in the elderly: A world-wide projection. Osteopor. Int. 2, 285 – 289 (1992).
Osteoporosis is a complex, multifactorial chronic disorder in which a variety of pathophysiologic mechanisms lead to a progressive reduction in bone strength and an increased risk of fracture. Although viewed for many years as a major public health problem, the exact burden posed by
21 Magnitude and Impact of Osteoporosis and Fractures
565
osteoporosis is only now being rigorously assessed. Whether the disorder is defined by low bone mass or by the occurrence of specific fractures, osteoporosis is clearly a common condition. Thus, a third of postmenopausal white women in the United States can be expected to have osteoporosis in the lumbar spine, proximal femur, or midradius at any point in time, while the lifetime risk of a hip, spine, or distal forearm fracture from age 50 years onward in this group approaches 40%. However, the relative absence of symptoms until fractures occur makes effective therapeutic intervention difficult to implement. The public health burden will worsen dramatically in future decades, and the evaluation of strategies to prevent these fractures, both in individuals and in populations, has become an urgent priority.
density in the postmenopausal estrogen/progestin interventions trial. J. Bone Miner. Res. 9, 1467–1476 (1994). T. Cundy, J. Cornish, M. C. Evans, G. Gamble, J. Stapleton, and I. R. Reid, Sources of interracial variation in bone mineral density. J. Bone Miner. Res. 10, 368–373 (1995). G. S. Bhudhikanok, M.-C. Wang, K. Eckert, C. Matkin, R. Marcus, and L. K. Bachrach, Differences in bone mineral in young Asian and Caucasian Americans may reflect differences in bone size. J. Bone Miner. Res. 11, 1545–1556 (1996). L. J. Melton III, How many women have osteoporosis now? J. Bone Miner. Res. 10, 175–177 (1995). M. Lunt, D. Felsenberg, J. Adams, L. Benevolenskaya, J. Cannata, J. Dequeker, C. Dodenhof, J. A. Falch, O. Johnell, K.-T., Khaw, P. Masaryk, H. Pols, G. Poór, D. Reid, C. Scheidt-Nave, K. Weber, A. J. Silman, and J. Reeve, Population-based geographic variations in DXA bone density in Europe: The EVOS Study. Osteoporos. Int. 7, 175–189 (1997). M. C. Hochberg, J. Williamson, E. A. Skinner, J. Guralnik, J. D. Kasper, and L. P. Fried, The prevalence and impact of self-reported hip fracture in elderly community-dwelling women: The Women’s Health and Aging Study. Osteoporos. Int. 8, 385–389 (1998). L. J. Melton III, T. M. Therneau, and D. R. Larson, Long-term trends in hip fracture prevalence: The influence of hip fracture incidence and survival. Osteoporos. Int. 8, 68–74 (1998). T. W. O’Neill, D. Felsenberg, J. Varlow, C. Cooper, J. A. Kanis, and A. J. Silman, The prevalence of vertebral deformity in European men and women: The European Vertebral Osteoporosis Study. J. Bone Miner. Res. 11, 1010 – 1018 (1996). G. F. Jensen, C. Christiansen, J. Boesen, V. Hegedüs, and I. Transbøl, Epidemiology of postmenopausal spinal and long bone fractures: A unifying approach to postmenopausal osteoporosis. Clin. Orthop. 166, 75 – 81 (1982). B. Ettinger, D. M. Black, M. C. Nevitt, A. C. Rundle, J. A. Cauley, S. R. Cummings, and H. K. Genant, The Study of Osteoporotic Fractures Research Group, Contribution of vertebral deformities to chronic back pain and disability. J. Bone Miner. Res. 7, 449 – 456 (1992). L. J. Melton III, A. W. Lane, C. Cooper, R. Eastell, W. M. O’Fallon, and B. L. Riggs, Prevalence and incidence of vertebral deformities. Osteoporos. Int. 3, 113 – 119 (1993). G. Jones, C. White, T. Nguyen, P. N. Sambrook, P. J. Kelly, and J. A. Eisman, Prevalent vertebral deformities: Relationship to bone mineral density and spinal osteophytosis in elderly men and women. Osteoporos. Int. 6, 233 – 239 (1996). L. J. Melton III, E. A. Chrischilles, C. Cooper, A. W. Lane, and B. L. Riggs, How many women have osteoporosis? J. Bone Miner. Res. 7, 1005 – 1010 (1992). C. Cooper, Epidemiology of osteoporosis. Osteoporos. Int. 9 (Suppl. 2), S2 – S8 (1999). W. M. Garraway, R. N. Stauffer, L. T. Kurland, and W. M. O’Fallon, Limb fractures in a defined population. I. Frequency and distribution. Mayo Clin. Proc. 54, 701 – 707 (1979). Editorial, Fracture patterns revisited. Lancet 336, 1290 – 1291 (1990). L. J. Donaldson, A. Cook, and R. G. Thomson, Incidence of fractures in a geographically defined population. J. Epidemiol. Community Health 44, 241 – 245 (1990). J. L. Kelsey, Osteoporosis: Prevalence and incidence. In “Osteoporosis,” pp. 25 – 28. Proceedings of the NIH Consensus Development Conference, April 2 – 4, 1984. L. J. Melton III, Epidemiology of fractures. In “Osteoporosis: Etiology, Diagnosis, and Management” (B. L. Riggs and L. J. Melton III, eds.), pp. 133 – 154. Raven Press, New York, 1988. C. Cooper, G. Campion, and L. J. Melton III, Hip fractures in the elderly: A worldwide projection. Osteoporos. Int. 2, 285 – 289 (1992).
CHAPTER
12.
13.
14. 15.
16.
Acknowledgments
17.
The authors thank Mrs. Gill Strange for her assistance in preparing the manuscript. This work was supported in part by Research Grants AG04875 and AR27065 from the National Institutes of Health, U.S. Public Health Service, and by the Medical Research Council of Great Britain.
18.
19.
References 1. D. Schapira and C. Schapira, Osteoporosis: The evolution of a scientific term. Osteoporos. Int. 2, 164–167 (1992). 2. J. A. Kanis and the WHO Study Group, Assessment of fracture risk and its application to screening for postmenopausal osteoporosis: Synopsis of a WHO report. Osteoporos. Int. 4, 368–381 (1994). 3. C. Cooper, The epidemiology of fragility fractures: Is there a role for bone quality? Calcif. Tissue Int. 53 (Suppl. 1), S23–S26 (1993). 4. Anonymous, Consensus Development Statement: Who are candidates for prevention and treatment for osteoporosis? Osteoporos. Int. 7, 1–6 (1997). 5. A. Cooper, “A Treatise on Dislocations and Fractures of the Joints” (B. B. Cooper, ed.). John Churchill, London, 1842 6. World Health Organization, “Assessment of Fracture Risk and Its Application to Screening for Postmenopausal Osteoporosis,” WHO Technical Report Series. WHO, Geneva, 1994. 7. A. C. Looker, E. S. Orwoll, C. C. Johnston, Jr., R. L. Lindsay, H. W. Wahner, W. C. Dunn, M. S. Calvo, T. B. Harris, and S. P. Heyse, Prevalence of low femoral bone density in older US adults from NHANES III. J. Bone Miner. Res. 12, 1761–1768 (1997). 8. E. Seeman, Growth in bone mass and size: Are racial and gender differences in bone mineral density more apparent than real? J. Clin. Endocrinol. Metab. 83, 1414–1419 (1998). 9. R. A. Faulkner, R. G. McCulloch, S. L. Fyke, W. E. De Couteau, H. A. McKay, D. A. Bailey, C. S. Houston, and A. A. Wilkinson, Comparison of areal and estimated volumetric bone mineral density values between older men and women. Osteoporos. Int. 5, 271–275 (1995). 10. L. J. Melton III, E. J. Atkinson, M. K. O’Connor, W. M. O’Fallon, and B. L. Riggs, Bone density and fracture risk in men. J. Bone Miner. Res. 13, 1915–1923 (1998). 11. R. Marcus, G. Greendale, B. A. Blunt, T. L. Bush, S. Sherman, R. Sherwin, H. Wahner, and B. Wells, Correlates of bone mineral
20.
21.
22.
23.
24. 25.
26. 27.
28.
29.
30.
566 31. L. J. Melton III, E. Y. S. Chao, and J. Lane, Biomechanical aspects of fractures. In “Osteoporosis: Etiology, Diagnosis, and Management” (B. L. Riggs and L. J. Melton III, eds.), pp. 111 – 131. Raven Press, New York, 1988. 32. B. L. Riggs, H. W. Wahner, E. Seeman, K. P. Offord, W. L. Dunn, R. B. Mazess, K. A. Johnson, and L. J. Melton III, Changes in bone mineral density of the proximal femur and spine with aging: Differences between the postmenopausal and senile osteoporosis syndromes. J. Clin. Invest. 70, 716 – 723 (1982). 33. D. Marshall, O. Johnell, and H. Wedel, Meta-analysis of how well measures of bone mineral density predict occurrence of osteoporotic fractures. Br. Med. J. 312, 1254 – 1259 (1996). 34. S. J. Winner, C. A. Morgan, and J. G. Evans, Perimenopausal risk of falling and incidence of distal forearm fracture. Br. Med. J. 298, 1486 – 1488 (1989). 35. M. J. Gibson, The prevention of falls in later life. Dan. Med. Bull. 34 (Suppl. 4), 1 – 24 (1987). 36. L. J. Melton III, Differing patterns of osteoporosis across the world. In “New Dimensions in Osteoporosis in the 1990s” (C. H. Chesnut III, ed.), pp. 13 – 18. Asia Pacific Conference Series 125, Excerpta Medica, Hong Kong, 1991. 37. O. Johnell, B. Gullberg, E. Allander, J. A. Kanis, and the MEDOS Study Group, The apparent incidence of hip fracture in Europe: A study of national register sources. Osteoporos. Int. 2, 298 – 302 (1992). 38. P. Gärdsell, O. Johnell, B. E. Nilsson, and I. Sernbo, Bone mass in an urban and a rural population: A comparative, population-based study in southern Sweden. J. Bone Miner. Res. 6, 67 – 75 (1991). 39. S. J. Jacobsen, J. Goldberg, T. P. Miles, J. A. Brody, W. Stiers, and A. A. Rimm, Regional variation in the incidence of hip fracture: US white women aged 65 years and older. JAMA 264, 500 – 502 (1990). 40. F. Albright, P. H. Smith, and A. M. Richardson, Postmenopausal osteoporosis: Its clinical features. JAMA 116, 2465 – 2474 (1941). 41. C. Cooper, E. J. Atkinson, W. M. O’Fallon, and L. J. Melton III, The incidence of clinically diagnosed vertebral fractures: A populationbased study in Rochester, Minnesota, 1985 – 1989. J. Bone Miner. Res. 7, 221 – 227 (1992). 42. S. J. Jacobsen, C. Cooper, M. S. Gottlieb, J. Goldberg, D. P. Yahnke, and L. J. Melton III, Hospitalization with vertebral fracture among the aged: A national population-based study, 1986 – 1989. Epidemiology 3, 515 – 518 (1992). 43. J. A. Kanis and E. V. McCloskey, Epidemiology of vertebral osteoporosis. Bone 13, S1 – S10 (1992). 44. R. D. Wasnich, P. D. Ross, C. J. MacLean, J. W. Davis, and J. M. Vogel, The relative strengths of osteoporotic risk factors in a prospective study of postmenopausal osteoporosis. In “Osteoporosis 1987” (C. Christiansen, J. S. Johansen, and B. J. Riis, eds.), Vol. 1, pp. 394 – 395. Osteopress ApS, Copenhagen, 1987. 45. C. Cooper, S. Shah, D. J. Hand, J. Adams, J. Compston, M. Davie, A. Woolf, and the Multicentre Vertebral Fracture Study Group, Screening for vertebral osteoporosis using individual risk factors. Osteoporos. Int. 2, 48 – 53 (1991). 46. C. Huang, P. D. Ross, S. Fujiwara, J. W. Davis, R. S. Epstein, K. Kodama, and R. D. Wasnich, Determinants of vertebral fracture prevalence among native Japanese women and women of Japanese descent living in Hawaii. Bone 18, 437 – 442 (1996). 47. M. Naves Diaz, T. W O’Neill, and A. J. Silman, The influence of alcohol consumption on the risk of vertebral deformity: European Vertebral Osteoporosis Study Group. Osteoporos. Int. 7, 65 – 71 (1997). 48. A. J. Silman, T. W. O’Neill, C. Cooper, J. Kanis, and D. Felsenberg, Influence of physical activity on vertebral deformity in males and women: Results from the European Vertebral Osteoporosis Study. J. Bone Miner. Res. 12, 813 – 819 (1997). 49. T. W. O’Neill, A. J. Silman, M. Naves Diaz, C. Cooper, J. Kanis, and D. Felsenberg, Influence of hormonal and reproductive factors on the
MELTON AND COOPER
50.
51.
52.
53. 54.
55.
56.
57.
58.
59.
60. 61.
62. 63. 64.
65.
66.
67.
68. 69.
70.
risk of vertebral deformity in European women: European Vertebral Osteoporosis Study Group. Osteoporos. Int. 7, 72 – 78 (1997). L. J. Melton III, E. J. Atkinson, S. Khosla, W. M. O’Fallon, and B. L. Riggs, Secondary osteoporosis and the risk of vertebral deformities in women. Bone 24, 49 – 55 (1999). M. C. Nevitt and S. R. Cummings, Type of fall and risk of hip and wrist fractures: The Study of Osteoporotic Fractures Research Group. J. Am. Geriatr. Soc. 41, 1226 – 1234 (1993). W. C. Graafmans, M. E. Ooms, P. D. Bezemer, L. M. Bouter, and P. Lips, Different risk profiles for hip fractures and distal forearm fractures: A prospective study. Osteoporos. Int. 6, 427 – 431 (1996). C. Cooper, Osteoporosis: An epidemiological perspective: A review. J. R. Soc. Med. 82, 753 – 757 (1989). S. J. Jacobsen, D. J. Sargent, E. J. Atkinson, W. M. O’Fallon, and L. J. Melton III, Contribution of weather to the seasonality of distal forearm fractures: A population-based study in Rochester, Minnesota. Osteoporos. Int. 9, 254 – 259 (1999). D. G. Seeley, W. S. Browner, M. C. Nevitt, H. K. Genant, J. C. Scott, and S. R. Cummings, Which fractures are associated with low appendicular bone mass in elderly women? The Study of Osteoporotic Fractures Research Group. Ann. Intern. Med. 115, 837 – 842 (1991). S. H. Rose, L. J. Melton III, B. F. Morrey, D. M. Ilstrup, and B. L. Riggs, Epidemiologic features of humeral fractures. Clin. Orthop. 168, 24 – 30 (1982). B. Kristiansen, G. Barfod, J. Bredesen, J. Erin-Madsen, B. Grum, M. W. Horsnaes, and J. R. Aalberg, Epidemiology of proximal humeral fractures. Acta Orthop. Scand. 58, 75 – 77 (1987). J. L. Kelsey, W. S. Browner, D. G. Seeley, M. C. Nevitt, and S. R. Cummings, Risk factors for fractures of the distal forearm and proximal humerus: The Study of Osteoporotic Fractures Research Group. Am J. Epidemiol. 135, 477 – 489 (1992). L. J. Melton III, J. M. Sampson, B. F. Morrey, and D. M. Ilstrup, Epidemiologic features of pelvic fractures. Clin. Orthop. 155, 43 – 47 (1981). A. J. Buhr and A. M. Cooke, Fracture patterns. Lancet 1, 531 – 536 (1959). U. Bengnér, O. Johnell, and I. Redlund-Johnell, Increasing incidence of tibia condyle and patella fractures. Acta Orthop. Scand. 57, 334 – 336 (1986). E. Foltin, Bone loss and forms of tibial condylar fracture. Arch. Orthop. Trauma Surg. 106, 341 – 348 (1987). R. Hedlund and U. Lindgren, Epidemiology of diaphyseal femoral fracture. Acta Orthop. Scand. 57, 423 – 427 (1986). T. J. Arneson, L. J. Melton III, D. G. Lewallen, and W. M. O’Fallon, Epidemiology of diaphyseal and distal femoral fractures in Rochester, Minnesota, 1965 – 1984. Clin Orthop. 234, 188 – 194 (1988). L. Forsén, A. J. Søgaard, H. E. Meyer, T.-H. Edna, and B. Kopjar, Survival after hip fracture: Short- and long-term excess mortality according to age and gender. Osteoporos. Int. 10, 73 – 78 (1999). G. Poór, E. J. Atkinson, W. M. O’Fallon, and L. J. Melton III, Determinants of reduced survival following hip fractures in men. Clin. Orthop. 319, 260 – 265 (1995). M. T. Clayer and R. J. Bauze, Morbidity and mortality following fractures of the femoral neck and trochanteric region: Analysis of risk factors. J. Trauma 29, 1673 – 1678 (1989). G. Poór, S. J. Jacobsen, and L. J. Melton III, Mortality following hip fracture. Facts Res. Gerontol. 7, 91 – 109 (1994). C. Cooper, E. J. Atkinson, S. J. Jacobsen, W. M. O’Fallon, and L. J. Melton III, Population-based study of survival following osteoporotic fractures. Am. J. Epidemiol. 137, 1001 – 1005 (1993). D. M. Kado, W. S. Browner, L. Palermo, M. C. Nevitt, H. K. Genant, and S. R. Cummings, Vertebral fractures and mortality in older women: A prospective study. Arch. Intern. Med. 159, 1215 – 1220 (1999).
21 Magnitude and Impact of Osteoporosis and Fractures
567
71. W. S. Browner, D. G. Seeley, T. M. Vogt, and S. R. Cummings, Nontrauma mortality in elderly women with low bone mineral density: Study of Osteoporotic Fractures Research Group. Lancet 338, 355 – 358 (1991). 72. C. Johansson, D. Black, O. Johnell, A. Odén, and D. Mellström, Bone mineral density is a predictor of survival. Calcif. Tissue Int. 63, 190 – 196 (1998). 73. P. von der Recke, M. A. Hansen, and C. Hassager, The association between low bone mass at the menopause and cardiovascular mortality. Am. J. Med. 106, 273 – 278 (1999). 74. E. A. Chrischilles, C. D. Butler, C. S. Davis, and R. B. Wallace, A model of lifetime osteoporosis impact. Arch. Intern. Med. 151, 2026 – 2032 (1991). 75. W. Miller, Survival and ambulation following hip fracture. J. Bone Joint. Surg. 60A, 930 – 934 (1978). 76. J. S. Jensen and J. Bagger, Long-term social prognosis after hip fractures. Acta Orthop. Scand. 53, 97 – 101 (1982). 77. S. K. Bonar, M. E. Tinetti, M. Speechley, and L. M. Cooney, Factors associated with short- versus long-term skilled nursing facility placement among community-living hip fracture patients. J. Am. Geriatr. Soc. 38, 1139 – 1144 (1990). 78. N. F. Ray, J. K. Chan, M. Thamer, and L. J. Melton III, Medical expenditures for the treatment of osteoporotic fractures in the United States in 1995: Report from the National Osteoporosis Foundation. J. Bone Miner. Res. 12, 24 – 35 (1997). 79. T. L. Holbrook, K. Grazier, J. L. Kelsey, and R. N. Stauffer, “The Frequency of Occurrence, Impact and Cost of Selected Musculoskeletal Conditions in the United States.” American Academy of Orthopedic Surgeons, Chicago, 1984. 80. P. D. Ross, J. W. Davis, R. S. Epstein, and R. D. Wasnich, Pain and disability associated with new vertebral fractures and other spinal conditions. J. Clin. Epidemiol. 47, 231 – 239 (1994). 81. G. Leidig, H. W. Minne, P. Sauer, C. Wuster, M. Logen, F. Raue, and R. Ziegler, A study of complaints and their relation to vertebral destruction in patients with osteoporosis. Bone Miner. 8, 217 – 229 (1990). 82. W. M. Garraway, R. N. Stauffer, L. T. Kurland, and W. M. O’Fallon, Limb fractures in a defined population. II. Orthopedic treatment and utilization of health care. Mayo Clin. Proc. 54, 708 – 713 (1979). 83. J. A. Kanis and F. A. Pitt, Epidemiology of osteoporosis. Bone 13 (Suppl. 1), S7 – S15 (1992). 84. R. M. Atkins, T. Duckworth, and J. A. Kanis, Features of algodystrophy following Colles fracture. J. Bone Joint Surg. 72B, 105 – 110 (1990). 85. H. P. de Bruijn, The Colles fracture, review of literature. Acta Orthop. Scand. 58 (Suppl. 223), 7 – 25 (1987). 86. J.-P. Kaukonen, E. O. Karaharju, M. Porras, P. Lüthje, and A. Jakobsson, Functional recovery after fractures of the distal forearm. Ann. Chir. Gynaecol. 77, 27 – 31 (1988). 87. A. Praemer, S. Furner, and D. P. Rice, “Musculoskeletal Conditions in the United States.” American Academy of Orthopaedic Surgeons, Park Ridge, IL, 1992. 88. S. Eiskjaer, S. E. Østgård, B. W. Jakobsen, J. Jensen, and U. Lucht, Years of potential life lost after hip fracture among postmenopausal women. Acta Orthop. Scand. 63, 293 – 296 (1992). 89. K. Lippuner, J. von Overbeck, R. Perrelet, H. Bosshard, and Ph. Jaeger, Incidence and direct medical costs of hospitalizations due to osteoporotic fractures in Switzerland. Osteoporos. Int. 7, 414 – 425 (1997). 90. Royal College of Physicians of UK. “Guidelines for the Prevention and Treatment of Osteoporosis.” Royal College of Physicians of UK, London, 1999. 91. E. Levy, Cost analysis of osteoporosis related to untreated menopause. Clin. Rheumatol. 8 (Suppl. 2), 76 – 82 (1989). 92. B. Gullberg, O. Johnell, and J. A. Kanis, World-wide projections for hip fracture. Osteoporos. Int. 7, 407 – 413 (1997). 93. L. J. Melton III, E. J. Atkinson, and R. Madhok, Downturn in hip fracture incidence. Public Health Rep. 111, 146 – 150 (1996).
94. T. Naessen, R. Parker, I. Persson, M. Zack, and H. O. Adami, Time trends in incidence rates of first hip fracture in the Uppsala Health Care Region, Sweden, 1965 – 1983. Am. J. Epidemiol. 130, 289 – 299 (1989). 95. L. Rehnberg, S. Nungu, and C. Olerud, The incidence of femoral neck fractures in women is decreasing. Acta Orthop. Scand. 63 (Suppl. 248), 92 – 93 (1992). 96. T. M. Huusko, P, Karppi, V. Avikainen, H. Kautiainen, and R. Sulkava, The changing picture of hip fractures: Dramatic change in age distribution and no change in age-adjusted incidence within 10 years in central Finland. Bone 24, 257 – 259 (1999). 97. C. Rogmark, I. Sernbo, O. Johnell, and J.-Å. Nilsson, Incidence of hip fractures in Malmö, Sweden, 1992 – 1995: A trend-break. Acta Orthop. Scand. 70, 19 – 22 (1999). 98. T. D. Spector, C. Cooper, and A. F. Lewis, Trends in admissions for hip fracture in England and Wales, 1968 – 85. Br. Med. J. 300, 1173 – 1174 (1990). 99. E. M. Lau and C. Cooper, The epidemiology of osteoporosis: The Oriental perspective in a world context. Clin. Orthop. 323, 65 – 74 (1996). 100. R. Hoffenberg, O. F. W. James, J. C. Brocklehurst, I. D. Green, P. Horracks, J. A. Kanis, N. J. Wald, G. E. MacLellan, and R. H. Vickers, Fractured neck of femur: Prevention and management. Summary and recommendations of a report of the Royal College of Physicians. J. R. Coll. Phys. London 23, 8 – 12 (1989). 101. S. R. Lord and P. F. Sinnett, Femoral neck fractures: Admissions, bed use, outcome and projections. Med. J. Aust. 145, 493 – 496, (1986). 102. P. Kannus, S. Niemi, J. Parkkari, M. Palvanen, I. Vuori, and M. Järvinen, Hip fractures in Finland between 1970 and 1997 and predictions for the future. Lancet 353, 802 – 805 (1999). 103. E. A. Papadimitropoulos, P. C. Coyte, R. G. Josse, and C. E. Greenwood, Current and projected rates of hip fracture in Canada. Can. Med. Assoc. J. 157, 1357 – 1363 (1997). 104. K. J. Obrant, U. Bengnér, O. Johnell, B. E. Nilsson, and I. Sernbo, Increasing age-adjusted risk of fragility fractures: A sign of increasing osteoporosis in successive generations? Calcif. Tissue Int. 44, 157 – 167 (1989). 105. L. J. Melton III, W. M. O’Fallon, and B. L. Riggs, Secular trends in the incidence of hip fractures. Calcif. Tissue Int. 41, 57 – 64 (1987). 106. C. Cooper, D. J. P. Barker, and C. Wickham, Physical activity, muscle strength and calcium intake in fracture of the proximal femur in Britain. Br. Med. J. 297, 1443 – 1446 (1988). 107. E. Lau, S. Donnan, D. J. P. Barker, and C. Cooper, Physical activity and calcium intake in fracture of the proximal femur in Hong Kong. Br. Med. J. 297, 1441 – 1443 (1988). 108. C. Wickham, K. Walsh, C. Cooper, D. J. Barker, B. M. Margetts, J. Morris, and S. A. Bruce, Dietary calcium, physical activity, and risk of hip fracture: A prospective study. Br. Med. J. 299, 889 – 892 (1989). 109. E. W. Gregg, J. A. Cauley, D. G. Seeley, K. E. Ensrud, and D. C. Bauer, Physical activity and osteoporotic fracture risk in older women: Study of Osteoporotic Fractures Research Group. Ann. Intern. Med. 129, 81 – 88 (1998). 110. C. Wickham, C. Cooper, B. M. Margetts, and D. J. P. Barker, Muscle strength, activity, housing and the risk of falls in elderly people. Age Ageing 18, 47 – 51 (1989). 111. I. R. Reid, K. Chin, M. C. Evans, and J. G. Jones, Relation between increase in length of hip axis in older women between 1950s and 1990s and increase in age specific rates of hip fracture. Br. Med. J. 309, 508 – 509 (1994). 112. J. G. Evans, V. Seagroatt, and M. J. Goldacre, Secular trends in proximal femoral fracture, Oxford record linkage study area and England, 1968 – 86. J. Epidemiol. Community Health 51, 424 – 429 (1997). 113. A. McColl, P. Roderick, and C. Cooper, Hip fracture incidence and mortality in an English region: A study using routine National Health Service data. J. Public Health Med. 20, 196 – 205 (1998).
CHAPTER
569
CHAPTER 22 Race, Ethnicity, and Osteoporosis
CHAPTER 22
CHAPTER 22
Race, Ethnicity, and Osteoporosis MARIE LUZ VILLA LORENE NELSON DOROTHY NELSON
Department of Medicine, Division of Gerontology and Geriatrics, University of Washington School of Medicine, Seattle, Washington 98104 Department of Health Research and Policy, Division of Epidemiology, Stanford University School of Medicine, Stanford, California 94305 Department of Internal Medicine, Wayne State University, Detroit, Michigan 48201
IV. Racial and Ethnic Influences on Risk for Osteoporosis V. Summary References
I. We Are All Individuals II. Defining Terms III. Ethnoepidemiology of Osteoporosis
I. WE ARE ALL INDIVIDUALS Marked differences exist between groups that can be sorted according to descriptors such as race or ethnicity. Within racial and ethnic groups, large degrees of individual variation exist; this chapter attempts to address observed trends in bone dynamics that can be sorted by race or ethnic grouping. Sorting of humans, however, suffers from the inconstancy of racial and ethnic definitions. This chapter therefore defines terms used to describe human groups and then explores their application to the epidemiology of hip fracture and parameters that affect skeletal health. The diagnostic term “osteoporosis” refers to a skeletal condition that predisposes to multisite fragility fractures. Because only hip fractures reliably lead to hospitalization
OSTEOPOROSIS, SECOND EDITION VOLUME 1
and therefore documentation, we present hip fracture data in various populations as a reflection of morbidity resulting from osteoporosis.
II. DEFINING TERMS A. Race and Ethnicity “Ethnicity” and “race” appear interchangeably in many publications. “Race” in the United States reflects the belief that a limited number of genetically characterized human groups exist, exemplified by the list used by the U.S. Census: White/Caucasian, Black/African American, Native American/American Indian, Alaskan native/Eskimo/Aleut,
569
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
570
VILLA, NELSON, AND NELSON
Asian/Pacific Islander, and other (Spanish or Hispanic origin is asked separately). Most investigators recognize, however, that distinct racial lines may not be drawn due to significant genetic admixture that has occurred over time. Also, because environments change, and because populations move and interbreed, it is difficult if not impossible to identify discrete, biologically meaningful subgroups of humans. A factor that reflects cultural, religious, dietary, geographic, and other differences among races, known as ethnicity, then becomes important [1]. In fact, ethnicity plays an important role in disease prevalence even within races. For example, Hispanic Caucasians show different trends in disease incidence when compared to non-Hispanic Caucasians (an ethnic dichotomization of the Caucasian race specific to the Americas): Mexican Americans have a twoto fivefold greater risk of developing noninsulin dependent diabetes mellitus (NIDDM) than does the majority of the U.S. population [2]. Lack of ethnic definition of study groups affects the general applicability of data. A study reporting hip fracture rates of “Asians” does not help determine a Korean woman’s risk of suffering a hip fracture. Ethnic-specific data would be more valuable than a broad, racial summarization because bone mineral densities (BMD) and fracture rates vary among countries, as well as among ethnic subgroups. In a study comparing the average BMD of Japanese, Korean, and Taiwanese women, the Taiwanese had consistently greater BMD at the lumbar spine at almost every age [3]. In a study of ethnic/racial BMD differences among children, a large subgroup of “Whites” considered themselves Chaldean, an Iraqi ethnic group [4]. The Chaldean children’s whole body bone mass was significantly higher than– non-Chaldean White children and was not different from other study subjects who considered themselves Black. Because Middle Easterners are included in the U.S. Census category “White/Caucasian,” such a difference would not be expected a priori and would affect the results of the study if the Chaldeans were analyzed together with other White children.
B. Acculturation In addition to race and ethnicity, another factor known as acculturation contributes to nuances in measured variables. Acculturation scales measure how much an ethnic group assimilates the language, habits, and cultural values of the country or area to which it migrates [5 – 7]. Thus, a higher acculturation score means greater adoption of the dominant culture of a region. Degree of acculturation
may also affect dietary and lifestyle habits of people inhabiting a given region for many years: African Americans preserve many distinct customs and dietary habits when moving to regions outside the southern United States [8]. Returning to the example of ethnic differences in the incidence of diabetes, the prevalence of NIDDM varies within Hispanic ethnic groups, depending on the degree of acculturation: for Mexican Americans, the rate of NIDDM decreases with increasing acculturation [9]. Sometimes acculturation may serve as a proxy for factors that affect risk for osteoporosis but are themselves not easily measured, as the environment in which a group is born or raised can affect variance in observed fracture rates [10]. Hip fracture incidence for elderly Japanese women ranges from 450 to 1011 per 100,000 person-years/year [11], depending on their region of origin. Tools for retrospectively assessing fracture risk factors in elders suffer from substantial inaccuracy. If, in addition, the study group originated in another country or was reared under very different cultural conditions than the group for which the tool was validated, even greater measurement error may be introduced. Retrospectively recalled teenage physical activity and milk intake have no relationship to lumbar bone mass of postmenopausal Mexican American women, yet lumbar BMD is independently and positively related to acculturation, even when controlling for age and obesity [12]. Assessment of ethnicity and acculturation helps describe disease occurrence among human groups, contributing greatly to the description of factors that may affect observed differences in fracture rates. For this reason, as well as the pitfalls associated with “race” described earlier, we attempt to use “ethnic” instead of “racial” in most of the contexts that follow. However, data frequently are reported using “racial” categorizations without indication of ethnic grouping. We present those data as reported.
III. ETHNOEPIDEMIOLOGY OF OSTEOPOROSIS Great variation in the occurrence of osteoporotic fractures exists within and among different racial and ethnic groups. Some studies report wide ranges in hip fracture incidence rates within a given racial group, probably due to regional and/or cultural (ethnic and lifestyle) factors. (Geographic variation in fracture rates is discussed elsewhere in the text.) Fracture incidence rates in Caucasians vary among ethnic groups, as demonstrated in Table 1. The magnitude of these within-Caucasian differences approaches that noted between Caucasians and other racial groups [13 – 15].
571
CHAPTER 22 Race, Ethnicity, and Osteoporosis
A. Methodologic Issues We reviewed information on hip fracture incidence rates obtained from studies conducted among different racial and ethnic groups. Prior to summarizing these data, several methodologic issues are discussed that have important bearing on the ability to compare incidence rates among studies, including the need for age and sex standardization
to adjust for differences in the age and gender composition of the populations under study, differences between studies in definitions of hip fracture; and differences between studies in the methods that were used to identify individuals with hip fracture. Several recent articles provide insights into other methodologic issues that may also affect the ability make cross-national comparisons of hip fracture incidence rates [13,24,25,49].
TABLE 1 Age-Adjusted Ratesa of Hip Fracture per 100,000 Population for Females, Males, and Totalb and Year of Study Ethnic group Blacks
Hispanicsc Asians
Caucasiansd
Site (reference)
Years of study
Female
Male
Total
Female:male
USA [16]
1986 – 1989
214
179
200
1.2
Maryland [17]
1979 – 1988
345
191
283
1.8
USA [18]
1984 – 1985
344
235
300
1.5
California [19]
1983 – 1984
241
153
202
1.6
Texas [20]
1980
243
13
141
18.7
USA [14]
1974 – 1979
174
108
137
1.6
Johannesburg, South Africa [21]
1950 – 1964
26
20
23
1.3
California [19]
1983 – 1984
219
97
165
2.3
Texas [20]
1980
305
128
227
2.4
Tottori, Japan [22]
1994
342
136
249
2.5
Tottori, Japan [23]
1986 – 1987
227
79
163
2.9
Okinawa, Japan [11]
1984 – 1985
325
86
219
3.8
Beijing, China [24]
1990 – 1992
97
101
99
1.0
Beijing, China
1990 – 1992
96
107
**
0.9
Hong Kong [25]
1990 – 1992
428
270
**
1.4
Hong Kong [26]
1985
389
196
304
2.0
Hong Kong [27]
1965 – 1967
179
113
150
1.6
Singapore [28]
1955 – 1962
83
111
95
0.7
Kuwait [29]
1992 – 1995
378
279
333
1.4
New Zealand [30]
1973 – 1976
212
121
172
1.8
California [19]
1983 – 1984
383
116
265
3.3
Hawaii [11]
1979 – 1981
224
66
153
3.4
Sweden [31]
1985
714
268
517
2.7
Sweden [32]
1980
432
199
517
2.7
Sweden [33]
1980
984
338
705
2.9
Sweden [34]
1972 – 1981
714
319
540
2.2
Sweden [35]
1972 – 1981
730
581
664
1.3
Malmo, Sweden [36]
1950 – 1960
468
153
329
3.1
Norway [37]
1983 – 1984
737
298
543
2.5
Oslo, Norway [15]
1978 – 1979
850
329
620
2.6
1970
377
142
273
2.7
1990 – 1992
697
349
**
2.0 2.6
Finland [32] Reykjavik [25] Kuopio, Finland [38] Edin, Scotland [39] Oxford, England [40] Yorkshire, UK [41]
1968
280
107
204
1978 – 1979
529
174
376
3.0
1983
603
114
392
5.3
1973 – 1977
310
102
218
3.0 (continues)
572
VILLA, NELSON, AND NELSON TABLE 1
Ethnic group
(continued)
Site (reference)
Years of study
Alicante, Spain [42]
1974 – 1984
90
57
75
1.6
Italy [43]
1988 – 1989
287
110
207
2.6
Jerusalem, Israel [44]
1957 – 1966
355
168
272
2.1
Canada [45]
1976 – 1985
788
307
595
2.6
USA [16]
1986 – 1989
968
396
738
2.4
USA [18]
1984 – 1985
845
350
645
2.4
California [19]
1983 – 1984
617
215
439
2.9
Texas [20]
Female
Male
Total
Female:male
1980
593
223
430
2.7
1979 – 1988
950
358
712
2.7
Hawaii [11]
1979 – 1981
645
205
451
3.1
Minnesota [11]
1978 – 1982
613
285
468
2.2
USA [14]
1974 – 1979
422
151
285
2.8
USA [46]
1970 – 1983
705
244
506
2.9
Rochester [47]
1965 – 1974
559
191
396
2.9
New Zealand [30]
1973 – 1976
466
139
321
3.4
Australia [48]
1994 – 1996
575
244
425
2.4
Maryland [17]
a
Rates were age and gender adjusted to the 1990 U.S. non-Hispanic Caucasian population. Both age and gender adjusted. c Hispanic Caucasians. d Non-Hispanic Caucasians. b
1. NEED FOR AGE AND SEX STANDARDIZATION Because the number of elderly in the world’s population is increasing rapidly with time, and because studies differ with respect to the age and gender composition of the populations under study, hip fracture incidence rates obtained from different time periods and from different populations are not strictly comparable unless the age and gender differences between study populations have been taken into account. A method called standardization is used as a means to provide an estimate of the incidence rate in a given population if that population had the same gender and age composition as an arbitrarily selected standard population. The 1990 U.S. non-Hispanic Caucasian population served as the standard population for the age- and sex-adjusted incidence rates for hip fracture that are presented in Table 1. The studies summarized in Table 1 all contained information regarding age- and sex-specific incidence rates of hip fracture for individuals above 50 years of age so that age- and sex-adjusted incidence rates could be calculated. The differences among study populations in adjusted hip fracture incidence rates that exist after standardization are unlikely to be attributed to differences in age or gender composition that exist between the study populations. Of note, however, many studies treat individuals aged 80 and older as one group. Because fracture incidence rises steeply with age, the standardization process cannot adequately adjust for age if studies do not provide enough detailed data for the older age groups.
2. DIFFERENCES IN THE DEFINITION OF HIP FRACTURE Studies differ with respect to the amount of detail provided regarding the exact anatomic locations of the fractures. Each study summarized in Table 1 used one of the following definitions of hip fracture: (1) fracture of the femoral neck or proximal femur; (2) cervical, trochanteric, intracapsular, extracapsular, or intertrochanteric fracture; (3) hip fracture defined on the basis of International Classification of Disease (ICD) codes; or (4) hip fracture with no specification of fracture location. A small percentage of hip fractures result from severe trauma or from underlying pathology. In countries where hip fractures are uncommon, fractures due to severe trauma account for a larger proportion of the total number of hip fractures [50]. While some of the studies in Table 1 excluded fractures due to severe trauma, tumors, or metabolic bone diseases, others did not specify that these exclusions had occurred. 3. DIFFERENCES IN CASE ASCERTAINMENT METHODS Most studies of hip fracture suffer some degree of underascertainment due to the difficulties of identifying every person with hip fracture. All studies included in Table 1 ascertained cases of hip fracture through hospital records, usually through hospital discharge diagnoses. Reasons for underascertainment are that some fractures are misclassified as femoral shaft fractures, and some individuals with hip fracture are not hospitalized either because health services are not available or because they are treated in
CHAPTER 22 Race, Ethnicity, and Osteoporosis
another setting (i.e., at home, in a chronic care facility, or by a native healer). Although the latter sources of bias differ substantially from one country or study to another, they may not contribute substantially to fracture statistics in those countries listed in Table 1.
B. Racial and Ethnic Differences in Rates of Hip Fracture Despite methodologic difficulties that affect the comparison of hip fracture rates among studies, broad conclusions can be drawn regarding differences in hip fracture incidence rates for members of different races and ethnic groups. Age- and sex-adjusted incidence rates of hip fracture in Blacks, Hispanic Caucasians, Asian or Pacific Islanders, and non-Hispanic Caucasians are presented in Table 1. Within each race or ethnic group and each country or geographic region, studies are arranged in order of most recent to least recent so that cross-ethnic comparisons can be made between studies that have been conducted in similar time periods. In addition, many authors do not give detailed information about the racial and ethnic backgrounds of groups studied so data presented reflect use of ethnic and racial categories as published. Since few studies prior to 1980 were conducted in groups other than non-Hispanic Caucasians, our discussion will be largely focused on studies conducted in the 1980s and 1990s. Caucasians have the highest hip fracture incidence rates of any race or ethnic group, and this is particularly striking in northern Europe and North America. Studies in the United States demonstrate that hip fracture incidence in Asian Americans is intermediate to those of non-Hispanic Caucasians and Blacks. Although the number of studies in Hispanics is small, estimates of hip fracture incidence in this group are close to (and in some cases lower than) rates among Black Americans [19 – 21]. Hip fracture incidence among Black South Africans was reported to be very low in a study conducted between 1950 and 1964 [21], but there are no recent studies in African populations. Studies conducted in racially diverse populations using the same methodology for ascertaining hip fractures in all groups are particularly valuable for making inferences about racial differences in hip fracture incidence [16 – 20,30]. A cross-national study carried out in five geographic areas in the years 1990 – 1992 used similar methodologies in all areas and corrected for methodologic differences between studies [25]. The highest hip fracture incidence rates occurred in Iceland, intermediate rates in Hungary and Hong Kong, and lowest rates in Beijing. Other studies found that hip fracture incidence in Beijing and Singapore appears to be much lower than rates in Hong Kong or other Asian countries; reasons for these differences are not understood [24,28]. Another cross-national study re-
573 lied solely on hospital discharge diagnoses to identify hip fracture cases [49]. The highest rates were reported for the European and North American countries, and the lowest fractures rates were observed in Venezuela and Chile; however, sole reliance on hospital discharge data likely resulted in the underascertainment of hip fracture cases [25]. Many studies of racially diverse populations have been conducted in the United States and they have consistently reported higher rates among Caucasians than among other racial and ethnic groups. In a study conducted in Bexar County, Texas, age- and sex-adjusted hip fracture incidence was lowest in Blacks and highest in non-Hispanic Caucasians, with intermediate rates for Hispanics [20]. Studies in the United States have utilized Medicare data from the Health Care Financing Administration to estimate hip fracture incidence rates among elderly individuals (ages 65 and older). These studies have confirmed that U.S. Hispanic rates are intermediate between the lower Black rates and the higher Caucasian rates [51]. In contrast, Silverman et al. [19] found that age- and sex-adjusted incidence of hip fracture in California was lower among Hispanic Caucasians than among all other groups, including non-Hispanic Caucasians, Blacks, and Asians. A study comparing hip fracture incidence among native Japanese, Japanese Americans, and non-Hispanic Caucasian Americans [11] reported the lowest rates among Japanese Americans and the highest rates among non-Hispanic Caucasians. The analysis of Medicare data for individuals ages 65 and older found that age-adjusted hip fracture incidence rates were lower for all Asian groups (Chinese, Japanese, and Korean Americans) than for Caucasians [52]. Also of interest are racial and ethnic differences with respect to the magnitude of the female:male ratio of hip fracture incidence rates, which usually exceeds two in non-Hispanic Caucasian populations. Also of interest are racial and ethnic differences with respect to the magnitude of the female:male ratio of hip fracture incidence rates. Among Black Americans, hip fracture is more common among women than men; however, the female:male ratio is usually below two in contrast to non-Hispanic Caucasian populations where the gender ratio usually exceeds two. In the studies conducted in Hispanic populations, the gender ratio is close to two [19,20]. All studies but two [24,28] in Asian populations demonstrate a higher incidence of hip fracture in Asian women than men, with two studies reporting female:male ratios that exceed 3.0 [11,19]. Increasing age is an established risk factor for hip fracture in all racial and ethnic groups. An excellent review of differences in age-specific incidence rates of hip fracture between racial and ethnic groups is provided by Maggi et al. [13]. Although hip fracture incidence increases with age in all ethnic groups, the increase occurs earlier in non-Hispanic Caucasian populations than in
574
VILLA, NELSON, AND NELSON
Black, Asian, and Hispanic populations [13]. Studies conducted in non-Hispanic Caucasian populations report higher rates of hip fracture among men than women before 50 years of age, whereas after age 50, women have higher rates than men. Environmental factors such as diet, level of physical activity, frequency of cigarette smoking, and use of hormonal medications may explain some of the differences in hip fracture incidence observed between racial and ethnic groups. Factors that may contribute to racial and ethnic differences in skeletal health and risk for hip fracture are discussed in the remainder of this chapter.
IV. RACIAL AND ETHNIC INFLUENCES ON RISK FOR OSTEOPOROSIS Many factors affect the risk of developing osteoporosis or suffering nontraumatic hip fracture, as described elsewhere in this volume. It is not clear that all populations are similarly characterized with respect to osteoporotic risk factors, as most were established from studies of non-Hispanic Caucasians. It does seem intuitive that most people should respond similarly to factors such as reproductive hormone status, medication use, and physical activity. Other factors, such as calcium metabolism, bone mass, and body composition, may have different effects from one ethnic group to the next.
A. Bone Mass The term bone mass can refer to a variety of measurements, including bone mineral content (BMC in g), areal bone mineral density (BMD in g/cm2), and volumetric bone density (BMD in g/cm3). The degree of ethnic differences in bone mass reported by various investigators varies with the measurement used as well as other factors. American Blacks have significantly greater areal bone mass than Caucasians [53 – 59], which is thought to contribute to their lower rate of hip fracture. Kleerekoper and colleagues have shown that volumetric BMD measured by quantitative computed tomography (QCT) is 40% higher in African American compared with White women [60], considerably greater than the 5 – 15% difference in areal BMD generally reported for African American versus Caucasian adults. It is not well understood whether racial differences in bone mass exist at birth or develop at some point thereafter. Most studies based on absorptiometry (SPA, DXA) find higher bone mass (bone mineral content or areal bone density) in African Americans throughout childhood [61 – 66]. Some studies of volumetric bone density (g/cm3),
based on QCT, find no distinction in skeletal status between young Black and Caucasian children [67,68]. Gilsanz and colleagues [69] examined bone mineral density in Black and non-Hispanic Caucasian children at different stages of sexual development and found that significant racial differences did not occur until late puberty. In contrast, recent investigations describe significant differences in volumetric (BMAD) femoral neck bone density, based on DXA, at all stages of puberty [70,71]. It has been hypothesized that the higher bone mass seen in North American Blacks stems in part from genetic factors. However, the U.S. Blacks’ gene pool is very heterogeneous and is the result of much admixture over several centuries. It might be assumed that any population of African origin would have a high bone mass similar to African Americans, but this has not been borne out. Investigations of Blacks in South Africa [53,72 – 74] and the Gambia [67,75] showed that their bone mass does not exceed and, in some cases, is lower than that of age-matched African Caucasians. These data illustrate the difficulty in generalizing about a “racial” group, when obviously ethnic gradations in bone mass exist within people of African descent, with further differences introduced by acculturation in areas to which Black Africans migrated. The positive relationship between high bone mass and low incidence of hip fracture is seen in New Zealand Polynesians [30,76]. However, Hispanics have hip fracture incidence rates comparable to those of African Americans, yet have bone mass values closer to those of non-Hispanic Caucasians [12,58,77,78]; Asian Americans demonstrate a similar relationship [52,79 – 81]. One South American ethnic group in Vilcabamba, Ecuador, enjoys an extremely low rate of hip fracture despite bone mineral density values much lower than that of non-Hispanic Caucasians [82]. Few studies exist investigating bone health in American Indian tribes [83 – 86]. Data presented in abstract form by Chen and colleagues [87] showed BMD of Native Americans residing in the Southwest to be significantly lower than that of Whites, although the study groups were not well matched for age. Unfortunately, hip fracture studies entirely exclude Native Americans so that relationships between bone mass and fracture risk cannot be drawn. Bone mass therefore may not be the factor that best predicts fracture risk in many racial and ethnic groups, rendering World Health Organization guidelines for diagnosis of osteoporosis [88] only narrowly applicable.
B. Bone Turnover Risk for osteoporotic fracture depends not only on the mass of bone, but its quality as well. The rate and efficiency of bone turnover (affected by reproductive hormone status,
575
CHAPTER 22 Race, Ethnicity, and Osteoporosis
body composition, vitamin D/calcium nutriture, and physical activity) affects bone architecture, which is an essential component of skeletal strength. Low bone turnover in Black adults may partially explain their greater lifelong bone mass and lower fracture risk than non-Hispanic Caucasians. Some studies found biochemical evidence that African Americans have lower rates of bone turnover than non-Hispanic Caucasians [60,89,90]. In addition, histomorphometric studies showed mean rates of bone formation in American Blacks to be significantly lower than those of non-Hispanic Caucasians [90,91]. These authors concluded that most, if not all, ethnic differences observed in bone cell function could be the result of differences in bone accumulation during growth: higher bone mass would result in less fatigue damage and less need for repair by directed bone remodeling. However, studies comparing South African Blacks and Caucasians suggested higher bone turnover in Blacks, which was hypothesized to lead to fewer fractures because of better trabecular bone quality and less skeletal fragility [92,93]. Evidence of no black/white differences in bone turnover or mass has been presented as well [94]. These contrasting studies again highlight the pitfalls associated with assuming that subgroups (such as geographically different populations) of a “racial” group will be biologically similar. No studies have directly compared bone turnover of Asians, Blacks, and Caucasians, but it appears that the reported normal values for circulating osteocalcin in Japanese women are lower than those of non-Hispanic Caucasian women [95,96]. However, Polynesians and Caucasians do not manifest different serum concentrations of osteocalcin or parathyroid hormone (PTH), or urinary excretion of hydroxyproline, despite significant differences in BMD [97]. In one study of young Mexican American and non-Hispanic Caucasians, osteocalcin concentrations did not differ significantly between the two groups, despite differences in 25hydroxyvitamin D and PTH values [98]. Differences in reproductive hormone status may contribute to ethnic and racial variation in bone turnover, skeletal quality, and subsequent fracture risk. Androgens and estrogens contribute positively and independently to attainment of peak bone mass [99,100], and adult bone loss often stems from the increased bone turnover associated with decreased levels of reproductive hormones [95]. Furthermore, serum unbound sex steroid concentrations are lower in women with hip fracture than in controls [101]. Gilsanz and colleagues found that racial differences in bone mass develop during late stages of puberty (69), perhaps related to differences in serum sex hormone status. Bone loss in Japanese women appears to be greatest in the early postmenopausal period, but subsequently declines at rates similar to those for non-Hispanic Caucasians [95,102]. In a large study of postmenopausal women con-
ducted in the northeastern United States, race and serum estrone concentrations contributed independently to observed racial differences in bone mass [103]. Serum estrone values were significantly higher in African Americans, but this difference disappeared when analyses were adjusted for obesity (as determined by body mass index 27.3 kg/m2).
C. Body Size and Composition Body size appears to be an independent contributor to variance in BMD. Therefore, use of a mathematical correction for differences in densitometric bone size [104] from one population to another might correct for differences in body habitus and shed some light on the seeming discrepancies between bone mass and fracture risk across ethnic groups. In a multisite study of hormone replacement and its effects on bone mass in postmenopausal women, it was noted that although African Americans had the highest measured bone mass, when adjustments were made for bone size, the ethnic differences in bone density were significantly attenuated [55]. Analysis of Eskimo bone mineral content showed it to be lower than that of non-Hispanic Caucasians, but their low rate of fracture is thought to be due to relatively larger bone size [84]. Despite the widely accepted axiom that Asians have lower bone mass than non-Hispanic Whites, a recent comparison of closely matched non-Hispanic White and Chinese women found slightly higher bone mass in the Chinese when height and weight (and theoretically differences in bone size) were controlled [105]. Data suggest that differences in bone accumulation in multiethnic teenagers may predominantly reflect bone size [70,106]. Body weight factors importantly in the maintenance of bone density, and thinness is an important risk factor for hip fracture in Black, Caucasian, and Asian women [107 – 109]. However, it appears that both fat and lean body mass contribute to preservation of the skeleton [110,111], perhaps due in part to peripheral aromatization of androgen to estrogen that occurs in adipose tissue and skeletal muscle [101,112]. Serum estrone concentrations relate positively to degree of obesity, and bone mass correlates positively with body weight in both Black and Caucasian women [113]. However, differences in body weight do not explain the differences in bone mass between Blacks and Caucasians [114]. Epidemiologic studies indicate that African American women are classified as overweight twice as frequently as non-Hispanic Caucasian women [115], but that differences in body mass index (BMI) do not develop until after adolescence (interestingly, that is about the time differences in bone mass become apparent as well). Obesity classification typically depends on self-report of weight and height, with
576 subsequent computation of BMI. Obesity is then defined as BMI 27.3 kg/m2 [116]. This definition is based on statistics from the second National Health and Nutrition Examination Survey (NHANES II), using the 85th percentile of BMI for 20 to 29-year-old non-Hispanic Caucasians as an obesity cutoff point. Application of this definition for obesity has not been race or ethnicity adjusted in the majority of studies. Lopez and Masse [117] compared anthropometric data from NHANES II to that acquired during the 1982 – 1984 Hispanic Health and Nutrition Examination Survey (HHANES). They found that use of the NHANES II obesity cutoff consistently labeled 12 – 14% more Puerto Rican and Mexican American women as obese than if ethnicity-specific cutoff data from HHANES were used [117]. Cuban American women’s BMIs reflected those of NHANES II data, rendering the use of a “Hispanic”-specific cutoff meaningless because there are differences in height and weight distribution among the Hispanic ethnic groups [118]. The optimal formula for a body mass index, which is conventionally expressed as weight/height2, varies among populations. If the purpose of using a weight for height index is to minimize the effect of height on body mass, then one approach to identifying an appropriate index is to find the exponent for the denominator that minimizes the correlation between weight and height (such that correlation r 0) in a given population [119]. Kleerekoper et al. [120] applied this approach to 201 White and 77 African American postmenopausal women participating in a longitudinal study of bone mass and biochemical markers of bone remodeling. The African American women had a significantly greater BMI based on the conventional formula. However, the formula that best provided a height-free measure of weight was different for the White [weight (kg)/height(m)1.17] and the African American women [weight(kg)/height(m)1.30]. When BMI was calculated using these population-specific formulae, there was no significant difference between the two groups, suggesting that these African American women were not more “obese.” This underscores the need to consider population-specific approaches to studying human biologic phenomena. Although BMI is often used to estimate degree of obesity, body composition is much more complex. The simplest model uses bone, fat, and lean body mass as the three major components. Because bone and lean body mass are closely related [121], it stands to reason that groups with increased muscle mass, such as African Americans and Polynesians [97,122], have higher bone mass. Therefore, total body weight will be greater in those with higher bone mass, not necessarily because of obesity but also due to the contribution of bone and muscle weight. Body composition comparisons made between African American and Caucasian adults demonstrate consistently greater muscle and
VILLA, NELSON, AND NELSON
bone mass in the former [123 – 125], underscoring the importance of using race- and ethnicity-specific reference populations when interpreting body habitus data.
D. Calcium Nutriture Calcium intake affects attainment of peak bone mass as well as the ability to preserve skeletal calcium throughout life [126]. In an early study, Matkovic and colleagues demonstrated that hip fracture rates differed significantly within an ethnic group living in two regions of Croatia with divergent levels of dietary calcium intake [127] and concluded that these differences were due to differences in attainment of peak bone mass. Gradations in bone mass related to calcium intake are also observed in other racial and ethnic groups. In an excellent dietary study of Chinese women with similar ethnic backgrounds, Hu and collaborators demonstrated a wide range in BMD depending on dietary calcium intake [128]. In this group, although the women with higher calcium intake had higher BMD, the rate of bone loss with age was not affected by dietary calcium, supporting the hypothesis that the differences in bone mass observed in older women were realized earlier in life. Calcium nutriture may contribute to differences in the bone mass of Japanese and Japanese American groups as well [129,130], and calcium supplementation has been shown to reduce bone loss in elderly Chinese women [131]. 1. LACTOSE INTOLERANCE Dairy products serve as a major source of dietary calcium in many places. Because the majority of the world’s non-Caucasian population develops lactase deficiency relatively early in life, it would seem that impaired dairy tolerance should lead to low calcium intake and therefore suboptimal development and preservation of bone. Many investigators cite this rationale when attempting to explain ethnic differences in calcium and vitamin D metabolism. However, the presence of lactose malabsorption does not predict milk consumption in Mexican Americans, Blacks, or the elderly [132,133], and lactase deficiency does not cause adult bone loss [134]. In fact, milk is the primary source of calcium in the diets of three Hispanic ethnic groups living in North America [135]. When analyzing NHANES and NHANES II data, Looker and co-workers [135] found that calcium intakes in Hispanic diets paralleled those of non-Hispanic Caucasians and were somewhat higher than those of African Americans. Striking differences existed in the dietary sources of calcium for the three Hispanic groups, although total calcium intakes did not vary significantly. Milk was the single greatest contributor for all three, but corn tortillas were
CHAPTER 22 Race, Ethnicity, and Osteoporosis
second in importance for Mexican Americans alone. A listing of the top 10 contributors to dietary calcium for Mexican Americans also included flour tortillas and pinto beans, whereas for Cuban Americans and Puerto Ricans, pizza and rice were major sources. The bioavailability of calcium may vary widely among these diverse foods. For some, such as pinto beans, calcium absorbability is less than that of milk [136], but the total intake of calcium from all sources may be sufficient to exceed gastrointestinal and renal losses. Questionnaire assessment of dietary calcium intake should therefore be tailored to include ethnic foods in order to collect representative data. 2. CALCIUM METABOLISM Racial or ethnic differences in the absorption and excretion of calcium may affect overall calcium balance. One study conducted under conditions of severe calcium restriction found African Americans to have low vitamin D levels and compensatory hypersecretion of parathyroid hormone, theoretically maximizing urinary retention of calcium and thereby contributing to greater bone mass [96]. Another study that provided adequate dietary calcium found no evidence of a racial alteration in the vitamin D endocrine system [137]. However, despite lack of racial differences in dietary calcium and vitamin D intake, it was found that Blacks had significantly lower urinary calcium excretion and that calcium excretion was related inversely to radial BMD. Studies comparing postmenopausal African American and White women reported statistically significant differences in calciotropic hormones and biochemical markers of bone remodeling. PTH and 1,25 vitamin D concentrations tend to be higher in African Americans, and 25 hydroxyvitamin D(25OHD), osteocalcin, hydroxyproline, and bonespecific alkaline phosphatase values lower in AfricanAmerican compared to White women (p0.05), suggesting possible resistance to the skeletal actions of PTH in African-Americans [60,138]. Skeletal attributes develop early in life, as mentioned previously, so Abrams and colleagues investigated aspects of calcium metabolism in Mexican American and White children. They found higher PTH concentrations in Mexican American girls despite lack of vitamin D deficiency, although ethnic differences in 25OHD and PTH concentrations did not significantly affect calcium absorption, excretion, or bone calcium kinetics [139].
E. Vitamin D Exposure Hypovitaminosis D, when present in non-Hispanic Caucasian populations, predicts low bone mass and increased risk for hip fractures [140,141]. A similar relationship is
577 seen among inhabitants of Hong Kong, where hypovitaminosis D appears as a common problem in elders with hip fracture [142]. In this group, subclinical vitamin D deficiency is also associated with muscle weakness and increased risk of falling. Studies conducted in Japan note a marked beneficial effect on BMD and spinal fracture rate in patients treated with vitamin D [143]. It appears, therefore, that individuals from divergent racial and ethnic backgrounds respond similarly to the influence of circulating 25OHD. Regional differences in vitamin D status exist, perhaps due to differences in levels of solar radiation, individuals’ exposure to sunlight, and skin color. Differences in skin pigmentation reflect an evolutionary adaptation to solar radiation, according to one theory, so that those living closer to the equator would have greater amounts of skin pigment in order to reduce the risk of hypervitaminosis D [144]. Conversely, those living in higher latitudes would have fairer skin in order to make sufficient amounts of vitamin D. Data suggest that there is a difference in the gradient of skin color south of the equator compared with north of the equator [145]; this would suggest that factors other than solar radiation affect skin pigmentation. Some investigators find that individuals with high and low skin pigmentation possess similar 25OHD synthetic abilities in response to ultraviolet exposure [146]. Others imply that increased pigment reduces the capacity of skin to synthesize vitamin D [96,98,147]. Data presented in abstract form found a high prevalence of inadequate vitamin D nutriture in a group of elderly Mexican American women, but this was related to vitamin D exposure rather than skin pigmentation [148]. In a multiracial/ethnic study of the relationship between skin pigment and cutaneous synthesis of vitamin D, Matsuoka and colleagues [49] found that increased skin pigmentation had a photoprotective effect but did not impair adequate formation of vitamin D. Comparison of Blacks and Caucasians living in Zaire or Belgium found no racial differences when the study was conducted in Zaire. Evaluation of respective groups in Belgium, however, demonstrated lower vitamin D values in Blacks, with an inverse relationship noted between serum 25OHD and length of stay in Belgium [150]. Blacks living in the United States showed lower concentrations of 25OHD and higher levels of 1,25-dihydroxyvitamin D when compared with Whites in the same geographic area, but without discernible effects on levels of calciotropic hormones or renal calcium excretion [151]. It is possible that at northern latitudes, avoidance of lifestyle habits such as sun-seeking behavior and intake of vitamin D-fortified foods will result in lower levels of circulating 25OHD. This could manifest itself in the protective response of mild secondary hyperparathyroidism seen by some investigators [152].
578
VILLA, NELSON, AND NELSON
F. Vitamin D Receptor Gene In 1992, Morrison et al. [153] reported that allelic variation in several polymorphisms at the vitamin D receptor (VDR) gene locus could be used to predict bone turnover and later reported an association with bone mass in a large group of white Australian women [154]. Attempts to corroborate these findings in other populations have yielded variable results, as well as investigations of other restriction fragment length polymorphisms (RFLPs). Investigations of the VDR gene in U.S. populations mainly focus on white women, but occasionally include other ethnic groups, such as African Americans [155 – 158] and Mexican Americans [159]. Results of these are somewhat contradictory, as discussed later, which may reflect ethnic and/or environmental differences in the genotype frequencies as well as on the expression of the VDR. Outside the United States, data from a study of Japanese women suggest an association between VDR gene polymorphisms and both BMD [160] and the rate of postmenopausal bone loss [161]. Studies of VDR in Chinese [162,163] and Korean [164] women do not find an association with bone mass, although the study groups may have had low dietary calcium intake, which could have independently affected attainment of peak bone mass. A significant correlation of the BsmI VDR and BMD [165] was observed among premenopausal Brazilian women living in Sao Paulo, a population characterized by a high degree of miscegenation. Fleet et al. [156] reported no significant ethnic difference in genotype distribution in adult White and African American women. They noted no significant interaction of ethnicity and genotype on BMD of the femoral neck and lumbar spine, although a significant relationship between the genotypes and bone density existed in the group as a whole. Using a start codon polymorphism detected with the endonuclease FokI, Harris et al., found Black/White differences in its distribution among premenopausal women [157]. They suggested that this polymorphism may influence peak bone density and that ethnic differences in genotype frequencies may explain some ethnic differences in femoral neck BMD. Nelson et al. [158] drew a similar conclusion using the BsmI polymorphism, reporting a significant difference in genotype distribution between premenopausal African American and White women, as well as a significantly higher mean whole body bone mass in the high bone mass (bb) genotype in the groups combined. It is notable that the low bone mass genotype (BB) was absent among the African American women. Their data suggested that ethnic differences in the distribution of the BsmI genotypes may help explain observed ethnic difference in whole body bone mass in younger adult women. Zmuda et al. [155,166] investigated four VDR gene polymorphisms (BsmI, ApaI, FokI, and TaqI), bone
turnover, and rates of bone loss in older African American women. They did not find an association between VDR gene polymorphisms and BMD or indices of bone turnover in this group. McClure et al. [159] studied three RFLPs (BsmI, ApaI, and TaqI) for VDR in postmenopausal Mexican American women and did not find significant associations with BMD, but there were trends suggesting that a larger sample size may reveal such associations. It may also be that lifelong habits and exposures (such as vitamin D and calcium intake) muddy the relationship between VDR and BMD when studied in older adults. Interestingly, Sainz et al. conducted a study looking at BsmI, ApaI, and TaqI VDR polymorphisms in prepubertal Mexican American girls, finding a strong relationship between VDR and both femoral and vertebral bone density.
G. Physical Activity and Other Lifestyle Factors Physical activity benefits the skeleton throughout all stages of life (physical activity and risk for osteoporosis are discussed elsewhere in this publication) and is a lifestyle habit subject to great variation among different groups. Despite varying cultural values and attitudes toward physical activity throughout life, its beneficial effect on bone mass appears ubiquitous. Ethnic differences in bone mass and fracture risk related to physical activity habits are seen in studies from Sweden [10], Japan [168], Mexico [169], and Hawaii [170]. Although multicultural studies concerning type of physical activity and accretion of bone density are lacking, it would be interesting to see whether culturally determined activities during adolescence affect lifetime risk for fracture. It could be that habitual activities or work during youth affect adult bone mass and possibly skeletal structure. Physical activity and muscle strength also affect the risk of falling, another contributor to overall fracture risk. Tobacco use contributes to risk of osteoporotic fracture (discussed in greater detail elsewhere in this volume). Smoking is thought to reduce calcium absorption [171], and heavy smokers appear to have lower BMD than light smokers or nonsmokers [172]. Elderly Japanese American men who smoke have significantly faster bone loss rates than those who do not [173,174], but the relationship among smoking, bone mass, race, and ethnicity has not been widely studied. Alcohol intake contributes to risk for low bone density [175], but its relationship to fracture is not well established in multiethnic populations. Some data suggest that noninsulin dependent diabetes mellitus is positively associated with bone mineral density in women [176], and a positive association between serum glucose concentrations and bone mass in Mexican American women has been described [12]. Mexican Americans and Blacks experience a proportionately greater amount of diabetes than non-Hispanic Caucasians [2,177]. This may
579
CHAPTER 22 Race, Ethnicity, and Osteoporosis
contribute to differences in bone mineral density or rates of bone loss.
H. Falls Falling down contributes greatly to risk of suffering a hip fracture. Few studies of racial or ethnic differences in falls incidence exist. In a study of risk factors for hip fracture in African American women, Grisso and colleagues found that variables associated with risk for falling (previous stroke, use of ambulatory aids, lower-limb dysfunction, history of seizures) also predicted hip fracture risk [107]. A study of postmenopausal Mexican American Caucasians [178] found the falls rate and risk factors contributing to falls to be comparable to that reported for non-Hispanic Caucasians [179]. Rate of falls in Japanese Americans living in Hawaii appear to be lower than that published for Caucasians [180], but that risk for injury following a fall to be no different.
I. Bone Geometry Bone mineral density predicts fracture risk, but there is considerable overlap of BMD in non Hispanic Caucasian controls and hip fracture cases. Also, as noted previously, there is considerable variation in fracture risk among ethnic and racial groups with similar BMD values. Thus, factors other than bone mass also affect risk for osteoporotic fracture. Analysis of bone densitometry data collected in a large study of osteoporotic fracture suggests that a simple geometric measurement of femoral size, hip axis length (HAL), is related to hip fracture risk [181]. In this study, shorter HAL was associated with a decreased risk of hip fracture. Of the three racial groups represented in this study, African Americans and Asian Americans had a significantly lower HAL than the fracture group (which was predominantly non-Hispanic Caucasian) [182]. Additionally, the HAL of Mexican American Caucasian women, another ethnic group with a relatively low risk for hip fracture (see Table 1), averages about the same as for Asian and African Americans [12]. Radiographic studies of hip geometry have shown crossnational as well as ethnic differences that may relate to hip fracture risk. One study comparing Japanese and Caucasian differences in geometric properties of the femoral neck demonstrated an association between low fracture risk and short femoral neck [183], whereas a study of African Americans and U.S. Whites found significant ethnic differences in various measurements of hip geometry [184]. Other potentially important geometric variables have been assessed in the proximal femur using DXA data to describe cross-sectional geometry, using the method of “hip structure analysis” developed by Beck et al. [185].
Measurements include bone width (subperiosteal diameter), cortical thickness, cross-sectional bone area, cross-sectional moment of inertia, and section modulus, which contribute to the biomechanical strength of the hip. An investigation of data from NHANES III showed significant sex and ethnic differences in many of these variables [186]. Nelson et al. [187] used this method when analyzing data from a group of postmenopausal African American and white women, showing that the spatial distribution of bone in the femoral neck is arranged to resist greater loading in the African American women. A comparison of these results with hip structure analysis of data from Black and White postmenopausal women in Johannesburg showed that both U.S. ethnic groups have significantly greater indices than their South. African counterparts, although the Black women in both countries have a higher section modulus — an index of bending strength — in the femoral neck compared with White women (unpublished data). Thus it is clear that factors other than bone mass may be important in determining the biomechanical strength of the hip and that these may differ across ethnic groups. The same cautionary rule applies to assessment of HAL in determining risk for fracture: its application to risk may vary with ethnicity. In a comparison of bone geometric properties as risk factors for hip fracture in European, Chinese, Indian, and Polynesian premenopausal women, Chin and colleagues [188] found shorter HAL in the Chinese and Indian groups, but longer HAL in European and Polynesian women. Because Polynesian women enjoy a very low rate of hip fracture, osteoporotic risk factors other than HAL must be considered.
V. SUMMARY The international range in hip fracture incidence confirms the widely held notion that many factors enter into determination of skeletal health. Much information may be gleaned from interracial and interethnic studies that may help elucidate possible etiologies in the pathogenesis of osteoporosis. In order to best delineate these factors, investigators must explore contributions from the environmental and cultural milieu in which different groups of people reside. Description of study groups would ideally describe the criteria on which categories such as race and ethnicity were determined and would include a discussion of the degree of acculturation when appropriate. Description of study groups would ideally include race, ethnicity, and acculturation. Bone mass, in and of itself, is not the best predictor of fracture risk in all groups. Variables such as bone geometry, reproductive history, physical activity, dietary exposures, body composition, and others all contribute to fracture risk, reflecting the rich diversity peculiar to the human race.
580
VILLA, NELSON, AND NELSON
References 1. M. L. Villa, Cultural determinants of skeletal health: The need to consider both race and ethnicity in bone research. J. Bone Miner. Res. 9(9), 1329 – 1332 (1994). 2. C. L. Hanis, R. E. Ferrell, S. A. Barton, L. Aguilar, A. Garza-Ibarra, B. R. Tulloch, C. A. Garcia, and W. J. Schull, Diabetes among Mexican Americans in Starr County, Texas. Am. J. Epidemiol. 118(5), 659 – 672 (1983). 3. T. Sugimoto, M. Tsutsumi, Y. Fujii, M. Kawakatsu, H. Negishi, M. C. Lee, K. S. Tsai, M. Fukase, and T. Fujita, Comparison of bone mineral content among Japanese, Koreans, and Taiwanese assessed by dual-photon absorptiometry. J. Bone Miner. Res. 7(2), 153 – 159 (1992). 4. D. A. Nelson and D. A. Barondess, Whole body bone, fat and lean mass in children: Comparison of three ethnic groups. Am. J. Phys. Anthropol. 103(2), 157 – 62 (1997). 5. J. Anderson, M. Moeschberger, M. J. Chen, P. Kunn, M. E. Wewers, and R. Guthrie, An acculturation scale for Southeast Asians. Soc. Psychiatry Psychiatr. Epidemiol. 28(3), 134 – 141 (1993). 6. I. Cuellar, L. C. Harris, and R. Jasso, An acculturation scale for Mexican-American normal and clinical populations. Hisp. J. Behav. Sci. 2, 199 – 217 (1980). 7. H. P. Hazuda, M. P. Stern, and S. M. Haffner, Acculturation and assimilation among Mexican Americans: scales and population-based data. Soc. Sci. Quart. 69(3), 687 – 706 (1988). 8. N. W. Jerome, Diet and acculturation. In “Nutritional Anthropology: Contemporary Approaches to Diet and Culture” (N. W. Jerome, R. F. Kandel, and G. H. Pelto, eds.), pp 275 – 325. Redgrave Publishing Company, (1980). 9. H. P. Hazuda, S. M. Haffner, M. P. Stern, and C. W. Eifler, Effects of acculturation and socioeconomic status on obesity and diabetes in Mexican Americans: The San Antonio Heart Study. Am. J. Epidemiol. 128(6), 1289 – 1301 (1988). 10. B. Jónsson, P. Gärdsell, O. Johnell, I. Sernbo, and B. Gullberg, Lifestyle and different fracture prevalence: a cross-sectional populationbased study. Calcif. Tissue Int. 52, 425 – 433 (1993). 11. P. D. Ross, H. Norimatsu, J. W. Davis, K. Yano, R. D. Wasnich, S. Fujiwara, Y. Hosoda, and L. Melton, A comparison of hip fracture incidence among native Japanese, Japanese Americans, and American Caucasians. Am. J. Epidemiol. 133(8), 801 – 809 (1991). 12. M. L. Villa, R. Marcus, R. R. Delay, and J. L. Kelsey, Factors contributing to skeletal health of postmenopausal Mexican-American women. J. Bone Miner. Res. 10(8), 1233 – 1242 (1995). 13. S. Maggi, J. L. Kelsey, J. Litvak, and S. P. Heyse, Incidence of hip fractures in the elderly: A cross-national analysis. Osteopor. Int. 1(4), 232 – 241 (1991). 14. M. E. Farmer, L. R. White, J. A. Brody, and K. R. Bailey, Race and sex differences in hip fracture incidence. Am. J. Public Health 74 (12), 1374 – 1380 (1984). 15. J. A. Falch, A. Ilebekk, and U. Slungaard, Epidemiology of hip fractures in Norway. Acta Orthop. Scand. 56, 12 – 16 (1985). 16. J. A. Baron, J. Barrett, D. Malenka, E. Fisher, W. Kniffin, T. Bubolz, and T. Tosteson, Racial differences in fracture risk. Epidemiology 5, 42 – 47 (1994). 17. E. S. Fisher, J. A. Baron, D. J. Malenka, J. A. Barrett, W. D. Kniffin, F. S. Whaley, and T. A. Bubolz, Hip fracture incidence and mortality in New England. Epidemiology 2, 116 – 122 (1991). 18. R. Y. Hinton and G. S. Smith, The association of age, race, and sex with the location of proximal femoral fractures in the elderly. J. Bone Joint Surg. Am. 75(5), 752 – 759 (1993). 19. S. L. Silverman and R. E. Madison, Decreased incidence of hip fracture in Hispanics, Asians, and blacks: California Hospital Discharge Data. Am. J. Public Health 78(11), 1482 – 1483 (1988).
20. R. L. Bauer, Ethnic differences in hip fracture: A reduced incidence in Mexican Americans. Am. J. Epidemiol. 127(1), 145 – 149 (1988). 21. L. Solomon, Osteoporosis and fracture of the femoral neck in the South African Bantu. J. Bone Jt. Surg. 50B, 2 – 11 (1968). 22. H. Hagino, K. Yamamoto, H. Ohshiro, T. Nakamura, H. Kishimoto, and T. Nose, Changing incidence of hip, distal radius, and proximal humerus fractures in Tottori Prefecture, Japan. Bone 24, 265 – 270 (1999). 23. H. Hagino, K. Yamamoto, R. Teshima, H. Kishimoto, K. Kuranobu, and T. Nakamura, The incidence of fractures of the proximal femur and the distal radius in Tottori prefecture, Japan. Arch. Orthop. Trauma Surg. 109(1), 43 – 44 (1990). 24. X. Ling, L. Aimin, C. Xiaoshu, and S. R. Cummings, Very low rates of hip fracture in Beijing, People’s Republic of China: The Beijing Osteoporosis Project. Am. J. Epidemiol. 144, 901 – 907 (1996). 25. A. V. Schwartz, J. L. Kelsey, S. Maggi, M. Tuttleman, S. C. Ho, P. V. Jonsson, G. Poor, J. A. Sisson de Castro, L. Xu, C. C. Matkin, L. M. Nelson, and S. P. Heyse, International variation in the incidence of hip fractures: Cross-national project on osteoporosis for the World Health Organization Program for Research on Aging. Osteopor. Int. 9, 242 – 253 (1999). 26. E. M. C. Lau, C. Cooper, C. Wickham, S. Donnan, and D. J. P. Barker, Hip fracture in Hong Kong and Britain. Int. J. Epidemiol 19, 1119 – 1121 (1990). 27. J. Chalmers and K. C. Ho, Geographical variations in senile osteoporosis: The association with physical activity. J. Bone Jt. Surg. 52B, 667 – 675 (1970). 28. P. C.N. Wong, Fracture epidemiology in a mixed southeastern Asian community (Singapore). Clin. Orthop. 45, 55 – 61 (1966). 29. A. Memon, W. M. Pospula, A. Y. Tantawy, S. Abdul-Ghafar, A. Suresh, and A. Al Rowaih, Incidence of hip fracture in Kuwait. Int. J. Epidemiol. 27, 860 – 865 (1998). 30. S. Stott, D. H. Gray, and W. Stevenson, The incidence of femoral neck fractures in New Zealand. New Zeal. Med. J. 91, 6 – 9 (1980). 31. R. Nilsson, O. Lofman, K. Berglund, L. Larson, and G. Toss, Increased hip-fracture incidence in the county of Ostergotland, Sweden, 1940-1986, with forecasts up to the year 2000: An epidemiological study. Int. J. Epidemiol. 20, 1018 – 1024 (1991). 32. P. Luthje, Incidence of hip fracture in Finland: A forecast for (1990). Acta Orthop. Scand. 56, 223 – 225 (1985). 33. B. S. Z. Elabdien, S. Olerud, G. Karlstrom, and B. Smedby, Rising incidence of hip fracture in Uppsala, 1965 – 1980. Acta Orthop. Scand. 55, 284 – 289 (1984). 34. R. Hedlund, U. Lindgren, and A. Ahlbom, Age- and sex-specific incidence of femoral neck and trochanteric fractures. Clin. Orthop. Rel. Res. 222, 132 – 139 (1987). 35. R. Hedlund, A. Ahlbom, and U. Lindgren, Hip fracture incidence in Stockholm 1972 – 1981. Acta Orthop. Scand. 57, 30 – 34 (1985). 36. P. A. Alffram, An epidemiologic study of cervical and trochanteric fractures of the femur in an urban population. Acta Orthop. Scand. Suppl. 65, 101 – 109 (1964). 37. V. Finsen and P. Benum, Changing incidence of hip fractures in rural and urban areas of central Norway. Clin. Orthop. Rel. Res. 218, 104 – 110 (1987). 38. E. M. Alhava and J. Puittinen, Fractures of the upper end of the femur as an index of senile osteoporosis in Finland. Ann. Clin. Res. 5, 398 – 403 (1973). 39. A. L. Currie, D. M. Reid, and N. Brown, An epidemiological study of fracture of the neck of femur. Health Bull. 44, 143 – 148 (1986). 40. W. J. Boyce, and P. M. Vessey, Rising incidence of fracture of the proximal femur. Lancet 1, 150 – 151 (1985). 41. M. R. Baker, An investigation into secular trends in the incidence of femoral neck fracture using hospital activity analysis. Public Health 94, 368 – 374 (1980).
CHAPTER 22 Race, Ethnicity, and Osteoporosis 42. A. Lizaur-Utrilla, P. Orts, F. Sanchez Del Campo, A. A. Barrio, and P. Gutierrez-Carbonell, Epidemiology of trochanteric fractures of the femur in Alicante, Spain, 1974 – 1982. Clinic. Orthop. Rel. Res. 218, 24 – 31 (1987). 43. G. F. Mazzuoli, C. Gennari, M. Passeri, F. S. Celi, M. Acca, A. Camporeale, G. Pioli, and M. Pedrazzoni, Incidence of hip fracture: An Italian survey. Osteopor. Int. 3 (Suppl 1), 8 – 9 (1993). 44. S. Levine, M. Makin, J. Menczel, G. Robin, E. Naor, and R. Steinbert, Incidence of fractures of the proximal end of the femur in Jerusalem: A study of ethnic factors. J. Bone Jt. Surg. 52, 1193 – 1202 (1970). 45. W. A. Ray, M. R. Griffin, R. West, L. Strand, and L. J. Melton, Incidence of hip fracture in Saskatchewan, Canada, 1976 – 1985. Am. J. Epidemiol. 131, 502 – 509 (1990). 46. J. G. RodrÌguez, R. W. Sattin, and R. J. Waxweiler, Incidence of hip fractures, United States, 1970 – 83. Am. J. Prev. Med. 5(3), 175 – 181 (1989). 47. J. C. Gallagher, L. J. Melton, B. L. Riggs, and E. Bergstrath, Epidemiology of fractures of the proximal femur in Rochester, Minnesota. Clin. Orthop. Rel. Res. 150, 163 – 171 (1980). 48. K. M. Sanders, E. Seeman, A. M. Ugoni, J. A. Pasco, T. J. Martin, B. Skoric, G. C. Nicholson, and M. A. Kotowicz, Age- and genderspecific rate of fractures in Australia: A population-based study. Osteopor. Int. 10, 240 – 247 (1999). 49. W. E. Bacon, S. Maggi, A. Looker, T. Harris, C. R. Nair, J. Giaconi, R. Honkanen, S. C. Ho, K. A. Peffers, O. Torring, R. Gass, and N. Gonzalez, International comparison of hip fracture rates in 1988 – 89. Osteopor. Int. 6, 69 – 75 (1996). 50. L. J. Melton, Epidemiology of fractures. In “Osteoporosis: Etiology, Diagnosis and Management” (B. L. Riggs, L. J. D. Melton, eds.), pp. 111 – 154. Raven Press, New York, (1988). 51. D. S. Lauderdale, S. J. Jacobsen, S. E. Furner, P. S. Levy, J. A. Brody, and J. Goldberg, Hip fracture incidence among elderly Hispanics. Am. J. Public Health 88(8), 1245 – 1247 (1998). 52. D. S. Lauderdale, S. J. Jacobsen, S. E. Furner, P. S. Levy, J. A. Brody, and J. Goldberg, Hip fracture incidence among elderly Asian-American populations. Am. J. Epidemiol. 146(6), 502 – 509 (1997). 53. L. Solomon, Bone density in ageing Caucasian and African populations. Lancet 2, 1326 – 1330 (1979). 54. M. M. Luckey, D. E. Meier, J. P. Mandeli, M. C. DaCosta, M. L. Hubbard, and S. J. Goldsmith, Radial and vertebral bone density in white and black women: Evidence for racial differences in premenopausal bone homeostasis. J. Clin. Endocrinol. Metab. 69(4), 762 – 770 (1989). 55. R. Marcus, G. Greendale, B. A. Blunt, T. L. Bush, S. Sherman, R. Sherwin, H. Wahner, and B. Wells, Correlates of bone mineral density in the Postmenopausal Estrogen/Progestin Interventions Trial. J. Bone Miner. Res. 9(9), 1467 – 1476 (1994). 56. Y. Liel, J. Edwards, J. Shary, K. M. Spicer, L. Gordon, and N. H. Bell, The effects of race and body habitus on bone mineral density of the radius, hip, and spine in premenopausal women. J. Clin. Endocrinol. Metab. 66(6), 1247 – 1250 (1988). 57. J. F. Aloia, A. Vaswani, M. Mikhail, M. Badshah, and E. Flaster, Cancellous bone of the spine is greater in black women. Calcif. Tissue Int. 65(1), 29 – 33 (1999). 58. A. C. Looker, H. W. Wahner, W. L. Dunn, M. S. Calvo, T. B. Harris, S. P. Heyse, C. C. Johnston, Jr., and R. L. Lindsay, Proximal femur bone mineral levels of US adults. Osteopor. Int. 5(5), 389 – 409 (1995). 59. B. Ettinger, S. Sidney, S. R. Cummings, C. Libanati, D. D. Bikle, I. S. Tekawa, K. Tolan, and P. Steiger, Racial differences in bone density between young adult black and white subjects persist after adjustment for anthropometric, lifestyle, and biochemical differences. J. Clin. Endocrinol. Metab. 82(2), 429 – 434 (1997).
581 60. M. Kleerekoper, D. A. Nelson, E. L. Peterson, M. J. Flynn, A. S. Pawluszka, G. Jacobsen, and P. Wilson, Reference data for bone mass, calciotropic hormones, and biochemical markers of bone remodeling in older (55-75) postmenopausal white and black women. J. Bone Miner. Res. 9(8), 1267 – 1276 (1994). 61. J. Y. Li, B. L. Specker, M. L. Ho, and R. C. Tsang, Bone mineral content in black and white children 1 to 6 years of age: Early appearance of race and sex differences. Am. J. Dis. Child. 143(11), 1346 – 1349 (1989). 62. N. H. Bell, J. Shary, J. Stevens, M. Garza, L. Gordon, and J. Edwards, Demonstration that bone mass is greater in black than in white children. J. Bone Miner. Res. 6(7), 719 – 723 (1991). 63. D. P. McCormick, S. W. Ponder, H. D. Fawcett, and J. L. Palmer, Spinal bone mineral density in 335 normal and obese children and adolescents: Evidence for ethnic and sex differences. J. Bone Miner. Res. 6(5), 507 – 513 (1991). 64. D. Laraque, L. Arena, J. Karp, and D. Gruskay, Bone mineral content in black pre-schoolers: Normative data using single photon absorptiometry. Pediatr. Radiol. 20(6), 461 – 463 (1990). 65. S. M. Garn, S. T. Sandusky, J. M. Nagy, and M. B. McCann, Advanced skeleton development in low-income and Negro children. J. Pediatr. 80, 965 – 969 (1972). 66. D. A. Nelson, P. M. Simpson, C. C. Johnson, D. A. Barondess, and M. Kleerekoper, The accumulation of whole body skeletal mass in third- and fourth-grade children: Effects of age, gender, ethnicity, and body composition. Bone 20(1), 73 – 78 (1997). 67. A. Prentice, M. A. Laskey, J. Shaw, T. J. Cole, and D. R. Fraser, Bone mineral content of Gambian and British children aged 0 – 36 months. Bone Miner. 10, 211 – 224 (1990). 68. R. N. Southard, J. D. Morris, J. D. Mahan, J. R. Hayes, M. A. Torch, A. Sommer, and W. B. Zipf, Bone mass in healthy children: Measurement with quantitative DXA. Radiology 179, 735 – 738 (1991). 69. V. Gilsanz, T. F. Roe, S. Mora, G. Costin, and W. G. Goodman, Changes in vertebral bone density in black girls and white girls during childhood and puberty [see comments]. N. Engl. J. Med. 325(23), 1597 – 600 (1991). 70. M. C. Wang, M. Aguirre, G. S. Bhudhikanok, C. G. Kendall, S. Kirsch, R. Marcus, and L. K. Bachrach, Bone mass and hip axis length in healthy Asian, black, Hispanic, and white American youths. J. Bone Miner. Res. 12(11), 1922 – 1935 (1997). 71. L. K. Bachrach, T. Hastie, M. C. Wang, B. Narasimhan, and R. Marcus, Bone mineral acquisition in healthy Asian, Hispanic, black, and Caucasian youth: A longitudinal study. J. Clin. Endocrinol. Metab. 84(12), 4702 – 4712 (1999). 72. D. N. Patel, J. M. Pettifor, P. J. Becker, C. Grieve, and K. Leschner, The effect of ethnic group on appendicular bone mass in children. J. Bone Miner. Res. 7(3), 263 – 272 (1992). 73. E. D. Daniels, J. M. Pettifor, C. M. Schnitzler, S. W. Russell, and D. N. Patel, Ethnic differences in bone density in female South African nurses. J. Bone Miner. Res. 10(3), 359 – 367 (1995). 74. E. D. Daniels, J. M. Pettifor, C. M. Schnitzler, G. P. Moodley, and D. Zachen, Differences in mineral homeostasis, volumetric bone mass and femoral neck axis length in black and white South African women. Osteopor. Int. 7(2), 105 – 112 (1997). 75. T. J. Aspray, A. Prentice, T. J. Cole, Y. Sawo, J. Reeve, and R. M. Francis, Low bone mineral content is common but osteoporotic fractures are rare in elderly rural Gambian women. J. Bone Miner. Res. 11(7), 1019 – 1025 (1996). 76. I. R. Reid, M. Mackie, and H. K. Ibbertson, Bone mineral content in Polynesian and white New Zealand women. Br. Med. J. 292, 1547 – 1548 (1986). 77. B. W. Benson, T. J. Prihoda, and B. J. Glass, Variations in adult cortical bone mass as measured by a panoramic mandibular index. Oral Surg. Oral Med. Oral Pathol. 71(3), 349 – 356 (1991).
582 78. L. Haddock, V. Ortiz, M. D. Vazquez, F. Aguilo, E. Bernard, A. Ayala, and N. Mejias, The lumbar and femoral bone mineral densities in a normal female Puerto Rican population. P. R. Health Sci. J. 15(1), 5 – 11 (1996). 79. M. Russell-Aulet, J. Wang, J. Thornton, E. W. Colt, and R. J. Pierson, Bone mineral density and mass by total-body dual-photon absorptiometry in normal white and Asian men. J. Bone Miner. Res. 6(10), 1109 – 1113 (1991). 80. K. Yano, R. D. Wasnich, J. M. Vogel, and L. K. Heilbrun, Bone mineral measurements among middle-aged and elderly Japanese residents in Hawaii. Am. J. Epidemiol. 119(5), 751 – 764 (1984). 81. S. Hagiwara, T. Miki, Y. Nishizawa, H. Ochi, Y. Onoyama, and H. Morii, Quantification of bone mineral content using dual-photon absorptiometry in a normal Japanese population. J. Bone Miner. Res. 4(2), 217 – 222 (1989). 82. R. B. Mazess, Bone mineral in Vilcabamba, Ecuador. Am. J. Roentgenol. 130, 671 – 674 (1978). 83. H. M. Perry III, M. Bernard, M. Horowitz, D. K. Miller, S. Fleming, M. Z. Baker, J. Flaherty, R. Purushothaman, R. Hajjar, F. E. Kaiser, P. Patrick, and J. E. Morley, The effect of aging on bone mineral metabolism and bone mass in Native American women. J. Am. Geriatr. Soc. 46(11), 1418 – 1422 (1998). 84. A. B. Harper, W. S. Laughlin, and R. B. Mazess, Bone mineral content in St Lawrence Island Eskimos. Hum. Biol. 56(1), 63 – 78 (1984). 85. R. B. Mazess and W. Mather, Bone mineral content of North Alaskan Eskimos. Am. J. Clin. Nutr. 27, 916 – 925 (1974). 86. D. McHugh, R. N. Baumgartner, P. M. Stauber, S. Wayne, V. L. Hicks, and V. H. Heyward, Bone mineral in southwest Native American women. Basic Life Sci. 60(87), 87 – 88 (1993). 87. Z. Chen, M. J. Maricic, S. B. Going, T. G. Lohman, B. R. Altimari, and T. L. Bassford, Comparative findings in bone mineral density among postmenopausal Native American women and postmenopausal White women residing in Arizona. Bone 23(5), S592 (1998). 88. J. A. Kanis, L. J. d. Melton, C. Christiansen, C. C. Johnston, and N. Khaltaev, The diagnosis of osteoporosis. J. Bone Miner. Res. 9(8), 1137 – 1141 (1994). 89. D. E. Meier, M. M. Luckey, S. Wallenstein, R. H. Lapinski, and B. Catherwood, Racial differences in pre- and postmenopausal bone homeostasis: Association with bone density. J. Bone Miner. Res. 7(10), 1181 – 1189 (1992). 90. Z. H. Han, S. Palnitkar, D. S. Rao, D. Nelson, and A. M. Parfitt, Effects of ethnicity and age or menopause on the remodeling and turnover of iliac bone: Implications for mechanisms of bone loss [published erratum appears in J. Bone Miner. Res. 14(4), 660 (1999)]. J. Bone Miner. Res. 12(4), 498 – 508 (1997). 91. R. S. Weinstein and N. H. Bell, Diminished rates of bone formation in normal Black adults. N. Engl. J. Med. 319(26), 1698 – 1701 (1988). 92. C. M. Schnitzler, Bone quality: A determinant for certain risk factors for bone fragility. Calcif. Tissue Int. 53 (Suppl 1), S27 – 531 (1993). 93. C. M. Schnitzler, J. M. Pettifor, J. M. Mesquita, M. D. Bird, E. Schnaid, and A. E. Smyth, Histomorphometry of iliac crest bone in 346 normal black and white South African adults. Bone Miner. 10(3), 183 – 199 (1990). 94. B. Dibba, A. Prentice, M. A. Laskey, D. M. Stirling, and T. J. Cole, An investigation of ethnic differences in bone mineral, hip axis length, calcium metabolism and bone turnover between West African and Caucasian adults living in the United Kingdom. Ann. Hum. Biol. 26(3), 229 – 242 (1999). 95. H. Hagino, K. Yamamoto, R. Teshima, H. Kishimoto, and T. Kagawa, Radial bone mineral changes in pre- and postmenopausal healthy Japanese women: Cross-sectional and longitudinal studies. J. Bone Miner. Res. 7(2), 147 – 152 (1992).
VILLA, NELSON, AND NELSON 96. N. H. Bell, A. Greene, S. Epstein, M. J. Oexman, S. Shaw, and J. Shary, Evidence for alteration of the vitamin D-endocrine system in Blacks. J. Clin. Invest. 76, 470 – 473 (1985). 97. I. R. Reid, S. Cullen, B. A. Schooler, N. E. Livingston, and M. C. Evans, Calcitropic hormone levels in polynesians: Evidence against their role in interracial differences in bone mass. J. Clin. Endocrinol. Metab. 70(5), 1452 – 1456 (1990). 98. C. Reasner, J. F. Dunn, D. A. Fetchick, Y. Liel, B. W. Hollis, S.Epstein, J. Shary, G. R. Mundy, and N. H. Bell, Alteration of vitamin D metabolism in Mexican-Americans. J. Bone Miner. Res. 5(1), 13 – 17 (1990). 99. J. R. Buchanan, C. A. Myers, T. Lloyd, P. Leuenberger, and L. M. Demers, Determinants of peak trabecular bone density in women: The role of androgens, estrogen, and exercise. J. Bone Miner. Res. 3(6), 673 – 680 (1988). 100. J. S. Finkelstein, R. M. Neer, B. M.K. Biller, J. D. Crawford, and A. Klibanski, Osteopenia in men with a history of delayed puberty. N. Engl. J. Med. 326(9), 600 – 604 (1992). 101. B. J. Davidson, R. K. Ross, A. Paganini-Hill, G. D. Hammond, P. K. Siiteri, and H. L. Judd, Total and free estrogens and androgens in postmenopausal women with hip fractures. J. Clin. Endocrinol. Metab. 54(1), 115 – 120 (1982). 102. K. Kin, K. Kushida, K. Yamazaki, S. Okamoto, and T. Inoue, Bone mineral density of the spine in normal Japanese subjects using dualenergy X-ray absorptiometry: Effect of obesity and menopausal status. Calcif. Tissue Int. 49(2), 101 – 106 (1991). 103. J. A. Cauley, J. P. Gutai, L. H. Kuller, J. Scott, and M. C. Nevitt, Black-white differences in serum sex hormones and bone mineral density. Am. J. Epidemiol. 139(10), 1035 – 1046 (1994). 104. D. K. Katzman, L. K. Bachrach, D. R. Carter, and R. Marcus, Clinical and anthropometric correlates of bone mineral acquisition in healthy adolescent girls. J. Clin. Endocrinol. Metab. 73(6), 1332 – 1339 (1991). 105. M. Russell-Aulet, J. Wang, J. C. Thornton, E. W. Colt, and R. J. Pierson, Bone mineral density and mass in a cross-sectional study of white and Asian women. J. Bone Miner. Res. 8(5), 575 – 582 (1993). 106. G. S. Bhudhikanok, M. C. Wang, K. Eckert, C. Matkin, R. Marcus, and L. K. Bachrach, Differences in bone mineral in young Asian and Caucasian Americans may reflect differences in bone size. J. Bone Miner. Res. 11(10), 1545 – 1556 (1996). 107. J. A. Grisso, J. L. Kelsey, B. L. Strom, L. A. O’Brien, G. Maislin, K. LaPann, L. Samelson, and S. Hoffman, Risk factors for hip fracture in black women. N. Engl. J. Med. 330(22), 1555 – 1559 (1994). 108. E. M. Lau, J. Woo, P. C. Leung, and R. Swaminthan, Low bone mineral density, grip strength and skinfold thickness are important risk factors for hip fracture in Hong Kong Chinese. Osteopor. Int. 3(2), 66–70 (1993). 109. M. E. Pruzansky, M. Turano, M. Luckey, and R. Senie, Low body weight as a risk factor for hip fracture in both black and white women. J. Orthop. Res. 7(2), 192 – 197 (1989). 110. I. R. Reid, L. D. Plank, and M. C. Evans, Fat mass is an important determinant of whole body bone density in premenopausal women but not in men. J. Clin. Endocrinol. Metab. 75(3), 779 – 782 (1992). 111. J. F. Aloia, D. M. McGowan, A. N. Vaswani, P. Ross, and S. H. Cohn, Relationship of menopause to skeletal and muscle mass. Am J. Clin. Nutr. 53, 1378 – 1383 (1991). 112. H. Matsumine, K. Hirato, T. Yanaihara, T. Tamada, and M. Yoshida, Aromatization by skeletal muscle. J. Clin. Endocrinol. Metab. 63(3), 717 – 720 (1986). 113. D. P. DeSimone, J. Stevens, J. Edwards, J. Shary, L. Gordon, and N. H. Bell, Influence of body habitus and race on bone mineral density of the midradius, hip, and spine in aging women. J. Bone Miner. Res. 4(6), 827 – 830 (1989). 114. D. A. Nelson, M. Kleerekoper, and A. M. Parfitt, Bone mass, skin color and body size among black and white women. Bone Miner. 4, 257 – 264 (1988).
583
CHAPTER 22 Race, Ethnicity, and Osteoporosis 115. S. Kumanyika, Obesity in Black women. Epidemiol. Rev. 9, 31 – 51 (1987). 116. Anonymous, Anthropometric reference data and prevalence of overweight, United States, 1976-1980. Dept of Health and Human Services, Washington, DC. 117. L. M. Lopez and B. Masse, Comparison of body mass indexes and cutoff points for estimating the prevalence of overweight in Hispanic women. J. Am. Diet. Assoc. 92(11), 1343 – 1347 (1992). 118. L. M. López and B. Masse, Comparison of body mass indexes and cutoff points for estimating the prevalence of overweight in Hispanic women. J. Am. Diet. Assoc. 92(11), 1343 – 1347 (1992). 119. J. Lee, L. Kolonel, and M. Hinds, Relative merits of the weight-corrected-for height indices. Am. J. Clin. Nutr. 34, 2521 – 2529 (1981). 120. M. Kleerekoper, D. Nelson, E. Peterson, P. Wilson, G. Jacobsen, and C. Longcope, Body composition and gonadal steroids in older white and black women. J. Endocrinol. Metab. 79, 775 – 779 (1994). 121. K. J. Ellis and S. H. Cohn, Correlation between skeletal calcium mass and muscle mass in man. J. Appl. Physiol. 38(3), 455 – 460 (1975). 122. S. H. Cohn, C. Abesamis, I. Zanzi, J. F. Aloia, S. Yasumura, and K. J. Ellis, Body elemental composition: Comparison between black and white adults. Am. J. Physio. 232, E419 – 422 (1977). 123. K. D. Côté and W. C. Adams, Effect of bone density on body composition estimates in young adult black and white women. Med. Sci. Sports Exerc. 25(2), 290 – 296 (1993). 124. O. Ortíz, M. Russell, T. L. Daley, R. N. Baumgartner, M. Waki, S. Lichtman, J. Wang, R. J. Pierson, and S. B. Heymsfield, Differences in skeletal muscle and bone mineral mass between black and white females and their relevance to estimates of body composition. Am. J. Clin. Nutr. 55(1), 8 – 13 (1992). 125. K. J. Ellis, S. A. Abrams, and W. W. Wong, Body composition of a young, multiethnic female population. Am. J. Clin. Nutr. 65(3), 724 – 731 (1997). 126. R. P. Heaney, Calcium, bone health and osteoporosis. In “Bone and Mineral Research” (W. A. Peck, ed.) Vol. 4, pp. 255 – 301. Elsevier, Amsterdam. 127. V. Matkovic, K. Kostial, I. Simonovic, R. Buzina, A. Brodarec, and B. E. C. Nordin, Bone status and fracture rates in two regions of Yugoslavia. Am. J. Clin. Nutr. 32, 540 – 549 (1979). 128. J. F. Hu, X. H. Zhao, J. B. Jia, B. Parpia, and T. C. Campbell, Dietary calcium and bone density among middle-aged and elderly women in China. Am. J. Clin. Nutr. 58(2), 219 – 227 (1993). 129. T. Fujita, Y. Okamoto, T. Tomita, Y. Sakagami, K. Ota, and M. Ohata, Calcium metabolism in aging inhabitants of mountain versus seacoast communities in the Kii peninsula. J. Am. Geriatr. Soc. 25, 254 – 258 (1977). 130. K. Yano, L. K. Heilbrun, R. D. Wasnich, J. H. Hankin, and J. M. Vogel, The relationship between diet and bone mineral content of multiple skeletal sites in elderly Japanese-American men and women living in Hawaii. Am. J. Clin. Nutr. 42, 877 – 888 (1985). 131. E. M.C. Lau, J. Woo, P. C. Leung, R. Swaminathan, and D. Leung, The effects of calcium supplementation and exercise on bone density in elderly Chinese women. Osteopor. Int. 2, 168 – 173 (1992) 132. C. E. Woteki, E. Weser, and E. A. Young, Lactose malabsorption in Mexican-American adults. Am. J. Clin Nutr. 30, 470 – 475 (1977). 133. M. H. Rorick and N. S. Scrimshaw, Comparative tolerance of elderly from differing ethnic backgrounds to lactose-containing and lactosefree dairy drinks: A double-blind study. J. Gerontol. 34(2), 191 – 196 (1979). 134. C. W. Slemenda, J. C. Christian, S. Hui, J. Fitzgerald, and C. C. Johnston, No evidence for an effect of lactase deficiency on bone mass in pre- or postmenopausal women. J. Bone Miner. Res. 6(12), 1367 – 1371 (1991). 135. A. C. Looker, C. M. Loria, M. D. Carroll, M. A. McDowell, and C. L. Johnson, Calcium intakes of Mexican Americans, Cubans,
136. 137.
138.
139.
140.
141.
142.
143. 144. 145. 146.
147.
148.
149.
150.
151.
152.
153.
154.
Puerto Ricans, non-Hispanic whites, and non-Hispanic blacks in the United States. J. Am. Diet. Assoc. 93(11), 1274 – 1279 (1993). R. P. Heaney and C. M. Weaver, Plant constituents and food calcium absorbability. J. Bone Miner. Res. 7 (Suppl), S136 (1992). [Abstract] D. E. Meier, M. M. Luckey, S. Wallenstein, T. L. Clemens, E. S. Orwoll, and C. I. Waslien, Calcium, vitamin D, and parathyroid hormone status in young white and black women: Association with racial differences in bone mass. J. Clin. Endocrinol. Metab. 72(3), 703 – 710 (1991). H. M. Perry III, M. Horowitz, J. E. Morley, S. Fleming, J. Jensen, P. Caccione, D. K. Miller, F. E. Kaiser, and M. Sundarum, Aging and bone metabolism in African American and Caucasian women. J. Clin. Endocrinol. Metab. 81(3), 1108 – 1117 (1996). S. A. Abrams, K. C. Copeland, S. K. Gunn, J. E. Stuff, L. L. Clarke, and K. J. Ellis, Calcium absorption and kinetics are similar in 7- and 8-year-old Mexican-American and Caucasian girls despite hormonal differences. J. Nutr. 129(3), 666 – 671 (1999). D. T. Villareal, R. Civitelli, A. Chines, and L. V. Avioli, Subclinical vitamin D deficiency in postmenopausal women with low vertebral bone mass. J. Clin. Endocrinol. Metab. 72(3), 628 – 634 (1991). M. C. Chapuy, M. E. Arlot, F. Duboeuf, J. Brun, B. Crouzet, S. Arnaud, P. D. Delmas, and P. J. Meunier, Vitamin D3 and calcium to prevent hip fractures in elderly women. N. Engl. J. Med. 327(23), 1637 – 1642 (1992). K. K. Pun, F. H. W. Wong, C. Wang, P. Lau, P. W. M. Ho, W. K. Pun, S. P. Chow, C. L. Cheng, J. C. Y. Leong, and R. T. T. Young, Vitamin D status among patients with fractured neck of femur in Hong Kong. Bone 11, 365 – 368 (1990). T. Fujita, Studies of osteoporosis in Japan. Metabolism 39(4), 39 – 42 (1990). W. F. Loomis, Skin-pigment regulation of vitamin-D biosynthesis in man. Science 157, 501 – 506 (1967). J. Relethford, Hemispheric differences in human skin color. Am. J. Phys. Anthropol. 104, 449 – 457 (1998). W. F. Brazerol, A. J. McPhee, F. Mimouni, B. L. Specker, and R. C. Tsang, Serial ultraviolet B exposure and serum 25-hydroxyvitamin D response in young adult American Blacks and Whites: No racial differences. J. Am. Coll. Nutr. 7(2), 111 – 118 (1988). T. L. Clemens, S. L. Henderson, J. S. Adams, and M. F. Holick, Increased skin pigment reduces the capacity of skin to synthesise vitamin D3. Lancet 1, 74 – 76 (1982). M. Villa, J. Kelsey, J. Chen, and R. Marcus, Skin pigmentation does not affect vitamin D status in community-dwelling Mexican-American women J. Bone Miner. Res. 9 (Suppl 1), S418 (1994). [Abstract] L. Y. Matsuoka, J. Wortsman, J. G. Haddad, P. Kolm, and B. W. Hollis, Racial pigmentation and the cutaneous synthesis of vitamin D. Arch. Dermatol. 127(4), 536 – 538 (1991). J. R. M’Buyamba-Kabangu, R. Fagard, R. Lijnen, R. Bouillon, W. Lissens, and A. Amery, Calcium, vitamin D-endocrine system, and parathryroid hormone in black and white males. Calcif. Tissue Int. 41, 70 – 74 (1987). D. D. Bikle, B. Ettinger, S. Sidney, I. S. Tekawa, and K. Tolan, Differences in calcium metabolism between black and white men and women. Miner. Electrolyte Metab. 25(3), 178 – 184 (1999). H. M. I. Perry, D. K. Miller, J. E. Morley, M. Horowitz, F. E. Kaiser, H. M. J. Perry, J. Jensen, J. Bentley, S. Boyd, and D. Kraenzle, A preliminary report of vitamin D and calcium metabolism in older African Americans. J. Am. Geriatr. Soc. 41(6), 612 – 616 (1993). N. A. Morrison, R. Yeoman, P. J. Kelly, and J. A. Eisman, Contribution of trans-acting factor alleles to normal physiological variability: Vitamin D receptor gene polymorphism and circulating osteocalcin. Proc. Natl. Acad. Sci. USA 89(15), 6665 – 6669 (1992). N. A. Morrison, J. C. Qi, A. Tokita, P. J. Kelly, L. Crofts, T. V. Nguyen, P. N. Sambrook, and J. A. Eisman, Prediction of bone density from vitamin D receptor alleles [published erratum
584
155.
156.
157.
158.
159.
160.
161.
162.
163.
164.
165.
166.
167.
168.
169.
170.
VILLA, NELSON, AND NELSON appears in Nature 387 (6628),106 (1997)] Nature 367(6460), 284 – 287 (1994). J. M. Zmuda, J. A. Cauley, M. E. Danielson, R. L. Wolf, and R. E. Ferrell, Vitamin D receptor gene polymorphisms, bone turnover, and rates of bone loss in older African-American women. J. Bone Miner. Res. 12(9), 1446 – 1452 (1997). J. C. Fleet, S. S. Harris, R. J. Wood, and B. Dawson-Hughes, The BsmI vitamin D receptor restriction fragment length polymorphism (BB) predicts low bone density in premenopausal black and white women. J. Bone Miner. Res. 10(6), 985 – 990 (1995). S. S. Harris, T. R. Eccleshall, C. Gross, B. Dawson-Hughes, and D. Feldman, The vitamin D receptor start codon polymorphism (FokI) and bone mineral density in premenopausal American black and white women. J. Bone Miner. Res. 12(7), 1043 – 1048 (1997). D. Nelson, P. VandeVord, and P. Wooley, Polymorphism in the vitamin D receptor gene and bone mass in African-American and white mothers and children: A preliminary report. Ann. Rheum Dis. 59, 626 – 630 (2000). L. McClure, T. R. Eccleshall, C. Gross, M. L. Villa, N. Lin, V. Ramaswamy, L. Kohlmeier, J. L. Kelsey, R. Marcus, and D. Feldman, Vitamin D receptor polymorphisms, bone mineral density, and bone metabolism in postmenopausal Mexican-American women. J. Bone Miner. Res. 12(2), 234 – 240 (1997). M. Tamai, M. Yokouchi, S. Komiya, K. Mochizuki, S. Hidaka, S. Narita, A. Inoue, and K. Itoh, Correlation between vitamin D receptor genotypes and bone mineral density in Japanese patients with osteoporosis. Calcif. Tissue Int. 60(3), 229 – 232 (1997). Z. Yamagata, T. Miyamura, S. Iijima, A. Asaka, M. Sasaki, J. Kato, and K. Koizumi, Vitamin D receptor gene polymorphism and bone mineral density in healthy Japanese women. Lancet 344(8928), 1027 (1994). A. W. Kung, S. S. Yeung, and K. S. Lau, Vitamin D receptor gene polymorphisms and peak bone mass in southern Chinese women. Bone 22(4), 389 – 393 (1998). K. S. Tsai, S. H. Hsu, W. C. Cheng, C. K. Chen, P. U. Chieng, and W. H. Pan, Bone mineral density and bone markers in relation to vitamin D receptor gene polymorphisms in Chinese men and women. Bone 19(5), 513 – 518 (1996). S. K. Lim, Y. S. Park, J. M. Park, Y. D. Song, E. J. Lee, K. R. Kim, H. C. Lee, and K. B. Huh, Lack of association between vitamin D receptor genotypes and osteoporosis in Koreans. J. Clin. Endocrinol. Metab. 80(12), 3677 – 3681 (1995). M. Lazaretti-Castro, M. A. Duarte-de-Oliveira, E. M. Russo, and J. G. Vieira, Vitamin D receptor alleles and bone mineral density in a normal premenopausal Brazilian female population. Braz. J. Med. Biol. Res. 30(8), 929 – 932 (1997). J. M. Zmuda, J. A. Cauley, M. E. Danielson, T. M. Theobald, and R. E. Ferrell, Vitamin D receptor translation initiation codon polymorphism and markers of osteoporotic risk in older African-American women. Osteopor. Int. 9(3), 214 – 219 (1999). J. Sainz, J. M. Van Tornout, M. L. Loro, J. Sayre, T. F. Roe, and V. Gilsanz, Vitamin D-receptor gene polymorphisms and bone density in prepubertal American girls of Mexican descent. N. Engl. J. Med. 337(2), 77 – 82 (1997). T. Hirota, M. Nara, M. Ohguri, E. Manago, and K. Hirota, Effect of diet and lifestyle on bone mass in Asian young women. Am. J. Clin. Nutr. 55(6), 1168 – 1173 (1992). S. Parra-Cabrera, M. Hernandez-Avila, J. Tamayo-y-Orozco, L. Lopez-Carrillo, and F. Meneses-Gonzalez, Exercise and reproductive factors as predictors of bone density among osteoporotic women in Mexico City. Calcif. Tissue Int. 59(2), 89 – 94 (1996). P. D. Ross, H. Orimo, R. D. Wasnich, J. M. Vogel, C. J. MacLean, J. W. Davis, and A. Nomura, Methodological issues in comparing genetic and environmental influences on bone mass. Bone Miner. 7, 67 – 77 (1989).
171. E. A. Krall and B. Dawson-Hughes, Smoking and bone loss among postmenopausal women. J. Bone Miner. Res. 6(4), 331 – 338 (1991). 172. C. W. Slemenda, S. L. Hui, C. Longcope, and C. J. Johnston, Cigarette smoking, obesity, and bone mass. J. Bone Miner. Res. 4(5), 737 – 741 (1989). 173. J. W. Davis, P. D. Ross, J. M. Vogel, and R. D. Wasnich, Effects of smoking and alcohol on the rates of bone loss among elderly JapaneseAmerican men. J. Bone Miner. Res. 7(Suppl), S138 (1992). [Abstract] 174. J. M. Vogel, J. W. Davis, A. Nomura, R. D. Wasnich, and P. D. Ross, The effects of smoking on bone mass and the rates of bone loss among elderly Japanese-American men. J. Bone Miner. Res. 12(9), 1495 – 501 (1997). 175. P. Clark, F.de la Pena, F. Gomez Garcia, J. A. Orozco, and P. Tugwell, Risk factors for osteoporotic hip fractures in Mexicans. Arch. Med. Res. 29(3), 253 – 257 (1998). 176. E. Barrett-Connor and T. L. Holbrook, Sex differences in osteoporosis in older adults with noninsulin-dependent diabetes mellitus. JAMA 268(23), 3333 – 3337 (1992). 177. L. S. Lieberman, Diabetes and obesity in elderly black Americans. In “The Black American Elderly” (J. S. Jackson, ed.). Springer, New York, (1988). 178. A. V. Schwartz, M. L. Villa, M. Prill, J. A. Kelsey, J. A. Galinus, R. R. Delay, M. C. Nevitt, D. A. Bloch, R. Marcus, and J. L. Kelsey, Falls in older Mexican-American women. J. Am. Geriatr. Soc. 47(11), 1371 – 1378 (1999). 179. J. L. O’Loughlin, Y. Robitaille, J-F. Boivin, and S. Suissa, Incidence of and risk factors for falls and injurious falls among the communitydwelling elderly. Am. J. Epidemiol. 137(3), 342 – 354 (1993). 180. J. W. Davis, P. D. Ross, M. C. Nevitt, and R. D. Wasnich, Incidence rates of falls among Japanese men and women living in Hawaii. J. Clin. Epidemiol. 50(5), 589 – 594 (1997). 181. K. G. Faulkner, S. R. Cummings, D. Black, L. Palermo, C-C. Glüer, and H. K. Genant, Simple measurement of femoral geometry predicts hip fracture: The study of osteoporotic fractures. J. Bone Miner. Res. 8(10), 1211 – 1217 (1993). 182. S. R. Cummings, J. A. Cauley, L. Palermo, P. D. Ross, R. D. Wasnich, D. Black, and K. G. Faulkner, Racial differences in hip axis length might explain racial differences in rates of hip fracture. Osteopor. Int. 4, 226 – 229 (1994). 183. T. Nakamura, C. H. Turner, T. Yoshikawa, C. W. Slemenda, M. Peacock, D. B. Burr, Y. Mizuno, H. Orimo, Y. Ouchi, and C. C. I. Johnston, Do variations in hip geometry explain differences in hip fracture risk between Japanese and White Americans? J. Bone Miner. Res. 9(7), 1071 – 1076 (1994). 184. T. M. Theobald, J. A. Cauley, C. C. Gluer, C. H. Bunker, F. A. Ukoli, and H. K. Genant, Black – white differences in hip geometry: Study of osteoporotic fractures Research group. Osteopor. Int. 8(1), 61 – 67 (1998). 185. T. Beck, C. B. Ruff, W. W. Scott, C. C. Pluto, J. D. Tobin, and C. A. Quan, Sex differences in geometry of the femoral neck with aging: A structural analysis of bone mineral data. Calcif. Tissue Int. 50, 24 – 29 (1992). 186. T. J. Beck, A. C. Looker, C. B. Ruff, H. Sievanen, and H. W. Wahner, Structural trends in the aging femoral neck and proximal shaft: Analysis of the Third National Health and Nutrition Examination Survey dual-energy x-ray absorptiomerty data. J. Bone Miner. Res. 15, 2297 – 2304 (2000). 187. D. Nelson, D. Barondess, S. Hendrix, and T. Beck, Cross-sectional geometry and bone mass in the proximal femur in African-American and white postmenopausal women. J. Bone Miner. Res. 188. K. Chin, M. C. Evans, J. Cornish, T. Cundy, and I. R. Reid, Differences in hip axis and femoral neck length in premenopausal women of Polynesian, Asian and European origin. Osteopor. Int. 7(4), 344 – 347 (1997).
CHAPTER 23
Epidemiology of Osteoporotic Fractures in Europe CHRIS DE LAET JONATHAN REEVE
Institute for Medical Technology Assessment, Erasmus University, Rotterdam 3000 DR, The Netherlands Department of Medicine, Addenbrooke’s Hospital, University of Cambridge, Cambridge CB2 2QQ, England
I. Introduction II. Hip Fractures III. Vertebral Fractures
IV. General Conclusions References
I. INTRODUCTION
obtained from including subjects and/or populations at extremes of the range of experienced risk, as this tends to improve the power of any such study provided it is properly designed and executed. Partly with this in mind, the Mediterranean Osteoporosis Study (MEDOS) included a section in which 2816 cases of hip fracture were compared with 5369 controls. It was noteworthy that rates of hip fracture in the highest decile of age varied according to investigational center by an order of magnitude, while at the same time rates for men and women were highly correlated within the participating centers. All participating countries in MEDOS, apart from Portugal, have a Mediterranean seaboard [6]. At about the same time, methods were developed for the quantitative assessment of vertebral deformity, which made it possible to envisage large multicenter studies of vertebral fracture. This allowed a substantial group of European investigators from 36 centers in 18 countries to launch the European Vertebral Osteoporosis Study (EVOS) as a prevalence study [7] in population-based, age-stratified
The epidemiology of osteoporosis in Europe has been the subject of increasingly intensive investigational effort since mid-1970s [1]. Earlier work was largely based in single countries, particularly in northern Europe. Work such as that by Fenton-Lewis and colleagues [2] in Britain established that age-specific rates of hip fracture increased substantially between the 1950s and the 1980s and this has been confirmed for other European countries within a sometimes different time frame. It seems likely that a similar increase in age-specific rates of vertebral fracture has occurred in parallel [3]. This contrasts remarkably with the downward secular trend in vertebral fracture rates observed in the studies in Hiroshima, Japan, of Fujiwara [4]. Other work carried out in the 1970s and 1980s showed that there were considerable between-country differences in rates of osteoporotic fracture, which extended to differences between individual European countries [5]. Scientific advantage in cohort and case-control studies is to be
OSTEOPOROSIS, SECOND EDITION VOLUME 1
585
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
586
DE
cohorts of both sexes in the age range of 50 – 80. Most of the original EVOS centers, together with several new centers, have proceeded to follow their cohorts prospectively in the European Prospective Osteoporosis Study (EPOS), which has now been collecting fracture data for over 6 years. This chapter summarizes results of the considerable volume of recent epidemiological work done on hip and vertebral fractures in Europe. Europe is heterogeneous not only genetically but also because of the large variation in lifestyles, often different from those seen in the United States and Canada. However, it has been unclear whether there are true differences in fracture rates between North America and Europe or whether risk factors for fracture differ between continents. This is important, not so much because it may help us understand the etiology of fracture (comparative ecological studies of geographical differences in incidence often do not provide a very efficient platform for identifying true causes of disease). The importance of studying European data for Europeans is because the management of osteoporosis requires knowledge of the burden of disease as it affects the individuals and populations whose health is being cared for. Knowledge of the impact on quality of life of fractures is particularly important with the increasing competition for scarce health care resources. From the public health perspective, identifying risk factors for fracture is important in the near term for devising strategies for identifying individuals at high risk and in the longer term as part of a research strategy aimed at identifying and counteracting the biological pathways leading to osteoporotic fractures with populationbased approaches to prevention. For North Americans and others, there is much to be learned from published European data. This is for several reasons. Not only is the epidemiology of osteoporosis in men quite well advanced in Europe, but also some parts of Europe are given habitually to high rates of compliance with epidemiological studies, increasing confidence that such studies are reasonably representative of the populations from which the study subjects were drawn. Also, by comparing European with non-European studies it should be possible to derive new perspectives on the general validity of the conclusions and recommendations for action derived from both European and non-European studies.
II. HIP FRACTURES A. Current Incidence Currently, most hip fractures occur in Western industrialized countries, and Europe accounts for a large proportion of these. While the total number of hip fractures worldwide in 1990 was estimated at 1.7 million [8], 560,000 of those
LAET AND REEVE
occurred in Europe (including the former Soviet Union) and 360,000 in North America. Lifestyle and living conditions are more diverse in different countries around the European continent than they are in the United States. However, little information is available about the influence of racial differences on the incidence of hip fractures in Europe whether they might act through genetic or environmental mechanisms. Information about hip fractures is available for most countries, as virtually all hip fractures are treated during a hospital admission, making these fractures relatively easy to detect using discharge registries. The age-specific incidence of hip fractures in Europe shows a similar exponential increase with age as observed in other parts of the world, and in general, age-specific incidence is higher in women than in men after the age of 50. This higher incidence in women is reinforced by the fact that there are more elderly women than men who currently only account for approximately a quarter of the total. In 1998 the European Commission reported on the incidence of hip fractures in the 15 member states of the European Union (EU) [9]. Data were compiled from the most recent studies, including the previously mentioned MEDOS study, focusing on countries around the Mediterranean Sea and also from national studies[6,10 – 12]. Those data were then further extrapolated to the European population. Whenever country-specific data were lacking, information from neighboring countries was used [9]. For the year 2000, the total number of hip fractures in those 15 countries was estimated at over 400,000 [9]. Incidence showed a north – south gradient, with the highest incidence in Scandinavia and much lower incidences in the countries around the Mediterranean Sea. The incidence in Sweden was the highest; compared to the incidence of hip fractures in the United States [5], the relative incidence was 1.3 in women and 1.7 in men (own calculations based on the European Commission report). In Finland, by slight contrast, rates are comparable to U.S. rates in women but slightly higher in men. In the United Kingdom, the Netherlands, and Germany, the calculated incidence is very similar to the incidence in the United States. In southern European countries, the incidence was much lower: In France, Greece, and Spain, the incidence was about 70 % of the U.S. incidence, while in Italy and Portugal incidence was as low as 50% of comparable U.S. rates. Figures 1 and 2 (see also color plates) give yearly incidence rates/100,000 for a few selected countries. The MEDOS study also contained data on Turkey, a non-EU country with a highly diverse gene-pool, judged by its history of repeated population mixture in the last two millennia. Turkish data are quite different from data from the remainder of Europe. The low absolute and age-standardized rates approach the levels of some less developed countries and the increase with age was less pronounced than in European countries [13]. Especially in the rural
587
CHAPTER 23 Epidemiology of Osteoporotic Fractures in Europe
FIGURE 1
Age-specific incidence rates of hip fracture per 100,000 women in a few selected countries. (See also color plate.)
(Asia Minor) areas, but even in the principal cities of Istanbul and Ankara, incidence was very low. Incidence rates were about 10 – 20 % of comparable rates in U.S. in women and about 20 – 30% in men.
respectively). Again, Turkish data were strikingly different with a female to male ratio of about 1 in the large cities of Istanbul and Ankara, and a reversed ratio was found in rural areas (around 0.4), especially in older participants.
B. Gender Ratio
C. Future Trends
The global European female to male ratio for the total number of fractures was 3.7 for the year 2000. This high female to male ratio was largely due to demography, as women live longer than men. The female to male ratio of incidence rates was between 1.6 and 2 in most countries. Only in Sweden and Finland was a lower ratio found (1.3 and 1.4,
Future trends are influenced by both demography and trends in age-adjusted incidence. Demography is relatively easy to predict for the near future, all affected persons have already been born; but trends in hip fracture incidence are more difficult to predict and there are conflicting findings from different countries. An increase in the age-adjusted
FIGURE 2 Age-specific incidence rates of hip fracture per 100,000 men in a few selected countries. (See also color plate.)
588
DE
incidence between 1930 and 1980 was described in several countries and although the increase of those rates appeared to have leveled off in the 1980s in parts of the United States [14,15] and also in Sweden and the United Kingdom [2,5, 16], in other European countries, such as the Netherlands and Italy, the age adjusted incidence continued to rise during the beginning of the 1990s [5,17,18]. Because of this uncertainty, most predictions for the future are based on current incidence rates taking only demography into account. If the incidence continues to rise, the picture can only become bleaker. It is thought that the global number of hip fractures worldwide will increase to over 6 million by the year 2050. Although the high proportion of the world’s total of hip fractures that occur in Europe will decline in the next decades due to demographic growth in other parts of the world, the absolute number of hip fractures will also continue to rise in Europe to over 1 million in 2050. As in other areas of the world, the population in Europe is aging, mainly due to the post-World War II baby boom, which was followed by record low fertility rates, most extreme in Spain, Italy, and Germany. Moreover, with the continuing increase in life expectancy in both men and women there is a remorseless increase in the proportion of very elderly people. Over the next decades these trends will continue. Estimates of lifetime risk for hip fracture range from 13 to 18% in northern Europe and the United States, but taking into account expanding life expectancy, these rates are very likely underestimated [19]. The report from the European Commission [9] estimated that while the yearly number of hip fractures would increase to almost 1 million by the year 2050, the global female to male ratio within the European Union for the total number of fractures would decline from 3.7 for the year 2000 to about 3.2 in the year 2050.
D. Risk Indicators Although there is a multitude of known risk indicators for hip fractures, the real occurrence of a fracture remains a chance event and, by definition, chance events cannot be predicted accurately. Only the risk for the occurrence of fractures can be estimated. The list of known risk factors or indicators for hip, as it is for other fractures, is very long [20 – 24]. Broadly speaking, these may be grouped into indicators related to bone fragility and indicators related to the risk of a fall sideways onto the greater trochanter. If these are to be applied to selecting patients at high risk in populations a pragmatic approach should be followed, limiting this list of risk indicators to only those that are most important, both in terms of relative risk and in terms of their prevalence. In the setting of case finding, up until now the normal modus operandi of practicing physicians in this
LAET AND REEVE
field, success depends on an appreciation of the level of absolute risk faced by the individual patient in the consulting room. Thus a rare risk indicator that has a large impact on individual risk may be extremely important to the individual patient who is positive for that indicator. The successful clinician will identify such rare individuals, who may be missed by so-called “screening” of populations and therefore be at risk of not receiving counseling or treatment. If consensus is reached in the future that the screening of older populations is justifiable, the case-finding approach will remain important for this reason. Just as bone density measurements at different skeletal sites are correlated, many risk indicators of frailty are correlated, so that even in case finding one risk indicator can sometimes serve for several. The need in case finding or screening is to summate the overall risk faced by the individual and develop a risk reduction strategy if the absolute risk level is unacceptable. In its 1998 report [25], the American National Osteoporosis Foundation (NOF) proposed the use of a limited set of five risk indicators for the estimation of fracture risk in the setting of a population-based approach to screening women from the seventh decade of age and older. Those indicators were bone mineral density (BMD), a history of a prior fracture after the age of 40, a family history of fracture, “thinness” (which was categorized as the lowest quartile of body weight in U.S. white women), and current cigarette smoking. This selection was based on the large North American Study of Osteoporotic Fractures (SOF) [24] and so far has not been validated elsewhere. It also failed to include any validated indicator of the risk of falling. There is generally an inherent weakness in any such approach, which is that statistical analyses based on individual cohort studies tend to overestimate collectively the predictive power of the risk indicators derived from them. Because risk indicators are chosen purely because of their statistical associations with risk in the given study, each individual association has a certain level of imprecision or uncertainty. Also, in extrapolating the use of risk indicators to populations other than those from which the study population was drawn, differences between populations are likely to affect the relative strengths of different risk indicators. It is of considerable importance therefore to validate in other studies the strengths of the various associations selected, e.g., from the SOF study, for use in screening instruments. This has been addressed in the large Dutch ERGO (Rotterdam) study, in which family history and smoking were found not to be significant risk indicators [26], whereas other indicators, such as the use of a walking aid, were of increased apparent importance. Other groups have proposed alternative hip fracture risk assessment schemes, also based on specific cohorts [27–29]. While age and gender should be a necessary ingredient of any hip fracture prediction score, most such scores also include a measure of BMD and either low body weight or low body mass index (BMI:
CHAPTER 23 Epidemiology of Osteoporotic Fractures in Europe
weight/height2). Some studies, such as the European EPIDOS and Rotterdam Studies, captured the risk of falling by gait speed or by the use of walking aids, respectively, and these could be incorporated directly into a practical scoring instrument. Some scores, such as the NOF instrument, tried to incorporate the component of skeletal fragility, which is independent of bone mineral density: the indicator of this, which is usually preferred, is a prior fragility fracture, but ultrasound BUA has also been suggested for this purpose [27,30] and even biochemical markers of bone turnover [31–33]. The development of genetic markers is promising, and several candidate genes have been identified [34–37], but at the moment, apart from parental history of hip fracture, there is no practical routine way of capturing the genetic component of hip fracture risk. In any scoring system, a combination of these indicators will have to be used. A truly working instrument will probably have to include individual indicators of the principal etiological axes. Further, not only must risk indicators be strong, but in applying them to unselected populations they must be both universally applicable and of reasonably high prevalence. Moreover, it is essential for simplicity that these risk indicators each individually continue to have a significant contribution to fracture risk in a multivariate model, even when they are used in conjunction. These models will then have to be tested and validated in separate cohorts before any universal application can be recommended. Finally, in devising future efficient combinations of risk indicators for predicting hip fracture, more attention may have to be paid to the differences in etiology between intracapsular (femoral neck) and extracapsular (intertrochan teric) hip fracture. It seemed possible that current antiresorptive treatments, such as estrogens and bisphosphonates, may be more efficient at preventing extra- than intracapsular hip fractures. In the case of bisphosphonates this is because extracapsular hip fractures are associated more closely with low bone density and bisphosphonates appear decreasingly effective as bone density increases; and in the case of estrogens because their cessation leads to a more rapid increase in risk or intra- than extracapsular fractures [38]. Intriguingly, when Cummings and Palermo [38a] addressed this hypothesis they found the reverse to be the case. Also, comorbidities may affect relative risks, particularly coxarthrosis (hip osteoarthritis), which reduces the risk of intra- but not extracapsular fracture [39].
III. VERTEBRAL FRACTURES A. Pre-morphometry Studies of the Epidemiology of Vertebral Fracture Studies undertaken before the development of modern approaches to vertebral morphometry included some very
589 large studies, such as that of Santavirta et al. [40], who identified cases of thoracic vertebral fracture on miniature radiographs taken during screening for tuberculosis. The problems with these early studies were summarized succinctly by Kanis and McCloskey [41]. They centered on the variety of mutually incompatible methods used to identify cases. Sometimes clinical readings were used; otherwise one of a range of morphometric approaches was adopted. Concerning the morphometric methods, it was not always appreciated that when up to 39 vertebral height measurements are made on thoraciclumbar vertebral images from a single spine there is a 50%, not the anticipated 2.5%, chance that one of the heights that is in fact normal will be 2 SDs below the expected value. Therefore, many early studies using morphometric methods generated falsely high prevalence rates. The problems with using clinical readings were starkly revealed by the substantial disagreements between observers in a study comparing diagnoses in a representative sample of test X-rays sent blind to a number of expert radiologists and clinicians [42]. So it was generally agreed some 10 years ago that new studies would have to be undertaken using novel methodologies for identification of vertebral fracture cases. Against this background the European Vertebral Osteoporosis Study (EVOS) was conceived and implemented by a large group of European investigators.
B. Methodological Issues: The EVOS Study Because there is no central European source for comprehensive funding of multinational epidemiological studies for most categories of disease, central funding was available only for the design and analysis of results. The individual centers had to raise their own funds for participation, notably for recruitment, interview, and the X-ray of their subjects. A significant proportion of centers (13 of 36 in 19 countries) also performed dual X-ray absorptiometry (DXA) on the spine and/or hip of between 20 and 100% of their subjects. This was one practical example of the autonomy within the study of the individual centers, who all agreed to operate within a framework set by the study design partnership. At the study’s initiation the previously mentioned investigation of between-observer variation in the clinical identification of deformities on X-rays [42] made it mandatory that all X-rays should be assessed in a single center and that vertebral morphometry as well as clinical identification[43] should be carried out. In evaluating the vertebral morphometry results, the Eastell – Melton algorithm [44] was used alongside the McCloskey – Kanis algorithm [45]. In a pilot study, O’Neill et al. [46] showed that Eastell grade 1 (3 SD) deformities were more prevalent but less specific than McCloskey – Kanis 3 SD deformities and that Eastell grade
590
DE
2 deformities were of the same specificity but less prevalent. It was demonstrated by a three-center analysis that the use of a 3 SD cutoff to define a biologically significant deformity required the use of population- as well as vertebra level-specific normal ranges to define the appropriate cutoff [46]. At the same time as methods for defining vertebral deformities were being refined, considerable effort was devoted to validating the half-hour questionnaire for epidemiological risk factors [47], to establishing whether significant bias was likely to be introduced by varying nonresponse levels in different centers [48,49], and in demonstrating that the answers to the questions asked were reproducible on different occasions to a level that was scientifically acceptable [50]. In general, the questionnaire, which was constructed in close consultation with colleagues undertaking similar work in North America and elsewhere, performed well against these tests. There was no evidence of nonresponse bias substantially affecting the main findings of the study [48].
C. Vertebral Deformity Prevalence Results The main results of the study showed that there was significant variation in prevalence between centers, with rates for both men and women extending over a threefold range after age standardization [51]. Surprisingly, male and female rates were rather similar at over 10% using the McCloskey – Kanis 3 SD definition, with male rates being higher at the age of 50 and female rates being substantially higher than the male rates after the age of 75. Rates in Scandinavia were significantly higher than in other parts of Europe [51].
D. Vertebral Fracture Incidence Rates At the time of writing, the incidence data for vertebral fracture in the EPOS study have only been published in abstract form. An extremely rigorous quality assurance procedure was adopted to ensure that the number of false-positive fractures reported on the second (and the first) films was minimized. The approach adopted was essentially that described by Weber et al. [52] in the pilot study, with careful attention paid to ensuring that pairs of films were adjusted to the same magnification to avoid the artifactual appearance of vertebral height changes [53] and careful visual scrutiny by a senior radiologist of all candidate fractures for (i.e., point placement) diagnostic errors [54]. At age 65, women had a 1.0 – 1.1% risk of suffering an incident vertebral deformity in a 12 month period and this risk increased two-N-fold with each decade increase in age. Based on available North American data, these results do not so far support the hypothesis that American and
LAET AND REEVE
European incidence rates of vertebral fracture in women after the age of menopause are substantially different. Their rates of increase with age also appear similar. The principal difference between the results of the European incidence and prevalence studies concerned men. Before age 60, in the European prevalence study [51], male deformity rates were actually higher than those of women. This difference contributed to the overall similarity of male and female prevalence rates from age 50 to 80, since as expected female rates were substantially higher than male rates over age 70. When risk factors for prevalent deformity were incorporated into models of deformity risk, men appeared to be at twice the risk compared to women of a prevalent vertebral fracture for any given BMD value. In contrast, after adjusting for age and bone density, the risk of an incident fracture was similar in the two genders. The most plausible explanation for these findings is that the gender difference in incident fracture rates in the over 50s depends principally on the postmenopausal bone loss seen in women [55]. The higher prevalence of deformities experienced by men compared to women at younger ages, which is supported by previous single country studies [4, 40], suggests that men are at increased risk of vertebral fractures under the age of 50.
E. Epidemiological Risk Factors So far, the principal sources of data have been prevalence studies, in particular the EVOS study. In EVOS, analyses have been undertaken of both risk of vertebral deformities and risk of low bone density. Concerning risk of vertebral deformity, an important determinant of risk was found to be body mass index (BMI weight /height2), with leanness being a significant positive risk factor [56]. A long fertile period (interval between menarche and menopause) was found to be significantly protective [57]. Also protective was past use of the oral contraceptive pill [57]. The protective effect of these estrogen-related factors on bone density was also demonstrated [58]. In women, measures of physical activity demonstrated a protective effect of moderate exercise against vertebral deformity, but in men, data suggested that this effect was biphasic. Intensive occupational physical activity was associated with increased risk of a vertebral deformity, whereas physical inactivity was associated with a (nonsignificant) increase in risk by comparison with men who walked or cycled for half an hour or more a day [59]. This suggested the possibility that trauma may play a role in the development of vertebral deformity in men younger than 50 years of age if they had previously experienced very high levels of load-bearing physical activity, perhaps applied to their vertebral column. Moderate alcohol
CHAPTER 23 Epidemiology of Osteoporotic Fractures in Europe
consumption, as in previous studies, was found to be somewhat protective against spinal deformity [60]. However, the presence of menopausal symptoms was unrelated to risk of vertebral deformity [61]. Lifestyle, gynecological, and environmental risk factors for incident vertebral fractures at the time of writing were still being evaluated in the EPOS incidence study. In confirmation of previous studies [62,63], a prevalent fracture was a strong determinant of incident vertebral fracture [54], even after adjusting for BMD. Large fractures (e.g., crush fractures as contrasted with smaller wedge deformities) conveyed a greater increase in risk of future fractures. Beyond that, a new finding, which might be important in relation to the clinical impact on future quality of life, was that the size of a new fracture in terms of vertebral body height lost (calculated as the sum of losses of all three measured heights) could be predicted from the clinical classification of a previous fracture as a wedge, a single end plate concavity, a double end plate biconcavity, or a complete crush fracture [64].
F. Genetic Determinants of Risk Twin studies and studies of mother daughter pairs have made it clear that there is a substantial heritable component to both fracture risk and adult bone density[65]. It is not clear, however, to what extent these associations depend on gene – environment interactions, for which there is some preliminary evidence [66]. The recent doubling of age-specific hip fracture rates in one generation in Europe clearly has environmental or lifestyle rather than directly genetic causes. However, this must not be allowed to detract from our understanding of the potential of heritable factors to explain variations in fracture risk within the present generation. Efforts are currently underway to use candidate gene and genome searching techniques to identify genes of potential importance. We know that concerning risk of vertebral fractures, fractures in a close relative are of predictive value, as a parental history of hip fracture is a positive risk factor for vertebral deformity [67]. So far there have been no large European multinational studies of the genetic determinants of vertebral fractures, but in two smaller single country studies [68,69], the SP1 polymorphism (genotypes sS and SS combined) in the regulatory region of the COL1A1 locus has been associated with a markedly increased risk of vertebral deformity of the order of two-to three-fold. This is particularly intriguing because this polymorphism is only weakly associated, if at all, with low bone density, and most of the known environmental and endocrine risk factors for fracture are associated with equivalent effects on bone density.
