Current Topics in Developmental Biology

Series Editor Gerald P. Schatten Director, PITTSBURGH DEVELOPMENTAL CENTER Deputy Director, Magee-Women’s Research Institute Professor and Vice-Chair of Ob-Gyn-Reproductive Sci. & Cell Biol.-Physiology University of Pittsburgh School of Medicine Pittsburgh, Pennsylvania 15213

Editorial Board Peter Gru¨ss Max-Planck-Institute of Biophysical Chemistry Go¨ttingen, Germany

Philip Ingham University of Sheffield, United Kingdom

Mary Lou King University of Miami, Florida

Story C. Landis National Institutes of Health National Institute of Neurological Disorders and Stroke Bethesda, Maryland

David R. McClay Duke University, Durham, North Carolina

Yoshitaka Nagahama National Institute for Basic Biology, Okazaki, Japan

Susan Strome Indiana University, Bloomington, Indiana

Virginia Walbot Stanford University, Palo Alto, California

Founding Editors A. A. Moscona Alberto Monroy

Contributors Numbers in parentheses indicate the pages on which the authors’ contributions begin.

Grant W. Anderson (135), College of Pharmacy, University of Minnesota, Duluth, Minnesota 55812 Christian Braendle (171), Institut Jacques Monod, CNRS-Universities of Paris 6/7, Tour 43, 2 Place Jussieu, 75251 Paris Cedex 05, France Peter Carmeliet (1), Department of Transgene Technology and Gene Therapy, VIB, Leuven, Belgium and Department of Transgene Technology and Gene Therapy, K. U. Leuven, Leuven, Belgium Jan DeNofrio (311), Curriculum in Genetics and Molecular Biology, University of North Carolina, Chapel Hill, North Carolina 27599 Richard D. Egleton (277), Department of Pharmacology, Physiology and Toxicology, Joan C. Edwards Medical School, Marshall University, Huntington, West Virginia 25755 Marie-Anne Fe´lix (171), Institut Jacques Monod, CNRS-Universities of Paris 6/7, Tour 43, 2 Place Jussieu, 75251 Paris Cedex 05, France Galina N. Filippova (337), Human Biology Division, Fred Hutchinson Cancer Research Center, Seattle, Washington 98109 Brian T. Hawkins (277), Laboratory of Pharmacology and Chemistry, National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina 27709 Jun Liu (311), Departments of Biochemistry and Biophysics, and Pharmacology, University of North Carolina, Chapel Hill, North Carolina 27599 Gilles Luquet (209), UMR CNRS 5561 ‘Bioge´osciences,’ Universite´ de Bourgogne, 6 Boulevard Gabriel, 21000 DIJON, France Benjamin Marie (209), UMR CNRS 5561 ‘Bioge´osciences,’ Universite´ de Bourgogne, 6 Boulevard Gabriel, 21000 DIJON, France Fre´de´ric Marin (209), UMR CNRS 5561 ‘Bioge´osciences,’ Universite´ de Bourgogne, 6 Boulevard Gabriel, 21000 DIJON, France Andrew W. McFadden (311), Department of Pharmacology, and Lineberger Comprehensive Cancer Center, University of North Carolina, Chapel Hill, North Carolina 27599 ix

x

Contributors

Davorin Medakovic (209), Center for Marine Research Rovinj, Ruder Boskovic Institute, 5 Giordano Paliaga, 52210 ROVINJ, Croatia Josselin Milloz (171), Institut Jacques Monod, CNRS-Universities of Paris 6/7, Tour 43, 2 Place Jussieu, 75251 Paris Cedex 05, France Jackob Moskovitz (93), Department of Pharmacology and Toxicology, School of Pharmacy, University of Kansas, Lawrence, Kansas 66045 Dolores D. Mruk (57), Center for Biomedical Research, Population Council, New York, New York 10021 Derek B. Oien (93), Department of Pharmacology & Toxicology, School of Pharmacy, University of Kansas, Lawrence, Kansas 66045 Leslie V. Parise (311), Departments of Biochemistry and Biophysics, and Pharmacology, and Carolina Cardiovascular Biology Center, and Lineberger Comprehensive Cancer Center, University of North Carolina, Chapel Hill, North Carolina 27599 Timothy P. Rich (135), Department of Medicine, University of Minnesota, Duluth, Minnesota 55812 Carmen Ruiz de Almodovar (1), Department of Transgene Technology and Gene Therapy, VIB, Leuven, Belgium, and Department of Transgene Technology and Gene Therapy, K. U. Leuven, Leuven, Belgium Jon N. Rumbley (135), Chemistry and Biochemistry, University of Minnesota, Duluth, Minnesota 55812 David R. Salo (135), College of Pharmacy, University of Minnesota, Duluth, Minnesota 55812 Zhengyan Wang (311), Clinical Research Scholar Program, School of Dentistry, University of North Carolina, Chapel Hill, North Carolina 27599 Daniel E. Westholm (135), College of Pharmacy, University of Minnesota, Duluth, Minnesota 55812 C. Yan Cheng (57), Center for Biomedical Research, Population Council, New York, New York 10021 Helen H. N. Yan (57), Center for Biomedical Research, Population Council, New York, New York 10021 Weiping Yuan (311), Departments of Biochemistry and Biophysics, and Pharmacology, University of North Carolina, Chapel Hill, North Carolina 27599 Serena Zacchigna (1), Department of Transgene Technology and Gene Therapy, VIB, Leuven, Belgium and Department of Transgene Technology and Gene Therapy, K. U. Leuven, Leuven, Belgium

Preface It is my great pleasure to announce that eVective from the next volume, Current Topics in Developmental Biology will benefit from the insightful partnership of two truly superb scholars as the next set of coeditors: Paul Wassarman from the Mount Sinai School of Medicine and Olivier Pourquie´ from the Stowers Institute for Medical Research have agreed to serve as coeditors. Together they provide comprehensive and complementary perspectives on the most innovative research breakthroughs and exciting newest conceptual and technological directions in developmental biology. This transition will ensure that Current Topics in Developmental Biology remains an invaluable serial for our scientific community and that it will continue to prosper as the longest-running public forum for contemporary issues in developmental biology. In 1966, now over forty years ago, Current Topics in Developmental Biology was launched by the inspired partnership of Alberto Monroy and Aaron Moscona. They established Current Topics in Developmental Biology with the tradition of coeditors within the Academic Press family—that marvelous enterprise for scholarly publications within Elsevier. I, too, enjoyed that sibling relationship when Roger Pedersen (Cambridge University) invited me to join him as coeditor from 1989 to 1999 starting with volume 30. Current Topics in Developmental Biology continues to hold many unique distinctions including its now historic reputation in our field of developmental biology of providing up-to-date scholarly reviews encompassing the entire breadth and depth of research in all aspects of our understanding the expanding field. I am also pleased to report that also eVective with the next volume, the Society for Developmental Biology (SDB) has agreed to incorporate Current Topics in Developmental Biology, together with the journal Developmental Biology, as their society-sponsored publications. This relation further solidifies the complementary benefits between the membership of the SDB, SDB’s leadership, and Academic Press and its publisher Elsevier. I’m grateful to everyone involved in this constructive tradition— especially SDB Past President Gail Martin (University of California, San Francisco), SDB President Eric Wieschaus (Princeton), SDB President Elect Marianne Bronner-Fraser (Cal Tech), SDB Executive OYcer Ida Chow, and former Academic Press/Elsevier’s Jasna Markovac. Current Topics in Developmental Biology has grown and stayed abreast of the enormous explosions in developmental biology largely because our scientific community has participated actively in both writing important and xi

xii

Preface

articulate reviews, as well as in subscribing to the series. These volumes have benefited from the ongoing cooperation of a team of participants who are jointly responsible for the content and quality of its material. The authors deserve the full credit for their success in covering their subjects in depth yet with clarity, and for challenging the reader to think about these topics in new ways. Volume 80 of Current Topics in Developmental Biology is remarkable in many ways, including the breadth and depth of its consideration of various timely challenges in development. From spermatogenesis and male contraceptive considerations, through diVerentiation of neurons and blood vessels, and also platelets; cellular regulation by posttranslational modifications, ion transporters and channels, environmental factors, cell surface signaling, and junctional communications; and tissue interactions with the extracellular environment ranging from molluscan shell formation, the blood–brain barrier in the brain as well as in the testes, volume 80 provides the reader up-to-date reviews on emerging topics in development. Conceptual problems including gene–environment interaction, to epigenetics, to cell surface and extracellular function and dynamics, are addressed; and this volume is exceptional for its ‘‘micro’’ approach to ‘‘macro’’ issues as well as its consideration of the ideal experimental system in which to investigate contemporary issues. Developmental biologists are fortunate to enjoy a rich and important scientific history now a few centuries old, culminating here in outstanding expositions on our science. On a personal note, it has been a distinct privilege to serve as the editor of Current Topics in Developmental Biology since 2000. My contributions to Current Topics in Developmental Biology have indeed been a tremendous pleasure because of the stellar foundations built by the Founding Coeditors Alberto Monroy and Aaron Moscona, as well as Roger Pedersen and the numerous editors of special thematic volumes. I am pleased to oYcially thank the members of the Editorial Board: Peter Gruess at the Max Planck Institute for Biophysical Chemistry at Goettingen; Philip Ingham at the University of SheYeld; Mary Lou King from the University of Miami and Story Landis at the National Institutes of Health; David McClay from Duke University and Yoshitaka Nagahama from the National Institute for Basic Biology in Okazaki; and Susan Strome from Indiana University and Virginia Walbot from Stanford University. Also, I am grateful to everyone at the Pittsburgh Development Center of Magee-Womens Research Institute and in the Division of Developmental and Regenerative Medicine in the Department of Obstetrics, Gynecology and Reproductive Sciences here at the University of Pittsburgh School of Medicine for providing intellectual and infrastructural support for Current Topics in Developmental Biology. Finally, I also would like to oVer a special word of appreciation to everyone at Academic Press and Elsevier—and most especially Cindy

Preface

xiii

Minor, who together with Jasna Markovic, have been the enduring heart and soul of Current Topics in Developmental Biology. Lastly, I thank my children, Samantha, Madeline, and Daniel, who have taught me firsthand about development


Jerry Schatten Pittsburgh Development Center, Pennsylvania July 2007

Contents

Contributors Preface xi

ix

1 Similarities Between Angiogenesis and Neural Development: What Small Animal Models Can Tell Us Serena Zacchigna, Carmen Ruiz de Almodovar, and Peter Carmeliet I. II. III. IV. V. VI. VII.

Introduction 2 Small Animal Models to Study Blood and Vessel Guidance 4 Vascular and Neural Cell-Fate Specification 20 Molecular Links Between Angiogenesis and Neurogenesis 22 Similarities in the Organization of Vascular and Neural Boundaries Molecular Cues Involved in Nerve and Vessel Guidance 26 Perspectives 41 Acknowledgments 42 References 42

24

2 Junction Restructuring and Spermatogenesis: The Biology, Regulation, and Implication in Male Contraceptive Development Helen H. N. Yan, Dolores D. Mruk, and C. Yan Cheng I. II. III. IV. V.

Introduction 58 Anchoring Junctions in the Testes: An Update 60 Roles of ECM Proteins in Junction Dynamics in the Testes 63 Role of Androgens in Junction Dynamics in Testes 72 Regulation of Junction Turnover by Protein Endocytosis and Recycling 76 VI. Regulation of Junction Dynamics by Myoid Cells 79 VII. Environmental Toxicants: Are They Targeting the Tight and/or Anchoring Junction? 81

v

vi

Contents VIII. Concluding Remarks 82 Acknowledgments 83 References 83

3 Substrates of the Methionine Sulfoxide Reductase System and Their Physiological Relevance Derek B. Oien and Jackob Moskovitz I. II. III. IV. V.

Introduction 94 Regulated Substrates 96 Scavenging Substrates 103 Modified Substrates with ‘‘Damaged’’ EVects Discussion 123 References 125

109

4 Organic Anion-Transporting Polypeptides at the Blood–Brain and Blood–Cerebrospinal Fluid Barriers Daniel E. Westholm, Jon N. Rumbley, David R. Salo, Timothy P. Rich, and Grant W. Anderson I. II. III. IV. V. VI. VII. VIII. IX. X. XI.

Introduction 136 BBB Structure and Function 136 BCSFB Structure and Function 138 The OATP/Oatp Superfamily 140 Molecular Architecture of the Oatp Superfamily 142 Oatp Substrate Structural Features 145 OATP/Oatp Expression and Action at the BBB and BCSFB 148 Specific Oatps/Oatps Expressed at BBB and BCSFB 151 PG Metabolism and Oatps 155 Oatp-Mediated Transport of Conjugated Endobiotics 158 Oxidation, Conjugation, and Transport Metabolism of DHEA and Estradiol (E2) in the Brain 159 XII. Summary 163 Acknowledgments 164 References 164

vii

Contents

5 Mechanisms and Evolution of Environmental Responses in Caenorhabditis elegans Christian Braendle, Josselin Milloz, and Marie-Anne Fe´lix I. II. III. IV. V. VI. VII.

Introduction 172 Interactions Between Organism and Environment 173 The Nematode C. elegans 176 Overview of C. elegans Responses to the Environment 180 Phenotypic Plasticity of C. elegans Dauer Formation 186 Environmental Robustness of C. elegans Vulva Formation 190 Conclusion 198 Acknowledgments 198 References 199

6 Molluscan Shell Proteins: Primary Structure, Origin, and Evolution Fre´de´ric Marin, Gilles Luquet, Benjamin Marie, and Davorin Medakovic I. II. III. IV.

Introduction: The Shell, a Biologically Controlled Mineralization Molluscan Shell Formation: Developmental Aspects 212 The Topographic Models of Shell Mineralization 221 Molluscan Shell Proteins: Characterization of Their Primary Structure 229 V. Origin and Evolution of Molluscan Shell Proteins 254 VI. Concluding Remarks 262 Acknowledgments 263 References 263

210

7 Pathophysiology of the Blood–Brain Barrier: Animal Models and Methods Brian T. Hawkins and Richard D. Egleton I. II. III. IV. V.

The Blood–Brain Barrier 278 Animal-Based Methods in BBB Pathophysiology 285 BBB Dysfunction as a Complication of Peripheral Disease The BBB in Disease Etiology 295 Concluding Remarks 297 Acknowledgments 297 References 297

292

viii

Contents

8 Genetic Manipulation of Megakaryocytes to Study Platelet Function Jun Liu, Jan DeNofrio, Weiping Yuan, Zhengyan Wang, Andrew W. McFadden, and Leslie V. Parise I. II. III. IV.

Introduction 312 Culture and DiVerentiation of Megakaryocytes 313 Genetic Manipulation of Megakaryocytes 320 Current Use and Future Application of Megakaryocytes Acknowledgments 331 References 332

325

9 Genetics and Epigenetics of the Multifunctional Protein CTCF Galina N. Filippova I. History of CTCF Discovery 338 II. Multifunctional Nature of CTCF Versus Its Multiple Sequence Specificity 338 III. CTCF Functions in Epigenetic Regulation in Development 344 IV. CTCF Function in Chromatin Organization of Repetitive Elements V. Is CTCF a Tumor Suppressor Gene? 348 VI. Concluding Remarks 353 Acknowledgments 353 References 354 Index 361 Contents of Previous Volumes

373

346

1

____________________________________________________________________________

Similarities Between Angiogenesis and Neural Development: What Small Animal Models Can Tell Us Serena Zacchigna,* ,{,1 Carmen Ruiz de Almodovar,* ,{,1 and Peter Carmeliet * ,{ * Department of Transgene Technology and Gene Therapy, VIB, Leuven, Belgium { Department of Transgene Technology and Gene Therapy, K.U. Leuven Leuven, Belgium

I. Introduction II. Small Animal Models to Study Blood and Vessel Guidance A. Caenorhabditis elegans (Nematode Worm) B. D. melanogaster (Fruit Fly) C. Zebrafish D. Xenopus III. IV. V. VI.

Vascular and Neural Cell‐Fate Specification Molecular Links Between Angiogenesis and Neurogenesis Similarities in the Organization of Vascular and Neural Boundaries Molecular Cues Involved in Nerve and Vessel Guidance A. Axon Growth Cones and Endothelial Tip Cells B. Common Signals for Axon and Blood Vessel Wiring

VII. Perspectives Acknowledgments References

During evolution vertebrates had to evolve in order to perform more and more complex tasks. To achieve this goal, they developed specialized tissues: a highly branched vascular system to ensure that all tissues receive adequate blood supply, and an intricate nervous system in which nerves branch to transmit electrical signals to peripheral organs. The development of both systems is tightly controlled by a series of developmental cues, which ensure the accomplishment of a complex and highly stereotyped mature network. Vessels and nerves use similar signals and principles to grow, diVerentiate, and navigate toward their final targets. Both systems share several molecular pathways, highlighting an important link between vascular biology and neuroscience. Moreover, the vascular and the nervous system crosstalk and, when deregulated, contribute to medically relevant diseases. This new phenomenon, named the neurovascular link, promises to accelerate the discovery of new pathogenetic 1

Both authors contributed equally to this manuscript.

Current Topics in Developmental Biology, Vol. 80 Copyright 2008, Elsevier Inc. All rights reserved.

1

0070-2153/08 $35.00 DOI: 10.1016/S0070-2153(07)80001-9

2

Zacchigna et al.

insights and therapeutic strategies for the treatment of both vascular and neurological diseases. To study the development of both systems, scientists are taking advantage of the use of several vertebrate and invertebrate animal models. In the first part of this chapter, we will discuss the more commonly used animal models; in the second part, the striking similarities occurring during the development of the vascular and the neural systems will be revised. ß 2008, Elsevier Inc.

I. Introduction During the course of evolution, vertebrates have learned how to perform more complex and sophisticated tasks—this challenge could only be met by the coincident development of two intertwined anatomic systems, blood vessels and nerves: the former providing nutrients and the latter transmitting electrical signals required for coordination. The vital importance of these two systems was already acclaimed two millennia ago, when two major schools of thought in the ancient Greece debated about the relative role of the brain or the heart as the central source of life. The first school, headed by Plato, supported the concept that the brain harbored the soul, whereas Aristotle and his followers, in the second school, considered the heart and blood vessels of major importance for life. As Aristotle stated, ‘‘The blood vessel system can be compared to those of watercourses in gardens: they start from one source and branch oV into numerous channels, so as to carry a supply to every part of the garden.’’ Nowadays, we realize that there is no reason to consider blood vessels and nerves as antagonists, as the Greeks once thought. Instead, they have been recognized to share much more in common than originally anticipated, in terms of development, molecular mechanisms of wiring, and pathogenesis of disease. Thanks to the use of a variety of animal models we have started to dissect the molecular pathways underlying blood vessel and neuronal circuitry formation. Surprisingly, we are discovering that both blood vessels and nerves use similar signals to grow, diVerentiate, and navigate toward their final targets. Moreover, the vascular and the nervous systems have been shown to crosstalk with each other and, when deregulated, contribute to medically relevant diseases. In this chapter, we will first provide an overview of the animal models that have mainly contributed to our current understanding of the neurovascular link. We will focus on the advantages of using either invertebrate (such as worms and flies) or vertebrate (like fish and frogs) models to study the development and function of nerves and blood vessels. Then, we will focus our attention in the parallelism between neurogenesis and vasculogenesis, as well as between vessel and nerve growth and navigation. Finally, we will discuss our current knowledge about the molecular players certainly or possibly involved in blood vessel and axon guidance.

3

1. Angiogenesis and Neurogenesis A

STANDARD TRANSGENIC APPROACH

Transgene DNA is microinjected into the male pronucleus of a fertilized murine oocyte

B

Injected oocytes are transferred to a pseudopregnant recipient mouse

Offsprings are screened for the transgene by DNA analysis

GENE-TARGETED TRANSGENIC APPROACH

Isogenic transgene DNA is introduced into ES cells (e.g., by electroporation)

Drug selection is used and the surviving colonies are screened for the transgene

Characterized targeted cells are microinjected into 3.5-day mouse blastocyst

Blastocysts are transferred to a pseudopregnant recipient mouse

Chimeric offsprings are identified and mated to test for germ line transmission of the transgene

C wt

LacZ

Figure 1 Transgenic mice technology. (A) The standard transgenic approach is based on the microinjection of the transgene into the male pronucleus of a murine oocyte, which is then transferred to a 0.5‐day‐pseudopregnant recipient mouse. OVspring are screened for the presence of the transgene. (B) In the more recent gene‐targeted transgenic approach, isogenic DNA [to the embryonic stem (ES) cells being targeted] containing the transgene is introduced into the ES cells, for instance, by electroporation. Drug selection is used and surviving colonies are screened for the presence of the transgene. Targeted ES cells are then injected into 3.5‐day mouse blastocysts and transferred to 2.5‐day‐pseudopregnant recipient mice. The incorporation of targeted ES cells into the oVspring is determined by coat colour—that is, chimeric mice are generated that display coat color of both mice, from which either blastocysts or the ES cells were derived. Chimeric mice

4

Zacchigna et al.

II. Small Animal Models to Study Blood and Vessel Guidance Many aspects of biology are similar in most or all organisms, but it is frequently more straightforward to study particular aspects of development in specific organisms. Over the past century, the mouse has become the premier mammalian model system for genetic research. Scientists working in a wide range of biomedical fields have gravitated to the mouse because of its close genetic and physiological similarities to humans, as well as the relative ease with which its genome can be manipulated and analyzed. Mouse models currently used in biomedical research include thousands of unique inbred strains and genetically engineered mutants. There are mice prone to develop diVerent kinds of cancers, diabetes, obesity, blindness, Amyotrophic Lateral Sclerosis (ALS), Huntington’s disease, anxiety, alcoholism, and drug addiction. Innovative genetic technologies have led to the production of custom‐made mouse models for the study of a wide array of both disease and developmental processes. Undoubtedly, among the most important advances has been the possibility to create transgenic mice, in which a foreign gene is inserted into the animal’s germ line (Fig. 1). Alternative approaches, relying on homologous recombination, have permitted the development of tools to ‘‘knock out’’ genes by disrupting existing genes, even tissue specifically, or to ‘‘knock in’’ genes by altering a mouse gene in its natural location. However, these techniques, which accounted for an enormous progress in our understanding of the molecules implicated in embryonic development as well as in the pathogenesis of diVerent diseases, are extremely expensive, laborious, and time consuming. In contrast, smaller animal models, such as worms, flies, fishes, and frogs, oVer the advantage to study gene function with a much smaller budget and much shorter time frame, even at a high‐throughput scale. This is probably the reason why nonmammalian, small animal models have been extensively used in developmental biology, quickly providing useful information about gene function, and pioneering medical research to define novel therapeutic entry points. Indeed, genetic studies are easier in small animals and much more complex in mammals. With the fact that the genomes from many diVerent species are already sequenced, there is currently a general eVort to functionally link genes from model organisms to their counterparts in humans, in order to expand their utility for understanding biological processes and diseases. In this perspective, only after confirmation of the importance of one or more candidate genes in the small animal models, they will ultimately be evaluated in the mouse or other larger animals. are mated to determine whether the targeted ES cells have contributed to the germ line. Germ line oVspring are finally screened for the presence of the transgene and mated to establish the transgenic line. (C) The expression of the LacZ reporter gene, whose protein product ‐galactosidase can be detected histochemically by X‐gal staining, is evident in the LacZ transgenic mouse (right) but not in the wilt type (wt) mouse (left).

1. Angiogenesis and Neurogenesis

5

In the following section, we will discuss the major advantages of the most popular small animal models used in the study of angiogenesis and neurobiology, together with the techniques that can be more suitably applied to each model. However, it is important to mention that reliance of one model can result in several experimental limitations, and can even distort our view of the problem. For instance, some molecular pathways simply do not exist in smaller animals, and thus cannot be studied there. In addition, small organisms usually do not recreate the entire pathophysiology of complex human diseases. Therefore, it is critical that diVerent models are considered at the same time, as none of them can address all issues. Indeed, these models are complementary, and only an integrated view of the information obtained by the use of diVerent animals, further supplemented by human genetics, is likely necessary for a correct understanding of any biological process.

A. Caenorhabditis elegans (Nematode Worm) In the second half of the twentieth century, Sydney Brenner adopted the nematode C. elegans as a laboratory animal model with the specific purpose of studying the genetics of development, its nervous system, and its behavior. Today, C. elegans is used to study a much larger variety of biological processes, including apoptosis, cell signaling, cell cycle, cell polarity, gene regulation, metabolism, ageing, and sex determination (Kaletta and Hengartner, 2006). Several key discoveries for basic biology as well as for medically relevant areas were first made in this tiny worm. For instance, in 1993, the first presenilin gene was discovered in C. elegans (Sundaram and Greenwald, 1993), and only 2 years later, mutations in the human presenilin‐1 gene were associated with early onset of familial Alzheimer’s disease (Sherrington et al., 1995). Another example refers to type 2 diabetes. In 1997, genetic studies in C. elegans identified negative regulators of the insulin signaling pathway, among daf‐16 (the C. elegans orthologue of the forkhead transcription factor Forkhead box 01 transcription factor (FOXO) (Ogg et al., 1997)). Five years later, FOXO loss of function was found to rescue the diabetic phenotype of insulin‐resistance mice (Nakae et al., 2002). Finally, C. elegans is not only an established genetic model but can also be exploited to investigate the underlying mechanisms of whole animal pharmacology. For instance, the antidepressant fluoxetine has been shown to increase serotonergic signaling in C. elegans by inhibiting the orthologue of the serotonin reuptake transporter (SERT) (Ranganathan et al., 2001). This has stimulated a number of investigations to identify additional modes of action of antidepressants and to further elucidate the molecular mechanisms of depression. C. elegans has a number of features that make it particularly attractive for several areas of research. First, it is easy to grow as, although it usually lives in the soil and feeds on various bacteria, it can readily be raised in the

6

Zacchigna et al.

laboratory on a diet of Escherichia coli. Second, it reproduces rapidly and prolifically, developing from an egg to an adult worm of 1.3 mm within 3 days. Third, because of its small size, several assays can be carried out in microtiter plates. Fourth, C. elegans genome was sequenced completely at the end of 1998. It has about a hundred million base pairs and is surprisingly similar (40% homology) to that of humans, which renders it a particularly attractive model in the study of human development and disease. Overall, genetic research is quite straightforward. C. elegans has five pairs of autosomes and one pair of sex chromosomes, whose ratios determine the sex: if the sixth chromosome pair is XX, then C. elegans will be a hermaphrodite, while an X0 combination in the sixth chromosome pair will produce a male (in nature, hermaphrodites are the most common sex). Hermaphrodites can self‐fertilize or mate with males but cannot fertilize each other. In the laboratory, self‐ fertilization of hermaphrodites or crossing with males can be manipulated to produce progeny with the desired genotype. Finally, what is unique about the worm is that it is transparent at all developmental stages, facilitating the identification of anatomical aberrations (e.g., as a consequence of mutations), simply by inspecting the living animal. 1. C. elegans and Its Nervous System C. elegans displays an invariant lineage, with exactly 959 somatic cells, of which we know both the cellular anatomy and, in many cases also, the function (Fig. 2). Among these, there are only 302 neurons, which are classified in 118 functional classes, according to their involvement in diVerent behavioral aspects, including the sensation of mechanical, chemical, olfactory, and thermal stimuli; the movement pattern during mating; and, quite surprisingly for such a scanty number of cells, diVerent kinds of associative learning. From an anatomical point of view, C. elegans lacks of a vascular system, therefore allowing the study of several developmental processes independently of blood vessel growth. In contrast, it has a prominent ventral and a minor dorsal nerve cord, running along its longitudinal axis. The centerpiece of its nervous system is the circumpharyngeal nerve ring, surrounded by six ganglia; this structure is sometimes referred as the ‘‘brain’’ of the worm. With a few exceptions, neurons in C. elegans have a simple uni‐ or bipolar morphology, typical of invertebrate animals. As discussed later more extensively, in order to acquire a functional neuronal network, specific connections (synapses) need to be formed between the growth cone of nerve cells and their target cells. For that, neurons send axons that are guided to their final destination by a process called axon guidance. Finally, once the axon reaches its final target, the diVerentiation of presynaptic terminals occurs. Molecules involved in axon guidance and synapse formation remain extremely interesting subjects of investigation. At the neuromuscular

7

1. Angiogenesis and Neurogenesis A

1

Distal gonad

23

4

5

Pharynx Intestine

Anus Uterus Proximal gonad

B Body wall muscle

Lateral ganglion

Amphid nerve

Dorsal ganglion

Nerve

3

2

1 Hypodermis

Ring DNC Labial nerve

Pharynx

Ventral ganglion Dorsorectal ganglion 4

5

Intestine

Distal gonad

Anal depressor muscle

Pseudocoelom Seam cell Lumbar ganglion VNC

Proximal gonad

Rectum

Figure 2 Anatomy of an adult C. elegans hermaphrodite. (A) DIC image of an adult hermaphrodite, left lateral side. Scale bar 0.1 mm. (B) Schematic drawing of anatomical structures, left lateral side. Dotted lines and numbers mark the level of each section in the lower part of the figure. (1) Section through anterior head. (2) Section through the middle of head. (3) Section through posterior head. (4) Section through posterior body. DNC, dorsal nerve cord; VNC, ventral nerve cord. (5) Section through tail, rectum area (adapted from Altun, Z. F. and Hall, D. H. 2005. ‘‘Handbook of C. elegans Anatomy.’’ In WormAtlas http://www.wormatlas.org/handbook/ contents.htm).

junction (NMJ), nematode muscle cells are unusual: instead of motoneuron axons navigating to contact the muscle, the situation is turned upside down, and in C. elegans, muscle cells send cellular processes (muscle arms) to contact motoneurons.

8

Zacchigna et al.

Genetic studies in the worm, by forward genetics or other manipulation techniques, have shed light on several aspects of the nervous system development. Furthermore, a plethora of components of synaptic neurotransmission have been identified in the worm and linked to specific behavioral functions, with similar approaches currently only carried out in the fruit fly Drosophila melanogaster, as discussed later. In contrast, only few studies in C. elegans have been based on electrophysiological recordings, with a still poor understanding of the molecular details of neurotransmitter function at the level of the synapsis. Recently, the introduction of new techniques has rendered the worm even more attractive for the neurobiology field. The discovery of RNA interference (RNAi), first in C. elegans, currently allows genome‐wide screening for genetic components of specific neural functions. Notably, the 2006 Nobel Prize in Physiology or Medicine has been awarded to the two American scientists, Andrew Fire and Craig Mello, who first published their discovery of the RNAi mechanism in 1998 (Fire et al., 1998). Although the system is not equally reliable for all neural genes (Tavernarakis et al., 2000), the development of sensitized genetic backgrounds has significantly improved the eYcacy of gene knockdown by simply feeding or injecting double‐ stranded RNA (dsRNA) targeting individual worm genes. In mammalian systems, only short 22‐nucleotide dsRNA molecules are used, in order to avoid an interferon response or nonspecific inhibition of protein synthesis through dsRNA‐dependent protein kinases (Elbashir et al., 2001; Yang et al., 2001). However, since neither of those responses occurs in C. elegans, it is possible to use long dsRNA, which will give rise to many diVerent siRNA molecules and attack the target mRNA at several points, thus enormously increasing the eYciency of RNAi. Of notice, RNAi can also be induced at any time during the animal’s life cycle, therefore oVering the opportunity to study gene function at all stages. Another advantage of the use of RNAi in C. elegans is that RNAi‐induced phenotype can be maintained over several generations, simply by continuously feeding the worm on bacteria producing the relevant dsRNA. Thus, the ease of the system, together with the availability of the relevant genome data, has enabled a novel, high‐throughput, systematic reverse genetic approach, known as genome‐wide RNAi, based on libraries of either in vitro synthesized dsRNA or bacteria that produce dsRNA (Mello and Conte, 2004; Tabara et al., 1998). Finally, the introduction of the green fluorescent protein (GFP), also first in C. elegans (Chalfie et al., 1994), now allows us to follow neural development, axon migration, and synaptogenesis in vivo simply by looking at cell fluorescence. In this context, neuronal specific green fluorescent transgenic worms are being used (Christensen et al., 2002), thus enabling a number of screens that have established milestones in our understanding of neural patterning (Seifert et al., 2006).

1. Angiogenesis and Neurogenesis

9

B. D. melanogaster (Fruit Fly) The fruit fly D. melanogaster, a small insect 3‐mm long, has been used for decades to elucidate various developmental processes through its powerful genetics. Its importance for biological research and human health was recognized by the award of the Nobel Prize in Physiology and Medicine to E. Lewis, C. Nusslein‐Volhard, and E. Wieschaus in 1995. Part of the reason why several scientists use this insect as a model for their research is historical (so much is known about it that it is easy to handle and manipulate) and part is practical: it is a small animal, with a short life cycle of about 2 weeks, and is cheap and easy to keep in large numbers. The Drosophila egg is about half a millimeter long. It takes 1 day for the embryo to develop and hatch into a warm‐like larva. The larva eats and grows continuously, molting 1, 2, and 4 days after hatching (first, second, and third instars). After 2 additional days, it molts one more time to form an immobile pupa. Over the next 4 days, the body is entirely remodeled to give origin to the adult winged form, which then hatches from the pupal case and is fertile in about 12 hours (this timing refers to 25  C; at 18  C, development takes twice as long). Mutant flies, with defects in any of several thousand genes, are available, and the entire genome, containing 14,000 genes, has recently been sequenced. Drosophila has four pairs of chromosomes: the X/ Y sex chromosomes and the 2, 3, and 4 autosomes. The magic markers that first put Drosophila in the spotlight are polytene chromosomes. During larval growth, the number of cells is kept constant but gene expression increases. As a consequence, cells get much bigger and each chromosome divides hundreds of times, but all the strands stay attached to each other. The result is a massively thick polytene chromosome, which can be easily seen under the microscope. Even better, these chromosomes have a pattern of dark and light bands, reminiscent of a bar code, which is unique for each chromosome section. As a consequence, any large deletion of major rearrangement can be identified, and by the use of nucleic acid probes, individual cloned genes can be placed on a polytene map. 1. Drosophila and Its Use to Study the Nervous System As happens for C. elegans, Drosophila also lacks of a proper vascular system. It only has a primitive heart, and hemolymph circulates through tissue spaces devoid of any endothelial lining (HoVmann, 1995). The analysis of mutants and the possibility to conduct genetic screening for particular phenotypes have been extensively applied to decipher the molecular mechanisms leading to a functional neuronetwork in Drosophila. Moreover, the progressive development of genetic and cell imaging techniques has allowed neurobiologists to generalize the use of various neuronal models at diVerent developmental

10

Zacchigna et al.

stages. It also allowed the execution of single‐cell analysis, as well as single‐cell genetic manipulations. Several neuronal models have been chosen for their best adequacy to study specific steps in the formation of functional networks, highlighting some constants in the programming of neuronal connectivity. In this respect, to study axon guidance, Drosophila oVers one of the most useful models: the checkpoint decision at the midline. As further described below in more detail, the central nervous system (CNS) of bilateral animals is physically divided by the midline, through which neurons have to establish communications in order to ensure proper coordination between the two sides of the brain. Indeed, specific neurons send their processes across the midline, creating communication lines, displayed as commissures. Therefore, several CNS neurons face two choices regarding their projections, remaining ipsilateral, thus avoiding the midline, or projecting contralaterally and crossing the midline. This is the reason why the midline, with its ‘‘yes or no’’ choice is an ideal paradigm to study how neurons make a directionality decision at a checkpoint. In Drosophila, the embryonic abdominal CNS is composed of six repetitive identical segments (A2–A7), each of which in turn can be divided into two mirroring hemisegments. Each abdominal hemisegment contains 342 neurons, including 34 motoneurons leaving the CNS to innervate 30 abdominal muscles (Landgraf et al., 1997; Schmid et al., 1999). Staining of the whole embryonic neuropile reveals a ladderlike structure with longitudinal fascicles connected by a pair of anterior and posterior commissures in each segment (Seeger et al., 1993; Fig. 3). In this context, two main groups A

B

VNC

B

Figure 3 CNS development in Drosophila embryo. (A) Ventral view of a whole Drosophila embryo stained with the monoclonal antibody BP104, which binds to and marks axons of all neurons in the CNS. The outlines of the brain (B) and the ventral nerve cord (VNC) are shown in (B).

1. Angiogenesis and Neurogenesis

11

of mutations have been identified: mutants in which commissures appear thickened by an excessive number of crossing fibers, and mutants with thinner commissures due to a reduced amount of fibers crossing the midline. These opposite phenotypes lead to a model in which excess crossing is due to a lack of repulsion, while reduced crossing depends on absence of attraction (Tessier‐ Lavigne and Goodman, 1996). Several molecules that generate this disorganization of midline crossing have been identified so far, and most of them will be described and discussed in the following sections. Here, we want to emphasize how these studies highlighted the importance of midline glial cells as an organizing center of neuronal projections. In fact, as further discussed later, midline glia are responsible for the secretion of the major guidance cues for commissural axons, including Netrin and Slit, which provide attractive and repulsive signals by interacting with frazzled and Robo receptors, respectively (Battye et al., 1999; Harris et al., 1996; Mitchell et al., 1996). With its stereotypical organization, the ventral nerve cord of Drosophila has allowed single‐cell analysis manipulation, mainly through the use of the Gal4/UAS system (Brand and Perrimon, 1993). In this system, Gal4 drives the expression of a specific transgene in a restricted subset of cell types. Alternatively, fluorescent proteins can be expressed in a cell of interest, allowing its complete morphological visualization in vivo. In this respect, the generation of transgenic flies expressing fluorescent reporters in specific subset of neurons has been extremely useful for the visualization of axon pathfinding in vivo (Murray et al., 1998; Salvaterra and Kitamoto, 2001). In addition, such approaches have shown that axons of a specific neuron can respond in a diVerent manner to mutations of robo (the receptor for the midline axon‐repellent molecule Slit in flies) or frazzled (the Drosophila receptor for the midline axon‐attracting molecule Netrin). More precisely, the anterior corner cell (aCC) motoneuron normally displays an ipsilateral axon and two groups of dendrites, one that crosses the midline and the other one that remains ipsilateral. In frazzled mutants, aCC axonal projection remains unaVected, while the dendrites are not any more able to cross the midline, indicating a role for frazzled in midline dendritic crossing. In robo mutants, both aCC axons and dendrites remain normal. The story appears completely diVerent for another motoneuron (RP3), which normally projects its axon contralaterally and has two dendrite groups, growing away on each side of the midline. In this case, frazzled mutants prevent the axon to cross the midline, whereas robo mutants have normal axons but dendrites fail to escape the midline (Furrer et al., 2003; Wolf and Chiba, 2000). These results show that neurons are able to subcellularly integrate divergent signals in axons and dendrites and make Drosophila an invaluable tool to visualize and genetically manipulate single cells, thus shedding the light on the subcellular regulation involved in axon guidance at the midline.

12

Zacchigna et al.

Another research field that has extensively exploited Drosophila as a model organism is the study of synaptic contacts between neurons. In fact, the precise description of single brain cells and synapses and their amenability to genetic analysis has provided a useful platform to unravel the mechanisms and principles of synapse formation, which find many counterparts in other animals. In particular, a detailed description of synaptic development and structure has been reported for the Drosophila NMJ, which is easily accessible to manipulation and visualization. Fly’s NMJs are established in the periphery in predictable combinations between individual motoneurons and muscles (Landgraf et al., 1997), in a stepwise process. First, during the period of pathfinding, each motor axon grows to its appropriate exit point from the CNS, and either pioneers or joins the correct nerve branch in order to reach its target muscle (Prokop and Meinertzhagen, 2006). Second, a precise number of synapses form at each contact site, acquiring an appropriate neurotransmitter composition and spatial distribution. More precisely, synapses start to assemble at 13 hours of embryonic development, with most of the contacts reaching structural and functional maturity at the time of hatching (Prokop, 1999). However, during larval stages, NMJs dramatically increase in size, and an additional de novo formation of NMJs occurs during metamorphosis at the pupal stage. At each phase of postembryonic development, NMJs adopt a stereotypic morphology, defined by nerve entry points, branching pattern, and terminal size (Johansen et al., 1989). While the mechanisms controlling NMJ formation and diVerentiation during embryonic development still remain largely unknown, a huge amount of information has been accumulated on larval NMJ, in which synapse assembly clearly requires a coordinated regulation by pre‐ and postsynaptic cells (Ashley et al., 2005; Paradis et al., 2001). In contrast to NMJs, synapse assembly at photoreceptor contacts only occurs during pupal development, although some plasticity persists throughout adult life. The Drosophila visual organ represents another paradigm model system for the understanding of neuronal connectivity. Every photoreceptor terminal has to establish about 50 presynaptic sites, each arranging in a tetrad (a constellation of 4 postsynaptic sites facing a single presynaptic release site), with a predictable blend of postsynaptic cells (L1, L2, L3, and amacrine cells). Conversely, each L1 or L2 lamina cell has to connect with the six photoreceptor terminals of its lamina module, in order to pool the information from all the photoreceptors of the same field of view (Meinertzhagen, 2000). Hence, Drosophila represents an outstanding model to dissect the mechanisms ensuring the correct number of synapses and their specificity in order to establish reproducible and functional microcircuits in the lamina. Although the anatomy of the fly nervous system, consisting of 100,000 neurons, diVers significantly from that of vertebrates, many fundamental cellular and molecular features of neuronal development and patterning are

1. Angiogenesis and Neurogenesis

13

conserved between vertebrates and invertebrates. This conservation makes Drosophila a powerful system for basic studies of neuronal development and function and, more recently, also for studies of neuronal dysfunction. In fact, despite the prevalence and the severity of Parkinson’s disease, Alzheimer’s disease, and multiple sclerosis, little is known about their molecular etiology; therefore, disease‐modifying therapies have been so far remained largely elusive. Although recent linkage studies identified a few genes responsible for rare, heritable forms of neurodegenerative disorders, we still know very little about the biological functions of those genes and how their mutations lead to neuronal death. The completion of both human and Drosophila genome‐sequencing projects has also revealed that a large fraction of human genes involved in neurodegenerative disorders have highly conserved counterparts in Drosophila. For instance, the Drosophila genome encodes homologues of five of the six Parkinson‐related genes identified so far (Whitworth et al., 2006). In this respect, an important advantage of using Drosophila to understand human disease is the possibility to perform genome‐wide genetic screens for mutations in other genes able to modulate the phenotype associated with a certain disease model. The power of this approach is the potential to identify genetic pathways that cause the disease, as well as those that can influence its progression, without requiring a priori knowledge of the function of the disease gene. In this way, the human counterparts of suppressors identified from screens using Drosophila define potential targets for therapeutic interventions.

C. Zebrafish The zebrafish (Danio rerio) is a small tropical freshwater fish, which lives in rivers of northern India, northern Pakistan, Nepal, and Bhutan in South Asia. It possesses a unique combination of features that make it particularly well suited for experimental and genetic analyses of early vertebrate development (Anderson and Ingham, 2003; Kimmel, 1989; Ny et al., 2006). Adult zebrafish are only 3‐ to 4‐cm long, so large numbers can be maintained relatively inexpensively in a small space (Fig. 4). Furthermore, zebrafish reaches sexual maturity in about 3 months, and a pair of zebrafish can generate hundreds of embryos every few weeks, making it possible to generate thousands of progeny from a single breeding pair of fishes. Zebrafish eggs are externally fertilized, providing readily access to the developing embryos at all stages of development (Fig. 4). The fertilized embryos develop rapidly making it possible to observe the entire course of early development in a short time. As zebrafish embryos are optically clear, it is possible to directly, noninvasively observe the major parts of the neural and vascular systems (Ny et al., 2006). For all these reasons, zebrafish has recently emerged as an

14

Zacchigna et al.

Figure 4 Zebrafish as a model to study angiogenesis and neurogenesis. Image of a typical adult zebrafish (A). Morpholino injection into one cell stage zebrafish embryo (B). Confocal image of the vasculature of a 48‐hpf Fli:GFP transgenic zebrafish (C). Confocal image of a transverse section of a 48‐hpf Fli:GFP transgenic zebrafish. Note ISVs growing in close apposition to somites and neural tube (D). Confocal image of a 48‐hpf zebrafish immunostained with anti‐ acetylated tubulin to detect axons. Note motoneuron axons growing out of the spinal cord and commissural axons crossing the midline (E). DA, dorsal aorta; PCV, posterior cardinal vein; DLAV, dorsal longitudinal anastomic vessel; PAV, paracordal vessel; ISV, intersegmental vessel; SOM, somite; SC, spinal cord; MNA, motoneuron axon; CA, commissural axons.

advantageous model organism for the study of the stereotypic and evolutionary conserved development of blood vessels and nerves. 1. Zebrafish and Its Use to Study Vascular Development The vasculature is often diYcult to visualize, manipulate, and analyze in higher vertebrates, mainly because it is deeply dispersed within other, fairly opaque tissues, and its function is essential early in development.

1. Angiogenesis and Neurogenesis

15

Intersegmental vessels (ISVs) in zebrafish embryos develop within the first 2 days of life. Pathfinding of these vessels is stereotyped and likely genetically programmed by an interaction of attractive and repulsive cues. In control embryos, ISVs sprout from the dorsal aorta and grow dorsally between the somites and neural tube; eventually they elongate and fuse with vessels from the adjacent segments to form the dorsolateral anastomotic vessel (DLAV) (Lawson and Weinstein, 2002). Secondary sprouts then come out from the posterior cardinal vein (PCV) and migrate dorsally up to the horizontal myoseptum to form the parachordal vessel (PAV) (Lawson and Weinstein, 2002; Fig. 4). As described above, the zebrafish embryos are optically transparent allowing the visualization of the vascular system. In addition, fish embryos are small enough that they can receive suYcient oxygen via passive diVusion to develop normally for a few days in the absence of blood circulation, on perturbation of angiogenic processes. This is a unique advantage since perturbation of angiogenic processes in mice leads to early embryonic lethality, complicating or impeding phenotypic analysis. Moreover, zebrafish is easily amenable to forward and large‐scale genetic analyses. In this respect, an angiogenesis assay in the regenerating fin of adult zebrafish was used to screen for antiangiogenic activity of chemical compounds (De Smet et al., 2006). For the purpose of investigating the mechanisms of angiogenesis in zebrafish, a variety of tools and methodologies have been recently developed, thus enormously amplifying the intrinsic advantages of the fish model. These include cell‐fate and lineage analysis techniques, microinjection of biologically active molecules, gene knockdown by injection of morpholino antisense oligonucleotides, transgenic zebrafish lines expressing fluorescent proteins under the control of vascular‐specific promoters such as the promoter for fli1 (Lawson and Weinstein, 2002) or flk1 (Jin et al., 2005; Fig. 4), as well as advanced microscopy technologies such as confocal microangiography and high‐resolution mutliphoton time‐lapse imaging (Lawson and Weinstein, 2002; Motoike et al., 2000; Weinstein et al., 1995). As further discussed later, the latter technique has supplied our first glimpses of the highly dynamic, growth cone‐like behavior of growing endothelial cells (ECs) in vivo (Lawson and Weinstein, 2002).

2. Zebrafish and Its Use to Study the Development of the Nervous System Parallel to its wide use in cardiovascular developmental biology, many zebrafish mutants have been characterized, which exhibit specific defects in axon guidance and/or synaptogenesis, thus making the fish an excellent model in neurobiology as well (Hutson and Chien, 2002). Indeed, its nervous system is simple and well characterized (Beattie, 2000). Its optical transparency allows to image cell movement in vivo, or to ablate specific cells in order to screen for

16

Zacchigna et al.

defects in development or behavior (Liu and Fetcho, 1999; Myers et al., 1986). In addition, cells or tissues can be easily transplanted to test the autonomy of gene function (Fricke et al., 2001). Like ISVs, motoneuron axons in zebrafish follow a highly stereotyped pattern to navigate to their final destination. Therefore, their migration during development has been analyzed to identify new axon guidance cues (Beattie et al., 2002). Transgenic zebrafish expressing GFP under specific motoneuron promoters, such as islet1 (Higashijima et al., 2000) or hb9 (Flanagan‐Steet et al., 2005), have been developed and used to study motoneuron axonal growth and guidance in vivo. Furthermore, GFP has been cloned under several zebrafish neuronal promoters to follow the development of the nervous system in vivo (Park et al., 2004a; Yoshida and Mishina, 2003). Finally, zebrafish has also been used as an animal model to study commissural axon fasciculation and midline crossing at the hindbrain (Marx et al., 2001). Of notice, instead of characterizing neurons based on their gene expression, large genetic screens can be successfully used to search for genes according to their function in zebrafish (Driever et al., 1996; HaVter et al., 1996). An interesting example refers to space cadet, a mutant with abnormal locomotion, both in response to escape stimuli and during normal swimming (Granato et al., 1996). In these mutants, some hindbrain commissural fibers fail to make normal connections to Mauthner neurons, the largest neurons in the fish hindbrain, which are an essential component of the escape response circuit (Lorent et al., 2001). Interestingly, another mutant, deadly seven (des)/ Notch‐1a, which has five times the normal number of Mauthner neurons, does not display any defect in escape response (Gray et al., 2001). Further analysis of these mutant larvae revealed that all the excess Mauthner neurons were functionally incorporated into the escape circuitry, but with compensatory decrease in the number of collateral departing from each Mauthner neuron, as well as in the overlap of target innervation (Liu et al., 2003). These data show that, similar to humans, neuronal circuitries in the zebrafish embryo display a remarkable degree of plasticity, and it is tempting to speculate that such developmental plasticity has been the prerequisite for the evolution of new and more specialized neural networks.

D. Xenopus The fundamental work of Hans Spemann on the Organizer, which entitled him for the 1935 Nobel Prize, marked the emergence of the amphibian as an important model system for the study of development. In most laboratories of the molecular era, amphibian embryology is taken as a synonym of the study of the African clawed frog Xenopus. This anuran has actually served as a treasure trove for biologists to plunder in search of novel developmental

1. Angiogenesis and Neurogenesis

17

regulators. For instance, the definition of the role of transforming growth factor‐ (TGF‐ ) in mesoderm induction, as well as the identification of several patterning genes, such as noggin, cerebrus, or chordin, has been accomplished in Xenopus (Callery, 2006). The most used frog so far is the tetraploid X. laevis, although many groups are now also adopting its more genetically amenable diploid relative, X. tropicalis (Khokha et al., 2002). Indeed, while the slower development and larger embryos of X. laevis make them preferable for experimental embryology, the shorter generation time of X. tropicalis renders genetics more feasible in this species, and several mutants have been already identified (Noramly et al., 2005). The recent availability of X. tropicalis genome sequence will result in great use for several aspects of developmental biology. First, the analysis of cis‐regulator sequences can be identified in silico and subsequently tested in vivo for their capacity to drive the expression of appropriate reporter genes in transgenic frogs. Second, there will be a great advantage for loss‐of‐function studies, as morpholinos targeting candidate genes are likely to work more eYciently in X. tropicalis than in X. laevis, because of the allelic redundancy in the tetraploid. Finally, the sequenced genome will allow the use of additional techniques such as chromatin immunoprecipitation and microarray analysis. 1. Xenopus and Its Use to Study the Vascular System Of notice, both Xenopus and zebrafish models have been elegantly exploited to better understand several aspects of cardiovascular physiology. Indeed, Xenopus has been used for the screening of chemical compounds for cardiovascular development (De Smet et al., 2006). In fact, for many years, the mouse has been the preferred model because of the known genomic sequence and the possibility to create knockout as well as transgenic animals. However, genetic mutations linked to severe cardiovascular dysfunction usually die very early, making functional analysis diYcult, if not impossible. In contrast, as already discussed for zebrafish, also in Xenopus, mutations aVecting the heart at early stages do not aVect survival because they are independent of blood flow for oxygen delivery early on (Territo and Burggren, 1998). There are additional advantages in the use of these small animal models for studies of embryonic cardiovascular function and development. Measurements of heart rate, oxygen consumption, blood pressure, and hematocrit have all been performed in developing embryos of both species (Bagatto et al., 2001; Fritsche et al., 2000; Jacob et al., 2002; Pelster and Burggren, 1996; Schwerte and Fritsche, 2003). Most of the times, as the larvae of both animals are transparent, these parameters can be monitored noninvasively, by means of a variety of video techniques. While the cardiovascular system becomes functional at 24 hours postfertilization (hpf) in zebrafish, in Xenopus this occurs at NF stage 33/34 (48 hpf).

18

Zacchigna et al.

In addition, to study vascular development in Xenopus, transgenic frogs expressing fluorescent proteins under the specific vascular promoter Tie2 have been developed (Ross Breckenridge and Timothy Mohun, personal communication). Moreover, the use of this small animal model has also been proven as a powerful model to study lymphangiogenesis (Ny et al., 2005).

2. The Use of Xenopus to Study the Development of the Nervous System The fact that, in Xenopus, neural development occurs much earlier than vascular development renders the frog particularly useful for the study and the dissection of the molecules involved in the neurovascular link. In fact, during Xenopus embryogenesis, there is an early window (12–48 hpf ) in which the eVect of a candidate gene directly on the nervous system can be studied independently from any possible vascular eVect. The embryonic Xenopus spinal cord has been proven to be an excellent model to follow the growth pattern of living neurons (Moon and Gomez, 2005; Shim et al., 2005). In addition, as mentioned above, it provides the advantage to study the possible role of diVerent angiogenic molecules in midline axon guidance without any primary vascular defect: commissural axons start crossing the midline at approximately stage 23 and are finished by stage 28, when the first angioblasts start to assemble in order to form the vascular system. Moreover, by the blastomere injection method, it is possible to knock down a gene specifically at the dorsal part of the neural tube, thereby targeting specifically commissural neurons (Robles and Gomez, 2006). A representative picture of a neuronal staining of a spinal cord from a stage 28 Xenopus embryo is shown in Fig. 5. An additional system that has been providing increasing insight on axonal wiring is the amphibian visual system, which undergoes profound remodeling during metamorphosis. In tadpoles, the two eyes are laterally placed, with retinal cells projecting exclusively to the contralateral side of the brain and no binocular overlap (Holt and Harris, 1983). During metamorphosis, as the skull changes its shape, the eyes migrate frontally, leading to a substantial degree of binocular overlap, which will be essential to the predatory lifestyle of the adult frog (Grant and Keating, 1986). Concomitantly, a new pattern of retinal projections has to be established, connecting each retina to ipsilateral thalamic nuclei (Hoskins and Grobstein, 1985). The pattern of cell production in the retina is also completely changed during metamorphosis. Whereas in tadpoles retinal stem cells proliferate symmetrically, a sudden shift toward asymmetrical growth occurs at the beginning of metamorphosis, with more cells being added at the ventral and temporal margin than at the dorsal and

1. Angiogenesis and Neurogenesis

19

Figure 5 Xenopus as a model to study angiogenesis and neurogenesis. Image of a X. laevis egg at one cell stage (A). Stage 26–28 of a X. laevis tadpole (B). Fluorescent image of an Flk1:GFP transgenic tadpole at stage 46–47, as observed GFP is expressed throughout the whole vascular system (C). Whole mount neurofilament staining of a dissected spinal cord from a stage 28 tadpole showing commissural axons crossing the midline (D). Transverse section of a stage 28 tadpole showing commissural neuron cell bodies and axons (neurofilament staining in red) crossing the midline (E). Nuclear staining (DAPI) is in blue. DA, dorsal aorta; PCV, posterior cardinal vein; DLAV, dorsal longitudinal anastomotic vessel; H, heart; SC, spinal cord; CN, commissural neuron cell bodies; CA, commissural axons; NC, notochord.

nasal margin (Beach and Jacobson, 1979). This is attributed to an increased number of progenitor cells in the ventral as compared to the dorsal margin, and it has been proposed that the extensive proliferation in the ventral retina serves the production of neurons for the new binocular visual field and compensates for the change in eye position (Mann and Holt, 2001). This model system has been of particular relevance to study the role of ehprin (Eph) family molecules in axon guidance (Mann and Holt, 2001), as discussed later in more detail.

20

Zacchigna et al.

III. Vascular and Neural Cell‐Fate Specification In this part of the chapter, we will focus on recent evidences demonstrating that cell‐fate determination of vascular and neuronal precursors is regulated in part by common signals and genetic pathways. We will also describe how vascular and neuronal precursors influence the cell‐fate decision making of one another. The main types of vascular cells, ECs, arise from mesodermal angioblasts or hemangioblasts (Vogeli et al., 2006). The fact that diVerent ECs arise from diVerent angioblasts (Rovainen, 1991) indicates that angioblasts might have or acquire positional and temporal identity during development. One of the most important EC‐fate determinations is the diVerentiation of ECs into arterial or venous ECs. In this context, molecules and pathways known to be crucial for neural cell‐fate specification have also been shown to participate in arteriovenous cell fate. Although blood pressure can influence this cell‐fate decision, recent findings indicate that arteriovenous EC fate is specified before the onset of circulation. An example of common mechanisms regulating vascular and neural cell fate is shown in zebrafish embryos: during development, angioblast precursors for the dorsal aorta and the PCV are mixed in the lateral posterior mesoderm (Zhong et al., 2001); the decision of those angioblasts to form either the aorta or the vein is induced by signals released by the ventral endoderm and the notochord (Fouquet et al., 1997; Sumoy et al., 1997). Similarly, neural progenitors also receive signals from the notochord and the endoderm in order to diVerentiate. Genetic studies in mice, zebrafish, and Xenopus have started to define the transcriptional code that determines EC fate (Brown et al., 2000; Liao et al., 2000; Mikkola and Orkin, 2002). As in neuronal cell fate, this code involves basic Helix‐loop‐Helix (bHLH) transcription factors (Carmeliet, 1999) and inhibitors of DNA biding and inhibitors of differentiation (Id) repressors (Lyden et al., 1999). During embryonic development, neurons and glial cells arise from neuroectodermal stem cells (NSCs) located in the neural tube. In order to diVerentiate, NSCs have to pass through three main phases: (1) the decision to commit to a neural cell phenotype, (2) the determination of positional identity (anteroposterior and dorsolateral), and (3) the developmental decision to diVerentiate (Osterfield et al., 2003; Panchision and McKay, 2002; Temple, 2001). Notch expression in the neural tube is responsible for maintaining the NSC potential by upregulating Hes1/5 type repressors (Gaiano and Fishell, 2002; Hitoshi et al., 2002). Vascular endothelial growth factor (VEGF), the key player of angiogenesis (Lawson and Weinstein, 2002), has been implicated also in maintaining NSC potential. In the avascular chicken retina, VEGF is secreted by postmitotic retinal neurons and, by binding to its receptor Flk1 expressed in neuronal progenitor cells, influences cell proliferation and suppresses retinal ganglion diVerentiation (Hashimoto et al., 2006).

1. Angiogenesis and Neurogenesis

21

The activation of the mitogen‐activated protein kinase MEK–ERK pathway by VEGF was shown to be responsible for the proliferation of retinal neuronal precursors, while the induction of a Hes1 response was responsible for the blockage of the diVerentiation of these cells to retinal ganglion cells (RGCs) (Hashimoto et al., 2006). When pro‐proliferation signals are downregulated, neural cells start to diVerentiate. The subsequent positional identity of the diVerentiating neural cells is determined by gradients of the morphogens fibroblast growth factor (FGF), bone morphogenetic protein (BMP), and sonic hedgehog (Shh), which are secreted by tissues adjacent to the neural tube (Osterfield et al., 2003; Patten and Placzek, 2000; Temple, 2001). Furthermore, they induce distinct neuronal cell subtypes in a dose‐dependent manner among the morphogen gradient. Shh, by regulating the levels of VEGF, is also involved in vascular cell‐fate determination. This was shown by experiments performed in zebrafish mutants of sonic you (the homologue of Shh in mammals), where the formation of the aorta was impaired (Brown et al., 2000; Chen et al., 1996). Shh was shown to induce expression of VEGF in the adjacent somites, which in turn drove the arterial diVerentiation of angioblasts (Lawson and Weinstein, 2002; Lawson et al., 2002). Parallel to its role in neuronal cell fate, Notch signals may also regulate the decision of hemangioblasts to diVerentiate into either endothelial or hematopoietic cells. Furthermore, Notch signaling also influences arterial EC‐fate specification by acting downstream of VEGF (Lawson and Weinstein, 2002). When Notch signaling is knocked down in zebrafish embryos, there is a loss of artery‐specific markers and an ectopic vein marker‐gene expression in the dorsal aorta. Consistently, the ectopic activation of Notch signaling represses venous cell fate (Lawson et al., 2001, 2003). Recent insights have shown that the orphan nuclear receptor, COUP‐ TFII, has a critical role in repressing Notch signaling to maintain vein identity. COUP‐TFII is expressed specifically in venous endothelium, and its mutation leads to activation of arterial markers in veins (You et al., 2005). The neurovascular link is further supported by the cell‐fate specification of neural crest (NC) cells. NC cells segregate from the dorsal portion of the neural tube and migrate as a pluripotent cell population to several regions in the embryo. NC cell diVerentiation is induced by a combination of a mediolateral gradient of BMP and an anteroposterior gradient of Wnt, FGF, and retinoic acid (Aybar and Mayor, 2002; Etchevers et al., 2002; Knecht and Bronner‐Fraser, 2002). NC cells are able to diVerentiate to cells of the peripheral nervous system, melanocytes, and mesectodermal derivatives like the craniofacial cartilage and bone (Dupin et al., 2001; Knecht and Bronner‐ Fraser, 2002). Interestingly, NC cells also diVerentiate to the smooth muscle cells (SMCs) that cover the blood vessels of the pharyngeal arch arteries, and vessels in the jaws and in the forebrain (Dupin et al., 2001; Etchevers et al., 2002). Migration and diVerentiation of NC cells then depend on intrinsic

22

Zacchigna et al.

cascades of transcription factors, receptors, and ligands (MaschhoV and Baldwin, 2000) such as TGF‐ 1 which directs NC cells to an SMC fate (Shah et al., 1996). Notch‐3 signaling is also crucial for maintaining SMC fate and, when mutated, causes cerebral arteriopathy and stroke (Kalimo et al., 2002). Besides the NC, SMCs in other regions of the organism arise from mesodermal or epicardial‐derived cells (Carmeliet, 2000; Etchevers et al., 2002; Mikkola and Orkin, 2002). Examples of crosstalk between both systems are given by the fact that VEGF, when released from Schwann cells, induces arterial specification of vessels tracking alongside these nerves (Mukouyama et al., 2002) in the embryonic limb skin. The induction of nerve‐mediated arterial diVerentiation as well as the patterning of blood vessels along peripheral nerves is beneficial for both systems. On one side, apart from supplying the nerves with oxygen and nutrients, arterial vessels express neurotrophic factors such as nerve growth factor (NGF), neurotrophin‐3 (NT‐3), and brain derived neurotrophic factor (BDNF), which might be important to maintain the survival of growing axons before arriving at their final peripheral destination. On the other side, nerves control vasoconstriction and dilation of their flanking arteries. In a study, Mukouyama et al. (2005) used Cre‐specific mouse lines to demonstrate that, through binding to Np‐1, VEGF derived from motoneurons, sensory neurons, and Schwann cells is required for in vivo arterial diVerentiation. Another example of crosstalk is the induction of specific blood–brain barrier (BBB) ECs by glia‐derived neurotrophic factor (GDNF) and other glial cell‐ derived factors (Carmeliet, 2003; Orte et al., 1999). The BBB is a nonfenestrated EC barrier where tight junctions seal oV the vascular lumen. It can be considered as a functional neurovascular unit, constructed by ECs, astrocytes, and neurons, where mutual interactions between each component contribute to the formation, maintenance, and function of the BBB (Kim et al., 2006).

IV. Molecular Links Between Angiogenesis and Neurogenesis We will now discuss the direct link between angiogenesis and neurogenesis. ECs have other functions than constituting pipelines for the supply of oxygen. Instead, they are now known to release inductive cues for organogenesis and morphogenesis of various organs during development (Cleaver and Melton, 2003; Compernolle et al., 2002; Eremina et al., 2003; Gerber et al., 1999), as well as for neurogenesis and neural cell fate. ECs are present at similar sites as NSCs and astroglial cells, and interact with these cell types in a temporospatial manner (Huxlin et al., 1992; Zerlin and Goldman, 1997). In specific areas of the CNS in mammals, NSCs proliferate in small clusters around dividing capillaries—termed the vascular niche (Palmer et al., 2000). Furthermore, ECs release factors such as BMP‐2, BDNF, and FGF, which

23

1. Angiogenesis and Neurogenesis NSC Neural tube

Notch

BMP

Somite

Shh

VEGF N Dorsal aorta Notch Axial vein Angioblast Figure 6 Vascular and neural cell fate. Notch expression in the neural tube maintains the NSC potential. When Notch is downregulated NSC start to diVerentiate, then, gradients of morphogens such as Shh, which is secreted by the notochord, and BMP, determine the ventral and dorsal identity of neural progenitors cells and induce distinct neuronal cell subtypes in a dose‐dependent manner along the morphogen gradient. Shh also induces the release of VEGF from the somites, which in turn acts on angioblasts to induce arterial and venous endothelial cell fate.

induce the diVerentiation of astrocyte precursors or NSCs (Mi et al., 2001; Fig. 6). Additional evidence for a crosstalk between neural and vascular cells is supported by the fact that VEGF, and semaphorin‐3A (Sema‐3A) antagonistically aVect neural progenitor cells (Bagnard et al., 2001) and ECs (Miao et al., 1999). The link between the development of nerves and blood vessels is also strengthened by the observation that conditions that increase neural activity and stimulate neurogenesis also trigger angiogenesis (Kokaia and Lindvall, 2003; Monje and Palmer, 2003). New neurons are continuously being generated in the adult brain in localized discrete regions such as the rostral subventricular zone (SVZ) and the subgranular zone (SGZ) of the hippocampal dentate gyrus (DG). In these two neurogenic areas of the adult brain, neural stem cells occupy niches formed by both astrocytes and ECs (Lledo et al., 2006; Fig. 6). ECs are critical components of these niches as they secrete soluble factors that maintain CNS stem cell self‐renewal and neurogenic potential (Shen et al., 2004). When SVZ explants are cocultured with ECs, maturation, neurite outgrowth, and migration of neurons were enhanced (Leventhal et al., 1999), indicating a role for ECs in neurogenesis. Furthermore, a recent report

24

Zacchigna et al.

showed that when coculturing NSCs with human ECs, a significant percentage of the NSC population converted to cells that did not express neuronal or glial markers, but instead had a stable expression of multiple EC markers and the capacity to form capillary networks (Wurmser et al., 2004). Recent evidences show that VEGF is again the molecule that links angiogenesis and neurogenesis. For example, in the adult songbird brain, neurogenesis proceeds throughout life in the higher vocal center (HVC) of the neostriatum. Using this animal model, Louissaint et al. (2002) could show that testosterone‐induced angiogenesis as well as neurogenesis were associated with increased VEGF expression within the HVC and BDNF production by ECs. In addition, they showed that an inhibitor of VEGF receptor‐2 (VEGFR‐2) blocked both the angiogenesis and the neurogenesis‐promoting eVects of testosterone (Louissaint et al., 2002). Another illustration of the role of VEGF in adult neurogenesis was shown in murine cerebrocortical cultures as well as in the adult rat brain in vivo, VEGF stimulated proliferation of neuronal stem cells in the SVZ and in the SGZ of the DG (Jin et al., 2002). VEGF was also shown to induce neuroprotection, neurogenesis, and angiogenesis after focal cerebral ischemia (Sun et al., 2003). Conversely, transient forebrain ischemia‐induced cell proliferation and diVerentiation to mature neurons in the hippocampal DG was shown to be attenuated when a VEGF receptor tyrosine kinase inhibitor was administered intracerebroventricularly after the induction of transient forebrain ischemia (Kawai et al., 2006). Several evidences suggest that VEGF‐B, a VEGF homologue, could also play a role in adult neurogenesis: (1) VEGF‐B is expressed in the brain, and its expression is induced after brain injury (Nag et al., 2002); (2) VEGF‐B knockout mice have been shown to exhibit increased infarct size and more severe neurological deficits after stroke (Sun et al., 2004); and (3) VEGF‐B also reduced hypoxia‐induced cell death of cultured cerebrocortical neurons in vitro (Sun et al., 2004). It was indeed demonstrated that VEGF‐B stimulates neurogenesis in the adult brain (Sun et al., 2006). Sun et al. showed that, in VEGF‐B knockout mice, adult neurogenesis is reduced. Furthermore they showed that the addition of VEGF‐B to neuronal cultures induced neurogenesis in vitro, and that intracerebroventricular administration of VEGF‐B in rats promoted adult neurogenesis in the SGZ of the hippocampal DG and in the forebrain SVZ (Sun et al., 2006).

