Calcium and Bone Disorders in Children and Adolescents - Endocrine Development Vol 16

Calcium and Bone Disorders in Children and Adolescents

Endocrine Development Vol. 16

Series Editor

P.-E. Mullis

Bern

Calcium and Bone Disorders in Children and Adolescents Volume Editors

Jeremy Allgrove London Nick Shaw Birmingham 89 figures, 13 in colour, and 25 tables, 2009

Basel · Freiburg · Paris · London · New York · Bangalore · Bangkok · Shanghai · Singapore · Tokyo · Sydney

Endocrine Development Founded 1999 by Martin O. Savage, London

Jeremy Allgrove

Nick Shaw

Department of Paediatric Endocrinology Royal London Hospital London, UK

Clinical Lead for Endocrinology Birmingham Children's Hospital Birmingham, UK

Library of Congress Cataloging-in-Publication Data Calcium and bone disorders in children and adolescents / volume editors, Jeremy Allgrove, Nick Shaw. p. ; cm. -- (Endocrine development ; v. 16) Includes bibliographical references and indexes. ISBN 978-3-8055-9161-4 (hard cover : alk. paper) 1. Bone diseases in children--Case studies. 2. Calcium metabolism disorders in children--Case studies. I. Allgrove, Jeremy. II. Shaw, Nick. III. Series: Endocrine development ; v. 16. [DNLM: 1. Calcium Metabolism Disorders--physiopathology. 2. Adolescent. 3. Bone Diseases--physiopathology. 4. Child. W1 EN3635 v.16 2009 / WD 200.5.C2 C1433 2009] RJ482.B65C35 2009 618.92⬘71--dc22 2009018333

Bibliographic Indices. This publication is listed in bibliographic services, including Current Contents® and Index Medicus Disclaimer. The statements, opinions and data contained in this publication are solely those of the individual authors and contributors and not of the publisher and the editor(s). The appearance of advertisements in the book is not a warranty, endorsement, or approval of the products or services advertised or of their effectiveness, quality or safety. The publisher and the editor(s) disclaim responsibility for any injury to persons or property resulting from any ideas, methods, instructions or products referred to in the content or advertisements. Drug Dosage. The authors and the publisher have exerted every effort to ensure that drug selection and dosage set forth in this text are in accord with current recommendations and practice at the time of publication. However, in view of ongoing research, changes in government regulations, and the constant flow of information relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for any change in indications and dosage and for added warnings and precautions. This is particularly important when the recommended agent is a new and/or infrequently employed drug. All rights reserved. No part of this publication may be translated into other languages, reproduced or utilized in any form or by any means electronic or mechanical, including photocopying, recording, microcopying, or by any information storage and retrieval system, without permission in writing from the publisher. © Copyright 2009 by S. Karger AG, P.O. Box, CH–4009 Basel (Switzerland) www.karger.com Printed in Switzerland on acid-free and non-aging paper (ISO 9706) by Reinhardt Druck, Basel ISSN 1421–7082 ISBN 978–3–8055–9161–4

Contents

VII Foreword Glorieux, F.G. (Montréal, Qué.) IX Preface Allgrove, J. (London); Shaw, N. (Birmingham) 1 8 32 49 58 73 93 115 133 157 170

Voyages of Discovery Allgrove, J. (London) Physiology of Calcium, Phosphate and Magnesium Allgrove, J. (London) Physiology of Bone Grabowski, P. (Sheffield) Bone Biopsy: Indications and Methods Rauch, F. (Montréal, Qué.) Bone Densitometry: Current Status and Future Perspectives Crabtree, N. (Birmingham); Ward, K. (Cambridge) A Practical Approach to Hypocalcaemia in Children Shaw, N. (Birmingham) A Practical Approach to Problems of Hypercalcaemia Davies, J.H. (Southampton) A Practical Approach to Rickets Allgrove, J. (London) Disorders of Phosphate Homeostasis and Tissue Mineralisation Bergwitz, C.; Jüppner, H. (Boston, Mass.) Primary Osteoporosis Bishop, N. (Sheffield) Secondary Osteoporosis Ahmed, S.F.; Elmantaser, M. (Glasgow)

V

191 218 233 246

281

292 293

VI

Miscellaneous Bone Disorders Mughal, M.Z. (Manchester) Drugs Used in Paediatric Bone and Calcium Disorders Cheung, M. (London) Skeletal Aspects of Non-Accidental Injury Johnson, K. (Birmingham) Case Histories Howard, S. (London); Lyder, G. (Birmingham); Allgrove, J. (London); Shaw, N. (Birmingham) Appendix Author Index Subject Index

Contents

Foreword

The understanding of the biology and pathophysiology of the human skeleton has progressed at a remarkably fast pace in the past 35 years. Investigation and management of bone diseases were only chapters in textbooks on endocrinology or nephrology. Not anymore. Based on innovative diagnostic technologies, the development of new therapeutic modes, and the progress in medical genetics, the field of bone disease has established itself as a stand-alone specialty. As a consequence, a large number of books, thick and thin, have been and continue to be published on various aspects of skeletal diseases. In some, there are chapters on paediatric bone diseases, but it is only of late that entire books devoted to the study of the growing skeleton and its abnormalities have emerged, based on the impressive growth of knowledge in all aspects (cellular, organic, hormonal, structural) of bone metabolism and the diseases that interfere with the development, growth, remodelling and mineralisation of bone. The precursor opus was published in 1974 by Maroteaux [1]. It is a most extraordinary collection of radiographs, but with a timid approach at basic mechanisms. In 1980, there was also the book honouring Helen and Harold Harrison’s contribution to the understanding of paediatric bone diseases. It focused on biochemistry and clinical studies [2]. Then in 2003, an elaborate treatise was published as an attempt to organise the sum of current knowledge on paediatric bone diseases [3]. The editors of the present work have pursued the same goal in a concise and practical manner. They have recruited experts on various aspects of the biology of bone and the diseases affecting its structure and function. The field is well covered, but the true originality of the book lies within its last section where a series of case histories is presented. No doubt, the readers will appreciate this real-life approach to the problems discussed and, hopefully, connect them with their own experience. Francis H. Glorieux, OC, MD, PhD Shriners Hospital for Children, Montréal, Canada

VII

References 1 2 3

VIII

Maroteaux P: Maladies osseuses de l’ enfant. Paris, Flammarion, 1974. DeLuca HF, Anast CS (eds): Pediatric Diseases Related to Calcium. New York, Elsevier, 1980. Glorieux FH, Pettifor JM, Juppner H (eds): Pediatric Bone: Biology and Diseases. San Diego, Academic Press, 2003.

Glorieux

Preface

The idea for this book arose as a result of a number of initiatives. Several paediatricians in the UK with a clinical and research interest in calcium and bone disorders in children started to establish dedicated Metabolic Bone Disease clinics about 10 or more years ago. These clinics have attracted a wide variety of different conditions. Many of these are unusual but not really of sufficient rarity to warrant separate case reports. Others have already been published. There is, however, no publication that specifically brings together case histories related to bone and calcium disorders placed in the context of the background physiology and pathology, so, consequently, the idea arose of providing one. As a consequence of the rarity of some of the conditions we were seeing, a group of us started to meet twice a year to discuss challenging cases and share knowledge and expertise. Out of these initial meetings developed the British Paediatric and Adolescent Bone Group (BPABG), which is now affiliated to the Royal College of Paediatrics and Child Health as a speciality group with our own scientific session at the annual conference. The aim of this group is to promote knowledge and understanding of paediatric bone disease. As an additional development to fulfil this objective it has, for the past 3 years, run an annual postgraduate teaching course for senior trainees and consultants who are interested in learning more about the subject. An important aspect of these courses is that the delegates have been asked to bring case histories for presentation and discussion. Many of the cases described in the final chapter have been presented at these courses. Many of the chapters in this book are based on the lectures given at them and all of the founder members of BPABG have contributed. Additional contributions have come from other members of the BPABG together with some international authors.

IX

We are grateful to Karger for inviting us to edit this book, which is published as part of their Endocrine Development series under the overall editorship of Professor Primus Mullis. We are grateful to all of the authors for sending us their manuscripts in a timely manner. We wish to thank those clinicians whose cases are presented and who have allowed us to report them. Finally, we wish to thank our wives, Natalie and Vicki, for their long-suffering approach to our slaving over hot computers when more sociable activities beckoned. Jeremy Allgrove Nick Shaw

X

Allgrove · Shaw

Chapter 1 Allgrove J, Shaw NJ (eds): Calcium and Bone Disorders in Children and Adolescents. Endocr Dev. Basel, Karger, 2009, vol 16, pp 1–7

Voyages of Discovery Jeremy Allgrove Royal London Hospital, and Great Ormond Street Hospital, London, UK

Abstract The metabolism of calcium and bone is controlled by five principal hormones: parathyroid hormone, 1,25-dihydroxyvitamin D, calcitonin, parathyroid hormone-related peptide and fibroblast growth factor 23, some of which have been known for several decades and some of which have only recently been identified. The stories of discovery of these hormones have constituted a series of complex journeys which have been undertaken over the past century or so and none of which has yet been completed. The complexities of bone and calcium metabolism have been and remain, to many people, somewhat mysterious and a daunting task to understand. This book is designed to try to unravel those mysteries and present them in an interesting and comprehensible manner. Copyright © 2009 S. Karger AG, Basel

The study of the diseases that affect bone and calcium has made huge strides over the past few decades. The initial realisation that rickets, which was rife in industrial cities, particularly in the UK, could be cured by exposing children to sunlight or supplementing them with foods such as cod liver oil was a major step in improving the health of children in the early part of the twentieth century. Subsequently, the discovery of other hormones that are involved in mineral metabolism, both calcium and phosphate, has enabled much wider understanding of the mechanisms of disease to be gained. This has led to the introduction of logical treatments based on this scientific understanding. There are five major hormones, vitamin D and its metabolites, parathyroid hormone (PTH), calcitonin (CT), parathyroid hormone-related peptide (PTHrP) and fibroblast growth factor 23 (FGF23), that are involved directly in the control of mineral metabolism in man. In addition, several other hormones, such as oestrogens and androgens, cortisol, growth hormone and thyroxine, have modifying effects. The story of the unravelling of these hormones is a long and complicated one that has gradually revealed itself over the past century or so. For each there has been a long voyage of discovery, some lasting longer than others, but each is still a journey in progress.

Vitamin D and Rickets

Rickets is an ancient disease. It was probably known in the ancient world, but is recorded in the UK since the 17th Century [1]. It became widespread with the increase in industrialisation during the 19th and 20th centuries. The first breakthrough in treatment came with the realisation, shortly after the end of the First World War, that most rickets could be cured either by exposure to sunlight or with supplements of cod liver oil [2]. Vitamin D was discovered to be the agent that effected the cure. As a consequence of this, rickets virtually disappeared in the UK until the first major wave of immigration, mainly from the old commonwealth countries. Most of this immigration was from either south Asia or the Caribbean and brought with it a greater predisposition to rickets than was present within the white population because of the need for greater sunlight exposure of more darkly pigmented skin in order to synthesise sufficient vitamin D [3]. This resulted in a second wave of rickets that again occurred mainly in the industrialised cities. Following a pilot study, it was demonstrated that the incidence of rickets in Glasgow could be effectively reduced by a campaign of supplementation [4]. Since then, the incentives to persist in such a campaign appear to have been lost and a third wave resurgence of rickets has been seen in many countries of the world [5]. In the meantime, during the 1960s it was discovered that vitamin D required metabolism in order to become effective and elevated it from the status of ‘vitamin’ to one of ‘hormone’. As a consequence, some forms of rickets that had previously been thought to be caused by vitamin D deficiency were now understood to result from inborn errors of metabolism and explained why some children had not previously responded to vitamin D treatment. Since then the metabolism of vitamin D has been well worked out and provides a logical basis for treatment. The third stage of investigation of vitamin D has been the demonstration that vitamin D deficiency may play an important part in contributing to the aetiology of a number of common diseases that previously had not been associated with vitamin D deficiency. These include certain cancers, especially of the breast and colon, diabetes, both type 1 and type 2, and coronary heart disease. These relationships remain to be worked out although increasing evidence is accumulating that vitamin D plays a part in the prevention of many of these diseases [6]. Whilst they are generally diseases of adulthood, it is conceivable that their origins lie in childhood. Vitamin D remains the preferred treatment of vitamin D deficiency but it seems extraordinary that, in modern societies, the ability to eliminate a fully preventable disease eludes us. It is arguable that effective vitamin D supplementation is the single most cost-effective treatment that could be given, at least to ‘at risk’ populations.

2

Allgrove

Parathyroid Hormone

During the 1920s the role of the parathyroid glands in secreting a calcium-raising hormone was clarified [7]. The first description of hypoparathyroidism was made in 1929 [8] and this was followed in 1942 by the description of parathyroid hormone resistance in pseudohypoparathyroidism [9]. Fuller Albright was an early pioneer of parathyroid physiology and pathology who laid the basis of much of what we know about basic parathyroid actions. He was correct in describing pseudohypoparathyroidism as a hormone resistance syndrome. However, as it turns out, he was incorrect in referring to it as ‘an example of Seabright bantam syndrome’. The ‘Seabright bantam’, named after Sir John Sebright (sic), was misspelt in the original paper. It is characterised by the cock birds having a ‘hen-feathering’ appearance which led to the misapprehension that they were resistant to testosterone. In fact, they have excessive activity of aromatase P450 in extragonadal tissues that converts testosterone to oestrogen [10]. It is this that causes the characteristic feathering pattern. In most other aspects, he was correct and he made a huge contribution to our understanding of bone disease and several conditions bear his name eponymously. The first immunoassays for PTH were described in 1969 [11]. These had been developed in the wake of other immunoassays such as those for insulin and growth hormone. However, it rapidly became apparent that these were not straightforward since PTH, a large molecule containing 84 amino acids, circulates as a number of fragments [12]. These are particularly problematic in the presence of renal failure when the inactive fragments tend to circulate in higher quantities than normal. Since the original assays were developed, further refinements have been made that now allow measurement of physiological levels of intact hormone. The structure of PTH was difficult to establish and different structures were proposed initially. Once these were resolved, it became apparent that, although PTH contains 84 amino acids, only the first 34 are required for full biological activity [13]. The function of the remainder of the molecule remains unclear. Subsequent work revealed the mechanism of action of PTH via the Gsα second messenger which is common to a number of polypeptide hormones and which provides an explanation for the hormone resistance state known as pseudohypoparathyroidism that was originally described by Fuller Albright.

Calcitonin

This hormone was first described in 1963 [14] and its structure elucidated in 1968 [15]. It is a 32 amino acid protein and, unlike PTH, has a disulphide bond between the cystine residues at positions one and seven. It has an action that is largely opposite to that of parathyroid hormone, i.e. it has a calcium-lowering effect. There are considerable interspecies differences in structure [16] and, interestingly, the salmon

Voyages of Discovery

3

hormone has considerably greater activity than its human counterpart in humans. For this reason, it has been used as a therapeutic agent to lower calcium in certain hypercalcaemic conditions although its use in this respect has been largely superseded by the introduction of the bisphosphonates. CT is now known to be one product of the α-CT/CT gene-related peptide (CALCA) which, as a result of alternative splicing, gives rise to at least two products, α-CT and CT gene-related peptide (CGRP) [17]. Each is produced mainly by different tissues, CT by the C cells of the thyroid and CGRP by the hypothalamus. The physiological role of these proteins, together with those of two other closely related proteins, amylin and adrenomedullin, have yet to be identified precisely, but CT probably makes a contribution to bone formation and CGRP is mainly a neuropeptide which plays a part in vascular tone. It may have a role in the pathogenesis of migraine. Nevertheless, it is also thought that all four proteins may play some part in bone formation [18, 19] possibly via a network of neurones that exists in bone. However, pathological states in man in which CT is produced in excess, such as medullary carcinoma of the thyroid (MCT), do not result in hypocalcaemia and the principal significance of CT is as a marker of MCT and as a therapeutic agent.

Parathyroid Hormone-Related Peptide

The first indication that there is a substance that has PTH-like activity but is not PTH came with a publication in 1985 demonstrating that human umbilical cord blood contained a compound which had PTH-like bioactivity and yet could not be identified as PTH on immunoassay [20]. The calcium concentration in foetal cord blood is unusual in being one of the few substances which is present at higher levels than in the mother, i.e. there is a positive gradient across the placenta. Whilst it had previously been suggested that foetal PTH levels are suppressed because of these relatively high levels of calcium, the question had never been asked as to what maintains the gradient. It seems that it is PTHrP that is responsible. Subsequently, a humoral factor was identified and purified from malignant tissue that was found to be responsible for some instances of humoral hypercalcaemia of malignancy [21]. This was a PTH-like factor that shared some properties with PTH, including its ability to stimulate cyclic-AMP, but was sufficiently dissimilar as to be undetectable on standard PTH immunoassays. It is a considerably larger molecule than PTH itself and has some limited homology with PTH such that it binds to the PTH1 receptor. The role of PTHrP in man seems to be principally in the foetus to maintain the calcium gradient across the placenta and to have a paracrine function in promoting cartilage development. In postnatal life, it seems not to have a classical endocrine role but is important as a mediator of humoral hypercalcaemia of malignancy [22].

4

Allgrove

Fibroblast Growth Factor 23

The factors controlling phosphate metabolism have, until relatively recently, not been well understood. The discovery of FGF23 in 2000 [23] led to an explosion of discoveries relating to phosphate. The relationship of this to PHEX, DMP1, GALNT3, FGFR1, Klotho and the sodium/phosphate co-transporter in renal tubular cells has widened our understanding considerably and led to a much greater knowledge of the pathological processes that go towards explaining the mechanisms of disorders of phosphate metabolism. Further details of all these hormones are given in the relevant chapters. The discovery of DNA in the early 1950s and its role as a genetic blueprint has allowed the identification of a whole host of diseases that are genetically based. The diseases related to bone and calcium are no exception and, if one excludes vitamin D deficiency and secondary osteoporosis, it is the case that virtually all other causes of bone and calcium diseases have a genetic origin. Indeed, they encompass the full gamut of genetic conditions including autosomal and X-linked dominant and recessive, mitochondrial and imprinting disorders. It is therefore necessary to have at least a modicum of understanding of genetics in order to be able fully to understand their mechanisms. Fortunately, modern technology allows the rapid advances in genetics to be recorded electronically without having to ‘go back to the books’ all the time. It also ensures that updates to discoveries can be made available to a wide audience more rapidly than previously. The most useful tool is the creation of the Online Mendelian Inheritance in Man (OMIM) website which was the brainchild of the late Victor McKusick [24] when his original paper version became too unwieldy and difficult to update. The website is accessible at http://www.ncbi.nlm.nih.gov/sites/entrez?db = omim and gives details of all disorders that are or are thought to be genetically based, together with the genes involved. Because of the diversity of the disorders of bone and calcium metabolism, each of the clinical chapters in this book is accompanied by at least one table that gives the OMIM reference numbers for these disorders and their genes. Each entry is accompanied by a detailed bibliography which is regularly updated. Hopefully, this will prove useful to readers. An explanation of any abbreviations appearing in the text and which are not defined at the time can be found in the Appendix together with relevant conversion factors for readers who are not familiar with either SI or ‘conventional’ units of measurement. The final chapter in this book is a series of case histories. These are intended to illustrate some of the problems that are discussed in the previous chapters. It is not a comprehensive coverage of all the conditions mentioned but, since this book will be available on line, it will be possible in the future to add further cases. When the text describes a case that is included in the case history section, the number of that case is shown in the text. References to these cases may appear in more than one chapter.

Voyages of Discovery

5

‘Dr. Donne’s verses are like the peace of God; they pass all understanding’. With these words, King James I of England and VI of Scotland is said to have replied to Archdeacon Plume when asked to comment on the poetry of John Donne [25]. There are many, even in the world of paediatric endocrinology, for whom the same is true of the study of bone and calcium disorders in children and for whom they remain a mystery. This was recognised by the late Graham Chapman, of Monty Python fame, comedian, bon viveur and erstwhile medical student, who, in a book of collected sketches, letters and essays, wrote a brief essay entitled ‘Calcium Made Interesting’ [26]. This book is designed not only to enable mineral metabolism to be understood, but to ‘make calcium interesting’.

References 1 O’Riordan JL: Rickets in the 17th century. J Bone Miner Res 2006;21:1506–1510. 2 DeLuca HF: The vitamin D story: a collaborative effort of basic science and clinical medicine. FASEB J 1988;2:224–236. 3 Parra EJ: Human pigmentation variation: evolution, genetic basis, and implications for public health. Am J Phys Anthropol 2007;(suppl 45):85–105. 4 Dunnigan MG, Glekin BM, Henderson JB, McIntosh WB, Sumner D, Sutherland GR: Prevention of rickets in Asian children: assessment of the Glasgow campaign. Br Med J (Clin Res Ed) 1985;291:239– 242. 5 Chesney RW: Rickets: the third wave. Clin Pediatr (Phila) 2002;41:137–139. 6 Bouillon R, Eelen G, Verlinden L, Mathieu C, Carmeliet G, Verstuyf A: Vitamin D and cancer. J Steroid Biochem Mol Biol 2006;102:156–162. 7 Collip JB: The internal secretion of the parathyroid glands. Proc Natl Acad Sci USA 1925;11:484–485. 8 Albright F, Ellsworth R: Studies on the physiology of the parathyroid glands. I. Calcium and phosphorus studies on a case of idiopathic hypoparathyroidism. J Clin Invest 1929;7:183–201. 9 Albright FBC, Smith PH, Parson W: Pseudohypoparathyroidism: an example of ‘Seabright-Bantam syndrome’. Endocrinology 1942;30:922–932. 10 Matsumine H, Wilson JD, McPhaul MJ: Sebright and Campine chickens express aromatase P-450 messenger RNA inappropriately in extraglandular tissues and in skin fibroblasts. Mol Endocrinol 1990; 4:905–911. 11 Berson SA, Yalow RS, Aurbach GD, Potts JT: Immunoassay of Bovine and Human Parathyroid Hormone. Proc Natl Acad Sci USA 1963;49:613– 617.