591
G. Low Bone Density in the Assessment of Fracture Risk Prevalence studies such as EVOS can provide important clues as to the true determinants of osteoporosis but cannot prove a causal link between risk factors and outcomes. For this reason, 13 out of 36 centers in EVOS explored the role of low bone density in the determination of vertebral deformity risk by measuring spine and hip bone density using DXA in as many as possible of their subjects. The 13 centers were equipped with bone densitometers from four different manufacturers: Hologic, Lunar, Norland, and Sopha. Because these machines all reported bone density in units that are normally not cross-calibrated between machine brands, the bone density studies in EVOS were undertaken in close collaboration with a contemporary concerted action: “The Quantitative Assessment of Osteoporosis” (QAO) [70]. The QAO study developed a semianthropomorphic phantom, the European Spine Phantom Prototype (ESPp), for the cross-calibration of bone densitometers [71 – 73]. With this phantom, most of the systematic differences between the different brands of densitometer were eliminated and, at the same time, differences in initial setting up between different densitometers of the same brand were removed [72]. The EVOS bone density study was one of the first to apply DXA bone densitometry to population samples of men and women aged 50 – 80. We found first that there were substantial differences between centers in mean bone density in the spine, the femoral trochanter, and the femoral neck even after adjusting for age, sex, height, and weight [74]. Apparent rates of change of bone density with age in this crosssectional study were also very different between centers. The size of these differences was surprisingly large; in relation to the overall standard deviation of the combined European population of subjects studied, the range of mean values at these three sites in women typically extended over half a standard deviation (Fig. 3) The size of these betweencenter differences had major implications for the diagnosis of osteoporosis using densitometry cutoffs such as the one proposed by the WHO study group [75]. Another important finding was that the population SD of bone density was dependent on both age and weight. Not only did the spread of values seen in our populations increase with age (with the exception of the femoral neck in women), but weight variation had a surprising and potentially important effect. A low body weight in women was associated with an increase in the standard deviation of bone density as well as being associated with lower mean bone density. This effect is illustrated in Fig. 4. The increasing prevalence of osteoporosis (as defined using the WHO cutoff of 2.5 SDs in relation to a young normal population) increases progressively with declining body weight as the result of a combination of the
592
DE
LAET AND REEVE
Mean density at the spine ( ), femoral neck (), and trochanter () in women in each center, after adjusting to age 65, height 1.65 m and weight 70 kg. Bone densitometers were cross-calibrated with the European Spine Phantom prototype (ESPp). Centers are shown in order of decreasing femoral neck bone density. Ab, Aberdeen (GB); Be, Berlin (DE); Bu, Budapest (HU); Ca, Cambridge (GB); Er, Erfurt (DE); Gr, Graz (AT); Ha, Harrow (GB); He, Heidelberg (DE); Le, Leuven (BE); Mal, Malmo (SE); Man, Manchester (GB); Mo, Moscow (RU); Os, Oslo (NO); Ov, Oviedo (ES); Pi, Piestany (SK); Ro, Rotterdam (NE). From Lunt et al. [74], with permission.
FIGURE 3
effects of weight on mean bone density and on the SD of bone density. These bone density studies also allowed us to explore the association between low bone density and vertebral deformity risk [55]. First we classified the subjects in the study according to whether they had a deformity or whether they were free of vertebral deformities according to the results of using one of the two algorithms [44,45] adopted for use in the study. In a preliminary evaluation we found the risk of being a ‘case’ defined by the McCloskey – Kanis algorithm was more strongly predicted by low bone density [76]; therefore, in the discussion that follows, we shall refer to the use of the McCloskey – Kanis algorithm with a 3 SD population-specific cutoff. In the morphometric assessment of vertebral deformity, three heights are measured per vertebra. These are placed anteriorly, posteriorly, and approximately midway between vertebral endplates [43]. This allows the classification of vertebral deformities into six groups: vertebrae with anterior height loss, those with loss of anterior midbody height, those with loss of all three heights, those with loss of midbody height only, and, less frequently, those with posterior and midbody height loss or (unusually) posterior anterior height loss without apparent loss of midbody height. It was found that the principal difference between genders was that bone density was related more strongly to loss of all three vertebral heights (crush fractures) in
FIGURE 4 Predicted prevalence of female femoral neck bone density values below 0.580 g/cm2 after adjustment to age 65 according to the statistical model. This is the value presented by Pearson et al. [71] for a mixed population of normal young European women, which represents a T score of 800 mg/day) and high ( 800 mg/day) calcium intake [1]. Users of thiazide diuretics tend to be more obese than nonusers, and because the degree of obesity is inversely associated with the risk of hip fracture, the effect of thiazide diuretics on fracture risk may be obscured by inclusion of the extremely obese. To test this hypothesis, we excluded obese women (body mass index of 31.5 kg/m2, which corresponds to the 85th percentile) [1]. Results were similar to those derived from the analysis of the whole cohort. None of the relative risks was statistically significant, but they did show a trend toward a protective effect of thiazide diuretics against hip and wrist fracture, suggesting that inclusion of the very obese had little effect on the association between TD and fracture.
CAULEY AND SALAMONE
C. Underlying Mechanisms The mechanism by which TD may preserve bone mass and protect against hip fractures is not known. Thiazide diuretics have been shown to influence calcium metabolism. A reduction in the urinary excretion of calcium [14,32], increases in serum calcium [33], and decreased parathyroid hormone (PTH) [34] concentrations, accompanied by reduced serum concentrations of 1,25-dihydroxyvitamin D [(1,25(OH)2D)] and alterations in calcium absorption, have all been reported [35]. Thiazide diuretics may also decrease bone resorption [32] and bone turnover [36]. A recent animal study suggested that thiazide diuretics cause an increase in bone mineral crystallization [37]. Thiazide diuretics may decrease fracture risk by preserving bone mass. As reviewed earlier, cross-sectional studies have shown that TD users have slightly higher cortical and trabecular bone mass than nonusers. If thiazide diuretics reduce the risk for fracture by preserving bone mass, then a statistical adjustment for bone mass should attenuate the relation between thiazide use and the decreased risk for fractures. However, inclusion of the distal radius bone mass did not substantially change the relation between thiazide use and the reduced risk for most fractures, including hip fracture, despite the fact that distal radius bone mass predicts hip fracture and osteoporotic fractures [1]. For wrist fractures, a modest change was observed in the relative risk when the distal radius bone mass measured near the site of wrist fracture was included in the model. It is possible that the inclusion of bone density at the hip may have attenuated the effect of thiazide diuretics on hip fracture. However, TD could aggravate the risk for fractures by increasing the risk for falls. However, the association of diuretic use with falling is not consistent [38,39]. The Study of Osteoporotic Fractures found no association between TD use and risk of falls [1]. The age-adjusted odds ratio for experiencing two or more falls in the first year after baseline current users of TD compared with never users was 1.06; CI, 0.90 to 1.20.
D. Summary and Future Research Cross-sectionally, current thiazide diuretic users have higher axial and appendicular bone mass than either past or never users. This observation has been demonstrated consistently for both axial and appendicular bone sites and for both men and women. The magnitude of the effect of TD on bone mass is about 4 to 5%. Prospectively, thiazide diuretics preserve bone mass, although this observation is limited to appendicular bone mass. Over the short term, TD may abate the seasonal decrease in BMD observed in the fall and winter months.
749
CHAPTER 30 Postmenopausal Endogenous/Exogenous Hormones
The effect of TD on all fractures, including all nonspinal or osteoporotic, i.e., those fractures associated with low bone mass or prevalent vertebral fractures, has not been consistent. For hip fractures, thiazide diuretic users tend to be at reduced risk. The underlying mechanism(s) whereby TD preserve bone mass and prevent fractures has not been established. The effect of TD on hip fractures could not be explained by the higher bone mass of longterm users. There is a paucity of therapeutic agents currently approved for the prevention of osteoporotic fractures. Based on the literature it appears that thiazide diuretics may prevent hip fractures and preserve bone mass at least in elderly women. Jones et al. [31] have suggested that the results of the meta-analysis may preclude the necessity of doing a clinical trial on thiazide diuretics and hip fracture. However, we have very little data on the effect of TD on bone mass and fracture among normotensive women. We and others [1,40] believe that a randomized clinical trial on the effect of TD on the incidence of fractures is the only way to resolve the issue. It is conceivable that this trial could be done in osteopenic women with prevalent vertebral fractures. Focusing on these high-risk women would lead to a decrease in the required sample size and make the trial more feasible. Until such a trial is completed, it is premature to recommend thiazide diuretics for the prevention of osteoporosis and fractures.
III. EXOGENOUS ESTROGEN An estimated 31.7 million prescriptions for oral menopausal estrogens were dispensed in 1992, representing a greater than twofold increase since 1982 [41]. The use of transdermal estrogen also increased, although the absolute number of prescriptions was considerably lower (4.7 million in 1992). Prescriptions for medroxyprogesterone increased almost fivefold from 2.3 million in 1982 to 11.3 million in 1992. These trends demonstrate the vast exposure to postmenopausal estrogens and progestins. Despite the large number of prescriptions that are written, it is estimated that 20 – 30% of postmenopausal women never filled their prescriptions, 10% of users reported only intermittent use, and another 20% discontinued their therapy within 8 months [42] Compliance to HRT remains an issue. A more recent estimate is that fewer than 20% of postmenopausal women take HRT [43]. Estrogen is the cornerstone of preventive therapy for osteoporosis in menopausal women. It is endorsed by many expert panels as the first line of therapy for osteoporosis, including the National Osteoporosis Foundation [44]. More recently, however, the package insert was changed to reflect an approved indication for prevention and not treatment.
A. Estrogen and Bone Mass The effectiveness of estrogen in reducing bone loss is well established [45,46] (see also Chapter 69). Estrogens reduce bone loss by slowing bone turnover, which may be mediated by estrogen receptors in osteoblasts [47]. Most research has examined the effect of oral estrogens [45,46], but estrogens given transdermally [48,49] or by implants [50,51] are also effective. The estrogen dose, but not route of administration, influences the effectiveness of estrogens [45]. Combination (estrogen plus progestins) therapy and monotherapy (unopposed estrogens) have similar effects on bone mass [52]. Analyses of the correlates of baseline BMD among the 875 women recruited into the Postmenopausal Estrogen/Progestin Intervention Trial (PEPI) showed no additional benefit of progestins taken in conjunction with replacement estrogens [53]. The women in PEPI were more likely to have taken medroxyprogesterone acetate, and the authors point out that the effect of contraceptive androgenic progestins on BMD is unknown [53]. Progestins given alone have also been shown to reduce bone loss, particularly loss of cortical bone, but there is little evidence that combination therapy has additive or synergistic effects on bone mass [54]. The minimum duration of estrogen therapy for the prevention of bone loss is not known, but Lindsay [45] recommended a duration of 10 or more years or lifelong for those with established disease. Estrogen appears to be an effective therapy for women with established osteoporosis [55 – 59]. In one study the effect was greatest among those with the lowest bone mineral density [57]. The effect was similar for women on combination therapy [59] and in women receiving transdermal estrogen [58]. An important controversy regarding estrogen and bone mass is whether it is effective in very old women. Data from Framingham [60] showed little protective effect from an average of 10 years estrogen therapy on bone density among women 75 years of age and older. However, failure to see an effect of estrogen in the Framingham women may have reflected the fact that most of the estrogen use occurred in the past. Resnick and Greenspan [61] suggested that therapies that slow bone loss may be less effective among older women whose bone loss has slowed and whose bone is of poorer quality. Estrogen has been shown to be effective in preserving bone mass in elderly women [62,63]. As part of the Study of Osteoporotic Fractures (SOF), rates of bone loss increased among the very old women (age 80), but even among this group, rates were significantly lower in women who reported estrogen use [64]. There is little agreement on when to initiate hormone therapy. Initiation of hormone therapy at menopause effectively prevents bone loss; however, once a woman
750 discontinues her estrogen, bone loss quickly resumes. If a woman continues her estrogen, concern about breast cancer associated with long-term use may cause her to discontinue her therapy. Schneider et al. [65] demonstrated that women who initiated estrogen before age 60 and continued had the highest BMD. Of importance, however, women who initiated therapy after age 60 and continued had BMD that was significantly higher than women who had never used estrogen or discontinued estrogen. These data suggest that women can delay their use of estrogen and still receive a bone-sparing effect on BMD. This study, however, was observational and did not include data on fractures. 1. ESTROGEN AND BONE MASS: RANDOMIZED TRIALS The majority of studies of estrogen and BMD data published prior to 1990 relied on observational or uncontrolled studies. More recently, there have been nine randomized placebo-controlled trials of estrogen and BMD. The largest of these studies was the PEPI study, which enrolled 875 women, mean age 56 years, into a 3-year trial [66]. There were five treatment arms: placebo, conjugated equine estrogen (CEE) 0.625 mg/day, CEE with medroxyprogesterone acetate (MPA) given cyclically (5 mg, 12 days per month), CEE with MPA given continuously (2.5 mg/daily), and CEE with micronized progesterone (200 mg/day) given 12 days per month [66]. Women randomized to placebo lost hip and spine BMD. Hip BMD increased about 1.7% in all active treatment groups (all p0.05 versus placebo). Spine BMD also increased in all active treatment groups. The percentage increase in hip BMD was significantly greater for the CEE-MPA (cont) at 5% compared with an increase of 3.8% in the other treatment groups. Among a group of older women, all age 65 years, a lower dose of CEE (0.3 mg/day) was found to significantly increase spine BMD 3.5% over 3.5 years of observation in comparison to placebo [67]. All of the women had adequate calcium and vitamin D intake. Two randomized dose-ranging trials of estrone-sulfate (estropipate) in newly postmenopausal women (mean 51 years) demonstrated a significant increase in spine BMD measured by quantative computed tomograph (QCT) [68,69]. The increase in spine BMD was primarily observed in women who were randomized to higher doses of estrone-sulfate (0.625 or 1.25 mg/day). Oral micronized estradiol given to 63 postmenopausal women at three different doses resulted in a significant increase in spine BMD of 1% and hip BMD of 2% compared to placebo after 12 months of therapy [70]. There have been two randomized controlled trials (RCTs) of transdermal estradiol. Thirty-four women, followed for 18 months, were randomized to either placebo or 50 mg transdermal estradiol [71]. Forearm BMD increased about 4% in the estrogen arms and decreased 3.5% in the
CAULEY AND SALAMONE
placebo arm. In a dose-ranging study of transdermal estradiol, a minimum dose of 0.05 mg estradiol transdermal is needed to influence spine BMD [72]. A total of 406 women, mean age of 51 years, were randomized to ethinyl estradiol (Ee) — 0.3, 0.625, 1.25 mg, or placebo — and followed for 24 months. Of importance, all doses of Ee showed a significant increase in BMD compared to bone loss in the placebo group [73]. Combinations of Norethindrone acetate (NETA) and ethinyl estradiol were evaluated in 1265 women, mean age 52, for their effects on spine BMD (QCT) after 24 months [74]. The placebo group lost 7.4% of their BMD. All combination groups except the lowest dose group had lost significantly less BMD than the placebo group. An increase in BMD was observed for the combination 1.0 mg NETA and 5g Ee and 1.0 mg NETA and 10g Ee groups.
B. Estrogen and Fractures: Observational Data Current users of estrogen have a statistically significant decreased risk of hip, wrist, and spine fractures [75 – 86]. A recent meta-analysis suggested a 25% decrease in the risk of hip fracture in postmenopausal women who reported ever using estrogen [87]. Several studies reported the relative risk for hip and wrist fracture combined [75,76]. Examination of the effect of ERT on wrist fracture alone revealed about a 60% reduction in the risk of wrist fracture [81,84,86]. The relative risk of vertebral fractures among estrogen users was 0.60 (95% CI, 0.36 to 0.99) [84]. For all fractures, the relative risk was reduced by about 35% [86] to 50% [80] among current ERT users. Most studies have examined the relation between unopposed estrogen and fractures [75 – 79,81,82]. In one cohort study, women who reported using the combined estrogen and progestin regimen constituted about 40% of their users who initiated use before age 60 and 20% of users who initiated use after age 60 [85]. Results of this study showed about a 30% reduction in the risk of hip fracture among women using estrogen, suggesting that the high prevalence of combination users did not influence their results. This study did not however, compare the relative risks separately for unopposed estrogens and combination therapy. In the Study of Osteoporotic Fractures, we compared the relative risk of all nonspine fractures and wrist fractures between current users of unopposed estrogen and current users of estrogen plus progestin [86]. Among current users, the effect of unopposed estrogen was similar to that of estrogen plus progestin for wrist fracture and all nonspinal fractures. The multivariate-adjusted relative risk for wrist fracture among current unopposed estrogen users was 0.42 (CI, 0.25 to 0.70), and this risk among current users of estrogen plus progestin was 0.31 (CI, 0.11 to 0.84). For all nonspinal fractures, the multivariate-adjusted relative risk
751
CHAPTER 30 Postmenopausal Endogenous/Exogenous Hormones
was 0.69 (CI, 0.56 to 0.86) among users of unopposed estrogen and 0.51 (CI, 0.33 to 0.78) among users of estrogen plus progestin. We were unable to look specifically at hip fractures because the prevalence of use of the combined regimen was low in our cohort. The decrease in fracture risk associated with estrogen use is greatest among current or recent users, and the decreased risk tends to diminish with time after stopping estrogen [76,78,83]. In the Study of Osteoporotic Fractures cohort, we found no association between previous use of estrogen and the risk of either hip fracture or all nonspinal fractures [86]. Although previous users had a 20% decrease in the risk of wrist fracture, the confidence intervals included 1.0. Moreover, there was no association with fracture even among previous users who had used estrogen for 10 or more years. Previous users who initiated use within 5 years of menopause had a slightly lower risk of wrist and all nonspinal fractures, but these results were not statistically significant. These results imply that women need to continue estrogen therapy to reduce their risk of fracture. In most studies, a longer duration was more beneficial in reducing fracture than a short duration [76,77,79,86]. The protective effect of estrogen was more pronounced if estrogen was initiated around menopause [75,82,86]. In the large cohort study from Sweden, initiation of use after age 60 was not associated with a reduced risk of hip fracture, although these women tended to take less potent estrogens [85]. Initiation of use within 5 years of menopause among long-term users was associated with a 50% reduction in the risk for all nonspinal fracture, but no effect was noted if estrogen was initiated more than 5 years after menopause, even among long-term users [86]. The age-adjusted risk for all nonspinal fractures among current long-term users of estrogen who initiated estrogen within 5 years of menopause was 0.53 (CI, 0.37 to 0.74), and this risk was 0.97 (CI, 0.56 to 1.68) in current long-term users who initiated estrogen more than 5 years after menopause [86]. Relatively few studies have examined whether the dose of estrogen influenced the results on fractures. Three studies examined the effect of estrogen dose and fracture and found that the dose of estrogen had little effect on the degree of protection [20,76,83]. In the Leisure World Study, the age-adjusted relative risk of hip fracture was similar among women reporting a dose of 0.625 mg (RR 0.84; 0.58 to 1.21) and women reporting a dose of 1.25 mg (RR 0.91; 0.64 to 1.29) compared with never users [83]. Krieger et al. [78] combined dose with duration of use and computed a milligram month variable, which was defined as the total number of months of use times the milligrams of estrogen in that preparation [78]. Women who were exposed at a level of 50 mgmonth or greater were significantly less likely to fracture.
Several factors that could potentially modulate the effect of ERT on fractures have been examined: age, cigarette smoking, degree of obesity, type of menopause, and history of osteoporosis. 1. AGE In two prospective studies [82,85] and one case-control study [79] the protective effects of estrogen on fracture were greater in younger women (defined as either age 80 [79] or age 75 [82]) and weaker [79,82] among the older women. However in both the Framingham Study [82] and the Mediterranean Osteoporosis Study [79], the point estimate of the relative risk suggested a 20 – 30% reduced risk of hip fractures among the oldest women, but confidence intervals were wide and included 1.0. The failure to reach statistical significance may have reflected reduced power in the oldest age group, although it is possible that the degree of protection could still be reduced in older women. In the Study of Osteoporotic Fractures cohort, the association between estrogen use and risk for all nonspinal fractures was similar in those younger (RR 0.66, 0.53 to 0.84) and older than 75 years of age (RR 0.65; 0.43 to 0.97) [86]. The effect of estrogen on wrist fracture was also similar in younger and older women, but the confidence intervals were wider in the older age stratum because of the smaller number of women. We found an 80% decrease in the risk for hip fractures among women older than 75 years of age, and we found no effect on hip fracture in those 75 years age or younger. Once we had excluded women with a history of osteoporosis, the relative risk for hip fracture associated with current estrogen use was decreased in older and younger women, but confidence intervals included 1.0. 2. SMOKING Data from Framingham showed that among current smokers, estrogen use was not associated with a reduced risk of hip fracture (OR, 1.26; CI, 0.29 to 5.45) [88]. This is in contrast to an odds ratio for hip fracture among current nonsmoking estrogen users of 0.37 (CI, 0.19 to 0.75) [88]. In the Study of Osteoporotic Fractures, we did not confirm this finding. Current smoking did not attenuate the effect of estrogens on fractures [86]. Similarly, in the casecontrol study of Williams et al. [89], the benefit in preventing hip fractures was greatest in thin women who reported current cigarette smoking. 3. DEGREE OF OBESITY Estrogen users tend to be less obese than nonusers [2] and thin women are at increased risk of fracture [82]. Hence, there may be an interaction between estrogen use and obesity. The beneficial effects of estrogen on hip and wrist fractures varied by the level of obesity and smoking, with the greatest benefit achieved in thin nonsmoking women [89]. This is inconsistent with the observation of
752 the women recruited into the PEPI trial that the increase in BMD with increasing body mass index (BMI) was similar among estrogen users and nonusers. There was no interaction between estrogen and the degree of obesity as measured by the body mass index [53]. However, women with more than moderate obesity (BMI 35) were excluded from PEPI. 4. TYPE OF MENOPAUSE The case-control study of Paganini-Hill and colleagues [20] suggested that the effect of estrogen was greatest among oophorectomized women. The relative risk of hip fracture among oophorectomized women who reported using ERT for more than 5 years was 0.14 compared to 0.86 among women who had intact ovaries. To our knowledge, none of the other studies have been stratified by history of oophorectomy. These oophorectomy data were not adjusted for any other factors. Hence, this observation may reflect a failure to control for other differences between women with a history of oophorectomy and women with intact ovaries such as age at first use, or total duration of use. It is possible that women undergoing a surgical menopause initiate use soon after the surgery and hence earlier in their menopausal transition. The total duration of use may also be longer among oophorectomized women. 5. HISTORY OF OSTEOPOROSIS In The Study of Osteoporotic Fractures, we found that current use of estrogen was associated with similar reductions in the risk for wrist and all nonspinal fractures among women with and without a history of osteoporosis [86]. Current use of estrogen was not associated with a statistically significant reduction in the risk for hip fracture in women reporting a history of osteoporosis, although the reduction in risk in those without osteoporosis was statistically significant. Women with osteoporosis are more likely to take estrogen. Use of estrogen may be a marker of more severe osteoporosis, which could lead to an underestimation of the effect of estrogen in that group. 6. ESTROGEN AND FRACTURES: RANDOMIZED TRIALS There have been few randomized trials of hormone replacement and fractures (Table 3). Nachtigall et al. [90] followed 84 pairs of institutionalized women who were treated daily with placebo or conjugated estrogen 2.5 mg/day and medroxyprogesterone, 10 mg per day, 7 days per month for 10 years: there were seven fractures in the control group and none in the treatment group. Lindsay and colleagues [91] studied 100 women who were enrolled in a longitudinal study of bone loss following oophorectomy; about onehalf of the women were prescribed mestranol (average dose, 23 g/day) and half placebo. After a median followup of 9 years (range 6 – 12 years), estrogen treatment significantly reduced the incidence of vertebral compression.
CAULEY AND SALAMONE
A 1-year trial of 75 postmenopausal women with established osteoporosis reported a 40% reduction in the radiographic vertebral fracture rate in the group treated with 0.1 mg/day of transdermal estrogen compared with placebo [58]. However, a reanalysis of these data based on the number of subjects with at least one new vertebral fracture (not fracture rate) produced a nonsignificant result [92]. In a randomized trial of 464 nonosteoporotic Finnish women, HRT significantly reduced the risk of nonvertebral fractures by about 60%, after adjusting for BMD and history of previous fractures [93]. To our knowledge, this is the first randomized trial of HRT on clinical fractures. Nevertheless, the hormone regimen used in this study is not approved in the United States. The Heart and Estrogen/progestin Replacement Study (HERS) was a randomized, double-blind, placebo-controlled secondary prevention trial of daily use of conjugated equine estrogen plus medroxyprogesterone acetate (progestin) (Prempro) on the occurrence of nonfatal myocardial infarction or coronary heart disease death among women with documented heart disease [94]. Fractures were a secondary end point to the HERS trial. There was no evidence that 4 years of treatment with Prempro substantially reduced the incidence of fractures or height loss among these older, nonosteoporotic women who all had coronary heart disease [94,95]. How do we interpret the differences in the results of the randomized trials of HRT on fractures and observational studies? First, HERS is the largest RCT trial of HRT but was not designed to look at the primary effect of hormone therapy on fractures. Of importance, HRT may be more effective in preventing fractures among women who have osteoporosis. Second, the duration of the trials were relatively short, 4 years. A longer duration of HRT use may be needed to prevent fractures. Finally, the effect of HRT on vertebral fractures in HERS was limited to painful, clinical fractures, which represent only about one-third of the total number of fractures [95]. The earlier study of Lindsay et al. [91] did report an effect of estrogen on vertebral compression. The lack of randomized trials of HRT on fractures emphasize the need to clarify the effect of HRT on fractures in women, both those with and without osteoporosis. Nevertheless, given the data from the RCT of HRT and bone density, it seems reasonable to assume that HRT will preserve bone density, which should ultimately result in a reduction in fractures.
C. Summary and Future Research Current use of estrogen reduces the risk of all fractures; past use has little effect on fracture risk. Estrogen appears to reduce the risk of fracture among older women who smoke. The effects of unopposed estrogen and the
TABLE 3 Summary of Randomized Clinical Trials of Hormone Therapy and Fracture First author (year)
Intervention
Subject charactistics
Results
CEE 2.5mg daily and MPA 10 mg daily for 7 days (CYC) (n 67) Placebo (n 62) 10 years
Age matched 76.7% completed study 40%, 3 years since menopause 10 years
Number of fractures CEE and MPA 0 Placebo, 7
Lindsay (1980)
Mestranol average dose 23.3 g day (range 10 to 38 g) (n 58) Placebo (n 42) 9 years (range: 6 to 12)
Oophorectomized Subset of original trial: women who attended clinic visits consistently and for longest period of time Mean age about 48 years
Vertebral morphometry Mestranol: 4 – 7% greater anterior height (p 0.10) Mestranol: 34 – 56% smaller wedge angle (p 0.05) Mestranol: 9 – 13% greater ratio of central height to anterior height (p 0.06)
Lufkin (1992) Windeler (1995)
Estraderm 0.1 mg estradiol plus MPA 10 mg/day for 10 days (n 36) Placebo (n 39) 1 year
Osteoporotic women: prevalent vertebral fracture or BMD 10 percentile of normal premenopausal women All Caucasian Mean age 65 years (55 to 72)
Vertebral morphometry Number of new fractures PBO: 20 new fractures in 12 women E2 MPA: 8 new fractures in 7 women Fracture rate per 100 PY PBO: 58 E2 MPA: 23 (p 0.04) Number of women with 1 new fracture PBO: 12 E2 MPA: 7 (p 0.28)
Komulainen (1998)
E2Valerate 2 mg (days 1 – 21) and CPA 1 mg days 12 – 21 (n 116) Vitamin D 300 IU and 93 mg calcium (n 116) E2Valerate plus vitamin D (n 116) Placebo (n 116) 4.3 years
Subset of Kuopio Osteoporosis Study 6 – 24 months since last menstruation Nonosteoporotic: t score 2.5 Mean age 53 years (47 to 56)
All symptomatic nonvertebral fractures ITT RR (95% CI)a E2Valerate CPA 0.38 (0.15 – 0.99) Vitamin D 0. 64 (0.29 – 1.42) E2Valerate CPA vitamin D 0.48 (0.19 – 1.18) Placebo 1.0
Cauley (2000)
0.625 mg CEE in combination with 2.5 MPA daily (n 1380) Placebo (n 1383) 4.1 years
Documented Coronary Heart Disease 89% Caucasian Postmenopausal Intact uterus Mean age 67 years (44 to 79)
ITT RR (95% CI) All clinical fractures 0.94 (0.75 – 1.19) Hip 1.09 (0.51 – 2.31) Wrist 0.97 (0.58 – 1.62) Spine 0.69 (0.34 – 1.40)
753
Nachtigall (1979)
Notes: ITT, intention to treat; CEE, conjugated equine estrogen; MPA, medroxy progesterone acetate; PBO, lacebo; E2, estradiol; CPA, cyproterone acetate. a Adjusting for BMD and previous fractures.
754 combination of estrogen plus progestin are similar. For optimal protection from fractures, estrogen use should be initiated early in menopause and continued indefinitely. If women initiate estrogen at menopause and continue it indefinitely they may be exposed to it for about 30 years. Further studies are needed to evaluate the overall risks and benefits of such a prolonged use of estrogen. In addition, further studies are also needed on the effect of estrogen on fracture risk if estrogen is initiated much after menopause. Ettinger and Grady [96] have suggested that initiating estrogen at age 65 will result in a reduced risk of hip fracture. Some of the protective effects of estrogen observed among long-term users may reflect a compliance bias. These longterm users by definition are compliant women. Dr. Petititi has shown that among women assigned to receive a beta blocker or placebo, the mortality rate was lower among the women who had 75% compliance irrespective of whether they were assigned to either the study medication or the placebo [97]. Thus, hormone therapy,-especially long-term therapy,may be a marker of socioeconomic, clinical, or lifestyle factors that place estrogen users at a lower risk of fracture. Compliance to estrogen therapy is also important when considering its long-term effectiveness. The low compliance rate to estrogen therapy has been demonstrated in the United States [42,98] and elsewhere [99]. Furthermore, according to the Massachusetts Women’s Health survey of women receiving estrogen therapy for the first time, 20% stopped taking it within 9 months of initiation, 10% used it intermittently, and 20 – 30% never filled their prescriptions [98]. In a follow-up study of 1689 women in the Netherlands, 66% of women were no longer taking estrogen 1 year and 9 months after the initial survey [100]. In the Study of Osteoporotic Fractures cohort, a substantial proportion of women reported stopping estrogen because they felt they did not need it simply because their symptoms abated [101]. Bone loss occurs asymptomatically until a fracture occurs. Hence, women need to be educated about the long-term benefits and risks of estrogen. This education program may result in improved compliance. Finally, 1-year compliance rates were greater among women prescribed transdermal estrogens than among women taking oral estrogens [102]. Future research needs to identify the best hormonal regimen that will achieve maximum benefit, minimum risks, and maximum compliance.
IV. ENDOGENOUS ESTROGEN, BONE MASS, AND FRACTURES Almost any study of endogenous estrogen, bone density, or fracture depends on the measurement of hormonal milieu at a single point in time, revealing little about the exposure over the course of a lifetime, due to both the daily variability of such measurements during the menstrual cycle and
CAULEY AND SALAMONE
the lack of correlation of postmenopausal measures with premenopausal levels. Also, there is disagreement as to whether estrogen-responsive tissues respond to absolute levels of estrogen or some other parameter such as unbound estrogen, which estrogens are important (e.g., estrone versus estradiol), and how hormone binding and metabolism influence the biological effects on bone. In contrast to the well-documented protective effect of exogenous estrogens on bone, bone loss, and fracture, less is known about the relationship of endogenous estrogens to bone. Several approaches have been taken to examine the association among endogenous estrogens, BMD, and fracture: cross-sectional studies have examined serum estrogens and bone density; longitudinally, serum estrogens have been compared among fast and slow bone losers or correlated with rates of bone loss; case-control studies have compared serum estrogens in fracture cases and controls.
A. Cross-Sectional Studies The majority of the cross-sectional studies reported positive associations between various measures of bone density and serum estrogens [103 – 111]. One study reported positive correlations between radial bone mineral content and estrogen excretion [112]. Two studies that failed to report an association between serum estrogens and bone mass had very small sample sizes and may have had little power to see an effect [113,114]. Johnston et al. [115] studied 40 postmenopausal women with type II diabetes and, after controlling for obesity, age, and diabetic control, neither estrone nor estradiol was a significant predictor of radial bone mass. In this study, failure to see an association may have reflected the small sample size. In addition, the distribution of the degree of obesity was narrow and could have influenced the homogeneity of estrogen concentrations. In a cross-sectional analysis of a random subset of 274 women enrolled in the Study of Osteoporotic Fractures, women who had estradiol levels 10 – 25 pg/ml had 4.9, 9.6, 7.3, and 6.8% greater BMD at the total hip, calcaneous, radius, and spine than those with levels below 5 pg/ml [116]. After multiple adjustments, BMD differences remained statistically significant and corresponded to about 0.4 standard deviation limits. Bioavailable estradiol concentrations are also important determinants of BMD in men and women, with correlation coefficients of 0.31 to 0.38 in men and 0.38 to 0.45 in women [117].
B. Longitudinal Studies For the most part, an association between serum estrogens and rates of bone loss was reported in both peri- or
755
CHAPTER 30 Postmenopausal Endogenous/Exogenous Hormones
early postmenopausal women and older women [118 – 128]. Several studies compared estrogen concentrations in fast and slow users [119 – 122]. Two of these studies focused on early postmenopausal women and found higher estrogens among the slow losers [121,122]. One of the best studies in this group was the study by Slemenda et al. [124] in which 84 peri- and postmenopausal women were followed over 3 years. Both estrone and estradiol concentrations were significantly correlated with change in radial BMD, with correlation coefficients ranging from 0.24 to 0.36. No association was found between estrone or estradiol and bone loss in a larger population of Dutch women [125]. These women were much older than the women in the Slemenda et al. [124] studies and it is possible that estrogen levels may have a lower impact on very old women. However, it has been shown that lower levels of endogenous estrogens and higher sex hormone binding globulin concentrations were associated with more rapid bone loss from the calcaneus and hip [128]. The women in this study were, on average, 72 years old. Women with serum estradiol values 10 pg/ml averaged only 0.1% [95% confidence interval ( 0.7%, 0.5%)] annual hip bone loss, whereas women with values below 5 pg/ml averaged 0.8% (0.3, 1.2) hip bone loss per year. These associations were independent of age and body weight.
C. Fracture Studies Evaluation of the relationship between endogenous estrogen and fracture have inherent shortcomings. Fractures are caused by both bone-related and fall-related factors. Fall-related factors may be influenced by estrogen. Second, vertebral fractures are difficult to define, do not always come to medical attention, and, consequently, are not fully represented in many studies. Third, there may be metabolic and binding factors affecting bone that are not captured by a simple assessment of estrogen concentration. Finally, in many cases, measurement, of estrogen concentrations were taken after the occurrence of the fracture and may reflect postfracture changes in body weight, physical activity, and other factors. Comparison of estrogen concentrations in osteoporotic fracture cases and controls without fracture yields inconsistent results, with some studies reporting higher levels of estrogens among controls [129 – 134] and others reporting no difference [121,122,125,135 – 138]. Older women with estradiol concentrations 5 pg/ml were 60% (20% to 80%) less likely to have a prevalent vertebral deformity than women with lower levels [116]. In the Rancho Benardo cohort, low values estradiol were associated with prevalent vertebral fractures in older men, but not women [139]. Also, while a few of the studies reviewed here assess
estrogen binding and metabolism and, by implication, its availability to target tissues, there is not enough information from which to draw conclusions regarding these issues. While Bartizal et al. [136] found similar concentrations of estrogens in vertebral fracture cases and controls, estradiol binding was significantly higher in controls, consistent with the hypothesis that the apparent decrease in estrogen binding activity may reflect decreased tissue sensitivity to estrogen rather than a lower serum hormone level. More data are needed to support this hypothesis. Another important consideration when evaluating inconsistencies is uncontrolled confounding. Few of these studies addressed the issue of confounding by body weight, probably the most important confounder of the relationship between estrogen levels and fracture. For instance, Aloia and associates [134] acknowledged case-control differences in weight and other potential confounders, but the analysis does not control for them. Another difficulty in interpretation of these studies is the clinical nature of most of the samples. Many of the earlier studies (published before 1990) are difficult to interpret because of alterations in hormones after the fracture, nonrepresentative cases and controls, and use of insensitive assays. A much stronger design is a prospective study design, where hormones are measured prior to the incident of fracture. In a nested case-cohort study within SOF, women with undetectable serum estradiol (5 pg/ml) had a relative risk of 2.5 for subsequent hip fractures (95% CI, 1.4 – 4.6) and subsequent vertebral fracture (95% CI, 1.4 – 4.2) as compared to women with detectable serum estradiol concentrations [140].
D. Summary and Future Research Estrogen concentrations are related to BMD and to longitudinal changes in BMD. Postmenopausal women who have low levels of estradiol have lower BMD and experience faster rates of bone loss. These observations have been reported in both peri- and newly postmenopausal women as well as much older women. Women with low levels of estradiol are also 2.5 times more likely to experience a hip or vertebral fracture. These findings are remarkable. The associations between endogenous estradiol and fractures are much stronger than we observed between serum cholesterol and death from cardiovascular disease [141]. These findings raise the possibility that a single measurement of estradiol could be used to estimate a woman’s risk of suffering accelerated bone loss and fractures. Women with very low levels of estradiol could consider estrogen replacement therapy. The dose of estrogen replacement therapy could be titrated to the endogenous concentration of estrogen. A lower dose of estrogen could be considered [142]. It is also
756 possible that nonpharmacologic interventions that either reduce or raise serum estrogen concentrations could thereby influence the risk of disease.
V. OBESITY The epidemic of overweight in the United States is welldocumented. The third National Health and Nutrition Examination Survey [NHANES III (phase 1)] of 1988 – 1991 showed the age-specific prevalence of overweight, as defined by a body mass index (BMI) of 27.3 kg/m2, in women ages 50 – 59 years and 60 – 74 years of 52.0 and 41.3%, respectively [143]. This value represents approximately 120% of the desirable weight for women defined by the midpoint of the range of weights for a medium frame from the 1983 Metropolitan Height and Weight tables [144,145].
A. Degree of Obesity and Bone Mineral Density Evidence from both cross-sectional and longitudinal studies supports the generally held beliefs that obesity confers higher BMD and protects against postmenopausal bone loss. 1. CROSS-SECTIONAL STUDIES Numerous cross-sectional studies have demonstrated that a variety of measures of body size are positively correlated with BMD in postmenopausal women [146 –161]. In one of the earlier studies, Dalen et al. [160] found that obese women had an 11% greater metacarpal cortical area than the nonobese. Correlations with body size and either BMD or BMC measures range from 0.00 to 0.42 at the lumbar spine [152–156,158], 0.195 to 0.27 at the femoral neck [154–156,158], 0.00 to 0.44 at the radius [152–154,156,158], and 0.21 to 0.54 for the whole body [150,161]. Although no ideal weight-to-height ratio has been set for reducing osteoporosis and fracture risk, a higher BMI appears to confer protection; accordingly, a BMI 26 – 28 kg/m2 offers protection, whereas a slender figure of 22 – 24 kg/m2 increases risk [162]. A BMI of about 30 kg/m2 is associated with a 4 – 8% greater spine BMD, 8 – 9% greater hip BMD, and 25% greater radial BMD compared to a BMI of 20 kg/m2; furthermore, a BMI of 30 kg/m2 is associated with one-half the loss in spine BMD in early postmenopausal years than a BMI of 20 kg/m2 [148, 157, 163, 164]. By stratifying postmenopausal women by those 115% or above their ideal body weight (IBW) and those below 115%, Dawson-Hughes et al. [156] demonstrated that heavier women had significantly greater BMD at the spine, hip, and radius than normal weight women. Among the obese women, BMD declined only at the radius with increasing years since menopause, whereas declines in BMC were
CAULEY AND SALAMONE
shown at both the radius and the hip. The observation that hip BMC declined while hip BMD was maintained suggests that obese women may experience an adaptive bone remodeling response after menopause, translating into improved bone density. In another study conducted in obese peri- and postmenopausal women and age-matched normals, obesity was defined as an excess of body weight of more than 10% from normal weight, as calculated from Lorentz’s equation [157]. Obese postmenopausal women had significantly higher vertebral BMD than nonobese women (1.056 – 0.12 and 0.974 – 0.14 g/cm2, respectively, P 0.0001), whereas no significant differences in BMD were observed among perimenopausal women. This implies that obesity may be more instrumental in attenuating postmenopausal bone loss rather than in achieving higher peak BMD. From prospective cohort studies, such as the Framingham cohort’s biennial examination (1988 – 1989) of 693 postmenopausal women, both body weight and BMI explained a substantial proportion of the variance in BMD for both weight (lumbar spine, femoral neck) and nonweightbearing sites (radius), accounting for 8.9 to 19.8% of the total variance (P 0.01 for all) [165]. In the Rancho Bernardo population, which examined the association between BMD of the spine, hip, and radius and measures of body size, both body weight and BMI consistently contributed to BMD at all sites [166]. It is important to note that measures of body size explained a much larger percentage of the variance in BMD in the axial weight-bearing spine and hip (12.4 and 16.5%, respectively) than in the nonweight-bearing midshaft and ultradistal radial sites (8.1 and 3.9%, respectively). Similarly, in a Japanese-American population using multiple regression analysis, body weight was related more strongly to BMC at the os calcis than at nonweight-bearing radial sites [167]. It is well established that obesity is associated with higher BMD, at least from cross-sectional evidence. What remains unclear is whether lifetime body weight or weight change in adulthood is a significant determinant of BMD. It is logical that both the degree and the duration of obesity may be important to consider when evaluating the association between body weight and BMD. From a cross-sectional evaluation of the Rancho Bernardo cohort, current BMI explained more than 29% of the variance in BMD at the radial shaft and total hip, 11% at the total spine, and 17% at the ultradistal radius [147]. When weight change was considered from self-reported weight at age 18 and present weight, greater weight gain was associated with higher BMD, explaining approximately 26% of the variance at the total hip. Therefore, it is possible that weight change or weight fluctuation in adulthood may be a most important determinant of BMD. Longitudinal evaluations of obese and normal weight women are necessary to confirm these cross-sectional observations.
CHAPTER 30 Postmenopausal Endogenous/Exogenous Hormones
2. LONGITUDINAL STUDIES To date, few longitudinal studies have addressed the influence of body weight or degree of obesity on postmenopausal bone loss. The longitudinal rates of bone loss over 31 months in a population of overweight postmenopausal women (BMI 25.0, n 40) were compared to those of women of normal weight women (BMI 25.0, n 115) [148]. The annual rate of vertebral bone loss was reduced significantly among the overweight compared with normal weight women (0.54 – 1.1 and 1.46 – 1.6%, respectively, P 0.05). These results are in agreement with those of Harris et al. [149], who demonstrated that increased body weight up to 106% of IBW (corresponding to a BMI of 23.6 kg/m2) was protective against postmenopausal vertebral bone loss, but not against radial or femoral bone loss. Additional weight beyond 106% did not appear to provide any further protection from vertebral bone loss. From the Framingham cohort, the change in body weight from the biennial examination (1948 – 1951 to 1988 – 1989) in 693 women was the strongest explanatory factor for BMD among women at the spine, the femoral neck, and radius (10.7, 8.2, and 6.5%, respectively, of the total variance, P 0.01 for all) [165]. Of interest, the relationship between weight change and BMD was strongest in women not using estrogen.
B. Weight Loss and Bone Mineral Density The consequences of weight reduction on BMD in obese subjects remains, at present, unknown. Cross-sectional studies performed on obese subjects undergoing surgical procedures for weight reduction, including jejuno-ileal bypass, gastroplasty, and biliopancreatic bypass, are associated with some reduction in BMD, albeit not consistently [168 – 172]. Early longitudinal evidence was unable to demonstrate any significant declines in BMD following weight loss [171,172]. For example, in a study by Rickers et al., [171], there was no evidence of radial BMD loss in obese subjects at 12 months (n 11; 114% overweight) or 8 years after intestinal bypass (n 12; 38% overweight). These discordant results may be ascribed to varying amounts of dietary calcium and vitamin D provided, methodological problems in assessing bone in obese subjects, or perhaps inherent differences in cross-sectional and longitudinal study designs. More recently, changes in total and regional BMD, as measured by dual-energy X-ray absorptiometry (DXA), were assessed in obese patients undergoing rapid weight loss on a low-calorie diet for 2 weeks [mean 1.9 MJ (452 kcal) for women and 2.4 MJ (571 kcal) for men] and thereafter up to 4.2 MJ (1000 kcal) for women and 4.7 MJ (1119 kcal) for men [174]. After 15 weeks, whole body BMC was reduced significantly ( 5.9%), with the largest declines in
757 the trunk ( 6.9%) and the smallest in the arms ( 4.0%), suggesting greater losses in weight-bearing than in nonweight-bearing sites. In addition, losses in whole body BMC were correlated positively and strongly with fat mass losses (r 0.86, P 0.01) after 15 weeks and at 9 months (r 0.94, P 0.01), but not with lean mass (r 0.20, P 0.15). Compston et al. [175] found inconsistent losses in BMD of approximately 1 to 2% at weight-bearing sites in obese postmenopausal women participating in a low-calorie dietary intervention, but not at the radius [175]. It is of interest that after 10 months of follow-up, subjects returned to their baseline weight as well as baseline BMD values. Given the fact that body weight and BMD changed in the same direction during weight loss and gain suggests that observed changes in BMD may indeed be real. In contrast, Avenell et al. [176] did not observe parallel gains in weight and BMD in the second 6 months of a 1-year dietary study. These studies raise important questions about whether weight regain is accompanied by bone regain and whether the bone that is regained is of similar quality. More recently, Salamone et al. [177] examined the effect of a lifestyle intervention aimed at lowering dietary fat intake and increasing physical activity to produce modest weight loss or prevent weight gain on BMD in premenopausal women participating in the Women’s Healthy Lifestyle Project (WHLP). Of the 236 women enrolled, the intervention group experienced a mean weight loss of 3.2 4.7 kg over the 18-month study period (n 115) compared to a weight gain of 0.42 3.6 kg in the controls (n 111). The annualized rate of hip BMD loss was twofold higher in the intervention group (0.81 1.3%/year) than in the control group (0.42 1.1%/year); a similar although nonsignificant pattern was observed at the spine (0.70 1.4%/year vs 0.37 1.5%/year). Evidence from this study suggests that the overall risks and benefits of weight reduction among premenopausal women need to include effects on BMD and risks of osteoporosis. One would hypothesize that postmenopausal women would be at an even greater risk of osteoporosis given the additive effect of weight loss-induced bone loss and normal postmenopausal bone loss. Current studies are underway to determine the longer term repercussions of weight loss-induced bone loss at the completion of the 5-year intervention study. There is limited evidence to examine whether similar changes in BMD occur when exercise is included as a component of the intervention. In the WHLP study population, large increases in physical activity (4180 kJ/week) during the study attenuated spinal bone loss but had no impact on hip bone loss [177]. However, Svendsen et al. [178] compared an energy-restrictive dietary intervention with and without aerobic exercise and weight training and found no additional benefit of exercise on weight loss-induced changes in BMD. Further studies considering the type,
758
CAULEY AND SALAMONE
duration, and intensity of physical activity that confers protection against BMD during weight loss are needed. A few mechanisms have been postulated to explain the association between changes in weight and changes in BMD. First, it is possible that during weight loss and subsequent loss of fat tissue, there is a marked reduction in endogenous estrogen production. Such declines in fat mass with weight loss result in lower concentrations of androgen precursors for conversion to estrogens in peripheral tissues. Second, during weight loss, the rate of bone turnover may be enhanced such that changes in total body BMD are masked by larger ones in trabecular bone such as in the spine or hip [161], indicating that more region-specific measures are necessary to assess the clinical significance of weight loss on BMD. Third, changes in BMD may be artifactual. Because of the discordant results across studies on whether bone loss accompanies weight loss in obese subjects, it is possible that these differences reflect methodological issues in bone measures in obese subjects [173,179]. Earlier techniques to measure BMD, such as photon absorptiometry, might have been compromised by the influence of fat tissue, whereas the more recent DXA measurements, which are less dependent on adjacent fat tissue, may be more effective in monitoring changes in BMD. Fourth, with weight loss there is a decline in the mechanical stress on the skeleton that may influence bone remodeling. Next, nonsignificant reductions in markers of bone formation [178] and elevations in markers of bone resorption [180] have been reported among women consuming an energy-restrictive diet compared to controls. Similar results were shown by Salamone et al. [177] in which premenopausal women who lost the most weight ( 8%) experienced the largest increase in the percentage change in N-telopeptide, a marker of bone resorption. These parallel changes in bone resorption and BMD loss support the hypothesis that weight loss-induced bone loss may be mediated, at least in part, by alterations in bone remodeling. Finally, weight control programs may compromise the adequacy of nutrient intake, especially that of calcium and vitamin D – both of which are integral to bone health.
C. Methodological Issues Measurements of body composition and BMD by absorptiometric techniques such as dual photon absorptiometry (DPA) or DXA are generally limited by the basic assumption of a three-compartment model (bone, bone-free lean mass, and fat), which does not include water as a separate component [181]. Such compartmental measures assume uniform hydration, particularly of bone-free lean mass at 0.73 ml/g [182, 183], when indeed this value can vary by as much as 67 – 85% [183]. Because lean and fat mass are summed to give total weight, errors in lean mass
are propagated to the fat compartment, perhaps resulting in distorted values [184]. The extent to which absorptiometric measurements of soft tissue body composition are sensitive to variations in hydration status, particularly in the obese, is presently unknown. Many studies have examined both the accuracy and the precision of absorptiometric measures in populations of normal weight, although the validity of such measures is limited in the obese. In vitro studies have tried to simulate “fatness” by measuring aluminum phantoms at varying thicknesses. There was some suggestion of machine drift over time in the thickest phantom [185]. Similarly, evidence from in vitro phantoms scanned by DXA in water and simulated tissue (ground beef and fat) supports a significant effect of “thickness” on total body BMC, measured by DXA and a similar, although less conclusive, effect on BMD [186]. Others have shown no significant effect of increasing tissue equivalent thickness on DXA phantom measurements, whereas with DPA, significant reductions in phantom BMD with increasing thickness are observed [187]. Such differences in phantom measurements between DPA and DXA may explain, in part, differences observed in the strength of the relationship between body weight and BMD across studies. DXA measurements appear to be sensitive to the anteroposterior thickness of the body, as supported largely from in vitro studies. The influence of thickness on BMD measurements by DPA and DXA is reported as a consistent under- or overestimation [188 – 191]. Anterior – posterior bone measurements by DXA do not account for bone thickness, which is related to overall body size, leading to exaggerated results. This modest thickness effect may have the most significant implications for subjects having extreme truncal thicknesses or those who experience changes in thickness or body weight [184]. In theory, measurement errors by absorptiometry between thin and obese subjects may result from either the scanner’s misinterpretation of soft tissue pixels as bone, leading to reduced measures of BMD [186], or a reduction in the proportion of X-rays surviving attenuation, resulting in errors in bone edge delineation and subsequently, enlarged bone with reduced density. Svendsen et al. [192], as did Van Loan and Mayclin [193], suggested that large changes in body mass and composition may influence the ability of DXA to detect bone edges, thereby changing the estimation of bone area. Because of these limitations, DXA measures are truncated in women weighing up to 275 lbs. It is likely that systematic errors in BMD measurements by absorptiometry exist between thin and obese subjects, although the degree to which this occurs is not well established. The overall lack of instrument sensitivity in measuring changes in BMC and BMD has also been suggested. After a 15-week weight loss program, nonsignificant changes in BMC and bone area measurements occurred
CHAPTER 30 Postmenopausal Endogenous/Exogenous Hormones
in opposite directions, resulting in a significant decline in BMD without a loss in BMC [194]. These observations suggest that changes in BMD observed in weight loss studies may result from the lack of instrument sensitivity when body weight and body composition change and may indeed be an artifact and not a real physiologic change. Ritchie et al. [195] also showed that changes in BMD after pregnancy were the result of changes in the measurement of bone area, which in turn impacts the calculation of BMD. Clearly, there are significant methodological limitations in measuring bone density in the obese. Interpretation of such studies is further complicated by the fact that the definition of obesity is imprecise, often defined by different weight to height ratios.
D. Body Composition There is little disagreement that body weight plays a significant role in determining BMD, particularly at the axial skeleton. However, what remains less clear is what the relative contributions are of the fat and lean mass compartments of that body weight to BMD. While there is strong evidence to support an association between body weight and BMD, there is considerably less evidence relating the components of body weight, lean and fat mass, to bone mass. Among postmenopausal women, some studies attribute most or all of the weight effect at the weight-bearing skeleton to the fat component [188,196,197], whereas others assign a greater role to lean mass [153,198]. The percentage of variability of bone mass explained by fat-free mass ranged from 38 to 55% in premenopausal women and from 28 to 52% in postmenopausal women, whereas the percentage of variability explained by fat mass was only 1 to 9% in premenopausal women and 12 to 21% in postmenopausal women [198]. In postmenopausal women, Reid and colleagues [196] demonstrated that total fat mass, as measured by DXA, is a consistent predictor of spinal and femoral neck BMD. Similar correlations were reported between total fat mass and femoral neck BMD (r 0.38) as with body weight and BMD (r 0.39, P 0.05 for both) [196]. Analogous results were presented by Compston et al. [197], who showed that fat mass, after adjustment for lean mass and age, was significantly correlated with bone mineral measurements at the spine (r 0.30), femoral neck (r 0.31), and whole body BMD (r 0.48, P 0.01 for all). Low fat mass is considered to be a risk factor for osteoporotic fractures, as 11 – 15% of fracture patients had fat mass values, as measured by DPA, below the lowest value in the controls, whereas lean mass was essentially equal in both groups [188]. The fact that fast bone losers have a lower anthropometrically determined fat mass than slow
759 bone losers in early postmenopause further adds to the importance of fat mass in determining BMD in postmenopause [122]. Whether lean mass is a significant determinant of BMD in the postmenopausal years is, at present, unresolved. The importance of lean mass or muscularity to BMD is supported by both in vitro and in vivo studies. From human autopsy specimens, after controlling for age, weight, and height, there was a significant correlation between the ash weight of the third lumbar vertebrae and the weight of the left psoas muscle [199]. Sinaki et al. [200] reported a significant correlation between the isometric strength of the back extensor muscles and BMD of the lumbar vertebrae, accounting for 9% of the variance in spine density. In a study by Bevier et al. [153] conducted in women greater than 60 years of age, fat mass was calculated from skinfold thicknesses and bioelectrical impedance analysis (BIA), and lean mass was calculated by subtracting fat mass from body weight. Spinal density correlated with both fat and lean mass, but stepwise multiple regression analyses found that only lean mass contributed significantly to the prediction of spinal BMD. This finding was not confirmed by Aloia et al. [201], who found no significant correlations between total body potassium and trochanteric BMD or Ward’s triangle BMD. Indeed, this may be attributed to regional differences in the quantities of fat or lean mass, similar to regional differences in bone remodeling and architecture within the femur. More recent studies have shown that once bone thickness and body weight are taken into account, body composition appears to have little independent effect on BMD at spine, femoral neck, or total body. This observation is supported by Carter et al. [202], who introduced the term bone mineral apparent density by dividing BMC by an estimate of bone volume, and by Reid et al. [203], who divided BMD by height to adjust for differences in bone size. In a study by Harris and Dawson-Hughes [204], the association between nonfat soft tissue (NFST) and BMD was examined in 261 postmenopausal women. A height-dependent variable was calculated to correct for differences in bone thickness by dividing BMD by height [204]. Body weight was correlated positively with height-adjusted BMD at the spine, hip, and total body (r 0.22 – 0.26). However, the percentage of NFST was not associated with height-adjusted BMD at the spine or femoral neck and only weakly at total body (r 0.12). This finding supports the hypothesis that the protective effect of body weight is predominantly through its mechanical force on the skeleton; thus, previously reported associations between bone mineral and fat-free mass in postmenopausal women may be an artifact attributable to incomplete control for bone and body size and the inclusion of bone mineral in fat-free mass [205]. Mechanisms other than just the simple mechanical loading on the skeleton, however, have been postulated to
760 explain the association between body composition or the relative proportions of body fat or lean mass on BMD. Different mechanisms predominate at different stages of life. In premenopausal years, lean mass, a major component of weight, becomes more important in the maintenance of BMD. With increasing age and especially after menopause, fat mass and weight increase, while lean mass declines, changing the relative proportion of lean mass and fat mass to weight. Thus, the association between fat mass and bone mass increases with age as more of the weight loading on bones is attributable to fat mass [205]. Such increases in fat mass may promote greater estrogen production. Because the conversions of androgen to estrogen become a major source of estrogen after menopause, the endocrine effect of fat mass may be minimal during bone accretion. It is reasonable to conclude that lean mass may also be an important determinant of bone mass [206]. The positive association between lean mass and bone mass may indicate a potential genetic association between higher peak bone mass and higher lean tissue, a greater mechanical loading on the skeleton, or perhaps indicate higher levels of physical activity. Evidence from the Sydney Twin Study of Osteoporosis demonstrated that lean mass was a significant determinant of areal BMD, whereas fat mass was a significant determinant of volumetric BMD. From this study, it was concluded that lean mass, fat mass, and bone mass are under strong genetic control; although it is plausible that the same genes modulate bone mass, the association between bone mass and body composition appeared to be mediated primarily through common environmental factors [208]. Both fat and lean mass have independent influences on bone mass, but their relative contributions appear to vary by bone site based on trabecular content, physical mobility, and muscularity of the site [209]. It is possible that more trabecular bone sites are influenced more strongly by fat mass because they may respond more to the hormonal effects of fat mass and it is metabolically more active bone. In contrast, lean mass may act via mechanical stress of muscular contractions that result from physical activity. Lean mass may play a larger role at the femur, for example, because of the muscular activity in this portion of the body. Based on the available evidence mostly from cross-sectional studies, there is some consensus on the relative contributions of lean and fat mass to BMD in postmenopausal women. From cross-sectional data in postmenopausal women, lean mass explained a greater proportion of the variance than fat mass; longitudinally, however, annual changes in fat mass were associated with regional changes in BMD [206]. Whether loss of lean mass with aging contributes to loss of skeletal mass simply by reducing mechanical load on the skeleton is not established [210].
CAULEY AND SALAMONE
Further longitudinal studies on changes in BMD in the postmenopausal years are important in order to determine whether changes in BMD are sensitive to changes in body composition and, furthermore, to delineate the mechanism of such an association.
E. Degree of Obesity and Fractures The inverse association between body weight and risk of fractures has been documented extensively by cross-sectional [20,75,78,82,89] and case-control studies [211 – 219], although only three prospective studies have examined this association [213,215,216]. As discussed previously, obese women tend to have higher BMD than nonobese as well as lower rates of postmenopausal bone loss as compared to slender women [149]. Studies of hip fracture [75] and forearm fracture [75,213,217,218] find the thinnest women at the higher risk of fracture. From the Nurse’s Health Study, there was a 20% reduction in fracture incidence when women with a BMI 21 kg/m2 were compared to those 21 kg/m2 [217]. Similarly, in a study of hip fracture cases and trauma controls. From the Study of Osteoporotic Fractures, higher body weight was associated with a significant reduction in fractures of the distal forearm (RR 0.89; 95% CI, 0.79 to 1.01), but no real differences in fractures of the proximal humerus (RR 0.97; 95% CI, 0.82 to 1.15) [213]. Accordingly, heavy (as compared to thin) women appear less likely to develop osteoporosis and osteoporosis-related fractures. Kiel and colleagues [82] confirmed the strong inverse association between body weight and hip fracture using weight categories of Metropolitan weight tables (1959). With self-reported weight, Alderman et al. [218] demonstrated that the risk of hip or forearm fracture declined with increasing increments of weight. Another case-control study showed that the effect of obesity on fractures appears to be limited to women not taking replacement estrogens, with a reduced risk of hip fracture in the moderately obese groups, yet no protective effect in the most obese [20]. However,among obese women, as defined by a ponderal index between 9.6 and 12.5, with little or no use of estrogen replacement therapy (less than 1 year), the risk of hip fracture was greater in thin women than in obese women. Among the obese women, neither the risk of hip fracture nor that of forearm fracture was affected by the use of estrogens, with relative risks of 1.9; 0.7 to 5.4 and 0.8; 0.4 to 1.6, respectively [89]. From these results, it appears that the use of replacement estrogens changes fracture risk little, if any, in obese postmenopausal women, although as adiposity declines, the beneficial effect of estrogen therapy becomes more pronounced. More recently, Cumming and Klineberg [211]
CHAPTER 30 Postmenopausal Endogenous/Exogenous Hormones
considered the influence of both past weight (age 20 years) and recent weight (age 50 years) on the risk of hip fracture. Current high BMI was associated with a reduced risk of hip fracture, whereas high BMI at age 20 was associated with an increased risk of hip fracture. Comparing individuals who maintained or gained weight between age 50 and time of the fracture, the odds ratio for risk of fracture for those who lost weight was 1.9; 1.1 to 3.3, whereas for those who lost weight after age 20 was 3.4; 1.8 to 6.4, suggesting that weight loss may be associated with an increased risk of fracture. These results are consistent with those reported in the Study of Osteoporotic Fractures, indicating that weight loss after age 50 years is associated with reductions in BMD [220] and increased risk of fracture [221]. Cases of vertebral fractures between otherwise matched subjects were more likely seen in subjects with a BMI 24 kg/m2 than in subjects 26 kg/m2 [222,223]. Another study showed prevalences of vertebral fractures in women 60 – 80 years of 79, 48, and 27% for BMIs of 19, 22, and 28 kg/m2 [224]. Clearly, low body weight increases osteoporosis risk and a more generous body weight confers some protection against osteoporosis and related fracture risk to a point. Few studies, only one of which is prospectively designed [214], have examined the association between body composition and fracture risk [199,214,219]. Because body weight is composed of muscle and adipose tissue, it is reasonable that the components of this weight may offer some explanation of the association between body weight and fractures. In a Swedish study comparing urban and rural populations, lower muscle mass, as measured by BIA, was found in women experiencing fragility fractures after age 70. Muscle mass was correlated more strongly with forearm BMC than body weight, suggesting the importance of muscle mass as a primary determinant of bone mass and fracture risk [225]. There is additional evidence from the NHANES I cohort of 3595 women, ages 40 – 77 years, who were followed prospectively for 10 years, that muscle and adipose tissue contribute independently to the risk of hip fractures [214]. Of interest, women with hip fractures tended to weigh less and have lower triceps skinfold thickness and muscle arm area. These differences translated into a twofold reduction in the relative risk of hip fracture for BMI, tricep skinfolds, and muscle arm area (2.6; 1.8 to 3.9, 2.0; 1.4 to 2.9, and 1.7; 1.2 to 2.4), respectively, even after adjusting for potential confounders. Comparisons of subscapular measures and triceps skinfold thicknesses indicated that the risk of hip fractures was not associated with regional differences in adiposity. Whether body weight alone or the individual components of that weight are more significant in explaining the association between obesity and fracture risk has not been fully elucidated.
761
F. Mechanisms of the Association between Obesity and Bone Mineral Density Obesity, even if moderate, appears to protect against osteoporosis. The relationship between body size and BMD has not been fully explained, although several potential mechanisms have been presented. As shown, obese women tend to have higher BMD, particularly in the axial (weight-bearing) skeleton, than nonobese women [151 – 158]. Such differences in skeletal mass are largely ascribed to the increased mechanical load on the skeleton [210]. Indeed, if obesity persists from early adulthood, it is likely that excess body weight is a primary factor in both maximizing peak bone mass and maintaining BMD as an older adult. This mechanical hypothesis, however, does not entirely explain the association between obesity and BMD. In a study by Ribot et al. [157], vertebral BMD of slightly obese postmenopausal women was higher than that of normal weight women, although a similar trend was not demonstrated among perimenopausal women, suggesting the importance of hormonal factors in the relationship between obesity and BMD. In postmenopausal women, most circulating estrogens are derived from the peripheral conversion of androstenedione to estrone in adipose tissue (10 – 15%) [226 – 229] and muscle (25 – 30%) [230]. The adrenal production of androgens tends to be higher in obese than nonobese women, resulting in more available androgen precursors for conversion by the aromatase enzyme to estrogens in peripheral tissues. This peripheral aromatization seems to be accelerated in obese women [229], as shown by the increased percentage of androstenedione converted to estrone, ranging from 1 – 2% in normal weight women to 12 – 15% in women weighing between 300 and 400 lbs. [227]. Moreover, the conversion of androgens to estrogens is increased with age [230] and may be mediated by the type of obesity present (upper vs lower) [151,231,232]. Moreover, obese postmenopausal women may have lower concentrations of sex hormone-binding globulin (SHBG), which may result in a greater proportion of free, metabolically active estrogens as well as androgens [233]. Estrogen deficiency is integral in the development of postmenopausal osteoporosis and fractures. Because obese women have greater amounts of both adipose and muscle mass than the nonobese, it is possible that adiposity protects against osteoporosis by providing higher amounts of circulating estrogens, which may impact BMD through osteoclastic and osteoblastic activity, as well as in alterations in bone remodeling. This is supported by most [235], but not all [196], studies. Reid and colleagues [196] maintained that the conversion of adrenal androgens to estrogen in adipose tissue does not explain the relationship between adiposity and BMD, as this association was found to be independent of serum estrone levels. From Framingham data,
762 adiposity was correlated positively with estrogen levels in postmenopausal women, but related inversely to the rate of postmenopausal bone loss [165]. Moreover, women who never used replacement estrogens had a stronger association with BMD and body weight than women who used estrogen replacement therapy. Because estrogen users are generally less obese than nonusers [2], this lower association may rather reflect the narrow distribution of body weight among estrogen users. The observation that obese women experience attenuated rates of bone loss as compared to their normal weight counterparts might be explained by their enhanced bone formation, reduced bone resorption, or a combination of the two. Frumar et al. [236] demonstrated a negative correlation between fasting urinary calcium/creatinine ratio (Ca/Cr) and percentage ideal body weight (r 0.43, P 0.01) and a reduced Ca/Cr in obese than in nonobese postmenopausal women. As fat mass increases, bone resorption tends to decline, as measured by Ca/Cr, without a simultaneous decline in bone formation, measured by bone gla protein [237]. This adaptive mechanism in bone remodeling, which supports greater bone formation than bone resorption, might inhibit net bone loss. It is possible that calcium and vitamin D homeostasis is altered in obese women, resulting in reduced ionized calcium activity, increased PTH concentration, reduced 25-hydroxyvitamin D, and increased 1,25(-OH)2D [238 – 240]. Bell and colleagues [238] support alterations in the vitamin D endocrine system resulting from the greater mechanical strain on the axial skeleton. In obese subjects, greater than 30% or more above ideal body weight, 25-hydroxyvitamin D was significantly lower in the obese [238 – 241], whereas 1,25-dihydroxyvitamin D and serum PTH were significantly higher in obese than in age-matched, nonobese women [242]. Of interest, PTH and IBW were positively correlated (r 0.72, P 0.01). Given this, Bell et al. [238] asserted that the vitamin D endocrine system is altered in obese individuals; this alteration is characterized by secondary hyperparathyroidism, enhanced renal tubular reabsorption of calcium, and increased circulating 1,25(OH)2D. The consequent reduction in serum 25-hydroxyvitamin D is ascribed to a feedback inhibition on its hepatic synthesis, as determined by the increased concentration of 1,25-dihydroxyvitamin D [238]. Such alterations in the vitamin D endocrine system were confirmed by histomorphometric analyses of iliac crest biopsies in morbidly obese subjects [243].
G. Mechanisms for the Association between the Degree of Obesity and Fracture Risk Potential mechanisms for the protective effect of body weight on fracture risk have been postulated. First, as expected with higher body weight, there are greater amounts
CAULEY AND SALAMONE
of both muscle and adipose tissue, thereby increasing the mechanical stress on the weight-bearing skeleton. Greater weight-bearing stress present during early adulthood can maximize peak skeletal density and, in later life, may attenuate postmenopausal bone loss. Indeed, larger-boned women have more bone mass to lose than small-boned women before reaching critical fracture threshold levels. An alternative explanation may be that the obese have improved bone strength or bone quality, although to date, no studies have examined the role of obesity on bone quality. Second, heavy postmenopausal women have higher levels of endogenous estrogen because of aromatization of androstenedione to estrone in muscle and adipose tissue. In absence of ovarian function, the conversion of estrone to estradiol represents the primary source of circulating estradiol. In addition, because SHBG is higher in hip fracture patients than in age-matched controls [132], a difference largely ascribed to differences in body size, there is a lower concentration of biologically available estradiol and testosterone in hip fracture patients. Taken together, obese postmenopausal women have a significantly higher production of estradiol than their thinner counterparts, the availability of which is increased by lower concentrations of SHBG [244,245]. Accordingly, because estrogen therapy is known to prevent bone loss in postmenopausal women [246], an important mechanism by which obesity is protective against fracture risk is through both increased estrogen production and availability after menopause. Third, because obese women tend to be less active than thin women, they may be less likely to fall and sustain a fracture. However, in the event of a traumatic fall, higher amounts of muscle and adipose tissue in the obese may offer a cushioning protection for the skeleton [234]. In the event of bone trauma, some researchers have even advocated that slim elderly wear protective padding in the hip region to compensate for the loss of fat protection [247]. Finally, it is reasonable that obesity may protect against osteoporotic fracture through alterations in the vitamin D endocrine system, as suggested previously by Bell and colleagues [238].
H. Future Directions Although much progress has been made in elucidating the mechanisms associated with obesity, BMD, and fractures, much remains to be learned, particularly in the area of body composition and BMD. There is a major need for additional prospective cohort studies to examine changes in body weight and weight cycling in relation to BMD and fracture risk. Because the contribution of lifetime weight and weight change to BMD and fracture risk is unclear, further examination of both the duration and the degree of obesity on such
763
CHAPTER 30 Postmenopausal Endogenous/Exogenous Hormones
end points is important. Of particular interest will be determining the relative importance of weight gain and weight loss on fracture risk among the elderly and whether weight change or fluctuation in early adulthood is an important determinant of BMD and fracture risk in later years. Additional studies are necessary to examine whether individual compartments of body weight, specifically lean and fat mass, are more important determinants of BMD and fracture risk than body weight alone. Finally, because of the substantially higher prevalence of overweight in African-American and Hispanic women as compared to Caucasian women [248], it is important to consider the potential effect of obesity on racial differences in BMD and fracture risk. Further explorations into genetic/environmental and their interaction contributions to BMD are integral to gaining a better understanding of the associations between obesity and both BMD and fractures.
3.
4.
5.
6.
7.
8.
9.
VI. OVERALL SUMMARY This chapter reviewed the relationship among three important determinants of osteoporotic risk. Estrogen replacement therapy has been consistently shown to preserve bone mass and prevent fractures. The biggest question surrounding estrogens relates to the risks and benefits of long-term use, as data suggest that women need to continue estrogens indefinitely. In addition, compliance to estrogen therapy is low; future studies are needed to identify ways to maximize compliance. Thiazide diuretics may prevent hip fractures, but the underlying mechanisms for this effect are not known. More data are needed in normotensives before thiazide diuretics can be advocated for the prevention of osteoporotic fractures. Data are needed not only on the efficacy of thiazide diuretics on fracture, but also careful documentation of the risks and side effects of this therapy. Obesity may be in the causal pathway among endogenous estrogens, bone mass, and fractures. In addition, there are other mechanisms for an effect of obesity on osteoporotic risk. Future research needs to identify this causal pathway to improve our understanding of the etiology of osteoporotic risk. In addition, prospective studies are needed to test the hypothesis that low serum estrogens predict individuals at risk for fracture, particularly hip fracture.
10.
11.
12. 13.
14.
15.
16.
17.
18.
19.
References 20. 1. J. A. Cauley, S. R. Cummings, D. G. Seeley, D. Black, W. Browner, L. H. Kuller, and M. C. Nevitt, Effects of thiazide diuretic therapy on bone mass, fractures, and falls. Ann. Intern. Med. 118, 666 – 673 (1993). 2. J. A. Cauley, S. R. Cummings, D. M. Black, S. R. Mascioli, and D. G. Seeley, Prevalence and determinants of estrogen replacement
21. 22.
therapy in elderly women. Am. J. Obstet. Gynecol. 163(5), 1438 – 1444 (1990). J. A. Cauley, J. P. Gutai, L. H. Kuller, and J. G. Powell, The relationship of endogenous sex steroid hormone concentrations to serum lipid and lipoprotein levels in postmenopausal women. Am. J. Epidemiol. 132, 884 – 894 (1990). R. D. Wasnich, R. J. Benfante, K. Yano, L. K. Heilburn, and J. M. Vogel, Thiazide effect on the mineral content of bone. N. Engl. J. Med. 309, 344 – 347 (1983). M. R. Sowers, R. B. Wallace, and J. H. Lemke, Correlates of mid-radius bone density among postmenopausal women: A community study. Am. J. Clin. Nutr. 41, 1045 – 1053 (1985). R. D. Wasnich, R. D. Philip, L. K. Heilbrun, J. M. Vogel, K. Yano, and R. J. Benefante, Differential effects of thiazide and estrogen upon bone mineral content and fracture prevalence. Obstet. Gynecol. 67, 457 – 462 (1986). P. Adland-Davenport, M. W. McKenzie, M. Notelovitz, L. C. McKenzie, and J. F. Pendergast, Thiazide diuretics and bone mineral content in postmenopausal women. Am. J. Obstet. Gynecol. 152, 630 – 634 (1985). R. Lindsay, B. Marshall, A. Haboubi, B. S. Herrington, and J. Tohme, Increased axial bone mass in women with hypertension: Role of thiazide therapy. J. Bone Miner. Res. 2, 223 (1987). N. W. Glynn, E. N. Meilahn, M. Charron, S. J. Anderson, L. H. Kuller, and J. A. Cauley, Determinants of bone mineral density in older men. J. Bone Miner. Res. 10, 1769 – 1777 (1995). D. J. Morton, E. L. Barrett-Connor, and S. L. Edelstein, Thiazides and bone mineral density in elderly men and women. Am. J. Epidemiol. 139, 1107 – 1115 (1994). R. Wasnich, J. Davis, P. Ross, and J. Vogel, Effects of thiazide on rates of bone mineral loss: A longitudinal study. Br. Med. J. 301, 1303 – 1305 (1990). B. Dawson-Hughes and S. Harris, Thiazides and seasonal bone change in healthy postmenopausal women. Bone Miner. 21, 41 – 51 (1993). M. R. Sowers, M. K. Clark, M. L. Jannausch, and R. B. Wallace, Body size, estrogen use and thiazide diuretic use affect 5-year radial bone loss in postmenopausal women. Osteopor. Int. 3, 314 – 321 (1993). J. S. Adams, C. F. Song, and V. Kantorovich, Rapid recovery of bone mass in hypercalciuric men treated with hydrochlorothiazide. Ann. Intern. Med. 130, 658 – 660 (1999). C. Christiansen, M. S. Christiansen, P. McNair, C. Hagen, K. E. Stocklund, and I. Transbol, Prevention of early postmenopausal bone loss: Controlled 2-year study in 315 normal females. Eur. J. Clin. Invest. 10, 273 – 279 (1980). I. Transbol, M. S. Christiansen, G. F. Jensen, C. Christiansen, and P. McNair, Thiazide for the postponement of postmenopausal bone loss. Metabolism 31, 383 – 386 (1982). T. D. Giles, G. E. Sander, L. E. Roffidal, A. C. Quiroz, and A. L. Mazzu, Comparative effects of nitrendipine and hydrochlorothiazide on calciotropic hormones and bone density in hypertensive patients. Am. J. Hypertens. 5, 875 – 879 (1992). R. D. Wasnich, J. W. Davis, Y. F. He, H. Petrovich, and P. D. Ross, Thiazides reduce bone loss: A randomized double-blind, placebo controlled trial. J. Bone Miner. Res. 8, 123 (1993). [Abstract] W. E. Hale, R. B. Steward, and R. G. Marks, Central nervous system symptoms of elderly subjects using antihypertensive drugs. J. Am. Geriatr. Soc. 32, 5 – 10 (1984). A. Paganini-Hill, R. K. Ross, V. R. Gerkins, B. E. Henderson, M. Arthur, and T. M. Mack, Menopausal estrogen therapy and hip fractures. Ann. Intern. Med. 95, 28 – 31 (1981). S. Rashiq and R. F. Logan, Role of drugs in fractures of the femoral neck. Br. Med. J. 292, 861 – 863 (1986). H. M. Taggart, Do drugs affect the risk of hip fracture in elderly women? J. Am. Geriatr. Soc. 36, 1006 – 1010 (1988).
764 23. A. Stevens and C. Mulrow, Drugs affecting postural stability and other risk factors in the hip fracture epidemic-case-control study. Community Med. 11, 27 – 34 (1989). 24. D. T. Felson, D. Sloutskis, J. J. Anderson, J. M. Anthony, and D. P. Kiel, Thiazide diuretics and the risk of hip fracture. JAMA 265, 370 – 373 (1991). 25. F. E. Heidrich, A. Stergachis, and K. M. Gross, Diuretic drug use and the risk for hip fracture. Ann. Intern. Med. 115, 1 – 6 (1991). 26. R. G. Cumming and R. J. Klineberg, Psychotropics, thiazide diuretics and hip fractures in the elderly. Med. J. Aust. 158, 414 – 417 (1993). 27. W. A. Ray, W. Downey, M. R. Griffin, and L. J. Melton, Long-term use of thiazide diuretics and risk of hip fracture. Lancet April, 687 – 690 (1989). 28. A. Z. LaCroix, J. Wienpahl, L. R. White, R. B. Wallace, P. A. Scherr, L. K. George, J. Cornoni-Huntley, and A. M. Ostfeld, Thiazide diuretic agents and the incidence of hip fracture. N. Engl. J. Med. 322, 286 – 290 (1990). 29. D. Feskanich, W. C. Willett, M. J. Stampfer, and G. A. Colditz, A prospective study of thiazide use and fractures in women. Osteopor. Int. 7, 79 – 84 (1997). 30. Z. Guo, P. Wills, M. Vitanen, J. Fastbom and B. Winblad, Cognitive impairment, drug use, and the risk of hip fracture in persons over 75 years old: A community-based prospective study. Am. J. Epidemiol. 148, 887 – 92 (1998). 31. G. Jones, T. Nguyen, P. N. Sambrook, and J. A. Eisman, Thiazide diuretics and fractures:Can meta-analysis help? J. Bone Miner. Res. 10, 106 – 111 (1995). 32. S. Middler, C. Y. Pak, F. Murad, and F. Bartter, Thiazide diuretics and calcium metabolism. Metabolism 22, 139 – 146 (1973). 33. E. R. Yendt and M. Cohanim, Prevention of calcium stones with thiazides. Kidney Int. 13, 397 – 409 (1978). 34. P. N. Singhellakis, A. E. Nikov, C. Nicolou et al., Effect of thiazide diuretics on parathyroid function in humans. Acta Endocrinol. Suppl. 261, 32 – 35 (1983). 35. K. Sakhaee, M. J. Nicar, K. Glass, J. E. Zerwekh, and C. Y. Pak, Reduction of intestinal calcium absorption by hydrochlorothiazide in postmenopausal osteoporosis. J. Clin. Endocrinol. Metab. 59, 1037 – 1043 (1984). 36. T. Steiniche, L. Mosekilde, M. S. Christensen, and F. Melsea, Histomorphometric analysis of bone in idiopathic hypercalciuria before and after treatment with thiazide. APMIS 97, 302 – 308 (1989). 37. R. O. Pierce and A. Perry, The effects of thiazide diuretics on bone mineral metabolism. J. Natl. Med. Assoc. 90, 46 – 50 (1998). 38. A. Z. LaCroix, Thiazide diuretic agents and prevention of hip fracture. Compr. Ther. 17, 30 – 39 (1991). 39. M. Nevitt, Falls in older persons: Risk factors and prevention. In “The Second Fifty Years: Promoting Health and Preventing Disability” (R. L. Berg and J. S. Cassells, eds.). National Academy Press, Washington, DC, 1990. 40. W. A. Ray, Thiazide diuretics and osteoporosis: Time for a clinical trial? Ann. Intern. Med. 115, 64 – 65 (1991). 41. D. K. Wysowski, L. Golden, and L. Burke, Use of menopausal estrogens and medroxyprogesterone in the United States, 1982 – 1992. Obstet. Gynecol. 85, 6 – 10 (1995). 42. V. A. Ravnikar, Compliance with hormone therapy. Am. J. Obstet. Gynecol. 156, 1332 (1987). 43. S. J. Birge, Soy phytoestrogens: An adjunct to hormone replacement therapy? Maturitas 7, 209 – 212 (2000). 44. National Osteoporosis Foundation, Osteoporosis: Review of the evidence for prevention, diagnosis and treatment and cost-effectiveness analysis. Osteopor. Int. Suppl. 4, S1 – 86 (1998). 45. R. Lindsay, Hormone replacement therapy for prevention and treatment of osteoporosis. Am. J. Med. 95, 37S–39S (1993). 46. C. Christiansen, Hormonal prevention and treatment of osteoporosisstate of art 1990. J. Steroid Biochem. Mol. Biol. 37, 447 – 449 (1990).
CAULEY AND SALAMONE 47. E. F. Eriksen, D. S. Colvard, N. J. Berg et al., Evidence of estrogen receptors in normal human osteoblast-like cells. Science 241, 84 – 86 (1988). 48. S. L. Corson, A decade of experience with transdermal estrogen replacement therapy: Overview of key pharmacologic and clinical findings. Int. J. Fertil. 38(2), 79 – 91 (1993). 49. J. C. Stevenson, M. P. Cust, K. F. Gangar, T. C. Hillard, B. Lees, and M. I. Whitehead, Effects of transdermal versus oral hormone replacement therapy on bone density in spine and proximal femur in postmenopausal women. Lancet 335, 265 – 269 (1990). 50. E. F. Holland, A. T. Leather, and J. W. Studd, The effect of 25-mg percutaneous estradiol implants on the bone mass of postmenopausal women. Obstet. Gynecol. 83, 43 – 46 (1994). 51. S. J. S. Ryde, K. Bowen-Simpkins, P. Bowen-Simpkins, W. D. Evans, W. D. Morgan, and J. E. Compston, The effect of oestradiol implants on regional and total bone mass: A three-year longitudinal study. Clin. Endocrinol. 40, 33 – 38 (1994). 52. B. J. Riis, K. Thomsen, V. Strom, and C. Christiansen, The effect of percutaneous estradiol and natural progesterone on postmenopausal bone loss. Am. J. Obstet. Gynecol. 156, 61 – 65 (1987). 53. R. Marcus, G. Greendale, B. A. Blunt, T. L. Bush, S. Sherman, R. Sherwin, H. Wahner, and B. Wells, Correlates of bone mineral density in the postmenopausal estrogen/progestin interventions trial. J. Bone Miner. Res. 9, 1467 – 1476 (1994). 54. J. C. Gallagher, W. T. Kable, and D. Goldgar, Effect of progestin therapy on cortical and trabecular bone: Comparison with estrogen. Am. J. Med. 90, 171 – 178 (1991). 55. R. Civitelli, D. Agnusdei, P. Nardi, F. Zacchei, L. V. Avioli, and C. Gennari, Effects of one-year treatment with oestrogens on bone mass, intestinal calcium absorption and 25-hydroxyvitamin D-1 hydroxylase reserve in postmenopausal osteoporosis. Calcif. Tissue Int. 42, 77 – 86 (1988). 56. C. Christiansen and B. J. Riis, 17b-Oestradiol and continuous norethisterone: A unique treatment for established osteoporosis in elderly women. J. Clin. Endocrinol. Med. 71, 836 – 841 (1990). 57. R. Lindsay and J. F. Tohme, Oestrogen treatment of patients with established postmenopausal osteoporosis. Obstet. Gynecol. 76, 290 – 295 (1990). 58. E. G. Lufkin, H. W. Wahner, W. M. O’Fallon, S. F. Hodgson, M. A. Kotowicz, A. W. Lane, H. L. Judd, R. H. Caplan, and B. L. Riggs, Treatment of postmenopausal osteoporosis with transdermal oestrogen. Ann. Int. Med. 117, 1 – 9 (1992). 59. A. B. Grey, T. F. Cundy, and I. R. Reid, Continuous combined oestrogen/progestin therapy is well tolerated and increases bone density at the hip and spine in post-menopausal osteoporosis. Clin. Endocrinol. 40, 671 – 677 (1994). 60. D. T. Felson, Y. Zhang, M. T. Hannon, D. T. Kiel, P. W. F. Wilson, and J. J. Anderson, The effect of postmenopausal estrogen therapy on bone density in elderly women. N. Engl. J. Med. 329, 1141 – 1146 (1993). 61. N. M. Resnick and S. L. Greenspan, “Senile” osteoporosis reconsidered. JAMA 261, 1025 – 1029 (1989). 62. M. E. Quigley, P. L. Martin, A. M. Burnier, and P. Brooks, Estrogen therapy arrests bone loss in elderly women. Am. J. Obstet. Gynecol. 156, 1516 – 1523 (1987). 63. C. W. Marx, G. E. Dailey, C. Cheney, V. C. Vint, and D. B. Muchmore, Do estrogens improve bone mineral density in osteoporotic women over age 65? J. Bone Miner. Res. 7, 1275 – 1279 (1992). 64. K. E. Ensrud, L. Palermo, D. M. Black, J. Cauley, M. Jergas, E. S. Orwoll, M. C. Nevitt, K. M. Fox, and S. R. Cummings, Hip and calcaneal bone loss increase with advancing age: Longitudinal results from the Study of Osteoporotic Fractures. J. Bone Miner. Res. 11, 1778 – 87 (1995). 65. D. L. Schneider, E. L. Barrett-Connor, and D. J. Morton, Timing of postmenopausal estrogen for optimal bone mineral density. JAMA 277, 543 – 547 (1997).
CHAPTER 30 Postmenopausal Endogenous/Exogenous Hormones 66. The Postmenopausal Estrogen/Progestin Interventions (PEPI) Trial Writing Group, Effects of hormone therapy on bone mineral density: Results from the Postmenopausal Estrogen/Progestin Interventions (PEPI) trial. JAMA 276, 1389 – 96 (1996). 67. R. R. Recker, M. Davies, R. M. Dowd, and R. P. Heaney, The effect of low-dose continuous estrogen and progesterone therapy with calcium and vitamin D on bone in elderly women: A randomized controlled trial. Ann. Intern. Med. 130, 897 – 904 (1999). 68. H. K. Genant, D. J. Baylink, J. C. Gallagher, S. T. Harris, P. Steiger, and M. Herber, Effect of estrone sulfate on postmenopausal bone loss. Am. J. Obstet Gynecol. 76, 579 – 84 (1990). 69. S. T. Harris, H. K. Genant, D. J. Baylink, J. C. Gallagher, S. K. Karp, M. A. McConnell et al., The effects of estrone (Ogen) on spinal bone density of postmenopausal women. Arch. Intern. Med. 151, 1980 – 4 (1991). 70. B. Ettinger, H. K. Genant, P. Steiger, and P. Madvig, Low-dosage micronized 17 -estradiol prevents bone loss in postmenopausal women. Am. J. Obstet. Gynecol. 6, 479 – 88 (1992). 71. S. Adami, R. Suppi, F. Bertoldo, M. Rossini, M. Residori, V. Maresca, and V. LoCascio, Transdermal estradiol in the treatment of postmenopausal bone loss. Bone Miner. 7, 79 – 86 (1989). 72. C. S. Field, S. J. Ory, H. W. Wahner, R. R. Herrmann, H. L. Judd, and B. L. Riggs, Preventive effects of transdermal 17 beta-estradiol on osteoporotic changes after surgical menopause: A two-year placebo-controlled trial. Am. J. Obstet. Gynecol. 168, 114 – 21 (1993). 73. H. K. Genant, J. Lucas, S. Weiss, M. Akin, R. Emkey, H. McNancyFlint, R. Downs, J. Mortola, N. Watts, H. M. Yang, N. Banav, J. J. Brennan, and J. C. Nolan, for the Estratab/Osteoporosis Study Group, Low-dose estrified estrogen therapy: Effects on bone, plasma estradiol concentrations, endometrium and lipid levels. Arch. Intern. Med. 157, 2609 – 15 (1997). 74. L. Speroff, J. Rowan, J. Symons, H. Genant, and W. Wilborn for the CHART study group, The comparative effect on bone density, endometrium and lipids of continuous hormones as replacement therapy (CHART study). JAMA 276, 1397 – 1403 (1996). 75. T. A. Hutchinson, S. M. Polansky, and A. R. Feinstein, Postmenopausal estrogens protect against fractures of hip and distal radius: A case-control study. Lancet 6, 705 – 709 (1979). 76. N. S. Weiss, C. L. Ure, J. H. Ballard, A. R. Williams, and J. R. Daling, Decreased risk of fractures of the hip and lower forearm with postmenopausal use of estrogen. N. Engl. J. Med. 303, 1195 – 1198 (1980). 77. R. E. Johnson and E. E. Specht, The risk of hip fracture in postmenopausal females with and without estrogen drug exposure. Am. J. Public Health 71, 138 – 144 (1981). 78. N. Krieger, J. L. Kelsey, T. R. Holford, and T. O’Connor, An epidemiologic study of hip fracture in postmenopausal women. Am. J. Epidemiol. 116, 141 – 148 (1982). 79. J. A. Kanis, O. Johnell, B. Gullberg, E. Allander, G. Dilsen, C. Gennari et al., Evidence for efficacy of drugs affecting bone metabolism in preventing hip fracture. Br. Med. J. 305, 1124 – 1128 (1992). 80. C. B. Hammond, F. R. Jelovsek, K. L. Lee, W. T. Creasman, and R. T. Parker, Effects of long-term estrogen replacement therapy. Am. J. Obstet. Gynecol. 122, 525 – 536 (1979). 81. B. Ettinger, H. K. Genant, and C. E. Cann, Long-term estrogen replacement therapy prevents bone loss and fractures. Ann. Intern. Med. 102, 319 – 324 (1985). 82. D. P. Kiel, D. T. Felson, J. J. Anderson, P. W. Wilson, and M. A. Moskowitz, Hip fracture and the use of estrogens in postmenopausal women: The Framingham Study. N. Engl. J. Med. 317, 1169 – 1174 (1987). 83. A. Paganini-Hill, A. Chao, R. K. Ross, and B. E. Henderson, Exercise and other factors in the prevention of hip fracture: The Leisure World Study. Epidemiology 2, 16 – 25 (1991).
765 84. P. Maxim, B. Ettinger, and G. M. Spitalny, Fracture protection provided by long-term estrogen treatment. Osteopor. Int. 5, 23 – 29 (1995). 85. T. Naessen, I. Persson, H. O. Adama, R. Bergstrom, and L. Bergkvist, Hormone replacement therapy and the risk for the first hip fracture: A prospective, population-based cohort study. Ann. Intern. Med. 113, 95 – 103 (1990). 86. J. A. Cauley, D. G. Seeley, K. Ensrud, B. Ettinger, D. Black, and S. R. Cummings, Estrogen replacement therapy and fractures in older women. Ann. Intern. Med. 122, 9 – 16 (1995). 87. D. Grady, S. M. Rubin, D. B. Petitti, C. S. Fox, D. Black, B. Ettinger et al., Hormone therapy to prevent disease and prolong life in postmenopausal women. Ann. Intern. Med. 117, 1016 – 1037 (1992). 88. D. P. Kiel, J. A. Baron, J. J. Anderson, M. T. Hannan, and D. T. Felson, Smoking eliminates the protective effect of oral estrogens on the risk for hip fracture among women. Ann. Intern. Med. 116, 716 – 721 (1992). 89. A. R. Williams, N. S. Weiss, C. L. Ure, J. Ballard, and J. R. Daling, Effect of weight, smoking, and estrogen use on the risk of hip and forearm fractures in postmenopausal women. Obstet. Gynecol. 60, 695 – 696 (1992). 90. L. E. Nachtigall, R. H. Nachtigall, R. D. Nachtigall, and E. M. Beckman, Estrogen replacement therapy I: A 10-year prospective study in the relationship to osteoporosis. Am. J. Obstet. Gynecol, 53, 277 – 81 (1979). 91. R. Lindsay, D. M. Hart, C. Forrest and C. Baird, Prevention of spinal osteoporosis in ooporetctomized women. Lancet 2, 1151 – 53 (1980). 92. J. Windeler and S. Lange, Events per person year: A dubious concept. Br. Med. J. 310, 454 – 56 (1995). 93. M. H. Komulainen, H. Kroger, M. T. Tuppurainen, A. M. Heikkinen, E. Alhava, R. Honkanen, and S. Saarikowski, HRT and vitamin D in prevention of nonvertebral fractures in postmenopausal women: A 5year randomized trial. Maturitas 31, 45 – 54 (1998). 94. S. Hulley, D. Grady, T. Bush et al., Randomized trial of estrogen plus progestin for secondary prevention of coronary heart disease in postmenopausal women: Heart and Estrogen/progestin Replacement Study (HERS) research group. JAMA 380, 605 – 13 (1998). 95. J. A. Cauley, D. M. Black, E. Barrett-Connor, F. Harris, K. Shields, W. Applegate, and S. R. Cummings, Effects of hormone replacement therapy on clinical fractures and height loss: The Heart Estrogen/progestin Replacement Study (HERS). Submitted for publication. 96. B. Ettinger and D. Grady, Maximizing the benefit of estrogen therapy for prevention of osteoporosis. Menopause 1, 19 – 36(1994). 97. D. B. Petititi, Coronary heart disease and estrogen replacement therapy: Can compliance bias explain the results of observational studies? Ann. Epidemiol. 4, 115 – 118 (1994). 98. C. B. Hammond, F. R. Jelovsek, K. L. Lee, W. T. Creasman, and R. T. Parker, Effects of long-term estrogen replacement therapy. II. Neoplasia. Am. J. Obstet. Gynecol. 233, 537 – 547 (1979). 99. A. H. MacLennan, A. MacLennan, and D. Wilson, The prevalence of estrogen replacement therapy in South Australia. Maturitas 16, 175 – 183 (1993). 100. F. P. Groeneveld, F. P. Bareman, R. Barentsen, H. J. Dokter, A. C. Drogendijk, and A. W. Hoes, Determinants of first prescription of hormone replacement therapy: A follow-up study among 1689 women aged 45 – 60 years. Maturitas 20, 81 – 89 (1995). 101. L. M. Salamone, A. R. Pressman, D. G. Seeley, and J. A. Cauley, Estrogen replacement therapy: A survey of elderly women. Submitted for publication. 102. A. Cano, Compliance to hormone replacement therapy in menopausal women controlled in a third level academic centre. Maturitas 20, 91 – 99 (1995). 103. C. C. Johnston, S. L. Hui, R. M. Witt et al., Early menopausal changes in bone mass and sex steroids. J. Clin. Endocrinol. Metab. 61, 905 – 911 (1985).
766 104. J. A. Cauley, J. P. Gutai, R. V. Sandler et al., The relationship of endogenous estrogen to bone density and bone area in normal postmenopausal women. Am. J. Epidemiol. 124, 752 – 761 (1986). 105. J. A. Cauley, J. P. Gutai, L. H. Kuller, J. Scott, and M. C. Nevitt, Black-white differences in serum sex hormones and bone mineral density. Am. J. Epidemiol. 139, 1035 – 1046 (1994). 106. K. K. Steinberg, L. W. Freni-Titulaer, E. G. DePuey et al., Sex steroids and bone density in premenopausal and perimenopausal women. J. Clin. Endocrinol. Metab. 69, 533 – 539 (1989). 107. S. Rozenberg, H. Ham, D. Bosson et al., Age, steroids and bone mineral content. Maturitas 12, 137 – 143 (1990). 108. S. Murphy, K. Khaw, M. J. Sneyd et al., Endogenous sex hormones and bone mineral density among community-based postmenopausal women. Postgrad. Med. J. 68, 908 – 913 (1992). 109. G. M. Hall, L. A. Perry, and T. D. Spector, Depressed levels of dehydroepiandrosterone sulphate in postmenopausal women with rheumatoid arthritis but no relation with axial bone density. Ann. Rheum. Dis. 52, 211 – 214 (1993). 110. I. S. Fentiman, D. Y. Wang, D. S. Allen et al., Bone density of normal women in relation to endogenous and exogenous oestrogens. Br. J. Rheumatol. 33, 808 – 815 (1994). 111. T. Murakami, M. Shiraki, H. Orimo et al., Serum estradiol and radial mineral content in postmenopausal females. Endocrinol. Jpn. 26, 635 – 636 (1979). 112. M. H. Rae, P. A. Mole, and C. R. Paterson, Endogenous factors affecting bone mineral content in post-menopausal women. Maturitas 13, 319 – 324 (1991). 113. S. Deutsch, F. Benjamin, V. Seltzer et al., The correlation of serum estrogens and androgens with bone density in the late postmenopause. Int. J. Gynaecol. Obstet. 25, 217 – 222 (1987). 114. M. E. Nelson, C. N. Meredith, B. Dawson-Hughes et al., Hormone and bone mineral status in endurance-trained and sedentary postmenopausal women. J. Clin. Endocrinol. Metab. 66, 927 – 933 (1988). 115. C. C. Johnston, S. L. Hui, and C. Longcope, Bone mass and sex steroid concentrations in postmenopausal caucasian diabetics. Metabolism 34, 544 – 550 (1985). 116. B. Ettinger, A. Pressman, P. Sklarin, D. C. Bauer, J. A. Cauley, and S. R. Cummings, Associations between low levels of serum estradiol, bone density, and fractures among elderly women: The Study of Osteoporotic Fractures. J. Clin. Endocrinol. Metab. 83, 2239 – 2243 (1998). 117. S. Khosla, L. J. Melton III, E. J. Atkinson, W. M. O’Fallon, G. G. Klee, and B. L. Riggs, Relationship of serum sex steroid levels and bone turnover markers with bone mineral density in men and women: A key role for bioavailable estrogen. J. Clin. Endocrinol. Metab. 83, 2266 – 2274 (1998). 118. R. Lindsay, D. M. Hart, A. Sweeney et al., Osteopenia and osteoporosis. Calcif. Tissue Res. 22 (Suppl.), 213 – 216 (1977). 119. J. A. Johnston, R. A. Khairi, and C. Longcope, Age-related bone loss. In “Osteoporosis II”, (U.S. Barzel, ed.), pp. 91 – 100. Grune and Stratton, New York, 1979. 120. S. C. Manolagas, D. C. Anderson, and R. Lindsay, Adrenal steroids and the development of osteoporosis in oophorectomised women. Lancet 3143, 597 – 600 (1979). 121. B. J. Riis, C. Christiansen, L. J. Deftos et al., The role of serum concentrations of estrogens on postmenopausal osteoporosis and bone turnover. In “Osteoporosis I” (C. Christiansen, ed.), pp. 333 – 336. Stifsbogtrykkeri Printers, Aalborg, 1984. 122. B. J. Riis, P. Rodbro, and C. Christiansen, The role of serum concentrations of sex steroids and bone turnover in the development and occurrence of postmenopausal osteoporosis. Calcif. Tissue Int. 38, 318 – 322 (1986). 123. J. F. Aloia, A. N. Vaswani, J. K. Yeh et al., Determinants of bone mass in postmenopausal women. Arch. Intern. Med. 143, 1700 – 1704 (1983).
CAULEY AND SALAMONE 124. C. Slemenda, S. L. Hui, C. Longcope et al., Sex steroids and bone mass: A study of changes about the time of menopause. J. Clin. Invest. 80, 1261 – 1269 (1987). 125. A. M. van Hemert, J. C. Birkenhager, F. H. De Jong et al., Sex hormone binding globulin in postmenopausal women: A predictor of osteoporosis superior to endogenous oestrogens. Clin. Endocrinol. 31, 499 – 509 (1989). 126. S. Harris, G. E. Dallal, and B. Dawson-Hughes, Influence of body weight on rates of change in bone density of the spine, hip and radius in postmenopausal women. Calcif. Tissue Int. 50, 19 – 23 (1992). 127. E. C. H. Van Beresteijn, J. P. R. M. Van Laarhoven, and A. G. H. Smals, Body weight and/or endogenous estradiol as determinants of cortical bone mass and bone loss in healthy early postmenopausal women. Acta Endocrinol. 127, 226 – 230 (1992). 128. K. Stone, D. C. Bauer, D. M. Black, P. Sklarin, K. E. Ensrud, and S. R. Cummings, for the Study of Osteoporotic Fractures research group, Hormonal predictors of bone loss in elderly women: A prospective study. J. Bone. Miner. Res. 13, 1167 – 1174 (1998). 129. D. H. Marshall, R. G. Crilly, and B. E. C. Nordin, Plasma androstenedione and oestrone levels in normal and osteoporotic postmenopausal women. Br. Med. J. 2, 1177 – 1179 (1977). 130. B. Pelc, D. H. Marshall, P. Guha et al., The relation between plasma androstenedione, plasma oestrone and androstenedione to oestrone conversion rates in post-menopausal women with and without fractures. Clin. Sci. Mol. Med. 54, 125 – 131 (1978). 131. B. J. Davidson, B. L. Riggs, C. B. Coulam et al., Concentration of cytosolic estrogen receptors in patients with postmenopausal osteoporosis. Am. J. Obstet. Gynecol. 136, 430 – 434 (1980). 132. B. J. Davidson, R. K. Ross, A. Paganini-Hill et al., Total and free estrogens and androgens in postmenopausal women with hip fractures. J. Clin. Endocrinol. Metab. 54, 115 – 120 (1982). 133. L. R. Laufer, B. J. Davidson, R. K. Ross et al., Physical characteristics and sex hormone levels in patients with osteoporotic hip fractures or endometrial cancer. Am. J. Obstet. Gynecol. 145, 585 – 590 (1983). 134. J. F. Aloia, S. F. Cohn, A. Vaswani et al., Risk factors for postmenopausal osteoporosis. Am. J. Med. 78, 95 – 100 (1985). 135. B. L. Riggs, R. J. Ryan, H. W. Wahner et al., Serum concentrations of estrogen, testosterone and gonadotropins in osteoporotic and nonosteoporotic postmenopausal women. J. Clin. Endocrinol. Metab. 36, 1097 – 1099 (1973). 136. F. J. Bartizal, C. B. Coulam, T. A. Gaffey et al., Impaired binding of estradiol to vaginal mucosal cells postmenopausal osteoporosis. Gynecol. Invest. 7, 330 – 336 (1976). 137. B. J. Davidson, B. L. Riggs, H. W. Wahner et al., Endogenous cortisol and sex steroids in patients with osteoporotic spinal fractures. Obstet. Gynecol. 61, 275 – 278 (1983). 138. C. Longcope, R. S. Baker, S. L. Hur et al., Androgen and estrogen dynamics in women with vertebral crush fractures. Maturitas 6, 309 – 318 (1984). 139. E. Barrett-Connor, J. E. Mueller, D. G. Von Mühlen, G. A. Laughlin, D. L. Schneider, and D. J. Sartoris, Low levels of estradiol are associated with vertebral fractures in older men, but not women: The Rancho Bernardo study. J. Clin. Endocrinol. Metab. 85, 219 – 223 (2000). 140. S. R. Cummings, W. S. Browner, D. Bauer, K. Stone, K. Ensrud, S. Jamal, and B. Ettinger for the Study of Osteoporotic Fractures research group, Endogenous hormones and the risk of hip and vertebral fractures among older women. N. Engl. J. Med. 339, 733 – 8 (1998). 141. N. F. Santoro, N. F. Col, M. H. Eckman, J. B. Wong, S. G. Pauker, J. A. Cauley, J. Zmuda, S. Crawford, C. B. Johannes, J. E. Rossouw, and C. N. Merz, Therapeutic controversy: Hormone replacement therapy: Where are we going? J. Clin. Endocrinol. Metab. 84(6), 1798 – 812 (1999).
CHAPTER 30 Postmenopausal Endogenous/Exogenous Hormones 142. J. C. Gallagher, Moderation of the daily dose of HRT: Prevention of osteoporosis. Maturitas 33, S57 – S63 (1999). 143. R. J. Kuczmarski, K. M. Flegal, S. M. Campbell, and C. L. Johnson, Increasing prevalence of overweight among US adults: The National Health and Nutrition Examination Surveys. JAMA 272, 205 – 211 (1994). 144. Metropolitan Life Insurance Company, Metropolitan height and weight tables. Stat. Bull. Metrop. Insur. Co. 64, 2 – 9 (1983). 145. National Institutes of Health, Health implications of obesity: National Institutes of Health Consensus Development Conference Statement. Ann. Intern. Med. 103, 1073 – 1077 (1985). 146. A. B. Grey, M. C. Evans, J. P. Stapleton, and J. R. Reid, Body weight and bone mineral density in postmenopausal women with primary hyperparathyroidism. Am. College Phys. 121, 745 – 749 (1994). 147. T. L. Holbrook and E. Barrett-Connor, The association of lifetime weight and weight control patterns with bone mineral density in an adult community. Bone Miner. 20, 141 – 149 (1993). 148. F. A. Tremollieres, J. M. Pouilles, and C. Ribot, Vertebral postmenopausal bone loss is reduced in overweight women: A longitudinal study in 155 early postmenopausal women. J. Clin. Endocrinol. Metab. 77, 683 – 686 (1993). 149. S. Harris, G. E. Dallal, and B. Dawson-Hughes, Influence of body weight on rates of change in bone density of the spine, hip, and radius in postmenopausal women. Calcif. Tissue Int. 50, 19 – 23 (1992). 150. R. Lindsay, F. Cosman, B. S. Herrington, and S. Himmelstein, Bone mass and body composition in normal women. J. Bone Miner. Res. 7, 55 – 63 (1992). 151. K. Kin, K. Kushida, K. Yamazaki, S. Okamoto, and T. Inoue, Bone mineral density of the spine in normal Japanese subjects using dualenergy X-ray absorptiometry: Effect of obesity and menopausal status. Calcif. Tissue Int. 49, 101 – 106 (1991). 152. E. C. H. van Beresteijn, P. R. M. Laarhoven, and A. G. H. Smals, Body weight and/or endogenous estradiol as determinants of cortical bone loss in healthy early postmenopausal women. Acta Endocrinol. 127, 226 – 230 (1992). 153. W. C. Bevier, R. A. Wiswell, G. Pyka, K. C. Kozak, K. M. Newhall, and R. Marcus, Relationship of body composition, muscle strength, and aerobic capacity to bone mineral density in older men and women. J. Bone Miner. Res. 4, 421 – 432 (1989). 154. N. Pocock, J. Eisman, T. Gwinn, P. Sambrook, P. Kelly, J. Freun, and M. Yates, Muscle strength, physical fitness and weight but not age predict femoral neck bone mass. J. Bone Miner. Res. 4, 441 – 447 (1989). 155. J. C. Stevenson, B. Lees, M. Devenport, M. P. Cust, and K. F. Ganger, Determinants of bone density in normal women: Risk factors for future osteoporosis? Br. Med. J. 298, 925 – 928 (1989). 156. B. Dawson-Hughes, C. Shipp, L. Sadowski, and G. Dallal, Bone density of the radius, spine and hip in relation to percent ideal body weight in postmenopausal women. Calcif. Tissue Int. 40, 310 – 314 (1987). 157. C. Ribot, F. Tremollieres, J. M. Pouilles, M. Bonneu, F. German, and J. P. Louvet, Obesity and postmenopausal bone loss: The influence of obesity on vertebral density and bone turnover in postmenopausal women. Bone 8, 327 – 331 (1988). 158. N. A. Pocock, J. A. Eisman, M. G. Yeates, P. N. Sambrook, and S. Eberl, Physical fitness is a major determinant of femoral neck and lumbar spine bone mineral density. J. Clin. Invest. 78, 618 – 621 (1986). 159. H. W. Daniell, Osteoporosis of the slender smoker. Arch. Intern. Med. 136, 298 – 304 (1976). 160. N. Dalen, D. Hallberg, and B. Lamke, Bone mass in obese subjects. Acta Med. Scand. 197, 353 – 355 (1975). 161. Y. Nishizawa, H. Koyama, T. Shoji, H. Aratani, S. Hagiwara, T. Miki, and H. Morii, Obesity as a determinant of regional bone mineral density. J. Nutr. Sci. Vitaminol. 37, S65 – S70 (1991).
767 162. G. M. Wardlaw, Putting body weight and osteoporosis into perspective. Am. J. Clin. Nutr. 63(Suppl.), 433S – 436S (1996). 163. S. M. Haffner and R. L. Bauer, The association of obesity and glucose and insulin ocncentrations with bone density in premenopausal and postmenopausal women. Metabolism. 42, 735 – 8 (1993). 164. M. Shiraki, H. Ito, H. Fujimaki, and T. Higuchi, Relation between body size and bone mineral density with special reference to sex hormones and calcium regulating hormones in elderly females. Endocrinol. Jpn. 38, 343 – 9 (1996). 165. D. T. Felson, Y. Zhang, M. T. Hannan, and J. J. Anderson, Effects of weight and body mass index on bone mineral density in men and women: The Framingham study. J. Bone Miner. Res. 8, 567 – 573 (1993). 166. S. L. Edelstein, and E. Barrett-Connor, Relation between body size and bone mineral density in elderly men and women. Am. J. Epidemiol. 138, 160 – 169 (1993). 167. K. Yano, R. D. Wasnich, J. M. Vogel, and L. K. Heilbrun, Bone mineral measurements among middle-aged and elderly Japanese residents in Hawaii. Am. J. Epidemiol. 119, 751 – 764 (1984). 168. P. Dano and C. Christiansen, Calcium absorption and bone mineral content following intestinal shunt operation in obesity: A comparison of three types of operation. Scand. J. Gastroenterol. 9, 775 – 779 (1974). 169. J. E. Compston, S. Vedi, E. Giantta, G. Watson, D. Civalleri, and N. Scopinaro, Bone histomorphometry and vitamin D status after biliopancreatic bypass for obesity. Gastroenterology 87, 350 – 356 (1984). 170. J. M. Campbell, T. K. Hunt, J. H. Karam, and P. H. Forsham, Jejunoilieal bypass as a treatment of morbid obesity. Arch. Intern. Med. 137, 602 – 610 (1977). 171. H. Rickers, I. Balslev, H. Foltved, and P. Rodbro, Bone mineral content before and after intestinal bypass operation in obese patients. 209, 203 – 207 (1981). 172. H. Hey, B. Lund, O. H. Sorensen, B. Lund, and M. S. Christensen, Impairment of vitamin D and bone metabolism in patients with bypass operation for obesity. Acta Med. Scand. (Suppl.) 624, 73 – 78 (1979). 173. A. M. Parfitt, M. J. Miller, B. Frame, A. R. Villanueva, I. Olivier, and D. L. Thomson, Metabolic bone disease after intestinal bypass for treatment of obesity. Ann. Intern. Med. 89, 193 – 199 (1978). 174. L. B. Jensen, F. Quaade, and O. L. Sorensen, Bone loss accompanying voluntary weight loss in obese humans. J. Bone Miner. Res. 9, 459 – 463 (1994). 175. J. E. Compston, M. A. Laskey, P. I. Croucher, A. Coxon, and S. Keitzman, Effect of diet-induced weight loss of total body bone mass. Clin. Sci. 82, 429 – 432 (1992). 176. A. Avenell, P. R. Richmond, M. E. J. Jean, and D. M. Reid, Bone loss associated with a high-fibre weight reduction diet in postmenopausal women. Eur. J. Clin. Nutr. 48, 561 – 6 (1994). 177. L. M. Salamone, J. A. Cauley, D. M. Black, L. Simkin-Silverman, W. Lang, E. Gregg, L. Palermo, R. S. Epstein, L. H. Kuller, and R. Wing, Effect of a lifestyle intervention on bone mineral density in premenopausal women: A randomized trial. Am. J. Clin. Nutr. 70, 97 – 103 (1999). 178. O. Svendsen, C. Hassager, and C. Christiansen, Effects of an energy restrictive diet, with or without exercise on lean tissue mass, resting metabolic rate, cardiovascular risk factors and bone in overweight postmenopausal women. Am. J. Med. 95, 131 – 140 (1993). 179. B. Mazess, P. R. Judy, C. R. Wilson, and J. R. Cameron, Progress in clinical use of photon absorptiometry: Clinical aspects of metabolic bone disease. Excerpta Med. Amst. 37 – 41 (1973). 180. L. Hyldstrup, T. Andersen, P. McNair, L. Breum, and I. Transbol, Bone metabolism in obesity: Changes related to severe overweight and dietary weight reduction. Acta Endocrinol. 129, 393 – 398 (1993).
768 181. J. E. Pritchard, C. A. Nowson, B. J. Strauss, J. S. Carlson, B. Kaymakci, and J. D. Wark, Evaluation of dual energy x-ray absorptiometry as a method of measurement of body fat. Eur. J. Clin. Nutr. 47, 216 – 228 (1993). 182. R. Roubenoff and J. J. Kehayias, The meaning and measurements of lean body mass. Nutr. Rev. 46, 163 – 175 (1991). 183. S. B. Heymsfield and M. Waki, Body composition in humans: Advances in the development of multicompartment chemical models. Nutr. Rev. 49, 97 – 108 (1991). 184. R. Roubenoff, J. J. Kehayias, B. Dawson-Hughes, and S. Heymsfield, Use of dual-energy x-ray absorptiometry in body composition studies: Not yet a “gold standard.” Am. J. Clin. Nutr. 58, 589 – 591 (1993). 185. J. Johnson and B. Dawson-Hughes, Precision and stability of dualenergy x-ray absorptiometry measurements. Calcif. Tissue Int. 49, 174 – 178 (1991). 186. K. Buhl, S. B. Heymsfield, M. Russel-Aulet, J. Wang, R. N. Pierson, and S. Lichtman, Effect of tissue thickness on bone density and bone mineral by dual energy x-ray absorptiometry. FASEB J. 5, A924 (1991). [Abstract] 187. P. Martin, C. Verhas, C. Als, I. Geerts, J. Paternot, and P. Bergmann, Influence of patient’s weight on dual-photon absorptiometry and dual-energy x-ray absorptiometry measurements of bone mineral density. Osteopor. Int. 3, 198 – 203 (1993). 188. C. Hassager and C. Christiansen, Influence of soft-tissue body composition on bone mass and metabolism. Bone 10, 415 – 419 (1989). 189. H. Rickers, I. Balstev, H. Foldved, and P. Rodbro, Bone mineral content before and after intestinal bypass operation in obese patients. Acta Med. Scnd. 209, 203 – 207 (1981). 190. T. N. Hangartner and C. C. Johnston, Influence of fat on bone measurements with dual-energy absorptiometry. Bone Miner. 9, 71 – 81 (1990). 191. J. Haarbo, A. Godtfredsen, C. Hassager, and C. Christiansen, Validation of body composition by dual-energy x-ray absorptiometry (DEXA). Clin. Physiol. 11, 331 – 341 (1991). 192. O. L. Svendsen, C. Haarbo, Hassager, and C. Christiansen, Accuracy of measurements of body composition by dual energy x-ray absorptiometry in vivo. Am. J. Clin. Nutr. 57, 605 – 8 (1993). 193. M. D. Van Loan, and P. L. Mayclin, Body composition assessment: Dual energy x-ray absorptiometry (DXA) compared to reference methods. Eur. J. Clin. Nutr. 46, 125 – 30 (1992). 194. M. D. Van Loan, H. L. Johnson, and T. F. Barbieri, Effect of weight loss on bone mineral content and bone mineral density in obese women. Am. J. Clin. Nutr. 67, 734 – 8 (1998). 195. L. D. Ritchie, J. C. King, E. B. Fung, M. D. Van Loan, and C. E. Cann, Pregnancy and lactation cause spinal bone mineral loss in humans. FASEB J. 7, A67 (1993). (abstract) 196. I. R. Reid, R. Ames, M. C. Evans, S. Sharpe, G. Gamble, J. T. France, T. M. T. Lim, and T. F. Cundy, Determinants of total body and regional bone mineral density in normal postmenopausal women: A key role for fat mass. J. Clin. Endocrinol. Metab. 75, 45 – 51 (1992). 197. J. E. Compston, M. Bhambhani, M. A. Laskey, S. Murphy, and K. T. Khaw, Body composition and bone mass in postmenopausal women. Clin. Endocrinol. 37, 426 – 431 (1992). 198. J. F. Aloia, A. Vaswani, R. Ma, and E. Flaster, To what extent is bone mass determined by fat-free or fat mass? Am. J. Clin. Nutri. 61, 1110 – 1114 (1995). 199. F. Doyle, J. Brown, and C. Lachance, Relation between bone mass and muscle weight. Lancet 1, 391 – 393 (1970). 200. M. Sinaki, M. C. C. McPhee, S. F. Hodgson, J. M. Merrit, and M. S. Offord, Relationship between bone mineral density of spine and strength of back extensors healthy postmenopausal women. Mayo Clin. Proc. 61, 116 – 122 (1986). 201. J. F. Aloia, D. M. McGowan, A. N. Vaswani, P. Ross, and S. H. Cohn, Relationship of menopause to skeletal and muscle mass. Am. J. Clin. Nutr. 53, 1378 – 1383 (1991).
CAULEY AND SALAMONE 202. D. R. Carter, M. L. Bouxsein, and R. Marcus, New approaches for interpreting projected bone densitometry data. J. Bone. Miner. Res. 7, 137 – 145 (1992). 203. I. R. Reid, M. C. Evans, and R. W. Ames, Volumetric bone density of the lumbar spine is related to fat mass but not lean mass in normal postmenopausal women. Osteopor. Int. 4, 362 – 367 (1994). 204. S. S. Harris, and B. Dawson-Hughes, Weight, body composition, and bone density in postmenopausal women. Calcif. Tissue. Int. 59, 428 – 432 (1996). 205. R. N. Baumgartner, P. M. Stauber, K. M. Koehler, L. Romero, and P. J. Garry, Associations of fat and muscle masses with bone mineral in elderly men and women. Am. J. Clin. Nutri. 63, 365 – 72 (1996). 206. Z. Chen, T. G. Lohman, W. A. Stini, C. Ritenbaugh, and M. Aickin, Fat or lean tissue mass: Which one is the major determinant of bone mineral mass in healthy postmenopausal women? J. Bone Miner. Res. 12, 144 – 151 (1997). 207. R. B. Sandler, Muscle strength assessments and the prevention of osteoporosis. JAGS 37, 1192 – 1197 (1989). 208. T. V. Nguyen, G. M. Howard, P. J. Kelly, and J. A. Eisman, Bone mass, lean mass and fat mass: Same genes or same environment? Am. J. Epidemiol. 147, 3 – 16 (1998). 209. M. M. Hla, J. W. Davis, P. D. Ross, R. D. Wasnich, J. A. Yates, P. Ravn, D. J. Hosking, and M. R. McClung, A multicenter study of the influence of fat and lean mass on bone mineral content: Evidence for differences in their relative influence at major fracture sites. Am. J. Clin. Nutr. 64, 354 – 360 (1996). 210. G. A. Rodan, Mechanical loading, estrogen deficiency and the coupling of bone formation and bone resorption. J. Bone Miner. Res. 6, 527 – 530 (1991). 211. R. G. Cumming and R. J. Klineberg, Case-control study of risk factors for hip fractures in the elderly. Am. J. Epidemiol. 139, 493 – 503 (1994). 212. S. B. Jaglal, N. Kreiger, and G. Darlington, Past and recent physical activity and risk of hip fracture. Am. J. Epidemiol. 138, 107 – 118 (1993). 213. J. L. Kelsey, W. S. Browner, D. G. Seeley, M. C. Nevitt, and S. R. Cummings, Risk factors for fractures of the distal forearm and proximal humerus. Am. J. Epidemiol. 135, 477, 489 (1992). 214. M. E. Farmer, T. Harris, J. H. Madans, R. B. Wallace, J. CornoniHuntley, and L. R. White, Anthropometric indicators and hip fracture. The NHANES I epidemiologic follow-up study. J. Am. Geriatr. Soc. 37, 9 – 16 (1989). 215. P. Gardsell, O. Johnell, and B. E. Nilsson, Predicting fractures in women by using forearm densitometry. Calcif. Tissue Int. 44, 235 – 242 (1989). 216. M. E. Pruzansky, M. Turano, M. Luckey, and R. Senie, Low body weight as a risk factor for hip fracture in both black and white women. J. Orthop. Res. 7, 192 – 197 (1989). 217. D. Hemenway, G. A. Colditz, W. C. Willet, M. J. Stampfer, and F. E. Speizer, Fractures and lifestyle: Effect of cigarette smoking, alcohol intake, and relative weight on the risk of hip and forearm fractures in middle-aged women. Am. J. Public Health 78, 1554 – 1558 (1988). 218. B. W. Alderman, N. S. Weiss, J. R. Daling, C. L. Ure, and J. H. Ballard, Reproductive history and postmenopausal riskofhip and forearm fracture. Am. J. Epidemiol. 124, 262, 267 (1986). 219. R. Wooton, E. Breyson, U. Elsasser, H. Freeman, J. R. Green, and R. Hesp, Risk factors for fractured neck of femur in the elderly. Age Ageing 11, 160 – 168 (1982). 220. D. C. Bauer, W. S. Browner, J. A. Cauley et al., Factors associated with appendicular bone mass in older women. Ann. Int. Med. 118, 657 – 665 (1993). 221. S. R. Cummings, M. C. Nevitt, W. S. Browner, K. Stone, K. M. Fox, K. E. Ensrud, J. A. Cauley, D. Black, and T. M. Vogt, Risk factors for hip fracture in white women. N. Engl. J. Med. 332, 767 – 773 (1995).
CHAPTER 30 Postmenopausal Endogenous/Exogenous Hormones 222. R. Nuti, G. Martini, Measurements of bone mineral density by DXA total body absorptiometry in different skeletal sites in postmenopausal osteoporosis. Bone 13, 173 – 178 (1992). 223. H. Rico, M. Revilla, L. F. Villa, E. R. Hernandez, and J. P. Fernandez, Crush fracture syndrome in senile osteoporosis: A nutritional consequence? J. Bone Miner. Res. 7, 317 – 319 (1992). 224. M. E. Ooms, P. Lips, A. Van Lingen, and H. A. Valkenburg, Determinants of bone mineral density and risk factors for osteoporosis in healthy elderly women. J. Bone Miner. Res. 8, 66 – 75 (1993). 225. S. Elmstahl, P. Gardsell, K. Ringsberg, and I. Sernbo, Body composition and its relation to bone mass and fractures in an urban and rural population. Aging Clin. Exp. Res. 5, 47 – 54 (1993). 226. C. Longcope, H. Pratt, S. H. Schneider, and S. E. Fineberg, Aromatization of androgens by muscle and adipose tissue in vivo. J. Clin. Endocrinol. Metab. 46, 146 – 152 (1978). 227. P. Siiteri, Adipose tissue as a source of hormones. Am. J. Clin. Nutr. 45, 277 – 282 (1987). 228. J. M. Grodin, P. K. Siiteri, and P. C. MacDonald, Source of estrogen production in postmenopausal women. J. Clin. Endocrinol. Metab. 36, 207 – 214 (1973). 229. P. C. MacDonald, C. D. Edman, D. L. Hemsell, J. C. Porter, and P. K. Siiteri, Effect of obesity on conversion of plasma androstenedione to estrone in postmenopausal women with endometrial cancer. Am. J. Obstet. Gynecol. 130, 448 – 455 (1978). 230. C. Longcope and S. Baker, Androgen and estrogen dynamics: Relationships with age, weight and menopausal status. J. Clin. Endocrinol. Metab. 76, 601 – 604 (1993). 231. M. A. Kirshner, E. Samojlik, M. Drejka, E. Szmal, G. Schneider, and N. H. Ertel, Androgen-estrogen metabolism in women with upper body vs. lower body obesity. J. Clin. Endocrinol. Metab. 70, 473 – 478 (1990). 232. E. Samojlik, M. A. Kirschner, D. Silber, G. Schneider, and N. H. Ertel, Elevated production and metabolic clearance rates of androgen in morbidity obese women. J. Clin. Endocrinol. Metab. 49, 949 – 954 (1984). 233. D. Goodman-Gruen, and E. Barrett-Connor, Total but not bioavailable testosterone is a predictor of central adiposity in postmenopausal women. Int. J. Obesity 19, 293 – 298 (1995). 234. S. R. Cummings, J. L. Kelsey, M. C. Nevitt, and K. J. O’Dowd, Epidemiology of osteoporosis and osteoporotic fractures. Epidemiol. Rev. 7, 178 – 208 (1985). 235. C. Christiansen and R. Lindsay, Estrogens, bone loss and preservation. Osteopor. Int. 1, 7 – 13 (1987).
769 236. A. M. Frumar, D. R. Meldrum, F. Geola, I. M. Shamonki, I. V. Tataryn, L. J. Deftos, and H. L. Judd, Relationship of fasting urinary calcium to circulating estrogen and body weight in postmenopausal women. J. Clin. Endocrinol. Metab. 50, 70 – 75(1980). 237. C. Hassager and C. Christiansen, Influence of soft-tissue body composition on bone mass and metabolism. Bone 10, 415 – 419 (1989). 238. N. H. Bell, S. Epstein, A. Greene, J. Shary, M. J. Oexmann, and S. Shaw, Evidence for alteration of the vitamin D endocrine system in obese subjects. J. Clin. Invest. 76, 370 – 373(1985). 239. T. Anderson, P. McNair, N. Fogh-Andersen, T. T. Nielson, L. Hyldstrup, and I. Transbol, Increased parathyroid hormone as a consequence of changed complex binding of plasma calcium in morbid obesity. Metabolism 35, 147 – 151 (1986). 240. Y. Liel, E. Ulmer, J. Shary, B. W. Hollis, and N. Bell, Low circulating vitamin D in obesity. Calcif. Tissue Int. 43, 199 – 201 (1988). 241. S. L. Teitelbaum, J. D. Halverson, M. Bates, L. Wise, and J. G. Haddad, Abnormalities of circulating 25OH vitamin D after jejunal bypass for obesity: Evidence of an adaptive response. Ann. Intern. Med. 86, 289 – 293 (1977). 242. R. L. Atkinson, W. T. Dahms, G. A. Bray, and A. A. Schwartz, Parathyroid hormone levels in obesity: Effects of intestinal bypass surgery. Miner. Electrolyte Metab. 1, 315 – 320 (1978). 243. J. E. Compston, Vedis, J. E. Ledger, J. C. Gazet, and R. E. Pilkington, Vitamin D status and bone histomorphometry in gross obesity. Am. J. Clin. Nutr. 34, 2359 – 2363 (1981). 244. J. A. Nisker, G. L. Hammond, B. J. Davidson, A. M. Frumar, N. K. Takaki, H. L. Judd, and P. K. Siiteri, Serum sex-hormone binding globulin capacity and the percentage of free estradiol in postmenopausal women with and without endometrial carcinoma. Am. J. Obstet. Gynecol. 138, 637 – 641 (1980). 245. B. J. Davidson, J. C. Gambone, L. D. Lagasse, T. W. Castaldo, G. L. Hammond, P. K. Siiteri, and H. L. Judd, Free estradiol in postmenopausal women with and without endometrial cancer. J. Clin. Endocrinol. Metab. 52, 404 – 408 (1981). 246. R. Lindsay, D. M. Hart, and D. M. Clark, The minimum effective dose of estrogen for the prevention and treatment of postmenopausal bone loss. Obstet. Gynecol. 63, 759 – 763 (1984). 247. S. L. Greenspan, E. R. Myers, L. A. Maitland, N. M. Resnick, and W. C. Hayes, Fall severity and bone mineral density as risk factors for hip fracture in ambulatory elderly. JAMA 271, 128 – 133 (1994). 248. D. F. Williamson, Descriptive epidemiology of body weight and weight change in U.S. adults. Ann. Intern. Med. 119, 646 – 649 (1993).
771
CHAPTER 31 Effects of Tobacco and Alcohol Use on Bone
CHAPTER 31
Effects of Tobacco and Alcohol Use on Bone EGO SEEMAN
Austin and Repatriation Medical Centre, University of Melbourne, Heidelberg, Melbourne 3084, Australia
IV. Conclusions and Questions References
I. Introduction II. Tobacco and Bone III. Alcohol and Bone
I. INTRODUCTION
contradictory data are to be understood. Individuals rarely, if ever, differ in their exposure by only one factor. Confounding by other important covariates such as age, weight, exercise, nutritional factors, and drug therapy must be measured and taken into account in analyses (see Chapter 20). Thus, a difference in BMD between a smoker and nonsmoker may be due to differences in lifestyle factors other than tobacco consumption. Likewise, a true difference attributable to tobacco consumption may be obscured by differences in other factors unless these are taken into account in the analysis. Small sample sizes, biological variation, and difficulties in quantitation of the exposure “dose” of the risk factor reduce the power of studies to detect small but important effects. A true association between smoking and fracture rates is likely to be difficult to demonstrate in small studies because fractures are uncommon annual events in individuals. Fractures occur at a peak incidence of about 3 to 4 per 100 per year in women over 75 to 80 years of age. In younger groups, the incidence is 1 to 2 per 1000 or 10,000 per year depending on the decades studied. Thus, even if 20 – 30% of the risk for fracture was attributable to tobacco use, this risk may remain undetectable unless the samples sizes are very large or the prevalence of exposure was high. This problem is compounded further when end points are
To warrant study from a public health point of view, risk factors for osteoporosis should occur commonly and be amenable to intervention safely and at low cost. Modifying a risk factor that is uncommon and has a small effect on bone mineral density (BMD) is unlikely to influence fracture risk in the individual or the public health problem of fractures. Modifying a risk factor that is common and has a large effect (such as estrogen deficiency or falls) may help the individual and the public health problem. Modifying a risk factor that is uncommon but has a large effect (such as corticosteroid treatment) may help the individual, but not the public health problem. Modifying a risk factor that is common but has a small effect may not help the individual greatly, but may help the public health problem of osteoporosis. Tobacco and alcohol use fulfill several of these criteria. They are used by many individuals during a large proportion of their lives. The effect may be small in the short term but may become clinically important when exposure is prolonged. These are risk factors that can be modified at little, or no, expense. The difficulties in quantifying the effects of risk factors such as tobacco and alcohol should be acknowledged if
OSTEOPOROSIS, SECOND EDITION VOLUME 1
771
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
772
EGO SEEMAN
difficult to define such as a vertebral “fracture” or “deformity.” Similarly, hip fractures are the result of a fall as well as bone fragility so that identifying and measuring the risk conferred by tobacco use are difficult when multiple risk factors for falls and bone fragility are present in patients with hip fractures. Moreover, ascertainment bias may occur when patients with fractures are interviewed after the fracture. Circulating levels of some hormones exhibit diurnal variation or pulsatility. Estrogen concentrations vary from day to day by 50 to 200%, raising questions about the interpretation of negative studies and the usefulness of single measurements. If exposure to tobacco reduces estrogen concentration by 20 to 30%, this will be difficult to detect with single measurements taken at different times of day in different people. Rates of bone loss are around 1 to 2% per year, similar to the coefficient of variation of the methods used to measure BMD. In other words, the effect of exposure may be important but small in relation to the biological variation of the measurement and the precision of the technique. Thus, differences in BMD between a smoker and a nonsmoker or between a consumer of alcohol or an abstainer may be due to differences in concomitant lifestyle factors rather than any direct effect of tobacco or alcohol on bone. These covariates must be measured and their effects taken into account before an association is identified or concluded not to exist. Few studies pay meticulous attention to these methodological issues. Even when they are taken into account, the studies document associations, but causality can only be inferred from the observed association.
II. TOBACCO AND BONE A. The Problem Tobacco use is the single greatest preventable cause of premature death in the United States. Nearly 1 in 4 women age 18 years and over in the United States was a smoker in 1991. Thus, about 22 million women in the United States currently smoke cigarettes. Ninety percent of smokers begin to do so before the age of 20 years. Twenty-seven percent of women with 12 or fewer years of schooling were smokers compared to 12.5% with 16 or more years of schooling. Among girls aged 17 to 18 years, 33% who dropped out of school were current smokers compared to 17% who were still in high school or had graduated. Among adolescent girls aged 12 to 17 years, the reported smoking in the past month increased from 8.7% in 1990 to 9.4% in 1992. Smoking-related disease accounted for about 150,000 deaths among U.S. women in 1988. Lung cancer now surpasses breast cancer as a leading cause of cancer death
among women, accounting for 22% of female cancer deaths compared to 18% for breast cancer. Other tobaccorelated diseases include cancers of the oral cavity, esophagus, larynx, bladder, and pancreas, heart disease, stroke, emphysema, and bronchitis. Women smokers are at increased risk for cervical cancer, early menopause, complications of the oral contraceptive pill, and unfavorable pregnancy outcomes. There are 3000 new users daily in the United States; many of these are teenagers [1 – 3].
B. Fractures 1. HIP FRACTURES In most, but not all, studies, tobacco use is associated with an increased risk for hip fractures in women and in men. In general, the relative risks (RR) for hip fracture associated with tobacco use are about 1.2 to 1.5 with confidence intervals (CI) that include, or almost include, unity. The serious methodological problems associated with studying hip fracture pathogenesis should be recognized. In many studies, the sample sizes are small, around 100 cases or fewer. Survivors are interviewed whereas subjects with dementia and other serious comorbidity are omitted. This is a concern because a substantial proportion of the hip fracture cases will not be assessed. Data obtained cannot be verified and may be inaccurate given that the patients are elderly. In virtually all studies, exposure is assessed after the fracture has occurred. This may introduce ascertainment bias. In many studies, smoking was assessed dichotomously (ever versus never), without adjustment for age or of other potential confounding factors. Baker [4] found no association between smoking and hip fractures among 189 cases and 95 community controls. Paganini-Hill and colleagues [5] reported a RR of 1.5 (95% CI 0.8 to 3) among 91 cases less than 80 years old and 182 community controls. Williams et al. [6] reported a RR of 2.2 (95% CI, 0.6, 8.2) in 160 cases and 567 community controls. Kreiger and colleagues [7] reported a RR of 1.3 (95% CI, 0.8, 2.1) in 98 cases and 884 hospital controls. Boyce [8] reported a RR of 1.2 among 139 cases and 139 controls [8], whereas Alderman et al. [9] reported a RR of 1.2 (95% CI, 0.9, 1.6) in 160 cases aged 50 to 74 years and 567 controls. Cooper and Wickham [10] found that the RR for hip fracture in women was 1.7 (95% CI, 1.2, 2.4). This remained significant after adjusting for alcohol, body mass, activity, and calcium intake. The study was based on 240 cases (100 “ever,” 140 “never” smokers) compared with 480 controls. Among 60 men with hip fractures (45 “ever,” 15 “never” smokers, and 120 controls) the RR was 1.5 (95% CI, 0.7, 3.1). After adjusting for alcohol intake the RR was 0.7 (95% CI 0.3, 1.8). Johnell and colleagues [11] studied risk factors for hip fracture in 2086 women (mean age 78 years) with hip
773
CHAPTER 31 Effects of Tobacco and Alcohol Use on Bone
fracture and 3532 controls from six countries. No detectable effect of tobacco use was identified. Alcohol consumption was protective in young adulthood. On multivariate analysis, late menarche, poor mental score, low body mass index (BMI) and physical activity, low exposure to sunlight, and low consumption of calcium and tea were independent risk factors, accounting for 70% of hip fractures. Cornuz and colleagues [12] followed 116,229 women aged 34 to 59 years for 12 years. At baseline, 31% were current smokers and 26% were current smokers and 26% former smokers; after 12 years, another 11% were former smokers. During follow-up there were 377 hip fractures, with an average age of 60 years. Risk increased linearly with cigarette consumption (RR 1.6; CI 1.1, 2.3). The age adjusted RR of hip fracture in current versus never smokers was 1.3 (95% CI 1.0, 1.7) and remained elevated, but not significantly so, after adjusting for menopausal status, use of estrogen, activity, and intakes of calcium, alcohol caffeine, and BMI. Melhus and colleagues [13] reported that an insufficient dietary intake of vitamins E and C may increase the risk of hip fracture in smokers. In 205 women aged 40 to 76 years who sustained a hip fracture (44 current and 161 never smokers) and in 746 controls (93 current and 653 never smokers), after adjustment for risk factors, the odds ratio (OR) for hip fracture among current smokers with a low vitamin E intake (below the median in controls) was 3.0 (95% CI 1.6, 5.4) and with a high vitamin E intake was 1.1 (CI 0.5, 2.4). Equivalent ORs for low and high vitamin C intakes were 3.0 (CI 1.6, 5.6) and 1.4 (CI 0.7, 3.0). The OR in current smokers with low intakes of both vitamin E and C was 4.9 (CI 2.2, 11.0). The influence of vitamin E and C intake was less pronounced in former than in current smokers. Thus, data support an association between tobacco use and hip fractures but confidence intervals often cross unity. There may be an interaction involving tobacco use, body weight, estrogen exposure, and fracture risk. Tobacco use may negate the protective effect of estrogen replacement therapy against hip fracture. Kiel et al. [14] observed 207 hip fractures among 34,700 women. Smoking was not associated with an increased risk of hip fracture (RR 1.22, 95% CI 0.76, 1.95). Data suggest that estrogen use was protective in nonsmokers (RR 0.37, CI 0.19, 0.75) but not in current smokers (RR 1.26, CI 0.29, 5.45). Forsen and colleagues [15] found that 421 hip fractures occurred during 1986 – 1989 among 34,856 adults over 50 years of age who had attended health screening between 1984 and 1986 (91% of the eligible population). For women who smoked, the RR of hip fracture was 1.5 (95% CI, 1.0, 2.4), 3.0 (95% CI, 1.8, 5.0) in thinner women (BMI 20kg/m2) and 1.8 (95% CI, 1.2, 2.9) in men (independent of body mass). Lack of physical activity increased the RR of hip fracture to 1.4 in women and 2.3 in men after adjustment for ill health.
Williams and colleagues [6] studied 353 women aged 50 – 74 with hip or forearm fractures and 567 communitybased controls. The RR of hip fracture was increased in thin women who smoked, particularly among nonusers of estrogen. The RR of hip fracture in women who were thin, never smoked, and did not use estrogens was 13.5 (95% CI, 2, 35.5) and 6.4 (95% CI, 2.1, 19.4) in those who had used estrogens for more than 1 year. Women who were obese were not at increased risk for hip fracture whether they smoked or not. The benefit of using estrogens in preventing hip fractures was greatest in the thin smoker. In contrast, Hemenway et al. [16] studied 96,508 middle-aged nurses 35 – 59 years of age. Hip or forearm fractures occurred in 925 of the women. No association with the risk for hip fracture and smoking was detected, perhaps because of the relatively young age of this cohort. Kreiger and colleagues [7] reported that of 98 hip fractures and 884 hospital controls, tobacco use was associated with an age adjusted RR of 1.27 (95% CI, 0.63, 2.56), an estimate that included unity, perhaps because of the small sample size. Law and colleagues [17] conducted a meta-analysis of 29 published cross-sectional studies including 2156 smokers and 9705 nonsmokers and 19 cohort and case control studies recording 3889 hip fractures. Risk of hip fracture was 17% greater at 60 years, 41% greater at 70 years, 71% greater at 80 years, and 108% greater at 90 years. An estimated 19% of current smokers and 12% nonsmokers would have a hip fracture by 85 years of age, and 37% of current smokers and 22% of nonsmokers by the age of 90 years. One hip fracture in eight was attributable to smoking in women (Fig. 1). 2. SPINE, FOREARM, AND OTHER FRACTURES Daniell [18] reported that 76% of 38 women with one or more vertebral fractures smoked at least 10 cigarettes per day for 5 or more years compared with 43% of 572 controls. Aloia et al. [19] reported that 58 postmenopausal women with crush fractures smoked twice as much as 58 age-matched controls. Seeman and colleagues [20] reported an odds ratio of 2.3 in men who were current smokers. Smoking, body weight, alcohol, and underlying comorbidity were independent predictors of vertebral fracture risk. The risk associated with smoking increased by 1.009 for each pack year of exposure. Seventy-nine percent of the cases and 63% of controls smoked (P 0.009), whereas 82% of cases and 70% of controls drank alcohol (P 0.02). The risk of osteoporosis increased by 1.007 for each ounce year of cumulative exposure (P 0.01). The relationship among these factors is shown in Fig. 2. In those with no underlying disease, the relative risk was 0.3 (Fig. 2) (hatched bar on the left), reaching 30 in an individual who drank alcohol, smoked, was not obese, and had an underlying illness.