V. Similarities in the Organization of Vascular and Neural Boundaries The physical segregation of distinct diVerentiated cell populations is a requirement for the organization of both the vascular and nervous network. In both systems, similar families of molecules are responsible for the proper cell

1. Angiogenesis and Neurogenesis

25

separation. Accordingly, cells with common functions are attracted to each other and cell boundaries are established between cells with diVerent properties. In that way, the identity of the many individual cell types in the complex architecture is secured and organized. Members of the Eph family of receptor tyrosine kinases and their membrane‐bound ephrin ligands (Davy and Soriano, 2005; Pasquale, 2005) have been shown to be the molecules responsible for establishing these boundaries through bidirectional cell responses (Mellitzer et al., 1999). Typical examples of boundary formation are the segmentation of the vertebral hindbrain in rhombomeres in the nervous system (Cooke and Moens, 2002; Krull, 2001; Tepass et al., 2002) and the separation between arterial and venous ECs in the vascular system. The hindbrain is transiently subdivided during development into repeated segments called rhombomeres. Due to expression of ephrins and Eph receptors in alternating segments (Cooke and Moens, 2002), rhombomeres are maintained as lineage‐restricted compartments. Ephrin–Eph signaling induces cell repulsion between cells from diVerent rhombomeres, therefore maintaining rhombomeres separated from each other (Cooke et al., 2001). In Xenopus and zebrafish, it was shown that ephrin‐B signaling through Eph‐A4 is necessary for rhombomere boundary formation (Cooke et al., 2005; Xu et al., 1995). Furthermore, mutant zebrafish embryos that lack rhombomere boundaries, due to a null mutation in the val gene, show a failure to establish complementary expression pattern of Eph‐A4 and ephrin‐B2a between rhombomeres. In addition, a publication showed that in Eph‐A4 mosaic zebrafish embryos, Eph‐A4‐knockdown cells and Eph‐A4‐expressing cells segregate from each other, suggesting that Eph‐A4 also regulate boundary formation by promoting cell adhesion within cells of the same rhombomere (Cooke et al., 2005). Taken everything together, Eph–ephrin interactions seem to contribute to the sharpening of segments by regulating both repulsion at interfaces and cell aYnity within rhombomeres. Similar to the nervous system, ephrin–Eph signaling in the vascular system establishes the boundaries between arterial and venous ECs (Adams and Klein, 2000; Brantley et al., 2002; Torres‐Vazquez et al., 2003). Ephrin‐B2 and Eph‐B4 are selectively and respectively expressed in arteries and veins of mouse embryos, and this expression pattern seems to persist in the adult as well (Gale et al., 2001; Shin et al., 2001). Ephrin‐B2‐null mice die at E10.5 as a consequence of impaired vascular diVerentiation and arteriovenous remodeling, which results in a failure to form a properly branched capillary network (Wang et al., 1998). Eph‐B4‐null mice essentially phenocopy ephrin‐B2‐null mice, thus defining Eph‐B4/ephrin‐B2 as principal regulators of vascular morphogenesis (Gerety et al., 1999). In a study with patients suVering from venous malformations, it was found that while in normal subjects, ephrin‐B2 and Eph‐B2 expression was restricted to arterial ECs, in these patients, both molecules were found ectopically expressed in venous

26

Zacchigna et al.

ECs (Diehl et al., 2005). Parallel to what was described for Eph‐A4 as a promoter of cell adhesion between cells within the same rhombomere, in the vascular system, apart from a repulsive role for ephrin‐B2/Eph‐B4 signaling in demarcating arteriovenous cell boundaries, ephrin signaling also regulates the coherence of vascular cell subtypes of the same class. For instance, conditional loss of ephrin‐B2 in vascular SMCs causes the mural vascular smooth muscle coat to loosen up and mural cells to detach from each other (Foo et al., 2006). Finally, ephrin‐B2 also controls interactions between mural cells as well as between pericytes and the endothelium (Foo et al., 2006). These interactions may reduce mural cell migration to ensure a proper cover of mature vascular beds.

VI. Molecular Cues Involved in Nerve and Vessel Guidance Understanding how axons can navigate throughout the body, eventually reaching their final target, is one of the most interesting topics in the neurobiology field today. As they move, growth cones encounter and respond diVerently to a complex array of attractive and repulsive signals at diVerent points along their pathway. Several families of axon guidance signals, along with their cognate receptors, have been described so far. In this section, we will present these molecules and the way by which their role in axon guidance has been discovered. However, even if these ligand–receptor pairs have been the stars of the show, there are probably more molecules in this ‘‘theater troupe.’’ A main challenge for the future will be to understand how these secondary players can help and cooperate to finely modulate the axon guidance process.

A. Axon Growth Cones and Endothelial Tip Cells Andreas Vesalius illustrated already five centuries ago that, although functionally distinct, the vascular and the nervous systems are architecturally similar and are structured into a ramifying and a hierarchically ordered network. Both systems are composed of largely separate eVerent and aVerent networks (i.e., motor and sensory nerves in the nervous system, and arteries and veins in the vasculature) (Carmeliet and Tessier‐Lavigne, 2005). Only today, scientific evidence is emerging that vessels, which arose later in evolution than nerves, co‐opted several organizational principles and molecular mechanisms that evolved to wire up the nervous system. Apart from the similarities already described for neural and EC cell fate, neuro‐ and angiogenesis, and tissue boundary formation, we can also find striking similarities in the way axons and ECs find their way to their final

27

1. Angiogenesis and Neurogenesis A

Eph Ephrin

Axonal growth cone Ephrin Eph

Robo Slit

Npn/Plexin UNC5/DCC

Semaphorins Npn/Plexin

Robo

Ephrin Eph

Netrins UNC5

B

Endothelial tip cell

C

Figure 7 Morphological and molecular similarities between axonal growth cones and endothelial tip cells. (A) The guidance of both axons (gray) and endothelial tip cells (green) is directed by four major classes of ligands and their cognate receptors. (B) Scanning electron micrograph of an axonal growth cone, terminating with numerous filopodial extensions (reproduced from Wessells and Nuttall, 1978). (C) Multiphoton imaging of tip cell filopodia extending from the dorsal aorta in a 22‐hour‐old zebrafish embryo.

destination. A highly motile structure located at the tip of the axon, the growth cone, is the key player in axon pathfinding (Chilton, 2006; Fig. 7). By extending filopodia and lamellipodia, the growth cone senses the microenvironment and subsequently responds to a variety of guidance cues. Like that, the growth cone reassesses its spatial environment and accurately selects a correct trajectory among the maze of possible routes (Wen and Zheng, 2006). In the vascular system, an emerging vessel is composed of specialized endothelial ‘‘tip’’ cells present at the forefront and stalk cells located behind (Gerhardt et al., 2003; Fig. 7). Notch signaling has been shown to participate in the specification of endothelial tip cells at the forefront of the vascular

28

Zacchigna et al.

sprouts versus endothelial stalk cells trailing these tip cells in the nascent vessels (Sainson et al., 2005). Endothelial tip cells share many similarities with axonal growth cones, and also explore their environment by extending and retracting numerous filopodia in saltatory fashion, suggesting that they direct the extension of vessel sprouts (Gerhardt et al., 2003). Endothelial tip cells, which proliferate minimally, ‘‘pave the path’’ for the subjacent ‘‘stalk’’ ECs of the growing vessel. In contrast, stalk cells proliferate extensively while migrating in the wake of the tip cell, thus permitting extension of the nascent vessel.

B. Common Signals for Axon and Blood Vessel Wiring Axons and vessels migrate to their final destination in a highly stereotyped pattern. The long trajectory that growing axons follow to reach their final target is split into multiple checkpoints that divide the path into a series of shorter decision‐making events (Autiero et al., 2005). Like that, axons simplify their task and navigate from one ‘‘intermediate target’’ or ‘‘choice point’’ to the other (Chilton, 2006). They are usually attracted to a choice point by long‐ range attracting signal produced by the intermediate target; once there, they are expelled by short‐ or long‐range repellents also produced by cells at the choice point (Chilton, 2006), in order to continue their trajectory. Like axons, ECs have to migrate over short distances (similar to the short segments in axons) to reach their final destination. Findings show that molecules that were originally identified as axon guidance cues also play a role in blood vessel guidance (Carmeliet and Tessier‐Lavigne, 2005; Fig. 8). Furthermore, axons and vessels often take advantage of one another to follow the same path. In some cases, vessels produce signals (such as artemin and NT‐3) that attract axons to track alongside the pioneer vessel (Honma et al., 2002; Kuruvilla et al., 2004). Conversely, nerves may also produce signals such as VEGF to guide blood vessels (Mukouyama et al., 2002). Four families of axon guidance cues, acting over a short range (cell‐ or matrix‐associated signals) or long range (secreted diVusible signals), were identified in the 1990s by genetic, biochemical, and molecular studies (Fig. 8). These families are Netrins and their deleted colorectal cancer (DCC) and uncoordinated 5 (Unc‐5) receptors, Semas and their neuropilin (Npn) and plexin receptors, Slits and their Robo receptors, and ephrins and their Eph receptors (reviewed in Carmeliet, 2003; Chilton, 2006; Dickson, 2002; Huber et al., 2003). Additionally, over the last few years, members from at least three families of morphogens, previously described for their role in controlling cell fate and tissue patterning, have been shown to act also as guidance cues: the wingless/Wnt, hedgehog (Hh), and the decapentaplegic (Dpp)/BMP/TGF‐ families. The identification of these families of guidance cues has greatly improved our understanding on the mechanisms of axon guidance and on the

29

1. Angiogenesis and Neurogenesis BMP-7

rp 1

2

5

3 4 fp

Shh

1. Initial growth away from the roof plate: BMP-7/GDF-7; CS-PG 2. Projection along the lateral side of the spinal cord: TAG-1 3. Ventral and medial extension toward the floor plate: Shh/BOC; Netrin-1/DCC; Rig-1/Robo-3 4. Midline crossing toward the controlateral floor plate border: Slits/Robos; Sema-3B/Npn-2; Eph-B/ephrin-B; NrCAM; F-spondin 5. Anterior turn to rostral projection: Wnt4/fz3

Figure 8 Major guidance cues acting at the developing spinal cord midline. In the upper part, the drawing on a transverse section of the spinal cord from a mouse embryo shows the pathway of commissural axons, divided in five major segments, indicated by numbers. BMP-7 and Shh, produced by the roof plate (rp) and by the floor plate (fp), respectively, create opposite morphogen gradients, which contribute to commisural axon pathfinding in the developing spinal cord. On the bottom, a list of the known molecules that have been implicated in the diVerent phases of commissural axon guidance, as indicated by the numbers above.

wiring of the nervous system. However, many guidance events still remain weakly understood and the number of guidance molecules identified seems small, relative to the complexity of the nervous system. In the next paragraphs, we will describe the role of these cues in the process of axon guidance and their newly identified role as blood vessel guidance molecules.

1. Morphogens Morphogens are signaling molecules that are produced locally, yet act at a distance to pattern the surrounding field of cells in a concentration‐ dependent manner. Members of the Wnt, TGF‐ , Hh, and more recently

30

Zacchigna et al.

FGF families have been shown to act as morphogens in both vertebrate and invertebrate model organisms. The morphogen concentration gradient usually translates into an activity gradient, which results in the diVerential activation of target genes in diVerent cells as a function of their distance from the morphogen source. In this way, the so‐called ‘‘high‐threshold targets’’ respond only to high levels of morphogen signaling, while ‘‘low‐threshold targets’’ respond to low levels. During vertebrate CNS development, the paradigmatic example is provided by the graded function of Wnt in patterning the neural plate along its anteroposterior axis, and the antagonistic role of BMPs and Shh to specify neuronal identity along the dorsoventral axis of the entire neural tube (Bovolenta, 2005; Wilson and Houart, 2004). Interestingly, studies implicating Shh, BMP, and Wnt in the control of growth cone movement have been pushing the idea that morphogens can be reused later in development as axon guidance cues (Butler and Dodd, 2003; Charron et al., 2003; Lyuksyutova et al., 2003; Trousse et al., 2001; Wilson and Houart, 2004). However, the first evidence that morphogenetic signaling further contributes to growth cone steering originated from studies on FGF signaling on target recognition. a. Fibroblast Growth Factor. FGF proteins constitute a large family of secreted factors composed of at least 23 members, which give rise to a series of splice variants and bind to cell surface tyrosine kinase receptors, encoded by four genes (FGFR1‐4). Several FGFs and at least three receptors are expressed in the developing CNS. For instance, FGFR1 is already expressed in the cell body and growth cone of RGCs from the earliest developmental stages (Brittis et al., 1996). When leaving the eye, RGC axons form the optic nerve and reach the midline at the level of the optic chiasm, where in animals with binocular vision, they face the choice to project contralaterally or stay ipsilaterally. Extending through the optic tract, they finally establish synaptic contacts with neurons in the optic tectum or superior colliculus (in mammals). FGF signaling contributes to the proper RGC growth cone navigation in diVerent segments of this pathway, as proven by the defasciculation and growth cone dispersion observed in rats treated with FGFR‐blocking antibodies (Brittis et al., 1996). In accordance, FGF8 and FGF19 have been shown to be expressed by the optic fissure exactly at the time when RGC axons are navigating through this region (Kurose et al., 2004). An analogue mechanism seems operational in frogs, at least in the final steps of the RGC trajectory. In fact, addition of FGF stimulated neurite extension from cultured retinal neurons and induced aberrant axonal overgrowth in exposed brain preparations of Xenopus embryos (McFarlane et al., 1995). Evidence of a direct eVect of FGFs on RGC axons come from in vitro studies showing that RGC growth cones are repelled by high levels of FGF2 (Webber et al., 2003). Similar changes have been described for cortical neurons that, in response to

1. Angiogenesis and Neurogenesis

31

FGF2, increased the size of their growth cone, slowed their migration, and increased axon branching (Szebenyi et al., 2001). Finally a peculiar role of FGF in the regulation of axonal growth has been shown during trochlear motoneuron pathfinding (Irving et al., 2002). Trochlear motoneurons, which innervate eye muscles, send axons that project toward and extend within the isthmus at the midbrain–hindbrain boundary, where Fgf8 expression is required for a correct morphogenetic patterning in all vertebrates (Martinez et al., 2001). Notably, FGF8 is later used to guide trochlear axons out of the neural tube to form the fourth cranial nerve, as proven by the observation that an antiserum against FGF8 or FGFR blockers causes axon misguidance and defasciculation (Irving et al., 2002).

b. Sonic Hedgehog. Another morphogen involved in midline crossing of diVerent types of axons is Shh. As described later in more detail, the major chemoattractant for commissural axons at the ventral midline of spinal cord is Netrin‐1. Indeed, in mice mutants for Netrin‐1 or its receptor DCC, many commissural axon trajectories are foreshortened and misguided, failing to invade the ventral spinal cord (Fazeli et al., 1997; Serafini et al., 1996). However, some of them still reach the midline, indicating that other guidance cues cooperate with Netrin‐1 to guide these axons. Further analysis of Netrin‐1 knockout mice suggested that the floor plate produces additional diVusible attractants for commissural axons. Given its expression by the floor plate and its long‐range eVects already established in the spinal cord, Shh was a candidate for a midline‐derived axonal guidance cue and it was indeed shown to mimic the Netrin‐1 axonal chemoattractant activity of the floor plate in vitro and in vivo (Charron et al., 2003). A putative receptor for Shh, named biregional Cdon‐binding protein (Boc), has been identified on commissural axons (Okada et al., 2006). Shh has also been proposed to act as a negative regulator of RGC axon growth (Torres et al., 1996). Inactivation of the murine Pax2 gene alters the development of the optic chiasm in a way that RGC axons never cross the midline. Notably, while Shh is downregulated in the chiasm during RGC axon migration in wild‐type mice, Shh expression is maintained in the Pax2‐mutant mice (Torres et al., 1996), suggesting that the continuous expression of Shh in the midline region might impair RGC axon growth, preventing them from crossing the midline. In accordance, the ectopic expression of Shh at the midline is able to prevent RGC axons from crossing the midline, without aVecting patterning and neural diVerentiation in the eye (Trousse et al., 2001). The apparent contradiction between the eVects of Shh on commissural and retinal axons (attraction vs repulsion, respectively) may be related to the involvement of distinct signaling pathways, resulting in opposite guidance eVects, as we will further discuss Netrins in the following paragraphs.

32

Zacchigna et al.

c. Bone Morphogenetic Proteins. An additional intriguing observation in the Netrin‐1 and DCC mutants was that commissural axons initially migrated ventrally for one‐third of their normal trajectory before becoming misrouted (Fazeli et al., 1997; Serafini et al., 1996), indicating that additional cues might guide their migration in the dorsal part. The proximity of commissural neurons to the roof plate suggested that the roof plate itself might repel commissural axons. Among a battery of candidate molecules, BMP‐7 and BMP‐6, both expressed by the roof plate, can indeed mimic the chemorepellent activity of the roof plate in vitro (Augsburger et al., 1999). Further genetic studies indicated that the roof plate BMP‐related chemorepellent activity, which guides the initial trajectory of commissural axons in the developing spinal cord, consists of a BMP‐7 and growth diVerentiation factor‐7 (GDF‐7) heterodimer (Charron and Tessier‐Lavigne, 2005). BMPs and the related GDFs are signaling molecules of the TGF‐ superfamily, required for the specification of the spinal cord dorsal interneurons, including commissural neurons, the choroid plexus in the forebrain, and granule cells in the cerebellum (Alder et al., 1999; Hebert et al., 2002; Lee et al., 1998). Thus, gradients of BMPs ad Shh appear to cooperate at least twice during the development of the neural tube: first in the specification of cell fate, and later to guide commissural axons to the ventral midline. Whereas a single Shh molecule seems to play both roles, it remains to be determined whether the same BMP family members can accomplish both functions or, instead, whether diVerent BMP molecules independently perform each role. d. Wnt. The last class of morphogens implicated in axon guidance is the Wnt family of proteins, which are secreted molecules implicated in tissue patterning, as well as in cell proliferation and diVerentiation in a variety of tissues. By binding to the frizzled (Fz) receptors, Wnt activates several signaling pathways, which ultimately result in changes in both gene expression and cell adhesion. The property of Wnt proteins to rearrange the cytoskeleton during axonal growth cone extension suggested that they might also be involved in axon guidance. Intriguing evidence of a guidance role for Wnt protein was obtained by studying the nervous system development in Drosophila, where commissural axons have to choose between projecting either through the anterior or through the posterior commissure. A major role of the derailed (Drl) receptor in this decision has been established, based on the initial observation that its expression is restricted to the growth cones of axons that project into the anterior commissure (Callahan et al., 1995). Indeed, in the absence of Drl receptor, neurons that normally cross the anterior commissure often project to the posterior commissure. Conversely, ectopic expression of Drl receptor in posterior neurons forces them to project in the anterior commissure (Bonkowsky et al., 1999). Thus, Drl receptor seems to be necessary and suYcient to direct axons into the anterior commissure.

1. Angiogenesis and Neurogenesis

33

The fact that Drl receptor contains a Wnt inhibitory factor domain led to propose a model in which a Wnt protein might act as a repellent at the posterior commissure, forcing axons to project anteriorly by binding Drl receptor. In accordance with this idea, loss of Wnt5 function resulted in commissural axon defects similar to those observed in drl mutants, whereas overexpression of Wnt5 throughout the midline prevented the formation of the anterior commissure, but not in drl mutants. Subsequent biochemical work demonstrated that Drl receptor actually functions as a receptor for Wnt5 (Yoshikawa et al., 2003), providing the first evidence of a ligand for the Drl family of receptors and suggesting that other members of the family might also act as Wnt receptors. Once at the contralateral side of the floor plate, commissural axons have to make a sharp turn projecting in a rostral direction—which are the cues that control this step? Wnt4, which is expressed in an increasing posterior to anterior gradient, at least at the mRNA level, appears to be a main player in this pathway. Ectopic expression of Wnt4 was found to redirect postcrossing axons in vitro, and soluble Wnt inhibitors induced postcrossing commissural axons to stall and turn randomly along the anteroposterior axis (Lyuksyutova et al., 2003). In addition to Wnt4, Shh has also been identified as a major candidate to guide commissural axons in the rostral direction along the longitudinal axis of the spinal cord in the chick embryo (Bourikas et al., 2005). Indeed, inhibiting Shh activity by RNAi or blocking antibodies led to axon stalling, with some axons turning caudally or rostrally, apparently in a random manner. Finally, postcrossing commissural axons were shown to avoid ectopic Shh in vivo (Bourikas et al., 2005), providing strong evidence that Shh is essential for the normal guidance of commissural axons along the longitudinal axis of the spinal cord. Although it is not yet known whether Shh guides postcrossing commissural axons in rodents, nor whether Wnt4 guides postcrossing commissural axons in chick, it seems particularly interesting to note that complementary Wnt4 and Shh gradients might act in a cooperative manner in the rostral guidance of commissural axons (Charron and Tessier‐Lavigne, 2005). 2. Netrins and Their Unc‐5 and DCC Receptors Netrins are a family of proteins highly conserved from C. elegans to mammals. In mammals, there are four Netrin homologues (Netrin‐1, ‐2, ‐3, and ‐4). Their structures share similarities with the short arms of laminin‐ (Netrin‐1 and ‐3) or ‐chains (Netrin‐4). They contain a laminin VI domain, three EGF‐ like repeats, and a C‐terminal domain that can bind heparin, heparan sulfate proteoglycans, or membrane glucolipids, thereby allowing interaction with components of the extracellular matrix or the cell surface (Barallobre et al., 2005). The extent of their diVusion is determined by both their expression level

34

Zacchigna et al.

and the concentration of binding sites in the surrounding tissue. Netrins bind two families of receptors, each with a single transmembrane domain. In vertebrates, the DCC receptor family comprises DCC and Neogenin, which share homology to Unc‐40 in C. elegans and frazzled in Drosophila (Barallobre et al., 2005). The other family of receptors that bind Netrins is the Unc‐5 family. Vertebrates have four homologues, Unc‐5‐A, ‐B, ‐C, and ‐D, which are orthologues of Unc‐5 in C. elegans (Barallobre et al., 2005). Netrins have a dual role in axon guidance; they can act either as attractant or as repellent molecules (Barallobre et al., 2005). Signaling of Netrins through DCC receptors induces axon attraction, while repulsion is generated by the binding of Netrins to Unc‐5 receptors (short‐range repulsion) or to a combination of Unc‐5 and DCC receptors (long‐range repulsion) (Hong et al., 1999; Keleman and Dickson, 2001). As mentioned previously, in species with bilateral symmetry, neurons connect from one side of the CNS to the other by projecting axons across the midline via commissures. In this way, a proper and coordinated function of the brain is ensured. Netrin‐1 is expressed in the floor plate and in neuroepithelial cells of the ventral region of the spinal cord during development (Barallobre et al., 2005). The analysis of Netrin‐1‐deficient mice showed that commissural axons, which express DCC on their surface, although they started to grow toward the floor plate later on their trajectory they were stalled or misrouted on their way to the midline (Serafini et al., 1996). A similar phenotype was found with DCC (Fazeli et al., 1997; Serafini et al., 1996). In wild type embryos, once axons have reached the midline, their response to the chemoattractant activity of Netrin‐1 is silenced to avoid stalling at the midline. This is achieved by the commissureless (Comm) receptor in Drosophila and by Robo‐3 in mouse (Section 3), whose expression levels become higher in postcrossing commissural axons, by forming a complex with DCC that inactivates Netrin‐1 attractant activity (Stein et al., 2001). The analysis of developing ISVs in zebrafish embryos revealed an unexpected role for Unc‐5b and Netrin‐1a in vessel guidance. The initial sprouting of the ISVs in zebrafish knockdown of either Unc‐5b or Netrin‐1a was unaVected, but when they reached the level of either the horizontal myoseptum or the floor plate (which normally express Netrin‐1a), they deviated laterally instead of extending dorsally (Lu et al., 2004). Furthermore, capillary branching was increased, which together with the ISV defects resembles the phenotype observed in Unc‐5b‐deficient mice (Lu et al., 2004). When injecting recombinant Netrin‐1 into hindbrains of E10.5 wild‐type mouse embryos, a marked retraction of the tip cell filopodia occurred, compared to injected control. In addition, this eVect was abolished in Unc‐5b knockout mice (Lu et al., 2004). These results suggest that Netrin‐1, by binding to Unc‐5B, inhibits vessel branching at specific ‘‘signaling points.’’ In contrast, other

1. Angiogenesis and Neurogenesis

35

studies proposed an attractive role for Netrin‐1. In contrast, using also zebrafish as animal model, it was shown that, after knockdown of Netrin‐1a, the ISVs and DLAVs formed normally but the formation of the PAVs was inhibited (Park et al., 2004b; Wilson et al., 2006). The authors proposed that Netrin‐1a is required to induce EC migration alongside muscle pioneer cells when forming the PAV (Park et al., 2004b; Wilson et al., 2006). In the same study, Netrins also promoted neovascularization and reperfusion in a murine model of peripheral vascular disease (hind limb ischemia) (Wilson et al., 2006). Why has Netrin‐1 been described to have repulsive (Lu et al., 2004) and attractive activities (Wilson et al., 2006) in blood vessel guidance? One possible reconciling hypothesis is that depending on the receptor to which it binds, Netrin‐1 may act as a repulsive or attractant cue for ECs. 3. Slits and Their Roundabout Receptors Slits are proteins that are highly conserved from C. elegans to vertebrates (Brose and Tessier‐Lavigne, 2000). They have multiple binding domains, including four leucine‐rich repeats (LRRs), nine EGF‐like repeats (seven in Drosophila), and a C‐terminal cystenin knot. In Drosophila, there is only one Slit; in contrast in mammals, three family members have been identified (Slit‐1, ‐2, and ‐3). Slits signal through binding single transmembrane receptors of the roundabout or Robo family (Kidd et al., 1998). These receptors contain an extracellular region with five immunoglobulin (Ig) domains and three fibronection type III repeats. In Drosophila, Slit binds to one Robo receptor; in vertebrates, four Robo receptors (Robo‐1, ‐2, ‐3, and ‐4) are known, with Robo‐4 (also known as magic roundabout) being structurally divergent from the other Robos. Slits are expressed in the nervous system midline (Brose and Tessier‐ Lavigne, 2000). They have been described to repel certain axons but, conversely, also stimulate branching and elongation of others (Kidd et al., 1999; Li et al., 1999; Wang et al., 1999). Slit proteins have been shown to regulate midline guidance in Drosophila and vertebrates. In flies, Slit is expressed at the ventral midline, where it acts (through Robo) as a short‐range repellent to prevent ipsilateral axons from crossing the midline and commissural axons from recrossing (Kidd et al., 1999). In flies lacking Slit, axons that normally do not cross the midline do so, and axons that cross it only once can then cross it several times. Midline defects at the optic chiasm and in major forebrain tracts were also observed in mice lacking Slit‐1 and Slit‐2, yet spinal commissural axons appeared unaVected (Plump et al., 2002). The analysis of a triple mouse Slit knockout showed that when all six Slit alleles are removed, commissural axons reached the midline, but then many of them failed to leave it, while others recross it (Long et al., 2004).

36

Zacchigna et al.

If commissural axons are attracted to the midline by Netrin‐1, how can they cross it if they are also repelled by Slits? A highly controlled mechanism guarantees that Slits expel crossing axons only after and not before they cross the midline. Several mechanisms have been proposed to underlay this switch. First, Robo receptors are expressed at low levels in precrossing commissural axons (Kidd et al., 1998; Long et al., 2004). Second, Robo receptors are expressed in precrossing commissural axons but are forming inactive complexes (Sabatier et al., 2004). In both ways, precrossing commissural axons are insensitive to Slit repulsion. In Drosophila, the regulatory protein Comm keeps the Robo receptor intracellularly away from the plasma membrane, thereby lowering Robo surface expression in precrossing commissural axons (Keleman et al., 2002, 2005). Once commissural axons have crossed the midline, Comm repression is lost and Robo becomes expressed at the axon surface. Consequently, axons become sensitive to Slits and are expelled from the midline. In mammals, Robo‐1 and Robo‐3 are expressed in precrossing commissural axons. Robo‐3 functions as an ‘‘anti‐Robo’’ as it plays a similar role as Comm in Drosophila: it silences Robo‐1 and blocks the binding of Slits to Robo‐1, thereby eliminating its repulsive activity (Sabatier et al., 2004). After crossing, Robo‐3 remains expressed but Robo‐1 and Robo‐2 become upregulated (Sabatier et al., 2004). Like this, they overwhelm the negative regulation of Robo‐3 and ensure that Slit‐mediated repulsion in axons starts only after they have crossed the midline (Sabatier et al., 2004). Robo‐4 was described as a vascular‐specific Robo homologue (Park et al., 2003) which is selectively expressed in developing blood vessels during embryonic development (Park et al., 2003) and expressed only at sites of active angiogenesis including tumor vessels during adulthood (Huminiecki et al., 2002). A Robo‐4 knockdown study in zebrafish showed that some Robo‐ 4‐expressing ISVs failed to sprout from the dorsal aorta or arrested midway through their dorsal migration path, whereas others deviated from their normal dorsal trajectory (Bedell et al., 2005). By in vitro gain of function approaches, Kaur et al. (2006) showed that Robo‐4 activates Cdc42 and Rac Rho GTPases in ECs. When they knocked down Robo‐4 in zebrafish, they could observe lower amounts of active Cdc42 and Rac1 as well as a lack of direction of isolated Robo‐4 knock down (KD) angioblasts (Kaur et al., 2006). Although it is not clear what ligand is signaling through Robo‐4 to create that phenotype, in vitro experiments showed that Robo‐4 on human ECs bound soluble Slit‐2 and that this binding inhibited EC migration, suggesting a repulsive role for Robo‐4/Slit‐2 in angiogenesis (Park et al., 2003). However, another study failed to detect binding of Slit‐2 to Robo‐4 (Suchting et al., 2005). More evidences demonstrating a role for Slits and Robos in angiogenesis come from in vitro studies in which the exposure of Robo‐1‐positive human umbilical vein ECs (HUVEC) to a Slit‐2 source stimulated their chemotaxis. In vivo experiments also supported an attractive role for Slit‐2 on

1. Angiogenesis and Neurogenesis

37

Robo‐1‐expressing vessels (Wang et al., 2003). These results are thus in disagreement with the documented repulsive activity of Slit‐2 when binding to Robo‐4, as mentioned above (Park et al., 2003). Additional studies will help to clarify the role of Slit‐2 in developmental and pathological angiogenesis and explain whether Slit‐2 could have opposite eVects (repulsion or attraction) depending on the Robo receptor subtype to which it binds. 4. Semas and Their Npn and Plexin Receptors To date, more than 20 Semas have been identified and categorized, according to sequence similarities and structural properties, into 8 classes. They belong to a large family of both membrane and secreted proteins, characterized by the presence of a highly conserved 500‐amino acid extracellular domain (Sema domain), which mediates the binding to multimeric receptor complexes, mainly composed of plexins and Npns, but often including additional molecules (Suchting et al., 2006). Generally, membrane‐associated Semas bind directly to plexin receptors, while class III secreted Semas (Sema‐3A‐F) require Npn receptors, which seem not to signal themselves but act as coreceptors for plexin signaling. Originally, genetic studies in Drosophila and mice implied Semas as major cues in axon guidance and neuronal cell migration. In general, they are considered to act as repellents, though Sema‐3A can also function as a chemoattractant, depending on the intracellular levels of cyclic nucleotides (Carmeliet and Tessier‐Lavigne, 2005). The main receptors for Semas in the nervous system are plexins, either alone or complexed with Npn receptors. As the intracellular domain of Npns is extremely short, they associate with plexins in order to induce an intracellular signaling mechanism. While Sema‐3A binds only Npn‐1, other members of the family, such as Sema‐3B, Sema‐3C, and Sema‐3F bind both Npn‐1 and Npn‐2 (Chen et al., 1998; Takahashi et al., 1998). Surprisingly, Npn‐1 and Npn‐2 were also found to be expressed by ECs and to associate as coreceptors with VEGFR‐1 and ‐2. The expression of Npn‐1 in ECs increases the aYnity of VEGF164 for VEGFR‐2, thus enhancing VEGFR‐2 signaling, leading to EC chemotaxis and other angiogenic steps (Miao et al., 1999). In contrast, when complexed with VEGFR‐1, Npn‐1 seems to prevent the binding of VEGF to this receptor (Fuh et al., 2000), but the general relevance of this finding remains to be determined. The heparin‐binding form of PlGF (PlGF‐2) and VEGF‐B, two additional members of the VEGF family, also bind Npn‐1 (Makinen et al., 1999; Migdal et al., 1998). Npn‐2 was shown to interact with VEGFR‐2 and ‐3 and act as a coreceptor to enhance EC response to VEGF and VEGF‐C (Favier et al., 2006). Therefore, Npns bind to two unrelated ligand families, the Sema family and the VEGF family, which suggests the existence of common molecular mechanisms in these two biological processes.

38

Zacchigna et al.

Sustaining such a neurovascular link is the fact that VEGF antagonizes the proapoptotic and collapsing eVect of Sema‐3A on axons (Gu et al., 2002). On the other hand, ECs respond to Sema‐3A by decreasing their migratory capacity, as well as microvessels and lamellipodia formation, eVects can be reversed by VEGF (Miao et al., 1999). Moreover, VEGF induces the proliferation of diVerent tumor cell lines, while Sema‐3A exerts a proapoptotic eVect on the same cells (Guttmann‐Raviv et al., 2006). The opposing eVect of VEGF and Sema‐3A suggests that (1) they compete for overlapping binding sites in the extracellular domain of a series of shared receptors or (2) they provide independent, opposite intracellular signals to their target cells. Using transgenic mice which express a mutant Npn‐1 that is unable to bind to Sema‐3A, but still able to bind VEGF (npn‐1Sema mice), it was shown that neural morphogenesis was severely aVected without any eVect in the vascular system, indicating that Sema‐3A/ Npn‐1 is dispensable for vascular development (Gu et al., 2003). Moreover, vascular malformations were caused by conditional silencing of Npn‐1 in ECs presumably because of an impaired VEGF/ Npn‐1 signaling. A role for Sema‐3C in the vascular system has also been described (Banu et al., 2006). Sema‐3C was shown to regulate glomerular EC function by stimulating integrin phosphorylation and VEGF120 secretion (Banu et al., 2006). The receptor Plexin‐B1, which is widely expressed in the nervous tissues, is also expressed in adult ECs (Basile et al., 2004). The signaling Sema‐4D/ Plexin‐B1 induces repulsion in developing axons and maintenance of established neural pathways in the adult (Kruger et al., 2005). In contrast, the binding of Sema‐4D to Plexin‐B1 induces tubulogenesis and migration of ECs, and angiogenesis in vivo (Basile et al., 2004). Furthermore, Sema‐4D was shown to be expressed in head and neck squamous cell carcinoma (HNSCCs) and to stimulate tumor angiogenesis (Basile et al., 2006). Results obtained in zebrafish and mouse embryos confirmed that Plexin‐D1 is involved in vessel morphogenesis (Weinstein, 2005). Plexin‐D1 is expressed in ECs in zebrafish embryos; complementary, class III Semas are expressed in somites (Weinstein, 2005). These Semas, by binding to Plexin‐D1, act as a repulsive cue for ECs, permitting them to select the appropriate ISV branching site (Weinstein, 2005). Interestingly, Sema‐3E and Plexin‐D1 mouse mutant embryos exhibit similar vascular phenotypes, suggesting that Sema‐3E signals through Plexin‐D1 to restrict blood vessel growth to the intersomitic boundaries. However, Plexin‐ D1‐deficient mice die shortly after birth due to major defects in the cardiac outflow tract, but Sema‐3E‐deficient mice are viable. These phenotypic diVerences suggest that, apart from Sema‐3E, other ligands are required for proper cardiovascular patterning. In this term, it was proposed that morphogenesis of the outflow tract requires coordinated signaling of VEGF through VEGFR‐2/ Npn‐1, and of Sema‐3A and Sema‐3C through Plexin‐D1/ Npn‐1 and Plexin‐D1/ Npn‐2, respectively (Eichmann et al., 2005).

1. Angiogenesis and Neurogenesis

39

5. Ephrins and Their EPH Receptors Eph receptors and ephrins are membrane‐bound proteins that function as a receptor–ligand pair, with 16 Eph receptors and 9 ephrins identified so far in mammals. According to distinct structural properties, Eph receptors and their ephrin ligands are classified into A and B subfamilies. Although there is a considerable crosstalk between A and B family members, type A ephrins preferentially bind Eph‐A receptors and type B ephrins bind Eph‐B receptors (Heroult et al., 2006). Ephs are generally described as receptors and ephrins as ligands, but it is now known that their interaction initiates bidirectional signals in both the Eph‐ and the ephrin‐expressing cells (forward and reverse signaling, respectively) (Kullander and Klein, 2002). The direct interaction between ephrins and Eph receptors provides adhesive forces between cells, whereas more complex interactions and coupling with intracellular signaling molecules translate such contacts into both repulsive signals between adjacent cells and attractive guidance cues for cell migration (Janes et al., 2005; Zimmer et al., 2003). It has been proposed that Eph receptors might act in a bimodal manner, being capable of transmitting both proadhesive as well as antiadhesive signals. In particular, reverse ephrin‐B signaling has been implicated in both attractive and repulsive functions (Kullander and Klein, 2002), suggesting that Eph‐B receptors are able to transmit both propulsive and repulsive signals on Eph‐B/ephrin‐B interacting cells. The Eph receptors and ephrins were first identified as repellent axon guidance molecules in the retinotectal projection system. Axons from the temporal retina were shown to express a high density of Eph‐A receptors and to project to the anterior colliculus, where expression of the ephrin‐A repellent is low (Carmeliet and Tessier‐Lavigne, 2005). On the other hand, axons from the nasal retina were shown to express low Eph‐A levels and to project to the posterior colliculus, where abundant expression levels of the ligand‐ repellent ephrin‐A were detected. Later, they have been implicated in cell migration and positioning, axonal outgrowth, axon guidance, axon fasciculation, and also angiogenesis (O’Leary and Wilkinson, 1999). Apart from the role of ephrins–Ephs in demarcating arteriovenous cell boundaries described previously, they have also been involved in blood vessel guidance. EC mixing experiments support a model for the action of ephrin‐B2 and Eph‐B4 in blood vessel guidance, whereby signaling via ephrin‐B2 and Eph‐B4 leads to propulsive and repulsive eVects on ECs, respectively (Hamada et al., 2003). In zebrafish, ephrin‐B2 is expressed in somites, where it prevents Eph‐B3/Eph‐B4‐expressing ISVs from entering somites, thus providing short‐range guidance cues for vessels to navigate through tissue boundaries (Adams and Klein, 2000; Oike et al., 2002; Wang et al., 1998). It was found that the zebrafish homologue for the C. elegans max‐1 protein, which is mainly expressed in neuronal tissue and somites during development,

40

Zacchigna et al.

acts upstream of the ephrin pathway to regulate vascular patterning of the ISVs (Zhong et al., 2006). Max‐1 appears to regulate membrane localization of Ephs since, in max‐1 knockdown zebrafish cells, Eph‐B3 was shown to form abnormal aggregates within the cytoplasm, instead of becoming translocated to the cell surface (Zhong et al., 2006). 6. Vascular Endothelial Growth Factor Originally identified as a growth factor able to stimulate vascular permeability and EC proliferation, VEGF represents perhaps the best illustration of how the crosstalk between vessels and nerves aVects development and disease. Indeed, as described in the previous sections, VEGF is critically involved in the development and homeostasis of various tissues and organs. This is accomplished in part by its potent eVect on the vasculature, but certain novel functions of VEGF, such as those influencing morphogenesis and tissue survival, are probably independent from its ability to stimulate new vessel growth. Over the last 5 years, a growing body of evidence has been accumulated showing that VEGF has direct eVects on neural cells and is a critical player in neurodegeneration (Oosthuyse et al., 2001; Storkebaum and Carmeliet, 2004; Storkebaum et al., 2005). Could VEGF also have a role in neuronal cell and axon guidance? What is known so far is that VEGF (1) stimulates axonal outgrowth and improves the survival of cultured superior cervical and dorsal root ganglion neurons (Sondell et al., 1999); (2) induces neurite outgrowth from cerebrocortical neurons (Jin et al., 2006); (3) promotes the survival of mesencephalic, hippocampal, and cerebrocortical neurons (Jin et al., 2000; Matsuzaki et al., 2001; Silverman et al., 1999); and (4) promotes neurogenesis in vitro and in vivo (Jin et al., 2002). In addition, a peculiar role of VEGF in supporting the correct positioning of the facial motoneuron somata, but not of their axons, has been established (Schwarz et al., 2004). Interestingly, Sema‐3A was shown to be essential for guiding the axons of the same neurons, but not for their cell body pathfinding, indicating that VEGF and Sema‐3A, instead of competing, cooperate by patterning diVerent compartments of the same cell, to properly guide and position the somata and the axons of facial motoneurons (Schwarz et al., 2004). Additional evidence that VEGF can act as a chemoattractant on neurons comes from studies on neural progenitor cells. As already mentioned above, angiogenesis and neurogenesis occur concomitantly in the mammalian adult DG (Palmer et al., 2000), as well as in the songbird brain (Louissaint et al., 2002), where neurogenesis and neuronal migration are required for structural plasticity and learning throughout adulthood (Goldman and Nottebohm, 1983). Interestingly, clusters of proliferating cells in these neurogenic niches were found to be positive for VEGFR‐2, whereas VEGF immunoreactivity was detected in the surrounding tissue (Palmer

1. Angiogenesis and Neurogenesis

41

et al., 2000). This raised the intriguing hypothesis that VEGF could function as a common guidance cue to recruit both endothelial and neural progenitor cells to the correct sites for their final diVerentiation. In retrospect, this concept should not be particularly surprising, as VEGF and its receptors appeared first in evolution in the CNS of invertebrate species, such as worms and flies, which lack a well‐developed vascular network. In C. elegans, a family of four tyrosine kinase receptors, structurally similar to VEGFRs, has been identified. These receptors (vascular endothelial growth factor receptor or ver genes) appeared to be expressed by specialized cells of neural origin, such as glial cells, chemosensorial neurons, and neurons of the dorsal ganglia (Popovici et al., 2002). A platelet‐derived growth factor (PDGF)/ VEGF‐like growth factor (PVF‐1) has also been characterized in the worm, which has biochemical properties similar to the vertebrate PDGF and VEGF, and is able to bind to the human VEGFRs Flt‐1 and KDR, inducing angiogenesis in vertebrate model systems (Tarsitano et al., 2006). Important questions for the future are where and when the pvf‐1 gene is expressed in the worm and whether PVF‐1 binds to the four neuronal VEGFRs. Similarly, the fruit fly also expresses a receptor tyrosine kinase related to mammalian PDGF and VEGF receptors (PVR), which is required for hemocytes (primitive blood cells) to migrate in response to three VEGF hortologs (Cho et al., 2002). Loss of PVR function induces important defects in axon tract patterning and positioning of glial cells, thus further supporting the idea that VEGF and its receptors might have originated from the CNS (Olofsson and Page, 2005; Sears et al., 2003). It is predictable that these small animal model systems, devoid of a well‐established vascular network, will turn out to be extremely powerful to unravel a possible novel eVect of VEGF, as well as of other traditional angiogenic molecules, in axon guidance.

VII. Perspectives Emerging evidence has highlighted the importance of the neurovascular link. Not only blood vessels and nerves originate, develop, and branch topographically in a similar manner, but they also share common mechanisms for cell signaling and pathfinding. Perhaps the most striking similarity is between the growth cone of axons and the endothelial tip cells in blood vessels. Both play a similar role in exploring the environment and function to define the direction in which the axon or the new vascular sprout grows. Interestingly, initial observations suggest that molecular cues, previously described as axon guidance signals, might also be implied in angiogenesis. For instance, Netrins or Slits seem to act as attractant or as repellent cues for ECs, but the underlying molecular mechanisms for these eVects remain to be further resolved: Is this dual function dependent on the cellular context or the

42

Zacchigna et al.

receptor type to which they bind? In addition, more research is required to elucidate the intracellular pathways linking guidance receptor activation to cytoskeletal changes in ECs. Are the signaling cascades in neuronal guidance systems similar in ECs? Another exciting new possibility is that at least some of the molecules involved in vessel guidance similarly regulate axon guidance. For instance, VEGF, the key angiogenic factor, has been recognized to be involved in many neurobiological processes, including growth cone movement, neuronal survival, and maintenance of neuronal circuitries, suggesting a possible role for VEGF in axon guidance. From a therapeutic perspective, the discovery of the striking parallelism between vessels and nerves might also pave the way for the development of novel pro‐ and antiangiogenic therapeutic strategies. The increasing evidence that several of the molecules involved in the pathfinding of vessels and nerves are also expressed by diVerent tumor cells, and regulate tumor cell growth, motility, and invasion (Klagsbrun and Eichmann, 2005), oVers new therapeutic concepts. Initial evidence that interfering with Robo, Sema, or ephrin signaling inhibits tumor angiogenesis in diVerent animal models provides a first glimpse of this therapeutic potential (Bielenberg et al., 2006; Heroult et al., 2006; Wang et al., 2003). It also remains to be determined whether some of these molecules will be useful for stimulating the reperfusion of ischemic tissues in the clinic, an unmet medical need to date.

Acknowledgments The authors thank A. Ny for kindly providing the pictures for Figs. 5A–C. Zacchigna, S. is supported by the European Union seventh framework program via a Marie Curie intra‐ European fellowship. Ruiz de Almodovar, C. is a recipient of a fellowship from the Federation of European Biochemical Societies (FEBS).

References Adams, R. H., and Klein, R. (2000). Eph receptors and ephrin ligands. Essential mediators of vascular development. Trends Cardiovasc. Med. 10, 183–188. Alder, J., Lee, K. J., Jessell, T. M., and Hatten, M. E. (1999). Generation of cerebellar granule neurons in vivo by transplantation of BMP‐treated neural progenitor cells. Nat. Neurosci. 2, 535–540. Anderson, K. V., and Ingham, P. W. (2003). The transformation of the model organism: A decade of developmental genetics. Nat. Genet. 33(Suppl.), 285–293. Ashley, J., Packard, M., Ataman, B., and Budnik, V. (2005). Fasciclin II signals new synapse formation through amyloid precursor protein and the scaVolding protein dX11/Mint. J. Neurosci. 25, 5943–5955. Augsburger, A., Schuchardt, A., Hoskins, S., Dodd, J., and Butler, S. (1999). BMPs as mediators of roof plate repulsion of commissural neurons. Neuron 24, 127–141.

1. Angiogenesis and Neurogenesis

43

Autiero, M., De Smet, F., Claes, F., and Carmeliet, P. (2005). Role of neural guidance signals in blood vessel navigation. Cardiovasc. Res. 65, 629–638. Aybar, M. J., and Mayor, R. (2002). Early induction of neural crest cells: Lessons learned from frog, fish and chick. Curr. Opin. Genet. Dev. 12, 452–458. Bagatto, B., Pelster, B., and Burggren, W. W. (2001). Growth and metabolism of larval zebrafish: EVects of swim training. J. Exp. Biol. 204, 4335–4343. Bagnard, D., Vaillant, C., Khuth, S. T., Dufay, N., Lohrum, M., Puschel, A. W., Belin, M. F., Bolz, J., and Thomasset, N. (2001). Semaphorin 3A‐vascular endothelial growth factor‐165 balance mediates migration and apoptosis of neural progenitor cells by the recruitment of shared receptor. J. Neurosci. 21, 3332–3341. Banu, N., Teichman, J., Dunlap‐Brown, M., Villegas, G., and Tufro, A. (2006). Semaphorin 3C regulates endothelial cell function by increasing integrin activity. FASEB J. 20, 2150–2152. Barallobre, M. J., Pascual, M., Del Rio, J. A., and Soriano, E. (2005). The Netrin family of guidance factors: Emphasis on Netrin‐1 signalling. Brain Res. Brain Res. Rev. 49, 22–47. Basile, J. R., Barac, A., Zhu, T., Guan, K. L., and Gutkind, J. S. (2004). Class IV semaphorins promote angiogenesis by stimulating Rho‐initiated pathways through plexin‐B. Cancer Res. 64, 5212–5224. Basile, J. R., Castilho, R. M., Williams, V. P., and Gutkind, J. S. (2006). Semaphorin 4D provides a link between axon guidance processes and tumor‐induced angiogenesis. Proc. Natl. Acad. Sci. USA 103, 9017–9022. Battye, R., Stevens, A., and Jacobs, J. R. (1999). Axon repulsion from the midline of the Drosophila CNS requires slit function. Development 126, 2475–2481. Beach, D. H., and Jacobson, M. (1979). Patterns of cell proliferation in the retina of the clawed frog during development. J. Comp. Neurol. 183, 603–613. Beattie, C. E. (2000). Control of motor axon guidance in the zebrafish embryo. Brain Res. Bull. 53, 489–500. Beattie, C. E., Granato, M., and Kuwada, J. Y. (2002). Cellular, genetic and molecular mechanisms of axonal guidance in the zebrafish. Results Probl. Cell DiVer. 40, 252–269. Bedell, V. M., Yeo, S. Y., Park, K. W., Chung, J., Seth, P., Shivalingappa, V., Zhao, J., Obara, T., Sukhatme, V. P., Drummond, I. A., Li, D. Y., and Ramchandran, R. (2005). roundabout4 is essential for angiogenesis in vivo. Proc. Natl. Acad. Sci. USA 102, 6373–6378. Bielenberg, D. R., Pettaway, C. A., Takashima, S., and Klagsbrun, M. (2006). Neuropilins in neoplasms: Expression, regulation, and function. Exp. Cell Res. 312, 584–593. Bonkowsky, J. L., Yoshikawa, S., O’Keefe, D. D., Scully, A. L., and Thomas, J. B. (1999). Axon routing across the midline controlled by the Drosophila Derailed receptor. Nature 402, 540–544. Bourikas, D., Pekarik, V., Baeriswyl, T., Grunditz, A., Sadhu, R., Nardo, M., and Stoeckli, E. T. (2005). Sonic hedgehog guides commissural axons along the longitudinal axis of the spinal cord. Nat. Neurosci. 8, 297–304. Bovolenta, P. (2005). Morphogen signaling at the vertebrate growth cone: A few cases or a general strategy? J. Neurobiol. 64, 405–416. Brand, A. H., and Perrimon, N. (1993). Targeted gene expression as a means of altering cell fates and generating dominant phenotypes. Development 118, 401–415. Brantley, D. M., Cheng, N., Thompson, E. J., Lin, Q., Brekken, R. A., Thorpe, P. E., Muraoka, R. S., Cerretti, D. P., Pozzi, A., Jackson, D., Lin, C., and Chen, J. (2002). Soluble Eph A receptors inhibit tumor angiogenesis and progression in vivo. Oncogene 21, 7011–7026. Brittis, P. A., Silver, J., Walsh, F. S., and Doherty, P. (1996). Fibroblast growth factor receptor function is required for the orderly projection of ganglion cell axons in the developing mammalian retina. Mol. Cell Neurosci. 8, 120–128. Brose, K., and Tessier‐Lavigne, M. (2000). Slit proteins: Key regulators of axon guidance, axonal branching, and cell migration. Curr. Opin. Neurobiol. 10, 95–102.

44

Zacchigna et al.

Brown, L. A., Rodaway, A. R., Schilling, T. F., Jowett, T., Ingham, P. W., Patient, R. K., and Sharrocks, A. D. (2000). Insights into early vasculogenesis revealed by expression of the ETS‐ domain transcription factor Fli‐1 in wild‐type and mutant zebrafish embryos. Mech. Dev. 90, 237–252. Butler, S. J., and Dodd, J. (2003). A role for BMP heterodimers in roof plate‐mediated repulsion of commissural axons. Neuron 38, 389–401. Callahan, C. A., Muralidhar, M. G., Lundgren, S. E., Scully, A. L., and Thomas, J. B. (1995). Control of neuronal pathway selection by a Drosophila receptor protein‐tyrosine kinase family member. Nature 376, 171–174. Callery, E. M. (2006). There’s more than one frog in the pond: A survey of the Amphibia and their contributions to developmental biology. Semin. Cell Dev. Biol. 17, 80–92. Carmeliet, P. (1999). Developmental biology. Controlling the cellular brakes. Nature 401, 657–658. Carmeliet, P. (2000). Developmental biology. One cell, two fates. Nature 408, 43–45. Carmeliet, P. (2003). Blood vessels and nerves: Common signals, pathways and diseases. Nat. Rev. Genet. 4, 710–720. Carmeliet, P., and Tessier‐Lavigne, M. (2005). Common mechanisms of nerve and blood vessel wiring. Nature 436, 193–200. Chalfie, M., Tu, Y., Euskirchen, G., Ward, W. W., and Prasher, D. C. (1994). Green fluorescent protein as a marker for gene expression. Science 263, 802–805. Charron, F., and Tessier‐Lavigne, M. (2005). Novel brain wiring functions for classical morphogens: A role as graded positional cues in axon guidance. Development 132, 2251–2262. Charron, F., Stein, E., Jeong, J., McMahon, A. P., and Tessier‐Lavigne, M. (2003). The morphogen sonic hedgehog is an axonal chemoattractant that collaborates with netrin‐1 in midline axon guidance. Cell 113, 11–23. Chen, H., He, Z., Bagri, A., and Tessier‐Lavigne, M. (1998). Semaphorin‐neuropilin interactions underlying sympathetic axon responses to class III semaphorins. Neuron 21, 1283–1290. Chen, J. N., HaVter, P., Odenthal, J., Vogelsang, E., Brand, M., van Eeden, F. J., Furutani‐ Seiki, M., Granato, M., Hammerschmidt, M., Heisenberg, C. P., Jiang, Y. J., Kane, D. A., et al. (1996). Mutations aVecting the cardiovascular system and other internal organs in zebrafish. Development 123, 293–302. Chilton, J. K. (2006). Molecular mechanisms of axon guidance. Dev. Biol. 292, 13–24. Cho, N. K., Keyes, L., Johnson, E., Heller, J., Ryner, L., Karim, F., and Krasnow, M. A. (2002). Developmental control of blood cell migration by the Drosophila VEGF pathway. Cell 108, 865–876. Christensen, M., Estevez, A., Yin, X., Fox, R., Morrison, R., McDonnell, M., Gleason, C., Miller, D. M., 3rd, and Strange, K. (2002). A primary culture system for functional analysis of C. elegans neurons and muscle cells. Neuron 33, 503–514. Cleaver, O., and Melton, D. A. (2003). Endothelial signaling during development. Nat. Med. 9, 661–668. Compernolle, V., Brusselmans, K., Acker, T., Hoet, P., Tjwa, M., Beck, H., Plaisance, S., Dor, Y., Keshet, E., Lupu, F., Nemery, B., Dewerchin, M., et al. (2002). Loss of HIF‐ 2alpha and inhibition of VEGF impair fetal lung maturation, whereas treatment with VEGF prevents fatal respiratory distress in premature mice. Nat. Med. 8, 702–710. Cooke, J., Moens, C., Roth, L., Durbin, L., Shiomi, K., Brennan, C., Kimmel, C., Wilson, S., and Holder, N. (2001). Eph signalling functions downstream of Val to regulate cell sorting and boundary formation in the caudal hindbrain. Development 128, 571–580. Cooke, J. E., and Moens, C. B. (2002). Boundary formation in the hindbrain: Eph only it were simple. Trends Neurosci. 25, 260–267. Cooke, J. E., Kemp, H. A., and Moens, C. B. (2005). EphA4 is required for cell adhesion and rhombomere‐boundary formation in the zebrafish. Curr. Biol. 15, 536–542.

1. Angiogenesis and Neurogenesis

45

Davy, A., and Soriano, P. (2005). Ephrin signaling in vivo: Look both ways. Dev. Dyn. 232, 1–10. De Smet, F., Carmeliet, P., and Autiero, M. (2006). Fishing and frogging for anti‐angiogenic drugs. Nat. Chem. Biol. 2, 228–229. Dickson, B. J. (2002). Molecular mechanisms of axon guidance. Science 298, 1959–1964. Diehl, S., Bruno, R., Wilkinson, G. A., Loose, D. A., Wilting, J., Schweigerer, L., and Klein, R. (2005). Altered expression patterns of EphrinB2 and EphB2 in human umbilical vessels and congenital venous malformations. Pediatr. Res. 57, 537–544. Driever, W., Solnica‐Krezel, L., Schier, A. F., Neuhauss, S. C., Malicki, J., Stemple, D. L., Stainier, D. Y., Zwartkruis, F., Abdelilah, S., Rangini, Z., Belak, J., and Boggs, C. (1996). A genetic screen for mutations aVecting embryogenesis in zebrafish. Development 123, 37–46. Dupin, E., Real, C., and Ledouarin, N. (2001). The neural crest stem cells: Control of neural crest cell fate and plasticity by endothelin‐3. An. Acad. Bras. Cienc. 73, 533–545. Eichmann, A., Le Noble, F., Autiero, M., and Carmeliet, P. (2005). Guidance of vascular and neural network formation. Curr. Opin. Neurobiol. 15, 108–115. Elbashir, S. M., Harborth, J., Lendeckel, W., Yalcin, A., Weber, K., and Tuschl, T. (2001). Duplexes of 21‐nucleotide RNAs mediate RNA interference in cultured mammalian cells. Nature 411, 494–498. Eremina, V., Sood, M., Haigh, J., Nagy, A., Lajoie, G., Ferrara, N., Gerber, H. P., Kikkawa, Y., Miner, J. H., and Quaggin, S. E. (2003). Glomerular‐specific alterations of VEGF‐A expression lead to distinct congenital and acquired renal diseases. J. Clin. Invest. 111, 707–716. Etchevers, H. C., Couly, G., and Le Douarin, N. M. (2002). Morphogenesis of the branchial vascular sector. Trends Cardiovasc. Med. 12, 299–304. Favier, B., Alam, A., Barron, P., Bonnin, J., Laboudie, P., Fons, P., Mandron, M., Herault, J. P., Neufeld, G., Savi, P., Herbert, J. M., and Bono, F. (2006). Neuropilin‐2 interacts with VEGFR‐2 and VEGFR‐3 and promotes human endothelial cell survival and migration. Blood 108, 1243–1250. Fazeli, A., Dickinson, S. L., Hermiston, M. L., Tighe, R. V., Steen, R. G., Small, C. G., Stoeckli, E. T., Keino‐Masu, K., Masu, M., Rayburn, H., Simons, J., Bronson, R. T., et al. (1997). Phenotype of mice lacking functional Deleted in colorectal cancer (Dcc) gene. Nature 386, 796–804. Fire, A., Xu, S., Montgomery, M. K., Kostas, S. A., Driver, S. E., and Mello, C. C. (1998). Potent and specific genetic interference by double‐stranded RNA in Caenorhabditis elegans. Nature 391, 806–811. Flanagan‐Steet, H., Fox, M. A., Meyer, D., and Sanes, J. R. (2005). Neuromuscular synapses can form in vivo by incorporation of initially aneural postsynaptic specializations. Development 132, 4471–4481. Foo, S. S., Turner, C. J., Adams, S., Compagni, A., Aubyn, D., Kogata, N., Lindblom, P., Shani, M., Zicha, D., and Adams, R. H. (2006). Ephrin‐B2 controls cell motility and adhesion during blood‐vessel‐wall assembly. Cell 124, 161–173. Fouquet, B., Weinstein, B. M., Serluca, F. C., and Fishman, M. C. (1997). Vessel patterning in the embryo of the zebrafish: Guidance by notochord. Dev. Biol. 183, 37–48. Fricke, C., Lee, J. S., Geiger‐Rudolph, S., BonhoeVer, F., and Chien, C. B. (2001). astray, a zebrafish roundabout homolog required for retinal axon guidance. Science 292, 507–510. Fritsche, R., Schwerte, T., and Pelster, B. (2000). Nitric oxide and vascular reactivity in developing zebrafish, Danio rerio. Am. J. Physiol. Regul. Integr. Comp. Physiol. 279, R2200–R2207. Fuh, G., Garcia, K. C., and de Vos, A. M. (2000). The interaction of neuropilin‐1 with vascular endothelial growth factor and its receptor flt‐1. J. Biol. Chem. 275, 26690–26695. Furrer, M. P., Kim, S., Wolf, B., and Chiba, A. (2003). Robo and Frazzled/DCC mediate dendritic guidance at the CNS midline. Nat. Neurosci. 6, 223–230.

46

Zacchigna et al.

Gaiano, N., and Fishell, G. (2002). The role of notch in promoting glial and neural stem cell fates. Annu. Rev. Neurosci. 25, 471–490. Gale, N. W., Baluk, P., Pan, L., Kwan, M., Holash, J., DeChiara, T. M., McDonald, D. M., and Yancopoulos, G. D. (2001). Ephrin‐B2 selectively marks arterial vessels and neovascularization sites in the adult, with expression in both endothelial and smooth‐muscle cells. Dev. Biol. 230, 151–160. Gerber, H. P., Vu, T. H., Ryan, A. M., Kowalski, J., Werb, Z., and Ferrara, N. (1999). VEGF couples hypertrophic cartilage remodeling, ossification and angiogenesis during endochondral bone formation. Nat. Med. 5, 623–628. Gerety, S. S., Wang, H. U., Chen, Z. F., and Anderson, D. J. (1999). Symmetrical mutant phenotypes of the receptor EphB4 and its specific transmembrane ligand ephrin‐B2 in cardiovascular development. Mol. Cell 4, 403–414. Gerhardt, H., Golding, M., Fruttiger, M., Ruhrberg, C., Lundkvist, A., Abramsson, A., Jeltsch, M., Mitchell, C., Alitalo, K., Shima, D., and Betsholtz, C. (2003). VEGF guides angiogenic sprouting utilizing endothelial tip cell filopodia. J. Cell. Biol. 161, 1163–1177. Goldman, S. A., and Nottebohm, F. (1983). Neuronal production, migration, and diVerentiation in a vocal control nucleus of the adult female canary brain. Proc. Natl. Acad. Sci. USA 80, 2390–2394. Granato, M., van Eeden, F. J., Schach, U., Trowe, T., Brand, M., Furutani‐Seiki, M., HaVter, P., Hammerschmidt, M., Heisenberg, C. P., Jiang, Y. J., Kane, D. A., Kelsh, R. N., et al. (1996). Genes controlling and mediating locomotion behavior of the zebrafish embryo and larva. Development 123, 399–413. Grant, S., and Keating, M. J. (1986). Ocular migration and the metamorphic and postmetamorphic maturation of the retinotectal system in Xenopus laevis: An autoradiographic and morphometric study. J. Embryol. Exp. Morphol. 92, 43–69. Gray, M., Moens, C. B., Amacher, S. L., Eisen, J. S., and Beattie, C. E. (2001). Zebrafish deadly seven functions in neurogenesis. Dev. Biol. 237, 306–323. Gu, C., Limberg, B. J., Whitaker, G. B., Perman, B., Leahy, D. J., Rosenbaum, J. S., Ginty, D. D., and Kolodkin, A. L. (2002). Characterization of neuropilin‐1 structural features that confer binding to semaphorin 3A and vascular endothelial growth factor 165. J. Biol. Chem. 277, 18069–18076. Gu, C., Rodriguez, E. R., Reimert, D. V., Shu, T., Fritzsch, B., Richards, L. J., Kolodkin, A. L., and Ginty, D. D. (2003). Neuropilin‐1 conveys semaphorin and VEGF signaling during neural and cardiovascular development. Dev. Cell 5, 45–57. Guttmann‐Raviv, N., Kessler, O., Shraga‐Heled, N., Lange, T., Herzog, Y., and Neufeld, G. (2006). The neuropilins and their role in tumorigenesis and tumor progression. Cancer Lett. 231, 1–11. HaVter, P., Granato, M., Brand, M., Mullins, M. C., Hammerschmidt, M., Kane, D. A., Odenthal, J., van Eeden, F. J., Jiang, Y. J., Heisenberg, C. P., Kelsh, R. N., Furutani‐ Seiki, M., et al. (1996). The identification of genes with unique and essential functions in the development of the zebrafish, Danio rerio. Development 123, 1–36. Hamada, K., Oike, Y., Ito, Y., Maekawa, H., Miyata, K., Shimomura, T., and Suda, T. (2003). Distinct roles of ephrin‐B2 forward and EphB4 reverse signaling in endothelial cells. Arterioscler. Thromb. Vasc. Biol. 23, 190–197. Harris, R., Sabatelli, L. M., and Seeger, M. A. (1996). Guidance cues at the Drosophila CNS midline: Identification and characterization of two Drosophila Netrin/UNC‐6 homologs. Neuron 17, 217–228. Hashimoto, T., Zhang, X. M., Chen, B. Y., and Yang, X. J. (2006). VEGF activates divergent intracellular signaling components to regulate retinal progenitor cell proliferation and neuronal diVerentiation. Development 133, 2201–2210.