6

12 Berson SA, Yalow RS: Immunochemical heterogeneity of parathyroid hormone in plasma. J Clin Endocrinol Metab 1968;28:1037–1047. 13 Rosenblatt M, Segre GV, Tregear GW, Shepard GL, Tyler GA, Potts JT Jr: Human parathyroid hormone: synthesis and chemical, biological, and immunological evaluation of the carboxyl-terminal region. Endocrinology 1978;103:978–984. 14 Copp DH: Calcitonin: a new hormone from the parathyroid which lowers blood calcium. Oral Surg Oral Med Oral Pathol 1963;16:872–877. 15 Neher R, Riniker B, Rittel W, Zuber H: Human calcitonin: structure of calcitonin M and D. Helv Chim Acta 1968;51:1900–1905. 16 Niall HD, Keutmann HT, Copp DH, Potts JT Jr: Amino acid sequence of salmon ultimobranchial calcitonin. Proc Natl Acad Sci USA 1969;64:771– 778. 17 Rosenfeld MG, Lin CR, Amara SG, Stolarsky L, Roos BA, Ong ES, et al: Calcitonin mRNA polymorphism: peptide switching associated with alternative RNA splicing events. Proc Natl Acad Sci USA 1982; 79:1717–1721. 18 Huebner AK, Keller J, Catala-Lehnen P, et al: The role of calcitonin and alpha-calcitonin gene-related peptide in bone formation. Arch Biochem Biophys 2008;473:210–217. 19 Naot D, Cornish J: The role of peptides and receptors of the calcitonin family in the regulation of bone metabolism. Bone 2008;43:813–818. 20 Allgrove J, Adami S, Manning RM, O’Riordan JL: Cytochemical bioassay of parathyroid hormone in maternal and cord blood. Arch Dis Child 1985;60: 110–115.

Allgrove

21 Burtis WJ, Wu T, Bunch C, et al: Identification of a novel 17,000-dalton parathyroid hormone-like adenylate cyclase-stimulating protein from a tumor associated with humoral hypercalcemia of malignancy. J Biol Chem 1987;262:7151–7156. 22 Kaiser SM, Goltzman D: Parathyroid hormone-related peptide. Clin Invest Med 1993;16:395–406. 23 Yamashita T, Yoshioka M, Itoh N: Identification of a novel fibroblast growth factor, FGF-23, preferentially expressed in the ventrolateral thalamic nucleus of the brain. Biochem Biophys Res Commun 2000; 277:494–498.

24 Obituary: Professor Victor McKusick: advocate of the Human Genome Project. The Times 2008 Aug 1. 25 Oxford Dictionary of Quotations, ed 2. London, Oxford University Press, 1953. 26 Chapman G: Calcium Made Interesting; in Yoakum J (ed): Calcium Made Interesting: Sketches, Letters Essays and Gondolas. London, Sidgwick & Jackson, 2005, pp 88–89.

Jeremy Allgrove, MD Department of Paediatric Endocrinology, David Hughes Building, First Floor Royal London Hospital, Whitechapel London E1 1BB (UK) Tel. +44 20 7377 7468, Fax +44 20 7943 1353, E-Mail [email protected]

Voyages of Discovery

7

Chapter 2 Allgrove J, Shaw NJ (eds): Calcium and Bone Disorders in Children and Adolescents. Endocr Dev. Basel, Karger, 2009, vol 16, pp 8–31

Physiology of Calcium, Phosphate and Magnesium Jeremy Allgrove Royal London Hospital, and Great Ormond Street Hospital, London, UK

Abstract The physiology of calcium and the other minerals involved in its metabolism is complex and intimately tied in with the physiology of bone. Five principal humoral factors are involved in maintaining plasma levels of calcium, magnesium and phosphate and coordinating the balance between these and their content in bone. The transmembrane transport of these elements is dependent on a series of complex mechanisms that are controlled by these hormones. The plasma concentration of calcium is initially sensed by a calcium-sensing receptor which then sets up a cascade of events that initially determines parathyroid hormone secretion and eventually results in a specific action within the target organs, mainly bone and kidney. This chapter describes the physiology of these humoral factors and relates them to the pathological processes that give rise to disorders of calcium and bone metabolism. It details the stages in the calcium cascade and describes the effects on the various target organs. The pathology of disorders of bone and calcium metabolism is described in detail Copyright © 2009 S. Karger AG, Basel in the relevant chapters.

Calcium, phosphate and magnesium metabolism is intimately bound and it is necessary to discuss all three together. Furthermore, this metabolism is, in many ways, different from that of most other substances by virtue of the fact that the majority of each is contained within bone which acts as a structural material, as well as a reservoir, whilst also acting as an important physiological regulator. Thus, calcium is required to be kept within narrow limits within plasma in order to maintain optimum neuromuscular function and phosphate is involved in virtually all metabolic processes whilst magnesium is required to ensure optimum parathyroid hormone secretion. The mechanisms required to maintain these levels are complex and dependent on a number of factors. It is the purpose of this chapter to describe these factors and to indicate how disorders of function give rise to clinical problems.

Calcium Physiology

A fully grown adult contains approximately 1,200 g of calcium. In foetal and neonatal life, the total calcium content is related to body weight and a very close relationship exists between the two under normal circumstances. This relationship is expressed by the formula: Ca = 0.00075*BWt1.3093,

where Ca and BWt are both expressed in grams [1]. The relationship has been observed during the foetal and neonatal period and probably largely holds true throughout the period during which bone accretion is occurring. About 99% of calcium is normally contained within bone, the remainder being present either as an intracellular cation or circulating in plasma. Within plasma, there are three main fractions: ionised, protein bound and complexed, mainly to citrate or sulphate. The ionised fraction constitutes approximately 50% of the total. Most blood gas machines found within critical or intensive care units can measure ionised calcium directly. Of the remainder, most circulates bound to albumin and plasma albumin levels affect the total concentration of calcium. Various formulae are used to ‘correct’ total calcium to allow for this and many laboratories automatically provide a value for ‘corrected’ calcium (see chapter 6). The concentration of ionised calcium is normally kept within very narrow limits (1.1–1.3 mmol/l), a level which is necessary to maintain normal neuromuscular activity. Complex mechanisms are involved in maintaining this level. These involve altering calcium absorption in the gut, changing excretion within the renal tubules and balancing the rate of deposition into or removal from bone. If calcium levels vary significantly from this, either upwards or downwards, symptoms may develop. These are discussed in more detail in the relevant chapters.

Control of Plasma Calcium

Five principal humoral factors are involved in the maintenance of normal levels of calcium and phosphate in plasma. Plasma calcium is mainly influenced by parathyroid hormone (PTH) and the active form of vitamin D, 1α,25-dihydroxyvitamin D (1,25(OH)2D). In addition, calcitonin (CT) and parathyroid hormone-related peptide (PTHrP) play a more minor role, at least in postnatal life, but attain greater significance in a number of pathological situations. Plasma phosphate is also influenced by PTH and 1,25(OH)2D, but another factor, fibroblast growth factor 23 (FGF23) also plays an important part in its metabolism. Magnesium is influenced, though to a lesser degree, by the same factors that control calcium but, in addition, itself influences calcium indirectly by altering the PTH secretion in response to hypocalcaemia.

Calcium Physiology

9

Calcium TRPV5

TRPV6 Claudin 16

Mainly GI

Mainly renal

Na+

Paracellular

CB28k CB9k

Ca2+ PMCA1b

NCX1 Ca2+

Fig. 1. Schematic representation of the mechanisms of calcium transport in the gut and renal tubules. Similar mechanisms are present in both tissues although the importance of each differs between them. The principal mechanisms in the gut are shown on the right hand side and the more important ones in the renal tubules are shown on the left. Abbreviations are explained in the Appendix.

Transmembrane Calcium Transport

Calcium balance is controlled principally by transport across membranes in the gastrointestinal tract and in renal tubules. The mechanisms for both are similar but with differences of emphasis depending on which organ is affected. Transport occurs by both paracellular and transcellular mechanisms. They are summarised in figure 1. Paracellular transport occurs through the tight junctions between cells and is facilitated by a number of proteins including, amongst others, the claudins. The most important of these is claudin 16 (*603959) (also known as paracellin 1). It is coded for by a gene on chromosome 3q27 and has its action mainly in renal tubules where it also facilitates passive transport of magnesium. Mutations in this gene cause the hypomagnesaemia, hypercalciuria and nephrocalcinosis (HOMG3) syndrome (#248250) in which both calcium and magnesium are poorly reabsorbed (see chapter 6). The most important mechanism for calcium absorption in the gut is via active transport. Three steps are involved in this process [2]. There is initial absorption of calcium from the lumen followed by transcellular transport and lastly extrusion of

10

Allgrove

calcium across the basolateral membrane. A similar process involving related proteins is present in renal tubules (see below). Two recently discovered proteins, TRPV5 (*606679) and TRPV6 (*606680), which are members of the transient receptor potential (TRP) channel protein family, are thought to play an important role in promoting active calcium transport [2, 3]. In the gut the most important of these is TRPV6 whilst TRPV5 plays a larger role in renal tubules. They are inwardly rectifying calcium channels whose affinity is greater for calcium than for magnesium. Once calcium reaches the intracellular compartment, cytosolic diffusion across the cell is facilitated by two further proteins, calbindin28K (*114050) and calbindin9K (*302020). These bind calcium and transport it across the cytoplasm. At the basolateral surface, extrusion of calcium is facilitated by both an ATP-dependent Ca+-transporting ATPase (PMCA1b) (*108731) and a Na+/ Ca+ exchanger (NCX1) (*182305). The former of these is more important in the gut. There are vitamin D receptors in the small intestinal cells and both this and TRPV6 are stimulated by it. If these receptors are defective, as in hereditary 1α,25(OH)2Dresistant rickets (HVDRR) (#277440), calcium cannot be absorbed properly and rickets results (see chapter 8 for further details). Calcium absorption is also influenced by a number of other factors. In particular, absorption can be reduced in the presence of large quantities of calcium-binding agents such as phytate or oxalate [4]. Bisphosphonates also bind to calcium in the gut and, if used orally for therapeutic purposes, should be taken as far away from meals as possible. Alternatively, a high calcium intake helps to protect from the effects of vitamin D deficiency [5], presumably by increasing passive absorption. Calcium reabsorption in renal tubules is largely passive. It is influenced by a number of dietary factors including a high sodium, protein or acid load, all of which increase calcium excretion. About 70% of filtered load is reabsorbed passively in the proximal tubule in conjunction with sodium. A further 20% is reabsorbed in the loop of Henle by paracellular processes under the influence of claudin 16 (paracellin 1). Paracellular reabsorption does not occur unless claudin 16 is present. The remaining 5–10% is reabsorbed in the distal tubule. Similar mechanisms to those in the gut are present although TRPV5 is thought to be the major influence. Transcellular transport is facilitated by calbindins. At the basolateral surface NCX1 is the more important mechanism. This is under hormonal influence, mainly PTH, and, in the presence of hypoparathyroidism, treatment with active vitamin D analogues must be monitored carefully to prevent hypercalciuria.

Magnesium Metabolism

Magnesium is, like calcium, a divalent cation that is important for bone and calcium metabolism. It is normally present in plasma at a concentration of between 0.7 and 1.2 mmol/l. Adequate plasma magnesium is required for normal secretion of PTH

Calcium Physiology

11

Magnesium TRPM6

TRPM7

Claudin 16 Claudin 19

Paracellular

+

Na+ Pro-EGF ? EGFR

? ␣ ␤ Y

Mg2+ EGF

Na-K ATPase Mg2+

Fig. 2. Schematic representation of the mechanisms of magnesium transport in the gut and renal tubules. Similar mechanisms are present in both tissues although the importance of each differs between them. Abbreviations are explained in the Appendix.

which occurs in response to a magnesium-dependent adenylate cyclase system. In the presence of hypomagnesaemia, PTH secretion is inadequate and correction of hypocalcaemia cannot occur normally [6]. Magnesium absorption occurs in the small intestine by mechanisms that are very similar to those of calcium although these mechanisms are not so well understood [3]. They are summarised in figure 2. Two proteins, TRPM6 and TRPM7, which are related to the corresponding proteins involved in calcium absorption, are present in renal tubules and intestinal cells. Mutations in TRPM6 cause hypomagnesaemia with secondary hypocalciuria (HOMG1) (#602014) as a result of impaired magnesium absorption in the gut (see chapter 6). Transcellular transport of magnesium probably involves similar calbindin proteins to those of calcium. Renal tubular reabsorption of magnesium mostly occurs by passive reabsorption in the ascending loop of Henle along with calcium. The tight junction protein claudin 16 (Paracellin 1) (*603959) is principally responsible for this and mutations in its gene cause the HOMG3 syndrome (#248250). Because calcium reabsorption is also impaired, this is accompanied by hypercalciuria and nephrocalcinosis (see chapter 6).

12

Allgrove

Further along the renal tubule, active reabsorption takes place in the DCT where TRPM6 is situated. It is under the influence of epidermal growth factor (EGF) which is present on the basolateral membrane of the renal cells where it is processed from Pro-EGF [7]. Following cleavage from Pro-EGF, it interacts with its receptor, the EGFR (*131550), which, amongst its other actions, stimulates magnesium absorption via TRPM6 on the luminal surface. It has recently been shown that mutations in the EGF gene disrupt the basolateral sorting of Pro-EGF resulting in understimulation of TRPM6 and impaired magnesium reabsorption [8]. The resulting condition is known as isolated recessive renal hypomagnesaemia (IRH) (see chapter 6). Transcellular transport of magnesium is probably effected by proteins similar to the calbindins involved in calcium transport but these mechanisms are not well understood. At the basolateral membrane, magnesium is transported partly by a mechanism that involves Na+/K+ ATPase. This has three subunits, α, β and γ, the latter of which is coded for by the FXYD2 (*601814) gene. Mutations in this gene result in defective magnesium reabsorption in the autosomal-dominant renal hypomagnesaemia associated with hypocalciuria (HOMG2) (#154020) syndrome (see chapter 6). The thiazide-sensitive sodium chloride co-transporter (NCC) is also involved in magnesium transport and mutations in the coding gene, SLC12A3 (*600968), cause Gitelman’s syndrome (#263800) in which hypermagnesuria is a feature. Raised urinary magnesium excretion is also present in some cases of Bartter’s syndrome which is caused by a variety of mutations affecting chloride and sodium reabsorption in the loop of Henle. In the last part of the renal tubules, the collecting ducts, both calcium and magnesium are again reabsorbed passively via the tight junction protein, claudin 19 (*610036). Mutations in this gene cause renal hypomagnesaemia with ocular involvement (#248190) (see chapter 6). Renal tubular transport of magnesium can also be increased by several non-genetic causes including diuretics, diabetic ketoacidosis, gentamicin, mercury-containing laxatives, transplanted kidney, urinary tract obstruction, the diuretic phase of acute renal failure and cisplatin.

Phosphate Metabolism

A fully grown adult contains approximately 700 g phosphate. As with calcium, the total body content of phosphate is closely related to body weight and is expressed by the formula: PO4 = 0.00037*BWt1.2409.

Approximately 80% of phosphate is contained in bone. Of the remainder, 45% (9% of the total) is present in skeletal muscle, 54.5% in the viscera and only 0.5% in extracellular fluid. Most of the phosphate is present in inorganic form but plays a crucial

Calcium Physiology

13

part in many intracellular processes. In plasma, phosphate circulates in the form of phospholipids, phosphate esters, and free inorganic phosphate (Pi). Plasma Pi concentrations are not as tightly controlled as those of calcium and reflect the fluxes of phosphate entering and leaving the extracellular pool. In contrast to calcium, phosphate concentrations in plasma vary considerably during life, being highest during phases of rapid growth. Thus, phosphate concentrations in premature infants are normally above 2.0 mmol/l (6.4 mg/dl), falling to 1.3–2.0 mmol/l (4.2–6.4 mg/dl) during infancy and childhood and to 0.7–1.3 mmol/l (2.2–4.3 mg/dl) in young adults. Phosphate transport across membranes is controlled by a series of sodium-dependent active transport mechanisms (Na/Pi co-transporters). Three classes are known to exist. Type 1 is present in renal tubular brush borders but is not thought to have a major role in renal tubular reabsorption of phosphate. Type 2, which has three subtypes, 2a, 2b and 2c, are probably the most important in regulating phosphate absorption and reabsorption. Type 3 is present in many tissues but is thought to have more of a ‘gatekeeping’ role. Phosphate is readily absorbed throughout the small bowel by both passive and active mechanisms. Approximately 70% is absorbed via type 2b Na/Pi co-transporter, the remainder being by passive absorption. This active transport is stimulated directly by 1,25(OH)2D and therefore indirectly by hypocalcaemia and PTH [9]. Since hypophosphataemia is a powerful stimulant of 25-hydroxyvitamin D-1-alphahydroxylase (1α-hydroxylase), phosphate deficiency itself stimulates increased absorption. However, the total amount absorbed is dependent on the dietary phosphate load and may be inhibited by phosphate-binding agents such as calcium acetate (Phosex®), calcium carbonate (Tetralac®) or sevelamer (Renagel®). These are of value in hyperphosphataemic states such as chronic renal failure when phosphate absorption needs to be limited. The metabolism of phosphate has, until recently, been relatively poorly understood. However, in 2000, a new member of the fibroblast growth factor family, FGF23 was discovered [10]. This was subsequently shown to be mutated in cases of autosomaldominant hypophosphataemic rickets (ADHR) (#193100) [11] and it is now thought to play a key role in phosphate metabolism. FGF23 is derived from bone cells, particularly osteocytes, circulates in plasma, and is subject to a variety of feedback mechanisms. As a result, it is now considered to be a classic hormone. Its synthesis and secretion are modified by several factors, especially PHEX and DMP1 (see below), and it undergoes post-translational modification, which results in either inactivation (cleavage) or activation (O-glycosylation) before becoming active. Once activated, its principal target organ is the renal tubule where it stimulates renal phosphate excretion and inhibits 1α-hydroxylase activity to reduce levels of 1,25(OH)2D. Hypophosphataemia also inhibits FGF23 secretion. These actions are summarised in figure 3. FGF23 (*605380) is a 251 amino acid protein which includes a 24 amino acid signal sequence. It is coded for by a gene on chromosome 12p13.3. It has a crucial cleavage site between residues arginine179 and serine180 where it is cleaved by a subtilisin/ furin-like enzyme that renders it inactive [12]. A second arginine residue is present at

14

Allgrove

DMP1

BMP1 etc

Gs␣

GALNT3 Endopeptidase

FGFR1c

FGF23

Klotho

Renal tubule Subtilisin/furin Na/Pi PHEX

N-terminal

C-terminal

f1␣-hydroxylase Degradation products

DPhosphaturia DPhosphate absorption

fPlasma phosphate

f1,25(OH)2D

Fig. 3. Schematic representation of the control of phosphate metabolism. Fibroblast growth factor 23 (FGF23) sits at the centre. Its secretion is influenced by several other factors and it has to undergo modification before becoming active. Its receptor on the renal tubule enables it to promote phosphate excretion. Solid lines represent stimulatory effects and interrupted lines inhibitory actions. Abbreviations are explained in the Appendix.

position 176 and mutations in either of these arginine residues renders FGF23 resistant to cleavage without affecting its intrinsic activity. As a consequence, circulating FGF23 levels remain high and result in the excessive renal phosphate loss in ADHR (#193100) (see chapter 9 for further details). Activation of FGF23 occurs by O-glycosylation. This occurs under the influence of UDP-nacetyl-alpha-d-galactosamine:polypeptide N-acetylgalactosaminyltransferase 3 (GALNT3) (*601756). This is a 633 amino-acid protein that is coded for by a gene on chromosome 2q24-q31 [13]. The gene has ten exons and GALNT3 itself has a single transmembrane spanning region. It catalyses the O-glycosylation of serine and threonine residues on the native protein. Its crucial role in phosphate metabolism is demonstrated by the fact that inactivating mutations in this gene result in hyperphosphataemic familial tumoral calcinosis type 1 (HFTC1) (#211900) [14] in which, although FGF23 levels are high, these are inactive and hyperphosphataemia and soft tissue calcinosis occurs (see chapter 9). The principal target organ of FGF23 is the renal tubule. Once it has been activated, it acts on a receptor on the surface of the tubules. This receptor is part of the fibroblast growth factor receptor family, FGFR1(IIIc) (*136350), which is coded for by