774
EGO SEEMAN
FIGURE 1 Relative risk (95% confidence intervals) of hip fracture in smokers compared to nonsmokers in postmenopausal women according to age in cohort studies and case-control studies (open circles). Reproduced from Law and Hackshaw [17], with permission from the authors and publishers. The risk conferred by exposure to tobacco, alcohol, and underlying disease and the protective effect of obesity was examined in a multiple logistic model. The extent of the elevation in risk was age dependent. The effects of smoking and alcohol were not evident in persons under 60 years old, emerged in 60- to 69-year olds, and were strongest in those 70 years and over. In nonobese individuals over 70 years of age who drank alcohol, smoked, but had no dis-
FIGURE 2
ease, the relative risk was 20.2 (P 0.05). This risk was reduced to 6.9 in the presence of obesity. In a nonobese individual who drank, smoked, and had an underlying illness, the relative risk for vertebral fractures was 192.5 (P 0.05). In the study by Williams et al. [6] of women aged 50 – 74 with forearm fractures and 567 community-based controls, nonestrogen users had a RR for forearm fractures
Relative risk of spinal osteoporosis with vertebral fractures in men 60 to 69 years old with various combinations of tobacco use, alcohol use, obesity, and presence of underlying medical illness. From Seeman and colleagues [20]. Reproduced with permission from the authors and publisher.
CHAPTER 31 Effects of Tobacco and Alcohol Use on Bone
of 5.4 (95% CI, 2.5, 11.3) if they were thin and ever smoked and 0.7 (95% CI, 0.2, 2.5) if they had used estrogens for more than 1 year. In estrogen users, there was no relationship between the risk of forearm fractures and a history of smoking and body weight. The benefit of using estrogens in preventing forearm fractures was greatest in the thin smoker. In the study by Hemenway et al. [16], no association between risk for forearm fracture and smoking was detected. Jensen [21] studied 285 women 70 years of age, of whom 77 were smokers and 103 were nonsmokers. A history of fractures was obtained in 40.3% of the smokers and 44.7% of the nonsmokers. The authors concluded that osteoporosis was of the slender, rather than the slender smoker, and that smokers have a reduced BMD appropriate for their slenderness. Thus, most studies suggest that tobacco use increases the risk for fracture, particularly in thinner individuals. This risk appears to emerge in advanced age and may reduce the protective effect of obesity and estrogen exposure for fracture, whereas higher body weight may protect from the effects of tobacco.
C. Bone Mineral Density The increased risk for fractures associated with tobacco use is likely to be partly conferred by a reduction in BMD. This effect may even be conferred before birth. Jones and colleagues [22] studied maternal smoking habits related to bone mass and growth in 330 children aged 8 years. For children born at term, smoking during pregnancy was associated with a lower height (1.53 cm; 95% CI 3.03, 0.03) and a trend to lower weight (1.35 kg; CI 2.75, 0.11). After adjustment for body size, BMD was lower at the lumbar spine (0.019 g/cm2; CI 0.033, 0.005) and femoral neck (0.018 g/cm2; CI 0.034, 0.002) but not total body (0.005 g/cm2; CI 0.015, 0.005) in the maternal smoking group. Placental weight was lower in smoking mothers (56 g; CI 95, 17), and adjustment for placental weight removed the effects of smoking on BMD and growth. In children born prior to 37 weeks, maternal smoking was not associated with deficits in BMD. The influence of smoking was not altered by adjustment for sports activity, dietary calcium intake, or sunlight exposure. Children who were breast-fed and had nonsmoking mothers had higher BMD at all sites (3.3 to 7.7%) than children who were not breast-fed and whose mothers smoked. The authors concluded that maternal smoking during pregnancy is associated with growth and bone mass in children at age 8 years. These associations may be mediated through placental size and function. Thus, the skeleton of offspring of smokers may be affected in early development.
775 Because tobacco users often start smoking by 10 to 12 years of age, the reduced peak BMD may contribute to any deficit in BMD in adulthood. Even though over 90% of peak BMD is achieved by about 15 to 18 years of age, exposure to tobacco may have important effects, as mineral accrual is very rapid, particularly in the peripubertal period, and if interrupted may result in biologically important deficits in BMD. No studies have been done to examine this possibility. Studies in children will be difficult to perform and interpret because information regarding exposure from children may be unreliable. In addition, socioeconomic factors must be taken into consideration when these studies are designed, as children from lower socioeconomic groups are more likely to take up smoking and may have lower BMD at the time that tobacco exposure begins. In addition, reduced bone size may contribute to any deficit in BMD because this measurement of “density” only partly takes bone size into account. There are no data available addressing the question of the effects of tobacco on stature and bone size in children or adults. Few prospective studies have examined the association between tobacco use and rates of bone loss. Krall and Dawson-Hughes [23] found that bone loss (percent per year) from the radius was greater in 34 smokers than in 278 nonsmokers ( 0.91 2.6 vs 0.004 2.57, P 0.05). Similar trends were found at the femoral neck (0.58 2.9 vs 0.16 2.84), oscalcis (1.42 3.32 vs 1.19 3.26), and spine (1.31 2.19 vs 0.97 2.16). More recently, Krall et al. [24] measured rates of change in BMD over 3 years in 402 men and women aged 65 years or above. One half of participants received daily calcium (500 mg elemental calcium) and cholecalciferol (700 IU) supplements and the remainder received placebo. The annualized rate of change in BMD was higher in the 32 smokers than in the nonsmokers at the femoral neck (0.71 0.29 vs 0.04 0.08%, P 0.02) and total body (0.36 0.10 vs 0.15 0.03%, P 0.05). Mean calcium absorption was lower in smokers after adjusting for gender, age, supplementation status, and dietary calcium and vitamin D intakes (12.9 0.8 vs 14.6 0.2%, P 0.05). Lowest absorption was in smokers of at least 20 cigarettes/day (12.1 1.1%). In the subgroup receiving calcium and cholecalciferol, urinary calcium/creatinine excretion was lower in smokers than in nonsmokers (44 12 vs 79 9%, P 0.05). In contrast, Slemenda et al. studied 84 peri- and postmenopausal women and found no association between rates of bone loss and tobacco consumption. However, there were 13 subjects who smoked over 20 pack years and 8 who smoked under 20 pack years. Increased rates of bone loss, determined by changes in metacarpal morphometry, have been reported in men [26]. Most studies have been cross-sectional and of these, most, but not all, report an association between tobacco exposure and reduced BMD. In general, no deficits in BMD
776
EGO SEEMAN
FIGURE 3 Differences (95% confidence intervals) as a proportion of 1 SD in BMD between female smokers and nonsmokers according to age and menopausal status. Open circles refer to two studies and solid circles refer to other studies cited. From Law and Hackshaw [17], reproduced with permission from the authors and publisher. are observed in smokers compared to premenopausal and perimenopausal “never” smokers [27 – 35]. As shown in Fig. 3, in the study by Law and Hackshaw [17], premenopausal smokers and nonsmokers had similar BMD and postmenopausal bone loss was greater in current smokers, diminishing 2% more for every 10 years increase in age with a difference of 6% at 80 years of age. By the age of 80 years, BMD was 0.45 SD lower in smokers than in nonsmokers. The estimate in men was 0.32 SD lower BMD in smokers relative to nonsmokers. The association was independent of body weight. Hansen [36] studied 249 healthy premenopausal women 39 6 years of age and found no association between BMD and tobacco use. Stevenson and colleagues [37] reported an association between BMD and tobacco use at the spine but not at the proximal femur, where McCulloch et al. reported lower BMD in heavy smokers. McDermott and Witte [39] reported no difference in the mid- and distal radius BMD in 35 smokers and 35 nonsmokers. Grainge et al. reported a negative correlation between BMD and months of smoking in 580 postmenopausal women aged 45 to 59 years. However, smoking duration accounted for only 1% of variance in BMD. May and colleagues [41] reported no association between past or present smoking and BMD among 453 men ages 65 to 76 years, after adjusting for age and weight. Previous analyses of women from the same population showed a dose – response relation between smoking and BMD. In a study of 186 women and 224 men aged 61 to 73 years, Egger et al. reported current male smokers had 7.3% lower lumbar spinal BMD than never smokers (95% CI 0.4, 14.2). For women, BMD was 7.7%
lower (CI 0.3, 15.6). Each decade of smoking reduced lumbar spinal BMD by 0.015 g/cm2. Smaller effects of smoking on BMD were observed at the femoral neck. Thus, deficits of about 0.5 to 1.0 SD are associated with prolonged exposure in postmenopausal smokers and in men. This is in accord with finding an increased relative risk for fractures in advanced age. Jones and Scott [43] compared 118 current smokers and 158 nonsmokers (mean ages 33 and 34 years, respectively). Smokers had lower BMD (femoral neck, 0.32 SD; 95% CI 0.60, 0.04; lumbar spine, 0.49 SD; CI 0.76, 0.22; and total body, 0.40 SD; CI 0.66, 0.14) among women with BMI 25 kg/m2, in whom there was a dose – response relation between cigarettes smoked and BMD. Relative to current smokers, women who ceased smoking between 1988 and 1996 (n 27) had higher BMI (by 2.5 kg/m2), body weight (7.2 kg), and lumbar BMD (0.52 SD). Smokers who had breast-fed at least one child had an additional deficit in BMD (femoral neck, 0.48 SD; CI 0.89, 0.07; lumbar spine, 0.39 SD; CI 0.80, 0.02; and total body, 0.37 SD; CI 0.77, 0.06). Smokers who participated in competitive sport had increments in bone mass (0.74 SD; CI 0.31, 1.17; 0.48 SD; CI 0.03, 0.93; and 0.42 SD; CI 0.00, 0.84, at respective sites). The cotwin control model overcomes several of the difficulties discussed in Section I by controlling for age, sex, and genetic composition, which are major determinants of BMD. To examine the effects of cigarette smoking on BMD, Hopper and Seeman [44] studied 41 female twin pairs (21 monozygotic), ages 27 to 73 years (mean 49), discordant for at least 5 pack years of smoking (mean 23,
777
CHAPTER 31 Effects of Tobacco and Alcohol Use on Bone
FIGURE 4
Bone density (g/cm2) at the lumbar spine in the greater smoker and lesser smoker as a function of age. From E. Seeman, unpublished data.
maximum 64). Each twin was classified as either the “greater” or “lesser” smoker of a pair. The mean difference in smoking was 23 – 15 pack years (range 5 to 64). The utility of this model is illustrated in Fig. 4; BMD in the greater smoker (closed symbol) and lesser smoker (open symbol) diminish as age advances, but the difference in BMD between them is not readily apparent because genetic and other environmental factors also influence an individual’s BMD. As shown in Fig. 5, when the within-pair difference in BMD between the greater and the lesser smoking twin is plotted as a function of increasing discordancy in pack years smoking, the effect of smoking is apparent. As the difference in pack years between greater and lesser smoker increases, the BMD deficit increases so that after about 20 pack years of discordancy, few data points are above the x axis. The BMD deficit in the smoker is about 5 to 10%, or over a half a standard deviation, with over 20 pack years of difference in tobacco exposure. For the 20 most discordant pairs, BMD was lower in the greater smoking twin at the lumbar spine by 9.3 to 3.1% (P 0.008), at the femoral neck by 5.8 to 2.9% (P 0.06), and at the femoral shaft by 6.5 to 3.2% (P 0.05). Across all 41 twin pairs, for each 10 pack years of smoking, the deficit in BMD increased by 2% at the lumbar spine and by about 1% at the femoral sites. Women who smoke one pack of cigarettes each day through adult life will, by menopause, have an average deficit in BMD of 0.5 to 0.8 of a population standard deviation or a 5 to 8% deficit, an amount sufficient to increase their risk for fracture. In vitro, bone strength decreases threefold with a 10% decrease in mineral content. A 10% difference in BMD is equivalent to almost one population standard deviation, 10 years of age-related bone loss, and about one-half the diminution found in postmenopausal women relative to premenopausal women. A decrease of
FIGURE 5
Within-pair difference (greater minus less smoker) in bone density expressed as a percentage of the pair mean and plotted as a function of pack years discordancy at the lumbar spine, femoral neck, and femoral shaft. Monozygotic pairs are represented by a closed symbol and dizygotic pairs by an open symbol. From J. L. Hopper and E. Seeman [44]. Reproduced with permission from the authors and publisher.
10% over 10 years may confer a 44% increase in hip fracture risk. The greater smoker consumed more coffee (P 0.02), and among the 20 most discordant pairs, the twin who smoked more walked less (P 0.004). Of the 16 postmenopausal pairs, menopausal age was 1.5 – 1.2 years earlier in the greater smoker (P 0.20). Allowing for life style factors resulted in only minor changes in the strengths of the association between BMD and smoking. At the lumbar spine, the difference in BMD per 10 pack years decreased from 2.0 to 1.7% after adjusting for a positive association between that difference in BMD and the difference on the walking scale (P 0.03). In the 20 most discordant
778
EGO SEEMAN
pairs, after adjusting for discordancy in estrogen replacement therapy, the difference in BMD increased from 9.3 to 10.4% (P 0.004). Further multivariate analyses that allowed for differences in other measured life style factors, including alcohol and coffee consumption, did not influence the strength of association with smoking. The relationships between deficits in BMD at all sites and difference in pack years smoking were not eliminated by allowing for biochemical factors. Syversen and associates [45] reported that female rats aged 2 months exposed to nicotine vapor ( 100 ng/ml) for 20 days, 5 days/week for 2 years had 10% lower body weight than controls throughout the study but no differences in femoral BMC or BMD; femoral length; ultimate bending moment, ultimate energy absorption, stiffness, or deflection in the femoral shaft and femoral neck; or midshaft cortical area and stress.
D. Bone Loss: Increased Bone Resorption Bone loss occurs if there is an imbalance between the amount of bone resorbed and the amount of bone formed. The evidence available examining whether one or both of these mechanisms contributes to the bone loss associated with smoking is limited. De Vernejoul and colleagues [46] reported histomorphometric changes in 11 men ages 35 – 50 years with idiopathic osteoporosis who were mild alcoholics and heavy smokers. In patients vs controls, trabecular bone volume was 15.5 3.4% vs 24.3 5.4% (mean SD). Mean wall thickness was 50.6 6.7 mm vs 59.9 5.6 mm, and mean trabecular plate thickness was 159 27.4 mm vs 193 37.2 mm. There was no evidence of osteomalacia. Bone resorption was not increased relative to controls. These data are consistent with reduced bone formation accounting for the deficit in bone volume. Hansen [36] studied 249 healthy premenopausal women 39 6 years of age and found that serum osteocalcin was reduced in smokers compared to nonsmokers (P 0.01). Leino et al. studied 519 women and found with multivariate analysis that serum osteocalcin was associated with alcohol consumption but not tobacco use; alkaline phosphatase and total ionized calcium were associated with cigarette smoking. In the cotwin control study [44], while no evidence for reduced bone formation was detected, there was evidence for increased bone resorption. Serum calcium concentrations were higher in the greater smoker in 17 of the 26 pairs (P 0.08). For the 11 most discordant pairs, serum phosphate and serum alkaline phosphatase activity were higher in the greater smoker (P 0.05, P 0.006, respectively). Pair differences in serum calcium were positively associated with pair differences in urinary hydroxyproline excre-
tion (r 0.8, P 0.001) and urine pyridinoline to creatinine ratio (measured in 17 pairs) (r 0.8, P 0.001). At the lumbar spine, the lower BMD in the greater smoker was associated with higher serum calcium and urine pyridinoline values, consistent with increased bone resorption. As the within-pair smoking difference increased, the differences in serum parathyroid hormone concentrations decreased by 5.0 2.5% for each 10 pack years (P 0.05). Pair difference in serum parathyroid hormone was negatively associated with pair difference in BMD at the lumbar spine (r 0.60; P 0.005), femoral neck (r 0.75; P 0.001), and femoral shaft (r 0.53; P 0.01). Allowing for pack years of smoking and serum parathyroid hormone, which together accounted for 70% of the variation in pair differences in BMD at the lumbar spine, the deficit in spinal BMD increased with pair difference in serum calcium (P 0.05) and, in the 17 pairs measured, with the pair difference in urine pyridinoline (P 0.06). Smoking was associated with higher serum folliclestimulating hormone (P 0.02) and luteinizing hormone (P 0.03) concentrations and lower serum parathyroid hormone concentrations (P 0.05). Several of these relationships are shown in Fig. 6.
E. Bone Loss: Decreased Bone Formation Reduced bone formation may, in part, be due to a reduction in osteoblast numbers. Fang et al. [48] colleagues examined the effects of nicotine on UMR (University, Melbourne, Repatriation) 106-01 rat osteosarcoma cell proliferation and activity. Nicotine produced a dose-dependent suppression of thymidine incorporation with maximum suppression seen at 10 mM. One micromolar nicotine decreased cell number by 20 – 30%. There was a dose-dependent increase in alkaline phosphatase activity with a maximum stimulation of 189% relative to control at 1mM, suggesting that nicotine may promote the differentiation of osteoblasts. In contrast, Lenz and colleagues [49] reported that nicotine stimulated DNA synthesis in osteoblast-like cells from embryonic chick calvaria and inhibited collagen synthesis and alkaline phosphatase activity. Moreover, Galvin et al. [50] reported that smokeless tobacco extract did not affect thymidine incorporation and inhibited collagen synthesis and alkaline phosphatase activity in chick embryo calvarial cells. These disparate observations may be related to species, cell culture conditions, or the type of osteoblast model used.
F. Effects of Tobacco on Estrogen Metabolism The increased bone resorption is, in part, likely to be mediated by a reduction in circulating estrogen in women. This reduction in circulating estrogen is probably due to
CHAPTER 31 Effects of Tobacco and Alcohol Use on Bone
FIGURE 6
Within-pair difference (greater minus lesser smoker) in serum-luteinizing hormone, follicle-stimulating hormone, parathyroid hormone, and calcium expressed as a percentage of the pair mean and plotted as a function of pack years discordancy. Monozygotic pairs are represented by a closed symbol and dizygotic pairs by an open symbol. From J. L. Hopper and E. Seeman [44]. Reproduced with permission from the authors and publisher.
decreased production and increased degradation of circulating estrogen. The inferences are based mainly on crosssectional studies. Postmenopausal smokers have lower estrogen concentration than nonsmokers. Smoking is associated with an earlier menopause. Pregnant women who smoke have lower estrogen levels than nonsmokers. McMahon and col-
779 leagues [51] showed that premenopausal women who smoke have lower luteal phase excretion of estrone, estradiol, and estriol than nonsmokers. There were no differences detected in the follicular phase in premenopausal women or in any measures in postmenopausal women. Barbieri and colleagues [52] examined the effects of constituents of tobacco on estrogen production by human choriocarcinoma cells and placental microsomes. Nicotine, cotinine, and anabasine inhibited the conversion of androstenedione to estrogen in a dose-dependent fashion. The inhibition of aromatase was reversible and competitive as supraphysiological doses of androstenedione block the inhibition of aromatase by these tobacco products. Figure 7 shows the effects of nicotine, cotinine, and anabasine on estrogen and progesterone accumulation in choriocarcinoma cell cultures. The inhibition of estrogen production was specific. There was no change in progesterone accumulation. Figure 8 shows that nicotine inhibition of aromatase activity in human placental microsomes was competitive. The mechanism probably reflects inhibition of the aromatase enzyme system. It is likely that nicotine interacts directly with placental microsomal cytochrome P450 and may reversibly alter the function of active sites of the cytochrome P450 component of the aromatase enzyme system. Aminoglutethamide inhibits that aromatization of testosterone and shares structural similarities to nicotine. Increased degradation of endogenous and exogenous estrogens may contribute to the lower levels of circulating estrogen. Estradiol is reversibly oxidized to estrone, which is then irreversibly hydroxylated in the C-2 position to form 2-hydroxyestrone and 2-methoxestrone or is hydroxylated in the 16 position to form 16-hydroxyestrone and -estriol. Michnovicz and colleagues [53,54] studied 14 premenopausal women who smoked 15 to 30 cigarettes per day and 13 premenopausal nonsmokers aged 21 – 44 years. Increased 2-hydroxylation of estrogens was observed. Total estrogen excretion did not differ in smokers and nonsmokers; however, the metabolite 2-hydroxyestrone constituted 53.6 2.2% vs 35.1 1.8% of the total estrogen excretion (P 0.02). As shown in Fig. 9, the extent of 2-hydroxylation was increased. The metabolite 2-hydroxyestrone was elevated in smokers (17.2 2.4 mg/g vs 9.4 1.2 mg/g creatinine, P 0.02) and there was a reduction in urinary estriol in smokers (11.9 1.2 mg/g vs 35.2 10.7 mg/g creatinine). Jensen and colleagues [55] studied 136 postmenopausal women. Of these, 63 received 4 mg estradiol, 42 received 2 mg, 23 received 1 mg, and 23 received placebo. Serum estradiol did not differ at baseline. Figure 10 shows that both circulating estrone and estradiol were lower in smokers than in nonsmokers, being half that of nonsmokers in the high dose group. In the 114 women receiving hormone therapy, there was an inverse relationship between the number of cigarettes smoked and the change in estrone (r 0.34, P 0.001)
780
EGO SEEMAN
FIGURE 7
Nicotine, cotinine, and anabasine reduced estrogen accumulation but not progesterone accumulation in choriocarcinoma cell cultures. From Barbieri and colleagues [52]. Reproduced with permission from the authors and publisher.
and estradiol (r 0.31, P 0.001). Body weight was 59.2 kg (range 39.5 to 86.5) in smokers and 67.3 kg (range 46 to 110) in nonsmokers. Whether this may have confounded the observations is uncertain. Data are consistent with increased metabolism of the administered estrogen. Cassidenti and colleagues [56] studied the baseline estrogen status of 13 postmenopausal smokers and 12 nonsmokers. There was no difference in baseline measurements of estrone, estradiol, estrone glucuronide, or sulfate. The unbound estradiol concentration was lower in smokers than in nonsmokers 8 h after the administration of micronized estradiol (13.3 1.4 pg/ml vs 25.9 4.8 pg/ml, P 0.02) with the higher dose (2 mg) of micronized estradiol. The same pattern was present with lower doses in the smokers at 4 and 8 h after administration of estrogen. Sex hormone-binding globulin was higher in smokers than in nonsmokers with the 1-mg (136.7 8.2 nmol/liter vs 61.6 8.3 mol/liter, P 0.0001). Serum estrone sulfate levels were higher smokers receiving the larger dose of estrogen. Data are compatible with tobacco increasing the hepatic synthesis of sex hormone binding globulin. Not all investigators document an association between tobacco consumption and estrogen. Cauley and colleagues studied 176 white postmenopausal women mean age 58 years. These investigators found the degree of obesity to be a major determinant of estrone and estradiol. The estrone was 40% higher than in nonobese, Smokers had higher androstenedione than nonsmokers estrone (82.1. 52.3 vs 65.9 42.1, mean SD, P 0.11), but there was little
FIGURE 8
Nicotine inhibition of aromatase activity in human placental microsomes using testosterone substrate concentrations of 4.9, 7.4 and 12.5. From Barbieri and colleagues [52]. Reproduced with permission from the authors and publisher.
CHAPTER 31 Effects of Tobacco and Alcohol Use on Bone
781
FIGURE 9
The extent of estradiol 2-hydroxylation as determined by a radiometric assay in 14 premenopausal female smokers and 13 nonsmoking controls. Bars indicate means ISEM. From Michnovicz and colleagues [54]. Reproduced with permission from the authors and publisher.
difference in serum estrogens between the 26 smokers and 150 nonsmokers. The small sample size may have limited the power of the study to detect a difference. Crawford et al., found 33% lower ethinyl estradiol levels in smokers versus nonsmokers 16 – 19 h after ingestion of that drug. This difference did not reach statistical significance. Geilser and colleagues, reported that plasma estrone, estradiol, and estrone sulphate were 40 – 70% lower in smokers when estrogen replacement therapy (ERT) was given orally but not transdermaly. Most results are consistent with the notion that tobacco use reduces circulating estrogens, leading to increased serum concentrations of follicle-stimulating hormone and luteinizing hormone and to increased bone resorption. The latter results in increased circulating calcium and a subsequent lowering of serum parathyroid hormone concentration and in increased urine hydroxyproline and prydinoline excretion. There is independent support for an effect of smoking on the production and degradation of estrogens.
G. Effects of Tobacco on Testosterone and Other Steroids Testosterone has been reported to be elevated, unchanged, or decreased in men who smoke. For example, in a population-based study of 590 men ages 30 to 79
FIGURE 10 Serum concentrations of estrone (E1) and estradiol (E2) in smokers and nonsmokers after treatment with low, medium, and high doses of estradiol. Values are means ISEM. From Jensen and colleagues [55]. Reproduced with permission from the authors and publisher.
years, serum testosterone was modestly increased in current smokers. Current smokers also had higher endogenous androstenedione, estrone, and estradiol levels than nonsmokers [60,61]. The increased estrogens were explained by neither adiposity nor alcohol consumption. In contrast, Handelsman et al. [62] studied 71 nonsmokers and 23 smokers ages 16 to 47 years and found no difference in plasma testosterone, follicle-stimulating hormone (FSH), leuteinizing hormone (LH), or prolactin concentration. Total sperm output, sperm motility, total motile sperm, and oval sperm density were reduced relative to nonsmokers. Briggs [63] reported changes in plasma testosterone after 7 days of abstinence in men smoking 30 cigarettes per day. Smokers showed a rise of 1.65 0.5 ng/ml after 7 days of abstinence.There was no change in plasma testosterone in controls. Briggs suggested that carbon monoxide may inhibit testosterone formation due to the blockade of 17 hydroxylation of progesterone. The enzyme catalyzing
782 this reaction, NADP: pregnane-17-oxo-reductase, located in the microsomal fraction, requires a cofactor related to cytochrome P450. Carbon monoxide inhibits microsomal cytochrome P450. Meikle and colleagues [64] studied the effects of nicotine and cotinine on testosterone metabolism. Both competitively inhibit 3-hydroxysteroid dehydrogenase, an enzyme that converts dihydrotestosterone to 3-androstanediol in dog prostate microsomes. Microsomal fractions incubated for 1 h with nicotine and cotinine resulted in elevated dishydrotestosterone and suppressed 3-androstanediol. (The 5reductase activity is unaffected by these metabolites of tobacco.) The concentrations used in these experiments were similar to those achieved in humans who smoke. Elevated testosterone levels have also been reported in women smokers. Friedman and colleagues [35] studied 9 postmenopausal smokers and 16 nonsmokers and found elevated testosterone (360 111.7 pg/ml versus 244.9 128.9 pg/ml, mean SD, P 0.05) and nonsignificantly higher dihydrotestosterone (135.6 81.9 pg/ml versus 88.4 41.2 pg/ml). Progesterone levels were elevated (107.8 59.9 pg/ml versus 65.1 34.8 pg/ml, P 0.05) as were those of 17-hydroxyprogesterone (708.4 418.7 pg/ml versus 197.8 74.1 pg/ml, P 0.0005) and androstenedione (844.8 503.6 pg/ml versus 486.8 287.8 pg/ml, P 0.05). The differences in dehydroepiandrostenedione (DHEA) sulfate failed to reach statistical significance (1622.2 1503.02 pg/ml versus 864.6 428.4pg/ml). Serum cortisol was higher (169.2 35.2 ng/ml versus 115.3 27.6 ng/ml. P 0.0001). In contrast, Cauley et al. [57] found no relationship between testosterone and cigarette smoking. Smokers had higher androstenedione levels (82.1 52.3 versus 65.9 42.1, P 0.11) (mean SD) than nonsmokers. This was independent of obesity, alcohol consumption, and other confounder. Higher circulating progesterone may protect from osteoporosis whereas hypercortisolism may increase the risk for osteoporosis in women who smoke. Adrenal cortisol and androgens increase due to the stimulation of ACTH release. Nicotine inhibits 11-hydroxylase and 21hydroxylase, causing the increase in DHEA and androstenedione. Higher androstenedione may also be the result of aromatase inhibition. Inhibition of desmolase also reduces estrogen and progestational hormones. 1. CESSATION OF SMOKING There are scant data concerning the effects of cessations of tobacco use on the skeleton. In the study by Cornuz and colleagues [12], the risk of fracture in former smokers did not decrease until 10 years after quitting: at 5 years, RR 1.1 (CI 0.8, 1.7); at 5 – 9 years, RR 1.1 (CI 0.7, 1.7): and at 10 years, RR 0.7 (CI 0.5, 0.9). Adjustment for BMI reduced the benefit of quitting smoking at 10 years (RR 0.8; CI 0.5, 1.0).
EGO SEEMAN
H. Summary Tobacco is used by many individuals during a large proportion of their lives and apears to be associated with an increased risk of fracture of the axial and appendicular skeleton in women and in men. This risk emerges in advanced age and reduces the skeletal protective effects of obesity and estrogen exposure. The effect is mediated, in part, by a reduction in BMD, which is probably due primarily to increased bone loss. Decreased bone formation and increased bone resorption are responsible for bone loss. Increased bone resorption associated with smoking is, in part, due to a reduction in the production and an acceleration in the degradation of estrogen. The mechanism of reduced bone formation is uncertain. The effects of tobacco use on BMD are usually undetectable until late in adulthood because smoking one pack per day on average results in a deficit in spinal BMD of 2% per decade. However, a deficit of 0.5 to 0.8 standardized deviations directly attributable to tobacco use may be incurred over the three decades from age 20 to 50 years, a change that may double fracture risk.
III. ALCOHOL AND BONE A. Fractures 1. ALCOHOL ABUSE Alcohol abuse appears to confer a high risk for fracture in women and men. When present, it is commonly associated with fractures of the axial and appendicular skeleton. However, in women, only a small proportion of fractures are attributable to alcohol abuse. In contrast, alcohol abuse should be strongly suspected in men presenting with fractures. Johnell and colleagues [65] found that 25% of all men and 37% of men over 30 years of age admitted to the hospital for lower extremity fractures had a history of alcohol abuse, [65]. This was found in only 4% of women admitted for lower limb fractures. Most of the studies of fracture prevalence in alcoholic men are based on small series, many of which are uncontrolled. Nevertheless, the majority suggest an unusually high prevalence of fractures, particularly given the age distribution of subjects. De Vernejoul and colleagues [46] reported that 11 men with osteoporosis had a history of heavy alcohol and tobacco use. Bikle [66] reported that 25% of alcoholic subjects under 45 years of age had fractures of the spine. Lindsell and colleagues [67] found rib fractures in 28% of 72 patients with alcoholic cirrhosis, 1% of 77 with non-alcohol-related liver disease, and 7% of 149 controls. Israel and colleagues observed that 29% of 198 male alcoholics had rib or spine fractures compared to 2% of 218 controls. The authors suggest that the presence of a verte-
CHAPTER 31 Effects of Tobacco and Alcohol Use on Bone
bral fracture in men should be regarded as a sign of alcohol consumption. Diamond et al. [69] found that 30% of the 40 subjects who drank alcohol had a history of one or more fractures (wrist, 6; femoral neck, 4; ribs, 4; others, 7). Only 6 (15%) had radiological evidence of crush fractures. Unilateral and bilateral femoral neck fractures occur with increased frequency in alcohol abusers [71 – 74]. 2. MODERATE ALCOHOL CONSUMPTION Evidence for an association between moderate alcohol intake and fractures is contradictory. Men with spine fractures report a higher prevalence of alcohol and tobacco consumption. Seeman et al. [10] reported that moderate alcohol use was associated with a relative risk of 2.4 in 105 men with spine fractures and controls. The risk of osteoporosis increased by 1.007 for each ounce year of cumulative exposure (P 0.01). The relationship between these factors is shown in Fig. 2. The risk was age dependent. The effects of alcohol were not evident in men under 60 years old, emerged in 60 to 69 years old, and were strongest in those 70 years and over. In nonobese individuals over 70 years who drank alcohol and smoked but had no underlying disease, the relative risk was 20.2 (P 0.05). In a nonobese individual who drank, smoked, and had an underlying illness, the relative risk for osteoporosis and vertebral fractures was 192.5 (P 0.05). In contrast, Hemenway and co-workers [74] observed 271 wrist fractures in a study of 51,529 men aged 50 to 75 years during 271,552 person years of observation. The risk of fracture was unrelated to alcohol or tobacco onsumption, age, height, or weight. Paganini-Hill et al. [5] observed a trend for higher risk for fractures with an increasing number of “shots” of liquor per week after menopause. Women drinking more than eight shots per week had a relative risk of 1.85 compared to nondrinkers. Hernandez-Avila and colleagues [75] surveyed 84,484 women in the United States aged 34 – 59 years and found that 5934 forearm and hip fractures occurred during 6 years. A daily consumption of 25 g alcohol (one to two whiskies) was associated with a risk of 2.33 (95% CI, 1.18, 4.57) for hip fractures and 1.38 (95% CI, 1.09, 1.74) for forearm fractures. Significant trends were found for beer and liquor, but not wine. Hemenway and colleagues [74] studied 96,508 nurses aged 35 – 59 years. There was an interaction between body weight and alcohol intake. Approximately 30% took no alcohol, 12% drank between 0 and 1.4 g per day, 20% drank 1. 5 – 4.9 g per day, 19% drank 5 – 15 g per day, and 13% drank more than 15 g per day. Under 1% of respondents reported a fracture during the 4 years (925 of 96,508). Increased fracture risk was found in women who drank more than 15 g of alcohol per day and had a relative weight of less than 21 kg/m2. The increased risk in lean women was confined to those over 50 years of age. Neither factor was independently associated with fracture risk. Women who
783 drank more than 15 g of alcohol per day were not at increased risk unless they were thin and women who were thin were not at increased risk unless they drank more than 15 g of alcohol per day. Those with low body weight who drank most heavily had an age-adjusted risk of 1.73 (95% CI, 1.3, 2.29). In terms of absolute risk, heavy drinkers with low body weight accounted for less than 4% of the population but had 6% of the fractures. That is, only a small proportion of fractures in women are attributable to alcohol. It is important to note that the study cohort was young with only 17% in the oldest age bracket of 55 – 60 years. Felson and colleagues [76] examined the association between alcohol consumption and hip fractures using a retrospective cohort design. During 117,224 person years, 217 hip fractures occurred (174 in women, 43 in men). In the women, the relative risks were 1, 1.34 (95% CI, 0.91, 1.95), and 1.54 (95% CI, 0.95% CI, 0.92, 2.58) for light, moderate (2 to 6 oz. per week), and heavy (more than 7 oz. per week) consumption categories. In men, the age-adjusted relative risks were 1.0 with light, 0.78 (95% CI, 0.34, 1.78) with moderate, and 1.26 (95% CI, 0.62, 2.55) with heavy intake. All confidence intervals included unity. For the entire group, the relative risk was 1.28 per 7 oz. per week of consumption. In those less than 65 years of age, the relative risk with heavy alcohol consumption was 1.4 (95% CI, 1.07, 1.84). In those aged 65 years and over, the relative risk was 1.17 (95% CI, 0.9, 1.53). Stratification acording to drinking, age, and gender resulted in small numbers of cases in each cell and wide confidence inervals. Moderate alcohol intake (4 – 12 drinks per week) increased the risk for hip fracture in women under 65 years of age. Holdrup and collegues [77] recorded alcohol intake in 17,868 men and 13,917 women in three studies between 1964 and 1992. During follow-up, there were 500 first hip fractures in women and 307 in men. Low to moderate alcohol intake (1 – 13 drinks/week in women; 1 – 27 in men) was not associated with hip fracture. In women who drank 14 – 27 drinks/week, the age-adjusted RR of hip fracture was 1.44 (95% CI 1.03, 2.03) and 1.32 (CI 0.92, 1.87) after adjustment for confounder. In men, the RR of hip fracture increased as intake increased above 28 drinks/week: from 1.75 (CI 1.06, 2.89) for 28 – 41 drinks to 1.84 (CI 1.00, 3.41) for 42 – 69 drinks to 5.28 (CI 2.60, 10.70) for 70 or more drinks, after adjustment for confounder. RR of hip fracture in men and women was higher for those preferring beer (1.46; CI 1.11, 1.91) than wine (0.77; CI 0.58, 1.03) and spirits (0.82; CI 0.58, 1.14). The authors conclude different thresholds exist for the harmful effects of drinking on hip fracture between men and women. Navez Diaz and colleagues [78] reported no detectable association between frequency of alcohol intake and vertebral deformity in 14,237 men and women aged 50 years or greater from 19 European countries. Vertebral
784
EGO SEEMAN
deformity was present in 809 men and 884 women. On stratification by age, women 65 years and over who drink alcohol more than 5 days a week had a reduced risk of vertebral deformity; adjusted for age, BMI smoking, current physical activity, and previous fractures, the OR was 0.65 (95% Ci 0.43, 0.99). Thus, moderate alcohol consumption may be a risk factor for fractures, but available data are inconclusive. A risk associated with alcohol consumption may be seen primarily among those with lower body weight. In women, only a small proportion of the fractures are explained by alcohol exposure.
B. Bone Mineral Density 1. ALCOHOL ABUSE There is great deal of evidence for an association between reduced BMD and alcoholism. Alcoholism is associated with other risk factors for osteoporosis and fractures. Poor nutrition, leanness, liver disease, malabsorption, vitamin D deficiency, hypogonadism, hemosiderosis, parathyroid dysfunction, and tobacco use may contribute to the pathogenesis of bone disease in alcoholism. Saville [79] reported reduced trabecular bone volume in iliac crest samples from 39 alcoholics. Diez and colleagues studied 20 men and 6 women 47.7 11.7 years of age without liver disease. An intake of at least 150 g alcohol per day for 8 years or more was associated with decreased bone volume (RR 0.06; 95% CI, 0.01, 0.34). Dalen and Lamke [81] found that well nourished alcoholics had a deficit in BMD of 4.7% at the proximal femur and 8.2% at the calcaneus, with more rapid bone loss at these sites during 40 months of follow-up. Chon et al. [81] reported reduced BMD by about 0.5 to 0.7 SD in chronically alcoholic, but otherwise healthy, men abstinent from alcohol for a median of 4 months. Reduced BMD is not always reported. Harding and colleagues [83] studied alcoholics ages 20 to 40 years and found no reduction in BMD. Laitinen and colleagues found no deficit in BMD at the spine or proximal femur in 27 eugonadal noncirrhotic alcoholic men. The man duration of drinking was 17 years, ranging from 6 to 30 years. Those with longer history of alcohol consumption had lower BMD after adjusting for age and weight. Peris and associates [85] investigated the effect of trauma and/or ethanol induced osteopenia in 76 chronic male alcoholics and in 62 matched controls. Twenty-seven alcoholics (36%) had 41 vertebral fractures and 46 (61%) had a history of nonvertebral fractures. Lumbar spinal BMD was lower in alcoholics than controls (1.12 0.2 vs 1.19 0.1g/cm2, (P 0.009) but did not differ between those with and without fractures. By densitometric criteria, 22 alcoholics (29%) and 5 controls (8%) had osteoporosis. Eight of the former had vertebral fractures and 5 had a lumbar BMD below the vertebral threshold. Previous trauma
was reported by 24 (89%) alcoholics with vertebral fracture and by 28 (57%) without. Alcoholic men frequently have vertebral fractures despite a normal BMD. 2. MODERATE ALCOHOL CONSUMPTION Data on the association between moderate alcohol consumption and BMD are difficult to interpret. In one study of 142 men and 220 women, BMD was measured 12 years after documentation of alcohol intake by a questionnaire [86]. Intake was recorded as low (less than 87.3 g per week or 19.1 g per week or 19.1 g per day), medium (87. 4 – 180.9 g per week or 19. 2 – 41.1 g per day), or high (greater than 181 g per week or 41.2 g per day). Increasing alcohol consumption was associated with higher BMD at the proximal femur in men and at the spine in women. In relation to 1 week of alcohol intake, femoral neck BMD in men and lumbar spine BMD in women were higher in those with a higher alcohol intake (P 0.01). In relation to 24-h intake, BMD at the radius and spine increased in women, not men. In the study by Grainge and colleagues [40] alcohol consumption was not associated with BMD in postmenopausal women. Heavy beer drinkers had lower BMD than non/moderate drinkers; heavy wine drinkers appeared to have higher BMD, while for spirits the pattern was unclear. Hansen and colleagues [87] reported a decreased rate of bone loss in 121 postmenopausal women with moderate alcohol consumption followed for 12 years. Angus and colleagues [88] and Laitinen and colleagues [89] have also reported higher BMD in association with moderate alcohol consumption. Gonzalez-Galkin et al. [90] studied 26 heavy drinkers (more than 100 g ethanol per day for more than 10 years), 13 moderate drinkers (60 – 100 g ethanol per day), and 19 controls. BMD was reduced and correlated with cumulative alcohol intake. These subjects had no underlying cirrhosis. BMD was reduced by about 1 SD at the lumbar spine and 0.6 SD at the femoral neck in 51 heavy drinkers and by about 0.5 SD in 51 in moderate drinkers. Slemenda and colleagues [91] studied 111 U.S. male veterans and found that those drinking more than 1.5 drinks per day lost more bone than nondrinkers during 16 years of follow-up. May and associates [41] studied 458 men with a mean age of 69.1 years (range 64 – 76). BMD at the proximal femur was higher in drinkers than in nondrinkers before and after adjusting for age and weight. After further adjustment for tobacco use, caffeine intake, and activity, the differences remained significant at the trochanter only. The authors classified intake into units, with one unit being equivalent to 9 g of alcohol. The mean alcohol intake was 10 units ranging from 1 to 100 units per week. Among the 335 drinkers, BMD was 0.79 0.13 g/cm2 and among the 123 nondrinkers, BMD was 0.75 0.12 g/cm2 at the femoral neck (P 0.008). The adjusted differences were nonsignificant. At the trochanter the values were 0.75 0.13 g/cm2 versus 0.71 0.12 g/cm2 (P 0.007). This remained significant
CHAPTER 31 Effects of Tobacco and Alcohol Use on Bone
after adjustment. BMD at Ward’s triangle in drinkers was 0.54 0.13 g/cm2 versus 0.50 0.12 g/cm2 (P 0.006). This was not significant after adjustment. Differences in BMD in the lumbar spine were not significant. Men in the highest tertiles of alcohol intake (more than 11 units per week) had higher glutamyl transferase levels and a lower caffeine intake and were younger. Alcohol intake of one to two drinks per day did not appear to have a detrimental effect on BMD. The results expressed as a function of increasing alcohol consumption showed a trend for higher BMD in those consuming a greater number of units per alcohol per week but at no site were these increments significant after adjusting for covariates. Although the authors concluded that alcohol intake is associated with higher BMD, the evidence for this is not compelling. Thus, available data suggest that moderate alcohol intake is unlikely to be associated with lower BMD. When an association is found, the deficits appear to be explained by confounding factors. The association between moderate alcohol intake and higher BMD remains a possibility, but should be interpreted with caution. Although these results were adjusted for covariates, moderate alcohol intake may be associated with higher socioeconomic groups with better nutrition and lifestyle during growth and adulthood.
C. Histomorphometry before and after Abstention from Alcohol Osteoporosis is common in alcoholics. Osteomalacia may be found but is less common. Arlot and colleagues studied 77 French patients with aseptic necrosis of the hip. Sixty-eight had low trabecular bone volume and of these, 29 were alcoholics. Only 9 had osteomalacia and of these, 4 were alcoholics. Bikle reported a high bone turnover osteoporosis in younger alcoholics. Low turnover osteoporosis is common in chronic alcoholism. Saville [79] reported reduced trabecular bone volume in iliac crest samples from 39 alcoholics. Diez and colleagues [80] studied 20 men and 6 women 47. 7 11.7 years of age without liver disease. An intake of at least 150 g alcohol per day for 8 years or more was associated with decreased bone volume (RR 0.06; 95% CI, 0.01, 0.34). Static and dynamic histomorphometric measures show that the reduction in BMD is due to reduced bone formation with a contribution of increased bone resorption. In the study by Diez et al. [80] data were reported relative to biopsies from 26 normal subjects (kidney donors), 8 male, 18 female, mean age 44.8 16 years (SD). The following parameters of bone formation were reported: decreased bone formation rate (0.023 0.028 mm3/mm2/day vs 0.0409 0.020 mm3/mm2/day, P 0.013), decreased mineralized surfaces (3.75 4.4% vs 5.9 2.9%, P 0.01), decreased osteoid maturation rate (2.4
785 2.8%/day vs 3.1 0.1%/day, P 0.02), increased mineralization lag time (175.5 405.5 days vs 34.5 8.1 days, P 0.01, P 0.008), and reduced mineral appositional rate (0.309 0.269 mm/day vs 0.771 0.529 mm/day, P 0.03). Increased resorption surface and increased osteoclast number were observed. In a study of female rats by Diez and associates [93], intraperitoneal ethanol (2 g/kg BW) or saline was administered followed by sacrifice at 1, 4, 8, and 24. Ethanol decreased osteoid surface with osteoblasts (42.86 15.61 vs 64.57 6.24%, P 0.05) and osteoclast number (0.05 0.02 vs 0.17 0.09 n/mm2, P 0.05). Osteoclast surfaces decreased at 4 (0.129 0.09 vs 0.425 0.26%, P 0.05), with an increase at 8 (0.765 0.24 vs 0.226 0.17, P 0.01). The osteoblast surface was unchanged. Thus, measures of bone formation are reduced, with a reduction in remodeling consistent with a reduction in bone turnover. Chappard and co-workers [94] reported reduced trabecular bone volume (14.2 4.6% vs 18.8 4.8%, P 0.01), mean wall thickness (39.6 8.1 mm vs 50.2 8.7 mm, P 0.001), and bone formation rate as assessed by double tetracycline labelling in 20 patients 59.1 10.1 years of age with alcoholic cirrhosis. The mineral appositional rate was reduced by about 50%, where mineralizing surfaces were reduced and the bone formation rate was half that found in controls (0.009 0.001 mm3/mm2/year vs 0.0175 0.0125 mm3/mm2/year, P 0.001). Trabecular thickness was reduced (106 25.2 mm vs 131 19 mm, P 0.05). There was no change in trabecular number. Eroded surfaces were increased (8.1 5.2% vs 3.7 1.1%, P 0.001). Osteoclast numbers were not increased. Schnitzler and associates [95] examined the contributions of alcohol and iron in osteoporosis associated with hemosiderosis. Iliac crest bone biopsies were taken from 53 black male drinkers: 38 with and 15 without iron overload. Bone volume and trabecular thickness were lower in both groups than in matched controls and were attributed to alcohol. Mineralization lag time was increased in 34% of patients with iron overload and in 27% of those without, combined with low normal or subnormal osteoid thickness; this evidence of osteoblast dysfunction is also attributed to alcohol. Erosion depth (r 0.373), trabecular number (r 0.295), and trabecular separation (r 0.347) correlated with iron granule number in marrow and are attributed mainly to iron overload. Crilly and colleagues [94] examined the effects of abstention from alcohol on bone histomorphometry in men (16 current drinkers and 9 abstainers). Osteoid seams were increased in abstainers (11.4 0.68 mm vs 7.95 0.48 mm, P 0.001). The osteoid surface with osteoblasts was 20.1 4.3% vs 5.5 1.7% (P 0.01). Mean wall thickness was higher but not significantly so in abstainers (41.0 2.8 mm vs 34.7 2.3 mm, NS). Likewise, seams with double labels were higher but not significantly so in
786 abstainers (0.58 0.07 vs 0.37 0.10, NS). Mineralization rate was 0.52 0.1 mm/day vs 0.26 0.07 mm/day (P 0.04). Mineralization lag time was 28 4 days vs 62 10 days, (P 0.006). Resorption parameters were no different. Diamond and associates [97] studied 28 current drinkers and 12 subjects who had not taken alcohol for at least 6 months. Thirty-five controls were studied without clinical or biochemical evidence of liver disease. Reduced BMD was found in both drinkers and abstainers, but abstainers had higher osteoblast perimeters, mineralizing perimeters, and bone formation rate with a shorter mineralization lag time than patients who continued to drink. Serum testosterone concentrations were higher in abstainers than in drinkers (18.5 3.1 pg/ml vs 14.7 1.7 pg/ml) and did not differ from controls. Lindholm and co-workers [98] studied the reversibility of the effects of alcohol abuse on bone histomorphometry in men who had abstained from alcohol for at least 2 years. There were eight abstainers and nine drinkers. Mineralizing surfaces, bone formation rate, osteoid width and wall thickness, adjusted appositional rate, and formation period were greater in abstainers than in drinkers. There were no significant differences in resorptive indices between drinkers and abstainers. Median activation frequency was higher in abstainers than in drinkers. Although there was no difference in 25-hydroxyvitamin D, 1,25-dihydroxyvitamin D was higher in abstainers (111.6 vs 56.4, median values, P 0.01). Data were consistent with reversible histological changes, although ideally paired biopsies would have provided more convincing evidence, as disease severity may have been less in those able to abstain from alcohol. There was no evidence of osteomalacia. In growing animals, alcohol appears to reduce growth in size and volumetric density, effects that are partially restored by resumption of growth with cessation of alcohol. For example, Sampson and colleagues [99] studied 4-weekold female rats fed a liquid diet containing 35% ethanol-derived calories, a liquid control diet, or standard chow for 2 or 4 weeks. The alcohol group gained less weight than chow controls at 2 and 4 weeks. Relative to liquid controls, the alcohol group had lower femoral length, diameter, volume, wet weight, dry weight, ash weight, BMD, and percentage mineral at 4 weeks. Chow-fed animals had similar or greater values for these parameters than liquid controls. Tibial trabecular area and thickness in the alcohol group were lower than in chow controls but not in liquid controls. Serum insulin-like growth factor-I (IGF-I) was reduced at 2 and 4 weeks by the alcohol diet relative to chow controls, with intermediate levels in the liquid controls. In animals fed alcohol for 2 weeks then chow for 2 weeks, relative to the group alcohol fed for 4 weeks, the recovery group had greater weight; greater femoral length, volume, wet weight, dry weight, and ash weight; lower tib-
EGO SEEMAN
ial trabecular area; and greater tibial trabecular number and thickness. Femoral density, percentage mineral, and percentage water did not differ. The recovery group was below pair-fed controls for femoral length, wet weight, dry weight, ash weight, and density. The recovery group had IGF-I levels close to, but below, controls. The authors concluded that animals removed from the alcohol diet improved incompletely in all parameters, probably due to the continued growth of young bones rather than to regaining bone lost during alcohol consumption. Kidder and colleagues [100] studied 6-month-old ovariectomized rats receiving a liquid diet supplemented isocalorically with 13 or 35% ethanol for 2 months. Ethanol at both concentrations reduced the ovariectomy-related increase in body weight. The 35% ethanol group increased tibial cortical medullary area (1.31 0.04 vs 1.08 0.07 mm2) relative to ovariectomy alone. Cortical bone area and periosteal perimeter were unchanged. Dynamic parameters in cortical bone and static and dynamic parameters in cancellous bone were unaltered. The authors concluded that a chronic ingestion of high doses of alcohol does not accentuate bone loss in ovariectomized rats. Thus, the main histomorphometric abnormality is a reduction in parameters of bone formation, although eroded surfaces may be increased. A reduction in bone formation may be a direct toxic effect of alcohol causing a reduction in osteoblast life span or a reduction in the activity of osteoblasts. A reduction in serum testosterone concentration may contribute, as there is an association between bone volume and serum testosterone in alcoholic cirrhosis [100].
D. Biochemical Measures of Bone Remodeling In general, measurement of biochemical markers of bone remodeling confirms that bone loss is due to a reduction in bone formation rather than increased bone resorption. In the vast majority of studies, circulating osteocalcin, a marker of bone formation, is reduced and increases with abstention from alcohol. Jaouhari and co-workers [101] studied 40 chronic alcoholics before and after 3 weeks of ethanol withdrawal and compared the results with 50 nonalcoholic controls matched for age and gender. Plasma osteocalcin was reduced in the cases (3.0 2.6 mg/liter vs 4.7 2.8 mg/liter). After 21 days of withdrawal, osteocalcin levels increased (5.8 3.5 mg/liter) and were no different from those of controls. The hydroxyapatite-binding capacity of the plasma osteocalcin before and after withdrawal was also not different from that of controls. These features support the view that the reduction in bone formation may be reversible. Pepersack and colleagues [102] studied the effects of chronic alcoholism on bone turnover markers in 12 alco-
787
CHAPTER 31 Effects of Tobacco and Alcohol Use on Bone
holic men before and during a 2-week period of alcohol withdrawal. Osteocalcin was reduced and urinary excretion of hydroxyproline was increased, suggesting an imbalance in bone turnover. An increased renal threshold for phosphate (TmP/GFR) was present without a change in serum (PTH). Following alcohol withdrawal, serum osteocalcin increased from 1.8 0.3 to 2.6 0.3 ng/ml, no different from the control value of 2.6 0.2 ng/ml. TmP/GFR was 3.8 0.3 mg/dl and increased to 4.3 0.2 mg/dl after 2 weeks. Fasting hydroxyproline excretion was elevated compared to that of controls (19.9 1.7 mg/mg versus 13.4 1.8 mg/mg creatinine) and did not fall to the normal range after 2 weeks. No changes were observed in serum concentrations of calcium, phosphate, magnesium, albumin, (PTH), 25-hydroxyvitamin D, or 1,25-dihydroxyvitamin D during the observation period. Preedy and colleagues [103] studied the urinary excretion of collagen degradation products using a rat model of alcoholic bone disease and suggested that there is a reduction in bone resorption. Six weeks of ethanol feeding showed that there was a reduction in total and conjugated deoxypyridinoline, a marker found in only type I collagen of bone and dentine. There was a 20% reduction in free deoxypyridinoline (NS). The 24-h urinary excretion of total, free, and conjugated deoxypyridinoline was reduced by 25 – 55%. Total, free, and conjugated pyridinoline were unaltered. There was no effect on the 24-h urinary excretion of pyridinoline. (Pyridinoline is found in cartilage and bone and some other tissues.) These observations are consistent with low bone remodeling associated with alcohol consumption. Hydroxyproline may be increased in current users of alcohol, consistent with increased bone resorption [104]. However, hydroxyproline may not be from bone. The increase in hydroxyproline excretion reported by Diamond and colleagues [97] may reflect a change in nutritional status, as hydroxyproline excretion is not only determined by bone resorption. Pyridinoline is found largely in type II and type IX collagens of cartilage and to a lesser extent, in type I collagen of bone. Deoxypyridinoline is found only in type I collagen of bone and dentine. Neither cross-links are found in the skin. Laitinen and co-workers [89] studied 27 noncirrhotic male alcoholics in a hospital for 2 weeks. Osteocalcin was reduced by 28% at the time of admission. The procollagen 1 carboxyterminal propeptide was reduced by 17% and normalized with abstention. Ionized calcium increased. PTH was unchanged. Intestinal calcium absorption measured by stable strontium was 37% higher than controls and decreased over time. Rico et al. [104] studied 15 patients with acute alcohol intoxication. Serum osteocalcin was lower than in controls of similar age and sex (2.7 0.9 ng/ml vs 6.6 0.8 ng/ml). There was no correlation between serum osteocal-
cin and alkaline phosphatase. However, there was a correlation between osteocalcin and glutamyl transferase activity (r 0.78, P 0.001).
E. Alcohol and Cellular Function There is evidence based on calvarial cultures that alcohol or acetaldehyde may reduce bone formation. Hurley et al. [105] measured bone formation using tritiated proline incorporation into collagenase-digestible protein and noncollagen protein. One percent ethanol decreased tritiated thymidine incorporation by 31% at 24 h whereas 0.1% ethanol increased tritiated thymidine incorporation by 22%, collagenase digestible protein by 73%, and noncollagen protein by 67% at 24 h. Prostaglandin release was decreased by 88 and 75% using 1 and 0.3% ethanol, respectively. Acetaldehyde inhibited tritiated thymidine and proline incorporation and inhibited PTH-stimulated bone resorption whereas ethanol had no effect. The authors concluded that ethanol has little effect on bone formation or resorption; however, acetaldehyde is a potent inhibitor of both. Giuliani and colleagues [106] reported that ethanol and acetaldehyde inhibit osteoblastogenesis of bone marrow cells, an effect that may contribute to the reduced bone formation found in alcoholics. Farley and colleagues [107] showed that ethanol reduced tritiated thymidine incorporation into monolayer cultures of calvarial cells in a time and dose-dependent manner. Ethanol also reduced the mitogenic action of human skeletal growth factor, sodium fluoride, and PTH. Alkaline phosphatase activity per cell was decreased. Tritiated hydroxyproline synthesis from proline was unaffected by ethanol, but the effect of PTH and sodium fluoride on hydroxyproline synthesis was inhibited with 0.2% ethanol. These features are consistent with the view that ethanol may reduce bone formation directly and indirectly. Ethanol increased cAMP and prostaglandin E2, (PG2E) production by calvarial cells and increased skeletal collagen resorption reflected in the release of tritiated hydroxyproline from intact embryonic chick tibiae prelabeled with tritiated proline in vitro. These authors also provided evidence to suggest that ethanol affects bone cells by changing membrane fluidity. Friday and Howard [108] showed that the synthesis of DNA measured by tritium-labeled thymidine incorporation was reduced in human osteoblasts derived by the collagenase digestion of trabecular bone after exposure to 0.0 5 – 1% ethanol. Protein synthesis measured by tritiated proline incorporation into trichloroacetic acid-precipitable material was reduced. Human bone cell protein concentrations and alkaline phosphatase activity were reduced after exposure to 1% ethanol but not lower doses. Cheung and associates [109] studied the direct effects of ethanol on osteoclasts from long bones of 19-day
788 prehatched chicks. The osteoclasts were seeded onto dentine slices and cultured with and without ethanol. Resorption pits in dentine were measured using confocal laser reflection microscopy. Increased pit numbers, areas, volumes, and volume/area ratios were observed with 0.001 and 0.01% ethanol. Greatest mean volume and area resorbed per pit and number of pits occurred with 0.01% ethanol. Volume/area (mean depth) per pit was greatest with 0.001% ethanol. These are ethanol concentrations encountered in social drinkers. Osteoclasts are derived from a pluripotent stem cell most likely to be a colony-forming unit from the granulocyte – macrophage series (CFU-GM). Ethanol and its metabolites impair GFU-GM formation. Tisman and Herbet [110] showed that ethanol inhibits the proliferation of granulocyte – macrophage progenitor cells. Meagher and colleagues [111] showed that erythroid progenitor cells were suppressed by 0. 0 5 – 0.2% ethanol and 0. 001 – 0.01% acetaldehyde. Granulocyte – macrophage progenitor cell suppression required 3% ethanol or 0.03% acetaldehyde. These suppressive effects were partly reversed by folinic acid or pyridoxine. The lowest concentration of acetaldehyde that inhibited colony formation was 0.001% for burst-forming units – erthyroid (BFU-E) and colony-forming units – erthyroid (CFU-E) and 0.03% for CFU-GM. Balsinde and co-workers [112] showed that ethanol inhibits the mouse peritoneal macrophage superoxide anion response to phorbol myristate acetate, perhaps through disorganization of the plasma membrane. These investigators studied the interaction of ethanol and exogenous arachadonic acid in the generation of extracellular messengers by mouse peritoneal macrophages. Ethanol caused a dose-dependent decrease in the production of cyclooxygenase and lipooxygenase metabolites. Hydroxy-eicostetranoic acid metabolites and 6-ketoprostaglandinl levels were reduced in the presence of ethanol. The effect of ethanol on the metabolism of exogenous arachadonic acid to 5 hydroxy-eicostetranoic acid by macrophages was dose dependent. The failure of generation of oxygenated metabolites of arachadonic acid may be due to the impairment of access of arachadonic acid into the cell. Gilhus and Matre [113] studied mononuclear cells from 10 healthy blood doners. Incubation with ethanol at 37°C and 10 liter reduced active E rosette-forming cells. The percentage of cells with phagocytic capacity was reduced from 11.7 5.4 to 7.0 5.0 (P 0.01). Incubation at 1.0 g/liter had similar effects, but not at 0.1 g/liter. Ethanol reduces the mobilization and phagocytic capacity of macrophages in the lung, in the peritoneal cavity, and in the lining of vascular sinusoids. Fixed macrophages in intoxicated patients have depressed phagocytic activity, but the function returns to normal after abstention. Ethanol impairs the phagocytic ability of monocytes and macrophages and enhances the ability to produce superox-
EGO SEEMAN
ide anions and ruffled borders in hepatic macrophages [94 – 96]. Whether these models apply to the cells of bone is uncertain. Changes in membrane fluidity have been reported with ethanol in vivo and in vitro. Dimethyl sulfoxide, ethylene glycol, and lethicin increase membrane fluidity and mimic the effects of ethanol on bone cell proliferation. The changes in membrane fluidity may contribute to changes in membrane-bound enzymes such as skeletal alkaline phosphatase, changes in bone cell responsiveness to parathyroid hormone, and changes in cyclic AMP production [114 – 116].
F. Gonadal Function Gonadal dysfunction in men may be due to concurrent direct effects of alcohol on testicular function, hypothalamic pituitary function, altered clearance by the liver due to increased hepatic testosterone A ring reductase activity, increased hepatic blood flow, and interference with retinonic acid metabolism. Clarification of whether one or more of these mechanisms is responsible will require studies designed specifically to address this question. Alcoholics may have reduced serum testosterone concentrations. Diamond and associates [69,97] studied 115 consecutive ambulant patients with histologically proven chronic liver disease and 113 age- and gender-matched controls. The 40 alcoholic patients had a higher prevalence of peripheral fractures than patients with other liver disorders. The presence of hypogonadism was an important risk factor. Among 81 eugonadal subjects, 12 had fractures. Among 34 hypogonadal subjects, 21 had fractures. Van Thiel and colleagues [117,118] studied 40 men with alcoholic liver disease and reported reduced serum testosterone with more severe derangements of liver histology. FSH and LH were elevated, consistent with a primary abnormality of testicular function. Sex hormone-binding globulin was elevated eightfold. However, the response to clomiphene stimulation was diminished, suggesting that both a hypothalamic – pituitary and a local testicular defect were responsible for the hypogonadism in these subjects. Mendelson and colleagues [119] showed that acute alcohol intoxication in 16 healthy nonalcoholic males was associated with a reduction in serum testosterone and a rise, not fall, in LH concentration. A rapid fall in testosterone during acute alcohol administration may be the result if hepatic blood flow and liver metabolism of the testosterone increase. Rubin et al. [120] report increased hepatic metabolism via testosterone A ring reductase. Gordon and colleagues [121] found increased metabolic clearance and reduced production of testosterone in 11 healthy men given ethanol for 4 weeks. The male volunteers were ages 20 – 40 years with no evidence of liver dis-
789
CHAPTER 31 Effects of Tobacco and Alcohol Use on Bone
ease. Nine were social drinkers consuming no more than 70 g per week of alcohol and 2 were chronic alcoholics. Subjects were given alcohol for 4 weeks, with liver biopsy before and at the end of the experiment in 5 of the subjects. Hepatic testosterone A ring reductase activity increased in the 5 subjects tested. The mean plasma testosterone concentration decreased in the 6 subjects tested, based on 49 observations in each of 4 subjects and 25 observations in each of 2 subjects per 24 h. The metabolic clearance rate of testosterone increased, while the production rate decreased in 3 of the 4 subjects. Inconsistent results were obtained in measured gonadotropins. Testosterone-binding capacity fell during the study. Gordon and co-workers [122] showed that hepatic aromatase activity is increased in rats exposed to ethanol, and a reduction in plasma testosterone may be due to an increased conversion to estrogen. Plasma testosterone decreased by 55% (318 48 ng/dl vs 144 22 ng/dl), Whereas plasma estrogen increased by 60% (1.5 0.1 ng/dl vs 2.4 0.4 ng/dl). There were no significant differences in testosterone within the testis (assessed in vitro). Hepatic aromatase activity increased with testosterone substrate in the assay (4.8 0.5 pmol/h/10 mg protein vs 6.3 0.7 pmol/h/10 mg protein). Van Thiel and colleagues [123] showed that alcohol-fed rats had a marked reduction in testicular mass compared to controls and that this was caused by a reduction in the mean seminiferous tubular diameter and a reduction in the amount of germinal epithelium. Serum testosterone decreased compared to isocaloric controls (291.0 38.1
FIGURE 11
pg/ml vs 1619.3 283.3 pg/ml). There were no differences in the histology of the pituitary gland in the two groups of animals. Microsomal reductase activity more than doubled in ethanol-fed ratscompared to that seen in controls (43.4 3.1 vs 22.3 1.9, P 0.005). Vitamin A is essential for spermatogenesis [124]. It is ingested as retinol and oxidized to the active retinal by alcohol dehydrogenase. Alcohol dehydrogenase metabolizes ethanol to acetaldehyde. Van Thiel and colleagues [123] showed that ethanol inhibited the oxidation of retinol by testicular homogenates containing alcohol dehydrogenase. Ethanol inhibits testicular retinal formation. Complete inhibition of retinol oxidation occurred at molar concentrations of ethanol greater than 2 105. The hypothesis was not directly tested by measuring either spermatogenesis or testosterone metabolism. Ginsburg et al. [125] studied 12 healthy postmenopausal women aged 54.5 4.0 years receiving estradiol (1 mg/day) and progestin and 12 women aged 61.7 6.4 years not receiving hormone replacement therapy (HRT). Each took 0.7 g/kg alcohol and isoenergetic placebo in random sequence on consecutive days. After alcohol ingestion, circulating alcohol levels were similar in both groups. In the non-HRT group, circulating estradiol was unchanged by alcohol. Circulating estradiol increased three fold in the HRT group (from 297 to 973 pmol/l) and remained elevated for 5 h (Fig. 11). Estradiol and alcohol levels correlated during the ascending and descending phases of the blood alcohol curve (r 0.9, both). Acute alcohol may thus result in sustained elevations in estradiol.
Estradiol levels after alcohol and placebo remain low and not different in women not receiving estrogen. In women receiving estrogen, estradiol increased and remained elevated after alcohol but not placebo. From Ginsburg and colleagues [125], with permission from the authors and publisher.
790
EGO SEEMAN
G. Alcohol and Mineral Metabolism Vitamin D deficiency may result from a lack of sunlight exposure, malabsorption due to gastrectomy, pancreatitis, liver disease, or protein malnutrition. Reduced vitamin D and 1,25-dihydroxyvitamin D levels are reported in alcoholics with or without liver disease. Circulating free hormones are usually normal [66]. The deficits are reversible with the cessation of alcohol [66,97,98] and are partly due to reduced production of the vitamin D-binding proteins. Serum calcium concentrations may be reduced in alcoholics. Ionized calcium activity is usually normal or slightly reduced but may be increased. For example, Diez and colleagues [80] found increased plasma calcium and decreased PTH after adjustment for age and gender in 20 men and 6 women (47.7 11.7 years of age) without liver disease and with a daily intake of at least 150 g alcohol for 8 years or more. In this study, the ethanol group showed reduced plasma calcium (8.51 0.23 vs 9.10 0.29 mg/dl, P 0.01) and reduced PTH (23.51 5.72 vs 76.39 11.66 pg/ml, P 0.001) at 1 h and reduced osteocalcin at 24 h (36.98 2.21 vs 45.77 5.05 ng/ml, P 0.05). Magnesium deficiency is common in alcoholics and is due to malabsorption, renal wasting, and the use of diuretics [93]. It is associated with hypocalcemia, hypoparathyroidism, and PTH resistance with an impaired cAMP response and 1,25-dihydroxyvitamin D response to PTH. Magnesium deficiency may reduce PTH secretion and contribute to a rapid fall of serum calcium with acute alcohol administration [93]. Phosphate deficiency occurs in alcoholics, and reduced muscle phosphate may contribute to the myopathy. PTH may be normal, reduced, or increased [93]. PTH may decrease in response to acute alcohol consumption. This may precede, rather than follow, a fall in ionized calcium activity, so the effect may be direct. Elevated PTH values have been found in alcoholic men [98], but Diamond and colleagues [97] found no difference between drinkers and abstainers. The reasons for the variable observations are uncertain.
cell, nutritional deficiency, depressed glycolytic enzyme activity, direct inhibition of muscle carbohydrate metabolism, and potassium deficiency [126]. Myopathy is characterized by a decreased diameter of type II muscle fibers (fast twitch). The II B fibers, which have no or few mitochondria, are more affected than the type II A fibers. Type I fibers, which are rich in mitochondria and are slow twitch, with aerobic or oxidative metabolism, are less sensitive and may show compensatory hypertrophy. The decrease in type II fibers is responsible for the loss of muscle mass, which may at least in part contribute to the frequency of falls, difficulties in gait, proximal muscle weakness, and perhaps muscle cramps. The acute form of rhabdomyolysis occurs in less than 1% of alcoholics. The pathogenesis is likely to be due to free radical damage. Type I fibers have a higher antioxidant capacity than type II fibers. Alcoholics have a reduced antioxidant status. A deficiency in antioxidants is associated with myopathy in animal models [127,128].
I. Summary Alcohol abuse appears to be associated with an increased risk for fracture in women and men. However, few fractures in women are attributable to alcohol abuse, wheras alcohol abuse with or without concomitant hypogonadism is reported to be much more common in men presenting with fractures. This increased risk may be at least, in part, attributable to a reduction in BMD, which is due to a reduction in bone formation. Increased bone resorption probably contributes, but evidence for this is less compelling. Moderate alcohol consumption may be associated with a higher risk for fractures, but data are not consistent. Some evidence suggests that moderate alcohol consumption may be associated with a higher BMD. A concomitant lifestyle or socioeconomic factors may explain this association.
IV. CONCLUSIONS AND QUESTIONS H. Alcohol and Myopathy Alcohol myopathy is characterized by acute muscle tenderness, swelling, pain, and myoglobinuria. Type I myopathy is a subclinical disorder manifested by biochemical changes only (elevated creatinine phosphokinase). Type II disease is an acute variety characterized by muscle cramps, diffuse muscle weakness, rhabdomyolysis, and myoglobinuria. Type III myopathy is a chronic form associated with proximal muscle wasting and weakness. Several mechanisms are involved in the pathogenesis, including prolonged ischemia, direct injury to the sarcolemmal membrane, toxic inhibition of active transport by the muscle
Tobacco use is an important risk factor for osteoporosis in women. This information should be used by health organizations as a means of dissuading women from taking up smoking or continuing to smoke. Bone loss is likely to be responsible for the low BMD in smokers. The effects of tobacco use on mineral accrual during growth are unknown. Tobacco use is probably a risk factor for osteoporosis in men as well. Alcohol excess appears to be a cause of fractures, but from a public health point of view, this is a problem in men, not in women. Specific studies should be designed to address the question of whether moderate alcohol consumption is a protective factor against fractures, an observation that has been made by several groups. Whether
CHAPTER 31 Effects of Tobacco and Alcohol Use on Bone
confounding by socioeconomic and other factors explain this observation is uncertain. If moderate alcohol use is associated with higher BMD, then the mechanism of this effect needs to be understood.
References 1. V. L. Ernster, Women and smoking. Am. J. Public Health 83(9), 1202 – 1204 (1993). 2. CDC, “Reducing the Health Consequences of Smoking: 25 Years of Progress – A Report of the Surgeon General,” DHHS publication No. (C.D.C.) 89-8411. U.S. Department of Health and Human Services, Public Health Service, Rockville, Md, 1989. 3. A.C. King, C. B. Taylor, and W. L. Haskell, Smoking in older women. Is being a female a “risk factor” for continued cigarette use? Arch. Intern. Med. 150, 1841 – 1846 (1990). 4. M. R. Baker, “The Epidemiology and Etiology of Femoral Neck Fracture,” M.D. thesis, University of Newcastle-upon-Tyne, 1980. 5. A. Paganini-Hill, R. K. Ross, V. R. Gerkins, B. E. Henderson, M. Arthur, and T. M. Mack, Menopausal estrogen therapy and hip fractures. Ann. Intern. Med. 95, 28 – 31 (1981). 6. A. R. Williams, N. S. Weiss, C. L. Ure, J. Ballard, and J. R. Daling, Effect of weight, smoking and estrogen use on the risk of hip and forearm fractures in postmenopausal women. Obstet. Gynaecol. 60, 695 – 699 (1982). 7. N. Kreiger, J. L. Kelsey, T. R. Holford, and T. O’Connor, An epidemiologic study of hip fracture in postmenopausal women. Am. J. Epidemiol. 116, 141–148 (1982). 8. W. J. Boyce, “Fracture of the Proximal Femur: Studies of the Distribution, Risk Factors, Care and Outcome in Oxford, 1983 – 1984,” MFCM thesis, Royal College of Physicians, London, 1985. 9. B. W. Alderman, N. S. Weiss, J. R. Daling, C. L. Ure, and J. H. Ballard, Reproductive history and postmenopausal risk of hip and forearm fracture. Am. J. Epidemiol. 124, 262 – 267 (1986). 10. C. Cooper and C. Wickham, “Smoking and Hormone-Related Disorders” (N. Wald and J. Baron, eds.), Chapter 8, pp. 93 – 99. Oxford Univ. Press, 1990. 11. O. Johnell, B. Gullberg, J. A. Kanis, E.Allander, L. Elffors, J. Dequeker, G. Dilsen, C. Gennari, A.L. Vaz, G.Lyritis, G. Mazzuoli, L.Miravet, M. Passeri, R.P. Cano, A. Rapado, and C. Ribot, Risk factors for hip fracture in European women: The MEDOS study. J. Bone Miner. Res. 10, 1802 – 15 (1995). 12. J. Cornuz, D. Feskanich, W. C. Willett, and G. A. Colditz, Smoking, smoking cessation, and risk of hip fracture in women. Am. J. Med, 106, 311 – 4 (1999). 13. H. Melhus, K. Michaelsson, L. Holmberg, A. Wolk, and S. Ljunghall, Smoking, antioxidant vitamins, and the risk of hip fracture. J. Bone Miner. Res. 14, 129 – 35 (1999). 14. D. P. Kiel, J. A. Baron, J. J. Anderson, M. T. Hannan, and D. T. Felson, Smoking eliminates the protective effect of oral estrogens on the risk for hip fracture among women. Ann. Int. Med. 116, 716 – 721 (1992). 15. L. Forsen, A. Bjorndal, K. Bjartveit, T. H. Edna, J. Holmen, V. Jessen, and G. Westberg, Interaction between current smoking, leanness, and physical inactivity in the prediction of hip fracture. J. Bone Miner. Res. 9, 1671 – 1678 (1994). 16. D. Hemenway, G. A. Colditz, W. C. Willett, M. J. Stampfer, and S. E. Steizer, Fractures and lifestyle: Effect of cigarette smoking, alcohol intake, and relative weight on the risk of hip and forearm fractures in middle aged women. Am. J. Public Health 78, 1554 – 1558 (1988). 17. M. R. Law and A. K. Hackshaw, A meta-analysis of cigarette smoking, bone mineral density and risk of hip fracture: Recognition of a major effect. Br Med. J. 315, 841 – 846 (1997).
791 18. H. W. Daniell, Osteoporosis of the slender smoker: Vertebral compression fractures and loss of metacarpal cortex in relation to postmenopausal cigarette smoking and lack of obesity. Arch. Intern. Med. 136, 298 – 304 (1976). 19. J. Aloia, S. H. Cohn, A. Vaswani, J. K. Yeh, and K. Ellis, Risk factors for postmenopausal osteoporosis. Am. J. Med. 78, 95 – 100 (1985). 20. E. Seeman, L. J. Melton III, T. O’Fallon, and B. L. Riggs, Risk factors for spinal osteoporosis in men. Am. J. Med. 75, 977 – 983 (1983). 21. G.F. Jensen, Osteoporosis of the slender smoker revisited by epidemiologic approach. Eur. J. Clin. Invest. 16, 239 – 242 (1986). 22. G. Jones, M. Riley, and T. Dwyer, Maternal smoking during pregnancy, growth, and bone mass in prepubertal children. J. Bone Miner. Res. 14, 146 – 51 (1999). 23. E. A. Krall and B. Dawson-Hughes, Smoking and bone loss among postmenopausal women. J. Bone Miner. Res. 6, 331 – 338 (1991). 24. E. A. Krall and B. Dawson-Hughes, Smoking increases bone loss and decreases intestinal calcium absorption. J. Bone Miner. Res. 14, 215 – 20 (1999). 25. C. W. Slemenda, S. L. Hui, C. Longcope, and C. Johnston, Jr., Cigarette smoking, obesity, and bone mass. J. Bone Miner. Res. 4, 737 – 741 (1989). 26. D. Sparrow, N. I. Beausoleil, A. J. Garvey, B. Rosner, and J. E. Silbert, The influence of cigarette smoking and age on bone loss in men. Arch. Environ. Health 37, 246 – 249 (1982). 27. A. Rundgren and D. Mellstrom, The effect of tobacco smoking on the bone mineral content of the ageing skeleton. Mechanisms Ageing Dev. 28, 272 – 277 (1984). 28. O. Johnell and B. E. Nilsson, Life-style and bone mineral mass in perimenopausal women. Calcif. Tissue Int. 35, 354–356 (1984). 29. O. Lindquist and C. Bengtsson, Menopausal age in relation to smoking. Acta Med. Scand. 205, 73 – 77 (1979). 30. I. Hollo, I. Gergely, and M. Boross, Influence of heavy smoking upon the bone mineral content of the radius of the aged and effect of tobacco smoke on the sensitivity to calcitonin of rats. Aktuel. Gerontol. 9, 365 – 368 (1979). 31. R. Lindsay, The influence of cigarette smoking on bone mass and bone loss. “Osteoporosis: Recent Advances in Pathogenesis and Treatment” (H. F. Deluca, H. M. Frost, C.C. Johnston, Jr., and A. M. Parfitt, eds.), p. 481. Univ. Park Press, Baltimore, MD, 1981. 32. M. R. Sowers, R. B. Wallace, and J. H. Lemke, Correlates of mid-radius bone density among postmenopausal women: A community study. Am. J. Clin. Nutr. 41, 1045 – 1053 (1985). 33. H. Suominen, E. Heikkinen, P. Vainio, and T. Laitinen, Mineral density of calcaneous in men at different ages: A population study with special references to life-style factors. Age Ageing 13, 273 – 281 (1984). 34. P. McNair, M. S. Christiansen, S. Madsbad, C. Christiansen, C. Binder, and I. Transbol, Bone loss in patients with diabetes mellitus: Effects of smoking. Miner. Electrolyte Metab. 3, 94 – 97 (1980). 35. A. J. Friedman, V. A. Ravnikar, and R. L. Barbieri, Serum steroid hormone profiles in postmenopausal smokers and nonsmokers. Fertil. Steril. 47(3), 398 – 401 (1987). 36. M. A. Hansen, Assessment of age and risk factors on bone density and bone turnover in healthy premenopausal women. Osteopor. Int. 4, 123 – 128 (1994). 37. J. C. Stevenson, B. Lees, M. Devenport, M. P. Cust, and K. F. Ganger, Determinants of bone density in normal women: Risk factors for future osteoporosis? Br. Med. J. 298, 924 – 928 (1989). 38. R. G. McCulloch, D. A. Bailey, C. S. Houston, and B. L. Dodd. Effects of physical activity, dietary calcium intake, and selected lifestyle factors on bone density in young women. Can. Med. Assoc. J. 142, 221 – 227 (1990). 39. M. T. McDermott and M. C. Witte, Bone mineral content in smokers. South. Med. J. 81, 477 – 480 (1988).
792 40. M. J. Grainge, C. A. C. Coupland, S. J. Cliffe, C. E. D. Chilvers, and D. J. Hosking, Cigarette smoking, alcohol and caffeine consumption, and bone mineral density in postmenopausal women. Osteopo. Int. 8, 355 – 63 (1998). 41. H. May, S. Murphy, and K. T. Khaw, Alcohol consumption and bone mineral density in older men. Gerontology 41, 152 – 158 (1995). 42. P. Egger, S. Duggleby, R. Hobbs, C. Fall, and C. Cooper, Cigarette smoking and bone mineral density in the elderly. J. Epidemiol. Community Health 50, 47 – 50 (1996). 43. G. Jones and F. S. Scott, A cross-sectional study of smoking and bone mineral density in premenopausal parous women: Effect of body mass index, breastfeeding, and sports participation. J. Bone Miner. Res. 14, 1628 – 33 (1999). 44. J. L. Hopper and E. Seeman, Bone density in twins discordant for tobacco use. N. Engl. J. Med. 330, 387– 392 (1994). 45. U. Syversen, L. Nordsletten, J. A. Falch, J. E. Madsen, O. G. Nilsen and H. L. Waldum, Effect of lifelong nicotine inhalation on bone mass and mechanical properties in female rat femurs. Calcif. Tissue Int. 65, 246 – 9 (1999). 46. M. C. DeVernejoul, J. Bielakoff, M. Herve, J. Gueris, M. Hott, D. Modrowski, D. Kuntz, L. Miravet, and A. Ryckewaert, Evidence of defective osteoblastic function. A role for alcohol and tobacco consumption in osteoporosis in middle-aged men. Clin. Orthop. Rel. Res. 179, 1107 – 1115 (1983). 47. A. Leino, J. Jorvisalo, O. Impivaara, and M. Kaitsaari, Ovarian hormone status, life-style factors, and markers of bone metabolism in women aged 50 years. Calcif. Tissue Int. 54, 262–267 (1994). 48. M. A. Fang, P. J. Frost, A. Lida-Klein, and T. J. Hahn, Effects of nicotine on cellular function on UMR 106-01 osteoblast-like cells. Bone 12, 283 – 286 (1991). 49. L. G. Lenz, R. G. S. Galvin, and W. K. Ramp, Nicotine inhibits collagen synthesis and alkaline phosphatase activity but stimulates DNA synthesis in osteoblast-like cells. J. Bone Miner. Res. 5, S159 (1990). 50. R. J. S. Galvin, L. G. Lenz, and W. K. Ramp, Effects of smokeless tobacco extracts on DNA and collagen synthesis in fibroblasts and osteoblasts. J. Bone Miner. Res. 5, S159 (1990). 51. B. MacMahon, D. Trichopoulos, P. Cole, and J. Brown, Cigarette smoking and urinary estrogens. N. Engl. J. Med. 307, 1062–1065 (1982). 52. R. L. Barbieri, J. Gochberg, and K. J. Ryan, Nicotine, cotinine, and anabasine inhibit aromatase in human trophoblast in vitro. J. Clin. Invest. 77, 1727 – 1733 (1986). 53. J. Michnovicz, H. Naganuma, R. J. Hershcopf, H. L. Bradlow, and J. Fishman, Increased urinary catechol estrogen excretion in female smokers. Steroids 52, 69 – 83 (1988). 54. J. Michnovicz, R. J. Herschcopf, H. Naganuma, H. L. Bradlow, and J. Fishman, Increased 2-hydroxylation of estradiol as a possible mechanism for the anti-estrogenic effect of cigarette smoking. N. Engl. J. Med. 315, 1305 – 1309 (1986). 55. J. Jensen, C. Christiansen, and P. Rodbro, Cigarette smoking, serum estrogens, and bone loss during hormone-replacement therapy early after menopause. N. Engl. J. Med. 313, 973 – 975 (1985). 56. D. L. Cassidenti, A. G. Vijod, M. A. Vijod, F. Z. Stanczyk, and R. A. Lobo, Short-term effects of smoking on the pharmacokinetic profiles of micronized estradiol in postmenopausal women. Am. J. Obstet. Gynecol. 163, 1953 – 1960 (1990). 57. J. A. Cauley, J. P. Gutai, L. H. Kuller, D. LeDonne, and J. G. Powell, The epidemiology of serum sex hormones in postmenopausal women. Am. J. Epidemiol. 124, 752 – 761 (1989). 58. F. E. Crawford, D. J. Back, M. L’E. Orme, and A. M. Breckenridge. Oral contraceptive steroid plasma concentrations in smokers and non-smokers. Br. Med. J. 282, 1829 – 1830 (1981). 59. J. Geisler, I.H. Omsjo, S.I. Helle, D. Ekse, T. Silsand, and PE Lonning, Plasma ostrogen fractions in postmenopausal women receiving hormone replacement therapy: Influence of route of ad-
EGO SEEMAN
60.
61.
62. 63. 64.
65.
66.
67.
68.
69.
70. 71. 72. 73.
74.
75.
76.
77.
78.
79. 80.
81.
ministration and cigarette smoking. J. Endocrinol. 162, 265 – 270 (1999). E. Barrett-Connor and Kay-Tee Khaw, Cigarette smoking and increased endogenous estrogen levels in men. Am. J. Epidemiol. 126, 187 – 192 (1987). K-T. Khaw, S. Tazuke, and E. Barrett-Connor, Cigarette smoking and levels of adrenal androgens in postmenopausal women. N. Engl. J. Med. 318, 1705 – 1709 (1988). D. J. Handelsman, J. Conway, L. M. Boylan, and J. R. Turtle, Int. J. Androl. 7, 369 – 382 (1984). M. H. Briggs, Med. J. Aust. March 24 (1973). A. W. Meikle, X. H. Liu, G. N. Taylor, and J. D. Stringham, Nicotine and cotonine effects on 3 alpha hydroxysteroid dehydrogenase in canine prostate. Life Sci. 43, 1845 – 1850 (1998). O. Johnell, H. Kristenson, and I. Redlund-Johnell, Lower limb fractures and registration for Alcoholism. Scand J. Soc. Med. 13, 95–97 (1985). D. D. Bikle, Alcohol-induced bone disease. “Osteoporosis: Nutritional Aspects” (A. P. Simopoulous and C. Galli, eds.), Vol. 73, pp. 53 – 79. Basel, Karger, 1993. D. R. M. Lindsell, A. G. Wilson, and J. D. Maxwell, Fractures on the chest radiograph in detection of alcoholic liver disease. Br. Med. J. 285, 597 – 599 (1982). Y. Israel, H. Orrego, S. Holt, D. W. Macdonald, and H. E. Meema, Identification of alcohol abuse: Thoracic fractures on routine chest xrays as indicators of alcoholism. Alcohol. Clin. Exp. Res. 4, 420 – 422 (1980). T. Diamond, D. Stiel, M. Lunzer, M. Wilkinson, J. Roche, and S. Posen, Osteoporosis and skeletal fractures in chronic liver disease. Gut 31, 82 – 87 (1990). B. E. Nilsson, Conditions contributing to fracture of the femoral neck. Acta Chir. Scand. 136, 383 – 384 (1970). N. Dalen and A. L. Feldreich, Osteopenia in alcoholism. Clin. Orthop. 99, 201 – 202 (1974). C. M. Schnizler and L. Solomon, Bone changes after alcohol abuse. S. Afr. Med. J. 66, 730 (1984). I. Elffors, E. Allander, J. A. Kanis, B. Gullberg, O. Johnell, J. Dequeker, G. Dilsen, C. Gennari, A. A. Vaz Lopez, G. Lyritis, G. F. Mazzuoli, L. Miravet, M. Passeri, R. Cano Perez, A. Rapado, and C. Ribot. The variable incidence of hip fracture in southern Europe: The MEDOS study. Osteopor Int. 4, 253 – 263 (1994). D. ‘Hemenway, D. R. Azrael, E. B. Rimm, D. Feskanich, and W. C. Willett, Risk factors for wrist fracture: Effect of age, cigarettes, alcohol, body height, relative weight, and handedness on the risk for distal forearm fractures in men. Am. J. Epidemiol. 140, 361 –367 (1994). M. Hernandez-Avila, G. A. Colditz, M. J. Stampfer, B. Rosner, F. E. Speizer, and W. C. Willet, Caffeine, moderate alcohol intake, and risk of fractures of the hip and forearm in middle-aged women. Am. J. Clin. Nutr. 54, 157 – 163 (1991). D. T. Felson, D. P. Kiel, J. J. Anderson, and W. B. Kannel, Alcohol consumption and hip fractures: The Framingham Study. Am. J. Epidemiol. 128, 1102 – 1110 (1988). S. Holdrup, M. Gronbaek, A. Gottschau, J.B. Lauritzen, M. Schroll, Alcohol intake, beverage preference, and risk of hip fracture in men and women. Am. J. Epidemiol. 149, 993 – 1001 (1999). M. Navez Diaz, T.W. O’Neill and A.J. Silman, The influence of alcohol consumption on the risk of vertebral deformity. Osteopor Int., 7, 65 – 71 (1997). P. D. Saville, Changes in bone mass with age and alcoholism. J. Bone J. Surg. 47, 492 – 499 (1965). A. Diez, J. Puig, S. Serrano, M-L. Marinoso, J. Bosch, J. Marrugat. et al., Alcohol-induced bone disease in the absence of severe chronic liver damage. J. Bone Miner. Res. 9, 825 – 831 (1994). N. Dalen and B. Lamke, Bone mineral losses in alcoholics. Acta Orthop. Scand. 47, 469 – 471 (1976).
CHAPTER 31 Effects of Tobacco and Alcohol Use on Bone 82. K. S. Chon, D. J. Sartoris, S. A. Brown, and P. Clopton, Alcoholismassociated spinal and femoral bone loss in abstinent male alcoholics, as measured by dual x-ray absorptiometry. Skel. Radiol. 21, 431 – 436 (1992). 83. A. Harding, J. Dunlap, S. Cook, A. Mattalino, F. Azar, M. O’Brien, and M. Kester, Osteoporotic correlates of alcoholism in young males. Orthopedics 11, 279–282 (1988). 84. K. Laitenen, C. Lamberg-Allardt, R. Tunninen, M. Harkonen, and M. Valimaki, Bone mineral density and abstention-induced changes in bone and mineral metabolism in noncirrhotic male alcoholics. Am. J. Med. 93, 642 – 650 (1992). 85. P. Peris, N. Guanabens, A. Pares, F. Pons, L. del Rio, A. Monegal, X. Suris, J. Caballeria, J. Rodes and J. Munoz-Gomez, Vertebral fractures and osteopenia in chronic alcoholic patients. Calcif, Tissue Int. 57, 111 – 4 (1995). 86. T. L. Holbrook and E. Barrett-Connor, A prospective study of alcohol consumption and bone mineral density. Br. Med. J. 306, 1506 – 1509 (1993). 87. M. A. Hansen, K. Overgaard, B. J. Riis, and C. Christiansen, Potential risk factor for development of postmenopausal osteoporosis examined over a 12-year period. Osteopor. Int. 1, 95 – 102 (1991). 88. R. M. Angus, P. N. Sambrook, and N. A. Pocock, Dietary intake and bone mineral density. Bone Miner. 4, 265 – 277 (1988). 89. K. Laitinen, M. Valimaki, and P. Keto, Bone mineral density measured by dual-energy x-ray absorptiometry in healty Finnish women. Calcif. Tissue Int. 48, 224 – 231 (1991). 90. J. L. Gonzales-Calvin, A. Garcia-Sanchez, V. Bellot, M. Munoz-Torres, E. Raya-Alvares, and D. Salvatierra-Rios, Mineral metabolism, osteoblastic function and bone mass in chronic alcoholism. Alcohol Alcoholism 28, 571 – 579 (1993). 91. C. W. Slemenda, S. L. Hui, C. Longcope, and C. Johnston, Jr., Cigarette smoking, obesity, and bone mass. J. Bone Miner. Res. 4, 737 – 741 (1989). 92. M. E. Arlot, M. Bonjean, and P. M. Chavassieux, Bone histology in adults with aseptic necrosis. J. Bone J. Surg. 65A, 1319 – 1327 (1983). 93. A. Diez, S. Serrano, J. Cucurull, L. L. Marinoso, J. Bosch, J. Puig, X. Nogues, J. Aubia, Acute effects of ethanol on mineral metabolism and trabecular bone in Sprague-Dawley rats. Calcif-Tissue Int. 61, 168 – 71 (1997). 94. D. Chappard, B. Plantard, M. Petitjeqan, C. Alexandre, and G. Riffat, Alcoholic cirrhosis and osteoporosis in men: A light and scanning electron microscopy study. J. Stud. Alcohol 52, 269 – 274 (1991). 95. C. M. Schnitzler, A. P. MacPhail, R. Shires, E. Schnaid, J. M. Mesquita, and H. J. Robson, Osteoporosis in African hemosiderosis: role of alcohol and iron. J. Bone Miner, Res. 9, 1865 – 1873 (1994). 96. R. G. Crilly, C. Anderson, D. Hogan, and L. Delaquerriere-Richardson, Bone histomorphometry, bone mass, and related parameters in alcoholic males. Calcif. Tissue Int. 43, 269 – 276 (1988). 97. T. Diamond, D. Stiel, M. Lunzer, M. Wilkinson, and S. Posen, Ethanol reduces bone formation and may cause osteoporosis. Am. J. Med. 86, 282 – 288 (1989). 98. J. Lindholm, T. Steiniche, E. Rasmussen, G. Thamsborg, I. Nielsen, H. Brockstedt-Rasmussen, T. Storm, L. Hyldstrup, and C. Schou, Bone disorder in men with chronic alcoholism: A reversible disease? J. Clin. Endocrinol. Metab. 73, 118 – 124 (1991). 99. H. W. Sampson and H. Spears, Osteopenia due to chronic alcohol consumption by young actively growing rats is not completely reversible. Alcohol Clin Exp. Res. 23, 324 – 327 (1999). 100. L. S. Kidder and R. T. Turne, Dietary ethanol does not accelerate bone loss in ovarietomized rats. Alcohol Clin. Exp. Res. 22, 2159 – 2164 (1998). 101. J. Jaouhari, F. Schiele, P. Pirollet, E. Lecomte, F. Paille, and Y. Artur, Concentration and hydroxyapatite binding capacity of plasma osteocalcin in chronic alcoholic men: Effect of a three-week withdrawal therapy. Bone Miner. 21, 171 – 178 (1993).
793 102. T. Pepersack, M. Fuss, J. Otero, P. Bergmann, J. Valsamis, and J. Corvilain, Longitudinal study of bone metabolism after ethanol withdrawal in alcoholic patients. J. Bone Miner. Res. 7, 383, 387 (1992). 103. V. R. Preedy, R. A. Sherwood, C. I. O. Akpoguma, and D. Black, The urinary excretion of the collagen degradation markers pyridinoline and deoxypyridinoline in an experimental rat model of alcoholic bone disease. Alcohol Alcohol. 26, 191–198 (1991). 104. H. Rico, J. A. Cabranes, J. Cabello, F. Gomez-Casresana, and E. R. Hernandez, Low serum osteocalcin in acutre alcohol intoxication: A direct toxic effect of alcohol on osteoblasts. Bone Miner. 2, 221 – 225 (1987). 105. M. M. Hurley, D. L. Martin, B. E. Kream, and L. G. Raisz, Effects of ethanol and acetaldehyde on collagen synthesis, prostaglandin release, and resorption of fetal rat bone in organ culture. Bone 11, 47 – 51 (1990). 106. N. Giuliani, G. Girasole, P. P. Vescovi, G. Passeri, M. Pedrazzoni, Ethanol and acetaldehyde inhibit the formation of early osteoblast progenitors in murine and human bone marrow cultures. Alcohol Clin. Exp. Res. 23, 381 – 385 (1999). 107. J. R. Farley, R. Fitzsimmons, A. K. Taylor, U. M. Jorch, and K-H. W. Lau, Direct effects of ethanol on bone resorptionand formation in vitro. Arch. Biochem. Biophys 238, 305 – 314 (1985). 108. K. E. Friday and G. A. Howard, Ethanol inhibits human bone cell proliferation and function in vitro. Metabolism 40, 562 – 565 (1991). 109. R. C-Y. Cheung, C. Gray, A. Boyde, and S. J. Jones. Effects of ethanol on bone cells in vitro resulting in increased resorption. Bone 16, 143 – 147 (1995). 110. G.Tisman and V. Herbert, In vitro myelosuppression and immunosupression by ethanol. J. Clin. Invest. 52, 1410 – 1414 (1973). 111. R. C. Meagher, F. Sieber, and J. L. Spivak, Suppression of hematopoietic-progenitor-cell proliferation by ethanol and acetaldehyde. N. Engl. J. Med. 307, 845 – 849 (1982). 112. J. Balsinde, A. Schueller, and E. Diez, The interaction of ethanol and exogenous arachidonic acid in the generation of extracellular messengers by mouse peritoneal macrophages. Biochem. Biophys. Acta 970, 83 – 89 (1988). 113. N. E. Gilhus and R. Matre, In vitro effect of ethanol on subpopulations of human blood mononuclear cells. Int. Arch. Allergy Appl. Immunol. 68, 382 – 386 (1982). 114. A. Zuiable, E. Wiener, and S. N. Wickramasinghe, In vitro effects of ethanol on the phagocytic and microbial killing activities of normal human monocytes and monocyte-derived macrophages. Clin. Lab. Haematol. 14, 137 – 147 (1992). 115. A. Castro, D. L. Lefkowitz, and S. S. Lefkowitz, The effects of alcohol on murine macrophage function. Life Sci. 52, 1585 – 1593 (1993). 116. S. Yamada, S. Mochida, A. Ohno, K. Hirata, I. Ogata, Y. Ohta, and K. Fujiwara, Evidence for enhanced secretory function of hepatic macrophages after longterm ethanol feeding in rats. Liver 11, 220 – 224 (1991). 117. D. H. Van Thiel, J. S. Gavaler, R. Lester, and M. D. Goodman, Alcohol-induced testicular atrophy. An experimental model for hypogonadism occurring in chronic alcoholic men. Gastroenterology 69, 326 – 332 (1975). 118. D. H. Van Thiel, R. Lester, and R. J. Sherins, Hypogonadism in alcoholic liver disease: Evidence for a double defect. Gastroenterology 67, 1188 – 1199 (1974). 119. J. H. Mendelson, N. K. Mello, and J. Ellingboe, Effects of acute alcohol intake on pituitary – gonadal hormones in normal human males. J. Pharmacol. Exp. Ther. 202, 676 – 682 (1977). 120. E. Rubin, C. S. Lieber, K. Altman, G. G. Gordon, and A. L. Southren, Testosterone metabolism in the liver. Science 191, 563 – 564 (1976).
794 121. G. G. Gordon, K. Altman, A. L. Southren, E. Rubin, and C. S. Lieber, Effect of alcohol (ethanol) administration on sex-hormone metabolism in normal men. N. Engl. J. Med. 295, 793 – 797 (1976). 122. G. G. Gordon, A. L. Southren, J. Vittek, and C. S. Lieber, The effect of alcohol ingestion on heptic aromatase activity and plasma steroid hormones in the rat. Metabolism 28, 20 – 24 (1979). 123. D. H. Van Thiel, J. Gavaler, and R. Lester, Ethanol inhibition of vitamin A metabolism in the testes: Possible mechanism for sterility in alcoholics. Science 186, 941 – 942 (1974). 124. J. L. Leathem, In “The Testes” (A.D. Johnson, W. R. Gomes, and N. L. Vandermark, eds.) Vol. 3, pp. 188 – 190. Academic Press, New York, 1970.
EGO SEEMAN 125. E. S. Ginsburg, N. K. Mello, J. H. Mendelson, R. L. Barbieri, S. K. Teoh, M. Rothman, X. Gao, and J. W. Sholar, Effects of alcohol ingestion on estrogens in postmenopausal women. JAMA 276, 1747 – 51 (1996). 126. S. Feitberg, S. Epstein, F. Ismail, and C.D’Amanda, Deranged bone mineral metabolism in chronic alcoholism. Metabolism 36, 322 – 326 (1987). 127. J. P. Knochel, G. L. Bilbrey, T. J. Fuller, and N. W. Carter, The muscle cell in chronic alcoholism: The possible role of phosphate depletion in alcoholic myopathy. Ann. N.Y. Acad. Sci. 252, 274 – 286 (1975). 128. V. R. Preedy and T. J. Peter, Alcohol and muscle disease. J. R. Soc. Med. 87, 188–189 (1994).
795
CHAPTER 32
Falls as Risk Factors for Fractures CHAPTER 32
Falls as Risk Factors for Fractures ANN V. SCHWARTZ ELIZABETH CAPEZUTI JEANE ANN GRISSO
Department of Epidemiology and Biostatistics, University of California, San Francisco, School of Medicine, San Francisco, California 94143 Nell Hodgson Woodruff School of Nursing, Emory University, Atlanta, Georgia 30322 Center for Clinical Epidemiology and Biostatistics, Division of General Internal Medicine, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104
IV. Prevention of Falls and Fall-Related Fractures V. Summary and Directions for Future Research References
I. Introduction II. Risk Factors for Falls III. Risk Factors for a Fall-Related Injury
I. INTRODUCTION Falls are a common event among the elderly. Population-based surveys and prospective studies indicate that around 30% of community-dwelling elderly fall one or more times each year. Of those who report falling, about half report more than one fall during the year [1 – 10]. In residential institutions, the proportion of those who fall is higher, about 40 – 50% [11 – 13]. Fall-related injuries are the leading cause of mortality due to unintentional injuries among adults 65 and older in the United States [14]. In 1995, 11,057 deaths among the elderly were attributed to falls [14]. This figure, which is derived from death certificates, is likely to underestimate the extent to which falls play a role in fatalities. About one-third of all deaths from falls occur among those 85 years of age or older [15]. Falls are also the leading cause of nonfatal injuries among the elderly [16]. An estimated 5 – 10% of those over age 75 visit a hospital emergency department each year for
OSTEOPOROSIS, SECOND EDITION VOLUME 1
treatment of a fall-related injury; about one-third are subsequently hospitalized [17 – 20]. Fall-related injuries include fractures and other serious injuries (dislocated joints, subdural hematomas, and lacerations requiring sutures) and minor injuries (bruises, abrasions, certain sprains, and other soft tissue injuries). Fractures alone account for 30 – 40% of fall injuries among the elderly [19,21 – 23]. One-fourth to one-third of these fractures are of the hip, resulting in over 250,000 hospital admissions for hip fracture each year in the United States [24]. Hip fractures in particular are associated with increased mortality and disability. Of those who can walk without assistance at the time of hip fracture, half cannot resume this level of independence afterward [25]. Several types of fractures among the elderly are usually due to a fall. It is well documented that over 90% of hip fractures in the elderly occur as a result of a fall. Other fractures that are largely due to falls include those of the distal forearm (Colles’ or wrist fracture), pelvis, humerus, rib, leg, hand, patella, ankle, elbow, and face [26 – 30].
795
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
796
SCHWARTZ, CAPEZUTI, AND GRISSO
Although fall-related injuries can be devastating, most falls do not result in serious injury. In prospective studies of community-dwelling elderly, about 10% of falls result in an injury requiring medical attention. About 3 – 5% of falls in the elderly result in a fracture; about 1% result in hip fracture [23]. However, even falls that do not result in serious injury may have long-term health consequences for the elderly. Fear of falling can lead to decreased independence and mobility [31]. Multiple falls are often a reason for admission to a nursing home [32]. The percentage of the elderly who fall increases with age, more steeply after age 75. Women are somewhat more likely to fall than men until age 85 when the percentages who report a fall are nearly equal [33]. These estimates of fall frequency are based on studies among mainly non-Hispanic Caucasian women. Investigation of fall rates in other groups is limited. Mexican-American women appear to fall at a similar rate [34]. A study of older Japanese-American women found that their age-adjusted fall rate was about half the rates reported for white women [35]. Two prospective studies indicate that fall frequency may be somewhat lower among black women than among white women [5,8].
II. RISK FACTORS FOR FALLS A. Introduction Most falls in the elderly are probably due to both intrinsic (host) and extrinsic (environmental) factors. Although there has been a great deal of progress in the identification of intrinsic risk factors, investigation of extrinsic factors remains more limited. Environmental factors are thought to be particularly important in falls among the more active elderly, whereas intrinsic factors may play more of a role among the frail elderly [33]. In studies of risk factors for falls, a fall is usually defined as “an event that results in a person coming to rest inadvertently on the ground or other lower level.” Falls due to an overwhelming force or event, such as a motor vehicle accident or loss of consciousness, are usually excluded [32].
B. Intrinsic (Host) Factors
TABLE 1 Intrinsic Risk Factors for Falls among the Elderlya Risk factor
Evidence for association
Demographic characteristics Older age
Gender, Women
Race, White
Functional level ADL/IADL
Cane/Walker use
History of falls
Gait, balance, strength Walking speed
Lower extremity strength
Upper extremity strength
Postural sway
Impaired reflexes
Sensory Vision
Lower extremity sensory perception
Chronic illnesses Heart disease
/
Parkinson’s disease
Other neuromuscular disease
Stroke
Urinary incontinence
Arthritis
Acute illness
Medications, alcohol No. medications used
Hypnotics
Sedatives
Antipsychotics
Antidepressants
Antiparkinson drugs
Cardiac
/
Diuretics
/
Antihypertensives
/
Alcohol
/
Mental status Cognitive impairment Depression
, strong; , moderate; /, inconsistent.
a
Falls among the elderly are generally associated with frailty and poor health. Table 1 summarizes major intrinsic factors that have been identified as risk factors for falls. A person’s chances of falling increase with age but the functional level appears to be a better predictor of falls than age alone [5,8,36]. Limitations in both activities of daily living (ADL) and instrumental activities of daily living (IADL)
are associated with an increased risk of falling [5,8,13,36 – 38]. Functional or performance-based measures indicating deficits in balance, gait, and strength are strongly associated with falls [2,4,5,8,9,13,36,38 – 41], with evidence that fall risk increases with the number of disabilities [38]. These
CHAPTER 32
797
Falls as Risk Factors for Fractures
measures include tests such as standing up from a chair, walking speed, step length, and postural sway. Reduced grip and lower extremity strength are also correlated with falls [1,2,5,13,36,39,42,43]. Maintaining postural control and avoiding environmental obstacles depend on proprioceptive, vestibular, and visual input translated into appropriate motor responses [44]. Both impaired visual acuity [5,8,12,37,38,45] and diminished sensory function in the lower extremities [8,13,42,46] have been shown to be associated with falls in many, but not all [2,4,5,13,36,47], studies. It may be that visual or somatosensory impairment is most important for those who have other impairments, such as gait abnormalities or muscle weakness [23]. The ability to translate perceptions into appropriate motor responses diminishes with age [48]. Laboratory-based studies indicate that older adults have more difficulty maintaining balance when their attention is divided [49]. However, detailed measurements of central neurological processes have not been included in larger epidemiological studies of risk factors for falls. Reaction time has been assessed in a few studies with varying results [5,50,51]. Global cognitive impairment is associated with an increased risk of falls [2,4,7,8,37,52]. Impaired cognitive functioning is a significant predictor of fall-related injury in both nursing home residents [22,53] and community-residing older adults [54,55]. Hypothesized causal connections include a neurologically based reduction in the ability to maintain balance and behavioral changes, such as wandering, associated with impaired judgement [52,56]. Depression is also found in association with increased falls [5,8,37,40,57]. Depression is associated with disability and poor health and may lead to decreased attention to environmental hazards. In addition, antidepressant medications are associated with falls [8,58 – 60]. The presence of certain chronic medical conditions has been found to increase the risk of a fall, including Parkinson’s disease [2,5,12], urinary incontinence [5,8,13], dementia [12,52], history of stroke [2,7,61], and arthritis [1,2,5,58,62]. A period of acute illness may also lead to a greater risk of a fall [8,38,63]. Use of a greater number of medications is associated with falls, although it has been difficult to distinguish an effect of medications from the chronic condition that is being treated [64]. Specific medications may increase the risk of falls through depressed psychomotor function, reduced alertness, greater fatigue, or postural hypotension [58]. For psychotropic medication [65,66], an increase in the risk of falls seems to be well established. A recent meta-analysis found that among 14 cardiac and analgesic drugs or drug groups, only diuretics, digoxin, and type IA antiarrhythmic agents significantly increased the risk of falling [64]. Alcohol use has been postulated to increase the risk of falls and has been evaluated in a range of studies [67].
However, most studies, including community-based prospective studies, have not found an association [2,5,7,8,68].
C. Extrinsic Factors Environmental or extrinsic risk factors include situational or activity related factors as well as environmental hazards. Most falls occur when older persons are performing their usual activities, such as rising from a chair or ambulating [8,9,11,38,69 – 71]. The relationship between environmental hazards and falls for community-residing older adults is most frequently ascertained by self-report [5,72,73] with subjects implicating environmental factors in one-third to one-half of their falls [32,74,75]. Home hazards include inappropriately placed furniture or objects, scatter rugs, and slippery surfaces [74,76,77]. Environmental hazards outside the home include irregular sidewalk surfaces, ill-repaired stairs, and traffic lights that do not allow sufficient time to cross the street [74]. Table 2 lists potential hazards that can be found in the home. Observational studies have not identified strong associations between environmental hazards and falls, but this may be the result of inherent difficulties in study design [5,8,78 – 80]. Assessing environmental influences is difficult because (i) environmental conditions may change over time, (ii) only a fraction of falls are likely to occur in any particular location, and (iii) there are no standardized methods developed to use in attributing a fall to a specific environmental hazard. Environmental hazards may be more important for healthy and mobile elders [81]. In a prospective study of falls in 325 community-residing older adults, Northridge et al. [82] found an association between falls and the home environment for vigorous but not frail participants. Certain functional characteristics (notably arthritis and poor depth perception) increased the likelihood of falls in the presence of environmental hazards [83]. Although observational studies have not provided strong evidence for a role of environmental hazards in falls, a prevention trial focused on home hazard reduction demonstrated some success in decreasing falls (see Section IV,A,5) [76]. For hospitalized and institutionalized elderly, traveling to the bathroom is the most frequently reported activity-related risk for falls [84 – 88]. In the hospital, disorienting effects of an unfamiliar environment are considered a major causative factor for increased fall risk. Similarly, for nursing home residents, high fall rates are observed most often in the first week after admission or transfer to a different unit in the same nursing home [70,89]. Similar to the community setting, general environmental factors contributing to falls in nursing homes have not been well studied. One review of the falls literature proposes an
798
SCHWARTZ, CAPEZUTI, AND GRISSO
TABLE 2
Environmental Home Hazards
Ground surfaces Throw rugs Loose carpets (not taped or tacked down) Slippery floors Cords and wires on the floor Low-lying objects, e.g., toys Stairs with rugs or in poor repair Furniture
injury [22]. Another study examining restraint use in confused ambulatory nursing home residents found an increased risk of falling among restrained residents compared to nonrestrained residents [93]. Accurate information on whether restraints were being used at the time of a fall is difficult to obtain [96,97] so studies to date have not considered whether restraints have a direct effect on falls and injuries. Most importantly, studies examining the effect of reducing restraint use in nursing homes report no significant increase in falls or fall-related injuries [98 – 101].
Clutter, especially if there is insufficient space for unobstructed mobility Unstable furniture Low-lying furniture, e.g., coffee table
III. RISK FACTORS FOR A FALL-RELATED INJURY
Low chairs without armrest support or seat backs Beds that are too high or too low Cabinets that are either too high or too low Lighting Glare from unshielded windows or lamps or highly polished floors Low or dim compounded by dark-colored walls Absence of night-lights Bathroom Low toilet seats and/or no secure grab bars
Early research on the identification of fall risk factors has been focused on all falls, despite the fact that the majority of falls do not result in injury. Some researchers suggest that fall-related serious injury, not falls per se, is the significant outcome measure and, thus, research and preventive actions should focus on risk factors for injurious falls [70,102,103]. However, it is plausible that efforts to reduce falls in general would also reduce injuries [104]. The most effective approach remains to be clearly established.
Absence of slip-resistant, strongly secured grab bars Absence of nonslip surfaces or assistive devices (e.g., tub chairs in bathtub) Door jambs Other Poorly maintained walking aids and equipment Improper shoes (not slip-resistant, high-heeled, too large)
association between long trouser legs and ill-fitting shoes and increased fall risk [90]. Other environmental factors are generally regarded as less problematic for nursing home residents, as nursing home regulations mandate specific environmental adaptations meant to prevent falls. Physical restraints, including side rails, have been considered an environmental risk factor for falls due to a significant incidence of falls in restrained elders [22,91 – 94]. While restraint use was traditionally viewed as a preventive strategy, observational studies have indicated that restraint use does not reduce fall risk and may instead contribute indirectly to an increase in falls and injuries. Physical restraint use, by immobilizing older persons, can result in reduced muscle mass, strength, joint flexibility, and vasomotor stability [22,95]. In a prospective study, nursing home residents who were restrained for any amount of time during followup had an increased risk of fall-related serious injury even after controlling for risk factors highly associated with
A. Differentiating Risk Factors for Falls from Risk Factors for Injurious Falls Table 3 lists risk factors for injurious falls. Many of the risk factors are the same as those in Table 1 for falls in general, including slow gait, lower and upper extremity weakness, and use of psychotropic medications. However, there are some risk factors specific for injurious falls. Circumstances of the fall, including the height of the fall and orientation of the faller, influence the risk of injury. Certain host factors such as low bone density increase the risk of injury in a fall. Recurrent fallers are, not surprisingly, more likely to sustain a fall-related injury [22,36,105 – 108]. In a population-based case-control study, a strong association was found between reported numbers of falls in the past year and risk of hip fracture, especially among men [109]. Lipsitz et al. [36] contended that “because recurrent fallers are most likely to experience injury from repeated episodes, they constitute an important group to target for diagnostic and preventive efforts.” There is, however, also evidence that factors beyond the greater frequency of falls among the elderly are associated with a greater risk of fall injuries. Speechley and Tinetti found that although “frail” elderly fell more often, they had a lower rate of serious injury per fall than the “vigorous” elderly, leading to a similar rate of serious injury due to
CHAPTER 32
799
Falls as Risk Factors for Fractures
TABLE 3
Risk Factors for Injurious Falls
Risk factor
Evidence for association
Orientation of fall Slow gait
Falling sideways
Falling on/near hip
Falling while turning
Falling while reaching
Falls from a standing height
Protective responses Neurological disease
Upper extremity weakness
Lower extremity weakness
Use of ambulatory aids
Syncopal fall or “drop attack”
Long-acting hypnotic-anxiolytic (benzodiazepines) use
Tricyclic antidepressant use
Antipsychotic use
Impaired cognition
Physical restraint-induced deconditioning
Local shock absorbers Lower body weight/skinfold thickness
Weakness of hip abductors
Falls on hard surface
Bone strength Low bone mineral density/osteoporosis
Caucasians
Women
falls in these two groups [81]. Older women who experience a rapid increase in falls appear to be at higher risk of fracture, independent of the total number of falls (A. Schwartz, personal communication). Cummings and Nevitt [110] noted that the risk of hip fracture among elderly white women increases very steeply with age, whereas the risk of falling rises only moderately.
B. Risk Factors for Injurious Falls 1. HIP FRACTURE MODEL The most comprehensive model for causes of hip fracture identifies and categorizes risk factors according to orientation of the fall, protective responses, local shock absorbers, and bone strength [110]. It is hypothesized “that risk factors for each step in the sequence will
interact in a multiplicative fashion” [110]. Although hip fractures (as opposed to other injuries) are the focus of this model, it is appropriate since the most devastating consequence of injurious falls is hip fracture [108] (see also Chapter 19). 2. ORIENTATION OF THE FALL The points of impact of a fall influence the type and extent of injury. The elderly may be more likely to land on the hip in a fall, thus increasing their risk of hip fracture, whereas, most injurious falls in young and middle-aged adults result in wrist fractures. This is most likely a result of the more rapid gait speed in the young in comparison to that in older persons, which tends to propel the faller forward, resulting in principal impact on the wrist (Fig. 1). In contrast, injurious falls in older persons generally occur while the individual is standing still, during transfer, or while walking slowly with little forward momentum [2,110]. Risk of hip fracture is reported as significantly increased in women falling sideways, straight down, or those who fall on or near the hip; falling backward may be associated with a decreased risk of hip fracture [111 – 113]. Falling during a displacing activity, such as turning, is more likely to result in a serious injury than falling while walking in one direction [81,110]. Certain circumstances are more likely to increase loading forces on the proximal femur, thus resulting in hip fracture. Falls from a standing height or higher are more likely to result in injury [114,115]. Similarly, Tinetti’s study of nursing home residents reported that those who fell while rising from a chair were unlikely to be injured [102]. The risk of fall-related injury in community-residing elderly is higher during stair climbing and when turning around or reaching for objects [55,115]. Risk of hip fracture is increased in taller subjects [112], presumably due to falling from a greater height as well as longer hip axis length [116]. 3. PROTECTIVE RESPONSES Several reflexes and postural responses are initiated during a fall, which potentially prevent or reduce injury. These responses are protective if they can “change the orientation of the faller or reduce the energy of a fall” [110]. The effectiveness of reflex actions depends on the speed of execution and the strength of the muscles initiating the protective movement [110]. Research links impaired protective responses with an increased likelihood of hip fracture. These factors include preexisting neurologic conditions, both upper and lower limb dysfunction or weakness, and use of ambulatory aids [22,50,54,55,102,106,112,117 – 121]. Fallers who initiated protective responses, such as grabbing or hitting an object before landing and those who
800 landed on their hand, were less likely to sustain a hip fracture than those who did not [112]. During syncopal falls, the faller is unable to initiate protective responses, which contributes to a greater likelihood of fractures [19,55]. Likewise, during a “drop attack” a person is unable to recognize the sensation of falling and thus will not display an appropriate protective response [106]. The sedating effect of certain drugs can impair protective responses. Thus, in addition to increasing the risk of falls generally, long-acting hypnotics – anxiolytics (including benzodiazepines), tricyclic antidepressants, and antipsychotics may increase the risk of a fall-related injury [65]. 4. LOCAL SHOCK ABSORBERS The impact of falls can be potentially absorbed by surrounding soft tissue (skin, fat, and muscles), thereby reducing the likelihood of an injury. A lower prevalence of injury occurs in fallers with more skinfold thickness [106] or overall higher body weight [52,54,112,122 – 124]. The apparent protective effect of higher body weight on reducing the risk of hip fracture may be due to increased bone mineral density, local shock-absorbing capacity of muscle and fat, or both. The strength of hip abductors, the largest group of muscles surrounding the hip, is postulated to affect absorption on impact [110]. The risk of fall-related serious injury in older adults is increased when falls occur on a hard surface [55,112,123]. Grisso et al. [125] reported falling on a hard surface as significantly associated with hip fracture; however, this was more likely in women 55 to 74 years of age than in those 75 or older. Flooring specifically designed to reduce the impact of a fall may reduce the likelihood of hip fracture [126]. 5. BONE STRENGTH Bone strength is considered the “last defense” against hip fractures. A fracture will occur if the residual energy of the fall applied to the proximal femur exceeds the strength of the bone [110]. Bone mineral density is the usual measure of bone strength, although other factors, such as geometry and microarchitecture of the bone, may also be important [116]. Nevitt and colleagues [112] found substantial increases in hip and wrist fractures in those with lower bone density, including a nearly sevenfold increase in the risk of a hip fracture when a woman with low bone density falls on the hip. Numerous studies indicate a direct relationship between fracture risk and osteoporosis [27,127,128]. Race, size, and gender, indirect indicators of osteoporosis, also are frequently cited [19,22,27,55,120,129]. The Cummings et al. model provides a useful framework for understanding how numerous risk factors may contribute to injurious falls [110].
SCHWARTZ, CAPEZUTI, AND GRISSO
IV. PREVENTION OF FALLS AND FALL-RELATED FRACTURES A. Prevention of Falls 1. MULTIFACTORIAL INTERVENTIONS Randomized trials have demonstrated the possibility of reducing the rate of falls among the elderly. The most promising results to date have been obtained with multifactorial interventions [130]. Tinetti et al. [131] randomized elderly members of an HMO to either interventions for a range of identified problems or social visits with research staff. The interventions targeted postural hypotension, use of sedative – hypnotics, multiple prescription drugs, environmental hazards, balance and gait impairments, and decreased muscle strength [131]. Results showed a significantly lower fall rate among those persons receiving the intervention (35%) than among the social visit group (47%) [132]. Targeted risk factors were also significantly reduced, with the intervention group showing decreases in the use of multiple medications as well as improvement in measures of balance, gait, and functional impairment [133]. In a later study, Close and co-workers [134] followed communitydwelling older patients who had presented to an emergency room (ER) with a fall. They demonstrated a reduced risk of falling after the ER visit (OR 0.39; 95% CI 0.23 – 0.66) in the group that received a detailed medical and physical therapy assessment with referral as indicated compared to usual care. The benefits of a comprehensive individual assessment were also demonstrated in a nursing home setting by Ray et al. [135], who found a 19.1% (95% CI 2.4 – 35.8%) reduction in the proportion of re-current fallers at intervention facilities. 2. EXERCISE Exercise alone has been shown to reduce falls in three studies using different exercise programs: Tai Chi [136], strength and endurance training [137], and strength and balance exercises combined with walking [138]. A metaanalysis of FICSIT (Frailty and Injuries: Cooperative Studies of Intervention Techniques) studies found that the exercise component of the tested interventions reduced falls by about 10% [139]. In other studies, however, exercise has not been shown to be effective in reducing falls [140 – 143]. To date, clinical trials of exercise with falls as an outcome have not targeted those over 75 years of age. However, evidence shows that exercise programs are effective in improving strength and balance, and thus reducing the risk factors for falls, among older individuals. In the Boston FICSIT study, Fiatarone and co-workers [144] demonstrated, in a nursing home population whose mean age was 87 years, that a progressive resistance exercise training
CHAPTER 32
801
Falls as Risk Factors for Fractures
intervention significantly increased muscle strength, gait velocity, the ability to climb stairs, and the general level of physical activity. 3. MEDICATIONS Campbell and colleagues [145] conducted a trial of withdrawal of psychotropic medication among communitydwelling elderly and reported a 66% reduction in fall risk for the intervention group. Some multifactorial interventions have also included a component designed to reduce medication use. Reducing inappropriate drug prescribing and prescribing of multiple medications may be a particularly promising way of preventing falls for several reasons. First, the problem is extremely common. In a recent analysis of a national survey of community-dwelling persons 65 years or older, 25 to 32% of those surveyed were taking potentially inappropriate medications [146]. This means that up to 9 million older persons may be at risk of adverse effects of drugs, including falls. Second, it is relatively easy to identify individuals who are taking these medications through clinical practices and nursing homes. Extensive screening procedures are not required. Third, interventions targeting physician prescribing patterns could potentially affect large numbers of older persons, given that most older persons have primary care physicians whom they visit regularly. However, changing physician prescribing patterns is not as simple as might be supposed. Studies have documented that simply providing information is not adequate [147]. Promising methods include training “opinion leaders” who then influence their colleagues, providing feedback to individual physicians, and instituting administrative interventions that include incentives or disincentives. Two controlled trials have demonstrated substantial reductions in the use of psychoactive drugs. These studies, however, were conducted in nursing homes where it may be easier to affect prescribing patterns. Ray et al. [148] reported that an intervention including in-service education of administrative, physician, and nursing staff resulted in a 72% reduction in the days of antipsychotic use compared with only a 13% reduction in the control nursing homes (P 0.001). Avorn et al. [149] conducted a randomized controlled trial of an educational intervention in which a clinical pharmacist conducted multiple interactive visits with the physicians who commonly prescribed psychoactive drugs. Psychoactive drug use declined by 27% in experimental nursing homes compared with 8% in control homes (P 0.02) [149]. There has also been documentation that regulations or administrative procedures can result in decreased antipsychotic drug use [150,151]. Shorr and co-workers [150] reported that antipsychotic drug use decreased by 27% (P 0.001) after implementation of the Nursing Home Reform Act [Omnibus Budget Reconciliation Act (OBRA) of 1987]. This federal
legislation was designed in part to reduce unnecessary antipsychotic drugs through establishing clinical practice guidelines and follow-up inspections of Medicaid/ Medicare certified nursing homes. In addition, permanent withdrawal from certain medications may be difficult to achieve. In the clinical trial of withdrawal from psychotropic medications reported by Campbell et al. [145], the authors noted that about half of the participants who withdrew from psychotropics had restarted medication within a month of the trial’s completion. 4. VISION Visual impairment has been found to be a risk factor for falls and hip fracture [8,45,120,127,152,153]. Although correction of vision problems has been included in some multifactorial interventions [134,154,155], it has not been the specific focus of a prevention trial to date. Numerous age-related physiological changes occur in the eye that commonly affect vision in older persons, including cataracts, glaucoma, diabetic retinopathy, and macular degeneration [8]. Many of these conditions, if detected early, can be treated and improved by measures ranging from providing appropriate eyewear to surgery for cataracts. In addition, it is also important to develop interventions that help older persons adapt to functional limitations resulting from decreased vision, such as visual rehabilitation programs, or environmental adaptations, such as adequate lighting, placement of bright-colored tape on steps, and other home modifications. Studies are needed to evaluate the impact of interventions to improve visual function on the risk of falls. 5. EXTRINSIC FACTORS Although fall prevention programs usually advocate environmental modifications, few studies have evaluated the impact of environmental hazard reduction on fall occurrence. Hornbrook et al. [76] conducted a randomized trial of a falls prevention program that included home safety as part of the intervention. The study demonstrated a reduction in the proportion of fallers of about 15% (P 0.05), although the number of falls was only reduced by 7%. Home safety hazards were identified using a standard hazards protocol. Participants were provided information on procedures for obtaining technical and financial assistance in making safety repairs and were invited to attend fall prevention classes. Sixty-two percent of intervention households received financial assistance to make safety repairs, and some modifications were achieved in every intervention household (including removal of throw rugs, objects in pathways, installation of night-lights, and installation of bathtub nonskid mats or strips). Schwarz and colleagues [156] conducted a controlled trial in low-income
802
SCHWARTZ, CAPEZUTI, AND GRISSO
innercity neighborhoods that reported a statistically significant reduction in home hazards. However, no reduction in fall injury rates was observed through ongoing surveillance of emergency department records (D. Schwarz, personal communication).
B. Prevention of Fall-Related Injuries Although preventing fall-related injuries and subsequent disability is most important, to date no published studies have addressed the impact of preventive measures on injurious falls as a primary outcome. One of the reasons is that serious injuries are less common, and thus large sample sizes would be required to detect clinically important effects. Nevertheless, many interventions designed to prevent fall occurrence may also be effective in preventing fall-related injuries. A pooled estimate from five multifactorial interventions indicated a possible reduction of 30% in falls requiring medical care (OR 0.70; 95% CI 0.47 – 1.04) [130]. One of the most disabling of fall outcomes is the occurrence of a fracture. The probability that a given fall results in fracture is, in part, dependent on the level of bone mass and perhaps the quality of the bone structure. Several medications have been demonstrated to prevent loss of bone mass and reduce fracture risk, including alendronate [157,158], estrogen [159], and raloxifene [160] (see also Chapters 69 and 72). Clinical trials have demonstrated that calcium supplementation is somewhat effective in preserving bone mass, particularly among older women with low calcium intakes (no greater than 400 mg/day) [161]. Vitamin D supplementation may also be a valuable preventive measure. A randomized double-blind, placebo-controlled clinical trial of vitamin D and calcium supplements demonstrated a 26% reduction in the incidence of nonvertebral fractures [162] (see also Chapters 67 and 68). Extrinsic factors may also be effective in reducing the likelihood that a fall will result in serious injury. In laboratory simulations of a fall on the hip, polyurethane foam pads can reduce the peak impact force by nearly 20%. Although the peak femoral force remains greater than the force needed to cause a femoral fracture for the currently available hip pads, at least one trial has demonstrated a 53% reduction in hip fractures among persons randomized to receive external hip protectors [163]. Of note, during this trial no hip fracture was sustained by any of the six residents wearing a hip protector at the time of a fall, and none of the eight residents in the intervention group who had a hip fracture was wearing a hip protector at the time of the fracture. However, the authors found that only 24% of residents given hip protectors wore them regularly. Improvements in the design of hip protectors may increase compliance as well as effectiveness [164].
FIGURE 1
(A) fall occurring while standing still, walking slowly, or slowly descending a step has little forward momentum. With little forward momentum, the principal point of impact will be near the hip. (B) A fall occurring during rapid walking has enough forward momentum to carry the faller onto the hands or knees instead of the hip.
Other environmental modifications are also promising. Experts in biomechanics and engineering have developed a new kind of flexible flooring that may reduce the force of the impact from a fall by as much as 30% [165]. Beds that can be adjusted electrically from 23 inches down to 7 inches above the floor have become readily available [166]. A pilot study demonstrated less risk of falls, recurrent falls, and injuries among frail nursing home residents using a low-height bed (10 inches above the floor) compared to residents using a standard bed (21 inches above the floor) [167]. In addition, traditional environmental modifications have not been evaluated but could assist in protective responses (such as grab bars in bathrooms or second hand rails on staircases) or diminish the force of the impact of a fall (such as wall-to-wall carpeting or flexible flooring). Finally, interventions should be considered on a community or governmental level. Although investigators and health care providers had been trying to reduce the use of psychotropic medications in nursing homes for years, the implementation of the federal regulations promulgated by the Nursing Home Reform Act (OBRA-87) resulted in significant reductions in a very short period of time. In addition, implementation of the OBRA-87 legislation reduced the use of physical restraints without an increase in fall-related injuries [101]. Similarly, instituting safe access
CHAPTER 32
Falls as Risk Factors for Fractures
for older persons and the disabled in public areas may be effective in preventing falls and fall-related injuries that occur outside the home.
V. SUMMARY AND DIRECTIONS FOR FUTURE RESEARCH In recent years, we have learned a great deal about the importance of falls in relation to the risk of osteoporotic fractures. There have been significant advances in knowledge concerning which fall risk factors are important determinants of osteoporotic fractures. We have learned that multiple factors affect fall risk and that the most effective interventions will probably need to address several types of risk factors. Although we have made significant progress in understanding the pathogenesis of falls and fall-related injuries, there is much still to be learned. Future research to test the effectiveness of various interventions in reducing fall injuries is especially needed.
References 1. A. J. Blake, K. Morgan, and M. J. Bendall, Falls by elderly people at home: Prevalence and associated factors. Age Ageing 17, 365 – 372 (1988). 2. A. J. Campbell, M. J. Borrie, and G. F. Spears, Risk factors for falls in a community-based prospective study of people 70 years and older. J.Gerontol. 44, M112 – M117 (1989). 3. W. A. Hale, M. J. Delaney, and W. C. McGaghie, Characteristics and predictors of falls in elderly patients. J. Fam. Pract. 34, 577 – 581 (1992). 4. S. R. Lord, R. D. Clark, and I. W. Webster, Physiological factors associated with falls in an elderly population. J. Am. Geriatr. Soc. 39, 1194 – 1200 (1991). 5. M. C. Nevitt, S. T. Cummings, and S. Kidd, Risk factors for recurrent nonsyncopal falls: A prospective study. JAMA 261, 2663 – 2668 (1989). 6. B. C. Perry, Falls among the elderly living in high-rise apartments. J. Fam. Pract. 14, 1069 – 1073 (1982). 7. D. Prudham, Factors associated with falls in the elderly: A community study. Age Ageing 10, 141 – 146 (1981). 8. M. E. Tinetti, M. Speechley, and S. F. Ginter, Risk factors for falls among elderly persons living in the community. N. Engl. J. Med. 319, 1701 – 1707 (1988). 9. A. K. Topper, B. E. Maki, and P. J. Holliday, Are activity-based assessments of balance and gait in the elderly predictive of risk of falling and/or type of fall? J. Am. Geriatr. Soc. 41, 479 – 487 (1993). 10. C. Wickham, C. Cooper, and B. M. Margetts, Muscle strength, activity, housing and the risk of falls in elderly people. Age Ageing 18, 47 – 51 (1989). 11. J. H. Gurwitz, M. T. Sanchez-Croos, M. A. Eckler, and J. Matulis, The epidemiology of adverse and unexpected events in the longterm care setting. J. Am. Geriatr. Soc. 42, 33 – 38 (1994). 12. P. O. Jantti, V. I. Pyykko, A. L. J. Hervonen, V. I. Pyykko, and A.L. Hervonen, Falls among elderly nursing home residents. Public Health 107, 89 – 96 (1993). 13. A. S. Robbins, L. Z. Rubenstein, and K. R. Josephson, Predictors of falls among elderly people: Results of two population-based studies. Arch. Intern. Med. 149, 1628 – 1631 (1989).
803 14. National Safety Council, “Accident Facts.” National Safety Council, Itasca, IL (1998). 15. S. P. Baker, B. O’Neill, and R. S. Karpf, “The Injury Fact Book.” Oxford Univ. Press, New York (1992). 16. D. P. Rice, E.J. MacKenzie, et al., “Cost of Injury in the United States: A Report to Congress, 1989.” Johns Hopkins Univ. Baltimore MD (1989). 17. D. l. Covington, J. G. Maxwell, and T. V. Clancy, Hospital resources used to treat the injured elderly at North Carolina trauma centers. J. Am. Geriatr. Soc. 41, 847 – 852 (1993). 18. D. F. Schwarz, J. A. Grisso, C. G. Miles, et al., A longitudinal study of injury morbidity in an African-American population. JAMA 271, 755 – 760 (1994). 19. R. W. Sattin, D. A. L. Humber, C. A. deVito, et al., The incidence of fall injury events among the elderly in a defined population. Am. J. Epidemiol. 131, 1028 – 1037 (1990). 20. D. Fife, J. I. Barancik, and B. F. Chatterjee, Northeastern Ohio trauma study. II. Injury rates by age sex, and cause. Am. J. Public Health 74, 473 – 478 (1984). 21. J. A. Grisso, D. F. Schwarz, V. Wolfson, et al., The impact of falls in an inner-city elderly African-American population. J. Am. Geriatr. Soc. 40, 673 – 678 (1992). 22. M. E. Tinetti, W. L. Liu, and S. F. Ginter, Mechanical restraint use and fall-related injuries among residents of skilled nursing facilities. Ann. Intern. Med. 116, 369 – 374 (1992). 23. M. C. Nevitt, Falls in the elderly: Risk factors and prevention. In “Gait Disorders of Aging: Falls and Therapeutic Strategies” (J. C. Masdeu, L. Sudarsky, and L. Wolfson, eds.), pp. 13 – 36. LippincottRaven, Philadelphia (1997). 24. National Center for Health Statistics, National hospital discharge survey: Annual summary, 1991. In “Vital and Health Statistics, Series 13, No. 114.” Department of Health and Human Services, Hyattsville (1993). 25. Institute of Medicine, Osteoporosis. In “The Second Fifty Years: Promoting Health and Preventing Disability” (R. L. Berg and J. S. Cassells, Eds.). National Academy Press, Washington, DC (1990). 26. J. A. Cauley, S. R. Cummings, D. G. Seeley, et al., Effects of thiazide diuretic therapy on bone mass, fractures, and falls. Ann. Intern. Med. 118, 666 – 673 (1993). 27. S. R. Cummings, J. L. Kelsey, M. C. Nevitt, and K. J. O’Dowd, Epidemiology of osteoporosis and osteoporotic fractures. Epidemiol. Rev. 7, 178 – 208 (1985). 28. M. R. Griffin, W. A. Ray, R. L. Fought, and L. J. I. Melton, Blackwhite differences in fracture rates. Am. J. Epidemiol. 136, 1378 – 1385 (1992). 29. J. L. Kelsey, W. S. Browner, D. G. Seeley, et al., Risk factors for fractures of the distal forearm and proximal humerus. Am. J. Epidemiol. 135, 477 – 489 (1992). 30. D. G. Seeley, W. S. Browner, M. C. Nevitt, et al., Which fractures are associated with low appendicular bone mass in elderly women? Ann. Intern. Med. 115, 837 – 842 (1991). 31. M. E. Tinetti, C. F. Mendes de Leon, J. T. Doucette, et al., Fear of falling and fall-related efficacy in relationship to functioning among community-living elders. J. Gerontol. 49, M140 – M147 (1994). 32. Kellogg International Work Group on the Prevention of Falls by the Elderly, The prevention of falls in later life. Dan. Med. Bull. 34 (Suppl. 4), 1 – 24 (1987). 33. M. C. Nevitt, Falls in older persons: Risk factors and prevention. In “The Second Fifty Years: Promoting Health and Preventing Disability” (R. L. Berg and J. S. Cassells, eds.), National Academy Press, Washington, DC (1990). 34. A. V. Schwartz, M. L. Villa, M. Prill, J. A. Kelsey, J. A. Galinus, R. R. Delay, M. C. Nevitt, D. A. Bloch, R. Marcus, and J. L. Kelsey, Falls in older Mexican-American women. J. Am. Geriatr. Soc. 47, 1 – 8 (1999).
804 35. J. W. Davis, P. D. Ross, M. C. Nevitt, and R. D. Wasnich, Incidence rates of falls among Japanese men and women living in Hawaii. J. Clin. Epidemiol. 50, 589 – 594 (1997). 36. L. A. Lipsitz, P. V. Jonsson, M. M. Kelley, and J. S. Koestner, Causes and correlates of recurrent falls in ambulatory frail elderly. J. Gerontol. 46, M114 – M122 (1991). 37. A. J. Campbell, J. Reinken, and B. C. Allan, Falls in old age: A study of frequency and related clinical factors. Age Ageing 10, 264 – 270 (1981). 38. M. E. Tinetti, T. F. Williams, and R. Mayewski, Fall risk index for elderly patients based on number of chronic disabilities. Am. J. Med. 80, 429 – 434 (1986). 39. G. M. Gehlsen and M. H. Whaley, Falls in the elderly. II. Balance, strength, and flexibility. Arch. Phys. Med. Rehabil. 71, 739 – 741 (1990). 40. B. E. Maki, P. J. Holliday, and A. K. Topper, A prospective study of postural balance and risk of falling in an ambulatory and independent elderly population. J. Gerontol. 49, M72 – M84 (1994). 41. C. Ring, U. S. L. Nayak, and B. Isaac, Balance function in elderly people who have and who have not fallen. Arch. Phys. Med. Rehabil. 69, 261 – 264 (1988). 42. G. S. Sorock and D. M. Labiner, Peripheral neuromuscular dysfunction and falls in an elderly cohort. Am. J. Epidemiol. 136, 584 – 591 (1992). 43. R. H. Whipple, L. I. Wolfson, and P. M. Amerman, The relationship of knee and ankle weakness to falls in nursing home residents: An isokinetic study. J. Am. Geriatr. Soc. 35, 13 – 20 (1987). 44. N. B. Alexander, Postural control in older adults. J. Am. Geriatr. Soc. 42, 93 – 108 (1994). 45. R. Q. Ivers, R. G. Cumming, P. Mitchell, and K. Attebo, Visual impairment and falls in older adults: The Blue Mountains Eye Study. J. Am. Geriatr. Soc. 46, 58 – 64 (1998). 46. J. K. Richardson, C. Ching, and E. A. Hurvitz, The relationship between electromyographically documented peripheral neuropathy and falls. J. Am. Geriatr. Soc. 40, 1008 – 1012 (1992). 47. W. C. Graafmans, M. E. Ooms, H. M. Hofstee, P. D. Bezemer, L. M. Bouter, and P. Lips, Falls in the elderly: A prospective study of risk factors and risk profiles. Am. J. Epidemiol. 143, 1129 – 1136 (1996). 48. M. H. Woollacott and P. F. Tang, Balance control during walking in the older adult: Research and its implications. Phys. Ther. 77, 646 – 660 (1997). 49. L. A. Brown, A. Shumway-Cook, and M. H. Woollacott, Attentional demands and postural recovery: The effects of aging. J. Gerontol. A Biol. Sci. Med. Sci. 54, M165 – M17 (1999). 50. S. R. Lord, J. A. Ward, P. Williams, and K. J. Anstey, Physiological factors associated with falls in older community-dwelling women. J. Am. Geriatr. Soc. 42, 1110 – 1117 (1994). 51. S. M. Woolley, S. J. Czaja, and C. G. Drury, An assessment of falls in elderly men and women. J. Gerontol. A Biol. Sci. Med. Sci. 52, M80 – M87 (1997). 52. D. M. Buchner and E. B. Larson, Falls and fractures in patients with alzheimer-type dementia. JAMA 257, 1492 – 1495 (1987). 53. J. L. Melton III, C. M. Beard, E. Kokmen, et al., Fracture risk in patients with Alzheimer’s disease. J. Am. Geriatr. Soc. 42, 614 – 619 (1994). 54. J. A. Grisso, G. Y. Chiu, G. Maislin, W. C. Steinmann, and J. Portale, Risk factors for hip fractures in men: A preliminary study. J. Bone. Miner. Res. 6, 865 – 868 (1991). 55. M. C. Nevitt, S. R. Cummings, and E. S. Hudes, Risk factors for injurious falls: A prospective study. J. Gerontol. 46, M164 – M170 (1991). 56. E. H. Franssen, L. E. Souren, C. L. Torossian, and B. Reisberg, Equilibrium and limb coordination in mild cognitive impairment and mild Alzheimer’s disease. J. Am. Geriatr. Soc. 47, 463 – 469 (1999). 57. M. A. Whooley, K. E. Kip, J. A. Cauley, K. E. Ensrud, M. C. Nevitt, and W. S. Browner, Depression, falls, and risk of fracture in older
SCHWARTZ, CAPEZUTI, AND GRISSO
58.