1. Angiogenesis and Neurogenesis

47

Hebert, J. M., Mishina, Y., and McConnell, S. K. (2002). BMP signaling is required locally to pattern the dorsal telencephalic midline. Neuron 35, 1029–1041. Heroult, M., SchaVner, F., and Augustin, H. G. (2006). Eph receptor and ephrin ligand‐ mediated interactions during angiogenesis and tumor progression. Exp. Cell Res. 312, 642–650. Higashijima, S., Hotta, Y., and Okamoto, H. (2000). Visualization of cranial motor neurons in live transgenic zebrafish expressing green fluorescent protein under the control of the islet‐1 promoter/enhancer. J. Neurosci. 20, 206–218. Hitoshi, S., Alexson, T., Tropepe, V., Donoviel, D., Elia, A. J., Nye, J. S., Conlon, R. A., Mak, T. W., Bernstein, A., and van der Kooy, D. (2002). Notch pathway molecules are essential for the maintenance, but not the generation, of mammalian neural stem cells. Genes Dev. 16, 846–858. HoVmann, J. A. (1995). Innate immunity of insects. Curr. Opin. Immunol. 7, 4–10. Holt, C. E., and Harris, W. A. (1983). Order in the initial retinotectal map in Xenopus: A new technique for labelling growing nerve fibres. Nature 301, 150–152. Hong, K., Hinck, L., Nishiyama, M., Poo, M. M., Tessier‐Lavigne, M., and Stein, E. (1999). A ligand‐gated association between cytoplasmic domains of UNC5 and DCC family receptors converts netrin‐induced growth cone attraction to repulsion. Cell 97, 927–941. Honma, Y., Araki, T., Gianino, S., Bruce, A., Heuckeroth, R., Johnson, E., and Milbrandt, J. (2002). Artemin is a vascular‐derived neurotropic factor for developing sympathetic neurons. Neuron 35, 267–282. Hoskins, S. G., and Grobstein, P. (1985). Development of the ipsilateral retinothalamic projection in the frog Xenopus laevis. I. Retinal distribution of ipsilaterally projecting cells in normal and experimentally manipulated frogs. J. Neurosci. 5, 911–919. Huber, A. B., Kolodkin, A. L., Ginty, D. D., and Cloutier, J. F. (2003). Signaling at the growth cone: Ligand‐receptor complexes and the control of axon growth and guidance. Annu. Rev. Neurosci. 26, 509–563. Huminiecki, L., Gorn, M., Suchting, S., Poulsom, R., and Bicknell, R. (2002). Magic roundabout is a new member of the roundabout receptor family that is endothelial specific and expressed at sites of active angiogenesis. Genomics 79, 547–552. Hutson, L. D., and Chien, C. B. (2002). Wiring the zebrafish: Axon guidance and synaptogenesis. Curr. Opin. Neurobiol. 12, 87–92. Huxlin, K. R., Sefton, A. J., and Furby, J. H. (1992). The origin and development of retinal astrocytes in the mouse. J. Neurocytol. 21, 530–544. Irving, C., Malhas, A., Guthrie, S., and Mason, I. (2002). Establishing the trochlear motor axon trajectory: Role of the isthmic organiser and Fgf8. Development 129, 5389–5398. Jacob, E., Drexel, M., Schwerte, T., and Pelster, B. (2002). Influence of hypoxia and of hypoxemia on the development of cardiac activity in zebrafish larvae. Am. J. Physiol. Regul. Integr. Comp. Physiol. 283, R911–R917. Janes, P. W., Saha, N., Barton, W. A., Kolev, M. V., Wimmer‐Kleikamp, S. H., Nievergall, E., Blobel, C. P., Himanen, J. P., Lackmann, M., and Nikolov, D. B. (2005). Adam meets Eph: An ADAM substrate recognition module acts as a molecular switch for ephrin cleavage in trans. Cell 123, 291–304. Jin, K., Zhu, Y., Sun, Y., Mao, X. O., Xie, L., and Greenberg, D. A. (2002). Vascular endothelial growth factor (VEGF) stimulates neurogenesis in vitro and in vivo. Proc. Natl. Acad. Sci. USA 99, 11946–11950. Jin, K., Mao, X. O., and Greenberg, D. A. (2006). Vascular endothelial growth factor stimulates neurite outgrowth from cerebral cortical neurons via Rho kinase signaling. J. Neurobiol. 66, 236–242. Jin, K. L., Mao, X. O., and Greenberg, D. A. (2000). Vascular endothelial growth factor: Direct neuroprotective eVect in in vitro ischemia. Proc. Natl. Acad. Sci. USA 97, 10242–10247.

48

Zacchigna et al.

Jin, S. W., Beis, D., Mitchell, T., Chen, J. N., and Stainier, D. Y. (2005). Cellular and molecular analyses of vascular tube and lumen formation in zebrafish. Development 132, 5199–5209. Johansen, J., Halpern, M. E., and Keshishian, H. (1989). Axonal guidance and the development of muscle fiber‐specific innervation in Drosophila embryos. J. Neurosci. 9, 4318–4332. Kaletta, T., and Hengartner, M. O. (2006). Finding function in novel targets: C. elegans as a model organism. Nat. Rev. Drug Discov. 5, 387–398. Kalimo, H., Ruchoux, M. M., Viitanen, M., and Kalaria, R. N. (2002). CADASIL: A common form of hereditary arteriopathy causing brain infarcts and dementia. Brain Pathol. 12, 371–384. Kaur, S., Castellone, M. D., Bedell, V. M., Konar, M., Gutkind, J. S., and Ramchandran, R. (2006). Robo4 signaling in endothelial cells implies attraction guidance mechanisms. J. Biol. Chem. 281, 11347–11356. Kawai, T., Takagi, N., Mochizuki, N., Besshoh, S., Sakanishi, K., Nakahara, M., and Takeo, S. (2006). Inhibitor of vascular endothelial growth factor receptor tyrosine kinase attenuates cellular proliferation and diVerentiation to mature neurons in the hippocampal dentate gyrus after transient forebrain ischemia in the adult rat. Neuroscience 141, 1209–1216. Keleman, K., and Dickson, B. J. (2001). Short‐ and long‐range repulsion by the Drosophila Unc5 netrin receptor. Neuron 32, 605–617. Keleman, K., Rajagopalan, S., Cleppien, D., Teis, D., Paiha, K., Huber, L. A., Technau, G. M., and Dickson, B. J. (2002). Comm sorts robo to control axon guidance at the Drosophila midline. Cell 110, 415–427. Keleman, K., Ribeiro, C., and Dickson, B. J. (2005). Comm function in commissural axon guidance: Cell‐autonomous sorting of Robo in vivo. Nat. Neurosci. 8, 156–163. Khokha, M. K., Chung, C., Bustamante, E. L., Gaw, L. W., Trott, K. A., Yeh, J., Lim, N., Lin, J. C., Taverner, N., Amaya, E., Papalopulu, N., Smith, J. C., et al. (2002). Techniques and probes for the study of Xenopus tropicalis development. Dev. Dyn. 225, 499–510. Kidd, T., Brose, K., Mitchell, K. J., Fetter, R. D., Tessier‐Lavigne, M., Goodman, C. S., and Tear, G. (1998). Roundabout controls axon crossing of the CNS midline and defines a novel subfamily of evolutionarily conserved guidance receptors. Cell 92, 205–215. Kidd, T., Bland, K. S., and Goodman, C. S. (1999). Slit is the midline repellent for the robo receptor in Drosophila. Cell 96, 785–794. Kim, J. H., Kim, J. H., Park, J. A., Lee, S. W., Kim, W. J., Yu, Y. S., and Kim, K. W. (2006). Blood‐neural barrier: Intercellular communication at glio‐vascular interface. J. Biochem. Mol. Biol. 39, 339–345. Kimmel, C. B. (1989). Genetics and early development of zebrafish. Trends Genet. 5, 283–288. Klagsbrun, M., and Eichmann, A. (2005). A role for axon guidance receptors and ligands in blood vessel development and tumor angiogenesis. Cytokine Growth Factor Rev. 16, 535–548. Knecht, A. K., and Bronner‐Fraser, M. (2002). Induction of the neural crest: A multigene process. Nat. Rev. Genet. 3, 453–461. Kokaia, Z., and Lindvall, O. (2003). Neurogenesis after ischaemic brain insults. Curr. Opin. Neurobiol. 13, 127–132. Kruger, R. P., Aurandt, J., and Guan, K. L. (2005). Semaphorins command cells to move. Nat. Rev. Mol. Cell Biol. 6, 789–800. Krull, C. E. (2001). Segmental organization of neural crest migration. Mech. Dev. 105, 37–45. Kullander, K., and Klein, R. (2002). Mechanisms and functions of Eph and ephrin signalling. Nat. Rev. Mol. Cell Biol. 3, 475–486. Kurose, H., Bito, T., Adachi, T., Shimizu, M., Noji, S., and Ohuchi, H. (2004). Expression of Fibroblast growth factor 19 (Fgf19) during chicken embryogenesis and eye development, compared with Fgf15 expression in the mouse. Gene Expr. Patterns 4, 687–693.

1. Angiogenesis and Neurogenesis

49

Kuruvilla, R., Zweifel, L. S., Glebova, N. O., Lonze, B. E., Valdez, G., Ye, H., and Ginty, D. D. (2004). A neurotrophin signaling cascade coordinates sympathetic neuron development through diVerential control of TrkA traYcking and retrograde signaling. Cell 118, 243–255. Landgraf, M., Bossing, T., Technau, G. M., and Bate, M. (1997). The origin, location, and projections of the embryonic abdominal motorneurons of Drosophila. J. Neurosci. 17, 9642–9655. Lawson, N. D., and Weinstein, B. M. (2002). In vivo imaging of embryonic vascular development using transgenic zebrafish. Dev. Biol. 248, 307–318. Lawson, N. D., Scheer, N., Pham, V. N., Kim, C. H., Chitnis, A. B., Campos‐Ortega, J. A., and Weinstein, B. M. (2001). Notch signaling is required for arterial‐venous diVerentiation during embryonic vascular development. Development 128, 3675–3683. Lawson, N. D., Vogel, A. M., and Weinstein, B. M. (2002). sonic hedgehog and vascular endothelial growth factor act upstream of the Notch pathway during arterial endothelial diVerentiation. Dev. Cell 3, 127–136. Lawson, N. D., Mugford, J. W., Diamond, B. A., and Weinstein, B. M. (2003). phospholipase C gamma‐1 is required downstream of vascular endothelial growth factor during arterial development. Genes Dev. 17, 1346–1351. Lee, K. J., Mendelsohn, M., and Jessell, T. M. (1998). Neuronal patterning by BMPs: A requirement for GDF7 in the generation of a discrete class of commissural interneurons in the mouse spinal cord. Genes Dev. 12, 3394–3407. Leventhal, C., Rafii, S., Rafii, D., Shahar, A., and Goldman, S. A. (1999). Endothelial trophic support of neuronal production and recruitment from the adult mammalian subependyma. Mol. Cell Neurosci. 13, 450–464. Li, H. S., Chen, J. H., Wu, W., Fagaly, T., Zhou, L., Yuan, W., Dupuis, S., Jiang, Z. H., Nash, W., Gick, C., Ornitz, D. M., Wu, J. Y., et al. (1999). Vertebrate slit, a secreted ligand for the transmembrane protein roundabout, is a repellent for olfactory bulb axons. Cell 96, 807–818. Liao, W., Ho, C. Y., Yan, Y. L., Postlethwait, J., and Stainier, D. Y. (2000). Hhex and scl function in parallel to regulate early endothelial and blood diVerentiation in zebrafish. Development 127, 4303–4313. Liu, K. S., and Fetcho, J. R. (1999). Laser ablations reveal functional relationships of segmental hindbrain neurons in zebrafish. Neuron 23, 325–335. Liu, K. S., Gray, M., Otto, S. J., Fetcho, J. R., and Beattie, C. E. (2003). Mutations in deadly seven/notch1a reveal developmental plasticity in the escape response circuit. J. Neurosci. 23, 8159–8166. Lledo, P. M., Alonso, M., and Grubb, M. S. (2006). Adult neurogenesis and functional plasticity in neuronal circuits. Nat. Rev. Neurosci. 7, 179–193. Long, H., Sabatier, C., Ma, L., Plump, A., Yuan, W., Ornitz, D. M., Tamada, A., Murakami, F., Goodman, C. S., and Tessier‐Lavigne, M. (2004). Conserved roles for Slit and Robo proteins in midline commissural axon guidance. Neuron 42, 213–223. Lorent, K., Liu, K. S., Fetcho, J. R., and Granato, M. (2001). The zebrafish space cadet gene controls axonal pathfinding of neurons that modulate fast turning movements. Development 128, 2131–2142. Louissaint, A., Jr., Rao, S., Leventhal, C., and Goldman, S. A. (2002). Coordinated interaction of neurogenesis and angiogenesis in the adult songbird brain. Neuron 34, 945–960. Lu, X., Le Noble, F., Yuan, L., Jiang, Q., De Lafarge, B., Sugiyama, D., Breant, C., Claes, F., De Smet, F., Thomas, J. L., Autiero, M., Carmeliet, P., et al. (2004). The netrin receptor UNC5B mediates guidance events controlling morphogenesis of the vascular system. Nature 432, 179–186.

50

Zacchigna et al.

Lyden, D., Young, A. Z., Zagzag, D., Yan, W., Gerald, W., O’Reilly, R., Bader, B. L., Hynes, R. O., Zhuang, Y., Manova, K., and Benezra, R. (1999). Id1 and Id3 are required for neurogenesis, angiogenesis and vascularization of tumour xenografts. Nature 401, 670–677. Lyuksyutova, A. I., Lu, C. C., Milanesio, N., King, L. A., Guo, N., Wang, Y., Nathans, J., Tessier‐Lavigne, M., and Zou, Y. (2003). Anterior‐posterior guidance of commissural axons by Wnt‐frizzled signaling. Science 302, 1984–1988. Makinen, T., Olofsson, B., Karpanen, T., Hellman, U., Soker, S., Klagsbrun, M., Eriksson, U., and Alitalo, K. (1999). DiVerential binding of vascular endothelial growth factor B splice and proteolytic isoforms to neuropilin‐1. J. Biol. Chem. 274, 21217–21222. Mann, F., and Holt, C. E. (2001). Control of retinal growth and axon divergence at the chiasm: Lessons from Xenopus. Bioessays 23, 319–326. Martinez, G., Di Giacomo, C., Sorrenti, V., Carnazza, M. L., Ragusa, N., Barcellona, M. L., and Vanella, A. (2001). Fibroblast growth factor‐2 and transforming growth factor‐beta1 immunostaining in rat brain after cerebral postischemic reperfusion. J. Neurosci. Res. 63, 136–142. Marx, M., Rutishauser, U., and Bastmeyer, M. (2001). Dual function of polysialic acid during zebrafish central nervous system development. Development 128, 4949–4958. MaschhoV, K. L., and Baldwin, H. S. (2000). Molecular determinants of neural crest migration. Am. J. Med. Genet. 97, 280–288. Matsuzaki, H., Tamatani, M., Yamaguchi, A., Namikawa, K., Kiyama, H., Vitek, M. P., Mitsuda, N., and Tohyama, M. (2001). Vascular endothelial growth factor rescues hippocampal neurons from glutamate‐induced toxicity: Signal transduction cascades. FASEB J. 15, 1218–1220. McFarlane, S., McNeill, L., and Holt, C. E. (1995). FGF signaling and target recognition in the developing Xenopus visual system. Neuron 15, 1017–1028. Meinertzhagen, I. A. (2000). Wiring the fly’s eye. Neuron 28, 310–313. Mellitzer, G., Xu, Q., and Wilkinson, D. G. (1999). Eph receptors and ephrins restrict cell intermingling and communication. Nature 400, 77–81. Mello, C. C., and Conte, D., Jr. (2004). Revealing the world of RNA interference. Nature 431, 338–342. Mi, H., Haeberle, H., and Barres, B. A. (2001). Induction of astrocyte diVerentiation by endothelial cells. J. Neurosci. 21, 1538–1547. Miao, H. Q., Soker, S., Feiner, L., Alonso, J. L., Raper, J. A., and Klagsbrun, M. (1999). Neuropilin‐1 mediates collapsin‐1/semaphorin III inhibition of endothelial cell motility: Functional competition of collapsin‐1 and vascular endothelial growth factor‐165. J. Cell Biol. 146, 233–242. Migdal, M., Huppertz, B., Tessler, S., Comforti, A., Shibuya, M., Reich, R., Baumann, H., and Neufeld, G. (1998). Neuropilin‐1 is a placenta growth factor‐2 receptor. J. Biol. Chem. 273, 22272–22278. Mikkola, H. K., and Orkin, S. H. (2002). The search for the hemangioblast. J. Hematother. Stem Cell Res. 11, 9–17. Mitchell, K. J., Doyle, J. L., Serafini, T., Kennedy, T. E., Tessier‐Lavigne, M., Goodman, C. S., and Dickson, B. J. (1996). Genetic analysis of Netrin genes in Drosophila: Netrins guide CNS commissural axons and peripheral motor axons. Neuron 17, 203–215. Monje, M. L., and Palmer, T. (2003). Radiation injury and neurogenesis. Curr. Opin. Neurol. 16, 129–134. Moon, M. S., and Gomez, T. M. (2005). Adjacent pioneer commissural interneuron growth cones switch from contact avoidance to axon fasciculation after midline crossing. Dev. Biol. 288, 474–486.

1. Angiogenesis and Neurogenesis

51

Motoike, T., Loughna, S., Perens, E., Roman, B. L., Liao, W., Chau, T. C., Richardson, C. D., Kawate, T., Kuno, J., Weinstein, B. M., Stainier, D. Y., and Sato, T. N. (2000). Universal GFP reporter for the study of vascular development. Genesis 28, 75–81. Mukouyama, Y. S., Shin, D., Britsch, S., Taniguchi, M., and Anderson, D. J. (2002). Sensory nerves determine the pattern of arterial diVerentiation and blood vessel branching in the skin. Cell 109, 693–705. Mukouyama, Y. S., Gerber, H. P., Ferrara, N., Gu, C., and Anderson, D. J. (2005). Peripheral nerve‐derived VEGF promotes arterial diVerentiation via neuropilin 1‐mediated positive feedback. Development 132, 941–952. Murray, M. J., Merritt, D. J., Brand, A. H., and Whitington, P. M. (1998). In vivo dynamics of axon pathfinding in the Drosophilia CNS: A time‐lapse study of an identified motorneuron. J. Neurobiol. 37, 607–621. Myers, P. Z., Eisen, J. S., and Westerfield, M. (1986). Development and axonal outgrowth of identified motoneurons in the zebrafish. J. Neurosci. 6, 2278–2289. Nag, S., Eskandarian, M. R., Davis, J., and Eubanks, J. H. (2002). DiVerential expression of vascular endothelial growth factor‐A (VEGF‐A) and VEGF‐B after brain injury. J. Neuropathol. Exp. Neurol. 61, 778–788. Nakae, J., Biggs, W. H., 3rd, Kitamura, T., Cavenee, W. K., Wright, C. V., Arden, K. C., and Accili, D. (2002). Regulation of insulin action and pancreatic beta‐cell function by mutated alleles of the gene encoding forkhead transcription factor Foxo1. Nat. Genet. 32, 245–253. Noramly, S., Zimmerman, L., Cox, A., Aloise, R., Fisher, M., and Grainger, R. M. (2005). A gynogenetic screen to isolate naturally occurring recessive mutations in Xenopus tropicalis. Mech. Dev. 122, 273–287. Ny, A., Koch, M., Schneider, M., Neven, E., Tong, R. T., Maity, S., Fischer, C., Plaisance, S., Lambrechts, D., Heligon, C., Terclavers, S., Ciesiolka, M., et al. (2005). A genetic Xenopus laevis tadpole model to study lymphangiogenesis. Nat. Med. 11, 998–1004. Ny, A., Autiero, M., and Carmeliet, P. (2006). Zebrafish and Xenopus tadpoles: Small animal models to study angiogenesis and lymphangiogenesis. Exp. Cell Res. 312, 684–693. O’Leary, D. D., and Wilkinson, D. G. (1999). Eph receptors and ephrins in neural development. Curr. Opin. Neurobiol. 9, 65–73. Ogg, S., Paradis, S., Gottlieb, S., Patterson, G. I., Lee, L., Tissenbaum, H. A., and Ruvkun, G. (1997). The Fork head transcription factor DAF‐16 transduces insulin‐like metabolic and longevity signals in C. elegans. Nature 389, 994–999. Oike, Y., Ito, Y., Hamada, K., Zhang, X. Q., Miyata, K., Arai, F., Inada, T., Araki, K., Nakagata, N., Takeya, M., Kisanuki, Y. Y., Yanagisawa, M., et al. (2002). Regulation of vasculogenesis and angiogenesis by EphB/ephrin‐B2 signaling between endothelial cells and surrounding mesenchymal cells. Blood 100, 1326–1333. Okada, A., Charron, F., Morin, S., Shin, D. S., Wong, K., Fabre, P. J., Tessier‐Lavigne, M., and McConnell, S. K. (2006). Boc is a receptor for sonic hedgehog in the guidance of commissural axons. Nature 444, 369–373. Olofsson, B., and Page, D. T. (2005). Condensation of the central nervous system in embryonic Drosophila is inhibited by blocking hemocyte migration or neural activity. Dev. Biol. 279, 233–243. Oosthuyse, B., Moons, L., Storkebaum, E., Beck, H., Nuyens, D., Brusselmans, K., Van Dorpe, J., Hellings, P., Gorselink, M., Heymans, S., Theilmeier, G., Dewerchin, M., et al. (2001). Deletion of the hypoxia‐response element in the vascular endothelial growth factor promoter causes motor neuron degeneration. Nat. Genet. 28, 131–138. Orte, C., Lawrenson, J. G., Finn, T. M., Reid, A. R., and Allt, G. (1999). A comparison of blood‐brain barrier and blood‐nerve barrier endothelial cell markers. Anat. Embryol. (Berl.) 199, 509–517.

52

Zacchigna et al.

Osterfield, M., Kirschner, M. W., and Flanagan, J. G. (2003). Graded positional information: Interpretation for both fate and guidance. Cell 113, 425–428. Palmer, T. D., Willhoite, A. R., and Gage, F. H. (2000). Vascular niche for adult hippocampal neurogenesis. J. Comp. Neurol. 425, 479–494. Panchision, D. M., and McKay, R. D. (2002). The control of neural stem cells by morphogenic signals. Curr. Opin. Genet. Dev. 12, 478–487. Paradis, S., Sweeney, S. T., and Davis, G. W. (2001). Homeostatic control of presynaptic release is triggered by postsynaptic membrane depolarization. Neuron 30, 737–749. Park, H. C., Shin, J., and Appel, B. (2004a). Spatial and temporal regulation of ventral spinal cord precursor specification by Hedgehog signaling. Development 131, 5959–5969. Park, K. W., Morrison, C. M., Sorensen, L. K., Jones, C. A., Rao, Y., Chien, C. B., Wu, J. Y., Urness, L. D., and Li, D. Y. (2003). Robo4 is a vascular‐specific receptor that inhibits endothelial migration. Dev. Biol. 261, 251–267. Park, K. W., Crouse, D., Lee, M., Karnik, S. K., Sorensen, L. K., Murphy, K. J., Kuo, C. J., and Li, D. Y. (2004b). The axonal attractant Netrin‐1 is an angiogenic factor. Proc Natl. Acad. Sci. USA 101, 16210–16215. Pasquale, E. B. (2005). Eph receptor signalling casts a wide net on cell behaviour. Nat. Rev. Mol. Cell Biol. 6, 462–475. Patten, I., and Placzek, M. (2000). The role of Sonic hedgehog in neural tube patterning. Cell Mol. Life Sci. 57, 1695–1708. Pelster, B., and Burggren, W. W. (1996). Disruption of hemoglobin oxygen transport does not impact oxygen‐dependent physiological processes in developing embryos of zebra fish (Danio rerio). Circ. Res. 79, 358–362. Plump, A. S., Erskine, L., Sabatier, C., Brose, K., Epstein, C. J., Goodman, C. S., Mason, C. A., and Tessier‐Lavigne, M. (2002). Slit1 and Slit2 cooperate to prevent premature midline crossing of retinal axons in the mouse visual system. Neuron 33, 219–232. Popovici, C., Isnardon, D., Birnbaum, D., and Roubin, R. (2002). Caenorhabditis elegans receptors related to mammalian vascular endothelial growth factor receptors are expressed in neural cells. Neurosci. Lett. 329, 116–120. Prokop, A. (1999). Integrating bits and pieces: Synapse structure and formation in Drosophila embryos. Cell Tissue Res. 297, 169–186. Prokop, A., and Meinertzhagen, I. A. (2006). Development and structure of synaptic contacts in Drosophila. Semin. Cell Dev. Biol. 17, 20–30. Ranganathan, R., Sawin, E. R., Trent, C., and Horvitz, H. R. (2001). Mutations in the Caenorhabditis elegans serotonin reuptake transporter MOD‐5 reveal serotonin‐dependent and ‐independent activities of fluoxetine. J. Neurosci. 21, 5871–5884. Robles, E., and Gomez, T. M. (2006). Focal adhesion kinase signaling at sites of integrin‐ mediated adhesion controls axon pathfinding. Nat. Neurosci. 9, 1274–1283. Rovainen, C. M. (1991). Labeling of developing vascular endothelium after injections of rhodamine‐dextran into blastomeres of Xenopus laevis. J. Exp. Zool. 259, 209–221. Sabatier, C., Plump, A. S., Le, M., Brose, K., Tamada, A., Murakami, F., Lee, E. Y., and Tessier‐Lavigne, M. (2004). The divergent Robo family protein rig‐1/Robo3 is a negative regulator of slit responsiveness required for midline crossing by commissural axons. Cell 117, 157–169. Sainson, R. C., Aoto, J., Nakatsu, M. N., Holderfield, M., Conn, E., Koller, E., and Hughes, C. C. (2005). Cell‐autonomous notch signaling regulates endothelial cell branching and proliferation during vascular tubulogenesis. FASEB J. 19, 1027–1029. Salvaterra, P. M., and Kitamoto, T. (2001). Drosophila cholinergic neurons and processes visualized with Gal4/UAS‐GFP. Brain Res. Gene Expr. Patterns 1, 73–82.

1. Angiogenesis and Neurogenesis

53

Schmid, A., Chiba, A., and Doe, C. Q. (1999). Clonal analysis of Drosophila embryonic neuroblasts: Neural cell types, axon projections and muscle targets. Development 126, 4653–4689. Schwarz, Q., Gu, C., Fujisawa, H., Sabelko, K., Gertsenstein, M., Nagy, A., Taniguchi, M., Kolodkin, A. L., Ginty, D. D., Shima, D. T., and Ruhrberg, C. (2004). Vascular endothelial growth factor controls neuronal migration and cooperates with Sema3A to pattern distinct compartments of the facial nerve. Genes Dev. 18, 2822–2834. Schwerte, T., and Fritsche, R. (2003). Understanding cardiovascular physiology in zebrafish and Xenopus larvae: The use of microtechniques. Comp. Biochem. Physiol. A. Mol. Integr. Physiol. 135, 131–145. Sears, H. C., Kennedy, C. J., and Garrity, P. A. (2003). Macrophage‐mediated corpse engulfment is required for normal Drosophila CNS morphogenesis. Development 130, 3557–3565. Seeger, M., Tear, G., Ferres‐Marco, D., and Goodman, C. S. (1993). Mutations aVecting growth cone guidance in Drosophila: Genes necessary for guidance toward or away from the midline. Neuron 10, 409–426. Seifert, M., Schmidt, E., and Baumeister, R. (2006). The genetics of synapse formation and function in Caenorhabditis elegans. Cell Tissue Res. 326, 273–285. Serafini, T., Colamarino, S. A., Leonardo, E. D., Wang, H., Beddington, R., Skarnes, W. C., and Tessier‐Lavigne, M. (1996). Netrin‐1 is required for commissural axon guidance in the developing vertebrate nervous system. Cell 87, 1001–1014. Shah, N. M., Groves, A. K., and Anderson, D. J. (1996). Alternative neural crest cell fates are instructively promoted by TGFbeta superfamily members. Cell 85, 331–343. Shen, Q., Goderie, S. K., Jin, L., Karanth, N., Sun, Y., Abramova, N., Vincent, P., Pumiglia, K., and Temple, S. (2004). Endothelial cells stimulate self‐renewal and expand neurogenesis of neural stem cells. Science 304, 1338–1340. Sherrington, R., Rogaev, E. I., Liang, Y., Rogaeva, E. A., Levesque, G., Ikeda, M., Chi, H., Lin, C., Li, G., Holman, K., Tsuda, T., Mar, L., et al. (1995). Cloning of a gene bearing missense mutations in early‐onset familial Alzheimer’s disease. Nature 375, 754–760. Shim, S., Goh, E. L., Ge, S., Sailor, K., Yuan, J. P., Roderick, H. L., Bootman, M. D., Worley, P. F., Song, H., and Ming, G. L. (2005). XTRPC1‐dependent chemotropic guidance of neuronal growth cones. Nat. Neurosci. 8, 730–735. Shin, D., Garcia‐Cardena, G., Hayashi, S., Gerety, S., Asahara, T., Stavrakis, G., Isner, J., Folkman, J., Gimbrone, M. A., Jr., and Anderson, D. J. (2001). Expression of ephrinB2 identifies a stable genetic diVerence between arterial and venous vascular smooth muscle as well as endothelial cells, and marks subsets of microvessels at sites of adult neovascularization. Dev. Biol. 230, 139–150. Silverman, W. F., Krum, J. M., Mani, N., and Rosenstein, J. M. (1999). Vascular, glial and neuronal eVects of vascular endothelial growth factor in mesencephalic explant cultures. Neuroscience 90, 1529–1541. Sondell, M., Lundborg, G., and Kanje, M. (1999). Vascular endothelial growth factor has neurotrophic activity and stimulates axonal outgrowth, enhancing cell survival and Schwann cell proliferation in the peripheral nervous system. J. Neurosci. 19, 5731–5740. Stein, E., Zou, Y., Poo, M., and Tessier‐Lavigne, M. (2001). Binding of DCC by netrin‐1 to mediate axon guidance independent of adenosine A2B receptor activation. Science 291, 1976–1982. Storkebaum, E., and Carmeliet, P. (2004). VEGF: A critical player in neurodegeneration. J. Clin. Invest. 113, 14–18. Storkebaum, E., Lambrechts, D., Dewerchin, M., Moreno‐Murciano, M. P., Appelmans, S., Oh, H., Van Damme, P., Rutten, B., Man, W. Y., De Mol, M., Wyns, S., Manka, D., et al. (2005). Treatment of motoneuron degeneration by intracerebroventricular delivery of VEGF in a rat model of ALS. Nat. Neurosci. 8, 85–92.

54

Zacchigna et al.

Suchting, S., Heal, P., Tahtis, K., Stewart, L. M., and Bicknell, R. (2005). Soluble Robo4 receptor inhibits in vivo angiogenesis and endothelial cell migration. FASEB J. 19, 121–123. Suchting, S., Bicknell, R., and Eichmann, A. (2006). Neuronal clues to vascular guidance. Exp. Cell Res. 312, 668–675. Sumoy, L., Keasey, J. B., Dittman, T. D., and Kimelman, D. (1997). A role for notochord in axial vascular development revealed by analysis of phenotype and the expression of VEGR‐2 in zebrafish flh and ntl mutant embryos. Mech. Dev. 63, 15–27. Sun, Y., Jin, K., Xie, L., Childs, J., Mao, X. O., Logvinova, A., and Greenberg, D. A. (2003). VEGF‐induced neuroprotection, neurogenesis, and angiogenesis after focal cerebral ischemia. J. Clin. Invest. 111, 1843–1851. Sun, Y., Jin, K., Childs, J. T., Xie, L., Mao, X. O., and Greenberg, D. A. (2004). Increased severity of cerebral ischemic injury in vascular endothelial growth factor‐B‐deficient mice. J. Cereb. Blood Flow Metab. 24, 1146–1152. Sun, Y., Jin, K., Childs, J. T., Xie, L., Mao, X. O., and Greenberg, D. A. (2006). Vascular endothelial growth factor‐B (VEGFB) stimulates neurogenesis: Evidence from knockout mice and growth factor administration. Dev. Biol. 289, 329–335. Sundaram, M., and Greenwald, I. (1993). Suppressors of a lin‐12 hypomorph define genes that interact with both lin‐12 and glp‐1 in Caenorhabditis elegans. Genetics 135, 765–783. Szebenyi, G., Dent, E. W., Callaway, J. L., Seys, C., Lueth, H., and Kalil, K. (2001). Fibroblast growth factor‐2 promotes axon branching of cortical neurons by influencing morphology and behavior of the primary growth cone. J. Neurosci. 21, 3932–3941. Tabara, H., Grishok, A., and Mello, C. C. (1998). RNAi in C. elegans: Soaking in the genome sequence. Science 282, 430–431. Takahashi, T., Nakamura, F., Jin, Z., Kalb, R. G., and Strittmatter, S. M. (1998). Semaphorins A and E act as antagonists of neuropilin‐1 and agonists of neuropilin‐2 receptors. Nat. Neurosci. 1, 487–493. Tarsitano, M., De Falco, S., Colonna, V., McGhee, J. D., and Persico, M. G. (2006). The C. elegans pvf‐1 gene encodes a PDGF/VEGF‐like factor able to bind mammalian VEGF receptors and to induce angiogenesis. FASEB J. 20, 227–233. Tavernarakis, N., Wang, S. L., Dorovkov, M., Ryazanov, A., and Driscoll, M. (2000). Heritable and inducible genetic interference by double‐stranded RNA encoded by transgenes. Nat. Genet. 24, 180–183. Temple, S. (2001). The development of neural stem cells. Nature 414, 112–117. Tepass, U., Godt, D., and Winklbauer, R. (2002). Cell sorting in animal development: Signalling and adhesive mechanisms in the formation of tissue boundaries. Curr. Opin. Genet. Dev. 12, 572–582. Territo, P. R., and Burggren, W. W. (1998). Cardio‐respiratory ontogeny during chronic carbon monoxide exposure in the clawed frog Xenopus laevis. J. Exp. Biol. 201, 1461–1472. Tessier‐Lavigne, M., and Goodman, C. S. (1996). The molecular biology of axon guidance. Science 274, 1123–1133. Torres, M., Gomez‐Pardo, E., and Gruss, P. (1996). Pax2 contributes to inner ear patterning and optic nerve trajectory. Development 122, 3381–3391. Torres‐Vazquez, J., Kamei, M., and Weinstein, B. M. (2003). Molecular distinction between arteries and veins. Cell Tissue Res. 314, 43–59. Trousse, F., Marti, E., Gruss, P., Torres, M., and Bovolenta, P. (2001). Control of retinal ganglion cell axon growth: A new role for Sonic hedgehog. Development 128, 3927–3936. Vogeli, K. M., Jin, S. W., Martin, G. R., and Stainier, D. Y. (2006). A common progenitor for haematopoietic and endothelial lineages in the zebrafish gastrula. Nature 443, 337–339. Wang, B., Xiao, Y., Ding, B. B., Zhang, N., Yuan, X., Gui, L., Qian, K. X., Duan, S., Chen, Z., Rao, Y., and Geng, J. G. (2003). Induction of tumor angiogenesis by Slit‐Robo signaling and inhibition of cancer growth by blocking Robo activity. Cancer Cell 4, 19–29.

1. Angiogenesis and Neurogenesis

55

Wang, H. U., Chen, Z. F., and Anderson, D. J. (1998). Molecular distinction and angiogenic interaction between embryonic arteries and veins revealed by ephrin‐B2 and its receptor Eph‐B4. Cell 93, 741–753. Wang, K. H., Brose, K., Arnott, D., Kidd, T., Goodman, C. S., Henzel, W., and Tessier‐ Lavigne, M. (1999). Biochemical purification of a mammalian slit protein as a positive regulator of sensory axon elongation and branching. Cell 96, 771–784. Webber, C. A., Hyakutake, M. T., and McFarlane, S. (2003). Fibroblast growth factors redirect retinal axons in vitro and in vivo. Dev. Biol. 263, 24–34. Weinstein, B. M. (2005). Vessels and nerves: Marching to the same tune. Cell 120, 299–302. Weinstein, B. M., Stemple, D. L., Driever, W., and Fishman, M. C. (1995). Gridlock, a localized heritable vascular patterning defect in the zebrafish. Nat. Med. 1, 1143–1147. Wen, Z., and Zheng, J. Q. (2006). Directional guidance of nerve growth cones. Curr. Opin. Neurobiol. 16, 52–58. Wessells, N. K., and Nuttall, R. P. (1978). Normal branching, induced branching, and steering of cultured parasympathetic motor neurons. Exp. Cell Res. 115, 111–122. Whitworth, A. J., Wes, P. D., and Pallanck, L. J. (2006). Drosophila models pioneer a new approach to drug discovery for Parkinson’s disease. Drug Discov. Today 11, 119–126. Wilson, B. D., Ii, M., Park, K. W., Suli, A., Sorensen, L. K., Larrieu‐Lahargue, F., Urness, L. D., Suh, W., Asai, J., Kock, G. A., Thorne, T., Silver, M., et al. (2006). Netrins promote developmental and therapeutic angiogenesis. Science 313, 640–644. Wilson, S. W., and Houart, C. (2004). Early steps in the development of the forebrain. Dev. Cell 6, 167–181. Wolf, B. D., and Chiba, A. (2000). Axon pathfinding proceeds normally despite disrupted growth cone decisions at CNS midline. Development 127, 2001–2009. Wurmser, A. E., Nakashima, K., Summers, R. G., Toni, N., D’Amour, K. A., Lie, D. C., and Gage, F. H. (2004). Cell fusion‐independent diVerentiation of neural stem cells to the endothelial lineage. Nature 430, 350–356. Xu, Q., Alldus, G., Holder, N., and Wilkinson, D. G. (1995). Expression of truncated Sek‐1 receptor tyrosine kinase disrupts the segmental restriction of gene expression in the Xenopus and zebrafish hindbrain. Development 121, 4005–4016. Yang, S., Tutton, S., Pierce, E., and Yoon, K. (2001). Specific double‐stranded RNA interference in undiVerentiated mouse embryonic stem cells. Mol. Cell Biol. 21, 7807–7816. Yoshida, T., and Mishina, M. (2003). Neuron‐specific gene manipulations to transparent zebrafish embryos. Methods Cell Sci. 25, 15–23. Yoshikawa, S., McKinnon, R. D., Kokel, M., and Thomas, J. B. (2003). Wnt‐mediated axon guidance via the Drosophila Derailed receptor. Nature 422, 583–588. You, L. R., Lin, F. J., Lee, C. T., DeMayo, F. J., Tsai, M. J., and Tsai, S. Y. (2005). Suppression of Notch signalling by the COUP‐TFII transcription factor regulates vein identity. Nature 435, 98–104. Zerlin, M., and Goldman, J. E. (1997). Interactions between glial progenitors and blood vessels during early postnatal corticogenesis: Blood vessel contact represents an early stage of astrocyte diVerentiation. J. Comp. Neurol. 387, 537–546. Zhong, H., Wu, X., Huang, H., Fan, Q., Zhu, Z., and Lin, S. (2006). Vertebrate MAX‐1 is required for vascular patterning in zebrafish. Proc. Natl. Acad. Sci. USA 103, 16800–16805. Zhong, T. P., Childs, S., Leu, J. P., and Fishman, M. C. (2001). Gridlock signalling pathway fashions the first embryonic artery. Nature 414, 216–220. Zimmer, M., Palmer, A., Kohler, J., and Klein, R. (2003). EphB‐ephrinB bi‐directional endocytosis terminates adhesion allowing contact mediated repulsion. Nat. Cell Biol. 5, 869–878.

2

____________________________________________________________________________

Junction Restructuring and Spermatogenesis: The Biology, Regulation, and Implication in Male Contraceptive Development Helen H. N. Yan, Dolores D. Mruk, and C. Yan Cheng Center for Biomedical Research, Population Council, New York, New York 10021

I. Introduction II. Anchoring Junctions in the Testes: An Update A. Actin‐Based Adherens Junctions B. Testis‐Specific AJs: ES and TBC C. Intermediate Filament‐Based Anchoring Junctions III. Roles of ECM Proteins in Junction Dynamics in the Testes A. Collagens B. Laminins C. Laminin Receptors IV. Role of Androgens in Junction Dynamics in Testes A. Introduction B. Models to Study the Role of Androgen in Spermatogenesis V. Regulation of Junction Turnover by Protein Endocytosis and Recycling A. An Overview B. Recent Advances on Studies Investigating the Role of Endocytosis in Junction Dynamics in the Testis VI. Regulation of Junction Dynamics by Myoid Cells VII. Environmental Toxicants: Are They Targeting the Tight and/or Anchoring Junction? VIII. Concluding Remarks Acknowledgments References

Spermatogenesis that occurs in the seminiferous epithelium of adult mammalian testes is associated with extensive junction restructuring at the Sertoli–Sertoli cell, Sertoli–germ cell, and Sertoli–basement membrane interface. While this morphological phenomenon is known and has been described in great details for decades, the biochemical and molecular changes as well as the mechanisms/signaling pathways that define changes at the cell– cell and cell–matrix interface remain largely unknown until recently. In this chapter, we summarize and discuss findings in the field regarding the coordinated eVorts of the anchoring [e.g., adherens junction (AJ), such as basal ectoplasmic specialization (basal ES)] and tight junctions (TJs) that are present in the same microenvironment, such as at the blood–testis barrier (BTB), or at distinctly opposite ends of the Sertoli cell epithelium, such as Current Topics in Developmental Biology, Vol. 80 Copyright 2008, Elsevier Inc. All rights reserved.

57

0070-2153/08 $35.00 DOI: 10.1016/S0070-2153(07)80002-0

58

Yan et al.

between apical ectoplasmic specialization (apical ES) in the apical compartment, and the BTB adjacent to the basal compartment of the epithelium. These eVorts, in turn, regulate and coordinate diVerent cellular events that occur during the seminiferous epithelial cycle. For instance, the events of spermiation and of preleptotene spermatocyte migration across the BTB both take place concurrently at stage VIII of the epithelial cycle of spermatogenesis. Recent findings suggest that these events are coordinated by protein complexes found at the apical and basal ES and TJ, which are located at diVerent ends of the Sertoli cell epithelium. Besides, we highlight important areas of research that can now be undertaken, and functional studies that can be designed to tackle diVerent issues pertinent to junction restructuring during spermatogenesis. ß 2008, Elsevier Inc.

I. Introduction During spermatogenesis, spermatozoa (haploid, 1n) are formed from spermatogonial stem cells (diploid, 2n) in the testes. This event takes place in the seminiferous tubules of mammalian testes such as in rats and humans. This process completes in
58 days in rats when a single spermatogonium undergoes six sequential mitotic and two meiotic divisions to give rise to fully developed spermatids (spermatozoa) (de Kretser and Kerr, 1988; Leblond and Clermont, 1952). The morphological changes of germ cells in the seminiferous epithelium during spermatogenesis can be divided into 14 stages (stages I–XIV) in rats, which, in turn, constitute one seminiferous epithelial cycle. Each stage is typified by the association of germ cells at defined stages of their development with Sertoli cells (Cheng and Mruk, 2002; Hess et al., 1990; Leblond and Clermont, 1952; Mruk and Cheng, 2004a; Parvinen, 1982). Other than the morphological changes associated with germ cell development during spermatogenesis, some spermatogonial stem cells residing near the basement membrane, via a yet‐to‐be defined mechanism, transform into type B spermatogonia, which enter into the cell cycle by diVerentiating into preleptotene spermatocytes, traversing the blood–testis barrier (BTB) and migrate progressively across the seminiferous epithelium. However, germ cells remain attached to Sertoli cells for nourishment and structural supports at all time during the epithelial cycle while diVerentiating into spermatozoa (Russell, 1977b). During this active cell migration process, intermittent junction disassembly and reassembly occur at the Sertoli–Sertoli cell and Sertoli– germ cell interface (Mruk and Cheng, 2004b). If cross talk between these cells is disrupted, spermatogenic cells cannot migrate and/or orientate properly in the seminiferous epithelium. This thus leads to germ cell apoptosis, premature germ cell depletion from the epithelium, and infertility.

59

2. Junction Restructuring and Spermatogenesis

During the seminiferous epithelial cycle, two crucial cellular events take place simultaneously at stages VII–VIII which last for
3.5 days in rats: (1) the migration of preleptotene spermatocytes across the BTB, and (2) the depletion of mature spermatozoa from the luminal edge of the epithelium into the tubule lumen at spermiaton. However, the biochemical events and the molecular mechanisms by which these events are coordinated and regulated are virtually unknown until recently. Since the molecular architecture of diVerent cell junction types in the testis has been reviewed (Bart et al., 2002; Cheng and Mruk, 2002; Mruk and Cheng, 2004b; Toyama et al., 2003; Vogl et al., 2000, 2007) and summarized in Table I, this subject area is not covered in this chapter. Furthermore, the regulation of BTB dynamics during spermatogenesis, in particular the roles and the involvement of cytokines, proteases, and protease inhibitors has also been reviewed (Lui et al., 2003a; Wong and Cheng, 2005), readers are encouraged to seek information from these earlier reviews. Instead, we focus herein more recent findings in the field in particular the roles of extracellular matrix components, steroids, and GTPases that contributed to the two coordinated cellular events at stages VII–VIII of the epithelial cycle during spermatogenesis. We also provide a recent update on the anchoring junctions in the testis and present a hypothesis regarding the unusual vulnerability of the testes toward diVerent environmental toxiants. In particular, we highlight the significance of cross talk between diVerent junction types restricted to diVerent cellular compartments in the seminiferous epithelium. It is likely that these

Table I

Types of Junctions Found in the Seminiferous Epithelium of Adult Testes

Junction Type Occluding/TJ Anchroing junction 1. Actin filaments‐based cell–cell anchoring junctions i. Classic adheren junctions ii. Ectoplasmic specialization (ES) a. Basal ES b. Apical ES iii. TBC a. Basal TBC b. Apical TBC 2. Intermediate filament‐based cell–cell desmosome‐like junctions Communicating/GJs

Location Sertoli–Sertoli cells at the BTB

Sertoli–Sertoli cells and Sertoli–germ cells Sertoli–Sertoli cells at the BTB Sertoli cell‐elongating /elongated spermatid (from step 8 spermatid and beyond) Sertoli–Sertoli cells at the BTB Sertoli–elongated spermatids Sertoli–germ cells (spermatogonia, spermatocytes, and prior to step 8 spermatids) Between Leydig cells, Sertoli–Sertoli cells at the BTB, and Sertoli–germ cells

60

Yan et al.

cross talks play a crucial role in coordinating diVerent biochemical events during the epithelial cycle.

II. Anchoring Junctions in the Testes: An Update A. Actin‐Based Adherens Junctions Based on earlier studies that define the localization of actin‐based cytoskeletal filaments in the seminiferous epithelium of adult rodent testes (Vogl, 1989; Vogl et al., 2007), actin‐based adherens junctions (AJs) are mostly concentrated in two specific regions: (i) the BTB at the Sertoli–Sertoli cell interface, and (ii) the apical ectoplasmic specialization (ES) and the apical tubulobulbar complex (TBC) at the Sertoli–spermatid interface. Most of the reports in the field regarding AJ in the testis are focused on ES such as basal ES at the BTB, and apical ES (for reviews, see Mruk and Cheng, 2004a; Vogl et al., 2000). Thus, other actin‐ based AJ at the Sertoli–Sertoli and Sertoli–spermatogonia, –spermatocyte, or –round spermatid interface are less characterized and studied. Skeletal protein 4.1 G identified in the mouse testis (Terada et al., 2005) perhaps is one of the few AJ structural components that has been studied thus far in testes. Under electron microscope, it was found to localize to the plasma membrane of Sertoli cells where germ cells attached. It also colocalized with E‐cadherin but not at the site of basal or apical ES when visualized by immunofluorescent microscopy (Terada et al., 2005). B. Testis‐Specific AJs: ES and TBC ES is the best characterized AJ type in the testis. It is confined to the interface between Sertoli cells at the BTB known as the basal ES, as well as between Sertoli cells and elongating/elongated spermatids designated the apical ES (Mruk and Cheng, 2004a; Russell, 1977c; Toyama et al., 2003; Vogl et al., 2000). ES is diVerent from the classic AJ in many ways. For instance, ES is typified ultrastructurally by the presence of hexagonally‐packed actin bundles sandwiched between Sertoli cell plasma membrane and the cisternae of endoplasmic reticulum; these ultrastructures are readily visible by electron microscopy (Mruk and Cheng, 2004a; Vogl et al., 2007). Cadherins, nectins and integrins are the three classes of transmembrane proteins present at the site of ES in the rat testis (Mruk and Cheng, 2004b). These proteins are linked to actin filaments via adaptors. For example, catenins and afadins are adaptors of cadherins and nectins, respectively; whereas, vinculin links integrin to actin. Recent studies seem to question the concept that cadherins (e.g., N‐ and E‐cadherin) structurally interact with ‐ or ‐catenin at a stoichiometric ratio of 1:1, which, in turn, link to the actin cytoskeleton network through ‐catenin.

2. Junction Restructuring and Spermatogenesis

61

It was suggested that a stable E‐cadherin/ ‐catenin/‐catenin/actin quaternary protein complex did not exist in epithelial cells, instead ‐catenin served as a ‘‘switch’’ and dynamically shuZed between E‐cadherin/ ‐caterin complex and actin bundles (Drees et al., 2005; Yamada et al., 2005). For instance, ‐catenin was shown to interact with either actin filaments or the E‐cadherin/ ‐catenin complex; it failed to have simultaneous interaction with them, illustrating that the interaction between ‐catenin and actin or ‐catenin and E‐cadherin/ ‐catenin is mutually exclusive, even in the presence of adaptors such as vinculin and ‐actinin (Yamada et al., 2005). Besides, ‐catenin can either exist as a monomer or a dimer which binds to the E‐cadherin/ ‐catenin complex or actin filaments, respectively (Drees et al., 2005). While this issue remains controversial, another testis‐specific truncated T‐catenin isoform‐X was identified in mouse testes apart from the ubiquitously expressed E‐catenin (Goossens et al., 2006). This truncated T‐catenin isoform‐X, however, is lacking the ‐catenin‐ binding domain such that it does not mediate cell–cell adhesion via the traditional cadherin/catenin/actin complex, instead, it was found to bind L‐afadin as demonstrated using the yeast two‐hybrid system. By immunohistochemistry, truncated T‐catenin isoform‐X was localized to the apical ES illustrating stage specificity distribution at stages II–XIII (Goossens et al., 2006). In short, much work is needed to define the role of alpha‐catenin in linking the cadherin/catenin complex to the actin‐based cytoskeleton. Interestingly, the structural components of the apical ES seem to go beyond our understanding of the classical AJ. Recent studies have identified numerous proteins at the apical ES, which are usually restricted to cell–matrix interface to facilitate cell movement such as in cancer cells during tumorigenesis. These include 6 1‐integrin, laminin‐333, p‐FAK, vinculin, and paxillin (Siu and Cheng, 2004a; Yan et al., 2007). Besides, TJ integral membrane proteins such as JAM‐B/C and CAR which were thought to be the BTB site are shown to be a component of the apical ES (Gliki et al., 2004; Mirza et al., 2006); these proteins together with the polarization protein complexes such as Cdc42‐PAR3/6‐aPKC are suggested to assist the proper orientation of spermatids before spermiation (Gliki et al., 2004; Mirza et al., 2006). By using fluorescent microscopy, JAM‐C was shown to colocalize with JAM‐B at apical ES (Gliki et al., 2004), likely forming heterotypic protein–protein interactions to confer ‘‘TJ’’ and/or cell adhesion. A homophilic interaction of JAM‐C was found between tumor cells as well as between endothelial cells (Santoso et al., 2005), yet, it was suggested that JAM‐C prefers forming heterodimers with JAM‐B instead of homodimer (Lamagna et al., 2005). In the testis, it is presently not known whether there is homo‐ or heterodimer of JAM‐C and/or JAM‐B at the apical ES. Besides, the adaptor(s) that structurally links JAM‐B/C and CAR to the actin cytoskeleton at the apical ES is presently not known. Nonetheless, these latest findings regarding the structural protein complexes at the ES seem to favor the argument that ES is

62

Yan et al.

utilizing the best features of AJ, TJ, and focal contacts to maintain the dynamic interactions between Sertoli cells and developing spermatids (step 8 and beyond) to facilitate spermatid migration across the seminiferous epithelium during spermatogenesis (Yan et al., 2007). These results also illustrate that apical ES is a hybrid junction of AJ, TJ, and focal contacts. TBC is another testis‐specific AJ type. Basal TBC is restricted to the Sertoli– Sertoli cell interface at the BTB, coexisting with TJ, basal ES, desomsome‐like junction and gap junction (GJ) to constitute the BTB. Apical TBC, however, is confined to the concave side of elongated spermatids and appears only at late stage VIII of the seminiferous epithelial cycle prior to spermiation by replacing the apical ES (Russell and Clermont, 1976). As such, the presence of apical ES and apical TBC at the Sertoli–elongated spermatid interface is mutually exclusive. The transient appearance of TBC and the concomitant dissolution of apical ES were postulated to facilitate spermiation (Russell, 1979). In other epithelia, the cell–matrix actin‐based anchoring junction is known as focal contact or focal adhesion complex. However, it remains to be reported if focal contact is present in the testes. Ohtsuka et al. have identified two actin‐ binding proteins from rat brain and rat 3Y1 fibroblasts, known as b‐ and s‐nexilin. Both variants were shown to be expressed in the testis, and s‐nexilin was colocalized with vinculin, talin, and paxillin at the cell–matrix anchoring junctions, but not at cell–cell AJ, in 3Y1 cell monolayers as visualized by immunofluorescent microscopy (Ohtsuka et al., 1998). However, the precise localization of nexilin in testes is not known even though vinculin and paxillin have been shown to localize at the apical ES (PfeiVer and Vogl, 1991; Wine and Chapin, 1999), but talin was found in the basement membrane of human seminiferous tubules and at the cell‐matrix interface (Santoro et al., 2000). These biochemical studies seem to support the likely presence of focal contacts in the testes. C. Intermediate Filament‐Based Anchoring Junctions Desmosomes and hemidesmosomes are intermediate filament‐based anchoring junctions at the cell–cell and cell–matrix interface, respectively. The transmembrane proteins that constitute desmosomes in many epithelia known to date are desmosomal cadherin families: desmogleins (Dsg) and desmocollins (Dsc). Plakophilin and plakoglobin are the cytoplasmic proteins which connect desmosome integral membrane proteins to intermediate filaments via desmoplakin (Green et al., 2005; Yin and Green, 2004). In the testis, traditional desmosomes and GJs apparently are intermingled together forming the desmosome‐like junctions (Russell, 1977a), but putative GJs are also found (Pointis and Segretain, 2005). At present, it is not known if any of the Dsg and Dsc are indeed components of the desmosome‐like junctions except that several connexins have been found in the testis

2. Junction Restructuring and Spermatogenesis

63

(Pointis and Segretain, 2005). For hemidesmosomes, integrins and bullous permphigoid (BP) antigen BP180 (also known as type XVII collagen) are the major cell surface proteins. CD151, a tetraspanin membrane protein with four transmembrane domains, is known to be a binding partner of 6‐integrin (Litjens et al., 2006). These proteins coupled with two plakin family members: plectin and BP230, are presently known to constitute the hemidesmosomes (Litjens et al., 2006; Nievers et al., 1999). Hemidesmosomes remain rarely explored in the testes except some limited morphological data illustrating its presence in bovine and rat seminiferous tubules (Siu and Cheng, 2004a; Wrobel et al., 1979).

III. Roles of ECM Proteins in Junction Dynamics in the Testes Type IV collagen, laminin, heparin sulfate proteoglycan, and nidogen (also known as entactin) are the major components of the basement membrane in the seminiferous epithelium of mammalian testes, which is a modified form of ECM (for a review, see Dym, 1994). They are secretory products from diVerent cell types in the testis, mostly Sertoli and peritubular myoid cells (Dym, 1994). In the following section, we focus on the roles of collagens and laminins in the testes in particular their involvement in junction dynamics during spermatogenesis.

A. Collagens There are at least 19 diVerent collagens found in mammals and 4 types of collagens of type I, II, III and IV, with type IV collagen being the major building block of ECM network in the seminiferous tubule. For types I, II and III collagens they are proteins that confer tensile strength to epithelia in tissues. For instance, type I collagen fibrils form a non‐cellular zone underneath the basement membrane in adult rat testes (for a review, see Siu and Cheng, 2004a). Type IV collagen is a trimer composed of three  chains. Thus far, six distinct  chains designated 1–6 are known (Harvey et al., 2006; Siu and Cheng, 2004a). The six  chains can be self‐assembled into three basic functional units: 122, 526, and 345; and three distinct networks: 1/2, 3/4/5, and 1/2/5/6 are formed based on the available functional units. Hervey et al. (2006) has suggested that there is a sequential change in the content of collagen network at the basement membrane of seminiferous tubules in canine testes from birth to adulthood. It was indicated that diVerent combinations of collagens at the basement membrane were essential in nourishing cells in the seminiferous tubules during spermatogenesis. Type IV collagen was detected when isolated Sertoli cells were

64

Yan et al.

cultured in vitro, illustrating the de novo synthesis by Sertoli cells (Davis et al., 1990). Additionally, type I collagen 1, 2, and type IV collagen 3 are products of germ cells (He et al., 2005; Siu et al., 2003a). Type IV collagen 3 chain is present in germ cell‐conditioned medium (GCCM) using purified germ cells cultured in vitro (Siu et al., 2003a). Type IV collagen 3 and 4 were both colocalized to the basement membrane of the seminiferous epithelium and peritubular myoid cells in rat testes (Frojdman et al., 1998). Collectively, these data suggest that collagens are secreted by various cell types in the seminiferous tubules and each probably has its unique function(s). Apart from its role to provide scaVolding support to the overlying Sertoli cells in the seminiferous epithelium, collagens apparently are involving in junction restructuring events in the testes during spermatogenesis. This postulate was supported in part by the observation that the steady‐state mRNA level of collagens in rat testes peaked at 10–20 days after birth, coinciding with the time of BTB assembly at approximately days 16–18 (Enders et al., 1995; Siu et al., 2003a), illustrating collagens may be involved in BTB formation in vivo. Furthermore, in vitro studies have shown that the presence of an anti‐collagen 3 (IV) antibody in Sertoli cell cultures could compromise the TJ‐barrier permeability (Siu et al., 2003a). Besides, type I collagen 1 and 2 chains were suggested to induce germ cell detachment and migration on the basement membrane (He et al., 2005). By immunofluorescent microscopy, procollagen I, a precursor of type I collagen, was only found to associate with spermatogonia but not Sertoli cells in 6‐day‐old mice. In adult mice, specific staining of procollagen I was detected in undiVerentiated type A spermatogonia as well as preleptotene spermatocytes. These germ cell types have close contacts with the basement membrane. Interestingly, leptotene spermatocytes, round spermatids, and elongating/elongated spermatids which are no longer in contact with the basement membrane, the expression of procollagen I was virtually undetectable (He et al., 2005). Collectively, these recent findings illustrate that a disruption of extracellular matrix function by using an anti‐collagen antibody can lead to Sertoli cell TJ‐barrier malfunction, illustrating a cross talk between ECM and TJs (Fig. 1).

B. Laminins Laminin is a heterotrimeric glycoprotein composed of , , and chains. There are several steps involved in the formation of a functional laminin complex, which include chain selection, assembly, and stabilization. In brief, a stretch of peptide sequence near the C‐terminal of laminin chain initially forms a disulfide‐bonded dimer with chain. Laminin  chain will subsequently be incorporated to the dimer via the stretch of sequence near domain II, yielding a functional trimer (Aumailley and Smyth, 1998). Six domains, I through VI, are

65

2. Junction Restructuring and Spermatogenesis

Tubule lumen

Laminin fragments

Adluminal compartment

Apical ES

Sertoli cell

Round spermatid

Sertoli cell

?

Pachytene spermatocyte

Tight junction (cell−cell actin based) Basal TBC Basal ES

Basal TBC

BTB Basal compartment

Apical TBC

Apical TBC

?

Basement membrane Type I collagen Myoid cells Lymph Lymphatic endothelium

Elongated spermatid

Adherens junction Desmosome-like junction Gap junction

Preleptotene spermatocyte

Actin bundles Hemidesmosome Apical ES

Spermatogonium

_ Collagen IV Laminin

Tunica propria _

Figure 1 A schematic drawing illustrating ‘‘cross talks’’ between diVerent cell junction types at diVerent locations in the seminiferous epithelium. This drawing shows the relative location of diVerent junction types between Sertoli–Sertoli and Sertoli–germ cells where germ cells are at diVerent stages of their development using adult rat testes as the model. The BTB is composed of coexisting TJs, basal TBC, basal ES, desmosome‐like junctions, and GJs. It physically divides the seminiferous epithelium into the basal and adluminal compartments. The base of the Sertoli cell is in direct contact with the basement membrane (a modified form of extracellular matrix, ECM) which is tightly associated with the type I collagen network, and, in turn, they constitute the acellular component of the tunica propria. In rodents, peritubular myoid cells and the lymphatic vessel constitute the cellular zone of the tunica propria. Each of the colored arrows represents diVerent possible ‘‘cross talk’’ which likely facilitates the event of germ cell migration across the seminiferous epithelium, perhaps also important for morphological and molecular changes of germ cells during their development. These include: (1) cross talk between apical ES and BTB (dark‐blue arrow): it is postulated that during spermiation, proteolysis of laminins generates biologically active fragments by proteases such as MMPs that occurs at the apical ES, which, in turn, mediates junction disassembly at the both apical ES and BTB. This leads to spermiation and the opening of the BTB to facilitate preleptotene spermatocyte migration that occurs at stage VIII of the seminiferous epithelial cycle; (2) cross talk between peritubular myoid cells and junctions inside the seminiferous epithelium (orange‐yellow arrow): specific cellular knockout of

66

Yan et al.

found in each of the laminin subunits except for certain subunits, such as laminin 3 and 2, in which the domain VI is lacking. Domains I and II are important to guide the intramolecular assembly of the functional laminin complex while domain VI facilitates chain polymerization and is crucial to establish a network. Laminin  chains have a specific domain G which is used for the interaction with cell surface receptor integrins (Engvall and Wewer, 1996). There are at least 11 laminin subunits identified in mammals including five ‐, three ‐, and three ‐chains. Some of these laminin subunits have been found in either Sertoli or germ cells, or both (for a review, see Yan et al., 2007). Among the available literature that documented the function of laminins in the testes, all these studies were performed by injecting an anti‐laminin antiserum intratesticularly to assess any changes in the phenotypes (Denduchis et al., 1985; Lustig et al., 1987, 2000). The most obvious eVect was the sloughing of germ cells from the epithelium following blocking antibody administration locally, illustrating a disruption of junctions at the Sertoli–germ cell interface had occurred by perturbing laminins in the ECM (Fig. 1). Other in vitro studies have shown that the addition of a pentapeptide of YIGSR corresponding to the domain III of laminin 1 (Graf et al., 1987) can lead to Sertoli cell detachment from the cell matrix within 2–3 hours (Tung and Fritz, 1993). Besides, the synthetic peptides NH2–SDGR–COOH and NH2–RSGIY–COOH corresponding to the sequences found in the cell‐binding domain of laminin ‐ and 1‐chain, respectively, have been shown to inhibit Sertoli cell cord structures formation in vitro (Hadley et al., 1990). However, these studies were focused on laminin localized at the basement membrane of the seminiferous epithelium. Recent studies have demonstrated that laminin 3, 3, and 3 subunits form a functional laminin protein complex restricted to developing spermatids, most notably at approximately step 8 spermatids and beyond, at stages VI to early VIII tubules, and this laminin‐333 was shown to associate with 6 1‐integrin residing on the Sertoli cell, forming a bona fide adhesion complex at the apical ES (Yan and Cheng, 2006). When either anti‐laminin 3 or 3 IgG was injected to the rat testes locally, a blockage of this laminin can mediate a transient TJ disassembly (Yan and Cheng, 2006) as well as germ cell depletion from the epithelium by day 4 (see Fig. 2). Besides these obvious phenotypic changes, multinucleated germ cells were also detected in the androgen receptor in peritubular myoid cells led to a lowering in the steady‐state mRNA levels of TJ, GJ, and apical ES; (3) cross talk between extracellular matrix proteins and diVerent cell junctions in the seminiferous epithelium (purple arrows): anti‐collagen IV and anti‐laminin antibodies were shown to perturb TJ‐barrier permeability in vitro, and inducing germ cell sloughing from the epithelium in vivo, respectively. It remains to be determined if there is any cross talk between basal TBC and apical TBC as well as between desmosome‐like junction and AJ (see red and blue arrows with a question mark). This figure was prepared based on reports and reviews in the field (Siu and Cheng, 2004a, b; Siu et al., 2003a; Yan and Cheng, 2006; Yan et al., 2007; Zhang et al., 2006) (see also discussion in text).