Calcium Physiology

15

a gene on chromosome 8p11.2-p11.1. The FGF receptors, of which four classes are described, are involved in a wide variety of functions, and mutations in FGFR1 give rise to several conditions including Kallmann syndrome, hypogonadotrophic hypogonadism and a number of skeletal abnormalities [15]. FGFR1 is therefore not specific for FGF23. However, in 2006 another factor, α-Klotho (KL) (+604824) was found to act as a cofactor which confers specificity of FGFR1(IIIc) for FGF23 [16]. Klotho (named after the Greek Fate who spins the thread of life) is coded for by a gene on chromosome 13q12. Patients with chronic renal failure (CRF) have low renal expression of KL and it has been suggested that this may accelerate the degenerative processes, such as atherosclerosis, osteoporosis and skin ageing, seen in CRF [17]. KL is not capable of acting as an FGF23 receptor on its own, but requires FGFR1. Similarly, FGFR1 is not active as an FGF23 receptor if KL is inhibited or mutated [16]. Rare patients have been described who have inactivating mutations of KL which cause a form of tumoral calcinosis and activating mutations have been associated with hypophosphataemic rickets that is associated with hyperparathyroidism (see chapter 9). Once the FGFR1-KL complex has been activated by FGF23, it increases renal tubular phosphate excretion by means of the NaPi-IIc/SLC34A3 (*609826) exchanger at the luminal surface of the cells. Mutations in the NaPi-IIc exchanger result in hereditary hypophosphataemic rickets with hypercalciuria (HHRH) (#241530). However, unlike those conditions which are associated with high FGF23 levels, low levels of active FGF23 are present. 1α-Hydroxylase activity is therefore not inhibited. The resulting raised levels of 1,25(OH)2D not only stimulate calcium and phosphate absorption in the gut but also increase calcium excretion in renal tubules, as a consequence of which hypercalciuria and renal stones are present. The secretion and initial processing of FGF23 is under the influence of several other factors. The most important of these is the phosphate-regulating gene with homologies to endopeptidases on the X chromosome (PHEX) (*300550). Several studies in the hyp-mouse, an animal model of X-linked dominant hypophosphataemic rickets (XLH), have demonstrated that PHEX is somehow involved in the regulation of FGF23 despite the fact that it is not present in renal tubules. The precise mechanisms by which this occurs are not fully understood but may involve either modification of the activity of the subtilisin/furin enzyme activity that cleaves FGF23, or by modifying dentin matrix protein 1 (DMP1) that also affects FGF23 (see below). Whatever the precise mechanism, mutations in PHEX result in failure of cleavage of FGF23 which therefore causes hyperphosphaturia and hypophosphataemia. DMP1 (*600980) is another protein synthesised by osteocytes. It is one of a number of small integrin-binding ligand, N-linked glycoproteins (SIBLING) that are involved in bone mineralisation. It may act as a mechanostat that responds to changes in stresses within bone transmitted via the fluid filled canaliculi within bone in which the osteocytes lie. It is cleaved into two fragments of 35 and 57 kDa, respectively, possibly by the action of bone morphogenic protein 1 (BMP1). The latter of these is an active inhibitor of FGF23 secretion. Thus, DMP1 has a controlling effect on the

16

Allgrove

action of FGF23. Homozygous or compound heterozygous mutations in DMP1 result in excess FGF23 secretion which results in autosomal-recessive hypophosphataemic rickets (ARHR) (#241520) (see chapter 9). Other SIBLING proteins include bone sialoprotein (BSP) (*166490), osteopontin (OPN) (*166490), dentin sialophosphoprotein (DSPP) (*125485) and matrix extracellular phosphoglycoprotein (MEPE) (*605912). Some of these are upregulated in certain forms of cancer and may be responsible for alterations in FGF23 secretion that causes TIO. Some individuals with the McCune-Albright polyostotic fibrous dysplasia, caused by somatic mutations in the alfa subunit of the stimulatory G-protein (Gsα), have an associated excess phosphate excretion secondary to increased FGF23 by an, as yet, ill understood mechanism. Hypophosphataemia and rickets are also seen in several primary renal tubular abnormalities, such as the Fanconi syndrome (whatever the cause), in which a generalised proximal renal tubular defect, which results in bicarbonaturia, glycosuria and amino aciduria as well as a phosphate leak, is present. The most common inherited cause of Fanconi syndrome is cystinosis (#219800) and rickets may be the presenting feature of this condition. Hyperparathyroidism also causes a mild form of Fanconi syndrome and patients with parathyroid tumours may have a mild metabolic acidosis and aminoaciduria in addition to hypercalcaemia. These features resolve when the tumour is removed.

The Calcium Cascade

The concentration of calcium in plasma is normally maintained within very narrow limits. The initial stage of this process is binding of calcium to a specific calciumsensing receptor. This then initiates a cascade of events that terminates in the action of PTH on its target organs (fig. 4).

The Calcium Sensing Receptor

The calcium-sensing receptor (CaSR) (+601199) is a large molecule consisting of 1078 amino acids. It is coded for by a gene on chromosome 3q13-q21. It has a large extracellular calcium-binding domain consisting of approximately the first 610 residues followed by a seven-transmembrane domain consisting of the next 250 residues with a further 210 residues making up the intracellular component. The receptor is present in many tissues, especially the parathyroid glands and renal tubules, but also in bone and cartilage as well as other tissues [18]. When calcium binds to the extracellular domain, it alters PTH secretion via both phospholipase Cb and G-protein second messengers. As a consequence, PTH secretion changes in a sigmoidal fashion in response to acute changes in plasma calcium (fig. 5), and there is a continuous tonic secretion of PTH, which maintains plasma-ionised calcium at whatever level is ‘set’

Calcium Physiology

17

Ca2+ CaSR

Parathyroid Glands PTH

PTH1R

Gs␣,␤␥

Target Organs – Kidney Bone (Gut)

Fig. 4. The calcium cascade. Plasma calcium levels are controlled by a series of events that begin with the effect of calcium on the calcium sensing receptor and end with the response of the target organs.

by the CaSR [19]. Magnesium also binds to the CaSR and influences PTH secretion in a similar, but less potent, manner to that of calcium. However, severe magnesium deficiency inhibits PTH secretion, probably because the adenylate cyclase coupled to the G-protein is itself magnesium dependent [6]. Mutations within the CaSR gene result in either inactivation or activation of the receptor, which result in hyper- and hypocalcaemia respectively. Inactivating mutations cause insensitivity to calcium, which shifts the curve of PTH secretion in response to plasma calcium to the right (fig. 5). As a consequence, PTH secretion is switched off at a higher concentration than normal, and hypercalcaemia results [18]. The receptors are also present in the renal tubule, and renal calcium excretion is thereby reduced. The resulting condition is known as familial benign hypercalcaemia (FBH) or familial hypocalciuric hypercalcaemia (FHH) (#145980) (see chapter 7). In contrast, activating mutations of the receptor shift the PTH secretion curve to the left (fig. 5) causing chronic hypocalcaemia and hypercalciuria, a condition known as autosomal-dominant hypocalcaemia (ADH) (#146200). One particular mutation causes a constitutive activation of the receptor which remains constantly active whatever the calcium concentration so that, rather than shifting the curve to the left, PTH secretion remains permanently switched off [20] (see chapter 6).

18

Allgrove

80 70 Inactivating mutations 60 50 40

Intact PTH

30 20

Activating mutations

10 Ca2+ 0 0.95

1.00

1.05

1.10

1.15

1.20

1.25

1.30

1.35

1.40

1.45

1.50

Fig. 5. Schematic representation of the relationship between plasma ionised calcium and PTH secretion as determined by the calcium sensing receptor. Inactivating mutations generally shift the curve to the right whilst activating mutations do so to the left.

Many of the mutations found in FBH are clustered around the aspartate- and glutamate-rich regions of the extracellular domain of the molecule, and it has been postulated that this region contains low-affinity binding sites for calcium. Many of the FBH kindreds have been found to have unique mutations. Mutations have also been detected within the transmembrane domain but only rarely within the intracellular domain. Mutations within this latter domain may have a greater effect on the CaSR in the parathyroid glands than in the renal tubules and patients are described in whom inactivating mutations are associated with hypercalciuria and PT gland hyperplasia necessitating parathyroidectomy [21]. Similarly, most activating mutations that cause ADH are present within the extracellular calcium-binding domain. One hundred and twenty-eight mutations, three of which are polymorphisms, have been described so far, and an online database has been established to keep track of them (http://www. casrdb.mcgill.ca). Two other loci, located on chromosomes 19p and 19q13, respectively, have been identified by family linkage studies. The precise nature of the gene products of these loci remains uncertain, but mutations within them result in clinical syndromes that are very similar to those resulting from inactivating mutations of the CaSR itself. Not all families with FBH have mutations within the CaSR gene, and it has been suggested that there may be abnormalities either within the CaSR gene promoter or within one

Calcium Physiology

19

of these two loci found on chromosome 19. The three variants of FBH linked to chromosome 3q, 19p and 19q, have therefore been referred to as FBH types 1–3.

The Parathyroid Glands

The parathyroid (PT) glands, usually four in number but sometimes as many as seven, are derived embryologically from the third (lower glands) and fourth (upper glands) branchial arches. Several transcription factors are involved in their development [22]. Some, such as Hoxa3 (thyroid and thymus, chromosome 7p15-p14.2) (*142954), GATA3 (hearing sensation and kidney, chromosome 10p13–14) (*131320), Tbx1 (thymus, cardiac outflow tract and the face, chromosome 22q11) (*602054) and UDF1L are involved in the development of other structures. The latter two genes are located on the long arm of chromosome 22. Mutations within or deletion of the genes responsible for these factors result in congenital hypoparathyroidism that is associated with other conditions such as the hypoparathyroidism, deafness, renal anomalies (HDR) (#146255) syndrome and the 22q deletion complex, of which the DiGeorge syndrome (DGS) (#188400) is part. The homologue of drosophila glial cells missing 2 (GCM2) (*603716) is a highly conserved gene that is necessary for PT gland development. It has no other known function in man. Mutations in this gene cause autosomal recessive familial isolated hypoparathyroidism (FIH) (#146200). It is also thought that the SRY-related HMGbox gene 3 (SOX3) (*313430), located on the X-chromosome, is involved in PT gland development and mutations in this gene may be responsible for X-linked recessive familial isolated hypoparathyroidism (%307700). Apart from these autosomal- and X-linked syndromes, there are several mitochondrial genes that are involved in PTG development. Mutations in these genes give rise to a variety of syndromes in which hypoparathyroidism is a feature. Because the genes are mitochondrial, these syndromes are maternally inherited (for full details of these conditions, see chapter 6). In addition to these genetic causes, destruction of the glands may occur as a result of surgery (e.g. following thyroidectomy), infiltration (e.g. with iron in β-thalassaemia) or antibodies. These may either be isolated or associated with autoantibodies to other organs as in the polyendocrinopathy type 1 syndrome (APS1), also known as the APECED syndrome (#240300) (see chapter 6 for further details).

Parathyroid Hormone

PTH (*168450) is a single-chain polypeptide hormone containing 84 amino acids. It is encoded by a gene on chromosome 11. It is synthesised by the parathyroid glands from prepro-PTH, which has an additional 31 N-terminal amino acids. Synthesis occurs in

20

Allgrove

the ribosomes, where the initial 25-amino-acid ‘pre’ sequence acts as a signal peptide to aid transport through the rough endoplasmic reticulum [23, 24]. The ‘pre’ sequence is cleaved, and pro-PTH then travels to the Golgi apparatus where the 6-amino-acid ‘pro’ sequence is cleaved to yield the mature hormone, which is stored in secretory vesicles that fuse with the plasma membrane prior to secretion of the hormone [25]. Very little PTH is stored within the glands, and most of the secreted hormone is newly synthesised. Mutations in the PTH gene, involving the pre-pro- sequence have been described that result in inherited hypoparathyroidism (see chapter 6). Only the first 34 N-terminal amino acids are required for full activity, and the function of the remainder of the molecule is not understood. The half-life of PTH in the circulation is 1–2 min [25]. The molecule is cleaved at various sites, which results in a number of fragments that can be identified in the circulation. The best modern assays of PTH measure ‘intact’ PTH, are able to measure physiological concentrations of PTH, correlate well with bioactivity and ignore the inactive fragments. This is particularly important in conditions such as chronic renal failure where inactive fragments are cleared less rapidly than normal. Normal levels of PTH in the circulation are about 1–6 pmol/l (10–60 pg/ml) but vary depending on the assay used.

The PTH Receptors

PTH acts via two receptors. The first and principal one is PTH1R (also called PTH/ PTHrP) receptor (*168468), which has equal affinity for both PTH and PTHrP. It consists of 593 amino acids coded by a gene on the long arm of chromosome 3 [26]. It has an extracellular binding domain of 190 residues, a seven-transmembrane domain, and a cytosolic component of 134 residues. Both inactivating and activating mutations of the PTH1R have been described. These result in the very rare conditions of Blomstrand lethal chondrodysplasia (#215045 ) and Jansen disease (#156400) respectively. A second PTH2 receptor (PTH2R) is present in the central nervous system. PTHrP is not a ligand for it.

Intracellular Signalling

Intracellular signalling occurs principally by coupling of the cytosolic component of the PTH1R to G-protein second messengers, Gs and Gq [27]. These are heterotrimeric, consisting of α, β, and γ subunits. In the resting state, they are associated, and the Gsα subunit is bound to GDP. Binding of the ligand with the receptor results in GDP being exchanged for GTP and dissociation of the Gsα subunit from the β,γ complex. The Gsα is then free to stimulate adenylate cyclase, which results in an increase in intracellular cAMP, which activates the various actions of PTH via specific protein kinases. Intrinsic GTPase activity associated with the Gsα subunit hydrolyses GTP to GDP, which causes

Calcium Physiology

21

reassociation of the components of the G-protein. At the same time, phosphodiesterases inactivate the cAMP to AMP, and the cell reverts to its resting state. This mechanism is common to several hormones, including thyroid-stimulating hormone (TSH), gonadotropins, and growth hormone-releasing hormone (GHRH) [27]. The Gsα subunit is coded by a gene, GNAS1, (+139320) located on chromosome 20q13.3. This complex gene contains 13 exons that code for the Gsα subunit itself plus several other exons, known as A/B, XL, NESPAS (*610540) (which is an antisense transcript) and NESP55 which is only expressed in renal tubules. Alternative promoter use and splicing results in several different mRNA transcripts. In most tissues, these show biallelic expression, but the transcripts arising from the A/B, XL and NESPAS exons are paternally derived whilst those arising from the NESP55 exon are maternally expressed. This results from methylation of these uniparental alleles that either switches on or switches off the activity of those alleles in an epigenetic manner (fig. 6). In addition, there is a further gene, Syntaxin (STX16) (*603666), upstream of the GNAS complex which appears to influence the methylation of the A/B exon. Mutations within the biallelic coding region (exons 2–13) of the gene result in resistance to the action of PTH which clinically causes pseudohypoparathyroidism type 1a (PHP1A) (#103580) if they are associated with the maternally derived transcripts but cause pseudopseudohypoparathyroidism (PPHP) (#612463) and/or progressive osseous heteroplasia (POH) (#166350) if derived from paternal sources [28]. These patients frequently have resistance to other hormones whose action is mediated via the Gsα second-messenger mechanism and many display features of Albright’s hereditary osteodystrophy (AHO). Somatic activating mutations in the GNAS complex are responsible for the McCune-Albright syndrome (MAS) (#174800) (see chapter 12 for more details). Alterations in the methylation patterns of the monallelic exons, particularly A/B, cause pseudohypoparathyroidism type 1b (PHP1B) (#603233) when they are on the maternally derived alleles. Under these circumstances there are no mutations found in the coding regions of the GNAS gene and the patients do not usually have evidence of AHO. Mutations in STX16 have also been associated with some forms of pseudohypoparathyroidism type 1b probably by influencing the methylation of the A/B exon (see chapter 6).

The Target Organs

The principal target organs of PTH are bone and kidney. In bone, PTH has two main effects. Under physiological conditions, it promotes bone formation via receptors on the osteoblasts. Under circumstances of hypocalcaemia, PTH stimulates bone resorption in order to retrieve calcium from the large reservoir within bone so that normocalcaemia can be restored. There are very few receptors for PTH in osteoclasts and bone resorption occurs as a result of changes within the relationship between osteoblasts and osteoclasts. Both RANKL and osteoprotegerin (OPG) are produced by osteoblasts. RANKL stimulates osteoclast differentiation whilst OPG acts as a decoy ligand

22

Allgrove

STX1 M P

Nesp55 – +

Nespas + –

XL + –

A/B + –

1

2

3 3N

4 5

6

7

13

Allele-specific methylation

Bi-allelic (most tissues)

Exons 2–13 Exon1

Gsα

Exons 2–13 A/B Paternal

Exons 2–13 XL Exons 2–13 Nespas

Maternal

Exons 2–13 Nesp55

Fig. 6. Diagrammatic representation of the GNAS gene showing the different products that result from alternative splicing. Native Gsα is expressed biallelically. The A/B, XL and Nespas transcripts are principally expressed in the paternal allele whilst the Nesp55 transcript is mainly expressed in the maternal allele. Since the latter is present in renal tubules and the others only in other tissues, mutations in exons 2–13 result in AHO and, if derived from the maternal allele, are associated with pseudohypoparathyroidism. If the paternal allele is the origin, pseudopseudohypoparathyroidism is the result. Alterations to the methylation pattern of the various alternative splicing products without mutations in exons 2–13 result in pseudohypoparathyroidism type 1b if they are of maternal origin. (Adapted and reprinted from Bastepe M, Jűppner H, Thakker RV: Parathyroid disorders; in Glorieux FH, Pettifor JM, Jűppner H (eds): Pediatric Bone: Biology and Diseases. Academic Press, San Diego, 2003, p 493, with permission from Elsevier.)

for RANKL and inhibits its action. PTH alters the balance between the two in such a way as temporarily to change the balance in favour of bone resorption (see chapter 3 for further details). In the absence of PTH over long periods, such as in unrecognised hypoparathyroidism, bone becomes undermineralised (see chapter 15, case 13).

Calcium Physiology

23

In the nephron, PTH has three main actions. In the convoluted and straight parts of the proximal tubule, it stimulates the conversion of 25OHD to 1,25(OH)2D. In the distal tubules, it promotes the reabsorption of both calcium and magnesium. It also promotes the excretion of phosphate which allows excess phosphate that is resorbed from bone by PTH to be excreted. There is also an effect of PTH on bicarbonate and amino acid reabsorption in the proximal tubule that results in a mild form of Fanconi syndrome in hyperparathyroidism. This resolves when the hyperparathyroidism is reversed.

Parathyroid Hormone-Related Peptide

The presence of a PTH-like substance with similar biological activity but different immunological properties was originally suggested in 1985 [29]. These studies showed that neonatal cord blood contained high PTH-like activity although N-terminal immunoreactivity was absent. The bioactivity was related to the positive gradient of calcium across the placenta and the authors suggested that it was PTHrP that maintained this gradient. It had also been recognised for some time that some patients with malignancy developed hypercalcaemia with undetectable levels of PTH. Subsequently, a protein was purified from lung cancer cells that had similar biological properties to PTH but which was clearly different from PTH itself [30]. This protein was subsequently identified as PTHrP (+168470). PTHrP is a 141 amino acid polypeptide that is coded for by a gene on chromosome 12p12.1-p11.2. It has some homology with PTH in its N-terminal end but diverges from PTH after residue 13. PTHrP cannot normally be measured in the circulation and has no significant classical hormone action in post natal life but does have an important paracrine role in chondrocyte proliferation and maturation. PTHrP has equal activity with PTH on the PTH1R, and some of the changes seen in Jansen’s metaphyseal chondrodysplasia (#156400) are thought to be related to overactivity of these receptors. PTHrP is not a ligand for the PTH2R which is mainly present in brain. However, PTHrP is secreted by the lactating breast and women with hypoparathyroidism who are breast feeding may need to reduce their dose of vitamin D analogues. The principal pathological importance of PTHrP in postnatal life is as a cause of hypercalcaemia of malignancy (see chapter 6).

Calcitonin

CT (*114130) is a 31 amino acid protein that is synthesised by the C cells of the thyroid gland. It is coded for by a gene on chromosome 11p15.2-p15.1 which, by alternative splicing, also results in another protein, calcitonin gene-related peptide (CGRP). CT is mainly active in the thyroid gland whilst CGRP plays more of a role in the hypothalamus.

24

Allgrove

It is secreted in response to hypercalcaemia and acts via a specific receptor that is coded for by a gene on chromosome 7q21.3 (*114131). The principal action of CT is to lower plasma calcium in a manner opposite to that of PTH. It may have a role in promoting skeletal mineralisation in the foetus but has little physiological role in postnatal life. It is sometimes used therapeutically to reduce plasma calcium in symptomatic hypercalcaemia, although bisphosphonates are now used more frequently for this purpose, but its principal value is as a marker of malignancy in familial medullary carcinoma of the thyroid (MTC) (#155240).