59.
60.
61.
62.
63.
64.
65. 66.
67. 68.
69.
70.
71. 72. 73. 74.
75. 76.
77. 78.
79.
80.
women: Study of Osteoporotic Fractures Research Group. Arch. Intern. Med. 159, 484 – 490 (1999). W. A. Ray and M. R. Griffin, Prescribed medications, falling, and fall-related injuries. In “Reducing Frailty and Falls in Older Persons” (R. Weindruch, E. C. Hadley, and M. G. Ory, eds.), Charles C. Thomas, Springfield, III (1991). R. G. Cumming, J. P. Miller, J. L. Kelsey, et al., Medications and multiple falls in elderly people: The St. Louis OASIS study. Age Ageing 20, 455 – 461 (1991). P. B. Thapa, P. Gideon, T. W. Cost, A. B. Milam, and W. A. Ray, Antidepressants and the risk of falls among nursing home residents. N. Engl. J. Med. 339, 875 – 882 (1998). N. E. Mayo, N. Korner-Bitensky, and R. Becker, Predicting falls among patients in a rehabilitation hospital. Am. J. Phys. Med. Rehabil. 68, 139 – 146 (1989). E. Granek, S. P. Baker, H. Abbey, et al., Medications and diagnoses in relation to falls in a long-term care facility. J. Am. Geriatr. Soc. 35, 503 – 511 (1987). J. Mahoney, M. Sager, N. C. Dunham, and J. Johnson, Risk of falls after hospital discharge. J. Am. Geriatr. Soc. 42, 269 – 274 (1994). R. M. Leipzig, R. G. Cumming, and M. E. Tinetti, Drugs and falls in older people: A systematic review and meta-analysis. II. Cardiac and analgesic drugs. J. Am. Geriatr. Soc. 47, 40 – 50 (1999). R. G. Cumming, Epidemiology of medication-related falls and fractures in the elderly. Drugs Aging 12, 43 – 53 (1998). R. M. Leipzig, R. G. Cumming, and M. E. Tinetti, Drugs and falls in older people: A systematic review and meta-analysis. I. Psychotropic drugs. J. Am. Geriatr. Soc. 47, 30 – 39 (1999). J. E. Carlson, Alcohol use and falls. J. Am. Geriatr. Soc. 41, 346 – 347 (1993). D. E. Nelson, R. W. Sattin, J. A. Langlois, et al., Alcohol as a risk factor for fall injury events among elderly persons living in the community. J. Am. Geriatr. Soc. 40, 658 – 661 (1992). W. P. Berg, H. M. Alessio, E. M. Mills, and C. Tong, Circumstances and consequences of falls in independent community-dwelling older adults. Age Ageing 26, 261 – 268 (1997). P. T. van Dijk, O. G. Meulenberg, H. J. van de Sande, and J. D. Habbema, Falls in dementia patients. Gerontologist 33, 200 – 204 (1993). H. Nickens, Intrinsic factors in falling among the elderly. Arch. Intern. Med. 145, 1089 – 1093 (1985). E. M. Lau and S. P. Donnan, Falls and hip fracture in Hong Kong Chinese. Public Health 104, 117 – 121 (1990). R. Tideiksaar, Geriatric falls in the home. Home Healthc. Nurse 4, 14 – 23 (1986). L. Z. Rubenstein, A. S. Robbins, B. L. Schulman, J. Rosado, D. Osterweil, and K. R. Josephson, Falls and instability in the elderly. J. Am. Geriatr. Soc. 36, 266 – 278 (1988). B. E. Maki, P. J. Holliday, and A. K. Topper, Fear of falling and postural performance in the elderly. J. Gerontol. 46, M123 – M131 (1991). M. C. Hornbrook, V. J. Stevens, D. J. Wingfield, J. F. Hollis, M. R. Greenlick, and M. G. Ory, Preventing falls among communitydwelling older persons: Results from a randomized trial. Gerontologist 34, 16 – 23 (1994). R. Tideiksaar, “Falling in Old Age: Prevention and Management.” Springer, New York (1997). L. Clemson, R. G. Cumming, and M. Roland, Case-control study of hazards in the home and risk of falls and hip fractures. Age Ageing 25, 97 – 101 (1996). S. Studenski, P. W. Duncan, J. Chandler, et al., Predicting falls: The role of mobility and nonphysical factors. J. Am. Geriatr. Soc. 42, 297 – 302 (1994). A. J. Campbell, M. J. Borrie, G. F. Spears, S. L. Jackson, J. S. Brown, and J. L. Fitzgerald, Circumstances and consequences of
CHAPTER 32
81.
82.
83.
84.
85. 86. 87.
88. 89. 90. 91.
92.
93.
94. 95.
96.
97.
98.
99.
100.
101.
Falls as Risk Factors for Fractures
falls experienced by a community population 70 years and over during a prospective study [published erratum appears in Age Ageing 19(5), 345 – 346 (1990)]. Age Ageing 19, 136 – 141 (1990). M. Speechley and M. Tinetti, Falls and injuries in frail and vigorous community elderly persons. J. Am. Geriatr. Soc. 39, 46 – 52 (1991). M. E. Northridge, M. C. Nevitt, J. L. Kelsey, and B. Link, Home hazards and falls in the elderly: The role of health and functional status. Am. J. Public Health 85, 509 – 515 (1995). M. E. Northridge, M. C. Nevitt, and J. L. Kelsey, Non-syncopal falls in the elderly in relation to home environments. Osteopor. Int. 6, 249 – 255 (1996). M. J. Ashley, C. I. Gryfe, and A. Amies, A longitudinal study of falls in an elderly population II. Some circumstances of falling. Age Ageing 6, 211 – 220 (1977). N. Coyle Cancer audit. A problem-focused approach to nursing audit: Patient falls. Cancer Nurs. 2, 389 – 391 (1979). J. Foerster, A study of falls: The elderly nursing home resident. J. N. Y. State Nurses Assoc. 12, 9 – 17 (1981). A. Halpert and J. P. Connors, Prevention of patient falls through perceived control and other techniques. Law Med. Health Care 14, 20 – 4, 12 (1986). N. G. Rainville, Effect of an implemented fall prevention program on the frequency of patient falls. QRB Qual. Rev. Bull. 10, 287–291 (1984). G. M. Tinker, Accidents in a geriatric department. Age Ageing 8, 196 – 198 (1979). M. E. Tinetti and M. Speechley, Prevention of falls among the elderly. N. Engl. J. Med. 320, 1055 – 1059 (1989). Y. T. Gross, Y. Shimamoto, C. L. Rose, and B. Frank, Why do they fall? Monitoring risk factors in nursing homes. J. Gerontol. Nurs. 16, 20 – 25 (1990). L. C. Mion, J. D. Frengley, C. A. Jakovcic, and J. A. Marino, A further exploration of the use of physical restraints in hospitalized patients [published erratum appears in J. Am. Geriatr. Soc. 38(2), 181 (1990)]. J. Am. Geriatr. Soc. 37, 949 – 956 (1989). E. Capezuti, L. Evans, N. Strumpf, and G. Maislin, Physical restraint use and falls in nursing home residents. J. Am. Geriatr. Soc. 44, 627 – 633 (1996). E. Capezuti, N. Strumpf, L. Evans, and G. Maislin, Gerontologist 38(Spec. No.1), 226 (1998). [Abstract] L. K. Evans and N. E. Strumpf, Tying down the elderly: A review of the literature on physical restraint. J. Am. Geriatr. Soc. 37, 65 – 74 (1989). D. M. Buchner, M. E. Cress, E. H. Wagner, B. J. de Lateur, R. Price, and I. B. Abrass, The Seattle FICSIT/MoveIt study: The effect of exercise on gait and balance in older adults. J. Am Geriatr. Soc. 41, 321 – 325 (1993). D. N. Kanten, C. D. Mulrow, M. B. Gerety, M. J. Lichtenstein, C. Aguilar and J. E. Cornell, Falls: An examination of three reporting methods in nursing homes. J. Am. Geriatr. Soc. 41, 662 – 666 (1993). F. K. Ejaz, J. A. Jones, and M. S. Rose, Falls among nursing home residents: An examination of incident reports before and after restraint reduction programs. J. Am. Geriatr. Soc. 42, 960 – 964 (1994). L. K. Evans, N. E. Strumpf, S. L. Allen-Taylor, E. Capezuti, G. Maislin, and B. Jacobsen, A clinical trial to reduce restraints in nursing homes. J. Am. Geriatr. Soc. 45, 675 – 681 (1997). E. Capezuti, N. Strumpf, L. Evans, and G. Maislin, Outcomes of nighttime physical restraint removal for severely impaired nursing home residents. Am. J. Alzheim. Dis. 14, 157 – 164 (1999). E. Capezuti, N. E. Strumpf, L. K. Evans, J. A. Grisso, and G. Maislin, The relationship between physical restraint removal and falls and injuries among nursing home residents. J. Gerontol. A Biol. Sci. Med. Sci. 53, M47 – M52 (1998).
805 102. M. E. Tinetti, Factors associated with serious injury during falls by ambulatory nursing home residents. J. Am. Geriatr. Soc. 35, 644 – 648 (1987). 103. S. F. Ginter and L. C. Mion, Falls in the nursing home: Preventable or inevitable? J. Gerontol. Nurs. 18, 43 – 48 (1992). 104. M. E. Tinetti, J. Doucette, E. Claus, and R. Marottoli, Risk factors for serious injury during falls by older persons in the community. J. Am. Geriatr. Soc. 43, 1214 – 1221 (1995). 105. C. I. Gryfe, A. Amies, and M. J. Ashley, A longitudinal study of falls in an elderly population. I. Incidence and morbidity. Age Ageing 6, 201 – 210 (1977). 106. J. C. Brocklehurst, A. N. Exton-Smith, B. S. Lempert, L. P. Hunt, and M. K. Palmer, Fracture of the femur in old age: A two-centre study of associated clinical factors and the cause of the fall. Age Ageing 7, 2 – 15 (1978). 107. G. Berry, R. H. Fisher, and S. Lang, Detrimental incidents, including falls, in an elderly institutional population. J. Am. Geriatr. Soc. 29, 322 – 324 (1981). 108. J. M. Mossey, Social and psychologic factors related to falls among the elderly. Clin. Geriatr. Med. 1, 541 – 553 (1985). 109. R. G. Cumming, and R. J. Klineberg, Fall frequency and characteristics and the risk of hip fractures. J. Am. Geriatr. Soc. 42, 774 – 778 (1994). 110. S. R. Cummings and M. C. Nevitt, A hypothesis: The causes of hip fractures. J. Gerontol. 44, M107 – M111 (1989). 111. W. C. Hayes, E. R. Myers, J. N. Morris, T. N. Gerhart, H. S. Yett, and L. A. Lipsitz, Impact near the hip dominates fracture risk in elderly nursing home residents who fall. Calcif. Tissue Int. 52, 192 – 198 (1993). 112. M. C. Nevitt, S. R. Cummings, et al., Type of fall and risk of hip and wrist fractures: The study of osteoporotic fractures. J. Am. Geriatr. Soc. 41, 1226 – 1234 (1993). 113. S. L. Greenspan, E. R. Myers, L. A. Maitland, N. M. Resnick, and W. C. Hayes, Fall severity and bone mineral density as risk factors for hip fracture in ambulatory elderly. JAMA 271, 128 – 133 (1994). 114. J. A. Grisso, A. R. Wishner, D. F. Schwarz, B. A. Weene, J. H. Holmes, and R.L. Sutton, A population-based study of injuries in inner-city women. Am. J. Epidemiol. 134, 59 – 68 (1991). 115. M. E. Tinetti, J. T. Doucette, and E. B. Claus, The contribution of predisposing and situational risk factors to serious fall injuries. J. Am. Geriatr. Soc. 43, 1207 – 1213 (1995). 116. K. G. Faulkner, S. R. Cummings, D. Black, L. Palermo, C. C. Gluer, and H. K. Genant, Simple measurement of femoral geometry predicts hip fracture: The Study of Osteoporotic Fractures. J. Bone Miner. Res. 8, 1211 – 1217 (1993). 117. T. E. Cohn and D. J. Lasley, Visual depth illusion and falls in the elderly. Clin. Geriatr. Med. 1, 601 – 620 (1985). 118. C. Cooper, D. J. Barker, and C. Wickham, Physical activity, muscle strength, and calcium intake in fracture of the proximal femur in Britain. Br. Med. J. 297, 1443 – 1446 (1988). 119. D. T. Felson, J. J. Anderson, M. T. Hannan, R. C. Milton, P. W. Wilson, and D. P. Kiel, Impaired vision and hip fracture. The Framingham Study. J. Am. Geriatr. Soc. 37, 495 – 500 (1989). 120. J. A. Grisso, J. L. Kelsey, B. L. Strom, L. A. O’Brien, G. Maislin, K. LaPann, L. Samelson, and S. Hoffman, Risk factors for hip fracture in black women: The Northeast Hip Fracture Study Group. N. Engl. J. Med. 330, 1555 – 1559 (1994). 121. J. A. Grisso, J. L. Kelsey, L. A. O’Brien, C. G. Miles, S. Sidney, G. Maislin, K. LaPann, D. Moritz, and B. Peters, Risk factors for hip fracture in men. Hip Fracture Study Group. Am. J. Epidemiol. 145, 786 – 793 (1997). 122. A. Paganini-Hill, R. K. Ross, V. R. Gerkins, B. E. Henderson, M. Arthur, and T. M. Mack, Menopausal estrogen therapy and hip fractures. Ann. Intern. Med. 95, 28 – 31 (1981). 123. R. Wootton, E. Bryson, U. Elsasser, H. Freeman, J. R. Green, R. Hesp, E. A. Hudson, L. Klenerman, T. Smith, and J. Zanelli, Risk
806
124.
125.
126.
127.
128. 129.
130.
131.
132.
133.
134.
135.
136.
137.
138.
139.
SCHWARTZ, CAPEZUTI, AND GRISSO factors for fractured neck of femur in the elderly. Age Ageing 11, 160 – 168 (1982). A. V. Schwartz, J. L. Kelsey, S. Sidney, and J. A. Grisso, Characteristics of falls and risk of hip fracture in elderly men. Osteopor. Int. 8, 240 – 246 (1998). J. A. Grisso, J. L. Kelsey, B. L. Strom, G. Y. Chiu, G. Maislin, L. A. O’Brien, S. Hoffman, and F. Kaplan, Risk factors for falls as a cause of hip fracture in women: The Northeast Hip Fracture Study Group. N. Engl. J. Med. 324, 1326 – 1331 (1991). B. E. Maki and G. R. Fernie, Impact attenuation of floorcoverings in simulated falling accidents. Appl. Ergonom. 21, 107 – 114 (1990). S. R. Cummings, M. C. Nevitt, W. S. Browner, K. Stone, K. M. Fox, K. E. Ensrud, J. Cauley, D. Black, and T. M. Vogt, Risk factors for hip fracture in white women: Study of Osteoporotic Fractures Research Group. N. Engl. J. Med. 332, 767 – 773 (1995). L. J. Melton and B. L. Riggs, Risk factors for injury after a fall. Clin. Geriatr. Med. 1, 525 – 536 (1985). M. E. Farmer, L. R. White, J. A. Brody, and K. R. Bailey, Race and sex differences in hip fracture incidence. Am. J. Public. Health 74, 1374 – 1380 (1984). L. D. Gillespie, W. J. Gillespie, R. Cumming, S. E. Lamb, and B. H. Rowe, Interventions for preventing falls in the elderly (Cochrane Review). In “The Cochrane Library,” 3, Ed., Update Software, Oxford, (1999). M. E. Tinetti, D. I. Baker, P. A. Garrett, M. Gottschalk, M. L. Koch, and R. I. Horwitz, Yale FICSIT: Risk factor abatement strategy for fall prevention. J. Am. Geriatr. Soc. 41, 315 – 320 (1993). M. E. Tinetti, D. I. Baker, G. McAvay, E. B. Claus, P. Garrett, M. Gottschalk, M. L. Koch, K. Trainor, and R. I. Horwitz, A multifactorial intervention to reduce the risk of falling among elderly people living in the community . N. Engl. J. Med. 331, 821 – 827 (1994). M. E. Tinetti, G. McAvay, and E. Claus, Does multiple risk factor reduction explain the reduction in fall rate in the yale ficsit trial? Frailty and injuries cooperative studies of intervention techniques. Am. J. Epidemiol. 144, 389 – 399 (1996). J. Close, M. Ellis, R. Hooper, E. Glucksman, S. Jackson, and C. Swift, Prevention of falls in the elderly trial (PROFET): A randomised controlled trial. Lancet 353, 93 – 97 (1999). W. A. Ray, J. A. Taylor, K. G. Meador, P. B. Thapa, A. K. Brown, H. K. Kajihara, C. Davis, P. Gideon, and M. R. Griffin, A randomized trial of a consultation service to reduce falls in nursing homes. JAMA 278, 557 – 562 (1997). S. L. Wolf, H. X. Barnhart, G. L. Ellison, and C. E. Coogler, The effect of Tai Chi Quan and computerized balance training on postural stability in older subjects. Atlanta FICSIT Group. Frailty and Injuries: Cooperative Studies on Intervention Techniques. Phys. Ther. 77, 371 – 81; discussion 382 – 384 (1997). D. M. Buchner, M. E. Cress, B. J. de Lateur, P. C. Esselman, A. J. Margherita, R. Price, and E. H. Wagner, The effect of strength and endurance training on gait, balance, fall risk, and health services use in community-living older adults. J. Gerontol. A Biol. Sci. Med. Sci. 52, M218 – M224 (1997). A. J. Campbell, M. C. Robertson, M. M. Gardner, R. N. Norton, M. W. Tilyard, and D. M. Buchner, Randomised controlled trial of a general practice programme of home based exercise to prevent falls in elderly women. Br. Med. J. 315, 1065 – 1069 (1997). M. A. Province, E. C. Hadley, M. C. Hornbrook, L. A. Lipsitz, J. P. Miller, C. D. Mulrow, M. G. Ory, R. W. Sattin, M. E. Tinetti, and S. L. Wolf, The effects of exercise on falls in elderly patients: A
140.
141.
142.
143.
144.
145.
146.
147. 148.
149.
150.
151.
152.
153.
154.
155.
preplanned meta-analysis of the FICSIT Trials. Frailty and Injuries: Cooperative Studies of Intervention Techniques. JAMA 273, 1341 – 1347 (1995). M. E. McMurdo, P. A. Mole, and C. R. Paterson, Controlled trial of weight bearing exercise in older women in relation to bone density and falls. Br. Med. J. 314, 569 (1997). S. R. Lord, J. A. Ward, P. Williams, and M. Strudwick, The effect of a 12-month exercise trial on balance, strength, and falls in older women: A randomized controlled trial. J. Am. Geriatr. Soc. 43, 1198 – 1206 (1995). S. Reinsch, P. MacRae, P. A. Lachenbruch, and J. S. Tobis, Attempts to prevent falls and injury: A prospective community study. Gerontologist 32, 450 – 456 (1992). C. D. Mulrow, M. B. Gerety, D. Kanten, J. E. Cornell, L. A. DeNino, L. Chiodo, C. Aguilar, M. B. O’Neil, J. Rosenberg, and R. M. Solis, A randomized trial of physical rehabilitation for very frail nursing home residents. JAMA 271, 519 – 524 (1994). M. A. Fiatarone, E. F. O’Neill, N. D. Ryan, K. M. Clements, G. R. Solares, M. E. Nelson, S. B. Roberts, J. J. Kehayias, L. A. Lipsitz, and W. J. Evans, Exercise training and nutritional supplementation for physical frailty in very elderly people. N. Engl. J. Med. 330, 1769 – 1775 (1994). A. J. Campbell, M. C. Robertson, M. M. Gardner, R. N. Norton, and D. M. Buchner, Psychotropic medication withdrawal and a homebased exercise program to prevent falls: A randomized, controlled trial. J. Am. Geriatr. Soc. 47, 850 – 853 (1999). S. M. Willcox, D. U. Himmelstein, and S. Woolhandler, Inappropriate drug prescribing for the community-dwelling elderly. JAMA 272, 292 – 296 (1994). P. J. Greco and J. M. Eisenberg, Changing physicians’ practices. N. Engl. J. Med. 329, 1271 – 1273 (1993). W. A. Ray, J. A. Taylor, K. G. Meador, M. J. Lichtenstein, M. R. Griffin, R. Fought, M. L. Adams, and D. G. Blazer, Reducing antipsychotic drug use in nursing homes: A controlled trial of provider education. Arch. Intern. Med. 153, 713 – 721 (1993). J. Avorn, S. B. Soumerai, D. E. Everitt, D. Ross-Degnan, M. H. Beers, D. Sherman, S. R. Salem-Schatz, and D. Fields, A randomized trial of a program to reduce the use of psychoactive drugs in nursing homes . N. Engl. J. Med. 327, 168 – 173 (1992). R. I. Shorr, R. L. Fought, and W. A. Ray, Changes in antipsychotic drug use in nursing homes during implementation of the OBRA-87 regulations . JAMA 271, 358 – 362 (1994). E. L. Siegler, E. Capezuti, G. Maislin, M. Baumgarten, L. Evans, and N. Strumpf, Effects of a restraint reduction intervention and OBRA ‘87 regulations on psychoactive drug use in nursing homes . J. Am. Geriatr. Soc. 45, 791 – 796 (1997). M. C. Hornbrook, V. J. Stevens, and D. J. Wingfield, Seniors’ Program for Injury Control and Education. J. Am. Geriatr. Soc. 41, 309 – 314 (1993). P. Dargent-Molina, F. Favier, H. Grandjean, C. Baudoin, A. M. Schott, E. Hausherr, P. J. Meunier, and G. Breart, Fall-related factors and risk of hip fracture: The EPIDOS prospective study. epidemiologie de l’osteoporose. Lancet 348, 145 – 149 (1996). L. Z. Rubenstein, A. S. Robbins, K. R. Josephson, B. L. Schulman, and D. Osterweil, The value of assessing falls in an elderly population: A randomized clinical trial. Ann. Intern. Med. 113, 308 – 316 (1990). E. H. Wagner, A. Z. Lacroix, L. Grothaus, S. G. Leveille, J. A. Hecht, K. Artz, K. Odle, and D. M. Buchner, Preventing disability and falls in older adults: A population-based randomized trial. Am. J. Public Health 84, 1800 – 1806 (1994).
CHAPTER 32
Falls as Risk Factors for Fractures
156. D. F. Schwarz, J. A. Grisso, C. Miles, J. H. Holmes, and R. L. Sutton, An injury prevention program in an urban African-American community. Am. J. Public Health 83, 675 – 680 (1993). 157. D. M. Black, S. R. Cummings, D. B. Karpf, J. A. Cauley, D. E. Thompson, M. C. Nevitt, D. C. Bauer, H. K. Genant, W. L. Haskell, R. Marcus, S. M. Ott, J. C. Torner, S. A. Quandt, T. F. Reiss, and K. E. Ensrud, Randomised trial of effect of alendronate on risk of fracture in women with existing vertebral fractures: Fracture Intervention Trial Research Group. Lancet 348, 1535 – 1541 (1996). 158. S. R. Cummings, D. M. Black, D. E. Thompson, W. B. Applegate, E. Barrett-Connor, T. A. Musliner, L. Palermo, R. Prineas, S. M. Rubin, J. C. Scott, T. Vogt, R. Wallace, A. J. Yates, A. Z. Lacroix, Effect of alendronate on risk of fracture in women with low bone density but without vertebral fractures: Results from the Fracture Intervention Trial. JAMA 280, 2077 – 2082 (1998). 159. M. H. Komulainen, H. Kroger, M. T. Tuppurainen, A. M. Heikkinen, E. Alhava, R. Honkanen, and S. Saarikoski, HRT and Vit D in prevention of non-vertebral fractures in postmenopausal women; a 5 year randomized trial. Maturitas 31, 45 – 54 (1998). 160. B. Ettinger, D. M. Black, B. H. Mitlak, R. K. Knickerbocker, et al., Reduction of vertebral fracture risk in postmenopausal women with osteoporosis treated with raloxifene: Results from a 3-year randomized clinical trial. JAMA 282, 637 – 645 (1999)
807 161. B. Dawson-Hughes, G. E. Dallal, E. A. Krall, L. Sadowski, N. Sahyoun, and S. Tannenbaum, A controlled trial of the effect of calcium supplementation on bone density in postmenopausal women. N. Engl. J. Med. 323, 878 – 883 (1990). 162. M. C. Chapuy, M. E. Arlot, F. Duboeuf, J. Brun, B. Crouzet, S. Arnaud, P. D. Delmas, and P. J. Meunier, Vitamin D3 and calcium to prevent hip fractures in the elderly women. N. Engl. J. Med. 327, 1637 – 1642 (1992). 163. J. B. Lauritzen, M. M. Petersen, and B. Lund, Effect of external hip protectors on hip fractures. Lancet 341, 11 – 13 (1993). 164. S. N. Robinovitch, W. C. Hayes, and T. A. McMahon, Energy-shunting hip padding system attenuates femoral impact force in a simulated fall. J. Biomech. Eng. 117, 409 – 413 (1995). 165. P. Cavanagh, Saving frequent fallers. The New York Times Magazine (1994). 166. E. Capezuti, K. A. Talerico, N. Strumpf, and L. Evans, Individualized assessment and intervention in bilateral siderail use. Geriatr. Nurs. 19, 322 – 330 (1998). 167. E. Capezuti, K. A. Talerico, L. Wagner, N. Strumpf, H. Becker, and G. Maislin, The effect of a low-height bed intervention on nightfalls among frail nursing home residents. Gerontologist 39 (Special issue 1), 196.
CHAPTER 33
Assessing Fracture Risk CHARLES W. SLEMENDA,† C. CONRAD JOHNSTON, AND SIU L. HUI Department of Medicine, Indiana University School of Medicine and the Regenstrief Institute for Health Care, Indianapolis, Indiana 46202
I. II. III. IV.
Introduction Skeletal Elements of Fracture Risk Nonskeletal Elements of Fracture Risk Assessment Measuring Bone Loss for Risk Assessment: Necessary or Useful?
V. Monitoring Bone Mass VI. The Expression of Risk VII. Summary References
I. INTRODUCTION
Thus, there will always be fractures that occur among people evaluated as being at low risk, and some at the highest risk will avoid fracture (primarily through the avoidance of falls). Second, there are events, some identifiable and some not (e.g., rapid bone loss, neurological disorders), that will substantially increase a patient’s risk after the initial assessment, but the high frequency of monitoring of bone mass and other risk factors necessary to detect such changes is impractical and potentially very expensive. It is also, for the majority of patients, uninformative. Third, the acceptable balance between the costs associated with diagnosis and treatment of osteoporosis and the prevention of the suffering and costs associated with osteoporotic fractures is variable, both as perceived by patients and as perceived by providers of health care. Whereas measurements of all skeletal sites every few months might provide virtually perfect understanding of this aspect of risk, it would be prohibitively expensive and would improve fracture prevention marginally, if at all. Also, it would increase cost:benefit ratios greatly. What follows is an evaluation of both skeletal and nonskeletal factors that influence risk for osteoporotic fractures, and of the consequences associated with various courses of action.
The identification of patients at highest risk of suffering osteoporotic fractures is an issue of both clinical and scientific importance. Risk factors for fracture include not only low bone mass and associated defects of macroand microarchitecture, but also factors that affect risk of falling and other as yet poorly understood mechanisms. From a scientific point of view, all of these elements deserve attention, whereas for clinical purposes a more narrow approach addressing factors influencing the interpretation of alterable risks may be appropriate. However, the divergence between these approaches is small, as will be shown. Inherent in the clinical approaches to identification of risks and the prevention of osteoporotic fractures are several factors that make this aspect of the practice of clinical medicine difficult. First, there is no level of bone mass and no identifiable phenotype that reduces fracture risk to zero.
†
Deceased.
OSTEOPOROSIS, SECOND EDITION VOLUME 1
809
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
810
SLEMENDA, JOHNSTON, AND HUI
II. SKELETAL ELEMENTS OF FRACTURE RISK A. Bone Mass Low bone mass increases the risk of osteoporotic fractures, and this risk is independent of the risk associated with increasing age [1 – 3]. The earliest prospective studies of bone mass and fractures showed that appendicular measurements of both bone mineral content (BMC) and bone mineral density (BMD) could identify those women at the highest risk of fractures and that the risk of nonspine fractures increased approximately 1.5- to 2-fold for each standard deviation (SD) decline in bone mass at the radius or calcaneus [3 – 6]. Furthermore, there was no evidence for a threshold effect because risk increased progressively with declining BMD. Later studies of axial skeletal sites, e.g., the hip and spine, demonstrated similar predictive value for these measurements [3,7,8]. A comprehensive review [9] concluded that all measurement sites had similar predictive abilities (relative risk, RR, about 1.5 per SD decrease in BMD) except for spine measurements predicting spine (relative risk of 2.3) [7] and hip measurements predicting hip (relative risk of 2.6) [8]. A measurement of BMD at any skeletal site provides information regarding bone mass and associated fracture risk that cannot be otherwise obtained. Numerous studies have shown that clinical risk factors, although significantly associated with bone mass, misclassify more than one-third of women with low bone mass [10], although some risk factors may influence fracture risk through mechanisms other than bone mass. In the absence of bone mass measurements, risk factor information may be useful, but this is an inferior approach to the classification of fracture risk. For the identification of patients with low bone mass for whom therapies to preserve bone mass are being considered, only direct measurement of bone mass provides accurate and specific data for assessment of this risk. It is also necessary to distinguish between risk factors for low bone mass and risk factors for fractures, which may differ, and in rare cases work in opposite directions, as discussed later.
B. Which Skeletal Site(s) Should Be Measured? Bone mass can be measured at central sites (spine and hip) or peripheral sites (forearm, heel, etc.). Almost any measurement predicts fractures at any site, but generally measurements at a given site best predict fractures at the same site. The ideal bone mass measurement would produce the largest gradient of fracture risk from high to low bone mass and would classify patients with the least error. If only one measurement could be made, a scan of the proximal femur appears to approach most closely this ideal for
hip fractures. First, the higher relative risk of hip fractures associated with declines in BMD at this site (RR = 2.6 per SD difference in BMD) confers a considerable advantage in terms of risk gradient compared with measurements of other sites for which relative risks have been found to be 1.5 – 2.0. For example, given these relative risks, at any given age, a woman in the 10th percentile of proximal femur BMD (RR = 2.6) has 11.5 times the fracture risk of a woman with BMD in the 90th percentile (1.28 SD above the mean vs 1.28 SD below the mean); for the lower RR of 2 associated with measurements at other sites, this gradient is only six-fold. Second, although for cross-sectional measurements the error at the hip is slightly higher than at the spine, this is trivial relative to the increased risk gradient. Further, spine measurements are associated with other problems, such as the possibility of deformed of fractured vertebrae in the scan area, the development of osteophytes, and other artifacts among the elderly. Fortunately, there are fewer problems with artifacts in measurements of the spine for women between 50 and 65. Clinically, spine BMD is therefore most useful in perimenopausal women who might experience rapid bone loss due to the abrupt changes in sex hormone concentrations. Peripheral measurements have been gaining popularity because of the lower cost and portability of the instruments. Generally, a low peripheral measurement would trigger a central measurement for consideration of treatment and establishing a baseline for subsequent monitoring. However, the trigger point for optimizing cost effectiveness has not been established for any of the peripheral devices. Besides the choice of skeletal sites, the choice of instrument for measuring bone mass is important. Dual-energy X-ray absorptiometry (DXA) is the most popular method for measuring central sites; its high precision and ability to predict fractures are well established. Recently, more portable DXA scanners for measuring peripheral sites are also available; these instruments demonstrate adequate precision but their ability to predict fractures needs to be established. Also gaining popularity are quantitative ultrasound (QUS) measurements of the calcaneus; the scanners are portable and do not involve radiation. QUS has been shown to predict hip fractures almost as well as hip measurements [11 – 13]. In addition, it improves fracture prediction even after accounting for BMD measured by DXA. This observation has led some people to believe that QUS may provide a measure of bone “quality” not captured by the quantitative measurement of BMD. Finally, advocates of quantitative computed tomography (QCT) prefer this measurement because it measures volumetric density in the vertebrae and, more recently, at the hip [14]. However, data for predicting fracture based on QCT measurements are not as extensive as for DXA measurements. Furthermore, the higher level of radiation makes QCT less desirable in clinical settings.
811
CHAPTER 33 Assessing Fracture Risk
Is the measurement of a single site adequate for decisions regarding the initiation of therapy? Yes and no. Extensive clinical experience and many studies have shown that the correspondence in BMD among skeletal sites is modest, and thus some patients with adequate bone mass at one site will have low bone mass at another. Given the lack of prospective data for predicting fractures, low BMD at a peripheral site should be corroborated by a central measurement. The moderate correlation between these two sites [15,16] would suggest that measurements of both the spine and the hip are needed. It has been argued that, from a practical point of view, the incremental costs associated with a second measurement of a patient already on a scanning table are very small, and thus these measurements should be done then. However, for the measurement of a second site to have value it should be shown to improve fracture prediction. Data from the Study of Osteoporotic Fractures (SOF) have important implications in this regard [17]. Using a cut off value of 2.5 SD from the mean for young normal women, a hip measurement alone yielded a higher gradient of risk (the ratio of fractures in the low relative to the high bone mass group equaled about 5) than using both hip and spine measurements (i.e., for either site below 2.5 SD, the fracture ratio was about 2.6). These data probably reflect in part the increasing prevalence of arthritis-related artifacts in spine scans of this older, mostly 65- to 74-yearold, population, which diminishes the value of spine measurements for older people. In contrast, this and other studies, such as the EPIDOS study in Europe, find that the combination of DXA and QUS measurements best predicts fractures [18]. For which patients is one bone mass measurement enough? First, those with low bone mass detected in their first scan to not need a second baseline scan. These individuals will be treated, and for the treatment to be considered effective it will require that this first site respond to the chosen therapy. It should be noted that some therapies may have differing effects depending on skeletal site; few data exist, however, regarding clinical approaches to this issue. It is obvious that patients should be measured at least at the site expected to respond favorably to therapy. Second, those with very high proximal femur bone mass, e.g., / 1 SD compared with young normal values, have a very low probability of having low BMD at the spine, probably do not need a second baseline scan at some other site, and probably will not reach the treatment threshold until late in life. Third, individuals with intermediate BMD values but very poor risk factor profiles, as discussed later, can also probably be treated without further measurements. The fact that there are a number of factors that increase risk independently of BMD makes this feasible. There will then remain a small group with marginally low BMD values for the proximal femur and without an adequately informative risk factor profile to make a well-informed clinical
judgment. For these, a second measurement will be informative when the second site measured is below the treatment threshold. Other practical issues regarding the approach to this problem have been raised. Some assessment of the patients’ risk factors will have to be made prior to scanning so that the technicians performing scans will know whether one or two scans are needed. The use of more data than the scan itself will also require additional planning, as described later.
C. At What Level of Bone Mass Should Treatment Be Considered? Choosing a point at which to institute therapy to prevent fractures consists of balancing risks, costs, and benefits. This balance will surely differ depending on the patients’ and physicians’ views, and the largest differences between individual’s perceptions of the risk:benefit ratio will probably be in the perception of risks. Regarding benefits, from the standpoint of an individual patient, long-term (5 – 10 years or more) hormone replacement therapy will diminish fracture risk by about half [19], but if fracture risk is very low this may be perceived as a trivial change. From a public health perspective the most efficient use of therapy is in the treatment of the highest risk subjects. The risks associated with hormone replacement and other therapies can be viewed both objectively and as a matter of personal perspective. For hormone replacement therapy as an example, in addition to the clear benefits in terms of reduced fracture risk, there is the disputed benefit of reduced risk for coronary heart disease. There are side effects of treatment, such as bleeding and breast tenderness for some women, but the major concerns relate to cancer risks. In particular, breast cancer risks remain poorly defined but hold great importance for many women. Even small increases in breast cancer risk may be unacceptable for many patients. Thus, the recommendations for treatment cut offs that follow are obviously guidelines to be altered as patients and their physicians choose. The World Health Organization defines osteopenia as hip BMD of 1 to 2.5 SD relative to young normals and osteoporosis as more than 2.5 SD [20] below young normals. The number of standard deviations below young normals has now been widely accepted as the T score. The National Osteoporosis Foundation (NOF) guidelines recommend treatment for women with BMD T scores below 2 in the absence of osteoporosis risk factors (to be discussed later), and with T scores below 1.5 otherwise [21], Implementation of these guidelines, however, is not always straightforward because the same individual may have very different T scores based on different measurements of BMD.
812
SLEMENDA, JOHNSTON, AND HUI
In the United States, most manufacturers have adopted normal reference values for the hip from the Third National Health and Nutrition Survey (NHANES III) based on a large representative sample of the U. S. population. The means and standard deviations of DXA measurements at the hip between ages 20 and 29 years in the survey are adopted as young normal values. Although these values are available for men and women, white, black, and MexicanAmericans, it is not clear whether gender- and race-specific norms should be used for calculating T scores in assessing fracture risk. Fracture risk assessment based on other bone measurements is even more problematic because data from nationally representative samples do not exist. For the spine it appears that the manufacturers have similar databases to define “young normal.” The databases for establishing norms for peripheral measurements are generally smaller and more subject to biased selection of subjects. There is an initiative underway in the bone research community for a normative study in which a large number of representative subjects are measured on most of the instruments in the market. However, even if there is a large representative sample for establishing young normals for different instruments, inherent problems exist in defining T scores for other instruments or other skeletal sites in the same way as for DXA measurements at the hip. It has been shown that the prevalence of osteoporosis (based on T 2.5) in any age group varies greatly depending on which instrument the T score is based [22]. For example, at age 60, the prevalence of osteoporosis in white women ranges from 3% based on peripheral DXA at the heel to 50% based on spine QCT. This also implies that two equal T scores based on two instruments and/or skeletal sites do not reflect the same risk of fracture. In order to correct this problem, the National Osteoporosis Foundation, the International Society for Clinical Densitometry, and the American Society for Bone and Mineral Research have joined forces to derive T score equivalents for most densitometers in the market. In the near future, clinicians will be able to obtain measurements reported as T score equivalents on different instruments, and the same T score equivalents will reflect the same level of fracture risk. In the meantime, it is probably prudent to obtain a central measurement of BMD to establish a high risk of fracture prior to starting treatment.
D. Skeletal Structure Personal history of fracture as an adult is a strong predictor of future fractures [23 – 25]. The existence of a previous fracture since age 40 (or 50) confers a significant increase in risk (RR 1.5 to 2.0) above and beyond that associated with low bone mass [26]. In particular, those with existing vertebral deformities have an increased risk of
fractures of other vertebral bodies [23]. It is almost certain that part of this increase in risk results from local changes in load bearing due to the initial fracture, but the presence of a fracture at any site also increases the risk for fractures elsewhere, again independent of BMD. One implication of these data may be that a fracture also indicates structural defects not captured by bone mass, e.g., flaws in skeletal microarchitecture. Unfortunately, no technique can measure the quality of bone microarchitecture in vivo, other than possibly QUS. Therefore, at the time of assessment of fracture risk, a question regarding fracture history since age 40 (or 50) is essential and that an affirmative response should be considered to increase risk approximately twofold, similar to a 1 SD further decline in BMD; i.e., a patient with a positive fracture history would reach the treatment threshold earlier than a patient without. Studies have addressed the role of skeletal geometry, particularly in the proximal femur, in the etiology of hip fractures. Faulkner and colleagues [27] observed that a longer distance from the inner edge of the pelvic brim to the outer edge of the greater trochanter, as measured along the axis of the femoral neck, called hip axis length (HAL), was associated with an increased risk of hip Fractures, Duboeuf et al. [28] later found that HAL is associated with cervical but not trochanteric fractures of the hip. Other measurements have also been suggested [30]. Although broadly consistent, these studies differed in the elements of femoral geometry that contributed to the alteration of fracture risk [30]. Each of these will require further study before their practical application to hip fracture risk assessment is possible. It is also unclear whether hip geometry can be altered by any intervention.
III. NONSKELETAL ELEMENTS OF FRACTURE RISK ASSESSMENT A. Lifestyle Factors Although much research has focused on risk factors for hip fractures, available evidence suggests that the same risk factors apply to fractures at other sites. Large prospective studies, including SOF and EPIDOS, have published a number of risk factors for hip fractures [26,31]. While most studies have focused on hip fractures in old women, some studies have included all fractures in perimenopausal women and men. The elements of lifestyle considered generally include smoking, alcohol intake, caffeine and calcium consumption, and work and physical activities. Although moderate alcohol consumption is harmless or even beneficial, caffeine consumption has been associated with a small increase in fracture risk. The combined evidence from multiple studies suggests that smoking from menopause onward increases the risk of hip fracture by
813
CHAPTER 33 Assessing Fracture Risk
about half, and this increase cannot be totally explained by the earlier onset of menopause or by the reduction in body weight or circulating estrogen [32]. Physical activities generally are shown be protective against fractures, but there is no strong relationship between fractures and calcium intake beyond adolescence [33,34].
B. Health and Medications Besides the well-known risk factors of menopause and low estrogen, self-rated health has been shown to be a strong independent predictor of hip fractures [26]. A selfreported history of a prior fracture or a family history of fractures also increases the risk of future fractures. In multiple studies [31], fracture can be predicted by various functional assessments, including elevated resting pulse rates, inability to rise from a chair, slower gait speed, and difficulty in tandem walk. Visual acuity and depth perception, as well as dementia, were also associated with significant increases in fracture incidence in different studies. The use of long-acting benzodiazepines and other psychotropic drugs has been shown repeatedly to be associated with increased risk of falls [35] and, as a consequence, fractures [26].
C. Body Size The stereotype of the patient with osteoporosis is a frail, thin, short woman, and this description contains an element of truth, but fracture risk assessment requires closer examination. Body weight, apart from its effects on bone mass, has not been shown to affect fracture risk independently, despite plausible mechanisms for fatness to exert additional protective effects (e.g., soft tissue energy absorption during falls) [36]. The part of the weight made up of lean body mass may be correlated with stronger muscles that help maintain balance and prevent falls. Increase in weight since age 25 has also been suggested as protective, whereas weight loss predicts fractures [37]. The mechanism for this effect is unclear, although it is possible that women who lose weight also loss bone, and there are resulting microarchitectural deficits that weaken the femur. It is also possible that loss of muscle mass during weight loss weakens the balance and increases the propensity to fall.
IV. MEASURING BONE LOSS FOR RISK ASSESSMENT: NECESSARY OR USEFUL? A separate issue in the decision regarding how many and which sites to measure is the role of rapid bone loss,
particularly from the spine, in early postmenopausal women. Obviously, without a spine measurement, assessment of bone loss from the spine is impossible. The genesis of this concept arises both from theoretical considerations, which suggest that rapid bone loss may lead to perforations of trabeculae, and from the data of Christiansen and colleagues [38] suggesting that those women with the most rapid early postmenopausal bone loss have an increased risk of vertebral crush fractures and deformities, although without an increase in risk of nonspine fractures. The issues of assessing rates of bone loss, monitoring the effectiveness of therapies, and remeasurement of patients with initial measurements above but near the threshold for treatment are related and are discussed further later.
A. Bone Loss and Fracture Risk It is obvious that bone loss contributes to fracture risk through its effect on bone mass, but the importance of estimating rates of bone loss depends on whether loss itself contributes to risk beyond its effects on bone mass. Before proceeding, it should be noted that this discussion does not address whether a patient will ever require a second measurement of bone mass; this is generally true and is considered further, later in this chapter. The primary issue here is whether it is necessary to identify that group of women with adequate bone mass, but very rapid bone loss, who might, as a result of this loss, produce structural defects in bone that make it weaker than would be estimated based on its mass alone. In other words, apart from the interpretation of later bone mass measurements themselves, is there a reason to interpret the rate of bone loss in assessing fracture risk? Danish data showing that rates of loss predict vertebral fractures independent of bone mass are one source of evidence supporting this concept [38], and the finding that nonspine fractures predict spine fractures independently may also be thought to support such an idea [23]. If it is assumed that these data are critical in the assessment of spine (and perhaps other) fracture risks, how should this be approached? With the improved precision of densitometers it can be expected that with excellent quality control the standard deviation of repeated measurements of an individual at the same site will approach 0.01 g/cm2, i.e., something near a 1% coefficient of variation (SD/mean of repeated measures); for those with lower bone mass, precision is probably worse (due to problems with edge detection in low bone mass regions, and perhaps other factors). To address the issue of how measurement precision affects the estimation of rates of bone loss, assume, for example, that a group of women all are losing bone from some skeletal site at exactly 3% per year (relatively rapid bone loss of
814 the sort that might produce microarchitectural deficiencies). If two measurements of that site taken 1 year apart, 95% of the apparent rates of bone loss would be between 0.2 and 5.8% per year [based on var (A-B) varA varB], and thus, virtually all of these women would be classified correctly as losing bone mass, although half would have apparent rates of loss less than 3% per year and approximately 16% would have rates less than 1.5% per year. Thus, if one could be sure that only those people with substantial rates of loss were being measured, rates of bone loss could be estimated quite well with measurements after 1 year. Of course, if one had this knowledge, measurements would not be necessary. For more modest rates of loss or for machines with poorer precision (or when software changes occur), the misclassification errors would be markedly higher. For example, for a machine with 1.5% error (CV) the range of estimated rates of loss (for the 3% per year group) would be 1.2% to 7.2% per year. That is, even with substantial bone loss and very good instrument precision (1.5% CV), some women losing 3% per year would appear to be gaining bone. Where the errors in estimates of rates of bone loss become especially important is when no bone loss is occurring, e.g., with effective therapies. With zero bone loss, half of women will still be seen as losing bone due to variability in estimates of bone loss rates and some of these rates will be substantial (with 1.5% CV, approximately one-sixth of the women will appear to have bone loss greater than 2% per year at 1 year), but these errors, when expressed as loss per year, will be smaller with longer intervals between measurements. In contrast to the less realistic example of a sample of women where all are losing bone at 3% per year, measurements at 1-year intervals appear less valuable for the vast majority of patients. The ultimate implication of these data are that for women losing substantial amounts of bone (3% per year), measurements at 1-year intervals (with 1% CV) would allow most to be recognized as losing bone mass. Unfortunately, half of the women losing no bone at all would also be classified as losing bone mass. In practice, patients are a mix of these various groups that are very difficult to separate in the short term. With measurements at 2-year intervals, the overlap between these groups would be much smaller and would be smaller yet with greater intervals between measurements. The greatest danger in misclassification is probably in the failure to treat rapid bone loss, but with current techniques it is unlikely that measurements at intervals of 2 years or less would provide adequately precise data for this purpose. It should also be recognized that although there are a few women with rapid, persistent bone loss, rapid short-term bone loss is a poor predictor of who will lose bone rapidly
SLEMENDA, JOHNSTON, AND HUI
over a long period of time [39]. Thus, there will be far more women not losing bone who would be misclassified than the reverse. Although rapid bone loss is a problem deserving clinical attention, identification of these patients is very difficult. Hence, for women not initially treated, bone loss estimates are probably not necessary. Except in very rare circumstances, subsequent measurement of bone mass in these women should be interpreted simply on the basis of whether or not they fall below the treatment cutoff.
B. Biochemical Measurements for the Identification of Rapid Bone Losers and Nonresponders to Therapy It has been suggested that biochemical measurements may be useful in either monitoring responses to therapy or identifying those with rapid rates of bone loss. Generally, these biochemical markers have included those which reflect aspects of skeletal metabolism. Although some are thought to reflect primarily bone formation or osteoblast activity (e.g., osteocalcin) and others bone resorption or osteoclast activity [e.g., tartrate-resistant acid phosphatase (TRAP) or collagen cross-links], their primary value is in estimating rates of skeletal turnover. More rapid turnover, reflected by higher concentrations of any of these markers, is generally associated with more rapid rates of bone loss in perimenopausal women [40,41] and in subjects under treatment [42,43]. However, the existence of significant correlations with rates of bone loss does not necessarily imply clinical utility. A comprehensive review by the NOF concludes that bone turnover markers are valuable in research on metabolic bone diseases and in the clinical management of Paget’s disease. However, these measurements cannot be recommended for clinical use until substantial positive and negative predictive values can be established from more studies with fracture end points [41].
V. MONITORING BONE MASS A. Bone Mass: Whom Not to Measure Patients whose femoral neck bone mass is more than 1 SD above the young normal mean would need to lose bone at 0.03 g/cm2 for nearly 9 years to achieve a bone mass at least 1 SD below the young normal mean. This rate of loss (almost 4% per year) for a sustained period of time is an unlikely occurrence. These individuals are likely to remain in the lowest age-specific risk group and the efficiency both of therapies and of monitoring bone mass will be very poor.
815
CHAPTER 33 Assessing Fracture Risk
With the development of new drugs, therapeutic options are likely to increase in the next few years, although there will remain a group of individuals for whom there are no options available or acceptable. Although it has been suggested that bone mass measurements may be useful in motivating patients to more appropriate decisions regarding elements of fracture risk and the adherence to therapy, this has not been proven and would require further study before bone mass measurements could be recommended on this basis. Therefore, for patients without available therapeutic options, repeated bone mass measurements are not indicated.
B. Whom to Monitor and When Among the untreated with initial bone mass measurements above the treatment threshold, second and subsequent measurements should be made based on the probability that the patient would have lost enough bone to require therapy. However, the reassessment would not be done to estimate bone loss, but only to determine whether bone mass was now below the treatment level. If the chosen treatment level is some value relative to young normals, then this process is somewhat easier. Assuming mean rates of bone loss from the hip of 1 – 1.5% per year, approximately half of the patients with BMD 1 SD above the treatment threshold would reach this threshold in 7 – 10 years. Of course some would achieve this level sooner and some would never reach it. Those closer to the threshold would require second measurements sooner. However, although bone mass measurements must be the cornerstone of assessment when considering therapies that work through the preservation of bone mass, many other data remain that are of value in assessing fracture risk. If an untreated patient suffers a fracture, this would alter risk assessment in favor of earlier treatment. Similarly, some acute changes should prompt earlier reassessment. For example, prolonged illness associated with bed rest or forced immobility should be considered reasons for earlier reassessment, as would the use of medications known to influence fracture risk. In contrast, changes in risk factors that would diminish fracture risk should increase the interval between measurements.
C. Monitoring Therapy The quantitative analysis in Section IV,A also applies to the monitoring of therapy. Over the short term, it is difficult to determine whether an individual patient is responding to therapy because of the measurements error. Given misinformation on a patient’s rate of bone loss, the clinician can
actually be misled into making the wrong clinical decision, such as stopping or switching the treatment.
VI. THE EXPRESSION OF RISK A. Absolute and Relative Risk Fracture risk, as discussed thus far, is expressed as relative to young normals with no risk factors, but several other ways of expressing risk may be equally or more useful. Absolute risk, i.e., fractures per 1000 women per year, or lifetime fracture risk may be preferred. The probability of any event, e.g., a fracture, may be expressed in many ways. With 300,000 hip fractures among 250,000,000 Americans each year, the probability of a randomly chosen person suffering such a fracture is 0.0012, 12 chances in 10,000 each year. Improving on this estimate of risk was the purpose of the preceding discussion. Women of age 72 with bone mass near the median for their age and three to four additional risk factors suffered hip fractures at a rate of 56/10,000 per year in SOF. Extrapolating from the SOF data [24], women with similar bone mass who walked for exercise (relative hazard 0.7 compared with nonwalkers) would be expected to suffer fractures at a rate of 39/10,000 per year. Of course, this does not address the efficacy of recommending that women begin walking to change their risk profile, but it gives a rough estimate of how absolute risk differs depending on risk profile and bone mass. Absolute risks are small over any short time period. However, cumulative risk over a patient’s remaining life span can be considerable.
B. Lifetime Risk Melton and colleagues [44] explored the concept of lifetime fracture risk, and several other groups have published data on the probability of a fracture in one’s remaining lifespan. Approximately 40% of women suffer some sort of fracture between ages 50 and 85, and roughly one woman in six will experience a hip fracture [44]. How BMD and other potential risk factors affect lifetime fracture risk has been explored less thoroughly because of the lack of studies of sufficient duration and the assumptions required to extrapolate from short-term risk to lifetime risk. In one study where some subjects were followed for up to 15 years, 44% of those with radius BMD below 0.6 g/cm had experienced a nonspine fracture by 7.5 years follow-up, in contrast to 15% of those with bone mass above 0.6 g/cm. Melton and associates [45] have explored potential models for estimating lifetime fracture risk, but the lifetime risk estimates require
816 numerous assumptions [e.g., remaining years of life, longterm effects of risk factors that might change (e.g., BMD)] and the development of appropriate models. As an example, the projected decrease in future mortality increases the lifetime risk of hip fracture substantially [46]. Perhaps more importantly, the application of multiple risk factor data to lifetime risk estimates is more complex (although possible).
VII. SUMMARY Assessment of the risk for osteoporotic fractures is a complex issue. The prevention of fractures through selective application of therapeutic interventions aimed at preserving bone mass is intuitively appealing, but difficult to apply. However, several issues are clear. First, if therapy to prevent bone loss is being considered, then bone mass measurements are probably necessary for several reasons: (i) it is inappropriate to treat someone for low bone mass if bone mass is high, (ii) it is impossible to monitor the effectiveness of therapy without a baseline bone mass measurement, and (iii) there is currently no method of identifying those at either high or low risk of suffering fractures due to osteopenia without direct measurement of bone mass. Second, numerous factors appear to predict fracture risk independently of bone mass. Most of these factors have been studied in relation to hip fractures, but several seem to be associated with fractures at multiple sites. For example, a fracture after age 40 (or 50) is associated with a higher risk of other fractures later in life. In the absence of bone mass measurements, these additional risk factors could provide some basis for risk assessment, but their ideal use is in the context of a complete clinical assessment with bone mass measurements as a cornerstone. In this same vein, some women will have bone mass that is so low that interventions aimed at the preservation of bone mass may be of little value without additional protection (e.g., pads for the hips), and bone mass measurements are the only way of identifying such women. Finally, there is general agreement that fracture risk can be altered substantially with pharmacological intervention, especially with the newest generation of drugs coming onto the market. Precise application of bone mass measurements and additional risk factors will benefit from further study, but this should not cause delay in the use of these data now. It is inevitable that physicians and their patients will interpret the risks and benefits of therapies in ways appropriate to their own circumstances. However, there are now adequate data to move the assessment of fracture risk onto firmer scientific ground and to quantify some of the elements of risk. Whether a physician chooses to recommend treatment at 1 SD compared to young normal BMD or at the bottom one-third of patients at any given age, it is certain that those at highest risk are being
SLEMENDA, JOHNSTON, AND HUI
treated. Moreover, the increased information now available regarding risk factors can also be used to identify those individual at highest risk of suffering osteoporotic fractures with greater success.
References 1. S. L. Hui, C. W. Slemenda, and C. C. Johnston, Age and bone mass as predictors of fracture in a prospective study. J. Clin. Invest. 81, 1804 – 1809 (1988). 2. S. R. Cummings, D. M. Black, M. C. Nevitt et al., Appendicular bone density and age predict hip fracture in women. JAMA 262, 665 – 668 (1989). 3. L. J. Melton, E. J. Atkinson, W. M. O’Fallan, H. W. Wahner, and B. L. Riggs, Long-term fracture prediction by bone mineral assessed at different skeletal sites. J. Bone Miner. Res. 10, 1227 – 1233 (1993). 4. S. L. Hui, C. W. Slemenda, and C. C. Johnston, Baseline measurement of bone mass predicts fractures in white women, Ann. Intern. Med. 111, 355 – 361 (1989). 5. R. Wasnich, P. D. Ross, L. K. Heibrun, and J. M. Vogel, Prediction of postmenopausal fracture risk with use of bone mineral measurements. Am. J. Obstet. Gynecol. 153, 745 – 751 (1985). 6. P. Gardsell, O. Johnell, B. E. Nilsson, and B. Gullberg, Predicting various fragility fractures in women by forearm bone densitometry: A follow-up study. Calcif. Tissue Int. 52, 348 – 353 (1993). 7. P. D. Ross, R. D. Wasnich, and J. M. Vogel, Detection of prefracture osteopororis using bone mineral absorptiometry. J. Bone Miner. Res. 3, 1 – 11 (1998). 8. S. R. Cummings, D. M. Black, M. C. Nevitt et al., Bone density at various sites for the prediction of hip fractures. Lancet 341, 72 – 75 (19993). 9. D. Marshall,O. Johnell, and H. Wedel, Meta-analysis of how well measures of bone mineral density predict occurrence of osteoporotic fractures. Br. Med. J. 312, 1254 – 1259 (1996). 10. C. W. Slemenda, S. L. Hui, C. Longcope, H. Wellman, and C. C. Johnston, Predictors of bone mass in perimenopausal women: A prospective study of clinical data using photon absorptiometry. Ann. Intern. Med. 112, 96 – 101 (1990). 11. D. T. Baran, A.M. Kelly, A. Karellas, M. Gionet, M. Price, D. Leahey, S. Steuterman, B. Mcsherry, and J. Roche, Ultrasound attenuation of the os calsis in women with osteoporosis and hip fractures. Calcif. Tissue Int. 43, 138 – 142 (1998). 12. D. Hans, P. Dargent-Molina, A. M. Schoot, J. L. Sebert, C. Cormier, P. O. Kotzki, P. D. Delmas, J. M. Pouilles, G. Breart, and P. J. Meunier, Ultrasonoraphic heel measurements to predict hip fracture in elderly women: The EPIDOS prospective study. Lancet 308, 511 – 514 (1996). 13. C. F. Njeh, C. M. Boivin, and C. M. Langton, The role of ultrasound in the assessment of osteoporosis: A review. Osteopor. Int. 7(1), 7 – 22 (1997). 14. T. F. Lang, J. H. Keyak, M. W. Heitz, P. Augat, Y. Lu, A. Mathur, and H. K. Genant, Volumetric quantitative computed tomography of the proximal femur: Precision and relation to bone strength. Bone 21(1), 101 – 108 (1997). 15. K. Lai, M. Rencken, B. L. Drinkwater, and C. H. Chestnut III, Site of bone density measurement may affect therapy decision. Calcif. Tissue Int. 53, 225 – 228 (1993). 16. P. J. Ryan, G. M. Blake, R. Herd, J. Parker, and I. Foegelman, Spine and Femur BMD by DXA in patients with varying severity spinal osteoporosis. Calcif. Tissue Int. 52, 263 – 268 (1993).
CHAPTER 33 Assessing Fracture Risk 17. S. R. Cummings, D. M. Black, M. C. Nevitt, W. Browner, J. Cauley, K. Ensrud, H. K. Genant, L. Palermo, J. Scott, and T. M. Vogt, Bone density at various sites for prediction of hip fractures. Lancet 341, 72 – 75 (1993). 18. D. M. Black, S. R. Cummings, H. K. Genant, M. C. Nevitt, L. Palermo, and W. Browner, Axial and appendicular, bone density predict fractures in older women. J. Bone & Miner. Res. 7, 633 – 638 (1992). 19. A. Paganini-Hill, R. K. Ross, V. R. Gerkins, B. E. Henderson, M. Arthur, and T. M. Mack, Menopausal estrogen therapy and hip fractures. Ann. Intern. Med. 95, 28 – 31 (1981). 20. J. A. Kanis and WHO study group, “Assessment of Fracture Risk and Its Application to Screening for Postmenopausal Osteoporosis,” WHO technical report series 843, 1994. 21. NOF guideline, National Osteoporosis foundation, Washington, DC, 1998. 22. K. G. Faulkner, E. V. Stetten, and P. Miller, Discordance in patient classification using t-scores. J. Clin. Desitomet. 2, 343 – 350 (1999). 23. P. D. Ross, J. W. Davis, R. Epstein, and R. D. Wasnich, Pre-existing fractures and bone mass predict vertebral fracture incidence in women. Ann. Intern. Med. 114, 919 – 923 (1991). 24. S. R. Cummings, M. C. Nevitt, W. S. Browner, K. Stone, K. M. Fox, K. E. Ensrud, J. Cauley, D. Black, and T. M. Vogt, Risk factors for hip fracture in white women. N. Engl. J. Med. 322, 767 – 773 (1995). 25. O. Johnell, Risk factors, for osteoporosis fractures. In “Bone Densitometry and Osteoporosis” H. K. Genant, G. Guglielmi, M. Jergas, Springer Verlag, Berlin, 1998. 26. S. R. Cummings, M. C. Nevitt, W. S. Browner, K. Stone, K. M. Fox, K. E. Ensrud, J. Cauley, D. Black, and T. M. Vogt, Risk factors for hip fracture in white women. N. Eng1. J. Med. 332, 767 – 773 (1995). 27. K. G. Faulkner, S. R. Cummings, D. Black, L. Palermo, C. C. Gluer, and H. K. Genant. simple measurement of femoral geometry predicts hip fracture: The study of osteoporotic fractures. J. Bone Miner. Res. 8, 1211 – 1217 (1993). 28. F. Duboeuf, D. Hans, A. M. Schott, P. O. Kotzki, F. Favier, C. Marcelli, P. F. Meunier, and P. D. Delmas, Different morphometric and densitometric parameters perdict cervical and trochanteric hip fracture: The EPIDOS study. J. Bone Miner. Res. 12(11), 1995 – 1902 (1997). 29. T. Yoshikawa, C. H. Turner, M. Peacock, C. W. Slemenda, C. M. Weaver, D. Teegarden, P. Markwardt, and D. B. Burr, Geometric structure of the femoral neck measured using dula-energy x-ray absorptiometry. J. Bone Miner. Res. 9, 1053 – 1064 (1994). 30. C. C. Gluer, S. R. Cummings, A. Pressman, J. Li, K. Gluer, K. G. Faulkner, S. Grampp, and H. K. Genant, Prediction of hip fractures from pelvic radioraphs: The study of osteoporotic fractures. J.Bone Miner. Res. 9, 671 – 677 (1994). 31. P. Dargent-Molina, F. Favier, H. Grandjean, C. Baudoin, A. M. Schott, E. Hausherr, P. J. Meunier, and G. Breart, Fall-related factors and risk of hip fracture: The EPIDOS prospective study. Lancet 348, 416 (1996).
817 32. M. R. Law and A. K. Hackshaw, A meta-analysis of cigarette smoking, bone mineral density and risk of hip fracture: Recognition of major effects. Br. Med. J. 315, 841 – 846 (1997). 33. D. Feskanich, W. C. Willett, M. J. Stampfer, and G. A. Colditz, Milk dietary calcium and bone fractures in women: A 12-year prospective study. Am. J. Public Health 87, 992 – 997 (1997). 34. R. G. Cumming, S. R. Cummings, M. C. Nevitt, J. Scott, K. E. Ensrud, T. M. Vogt, and K. Fox, Calcium intake and fracture risk: Results from the study of osteoporotic fractures. Am. J. Epidemiol. 145, 926 – 934 (1997). 35. W. A. Ray, M. R. Griffen, and W. Downey, Benzodiazepines of long and short elimination half-life and the risk of hip fracture. JAMA 262, 3303 – 3307 (1989). 36. S. L. Greenspan, E. R. Myers, L. A. Maitland, N. M. Resnick, and W. C. Hayes, Fall severity and bone mineral, density as risk factors for hip fracture in ambulatory elderly. JAMA 271, 128 – 133 (1994). 37. K. E. Ensrud, J. Cauley, R. Lipschutz, and S. R. Cummings, Weight change and fractures in older women. Arch. Intern. Med. 157, 857 – 863 (1997). 38. C. Christiansen, M. A. Hansen, K. Overgaard, and B. J. Riis, Prediction of future fracture risk, In “Proceedings of the Fourth International symposium on Osteoporosis and Concensus Development Conference” (C. Christiansen and B.J. Riis, eds.), pp. 52 – 54. Handelstrykkeriet Aalborg Aps, Aalborg, Denmark, 1993. 39. S. L. Hui, C. W. Slemenda, and C. C. Johnston, The contribution of bone loss to postmenopausal osteoporosis. Osteopor. Int. 1, 30 – 34 (1990). 40. C. W. Slemenda, S. L. Hui, C. Longcope, and C. C. Johnston, Sex steroids and bone mass: A study of changes about the time of menopause. J. Clin. Invest. 80, 1261 – 1269 (1987). 41. A. C. Looker, D. C. Bauer, C. H. Chestnut III, C. M. Gundberg, M. C. Hochberg, G. Klee, M. Kleerekoper, N. B. Watts, and N. H. Bell, Clinical use of biochemical markers of bone remodeling. Current status and furture direction. Osteopor. Int. 11(6), 467 – 480 (2000). 42. P. Ravan, M Bidstrup, C. Christiasen, and H. Lawaetz, Prediction of long term response in bone mass by markers: 4-year results from the Danish cohort of the EPIC study. J. Bone Miner. Res. 14, S370 (1999). 43. R. Branton and D. A. Percival, Measurement of urinary CTx using osteosal in patients on long term fosamax therapy. J. Bone Miner. Res. 14, S162 (1999). 44. L. J. Melton et al., How many women have osteoporosis. J. Bone Miner. Res. 7, 1005 – 1010 (1992). 45. L. J. Melton, S. H. Kan, H. W. Wahner, and B. L. Riggs, Lifetime fracture risk: An approach to hip fracture risk assessment based on bone mineral density and age. J. Clin. Epidemiol. 41, 985 – 994 (1998). 46. A. Oden, A. Dawson, W. Dere, O. Johnell, B. Jonsson, and J. A. Kanis, Lifetime risk of hip fractures is underestimated. Osteopor. Int. 8(6), 599 – 603 (1998).
CHAPTER 34
Outcomes of Osteoporotic Fractures
I. II. III. IV.
GAIL A. GREENDALE
Department of Medicine, Division of Geriatrics, University of California, Los Angeles, School of Medicine, Los Angeles, California 90095
ELIZABETH BARRETT-CONNOR
Department of Family and Preventive Medicine, University of California, San Diego, La Jolla, California 90293
V. Outcomes of Hip Fractures VI. Conclusions References
Introduction Definitions of Functional Outcomes Related to Fracture Outcomes of Wrist Fractures Outcomes of Vertebral Fractures
I. INTRODUCTION
tasks cover a wide range of activities from simple self-care to complex occupational duties. Functional tasks are generally classified by the level of difficulty into basic (BADL), intermediate or instrumental (IADL), and advanced (AADL) activities of daily living. BADLs are the rudiments of personal care, such as eating, dressing, and bathing. IADLs constitute those activities required to maintain independent living, e.g., cooking, shopping, and transportation [2 – 4]. AADLs are elective, vary by individual, and may be important components of the maintenance of personal satisfaction and well-being [5,6]. Examples of AADLs include recreational and intellectual pursuits. Deterioration of functional capability at any of these levels often follows osteoporotic fracture. Other outcomes of osteoporotic fracture include declines in physical capabilities (e.g., walking, bending, and other movements required in daily life), increased social support requirements, worsened economic status, and diminished quality of life, including depression and deterioration in perceived health. These outcomes can have an impact on
The consequences of osteoporotic fracture, particularly of the hip, are often calculated in economic terms. In the United States alone, the annual hospital, nursing home, and lost wage cost attributable to hip fracture may be as high as $10 billion [1]. However, there are other important ramifications of osteoporotic fractures. The human costs of osteoporosis include diminution in functional status, health status, and independence. From the individual’s perspective, quality of life is threatened by these declines; from a societal point of view, loss of functional independence is a major determinant of the need for in-home assistance or institutionalization.
II. DEFINITIONS OF FUNCTIONAL OUTCOMES RELATED TO FRACTURE Functional competence is defined as the ability to complete functional tasks and to discharge social roles. These
OSTEOPOROSIS, SECOND EDITION VOLUME 1
819
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
820
GREENDALE AND BARRETT-CONNOR
TABLE 1 Design
Case Series of Outcome after Wrist Fracture Sample size
Results
Reference
Patients treated with closed reduction under anesthesia 10 years postfracture
55 of 100 treated cases
Gartland and Werley rating: 47% excellent, 38% good, 11% fair, 4% poor 44% hand pain by analog scale 40% weaker in fractured than nonfractured hand 27% one or more components of algodystrophy
[8]
Patients treated at a university-based orthopedic clinic 2 – 6 weeks after cast removal
59 of 60 treated cases
29% subjective hand pain 39% tender by dolorimetry 5% hand swelling; 15% finger swelling 41% symptoms of vasomotor instability
[7]
Patients treated at a university-based orthopedic clinic 10 years postfracture
55 of 100 treated cases (85% of 10-year survivors)
Gartland and Werley rating: 49% excellent, 36% good, 11% fair, 4% poor 27% one or more features of algodystrophy
[9]
Case series of fractures treated nonsurgically between 1977 and 1980 at a university-based orthopedic clinic 1.5 – 6 years postfracture
297 of 640 cases
Gartland and Werley rating: 38% excellent, 49% good, 11.5% fair, 1.5% poor
[11]
Case series of casted or reduced fractures at a university-based orthopedic clinic 3 and 6 month follow-up
215 of 235 at 3 months 209 of 235 at 6 months
36% diminished grip strength 36% radio-ulnar joint pain Patients perceived outcome, 48% excellent, 32% good, 18% fair, 2% poor.
functional competence, but are not synonymous with it. For example, a wrist fracture may profoundly limit the job performance of a craftsman, but have minimal occupational consequence to an executive. This chapter summarizes information about the outcomes of wrist, vertebral, and hip fracture because they are common and have been studied with respect to functional impact. The majority of included studies did not specify the level of trauma associated with the fracture; however, because the preponderance of subjects were older men and women, it is likely that most fractures were related to skeletal fragility.
III. OUTCOMES OF WRIST FRACTURES Most studies of function after wrist fracture are clinical case series (Table 1). The length of follow-up has ranged from weeks [7] up to a decade [8,9]. Thus some short-term and some longer term data are available on the prevalence of symptoms and disabilities in select groups of patients.
Gartland and Werley rating: 41% excellent or good at 3 months 69% excellent or good at 6 months
[13]
A surprising proportion of patients report symptoms after wrist fracture. Across series, the prevalence of hand pain ranges from 29 to 44% and hand weakness between 36 and 40%. Algodystrophy (also termed Sudeck’s atrophy or reflex sympathetic dystrophy) also occurs after wrist fracture. While the definition of algodystrophy is not uniform, it usually is composed of hand pain, limited finger movement, and vasomotor instability [7]. The reported prevalence of postfracture algodystrophy varies widely, in part due to differing definitions, ranging from 0.1 to 47% [7,8]. Table 1 shows results from studies of wrist fracture outcome classified by the Gartland and Werley [10] scoring system. This scale records self-reported symptoms, such as pain and limitation of movement or function, and physician-assessed objective measures, such as range of motion and muscle strength. Items included in the Gartland and Werley scale are reproduced in modified form in Table 2. By these criteria, most patients with wrist fracture are classified as having a good or excellent outcome [8,9,11 – 13]. However, two major shortcomings of this scale must be noted. First, good results are overestimated; e.g., a patient with hand pain and diminished grip strength (and no other symptom or sign) would be
821
CHAPTER 34 Outcomes Of Osteoporotic Fractures
TABLE 2 Items Included in the Gartland and Werley Rating System for Evaluation of Healed Wrist Fracturesa Item
Demeritsb
Subjective evaluation (by patient) No pain, disability, or limitation of motion
0
Occasional pain, slight limitation of motion, no disability
2
Occasional pain, some limitation of motion, feeling of weakness in wrist, no particular disability if careful; activities slightly restricted
4
Pain, limitation of motion, disability, activities more or less markedly restricted
6
Objective evaluation (by physician) Dorsiflexion 45°
5
Palmar flexion 30°
1
Ulnar deviation 15°
3
Radial deviation 15°
1
Supination 50°
2
Pronation 50°
2
Circumduction
1
Pain in distal radionavicular joint
1
Grip strength: 60% or less of opposite side
1
Residual deformity (by X-ray) Prominent ulnar styloid Residual dorsal tilt
1 2
Radial deviation of hand
2 – 3c
Finger stiffness
1 – 2c
Nerve complications
1 – 2 – 3c
Arthritis
1 – 2 – 3c
a
Modified from Gartland and Werley [10]. Demerits are given for each sign or symptom. Scores are excellent (0 – 2), good (3 – 8), fair (9 – 20), and poor (21). c Can give higher demerits according to degree of severity. b
classified in the “good” outcome category. Second, there are no reported validation studies of the scale, constructed in 1951. One predictor of poor (Gartland and Werley) outcome is the appearance of radial shortening by X-ray [9,11,14,15]. A shortened radius may reduce the mechanical function of extensor tendons [14]. The presence of postfracture algodystrophy also correlates with a poorer Gartland and Werley outcome [7,8,12]. Impairments in activities of daily living are also related to wrist fracture. In one population-based cohort, after controlling for other diseases, women with a history of wrist fracture were three times more likely to report difficulty with shopping for groceries or clothing, nine times more likely to have difficulty cooking, and two to three times more likely to have difficulty getting into and out of a car or descending stairs than women who have never fractured their wrists [16].
IV. OUTCOMES OF VERTEBRAL FRACTURES A. Pain and Functional Outcomes Studies of pain and functional outcomes after vertebral fracture can be classified into two major categories: those concerned with the consequences of vertebral deformities and those that focus on the effects of clinical fractures. Vertebral deformity studies examine the associations between radiographic evidence of vertebral fracture (loss of vertebral height), only 30% of which are severe enough to be recognized clinically [17], and clinical outcomes. Studies of patients with clinically evident fractures investigate the functional impairments, physical limitations, and symptoms of clinically diagnosed osteoporotic vertebral fracture; patients presenting for the evaluation of a problem thought to be related to vertebral fracture (e.g., back pain) are typically the subjects of these studies. Outcomes associated with radiographically determined vertebral deformity are summarized in Table 3; overall, cross-sectional surveys find that symptoms and disabilities are more pronounced when the degree of deformity is high. Women with moderate to severe prevalent vertebral deformities (see grading criteria, Table 3) report more back pain, general disability, disability specifically attributed to back problems, poorer self-rated health, and greater embarrassment about their appearance than women without vertebral deformity [18,19,20]. Physical limitation in at least one of six movements (e.g., bending) due to back pain, pain-associated activity limitations, and doctor visits for back problems occur with increased frequency among women with severe prevalent vertebral deformity compared to those without such deformity [21]. Longitudinal studies suggest that relatively recent vertebral deformities, rather than remote ones, produce negative health outcomes [22,23]. The Honolulu Osteoporosis Study classified recent deformities as those that occurred in the past 4 years. The number of recent vertebral deformities was significantly related to the odds of functional impairment (difficulty in performing three or more basic or intermediate level activities of daily living) [22]. In contrast, deformities that occurred more than 4 years previously were unrelated to functional limitations. Using a similar time frame (about 4 years) to define incident fractures, the Study of Osteoporotic Fractures (SOF) also found a differential impact of incident compared to prevalent deformities [23]. Even severe (greater than 4 SD) prevalent vertebral deformity was unrelated to pain or limited function. However, one or more incident deformities significantly increased the odds of back pain, back-related disability, annual days of bed rest, and number of limited activity days. Adverse health outcomes may be associated with clinical vertebral fractures, but data are sparse (Table 4). In one
822
GREENDALE AND BARRETT-CONNOR
TABLE 3
Pain and Functional Outcomes Associated with Radiographically Determined Vertebral Deformity
Design/subjects
Sample size
Vertebral deformity grading
Results
Reference
Cross-sectional study of volunteers 55 – 75 years of age
204
Normal: 15% anterior height loss Minimal 15 – 20% anterior height loss Mid 20 – 25% anterior height loss Moderate: 25% anterior height loss; mid or posterior losses with end plate deformity Severe: Marked crush fracture Score: Fractures weighted by severity and summed
Minimal and mild deformities were not associated with physical, functional, or emotional limitations Moderate to severe deformity was associated with higher general disability scores, some specific disability attributed to back problems, and embarrassment
[18]
Same subjects as Ettinger et al. [18]
204
Same vertebral X-ray readings and scoring method as original Ettinger study with one additional classification: women were assigned to a severity category based on the highest grade of fracture they manifested
Risk of back pain and disability attributed to back problems was associated with higher fracture summary score and higher fracture severity classification
[19]
Cross-sectional from a population-based recruitment 65 – 70 years of age
2,992
Digitized computer reading used to grade each vertebra between T2 and L4 Severity was graded by standard deviation departures from normative values for each vertebral level in the cohort Classification based on (i) worst deformity by vertebral level. (ii) number of severe deformities, and (iii) worst deformity by wedge, end plate, and crush deformity
No increase in frequency of back pain was evident until grade of vertebral deformity reached 4 SD If at least one 4 SD deformity was present, more difficulty performing activities related to back function and a higher back disability score was found
[21]
Community-based cohort of men and women from 19 European countries 50 – 79 years of age
15,570
McKloskey method and Eastell and Melton method: Grade 1 3 SD and grade 2 4 SD compared to normative values
Limitations in back-related activities of daily living, back pain, and poorer self-related health associated with one prevalent deformity. More severe deformity and greater number of deformities associated with increasingly worse function, greater pain, and poorer health status.
[20]
Community-based cohort of Japanese-American women in Hawaii 55 – 93 years of age
569
Prevalent: 3 SD deformity Incident: 20% height loss in one dimension
Number of recent (within 4 years) vertebral deformities positively associated with greater impairment in activities of daily living and with more physician visits for back pain.
[22]
Community-based cohort of white women 65 years of age
7,223
Prevalent: 3 SD deformity compared to normal mean values for anterior, mid, or posterior dimensions Prevalent, severe: 4 SD deformity Incident: 20% and at least 4 mm vertebral height loss
Greater number of incident deformities positively related to greater back-related disability, annual number of beddays, and annual number of limited activity days
[23]
(continues)
823
CHAPTER 34 Outcomes Of Osteoporotic Fractures
TABLE 3 Design/subjects
Sample size
Volunteers for study of osteoporosis risk factors recruited from general practitioner’s offices 50 – 82 years of age
222
(continued )
Vertebral deformity grading Wedge (anterior) concavities and compressions (posterior) graded within vertebral level by comparing posterior height to inferior-antero-posterior dimension Vertebrac also graded relative to the adjacent superior and inferior vertebra (except T4 and L4)
Results
Reference
No association was found between number of deformities and back pain Grade 2 or 3 deformity and low BMD was not associated with more back pain
[61]
A higher prevalence of back pain was not reported by the women with severe deformity
[62]
Grade 1: 2 Sd deformity (N 120) Grade 2: 3 SD deformity (N 27) Grade 3: 4 SD deformity (N 8) 77% of 1307 age-eligible members of a Londonbased general practice 45 – 69 years of age
1035
McCluskey method of grading vertebral deformity based on up to four adjacent vertebrac Mild 2-2.99 SD deformity Severe 3 SD deformity (N 20)
series of patients hospitalized for evaluation and treatment of vertebral fracture, 64% had pain and 70% had difficulty bending and rising [24]. Even within this highly selected, symptomatic sample, the number and severity of vertebral deformities (assessed by the Spinal Deformity index, Table 4) correlated modestly (0.29 to 0.44) with pain intensity, dysphoric mood, and degree of limitation in six physical movements (e.g., bending, lifting). In another study in the outpatient setting, women with one or more vertebral fractures had poorer self-rated function and measured physical performance (time to stand from a seated position) compared to a comparison group of patients with low back pain who did not have fractures [25]. The prevalence of symptoms, functional impairments, and negative emotional consequences in a group of 100 women, all of whom had chronic pain attributed to osteoporosis and at least one clinical vertebral fracture, is summarized in Table 5 [26]. The average time since fracture diagnosis in this study was 3.8 years, suggesting that some women with fracture have prolonged pain and disability; whether this is attributable to the fracture cannot be determined on the basis of these data. No correlation between
any quality of life domain and the number of vertebral fractures (median 2, range 1 – 11) was found. Physical performance impairments in trunk strength, walking speed, range of motion, and functional reach are also more common among women with several vertebral fractures than among nonfractured controls matched on age and comorbidity [27]. Instrumental activities of daily living may also be curtailed by vertebral fracture. In a cross-sectional population survey, women clinically diagnosed with vertebral fractures an average of 6 years previously were three times more likely to have difficulty cooking and shopping than comparable women without clinical vertebral fractures [16].
B. Illness, Hospitalization, and Mortality Outcomes Although the association between vertebral fractures and hospitalization has received little attention, one U.S. study found that hospitalization for vertebral fractures is not rare [28]. The sex-and race-specific incidence of
824
GREENDALE AND BARRETT-CONNOR
TABLE 4
Outcomes Associated with Clinically Diagnosed Vertebral Fractures
Design/subjects Cross-sectional clinical sample of patients admitted to hospital for evaluation or treatment 51 women, 61 ( 11) years 19 men, 52 ( 12) years
Cross-sectional survey selected from osteoporosis practice
Inclusions (i) postmenopausal, over 50; (ii) at least one vertebral fracture; (iii) clinical diagnosis of osteoporosis back pain Exclusions. (i) severe concomitant disease, (ii) unable to complete questionnaire Of 122 contacted women. 100 (82%) agreed to participate, average age 69 ( 8) years
Case-control study selected from a geriatric practice Cases: 10 women with documented vertebral fractures, 82 ( 6) years
Vertebral deformity grading
Results
Reference
Spinal deformity index (SDI): T3 to L5 anterior, posterior, and central heights calculated relative to T4, expressed as SD units SDI calculated by summing deviations between T3 and L5
Patients reported the following symptoms pain with activity (64%), limited bending (71%), and limited rising (70%) 41% needed help in self-care Correlation between SDI and physical limitations, 0.44
[24]
Vertebral fracture minimum criterion was 15% reduction in anterior compared to posterior height Anterior vertebral height loss graded as minimal, mild, moderate, or severe (grading criteria not given)
High prevalence of quality of life impairment (see Table 5)
[26]
No association between number of severity of fractures and severity of symptoms
Fracture criterion was a minimum of 30% or greater anterior wedge deformity Cases were required to have at least two vertebral fractures
Cases performed more poorly on all physical performance measures (e.g., lumbar spine motion, functional reach) Cases had more difficulty with activities of daily living and had more pain with activity
[27]
Height reduction of 20% in at least one vertebra
Cases were more dependent in self care, had more limitations in basic ADLS, and had poorer measured lower extremity strength
[25]
Controls: nonfractured female patients, 80 ( 6) years Major exclusions: inability to ambulare independently, poor visual acuity, failed cognitive status screen
Case-control study from an endocrinology practice Cases: 63 women with vertebral fracture, aged 65 ( 7.9) years Controls: 77 women with chronic low back pain and no spinal fracture, aged 56 ( 6.5) years
hospitalization due to osteoporotic vertebral fracture (1986 – 1989) was estimated using Medicare principal diagnosis of vertebral fracture, excluding persons under 65, and fracture diagnostic codes consistent with injury. Rates were calculated using the 1985 census as the de-
nominator. In women, there were 111,999 hospitalizations for vertebral fracture annually, a rate of 17.1 per 10,000. The annual rate in men was 3.7 per 10,000. Age-, gender-, and race-related differences in hospitalization rates were similar to those observed with hip fracture.
825
CHAPTER 34 Outcomes Of Osteoporotic Fractures
TABLE 5
Quality of Life in 100 Women with Osteoporotic Fracturesa Frequencyb
Importancec
Impactd
Pain
95
3.40
323.00
Pain when standing
85
3.47
295.00
Fatigue
73
3.40
248.00
pain when carrying
75
3.27
245.00
Pain when sitting
73
3.15
230.00
Difficulty carrying
87
3.23
281.00
Difficulty lifting
81
3.20
259.00
Difficulty walking
60
3.50
210.00
Difficulty bending
70
2.97
208.00
68
2.97
202.00
Vacuuming
81
2.95
239.00
Housework
70
3.19
223.00
Shopping for food
60
3.38
203.00
Shopping for clothes
61
3.30
201.00
Cleaning a bathtub
61
3.21
196.00
Afraid of falling
82
3.32
272.00
Afraid of fractures
74
3.57
264.00
Frustration
66
3.15
208.00
Anger
53
3.40
180.00
Overwhelmed
49
3.12
153.00
Traveling
57
3.30
188.00
Vacationing
41
3.41
140.00
Sports
43
3.07
132.00
Dancing
43
3.05
131.00
Attending church
35
2.94
103.00
Item Symptoms
Physical functioning
Difficulty finding comfortable chair Difficulty with activities of daily living
Emotions
Difficulty with leisure/social activities
a At least one clinically diagnosed vertebral fracture. From D. J. Cook, Arthritis Rheum. 36, 750–756 (1993). b The number of patients who cited the item as being a problem they experienced as a result of osteoporosis. c Rated by each patient, using a scale of 1 – 5, where 1 represents not at all important and 5 represents very important. d The product of frequency values times mean importance values (maximum possible score 500).
The consequences of vertebral fracture may include compromised respiratory function. In a consecutive case series of 132 women referred to a Canadian osteoporosis clinic, measured lung function was inversely related to the
number of prevalent vertebral fractures; forced vital capacity declined by about 10% for each prevalent thoracic anterior wedge deformity [29]. Independent of fracture number, the degree of kyphosis (assessed by Cobb’s angle) was associated with poorer lung function. (The correlation between Cobb’s angle and presence of vertebral fracture was only 0.5, suggesting that nonvertebral fracture-related kyphosis is also an important determinant of lung volumes [29].) Similarly, in a case-control study, women with prevalent vertebral fractures demonstrated forced expiratory flows (FEV1) that were 80% of predicted values (when standardized to FEV1 expected for their 25-year-old height) [30]. A comparison group of low back pain patients had significantly higher FEV1 values (92% of predicted) [30]. A decrease in survival has been noted after a clinically diagnosed vertebral fracture [31]. In one population-based cohort, the survival of patients with a clinical vertebral fracture was 61% compared to the expected population survival rate of 76% (relative survival 61/76, or 0.81). The survival rate of individuals with vertebral fractures diverged steadily from the expected rate over the course of follow-up. In the same population, the relative survival after hip fracture was similar (0.82), but the excess mortality was concentrated in the first 5 years after fracture, suggesting the excess early mortality after hip fracture likely represents the effect of fracture complications and comorbidity. Gradual widening over time between observed and expected mortality following vertebral fracture suggests that vertebral fracture is a marker for other conditions that increase the risk of death. Browner and colleagues [32] have shown that low bone mineral density (without fracture) predicts higher mortality, further supporting the concept that comorbidity and frailty often accompany osteoporosis. Women with radiographic vertebral deformities also experience higher all-cause mortality rates and are at higher risk of death due to pulmonary disease and cancer [33]. In SOF, compared to women without vertebral deformity, those with one or more fractures had a 23% greater ageadjusted mortality; the risk of death increased with the number of fractures. The significance of the relation between vertebral fractures and mortality was maintained after adjustment for numerous potential confounders, including smoking, bone density, self-reported health status, physical activity, and estrogen use (relative risk for one or more fractures 1.16) [33]. A large, multisite European cohort study found similar associations between vertebral deformity and mortality in both women and men [34]. The age-adjusted relative risk of death was 1.9 and 1.3 in women and men, respectively. After multiple adjustments for alcohol use, smoking, body mass index, self-rated health, and steroid use, relative risks or mortality was only slightly reduced 1.6 and 1.2 in women and men [34].
826
GREENDALE AND BARRETT-CONNOR
TABLE 6
Studies of Functional Outcome of Hip Fracturea
Design
Sample size/subjects
Type of fracture
Reference
Consecutive case series of subjects admitted to a rehabilitation ward whose hip fractures had occurred within 6 months or less
118 women 12 men Age range: 50 – 98 years
56% trochanteric 44% femoral neck
[35]
Consecutive patients over 54 years old with new hip fractures admitted to the university hospital
50 women 25 men
68% intertrochanteric 22% femoral neck
[37]
84% admitted from home
Age not reported
Consecutive case series of newly diagnosed hip fractures at one county and two private hospitals
63 women 29 men
51% femoral neck 45% intertrochanteric
[38]
Able to speak English; not disoriented
Mean age: 71 years
4% greater trochanter
Consecutive cases of fracture admitted to seven Baltimore area hospitals
442 women 94 men mean age: 78 years
53% extracapsular
[36]
a
Results of studies are shown in Table 7.
V. OUTCOMES OF HIP FRACTURES Functional competence in basic and intermediate activities of daily living and physical functioning is markedly diminished after hip fracture. Most studies have assessed postfracture declines in function compared to patient or proxy recall of function prior to the fracture [35 – 38]. One report measured function prior to and after hip fracture [39]. As shown in Tables 6 and 7, return to prefracture competence in ADL occurs in less than 50% of patients by 6 months after fracture; little further improvement in ADL is made by 1 year postfracture. Hip fractures also have a devastating effect on IADL: at 6 months postfracture, approximately one-fourth of patients regain their prefracture functional status, with no further recovery evident by 1 year after the fracture event. A similarly modest return to prior social/role function is obtained by most hip fracture TABLE 7
patients, with only 26% returning to premorbid levels of function. In these instances, recovery refers to a resumption of prior function, not to attainment of independent function. Patients who were independent in ADL and IADL prior to their hip fracture also suffer marked deterioration in functional status after fracture. For example, about half of those who dressed independently before fracture regained this ability and only one-third of patients resumed independent transferring (i.e., the ability to move from bed to chair or from chair to upright posture) [39]. Considering patients who reported independence in several ADL and basic mobility (bed to chair and toilet transfers, putting on socks and shoes, and indoor walking), 33% recovered independence in all functions after hip fracture. At 1 year after fracture, 40% of patients were independent in all BADL, compared to 70% prior to fracture. Physical performance deteriorates significantly after hip fracture. One study that recorded functional status prior to
Results of Studies Summarized in Table 6a
Walking (%)b
Basic ADL (%)b
6 months
1 year
6 months
1 year
24
40
43
48
53 (intertrochanteric) 79 (subcapital)
Intermediate ADL (%)b 6 months
1 year
[35]
33 (all fractures)
21
65 62 a
64
[37]
[38] 46
48
27
29
Marottoli et al. [39] study not included, as results are reported only for return to independent status. Percentages are those returning to prefracture levels, not necessarily to independent function.
b
Reference
[36]
827
CHAPTER 34 Outcomes Of Osteoporotic Fractures
hip fracture occurrence reported poorer physical function outcomes than those studies that estimated prefracture function by recall: at 6 months, only 15% of subjects with hip fracture could walk across a room independently compared to 75% at baseline. While 41% could walk one-half mile at baseline, 6% could do so 6 months after hip fracture. The ability to climb stairs was regained by only 8%; prior to fracture, 63% had been able to perform this activity. Other studies that employed recall of function prior to fracture found less severe, but quite substantial losses of mobility. By 66 months after fracture, between 24 and 62% regained prior capability in walking; these figures improved to 40 – 79% at 1 year. Hip fracture results in functional limitations in both intermediate and advanced activities of daily living. These limitations include diminished competence in money management, cooking, performance of housework, use of transportation, grocery shopping, carrying bundles, taking medications, visiting friends and engaging in community activities [16,35 – 37,40 – 42].
A. Predictors of Recovery after Hip Fracture Elders who fracture their hips tend to have other diseases and functional limitations prior to the occurrence of the fracture. Thus, the hip fracture is often not the sole factor leading to functional decline; rather, it is maybe the “last straw” effect of the hip fracture in the setting of other comorbidities and limitations that leads to the profound decrease in functional capability. Factors that predict better recovery from hip fracture are consistent with the concept that the outcome of hip fracture depends largely on the prior condition of individuals who suffer the fractures. Recovery of prefracture functional status and/or return to living at home is more common in patients who were younger [36,37,43], in better general health [37,41,45], not demented [36,43 – 45], and had larger social networks [36,44]. In some [36,37,47], but not all [47], studies intertrochanteric fractures were associated with better functional outcomes, fewer postoperative complications, and lower mortality than femoral neck fractures.
from 5 to 9% compared to 1 to 3% for similarly aged women [48]. The effect of race on mortality varies by gender; white men have a slightly higher mortality rate than black men, while the converse is observed for women [53]. Nonoperative treatment of hip fracture is associated with higher mortality [49,54], but the frailest patients are those most likely to be treated in this fashion. Length of hospitalization for hip fracture has declined substantially in the United States over the past decade, from 20 days in 1981 to 13 days in 1990 [48]. In parallel, in-hospital hip fracture mortality has diminished from 11% in 1970 to 3 – 4% in 1991, perhaps reflecting a shift in deaths to the nursing home setting, as well as improved medical and surgical care of fractures. Nevertheless, 1-year mortality after hip fracture in the United States has remained stable at 20 – 25% [48]. Similar to the pattern seen with in-hospital deaths, older age, male sex, and comorbidity are associated with higher mortality [36]. Fracture treatment also appears to be related to longterm survival; however, underlying differences in patients chosen for each operative procedure and for nonoperative treatment make the interpretation of differences difficult. Neither acute nor long-term mortality rates after hip fracture are attributable solely to hip fracture, as they represent all-cause mortality estimates. In general, age-specific mortality during the year after hip fracture exceeds by 6 – 14% all-cause mortality in comparable age groups. Estimates of the time after fracture during which mortality exceeds age-specific population norms range from 6 months to 4 years [36,55,56,47]. If hip fracture leads to excess mortality, would its prevention save lives? A recent analysis from SOF suggests that the answer to this question may be no [47]. After adjustment for many other predictors of mortality (such as age, health status, smoking, weight, exercise) women with hip or pelvic fracture were 2.4 times more likely to die compared to nonfracture controls. However, a detailed review of death certificates and hospital records of the 64 cases of hip or pelvic fracture revealed that only 14% of deaths were caused or hastened by the fracture. More commonly, the fractures were markers of underlying chronic diseases.
C. Institutionalization B. Mortality from Hip Fracture Early mortality occurring during the acute hospitalization for hip fracture is relatively uncommon. According to the U.S. National Hospital Discharge Surveys of 1988 and 1991, between 3 and 4% of patients died during hospital admission for hip fracture [48]. Older age, male sex, and poorer general health are related to higher in-hospital mortality [49 – 52]. Death rates for men ages 50 to 99 range
Wide variation in practice patterns and availability of services by region and country make it difficult to estimate the proportion of patients who are transferred to nursing homes after acute hospitalization for hip fracture. In the United States, this figure varies from one-fourth to three-fourths of hospital discharges after hip fracture [57,58]. In England, up to one-third of the hospital stay may be accounted for by the unavailability of nursing
828
GREENDALE AND BARRETT-CONNOR
TABLE 8
Sex
No.
Discharge Status and Destination for People with a Hip Fracture Treated in Short-Stay Nonfederal Hospitals in 1991a
Died in hospital (%)
Left against Discharged to home (%)
Discharged to another medical advice (%)
Discharged to a long-term care short-stay hospital (%)
Discharged alive; institution (%)
Destination not stated (%)
Discharged status not stated (%)
Both sexes
281,685
10,535 (4)
90,889 (32)
1390 (1)
36,278 (13)
108,756 (39)
29,079 (10)
4758 (2)
Men
68,541
5114 (7)
22,888 (33)
224 (1)
8866 (13)
24,467 (36)
6454 (9)
528 (1)
Women
213,144
5421 (3)
68,001 (32)
1166 (1)
27,412 (13)
84,289 (40)
22,625 (11)
4230 (2)
a
Modified from U. S. Congress, OTA, 1994.
home placements; it is possible that longer hospitalization would lead to fewer nursing home admissions, but this has received limited study. Using the 1988 and 1991 National Hospital Discharge Surveys, the U.S. Office of Technology Assessment [48] has compiled overall estimates of discharges to nursing homes from nonfederal hospitals (Table 8). The Post Acute Care Study estimated that the duration of nursing home placements after hip fracture is as follows: 1 month, 24%; 2 months, 8%; 3 months, 8%; 4 months, 8%; 5 months, 8%; 6 months, 10%; and 1 year or more, 34% [48]. However, characteristics other than hip fracture, such as old age, ADL and IADL dependence, and impaired mental status, predict a long nursing home stay [59,60]. Therefore, to estimate the use of services and expenditures related to hip fracture, the Office of Technology Assessment concluded that a maximum of 1 year’s nursing home stay should be attributed to hip fracture and that subsequent time in the nursing home is largely due to other factors such as dementia, comorbidity, frailty, and lack of social support.
VI. CONCLUSIONS Mortality and nursing home placement are well-recognized outcomes of osteoporotic hip fracture. The consequences of other types of osteoporotic fractures are less well studied and not widely appreciated. Wrist and vertebral fractures are associated with pain and limitations in multiple activities of daily living. Additional studies are required to further delineate the functional sequelae of nonhip fractures. The development of strategies to prevent loss of independence and function after osteoporotic fracture is an important goal for future research.
References 1. M. J. Parker, J. W. Myles, J. K. Anand et al., Cost – benefit analysis of hip fracture treatment. J. Bone J. Surg. 74B, 261 – 264 (1992). 2. S. Katz, T. D. Downs, H. R. Cash et al., Progress in development of the index of ADL. Gerontology, 10, 20 (1970). 3. M. P. Lawton and E. M. Brody, Assessment of older people: Selfmaintaining and instrumental activities of daily living measure. Gerontology 9, 179 (1969). 4. G. G. Fillebaum, Screening the elderly: A brief instrumental activities of daily living measure. J. Am. Geriatr. Soc. 33, 698 (1985). 5. D. B. Reuben and D. H. Solomon, Assessment in geriatrics: Of caveats and names. J. Am. Geriatr. Soc. 37, 570 – 572 (1989). 6. D. B. Reuben, D. L. Weiland, and L. Z. Rubenstein, Functional assessment in older persons: Concepts and implications. Facts Res. Gerontol. 7, 231 – 240 (1993). 7. R. M. Atkins, T. Duckworth, and J. A. Kanis, Features of algodiptrophy after Colles’ fracture. J. Bone J. Surg. 72(8), 105 – 110 (1990). 8. J. Field, D. Warwick, C. G. Bannister et al., Long-term prognosis of displaced Colles’ fracture: A 10-year prospective review. Injury 23, 529 – 532 (1992). 9. D. Warwick, J. Field, D. Prothero et al., Function after ten years after Colles’ fracture. Clin. Orthop. Rel. Res. 295, 270 – 274 (1993). 10. J. J. Gartland and C. W. Werley, Evaluation of healed Colles’ fractures. J. Bone J. Surg. 33A, 895 – 907 (1951). 11. M. Altissimi, R. Antenucci, C. Fiacca et al., Long-term results of conservative treatment of fractures of the distal radius. Clin. Orthop. Rel. Res. 206, 202 – 210 (1986). 12. R. H. Roumen, W. L. E. M. Hesp, and E. D. M. Bruggink, Unstable Colles’ fracture in elderly patients. J. Bone J. Surg. 73B, 307 – 311 (1991). 13. H. D. Stewart, A. R. Innes, and F. D. Burke, Factors affecting the outcome of Colles’ fracture: An anatomical and functional study. Injury 16, 289 – 295 (1985). 14. M. McQueen and J. Caspers, Colles’ fracture: Does the anatomical result affect the final function? J. Bone J. Surg. 70B, 649 – 651 (1988). 15. R. N. Villar, D. Marsh, N. Rushton et al., Three years after Colles’ fracture. J. Bone J. Surg. 69B, 635 – 638 (1987). 16. G. A. Greendale, E. Barrett-Connor, S. Ingles et al., Late physical and functional effects of osteoporotic fractures in women: The Rancho Bernardo Study. J. Am. Geriatr. Soc. 43, 955 – 961 (1995).
CHAPTER 34 Outcomes Of Osteoporotic Fractures 17. C. Cooper, G. Campion, L. J. Melton III, Hip fractures in the elderly: A world-wide projection. Osteopor. Int. 2, 285 – 289 (1992). 18. B. Ettinger, J. E. Block, R. Smith et al., An examination of the association between vertebral deformities, physical disabilities and psychosocial problems. Maturitus 10, 283 – 296 (1988). 19. P. D. Ross, B. Ettinger, J. W. Davis et al., Evaluation of adverse health outcomes associated with vertebral fractures. Osteopor. Int. 1, 134 – 140 (1991). 20. C. Mathis, U. Weber, T. W. O’Neill et al., Health impact associated with vertebral deformities: Resulta from the European vertebral osteoporosis study (EVOS). Osteopor. Int. 8, 364 – 372 (1988). 21. B. Ettinger, D. M. Black, M. C. Nevitt et al., Contribution of vertebral deformities to chronic back pain and disability. J. Bone Miner. Res. 7, 449 – 456 (1992). 22. C. Huang, P. D. Ross, and R. D. Wasnich, Vertebral fracture and other predictors of physical impairment and health care utilization. Arch. Intern. Med. 156, 2469 – 2475 (1996). 23. M. C. Nevitt, B. Ettinger, D. M. Black et al., The association of radiographically detected vertebral fractures with back pain and function: A prospective study. Ann. Intern. Med. 128, 793 – 800 (1998). 24. G. Leidig, H. W. Minne, P. Sauer et al., A study of complaints and their relation to vertebral destruction in patients with osteoporosis. Bone Miner. 8, 217 – 229 (1990). 25. G. Leidig-Bruckner, H. W. Minne, C. Schlaich et al., Clinical grading of spinal osteoporosis of life components and spinal deformity in women and chronic low back pain and women with vertebral osteoporosis. J. Bone Miner. Res. 12, 663 – 675 (1997). 26. D. J. Cook, G. H. Guyatt, J. D. Adachi et al., Quality of life issues in women with vertebral fractures due to osteoporosis. Arthritis Rheum. 36, 750 – 756 (1993). 27. K. W. Lyles, D. T. Gold, K. M. Shipp et al., Association of osteoporotic vertebral compression fractures with impaired functional status. Am. J. Med. 94, 595 – 601 (1993). 28. S. J. Jacobsen, C. Cooper, M. S. Gottlieb et al., Hospitalization with vertebral fracture among the aged: A national population-based study, 1986 – 89. Epidemiology 3, 515 – 518 (1992). 29. J. A. Leech, C. Dulberg, S. Kellie et al., Relationship of lung function to severity of osteoporosis in women. Am. Rev. Respir. Dis. 141, 68 – 71 (1990). 30. C. Schlaich, H. W. Minne, T. Bruckner et al., Reduced pulmonary function in patients with spinal osteoporotic fractures. Osteopor. Int. 8, 261 – 267 (1998). 31. C. Cooper, E. S. Atkinson, S. J. Jacobsen et al., Population-based study of survival after osteoporotic fractures. Am. J. Epidemiol. 137(9), 1001 – 1005 (1993). 32. W. S. Browner, D. G. Seeley, T. M. Vogt et al., Non-trauma mortality in elderly women with low bone mineral density. Lancet 338, 355 – 358 (1991). 33. D. M. Kado, W. S. Browner, L. Palermo et al., Vertebral fractures and mortality in older women: A prospective study. Arch. Intern. Med. 159, 1215 – 1220 (1999). 34. A. A. Ismail, T. W. O’Neill, C. Cooper et al., Mortality associated with vertebral deformity in men and women: Results from the European prospective osteoporosis study (EPOS). Osteopor. Int. 8, 291 – 297 (1998). 35. S. Katz, K. Heiple, T. Downs et al., Long-term course of 147 patients with fracture of the hip. Surg. Gynecol. Obstet. 124, 1219 – 1230 (1967). 36. J. Magaziner, E. M. Simonsick, T. M. Kashner et al., Predictors of functional recovery one year following hospital discharge for hip fracture: A prospective study. J. Gerontol. 45(3), M101 – M107 (1990). 37. A. M. Jette, B. A. Harris, P. D. Cleary et al., Functional recovery after hip fracture. Arch. Phys. Med. Rehabil. 68, 737 – 740 (1987).
829 38. S. R. Cummings, S. L. Phillips, M. E. Wheat et al., Recovery of function after hip fracture: The role of social supports. J. Am. Geriatr. Soc. 36(9), 801 – 806 (1988). 39. R. A. Maritolli, L. F. Berkman, and L. M. Cooney, Decline in physical function following hip fracture. J. Am. Geriatr. Soc. 40, 861–866 (1992). 40. G. Jarnlo, Hip fracture patients: Background factors and function. Scand. J. Rehabil. Med. Suppl. 24, 1 – 31 (1990). 41. L. Ceder, K. Svensson, and K. G. Thorngren, Statistical prediction of rehabilitation in elderly patients with hip fractures. Clin. Orthop. Rel. Res. 152, 185 – 190 (1980). 42. A. A. Guccione, D. T. Felson, J. J. Anderson et al., The effects of specific medical conditions on the functional limitations of elders in the Framingham Study. Am. J. Public Health 84, 351 – 358 (1994). 43. J. M. Mossey, E. Mutran, K. Knott et al., Determinants of recovery 12 months after hip fracture: The importance of psychosocial factors. Am. J. Public Health 79(3), 279 – 296 (1989). 44. J. A. Van der Sluijs and G. H. I. M. Walenkamp, How predictable is rehabilitation after hip fracture? Acta Orthop. Scand. 62, 567 – 572 (1991). 45. S. R. Cummings, J. L. Kelsey, M. C. Nevitt, and K. O’Dowd, Epidemiology of osteoporosis and osteoporotic fractures. Epidemiol. Rev. 7, 178 – 208 (1985). 46. Keene (1993). 47. W. S. Browner, A. R. Pressman, M. C. Nevitt et al., Mortality following fractures in older women. Arch. Intern. Med. 156, 1521 – 1525 (1996). 48. U. S. Congress Office of Technology Assessment, “Hip Fracture Outcomes in People Age 50 and Over: Background Paper,” OTA-BP-H120. U. S. Government Printing Office, Washington, DC, 1994. 49. A. H. Myers, E. G. Robinson, M. L. Van Natta et al., Hip fractures among the elderly: Factors associated with in-hospital mortality. Am. J. Epidemiol. 134(10), 1128 – 1137 (1991). 50. T. I. Davidson and W. N. Bodey, Factors influencing survival following fractures of the upper end of the femur. Injury 17, 12 – 14 (1986). 51. T. B. Young and A. C. C. Gibbs, Prognosis factors for the elderly with proximal femoral fracture. Arch. Emerg. Med. 1(4), 215 – 224 (1984). 52. J. G. Crane and C. B. Kernek, Mortality associated with hip fractures in a single geriatric hospital and residential health faciltiy: A ten-year review. J. Am. Geriatr. Soc. 31, 472 – 475 (1983). 53. S. E. Kellie and J. A. Brody, Sex-specific and race-specific hip fracture rates. Am. J. Public Health. 80(3), 326 – 328 (1990). 54. L. Matheny, T. F. Scott, C. M. Craythorne et al., Hospital mortality in 342 hip fractures. WV Med. J. 76(8), 188 – 190 (1980). 55. E. S. Fisher, J. A. Baron, D. J. Malenka et al., Hip fracture incidence and mortality in New England. Epidemiology 32(2), 116 – 122 (1991). 56. J. E. Kenzora, R. E. Mc Carthy, and J. D. Lowell, Hip fracture mortality. Clin. Orthop. Rel. Res. 186, 45 – 56 (1984). 57. J. F. Fitzgerald, L. F. Fagan, W. M. Tierney et al., Changing patterns of hip fracture care before and after implementation of the prospective payment system. JAMA 258(2), 218 – 221 (1987). 58. M. B. Gerety, V. Soderholm-Difatte, and C. H. Winograd, Impact of prospective payment and discharge location on the outcome of hip fracture. J. Gen. Intern. Med. 4, 388 – 391 (1989). 59. K. L. Kahn, E. B. Keeler, and M. J. Sherwood, Comparing outcomes of care before and after implementation of the DRG-based prospective payment system. JAMA 264(15), 1984 – 1988 (1990). 60. E. A. Chrischilles, C. D. Butler, C. S. Davis et al., A model of lifetime osteoporosis impact. Arch. Intern. Med. 151, 2026 – 2032 (1991). 61. P. H. Nicholson, M. J. Haddaway, M. W. Davie et al., Vertebral deformity, bone mineral density, back pain and height loss in unscreened women over 50 years. Osteopor. Int. 3, 300 – 307 (1993). 62. T. D. Spector, E. V. McCloskey, D. V. Doyle et al., Prevalence of vertebral fracture in women and the relationship with bone density and symptoms: The Chingford Study. J. Bone Miner. Res. 8, 817 – 822 (1993).
CHAPTER 35
The Nature of Osteoporosis ROBERT MARCUS SHARMILLA MAJUMDER
Veterans Affairs Medical Center, Palo Alto, California 94304 Department of Radiology, University of California, San Francisco, San Francisco, California 94143
I. Defining Osteoporosis II. The Nature of Osteoporotic Bone
III. Conclusions References
I. DEFINING OSTEOPOROSIS
related to menopausal estrogen loss and the other to aging. This concept has been elaborated upon by Riggs and associates [2] who suggested the terms Type I osteoporosis, to signify a loss of trabecular bone after menopause, and Type II osteoporosis, to represent a loss of cortical and trabecular bone in men and women as the end result of age-related bone loss. Whereas the Type I disorder directly results from lack of endogenous estrogen, Type II osteoporosis reflects the composite influences of long-term remodeling efficiency, adequacy of dietary calcium and vitamin D, intestinal mineral absorption, renal mineral handling, and parathyroid hormone (PTH) secretion. This model is discussed in detail in Chapter 38. Although there is heuristic value to defining subsets of patients in this manner, compelling validation of the model remains to be offered. Iliac crest biopsies do not show a characteristic histomorphometric profile of patients whose clinical status suggests type I osteoporosis. Most importantly, the model suffers by adhering to a bone loss paradigm. Postmenopausal women with low bone mass are assumed to have achieved that status because they experienced a drastic menopausal loss of bone. However, we understand that bone mass at any time in adult life reflects the peak investment in bone mineral at skeletal maturity minus that which has been subsequently lost. A woman who experienced interruption of menses, extended bed rest, eating disorder, or systemic illness during her
This chapter introduces a series of contributions dealing with osteoporosis in a wide variety of clinical settings. Its purpose is to consider the definition of osteoporosis and to discuss the nature of osteoporotic bone, its mass and distribution, its microscopic architecture, and other aspects of its quality and strength. Osteoporosis is a condition of generalized skeletal fragility in which bone strength is sufficiently weak that fractures occur with minimal trauma, often no more than is applied by routine daily activity. “Primary” osteoporosis is by tradition a skeletal disorder of postmenopausal women (“postmenopausal” osteoporosis) or of older men and women (“senile” osteoporosis). The term “secondary” osteoporosis refers to bone loss resulting from specific, defined clinical disorders, such as thyrotoxicosis or hyperadrenocorticism. Because of the intimate relationship of reproductive hormone deficiency to the development of postmenopausal osteoporosis, several conditions that could legitimately be considered secondary forms of osteoporosis are commonly treated as variants of primary osteoporosis. These include osteoporosis resulting from exercise-related amenorrhea and from prolactinsecreting tumors. Albright and Reifenstein [1] proposed in 1948 that primary osteoporosis consists of two separate entities, one
OSTEOPOROSIS, SECOND EDITION VOLUME 2
3
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
4 adolescent growth years might enter adult life having failed to achieve the bone mass that would have been predicted from her genetic or constitutional profile. If she then underwent a perfectly normal rate of bone loss, her skeleton would still be in jeopardy simply due to the deficit in peak bone mass. Thus, for the present, it may be most appropriate to consider osteoporosis the consequence of a stochastic process, that is, multiple genetic, physical, hormonal, and nutritional factors acting alone or in concert to diminish skeletal integrity. Although historical artifacts show that characteristic deformities of vertebral osteoporosis were recognized in antiquity [3], broad awareness of this condition has come about only during the past few decades, catalyzed in particular by the work of Albright and colleagues [1]. Unfortunately, because traditional radiographic techniques cannot distinguish osteoporosis until it is severe, confirmation of the diagnosis remains problematic. Until recently, diagnosis was by necessity clinical, requiring a history of one or more low-trauma fractures. Although highly specific, such a grossly insensitive diagnostic criterion offers no assistance to physicians who hope to identify and treat affected individuals who have been fortunate not yet to have sustained a fracture. The introduction of accurate noninvasive bone mass measurements afforded the opportunity to make an early diagnosis of osteoporosis. Bone mineral density (BMD) of patients with osteoporotic fractures was generally found to be lower than that of age-matched nonfractured controls. However, it soon became evident that substantial overlap exists in the distribution of BMD of patients with and without osteoporotic fracture [2,4 – 6] (Fig. 1) and that BMD does not accurately predict the presence of osteoporotic fractures (Fig. 2) [6].
MARCUS AND MAJUMDER
Several factors underlie the poor ability of BMD measurements to predict fracture prevalence. The normative data against which BMD comparisons are most often made have been determined for Caucasian men and women, and do not necessarily apply to other ethnic groups. BMD is clearly related to body weight, yet routine clinical bone mass assessments are not weight-adjusted. Various features of bone geometry that affect bone strength and fracture risk are not generally considered in the clinical interpretation of bone mass measurements. These include bone size, the distribution of bone mass around its bending axis (moments of inertia), and some derivative functions, such as the hip axis length [7]. Moreover, bone mass determinations cannot distinguish individuals with low mass and intact microarchitecture from those with equal mass who have trabecular disruption and cortical porosity. They also cannot distinguish other aspects of bone quality. Finally, some patients with skeletal fragility have enjoyed the good fortune not to experience a fracture by the time of their bone mass measurement. These individuals, who would be designated “nonfractured,” may nonetheless be in severe jeopardy for fracture in the near future. Thus, as a means of diagnosing the presence or absence of osteoporotic fracture, BMD assessment is not useful.
FIGURE 2 FIGURE 1
Individual lumbar spine BMD values for 84 women with one or more vertebral compression fractures compared to age-predicted mean values (regression line) and 90% confidence interval (shaded region). Results show substantial BMD overlap of fracture patients with the normal range. Reproduced from The Journal of Clinical Investigation (2) by copyright permission of the Society for Clinical Investigation.
Receiver-operated curve (ROC) relating the sensitivity and specificity of bone density measurement techniques for diagnosis of prevalent fractures. DPA, dual photon absorptiometry; QCT, quantitative computed tomography; TBC, total body calcium; SPA, single photon absorptiometry. A perfect diagnostic instrument would give a line that reaches the top left-hand corner of the box. Reprinted from Ott et al., with permission (6).
5
CHAPTER 35 Nature of Osteoporosis
A very different conclusion is reached when bone mass is used to assess an individual’s prospective long-term risk for fracture. This topic is fully discussed in Chapters 33 and 59. Briefly stated, large prospective studies have shown that a reduction in BMD of 1 standard deviation from the mean value for an age-specific population confers a two- to three fold increase in long-term fracture risk [8 – 11]. In a manner similar to that by which serum cholesterol concentration predicts risk for heart attack or blood pressure predicts risk for stroke, BMD measurements can successfully identify subjects at risk of fracture and can help physicians select those individuals who will derive greatest benefit for initiation of therapy. In 1994, a group of senior investigators in this field offered a working definition of osteoporosis based exclusively on bone mass [12]. The reasoning behind this proposal, made on behalf of the World Health Organization (WHO), was that the clinical significance of osteoporosis lies exclusively in the occurrence of fracture, that bone mass predicts long-term fracture risk, and that selection of rigorous diagnostic criteria would minimize the number of patients who are incorrectly diagnosed. The authors suggested a cutoff BMD value of 2.5 standard deviations below the average for healthy young adult women. Using this value, approximately 30% of postmenopausal women would be designated as osteoporotic (Fig. 3), which gives a realistic projection of lifetime fracture rates. In addition, Kanis et al. [12] proposed that BMD values of 1 – 2 standard deviations below the young adult mean be designated as “osteopenic.” Such values identify individuals at increased risk for fracture, but for whom a diagnosis of osteoporosis would not be justified since it would mislabel far more individuals than would actually be expected ever to fracture. This approach has proven useful for clinical management, but has several limitations, all of which were acknowledged by the authors. Although a significant relationship between BMD and fracture is likely to hold for men, no evidence yet suggests that the cutoff of 2.5 SD carries the same risk as for women. The applicability of this criterion to young people prior to their acquisition of peak bone mass would be inappropriate. The BMD measurement is itself subject to several confounding factors, including bone size and geometry. As BMD correlations among skeletal sites are not strong, designating a person “normal” based on a single site, for example the lumbar spine, necessarily overlooks individuals with low bone density elsewhere, such as the hip. It seems reasonable to suppose that adjustment of bone density readings for such factors as body size, bone geometry, and ethnic background might improve the accuracy of this technique. It should be evident that the WHO proposal [12] will be of limited use to investigators whose interest is the nature and causes of osteoporosis. Knowledge of a low bone density at a particular point in
FIGURE 3
Bone mineral density in women at different ages and the prevalence of osteoporosis, defined as a BMD value 2.5 standard deviations below that for a healthy 25-year-old woman. Reprinted from Kanis et al., with permission (12).
time offers no information regarding the adequacy of peak bone mass attained, the amount of bone that may have been lost, or the quality of bone that remains.
II. THE NATURE OF OSTEOPOROTIC BONE At a recent consensus development conference [13], osteoporosis was defined as “a disease characterized by low bone mass and microarchitectural deterioration of bone tissue, leading to enhanced bone fragility and a consequent increase in fracture risk.” Implicit in this definition is the view that osteoporosis results from deficits in normally composed bone and that the residual bone is defective in amount and distribution, but not in matrix composition or mineralization. In contrast, osteomalacic bone matrix is grossly undermineralized. Moreover, for many years the prevailing view has been that osteoporosis develops
6
MARCUS AND MAJUMDER
through excessive loss of bone. Only recently has attention been drawn to abnormalities in bone acquisition as a basis for subsequent bone fragility (see Chapter 25). This latter issue notwithstanding, the dominant model of osteoporosis among workers in the field has, until recently, emphasized only the amount and distribution of bone substance. However, the great overlap in bone density between individuals with and without fracture indicates the limitations of such a model to account adequately for individual differences in fracture susceptibility. In other words, additional properties of bone quality may contribute to skeletal fragility. Several lines of evidence support such a view. McCabe et al. [14] used a technique called Mechanical Resistance Tissue Analysis to show an age-related decrease in ulnar bending stiffness in women that was associated with, but not completely accounted for by changes in either bone mineral or gross bone geometry. Kann et al. [15] demonstrated an age-related decrease in the resonance frequency of cortical bone in women. Transmission velocity of ultrasound, a property related to bone material and structural properties, degrades with age, and has a relationship to fracture risk that is not fully accounted for by bone mass itself [16]. In 1993, the National Institute on Aging reported the proceedings of a multidisciplinary review of bone quality in osteoporosis [17], from which one may safely conclude that very little was known in this area. During the past several years, considerable effort has been focused on adapting histomorphometric analysis as well as developing noninvasive methods to the study of bone quality. In the remainder of this chapter, we review several aspects of this problem (Table 1). At the outset, however, we emphasize that such qualitative abnormalities, like decreased bone mass, are commonly associated with normal human aging and do not appear to be specific markers for osteoporosis.
A. Alterations in Bone Composition Considerable evidence indicates that the mineral composition of bone is neither homogeneous nor constant throughout life, and substantial heterogeneity in the mineralization of bone matrix can easily be shown using microra-
TABLE 1
Aspects of Bone Quality in Osteoporosis
Cortical porosity Undermineralized matrix Cement line accumulation Trabecular thinning, perforation and disruption Fatigue accumulation
FIGURE 4
Microradiograph of a section from the femoral shaft of a 62year-old man demonstrating microheterogeneity of matrix mineralization. Different gray levels reflect differences in mineralization, white representing the highest level. From Grynpas (18) with permission.
diography [18] (Fig. 4). As early as 1960, Jowsey [19] reported that the abundance of “low density” osteons was increased in biopsies obtained from subjects older than 50 years. This observation, as well as others [20], suggests either that a greater fraction of the total osteonal pool is actively engaged in remodeling, so that at any given time more remodeling units are encountered early in the mineralization process, or that completion of osteonal mineralization somehow deteriorates with age. As discussed by Crofts et al. [21], the modest rise in remodeling activation frequency observed after age 50 does not adequately explain the higher prevalence of lower-density osteons. 1. DECREASED MATRIX MINERALIZATION The first systematic effort to assess the composition of human osteoporotic bone was that of Burnell et al. [22], who compared iliac crest biopsies from osteoporotic postmenopausal women with vertebral compression fractures to biopsies from normal controls. As expected, osteoporotic bone was less dense. However, the fraction of mineral per gram of bone tissue was also reduced. Moreover, within the mineral phase, carbonate and the calcium-to-phosphorus ratio were decreased, while sodium and magnesium content were increased, yet the same biopsies gave no hint of osteomalacia. Although these results describe average values for the entire study cohort, they reveal considerable heterogeneity in bone composition, even within this group of clinically homogeneous patients. Most patients had normal results; one quarter showed undermineralized matrix, and only a few showed decreased matrix but normal mineralization. The authors also found that the subjects with decreased mineral fraction were those who also had an increased content of sodium and magnesium in the
7
CHAPTER 35 Nature of Osteoporosis
mineral phase, suggesting the presence of skeletal calcium deficiency. What is the additional mechanical consequence of undermineralization to an already porotic bone? While no published clinical fracture data address this question, mineralization is known to contribute importantly to bone structural strength [23], and mineral content strongly affects fundamental bone material properties, such as Young’s modulus of elasticity. In fact, a modest 7% increase in bone mineral content is associated with a threefold increase in bone stiffness and a doubling in breaking strength [24]. Thus, it seems inescapable that undermineralization would promote bone fragility. In a follow-up study, Burnell et al. [25] showed that daily supplements of calcium and vitamin D improved the calcium content of bone tissue while simultaneously decreasing its sodium content. Given prevailing views about pathogenesis of osteoporosis, in which simultaneous turnover of matrix and mineral is the consequence of bone remodeling, it is not clear how alterations in remodeling activity alone would eventuate in a relatively undercalcified bone matrix. The authors offered two explanations for this finding. First, they proposed that bone formation occurring in an environment of restricted calcium supply could lead to failure of bone to achieve its full mineralized potential. Alternatively, given the extensive hydroxyapatite surface area that reaches chemical equilibrium with bone interstitial and marrow fluids, a deficiency in dietary calcium or a decrease in calcium absorption efficiency, perhaps brought about by marginal vitamin D status, could slightly decrease bone fluid Ca2, with a shift in chemical equilibrium that favors calcium loss. To maintain electrical neutrality, bone calcium would then be replaced by sodium or magnesium. There is at present no basis for favoring either of these models, and the results of the calcium intervention study [25] are compatible with both. In fact, two other mechanisms also merit consideration. To provide additional carbonate and phosphate buffer to the extracellular fluid, the labile bone mineral pool is mobilized during sustained metabolic acidosis, a compensation that passively increases urinary calcium excretion. The concept that skeletal depletion reflects lifelong buffering of diets rich in animal protein (so-called “acid-ash diets”) persists as a minority view regarding the pathogenesis of osteoporosis [26]. It is possible that the undermineralized matrix in some of the patients of Burnell et al. [22] reflects such a buffering process. Alternatively, in the presence of an overall increase in remodeling activity, newly remodeled osteons might not have sufficient time to become fully mineralized prior to the next wave of remodeling at the same site. If that were the case, the average mineral content of each completed osteon would be decreased, and if this process continued for a sustained period, the average degree of matrix mineralization would be reduced. With calcium and vitamin D therapy
remodeling would be suppressed and newly completed osteons would have the chance to become fully mineralized. 2. SPATIAL HETEROGENEITY OF MINERALIZATION The work of Crofts et al. [21] emphasizes the analytical complexity that must be brought to compositional studies, particularly of cortical bone, and indicates heterogeneous mineralization within the femur, and indeed, within individual osteons, that may affect bone material properties and strength. Using backscattered electron imaging, the authors described age- and locational differences in osteonal mineralization among different regions of femoral cortex. They noted a 12% decrease in ash content of bones from an older group compared to those from young adults, and reported that mineralization decreased with distance from the central Haversian canal. These results indicate a decrease in mineral content of equivalent magnitude to that shown to have substantial mechanical consequences [24]. Moreover, they support previous work [27,28] demonstrating a consistent pattern in cortical bone of relatively increased mineralization of regions immediately subtending central Haversian canals. These results have major, although imperfectly understood, implications regarding the site within bone where fractures are likely to be initiated [29]. 3. FLUORIDE ACCUMULATION IN BONE CRYSTAL Incorporation of fluoride into the hydroxyapatite crystal increases crystal brittleness, so lifelong fluoride exposure is potentially another influence on age-related changes in bone mineral properties. The fluoride content of potable water varies from almost 0 to as much as 3 – 4 mg/liter. Communities that regulate the fluoride content of drinking water to minimize dental carries provide a level of ~1 mg/liter (1 ppm). Concentrations of 3 mg/liter may be associated with mottling of dental enamel. When fluoride content is higher, the risk of skeletal fluorosis increases, particularly in hot environments where increased daily water consumption may be required. The relationship of lifelong fluoride exposure to osteoporosis is not certain. The earliest epidemiological studies suggested that higher exposure to fluoride in water was associated with fewer consequences of osteopenia [30]. Subsequent work has not confirmed this view [31], and even suggests higher fracture rates in populations exposed to high [32] or even apparently optimally regulated fluoride exposure [33]. Richards et al. [34] evaluated vertebral trabecular bone mass, mechanical strength, mineral content, and fluoride burden in cadaveric bone samples. As one might predict, bone mass and strength both decreased with age. Bone fluoride content, however, more than tripled from age 20 to 80 years. However, no independent effect of bone fluoride content could be shown on mechanical strength. The authors concluded that long-term fluoride consumption does not independently affect bone quality once the effects of age and
8
MARCUS AND MAJUMDER
sex on bone mass are taken into account. Discussion of the effect of fluoride on bone mass and rates of loss, as well as its therapeutic effects at pharmacologic amounts is presented elsewhere in Chapters 27, 74, and 75. As pointed out above, traditional descriptions of osteoporosis reflect the belief that no systematic differences can be consistently demonstrated between osteoporotic and normal bone. From the evidence described here it can be seen that although this view may be true when bone specimens are grossly compared, careful examination of biopsy material reveals mineralization to be spatially heterogeneous and variable with age and demonstrates subtle, but clinically meaningful, alterations in bone composition for at least a subgroup of osteoporotic patients.
B. Loss of Trabecular Connectivity Normal trabecular bone resembles a honeycomb. It consists of a highly connected network of vertical and horizontal plates, called trabeculae. By contrast, osteoporotic trabecular bone shows apparent replacement of plates by rods, and obvious trabecular disruption (Fig. 5), giving rise to the view that loss of connectivity is an important component of skeletal fragility in osteoporosis. In this section we review trabecular bone structure, the methods by which connectivity has been traditionally assessed, attempts to evaluate the role of connectivity in osteoporosis, and new developments in the three dimensional analysis of trabecular bone. Human trabecular bone is generally anisotropic with respect to its mechanical properties and architecture. That is
FIGURE 5
to say, the bone does not look or behave identically when held in one position compared to its appearance and behavior following rotation by 90°. Trabecular anisotropy reflects the manner in which gravitational stresses are transmitted through the skeleton. Because human locomotion is bipedal and upright, vertical trabeculae are thicker (approximately 200 m) than horizontal trabeculae [35]. By contrast, arboreal primates show isotropic trabecular bone; i.e., the bone appears identical regardless of which side is up. Trabecular distribution is heterogeneous within each vertebral body, so that central regions are dominated by thinner vertical plates. Although the vertebral shell is called cortical, it is not histologically identical to compact bone and may more properly be considered a condensation of trabecular bone [35]. Although horizontal trabeculae are shorter and thinner than vertical trabeculae, they make an important contribution to trabecular strength. Figure 6 illustrates the difference in ultimate breaking strength (called the Euler buckling load) of supported and unsupported columns of similar dimension. In this example, a single horizontal connecting element confers a fourfold increase in load-bearing capacity [36]. 1. ASSESSMENT OF TRABECULAR CONNECTIVITY ANATOMIC SPECIMENS
IN
Several laboratories have brought to bear multiple techniques for characterizing normal trabecular architecture and its changes with age. Most work has focused on vertebral microarchitecture, because vertebrae constitute an important site of clinical fracture, and because they, along with
Scanning electron micrograph of normal (left) and osteoporotic (right) vertebral trabecular bone. Note transformation of trabecular plates to rods and trabecular perforations (photograph by Dr. Jon Kosek, reproduced with permission).
CHAPTER 35 Nature of Osteoporosis
FIGURE 6
Effect of horizontal trabeculae on resistance to buckling. The pillar on the left is not stabilized by connecting elements and is assigned a critical buckling load (Pcr) of 1.0. The pillar on the right has a single horizontal connection approximately halfway along its length. Buckling strength is four-fold greater (Pcr 4.0). Reproduced from Snow-Harter and Marcus (36) with permission.
the pelvis and proximal femur, are the sole repository of red marrow in adults, bringing their trabecular surfaces into close proximity to the cells that initiate bone remodeling. With age, characteristic changes first appear in the vertebral centrum and spread radially. The trabecular network, the superior and inferior vertebral endplates, and the cortical rim all attenuate. This, together with progressive loss of trabecular connections, results in a severe loss of mechanical competence for individual vertebral bodies. The loadbearing capacity of a vertebral body approximates 1000 kg in young adults, but is markedly reduced in older people to about 100 kg [37]. Evidence has been presented that loss of horizontal trabeculae occurs earlier and to a greater extent than vertical trabeculae [38,39], a view that dominates thinking in this area. However, some reports have not confirmed preferential loss of horizontal trabeculae [40,41]. For example, Vogel et al. [41] showed that the number of vertical trabeculae is approximately twice that of horizontal trabeculae throughout adult life, and that the slopes of loss for each are parallel. Although these studies appeared to be fastidiously carried out, the number of measurements was insufficient to permit definite conclusions. Changes in trabecular structure with aging include thinning and loss of contiguity. The first systematic approach to analyzing the relative contributions of these two processes was reported by Parfitt et al. [42], who found that trabecular thickness did not decrease with age in women, forcing the conclusion that loss of trabecular connections was the major event. Subsequent work from Weinstein and Hutson [43] indicated that thinning and disruption are both important aspects of bone loss, whereas that from other laboratories [44,45] suggested that trabecular thinning occurs with age in men, but not in women. The weight of evidence, therefore, indicates the primary feature in women to be loss of entire trabecular elements [42,44 – 46].
9 Stereologic assessment of trabecular connectivity is not a trivial undertaking and involves a series of assumptions regarding isotropy and obliquity of tissue sections. Vesterby and colleagues [47 – 50] pointed out difficulties with these assumptions and proposed a new variable called the Star Volume, which they define as “the mean volume of all the parts of an object which can be seen unobscured from a random point inside the object in all possible directions.” It is thus a stereological estimate that applies without the aforementioned assumptions regarding isotropy that limit earlier methods. The mean star volume is obtained from a sufficient number of measurements to provide statistical power, generally about 200. Very long values for individual measurements are obtained only when there is no trabecular bone to obliterate the view. Using a limited number of specimens, Vesterby et al. [49] showed a striking agerelated increase in marrow space star volume of both the lumbar vertebrae and the iliac crest in men and women. Using a different approach, Vedi et al. [51] employed a technique of strut analysis that is based on the classification of trabeculae, or struts, and on the recognition of junctions between three or more struts (called nodes) and of free ends. By this method, the number of free ends is inversely related to trabecular connectivity, whereas nodes are indicative of trabecular connections. Vedi et al. [51] showed impressive correlations between the density of nodes and struts and trabecular bone volume on iliac crest biopsies. All methods discussed to this point require extrapolation of 2-D biopsy information to 3-D. Vogel et al. [41] pointed out that results will vary by up to 100% if trabeculae resemble rods as opposed to plates. They also introduced an ingenious direct approach to measure both 2-D and 3-D architecture in the same biopsy samples. The results of a small number of samples indicate that age-related decreases in trabecular bone volume are due primarily to transformation of trabecular plates into rods by multiple perforations, a process discussed by Parfitt [52]. Direct visualization of perforations in the 3-D specimens showed them to be confined to trabecular plates at younger ages, but present also in rods at older ages. Another novel technique, fractal geometry, has been widely used to characterize topological features of objects in astronomy, chemistry, and other physical sciences, and has recently been applied to bone structure [53 – 55]. Fractal analysis describes objects with rough surface features, such as seacoasts, and may be ideally suited to assess the microstructure of trabecular bone. Similarity of surface topology over a range of magnification is a fundamental property of fractal objects, that is, any piece of a fractal, appropriately magnified, resembles the whole. The surface ruggedness of a fractal may be defined by its fractal dimension (D), a mathematical term describing the manner in which the object fills space. Buckland-Wright et al. [54] combined high-resolution radiographs and fractal analysis
10
MARCUS AND MAJUMDER
to assess lumbar spine structure in a large group of postmenopausal women who were stratified into high and low BMD groups. In radiographs of healthy vertebral bodies, the presence of large vertical plates reduced the number of visible structures, so that the derived fractal dimension differed from that of porotic vertebrae, where vertical plates were reduced to narrow rods, intertrabecular space was enlarged, and previously hidden posterior structures, such as fragmented trabeculae, became increasingly visibile. The results were compatible with the view that trabecular bone loss consists of focal fenestration of vertical plates into a lattice of bars and rods [42]. Gundersen et al. [56] took fundamental issue with the validity of even attempting to estimate the connectivity of three-dimensional objects by anything less than a threedimensional sample. They adapted the Euler number (a topological feature based on the number of holes and connected elements in an object) and a method for calculating the Euler number in thin sections (the Conneuler) to skeletal analysis. Although this technique shows promise, no data concerning age- or osteoporosis-related changes have yet been presented. 2. RECENT DEVELOPMENTS TRABECULAR STRUCTURE
IN
THE
ANALYSIS
OF
As described above, evidence suggesting an important role for trabecular structure in determining bone strength coupled with ambiguities in the application of stereologic methods to this problem has led a number of investigators to invest considerable effort in developing 3-D imaging techniques to assess trabecular bone microarchitecture. In this section we summarize recent progress in this area. Three-dimensional evaluation of trabecular bone structure has received great impetus from the advent of CT scanners. These machines enable imaging of trabecular bone specimens at resolutions of 14 – 50 m, and in vivo scanners produce images at resolutions of 100 m. Figure 7 represents a three-dimensional image from a human iliac crest biopsy. As seen from the image, apart from obtaining routine measures of histomorphometry, such images provide a volumetric rendition of the three-dimensional structure. Peripheral computed tomography equipment under development in a research setting [57] has also led to the generation of in vivo images of trabecular bone architecture in the distal radius (Fig. 8). Goldstein and colleagues [58,59] prepared 8-mm cubic specimens of cadaveric trabecular bone for microCT scanning followed by mechanical testing. Results showed highly significant relationships of both the trabecular plate number and connectivity with the trabecular bone volume and that the bone volume explained 90% of the variance in bone strength. Consequently, although they found connectivity to be an important feature of skeletal integrity, its contribution to bone strength was contained within the information
FIGURE 7
A microcomputed tomography image of a specimen from a human iliac crest. The image is a three-dimensional rendering showing the rod-and-plate-like structure of the iliac crest. Image courtesy of Andres Laib, data acquired in laboratory of Dr. Peter Ruegsegger (ETH Zurich). Reproduced with permission.
provided by bone volume itself. One must keep in mind that these interesting results were based solely on material from four skeletons, none of which were osteoporotic. In recent years, the noninvasive and nonionizing nature of magnetic resonance (MR) has also been exploited to produce images of trabecular bone both in vitro and in vivo. MR techniques have been used in vitro to obtain images at resolutions as high as 50 µm isotropic [60 – 62]. Hipp et al.
11
CHAPTER 35 Nature of Osteoporosis
FIGURE 8
A microcomputed tomography image of the distal radius in vivo. The image is a three-dimensional rendering showing the cortical shell and trabecular bone. Image courtesy of Andres Laib, data acquired in laboratory of Dr. Peter Ruegsegger (ETH Zurich). Reproduced with permission.
[63] compared MR-derived stereology measures, determined from isotropic image voxels of 50 m, to those obtained using optical imaging and found good correlations, while Chung et al. [61] found good correlations of the estimated apparent trabecular bone volume (BV/TV) determined from MR images to those determined using displacement techniques. Antich et al. [60] confirmed that application of MR techniques to iliac crest biopsy specimens may be used to monitor changes in trabecular bone structure after fluoride therapy. In a recent study involving a set of 94 trabecular bone cubes from 13 cadavers and specimens from the calcaneus , distal femur, proximal femur, and vertebral bodies, the relationship between MRderived 3D measures of trabecular architecture, bone mineral density and biomechanical properties have been demonstrated [64].
Shown in Fig. 9, reverse gray scale images of two specimens from the vertebral body show clear visual differences in structure and density. The spatial resolution (117 117 300 m) of these images is typically greater than, or on the order of, trabecular bone dimensions. This gave rise to partial volume effects, but although the MR image derived 3D measures of trabecular architecture differ from those measured using 20-m optical images, there is good correlation between the two sets of measures [65]. The correlation coefficients are 0.69, 0.89, and 0.78 (P 0.01) for the apparent BV/TV, trabecular spacing (Tb.Sp), and trabecular number (Tb.N), respectively. The correlation between the trabecular thickness measure was poor, R 0.06 (P 0.84) [65]. The lower correlations specifically for measures such as the apparent Tb.Th, are expected based on the limited spatial resolution of the MR images compared to the trabecular sizes [66,67].
12
MARCUS AND MAJUMDER
FIGURE 9
Two vertebral bodies imaged using magnetic resonance imaging at a resolution of 117 117 300 m using a 1.5 T scanner.
Recent introduction of new micro-finite element (microFE) techniques should permit calculatation of cancellous bone mechanical properties directly from high-resolution images of its internal architecture. Computer reconstructions made from MR images were converted to micro-FE models from which bone elastic properties have been calculated [68]. The results of the MR-based FE models correlated well to those of the more accurate micro-CT based models, thus emphasizing the ability not only to compute architectural features from these images, but also to translate them into biomechanical parameters. In human MR studies, the spatial image resolution has ranged from 78 to 195 m in plane to 300 to 1000 m in the slice orientation depending on the anatomical site being examined [66,69 – 71]. Figure 10 illustrates representative calcaneal images, clearly depicting the
FIGURE 10
spatial heterogenity of trabecular bone microarchitecture. By selecting regions of interest extending from the joint line and 3 cm into the shaft in the radius and the posterior region in the calcaneus differences in trabecular architecture between patients with and without hip fractures have been demonstrated [72]. 3. RELATIONSHIP OF TRABECULAR CONNECTIVITY OSTEOPOROTIC FRACTURE
TO
Limited information is available concerning the contribution of trabecular connectivity to osteoporotic fractures. Kleerekoper et al. [73] compared indices of connectivity on iliac crest biopsy specimens from patients with osteoporotic fracture to those from nonfracture controls who had approximately the same trabecular and cortical bone mass.
Representative radius and calcaneus images. Representative radiograph (lateral projection) of a calcaneus reflects the heterogenity in bone content at this skeletal site (left). (Right) A sagittal MR section (500 m thick, 195 195 in plane resolution) showing trabecular bone microarchitecture in the calcaneus. The heterogeneity in trabecular structure seen as thicker, and a greater number of trabeculae in the subtalar portion is visually seen from the MR image.