A Ctrl

B6h

C1d

D2d

E4d

F7d

G 15 d

Gi

i

H ii

H 25 d

ii

iii

J iv

J 65 d

I iii

I 45 d

Kv

K 100 d

iv

v

68

Yan et al.

seminiferous epithelium following anti‐laminin IgG administration, which is likely the result of necrosis (Fig. 2). Based on these limited available data, it is tempting to speculate that a blockage of laminin function at the apical ES can lead to a disruption of cell adhesion, inducing premature release of elongate spermatids into the tubule lumen, and mimicking the phenotype of a stage VIII tubule. This event may also lead to a disruption of the neighboring AJ protein complexes, such as the cadherins/ ‐catenin. Besides, this event also induces other pertinent signaling events that mediate changes at the BTB at the Sertoli–Sertoli interface, lowering the steady‐state protein level of occludin. This thus facilitates the migration of preleptotene spermatocytes across the BTB. It is likely that the cleavage of laminins by proteases, such as MMPs, at the apical ES generate active biological peptides that exert their eVects on the Sertoli cell TJ‐barrier at the BTB. In other epithelia, laminins are cleaved by MMPs into smaller fragments which can enhance cell migration or cause blood–brain barrier disruption (Giannelli et al., 1997; Gurney et al., 2006; Udayakumar et al., 2003). Furthermore, MMP2 has been colocalized with laminin 3 to the apical ES in vivo, and its expression was shown to be upregulated when germ cells were cocultured with Sertoli cells in vitro (Longin et al., 2001; Siu and Cheng, 2004b), illustrating the machineries need to generate the biologically active peptides are present at the apical ES. Recent studies from our laboratory have shown that when fragments of laminin 3 or 3 recombinant protein (containing part of the domain I of the laminin subunits) were added to Sertoli–germ cell coculture in vitro, this led to a significant decline in the expression of 1‐integrin, ‐catenin, and occludin (Yan and Cheng, unpublished observations). When the Sertoli cell TJ‐barrier permeability was monitored by quantifying the transepithelial electrical resistant (TER) across the cell epithelium, there was a
40% decline in TER after adding laminin fragments to Sertoli cells versus cells received a control fragment (Yan and Cheng, unpublished observations). Figure 2 (A–K) EVects on cell adhesion in the seminiferous epithelium following a blockage of the laminin‐333 function by using specific anti‐laminin 3 antibodies. Micrographs of paraYn sections of testes stained with hematoxylin–eosin from rats received normal rabbit IgG or preimmune rabbit IgG (A) at a dose of 75 mg IgG/testis via intratesticular administration (Yan and Cheng, 2006), and terminated by day 15 thereafter. Results were compared to rats received 75 mg/testis of anti‐laminin 3 IgG and terminated on 6 h (B), day 1 (C), day 2 (D), day 4 (E) (received one injection), day 7 (F), day 15 (G, G: i) (received two IgG injections, one each on day 0 and day 7), day 25 (H, H: ii), day 45 (I, I: iii), day 65 (J, J: iv) and day 100 (K, K: v) (received three IgG injections, one each on at day 0, day 7, and day 15). Germ cell depletion was detected by day 7 after an initial administration of anti‐laminin 3 IgG, and recovery was not apparent by up to day 100. Multinucleated germ cells were also found (F and G) which is likely the result of necrosis; however, blood vessels remained relatively intact (see H and I). Scale bar: (A) 120 mm, which applies to (B)–(H), (K: v); (G: i), 40 mm, which applies to (H: ii), (I: iii), (J: iv); (I), 180 mm, which applies to (J) and (K). Details of this experiment can be found in a recent report (Yan and Cheng, 2006), and this experiment was performed after the publication of the earlier report to assess recovery following treatment with the blocking antibody.

2. Junction Restructuring and Spermatogenesis

69

Taking collectively, these data provide concrete evidence that illustrate the presence of cross talk between apical ES and BTB (Fig. 1). Perhaps, this cross talk is essential to govern and coordinate the two crucial events that take place simultaneously at stage VIII of the epithelial cycle, namely: (i) the preleptotene spermatocyte migration across the BTB and (ii) sloughing of fully developed spermatids (i.e., spermatozoa) into the tubule lumen at spermiation.

C. Laminin Receptors Receptors for functional laminin protein complexes can be divided into integrin and non‐integrin‐based proteins (for reviews, see Ekblom, 1996; Mecham, 1991; Mercurio, 1995). Previous studies from our laboratory have identified laminin‐333 as the ligand of receptor 6 1‐integrin at the apical ES (Yan and Cheng, 2006). Since laminin‐333 is a germ cell product limited to elongating/elongated spermatids without a transmembrane domain, it seems logical to speculate that a yet‐to‐be identified transmembrane protein serving as a scaVolding protein for laminin‐333 is present in spermatids. Table II and this section summarize diVerent types of laminin receptors and their localization in the testes.

1. Integrin The best studied integrin receptor found in the testis is 6 1‐integrin; it is localized almost exclusively on the Sertoli cell side at the apical ES with laminin‐333 as one of its putative ligands (Salanova et al., 1995; Yan and Cheng, 2006). At the basement membrane, 6 4‐integrin is speculated to be another receptor of laminin at the basal ES; however, its presence remains to be confirmed (Mulholland et al., 2001). In other epithelia, such as skin, 6 4‐ integrin is the receptor of laminin‐332 which confers cell adhesion and facilitates cell migration. Besides these integrin subunits, integrins 3 and 6 have been identified in the plasma membrane of spermatogenic cells (Kierszenbaum et al., 2006), thus, it is highly likely that other integrins may be present in elongating/elongated spermatids at the apical ES for anchoring laminin‐333 besides 6 1‐integrin.

2. 67‐kDa Laminin Receptor 67‐kDa laminin receptor (67‐LR) is an integral membrane protein known to have a high aYnity for laminins (Mecham, 1991). Using aYnity chromatography, the 67‐LR was shown to coelute with the laminin complex (Mecham, 1991). The YIGSR sequence located on domain III of laminin 1 subunit is

Table II

Laminins (Ligands) and Their Receptors

Receptor Integrin 1 1a

2 1a

3 1

6 1 6 4 A7X1 1a A7X2 1a Non‐integrin Entactin/Nidogen1 Dystroglycan Syndecan‐2,4 67‐kDa laminin receptor Lutheran/B‐CAM

Ligand

Localization of Laminin Receptors in Testes

References

Laminin‐111 (domain VI of laminin 1 chain), ‐211 (domain VI of laminin 2 chain)

Integrin 1 was localized to the BM of the seminiferous epithelium, while at stages V to VII of epithelial cycle, it was localized at apical ES n.d.

Colognato et al., 1997; Colognato‐Pyke et al., 1995; Giebel et al., 1997

BM

Kim et al., 2005; Nakamura et al., 2001

Sertoli cell side of apical ES

Ekblom, 1996; Geberhiwot et al., 1999; Salanova et al., 1995; Yan and Cheng, 2006 Kikkawa et al., 2000; Nievers et al., 1999; Rousselle et al., 1991 Nishiuchi et al., 2006 Nishiuchi et al., 2006

Laminin‐111 (LG4 and domain VI of laminin 1 chain), ‐211 (LG4 and domain VI of laminin 2 chain) Laminin‐332 (PPFLMLLKGSTR of LG3 of laminin 3 chain), ‐511/521 (LG3 of laminin 5 chain) Laminin‐111, ‐211/221, ‐332, ‐333, ‐411, ‐511/521 (LG3 of laminin 5 chain) Laminin‐332 (first repeat of domain G of laminin 3 chain), ‐511/521 Laminin‐111, ‐211/221, ‐511/521 Laminin‐111, ‐211/221

Speculate to be localized at basal ES n.d. n.d.

Colognato et al., 1997; Hozumi et al., 2006

Lamin‐111 (epidermal growth factor‐like III4 of laminin 1 chain) LG4 of laminin 1 chain Laminin‐332 (NSFMALYLSKGR of LG4 of laminin 3 chain) YIGSR of laminin 1 chain

BM In Sertoli cells, precise localization is not known

GersdorV et al., 2005; Lian et al., 1992; Mann et al., 1989; Mayer et al., 1993 Andac et al., 1999; Durbeej et al., 1998, 2001 Brucato et al., 2000; Utani et al., 2001

Pachytene spermatocytes, round spermatids

Fulcher et al., 1993; Graf et al., 1987

Laminin‐511/‐521 (LG3 of laminin 5 chain)

n.d.

Kikkawa and Miner, 2005; Kikkawa et al., 2002

BM

a Integrin 1 subunit has been localized to apical ES and possibly basal ES of rat testes (Salanova et al., 1995; Siu and Cheng, 2004b). BM ¼ basement membrane; n.d. ¼ not determined; CAM, cell adhesion molecule.

71

2. Junction Restructuring and Spermatogenesis ate

A a

tis

B

lys

s Te

b XIV

100 kDa 72 kDa -

VIII

- 67 kDa LR

55 kDa 40 kDa -

c

d

V

Mr

33 kDa V

24 kDa 17 kDa -

VIII

Laminin γ 3 e

f

Figure 3 Immunofluorescent localization of the 67‐kDa laminin receptor in the seminiferous epithelium of adult rat testes. (A: a–c) ParaYn sections of adult rat testes were incubated with a mouse anti‐67‐kDa laminin receptor antibody (Abcam, Cat No.: ab 3099–500, lot: 158206), followed by a goat anti‐mouse IgG secondary antibody conjugated with FITC. Specific staining was detected at the spermatogonia, pachytene spermatocytes, round spermatids. However, at stage VIII, no staining was found near the head of elongated spermatids at the site of apical ES (b), this is in contrast with the localization of laminin 3 in the seminiferous epithelium of adult rat testes (see d) in which maximum intensity of laminin 3 was detected at the apical ES at stage VIII of the epithelial cycle. These results thus suggest that the 67‐kDa laminin receptor is not likely to be the anchoring binding partner of the laminin‐333 (see text for discussion). (A: e–f) These two micrographs illustrate the localization of 67‐kDa laminin receptor in the human breast tumor tissue (Abcam, Cat No.: ab4697, lot: A606083) that serves as positive control. Scale bar: (a) 80 mm, which applies to (b) and (d); (c) 40 mm, which applies to (f); (e) 60 mm. (B) An immunoblot in which the anti‐67‐kDa laminin receptor antibody was used to stain lysates of rat testes illustrating the specificity of the antibody since one prominent band with an electrophoretic mobility of 67 kDa was detected, illustrating the staining seen in (A: a–f) is specific for this 67‐kDa laminin receptor.

the putative binding site for 67‐LR (Graf et al., 1987). In mouse testes, 67‐LR was detected in spermatogenic cells with predominantly expression restricted to round spermatids and absent in elongated spermatids (Fulcher et al., 1993). By immunofluorescent microscopy as shown in Fig. 3, 67‐LR was

72

Yan et al.

localized to the spermatogonia at the basement membrane, as well as the heads of round spermatids, apparently at the site of early apical ES. However, in stage VIII seminiferous tubules, no staining was found at the apical ES in elongated spermatids (Fig. 3), consistent with the results of an earlier report (Fulcher et al., 1993). Based on these preliminary findings, it is apparent that another yet‐to‐be identified laminin(s) is present in developing round spermatids that serves as an anchoring device at the spermatid–Sertoli cell interface. 3. Nidogen Nidogen is an ubiquitously expressed protein in the basement membrane of multiple epithelia with two known isoforms. Nidogen‐1, also called entactin, is a 150‐kDa rod‐shaped molecule. Nidogen‐2, 200 kDa, shares 46% sequence homology with nidogen‐1, but the aYnity of nidogen‐2 toward laminin is about 100‐ to 1000‐fold less than nidogen‐1. Nidogen binds to laminin

1 chain via the epidermal growth factor‐like (LE) module, 1III4 (Mayer et al., 1993). The G2 domain of nidogen binds to type IV collagen, stabilizing the collagen network formation of the basement membrane (Fox et al., 1991). The interaction between nidogen and laminin is physiologically significant since gene targeting deletion of the 1III4 sequence in mice led to embryonic lethality due to severe defects in kidney and lung development caused by basement membrane ruptures (Willem et al., 2002). In testes, entactin was localized to the basement membrane of seminiferous tubules (Lian et al., 1992). It is secreted by Sertoli, peritubular myoid, and Leydig cells but not by germ cells (Lian et al., 1992).

IV. Role of Androgens in Junction Dynamics in Testes A. Introduction Testosterone, the main androgenic steroid, is produced by Leydig cells in the interstitium under the regulation of luteinizing hormone (LH) released from the pituitary gland. In rats and humans, the normal intratesticular testosterone concentration is,
50–70 and 600 ng/ml while the total testosterone concentration in normal serum is, 2 and 5 ng/ml, respectively (Jarow et al., 2005; Turner et al., 1984; Zirkin et al., 1989). This thus illustrates that the level of intratesticular testosterone is
50‐ to 100‐folds higher than the systemic circulation. While the precise physiological function for such drastic diVerences between the two compartments is not entirely known, recent studies have shown that this high level of androgen in the testis is crucial to maintain spermatogenesis and other pertinent testicular function such as Sertoli–germ cell adhesion.

2. Junction Restructuring and Spermatogenesis

73

Androgen exerts its eVect by diVusing to the target cells following its binding onto its specific androgen receptor (AR) (Heinlein and Chang, 2002; Tsai and O’Malley, 1994). It is generally accepted that AR is found on Sertoli, peritubular myoid, and Leydig cells while the presence of AR on germ cells remains controversial. Besides this classical action of androgen via ligand–receptor interactions, there is a suggestion that androgen can exert its eVects in target cells through a more rapid and non‐AR‐dependent mechanism, which is the MAP kinase pathway by stimulating the CREB and Src/ERK phosphorylation in Sertoli cells (Cheng et al., 2007; Fix et al., 2004; Walker, 2003). Apart from regulating the development and diVerentiation of male reproductive organs during embryogenesis and at puberty, androgen was recently shown to regulate junction restructuring events during spermatogenesis. Also, androgen was shown to aVect GJs in prostate cancer cells (Mitra et al., 2006). For instance, in the absence of androgen, rapid degradation of connexin 32 and 43, but not other TJ or AJ molecules, were detected; whereas androgen replacement rescued such unwanted degradation (Mitra et al., 2006).

B. Models to Study the Role of Androgen in Spermatogenesis The significance of androgen in spermatogenesis has been widely studied utilizing several well‐established in vivo and in vitro models. For instance, the cell‐specific knockout of androgen receptor in diVerent testicular cell types allowed investigators to identify more precisely the functional significance of androgen in testes (Chang et al., 2004; de Gendt et al., 2004; Holdcraft and Braun, 2004; Tsai et al., 2006; Yeh et al., 2002; Zhang et al., 2006). It is apparent that the primary cellular target of androgen in the testis is the Sertoli cell (Table III). For instance, specific knockout of AR in Sertoli and Leydig cells, but not germ and peritubular myoid cells, led to infertility in mice (Chang et al., 2004; de Gendt et al., 2004; Holdcraft and Braun, 2004; Tsai et al., 2006; Zhang et al., 2006). Besides, S‐AR/y (S: Sertoli cell) mice displayed the phenotype with BTB impairment, which was accompanied by a significant drop in the expression of claudin‐3 (Meng et al., 2005), caludin‐11, and occludin (Wang et al., 2006), yet other AJ markers such as N‐cadherin, nectin‐2, and laminin 3 were not aVected (Wang et al., 2006). In reviewing the knockout models published in the field, several crucial points are highlighted here. First, the serum testosterone levels of T‐AR/y, S‐AR/y, and L‐AR/y (T: testis, L: Leydig cell) tended to be significantly reduced as compared to wild types (Tsai et al., 2006), yet the serum testosterone level of S‐AR/y was considered as ‘‘no significant diVerence’’ and ‘‘significantly elevated’’ when compared with wild types in the knockout models of de Gendt et al. (2004) and Holdcraft and Braun (2004), respectively. Second, when the infertility of mice in the S‐AR/y models was carefully examined,

Table III EVects of Androgen Receptor Knockout on Spermatogenesis in Mice

Fertility

Testis Weight (% of Wild Type)

Testis (T‐AR/y) SC (S‐AR/y)

Infertile Infertile

6.9 23.4–28

GC (G‐AR/y) Myoid cell (PM‐AR/y)

Fertile Fertile

96.3 76a

Leydig cell (L‐AR/y)

Infertile

31.1

Sperm Count

Serum Testosterone Level

References

Azoospermia Azoospermia

Decrease Varies (see text for discussion)

Normal Oligozoospermia, 43% as wild type Azoospermia

Within normal range Within normal range

Yeh et al., 2002 Chang et al., 2004; de Gendt et al., 2004; Holdcraft and Braun, 2004 Tsai et al., 2006 Zhang et al., 2006

Decrease

Tsai et al., 2006

a Percentage of testis size, instead of weight, as of wild types. T, testis; SC, Sertoli cell; GC, germ cell; PM, peritubular myoid cell; L, Leydig cell; AR, androgen receptor.

2. Junction Restructuring and Spermatogenesis

75

both models from Chang et al. (2004) and de Gendt et al. (2004) illustrated spermatogenic arrest before the first meiosis in which there was an increase in ratio of tetraploid to haploid cells as well as a drastic decrease in number of round spermatids in the S‐AR/y mice. However, from the knockout model of Holdcraft and Braun (2004), meiosis was not aVected, instead, the development of round spermatids to elongating spermatids was disrupted in their S‐AR/y mice. Third, it must be cautious to note that in this latter model of Holdcraft and Braun, an attempt was made to flox exon 1 of the androgen receptor gene (note: for the other models, exon 2 was floxed) but the floxed animals had already displayed a marked hypomorphic phenotype and the knockout of AR in Sertoli cells was neither complete nor Sertoli cell selective (Holdcraft and Braun, 2004). Taking another well‐characterized in vivo model into consideration in which rats received low doses of testosterone and estradiol implants to induce an intratesticular androgen suppression, it was shown that this led to a premature detachment of developing spermatids (step 8 and beyond) from the Sertoli cells in the seminiferous epithelium (O’Donnell et al., 1996). While intact ES structure was still observed by electron and confocal microscopy in the Sertoli cell adluminal cytoplasm during and after round spermatid detachment (O’Donnell et al., 1996), this finding does not negate the fact that the adhesion functionality of the apical ES has been disrupted. For instance, a loss of adhesion between cadherins in MDCK can be rapidly induced as a result of [Ca2þ] depletion since cadherins are calcium‐dependent cell adhesion molecules, and this calcium depletion‐ induced AJ disruption also failed to elicit a dissolution of cadherins (for a review, see Cheng and Mruk, 2002). By immunohistochemistry, espin, an apical ES marker, was also detected even when there was no spermatid at the apical ES site in the seminiferous epithelium. Therefore, O’Donnell et al. (2000) suggested that there may be defects in adhesive molecules such as integrin between round spermatids and Sertoli cells (O’Donnell et al., 2000) which was disrupted as the result of a declining intratesticular androgen level. Also, this maybe the result of a degradation of vimentin into monomers in the Sertoli cells (Show et al., 2003). These findings thus illustrate that a disruption of the androgen function in adult rats can lead to a compromise of apical ES function as early as in step 8 spermatids. Collectively, these data seemingly suggest that the action of androgen and AR on spermatogenesis in adult animals is at the early stage of meiosis rather than the postmeiotic development of spermatids. In this context, it is of interest to note that the androgen suppression in vivo model has been used by two diVerent groups to study ES dynamics. It was shown that the level of Tyr‐phosphorylated FAK (p‐FAK) was induced and the association with 6 1‐integrin persisted even after the disappearance of apical ES during germ cell depletion (Beardsley et al., 2006) and earlier studies have shown that FAK and p‐FAK are restricted to basal and apical ES, respectively (Siu et al., 2003b). Similarly, there was an

76

Yan et al.

induction in the protein levels of 1‐integrin, p‐FAK and c‐Src in the androgen suppression rats, which was coupled with an increase in the association of p‐FAK/c‐Src protein complex at the apical ES (Siu et al., 2003b; Wong et al., 2005b), illustrating such an increase in kinases may somehow alter the cell adhesion functionality of proteins at the Sertoli–spermatid interface. There are also reports in the literature in which studies were conducted using the in vitro models to examine the role of testosterone in junction dynamics. For instance, Sertoli cells were cultured on Matrigel‐coated dishes to allow the establishment of functional TJ and AJ. The inclusion of testosterone at 2  107 M in these cultures was found to enhance TJ and AJ formation between Sertoli cells with an induction of steady‐state mRNA and/ or protein levels of occludin (Chung and Cheng, 2001), claudin‐1, 11 (Florin et al., 2005; Gye, 2003), E‐cadherin, and catenin (Lee et al., 2003). However, in a study, it was claimed that the inter‐Sertoli cell TJ and ES were stimulated by only FSH but not testosterone (Sluka et al., 2006). Furthermore, cocultures of Sertoli and germ cells were also used to examine the role of androgen on apical ES function. For instance, it was shown that maximal spermatid binding to Sertoli cells required the presence of both FSH and testosterone (Cameron and MuZy, 1991). Also, treatment of Sertoli–germ cell cocultures with 2  107 M testosterone was shown to stimulate AJ formation between Sertoli and germ cells which was accompanied by an increase in the steady‐ state protein levels of adaptors and signaling molecules, such as ‐catenin and c‐Src, respectively (Zhang et al., 2005). This androgen eVect appears to be specific since the presence of a specific anti‐androgen, cyproterone acetate (1  106 M), could abolish the stimulating eVects of androgen in the same experiment (Zhang et al., 2005). In short, it is increasingly clear that testosterone promotes cell adhesion at the Sertoli–germ or Sertoli–Sertoli cell interface in the seminiferous epithelium of adult testes.

V. Regulation of Junction Turnover by Protein Endocytosis and Recycling A. An Overview Endocytosis is a multistep cellular process. It includes the budding/invaginations of plasma membrane, followed by vesicle formation, and delivery of vesicles into diVerent specific intracellular compartments (Maxfield and McGraw, 2004). Cell surface proteins can be endocytosed via three diVerent endocytic structures: clathrin‐coated pits, caveolae, or macropinosomes (Ivanov et al., 2005; Maxfield and McGraw, 2004; Mukherjee et al., 1997). Internalized proteins are first sent to early endosomes, and then to the

2. Junction Restructuring and Spermatogenesis

77

recycling endosomes. At this stage, proteins will either be recycled back to cell surface or be sent to the late endosomes for degradation (Ivanov et al., 2005). Endocytic recycling of junction proteins plays a pivotal role in junction restructuring events pertinent to cell migration, epithelial proliferation, and tumorigenesis (Fujito et al., 2005; Janda et al., 2006; Matsuda et al., 2004). In a study using mouse Eph4 cells, intercellular movement was shown to be enhanced by coendocytosis of apposing TJ‐integral membrane proteins such as claudin‐3 into the adjacent cells in vitro (Matsuda et al., 2004). In high cell density cultures of NIH3T3 cells, downregulation of Necl‐5, an AJ transmembrane protein, via endocytosis can lead to inhibition of cell movement and proliferation, which apparently is the means used by these cells as a protective measure to avoid over‐crowding (Fujito et al., 2005). In the epithelial–mesenchymal transition (EMT) stage of tumor progression induced by Raf and transforming growth factor‐ (TGF‐ ), there was a gradual loss in the levels of junction molecules such as E‐cadherin, possibly by enhancing lysosomal degradation instead of its recycling (Janda et al., 2006). Furthermore, intracellular traYcking of junction molecules provides an eYcient mechanism in regulating the barrier permeability without requiring de novo protein synthesis (Harhaj and Antonetti, 2004; Ivanov et al., 2004; Laukoetter et al., 2006). Also, there is growing evidence that intestinal inflammatory diseases, such as Crohn’s disease (Laukoetter et al., 2006), are caused by the abnormal internalization of integral membrane proteins mediated by cytokines. For instance, interferon‐ (IFN ), a proinflammatory cytokine, was shown to increase the intestinal permeability by accelerating the kinetics of internalization of integral membrane TJ proteins, such as occludin, JAM‐A, and claudin‐1, in T84 cells (Bruewer et al., 2005; Utech et al., 2005).

B. Recent Advances on Studies Investigating the Role of Endocytosis in Junction Dynamics in the Testis In the testis, it is expected that endocytosis, similar to other epithelia, is crucial in the regulation of BTB dynamics, which facilitates preleptotene spermatocyte migration that occurs at stage VIII of the seminiferous epithelial cycle of spermatogenesis. For instance, cytokines, such as TGF‐ 3 and TNF‐, have been shown to increase the BTB permeability with a reduction in the steady‐ state integral membrane protein levels of TJ such as occludin and claudin‐11 via diVerent signaling pathways both in vitro (Lui et al., 2001; Siu et al., 2003a) and in vivo (Li et al., 2006; Lui et al., 2003b). Testosterone and FSH have been shown to tighten the TJ‐permeability barrier integrity by enhancing the TJ and AJ protein expression (Chung and Cheng, 2001; Gye, 2003; Sluka et al., 2006). However, the underlying mechanism(s) that up‐ and/or downregulate the

78

Yan et al.

steady‐state protein levels following cytokine/steroid treatments remains to be determined. Is this simply the result of a decline in de novo protein synthesis or of a change in the kinetics of endocytosis/endocytic recycling of TJ‐ and AJ‐ integral membrane proteins or both? In other epithelia, GTPase is a key player of intracellular traYcking. For instance, Morimoto et al. (2005) have showed that endocytic recycling of TJ proteins (e.g., occludin) was mediated by Rab 13, while internalization of AJ proteins (e.g., E‐cadherin) was shown to activate Rap1, enhancing integrin‐based adhesion (Balzac et al., 2005). In the testis, Rab8B was shown to structurally associate with E‐cadherin at both the BTB and the apical ES (Lau and Mruk, 2003). Since the colocalization of E‐cadherin and Rab5 in early endosomes has been demonstrated in MDCK cells (Le et al., 1999), it is logical to speculate that Rab8B may assist the intracellular traYcking of E‐cadherin, which, in turn, facilitates AJ restructuring during spermatogenesis. Recent studies from our laboratory using cell surface biotinylation coupled with immunoblotting techniques in primary Sertoli cell cultures with functional BTB have illustrated that BTB‐associated integral membrane proteins, such as occludin, JAM‐A, and N‐cadherin, were capable of being internalized and recycled back to the cell surface (Yan and Cheng, unpublished observations). The pathway(s) by which these BTB‐integral membrane proteins was utilized for their internalization is not known; however, based on a recent study, apical junction proteins, such as E‐cadherin, occludin, claudins, and JAM‐A, at the intestinal epithelia were being internalized via the clathrin‐ mediated pathway (Ivanov et al., 2004). Dynamins, members of a large GTPase family, are involved in the clathrin‐mediated endocytosis. For instance, dynamins serve as the mechanoenzyme that wrap around the neck of clathrin‐coated pits involving in the event of membrane fission (Hinshaw, 2006; Thompson and McNiven, 2001). Among the three isoforms of dynamin, dynamins‐2 and ‐3 are highly expressed in the testis, particularly in Sertoli and germ cells (Iguchi et al., 2002; Kamitani et al., 2002; Lie et al., 2006) while dynamin‐1 is neuron‐specific (Cao et al., 1998). Dynamin‐2 was shown to interact with occludin‐ and cadherin‐based protein complexes at the BTB (Lie et al., 2006). Using Adjudin to induce anchoring junction restructuring in the seminiferous epithelium in vivo, an increase in protein–protein association between dynamin‐2 and adaptors such as ‐catenin and ZO‐1 was detected (Lie et al., 2006). These changes, however, were accompanied by a loss in the interaction between dynamin‐2 and integral membrane proteins N‐cadherin and occludin (Lie et al., 2006). It is postulated that dynamin‐2 facilitates the disengagement of ZO‐1 and catenins such that these adaptors only bind to their corresponding integral membrane proteins. These findings thus support a recent model illustrating the timely disengagement between ZO‐1 and catenins that was used to reinforce the BTB integrity to facilitate preleptotene spermatocyte migration across the BTB (Yan and Cheng, 2005). Nonetheless, there is still a lack of functional experiments in the field to elucidate the precise role of dynamins in

2. Junction Restructuring and Spermatogenesis

79

testes unless there are testis‐specific dynamin knockout mouse models available or cell‐specific silencing of dynamins in Sertoli cells in vitro. Still, it is likely that dynamin‐2 may be involved in the endocytosis of the BTB‐associated integral membrane proteins. If this is indeed the case, this would be a rather complicated event since TJ, AJ (e.g., basal ES and basal TBC), and desmosome‐like junctions are also coexisting at the BTB, and some diVerential mechanisms must be in place to orchestrate these cellular processes. Even though the role of protein endocytosis and recycling in BTB dynamics remains obscure, recent findings have illustrated TBC is utilizing endocytosis to facilitate spermiation. Earlier studies have implicated the role of TBC in facilitating the release of mature spermatids via internalization of TBC‐ junction molecules (Pelletier and Byers, 1992). This postulate was supported by findings of Guttman et al. (2004) who demonstrated TBC indeed appeared at the concave surface of the head of elongated spermatids that was previously occupied by apical ES. More important, it was shown that the adhesion domains of nectin‐2 and ‐3 were found to be internalized as membrane vesicles near the TBC at spermiation (Guttman et al., 2004). The formation of membrane vesicles is probably mediated by the recruitment of dynamin‐3 to the TBC site as dynamin‐3 is an integral TBC component (Vaid et al., 2007). The fate of these internalized adhesion molecules at TBC remains to be determined. They can either be degraded intracellularly via endosomes or be recycled back to the cell surface. In other epithelia, however, most of the internalized integral membrane proteins, such as E‐cadherin and occludin, enter the recycling pathway so that they can be rapidly recycled back to the cell surface to maintain junction integrity, especially in unstable cell–cell contacts or in response to rapid changes in the cellular microenvironment (Le et al., 1999; Morimoto et al., 2005). For instance, a significant increase in E‐cadherin recycling was detected in MDCK cells following calcium replacement after [Ca2þ] depletion‐induced loss of cell adhesion (Le et al., 1999). Obviously, much research is needed in this area to investigate the role(s) of integral membrane protein internalization and recycling at the BTB, TBC, and perhaps ES to facilitate germ cell movement and/or spermatid orientation during spermatogenesis. These studies can also identify new targets for male contraceptive research, such as by disrupting the events of protein internalization and/or recycling at the BTB and/or apical ES, compromising cell adhesion and spermatogenesis.

VI. Regulation of Junction Dynamics by Myoid Cells It is of interest to note that the events of junction restructuring that occur in the seminiferous epithelium during the seminiferous epithelial cycle are regulated, at least in part, by components of the tunica propria, such as

80

Yan et al.

peritubular myoid cells. Myoid cells form a continuous single cell layer surrounding each seminiferous tubule in rat testes (Maekawa et al., 1996). There are some evidences that the myoid cell layer contributes partially to the integrity of the BTB since it functions as a ‘‘second’’ barrier in the testes besides the BTB, which is assembled by adjacent Sertoli cells near the basement membrane (Dym and Fawcett, 1970). For instance, it is known for decades that molecules such as colloidal carbon or thorium dioxide were excluded by the myoid cell barrier at the tunica propria, and lanthanum penetrated beyond the myoid cell layer in only
15% of the tubules in rodent testes (Dym and Fawcett, 1970; Fawcett et al., 1970), illustrating the peritubular myoid cell layer in the tunica propria contributes significantly to the BTB integrity. This also indicates the presence of possible cross talk between myoid and Sertoli cells (see Fig. 1). Indeed, myoid cells have been reported to secrete cytokines such as TGF‐ (Skinner and Moses, 1989). TGF‐ 3 has been shown to perturb AJ and TJ by regulating the steady‐state levels of integral membrane proteins at the BTB between Sertoli cells as well as at the apical ES between Sertoli cells and developing spermatids via the p‐38 MAPK or the Ras/ERK pathway (Lui et al., 2003b; Xia and Cheng, 2005). While these earlier studies focused mostly on cytokines secreted by Sertoli and/or germ cells, myoid cells can plausibly play a prominent role in determining the ‘‘opening’’ or the ‘‘closing’’ status of the BTB during the seminiferous epithelial cycle. Furthermore, other studies have shown that Sertoli cell secretory function is regulated by myoid cells, at least in vitro (Zwain et al., 1993). A study using a conditional knockout model of androgen receptors in myoid cells has illustrated that these mutant mice displayed oligozoospermia without causing infertility (Zhang et al., 2006). While there was no major defect in fertility when the androgen receptors were deleted in myoid cells, there were defects in several Sertoli cell junction proteins (Zhang et al., 2006). For instance, the steady‐state mRNA levels of AJ proteins such as N‐cadherin, vinculin, laminin 3, nectin; GJ proteins such as connexin 43; and TJ proteins such as occludin; were significantly reduced versus the wild type. The lack of a complete loss of fertility reflects the fact that even though the myoid cell layer contributes to the BTB function and other junction restructuring events in the testis, the myoid cell layer is playing only a supporting role in junction dynamics in the seminiferous epithelium, consistent with the functional tests assessing the diVusion of lanthanum across the BTB (see above). Nonetheless, these findings illustrate some unexplained cases of infertility such as oligozoospermia in men could be the result of a malfunctioning of peritubular myoid cells. It is likely that the myoid cell exerts its regulatory role via its secretory products (e.g., TGF‐ ) since this cell layer is separated from Sertoli cells and spermatogonia in the seminiferous epithelium by the acellular zone composed of basement membrane (e.g., type IV collagens) and type I collagen network. Perhaps some work is needed using

2. Junction Restructuring and Spermatogenesis

81

Sertoli‐myoid cell cocultures to further explore the functional relationship between these two cell types regarding their role in contributing to the BTB integrity.

VII. Environmental Toxicants: Are They Targeting the Tight and/or Anchoring Junction? In recent years, numerous reports are found in the literature citing the progressive decline of sperm counts and semen quality in men particularly for those residing in industrialized nations (Hess, 2003; Shaha, 2007; Sharpe, 2004). This by and large is the result of over exposure of men to environmental toxicants, which include heavy metals (e.g., cadmium, lead, mercury) (Henson and Chedrese, 2004; Pant et al., 2003; Waalkes et al., 1991), anti‐ androgens (e.g., vinclozolin, procymidone, linuron) (Gray et al., 2006; Kang et al., 2004), and estrogens (e.g., bisphenol A, phthalate) (MaYni et al., 2006; Welshons et al., 2006). Many of these toxicants are endocrine‐disrupting chemicals (EDCs) alike since the hypothalamic–pituitary–testicular axis is one of the known primary targets of these chemicals, possibly because of their androgenic or estrogenic activities that mimic the endogenous steroids in the systemic circulation. Some of these compounds thus create a negative feedback eVect on the hypothalamic gonadotropin secretion, perturbing the intratesticular androgen level, this, in turn, perturbs germ cell adhesion in the seminiferous epithelium. As such, premature release of germ cells from the testis causes a reduced sperm count in semen. For instance, recent studies have shown that environmental toxicants (e.g., CdCl2, bisphenol A, DDT) disrupt Sertoli–Sertoli cell junctions by either reducing the steady‐state levels or inducing the aberrant localization of occludin, ZO‐1, N‐cadherin, and connexin 43 (Fiorini et al., 2004; Wong et al., 2004). While these hazardous eVects to the reproductive function are alarming, not much can be done among government agencies for the general public since these toxicants are now an integrated part of our daily activities as they are present in water, air, foods (both raw and processed products, such as vegetables, meats, milk, fruits), cigarettes, beverages, plastics, and virtually every items found in the household (for further information, see http://www.e-b-i.net/ebi/contaminants/ cadmium; http://www.corrosion-doctors.org/Elements-Toxic/Cadmium.htm; http://www.environmentcalifornia.org/environmental-health/stop-toxic-toys/ bisphenol-a-overview). Reproductive organs, such as the testes, are sensitive to the toxic eVects of the EDCs (Garside and Harvey, 1992; Li and Heindel, 1998; Mantovani and Maranghi, 2005; Safe, 2005). The biochemical basis for this unusual vulnerability of the testes toward environmental toxicants, such as cadmium,

82

Yan et al.

bisphenol A, DDT, and others is not known. We postulate that this is likely the result of the unique arrangement of AJ and TJ in the seminiferous epithelium of the testes in particular at the BTB. For instance, studies have shown that E‐cadherin is the likely cellular target of cadmium toxicity (Prozialeck, 2000; Prozialeck et al., 2003). In other epithelia and endothelia, E‐cadherin, an integral membrane component of the AJ, is restricted to the adherens plague located behind the TJ barrier, such as in the blood vessel endothelium, skin, and intestinal epithelium. Thus, cadmium is being sealed oV from the E‐cadherin‐ based AJ because of the TJ‐barrier. However, BTB is formed by adjacent Sertoli cells at the basal compartment of the seminiferous epithelium. Additionally, TJ coexists structurally with basal ES, a testis‐specific AJ type (Yan and Cheng, 2005), at the BTB; thereby, E‐cadherin, a component of the basal ES, is present side‐by‐side with occludin and ZO‐1 which are components of the TJ. As such, environmental toxicants uptaken from the environment have immediate access to their primary targets once they reach the BTB in the testis when they leak through the endothelial TJ‐barrier in the microvessels at the interstitium, causing extensive damage there more rapidly than other blood– tissue barriers. This postulate is supported by a recent study that determined the kinetics of BTB damage versus microvascular damage, illustrating that BTB was damaged by cadmium prior to microvessel damage in the interstitium in adult rat testes (Wong et al., 2005a). Much work is needed in this area of research to define the molecular and biochemical basis of vulnerability of the testes toward environmental toxicants. In this context, it is of interest to note that the entry of toxicants (e.g., cadmium, mercury, lead) to an organ, such as the testis, via an epithelium (e.g., Sertoli cells) and/or endothelium (e.g., microvessels in the interstitium) is mediated via a transporter. Recent studies have identified the Slc39a8 gene that encodes SLC39A8 (ZIP8) facilitates the cadmium influx into the mouse testis, which is localized predominantly to the microvessels in the interstitium and it is also found sparingly in Sertoli cells near the basement membrane, in its absence, mice became resistant to cadmium toxicity (Dalton et al., 2005). It remains to be shown, however, whether this cadmium transporter gene is indeed required for Sertoli cell toxicity since environmental toxicants were shown to be toxic to germ or Leydig cells (Mantovani and Maranghi, 2005; McClusky et al., 2007; Svechnikov et al., 2005), yet, expression of SLC39A8 or other possible transporters was not visibly detected in either Leydig or germ cells (Dalton et al., 2005).

VIII. Concluding Remarks This chapter summarizes recent findings in the field regarding the regulation of junction dynamics during spermatogenesis. It is increasingly clear that ECM proteins, steroids, and GTPases are playing a crucial role in regulating junction

2. Junction Restructuring and Spermatogenesis

83

restructuring events in the seminiferous epithelium during the epithelial cycle of spermatogenesis. These molecules apparently are working in concert with cytokines, proteases, protease inhibitors as well as other yet‐to‐be identified biomolecules at the microenvironment of the BTB that determine the status of the cell–cell interface—whether the junctions should be restructured to induce their ‘‘opening’’ or remain ‘‘closed.’’ These junction restructuring events, however, must precisely coordinate with other molecular and cellular changes in the seminiferous epithelium, such as meiosis, spermiogenesis, spermatogonial proliferation, perhaps via paracrine and autocrine factors, and gap communicating junctions. Thus, precise ‘‘cross talks’’ occur between testicular cells to coordinate these events. Interestingly, very few biochemical and functional studies are found in the literature to address this important aspect of spermatogenesis. A recent report has demonstrated that a blockage of laminin function at the apical ES can lead to a transient opening of the BTB; implicating the presence of cross talk between the apical and basal compartments of the seminiferous epithelium (Yan and Cheng, 2006) (Fig. 1). Furthermore, the myoid cell‐specific knockout of androgen receptors can also lead to a declining expression of several TJ, AJ, and GJ proteins (Zhang et al., 2006), demonstrating the presence of cross talk between cells in the seminiferous epithelium and the cellular zone of the tunica propria even though they are physically segregated by the acellular zone of the basement membrane and the type I collagen fibril network (Fig. 1). It is apparent that this is a pivotal area of research deserving much attention by investigators in the field.

Acknowledgments This work was supported in part by grants from the National Institutes of Health (NICHD, U54 HD029990 Project 3, and U01 HD045908) and the CONRAD Program (CICCR, CIG 01‐72 to CYC and CIG 01‐74 to DDM).

References Andac, Z., Sasaki, T., Mann, K., Brancaccio, A., Deutzmann, R., and Timpl, R. (1999). Analysis of heparin, ‐dystroglycan and sulfatide binding to the G domain of the laminin 1 chain by site‐directed mutagenesis. J. Mol. Biol. 287, 253–264. Aumailley, M., and Smyth, N. (1998). The role of laminins in basement membrane function. J. Anat. 193, 1–21. Balzac, F., Avolio, M., Degani, S., Kaverina, I., Torti, M., Silengo, L., Small, J. V., and Retta, S. F. (2005). E‐cadherin endocytosis regulates the activity of Rap1: A traYc light GTPase at the crossroads between cadherin and integrin function. J. Cell Sci. 118, 4765–4783. Bart, J., Groen, H. J. M., van der Graaf, W. T., Hollema, H., Hendrikse, N. H., Vaalburg, W., Sleijfer, D. T., and de Vries, E. G. E. (2002). An oncological view on the blood–testis barrier. Lancet Oncol. 3, 357–363.

84

Yan et al.

Beardsley, A., Robertson, D. M., and O’Donnell, L. (2006). A complex containing 6/ 1 integrin and phosphorylated FAK between Sertoli cells and elongated spermatids during spermatid release from the seminiferous epithelium. J. Endocrinol. 190, 759–770. Brucato, S., Harduin‐Lepers, A., Godard, F., Bocquet, J., and Villers, C. (2000). Expression of glypican‐1, syndecan‐1 and syndecan‐4 mRNAs protein kinase C‐regulated in rat immature Sertoli cells by semi‐quantitative RT‐PCR analysis. Biochim. Biophys. Acta 1474, 31–40. Bruewer, M., Utech, M., Ivanov, A. I., Hopkins, A. M., Parkos, C. A., and Nusrat, A. (2005). Interferon‐gamma induces internalization of epithelial tight junction proteins via a macropinocytosis‐like process. FASEB J. 19, 923–933. Cameron, D. F., and MuZy, K. E. (1991). Hormonal regulation of spermatid binding. J. Cell Sci. 100, 623–633. Cao, H., Garcia, F., and McNiven, M. A. (1998). DiVerential distribution of dynamin isoforms in mammalian cells. Mol. Biol. Cell 9, 2595–2609. Chang, C., Chen, Y. T., Yeh, S. D., Xu, Q., Wang, R. S., Guillou, F., Lardy, H., and Yeh, S. (2004). Infertility with defective spermatogenesis and hypotestosteronemia in male mice lacking the androgen receptor in Sertoli cells. Proc. Natl. Acad. Sci. USA 101, 6876–6881. Cheng, C. Y., and Mruk, D. D. (2002). Cell junction dynamics in the testis: Sertoli‐germ cell interactions and male contraceptive development. Physiol. Rev. 82, 825–874. Cheng, J., Watkin, S. C., and Walker, W. H. (2007). Testosterone activates mitogen‐activated protein kinase via Src kinase and the epidermal growth factor receptor in Sertoli cells. Endocrinology 148, 2066–2074. Chung, N. P. Y., and Cheng, C. Y. (2001). Is cadmium chloride‐induced inter‐Sertoli tight junction permeability barrier disruption a suitable in vitro model to study the events of junction disassembly during spermatogenesis in the rat testis? Endocrinology 142, 1878–1888. Colognato, H., MacCarrick, M., O’Rear, J. J., and Yurchenco, P. D. (1997). The laminin 2‐chain short arm mediates cell adhesion through both the 1 1 and 2 1 integrins. J. Biol. Chem. 272, 29330–29336. Colognato‐Pyke, H., O’Rear, J. J., Yamada, Y., Carbonetto, S., Cheng, Y. S., and Yurchenco, P. D. (1995). Mapping of network‐forming, heparin‐binding, and 1 1 integrin‐ recognition sites within the ‐chain short arm of laminin‐1. J. Biol. Chem. 270, 9398–9406. Dalton, T. P., He, L., Wang, B., Miller, M. L., Jin, L., Stringer, K. F., Chang, X., Baxter, C. S., and Nebert, D. W. (2005). Identification of mouse SLC39A8 as the transporter responsible for cadmium‐induced toxicity in the testis. Proc. Natl. Acad. Sci. USA 102, 3401–3406. Davis, C. M., Papadopoulos, V., Sommers, C. L., Kleinman, H. K., and Dym, M. (1990). DiVerential expression of extracellular matrix components in rat Sertoli cells. Biol. Reprod. 43, 860–869. de Gendt, K., Swinnen, J. V., Saunders, P. T. K., Schoonjans, L., Dewerchin, M., Devos, A., Tan, K., Atanassova, N., Claessens, F., Lecureuil, C., Heyns, W., Carmeliet, P., et al. (2004). A Sertoli cell‐selective knockout of the androgen receptor causes spermatogenic arrest in meiosis. Proc. Natl. Acad. Sci. USA 101, 1327–1332. de Kretser, D. M., and Kerr, J. B. (1988). ‘‘The Cytology of the Testis.’’ Raven, New York. Denduchis, B., Satz, M. L., Sztein, M. B., Puig, R. P., Doncel, G., and Lustig, L. (1985). Multifocal damage of the testis induced in rats by passive transfer of antibodies prepared against non‐collagenous fraction of basement membrane. J. Reprod. Immunol. 7, 59–75. Drees, F., Pokutta, S., Yamada, S., Nelson, W. J., and Weis, W. I. (2005). ‐Catenin is a molecular switch that binds E‐cadherin‐ ‐catenin and regulates actin‐filament assembly. Cell 123, 903–915. Durbeej, M., Henry, M. D., Ferletta, M., Campbell, K. P., and Ekblom, P. (1998). Distribution of dystroglycan in normal adult mouse tissues. J. Histochem. Cytochem. 46, 449–457. Durbeej, M., Talts, J. F., Henry, M. D., Yurchenco, P. D., Campbell, K. P., and Ekblom, P. (2001). Dystroglycan binding to laminin 1LG4 module influences epithelial morphogenesis of salivary gland and lung in vitro. DiVerentiation 69, 121–134.

2. Junction Restructuring and Spermatogenesis

85

Dym, M. (1994). Basement membrane regulation of Sertoli cells. Endocr. Rev. 15, 102–115. Dym, M., and Fawcett, D. W. (1970). The blood‐testis barrier in the rat and the physiological compartmentation of the seminiferous epithelium. Biol. Reprod. 3, 308–326. Ekblom, P. (1996). Receptors for laminins during epithelial morphogenesis. Curr. Opin. Cell Biol. 8, 700–706. Enders, G. C., Kahsai, T. Z., Lian, G., Funabiki, K., Killen, P. D., and Hudson, B. G. (1995). Developmental changes in seminiferous tubule extracellular matrix components of the mouse testis: 3 (IV) collagen chain expressed at the initiation of spermatogenesis. Biol. Reprod. 53, 1489–1499. Engvall, E., and Wewer, U. M. (1996). Domains of laminin. J. Cell. Biochem. 61, 493–501. Fawcett, D. W., Leak, L. V., and Heidger, P. M. J. (1970). Electron microscopic observations on the structural components of the blood‐testis barrier. J. Reprod. Fertil. Suppl. 10, 105–122. Fiorini, C., Tilloy‐Ellul, A., Chevalier, S., Charuel, C., and Pointis, G. (2004). Sertoli cell junctional proteins as early targets for diVerent classes of reproductive toxicants. Reprod. Toxicol. 18, 413–421. Fix, C., Jordan, C., Cano, P., and Walker, W. H. (2004). Testosterone activates mitogen‐ activated protein kinase and the cAMP response element binding protein transcription factor in Sertoli cells. Proc. Natl. Acad. Sci. USA 101, 10919–10924. Florin, A., Maire, M., Bozec, A., Hellani, A., Chater, S., Bars, R., Chuzel, F., and Benahmed, M. (2005). Androgens and postmeiotic germ cells regulate claudin‐11 expression in rat Sertoli cells. Endocrinology 146, 1532–1540. Fox, J. W., Mayer, U., Nischt, R., Aumailley, M., Reinhardt, D., Weidemann, H., Mann, K., Timpl, R., Krieg, T., Engel, J., and Chu, M. L. (1991). Recombinant nidogen consists of three globular domains and mediates binding of laminin to collagen type IV. EMBO J. 10, 3137–3146. Frojdman, K., Pelliniemi, L. J., and Virtanen, I. (1998). DiVerential distribution of type IV collagen chains in the developing rat testis and ovary. DiVerentiation 63, 125–130. Fujito, T., Ikeda, W., Kakunaga, S., Minami, Y., Kajita, M., Sakamoto, Y., Monden, M., and Takai, Y. (2005). Inhibition of cell movement and proliferation by cell‐cell contact‐induced interaction of Necl‐5 with nectin‐3. J. Cell Biol. 171, 165–173. Fulcher, K. D., Welch, J. E., Davis, C. M., O’Brien, D. A., and Eddy, E. M. (1993). Characterization of laminin receptor messenger ribonucleic acid and protein expression in mouse spermatogenic cells. Biol. Reprod. 48, 674–682. Garside, D. A., and Harvey, P. W. (1992). ‘‘Endocrine Toxicity of the Male Reproductive System.’’ Cambridge University, Cambridge. Geberhiwot, T., Ingerpuu, S., Pedraza, C., Neira, M., Lehto, U., Virtanen, I., Kortesmaa, J., Tryggvason, K., Engvall, E., and Patarroyo, M. (1999). Blood platelets contain and secrete laminin‐8 (4 1 1) and adhere to laminin‐8 via 6 1 integrin. Exp. Cell Res. 253, 723–732. GersdorV, N., Kohfeldt, E., Sasaki, T., Timpl, R., and Miosge, N. (2005). Laminin 3 chain binds to nidogen and is located in murine basement membrane. J. Biol. Chem. 280, 22146–22153. Giannelli, G., Falk‐Marzillier, J., Schiraldi, O., Stetler‐Stevenson, W. G., and Quaranta, V. (1997). Induction of cell migration by matrix metalloprotease‐2 cleavage of laminin‐5. Science 277, 225–228. Giebel, J., Loster, K., and Rune, G. M. (1997). Localization of integrin 1, 1, 5, 9 subunits in the rat testis. Int. J. Androl. 20, 3–9. Gliki, G., Ebnet, K., Aurrand‐Lions, M., Imhof, B. A., and Adams, R. H. (2004). Spermatid diVerentiation requires the assembly of a cell polarity complex downstream of junctional adhesion molecule‐C. Nature 431, 320–324. Goossens, S., Janssens, B., Vanpoucke, G., de Rycke, R., van Hengel, J., and van Roy, F. (2006). Truncated isoform of mouse T‐catenin is testis‐restricted in expression and function. FASEB J. 21, 647–655.

86

Yan et al.

Graf, J., Iwamoto, Y., Sasaki, M., Martin, G. R., Kleinman, H. K., Robey, F. A., and Yamada, Y. (1987). Identification of an amino acid sequence in laminin mediating cell attachment, chemotaxis, and receptor binding. Cell 48, 989–996. Gray, L. E. J., Wilson, V. S., Stoker, T., Lambright, C., Furr, J., Noriega, N., Howdenshell, K., Ankley, G. T., and Guillette, L. (2006). Adverse eVects of environmental antiandrogens and androgens on reproductive development in mammals. Int. J. Androl. 29, 96–104. Green, K. J., Bohringer, M., Gocken, T., and Jones, J. C. R. (2005). Intermediate filament associated proteins. Adv. Protein Chem. 70, 143–202. Gurney, K. J., Estrada, E. Y., and Rosenberg, G. A. (2006). Blood‐brain barrier disruption by stromelysin‐1 facilitates neutrophil infiltration in neuroinflammation. Neurobiol. Dis. 23, 87–96. Guttman, J. A., Takai, Y., and Vogl, A. W. (2004). Evidence that tubulobulbar complexes in the seminiferous epithelium are involved with internalization of adhesion junctions. Biol. Reprod. 71, 548–559. Gye, M. C. (2003). Changes in the expression of claudins and transepithelial electrical resistance of mouse Sertoli cells by Leydig cell coculture. Int. J. Androl. 26, 271–278. Hadley, M. A., Weeks, B. S., Kleinman, H. K., and Dym, M. (1990). Laminin promotes formation of cord‐like structures by Sertoli cells in vitro. Dev. Biol. 140, 318–327. Harhaj, N. S., and Antonetti, D. A. (2004). Regulation of tight junctions and loss of barrier function in pathophysiology. Int. J. Biochem. Cell Biol. 36, 1206–1237. Harvey, S. J., Perry, J., Zheng, K., Chen, D., Sado, Y., JeVerson, B., Ninomiya, Y., Jacobs, R., Hudson, B. G., and Thorner, P. S. (2006). Sequential expression of type IV collagen networks: Testis as a model and relevance to spermatogenesis. Am. J. Pathol. 168, 1587–1597. He, Z., Feng, L., Zhang, X., Geng, Y., Parodi, D. A., Suarez‐Quian, C. A., and Dym, M. (2005). Expression of Col1a1, Col1a2 and procollagen I in germ cells of immature and adult mouse testis. Reproduction 130, 333–341. Heinlein, C. A., and Chang, C. (2002). Androgen receptor (AR) coregulators: An overview. Endocr. Rev. 23, 175–200. Henson, M. C., and Chedrese, P. J. (2004). Endocrine disruption by cadmium, a common environmental toxicant with paradoxical eVects on reproduction. Exp. Biol. Med. 229, 383–392. Hess, R. A. (2003). Estrogen in the adult male reproductive tract: A review. Reprod. Biol. Endocrinol. 1, 52–66. Hess, R. A., SchaeVer, D. J., Eroschenko, V. P., and Keen, J. E. (1990). Frequency of the stages in the cycles of the seminiferous epithelium in the rat. Biol. Reprod. 43, 517–524. Hinshaw, J. E. (2006). Filling the GAP for dynamin. Nat. Cell Biol. 8, 432–433. Holdcraft, R. W., and Braun, R. E. (2004). Androgen receptor function is required in Sertoli cells for the terminal diVerentiation of haploid spermatids. Development 131, 459–467. Hozumi, K., Suzuki, N., Nielsen, P. K., Nomizu, M., and Yamada, Y. (2006). Laminin 1 chain LG4 module promotes cell attachment through syndecans and cell spreading through integrin 2 1. J. Biol. Chem. 281, 32929–32940. Iguchi, H., Watanabe, M., Kamitani, A., Nagai, A., Hosoya, O., Tsutsui, K., and Kumon, H. (2002). Localization of dynamin 2 in rat seminiferous tubules during the spermatogenic cycle. Acta. Med. Okayama 56, 205–209. Ivanov, A. I., Nusrat, A., and Parkos, C. A. (2004). Endocytosis of epithelial apical junctional proteins by a clathrin‐mediated pathway into a unique storage compartment. Mol. Biol. Cell 15, 176–188. Ivanov, A. I., Nusrat, A., and Parkos, C. A. (2005). Endocytosis of the apical junctional complex: Mechanisms and possible roles in regulations of epithelial barriers. BioEssays 27, 356–365. Janda, E., Nevolo, M., Lehmann, K., Downward, J., Beug, H., and Grieco, M. (2006). Raf plus TGF ‐dependent EMT is initiated by endocytosis and lysosomal degradation of E‐cadherin. Oncogene 25, 7117–7130.

2. Junction Restructuring and Spermatogenesis

87

Jarow, J. P., Wright, W. W., Brown, T. R., Yan, X., and Zirkin, B. R. (2005). Bioactivity of androgens within the testes and serum of normal men. J. Androl. 26, 343–348. Kamitani, A., Yamada, H., Kinuta, M., Watanabe, M., Li, S. A., Matsukawa, T., McNiven, M., Kumon, H., and Takei, K. (2002). Distribution of dynamins in testis and their possible relation to spermatogenesis. Biochem. Biophys. Res. Commun. 294, 261–267. Kang, I. H., Kim, H. S., Shin, J. H., Kim, T. S., Moon, H. J., Kim, I. Y., Choi, K. S., Kil, K. S., Park, Y. I., Dong, M. S., and Han, S. Y. (2004). Comparison of anti‐androgenic activity of flutamide, vinclozolin, procymidone, linuron, and p,p‐DDE in rodent 10‐day Hershberger assay. Toxicology 199, 145–159. Kierszenbaum, A. L., Rosselot, C., Rivkin, E., and Tres, L. L. (2006). Role of integrins, tetraspanins, and ADAM proteins during the development of apoptotic bodies by spermatogenic cells. Mol. Reprod. Dev. 73, 906–917. Kikkawa, Y., and Miner, J. H. (2005). Lutheran/B‐CAM: A laminin receptor on red blood cells and in various tissues. Connect. Tissue Res. 46, 193–199. Kikkawa, Y., Sanzen, N., Fujiwara, H., Sonnenberg, A., and Sekiguchi, K. (2000). Integrin binding specificity of laminin‐10/11: Laminin‐10/11 are recognized by 3 1, 6 1 and 6 4 integrins. J. Cell Sci. 113, 869–876. Kikkawa, Y., Moulson, C. L., Virtanen, I., and Miner, J. H. (2002). Identification of the binding site for the Lutheran blood group glycoprotein on laminin 5 through expression of chimeric laminin chains in vivo. J. Biol. Chem. 277, 44864–44869. Kim, J. M., Park, W. H., and Min, B. M. (2005). The PPFLMLLKGSTR motif in globular domain 3 of the human laminin‐5 3 chain is crucial for integrin 3 1 binding and cell adhesion. Exp. Cell Res. 304, 317–327. Lamagna, C., Meda, P., Mandicourt, G., Brown, J., Gilbert, R. J. C., Jones, E. Y., Keifer, F., Ruga, P., Imhof, B. A., and Aurrand‐Lions, M. (2005). Dual interaction of JAM‐C with JAM‐B and M 2integrin: Function in junctional complexes and leukocyte adhesion. Mol. Biol. Cell 16, 4992–5003. Lau, A. S. N., and Mruk, D. D. (2003). Rab8B GTPase and junction dynamics in the testis. Endocrinology 144, 1549–1563. Laukoetter, M. G., Bruewer, M., and Nusrat, A. (2006). Regulation of the intestinal epithelial barrier by the apical junctional complex. Curr. Opin. Gastroenterol. 22, 85–89. Le, T. L., Yap, A. S., and Stow, J. L. (1999). Recycling of E‐cadherin: A potential mechanism for regulating cadherin dynamics. J. Cell Biol. 146, 219–232. Leblond, C. P., and Clermont, Y. (1952). Definition of the stages of the cycle of the seminiferous epithelium in the rat. Ann. N. Y. Acad. Sci. 55, 548–573. Lee, N. P. Y., Mruk, D. D., Lee, W. M., and Cheng, C. Y. (2003). Is the cadherin/catenin complex a functional unit of cell‐cell actin‐based adherens junctions in the rat testis? Biol. Reprod. 68, 489–508. Li, L. H., and Heindel, J. J. (1998). ‘‘Sertoli Cell Toxicants.’’ Marcel Dekker, New York. Li, M. W. M., Xia, W., Mruk, D. D., Wang, C. Q. F., Yan, H. H. N., Siu, M. K. Y., Lui, W. Y., Lee, W. M., and Cheng, C. Y. (2006). Tumor necrosis factor (alpha) reversibly disrupts the blood‐testis barrier and impairs Sertoli‐germ cell adhesion in the seminiferous epithelium of adult rat testes. J. Endocrinol. 190, 313–329. Lian, G., Miller, K. A., and Enders, G. C. (1992). Localization and synthesis of entactin in seminiferous tubules of the mouse. Biol. Reprod. 47, 316–325. Lie, P. P. Y., Xia, W., Wang, C. Q. F., Mruk, D. D., Yan, H. H. N., Wong, C. H., Lee, W. M., and Cheng, C. Y. (2006). Dynamin II interacts with the cadherin‐ and occludin‐based protein complexes at the blood‐testis barrier in adult rat testes. J. Endocrinol. 191, 571–586. Litjens, S. H. M., de Pereda, J. M., and Sonnenberg, A. (2006). Current insights into the formation and breakdown of hemidesmosomes. Trends Cell Biol. 16, 376–383.

88

Yan et al.

Longin, J., Guillaumot, P., Chauvin, M., Morera, A., and Le Magueresse‐Battistoni, B. (2001). MT1‐MMP in rat testicular development and the control of Sertoli cell proMMP‐2 activation. J. Cell Sci. 114, 2125–2134. Lui, W. Y., Lee, W. M., and Cheng, C. Y. (2001). Transforming growth factor‐ 3 perturbs the inter‐Sertoli tight junction permeability barrier in vitro possibly mediated via its eVects on occludin, zonula occludens‐1, and claudin‐11. Endocrinology 142, 1865–1877. Lui, W. Y., Mruk, D. D., Lee, W. M., and Cheng, C. Y. (2003a). Sertoli cell tight junction dynamics: Their regulation during spermatogenesis. Biol. Reprod. 68, 1087–1097. Lui, W. Y., Wong, C. H., Mruk, D. D., and Cheng, C. Y. (2003b). TGF‐ 3 regulates the blood‐ testis barrier dynamics via the p38 mitogen activated protein (MAP) kinase pathway: An in vivo study. Endocrinology 144, 1139–1142. Lustig, L., Doncel, G. F., Berensztein, E., and Denduchis, B. (1987). Testis lesions and cellular and humoral immune responses induced in rats by immunization with laminin. Am. J. Reprod. Immunol. Microbiol. 14, 123–128. Lustig, L., Denduchis, B., Ponzio, R., Lauzon, M., and Pelletier, R. M. (2000). Passive immunization with anti‐laminin immunoglobulin G modifies the integrity of the seminiferous epithelium and induces arrest of spermatogenesis in the guinea pig. Biol. Reprod. 62, 1505–1514. Maekawa, M., Kamimura, K., and Nagano, T. (1996). Peritubular myoid cells in the testis: Their structure and function. Arch. Histol. Cytol. 59, 1–13. MaYni, M. V., Rubin, B. S., Sonnenschein, C., and Soto, A. M. (2006). Endocrine disruptors and reproductive health: The case of bisphenol‐A. Mol. Cell. Endocrinol. 254–255, 179–186. Mann, K., Deutzmann, R., Aumailley, M., Timpl, R., Raimondi, L., Yamada, Y., Pan, T. C., Conway, D., and Chu, M. L. (1989). Amino acid sequence of mouse nidogen, a multidomain basement membrane protein with binding activity for laminin, collagen IV and cells. EMBO J. 8, 65–72. Mantovani, A., and Maranghi, F. (2005). Risk assessment of chemicals potentially aVecting male fertility. Contraception 72, 308–313. Matsuda, M., Kubo, A., Furuse, M., and Tsukita, S. (2004). A peculiar internalization of claudins, tight junction‐specific adhesion molecules, during the intercellular movement of epithelial cells. J. Cell Sci. 117, 1247–1257. Maxfield, F. R., and McGraw, T. E. (2004). Endocytic recycling. Nat. Rev. Mol. Cell Biol. 5, 121–132. Mayer, U., Nischt, R., Poschl, E., Mann, K., Fukuda, K., Gerl, M., Yamada, Y., and Timpl, R. (1993). A single EGF‐like motif of laminin is responsible for high aYnity nidogen binding. EMBO J. 12, 1879–1885. McClusky, L. M., de Jager, C., and Bornman, M. S. (2007). Stage‐related increase in the proportion of apoptotic germ cells and altered frequencies of stages in the spermatogenic cycle following gestational, lactational, and direct exposure of male rats to p‐Nonylphenol. Toxicol. Sci. 95, 249–256. Mecham, R. P. (1991). Receptors for laminin on mammalian cells. FASEB J. 5, 2538–2546. Meng, J., Holdcraft, R. W., Shima, J. E., Griswold, M. D., and Braun, R. E. (2005). Androgens regulate the permeability of the blood‐testis barrier. Proc. Natl. Acad. Sci. USA 102, 16696–16700. Mercurio, A. M. (1995). Laminin receptors: Achieving specificity through cooperation. Trends Cell Biol. 5, 419–423. Mirza, M., Hreinsson, J., Strand, M., Hovatta, O., Soder, O., Philipson, L., Pettersson, R. F., and Sollerbrant, K. (2006). Coxsackievirus and adenovirus receptor (CAR) is expressed in male germ cells and forms a complex with the diVerentiation factor JAM‐C in mouse testis. Exp. Cell Res. 312, 817–830.

2. Junction Restructuring and Spermatogenesis

89

Mitra, S., Annamalai, L., Chakraborty, S., Johnson, K., Song, X. H., Batra, S. K., and Mehta, P. P. (2006). Androgen‐regulated formation and degradation of gap junctions in androgen‐responsive human prostate cancer cells. Mol. Biol. Cell. 17, 5400–5416. Morimoto, S., Nishimura, N., Terai, T., Manabe, S., Yamamoto, Y., Shinahara, W., Miyake, H., Tashiro, S., Shimada, M., and Sasaki, T. (2005). Rab 13 mediates the continuous endocytic recycling of occludin to the cell surface. J. Biol. Chem. 280, 2220–2228. Mruk, D. D., and Cheng, C. Y. (2004a). Cell‐cell interactions at the ectoplasmic specialization in the testis. Trends Endocrinol. Metab. 15, 439–447. Mruk, D. D., and Cheng, C. Y. (2004b). Sertoli‐Sertoli and Sertoli‐germ cell interactions and their significance in germ cell movement in the seminiferous epithelium during spermatogenesis. Endocr. Rev. 25, 747–806. Mukherjee, S., Ghosh, R. N., and Maxfield, F. R. (1997). Endocytosis. Physiol. Rev. 77, 759–803. Mulholland, D. J., Dedhar, S., and Vogl, A. W. (2001). Rat seminiferous epithelium contains a unique junction (ectoplasmic specialization) with signaling properties both of cell/cell and cell/ matrix junctions. Biol. Reprod. 64, 396–407. Nakamura, Y., Handa, K., Iwamoto, R., Tsukamoto, T., Takahasi, M., and Mekada, E. (2001). Immunohistochemical distribution of CD9, heparin binding epidermal growth factor‐like growth factor, and integrin 3 1 in normal human tissues. J. Histochem. Cytochem. 49, 439–444. Nievers, M. G., Schaapveld, R. Q. J., and Sonnenberg, A. (1999). Biology and function of hemidesmosomes. Matrix Biol. 18, 5–17. Nishiuchi, R., Takagi, J., Hayashi, M., Ido, H., Yagi, Y., Sanzen, N., Tsuji, T., and Sekiguchi, K. (2006). Ligand‐binding specificities of laminin‐binding integrins: A comprehensive survey of laminin‐integrin interactions using recombinant 3 1, 6 1, 7 1 and 6 4 integrins. Matrix Biol. 25, 189–197. O’Donnell, L., MaLachlan, R. I., Wreford, N. G., de Kretser, D. M., and Robertson, D. M. (1996). Testosterone withdrawal promotes stage‐specific detachment of round spermatids from the rat seminiferous epithelium. Biol. Reprod. 55, 895–901. O’Donnell, L., Stanton, P. G., Bartles, J. R., and Robertson, D. M. (2000). Sertoli cell ectoplasmic specializations in the seminiferous epithelium of the testosterone‐suppressed adult rat. Biol. Reprod. 63, 99–108. Ohtsuka, T., Nakanishi, H., Ikeda, W., Satoh, A., Momose, Y., Nishioka, H., and Takai, Y. (1998). Nexilin: A novel actin filament‐binding protein localized at cell–matrix adherens junctions. J. Cell Biol. 143, 1227–1238. Pant, N., Upadhyay, G., Pandey, S., Mathur, N., Saxena, D. K., and Srivastava, S. P. (2003). Lead and cadmium concentration in the seminal plasma of men in the general population: Correlation with sperm quality. Reprod. Toxicol. 17, 447–450. Parvinen, M. (1982). Regulation of the seminiferous epithelium. Endocr. Rev. 3, 404–417. Pelletier, R. M., and Byers, S. W. (1992). The blood‐testis barrier and Sertoli cell junctions: Structural considerations. Microsc. Res. Tech. 20, 3–33. PfeiVer, D. C., and Vogl, A. W. (1991). Evidence that vinculin is co‐distributed with actin bundles in ectoplasmic (‘‘junctional’’) specializations of mammalian Sertoli cells Anat. Rec. 231, 89–100. Pointis, G., and Segretain, D. (2005). Role of connexin‐based gap junction channels in testis. Trends Endocrinol. Metab. 16, 300–306. Prozialeck, W. C. (2000). Evidence that E‐cadherin may be a target for cadmium toxicity in epithelial cells. Toxicol. Appl. Pharmacol. 164, 231–249. Prozialeck, W. C., Lamar, P. C., and Lynch, S. M. (2003). Cadmium alters the localization of N‐cadherin, E‐cadherin, and beta‐catenin in the proximal tubule epithelium. Toxicol. Appl. Pharmacol. 189, 180–195.