Alkaline Phosphatase

This enzyme is present in several tissues and exists in three main isoforms, intestinal (IAP) (*171740), placental (PLAP) (*171810), and liver (tissue non-specific) (LALP, TNSAP) (*171760). A gene on chromosome 2q34–37 codes for the first two, and a gene on chromosome 1p36.1-p34 codes for the last [31]. Different post-translational modifications of TNSAP enzyme result in three tissue-specific forms found in bone, liver, and kidney that can be distinguished by their different isoelectric points and heat lability, the bone-specific form (bTNAP) being the least stable. It has been suggested that there are three codominant alleles (HN, HC and HI) of this enzyme and that the presence or absence of hypophosphatasia and its severity depends on which alleles are present. The HN allele is by far the commonest and is homozygous in most individuals. The HI allele results in the most serious reduction in activity whilst the HC allele is intermediate. Homozygous HI alleles result in the perinatal lethal or infantile forms of hypophosphatasia (#241500), whilst heterozygous HN/HC or HN/HI cause the adult form (#146300). The intermediate childhood form (#241510) results from HC/HC or HC/HI combination [32]. For a clinical description of these conditions, see chapter 12. bTNSAP is secreted by osteoblasts and promotes bone mineralisation. Circulating TNSAP is largely derived from liver and bone. Levels in plasma during childhood reflect growth rate [33] and are also raised in the presence of rickets (see chapter 8), in juvenile Paget’s disease (#239000) and in fibrous dysplasia (see chapter 12). Low levels are seen in hypophosphatasia, which results from mutations in the TNSAP gene. A database that keeps track of these mutations (currently 194) has been established and can be accessed at http://www.sesep.uvsq.fr/Database.html.

Vitamin D Metabolism

Although referred to as a vitamin, vitamin D is mainly available not from dietary sources but as a result of the action of sunlight on 7-dehydrocholesterol. Ultraviolet light of wavelength 270–300 nm breaks the B-ring of the steroid molecule creating a

Calcium Physiology

25

secosteroid. Further rearrangement of the molecule occurs by the action of body heat to create cholecalciferol (vitamin D3). Vitamin D is also available from plant sources as ergocalciferol (vitamin D2), which is synthesised from ergosterol and which differs structurally from cholecalciferol only in the presence of an additional double bond in the side chain. Both compounds are then metabolised in a similar manner and are thought to be equipotent. Collectively, they are referred to as vitamin D or calciferol. Under normal circumstances, approximately 80% of vitamin D requirements are obtained from this action of sunlight but synthesis is dependent on the amount of sunlight exposure, the strength of the UV light in that sunlight and skin colour. Cultural practices that necessitate substantial covering of the skin limit sunlight exposure. In addition, in temperate climates there is insufficient UV light available in sunlight during winter months even if skin exposure is possible. Melanin absorbs UV light of the appropriate wavelength and, since the melanophores that determine skin colour are situated in the skin above the keratinocytes that synthesise vitamin D, darker skinned individuals require a greater degree of sunlight exposure to achieve the same effect as light-skinned people [34]. There may be as much as a sixfold difference in requirement to overcome this barrier. If this is achieved, darker-skinned individuals are equally capable of synthesising vitamin D. There is generally little vitamin D in food although some oily fish have a relatively high content and it is a common misconception that, because a child is taking a ‘healthy diet’, they are not at risk of vitamin D deficiency. If sunlight exposure is halved, vitamin D intake must be trebled to compensate for this and the only realistic way of achieving this is by giving adequate dietary supplementation. Vitamin D is stored in liver and adipose tissue. Obese subjects have lower circulating levels of vitamin D than non-obese subjects, possibly because they sequester more vitamin D into their fat stores [35]. Following synthesis, vitamin D is bound to a specific vitamin D binding protein (DBP) and passes to the liver. Native vitamin D has little biological activity and requires metabolism via two hydroxylation steps, firstly at the 25- and subsequently at the 1- position in order to become fully active [36]. All of the steps in vitamin D metabolism are catalyzed by cytochrome P450 enzymes (fig. 7). The first step is catalysed by vitamin D 25-hydroxylases. There are at least four different enzymes that have an influence on 25-hydroxylase activity. They are distinguishable by their different affinities and capacities and by their intracellular localisation. The first to be cloned, a low-affinity, high-capacity enzyme (CYP27A1) (*606530) is located in mitochondria. However, there are no reports of rickets resulting from mutations in this gene, but they do cause cerebrotendinous xanthomatosis (#213700). A second high-affinity, low-capacity enzyme (CYP2R1) (*608713), which is probably of greater physiological significance, is located within hepatic microsomes. It contains 501 amino acids and is coded for by a gene on chromosome 11p15.2 [37]. Rare cases are described of rickets associated with mutations in this gene (#600081) [38]. Two other enzymes, CYP3A4 (*124010) and CYP2J2 (*601258) probably also have some effect on 25-hydroxylase but are mainly involved in drug metabolism.

26

Allgrove

7-Dehydrocholesterol

Diet

Sunlight PrevitaminD Body heat

1␣-hydroxycholecalciferol (alfacalcidol)

Cholecalciferol/Ergocalciferol (Vitamin D) Vitamin D 25-hydroxylase

25-OH vitamin D 25-hydroxyvitamin D 1␣-hydroxylase

Vitamin D 25-hydroxylase

1,25(OH)2 vitamin D Vitamin D receptor

Vitamin D 24-hydroxylase

Peripheral action

24,25-Dihydroxyvitamin D 1,24,25-Trihydroxyvitamin D

Fig. 7. Diagrammatic representation of vitamin D metabolism.

The resulting product, 25-hydroxyvitamin D (25OHD), circulates in plasma bound to the DBP in nanomolar concentrations. Assay of 25OHD gives a measure of vitamin D status. Its level varies depending on the supply of vitamin D and shows a considerable annual variation with a peak about 6 weeks after maximal exposure to sunlight. It is now generally agreed that vitamin D sufficiency is defined by a plasma concentration above 50 nmol/l [39]. It has some weak activity, which is not normally of clinical significance, but may become so in the presence of vitamin D excess. Vitamin D 25-hydroxylase also catalyses the conversion of the synthetic vitamin D analogues, 1α-hydroxy-cholecalciferol (alfacalcidol) and 1α-hydroxy-ergocalciferol (doxercalciferol), to 1,25(OH)2D3 and 1,25(OH)2D2, respectively. 25OHD is metabolised to its active hormone 1,25(OH)2D by 25-hydroxyvitamin D 1α-hydroxylase, which is active only against metabolites that are already hydroxylated at position 25 [40]. A single enzyme has been identified located in convoluted and straight portions of the proximal renal tubule. Activity is also present in osteoblasts, keratinocytes, and lymphohaematopoietic cells, where 1,25(OH)2D may have an autocrine or paracrine role. During foetal life, 1α-hydroxylase activity is found in the placenta. In pathological states, it is present in the macrophages of sarcoid tissue and subcutaneous fat necrosis (see chapter 7). It is a mitochondrial enzyme

Calcium Physiology

27

(CYP27B1) (*609506) consisting of 508 amino acids with considerable homology to other P450 enzymes. It is encoded by a single gene on chromosome 12q13.1-q13.3. Mutations in this gene are responsible for the condition known variously as pseudovitamin D deficiency rickets (PDDR), vitamin D-dependent rickets type I (VDDR-I), Prader rickets or 1α-hydroxylase deficiency (#264700). Activity of 1α-hydroxylase is stimulated by PTH via its cAMP/protein kinase actions. Hypocalcaemia stimulates 1α-hydroxylase activity, but this effect is not a direct effect but mediated via PTH. Plasma phosphate has a direct effect on 1α-hydroxylase activity, although there is some evidence to suggest that this may be modulated by growth hormone (GH); calcitonin may also regulate the enzyme. Its activity is inhibited by FGF23. 1,25(OH)2D is a highly potent compound that circulates in picomolar concentrations. However, measurement of 1,25(OH)2D in plasma gives no measure of vitamin D status. Its synthesis is tightly controlled by the plasma calcium concentration. In order to enable changes in 1,25(OH)2D to occur rapidly, a second enzyme, 25-hydroxyvitamin D 24-hydroxylase (25OHD 24-OHase) (*126065), exists. This is yet another cytochrome P450 enzyme that can use both 25OHD and 1,25(OH)2D as substrates to form 24,25-dihydroxyvitamin D (24,25(OH)2D) and 1α,24,25-trihydroxyvitamin D (1,24,25(OH)3D) respectively. The role of this enzyme is probably to divert metabolism of 25OHD away from 1,25(OH)2D synthesis when this is not needed and to participate in the degradation of existing 1,25(OH)2D. It is inhibited by PTH and stimulated by 1,25(OH)2D and FGF23. 1,24,25(OH)3D has limited potency (about 10% of 1,25(OH)2D) and is probably an intermediate degradation metabolite of 1,25(OH)2D. The role, if any, of 24,25(OH)2D is uncertain. Some authors have argued that it has no role to play whereas others have suggested that it may influence bone mineralisation. In addition, people of South Asian origin possess higher 25OHD 24-OHase activity than those of European origin [41] and this seems to contribute to their susceptibility to vitamin D deficiency rickets. 1,25(OH)2D acts via a specific vitamin D receptor [42] (*601769). It is a member of the steroid-thyroid-retinoid superfamily of nuclear receptors and, in many respects, is typical of this group with ligand binding, DNA binding, dimerisation, and transcriptional activation domains. It is encoded by a gene on chromosome 12 near the 1α-hydroxylase gene. The receptors are widely distributed in gut, parathyroid glands, chondrocytes, osteoblasts, and osteoclast precursors. 1,25(OH)2D plays a critical role in promoting calcium absorption in the small intestine, suppresses PTH secretion from the parathyroids, influences growth plate mineralisation, and stimulates differentiation of osteoclasts. In addition, there are receptors present in many tissues that are not directly related to calcium homeostasis such as skin, breast, prostate, colon, etc., and it has been postulated that 1,25(OH)2D may play a part in preventing cancers of these tissues [43]. Mutations in the vitamin D receptor occur throughout the molecule but particularly in either the ligand-binding (ligand-binding-negative) or the DNA-binding

28

Allgrove

(ligand-binding-positive) domains [42]. These mutations cause severe rickets, and many individuals, especially those with defects in DNA binding, also have alopecia. Originally referred to as vitamin D-dependent rickets type II (VDRR-II), it is now more properly called hereditary 1,25(OH)2D-resistant rickets (HVDRR) (#277440). In another form of HVDRR, no mutations of the receptor have been identified, but is thought to be caused by overexpression of a nuclear ribonucleoprotein that binds with the hormone receptor complex to attenuate its action (%600785) (see chapter 8 for details).

Conclusions

The mechanisms that are involved in maintaining normal calcium, magnesium and phosphate are complex and involve several different hormonal mechanisms that influence both calcium, magnesium and phosphate in an independent but linked manner. Normal calcium and phosphate physiology demands that these mechanisms all function satisfactorily in order to maintain good bone health and a suitable milieu in which muscle and nerve function can be optimised. Disruptions to these mechanisms may be either environmental, principally due to vitamin D deficiency, or, in many instances, genetic. A thorough understanding of the physiology is required before a correct diagnosis can be made.

References 1 Allgrove J: Practical management of disorders of calcium metabolism; in Aynsley-Green A (ed): Paediatric Endocrinology in Clinical Practice. Lancaster, MTP Press, 1984, pp 241–263. 2 Hoenderop JG, Nilius B, Bindels RJ: Calcium absorption across epithelia. Physiol Rev 2005;85: 373–422. 3 Hoenderop JG, Bindels RJ: Epithelial Ca2+ and Mg2+ channels in health and disease. J Am Soc Nephrol 2005;16:15–26. 4 Heaney RP: Nutrition and risk of osteoporosis; in Marcus R, Feldman D, Kelsey J (eds): Osteoporosis. San Diego, Academic Press, 2001. 5 Khadilkar A, Das G, Sayyad M, et al: Low calcium intake and hypovitaminosis D in adolescent girls. Arch Dis Child 2007;92:1045. 6 Allgrove J, Adami S, Fraher L, Reuben A, O’Riordan JL: Hypomagnesaemia: studies of parathyroid hormone secretion and function. Clin Endocrinol (Oxf) 1984;21:435–449. 7 Muallem S, Moe OW: When EGF is offside, magnesium is wasted. J Clin Invest 2007;117:2086–2089.

Calcium Physiology

8 Groenestege WM, Thebault S, van der WJ, et al: Impaired basolateral sorting of pro-EGF causes isolated recessive renal hypomagnesemia. J Clin Invest 2007;117:2260–2267. 9 Wilz DR, Gray RW, Dominguez JH, Lemann J Jr: Plasma 1,25-(OH)2-vitamin D concentrations and net intestinal calcium, phosphate, and magnesium absorption in humans. Am J Clin Nutr 1979;32: 2052–2060. 10 Yamashita T, Yoshioka M, Itoh N: Identification of a novel fibroblast growth factor, FGF-23, preferentially expressed in the ventrolateral thalamic nucleus of the brain. Biochem Biophys Res Commun 2000; 277:494–498. 11 ADHR Consortium: Autosomal dominant hypophosphataemic rickets is associated with mutations in FGF23. Nat Genet 2000;26:345–348. 12 Fukumoto S: Post-translational modification of fibroblast growth factor 23. Ther Apher Dial 2005;9: 319–322.

29

13 Bennett EP, Hassan H, Clausen H: cDNA cloning and expression of a novel human UDP-N-acetylalpha-d-galactosamine: polypeptide N-acetylgalactosaminyltransferase, GalNAc-t3. J Biol Chem 1996; 271:17006–17012. 14 Topaz O, Indelman M, Chefetz I, et al: A deleterious mutation in SAMD9 causes normophosphatemic familial tumoral calcinosis. Am J Hum Genet 2006; 79:759–764. 15 Passos-Bueno MR, Wilcox WR, Jabs EW, Sertie AL, Alonso LG, Kitoh H: Clinical spectrum of fibroblast growth factor receptor mutations. Hum Mutat 1999; 14:115–125. 16 Urakawa I, Yamazaki Y, Shimada T, et al: Klotho converts canonical FGF receptor into a specific receptor for FGF23. Nature 2006;444:770–774. 17 Koh N, Fujimori T, Nishiguchi S, et al: Severely reduced production of Klotho in human chronic renal failure kidney. Biochem Biophys Res Commun 2001;280:1015–1020. 18 Brown EM, MacLeod RJ: Extracellular calcium sensing and extracellular calcium signalling. Physiol Rev 2001;81:239–297. 19 Conlin PR, Fajtova VT, Mortensen RM, LeBoff MS, Brown EM: Hysteresis in the relationship between serum ionized calcium and intact parathyroid hormone during recovery from induced hyper- and hypocalcemia in normal humans. J Clin Endocrinol Metab 1989;69:593–599. 20 Zhao XM, Hauache O, Goldsmith PK, Collins R, Spiegel AM: A missense mutation in the seventh transmembrane domain constitutively activates the human Ca2+ receptor. FEBS Lett 1999;448:180–184. 21 Carling T, Szabo E, Bai M, et al: Familial hypercalcemia and hypercalciuria caused by a novel mutation in the cytoplasmic tail of the calcium receptor. J Clin Endocrinol Metab 2000;85:2042–2047. 22 Parfitt AM: Parathyroid growth: Normal and abnormal; in Bilezikian JP, Marcus R, Levine MA (eds): The Parathyroids: Basic and Clinical Concepts. San Diego, Academic Press, 2001. 23 Habener JF, Potts JT Jr: Biosynthesis of parathyroid hormone (second of two parts). N Engl J Med 1978; 299:635–644. 24 Habener JF, Potts JT Jr: Biosynthesis of parathyroid hormone (first of two parts). N Engl J Med 1978; 299:580–585. 25 Kronenberg HM, Bringhurst FR, Segre GV, Potts JT: Parathyroid hormone biosynthesis and metabolism; in Bilezikian JP, Marcus R, Levine MA (eds): The Parathyroids: Basic and Clinical Concepts. San Diego, Academic Press, 2001.

30

26 Nissensen RA: Receptors for parathyroid hormone and parathyroid hormone-related protein: signaling and regulation; in Bilezikian JP, Marcus R, Levine MA (eds): The Parathyroids: Basic and Clinical Concepts. San Diego, Academic Press, 2001. 27 Farfel Z, Bourne HR, Iiri T: The expanding spectrum of G protein diseases. N Engl J Med 1999; 340:1012–1020. 28 Bastepe M: The GNAS locus and pseudohypoparathyroidism. Adv Exp Med Biol 2008;626:27–40. 29 Allgrove J, Adami S, Manning RM, O’Riordan JL: Cytochemical bioassay of parathyroid hormone in maternal and cord blood. Arch Dis Child 1985;60: 110–115. 30 Moseley JM, Kubota M, Diefenbach-Jagger H, et al: Parathyroid hormone-related protein purified from a human lung cancer cell line. Proc Natl Acad Sci USA 1987;84:5048–5052 31 Smith M, Weiss MJ, Griffin CA, et al: Regional assignment of the gene for human liver/bone/kidney alkaline phosphatase to chromosome 1p36.1-p34. Genomics 1988;2:139–143. 32 Igbokwe EC: Inheritance of hypophosphatasia. Med Hypotheses 1985;18:1–5. 33 Round JM: Changes in plasma urate, creatinine, alkaline phosphatase and the 24 hours excretion of hydroxyproline during sexual maturation in adolescents. Ann Hum Biol 1980;7:83–88. 34 Lo CW, Paris PW, Holick MF: Indian and Pakistani immigrants have the same capacity as Caucasians to produce vitamin D in response to ultraviolet irradiation. Am J Clin Nutr 1986;44:683–685. 35 Wortsman J, Matsuoka LY, Chen TC, Lu Z, Holick MF: Decreased bioavailability of vitamin D in obesity. Am J Clin Nutr 2000;72:690–693. 36 Okuda K, Usui E, Ohyama Y: Recent progress in enzymology and molecular biology of enzymes involved in vitamin D metabolism. J Lipid Res 1995; 36:1641–1652. 37 Cheng JB, Motola DL, Mangelsdorf DJ, Russell DW: De-orphanization of cytochrome P450 2R1: a microsomal vitamin D 25-hydroxilase. J Biol Chem 2003;278:38084–38093. 38 Cheng JB, Levine MA, Bell NH, Mangelsdorf DJ, Russell DW: Genetic evidence that the human CYP2R1 enzyme is a key vitamin D 25-hydroxylase. Proc Natl Acad Sci USA 2004;101:7711–7715. 39 Holick MF: Vitamin D: a D-lightful health perspective. Nutr Rev 2008;66:S182–S194. 40 St Arnaud R, Messerlian S, Moir JM, Omdahl JL, Glorieux FH: The 25-hydroxyvitamin D 1-alphahydroxylase gene maps to the pseudovitamin D-deficiency rickets (PDDR) disease locus. J Bone Miner Res 1997;12:1552–1559.

Allgrove

41 Awumey EM, Mitra DA, Hollis BW, Kumar R, Bell NH: Vitamin D metabolism is altered in Asian Indians in the southern United States: a clinical research center study. J Clin Endocrinol Metab 1998;83:169–173.

42 Haussler MR, Haussler CA, Jurutka PW, et al: The vitamin D hormone and its nuclear receptor: molecular actions and disease states. J Endocrinol 1997; 154(Suppl):S57–S73. 43 Holick MF: Vitamin D and sunlight: strategies for cancer prevention and other health benefits. Clin J Am Soc Nephrol 2008;3:1548–1554.

Jeremy Allgrove, MD Department of Paediatric Endocrinology, David Hughes Building, First Floor Royal London Hospital, Whitechapel London E1 1BB (UK) Tel. +44 20 7377 7468, Fax +44 20 7943 1353, E-Mail [email protected]

Calcium Physiology

31

Chapter 3 Allgrove J, Shaw NJ (eds): Calcium and Bone Disorders in Children and Adolescents. Endocr Dev. Basel, Karger, 2009, vol 16, pp 32–48

Physiology of Bone Peter Grabowski Faculty of Medicine, Dentistry and Health, Academic Unit of Child Health, University of Sheffield, Sheffield, UK

Abstract Bone serves three main physiological functions. Its mechanical nature provides support for locomotion and offers protection to vulnerable internal organs, it forms a reservoir for storage of calcium and phosphate in the body, and it provides an environment for bone marrow and for the development of haematopoietic cells. The traditional view of a passive tissue responding to hormonal and dietary influences has changed over the past half century to one of a dynamic adaptive tissue responding to mechanical demands. This chapter gathers together some recent advances in bone physiology and molecular cell biology and discusses the potential application of the bone’s functional adaptation to loading in enhancing bone strength during childhood and adolescence. Copyright © 2009 S. Karger AG, Basel

Bone is a dynamic mineralised connective tissue with multiple physiological functions. At the organ level, bones provide mechanical support for load bearing and locomotion, offer physical protection to vulnerable internal organs, form a mobilisable reservoir of calcium and phosphate ions, and provide an environmental niche for bone marrow and the development of haematopoietic cells. At the tissue level, the coordinated activities of bone formation and resorption provide mechanisms for bone modelling – i.e. the formation of new bone during growth and development – and remodelling – i.e. the coordinated process by which old bone is firstly removed then replaced during skeletal maintenance, and for responding rapidly to the body’s immediate calcium, phosphate and acid-base homeostatic requirements. At the cellular level, bone matrix formation and mineralisation are mediated by osteoblasts and bone resorption is mediated by osteoclasts, while at the molecular level, a range of systemic and local factors regulate cellular and tissue level processes in bone. Bones are highly dependent upon other organs for their growth and development, in particular the intestine and kidney, through which mineral and nutritional factors are absorbed, reabsorbed and excreted, as well as the hypothalamus, pituitary, gonads, parathyroid glands, liver and skin that produce hormonal factors regulating bone

growth and mineral homeostasis. Whereas in adults bone physiology is concerned with skeletal maintenance, in children the context is one of bones that are growing in size, mass and mineral density, while at the same time being modelled into their final adult shape and form. Much of what we know of basic skeletal biology and physiology derives from the study of adults or animals, but the study of children’s bone is an active and expanding area of research. Macroscopically, bone tissue is classified as either cortical or trabecular. Cortical bone is found most commonly in the shafts of long bones and consists of a dense compact tissue penetrated by blood vessels and canaliculi which surround osteocytes and their connecting cellular processes. Trabecular or cancellous bone is found at the ends of long bones, in vertebrae and near joint surfaces and consists of a network of thin plates and connecting struts surrounded by bone marrow. Cortical and trabecular bone are very similar in their cellular and molecular composition but differ significantly in their function and mechanical properties. For much of the 20th century, bone physiology largely centred on understanding the hormonal regulation of osteoblasts and osteoclasts in skeletal maintenance and to a lesser extent in bone growth. Since the mid-1960s bone physiology has seen a change of focus largely due to the efforts of Frost, Jee and others (reviewed in [1–3]), giving a vision of bone as a dynamic tissue that responds at the tissue level to the mechanical demands placed upon it, developing the concept of the mechanostat [2], and leading to an increased interest in the role of cells within the bone matrix and marrow as sensors of local mechanical stimuli and regulators of local bone turnover [4]. Advances in tissue and cell culture techniques have contributed to our understanding of the development, regulation and function of osteoblasts and osteoclasts and, more recently, the development of molecular genetics and the ability to generate targeted transgenic mice have revolutionised the study of individual gene product functions in bone.