13
CHAPTER 35 Nature of Osteoporosis
The fracture group had a 20% decrease in trabecular plate density and increases of equal magnitude in both the separation and thickness of trabecular plates. In other words, the negative effect of having fewer, more sparsely placed trabeculae was not counteracted by the relative thickness of the residual trabeculae. Recker [74] has reported a similar comparison. In his study, biopsies from fracture patients showed an 11% decrease in trabecular number, a 37% decrease in trabecular connecting nodes, and a 37% increase in trabecular free ends compared to the control biopsies. A report by Khosla et al. [75] suggests that trabecular disruption characterizes younger patients with osteoporosis as well. The authors evaluated histomorphometric features of iliac crest biopsies from 18 patients with idiopathic osteoporosis and 18 normal controls, average age 35 years. Although differences in the respective trabecular bone volumes for the two groups (16% vs 24%) were highly significant, differences in average trabecular wall thickness (40.0 mm vs 46.5 mm) were much smaller, implying that at least some of the deficit in trabecular bone mass must be due to dropout of entire trabecular units, i.e., a degradation in trabecular connectivity. Weinstein and Majumdar [53] applied fractal analysis to photographs of iliac crest biopsies from control subjects and from patients with osteoporosis. Although groups were not separable by traditional two-dimensional measures of microarchitecture, the fractal dimension, D, differed significantly. Taken together, results from these studies support a role for connectivity as a contributor to fracture. A dissenting view was registered by Compston et al. [76], who found no greater trabecular disruption in osteoporotic patients than in controls and suggested that structural changes in primary osteoporosis do not differ qualitatively from those of age-related bone loss. Microarchitectural comparisons have now been extended to the hip. Ciarelli et al. [77] prepared cubes of trabecular bone from the proximal femurs of women who had sustained hip fractures and from cadaveric nonfracture control bones. Control specimens had higher trabecular bone volume, trabecular number, and connectivity than did specimens from fracture patients. No differences were observed in mean trabecular thickness. Mechanical testing showed greater ultimate stress tolerance in the control bones. Furthermore, fracture specimens showed greater anisotropy of trabecular orientation, with fewer trabecular elements lying in a plane that was transverse to the primary load axis. It remains premature to offer a firm conclusion regarding an independent role for trabecular connectivity in osteoporotic fracture. Reasons for this uncertainty are multiple and include an apparently enormous bone-to-bone heterogeneity (variations up to 100%!) in trabecular volume and structure throughout the axial skeleton [78] as well as inadequacies of stereological methods. Ultimate resolution of this question will likely require more extensive application of the 3D techniques described above.
C. Accumulation of Cement Lines Secondary Haversian bone has poorer tensile strength and material properties than does primary lamellar bone [79,80]. In part, this reduced tensile strength may reflect the long-term accumulation of cement lines, which are the visible residua of previous bone remodeling events. Cement lines are observed as thin ribbons of loosely woven collagen fibers, distinguished easily under light microscopy from the surrounding lamellar bone (Fig. 11). They demarcate the area of deepest bone resorption and form the scaffold upon which new bone is deposited. Following completion of a remodeling cycle, the new, apparently pristine lamellar bone is interrupted by an area of structurally weaker woven bone. Carter and Hayes [81] have demonstrated debonding and disruption of bone at the cement line as a consequence of fatigue microdamage, indicating that cement lines represent loci of structural least resistance. As a consequence of many years of bone remodeling, both cortical and trabecular bone show a plethora of cement lines as well as evidence for previously remodeled areas
FIGURE 11
Cement lines. Reproduced from Carter and Hayes (81) with permission. (Bottom) Magnified view of boxed area in top panel illustrating debonding of the cement line following application of mechanical stress.
14
MARCUS AND MAJUMDER
that have been partly to completely removed (Fig. 12). For any given bone density, such highly remodeled bone is structurally weaker than is primary lamellar bone of younger adults. Choi and Goldstein [82] compared the fatigue and mechanical properties of single trabeculae to those of cortical bone specimens of equal size. The trabecular specimens had significantly lower elastic moduli and lower fatigue strength than cortical specimens, despite having higher mineral density. The authors proposed that cement line accumulation had decreased the mechanical properties of the trabecular bone.
D. Increased Cortical Porosity Strictly speaking, porosity is a measure of the prevalence and size of holes within bony cortex. Such holes represent Haversian canals, osteocyte lacunae, and new cutting cones that have been produced by alterations in systemic or local factors favoring resorption over bone forma-
tion (Fig. 13). Cortical porosity is difficult to assess, since most cortical holes are smaller than the resolution capability of the measuring instruments. Even with biopsy, this has been a difficult problem, since much of the process appears as “trabecularization” of the endocortex. In other words, porosity is sufficiently great that the histomorphometrist may read a highly porous endocortex as a simple extension of the trabecular bone. When bone is acquired during growth, primary Haversian canals constitute the major, if not exclusive, source of cortical porosity. Later, as a consequence of continuous remodeling, secondary Haversian systems gradually accumulate. Martin [83] used a rib model to analyze the relationships among the components of cortical porosity and age in humans. Total porosity actually decreased between ages 10 and 40, but rose progressively thereafter. The contribution of secondary Haversian canals to total porosity increased dramatically until age 40 years, following which it remained stable. Progressive cortical porosity after age 40 was attributable largely to an expanded remodeling space due to an increase in remodeling activation rate. Thus, increased cortical porosity is a feature of normal skeletal aging. Increased porosity is also characteristic of PTHdependent bone resorption. Therefore, to the extent that
FIGURE 12
Extensive cortical remodeling in trans-ileal biopsy specimen from an elderly woman (photograph courtesy of R. R. Recker with permission).
FIGURE 13
Porosity in human cortical bone (photograph courtesy of D. R. Carter with permission).
15
CHAPTER 35 Nature of Osteoporosis
hypersecretion of PTH in older individuals promotes increased remodeling activity, increased cortical porosity due to expanded remodeling space will occur. In theory, interventions designed to suppress PTH secretion and constrict the remodeling space should constrain the accumulation of cortical porosity with age.
E. Bone Fatigue Accumulation of fatigue with sustained use is a fundamental property of all materials. Although many investigators treat the terms “fatigue damage” and “microscopic damage” interchangeably, this is a simplistic construction, since cyclic loading produces functional, yet invisible, changes over time. Repetitive loading of compact bone leads to progressive deterioration in the modulus of elasticity, and ultimately to structural failure. Indeed, Carter and Hayes [84] showed that bone has relatively poor fatigue properties compared to a number of common building materials. Fatigue accumulates over time. Assuming that the intensity of loading remains constant, there is little difference in cumulative fatigue if materials are subjected to 10,000 cycles each day for 5 days or to 50,000 cycles at a single session. The importance of fatigue as a contributor to compact bone failure has been questioned by Schaffler et al. [85], who subjected standardized bone plugs to many loading cycles at relatively low, physiologic strains. A 5% deterioration in modulus of elasticity was observed after about 106 loading cycles, but no further change was observed even after 20 106 cycles, leading the authors to conclude that, under physiological conditions of low-magnitude repetitive strain, compact bone shows greater resistance to fatigue than would be predicted from the work of Carter and Hayes [84] and others, who carried out tests at higher strain magnitudes. Schaffler et al. [85] calculated that at the strains typically measured during running, 1200 – 1500 microstrain, cyclic loads equivalent to 11,000 miles of running could be sustained before repair processes would be necessary. Nonetheless, the relatively high frequency of clinical “stress” or fatigue fractures with overuse, particularly when habitual loading has been markedly and precipitously increased, certainly argues for the clinical relevance of fatigue accumulation in vivo. Schaffler et al. [86] suggest that vigorous activity might subject bone to brief episodes of very high strain, which would drastically reduce the number of cycles necessary to produce failure. Furthermore, if repetitive loading in vivo were to stimulate remodeling activity, the initial resorption phase would temporarily increase the remodeling space, thereby decreasing bone strength. Continued loading of such a bone, even at relatively low intensity, might then increase the risk of fracture. Recent evidence indicates that bone fatigue is as-
sociated with osteocytic apoptosis, which appears to be an important step in targeting bone for a subsequent remodeling event [87]. The contribution of fatigue to osteoporotic fracture is therefore complex. Certainly, it is important to the extent that by stimulating bone remodeling fatigue promotes the visible deterioration in cortical and trabecular microarchitecture described elsewhere in this chapter. Furthermore, cortical bones with extensive secondary Haversian remodeling, such as occurs normally with age, show fatigue earlier than does primary lamellar bone [86,88]. Beyond that, however, even prior to the emergence of visible damage, changes in fundamental bone material properties accumulate due to repetitive loading over time, and these subtle changes are likely contributors to overall bone fragility.
III. CONCLUSIONS At the beginning of this chapter we discussed the limitations of a bone mass-based diagnosis of osteoporosis [12]. A primary difficulty with such a definition is that its sensitivity to factors known collectively as “bone quality” has not been clarified, and it is tempting to attribute the diagnostic ambiguities of BMD measurements to their failure to account for these features. Although these concerns persist, the fact that information contained in the BMD estimate accounts in part for some of the important geometric, material, and microarchitectural properties, solidifies its rationale as a diagnostic criterion. Certainly, any substantial degree of matrix undermineralization would be reflected in a lower BMD, and the data of Goldstein et al. [59] suggest that trabecular disruption of sufficient magnitude to be mechanically important would also register as a bone mineral deficit, and therefore as a lower BMD. Qualitative features that would not be included in a BMD assessment include cement lines, unremodeled fatigue damage, and fluoride accumulation. The question remains whether osteoporosis should be viewed as one or more unique diagnostic entities, as is the case for Paget’s disease, or whether it is more useful to consider it a condition of skeletal fragility resulting from a stochastic process, in which contributory factors include age, body size, adequacy of peak bone mass, degree of adult bone loss, and accumulation of qualitative impairments. Since the overall trajectory over time of adolescent bone acquisition and adult bone loss appears to be universal, the only basis for considering osteoporosis one or more distinct entities would be a demonstration that its qualitative abnormalities, such as those discussed above, are restricted to those patients who have suffered a fragility fracture. Although evidence remains incomplete, it seems unlikely that such specificity will be validated for most of these abnormalities.
16
MARCUS AND MAJUMDER
References 1. F. Albright and E. C. Reifenstein, Jr., “The Parathyroid Glands and Metabolic Bone Disease: Selected Studies,” p. 162. Williams and Wilkins, Baltimore, 1948. 2. B. L. Riggs, H. W. Wahner, E. Seeman, K. P. Offord, W. L. Dunn, R. B. Mazess, K. A. Johnson, and L. J. Melton III, Changes in bone mineral density of the proximal femur and spine with aging. Differences between the postmenopausal and senile osteoporosis syndromes. J. Clin. Invest. 70, 716 – 723 (1982). 3. T. Appelboom and J-J Body, The antiquity of osteoporosis: More questions than answers. Calcif. Tissue Int. 53, 367–369 (1993). 4. B, Krølner and S.P. Nielsen, Bone mineral content of the lumbar spine in normal and osteoporotic women: Cross-sectional and longitudinal studies. Clin. Sci. 62, 329 – 336 (1982). 5. R. Pacifici, N. Susman, P. L. Carr, S. J. Birge, and L. V. Avioli, Single and dual energy tomographic analysis of spinal trabecular bone. A comparative study in normal and osteoporotic women. J. Clin. Endocrinol. Metab. 64, 209 – 214 (1987). 6. S. M. Ott, R. F. Kilcoyne, and C. H. Chesnut III, Ability of four different techniques of measuring bone mass to diagnose vertebral fractures in postmenopausal women. J. Bone. Miner. Res., 2, 201 – 210 (1987). 7. K. G. Faulkner, S. R. Cummings, D. Black, L. Palermo, C-C Glüer, and H. K. Genant, Simple measurement of femoral geometry predicts hip fracture: The study of osteoporotic fractures. J. Bone Miner. Res. 8, 1211 – 1218 (1993). 8. R. D. Wasnich, P. D. Ross, L. K. Heilbrun, and J. M. Vogel, Prediction of postmenopausal fracture risk with use of bone mineral measurements. Am. J. Obstet. Gynecol. 153, 745 – 751 (1985). 9. S. L. Hui, C. W. Slemenda, and C. C. Johnston, Jr., Baseline measurement of bone mass predicts fracture in white women. Ann. Intern. Med. 111, 355–361 (1989). 10. S. R. Cummings, D. M. Black, M. C. Nevitt, W. S. Browner, J. A. Cauley, H. K. Genant, S. R. Mascioli, J. C. Scott, D. G. Seeley, P. Steiger, and T. Vogt, for the SOF Research Group, Appendicular bone density and age predict hip fracture in women. JAMA 263, 665 – 668 (1990). 11. L. J. Melton III, E. J. Atkinson, W. M. O’Fallon, H. W. Wahner, and B. L. Riggs, Long-term fracture prediction by bone mineral assessed at different skeletal sites. J. Bone Miner. Res. 8, 1227 – 1234 (1993). 12. J. A. Kanis, L. J. Melton III, C. Christiansen, C. C. Johnston, Jr., and N. Haltaev, Perspective: The diagnosis of osteoporosis. J. Bone Miner. Res. 9, 1137 – 1142 (1994). 13. Consensus Development Conference,1991, Prophylaxis and treatment of osteoporosis. Am. J. Med. 90, 107–110 (1991). 14. F. McCabe, L.-J. Zhou, C. R. Steele, and R. Marcus, Non-invasive assessment of ulnar bending stiffness in women. J. Bone Miner. Res. 6, 53 – 59 (1991). 15. P. Kann, S. Graeben, and J. Beyer, Age-dependence of bone material quality shown by measurement of frequency of resonance in the ulna. Calcif. Tissue Int. 54, 96 – 100 (1994). 16. G. H. Brandenburger, Clinical determination of bone quality: Is ultrasound an answer? Calcif. Tissue. Int. 53 (Suppl. 1), S151 – S156 (1993). 17. S. Sherman, R. P. Heaney, A. M. Parfitt, E. C. Hadley, and C. Dutta, NIA Workshop on aging and bone quality. Calcif. Tissue Intl. 53, (Suppl. 1), S1 – S180 (1993). 18. M. Grynpas, Age and disease-related changes in the mineral of bone.Calcif. Tissue. Int. 53 (Suppl. 1), S57–S64 (1993). 19. J. Jowsey, Age changes in human bone. Clin. Orthop. 17, 210 – 218 (1960). 20. D. J. Ortner, Aging effects on osteon remodeling. Calcif. Tissue Res. 18, 27 – 36 (1975). 21. R. D. Crofts, T. M. Boyce, and R. D. Bloebaum, Aging changes in osteon mineralization in the human femoral neck. Bone 15, 147– 52 (1994).
22. J. M. Burnell, D. J. Baylink, C. H. Chesnut III, M. W. Mathews, and E. J. Teubner, Bone matrix and mineral abnormalities in postmenopausal osteoporosis. Metabolism 31, 1113 – 1120 (1982). 23. G. P. Vose and A. L. Kubala, Jr.,. Bone strength-its relationship to x-ray determined ash content. Hum. Biol. 31, 261 – 270 (1959). 24. J. D. Currey, The mechanical consequences of variation in the mineral content of bone. J. Biomech. 2, 1 – 11 (1969). 25. J. M. Burnell, D. J. Baylink, C. H. Chesnut III, and E. J. Teubner, The role of skeletal calcium deficiency in postmenopausal osteoporosis. Calcif. Tissue Int. 38, 187 – 92 (1986). 26. U. Barzel, The effect of excessive acid feeding on bone. Calcif. Tissue Res. 4, 94 – 100 (1969). 27. R. Amprino and A. Engstrom, Studies on x-ray absorption and diffraction of bone tissue. Acta Anat. 15, 1 – 22 (1952). 28. D. J. Ortner and D. W. Von Endt, Microscopic and electron microprobe characterization of the sclerotic lamellae in human osteons. Isr. J. Med. Sci. 7, 480 – 482 (1971). 29. T. M. Boyce and R. D. Bloebaum, Cortical aging differences and fracture implications for the human femoral neck. Bone, 14, 769 –778 (1993). 30. N. C. Leone, C. A. Stevenson, T. F. Hilbish, and M. C. Sosman, A roentgenologic study of a human population exposed to high-fluoride domestic water (a 10-year study). Am. J. Roentg 74, 874 – 885 (1955). 31. J. Avorn and L. C. Niessen, Relationship betwen long bone fractures and water fluoridation. Gerodontology 2, 175 – 179 (1986). 32. M. R. Sowers, M. K. Clark, M. L. Jannausch, and R. B. Wallace, A prospective study of bone mineral content and fracture in communities with differential fluoride exposure. Am. J. Epidemiol. 133, 649 – 660 (1991). 33. C. Cooper, C. A. C. Wickham, D. J. R. Barker, and S. J. Jacobsen, Water fluoridation and hip fracture. JAMA 266, 513 – 514 (1991). 34. A. Richards, Li. Mosekilde, and C. H. Søgaard, Normal age-related changes in fluoride content of vertebral trabecular bone-relation to bone quality. Bone 15, 21–26 (1994). 35. L. Mosekilde, Vertebral structure and strength in vivo and in vitro. Calcif. Tissue Res. 53 (Suppl. 1), S121 – S126 (1993). 36. C. Snow-Harter, and R. Marcus, Exercise, bone density, and osteoporosis. Exercise Sport Sci. Rev. 19, 351 – 388 (1991). 37. Li. Mosekilde and Le. Mosekilde, Sex differences in age-related changes in vertebral size, density and biomechanical competence in normal individuals. Bone 11, 67 – 73 (1990). 38. P. J. Atkinson, Variation in trabecular structure of vertebrae with age. Calcif. Tissue Res. 1, 24 – 32 (1967). 39. Li. Mosekilde, Age-related changes in vertebral trabecular bone architecture-assessed by a new method. Bone 9, 247 – 250 (1988). 40. B. D. Snyder, S. Piazza, W. T. Edwards, and W.C. Hayes, Role of trabecular morphology in the etiology of age-related vertebral fractures. Calcif. Tissue Int. 53 (Suppl. 1), S14 – S22 (1993). 41. M. Vogel, M. Hahn, and G. Delling, Relation beween 2- and 3-dimensional architecture of trabecular bone in the human spine. Bone 14, 199 – 203 (1993). 42. A. M. Parfitt, C. H. E. Mathews, A. R. Villanueva, M. Kleerekoper, B. Frame, and D. S. Rao, Relationship between surface, volume and thickness of iliac trabecular bone in aging and in osteoporosis: Implications for the microanatomic and cellular mechanisms of bone loss. J. Clin. Invest. 72, 1396 – 1409 (1983). 43. R. S. Weinstein and M. S. Hutson, Decreased trabecular width and increased trabecular spacing contribute to bone loss with ageing. Bone 8, 137 – 142 (1987). 44. J. E. Aaron, N. B. Makins, and K. Sagreiya, The microanatomy of trabecular bone loss in normal aging in men and women. Clin. Orthop. Rel. Res. 215, 260 – 271 (1987). 45. R. W. E. Mellish, N. J. Garrahan, and J. E. Compston, Age-related changes in trabecular width and spacing in human iliac crest biopsies. Bone and Miner. 6, 331 – 338 (1989). 46. D. H. Birkenhäger-Frenkel, P. Courpron, E. A. Hüpscher, E. Clermonts, M.F. Coutinho, P. I. M. Schmitz, and P.J. Meunier,
17
CHAPTER 35 Nature of Osteoporosis
47.
48.
49.
50.
51.
52. 53. 54.
55. 56.
57. 58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
Age-related changes in trabecular bone structure. Bone and Miner. 4, 197 – 216 (1988). A. Vesterby, H. J. G. Gundersen, and F. Melsen, Star volume of marrow space and trabeculae of the first lumbar vertebra: Sampling efficiency and biological variation. Bone 10, 7 – 13 (1989). A. Vesterby, H. J. G. Gundersen, F. Melsen, and L. Mosekilde, Normal postmenopausal women show iliac crest trabecular thickening on vertical sections. Bone, 10, 333 – 339 (1989). A. Vesterby, Star volume of marrow space and trabeculae in iliac crest: Sampling procedure and correlation to star volume of first lumbar vertebrae. Bone 11, 149 – 55 (1990). A. Vesterby, H. J. G. Gundersen, F. Melsen, and L. Mosekilde, Marrow space star volume in the iliac crest. Decreases in osteoporotic patients after continuous treatment with fluoride, calcium, and vitamin D2 for 5 years. Bone 12, 33 – 7 (1991). S. Vedi, J. E. Compston, A. Webb, and J. R. Tighe, Histomorphometric analysis of bone biopsies from the iliac crest of normal British subjects. Metab. Bone Dis and Relat. Res. 4, 231 – 236 (1982). A. M. Parfitt, Trabecular bone architecture in the pathogenesis and prevention of fracture. Am. J. Med. 82(Suppl. 1B), 68 – 72 (1987). R. S. Weinstein and S. Majumdar, Fractal geomery and vertebral compression fractures. J. Bone Miner. Res. 9, 1797 – 1802 (1994). J. C. Buckland-Wright, J. A. Lynch, J. Rymer, and I. Fogelman, Fractal signature analysis of macroradiographs measures trabecular organization in lumbar vertebrae of postmenopausal women. Calcif. Tissue Int. 54, 106 – 112 (1994). P. Caligiuri, M. L. Giger, and M. Favus, Multifractal radiographic analysis of osteoporosis. Med. Phys. 21, 503 – 508 (1994). H. J. G. Gundersen, R. W. Boyce, J.R. Nyengaard, and A. Odgaard, The conneulor: Unbiased estimation of connectivity using physical disectors under projection. Bone 14, 217 – 222 (1993). R. Muller, T. Hildebrand, H. J. Hauselmann, and P. Ruesegger, J. Bone Miner. Res. 11, 1745 – 1750 (1996). L.A. Feldkamp, S. A. Goldstein, A. M. Parfitt, G. Jesion, and M. Kleerekoper, The direct examination of three– dimensional bone architecture in vitro by computed tomography. J. Bone Miner. Res. 4, 3 – 11 (1989). S. A. Goldstein, R. Goulet, and D. McCubbrey, Measurement and significance of three-dimensional architecture to the mechanical integrity of trabecular bone. Calcif. Tissue Int. 53 (Suppl.), S127 – S133 (1993). P. P. Antich, R. P. Mason, R. McColl, J. Zerwech, and C. Y. C. Pak, Trabecular architecture studies by 3D MRI microscopy in bone biopsies. J. Bone Miner. Res. 9S1, 327 (1994). H. W. Chung, F. W. Wehrli, J. L. Williams, and S. L. Wehrli, Three dimensional nuclear magnetic resonance micro-imaging of trabecular bone J. Bone Miner. Res. 10, 1452 – 1461 (1995). R. D. Kapadia, W. B. High, H. A. Soulleveld, D. Bertolini, and S. K. Sarkar, Magnetic resonance microscopy in rat skeletal research Magn. Reson. Med. 30, 247 – 250 (1993). J. A. Hipp, A. Jansujwicz, C. A. Simmons, and B. Snyder, Trabecular bone morphology using micro-magnetic resonance imaging. J. Bone Miner. Res. 11, 286 – 292 (1996). S. Majumdar, M. Kothari, P. Augat, D. C. Newitt, T. M. Link, J. C. Lin, T. Lang, Y.Lu, and H. K. Genant, High resolution magnetic resonance (MR): Three dimensional trabecular bone architecture and biomechanical properties. Bone May 22, 445 – 454 (1998). M. Kothari,T. Keaveny, J. Lin, D. C. Newitt, H. K Genant, and S. Majumdar, Impact of spatial resolution on the prediction of trabecular architecture parameters. Bone 22, 437 – 443 (1998). S. Majumdar, D. C. Newitt, A. Mathur, et al., Magnetic resonance imaging of trabecular bone structure in the distal radius: Relationship with x-ray tomographic microscopy and biomechanics. Osteoporosis Int. 6, 376 – 385 (1996). R. Muller, M. Hahn, M. Vogel, G. Delling, and P. Ruegsegger, Morphometric analysis of non-invasively assessed bone biopsies:
68.
69.
70.
71.
72.
73.
74. 75.
76.
77.
78.
79. 80.
81. 82.
83. 84. 85.
86. 87.
88.
Comparison of high resolution computed tomography and histologic sections. Bone , 8, 215 – 220 (1996). B. van Rietbergen, S. Majumdar, W. Pistoia, D. C. Newitt, M. Kothari, A. Laib, and P. Rüegsegger, Assessment of cancellous bone mechanical properties from micro-Fe models based on micro-Ct, Pqct and Mr images. Stud. Technol. Healthcare 6(5 – 6), 413 – 20 (1998). H. Jara, F. W. Wehrli, H. Chung, and J. C. Ford, High-resolution variable flip angle 3D MR imaging of trabecular microstructure in vivo. Magn. Reson. Med. 29, 528 – 539 (1993). B. Kuehn, B. Stampa, M. Heller, and C. Glueer, In vivo assessment of trabecular bone structure of the human phalanges using high resolution magnetic resonance imaging. Osteoporosis Int. 7, 291 (1997). X. Ouyang, K. Selby, P. Lang, et al., High resolution MR imaging of the calcaneus: Age-related changes in trabecular structure and comparison with DXA measurements. Calcif. Tissue Int. 60, 139 – 147 (1997). T. Link, S. Majumdar, J. Lin, et al., Can texture analysis of high resolution MRI of calcaneus be used to predict osteoporotic hip fractures? J. Bone Miner. Res. (2001) (in press). M. Kleerekoper, A. R. Villanueva, J. Stanciu, D. S. Rao, and A. M. Parfitt, The role of three-dimensional trabecular microstructure in the pathogenesis of vertebral compression fractures. Calcif. Tissue Int 37, 594 – 597 (1985). R. R. Recker, Architecture and vertebral fracture. Calcif. Tissue Int., 53 (Suppl.), S139 – S142 (1993). S. Khosla, E. G. Lufkin, S. F. Hodgson, L. A. Fitzpatrick, and L. J. Melton III, Epidemiology and clinical features of osteoporosis in young individuals. Bone 15, 551 – 555 (1994). J. E. Compston, N. J. Garrahan, P. I. Croucher, C. D. P. Wright, and K. Yamaguchi, Quantitative analysis of trabecular bone structure: Studies in patients with osteoporosis and in women treated with tamoxifen. Bone 13, A8 (1992). T. E. Ciarrelli, D. P. Fyhrie, M. B. Schaffler, and S. A. Goldstein, Variations in three-dimensional cancellous bone architecture of the proximal femur in female hip fractures and in controls. J. Bone Miner. Res. 15, 32 – 40 (2000). H Ritzel, M. Amling, M. Vogel, M. Posl, M. Hahn, M. Werner, and G. Delling, Spongiosa structure and polyostotic heterogeneity in osteoporosis. Mechanism of bone transformation, morphology, clinical significance. Pathologe (German) 17, 68 – 77 (1996). S. Saha and W. C. Hayes, Relations between tensile impact properties and microstructure of compact bone. Calcif. Tissue Res. 24, 65 – 72 (1977). R. Vincentelli and M. Grigorov, The effect of Haversian remodeling on the tensile properties of human cortical bone. J. Biomech. 18, 201 – 207 (1985). D. R. Carter and W. C. Hayes, Compact bone fatigue damage. A microscopic examination. Clin. Orthop. Relat. Res. 127, 265 – 274 (1977). K. Choi and S. A. Goldstein, A comparison of the fatigue behavior of human trabecular and cortical bone tissue. J. Biomech. 25, 1371 – 1381 (1992). R. B. Martin, On the signifiance of remodeling space and activation rate changes in bone remodeling. Bone 12, 391 – 400 (1991) D. R. Carter and W. C. Hayes, The compressive behavior of bone as a two phase porous structure. J. Bone Joint Surg. 59A, 954 – 962 (1977). M. B. Schaffler, E. L. Radin, and D. B. Burr, Long-term fatigue behavior of compact bone at low strain magnitude and rate. Bone 11, 321 – 326 (1990). M. B. Schaffler and D. B. Burr, Stiffness of compact bone: Effects of porosity and density. J Biomech, 21, 13 – 16 (1988). O. Verborgt, G. J. Gibson, and M. B. Schaffler, Loss of osteocyte integrity in association with microdamage and bone remodeling after fatigue in vivo. J. Bone Miner. Res. 15, 60 – 67 (2000). D. B. Burr, M. B. Schaffler, and R. G. Frederickson, Composition of the cement line and its mechanical role as an interface in human compact bone. J. Biomech. 21, 939–945 (1988).
CHAPTER 36
Local and Systemic Factors in the Pathogenesis of Osteoporosis LAWRENCE G. RAISZ
I. II. III. IV. V. VI.
Department of Endocrinology and Metabolism, University of Connecticut Health Center, Farmington, Connecticut 06032
VII. Colony-Stimulating Factors VIII. Further Considerations of Interactions of Systemic Hormones and Local Mediators IX. Conclusions References
Introduction Limited Role of Systemic Hormones Local Factors Cytokines Prostaglandins Growth Factors
I. INTRODUCTION
The action of estrogen on cytokine production has been reviewed elsewhere in this volume (Chapters 10, 13, and 41), as have the effects of calcium regulating and other systemic hormones. In this chapter, I will summarize current evidence that supports a role for local factors acting both directly and as mediators of systemic hormones. Syndromes involving cancellous bone loss, which include vertebral crush fractures and Colles’ fractures, as well as fractures of the pelvis and trochanteric fractures of the hip, are most likely to involve local factors [5,6]. These comprise a substantial proportion of osteoporotic fractures and often occur with minimal or no trauma. Fractures involving cortical bone, although usually associated with some trauma, are also more likely to occur in individuals with bone loss. Thus, high bone turnover is associated with cortical bone loss and an increased risk of hip fracture [7,8]. The relative importance of high bone turnover with increased remodeling and increased resorption, as opposed to low bone turnover with decreased formation in the
The concept that local bone factors play an important role in the pathogenesis of osteoporosis has developed from several lines of evidence [1 – 4]. Among these are the following: (i) it has been difficult to demonstrate relevant differences in the production of systemic hormones between osteoporotic patients and matched controls, (ii) many local factors which regulate bone metabolism have been identified, and (iii) estrogen deficiency, as well as changes in other systemic hormones which are thought to play pathogenetic roles in osteoporosis, have marked effects on local factors. Since this chapter was written for the first edition of Osteoporosis there have been many new observations which increase our understanding of the interaction between local and systemic factors and expand our concepts of the cellular mechanisms by which these interactions can result in decreased bone mass and increased bone fragility.
OSTEOPOROSIS, SECOND EDITION VOLUME 2
19
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
20 pathogenesis of osteoporosis, is still not fully resolved, although a number of studies indicate that rapid bone loss is likely to be associated with high turnover [8,9].
II. LIMITED ROLE OF SYSTEMIC HORMONES There is evidence to support a primary role for either calcium regulating or systemic hormones other than sex hormones in the pathogenesis of cancellous bone loss. Agerelated calcium deficiency, which involves decreased intake of both calcium and vitamin D, as well as decreased formation of, and responsiveness to, calcitriol, results in secondary hyperparathyroidism, and this probably plays a role in cortical bone loss with age [10]. However, in vertebral crush fracture patients, serum parathyroid hormone (PTH) concentrations are not increased. On the contrary, there is a blunted PTH response to hypocalcemia [11]. It is plausible that this blunted response, reflecting the enhancement of bone resorption by local factors, moves calcium from bone to blood without requiring PTH, and hence parathyroid responsiveness is decreased. Calcitonin deficiency has not been demonstrated in vertebral crush fracture patients [12]. Circulating IGF-I and IGF-I-BP3 levels decrease with age [13]. These changes could aggravate bone loss, but have not been found to be greater in most osteoporotic patients, although they have been implicated in idiopathic osteoporosis in men [14]. Glucocorticoid excess can clearly produce osteoporosis, largely by inhibiting bone formation [15]. There is little evidence that glucocorticoid excess is important in primary osteoporosis, although high trough glucocorticoid levels in the evening have been associated with decreased bone mass in older men [16]. Sex hormones are critically important in the development and maintenance of the skeleton, but their mechanisms of action are still not clear. Estrogen is important in both sexes [17]. Males with a defective estrogen receptors or lack of aromatase, which converts androgens to estrogen, show failure of epiphyseal closure and high bone turnover with low bone mass, despite full responsiveness to androgen [18,19]. In male rats with androgen resistance, there is bone loss after orchiectomy, presumably due to loss of estrogen formed from androgen by aromatization [20]. In men with aromatase deficiency treatment with estrogen can decrease bone turnover and increase bone mass [19]. There is also evidence that androgens have direct effects on bone [21]. Thus, the sex hormones appear to work in concert, probably through separate but overlapping pathways. A generalization, consistent with current data, is that androgens increase bone mass indirectly by increasing muscle mass and directly by stimulating bone formation, particularly in the periosteum, while estrogens prevent bone loss by decreasing trabecular and endocortical bone resorption and decreasing turnover.
LAWRENCE G. RAISZ
A role for progesterone in the skeleton has been proposed but is not established. Progesterone is a potent mitogen in bone cell cultures [22,23], but there is little direct evidence for an anabolic action in adult humans [24]. The possibility that other reproductive hormones, including gonadotropins, inhibin, and prolactin, have skeletal effects has not been adequately explored.
III. LOCAL FACTORS The interaction of local and systemic factors in regulating bone metabolism is essential for the skeleton to play its dual roles as a structure for locomotion and protection of internal organs and as a storehouse for mineral. The structure of the skeleton is determined by mechanical forces. Recent studies on the effects of loading and unloading on the skeleton have identified a number of local factors which may mediate cellular responses, including nitric oxide, prostaglandins insulin-like growth factor-I (IGF-I) and glutamate [25 – 27]. Production of these factors may be initiated by fluid shear stress exerted on the osteocyte-osteoblast canalicular network [28]. Local regulation also involves a complex interplay between cells in the marrow and in bone. Both hematopoietic precursors for osteoclasts and mesenchymal precursors for osteoblasts are present in the marrow. Marrow cells may produce some of the local factors which act on bone cells [29,30]. Other tissues, including the vasculature, cartilage, muscle, tendon, and synovium, as well as blood-borne elements, including platelets and leukocytes, can produce factors that regulate bone metabolism [31]. There is as yet no evidence for a role for extraskeletal sources of local factors in the pathogenesis of osteoporosis, but this has not been adequately explored. Since the previous version of this chapter was written there have been exciting new findings on the mechanism of the interaction between cells of the osteoblastic and osteoclastic lineage in regulating bone resorption. As discussed elsewhere in this volume (Chapters 3, 12, and 13), osteoclasts and their precursors express receptor activator of NFB (RANK) for which the ligand, variously termed RANK ligand (RANKL), osteoclast differentiation factor (ODF), TRANCE, or OPGL, is expressed on stromal cells of the osteoblast lineage. This interaction is critical for the formation and activity of osteoclasts. It can be inhibited by a decoy receptor osteoprotegerin (OPG), which is produced by many cells. Both RANKL and OPG are regulated by both the systemic and local factors which influence bone resorption [32]. In reviewing the role of local factors, it is important to recognize that the data are derived largely from animal models and that there is little information on humans. Nevertheless, the fact that different portions of the skeleton are
CHAPTER 36 Local and Systemic Factors in the Pathogenesis of Osteoporosis
differentially affected in osteoporosis and that there are marked local changes in cancellous bone structure in osteoporosis supports a role for local factors. Further support comes from the clinical observation that a disease in which bone loss is clearly due to the secretion of local factors, multiple myeloma, can produce a rapidly progressive vertebral crush fracture syndrome which resembles severe primary osteoporosis. Multiple local factors are probably responsible for the intense osteoclastic activity produced by myeloma [33]. Other marrow disorders can also produce bone loss, presumably by local mechanisms.
21
In humans, increased IL-1 and IL-1ra production by circulating macrophages has been described in postmenopausal women, with inhibition by estrogen and greater persistence in osteoporotic patients, but not all studies confirm this observation [51,52]. Studies of cytokine production by marrow cultures have also produced variable results [53]. Direct measurement in bones from healthy and affected humans will be necessary before the roles of individual cytokines can be fully defined. However, the current availability of specific antagonists such as IL-1-receptor antagonist and TNF binding protein for clinical trials may also provide clues as to the role of these factors.
IV. CYTOKINES V. PROSTAGLANDINS The role of cytokines in bone resorption was first suggested by the finding that mononuclear leukocytes could produce an osteoclast activating factor [34]. The bone resorbing activity in the supernatants of mitogen or antigenactivated leukocytes ultimately turned out to be interleukin(IL)-1 [35,36]. However, other cytokines are also active. Tumor necrosis factor (TNF) is a potent bone resorber [37]. IL-6 and its soluble receptor may be cofactors for osteoclast generation [38]. IL-7 and IL-11 can increase bone resorption [39,40]. IL-4, IL-13, and interferon gamma can inhibit bone resorption in part by decreasing prostaglandin production [41,42]. However, IL-4 overexpression in mice produces osteoporosis associated with decreased osteoblast function [43]. Direct evidence for involvement of cytokines in the pathogenesis of osteoporosis is derived largely from studies in ovariectomized rodents. Estrogen can inhibit IL-6 production [44], and the increase in bone turnover after orchidectomy is inhibited by IL-6 antibodies [45]. Bone loss following ovariectomy in rats can be abrogated by administration of a combination of the IL-1 receptor antagonist and TNF, soluble binding protein [46]. Moreover, animals in whom the IL-1 activating receptor (IL-1-R1) has been knocked out do not show bone loss after ovariectomy [47]. Resorptive factors may be derived from the marrow. Bone marrow supernatant fractions from oophorectomized animals stimulate bone resorption and increase prostaglandin production in bone by a mechanism which can be blocked by IL-1ra as well as indomethacin [30,38]. However, it is not clear that IL-1 itself is the agonist, since the concentrations measured by immunoassay are not increased in the marrow supernatant fractions from ovariectomized animals compared to those from sham-operated controls. One possibility is that the effects of estrogen deprivation results in reciprocal increases in the activating receptor, IL-1-R1, and decreases in the decoy receptor IL-1-R2 [48]. Another possibility is that estrogen deficiency results in a decrease in OPG production since estradiol can decrease OPG levels in osteoblastic cell cultures [49].
Prostaglandins are potent, multifunctional regulators of bone metabolism, and their production by both hematopoietic cells and bone cells is abundant and highly regulated [54 – 56]. Most of the systemic hormones, cytokines, and growth factors which affect bone metabolism also affect prostaglandin production; however, the importance of this prostaglandin production in the response to hormones and local factors varies greatly. Agents that stimulate bone resorption also stimulate prostaglandin production, while inhibitors of bone resorption, such as glucocorticoids, IL-4, and interferon gamma, can inhibit prostaglandin production. In cell cultures which can produce osteoclasts in response to various stimuli, inhibition of prostaglandin synthesis decreases osteoclast formation regardless of the stimulator, although to a variable degree. The role of prostaglandin is probably dependent upon its synthesis by osteoblasts and this in turn appears to depend on the de novo synthesis of inducible prostaglandin GH synthase or cyclooxygenase-2 (COX-2) which has been shown to be highly regulated in osteoblastic cells [56]. Prostaglandins stimulate bone resorption in organ culture and in vivo [57,58]. This effect is probably mediated through cyclic AMP and prostaglandins of the E series are most potent. However, prostaglandins can also produce transient direct inhibition of the function of isolated osteoclasts [59]. In vivo, prostaglandins of the E series are potent stimulators of both endosteal and periosteal bone formation [60]. High concentrations of prostaglandins can inhibit collagen synthesis in vitro. This effect was greatest with prostaglandins of the F series and was mimicked by activators of protein kinase C [61]. Prostaglandin F2 may also stimulate bone resorption in part due to its ability to increase endogenous prostaglandin production [62]. This latter effect of prostaglandins, which we have termed “autoamplification’’ can occur with many prostanoids and may be important in maintaining and enhancing small signals such as fluid shear stress [63,64]. There is evidence that impact loading can stimulate prostaglandin production and that prostaglandin production
22
LAWRENCE G. RAISZ
is required for the increase in bone formation that occurs in response to mechanical forces [25]. As noted above this may be due to fluid shear stress on the syncytium of osteoblasts and osteocytes in bone [28]. Small strains can affect fluid flow in the canaliculi connecting osteocytes to each other and to surface osteoblasts and lining cells. The effects of fluid shear stress may be initiated by release of arachidonic acid and prostaglandin production by constitutive cyclooxygenase (COX-1) and sustained and amplified by induction of COX-2 [64]. The initial bone loss that occurs after immobilization may also be prostaglandin dependent, since it is inhibited by nonsteroidal anti-inflammatory drugs (NSAIDs) [65,66]. However, a sustained decrease in bone formation occurs after immobilization which is not reversed by indomethacin [67]. As noted above, marrow supernatant fractions from oophorectomized animals can stimulate bone resorption by a prostaglandin-dependent mechanism [48]. This effect is associated with a induction of COX-2. NSAIDs can also inhibit bone resorption in postmenopausal women [68]. Epidemiologic studies have demonstrated an increase in bone mineral density in some patients on nonselective NSAIDs [69], but these drugs are seldom taken continuously at high doses. The recent development of COX-2 selective inhibitors may make it possible to test the role of endogenous prostaglandins in human bone remodeling. In addition to prostaglandins produced by the cyclooxygenase pathway, metabolism of arachidonic acid by the lipoxygenase pathway may produce leukotrienes which can stimulate osteoclastic resorption [70]. Another potential regulator is nitric oxide (NO). NO has complex effects on bone resorption, although its most prominent effect is direct inhibition of osteoclastic activity [71].
VI. GROWTH FACTORS The families of insulin-like growth factors and of transforming growth factor (TGF-) and related bone morphogenetic proteins (BMPs) have been extensively studied for their role in bone growth [72 – 74] (see Chapter 14). It is possible that deficiency in the production of, or response to, these growth factors is important in the pathogenesis of decreased bone formation in osteoporosis. Studies of the insulin-like growth factor (IGF) family are complicated by the fact that the binding proteins (IGFBPs) are also regulated and have both inhibitory and stimulatory effects on IGF responses [75]. IGFBP-5 is present in bone matrix and is regulated by PTH and prostaglandins [76]. Both IGFBP5 and the inhibitory binding protein, IGFBP-4 may play a role in bone loss [77,78]. Transforming growth factor beta (TGF-) is a multifunctional regulator of bone, which can both stimulate and inhibit resorption, but is largely stimulatory for bone
formation [74]. The inhibition of osteoclast activity by estrogen has been attributed to an increase in TGF- which enhances osteoclast apoptosis [79]. Paradoxically overexpression of TGF- in osteoblasts results in osteoporosis-like phenotype in young mice [80], while diminished TGF activity has been associated with spontaneous bone loss in old male mice [81]. Bone morphogenetic proteins are likely to be involved in osteoporosis, but have not been studied extensively [73]. Among the heparin binding growth factors, basic fibroblast growth factor (FGF) has been shown to increase bone formation in vivo although it inhibits collagen synthesis in vitro [82,83]. FGF can increase bone resorption by both prostaglandin-dependent and independent mechanisms [84,85]. Platelet-derived growth factor is also a potent mitogen in bone and a stimulator of bone resorption [86,87].
VII. COLONY-STIMULATING FACTORS The critical role of macrophage colony-stimulating factor (M-CSF or CSF-1) in bone resorption was first suggested by the fact that mice deficient in this factor were osteopetrotic. Subsequent studies showed that M-CSF could restore osteoclastic activity in such animals [88]. These studies have been amplified by gene knock-out experiments in which animals lacking c-fos, a critical transcription factor for macrophages, also show osteopetrosis, which could be reversed by transplanting marrow cells from wild-type mice [89]. M-CSF is probably critical for the replication of osteoclast precursors and when added with RANKL can stimulate osteoclast production in hematopoietic cell cultures devoid of osteoblasts or their precursors [90]. Granulocyte/monocyte-CSF (GM-CSF) is also important in the regulation of bone resorption. Its affects appear to be inhibitory, probably due to a stimulation of cells along alternative pathways such as macrophage formation, which prevents them from differentiating into osteoclasts. Addition of GM-CSF to cell cultures can block osteoclast formation. IL-18 probably acts by increasing GM-CSF production in such cultures [91].
VIII. FURTHER CONSIDERATIONS OF INTERACTIONS OF SYSTEMIC HORMONES AND LOCAL MEDIATORS A number of interactions of systemic hormones, particularly sex hormones, with local factors have already been discussed. Our understanding in this area is still quite fragmentary, but it is clear that multiple systemic hormones have effects on multiple local factors. Thus, knock-out,
CHAPTER 36 Local and Systemic Factors in the Pathogenesis of Osteoporosis
overexpression, or inhibition experiments involving a single factor may not fully elucidate the role of that factor because other factors can substitute for its action. The major systemic mediators of bone resorption, PTH and 1,25-dihydroxyvitamin D stimulate prostaglandin production in bone and their ability to stimulate osteoclastogenesis in cell culture is partially prostaglandin dependent [92]. However, these hormones are not dependent on prostaglandins for their ability to stimulate bone resorption in organ culture. Indeed, the only evidence for a prostaglandindependent resorption pathway for a systemic hormone is that of thyroid hormone in mouse calvariae [93]. PTH and thyroid hormone, as well as IGF-I and prostaglandin E2, can increase interleukin-6 production [94,95]. Bone resorbing hormones, as well as local factors that stimulate bone resorption, can increase TGF- activity in organ cultures [96]. This is probably largely the result of activation and release from the matrix rather than new synthesis [97]. Activation of TGF- could play an important role in limiting resorption and initiating the coupled formation response. Glucocorticoids have complex interactions with local factors [15]. Glucocorticoids can inhibit production of interleukin-1, interleukin-6, prostaglandins, insulin-like growth factors, and IGF binding proteins from hematopoietic or bone cells. However, glucocorticoids can also increase the sensitivity to local factors, particularly prostaglandins and IGFs. The diurnal rhythm of glucocorticoid secretion is critical for normal skeletal metabolism. A decrease in glucocorticoid secretion in the afternoon and night may be permissive for the nocturnal increase in bone turnover. Small doses of glucocorticoids given in the evening can prevent the nocturnal rise in osteocalcin, and a block of the morning increase in glucocorticoid secretion can result in a sustained increase in this marker of osteoblastic activity [98,99]. One area of interaction between systemic and local factors which needs further exploration is the growth hormone/IGF-I system. Growth hormone can affect IGF production in skeletal tissue as well as in the liver. Since circulating IGF-I produces feedback inhibition of growth hormone secretion, alterations in hepatic production could have important indirect effects on skeletal production. Oral estrogen therapy produces a decrease in hepatic IGF-I production which could result in an increased growth hormone secretion which in turn would increase local IGF-I production [100]. In contrast, transdermal estrogen produces no change or an increase in circulating IGF-I and could have the opposite effect.
IX. CONCLUSIONS Even though the available data are limited, it seems likely that differences in the production of, or response to, local factors will be important in the pathogenesis of
23
osteoporosis. Since these local factors often act in concert and are often coordinately regulated, it seems likely that the relevant differences will involve multiple local factors in most patients. This hypothesis is particularly attractive because there is so little difference in the levels of systemic hormones in osteoporotic patients compared to that seen in age-matched controls. A plausible hypothesis would be that the production or activity of local factors changes to varying degrees with age and estrogen deficiency and that the individuals with severe osteoporosis are those who have (i) large increases in bone resorbing factors, (ii) loss of inhibitors of resorption, (iii) large increases in inhibitors of formation, or (iv) loss of stimulators of bone formation [101,102]. Another critical factor in the pathogenesis of osteoporosis is peak bone mass. It is quite possible that the genetic determinants of peak bone mass involve effects on local factors, particularly growth factors. Moreover, the effect of weight-bearing activity on peak bone mass is likely to be mediated by local factors. Regulation involves contributions from marrow cells of both the hematopoietic and the mesenchymal lineages as well as bone cells themselves. There is some evidence for a decrease in the number of osteogenic stem cells in the marrow with age, and this could account in part for age-related decreases in bone formation [103 – 105]. On the other hand, the differentiation of cells of the osteoclastic lineage may be maintained with age, thus maintaining high rates of bone resorption (J. A. Lorenzo, unpublished observations). An interaction between marrow and bone could explain the fact that bone loss in osteoporosis largely involves cancellous bone and endosteal surfaces of cortical bone [48,106]. Finally it is likely that the pathogenesis of severe osteoporosis will be heterogeneous; that is, there will be patients with different patterns of abnormalities involving different systemic hormones and local factors. The division into Type I, postmenopausal osteoporosis, and type II, senile osteoporosis, represent a simplified version of this concept. The broad range of values for bone resorption and formation in bone biopsies in osteoporosis supports the concept of heterogeneity [107]. Thus, rather than a division into high and low turnover forms, it is more likely that there will be a spectrum of relative contributions of abnormalities of resorption and formation in different patients. Moreover, these changes may vary over the course of the disease. Confirmation of the hypothesis that local factors play a role in osteoporosis and identification of specific factors should now be possible through the application of newer methods of molecular and cellular biology as well as new pharmacologic approaches. A combination of the use of microarray technology and quantitative reverse-transcriptasepolymerase chain reaction methodology should allow us to identify and quantify specific local factors and their receptors. It may be more difficult to prove that they play a role in the pathogenesis. Selective inhibitors of cytokines,
24
LAWRENCE G. RAISZ
prostaglandins, and other local factors are being developed and could be tested in appropriately selected patients. Such an approach will not only improve our understanding of the pathogenetic mechanisms in osteoporosis, but might also lead to a more specific preventive and therapeutic measures in this disorder.
References 1. B. L. Riggs and L. J. Melton III, Involutional osteoporosis. N. Engl. J. Med. 314, 1676–1686 (1986). 2. L. G. Raisz, Local and systemic factors in the pathogenesis of osteoporosis. N. Engl. J. Med. 318, 818 – 828 (1988). 3. L. G.Raisz, The osteoporosis revolution. Ann. Intern. Med. 126, 458 – 462 (1997). 4. J. Center and J. Eisman, The epidemiology and pathogenesis of osteoporosis. Baillieres Clin. Endocrinol. Metab. 11, 23 – 62 (1997). 5. T. Steiniche, P. Christiansen, A. Vesterby, C. Hasling, R. Ullerup, L. Mosekilde, and F. Melsen, Marked changes in iliac crest bone structure in postmenopausal osteoporotic patients without any signs of disturbed bone remodeling or balance. Bone 15, 73 – 80 (1994). 6. J. A. P. Jayasinghe, S. J. Jones, and A. Boyde, Three dimensional photographic study of cancellous bone in human fourth lumbar vertebral bodies. Anat. Embryol. 189, 259 – 274 (1994). 7. P. Garnero, P. Dargent-Molina, D. Hans, A. M. Schott, G. Breart, P. J. Meunier, and P. D. Delmas, Do markers of bone resorption add to bone mineral density and ultrasonographic heel measurement for the prediction of hip fracture in elderly women? The EPIDOS prospective study. Osteoporosis Int. 8, 563 – 569 (1998). 8. P. Garnero, E. Sornay-Rendu, F. Duboeuf, and P. D. Delmas, Markers of bone turnover predict postmenopausal forearm bone loss over 4 years: The OFELY study. J. Bone Miner. Res. 14, 1614 – 1621 (1999). 9. M. A. Hansen, K.Overgaard, B. J. Riis, and C. Christiansen, Role of peak bone mass and bone loss in postmenopausal osteoporosis: 12 Year study. Br. Med. J. 303, 961 – 964 (1991). 10. S. Boonen, P. Broos, G. Verbeke, J. Aerssens, E. Van Herck, I. Jans, J. Dequeker, and R. Buillon, Calciotropic hormones and markers of bone remodeling in age-related (type II) femoral neck osteoporosis: Alterations consistent with secondary hyperparathyroidism-induced bone resorption. J. Gerontol. A. Biol. Sci. Med. Sci. 52, M286 – M293 (1997). 11. S. J. Silverberg, E. Shane, L. de la Cruz, et al., Abnormalities in parathyroid hormone secretion and 1,25-dihydroxyvitamin D3 formation in women with osteoporosis. N. Engl. J. Med. 320, 277– 281 (1989). 12. R. D. Tiegs, J. J. Body, H. W. Wahner, J. Barta, B. L. Riggs, and H. Heath III, Calcitonin secretion in postmenopausal osteoporosis. N. Engl. J. Med. 312, 1097 – 1100 (1985). 13. C. A. Benbassat, K. C. Maki, and T.G. Unterman, Circulating levels of insulin-like growth factor (IGF) binding protein-1 and -3 in aging men: Relationships to insulin, glucose, IGF, and dehydroepiandrosterone sulfate levels and anthropometric measures. J. Clin. Endocrinol. Metab. 82, 1484 – 1491 (1997). 14. E. S. Kurland, F. K. Chan, C. J. Rosen, and J. P. Bilezikian, Normal growth hormone secretory reserve in men with idiopathic osteoporosis and reduced circulating levels of insulin-like growth factor-I. J. Clin. Endocrinol. Metab. 83, 2576 – 2579 (1998). 15. B. P. Lukert, and L. G. Raisz, Glucocorticoid-induced osteoporosis. J. Clin. Endocrinol. Metab. 84, 3058 – 3063 (1994).
16. E. Dennison, P. Hindmarsh, C. Fall, S. Kellingray, D. Barker, D. Phillips, and C. Cooper, Profiles of endogenous circulating cortisol and bone mineral density in healthy elderly men. J. Clin. Endocrinol. Metab. 33, 3058 –3063 (1999). 17. S. Khosla and L. J. Melton III, E. J. Atkinson, W. M. O’Fallon, G. G. Klee, and B. L. Riggs, Relationship of serum sex steroid levels and bone turnover markers with bone mineral density in men and women: A key role for bioavailable estrogen. J. Clin. Endocrinol. Metab. 83, 2266 – 2274 (1998). 18. E. P. Smith, J. Boyd, G. R. Frank, H. Takahashi, R. M. Cohen, B. Specker, T. C. Williams, D. B. Lubahn, and Korach, Estrogen resistance caused by mutation in the estrogen receptor gene in a man. N. Engl. J. Med. 331, 1056 – 1061 (1994). 19. J. P. Bilezikian, A. Morishima, J. Bell, and M. M. Grumbach, Increased bone mass as a result of estrogen therapy in a man with aromatase deficiency. N. Engl. J. Med. 339, 599 – 603 (1998). 20. D. Vanderschueren, E. Vanherck, P. Geusens, A. Suiker, W. Visser, K. Chung, and R. Bouillon, Androgen resistance and deficiency have different effects on the growing skeleton of the rat. Calcif. Tissue Int. 55, 198 – 203 (1994). 21. L. C. Hofbauer and S. Khosla, Androgen effects on bone metabolism: Recent progress and controversies. Eur. J. Endocrinol. 140, 271 – 286 (1999). 22. F. A. Tremollieres, D. D. Strong, D. J. Baylink, and S. Mohan, Progesterone and promegestone stimulate human bone cell proliferation and insulin-like growth factor – 2 production. Acta Endocrinol. 126, 329 – 337 (1992). 23. D. L. Manzi, C. C. Pilbeam, and L. G. Raisz, The anabolic effects of progesterone on fetal rat calvaria in tissue culture. J. Soc. Gynecol. Invest. 1, 302 – 309 (1994). 24. Z. Ikram, L. Dulipsingh, and K. M. Prestwood, Lack of effect of short-term micronized progesterone on bone turnover in postmenopausal women. J. Women’s Health & Gender-Based Med. 8, 973–978 (1999). 25. J. W. Chow, S. W. Fox, M. J. Lean, and T. J. Chambers, Role of nitric oxide and prostaglandins in mechanically induced bone formation. J. Bone Miner. Res. 13, 1039 – 1044 (1998). 26. J. M. Lean, C. J. Jagger, T. J. Chambers, and J. W. Chow, Increased insulin-like growth factor I mRNA expression in rat osteocytes in resonse to mechanical stimulation. Am. J. Physiol. 268, E318 – E327 (1995). 27. T. M. Skerry, Identification of novel signaling pathways during functional daptation of the skeleton to mechanical loading: The role of glutamate as a paracrine signaling agent in the skeleton. J. Bone Miner. Res. 17, 66 – 70 (1995). 28. E. J. Burger and J. Klein-Nulend, Mechanotransduction in bone–– Role of the lacuno-canalicular network. FASEB J. 13, S101 – 112 (1999). 29. S. V. Reddy and G. D. Roodman, Control of osteoclast differentiation. Crit. Rev. Eukaryotic Gene Expression. 8, 1 – 17 (1998). 30. H. Kawaguchi, C. C. Pilbeam, S. J. Vargas, E. E. Morse, J. A. Lorenzo, and L. G. Raisz, Ovariectomy enhances and estrogen replacement inhibits the activity of bone marrow factors that stimulate prostaglandin production in cultured mouse calvariae. J. Clin. Invest. 96, 539 – 548 (1995). 31. P. Collin-Osdoby, Role of vascular endothelial cells in bone biology. J. Cell. Biochem. 55, 304 – 309 (1994). 32. T. Suda, N. Takahashi, N. Udagawa, E. Jimmi, M. T. Gillespie, and T. J. Martin, Modulation of osteoclast differentiation and function by the new members of the tumor necrosis factor receptor and ligand families. Endocrinol. Rev. 20, 345 – 357 (1999). 33. G. D. Roodman, Mechanisms of bone lesions in multiple myeloma and lymphoma. Cancer 80, 1557 – 1563 (1997). 34. L. G. Raisz, R. A. Luben, G. R. Mundy, J. W. Dietrich, J. E. Horton, and C. L. Trummel, Effect of osteoclast activating factor from
CHAPTER 36 Local and Systemic Factors in the Pathogenesis of Osteoporosis
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
human leukocytes on bone metabolism. J. Clin. Invest. 56, 408 – 413 (1975). F. E. Dewhirst, P. P. Stashenko, J. E. Mole, and T. Tsurumachi, Purification and partial sequence of human osteoclast-activating factor: Identity with interleukin 1b. J. Immunol. 135, 2562 – 2568 (1985). J. A. Lorenzo, S. L. Sousa, C. Alander, L. G. Raisz, and C. A. Dinarello, Comparison of the bone-resorbing activity in the supernatants from phytohemagglutinin-stimulated human peripheral blood mononuclear cells with that of cytokines through the use of an antiserum to interleukin 1. Endocrinology 121, 1164 – 1170 (1987). D. R. Bertolini, G. E. Nedwin, T. S. Bringman, D. D. Smith, and G. R. Mundy, Stimulation of bone resorption and inhibition of bone formation in vitro by human tumour necrosis factors. Nature 319, 516 – 518 (1986). C. Miyaura, K. Kusano, T. Masuzawa, O. Chaki, Y Onoe, M. Aoyagi, T. Sasaki, T. Tamura, Y. Koishihara, Y. Ohsugi, et al., Endogenous bone-resorbing factors in estrogen deficiency: Cooperative effects of IL–1 AND IL–6. J. Bone Miner. Res. 10, 1365 – 1373 (1995). C. Miyaura, Y. Onoe, M. Inada, K. Maki, K. Ikuta, M. Ito, and T. Suda, Increased B-lymphopoiesis by interleukin 7 induces bone loss in mice with intact ovarian function: Similarity to estrogen deficiency. Proc. Natl. Sci. USA 94, 2581 – 1591 (1996). E. Romas, N. Udagawa, H. Zhou, T. Tamura, M Saito, T. Taga, D. J. Hilton, T. Suda, K. W. Ng, and T. J. Martin, The role of gp 130-mediated signals in osteoclast development: Regulation of interleukin 11 production by osteoblasts and its distribution of its receptor in bone marrow cultures. J. Exp. Med. 183, 2581 – 2591 (1996). Y. Onoe, C. Miyaura, T. Kaminakayashiki, Y. Nagai, K. Noguchi, Q. R. Chen, H. Seo, H. Ohta, S. Nozawa, I. Kudo, and T. Suda, IL-4 and IL-13 inhibit bone resorption by suppressing cyclooxygenase – 2-dependent prostaglandin synthesis in osteoblasts. J. Immunol. 156, 758 – 764 (1996). O. Hoffman, K. Klaushofer, H. Gleispach, H. J. Leis, T. Luger, K. Koller, and M. Peterlik, ?Interferon inhibits basal and interleukin 1induced prostaglandin production and bone resorption in neonatal mouse calvaria. Biochem. Biophys. Res. Commun. 143, 38 – 43 (1987). D. B. Lewis, H. D. Liggitt, E. L. Effmann, S. T. Motley, S. L. Teitelbaum, K. J. Jepsen, S. A. Goldstein, J. Bonadio, J. Carpenter, and R. M. Perlmutter, Osteoporosis induced in mice by overproduction of interleukin-4. Proc. Natl. Acad. Sci. USA 90, 11618 – 11622 (1993). S. T. Pottratz, T. Bellido, H. Mocharlia, D. Crabb, and S. C. Manolagas, 17b-Estradiol inhibits expression of human interleukin – 6 promoterreporter constructs by a receptor dependent mechanism. J. Clin. Invest. 93, 944 – 950 (1994). T. Bellido, R. L. Jilka, B. F. Boyce, G. Girasole, H. Broxmeyer, S. A. Dalrymple, R. Murray, and S. C. Manolagas, Regulation of interleukin – 6, osteoclastogenesis, and bone mass by androgens. The role of the androgen receptor. J. Clin. Invest. 95, 2886 – 2895 (1995). R. B. Kimble, A. B. Matayoshi, J. L. Vannice, V. T. Kung, C. Williams, and R. Pacifici, Simultaneous block of interleukin – 1 and tumor necrosis factor is required to completely prevent bone loss in the early postovariectomy period. Endocrinology 136, 3054 – 3061 (1995). J. A. Lorenzo, A. Naprta, Y. Rao, C. Alander, M. Glaccum, M. Widmer, et al., Mice lacking the type I interleukin – 1 receptor do not lose bone mass after ovariectomy. Endocrinology 139, 3022 – 3025 (1998). T. Sunyer, J. Lewis, and P. Collin-Osdoby, Estrogen’s bone-protective effects may involve differential IL-1 receptor regulation in human osteoclast-like cells. J. Clin. Invest. 103, 1409 – 1418 (1999). L. C. Hofbauer, S. Khosla, C. R. Dunstan, D. L. Lacey, T. C. Spelsberg, and B. L. Riggs, Estrogen stimulates gene expression and
50.
51.
52.
53.
54.
55. 56.
57. 58.
59.
60. 61.
62.
63.
64.
65.
66.
67.
68.
25
protein production of osteoprotegerin in human osteoblastic cells. Endocrinology 140, 4367 – 4370. (1999). R. R. Pacifici, L. Rifas, R. McCracken, I. Vered, C. McMurtry, L. V. Avioli, and W. A. Peck, Ovarian steroid treatment blocks a postmenopausal increase in blood monocyte interleukin 1 release. Proc. Natl. Acad. Sci. USA 86, 2398 – 2402 (1989). R. Pacifici, J. L. Vannice, L. Rifas, and R. B. Kimble, Monocytic secretion of interleukin – 1 receptor antagonist in normal and osteoporotic women: Effects of menopause and estrogen/progesterone therapy. J. Clin. Endocrinol. Metab. 77, 1135 – 1141 (1993). W. McKane, S. Khosla, J. M. Peterson, K. Egan, and B. L. Riggs, Circulating levels of cytokines that modulate bone resorption: Effects of age and menopause in women. J. Bone Miner. Res. 9(8), 1313 – 1318 (1994). M. Kassem, S. Khosla, T. C. Spelsberg, B. L. Riggs, Cytokine production in the bone marrow microenvironment: Failure to demonstrate estrogen regulation in early postmenpausal women. J. Clin. Endocrinol. Metal. 81, 513 – 518 (1996). H. Kawaguchi, C. C. Pilbeam, J. R. Harrison, and L. G. Raisz, The role of prostaglandins in the regulation of bone metabolism. Clin. Orthop. Rel. Res. 313, 36 – 46 (1995). L. G. Raisz, Prostaglandins and bone: Physiology and pathophysiology. Osteoarthritis Cartilage 7, 419 – 421 (1999). C. C. Pilbeam, H. Kawaguchi, Y. Hakeda, O. Voznesensky, C. B. Alander, and L. G. Raisz, Differential regulation of inducible and constitutive prostaglandin endoperoxide synthase in osteoblastic MC3T3E1 cells. J. Biol. Chem. 268, 25643 – 25649 (1993). D. C. Klein and L. G. Raisz, Prostaglandins: Stimulation of bone resorption in tissue culture. Endocrinology 86, 1436 – 1440 (1970). B. F. Boyce, T. B. Aufdemorte, I. R. Garrett, A. J. P. Yates, and G. R. Mundy, Effects of interleukin – 1 on bone turnover in normal mice. Endocrinology 125, 1142 – 1150 (1989). K. Fuller and T. J. Chambers, Effects of arachidonic acid metabolites on bone resorption by isolated rat osteoclasts. J. Bone Miner. Res. 4, 209 – 215 (1989). W. J. Jee and Y. F. Ma, The in vivo anabolic actions of prostaglandins in bone. Bone 21, 297 – 304 (1997. P. M. Fall, D. T. Breault, and L. G. Raisz, Inhibition of collagen synthesis by prostaglandins in the immortalized rat osteoblastic clonal cell line, Py1a: Structure activity relations and signal transduction mechanisms. J. Bone Miner. Res. 9, 1935 – 1943 (1994). L. G. Raisz, C. B. Alander, P. M. Fall, and H. A. Simmons, Effects of prostaglandin F2 on bone formation and resorption in cultured neonatal mouse calvariae: Role of prostaglandin E2 production. Endocrinology 126, 1076 – 1079 (1990). K. M. Reich and J. A. Frangos, Protein kinase-C mediates flowinduced prostaglandin E2 production in osteoblasts. Calcif. Tissue Int. 52, 62 – 66 (1993). J. Klein-Nulend, E. H. Burger, C. M. Semeins, L. G. Raisz, and C. C. Pilbeam, Pulsating fluid flow stimulates prostaglandin release and inducible prostaglandin G/H synthase mRNA expression in primary mouse bone cells. J. Bone Miner Res. 12, 45 – 51 (1997). D. D. Thompson and G. A. Rodan, Indomethacin inhibition of tenotomy-induced bone resorption in rats. J. Bone Miner. Res. 3, 409 – 414 (1988). D. J. Waters, D. D. Caywood, G. J. Trachte, R. T. Turner, and S. F. Hodgson, Immobilization increases bone prostaglandin E. Effect of acetylsalicylic acid on disuse osteoporosis studied in dogs. Acta Orthop. Scand. 62, 238 – 243 (1991). M. Weinreb, G. A. Rodan, and D. D. Thompson, Osteopenia in the immobilized rat hind limb is associated with increased bone resorption and decreased bone formation. Bone 10, 187 – 194 (1989). N. H. Bell, B. W. Hollis, J. R. Shary, D. R. Eyre, R. Eastell, A. Colwell, and R. G. G. Russell, Diclofenac sodium inhibits bone
26
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
79.
80.
81.
82.
83.
84.
85.
LAWRENCE G. RAISZ resorption in postmenopausal women. Am. J. Med. 96, 349 – 353 (1994). D. J. Morton, E. L. Barrett-Connor, D. L. Schneider, Nonsteroidal anti-inflammatory drugs and bone mineral density in older women: The Rancho Bernardo study. J. Bone Miner. Res. 13, 1924 – 1931 (1998). M. A. Flynn, M. Qiao, C. Garcia, M. Dallas, and L. F. Bonewald, Avian osteoclast cells are stimulated to resorb calcified matrices by and possess receptors for leukotriene B–4. Calcif. Tissue Intl. 64, 154 – 159 (1999). R. J. van’t Hof, S. H. Ralston, Cytokine-induced nitric osxide inhibits bone resorption by inducing apoptosis of osteoclast progenitors and suppressing osteoclast activity. J. Bone Miner. Res. 12, 1797 – 1804 (1997). G. Crawford-Sharpe and C. J. Rosen, Insulin-like growth factor-I and the skeleton: New perspectives. Endocrinologist 9, 81 – 86 (1999). J. M. Wozney and V. Rosen, Bone morphogenetic protein and bone morphogenetic protein gene family in bone formation and repair. Clin. Orthop. 346, 26 – 37 (1998). L. F. Bonewald and S. L. Dallas, Role of active and latent transforming growth factor b in bone formation. J. Cell. Biochem. 55, 350 – 357 (1994). C. A. Conover Regulation and physiological role of insulin-like growth factor binding proteins. Endocrinol J. 43, S43 – S48 (1996). Y. Hakeda, H. Kawaguchi, M. Hurley, C. C. Pilbeam, C. Abreu, T. A. Linkhart, S. Mohan, M. Kumegawa, and L. G. Raisz, Intact insulin-like growth factor binding protein-5 (IGFBP-5) associates with bone matrix and soluble fragments of IGFBP-5 accumulate in culture medium of neonatal mouse calvariae treated with parathyroid hormone or prostaglandin E2. J. Cell. Physiol. 166, 370 – 379 (1996). C. Rosen, L. R. Donahue, S. Hunter, M. Holick, H. Kavookjian, A. Kirschenbaum, S. Mohan, and D. J. Baylink, The 24/25-kDa serum insulin-like growth factor-binding protein is increased in elderly women with hip and spine fractures. J. Clin. Endocrinol. Metab. 74, 24 – 27 (1992). S. Mohan and D. J. Baylink, Serum insulin-like growth factor binding protein (IGFBP)-4 and IGFBP-5 levels in aging and age-associated diseases, Endocrinology 7, 87 – 91 (1997). D. E. Hughes, A. Dai, J. C. Tiffee, H. H. Li, G. R. Mundy, and B. F. Boyce, Estrogen promotes apoptosis of murine osteoclasts mediated by TGF-beta. Nat. Med. 2, 1132 – 1136 (1996). A. Erlebacher and R. Derynck, Increased expression of TGF-beta 2 in osteoblasts results in an osteoporosis-like phenotype. J. Cell. Biol. 132, 195 – 210 (1996). D. Gazit, Y. Zilberman, R. Ebner, and A. Kahn, Bone loss (osteopenia) in old male mice results from diminished activity and availability of TGF-beta. J. Cell Biochem. 70, 478 – 488 (1998). C. R. Dunstan, R Boyce, I. R. Garrett, E. Izbicka, W. H. Burgess, and G. R. Mundy, Systemic administration of acidic fibroblast growth factor (FGF-1) prevents bone loss and increases new bone formation in ovariectomized rats, J. Bone Miner. Res. 14, 953 – 959 (1999). M. M. Hurley, C. Abreu, J. R. Harrison, A. C. Lichtler, L. G. Raisz, and B. E. Kream, Basic fibroblast growth factor inhibits type-I collagen gene expression in osteoblastic MC3T3-E1 cells. J. Biol. Chem. 268, 5588 – 5593 (1993). H. A. Simmons and L. G. Raisz, Effects of acid and basic fibroblast growth factor and heparin on resorption of cultured fetal rat long bones. J. Bone Miner. Res. 6, 1301 – 1305 (1991). M. M. Hurley, S. K. Lee, L. G. Raisz, P. Bernecker, and J. Lorenzo, Basic fibroblast growth factor induces osteoclast formation in murine bone marrow cultures, Bone 22, 309 – 316 (1998).
86. J. M. Hock and E. Canalis, Platelet-derived growth factor enhances bone cell replication, but not differentiated function of osteoblasts, Endocrinology 134, 1423 – 1428 (1994). 87. Z. Zhang, J. Chen, and D. Jin, Platlet-derived growth factor (PDGF)BB stimulates osteoclastic bone resorption directly: The role of rector beta, Biochem. Biophys. Res. Commun. 251, 190 – 194 (1998). 88. R. Felix, M. G. Cecchini, and H. Fleisch, Macrophage colony stimulating factor restores in vivo bone resorption in the OP/OP osteopetrotic mouse. Endocrinology 127, 2592 – 2594 (1990). 89. A. E. Grigoriadis, Z.-Q. Wang, M. G. Cecchini, W. Hofstetter, R. Felix, H. A. Fleisch, and E. F. Wagner, c-Fos: A key regulator of osteoclast – macrophage lineage determination and bone remodeling. Science 266, 443 – 448 (1994). 90. J. M. Quinn, J. Elliott, M. T. Gillespie, and T. J. Martin, A combination of osteoclast differentiation factor and macrophage-colony stimulating factor is sufficient for both human and mouse osteoclast formation in vitro. Endocrinology 139, 4424 – 4427 (1998). 91. N. J. Horwood, N. Udagawa, J. Elliott, D. Grail, H. Okamura, and M. Kurimoto, Interleukin 18 inhibits osteoclast formation via T cell production of pranulocyte macrophage colony-stimulating factor J. Clin Invest. 101, 595 – 603 (1998). 92. H. Inoue, T. Tsujisawa, T. Fukuizumi, S. Kawagishi, and C. Uchiyama, SC-19220, a prostaglandin E2 antagonist, inhibits osteoclast formation by 1,25-dihydroxyvitamin D3 in cell cultures. J. Endocrinol. 161, 231 – 236 (1999). 93. K. Klaushofer, O. Hoffman, H. Gleispach, et al., Bone-resorbing activity of thyroid hormones is regulated to prostaglandin production in cultured neonatal mouse calvaria. J. Bone Miner. Res. 4, 305 – 312 (1989). 94. I. Holt, M. W. J. Davie, I. P. Braidman, and M. J. Marshall, Prostaglandin E (2) stimulates the production of interleukin – 6 by neonatal mouse parietal bones. Bone Miner. 25, 47 – 58 (1994). 95. M. C. Slootweg, W. W. Most, E. Vanbeek, L. P. C. Schot, S. E. Papapoulos, and C. W. G. M. Lowik, Osteoclast formation together with interleukin-6 production in mouse long bones is increased by insulinlike growth factor-I. J. Endocrinol. 132, 433 – 438 (1992). 96. J. Pfeilschifter and G. R. Mundy, Modulation of type b transforming growth factor activity in bone cultures by osteotropic hormones. Proc. Natl. Acad. Sci. USA 84, 2024 – 2028 (1987). 97. S. L. Dallas, S. Park-Snyder, K. Miyazono, D. Twardzik, G. R. Mundy, and L. F. Bonewald, Characterization and autoregulation of latent transforming growth factor b (TGFb) complexes in osteoblastlike cell lines — Production of a latent complex lacking the latent TGFb-binding protein. J. Biol. Chem. 269, 6815 – 6822 (1994). 98. H. K. Nielsen, P. Charles, and L. Mosekilde, The effect of single oral doses of prednisone on the circadian rhythm of serum osteocalcin in normal subjects. J. Clin. Endocrinol. Metab. 67, 1025 – 1030 (1988). 99. H. K. Nielsen, K. Brixen, M. Kassem, P. Charles, and L. Mosekilde, Inhibition of the morning cortisol peak abolishes the expected morning decrease in serum osteocalcin in normal males — Evidence of a controlling effect of serum cortisol on the circadian rhythm in serum osteocalcin. J. Clin. Endocrinol. Metab. 74, 1410 – 1414 (1992). 100. K. K. Y. Ho and A. J. Weissberger, Impact of short-term estrogen administration on growth hormone secretion and action: Distinct routedependent effects on connective and bone tissue metabolism. J. Bone Miner. Res. 7, 821 – 827 (1992). 101. M. E. Cohensolal, A. M. Graulet, M. A. Denne, J. Gueris, D. Baylink, and M. C. Devernejoul, Peripheral monocyte culture supernatants of menopausal women can induce bone resorption: Involvement of cytokines. J. Clin. Endocrinol. 77(6), 1648–1653 (1993). 103. J. Pfeilschifter, I. Diel, U. Pilz, K. Brunotte, A. Naumann, and R. Ziegler, Mitogenic responsiveness of human bone cells in vitro to hormones and growth factors decreases with age. J. Bone Miner. Res. 8, 707 – 718 (1993).
CHAPTER 36 Local and Systemic Factors in the Pathogenesis of Osteoporosis 104. R. L. Jilka, R. S. Weinstein, K, Takahashi, A. M. Parfitt, and S. C. Manolagas, Linkage of decreased bone mass with impaired osteoclastogenesis in a murine model of accelerated senescence. J. Clin Invest. 97, 1732 – 1740 (1996). 105. G. D’Ippolito, P. C. Schiller, C. Ricordi, B. A. Roos, and G. A. Howard, Age-related osteogenic potential of mesenchymal stromal stem cells from human vertebral bone marrow. J Bone Miner. Res. 14, 1115 – 1122 (1999).
27
106. H. Bismar, I. Diel, R. Ziegler, and J. Pfeilschifter, Increased cytokine secretion by human bone marrow cells after menopause or discontinuation of estrogen replacement. J. Bone Miner. Res. 9, S158 (1994). 107. E. F. Eriksen, S. F. Hodgson, R. Eastell, S. L. Cedel, W. M. O’Fallon, and B. L. Riggs, Cancellous bone remodelling in type I (postmenopausal) osteoporosis: Quantitative assessment of rates of formation, resorption and bone loss at tissue and cellular levels. J. Bone Miner. Res. 5, 311 – 319 (1990).
CHAPTER 37
Animal Models for in Vivo Experimentation in Osteoporosis Research DONALD B. KIMMEL
Osteoporosis Research Center, Creighton University, Omaha, Nebraska 68131
I. Introduction II. A Perspective on in Vivo Animal Experimentation III. Criteria for Animal Models in Osteoporosis: A Time for Compromise
IV. The Criteria V. Animal Models for Human Osteoporosis References
of animals that are either losing bone or that have become osteopenic following ovariectomy because of the relationship of human osteoporosis to estrogen depletion. The FDA requires data from both the rat and a larger species. Studies of all rats must be of 12 months duration, but those in other species must be of 16 months duration, ostensibly equivalent to 4 years of human treatment, and include histologic evaluation. The guidelines also suggest measuring not only bone density and biochemical markers of bone turnover, but also bone strength by biomechanical testing, as surrogates for the propensity to develop fragility fractures. These guidelines are based principally on past experiments with current bone-active agents. Since these agents mainly slow bone turnover (e.g., estrogens, calcitonin, and bisphosphonates), the requirements are structured to find relatively modest changes over sustained periods. When published experience with agents that stimulate bone formation accumulates, it is likely that the FDA guidelines for testing such agents will evolve appropriately.
I. INTRODUCTION This chapter presents criteria for evaluating animal models of osteoporosis and then applies them using today’s facts. The criteria are based on: (i) knowledge about human osteoporosis, (ii) fundamental understanding of human and animal skeletons, (iii) experiments that use the same agent in different species, and (iv) recognition of unmet needs in osteoporosis research. Both the criteria and their evaluation should evolve with time as data about osteoporosis, each animal model, and new evaluation methods accumulate.
A. FDA Recommendations for Animal Models of Osteoporosis The Food and Drug Administration (FDA) has established guidelines for using animals in preclinical testing of agents intended to treat osteoporosis [1]. It recommends use
OSTEOPOROSIS, SECOND EDITION VOLUME 2
29
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
30
II. A PERSPECTIVE ON IN VIVO ANIMAL EXPERIMENTATION In vivo animal experimentation is overused in osteoporosis research. Though this may signal nothing more than an imbalance between the numbers of bench and clinical investigators, it may also mark the existence of something more alarming, misconceptions about efficient experimentation. In vivo adult human experimentation is desirable because its results require minimal extrapolation to predict the outcome in afflicted adult humans. Epidemiologic data bases that contain information about human osteoporosis should be mined exhaustively [2 – 6]. Convenient opportunities to add critical information to those data bases should be leveraged. When natural products or agents already approved for human use have a sound basis for efficacy in new clinical situations, a thorough examination of existing human data and a Phase I human trial, not an animal experiment, are likely to be the next logical steps. When an understanding of the inheritance of bone characteristics in humans is desired, studies of human pedigrees or affected human sib-pairs are the most efficient way of establishing the linkage of specific skeletal phenotypes to genetic loci [7 – 9]. In summary, despite the apparent emphasis of this chapter, skeletal investigators should heed the adage, “When possible, do it in humans,” more often. However, when experimental designs or new therapeutic agents that pose overt short- or long-term risks are to be evaluated, in vivo animal experimentation is the main option. For instance, when critical questions about the effects of agents on turnover, structure, and cellular activity in the sites in humans most prone to osteoporotic fracture (e.g., vertebrae, hip, and wrist) arise, it is apparent that human experimentation will not generate suitable samples. An appropriately designed animal experiment is the only approach. When animal experimentation is indicated and an in vivo animal model that accurately reproduces human data is available, an animal experiment is the best solution. The lack of such a strategy, including the availability and proper use of animal models in the late 1970s, might have played a role in the unfavorable impression left by an osteoporosis treatment trial involving sodium fluoride [10 – 12]. That outcome led to the FDA’s cautious and somewhat onerous requirement that investigators of future osteoporosis treatments demonstrate both bone mass increases and eventual anti-fracture efficacy in clinical trials to achieve full approval.
III. CRITERIA FOR ANIMAL MODELS IN OSTEOPOROSIS: A TIME FOR COMPROMISE Expecting full parallelism of human symptoms with in vivo animal models is unrealistic. These criteria are flexible, occasionally creatively applied, and clearly able to
DONALD B. KIMMEL
evolve as new evidence or needs appear. They place the highest value on animal models that match the clinically apparent behavior of osteoporosis, treating the detailed match of mechanisms as a fine point. All bone researchers can and should participate in the refinement of these criteria. Clinical investigators can facilitate the development of in vivo animal models by both providing details of the behavior of the human osteoporotic skeleton when new data become available and ranking the importance of various clinical characteristics of osteoporosis. In vivo animal investigators can help by doing experiments that show how their models fit important human symptoms. The recent history of the ovariectomized rat model is a textbook example of bench scientists doing experiments to validate a highly relevant preclinical in vivo animal model [13 – 16]. That effort, largely completed in the 1980s, was made possible by the bone mass and metabolic data collected about human osteoporosis during the 1970s. Recent efforts to develop standardized tests of bone fragility in small and large animals as surrogates for osteoporotic fracture are another example [17 – 20]. In vivo animal scientists can also help by developing more animal models that employ relevant physiologic conditions that can be readily applied by capable investigators. For example, when the goal is to characterize better an adult disease process, choosing a growing animal, simply to see quicker or more marked changes, makes little sense. Doing nerve resection to study disuse [21,22] may really model only the motionless, denervated limb, not the motionless limb. Combining extreme calcium deprivation with estrogen depletion to accelerate (or even just create) bone loss, may produce a confused situation that bears little resemblance to the clinical picture [23]. For osteoporosis research, using relevant methods in an adult animal to produce a consistent, albeit incomplete set of symptoms, is more acceptable than using nonphysiologic circumstances to develop a full set of symptoms. A model that gives sporadic results in the hands of numerous investigators or requires convoluted manipulations is not likely to become widely accepted.
A. Summary Today’s knowledge of animal models and osteoporosis means that using an irrelevant animal model in osteoporosis research is inexcusable and wasteful. Imposing nonphysiologic conditions to create desired symptoms is undesirable and always decreases a model’s relevance. Using a relevant, though costly model is occasionally necessary when unique information is needed. The continuous evaluation of new animal models for relevance and cost-effectiveness benefits all.
31
CHAPTER 37 Animal Models for in Vivo Experimentation
IV. THE CRITERIA Application of in vivo animal models for the skeleton has been reviewed elsewhere [24 – 27]. This chapter not only emphasizes many of their important points, but also widens the perspective by encompassing additional animal models. Its fundamental approach is to find animal models that match tissue behaviors that can be measured in vivo in osteoporotic humans, rather than to assure identical cellular mechanisms.
A. Existence of Growth and Adult Phases An in vivo animal model of osteoporosis should exhibit both growing and adult skeletal phases of meaningful duration. Peak bone mass is a concept in osteoporosis etiology that is receiving increased attention [28 – 30]. Its importance is clear because a single bone mass measurement at menopause is the best predictor of future fracture in healthy persons [31 – 33]. That value approximates peak bone mass, because bone mass changes before menopause are minor [34 – 38]. Growth processes, principally bone modeling and its modifiers like nutrition, physical activity, and heredity [30,39 – 55], determine peak bone mass. Adult phase skeletal processes (predominantly bone remodeling, but some bone modeling to effect shape changes in response to changing physical activity patterns) determine bone mass after attainment of peak bone mass [56]. The best animal models of osteoporosis would have both growth and adult phases of duration that allow both accuracy and time frame compression.
B. Menstrual/Estrus Cyclicity Humans not only have a menarche and regular, frequent ovulatory cycles, but also generally experience bone loss at cessation of ovarian function. The linkage of these two facts may seem weak, but a strong hint of the importance is the low bone mass that exists long term or transiently in amenorrheic individuals [58 – 64] and the bone accumulation that occurs upon resumption of normal menses [63 – 65]. Mammals (and humans) that have regular, frequent ovulatory cycles with high peaks of estradiol may be the only ones that suffer estrogen-depletion bone loss. Some hold that regularly cycling female mammals accumulate an estrogen-related component of bone that is integrated into the skeleton [66,67] and lost summarily at menopause, as if estrogen regulates bone mass [57]. It would follow that animals with infrequent cycles and low estradiol peaks might develop only a small estrogen-related bone compartment and show little estrogen depletion bone loss.
C. Natural Menopause Most women experience a natural menopause of 2 – 7 years duration [68]. Only 25 – 30% of women experience surgical menopause and many of those have preserved ovarian function or prompt estrogen replacement [69]. Most animal skeletal models of estrogen depletion invoke surgical or medical oophorectomy [13,70,71]. No meaningful differences in bone behavior between surgical and natural menopause are known [72,73]. Nonetheless, animals with a natural menopause with hormone changes like that of humans might be the best models for human osteoporosis. For instance the gradually increasing intermittency of estrogen peaks over a long period may elicit a different bone adaptive response than the precipitous removal of ovaries without estrogen replacement.
D. Bone Loss and Rise in Turnover Rate after Estrogen Depletion Following estrogen depletion, bone loss accelerates [74 – 77] for a time in multiple sites [78] and then decelerates and enters a semiplateau phase [79,80]. These changes are most pronounced in cancellous regions and at endocortical surfaces [81,82]. Estrogen status plays a much more important role in determining bone quantity in the aged female skeleton than does age [83 – 85]. Estrogen depletion changes in intracortical (Haversian) remodeling in humans are poorly documented. This cancellous and endocortical bone loss is accompanied by an increase in bone turnover [86,87] and a marked, transient negative calcium balance [88]. Histomorphometric changes of increased turnover across menopause are readily demonstrable within individual humans [89]. These behaviors should be easily confirmed in an accurate animal model of postmenopausal osteoporosis. It is even reasonable that the same measuring techniques applied in humans be expected to work for animals.
E. Skeletal Reponse to Estrogen Replacement Oophorectomized or menopausal women given prompt estrogen replacement experience a smaller rise in turnover [86,90], less bone loss [91 – 93], and fewer fractures [91,94 – 96] than those who receive no estrogen replacement [97]. This response is demonstrated well by histomorphometric technqiues in humans and animals [16,90,98]. This response is so stereotyped in adult women that one should expect an accurate ovariectomized animal model to experience an identical response to estrogen replacment.
32
DONALD B. KIMMEL
F. Development of Osteoporotic Fractures and Steady-State Osteopenia
H. Remodeling 1. CANCELLOUS
In its strictest definition, osteoporosis is marked by the occurrence of low trauma fractures of the spine and hip [99 – 101]. Because of the excellent ability to quantitate bone mass [102], the World Health Organization has recently provided a numeric definition of osteoporosis, as all women with bone mass at one or more bone sites 2.5 standard deviations below the young adult normal [103]. In the broadest sense, this is rational because women with osteoporotic fractures tend to be osteopenic [104 – 106]. While the WHO definition of osteoporosis should prove helpful for patients, clinicians, and clinical researchers alike, it is unlikely to help in animal experimentation. Though one might imagine experiments on osteopenic subpopulations of animals, the population-based studies necessary to screen for appropriatelyosteopenic animals for a WHO-like definition of osteoporosis are impractical. Enrolling only extremely osteopenic animals in trials, while appealing as a means of maximizing the chance of fragility fracture, is simply not feasible. Along those lines, it would be extremely helpful to have an accurate animal model that developed fragility fractures after estrogen depletion, because preclinical trials of the ability of agents to halt fragility fracture could be done. Despite the existence of good animal models for postmenopausal bone loss, none exhibit low trauma fragility fracture. This fact may speak to the importance of humans’ propensity to have both low peak bone mass and late-life bone loss, as necessary ingredients in the development of an osteoporotic fracture syndrome. Animals, in whom latelife bone loss can be readily created, may simply lack the contribution of low peak bone mass that is necessary to put them below their fracture threshold. The one animal model with low peak bone mass, the SAM/P6 mouse, has fragility fractures [107].
G. Bone Loss and Decreased Formation after Decreased Mechanical Usage Older humans experience loss of both cancellous and cortical bone as well as declines in bone formation unrelated to estrogen depletion [108,109]. These changes come during a life phase when a generalized decline in physical activity also occurs. While extreme physical inactivity, as during bedrest or paraplegia, causes marked bone loss [110 – 112], the impact of long-term, mild inactivity is neither well-understood nor easily assessed. This bone behavior should also be easily demonstrable in an accurate animal model of osteoporosis.
Cancellous bone remodeling, the in situ removal and replacement of aged bone tissue with new bone tissue, is an important process [56]. It has frequently been mentioned that bone surfaces adjacent to marrow cells experience high remodeling rates. This process is also expressed at endocortical surfaces, quite possibly initiated from marrow cavities [82]. An accurate animal model would display such activity in its skeleton. 2. CORTICAL Cortical bone plays a dominant role in skeletal structure. Adult humans have Haversian or intracortical bone remodeling, the process of in situ removal and replacement of aged cortical bone tissue [56]. Though the extent to which this process is affected by estrogen depletion has never been clearly defined, postmenopausal osteoporosis is normally characterized by only minimal cortical porosity. Nonetheless, an accurate animal model should display considerable steady-state Haversian remodeling, because Haversian remodeling is extremely important to the maintenance of cortical bone strength. Adverse changes in Haversian remodeling caused by agents being considered for treatment of osteoporosis can and should be revealed first in animal studies, not in Phase III or Phase IV human studies.
I. Time Frame Compression In adult, estrogen-depleted women, the phase of accelerated estrogen-depletion bone loss lasts 5 – 8 years. The time from attainment of peak bone mass until the development of fragility fractures is 30 or more years. An effective animal model known to experience peak bone mass followed by postovariectomy bone loss should compress both times by an order of magnitude or more.
J. Convenience Convenience for animal models is denominated as cost of purchase, availability, housing requirements, and handling difficulties and the necessity for designing/implementing/validating new analysis procedures. Using an animal model with the highest degree of accuracy can be so inconvenient that it is not worthwhile. It may occasionally be worse than doing a human study. For example, if an intervention requires active subject cooperation, animals may not be capable of that cooperation. On the other hand, in animal experiments, recruitment, lost sampling units, and compliance to pharmaceutical regimens, all nagging problems in clinical research, are nonissues.
33
CHAPTER 37 Animal Models for in Vivo Experimentation
Animals that seem convenient to some investigators because of specialized facilities and expertise at their organization are not the best choice for others lacking those tools. In today’s rapidly changing scientific environment, investigators tempted to maintain familiar, but outdated techniques often impede their own progress in the name of “convenience.” New enterprises, with the opportunity to allocate resources freshly, enjoy both the advantages and the burdens of the chance to define their own level of convenience. When an organization needs information that can be obtained only from experimental animals that require specialized housing and care, it often makes sense to contract externally for that phase of the project in the same manner that expert laboratories are asked to perform specialized analysis of various tissues. However one views it, the worst scenario is to have wasted scarce resources on an irrelevant model in the name of convenience.
V. ANIMAL MODELS FOR HUMAN OSTEOPOROSIS Animals that should receive initial consideration in osteoporosis research are bird, mouse, rat, rabbit, dog, pig, sheep, and nonhuman primate species. Because the above criteria are directed at clinical symptoms and outcomes in humans, the animal models will generally be evaluated as to how they duplicate those outcomes. Dissimilarities from the human condition in detailed mechanisms of development of one or more of the conditions may well exist. The
TABLE 1 Attribute
fit of each animal model to the above criteria is summarized in Table 1.
A. Avian Birds have growing and adult skeletal phases. Adult female birds have daily egg-laying cycles that correspond to alternating deposition and resorption of medullary bone during cyclic oviposition and egg calcification [113]. The bone accumulation phase occurs with rising serum estradiol and the removal phase accompanies falling serum estradiol. Estradiol treatment of male birds also causes medullary bone accumulation [114]. The bird skeleton experiences localized bone loss during immobilization and increased bone mass during applied mechanical loads [115 – 118]. Though avian models have broken new ground in understanding bone cell origin [119,120] and bone responses to the mechanical environment, they are not now known to be models for studying the estrogen/fracture-centered disease of osteoporosis. Furthermore, current data, mostly observational in nature, suggest that birds normally have little cancellous or Haversian remodeling. The interesting bone response to estrogen in birds provides many opportunities for experiments that bear on bone biology [121], but the estrogen-related bone buildup suggests a fundamental dissimilarity to adult mammalian physiology. While it can be loosely inferred that hypoestrogenemia in birds is associated with medullary bone loss, just as
Summary of in Vivo Animal Models for Osteoporosis
Human
Avian
Mouse
Rat
Dog
Pig
Sheep
Primate Yes
Growth and adult phases?
Yes
OK
OK
OK
Yes
Yes
Yes
Menstrual/estrus cyclicity
28 Days
Daily
Inducible
4 – 5 Days
205 Days
21 Days
21 Days 21 – 28 seasonal Days
Natural menopause
Yes
No
Yes*
Yes*
No
?
?
Yes
Bone loss after estrogen depletion
Yes
?
Probably
Yes
Not consistent
Weak
Weak
Yes
Response to estrogen
Turnover z4
Formation z3
Formation z3
Turnover z4
Not consistent
?
?
Turnover z4
Development of osteoporotic fractures
Yes
No
No
No
No
?
?
No
Cancellous remodeling
Yes
No
No
Some
Yes
Yes
Yes
Yes
Haversian remodeling
Present (study site difficult)
No
No
Low levels; inducible
Yes
Yes
Yes
Yes
Time frame compression
No
?
Yes
Yes
No
Some
Some
Some
*
Convenience
OK
Yes
Yes
Yes
Weak
Poor
Poor
Depends
Drug dose range like humans?
Yes
?
No
No (1/100)
Close
?
?
Yes
Cost effectiveness
Yes*
No
Yes
Yes
Weak
?
?
?
34
DONALD B. KIMMEL
estrogen depletion in mammals is associated with osteopenia, the course of bone mass following oophorectomy, an event more relevant to osteoporosis and osteopenia, is not known. The estrogen-related bone accumulation may actually prove important in understanding the accumulation of peak bone mass in pubertal humans [66,67,122 – 124], but seems likely to hinder the proper interpretation of experiments about osteoporosis, an adult disease. Birds are an example of a model that is convenient with low cost, but is largely irrelevant for osteoporosis research, because their skeletal behavior does not mimic features associated with adult human osteoporosis.
B. Mouse Though past success with using mice in physiologic aspects of skeletal research is encouraging, the future is even brighter because of the relative ease with which the mouse can be applied in genetic studies. It has been spectacular as a model for the human in osteopetrosis [125 – 129], osteoclast and stromal cell ontogeny [130], and cytokine and marrow studies [131 – 134]. The mouse has become even more popular for the ease with which its genome can be manipulated [135 – 138] and investigated. For example, C57Bl/6J and C3H/HeJ mice have low and high peak bone mass, respectively, which lend them to genetic investigations using F2 cross and F1 backcross studies of extremes of a continuous phenotype [139]. It is logical that the mouse be considered as an animal model that can solve new problems in the osteoporosis field. However, data that would validate the mouse as an in vivo model for osteoporosis research are in short supply. The few existing studies suggest that, much like in the rat, cancellous [138,140], but not cortical [141], bone loss occurs soon after ovariectomy. Genetically hypogonadal female mice are osteopenic [142]. Estrogen-depletion bone loss appears to be prevented by estrogen replacement [140]. Ectopic bone ossicles are preserved by estradiol treatment [143]. Mice experience cancellous and cortical bone loss in the vertebrae [144,145] and femur [146] during the second year of life. However, increased bone formation with new woven bone deposition after estrogen administration, as in birds, is a routine finding [140,141,147 – 150] and may depend upon the presence of the uterus [149]. The mouse’s inducible estrus cycle, rather than a hypothalamic – pituitary-regulated cycle, is another important dissimilarity to humans. Validation of the mouse as a model for osteoporosis will take some targeted experimental work. The few available reports suggest that the mouse suffers estrogen-depletion bone loss which is stopped by estrogen replacement. Its time course and the site specificity for the development of estrogen-depletion osteopenia must be established, perhaps
in a strain-specific fashion, as it has been for the rat [13,14]. The aberrant formation response to estradiol at endocortical surfaces must be considered. This unique response may be age-related. Until it is shown either that an older mouse has no such anabolic response to estradiol or that estradiol can elicit a bone formation response in larger adult mammals or adult humans, the chances for an acceptable murine model for osteoporosis seem remote. Furthermore, the similar life span of mice and rats, the smaller, though adequate, bone specimens provided by the mouse, and the overall similar cost for experimentation suggests that it will take an application of the mouse not possible in the rat, like gene manipulation or genetic marker studies, to induce investigators to choose the mouse. 1. THE SAM MOUSE The SAM/P6 (senescence-accelerated mouse) mouse has low peak bone mass and develops fractures in middle and old age [107,152 – 154]. It is the only experimental animal with documented fragility fractures of aging. The SAM mouse needs a full genetic [155], hormonal [156], and biomechanical characterization, including site specificity for fractures. If it does not suffer from collagen defects like those in osteogenesis imperfecta [135,157], it may, when combined with standard osteopenic prevention approaches like estrogen or bisphosphonate treatment, provide an opportunity to study the role of low peak bone mass in causing late-life fractures. It may provide further opportunity to study treatments that would enhance peak bone mass. Such studies must include testing of the material properties of the bone tissue itself, as has been done for MOV-13 osteogenesis imperfecta mice [137]. In the near term, considering the availability of SAM/P6, C57Bl/6J, and C3H/HeJ mice and the existing knowledge of the mouse genome [158], using proper crossing and back-crossing techniques [9], and cDNA probing of previously mapped genetic loci should be fruitful. A whole genome search should be able to identify polymorphisms at one or more genetic loci that are linked to low or high bone mass. The disclosure of such markers can lead, at the very least, to an increased understanding of peak bone mass control in one species, the mouse. Paradigms developed in small animals can subsequently be tested in larger animals or humans.
C. Rat The rat has a long history of providing accurate fundamental data that apply to the adult human skeleton. It gave the first evidence that osteoclasts ingest bone substance [159] and early evidence about osteoclast origin [160 – 163]. The rat skeleton was once held unsuitable as an adult human skeletal model because many epiphyseal
CHAPTER 37 Animal Models for in Vivo Experimentation
growth cartilages in male rats remain open past age 30 months [164]. The past unintentional focus on male rats masked important sex differences in epiphyseal closure times. In recent osteoporosis research projects using female rats, it became apparent that anatomically identical growth cartilages close earlier in females than in males [165 – 168]. Studies of the effects of gonadal hormones on growing male and female rats also suggest that intact females cease growing earlier than males [169]. When studies of rat models of male osteoporosis are undertaken, the results may be more difficult to interpret than in female rats [25]. Bone elongation ceases and effective epiphyseal closure ensues at important sampling sites in female rats by age 6 – 9 months, an age after which considerable useful experimental lifespan remains [170]. Periosteal expansion continues until about age 10 months, marking the age of peak bone mass in the adult female rat [171,172]. The mean healthy life span for these rats is 21 – 24 months, after which the incidence of estrogen-dependent mammary tumors rises to unacceptable levels. Thus the female rat has an appreciable life span both before and after attainment of peak bone mass. The lengthy bone accumulation period in the rat presents ample untapped opportunities for the study of both internal and external determinants of peak bone mass. Even those holding strong opinions about the inappropriateness of the rat as a model of adult human skeletal disease because of its “continuous growth and lack of remodeling,” have relented. They now only caution to use female rats of at least 6 – 9 months age and avoid studying Haversian remodeling [173]. Adult female rats have a regular estrus cycle in which E2 levels spike for 18 h every 4 days [174]. During the second year of life, the fraction of rats found in constant diestrus rises gradually [175], and cancellous bone loss is frequently observed [171]. While this is not a true menopause, spikes in E2 cease as bone loss occurs, making a linkage of rat “menopause” to cancellous bone loss possible. Following ovariectomy, loss of cancellous bone mass and strength occurs, then decelerates in a site-specific fashion, to enter a plateau phase [13,14,176 – 181]. This cancellous bone loss is accompanied by an increase in the rate of bone turnover [14,182 – 183]. These features mimic well the bone changes that follow oophorectomy or natural menopause in humans. Not all cancellous bone sites in the rat experience such bone loss [184], further tightening the parallel of the rat and the human skeletons, since osteoporotic fragility fractures and osteopenia are limited to a few sites in humans [78,185]. Dual energy X-ray absorptiometry (DXA), the current state-of-the-art for measuring bone mass in humans, can also be readily applied for rats [179,186 – 188]. Ovariectomized rats given prompt estrogen replacement experience no rise in turnover [16,98,189] and no bone loss [16,190 – 192]. This response fully parallels that seen in
35 postmenopausal women receiving estrogen replacement. Agents like bisphosphonates [15,180,193,194] and calcitonin [195,196] also block the rise in turnover and bone loss in rats, just as they do in humans [197 – 203]. Though interpretations of data from one laboratory suggest that high doses of estradiol stimulate bone formation in the rat [204,205], this contention has been largely discounted by others using similar experimental designs who offer alternative explanations that correctly consider the use of growing rats and an unusually long fluorochrome interval [206]. The rat, like other experimental animal models of osteoporosis, has no fragility fractures associated with the development of osteopenia. This apparent shortcoming can be overcome by mechanical tests of various bone through servohydraulic test systems. Such assays now exist for the vertebral body [17,18,191], femoral shaft [207], and proximal femur [18,20,208,209]. Prompt estrogen replacement in 1year-old ovariectomized rats preserves vertebral body strength in rats [210], much as it reduces fracture incidence in older women [6,91,92,94,96,214,215]. Using rat vertebral body strength as a surrogate for human vertebral body strength (implied from fracture likelihood) is likely to enable additional preclinical evaluation end points for agents being considered as treatments that can decrease osteoporotic fracture incidence. Such preclinical testing may avoid the problems encountered during the investigation of sodium fluoride as a treatment for osteoporosis [10,11]. Rats lose bone following immobilization. Several methods of permanent and temporary immobilization are available [21,22,216 – 220]. Acute phase [218], chronic phase [219], and recovery phase [220,221,223] bone changes related to disuse are easily studied. Rats are a weak model of glucocorticoid osteopenia, because they exhibit decreased bone formation, but do not consistently develop osteopenia [224,226]. Adult rats have adequate amounts of cancellous bone remodeling to permit useful experiments [173,227,228]. Controversy exists as to the relative frequency of modeling and remodeling activity. Classic reversal lines at the base of cancellous osteons are not present in 4-month-old rats [204], suggesting that modeling predominates at this age. While modeling would indeed be expected to be most frequent in such young rats, studies of cement line morphology in older rats are needed to confirm that bone activity in rats does undergo the expected transition to remodeling activity during adulthood [56]. In spite of the lack of studies of cement line morphology, indirect reasoning strongly supports the existence of remodeling in rat cancellous bone. Many cancellous bone regions in adult rats maintain stable bone mass for long periods of time while showing abundant bone formation and resorption. This neutral balance for resorption and formation [13,178] suggests the presence of remodeling. Whether the resorption and formation activity is locally or nonlocally linked, the regional tendency
36 toward mass preservation is more suggestive of remodeling than modeling, which is usually associated with changes in bone mass and shape [56]. In most of its cortical bone, the rat displays low levels of Haversian remodeling, often not detectably different from zero. However, processes resembling intracortical remodeling are induced by strong anabolic agents [229] or stressful metabolic conditions [230,235]. Unfortunately, it is not known whether these same agents and conditions also accelerate Haversian remodeling in humans. Current data suggest that the rat has such low levels of Haversian remodeling that it is impractical to use it for analysis of Haversian remodeling behavior, especially when evaluating agents that suppress Haversian remodeling. These data come from cross sections taken at the tibio – fibula junction or mid-femur [231,232]. However, cortical bone regions that surround cancellous bone, as in long bone metaphyses, are a reasonable, and currently uninvestigated, site to check to find higher levels of Haversian remodeling. In 3-month-old ovariectomized rats, the phase of accelerated estrogen-depletion bone loss lasts 3 – 4 months in the proximal tibial metaphysis [13], a 20-fold time frame compression when compared to that of estrogen-deplete women. The rat reaches peak bone mass by age 10 months, a 30-fold time frame compression when compared to that of the adult human. The rat is also among the most convenient of experiment animals to handle and house. In executing experiments with ovariectomized rats, it is usually wise to make some attempt at food restriction to limit weight gain in the ovariectomized groups. This may take the form of pair feeding to a sham group (conservative) or weight restriction to a sham group (aggressive). Either type of food restriction will accentuate bone loss [233,234], possibly creating more reliable cortical bone loss and certainly quickening the development of osteopenia. When selecting aged female rats for study, investigators are often tempted to choose “retired breeders” because of their low cost. However, the highly variable skeletal status of retired breeders is likely to influence experimental outcomes. Most female rats are bred for the first time at age 3 – 4 months, before bone elongation is completed and long before peak bone mass is achieved. They often eat a diet that is 0.6% calcium and 0.55% phosphorus. The calcium demands of pregnancy and lactation inhibit normal patterns of skeletal growth. Thus, rats delivered to investigators as retired breeders at age 6 or 7 months have usually had two or three litters plus calcium-stressful lactation periods [235,236], causing them to have abnormally immature skeletons with long bone metaphyses that are relatively depleted of cancellous bone. When they are taken from breeding and placed in an experiment, they experience a period of “catch-up” growth during which cancellous bone of the metaphysis is repleted by resumption of endochondral processes and considerable cortical bone enhancement
DONALD B. KIMMEL
occurs, such that they reach a usual peak bone mass by age 10 – 12 months. This period of catch-up growth might be confused with the anabolic effect of an agent. Thus, the retired breeder female rat is somewhat acceptable when experiments to add bone to the osteopenic skeleton are planned, because its skeleton is fundamentally immature and osteopenic. The interaction of delayed growth processes with potentially anabolic agents is likely to be a problem for quantitative interpretation. However, when doing prevention experiments, the retired breeder female rat is generally unreliable because of both the (likely) already osteopenic condition of the skeleton, from which little more bone could be lost, and the period of catch-up growth. Virgin female rats that are more skeletally mature and not osteopenic in their long bone metaphyses, are much more acceptable for prevention studies, because they have considerably more bone in their long bone metaphyses that can be lost and whose loss can be prevented after ovariectomy [237]. 1. SUMMARY The FDA has made a wise choice in requiring rat experiments during osteoporosis research. The ovariectomized rat is an excellent model that correctly emulates the most important clinical features of the estrogen-deplete adult human skeleton. Its site-specific development of cancellous osteopenia is one of the most certain physiologic responses in skeletal research. Ample time exists for experimental designs that either preventestrogen-depletion bone loss or restore bone lost after estrogen depletion. Its response to estrogen replacement closely parallels that seen in the human. The rat’s low levels of Haversian remodeling present little immediate problem when testing agents for their ability to prevent the loss of cancellous bone or rebuild lost cancellous bone. Investigators can compensate for the lack of fragility fractures by biomechanical testing. Rats are convenient for most investigators; unbred females ages 6 – 10 months are the optimal choice. Existing laboratory measurement tools of biochemistry, histomorphometry, densitometry, and mechanical testing are readily applicable.
D. Guinea Pig, Rabbit, Ferret, and Cat Though occasional reports using guinea pigs, rabbits, ferrets, and cats in osteoporosis research have appeared [238 – 240], few studies of estrogen depletion bone loss exist. Too few experiments generally exist to assess properly their validity. Seven-month-old ovariectomized guinea pigs do not lose bone by 4 months postsurgery [239]. Adult rabbits have active Haversian remodeling and might serve as a model for testing Haversian remodeling after treatment with agents that have strong anabolic effects on cancellous bone. Their reproductive cycle bears dissimilarity to that of
37
CHAPTER 37 Animal Models for in Vivo Experimentation
humans. The success of the rabbit and dog [238,241], animals with significant Haversian remodeling, as models for glucocorticoid-induced osteopenia, coupled with the failure of the rat, an animal with mimimal Haversian remodeling, suggests that glucocorticoid osteopenia is a disease of deranged Haversian remodeling. The ferret, weighing less than 1 kg, has Haversian remodeling [242]. Its normal skeletal physiology, including the accumulation of estrogen-dependent bone that seems to accompany normal cyclicity in other mammals, is dependent on a regular light cycle [243]. It exhibits expected changes in bone remodeling rate and bone volume during treatment with parathyroid hormone [244,245]. Though a relatively small animal with appreciable quantities of Haversian remodeling would be welcome, the general appeal of other animal models with both Haversian remodeling and already proven human-like endocrine characteristics cannot be overlooked.
E. Dog The adult dog is a reliable model for the adult human skeleton that is generally similar in both metabolic and structural characteristics. Studies of 239Pu-injected beagles not only first proved the existence of adult cancellous bone remodeling [159], but also contributed one of the earliest indications of the hematogenous origin of osteoclasts [246]. The ratio of cortical to cancellous bone is similar to that in humans [247 – 249]. Haversian and cancellous osteons remodel with similar morphology, though more rapidly in dogs [250,251]. Skeletal responsiveness of the adult dog parallels that of the adult human for corticosteroids [241,252], uremia [253,254], bisphosphonates [255,256], disuse [110 – 112,257 – 260], and parathyroid hormone excess [245,261]. In contrast to all other applications for adult beagles as a model of the adult human skeleton, the oophorectomized beagle is controversial. Many individual studies lack significant findings, but a metaanalysis of the overall data [262,263] suggests that estrogen-depletion bone loss of perhaps 8 – 10% annually occurs in oophorectomized beagles. This conclusion is based on the three most prominent types of studies: (i) those that find a significant decline [264 – 266,270 – 273], (ii) those that find a nonsignificant decline [71,267,268,274], and (iii) those that find no change in bone mass or strength [23,269,275 – 282]. Increases rarely occur [283,284], usually in experiments of very small sample size (N 3). One strongly positive paper showing bone loss [264] uses spinal densitometry, a tool of high precision [285]. A mildly positive paper found some decline in vertebral trabecular strength, using mechanical testing, a tool with good specificity, though weak precision for trabecular bone evaluation [267,268].
The temporal pattern in bone formation after oophorectomy suggests an early rise with a later return to baseline or subbaseline levels. Work from one group [270 – 272] suggests that formation falls rapidly to 50% of baseline, with no increased turnover phase. This suggests the possibility of a marked dissimilarity to histomorphometric findings in transmenopausal humans, where the early rise in both formation and individual cell activity [89] has been well documented [86]. OF
1. THE CANINE ESTRUS CYCLE COMPARED TO THOSE HUMANS AND OTHER ANIMALS
E2 levels are usually very low in the dog, rising twice yearly for several weeks [286 – 288]. In rats, E2 spikes for 18 h every 4 days [174]. In women, E2 spikes for 1 or 2 days monthly [289,290]. The estrus cycle in monkeys has a similar frequency to that in humans, but reaches E2 peaks only about half as high [291]. Integrated estrogen exposure in dogs, though only marginally less than in rats, is only onefourth that in humans. It is similar to that in primates, except during the peak periods. E2 peaks in canines are one-sixth as frequent and one-sixth as high as in humans. This difference could contribute to the dog’s developing only a small estrogen-dependent compartment of cancellous bone. Because of the inconsistency of the ovariectomized dog model, predicting the outcome of experiments with it is difficult. The data suggest that estrogen treatment suppresses turnover, but that its bone mass sparing effects are uncertain [279 – 282]. 2. SUMMARY The adult beagle is an excellent model of the adult human skeleton except for estrogen depletion. Its principal advantage over smaller animals is its Haversian remodeling. Though the oophorectomized beagle has estrogen-depletion osteopenia, poor interlaboratory reproducibility has caused skepticism. The main problem has been that most of the individual studies contain insufficient power to detect the expected bone loss [265 – 284]. Beagles are less estrogen-replete than women and may have a smaller estrogendependent compartment of bone in their skeleton. Despite its inconsistency for developing estrogen-depletion bone loss, the dog remains an excellent model for testing the effects on Haversian remodeling of agents that have strong anabolic effects on cancellous bone. However, a general strategy of using one animal model displaying both Haversian remodeling and consistent estrogen-depletion bone loss has high appeal.
F. Pig The pig has both growing and adult skeletal phases. The oophorectomized pig has been tested a few times
38
DONALD B. KIMMEL
[292 – 294]. In one study, though minor structural deterioration was noted, no differences in bone mass, either by densitometry or histomorphometry, were seen. In a second study, minor bone loss from the fourth lumbar vertebra was noted at 3 months postovariectomy. The pig has a regular estrus cycle that is somewhat shorter than the human menstrual cycle. Pigs have been used successfully to study fluoride and exercise effects on the skeleton [295,296]. It appears that the age of peak bone mass in the pig is older than 2 – 3 years, a factor that has confounded experimental efforts thus far and seems destined to introduce logistical difficulty into designing experiments with the estrogen-deplete pig. More work will be necessary before the pig can gain acceptance as a model of estrogen-depletion bone loss.
G. Sheep The ewe has both growing and adult skeletal phases, but the age of peak bone mass is not known. Some species have a regular estrus cycle during the short days of winter and experience anestrus when days are longer [297,298]. The ewe has also been used to study fluoride effects in the adult skeleton. The findings seem to parallel histologic changes in humans with frequent signs of increased formation, sluggish mineralization, and toxicity to bone forming cells [26,299 – 301]. Skeletal behavior over oophorectomy has only recently been addressed [302]. Early data suggest a postoophorectomy picture that is considerably less straightforward than that seen in postmenopausal women, rats, or monkeys. Glucocorticoid data seem consistent with findings in other animal models and humans [252,304]. Sheep can be housed readily at most vivariums and pose little problem for handling. More data are needed to determine the age of peak bone mass, and the role of seasonal variation in bone mass, before the usefulness of the adult ewe as an in vivo model of osteoporosis can be validated.
H. Nonhuman Primate The nonhuman primate has both growing and adult skeletal phases. Peak bone mass occurs around age 10 – 11 years in cynomolgous and rhesus monkeys and baboons [305 – 307]. All nonhuman primates have a regular menstrual cycle with an approximate duration of 28 days, an excellent analog of adult women. Nonhuman primates experience a natural menopause near the end of the second decade of life. Nonhuman primates in captive environments are ideal candidates for family pedigree analysis, the most powerful way of establishing genetic linkage to phenotypic traits [9]. The ability to study DNA polymorphisms in nonhuman primates has also been demonstrated [308,309]. Combining
genetic studies of the mouse and nonhuman primate offers a multispecies approach to understanding the genetic control of peak bone mass that can lead to the development of concepts that could be applied in humans. Nonhuman primates experience decreased bone mass and bone strength with increased turnover [310], after ovariectomy [70,310 – 315], premature menopause [316], or GnRH agonist treatment [70]. The response to estrogen replacement has not been well-characterized. Nonhuman primates experience bone loss with age [317 – 318] and after prolonged immobilization [319]. Histomorphometric studies of primates and humans yield remarkably similar values [306 – 310,320,321]. Postovariectomy changes in bone mass are frequently masked by the selection of animals of an age that they are still acquiring peak bone mass [322 – 324]. Late life spinal pathology in baboons [326,327] and rhesus monkeys [318,325] is mostly osteoarthritis. Baboons experience osteopenia [317] with an age-related decline in anterior vertebral height that bears more similarity to that accompanying osteoarthritis than to vertebral crush fractures [326 – 328]. This means that nonhuman primates are not likely to be a model of fragility fractures, because in humans osteoarthritis and osteoporosis tend to be mutually exclusive conditions [328 – 330]. In addition, spinal bone mass measurements in older primates may be affected because the osteophytes that occur in osteoarthritis have enough calcium to change spinal bone mass values meaningfully and obscure proper interpretation [331 – 334]. Like the dog, nonhuman primates offer cancellous and Haversian remodeling that is directly comparable to that found in the adult human skeleton. The time course of bone loss after ovariectomy has not been well-characterized, but appears to offer significant time frame compression when compared to that in humans. Like all skeletal experiments intended to yield information about osteoporosis, care should be exercised in choosing animals that have achieved peak bone mass. Extreme requirements for housing and care of nonhuman primates limit their use to relatively small numbers of facilities. However, when handled by experienced staff in an appropriate environment, they present few care problems.
I. Summary of All in Vivo Animal Models When an in vivo osteoporosis research project cannot be done in humans, the 10-month-old female rat is the first animal model of choice. It has reached peak bone mass and can be manipulated to accurately simulate most clinical findings of osteoporosis in the adult female skeleton. Methods like serum biochemistry, histomorphometry, and densitometry that are routinely used in humans, are applicable in rats. Like all animal models of osteoporosis, the rat
CHAPTER 37 Animal Models for in Vivo Experimentation
develops no fragility fractures. However, mechanical testing of rat bones substitutes well as an accurate predictor of bone fragility. However, the rat’s low levels of Haversian remodeling do not permit accurate evaluation of intracortical bone behavior, making studies of larger animals a must. The dog is generally an accurate model of the adult human skeleton, but has yielded inconsistent results in the acutely estrogen-deplete state that differ from those in humans. Data on all other oophorectomized species except nonhuman primates, are scarce. Estrogen-deplete nonhuman primates are the large animal of choice when Haversian remodeling outcomes are to be studied. Avians have a metabolically active skeleton whose behavior does not seem directly relevant to osteoporosis research. Data about estrogen-deplete mice now seem promising, but the high state of development of the ovariectomized rat model suggests that developing an ovariectomized mouse model as an alternative is not urgent. When studied before age 10 months, female rats and mice have skeletal growth and maturation phases that are useful for experiments about peak bone mass. Mice are likely to be useful in revealing both genetic markers that control peak bone mass and gene manipulations that affect both bone mass and structure. The rabbit is the animal model of choice for studying glucocorticoid osteoporosis.
References 1. “Guidelines for Preclinical and Clinical Evaluation of Agents Used in the Prevention or Treatment of Postmenopausal Osteoporosis.” Division of Metabolism and Endocrine Drug Products, Food and Drug Administration, Washington, DC, 1994. 2. K. Faulkner, S. R. Cummings, D. Black, L. Palermo, C. C. Gluer, and H. K. Genant, Simple measurement of femoral geometry predicts hip fracture: The study of osteoporotic fractures. J. Bone Miner. Res. 8, 1211 – 1218 (1993). 3. S. B. Soroko, E. L. Barrett-Conner, S. L. Edelstein, and D. Kritz-Silverstein, Family history of osteoporosis and bone mineral density at the axial skeleton: The Rancho Bernardo study. J. Bone Miner. Res. 9, 761 – 770 (1994). 4. I. Elffors, E. Allander, J. A. Kanis, B. Gullberg, O. Johnell, J. Dequeker, G. Dilsen, C. Gennari, A. A. Lopes-Vaz, G. Lyritis, G. F. Mazzuoli, L. Miravet, M. Passeri, R. Perez Cano, A. Rapado, and C. Ribot, The variable incidence of hip fracture in Southern Europe: The MEDOS study. Osteoporosis Int. 4, 253 – 263 (1994). 5. G. Jones, T. Nguyen, P. N. Sambrook, P. J. Kelly, C. Gilbert, and J. A. Eisman, Symptomatic fracture incidence in elderly men and women: The Dubbo osteoporosis epidemiology study (DOES). Osteoporosis Int. 4, 277 – 282 (1994). 6. D. P. Kiel, D. T. Felson, J. J. Anderson, P. W. F., and Wilson, M. A. Moskowitz, Hip fracture and the use of estrogens in postmenopausal women. The Framingham Study. N. Engl. J. Med. 317, 1169 – 1174 (1987). 7. B. K. Suarez, J. Rice, and T. Reich, The generalized sib pair IBD distribution: Its use in the detection of linkage. Ann. Hum. Genet. 42, 87 – 92 (1978). 8. N. Risch, Linkage strategies for genetically complex traits. II. The power of affected relative pairs. Am. J. Hum. Genet. 46, 229–241 (1990).
39 9. E. S. Lander and N. J. Schork, Genetic dissection of complex traits. Science 263, 2037 – 2048 (1994). 10. M. Kleerekoper, E. L. Peterson, D. A. Nelson, E. Phillips, M. A. Schork, B. C. Tilley, and A. M. Parfitt, A randomized trial of sodium fluoride as a treatment for postmenopausal osteoporosis. Calcif. Tissue Int. 49, 155 – 161 (1991). 11. B. L. Riggs, S. F. Hodgson, W. M. O’Fallon, E. Y. Chao, H. W. Wahner, J. M. Muhs, S. L. Cedel, and L. J. Melton, Effect of fluoride treatment on the fracture rate in postmenopausal women with osteoporosis. N. Engl. J. Med. 322, 802 – 809 (1990). 12. B. L. Riggs, W. M. O’Fallon, A. Lane, S. F. Hodgson, H. W. Wahner, J. Muhs, E. Chao, and L. J. Melton III, Clinical trial of fluoride therapy in postmenopausal osteoporotic women: Extended observations and additional analysis. J. Bone Miner. Res. 9, 265 – 275 (1994). 13. T. J. Wronski, L. M. Dann, K. S. Scott, and M. Cintron, Long-term effects of ovariectomy and aging on the rat skeleton. Calcif. Tissue Int. 45, 360 – 366 (1989). 14. T. J. Wronski, M. Cintron, and L. M. Dann, Temporal relationship between bone loss and increased bone turnover in ovariectomized rats. Calcif. Tissue Int. 42, 179 – 183 (1988). 15. T. J. Wronski, L. M. Dann, K. S. Scott, and L. R. Crooke, Endocrine and pharmacological suppressors of bone turnover protect against osteopenia in ovariectomized rats. Endocrinology 125, 810 – 816 (1989). 16. T. J. Wronski, M. Cintron, A. L. Doherty, and L. M. Dann, Estrogen treatment prevents osteopenia and depresses bone turnover in ovariectomized rats. Endocrinology 123, 681 – 686 (1988). 17. Li Mosekilde, C. H. Sogaard, C. C. Danielson, O. Torring, and M. H. L. Nilsson, The anabolic effects of human parathyroid hormone (hPTH) on rat vertebral body mass are also reflected in the quality of bone, assessed by biomechanical testing: A comparison study between hPTH-(1 – 34) and hPTH-(1 – 84) Endocrinology 129, 421 – 428 (1991). 18. B. C. Toolan, M. Shea, E. R. Myers, R. E. Borchers, J. G. Seedor, H. Quartuccio, G. Rodan, and W. C. Hayes, Effects of 4-amino-1-hydroxybutylidene bisphosphonate on bone biomechanics in rats. J. Bone Miner. Res. 7, 1399 – 1406 (1992). 19. M. Kasra and M. D. Grynpas, Effect of long-term ovariectomy on bone mechanical properites in young female cynomolgus monkeys. Bone 15, 577 – 561. (1994). 20. C. H. Sogaard, T. J. Wronski, J. E. McOsker, and Li Mosekilde, The positive effect of PTH on femoral neck bone strength in ovariectomized rats is more pronounced than that of estrogen or bisphosphonate. Endocrinology 134, 650 – 657 (1994). 21. J. A. Svesatikoglou and S. Mattson, Changes in composition and metabolic activity of the skeletal parts of the extremity of the adult rat following resection of the sciatic nerve. Acta Chir. Scand. (Suppl. 476), 16 – 20 (1976). 22. G. K. Wakley, B. L. Baum, K. S. Hannon, and R. T. Turner, The effects of tamoxifen on the osteopenia induced by sciatic neurotomy in the rat: A histomorphometric study. Calcif. Tissue Int. 43, 383 – 388 (1988). 23. P. Geusens, L. P. C. Schot, J. Nijs, and J. Dequeker, Calcium-deficient diet in ovariectomized dogs limits the effects of 17-b-estradiol and nandrolone decanoate on bone. J. Bone Miner. Res. 6, 791 – 798 (1991). 24. J. B. Rodgers, M. C. Monier-Faugere, and H. H. Malluche, Animal models for the study of bone loss after cessation of ovarian function. Bone 14, 369 – 377 (1993). 25. D. Vanderschueren, E. Van Herck, P. Schot, E. Rush, T. Einhorn, P. Geusens, and R. Bouillon, The aged male rat as a model for human osteoporosis: Evaluation by nondestructive measurements and biomechanical testing. Calcif. Tissue Int. 53, 342 – 347 (1993). 26. P. Chavassieux, Bone effects of fluoride in animal models in vivo. A review and a recent study. J. Bone Miner. Res. 5 (Suppl. 1), S95 – S100 (1990).
40 27. M. Cesnjaj, A. Stavljenic, and S. Vukicevic, In vivo models in the study of osteopenias. Eur. J. Clin. Chem. Clin. Biochem. 29, 211 – 219 (1991). 28. C. H. Chesnut, Theoretical overview: Bone development, peak bone mass, bone loss, and fracture risk. Am. J. Med. 91 (Suppl. 5B), 2S – 4S (1991). 29. V. Matkovic, D. Fontana, C. Tominac, P. Goel, and C. H. Chesnut III, Factors that influence peak bone mass formation: A study of calcium balance and the inheritance of bone mass in adolescent females. Am. J. Clin. Nutr. 52, 878 – 888 (1990). 30. V. Matkovic, Calicum intake and peak bone mass. N. Engl. J. Med. 327, 119 – 120 (1992). 31. S. L. Hui, C. W. Slemenda, and C. C. Johnston, Jr., Baseline measurement of bone mass predicts fracture in white women. Ann. Int. Med. 111, 355 – 361 (1989). 32. P. Gardsell, O. Johnell, and B. E. Nilsson, Predicting fractures in women by using forearm bone densitometry. Calcif. Tissue Int. 45, 235 – 242 (1989). 33. P. D. Ross, R. D. Wasnich, and J. M. Vogel, Detection of prefracture spinal osteoporosis using bone mineral absorptiometry. J. Bone Miner. Res. 3, 1 – 11 (1988). 34. R. R. Recker, J. M. Lappe, K. M. Davies, and D. B. Kimmel, Change in bone mass immediately before menopause. J. Bone Miner. Res. 7, 857 – 862 (1992). 35. A. Rodin, B. Murby, M. A. Smith, M. Caleffi, I. Fentimen, M. G. Chapman, and I. Fogelman, Premenopausal bone loss inthe lumbar spine and femoral neck: A study of 225 Caucasian women. Bone 11, 1 – 15 (1990). 36. J. Schlechte, L Walkner, and M. Kathol, A longitudinal analysis of premenopausal bone loss in healthy women and women with hyperprolactinemia. J. Clin. Endocrinol. Metab. 75, 698 – 703 (1992). 37. R. Marcus, J. Kosek, A. Pfefferbaum, and S. Horning, Age-related loss of trabecular bone in premenopausal women: A biopsy study. Calcif. Tissue Int. 35, 406 – 409 (1983). 38. H. E. Meema and S. Meema, Longitudinal microradioscopic comparisons on endosteal and juxtaendosteal bone loss in premenopausal and postmenopausal women, and in those with end stage renal disease. Bone 8, 343 – 350 (1988). 39. C. C. Johnston, J. Z. Miller, C. W. Slemenda, T. K. Reister, S. Hui, J. C. Christian, and M. Peacock, Calcium supplementation and increases in bone mineral density in children. N. Engl. J. Med. 327, 82 – 87. (1992). 40. H. H. Jones, J. D. Priest, W. C. Hayes, C. C. Tichenor, and D. A. Nagel, Humeral hypertrophy in response to exercise. J. Bone Jt. Surg. 59A, 204 – 208 (1977). 41. H. Haapasalo, P. Kannus, H. Sievanen, A. Heinonen, P. Oja, and I. Vuori, Long-term unilateral loading and bone mineral density and content in female squash players. Calcif. Tissue Int. 54, 249 – 255 (1994). 42. T. J. Beck, C. B. Ruff, W. W. Scott, C. C. Plato, J. D. Tobin, and C. A. Quan, Sex differences in geometry of the femoral neck with aging: A structural analysis of bone mineral data. Calcif. Tissue Int. 50, 24 – 29. 43. P. J. Kelly, L. Twomey, P. N. Sambrook, and J. A. Eisman, Sex differences in peak adult bone mineral density. J. Bone Miner. Res. 5, 1169 – 1175 (1990). 44. Y. Liel, J. Edwards, J. Shary, K. M. Spicer, L. Gordon, and N. H. Bell, The effects of race and body habitus on bone mineral density of the radius, hip, and spine in premenopausal women. J. Clin. Endocrinol. Metab. 66, 1247 – 1250 (1988). 45. D. P. McCormick, S. W. Ponder, H. D. Fawcett, and J. L. Palmer, Spinal bone mineral density in 335 normal and obese children and adolescents: Evidence for ethnic and sex differences. J. Bone Miner. Res. 6, 507 – 525 (1991). 46. M. Moller, A. Horsman, B. Harvald, M. Hauge, K. Henningsen, and B. E. C. Nordin, Metacarpal morphometry in monozygotic and dizygotic elderly twins. Calcif. Tissue Res. 25, 197 – 201 (1978).
DONALD B. KIMMEL 47. J. Dequeker, J. Nijs, A. Verstraeten, P. Geusens, and G. Gevers, Genetic determinants of bone mineral content at the spine and radius: A twin study. Bone 8, 207 – 209 (1987). 48. D. M. Smith, W. E. Nance, W. K. Ke, J. C. Christian, and C. C. Johnston, Genetic factors in determining bone mass. J. Clin. Inv. 52, 2800 – 2808 (1973). 49. N. A. Pocock, J. A. Eisman, J. L. Hopper, M. G. Yeates, P. N. Sambrook, and S. Ebert, Genetic determinants of bone mass in adults: A twin study. J. Clin. Inv. 80, 706 – 710 (1991). 50. E. Seeman, J. L. Hopper, L. A. Bach, M. E. Cooper, E. Parkinson, J. McKay, and G. Jerums, Reduced bone mass in daughters of women with osteoporosis. N. Engl. J. Med. 320, 554 – 558 (1989). 51. R. A. Evans, G. M. Marel, E. K. Lancaster, S. Kos, M. Evans, and S. Y. P. Wong, Bone mass is low in relatives of osteoporotic patients. Ann. Int. Med. 109, 870 – 873 (1988). 52. F. A. Tylavsky, A. D. Bortz, R. L. Hancock, and J. J. B. Anderson, Familial resemblance of radial bone mass between premenopausal mothers and their college-age daughters. Calcif. Tissue Int. 45, 265 – 272 (1989). 53. N. A. Morrison, J. C. Qi, A. Tokita, P. J. Kelly, L. Crofts, T. V. Nguyen, P. N. Sambrook, and J. A. Eisman, Prediction of bone density from vitamin D receptor alleles. Nature 367, 284 – 287 (1994). 54. N. A. Morrison, R. Yeoman, P. J. Kelly, and J. A. Eisman, Contribution of trans-acting factor alleles to normal phsyiological variability: Vitamin D receptor gene polymorphisms and circulating osteocalcin. Proc. Natl. Acad. Sci. USA 89, 6665 – 6669 (1992). 55. F. G. Hustmyer, M. Peacock, S. Hui, C. C. Johnston, and J. Christian, Bone mineral density in relation to polymorphism at the vitamin D receptor gene locus. J. Clin. Inv. 94, 2130 – 2134 (1994). 56. H. M. Frost, The skeletal intermediary organization. Metab. Bone Dis. Rel. Res. 4, 281 – 290 (1983). 57. H. M. Frost, The mechanostat: A proposed pathogenic mechanism of osteoporoses and the bone mass effects of mechanical and nonmechanical agents. Bone Miner. 2, 73 – 85 (1987). 58. B. L. Drinkwater, K. Nilson, C. H. Chesnut, W. J. Brenner, S. Shainholtz, and M. B. Southworth, Bone mineral content of amenorrheic and eumenorrheic athletes. N. Engl. J. Med. 311, 277 – 281 (1984). 59. R. Marcus, C. Cann, P. Madvig, J. Minkoff, M. Goddard, M. Bayer, M. Martin, L. Gaudiani, W. Haskell, and H. Genant, Menstrual function and bone mass in elite women distance runners. Ann. Intern. Med. 102, 158 – 163 (1985). 60. C. E. Cann, M. C. Martin, H. K. Genant, and R. B. Jaffe, Decreased spinal mineral content in amenorrheic women. J. Am. Med. Assoc. 251, 626 – 629 (1984). 61. B. M. K. Biller, H. B. A. Baum, D. I. Rosenthal, V. C. Saxe, P. M. Charpie, and A. Klibanski, Progressive trabecular osteopenia in women with hyperprolactinemic amenorrhea. J. Clin. Endocrinol. Metab. 75, 692 – 697 (1992). 62. H. Rico, F. Arnanz, M. Revilla, S. Perera, M. Iritia, L. F. Villa, and I. Arribas, Total and regional bone mineral content in women treated with GnRH agonists. Calcif. Tissue Int. 52, 354 – 357 (1993). 63. A. Klibanski and S. L Greenspan, Increase in bone mass after treatment of hyperprolactinemic amenorrhea. N. Engl. J. Med. 315, 542 – 546 (1986). 64. Y. M. Dawood, V. Lewis, and J. Ramos, Cortical and trabecular bone mineral content in women with endometriosis: Effect of gonadotropin-releasing hormone agonist and danazol. Fertil. Steril. 52, 21 – 27 (1989). 65. B. L. Drinkwater, K. Nilson, and S. Ott, Bone mineral density after resumption of menses in amenorrheic athletes. J. Am. Med. Assoc. 256, 380 – 382 (1986). 66. S. M. Garn, “The Early Gain and Later Loss of Cortical Bone. Thomas, Springfield, IL, 1970. 67. V. Gilsanz, D. T. Gibbens, T. F. Roe, M. Carlson, M. O. Senac, M. I. Boechart, H. K. Huang, E. E. Schulz, C. R. Libanati, and C. C. Cann,
CHAPTER 37 Animal Models for in Vivo Experimentation
68. 69. 70.
71.
72.
73.
74.
75.
76.
77.
78.
79. 80.
81.
82.
83.
84. 85. 86. 87.
88.
89.
Vertebral bone density in children: Effect of puberty. Radiology 166, 847 – 850 (1988). A. E. Treloar, Menstrual cyclicity and the pre-menopause. Maturitas 3, 249 – 264 (1981). B. McMahon and B. Worcester, “Age at Menopause,” Series 11:19. National Center for Health Statistics, Washington, DC, 1966. D. R. Mann, K. G. Gould, and D. C. Collins, A potential primate model for bone loss resulting from medical oophorectomy or menopause. J. Clin. Endocrinol. Metab. 71, 105 – 110 (1990). R. B. Martin, R. L. Butcher, L. L. Sherwood, P. Buckendahl, R. D. Boyd, D. Farris, H. Sharkey, and G. Dannucci, Effects of ovariectomy in beagle dogs. Bone 8, 23 – 31 (1987). C. E. Cann, H. K. Genant, B. Ettinger, and G. S. Gordan, Spinal mineral loss in oophorectomized women. J. Am. Med. Assoc. 244, 2056 – 2059 (1980). D. Hartwell, B. J. Riis, and C. Christiansen, Changes in vitamin D metabolism during natural and medical menopause. J. Clin. Endocrinol. Metab. 71, 127 – 132 (1990). B. E. C. Nordin, A. Horsman, R. Brook, and D. A. Williams, The relationship between oestrogen status and bone loss in postmenopausal women. Clin. Endocrinol. 5, 353S – 361S (1976). O. Lindquist, C. Bengtsson, T. Hansson, and R. Jonsson, Changes in bone mineral content of the axial skeleton in relation to aging and the menopause. Scand. J. Clin. Lab. Invest. 43, 333 – 338 (1983). J. C. Gallagher, D. Goldgar, and A. Moy, Total bone calcium in normal women: Effect of age and menopause status. J. Bone Miner. Res. 2, 491 – 496 (1987). L. R. Hedlund and J. C. Gallagher, The effect of age and menopause on bone mineral density of the proximal femur. J. Bone Miner. Res. 4, 639 – 642 (1989). R. Wasnich, K. Yano, and J. Vogel, Postmenopausal bone loss at multiple skeletal sites: Relationship to estrogen use. J. Chron. Dis. 36, 781 – 790 (1983). R. Lindsay, D. M. Hart, and A. MacLean, Bone response to termination of estrogen treatment. Lancet 1, 1325 – 1327 (1978). A. Horsman, M. Simpson, P. A. Kirby, and B. E. C. Nordin, Non-linear bone loss in oophorectomized women. Br. J. Radio 50, 504 – 507 (1977). L. Nilas, J. Borg, and C. Christiansen, Different rates of loss of trabecular and cortical bone after the menopause. In “Osteoporosis” (C. Christiansen, C. D. Arnaud, B. E. C. Nordin, A. M. Parfitt, W. A. Peck, and B. L. Riggs, Eds.), pp. 161 – 163. Glostrup Hospital, Copenhagen, 1984. N. M. Keshawarz and R. R. Recker, Expansion of the medullary cavity at the expense of cortex in postmenopausal osteoporosis. Metab. Bone Dis. Relat. Res. 5, 223 – 228 (1984). L. S. Richelson, H. W. Wahner, L. J. Melton, and B. L. Riggs, Relative contributions of aging and estrogen deficiency to postmenopausal bone loss. N. Engl. J. Med. 311, 1273 – 1275 (1984). J. Dequeker and P. Geusens, Contributions of aging and estrogen deficiency to postmenopausal bone loss. N. Engl. J. Med. 313, 453 (1985). L. Nilas and C. Christiansen, Bone mass and its relationship to age and the menopause. J. Clin. Endocrinol. Metab. 65, 697 – 702. (1987). R. R. Recker, R. P. Heaney, and P. D. Saville, Menopausal changes in remodeling. J. Lab. Clin. Med. 92, 964 – 971 (1978). J. J. Stepan, J. Pospichal, J. Presl, and V. Pacovsky, Bone loss and biochemical indices of bone remodeling in surgically-induced postmenopausal women. Bone 8, 279 – 284 (1987). R. P. Heaney, R. R. Recker, and P. D. Saville, Menopausal changes in calcium balance performance. J. Lab. Clin. Med. 92, 953 – 963 (1978). T. Coble, D. B. Kimmel, J. M. Lappe, and R. R. Recker, Association of bone markers with cancellous bone behavior in normal premanopausal women. J. Bone Miner. Res. 9 (Suppl. 1), S260[B122] (1994).