90

Yan et al.

Rousselle, P., Lunstrum, G. P., Keene, D. R., and Burgeson, R. E. (1991). Kalinin: An epithelium‐specific basement membrane adhesion molecule that is a component of anchoring filaments. J. Cell Biol. 114, 567–576. Russell, L. D. (1977a). Desmosome‐like junctions between Sertoli and germ cells in the rat testis. Am. J. Anat. 148, 301–312. Russell, L. D. (1977b). Movement of spermatocytes from the basal to the adluminal compartment of the rat testis. Am. J. Anat. 148, 313–328. Russell, L. D. (1977c). Observations on rat Sertoli ectoplasmic (‘junctional’) specializations in their association with germ cells of the rat testis. Tissue Cell 9, 475–498. Russell, L. D. (1979). Further observations on tubulobulbar complexes formed by late spermatids and Sertoli cells in the rat testis. Anat. Rec. 194, 213–232. Russell, L., and Clermont, Y. (1976). Anchoring device between Sertoli cells and late spermatids in rat seminiferous tubules. Anat. Rec. 185, 259–278. Safe, S. (2005). Clinical correlates of environmental endocrine disruptors. Trends Endocrinol. Metab. 16, 139–144. Salanova, M., Stefanini, M., de Curtis, I., and Palombi, F. (1995). Integrin receptor 6 1 is localized at specific sites of cell‐to‐cell contact in rat seminiferous epithelium. Biol. Reprod. 52, 79–87. Santoro, G., Romeo, C., Impellizzeri, P., Cutroneo, G., Micali, A., Trimarchi, F., and Gentile, C. (2000). Immunofluorescence distribution of actin‐associated proteins in human seminiferous tubules of adolescent testes, normal and pathologic. J. Endocrinol. Invest. 23, 369–375. Santoso, S., Orlova, V. V., Song, K., Sachs, U. J., Andrei‐Selmer, C. L., and Chavakis, T. (2005). The homophilic binding of junctional adhesion molecule‐C mediates tumor cell‐endothelial cell interactions. J. Biol. Chem. 43, 36326–36333. Shaha, C. (2007). Estrogens and spermatogenesis. ‘‘Molecular Mechanisms in Spermatogenesis’’ (C.Y. Cheng, Ed.). Landes Bioscience, Austin, TX (in press). Sharpe, R. M. (2004). Environmental oestrogens: Foe or friend? J. Neuroendocrinol. 16, 867–868. Show, M. D., Anway, M. D., Folmer, J. S., and Zirkin, B. R. (2003). Reduced intratesticular testosterone concentration alters the polymerization state of the Sertoli cell intermediate filament cytoskeleton by degradation of vimentin. Endocrinology 144, 5530–5536. Siu, M. K. Y., and Cheng, C. Y. (2004a). Dynamic cross‐talk between cells and the extracellular matrix in the testis. BioEssays 26, 978–992. Siu, M. K. Y., and Cheng, C. Y. (2004b). Interactions of proteases, protease inhibitors, and the 1 integrin/laminin 3 protein complex in the regulation of ectoplasmic specialization dynamics in the rat testis. Biol. Reprod. 70, 945–964. Siu, M. K. Y., Lee, W. M., and Cheng, C. Y. (2003a). The interplay of collagen IV, tumor necrosis factor‐, gelatinase B (matrix metalloprotease‐9), and tissue inhibitor of metalloproteases‐1 in the basal lamina regulates Sertoli cell‐tight junction dynamics in the rat testis. Endocrinology 144, 371–387. Siu, M. K. Y., Mruk, D. D., Lee, W. M., and Cheng, C. Y. (2003b). Adhering junction dynamics in the testis are regulated by an interplay of 1‐integrin and focal adhesion complex‐ associated proteins. Endocrinology 144, 2141–2163. Skinner, M. K., and Moses, H. L. (1989). Transforming growth factor‐beta gene expression and action in the seminiferous tubule: Peritubular cell‐Sertoli cell interactions. Mol. Endocrinol. 3, 625–634. Sluka, P., O’Donnell, L., Bartles, J. R., and Stanton, P. G. (2006). FSH regulates the formation of adherens junctions and ectoplasmic specialisations between rat Sertoli cells in vitro and in vivo. J. Endocrinol. 189, 381–395. Svechnikov, K., Supornsilchai, V., Strand, M. L., Wahlgren, A., Seidlova‐Wuttke, D., Wuttke, W., and Soder, O. (2005). Influence of long‐term dietary administration of procymidone, a fungicide with anti‐androgenic eVects, or the phytoestrogen genistein to rats on the pituitary‐gonadal axis and Leydig cell steroidogenesis. J. Endocrinol. 187, 117–124.

2. Junction Restructuring and Spermatogenesis

91

Terada, N., Ohno, N., Yamakawa, H., Ohara, O., Liao, X., Baba, T., and Ohno, S. (2005). Immunohistochemical study of a membrane skeletal molecule, protein 4.1G, in mouse seminiferous tubules. Histochem. Cell Biol. 124, 303–311. Thompson, H. M., and McNiven, M. A. (2001). Dynamin: Switch or pinchase? Curr. Biol. 11, R850. Toyama, Y., Maekawa, M., and Yuasa, S. (2003). Ectoplasmic specializations in the Sertoli cell: New vistas based on genetic defects and testicular toxicology. Anat. Sci. Int. 78, 1–16. Tsai, M. J., and O’Malley, B. W. (1994). Molecular mechanisms of action of steroid/thyroid receptor superfamily members. Annu. Rev. Biochem. 63, 451–486. Tsai, M. Y., Yeh, S. D., Wang, R. S., Yeh, S., Zhang, C., Lin, H. Y., Tzeng, C. R., and Chang, C. (2006). DiVerential eVects of spermatogenesis and fertility in mice lacking androgen receptor in individual testis cells. Proc. Natl. Acad. Sci. USA 103, 18975–18980. Tung, P. S., and Fritz, I. B. (1993). Interactions of Sertoli cells with laminin are essential to maintain integrity of the cytoskeleton and barrier functions of cells in culture in the two‐chambered assembly. J. Cell. Physiol. 156, 1–11. Turner, T. T., Jones, C. E., Howards, S. S., Ewing, L. L., Zegeye, B., and Gunsalus, G. L. (1984). On the androgen microenvironment of maturing spermatozoa. Endocrinology 115, 1925–1932. Udayakumar, T. S., Chen, M. L., Bair, E. L., von Bredow, D. C., Cress, A. E., Nagle, R. B., and Bowden, G. T. (2003). Membrane type‐1‐matrix metalloproteinase expressed by prostate carcinoma cell cleaves human laminin‐5 3 chain and induces cell migration. Cancer Res. 63, 2292–2299. Utani, A., Nomizu, M., Matsuura, H., Kato, K., Kobayahsi, T., Takeda, U., Aota, S., Nielsen, P. K., and Shinkai, H. (2001). A unique sequence of laminin 3G domain binds to heparin and promotes cell adhesion through syndecan‐2 and ‐4. J. Biol. Chem. 276, 28779–28788. Utech, M., Ivanov, A. I., Samarin, S. N., Bruewer, M., Turner, J. R., Mrsny, R. J., Parkos, C. A., and Nusrat, A. (2005). Mechanism of IFN‐gamma‐induced endocytosis of tight junction proteins: Myosin II‐dependent vacuolarization of the apical plasma membrane. Mol. Biol. Cell. 16, 5040–5052. Vaid, K. S., Guttman, J. A., Babyak, N., Deng, W., McNiven, M. A., Mochizuki, N., Finlay, B. B., and Vogl, A. W. (2007). The role of dynamin 3 in the testis. J. Cell. Physiol. 210, 644–654. Vogl, A. W. (1989). Distribution and function of organized concentrations of actin filaments in mammalian spermatogenic cells and Sertoli cells. Int. Rev. Cytol. 119, 1–56. Vogl, A. W., PfeiVer, D. C., Mulholland, D. J., Kimel, G., and Guttman, J. (2000). Unique and multifunctional adhesion junctions in the testis: Ectoplasmic specializations. Arch. Histol. Cytol. 63, 1–15. Vogl, A. W., Vaid, K. S., and Guttman, J. A. (2007). The Sertoli cell cytoskeleton. In: ‘‘Molecular Mechanisms in Spermatogenesis’’ (C. Y. Cheng, Ed.). Landes Bioscience, Austin, TX (in press). Waalkes, M. P., Wahba, Z. Z., and Rodriguez, R. E. (1991). ‘‘Cadmium.’’ Williams and Wilkins, Baltimore. Walker, W. H. (2003). Nongenomic actions of androgen in Sertoli cells. Curr. Top. Dev. Biol. 56, 25–53. Wang, R. S., Yeh, S., Chen, L., Lin, H., Zhang, C., Ni, J., Wu, C., di Sant’ Agnese, P. A., deMesy‐Bentley, K. L., Tzeng, C., and Chang, C. (2006). Androgen receptor in Sertoli cell is essential for germ cell nursery and junctional complex formation in mouse testes. Endocrinology 147, 5624–5633. Welshons, W. V., Nagel, S. C., and vom Saal, F. S. (2006). Large eVects from small exposures. III. Endocrine mechanisms mediating eVects of bisphenol A at levels of human exposure. Endocrinology 147, S56–S69. Willem, M., Miosge, N., Halfter, W., Smyth, N., Jannetti, I., Burghart, E., Timpl, R., and Mayer, U. (2002). Specific ablation of the nidogen‐binding site in the laminin 1 chain interferes with kidney and lung development. Development 129, 2711–2722.

92

Yan et al.

Wine, R. N., and Chapin, R. E. (1999). Adhesion and signaling proteins spatiotemporally associated with spermiation in the rat. J. Androl. 20, 198–213. Wong, C. H., and Cheng, C. Y. (2005). The blood‐testis barrier: Its biology, regulation and physiological role in spermatogenesis. Curr. Top. Dev. Biol. 71, 263–296. Wong, C. H., Mruk, D. D., Lui, W. Y., and Cheng, C. Y. (2004). Regulation of blood‐testis barrier dynamics: An in vivo study. J. Cell Sci. 117, 783–798. Wong, C. H., Mruk, D. D., Siu, M. K. Y., and Cheng, C. Y. (2005a). Blood‐testis barrier dynamics are regulated by 2‐macroglobulin via the c‐Jun N‐terminal protein kinase pathway. Endocrinology 146, 1893–1908. Wong, C. H., Xia, W., Lee, N. P. Y., Mruk, D. D., Lee, W. M., and Cheng, C. Y. (2005b). Regulation of ectoplasmic specialization dynamics in the seminiferous epithelium by focal adhesion‐associated proteins in testosterone‐suppressed rat testes. Endocrinology 146, 1192–1204. Wrobel, K. H., Mademann, R., and Sinowatz, F. (1979). The lamina propria of the bovine seminiferous tubule. Cell Tissue Res. 202, 357–377. Xia, W., and Cheng, C. Y. (2005). TGF‐ 3 regulates anchoring junction dynamics in the seminiferous epithelium of the rat testis via the Ras/ERK signaling pathway: An in vivo study. Dev. Biol. 280, 321–343. Yamada, S., Pokutta, S., Drees, F., Weis, W. I., and Nelson, W. J. (2005). Deconstructing the cadherin‐catenin‐actin complex. Cell 123, 889–901. Yan, H. H. N., and Cheng, C. Y. (2005). Blood‐testis barrier dynamics are regulated by an engagement/disengagement mechanism between tight and adherens junctions via peripheral adaptors. Proc. Natl. Acad. Sci. USA 102, 11722–11727. Yan, H. H. N., and Cheng, C. Y. (2006). Laminin 3 forms a complex with 3 and 3 chains that serves as the ligand for 6 1‐integrin at the apical ectoplasmic specialization in adult rat testes. J. Biol. Chem. 281, 17286–17303. Yan, H. H. N., Mruk, D. D., Lee, W. M., and Cheng, C. Y. (2007). Ectoplasmic specialization: A friend or a foe of spermatogenesis? BioEssays 29, 36–48. Yeh, S., Tsai, M. Y., Xu, Q., Mu, X. M., Lardy, H., Huang, K. E., Lin, H., Yeh, S. D., Altuwaijri, S., Zhou, X., Xing, L., Boyce, B. F., et al. (2002). Generation and characterization of androgen receptor knockout (ARKO) mice: An in vivo model for the study of androgen functions in selective tissues. Proc. Natl. Acad. Sci. USA 99, 13498–13503. Yin, T., and Green, K. J. (2004). Regulation of desmosome assembly and adhesion. Semin. Cell Dev. Biol. 15, 665–677. Zhang, J., Wong, C. H., Xia, W., Mruk, D. D., Lee, N. P. Y., Lee, W. M., and Cheng, C. Y. (2005). Regulation of Sertoli–germ cell adherens junction dynamics via changes in protein– protein interactions of the N‐cadherin‐ ‐catenin complex which are possibly mediated by c‐Src and myotubularin‐related protein 2: An in vivo study using an androgen suppression model. Endocrinology 146, 1268–1284. Zhang, C., Yeh, S., Chen, Y., Wu, C., Chuang, K., Lin, H., Wang, R., Chang, Y., Mendis‐ Handagama, C., Hu, L., Lardy, H., and Chang, C. (2006). Oligozoospermia with normal fertility in male mice lacking the androgen receptor in testis peritubular myoid cells. Proc. Natl. Acad. Sci. USA 103, 17718–17723. Zirkin, B. R., Santulli, R., Awoniyi, C. A., and Ewing, L. L. (1989). Maintenance of advanced spermatogenic cells in the adult rat testis: Quantitative relationship to testosterone concentration within the testis. Endocrinology 124, 3043–3049. Zwain, I. H., Grima, J., Stahler, M. S., Saso, L., Cailleau, J., Verhoeven, G., Bardin, C. W., and Cheng, C. Y. (1993). Regulation of Sertoli cell 2‐macroglobulin and clusterin (SGP‐2) secretion by peritubular myoid cell. Biol. Reprod. 48, 180–187.

3

____________________________________________________________________________

Substrates of the Methionine Sulfoxide Reductase System and Their Physiological Relevance Derek B. Oien and Jackob Moskovitz Department of Pharmacology & Toxicology, School of Pharmacy University of Kansas, Lawrence, Kansas 66045

I. Introduction A. Msr System II. Regulated Substrates A. Alpha1‐Antitrypsin B. Calmodulin C. High‐Density Lipoprotein D. Inhibitor of Kappa B‐Alpha E. Potassium Channels F. Thrombomodulin G. Tissue Plasminogen Activator III. Scavenging Substrates A. Blood Clotting Cascade/Fibrinolysis B. Cytokines C. Enzymes D. Heat Shock Proteins E. Hormones F. Mucus Protease Inhibitor IV. Modified Substrates with ‘‘Damaged’’ EVects A. Enzymes B. Heme Proteins C. Hormones D. Neurodegenerative Disease Associated E. Serine Protease Inhibitors (Serpins) F. Snake Venom Toxins G. Miscellaneous Substrates V. Discussion References

Posttranslational modifications can change a protein’s structure, function, and solubility. One specific modification caused by reactive oxygen species is the oxidation of the sulfur atom in the methionine (Met) side chain. This modified amino acid is denoted as methionine sulfoxide (MetO). MetOs in proteins are of considerable interest as they are involved in early posttranslational modification events. Thus, various organisms produce specific enzymes that Current Topics in Developmental Biology, Vol. 80 Copyright 2008, Elsevier Inc. All rights reserved.

93

0070-2153/08 $35.00 DOI: 10.1016/S0070-2153(07)80003-2

94

Oien and Moskovitz

can reverse these modifications. MetO reductases, known collectively as the methionine sulfoxide reductase (Msr) system, are the only known enzymes that can reduce MetOs. The current research field of Met redox cycles is consumed with elucidating its role in regulation, redox homeostasis, prevention of irreversible modifications, pathogenesis, and the aging process. Substrates of the Msr system can be loosely classified by the overall eVect of the MetO on the protein. Regulated substrates utilize Met as a molecular switch to modulate activation; scavenging substrates use Mets to detoxify oxidants and protect important regions of the protein; and modified substrates are altered by Met oxidation resulting in various changes in their properties, including function, activity, structure, and degradation resistance. ß 2008, Elsevier Inc.

I. Introduction In 1956, Denham Harman published a theory on free radical damage to organisms and its relation to aging (Harman, 1956). This landmark article has evolved to what is now known as the free radical theory of aging (Harman, 1973). The free radical theory of aging is in accordance with the common hypothesis that aging is a process mediated by oxidative damage, and the controversial idea that age‐related disease is linked to oxidative damage. Oxidative damage to cells can aVect multiple cellular components, including lipids and proteins. In general, a chemical modification forming after the genetic expression of a protein is denoted as posttranslational modification. Oxidative modifications to proteins are caused by reactive species, such as reactive oxygen species (ROS) and reactive nitrogen intermediates. These oxidants are formed by a variety of biological processes within cells and tissues at varying rates, and specific toxins can induce their production. Examples include hydrogen peroxide, superoxide, oxygen, ozone, hypochlorous acid, chloramine T (sodium N‐chloro‐p‐toluenesulfonchloramide), N‐chlorosuccinimide, hydroxyl radicals, and peroxynitrite (Savige and Fontana, 1977). Modifications via ROS can aVect virtually all amino acid side chains and the backbone of the peptide linkage. However, amino acids vary in their vulnerability to become oxidized. Furthermore, the particular location of amino acid residues within the structure of a protein may determine the sensitivity level toward the oxidants, which contributes to even greater variability. Common ROS‐mediated side chain modifications can be in the form of a carbonyl or the oxidized sulfur atom of methionine (Met) or cysteine (Cys). Protein‐carbonyls are nonenzymatic irreversible modifications. They are often catalyzed by metal ions and tend to form on the side chains of the amino acids proline, arginine, threonine, and lysine (Dalle‐Donne et al., 2006). In contrast, the sulfur‐containing amino acids are readily oxidized and most modifications are reversible. Modifications of Cys residues have various consequences including the alteration of protein structure, inactivation of

3. Substrates of Methionine Sulfoxide Reductase

95

proteins, and formation of disulfide bridges (Jacob et al., 2006). Reversal of the posttranslational modifications to Cys residues can be achieved by several nonenzymatic reductants and reducing enzymes. However, the reduction of methionine sulfoxide (MetO) is exclusively performed by the methionine sulfoxide reductase (Msr) system. A. Msr System The oxidation of a Met side chain can form either a MetO or a Met sulfone. The former contains an oxygen atom double‐bonded to the sulfur atom, and the latter contains two oxygen atoms bonded to the sulfur atom. Met sulfones are formed in more extreme oxidant conditions following the formation of MetO. They are considered irreversible and cannot be reduced under physiological conditions. In contrast, MetO is readily formed relative to other posttranslational modifications and can be reduced by Msr. Oxidation of the Met sulfur atom can form either one of two possible enantiomers, denoted as S and R. With some specific exceptions, there is no general evidence to date predicting which enantiomer of MetO will be formed when the side chain of Met is oxidized. The S and R enantiomers can be reduced by the enzymes MsrA and MsrB, respectively. Msrs reduce MetO at the expense of becoming oxidized, and thus inactivate themselves. The first evidence of a reductase that reduced MetO in peptides was shown in 1981 (Brot et al., 1981). The enzymes are reduced and reactivated primarily by thioredoxin (Trx), thioredoxin reductase (Trr), and NAD(P)H. The reduction sequence can be illustrated by the equation: NADPH þ Trr ! Trx ! Msr ! MetðMethyl  SO to  Methyl  SÞ ð1Þ

where the arrows depict reduction of each oxidized form of the molecule by the former substance. After the reducing enzymes bind to their substrate, they become oxidized and thus reenter the reduction cycle. In some forms of human MsrB, specifically hMsrB2 and hMsrB3, Trx has been found less eYcient for reduction. In addition, a protein called thionein, especially for hMsrB3 (Sagher et al., 2006b), and certain selenium compounds make better reducing enzymes to activate these reductases (Sagher et al., 2006a). In experimental conditions, dithiothreitol is often used to reduce Msr, and lipoic acid also can act as a reductant (Biewenga et al., 1998). The Cys at position 72 is vital for Msr activity and is highly conserved in most species (Moskovitz et al., 2000). When Msr is inactivated, the Cys residue is oxidized and forms a disulfide bridge with another Cys near the C‐terminus (Brot and Weissbach, 2000). The Msr enzymes are found in many organisms in nature. In a bacterial strain like Escherichia coli, there is one form of MsrA and one form of MsrB. However, in mammals, there is one form of MsrA and four forms of MsrB:

96

Oien and Moskovitz

denoted MsrB1, MsrB2, MsrB3A, and MsrB3B. The MsrB1 form is a cytoplasmic selenoprotein (Bar‐Noy and Moskovitz, 2002; Moskovitz et al., 2002), MsrB2 and MsrB3B can be targeted to the mitochondria, and MsrB3A is targeted to the endoplasmic reticulum (Kim and Gladyshev, 2004). In addition, MsrA also has a long spliced variant form that is transported to the mitochondria (Balog et al., 2003). The msr genes are ubiquitously expressed in many organisms, and highly expressed in the liver and kidneys of mammals (Moskovitz et al., 1996). The Msr system is considered to be important for survival under oxidative stress conditions as manifested by msrA gene knockouts in various organisms (Moskovitz, 2005a). Oxidation of Met residues can be readily reduced by the Msr system. The latter may prevent changes resulting from Met oxidation, for example, alterations in the properties of methyl sulfoxide‐containing peptides, protein structure, biological function, or a combination of all possibilities. Met is a hydrophobic amino acid, and the hydrophobicity decreases when the sulfur atom is oxidized. However, the sulfoxide formation can alter the native folding and create a more hydrophobic protein (Chao et al., 1997). When Met is present in a protein, it can scavenge/detoxify the ROS, and this may occur with impunity if Msr is present. Thus, it is further speculated that Mets play a protective role from ROS by their placement on outer faces of certain proteins. In this situation, they may serve as a shield from oxidative modification, protecting the active site of the protein. Furthermore, lack of MsrA has been shown in some organisms to increase protein carbonyl accumulation, an irreversible form of protein oxidation (Moskovitz et al., 2001). There is also evidence that Msr levels decrease in various rat tissues as they age (Petropoulos et al., 2001), supporting the link between enhanced free radical damaging action and the aging process (Stadtman et al., 2002). The role of MetO in proteins and the subsequent reduction in the pathogenesis of disease and general aging of cells is not completely known. Hence, the Msr system may prove to be a key component of oxidative prevention, regulation, or a combination of both. The extent of the consequences of these posttranslational modifications and their complete physiological significance still require more research.

II. Regulated Substrates Cell signaling is an important biological mechanism and a critical component of cellular regulation. Signaling pathways consist of proteins that can be activated and inactivated to modulate a response. Activity of the protein is often determined by its structure, and overall by the accessibility of reactive site. Modifications that add to cell signaling are covalently bonded and reversible. Well‐known examples of this phenomenon include tyrosine

3. Substrates of Methionine Sulfoxide Reductase

97

and serine phosphorylation/dephosphorylation and Cys redox modulation (Ciorba et al., 1997). It is suggested that the redox properties of Met residues also can contribute to signaling pathways. Msrs are generally regarded as repair enzymes, but the following examples show evidence of the enzymes contributing to cellular regulation.

A. Alpha1‐Antitrypsin Alpha1‐antitrypsin (1‐AT) is a serine protease inhibitor (serpin) that inhibits the protease activity of trypsin. It is mainly produced by the liver and is associated with protection of the lungs by preserving antineutrophil elastase activity. Neutrophil elastase has destructive consequences in the alveolar matrix and is associated with the risk of developing emphysema. Oxidants can inactivate the serpin by creating MetOs, and addition of Msr reduces the modifications in vitro (Abrams et al., 1981). The serpin has surface exposed Mets that form sulfoxides when exposed to hydrogen peroxide, a component of cigarette smoke (Taggart et al., 2000). Oxidation of Met351 or Met358 in the serpin causes a loss of the antineutrophil elastase activity associated with emphysema. The antielastase activity can be partly restored with the addition of MsrA in vitro, but contrasting evidence is found in rats when oxidized inhibitor is injected (Glaser et al., 1987). Cultured human monocytes stimulated with oxidized 1‐AT increase cytokine expression and ROS production and decrease low‐density lipoprotein (LDL) actions (Moraga and Janciauskiene, 2000). These results suggest the serpin may be part of a positive‐feedback immune defense against pathogens. Specifically, the cytokine expression was noted in interleukin‐6 (IL‐6) and tumor necrosis factor‐, both proinflammatory cytokines. Production of ROS often accompanies an immune response, where it is thought to be a defense against pathogens. In addition, oxidized LDL can mediate monocyte migration. Despite these findings, there is lack of suYcient in vivo evidence to support this theory.

B. Calmodulin Calmodulin (CaM) is a calcium‐binding protein known to react with many diVerent proteins, and it serves as a regulatory protein to interpret calcium signals. Various calcium pumps and channels use CaM as a sensor of intracellular calcium concentration. Calcium modulates many intracellular processes including metabolism, which in turn produce ROS in the mitochondria (examples include superoxide, hydrogen peroxide, and peroxynitrite) (Bigelow and Squier, 2005). There are nine Met residues in CaM. CaM activity on plasma membrane calcium‐ATPase decreases with oxidized Mets in older rats

98

Oien and Moskovitz

(Michaelis et al., 1996). Approximately six Mets form sulfoxides, and no other amino acids are modified. While oxidation of seven of them has minimal eVects, oxidation of Met144 and Met145 results in a 50–60% reduction in plasma membrane calcium‐ATPase activation (Bartlett et al., 2003). The critical Mets may serve as redox sensors to modulate calcium concentrations. However, they may also contribute to higher calcium concentrations in the cytosol during conditions of oxidative stress. Moreover, oxidized CaM binds to a site on the calcium‐ATPase near the binding site of unoxidized CaM (Gao et al., 2001). Thus, the distinct binding site of oxidized CaM in a stabilized inhibitory state suggests a modulation role by MetO formation. Furthermore, function can be restored in oxidized CaM by incubation with MsrA in vitro (Sun et al., 1999). Thus, MsrA (and probably MsrB) may add another degree to calcium regulation in vivo. The Mets may act as a switch in CaM signaling, causing a stable inhibited state when oxidized and reverting to normal activity when reduced by the Msr system (Bigelow and Squier, 2005).

C. High‐Density Lipoprotein The high‐density lipoproteins (HDLs) are the smallest in size of the lipoproteins and contain the A class of apolipoproteins. Apolipoprotein‐AI (apo‐ AI) and apolipoprotein‐AII (apo‐AII) are the most abundant, and the other A class apolipoproteins are only present in relatively small amounts (Brouillette et al., 2001). They are mostly known for their binding and removal of cholesterol from arteries and subsequent transport of the cholesterol to the liver, also termed reverse cholesterol transport (Fielding and Fielding, 1995). Not all HDLs perform this function, but the ones that bind to cholesterol tend to be the largest. Furthermore, oxidized HDL has a decreased ability to transport cholesterol to the liver (Panzenbock and Stocker, 2005). 1. Apolipoprotein‐AI Apo‐AI is the most abundant of the A class apolipoproteins and contains three Met residues at the positions 86, 112, and 148 (Panzenbock and Stocker, 2005). The Met residues are located in the lipid‐binding domains of the protein. Met oxidation alters the secondary structure and lipid aYnity of the protein (Anantharamaiah et al., 1988). A Met residue at position 86 or 112 is readily oxidized, and Met148 is oxidized after the two other Met residues form sulfoxides (Panzenbock and Stocker, 2005). Furthermore, studies of plasma HDLs in healthy human patients showed no modifications (Sattler et al., 1994), while patients diagnosed with coronary artery disease contained nitrotyrosine residues, an irreversible modification that is thought

3. Substrates of Methionine Sulfoxide Reductase

99

to follow Met oxidation (the presence of MetO was not analyzed in these studies) (Pennathur et al., 2004; Zheng et al., 2004). The addition of MsrA can restore structure and lipid association in vitro (Sigalov and Stern, 1998). In theory, MsrB with MsrA could reduce all of the sulfoxides in vitro and in vivo. However, it is still unclear if and how the Msr system reduces oxidized HDL in vivo when the oxidation occurs extracellularly. Further studies are required to determine the role of the Msr system in the prevention of the apolipoprotein oxidation.

2. Apolipoprotein‐AII Apo‐AII is a dimer and contains one Met amino acid at position 26 of each monomer, and thus has a total of two Mets (Panzenbock and Stocker, 2005). Similar to apo‐AI, the Met residues are located in the lipid‐binding domains of the protein. Apo‐AII contains only one oxidized Met residue under relatively low oxidative stress conditions in vitro (Anantharamaiah and Garber, 1996). Similarly to the case of apo‐AI, Msr is suggested to reduce the MetOs, but it has not been demonstrated in apo‐AII.

3. Met Oxidation in Atherosclerosis Met oxidation in atherosclerosis may act as a local antioxidant (in the presence of the Msr system), preventing further modification to apolipoproteins. Atherosclerosis is an arterial blood vessel disease thought to be caused by oxidized LDL accumulation. The accumulation in blood vessels attracts monocytes, which then diVerentiate into macrophages and form foam cells (Pennings et al., 2006). The foam cells presence leads to the formation of a fibrous plaque that thickens and hardens the vessel walls. Macrophages secrete ROS, such as hydrogen peroxide and superoxide, and also myeloperoxidase (an enzyme that catalyzes the formation of hypochlorous acid from hydrogen peroxide and chloride). The produced ROS is part of an immune response against antigens, additionally causing inflammation. HDL, including apo‐AI, may play an antioxidant role of regulating inflammation and preventing atherosclerosis (Shao et al., 2006). Myeloperoxidase can cause chlorination of tyrosine residues, but Tyr115 is resistant to chlorination by the neighboring Met112 residue that is more readily oxidized. In contrast, Tyr192 can be chlorinated by myeloperoxidase, diminishing apo‐AI activity. However, it is unknown if the oxidation is part of a direct reaction from myeloperoxidase, or a reaction with an intermediate such as hydrogen peroxide. In the same study, application of MsrA and MsrB restores most of the activity to promote cholesterol eZux in vitro, demonstrating a potential regulation role in preventing atherosclerosis.

100

Oien and Moskovitz

D. Inhibitor of Kappa B‐Alpha The protein termed inhibitor of kappa B‐alpha (IB) is named for its inhibition activity of nuclear factor B. Nuclear factor B is a transcription factor known to regulate expression of many immune system genes in response to various stress‐related stimuli. If the Met at position 45 in the IB is modified to a sulfoxide, the protein is resistant to degradation (Kanayama et al., 2002). When the inhibitor is oxidized by taurine chloride, the modified inhibitor cannot dissociate from the nuclear factor. The nuclear factor is prevented from translocating to the nucleus, and the inhibition may limit inflammation. Furthermore, both proteins remain in a cytoplasmic complex until the cell is stimulated and the inhibitor protein is degraded by the proteosome (Midwinter et al., 2006). The inhibitor protein can be oxidized by glycine chloramine, which causes a mobility shift when electrophoresed on a polyacrylamide gel. The shift is reversed with the addition of Msrs, both MsrA and MsrB. This provides evidence that (1) glycine chloramine causes Met oxidation to form in a selective manner, specifically at position 45, and (2) the MetO causes a change in the proteins conformation. Moreover, there is no eVect with dithiothreitol treatment, indicating that there are no significant conformational changes from Cys modifications. It is unclear what impedes the protein degradation, but the evidence suggests that the Msr system can prevent the inhibition. E. Potassium Channels Potassium channels regulate many cellular processes and are found in most cells. They have important roles in neuronal action potentials and hormone secretion. Potassium channels are classified into four main families. The families are voltage‐gated, calcium‐activated, inward‐rectifier, and tandem‐ pore‐domain potassium channels. Voltage‐gated channels respond to the membrane potential of the cell (Trauner and Kramer, 2004). Calcium‐activated channels are activated by the concentration of intracellular calcium. Inward‐ rectifier channels are mostly unidirectional, allowing more potassium to move into the cell than out (Butt and Kalsi, 2006). Tandem‐pore‐domain channels, also known as leak channels, respond to various mechanisms in an electrodiVusion process (Lesage and Lazdunski, 2000). Thus far, evidence in regulation via Met redox cycles has only been shown in voltage‐gated and calcium‐activated potassium channels. 1. Voltage‐Gated Potassium Channel Hoshi and colleagues have demonstrated in vivo evidence of protein regulation via Met oxidation and reduction by MsrA (Ciorba et al., 1997). They expressed spliced variant genes of Shaker channels in Xenopus oocytes.

3. Substrates of Methionine Sulfoxide Reductase

101

Shaker channels are a member of the voltage‐gated potassium channel (Kv) family and are essential to cellular potassium regulation. The channel variants, ShC/B and ShB, mainly diVered in an additional Met residue at position 3 in ShC/B. Oxidants (such as chloramine T) caused MetO formation on Met3 of ShC/B. This caused a slowing of the N‐type inactivation process, a process of rapid inactivation by blocking the intracellular pore of the channel. Coexpression of the ShC/B under oxidative stress conditions with MsrA results in an acceleration of the inactivation time course. The eYcient manner in which these channels can alter their redox state provides evidence for regulation via Met residues. A later study shows that nitric oxide (a weak oxidant important in blood vessel regulation) can cause the same Met oxidation and can slow the ShC/B inactivation time course (Ciorba et al., 1999). In a similar study, the N‐type inactivation is eliminated, allowing the Shaker channel to independently display the less stable P‐type and relatively stable C‐type inactivation (Chen et al., 2000). In the segment regulating P/C‐ type inactivation, Met466 can modulate the inactivation time course when oxidized. High oxygen levels can cause this oxidation. Moreover, the channel is less sensitive to oxygen when Met466 is mutated to a leucine. The mutation of Met to leucine is not considered to create significant changes in the protein in comparison to the native form of the channel. The evidence suggests that oxygen levels may aVect the cellular excitability by oxidation/reduction of Met466.

2. Calcium‐Activated Potassium Channel Calcium‐activated potassium channels are voltage‐dependent and regulated by intracellular calcium concentrations. The Slo channel is calcium activated and has a large single‐channel opening that contributes to smooth muscle tone and neuronal excitability. The human Slo channel is suggested to be modulated by Cys and Met oxidation when expressed in mammalian cells (Tang et al., 2001). Under very low calcium concentration, the addition of chloramine T increases the channel current. The inactivation is partially restored by MsrA and dithiothreitol, but not without MsrA. This suggests that MetOs are a key element for inhibiting inactivation of the Slo channel via oxidants. The source of ROS is unknown and may give more evidence of in vivo regulation. Moreover, Met residues of the Slo channel at position 536, 712, and 739 are sensitive to oxidants (Santarelli et al., 2006). The sensitivity is diminished when the residues are mutated to leucines. However, the channel is still receptive to calcium concentration. At low intracellular calcium concentrations, the channel remains open and allows potassium to eZux into the cell. In contrast, the oxidized Mets contribute to less of an eVect at

102

Oien and Moskovitz

high intracellular calcium concentrations. Thus, the Met redox state may be important for cellular regulation in Slo channels.

F. Thrombomodulin Thrombomodulin (TM) is an integral membrane protein expressed on the surface of endothelial cells. It functions as a cofactor in the thrombin‐ induced activation of protein C in the anticoagulant pathway by forming a complex with the coagulation protein thrombin (factor II) (Esmon, 1989). Protein C is an anticoagulant that is activated by thrombin to degrade clotting factors such as factor Va and factor VIIIa. There are six epidermal growth factor‐like domains in TM, and the fourth and fifth domains contain the smallest active fragment (White et al., 1995). For TM to function as a cofactor, the fourth and fifth epidermal growth factor‐like domains must work together, and they are bound by a short sequence containing a single Met, Met388 (Fuentes‐Prior et al., 2000). Oxidation of Met388 by hydrogen peroxide or chloramine T inhibits cofactor activity by 90% in cultured cells, and thus overall anticoagulant activity (Glaser et al., 1992). Only substitution of a leucine results in impunity. The latter result not only demonstrates the significance of the Met residue but additionally suggests a role of Met being evolutionarily conserved as a regulator (Wood et al., 2005). In theory, oxidants produced by the immune system prevent bleeding by inhibition of TM anticoagulant activity. It is speculated that the decrease in hydrophobicity creates a situation where Met388 does not interact with the fifth domain, which may weaken the TM–thrombin interaction (Wood et al., 2003, 2005). There is no evidence of Msr system reduction, which could theoretically reactivate TM. Future in vivo evidence is predicted to demonstrate the Msr system involved in restoring the coagulant activity of TM, and thus add an additional component of coagulant regulation.

G. Tissue Plasminogen Activator Tissue plasminogen activator (t‐PA) is a serine protease that converts plasminogen to plasmin. Plasmin is a serine protease known to degrade blood plasma proteins in intravascular fibrinolysis. t‐PA can be secreted by macrophages and stimulated by fibrinogen degradation products. It contains five Met residues. The addition of chloramine T has no significant aVect on plasminogen activation in vitro (Stief et al., 1991). However, the aYnity for fibrinogen degradation products is significantly impaired. The Met residue at position 207 (that resides in the known binding region of these products) is suggested to mediate the binding with fibrinogen degradation products.

3. Substrates of Methionine Sulfoxide Reductase

103

Thus, oxidants produced by macrophages might regulate t‐PA activity and selectivity for fibrin.

III. Scavenging Substrates Scavenging substrates do not display any significant loss of biological activity when Met are oxidized under physiological conditions. This section is dedicated to substrates containing Mets that are hypothesized to play a protective role against the oxidation of proteins, or other residues of the same protein. Thus, by Met cyclic oxidation and reduction via the Msr system, these substrates are considered to be endogenous antioxidants, or oxidant scavengers. Theoretically, Met residues can eVectively scavenge oxidants before reactive species modify residues critical to the structure and function of the substrate (Levine et al., 1996). In this perspective, the Met residues act as ‘‘shields’’ to the active site or ‘‘molecular sinks’’ for reactive species balance. Evidence to support this idea is found in the placement of Mets. They are often described as surface‐exposed, protecting an opening to the active site, flanking the active site, or a combination. A. Blood Clotting Cascade/Fibrinolysis Blood clots, also known as fibrin clots, are the end products of the vascular injury‐activated coagulation cascade. The final step in the cascade is the creation of thrombin, a serine protease that cleaves fibrinogen into fibrin (Gorlach, 2005). Insoluble strands of fibrin form the clot, creating platelet stabilization and formation of a barrier at the site of injury. Fibrinolysis is the process of breaking down blood clots. Plasmin is a serine protease that cleaves the fibrin protein meshwork into fibrin degradation products. Protease inhibitors protect the fibrin clot from degradation. In addition, clotting factor‐ producing phagocytes can produce reactive species, which may create a need for the presence of oxidant scavengers in fibrin clotting and fibrinolysis (Stief and Fareed, 2000). 1. Alpha‐2‐Macroglobulin Alpha‐2‐macroglobulin (2M) is a broad spectrum protease inhibitor that is often present at the sites of inflammation and known to inhibit plasmin activity (Zorin et al., 2006). It is tetrameric and forms a covalent complex entrapping a broad range of proteases. The inhibitor is found at higher levels in tissue fluids than plasma levels (Swaim and Pizzo, 1988). Thus, the broad spectrum eVects are not limited to fibrinolysis. The antiprotease activity of the protein remains unaVected unless there is a further oxidation of a single

104

Oien and Moskovitz

tryptophan residue (Reddy et al., 1994). This causes the tetramer to dissociate into dysfunctional dimers. 2M has surface‐exposed Met that serve as an antioxidant region that can scavenge oxidant produced in the inflammation process (Levine et al., 2000). The further oxidation of the tryptophan residue is delayed and may be prevented all together by possible concurrent reduction from Msr both in vitro and in vivo. 2. Serine Protease Inhibitors (Serpins) Serpins are structurally related serine protease inhibitors and use conformational change to inhibit serine proteases (Law et al., 2006). They regulate proteolytic cascades, including coagulation. There are over 500 known serpins, and the majority of them specifically inhibit one serine protease. Usually, when a disease is linked to altered serpin function, it is often described as a serpinopathy. a. Alpha‐2‐Antiplasmin. Alpha‐2‐antiplasmin (2‐PI) is known to inactivate plasmin, an enzyme that degrades blood plasma protein, such as fibrin clots. 2‐PI contains 10 Met residues, 1 near its reactive center. Two of the outer Mets are oxidized with relatively high concentrations of chloramine T or hydrogen peroxide, similar to the concentration of a neutrophil or macrophage environment (Stief et al., 1988). However, no loss of binding activity to the relevant substrates plasmin, trypsin, or chymotrypsin was observed in vitro. b. Antithrombin III (AT III). Antithrombin III (AT III) is known to inactivate many coagulation factors, including factor II (thrombin). There are no Mets located in the reactive center of AT III. Under the same oxidative conditions applied to 2‐PI (Stief et al., 1988), AT III was shown to be resistant to inactivation when Mets formed sulfoxides. A later study reported supporting evidence that oxidation of Met314 and Met315 does not aVect thrombin‐inhibitory activity (Van Patten et al., 1999). The proximity of these Met residues suggests a role in preventing further modifications to other residues by scavenging ROS mediated by the Msr system.

B. Cytokines Cytokines are intercellular signaling compounds that bind to specific receptors on the cell surface. They produce a variety of eVects, most notably in the immune system and hematopoiesis. The scavenging properties of cytokine Met residues may be to protect critical regions of the protein in potentially harsh microenvironments or to maintain redox homeostasis.

3. Substrates of Methionine Sulfoxide Reductase

105

1. Interferon Alpha‐2b Interferon alpha‐2b (IFN‐2b) is expressed by many cells and is known to stimulate macrophages and natural killer cells in an immune response. Recombinant IFN‐2b is used to treat chronic hepatitis C and is also used as an antineoplastic agent. In a storage solution, the drug was reported to contain traces of MetO on position 111 (Gitlin et al., 1996). This oxidation has no significant eVect on receptor binding in bioassays. 2. Interferon Gamma Interferon gamma (IFN‐ ) is secreted by natural killer cells and T lymphocytes. It has significant roles in viral and tumor defense and can regulate other immune responses. Recombinant IFN‐ is used as a treatment for chronic granulomatous disease. It contains five Met residues, and the two Mets at positions 1 and 135 can be oxidized with low levels of hydrogen peroxide (Keck, 1996). The oxidation of Met1 and Met135 does not significantly inhibit binding (measured by antigen production in cell cultures). 3. Interleukin‐6 Interleukin‐6 (IL‐6) is a cytokine produced mainly by macrophages, endothelial cells, and T lymphocytes. It has multiple eVects including the stimulation of antibody‐producing cells. Incubation with chloramine T can oxidize five Met residues and cause a decrease in aYnity for the IL‐6 receptor (Nishimura et al., 1991). The Met162 residue is more resistant to oxidation than three of the other Met residues. The oxidation of Met162 is relatively consistent with the loss of receptor binding. However, the three Met residues that are readily oxidized have less of an impact on receptor binding when oxidized. Thus, this phenomenon suggests that three Mets may serve as oxidant scavengers to protect other residues from modification, including Met162, which may be localized near the active site. 4. Stem Cell Factor Stem cell factor (SCF) is a dimeric glycoprotein that regulates hematopoietic progenitor cells development by binding to the c‐kit receptor, often in synergy with other cytokines. It also plays a key role in the development and function of mast cells, germ cells, and melanocytes. SCF contains five Mets. When four or five are oxidized in vitro, the dimer dissociation rate constant increases and corresponds to a loss of c‐kit receptor binding (Hsu et al., 1996). Under higher oxidation rates, the three Met residues have insignificant eVects when sulfoxides are formed. Among these Mets, two of

106

Oien and Moskovitz

the Mets were more resistant to oxidation and were in proximity to each other. This observation indicates that the three Met residues may serve as a shield, protecting the region from modification of other residues. The protected region probably is important for the biological activity of SCF. C. Enzymes 1. Glutamine Synthetase Glutamine synthetase (GS) is an enzyme that catalyzes the reaction of glutamate and ammonia to form glutamine. It was chosen by Stadtman and coworkers as a model to study the eVect of Met oxidation on its activity and scavenging properties (Levine et al., 1996). To examine MetO formation, the enzyme was exposed to incrementing concentrations of hydrogen peroxide. GS contains 16 Mets, of which eight are exposed on the surface. It was shown that eight surface‐exposed Mets could be oxidized, and no other modifications were present. The surface exposed Mets were located around the entrance to the active site of the enzyme. Interestingly, Met oxidation decreased magnesium‐dependent ‐glutamyl transferase activity, but seemingly did not aVect the manganese‐dependent activity. This eVect is known to happen in bacteria under physiological conditions, and Met oxidation may play a role in the actual mechanism. 2. 15‐Lipoxygenase Lipoxygenases catalyze the oxygenation of polyunsaturated fatty acids and hydroperoxy compounds formed by the oxygenation reaction. The 15‐lipoxygenase contains about 16 Met residues (Levine et al., 1996). One oxidized Met can inactivate the enzyme, and products of the enzymatic reaction are suYcient to cause this oxidation (Rapoport et al., 1984). This auto‐oxidation leads to self‐inactivation, which is a possible evidence of a regulatory mechanism. However, a later study showed that some of the Met residues are located near the active site (Gan et al., 1995), and oxidation of Met590 can inactivate binding to the linoleate substrate. Oxidants can inactivate the enzyme when Met590 is substituted with a leucine. Thus, the Met590 residue may be acting as an antioxidant to protect the substrate‐binding pocket, which could theoretically be reduced by Msr in vivo and restore binding capabilities. D. Heat Shock Proteins Heat shock proteins (Hsps) are found in all cells and experience increased expression when the host cell is exposed to an elevated temperature. Consequently, they are termed as Hsps. The same group of proteins is also known as

3. Substrates of Methionine Sulfoxide Reductase

107

stress proteins or chaperone proteins. They are referred to as stress proteins because their expression also increases in response to stress conditions other than elevated temperatures. In addition, they also assist in folding and confirmation of the proteins and can move proteins across membranes and around the cell (Kedzierska, 2005).

1. GroEL GroEL (GrL) is a prokaryotic, Met‐rich molecular chaperone protein similar to the eukaryotic Hsp60. It assists in folding of newly synthesized proteins by creating a protective cage. GrL is not aVected by endogenous metabolic by‐products like hydrogen peroxide and nitric oxide in E. coli (Khor et al., 2004). In contrast, it is inactivated by peroxynitrite and hypochlorous acid, which can be produced by phagocytes. Of the 23 Met residues in the protein, 12 are oxidized by hypochlorous acid (1 mM). The inactivation may be from other identified modifications including cysteic acid, a negligible amount of chlorotyrosine, and Met sulfones on Met111 and Met114. Furthermore, incubation with MsrA and MsrB restores over 60% of GrL activity. A complete recovery may not be possible because of irreversible modifications. In addition, Msr enzymes also bind to GrL in Helicobacter pylori (Alamuri and Maier, 2006).

2. Small Hsps Small Hsps (sHsps) usually form larger oligomeric complexes and are important for cell stress protection. The larger complexes have an increased ability to expose hydrophobic regions. These regions can then bind to similar surfaces of denatured proteins and prevent possible aggregation (Sundby et al., 2005). Hsp21 is an sHsps found in the chloroplasts of plants. Relatively low levels of hydrogen peroxide can oxidize six of the nine Mets in the conserved regions of the Arabidopsis thaliana Hsp21 protein, and two additional Mets are oxidized with a small increase in oxidant concentration (Gustavsson et al., 1999). The sulfoxide formation coincides with a significant conformational change. Also, conformational changes may cause the oligomer size to be reduced to tetramers of Hsp21, causing an observed complete loss of chaperone‐like activity (Harndahl et al., 2001). However, the oxidation‐ induced eVects of MetO‐containing Hsp21 are not apparent when the Mets are substituted with leucines (Sundby et al., 2005). Furthermore, addition of dithiothreitol and a plastidial form of Msr (PMSR4) to Hsp21 reduces Mets in chloroplasts (Gustavsson et al., 2002). The latter experiments suggest a scavenging role of Mets and the Msr system to maintain redox homeostasis in A. thaliana and protect the chaperone‐like properties of Hsp21, respectively.

108

Oien and Moskovitz

E. Hormones 1. Growth Hormone Growth hormone (GH) is a protein hormone secreted by the anterior pituitary gland and stimulates growth and cell reproduction. Human GH has three Mets. One report indicates a 25% loss of potency when two Mets are oxidized in human GH (Houghten et al., 1977). The loss of potency was determined by loss of growth‐promoting activity using the outdated rat tibia test. However, a later study demonstrated that Met170 is resistant to sulfoxide formation, and oxidation of Met14 and Met125 has no significant eVects on a lactogenic receptor bioassay in vitro (Teh et al., 1987). The reasons for this discrepancy are unknown, and may be diVerent results from discrete tests. Research on bovine and sheep GHs also reported no significant eVects when three out of four Mets formed sulfoxides (Glaser and Li, 1974). 2. Chorionic Somatomammotropin Chorionic somatomammotropin (CS) is a placental hormone that assists in metabolism regulation of pregnant women. It has a similar sequence to GH and glucagon. The human hormone contains six Mets, of which five can be oxidized by hydrogen peroxide (Houghten et al., 1977). This oxidation causes a significant loss of biological activity when assayed by mouse mammary glands in vitro. The Met170 residue is resistant to oxidation, similar to the Met170 resistance reported in human GH. The oxidation of Met64 and Met179 impairs binding in lactogenic receptor assays in vitro (Teh et al., 1987). These residues also have a higher oxidation rate constant when compared to the three other residues. The three Mets are readily oxidized, probably as scavengers protecting the active sites of the binding protein. F. Mucus Protease Inhibitor Mucus protease inhibitor (MPI) is produced by secretory cells and is resistant to acidic conditions. The myeloperoxidase enzyme and cigarette smoke components were reported to decrease the ability to inhibit leukocyte elastase activity after formation of a MetO (Carp and JanoV, 1980). Treatment of the inhibitor with N‐chlorosuccinimide can oxidize the Met at position 73 in vitro (Boudier and Bieth, 1994). It has four Met residues and Met73 is exposed to solvents and in the region that reacts with the elastase. Conversely, the latter experiment showed that oxidized MPI is still potent in neutrophil elastase inhibition. The authors suggest the discrepancies may be from a small neutrophil elastase concentration previously used. Furthermore, the inactivation of protease inhibitors that act on neutrophil elastase may lead the lung tissue destruction in diseases such as cystic fibrosis. When leucines

3. Substrates of Methionine Sulfoxide Reductase

109

are substituted for Mets in the inhibitor, elastase inhibition is not decreased (Rudolphus et al., 1991). This observation suggests a possible scavenging role of the Mets to protect other residues in a region that is critical for the inhibiting activity.

IV. Modified Substrates with ‘‘Damaged’’ Effects The proteins in this section display altered functions, a loss of biological activity, a change in structure, or display resistance to degradation after the oxidation of the Met residue(s). The substrates are often linked to respective diseases, where they are discovered in a ‘‘damaged’’ state. Evidence demonstrates that MetOs may contribute to this state. The role that MetOs may play in the physical and chemical properties of the protein is based solely on their availability to oxidation. Additionally, it is not always clear to what extent the substrate plays a role in the pathogenesis development of a respective disease, or rather if it is a consequence of a mature pathological state. The Msr system is proposed to play a role in maintaining the integrity of these substrates in their reduced form. The role of the Msr system may include repairing Met‐ oxidized proteins by preventing malfunction due to MetO, protecting specific regions of the protein from further irreversible modifications and structural changes, or a combination of both. The ultimate question is, if indeed MetOs are responsible for this functional alteration, what makes Msr insuYcient to counteract this phenomenon? Possibilities include, but are not limited to, the decline of Msr presence, decline of Msr activity, and noncombatable concentrations of reactive species. In some cases, Met oxidation may serve a modulation role, or simply be an ROS scavenger. Evidence may be either lacking in support of these roles or the actual functions of the proteins may still be unknown.

A. Enzymes 1. Antiflammin‐2 Antiflammin‐2 (AF2) is a 9‐amino acid peptide synthetically derived from the active center of lipocortin I. Lipocortin I is an enzyme known to inhibit the phospholipase A2 inflammation pathway (Moreno, 2000), and the protein can also be induced by corticosteroids. Thus, AF2 is a pharmaceutical alternative to corticosteroids. AF2 contains one Met at position 3, which can be oxidized by hydrogen peroxide in vitro and cause inactivation of the peptide (Ye and Wolfe, 1996). This result suggests that ROS produced during an inflammatory response can inactivate AF2; but there is a lack of evidence to support this in vivo.

110

Oien and Moskovitz

2. Chymotrypsin Chymotrypsin is a digestive enzyme formed by a precursor protein secreted from the pancreas. It is known to cleave peptides where a tyrosine, tryptophan, or phenylalanine amino acid is present. Under oxidative conditions in vitro, the Met residue at position 192 has been shown to form a sulfoxide (Cutruzzola et al., 1993). The sulfoxide modification does not induce any significant changes in substrate binding. However, it does interrupt protease inhibitor binding, specifically from the protease inhibitor eglin c. This result may be from the mild structural changes in the region where protease inhibitors are known to bind. 3. H. pylori Msr Substrates The bacterium H. pylori is associated with gastritis and peptic ulcer disease (Enserink, 2005). The H. pylori Msr protein is fused with MsrA‐like and MsrB‐ like domains (Tomb et al., 1997). The Msr protein may be essential for persistent colonization in the host, a role that has been demonstrated in mice (Alamuri and Maier, 2004). The Met‐rich proteins catalase, GrL, Trx‐1, and site‐specific recombinase (SSR) are substrates of Msr under oxidative stress conditions in H. pylori (Alamuri and Maier, 2006). Trx is a well‐known reducing agent for the Msr system (Moskovitz, 2005b), while GrL is mentioned above as a scavenging substrate. a. Catalase. Catalase (KatA) is found in many living organisms where it catalyzes the reaction of hydrogen peroxide to water. Under increased levels of oxygen, catalase activity decreases by 50% in H. pylori lacking Msr (Alamuri and Maier, 2006). Addition of pure Msr can restore the majority of activity. The Msr system may be important to maintain the function of catalase. Furthermore, the Msr system and catalase may be part of a mechanism to prevent ROS toxicity, and thus may be a potential drug target. b. Site‐Specific Recombinase. SSR is a protein that facilitates the recombination of specific DNA sequences. It is a Met‐rich Msr substrate (Alamuri and Maier, 2006). Presumably it could serve as a repair substrate of Msr when oxidized or may serve as secondary role as a scavenger for ROS. 4. Human Immunodeficiency Virus 2 Protease Human Immunodeficiency Virus 2 (HIV‐2) protease is a dimeric enzyme that performs a peptide cleavage of proteins involved in the HIV‐2 reproductive cycle. Both Mets in the protease can be oxidized with high levels of hydrogen peroxide and inhibit the proteolytic activity (Davis et al., 2000).

3. Substrates of Methionine Sulfoxide Reductase

111

The addition of MsrA can restore about 25% of the protease activity. Additionally, the viral polyprotein processing, specifically p26, is inhibited with added oxidant in cultured H9 cells. Acidic conditions and a temperature of 37  C are ideal for Met oxidation (Davis et al., 2002). In this case, the addition of MsrA recovers about 40% of protease activity. Moreover, MsrA preferentially reduces Met95, indicating that oxidation of Met95 has a tendency to form the S‐enantiomer sulfoxide, and Met76 may have a tendency to form R‐MetO. 5. Lysozyme Lysozyme is a mucosal secretion enzyme also found in chicken egg cytoplasm (egg white). It is part of the nonspecific opsonin response that is associated with defense against bacterial pathogens. Two Met residues can be selectively oxidized by photooxidation in acidic conditions (Jori et al., 1968). MetO modification drastically reduces the enzymatic activity to 5% in dead Micrococcus lysodeikticus cells, which correlates with a conformational change in the protein. Addition of 2‐mercaptoethanol recovers most of the activity by restoring the conformation. 6. Pepsin Pepsin is a protease secreted by the stomach to cleave proteins in digestion. It cleaves digested proteins preferentially at aromatic amino acids such as tyrosine and phenylalanine. Hydrogen peroxide treatment in acidic conditions increases Met oxidation on bovine and porcine pepsin (Kido and Kassell, 1975). One Met can be oxidized in both pepsins under acidic conditions. However, the addition of hydrogen peroxide increases the MetO content to two or three residues out of four in porcine pepsin, and increases the content to one or two oxidized Mets out of three Mets residues in bovine pepsin. Oxidation decreases the enzymatic activity of porcine pepsin by varying degrees on diVerent substrates. Bovine pepsin oxidation behaves is a similar way, but to a lesser extent. Furthermore, the Met that is closest to the C‐terminus is the most resistant to oxidative modification. 7. Phosphoglucomutase Phosphoglucomutase is an enzyme involved in glycogenolysis, the breaking down of glycogen into glucose monomers. Phosphoglucomutase can catalyze the reaction of glucose‐1‐phosphate to glucose‐6‐phosphate. Glucose‐6‐ phosphate is a key intermediate in glycolysis. Accessible Met residues in phosphoglucomutase oxidize at a higher rate than other residues when photooxidized, which correlate with a loss of enzymatic activity (Ray et al., 1960).

112

Oien and Moskovitz

8. Ribonuclease A Ribonuclease A (RNase A) is an endonuclease that cleaves the 30 end of single‐stranded ribonucleic acid. Oxidation of RNase A by iodoacetate causes MetO formation and reduces binding activity to the cytidylic acid substrate (Brock et al., 2007). MetO formation in acidic conditions is in the order of Met29 and Met30 being the most sensitive, followed by Met13. The Met79 residue is resistant to oxidation and proposed to be buried inside the structure. The Met29 and Met30 residues are near the active site and may be intrapeptide antioxidants. Met‐oxidized RNase is targeted for degradation by the sulfoxide presence in rabbit reticulocyte extract (Dunten and Cohen, 1989). The combined evidence suggests that the protein is altered by MetO and directly targeted for disposal. 9. Tryptophanase Tryptophanase is an L‐tryptophan indole lyase that works with the coenzyme pyridoxal‐50 ‐phosphate. Pyridoxal‐50 ‐phosphate is the active form of vitamin B6 and a cofactor enzyme for transamination reactions. Chloramine T oxidation rapidly inactivates the enzyme to the synthetic substrate S‐o‐nitrophenyl‐L‐cysteine (Oda and Tokushige, 1988). Four oxidized Mets can inactivate the enzyme, which contains 16 Met residues. The sulfoxide formation is suggested to be part of the slower inactivation process, following the fast inactivation of sulfur–hydrogen groups. Binding of the pyridoxal‐50 ‐phosphate cofactor is only slightly aVected. Evidence suggests that Mets are in or near the active site, possibly playing a protective role. It is unclear if they are important for substrate binding. B. Heme Proteins Heme proteins have a cylic non‐amino acid heme group that often contains an iron atom. Hemoglobin (Hb) is an oxygen transport protein. Cytochromes and cytochrome peroxidases have critical roles in electron transferring and metabolic oxidation (Guallar and Olsen, 2006). 1. Hemoglobin Hb is a globular metalloprotein that transports oxygen from the lungs to tissues. It consists of two ‐ and two ‐subunits. Oxidation of the single ‐subunit Met (in both subunits) by mild chloramine T exposure maintains the structural and chemical properties of oxy‐Hb, but alters the functions of the protein in vitro (Amiconi et al., 1989). The oxidized tetramer has a higher aYnity for oxygen and behaves noncooperatively. In addition, low levels of hydrogen peroxide cause the oxidation of Cys and Mets in vitro (Jia et al., 2007).

3. Substrates of Methionine Sulfoxide Reductase

113

This oxidation may cause structural alterations and decrease biological function, and may be associated with diseases such as hemolytic disease and ischemia reperfusion. 2. Cytochrome c Cytochrome c is a small heme protein and a component of the electron transfer chain in the mitochondria, as well as an intermediate protein in the apoptosis pathway. Photooxidation of the protein can convert the Met side chains to sulfoxides, which results in a decreased redox potential (Ivanetich et al., 1976). Met oxidation is the only modification with these conditions and does not significantly alter binding to cytochrome c oxidase. Excess amounts of hypochlorous acid causes an increase of oxidase activity and also covalently incorporates an oxygen atom into the Met80 heme ligand (Chen et al., 2002). In addition, a sulfoxide can form on Met65 with increased oxidant. Oxidation of cytochrome c by hypochlorous acid causes an impaired ability to support oxygen consumption by cytochrome c oxidase, suggesting a possible break in the electron transport chain. Evidence suggests that there is no significant structural change to the protein accompanying Met oxidation with exception of the heme‐binding pocket. Moreover, oxidized cytochrome c exposed to hydrogen peroxide can form a tyrosyl radical, which may cause further damage and may indirectly be associated with the apoptotic pathway. 3. Cytochrome c Oxidase Cytochrome c oxidase is the enzyme that transfers electrons from cytochrome c to oxygen, at the expense of oxidizing iron (II) to iron (III). Hydrogen peroxide can oxidize three Met residues at positions 119, 230, and 231 (Kim and Erman, 1988). The three modifications do not significantly aVect cyanide binding, suggesting that the oxidation is causing a major structural diVerence in the protein. Loss of activity to react with hydrogen peroxide to from enzymatic intermediate product with oxidized iron, termed compound I, correlates with one of the Met230/231 pair forming a sulfoxide. This suggests that the heme microenvironment is altered and does not react to hydrogen peroxide. C. Hormones 1. Adrenocorticotropic Hormone Adrenocorticotropic hormone (ATCH) is a hormone secreted from the anterior pituitary gland that stimulates the cortex of the adrenal gland to increase corticosteroid production. Corticosteroids are involved in various roles in the body, including stress and immune responses, inflammation regulation,

114

Oien and Moskovitz

metabolism, and the breakdown of proteins. The hormone contains a Met in position 4 that can form a sulfoxide in the presence of hydrogen peroxide or chloramine T in vitro (Antonini and Spoto, 1988). Oxidants can cause Met oxidation and biological inactivation of the peptide in the adrenal‐ascorbic acid depletion method (Dedman et al., 1961). 2. Bombesin Bombesin is a peptide found in frog skin containing 14 amino acids and is homologous to the mammalian neuromedin B and gastrin‐releasing peptide. It is known to interact with gastrointestinal tract neuroendocrine hormone and stimulate gastrin release from G cells. Oxidative conditions decrease the binding activity to G cells in vitro (Vigna et al., 1988). There are no Cys in bombesin, and the binding activity can be restored with incubation in reducing chemicals. The protein structure suggests that the binding activity is reduced by ROS activity forming MetOs. 3. Calcitonin Calcitonin (CT) is a peptide hormone produced by the parafollicular cells of the thyroid gland and is involved in calcium regulation. It is also a pharmaceutical treatment for osteoporosis and hypercalcemia. The peptide contains two Met residues, or four in the dimeric form. Hydrogen peroxide can oxidize Met8 and alter the conformation of human CT in vitro (Nabuchi et al., 2004). The rate of oxidation correlates with a loss of hypocalcemic activity. The sulfoxide causes the protein to unfold and is predicted to be in a rigid conformation. 4. Cholecystokinin Cholecystokinin (CCK) is a peptide hormone secreted by the duodenum to stimulate the digestion of fat and protein. It can be oxidized with a mild amount of hydrogen peroxide that reduces the activity of gall bladder contractions in situ (Mutt, 1981). The activity is restored by about 75% with N‐methylmercaptoacetamide (chemical reductant) incubation. The hormone does not contain Cys, and the evidence of oxidized peptide reduction suggests that the Met oxidation causes the inactivation. 5. Glucagon Glucagon is a small hormone protein secreted by the pancreas to stimulate glyconeogenesis in response to low blood sugar. The single Met in position 27 can be oxidized during chloramine T iodination and causes a one‐third decrease in binding aYnity to rat adipocytes and hepatocytes versus a lactoperoxidase control lacking Met residues (Sonne et al., 1982). This result suggests that glucagon activity is altered by the oxidized Met.

3. Substrates of Methionine Sulfoxide Reductase

115

6. Luteinizing Hormone Luteinizing hormone (LH) is secreted by the anterior pituitary gland to regulation reproductive function. Chloramine T oxidation of bovine LH modifies five to six out of seven Mets (De la Llosa et al., 1980). Met oxidation correlates with a complete loss of biological activity in the ovarian ascorbic acid depletion test. In addition, peroxides can oxidize all seven Mets in the hormone. 7. Parathyroid Hormone Parathyroid hormone (PTH) is a small protein released by the parathyroid glands and is associated with calcium regulation. It can act to increase blood calcium in response to hypocalcemia, and inverse action of CT. Hydrogen peroxide can oxidize one to two Mets in the bovine hormone and decrease its potency in bovine kidney membrane adenylyl cyclase assays (Frelinger and Zull, 1984). Oxidation of Met18 causes the least decrease in potency, while oxidation of Met8 decreases potency at relatively moderate level. Also, oxidation of both Met residues causes the largest decrease in potency. Met was the only modified residue found after low levels of hydrogen peroxide exposure, and the MetOs can be reduced with mercaptoethylamine in vitro. In addition, hydrogen peroxide can also oxidize the two Mets in recombinant human PTH, which correlates to a loss of biological activity in rat osteosarcoma adenylate cyclase assays (Nabuchi et al., 1995). This observation suggests that the region near the Mets, especially Met8, is important for PTH activity. 8. Prolactin Prolactin is a peptide hormone secreted by the anterior pituitary gland and is mainly involved in breast milk secretion. Four of the seven Mets in the hormone can be oxidized with moderate levels of hydrogen peroxide (Houghten and Li, 1976). Under this oxidation state, ovine prolactin has a slight reduction of biological activity after injection in a pigeon crop‐sac assay (response by dry mucosal weight). Moreover, all seven Met residues can be oxidized in more extreme oxidant conditions. These data may indicate that the former Met residues are surface exposed. Thus, the latter residues (Met36, Met81, and Met132) may be partially buried and resistant to sulfoxide formation. 9. Vasoactive Intestinal Peptide Vasoactive intestinal peptide (VIP) is a hormone produced in the pancreas. It is found in the gastrointestinal tract, heart, lungs, thyroid, kidney, urinary bladder, genital organs, and the brain at significant levels (Henning and Sawmiller, 2001).