Skeletal Development

The evolutionary landmarks giving rise to the vertebrate skeleton are reflected in the developmental biology of bone tissue. Axial skeletal patterning, segmentation, growth and condensation are regulated by homeobox (hox) genes (reviewed in [5]), bone morphogenic proteins (BMPs) and other members of the transforming growth factor-β (TGF-β) superfamily (reviewed in [6]), fibroblast growth factors (FGFs) [7], hedgehog [8] and Wnt proteins [9]. Many of these factors act not only in skeletal patterning and development but also in the recruitment and differentiation of osteoblasts during bone modelling and remodelling throughout life. The axial skeleton is laid down initially as a cartilaginous matrix model by chondrocytic cells of mesenchymal origin (fig. 1). The chondrocytes mineralise the matrix and, through

Bone Physiology

33

GP

GP a

b

c

Fig. 1. Early stages of long bone development by endochondral ossification. a A condensation of mesenchymal cells leads to the formation of a cartilage model of the bone. b Chondrocytes beneath the perichondrium (black) differentiate, become hypertrophic and eventually undergo apoptosis, resulting in cartilage mineralisation (shaded) and release of metalloproteinases. c Angiogenic factors released by chondrocytes encourage vascular invasion, recruiting osteoclast and osteoblast precursors that differentiate and convert calcified cartilage into true bone. Growth plates (GP) establish from chondrocytes in the epiphyseal regions.

a process of hypertrophy, eventually die by apoptosis. Metalloproteinases released by the chondrocytes dissolve some of the matrix and generate angiogenic signals promoting vascularisation and the influx of osteoclasts which begin to resorb the mineralised cartilage. Along with the osteoclasts, osteoblast precursors enter the primitive bone and begin to form true bone behind the advancing osteoclasts, giving rise to the primary spongiosa under the growth plates. In the growth plate, gradients of Indian hedgehog (IHH) and parathyroid hormone-related peptide (PTHrP) regulate the directional proliferation and differentiation of chondrocytes to a hypertrophic phase characterised by mineralisation and metalloproteinase secretion, leading to longitudinal bone growth. In contrast, bones of the cranial vault form through a process of intramembranous ossification in membranes of mesenchymal condensations which progress to bone formation without chondrocyte involvement. Bone formed through both mechanisms undergoes remodelling, a process initiated by activation of osteoclasts, followed by a period of resorption, a reversal phase in which osteoclasts die and osteoblasts are activated, a period of matrix formation followed by a mineralisation phase and a return to the resting state. Communication between osteoblasts and osteoclasts coordinates this series of events, known as the bone remodelling cycle, within a basic multicellular unit (BMU). For a more detailed description of skeletal development see Karaplis [10]. Hormonal influences on bone development are described elsewhere in this book.

34

Grabowski

Mesenchymal stem cell

Osteochondroprogenitor

Osteoblasts Osteocytes Lining cells

Sox9

Runx2

Osterix

Fig. 2. Osteoblast differentiation. Osteoblasts arise from a multipotent precursor cell of mesenchymal origin (mesenchymal stem cell). An osteochondrogenitor cell capable of forming both chondrocytes and osteoblasts arises under the control of the transcription factor Sox9. Runx2 is the key regulator of osteoblast differentiation and is constitutively expressed in osteoblasts at all stages of differentiation. A second transcription factor, Osterix, acts downstream of Runx2 in osteoblast differentiation. Osteoblasts can further differentiate into osteocytes that become embedded in the bone matrix or into lining cells on bone surfaces.

Osteoblast Differentiation and Function

Osteoblasts, bone-lining cells and osteocytes all arise from a multipotent precursor of mesenchymal origin that also gives rise to chondrocytes, adipocytes, myocytes and fibroblasts, most commonly called a mesenchymal stem cell [11] (fig. 2). The early differentiation process leading to osteochondroprogenitor cells involves Sox9, the key transcriptional regulator of chondrogenesis. Significant advances in the last 10 years have defined Runx2 [12, 13] and Osterix [14] as the two critical transcription factors determining osteoblast lineage differentiation. Runx2 (also called CBFA1) is a member of the Runx transcription factor family that is characterised by a DNA binding domain homologous with the Drosophila gene runt. It was identified as a causative gene for cleidocranial dysplasia (table 1) [12, 13]. It is expressed in mouse embryonic tissues in cells destined to become osteoblasts or chondrocytes in the developing embryo, and in all osteoblasts regardless of their differentiation stage [15]. Runx2–/– mice are unable to produce either endochondral or intramembranous bone [13, 16] but can produce adipocytes and chondrocytes [17]. Osterix is a zinc finger containing transcription factor of the SP transcription factor family. Osterix–/– mice are deficient in osteoblasts and do not form intramembranous bone [14]. They do express Runx2, but Runx2–/– mice do not express Osterix, indicating that Osterix acts downstream of Runx2. The Wnt signalling pathway in osteoblasts contains a number of molecules that are now viewed as amongst the most important regulators of bone formation during growth and development, mediating some of the regulatory dialogue between osteoblasts and osteoclasts [18]. Wnts are glycoproteins that in osteoblasts act on receptors composed of a Frizzled (Fz, a G protein coupled receptor-like protein) and one of the low density lipoprotein receptor related proteins LRP5 or LRP6. Activation

Bone Physiology

35

Table 1. Regulatory, structural and processing genes in bone and cartilage with known skeletal disorders in children Gene/protein

OMIM

Disorder

OMIM

Inheritance

*604539 Ehlers-Danlos syndrome, type VIIC #225410 AR ADAMTS2 Procollagen I N-proteinase ADAMTS10 A disintegrin-like metalloproteinase with thrombospondin type 1 motif, 10

*608990 Weill-Marchesani syndrome

#277600 AR

ALPL (TNSALP)* Alkaline phosphatase, liver/bone/kidney

*171760 hypophosphatasia, perinatal and infantile hypophosphatasia, childhood hypophosphatasia, adult

#241500 ?AR

CA2* Carbonic anhydrase II

*611492 osteopetrosis, autosomal recessive type III

#259730 AR

CASR* Calcium sensing receptor

+601199 severe neonatal hyperparathyroidism familial hypocalciuric hypercalcaemia with hyperparathyroidism

#239200 AR

*602727 osteopetrosis, autosomal recessive type IV osteopetrosis, autosomal dominant type II

#611490 AR

COL1A1* Pro-α1 collagen type I

+120150 osteogenesis imperfecta type IA osteogenesis imperfecta type IIA osteogenesis imperfecta type III osteogenesis imperfecta type IV Caffey disease

#166200 #166210 #259420 #166220 #114000

AD AD AD AD AD

COL1A2* Pro-α2 collagen type I

*120160 osteogenesis imperfecta type IB osteogenesis imperfecta type II osteogenesis imperfecta type III osteogenesis imperfecta type III osteogenesis imperfecta type IV

166240 #166210 #259420 203760 #166220

AD AD AD AR AD

CLCN7* Chloride channel 7

36

#241510 AD #146300 AD

#145980 AD

#166600 AD

Grabowski

Table 1. Continued Gene/protein

OMIM

COL2A1 Pro-α1 collagen type II

+120140 achondrogenesis type 2 (Langer-Saldino) platyspondylic dysplasia (Torrance) hypochondrogenesis spondyloepiphyseal dysplasia, congenital spondyloepimetaphyseal dysplasia (Strudwick) Kniest dysplasia spondyloperipheral dysplasia Stickler syndrome type 1 otospondylomegaepiphyseal dysplasia

#200610 AD

COL5A1 Pro-α1 collagen type V

*120215 Ehlers-Danlos syndrome type I Ehlers-Danlos syndrome type II

#130000 AD #130010 AD

COL5A2 Pro-α2 collagen type V

*120190 Ehlers-Danlos syndrome type I Ehlers-Danlos syndrome type II

#130000 AD #130010 AD

COL9A1 Pro-α1 collagen type IX

+120210 multiple epiphyseal dysplasia type 6 Stickler syndrome, autosomal recessive, Col9a1-related

#120210 AD

COL9A2 Pro-α2 collagen type IX

*120260 multiple epiphyseal dysplasia type 2

#600204 AD

COL9A3 Pro-α3 collagen type IX

*120270 multiple epiphyseal dysplasia type 3

#600969 AD

COL10A1 Pro-α1 collagen type X

*120110 metaphyseal chondrodysplasia (Schmid)

#156500 AD

COL11A1 Pro-α1 collagen type XI

*120280 Stickler syndrome type 2 Marshall syndrome

#604841 AD #154780 AD

COL11A2 Pro-α1 collagen type XI

*120290 otospondylomegaepiphyseal dysplasia

#215150 ?AR

CRTAP* Cartilage-associated protein

*605497 osteogenesis imperfecta type IIB osteogenesis imperfecta type VII

#610854 AR #610682 AR

CTSK* Cathepsin K

*601105 pycnodysostosis

#265800 AR

Bone Physiology

Disorder

OMIM

Inheritance

#151210 ? usually lethal #200610 AD #183900 AD #184250 AD #156550 #271700 #108300 #215150

AD AD AD AD

#108300 AR

37

Table 1. Continued Gene/protein

OMIM

Disorder

OMIM

Inheritance

FGF23* *605380 hypophosphataemic rickets, Fibroblast growth factor 23 ADHR familial hyperphosphataemic tumoral calcinosis

#193100 AD

FGFR3 Fibroblast growth factor receptor 3

*134934 achondroplasia thanatophoric dysplasia type I thanatophoric dysplasia type II hypochondroplasia

#100800 #187600 #187601 #146000

LEPRE1* Collagen prolyl 3-hydroxylase 1

*610339 osteogenesis imperfecta type VIII

#610915 AR

LRP5* Low-density lipoprotein receptor-related protein 5

*603506 osteoporosis-pseudoglioma syndrome high bone mass osteopetrosis, autosomal dominant type I endosteal hyperostosis, autosomal dominant

#259770 AR

*120360 Torg-Winchester syndrome MMP2 multicentric osteolysis, nodulosis Matrix metalloproteinase 2 and arthropathy

#211900 AD AD AD AD AD

#601884 AD #607634 AD #144750 AD #259600 AR #605156 AR

MMP13 Matrix metalloproteinase 13

*600108 spondyloepimetaphyseal dysplasia type II metaphyseal anadysplasia

#602111 AD

OSTM1* Osteopetrosis-associated transmembrane protein 1

*607649 osteopetrosis, autosomal recessive type V

#259720 AR

PHEX* Phosphate-regulating endopeptidase homologue, X-linked

*300550 X-linked hypophosphataemia

#307800 XLD

PLEKHM1* *611466 osteopetrosis, autosomal Pleckstrin homology recessive type VI domain-containing, family M (with RUN domain) member 1 PLOD1 Procollagen lysyl hydroxylase 1

38

*153454 Ehlers-Danlos syndrome type VIA Nevo syndrome

309645

AD

#611497 AR

#225400 ?AR #601451 ?AR

Grabowski

Table 1. Continued Gene/protein

OMIM

Disorder

OMIM

Inheritance

PLOD2* Procollagen lysyl hydroxylase 2

*601865 Bruck syndrome type 2

#609220 AR

PLOD3 Procollagen lysyl hydroxylase 3

*603066 bone fragility with contractures, arterial rupture and deafness

#612394 AR

PTHR1* Parathyroid hormone receptor 1

*168468 metaphyseal chondrodysplasia (Jansen) chondrodysplasia (Blomstrand) endochondromatosis (Ollier) Eiken syndrome

#156400 AD

RUNX2* Runt-related transcription factor 2

*600211 cleidocranial dysplasia

#119600 AD

SOST* Sclerostin

*605740 sclerosteosis endosteal hyperostosis (Van Buchem)

#269500 AR #239100 AR

SOX9 Sex-determining region Y-related homeobox gene 9

*608160 campomelic dysplasia

#114290 XLR

TCIRG1 Vacuolar proton pump α-subunit 3

*604592 osteopetrosis, autosomal recessive type I

#259700 AR

TGFβ1* Transforming growth factor-β1

*190180 Camurati-Engelmann disease

#131300 AD

TNFRSF11A* Receptor activator of NF-κB (RANK)

*603499 familial expansile osteolysis Paget disease of bone osteopetrosis, autosomal recessive type VII

#174810 AD #602080 AR #612301 AR

TNFRSF11B* Osteoprotegerin (OPG)

*602643 juvenile Paget disease

#239000 AR

TNFSF11* Receptor activator of NF-κB ligand (RANKL)

*602642 osteopetrosis, autosomal recessive type II

#259710 AR

#215045 ?AR #166000 ?AD #600002 ?AR

The genes are listed in alphabetical order and those genes associated with conditions that are described in more detail in the book are indicated by (*). The genes involved and the conditions arising from mutations in those genes are shown with their appropriate OMIM numbers. See also Superti-Furga and Unger [55] for a comprehensive classification of genetic skeletal disorders.

Bone Physiology

39

a

b

c

Fig. 3. Wnt signalling in osteoblasts. a Osteoblasts express canonical Wnt pathway coreceptor molecules including LRPs and Frizzled family members. Wnts interact with LRP5/6 and Frizzled proteins to form a complex that leads to the inhibition of β-catenin destruction mediated by Dishevelled. As a result, β-catenin accumulates and is translocated to the nucleus where it interacts with transcription factors to modulate gene expression. b Dickkopf (Dkk) inhibits Wnt signalling by binding LRPs to Kremen, enhancing their internalisation and destruction. c Sclerostin (Sost) competes with Wnt for binding to LRPs, preventing the interaction of LRPs with Frizzled proteins.

of the Wnt receptor results in dephosphorylation and accumulation of intracellular β-catenin and its translocation into the nucleus, where it interacts with transcription factors to control osteoblast gene expression (fig. 3). The Dickkopf family of proteins act as negative regulators of Wnt signalling by binding to LRP5/6 and another cell surface co-receptor Kremen, causing internalisation and destruction of the resulting complex and reducing the density of Wnt receptors at the cell surface (fig. 3). Wnt signalling in osteoblasts exerts its effects in bone primarily by regulating osteoclast formation through modulating the production of osteoprotegerin (OPG), the soluble inhibitor of the RANK signalling pathway [19]. Gain of function mutations in LRP5 result in high bone mass disorders (table 1) while loss of function mutations in LRP5 result in the low bone mass disorder osteoporosis-pseudoglioma syndrome (table 1).

Bone and Cartilage Matrix Collagens and Their Modifying Enzymes

The major function of osteoblasts is to create a mineralised bone matrix which, until mineralised, is called osteoid. Type 1 collagen accounts for about 90% of osteoid content, with the remainder composed largely of glycoproteins and proteoglycans. Other proteins that are important for mineralisation, including alkaline phosphatase, osteocalcin and osteopontin, are also secreted by osteoblasts into the newly forming matrix. The process of bone mineralisation is poorly understood. Collagens are a diverse family of structural proteins found in extracellular matrices. They are the most abundant proteins in the body and there are at least twenty-eight

40

Grabowski

different collagens found in vertebrates. The characteristic feature of all collagens is a triple helical structure consisting of three interwoven α-chain polypeptides. The triple helical region of collagen α-chains consists of a repetitive series of amino-acid triplets – [Gly-X-Y]n – where Gly is glycine and X and Y are commonly proline or hydroxyproline. Collagens can be homo- or hetero-trimeric proteins, i.e. with three identical α-chains or with three α-chains encoded by either two or three unique genes respectively. During synthesis, carboxy pro-peptide regions of three pro-α-chains associate within the endoplasmic reticulum to initiate the formation of the triple helix which propagates towards the amino terminus. During assembly, many proline and lysine residues in fibrillar collagens become hydroxylated and some of the hydroxylysine residues are further modified by glycosylation. Intra- and inter-chain disulphide bonds are also formed during synthesis. Carboxy- and amino-propeptides are proteolytically cleaved extracellularly after secretion and the released monomers assemble into highly orientated, quarter-staggered fibrils which are held together through covalent cross-links promoted by the action of lysyl oxidase. The collagen triple helix is highly resistant to proteolytic cleavage by pepsin, trypsin and papain, and degradation of collagens is mediated by matrix metalloproteinases, cysteine proteinases (especially cathepsins B, K and L) and serine proteinases [20]. In bone, type I collagen is the most abundant fibrillar collagen. It is normally heterotrimeric, consisting of two α1(I) chains and one α2(I) chain. In the absence of pro-α2(I) chains, type I collagen α1(I) homotrimers can form. In bone, type I collagen forms heterotypic fibrils with type V collagen, a low-abundance fibrillar protein with three distinct α-chains – α1(V), α2(V) and α3(V). While the COL1A2 gene is not essential for survival, homozygous COL1A1 null mutations are not seen clinically. Skeletal phenotypes arising from mutations in type I collagen give rise to osteogenesis imperfecta and various forms of Ehlers-Danlos syndrome (EDS) (table 1), with mutations in type V collagen also giving rise to EDS (table 1). In cartilage, type II collagen is the most abundant fibrillar collagen, consisting of α1(II) homotrimers. Additionally, in cartilage, the pro-α1(II) chain is incorporated into heterotrimeric type XI collagen along with a pro-α1(XI) and a pro-α2(XI) chain. Type XI collagen is a low-abundance fibrillar collagen that forms heterotypic fibrils with type II collagen in cartilage. Yet another low abundance collagen – type IX, composed of three distinct α-chains – α1(IX), α2(IX) and α3(IX) – also forms heterotypic fibrils with type II collagen in cartilage. Type IX collagen has a triple helix that is interrupted by short non-helical regions which give the molecule some flexibility and it is classed as a fibril associated collagen with interrupted triple helices (FACIT). Mutations in type II collagen give rise to a variety of chondrodysplasias (table 1). Type IX collagen mutations give rise to various multiple epiphyseal dysplasias and Stickler syndrome, while mutations in type XI collagen give rise to Stickler and Marshall syndromes (table 1). Type X collagen is a homotrimeric protein found in the hypertrophic cartilage of the growth plate. Mutations in type X collagen result in Schmidt metaphyseal chondrodysplasia (table 1). Mutations that give rise to skeletal

Bone Physiology

41

phenotypes have also been identified in some collagen modifying and processing enzyme genes including ADAMTS2, ADAMTS10, CRTAP, CTSK, LEPRE1, MMP2, MMP13, PLOD1, PLOD2 and PLOD3 (table 1). For a more comprehensive review of the clinical genetics of collagen disorders, see Byers [21].

Osteocytes and Bone Lining Cells

As bone matrix is being formed, some osteoblasts undergo a terminal differentiation event (fig. 2) and, instead of continuing to produce matrix, they become osteocytes, encapsulated in concentric layers within lacunae in the osteoid. The signals that initiate and control terminal osteocytic differentiation are not known. Numerous dendritic cellular processes connect osteocytes to each other through canaliculi, both laterally and between cell layers within bone. When osteoblasts stop creating bone, they turn into bone lining cells and remain on the bone surface. The transition from osteoblasts into bone lining cells is poorly understood. Osteocytes and bone lining cells account for the largest proportion of cells in mineralised bone but are probably the least characterised and understood cells of bone. An important development in osteocyte biology has been the identification of sclerostin, mutations in which cause sclerosteosis and Van Buchem disease (table 1), both characterised by progressive bone thickening [22, 23]. Sclerostin binds to a number of BMP growth factors and has been shown to inhibit BMP-mediated osteoblast differentiation [24]. It also binds directly to LRP5/6, preventing activation of the Wnt signalling pathway (fig. 3) [18].