41 90. T. Steiniche, C. Hasling, P. Charles, E. F. Eriksen, L. Mosekilde, and F. Melson, A randomized study on the effects of estrogen/gestagen or high dose oral calcium on trabecular bone remodeling in postmenopausal osteoporosis. Bone 10, 313 – 320 (1989). 91. Deleted in proof. 92. C. Christiansen, M. S. Christensen, P. McNair, C. Hagen, K. E. Stocklund, and I. Transbol, Prevention of early postmenopausal bone loss: Controlled 2-year study in 315 normal females. Eur. J. Clin. Invest. 10, 273 – 279 (1980). 93. S. Meema and H. E. Meema, Menopausal bone loss and estrogen replacement. Isr. J. Med. Sci. 12, 601 – 606 (1976). 94. B. Ettinger, H. K. Genant, and C. E. Cann, Long-term estrogen replacement therapy prevents bone loss and fractures. Ann. Int. Med. 102, 319 – 324 (1985). 95. R. Lindsay, D. M. Hart, and J. M. Aitken, Long-term prevention of osteoporosis by estrogen. Lancet 1, 1038 – 1041 (1976). 96. Deleted in proof. 97. C. Christiansen and R. Lindsay, Estrogens, bone loss and preservation. Osteoporosis Int. 1, 7 – 13 (1990). 98. R. T. Turner, J. J. Vandersteenhover, and N. H. Bell, The effects of ovariectomy and 17-b-estradiol on cortical bone histomorphometry in growing rats. J. Bone Miner. Res. 2, 115 – 122 (1987). 99. J. A. Kanis and E. V. McCloskey, Epidemiology of vertebral osteoporosis. Bone 13, S1 – S10 (1992). 100. S. R. Cummings, J. L. Kelsey, M. C. Nevitt, and K. J. O’Dowd, Epidemiology of osteoporosis and osteoporotic fractures. Epidemiol. Rev. 7, 178 – 208 (1985). 101. G. F. Jensen, C. Christiansen, J. Boesen, V. Hegedus, and I. Transbol, Epidemiology of postmenopausal spinal and long bone fractures. Clin. Orthop. Relat. Res. 166, 75 – 81 (1982). 102. C. C. Johnston, C. W. Slemenda, and L. J. Melton, Clinical use of bone densitometry. N. Engl. J. Med. 324, 1105 – 1109 (1991). 103. J. A. Kanis, L. J. Melton, C. Christiansen, C. C. Johnston, and N. Khalttaev, The diagnosis of osteoporosis. J. Bone Miner. Res. 9, 1137 – 1141 (1994). 104. B. E. Nilsson and N. E. Westlin, Bone mineral content and fragility fractures. Clin. Orthop. Rel. Res. 125, 196 – 199 (1977). 105. J. F. Aloia, A. Vaswani, D. McGowan, and P. Ross, Preferential osteopenia in women with osteoporotic fractures. Bone Miner. 18, 51 – 63 (1992). 106. I. Eventon, B. Frisch, Z. Cohen, and I. Hammel, Osteopenia, hematopoiesis, and bone remodeling in iliac crest and femoral biopsies: A prospective study of 102 cases of femoral neck fractures. Bone 11, 1 – 6 (1991). 107. M. Matsushita, T. Tsyboyama, R. Kasai, H. Okumura, T. Yamamuro, K. Higuchi, K. Higuchi, A. Coin, T. Yonezu, A. Utani, M. Umezawa, and T. Takeda, Age-related changes in bone mass in the senescenceaccelerated mouse (SAM). Am. J. Pathol. 125, 276 – 283 (1986). 108. O. Schaadt and H. Bohr, different trends of age-related diminution of bone mineral content in the lumbar spine, femoral neck, and femoral shaft in women. Calcif. Tissue Int. 42, 71 – 76 (1988). 109. B. L. Riggs and L. J. Melton III, Heterogeneity of involutional osteoporosis: Evidence for two osteoporosis syndromes. Am. J. Med. 75, 899 – 901 (1983). 110. C. L. Donaldson, S. B. Hulley, J. M. Vogel, R. S. Hattner, J. H. Bayers and D. E. McMillan, Effect of prolonged bedrest on bone mineral. Metabolism 19, 1071 – 1084 (1970). 111. P. Minaire, P. J. Meunier, C. Edouard, J. Bernard, P. Courpron, and J. Bourret, Quantitative histologic data on disuse osteoporosis. Calcif. Tissue Res. 17, 57 – 73 (1974). 112. B. Krolner and B. Toft, Vertebral bone loss: An unheeded side effect of therapeutic bed rest. Clin. Sci. 64, 537 – 540 (1983). 113. W. Bloom, M. A. Bloom, and F. C. McLean, Calcification and ossification: Medullary bone changes in the reproductive cycle of female pigeons. Anat. Rec. 79, 443 – 466 (1941).
42 114. S. C. Miller and B. M. Bowman, Medullary bone osteogenesis following estrogen administration to mature male Japanese quail. Dev. Biol. 87, 52 – 63 (1981). 115. C. T. Rubin and L. E. Lanyon, Regulation of bone formation by applied dynamic loads. J. Bone Jt. Surg. 66A, 397 – 402 (1984). 116. C. T. Rubin and L. E. Lanyon, Regulation of bone mass by mechanical strain magnitude. Calcif. Tissue Int. 37, 411 – 417 (1985). 117. J. J. Matsuda, R. K. Zernicke, A. C. Vailas, V. A. Pedvini,A. PedriniMille, and J. A. Maynard, Structural and mechanical adaptation of immature bone to strenuous exercise. J. Appl. Physiol. 60, 2028–2034 (1986). 118. L. E. Lanyon, C. T. Rubin, and G. Baust, Modulation of bone loss during calcium insufficiency by controlled dynamic loading. Calcif. Tissue Int. 38, 209 – 216 (1986). 119. A. J. Kahn and D. J. Simmons, Investigation of cell lineage in bone using a chimera of chick and quail embyronic tissue. Nature 258, 325 – 326 (1975). 120. F. V. Jotereau and N. M. Ledouarin, The developmental relationship between osteocytes and osteoclasts: A study using the quail-chick nuclear marker in endochondral ossification. Dev. Biol. 63, 263 – 265 (1978). 121. S. C. Miller, Osteoclast cell-surface changes during the egg-laying cycle in Japanese quail. J. Cell Biol. 75, 104 – 118 (1977). 122. K. M. Attie, N. R. Ramirez, F. A. Conte, S. L. Kaplan, and M. M. Grumbach, The pubertal growth spurt in eight patients with true precocious puberty and growth hormone deficiency: Evidence for a direct role of sex steroids. J. Clin. Endocrinol. Metab. 71, 975 – 983 (1990). 123. V. Gilsanz, D. T. Gibbens, T. F. Roe, M. Carlson, M. O. Senac, M. I. Boechart, H. K. Huang, E. E. Schulz, C. R. Libanati,and C. C. Cann, Vertebral bone density in children: Effect of puberty. Radiology 166, 847 – 850 (1988). 124. J. P. Bonjour, G. Theintz, B. Buchs, D. Slosman, and R. Rizzoli, Critical years and stages of puberty for spinal and femoral bone mass accumulation during adolescence. J. Clin. Endocrinol. Metab. 73, 555 – 563 (1991). 125. D. G. Walker, Congenital osteopetrosis in mice cured by parabiotic union with normal siblings. Endocrinology 91, 916 – 920 (1972). 126. D. G. Walker, Osteopetrosis cured by temporary parabiosis. Science 180, 75 (1973). 127. D. W. H. Barnes, J. F. Loutit, and J. M Sansom, Histocompatible cells for the resolution of osteopetrosis in microophthalmic mice. Proc. R. Soc. London 188, 501 – 505 (1975). 128. D. G. Walker, Bone resorption restored in osteopetrotic mice by transplants of normal bone marrow and spleen cells. Science 190, 784 – 785 (1975). 129. D. G. Walker, Spleen cells transmit osteopetrosis in mice. Science 190, 785 – 786 (1975). 130. J. E. Lennon and H. S. Micklem, Stromal cells in long-term murine bone marrow culture: FACS studies and origin of stromal cells in radiation chimeras. Exp. Hematol. 14, 287 – 292 (1986). 131. S. Kyoizumi, C. M. Baum, H. Kaneshima, J. M. McCune, E. J. Yee, and Namikawa, Implantation and maintenance of functional human bone marrow in SCI-hu mice. Blood 79, 1704, 1711 (1992). 132. N. A. Nicola, C. G. Begley, and D. Metcalf, Identification of the human analogue of a regulator that induces differentiation in murine leukemic cells. Nature 314, 625 – 627 (1985). 133. J. F. Krowka, S. Sarin, R. Namikawa, J. M. McCune, and H. Kaneshima, Human T cells in the SCID-hu mouse are phenotypically normal and functionally competent. J. Immunol. 146, 3751 – 3757 (1991). 134. C. B. Shoemaker and L. D. Mitsock, Murine erythropoietin gene: Cloning, expression, and human gene homology. Mol. Cell Biol. 6, 849 – 856 (1986). 135. J. P. Cassella, R. Pereira, J. S. Khillan, D. J. Prockop, N. Garrington, and S. Y. Ali, An ultrastructural, micoranalytical, andspectroscopic study of bone from a transgenic mouse with aCOLAL1.A1 pro-Alpha-1 mutation. Bone 15, 611 – 619 (1994).
DONALD B. KIMMEL 136. S. C. Marks and P. W. Lane, Osteopetrosis, a new recessive skeletal mutation on chromosome 12 of the mouse. J. Hered. 67, 11 – 18 (1976). 137. R. A. Altschuler, D. F. Dolan, M. Ptok, G. Gholizadeh, J.Bonadio, and J. E. Hawkins, An evaluation of otopathology in the MOV-13 transgenic mutant mouse. Ann. NY Acad. Sci. 630, 249 – 252 (1991). 138. R. Balena, F. Costantini, M. Yamamoto, A. Markatos, R. Cortese, G. A. Rodan, and V. Poli, Mice with IL-6 gene knock-out do not lose cancellous bone after ovariectomy. J. Bone Miner. Res. 8 (Suppl. 1), S130 (1993). 139. E. S. Lander and D. Botstein, Mapping Mendelian factors underlying quantitative traits using RFLP linkage maps. Genetics 121, 185 – 199 (1989). 140. H. K. Suzuki, Effects of estradiol-17-b-n-valerate on endosteal ossification and linear growth in the mouse femur. Endocrinology 60, 743 – 747 (1958). 141. M. W. Edwards, S. D. Bain, M. C. Bailey, M. M. Lantry, and G. A. Howard, 17-b-Estradiol stimulation of endosteal bone formation in the ovariectomized mouse: An animal model for the evaluation of bone-targeted estrogens. Bone 13, 29 – 34 (1992). 142. L. R. Donahue, C. J. Rosen, and W. G. Beamer, Reduced bone density in hypogonadal mice. J. Bone Miner. Res. 9 (Suppl. 1), S193 (1994). 143. J. Hashimoto, K. Takaoka, H. Yoshikawa, S. Miyamoto, S. Suzuki, and K. Ono, Preservation of ectopically-induced bone in the mouse by estradiol. Bone 12, 249 – 255 (1991). 144. B. Bar-Shira-Maymon, R. Coleman, A. Cohen, E. SteinhagenThiessen, and M. Silbermann, Age-related bone loss in lumbar vertebrae of CW-1 female mice: A histomorphometric study. Calcif. Tissue Int. 44, 36 – 45 (1989). 145. B. Bar-Shira-Maymon, R. Coleman, E. Steinhagen-Thiessen, and M. Silbermann, Correlation between alkaline and acid phosphatease activities and age-related osteopenia in murine vertebrae. Calcif. Tissue Int. 44, 99 – 207 (1989). 146. A. Weiss, I. Arbell, E. Steinhagen-Thiessen, and M. Silbermann, Structural changes in aging bone: Osteopenia in the proximal femurs of female mice. Bone 12, 165 – 172 (1991). 147. M. R. Urist, A. M. Budy, and F. C. McLean, Endosteal bone formation in estrogen-treated mice. J. Bone Jt. Surg. 32A, 143 – 162 (1950). 148. S. D. Bain, M. C. Bailey, D. L. Celino, M. M. Lantry, and M. W. Edwards, High dose estrogen inhibits bone resorption and stimulates bone formation in the ovariectomized mouse. J. Bone Miner. Res. 8, 435 – 442 (1993). 149. S. D. Bain, M. C. Bailey, and M. W. Edwards, The anabolic effect of estrogen on endosteal bone formation in the mouse is attenuated by ovariohysterectomy: A role for the uterus in the skeletal response to estrogen? Calcif Tissue Int. 51, 223 – 228 (1992). 150. S. D. Bain, E. Jensen, D. L. Celino, M. C. Bailey, M. M. Lantry, and M. W. Edwards, High-dose gestagens modulate bone resorption and formation and enhance estrogen-induced endosteal bone formation in the ovariectomized mouse. J. Bone Miner. Res. 8, 291 – 230 (1993). 151. Deleted in proof. 152. T. Tsuboyama, K. Takahashi, M. Matsushita, H. Okumura, T. Yamamuro, M. Umezawa, and T. Takeda, Decreased endosteal formation during cortical bone modeling in SAM-P/6 mice with a low peak bone mass. Bone Miner. 7, 1 – 12 (1989). 153. T. Tsuboyama, M. Matsushita, H. Okumura, T. Yamamuro, K. Hanada, and T. Takeda, Modification of strain-specific femoral bone density by bone marrow chimerism in mice: A study on the spontaneously osteoporotic mouse (SAM-P/6). Bone 10, 269 – 277 (1989). 154. K. Takahashi, T. Tsuboyama, M. Matsushita, R. Kasai, H. Okumura, T. Yamamuro, Y. Okamoto, K. Toriyama, K. Kitagawa, and T. Takeda, Modification of strain-specific femoral bone density by
CHAPTER 37 Animal Models for in Vivo Experimentation
155.
156.
157.
158.
159.
160. 161.
162.
163. 164. 165.
166.
167.
168.
169.
170.
171.
172.
173. 174.
bone marrow-derived factors administered neonatally: A study on the spontaneously osteoporotic mouse, SAMP6. Bone Miner. 24, 245 – 255 (1994). T. Tsuboyama, K. Takahashi, T. Tamamuro, M. Hosokawa, and T. Takeda, Cross-mating study on bone mass in the spontaneously osteoporotic mouse (SAM-P/6) Bone Miner. 23, 57 – 64 (1993). K. Takahashi, T. Tsuboyama, M. Matsushita, R. Kasai, H. Okumura, T. Yamamuro, Y. Okamoto, K. Kitagawa, and T. Takeda, Effective intervention of low peak bone mass and bone modeling in the spontaneous murine model of senile osteoporosis, SAM-P/6, by calcium supplement and hormone treatment. Bone 15, 209 – 215 (1994). L. D. Spotila, C. D. Constantinou, L. Sereda, A. Ganguly, B. L. Riggs, and D. J. Prockop, Mutation in a gene for type I procollagen (COL1A2) in a woman with postmenopausal osteoporosis: Evidence for phenotypic and genotypic overlap withmild osteogenesis imperfecta. Proc. Natl. Acad. Sci. USA 88, 5423 – 5427. (1991). M. Ringwald, R. Baldock, J. Bard, M. Kaufman, J. T. Eppig, J. E. Richardson, J. H. Nadeau, and D. Davidson, A databasefor mouse development. Science 265, 2033 – 2034 (1994). J. S. Arnold and W. S. S. Jee, Bone growth and osteoclastic activity as indicated by radioautographic distribution of 239Pu. Am. J. Anat. 101, 367 – 417 (1957). G. Gothlin and J. L. E. Ericsson, On the histogenesis of the cells in fracture callus. Virchows Arch. Abt. B Zellpath. 12, 318 – 329 (1973). H. Andersen and M. E. Matthiessen, The histiocyte in human fetal tissues: Its morphology, cytochemistry, origin, function, and fate. Z. Zellforsch. 72, 193 – 211 (1966). S. C. Marks and G. B. Schneider, Evidence for a relationship between lymphoid cells and osteoclasts: Bone resorption restored in ia (osteopetrotic) rats by lymphocytes, monocytesand macrophages from a normal littermate. Am. J. Anat. 152, 331 – 342 (1978). K. Buring, On the origin of cells in heterotopic bone formation. Clin. Orthop. Relat. Res. 110, 293 – 302 (1973). A. B. Dawson, The age order of epiphyseal union in the long bones of the albino rat. Anat. Rec. 31, 1 – 17 (1925). R. M. Acheson, M. N. MacIntyre, and E. Oldham, Techniques in longitudinal studies of the skeletal development of the rat. Br. J. Nutr. 13, 283 – 296 (1959). E. E. Joss, E. H. Sobel, and K. A. Zuppinger, Skeletal maturation in rats with special reference to order and time of epiphyseal closure. Endocrinology 72, 117 – 122 (1963). C. Spark and A. B. Dawson, The order and appearance of centers of ossification in the fore and hind limbs of the albino rat, with reference to the possible influence of the sex factor. Am. J. Anat. 41, 411 – 445 (1928). H. E. Swanson and J. J. van der Werff Ten Bosch, Sex differences in growth of rats, and their modification by a single injection of testosterone propionate shortly after birth. J. Endocrinol. 26, 197 –207 (1963). R. T. Turner, K. S. Hannon, L. M. Demers, J. Buchanan, and N. H. Bell, Differential effects of gonadal function on bone histomorphometry in male and female rats. J. Bone Miner. Res. 4, 557 – 563 (1989). D. B. Kimmel, Quantitative histologic changes in the proximal tibial epiphyseal growth cartilage of aged female rats. Cells Mater. 1 (Suppl.), 11 – 18 (1992). D. Schapira, R. Laton-Miller, D. Barzilai, and M. Silbermann, The rat as a model for studies of the aging skeleton. Cells Mater. (Suppl.) 1, 181 – 188 (1992). X. J. Li, W. S. S. Jee, H. Z. Ke, S. Mori, and T. Akamine, Age related changes of cancellous and cortical bone histomorphometry in female Sprague – Dawley rats. Cells Mater. (Suppl.) 1, 25 – 37 (1992). H. M. Frost and W. S. S. Jee, On the rat model of human osteopenias and osteoporoses. Bone Miner. 18, 227 – 236 (1992). R. L. Butcher, W. E. Collins, and N. W. Fugo, Plasma concentration of LH, FSH, prolactin, progesterone, and estradiol-17-b throughout the four day cycle of the rat. J. Endocrinol. 94, 1704 – 1708 (1974).
43 175. K. H. Lu, B. R. Hopper, T. M. Vargo, and S. S. C. Yen, Chronological changes in sex steroid, gonadotropin, and prolactin secretion in againg female rats displaying different reproductive states. Biol. Reprod. 21, 193 – 203 (1979). 176. D. N. Kalu, The ovariectomized rat as a model of postmenopausal osteopenia. Bone Miner. 15, 175 – 191 (1991). 177. T. J. Wronski, The ovariectomized rat as an animal model for postmenopausal bone loss. Cells Mater. (Suppl. 1)1, 69 – 74 (1992). 178. T. J. Wronski, L. M. Dann, and S. L. Horner, Time course of vertebral osteopenia in ovariectomized rats. Bone 10, 295 – 301 (1989). 179. D. B. Kimmel and T. J. Wronski, Non-destructive measurement of bone mineral in femurs from ovariectomized rats. Calcif. Tissue Int. 46, 101 – 110 (1990). 180. J. G. Seedor, H. A. Quartuccio, and D. D. Thompson, The bisphosphonate alendronate (MK-217) inhibits bone loss due to ovariectomy in rats. J. Bone Miner. Res. 6, 339 – 346 (1991). 181. Li Mosekilde, C. C. Danielsen, and U. B. Knudsen, The effect of aging and ovariectomy on the vertebral bone mass and biomechanical properties of mature rats. Bone 14, 1 – 6 (1993). 182. T. J. Wronski, C. C. Walsh, and L. A. Ignaszewski, Histologic evidence for osteopenia and increased bone turnover in ovariectomized rats. Bone 7, 119 – 124 (1986). 183. D. Black, C. Farquharson, and S. P. Robins, Excretion of pyridinium cross-links of collagen in ovariectomized rats as urinary markers for increased bone resorption. Calcif. Tissue Int. 44, 343 – 347 (1989). 184. H. Ito, H. Z. Ke, W. S. S. Jee, and T. Sakou, Anabolic responses of an adult cancellous bone site to prostaglandin E2 in the rat. Bone Miner. 21, 219 – 236 (1993). 185. S. R. Cummings, Epidemiologic studies of osteoporotic fractures: Methodologic issues. Calcif. Tissue Int. 49 (Suppl.), S15 – S20 (1991). 186. P. Amman, R. Rizzoli, D. Slosman, and J. P. Bonjour, Sequential and precise in vivo measurement of bone mineral density in rats using dual energy x-ray absorptiometry. J. Bone Miner. Res. 7, 311 – 317 (1992). 187. M. G. Griffin, R. Kimble, W. Hopfer, and R. Pacifici, DXA of the rat: Accuracy, precision, and measurement of bone loss. J. Bone Miner. Res. 8, 795 – 800 (1993). 188. B. H. Mitlak, D. Schoenfeld, and R. M. Neer, Accuracy, precision, and utility of spine and whole-skeleton mineral measurements by DXA in rats. J. Bone Miner. Res. 9, 119 – 126 (1994). 189. D. N. Kalu, C. C. Liu, E. Salerno, B. W. Hollis, R. Echon, and M. Ray, Skeletal response of ovariectomized rats to low and high doses of 17-b-estradiol. Bone Miner. 14, 175 – 187 (1991). 190. T. J. Wronski, C. F. Yen, and K. S. Scott, Estrogen and diphosphonate treatment provide long-term protection against osteopenia in ovariectomized rats. J. Bone Miner. Res. 6, 387 – 394 (1991). 191. P. T. Beall, L. K. Misra, R. L. Young, H. J. Spjut, H. J Evans, and A. LeBlanc, Clomiphene protects against osteoporosis in the mature ovariectomized rat. Calcif. Tissue Int. 36, 123 – 125 (1984). 192. S. C. Garner, J. J. B. Anderson, M. H. Mar, and I. Parikh, Estrogens reduce bone loss in the ovariectomized, lactating rat model. Bone Miner. 15, 19 – 31 (1992). 193. R. Schenk, W. A. Merz, H. A. Fleisch, R. C. Muhlbauer, and R. G. G. Russell, Effects of ethane-1-hydroxy-1,1-diphosphonate (EHDP) and dichloromethylene diphosphonate (C12MDP) on the calcification and resorption of cartilage and bone in thetibial epiphysis and metaphysis of rats. Calcif. Tissue Res. 11, 196 – 214 (1973). 194. R. Schenk, P. Eggli, H. Fleisch, and S. Rosini, Quantitative morphometric evaluation of the inhibitory activity of new aminobisphosphonates on bone resorption in the rat. Calcif. Tissue Int. 38, 342 – 349 (1986). 195. K. Nakatsuka, Y. Nishizawa, S. Hagiware, H. Koyama, T. Miki, H. Ochi, and H. Morii, Effect of calcitonin on total body bone mineral contents of experimental osteoporotic rats determined by dual photon absorptiometry. Calcif. Tissue Int. 47, 378 – 382 (1990).
44 196. G. F., Mazzuoli, S. Tabolli, F. Bigi, C. Valtorta, S. Minisola, D. Diacinti, L. Scarnecchia, G. Bianchi, M. Piolini, and S. Dell’Acqua, Effects of salmon clacitonin on the bone loss induced by ovariectomy. Calcif. Tissue Int. 47, 209 – 214 (1990). 197. S. T. Harris, B. J. Gertz, H. K. Genant, D. R. Eyre, T. T. Survill, J. N. Ventura, J. DeBrock, E. Ricerca, and C. H. Chesnut, The effect of short term treatment with alendronate on vertebral density and biochemical markers of bone remodeling in early postmenopausal women. J. Clin. Endocr. Metab. 76, 1399 – 1406 (1993). 198. N. B. Watts, S. T. Harris, H. K. Genant, R. D. Wasnich, P. D. Miller, R. D. Jackson, A. A. Licata, P. Ross, G. C. Woodson, M. J. Yanover, W. J. Mysiw, L. Kohse, M. B. Rao, P. Steiger, B. Richmond, and C. H. Chesnut, Intermittent cyclical etidronate treatment of postmenopausal osteoporosis. N. Engl. J. Med. 323, 73 – 79 (1990). 199. G. A. Fromm, E. Vega, L. Plantalech, A. M. Galich, and C. A. Mautalen, Differential action of pamidronate on trabecular and cortical bone in women with involutional osteoporosis. Calcif Tissue Int. 49, 129 – 133 (1991). 200. M. L. Smith, I. Fogelman, D. M. Hart, E. Scott, J. Bevan, and I. Leggate, Effect of etidronate disodium on bone turnover following surgical menopause. Calcif. Tissue Int. 44, 74 – 79 (1989). 201. J. Y. Reginster, R. Deroisy, D. Denis, J. Collette, M. P. Lecart, N. Sarlet, D. Ethgen, and P. Franchimont, Prevention of postmenopausal bone loss by tiludronate. Lancet 2, 1469 – 1470 (1989). 202. I. MacIntyre, M. I. Whitehead, L. M. Banks, J. C. Stevenson, S. J. Wimalawansa, and M. J. R. Healy, Calcitonin for prevention of postmenopausal bone loss. Lancet 1, 900 – 901 (1988). 203. J. Szucs, C. Horvath, E. Kollin, M. Szathmari, and I. Hollo, Threeyear calcitonin combination therapy for postmenopausal osteoporosis with crush fractures of the spine. Calcif. Tissue Int. 50, 7 – 10 (1992). 204. J. M. Lean, J. W. M. Chow, and T. J. Chambers, Estrogen induces bone formation on non-resorptive surfaces in the rat. Bone 14, 297 – 302 (1993). 205. J. W. M. Chow, J. M. Lean, and T. J. Chambers, 17-b-Estradiol stimulates cancellous bone formation in female rats. Endocrinology 130, 3025 – 3032 (1992). 206. R. T. Turner, B. L. Riggs, and T. C. Spelsberg, Skeletal effects of estrogen. Endocrinol. Rev. 15, 275 – 300 (1994). 207. J. L. Ferretti, R. D. Tessaro, C. J. Delgado, C. E. Bozzini, R. M. Alippi, and A. C. Barelo, Biomechanical performance of diaphyseal shafts and bone tissue of femurs from protein-restricted rats. Bone Miner. 4, 329 – 339 (1988). 208. C. H. Sogaard, C. C. Danielsen, E. B. Thorling, and Li Mosekilde, Long-term exercise of young and adult female rats: Effect on femoral neck biomechanical competence and bone structure. J. Bone Miner. Res. 9, 409 – 416 (1994). 209. Z. Peng, J. Tuukkanen, and H. K. Vaananen, Exercise can provide protection against bone loss and prevent the decrease in mechanical strength of femoral neck in ovariectomized rats. J. Bone Miner. Res. 9, 1559 – 1564 (1994). 210. T. Okano, M. P. Akhter, T. Howard, and D. B. Kimmel, Estrogen effects on bone mass, bone strength, and bone turnover in aged ovariectomized rats. J. Bone Miner. Res. 10 (Suppl. 1), S456, (T407) (1995). 211. Deleted in proof. 212. Deleted in proof. 213. Deleted in proof. 214. L. E. Nachtigall, R. H. Nachtigall, and R. D. Nachtigall, Estrogen replacement therapy. I. A 10-year prospective study in the relationship to osteoporosis. Obstet. Gynecol. 53, 277 – 281 (1979). 215. A. Paganini-Hill, R. K. Ross, V. R. Gerkins, B. E. Henderson, M. Arthur, and T. M. Mack, Menopausal estrogen therapy and hip fractures. Ann. Intern. Med. 95, 28 – 31 (1981). 216. J. A. Svesatikoglou and S E. Larsson, Changes in composition and metabolic activity of the skeletal parts of the extremity of the adult
DONALD B. KIMMEL
217.
218.
219.
220. 221. 222.
223. 224.
225.
226.
227. 228.
229.
230. 231.
232. 233.
234.
235.
236.
237.
238.
rat following below-knee amputation. Acta Chir. Scand. (Suppl. 476), 9 – 15 (1976). V. T. Thomaidis and T. S. Lindholm, The effect of remobilization on the extremity of the adult rat after short-term immobilization in a plaster cast. Acta Chir. Scand. (Suppl. 476), 36 – 39 (1976). D. D. Thompson and G. A. Rodan, Indomethacin inhibition of tenotomy-induced bone resorption in rats. J. Bone Miner. Res. 3, 409 – 414 (1988). X. J. Li, W. S. S. Jee, S. Y. Chow, and D. M. Woodbury, Adaptation of cancellous bone to aging and immobilization in the adult rat: A single photon absorptiometry and histomorphometry study. Anat. Rec. 227, 12 – 24 (1990). J. U. Lindgren and S. Mattson, The reversibility of disuse osteoporosis. Calcif. Tissue Res. 23, 179 – 184 (1977). H. Maeda, D. B. Kimmel, N. Lane, and D. Raab, The musculoskeletal response to immobilization and recovery. Bone 14, 153 – 159 (1993). N. Lane, H. Maeda, D. M. Cullen, and D. B. Kimmel, Cancellous bone behavior in hindlimb immobilized rats during and after naproxen treatment. Bone Miner. 26, 43 – 59. (1994). S. Mattson, The reversibility of disuse osteoporosis. Acta Orthop. Scand. (Suppl. 144), 59 – 88 (1972). R. H. Young and W. A. J. Crane, Effect of hydrocortisone on the utilization of tritiated thymidine for skeletal growth in the rat. Ann. Rheum. Dis. 23, 163 – 167 (1964). D. J. Simmons and A. S. Kunin, Autoradiographic and biochemical investigations of the effect of cortisone on the bones of the rat. Clin. Orthop. Rel. Res. 55, 201 – 215 (1967). J. L. Ferretti, S. O. Vazquez, C. J. Delgado, R. Capozza, and G. Cointry, Biphasic dose-response curves of cortisol effects on rat diaphyseal bone biomechanics. Calcif. Tissue Int. 50, 49 – 54 (1992). A. Vignery and R. Baron, Dynamic histomorphometry of alveolar bone remodeling in the adult rat. Anat. Rec. 196, 191 – 200 (1980). R. Baron, R. Tross, and A. Vignery, Evidence of sequential modeling in rat trabecular bone: Morphology, dynamic histomorphometry, and changes during skeletal maturation. Anat. Rec. 208, 137 – 145 (1984). W. S. S. Jee, S. Mori, X. J. Li, and S. Chan, Prostaglandin E2 enhances cortical bone mass and activates intracortical bone remodeling in intact and ovariectomized female rats. Bone 11, 253 – 266 (1990). E. Ruth, An experimental study of the Haversian-type vascular channels. Anat. Rec. 112, 429 – 455 (1953). D. Baylink, M. Stauffer, J. Wergedal, and C. Rich, Formation, mineralization, and resorption of bone in vitamin D-deficient rats. J. Clin. Invest. 49, 1122 – 1134 (1970). E. R. Morey and D. J. Baylink, Inhibition of bone formation during spaceflight. Science 201, 1138 – 1141 (1978). T. J. Wronski, P. A. Schenck, M. Cintron, and C. C. Walsh, Effect of body weight on osteopenia in ovariectomized rats. Calcif. Tissue Int. 43, 179 – 183 (1987). R. E. Roudebush, D. E. Magee, D. N. Benslay, A. M. Bendele, and H. U. Bryant, Effect of weight manipulation on bone loss due to ovariectomy and the protective effects of estrogen in the rat. Calcif. Tissue Int. 53, 61 – 64 (1993). F. R. deWinter and R. Steendijk, The effect of a low-calcium diet in lactating rats: Observations on the rapid development and repair of osteoporosis. Calcif. Tissue Res. 17, 303 – 316 (1975). S. C. Miller, J. G. Shupe, E. H. Redd, M. A. Miller, and T. H Omura, Changes in bone mineral and bone formation rates during pregnancy and lactation in rats. Bone 7, 283 – 287 (1986). N. Binkley and D. Kimmel, Effect of age and parity on skeletal response to ovariectomy in rats. J. Bone Miner. Res. 9 (Suppl. 1), S197 (1994). H. Duncan, C. A Hanson, and A. Curtiss, The different effects of soluble and crystalline hydrocortisone on bone. Calcif. Tissue Res. 12, 159 – 168 (1973).
CHAPTER 37 Animal Models for in Vivo Experimentation 239. D. Vanderschueren, E. Van Herck, A. M. Suiker, K. Allewaert, W. J. Visser, P. Geusens, and R. Bouillon, Bone and mineral metabolism in the adult guinea pig: Long-term effects of estrogen and androgen deficiency. J. Bone Miner. Res. 7, 1407 – 1415 (1992). 240. J. Jowsey and J. Gershon-Cohen, Effect of dietary calcium levels on production and reversal of experimental osteoporosis in cats. Proc. Soc. Exp. Biol. Med. 116, 437 – 441 (1964). 241. S. Jett, K. Wu, H. Duncan, and H. M. Frost, Adrenalcorticosteroid and salicylate actions on human and canine Haversian bone formation and resorption. Clin. Orthop. Relat. Res. 68, 310 – 315 (1970). 242. M. S. Mackey, M. L. Stevens, X. J. Li, D. L Tressler, K. S. Combs, and J. E. McOsker, The ferret: A small animal model with BMUbased remodeling for skeletal research. J. Bone Miner. Res. 9 (Suppl. 1), S319 (1994). 243. X. J. Li, M. S. Mackey, M. L. Stevens, D. L. Tressler, K. S. Combs, and J. E. McOsker, The ferret: Skeletal response to estrogen deficiency. J. Bone Miner. Res. 9 (Suppl.1), S397 (1994). 244. X. J. Li, M. L. Stevens, M. S. Mackey, K. S. Combs, D. L. Tressler, and J. E. McOsker, The ferret: Skeletal responses to treatment with PTH. J. Bone Miner. Res. 9 (Suppl. 1), S258 (1994). 245. A. M. Parfitt, The actions of PTH on bone — Relation to bone remodeling and turnover, calcium homeostasis, and metabolic bone disease, PTH and osteoblast relationships between bone turnover and bone loss, and the state of bones in primary hyperparathyroidism. Metabolism 25, 1033 – 1069 (1976). 246. W. S. S. Jee and P. D. Nolan, Origin of osteoclasts from the fusion of phagocytes. Nature 200, 225 – 226 (1963). 247. J. K. Gong, J. S. Arnold, and S. H. Cohn, Composition of trabecular and cortical bone. Anat. Rec. 149, 325 – 331 (1964). 248. L. C. Johnson, Composition of trabecular and cortical bone in humans. In “Bone Biodynamics” (H. M. Frost, Ed.), pp. 543 – 654. Little, Brown, Boston, 1964. 249. N. J. Parks, W. S. S. Jee, R. B. Dell, and G. E. Miller, Assessment of cortical and trabecular bone distribution in the beagle skeleton by neutron activation analysis. Anat. Rec. 215, 230 – 250 (1986). 250. H. M. Frost, Tetracycline-based histological analysis of bone remodeling. Calcif. Tissue Res. 3, 211 – 239 (1969). 251. D. B. Kimmel and W. S. S. Jee, A quantitative histologic study of bone turnover in young adult beagles. Anat. Rec. 203, 35 – 51 (1982). 252. C. Bressot, P. J. Meunier, M. C. Chapuy, E. Lejeune, C. Edouard, and A. J. Darby, Histomorphometric profile, pathophysiology and reversibility of corticosteroid-induced osteoporosis. Metab. Bone Dis. Relat. Res. 1, 303 – 311 (1979). 253. E. Ritz, B. Krempien, O. Mehls, and H. H. Malluche, Skeletal abnormalities in chronic renal insufficiency before and during maintenance hemodialysis. Kidney Int. 4, 116 – 127 (1973). 254. H. H. Malluche, M. C. Faugere, R. M. Friedler, C. Matthews, and P. Fanti, Calcitriol, parathyroid hormone, and accumulation of aluminum in dogs with renal failure. J. Clin. Inv. 79, 754 – 761 (1987). 255. P. D. Delmas, S. Charhon, M. C. Chapuy, E. Vignon, D. Briancon, C. Edouard, and P. J. Meunier, Long-term effect of dichloromethylene disphosphonate on skeletal lesions in multiple myeloma. Metab. Bone Dis. Relat. Res. 4, 163 – 168 (1982). 256. L. Flora, G. S. Hassing, A. M. Parfitt, and A. R. Villanueva, Comparative skeletal effects of two diphosphonates in dogs. In “Bone Histomorphometry III” (W. S. S. Jee and A. M. Partiff, Eds.), pp. 389 – 407. Armour Montagu, Paris, 1981. 257. H. K. Uhthoff, G. Sekaly, and Z. F. G. Jaworski, Effect of long-term nontraumatic immobilization on metaphyseal spongiosa in young adult and old beagle dogs. Clin. Orthop. Relat. Res. 192, 279 – 283 (1985). 258. H. K. Uhthoff and Z. F. G. Jaworski, Bone loss in response to longterm immobilisation. J. Bone Jt. Surg. 60B, 420 – 429 (1978). 259. Z. F. G. Jaworski, M. Liskova-Kiar, and H. K Uhthoff, Effect of long-term immobilisation on the pattern of bone loss in older dogs. J. Bone Jt. Surg. 62B, 104 – 110 (1980).
45 260. Z. F. G. Jaworski and H. K. Uhthoff, Reversibility of nontraumatic disuse osteoporosis during its active phase. Bone 7, 431 – 439 (1986). 261. R. D. Podbesek, C. Edouard, P. J. Meunier, J. A. Parsons, J. Reeve, R. W. Stevenson, and J. M Zanelli, Effect of two treatment regimes with synthetic human parathyroid hormone fragment on bone formation and the tissue balance of trabecular bone in greyhounds. Endocrinology 112, 1000 – 1006 (1983). 262. D. B. Kimmel, The oophorectomized beagle as an experimental model for estrogen-depletion bone loss in the adult human. Cells Mater. (Suppl.) 1, 75 – 84 (1992). 263. R. G. Cumming, Calcium intake and bone mass: A quantitative review of the evidence. Calcif. Tissue. Int. 47, 194 – 201 (1990). 264. M. K. Drezner and T. Nesbitt, Role of calcitriol in prevention of osteoporosis: Part I. Metabolism 39 (Suppl. 1), 18 – 23 (1990). 265. A. Barbier, J. P. Bonjour, P. Geusens, M. C. deVernejoul, and F. Lacheretz, Tiludronate: A bisphosphonate with a positive effect on bone quality in experimental models. J. Bone Min. Res. 6 (Suppl. 1), S217 (1991). 266. J. A. Shaw, S. C. Wilson, E. M. Paul, and A. Bruno, Vertebral trabecular bone volume in dogs: Comparison of oophorectomized females, sham females, and males. Trans. Orthop. Res. Soc. 17(135) (1992). 267. W. R. Lynch, S. A. Goldstein, A. M. Schulick, and C. A. Mase, The effects of calcitonin on torsional properties of the long bones of ovariohysterectomized beagles. Trans. 37th Annu. Meeting Orthop. Res. Soc. 16, 430 (1991). [Abstract] 268. D. A. McCubbrey, R. W. Goulet, S. A. Goldstein, and A. M. Shulick, Vertebral body architecture and mechancial properties alterations due to ovariectomy in the beagle. Trans. 37th Annu. Meeting Orthop. Res. Soc. 16, 284 (1991). [Abstract] 269. R. W. Boyce, A. F. Franks, M. L. Jankowsky, C. M. Orcutt, A. M. Piacquadio, J. M. White, and J. M. Bevan, Sequential histomorphometric changes in cancellous bone from ovariohysterectomized dogs. J. Bone Miner. Res. 5, 947 – 953 (1990). 270. H. H. Malluche, M. C. Faugere, M. Rush, and R. M. Friedler, Osteoblastic insufficiency is responsible for maintenance of osteopenia after loss of ovarian function in experimental beagle dogs. Endocrinology 119, 2649 – 2655 (1986). 271. M. C. Faugere, P. Fanti, R. M. Friedler, and H. H. Malluche, Bone changes occurring early after cessation of ovarian function in beagle dogs: A histomorphometric study employing sequential biopsies. J. Bone Miner. Res. 5, 263 – 272 (1990). 272. H. H. Malluche, M. C. Faugere, R. M. Friedler, and P. Fanti, 1,25Dihydroxyvitamin D3 corrects bone loss but suppresses bone remodeling in ovariohysterectomized beagle dogs. Endocrinology 122, 1988 – 2006 (1988). 273. T. Nakamura, Y. Nagai, H. Yamato, K. Suzuki, and H. Orimo, Regulation of bone turnover and prevention of bone atrophyin ovariectomized beagle dogs by the adminsitration of 24R,25(OH)2D3. Calcif. Tissue Int. 50, 221 – 227 (1992). 274. G. A. Dannucci, R. B. Martin, and P. Patterson-Buckendahl, Ovariectomy and trabecular bone remodeling in the dog. Calcif. Tissue Int. 40, 194 – 199 (1987). 275. K. K. Karambolova, G. R. Snow, and C. Anderson, Differences in periosteal and corticoendosteal bone envelope activities in spayed and intact beagles: A histomorphometric study. Calcif. Tissue Int. 41, 665 – 668 (1985). 276. V. Shen, D. W. Dempster, R. Birchman, R. W. E. Mellish, E. Church, D. Kohn, and R. Lindsay, Lack of changes in histomorphometric, bone mass, and biochemical parameters in ovariohysterectomized dogs. Bone 13, 311 – 316 (1992). 277. K. K. Karambolova, G. R. Snow, and C. Anderson, Surface activity on the periosteal and corticoendosteal envelopes following continuous progestogen supplementation in spayed beagles. Calcif. Tissue Int. 38, 239 – 243 (1986).
46 278. K. K. Karambolova, G. R. Snow, and C. Anderson, Effects of continuous 17-b-estradiol administration on the periosteal and corticoendosteal envelope activity in spayed beagles. Calcif. Tissue Int. 40, 12 – 15 (1987). 279. G. R. Snow and C. Anderson, The effects of 17-b-estradiol and progestagen on trabecular bone remodeling in oophorectomized dogs. Calcif. Tissue. Int. 39, 198 – 205 (1986). 280. G. R. Snow and C. Anderson, The effects of continuous estradiol therapy on cortical bone remodeling activity in the spayed beagle. Calcif. Tissue Int. 37, 437 – 440 (1985). 281. G. R. Snow, M. A. Cook, and C. Anderson, Oophorectomy and cortical bone remodeling in the beagle. Calcif. Tissue Int. 36, 586 – 590 (1984). 282. G. R. Snow and C. Anderson, The effects of continuous progestogen treatment on cortical bone remodeling activity in beagles. Calcif. Tissue Int. 37, 282 – 286 (1985). 283. T. Koyama, H. Uno, A. Nagaami, T. Makita, T. Konno, and H. Takahashi, Effects of active vitamin D3 derivatives on bone remodeling in ovariectomized beagle dogs — Two year study — II Histomorphometric analysis of trabecular bone. Jpn. J. Bone Miner. 2, 268 – 275 (1984). 284. T. Koyama, A. Nagaami, T. Makita, T. Konno, H. Akazawa, and H. Takahashi, Effect of 1-,24(R)-dihydroxyvitamin D3 on trabecular bone remodeling in adult beagle dogs. Jpn. J. Bone Miner. 2, 249 – 257 (1984). 285. C. C. Shipp, P. S. Berger, M. S. Deehr, and B. Dawson-Hughes, Precision of dual photon absorptiometry. Calcif. Tissue Int. 42, 287 – 282 (1988). 286. P. W. Concannon, W. Hansel, and W. J. Visek, The ovarian cycle of the bitch: Plasma estrogen, LH and progesterone. Biol. Reprod. 13, 112 – 121 (1975). 287. G. E. Jones, A. R. Boyns, E. H. D. Cameron, E. T. Bell,D. W. Christie, and M. F. Parkes, Plasma oestradiol, luteinizing hormone and progesterone during the oestrous cycle in the beagle bitch. J. Endocrinol. 57, 331 – 332 (1973). 288. E. T. Bell, D. W. Christie, and E. V. Younglai, Plasma oestrogen levels during the canine oestrous cycle. J. Endocrinol. 51, 225 – 226 (1971). 289. D. T. Baird and A. Guevara, Concentration of unconjugated estrone and estradiol in peripheral plasma in nonpregnant women throughout the menstrual cycle, castrate and postmenopausal women and in men. J. Clin. Endocrinol. 29, 149 – 156 (1969). 290. M. J. Reed, P. A. Beranek, and V. H. T. James, Oestrogen production and metabolism in perimenopausal women. Maturitas 8, 29 – 34 (1986). 291. C. Longcope, L. Hoberg, S. Steuterman, and D. Baran, The effect of ovariectomy on spine bone mineral density in rhesus monkeys. Bone 10, 341 – 344 (1989). 292. Li Mosekilde, S. E. Weisbrode, J. A. Safron, H. F. Stills, M. L. Jankowsky, D. C. Ebert, C. C. Danielsen, C. H. Sogaard, A. F. Franks, M. L. Stevens, C. L. Paddock, and R. W. Boyce, Calcium-restricted ovariectomized Sinclair S-1 minipigs: An animal model of osteopenia and trabecular plate perforation. Bone 14, 379 – 382 (1993). 293. M. L. Stevens, N. Sacco-Gibson, K. S. Combs, C. Lowry, and X. J. Li, Evaluation of skeletal parameters in the Sinclair S-1 minipig: Histomorphometric assessment of skeletal changes at 3 months postovariectomy. J. Bone Miner. Res. 9(Suppl. 1), S258 (1994). 294. A. F. Franks, W. E. Wyder, X. J. Li, M. W. Lundy, M. L. Pfennig, N. Sacco-Gibson, R. M. Roberts, W. F. Trout, W. R. Myers, and G. A. Bouchard, Evaluation of skeletal parameters in the Sinclair S-1 minipig: Longitudinal analysis of ovariectomy-induced skeletal changes. J. Bone Miner. Res. 9 (Suppl. 1), S191 (1994). 295. J. Kragstrup, A. Richards, and O. Fejerskov, Effects of fluoride on cortical bone remodeling in the growing domestic pig. Bone 10, 421 – 424 (1989).
DONALD B. KIMMEL 296. D. M. Raab, T. Crenshaw, D. B. Kimmel, and E. L. Smith, Cortical bone response in adult swine after exercise. J. Bone Miner. Res. 6, 741 – 749 (1991). 297. R. Webb, G. Baxter, D. McBride, M. Ritchie, and A. J. Springbett, Mechanism controlling ovulation rate in ewes in relation to seasonal anoestrus. J. Reprod. Fertil. 94, 141 – 151 (1992). 298. B. Malpaux and F. J. Karsch, A role for short days in sustaining seasonal reproductive activity in the ewe. J. Reprod. Fertil. 90, 555 – 562 (1990). 299. E. F. Eriksen, L. Mosekilde, and F. Melsen, Effect of sodium fluoride, calcium, phosphate, and vitamin D2 on trabecular bone balance and remodeling in osteoporotics. Bone 6, 381 – 389 (1985). 300. P. Chavassieux, P. Pastoureau, G. Boivin, M. C. Chapuy, P. D. Delmas, and P. J. Meunier, Dose effects on ewe bone remodeling of short-term sodium fluoride administration — A histomorphometric and biochemical study. Bone 12, 421 – 427 (1991). 301. P. Chavassieux, P. Pastoureau, G. Boivin, M. C. Chapuy, P. D. Delmas, G. Milhaud, and P. J. Meunier, Fluoride-induced bone changes in lambs during and after exposure to sodium fluoride. Osteoporosis Int. 2, 26 – 33 (1991). 302. S. B. Hornby, S. L. Ford, N. J. Timpson, A. K. Ancill, and G. P. Evans, Skeletal changes in the ewe after ovariectomy. J. Bone Miner. Res. 9 (Suppl. 1), S258 (1994). 303. C. Bressot, P. J. Meunier, M. C. Chapuy, E. Lejeune, C. Edouard, and A. J. Darby, Histomorphometric profile, pathophysiology and reversibility of corticosteroid-induced osteoporosis. Metab. Bone Dis. Rel. Res. 1, 303 – 311 (1979). 304. P. Chavassieux, P. Pastoureau, M. C. Chapuy, P. D. Delmas, and P. J. Meunier, Glucocorticoid-induced inhibition of osteoblastic bone formation in ewes: A biochemical and histomorphometric study. Osteroporosis Int. 3, 97 – 102 (1993). 305. M. J. Jayo, D. S. Weaver, S. E. Rankin, and J. R. Kaplan, Accuracy and reproducibility of lumbar bone mineral status determined by DPA in live male cynomolgous macaques. Lab. Anim. Sci. 40, 266 – 269 (1990). 306. N. S. Pope, K. G. Gould, D. C. Anderson, and D. R. Mann, Effects of age and sex on bone density in the rhesus monkey. Bone 10, 109 – 112 (1989). 307. M. J. Jayo, S. E. Rankin, D. S. Weaver, C. S. Carlson, andT. B. Clarkson, Accuracy and precision of lumbar bone mineral content by DXA in live female monkeys. Calcif. Tissue Int. 49, 438 – 440 (1991). 308. J. Rogers and K. K. Kidd, Nuclear DNA polymorphisms in a wild population of yellow baboons (Papio hamadryas cynocephalus) from Mikumi National Park Tanzania. Am. J. Phys. Anthropol. 90, 477 – 486 (1993). 309. J. Rogers, G. Guano, and K. K. Kidd, Variability in nucelar DNA among non-human primates: Application of molecular genetic techniques to intro- and inter-species genetic analyses. Am. J. Primatol. 27, 93 – 105 (1992). 310. C. Jerome, D. B. Kimmel, J. A. McAlister, and D. S. Weaver, Effects of ovariectomy on iliac trabecular bone in baboons (Papio anubis). Calcif. Tissue Int. 39, 206 – 208 (1986). 311. C. Miller and D. Weaver, Bone loss in ovariectomized monkeys. Calcif. Tissue Int. 38, 62 – 65 (1986). 312. L. C. Miller, D. S. Weaver, J. A. McAlister, and D. R. Koritnik, Effects of ovariectomy on vertebral trabecular bone in the cynomolgus monkey. Calcif. Tissue Int. 38, 62 – 65 (1986). 313. K. Lundon and M. Grynpas, The long term effect of ovariectomy on the quality and quantity of cortical bone in the young cynomolgus monkey: A comparison of density fractionation and histomorphometric techniques. Bone 14, 389 – 395 (1993). 314. K. Lundon, M. Dumitriu, and M. Grynpas, The long-term effect of ovariectomy on the quality and quantity of cancellous bone in young macaques. Bone Miner. 24, 135 – 149 (1994).
CHAPTER 37 Animal Models for in Vivo Experimentation 315. D. R. Mann, C. G. Rudman, M. A. Akinbami, and K. G. Gould, Preservation of bone mass in hypogonadal female monkeys with recombinant human growth hormone administration. J. Clin. Endocrinol. Metab. 74, 1263 – 1269 (1992). 316. C. A. Shively, M. J. Jayo, D. S. Weaver, and J. R. Kaplan, Reduced vertebral bone mineral density in socially subordinate female cynomolgus macaques. Am. J. Primatol. 24, 135 – 138 (1991). 317. T. B. Aufdemorte, W. C. Fox, D. Miller, K. Buffum, G. R. Holt, and K. D. Carey, A non-human primate model for the study of osteoporosis and oral bone loss. Bone 14, 581 – 586 (1993). 318. M. D. Grynpas, C. B. Huckell, K. J. Reichs, C. J. Derousseau, C. Greenwood, and M. J. Kessler, Effect of age and osteoarthritis on bone mineral in rhesus monkey vertebrae. J. Bone Miner. Res. 8, 909 – 917 (1993). 319. D. R. Young, W. J. Niklowitz, and C. R. Steele, Tibial changes in experimental disuse osteoporosis in the monkey. Calcif. Tissue Int. 35, 304 – 308 (1983). 320. C. M. Schnitzler, U. Ripamonti, and J. M. Mesquita, Histomorphometry of iliac crest trabecular bone in adult male baboons in captivity. Calcif. Tissue Int. 52, 447 – 454 (1993). 321. R. R. Recker, D. B. Kimmel, A. M. Parfitt, K. M. Davies, N. Keshawarz, and S. Hinders, Static and tetracycline-based bone histomorphometric data from 34 normal post-menopausal females. J. Bone Miner. Res. 3, 133 – 144 (1988). 322. C. P. Jerome, C. J. Lees, T. C. Register, and R. P. Holmes, Bone loss in skeletally mature female monkeys fed minimal requirement of calcium is accelerated by ovariectomy. J. Bone Miner. Res. 9 (Suppl. 1), S259 (1994). 323. D. S. Weaver, R. A. Power, C. P. Jerome, B. M. Phifer, and T. C. Register, The effects of immediate and delayed treatment of ovariectomized cynomolgus monkeys with nandrolone decanoate on densitometry, bone biomarkers, sex steroids and body weight. J. Bone Miner. Res. 9 (Suppl. 1), S393 (1994).
47 324. R. A. Power, D. S. Weaver, C. P. Jerome, B. M. Phifer, and T. C. Register, Histomorphometry and serum bone biomarkersin ovariectomized nandrolone-treated cynomolgus monkeys. J. Bone Miner. Res. 9 (Suppl. 1), S393 (1994). 325. C. S. Carlson, R. F. Loeser, M. J. Jayo, D. S. Weaver, M. R. Adams, and C. P. Jerome, Osteoarthritis in cynomolgus macaques: A primate model of naturally occurring disease. J. Orthop. Res. 12, 331 – 339 (1994). 326. D. B. Kimmel, N. E. Lane, C. M. Kammerer, M. R. Stegman, K. S. Rice, and R. R. Recker, Spinal pathology in adult baboons. J. Bone Miner. Res. 8 (Suppl. 1), S279 (1993). 327. K. P. Hughes, D. B. Kimmel, C. M. Kammerer, K. M. Davies, K. S. Rice, and R. R. Recker, Vertebral morphometry in adult female baboons. J. Bone Miner. Res. 9 (Suppl. 1), S209 (1994). 328. J. Dequeker, The relationship between osteoporosis and osteoarthritis. Clin. Rheum. Dis. 11, 271 – 296 (1985). 329. H. Pogrund, M. Rutenberg, M. Makin, G. C. Robin, R. Steinberg, and R. Bloom, Osteoarthritis of the hand and osteoporosis. Clin. Orthop. Relat. Res. 203, 239 – 243 (1986). 330. K. C. Ng, P. A. Revell, M. Beer, B. J. Boucherf, R. D. Cohen, and H. L. F. Currey, Incidence of metabolic bone disease in the rheumatoid arthritis and osteoarthritis. Ann. Rheum. Dis. 43, 370 – 377 (1984). 331. A. Hopkins, S. Zylstra, M. M. Hreshchyshyn, and M. Anbar, Normal and abnormal features of the lumbar spine observed in dual energy absorptiometry scans. Clin. Nucl. Med. 14, 410 – 414 (1989). 332. E. S. Orwoll, S. K. Oviatt, and T. Mann, The impact of osteophytic and vascular calcifications on vertebral mineral density measurements in men. J. Clin. Endocrinol. Metab. 70, 1202 – 1207 (1990). 333. I. R. Reid, M. C. Evans, R. Ames, and D. J. Wattie, The influence of osteophytes and aortic calcification on spinal mineral density in postmenopausal women. J. Clin. Endocrinol. Metab. 72, 1372 – 1374 (1991). 334. P. J. Drinka, A. A. DeSmet, S. F. Bauwens, and A. Rogot, The effect of overlying calcification on lumbar bone densitometry. Calcif. Tissue Int. 50, 507 – 510 (1992).
CHAPTER 38
The Type I/Type II Model for Involutional Osteoporosis Update and Modification Based on New Observations B. LAWRENCE RIGGS,* SUNDEEP KHOSLA,* AND L. JOSEPH MELTON III,*† * Department of Internal Medicine, Division of Endocrinology and Metabolism, and †Department of Health Sciences Research, Section of Clinical Epidemiology, Mayo Clinic and Foundation, Rochester, Minnesota 55905
I. Introduction II. Validating Evidence III. Tests of Validity
IV. Conceptual Problems V. Summary and Conclusions References
I. INTRODUCTION
genders and is strongly related to age, we prefer to use the term involutional osteoporosis in place of primary osteoporosis. Riggs and Melton in 1983 [1], and in subsequent publications [2,3], proposed that involutional osteoporosis could be subdivided into two distinct syndromes — type I osteoporosis and type II osteoporosis — that differed with respect to changes in regional bone mineral density (BMD), pattern of fractures, hormonal changes, and causal mechanisms (Table 1). In this chapter, we review this model and update and modify it based on new observations, especially the findings that estrogen deficiency is a major cause of bone loss in elderly women and, perhaps also, in aging men [4].
A. Background Osteoporosis often is divided into primary osteoporosis and secondary osteoporosis syndromes, depending on whether or not the patient has a recognizable disease or is using a drug that can cause bone loss. Additionally, there is a general agreement that the rare syndromes of juvenile osteoporosis that occur in children near puberty and idiopathic osteoporosis that occurs in young adults of both sexes should be considered as separate entities. Because it occurs in both
OSTEOPOROSIS, SECOND EDITION VOLUME 2
49
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
50
RIGGS, KHOSLA, AND MELTON
TABLE 1
Characterization of the Two Main Types of Involutional Osteoporosis Type I
Type II
Age (years)
51 – 75
70
Sex ratio (F:M)
6:1
2:1
Type of bone loss
Mainly trabecular
Trabecular and cortical
Rate of bone loss
Accelerated
Not accelerated
Major fracture sites
Vertebrae (crush) and distal radius
Vertebrae (multiple wedge) and hip
Parathyroid function
Decreased
Increased
Estrogen effects
Mainly skeletal
Mainly extraskeletal
Main causes
Menopause plus individual predisposing factor(s)
Factors related to aging including late effects of estrogen deficiency
B. Clinical Characteristics Type I osteoporosis typically affects women within 15 to 20 years after menopause. It is characterized by fractures occurring at sites that contain relatively large amounts of cancellous bone, such as the vertebral body, distal forearm, and ankle. The mandible and the maxilla also contain substantial amounts of cancellous bone, and, therefore, increased tooth loss occurs in type I osteoporosis. The vertebral fractures are of the crush or collapse type and usually are associated with a reduction of more than 25% of vertebral height. These are commonly acutely painful [5], and the back pain may require up to 6 months to subside. Type II osteoporosis occurs in both sexes but is twice as common in women as in men. Although type II osteoporosis can occur at any age, it is the predominant form of osteoporosis in women and men over the age of 70 years. Fractures associated with type II osteoporosis occur at sites that contain both cancellous and cortical bone. The most typical type II osteoporotic fracture is hip fracture. Also, fractures of the pelvis, proximal humerus, and proximal tibia are commonly associated with type II osteoporosis. Typically, a specific type of vertebral fracture can be included as part of the syndrome of type II osteoporosis. Rather than the acutely painful crush or collapse fractures observed in type I osteoporosis, there usually is a gradual and progressive deformation of the vertebrae, leading to dorsal kyphosis, often referred to as “the dowager’s hump.” Such fractures are painless or are associated with minimal aching pain, are manifest as anterior wedge deformities usually with less than a 25% reduction in vertebral height, are located almost exclusively in the mid-thoracic area, and generally occur in a series of several adjacent vertebrae.
C. Relationship to Patterns of Age-Related Bone Loss The manifestations of type I and type II osteoporosis are closely related to underlying patterns of age-related bone
loss. Based on cross-sectional and longitudinal bone densitometric studies [6 – 9], two distinct phases of age-related bone loss have been recognized — a slow age-related phase that occurs in both sexes and an accelerated phase that occurs only in postmenopausal women and, more rarely, in hypogonadal men. These are shown diagrammatically in Fig. 1. The slow phase of bone loss begins about age 40 and continues throughout life, has a similar rate in both sexes, and results in loss of similar amounts of cortical and cancellous bone. A transient accelerated phase beginning at the menopause is superimposed in women and results in a loss of disproportionately more cancellous bone than cortical bone. The accelerated postmenopausal loss lasts about 10 years, although most of the bone is lost in the first 3 to 4 years, and it declines exponentially with time until it
FIGURE 1
Changes in bone mass with aging in men and women showing patterns of bone loss. (I) Peak bone mass, (II) rapid phase of bone loss seen in women around the menopause, (III) is the age-related bone loss which is similar in both men and women. Modified from B. L. Riggs and L. J. Melton III, Involutional osteoporosis. In “Oxford Textbook of Geriatric Medicine, (J. G. Evans and T. F. Williams, eds.), pp 405 – 411. Oxford University Press, Oxford.
51
CHAPTER 38 Type I/Type II Model for Involutional Osteoporosis
merges asymptotically with the continuing slow phase of loss.
II. VALIDATING EVIDENCE The type I/type II model of involutional osteoporosis is supported by four different types of data: — differences in fracture patterns as assessed by epidemiologic studies, patterns of bone loss as assessed by bone densitometry, differences in parathyroid function, and differences in hormonal mechanisms of bone loss. That these data are largely independent of each other and yet lead to the same conclusion is, we believe, strongly supportive of the central hypothesis.
A. Differences in Fracture Pattern Striking differences have been noted with regard to the distribution by gender, frequency of occurrence, and location of fractures. As the epidemiological characteristics of a disease reflect its underlying pathophysiology, these observed differences in fracture pattern suggest that the pathophysiological mechanisms responsible for type I and type II osteoporosis are different. The female-to-male incidence ratio is about 6:1 for vertebral fractures occurring between ages 51 and 65 years, whereas it is only 2:1 for hip fractures occurring after 75 years [1,2]. Among women, the incidence of relatively “pure” fractures of type I osteoporosis, such as fracture of the distal forearm, rises soon after menopause, continues to
FIGURE 2
rise for another 10 to 15 years, and then plateaus. In contrast, the incidence of fractures of the femoral neck, a relatively “pure” type II osteoporotic fracture, increases exponentially throughout life. These contrasting patterns are shown in Fig. 2. In contrast to these clearly different fracture patterns for the hip and wrist, the pattern for vertebral fracture appears to be intermediate between the type I and type II patterns. Like type I fractures, the incidence of vertebral fractures increases soon after menopause and by age 65 only 2% of women will have had a hip fracture, whereas 11% will have had a vertebral fracture. Thereafter, the ratio of incidence for vertebra and hip fractures declines steadily until, among women who are 85 years of age or older, it is 1:1 [1,2]. We believe that this intermittent pattern occurs because vertebral fractures are, in fact, an admixture of two different types — crush fractures due to excessive osteoclastic activity and perforative resorption of cancellous plates occurring soon after menopause and wedge fractures due to trabecular thinning from impaired osteoblastic activity later in life. Direct evidence for such an admixture is supported by the study of Jensen et al. [10] who found in a random sample of 70-year-old Danish women, onefifth of whom had vertebral fractures, that 20% of them were crush fractures whereas 80% of them were wedge deformities. Finally, the proportional content of cancellous bone determines the location of fracture in the two osteoporotic syndromes. In patients with type I osteoporosis, fractures occur at sites that contain 50 to 75% cancellous bone, whereas, in patients with type II osteoporosis, fractures occur at sites that contain only 25 to 50% cancellous bone.
Incidence rates for a “pure” type I osteoporotic fracture (Colles’ fracture of distal forearm) and for a “pure” type II osteoporotic fracture (fracture of proximal femur) plotted as a function of age at the time of the fracture. Data are for women from the community population of Rochester, Minnesota (L. J. Melton and B. L. Riggs, unpublished data).
52
RIGGS, KHOSLA, AND MELTON
TABLE 2 Clinical syndrome Location pattern
Age-Related Fractures: Types of Patterns
Type I osteoporosis
Type II osteoporosis
Distal forearm
Proximal femur
Distal tibia
Proximal humerus
Vertebrae (crush)
Traumatic only Shafts of limb bones
Proximal tibia Pelvis Vertebrae (multiple wedge)
Sex ratio (F:M)
6:1
Bone composition
85% Trabecular
2:1
1:1
Cortical 50 – 70%
85% Cortical
Trabecular 30 – 50% Note. F, female; M, male.
These complex relationships are shown in Table 2. A third pattern is exhibited by fractures of the shafts of the long bones that are composed almost entirely of cortical bone. These fractures show neither a definite female preponderance nor an increase with age, and they occur almost entirely in association with severe trauma. Thus, this third pattern of fractures does not seem to be directly related to osteoporosis.
B. Differences in Patterns of Bone Loss Women with vertebral crush fractures due to type I osteoporosis have a mean BMD for lumbar spine, a site containing predominately cancellous bone, that is about 2 standard deviations (SD) below the mean for age-matched normal women, and about half of the individual values are below the 10th percentile of normal. However, for the radius shaft, a site containing predominately cortical bone, the mean BMD is only about 0.5 SD below the normal mean, and only a few individual values are below the 10th percentile [7]. In contrast, elderly patients with hip fracture due to type II osteoporosis have only small decreases below the age-adjusted normal mean for BMD at any site, although individual values tend to cluster in the lower part of the range [8]. These data suggest that patients with type I osteoporosis have lost disproportionately more cancellous bone than cortical bone, whereas patients with type II osteoporosis have lost similar amounts of both types of bone (Fig. 3).
C. Differences in Parathyroid Function An important difference that is qualitative, rather than quantitative, is parathyroid function as it changes in opposite directions in the two syndromes. As previously reviewed [4], during the accelerated phase of bone loss
early after menopause and in established osteoporosis, parathyroid function is normal or slightly decreased due to compensatory suppression to maintain normal levels of serum ionized calcium activity in the presence of the primary increased osteoclastic activity. When bone resorption is decreased by estrogen treatment, however, there are significant increases in circulating PTH, both in the early postmenopausal accelerated phase of bone loss and in established type I osteoporosis [11], consistent with a partial suppression of parathyroid function. In contrast, PTH concentrations increase with aging and its levels are higher in some patients with hip fracture than their age-matched controls [12]. Moreover, elderly women have abnormal PTH secretory dynamics that are consistent with functional parathyroid gland hyperplasia. Therefore, in type I osteoporosis, decreased PTH secretion may be the result of the increase in bone loss, whereas, in type II osteoporosis, increased PTH secretion may be driving the bone loss.
D. Differences in Causal Mechanisms The data thus far cited show that there are major differences in the presentation of the two syndromes suggesting that there are major differences in pathophysiologic mechanisms. We will review below the evidence that such differences do exist. 1. TYPE I OSTEOPOROSIS The predilection for women and the temporal proximity to menopause implicate estrogen deficiency as the etiologic agent for this type of osteoporosis. The relatively rapid and large decrease in estrogen secretion at menopause leads to increased bone turnover, and bone resorption increases to a greater extent than does bone formation [4,12], leading to rapid bone loss. However, although all postmenopausal women are estrogen deficient, type I osteoporosis occurs in only about 10 to 20% of them. To reconcile this discrepancy,
53
CHAPTER 38 Type I/Type II Model for Involutional Osteoporosis
FIGURE 3
Bone mineral density (BMD) levels for vertebrae and femoral neck, plotted as a function of age for 111 patients with vertebral fractures () and 49 patients with hip fractures (). The line represents the regression on age; the cross-hatched area shows the 90% confidence limits for 166 normal women. Note that the fracture threshold (90th percentile of the measurements for patients with fractures) is about 1.0 g/cm2 and is independent of age. Data are expanded from those reported by B. L. Riggs, H. W. Wahner, E. Seeman, K. P. Offord, W. L. Dunn, R. B. Mazess, K. A. Johnson, and L. J. Melton, III, Changes in bone mineral density of the proximal femur and spine with aging: Differences between the postmenopausal and senile osteoporosis syndromes. J. Clin. Invest. 70, 716 – 723 (1982).
we have suggested that the type I osteoporosis syndrome requires not only estrogen deficiency, but some additional factor or factors that operate only in the presence of estrogen deficiency and lead to an exacerbation and prolongation of the rapid phase of bone loss [1 – 3]. In normal women, this rapid phase lasts only from 4 to 8 years, whereas in women who develop type I osteoporosis it may last for 15 to 20 years. It is unclear at present what are the factors enhancing bone loss in this subset of early postmenopausal women. Perhaps a genetically determined increased responsiveness of bone in the presence of estrogen deficiency may predispose some postmenopausal women, but not others, to excessive bone loss. Much has been learned in the past decade about the paracrine mediators of the estrogen effect on bone. A number of proinflammatory cytokines, including interleukin-1 (IL-1), IL-6, tumor necrosis factor (TNF), and prostaglandin E2, increase the abundance of osteoclast precursors in bone marrow; and the presence of macrophage colony-stimulating factor(M-CSF) and newly discovered members of the TNF and TNF receptor superfamilies, osteoprotegerin and receptor activator of NF-B(RANK) ligand [12], allow these to be differentiated into active mature osteoclasts. Estrogen appears to act at multiple levels in this regulatory system to reduce osteoclast formation. Polymorphisms of genes coding for these paracrine effectors could lead to enhanced resorptive activity when estrogen is deficient. Genetic polymorphisms resulting in differences in the
number or the function of estrogen receptors could also augment the effect of estrogen deficiency on bone cell activity, although this has not yet been convincingly demonstrated. Another potential predisposing factor is the presence of impaired renal tubular calcium transport leading to chronic renal calcium losses in women with type I osteoporosis [14]. The skeletal and renal abnormalities would not be mutually exclusive and both could result from local increases of the same cytokine(s) in bone and kidney. 2. TYPE II OSTEOPOROSIS IN WOMEN This form of osteoporosis is caused by age-related factors occurring throughout the entire population of aging women and men. Of these, the two most important proximate causes appear to be secondary hyperparathyroidism and an age-related impairment in osteoblast function. In addition, some elderly housebound men and women develop nutritional deficiency of vitamin D [15], which exacerbates the secondary hyperparathyroidism and bone loss. Serum intact PTH concentrations and indices of bone turnover increase pari passu in aging women, and these increases correlate directly with each other, even after adjusting for the effect of age. As previously reviewed [4], suppression of PTH secretion by intravenous calcium infusion abolished the differences in bone resorption markers between young and elderly women. This suggests strongly that the increase in bone resorption in aging women is PTH-dependent. Concentrations of serum PTH and bone
54 resorption markers also increase in aging men and the pattern of increase closely resembles that in aging women. Thus, a considerable body of data implicates secondary hyperparathyroidism as a major cause of the slow, age-related phase of bone loss in both genders. Although the original 1983 hypothesis attributed the increase in bone resorption and secondary hyperparathyroidism to a primary age-related defect in intestinal calcium absorption and renal calcium conservation, this part of the hypothesis now must be modified to accommodate the new findings of the continued effects of estrogen deficiency during the slow phase of bone loss in both aging women and men [4]. Both McKane et al. [16], in experimental studies, and Khosla et al. [17], in population-based observational studies, demonstrated that the increases in both serum PTH and in bone resorption in postmenopausal women could be restored to the level found in premenopausal women by estrogen replacement therapy. Moreover, it has been recently demonstrated that, in addition to direct effects on bone cells, estrogen has potent extraskeletal effects on calcium homeostasis. The intestine contains estrogen receptors. Estrogen acts through these to increase intestinal calcium absorption [18,19], possibly by enhancing the responsiveness of the intestine to 1,25-dihydroxyvitamin D [19]. Estrogen also acts to increase renal calcium conservation [20,21] by enhancing tubular reabsorption of calcium through a PTHindependent mechanism [21]. In the presence of estrogen deficiency, the loss of these extraskeletal effects leads to calcium wasting and to a substantial increase in the level of dietary calcium required to prevent negative calcium balance. That the effects of estrogen in postmenopausal women are mainly extraskeletal, rather than skeletal, was further demonstrated by McKane et al. [22] who demonstrated that chronic ingestion of a very high calcium intake, 2400 mg/day, also restored bone resorption and serum PTH to premenopausal levels. Thus, the increases in bone resorption in elderly women can be normalized either by estrogen replacement (which restores the extraskeletal calcium fluxes to premenopausal levels) or by large increases in calcium intake (which offsets the net calcium losses induced by postmenopausal abnormalities in extraskeletal calcium fluxes). The cause of the impaired osteoblast function in elderly women is unclear. This abnormality exacerbates bone loss by preventing a compensatory increase in bone formation to offset the postmenopausal increases in bone resorption. As previously reviewed [4], histomorphometric data have demonstrated decreased bone formation at the cellular level with aging, which is decreased even more in patients with osteoporosis than in age- matched normals. These changes generally have been attributed to age-related factors, particularly to decreased paracrine production of growth factors or to decreased circulating levels of growth hormone and insulin-like growth factor-I (IGF-I).
RIGGS, KHOSLA, AND MELTON
It is also possible that this late occurring defect in osteoblast function could be partly or entirely the result of estrogen deficiency. Several observations are consistent with this possibility. Estrogen treatment increases production of IGF-I [23] and transforming growth factor- [24] by osteoblastic cells in vitro and acts to decrease osteoblast apoptosis [25]. Also, as previously reviewed [4], an effect of estrogen to stimulate skin collagen synthesis by fibroblasts has been repeatedly demonstrated and, as with fi broblasts, collagen production by osteoblasts represents 90% of cell biosynthetic capacity. However, estrogen treatment also has been reported both to stimulate and to inhibit bone formation in experimental animals and to stimulate and inhibit proliferation of human osteoblastic cells in vitro, so the issue remains unresolved. 3. TYPE II OSTEOPOROSIS IN MEN Men also lose bone with aging and have a third as many fractures as women. After accounting for their lack of the rapid phase of early postmenopausal bone loss, men exhibit the same pattern of slow bone loss and similar increases in bone resorption and serum PTH as aging women do [26]. These observations suggest that the causal mechanism for slow bone loss in aging men may be the same as or very similar to that of aging women. Recently reported data suggest that estrogen deficiency is the main cause for the secondary hyperparathyroidism and bone loss in elderly women [4]. However, because only a few aging men develop overt hypogonadism, how can estrogen deficiency be implicated? Recent studies in which either free or bioavailable sex steroids have been measured show that aging men do in fact have substantial decreases in both bioavailable estrogen and testosterone [4,26] (Table 3). In contrast to women, in whom the major cause of the decrease is a sharp
TABLE 3 Changes in Sex Steroids and Related Factors in Men and Women over Life in Random Age-Stratified Samples of 350 Women and 350 Men Men % change
Women % change
27**
45**
Bioavailable estrogen
47**
83**
Bioavailable testosterone
64**
28*
Sex hormone binding globulin
124**
1
Luteinizing hormone
285**
731**
Follicle stimulating hormone
505**
1805**
Lateral spine BMD Serum
Note. Data are from Khosla et al. (J. Clin. Endocrinol. Metab. 83 2266 – 2274, 1998) *P 0.05 **P 0.005.
55
CHAPTER 38 Type I/Type II Model for Involutional Osteoporosis
drop in their production rate due to menopause, the major cause of the decreases in both estrogen and testosterone in men is a large age-related increase in sex hormone binding globulin [26], which binds the sex steroids in an inactive complex. Recent clinical reports that young adult men with either null mutations of the estrogen receptor- gene (who cannot respond to estrogen) [27] or the aromatase gene (who cannot synthesize estrogen) [28,29] have osteoporosis despite normal serum testosterone concentrations suggest that estrogen plays a major role in maintaining bone mass in men. As previously reviewed [4], this is further corroborated by four large, population-based observational studies that have shown that serum free or bioavailable estrogen correlates better with BMD in men than does free or bioavailable testosterone. Nonetheless, these findings do not exclude a substantial effect of the decrease of bioavailable testosterone in mediating bone loss in aging men. Further studies are needed to quantify the relative effects of reduced levels of both sex steroids.
III. TESTS OF VALIDITY For any hypothesis to be useful, it should lead to predictions that can be tested experimentally. The limited number of tests that have been made thus far have all supported the hypothesis and are reviewed below. First, Owen et al. [30] and Kotowicz et al. [31] carried out epidemiologic studies of the hypothesis that fracture occurrence in involutional osteoporosis is the result of two distinct syndromes. Using the resources of the Rochester Epidemiology Project, they followed subjects through time by reviewing their medical records to determine whether distal forearm fractures or vertebral crush fractures occurring in women below age 65 would be subsequently associated with a small or a large increase in hip fractures late in life. If the null hypothesis that there is but one osteoporotic syndrome is correct, women having early postmenopausal fractures would have the most severe disease and should have a large subsequent increase in hip fractures late in life as compared to the overall Rochester population, possibly a 5- to 10-fold increase. In contrast, the increase in hip fractures late in life would be relatively small if the hypothesis of separate syndromes were correct. Using this populationbased retrospective cohort study design, these investigators found that the relative risk of hip fracture in women with previous distal forearm fractures was 1.2 and for women with vertebral fractures was 1.8. Thus, these results strongly support the type I/type II model for osteoporosis. Second, if involutional osteoporosis is a single entity, there should be no differences in the type of bone lost in patients with typical type I and type II fractures. This was tested experimentally by Johnston et al. [32] who studied
iliac bone biopsies from 32 women with vertebral fractures (type I osteoporosis) and compared them with biopsies obtained from 27 patients with hip fractures (type II osteoporosis). Patients with vertebral fractures had decreased trabecular bone volume compared to age-matched controls, and patients with hip fractures had a deficit of both cortical and trabecular bone compared to perimenopausal women, but not different from age-matched controls. These findings of histological differences in the proportion of cortical and frabicular bone type I and type II osteoporosis is consistent with different pathophysiological mechanisms.
IV. CONCEPTUAL PROBLEMS Since the type I/type II model of osteoporosis was proposed, questions and concerns have been raised. We feel that they can be answered and resolved by existing data and by a clearer understanding of the hypothesis. These are discussed below.
A. Role of Falls It has been suggested that both type I and type II osteoporosis are characterized by osteopenia and that apparent differences in fracture patterns might be explained by agerelated differences in the frequency and types of falls [33]. While the increased propensity of the elderly to fall contributes to fracture risk, it is the difference in the relative proportions of cancellous and cortical bone mass that mainly determine the differences in fracture pattern between the two types of disorders. Thus, it is clear that falls onto the hip lead to an increased risk of hip fractures [34] but comparable age-related increases in distal femur fracture [35] cannot be explained in this manner. Although certain falls predispose to certain types of fractures [33], the risk of fracture is proportional to the bone mass at the fracture site, and this in turn is a function of the overall pattern of cancellous and cortical bone loss.
B. Role of Peak Bone Mass The type I/type II model for involutional osteoporosis explains pathogenesis mainly in terms of patterns of agerelated bone loss rather than of differences in peak bone mass. Most estimates are that the variance in BMD for 70year-old women is approximately equally due to peak bone mass and to subsequent rates of bone loss [12,36 – 38]. Thus, the rank order of individuals in the Gaussian distribution of BMD values at the onset of involutional bone loss greatly influences their risk of falling below the threshold for fracture as age-related bone loss ensues. This is
56
RIGGS, KHOSLA, AND MELTON
supported by the demonstration of reduced bone mass in premenopausal daughters of osteoporotic women [39]. For women with type I osteoporosis, however, the greater bone loss rather than a greater BMD at the onset of the bone loss clearly is playing the dominant role. There is large variability in the rate of bone loss in the early years after menopause [38,40]. Also, women with established type I osteoporosis have higher turnover than their peers matched for years after menopause [41,42], and these differences are reflected by their much lower BMD values (Fig. 3). However, the contribution of peak bone mass to fracture risk probably is greater for type II osteoporosis because high variability in the rates of the slow phase of bone loss has not been demonstrated and because there is a large overlap between individuals with and without fractures in this syndrome. These observations suggest that the rank order of individuals in the Gaussian distribution of BMD values at the onset of involutional bone loss may be the main determinant of fracture risk in type II osteoporosis, consistent with the hypothesis proposed by Newton-John and Morgan [43] over 40 years ago.
C. Overlap of Causal Processes in Both Syndromes Another objection is that the menopause in women and the aging processes in both sexes, the proposed causes of type I osteoporosis and type II osteoporosis respectively, are universal processes. Therefore, how could these result in two separate syndromes? As illustrated in Fig. 4, specificity occurs because type I osteoporosis is the result of not just menopause but also of one or more additional factors that are present only in some postmenopausal women and that amplify and extend the bone loss induced by estrogen deficiency. Thus, women with type I osteoporosis may represent a separate population, distinct from their postmenopausal peers, who are losing or have lost bone at a more rapid rate or for a longer duration. In contrast, the processes causing type II osteoporosis appear to involve the entire population of aging men and women and, as the slow phase of bone loss progresses, an increasing number of them will have BMD values below the fracture threshold. The occurrence of an accelerated phase of bone loss after menopause in all women, however, including those destined to develop type II osteoporosis, explains why elderly women have a twofold greater increase in hip fracture than elderly men even though the rates of slow bone loss are similar in both sexes.
D. Causal Role of Estrogen Deficiency Because we now consider estrogen deficiency to be the dominant mechanism for both the rapid accelerated phase
FIGURE 4
Model for development of type I osteoporosis and type II osteoporosis. Dark line, represents regression of BMD on age for women. Degree of fracture risk for levels of BMD are shown in inset. Slow bone loss begins in the fourth decade and accelerates transiently after menopause. Light line, represents regression of BMD on age for the subset of women in whom type I osteoporosis develops. In these, an additional factor(s) causes the postmenopausal acceleration of bone loss to be exaggerated and prolonged so that BMD falls below the fracture threshold within 15 to 20 years after menopause. In contrast, patients with type II osteoporosis represent the entire population of aging women or men (Gaussian distribution shown by top-shaped figure). As slow bone loss progresses, increasing proportions of the general population fall below the fracture threshold late in life. Although all eventually will become at risk for fractures (distribution of fracture patients shown by small solid circles), those with the lowest BMD are at the greatest risk (From B. L. Riggs and L. J. Melton, III Clinical heterogeneity of involutional osteoporosis: Implications for prevention therapy. J. Clin. Endocrinol. Metab 70, 1229 – 1232 (1990)).
of bone loss in early postmenopausal women and for the slow phase of bone loss in elderly women and aging men [4], how is it possible for a single process to produce different clinical manifestations? New information suggests that estrogen-deficiency causes bone loss both through the loss of the direct action of estrogen on bone cells (that restrains bone turnover) and through the loss of the action of estrogen on the intestine and kidney (that maintain extraskeletal calcium fluxes) [4]. In the rapid phase of bone loss in early postmenopausal women, the contribution of both processes are balanced with a slight predominance of the loss of the direct effects on bone cells that restrain their activity. This slight imbalance leads to a net transfer of calcium from bone into the extracellular fluids that causes minimal suppression of parathyroid function. As the response of bone cells to estrogen deficiency wanes, the rapid phase of bone loss subsides over 4 to 8 years. Thereafter, the effect of estrogen loss on extraskeletal calcium homeostasis leads to secondary hyperparathyroidism that then becomes the predominant mechanism driving bone loss. The presence of estrogen-independent causes of decreased osteoblast function, if these exist, and nutritional vitamin D
CHAPTER 38 Type I/Type II Model for Involutional Osteoporosis
deficiency in some elderly persons could further aggravate this loss.
V. SUMMARY AND CONCLUSIONS Evidence from multiple sources suggests that involutional osteoporosis can be subdivided into distinct syndromes of type I (postmenopausal) osteoporosis and type II (age-related) osteoporosis. Although estrogen deficiency is the dominant cause of bone loss in both types of osteoporosis, the mechanisms by which it induces bone loss differs. The type I/type II model has proved to be useful in organizing the large amount of information that has developed on osteoporosis during the past 15 years and has been helpful in examining the pathogenesis of the disease and in evaluating therapeutic results. We believe that it still provides these useful functions. Undoubtedly, however, further modifications will be required as new information is accrued on pathogenesis, particularly that relating to underlying genetic and molecular mechanisms.
Acknowledgments This work was supported by NIH Grants AG – 04875 and AR – 27065.
References 1. B. L. Riggs and L. J. Melton, Evidence for two distinct syndromes of involutional osteoporosis. Am. J. Med. 75, 899 – 901 (1983). 2. B. L. Riggs and L. J. Melton. Medical progress series: Involutional osteoporosis. N. Engl. J. Med. 314, 1676 – 1686 (1986). 3. B. L. Riggs and L. J. Melton, Clinical heterogeneity of involutional osteoporosis: Implications for preventive therapy. J. Clin. Endocrinol. Metab. 70, 1229 – 1232 (1990). 4. B. L. Riggs, S. Khosla, and L. J., Melton III, A unitary model for involutional osteoporosis: Estrogen deficiency causes both Type I and Type II osteoporosis in postmenopausal women and contributes to bone loss in aging men. J. Bone Miner. Res. 13, 763 – 773 (1998). 5. C. Cooper, E. J. Atkinson, W. M. O’Fallon, and L. J., Melton III, The incidence of clinically diagnosed vertebral fractures: A populationbased study in Rochester, Minnesota, 1985 – 1989. J. Bone Miner. Res. 7, 221 – 227 (1992). 6. R. Lindsay, J. M. Aitken, J. B. Anderson, D. M. Hart, E. B. MacDonald, and A. C. Clarke. Long-term prevention of postmenopausal osteoporosis by oestrogen: Evidence for an increased bone mass after delayed onset of oestrogen treatment. Lancet 1, 1038 – 1041 (1976). 7. B. L. Riggs, H. W. Wahner, W. L. Dunn, R. B. Mazess, K. P. Offord, and L. J., Melton III, Differential changes in bone mineral density of the appendicular and axial skeleton with aging: Relationship to spinal osteoporosis. J. Clin. Invest. 67, 328 – 335 (1981). 8. B. L. Riggs, H. W. Wahner, E. Seeman, K. P. Offord, W. L. Dunn, R. B. Mazess, K. A. Johnson, and L. J., Melton III, Changes in bone mineral density of the proximal femur with aging: Differences between the postmenopausal and senile osteoporosis syndromes. J. Clin. Invest. 70, 716 – 723 (1982).
57 9. J. M. Pouilles, F. Tremollieres, and C. Ribot. The effects of menopause on longitudinal bone loss from the spine. Calcif. Tissue Int. 52, 340 – 343 (1993). 10. G. F. Jensen, C. Christiansen, J. Boesen, V. Hegedus, and I. Transbøl. Epidemiology of postmenopausal spinal and long bone fractures: A unifying approach to postmenopausal osteoporosis. Clin. Orthop. Relat. Res. 166, 75 – 81 (1982). 11. B. l. Riggs, C. D. Arnaud, J. Jowsey, R. S. Goldsmith, and P. J. Kelly, Parathyroid function in primary osteoporosis. J. Clin. Invest. 52, 181 – 184 (1973). 12. P. Garnero, Sornay-E. Rendu, M. Choppy, and P. D. Delmas, Increased bone turnover in late postmenopausal women is a major determinant of osteoporosis. J. Bone Miner. Res. 11, 337 – 349 (1996). 13. L. C. Hofbauer, S. Khosla, C. R. Dunstan, D. L. Lacey, W. J. Boyle, and B. L. Riggs, The roles of osteoprotegerin and osteoprotegerin ligand in the paracrine regulation of bone resorption. J. Bone Miner. Res. 15, 2 – 12 (2000). 14. H. M. Heshmati, S. Khosla, M. F. Burritt, W. M. O’Fallon, and B. L. Riggs , Primary defect in renal calcium conservation may contribute to the pathogenesis of postmenopausal osteoporosis. J. Clin. Endocrinol. Metab. 83, 1916 – 1920 (1988). 15. M. C. Chapuy, M. E. Arlot, F. Duboeuf, J. Brun, B. Crouzet, S. Arnaud, P. D. Delmas, and P. J. Meunier, Vitamin D3 and calcium to prevent hip fractures in elderly women. N. Engl. J. Med. 327, 1637 – 1642 (1992). 16. R. W. McKane, S. Khosla, J. Risteli, S. P. Robins, J. M. Muhs, and B. L. Riggs, Role of estrogen deficiency in pathogenesis of secondary hyperparathyroidism and increased bone resorption in elderly women. Proc. Assoc. Am. Physicians 109, 174 – 180 (1997). 17. S. Khosla, E. J. Atkinson, L. J. Melton, III, and B. L. Riggs, Effects of age and estrogen status on serum parathyroid hormone levels and biochemical markers of bone turnover in women: A population-based study. J. Clin. Endocrinol. Metab. 82, 1522 – 1527 (1997). 18. J. C. Gallagher, B. L. Riggs, and H. F. DeLuca, Effect of estrogen on calcium absorption and serum vitamin D metabolites in postmenopausal osteoporosis. J. Clin. Endocrinol. Metab. 51, 1359 – 1364 (1980). 19. C. Gennari, D. Agnusdei, P. Nardi, and R. Civitelli, Estrogen preserves a normal intestinal responsiveness to 1,25-dihydroxyvitamin D3 in oophorectomized women. J. Clin. Endocrinol. Metab. 71, 1288 – 1293 (1990). 20. B. E. C. Nordin, A. G. Need, H. A. Morris, M. Horowitz, and W. G. Robertson, Evidence for a renal calcium leak in postmenopausal women. J. Clin. Endocrinol. Metab. 72, 401 – 407 (1991). 21. W. R. McKane, S. Khosla, M. F. Burritt, P. C. Kao, D. M. Wilson, S. J. Ory, and B. L. Riggs, Mechanism of renal calcium conservation with estrogen replacement therapy in women in early postmenopause: A clinical research center study. J. Clin. Endocrinol. Metab. 80, 3458 – 3464 (1995). 22. W. R. McKane, S. Khosla, K. S. Egan, S. P. Robins, M. F. Burritt, and B. L. Riggs, Role of calcium intake in modulating age-related increases in parathyroid function and bone resorption. J. Clin. Endocrinol. Metab. 81, 1699 – 1703 (1996). 23. M. Ernst, J. K. Heath, and G. A. Rodan, Estradiol effects on proliferation, messenger ribonucleic acid for collagen and insulin-like growth factor-I, and parathyroid hormone-stimulated adenylate cyclase activity in osteoblastic cells from calvariae and long bones. Endocrinology 125, 825 – 833 (1989). 24. M. J. Oursler, C. Cortese, P. E. Keeting, M. A. Anderson, S. K. Bonde, B. L. Riggs, and T. C. Spelsberg, Modulation of transforming growth factor production in normal human osteoblast-like cells by 17-estradiol and parathyroid hormone. Endocrinology 129, 3313 – 3320 (1991). 25. S. C. Manolagas, Birth and death of bone cells: Basic regulatory mechanisms and implications for the pathogenesis and treatment of osteoporosis. Endocrinol. Rev. (in press).
58 26. S. Khosla, L. J. Melton, III, E. J. Atkinson, W. M. O’Fallon, G. G. Klee, and B. L. Riggs, Relationship of serum sex steroid levels and bone turnover markers with bone mineral density in men and women: A key role for bioavailable estrogen. J. Clin. Endocrinol. Metab. 83, 2266 – 2274 (1998). 27. E. P. Smith, J. Boyd, G. R. Frank, H. Takahashi, R. M. Cohen, B. Specker, T. C. Williams, D. B. Lubahn, and K. S. Korach, Estrogen resistance caused by a mutation in the estrogen-receptor gene in a man. N. Engl. J. Med. 331, 1056 – 1061 (1994). 28. C. Carani, K. Qin, M. Simoni, M. Faustini-Fustini, S. Serpente, J. Boyd, K. S. Korach, and E. R. Simpson, Effect of testosterone and estradiol in a man with aromatase deficiency. N. Engl. J. Med. 337, 91 – 95 (1997). 29. A. Morishima, M. M. Grumbach, and J. P. Bilezikian, Estrogen markedly increases bone mass in an estrogen deficient young man with aromatase deficiency. J. Bone Miner. Res. 12, S126. 30. R. A. Owen, L. J. Melton, III, D. M. Ilstrup, K. A. Johnson, and B. L. Riggs, Colles’ fracture and subsequent hip fracture risk. Clin. Orthop. 171, 37 – 44 (1982). 31. M. A. Kotowicz, L. J. Melton III, C. Cooper, E. J. Atkinson, W. M. O’Fallon, and B. L. Riggs, Risk of hip fracture in women with vertebral fracture. J. Bone Miner. Res. 9, 599 – 605 (1994). 32. C. C. Johnston, J. Norton, M. R. A. Khairi, C. Kernek, C. Edouard, M. Arlot, and P. J. Meunier, Heterogeneity of fracture syndromes in postmenopausal women. J. Clin. Endocrinol. Metab. 61, 551 – 556 (1985). 33. M. C. Nevitt, S. R. Cummings, and the Study of Osteoporotic Fractures Research Group, Type of fall and risk of hip and wrist fractures: The study of osteoporotic fractures. J. Am. Geriatr. Soc. 41, 1226 – 1234 (1993).
RIGGS, KHOSLA, AND MELTON 34. S. L. Greenspan, E. R. Myers, L. A. Maitland, N. M. Resnick, and W. C. Hayes, Fall severity and bone mineral density as risk factors for hip fracture in ambulatory elderly. JAMA 271, 128 – 133 (1994). 35. T. J. Arneson, L. J. Melton III, D. G. Lewallen, and W. M. O’Fallon, Epidemiology of diaphyseal and distal femoral fractures in Rochester, Minnesota, 1965 – 1984. Clin. Orthop. 234, 188 – 194 (1988). 36. D. E. Meier, E. S. Orwoll, and J. M. Jones, Marked disparity between trabecular and cortical bone loss with age in healthy men. Ann. Intern. Med. 101, 605 – 612 (1984). 37. J. E. Block, R. Smith, C. Glueer, P. Steiger, B. Ettinger, and H. K. Genant, Models of spinal trabecular bone loss as determined by quantitative computed tomography. J. Bone Miner. Res. 4, 249 – 257 (1989). 38. M. A. Hansen, K. Overgaard, B. J. Riis, and C. Christiansen, Role of peak bone mass and bone loss in postmenopausal osteoporosis: 12 Year study. Br. Med. J. 303, 961 – 964 (1991). 39. E. Seeman, J. L. Hopper, L. A. Bach, M. E. Cooper, E. Parkinson, J. McKay, and G. Jerums, Reduced bone mass in daughters of women with osteoporosis. N. Engl. J. Med. 320, 554 – 558 (1989). 40. A. Laib, H. J. Hauselmann, and P. Ruegsegger, In vivo high resolution 3D-QCT of the human forearm. Technol. Health Care 6, 329–337 (1998). 41. E. F. Eriksen, S. F. Hodgson, R. Eastell, W. F. O’Fallon, and B. L. Riggs, Trabecular bone formation and resorption rates in type I (postmenopausal) osteoporosis. J. Bone Miner. Res. 3, S203 (1988). 42. R. Eastell, S. P. Robins, T. Colwell, A. M. A. Assiri, B. L. Riggs, and R. G. G. Russell, Evaluation of bone turnover in type I osteoporosis using biochemical markers specific for both bone formation and bone resorption. Osteoporos. Int. 3, 255 – 260 (1993). 43. H. F. Newton-John and D. B. Morgan, Osteoporosis: Disease or senescence? Lancet 1, 232 – 233.
CHAPTER 39
Bone Remodeling Findings in Osteoporosis ROBERT R. RECKER AND M. JANET BARGER-LUX Osteoporosis Research Center, Creighton University, Omaha, Nebraska 68131
I. Introduction II. Background III. Histomorphometric Findings in Osteoporosis
IV. Bone Remodeling at the Whole-Organism Level V. Summary References
I. INTRODUCTION
II. BACKGROUND
This chapter will consider the role of bone remodeling — chiefly as disclosed by histomorphometry of transilial bone biopsies from human subjects — in the pathogenesis of osteoporosis. The following pages will examine the findings that are plausible concomitants of osteoporosis, understood here as the loss of skeletal adaptation to normal mechanical usage. A recent consensus conference on osteoporosis [1] has redefined osteoporosis as “a skeletal disorder characterized by compromised bone strength [emphasis added] predisposing to an increased risk of fracture.” By focusing on compromised bone strength, not simply bone mass, this consensus definition has converged with the focus of this chapter. Comparisons of findings from healthy subjects and subjects with osteoporosis will be affected by the fact that “osteoporotic” specimens have been collected after the disease has become identifiable, usually after fragility fractures have occurred. The processes that led to bone fragility (whether approached via histomorphometric, kinetic, or biochemical methods) may no longer be present.
OSTEOPOROSIS, SECOND EDITION VOLUME 2
Other chapters in this volume describe bone anatomy, the bone remodeling system, the remodeling transient, and markers of bone turnover in detail. The following section is limited to those features that are particularly relevant to the present discussion.
A. Functional Organization of Bone 1. BONE ENVELOPES In a recent commentary and review [2], Seeman reminded researchers that “bone biology is the study of the behavior of the surfaces of the skeleton [emphasis added].” Each of five bone envelopes identifies one of these surfaces, layers in which groups of cells perform a similar function within a similar relative location. Each bone envelope has characteristic metabolic activities and characteristic systems for organizing the function of its bone cells. Bone gain is mediated by the outermost (periosteal) bone envelope, and bone loss occurs in the innermost (transitional and endosteal) envelopes. The following paragraphs
59
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
60
RECKER AND BARGER-LUX
identify the bone envelopes, from the outer to the inner margin of bone tissue. a. Periosteal Envelope The periosteal envelope or periosteum covers most of the skeleton on its outer margin. It includes a sheath of fibrous connective tissue and an underlying layer of undifferentiated cells (cambium). Periosteum is absent from the attachment points of tendons and ligaments, areas covered by articular cartilage, the neck of the femur, the subscapular area, and sesamoid bones. Expansion of the periosteal envelope corresponds to changes in skeletal dimensions that occur throughout life. The periosteal envelope functions in fracture healing and in adaptation to intense mechanical usage [3 – 5]. b. Haversian Envelope The cells and surfaces of cortical bone delineate the Haversian envelope [6]. It includes the Haversian systems, their vessels and nerves, the osteocytes and their cytoplasmic extensions, the adjacent mineralized tissue, and the walls of the Volkmann’s canals. Thinning of the Haversian envelope (i.e., cortical thinning) corresponds to movement of the cortico-endosteal envelope toward the periosteal envelope [7]. c. Cortico-endosteal Envelope This envelope delimits the outermost boundary of the medullary canal, where trabeculae connect to the inner wall of the cortex. In adolescent girls, this envelope moves centrally toward the marrow cavity for a brief period; otherwise, the cortico-endosteal envelope expands throughout life [8]. d. Transitional Envelope This envelope is immediately adjacent to the cortico-endosteal envelope. Keshawarz and Recker identified this area as the transitional zone [7]. Though most workers have included it within the endosteal envelope, the transitional envelope can be considered separately because it is there that trabeculation of cortical bone has occurred. e. Endosteal Envelope This is the interface between bone marrow and trabecular (cancellous) bone [6]. The ratio of endosteal surface to trabecular bone volume is high compared to surface-to-volume ratios of the other envelopes. 2. INTERMEDIARY ORGANIZATION: REMODELING Frost [9] coined the term “intermediary organization” (IO) of the skeleton to describe the regulation of bone cell activity. He recognized that bone cells do not function individually and that the end product of their group activity is not disrupted by interventions that change the work or life span of a single class of bone cells such as osteoclasts. He inferred that bone cell activity was regulated by an order of control higher than a single cell (osteoclast or osteoblast) or a single function (resorption or formation). Four functional
subdivisions of the IO paradigm can be recognized: growth, modeling, remodeling, and fracture healing. Each has its own discreet IO but utilizes the same bone cells to accomplish its work. This discussion will focus on the remodeling IO. a. Activation, Resorption, and Formation Remodeling is the predominant metabolic activity of the skeleton in adult human life. It functions to remove and replace bone. At each remodeling site, the ARF sequence of activity takes place, with activation of the remodeling process, resorption of existing bone, and formation of new bone [10]. Osteoclasts excavate resorption tunnels (in cortical bone) and surface pits (in trabecular bone), osteoblasts fill the tunnels and pits with matrix, and the matrix mineralizes to become new bone. A complete cycle takes about 6 months in trabecular bone [11] and longer in cortical bone. The group of cells which carry out the work of a remodeling site has been called the basic multicellular unit (BMU), and the quantum of bone formed by a BMU has been called the basic structural unit (BSU) [12] (see Figs. 1 – 3). b. Coupling of Resorption and Formation The tight link between osteoclast and osteoblast function in bone remodeling is referred to as “coupling” [13]. In Frost’ paradigm, coupling is part of the remodeling IO. Both resorption and formation occur at the same site and, absent problems (e.g., calcium deficiency or a reduction in physical activity), the amount formed is almost always very nearly equal to the amount resorbed. There has been little success in finding ways of uncoupling resorption and formation. Antiresorptive agents, rather than uncoupling resorption and formation, function by inhibiting osteoclast activation. c. The Remodeling Transient After administration of an antiresorptive agent, bone mass increases until the remodeling projects already underway have been completed. The gain in bone mass that corresponds to filling of the remodeling space with new bone is the remodeling transient [14]. This phenomenon does not indicate “cure” of a skeletal disease. To determine whether a treatment can raise bone mass beyond filling the remodeling space requires very long periods of observation (as long as several years in humans): after a complete remodeling cycle has passed, one must observe the new steady state for two or more cycles.
B. Sources of Bone Fragility 1. LOW BONE MASS Low bone mass has long been recognized as a major component of skeletal fragility, and techniques for measuring bone mass in vivo have established this link [15].
CHAPTER 39 Bone Remodeling Findings in Osteoporosis
Basic multicellular unit (BMU). Photomicrograph (Goldner stain, 325 original magnification), of trabecular resorbing surface from a thin section of an undemineralized human transilial biopsy spoecimen. The scalloped resorption surface is occupied on the left by a multinucleated osteoclast. (Copyright, Dr. Robert R. Recker, used with permission.)
FIGURE 1
FIGURE 2 Basic multicellular unit (BMU). Photomicrograph (Goldner stain, 325, original magnification), of trabecular forming surface from a thin section of an undemineralized human transilial biopsy specimen. The plump osteoblasts with eccentric nuclei are lined up on a dark osteoid surface. (Copyright, Dr. Robert R. Recker, used with permission.)
61
62
RECKER AND BARGER-LUX
Basic structural unit (BSU). Photomicrograph (Toluidine blue stain, 150 original magnification), of completed remodeling sites from a thin section of an undemineralized human transilial biopsy specimen. Dark stained cement lines demarcate the limit of resorption and the start of formation. (Copyright, Dr. Robert R. Recker, used with permission.)
FIGURE 3
a. Earlier Acquisition vs Later Conservation Low bone mass can originate from failure to acquire sufficient bone mass during skeletal growth and consolidation and/or failure to conserve bone later on [1]. However, much of the literature has uncritically attributed low bone mass to bone loss, sometimes by using the two terms almost interchangeably. In a young subject, there usually is no basis for attributing a finding of low bone mass to loss. It is, of course, well-established that bone loss occurs in association with menopause, disease, and the multiple concomitants of aging. In an older individual, however, low peak bone mass earlier in adult life may well exaggerate the bone deficit now attributed to loss. Bone loss in adults occurs via the bone remodeling mechanism. Failure to conserve bone during adult life could be due to intrinsic defect(s) in the remodeling IO, its physiologic functioning (e.g., in response to calcium deficiency or skeletal disuse), or some combination of the two. b. The Bone Deficit in Osteoporosis Cortical bone comprises about 80% of the mineralized skeleton of adults, with the remaining 20% present as trabeculae. Disappearance of half the trabeculae in the entire skeleton, therefore, would decrease total-body bone mineral by only about 10%. (In whole human vertebrae, the trabecular proportion is only modestly higher, about 26% [16].) However, bone mass measurements from patients with osteoporosis and
fragility fractures are typically 25 to 40% lower than corresponding values from ostensibly healthy young women [17]. These findings suggest that much of bone deficit in osteoporosis is a deficit of cortical bone. 2. MICROARCHITECTURAL DETERIORATION OF BONE Skeletal fragility out of proportion to low bone mass has been attributed to microarchitectural defects such as loss of trabecular connectivity or accumulation of microdamage. a. Loss of Trabecular Connectivity Loss of trabecular elements would reduce the number of interconnections between trabeculae. This loss of connectivity would explain some of the exaggerated fragility of the skeleton in patients with osteoporosis, because buckling strength varies with the fourth power of the distance between cross-connections in a structure. b. Microdamage in Bone Loading of any structural material is countered by stress, that is, the bend-resisting force within the material. Repeated submaximal loading produces microdamage, which weakens the material. The gravitational environment of Earth, the material properties of bone, the loads placed on the skeleton, and the number of load cycles together indicate that microdamage does occur in the human skeleton during life. Frost has estimated that, without repair of microdamage by the remodeling IO, total skeletal failure would ensue within 2 years [9].
63
CHAPTER 39 Bone Remodeling Findings in Osteoporosis
Figure 4 shows the expected appearance of microdamage in bone: cracks that have extended and propagated according to the direction of stress and the microstructure of the area. However, the study of microdamage in bone has been quite difficult, and we do not know how much of the
skeletal fragility of osteoporosis is due to accumulation of microdamage. c. Ultrasound Evaluation of Bone Bone densitometry measured by the “gold standard” method, dual-energy
FIGURE 4 Light photomicrograph (A) and SEM (B,C) of progressively higher magnification of a microcrack in bovine cortical bone. The specimen had been loaded (flexed) in cycles until there was change in mechanical properties, but a complete fatigue had not yet occurred. The crack propagated along the cement line and was trapped by the oseon.
64
RECKER AND BARGER-LUX
FIGURE 4
X-ray absorptionetry (DXA), cannot yield information on the structural factors that contribute to bone strength (i.e., trabecular connectivity, microdamage, or molecular aspects of bone matrix). Early work with quantitative ultrasound (QUS) of bone suggested that this technology might permit assessment of all the determinants of bone strength [18]. In a comparative study of subjects matched for BMD of the spine or forearm, patellar ultrasound velocity was significantly lower in osteoporotic patients than in normals (see Table 1) [19]. QUS has since been used to evaluate fracture risk in elderly women [20] and to screen for osteoporosis in perimenopausal women [21]. In biomechanical terms, QUS is related to the elastic properties of bone, which should bear some relationship to its microscructure. To date, however, only one published
TABLE 1
(continued )
clinical study has combined ultrasound and histomorphometry to examine trabecular connectivity [22]. In that study, the investigators cut sections from methylmethacrylateembedded specimens and tested the remaining block with a specialized ultrasound device. They reported relationships between in vitro QUS (of trabecular plus cortical bone) and two indices of trabecular connectivity.
C. Bone Histomorphometry Histomorphometry of undecalcified bone biopsies has been the principal method used to study bone remodeling as it relates to osteoporosis. Prior to reviewing the findings, however, two caveats are in order:
Ultrasound Velocitya in Osteoporotic and Normal Subjects Matched for Bone Density Matched by BMD at the
N pairs
Osteoporotic subjects mean SD (m/s)
Normal subjects mean SD (m/s)
P value
BV/Tb.N
Spineb
38
1800 15
1854 10
0.005
BV/Sr.V
Forearmc
46
1783 12
1827 14
0.02
a
As AVU (apparent velocity of ultrasound) measured at the patella. Spine BMD by DPA 0.706 0.025 g/cm2 c Forearm BMD by SPA 0.768 0.023 g/cm. b
65
CHAPTER 39 Bone Remodeling Findings in Osteoporosis
1. Studies of iliac crest biopsies have generated nearly all of the published histomorphometric data. While there are correlations between iliac and spinal trabecular bone [23], these relationships are far from perfect. Extrapolation of iliac crest findings to bone at other sites must be made with caution (though, to estimate and illustrate, we do so in the section that follows). 2. Nearly all published data from patients with osteoporosis are cross-sectional. Attempts to understand the longitudinal course of the disease from cross-sectional data will almost certainly be confounded by cohort differences [24]. Despite these cautions, it is worthwhile to look at histomorphometric findings in search of clues.
III. HISTOMORPHOMETRIC FINDINGS IN OSTEOPOROSIS A. Findings in Transilial Bone Biopsies 1. CORTICAL BONE a. Thinner Cortices Table 2 presents results of a comparative study of women with osteoporosis and ostensibly healthy women [25]. The women in the osteoporotic group were older (67 vs 60 years); they were also somewhat shorter (1.56 vs 1.60 m), and they weighed about 20 % less (55.9 vs 70.2 kg; figures are medians). The core width (CW, the distance between the inner and outer periosteal surfaces of the ilium) was less in the women with osteoporosis than in the healthy women (8.4 2.3 mm vs 9.4 1.7 mm, P 0.02), indicating smaller skeletons in the osteoporotic group. In the women with osteoporosis, inner and outer cortices (shown as cortical width Ct.W1 and Ct.W2,
respectively) were about 30 and 37% narrower [25]. (Extrapolated to the entire skeleton, 33% less cortical bone translates into 26% less total-body bone mineral.) This finding is consistent with the statement that low bone mass in osteoporosis represents chiefly a smaller amount of cortical bone. b. Trabeculation of Cortical Bone The pattern of bone loss from the cortex in osteoporosis is from the corticoendosteal surface, where the medullary canal enlarges at the expense of the inner cortex [7]. Bone loss does not occur at the periosteal surface. In transilial biopsies from patients with osteoporosis, resorption cavities within the subendosteal area enlarge and coalesce, resulting in net loss of bone tissue; progressive trabeculation of the cortex through this process creates the transitional envelope identified earlier. The resorption spaces in this zone are not typical Haversian systems. They do not run parallel to the long axis of the bone; instead, from the two-dimensional perspective of microscopic sections, they appear to be oriented haphazardly. The ARF pattern is operative at the corticoendosteal surface; however, formation, though not absent, is incomplete. In the emerging transitional envelope, therefore, the net effect of remodeling is bone loss. As the process continues, the inner cortex becomes completely trabeculated, and the new cortico-endosteal surface is closer to the periosteum. The pattern of cortical thinning and trabeculation of the inner cortex resembles that seen in disuse in the adult remodeling animal [26] (see Fig. 5). The major difference is that the process proceeds at a pace slower than that seen after acute disuse. 2. TRABECULAR BONE Together, the findings described below indicate that the major structural feature of trabecular bone loss in
TABLE 2 Structural Data Osteoporotic subjects N 90a (mean SD) BV/TV
13.9 4.5
13.42
(mean SD) 21.2 4.90
P value
20.5
0.001
8.6
0.02
0.586 0.264 (85)
0.517
0.984 0.386
0.861
0.001
Ct.W2
0.383 0.174 (79)
0.315
0.600 0.251 (31)
0.526
0.001
Cn.Wi
7.4 2.3 (79)
7.1
NS
W.Th
28.0 4.44
Tb.Th
124 27
Tb.N
1.12 0.24
Tb.Sp
928 0.218
8.3
7.4 28.2 119 1.09 894
9.4 1.7 (31)
(median)
Ct.W1
C.W
8.4 2.3 (79)
(median)
Normal subjects N 34a
7.8 1.8 (31) 32.1 4.13 138 29 1.55 0.26 654 139
Note. Modified from [25] a Ns which vary from the totals of 90 or 34 are shown in parentheses.
31.9 132 1.59 600
0.001 0.02 0.001 0.001
66
RECKER AND BARGER-LUX
FIGURE 5
Pattern of bone loss in disuse. Unstained, undemineralized cross sections of both humeri of adult beagle. (Left) Cross section from the left forelimb which was immobilized for 40 weeks in a plater cast. (Right) Cross section from the right forelimb which was not immobilized. Note the enlargement of the medullary canal on the left, with no significant difference in the outer circumference. (Courtesy of Prof. J. W. Jaworski, Ottawa.)
osteoporosis is disappearance of entire trabecular elements [25,27]. a. Unremarkable Dynamic Measures Studies from at least three centers have failed to uncover any important differences in the remodeling dynamics of trabecular bone between healthy subjects and patients with established osteoporosis [25,28,29]. As suggested earlier, absence of dynamic evidence of the processes leading to low bone mass means simply that such evidence was no longer present at the time of biopsy. b. Reduction in Trabecular Bone Mass Trabecular bone volume (BV/TV), averages about 35% lower in patients with osteoporosis, as compared to healthy individuals [25]; but there is considerable overlap. However, BV/TV does not give the complete picture of the effects of osteoporosis on trabecular bone. c. Reduction in Trabecular Connectivity Trabecular thickness (Tb.Th), trabecular number (Tb.N), and trabecular separation (Tb.S) are measures that describe the configuration of trabeculae in space. In the study cited earlier [25], investigators
from our laboratory compared these variables in transilial biopsies from normal subjects and from patients with osteoporosis (see Table 2). Tb.Th was only about 10% lower in the patients, while Tb.N was 28% lower and Tb.S was 30% higher. Specimens taken at autopsy from the vertebral bodies of osteoporotic patients have also shown loss of trabecular connectivity [30]. We also compared the spatial arrangement of trabeculae in osteoporotic patients and normals matched by BV/TV (see Table 3) [31]. The patients were about 10% lower in Tb.N and about 10% higher in Tb.Sp and Tb.Th. However, star volume (Sr.V), a more sensitive measurement of connectivity [32], was about 36% higher in the patients. A recent paper utilized a direct method (“three-dimensional histology”) for studying trabecular discontinuity: the technique involves superficial staining of a thick (300 m) section from each transilial biopsy. By this technique, artifactual trabecular termini (created by transecting trabeculae during sectioning) on the surface take up the stain, but real termini within the section do not. The investigators compared osteopenic, postmenopausal women with and without vertebral fractures, and the fracture group had almost four times as many termini (as number per mm2 of section
67
CHAPTER 39 Bone Remodeling Findings in Osteoporosis
TABLE 3 Spatial Arrangement of Trabeculae in Osteoporotic and Normal Subjects Matched for BV/TV Osteoporotic subjects (mean SD)
Normal subjects (mean SD)
P value
BV/Tb.N
1.33 0.21
1.49 0.27
0.001
BV/Tb.Sp
753 117
686 171
0.05
BV/Tb.Th
144 27
130 26
0.05
9.661 1.053
7.128 0.825
0.03
BV/Sr.V
Note. N 23 pairs; BV/TV 19.1 4.2.
surface) as did the nonfracture group [33]. Conventional histomorphometry of the same biopsies failed to disclose a structural distinction between the fracture and nonfracture groups, which were similar in trabecular mass [34]. The finding that osteoporotic patients with vertebral fractures are lower in trabecular connectivity than other individuals with the same trabecular bone mass [33] probably does explain, at least in part, why some persons with low spine BMD sustain vertebral fractures and others do not. d. Trabecular Thinning Uniform loss of bone from all trabecular surfaces would be relatively easy to document from biopsy data: there would be general thinning of trabeculae. Normal values for wall thickness (W.Th) would suggest excessively deep erosion by osteoclasts. Reduced values for wall thickness would suggest incomplete replacement of eroded bone by the osteoblasts. A combination of these two mechanisms would be harder to detect, but still detectable. Wall thickness is reduced in biopsies from patients with osteoporosis [35], but trabecular thinning is a relatively minor part of the bone loss. Erosion depth, measured directly, is reportedly not increased in patients with osteoporosis [36,37]; however, the measurement itself is technically challenging, and other laboratories have had difficulty obtaining consistent results. Because resorption depth is apparently normal in osteoporosis but wall thickness is reduced, osteoblast work must be reduced. Parfitt et al. [35] have attributed reduced osteoblast work in osteoporosis to defective recruitment of osteoblast teams. 3. REPAIR OF MICRODAMAGE Certain changes in remodeling seen in osteoporotic patients imply inefficient or delayed repair of normally occurring microdamage. These changes might be associated with reduced mechanical competence of bone by allowing the accumulation of an excessive amount of unrepaired microdamage. a. Prolonged Remodeling Periods Prolonged remodeling periods might well be associated with inefficient or
delayed repair of microdamage. Accumulation of microdamage would probably compound skeletal weakness associated with low bone mass and/or deranged microstructure. Some patients with fragility fractures (and some normal persons, as well) do have prolonged remodeling periods combined with normal appositional rates [17]. This combination of findings results from excessive off time, that is, interruption of a remodeling cycle during bone formation; eventually the process resumes and the BMU completes its work. Local changes in remodeling may be even more important. For example, Eventov [38] examined biopsies obtained intraoperatively from the femoral neck of hip fracture patients; the numbers of osteoclasts and osteoblasts in these biopsies were extremely low compared to the numbers in iliac crest biopsies taken at the same time from the same patients. The paper concluded that slow remodeling rates had allowed unrepaired microdamage to accumulate at the hip, contributing to the hip fractures that later occurred. b. Inefficient Detection? It seems likely that the network of living osteocytes and their processes are involved in detection of microdamage. If so, loss of osteocytes or defects in their function could permit unrepaired microdamage to accumulate, undetected. Elucidation of the damage detection system, as well as identification of defects in that system, awaits further research. c. Suppression of Repair? A recent animal study [39] examines the question of whether suppression of bone turnover allows microdamage to accumulate. The investigators gave daily injections of etidronate (0.5 or 5 mg/kg) or vehicle to skeletally mature dogs for 12 months, with X-rays later in the treatment period and harvesting of bone samples from ribs, pelvis, femurs, and spine at its conclusion. At the lower dose, etidronate reduced trabecular bone turnover; at the high dose, activation frequency was zero in both cortical and trabecular bone. Fractures of ribs and/or thoracic spinous processes occurred in nearly all of the dogs in the high-dose group; however, microdamage (as numerical density of microcracks) was not greater at their sites of fracture. The investigators attributed the fractures that occurred to inhibition of mineralization (increased osteoid volume, especially in trabecular bone), rather than to accumulation of microdamage.
B. Findings in Rib Biopsies In comparative studies of rib biopsies from osteoporosis patients and normals, Frost [6,40] found clear, significant differences. In the osteoporotic specimens, appositional and bone formation rates were depressed, and remodeling
68
RECKER AND BARGER-LUX
periods were prolonged. These abnormalities were consistent with inhibited repair of microdamage in cortical bone [42], a situation that could be an important determinant of excess fragility in the appendicular skeleton. Because their cortical shells may be an important determinant of the compressive strength of whole vertebral bodies [37], microdamage in cortical bone could also contribute to excess vertebral fragility.
IV. BONE REMODELING AT THE WHOLE-ORGANISM LEVEL
B. Markers of Resorption and Formation Over the past decade, a number of biochemical markers of bone resorption and bone formation have been studied [e.g., 46,47]. Unlike histomorphometric studies, bone markers reflect remodeling at the whole-organism level; unlike bone kinetics studies, they do not require the extended presence of subjects or the in vivo use of radioactive materials. In a recent study [48], investigators measured the following bone markers in healthy pre- and postmenopausal women and in osteoporotic women before and during treatment with alendronate: Markers of bone resorptiona
Data that reflect bone remodeling at the level of the total skeleton complement the results of histomorphometric studies. Histomorphometric data come from small, local samples that cannot represent the average of all skeletal sites. Calcium kinetics studies and bone markers estimate remodeling rates for the skeleton as a whole, but they cannot provide information about the structural changes that remodeling activities effect.
A. Calcium Kinetics By following the disappearance of 45Ca given intravenously, resorption and accretion of bone mineral for the total skeleton can be calculated. The method requires diets matched to the habitual calcium intake of each subject and complete collection of excreta for an 8 to 12-day in-patient stay. In a recent study employing this method, Heaney and Draper [43] reported that, among 33 healthy women in early postmenopause, resorption exceeded accretion by a mean value of 66 mg/day. In 1968, Heaney [44] had reported results of calcium kinetics studies from a longitudinal, observational study of healthy women. They examined changes in accretion and resorption rates in the 115 subjects who were tested at entry and again about 5 years later and reported that: • When menopause occurred during the interval between studies, changes averaged 56 mg/day in accretion and 80 mg/day in resorption; however, • Changes among subjects who were postmenopausal for both tests were similar to those of subjects who were premenopausal for both tests. The latter finding is evidence that the process that produced the bone loss (here, menopause-related) could no longer be demonstrated. Other work published in 1978 by Heaney et al. [45] was also compatible with the same general conclusion; whereas variation in bone turnover was greater among women with established osteoporosis, their mean values were indistinguishable from those of normals.
CTx NTx fDpd tDpd
C-terminal telopeptide of type I collagen N-terminal cross-linked telopeptide free deoxypyridinoline total deoxypyridinoline
Markers of bone formationb bAP Oc PINP PICP
bone-specific alkaline phosphatase osteocalcin N-terminal propeptide of type I procollagen C-terminal propeptide of type I procollagen
a CTx can be measured in serum or urine; other resorption markers require urine specimens. b All these formation markers require serum specimens.
By measuring the above markers in specimens collected from the premenopausal group on 4 successive days, the investigators calculated both individual coefficient of variation (iCV) and least significant change (LSC) for each assay. The investigators assessed the usefulness of the above markers to classify and follow individual patients by use of these calculations. They reported higher iCVs for the resorption markers (14.6 to 22.3%) than the formation markers (7.2 to 14.4%), a finding that is probably associated with the specimen requirements of the assays. In 15 of 16 patients (one was a false positive) treated with alendronate, a drop in serum CTx (of at least its LSC) after 4 months of treatment correctly predicted a significant gain in BMD (i.e., at least 27 mg/cm2) at 12 months. This study identified serum CTx as “most effective of the markers tested” for longitudinal monitoring of individuals [48]. Many studies have used bone markers to examine differences between groups; however, this paper helps to resolve some uncertainty about the practical usefulness of bone markers to classify and follow individuals.
V. SUMMARY The present state of information allows us to conclude that: • Low bone mass in osteoporosis represents primarily
69
CHAPTER 39 Bone Remodeling Findings in Osteoporosis
a deficit of cortical bone. Bone is lost from the corticoendosteal surface, and eventually the inner cortex becomes completely trabeculated. • Important differences in the remodeling dynamics of trabecular bone between healthy subjects and patients with established osteoporosis are usually absent at the time of biopsy. However, a number of structural findings are of interest. • Trabecular bone mass is probably reduced in most women with osteoporosis, and trabecular thinning is relatively minor. However, osteoporotic patients with vertebral fractures are lower in trabecular connectivity than are others with the same trabecular bone mass. • It is unclear whether inefficient detection of microdamage contributes to fragility fractures. However locally inefficient repair of microdamage (e.g., markedly reduced numbers of osteoclasts and osteoblasts) may well be an important factor. The repair of microdamage is an important issue in both trabecular and cortical bone. • Calcium kinetics studies have shown clearly that, in association with menopause (i.e., estrogen deprivation), the relationship between bone resorption and bone accretion changes in favor of resorption. • Data that describe within-individual magnitude of change (both day-to-day and in response to treatment) permit evaluation of the utility of bone markers to classify and follow individuals.
References 1. Osteoporosis prevention, diagnosis, and therapy (March 27 – 29, 2000). NIH Consensus Statement Online 17(2), 1 – 36 (cited July 27, 2000). 2. E. Seeman, From density to structure: Growing up and growing old on the surfaces of bone. J. Bone Miner. Res. 12, 509 – 521 (1997). 3. M. P. Akhter, D. M. Raab, C. H. Turner, D. B. Kimmel, and R. R. Recker, Characterization of a non-invasive method of in vivo strain application. J. Biomech. 25, 1241 – 1246 (1992). 4. D. B. Burr, M. B Schaffler, K. H. Yang, M. Lukoschek, N. Sivaneri, J. D. Blaha and E. L. Radin, Skeletal change in response to altered strain environments: Is woven bone a response to elevated strain? Bone 10, 223 – 233 (1989). 5. C. T. Rubin and L. E. Lanyon, Regulation of bone mass by mechanical strain magnitude. Calcif. Tissue Int. 37, 411 – 417 (1985). 6. H. M. Frost, “Bone Remodeling and Its Relationship to Metabolic Bone Diseases,” Vol. 3. Thomas, Springfield, IL, 1973. 7. N. M. Keshawarz and R. R. Recker, Expansion of the medullary cavity at the expense of cortex in postmenopausal osteoporosis. Metab. Bone Dis. Relat. Res. 5, 223 – 228 (1984). 8. S. M. Garn, “The Earlier Gain and the Later Loss of Cortical Bone in Nutritional Perspective” Thomas, Springfield, IL, (1970). 9. H. M. Frost, “Intermediary Organization of the Skeleton,” Vols. 1 and 2 CRC Press, Boca Raton, FL, (1986). 10. H. M. Frost, Dynamics of bone remodeling. In “Bone Biodynamics” (H. M. Frost, ed.), pp. 315 – 333. Little, Brown, Boston, (1964). 11. R. R. Recker, D. B. Kimmel, A. M. Parfitt, K. M. Davies, N. Keshawarz, and S. Hinders, Static and tetracycline-based bone histomorphometric
12.
13.
14.
15.
16.
17.
18.
19. 20.
21.
22.
23.
24. 25.
26. 27.
28.
29.
30.
31.
data from 34 normal postmenopausal females. J. Bone Miner. Res. 3, 133 – 144 (1988). H. M. Frost, Analysis of trabecular bone dynamics. In “Bone Histomorphometry: Techniques and Interpretation” (R. R. Recker, ed.), pp. 109 – 142. CRC Press, Boca Raton, FL (1983). J. L. Ivey and D. J. Baylink, Postmenopausal osteoporosis: Proposed roles of defective coupling and estrogen deficiency. Metab. Bone Dis. Relat. Res. 3, 3 – 7 (1981). R. P. Heaney, The bone-remodeling transient: Implications for the interpretation of clinical studies of bone mass change. J. Bone Miner. Res. 9, 1515 – 1523 (1994). D. M. Black, S. R. Cummings, H. K. Genant, M. C. Nevitt, L. Palermo and W. Browner, Axial and appendicular bone density predict fractures in older women. J. Bone Miner. Res. 7, 633 – 638 (1992). S. Y. Nottestad, J. J. Baumel, D. B. Kimmel, R. R. Recker, and R. P. Heaney, The proportion of trabecular bone in human vertebrae. J. Bone Miner. Res. 2, 221 – 229 (1987). B. L. Riggs, H. W. Wahner, E. Seeman, K. P. Offord, W. L. Dunn, R. B. Mazess, K. A. Johnson, and L.G. Melton, III, Changes in bone mineral density of the proximal femur and spine with aging. J. Clin. Invest. 70, 716 – 723 (1982). R. P. Heaney, L. V. Avioli, C. H. Chesnut, J. Lappe, R. R. Recker, and G. H. Brandenburger, Osteoporotic bone fragility: Detection by ultrasound transmission velocity. J. Am. Med. Assoc. 261, 2986 – 2990 (1989). D. B. Kimmel, Unpublished observations (1993). C. C. Gluer and The International Quantitative Ultrasound Consensus Group, Quantitative ultrasound techniques for the assessment of osteoporosis: Expert agreement on current status. J. Bone Miner. Res. 8, 1280 – 1288 (1997). J. Nairus, S. Ahmadi, S. Baker, and D. Baran, Quantitative ultrasound: An indicator of osteoporosis in perimenopausal women. J. Clin. Densitom. 2, 141 – 147 (2000). J. E. Zerwekh, P. O. Antich, S. Mehta, K. Sakhaee, F. Gottschalk, and C. Y. Pak, Reflection ultrasound velocities and histomorphometric and connectivity analyses: Correlations and effect of slow release sodium fluoride. J. Bone Miner. Res. 12, 2068 – 2075 (1997). L. Mosekilde and L. Mosekilde, Iliac crest trabecular bone volume as predictor for vertebral compressive strength, ash density and trabecular bone volume in normal individuals. Bone 9, 195 – 199 (1988). R. R. Recker and R. P. Heaney, Letter to Editor. J. Bone Miner. Res. 5, 307 – 308 (1990). D. B. Kimmel, R. R. Recker, J. C. Gallagher, A. S. Vaswani, and J. F. Aloia, A comparison of iliac bone histomorphometric data in postmenopausal osteoporotic and normal subjects. Bone Miner. 11, 217 – 235 (1990). Z. F. G. Jaworski and H. K. Uhthoff, Reversibility of nontraumatic disuse osteoporosis during its active phase. Bone 7, 432 – 439 (1986). A. M. Parfitt, C. H. E. Mathews, A. R. Villanueva, M. Kleerekoper, B. Frame, and D. S. Rao, Relationships between surface, volume, and thickness of iliac trabecular bone in aging and in osteoporosis. J. Clin. Invest. 7, 1396 – 1409 (1983). M. E. Arlot, P. D. Delmas, D. Chappard, and P. J. Meunier, Trabecular and endocortical bone remodeling in postmenopausal osteoporosis: Comparison with normal postmenopausal women. Osteoporosis Int. 1, 41 – 49 (1990). T. Steiniche, P. Christiansen, A. Vesterby, C. Hasling, R. Ullerup, L. Mosekilde, and F. Melsen, Marked changes in iliac crest bone structure in postmenopausal osteoporotic patients without any signs of disturbed bone remodeling or balance. Bone 15, 73 – 79 (1994). A. Vesterby, Star volume of marrow space and trabeculae in iliac crest: Sampling procedure and correlation to star volume of first lumbar vertebra. Bone 11, 149 – 155 (1990). R. R. Recker, Architecture and vertebral fracture. Calcif. Tissue Int. 53, S139 – S142 (1993).
70 32. A. Vesterby, H. J. Gundersen, and F. Melsen, Star volume of marrow space and trabeculae of the first lumbar vertebra: Sampling efficiency and biological variation. Bone 10, 7 – 13 (1989). 33. J. E. Aaron, P. A. Shore, R. C. Shore, M. Beneton, and J. A. Kanis, Trabecular architecture in women and men of similar bone mass with and without vertebral fracture. II. Three-dimensional histology. Bone 27, 277 – 282 (2000). 34. L. Hordon, M. Raisi, J. E. Aaron, J. E. Paxton, S. K. Beneton, M. Beneton, and J. A. Kanis, Trabecular architecture in women and men of similar bone mass with and without vertebral fracture. I. Twodimensional histology. Bone 27, 271 – 276 (2000). 35. A. M. Parfitt, A. R. Villanueva, J. Foldes, and S. D. Rao, Relations between histologic indices of bone formation: Implications for the pathogenesis of spinal osteoporosis. J. Bone Miner. Res. 10, 466 – 473 (1995). 36. E. F. Eriksen, S. F. Hodgson, R. Eastell, S. L. Cedel, W. M. O’Fallon, and B. L. Riggs, Cancellous bone remodeling in type I (postmenopausal) osteoporosis: Quantitative assessment of rates of formation, resorption, and bone loss at tissue and cellular levels. J. Bone Miner. Res. 5, 311 – 319 (1990). 37. M. E. Cohen-Solal, M-S. Shih, M. W. Lundy, and A. M. Parfitt, A new method for measuring cancellous bone erosion depth: Application to the cellular mechanisms of bone loss in postmenopausal osteoporosis. J. Bone Miner. Res. 6, 1331 – 1338 (1991). 38. I. Eventov, B. Frisch, Z. Cohen, and I. Hammel, Osteopenia, hematopoiesis, and bone remodelling in iliac crest and femoral biopsies: A prospective study of 102 cases of femoral neck fractures. Bone 12, 1 – 6 (1991). 39. T. Hirano, C. H. Turner, M. R. Forwood, C. C. Johnston, and D. B. Burr, Does suppression of bone turnover impair mechanical proper-
RECKER AND BARGER-LUX
40.
41.
42.
43.
44. 45. 46.
47.
48.
ties by allowing microdamage accumulation? Bone 27, 13 – 20 (2000). S. Jett, K. Wu, and H. M. Frost, Tetracycline-based histological measurement of cortical-endosteal bone formation in normal and osteoporotic rib. Henry Ford Hosp. Med. J. 15, 325 – 344 (1967). H. M. Frost, Transient steady state phenomena in microdamage physiology: A proposed algorithm for lamellar bone. Calcif. Tissue Int. 44, 367 – 381 (1989). A. Vesterby, L. Mosekilde, H. J. Gundersen, F. Melsen, L. Mosekilde, and K. Holme, Biologically meaningful determinants of the in vitro strength of lumbar vertebrae. Bone 12, 219 – 224 (1991). R. P. Heaney and M. W. Draper, Raloxifene and estrogen: Comparative bone-remodeling kinetics. J. Clin. Endocrinol. Metab. 82, 3425 – 3429 (1997). R. P. Heaney, Kinetic studies of calcium in metabolic bone disease. Clin. Endocrinol. 2, 309 – 323 (1968). R. P. Heaney, R. R. Recker, and P. D. Saville, Menopausal changes in bone remodeling. J. Lab. Clin. Med. 92, 964 – 970 (1978). P. Garnero, W. J. Shih, E. Gineyts, D. B. Karpf, and P. D. Delmas, Comparison of new biochemical markers of bone turnover in late postmenopausal osteoporotic women in response to alendronate treatment. J. Clin. Endocrinol. Metab. 79, 1693 – 1700 (1994). R. Eastell, S. P. Robins, T. Colwell, A. M. A. Assiri, B. L. Riggs, and R. G. G. Russell, Evaluation of bone turnover in type I osteoporosis using biochemical markers specific for both bone formation and bone resorption. Osteoporosis Int. 3, 255 – 260 (1993). E. Fink, C. Cormier, P. Steinmetz, C. Kindermans, Y. Le Bouc, and J-C. Souberbielle, Differences in the capacity of several biochemical bone markers to assess high bone turnover in early menopause and response to alendronate therapy. Osteoporosis Int. 11, 295 – 303 (2000).
CHAPTER 40
The Role of Parathyroid Hormone and Vitamin D in the Pathogenesis of Osteoporosis JOHN P. BILEZIKIAN* AND SHONNI J. SILVERBERG† Departments of Medicine*† and Pharmacology,* College of Physicians and Surgeons, Columbia University, New York, New York 10032
I. Introduction II. Vitamin D and Osteoporosis III. Parathyroid Function in Osteoporosis
IV. Summary References
I. INTRODUCTION
relationship between bone resorption and formation [1]. Any imbalance favoring resorption over formation will lead to bone loss. The time line of this slow but inexorable process predicts that if a woman or man lives long enough, she or he will experience the disorder we call osteoporosis. Viewed in this context, osteoporosis is an intrinsic outcome of the aging process. Also intrinsic to the aging process are changes in the synthesis, metabolism, and responsiveness of vitamin D and parathyroid hormone. It is possible that the age-associated changes in vitamin D and parathyroid hormone are causally related to the age-associated changes in bone mass. On the other hand, some of the hormonal changes may be adaptive, serving to protect the aging skeleton from further weakening. Superimposed upon the age-related decline in bone mass is a set of other pathophysiological challenges, such as
Many factors contribute to the bone loss that characterizes the syndrome of osteoporosis. This chapter focuses primarily upon the two major calcium-regulating hormones, parathyroid hormone and vitamin D. Evidence has accumulated slowly in support of their potential involvement in the development of osteoporosis. This chapter reviews provocative literature that could accomodate hypotheses implicating these two hormones. Among the panopoly of functions that parathyroid hormone and vitamin D serve (see Chapters 7 and 9) none is more important than to maintain normal bone remodeling. After peak bone mass has been achieved and after a short period of neutral bone balance, most adult life is believed to be associated with a loosening of a rather tightly coupled
OSTEOPOROSIS, SECOND EDITION VOLUME 2
71
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
72
BILEZIKIAN AND SILVERBERG
estrogen deficiency, that add to the risk of osteoporosis in a given individual. Altered relationships between the calcium-regulating hormones and bone resulting from estrogen deficiency, for example, could lead to very different effects of these hormones on bone balance. Thus, the actions of parathyroid hormone on bone might be influenced by the presence or absence of estrogens. This chapter considers potential roles of vitamin D and parathyroid hormone in the context of the aging process per se and in association with selected other changes (e.g., estrogen deficiency). The classical metabolic bone diseases associated with gross deficiences or excesses of vitamin D or parathyroid hormone are not covered in this chapter.
II. VITAMIN D AND OSTEOPOROSIS Impaired calcium absorption from the gastrointestinal tract due to abnormalities at any point in the synthesis and metabolism of, or responsiveness to, vitamin D underlie the putative role of vitamin D in the development of osteoporosis. Possible abnormalities include vitamin D deficiency, abnormal production of 25-hydroxyvitamin D (25OHD) or 1,25-dihydroxyvitamin D (1,25(OH)2D), the active form of the vitamin, altered intestinal vitamin D receptor number or sensitivity, or acquired resistance to vitamin D (see Chapter 9). Individuals deficient in calcium as a result of these possible abnormalities will be at greater risk for an imbalance in the dynamic interplay between bone formation and bone resorption. The clinical consequences of these abnormalities in the context of osteoporosis differ from the classical disorder associated with overt vitamin D deficiency, namely osteomalacia. Most, but not all studies [2,3] report a fall in the circulating concentration of 1,25(OH)2D with advancing age [4 – 7]. Osteoporotic individuals have circulating 1,25(OH)2D levels that are even lower [4,8]. Reduced concentration of calcitriol thus is implicated as a cause of age-associated reductions in bone mass and of further reduction in bone mass in osteoporosis. Since 1,25(OH)2D concentrations reflect the availability of precursors and metabolic processes that lead to its formation and metabolism, a number of specific factors could account for reductions in circulating levels (Table 1).
A. Vitamin D Deficiency The early literature describing vitamin D deficiency came mainly from the United Kingdom where subclinical vitamin D deficiency has long been recognized to be a factor in the development of some forms of osteoporosis. In the past 25 years, reports have come from around the globe as well [9]. Inadequate supply of vitamin D from its two sources, diet and sunlight, leads to vitamin D deficiency. Vitamin D is
TABLE 1 Pathogenesis of Osteoporosis: Possible Roles of Vitamin D I. Vitamin D deficiency A. Dietary deficiency B. Insufficient ultraviolet B exposure II. Altered vitamin D metabolism A. Defect in the renal 1-hydroxylase B. Intestinal vitamin D receptor abnormality 1. Decreased receptor number 2. End organ resistance
derived from dietary sources and by supplementation in milk in the United States. It is not found as a supplement in other dairy products. Adult Americans obtain approximately 75 – 100 IU per day in their diets [10], a value that is below the former guidelines of 400 IU per day. Even this figure, which is generally not met, probably underestimates physiological requirements set at the current time by the U.S. Food and Nutrition Board to be 600 – 800 IU per day (11). In addition to possible reductions in dietary vitamin D intake with advancing years, its absorption from the gastrointestinal tract is affected by age. In women, a decline of up to 40% in absorption of vitamin D in the distal ileum has been reported to occur with advancing age [12]. Whether women who develop osteoporosis demonstrate even further reductions in dietary vitamin D absorption than expected for their age is not established (13 – 16). The other source of vitamin D is the skin, where ultraviolet B energy of 290 – 315 nm converts 7-dehydrocholesterol to previtamin D [17]. 7-Dehydrocholesterol levels in the skin fall by approximately 50% between 20 and 80 years of age [18]. It is not known whether the reduced quantity of skin substrate is exaggerated in those affected by osteoporosis. It is also not known whether the thermal-dependent conversion of previtamin D to vitamin D is impaired with aging or whether osteoporotic subjects are more deficient in this step. In addition to reduced levels of 7-dehydrocholesterol with aging, a comparison of vitamin D status from several global regions shows seasonal variation in concentrations of 25-hydroxyvitamin D. If dietary sources of and absorption of vitamin D do not vary substantially over a short period of time, from year to year, seasonal variations of 25-hydroxyvitamin D could reflect changes in UV exposure. In winter months at northern latitudes, decreased exposure of skin to UV sunlight of correct wavelength can lead to vitamin D insufficiency. It has long been assumed [19,20] that vitamin D deficiency was unique to individuals [21 – 26] living in northern latitudes in whom dietary sources and food supplements were inadequate. This was not considered to be common in the United States because of its generally lower latitude compared to countries where vitamin D deficiency has
CHAPTER 40 Role of Parathyroid Hormone and Vitamin D
been a problem and because of the fortification of milk with vitamin D. More recent studies have suggested that subclinical vitamin D deficiency may indeed play a role in bone loss in postmenopausal women in the United States. In Maine (45.5ø northern latitude) and Boston, Massachusetts (42.2ø northern latitude), no synthesis of previtamin D occurs in the skin during the late fall and winter months [27,28]. Bone density also exhibited a steep decline during those months, rising again during the summer [28]. The rise in bone density was associated with resolution of an accompanying secondary hyperparathyroidism. It did not, however, compensate for the bone loss in winter. Subclinical vitamin D deficiency has been shown in subpopulations of osteoporotic American women. Vallareal et al. reported low concentrations (less than 38 nmol/liter) of 25hydroxyvitamin D in 9% (49/539) of women referred for osteoporosis screening in St. Louis, Missouri [29]. These women had no symptoms of vitamin D deficiency, but they did have lower mean vertebral bone mineral density, lower serum calcium and phosphorus concentrations, and higher circulating parathyroid hormone and urinary calcium excretion than vitamin D-sufficient patients. Multivariate analysis suggested that the secondary hyperparathyroidism induced by the vitamin D-deficient state was a predominant factor leading to low bone density. Studies that have attempted to implicate subclinical vitamin D deficiency in the development of osteoporosis have depended upon measurement of circulating 25-hydroxyvitamin D. Since this is believed to be an accurate representation of the storage form of the vitamin, such measurements are useful. However, Peacock et al. suggested that a 25hydroxyvitamin D value of 50 nmol/L may be necessary to ensure sufficient 1,25(OH)2D production [30]. Thus, the standard published values for the “normal range” (25 – 125 nmol/L or 10 – 50 ng/mL) includes individuals who should really be regarded as having subnormal levels of 25hydroxyvitamin D. In an older population entering a New York nursing home, 86% had 25-hydroxyvitamin D levels below 50 nmol/L. In 41% of the group, 25-hydroxyvitamin D concentrations were below 25 nmol/L [31]. The bone densities among subjects with lower levels of 25hydroxyvitamin D were below the mean for age- and gender-matched controls. Gloth et al. also reported that vitamin D deficiency is found commonly in the elderly with estimates ranging from 38 to 54% (32,33), and Thomas reported that among an unselected free-living population living in the metropolitan area of Boston, over 50% had evidence for vitamin D deficiency (34).