116

Oien and Moskovitz

It is also associated with the biological clock (suprachiasmatic nucleus) (Buijs et al., 2003). Oxidation of Met17 in the porcine peptide has no eVect on the pancreatic response of an anesthetized cat (Mutt, 1981). The Met residue may not be near the active site or the oxidized residue may not cause a significant conformational change. However, it is also a likely possibility that the endogenous feline Msrs were able to eYciently reduce the peptide. D. Neurodegenerative Disease Associated A neurodegenerative disease is a condition resulting from the deterioration of neurons, ultimately aVecting brain function. They often present with problems in movement or memory and dementia. Certain neurodegenerative diseases are associated with a specific brain protein. Examples are amyloid‐ , alpha‐synuclein (S), and prion (Moskovitz, 2005a). The functions of these brain proteins are unknown, but they are each implicated in a respective neurodegenerative disease. Oxidant‐mediated accumulation of modified proteins may be one of the major causes of age‐related neurodegenerative disease (Moskovitz and Bush, 2005). In addition, the Msr system may prevent irreversible protein damage and thus accumulation of posttranslational modified proteins. 1. Alzheimer’s Disease Associated Alzheimer’s disease is a data‐processing disorder associated with neural cell neurofibrillary tangles and the extracellular deposition of the amyloid plaques. Amyloid plaques are extracellular plaque filaments with an amyloid structure, and contain the amyloid‐ peptide. The amyloid‐ peptide is formed by cleavage of the amyloid precursor protein (APP) (Goedert and Spillantini, 2006). This cleavage commonly results in an amyloid‐ peptide of 40 or 42 residues. Amyloid‐ ‐40 concentrates in plaques of cerebral arteries. Amyloid‐ ‐42 is more concentrated in neural plaques. It is neurotoxic and thought to cause neurodegeneration. The toxicity of amyloid‐ ‐42 is suggested to play a central role in Alzheimer’s disease. The formation of amyloid plaques is considered one possible pathway of Alzheimer’s disease progression. However, this theory is not conclusive. a. Amyloid‐b‐42. Amyloid‐ ‐42 (A ) is a neurotoxin found in extracellular amyloid plaques. It contains 42 amino acids, including one Met at position 35. The neurotoxic properties of amyloid‐ may be a result of reactive species and free radical formation (Pogocki, 2003). Oxidation of Met35 can reduce the amyloid‐ ‐42 fibrillation rate (Hou et al., 2002). The fibrillation rate may be associated with its ability to form the insoluble plaques, although the exact role is still unclear. The oxidized Met35 alters the peptide structure and

3. Substrates of Methionine Sulfoxide Reductase

117

decreases the hydrophobicity of the peptide (Barnham et al., 2003). The change in hydrophobicity decreases its ability to penetrate lipid membranes and may contribute to the peptides altered location. The Msr system may assist in retaining the structure of amyloid‐ ‐42 (Moskovitz and Bush, 2005). Moreover, the transduction of MsrA as a fusion protein to cells with amyloid‐ ‐42 increases the viability of the cultures (Jung et al., 2003). In addition, Msr activity was shown to decrease in Alzheimer’s brains (Gabbita et al., 1999). The higher levels of oxidative stress correlate with the presence of amyloid‐ in the hippocampus and cortex. b. Amyloid Precursor Protein. APP is a glycoprotein expressed throughout the plasma membrane. Mutations in the APP gene are thought to cause a familial form of Alzheimer’s disease. For example, Val717 can be changed to a phenylalanine or glycine. This substitution stimulates G‐protein binding and signal transduction in aged human brain membranes (Karelson et al., 2005). The stimulation is weakened in the temporal cortex when a Met at position 722 is oxidized to a sulfoxide (Reis et al., 2007). Furthermore, oxidized Met722 in the V717G mutated peptide causes less mitochondrial damage in neuron cell cultures. The neurotoxicity of this mutation may be mediated by Met oxidation; thus, the Msr system may also contribute to this regulation. 2. Parkinson’s Disease Associated Parkinson’s disease is a neurodegenerative disease that causes alterations in motor function and postural stability. The disease is characterized by a progressive loss of dopaminergic neurons in the substantia nigra (Wood‐ Kaczmar et al., 2006). The majority of Parkinson’s cases are sporadic, and the cause of the disease is unknown. However, over 5% of Parkinson’s disease cases are familial, causes by genetic mutations. These cases include mutations in the S and DJ‐1 genes. Hence, the S and DJ‐1 proteins may also be implicated in sporadic Parkinson’s disease. a. Alpha‐Synuclein. S is a neural protein that is to play a role in multiple diseases, collectively termed synucleinopathies. It is found in the lewy bodies of postmortem Parkinson’s disease patients. Lewy bodies are cytoplasmic inclusions mainly in nigral dopaminergic neurons. S is soluble, natively unfolded protein that has no ordered structure under physiological conditions (Uversky et al., 2000). S has four Mets at positions 1, 5, 116, and 127 (Glaser et al., 2005). All four Met residues can be oxidized with hydrogen peroxide, which inhibits S fibrillation (Uversky et al., 2002). Metals such as titanium, zinc, aluminum, and lead can overcome the inhibited fibrillation of the Met‐ oxidized protein (Yamin et al., 2003). If the reversal of Met oxidation can

118

Oien and Moskovitz

prevent further irreversible modification to other residues, it may be a new target for therapeutic intervention (Moskovitz and Bush, 2005). b. DJ‐1. DJ‐1 is ubiquitously expressed and has no known function. However, it is associated with various cellular processes, including a oxidative stress response (Lev et al., 2006). Mutations in DJ‐1 are associated with a genetic autosomal recessive form of Parkinson’s disease, and suggested to cause neurodegeneration. However, patients with sporadic Parkinson’s and Alzheimer’s disease can accumulate oxidized DJ‐1 in frontal cortex brain tissues (Choi et al., 2006). Oxidized modifications are found including Cys and Met residues and additionally irreversible carbonyls and Met sulfones. It is suggested that irreversible oxidations occur under more extreme oxidant conditions, after Cys and Met are oxidized in a reversible state. Moreover, the levels of DJ‐1 are increased in Parkinson’s and Alzheimer’s brains. Altogether, reduction systems such as Msr may protect DJ‐1 in healthy brains, and elevated oxidant conditions in diseased brains may lead to irreversible residue modifications.

3. Protease‐Resistant Protein‐C Protease‐resistant protein‐C (Prpc) is the protein found in prions, which are present in the cells of nervous and immune tissue. Prion is a short name for proteinaceous infectious particle, and C refers to cellular. PrpSc is a disease‐ associated isomer. The Sc refers to scapie, a prion disease in sheep. These proteins can be genetic, sporatic, or infectious. They are known to infect brain neural tissue and induce misfolding of other proteins. Diseases associated with prions are called spongiform encephalopathies, one example being Creutzfeldt‐Jakob disease (Moskovitz, 2005a). Creutzfeldt‐Jakob disease is a rapidly progressing neurodegenerative disease characterized by dementia and loss of memory. Prpc exists in an ‐helix structure. It can convert into the ‐sheet confirmation of the insoluble Sc isomer, probably by conformational changes involving posttranslational modification (Jackson et al., 1999). In addition, a polymorphism at residue 129 results in either a valine or Met surface‐exposed amino acid that can aVect the susceptibility of developing Creutzfeldt‐Jakob disease (Hill et al., 2003). Hydrogen peroxide can cause Met oxidation, which interferes with the conformation change (Breydo et al., 2005). Moreover, Mets in mouse and chicken prion protein can be selectively oxidized and can form free radicals by copper (II) ions when they refold in vitro (Wong et al., 1999). Furthermore, the cellular prion protein level is higher in msrA null mice (Williams et al., 2004). This may be a compensation pathway of additional oxidant stress from the lack of an endogenous antioxidant enzyme (Moskovitz and Bush, 2005). It has not been determined if

3. Substrates of Methionine Sulfoxide Reductase

119

Met29 and other Mets are involved in the prion cascade of events. Prion may function as an antioxidant (Wong et al., 2001) or as a signal to increase endogenous antioxidant systems (Rachidi et al., 2003). E. Serine Protease Inhibitors (Serpins) 1. Factor VIIa Factor VII is a vitamin K‐dependent serine protease in the coagulation cascade. Factor VIIa (FVIIa) is the activated form of factor VII that binds to many components including tissue factor (cofactor) and factor X. Out of four Mets in FVIIa, Met298 and Met306 can be oxidized with hydrogen peroxide (Kornfelt et al., 1999). Met oxidation weakens binding to soluble tissue factor causing a threefold decrease in the dissociation constant. Also, the amidolytic activity is significantly decreased in the presence of soluble tissue factor. Factor X activation is reduced by with oxidized FVIIa and lipid surface tissue factor, although the FVIIa binding to lipid surface tissue factor is not significantly altered. Evidence suggests that the Mets in FVIIa are surface exposed, and thus readily accessible to solvent. Furthermore, Met oxidation may be implicated in factor VII modulation.

2. Ovoinhibitor Ovoinhibitor is a serine protease found in chicken egg cytoplasm. Oxidation of Mets in chicken ovoinhibitor by N‐chlorosuccinimide at moderately low levels decreases inhibitory rate of activity on chymotrypsin (Shechter et al., 1977). In this case, two to three out of four Mets are oxidized. However, all four methionines can be oxidized with increased oxidant, suggesting that the former case may involve surface exposed Met residues. Furthermore, in either case, trypsin‐inhibiting activity is not aVected, while elastase activity decreases completely after the modification of Met residues.

3. Plasminogen Activator Inhibitor‐1 Plasminogen activator inhibitor‐1 (PAI‐1) is a serpin known to inhibit t‐PA and urokinase, both activators of plasminogen. Activated plasmin can break down fibrin clots, products of coagulation. PAI‐1 can be inactivated with chloramine T in vitro, causing a loss of inhibitory activity and binding aYnity to t‐PA (Lawrence and LoskutoV, 1986). Subsequent incubation with MsrA and dithiothreitol restores over 90% of this activity. Furthermore, oxidation of Mets in position 266, 347, or 354 can cause a conformational change, correlating with the loss of inhibitory activity.

120

Oien and Moskovitz

4. Secretory Leukocyte Protease Inhibitor Secretory leukocyte protease inhibitor (SLPI) is a serpin that is known to inhibit neutrophil elastase and cathepsin G proteases. Ozone at relatively stoichiometric levels can oxidize two of the three Met residues in the inhibitor (Smith et al., 1987). This oxidation decreases the ability to inhibit neutrophils elastase and cathepsin G activity. The unmodified Met can be oxidized with increased levels of ozone. In addition, tryptophan and tyrosine are modified at the same high levels. The relative susceptibility of two Mets suggests they may be surface exposed. Furthermore, the Mets may be preventing irreversible modifications to tryptophan and tyrosine residues. F. Snake Venom Toxins 1. Beta1‐Bungarotoxin Bungarotoxin is a neurotoxin found in krait snake venom. Beta1‐bungarotoxin ( 1‐Butx) binds to presynaptic terminal proteins and prevents the release of acetylcholine, which blocks neuromuscular transmission. The Met residue at position 8 is oxidized in vitro prior to the modification of other residues (Chu et al., 1993). Also, trace amounts of oxidized Met6 and other modifications are found. 1‐Butx contains a phospholipase A2 subunit, and the oxidation of Met6 reduces phospholipase A2 activity. The reduced activity causes a loss of lethality properties when injected into mice. 2. Cardiotoxin‐VII1 Cardiotoxin‐VII1 is a toxin found in the venom of the cobra snake, Naja melanoleuca. The toxin contains two Mets that can be oxidized with N‐chlorosuccinimide in vitro (Carlsson and Louw, 1978). The MetO formation diminishes activity with phospholipase A2 and is nonlethal to mice. The loss of activity suggests the Mets may be part of or near an active site that is implicated in the toxic eVects.

3. Cytotoxin‐III Cytotoxin‐III (CX3) is a toxin found in the venom of the Thai cobra snake, Naja naja atra. It contains two Met residues that can be oxidized with chloramine T in vitro (Stevens‐Truss and Hinman, 1996). Oxidation of Met26 is more susceptible to oxidation and decreases the toxicity of CX3 in tissue cultures of both human leukemic T lymphocytes and rat heart myoblasts. Furthermore, the removal of the peptide portion containing the Mets also decreases the toxic properties, but only has modest eVects of binding. The Mets may be important to the toxic eVects of the venom.

3. Substrates of Methionine Sulfoxide Reductase

121

4. Echistatin Echistatin is a peptide found in Echis carinatus snake venom and can block specific integrin receptors on the cell membrane. The Met28 residue near the integrin‐binding site can be oxidized by chloramine T, and this modification correlates with a lower aYnity for the integrin receptor (v 3) (Kumar et al., 1998). Furthermore, the oxygen in the sulfoxide can form hydrogen bonds with a glycine and aspartic acid in this binding site, indicating sulfoxide formation hinders receptor binding, and thus causing inactivation of this antagonist.

G. Miscellaneous Substrates 1. Actin G‐actin is a globular protein that polymerizes to form actin filaments (F‐actin) in the cytoskeleton of cells. The surface‐exposed Mets residues at positions 44, 47, and 355 can be readily oxidized with chloramine T (Dalle‐Donne et al., 2002). The Met modifications slightly decrease the capability of actin polymerization. However, increased chloramine T concentrations can cause oxidation of Mets at positions 176, 190, and 269, causing drastic depolymerization in F‐actin. The oxidized Mets increase surface hydrophobicity and indicate a conformational change of the protein, in addition to proteolysis resistance. Additionally, oxidized Mets, oxidized Cys, and carbonyls can form in harsh oxidant conditions using hypochloric acid (DalleDonne et al., 1999). The carbonyls form after oxidation of Cys and Met and indicate that the sulfur‐ containing amino acids may be scavenging ROS to protect the formation of irreversible carbonyls. 2. Ffh Ffh is part of the E. coli signal recognition particle (SRP), a protein‐RNA complex that transports proteins to the plasma membrane. It is similar to the human SRP54 protein, and both contain a Met‐rich domain (Ezraty et al., 2004). MetOs in Ffh inhibit its ability to bind RNA. The Msr system (MsrA and MsrB) can recover the RNA‐binding ability. Furthermore, the protein loses stability in bacteria lacking MsrA and MsrB, indicating that the Msr system is crucial for protection of protein transportation in E. coli. 3. Fibronectin Fibronectin (Fn) is made in the liver and is a relatively large glycoprotein composed of two subunits joined by a disulfide bond. It is can bind to integrin receptor proteins and also to collagen, fibrin, and heparin. Three of the nine Met residues in the collagen‐binding region can be oxidized with

122

Oien and Moskovitz

chloramine T in vitro (Miles and Smith, 1993). The MetOs prevent collagen binding, which is important for the formation of extracellular matrix. Also, Streptococcus gordonii lacking the msrA gene does not bind Fn in extracellular matrix bioassays (Giomarelli et al., 2006). Thus, oxidation of the Mets in Fn inhibits binding to the extracellular matrix. Moreover, MsrA may be critical for the infection of S. gordonii. 4. Fifth Component of Complement The fifth component of complement (C5) is a protein that is part of the complement cascade of defense against microbes. This protein family is activated sequentially to form a membrane attack complex on the cell membrane. The human fifth component is activated by C5 convertase, cleaving the protein into subunits. In addition, it can be activated by the oxidants hypochlorite and taurine chloramine in guinea pig red blood cells and is accompanied by the oxidation of half of the Met residues (Vogt and Hesse, 1994). However, the activation does not cleave the protein, but the oxidized protein does bind to the sixth component, and is present in the membrane attack complex. Furthermore, the altered activation is not an eVect caused by Cys and tryptophan residues (Vogt et al., 1992). The combined evidence suggests an alternative pathway to activate the fifth component, possibly by changing the structure and exposing a binding site for the sixth component. 5. High‐Mobility Group‐D High‐mobility group proteins are chromosomal proteins that assist in DNA transcription, replication, recombination, and repair functions. High‐mobility group‐D (HMD‐D) proteins are found in the early embryogenesis of Drosophila melanogaster. Purified recombinant HMD‐D can contain oxidized Met13 after the process of expression in E. coli (Dow et al., 1997). The oxidation of Met13 decreases the hydrophobicity and lowers the DNA‐binding aYnity of the protein. Furthermore, the structure of the oxidized protein may also be preventing DNA intercalation. The Met46 residue in the hydrophobic core is not oxidized in this process. 6. L7/L12 The E. coli protein L12 is located in the large subunit of the ribosome and can be acetylated to from L7. The L7/L12 proteins dimerize and form the stalk of the 50S ribosomal subunit. L12 will not dimerize and become acetylated in oxidative conditions (Caldwell et al., 1978). The biological activity is aVected by oxidation of Met14, Met17, and Met26 and can be restored by the addition of MsrA (Brot et al., 1981).

3. Substrates of Methionine Sulfoxide Reductase

123

7. Met‐Enkephalin An enkephalin is a 5‐amino acid peptide mainly associated with pain regulation. Met‐enkephalin (MENK) contains a C‐terminal Met. It is oxidized by phagocytosing neutrophils causing degradation resistance (Turkall et al., 1982), and oxidation be partly dependent on the myeloperoxidase system. The other four residues are not aVected in the presence of myeloperoxidase, hydrogen peroxide, and halides. The sulfoxide derivative has reduced opiate agonist activity (Beddell et al., 1977). This suggests that oxidation may contribute to pain from inflammation. It is also possible that MENK may assist in oxidant scavenging. Evidence demonstrating resistance to oxidative stress by MENK in mice brains supports this idea (Balog et al., 2004). 8. Neuropeptide Y Neuropeptide Y (NYP) is a neurotransmitting peptide found in the brain. It is expressed in regions associated with seizure generation and propagation, and additionally linked to epileptic activity regulation (Baraban, 2004). A pentylenetetrazole‐induced seizure in rats increases the NYP gene expression (McCarthy et al., 1998). The majority of the increased NYP in the hippocampal region contains MetOs. This modification alters the peptide release sensitivity of calcium concentration. However, the aYnity to receptors is similar to the native peptide. NYP may be part of a mechanism for excitability control. 9. Subtilisin Subtilisin is a serine protease secreted from the Bacillus bacterium. Met522 out of five Met residues is oxidized at moderate levels of hydrogen peroxide in vitro (StauVer and Etson, 1969). The MetO‐protease has decreased the enzyme activity eYciency with synthetic substrates such as N‐cinnamoylimidazole. The Met may be near the active site, possibly protecting the protease activity of the site by scavenging ROS.

V. Discussion In summary, we report here the broad spectrum of substrates of the Msr system. The physiological significance of the MetO in conjunction with the Msr system is based on the evidence presented. Furthermore, the substrates are classified into the three categories of regulated substrates, scavenging substrates, and modified substrates resulting in various changes in properties. The complete list of substrates is presented in Table I. In addition, the substrates are subcategorized depending on the experimental evidence of in vivo/ex vivo (including in situ) and in vitro.

124

Oien and Moskovitz

Table I Classification of Substrates for Met Oxidation Regulated in vivo/ex vivo 1‐ATa Apo‐AIa Calcium‐activated potassium channelsa IB TMa Voltage‐activated potassium channelsa Scavenging in vivo/ex vivo GSa GrL GHa IFN‐ a sHspsa Modified Substrates in vivo/ex vivo S APP A a Cardiotoxin‐VII1a Catalase CCKa CX3 DJ‐1 Ffha Fna Glucagona LHa MENKa NYPa PTHa Prolactina Prpc SSR VIPa Regulated in vitro Apo‐AIIa t‐PAa a

Scavenging in vitro 2‐PIa 2Ma AT IIIa CSa IFN‐2ba IL‐6a 15‐lipoxygenasea MPIa SCFa Modified Substrates in vitro Actin ATCHa AF2a 1‐Butxa Bombesina CTa Chymotrypsina Complement C5a Cytochrome ca Cytochrome c oxidasea Echistatina FVIIaa Hb HMD‐D HIV‐2 proteasea L12/L7 (ribosomal proteins)a Lysozymea Ovoinhibitora Pepsina Phosphoglucomutasea PAI‐1a RNase Aa Secretory leukocyte inhibitora Subtilisina Tryptophanasea

Substrates that are listed by the review of Levine et al. (2000).

The continuing research will provide new information on the role of the Met redox cycle in regulation, redox homeostasis, prevention of irreversible modifications, pathogenesis, and the aging process. Moreover, the characterization of Met oxidation and the Msr system may lead to new targets for therapeutic intervention.

3. Substrates of Methionine Sulfoxide Reductase

125

References Abrams, W. R., Weinbaum, G., Weissbach, L., Weissbach, H., and Brot, N. (1981). Enzymatic reduction of oxidized alpha‐1‐proteinase inhibitor restores biological activity. Proc. Natl. Acad. Sci. USA 78, 7483–7486. Alamuri, P., and Maier, R. J. (2004). Methionine sulphoxide reductase is an important antioxidant enzyme in the gastric pathogen Helicobacter pylori. Mol. Microbiol. 53, 1397–1406. Alamuri, P., and Maier, R. J. (2006). Methionine sulfoxide reductase in Helicobacter pylori: Interaction with methionine‐rich proteins and stress‐induced expression. J. Bacteriol. 188, 5839–5850. Amiconi, G., Ascoli, F., Barra, D., Bertollini, A., Matarese, R. M., Verzili, D., and Brunori, M. (1989). Selective oxidation of methionine beta(55)D6 at the alpha 1 beta 1 interface in hemoglobin completely destabilizes the T‐state. J. Biol. Chem. 264, 17745–17749. Anantharamaiah, G. M., and Garber, D. W. (1996). Chromatographic methods for quantitation of apolipoprotein A‐I. Methods Enzymol. 263, 267–282. Anantharamaiah, G. M., Hughes, T. A., Iqbal, M., Gawish, A., Neame, P. J., Medley, M. F., and Segrest, J. P. (1988). EVect of oxidation on the properties of apolipoproteins A‐I and A‐II. J. Lipid Res. 29, 309–318. Antonini, G., and Spoto, G. (1988). Reversed phase high performance liquid chromatography of adrenocorticotropin 1–39 and its fragments in the native and oxidized forms. Biochem. Int. 16, 1009–1017. Balog, E. M., Norton, L. E., Bloomquist, R. A., Cornea, R. L., Black, D. J., Louis, C. F., Thomas, D. D., and Fruen, B. R. (2003). Calmodulin oxidation and methionine to glutamine substitutions reveal methionine residues critical for functional interaction with ryanodine receptor‐1. J. Biol. Chem. 278, 15615–15621. Balog, T., Sobocanec, S., Sverko, V., and Marotti, T. (2004). Met‐enkephalin modulates resistance to oxidative stress in mouse brain. Neuropeptides 38, 298–303. Bar‐Noy, S., and Moskovitz, J. (2002). Mouse methionine sulfoxide reductase B: EVect of selenocysteine incorporation on its activity and expression of the seleno‐containing enzyme in bacterial and mammalian cells. Biochem. Biophys. Res. Commun. 297, 956–961. Baraban, S. C. (2004). Neuropeptide Y and epilepsy: Recent progress, prospects and controversies. Neuropeptides 38, 261–265. Barnham, K. J., Ciccotosto, G. D., Tickler, A. K., Ali, F. E., Smith, D. G., Williamson, N. A., Lam, Y. H., Carrington, D., Tew, D., Kocak, G., Volitakis, I., Separovic, F., et al. (2003). Neurotoxic, redox‐competent Alzheimer’s beta‐amyloid is released from lipid membrane by methionine oxidation. J. Biol. Chem. 278, 42959–42965. Bartlett, R. K., Bieber Urbauer, R. J., Anbanandam, A., Smallwood, H. S., Urbauer, J. L., and Squier, T. C. (2003). Oxidation of Met144 and Met145 in calmodulin blocks calmodulin dependent activation of the plasma membrane Ca‐ATPase. Biochemistry 42, 3231–3238. Beddell, C. R., Clark, R. B., Hardy, G. W., Lowe, L. A., Ubatuba, F. B., Vane, J. R., and Wilkinson, S. (1977). Structural requirements for opioid activity of analogues of the enkephalins. Proc. R. Soc. Lond. B Biol. Sci. 198, 249–265. Biewenga, G. P., Veening‐GriYoen, D. H., Nicastia, A. J., Haenen, G. R., and Bast, A. (1998). EVects of dihydrolipoic acid on peptide methionine sulfoxide reductase. Implications for antioxidant drugs. Arzneimittelforschung 48, 144–148. Bigelow, D. J., and Squier, T. C. (2005). Redox modulation of cellular signaling and metabolism through reversible oxidation of methionine sensors in calcium regulatory proteins. Biochim. Biophys. Acta 1703, 121–134. Boudier, C., and Bieth, J. G. (1994). Oxidized mucus proteinase inhibitor: A fairly potent neutrophil elastase inhibitor. Biochem. J. 303(Pt. 1), 61–68.

126

Oien and Moskovitz

Breydo, L., Bocharova, O. V., Makarava, N., Salnikov, V. V., Anderson, M., and Baskakov, I. V. (2005). Methionine oxidation interferes with conversion of the prion protein into the fibrillar proteinase K‐resistant conformation. Biochemistry 44, 15534–15543. Brock, J. W., Ames, J. M., Thorpe, S. R., and Baynes, J. W. (2007). Formation of methionine sulfoxide during glycoxidation and lipoxidation of ribonuclease A. Arch. Biochem. Biophys. 457, 170–176. Brot, N., and Weissbach, H. (2000). Peptide methionine sulfoxide reductase: Biochemistry and physiological role. Biopolymers 55, 288–296. Brot, N., Weissbach, L., Werth, J., and Weissbach, H. (1981). Enzymatic reduction of protein‐ bound methionine sulfoxide. Proc. Natl. Acad. Sci. USA 78, 2155–2158. Brouillette, C. G., Anantharamaiah, G. M., Engler, J. A., and Borhani, D. W. (2001). Structural models of human apolipoprotein A‐I: A critical analysis and review. Biochim. Biophys. Acta 1531, 4–46. Buijs, R. M., van Eden, C. G., Goncharuk, V. D., and Kalsbeek, A. (2003). The biological clock tunes the organs of the body: Timing by hormones and the autonomic nervous system. J. Endocrinol. 177, 17–26. Butt, A. M., and Kalsi, A. (2006). Inwardly rectifying potassium channels (Kir) in central nervous system glia: A special role for Kir4.1 in glial functions. J. Cell. Mol. Med. 10, 33–44. Caldwell, P., Luk, D. C., Weissbach, H., and Brot, N. (1978). Oxidation of the methionine residues of Escherichia coli ribosomal protein L12 decreases the protein’s biological activity. Proc. Natl. Acad. Sci. USA 75, 5349–5352. Carlsson, F. H., and Louw, A. I. (1978). The oxidation of methionine and its eVect of the properties of cardiotoxin VII1 from Naja melanoleuca venom. Biochim. Biophys. Acta 534, 322–330. Carp, H., and JanoV, A. (1980). Inactivation of bronchial mucous proteinase inhibitor by cigarette smoke and phagocyte‐derived oxidants. Exp. Lung Res. 1, 225–237. Chao, C. C., Ma, Y. S., and Stadtman, E. R. (1997). Modification of protein surface hydrophobicity and methionine oxidation by oxidative systems. Proc. Natl. Acad. Sci. USA 94, 2969–2974. Chen, J., Avdonin, V., Ciorba, M. A., Heinemann, S. H., and Hoshi, T. (2000). Acceleration of P/C‐type inactivation in voltage‐gated K(þ) channels by methionine oxidation. Biophys. J. 78, 174–187. Chen, Y. R., Deterding, L. J., Sturgeon, B. E., Tomer, K. B., and Mason, R. P. (2002). Protein oxidation of cytochrome C by reactive halogen species enhances its peroxidase activity. J. Biol. Chem. 277, 29781–29791. Choi, J., Sullards, M. C., Olzmann, J. A., Rees, H. D., Weintraub, S. T., Bostwick, D. E., Gearing, M., Levey, A. I., Chin, L. S., and Li, L. (2006). Oxidative damage of DJ‐1 is linked to sporadic Parkinson and Alzheimer diseases. J. Biol. Chem. 281, 10816–10824. Chu, S. T., Chu, C. C., Tseng, C. C., and Chen, Y. H. (1993). Met‐8 of the beta 1‐bungarotoxin phospholipase A2 subunit is essential for the phospholipase A2‐independent neurotoxic eVect. Biochem. J. 295(Pt. 3), 713–718. Ciorba, M. A., Heinemann, S. H., Weissbach, H., Brot, N., and Hoshi, T. (1997). Modulation of potassium channel function by methionine oxidation and reduction. Proc. Natl. Acad. Sci. USA 94, 9932–9937. Ciorba, M. A., Heinemann, S. H., Weissbach, H., Brot, N., and Hoshi, T. (1999). Regulation of voltage‐dependent Kþ channels by methionine oxidation: EVect of nitric oxide and vitamin C. FEBS Lett. 442, 48–52. Cutruzzola, F., Ascenzi, P., Barra, D., Bolognesi, M., Menegatti, E., Sarti, P., Schnebli, H. P., Tomova, S., and Amiconi, G. (1993). Selective oxidation of Met‐192 in bovine alpha‐ chymotrypsin. EVect on catalytic and inhibitor binding properties. Biochim. Biophys. Acta 1161, 201–208.

3. Substrates of Methionine Sulfoxide Reductase

127

DalleDonne, I., Milzani, A., and Colombo, R. (1999). The tert‐butyl hydroperoxide‐induced oxidation of actin Cys‐374 is coupled with structural changes in distant regions of the protein. Biochemistry 38, 12471–12480. Dalle‐Donne, I., Rossi, R., Giustarini, D., Gagliano, N., Di Simplicio, P., Colombo, R., and Milzani, A. (2002). Methionine oxidation as a major cause of the functional impairment of oxidized actin. Free Radic. Biol. Med. 32, 927–937. Dalle‐Donne, I., Aldini, G., Carini, M., Colombo, R., Rossi, R., and Milzani, A. (2006). Protein carbonylation, cellular dysfunction, and disease progression. J. Cell. Mol. Med. 10, 389–406. Davis, D. A., Newcomb, F. M., Moskovitz, J., Wingfield, P. T., Stahl, S. J., Kaufman, J., Fales, H. M., Levine, R. L., and Yarchoan, R. (2000). HIV‐2 protease is inactivated after oxidation at the dimer interface and activity can be partly restored with methionine sulphoxide reductase. Biochem. J. 346(Pt. 2), 305–311. Davis, D. A., Newcomb, F. M., Moskovitz, J., Fales, H. M., Levine, R. L., and Yarchoan, R. (2002). Reversible oxidation of HIV‐2 protease. Methods Enzymol. 348, 249–259. De la Llosa, P., El Abed, A., and Roy, M. (1980). Oxidation of methionine residues in lutropin. Can. J. Biochem. 58, 745–748. Dedman, M. L., Farmer, T. H., and Morris, C. J. (1961). Studies on pituitary adrenocorticotrophin. 3. Identification of the oxidation‐reduction centre. Biochem. J. 78, 348–352. Dow, L. K., Changela, A., Hefner, H. E., and Churchill, M. E. (1997). Oxidation of a critical methionine modulates DNA binding of the Drosophila melanogaster high mobility group protein, HMG‐D. FEBS Lett. 414, 514–520. Dunten, R. L., and Cohen, R. E. (1989). Recognition of modified forms of ribonuclease A by the ubiquitin system. J. Biol. Chem. 264, 16739–16747. Enserink, M. (2005). Physiology or medicine. Triumph of the ulcer‐bug theory. Science 310, 34–35. Esmon, C. T. (1989). The roles of protein C and thrombomodulin in the regulation of blood coagulation. J. Biol. Chem. 264, 4743–4746. Ezraty, B., Grimaud, R., El Hassouni, M., Moinier, D., and Barras, F. (2004). Methionine sulfoxide reductases protect Ffh from oxidative damages in Escherichia coli. EMBO J. 23, 1868–1877. Fielding, C. J., and Fielding, P. E. (1995). Molecular physiology of reverse cholesterol transport. J. Lipid Res. 36, 211–228. Frelinger, A. L., III, and Zull, J. E. (1984). Oxidized forms of parathyroid hormone with biological activity. Separation and characterization of hormone forms oxidized at methionine 8 and methionine 18. J. Biol. Chem. 259, 5507–5513. Fuentes‐Prior, P., Iwanaga, Y., Huber, R., Pagila, R., Rumennik, G., Seto, M., Morser, J., Light, D. R., and Bode, W. (2000). Structural basis for the anticoagulant activity of the thrombin‐thrombomodulin complex. Nature 404, 518–525. Gabbita, S. P., Aksenov, M. Y., Lovell, M. A., and Markesbery, W. R. (1999). Decrease in peptide methionine sulfoxide reductase in Alzheimer’s disease brain. J. Neurochem. 73, 1660–1666. Gan, Q. F., Witkop, G. L., Sloane, D. L., Straub, K. M., and Sigal, E. (1995). Identification of a specific methionine in mammalian 15‐lipoxygenase which is oxygenated by the enzyme product 13‐HPODE: Dissociation of sulfoxide formation from self‐inactivation. Biochemistry 34, 7069–7079. Gao, J., Yao, Y., and Squier, T. C. (2001). Oxidatively modified calmodulin binds to the plasma membrane Ca‐ATPase in a nonproductive and conformationally disordered complex. Biophys. J. 80, 1791–1801. Giomarelli, B., Visai, L., Hijazi, K., Rindi, S., Ponzio, M., Iannelli, F., Speziale, P., and Pozzi, G. (2006). Binding of Streptococcus gordonii to extracellular matrix proteins. FEMS Microbiol. Lett. 265, 172–177.

128

Oien and Moskovitz

Gitlin, G., Tsarbopoulos, A., Patel, S. T., Sydor, W., Pramanik, B. N., Jacobs, S., Westreich, L., Mittelman, S., and Bausch, J. N. (1996). Isolation and characterization of a monomethioninesulfoxide variant of interferon alpha‐2b. Pharm. Res. 13, 762–769. Glaser, C. B., and Li, C. H. (1974). Reaction of bovine growth hormone with hydrogen peroxide. Biochemistry 13, 1044–1047. Glaser, C. B., Karic, L., Parmelee, S., Premachandra, B. R., Hinkston, D., and Abrams, W. R. (1987). Studies on the turnover of methionine oxidized alpha‐1‐protease inhibitor in rats. Am. Rev. Respir. Dis. 136, 857–861. Glaser, C. B., Morser, J., Clarke, J. H., Blasko, E., McLean, K., Kuhn, I., Chang, R. J., Lin, J. H., Vilander, L., Andrews, W. H., and Light, D. R. (1992). Oxidation of a specific methionine in thrombomodulin by activated neutrophil products blocks cofactor activity. A potential rapid mechanism for modulation of coagulation. J. Clin. Invest. 90, 2565–2573. Glaser, C. B., Yamin, G., Uversky, V. N., and Fink, A. L. (2005). Methionine oxidation, alpha‐ synuclein and Parkinson’s disease. Biochim. Biophys. Acta 1703, 157–169. Goedert, M., and Spillantini, M. G. (2006). A century of Alzheimer’s disease. Science 314, 777–781. Gorlach, A. (2005). Redox regulation of the coagulation cascade. Antioxid. Redox Signal. 7, 1398–1404. Guallar, V., and Olsen, B. (2006). The role of the heme propionates in heme biochemistry. J. Inorg. Biochem. 100, 755–760. Gustavsson, N., Harndahl, U., Emanuelsson, A., RoepstorV, P., and Sundby, C. (1999). Methionine sulfoxidation of the chloroplast small heat shock protein and conformational changes in the oligomer. Protein Sci. 8, 2506–2512. Gustavsson, N., Kokke, B. P., Harndahl, U., Silow, M., Bechtold, U., Poghosyan, Z., Murphy, D., Boelens, W. C., and Sundby, C. (2002). A peptide methionine sulfoxide reductase highly expressed in photosynthetic tissue in Arabidopsis thaliana can protect the chaperone‐like activity of a chloroplast‐localized small heat shock protein. Plant J. 29, 545–553. Harman, D. (1956). Aging: A theory based on free radical and radiation chemistry. J. Gerontol. 11, 298–300. Harman, D. (1973). Free radical theory of aging. Triangle 12, 153–158. Harndahl, U., Kokke, B. P., Gustavsson, N., Linse, S., Berggren, K., Tjerneld, F., Boelens, W. C., and Sundby, C. (2001). The chaperone‐like activity of a small heat shock protein is lost after sulfoxidation of conserved methionines in a surface‐exposed amphipathic alpha‐helix. Biochim. Biophys. Acta 1545, 227–237. Henning, R. J., and Sawmiller, D. R. (2001). Vasoactive intestinal peptide: Cardiovascular eVects. Cardiovasc. Res. 49, 27–37. Hill, A. F., Joiner, S., Wadsworth, J. D., Sidle, K. C., Bell, J. E., Budka, H., Ironside, J. W., and Collinge, J. (2003). Molecular classification of sporadic Creutzfeldt‐Jakob disease. Brain 126, 1333–1346. Hou, L., Kang, I., Marchant, R. E., and Zagorski, M. G. (2002). Methionine 35 oxidation reduces fibril assembly of the amyloid abeta‐(1–42) peptide of Alzheimer’s disease. J. Biol. Chem. 277, 40173–40176. Houghten, R. A., and Li, C. H. (1976). Studies on pituitary prolactin. 39. Reaction of the ovine hormone with hydrogen peroxide. Biochim. Biophys. Acta 439, 240–249. Houghten, R. A., Glaser, C. B., and Li, C. H. (1977). Human somatotropin. Reaction with hydrogen peroxide. Arch. Biochem. Biophys. 178, 350–355. Hsu, Y. R., Narhi, L. O., Spahr, C., Langley, K. E., and Lu, H. S. (1996). In vitro methionine oxidation of Escherichia coli‐derived human stem cell factor: EVects on the molecular structure, biological activity, and dimerization. Protein Sci. 5, 1165–1173. Ivanetich, K. M., Bradshaw, J. J., and Kaminsky, L. S. (1976). Methionine sulfoxide cytochrome c. Biochemistry 15, 1144–1153.

3. Substrates of Methionine Sulfoxide Reductase

129

Jackson, G. S., Hosszu, L. L., Power, A., Hill, A. F., Kenney, J., Saibil, H., Craven, C. J., Waltho, J. P., Clarke, A. R., and Collinge, J. (1999). Reversible conversion of monomeric human prion protein between native and fibrilogenic conformations. Science 283, 1935–1937. Jacob, C., Knight, I., and Winyard, P. G. (2007). Aspects of the biological redox chemistry of cysteine: From simple redox responses to sophisticated signalling pathways. Biol. Chem. 387, 1385–1397. Jia, Y., Buehler, P., Boykins, R. A., Venable, R. M., and Alayash, A. I. (2007). Structural basis of peroxide mediated changes in human hemoglobin: A novel oxidative pathway. J. Biol. Chem. 282, 4894–4907. Jori, G., Galiazzo, G., Marzotto, A., and ScoVone, E. (1968). Selective and reversibe photo‐ oxidation of the methionyl residues in lysozyme. J. Biol. Chem. 243, 4272–4278. Jung, B., Lee, E. H., Chung, W. S., Lee, S. J., Shin, S. H., Joo, S. H., Kim, S. K., and Lee, J. H. (2003). Increased viability of PC12 cells exposed to amyloid‐beta peptide by transduction with human TAT‐methionine sulfoxide reductase. Neuroreport 14, 2349–2353. Kanayama, A., Inoue, J., Sugita‐Konishi, Y., Shimizu, M., and Miyamoto, Y. (2002). Oxidation of Ikappa Balpha at methionine 45 is one cause of taurine chloramine‐induced inhibition of NF‐kappa B activation. J. Biol. Chem. 277, 24049–24056. Karelson, E., Fernaeus, S., Reis, K., Bogdanovic, N., and Land, T. (2005). Stimulation of G‐proteins in human control and Alzheimer’s disease brain by FAD mutants of APP (714–723): Implication of oxidative mechanisms. J. Neurosci. Res. 79, 368–374. Keck, R. G. (1996). The use of t‐butyl hydroperoxide as a probe for methionine oxidation in proteins. Anal. Biochem. 236, 56–62. Kedzierska, S. (2005). [Role of Escherichia coli molecular chaperones in the protection of bacterial cells against irreversible aggregation induced by heat shock]. Postepy Biochem. 51, 146–153. Khor, H. K., Fisher, M. T., and Schoneich, C. (2004). Potential role of methionine sulfoxide in the inactivation of the chaperone GroEL by hypochlorous acid (HOCl) and peroxynitrite (ONOO). J. Biol. Chem. 279, 19486–19493. Kido, K., and Kassell, B. (1975). Oxidation of methionine residues of porcine and bovine pepsins. Biochemistry 14, 631–635. Kim, H. Y., and Gladyshev, V. N. (2004). Characterization of mouse endoplasmic reticulum methionine‐R‐sulfoxide reductase. Biochem. Biophys. Res. Commun. 320, 1277–1283. Kim, K., and Erman, J. E. (1988). Methionine modification in cytochrome‐c peroxidase. Biochim. Biophys. Acta 954, 95–107. Kornfelt, T., Persson, E., and Palm, L. (1999). Oxidation of methionine residues in coagulation factor VIIa. Arch. Biochem. Biophys. 363, 43–54. Kumar, C. C., Nie, H., Armstrong, L., Zhang, R., Vijay‐Kumar, S., and Tsarbopoulos, A. (1998). Chloramine T‐induced structural and biochemical changes in echistatin. FEBS Lett. 429, 239–248. Law, R. H., Zhang, Q., McGowan, S., Buckle, A. M., Silverman, G. A., Wong, W., Rosado, C. J., Langendorf, C. G., Pike, R. N., Bird, P. I., and Whisstock, J. C. (2006). An overview of the serpin superfamily. Genome Biol. 7, 216.1–216.11. Lawrence, D. A., and LoskutoV, D. J. (1986). Inactivation of plasminogen activator inhibitor by oxidants. Biochemistry 25, 6351–6355. Lesage, F., and Lazdunski, M. (2000). Molecular and functional properties of two‐pore‐domain potassium channels. Am. J. Physiol. Renal Physiol. 279, F793–F801. Lev, N., Roncevich, D., Ickowicz, D., Melamed, E., and OVen, D. (2006). Role of DJ‐1 in Parkinson’s disease. J. Mol. Neurosci. 29, 215–225. Levine, R. L., Mosoni, L., Berlett, B. S., and Stadtman, E. R. (1996). Methionine residues as endogenous antioxidants in proteins. Proc. Natl. Acad. Sci. USA 93, 15036–15040.

130

Oien and Moskovitz

Levine, R. L., Moskovitz, J., and Stadtman, E. R. (2000). Oxidation of methionine in proteins: Roles in antioxidant defense and cellular regulation. IUBMB Life 50, 301–307. McCarthy, J. B., Walker, M., Pierce, J., Camp, P., and White, J. D. (1998). Biosynthesis and metabolism of native and oxidized neuropeptide Y in the hippocampal mossy fiber system. J. Neurochem. 70, 1950–1963. Michaelis, M. L., Bigelow, D. J., Schoneich, C., Williams, T. D., Ramonda, L., Yin, D., Huhmer, A. F., Yao, Y., Gao, J., and Squier, T. C. (1996). Decreased plasma membrane calcium transport activity in aging brain. Life Sci. 59, 405–412. Midwinter, R. G., Cheah, F. C., Moskovitz, J., Vissers, M. C., and Winterbourn, C. C. (2006). IkappaB is a sensitive target for oxidation by cell‐permeable chloramines: Inhibition of NF‐kappaB activity by glycine chloramine through methionine oxidation. Biochem. J. 396, 71–78. Miles, A. M., and Smith, R. L. (1993). Functional methionines in the collagen/gelatin binding domain of plasma fibronectin: EVects of chemical modification by chloramine T. Biochemistry 32, 8168–8178. Moraga, F., and Janciauskiene, S. (2000). Activation of primary human monocytes by the oxidized form of alpha1‐antitrypsin. J. Biol. Chem. 275, 7693–7700. Moreno, J. J. (2000). Antiflammin peptides in the regulation of inflammatory response. Ann. N Y Acad. Sci. 923, 147–153. Moskovitz, J. (2005a). Methionine sulfoxide reductases: Ubiquitous enzymes involved in antioxidant defense, protein regulation, and prevention of aging‐associated diseases. Biochim. Biophys. Acta 1703, 213–219. Moskovitz, J. (2005b). Roles of methionine suldfoxide reductases in antioxidant defense, protein regulation and survival. Curr. Pharm. Des. 11, 1451–1457. Moskovitz, J., and Bush, A. I. (2005). Methionine sulfoxide reductase system: Possible roles in protection against neurodegenerative diseases. In ‘‘Oxidative Stress and Age‐Related Neurodegeneration’’ (Y. Luo and L. Packer, Eds.), pp. 197–210. CRC, New York, NY. Moskovitz, J., Jenkins, N. A., Gilbert, D. J., Copeland, N. G., Jursky, F., Weissbach, H., and Brot, N. (1996). Chromosomal localization of the mammalian peptide‐methionine sulfoxide reductase gene and its diVerential expression in various tissues. Proc. Natl. Acad. Sci. USA 93, 3205–3208. Moskovitz, J., Poston, J. M., Berlett, B. S., Nosworthy, N. J., Szczepanowski, R., and Stadtman, E. R. (2000). Identification and characterization of a putative active site for peptide methionine sulfoxide reductase (MsrA) and its substrate stereospecificity. J. Biol. Chem. 275, 14167–14172. Moskovitz, J., Bar‐Noy, S., Williams, W. M., Requena, J., Berlett, B. S., and Stadtman, E. R. (2001). Methionine sulfoxide reductase (MsrA) is a regulator of antioxidant defense and lifespan in mammals. Proc. Natl. Acad. Sci. USA 98, 12920–12925. Moskovitz, J., Singh, V. K., Requena, J., Wilkinson, B. J., Jayaswal, R. K., and Stadtman, E. R. (2002). Purification and characterization of methionine sulfoxide reductases from mouse and Staphylococcus aureus and their substrate stereospecificity. Biochem. Biophys. Res. Commun. 290, 62–65. Mutt, V. (1981). Additional observations on cholecystokinin and the vasoactive intestinal polypeptide. Peptides 2(Suppl. 2), 209–214. Nabuchi, Y., Fujiwara, E., Ueno, K., Kuboniwa, H., Asoh, Y., and Ushio, H. (1995). Oxidation of recombinant human parathyroid hormone: EVect of oxidized position on the biological activity. Pharm. Res. 12, 2049–2052. Nabuchi, Y., Asoh, Y., and Takayama, M. (2004). Folding analysis of hormonal polypeptide calcitonins and the oxidized calcitonins using electrospray ionization mass spectrometry combined with H/D exchange. J. Am. Soc. Mass Spectrom. 15, 1556–1564. Nishimura, C., Ekida, T., Masuda, S., Futatsugi, K., Itoh, S., Yasukawa, K., Kishimoto, T., and Arata, Y. (1991). Chemical modification and 1H‐NMR studies on the receptor‐binding region of human interleukin 6. Eur. J. Biochem. 196, 377–384.

3. Substrates of Methionine Sulfoxide Reductase

131

Oda, T., and Tokushige, M. (1988). Chemical modification of tryptophanase by chloramine T: A possible involvement of the methionine residue in enzyme activity. J. Biochem. (Tokyo) 104, 178–183. Panzenbock, U., and Stocker, R. (2005). Formation of methionine sulfoxide‐containing specific forms of oxidized high‐density lipoproteins. Biochim. Biophys. Acta 1703, 171–181. Pennathur, S., Bergt, C., Shao, B., Byun, J., Kassim, S. Y., Singh, P., Green, P. S., McDonald, T. O., Brunzell, J., Chait, A., Oram, J. F., O’Brien, K., et al. (2004). Human atherosclerotic intima and blood of patients with established coronary artery disease contain high density lipoprotein damaged by reactive nitrogen species. J. Biol. Chem. 279, 42977–42983. Pennings, M., Meurs, I., Ye, D., Out, R., Hoekstra, M., Van Berkel, T. J., and Van Eck, M. (2006). Regulation of cholesterol homeostasis in macrophages and consequences for atherosclerotic lesion development. FEBS Lett. 580, 5588–5596. Petropoulos, I., Mary, J., Perichon, M., and Friguet, B. (2001). Rat peptide methionine sulphoxide reductase: Cloning of the cDNA, and down‐regulation of gene expression and enzyme activity during aging. Biochem. J. 355, 819–825. Pogocki, D. (2003). Alzheimer’s beta‐amyloid peptide as a source of neurotoxic free radicals: The role of structural eVects. Acta Neurobiol. Exp. (Wars.) 63, 131–145. Rachidi, W., Vilette, D., Guiraud, P., Arlotto, M., Riondel, J., Laude, H., Lehmann, S., and Favier, A. (2003). Expression of prion protein increases cellular copper binding and antioxidant enzyme activities but not copper delivery. J. Biol. Chem. 278, 9064–9072. Rapoport, S., Hartel, B., and Hausdorf, G. (1984). Methionine sulfoxide formation: The cause of self‐inactivation of reticulocyte lipoxygenase. Eur. J. Biochem. 139, 573–576. Ray, W. J., Jr., Latham, H. G., Jr., Katsoulis, M., and Koshland, D. E., Jr. (1960). Evidence for involvement of a methionine residue in the enzymatic action of phosphoglucomutase and chymotrypsin. J. Am. Chem. Soc. 82, 4743–4744. Reddy, V. Y., Desorchers, P. E., Pizzo, S. V., Gonias, S. L., Sahakian, J. A., Levine, R. L., and Weiss, S. J. (1994). Oxidative dissociation of human alpha 2‐macroglobulin tetramers into dysfunctional dimers. J. Biol. Chem. 269, 4683–4691. Reis, K., Zharkovsky, A., Bogdanovic, N., Karelson, E., and Land, T. (2007). Critical role of methionine‐722 in the stimulation of human brain G‐proteins and neurotoxicity induced by London familial Alzheimer’s disease (FAD) mutated V717G‐APP(714–723). Neuroscience 144, 571–578. Rudolphus, A., Heinzel‐Wieland, R., Vincent, V. A., Saunders, D., SteVens, G. J., Dijkman, J. H., and Kramps, J. A. (1991). Oxidation‐resistant variants of recombinant antileucoprotease are better inhibitors of human‐neutrophil‐elastase‐induced emphysema in hamsters than natural recombinant antileucoprotease. Clin. Sci. (Lond.) 81, 777–784. Sagher, D., Brunell, D., Brot, N., Vallee, B. L., and Weissbach, H. (2006a). Selenocompounds can serve as oxidoreductants with the methionine sulfoxide reductase enzymes. J. Biol. Chem. 281, 31184–31187. Sagher, D., Brunell, D., Hejtmancik, J. F., Kantorow, M., Brot, N., and Weissbach, H. (2006b). Thionein can serve as a reducing agent for the methionine sulfoxide reductases. Proc. Natl. Acad. Sci. USA 103, 8656–8661. Santarelli, L. C., Wassef, R., Heinemann, S. H., and Hoshi, T. (2006). Three methionine residues located within the regulator of conductance for Kþ (RCK) domains confer oxidative sensitivity to large‐conductance Ca2þ‐activated Kþ channels. J. Physiol. 571, 329–348. Sattler, W., Maiorino, M., and Stocker, R. (1994). Reduction of HDL‐ and LDL‐associated cholesterylester and phospholipid hydroperoxides by phospholipid hydroperoxide glutathione peroxidase and Ebselen (PZ 51). Arch. Biochem. Biophys. 309, 214–221. Savige, W. E., and Fontana, A. (1977). Interconversion of methionine and methionine sulfoxide. Methods Enzymol. 47, 453–459. Shao, B., Oda, M. N., Bergt, C., Fu, X., Green, P. S., Brot, N., Oram, J. F., and Heinecke, J. W. (2006). Myeloperoxidase impairs ABCA1‐dependent cholesterol eZux through methionine

132

Oien and Moskovitz

oxidation and site‐specific tyrosine chlorination of apolipoprotein A‐I. J. Biol. Chem. 281, 9001–9004. Shechter, Y., Burnstein, Y., and Gertler, A. (1977). EVect of oxidation of methionine residues in chicken ovoinhibitor on its inhibitory activities against trypsin, chymotrypsin, and elastase. Biochemistry 16, 992–997. Sigalov, A. B., and Stern, L. J. (1998). Enzymatic repair of oxidative damage to human apolipoprotein A‐I. FEBS Lett. 433, 196–200. Smith, C. E., Stack, M. S., and Johnson, D. A. (1987). Ozone eVects on inhibitors of human neutrophil proteinases. Arch. Biochem. Biophys. 253, 146–155. Sonne, O., Larsen, U. D., and Markussen, J. (1982). The eVect of oxidation of the Met27 residue of [125I]monoiodoglucagon on receptor‐binding aYnity. Hoppe Seylers Z. Physiol. Chem. 363, 95–101. Stadtman, E. R., Moskovitz, J., Berlett, B. S., and Levine, R. L. (2002). Cyclic oxidation and reduction of protein methionine residues is an important antioxidant mechanism. Mol. Cell. Biochem. 234–235, 3–9. StauVer, C. E., and Etson, D. (1969). The eVect on subtilisin activity of oxidizing a methionine residue. J. Biol. Chem. 244, 5333–5338. Stevens‐Truss, R., and Hinman, C. L. (1996). Chemical modification of methionines in a cobra venom cytotoxin diVerentiates between lytic and binding domains. Toxicol. Appl. Pharmacol. 139, 234–242. Stief, T. W., and Fareed, J. (2000). The antithrombotic factor singlet oxygen/light (1O2/h nu). Clin. Appl. Thromb. Hemost. 6, 22–30. Stief, T. W., Aab, A., and Heimburger, N. (1988). Oxidative inactivation of purified human alpha‐ 2‐antiplasmin, antithrombin III, and C1‐inhibitor. Thromb. Res. 49, 581–589. Stief, T. W., Martin, E., Jimenez, J., Digon, J., and Rodriguez, J. M. (1991). EVect of oxidants on proteases of the fibrinolytic system: Possible role for methionine residues in the interaction between tissue type plasminogen activator and fibrin. Thromb. Res. 61, 191–200. Sun, H., Gao, J., Ferrington, D. A., Biesiada, H., Williams, T. D., and Squier, T. C. (1999). Repair of oxidized calmodulin by methionine sulfoxide reductase restores ability to activate the plasma membrane Ca‐ATPase. Biochemistry 38, 105–112. Sundby, C., Harndahl, U., Gustavsson, N., Ahrman, E., and Murphy, D. J. (2005). Conserved methionines in chloroplasts. Biochim. Biophys. Acta 1703, 191–202. Swaim, M. W., and Pizzo, S. V. (1988). Methionine sulfoxide and the oxidative regulation of plasma proteinase inhibitors. J. Leukoc. Biol. 43, 365–379. Taggart, C., Cervantes‐Laurean, D., Kim, G., McElvaney, N. G., Wehr, N., Moss, J., and Levine, R. L. (2000). Oxidation of either methionine 351 or methionine 358 in alpha 1‐antitrypsin causes loss of anti‐neutrophil elastase activity. J. Biol. Chem. 275, 27258–27265. Tang, X. D., Daggett, H., Hanner, M., Garcia, M. L., McManus, O. B., Brot, N., Weissbach, H., Heinemann, S. H., and Hoshi, T. (2001). Oxidative regulation of large conductance calcium‐ activated potassium channels. J. Gen. Physiol. 117, 253–274. Teh, L. C., Murphy, L. J., Huq, N. L., Surus, A. S., Friesen, H. G., Lazarus, L., and Chapman, G. E. (1987). Methionine oxidation in human growth hormone and human chorionic somatomammotropin. EVects on receptor binding and biological activities. J. Biol. Chem. 262, 6472–6477. Tomb, J. F., White, O., Kerlavage, A. R., Clayton, R. A., Sutton, G. G., Fleischmann, R. D., Ketchum, K. A., Klenk, H. P., Gill, S., Dougherty, B. A., Nelson, K., Quackenbush, J., et al. (1997). The complete genome sequence of the gastric pathogen Helicobacter pylori. Nature 388, 539–547.

3. Substrates of Methionine Sulfoxide Reductase

133

Trauner, D., and Kramer, R. H. (2004). Metabolic modulation of potassium channels. Sci. STKE 2004, pe22. Turkall, R. M., Denison, R. C., and Tsan, M. F. (1982). Degradation and oxidation of methionine enkephalin by human neutrophils. J. Lab. Clin. Med. 99, 418–427. Uversky, V. N., Gillespie, J. R., and Fink, A. L. (2000). Why are ‘‘natively unfolded’’ proteins unstructured under physiologic conditions? Proteins 41, 415–427. Uversky, V. N., Yamin, G., Souillac, P. O., Goers, J., Glaser, C. B., and Fink, A. L. (2002). Methionine oxidation inhibits fibrillation of human alpha‐synuclein in vitro. FEBS Lett. 517, 239–244. Van Patten, S. M., Hanson, E., Bernasconi, R., Zhang, K., Manavalan, P., Cole, E. S., McPherson, J. M., and Edmunds, T. (1999). Oxidation of methionine residues in antithrombin. EVects on biological activity and heparin binding. J. Biol. Chem. 274, 10268–10276. Vigna, S. R., Giraud, A. S., Reeve, J. R., Jr., and Walsh, J. H. (1988). Biological activity of oxidized and reduced iodinated bombesins. Peptides 9, 923–926. Vogt, W., and Hesse, D. (1994). Oxidants generated by the myeloperoxidase‐halide system activate the fifth component of human complement, C5. Immunobiology 192, 1–9. Vogt, W., Zimmermann, B., Hesse, D., and Nolte, R. (1992). Activation of the fifth component of human complement, C5, without cleavage, by methionine oxidizing agents. Mol. Immunol. 29, 251–256. White, C. E., Hunter, M. J., Meininger, D. P., White, L. R., and Komives, E. A. (1995). Large‐ scale expression, purification and characterization of small fragments of thrombomodulin: The roles of the sixth domain and of methionine 388. Protein Eng. 8, 1177–1187. Williams, W. M., Stadtman, E. R., and Moskovitz, J. (2004). Ageing and exposure to oxidative stress in vivo diVerentially aVect cellular levels of PrP in mouse cerebral microvessels and brain parenchyma. Neuropathol. Appl. Neurobiol. 30, 161–168. Wong, B. S., Wang, H., Brown, D. R., and Jones, I. M. (1999). Selective oxidation of methionine residues in prion proteins. Biochem. Biophys. Res. Commun. 259, 352–355. Wong, B. S., Chen, S. G., Colucci, M., Xie, Z., Pan, T., Liu, T., Li, R., Gambetti, P., Sy, M. S., and Brown, D. R. (2001). Aberrant metal binding by prion protein in human prion disease. J. Neurochem. 78, 1400–1408. Wood‐Kaczmar, A., Gandhi, S., and Wood, N. W. (2006). Understanding the molecular causes of Parkinson’s disease. Trends Mol. Med. 12, 521–528. Wood, M. J., Becvar, L. A., Prieto, J. H., Melacini, G., and Komives, E. A. (2003). NMR structures reveal how oxidation inactivates thrombomodulin. Biochemistry 42, 11932–11942. Wood, M. J., Helena Prieto, J., and Komives, E. A. (2005). Structural and functional consequences of methionine oxidation in thrombomodulin. Biochim. Biophys. Acta 1703, 141–147. Yamin, G., Glaser, C. B., Uversky, V. N., and Fink, A. L. (2003). Certain metals trigger fibrillation of methionine‐oxidized alpha‐synuclein. J. Biol. Chem. 278, 27630–27635. Ye, J. M., and Wolfe, J. L. (1996). Oxidative degradation of antiflammin 2. Pharm. Res. 13, 250–255. Zheng, L., Nukuna, B., Brennan, M. L., Sun, M., Goormastic, M., Settle, M., Schmitt, D., Fu, X., Thomson, L., Fox, P. L., Ischiropoulos, H., Smith, J. D., et al. (2004). Apolipoprotein A‐I is a selective target for myeloperoxidase‐catalyzed oxidation and functional impairment in subjects with cardiovascular disease. J. Clin. Invest. 114, 529–541. Zorin, N. A., Zorina, V. N., and Zorina, R. M. (2006). [The role of macroglobulin family proteins in the regulation of inflammation]. Biomed. Khim. 52, 229–238.

4

____________________________________________________________________________

Organic Anion‐Transporting Polypeptides at the Blood–Brain and Blood–Cerebrospinal Fluid Barriers Daniel E. Westholm,*,1 Jon N. Rumbley,{,1 David R. Salo,* Timothy P. Rich,{ and Grant W. Anderson * *College of Pharmacy, University of Minnesota, Duluth, Minnesota 55812 { Chemistry and Biochemistry, University of Minnesota, Duluth, Minnesota 55812 { Department of Medicine, University of Minnesota, Duluth, Minnesota 55812

I. II. III. IV. V. VI. VII. VIII.

Introduction BBB Structure and Function BCSFB Structure and Function The OATP/Oatp Superfamily Molecular Architecture of the Oatp Superfamily Oatp Substrate Structural Features OATP/Oatp Expression and Action at the BBB and BCSFB Specific Oatps/Oatps Expressed at BBB and BCSFB A. Oatp1a1 B. OATP1A2 C. Oatp1a4 D. Oatp1a5 E. Oatp1c1 F. Oatp2a1

IX. PG Metabolism and Oatps X. Oatp‐Mediated Transport of Conjugated Endobiotics XI. Oxidation, Conjugation, and Transport Metabolism of DHEA and Estradiol (E2) in the Brain A. Oxidative Metabolism B. Conjugation Metabolism C. Transport Metabolism XII. Summary Acknowledgments References

Organic anion‐transporting polypeptides (Oatps) are solute carrier family members that exhibit marked evolutionary conservation. Mammalian Oatps exhibit wide tissue expression with an emphasis on expression in barrier cells. In the brain, Oatps are expressed in the blood–brain barrier endothelial cells and blood–cerebrospinal fluid barrier epithelial cells. This expression profile 1

These authors contributed equally to this work.

Current Topics in Developmental Biology, Vol. 80 Copyright 2008, Elsevier Inc. All rights reserved.

135

0070-2153/08 $35.00 DOI: 10.1016/S0070-2153(07)80004-4

136

Westholm et al.

serves to illustrate a central role for Oatps in transporting endo‐ and xenobiotics across brain barrier cells. This chapter will detail the expression patterns and substrate specificities of Oatps expressed in the brain, and will place special emphases on the role of Oatps in prostaglandin synthesis and in the transport of conjugated endobiotics. ß 2008, Elsevier Inc.

I. Introduction The contents of the central nervous system (CNS) must be exquisitely regulated in order to maintain proper functioning of this critical system. The brain has a high metabolic demand, and thus requires constant influx of oxygen and nutrients as well as eZux of waste products (Takano et al., 2006). Tight regulation of this homeostasis is accomplished, in part, by the blood–brain barrier (BBB) and blood–cerebrospinal fluid barrier (BCSFB). Transport across these barriers is often a required operation during waste eZux (e.g., glucuronides), nutrient delivery (e.g., glucose), drug delivery (e.g., phenytoin), and hormonal and other biological signaling [e.g., via thyroxine and prostaglandins (PGs)] in the brain (Adachi et al., 2003; Borst et al., 2006; Potschka and Loscher, 2001; Sugiyama et al., 2003). This chapter will discuss the role of an important group of transporters at the BBB and BCSFB barrier: the superfamily of organic anion‐transporting polypeptides (humans: OATPS; rodents and other animals: Oatps). Special focus will be directed at the role OATPs/Oatps play in PG and metabolite transport at the BBB and BCSFB.

II. BBB Structure and Function The BBB exists as a selectively permeable barrier composed of a vast network of microvascular endothelium (Hawkins and Davis, 2005) (Fig. 1). Brain endothelia are distinguished from peripheral endothelia by minimal pinosytosis, a lack of fenestrations, and the presence of tight junctions (Fenstermacher et al., 1988; Kniesel and Wolburg, 2000; Sedlakova et al., 1999). Although cerebral microvasculatrue endothelia support the features of the BBB, astrocyte foot processes, pericytes, and neurons all surround individual endothelial cells of the capillary bed forming the complete neurovascular unit (Hawkins and Davis, 2005). Tight junctions restrict the entry of most solutes into the brain through junctional adhesion molecules such as claudin and occludin (Hawkins and Davis, 2005). Unlike other tissues where solutes can leave the stroma and enter the parenchyma through spaces between endothelial cells, BBB tight junctions prevent paracellular diVusion. Transport of a solute or ligand in order to enter or leave brain parenchyma requires crossing the BBB endothelial cell. Therefore, solutes or ligands are required to traverse both the luminal and abluminal membranes of polarized

4. OATPs at the BBB and BCSFB

137

Blood−brain barrier Brain capillary Astrocyte endothelial cell Abluminal membrane foot process Tight junction Basement Luminal membrane membrane

Figure 1 Diagram of BBB anatomy, including capillary endothelial cells sealed with tight junctions, a continuous basement membrane, and astrocyte foot processes. Tight junctions limit the paracellular diVusion of solutes from blood into brain parenchyma. Therefore, in order to cross the BBB, solutes must sequentially diVuse or be transported across both endothelial membranes.