Osteoclasts

Osteoclasts are large multinucleate cells found in close apposition to bone surfaces undergoing resorption. Osteoclast precursors share the same haematopoietic lineage as macrophages. One of the most significant breakthroughs in bone biology of the last fifteen years has been the identification and characterisation of the molecular pathway controlling osteoclastogenesis [25] (fig. 4). Osteoclast precursors express the receptor for macrophage colony stimulating factor (MCSF) which, when stimulated, promotes the expression of a TNF superfamily molecule, receptor activator of nuclear factor kappa-B (RANK). Osteoblasts control the differentiation of osteoclast precursors through production of RANK ligand (RANKL), a cell surface molecule that is the primary effector of the RANK receptor, and osteoprotegerin (OPG), a soluble decoy receptor for RANKL. The balance between RANKL and OPG concentrations regulates RANK activation. The RANK receptor can activate a network of intracellular pathways [25]. MCSF/RANKL stimulated osteoclast precursors form polykaryons through a process of cell membrane fusion that is poorly understood. RANK activation is also necessary for mature osteoclast activity. When settled onto bone, osteoclasts form a

42

Grabowski

Fig. 4. Osteoclast differentiation. Osteoclasts arise from circulating haematopoietic cells of the monocyte/macrophage lineage (osteoclast precursors) that express c-fms, the receptor for macrophage colony-stimulating factor (MCSF). MCSF, produced by osteoblasts, stimulates expression of receptor activator of nuclear factor κ-B (RANK) on osteoclast precursors. The key regulator of osteoclastogenesis, RANK ligand (RANKL), is expressed by osteoblasts. RANKL binds to RANK and stimulates the fusion of osteoclast precursors to form multinucleated immature osteoclasts. Mature osteoclasts form a tight seal and generate a ruffled membrane border against the bone surface through which they secrete acid and proteolytic enzymes forming a resorption lacuna. RANKL acts on osteoclasts at all stages of differentiation. Osteoclastogenesis is regulated by osteoblasts through balancing the production of RANKL and Osteoprotegerin (OPG), a decoy receptor produced by osteoblasts to inhibit the interaction of RANKL and its receptor RANK.

region of tight contact between the cell and the bone surface known as the sealing zone, creating a tightly enclosed area underneath the osteoclast where bone resorption takes place. The cellular membranes within the sealing zone develop into a ruffled border, a structure of deeply folded cellular membranes adjacent to the bone surface, through which are secreted acid and proteolytic enzymes to mediate bone resorption. Defects in genes encoding the molecular pathway controlling acid production (CA2) and its secretion through ion channels (TCIRG1, CLCN7, OSTM1) have been identified as frequent causes of osteoclast rich infantile onset osteopetrosis (table 1; fig. 1, chapter 12) (reviewed in [26]). More recently, defects in RANKL [27] and in RANK [28] have been identified in cases of infantile onset osteopetrosis in which osteoclasts are completely absent (table 1; fig. 1, chapter 12). The ultimate fate of the osteoclast is apoptosis.

Hormones and Mineral Homeostasis

Parathyroid hormone (PTH), parathyroid hormone-related peptide (PTH-rP) and vitamin D are the key hormonal regulators of mineral homeostasis. PTH acts in bone to stimulate the release of calcium and phosphate, while at the same time acting on

Bone Physiology

43

the kidney to enhance calcium reabsorption and to inhibit reabsorption of phosphate. PTH also enhances the renal production of 1,25(OH)2vitamin D, which acts on the intestine to enhance calcium absorption. The molecular mechanisms regulating intestinal absorption and renal reabsorption of calcium have recently been identified. Two calcium selective ion channels, TRPV5 and TRPV6, mediate epithelial transcellular transport of calcium from the intestinal or renal tubular lumen to the extracellular fluid, by way of intracellular calcium binding proteins and cell surface ion pumps (reviewed in Nijenhuis et al. [29]). FGF23 was identified as the gene responsible for autosomal dominant hypophosphataemic rickets (ADHR) (table 1) [30]. FGF23–/– mice have hyperphosphataemia, high renal reabsorption of phosphate and high circulating levels of 1,25(OH)2D [31, 32]. The receptor for FGF23 is a heterocomplex of the canonical FGF receptor and klotho [33]. Interestingly, klotho has β-gluguronidase enzymatic activity which is responsible for hydrolysing external sugar residues of TRPV5, resulting in trapping and activation of this calcium selective ion channel at the cell surface [34].

The Nervous System and Bone

Although nerves are found throughout the periosteum and near metabolically active parts of bone, until very recently little thought was given to the potential role of the nervous system in bone. Evidence for nerve fibres that signal through the neuropeptides calcitonin gene-related peptide (CGRP), vasoactive intestinal peptide (VIP) and substance P began to emerge in the 1980s (reviewed in Jones et al. [35]). The discovery of a glutamate/aspartate transporter molecule GLAST in bone, previously associated with glutamatergic neuronal signalling [36], energised the search for glutamate signalling mechanisms acting on bone formation and resorption. Osteoblasts, osteocytes and osteoclasts have all been shown to express ionotropic and metabotropic glutamate and NMDA receptors, and ion channel controlled electrical currents consistent with these receptors have been measured in osteoclasts [37]. The inhibitory action on bone formation of leptin, a hormone involved in controlling body mass, has been shown to be mediated through adrenergic signalling resulting from leptin acting in the hypothalamus [38, 39].

Functional Adaptation of Bone to Load Bearing

Bones need to be stiff enough to bear the loads they are commonly subjected to without deforming or breaking under load [40]. Bone strength is a function of its stiffness and is dependent on several factors including size, shape and material composition/ spatial distribution. The concept of functional adaptation of the skeleton, consolidated in Wolff ’s law [41], was developed into the mechanostat hypothesis [2] based

44

Grabowski

on observations that load bearing vertebrate bones undergo relatively few spontaneous as opposed to traumatic fractures, indicating that bones adapt their strength to be able to endure typical peak mechanical loads without fracturing while having a sufficiently large safety margin to endure occasional supra-normal loads. In the mechanostat hypothesis, Frost [2] describes a series of rules by which mechanical competence may be achieved and maintained in load bearing bones (which includes many non weight bearing bones e.g. mandibles), providing tissue-level mechanisms for functional adaptation of the skeleton to the loads placed upon it and presenting predictable hypotheses concerning the mechanical implications of bone disease. The key element is the biological machinery to sense the level of loading and to respond by increasing or decreasing the mechanical competence. To date, the nature of the mechanostat in bone is still unknown. Evidence dictates that such machinery is local to the bone under load, since bones respond to local or asymmetrical loads. This is typically seen in athletes such as tennis players in whom bones of the playing arm are strengthened preferentially [42]. A mechanostat model also implies upper and lower thresholds within which bone is sensed to be under normal loading, with loads above or below triggering a response to model or remodel the bone. Lanyon et al. [43] in the 1970s developed the concept that bone cells respond to the magnitude of the strain experienced by a bone under load (the ratio of change in bone length divided by original length) while, more recently, work in experimental animals and humans has shown that other factors relating to strain are also important, including the strain direction, rate of strain change, duration of loading, the number of loading cycles, frequency, repetition and rest within and between cycles [44]. Whether bone cells sense mechanical loads through direct cellular deformation or through shear strains induced by interstitial fluid flow resulting from bone deformation is unclear, but molecular mechanisms capable of strain detection have been described in osteoblasts, osteocytes, osteoclasts and vascular endothelial cells, with ion channels, integrins and associated proteins, connexins, cell surface structures, the cytoskeleton and nitric oxide as potential molecular mediators [45, 46]. The mechanostat hypothesis is consistent with observations that maternal environment and intrauterine muscular activity influence load bearing bones by the time of birth, and it makes provision for the influence of non-mechanical factors in bone growth, development and maintenance.

Healthy Bones for Life: Nutrition and Exercise in Bone Physiology

The impact of bone loss in the ageing population on quality of life and in health service provision has in the last 20–30 years introduced the consideration of public policy strategies to maximise bone mass accrual during childhood and adolescence by manipulation of diet or through exercise. Almost half of our adult bone mineral mass is accrued by the skeleton in the 3–4 years following the onset of puberty [47, 48],

Bone Physiology

45

making adolescence one of the most critical periods for skeletal development. Peak bone mass, achieved during the third decade, is a powerful predictor of postmenopausal osteoporosis [49, 50] and it is widely assumed that optimising peak bone mass accrual during childhood and adolescence will produce bones that are better equipped to handle the inevitable loss of bone in later life. The two simplest ways to influence bone mass accrual are through nutritional intervention and exercise. A recent systematic review of 22 controlled trials concluded that weight-bearing exercise in children and adolescents leads to modest increases in bone parameters over 6 months [51]. However, it is difficult to assess which exercise activities are most appropriate, or the time frame within which they should be undertaken. Growth occurs heterogeneously in the skeleton throughout childhood where, for example, the longitudinal growth velocity of the legs in infancy is about twice that in the spine until puberty [52]. Benefits of exercise are likely to be site specific, depend on the type of exercise (weight bearing, high impact) and may be influenced by dietary and hormonal factors [53], and even by conditions experienced while in utero [52]. The mechanostat theory predicts that bone strength will decrease on disuse, but evidence is emerging from studies in former athletes and animals that skeletal benefits may persist despite lack of exercise [54]. Well-designed long-term studies are needed to see if such benefits may be achievable in children.

Conclusions

Advances in understanding the molecular mechanisms and pathways regulating bone cell function bring with them opportunities for the development of novel, rational approaches to treat disorders of bone cell dysfunction in children. Knowledge gained from systematic study of children’s bone disorders also provides insights into the physiology of healthy bones. Improving our understanding of the physiology of bone growth and development during childhood will lead to better prospects for finding early pharmacological, physiotherapeutic and nutritional strategies to optimise and maintain bone health from childhood into adulthood, which in turn may help to reduce the burden of bone loss and its related ill-health in old age.

References 1 Frost HM: From Wolff ’s law to the Utah paradigm: insights about bone physiology and its clinical applications. Anat Rec 2001;262:398–419. 2 Frost HM: Bone’s mechanostat: a 2003 update. Anat Rec A Discov Mol Cell Evol Biol 2003;275:1081– 1101. 3 Jee WS: Principles in bone physiology. J Musculoskelet Neuronal Interact 2000;1:11–13.

46

4 Duncan RL, Turner CH: Mechanotransduction and the functional response of bone to mechanical strain. Calcif Tissue Int 1995;57:344–358. 5 Ducy P, Karsenty G: Genetic control of cell differentiation in the skeleton. Curr Opin Cell Biol 1998;10: 614–619. 6 Chen D, Zhao M, Mundy GR: Bone morphogenetic proteins. Growth Factors 2004;22:233–241.

Grabowski

7 Xu X, Weinstein M, Li C, Deng C: Fibroblast growth factor receptors (FGFRs) and their roles in limb development. Cell Tissue Res 1999;296:33–43. 8 Iwamoto M, Enomoto-Iwamoto M, Kurisu K: Actions of hedgehog proteins on skeletal cells. Crit Rev Oral Biol Med 1999;10:477–486. 9 Church VL, Francis-West P: Wnt signalling during limb development. Int J Dev Biol 2002;46:927–936. 10 Karaplis AC: Embryonic development of bone and the molecular regulation of intramembranous and endochondral bone formation; in Bilezikian JP, Raisz LG, Rodan GA (eds): Principles of Bone Biology. San Diego, Academic Press, 2002, pp 33–58. 11 Oreffo RO, Cooper C, Mason C, Clements M: Mesenchymal stem cells: lineage, plasticity, and skeletal therapeutic potential. Stem Cell Rev 2005;1: 169–178. 12 Mundlos S, Otto F, Mundlos C, Mulliken JB, Aylsworth AS, Albright S, et al: Mutations involving the transcription factor CBFA1 cause cleidocranial dysplasia. Cell 1997;89:773–779. 13 Otto F, Thornell AP, Crompton T, Denzel A, Gilmour KC, Rosewell IR, et al: Cbfa1, a candidate gene for cleidocranial dysplasia syndrome, is essential for osteoblast differentiation and bone development. Cell 1997;89:765–771. 14 Nakashima K, Zhou X, Kunkel G, et al: The novel zinc finger-containing transcription factor osterix is required for osteoblast differentiation and bone formation. Cell 2002;108:17–29. 15 Lee B, Thirunavukkarasu K, Zhou L, et al: Missense mutations abolishing DNA binding of the osteoblast-specific transcription factor OSF2/CBFA1 in cleidocranial dysplasia. Nat Genet 1997;16:307– 310. 16 Komori T, Yagi H, Nomura S, et al: Targeted disruption of Cbfa1 results in a complete lack of bone formation owing to maturational arrest of osteoblasts. Cell 1997;89:755–764. 17 Kobayashi H, Gao Y, Ueta C, Yamaguchi A, Komori T: Multilineage differentiation of Cbfa1-deficient calvarial cells in vitro. Biochem Biophys Res Commun 2000;273:630–636. 18 Baron R, Rawadi G: Targeting the Wnt/beta-catenin pathway to regulate bone formation in the adult skeleton. Endocrinology 2007;148:2635–2643. 19 Glass DA, Bialek P, Ahn JD, et al: Canonical Wnt signaling in differentiated osteoblasts controls osteoclast differentiation. Dev Cell 2005;8:751–764. 20 Kadler KE, Baldock C, Bella J, Boot-Handford RP: Collagens at a glance. J Cell Sci 2007;120:1955– 1958. 21 Byers PH: Collagens: building blocks at the end of the development line. Clin Genet 2000;58:270–279.

Bone Physiology

22 Balemans W, Ebeling M, Patel N, et al: Increased bone density in sclerosteosis is due to the deficiency of a novel secreted protein (SOST). Hum Mol Genet 2001;10:537–543. 23 Brunkow ME, Gardner JC, Van Ness J, et al: Bone dysplasia sclerosteosis results from loss of the SOST gene product, a novel cystine knot-containing protein. Am J Hum Genet 2001;68:577–589. 24 Kusu N, Laurikkala J, Imanishi M, et al: Sclerostin is a novel secreted osteoclast-derived bone morphogenetic protein antagonist with unique ligand specificity. J Biol Chem 2003;278:24113–24117. 25 Boyle WJ, Simonet WS, Lacey DL: Osteoclast differentiation and activation. Nature 2003;423:337– 342. 26 Del Fattore A, Cappariello A, Teti A: Genetics, pathogenesis and complications of osteopetrosis. Bone 2008;42:19–29. 27 Sobacchi C, Frattini A, Guerrini MM, et al: Osteoclast-poor human osteopetrosis due to mutations in the gene encoding RANKL. Nat Genet 2007;39:960–962. 28 Guerrini MM, Sobacchi C, Cassani B, et al: Human osteoclast-poor osteopetrosis with hypogammaglobulinemia due to TNFRSF11A (RANK) mutations. Am J Hum Genet 2008;83:64–76. 29 Nijenhuis T, Hoenderop JG, Bindels RJ: TRPV5 and TRPV6 in Ca(2+) (re)absorption: regulating Ca(2+) entry at the gate. Pflügers Arch 2005;451:181–192. 30 ADHR Consortium: Autosomal dominant hypophosphataemic rickets is associated with mutations in FGF23. Nat Genet 2000;26:345–348. 31 Shimada T, Kakitani M, Yamazaki Y, et al: Targeted ablation of Fgf23 demonstrates an essential physiological role of FGF23 in phosphate and vitamin D metabolism. J Clin Invest 2004;113:561–568. 32 Sitara D, Razzaque MS, Hesse M, et al: Homozygous ablation of fibroblast growth factor-23 results in hyperphosphatemia and impaired skeletogenesis, and reverses hypophosphatemia in Phex-deficient mice. Matrix Biol 2004;23:421–432. 33 Urakawa I, Yamazaki Y, Shimada T, et al: Klotho converts canonical FGF receptor into a specific receptor for FGF23. Nature 2006;444:770–774. 34 Chang Q, Hoefs S, van der Kemp AW, Topala CN, Bindels RJ, Hoenderop JG: The beta-glucuronidase klotho hydrolyzes and activates the TRPV5 channel. Science 2005;310:490–493. 35 Jones KB, Mollano AV, Morcuende JA, Cooper RR, Saltzman CL: Bone and brain: a review of neural, hormonal, and musculoskeletal connections. Iowa Orthop J 2004;24:123–132. 36 Mason DJ, Suva LJ, Genever PG, et al: Mechanically regulated expression of a neural glutamate transporter in bone: a role for excitatory amino acids as osteotropic agents? Bone 1997;20:199–205.

47

37 Laketic-Ljubojevic I, Suva LJ, Maathuis FJ, Sanders D, Skerry TM: Functional characterization of N-methyl-d-aspartic acid-gated channels in bone cells. Bone 1999;25:631–637. 38 Ducy P, Amling M, Takeda S, et al: Leptin inhibits bone formation through a hypothalamic relay: a central control of bone mass. Cell 2000;100:197– 207. 39 Takeda S, Elefteriou F, Levasseur R, et al: Leptin regulates bone formation via the sympathetic nervous system. Cell 2002;111:305–317. 40 Currey JD: Bones: Structure and Mechanics. Princeton, Princeton University Press, 2002. 41 Wolff J: Das Gesetz der Transformation der Knochen. Berlin, Hirschwald, 1892. 42 Haapasalo H, Kontulainen S, Sievanen H, Kannus P, Jarvinen M, Vuori I: Exercise-induced bone gain is due to enlargement in bone size without a change in volumetric bone density: a peripheral quantitative computed tomography study of the upper arms of male tennis players. Bone 2000;27:351–357. 43 Clark EA, Goodship AE, Lanyon LE: Locomotor bone strain as the stimulus for bone’s mechanical adaptability. J Physiol 1975;245:57P. 44 Skerry TM: One mechanostat or many? Modifications of the site-specific response of bone to mechanical loading by nature and nurture. J Musculoskelet Neuronal Interact 2006;6:122–127. 45 Rubin J, Rubin C, Jacobs CR: Molecular pathways mediating mechanical signaling in bone. Gene 2006; 367:1–16. 46 Malone AM, Anderson CT, Tummala P, Kwon RY, Johnston TR, Stearns T et al.: Primary cilia mediate mechanosensing in bone cells by a calcium-independent mechanism. Proc Natl Acad Sci USA 2007; 104:13325–13330.

47 Bailey DA: The Saskatchewan Pediatric Bone Mineral Accrual Study: bone mineral acquisition during the growing years. Int J Sports Med 1997; 18(suppl 3):S191–S194. 48 Bailey DA, Martin AD, McKay HA, Whiting S, Mirwald R: Calcium accretion in girls and boys during puberty: a longitudinal analysis. J Bone Miner Res 2000;15:2245–2250. 49 Hui SL, Slemenda CW, Johnston CC Jr: The contribution of bone loss to postmenopausal osteoporosis. Osteoporos Int 1990;1:30–34. 50 Seeman E: Reduced bone density in women with fractures: contribution of low peak bone density and rapid bone loss. Osteoporos Int 1994;4(suppl 1):15–25. 51 Hind K, Burrows M: Weight-bearing exercise and bone mineral accrual in children and adolescents: a review of controlled trials. Bone 2007;40:14–27. 52 Cooper C, Westlake S, Harvey N, Javaid K, Dennison E, Hanson M: Review: developmental origins of osteoporotic fracture. Osteoporos Int 2006;17:337– 347. 53 Loud KJ, Gordon CM: Adolescent bone health. Arch Pediatr Adolesc Med 2006;160:1026–1032. 54 Ducher G, Bass SL: Exercise during growth: compelling evidence for the primary prevention of osteoporosis? BoneKEy – Osteovision 2007;4:171– 180. 55 Superti-Furga A, Unger S: Nosology and classification of genetic skeletal disorders: 2006 revision. Am J Med Genet A 2007;143:1–18.

Peter Grabowski, PhD Faculty of Medicine, Dentistry and Health, Academic Unit of Child Health, University of Sheffield Sheffield S10 2RX (UK) Tel. +44 0114 2711798, Fax +44 0114 2755463, E-Mail [email protected]

48

Grabowski

Chapter 4 Allgrove J, Shaw NJ (eds): Calcium and Bone Disorders in Children and Adolescents. Endocr Dev. Basel, Karger, 2009, vol 16, pp 49–57

Bone Biopsy: Indications and Methods Frank Rauch Genetics Unit, Shriners Hospital for Children and McGill University, Montréal, Qué., Canada

Abstract In the context of metabolic bone disorders, obtaining biopsies of iliac bone can be useful for establishing a diagnosis in an individual patient or for investigating pathomechanisms when a series of samples is examined. Although bone specimens are usually decalcified for routine pathology to facilitate sample processing, when investigating metabolic bone disorders it is usually much more informative to analyse undecalcified samples. Biopsy samples can be assessed qualitatively and quantitatively. Quantitative analysis by computerised histomorphometry of undecalcified bone biopsy samples is a key tool for studying bone metabolism and, to a lesser extent, bone mass and structure. Standard histomorphometric analyses focuses on trabecular bone and therefore mainly provides information on trabecular remodelling. Remodelling activity changes markedly with age during development. This has to be taken into account when histomorphometry is used in the paediatric setting. Children and adolescents with severe bone fragility should have a bone biopsy for diagnostic purposes unless the diagnosis is obvious from non-invasive examinations. Quantitative histomorphometric analysis of transiliac bone biopsy samples is especially valuable in clinical studies, as this method provides safety and efficacy data that can not be obtained in any other way. Copyright © 2009 S. Karger AG, Basel

In the context of metabolic bone disorders, obtaining biopsies of iliac bone can be useful for establishing a diagnosis in an individual patient or for investigating pathomechanisms when a series of samples is examined. Biopsy samples can be used for qualitative assessment – similar to the pathologist’s evaluation of other tissue specimens – and for quantitative analysis, called histomorphometry. Bone tissue is very hard and for that reason is more difficult to process than soft tissue. In routine pathology, bone tissue is therefore often decalcified and thus converted into a soft tissue. However, this leads to the loss of important information about bone mineralisation and bone cell activity. To assess metabolic bone disorders, it is therefore generally more informative to analyse samples undecalcified. Computerised quantitative histomorphometry of undecalcified bone biopsy samples is a method to obtain direct quantitative information on bone tissue. When tetracycline labelling is performed prior to biopsy, bone cell function can be studied in

vivo. Important for paediatric use, bone histomorphometric results are not directly influenced by the growth process. In contrast to some currently popular indirect methods of bone analysis, histomorphometry yields results with a known meaning. Knowledge of bone tissue is also crucial for interpreting the findings of molecular and cellular studies. Despite these advantages, bone histomorphometry is underused in paediatrics. This may be partly due to the fact that histomorphometry requires an invasive procedure to obtain a bone sample, is labour intensive, and needs special equipment and expertise. Other reasons may include overestimation of the utility of non-invasive bone diagnostics and lack of information about what bone histomorphometry does. The present contribution tries to address this latter point. More detailed information on paediatric histomorphometry is available elsewhere [1].