B. Altered Vitamin D Metabolism There is no evidence that the liver loses its rather impressive capacity to convert vitamin D to 25-hydroxyvitamin D
73 with aging or in osteoporosis. When deficiencies in 25hydroxyvitamin D are observed, they are due either to limited availability of the substrate, vitamin D, to drug-related altered vitamin D metabolism [35,36], or to severe, advanced liver disease. One exception in this regard is primary biliary cirrhosis, a chronic liver disease in which there appears to be an impaired 25-hydroxylating ability that is disproportionately greater than the loss of functioning hepatic mass [37]. In contrast to the rather limitless capacity of the liver to form 25-hydroxyvitamin D, the kidney controls the formation of the active metabolite, 1,25(OH)2D, exquisitely well. Even in the setting of vitamin D toxicity, the kidney effectively controls excessive production of this active metabolite [38]. Factors that help to regulate the renal enzyme responsible for the formation of 1,25(OH)2D are parathyroid hormone, phosphorus, calcium, and 1,25(OH)2D itself (see Chapter 9). A decline in the ability of the kidney to form this active metabolite occurs with age and has been implicated as a possible mechanism for age-related osteoporosis [39]. Slovik et al. demonstrated that older individuals respond to the stimulating effects of parathyroid hormone on 1,25(OH)2D production less well than young normal controls [40]. This landmark study did not compare agematched subjects with and without osteoporosis and thus could not implicate a specific defect in osteoporosis. Nevertheless, the observations did support the notion that renal responsiveness to the stimulatory effects of parathyroid hormone is reduced with age. A possible specific defect in renal hydroxylating capacity was studied by Riggs et al., who infused 400 IU of parathyroid hormone on 3 consecutive days to osteoporotic and age-matched nonosteoporotic subjects. The results showed no differences between the two groups in the 1,25(OH)2D response to parathyroid hormone infusion [41]. Tsai et al. extended these observations by studying four different groups of women in a protocol that tested the possibility of an age-related decline in renal hydroxylating capability as well as a specific defect in osteoporotic individuals [7]. Older subjects had lower baseline 1,25(OH)2D concentrations and responded less well to the stimulatory effects of parathyroid hormone on formation. The data correlated best with diminishing responsiveness to parathyroid hormone as a function of declining renal function. Osteoporotic women responded less well than the age-matched nonosteoporotic group but their renal function was also substantially lower. Thus, it was not possible to conclude that the osteoporotic women had a significantly more sluggish response to parathyroid hormone than their nonosteoporotic counterparts. These studies support the generally accepted notion that the renal capacity to form 1,25(OH)2D diminishes with age. They do not support the idea that in osteoporotic subjects this age-related decline in 1—hydroxylating capacity is further compromised.
74
BILEZIKIAN AND SILVERBERG
The decline in renal hydroxylating capacity associated with advancing age could be due to the aging process per se or to declines in renal function. Halloran et al. [42] elegantly addressed this question, at least in men, by showing that the impaired responsiveness of the renal enzyme is due exclusively to a decline in renal function. Only in the sense that advancing age and declining renal function are mutually associated can the decline in renal enzymatic activity be said to be a function of age. In Halloran’s study of healthy 70-year-old men with completely normal endogenous creatinine clearance, the ability of the kidney to convert 25-hydroxyvitamin D to 1,25(OH)2D was normal. It can be concluded, therefore, that one pathophysiologic mechanism that accounts for reduced synthesis of 1,25(OH)2D is an age-related decline in renal function. One could advance the idea that those individuals who are able to maintain renal responsiveness do so either by being fortunate enough to maintain normal renal function or, more likely, by developing mechanisms to accomodate this agerelated decline (see below).
C. Altered Vitamin D Sensitivity In addition to possible changes in the quantity of vitamin D precursors and the capacity to generate the active metabolite of vitamin D, other hypotheses related to the role of vitamin D in the pathogenesis of osteoporosis focus on the reduced sensitivity to 1,25(OH)2D in the small intestine. Such Intestinal resistance is suggested to be primary alteration [2]. One theory proposes that osteoporosis is associated with an acquired alteration in binding of 1,25(OH)2D to its receptors in the small intestine [43]. Another attributes the apparent reduction in intestinal sensitivity to an age-related reduction in intestinal vitamin D receptor concentration, which has been observed in intestinal biopsies obtained from women spanning the wide age range of 20 – 87 years [44]. If vitamin D receptor concentrations are limiting, lower receptor abundance should lead to impaired calcium absorption. Presumably, this mechanism could not be satisfactorily overcome by increased calcium intake. In support of this hypothesis, Gennari et al. have shown impaired 1,25(OH)2D stimulation of fractional calcium absorption [45]. More recently, Pattanaungkul et al. revisited this issue in a protocol that correlated 1,25(OH)2D concentrations in young and elderly subjects with fractional fasting calcium absorption [46]. In the young subjects (mean age 28.7 5.3 years), fractional calcium absorption correlated significantly with the free 1,25(OH)2D index (r 0.63, P 0.01). In the elderly group, however (mean age 72.5 3.0 years), fractional calcium absorption was not significantly related to the free 1,25(OH)2D level. Moreover, the slopes of the relationship between these two indices were significantly greater in the young as than in
the elderly. These observations, therefore, support the idea that the elderly are relatively resistant to the physiological actions of 1,25(OH)2D on intestinal calcium absorption.
D. Genetic Variations in the Vitamin D Receptor (VDR) Genetic polymorphisms of VDR have been implicated in the development of osteoporosis [47] (see Chapter 26). In several studies a particular allele of the vitamin D receptor, as characterized by several restriction enzyme patterns, was associated with reduced bone mass [47,48]. Homozygous carriers of the allele designated “B” were found to have lower bone mass than homozygous carriers of the allele designated “b”. These studies remain controversial because they have not been confirmed in other investigations (49). Moreover, it has been very difficult to “link” identified VDR polymorphisms to a physiological abnormality in a system regulated by 1,25(OH)2D. These studies are nevertheless important because a predisposition to osteoporosis on the basis of vitamin D genetics may still be confirmed. More important, however, is the fact that these studies literally spawned an explosive search for genes that might contribute to osteoporosis. It is clear from these other studies that osteoporosis is a complex genetic trait, with the likelihood that a number of genes are going to be shown to be important in the overall disposition to this disease (50).
E. Overall View of Mechanisms of Vitamin D Alterations in Osteoporosis The importance of the various mechanisms by which alterations in the vitamin D system can lead to osteoporosis is unclear at this time. Any or all of the aforementioned potential pathophysiologic considerations could contribute to bone loss with aging and could be further deranged in osteoporosis. Hypotheses relating alterations in vitamin D to osteoporosis are based upon the idea that calcium absorption is ultimately impaired. There are other potential roles for vitamin D in bone health that could be important in the development of osteoporosis. If vitamin D is ultimately demonstrated to be important in the events leading to osteoporosis, its role is unlikely to be similar to its role in the events leading to osteomalacia, the principal clinical result of gross vitamin D deficiency. Alternatively, abnormalities in the vitamin D system, some of which have been demonstrated to occur with aging, may not be causative in the development of osteoporosis. Rather, these alterations in the vitamin D system could require compensatory adaptations in other systems, such as those related to parathyroid hormone. Osteoporosis could
75
CHAPTER 40 Role of Parathyroid Hormone and Vitamin D
result if the adaptive responses are inadequate, or if they themselves promote increased bone turnover. One possible compensatory mechanism is parathyroid hormone responsiveness, a subject that is covered in the next section of this chapter.
III. PARATHYROID FUNCTION IN OSTEOPOROSIS There is now general agreement that parathyroid function increases with advancing age. Initially this view was somewhat controversial due, in part, to the fact that declining renal function with age leads to an expected secondary increase in circulating parathyroid hormone. These secondary increases stemming directly from declining renal function are believed to be caused by several factors. Declining 1,25(OH)2D concentration as a function of impaired renal hydroxylating ability (see above) would relieve the inhibitory effects of this metabolite on the parathyroid hormone gene. They would also lead to a reduction in calcium absorption. Second, a very small but physiologically significant reduction in serum calcium concentration would be associated with slight increases in those of phosphorus resulting from declining renal function. These factors together would stimulate parathyroid hormone secretion. Although this view is entirely reasonable, it was hard to prove that parathyroid hormone concentrates actually increase because early generation assays for the hormone detected primarily circulating inactive hormone fragments. Such fragments normally accumulate when renal function declines. Thus, the increases in detectable parathyroid hormone may not have been due to the metabolic sequences described above, but merely to the reduced clearance of inactive hormone fragments. Immunoradiometric (IRMA) and immunochemiluminometric (ICMA) assays, which selectively measure circulating intact parathyroid hormone, as well as a large, inactive fragment [51] have helped to settle the point [52]. Biologically active concentrates of parathyroid hormone increase normally with age. Although this increase clearly reflects declining renal function, elevated values also occur even in older individuals who have no apparent decline in renal function. This point is illustrated well by Halloran et al. [42], who studied young and elderly men with normal renal function. Despite normal serum ionized calcium activity, serum 1,25(OH)2D, and urinary calcium excretion, basal serum parathyroid hormone was 1.5-fold higher in the elderly men than in the younger men. The increase in parathyroid hormone may be associated with a similar rise in bone turnover as assessed both by bone markers (53 – 56) and by histomorphometric indices (56 – 58). On the other hand, Gallagher et al. [59] could not show any increase in parathyroid hormone with age in a study of over 700 subjects.
FIGURE 1
Serum parathyroid hormone concentrations as a function of age among an age-stratified sample in Rochester, Minnesota. Men (solid line, squares) and women (dashed line, circles) both show an increase with age. The correlation with age was 0.30 for men and for women (P 0.001). Adapted from Khosla et al. [60] with permission.
Most studies have demonstrated the increases in serum parathyroid hormone concentration as a continuous relationship with age (60; Fig. 1) but more detailed analysis has shown that the major increase in circulating parathyroid hormone occurs in women greater than 70 years old. In fact, Koh et al. [61] showed a 1.9% decline per decade between 55 and 69 years. Between 70 and 81 years, there was a 3.8% increase per decade. Similarly, Prince et al. [62] showed that the age-related increase in parathyroid hormone becomes evident about 20 years after the menopause (Fig. 2), approximately 70 years of age. There are two fundamental hypotheses to explain the increase in parathyroid hormone concentrations with age. One links, in a causal way, the increase in parathyroid hormone to age-related bone loss (Table 2). An alternative hypothesis links the increase in hormone concentration to protection against age-related bone loss. This section will review the data in support of each of these hypotheses.
FIGURE 2
Serum parathyroid hormone concentrations in women as a function of time since menopause. Results with different letters are significantly different (P 0.05). Adapted from Prince et al. [62] with permission.
76
BILEZIKIAN AND SILVERBERG
TABLE 2 Changes in Serum PTH and in Biochemical Markers of Bone Turnover in 304 Women Residents of Rochester, MN, from the Third into the Tenth Decade of Life Spearman correlation coefficients
Increase with age (%)
vs age
vs PTH
PTH
54
0.354*
1.00
BSAP
38
0.329*
0.192†
OC
64
0.392*
0.206†
76
*
0.203†
*
0.190†
Variable
fPYD NTx
86
0.505
0.344
Note. BSAP, bone specific alkaline phosphatase; OC, osteocalcin; fPYD, urinary free pyridinoline; NTx, urinary N-telopeptide of type I collagen. Adapted from Riggs et al. [76] with permission. * P 0.0001. † P 0.001.
A. Parathyroid Hormone as a Contributing Factor to Osteoporosis 1. ALTERED PARATHYROID HORMONE SENSITIVITY The accelerated rate of bone loss in the early postmenopausal years has been explained by the local release of bone-resorbing cytokines [63,64]. Riggs and Melton [65] proposed that such local factors and the ensuing rapid loss of skeletal calcium could actually lead to suppression of parathyroid hormone. Reduced parathyroid hormone concentrations in osteoporotic women in the face of higher bone turnover could reflect enhanced skeletal sensitivity [66,67] to these local bone-resorbing factors. Although some studies have shown parathyroid hormone suppression in the early postmenopausal years, more often parathyroid hormone concentrations have not differed from those of age-matched controls. Despite normal circulating concentrations, parathyroid hormone could still induce bone loss if sensitivity is heightened in the postmenopausal state. One point in support of this idea comes from the literature on primary hyperparathyroidism, a clinical disorder seen often in women during the first decade after the menopause [68]. It seems that the hyperparathyroid process is unmasked in these women as a function of declining estrogen status. The reduced estrogen levels could also enhance skeletal sensitivity to parathyroid hormone. However, there are no prospective data showing that in these women, primary hyperparathyroidism existed in a subclinical form before menopause. When postmenopausal women with primary hyperparathyroidism are given estrogen, it is generally accepted that calcium concentrations decline somewhat without any change in the circulating concentration of parathyroid
hormone [69,70]. In the setting of the normocalcemic individual, however, McKane et al. [71] have demonstrated that within 6 months of estrogen replacement therapy in early postmenopausal women, and despite no change in circulating serum calcium concentration, parathyroid hormone concentrations increase by 38%. In the aggregate, these observations argue for an inhibitory effect of estrogen on parathyroid hormone action in bone, for enhanced sensitivity to the skeletal effects of parathyroid hormone when women become estrogen deficient, and for a physiological need for increased circulating parathyroid hormone when estrogen is provided to the early postmenopausal woman. The notion of enhanced skeletal sensitivity to parathyroid hormone is not new. Since the mid-1960s, this mechanism has been considered a possible explanation for the development of postmenopausal osteoporosis [72 – 74]. Kotowicz et al. [67] obtained histomorphometric data suggesting that in postmenopausal osteoporosis, the resorptive effects of parathyroid hormone are enhanced. For each picomole per liter rise in circulating parathyroid hormone, osteoporotic women had higher activation frequency (1.3% per year), bone resorption rate (3.9% per year), and cancellous bone loss (2.8% per year). On the other hand, enhanced sensitivity to parathyroid hormone has not been universally demonstrated. Tsai et al. [75] reported no difference between osteoporotic and normal women in urinary excretion of calcium or hydroxyproline in response to 400 units per day of bovine parathyroid hormone. Interpretation of the studies by Tsai et al. was limited by the pharmacologic amounts of hormone used and the relative insensitivity of the markers that were used to monitor the effect. Nevertheless, Ebeling et al. [66] reported similar results when calcium deprivation was used to stimulate endogenous parathyroid hormone secretion. The concept of parathyroid hormone as a pathophysiological culprit in postmenopausal osteoporosis has been promulgated even more boldly by Riggs and associates [76]. They recently proposed a “unitary model for involutional osteoporosis” according to which not only is parathyroid hormone involved in the early postmenopausal effects of estrogen deficiency (a relative suppression) but increases in parathyroid hormone also become important later in the setting of age-related bone loss. They proposed that in the second phase of bone loss (type II or age-related bone loss), the actions of estrogen deficiency predominate in nonskeletal sites, such as the intestine and kidney. Such extraskeletal actions would promote increased urinary calcium loss and reduce gut calcium absorption. The resulting secondary increase in parathyroid hormone is key in this explanation of the continued loss of bone mass with aging. Although Riggs et al. cite a number of studies in support of this hypothesis [77 – 80], our knowledge is admittedly still limited. We know little of the extraskeletal actions of estrogens with
CHAPTER 40 Role of Parathyroid Hormone and Vitamin D
regard to renal calcium handling and to gastrointestinal absorption of calcium. Also vexing is the hypothesis that only late-term estrogen deficiency is associated with an increase in parathyroid hormone. If one can account for these two separate actions of estrogens (skeletal vs extraskeletal] that are dichotomous in time, the secondary increase in parathyroid hormone occurring in this later phase would help to account for the more apparent cortical bone loss and an amelioration of the earlier rapid cancellous bone loss. Such observations gain support from data on primary hyperparathyroidism in which increased parathyroid hormone concentrations are associated with preservation of cancellous bone at the expense of cortical bone [81]. In fact, this and other observations have led many investigators to consider parathyroid hormone a potential therapy for postmenopausal osteoporosis [82] (see Chapter 77). Other confounding elements to the hypothesis put forth by Riggs et al. are noted by Bilezikian [83]. Nevertheless, circumstantial evidence does implicate parathyroid hormone in the pathogenesis of age-related osteoporosis. Positive correlations have been made between increases in parathyroid hormone and markers of bone turnover in elderly women [84,85]. As shown in Table 2, these correlations are modest, at best, with correlation coefficients generally in the neighborhood of 0.20, well below the more robust association of bone markers to age, per se. In support of a causal relationship, Ledger et al. showed that urinary collagen N-telopeptide excretion is suppressed to the young normal range after a 24-h calcium infusion [86]. Further evidence that argues for a negative effect of rising parathyroid hormone concentration with age comes from attempts to correlate parathyroid hormone with bone loss. Using peripheral quantitative computed tomography to distinguish cancellous from cortical elements [87], Boonen et al. showed a negative correlation between cortical bone loss and rising parathyroid hormone levels. Ledger et al. also showed that elevated parathyroid hormone concentrations in the elderly can be reduced to levels seen in young normals by administration of 1,25(OH)2D [88]. While these data in the aggregate argue for a role for parathyroid hormone in the pathogenesis of age-related osteoporosis, their indirect nature argues for caution in establishing a causal link at this time.
B. Parathyroid Hormone as a Protective Influence 1. ALTERED PARATHYROID HORMONE RESPONSIVENESS The observations reviewed above are consistent with a negative effect of parathyroid hormone on bone. Increased normal secretion or enhanced sensitivity to parathyroid
77 hormone could promote to bone loss on this basis. In contrast, it is possible that parathyroid hormone responsiveness is important to maintain bone health and that in osteoporosis this responsiveness is lost. In this context, which presents a completely different view, reduced responsiveness of the parathyroid glands contributes to the development of osteoporosis. Altered responsiveness could underlie changes in circadian rhythmicity of parathyroid hormone secretion. Daily parathyroid hormone secretion follows a biphasic profile with peaks at approximately 1800 and 0200 hours [89 – 91]. Presumably, the larger nocturnal peak represents a compsensation for mild hypocalcemia induced by night-time fasting. Calvo et al. reported that women exhibited a blunted parathyroid hormone peak concentration relative to that of men and, subsequently, a less dramatic decline in night-time urinary calcium excretion. Night-time urinary calcium excretion declined in men by 34%, whereas in women, it decreased by only 17%. This nocturnal calcium wasting could, over the years, contribute to the greater bone loss seen in women. Postmenopausal osteoporotic women showed a further blunting of their nocturnal parathyroid hormone peak, with no change in nocturnal fractional excretion of calcium [92]. The inefficient renal calcium conservation thus documented could contribute to the osteoporotic process. The cause of this blunted parathyroid hormone response to nocturnal fasting is unknown. More sophisticated pulsatility studies by Prank and colleagues [93] have shown that osteoporotic women demonstrate poorly predictable time series of pulses and patterns of parathyroid hormone secretion. Creating a discriminating statistic by fitting a time series model to pooled data from normal subjects, normal and osteoporotic subjects could be distinguished from each other [93,94]. In contrast, Samuels and colleagues [95] could not demonstrate any differences in amplitude or frequency, or pulsatile parathyroid hormone secretory parameters between osteoporotic and normal subjects. The lack of a difference was not influenced by the presence of estrogens. Further evidence for abnormalities in parathyroid hormone secretion in osteoporosis comes from the work of Silverberg et al. [96,97]. These studies were based on the premise that a mild hypocalcemic challenge should lead to age-appropriate increases in parathyroid hormone concentration. Oral phosphate was used to induce the hypocalcemic challenge. The first studies were conducted with two distinct groups of younger and older subjects who had no evidence for osteoporosis. In both cases, the serum phosphorus concentration rose and the serum calcium level fell to the same extent. Young subjects showed a 43% increase in parathyroid hormone concentration over baseline values, whereas older women showed a much more exuberant response to the same hypocalcemic stimulus, with a 2.5-fold increase over baseline levels. This protocol set up two
78 opposing stimuli with respect to 1,25(OH)2D phosphorus as an inhibitor and parathyroid hormone as a stimulus. In both cases, the opposing regulators were neutralized and 1,25(OH)2D concentration did not change. These data were interpreted to suggest that older, normal subjects require more parathyroid hormone for a given hypocalcemic challenge to maintain 1,25 (OH)2D status. Such a formulation is consistent with the reduced renal capacity to form this metabolite with age. It is also possible that the aging skeleton requires a greater amount of parathyroid hormone to achieve effects that are seen at lower levels in younger subjects. The same protocol was applied to a group of postmenopausal women with osteoporosis [97]. After phosphate administration, these women experienced the same increase in serum phosphorus concentration and the same reduction in serum calcium concentration that was observed for the young subjects and the age-matched older women. In contrast to the marked increase in parathyroid hormone in their age-matched counterparts, the osteoporotic women demonstrated only a modest 43% increase (Fig. 3). Although this was sufficient in younger individuals to prevent the inhibitory effects of phosphorus on 1,25(OH)2D production, it did not suffice in these osteoporotic women as 1,25(OH)2D concentrations fell by 50% (Fig. 4). These observations are consistent with the presence of an abnormality in parathyroid secretory function in osteoporosis. Osteoporotic women thus have both reduced ability to form 1,25(OH)2D and a superimposed deficiency in parathyroid responsiveness. The need for more parathyroid hormone with age could be achieved by altering the calcium set point. For any given serum calcium concentration, the parathyroid hormone concentration is higher in the elderly. This could account for the age-related increase in parathyroid hormone concentration in the absence of any change in circulating calcium.
BILEZIKIAN AND SILVERBERG
FIGURE 4 The effect of a hypocalcemic stimulus on the parathyroid hormone response (Fig. 3) and on 1,25(OH)2 D concentrations in osteoporotic and nonosteoporotic postmenopausal women and in young normal subjects. The details of the experimental protocol are given in the text (reprinted from [97] with permission). The means in patients with osteoporosis are represented by hatched bars, those in age-matched controls by open bars, and those in a comparison group of young, healthy subjects by solid bars. An asterisk denotes a significant change from baseline (P 0.05).
When Ledger et al. [88] studied this point with a provocative challenge, no age-related increase in the set point for parathyroid hormone secretion could be demonstrated. When postmenopausal women with osteoporosis were studied, however, differences did emerge. Cosman et al. [98] used infusions of the synthetic peptide, human parathyroid hormone(1 – 34) to assess suppressibility of endogenous parathyroid hormone secretion. It was possible to distinguish between exogenous human PTH (1 – 34) and endogenous human parathyroid hormone (1 – 84) by use of the immunoradiometric assay, which does not detect the 1 – 34 exogenously administered peptide. The data were consistent with a higher calcium set point in osteoporotic women. Similarly, Portale et al. showed in elderly men that the set point of parathyroid hormone responsiveness to calcium was “shifted” to the right [99]. Such results are consistent with a protective effect of parathyroid hormone in the pathogenesis of osteoporosis. 2. RACIAL DIFFERENCES IN THE PARATHYROID HORMONE–VITAMIN D AXIS
FIGURE 3 The effect of a hypocalcemic stimulus on the parathyroid hormone response in osteoporotic and nonosteoporotic postmenopausal women and in young normal subjects. The details of the experimental protocol are given in the text. Reprinted from [97] with permission.
The major difference in bone mass between Caucasian and Black individuals has provided another opportunity to assess whether or not increased exposure to parathyroid hormone preserves bone. Black individuals enjoy approximately 10% higher bone mass than Caucasians throughout adult life. Moreover, fracture incidence is lower in Blacks [100,101] (see Chapter 22). Higher peak bone mass is certainly a major reason for the relatively protected skeleton in Black individuals. The observation by Modlin and Bell et al. [102,103] that Blacks have lower urinary
CHAPTER 40 Role of Parathyroid Hormone and Vitamin D
calcium excretion than Caucasians has led to additional ethnic comparisons of the parathyroid – vitamin D axis. Although there is uniform agreement that urinary calcium excretion is lower in Blacks than in Whites, variable results have been reported concerning the hormones of mineral metabolism. Most, but not all, investigators have reported increased concentrations of parathyroid hormone, 1,25(OH)2D, or both in Black subjects. Once again, comparisons have been difficult due to use in some studies of parathyroid hormone assays that do not accurately reflect biologically active material [103 – 106]. Similarly, most investigators have reported that Blacks have lower serum levels of 25-hydroxyvitamin D. A prevailing hypothesis to explain this observation [105,106] is based on the fact that the increased skin pigment in Blacks leads to decreased dermal vitamin D synthesis [107,108]. This, in turn, leads to a secondary hyperparathyroidism with a resulting increase in 1,25(OH)2D concentration. The actions of the secondary hyperparathyroid state on renal tubular function leads to urinary calcium conservation. Bell postulated that increased 1,25(OH)2D and decreased urinary calcium could account for higher bone density observed in black individuals [105]. There is no difference between Black and White subjects with respect to calcium absorption from the gastrointestinal tract. These observations tend to support the idea that parathyroid hormone contributes to bone mass in Black individuals and is consistent with ideas about the anabolic properties of this hormone. Clearly, the mechanisms leading to greater bone mass in Blacks are more complex than differences in ambient concentrations of calcium-regulating hormones. For example, many studies have confirmed that bone turnover as assessed by histomorphometric analysis is reduced in Blacks. Osteocalcin [104 – 106], urinary hydroxyproline [104], and bonespecific alkaline phosphatase concentrations [104] are lower than in Caucasians, raising the possibility that high bone density in the face of increased parathyroid hormone and reduced bone turnover could reflect skeletal parathyroid hormone insensitivity [109]. Recent data in support of skeletal resistance to parathyroid hormone in Blacks come from El-Haji Fuleihan et al., who assessed responsiveness to hypo- and hypercalcemia [106]. They demonstrated higher maximal and minimal parathyroid hormone responses in Black subjects with no alteration in the set point or slope of the calcium – parathyroid hormone curve. However, just as baseline parathyroid hormone concentrations were higher and those of osteocalcin were lower in Blacks, hypocalcemia led to a less exuberant rise in osteocalcin in Blacks despite a vigorous rise in circulating parathyroid hormone. Although the differences in bone mass and susceptibility to osteoporosis among Blacks and Whites are incompletely understood, the possible role of the parathyroid hormone – vitamin D axis certainly deserves further investigation. A
79 recent report demonstrating by QCT analysis 40% more cancellous bone density at the lumbar spine in American Blacks than in Caucasians raises questions about racial differences in selected types of bone [104]. If, in fact, differences in the parathyroid – vitamin D axis underlie the relative protection from bone loss seen in Black subjects, differential effects in cancellous and cortical bone would not be surprising.
C. Parathyroid Hormone: A Positive or Negative Factor in Age-Related Bone Loss? The data reviewed in this chapter argue that parathyroid hormone can be viewed as either a negative or a positive factor in preservation of the postmenopausal skeleton. The age-related increase in parathyroid hormone may be adaptive or maladaptive. More information will be needed to sort out these diametrically opposite views. However, we already have abundant information about the skeleton in the classic condition of parathyroid hormone excess, primary hyperparathyroidism. In this disorder, excess parathyroid hormone leads to relative protection against bone loss in the lumbar spine [110]. By histomorphometric analysis, static, dynamic, and structural indices of the cancellous skeleton reveal maintenance of bone mass [111,112] and are contrasted dramatically with such indices in age-matched postmenopausal women with osteoporosis [113]. Moreover, longitudinal data confirm that the protective effect persists over time [114]. The data in primary hyperparathyroidism differ so strikingly from those in osteoporosis that one must conclude that the bone diseases of hyperparathyroidism and osteoporosis reflect two completely different disorders. In view of the protection accorded the very site that is at early risk for postmenopausal bone loss, namely the cancellous bone of the vertebral spine, parathyroid hormone’s actions on the aging skeleton are best viewed as protective, not deleterious. The hormone’s apparent anabolic effects on skeletal sites at risk in postmenopausal osteoporosis has led to studies designed to use it as a therapeutic agent. Clinical trials in postmenopausal women and in men have confirmed the hypothesis that parathyroid hormone increases bone mineral density of the lumbar spine [115,116] (see Chapter 77). Studies by Neer [117], Lindsay [118], Fujita [119], Kurland [120], Roe [121], Lane [122] and their colleagues have shown impressive effects of parathyroid hormone to increase bone density of the lumbar spine and hip without deleterious effects on cortical bone (Fig. 5). In other situations such as the bone loss due to the use of GnRH agonist [123] and when PTH is used as a combination therapy with bisphosphonate [124], parathyroid hormone has shown impressive anabolic potential. It is of interest that these increases are, in general, associated with enhanced bone turnover during the first 6 to 12 months of therapy.
80
BILEZIKIAN AND SILVERBERG
IV. SUMMARY This chapter has reviewed the evidence for different views of the parathyroid — vitamin D axis in postmenopausal osteoporosis. It is clear that age-related changes occur in this system, and it is likely that, to a certain extent, such changes are beneficial to an older individual who requires a greater amount of parathyroid hormone than is needed to achieve the same effect that can be reached at lower levels in youth. It is also evident that either insufficient or overexuberant adaptation could exert negative effects on the skeleton. Many other factors play a role in the common phenotype that we recognize as postmenopausal osteoporosis, and they are covered elsewhere in this book.
References
FIGURE 5
Changes in bone density after administration of parathyroid hormone (1 – 34) to men with idiopathic osteoporosis. The group receiving PTH is shown by the dark symbols; those receiving placebo are shown by the open symbols. The data are shown as percentage changes from baseline SEM for (A) lumbar spine, (B) femoral neck, and (C) distal (1/3) radius. Significant between group comparisons are indicated by the asterisk (P 0.05) and within group comparisons are indicated by the symbol (P 0.005). Reprinted from Kurland et al. (120) with permission.
The negative side of this hypothesis is that parathyroid hormone can have deleterious effects on cortical bone. In the mild form of hyperparathyroidism seen today, such demonstrable reductions are not severe [114,124]. In more severe forms of primary hyperparathyroidism, cortical bone loss is obvious. At some point in the spectrum of parathyroid hormone excess, markedly elevated hormone concentrations will lead to bone loss from the vertebral spine [125].
1. S. M. Ott, Bone remodeling and mineral homeostasis: Theoretical and methodological approach. In “Principles of Bone Biology” (J. P. Bilezikian, L. G. Raisz, and G. A. Rodan, eds), pp. 231 – 242. Academic Press, San Diego, 1996. 2. R. Eastell, A. L. Yergey, N. E. Viera, S. L. Cedel, R. Kumar, and B. L. Riggs, Interrelationship among vitamin D metabolism, true calcium absorption, parathyroid function and age in women: Evidence of an age-related intetinal resistance to 1,25-dihydroxyvitamin D action. J. Bone Miner. Res 6, 125 – 132, (1991). 3. S. Epstein, G. Bryce, J. Hinman, O. N. Miller, B. L. Riggs, S. L. Hui, and C. C. Johnston, Jr., The effect of age on bone mineral regulating hormones. Bone 7, 421 – 425 (1986). 4. J. Gallagher, B. Riggs, J. Eisman, A. Hamstra, S. Arnaud, and H. DeLuca, Intestinal calcium absorption and serum vitamin D metabolites in normal subjects and osteoporotic patients. J. Clin. Invest. 66, 729 (1979). 5. L. Avioli, J. McDonald, and S. Lee, The influence of age on the intestinal absorption of 47ca in women and its relation to calcium absorption in postmenopausal osteoporosis. J. Clin. Invest. 44, 1960 (1965). 6. O. Sorenson, B. Lumholtz, B. Lund, I. L. Hjelmshand, L. Mosekilde, F. Melsen, J. E. Bishop, and A. W. Norman, Acute effects of parathyroid hormone on vitamin D metabolism in patients with the bone loss of aging. J. Clin. Endocrinol. Metab. 54, 1258 – 1261 (1981). 7. K. Tsai, H. Heath, R. Kumar, and B. Riggs, Impaired vitamin D metabolism with aging in women. J. Clin. Invest. 3, 1668 – 1672 (1984). 8. J. F. Aloia, S. Cohn, and A. Vaswani, Risk factors for postmenopausal osteoporosis. Am. J. Med. 78, 95 – 100 (1985). 9. M. J. McKenna, Differences in vitamin D status between countries in young adults and the elderly. Am J. Med. 93, 69 – 77 (1992). 10. B. Dawson-Hughes, Regulation of parathyroid hormone by dietary calcium and vitamin D. In “The Parathyroids” (J. P. Bilezikian, R. Marcus, and M. A. Levine eds.), pp. 55 – 65. Raven Press, New York, 1994. 11. “National Academy of Science Recommended Dietary Allowances, 10th revised edition. The National Academy Press, Washington, DC, 1997. 12. J. M. Barragry, M. W. France, D. Corless, S. P. Gupta, S. Switala, B. J. Boucher, and R. D. Cohen, Intestinal cholecalciferol absorption in the elderly and in young adults. Clin. Sci. Mol. Sci. 55, 213 – 220 (1978). 13. M. J. Favus, Vitamin D. In “The Aging Skeleton” (C. J. Rosen, J. Glowacki, J. P. Bilezikian, eds), pp. 613 – 622. Academic Press, San Diego, CA, 2000.
CHAPTER 40 Role of Parathyroid Hormone and Vitamin D 14. O. H. Sorenson, B. Lumholtz, B. Lund, I. L. Hjelmstrand, L. Mosekilde, F. Melsen, J. E. Bishop, and A. W. Norman, Acute effects of parathyroid hormone on vitamin D. metabolism in patients with the bone loss of aging. J. Clin. Endocrinol. Metab. 54, 1258 – 1261 (1982). 15. B. L. Riggs, A. Hamstra, and H. F. Deluca, Assessment of 25hydroxyvitamin D 1alpha-hydroxylase reserve in postmenopausal osteoporosis by administration of parathyroid extract. J. Clin. Endocrinol. Metab. 53, 833 – 835 (1981) 16. K. S. Tsai, H. Heath III, R. Kumar, and B. L. Riggs, Impaired vitamin D metabolism with aging in women. Posssible role in pathogenesis of senile osteoporosis. J. Clin. Invest. 73, 1668 – 1672 (1984). 17. T. L. Clemens and J. L. H. O’Riordan, Vitamin D. In “Principles and Practice of Endocrinology and Metabolism” (K. L. Becker, ed.), 3rd ed. Lippincott, Philadelphia, 2001. 18. J. MacLaughlin and M. F. Holick, Aging decreases the capacity of human skin to produce vitamin D3. J. Clin. Invest. 76, 1536 – 1538 (1985). 19. B. Riggs and L. Melton, Medical progress: Involutional osteoporosis. N. Engl. J. Med. 314, 1676 – 1686 (1986). 20. A. R. Webb, C. Pilbeam, N. Hanafin, and M. Holick, An evaluation of the relative contributions of exposure to sunlight and of diet to the circulating concentrations of 25 hydroxyvitamin D in an elderly nursing home population in Boston. Am. J. Clin. Nutr. 51, 1075 – 1081 (1990). 21. H. M. Hodkinson, B. R. Stanton, P. Round, and C. Morgan, Sunlight, vitamin D and osteomalacia in the elderly. Lancet 1, 404 – 406 (1973). 22. A. R. Webb and M. F. Holick, The role of sunlight in the cutaneous production of vitamin D3. Annu. Rev. Nutr. 8, 375 – 399 (1988). 23. R. A. Bouillon, J. H. Auwers, W. D. Lissens, and W. K. Pelemans, Vitamin D status in the elderly: Seasonal substrate deficiency causes 1,25-dihyroxycholecaciferol deficiency. Am. J. Clin. Nutr. 45, 755 – 763 (1987). 24. T. C. B. Stamp and J. M. Round, Seasonal changes in human plasma levels of 25-hydroxyvitamin D. Nature 247, 1563 (1974). 25. M. A. Preece, S. Tomlinson, C. A. Ribot, J. Pietrek, H. T. Korn, D. M. Davies, J. A. Ford, M. G. Dunnijan, and I. L. O’Riordan, Studies of vitamin D deficiency in man. Q. J. Med. 44, 575 (1975). 26. M. F. Holick, The photobiology of vitamin D in man. In “Vitamin D. Basic and Clinical Aspects” (R. Kumar, ed.). Martinus Nijhoff, Boston, 1984. 27. A. R. Webb, L. Kline, and M. F. Holick, Influence of season and latitude on the cutaneous synthesis of vitamin D3: Exposure to winter sunlight in Boston and Edmonton will not promote vitamin D3 synthesis in human skin. J. Clin. Endocrinol. Metab. 67, 373 – 378 (1988). 28. C. J. Rosen, A. Morrison, H. Zhou, D. Storm, S. J. Hunter, K. Musgrave, T. Chen, W. Wei, and M. F. Holick, Elderly women in northern New England exhibit seasonal changes in bone mineral density and calciotropic hormones. Bone Miner. 25, 83 – 92 (1994). 29. D. Villareal, R. Civitelli, A. Chines, and L. Avioli, Subclinical vitamin D deficiency in postmenopausal women with low vertebral bone mass. J. Clin. Endocrinol. Metab. 72, 628 – 634 (1991). 30. M. Peacock, P. L. Selby, R. M. Francis, W. B. Brown, and L. Hordon, Vitamin D deficiency, insufficiency, sufficiency and intoxication: What do they mean? In “Vitamin D: Chemical, Biological and Clinical Update: 6th Workshop on Vitamin D” (A. W. Norman et al., eds.), pp. 569 – 570. Walter de Gruyter, New York, 1985. 31. L. Komar, J. Nieves, F. Cosman, A. Rubin, V. Shen, and R. Lindsay, Calcium homeostasis of an elderly population upon admission to a nursing home. J. Am. Geriatr. Soc. 41, 1057 – 1064 (1993). 32. F. M. Gloth, Vitamin D. In “The Aging Skeleton” (C. R. Rosen, J. Glowacki, and J. P. Bilezikian, eds), pp. 185 – 194. CA Academic Press, San Diego, 1999.
81 33. F. M. Gloth, C. M. Gundberg, B. W. Hollis, J. G. Haddad, J. D. Tobin, The prevalence of vitamin D deficiency in a cohort of homebound elderly subjects compared to a normative matched population in the United States. JAMA 274, 1683 – 1686 (1995). 34. M. K. Thomas, D. M. Lloyd-Jones, R. I. Thadhani, A. C. Shaw, D. J. Deraska, B. T. Kitch, E. C. Vamakis, I. M. Dick, R. L. Prince, and J. S. Finkelstein, Hypovitaminosis D in medical inpatients. N. Engl. J. Med. 338, 777 – 783 (1998). 35. D. S. Rao and M. Honasoge, Metabolic bone disease in gastrointestinal, hepatobiliary and pancreatic. In “Primer on the Metabolic Bone Diseases and Disorders of Mineral Metabolism” (M. J. Favus, ed.), 3rd ed., pp. 306 – 310. Lippinocott-Raven, New York, 1996. 36. L. M. Siegel and J. P. Bilezikian, Metabolic bone diseases and disorders of the gastrointestinal tract. In “Proceedings of the Falk Symposium: Gastrointestinal Tract and the Endocrine System.” Walters Kluwers, The Netherlands, 1995. 37. H. C. Mitchison, A. J. Malcolm, M. F. Bassendine, and O. F. W. James, Metabolic bone disease in primary biliary cirrhosis at presentation. Gastroenterology 94, 463 – 470 (1988). 38. D. P. DeSimone and N. H. Bell, Hypercalcemia and abnormal vitamin D metabolism. In “Disorders of Bone and Mineral Metabolism” (F. L. Coe and M. J. Favus, eds.), pp. 435 – 442. Raven Press, New York, 1992. 39. J. Gallagher, The pathogenesis of osteoporosis. Bone Miner. 9, 215 – 227 (1990). 40. D. M. Slovik, J. S. Adams, R. M. Neer, M. F. Hollick, and J. T. Potts, Jr., Deficient production of 1,25-dihydroxyvitamin D in elderly osteoporotic patients. N. Engl. J. Med. 305, 372 – 374 (1981). 41. B. Riggs, A. Hamstra, and H. DeLuca, Assessment of 25-hydroxyvitamin D 1-alpha hydroxylase reserve in postmenopausal osteoporosis by administration of parathyroid extract. J. Clin. Endocrinol. Metab. 53, 833 – 835 (1981). 42. B. P. Halloran, A. Portale, E. T. Lonergan, and R. C. Morris, Jr., Production and metabolic clearance of 1,25-dihydroxyvitamin D in men: Effect of advancing age. J. Clin. Endocrinol. Metab. 70, 3180 – 3323 (1990). 43. R. Francis, M. Peacock, G. Taylor, J. Storer, and B. Nordin, Calcium malaborption in elderly women with vertebral fractures: Evidence for resistance to the action of vitamin D metabolites on the bowel. Clin. Sci. 66, 103 – 107 (1984). 44. P. Ebeling, M. Sandgren, E. DiMango, A. Lane, H. DeLuca, and B. Riggs, Evidence of an age-related decrease in intestinal responsiveness to vitamin D: Relationship between serum 1,25-dihydroxyvitamin D and intestinal vitamin D receptor concentrations in normal women. J. Clin. Endocrinol. Metab. 75, 176 – 182 (1992). 45. C. Gennari, D. Agnusdei, P. Nardi, and R. Civitelli, Estrogen preserves a normal intestinal responsiveness to 1,25-dihydroxyvitamin D in oophorectomized women. J. Clin. Endocrinol. Metab. 71, 1288 – 1293 (1990). 46. S. Pattanaungkul, B. L. Riggs, A. L. Yergey, N. E. Vieira, W. M. O’Fallon, and S. Khosla, Relationship of intestinal calcium absorption to 1,25-dihydroxyvitamin D levels in young versus elderly women: Evidence for age-related intestinal resistance to 1,25-dihydroxyvitamin D action. J. Clin. Endocrinol. Metab. 85, 4023–4027 (2001). 47. N. A. Morrison, J. C. Qi, A. Tokita, P. J. Kelly, L. Crofts, T. V. Nguyen, P. N. Sambrook, and J. A. Eisman, Prediction of bone density from vitamin D receptor alleles. Nature (London) 367, 384 (1994). 48. N. A. Morrison, R. Yeoman, P. J. Kelly, and J. A. Eisman, Contribution of trans-acting factor alleles to normal physiological variability: Vitamin D receptor gene polymorphisms and circulating osteocalcin. Proc. Natl. Acad. Sci. USA 89, 6665 – 6669 (1992). 49. F. G. Hustmyer, M. Peacock, S. Hui, C. C. Johnston, and J. Christian, Bone mineral density in relation to polymorphism at the vitamin D receptor gene locus. J. Clin. Invest. 94, 2130 – 2134 (1994).
82 50. T. V. Nguyen and J. A. Eisman, Genetics of fracture: Challenges and opportunities. J. Bone Miner. Res. 15, 1253 – 1256 (2000). 51. S. J. Silverberg, I. N. Brown, J. P. Bilezikian, and L. J. Deftos, A new highly sensitive assay for parathyroid hormone in primary hyperparathyroidism. J. Bone Miner. Res. 15(Suppl. I), S167 (2000). 52. S. R. Nussbaum and J. T. Potts, Jr., Advances in Immunoassays for parathyroid hormone: Clinical applications in skeletal disorders of bone and mineral metabolism. In The Parathyroids (J. P. Bilezikian, R-Marcus, and M. Levine, eds), pp. 157 – 170. Raven Press, New York, 1994. 53. P. D. Delmas, D. Stenner, H. W. Wahner, K. G. Mann, and B. L. Riggs, Increase in serum BGP with aging in women: Implications for the mechanism of age-related bone loss. J. Clin. Invest. 71, 1316 – 1321 (1983). 54. R. J. Duda, J. F. O’Brien, J. A. Katzmann, J. M. Peterson, and K. G. Mann, B. L. Riggs, Concurrent assays of circulating bone GLA protein and bone alkaline phosphatase: Effects of age, sex and metabolic bone disease. J. Clin. Endocrinol. Metab. 66, 951 – 957 (1988). 55. R. Eastell, P. D. Delmas, S. F. Hodgson, E. F. Eriksen, K. G. Mann, and B. L. Riggs, Bone formation rate in older normal women: Concurrent assessment with bone histomorphometry, calcium kinetics, and biochemical markers. J. Clin. Endocrinol. 67, 741 – 748 (1988). 56. S. Vedi, J. E. Compston, A. Webb, and J. R. Tighe, Histomorphometric analysis of bone biopsies form the iliac crest of normal British subjects. Metab. Bone Dis. Relat. Res. 4, 231 – 236 (1982). 57. S. Vedi, J. E. Compston, A. Webb, and J. R. Tighe, Histomorphometric analysis of dynamic parameters of trabecular bone formation in the iliac crest of normal British subjects. Metab. Bone Dis. Relat. Res. 5, 69 – 74 (1983 – 1984). 58. S. Epstein, J. Poser, R. McClintock, C. C. Johnston, G. Bryce, and S. Hui, Differences in serum bone GLA protein with age and sex. Lancet 1, 307 – 310 (1987) 59. J. C. Gallagher, H. K. Kinyamu, S. E. Fowler et al., Calciotropic hormones and bone markers in the elderly. J. Bone Miner. Res. 13, 475 – 482 (1988). 60. S. Khosla, L. J. Melton III, and B. L. Riggs, Parathyroid function in the normal aging process. In “The Parathyroids” (J. P. Bilezikian, R. Marcus, and M. A. Levine, eds.). Academic Press, San Diego, in press. 61. L.K. H. Koh, D. C. Bauer, B. A. Forsyth, L. R. Gore, M. T. Vogt, and S. R. Cummings, PTH in postmenopausal osteoporotic women. J. Bone Miner. Res. 12, S166 (1997). 62. R. L. Prince, I. Dick, A. Devine, R. I. Price, D. H. Gutteridge, D. Kerr, A. Criddle, P. Garcia-Webb, and A. St. John, The effects of menopause and age on calcitropic hormones: A cross-sectional study of 655 healthy women aged 35 – 90. J. Bone Miner. Res. 10, 835 – 842 (1995). 63. R. G. H. Jilka, G. Girasole, G. Passeri, D. Williams, J. Abrams, B. Boyce, H. Broxmeyer, and S. Manolagas, Increased osteoclast development after estrogen loss: Mediation by interleukin-6. Science 257, 88 – 91 (1992). 64. S. C. Manolagas, R. L. Jilka, T. Bellido, C. A. O’Brien, and A. M. Parfitt, Interleukin-6-type cytokines and their receptors. In “Principles of Bone Biology” (J. P. Bilezikian, G. R. Raisz, and G. A. Rodan, Eds), pp. 701 – 714. Academic Press, San Diego, 1996. 65. B. Riggs and L. Melton, Evidence for two distinct syndromes of involutional osteoporosis. Am. J. Med. 75, 899 – 901 (1983). 66. P. Ebeling, J. Jones, M. Burritt, C. Duerson, A. Lane, C. Hassager, R. Kumar, and B. Riggs, Skeletal responsiveness to endogenous parathyroid hormone in postmenopausal osteoporosis. J. Clin. Endocrinol. Metab. 75, 1033 – 1038 (1992). 67. M. Kotowicz, G. Klee, P. Kao, S. Hodgson, S. Cedel, E. Eriksen, D. Gonchoroff, H. Judd, and B. L. Riggs, Relationship between serum intact PTH and bone remodeling in Type I osteoporosis: Evidence that skeletal sensitivity is increased. Osteoporosis 1, 14 – 20 (1990).
BILEZIKIAN AND SILVERBERG 68. J. P. Bilezikian, S. J. Silverberg, F. Gartenberg, T. S. Kim, T. Jacobs, E. Siris, and E. Shane, Clinical presentation of primary hyperparathyroidism. In “The Parathyroids” (J. P. Bilezikian, R. Marcus, and M. A. Levine, eds.), pp. 457 – 470. Raven Press, New York, 1994. 69. R. Marcus, P. Madvig, M. Crim, A. Pont, and J. Kosek, Conjugated estrogens in the treamtent of postmenopausal women with hyperparathyroidism. Ann. Int. Med. 100, 633 – 640 (1984). 70. P. Selby and M. Peacock, Ethinyl estradiol and norethindrone in the treatment of primary hyperparathyroidism in postmenopausal women. N. Engl. J. Med. 314, 1481 – 1485 (1986). 71. W. R. McKane, S. Khosla, M. F. Burritt, P. C. Kao, D. M. Wilson, S. J. Ory, and B. L. Riggs, Mechanism of renal calcium conservation with estrogen replacement therapy in women in early post menopause, J. Clin. Endocrinol. 80, 3458 – 3464 (1995). 72. R. Heaney, A unified concept of osteoporosis. Am. J. Med. 39, 377 – 380 (1965). 73. C. Jasani, B. Nordin, D. Smith, and I. Swanson, Spinal osteoporosis and the menopause. Proc. R. Soc. Med. 58, 441 – 444 (1965). 74. B. L. Riggs and L. J. Melton III, Evidence for two distinct syndromes of involutional osteoporosis. Am. J. Med. 75, 899 (1983). 75. K. Tsai, P. Ebeling, and B. Riggs, Bone responsiveness to parathyroid hormone in normal and osteoporotic postmenopausal women. J. Clin. Endocrinol. Metab. 69, 1024 – 1027 (1989). 76. B. L. Riggs, S. Khosla, and L. J. Melton III, A unitary model for involutional osteoporosis: estrogen deficiency causes both type I and type II osteoporosis in postmenopausal women and contributes to bone loss in aging men, J. Bone Miner. Res. 13, 763 – 773 (1998). 77. H. M. Heshmati, S. Khosla, M. F. Burritt, W. M. O’Fallon, and B. L. Riggs, A defect in renal calcium conservation may contribute to the pathogenesis of postmenopausal osteoporosis. J. Clin. Endocrinol. Metab. 83, 1916 – 1920 (1998). 78. S. Khosla, E. J. Atkinson, L. J. Melton III, and B. L. Riggs, Effects of age and estrogen status on serum parathyroid hormone levels and biochemical markers of bone turnover in women: A populationbased study. J. Clin. Endocrinol. Metab. 82, 1522 – 1527 (1997). 79. W. R. McKane, S. Khosla, J. Risteli, S. P. Robins, J. M. Muhs, and B. L. Riggs, Role of estrogen deficiency in pathogenesis of secondary hyperparathyroidism and increased bone resorption in elderly women. Proc. Assoc. Am. Phys. 109, 174 – 180 (1997). 80. B. L. Riggs, W. M. O’Fallon, J. Muhs, M. K. O’Conner, R. Kumar, and L. J. Melton, Long-term effects of calcium supplementation on serum parathyroid hormone level, bone turnover, and bone loss in elderly women. J. Bone Miner. Res. 13, 168 – 174 (1998). 81. M. Parisien, D. W. Dempster, E. Shane, and J. B. Bilezikian, Histomorphometric analysis of bone in primary hyperparathyroidism. In “The Parathyroids” (J. P. Bilezikian, R. Marcus, and M. A. Levine, eds.), pp. 567 – 574. Raven Press, New York, 1994. 82. J. S. Finkelstein, Pharmacological mechanisms of therapeutics: Parathyroid hormone. In “Principles of Bone Biology” (J. P. Bilezikian, L. G. Raisz, and G. A. Rodan, eds.), pp. 993 – 1006 Academic Press, San Diego, 1996. 83. J. P. Bilezikian, Estrogens and postmenopausal osteoporosis: Was Albright right after all? J. Bone Miner. Res. 13, 774 – 776 (1998). 84. J. P. Bilezikian and S. J. Silverberg, Parathyroid hormone: Does it have a role in the pathogenesis of osteoporosis. Clin. Lab. Med. 20, 559 – 567 (2000). 85. R. Eastell, A. L. Yergey, N. E. S. L. Veiria, R. Kumar, and B. L. Riggs, Interrelationships among vitamin D metabolism, true calcium absorption, parathyroid function, and age in women: Evidence of an age-related intestinal resistance to 1,25-dihydroxyvitamin D action. J Bone Miner. Res. 6, 125 – 132 (1991). 86. G. A. Ledger, M. F. Burritt, P. C. Kao, W. M. O’Fallon, B. L. Riggs, and S. Khosla, Role of parathyroid hormone in mediating nocturnal and age-related increases in bone resorption. J. Clin. Endocrin. Metab. 80, 3304 – 3310 (1995).
83
CHAPTER 40 Role of Parathyroid Hormone and Vitamin D 87. S. Boonen, X. G. Cheng, J. Nijs, P. H. Nicholson, G. Verbeke, E. Lesaffre, J. Aerssens, and J. Dequeker, Factors associated with cortical and trabecular bone loss as quantified by peripheral computed tomography (pQCT) at the ultradistal radius in aging women. Calcif. Tissue Int. 60, 164 – 170 (1997). 88. G. A. Ledger, M. F. Burritt, P. C. Kao, W. M. O’Fallon, B. L. Riggs, and S. Khosla, Abnormalities of parathyroid hormone secretion in elderly women that are reversible by short term therapy with 1,25dihydroxyvitamin D3. J. Clin. Endocrinol. Metab. 79, 211 – 216 (1994). 89. J. F. Tohme, F. Cosman, and R. Lindsay, Osteoporosis. In “Principles and Practice of Endocrinology and Metabolism” (K. L. Becker, ed.), 2nd ed. Lippincott, Philadelphia, 1995. 90. M. S. Calvo, R. Eastell, K. P. Offord, E. J. Bergsalh, and M. F. Burritt, Circadian variation in ionized calcium and intact parathyroid hormone: Evidence for sex differences in calcium homeostasis. J. Clin. Endocrinol. Metab. 72, 77 – 82 (1991). 91. M. E. Markowitz, S. Arnaud, J. F. Rosen, M. Thorpy, and S. Laximinarayan, Temporal relationships between the circadian rhythms of serum PTH and calcium concentrations. J. Clin. Endocrinol. Metab. 67, 1068 – 1073. 92. R. Eastell, M. S. Calvo, M. F. Burritt, K. P. Offord, R. G. G. Russell, and B. L. Riggs, Abnormalities in circadian patterns of bone resorption and renal calcium conservation in type I osteoporosis. J. Clin. Endocrinol. Metab. 74, 487 – 494 (1992). 93. K. Prank, S. Nowlan, H. Harms, et al., Time series prediction of plasma hormone concentration: Evidence for differences in predictability of parathyroid hormone secretion between osteoporotic patients and normal controls. J. Clin. Invest. 95, 2910 – 2919 (1995). 94. L. J. Deftos and S. J. Schiff, Predicting PTH pulses and patterns in osteoporosis. J. Clin. Invest. 95, 2433 – 2434 (1995). 95. M. H. Samuels, J. D. Veldhuis, P. Kramer, et al., Episodic secretion of parathyroid hormone in postmenopausal women: Assessment by deconvolution analysis and approximate entropy. J. Bone Miner. Res. 12, 616 – 623 (1997). 96. S. J. Silverberg, E. Shane, T. L. Clemens, D. W. Dempster, G. V. Segre, R. Lindsay, and J. P. Bilezikian, The effects of oral phosphate on major indices of skeletal metabolism in normal subjects. J. Bone Miner. Res. 1, 383 – 338 (1986). 97. S. J. Silverberg, E. Shane, L. de la Cruz, G. V. Segre, T. L. Clemens, and J. P. Bilezikian, Abnormalities in parathyroid hormone secretion and 1,25-dihydroxyvitamin D3 formation in women with osteoporosis. N. Engl. J. Med. 320, 277 – 281 (1989). 98. F. Cosman, V. Shen, B. Herrington, and R. Lindsay, Response of the parathyroid gland to infusion of human parathyroid hormone(1 – 34). J. Clin. Endocrinol. Metab. 73, 1345 – 1351 (1991). 99. A. A. Portale, E. T. Lonergan, D. M. Tanney, and B. P. Halloran, Aging alters calcium regulation of serum concentration of parathyroid hormone in healthy men. Am. J. Physiol. 35, E139 – E146 (1997). 100. Y. Liel, H. Edwards, J. Shary, K. M. Spicer, L. Gordan, and N. H. Bell, The effects of race and body habitus on bone mineral density of the radius, hip and spine in premenopausal women. J. Clin. Endocrinol. Metab. 66, 1247 – 1250 (1988). 101. M. E. Farmer, L. R. White, and J. A. Brody, Race and sex differences in hip fracture incidence. Am. J. Public Health 44, 1374 – 1380 (1984). 102. M. Modlin, Urinary calcium in normal adults and in patients with renal stones: An interracial study. Invest. Urol. 5, 49 – 57 (1967). 103. N. H. Bell, A. Greene, S. Epstein, M. J. Oexmann, S. Shaw, and J. Shary, Evidence for alteration of the vitamin D-endocrine system in blacks. J. Clin. Invest. 76, 470 – 473 (1985). 104. M. Kleerekoper, D. Nelson, E. L. Peterson, M. J. Flynn, A. S. Pawluszka, G. Jacobson, and P. Wilson, Reference data for bone mass, calcitropic hormones and biochemical markers of bone remodeling in
105.
106.
107.
108.
109. 110.
111.
112.
113.
114.
115.
116.
117.
118.
119.
older (55 – 75) postmenopausal white and black women. J. Bone Miner. Metab. 9, 1267 – 1276 (1994). D. E. Meier, M. M. Luckey, S. Wallenstein, T. L. Clemens, E. S. Orwoll, and C. I. Waslien, Calcium, vitamin D, and parathyroid hormone status in young white and black women: Association with racial differences in bone mass. J. Clin. Endocrinol. Metab. 72, 703 – 710 (1991). G. El-Haji Fuleihan, C. M. Gundberg, R. Gleason, E. M. Brown, M. E. Stronski, F. D. Grant, and P. R. Conlin, Racial differences in parathyroid hormone dynamics. J. Clin. Endocrinol. Metab. 79, 1642 – 1647 (1994). T. L. Clemens, S. L. Henderson, J. S. Adams, M. F. Holick, Increased skin pigment reduces the capacity of skin to synthesize vitamin D3. Lancet 1, 74 – 76 (1982). M. F. Holick, J. A. MacLaughlin, and S. H. Doppelt, Regulation of cutaneous previtamin D photosynthesis in man: Skin pigment is not an essential regulator. Science 211, 590 – 593 (1981). R. S. Weinstein and N. H. Bell, Diminished rates of bone formation in normal black adults. N. Engl. J. Med. 319, 1698 – 1701 (1988). S. J. Silverberg, E. Shane, L. de la Cruz, D. W. Dempster, F. Feldman, D. Seldin, T. P. Jacobs, E. S. Siris, M. Cafferty, M. V. Parisien, R. Lindsay, T. L. Clemens, and J. P. Bilezikian, Skeletal disease in primary hyperparathyroidism. J. Bone Miner. Res. 4, 283 – 291 (1989). M. Parisien, S. J. Silverberg, E. Shane, L. de la Cruz, R. Lindsay, J. P. Bilezikian, and D. W. Dempster, The histomorphometry of bone in primary hyperparathyroidism: Preservation of cancellous bone. J. Clin. Endocrinol. Metab. 70, 930 – 938 (1990). M. Parisien, R. Mellish, S. J. Silverberg, E. Shane, R. Lindsay, J. P. Bilezikian, and D. W. Dempster, Maintenance of cancellous bone connectivity in primary hyperparathyroidism: Trabecular and strut analysis. J. Bone Miner. Res. 7, 913 – 920 (1992). M. Parisien, F. Cosman, R. W. R. Mellish, M. Schnitzer, J. Nieves, S. J. Silverberg, E. Shane, D. Kimmel, R. R. Recker, J. P. Bilezikian, R. Lindsay, and D. W. Dempster, Bone structure in postmenopausal hyperparathyroid, osteoporotic and normal women. J. Bone Miner. Res. 10, 1393 – 1399 (1995). S. J. Silverberg, F. Gartenberg, T. P. Jacobs, E. Shane, E. Siris, R. B. Staron, D. McMahon, and J. P. Bilezikian, Longitudinal measurements of bone density and biochemical indices in untreated primary hyperparathyroidism. J. Clin. Endocrinol. Metab. 80, 723 – 728 (1995). D. M. Slovik, D. I. Rosenthal, S. H. Doppelt, J. T. Botts, M. A. Daly, J. A. Campbell, and R. M. Neer, Restoration of spinal bone in osteoporotic men by treatment with human parathyroid hormone (1 – 34) and 1,25-dihydroxyvitamin D. J. Bone Miner. Res. 1, 377 – 3811 (1986). J. Reeve, J. N. Bradbeer, M. Arlot, U. M. Davies, J. R. Green, L. Hampton, C. Edouard, R. Hesp, P. Hulme, and J. P. Ashby, hPTH1-34 treatment of osteoporosis with added hormone replacement therapy: Biochemical, kinetic and histological responses. Osteoporosis Int. 1, 162 – 170 (1991). R. Neer, C. Arnaud, J. R. Zanchetta, R. Prince, G. A. Gaich, and J. Y. Reginster, “Recombinant Human PTH [rhPTH(1 – 34)] Reduces the Risk of Spine and Non-spine Fractures in Postmenopausal Osteoporosis. 82nd Annual Meeting of the Endocrine Society, S193, 2000. [Abstract] R. Lindsay, J. Nieves, C. Formica, E. Henneman, L. Woelfert, V. Shen, D. Dempster, and F. Cosman, Randomised controlled study of effect of parathyroid hormone on vertebral-bone mass and fracture incidence among postmenopausal women on oestrogen with osteoporosis. Lancet 350, 550 – 555 (1997) T. Fujita, T. Inoue, M. Morii, R. Morita, H. Norimatsu, H. Orimo, H. E. Takahashi, K. Yamamoto, and M. Fukunaga, Effect of intermittent weekly dose of human parathyroid hormone (1 – 34) on osteoporosis: A randomized double-masked prospective study using three dose levels. Osteoporosis Int. 9, 296 – 306 (1999).
84 120. E. S. Kurland, F. Cosman, D. J. McMahon, C. J. Rosen, R. Lindsay, and J. P. Bilezikian, Therapy of idiopathic osteoporosis in men with parathyroid hormone: effects on bone mineral density and bone markers. J. Clin. Endocrinol. Metab. 85, 3069 – 3076 (2000). 121. E. B. Roe, S. D. Shanchez, G. A. del Puero, E. Pierini, P. Bacchetti, C. E. Cann, and C. D. Arnaud, Parathyroid hormone 1 – 34 (hPTH 1 – 34) and estrogen produce dramatic bone density increases in postmenopausal osteoporosis – Results from a placebo-controlled randomized trial. J. Bone Miner. Res. 14, S137 (1999). 122. N. E. Lane, S. Sanchez, G. W. Modin, H. K. Genant, E. Pierini, and C. D. Arnaud, Parathyroid hormone treatment can reverse corticosteroid-induced osteoporosis: Results of a randomized controlled clinical trial. J. Clin. Invest. 102, 1627 – 1633 (1998). 123. J. S. Finkelstein, A. Klibanski, E. H. J. Schaefer, M. D. Hornstein, I. Schiff, and R. M. Neer, Parathyroid hormone for the prevention of
BILEZIKIAN AND SILVERBERG bone loss induced by estrogen deficiency. N. Engl. J. Med. 331, 1618 – 1623 (1994). 124. R. S. Rittmaster, M. Bolognese, M. P. Ettinger, D. A. Hanley, A. B. Hodsman, D. L. Kendler, and C. J. Rosen, Enhancement of bone mass in osteoporotic women with parathyroid hormone followed by alendronate. J. Clin. Endocrinol. Metab. 85, 2129 – 2134 (2000). 125. D. S. Rao, R. J. Wilson, M. Kleerekoper, and A. M. Parfitt, Lack of biochemical progression or continuation of accelerated bone loss in mild asymptomatic primary hyperparathyroidism: Evidence for biphasic disease course. J. Clin. Endocrinol. Metab. 676, 1294 – 1298 (1988). 126. S. J. Silverberg, F. Locker, and J. P. Bilezikian, Vertebral osteopenia: A new indication for surgery in primary hyperparathyroidism, J. Clin. Endocrinol. Metab. 81, 4007 – 4012 (1996).
CHAPTER 41
Postmenopausal Osteoporosis How the Hormonal Changes of Menopause Cause Bone Loss ROBERTO PACIFICI
Division of Bone and Mineral Diseases, Washington University, St. Louis, Missouri 63110
I. Steroid Biosynthesis and Menopause II. Mechanism of Action of Estrogen in Bone
III. Summary and Conclusions References
I. STEROID BIOSYNTHESIS AND MENOPAUSE
postmenopausal period the major source of estradiol becomes the peripheral conversion of estrone and testosterone. This conversion takes places at many extraglandular sites, but mainly in the adipose tissue. The latter pathway is enhanced with aging and obesity. Due to the depletion of responsive follicles after menopause, the range of progesterone concentrations resembles that observed in premenopausal women during the proliferative phase. Small amounts of progesterone continue to be made by the adrenal glands, which also become the main source of androstenedione, the most abundant androgen in postmenopausal women [2]. The ovaries continue to account for about 20% of total androstenedione production. In contrast, ovarian production of testosterone does not decrease significantly after the menopause [2]. Dehydroepiandrosterone (DHEA) and DHEA-sulfate are mostly produced by the adrenal gland. The production of these androgens declines after age 30, independent of ovarian function. Additionally, it should be underlined that menopause is also the cause of an increased production of the pituitary
Menopause represents a critical life step characterized by complex endocrine changes which affect the musculoskeletal system and its neurological control. The hallmark of the menopausal transition is the cessation of menses. However, the hormonal changes that signal decreased ovarian function begin to occur in the decade prior to the development of frankly irregular cycles. Pathognomonic of the cessation of ovarian function is a marked decline in the 17-estradiol concentrations (Table 1), the major estrogen in women of reproductive age [1]. Menopause is also characterized by a marked decrease in estrone serum concentration. However, estrone produced as a result of the peripheral conversion of androstenedione and testosterone of adrenal and ovarian origins [1] becomes the most abundant estrogen after menopause. To a lesser extent estrone results from the hydrolysis of estrone sulfate and this represents a large and stable pool of estrogen in the body. In the
OSTEOPOROSIS, SECOND EDITION VOLUME 2
85
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
86
ROBERTO PACIFICI
TABLE 1
Serum Levels of Sex Steroid Hormones after Menopause
Hormone
Postmenopausal serum level
17 Estradiol (pg/ml) Estrone (pg/ml)
5 – 20 10 – 20
Progesterone (ng/ml) Androstenedione (pg/dl)
0–6 400 – 1100
DHEA (g/ml)
Major source after menopause Peripheral conversion of estrone and testosterone Peripheral conversion of androstenedione and testosterone Ovary and Adrenal Adrenal
0–3
Adrenal
DHEA-sulfate (g/ml)
0.82 – 3.38
Adrenal
Testosterone
144 – 252
Ovary
hormones follicle-stimulating hormone (FSH) and lutenizing hormone (LH). Although the role of these substances in maintaining bone health is unknown, evidence is beginning to emerge indicating that pituitary hormones may modulate the effects of estrogen in bone [3], perhaps by activating the estrogen receptor (ER) [4].
II. MECHANISM OF ACTION OF ESTROGEN IN BONE A. Introduction Postmenopausal osteoporosis is a heterogeneous disorder characterized by a progressive loss of bone tissue which begins after natural or surgical menopause and leads to fracture within 15 – 20 years from the cessation of the ovarian function [5]. Although suboptimal skeletal development (“low peak bone mass”) and age-related bone loss may be contributing factors, a hormone-dependent increase in bone resorption and accelerated loss of bone mass in the first 5 or 10 years after menopause appears to be the main pathogenetic factor [6,7] of this condition. That estrogen deficiency plays a major role in postmenopausal bone loss is strongly supported by the higher prevalence of osteoporosis in women than in men [8], the increase in the rate of bone mineral loss detectable by bone densitometry after artificial or natural menopause [9 – 11], the existence of a relationship between circulating estrogen and rates of bone loss [12 – 14], and the protective effect of estrogen replacement with respect to both bone mass loss and fracture incidence [15 – 17]. The potential fracture risk for any postmenopausal female depends on the degree of bone turnover, the rate and extent of bone loss, associated disease processes which induce bone loss, age of menarche and menopause, and the bone mass content achieved at skeletal maturity. The latter depends on the extent of estro-
gen exposure, habitual physical activity, quantity of calcium intake, and genetic predisposition (see Chapter 25). The manner with which the genetic “signal” conditions those biological mechanisms which are essential to achieve peak bone mass in adolescence is still unknown, although evidence is beginning to accumulate that low peak bone mass may be linked to a particular vitamin D receptor genotype [18] (see Chapter 26). The bone-sparing effect of estrogen is mainly related to its ability to block bone resorption [19], although stimulation of bone formation is likely to play a contributory role [20,21]. Estrogen-dependent inhibition of bone resorption is, in turn, due to both decreased osteoclastogenesis and diminished resorptive activity of mature osteoclasts. However, inhibition of osteoclast formation is currently regarded as the main mechanisms by which estradiol (E2 prevents bone loss [19,22].
B. Cells and Cytokines Which Regulate Osteoclast Formation Osteoclasts arise by cytokine-driven proliferation and differentiation of monocyte/macrophage precursors, a process facilitated by bone marrow stromal cells (Fig. 1) These cells provide a physical support for nascent osteoclasts and produce soluble and membrane-associated factors essential for the proliferation and/or the differentiation of osteoclast precursors [23]. During inflammation, activated T cells, a population of lymphocytes which does not participate in the regulation of physiologic bone turnover, assumes a key role in stimulating osteoclast formation and does so by producing potent membrane-bound and soluble cytokines [24]. Another cell lineage that may have an important role in the regulation of osteoclastogenesis is that of B cells although their exact role remains controversial. For example, B-cell-deficient mice have been found to display decreased trabecular area and increased bone resorption, as
87
CHAPTER 41 Postmenopausal Osteoporosis
FIGURE 1
Cell and cytokines critical for osteoclast formation. Estrogen decreases osteoclast formation by downregulating the monocytic production of IL-1 and TNF and the stromal cell production of M-CSF and IL-6.
compared to B-cell-replete mice of the same strain [25], suggesting that B cells inhibit bone resorption and osteoclastogenesis. In contrast, other studies have shown that estrogen deficiency up regulates B-lymphopoiesis in the bone marrow [26,27], suggesting that cells of the B lineage may contribute to the increased osteoclast [OC] production characteristic of estrogen deficient animals. In humans B cells are an important source of transforming growth factor (TGF), a factor that inhibits osteoclast formation by inducing apoptosis of early and late osteoclast precursors and mature osteoclasts. Among the cytokines involved in the regulation of osteoclast formation are (RANKL) (also known as OPGL, TRANCE, or ODF) and macrophage colony-stimulating factor (M-CSF) (Table 2) [28 – 33] (see Chapters 3, 6, 12, and 13). These factors are produced primarily by stromal cells and osteoblasts. However, T cells are an important source of both factors, especially in stimulated conditions [24]. RANKL is a member of the tumor necrosis factor (TNF) family which exists in a membrane-bound and in a soluble form. RANKL binds to the transmembrane receptor RANK, which is expressed on the surface of osteoclasts and osteoclast precursors of the monocytic lineage [28]. RANKL binds TABLE 2
also to osteoprotegerin (OPG), a soluble decoy receptor produced by numerous hematopoietic cells. Thus, OPG, by sequestering RANKL and preventing its binding to RANK, functions as a potent anti-osteoclastogenic cytokine [34]. In the presence of M-CSF, RANKL induces the differentiation of monocytic cells into osteoclasts [28] by activating the MAP kinase Jun terminal Kinase (JNK), an enzyme which enhances the production of two essential osteoclastogenic transcription factors: c-Fos and c-Jun [35]. RANKL binding to RANK also activates NFB, a family of transcription factors essential for osteoclast formation and survival. Under physiological conditions M-CSF and RANKL are the only factors produced in the bone marrow in an amount sufficient to induce osteoclast formation. Thus, MCSF and RANKL are regarded as the true essential physiologic osteoclastogenic cytokines. The critical role of each of these cytokines in the osteoclastogenic process is demonstrated by the finding that deletion of either gene prompts osteopetrosis due to absence of osteoclasts, a circumstance reversed by administration of the relevant cytokine [29,36,37]. M-CSF induces the proliferation of early osteoclast precursors, the differentiation of more mature osteoclasts, the fusion of mononucleated preosteoclasts, and it increases the survival of mature osteoclasts [38 – 40]. RANKL does not induce cell proliferation, but promotes the differentiation of osteoclast precursors from an early stage of maturation to fully mature multinucleated osteoclasts. RANKL is also capable of activating mature osteoclasts, thus rendering these cells capable of resorbing bone. While consensus exists that RANKL stimulates bone resorption in organ cultures, the effect of M-CSF on bone resorption is controversial, as both inhibitory and stimulatory effects on bone resorption have been reported [38 – 43]. Monocytes, stromal cells, osteoblasts, and lymphocytes produce inflammatory cytokines which have direct proosteoclastogenic effects. Among these factors are interleukin (IL)-1, IL-6, IL-11, and TNF [44 – 53]. These factors stimulate osteoclast formation by increasing the stromal cell
Effects of Cytokines on Bone Resorption
Cytokine
Stimulate osteoclast formation
Stimulate resorption activity of mature osteoclasts
Stimulate in vivo bone resorption under conditions of estrogen deficiency
RANKL
Yes
Yes
No
M-CSF
Yes
?
Yes
IL-1
No
Yes
Yes
IL-6
Yes
No
No
TNF
Yes
Yes
Yes
GM-CSF
Yes
No
No
88
ROBERTO PACIFICI
production of RANKL [30,54,55] and M-CSF [56,57]. Another factor relevant for osteoclastogenesis is TGF-. This cytokine stimulates OPG production [58], thus inhibiting osteoclast formation. While in physiological conditions IL-1, IL-6, and TNF are produced in the bone marrow at low concentration, their bone marrow levels increase both during inflammation and in conditions of estrogen deficiency [19,22]. Thus, IL-1, IL-6, and TNF play a critical role in enhancing osteoclast production, survival, and activity in pathological conditions. Recently Kobayashi et al. [59] have demonstrated that TNF, in the presence of M-CSF, induces the differentiation of monocytes into mature multinucleated osteoclasts, which are, however, incapable of resorbing bone. Neither IL-1 nor IL-6 are capable of directly promoting the differentiation of osteoclast precursors into multinucleated osteoclasts. However, the addition of IL-1 to cultures of osteoclasts generated using TNF and M-CSF induces in these cells the capacity of resorbing bone and increases their survival. Thus, TNF is a true osteoclastogenic cytokine which can induce osteoclast formation when present at high concentrations. In contrast, IL-1 is incapable of inducing osteoclast formation, although it promotes osteoclast activation and survival. Since M-CSF and RANKL are present in the bone marrow in physiological conditions, osteopetrosis is not a feature of transgenic mice lacking the capacity of producing and/or responding to either IL-1, IL-6, or TNF [60 – 62]. Thus, IL-1, IL-6, and TNF stimulate osteoclastogenesis in pathological conditions, but are not essential for baseline osteoclastogenesis. In summary, the accumulated data demonstrate that RANKL and M-CSF are the only two factors known at the present time which are absolutely critical for osteoclast formation in physiological conditions. In contrast, the inflammatory cytokines IL-1, IL-6, and TNF are not essential for the maintenance of baseline osteoclastogenesis, although they are key for enhancing osteoclast formation and osteoclast activity during inflammation [40,63] and in conditions of E2 deficiency [19,22].
C. Effects of Estrogen on the Production of Osteclastogenic Cytokines It is now recognized that estrogen downregulates the production of several proosteoclastogenic factors, including IL-1, IL-6, TNF, M-CSF, and Prostaglandin E2 (PGE2) In addition, estrogen stimulates the production of important anti-osteoclastogenic factors, including IL-1ra [64], OPG [65], and TGF- [66]. The expression of estrogen receptors (both and ) has been demonstrated in monocytes [67,68], osteoclasts [69], stromal cells, osteoblasts [70], and T cells [71]. Thus, all the major populations involved in the osteoclastogenesis
are a target of the sex steroid (see Chapter 10 for further discussion of estrogens and Chapter 13 for cytokines). The cytokines first recognized to be regulated by estrogen were IL-1 and TNF. This observation was prompted by the finding that monocytes of patients with “high turnover” osteoporosis, the histological hallmark of postmenopausal osteoporosis, secrete increased amounts of IL-1 [72]. Crosssectional and prospective comparisons of pre- and postmenopausal women revealed that monocytic production of IL-1 and TNF increases after natural and surgical menopause and is decreased by treatment with estrogen and progesterone [73,74]. Subsequent observations showed that the postmenopausal increase in IL-1 activity results from an effect of estrogen on the production of both IL-1 and IL-1ra [64]. Studies in normal women undergoing ovariectomy (ovx] [75,76] revealed that estrogen withdrawal is associated with an increased production of IL-1 and TNF, and also of granuloayte-macrophage colony-stimulating factor (GMCSF). The changes in these cytokine levels occur in a temporal sequence consistent with a causal role of IL-1, TNF, and GM-CSF in the pathogenesis of ovx-induced bone loss [75]. Estrogen and progesterone have been shown to decrease the secretion of IL-1 from peripheral blood and bone marrow monocytes and to decrease the steady-state expression of IL-1 mRNA in monocytes [77]. However, the exact molecular mechanism by which E2 decreases IL-1 production remains to be determined. Estrogen has been shown to increase the expression of the decoy type II, IL-1 receptor in bone marrow cells and osteoclasts [78]. Thus upregulation of cell responsiveness to IL-1 via downregulation of IL11RII is also likely to be a key mechanism by which estrogen deficiency induces bone loss. Recently we have clarified how estrogen downregulates the monocytic production of TNF. The genomic effects of E2 are mediated by a ligand-inducible transcription factor, known as ER. Following the discovery of a second ER (ER) [79 – 81], the first identified ER is now known as ER [79]. ER is expressed primarily in the uterus, testis, ovary, and pituitary, while ER is expressed mostly in the prostate, ovary, lung, bladder, brain, uterus, and testis [82]. Although ER and ER have similar DNA binding domains, their A/B domains and activation-function region 1 (AF-1) and quite different, suggesting that they may differentially regulate ER responsive genes [82]. Recent studies have indeed demonstrated that ER and ER respond differently to ligands, leading to opposite effects on AP-1-induced gene expression [83]. Specifically, while the ER mediated effects of E2 lead to stimulation of AP-1 induced gene expression, E2 acts as repressor of AP-1-induced transcription when bound to ER [83]. Cells of the monocytic lineage are known to express both ER and ER. Estrogen binding to ER leads to decreased activation of the Jun terminal kinase (JNK), a phenomenon which leads to decreased production of c-Jun
CHAPTER 41 Postmenopausal Osteoporosis
and JunD, two members of the AP-1 family of transcription factors [84]. Decreased AP-1 production results in decreased AP-1-induced TNF gene expression and lower TNF production [84, 85]. Studies conducted to determine if estrogen regulates the production of IL-6 revealed that in murine stromal and osteoblastic cells IL-6 production is inhibited by the addition of estrogen [49] and stimulated by estrogen withdrawal [51]. In vivo studies also revealed that the production of IL-6 is increased in cultures of bone marrow cells from ovx mice [50]. This effect is mediated, at least in the mouse, by an indirect effect of estrogen on the transcription activity of the proximal 225-bp sequence of the IL-6 promoter [86,87]. Interestingly, although studies with human cell lines demonstrated an inhibitory effect of estrogen on the human IL-6 promoter [88], three independent groups have failed to demonstrate an inhibitory effect of estrogen on IL-6 production from human bone cells and stromal cells expressing functional estrogen receptors [89 – 91]. These data raise the possibility that the production of human IL-6 protein does not increase under conditions of estrogen deficiency. This is further supported by a report that, in humans, surgical menopause is not followed by an increase in IL-6, although it causes an increase in soluble IL-6 receptor [92]. Recent studies have unveiled that one of the key mechanism by which estrogen regulates osteoclastogenesis is by modulating the stromal cell production of M-CSF. Under conditions of E2 deficiency the high bone marrow levels of IL-1 and TNF lead to the expansion of a stromal cell population which produces larger amounts of soluble M-CSF [93]. These high M-CSF producing stromal cells have an increased capacity to support osteoclastogenesis (Fig. 2). Interestingly, estrogen has no direct regulatory effects on the production of soluble M-CSF as it regulates M-CSF secretion exclusively by conditioning the differentiation of stromal cells toward a phenotype characterized by a lower production of M-CSF. The high M-CSF producing stromal cells found in estrogen deficient mice are characterized by increased phosphorylation of the transcription factor Egr-1. While Egr-1 binds and sequesters the nuclear protein Sp-1, phosphorylated Egr-1 does not bind to Sp-1. As a result, cells from estrogen-deficient mice are characterized by increased levels of free Sp-1. This protein binds to the M-CSF promoter and stimulates M-CSF gene expression [94] (Fig. 3). In addition to an indirect effect on soluble M-CSF, E2 has been shown to decrease the production of membrane-bound M-CSF via a direct effect on bone marrow cells [95,96]. However, the source of membrane-bound M-CSF which is under estrogen regulation remains to be defined. Regardless of the specific cell involved, estrogen regulates this key osteoclastogenic cytokine with at least two distinct mechanisms. Little information on the effects of menopause on the production of RANKL is currently available. However, the promoter region of the RANKL gene does not contain
89
FIGURE 2 Estrogen regulates the differentiation of stromal cell precursors and leads to the formation of “low” M-CSF producing stromal cells.
regions known to be repressed (directly or indirectly) by estrogen [97]. Therefore it is likely that future studies will confirm the preliminary observation available at the moment, which indicates that estrogen does not regulate RANKL. In contrast, estrogen has been shown to increase the production of OPG in osteoblastic cells [65]. Thus, estrogen enhancement of OPG secretion by osteoblastic cells is likely to represent another major mechanism in explaining the antiresorptive action of estrogen on bone. Another possible intermediate in estrogen action is TGF. This growth factor is a multifunctional protein that is produced by many mammalian cells including osteoblasts and has a wide range of biological activities (see Chapter 14). TGF is a potent osteoblast mitogen [98]. Under specific experimental conditions TGF- decreases both osteoclastic resorptive activity and osteoclast recruitment. Oursler et al. have reported that estrogen increases the steady-state level of TGF- mRNA and release of TGF- protein [66]. This mechanism provides the first example of “positive” effects of estrogen in bone which may result in decreased bone turnover.
D. Effects of Menopause on the Production of Bone Resorbing Cytokines An abundance of in vitro studies that demonstrated the potent effects of IL-1, TNF, and IL-6 on bone prompted a
90
ROBERTO PACIFICI
FIGURE 3 Mechanism by which stromal cells from estrogen deficient mice produce low levels of M-CSF. Stromal cells from estrogen deficient mice exhibit increased CKII-dependent phosphorylation of the nuclear protein Egr-1. Phosphorylated Egr-1 binds less avidly to the transcriptional activator Sp-1 and the resulting higher levels of free Sp-1 stimulates M-CSF gene expression. series of investigations on the relationship between bone remodeling, cytokine production, and osteoporosis (see Chapter 13). These studies were conducted using cultures of peripheral blood monocytes because these cells, when cultured in polystyrene plates with ordinary tissue culture medium (which contains small amounts of (LPS)), express IL-1 and TNF mRNA and secrete small quantities of IL-1 and TNF protein [99 – 101]. Another reason that prompted investigators to select this model is that the secretion of cytokines from peripheral blood monocytes reflects the secretory activity of bone marrow mononuclear cells [101]. This is not surprising because the in vitro production of cytokines from cultured monocytes is a reflection of phenotypic characteristics acquired in response to local stimuli during their maturation in the bone marrow, and these characteristics are maintained after release into the circulation [102]. This phenomenon is thought to play an important role in providing the basis for tissue and functional specificity. Consequently, monocyte cytokine secretion is relevant to postmenopausal bone loss as it mirrors cytokine secretion from marrow resident cells of the monocyte macrophage lineage or monocytes that have homed to bone [103]. This hypothesis was first proved correct by studies of Pioli et al. showing that in Pagetic patients the secretion of IL-1 from blood monocytes correlates with that from bone marrow mononuclear cells [104] as well as from observations in rats and mice, where ovariectomy and estrogen replacement have been found to regulate the bone marrow mononuclear cell production of IL-1 and TNF [101,105]. It is also important to recognize that monocytes are the major source of IL-1 and TNF in the bone marrow [100]. Moreover, the anatomical proximity of mononuclear cells to remodeling loci, the capacity to secrete numerous products
all recognized for their effects on bone remodeling, and the expression of integrin receptors [106], which make these cells capable of adhering to the bone matrix, make them likely candidates as participants in skeletal remodeling. Investigation of the monocytic production of IL-1 led to the discovery that monocytes of patients with “highturnover” osteoporosis secrete higher IL-1 activity than those from both patients with “low turnover” osteoporosis and, indeed, those from normal subjects [72]. Since increased bone turnover is characteristic of the early postmenopausal period, these data suggested the hypothesis that the bone sparing effect of estrogen is related to its ability to block the production of IL-1 from cells of the monocytic lineage. This hypothesis was first tested in studies designed to investigate the effect of natural menopause and estrogen/progesterone replacement on the monocytic production of IL-1. The results showed that IL-1 activity increases after menopause in both normal and osteoporotic women. However, whereas IL-1 activity in normal women spontaneously returned to premenopausal levels within 7 years after menopause, in osteoporotic subjects the increase in IL-1 activity lasted up to 15 years after menopause [73]. As a result, the finding of increased IL-1 activity 8 – 15 years after menopause is characteristic of women with postmenopausal osteoporosis. The data also showed that treatment of women in both the early and the late postmenopausal periods with estrogen and progesterone normalizes IL-1 activity within the first month of treatment. Similar effects of menopause have also been documented for TNF and GM-CSF. The latter is a cytokine recognized as a potent stimulator of osteoclastogenesis [74]. The increased production of cytokines associated with estrogen withdrawal occurs in a secular fashion consistent with
91
CHAPTER 41 Postmenopausal Osteoporosis
a direct causal role of these factors on postmenopausal bone loss. This was demonstrated by analyzing the time course of changes in cytokine secretion and markers of bone turnover in normal women undergoing bilateral ovariectomy. Using this strategy it was demonstrated that the monocytic secretion of GM-CSF increases within 1 week after ovariectomy.
This is followed by a marked increase in TNF and IL-1 at 2 weeks postsurgery. The increase in the latter two cytokines is associated with a concurrent increase in biochemical indices of bone resorption [75] (Fig. 4). Initiation of estrogen replacement therapy at 1 month after ovariectomy results in the rapid normalization of cytokine production [75]. Subsequent
FIGURE 4 Effect of ovariectomy and subsequent estrogen/progesterone replacement therapy on the monocytic production of cytokines and biochemical indices of bone turnover.
92
ROBERTO PACIFICI
studies confirmed that natural and surgical menopause are associated with an increased production of IL-1 and TNF from peripheral blood and bone marrow monocytes [76,107 – 109]. An increased mononuclear cell production of IL-6 has also been reported after ovariectomy [108]. Since IL-1 and TNF are powerful stimulators of IL-6 production [90,110], the latter is likely to reflect the impact of the higher levels of IL-1 and TNF induced by ovariectomy. That the increased monocytic production of cytokines plays a direct role in inducing bone resorption was later demonstrated by Cohen-Solal et al. [111] by examining the bone resorption activity of monocyte supernatants obtained from pre- and postmenopausal women. Using this approach it was found that the culture media of monocytes obtained from postmenopausal women has a higher in vitro bone resorption activity than that from either premenopausal women or estrogen treated postmenopausal women. The increased bone resorption activity of media from postmenopausal subjects is blocked by the addition of IL-1ra and anti-TNF antibody. Recent studies of bone marrow supernatants from estrogen deficient mice by Suda and coworkers have also indicated that IL-1 plays a dominant role in mediating the impact of estrogen withdrawal on bone resorption. Antibodies to IL-1a [the dominant IL-1 species in mice] but not antibodies to many other cytokines completely blocked the bone resorption activity of the monocyte conditioned medium. Antibodies to IL-1b, IL-6, and IL-6 receptor resulted in a partial neutralization of bone resorption activity. Thus, it is likely that IL-1 and TNF account for most of the resorption activity produced by cultured monocytes. Still to be determined is whether this effect is direct or mediated by other factors produced in response to IL-1 and TNF. More direct evidence in favor of cause – effect relationship between increased production of IL-1 and TNF (and IL-6) and postmenopausal osteoporosis is also provided by the findings of Ralston demonstrating that IL-1, TNF, and IL-6 mRNAs are expressed more frequently in bone cells from untreated postmenopausal women than in those from women on estrogen replacement [90].
E. Effect of Menopause on the Production of IL-1 Receptor Antagonist IL-1 bioassays are based on the measurement of the proliferation activity of IL-1-dependent cell lines. Thus, IL-1 bioactivity is stimulated by IL-1 and inhibited by IL-1ra. Therefore, IL-1 bioactivity reflects closely the IL-1/IL-1ra ratio. Since mammalian cells secrete IL-1ra along with IL-1, the regulatory effects of ovarian steroid on IL-1 bioactivity may involve both IL-1 ( or ) and IL-1ra. Studies have addressed this issue and revealed that estrogen and progesterone downregulate the production of both IL-1 and IL-1ra [64]. In contrast, estrogen and progesterone have no inhibitory effects on the secretion of IL-1. Interestingly, in
normal women the decrease in IL-1 bioactivity that accompanies the passage of time since menopause is associated with a parallel increase in the secretion of IL-1ra. Thus, in normal women the increasing production of IL-1ra which accompanies the passage of time since menopause is likely to help restore normal monocytic IL-1 bioactivity after the menopause. IL-1 is a powerful autocrine factor. In fact, the IL-1 produced by monocytes binds to IL-1 receptors expressed on the monocyte surface and further stimulates IL1 secretion [112]. Since this process is inhibited by IL-1ra, the progressive postmenopausal increase in IL-1ra secretion observed in nonosteoporotic women may also help to explain the parallel decrease in the secretion of IL-1b observed in these subjects as time elapses from menopause. As discussed above, in osteoporotic women the production of IL-1 bioactivity is increased for a length of time twice as long as in normal women. This is associated with an increased secretion of IL-1 which persists as long as the increase in IL-1 bioactivity. Interestingly, the levels of IL-1ra measured in osteoporotic women are higher than those of normal women, but do not change with the passage of time since menopause [64]. Thus, in osteoporotic women IL-1 bioactivity appears to be primarily regulated by changes in the production of IL-1. Since only a small fraction of the cytokines released into the bone microenvironment escape into the systemic circulation, studies based on the measurement of serum cytokine levels have been, for the most part, unrewarding. However, the recent development of supersensitive cytokine assays has made it possible to document that the serum IL-1/IL1ra ratio is significantly higher in women with postmenopausal osteoporosis than in their normal counterparts [113]. The use of these sensitive assays has also led to the demonstration that the rate of bone loss in osteoporotic women correlates inversely with serum IL-1ra levels [114]. Taken together, these data demonstrate that a modulatory action of estrogen and progesterone on the secretion of IL1ra contributes to the events of the menopause and the effects of hormone replacement on IL-1 bioactivity. The molecular mechanism by which estrogen and menopause regulate the monocytic production of IL-1ra remains to be defined. The local microenvironment is known to condition the production of IL-1ra. For example, alveolar macrophages from patients with interstitial lung disease produce more IL-1ra than those from normal controls [115]. It is likely, therefore, that the increased bone resorption induced by IL-1 and other cytokines after the menopause may lead to the release in the bone microenvironment of factors which, in turn, stimulate the secretion of IL-1ra. One such a factor is TGF- [116], a constituent of the bone matrix released locally upon activation of osteoclastic bone resorption [117,118]. The differences in the secretory pattern of IL-1ra observed between normal and osteoporotic women could, indeed, result from the more intense bone resorption and the resulting more
93
CHAPTER 41 Postmenopausal Osteoporosis
TABLE 3
Effect of Cytokine Inhibition on Bone Mass and Bone Turnover in Ovariectomized Rats and Mice in the First Month after Surgery (Early Postovariectomy Period) TNFbp
IL-1ra TNFbp
Anti-IL-6 Ab
Anti-M-CSF Ab
IL-1ra Prevents ovx-induced bone loss Blocks osteoclast formation Blocks mature osteoclasts
Stimulate bone formation
abundant release of TGF- which characterize the postmenopausal period of women with osteoporosis [119]. Since altered T4/T8 lymphocyte ratio [120,121] and abnormal mixed leukocyte reactions have been reported in osteoporotic patients [122], it is conceivable that specific monocyte phenotypes characterized by the ability to produce constitutively high amounts of IL-1ra may be preferentially expressed in osteoporotic patients. Should this be the case, the difference in IL-1ra levels observed between normal and osteoporotic patients could be related to intrinsic differences in the prevailing monocyte population.
F. Cytokine Inhibitors and Transgenic Mice: Tools For Investigating the Contribution of Candidate Factors to Ovariectomy-Induced Bone Loss Since several cytokines are under hormonal control and exhibit overlapping biological effects, analysis of cytokine
expression and secretion in bone and bone marrow cells is unlikely to provide definite evidence in favor of a cause – effect relationship between increased cytokine production and postmenopausal bone loss. However, direct demonstration that cytokines mediate the impact of estrogen deficiency on bone can be achieved with the use of genetic models and specific cytokine antagonists, such as the IL-1 antagonist, IL-1ra, and the TNF antagonist, TNF binding protein (TNFbp). Lorenzo et al. have shown that mice insensitive to IL-1 due to the lack of IL-1 receptor type I are protected against ovx-induced bone loss [60]. These findings confirmed earlier studies conducted by treating ovariectomized rats with IL-1ra beginning either at the time of surgery (early postovariectomy period) or 4 weeks later (late postovariectomy period) [107]. These experiments revealed that the functional block of IL-1 has distinct effects in both periods (see Fig. 5 and Tables 3 and 4). In fact, in the second month after ovariectomy, treatment with IL-1ra completely blocked bone loss, duplicating the effect of estrogen. In contrast, in the first month after ovariectomy, bone loss
Effect (mean SEM] of IL-1ra treatment on the bone mineral density (BMD) of the distal femur in ovx rats. Results are expressed as percentage change from baseline. When treatments were started at the time of ovariectomized (left) rats treated with IL-1ra () had a smaller decrease (*P 0.005) in BMD than rats treated with IL-1ra vehicle (). However, a complete prevention of bone loss was obtained with estrogen () but not IL-1ra. When treatments were started 4 weeks after ovariectomy (right) both IL-1ra and estrogen were equally effective in completely preventing additional bone loss.
FIGURE 5
94
ROBERTO PACIFICI
TABLE 4
Effect of Cytokine Inhibition on Bone Mass and Bone Turnover in Ovariectomized Rats and Mice in the Second Month after Ovariectomy (Late Postovariectomy Period) IL-1ra
TNFbp
IL-1ra TNFbp
Anti-IL-6 Ab
Anti-M-CSF Ab
Prevents ovx-induced bone loss
a
a
?
?
Blocks osteoclast formation
a
a
?
?
Blocks mature osteoclasts
?
?
Stimulate bone formation
?
?
a
Preliminary unpublished data.
was completely prevented by estrogen replacement therapy and decreased by about 40% by IL-1ra treatment. These findings indicated that cytokines produced independently of IL-1 contribute to induce bone loss in the early postovariectomy period. Since IL-1 and TNF have powerful additive and synergistic effects in many systems, TNF appeared to be the most likely candidate factor. That TNF contributes to bone loss in the early postovariectomy period was demonstrated by treating ovariectomized rats with IL-1ra, TNFbp, and a combination of the two inhibitors for 2 weeks starting at the time of surgery [123]. This critical experiment demonstrated that while treatment with either IL-1ra or TNFbp alone partially prevented ovariectomy-induced bone loss, complete bone sparing was achieved when ovariectomized rats were treated simultaneously with IL-1ra and TNFbp (Fig. 6). The mechanism responsible for the bone sparing effects of IL-1ra and TNFbp was elucidated by analyzing biochemical and histomorphometric indices of bone turnover in the early and late postovariectomy period. In the immediate postovariectomy period, urinary excretion of deoxypyridinium (DPD), a marker of bone resorption which reflects the number and the functional activity of mature osteoclasts [124], was decreased by both IL-1ra and TNFbp. However, an inhibitory effect comparable to that of estrogen was achieved only when ovariectomized rats were treated simultaneously with both inhibitors [123]. In contrast, in the second month after ovariectomy, treatment with IL-1ra alone was sufficient to normalize DPD excretion [107]. While trabecular bone loss prevails in the late postovariectomy period, cortical bone is preferentially lost in mature rats in the early postovariectomy period. This is typically associated with an increase in the number of osteoclasts adhering to endocortical, but not trabecular surfaces. This increase is completely prevented by both IL-1ra and TNFbp [123]. However, there is no additive effects of the two inhibitors, thus suggesting that inhibition of osteoclastogenesis is not the mechanism accounting for the more potent bone sparing effect of combined IL-1ra TNFbp treatment.
Since IL-1ra and TNFbp have significant additive antiresorptive effects, but have no additive effects in blocking osteoclastogenesis, inhibition of osteoclast activation appears to be the main mechanism of the potent antiresorptive effects of combined treatment with IL-1ra and TNFbp in the early postovariectomy period. In contrast, in the late postovariectomy period treatment with IL-1ra alone is sufficient for completely arresting bone loss and for normalizing the number of adherent osteoclasts [107]. Similarly, since inhibition of both IL-1 and TNF is required to prevent the effects of ovariectomy on bone
FIGURE 6 Effect [mean SEM] of IL-1ra and TNFbp treatment on the bone mineral density (BMD) of the distal femur in ovx rats. Results are expressed as percentage change from baseline. ****P 0.05 compared to baseline. *P 0.05 compared to BSA treated ovx rats. **P 0.05 compared to BSA-, TNFbp-, and IL-1ra-treated ovx rats.
CHAPTER 41 Postmenopausal Osteoporosis
resorption, IL-1 and TNF likely have independent and redundant stimulatory effects on osteoclast activation, the primary mechanism by which bone resorption acutely increases early after ovariectomy. Inhibition of either IL-1 or TNF is sufficient to block the postovariectomy increase in osteoclastogenesis. Thus, IL-1 and TNF must possess either synergistic or sequential effects on osteoclastogenesis, the primary mechanism underlying long-term elevations of bone resorption [125,126]. Analysis of changes in bone density which follow 1 month treatment with either estrogen or IL-1ra provides further insight on the mechanism of action of IL-1ra [126]. Cessation of estrogen therapy is followed by a rapid resumption of bone loss. In contrast, ovariectomized rats treated with IL-1ra maintain stable bone density for the first 4 weeks after the completion of the treatment. In these rats bone loss resumes only 6 weeks after discontinuation of the IL-1ra treatment. Estrogen regulates both osteoclastogenesis and osteoclast activation, although the short time between estrogen withdrawal and the onset of bone loss points to a “downstream” event, such as osteoclast activation, as the functionally dominant mechanism by which estrogen blocks bone resorption. In contrast, the long interval between stopping IL-1ra treatment and induction of bone loss suggests that in the late postovariectomy period the functional block of IL-1 with IL-1ra affects mainly “upstream” events, such as the proliferation and differentiation of osteoclast precursors. Consequently, it appears that in the last postovariectomy period, IL-1ra is more potent in blocking osteoclastogenesis than in inhibiting osteoclast activation. Since the active life span of rat osteoclasts is only 2.5 days [127], a constant supply of new precursor cells is needed to replenish aged osteoclasts at each resorptive site [128]. Thus, in ovariectomized rats treated with IL-1ra the persistent activation of mature osteoclasts is likely to lead to an impoverishment of those osteoclast precursors downstream from the IL-1-dependent step(s). As a result, the time necessary for reexpansion of the osteoclastic pool is likely to account for the lag time between discontinuation of IL-1ra treatment and the resumption of bone resorption. Histomorphometric studies also demonstrated important effects of IL-1ra and TNFbp on bone formation [107, 129].Two weeks after surgery there were no significant differences in trabecular and cortical bone formation rate between ovariectomy and sham operated rats. Interestingly, however, treatment with either IL-1ra or TNFbp induced a marked increase in bone formation rate in ovariectomized but not in sham-operated rats. This suggests that inhibition of endocortical bone formation resulting from high levels of IL-1 and TNF (characteristic of the early postovariectomy period) counteracts and masks direct stimulatory effects of ovariectomy on bone formation. In contrast, in the late postovariectomy period, bone formation is increased in
95 the trabecular but not in the cortical bone and in this time period neither IL-1ra nor TNFbp have significant effects on this index. Thus, when taken together, the data support the hypothesis that estrogen deficiency modulates bone resorption via an IL-1/TNF-dependent pathway and bone formation via a complex mechanism which involves an IL-1/TNF-independent stimulatory effect and an IL-1/TNF-mediated inhibitory effect. Early after ovariectomy the dominant phenomena mediated by IL-1 and TNF are the stimulation of osteoclast activity and the inhibition of bone formation. As time progresses from ovariectomy, the IL-1- and TNFdependent inhibition of bone formation subsides while the most important effect of these factors becomes the induction of osteoclastogenesis. These initial observations about the causal role of TNF were confirmed by Ammann et al., who reported that transgenic mice insensitive to TNF due to the overexpression of soluble TNF receptor, are also protected against ovxinduced bone loss [62]. Finally, an orally active inhibitor of IL-1 and TNF production was also shown to completely prevent bone loss in ovx rats [130]. Although the finding that functional block of either IL-1 or TNF is sufficient to prevent ovx-induced bone loss may appear to be difficult to reconcile, it should be emphasized that in most biological systems IL-1 and TNF have potent synergistic effects. Thus, the functional block of one of these two cytokines elicits biological effects identical to those induced by the block of both IL-1 and TNF. The longterm stimulation of bone resorption which follows ovx is sustained primarily by an expansion of the osteoclastic pool. Since OC formation is synergistically stimulated by IL-1 and TNF [101], it is not surprising that long-term inhibition of either IL-1 or TNF results in complete prevention of ovx induced bone loss. While studies with transgenic mice and inhibitors of IL-1 and TNF have consistently demonstrated that IL-1 and TNF are key inducers of bone loss in ovx animals, investigations aimed at assessing the contribution of IL-6 to ovxinduced bone loss have yielded conflicting results. In favor of a causal role for IL-6 in ovx induced bone loss is the report of Poli et al., indicating that IL-6 knock out mice are protected against the loss of trabecular bone induced by ovx [61]. By contrast are studies demonstrating that osteoporosis is not a feature of transgenic mice overexpressing IL-6 [131]. Studies have also been conducted by injecting an antibody neutralizing IL-6 in ovx mice. Neutralizing IL-6 prevents the increase in OC formation induced by estrogen deficiency [50,129] but does prevent ovx-induced bone loss and does not decrease in vivo bone resorption [129]. These findings confirm that IL-6 contributes to the expansion of the osteoclastic pool induced by ovx. However, this cytokine does not appear to be the dominant factor in inducing bone loss in estrogen deficient mice.
96
FIGURE 7 Treatment with the anti M-CSF Ab 5A1 Ab prevents ovx induced bone loss. Results (mean SEM) are expressed as percentage change from baseline *P 0.05 compared to baseline and to any other group. Recent studies have also demonstrated that the functional block of M-CSF by the anti M-CSF antibody 5A1 completely prevents ovx-induced bone loss in mice (Fig. 7) [132]. That M-CSF is another cytokine which plays a key role in ovx induced bone loss was further demonstrated by examining mice lacking the transcription factor Egr-1 [133]. Egr-1-deficient mice produce maximal amounts of M-CSF both in the presence and in the absence of estrogen [94]. Thus, ovx does not stimulate osteoclast formation in these mice, as it fails to further stimulate M-CSF production. Importantly, Egr-1 deficient mice are completely protected against ovx-induced bone loss, a finding that confirms the relevance of M-CSF [132]. No studies have been conducted to determine the effects of ovx in mice insensitive to RANKL, although one would predict that these animal will sustain significant bone loss due to the stimulated production of TNF. The role of IL-1, IL-6, TNF, M-CSF, and RANKL in osteoclastogenesis has been directly investigated using murine bone marrow cultures obtained from ovariectomized mice. Ovariectomy not only increases the number of bone marrow cells, but also increases the number of osteoclasts generated by ex vivo cultures of bone marrow cells [134]. IL-1ra and TNFbp both completely prevent the increase in osteoclastogenesis induced by ovariectomy [129]. Neither IL-1ra nor TNFbp decrease osteoclast formation in
ROBERTO PACIFICI
sham-operated mice. Osteoclast formation is also decreased, in part, by the anti-IL-6 antibody 20F3. However, the anti-IL-6 antibody is less effective than IL-1ra and TNFbp and, more importantly, decreases osteoclastogenesis in both ovariectomized and sham-operated mice [129], indicating that the contribution of IL-6 to osteoclastogenesis does not increase with estrogen deficiency. The formation of osteoclasts in ex vivo cultures of bone marrow cells from ovariectomized mice is also blocked by in vitro treatment with IL-1ra or TNFbp. In contrast, in vitro treatment with the anti-IL-6 antibody 20F3 has no effects in cultures from either ovariectomized or sham-operated mice. Another important difference between these inhibitors is that in vivo treatment with IL-1ra and TNFbp also decreases the urinary excretion of DPD in a manner similar to that of estrogen, whereas the anti-IL-6 antibody does not. In contrast, when in vitro bone resorption is evaluated by examining the effects of the three inhibitors on the formation of resorption lacunae, it appears that IL-1ra, anti-IL-6 antibody, and TNFbp all inhibit the formation of resorption pits [101]. Since the regulatory role of IL-6 is limited to the initial steps of the osteoclast differentiation process [135], it could be that the block of IL-6 in vivo is insufficient to prevent the complete maturation and activation of those cells which are downstream with respect to the IL-6 -dependent steps According to this hypothesis, the lack of change in DPD excretion with anti-IL-6 antibody treatment would reflect the maintenance of an unaltered pool of active, mature osteoclasts. Conversely, the decreased pit formation observed with the IL-6 block is likely to reflect the decreased bone marrow content of osteoclast precursors and the resulting decrease in the number of cells which reach functional maturity in vitro [32]. From these data it appears reasonable to hypothesize that inhibition of IL-1 and TNF blocks bone resorption in vivo and in vitro because, at least in rodent, these cytokines regulate early and late steps of osteoclast maturation. In vivo treatment of ovx mice with anti M-CSF antibody prevents the effects of ovx on ex-vivo osteoclast formation and bone resorption [132], in a manner similar to treatment with IL-1 and TNF antagonists. These data are consistent with the notion that estrogen deficiency increases M-CSF production indirectly, via an IL-1 and TNF-mediated mechanism [93, 94]. Recent studies have demonstrated that the presence of severe osteopetrosis due to complete lack of osteoclasts in mice lacking either RANKL or the RANKL receptor RANK. However, the effects of ovariectomy and/or estrogen deficiency in these animals remain to be investigated.
III. SUMMARY AND CONCLUSIONS The mechanism(s) of the bone sparing effects of estrogen appears to be particularly complex as it involves the
CHAPTER 41 Postmenopausal Osteoporosis
regulated production of cytokines from hematopoietic cells and bone cells [103,136] and the responsiveness of stromal cells to these cytokines. In addition, the contribution of specific factors to postmenopausal bone loss appears to vary as the system adapts over time to the hormonal withdrawal. Although many details of this process remain to be defined, it is now clearly established that estrogen downregulates the production of proosteoclastogenic and antiosteoclastogenic factors. Among them are IL-1, IL-6, TNF, M-CSF, OPG, and TGF-. Uncertainty remains over the role [if any] of RANKL in the mechanism by which estrogen prevents bone loss. The exact contribution of IL-6 also remains unclear because of insufficient data demonstrating that the block of IL-6 decreases bone resorption in vivo and bone loss in a bona fide experimental model of postmenopausal osteoporosis. However, the exact role of RANKL and IL-6 is likely to be defined in the near future. At the present time RANKL and M-CSF should be regarded as the factors responsible for osteoclast renewal under unstimulated conditions. The enhanced osteoclastogenesis and the increased osteoclastic bone resorption leading to postmenopausal bone loss results from stimulated production of inflammatory cytokines. These factors increase the production of M-CSF and RANKL and, in the case of TNF, directly synergize with RANKL to maximize osteoclast formation. Remarkable progress has been accomplished in clarifying the mechanism of the bone sparing effect of estrogen in animal models. A more challenging task will be to demonstrate the relevance of the mechanisms described above in human subjects.
References 1. G. A. Greendale and H. L. Judd, The menopause: Health implications and clinical management. Geriatrics Soc. 41, 426–436 (1993) [Review]. 2. R. J. Chang and H. L. Judd, The ovary after menopause. Clin. Obstet. Gynecol. 24, 181 – 191 (1981). 3. H. W. Cole, S. Chandrasekhar, H. M. Hsuing, M. L. Heiman, and M. Sato, Beneficial effects of ethynyl estradiol (EE2) and raloxifene in the ovariectomized (ovx) rat are dependent on an intact pituitary. J. Bone Miner. Res. 9, Abs 57 (1994). 4. R. F. Power, S. K. Mani, J. Codina, O. M. Conneely, and B. W. O’Malley, Dopaminergic and ligand-independent activation of steroid hormone receptors. Science 254, 1636 – 1639 (1991). 5. National Institutes of Health Consensus Conference. J. Am. Med. Assoc. 252, 799 – 802 (1984). 6. B. L. Riggs and L. J. Melton, Medical progress: Involutional osteoporosis. N. Engl. J. Med. 314, 1676 – 1684 (1986). 7. B. L. Riggs and L. J. Melton, Evidence for two distinct syndromes of involutional osteoporosis. Am. J. Med. 75, 899 – 901 (1986). 8. L. Nilas and C. Christiansen, Bone mass and its relationship to age and the menopause. J. Clin. Endocrinol. Metab. 65, 697 – 702 (1987). 9. B. L. Riggs, H. W. Wahner, W. L. Dunn, R. B. Mazess, K. P. Offord, and L. J. Melton, Differential changes in bone mineral density of the appendicular and the axial skeleton with aging. J. Clin. Invest. 67, 328 – 335 (1981).
97 10. C. Slemenda, S. L. Hui, C. Longcope, and C. C. Johnston, Sex steroids and bone mass. A study of changes about the time of menopause. J. Clin. Invest. 80, 1261 – 1269 (1987). 11. H. K. Genant, C. E. Cann, B. Ettinger, and G. S. Gordan, Quantitative computed tomography of vertebral spongiosa: A sensitive method for detecting early bone loss after oophorectomy. Ann. Intern. Med. 97, 699 – 705 (1982). 12. C. C. Johnston, Jr., S. L. Hui, R. M. Witt, R. Appledorn, R. S. Baker, and C. Longcope, Early menopausal changes in bone mass and sex steroids. J. Clin. Endocrinol. Metab. 61, 905 – 911 (1985). 13. H. Ohta, K. Makita, Y. Suda, T. Ikeda, T. Masuzawa, and S. Nozawa, Influence of oophorectomy on serum levels of sex steroids and bone metabolism and assessment of bone mineral density in lumbar trabecular bone by QCT-C value. J. Bone Miner. Res. 7, 659 – 665 (1992). 14. H. Ohta, T. Ikeda, T. Masuzawa, K. Makita, Y. Suda, and S. Nozawa Differences in axial bone mineral density, serum levels of sex steroids, and bone metabolism between postmenopausal and age- and body size-matched premenopausal subjects. Bone 14, 111 – 116 (1993). 15. B. Ettinger, H. K. Genant, and E. C. Cann, Long-term estrogen replacement therapy prevents bone loss and fractures. Ann. Intern. Med. 102, 319 – 324 (1985). 16. R. Lindsay, D. M. Hart, C. Forrest, and C. Baird, Prevention of spinal osteoporosis in oophorectomised women. Lancet 2, 1151 – 1154 (1980). 17. A. Horsman, M. Jones, R. Francis, and B. E. C. Nordin The effect of estrogen dose on postmenopausal bone loss. N. Eng. J. Med. 309, 1405 – 1407 (1983). 18. N. A. Morrison, J. C. Qi, A. Tokita, P. J. Kelly, L. Crofts, T. V. Nguyen, P. N. Sambrook, and J. A. Eisman, Prediction of bone density from vitamin D receptor alleles. Nature 367, 284 – 287 (1994). 19. S. C. Manolagas, and R. L. Jilka, Bone marrow, cytokines, and bone remodeling. N. Engl. J. Med. 332, 305 – 311 (1995). 20. J. Chow, J. H. Tobias, K. W. Colston, and T. J. Chambers, Estrogen maintains trabecular bone volume in rats not only by suppression of bone resorption but also by stimulation of bone formation. J. Clin. Invest. 89, 74 – 78 (1992). 21. S. D Bain, S. C. Bailey, D. L. Celino, M. M. Lantry, and M. W. Edwards, High-dose estrogen inhibits bone resorption and stimulates bone formation in the ovariectomized mouse. J. Bone Miner. Res. 8, 435 – 442 (1993). 22. R. Pacifici, Estrogen, cytokines and pathogenesis of postmenopausal osteoporosis. J. Bone Miner. Res. 11, 1043 – 1051 (1996). 23. G. D. Roodman, Advances in bone biology: the osteoclast. Endocr. Rev. 17, 308 – 332 (1996). 24. Y. Y. Kong, U. Feige, I. Sarosi, B. Bolon, A. Tafuri, S. Morony, C. Capparelli, J. Li, R. Elliott, S. McCabe, T. Wong, G. Campagnuolo, E. Moran, E. R. Bogoch, G. Van, L. T. Nguyen, P. S. Ohashi, D. L. Lacey, E. Fish, W. J. Boyle, and J. M. Penninger, Activated T cells regulate bone loss and joint destruction in adjuvant arthritis through osteoprotegerin ligand. Nature 402, 304 – 309 (1999). 25. I. R. Dissanayake, G. R. Goodman, E. R. Baker, S. Pun, L. Ciui, Y. A. Ma, S. S. Jee.W, S. N. Popoff, and S. Epstein, The role of the B lymphocyte in estrogen deficient and normal bone. J. Bone Miner. Res. 12 (Suppl. 1), F467-(Abs.) (1997). 26. R. G. Erben, S. Raith, J. Eberle, and M. Stangassinger, Ovariectomy augments B lymphopoiesis and generation of monocytemacrophage precursors in rat bone marrow. Am. J. Physiol. 274, E476 – 483 (1998). 27. T. Masuzawa, C. Miyaura, Y. Onoe, K. Kusano, H. Ohta, S. Nozawa, and T. Suda. Estrogen deficiency stimulates B lymphopoiesis in mouse bone marrow. J. Clin. Invest. 94, 1090 – 1097 (1994). 28. D. L. Lacey, E. Timms, H. L. Tan, M. J. Kelley, C. R. Dunstan, T. Burgess, R. Elliott, A. Colombero, G. Elliott, S. Scully, H. Hsu,
98
29.
30.
31.
32. 33.
34.
35.
36.
37.
38.
39.
40.
41.
ROBERTO PACIFICI J. Sullivan, N. Hawkins, E. Davy, C. Capparelli, A. Eli, Y. X. Qian, S. Kaufman, I. Sarosi, V. Shalhoub, G. Senaldi, J. Guo, J. Delaney, and W. J. Boyle, Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell 93, 165 – 176 (1998). Y. Y. Kong, H. Yoshida, I. Sarosi, H. L. Tan, E. Timms, C. Capparelli, S. Morony, A. J. Oliveira-dos-Santos, G. Van, A. Itie, W. Khoo, A. Wakeham, C. R. Dunstan, D. L. Lacey, T. W. Mak, W. J. Boyle, and J. M. Penninger, RANKL is a key regulator of osteoclastogenesis, lymphocyte development and lymph-node organogenesis. Nature 397, 315 – 323 (1999). H. Yasuda, N. Shima, N. Nakagawa, K. Yamaguchi, M. Kinosaki, S. Mochizuki, A. Tomoyasu, K. Yano, M. Goto, A. Murakami, E. Tsuda, T. Morinaga, K. Higashio, N. Udagawa, N. Takahashi, and T. Suda, Osteoclast differentiation factor is a ligand for osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical to TRANCE/RANKL. Proc. Natl. Acad. Sci. USA 95, 3597 – 3602 (1998). B. R. Macdonald, G. R. Mundy, S. Clark, E. A. Wang, T. J. Kuehl, E. R. Stanley, and G. D. Roodman, Effects of human recombinant CSF-GM and highly purified CSF-1 on the formation of multinucleated cells with osteoclast characteristics in long-term bone marrow cultures. J. Bone Miner. Res. 1, 227 – 232 (1986). T. Suda, N. Takahashi, and T. J. Martin, Modulation of osteoclast differentiation. Endocr. Rev. 13, 66 – 80 (1992). S. Tanaka, N. Takahashi, N. Udagawa, T. Tamura, T. Akatsu, E. R. Stanley, T. Kurokawa, and T. Suda, Macrophage colony-stimulating factor is indispensable for both proliferation and differentiation of osteoclast progenitors. J Clin. Invest. 91, 257 – 263 (1993). W. S. Simonet, D. L. Lacey, C. R. Dunstan, M. Kelley, M. S. Chang, R. Luthy, H. Q. Nguyen, S. Wooden, L. Bennett, T. Boone, G. Shimamoto, M. DeRose, R. Elliott, A. Colombero, H. L. Tan, G. Trail, J. Sullivan, E. Davy, N. Bucay, L. Renshaw-Gegg, T. M. Hughes, D. Hill, W. Pattison, P. Campbell, and W. J. Boyle, Osteoprotegerin: A novel secreted protein involved in the regulation of bone density. Cell 89, 309 – 319 (1997) [see comments]. H. Hsu, D. L. Lacey, C. R. Dunstan, I. Solovyev, A. Colombero, E. Timms, H. L. Tan, G. Elliott, M. J. Kelley, I. Sarosi, L. Wang, X. Z. Xia, R. Elliott, L. Chiu, T. Black, S. Scully, C. Capparelli, S. Morony, G. Shimamoto, M. B. Bass, and W. J. Boyle, Tumor necrosis factor receptor family member RANK mediates osteoclast differentiation and activation induced by osteoprotegerin ligand. Proc. Natl. Acad. Sci USA 96, 3540 – 3545 (1999) [in process citation]. R. Felix, M. G. Cecchini, and H. Fleish, Macrophage colonystimulating factor restores in vivo bone resorption in the op/op osteopetrotic mouse. Endocrinology 127, 2592 – 2597 (1990). H. Kodama, M. Ose, S. Niida, and A. Yamasaki, Essential role of macrophage colony-stimulating factor in the osteoclast differentiation supported by stromal cells. J. Exp. Med. 173, 1291 – 1294 (1991). K. Fuller, J. M. Owens, C. J. Jagger, A. Wilson, R. Moss, and T. J. Chambers, Macrophage colony-stimulating factor stimulates survival and chemotactic behavior in isolated osteoclasts. J. Exp. Med. 178, 1733 – 1744 (1993). U. Sarma and A. M. Flanagan, Macrophage colony-stimulating factor induces substantial osteoclast generation and bone resorption in human bone marrow cultures. Blood 88, 2531 – 2540 (1996). T. Suda, N. Takahashi, N. Udagawa, E. Jimi, M. T. Gillespie, and T. J. Martin, Modulation of osteoclast differentiation and function by the new members of the tumor necrosis factor receptor and ligand families. Endocr. Rev. 20, 345 – 357 (1999). G. Hattersley, E. Dorey, M. A. Horton, and T. J. Chambers, Human macrophage colony-stimulating factor inhibits bone resorption by osteoclasts disaggregated from rat bone. J. Cell. Physiol. 137, 199 – 203 (1988).
42. M. Edwards, U. Sarma, and A. M. Flanagan, Macrophage colonystimulating factor increases bone resorption by osteoclasts disaggregated from human fetal long bones. Bone 22, 325 – 329 (1998). 43. R. L. Lees and J.N. Heersche, Macrophage colony stimulating factor increases bone resorption in dispersed osteoclast cultures by increasing osteoclast size. J. Bone Miner. Res. 14, 937 – 945 (1999). 44. M. Gowen, D. D. Wood, E. J. Ihrie, M. K. B. McGuire, and R. G. G. Russell, An interleukin-1-like factor stimulates bone resorption in vitro. Nature 306, 378 – 380 (1983). 45. J. A. Lorenzo, S. L. Sousa, C. Alander, L. G. Raisz, and C. A. Dinarello, Comparison of the bone-resorbing activity in the supernatants from phytohemaglutinin-stimulated human peripheral blood mononuclear cells with that of cytokines through the use of an antiserum to interleukin 1. Endocrinology 121, 1164 – 1170 (1987). 46. M. Gowen, D. D. Wood, and R. G. G. Russell, Stimulation of the proliferation of human bone cells in vitro by human monocyte products with interleukin-1 activity. J. Clin. Invest. 75, 1223 – 1229 (1985). 47. E. Canalis, Interleukin-1 has independent effects on deoxyribonucleic acid and collagen synthesis in cultures of rat calvariae. Endocrinology 118, 74 – 81 (1986). 48. P. Stashenko, F. E. Dewhirst, M. L. Rooney, L. A. Desjardins, and J.D. Heeley, Interleukin-1 is a potent inhibitor of bone formation in vitro. J. Bone Miner. Res. 2, 559 – 565 (1987). 49. G. Girasole, R. L. Jilka, G. Passeri, S. Boswell, G. Boder, D. C. Williams, and S. C. Manolagas, 17-estradiol inhibits interleukin-6 production by bone marrow-derived stromal cells and osteoblasts in vitro: A potential mechanism for the antiosteoporotic effect of estrogens. J. Clin. Invest. 89, 883 – 891 (1992). 50. R. L. Jilka, G. Hangoc, G. Girasole, G. Passeri, D. C. Williams, J. S. Abrams, B. Boyce, H. Broxmeyer, and S. C. Manolagas, Increased osteoclast development after estrogen loss: mediation by Interleukin6. Science 257, 88 – 91 (1992). 51. G. Passeri, G. Girasole, R. L. Jilka, and S. C. Manolagas, Increased interleukin-6 production by murine bone marrow and bone cells after estrogen withdrawal. Endocrinology 133, 822 – 828 (1993). 52. D. R. Bertolini, G. E. Nedwin, T. S. Bringman, D. D. Smith, and G. R. Mundy, Stimulation of bone resorption and inhibition of bone formation in vitro by human tumor necrosis factor. Nature 319, 516 – 518 (1986). 53. B. M. Thomson, G. R. Mundy, and T. J. Chambers, Tumor necrosis factor and induce osteoblastic cells to stimulate osteoclastic bone resorption. J. Immunol. 138, 775 – 779 (1987). 54. L. C. Hofbauer, D. L. Lacey, C. R. Dunstan, T. C. Spelsberg, B. L. Riggs, and S. Khosla, Interleukin-1beta and tumor necrosis factoralpha, but not interleukin-6, stimulate osteoprotegerin ligand gene expression in human osteoblastic cells Bone 25, 255 – 259 (1999). 55. C. A. O’Brien, I. Gubrij, S. C. Lin, R. L. Saylors, and S. C. Manolagas, STAT3 activation in stromal/osteoblastic cells is required for induction of the receptor activator of NF-kappaB ligand and stimulation of osteoclastogenesis by gp130-utilizing cytokines or interleukin-1 but not 1,25-dihydroxyvitamin D3 or parathyroid hormone. J. Biol. Chem. 274, 19301 – 19308 (1999). 56. C. Thery, E. R. Stanely, and M. Mallat, Interleukin 1 and tumor necrosis factor-a stimulate the production of colony-stimulating factor 1 by murine astrocytes. J. Neurochem. 59, 1183 – 1186 (1992). 57. W. E. Fibbe, J. Van Damme, A. Billiau, H. M. Goselink, P. J. Vogt, P. R. Van Eeden, B. W. Altroch, and J.H.F. Falkenburg Intereleukin-1 induces human marrow stromal cells in long term culturs to produce G-CSF and M-CSF. Blood 71, 431 – 435 (1988). 58. H. Takai, M. Kanematsu, K. Yano, E. Tsuda, K. Higashio, K. Ikeda, K. Watanabe, and Y. Yamada, Transforming growth factor-beta stimulates the production of osteoprotegerin/osteoclastogenesis inhibitory factor by bone marrow stromal cells. J. Biol. Chem. 273, 27091 – 27096 (1998).
99
CHAPTER 41 Postmenopausal Osteoporosis 59. K. Kobayashi, N. Takahashi, E. Jimi, N. Udagawa, M. Takami, S. Kotake, N. Nakagawa, M. Kinosaki, K. Yamaguchi, N. Shima, H. Yasuda, T. Morinaga, K. Higashio, T. J. Martin, and T. Suda, Tumor necrosis factor alpha stimulates osteoclast differentiation by a mechanism independent of the ODF/RANKL – RANK interaction. J. Exp. Med. 191, 275 – 286 (2000). 60. J. A. Lorenzo, A. Naprta, Y. Rao, C. Alander, M. Glaccum, M. Widmer, G. Gronowicz, J. Kalinowski, and C. C. Pilbeam, Mice lacking the type I interleukin-1 receptor do not lose bone mass after ovariectomy. Endocrinology 139, 3022 – 3025 (1998). 61. V. Poli, R. Balena, E. Fattori, A. Markatos, M. Yamamoto, H. Tanaka, G. Ciliberto, G. A. Rodan, and F. Costantini. Interleukin6 deficient mice are protected from bone loss caused by estrogen depletion. EMBO J. 13, 1189 – 1196 (1994). 62. P. Ammann, R. Rizzoli, J. P. Bonjour, S. Bourrin, J. M. Meyer, P. Vassalli, and I. Garcia, Transgenic mice expressing soluble tumor necrosis factor-receptor are protected against bone loss caused by estrogen deficiency. J. Clin. Invest. 99, 1699 – 1703 (1997). 63. P. Isomaki and J. Punnonen, Pro- and anti-inflammatory cytokines in rheumatoid arthritis. Ann. Med. 29, 499 – 507 (1997). 64. R. Pacifici, J. L. Vannice, L. Rifas, and R. B, Kimble. Monocytic secretion of interleukin-1 receptor antagonist in normal and osteoporotic women: effect of menopause and estrogen/progesterone therapy. J. Clin. Endocrinol. Metab. 77, 1135 – 1141 (1993). 65. L. C. Hofbauer, S. Khosla, C. R. Dunstan, D. L. Lacey, T. C. Spelsberg, and B. L. Riggs, Estrogen stimulates gene expression and protein production of osteoprotegerin in human osteoblastic cells. Endocrinology 140, 4367 – 4370 (1999). 66. M. J. Oursler, C. Cortese, P. Keeting, M. A. Anderson, S. K. Bonde, B. L. Riggs, and T. C. Spelsberg, Modulation of transforming growth factor-beta production in normal human osteoblast-like cells by 17 beta-estradiol and parathyroid hormone. Endocrinology 129, 3313 – 3320 (1991). 67. J. J. Weusten, M. A. Blankenstein, F. H. Gmelig-Meyling, H. J. Schuurman, L. Kater, and J. H. Thijssen. Presence of estrogen receptors in human blood mononuclear cells and thymocytes. Acta Endocrinol. 112, 409 – 414 (1986). 68. S. Gulshan, A. B. McCruden, and W. H. Stimson, Estrogen receptor in macrophages. Scand. J. Immunol. 31, 691 – 697 (1990). 69. M. J. Oursler, P. Osdoby, J. Pyfferoen, B. L. Riggs, and T. C. Spelsberg, Avian osteoclasts as estrogen target cells. Proc. Natl. Acad. Sci. USA 88, 6613 – 6617 (1991). 70. E. F. Eriksen, D. S. Colvard, N. J. Berg, M. L. Graham, K. G. Mann, T. C. Spelsberg, and B. L. Riggs, Evidence of estrogen receptors in normal human osteoblast-like cells. Science 241, 84 – 86 (1988). 71. J. H. M. Cohen, L. Danel, G. Cordier, S. Saez, and J.-P. Revillard, Sex steroid receptor in peripheral T cells: Absence of androgen receptors and restriction of estrogen receptor to CD8 positive cells. J. Immunol. 131, 2767 – 2772 (1983). 72. R. Pacifici, L. Rifas, S. Teitelbaum, E. Slatopolsky, R. McCracken, M. Bergfeld, W. Lee, L. V. Avioli, and W. A. Peck, Spontaneous release of interleukin 1 from human blood monocytes reflects bone formation in idiopathic osteoporosis. Proc. Natl. Acad. Sci. USA 84, 4616 – 4620 (1987). 73. R. Pacifici, L. Rifas, R. McCracken, I. Vered, C. McMurtry, L. V. Avioli, and W. A. Peck, Ovarian steroid treatment blocks a postmenopausal increase in blood monocyte interleukin 1 release. Proc. Natl. Acad. Sci. USA 86, 2398 – 2402 (1989). 74. R. Pacifici, C. Brown, L. Rifas, and L. V. Avioli, TNF and GM-CSF secretion from human blood monocytes: Effect of menopause and estrogen replacement. J. Bone Miner. Res. 5, Abs 145 (1990). 75. R. Pacifici, C. Brown, E. Puscheck, E. Friedrich, E. Slatopolsky, D. Maggio, R. McCracken, and L. V. Avioli, Effect of surgical menopause and estrogen replacement on cytokine release from
76.
77.
78.
79.
80.
81.
82.
83.
84.
85.
86.
87.
88.
89.
90.
91.
human blood mononuclear cells. Proc. Natl. Acad. Sci. USA 88, 5134 – 5138 (1991). C. E. Fiore, Falcidia.E, R. Foti, M. Motta, and C. Tamburino, Differences in the time course of the effects of oophorectomy in women on parameters of bone metabolism and Interleukin-1 levels in the circulation. Bone Miner. 20, 79 – 85 (1993). M. L. Polan, A. Daniele, and A. Kuo, Gonadal steroids modulate human monocyte interleukin-1 (IL-1) activity. Fertil. Steril. 49, 964 – 968 (1988). T. Sunyer, J. Lewis, and P. Osdoby, Estrogen decreases the steady state levels of the IL-1 signaling receptor (type I) while increasing those of the IL-1 decoy receptor (type II)] in human osteoclast-like cells. J. Bone Miner. Res. 12(Suppl. 1), Abs131 (1997). G. G. Kuiper, E. Enmark, M. Pelto-Huikko, S. Nilsson, and J. A. Gustafsson, Cloning of a novel receptor expressed in rat prostate and ovary. Proc. Natl. Acad. Sci. U.S.A. 93, 5925 – 5930 (1996). S. Mosselman, J. Polman, and R. Dijkema, ER beta: Identification and characterization of a novel human estrogen receptor. FEBS Lett. 392, 49 – 53 (1996). G. B. Tremblay, A. Tremblay, N. G. Copeland, D. J. Gilbert, N. A. Jenkins, F. Labrie, and V. Giguere, Cloning, chromosomal localization, and functional analysis of the murine estrogen receptor beta. Mol. Endocrinol. 11, 353 – 365 (1997). J. F. Couse, J. Lindzey, K. Grandien, J. A. Gustafsson, and K. S. Korach, Tissue distribution and quantitative analysis of estrogen receptor- alpha (ERalpha) and estrogen receptor-beta (ERbeta) messenger ribonucleic acid in the wild-type and ERalpha-knockout mouse. Endocrinology 138, 4613 – 4621 (1997). K. Paech, P. Webb, G. G. Kuiper, S. Nilsson, J. Gustafsson, P. J. Kushner, and T. S. Scanlan, Differential ligand activation of estrogen receptors ERalpha and ERbeta at AP1 sites [see comments]. Science 277, 1508 – 1510 (1997). S. Srivastava, M. N. Weitzmann, S. Cenci, F. P. Ross, S. Adler, and R. Pacifici. Estrogen decreases TNF gene expression by blocking JNK activity and the resulting production of c-jun and junD. J. Clin. Invest. 104, 503 – 513 (1999). J. An, R. C. Ribeiro, P. Webb, J. A. Gustafsson, P. J. Kushner, J. D. Baxter, and D. C. Leitman, Estradiol repression of tumor necrosis factor-alpha transcription requires estrogen receptor activation function-2 and is enhanced by coactivators. Proc. Natl. Acad. Sci. USA 96, 15161 – 15166 (1999). S. T. Pottratz, T. Bellido, H. Mocharia, D. Crabb, and S. Manolagas, 17-Estradiol inhibits expression of human interleukin-6 promoterreporter constructs by a receptor-dependent mechanism. J. Clin. Invest. 93, 944 – 950 (1994). A. Ray, K. E. Prefontaine, and P. Ray, Down-modulation of interleukin-6 gene expression by 17 beta-estradiol in the absence of high affinity DNA binding by the estrogen receptor. J. Biol. Chem. 269, 12940 – 12946 (1994). B. Stein and M. X. Yang, Repression of the interleukin-6 promoter by estrogen receptor is mediated by NF-kappa-B and C/EBP-BETA. Mol. Cell. Biol. 15, 4971 – 4979 (1995). D. Rickard, G. Russell, and M. Gowen, Oestradiol inhibits the release of tumour necrosis factor but not interleukin 6 from adult human osteoblasts in vitro. Osteoporosis Int. 2, 94 – 102 (1992). L. R. Chaudhary, T. C. Spelsberg, and B. L. Riggs, Production of various cytokines by normal human osteoblast-like cells in response to interleukin-1 and tumor necrosis factor-: lack of regulation by 17-estradiol. Endocrinology 130, 2528 – 2534 (1992). L. Rifas, J. S. Kenney, M. Marcelli, R. Pacifici, L. L. Dawson, S. Cheng, and L. V. Avioli, Production of interleukin-6 in human osteoblasts and human bone marrow stromal cells: Evidence that induction by interleukin-1 and tumor necrosis factor- is not regulated by ovarian steroids. Endocrinology 136, 4056 – 4067 (1995).
100 92. G. Girasole, M. Pedrazzoni, N. Giuliani, G. Passeri, and M. Passeri, Increased serum soluble interleukin-6 receptors levels are induced by ovariectomy, prevented by estrogen replacement and reversed by alendronate administration. J. Bone Miner. Res. 10, A86 (1995). 93. R. B. Kimble, S. Srivastava, F. P. Ross, A. Matayoshi, and R. Pacifici, Estrogen deficiency increases the ability of stromal cells to support osteoclastogenesis via an IL-1 and TNF mediated stimulation of M-CSF production. J. Biol. Chem. 271, 28890 – 28897 (1996). 94. S. Srivastava, M. N. Weitzmann, R. B. Kimble, M. Rizzo, M. Zahner, J. Milbrandt, F. P. Ross, and R. Pacifici, Estrogen blocks M-CSF gene expression and osteoclast formation by regulating phosphorylation of Egr-1 and iots interaction with Sp-1. J. Clin. Invest. 102, 1850 – 1859 (1998). 95. U. Sarma, M. Edwards, K. Motoyoshi, and A. M. Flanagan, Inhibition of bone resorption by 17beta-estradiol in human bone marrow cultures. J. Cell Physiol. 175, 99 – 108 (1998). 96. C. K. Lea, U. Sarma, and A. M. Flanagan. Macrophage colony stimulating-factor transcripts are differentially regulated in rat bone-marrow by gender hormones. Endocrinology 140, 273 – 279 (1999). 97. R. Kitazawa, S. Kitazawa, and S. Maeda, Promoter structure of mouse RANKL/TRANCE/RANKL/ODF gene. Biochim. Biophys. Acta. 1445, 134 – 141 (1999). 98. M. J. Oursler, M. J. Osteoclast synthesis and secretion and activation of latent transforming growth factor beta. J. Bone Miner. Res. 9, 443 – 452 (1994). 99. R. C. Fuhlbrigge, D. D. Chaplin, J. -M. Kiely, and E. R. Unanue, Regulation of interleukin 1 gene expression by adherence and lipopolysaccharide. J. Immunol. 138, 3799 – 3802 (1987). 100. C. A. Dinarello, Interleukin-1 and its biologically related cytokines. Adv. Immunol 44, 153 – 205 (1989). 101. R. Kitazawa, R. B. Kimble, J. L. Vannice, V. T. Kung, and R. Pacifici, Interleukin-1 receptor antagonist and tumor necrosis factor binding protein decrease osteoclast formation and bone resorption in ovariectomized mice. J. Clin. Invest. 94, 2397 – 2406 (1994). 102. A. L. Witsell and L. B. Schook, Macrophage heterogeneity occurs through a developmental mechanism. Proc. Natl. Acad. Sci. USA 88, 1963 – 1967 (1991). 103. M. Horowitz, Cytokines and estrogen in bone: Anti-osteoporotic effects. Science 260, 626 – 627 (1993). 104. G. Pioli, G. Girasole, M. Pedrazzoni, P. Sansoni, L. Davoli, G. Ciotti, A. Mantovani, and M. Passeri, Spontaneous release of Interleukin 1 (IL-1) from medullary mononuclear cells of Pagetic subjects. Calcified Tissue International 45, 257 – 259 (1989). 105. R. B. Kimble, A. B. Matayoshi, J. L. Vannice, and R. Pacifici, Long-term treatment with IL-1 receptor antagonist (IL-1ra) blocks bone loss in ovariectomized rats. J. Bone Miner. Res. 55, 260 – 265 (1994). 106. R. O. Hynes, Integrins; Versatility, modulation, and signaling in cell adhesion, Cell 69, 11 – 25 (1992). 107. R. B. Kimble, J. L. Vannice, D. C. Bloedow, R. C. Thompson, W. Hopfer, V. Kung, C. Brownfield, and R. Pacifici, Interleukin-1 receptor antagonist decreases bone loss and bone resorption in ovariectomized rats. J. Clin. Invest. 93, 1959 – 1967 (1994). 108. G. Pioli, G. Basini, M. Pedrazzoni, G. Musetti, V. Ulietti, D. Bresciani, P. Villa, A. Bacci, D. Hughes, G. Russell, and M. Passeri, Spontaneous release of Interleukin-1 and interleukin-6 by peripheral blood monocytes after ovariectomy. Clin. Sci. 83, 503 – 507 (1992). 109. T. Matsuda, K. Matsui, Y. Shimakoshi, Y. Aida, and S. Hukuda, 1-Hydroxyethilidene-1, 1-bisphosphonate decreases the postovariectomy-enhanced interleukin 1 production by peritoneal macrophages in adult rats. Calc. Tissue Int. 49, 403 – 406 (1991). 110. D. L. Lacey, E. L. Grosso, S. A. Moser, J. Erdmann, H. L. Tan, R. Pacifici, and D. T. Villareal, IL-1 induced murine osteoblast IL-6
ROBERTO PACIFICI
111.
112.
113.
114.
115.
116. 117.
118.
119.
120.
121.
122.
123.
124.
125.
126.
127.