BBB endothelial cells, a distance separated by
110–300 nm of endothelial cytoplasm (Pardridge, 2003). At present, it is unknown how ligands traverse the cytoplasmic space between the luminal and abluminal membranes of BBB endothelial cells. Some ligands may passively diVuse, while others may bind carrier molecules that serve as intracellular transporters. As a result of the limitations imposed by the tight junctions at the BBB, solutes must be transported by transcellular diVusion, carrier‐mediated transport, or endocytosis (Wolka et al., 2003). Transcellular diVusion across the BBB is limited by the lipophilicity and hydrogen bonding potential of a particular solute. In general, solutes of low molecular weight, polarizability, and hydrogen bonding potential and high lipophilicity are associated with greater transcellular BBB diVusivity (Habgood et al., 2000). Much less energetic penalty is required for such molecules to cross the hydrophobic domain of the lipid bilayer than for large polar molecules. Endocytosis plays an important role, albeit reduced and modified compared to other tissues, in the delivery of various compounds to the brain parenchyma (Wolka et al., 2003). Receptor‐mediated, adsorptive, and fluid phase endocytosis all contribute significantly to the overall endocytosis at the BBB. One example of an endocytosed compound at the BBB is transferrin (Tf )‐ bound iron. Iron circulates in the blood bound to Tf (two Fe3þ atoms per protein) and is transported through the BBB via receptor‐mediated endocytosis after binding to the Tf receptor (Descamps et al., 1996). After endocytosis, the low pH of the late endosome causes iron to dissociate from Tf and enter

138

Westholm et al.

the cytosol. Other proteins endocytosed at the BBB include insulin and low‐ density lipoproteins and albumin (Dehouck et al., 1994; King and Johnson, 1985; Kumagai et al., 1987). The mechanism of BBB transport with the broadest spectrum of potential substrates is carrier‐mediated transport. Carrier‐mediated transport of solutes is facilitated by a specific substrate–transporter interaction and driven by direct energy conversion (e.g., ATP hydrolysis), concentration gradients, cotransport, or combination thereof. BBB transporters have been identified for amino acids, sugars, nucleosides, monocarboxylic acids, peptides, organic cations, and organic anions (Tamai and Tsuji, 2000). Transport families expressed at the BBB include monocarboxylic acid transporters (MCTs), organic anion transporters (OATs), organic cation transporters (OCTs), and OATPS/Oatps. In addition, multiple members of the ATP‐binding cassette (ABC) superfamily, including P‐glycoprotein and multidrug resistance associated proteins (MRPs), are expressed at the BBB (Girardin, 2006). These eZux transporters serve to export endo‐ and xenobiotics from the intracellular to extracellular milieu. Xenobiotics are substances that are present in an organism, but are not synthesized by the organism, whereas endobiotics are endogenously synthesized compounds. Both drugs and toxins are classified within the xenobiotic group of compounds. The physical barrier properties of the BBB, combined with the expression of various eZux ABC transporters, actively limit the uptake of certain solutes at the BBB. Endo‐ and xenobiotics that are not subjected to overcome, or work with, these eZux transporters are transported across the BBB in a vectoral fashion into or away from the brain parenchyma by OATPs, OCTs, OATs, and MCTs. It is often stated that the BBB in newborns and fetuses is immature or nonfunctional. Saunders et al. (1999) propose that the BBB is quite functional in newborns. Previous studies that found increased permeability in newborns likely disrupted the fragile neonatal BBB integrity with high volumes of injected tracers. In fact, the BBB obtains the ability to exclude proteins from brain at very early developmental time points. Tight junctions are present as soon as the embryonic brain begins to vascularize. Some maturing of the BBB does occur during development. For example, small lipid molecules are more permeable in neonates as compared to adults (Saunders et al., 1999). But some of the diVerences observed between neonatal and adult BBB have less to do with maturity than with divergent environmental requirements in the developing brain.

III. BCSFB Structure and Function The choroid plexuses are leaf‐like organs found in the median wall of each lateral ventricle, as well as the roof of the third and fourth ventricles of the brain (Kusuhara and Sugiyama, 2004; Strazielle and Ghersi‐Egea, 2000).

139

4. OATPs at the BBB and BCSFB Blood-CSF barrier Tight junction

Choroidal epithelial cell

CSF

CSF

Capillary (fenestrated)

Apical membrane

Basolateral membrane

Figure 2 Diagram of BCSFB anatomy consisting of polarized choroidal epithelial cells connected by tight junctions vascularized with a fenestrated capillary bed. The presence of tight junctions between choroidal epithelial cells forces solutes to sequentially cross both the apical (CSF‐facing) and basolateral (blood‐facing) membranes in order to pass from CSF to blood or vice versa.

Epithelial cells of these highly vascularized organs form the BCSFB (Strazielle and Preston, 2003). These cells are polarized, consisting of apical or brush border membranes [cerebrospinal fluid (CSF) facing] and basolateral membranes (blood facing) (Fig. 2). Invaginations within the choroid plexus organ combined with brush border membrane provide a very large surface area exposed to the CSF. Since the penetrating capillaries in choroid plexuses are fenestrated (leaky), plasma components have free access to the basolateral membrane. Like the junctions between endothelial cells of the BBB, choroid plexus epithelial cell junctions are tight. Thus, under most circumstances, there is no paracellular diVusion from the blood to the CSF. Some compounds are excreted from the CSF to the blood, such as drugs and endogenous waste, and while other solutes travel from the blood to CSF. In order to accomplish transport in either of these directionalities, these compounds must sequentially cross the apical and basolateral membranes of the choroidal epithelial cell. The BCSFB forms a dynamic interface between the dense, fenestrated choroidal stroma and the CSF. Choroid plexuses synthesize the majority of CSF, which in turn circulates from the choroid plexuses, into the subarachnoid spaces and then the spinal column. The diverse roles of CSF range from regulation of intracranial volume and buoyancy to buVering extracranial fluid (Strazielle and Ghersi‐Egea, 2000). In addition to synthesizing CSF, choroid plexuses influence the composition of the CSF through the expression of a plethora of transporters. Carrier‐mediated transport processes at the BCSFB involve many of the same families of transporters expressed

140

Westholm et al.

at the BBB. Finally, the choroid plexus plays a major role in the metabolism of endo‐ and xenobiotics. This topic will be expanded on in a later section of this chapter. Similar to the BBB, a functional BCSFB forms during early developmental stages (Saunders et al., 1999). Protein barriers formed by tight junctions between choroidal epithelial cells prevent paracellular diVusion between blood and CSF interfaces. However, high concentrations of plasma‐derived proteins are present in the CSF of the developing brain. These proteins are likely traYcked intracellularly via a tubuloendoplasmic reticulum system connected across choroidal epithelial cells (Saunders et al., 1999). This specialized system of endoplasmic reticulum (ER) disappears in the adult.

IV. The OATP/Oatp Superfamily OATPs/Oatps are involved in both influx and eZux across CNS and other cell barriers. OATPS/Oatps form a diverse and growing superfamily of solute carriers expressed throughout the body (Hagenbuch and Meier, 2003a). These transporters, with 11–12 membrane‐spanning domains, mediate transmembrane transport of various amphipathic organic anions including steroid conjugates, thyroid hormones, prostanoids, bile salts, oligopeptides, drugs, toxins, and other xenobiotics (Hagenbuch and Meier, 2004; Mikkaichi et al., 2004). The transport mechanism of OATPs/Oatps is sodium independent and likely anion exchange, as evidenced by bidirectional transport of bromosulfophthalein. The driving force is currently unknown, but intracellular glutathione is a possible candidate. A 1:1 stoichiometry of tarocholoate/glutathione exchange by Oatp1a1 (Jacquemin et al., 1994; Li et al., 1998) substrate specificity and tissue localization varies widely with diVerent OATP/Oatp members (Cheng et al., 2005; Hagenbuch and Meier, 2004). Some members are expressed ubiquitously and transport numerous substrates; other members are restricted to discrete tissue expression profiles and display a more specialized substrate specificity (Table I) (Fig. 3). Expression of multiple OATPs/Oatps in the liver, kidney, and BBB puts this superfamily of transporters in a powerful position of influence over the ultimate absorption, disposition, and excretion of drugs and endobiotics. OATPs/Oatps have suVered from inconsistent nomenclature and classification. Individual members often have multiple names used by various investigators. For example, in older literature, Oatp1c1 is variously referred as Oatp‐F or Oatp14. Hagenbuch and Meier proposed a new nomenclature system for members of the OATP/Oatp superfamily (Hagenbuch and Meier, 2004). With this new system, OATP/Oatp members are classified in the OATP/solute carrier Oatp (SLCO) superfamily and subdivided into families, subfamilies, and individual genes. In this chapter, we will use the new naming system, even if old

Table I New Gene Symbol

Oatps Expressed in the Brain Barrier Cells and Known Substratesa New Protein Nomenclature

Old Protein Nomenclature

Located

Oatp Isoform

Slco1a1

Oatp1a1

Oatp1, Oatp

Endothelial cell (luminal and abluminal); Choroid plexus (apical)

1a1

Slco1a4

Oatp1a4

Oatp2

Endothelial cells (luminal and abluminal); Choroid plexus (basolateral membrane)

1a4

Slco1a5

Oatp1a5

Oatp3

Endothelial (luminal and abluminal); Choroid plexus (apical membrane)

1a5

Slco1c1

Oatp1c1

Oatp14

Endothelial (luminal and abluminal), chorid plexus (Basolateral membrane) Endothelial (luminal and abluminal)

1c1

SLCO1A2

OATP1A2

OATP‐A, OATP

1A2

Transports (Increasing Km Values) Sulfotaurolithocholate, taurochenodeoxycholate (TCDCA), estradiol‐17 ‐glucuronide (E217 G, estrone‐3‐sulfate (E‐3‐S), taurolithocholic acid sulfate, cortisol, tauroursodeoxycholate (TUDCA), hyodeoxycholic acid, ochratoxin A, fexofenadine, pravastatin, taurocholate, CRC220, temocaprilat, DPDPE, cholate, glycocholate, deltorphin II, enalapril, BSP‐DNP‐SG, S‐dinitrophenyl glutathione, BQ123, ouabain, gadoxetate Digoxin, E217 G, fexofenadine, triiodothyronine (T3), thyroxine (T4), E‐3‐S, TCDCA, TUDCA, dehydroepiandrosterone sulfate (DHEAS), DPDPE, BQ123, taurocholate, pravastatin, glycocholate, cholate, ouabain Glycodeoxycholate, thyroxine (T4), GW4064, oleic acid, glycoursodeoxycholate (GUDCA), glycochenodeoxycholate (GCDCA), taurodeoxycholate, TUDCA, TCDCA, triiodothyronine (T3), ursodeoxycholic acid, bromosulfophthalein (BSP), cholate, dexamethasone, glycocholate, cortisol, taurocholate, prostaglandinE2 (PGE2), E217 G, DHEAS, E‐3‐S, BQ123 Thyroxine (T4), reverse T3 (rT3), troglitazone sulfate (TRO‐S), cerivastatin, E217 G Fexofenadine, triiodothyronine (T3), DHEAS, thyroxine (T4), TUDCA, BSP, microcystin‐LR, N‐methylquinine, N, methylquinidine, saquinavir mesylate (SQV), E‐3‐S, taurocholate, cholate, unoprostone, DPDPE, deltorphin II, ouabain

a An exhaustive literature review was performed to generate a list of known Oatp substrates. The identified substrates were listed in order of increasing Km values. The nomenclature used follows the suggested classifications of Hagenbuch and Meier (2004).

142

Westholm et al. S D

C K

H N- M 2

Extra

T S V Q P V G R P S F K T E Y P S S E E K Q P C C G E L H P L K P K A K G I V G I L F Y F S A C F V C G L A T V S V I V F M I G L Y G V N L A F G T K L L E I L A A I F G E G S M A D P Y Q I L F K V F S G M T V L I E Q T S Y Q S P K I I E Y R R F D E C E V R E N Y D S S T S I P S K S S S S K M N S S W I T K L Y E S V M F I V S S L P V G C P N L L L L L Q R E S G I S S

F L Q I N E K S S T D

I

II

III

X X X

L N V F G I D V Y L K A C S L G L L F F G I G P I A I V Q T V C G I Y F A A N D E S A F D D L Y A I G L P Q I P T E G

IV

D H I T

I

T Q Q Y G P E Q K I S D Y S P K S P Q R K W A Y N V F G T A V V M W I G G G L W L F I L N S Y I L F N P I A A Q T V I G A V S L I I G L S C T L F L A A G S L Y V F G F P V I V Y W P M K Y N L G K P F F C R K L I S V S N L K P L R s S P Q L S F R D E R D A S M N E S M S I S K E A K N S E K G F Q I P I T D Q D Y H D T

V

VI

VII

F L A R E H Y S V P K T G Q Y S V T L G A V D S N E C G L A L F S F L L Y L G V F S G S L L Y K A A G

VIII

C V

V S C N S R C K C S E T K

W

Y T I G N E G C M P E

A C L A G C Q T S N

T C N Y F I I N K G

G I A A S K S G N S R S S G I V R G S G C S R G R C L C Q R Y D K K S D F N N G W V G C K F R P L Q H C M S I F T Y L L D Y G I F L G V L V T V L F I I V S V Y G L I P T S T P A S V I Y L L I L T A G S S L R V A I G L I V G L A P I F L I T G L Y Y K I I L G K L L N L Y A F V R C S S K I K K P Q L H R L S Q F T I E T K K G R P E W R Y T N M P V Q S L T L R H F D Q S K T E T N Y

IX

X

XI

Intra

-COOH

Similar positions Conserved positions Invariable positions

Figure 3 Putative membrane topology of hOATP1C1 showing the amino acid conservation among all human Oatps. The membrane topology was predicted using the PHDhtm program. Sequences highlighted in red are variably predicted to be membrane spanning using several unique predictive algorithms. Mapped onto the topological profile is a multiple sequence alignment of all of the human OATPs using Probcons. This figure was made with Textopo.

protein and gene names are in common usage. Protein names are designated as OATP/Oatp, while gene names are designated as SLCO/slco. Family, subfamily, and individual gene designations are the same for both protein and gene names of an individual member. We will, next, briefly review current thoughts regarding the Oatp substrate structural characteristics required for recognition by Oatp transporters.

V. Molecular Architecture of the Oatp Superfamily Although no high‐resolution structure is available for any member of the Oatp superfamily, significant structural inferences have been made from sequence analysis and homology modeling. One common feature of the

4. OATPs at the BBB and BCSFB

143

Oatps is the presence of 12 putative transmembrane‐spanning ‐helices. It should be noted that this is not an empirically determined topology and predictions using the human OATP1C1 sequence and several predictive algorithms yield between 10 and 12 transmembrane spans. Figure 4 shows one example, using PHDhtm (Rost et al., 2004). The sequences highlighted by the red boxes represent the inconsistently predicted helices. The assignment of 12 transmembrane‐spanning helices seems reasonable by analogy and moderate sequence similarity to the major facilitator superfamily (MFS), two of which have known three‐dimensional structures. Recently, two groups have used homology modeling to construct three‐ dimensional structures of Oatps, using the known MFS structures. Meier‐ Abt et al. (2005) reported the first structural models, using OATP1B3 and OATP2B1. They identified the glycerol‐3‐phosphate transporter and the lactose permease as structural homologues to the Oatps. Both MFS structures, containing 12 transmembrane spans and a single proposed substrate binding site, were used in the comparative modeling. Following sequence alignment (two Oatps and two MFS proteins), the aligned amino acids of the OATP1B3 and OATP2B1 were mapped onto the three‐dimensional structures using the program MODELLER (Sali and Blundell, 1993). Meier‐Abt et al. used the calculated electrostatic surface potential to evaluate the reliability of the structures returned. This analysis showed the expected positive potential within the putative substrate binding site as well as a lack of electrostatic potential (positive or negative) on the helical surfaces predicted to face the lipid bilayer. Of the 12 transmembrane helices modeled, helices 1, 2, 4, 5, 7, 8, 10, and 11 all appear to make some contribution to the substrate pore. Several polar amino acids, conserved in the OATP1 family, were shown lining the pore. A similar result was obtained with the OATP2 family. They identified the conserved Arg181 (OATP1) and His579 (OATP2) as potential family‐specific contributors to the substrate binding site. A number of other conserved amino acids were proposed to play a more structural role, including the conserved glycines and prolines (Fig. 4) which may be important in packing and orienting helices toward the pore. The large hydrophilic domain between helices 9 and 10 was modeled independently, based on its sequence similarity to the Kazal‐type serine proteases. In this model, some of the cysteines, conserved in all Oatps, are proposed to be involved in disulfide bonds. Hanggi et al. (2006) proposed that all of the cysteines are involved in disulfide bonds using a cysteine labeling reagent. The structural comparison to the MFS was extended to include a similar proposed mechanism, an alternating two‐state mechanism of inward‐ and outward‐facing states referred to as a rocker switch type of movement (Abramson et al., 2003, Huang et al., 2003). The structure of the Oatps was consistent with single‐nucleotide polymorphisms reported for OATP1B3 and OATP2B1. In general, polymorphisms in the membrane domains, having eVects on transport and retaining protein expression levels, could be shown

144

Westholm et al.

Figure 4 Tissue distribution of human OATPs. The tissue distribution of human OATPs was analyzed with transcriptome data mined from the National Center for Biotechnology Information (NCBI) UniGene site (http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db¼unigene). Various tissues are represented on the x‐axis and specific human OATP isoforms are listed on the z‐axis. Tissue‐specific OATP mRNA expression levels are represented in transcripts per million on the y‐axis.

to be amino acids facing the pore in the proposed structures. Within this architecture, the amino acids defining the selectivity of the individual Oatps must exist. As more models are built and more diVerences in the substrate recognition site identified, the rules for discrimination will be elucidated, including the Oatps in the BBB that we are reviewing here. In a subsequent paper, Perry et al. (2006) presented a homology model for hOATP1. As above, the MFS protein glycerol‐3‐phosphate transporter was used as a template and MODELLER was used to build the structure. Again, 8 of the 12 helices face the putative substrate pore, and the pore is characterized by a positive electrostatic surface potential. The calculated volume of the ˚ . New mutations were made at Try230, substrate binding site was 830 A Lys431, and Phe438, based on the structure, all proposed to be facing the substrate channel. In each case, the transport of two unique substrates

4. OATPs at the BBB and BCSFB

145

confirmed that these amino acids play a role in the substrate pore and potentially in the substrate selectivity. Favorable comparisons were also made with similar amino acid positions in the OCT1 and OCT2. The Oatp models described here provide a scaVold to interpret the polymorphisms and the site‐specific mutations already characterized. Further, they will guide the construction of future Oatp mutations. Importantly, the rules (amino acid code) defining the substrate specificity of the Oatps can begin to be addressed. This crucial feature of the Oatp superfamily members dictates the tissue‐specific patterns of Oatp expression, including those in the BBB. It also has major implications on the pharmocogenomics of xenobiotic transport to the BBB and other tissues.

VI. Oatp Substrate Structural Features Clearly the disposition of small molecules in the brain is dependent on the specificity of the membrane transporters that transport them. The pathway in the transporter that the small molecules (substrates) must traverse necessarily contains a molecular surface complimentary to the substrate surface. This means that in three‐dimensional space, the transporter should contain a counter charge for electrostatic interactions, a hydrogen bond donor or acceptor to compliment an acceptor or donor, respectively, and nonpolar amino acids to interact with hydrophobic domains in the substrate. For the Oatps, very little is known about the protein determinants for substrate specificity. From sequence alignments, the most obvious conserved feature is that of an arginine, potential counter charge to the organic anions, in the putative transmembrane domain. Figure 4 shows the PHD predicted membrane topology of OATP1C1 (Rost et al., 2004). As indicated above, the number of transmembrane spans can vary from 10 to 12 depending on the method of prediction. Superimposed onto the topological map is the ProbCons alignment of all 11 human Oatps (Do et al., 2005). In this view, the conserved arginine can be seen in putative transmembrane helix 10. Further, helix 2 has a strong bias toward hydrophobic amino acids. Outside of this, the majority of the remaining conserved amino acids in the membrane spans are either prolines or glycines. Although the alignment is comparing Oatps having diVerent tissue distributions and somewhat diVerent substrate specificities, there remains considerable substrate overlap. The expectation for such a comparison is the identification of common rules for substrate discrimination. In this case, little can be discerned and while the homology models described above provide a preliminary three‐ dimensional picture of the Oatps, the method of substrate discrimination is still unclear. This will only be elucidated with complimentary mutational analysis

146

Westholm et al.

and detailed substrate/inhibitor kinetics measurement, including quantitative structure activity relationship (QSAR) analysis. Two groups have compiled three‐dimensional pharmacophore models of Oatp substrates from previously reported transport data. Yarim et al. (2005) carried out comparative molecular field analysis (CoMFA) on 18 Oatp1a5 substrates derived from two primary cell types, Madin‐Darby canine kidney (MDCK) epithelial cells and Xenopus laevis oocytes. Chang et al. (2005) used Catalyst (Accelrys Software, Inc.) to develop pharmacophore models for human OATP1B1 and rat Oatp1a1. In their study, Chang et al. compared the substrates of OATP1B1 and Oatp1a1 independently and with respect to the cell type the original data was obtained. Subsequently, the data from the unique cell types was merged in a meta analysis. Importantly, the combined data yielded more reliable predictive phamacophore models even though the Kms from diVerent cell types was disparate. In the Yarim et al. (2005) study, the substrate aYnities ranged from 4.3 to 417 M or 100‐fold. The highest aYnity substrates for Oatp1a5 were the bile salts (cholate derivatives) and thyroxine. To carry out the three‐dimensional QSAR analysis, the authors used the Genetic Algorithm Similarity Program (GASP) to flexibly align the training set of 18 molecules (Jones et al., 1995). The GASP algorithm resulted in a hypothetical pharmacophore model containing a negatively charged group on one end, an extended hydrophobic domain in the central part of the structure, and hydrogen bond donor at the opposite end. Therefore, as noted by the authors, a substrate containing all three of these properties would be expected to interact more strongly with the Oatp1a5 than one containing only one or two of the features. To correlate the aligned features to the measured Km values, a partial least square regression analysis was performed on the CoMFA fields (Tripos, Inc.) calculated from each aligned structure. This resulted in a three‐component analysis with an r2 value after bootstrapping of 0.935, suggesting good predictability. The overall contributions of steric and electrostatic interactions were 0.486 and 0.514, respectively. The analysis showed the importance of the electron density (negative charge) on one end of the molecule as expected. In addition, steric bulk around or near the negative end appeared to be favorable to binding. Steric bulk also appears to be favorable on the opposite end although the total end‐to‐end distance is limited by steric clash. Electron density on the opposite end of the substrate has a negative aVect on binding. In all, the CoMFA model had reasonable predictive power for a test set of five molecules. Chang et al. (2005) used Catalyst to generate 10 pharmacophore hypotheses based on a collection of conformers (up to 255) for each molecule in the training set. Although their data were derived from several cell types for both OATP1B1 and Oatp1a1, it was the meta‐analysis on the combined of data that resulted in the most reliable models. The combined training set for rat Oatp1a1 contained 26 molecules with a range of Km values from 0.015

4. OATPs at the BBB and BCSFB

147

to 3300 M. The observed versus predicted correlation was r ¼ 0.92. The pharmacophore generated in this case contained two hydrogen bond acceptors and three hydrophobes. Again, the polar groups (hydrogen bond acceptors) were located on the opposite ends of the molecules. This is consistent with the study of Yarim et al. (2005), showing the importance of the characteristic negative charge at one end of the substrate and a second polar group at the opposite end. Although, in their study, Yarim et al. suggest that the second polar group is a hydrogen bond donor. The distribution of hydrophobic determinants in the Oatp1a1 pharmacophore model was also consistent with the Oatp1a5 results of Yarim et al. In fact, there appears to be favorable hydrophobic interactions near one polar end similar to the steric bulk preference near the negative charge for 1a5. These similarities are not surprising, considering the extensive substrate overlap of these two Oatp variants reported in the literature (Hagenbuch and Meier, 2003b) and summarized in the next section. Further, although these two proteins have diVerent tissue distributions (Fig. 4), they share significant sequence homology. Combining data for human OATP1B1 resulted in a training set of 18 molecules, having a range of Km values from 0.0076 to 268 M. Chang et al. (2005) showed that this combined data set also generated a pharmacophore having two hydrogen bond acceptors and three hydrophobes just as the rat Oatp1a1. Unlike the previous examples, the three‐dimensional distribution of the pharmacophore features in OATP1B1 is significantly diVerent. In addition, the model changes drastically with the inclusion or omission of bilirubin, in question due to inconsistencies in the literature. In either case, the orientation of the two polar groups is not significantly diVerent from rat Oatp1a1 or Oatp1a5, but the position of the hydrophobic domains is quite diVerent. In the case of OATP1B1, a protruding hydrophobic domain exists roughly midway between and perpendicular to a line created between the two polar groups. The discrepancy is consistent with the diVerent substrate specificity of OATP1B1 and its limited localization to the liver. In summary, this type of analysis will continue to lead to new insights into the determinants of substrate specificity of the Oatps. The result of such three‐dimensional QSAR analysis suggests the types of amino acids that may be involved in the process of substrate discrimination and their relative orientation. With respect to the currently available data, this includes an arginine or lysine as a counter charge to the anionic charge characteristic of the Oatp substrates, a stabilizing pocket of hydrophobic interactions consisting of hydrophobic or aromatic amino acids, such as leucine, isoleucine, valine, phenylalanine, tyrosine, and tryptophan, and hydrogen bond acceptors or donors such as threonine, serine, glutamate, aspartate, backbone carbonyl, and amide groups. This implies that a reasonable model would bring helix II in proximity to helix X (helix XI if 12 transmembrane spans are assumed) from the conservation and type of amino acids shown in Fig. 4.

148

Westholm et al.

This is, in fact, the case in the reported homology models described above (Meier‐Abt et al., 2005; Perry et al., 2006). We will next turn our attention to a description of the Oatps expressed in the brain barrier cells and the substrates transported by these transporters.

VII. OATP/Oatp Expression and Action at the BBB and BCSFB Currently, there are over 30 mammalian members of the OATP/Oatp superfamily (Hagenbuch and Meier, 2003a). Of these, only a handful are expressed at the BBB and BCSFB. So far, Oatp1a2, Oatp1a4, Oatp1a5, Oatp1c1, and possibly Oatp1a1 have been identified at protein level at both the BBB and/or BCSFB (Kusuhara and Sugiyama, 2005). In addition, previous indirect evidence, in the form of inhibition studies and mRNA expression, suggested a localization of the PG transporter Oatp2a1 at the BBB (Taogoshi et al., 2005). Kis et al. (2006) demonstrated expression of Oatp2a1 at both the BBB and BCSFB. Other OATPs/Oatps, such as Oatp3a1 and Oatp4a1, have a ubiquitous expression pattern and may yet be identified at either the BBB or BCSFB (Hagenbuch and Meier, 2004). However, definitive localization of these transporters at the BBB and BCSFB awaits confirmation through protein detection. This list of OATP/Oatp members maintains a powerful influence over the CNS disposition and eZux of multiple compounds. Several from this group have high aYnities for specific substrates, with significant biological outcomes. For example, Oatp1c1 is a high‐aYnity thyroxine transporter and appears to account for a significant portion of the thyroxine that enters the brain (Tohyama et al., 2004). Thyroxine is a prohormone of triiodothyronine (T3), the active thyroid hormone. Most T3 is synthesized through the action of type II 50 ‐iodothyronine deiodinase (D2) (Anderson et al., 2003). T3 has profound impacts on the timing and initiation of multiple neurological development processes (Anderson, 2001). Oatp1c1‐dependent thyroxine transport accumulates thyroxine in the CNS at levels higher than would be possible with diVusion alone (Chen et al., 2006). Thus, diVerential expression of Oatp1c1 and other Oatp thyroxine transporters may allow an organism to meter CNS thyroxine concentrations throughout development or during times of thyroxine fluctuations. A hallmark of OATP/Oatp expression at the BBB and BCSFB is the asymmetric distribution of diVerent members at the polarized barrier cell membranes. Some members are expressed at both membranes at a certain barrier. For example, Oatp1a4 has been localized to both the luminal and abluminal membranes of the BBB, using immunofluorescence colocalization with von Willebrand factor (vWF) and glial fibrillary acid protein (GFAP)

4. OATPs at the BBB and BCSFB

149

under confocal microscopy (Gao et al., 1999). In contrast, other Oatp/ OATPs are expressed only on one membrane. For instance, Oatp1a5 is expressed exclusively at the brush border membrane of choroid plexus epithelial cells, as evidenced by immunohoistochemical analysis (Ohtsuki et al., 2004). Finally, some members display diVerent membrane expression patterns depending on what tissue they are expressed. For example, Oatp1c1 is expressed on both the luminal and abluminal membrane of BBB endothelial cells, but expressed at the basolateral membrane in choroid plexus epithelial cells (Sugiyama et al., 2003; Tohyama et al., 2004). OATP/Oatp traYcking signals may be diVerentially decoded by various host cells resulting in cell‐ specific membrane localization patterns (Ito et al., 2005; Sugiyama et al., 2003). The precise nature of these signals is unknown, but may include such identifiers as tyrosine motifs for basolateral sorting and N‐linked oliogosaccharides for apical sorting (Ito et al., 2005; Mostov et al., 2000). The specific localization and combined action of diVerent OATPs/Oatps likely have a significant impact on the overall net accumulation of a given substrate in either the plasma or brain parenchyma. Oatp1c1 and Oatp1a4, for example, largely contribute to the eZux of estradiol‐17 ‐glucuronide (E217 G) from the brain across the BBB (Kusuhara and Sugiyama, 2005; Sugiyama et al., 2001, 2003). E217 G is a substrate of both Oatp1a4 and OIatp1c1. Glucuronidation is a form of conjugative metabolism crucial to the xenobiotic detoxification systems of the body and will be covered in greater detail later in the chapter. Endogenous compounds are also glucuronidated, allowing directional transport and excretion of lipophilic compounds such as sterols (Sugiyama et al., 2001). EZux of E217 G out of the brain across the BBB by luminal and abluminal localized Oatp1a4 and Oatp1c1 helps prevent the accumulation of this metabolized hormone in the CNS. Under normal physiological conditions, most OATP/Oatp substrates are bound to protein, usually albumin, due to the hydrophobic component of these molecules (Hagenbuch and Meier, 2003a). It is widely assumed that substrates must first dissociate from the protein carrier to the free state before transport (Tanaka and Mizojiri, 1999). Thyroid hormone, a predominant substrate of Oatp1c1, circulates in the CSF bound to transthyretin, and thyroxine‐binding globulin (TBG) and albumin in the blood (Schussler, 2000). The free drug/hormone hypothesis states that the amount of drug or hormone available for uptake by intact tissues is equal to the amount of free drug or hormone determined under in vitro settings (Koch‐Weser and Sellers, 1976; Pardridge and Landaw, 1984). However, Pardridge showed that dissociation of ligands from proteins in the BBB capillary bed is significantly enhanced (Pardridge and Landaw, 1984). As a result, the amount of free hormone available for uptake is possibly much greater than concentration of free hormone under in vitro settings. Thus, the microenvironment of the BBB endothelial cell may enhance dissociation of thyroid hormone from TBG or

150

Westholm et al.

albumin. Interestingly, Oppenheimer proposed that thyroxine bound to protein in plasma was directly transferred to cellular binding sites (Oppenheimer et al., 1969). He hypothesized that competition between protein binding and cellular binding sites determined the extent of thyroxine transport into diVerent tissues. Members of the OATP/Oatp superfamily, including Oatp1c1, have large extracelluar domains. More research is needed to explore the possibility of these moieties interacting with protein bound ligand and direct transfer of ligand to transporter. Such a speculative model suggests that the enhanced dissociation of thyroid hormone from its binding protein as observed by Pardridge may be explained by direct interactions with extracellular domains of OATPs/Oatps. Future studies may determine the presence or extent of such interactions. OATPs/Oatps also play a significant role in transepithelial thyroxine transport in the choroid plexuses. Large pools of transthyretin (TTR) exist in the CSF. In the brain, TTR is synthesized exclusively by choroid plexus epithelial cells and unidirectionally secreted into the CSF (Cavallaro et al., 1993). The BCSFB is a major site of thyroxine transfer to the CNS, with Oatp1a4 and Oatp1c1 likely playing important roles in transport from the blood across the basolateral membrane of choroidal epithelial cells (Gao et al., 1999; Tohyama et al., 2004). Transport of thyroxine into the CSF across the brush border membrane may occur via membrane bound transporters or through TTR binding and concomitant secretion (Southwell et al., 1993). The large pool of TTR in the CSF is then able to bind and increase the concentration of thyroxine in CSF compared to plasma (Chen et al., 2006). Loss to the blood from CSF is minimal. As a result, concentrating thyroxine in the CSF may increase CNS retention and permit redistribution of the hormone in the brain and CSF (Chen et al., 2006). However, TTR null mice and human TBG mutations do not have functional eVects (Palha et al., 2000). The ontogenic expression of OATPs/Oatps has not been examined comprehensively. No published data exists on the diVerential developmental expression of OATPs/Oatps in the brain. Cheng et al. (2005) examined developmental expression of mouse Oatp1a1 and Oatp1a4 mRNA in kidney and liver. Oatp1a1 was not expressed in kidney in detectable amounts until postnatal (PN) 30. Adult levels were reached by PN 45, at which time male transcripts were expressed significantly higher than females. Oatp1a4 kidney levels did not change significantly during the time period analyzed (until PN 45). In liver, Oatp1a1 mRNA expression was very low until PN 23, at which time expression rapidly increased and reached adult levels by PN 30. At time points where Oatp1a1 mRNA was detectable, male transcripts levels were significantly higher than females. Oatp1a4 expression was low at birth, but increased steadily until PN 23. At this time, Oatp1a4 expression began to drop, with female expression remaining significantly higher than male expression.

4. OATPs at the BBB and BCSFB

151

Oatp1a4 and possibly Oatp1a1 are expressed at the BBB and/or BCSFB. Both of these transporters undergo substantial variation in mRNA expression in other tissues during early developmental time points. It remains to be elucidated whether similar fluctuations in ontogenic expression levels occur in the brain.

VIII. Specific Oatps/Oatps Expressed at BBB and BCSFB A. Oatp1a1 Oatp1a1 mediates the uptake of a wide range of bile salts (e.g., cholate), organic anions (e.g., monoglucuronosyl bilirubin), organic cations (e.g., rocuronium), and drugs (e.g., enalapril) (Hagenbuch and Meier, 2004). Oatp1a1 was the first identified Oatp and was initially cloned from rat liver and translates into a protein with 670 amino acids (Jacquemin et al., 1994). Northern blot analysis reveals mRNA expression in liver, kidney, brain lung, skeletal muscle, and proximal colon (Jacquemin et al., 1994). Oatp1a1 protein is highly expressed in the basolateral membrane of hepatocytes of liver and the apical membrane of the proximal tubule in kidney (Bergwerk et al., 1996). Later, using reverse‐transcriptase protein‐coupled receptors (PCR) total RNA and in situ hybridization, combined with fluorescence confocal microscopy using anti‐Oatp1a1 antibodies, Oatp1a1 expression was detected in choroid plexus, specifically the brush border membrane of epithelial cells (Angeletti et al., 1997). More recent publications have indicated Oatp1a5 is the most highly expressed Oatp member in the choroid plexus (Kusuhara et al., 2003; Ohtsuki et al., 2003). A high degree of homology between the Oatp1a1 and Oatp1a5 proteins has resulted in the creation of antibodies exhibiting cross‐reactivity (Kusuhara and Sugiyama, 2004). Thus, it is unclear whether the Oatp1a1 antibody is actually detecting the related Oatp1a5. As a result, the functional relevance of Oatp1a1 expression at the BCSFB remains unclear.

B. OATP1A2 OATP1A2, the first identified human OATP, was initially cloned from liver in an attempt to isolate the human orthologue to rat Oatp1a1. This 670 amino acid protein exhibits 67% amino acid identity to Oatp1a1, with expression in the luminal membrane of BBB endothelial cells and epithelial cells of liver, kidney, and intestine (Lee et al., 2005). Most OATPs/Oatps identified to date at the BBB and BCSFB are rodent. Thus knowledge of human expression of OATPs at the BBB and BCSFB represents a gap of

152

Westholm et al.

knowledge in the field. Of the currently identified OATP/Oatp superfamily members, OATP1A2 transports some of the most structurally varied compounds, including sterols, neutral and negatively charged peptides, bile salts, and bulky organic cations (Kullak‐Ublick et al., 2001). The wide acceptance of substrates and BBB expression make OATP1A2 of possible importance in determining drug and toxin disposition in the human CNS (Lee et al., 2005). For example, microcystins are cyclic heptapeptide toxins produced by freshwater cyanobacteria (Fischer et al., 2005). Microcystin toxicity manifests in cytoskeleton disruption and subsequent cell death. Target organs are primarily the liver and brain. Fisher et al. identified OATP1A2 as a transporter of variant microcystin‐LR with a Km of 20  8 M. Thus, OATP1A2‐mediated transport of microcystin across the BBB accounts for at least some of the delivery and resulting neurotoxicity associated with microcystin exposure (Fischer et al., 2005). This knowledge may provide therapeutic strategies for treating microcystin intoxication, such as using high doses of known OATP1A2 substrates to out compete microcystin for transporter access. OATP1A2 is also a known transporter of the folate antimetabolite methotrexate (MTX), a drug used to treat certain cancers and autoimmune diseases (Badagnani et al., 2006). Treatment with this chemotherapeutic can lead to, among other side‐eVects, severe CNS toxicity (Vezmar et al., 2003). OATP1A2 endothelial cell expression at the BBB may play a significant role in CNS delivery of this potent drug. Numerous nonsynonymous protein altering variants in multiple ethnicities have been identified in the intracellular, extracellular, and transmembrane domains of OATP1A2 (Badagnani et al., 2006; Lee et al., 2005). The OATP1A2 variants show a range MTX aYnities; hyper‐, hypo‐, and nonfunctional transport capacities have been observed. These genetic diVerences may account for diVering MTX‐induced toxicities and responses across individuals during treatment with this drug. Tumors often become refractory to chemotherapeutics. Bronger and colleagues hypothesized reduced expression of uptake transporters at the BBB or blood–tumor barrier contribute to insuYcient drug delivery to the neoplasms of the brain (Bronger et al., 2005). They examined expression and cellular distribution of ABC eZux transporters and OATPs in frozen glioma samples. ABCC4 and ABCC5 were detected in astrocytic gliomas. OATP1A2 expression was detected in endothelial cells of the BBB and blood–tumor barrier of gliomas, but not in glioma cells (Bronger et al., 2005). In addition, OATP2B1 expression was detected in blood vessels of gliomas and in endothelial cells of perilesional brain samples. Thus, OATP1A2 may be involved with MTX transport across blood–glioma barrier endothelia, but not into the glioma cells. This phenotype, combined with the presence of ABC eZux transporters in glioma cells, may contribute to the multidrug resistance character of gliomas.

4. OATPs at the BBB and BCSFB

153

C. Oatp1a4 The Oatp1a4 gene was isolated from rat brain through homology cloning and encodes a protein of 661 amino acids (Noe et al., 1997). Oatp1a4 is expressed at the luminal and abluminal membranes of BBB endothelial cells, basolateral membrane of choroid plexus epithelial cells as well as in the liver and retina (Gao et al., 1999; Reichel et al., 1999). A wide spectrum of substrates is transported by Oatp1a4 including bile salts, hormones and hormone conjugates, peptides, drugs, and large organic cations (Hagenbuch and Meier, 2003a). The cardiac glycoside digoxin is a high‐aYnity substrate (Km ¼ 0.24 M) of Oatp1a4 (Noe et al., 1997). Digoxin overdose leads to CNS toxicity, of which the pathogenesis is likely coupled to Oatp1a4‐mediated transport of this drug across the BBB (Noe et al., 1997). Oatp1a4 and Oatp1a1 have partial overlapping substrate specificities, and where coexpressed, may perform in tandem to facilitate the high‐aYnity transport of certain cholephilic substances from the liver sinusoidal blood plasma (Reichel et al., 1999). In addition, along with Oatp1a5 and Oatp1c1, Oatp1a4 is among the group of Oatps expressed at the BBB and BCSFB responsible for thyroid hormone uptake into the brain (Abe et al., 1998). [D‐penicillamine‐2,5]‐enkalphin (DPDPE) is an opioid receptor agonist and Oatp1a4 substrate (Gao et al., 2000). This cyclic pentapetide possesses potent antinociceptive eVects, but CNS distribution is limited in part by p‐glycoprotein (Mdr1a) expression at the BBB. Dagenais et al. (2001) performed in vivo transport assays in Mdr1a knockout mice using DPDPE and demonstrated saturable uptake of with a Km value of 24 M. This Km value is similar to the Km of Oatp1a4 (19 M)(Gao et al., 2000). Uptake was inhibited by the Oatp1a4 substrates E217 G and digoxin, supporting a role for Oatp1a4‐mediated brain uptake of DPDPE. Finally, Oatp1a4 has been implicated in E217 G and dehydroepiandrosterone sulfate (DHEAS) eZux across the BBB (Asaba et al., 2000; Sugiyama et al., 2001), a subject covered in greater detail later in this chapter.

D. Oatp1a5 Abe et al. (1998) used homology cloning to isolate Oatp1a5 from rat retina cDNA library. This 670 amino acid protein is expressed in the jejunum and at the brush‐border membrane of choroid plexus epithelial cells comprising the BCSFB (Hagenbuch and Meier, 2004; Kusuhara et al., 2003). Thus Oatp1a5 shows a polar opposite expression pattern at the BCSFB choroid epithelial cells compared to Oatp1c1 and Oatp1a4. Typical substrates of Oatp1a5 include bile salts, steroid conjugates, and thyroid hormones (Hagenbuch and Meier, 2003a).

154

Westholm et al.

Oatp1a4 and Oatp1a5 have a high degree of sequence homology and substrate specificity. The close proximity of these two genes on rat chromosome 4 and mouse chromose 6 suggests they may be paralogs resulting from gene duplication. These closely related transporters have substantial functional overlap. The polarized expression of Oatp1a4 and Oatp1a5 in choroidal epithelial cells could facilitate the vectoral transport of ligands into or out of the CSF. DiVerent driving forces and/or concentrations of cotransported molecules could influence Oatp1a4 and Oatp1a5‐mediated transport of a shared substrate at each unique membrane. However, the greatest limitation in building an accurate model is that the precise molar equivalence of each Oatp expressed in the brain barrier cells is unknown. The answer to this question is of paramount importance. E. Oatp1c1 Oatp1c1 is a 716 amino acid protein expressed in the luminal and abluminal membranes of BBB endothelial cells, the basolateral membrane of choroid plexus epithelial cells, human ciliary body epithelial cells, and Leydig cells of the testes (Gao et al., 2005; Sugiyama et al., 2003; Tohyama et al., 2004). Human OATP1C1 is also localized to, and highly expressed in the brain, but the precise cell types and membranes remain to be elucidated (Pizzagalli et al., 2002). Compared to other Oatps, Oatp1c1 has a relatively narrow range of substrates. However, Oatp1c1 is a high‐aYnity thyroxine transporter and has the lowest identified Km for thyroxine (0.18 M) of all Oatps expressed at the BBB (Sugiyama et al., 2003). Other substrates of Oatp1c1 include 3,30 ,2;50 ‐triiodothyronine (reverse T3), cerivastatin, and E217 G. The role of Oatp1c1 as a thyroxine transporter at the BBB and BCSFB is likely very important for overall thyroid hormone function in the CNS. Other Oatps that transport thyroid hormones, such as Oatp1a4 and Oatp1a5, transport thyroxine and T3 with roughly equal aYnity with Kms in the M range. Thus, Oatp1c1 has a thyroxine transport Km of at least an order of magnitude less than all other Oatp thyroxine transporters. This suggests Oatp1c1 is the most important thyroxine transporter expressed at the BBB and BCSFB. The preference for thyroxine transport of Oatp1c1 may oVer another level of control for thyroid hormone in the brain. Instead of delivery of active thyroid hormone directly to the CNS, transport of thyroxine allows T3 level modulation through control of D2 activity. F. Oatp2a1 Oatp2a1 is a PG transporter expressed at the BBB and BCSFB. Overall, Oatp2a1 is ubiquitously distributed and shares structural homology with other Oatp family members. Oatp2a1 was originally cloned as matrin F/G from

4. OATPs at the BBB and BCSFB

155

rat liver (Hakes and Berezney, 1991). At the time, no significant homologies were found with any other protein. On the basis of structure, localization, and DNA binding, it was hypothesized that matrin F/G had a novel regulatory role within the nucleus; however, no function was demonstrated. Later, Schuster’s demonstration of PG transport in Xenopus oocytes expressing matrin F/G suggested that the protein functioned as a PG transporter and thus the gene was renamed PG transporter (PGT) (Kanai et al., 1996; Schuster, 1998). Finally, subsequent hydropathy and homology analyses identified PGT as a member of the Oatp superfamily and was given the name Oatp2a1. Pharmacological characterization of Oatp2a1 reveals that this transporter exclusively transports PGE2. Additional Oatp family members, Oatp3a1 and Oatp4a1, transport PGs and also other substrates (Mikkaichi et al., 2004). Oatp2a1 expression in BBB endothelial cells and choroidal epithelial cells is predominantly luminal (Kis et al., 2006). Interestingly, in response to lipopolysaccharide (LPS) stimulation this expression pattern disappears and Oatp2a1 localization becomes primarily cytoplasmic. This phenomenon may exert a protective eVect by decreasing PG transport and subsequent fever inducing actions (Kis et al., 2006). PG transport by Oatp2a1 will be discussed in greater detail in a later section.

IX. PG Metabolism and Oatps Cytokine signaling induces cyclooxygenase (COX) to produce PGs in barrier cells. The COX‐1 isotype is responsible for baseline levels of PG, whereas COX‐2 isotype induces increased PG levels in response to cytokine induction (e.g., inflammatory signaling). Both BBB endothelia and choroid plexus epithelia express COX‐2 and synthesize PGs in response to inflammatory stimuli (Lacroix and Rivest, 1998; Mark et al., 2001). COX‐2 is localized on the luminal side of the ER membrane (Fig. 5). Crystallographic studies of COX‐2 have been used to assess the topological arrangement of COX‐2 when associated with ER membrane (Ruan, 2004). In the synthesis of PG, arachodonic acid (AA) passes from the cytoplasm through the ER membrane, binds COX‐2 on the luminal side, and is converted to the PG precursor, PGH2. PGH2 then diVuses back through the ER membrane to the cytoplasmic side to undergo PG subtype synthesis by specific PG synthases (Ruan, 2004; Fig. 5). Specific PG subtypes produced by barrier cells include PGE2 via PGE2 synthase causing fever, pain, and wakefulness; PGD2 via PGD2 synthase causing cooling and sleep; and PGE1 via PGE1 synthase causing vasodilation. Barrier cells produce PGs; however, the site of PG action is extracelluar. Therefore, newly synthesized cytoplasmic PG must be eZuxed from the cell to exert biological action. Barrier cell PG eZux is facilitated by MRPs.

156

Westholm et al. O

Extracelluar

OH H N

O

H

HO

H

COOH

Fenamate Oatp1a4 influx

CH3 H OH

Oatp2a1 influx MRP4 efflux

Oatp1c1 PG-15DH O

H

HO

H

COOH O

Cytoplasm

OH H N

PGE2 inactivation

CH3

PGE2

H OH

?

?

COX

ER lumen

OH

O O

NSAIDs

O OH CH3

COOH

Arachidonic acid

PGH2

CH3

Figure 5 Schematic of PG transport at the BBB. PGH2 is synthesized in the lumen of the ER by COX. PG is transported out of the ER by an unknown mechanism into the cytoplasm. In the cytoplasm, various PG synthases convert PGH2 into active PG subtypes, such as PGE2. MRP4, and to a lesser extent Oatp2a1 and Oatp1a4, contribute to the eZux of active PGs into the extracellular space. Oatp2a1 and Oatp1a4 influx PGs into the cytoplasm, where PG‐15DH catalyzes PG inactivation. Oatp1c1 transports NSAIDs, such as fenamates, into the cytoplasm. Fenamates are transported into the ER lumen where they inhibit the COX enzyme and disrupt PG synthesis.

MRP4 (ABCC4) eZuxes PGs from brain barrier cells. Two groups have demonstrated that MRP4 actively eZuxes PG out of both the BBB endothelia and choroid plexus epithelia (Kis et al., 2006; Reid et al., 2003). It is unlikely that MRP4 is the only transporter eZuxing PGs from barrier cells as bidirectional PGTs (Oatps) also transport PG (see later sections). Once outside the barrier cell membrane, PG exerts paracrine and/or autocrine eVects, or is targeted for enzymatic inactivation. PGs are potent biologic mediators, but exhibit a short half‐life. Therefore, they exert only an autocrine and/or paracrine action (Ashby, 1998). There are currently nine known PG receptors, each G‐protein‐coupled. Receptors for PGE2, PGD2, and PGE1 are located on neurons, on vascular smooth muscle for blood vessel constriction or dilation, and on circulating platelets to regulate aggregation or disaggregation (Ushikubi et al., 1998). It is not clear whether brain

4. OATPs at the BBB and BCSFB

157

endothelial cells themselves express PG receptors. Since a single PG subtype can signal diverse biological functions, depending on the target cell, signal termination is achieved locally to avoid undesired action at a distance (Schuster, 2002). Enzymatically, the first step in signal termination is oxidation of the 15‐hydroxyl group, which renders the PG subtypes unable to bind to their specific receptors. This activity is mediated by PG‐15‐dehydrogenase (PG‐15DH) (Schuster, 1998). Thus, a two‐step model of barrier cell autocrine/paracrine feedback regulation and enzymatic signal termination emerges: (1) selective, carrier‐mediated PG influx across the plasma membrane via PGTs; and (2) nonselective, intracellular oxidation of PG subtypes by PG‐15DH. Barrier cell influx of PG is facilitated primarily by Oatps, specifically Oatp2a1 (formerly PGT) and Oatp1a4. The discovery of Oatp2a1 demonstrated that diVusion alone cannot explain the passage of PG through the cellular membrane (Kanai et al., 1995; Schuster, 1998). Taogoshi et al. (2005) characterized transport of PGE1 across the BBB using an in situ rat brain perfusion technique. The luminal uptake of [3H]PGE1 was inhibited by bromocresol green (Oatp2a1 inhibitor), digoxin, and taurocholate (Oatp1a4 inhibitors) (Taogoshi et al., 2005). RT‐PCR reveals that Oatp2a1 mRNA is expressed in capillary‐rich fractions of the rat brain. Thus, these authors conclude that PGE1 transport across the BBB endothelium is mediated by Oatp family members. However, Oatps mediate both influx and eZux when assessed in vitro, thus it is likely that Oatp2a1 and Oatp1a4 also mediate some eZux of PGs from brain endothelia. Nonsteroidal anti‐inflammatory drugs (NSAIDS) function to inactive both COX‐1 and COX‐2. This class of drugs is transported into the target cell, passes through the cytoplasm, is transported into the ER, and binds to the cyclooxygenase active site thereby inhibiting the synthesis of PGH2 (Fig. 5). We have recently discovered that the fenamate class of NSAIDS are transported by the transporter Oatp1c1 (Westholm and Anderson, 2006). As discussed previously, Oatp1c1 is expressed luminally and abluminally on brain endothelial cells and apically on choroid plexus epithelial cells. We assessed whether isolated brain barrier cells transported this class of NSAIDS. We found that isolated rat brain microvessels demonstrated fenamate uptake. Microvessel uptake was inhibited by the Oatp1c1 substrate thryoxine. Thus, Oatps expressed in brain barrier cells transport not only PG but also drugs that inhibit PG synthesis. NSAIDS bind to the COX active site competing for binding with PG precursor AA. Thus, AA and NSAIDS likely share structural similarities recognized by the COX active site. Similarly, PGs and NSAIDS are both transported by Oatps and likely share structural similarities recognized by the Oatp substrate recognition pocket. This convergence of biological activity may provide guidance in the design and cellular targeting of new COX inhibitors.

158

Westholm et al.

Comparing respective Oatp/MRP PGT kinetics, membrane density of expressed tranporters, luminal versus abluminal, and subcellular localization are important factors for predicting the physiological significance of PG and NSAID traYcking. Already we are recognizing the importance of these pathways. In adults, epidemiologic evidence indicates that NSAID use is associated with a lower incidence risk of Alzheimer disease (AD) (Hendrie, 1997; McGeer, 2000; McGeer et al., 1996; O’Banion, 1999; Pasinetti, 2001). AD exhibits a strong inflammatory component initiated and/or exacerbated by fibrillar‐beta‐sheet beta‐amyloid deposits. Proinflammatory cytokines, PG, and other mediators of inflammation are elevated in and around the senile plaques present in AD brains. Thus, potentiating COX uptake to inhibit PG synthesis represents a therapeutic opportunity.

X. Oatp‐Mediated Transport of Conjugated Endobiotics One of the primary roles of the BBB and BCSFB is to form a biochemical barrier functioning to protect the brain through detoxification enzymes expressed in brain barrier cells. Endogenous chemicals and environmental toxins added selective pressure on the living systems to evolve oxidative and conjugative metabolic pathways with transporting capabilities to neutralize compounds. Thus, Oatps and the Mrps/MRPs work collectively with metabolic enzymes to protect the organism from harmful compounds and modulate levels of endogenous substrates (e.g., steroid hormones) within target tissues like the brain. The pathways of endo‐ and xenobiotic metabolism are divided into three major categories: (1) Oxidative metabolic reactions are accomplished by the cytochrome P450 enzymes and include oxidation, hydroxylation, reduction, and hydrolysis reactions. For each of these enzymatic reactions a new functional group is introduced into the substrate molecule, an existing functional group is modified, or a functional group or acceptor site for conjugation reactions is exposed. (2) Metabolic conjugation reactions are accomplished by syntheses such as UDP‐glucuronosyl transferases (UGTs), glutathione‐S‐tranferases (GSTs), and sulfotransferases (SULTs) whereby a functional group (e.g., alcohol, phenol, amine) is masked by the addition of a new group (e.g., acetyl, sulfate, glucuronic acid moieties) which further increase the polarity of the endo‐ or xenobiotic substrate. Most reactants undergo oxidation and conjugation reactions sequentially. Polar compounds produced by conjugation are more soluble in the aqueous portion of the blood, which facilitates traYcking of the conjugated compound to the bile duct, kidney, or other compartment for excretion out of the body. However, the polarity also limits substrate diVusion across lipid bilayers. Thus, these substrates require membrane spanning transporters, such as the Oatps, to move the conjugates between the blood and target

4. OATPs at the BBB and BCSFB

159

tissues (Strazielle et al., 2004). (3) The passage of substances across biological membranes is referred to as transport metabolism. The transport process can utilize energy (active transport) or may be energy independent (passive transport) (Strazielle et al., 2004). Thus, transporters such as the Oatps are integral partners for the production and elimination of metabolized substrates. For illustrative purposes we will next provide a detailed description of oxidation, conjugation, and transport metabolism for two Oatp substrates, DHEAS, and E217 G transported across brain barrier cells.

XI. Oxidation, Conjugation, and Transport Metabolism of DHEA and Estradiol (E2) in the Brain A. Oxidative Metabolism Cytochrome P450s (CYPs) are members of a superfamily of membrane bound heme‐containing monooxygenases found in the ER of the liver and other extrahepatic tissues including the brain (Hedlund et al., 2001). CYPs are integral membrane proteins imbedded in the ER membrane. The electron components and active site of CYPs are located on the cytoplasmic side of the ER, and a hydrophobic region is responsible for targeting, insertion, and retention in the ER membrane (Neve and Ingelman‐Sundberg, 2000). CYPs function as an electron transport system and can, among other actions, hydroxylate target substrates. Such modified substrates are now targets of conjugation reactions, which increase the aqueous solubility of the molecule for transport and excretion from the body. The endogenous hormones DHEA and estradiol undergo both oxidation and conjugation metabolism and the resultant metabolites are substrates for specific Oatps expressed in the brain barrier cells. These hormones are generated via oxidative metabolism of cholesterol (Hedlund et al., 2001) and are present in the mammalian brain. DHEA functions mainly as a precursor for estrogen and androgen production, but DHEA and DHEAS may also exert independent physiological eVects (Compagnone and Mellon, 2000; Gibbs et al., 2006). For example, DHEAS has been associated with growth of neocortical neurites during brain development (Compagnone and Mellon, 1998). CYP17 and CYP19 participate in the oxidation of cholesterol to synthesize DHEA and E2. Interestingly, CYP17 expression in the brain has been localized to the developing neocortical subplate (Compagnone et al., 1995) suggesting that these developing neurons may participate in the synthesis of brain neurosteroids. CYP19 is even more widely expressed in the mammalian brain (Compagnone and Mellon, 2000) and is most important in the synthesis of E2.

160

Westholm et al.

B. Conjugation Metabolism CYPs incorporate reactive groups into substrates that provide active sites for conjugative modification via SULT and glucuronidating enzymes. DHEA is a direct target of SULT conjugative modification. SULTs modify DHEA to form DHEAS. E2 is the target of conjugative modification by UGTs and is glucuronidated to form E217 G. SULTs sulfonate substrates through transfer of an active sulfate from 30 ‐phosphoadenosine‐50 ‐phosphosulfate (PAPS). Cytosolic SULTs are associated with the sulfation of phenolic steroids and xenobiotics (Wang et al., 2006). The sulfonated steroid conjugate is transported in the blood either to a primary site of action where it may be taken up by a target tissue such as the brain or excreted from the body in the urine or bile (Wang et al., 2006). DHEA and DHEAS are both present in the mammalian brain (Gibbs et al., 2006). Interestingly, expression of the DHEA SULT SULT2A1 has been detected in the rat brain (Gibbs et al., 2006; Shimada et al., 2001). These data suggest that DHEAS is produced in the brain and may exert local eVects. However, DHEAS is also produced in the liver through the actions of SULT1E1. Liver DHEA sulfation targets DHEAS for release into the bile for elimination but also into the circulation for possible uptake by other target tissues (Chen et al., 2005). Steroid sulfatases targeting DHEAS are also expressed in the human brain (Steckelbroeck et al., 2004) and may function to release DHEA for brain neurosteroid synthesis and action. Thus, brain DHEAS may either be synthesized locally within the brain or transported into the brain from the circulating blood. DHEAS transport across the brain barriers via Oatps will be addressed shortly. Glucuronidation is a mechanism to form water‐soluble metabolites targeted for ultimate elimination in the urine and bile. The glucuronidation reaction involves direct condensation of the xenobiotic/endobiotic (or its oxidative product) with the activated form of glucuronic acid, uridine diphosphate glucuronic acid (UDPGA). The reaction between UDPGA and the acceptor’s functional group (alcohol, phenol, or amine) is catalyzed by UGTs. UGTs are a multigene family of isozymes. The UGT active site is located along the luminal side of the ER membrane. UGTs are expressed in hepatocyes, epithelial cells of the intestine, and other extrahepatic tissues including the choroid plexus epithelial cells (King et al., 2000; Strazielle et al., 2004). There are three primary UGT isoforms expressed in the brain: UGT1A6, 2A1, and 2B7. UGT1A6 is also expressed in the liver and is considered the major UGT responsible for T4 glucuronidation (Vansell and Klaassen, 2002). However, the glucuronidation of thyroid hormone has not been assessed in brain barrier cells. Therefore, a functional relationship between UGT1A6 brain expression and Oatp expression in brain barrier cells is not known. Interestingly however, UGT1A6 is expressed in a developmentally regulated fashion in the choroid

4. OATPs at the BBB and BCSFB

161

plexus epithelial cell suggesting possible function in this brain barrier cell during development (Strazielle et al., 2004). UGT2B7 is expressed in human/rat brain, kidney, liver, esophagus, pancreas, and intestine. UGT2B7 substrates include morphine (yielding morphine‐6‐O‐glucuronide and morphine‐3‐O‐glucuronide), NSAIDS (naproxen, ketoprofen, ibuprofen), bile acids, T3, and estrogens. Importantly, UGT2B7 also glucuronidates 17‐ ‐estradiol to the more polar E217 G (King et al., 1999), an Oatp transport substrate (Table I).

C. Transport Metabolism Oatps in the BBB and BCSFB likely play essential roles in transporting conjugated endo‐ and xenobiotics across brain barrier cells (Table I; Fig. 6). Oatps1a1, 1a4, 1a5, and 1c1 are expressed in a polarized fashion on the choroid plexus epithelial cell (Table I). Oatp1a4 and Oatp1c1 are localized to the basolateral side of the epithelial cell, while Oatp1a5 is localized to the apical membrane. Expression of Oatps1a4, 1a5, and 1c1 is also observed in the endothelial cell. Protein is localized both luminally and abluminally for each transporter. Oatps are localized to both membranes of both choroid plexus epithelial and endothelial cells. Therefore, Oatps could theoretically participate in both the influx and eZux of E217 G and DHEAS. However, in the case of E217 G, in vitro and in vivo studies suggest that this substrate is rapidly eliminated from the CSF suggesting preferential eZux of E217 G from the brain. Initial studies suggested that the energy dependent Mrp1 transporter is involved in E217 G transport from the CSF to the blood across the BSCFB (Nishino et al., 1999). The authors noted an energy dependent uptake in vitro using isolated rat choroid plexus and detected Mrp1 expression on the basolateral membrane of the choroid plexus epithelial cell. They also noted rapid elimination of labeled E217 G from the brain into the blood when injected into the cerebral ventricles. The involvement of Mrp1 in vectoral transport of E217 G across brain barrier cells has recently been further supported by use of an Mrp1 null mouse model (Tohyama et al., 2004). In these studies the authors noted a significant decrease in E217 G eZux from the Mrp1 null mouse brain after cerebral microinjection. Importantly however, additional experimental evidence suggests that Oatps also drive the eZux transport of E217 G. This was demonstrated by experiments conducted by Sugiyama and colleagues using the brain eZux index method, in the presence of diVerent inhibitors, to identify the transporters responsible for E217 G eZux across the BBB (Sugiyama et al., 2001). Administration of p‐aminohippurate (PAH), a known inhibitor of the Oat/OAT family reduced E217 G eZux
20%, while administration of the high‐aYnity Oatp1a4 substrate, digoxin, reduced E217 G eZux
40%. Thus, Oatp1a4 is likely at

162

Westholm et al. Cholesterol

Brain

CYP17

CYP17

CYP19

CYP19

DHEA

E2 UGT2B7

SULT

DHEAS

E217bG

Brain Oatps Blood-brain-barrier endothelial cell

Nucleus

Mrp1 Blood (Circulation) DHEAS

E217bG

Figure 6 Oxidative, conjugative, and transport metabolism in the brain. CYPs contribute to the local production of neurosteroids. The neurosteroids are then subjected to conjugative metabolism by SULTs and UGTs to form polar compounds. These compounds are eZuxed out of the brain via Oatps and Mrps, and delivered to the circulation.

least partly responsible for E217 G elimination from the brain in conjunction with Oats/OATs. Importantly, Mrp1 is also expressed in the BBB endothelial cell (Zhang et al., 2000). A specific role for other Oatps in E217 G eZux is not known at this time. These data have together led to the development of a model for vectoral E217 G transport across brain barrier cells (Kusuhara and Sugiyama, 2005). Oatps expressed on the apical membrane of the choroid plexus epithelial cell, such as Oatp1c1 and Oatp1a4, transport E217 G from the CSF into the choroid plexus epithelial cell. Polarized basolateral expression of Mrp1 then allows energy driven transport of E217 G out of the epithelial cell and into the blood. As one of the transporters (Mrp1) uses energy to drive transport, the movement of substrate becomes vectoral. Similar studies have been performed to assess DHEAS traYcking across the brain barriers. As detailed previously, DHEAS is detected in the mammalian brain (Gibbs et al., 2006). The question remains whether DHEAS is actively influxed into the brain or eZuxed out of the brain. The expression of SULT2A1

4. OATPs at the BBB and BCSFB

163

in the brain suggests that DHEAS is synthesized within this tissue thus negating a requirement for active uptake of DHEAS into the brain and supporting a role for DHEAS eZux in controlling DHEAS brain levels. To test this hypothesis Asaba and colleagues (Asaba et al., 2000) utilized the BEI method to assess DHEAS transport across the BBB. They determined that brain eZux was saturable and that eZux transport was inhibited by Oatp substrates including taurocholate and estrone‐3‐sulfate. While they observed measurable DHEAS influx, the eZux rate was over tenfold greater thus suggesting that DHEAS is predominantly transported from the brain to the blood across the BBB. Interestingly, the in vivo Km value for DHEAS eZux from the brain is similar to that of Oatp1a4‐driven DHEAS eZux as measured in vitro (Asaba et al., 2000). The eZux process was also energy dependent suggesting the contributions of an energy‐dependent transporter as detailed above for E217 G. Together these data suggest that DHEAS eZux across the BBB is driven by the combined actions of Oatp and energy‐dependent transporters. Finally, in a study by Azuma and colleagues, intravenous daily administration of 400‐mg DHEAS to demented human subjects for 4 weeks resulted in increased serum and CSF levels of DHEAS and resultant improved mental capacity (Azuma et al., 1999). The increased CSF DHEAS levels suggest the possibility that DHEAS influx from the blood to the brain may also occur. EZux of the conjugated substrates DHEAS and E217 G from the brain targets them for elimination in the urine and bile (Fig. 7). Oatps participate in this elimination process as well and are expressed in the liver and kidney, tissues responsible for elimination of these conjugates. Specifically, brain barrier expressed Oatp1a1/1a4/1a5 protein is expressed on the basolateral membrane of liver hepatocytes and apical membrane of the kidney proximal tubule (Hagenbuch and Meier, 2003a, 2004). The polarized expression of these transporters on the kidney contributes to the directional eZux of E217 G and DHEAS metabolites out of the body and through the bile and urine (Chen et al., 2005). Figures 6 and 7 detail the suggested pathways for conjugated metabolite flux in the body.

XII. Summary The OATP/Oatp superfamily of solute carriers plays a pivotal role in the disposition and flux of multiple endo‐ and xenobiotics in the CNS. Polarized expression of multiple OATPs/Oatps at the BBB and BCSFB contributes to the net directional movement of these compounds. The exact stoichiometric contributions, structural determinants, and developmental expression patterns of individual OATPs/Oatps are currently undetermined. However, the potential for strong physiologic impacts in the CNS mediated by OATPs/ Oatps is clear. Further study, including the generation of Oatp knockout

164

Westholm et al. Circulation DHEAS

Liver

E217bG

Kidney

E217bG

DHEAS

E217bG

DHEAS

Basolateral membrane Proximal tubule

Hepatocyte

Apical membrane

E217bG

DHEAS

Bile

DHEAS

Oatp

E217bG

Urine

Figure 7 Xenobiotic metabolites circulating in the blood are eZuxed from the body into the bile or urine via Oatps and Mrps.

mice, will continue to build on the current knowledge of the role of OATPs/ Oatps at the BBB and BCSFB while possibly revealing novel functions of this versatile superfamily of transporters.