Bone Biopsy Procedure

Bone histomorphometry was first developed to study rib bone samples. This was soon abandoned because the ilium proved to be a much more convenient site for obtaining bone samples. In principle, histomorphometric analysis can be performed in any bone. In clinical paediatrics, however, the utility of samples from nonstandard sites is limited because reference data are only available for the ilium. Quantitative bone histomorphometry requires an intact biopsy specimen of good quality. This implies that the transiliac sample must be obtained under standardised conditions and with appropriate tools. It is essential that the sample is not fractured or crushed and contains two cortices separated by a trabecular compartment. These requirements are often quite difficult to meet in small or very osteopaenic children. Bone specimens for histomorphometric evaluation are horizontal, full-thickness biopsy samples of the ilium from a site 2 cm posterior from the anterior superior iliac spine. This bone is easily accessible, does not require extensive surgery, and is associated with few postoperative complications. Also, this is the only site for which paediatric histomorphometric reference data have been published [2]. A correctly performed biopsy procedure should yield a sample containing two cortices that are separated by a trabecular compartment (fig. 1). Vertical samples (from the iliac crest downwards, also called Jamshidi approach) are of questionable utility because of the presence of the growth plate at the top of the iliac crest. Turnover is very high and cortical thickness is very low in bone tissue below the growth plate and results are therefore not representative. Thus, the often-used term ‘iliac crest biopsy’ is a misnomer, as the iliac crest actually should be avoided during the biopsy procedure. The more accurate term is ‘transiliac biopsy’. The most widely used bone biopsy instrument is the Bordier needle. The inner diameter of the needle should be 5 mm (which we use in patients up to the age of 12 years) or 6 mm (for patients older than 12 years). The size of the needle diameter is

50

Rauch

Fig. 1. Section of an entire iliac biopsy specimen of a 9-year-old boy without metabolic bone disease. In this section core width is 7.2 mm, cortical width (i.e. the average length of the arrows indicated in the two cortical compartments) is 894 μm and bone volume per tissue volume in the trabecular compartment is 24.5%. Osteoid and cellular structures can not be identified at this magnification.

Cortical compartment

Trabecular compartment

Core width

important, because an appropriately large sample area must be available for histomorphometric analyses to obtain representative measures. A smaller needle diameter means that a smaller bone sample is obtained and that the margin of error of the histomorphometric analysis is wider. Most children younger than 14 years of age require general anaesthesia for the procedure. Local anaesthesia can be sufficient for older adolescents. Patients are allowed to get out of bed after 3 h and can usually be discharged on the same day. Another prerequisite for histomorphometric evaluation is bone labelling. Dynamic parameters of bone cell function can only be measured when the patient has received two courses of tetracycline label prior to biopsy. Tetracycline compounds form calcium chelate complexes that bind to bone surfaces. These complexes are buried within the bone at sites of active bone formation, whereas they redissociate from the other bone surfaces once serum tetracycline levels decrease. The tetracycline trapped at formation sites can then be visualised under fluorescent light (fig. 2). The most widely used tetracycline compound is demeclocycline hydrochloride (DeclomycinT, Ledermycin®). Two labelling cycles are given before the biopsy procedure, each one lasting for 2 days. Declomycin is given orally in two doses per day with a daily dose of 15–20 mg/kg body weight (maximum dosage: 900 mg/day). The first labelling course is given on days 17 and 16 before the biopsy procedure, the second course is given on days 5 and 4 before the procedure. The two courses are thus separated by an interlabel time of 10 days. Although children and adolescents generally tolerate tetracycline double labelling well, some side effects, such as allergic reactions, vomiting, and photosensitivity, might be observed. Administering the drug after meals can diminish gastrointestinal side effects. It is important that these meals do not include milk or other dairy products because tetracycline complexes with calcium contained in the food and is not absorbed adequately. Sun exposure must be

Bone Biopsy

51

Fig. 2. Trabecular remodelling. Multiple short tetracycline double-labels are present. Dynamic bone formation parameters are derived from the length of the tetracycline labels and the distance between the labels.

avoided while taking tetracyclines. Tetracycline use is generally not recommended for children younger than nine years of age because discolouration of teeth may occur. However, the previously mentioned schedule appears to be safe in this respect. At the Montreal Shriners Hospital, it has been used for more than 350 biopsies in children younger than 9 years of age and tooth discolouration has never been observed.

Sample Processing

The biopsy sample should be placed in a fixative solution as soon as possible after the procedure. The fixation process aims at the preservation of bone tissue constituents by inactivating lysosomal enzymes. The choice of fixative and temperature at which the sample should be kept depends on the planned staining techniques. For routine histomorphometry, 70% ethanol or 10% buffered formalin at room temperature can be used. The duration of fixation should be at least 48 h but should not exceed 10 days because the tetracycline labels are washed out when fixation is too long. Once in fixative, the sample can be sent to the laboratory where samples are cut and stained and where the histomorphometric analyses are performed.

Histomorphometric Parameters

Histomorphometrists use standardised terminology and clear definitions that were established by a working group of the American Society for Bone and Mineral Research [3]. According to these definitions, ‘bone’ is bone matrix, whether it is mineralised or not. Unmineralised bone matrix is called osteoid. The term ‘tissue’ refers to both bone and associated soft tissue, such as bone marrow. Histomorphometric measurements

52

Rauch

are performed in two-dimensional sections. Nevertheless, in order to stress the threedimensional nature of bone, the terminology committee favoured a three-dimensional nomenclature for reporting histomorphometric results [3]. Thus, what appears as a line in a microscopic bone section is called a surface, whereas what is visible as an area under the microscope is referred to as a volume. This is done simply by convention, and should not be mistaken as actual three-dimensional measurements. Histomorphometric parameters can be classified into four categories (table 1): Structural parameters, static bone formation parameters, dynamic formation parameters, and bone resorption parameters. Structural parameters describe the size and the amount of bone. The outer size of a transiliac biopsy specimen is called core width, a measure which reflects the thickness of the ilium. Cortical width is the average width of the two cortices. Bone volume per tissue volume of trabecular bone represents the proportion of the marrow cavity which is occupied by bone. In trabecular bone, bone volume per tissue volume can be schematically separated into two components, trabecular thickness and trabecular number. The group of static formation parameters comprises the surface extent, thickness and relative amount of osteoid, as well as the surface extent of osteoblasts (called osteoid surface per bone surface, osteoid thickness, osteoid volume per bone volume and osteoblasts surface per bone surface, respectively; table 1). Wall thickness reflects the amount of bone that is created by an osteoblast team during a remodelling event. Wall thickness should not be confused with cortical thickness, with which is does not have any relationship. Dynamic bone formation parameters yield information on in vivo bone cell function and can only be measured when patients have received two courses of tetracycline label prior to biopsy (table 1). The two basic parameters are the surface extent of mineralisation activity (mineralising surface per bone surface) and the speed of mineralisation in a direction perpendicular to the bone surface (mineral apposition rate). From these primary measures, mineralisation lag time and bone formation rate per bone surface are derived mathematically. It should be noted that a high bone formation rate does not necessarily lead to a net gain of bone. If the remodelling balance is zero, the amount of bone will remain unchanged even if bone formation rate is very high. The combination of a negative remodelling balance and high bone formation rate will even lead to rapid bone loss. Thus, bone formation rate per bone surface in trabecular bone indicates the activity of bone turnover rather than bone gain [4]. Bone resorption can only be quantified with static parameters, which makes evaluation of bone resorption the least informative aspect of histomorphometric analysis. It is possible to quantify the extent of bone surface that is covered by osteoclasts or which looks eroded (osteoclast surface per bone surface and eroded surface per bone surface, respectively), but it is not possible to tell from these measures how much bone resorption is actually going on. This may be an issue in the evaluation of renal bone disease. In chronic renal failure, osteoclasts resorb bone more slowly than normal, so that the extent of osteoclast and eroded surfaces overestimates the rate of bone resorption [5].

Bone Biopsy

53

Table 1. The most commonly used histomorphometric parameters Parameter Structural parameters Core width, mm Cortical width, μm Bone volume/tissue volume, %

Trabecular thickness, μm Trabecular number, /mm

Static formation parameters Osteoid thickness, μm Osteoid surface/bone surface, % Osteoid volume/bone volume, % Osteoblast surface/bone surface, % Wall thickness, μm Dynamic formation parameters Mineralising surface/bone surface, % Mineral apposition rate, μm/day Mineralisation lag time, days Bone formation rate/bone surface, μm3*μm-2*y-1 Static resorption parameters Eroded surface/bone surface, % Osteoclast surface/bone surface, %

Significance

overall size of the biopsy specimen distance between periosteal and endocortical surfaces space taken up by mineralised and unmineralised bone relative to the total size of a bone compartment self-explanatory number of trabeculae that a line through trabecular compartment would hit per millimetre of its length distance between the surface of the osteoid seam and mineralised bone percentage of bone surface covered by osteoid percentage of bone volume consisting of unmineralised osteoid percentage of bone surface covered by osteoblasts mean thickness of bone tissue that has been deposited at a remodelling site percentage of bone surface showing mineralising activity distance between two tetracycline labels divided by the length of the labelling interval time interval between the deposition and mineralisation of matrix amount of bone formed per year on a given bone surface percentage of bone surface presenting a scalloped appearance percentage of bone surface covered by osteoclasts

Bone Metabolism in Children and Adolescents

The volume of trabecular iliac bone increases markedly between 2 and 20 years of age [2]. This increase is entirely explained by trabecular thickening, whereas there is no change in trabecular number. Iliac trabeculae probably become thicker during development

54

Rauch

because bone remodels with a positive balance [6]. It has been estimated that, during a remodelling cycle, osteoblasts lay down about 5% more bone than osteoclasts resorb. In other words, 95% of the bone formation activity is required just to replace the bone that has been previously removed by the osteoclasts. Since the difference between resorption and formation is very small, a high remodelling activity is necessary to make trabeculae noticeably thicker. Remodelling activity is indeed elevated in young children, decreases until the age of 8 or 9 years, and increases again during puberty. After the age of puberty, remodelling activity declines into the much lower adult range [6].

Indications for Bone Biopsy in Paediatrics

The main use of iliac bone biopsies is to provide diagnostic clues in unclear bone fragility disorders. For example, some forms of osteogenesis imperfecta can be diagnosed on the basis of a characteristic histologic pattern [7]. Polyostotic fibrous dysplasia is sometimes difficult to distinguish from osteogenesis imperfecta on clinical grounds, but the diagnosis is usually quite obvious on bone histology. This has therapeutic implications, as children with osteogenesis imperfecta usually respond much better to bisphosphonate treatment than patients with fibrous dysplasia [7, 8]. Thus, children with multiple long-bone fractures or vertebral body compressions without adequate trauma should have a bone biopsy unless the diagnosis is obvious from non-invasive examinations. Another indication for bone biopsy is progressive bone deformity, which may sometimes arise without clear history of fractures. A bone biopsy sample allows evaluation of trabecular and cortical bone structure, the mineralisation process, bone lamellation (woven bone vs. lamellar bone, the appearance of lamellae), the presence of calcified cartilage, the activity of bone metabolism and the appearance of bone cells. All of this information is important in the assessment of skeletal disease processes, but none of it is reflected in ‘bone density’, whatever technique is used to measure it. These considerations are particularly relevant in the context of renal bone disease, which is a frequent but understudied problem after juvenile renal failure [9]. When the only aim is to assess an individual patient, it is not absolutely necessary to analyse the sample with quantitative histomorphometry. A qualitative evaluation of the histological appearance may be sufficient in such cases. However, a quantitative analysis is necessary in clinical research settings, when numbers are needed to describe the average effect of a disease or a treatment in a group of patients. Histomorphometric evaluation of bone biopsy samples should be a standard feature of studies that evaluate experimental drugs to treat bone disorders in children and adolescents. Current noninvasive methods for studying the amount, distribution, and metabolism of bone are fraught with technical limitations and uncertainties regarding the interpretation of results. The availability of histomorphometric data allows judging treatment effects in a rational way.

Bone Biopsy

55

The most important argument for performing bone biopsies in paediatric studies probably concerns patient safety. This is especially true when children and adolescents are treated with long-acting drugs, such as bisphosphonates. Analysis of bone samples provides safety measures that can not be obtained in any other way. For example, bone histologic studies have demonstrated that bisphosphonate treatment in children can lead to accumulation of calcified cartilage material in bone tissue, a disquieting finding that calls for caution in the use of these drugs in growing patients with minor skeletal symptoms [10, 11]. Thus, including bone biopsies in study protocols is crucial for documenting the efficacy of therapy as well as its safety.

Conclusions

Standard histomorphometric analysis of transiliac bone biopsies mainly provides information on trabecular remodelling. Assessment of cortical modelling processes is feasible but has rarely been used until now. When histomorphometric studies are performed in children and adolescents, it is important to take the age dependency of many histomorphometric parameters into account. When new treatments of bone disorders are studied, analysis of transiliac bone biopsy samples provides safety and efficacy data that can not be obtained in any other way. Children with multiple longbone fractures or vertebral body compressions that are not explained by adequate trauma should have a bone biopsy unless the diagnosis is obvious from non-invasive examinations.

Acknowledgements Thanks go to Mark Lepik for preparing the figures. The author is a Chercheur-Boursier Clinicien of the Fonds de la Recherche en Santé du Québec. This work was supported by the Shriners of North America.

References 1 Rauch F: Bone histomorphometry; in Glorieux FH, Pettifor J, Jueppner H (eds): Pediatric Bone. San Diego, Academic Press, 2003, pp 359–374. 2 Glorieux FH, Travers R, Taylor A, et al: Normative data for iliac bone histomorphometry in growing children. Bone 2000;26:103–109. 3 Parfitt AM, Drezner MK, Glorieux FH, et al: Bone histomorphometry: standardization of nomenclature, symbols, and units: report of the ASBMR Histomorphometry Nomenclature Committee. J Bone Miner Res 1987;2:595–610.

56

4 Parfitt AM: Renal bone disease: a new conceptual framework for the interpretation of bone histomorphometry. Curr Opin Nephrol Hypertens 2003;12: 387–403. 5 Jaworski ZF, Lok E, Wellington JL: Impaired osteoclastic function and linear bone erosion rate in secondary hyperparathyroidism associated with chronic renal failure. Clin Orthop Relat Res 1975; 107:298– 310.

Rauch

6 Parfitt AM, Travers R, Rauch F, Glorieux FH: Structural and cellular changes during bone growth in healthy children. Bone 2000;27:487–494. 7 Rauch F, Glorieux FH: Osteogenesis imperfecta. Lancet 2004;363:1377–1385. 8 Plotkin H, Rauch F, Zeitlin L, Munns C, Travers R, Glorieux FH: Effect of pamidronate treatment in children with polyostotic fibrous dysplasia of bone. J Clin Endocrinol Metab 2003;88:4569–4575.

9 Groothoff JW, Offringa M, Eck-Smit BLF, et al: Severe bone disease and low bone mineral density after juvenile renal failure. Kidney Int 2003;63:266– 275. 10 Rauch F, Travers R, Plotkin H, Glorieux FH: The effects of intravenous pamidronate on the bone tissue of children and adolescents with osteogenesis imperfecta. J Clin Invest 2002;110:1293–1299. 11 Whyte MP, Wenkert D, Clements KL, McAlister WH, Mumm S: Bisphosphonate-induced osteopetrosis. N Engl J Med 2003;349:457–463.

Frank Rauch Genetics Unit, Shriners Hospital for Children 529 Cedar Avenue Montréal, Qué., H3G 1A6 (Canada) Tel. +1 514 842 5964, Fax +1 514 842 5581, E-Mail [email protected]

Bone Biopsy

57

Chapter 5 Allgrove J, Shaw NJ (eds): Calcium and Bone Disorders in Children and Adolescents. Endocr Dev. Basel, Karger, 2009, vol 16, pp 58–72

Bone Densitometry: Current Status and Future Perspectives Nicola Crabtreea ⭈ Kate Wardb a Department of Nuclear Medicine, Queen Elizabeth NHS Foundation Trust Hospital, Birmingham, and bMRCHuman Nutrition Research, Elsie Widdowson Laboratory, Cambridge, UK

Abstract In this chapter we discuss the concept of what determines bone strength and fracture risk and how this can be quantified using current technologies. We describe bone densitometry measurement techniques that are currently available and consider the strengths and limitations of each technique, with particular relevance to paediatric scanning. Magnetic resonance imaging is reviewed as one of the newer technologies applied to the assessment of the growing skeleton. The role of dual energy X-ray absorptiometry (DXA) and quantitative computed tomography (QCT) in the clinical assessment of bone health in children is considered and current diagnostic application reviewed. Copyright © 2009 S. Karger AG, Basel

What Is Bone Mineral Density?

When describing the densitometric properties of bone the terms ‘bone mineralisation’ and ‘bone density’ are often interchanged and incorrectly used. The term ‘mineralisation’ refers to the incorporation of additional bone mineral (calcium, phosphorus and other minerals) into an existing extracellular matrix and the term ‘density’, as defined by Archimedes as the ratio of bone mass to the volume (g/cm3). Whilst there is a definite meaning to the term mineralisation, the term ‘bone density’ is less well defined and dependent on the region being assessed and the technique used. Quantitative computed tomography (QCT) measures volumetric density (BMDv) whereas dual energy X-ray absorptiometry (DXA) measures areal density (BMDa), which is not a physical density but the ratio of the total amount of bone tissue within the projected area of bone (g/cm2). The basic model of the bone describes three types of density, ‘material’, ‘compartment’ and ‘total’ all of which will have a significant function in the determination of bone strength and fracture risk (fig. 1; table 1) [1]. The three densities of bone are:

Fig. 1. Schematic representations of the ‘different’ densities of bone. Material density, calculated as the mass of bone illustrated by the grey shading divided by the volume defined by the outer pale grey border excluding Haversian canals, blood vessels and osteons (a). Compartmental density of cortical (b1) or trabecular (b2) bone density, calculated as the grey shaded mass of bone divided by the volume defined by the outer pale grey border. Total bone density calculated as the mass of both cortical and trabecular bone divided by the total volume of the bone (c). Adapted from Rauch and Schoenau [1].

(a) Material density

(b1) Cortical (b2) Trabecular compartmental compartmental bone density bone density

(c) Total bone density

Table 1. Assessment of the three ‘densities’ of bone by DXA and pQCT BMDtotal

BMDcompartment

BMDmaterial

DXA

 (estimate)

only in area of the long bone which are comprised predominantly of cortical bone (e.g. radius)

QCT/pQCT



 cortical and trabecular

the resolution required to measure BMDmaterial is not possible with current noninvasive densitometric techniques; BMDmaterial can only be determined from specimens taken at bone biopsy, which is an invasive procedure

• BMDtotal – the mineral density of all the material contained within the periosteal envelope and/or articular surfaces. • BMDcompartment – the amount of mineral contained within the trabecular, or the cortical, compartments, i.e. the mass of mineral per unit volume of trabecular or cortical bone. • BMDmaterial – the degree of mineralisation of the organic bone matrix. It is important to understand which ‘density’ each measurement technique is assessing and how it contributes to skeletal development, bone strength and fracture risk.

Bone Densitometry in Children

59

Assessment of Bone Density

The rationale for measuring bone strength in childhood is multi-factorial but can be broadly grouped into the following areas: • Assessment of the impact of chronic disease in childhood. • Monitoring the effect of modifiable factors in healthy growing children (e.g. physical activity, nutrient intake, etc.). • Predicting current and future risk of fragility fracture. Ideally, the measurement technique would be able to measure bone strength directly or, alternatively, important components of bone strength, i.e. bone mass/density, bone material and geometric properties. The technique would not be affected by growth, be safe and be readily performed in all children. Several techniques for measuring bone have been used over the past few decades, the most common of which are the technologies developed using X-rays (ionising radiation) which assess bone mass and bone density in vivo. These are DXA and QCT (both axial and peripheral).

Measurement Principles

The measurement of bone density and bone mass by ionising radiation is based upon the differential attenuation of the X-radiation beam through different tissues of the body. As X-rays pass through the body they will be attenuated (reduced in strength). The extent of attenuation varies with the energy of the photons and the nature and depth of the material they pass through. As such, the proportion of detected X-rays relates directly to the attenuation values and the area density of the body tissue through which it has passed i.e. bone or soft tissue. Using image processing techniques an attenuation/density map can be converted to a grey scale to produce an image which, with edge detection algorithms, can be used to calculate parameters such as bone size, bone area cortical width, etc. The most widely accessible and minimally invasive technique currently available to measure paediatric bone health is dual energy X-ray absorptiometry (DXA) which has the ability to measure bone mass and BMDatotal, both of which have been demonstrated to be associated with fracture risk [2].