128.
production is mediated by the type I IL-1receptor and is increased by 1,25 dihydroxyvitamin D3. J. Clin. Invest. 91, 1731 – 1742 (1993). M. E. Cohen-Solal, A. M. Graulet, M. A. Denne, J. Gueris, D. Baylink, and M. C. de Vernejoul, Peripheral monocyte culture supernatants of menopausal women can induce bone resorption: Involvement of cytokines. J. Clin. Endocrinol. Metab. 77, 1648 – 1653 (1993). C. L. Dinarello, T. Ikejima, S. J. C. Warner, S. F. Orencole, G. Lonneman, J. G. Cannon, and P. Libby, Interleukin-1 induces interleukin-1. I. Induction of circulating interleukin-1 in rabbits in vivo and in human mononuclear cells in vitro. J.Immunol. 139, 1902 – 1908 (1991). S. Khosla, J. M. Peterson, K. Egan, J. D. Jones, and B. L. Riggs. Circulating cytokine levels in osteoporotic and normal women, J. Clin. Endocrinol. Metab. 79, 707 – 711 (1994). R. A. Hannon, A. Blumshon, H. Galatti, H. Fenner, R. G. G. Russell, and R. Eastell, Cytokine receptors, receptor antagonists and postmenopausal bone loss. J. Bone Miner. Res. 8, 143A (1993). B. Galve-de Rochemonteix, L. P. Nicod, A. F. Junod, and J. M. Dayer, Characterization of a specific 20- to 25-kD interleukin-1 inhibitor from cultured human lung macrophages. Am. J. Respir. Cell Mol. Biol. 3, 355 – 361 (1992). W. P. Arend, Interleukin 1 receptor antagonist: A new member of the interleukin 1 family. J. Clin. Invest. 88, 1445 – 1451 (1991). J. Pfeilschifter, B. M. Seyedin, and G. R. Mundy, Transforming growth factor beta inhibits bone resorption in fetal rat long bone cultures. J. Clin. Invest. 82, 680 – 685 (1988). R. O. C. Oreffo, G. R. Mundy, S. M. Seyedin, and L.F. Bonewald, Activation of the bone derived TGF beta complex by isolated osteoclasts. Biochem. Biophys. Res. Commun. 158, 817 – 823 (1989). P. D. Delmas, Biochemical markers of bone turnover in osteoporosis. In “Osteoporosis” (B. L. Riggs and L. J. Melton, eds.) pp. Raven Press, New York, 297 – 316 (1988). C. J. Rosen, K. Usiskin, M. Owens, C. O. Barlascini, M. Belsky, and R. A. Adler, T lymphocyte surface antigen markers in osteoporosis. J Bone. Miner. Res. 5, 851 – 855 (1990). Y. Imai, T. Tsunenari, M. Fukase, and T. Fujita, Quantitative bone histomorphometry and circulating T lymphocyte subsets in postmenopausal osteoporosis. J. Bone Miner. Res. 5, 393 – 399 (1990). J. S. Duke-Cohan, H. Weinberg, R. Sharon, and D. Nahor, Immunological function in osteoporosis. Clin. Immunol. Immunopathol. 35, 125 – 129 (1985). R. B. Kimble, A. B. Matayoshi, J. L. Vannice, V. T. Kung, C. Williams, and R. Pacifici, Simultaneous block of interleukin1 and tumor necrosis factor is required to completely prevent bone loss in the early post-ovariectomy period. Endocrinology 136, 3054 – 3061 (1995). D. Uebelhart, A. Schlemmer, J. S. Johansen, E. Gineyts, C. Christiansen, and P. D. Delmas, Effect of menopause and hormone replacement therapy on the urinary excretion of pyridinium crosslinks. J. Clin. Endocrinol. Metab. 72, 367–373 (1992). T. J. Wronski, L. M. Dann, H. Qi, and C. F. Yen, Skeletal effects of withdrawal of estrogen and diphosphonate treatment in ovariectomized rats. Calcif. Tissue Int. 53, 210 – 216 (1993). R. B. Kimble, J. L. Vannice, C. Brownfield, and R. Pacifici, Persistent bone-sparing effect of interleukin-1receptor antagonist: A hypothesis on the role of IL-1 in ovariectomy induced bone loss. Calcif. Tissue Int. 55, 260 – 265 (1994). R. Baron, L. Neff, P. Tran Van, J. R. Nefussi, and A. Vignery, Kinetic and cytochemical identification of osteoclast precursors and their differentiation into multinucleated osteoclasts. Am. J. Pathol. 122(2), 36 3 – 378 (1986). A. M. Parfitt, Bone remodeling: Relationship to the amount and structure of bone, and the pathogenesis and prevention of fractures.
CHAPTER 41 Postmenopausal Osteoporosis In “Osteoporosis” (B. L. Riggs ed.), Raven Press, New York, pp. 45 – 93 (1988). 129. R. Kimble, S. Bain, and R. Pacifici, The functional block of TNF but not of IL-6 prevents bone loss in ovariectomized mice. J. Bone Miner. Res. 12, 935 – 941 (1997). 130. J. N. Bradbeer, S. J. Stroup, J. C. Hoffman, and M. Gowen, An orally active inhibitor of cytokine synthesis prevents bone loss in the ovariectomized rat. J. Bone Miner. Res. 11 (Suppl. 1), Abs123 (1996). 131. H. Kitamura, H. Kawata, F. Takahashi, Y. Higuchi, T. Furuichi, and H. Ohkawa, Bone marrow neutrophilia and suppressed bone turnover in human Interleukin-6 transgenic mice. Am. J. Pathol. 147, 1682 – 1692 (1995).
101 132. S. Cenci, M. N. Weitzmann, and R. Pacifici, M-CSF neutralization and Egr-1 deficiency prevent ovariectomy induced bone loss. J. Clin. Invest. 105, 1279 – 1287 (2000). 133. T. E. Wilson, M. L. Day, T. Pexton, K. A. Padgett, M. Johnston, and J. Milbrandt, In vivo mutational analysis of NGFI-A zinc fingers. J. Biol. Chem. 267, 3718 – 3724 (1992). 134. D. N. Kalu, Proliferation of tartrate-resistant acid phosphatase positive multinucleate cells in ovariectomized animals. Proc. Soc. Exp. Biol. Med. 195, 70 – 74 (1990). 135. G. D. Roodman, Interleukin-6: An osteotropic factor? J. Bone Miner. Res. 7, 475 – 478 (1992). 136. R. T. Turner, B. L. Riggs, and T. C. Spelsberg, Skeletal effects of estrogen. Endocr. Rev. 15, 275 – 300 (1994).
CHAPTER 42
Osteoporosis in Men Epidemiology, Pathophysiology, and Clinical Characterization ERIC S. ORWOLL AND ROBERT F. KLEIN Bone and Mineral Research Unit, Oregon Health Sciences University, and the Portland Veterans Affairs Medical Center, Portland, Oregon 97201
I. II. III. IV. V. VI.
VII. VIII. IX. X.
Introduction Fractures in Men The Major Determinants of Skeletal Health in Men Osteoporosis The Evaluation of Osteoporosis in Men Therapy
I. INTRODUCTION
II. FRACTURES IN MEN A. The Incidence of Fractures
Although osteoporosis has long been considered a disease of women, the earliest reports of the epidemiology of fractures associated with osteoporosis clearly showed that the classical age-related increase in fractures seen in women was also evident in men. In the past few years it has been recognized that the problem of osteoporosis in men represents an important public health issue [1] as well as a huge personal burden for those men who are affected [2]. It also presents a unique array of scientific challenges and opportunities [3 – 5]. Here we examine the issue of osteoporosis in men and compare its pathophysiology and clinical presentation to parallel processes in women.
OSTEOPOROSIS, SECOND EDITION VOLUME 2
Hypogonadism Alcoholism Tobacco Renal Stone Disease References
The incidence of all fractures is higher in men than women from adolescence through middle life [6 – 8] (Fig.1), and the personal and economic impact of these early life fractures is enormous. The average number of hospitalizations for fractures in men between the ages of 18 and 44 years was 181,000 in the United States (1985 – 1988), and the annual number of lost work days for men due to fractures was 17,543,000 [9]. Despite the importance of early life fractures in men, little has been done to understand their causation. Many result from serious trauma, but to some extent relative bone fragility may
103
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.
104
ORWOLL AND KLEIN
FIGURE 1 Average annual fracture incidence rate in males and females per 10,000 population, by age group [512].
contribute to fracture risk during this period. For instance, recent long-term follow-up of men who had sustained traumatic tibial or forearm fractures in early mid-life revealed that they were at much greater risk for later hip fracture [10]. At about age 40 – 50 years there is a reversal of this trend, with fractures in general, and in particular those of the pelvis, humerus, forearm, and femur, becoming much more common in women. However, the incidence of fractures due to minimal-to-moderate trauma (particularly hip and spine) also increases rapidly with aging in men (Fig. 2) and reflects an increasing prevalence of skeletal fragility. 1. PROXIMAL FEMUR The proximal femur is the most important site of osteoporotic fracture, about which the most complete epidemio-
FIGURE 2
logical data are available. The incidence of hip fracture rises exponentially in men with aging, as it does in women. However, the age at which the increase begins is slightly older (5 – 10 years) in men [11]. In U.S. men older than 65 years, the incidence of hip fracture is 4 – 5/1000 [12,13], compared to 8 – 10/1000 in similarly aged U.S. women. A 2 – 3:1 female:male ratio has also been reported in northern Europe and Australia, although in other geographic areas the ratio has been noted to be much lower [8,14]. In southern Europe and other areas, the incidence of hip fracture is relatively lower in both sexes, and men have as many hip fractures as do women [15 – 17], and in Asian populations the male:female incidence may actually be quite low or reversed [18,19]. Since there are fewer older men than women, the absolute number of hip fractures tends to be proportionately less in men (of those experiencing their first hip fracture 65 years or older, 165,000 in men vs 580,000 in women in the United States, in 1984 – 1987, or 22% of the total in men) [13]. In the United States (Rochester) the lifetime risk from age 50 onward of a hip fracture has been calculated to be 6% in men and 17.5% in women [20] and 2.4% in men and 9% in women in Canada (Saskatchewan and Manitoba) [21]. It is estimated that approximately 30% of hip fractures worldwide will occur in men [22]. Unfortunately, the number of hip fractures is projected to increase dramatically as the elderly population expands [23 – 25]. In some populations increasing fracture incidence in men has been noted (United States and northern Europe), whereas in many areas the rate of hip fracture in women and men appears to have stabilized, or even declined [15, 24,26 – 29]. Perhaps as a result of a higher prevalence of concomitant disease [30], mortality associated with a hip fracture in elderly men (75 years) is considerably higher than in
Age-specific incidence rates of hip, vertebral, and Colle’s fracture in Rochester, Minnesota, men and women [513].
105
CHAPTER 42 Osteoporosis in Men
FIGURE 3
Observed and expected race- and sex-specific survival following fracture of the hip, all ages combined [514].
women [31 – 34] (Fig. 3). In Europe the incidence of fracture is at least twice as great in women, but the death rates for femoral neck fractures are approximately equal, again suggesting a greater mortality risk in men [35]. Ethnic differences in the incidence of hip fracture in men are substantial. For instance, African-American men experience hip fractures at a rate only half that of Caucasians [13]. Interestingly, whereas African-American women are at significantly lower risk for hip fracture than Caucasian women, African-American men are less protected [11,36]. There are not extensive comparative data concerning other ethnicities, but Asian men have considerably lower incidence rates than in Caucasian populations [19,37]. Geography influences hip fracture rates for unclear reasons. The incidence of fracture is higher in urban than in rural men [38], and there are great variations in the incidence and sex ratio of hip fractures in southern Europe [39]. These differences are presumably the results of a mix of genetic and environmental factors. 2. VERTEBRAE Vertebral fracture is also an important sequel of osteoporosis. As in women, the presence of vertebral fracture in men is associated with loss of height, kyphosis, increased risk of other fractures, and increased disability [40, 41]. Since diagnostic criteria for a vertebral fracture are unsettled, and vertebral fracture infrequently results in hospitalization, consistent epidemiological information is somewhat limited. Previously considered uncommon in men, recent information suggests that the incidence of osteoporotic vertebral fracture in U.S. men is about half that in women (similar to hip fractures) [37,42 – 44]. Until about age 65, the prevalence of vertebral fracture is actually higher in men than women [45,46], and to some extent this
represents an increase in the occurrence of early life trauma in men [47]. In fact, vertebral and femoral bone mineral density values are lower in men with vertebral fractures than in nonfractured controls [42,48], indicating that vertebral fracture in men is not merely the result of a higher rate of trauma, but is also related to a low bone mass. Fractures are primarily in low thoracic vertebrae in men, but are found at all levels. Most fractures are anterior compression in type [42], with vertebral crush fractures occurring less frequently than in women. Vertebral epiphysitis (Scheuermann’s disease) is an uncommon cause of significant vertebral deformity in men [42]. As in women, the presence of vertebral deformity has adverse consequences on mortality [32], and the occurrence of a vertebral fracture indicates a much higher likelihood of sustaining other osteoporotic fractures [29]. 3. OTHER FRACTURES Other fractures (radius/ulna, humerus, pelvis, femoral shaft) share a common epidemiological pattern. Men experience more of these in youth, but with unusual exceptions (e.g., humerus) the incidence remains relatively stable during mid-life, while rising markedly in women [26,49 – 51]. It is only later (75 years) that the incidence of limb fractures begins to rise in men, and it then does so rapidly [49]. This increase is due primarily to an increase in lower limb fractures, while upper extremity fractures do not change as much. In older men, as in women, the likelihood of underlying bone pathology or propensity to fall (e.g., alcoholism) increased markedly [49]. Importantly, the occurrence of a distal forearm fracture [52,53] or a tibial fracture [10] in a man indicates a considerably increased risk of subsequent hip fracture, presumably as a result of low bone mass and/or an increased risk of falling.
106
ORWOLL AND KLEIN
B. The Determinants of Fracture 1. BONE MASS
FIGURE 4
Os calcis bone mineral content of men by 5-year age groups and grade of osteoporosis (n 821). Each point and bar represents the mean and SE, respectively. Grade 0, normal (no radiographic evidence of vertebral demineralization or fracture). Grades 1 – 3, progressively severe vertebral demineralization and fracture [58].
FIGURE 5
In women, bone mineral density (BMD) is clearly related to fracture risk in both retrospective case-controlled studies and prospective trials. There are few data available in men, but the available evidence is consistent with a similar inverse relationship of bone mass to fracture. For instance, in men with spinal fractures, measures of femoral cortical area and Singh index [54], proximal femoral dual photon absorptiometry (DPA) [42], vertebral quantitative computed tomography (QCT) [55], vertebral DPA [56,57], calcaneal SPA [58], and spine and hip dual energy X-ray absorptiometry (DXA) [48,59] have all revealed lower mean values than in control men (Fig. 4). In addition, the incidence of fracture is higher in men as bone density falls [60] (Fig. 5). In addition to lower BMD, men with vertebral fracture seem to have smaller vertebrae [61], probably reflecting the importance of size on biomechanical strength. Chevalley et al. and Karlsson et al. observed that hip and spine BMD are clearly reduced in a series of men with hip fracture compared to age-matched controls, and more recently several studies have documented the relationship between low BMD and increased risk of fracture at the hip and other appendicular sites [62 – 67]. In fact, the degree to which low BMD increases fracture risk appears to be similar between men and women. Although ultrasound measures are becoming more commonly used,
Incidence of new vertebral fractures (1981 – 1994) among men and women in the HOS, by quartiles of baseline calcaneus BMD [60].
107
CHAPTER 42 Osteoporosis in Men
only sparse data relate ultrasound measures to fracture risk in men. As in women, there is a clear overlap of bone density in men with fractures and nonfracture control subjects, indicating that bone density is not the sole determinant of vertebral fracture risk. Fracture is a somewhat chance event, and factors other than BMD (bone size, strength, propensity to fall, etc.) are also important variables [68]. When a man suffers a fracture, it implies a much higher risk of subsequent fracture [52,69,70], a relationship that may be even stronger in men than in women [71]. Even men who experience a traumatic fracture in mid-life are at higher risk of a later osteoporotic fracture [10,53]. 2. FALLS In addition to bone mass, the risk of falling has been identified as a major determinant of fracture in women. Although men have a somewhat lower risk of falls than women [72 – 74], a variety of factors indirectly related to risk of falling are associated with fracture. This is important, as the incidence of falls in the elderly appear to be increasing [75]. For instance, Nguyen et al. found that men who had experienced a nontraumatic fracture exhibit more body sway and lower grip strength (as well as lower bone density) than nonfracture controls [68]. In studies of men with hip fractures [76,77] a number of factors associated with falls were found to be more prevalent than in controls. These included receiving a disability pension, neurological disease, confusion, “ambulatory problems,” and alcohol use. The severity of falls, and the site of impact affect the likelihood of fracture [78 – 80]. Hemenway et al. reported that taller, heavier men have more hip fractures, possibly related to the force of impact involved in a fall [81]. As in women, the use of several classes of psychotropic drugs is associated with hip fracture risk in men [82,83]. Men with hip fracture weigh less, have lower fat and lean body mass, and more commonly live alone or are not married than control subjects [62,77,84]. These differences suggest a body habitus and lifestyle more conducive to falls and injury, as well as the possibility of other interacting risk factors (nutritional deficiencies, comorbidities). Finally, the characteristics of falls may be different in men and women, which in turn may influence the kinds if fractures that result [85]. 3. WHY ARE FRACTURES LESS COMMON IN MEN WOMEN?
THAN IN
The cause of the greater fracture rate in women is complex. First, accumulation of skeletal mass during growth, particularly in puberty, is greater in men than in women, resulting in larger bones. In tubular bones, there is a greater total width (20% in the second metacarpal) [86] and greater cortical width in early adulthood [87]. This difference persists throughout life. Since resistance to fracture in tubular bones is related both to total diameter and cortical thickness
it follows that long bone fractures should be less common in men [88]. Gender differences in the dimensions of axial bones may also contribute substantially to differences in mechanical competence [88]. For instance, compressive strength is strongly related to vertebral end plate area [89 – 91], and when bone density and body size are taken into account, fractures are more common in individuals with smaller vertebrae [92] (Fig. 6). From puberty on, mean vertebral cross-sectional area is 25% greater in men [92]. Moreover, in men vertebrae increase in cross-sectional area by 25 – 30% with aging as a result of periosteal apposition [93]. This process also occurs in women [94], but it may be more accentuated in men [95], thus amplifying the biomechanical advantage. Interestingly, the girth of the femur and other long bones increases with age in men more than in women [95 – 99]. These differences at the proximal femur and vertebrae may help explain the lower hip and vertebral fracture rates in men, particularly since the relative gender difference in peak hip and spine mineral density and in the rate of age-related decline in density are small. Other gender differences in skeletal anatomy may also provide a male advantage. Second, women lose more bone mineral with aging than do men, a phenomenon most apparent in long bones. Cortical porosity increases more in women [100], and women lose more at the endosteal surface and gain less periosteally than do men and thus accrue less biomechanical benefit [101 – 103]. At the proximal femur, as well, there is evidence that men lose bone less quickly than do women [104 – 106]. Moreover, as discussed above, a gender difference in the character of age-related changes in trabecular bone structure probably contributes to a greater fracture risk in women. Whereas in men the age-related fall in mineral density at trabecular sites (which is almost as impressive as in women) is the result of generalized trabecular thinning with some loss of trabeculae [96,107], in women there is a more marked loss of trabecular elements. Finally, elderly men fall less frequently than do women, reducing the risk of trauma as a cause of fracture [72 – 74,108].
III. THE MAJOR DETERMINANTS OF SKELETAL HEALTH IN MEN A. Peak Bone Mass Development In early childhood, there are few discernible differences between the skeletons of boys and girls [109]. During adolescence both sexes exhibit dramatic increases in bone mass which are closely related to pubertal stage and is almost complete when puberty ends [110]. In boys the achievement of peak bone mass is later, not only because puberty is later in onset but also because boys accrete bone for a
108
ORWOLL AND KLEIN
FIGURE 6
Axial cross-sectional area, transverse diameter, anteroposterior diameter, and vertebral body volume of the first (L1), second (L2), and third (L3) lumbar vertebral bodies in 12 pairs of men and women matched for age, vertebral body height, and bone density. Values are mean 1 SD, *P 0.001. Values in men are shown by the open bars; values in women are shown by shaded bars [515].
longer time during this crucial period [109] (Fig. 7). The adolescent development of adult bone mass depends upon changes in both density and size, with increases in size being quantitatively much more important [87,88,109]. During pubertal skeletal maturation obvious sexual differences in adult skeletal morphology emerge, and sexual differences appear to be related only to differences in size. Virtually all skeletal dimensions in men are larger than those in women. For example, radial width and cortical thickness are considerably larger in men [101], femoral neck cross-sectional area is larger [97], and vertebral body cross-sectional area is larger [111]. As a result, total body bone mineral is greater in men (3100 – 3500 g in young men vs 2300 – 2700 g in young women) [112,113]. The development of peak bone mass in boys is influenced by a variety of factors, among the most important being exercise, nutrition (calcium and vitamin D, protein), the adequacy and timing of puberty, and the presence of adverse medical events and lifestyles (e.g., smoking) [114 – 119]. The acquisition of peak bone mass in men, as in women, is also strongly influenced by genetic factors. Krall and Dawson-Hughes estimated heritability to be 40 – 83% at
several measurement sites in men [120]. Men with a family history of osteoporosis or fractures have lower bone mass and/or greater fracture risk than those without [121 – 124]. The specific genes responsible have yet to be identified, although several candidates have been evaluated, including the estrogen receptors, androgen receptor, vitamin D receptor, collagen type I1, insulin-like growth factor 1, aromatase, etc.
B. Age-Related Bone Loss 1. BONE MASS a. Appendicular Bone Cross-sectional studies suggest that age is associated with a fairly linear decrease in cortical bone mass [56,101,113,125 – 127], but some also indicate the BMD:age slope becomes more negative in men after 50 years [56,101,126,128]. This slope is not quite as steep as that in women [101,129], thereby accentuating the gender differences in cortical mass present in early adulthood. However, the rate of cortical bone loss in men as reported in longitudinal studies is considerably more rapid (5 – 10%/decade)
109
CHAPTER 42 Osteoporosis in Men
FIGURE 7 Bone mass gain at the lumbar spine during adolescence. The yearly increase in lumbar spine BMD and BMC is depicted in males and females (mean SE) [516]. [56,127,128,130] than previously estimated from crosssectional studies (1 – 3%/decade) [125,126,131]. b. Axial Bone The decline in axial bone density was initially considered to be relatively slow in men, primarily
FIGURE 8
because of cross-sectional studies using techniques that assess total spinal bone mass (DPA). Vertebral bone density as measured by QCT, however, suggested a much faster rate of bone loss with aging in normal men [132]. Subsequently, results derived from DPA were shown to be influenced by artifacts in measurement introduced by extravertebral calcifications. If men with such calcifications are excluded, the relationship of spinal bone density to age is similar in men and women [42]. Longitudinal studies verify a more rapid rate of vertebral bone loss with aging in normal men [56]. Moreover, bone volume in the iliac crest declines at very similar rates in both men and women. In cross-sectional studies the slope of density with age at proximal femoral sites is significantly negative in men, albeit somewhat less than that in women [125,131,133,134]. In longitudinal studies, femoral neck bone loss clearly occurs in men, at a rate approximately the same as that in women [104]. In both sexes, the rate of femoral bone loss accelerates with increasing age [104,105]. In the large Rotterdam Study Burger et al. noted that older women lose more bone between ages 55 and 70 (presumably reflecting the effects of menopause) but thereafter the rate of bone loss accelerates in both genders [105] (Fig. 8). Although the end result is a greater cumulative bone loss in older women, it is important to recognize the dramatic effect of aging on bone loss in men as well. Finally, cross-sectional studies using ultrasound measures of bone (calcaneus) also indicate a change with aging. Broadband ultrasound attenuation and speed of sound both decline with advancing age in men [135], at a rate clearly less than that in women. The interpretation of ultrasound measurements is yet somewhat uncertain, but they may reflect not only bone mass but also structural or material properties of bone. 2. BONE ARCHITECTURE a. Appendicular Bone In cortical bone, men experience an increase in porosity with aging, although at a rate somewhat slower than that seen in women [100,136]. This
Mean yearly rate of change in bone mineral density (BMD) and 95% confidence interval according to age group and sex, the Rotterdam Study, the Netherlands, 1990 – 1995. P values are for linear trends [105].
110
ORWOLL AND KLEIN
results in a reduction in density and mechanical strength [99] and probably increases fracture risk. Although men of greater weight and lean body mass have larger appendicular cortical areas, this does not protect them from age-related loss [101]. However, the decline in cortical mass is to some extent compensated by changes in cortical dimensions [137]. In both sexes, there is an age-related increase in cortical width, and since fracture resistance is so dependent on geometry, this change is beneficial. In a two-decade study Garn et al. found that the rate of metacarpal cortical loss in men was very similar in both men and women, but periosteal apposition was somewhat greater in men (and endocortical loss somewhat less), mitigating the loss of thickness and overall mass [101]. This gender difference can be observed in other long bones as well [99] (Fig. 9). The increased periosteal apposition rate and the somewhat lesser rate of cortical loss with age in men are in accord with the fracture patterns observed in the elderly, in whom the rate of appendicular fractures is less in men than in women. b. Axial Bone There are microscopic changes in axial bone architecture with aging that in all likelihood influence fracture risk independent of changes in bone mass. A decline in vertebral trabecular number and thickness with age is associated with a reduction in compressive strength [138], and men with vertebral and femoral fractures have a lower trabecular plate density [139]. In men and women there is a generalized loss of trabeculae, but loss of horizontal elements (number and thickness) is particularly marked, in turn resulting in less support to vertical, load bearing trabeculae [140]. Similar changes in trabecular structure in other locations (e.g., proximal femur) probably also contribute to fracture risk. In fact, the quantitation of proximal femoral trabecular patterns reveals a definite loss of trabeculation with age in men, and men with osteo-
FIGURE 10
FIGURE 9
Age-related changes in the calculated failure moment of male and female human femoral shafts in bending. The slope of the female data is significantly different from 0 [99].
porotic fractures have less trabeculation than control men [141]. In addition to trabecular loss, the appearance of microfractures increases with age and may also contribute to fracture risk [142]. Despite the basic similarities of these processes in men and women, the nature of trabecular bone loss may have gender differences. Using histomorphometry to analyze vertebral bone, Mosekilde found that while bone density is not particularly different between older men and women, the microarchitectural pattern of trabecular loss is distinct. Women tend to experience both trabecular thinning and trabecular loss (particularly horizontal elements) while men experience trabecular thinning with less trabecular dropout [140]. Similar results have been described in iliac crest biopsy specimens [107,139,143,144] (Fig. 10). Gender
Changes in trabecular number (solid lines) and trabecular width (dashed lines) with age in the iliac crest of men and women [143].
111
CHAPTER 42 Osteoporosis in Men
[160,161]. Ljunghall et al. [157] and Kurland et al. [162] noted lower IGF-I concentrations in men with idiopathic osteoporosis. Similarly, the skeletal content of important growth factors, including IGF-I and TGF- , decline with age in men and may herald major changes in growth factor action [156]. Some of the gender differences in bone biology may be due to the growth factor axis. The relationship between IGF-I and BMD may be different in men and women [161,163] and there are apparently gender differences in the skeletal response to growth hormone therapy [164,165]. 2. CALCIUM NUTRITION AND BONE LOSS FIGURE 11
The ratio of urinary pyridinoline to urinary creatinine excretion (mean SE) in 440 adults expressed as a percentage of values in young adults [517].
differences in the nature of structural changes with aging may have important biomechanical consequences, as load bearing of vertebral specimens appears to be better preserved in men than in women [140].
C. Causes of Age-Related Bone Loss Aging in normal men is associated with detectable appendicular and substantial axial bone loss. The cause of this loss in unknown but is speculated to be related to a number of factors. Histomorphometric techniques demonstrate a reduction in bone formation (mean wall thickness) in both sexes [143,145 – 149] that probably contributes to the decline in bone mass. An additional age-related increase in bone resorption in men is not apparent using these methods [145,148]. However, markers of bone remodeling increase with age in men [150 – 152] (Fig. 11), raising the possibility of an acceleration in bone turnover that contributes to bone loss. In addition to these putative influences, several other processes contribute to the pathophysiology of senile bone loss, including nutritional deficiencies, inactivity, and loss of gonadal function. 1. GROWTH FACTOR ACTION Growth factor concentrations decline with age [153, 154], and there are several reasons to link age-related bone loss in men to changes in growth factor or cytokine physiology [155 – 157]. Recently, Johannson et al. found a surprising correlation between insulin-like growth factor binding protein 3 (IGFBP-3) and bone mineral density in men [158], and similar relationships between insulin-like growth factor-I (IGF-I) and BMD have been noted by others [159]. Since there is a relationship between IGF-I and sex steroid concentrations, an effect of growth factor action and bone may be in part related to sex steroid action as well
Riggs and Melton [166] have suggested that senile (Type II) osteoporosis in men and women is due, at least in part, to alterations in calcium economy. The average level of dietary calcium necessary to maintain mineral balance is relatively low in young men (400 – 600 mg/day) but the range is large and there are data that suggest a higher requirement in older men [167,168]. Although U.S. men achieve a mean dietary calcium intake considerably greater than that of women (800 vs 500 mg/day in the 1978 NHANES survey) these data still indicate that about one half of men ingest less than the recommended daily allowance (800 mg), and many ingest much less. In addition, aging in men has been associated with increased parathyroid hormone (PTH) concentrations [169,170], reduced circulating 25-hydroxyvitamin D [171], and (in some studies) subnormal 1,25-dihydroxyvitamin D concentrations [172 – 175]. In the Baltimore Longitudinal Study of Aging [176,177], lower radial bone density in men was related to higher PTH and lower 25-hydroxyvitamin D concentrations. Halloran and Bikle [178] summarized the data relating age related changes in calcium homeostasis and bone health in men. Several reports have linked dietary calcium intake to levels of bone density in men, but the evidence is not yet conclusive. In a study of 222 subjects, Kroger and Laitinen found that men in the highest tertile of calcium intake (1200 mg/day) have higher proximal femoral BMD (but not spinal BMD) than those in the lowest tertile (800 mg/day) [177]. Similarly, in a cross-sectional study of a small group of men, Kelly and Pocock found that measures of axial BMD correlate with dietary calcium intake, but appendicular radial BMD do not [179]. Other groups who have examined appendicular bone mass in adult men in longitudinal studies have also found no clear relationship to calcium intake [101,104,130,180]. These results suggest that calcium intake may play a role in the determination of axial, but not (or to a lesser extent) appendicular bone mass. However, in the only published controlled trial of calcium supplementation in adult men, no beneficial effects were found on the rate of bone mineral loss from either spinal or radial sites [56], despite the fact that urine calcium
112 excretion increased and PTH concentrations were suppressed. Osteocalcin concentrations were not altered. The results of this trial are somewhat muted by the relatively large dietary calcium intake of the subjects prior to supplementation (1100 mg/day), and supplementation in a less calcium-replete population may prove to be more effective. In a longitudinal study of the Rotterdam population, Burger et al. found lower calcium intake to be associated with higher rates of bone loss [105]. There have been no studies of the relationship between calcium intake and skeletal structure (e.g., cortical thickness, trabecular architecture, remodeling rates, material properties). A variety of studies have examined the relationship between dietary calcium intake and hip fracture in men, with inconsistent results. In a small case-controlled study in Hong Kong, Lau et al. found that a very low calcium intake (75 mg/day) was associated with fracture risk [181]. In a British case-control trial, Cooper et al. found those men with the highest calcium intakes (1041 mg/day) to be significantly protected [182]. In several longitudinal observational trials (including the NHANES I follow-up study [183], the Rancho Bernardo study [184], and a study of eight communities in Britain [185], hip fracture risk in men was strongly suggested to be related to dietary calcium intake, but the relationships did not reach statistical significance. However, two other very large studies [68,186] found no relation between calcium intake and hip fracture risk in men. Looker et al. [183] have pointed out the pitfalls inherent in trials of this sort, including low power, difficulties in estimating calcium intake and the effects of confounding variables. In general, evaluations of the relationship between calcium nutrition and hip fracture in men are suggestive of a beneficial effect, but remain inconclusive. Moreover, there have been no attempts to examine the effects of calcium intake on other fractures, in particular vertebral fractures. Albeit incomplete, the data are probably consistent with a limited role for dietary calcium insufficiency in the determination of the rate of bone loss and fractures in men. 3. WEIGHT AND PHYSICAL ACTIVITY Mechanical force exerts major effects on bone mass in men [187], and it is probably one of the fundamental variables responsible for the gender dimorphism in bone mass and structure. In cross-sectional studies, bone mass is greater in physically active men [183,188 – 193], an effect that can be demonstrated at both the regional (i.e., the particular anatomic region affected) and the systemic levels. Muscle strength and lean body mass in men also correlate with bone density both regionally and systemically [183,188,194]. Furthermore, muscle strength is related to bone bending stiffness in men, an index of strength independent of mass, suggesting that mechanical force has effects not only on bone mass but also quality [195]. Longi-
ORWOLL AND KLEIN
tudinal studies tend to corroborate the effect of mechanical force on skeletal mass in men [104,196] but are very few in number. Finally, exercise has been strongly related to a reduction in hip fracture rates in men [185,186] an effect that may also relate to a reduced risk of falls. Unfortunately, the fairly consistent finding of positive correlations between exercise history and/or strength and bone mass in crosssectional studies have not been confirmed in longitudinal investigations. As in women, body weight is itself highly correlated with bone density in men [132], an effect that could be related to the mechanical effect exerted by mass alone or to a particular aspect of body composition (i.e., lean vs fat mass, adipose distribution). Reid et al. [197] suggested that there are gender differences in the relative effects of body composition on skeletal morphology. In their studies bone mass was associated with fat mass in women, but not in men (lean mass was not associated with bone mass in either sex). They speculated that androgens may contribute to the lack of fat – bone correlation in men, as androgen action is associated with an increase in bone mass but a fall in adiposity. Low body weight is also associated with increased rates of bone loss in men [105,198,199]. In sum, the available data strongly suggest a powerful effect of weight and mechanical force on the male skeleton. In view of the clear decline in physical activity and muscle strength with aging [200,201], bone loss in men may, in part, relate to a diminution of the trophic effects of mechanical force on skeletal tissues. Certainly, the character of agerelated bone loss closely mimics that of chronic disuse [202], but this tentative conclusion requires confirmation in longitudinal studies. 4. CHANGES IN GONADAL FUNCTION Aging in men is associated with changes in the hypothalamic – pituitary – gonadal axis that result in notable declines in total and free testosterone levels [203,204]. These changes have given rise to considerable speculation as to whether several of the concomitants of aging, are the result at least in part of the decline in testosterone levels [205,206]. For instance, the well documented declines in muscle strength and bone mass with aging have been suggested to be potential sequelae [207]. Indeed, there are several lines of evidence firmly linking androgen action to skeletal mass in men (vide infra), and there have been several attempts to link bone mass to testosterone levels. Kelly and Pocock [179] found that free testosterone levels correlated with ultra distal bone density (but not with a variety of other densitometric measurement sites) in a group of men ages 21 – 79 even after the effects of age were considered. Similarly, in several other studies androgen concentrations were found to correlate with (albeit weakly) bone mineral density [208 – 212]. However, these findings have not been corroborated by other investigators [158,
113
CHAPTER 42 Osteoporosis in Men
TABLE 1
Effect of T Administration on Biochemical Parameters Related to Bone Baseline
T
Placebo
Serum calcium (nmol/liter)
2.28 0.02
2.23 0.02
2.24 0.03
Serum phosphate (nmol/liter)
0.98 0.04
0.91 0.05
0.96 0.04
Creatinine clearance (ml/min)
104 5
110 8
103 4
54 4
51 3
52 3
Variable
Serum AP (U/liter) Osteocalcin (ng/ml) Calcium/GFR (mmol/dl)
4.3 0.6
4.3 0.7
4.1 0.7
4.87 0.52
4.43 0.59
4.59 0.49
HPro/GFR (mol/dl)
151 10
108 8*
142 13
TmP (mg/dl)
2.47 0.18
2.26 0.09
2.40 0.14
1,25-Dihydroxyvitamin D (pg/ml)
23.8 1.8
27.8 1.7
31.5 2.4
PTH (pg/ml)
216 17
188 18
200 19
Note. Values are the mean SE. AP, alkaline phosphatase; HPro, hydroxyproline. Reproduced with permission for J. S. Tenover [217]. *P 0.01.
213 – 216]. In an attempt to test the hypothesis that relative androgen deficiency has a skeletal impact in older men, Tenover reported in a small study (13 men) that parenteral testosterone supplementation reduced urinary hydroxyproline excretion [217] (see Table 1) and Snyder et al. found that testosterone administration increased spinal BMD slightly in older men [218]. Clarke et al. suggested that changes in testosterone are not as important as the skeletal effects of age-related declines in adrenal androgens [219]. The study of this and similar issues is made particularly difficult by the inability to adequately assess the long-term, integrated level of androgen action on bone with crosssectional or relatively short-term study designs. In general the correlations between serum androgen concentrations and bone mass in adult men have been weak or insignificant, but the role of androgen action on the skeleton remains unclear. The effects of estrogen on the male skeleton have become of great interest [220,221]. Several men with abnormalities of estrogen action, including one man with an abnormal estrogen receptor [222] and two others with aromatase deficiency [223,224], have presented as young adults with failure of skeletal maturation and low bone mass. Therapy with estrogen prompted a major increase in bone mineral density in those with aromatase deficiency [224,225]. Clearly, estrogen is important in male skeletal development, and these cases further raise the issue of whether estrogen is the more important sex steroid [226]. Additionally, estrogen has sometimes reported to be positively correlated with bone mineral density [227,228] or
fractures [229] in older men, even to a greater extent than is testosterone. These findings raise important issues related to the role and regulation of aromatase activity and the relative importance of estrogens vs androgens in skeletal homeostasis. Nevertheless, it is important to note that androgens clearly influence the skeleton, and that they probably play an independent, and coordinated, role with estrogens [230]. An elucidation of these interactions will be crucial for the understanding of bone biology in both genders. In addition, the appropriate use of estrogen measurements in the evaluation of men with osteoporosis and the use of estrogens or selective estrogen receptor modulators in the management of men with low bone mass require clarification. Obviously the issue of the importance of gonadal insufficiency in the genesis of senile bone loss in men remains unresolved. In addition to the decline in androgen levels associated with aging, overt gonadal insufficiency (reviewed below) would be expected to contribute to any aging effects in an additive fashion.
IV. OSTEOPOROSIS Osteoporosis in men is a heterogeneous condition, encompassing a wide variety of etiologies and clinical presentations. In practice, it is common to uncover several potential explanations for bone loss and fractures in a single patient.
A. Age-Related Osteoporosis Bone loss that occurs with aging is an important feature of osteoporosis in men and women (see above). In some men, age-related bone loss may alone suffice to cause nontraumatic fractures. Even when other causes of bone loss are present (i.e., hypogonadism, alcoholism), the universal loss of bone that accompanies aging unquestionably contributes to the eventual propensity for fractures.
B. Idiopathic (Primary) Osteoporosis Osteoporosis in men has been termed idiopathic if no known cause can be identified on clinical and laboratory grounds. Although metabolic bone disease in men has been traditionally considered to be more commonly related to “secondary” causes [166,231 – 234], this impression is difficult to substantiate. In fact, the frequency with which osteoporosis in men has been found to be idiopathic is significant. In large series of osteoporotic men, many patients were considered to have bone disease of unknown etiology (70 of 105 subjects [231], 40 of 94 subjects [235], and 60 of 95 subjects [232]).
114 The age of men with primary, or idiopathic, osteoporosis varies widely (23 – 86 years) with an average in the mid-60s. This age range overlaps that of “senile” osteoporosis, and differentiation of idiopathic and senile osteoporosis is somewhat arbitrary. Riggs and Melton [166] defined senile (or Type II) osteoporosis as occurring in either sex after age 70, but this definition obviously does not exclude the potential for pathophysiological overlap between older and younger patients. The universal decline in bone mass that happens as a concomitant of aging has the potential for eventually producing clinical osteoporosis in all individuals, and some idiopathic osteoporosis may represent this age-related process or its premature onset. Once again, it is important to emphasize that broad classifications of osteoporosis are of limited value in considering an individual patient, in whom several pathogenetic mechanisms (sometimes occult) may be operative. The character of idiopathic osteoporosis in men is relatively indistinct. After major secondary contributors to bone loss have been eliminated, more detailed biochemical and histomorphometric analyses of men with idiopathic disease fail to reveal consistent features [54,232,234,236]. Some patients have slightly increased serum alkaline phosphatase activity [54,232]. Reduced intestinal calcium absorption has been reported in the presence of lowered 1,25-dihydroxyvitamin D concentrations [54], but calcium balance has not been systematically examined in osteoporotics males. Osteoblastic dysfunction may contribute to osteoporosis in men [234,236]. Bordier et al. reported studies of a series of 11 patients with idiopathic osteoporosis, 10 of whom were men [237]. In those subjects histomorphometric parameters clearly suggested that a defect in bone formation contributed to loss of bone [235] Osteoblastic dysfunction, however, is not a consistent finding in idiopathic male osteoporosis [54,235]. Nordin et al. suggested that accelerated resorption may also be a primary mediator [238]. These apparently discrepant results may in part be explicable by findings reported by Aaron et al. [239], who noted that young men ( 49 years) with idiopathic osteoporosis had reductions in bone-forming parameters (osteoid surface, mean wall thickness), while older men had forming parameters similar to age-matched controls but evidence of slightly increased resorption. Khosla et al. also found the young patients with idiopathic osteoporosis (both men and women) had histomorphometric characteristics, suggesting a defect in bone formation [240]. Moreover, Marie et al. found that most men with eugonadal osteoporosis had evidence of decreased bone formation. When osteoblasts from these patients were compared to normal controls or to osteoporotics without reduced bone formation, a lower proliferation capacity was found [241] (Fig. 12). Thus the histomorphometric
ORWOLL AND KLEIN
FIGURE 12 The time course of osteoblastic cell proliferation evaluated by DNA synthesis in bone samples from osteoporotic men with low (open circles) or normal (open triangles) bone formation as determined by the extent of double-labeled surface, compared to normal bone cells (solid squares) (mean SE) [241]. mechanisms of early-onset primary osteoporosis may differ from those operating when the disorder appears later in life. The late-onset form appears to resemble the bone loss that occurs with normal aging, albeit at a more accelerated pace. A possible explanation for the osteoblastic defect postulated to be important in idiopathic osteoporosis is a reduction in growth factor action [242]. Several reports suggest that men with idiopathic osteoporosis have relatively low IGF-I, or IGFBP-3, levels [157,162,243], a finding that seems to relate to lower indices of osteoblast work. Other etiologies for idiopathic osteoporosis have been suggested, including abnormalities in cortisol dynamics [199]. It is likely that at least a fraction of the men who present with idiopathic osteoporosis have genetic underpinnings of their disorder. In fact, IGF-I concentrations are related to the presence of polymorphisms in the IGF-I gene [244]. Other genes have been implicated [245 – 249], but more definitive studies are needed. Certainly this is an area that needs additional development. Finally, the microarchitectural features of idiopathic osteoporosis in men have not been defined. Francis et al. [54] and Aaron et al. [239] did report that men with primary osteoporosis have a reduction in iliac crest bone volume and surface primarily because of a reduction in trabecular number rather than a decline in thickness. The issue of trabecular connectivity was not formally evaluated.
115
CHAPTER 42 Osteoporosis in Men
C. Osteoporosis Secondary to Other Disorders The pathophysiologic character of osteoporosis in men is minimally explored, but several reports have examined the risk factors present for bone disease in small patient populations. In the available series [54,231,232,250], 30 – 60% of men evaluated for vertebral fractures had “secondary” causes (underlying illness) contributing to the presence of bone disease. Most of these studies were of selected subjects – most commonly men presenting for health care in the United States or Great Britain because of vertebral fracture. Other series are small with heterogeneous patient groups [251]. Hence the findings may not accurately represent a generalizable spectrum of disease. There have not been adequate studies of the character of the bone disease present in men sustaining femoral or other fractures. The principal conditions found in men with reduced bone mass are shown in Table 2. Prominent are glucocorticoid excess, hypogonadism, alcoholism, gastrectomy and other gastrointestinal disorders, and hypercalciuria. These disorders will be the focus of this discussion. They and others have been recently reviewed [252]. Similar attempts to examine the contributing factors in osteoporotic women suggest that the spectrum of disorders differs somewhat [26,54,166,253], but glucocorticoid excess, premature hypogonadism, and gastrointestinal disease are prominent in women as well. It has been suggested that the number of men with “secondary” osteoporosis is higher than in women [166], but in other objective evaluations [253] the
TABLE 2
Osteoporosis in Men
I. Primary Senile Idiopathic II. Secondary Hypogonadism Glucocorticoid excess Alcoholism Gastrointestinal disorders Hypercalciuria Smoking Anticonvulsants Thyrotoxicosis Immobilization Osteogenesis imperfecta Homocystinuria Systemic mastocytosis Neoplastic diseases Rheumatoid arthritis
proportion of women with major illnesses contributing to the development of bone disease is actually very similar to that observed in male osteoporotics. 1. GLUCOCORTICOID EXCESS In the largest series of men evaluated for spinal osteoporosis, glucocorticoid excess (particularly exogenous) is the most prominent of the secondary causes identified, accounting for 16 – 18% of the men evaluated [54,231]. Commonly, glucocorticoid use is but one of several risk factors present in patients with chronic medical problems. For instance, there is widespread clinical recognition of the frequency of osteoporosis in older men with chronic obstructive pulmonary disease treated with glucocorticoids, often in the presence of tobacco and alcohol abuse. The pathophysiology of glucocorticoid-induced osteoporosis, although incompletely understood, is presumably similar in men and women (see also Chapter 44). The primary mechanism by which glucocorticoids cause bone loss appears to be via a direct receptor-medicated inhibitory effect on osteoblast activity. Physiologic concentrations of glucocorticoids enhance the function of differentiated osteoblasts [254,255], but extended exposure to supraphysiologic concentrations results in inhibition of osteoblastic synthetic function [256,257]. Recent evidence suggests that inhibition of osteoblastic collagen synthesis is mediated through glucocorticoid-induced inhibition of local synthesis of factors that promote collagen synthesis (e.g., insulin-like growth factor I, prostaglandin E2) [258,259]. Another skeletal effect that may be important in the pathogenesis of glucocorticoid-induced osteoporosis may be impairment of osteoblast recruitment [260]. Additionally, certain systemic effects of glucocorticoids can exert profound influence on skeletal health. Muscle weakness due to muscle fiber atrophy is a frequent concomitant to glucocorticoid administration [261 – 263] and a clear-cut association between steroid myopathy and osteoporosis has been observed [264]. The relative immobility that is imposed by such muscle weakness is likely to contribute to the bone loss in this setting. Most patients receiving pharmacologic doses of glucocorticoids exhibit impaired intestinal absorption of calcium and elevated urinary calcium excretion, resulting in a negative calcium balance [265]. Finally, glucocorticoids can alter gonadal hormone production. Exogenous glucocorticoids markedly reduce testosterone levels in men [266 – 268] by mechanisms that have not been fully defined, but which may include central inhibition of GnRH release, suppression of pituitary sensitivity to GnRH, and direct antagonism of testicular steroidogenesis [267 – 270]. Impotence and loss of libido frequently occur in the clinical settings in which glucocorticoids are administered and are attributed to the effects of the chronic illness. However, these symptoms
116
ORWOLL AND KLEIN
may actually be due to glucocorticoid-induced hypogonadism, which in turn may contribute substantially to the resulting low bone mass. Clinicians who care for men with osteoporosis should be aware of this phenomenon and recognize it as an important cause of a low serum testosterone value. Furthermore, because administration of testosterone to hypogonadal men improves bone mass, such therapy may be useful to prevent and treat glucocorticoid-induced osteoporosis in men. 2. HYPOGONADISM Sex steroids have major influences on the regulation of bone metabolism. The obvious importance of menopause to osteoporosis drew early attention to the role of estrogen, and both clinical and basic observations have also highlighted the importance of androgens in bone physiology in both sexes. There is an expanding understanding of the cellular effects of androgens on bone remodeling (see Chapter 11). Androgen receptors are present in physiologically relevant concentrations in osteoblastic cells [271 – 273], and androgens affect a variety of osteoblastic functions, including proliferation, growth factor and cytokine production, and bone matrix protein production (collagen, osteocalcin, osteopontin) [274 – 281]. Hence, there is an excellent foundation in basic research for the precept that androgens are active in bone. On the other hand, estrogen has assumed an even more important role as its importance in the control of male skeletal function has emerged. a. Pubertal Hypogonadism Since adolescence is so important for skeletal maturation, disorders of puberty have the potential to impair peak bone acquisition and thus to influence fracture risk throughout adulthood [282]. With adolescence, bone accretion (in both cancellous and cortical bone compartments) in both sexes is closely related to gonadal maturation [110,283] (Fig. 13). Testosterone has major effects on calcium kinetics and balance in boys [284,285] (Fig. 14). It is not known whether adrenarche affects the rate of skeletal maturation [286]. Strongly supporting the importance of androgen action in the achievement of peak bone mass in men is the fact that genetic males with complete androgen insensitivity (testicular feminization) experience increased pubertal growth but achieve a bone mass less than expected of androgen replete men [287 – 289]. Reduced bone mass is found in men who experienced an abnormal puberty (Klinefelter’s and Kallman’s syndromes) [290 – 292]. In Klinefelter’s syndrome radial bone mass is related to serum testosterone concentrations, and patients have lower circulating osteocalcin and higher rates of hydroxyproline excretion [293]. It is proposed that the failure to acquire peak bone mass with puberty is the primary abnormality in these forms of early-onset hypogonadism [290]. In fact, constitutionally delayed puberty is associated with permanent reductions in bone density
FIGURE 13 The relationships between bone mineral density of the lumbar spine, femoral neck, and femoral shaft and pubertal stages in male and female subject (*P 0.05) [110]. [294 – 296] (Fig. 15), although some have suggested that the effect may be more on bone size than density [297]. An apparent emphasis on cortical bone mass in males suggests that androgen action is especially important in the modeling
117
CHAPTER 42 Osteoporosis in Men
FIGURE 14 Changes in Va (dietary calcium absorption), Vbal (net calcium retention), and Vo (rate of bone accretion) in prepubertal children treated with testosterone (n 6). Solid bars, before testosterone; hatched bars, after testosterone. *P 0.05 [284].
drifts that add periosteal new bone during growth, and the animal model of testicular feminization supports this postulate in that cortical mass is disproportionately reduced [298]. A reasonable hypothesis is that androgens increase trabecular bone formation at epiphyseal areas, and strongly promote the addition of cortical thickness through periosteal and endosteal growth — processes that then are impaired in the presence of hypogonadism.
FIGURE 15
Radial bone mineral density in 23 men with a history of delayed puberty and 21 normal men [296].
Until recently there were no clear examples of estrogen deficiency in men, and the direct role of estrogen in skeletal metabolism in men remains controversial. As described above, there are now several examples of men with inadequate estrogen action (but adequate androgen effects) during development. From these individuals, it is clear that estrogen is needed for adequate epiphyseal function as well as bone mass development. The histological or biochemical nature of the skeletal effects of prepubertal hypogonadism in humans is unknown. Some animal studies have indirectly investigated the effects of prepubertal hypogonadism. Hock and Gera [299] examined trabecular and cortical bone mass in rats castrated (or sham operated) just before sexual maturation (4 weeks). In both compartments, bone mass in the hypogonadal rats continued to increase but at a rate considerably slower than in the controls, a finding consistent with the hypothesis that hypogonadism impairs the pubertal stimulus to bone mass accumulation. Another study of adolescent rats based on histomorphometric analysis of cortical bone suggests that periosteal bone formation and cortical thickness is decreased in adolescent males following castration [300], indicating that a reduction in bone formation contributes to low bone mass. Certainly the fact that male rats have markedly greater cortical mass than do females, and that the differences are eliminated by castration [300], strongly support a prominent role for androgens in this phase of cortical bone accumulation. b. Postpubertal Hypogonadism Androgens also appear essential for the maintenance of bone mass in adult men, as the development of hypogonadism in mature men is associated with low bone mass. Hypogonadism is present
118 in 5 – 33% of men evaluated for vertebral fractures and osteoporosis [54,231,233], and hip fractures in elderly men apparently occur more commonly in the setting of hypogonadism [301]. Reduced bone mass and fractures are associated with many forms of hypogonadism, including castration, hyperprolactinemia, anorexia, and hemachromatosis [302 – 305]. A very important group of men with hypogonadism are those treated with castration and/or GnRH agonists for prostate disease. Here bone loss is rapid and the development of osteoporosis at an accelerated rate can be expected [306 – 309]. Vertebral and appendicular bone mass are both reduced in hypogonadal men, but in adult-onset hypogonadism vertebral loss is relatively more pronounced. The degree of reduction in bone density has been correlated with levels of serum testosterone in some series [293,310,311], but in other groups no association between the two variables was detected [312]. There may be a threshold level of serum testosterone below which skeletal health is impaired, but at present it is not possible to establish that hypothesis. In addition to the link between primary testicular dysfunction and low bone mass, reduction in gonadal function secondary to several other conditions is now postulated to contribute to the development of bone loss. For instance, hypogonadism is suspected to contribute to the reduced bone mass associated with glucocorticoid excess, renal insufficiency, and other conditions [313]. The histological pattern of hypogonadal bone loss in adult men is inadequately described. A single report examines skeletal metabolism in the period immediately after gonadal failure. Stepan and Lachman [302] studied a small group of men in the years immediately following castration. The subjects lost bone rapidly (approximately 7%/year) and had clear biochemical indications of increased bone remodeling (increased serum osteocalcin levels and urinary hydroxyproline excretion). Similar evidence of an increase in remodeling was reported by Goldray et al. in studies of GnRH agonist administration [309]. Unfortunately, no direct histomorphometric analyses were reported. An early increase in remodeling after androgen withdrawal is also consistent with recent reports of the biochemical and cellular events that are associated with androgen action (a suppression of cytokine production and osteoclast formation) [314]. The histomorphometric data (primarily from patients with long standing hypogonadism), and the better documented sequence of events that follows gonadal hypofunction in menopause, suggest that this period of increased remodeling is followed by a subsequent phase of reduced turnover, possibly accompanied by a decline in bone formation. Several reports have described the histomorphometric character of hypogonadal, osteopenic men, but they are uncontrolled, and are from subjects in whom hypogonadism was of varied causation (both early and late onset) and of
ORWOLL AND KLEIN
long duration. For instance, in a study of 13 men with longstanding hypogonadism Francis and Peacock found that bone remodeling and formation were reduced, and 1,25-dihydroxyvitamin D concentrations were low in those with fractures [315]. With testosterone therapy, 1,25-dihydroxyvitamin D values increased and there was some indication of an increase in “formation” parameters. Similarly, Delmas et al. reported decreased rates of formation in a small group of hypogonadal men [316], and formation rates were low in a single case reported by Baran et al. [317] (unfortunately vitamin D metabolite levels were not reported in these studies). These data raise the issue of whether androgens provide an important stimulus to bone formation. In contrast, Jackson et al. [318] reported histomorphometric analyses of a small group of osteoporotic, chronically hypogonadal men with normal vitamin D status. In these patients no apparent defect in mineralization was observed, and the authors speculated that nutritional vitamin D defi ciency may have been a factor in the European studies of Francis and Delmas. Jackson et al. found a slight increase in mean remodeling rate and concluded that androgen defi ciency induces a remodeling defect similar to that of estrogen deficiency in the postmenopausal period. In both the series from Jackson and Francis [315,318], trabecular number was reduced in the hypogonadal men, indicating that trabecular loss is a feature of androgen deficiency as it is in estrogen deficiency [139]. There are no reports of cortical remodeling dynamics. Actually, the remodeling character of all these study populations was heterogeneous, and in view of the of the variability in the small groups, the presence of other confounding clinical conditions, and the lack of adequate controls, no firm conclusions can be drawn concerning the remodeling defect induced by hypogonadism in men. Animal studies provide some additional information (see also Chapter 11), but the current model of skeletal metabolism in hypogonadism remains quite incomplete. 3. ALCOHOLISM It is well established that long-term alcohol consumption can result in a host of abnormal clinical, biochemical, and physiologic findings that stem from the toxic effects of ethanol on the liver, gonads, marrow, heart, and brain. The fact that prolonged abuse of alcohol is also detrimental to skeletal integrity in men has only recently been recognized [319 – 321] (see Chapter 31). Numerous studies over the past quarter century have demonstrated a reduction in bone mass in alcoholics, especially in the iliac crest, calcaneus, vertebral column, and hip [231,234,322 – 335 ] — all areas with a high proportion of metabolically active, trabecular bone. It should be pointed out, however, these studies can be criticized for their small size and for being poorly controlled. In addition, recent reports suggest that modest alcohol intake may actually be associated with increases in bone density [336].
119
CHAPTER 42 Osteoporosis in Men
More extensive studies have attempted to estimate the prevalence of skeletal fracture in the alcoholic population. Reduced bone mass is evident on routine radiographs in a significant percentage of individuals (25 – 50%) whose drinking habits have prompted them to seek medical help [328 – 331]. The degree to which bone disease is present in the entire population of alcoholics remains uncertain, and determination of the true incidence of alcohol-induced reductions in bone mass in men must await a survey of a large number of cases. However, the habitual consumption of alcoholic beverages is clearly recognized as a significant negative determinant of bone mass in epidemiological surveys of men [130,231,330,334] and has been shown in longitudinal studies to be associated with increased rates of bone loss [130]. Thus, the link between alcohol abuse and bone disease is well established. Although there is a definite relationship between alcohol abuse and bone disease, the mechanisms by which alcohol induces bone disease remain unclear. An association between alcoholism and accidental injury is well recognized. However, emerging evidence now suggests that in addition to the increased incidence of trauma, the high incidence of fracture in this population may also stem from generalized skeletal fragility. The in vivo rate of skeletal protein synthesis of young male rats is reduced by 25 – 30% in response to acute ethanol exposure [337]. Furthermore, the chronic administration of ethanol to rats produces ultrastructural changes in bone morphology that compromise mechanical strength [338,339]. Bone histomorphometric studies have provided further insight into the specific nature of the skeletal disorder induced by ethanol. Iliac crest biopsy usually reveals significant reductions in trabecular bone
volume, osteoid matrix, number of osteoblasts, mineral apposition and overall rate of bone formation [331 – 333, 340 – 342] (Table 3). Marrow fibrosis is uncommon and there is generally no evidence of osteomalacia [234,332, 333,342], except in patients who have previously undergone gastric surgery [343]. Parameters reflecting osteoclast activity (e.g., eroded surface, resorption depth, and resorption period) are, for the most part, spared [331 – 333, 340 – 342] (Table 3). Since the osteoblast is the cell responsible for bone formation, the histomorphometric findings in alcoholic patients with bone disease suggest that the osteoblast may be specifically targeted by alcohol. Nutritional deficiencies are common in alcoholics. However, poor nutrition alone does not induce osteoporosis in experimental animals [344] and none of the histomorphometric studies cited above demonstrated any evidence for nutritional deficiency. Mild hypocalcemia, hypophosphatemia, and hypomagnesemia are frequently present in ambulatory alcoholic men because of poor dietary intake, malabsorption, and increased renal excretion [341, 345 – 349]. Hypocalcemia, if severe enough, could result in low bone mass by inducing a state of secondary hyperparathyroidism. However, evidence for hyperparathyroidism with accelerated bone remodeling is not seen on bone biopsies of affected patients [331 – 333,340]. Recently, acute alcohol intoxication has been reported to reduce PTH concentrations [350]. The relevance of this finding is unclear as the reduction in PTH lasts only a few hours, and low bone mass is not observed in hypoparathyroid patients. Early studies found circulating levels of vitamin D metabolites to be low [234,325,351], but subsequent investigation has excluded vitamin D deficiency as a major
TABLE 3 Abstainers (n 9)
Drinkers (n 16)
Significance (P value)
Parameters of Bone Formation Osteoid volume (%)
0.68 0.17
0.38 0.08
NS
Osteoid seam width (m)
11.4 0.68
7.95 0.48
0.001
Osteoid surface with osteoblasts (%)
20.1 4.3
5.5 1.7
0.01
Osteoblasts/10 cm surface
143.5 18.3
51.5 18
0.003
Parameters of bone resorption 6.9 1.03
5.5 0.7
NS
Extent of surface with osteoclasts (5)
0.76 0.24
1.11 0.22
NS
Osteoclasts/10 cm surface
17.1 3.4
21.3 5.1
NS
Mineralization rate (M/day)
0.52 0.1
0.26 0.07
0.04
Mineralization lag time (days)
28 4
62 10
0.006
Osteon remodeling time (days)
140 20
423 78
0.003
Extent of surface with lacunae (%)
Parameters of Mineralization
Note. Adapted from Crilly et al. [33].
120 cause of alcohol-induced bone disease by demonstrating normal vitamin D absorption [352] and conversion to 25hydroxyvitamin D [353] in alcoholic individuals and, more directly, by the measurement of normal free concentrations of 25-hydroxyvitamin D and 1,25-dihydroxyvitamin D in patients with alcoholic cirrhosis and alcoholic bone disease [333,354,355]. These findings do not exclude the possibility of an alcohol-induced vitamin D resistant state, but again the lack of histomorphometric evidence of osteomalacia in vitamin D-replete osteopenic alcoholic subjects [332] argues strongly against such a possibility. Hypogonadism is clearly a risk factor for osteoporosis in men (see above), and chronic alcoholic men suffer from impotence, sterility, and testicular atrophy [356]. Although most studies of alcoholic men with bone disease report normal androgen values [333,335,348], reduced serum free testosterone concentrations in alcoholic subjects with osteoporosis were reported by Diamond [332]. The testosterone levels were on average lower than those of the male control subjects but still fell within the normal range for the general male population overall. Recent studies suggest that ethanol also exerts toxic effects directly at the cellular level in bone. Ethanol induces a dose-dependent reduction in cellular protein and DNA synthesis in human osteoblasts in vitro [357,358]. Further evidence implicating a direct effect of ethanol on osteoblast activity comes from studies examining circulating bone Gla protein (BGP, osteocalcin) levels in alcoholic subjects. BGP is a small peptide synthesized by active osteoblasts, a portion of which is released into the circulation. BGP values are positively correlated with histomorphometric parameters of bone formation in normal individuals [359] and in patients with metabolic bone disease [360]. Chronic alcoholic patients exhibit significantly lower BGP concentrations than age-matched controls [361,362]. Moreover, alcohol has a dose-dependent suppressive effect on circulating BGP levels [350,363,364]. The consumption of 50 g of ethanol (equivalent to four “shots” of scotch whisky) over 45 min results in a 30% decrease in serum BGP concentration that is detectable 2 h later. Beyond these fragmentary attempts at characterization, however, little is known about phenotypic regulation by ethanol in the osteoblast. 4. TOBACCO USE Tobacco use is associated with lowered bone mass and fractures in women [365 – 367]. Tobacco was also linked to an increased prevalence of vertebral fractures in men in the cohort studies of Seeman [231], in which the relative risk of vertebral fracture in smokers was 2.3. This risk was independent of alcohol consumption, and in fact the risk imparted by the two variables was multiplicative. Hip fracture rates are higher in currently smoking men, particularly those smoking more heavily [77,186]. In support of the adverse effects of smoking on bone health in men, Slemenda
ORWOLL AND KLEIN
FIGURE 16 Bone loss in members of male twin pairs who differed in cigarette smoking behavior. Points above the line indicated pairs in which the twin who smoked more lost more bone [130].
et al. [130] found that the rate of bone loss from the radius was significantly greater (140%) in subjects who smoked than in nonsmokers (Fig. 16). In fact, there was a correlation between the number of cigarettes smoked and the rapidity of bone loss and the smokers lost more bone than did their nonsmoking twins. In this study the adverse effects of smoking and alcohol use were independent. A variety of other studies have noted the relationship between smoking and low bone mass, bone loss, and fractures in men [51,77,198,368 – 371] (see Chapter 31). The mechanism by which smoking affects bone in men is unclear. In women, tobacco use has been associated with lower weight, calcium absorption, and estrogen levels, all of which are negatively associated with bone mass [365,366,372]. Alternatively, smoking may impair respiratory function and hence bone metabolism, or a direct toxic effect of smoking on bone metabolism may exist. There are no data available concerning the effects of tobacco use in other forms (chewing, snuff). Whether smoking adversely affects androgen levels or other potential effectors of bone remodeling in men is unknown. 5. RENAL STONE DISEASE Several reports have linked hypercalciuria or nephrolithiasis in men to a reduction in BMD [231,373 – 377] and this issue was recently reviewed by Zerwekh [378]. It is not clear whether this apparent increase in bone disease results from a greater impact of nephrolithiasis in males [379] or perhaps reflects the fact that hypercalciuria is more than twice as common in men than women [380]. In fact,
121
CHAPTER 42 Osteoporosis in Men
hypercalciuria has also been linked to osteoporosis in women [381]. The etiology of the low bone mass observed in hypercalciuric patients is unclear, but has been postulated to involve an alteration in mineral metabolism. Somewhat surprisingly, osteopenia has been reported in patients with absorptive hypercalciuria [379,382 – 384]. In this setting, increased serum concentration of 1,25-dihydroxyvitamin D may increase bone resorption, but also the correlation of urinary sodium and sulfate levels with BMD in these patients implicate a contribution of dietary protein and sodium as well [382]. In renal hypercalciuria, a negative calcium balance with secondary hyperparathyroidism (including increased 1,25-dihydroxyvitamin D levels) is potentially important [383,385,386]. This hypothesis is supported by the finding that lower dietary calcium intakes in men with renal hypercalciuria are associated with further reductions in bone mass [387]. However, in some patients hypercalciuria may be only part of a more diffuse metabolic abnormality that affects bone metabolism in other ways. For instance, renal tubular acidosis may be present with hypercalciuria and low bone mass in the presence of complex abnormalities of mineral and bone metabolism [388], hypercalciuria has been linked to phosphate-wasting disorders causing low bone mass [389], medullary sponge kidney is not uncommonly associated with increased urinary calcium excretion and disordered parathyroid function [390 – 392], and Dent’s disease preferentially affects males and presents with metabolic bone disease, hypercalciuria and several other renal tubular abnormalities [393]. In most patients with idiopathic hypercalciuria, the bone deficit is quite modest, and of itself unlikely to result in clinically significant bone disease. On the other hand, renal lithiasis has been associated with symptomatic osteoporosis in men [231]. In a small series of relatively young men (five subjects ages 27 – 57 years), Perry et al. [394] found osteoporosis in association with moderate hypercalciuria in the absence of other risk factors for bone loss. In all patients, calcium hyperabsorption appeared to contribute to the hypercalciuria. Histomorphometric analysis revealed an increased rate of bone remodeling in the face of no apparent alteration in mineral metabolism. In a somewhat different experience, Zerwekh et al. [384] reported on 16 men (mean age 50 11) referred for evaluations of osteoporosis nine were hypercalciuric without any other obvious cause of bone disease. Further examination showed that all of the hypercalciuric group had evidence of an element of absorptive hypercalciuria, and four actually had increased gastrointestinal calcium absorption and 1,25-dihydroxyvitamin D concentrations. In the hypercalciuric subgroup, bone formation rates were depressed (reduced bone formation rate, increased mineralization lag time) in comparison to normals or normocalciuric osteoporotics, with no differences in indices of bone resorption. Similar findings were
reported in men with idiopathic hypercalciuria [395], and in a mixed population of absorptive and renal hypercalciurics [379]. In other groups of men with unexplained osteoporosis, some have been reported to be hypercalciuric [234,251, 396]. Thus, there are suggestive, but still preliminary data linking hypercalciuria to bone loss and osteoporosis in men. The specific pathophysiology involved and the clinical spectrum of resultant bone disease remain somewhat unclear. Although the relationship between hypercalciuria and osteopenia is relatively strong, and the pathophysiology at least superficially intuitive, some intriguing data suggest there may be other factors which are also important. Jaeger et al. studied a large group of renal stone formers and found low BMD not only in the hypercalciurics but also in the normocalciuric patients [377]. In their subjects, BMD was correlated with other factors which influence stone forming potential, including urinary sulfates, sodium, uric acid, and pH, raising the issue of whether the cause of osteopenia in stone formers is related to aspects of renal function aside from, or in addition to, calcium handling. 6. MISCELLANEOUS DISORDERS A variety of other illnesses or medications have been associated with bone loss or fractures in men, including anticonvulsant use [54,231,397], thyrotoxicosis [398], immobilization [399], liver and renal disease [232], homocystinuria [54], and others. However, there is little evidence to suggest that the skeletal abnormalities induced by these conditions affect men any differently (qualitatively or quantitatively) than women.
V. THE EVALUATION OF OSTEOPOROSIS IN MEN Guidelines for the efficient, cost-effective approach for the evaluation of patients with low bone mass, or patients suspected of having low bone mass, are poorly validated for either sex. Current recommendations are therefore based on existing knowledge of the epidemiology and clinical characteristics of osteoporosis [400,401] rather than upon models that have been carefully tested in prospective studies. Within these constraints it is possible to formulate an approach to the male osteoporotic (Fig. 17). Of necessity it derives from the more mature knowledge base available concerning osteoporosis in women, but may depart in several key areas.
A. The Diagnosis of Osteoporosis In some men (as in women) the diagnosis of an osteopenic metabolic bone disease can be made with basic
122
ORWOLL AND KLEIN
FIGURE 17
Scheme for the diagnosis and evaluation of osteoporosis in men [518].
clinical information. Most important is a clear history of low trauma fractures in the absence of evidence of a focal pathological process (malignancy, infection, Paget’s disease, etc.). There are several clinical situations in which the presence of osteoporosis cannot be confidently determined, but should be considered likely. In these circumstances further diagnostic steps are appropriate. These situations include the presence of suspicious fractures, the radiographic presence of low bone mass, and conditions known to be associated with increased risk of bone loss. 1. FRACTURES The presence of a low trauma fracture should raise the probability of metabolic bone disease and prompt further
evaluation (i.e., densitometry). Certainly the occurrence of classic osteoporotic fractures (vertebral, proximal femoral) in the absence of focal pathology should raise immediate concern. The incidental finding of vertebral deformity in men warrants comment, as their prevalence is relatively high [42,402]. It is frequently assumed that many of these result from excessive trauma, or developmental deformity (Scheuermann’s disease), and hence should not be considered the consequence of low bone mass. Davies et al. [403] reported that the prevalence of vertebral deformity is high in men, but does not increase with aging as it does in women, suggesting that these deformities may not be related to changes in bone mass or structure. However, other
123
CHAPTER 42 Osteoporosis in Men
studies show that men with even relatively small degrees of vertebral deformity (vertebral height reduction 2 standard deviations) generally have mean BMD values significantly below those of subjects without deformity [42] (Fig. 18). These observations argue that the finding of vertebral abnormality (for instance on chest radiography) should raise the concern that an osteopenic disorder is present, and further evaluation should be considered.
2. RADIOGRAPHIC DETECTION OF REDUCED BONE MASS Low bone mass detected radiographically, even in the absence of a fracture, is of concern because the loss of bone must be well advanced before it is detectable with routine x-ray procedures. If, in fact, a generalized reduction in bone mass is noted radiographically, it indicates the presence of clinically significant depression of bone mass, and should prompt a diagnostic evaluation. Although useful to observe when present, radiographic estimates of reduced bone mass are qualitative, and the actual quantitation of bone density provides a more definitive estimate of disease severity and a baseline from which to gauge subsequent change. 3. CLINICAL CONDITIONS ASSOCIATED LOW BONE MASS
WITH
As discussed above, it is apparent that there are a number of causes of increased risk for osteoporosis in men, including glucocorticoid excess, alcoholism, and hypogonadism (discussed above). The presence of one, or particularly several, of these conditions should prompt concern, and the consideration for further characterization of skeletal status. 4. BONE MINERAL DENSITY MEASUREMENTS
FIGURE 18
Lumbar spine (a) and femoral neck (b) bone mineral density in male subjects with and without vertebral deformity. Four grades of vertebral deformity are illustrated (anterior vertebral/posterior vertebral height ratio of 0.85 or 0.80, and anterior vertebral heights 2 SD or 3 SD of a population mean) [42].
In men who present with findings that suggest the presence of metabolic bone disease (low trauma fractures, radiographic criteria indicating the presence of a reduction in bone mass, or conditions associated with bone loss), the measurement of BMD should be strongly considered. These measurements can be useful in several ways, including cementing the diagnosis of low bone mass and gauging the severity of the process. Generalized screening of older men with bone mass measures is worth evaluating as a strategy, but should not yet be routinely adopted. For reasons that have been previously elaborated [404,405], bone density measures provide valuable data that cannot be deduced from other clinical information, and can solidify the diagnosis of low bone mass. Although this contention is derived from studies in women, the basic tenets should be applicable in men as well. Specifically, (i) low bone mass is related to fracture, (ii) bone mass measures predict fracture risk, (iii) bone mass can be accurately measured, (iv) an understanding of bone mass may influence the therapeutic approach, and (v) treatment of osteoporosis affects fracture risk. The diagnostic criteria that should be used to identify men with high fracture risk, and thus in need of intervention, is uncertain. Although it is clear that there is an inverse relationship between bone density and fracture risk [64,66], the relationship between bone mineral density and fracture risk is not well established in men [406]. Some have suggested that the relationship between the absolute
124
ORWOLL AND KLEIN
level of bone density and future fracture risk should be the same in men and women [60,66,407] while others have noted gender differences [408]. Until the fracture risk associated with any given level of bone density is established in prospective trials, it may be reasonable to utilize reference ranges based on young normal male values.
B. Differential Diagnosis The intent at this stage of the evaluation should be to determine with reasonable certainty the histological cause of the osteopenic disorder, and to identify the etiologic factors contributing to it. In women, the vast majority of patients with osteopenic fractures have histological osteoporosis, but a small proportion are found to be osteomalacic [409 – 411]. Similarly, a fraction of men with fracture have osteomalacia [409 – 411]. Osteomalacia is estimated to be present in 4 to 47% of men with femoral fractures, with most reports being 20% [409 – 413]. Since some foods are fortified with vitamin D, occult osteomalacia may be less frequent in the United States than in other areas (e.g., northern Europe). Increasing age is associated with an increasing prevalence of osteomalacia [411]. Some have suggested that women with femoral fracture are more frequently osteomalacic than men [409, 410], but others report no distinction [411]. Thus far, the only patients who have been carefully surveyed are those with proximal femoral fractures, and it is not known whether populations with other fractures (vertebral) would include similar proportions of osteoporotic and osteomalacic individuals. Although the exact magnitude of the problem presented by osteomalacia in men is uncertain, it is clear that any differential diagnosis of low bone mass and fractures in men must consider the possibility. This becomes particularly imperative because the treatment for osteomalacia differs considerably from that of osteoporosis [414]. 1. INITIAL EVALUATION: HISTORY, PHYSICAL, AND ROUTINE BIOCHEMICAL MEASURES The history, physical, and routine biochemical profile can be very helpful in directing a focused evaluation of a man with low bone mass. Several approaches for the differential diagnosis of low bone mass have been suggested using standard clinical and biochemical information [253,414,415]. The goals of this stage of the evaluation should be to determine the specific diagnosis (what is the cause of the low bone mass – osteoporosis or osteomalacia?) and to identify contributing factors in the genesis of the disorder. Of particular importance in the history and physical examination, therefore, are signs of genetic, nutritional/environmental, social (alcohol, tobacco), medical, or pharmacological factors that may be present. Routine laboratory testing should include serum creatinine, calcium,
phosphorus, alkaline phosphatase, and liver function tests, as well as a complete blood count. If, on the basis of these tests, there is evidence for medical conditions associated with bone loss (alcoholism, hyperparathyroidism, malignancy, Cushing’s syndrome, thyrotoxicosis, malabsorption, etc.), a definitive diagnosis should be pursued with appropriate testing. 2. EVALUATION OF THE PATIENT WITH “IDIOPATHIC” OSTEOPOROSIS In men with reduced bone mass in whom no clear pathophysiology is identified by the routine methods above, it has been considered appropriate to be diagnostically aggressive, primarily because the potential for occult “secondary” causes of osteoporosis may be higher in men. However, the incidence of occult causes of osteoporosis in men, or whether it is greater than in women, is poorly studied. The diagnostic yield and cost effectiveness of extensive biochemical studies in the man with apparently “idiopathic” osteoporosis is unknown. Nevertheless, lacking this information, a reasonable evaluation of the man without an obvious etiology for osteoporosis might include: • 24-h urine calcium and creatinine to identify idiopathic hypercalciuria • 24-h urine cortisol excretion to identify Cushing syndrome • serum 25-hydroxyvitamin D concentration • serum testosterone and LH 3. HISTOMORPHOMETRIC CHARACTERIZATION Transiliac bone biopsy is a safe and effective means of assessing bone histology and remodeling [416]. Some have suggested that a transiliac bone biopsy is indicated in those men in whom a thorough biochemical evaluation has failed to reveal an etiology for osteoporosis [233]. The rationale for this approach is based on the need to accomplish several objectives: (i) ensure that occult osteomalacia is not present; (ii) identify unusual causes of osteoporosis that may be revealed only by histological analysis, such as mastocytosis [417,418]; and (iii) to yield information concerning the remodeling rate, which in turn may further direct the differential diagnosis (e.g., unappreciated thyrotoxicosis or secondary hyperparathyroidism suggested by the presence of increased turnover) or may be helpful in designing the most appropriate therapeutic approach. However, considerable histologic heterogeneity exists among men with osteoporosis. Whether distinct histologic patterns represent different stages of a single disease entity, separate subtypes of the disease, or simply an arbitrary subdivision of a normal distribution of remodeling rates is unknown. Vigorous attempts have been made to substitute sensitive and specific biochemical markers of bone turnover for histomorphometric estimates of bone turnover. Serum
125
CHAPTER 42 Osteoporosis in Men
levels of osteocalcin and procollagen peptides, and urinary collagen crosslink excretion correlate well with remodeling rates [152], at least when patients with overt metabolic bone disorders are included in the analyses. In normals and most osteoporotic patients the usefulness of these measures in predicting bone turnover rates is less clear. Serum osteocalcin concentrations may be elevated in men with osteomalacia, but appear to offer no added sensitivity for making the diagnosis of osteomalacia beyond that of conventional biochemical tests (alkaline phosphatase, 25-hydroxyvitamin D) [419]. Similarly, urinary pyridinoline excretion may be increased in osteomalacia [420]. With this information, a reasonable approach to the evaluation of remodeling dynamics in men with idiopathic osteoporosis (no etiology apparent from non-invasive testing) may be to combine the advantages of the biochemical markers of bone turnover with those of bone biopsy. An initial biochemical assessment of bone turnover should provide an understanding of remodeling rate. In the presence of an increase in biochemical indices of remodeling (osteocalcin, pyridinoline), a bone biopsy may be appropriate to identify unusual causes of high-turnover osteoporosis (e.g., mastocytosis). Bone biopsy may also be particularly helpful if there is any clinical concern for occult osteomalacia. Although alkaline phosphatase activity is usually increased in osteomalacia [414], even in this situation, a bone biopsy can reveal unanticipated osteomalacia, particularly in older men [411]. Unfortunately, the diagnostic yield or clinical impact of the bone biopsy is unknown. There is concern that it may be low, thus detracting from its clinical applicability. Essentially, the decision to utilize a bone biopsy is not well codified, and remains a matter of expert judgment (see Chapter 63).
VI. THERAPY Therapy of osteoporotic disorders in men is relatively unexplored. There have been very few trials of osteoporosis therapies performed specifically in male populations, although some men with osteoporosis have been included in mixed populations treated with a variety of agents [421]. In general it is very difficult to assess independently the success of these approaches in the male subjects.
A. Calcitonin There has been one trial of calcitonin therapy in a small group of men with idiopathic osteoporosis [422] in which total body calcium tended to increase during a 24-month treatment interval (100 IU administered subcutaneously each day with a calcium and vitamin D supplement). However, the change was not significantly different from that
observed in the control groups (receiving calcium plus vitamin D supplements, or vitamin D alone), and there were no changes in radial bone mass. In another uncontrolled, 12month trial of subcutaneously administered cyclical calcitonin (100 IU three times per week for three months, followed by three months without calcitonin) in men with vertebral osteoporosis small benefits were noted in spinal and proximal femoral bone density (compared to baseline) [423]. Men have been included in several other trials of calcitonin therapy, but the results in men are not separable from those in women subjects. There are no published studies of the effectiveness of intranasal calcitonin in men. Although there are few data, from a theoretical perspective calcitonin should be useful in reducing osteoclastic activity in at least some patients with osteoporosis or in those at risk of continuing bone loss. Pain following vertebral fracture has been reported to be alleviated with calcitonin, and some reports of this benefit have included men [424]. Whether men can be expected to respond differently than women is unknown.
B. Bisphosphonates There have been few trials of bisphosphonates performed exclusively in men, and many have been reported only in preliminary form [421]. Nevertheless, there is no conceptual barrier to the use of bisphosphonates in men, and recent reports describe positive results. Male patients with osteoporosis have been included in mixed patient populations, and have seemed to experience beneficial effects on calcium balance and lumbar spine bone density during treatment with pamidronate [425]. Men were specifically reported to benefit (increased vertebral bone density, with no change in femoral density) from etidronate treatment in a 12-month study [426]. An uncontrolled observational experience with intermittent cyclical etidronate (with calcium supplementation) in men with idiopathic osteoporosis and vertebral fractures [427] recently found small increases in lumbar spine and proximal femoral bone mass (3.2 and 0.7% per year, respectively) (Fig. 19), but no data were presented concerning fracture occurrence. There was no change in alkaline phosphatase activity. In another uncontrolled trial, Geusens et al. reported a somewhat more robust response to cyclical etidronate in osteoporotic men [428]. In the first large controlled trial of a bisphosphonate (or any therapy) in men with primary osteoporosis, alendronate had positive results on bone mass and reduced the rate of vertebral fracture [429]. These results provide considerable support for the effectiveness of bisphosphonates in men with osteoporosis. Of interest, the increase in bone mass resulting from alendronate was as great as was previously reported in postmenopausal women with osteoporosis, and was as great
126
ORWOLL AND KLEIN
FIGURE 19
Men with idiopathic osteoporosis appeared to have a positive response to therapy with intermittent cyclical etidronate, especially subjects of greater age. The scatterplot is of change in lumbar spine bone mineral density (BMD) against age at first scan. The average change in BMD per year of treatment was 0.024 g/cm2, or 3.2% [427].
in men who began the trial with low free testosterone levels as in those with normal levels. Bisphosphonates have also been examined in men with secondary causes of osteoporosis, especially in the context of glucocorticoid therapy. Men were included in some studies that indicated a positive effect of etidronate in glucocorticoid-treated patients [430]. For instance, in a large trial of alendronate in men receiving glucocorticoids, positive effects were noted in lumbar spine BMD [431]. Similar results have been reported in trials with other bisphosphonates (etidronate, risedronate) [432,433] in which the increase in BMD at several skeletal sites, and the tendency toward a reduction in fracture risk, was similar in men and women. There are a variety of other situations in which bisphosphonates may be useful, but little experience is yet available. For instance, inhibitors of bone resorption have been considered attractive in states of immobilization and in inflammatory conditions (e.g., rheumatoid arthritis). Men who receive anti-androgen therapy for prostate carcinoma are at risk for bone loss, and anti resorptive therapy should provide some protection for those patients. In fact, Diamond et al. reported that intermittent cyclic etidronate therapy (plus calcium) reversed bone loss initially experienced in men following long acting gonadotropin-releasing hormone agonist plus androgen antagonist therapy [307]
(Table 2). Some early reports are available in other conditions [434], and more can be expected as the effects of bisphosphonates in men are further explored.
C. Thiazide Diuretics Evidence supports a beneficial effect of thiazide administration on bone mass, rates of bone loss, and hip fracture risk in men [435 – 437]. For instance, in case controlled trials the use of thiazides reduced the rate of loss in calcaneal bone density by 49% compared to controls [438] and the relative risk of hip fracture was halved by exposure to thiazides for more than 6 years [439]. In a trial of similar design, thiazide use in men was associated with an adjusted odds ratio of femur fracture of 0.2 (95% CI 0.1 – 0.7) [440]. Other diuretics did not seem to impart the same benefits. Unfortunately, none of the available studies has been randomized or controlled, so a confident estimate of the magnitude of the protective effect is not possible. Moreover, the available literature does not allow a comparison of the relative benefits in men and women [441]. The mechanism for the positive effect is unclear, but it has been postulated to stem from the hypocalciuric effects of thiazides. Although probably not appropriately considered a primary treatment modality, a thiazide is probably the diuretic of
127
CHAPTER 42 Osteoporosis in Men
Concentrations of osteocalcin (g/liter), carboxyl-terminal cross-linked telopeptide of type I collagen (ICTP; g/liter), carboxyl-terminal propeptide of type I procollagen (PICP, g/liter 10 – 1) and bone-specific alkaline phosphatase (b-ALP; kat/L 10 – 1) in 21 men and 15 women with GHD before and after 9 months of treatment with recombinant human growth hormone. Results are shown as the mean SD. **P 0.01; ***P 0.001. The P values in letters refer to the difference in response to treatment between men and women [164].
FIGURE 20
choice in osteoporotic patients (other considerations not withstanding).
D. Parathyroid Hormone Parathyroid hormone administration to osteoporotic subjects has been shown to increase trabecular bone formation and bone volume in concert with an increase in calcium balance [442 – 444]. Slovik et al. reported that in a small group of men with idiopathic osteoporosis, combined PTH and 1,25-dihydroxyvitamin D administration increased trabecular (spinal) bone mass and improved intestinal calcium absorption [443]. Recently, PTH has been shown to exert striking increases in BMD and decreases in fracture incidence in postmenopausal osteoporotic women (see Chapter 77). Although its ultimate role in the treatment of osteoporosis, either alone or in concert with other agents [444], remains unclear, the potential of PTH appears similar in men and women.
E. Growth Hormone Growth hormone (or other growth factors) [155] may have anabolic actions on the skeleton in the elderly and in subjects with osteoporosis, but the available data are inconclusive [445]. Low levels of IGF-I have been reported to be
present in men with idiopathic osteoporosis, and in a study of healthy older men with low IGF-I levels, Rudman et al. [446] found that in addition to positive effects on lean mass, fat mass, and skin thickness, vertebral bone mass was increased slightly (1.6%) by the administration of growth hormone for 6 months. Radial and proximal femoral densities were unaffected. There are a number of reports that growth hormone administration may improve bone mass in growth hormone deficient adults [447,448], and the treatment of adults with growth hormone provokes an increase in biochemical markers of bone remodeling. Men have been reported to be more responsive to replacement therapy with growth hormone than women (Fig. 20) [164,165], raising interesting questions of the relative importance of growth factors in men vs women and the role of sex steroids in growth factor action [449]. The potential benefits of growth hormone on body composition (increased lean mass and perhaps muscle strength) [450] have been suggested to have additional benefits in patients with osteoporosis, but the functional importance of those changes has been questioned [451,452]. In adults with growth hormone deficiency, growth hormone replacement may have a more promising role. In that context bone mass has been noted to increase in most [453 – 455] but not all [456] studies. The response might be dependent on sex steroid action [449]. Despite interesting preliminary findings, the use of growth hormone is fraught with a variety of uncertainties, the benefits remain inconsistent and experimental results have been
128
ORWOLL AND KLEIN
difficult to interpret [445]. In either sex, growth hormone therapy is thus of potential, but as yet unproven usefulness (see Chapter 78).
ularly those at risk of vertebral fracture (see Chapters 74 and 75).
G. Specific Therapeutic Approaches F. Fluoride The use of fluoride in the therapy of osteoporosis remains controversial. Consistent and sometimes dramatic increases in vertebral bone mass can be achieved with supplemental fluoride, but the effectiveness of fluoride therapy in reducing fracture rates is uncertain. Nevertheless, there is an active interest in refining the formulation and dose of fluoride in the hope of taking better advantage of its anabolic properties. In fact, a recent evaluation of cyclically administered slow-release form of fluoride was reported to increase bone mass and reduce fracture rates in older women [457]. As with many of the other therapies discussed, there have been few specific trials of fluoride administration in men. In some studies, osteoporotic men have been included in the treatment groups but it is difficult to ascertain whether responses were in any way sex-specific. Ringe et al. recently described a 36-month, controlled trial of intermittent monofluorophosphate and calcium (vs calcium alone) in a large group on men with idiopathic osteoporosis [458]. Whereas the calcium-treated men lost bone density at the spine, radius, and proximal femur, those treated with fluoride experienced increases. The most marked change was seen at the lumber spine (8.9%), with more modest increases at the other sites. Importantly, there were fewer patients who experienced vertebral fractures in the fluoride treated group (10% vs 40%, P 0.008) (Fig. 21), and back pain was reduced (P 0.0003). There were fewer non-vertebral fractures in the fluoridetreated men as well, but the difference was not significant. These results are similar to parallel studies in women and suggest that fluoride may have some benefit in men, partic-
FIGURE 21 The percentage of osteoporotic men who experience new vertebral fractures during 3 years of treatment with monofluorophosphate plus calcium (MFP/Ca) or calcium alone (Ca) [458].
As discussed above, there is some information that suggests that specific treatment of underlying conditions associated with low bone mass can be effective in stabilizing or improving skeletal mass in osteoporotic men, but even those data are scarce, and the effects of these therapies on fracture risk are unknown for most agents. A suggested approach for the treatment of osteoporosis in men is outlined in Figure 22. 1. PREVENTION OF AGE-RELATED BONE LOSS Although it has recently become apparent that bone loss and fractures with aging in men is an important public health issue, there is very little information available concerning its prevention. Reasonable guidelines can be developed on the basis of current pathophysiologic models and on experience in women, but these approaches lack validation. a. Exercise Whereas an exercise prescription is diffi cult to generate with currently available information, activity is probably beneficial in several ways. Reductions in strength and coordination contribute to fracture via an increased risk of falling [459]. In addition, inactivity is associated with bone loss, and exercise may increase or maintain bone mass. Specific exercise prescriptions to accomplish these goals have not been confirmed in men or women, although it is clear that strength can be dramatically increased, and risk of falls reduced, in the elderly with achievable levels of exercise [459 – 461]. That fracture rates are lower in elderly men who exercise modestly buttresses this contention [186]. Beck and Marcus have recently reviewed the issue of exercise, men, and skeletal health [187] (see Chapter 28). b. Calcium An area of obvious interest is the influence of calcium and vitamin D nutrition [462]. Calcium intake is probably important in the achievement of optimal peak bone mass in boys [119], as well as the prevention and therapy of osteoporosis later in life. Calcium absorption declines with aging in men as in women, particularly after the age of 60, and well-documented changes in mineral metabolism occur concomitantly with age in men [178]. These data suggest both that optimal levels of calcium intake may change with age and that inadequate calcium nutrition can have an adverse effect on skeletal mass. However, the level of calcium intake that should be recommended is unclear, as few prospective studies have addressed this issue. No bone density benefit was observed from calcium/ vitamin D supplementation in a very well nourished population (mean dietary calcium intake 1000 mg/day) [56], and no anti-fracture benefit was observed in a large trial of
129
CHAPTER 42 Osteoporosis in Men
FIGURE 22
Algorithm for the choice of men at risk of fracture who may benefit from pharmacological therapy [519].
vitamin D supplementation in older men and women with relatively high baseline intakes [463]. Dietary calcium intake was not found to be related to fracture rate in the men followed as part of the Health Professionals Follow-up Study [464]. On the other hand, an improvement in bone density was noted in healthy older men in response to a calcium and vitamin D supplement, while placebo treated men lost bone [465]. On the basis of the available information, and the likelihood of a high degree of safety, the U.S. Institute of Medicine recently recommended that men should have a calcium intake of 1200 mg/day, and a vitamin D intake of 800 IU. A reasonable approach, therefore, is to suggest a calcium intake of at least 1200 mg/day in both preventative as well as therapeutic situations. A NIH Consensus Development Conference has suggested the somewhat higher calcium intake of 1500 mg/day in men after 65 years [466]. Although these recommendations for supplemental calcium and vitamin D are reasonable, some attention to individual differences is probably important. For instance, the use of an invariant level of vitamin D supplementation
(e.g., 800 IU/day) may result in inadequate effects in some patients, especially those who have low levels of vitamin D at baseline. In a study of the effects of vitamin D (and calcium supplementation) in men, Orwoll et al. found that the average increase in 25(OH) vitamin D concentration in response to 25 g (1000 IU) per day of cholecalciferol was 30 nmol/liter (12 ng/dl). However, the increase was no greater in those who started with reduced 25(OH) vitamin D levels (Fig. 23), with the result that men who start with low vitamin D status could be inadequately treated with conventional amounts of supplement. Certainly vitamin D insufficiency is common in older men [467 – 470], and adjustments in the dose of supplements based on initial vitamin D values may be useful. The use of follow-up vitamin D measurements should provide assurance that adequate vitamin D status has been achieved. Similarly, in some special situations (e.g., glucocorticoid excess, malabsorption) dietary calcium requirements may be somewhat increased over those routinely recommended. At a given level of sodium excretion, elderly men were found to have a greater calcium excretion than women (despite similar dietary
130
ORWOLL AND KLEIN
FIGURE 23.
The relationship between baseline and follow-up 25(OH)D levels in normal men treated with 25 g (1000 IU) per day of cholecalciferol or placebo for 1 year. The regression line for each group is shown [520].
calcium intakes) [471], suggesting that excess dietary sodium intake should be especially avoided in men at risk for osteoporosis. One theoretical concern regarding dietary calcium/vitamin D supplementation has been the precipitation of calcium renal stones in susceptible individuals. Recent data suggest that dietary calcium intake actually correlates negatively with the risk of nephrolithiasis in men [472], potentially by increasing gastrointestinal oxalate binding. The Institute of Medicine recommendations suggest that intakes below 2000 mg/day are safe. 2. IDIOPATHIC OSTEOPOROSIS Idiopathic osteoporosis in men is a perplexing and difficult disorder for the clinician. In addition to the lack of knowledge of its pathophysiology, there have been essentially no attempts to define appropriate treatments. In the absence of contributory factors that can be addressed, basic issues of nutritional adequacy (calcium and vitamin D) and physical activity (both for its trophic effects on the skeleton as well as the desire to maintain strength and coordination to prevent falls) should be addressed. As low bone mass must have its genesis in a remodeling imbalance, it may be appropriate to consider antiresorptive agents (bisphosphonates, calcitonin). The efficacy of alendronate in the therapy of men with osteoporosis (without secondary causes) was recently demonstrated [429], and in that light bisphos-
phonates should be considered a mainstay of therapy. Since many men with idiopathic osteoporosis appear to have impaired osteoblastic function, the development of boneforming therapies (e.g., PTH) is very attractive. 3. GLUCOCORTICOID EXCESS The current clinical management of glucocorticoid-induced osteoporosis is based on limited data, not only with regard to the efficacy of preventive and therapeutic regimens, but also in terms of our limited understanding of the pathophysiology of the disease. Various therapies including, calcium, vitamin D, calcitonin, bisphosphonates, sex steroids, and fluoride have been examined, but usually in open studies of limited subjects measuring effects on bone mass rather than large scale investigations evaluating fracture risk (see Chapter 44). Certainly management of patients receiving long-term glucocorticoid therapy should include minimally effective doses, at all times; discontinuation of the drug, when practical; and topical administration, if possible. Although alternate-day glucocorticoid dosing preserves normal function of the hypothalamic – pituitary – adrenal axis, there is no evidence that such a regimen offers any advantage in terms of preventing bone loss [473,474]. Calcium supplements diminish indices of bone resorption [475] and thiazide diuretics combined with reduced dietary sodium intake improve gastrointestinal absorption
131
CHAPTER 42 Osteoporosis in Men
of calcium and attenuate urinary calcium losses [476,477]. Pharmacologic doses of vitamin D have been widely used to treat glucocorticoid-induced osteoporosis in the past. Such therapy is not justified on the basis of vitamin D deficiency [478] and has not consistently shown therapeutic benefit [479 – 484]. However, in some studies which included male patients, vitamin D therapy appeared beneficial. For instance, Sambrook et al. [484] demonstrated a preservation of lumbar spine (but not femoral or radial) BMD with calcium and calcitriol. In this study, the addition of nasal calcitonin to calcium and calcitriol had no additional benefit. Vitamin D toxicity frequently accompanies use of pharmacological doses of vitamin D [479,481,482, 484,485]. Supplementation with lower doses (800 – 1000 IU/day) is certainly safe. Because long-term glucocorticoid therapy reduces serum testosterone levels and administration of testosterone to hypogonadal men improves bone mass, such therapy may be helpful, but has not been adequately evaluated. Sodium fluoride stimulates replication and function of osteoblasts and, as such, might be particularly useful in overcoming the primary inhibitory effects of glucocorticoids on the osteoblast. Several open and uncontrolled studies have revealed variable responses to treatment. One study of only 6 months duration found no effect
on the rate of glucocorticoid-induced bone loss [483], while another study examining long-term fluoride therapy demonstrated marked histologic improvement in indices of bone formation and trabecular mass [486]. Agents that inhibit bone resorption, such as calcitonin and bisphosphonates, have also been shown to be of therapeutic benefit [484, 487 – 491]. However, the efficacy of these agents appears to be greatest when administered in a preventive fashion from the time of initial exposure to glucocorticoids [484,487].
VII. HYPOGONADISM A. Androgen Replacement in Hypogonadal Adult Men Androgen therapy in hypogonadal men has been shown to positively affect bone mass, at least in most patient groups [305,312,492,493]. For instance, Katznelson et al. recently reported an increase in spinal BMD of 5 – 6% in a group of adult men with hypogonadism treated with testosterone for 18 months [494], although there was an insignificant increase in radial BMD (Fig. 24). As in the experience reported by Katznelson et al., the increase in density
FIGURE 24 Changes in percentage body fat and BMD in hypogonadal men receiving testosterone replacement therapy. (a) Percentage body fat determined by bioelectric impedance analysis. (b) AP spinal BMD determined by dual-energy X-ray absorptiometry. (c) Radial BMD determined by single photon absorptiometry. Data are represented as the mean SEM percentage of the baseline. Statistical significance for analysis of the mean slope is shown in the bottom right-hand corner of each figure [494].
132
ORWOLL AND KLEIN
FIGURE 25
Increase in spinal BMD during long-term testosterone substitution therapy up to 16 years in 72 hypogonadal patients. Circles indicate hypogonadal patients with first quantitative computed tomography (QCT) measurement before institution of therapy, squares show those patients already receiving therapy at the first measurement. The dark shaded area indicates the range of high fracture risk, the top area shows the range without significant fracture risk, and the middle area indicates the intermediate range where fractures may occur [495].
following testosterone replacement generally appears to be most apparent in cancellous bone (e.g., lumbar spine), although the literature is not particularly consistent in this regard. Most reports indicate that the increase in bone mass with testosterone therapy can be expected to be modest in the short term (up to 24 months), but Behre et al. noted an increase in spinal trabecular BMD of 20% in the first year of testosterone therapy in a group of hypogonadal men, and further increases thereafter [495] (Fig. 25). The most marked increases were observed in those with the lowest testosterone levels before therapy. In men treated for at least 3 years, bone density was found to be at levels normally expected for their ages. Although the experience remains small, there is a suggestion that in older men with hypogonadism the response to therapy can be expected to be similar to that in younger adult patients [495,496]. The cellular mechanisms responsible for improvements in bone mass are unclear. As discussed above, in the early phases of androgen deficiency (e.g., following castration) there appears to be a phase of increased remodeling and resorption, so that therapy may be beneficial because of an inhibitory effect on osteoclastic activity. However, in most available clinical studies, treated patient populations have had well-established hypogonadism and were characterized by an array of remodeling states. In these subjects the cellular effects of androgen replacement are not well known. In
some reports testosterone therapy appeared to result in an increase in cancellous bone formation [315,317], but in other series there appeared to be no clear remodeling trend induced by therapy [492]. Most recently, several groups have reported that biochemical indices of remodeling decline in response to testosterone replacement [494,497], which is what might be predicted if sex steroid deficiency results in an increase in remodeling, and bone loss on that basis. Interestingly, some reports also suggest that osteocalcin levels may increase with androgen therapy [496,498], perhaps signaling an increase in bone formation. In addition to the generally positive effects of androgen replacement therapy in hypogonadal men, additional benefits may be gained from the increases that have been noted in strength and lean body mass in these patients [494,496,499,500]. Since lean body mass and strength have been correlated with bone mass and a reduced propensity to fall, they may further serve to promote bone health and reduce fracture risk. Despite the generally positive tenor of most studies of the skeletal effects of testosterone replacement, in some patient groups, for instance those with Kleinfelter’s syndrome, the advantage associated with androgen therapy is questionable, as the available studies report very mixed results [501,502]. This may be because the level of androgen deficiency in Kleinfelter’s (as in the case of some other
133
CHAPTER 42 Osteoporosis in Men
causes of hypogonadism) is quite variable. These findings suggest the need to carefully consider the potential benefits of androgen replacement in each patient individually. The most efficacious doses and routes of androgen administration for the prevention/therapy of bone loss in men remain uncertain. The specific testosterone values necessary for an optimal effect have not been defined, but current practice is to attempt to ensure testosterone concentrations similar to those of normal young men. Moreover, whether the pulsatile pattern of testosterone exposure characteristic of intramuscular administration is more or less conducive to skeletal health than the more stable pattern produced by transdermal administration is unknown. In some studies, transdermal testosterone therapy appeared to be as effective as was intramuscular administration in promoting bone mass [495]. Current oral preparations of androgens are not appropriate in view of the higher incidence of adverse effects associated with their use. However, newer androgenic compounds may obviate this concern. The follow-up of hypogonadal men treated with testosterone, although not well codified, should certainly include careful monitoring for adverse effects. The risk of prostate disease in androgen treated men is unknown, but regular prostate evaluations are necessary to ensure that any development of benign or malignant disease is detected early in its course. The development of errythrocytosis is not uncommon, particularly with intramuscular testosterone administration, and complete blood counts at 6 – 12 month intervals are useful to detect its appearance. Other problems that have been postulated to be of concern in androgentreated men are hyperlipidemia and sleep apnea [503]. In terms of skeletal disease, therapeutic success may be assessed via follow-up bone mass measures. In view of recent reports, increases in bone density can be anticipated in the average patient. Although the role of biochemical markers of remodeling is controversial, the available data suggest that an adequate androgen effect should be accompanied by a fall in indices of bone resorption, an effect that should be especially useful if resorption markers are increased at baseline. Markers of bone formation may be more difficult to use at present in routine clinical situations, as some reports suggest that increases follow therapy while others support a decline. The response may depend on the specific marker. Clinicians deciding on a follow-up strategy must be aware of the uncertainty currently inherent in the field and the vagaries of using the available tools (i.e., issues of measurement precision). There remain many additional unresolved issues concerning the role of androgen treatment in the prevention/ therapy of osteoporosis in hypogonadal men, including: • The degree of hypogonadism (level of testosterone) at which adverse skeletal effects begin to occur is undefined, and hence it is difficult to decide upon the usefulness of
therapy in many men with borderline levels of serum testosterone. • Because hypogonadism in men results in deficiencies of estrogen as well as testosterone, and since testosterone therapy results in increases in serum estrogen (as well as androgen) levels, the relative roles of estrogen vs testosterone in affecting skeletal health in hypogonadal men are unclear. It is unknown whether it is useful to assess estrogen concentrations in the diagnosis of hypogonadal bone disease in men, or whether using estrogen levels to monitor the success of testosterone therapy is beneficial. • In general, the available treatment studies are of relatively short duration, and it is unclear how long any increases in bone mass can be sustained and what eventual treatment effect can be expected. • As of yet, the increase in bone mass that appears to accompany testosterone therapy is of uncertain usefulness in preventing fractures. • Whether pretreatment age, duration of hypogonadism, degree of osteopenia, remodeling character, and associated medical conditions affect the therapeutic response is relatively unknown. • Potential adverse effects of androgen therapy (e.g., prostate, lipid) are not well delineated.
B. Androgen Therapy in Eugonadal Men It has been hypothesized that androgens may have positive effects on bone formation and resorption. The threshold level of androgens necessary to provide maximal skeletal benefits is unknown, and some have speculated that testosterone supplementation would benefit osteoporotic men even in the face of normal testosterone levels. The experience with this approach has been very limited, but Anderson et al. recently found in an uncontrolled trial that testosterone supplementation was associated with an increase in bone density, and a reduction in biochemical markers of remodeling, in a group of osteoporotic, eugonadal men [504]. This approach remains very much of uncertain benefit, and until its advantages are documented in controlled trials it cannot be recommended. This is particularly true in view of the lack of knowledge concerning the potential adverse effects that may be associated with testosterone supplementation.
C. Androgen Replacement in Adolescence Because adolescence is such a critically important part of the process of attaining optimal peak bone mass, it is also especially vulnerable to disruption by alterations in gonadal function. Even constitutional pubertal delay is associated with a reduction in peak bone mass development,
134
ORWOLL AND KLEIN
despite eventual full gonadal development [294, 296]. The impairment in bone mass in adolescence with organic hypogonadism (hypogonadotropic hypogonadism) is similar to that in patients with this form of hypogonadism studied later in life, suggesting that the detrimental effect suffered in adolescence is the major cause of osteopenia [290]. In view of the major effects of androgens on the skeleton during growth (whether direct or indirect, as discussed above), the response to therapy of gonadal dysfunction during this time would be expected to be brisk. Although studies are few, this would appear to be the case [505]. Finkelstein et al. reported that treatment of hypogonadal men with testosterone elicited the most robust skeletal response in those who were skeletally immature (open epiphyses) [492]. In young men considered to have constitutional delay of puberty, testosterone therapy results in a clear increase in bone mass, but whether this provides a solution to the problem of low peak bone mass in these patients is not yet known [506]. All this information suggests that the diagnosis of frank hypogonadism during childhood or adolescence carries with it the risk of impaired skeletal development, and that there is an opportunity to improve bone mass with androgen therapy. In fact, from a skeletal perspective, it appears that therapy should be initiated before epiphyseal closure to maximize bone mass accumulation. Issues that are unresolved include whether bone mass can be normalized with therapy, the most appropriate doses and timing of therapy, and the source of the beneficial effects (androgen vs estrogen, growth factor stimulation, etc.).
the levels (threshold concentrations) that are associated with adverse effects on bone.
E. Androgen Therapy in Secondary Forms of Metabolic Bone Disease A variety of system illnesses and medications are associated with lowered testosterone levels [508], and it has been postulated that relative hypogonadism may contribute to the bone loss that also accompanies many of these conditions. For instance, renal insufficiency, glucocorticoid excess, post-transplantation, malnutrition, and alcoholism are all associated with osteopenia and with low testosterone concentrations. Although there is little experience with testosterone supplementation in these patients, there may be advantages to skeletal health as well as to other tissues (muscle, red cells, etc.). In a randomized study of crossover design, Reid et al. [509] reported that testosterone therapy apparently improved bone density (and body composition) in a small group of men receiving glucocorticoids (Fig. 26). Similarly, testosterone therapy apparently improved forearm bone mass in a small group of men with hemochromatosis (treated simultaneously with venesection) [305]. The number of patients affected by conditions associated with low testosterone levels is potentially quite large, and more information is needed to understand the role of androgen replacement in the prevention/therapy of concomitant bone loss.
D. Androgen Replacement in Aging Men Old age is associated with a panoply of physical changes in men, many of which have been speculated to be related, either directly or indirectly, to the decline in androgens that accompanies aging [507]. A few small trials of androgen administration in older men have suggested that there may be beneficial effects (increased strength and improved body composition) [217,496,500], and some reports indicate that bone mass or biochemical indices of remodeling may improve [217,218,496]. Whether androgen replacement therapy can prevent or reverse bone loss in aging men is of enormous importance, but until more definitive data are available concerning both advantages and disadvantages, testosterone replacement should not be utilized in elderly patients unless there is convincing evidence for androgen deficiency. This decision is difficult in many older men who have symptoms that can be associated with androgen deficiency but which are also common in the aged regardless of gonadal status (weakness, loss of libido or sexual ability, etc.). The identification of hypogonadism in this group is made especially challenging by the expected decline in androgen levels with age and the dearth of data concerning
FIGURE 26
Rate of change in BMD of lumbar spine during control or testosterone treatment periods (each of 6 months duration) in men receiving glucocorticoid therapy. Data are given as the mean SEM. There was a significant difference between groups (P 0.05). The asterisk indicates a significant difference from 0 (P 0.005) [509].
135
CHAPTER 42 Osteoporosis in Men
F. New Development in Sex Steroid Therapy in Men Selective estrogen receptor modulators have considerably altered the concept of estrogen replacement therapy in postmenopausal women (see Chapter 70). Considerable interest has developed concerning the relevance of these compounds in the treatment of men as well. Animal studies suggest that selective estrogen receptor modulator may have encouraging effects in males [510], and clinical trials are underway in men. Finally, selective androgen receptor modulators are being developed, and promise to be useful as osteoprotective agents while reducing adverse effects on prostate, lipids, etc. [511].
reported. The toxic effects of alcohol and fluoride on the gastrointestinal tract may likely preclude its use in individuals that continue to drink.
IX. TOBACCO The ideal approach to osteoporosis associated with tobacco use is smoking cessation. Whether cessation leads to reduced rates of bone loss or to a gain in bone mass is unknown. In the studies by Slemenda et al. [130] there was apparently a protective effect of heavy physical activity on the bone loss induced by smoking. Other interventions (nutritional supplements, antiresorptive drugs) might potentially reduce the incidence of low bone mass in men who did not abstain, although these possibilities are untested.
VIII. ALCOHOLISM Osteoporosis should be suspected in every chronic alcohol abuser, and patients with “idiopathic” osteoporosis should be routinely and thoroughly questioned about drinking habits. Once the diagnosis of alcohol-induced bone disease has been established, a number of measures are recommended. Aggressive medical and psychiatric treatment should be pursued in the hopes of interrupting the cycle of chronic alcohol ingestion and thereby diminish the risk of further skeletal deterioration. A careful dietary history should be followed by an adequate well-balanced diet rich in calcium-containing products. Evidence that calcium supplementation will improve the bone disease of alcoholics has not been reported, but it is reasonable to minimize other potential risk factors for bone loss if possible. Adequate vitamin D nutrition and physical exercise should be encouraged. Tobacco use and excessive consumption of phosphate-binding antacids should be discouraged. Presumably, the cessation of alcohol intake will stop further progression of bone loss, but data are scant. Moreover, no evidence has been reported that bone, once lost, will be restored when alcohol abuse is discontinued. Studies on alcohol abstainers have demonstrated a rapid recovery of osteoblast function (as assessed histomorphometrically and by biochemical parameters of bone remodeling) within as little as 2 weeks after cessation of drinking, but no significant differences in bone mineral content were observed between abstainers and actively drinking men [302,326,332, 348,361]. The relatively short period of abstinence however, makes these results inconclusive. The challenge in alcohol-induced bone disease is to stimulate bone formation. Most drugs currently used to treat other forms of osteoporosis work primarily by inhibiting osteoclastic bone resorption. Agents such as fluoride, parathyroid hormone or growth hormone may stimulate bone formation, but such regimens remain investigational and no therapeutic trials in alcoholic men have been
X. RENAL STONE DISEASE Attempts to treat low bone mass associated with idiopathic hypercalciuria are not yet well developed, but have been summarized recently [378]. As the most likely cause of defects in bone remodeling result from the mineral abnormalities induced by the renal calcium disturbance, it seems prudent to prevent hypercalciuria. In patients with either renal or absorptive hypercalciuria, thiazide diuretics would be appropriate. It is unknown whether therapy of the metabolic disturbances present in some patients with hypercalciuric renal stones (e.g., acidosis) is associated with any benefit.
References 1. N. F. Ray, J. K. Chan, M. Thamer, and L. J. Melton, Medical expenditures for the treatment of osteoporotic fractures in the united states in 1995: Report from the national osteoporosis foundation. J. Bone Miner. Res. 12, 24 – 35 (1997). 2. D. T. Gold, Outcomes and the personal impact of osteoporosis. In “Osteoporosis in Men: The Effects of Gender on Skeletal Health” (E. S. Orwoll, ed.), pp. 51 – 63. Academic Press, San Diego, 1999. 3. C. Niewoehner, Osteoporosis in men: Is it more common than we think? Postgrad. Med. 93, 59 – 60, 63 – 70 (1993). 4. E. Seeman, Osteoporosis in men: Epidemiology, pathophysiology, and treatment possibilities. Am. J. Med. 95, 22S – 28S (1993). 5. A. C. Scane, A. M. Sutcliffe, and R. M. Francis, Osteoporosis in men. Clin. Rheum. 7, 589 – 601 (1992). 6. T. J. Arneson, L. J. I. Melton, D. G. Lewallen, and W. M. O’Fallon, Epidemiology of diaphyseal and distal femoral fractures in Rochester, Minnesota, 1965 – 1984. Clin. Orthop. Relat. Res. 234, 188 – 194 (1988). 7. L. J. I. Melton and S. R. Cummings, Heterogeneity of age-related fractures: Implications for epidemiology. J. Bone Miner. Res. 2, 321 – 331 (1987). 8. K. M. Sanders, E. Seeman, A. M. Ugoni, J. A. Pasco, T. J. Martin, B. Skoric, G. C. Nicholson, and M. A. Kotowicz, Age- and genderspecific rate of fractures in Australia: A population-based study. Osteoporosis Int. 10, 240 – 247 (1999).
136 9. A. A. o. Orthopaedic Surgeons, Musculoskeletal injuries. In “Musculoskeletal conditions in the United States” (A. A. o. O. Surgeons, ed.), pp. 85 – 124. American Academy of Orthopaedic Surgeons, Park Ridge, IL, 1992. 10. M. Karlsson, R. Hasserius, and E. J. Obrant, Individuals who sustain nonosteoporotic fractures continue to also sustain fragility fractures. Calcif. Tissue. Int. 53, 229 – 231 (1993). 11. M. E. Farmer, L. R. White, J. A. Brody, and K. R. Bailey, Race and sex differences in hip fracture incidence. Am. J. Public Health 74, 1374 – 1380 (1984). 12. W. E. Bacon, G. S. Smith, and S. P. Baker, Geographic variation in the occurrence of hip fractures among the elderly white US population. Am. J. Public Health 79, 1556 – 1558 (1989). 13. S. J. Jacobsen, J. Goldberg, T. P. Miles, J. A. Brody, W. Stiers, and A. A. Rimm, Hip fracture incidence among the old and very old: A population-based study of 745,435 cases. Am. J. Public Health 80, 871 – 873 (1990). 14. J. D. Gallagher, L. J. Melton, and B. L. Riggs, Epidemiology of fracture of the proximal femur in Rochester, MN. Clin. Orthop. Relat. Res. 150, 163 – 171 (1980). 15. J. A. Kanis, The incidence of hip fracture in Europe. Osteoporosis Int. 1, S10 – S15 (1993). 16. J. Chalmers and K. C. Ho, Geographical variations in senile osteoporosis: The association with physical activity. J. Bone Jt. Surg. 52B, 667 – 675 (1970). 17. L. Solomon, Osteoporosis and fracture of the femoral neck in the South African Bantu. J. Bone Jt. Surg. 50B, 2 – 13 (1968). 18. P. C. N. Wong, Fracture epidemiology in a mixed southeastern Asian community (Singapore). Clin. Orthop. 45, 55 – 61 (1966). 19. A. V. Schwartz, J. L. Kelsey, S. Maggi, M. Tuttleman, S. C. Ho, P. V. Jonsson, G. Poor, J. A. Sisson de Castro, L. Xu, C. C. Matkin, L. M. Nelson, and S. P. Heyse, International variation in the incidence of hip fractures: Cross-national project on osteoporosis for the world health organization program for research on aging. Osteoporosis Int. 9, 242 – 253 (1999). 20. L. J. Melton, III and E. A. Chrischilles, Perspective: How many women have osteoporosis? J. Bone Miner. Res. 7, 1005 – 1010 (1992). 21. A. D. Martin, K. G. Silverthorn, C. S. Houston, S. Bernhardson, A. Wajda, and L. L. Roos, The incidence of fracture of the proximal femur in two million Canadians from 1972 to 1984. Clin. Orthop. Rel. Res. 266, 111 – 118 (1991). 22. C. Cooper and G. Campion, Hip fractures in the elderly: a worldwide projection. Osteoporosis Int. 2(6), 285 – 289 (1992). 23. E. L. Schneider and J. M. Guralnik, The aging of America: impact on health care costs. JAMA 263, 2335 – 2340 (1990). 24. T. M. Huusko, P. Karppi, V. Avikainen, H. Kautiainen, and R. Sulkava, The changing picture of hip fractures: Dramatic change in age distribution and no change in age-adjusted incidence within 10 years in central Finland. Bone 24, 257 – 259 (1999). 25. B. Gullberg, O. Johnell, and J. A. Kanis, World-wide projections for hip fracture. Osteoporosis Int. 7, 407 – 413 (1997). 26. L. J. Melton, III, W. M. O’Fallon, and B. L. Riggs, Clinical investigations: Secular trends in the incidence of hip fractures. Calcif. Tissue Int. 41, 57 – 64 (1987). 27. J. A. Falch, T. S. Kaastad, and G. Bohler, Secular increase and geographical differences in hip fracture incidence in Norway. Bone 14, 634 – 645 (1993). 28. J. Parkkari, P. Kannus, S. Niemi, M. Pasanen, M. Jarvinen, P. Luthje, and I. Vuori, Increasing age-adjusted incidence of hip fractures in Finland: The number and incidence of fractures in 1970 – 1991 and prediction for the future. Calcif. Tissue Int. 55, 342 – 345 (1994). 29. L. J. Melton, C. S. Crowson, and W. M. O’Fallon, Fracture incidence in Olmsted County, Minnesota: Comparison of urban with rural rates and changes in urban rates over time. Osteoporosis Int. 9, 29 – 37 (1999).
ORWOLL AND KLEIN 30. G. Poor, E. J. Atkinson, W. M. O’Fallon, and L. J. Melton, Determinants of reduced survival following hip fractures in men. Clin. Orthop. Relat. Res. 319, 260 – 265 (1995). 31. L. J. Melton, III and B. L. Riggs, Epidemiology of age-related fractures. In “The Osteoporotic Syndrome” (L. V. Avioli, ed.), pp. 45 – 72. Grune & Stratton, New York, 1983. 32. J. R. Center, T. V. Nguyen, D. Schneider, P. N. Sambrook, and J. A. Eisman, Mortality after all mamor types of osteoporotic fracture in men and women: An observational study. Lancet 353, 878 – 882 (1999). 33. E. S. Fisher, J. A. Baron, D. J. Malenka, J. A. Barrett, W. D. Kniffin, F. S. Whaley, and T. A. Bubolz, Hip fracture incidence and mortality in new england. Epidemiology 2, 116 – 122 (1991). 34. I. Diamond and A. S. Gordon, Cellular and molecular neuroscience of alcoholism. Physiol. Rev. 77, 1 – 20 (1997). 35. S. P. Heyse, Epidemiology of hip fractures in the elderly: A crossnational analysis of mortality rates for femoral neck fractures. Osteoporosis Int. 1, S16 – S19 (1993). 36. J. A. Baron, M. Karagas, J. Barrett, W. Kniffin, D. Malenka, M. Mayor, and R. B. Keller, Basic Epidemiology of fractures of the upper and lower limb among Americans over 65 years of age. Epidemiology 7, 612 – 618 (1996). 37. P. D. Ross, H. Norimatsu, J. W. Davis, K. Yano, R. D. Wasnich, S. Fujiwara, Y. Hosoda, and L. J. I. Melton, A comparison of hip fracture incidence among native Japanese, Japanese Americans, and American Caucasians. Am. J. Epidemiol. 133, 801 – 809 (1991). 38. R. Madhok, L. J. Melton, E. J. Atkinsom, W. M. O’Fallon, and D. G. Lewallen, Urban vs. rural increase in hip fracture incidence. Acta. Orthop. Scand. 64, 543 – 548 (1993). 39. I. Elffors, E. Allander, J. A. Kanis, B. Gullberg, O. Johnell, J. Dequeker, G. Dilsen, C. Gennari, A. A. Lopes Vaz, G. Lyritis, G. F. Mazzuoli, L. Miravet, M. Passeri, R. P. Cano, A. Rapado, and C. Ribot, The variable incidence of hip fracture in Southern Europe: The MEDOS study. Osteoporosis Int. 4, 253 – 263 (1994). 40. H. Burger, P. L. A. Van Daele, K. Grashuis, A. Hofman, D. E. Grobbee, H. E. Schutte, J. C. Birkenhager, and H. A. P. Pols, Vertebral deformities and functional impairment in men and women. J. Bone Miner. Res. 12, 152 – 157 (1997). 41. A. C. Scane, A. M. Sutcliffe, and R. M. Francis, The sequelae of vertebral crush fractures in men. Osteoporosis Int. 4, 89 – 92 (1994). 42. T. Mann, S. K. Oviatt, D. Wilson, D. Nelson, and E. S. Orwoll, Vertebral deformity in men. J. Bone Miner. Res. 7, 1259 – 1265 (1992). 43. C. Cooper, E. J. Atkinson, W. M. O’Fallon, and L. J. Melton, Incidence of clinically diagnosed vertebral fractures: A population-based study in Rochester, Minnesota, 1985 – 1989. J. Bone Miner. Res. 7, 221 – 227 (1992). 44. J. A. Kanis and E. V. McCloskey, Epidemiology of vertebral osteoporosis. Bone 13, S1 – S10 (1992). 45. T. W. O’Neill, D. Felsenberg, J. Varlow, C. Cooper, J. A. Kanis, and A. J. Silman, The prevalence of vertebral deformity in european men and women: The eurpoean vertebral osteoporosis study. J. Bone Miner. Res. 11, 1010 – 1018 (1996). 46. K. M. Davies, M. R. Stegman, R. P. Heaney, and R. R. Recker, Prevalence and severity of vertebral fracture: The saunders county bone quality study. Osteoporosis Int. 6, 160 – 165 (1996). 47. A. J. Silman, T. W. O’Neill, C. Cooper, J. Kanis, and D. Felsenberg, Influence of physical activity on vertebral deformity in men and women: Results from the European vertebral osteoporosis study. J. Bone Miner. Res. 12, 813 – 819 (1997). 48. M. Lunt, D. Felsenberg, J. Reeve, L. Benevolenskaya, J. Cannata, J. Dequeker, C. Dodenhof, J. A. Falch, P. Masaryk, H. A. P. Pols, G. Poor, D. M. Reid, C. Scheidt-Nave, K. Weber, J. Varlow, J. A. Kanis, T. W. O’Neill, and A. J. Silman, Bone density variation and its effects on risk of vertebral deformity in men and women studied in thirteen european centers: The EVOS study. J. Bone Miner. Res. 12, 1883 – 1894 (1997).
CHAPTER 42 Osteoporosis in Men 49. W. M. Garraway, R. N. Stauffer, L. T. Kurland, and W. M. O’Fallon, Limb fractures in a defined population. I. Frequency and distribution. Mayo Clin. Proc. 54, 701 – 707 (1979). 50. A. J. Buhr and A. M. Cooke, Fracture patterns. Lancet 1, 531 – 536 (1959). 51. D. Hemenway, D. R. Azrael, E. B. Rimm, D. Feskanich, and W. C. Willett, Risk factors for wrist fracture: Effect of age, cigarettes, alcohol, body height, relative weight and handedness on the risk for distal forearm fractures in men. Am. J. Epidemiol. 140, 361 – 367 (1994). 52. M. T. Cuddihy, S. E. Gabriel, C. S. Crowson, W. M. O’Fallon, and L. J. Melton, Forearm fractures as predictors of subsequent osteoporotic fractures. Osteoporosis Int. 9, 469 – 475 (1999). 53. H. Mallmin, S. Ljunghall, I. Persson, T. Naessen, U. B. Krusemo, and R. Bergstrom, Fracture of the distal forearm as a forecaster of subsequent hip fracture: A population-based cohort study with 24 years follow-up. Calcif. Tissue Int. 52, 269 – 272 (1993). 54. R. M. Francis, M. Peacock, D. H. Marshall, A. Horsman, and J. E. Aaron, Spinal osteoporosis in men. Bone Miner. 5, 347 – 357 (1989). 55. H. K. Genant, G. S. Gordan, and P. G. J. Hoffman, Osteoporosis Part I. Advanced radiologic assessment using quantitative computed tomography — Medical Staff Conference, University of California, San Francisco. West J. Med. 139, 75 – 84 (1983). 56. E. S. Orwoll, S. K. Oviatt, M. R. McClung, L. J. Deftos, and G. Sexton, The rate of bone mineral loss in normal men and the effects of calcium and cholecalciferol supplementation. Ann. Int. Med. 112, 29 – 34 (1990). 57. B. L. Riggs, H. W. Wahner, E. Seeman, K. P. Offord, W. L. Dunn, R. B. Mazess, K. A. Johnson, and L. J. Melton, Changes in bone mineral density of the proximal femur and spine with aging. J. Clin. Invest. 70, 716 – 723 (1982). 58. P. D. Ross, R. D. Wasnich, and J. M. Vogel, Detection of prefracture spinal osteoporosis using bone mineral absorptiometry. J. Bone Miner. Res. 3, 1 – 11 (1988). 59. R. C. Hamdy, S. Moore, V. Cancellaro, K. Whalen, C. Landy, D. Chi, S. Reynolds, and G. Cambron, “Osteoporosis in Men — Prevalence and Biochemical Parameters.” 1992 Program and Abstracts of the 14th Annual Meeting of the American Society for Bone and Mineral Research, Liebert, Mary Ann, Inc., Minneapolis, MN, 1992. 60. P. D. Ross, A. Lombardi, and D. Freedholm, The assessment of bone mass in men. In “Osteoporosis in Men: The Effects of Gender on Skeletal Health” (E. S. Orwoll, ed.), pp. 505 – 525. Academic Press, San Diego, 1999. 61. E. Vega, C. Mautalen, H. Gomez, A. Garrido, L. Melo, and A. O. Sahores, Bone mineral density in patients with cervical and trochanteric fractures of the proximal femur. Osteoporosis Int. 1, 81 – 86 (1991). 62. M. E. Mussolino, A. C. Looker, J. H. Madans, J. A. Langlois, and E. S. Orwoll, Risk factors for hip fracture in white men: The NHANES I epidemiologic follow-up study. J. Bone Miner. Res. 13, 918 – 924 (1998). 63. C. E. D. H. De Laet, B. A. van Hout, H. Burger, A. Hofman, and H. A. P. Pols, Bone density and risk of hip fracture in men and women: Cross sectional analysis. Br. Med. J. 315, 221 – 225 (1997). 64. T. V. Nguyen, J. A. Eisman, P. J. Kelly, and P. N. Sambrook, Risk factors for osteoporotic fractures in elderly men. Am. J. Epidemiol. 144, 258 – 261 (1996). 65. H. Kroger, M. Lunt, J. Reeve, J. Dequeker, J. E. Adams, J. C. Birkenhager, M. Diaz Curiel, D. Felsenberg, L. Hyldstrup, P. Kotzki, A.-M. Laval-Jeantet, P. Lips, O. Louis, R. Perez Cano, C. Reiners, C. Ribot, P. Ruegsegger, P. Schneider, P. Braillon, and J. Pearson, Bone density reduction in various measurement sites in men and women with osteoporotic fractures of spine and hip: The european quantitation of osteoporosis study. Calcif. Tissue Int. 64, 191 – 199 (1999).
137 66. C. E. D. H. De Laet, B. A. Van Hout, H. Burger, A. E. A. M. Weel, A. Hofman, and H. A. P. Pols, Hip fracture prediction in elderly men and women: Validation of the Rotterdam Study. J. Bone Mineral. Res. 13, 1587 – 1593 (1998). 67. L. J. Melton, E. J. Atkinson, M. K. O’Connor, W. M. O’Fallon, and B. L. Riggs, Bone density and fracture risk in men. J. Bone Miner. Res. 13, 1915 – 1923 (1998). 68. T. Nguyen, P. Sambrook, P. Kelly, G. Jones, S. Lord, and J. Freund. Prediction of osteoporotic fractures by postural instability and bone density. Br. Med. J. 307, 1111 – 1115 (1993). 69. C. M. Klotzbuecher, P. D. Ross, P. B. Landsman, T. A. Abbott, and M. Berger, Patients with prior fractures have an increased risk of future fractures: A summary of the literature and statistical synthesis. J. Bone Miner. Res. 15, 721 – 739 (2000). 70. S. Fujiwara, F. Kasagi, M. Yamada, and K. Kodama, Risk factors for hip fracture in a Japanese cohort. J. Bone Miner. Res. 12, 998 – 1004 (1997). 71. L. J. Melton, E. J. Atkinson, C. Cooper, W. M. O’Fallon, and B. L. Riggs, Vertebral fractures predict subsequent fractures. Osteoporosis Int. 10, 214 – 221 (1999). 72. A. M. Tromp, J. H. Smit, D. J. H. Deeg, L. M. Bouter, and P. Lips, Predictors for falls and fractures in the Longitudinal Aging Study Amsterdam. J. Bone Miner. Res. 13, 1932 – 1939 (1998). 73. W. C. Graafmans, M. E. Ooms, H. M. A. Hofstee, P. D. Bezemer, L. M. Bouter, and P. Lips, Falls in the elderly: A prospective study of risk factors and risk profiles. Am. J. Epidemiol. 143, 1129 – 1136 (1996). 74. K. Aoyagi, P. D. Ross, J. W. Davis, R. D. Wasnich, T. Hayashi, and T. Takemoto, Falls amont community-dwelling elderly in japan. J. Bone Miner. Res. 13, 1468 – 1474 (1998). 75. P. Kannus, J. Parkkari, S. Koskinen, S. Niemi, M. Palvanen, M. Jarvinen, and I. Vuori, Fall-induced injuries and deaths among older adults. JAMA 281, 1895 – 1899 (1999). 76. J. A. Grisso, G. Y. Chiu, G. Maislin, W. C. Steinmann, and J. Portale, Risk factors for hip fractures in men: a preliminary study. J. Bone Miner. Res. 6, 865 – 868 (1991). 77. H. E. Meyer, A. Tverdal, and J. A. Falch, Risk factors for hip fracture in middle-aged Norwegian women and men. Am. J. Epidemiol. 137, 1203 – 1211 (1993). 78. A. V. Schwartz, J. L. Kelsey, S. Sidney, and J. A. Grisso, Characteristics of falls and risk of hip fracture in elderly men. Osteoporosis Int. 8, 240 – 246 (1998). 79. W. C. Hayes, E. R. Myers, J. N. Morris, T. N. Gerhardt, H. S. Yett, and L. A. Lipsitz, Impact near the hip dominates fracture risk in elderly nursing home residents who fall. Calcif. Tissue Int. 52, 192 – 198 (1993). 80. S. L. Greenspan, E. R. Myers, L. A. Maitland, N. M. Resnick, and W. C. Hayes, Fall severity and bone mineral density as risk factor for hip fracture in ambulatory elderly. JAMA 271, 128 – 133 (1994). 81. D. Hemenway, D. R. Azrael, and D. Feskanich, “The bigger they are . . . ”: The effect of height, relative weight, alcohol intake and cigarette smoking on the risk of hip fracture in middle-aged men. Am. J. Public Health (1994). 82. W. A. Ray, M. R. Griffin, W. Schaffner, D. K. Baugh, and L. J. Melton III, Psychotropic drug use and the risk of hip fracture. N. Engl. J. Med. 316, 363 – 370 (1987). 83. W. A. Ray, M. R. Griffin, and W. Downey, Benzodiazepines of long and short elimination half-life and the risk of hip fracture. JAMA 262, 3303 – 3307 (1989). 84. M. K. Karlsson, O. Johnell, B. E. Nilsson, I. Sernbo, and K. J. Obrant, Bone mineral mass in hip fracture patients. Bone 14, 161 – 165 (1993). 85. T. W. O’Neill, J. Varlow, A. J. Silman, J. Reeve, D. M. Reid, C. Todd, and A. D. Woolf, Age and sex influences on fall characteristics. Ann. Rheum. Dis. 53, 773 – 775 (1994).
138 86. S. M. Garn, “The Earlier Gain and the Later Loss of Cortical Bone in Nutritional Perspective.” Charles C. Thomas, Springfield, IL, 1970. 87. V. Gilsanz, Accumulation of bone mass during childhood and adolescence. In “Osteoporosis in Men: The Effects of Gender on Skeletal Health” (E. S. Orwoll, ed.), pp. 65 – 85. Academic Press, San Diego, 1999. 88. E. Seeman, Bone size, mass, and volumetric density: The importance of structure on skeletal health. In “Osteoporosis in Men: The Effects of Gender on Skeletal Health” (E. S. Orwoll, ed.), pp. 87 – 109. Academic Press, San Diego, 1999. 89. A. Vesterby, L. Mosekilde, and H. J. G. Gundersen, Biologically meaningful determinants of the in vitro strength of lumbar vertebrae. Bone 12, 219 – 224 (1991). 90. P. Brinckmann, M. Biggemann, and D. Hilweg, Prediction of the compressive strength of human lumbar vertebrae. Spine 14, 606 – 610 (1989). 91. S. D. Rockoff, E. Sweet, and J. Bleustein, The relative contribution of trabecular and cortical bone to the strength of human lumbar vertebrae. Calcif. Tissue Res. 3, 163 – 175 (1969). 92. V. Gilsanz, E. E. Shultz, and L. Loro, Smaller cross-sectional area of intact vertebral bodies in women with vertebral fractures. J. Bone Miner. Res. 8, S326 (1993). 93. L. Mosekilde and L. Mosekilde, Normal vertebral body size and compressive strength relations to age and to vertebral and iliac trabecular bone compression strength. Bone 7, 207 – 212 (1986). 94. V. Matkovic, T. Jelic, and G. M. Wardlaw, Timing of peak bone mass in caucasian females and its implication for the prevention of osteoporosis: Inference from a cross-sectional model. J. Clin. Invest. 93 (1994). 95. L. Mosekilde and L. Mosekilde, Sex differences in age-related changes in vertebral body size, density and biomechanical competence in normal individuals. Bone 11, 67 – 73 (1990). 96. R. W. J. Smith and R. R. Walker, Femoral expansion in aging women: implications for osteoporosis and fractures. Science 145, 156 – 157 (1964). 97. T. J. Beck, C. B. Ruff, W. W. J. Scott, C. C. Plato, J. D. Tobin, and C. A. Quan, Sex differences in geometry of the femoral neck with aging: A structural analysis of bone mineral data. Calcif. Tissue Int. 50, 24 – 29 (1992). 98. T. J. Beck, C. B. Ruff, and K. Bissessur, Age-related changes in female femoral neck geometry: Implications for bone strength. Calcif. Tissue Int. 53, S41 – S46 (1993). 99. B. Martin, Aging and strength of bone as a structural material. Calcif. Tissue Int. 53 (Suppl 1), S34 – S40 (1993). 100. H. Brockstedt, M. Kassem, and E. F. Eriksen, Age- and sex-related changes in iliac cortical bone mass and remodeling. Bone 14, 681 – 691 (1993). 101. S. M. Garn, T. V. Sullivan, S. A. Decker, F. A. Larkin, and V. M. Hawthorne, Continuing bone expansion and increasing bone loss over a two-decade period in men and women from a total community sample. Am. J. Hum. Biol. 4, 57 – 67 (1992). 102. C. B. Ruff and W. C. Hayes, Sex differences in age-related remodeling of the femur and tibia. J. Orthop. Res. 6, 886 – 896 (1988). 103. R. B. Martin and P. J. Atkinson, Age- and sex-related changes in the structure and strength of the human femoral shaft. J. Biomech. 10, 223 – 231 (1977). 104. G. Jones, T. Nguyen, P. Sambrook, P. J. Kelly, and J. A. Eisman, Progressive loss of bone in the femoral neck in elderly people: Longitudinal findings from the Dubbo osteoporosis epidemiology study. Br. Med. J. 309, 691 – 695 (1994). 105. H. Burger, C. E. D. H. de Laet, P. L. A. van Daele, A. E. A. M. Weel, J. C. M. Wittman, A. Hofman, and H. A. P. Pols, Risk factors for increased bone loss in an elderly population. Am. J. Epidemiol. 147, 871 – 879 (1998).
ORWOLL AND KLEIN 106. H. Burger, P. L. A. van Daele, D. Algra, F. A. van den Ouweland, D. E. Grobbee, A. Hofman, C. van Kuijk, H. E. Schutte, J. C. Birkenhager, and H. A. P. Pols, The association between age and bone mineral density in men and women aged 55 years and over: The Rotterdam Study. Bone Miner. 25, 1 – 13 (1994). 107. J. E. Compston, R. W. E. Mellish, P. Croucher, R. Newcombe, and N. J. Garrahan, Structural mechanism of trabecular bone loss in men. Bone 8, 339 – 350 (1989). 108. D. Prudham and J. G. Evans, Factors associated with falls in the elderly: A community study. Age Ageing 10, 141 – 146 (1981). 109. J.-P. Bonjour, G. Theintz, F. Law, D. Slosman, and R. Rizzoli, Peak bone mass. Osteoporosis Int. 1, S7 – S13 (1994). 110. J.-P. Bonjour, G. Theintz, B. Buchs, D. Slosman, and R. Rizzoli, Critical years and stages of puberty for spinal and femoral bone mass accumulation during adolescence. J. Clin. Endocrinol. Metab. 73, 555 – 563 (1991). 111. V. Gilsanz, M. I. Beochat, and T. Roe, Gender differences in vertebral dimensions in children and adolescents. Radiology 190, 673 – 677 (1993). 112. H. Rico, M. Revilla, E. R. Hernandez, L. F. Villa, and M. Alvarez del Buergo, Sex differences in the acquisition of total bone mineral mass peak assessed through dual-energy x-ray absorptiometry. Calcif. Tissue Int. 51, 251 – 254 (1992). 113. A. Gotfredsen, A. Hadberg, L. Nilas, and C. Christiansen, Total body bone mineral in healthy adults. J. Lab. Clin. Med. 110, 362 – 368 (1987). 114. C. M. Weaver, M. Peacock, and C. C. Johnston, Adolescent nutrition in the prevention of postmenopausal osteoporosis. J. Clin. Endocrinol. Metab. 84, 1839 – 1843 (1999). 115. G. Jones and T. Dwyer, Bone mass in prepubertal children: Gender differences and the role of physical activity and sunlight exposure. J. Clin. Endorinol. Metab. 83, 4274 – 4279 (1998). 116. M. J. Valimaki, M. Karkkainen, C. Lambergallardt, K. Laitinen, E. Alhava, J. Heikkinen, O. Impivaara, P. Makela, J. Palmgren, R. Seppanen, I. Vuori, H. K. Akerblom, M. Uhari, J. Viikari, M. Dahl, E. Pesonen, M. Pietikainen, M. Salo, A. Aromaa, L. Kannas, L. Keltikangasjarvinen, V. Kuusela, K. Pyorala, T. Ronnemaa, L. Rasanen, S. Sama, A. Seppanen, and R. Telama, Exercise, smoking, and calcium intake during adolescence and early adulthood as determinants of peak bone mass. Br. Med. J. 309, 230 – 235 (1994). 117. A. M. Fehily, R. J. Coles, W. D. Evans, and P. C. Elwood, Factors affecting bone density in young adults. Am. J. Clin. Nutr. 56, 579 – 586 (1992). 118. D. C. Welten, H. C. G. Kemper, G. B. Post, W. Van Mechelen, J. Twisk, P. Lips, and G. J. Teule, Weight-bearing activity during youth is a more important factor for peak bone mass than calcium intake. J. Bone Miner. Res. 9, 1089 – 1096 (1994). 119. C. C. Johnston, J. Z. Miller, C. W. Slemenda, T. K. Reister, S. Hui, J. C. Christian, and M. Peacock, Calcium supplementation and increases in bone mineral density in children. N. Engl. J. Me