Dual-Energy X-Ray Absorptiometry

Dual-energy X-ray absorptiometry (DXA) has been available since the late 1980s when it was principally introduced to measure and monitor the course of osteoporosis in post menopausal women [3]. Since the introduction and increased availability of DXA, there has been a dramatic rise in its use in paediatric research and clinical practice. The fundamental principle of DXA is the measurement of transmission of X-rays, produced from a stable X-ray source, at high and low energies. The two

60

Crabtree · Ward

Fig. 2. Lumbar spine and whole body DXA scan of a 9-year-old child with a history of multiple low trauma fractures.

energies allow the discrimination of soft tissue and bone and, by calculating the mass attenuation of the two materials and using sophisticated edge detection techniques, it is possible to measure BMDatotal, bone mineral content (BMC) and the projectional area of the bone. Modern-day DXAs achieve fast, precise measurements with low levels of radiation (0.1–6 μSv) and reasonable image resolution, all of which are of utmost importance when measuring children [4] (fig. 2). DXA measurements can be made at the spine, hip, whole body and forearm. For clinical practice in children, the spine and whole body less head are recommended sites. This chapter focuses on the whole body and spine as these are currently the most common sites used clinically. Although DXA is the most commonly employed method for the assessment of bone health, it does have several limitations, which are particularly relevant to the measurement of children. These include: • size dependence of BMDatotal, • inability to separate trabecular and cortical compartments, • inaccuracies due to changes in body composition,

Bone Densitometry in Children

61

2 × 2 × 2 cm

3 × 3 × 3 cm

Projected area

Bone mineral content Volume Projected area Volumetric density Areal density

Projected area

=8g = 8 cm3 = 4 cm2 = 1 g/cm3 = 2 g/cm2

Bone mineral content Volume Projected area Volumetric density Areal density

= 27 g = 27 cm3 = 9 cm2 = 1 g/cm3 = 3 g/cm2

Fig. 3. The two cubes represent vertebrae of different sizes and demonstrate the size dependence of areal BMD. The cubes have identical volumetric densities (1 g/cm3) but the smaller cube has considerably lower areal density than the larger cube on the right.

• software being designed for use in adults, • limited reference data. Additionally, there may be practical difficulties in acquiring DXA scans in young children, or children with marked learning difficulties as they may not be able to stay still for the duration of the scan [5]. The most significant of the limitations is that DXA is a projectional technique whereby a three-dimensional object is analysed using a two-dimensional projection. As mentioned, the attenuation of the X-rays is a function of the nature and depth of the material they pass through. Consequently, X-rays will be more highly attenuated in a large bone due to them travelling through more bone; therefore they will appear denser than a smaller bone with the same physical density (fig. 3). A direct consequence of the inability to measure bone depth is that the BMDatotal value calculated by DXA, is highly dependent on bone size [4, 6]. The consequence of the size dependence of BMDatotal means that DXA will inherently underestimate bone density in a short child, with smaller bones, and overestimate bone density in a tall child, with bigger bones, despite the fact that they may have identical volumetric densities. This technical problem can have serious consequences when measuring and monitoring children with chronic diseases, where the disease has also had an impact on their normal growth and development or their tempo of skeletal maturation, resulting in them being either tall or, more frequently, small for their chronological age. To avoid misdiagnosis of reduced BMDatotal due to variations in body stature rather than genuine deficiencies in bone mass, reference values for bone mineral density in children should either be independent of bone size,

62

Crabtree · Ward

if presented according to age, or, alternatively, account for the variations in bone and body size, to reduce any measurement ambiguity. Several size adjustment techniques have been developed and are listed below.

Volumetric Bone Density for Age

The most frequently used method of size adjustment in paediatric DXA is the calculation of bone mineral apparent density (BMAD) or estimated volumetric bone density. At the lumbar spine, BMAD is calculated by using either vertebral width or projected area from the DXA scan to estimate the vertebral depth and thereby enabling calculation of volumetric BMD (g/cm3) [6, 7]. The relative size independence of this parameter makes this a useful marker for reduced absolute bone density and, along with BMD for age, it has also been shown to be related to fracture risk in children [8]. This density is sometimes inappropriately referred to in the literature as ‘true bone density’. Currently, this method of size adjustment is not feasible for the whole body DXA scan.

Bone Mineral Content for Height

Reporting BMC for height is the simplest of all of the size adjustment methods since it requires no assumptions about bone size. Currently, this size adjustment has not been related to fracture risk, although it has been shown to be useful in comparing populations with diminished stature. It also correlates with the estimated bone strength measured by pQCT in children [9].

Allometric Approach

The ‘Mølgaard’ model is a simple allometric approximation that provides a threestage assessment to explain reduced bone density. The three-stage model assesses height for age (short bones), bone area for height (narrow bones), and BMC for bone area (light bones) [10]. The important diagnostic value of this is that it provides both geometric and densitometric information to give a better understanding of the bones underlying fragility. For example, thin gracile bones in cerebral palsy would have reduced strength due to reduced mineralisation and size.

Regression Models

‘Size-adjusted BMC’, is calculated using a regression, or a multivariate, statistical model to adjust BMC for confounders, such as projected BA, overall body height,

Bone Densitometry in Children

63

weight and pubertal stage [11–13]. This approach is often used in research studies. However, it should be borne in mind that body height and weight might not completely control for all relevant differences in size and shape of the skeletal region of interest. Furthermore, differences in bone size and shape may have important implications for bone strength, independently of adjusted or unadjusted BMC/ BMDatotal.

Mechanostat Functional Model

The mechanostat, or functional model, uses an alternative approach to size adjustment based on the relationship between muscle and bone [14]. The two-stage algorithm proposed by Schoenau et al. [15] was first applied to pQCT and then extended to DXA. Both techniques are based on the assumptions that BMC acts as a surrogate for bone strength and lean body mass as a surrogate for muscle load [16]. From this, four differential diagnoses can be made which potentially relate to risk of fracture. The two stages of assessment are: (1) whether the child has sufficient muscle for their height, and (2) whether they have sufficient bone for that muscle. This leads to four outcomes: (1) ‘Normal’ appropriate muscle mass for height and appropriate bone mass for muscle mass. (2) Primary bone defect, where the child has sufficient muscle for height but insufficient bone mass for muscle. (3) Primary muscle defect, where the child has reduced muscle for height but sufficient bone for muscle mass. (4) A mixed muscle and bone defect, where muscle and bone are both reduced. Currently, there is no consensus as to the best method of size adjustment, or which will best predict current or future fracture risk, or whether any of the different size adjustment techniques can improve the diagnostic capability of DXA. However, in a child with short stature, any of the size adjustment techniques will improve the diagnostic specificity of DXA and reduce the possibility of misdiagnosis due to body size [17].

Diagnostic Potential of Dual-Energy X-Ray Absorptiometry

The diagnostic potential of DXA was recently highlighted in a large prospective fracture study [18] and many other case control studies [2, 8, 19, 20]. The ISCD has recently produced recommendations for the clinical use of DXA as part of a comprehensive skeletal health assessment in patients with increased fracture. The guidelines recommend using DXA to measure the lumbar spine and whole body (less head) in patients

64

Crabtree · Ward

with primary bone diseases or potential secondary bone diseases, and that the term ‘osteoporosis’ should not be used in context of densitometry measurements alone. They also recommend that results should be size adjusted in children of reduced stature and reported as ‘low BMC or bone mineral density for chronological age’ if the BMC or BMDatotal Z-score is less than or equal to –2.0’ [21]. However, bone assessment by DXA is only part of a comprehensive skeletal assessment and, as such, the DXA results should be combined with a full clinical history and any other tests carried out.

Quantitative Computed Tomography

The main limitation of DXA in children is using information gained from a 2-dimensional projection to assess a 3-dimensional structure. Consequently, a measurement technique that collects information in all dimensions may potentially overcome this limitation. The demand for such information in paediatric bone density assessment has led to a renewed interest in computed tomography, particularly peripheral computed tomography. The main principle of CT measurements are that the linear X-ray absorption coefficient are transformed into a CT number (Hounsfield Units) which are reconstructed to form the CT image. The CT number is then transformed to BMD and geometry is calculated using image processing. For standalone CT scanners a bone equivalent phantom is required for quantification of mean volumetric BMDcompartment (mg/cm3) from the image. QCT can be applied to axial or peripheral skeletal sites. Radiation exposure is much lower in pQCT than axial QCT [5]. The most important feature of QCT is that it provides size-independent measures of volumetric BMDtotal or BMDcompartment of the trabecular and cortical bone. The trabecular BMDcompartment measured by QCT is a composite of the amount of bone and marrow per voxel. The reason for this is the relatively small size of trabeculae compared to the voxel, resulting in marrow being included in the measurement (fig. 4). As trabecular bone is generally more metabolically active than cortical bone, trabecular BMDcompartment, as measured by QCT, is likely be more sensitive to change than cortical BMDcompartment measurements [22]. QCT quantifies other important features of bone strength and measures bone size and geometry. It also provides estimates of in vivo bone strength, which relate well to fracture load [23]. More recently, high-resolution techniques (HR-CT) allow in vivo finite element modelling to be performed and to study the loading conditions to which the bone is subjected, which may improve the fracture prediction of the technique. To date, this has only been applied in adults [24, 25]. Axial QCT of the spine was first described in the late 1970s [26], and became more widely used during the 1980s [22] until DXA was introduced in 1988. The original body CT scanners used rotate-translate technology and permitted only 2D slices to be obtained and the procedure took about 15 min. The evolution of spiral and multi-

Bone Densitometry in Children

65

Fig. 4. Lumbar spine QCT and calibration block.

slice CT now allows the rapid, precise acquisition of volumetric scans of the spine, hip and peripheral sites. In a scan time as little as 10 s, a volume of data can be acquired. Quantitative skeletal assessment does not require the image quality required for conventional CT and therefore radiation dose can be minimised through use of a low dose technique [27]. Gathering a block of data makes the technique particularly useful in longitudinal studies as it make it easier to relocate the previous scan site. QCT has been applied in research studies to monitor response to intervention in central and peripheral sites [28, 29]. The limitations of axial QCT include: • approximately 10- to 12-fold greater dose of ionising radiation than DXA for spine scans, • the demand for CT equipment is much greater than for DXA, • specialist bone equivalent phantoms and software are required to perform bone measurements, • there is a lack of commercial analysis packages, • there is only one, small published reference dataset for paediatric use [30]. The high radiation dose but inherent advantages of QCT led to the development of stand alone pQCT scanners. Peripheral QCT first became commercially available in the early 1990s [31], the most commonly used technique is the single slice pQCT. The sites of measurement are the radius, tibia and femur and the radiation dose is much lower than axial QCT. For single slice techniques this is 200

Fully replete Replete Insufficient Deficient Seriously deficient

70–200 50–70 30–50 15–30 18 months. • School – are adaptations currently in place or needed?

Initial Examination From the top down look for: • Brachycephaly and occipital droop suggesting altered cranio-cervical junction anatomy and basilar invagination. • Large anterior fontanelle and sutural diastasis. • Blue scleral hue [17]. • Dentinogenesis imperfecta. • Ligamentous laxity – use the Beighton scale [18] for consistency, but also check for flat feet and genu valgum. • Muscle strength is also reduced in OI. This can be assessed clinically or may be quantified by the measurement of grip strength using a dynamometer.

Primary Osteoporosis

159

• Skin elasticity – may be reduced; note cross-over features with Ehlers-Danlos syndrome, some of whom have mutations in COL1A1 or 2 [19–21]. • Limb deformity – the degree of shortening and bowing reflects disease severity – forearm pronation and supination are reduced in OI V – check legs for discrepancy in length – assess mechanical axis – consider coax vara. • The back should be assessed standing (where possible) and bending forward for scoliosis and kyphosis. A ‘flat spot’ in the otherwise smooth curve of the spinous processes may be associated with underlying crush fracture, but is not an invariable sign. • Range of movement and strength are major determinants of both activities of daily living and the need for care-giver assistance and require expert assessment (usually by specialised therapists rather than doctors).

Differential Diagnosis In infancy, other causes of bone fragility leading to unexplained fractures include metabolic bone disease of prematurity, rare inherited metabolic diseases (e.g. I-cell disease) and non-accidental injury (NAI). Severe bone disease is radiologically apparent. The most difficult differentiation is between mild OI and NAI [22]. In older children, once malignancy, endocrine and inflammatory conditions are excluded, the diagnosis of idiopathic juvenile osteoporosis can be considered. Some of these children present in the classical fashion described by Dent et al. [23], but some present without the metaphyseal fractures and neo-osseous ossification features. Children with recurrent fractures and a bone mass that is more than two standard deviations below that predicted for body size are now defined as having osteoporosis by the ISCD criteria [24].

Investigation There are no definitive biochemical or imaging biomarkers for OI. However, bone turnover is typically elevated in the untreated state even in the absence of recent fracture. Plain X-rays may be used to assess the mechanical axis, the degree of bowing deformity and any coxa vara, the presence of vertebral crush fractures, scoliosis and spondylolisthesis in the spine. Wormian bones in the lambdoid suture may be normal; in the sagittal suture they are a strong indicator of skeletal abnormality. DXA scans can quantify the amount of mineralised bone at a given skeletal site – typically the spine, hip or whole body – and provide a means to assess some aspects of

160

Bishop

Table 1. Different types of OI and related conditions, together with the principal features, gene mutations (where known) and modes of inheritance: OMIM number for each condition and its associated gene is also shown OI type

OMIM

Phenotype (during childhood)

Genetic origin

Inheritance

OMIM

Chromosome location

I

#166200

Mild motor delay. Bowing of long bones. Vertebral crush fractures. Ligamentous laxity, hernias, mixed conductive/ sensorineural deafness, blue sclerae. Subdivided on the basis of the presence or absence of dentinogenesis imperfecta (A = absent, B = present).

Typically null allele of COL1A1, resulting from stop, frameshift or splice site mutations

AD

+120150

17q21.31-q22

IIA

#166210

Lethal. Subdivided by appearance of ribs.

Missense mutations in COL1A1 or COL1A2

AD +120150 *120160

17q21.31-q22 7q22.1

Complete loss of CRTAP

AR

*605497

3p22

+120150 *120160

17q21.31-q22 7q22.1

IIB

#610854

Lethal. Similar to type IIA.

III

#259420

Severe, progressively Missense mutations in deforming. Typically COL1A1, COL1A2; null have fractures in utero, allele of COL1A2 very poor post-natal growth. Characteristic facies with small mid face and pointed chin. Triangular facial appearance less noticeable with bisphosphonate treatment. Very delayed motor development. Almost all need intramedullary rodding. Blue sclerae remain. All have dentinogenesis imperfecta.

Primary Osteoporosis

AD usually new mutation or parental mosaicism

161

Table 1. Continued OI type

OMIM

Phenotype (during childhood)

IV

#166220

V

Inheritance

OMIM

Chromosome location

Moderately severe. Missense mutations in May have fractures in COL1A1, COL1A2 utero, but better postnatal growth than type III. Blue sclerae fade with age; may have dentinogenesis imperfecta

AD

+120150 *120160

17q21.31-q22 7q22.1

%610967

Moderately severe. Unknown Metaphyseal sclerosis in early life, followed by calcification of interosseous membranes in the forearm and lower leg. Characteristic bowing of the forearms. Hypertrophic callus formation following fractures and surgery.

AD

?

?

VI

%610968

Severe, progressively Unknown deforming. Osteomalacic on bone biopsy, possibly as a result of abnormal matrix deposition – normal lamellar structure is disrupted.

Possibly AR

?

?

VII

#610682

Moderately severe. Rhizomelic in both arms and legs; femurs and humeri are very bowed. White sclerae

Cryptic splice site in intron 1 of CRTAP

AR

*605497

3p22

VIII

#610915

Very severe/lethal. Round face, white sclerae, thin ribs (may be beaded). Most cases are reported in children whose families originate from West Africa, Pakistan and Ireland.

Deletion of LEPRE1

AR

*610339

1p34

162

Genetic origin

Bishop

Table 1. Continued OI type

OMIM

Phenotype (during childhood)

Genetic origin

Inheritance

OMIM

Chromosome location

Bruck 1

%259450

Contractures. White sclerae, mild DI. Moderately-severe bone disease.

Maps to 17p12

AR

?

17p12

Bruck 2

#609220

Clinical phenotype as for Bruck 1.

PLOD2; bone-specific telopeptide lysyl hydroxylase

AR

*601865

3q23-q24

Normal at birth; develop craniosynostosis, ocular proptosis, hydrocephalus and diaphyseal fractures.

Unknown

Unknown

?

?

112240 ColeCarpenter

therapeutic intervention. However, it is clear that the scans also measure calcified cartilage and, as such, may not accurately represent the bony response to therapy. QCT both of the vertebrae and at peripheral sites has been reported but remains a research tool at present (see chapter 5 for further details). MRI scans may be useful to detect crush fractured vertebrae in younger children whose vertebrae are difficult to visualise accurately using plain X-rays. MRI and CT may be required to visualise cranio-cervical junction anatomy in cases of basilar invagination, and to define any associated syrinx. Bone biopsy is undertaken either where there is difficulty in determining the underlying diagnosis [25] or, in some centres, before starting bisphosphonate therapy. If performed, double-labelling of bone using two brief (two day) courses of a tetracycline ten days apart will provide the opportunity to evaluate dynamic as well as static parameters of bone activity [26] (see chapter 4 for a more detailed discussion). Where mild OI is the differential diagnosis in unexplained fractures in infancy, a court may request genetic testing. Such testing is available at a number of centres worldwide. In the UK it comprises direct DNA sequencing with multiple ligationdependent probe amplification to detect large scale deletions and duplications. The detection rate for COL1A1 and COL1A2 mutations using this approach is thought to be >99%. However, it is important to inform those requesting the test that a significant proportion (10%) of individuals with the OI phenotype are negative on testing of COL1A1/2.

Primary Osteoporosis

163

Management The key to effective management in OI is the multidisciplinary approach. The initial consultation includes not only an assessment of severity and the need for medical therapy such as a bisphosphonate, but also of the functional and social needs of the child and their family. A comprehensive team should include specialist nursing, physiotherapy and occupational therapists, pain management, and input from orthopaedic and dental colleagues. Input from therapists is highly individualised according to patient need, as is nursing input. There is typically an important need for education of families concerning the pathology of the disease, the effects of intervention and likely medium to long term outcomes. The whole team is involved in this process, and families should also be directed to national (in the UK the Brittle Bone Society) and international organisations (the OI Foundation in the USA/Canada and OIFE in Europe). Managing severely affected infants can be particularly challenging. Such infants may have compromised respiratory function at birth because rib fractures in utero lead to lack of chest expansion and pulmonary hypoplasia, in addition to the impaired mechanics of the rib cage post-natally. In some severely affected infants, respiratory support, ranging from nasal oxygen to mechanical ventilation, is required. Infants with pre-existing respiratory compromise may deteriorate when infused with pamidronate [27] and should be carefully monitored in a setting where there is access to high dependency or intensive care. We have seen one infant in highoutput cardiac failure, probably secondary to extensive healing fractures. Frequent physiotherapy and occupational therapy input is required in order to enable the child to progress safely through their gross motor milestones. Side-lying initially of the head and subsequently the whole body may help maintain head shape and prevent severe brachycephaly. This in turn will help maintain normal cranio-cervical junction anatomy and normal orbital shape. Infants should be turned regularly and enabled to spend time prone where tolerated. For transportation the ideal car seat is one that allows maximum recline and can fit into the chassis of a pushchair in order to minimise handling. Careful management of positioning is required in severely affected infants in order to enable head control, sitting and crawling to proceed safely. Aids to mobility include orthoses, braces and appropriate wheelchair support. Therapists often start at the foot and work up in terms of providing aids to mobility. Even mildly affected children may benefit from insoles and supportive footwear. In more severely affected children, hip-knee-ankle-foot orthoses (HKAFOs) are often considered initially for children under five years of age who are ready to stand and walk. Once some pelvic control is achieved then a walking aid may be introduced. In the school setting particularly, joint hypermobility can impair the development of fine motor skills. Modified pencil holders for writing and wrist splints to reduce fatigue can be helpful.

164

Bishop

As with many chronic diseases, OI has a significant impact on daily life. How one is treated by others means that the psychosocial aspects to management are important to address, especially during adolescence. Our observation has been that, despite their often severe disability, children and adolescents with OI are often bright and sparky individuals with a large number of friends. Nevertheless, support and advice from the multidisciplinary team along with physical adaptations and other aids may be required to ensure integration into school. Concern may be expressed by schools about participation in sports and individual advice is often required. On-going monitoring for all children is required in a number of areas. Our practice is to obtain annual DXA and lateral spine films to evaluate bone accretion and check for occult vertebral crush fracture in mildly affected children not currently receiving bisphosphonate therapy. The frequency of review is dictated by the severity of the disease and the age of the child. We review severely affected infants six times in the first year; older severely affected children are typically seen four times each year, coinciding with their intravenous bisphosphonate therapy. For those on therapy, DXA is undertaken six monthly to evaluate the response to treatment. Therapy input may be required between clinic visits, tailored to individual need. All children with dentinogenesis imperfecta need annual dental review; some are seen more frequently if they have accelerated carious deterioration, or need specific intervention such as crowns, veneers or cosmetic work. Hearing should be evaluated every three years from age ten years so as to identify the minority of children (13 years old with testes 12 years with no breast development

Ht