Acknowledgments This work was supported, in part, by grants from NIH R01 DK054060 and the University of Minnesota Graduate School (GWA), NIH T32HL07741 (TPR), Research Corporation CC6681 (JNR), and a Melendy Summer Research Scholarship (DRS). Figures 1, 2, 5, 6, and 7 were created using ScienceSlides 2006 software from Visiscience. The authors would like to thank Amber Seys and Anna Malin for their contributions toward the data gathered in Table I.

References Abe, T., Kakyo, M., Sakagami, H., Tokui, T., Nishio, T., Tanemoto, M., Nomura, H., Hebert, S. C., Matsuno, S., Kondo, H., and Yawo, H. (1998). Molecular characterization and tissue distribution of a new organic anion transporter subtype (oatp3) that transports thyroid hormones and taurocholate and comparison with oatp2. J. Biol. Chem. 273, 22395–22401.

4. OATPs at the BBB and BCSFB

165

Abramson, J., Smirnova, I., Kasho, V., Verner, G., Kaback, H. R., and Iwata, S. (2003). Structure and mechanism of the lactose permease of Escherichia coli. Science 301, 610–615. Adachi, H., Suzuki, T., Abe, M., Asano, N., Mizutamari, H., Tanemoto, M., Nishio, T., Onogawa, T., Toyohara, T., Kasai, S., Satoh, F., Suzuki, M., et al. (2003). Molecular characterization of human and rat organic anion transporter OATP‐D. Am. J. Physiol. Renal Physiol. 285, F1188–F1197. Anderson, G. W. (2001). Thyroid hormones and the brain. Front. Neuroendocrinol. 22, 1–17. Anderson, G. W., Schoonover, C. M., and Jones, S. A. (2003). Control of thyroid hormone action in the developing rat brain. Thyroid 13, 1039–1056. Angeletti, R. H., NovikoV, P. M., Juvvadi, S. R., Fritschy, J. M., Meier, P. J., and WolkoV, A. W. (1997). The choroid plexus epithelium is the site of the organic anion transport protein in the brain. Proc. Natl. Acad. Sci. USA 94, 283–286. Asaba, H., Hosoya, K., Takanaga, H., Ohtsuki, S., Tamura, E., Takizawa, T., and Terasaki, T. (2000). Blood‐brain barrier is involved in the eZux transport of a neuroactive steroid, dehydroepiandrosterone sulfate, via organic anion transporting polypeptide 2. J. Neurochem. 75, 1907–1916. Ashby, B. (1998). Co‐expression of prostaglandin receptors with opposite effects: A model for homeostatic control of autocrine and paracrine signaling. Biochem. Pharmacol. 55(3), 239–246. Azuma, T., Nagai, Y., Saito, T., Funauchi, M., Matsubara, T., and Sakoda, S. (1999). The eVect of dehydroepiandrosterone sulfate administration to patients with multi‐infarct dementia. J. Neurol. Sci. 162, 69–73. Badagnani, I., Castro, R. A., Taylor, T. R., Brett, C. M., Huang, C. C., Stryke, D., Kawamoto, M., Johns, S. J., Ferrin, T. E., Carlson, E. J., Burchard, E. G., and Giacomini, K. M. (2006). Interaction of methotrexate with organic‐anion transporting polypeptide 1A2 and its genetic variants. J. Pharmacol. Exp. Ther. 318, 521–529. Bergwerk, A. J., Shi, X., Ford, A. C., Kanai, N., Jacquemin, E., Burk, R. D., Bai, S., NovikoV, P. M., Stieger, B., Meier, P. J., Schuster, V. L., and WolkoV, A. W. (1996). Immunologic distribution of an organic anion transport protein in rat liver and kidney. Am. J. Physiol. 271, G231–G238. Borst, P., Zelcer, N., and van de Wetering, K. (2006). MRP2 and 3 in health and disease. Cancer Lett. 234, 51–61. Bronger, H., Konig, J., Kopplow, K., Steiner, H. H., Ahmadi, R., Herold‐Mende, C., Keppler, D., and Nies, A. T. (2005). ABCC drug eZux pumps and organic anion uptake transporters in human gliomas and the blood‐tumor barrier. Cancer Res. 65, 11419–11428. Cavallaro, T., Martone, R. L., Stylianopoulou, F., and Herbert, J. (1993). DiVerential expression of the insulin‐like growth factor II and transthyretin genes in the developing rat choroid plexus. J. Neuropathol. Exp. Neurol. 52, 153–162. Chang, C., Pang, K. S., Swaan, P. W., and Ekins, S. (2005). Comparative pharmacophore modeling of organic anion transporting polypeptides: A meta‐analysis of rat Oatp1a1 and human OATP1B1. J. Pharmacol. Exp. Ther. 314, 533–541. Chen, R. L., Kassem, N. A., and Preston, J. E. (2006). Dose‐dependent transthyretin inhibition of T4 uptake from cerebrospinal fluid in sheep. Neurosci. Lett. 396, 7–11. Chen, Z. S., Guo, Y., Belinsky, M. G., Kotova, E., and Kruh, G. D. (2005). Transport of bile acids, sulfated steroids, estradiol 17‐beta‐D‐glucuronide, and leukotriene C4 by human multidrug resistance protein 8 (ABCC11). Mol. Pharmacol. 67, 545–557. Cheng, X., Maher, J., Chen, C., and Klaassen, C. D. (2005). Tissue distribution and ontogeny of mouse organic anion transporting polypeptides (Oatps). Drug Metab. Dispos. 33, 1062–1073. Compagnone, N. A., Bulfone, A., Rubenstein, J. L., and Mellon, S. H. (1995). Steroidogenic enzyme P450c17 is expressed in the embryonic central nervous system. Endocrinology 136, 5212–5223.

166

Westholm et al.

Compagnone, N. A., and Mellon, S. H. (1998). Dehydroepiandrosterone: A potential signalling molecule for neocortical organization during development. Proc. Natl. Acad. Sci. USA 95, 4678–4683. Compagnone, N. A., and Mellon, S. H. (2000). Neurosteroids: Biosynthesis and function of these novel neuromodulators. Front. Neuroendocrinol. 21, 1–56. Dagenais, C., Ducharme, J., and Pollack, G. M. (2001). Uptake and eZux of the peptidic delta‐ opioid receptor agonist. Neurosci. Lett. 301, 155–158. Dehouck, B., Dehouck, M. P., Fruchart, J. C., and Cecchelli, R. (1994). Upregulation of the low density lipoprotein receptor at the blood‐brain barrier: Intercommunications between brain capillary endothelial cells and astrocytes. J. Cell Biol. 126, 465–473. Descamps, L., Dehouck, M. P., Torpier, G., and Cecchelli, R. (1996). Receptor‐mediated transcytosis of transferrin through blood‐brain barrier endothelial cells. Am. J. Physiol. 270, H1149–H1158. Do, C. B., Mahabhashyam, M. S., Brudno, M., and Batzoglou, S. (2005). ProbCons: Probabilistic consistency‐based multiple sequence alignment. Genome Res. 15, 330–340. Fenstermacher, J., Gross, P., Sposito, N., AcuV, V., Pettersen, S., and Gruber, K. (1988). Structural and functional variations in capillary systems within the brain. Ann. N Y Acad. Sci. 529, 21–30. Fischer, W. J., Altheimer, S., Cattori, V., Meier, P. J., Dietrich, D. R., and Hagenbuch, B. (2005). Organic anion transporting polypeptides expressed in liver and brain mediate uptake of microcystin. Toxicol. Appl. Pharmacol. 203, 257–263. Gao, B., Stieger, B., Noe, B., Fritschy, J. M., and Meier, P. J. (1999). Localization of the organic anion transporting polypeptide 2 (Oatp2) in capillary endothelium and choroid plexus epithelium of rat brain. J. Histochem. Cytochem. 47, 1255–1264. Gao, B., Hagenbuch, B., Kullak‐Ublick, G. A., Benke, D., Aguzzi, A., and Meier, P. J. (2000). Organic anion‐transporting polypeptides mediate transport of opioid peptides across blood‐ brain barrier. J. Pharmacol. Exp. Ther. 294, 73–79. Gao, B., Huber, R. D., Wenzel, A., Vavricka, S. R., Ismair, M. G., Reme, C., and Meier, P. J. (2005). Localization of organic anion transporting polypeptides in the rat and human ciliary body epithelium. Exp. Eye Res. 80, 61–72. Gibbs, T. T., Russek, S. J., and Farb, D. H. (2006). Sulfated steroids as endogenous neuromodulators. Pharmacol. Biochem. Behav. 84, 555–567. Girardin, F. (2006). Membrane transporter proteins: A challenge for CNS drug development. Dialogues Clin. Neurosci. 8, 311–321. Habgood, M. D., Begley, D. J., and Abbott, N. J. (2000). Determinants of passive drug entry into the central nervous system. Cell. Mol. Neurobiol. 20, 231–253. Hagenbuch, B., and Meier, P. J. (2003a). The superfamily of organic anion transporting polypeptides. Biochim. Biophys. Acta 1609, 1–18. Hagenbuch, B., and Meier, P. J. (2003b). The superfamily of organic anion transporting polypeptides. Biochim. Biophys. Acta 1609, 1–18. Hagenbuch, B., and Meier, P. J. (2004). Organic anion transporting polypeptides of the OATP/ SLC21 family: Phylogenetic classification as OATP/ SLCO superfamily, new nomenclature and molecular/functional properties. Pflugers Arch. 447, 653–665. Hakes, D. J., and Berezney, R. (1991). Molecular cloning of matrin F/G: A DNA binding protein of the nuclear matrix that contains putative zinc finger motifs. Proc. Natl. Acad. Sci. USA 88, 6186–6190. Hanggi, E., Grundschober, A. F., Leuthold, S., Meier, P. J., and St‐Pierre, M. V. (2006). Functional analysis of the extracellular cysteine residues in the human organic anion transporting polypeptide, OATP2B1. Mol. Pharmacol. 70, 806–817. Hawkins, B. T., and Davis, T. P. (2005). The blood‐brain barrier/neurovascular unit in health and disease. Pharmacol. Rev. 57, 173–185.

4. OATPs at the BBB and BCSFB

167

Hedlund, E., Gustafsson, J. A., and Warner, M. (2001). Cytochrome P450 in the brain; a review. Curr. Drug Metab. 2, 245–263. Hendrie, H. C. (1997). Epidemiology of Alzheimer’s disease. Geriatrics 52(Suppl. 2), S4–S8. Huang, Y., Lemieux, M. J., Song, J., Auer, M., and Wang, D. N. (2003). Structure and mechanism of the glycerol‐3‐phosphate transporter from Escherichia coli. Science 301, 616–620. Ito, K., Suzuki, H., Horie, T., and Sugiyama, Y. (2005). Apical/basolateral surface expression of drug transporters and its role in vectorial drug transport. Pharm. Res. 22, 1559–1577. Jacquemin, E., Hagenbuch, B., Stieger, B., WolkoV, A. W., and Meier, P. J. (1994). Expression cloning of a rat liver Na(þ)‐independent organic anion transporter. Proc. Natl. Acad. Sci. USA 91, 133–137. Jones, G., Willett, P., and Glen, R. C. (1995). A genetic algorithm for flexible molecular overlay and pharmacophore elucidation. J. Comput. Aided Mol. Des. 9, 532–549. Kanai, N., Lu, R., Bao, Y., WolkoV, A. W., Vore, M., and Schuster, V. L. (1996). Estradiol 17 beta‐D‐glucuronide is a high‐aYnity substrate for oatp organic anion transporter. Am. J. Physiol. 270, F326–F331. Kanai, N., Lu, R., Satriano, J. A., Bao, Y., WolkoV, A. W., and Schuster, V. L. (1995). Identification and characterization of a prostaglandin transporter. Science 268, 866–869. King, C. D., Rios, G. R., Assouline, J. A., and Tephly, T. R. (1999). Expression of UDP‐ glucuronosyltransferases (UGTs) 2B7 and 1A6 in the human brain and identification of 5‐hydroxytryptamine as a substrate. Arch. Biochem. Biophys. 365, 156–162. King, C. D., Rios, G. R., Green, M. D., and Tephly, T. R. (2000). UDP‐glucuronosyltransferases. Curr. Drug Metab. 1, 143–161. King, G. L., and Johnson, S. M. (1985). Receptor‐mediated transport of insulin across endothelial cells. Science 227, 1583–1586. Kis, B., Isse, T., Snipes, J. A., Chen, L., Yamashita, H., Ueta, Y., and Busija, D. W. (2006). EVects of LPS stimulation on the expression of prostaglandin carriers in the cells of the blood‐ brain and blood‐cerebrospinal fluid barriers. J. Appl. Physiol. 100, 1392–1399. Kniesel, U., and Wolburg, H. (2000). Tight junctions of the blood‐brain barrier. Cell. Mol. Neurobiol. 20, 57–76. Koch‐Weser, J., and Sellers, E. M. (1976). Drug therapy. Binding of drugs to serum albumin (second of two parts). N. Engl. J. Med. 294, 526–531. Kullak‐Ublick, G. A., Ismair, M. G., Stieger, B., Landmann, L., Huber, R., Pizzagalli, F., Fattinger, K., Meier, P. J., and Hagenbuch, B. (2001). Organic anion‐transporting polypeptide B (OATP‐B) and its functional comparison with three other OATPs of human liver. Gastroenterology 120, 525–533. Kumagai, A. K., Eisenberg, J. B., and Pardridge, W. M. (1987). Absorptive‐mediated endocytosis of cationized albumin and a beta‐endorphin‐cationized albumin chimeric peptide by isolated brain capillaries. Model system of blood‐brain barrier transport. J. Biol. Chem. 262, 15214–15219. Kusuhara, H., and Sugiyama, Y. (2004). EZux transport systems for organic anions and cations at the blood‐CSF barrier. Adv. Drug Deliv. Rev. 56, 1741–1763. Kusuhara, H., and Sugiyama, Y. (2005). Active eZux across the blood‐brain barrier: Role of the solute carrier family. NeuroRx 2, 73–85. Kusuhara, H., He, Z., Nagata, Y., Nozaki, Y., Ito, T., Masuda, H., Meier, P. J., Abe, T., and Sugiyama, Y. (2003). Expression and functional involvement of organic anion transporting polypeptide subtype 3 (Slc21a7) in rat choroid plexus. Pharm. Res. 20, 720–727. Lacroix, S., and Rivest, S. (1998). EVect of acute systemic inflammatory response and cytokines on the transcription of the genes encoding cyclooxygenase enzymes (COX‐1 and COX‐2) in the rat brain. J. Neurochem. 70, 452–466.

168

Westholm et al.

Lee, W., Glaeser, H., Smith, L. H., Roberts, R. L., Moeckel, G. W., Gervasini, G., Leake, B. F., and Kim, R. B. (2005). Polymorphisms in human organic anion‐transporting polypeptide 1A2 (OATP1A2): Implications for altered drug disposition and central nervous system drug entry. J. Biol. Chem. 280, 9610–9617. Li, L., Lee, T. K., Meier, P. J., and Ballatori, N. (1998). Identification of glutathione as a driving force and leukotriene C4 as a substrate for oatp1, the hepatic sinusoidal organic solute transporter. J. Biol. Chem. 273, 16184–16191. Mark, K. S., Trickler, W. J., and Miller, D. W. (2001). Tumor necrosis factor‐alpha induces cyclooxygenase‐2 expression and prostaglandin release in brain microvessel endothelial cells. J. Pharmacol. Exp. Ther. 297, 1051–1058. McGeer, P. L. (2000). Cyclo‐oxygenase‐2 inhibitors: Rationale and therapeutic potential for Alzheimer’s disease. Drugs Aging 17, 1–11. McGeer, P. L., Schulzer, M., and McGeer, E. G. (1996). Arthritis and anti‐inflammatory agents as possible protective factors for Alzheimer’s disease: A review of 17 epidemiologic studies. Neurology 47, 425–432. Meier‐Abt, F., Mokrab, Y., and Mizuguchi, K. (2005). Organic anion transporting polypeptides of the OATP/SLCO superfamily: Identification of new members in nonmammalian species, comparative modeling and a potential transport mode. J. Membr. Biol. 208, 213–227. Mikkaichi, T., Suzuki, T., Tanemoto, M., Ito, S., and Abe, T. (2004). The organic anion transporter (OATP) family. Drug Metab. Pharmacokinet. 19, 171–179. Mostov, K. E., Verges, M., and Altschuler, Y. (2000). Membrane traYc in polarized epithelial cells. Curr. Opin. Cell Biol. 12, 483–490. Neve, E. P., and Ingelman‐Sundberg, M. (2000). Molecular basis for the transport of cytochrome P450 2E1 to the plasma membrane. J. Biol. Chem. 275, 17130–17135. Nishino, J., Suzuki, H., Sugiyama, D., Kitazawa, T., Ito, K., Hanano, M., and Sugiyama, Y. (1999). Transepithelial transport of organic anions across the choroid plexus: Possible involvement of organic anion transporter and multidrug resistance‐associated protein. J. Pharmacol. Exp. Ther. 290, 289–294. Noe, B., Hagenbuch, B., Stieger, B., and Meier, P. J. (1997). Isolation of a multispecific organic anion and cardiac glycoside transporter from rat brain. Proc. Natl. Acad. Sci. USA 94, 10346–10350. O’Banion, M. K. (1999). COX‐2 and Alzheimer’s disease: Potential roles in inflammation and neurodegeneration. Expert Opin. Investig. Drugs 8, 1521–1536. Ohtsuki, S., Takizawa, T., Takanaga, H., Terasaki, N., Kitazawa, T., Sasaki, M., Abe, T., Hosoya, K., and Terasaki, T. (2003). In vitro study of the functional expression of organic anion transporting polypeptide 3 at rat choroid plexus epithelial cells and its involvement in the cerebrospinal fluid‐to‐blood transport of estrone‐3‐sulfate. Mol. Pharmacol. 63, 532–537. Ohtsuki, S., Takizawa, T., Takanaga, H., Hori, S., Hosoya, K., and Terasaki, T. (2004). Localization of organic anion transporting polypeptide 3 (oatp3) in mouse brain parenchymal and capillary endothelial cells. J. Neurochem. 90, 743–749. Oppenheimer, J. H., Surks, M. I., and Schwartz, H. L. (1969). The metabolic significance of exchangeable cellular thyroxine. Recent Prog. Horm. Res. 25, 381–422. Palha, J. A., Fernandes, R., de Escobar, G. M., Episkopou, V., Gottesman, M., and Saraiva, M. J. (2000). Transthyretin regulates thyroid hormone levels in the choroid plexus, but not in the brain parenchyma: Study in a transthyretin‐null mouse model. Endocrinology 141, 3267–3272. Pardridge, W. M. (2003). Molecular biology of the blood‐brain barrier. Methods Mol. Med. 89, 385–399. Pardridge, W. M., and Landaw, E. M. (1984). Tracer kinetic model of blood‐brain barrier transport of plasma protein‐bound ligands. Empiric testing of the free hormone hypothesis. J. Clin. Invest. 74, 745–752.

4. OATPs at the BBB and BCSFB

169

Pasinetti, G. M. (2001). Cyclooxygenase and Alzheimer’s disease: Implications for preventive initiatives to slow the progression of clinical dementia. Arch. Gerontol. Geriatr. 33, 13–28. Perry, J. L., Dembla‐Rajpal, N., Hall, L. A., and Pritchard, J. B. (2006). A three‐dimensional model of human organic anion transporter 1: Aromatic amino acids required for substrate transport. J. Biol. Chem. 281, 38071–38079. Pizzagalli, F., Hagenbuch, B., Stieger, B., Klenk, U., Folkers, G., and Meier, P. J. (2002). Identification of a novel human organic anion transporting polypeptide as a high aYnity thyroxine transporter. Mol. Endocrinol. 16, 2283–2296. Potschka, H., and Loscher, W. (2001). In vivo evidence for P‐glycoprotein‐mediated transport of phenytoin at the blood‐brain barrier of rats. Epilepsia 42, 1231–1240. Reichel, C., Gao, B., Van Montfoort, J., Cattori, V., Rahner, C., Hagenbuch, B., Stieger, B., Kamisako, T., and Meier, P. J. (1999). Localization and function of the organic anion‐ transporting polypeptide Oatp2 in rat liver. Gastroenterology 117, 688–695. Reid, G., Wielinga, P., Zelcer, N., van der Heijden, I., Kuil, A., de Haas, M., Wijnholds, J., and Borst, P. (2003). The human multidrug resistance protein MRP4 functions as a prostaglandin eZux transporter and is inhibited by nonsteroidal antiinflammatory drugs. Proc. Natl. Acad. Sci. USA 100, 9244–9249. Rost, B., Yachdav, G., and Liu, J. (2004). The PredictProtein server. Nucleic Acids Res. 32, W321–W326. Ruan, K. H. (2004). Advance in understanding the biosynthesis of prostacyclin and thromboxane A2 in the endoplasmic reticulum membrane via the cyclooxygenase pathway. Mini Rev. Med. Chem. 4, 639–647. Sali, A., and Blundell, T. L. (1993). Comparative protein modelling by satisfaction of spatial restraints. J. Mol. Biol. 234, 779–815. Saunders, N. R., Habgood, M. D., and Dziegielewska, K. M. (1999). Barrier mechanisms in the brain, II. Immature brain. Clin. Exp. Pharmacol. Physiol. 26, 85–91. Schussler, G. C. (2000). The thyroxine‐binding proteins. Thyroid 10, 141–149. Schuster, V. L. (1998). Molecular mechanisms of prostaglandin transport. Annu. Rev. Physiol. 60, 221–242. Schuster, V. L. (2002). Prostaglandin transport. Prostaglandins Other Lipid Mediat. 68–69, 633–647. Sedlakova, R., Shivers, R. R., and Del Maestro, R. F. (1999). Ultrastructure of the blood‐brain barrier in the rabbit. J. Submicrosc. Cytol. Pathol. 31, 149–161. Shimada, M., Yoshinari, K., Tanabe, E., Shimakawa, E., Kobashi, M., Nagata, K., and Yamazoe, Y. (2001). Identification of ST2A1 as a rat brain neurosteroid sulfotransferase mRNA. Brain Res. 920, 222–225. Southwell, B. R., Duan, W., Alcorn, D., Brack, C., Richardson, S. J., Kohrle, J., and Schreiber, G. (1993). Thyroxine transport to the brain: Role of protein synthesis by the choroid plexus. Endocrinology 133, 2116–2126. Steckelbroeck, S., Nassen, A., Ugele, B., Ludwig, M., Watzka, M., Reissinger, A., Clusmann, H., Lutjohann, D., Siekmann, L., Klingmuller, D., and Hans, V. H. (2004). Steroid sulfatase (STS) expression in the human temporal lobe: Enzyme activity, mRNA expression and immunohistochemistry study. J. Neurochem. 89, 403–417. Strazielle, N., and Ghersi‐Egea, J. F. (2000). Choroid plexus in the central nervous system: Biology and physiopathology. J. Neuropathol. Exp. Neurol. 59, 561–574. Strazielle, N., and Preston, J. E. (2003). Transport across the choroid plexuses in vivo and in vitro. Methods Mol. Med. 89, 291–304. Strazielle, N., Khuth, S. T., and Ghersi‐Egea, J. F. (2004). Detoxification systems, passive and specific transport for drugs at the blood‐CSF barrier in normal and pathological situations. Adv. Drug Deliv. Rev. 56, 1717–1740.

170

Westholm et al.

Sugiyama, D., Kusuhara, H., Shitara, Y., Abe, T., Meier, P. J., Sekine, T., Endou, H., Suzuki, H., and Sugiyama, Y. (2001). Characterization of the eZux transport of 17beta‐ estradiol‐D‐17beta‐glucuronide from the brain across the blood‐brain barrier. J. Pharmacol. Exp. Ther. 298, 316–322. Sugiyama, D., Kusuhara, H., Taniguchi, H., Ishikawa, S., Nozaki, Y., Aburatani, H., and Sugiyama, Y. (2003). Functional characterization of rat brain‐specific organic anion transporter (Oatp14) at the blood‐brain barrier: High aYnity transporter for thyroxine. J. Biol. Chem. 278, 43489–43495. Takano, T., Tian, G. F., Peng, W., Lou, N., Libionka, W., Han, X., and Nedergaard, M. (2006). Astrocyte‐mediated control of cerebral blood flow. Nat. Neurosci. 9, 260–267. Tamai, I., and Tsuji, A. (2000). Transporter‐mediated permeation of drugs across the blood‐ brain barrier. J. Pharm. Sci. 89, 1371–1388. Tanaka, H., and Mizojiri, K. (1999). Drug‐protein binding and blood‐brain barrier permeability. J. Pharmacol. Exp. Ther. 288, 912–918. Taogoshi, T., Nomura, A., Murakami, T., Nagai, J., and Takano, M. (2005). Transport of prostaglandin E1 across the blood‐brain barrier in rats. J. Pharm. Pharmacol. 57, 61–66. Tohyama, K., Kusuhara, H., and Sugiyama, Y. (2004). Involvement of multispecific organic anion transporter, Oatp14 (Slc21a14), in the transport of thyroxine across the blood‐brain barrier. Endocrinology 145, 4384–4391. Ushikubi, F., Segi, E., Sugimoto, Y., Murata, T., Matsuoka, T., Kobayashi, T., Hizaki, H., Tuboi, K., Katsuyama, M., Ichikawa, A., Tanaka, T., Yoshida, N., et al. (1998). Impaired febrile response in mice lacking the prostaglandin E receptor subtype EP3. Nature 395, 281–284. Vansell, N. R., and Klaassen, C. D. (2002). Increase in rat liver UDP‐glucuronosyltransferase mRNA by microsomal enzyme inducers that enhance thyroid hormone glucuronidation. Drug Metab. Dispos. 30, 240–246. Vezmar, S., Becker, A., Bode, U., and Jaehde, U. (2003). Biochemical and clinical aspects of methotrexate neurotoxicity. Chemotherapy 49, 92–104. Wang, L. Q., Lehmler, H. J., Robertson, L. W., and James, M. O. (2006). Polychlorobiphenylols are selective inhibitors of human phenol sulfotransferase 1A1 with 4‐nitrophenol as a substrate. Chem. Biol. Interact. 159, 235–246. Westholm, D. E., and Anderson, G. W. (2006). Fenamic acid and indocyanine green inhibit organic anion transporting polypeptide 1c1 (Oatp1c1) mediated thyroxine transport. In ‘‘77th Annual Meeting of the American Thyroid Association’’ (T. F. Davies, Ed.), Vol. 16, p. 914. Mary Ann Liebert, Inc., Phoenix, Arizona. Wolka, A. M., Huber, J. D., and Davis, T. P. (2003). Pain and the blood‐brain barrier: Obstacles to drug delivery. Adv. Drug Deliv. Rev. 55, 987–1006. Yarim, M., Moro, S., Huber, R., Meier, P. J., Kaseda, C., Kashima, T., Hagenbuch, B., and Folkers, G. (2005). Application of QSAR analysis to organic anion transporting polypeptide 1a5 (Oatp1a5) substrates. Bioorg. Med. Chem. 13, 463–471. Zhang, Y., Han, H., Elmquist, W. F., and Miller, D. W. (2000). Expression of various multidrug resistance‐associated protein (MRP) homologues in brain microvessel endothelial cells. Brain Res. 876, 148–153.

5

____________________________________________________________________________

Mechanisms and Evolution of Environmental Responses in Caenorhabditis elegans Christian Braendle, Josselin Milloz, and Marie‐Anne Fe´lix Institut Jacques Monod, CNRS‐Universities of Paris 6/7, Tour 43 2 Place Jussieu, 75251 Paris Cedex 05, France

I. Introduction II. Interactions Between Organism and Environment A. Environmental EVects on the Phenotype B. Environmental Sensitivity of the Phenotype C. Role and Evolutionary Significance of Environmental Sensitivity of Development III. The Nematode C. elegans A. General Biology B. Natural Environment C. Laboratory Environment IV. Overview of C. elegans Responses to the Environment A. Perception and Transduction of Environmental Signals B. Global Responses in Physiology and Gene Expression C. Stress Responses D. Immune Responses E. Behavioral Responses F. Developmental, Morphological, and Life History Responses G. Evolution of Environmental Responses V. Phenotypic Plasticity of C. elegans Dauer Formation A. Characteristics of the Dauer Larva B. Environmental Cues Regulating Dauer Development C. Perception and Transduction of Environmental Cues D. Evolution of Dauer Formation VI. Environmental Robustness of C. elegans Vulva Formation A. Vulva Development B. Environmental Robustness of the Final Vulva Phenotype C. Environmental Sensitivity of Vulva Developmental Processes D. Developmental and Molecular Features Causing Robustness to Variation in the Environment E. Evolution of Vulval Development VII. Conclusion Acknowledgments References

Current Topics in Developmental Biology, Vol. 80 Copyright 2008, Elsevier Inc. All rights reserved.

171

0070-2153/08 $35.00 DOI: 10.1016/S0070-2153(07)80005-6

172

Braendle et al.

We review mechanistic and evolutionary aspects of interactions between the model organism Caenorhabditis elegans and its environment. In particular, we focus on environmental eVects aVecting developmental mechanisms. We describe natural and laboratory environments of C. elegans and provide an overview of the diVerent environmental responses of this organism. We then show how two developmental processes respond to changes in the environment. First, we discuss the development of alternative juvenile stages, the dauer and non‐dauer larva. This example illustrates how development responds to variation in the environment to generate complex phenotypic variation. Second, we discuss the development of the C. elegans vulva. This example illustrates how development responds to variation in the environment while generating an invariant final phenotype. ß 2008, Elsevier Inc.

I. Introduction Organisms develop and evolve in variable environments. Understanding how an organism responds to environmental change is therefore of central significance in developmental and evolutionary biology. During ontogeny, the environment modulates the translation of genotype into phenotype, but developmental biologists typically ignore how environmental factors aVect development and resulting phenotypes. Consequently, the detailed molecular and developmental mechanisms underlying environmental responses generally remain elusive. Similarly, whether and how organisms evolve adaptations to variable environments might depend on the mechanisms underlying genotype by environment interaction during ontogeny, yet evolutionary biologists typically ignore molecular and developmental mechanisms. An integrated understanding of the proximate and ultimate aspects of interactions between genes, development, and environment is therefore of mutual benefit to both developmental and evolutionary biologists. Here we review mechanistic and evolutionary aspects of how environmental factors aVect the nematode Caenorhabditis elegans, with a particular focus on the developmental role of the environment. We first introduce general aspects of interactions between organism and environment and discuss their evolutionary significance. After briefly introducing the C. elegans model, we then review how worms respond to diVerent environmental conditions and how such responses might evolve. Finally, we discuss how environmental factors influence the expression of two developmental phenotypes: the dauer larva and the vulva. Using dauer formation as an example, we illustrate how development integrates environmental variation to generate phenotypic variation (phenotypic plasticity). Using vulva formation as an example, we then discuss the opposite: how the interaction between development and the environment can result in the absence of phenotypic variation (environmental robustness).

5. Environmental Responses in C. elegans

173

II. Interactions Between Organism and Environment A. Environmental Effects on the Phenotype In this chapter, we primarily discuss the eVects of the exogenous environment on the organism, that is, abiotic and biotic factors external to the organism. Potentially, environmental factors aVect any level of biological organization; however, the extent and nature of such an eVect is highly variable depending on the examined phenotype and environment, respectively. Although eVects on biological systems are often categorized as genetic versus environmental eVects, it is generally diYcult to retrace the relative contribution of the two eVects on a phenotype: genetic and environmental eVects are not necessarily additive but can interact in aVecting the phenotype (Lewontin, 1974). Since biological systems underlying a given phenotype comprises genetic products that respond to environmental factors, such environmental responses generally have heritable genetic basis. Given the presence of genetic variation in the response, such a response may be subject to natural selection. Environmental eVects per se are classified as eVects that are nonheritable because the environment does not alter DNA sequence. Exceptions occur when environmental factors induce transgenerational eVects through particular epigenetic mechanisms, such as DNA methylation and other parental eVects (Jablonka and Lamb, 1995). In addition, particular environments may directly alter mutation rates of soma and germ line, for example, in response to stressful environmental conditions (Tenaillon et al., 2004). B. Environmental Sensitivity of the Phenotype Most organismic features are sensitive to some environmental factors. However, such sensitivity may or may not translate into phenotypic variation (or change). Adopting a simplified perspective, a given phenotype is thus either sensitive or insensitive to a given range of environmental conditions (Fig. 1). The term phenotypic plasticity describes the property of a genotype to generate phenotypic variation in response to a given range of environmental conditions (Pigliucci, 2001; Schlichting and Pigliucci, 1998; Stearns, 1992). When a single genotype produces diVerent phenotypes that vary as a continuous function of the environment, the relationship can be described by a reaction norm (Fig. 1A). The situation in which a single genotype expresses two or more discretely diVerent phenotypes in response to environmental cues is called polyphenism (Nijhout, 1999). In contrast, a phenotype may be insensitive to a given range of environmental conditions, so that a genotype always produces the same phenotype irrespective of environmental conditions (Fig. 1).

174

Braendle et al.

A

B

Phenotype

Phenotype

G1

G2

E1

E2

E3

Environment

E4

E1

E2

E3

E4

Environment

Figure 1 Environmental sensitivity of the phenotype. (A) Possible phenotypic responses to environmental variation visualized by reaction norms. Genotype G1 is environmentally sensitive and produces diVerent phenotypes in diVerent environments (phenotypic plasticity). Genotype G2 shows the same phenotype in all environments (environmental robustness). In these examples, the slope of the reaction norm is indicative of the degree of phenotypic sensitivity to environmental variation. Since the reaction norms of the two genotypes are nonparallel, they are indicative of a genotype‐by‐environment interaction, reflecting genetic variation in phenotypic plasticity. (B) Distinction of phenotypic variation generated within environments and between environments. The error bars of the phenotypic mean value represent stochastic phenotypic variation within a given environment, and correspond to the phenotypic variance. Both phenotypic mean and variance may or may not change across environments.

Such environmental insensitivity may be termed environmental robustness or environmental canalization (de Visser et al., 2003; Flatt, 2005; Waddington, 1942). Phenotypic plasticity and environmental robustness reflect environmental sensitivity and insensitivity, respectively. The two phenomena thus describe diVerent degrees of environmental sensitivity, and can be used to refer to both adaptive and nonadaptive phenotypic responses. While the categories of response represent opposite ends of phenotypic sensitivity to the environment, they are not mutually exclusive: the same phenotype may consist of both plastic and robust elements. In addition, a given phenotypic trait may show robustness in a range of environmental conditions, but plasticity in another range of conditions. Moreover, plasticity in one phenotype may lead to robustness in another phenotype. Note that phenotypic variation within an environment (i.e., stochastic variation) may be also referred to as environmentally induced phenotypic variation. Analogously, the term robustness is also used to refer to phenotypic insensitivity to stochastic variation within an environment. For a distinction of diVerent environmental responses of the phenotype, see Fig. 1 (for recent reviews, see Ancel Meyers and Bull, 2002; Badyaev, 2005; Flatt, 2005).

5. Environmental Responses in C. elegans

175

The degree of environmental sensitivity of a given phenotype can be influenced by genetic factors. DiVerent genotypes thus exhibit diVerences in the extent of their phenotypic response to the same range of environmental conditions (Stearns, 1992). Such a diVerential response is called a genotype‐by‐environment‐interaction (GxE) interaction, reflecting genetic variation in phenotypic plasticity. Graphically, diVerences in the slope of genotype‐specific reaction norms indicate the occurrence of GxE interactions (Fig. 1).

C. Role and Evolutionary Significance of Environmental Sensitivity of Development The environment may dramatically aVect the development and morphology of many organisms. With the rise of genetics during the early twentieth century, the developmental role of the environment and its evolutionary significance became largely ignored. Despite important exceptions (e.g., Goldschmidt, 1938; Waddington, 1942), it was only relatively recently that interactions between development and environment received increased attention, particularly in evolutionary biology. It has become apparent that adaptive plasticity of developmental mechanisms and resulting phenotypes is very common (Pigliucci, 2001; Schlichting and Pigliucci, 1998). Such phenotypic plasticity at the developmental level (developmental plasticity) has been proposed to play important roles in evolutionary diversification (West‐Eberhard, 2003). This idea is partly based on the observation that environmentally induced phenotypic variation can become genetically fixed through a selective process called genetic assimilation (a process which requires the presence of genetic variation in the environmental response) (Waddington, 1956). While developmental plasticity is a common phenomenon, many developmental processes generate little or no phenotypic variation across diVerent environments—they are robust. Such environmental robustness may evolve as a result of selection for a single invariant phenotype that is reproducible across environments. Environmental robustness of the phenotype is likely to correlate with mutational robustness of the phenotype (insensitivity to mutation) (Meiklejohn and Hartl, 2002). Thus, robustness of the phenotype can have seemingly paradoxical evolutionary consequences. On the one hand, robustness will reduce the capacity to evolve at the phenotypic level, since robustness decreases the expression of phenotypic variation. On the other hand, as the phenotype becomes robust to mutational variation, the underlying genotype accumulates genetic changes; thus, the capacity to evolve increases at

176

Braendle et al.

the genetic level. Such genetic variation that does not translate into phenotypic variation is often termed hidden or cryptic genetic variation (Gibson and Dworkin, 2004). Cryptic genetic variation might be uncovered (i.e., phenotypically expressed) by particular mutations or environmental changes; selection might then act on this variation (Rutherford and Lindquist, 1998; Waddington, 1956). Thus, cryptic genetic variation, once uncovered, could potentially play a role in phenotypic evolution. The degree of environmental sensitivity—ranging from extreme phenotypic plasticity to extreme environmental robustness—may aVect the capacity to evolve at genotypic and phenotypic levels. Yet, these inferences mainly stem from theoretical work and very little is understood about how environmental sensitivity is mediated by molecular and cellular mechanisms to interact with development. As a consequence, we have limited insights into how environmental sensitivity of the phenotype evolves and how it aVects evolutionary change. In this chapter, we discuss how the extensive research on the C. elegans model system can shed light on the mechanistic basis underlying phenotypic plasticity and environmental robustness, a subject of interest to both developmental and evolutionary biologists.

III. The Nematode C. elegans A. General Biology C. elegans has become a well‐established model organism for molecular genetic studies (www.wormbook.org), and is now increasingly used for evolutionary studies (Carvalho et al., 2006; Haag et al., 2007). This small nematode provides a simple metazoan system for research aimed at the cell level—the adult hermaphrodite contains 959 somatic cells and its virtually invariant cell lineage has been determined in its entirety (Kimble and Hirsh, 1979; Sulston and Horvitz, 1977; Sulston et al., 1983). C. elegans reproduces predominantly through self‐fertile hermaphrodites, yet outcrossing events are possible through male production. Sex determination is of the XX:XO type, with males having only a single X chromosome. ‘‘Spontaneous’’ males are generated through nondisjunction events of the X chromosome during meiosis, which are generally rare under laboratory conditions. After a mating event, however, about half of the hermaphrodite progeny will be male. Current evidence suggests that male production and outcrossing events are overall rare but can be variable among natural populations (Barrie`re and Fe´lix, 2005a, 2007; Sivasundar and Hey, 2005). The life cycle of both hermaphrodite and male includes five postembryonic stages: four juvenile (larval) stages and the adult stage, and may only take 3 days to complete in laboratory conditions (Fig. 2A). Depending on the environmental

5. Environmental Responses in C. elegans

177

Figure 2 (A) The life cycle of C. elegans in the laboratory (adapted from Riddle and Albert, 1997; Wood, 1988). (B) Phenotypes of non‐dauer and dauer larvae (from Altun and Hall, 2005).

conditions, individuals can adopt an alternative phenotype during the third larval stage, termed dauer larva. The dauer larva is morphologically and physiologically specialized, developmentally arrested, and is tolerant to harsh environments (Fig. 2B). The developmental divergence between non‐ dauer and dauer phenotypes represent a well‐studied example of phenotypic plasticity, which we discuss in detail later. This and many other biological features of C. elegans, as well as representing one of the anatomically and genetically best‐understood metazoans, provide an ideal basis to study an organism’s manifold interactions with the environment and how such interactions evolve.

B. Natural Environment Natural populations of C. elegans have been little studied and its ecology is not well understood. C. elegans was first found in humus samples collected in Algeria (Maupas, 1900) and the reference laboratory strain (N2) was isolated from humus used in a mushroom farm in United Kingdom (Hodgkin and Doniach, 1997; Staniland, 1957). Populations of C. elegans occur worldwide at geographical locations with cold (Que´bec, Canada) or very hot and dry (Southern California, United States) weather conditions (Barrie`re and Fe´lix, 2005b). C. elegans is often referred to as a soil‐dwelling nematode; however,

178

Braendle et al.

recent surveys indicate that natural populations also occur in a number of other habitats containing decomposing vegetal matter (fruits, compost, leaf litter), and to a lesser extent, decomposing invertebrates, such as snails. Most frequently, C. elegans was found in garden compost (Barrie`re and Fe´lix, 2005a, 2007; Caswell‐Chen et al., 2005; Cutter, 2006; Haber et al., 2005; Sivasundar and Hey, 2005). The characteristics of these natural habitats already indicate some of the likely environmental changes to be experienced by C. elegans in the wild, for example, changes in food availability, temperature, oxygen, and ethanol concentration. In addition, intra‐ and interspecific competition for food sources is probable as many other bacteriophagous nematode species, including the closely related species C. briggsae, occur in the same habitats as C. elegans (Barrie`re and Fe´lix, 2005b). The vast majority of isolated C. elegans individuals were found to be hermaphrodites in the dauer stage, indicating that environmentally sensitive dauer formation is common in nature. Rare proliferating populations with a large proportion of non‐dauers were found only on rotten fruit and in very fresh compost (Barrie`re and Fe´lix, 2005b, 2007). The typical C. elegans habitat therefore shows fluctuations between favorable and unfavorable growth conditions, and the dauer stage appears to allow survival in such ephemeral habitats. C. elegans individuals have also been found on various arthropods (isopods, millipedes, mites) as well as on snails (Barrie`re and Fe´lix, 2005a, 2007; Kiontke and Sudhaus, 2006). Such associations with invertebrates may reflect phoretic (dispersal) or necromenic (feeding on the dead, decomposing animal) behaviors. Dauer larvae show a stereotypical ‘‘waving behavior’’ which may serve to find or attract such hosts (Riddle, 1988; Riddle and Albert, 1997). Consistent with this hypothesis, laboratory observations show that C. elegans dauer larvae readily mount, and then stay attached to arthropods such as mites (M.‐A. F., unpublished data). Although rarely investigated, the microbial fauna associated with C. elegans appears to be extensive and diverse. A variety of bacterial species have been isolated from its natural habitat, which may serve as food source or act as pathogens (Grewal, 1991; M.‐A. F., unpublished data). In the laboratory, C. elegans is fed with Escherichia coli; however, the culture medium must be supplemented with cholesterol because C. elegans cannot synthesize cholesterol, required for steroid synthesis. This indicates that the natural diet of C. elegans may also include eukaryotes, such as unicellular slime moulds (Kessin et al., 1996). While C. elegans may feed on diVerent bacteria, many of them also represent pathogens, for example, Serratia marcescens, which has been isolated from compost samples containing C. elegans (M.‐A. F., unpublished data). In addition, certain bacterial spores cannot be digested and are accumulated in the intestine, which may inhibit food uptake. In natural C. elegans populations, many non‐dauer individuals exhibit apparent bacterial infections, such as accumulation of spores in the

5. Environmental Responses in C. elegans

179

intestinal lumen, bacterial invasion of the body cavity, or intracellular invasion (Barrie`re and Fe´lix, 2005a). While certain fungi may represent a food source, others exhibit specialized hyphae, allowing the trapping of nematodes (Dusenbery, 1996). Other likely C. elegans predators include nematophagous springtails (collembola) (Lee and Widden, 1996) or mites (acari), the latter of which have been found in C. elegans compost samples (M.‐A. F., unpublished data). Sampling and genetic analysis of natural populations confirm that the natural C. elegans habitat is often ephemeral, with populations undergoing frequent bottlenecks in size (Barrie`re and Fe´lix, 2005a, 2007). These observations suggest a distinct metapopulation structure with frequent extinction and recolonization of habitats. C. elegans life history further indicates that population expansion may be rapid under favorable conditions. Once a food source of the habitat has been exploited, the formation of dauer larvae may allow diapause or dispersal to new habitats.

C. Laboratory Environment C. elegans is usually maintained in Petri dishes filled with solidified agar made from a standardized nematode growth medium (NGM) (Hope, 1999). A laboratory strain of E. coli (OP50) is most commonly utilized as a food source. NGM plates are inoculated with E. coli by adding a small drop of a bacterial suspension on the surface of the agar. Subsequent growth of E. coli results in a viscous bacterial lawn on which the worms feed. The majority of C. elegans experiments are carried out using worms living on such monoxenic plates, within a temperature range of 15–25  C (generally 20  C). For certain experimental procedures, C. elegans is also reared in defined monoxenic or axenic liquid media (Hope, 1999; Szewczyk et al., 2003). The standard laboratory environment using NGM plates is very simple in structure and has been developed to maximize oVspring production, so that it can be regarded as a rather benign environment. However, E. coli, being a mammalian intestine symbiont, is unlikely to represent an important natural food source, and further may have pathogenic eVects (Herndon et al., 2002). Virtually all C. elegans research has made use of a single isogenic isolate, named N2. Biological observations of wild‐type and mutant animals are thus representative of a single genotype. N2 was isolated sometime before 1956 and was thereafter reared in various laboratory environments until the use of standardized NGM plates about 10 years later. Around 1970, the first frozen stocks of N2 were established (C. elegans survives freezing and can be cryopreserved over long time periods) (for details on the laboratory history of N2, see Hodgkin and Doniach, 1997). Because of the long maintenance of N2 in the laboratory environment, it is likely that this genotype has adapted to some

180

Braendle et al.

of these conditions, and may have evolved features atypical for other genotypes living in the natural environment. Such laboratory adaptation may aVect environmental responses in particular, for example, certain responses may not be maintained because of the absence of the relevant environmental stimuli. Nevertheless, in the laboratory, N2 generally shows no dramatic diVerences from freshly isolated wild isolates for most tested phenotypes. In the laboratory, environmental eVects on various C. elegans traits have been studied mostly in response to variation in specific abiotic or biotic factors, including temperature, chemical substances (e.g., toxins), oxygen availability, nutrition (e.g., starvation), culturing medium, population density, and pathogens. Most factors under study represent evident stressful conditions, defined as suboptimal conditions reducing viability, survival or reproduction relative to another environment. Very few studies included complex laboratory environments (Goranson et al., 2005; Van Voorhies et al., 2005) or environments that vary on a spatiotemporal scale (Friedenberg, 2003).

IV. Overview of C. elegans Responses to the Environment Although its ecology is poorly understood, it is evident that C. elegans inhabits a highly variable environment. Nevertheless, laboratory experiments analyzing mostly isolated environmental factors indicate that C. elegans tolerates many extreme environmental conditions due to physiological, morphological, and behavioral responses. Many of these responses reflect likely evolutionary adaptations allowing an adjustment to variable, and often nonoptimal, environments.

A. Perception and Transduction of Environmental Signals C. elegans is highly sensitive to environmental signals and responds to chemical, thermal, and mechanical stimuli. Both hermaphrodites and males possess a complex sensory system despite a relatively simple nervous system. The hermaphrodite has a total of 302 neurons, 30 of which represent sensory neurons (Hobert, 2005). Chemosensation plays a particularly important role in the mediation of environmental signals as C. elegans can detect hundreds of diVerent volatile or aqueous chemicals (Bargmann, 2006). The basic chemosensory system consists of ciliated neurons that are in direct or indirect contact with the environment. Laser cell ablation experiments have identified sensory neurons required for the response to specific environmental signals. Sensation of most signals occurs through the amphid sensory organ, a concentration of neurons located in the head region. Specific environmental variables that cause changes in chemical compounds, such as ions or amino

5. Environmental Responses in C. elegans

181

acids, are sensed by, sometimes overlapping, subsets of sensory neurons. Perception of molecules occurs through a large number of diVerent chemoreceptors required for subsequent signal transduction, with a single neuron expressing multiple chemoreceptor genes. Approximately 1300 G‐protein– coupled receptors (GPCRs) are encoded by the C. elegans genome, many of which are likely to play a role in chemoreception (Bargmann, 2006; Robertson, 2000). In total, several hundred GPCRs are expressed in the sensory system and alternative signaling components may complement signal sensation. Downstream signal transduction occurs through cGMP–cGMP channels and TRPV channels, as well as other signaling cascades (Bargmann, 2006). Signal transduction further underlies extensive regulation by kinases and phosphatases to ultimately aVect a potentially large number of genes containing signal‐specific response elements. On neuronal perception of environmental signals, neuropeptides, and neurohormones may be involved in the mediation of the signal. For example, the bioamine neurotransmitters serotonin and octopamine are involved in mediating responses to alternative feeding states. Exogenous application of these two substances can mimic well‐fed and starved behavioral states of the animal, respectively (Horvitz et al., 1982; Mohri et al., 2005; Suo et al., 2006). Perception of certain stimuli may also occur through neurons that are not in direct contact with the external environment, such as the PQR neurons of the body cavity, which sense changes in the body fluid and regulate the animal’s behavior (Cheung et al., 2004; Gray et al., 2004). Overall, neuronal sensation plays predominant and multiple roles in behavioral and physiological responses to the environment. Perception through neurons may lead to cell‐autonomous as well as systemic responses through secreted factors, which may act on a long distance to aVect both neuronal and nonneuronal cells. In addition, the environment may elicit responses by means diVerent from neuronal perception. For example, certain molecules (e.g., ions) may impact an animal by diVusion, or metabolic changes (e.g., due to changes in nutrition uptake) may cause environmental eVects independent of neuronal perception. Many environmental conditions seem likely to trigger a combination of eVects, involving both neuronal and nonneuronal perception. In addition, most conditions are likely to aVect multiple organismic features, and it is often unclear whether and how the environment acts on these features independently. As a consequence, it is diYcult to determine how environmental signals are integrated, especially for traits with complex genetic architecture and environmental sensitivity. Body size provides an example of such a complex composite trait, which is highly polygenic and sensitive to a wide range of environmental conditions. The developmental process underlying C. elegans body size determination involves multiple interdependent processes, such as chemosensory inputs, endocrine and metabolic signals, to ultimately aVect cell size (ploidy) and

182

Braendle et al.

number (Azevedo et al., 2002; Flemming et al., 2000; Fujiwara et al., 2002; Morck and Pilon, 2006; Savage‐Dunn et al., 2003; Wang et al., 2002). While individual genetic (mutant) or environmental factors on body size can be isolated, the complicated interplay between diVerent factors appears very diYcult to disentangle. B. Global Responses in Physiology and Gene Expression Physiological and molecular responses have been mainly examined in response to stressful conditions, and molecular genetic analyses have resulted in the detailed characterization of environmental signal transduction. Global analyses of diVerential gene expression and metabolic changes in response to diVerent environmental factors indicate extensive and widespread eVects on gene regulation and molecular processes downstream of signal perception (Jones et al., 2001; Li et al., 2006; Menzel et al., 2005; Reichert and Menzel, 2005; Szewczyk et al., 2006). The environmentally induced adoption of the alternative dauer developmental stage, for example, underlies expression changes in more than 2000 genes (Wang and Kim, 2003). A subset of these expression changes involved in dauer formation also occur in starved L1 larvae or old adults, suggesting that gene expression may respond similarly to diVerent environmental factors at multiple life stages (Cherkasova et al., 2000). The significance of such diVerential gene expression pattern is, however, for the most part diYcult to assess. Although in some cases RNAi‐based assays have been used with success to address the function of specific genes (Murphy, 2006; Shapira et al., 2006), it remains a main challenge to identify which of these genes are directly involved in the environmental signal transduction, or which gene expression changes translate into phenotypic changes.

C. Stress Responses Evolutionarily conserved stress response mechanisms have also been reported in C. elegans. Such mechanisms play fundamental roles in adjusting cellular and physiological functions to variation in critical environmental variables. Transcriptional activation of heat shock proteins (HSPs) at elevated temperature presents such a stress response conferring, for example, increased protein stability. Heat‐induced expression of HSPs, such as DAF‐21, a molecular chaperone of the HSP90 family, thus may be observed in virtually all cells of the organism (Inoue et al., 2003). However, HSP activation is not only thermosensitive, but may occur in response to other environmental stressors (e.g., hypo‐ and hyperoxia, toxins, pathogens, starvation) and may play additional roles during a number of developmental

5. Environmental Responses in C. elegans

183

processes in the absence of stress factors (Candido, 2002; Walker et al., 2001, 2003). Tolerant to a wide range of oxygen concentrations, C. elegans also survives both anoxic and hyperoxic conditions and shows physiological adaptations to low environmental oxygen levels (Lamitina et al., 2004; Van Voorhies and Ward, 2000). Such hypoxia responses are partly mediated by evolutionarily conserved hypoxia‐inducible factor (HIF) complexes (Jiang et al., 2001; Shen et al., 2005). Mutant analyses have implicated other diverse genetic components to be involved in a number of cellular and physiological stress responses. Many of these components belong to well‐characterized signaling cascades (Koga et al., 2000; Lehtinen et al., 2006). D. Immune Responses In the laboratory, C. elegans is readily infected by a variety of pathogens, some of which are known to occur in its natural environment (Darby, 2006). Current research increasingly focuses on such pathogen interactions to study the C. elegans immune response (Ewbank, 2006). Responses to pathogen infection may include previously mentioned stress response mechanisms. However, infection studies using diVerent pathogenic bacteria suggest the additional occurrence of specific immune responses, usually involving conserved signal transduction pathways, such as the ERK and p38 MAPK pathways (Ewbank, 2006). Apparent and potential defence responses include the production of antimicrobial peptides and proteins (Couillault et al., 2004; O’Rourke et al., 2006), morphological modification (Nicholas and Hodgkin, 2004), possibly RNA interference on viral infection (Lu et al., 2005; Wilkins et al., 2005), and behavioral avoidance (Pradel et al., 2007; Pujol and Ewbank, 2005). E. Behavioral Responses Behavioral flexibility represents one of the most fundamental adaptations to respond to variation in the environment. Behavioral change has the advantage that it allows a rapid, yet reversible adjustment to the current environmental situation. C. elegans behavior may respond to touch, smell, taste, or temperature. These responses primarily allow moving to favorable (e.g., food source) and avoiding unfavorable conditions (e.g., toxins). Chemosensation assays have revealed diverse chemical repellents and attractants (Bargmann, 2006). In C. elegans, in contrast to many other animals, it is a small set of neurons that determines the behavioral response to a given chemical. Each of these neurons expresses a repertoire of chemosensory receptors. Whether C. elegans is attracted or repulsed by an olfactory cue is determined by the cell(s) in which the corresponding chemosensory receptors are expressed. The pair of AWB neurons is associated with repulsion while AWC and AWA neurons are

184

Braendle et al.

associated with attraction; ectopic expression in AWB of a chemoreceptor (ODR‐10) normally exclusively expressed in AWA is suYcient to transform a normally attractive stimulus into a repulsive one (Troemel et al., 1997; Wes and Bargmann, 2001). Many other behavioral repertoires and their underlying neuronal basis have been studied in the context of their response to the environment, such as mechanosensation and locomotory responses to body or nose touch (Rankin, 2000), thermal preference and nutrition history (Mohri et al., 2005), egg laying and nutrition status (Schafer, 2006), or male mating and hermaphrodite presence (Emmons, 2005). In addition, C. elegans may adjust its behavior through learning and memory (Giles et al., 2005). For example, after exposure to pathogenic bacteria, animals modify their olfactory preferences by increased avoidance of pathogenic bacteria (Pujol and Ewbank, 2005; Zhang et al., 2005). F. Developmental, Morphological, and Life History Responses While behavioral change is a primary response to adjust to environmental change, many environmental factors also induce physiological modifications, which may have eVects on developmental decisions, overall phenotype (e.g., body size and form), and life history characteristics (e.g., fecundity or longevity). Some of these responses are transient or reversible, yet distinct developmental or morphological modifications occur in some environmental conditions, such as the adoption of the dauer phenotype (see Section VI), the developmental arrest of L1 larvae in the absence of food (Baugh and Sternberg, 2006; Fukuyama et al., 2006), or the stereotypical changes in body morphology and locomotion in liquid culture (Moghal et al., 2003; Szewczyk et al., 2006). A small number of studies have also examined C. elegans life history in laboratory conditions mimicking the natural environment, that is, in spatially complex, soil‐like environments (Goranson et al., 2005; Van Voorhies et al., 2005). These results show that many life history parameters, including oVspring production and longevity may drastically change under ecologically relevant conditions. G. Evolution of Environmental Responses Most of what we know about C. elegans, including its interactions with the environment, stems from studies using a single isogenic wild isolate, the standard laboratory strain N2. Therefore, little is known about how environmental responses evolve within C. elegans. Despite their small number, studies examining distinct C. elegans isolates clearly reveal evolution in interactions between organism and environment. Such comparative analyses reveal intraspecific variation in behavioral responses to diVerent environmental

5. Environmental Responses in C. elegans

185

stimuli (Davies et al., 2004; de Bono and Bargmann, 1998; Jovelin et al., 2003), survival and infection rate in response to pathogens (Schulenburg and Ewbank, 2004; Schulenburg and Muller, 2004), the propensity to form dauer larvae in response to dauer‐inducing conditions (Viney et al., 2003), or temperature‐dependent fecundity (Harvey and Viney, 2007). A number of other traits, known to be environmentally sensitive, show genetic variation among C. elegans wild isolates in a single environment. These traits include, for example, body size, fecundity, and longevity (Gems and Riddle, 2000; McCulloch and Gems, 2003). In addition, the temperature‐ dependent propensity to generate males through nondisjunction of the X‐chromosome varies among natural isolates (Hodgkin and Doniach, 1997; Nigon, 1949; Teotonio et al., 2006). Most of the examined isolates are geographically as well as genetically distinct. However, even within a single location, C. elegans genotypes may show genetic divergence (Barrie`re and Fe´lix, 2005a; Haber et al., 2005), which may correlate with diVerent responses to environmental factors, such as pathogens (Schulenburg and Ewbank, 2004). Evolutionary variation in environmental responses has also been detected by quantitative trait loci (QTL) analyses using recombinant inbred lines derived from crosses between distinct C. elegans natural isolates. Gutteling et al. (2006) conducted a QTL analysis at low and high temperatures, and revealed genetic variation and GxE interactions for a number of life history traits. A related study (Li et al., 2006) has applied QTL methods to map such temperature‐dependent diVerences to the gene expression level using oligonucleotide microarrays. The results suggest the presence of ample genetic variation in environmentally sensitive gene expression changes, but it is currently unclear which genetic factors have evolved to generate this diVerential response to the environment. Additional QTL studies have revealed GxE interactions for a number of life‐history traits (Ayyadevara et al., 2001, 2003; Ebert et al., 1996; Johnson and Hutchinson, 1993; Knight et al., 2001; Shook and Johnson, 1999; Shook et al., 1996). However, the results of these QTL studies should be interpreted with caution as they made use of C. elegans isolates which may have been particularly sensitive to environmental variation due to an increased mutational load caused by a high copy number of Tcl transposons. So far, basically all examined environmentally sensitive features of C. elegans show evolutionary variation. Moreover, several studies provide clear examples of GxE interactions, demonstrating that diVerent C. elegans genotypes show diVerential responses to the same environmental change (Harvey and Viney, 2007). The prevalence of extensive evolutionary variation in environmental responses is further supported by gene sequence analyses. Stewart et al. (2005) found high allelic diversity in chemoreceptor genes among 20 C. elegans wild isolates. A number of the examined genes correspond to nonfunctional

186

Braendle et al.

pseudogenes in the reference strain N2, but are apparently functional genes in other wild isolates. This and other sequence analyses indicate high polymorphism and rapid evolution of genes likely to be implicated in environmental interactions, including chemosensation, detoxification, and immune response (Thomas, 2006; Thomas et al., 2005).

V. Phenotypic Plasticity of C. elegans Dauer Formation Adoption of the alternative dauer developmental stage is highly sensitive to variation in the environment and represents an example of apparent adaptive phenotypic plasticity. In a favorable environment, normal development is maintained and reproduction is initiated rapidly. In an adverse environment, dauer development generates stress resistance and allows diapause, so that further postembryonic development can be postponed until a favorable environment is encountered. Entry and exit into dauer are cued by specific environmental signals indicative of future environmental conditions, allowing extensive, yet flexible phenotypic modifications. Dauer versus non‐dauer development generate distinctly diVerent phenotypes during the L3 stage and thus corresponds to a polyphenism, a special case of phenotypic plasticity. Dauer development represents the molecularly best‐understood example of how a genome can generate alternative developmental phenotypes in response to environmental signals.

A. Characteristics of the Dauer Larva Dauer larvae show a distinct overall morphology (Fig. 2B). They are thinner than regular L3 larvae, do not seem to feed (as their mouth is closed), have a constricted pharynx, exhibit a thin dark intestine, fat body accumulation, a specific cuticular pattern with lateral ridges (alae), and undergo remodeling of neurons, foregut, and other structures (Ao et al., 2004; Cassada and Russell, 1975; Riddle, 1988; Riddle and Albert, 1997). A large number of gene expression changes are associated with entry and exit into dauer (Dalley and Golomb, 1992; Jones et al., 2001; Wang and Kim, 2003). Some of these changes correspond to extensive metabolic changes with a partial shift to anaerobic fermentation on dauer entry (Holt and Riddle, 2003). Dauer larvae are usually immobile, yet may move very rapidly on touch. They may also display a specific waving behavior, which is thought to help finding a host for dispersal. Dauer larvae are resistant to a variety of stressors (e.g., starvation, desiccation, extreme temperatures, toxins) (Anderson, 1978; Cassada and Russell, 1975) and can survive up to several months in laboratory conditions

5. Environmental Responses in C. elegans

187

(Klass and Hirsh, 1976). The extended longevity of dauer larvae appears to be partly due to increased expression of stress resistance genes, metabolic changes, and insulin‐like signaling (Olsen et al., 2006).

B. Environmental Cues Regulating Dauer Development The known environmental conditions triggering entry into the dauer stage are: high dauer pheromone concentration (e.g., at high population density), low food concentration, and high temperature (Cassada and Russell, 1975; Golden and Riddle, 1984b). The propensity to enter dauer is very sensitive to variation in these conditions and usually depends on a combination of these cues. Dauer larvae readily occur in the laboratory when populations are maintained at high densities. The main environmental cue mediating this response is a pheromone. This dauer pheromone (or a component of it), termed daumone, has been purified and represents a pyranose sugar linked to a fatty acid chain (Jeong et al., 2005). Starvation conditions usually enhance the propensity to enter the dauer stage. In addition, dauer formation is highly sensitive to variation in temperature. The tendency to form dauers generally increases as temperature increases (Ailion and Thomas, 2000). The developmental decision to enter the dauer developmental pathway occurs around the late L1 stage. With the completion of the L1/L2 moult, the larva corresponds to a distinct pre‐dauer L2 stage (L2d), which diVers morphologically and metabolically from a normal L2 larva. The L2d then may enter the dauer stage at the next lethargic period of the L2/L3 moult; however, if dauer‐inducing conditions cease during the second larval stage, L2d larvae may develop into normal L3 larvae. Exit from the dauer stage is induced by a set of environmental cues indicating favorable growth conditions, that is, low pheromone levels, high food concentration, and low temperature. On encounter of such conditions, commitment to dauer recovery may occur within an hour and the animal shows increased pharyngeal pumping and locomotion (Golden and Riddle, 1984b). Within several hours, the animal may resume feeding and development to rapidly moult into a post‐dauer L4 larva (10 hours at 25  C) (Cassada and Russell, 1975). Post‐dauer development seems to be similar to non‐dauer development of corresponding life stage, but has been little studied. Although the dauer versus non‐dauer decision is usually depicted as a switch between two distinct developmental pathways, there may also be intermediate responses in some environmental conditions. First, larvae may adopt a partial dauer phenotype; for example, they exhibit no fat accumulation or pharynx constriction, which indicates that they may be able to feed (Ailion and Thomas, 2003). Second, a larva may enter and rapidly exit the dauer stage to resume development, as observed for the N2 strain at 27  C (Ailion and

188

Braendle et al.

Thomas, 2000). In addition, genetically identical individuals in the same environment (such as a pheromone plate at 25  C) do not all enter the dauer stage, which could imply stochastic variation in the response. Alternatively, this observation may be explained by variation in dauer‐inducing factors within a given environment. In general, the observation of alternative phenotypes may be due to developmental switches or to discrete occurrence of environmental factors (Nijhout, 2003); however, for many polyphenic organisms, intermediate phenotypes can be generated under some environmental circumstances.

C. Perception and Transduction of Environmental Cues Laser‐mediated cell ablations and mutant screens determined the role of specific sensory neurons and downstream neuroendocrine signaling cascades involved in dauer formation. Mutant screens yielded two opposite phenotypes with respect to dauer development: dauer‐defective (Daf‐d) mutants that show no or decreased sensitivity to dauer‐inducing conditions, while dauer‐constitutive (Daf‐c) mutants may undergo dauer formation even in the absence of dauer‐inducing conditions (Riddle and Albert, 1997). Genetic analysis of these mutants revealed a complex genetic network involved in dauer development (Fig. 3A). However, the nature of cellular interactions involved in this signaling process is not fully understood. Amphid sensory neurons may sense diVerent combinations of chemical cues to regulate dauer entry (Bargmann and Horvitz, 1991). In general, the ASI, ADF, and ASG amphid neurons repress dauer entry in good growth environments, while the ASJ neuron is required for dauer entry in unfavorable environments. (In addition, the ASJ neuron is required for exit from dauer.) Chemosensory reception in sensory neurons may involve cGMP signaling. Under good growth conditions (i.e., low pheromone but high food concentration), such molecules mediate the transcriptional activation of two downstream‐secreted neuroendocrine signals, DAF‐7/TGF‐ in ASI and DAF‐28/insulin (and possibly other insulins) in ASI and ASJ neurons (Li et al., 2003; Ren et al., 1996; Schackwitz et al., 1996; Fig. 3B). Although insulin signaling may have some cell‐autonomous eVects, it aVects another cell‐nonautonomous endocrine signal that regulates dauer entry downstream of TGF‐ and insulin reception. In epidermal and neuronal cells, TGF‐ and insulin signaling indeed converge in the transcriptional upregulation of DAF‐9 (a cytochrome P450 related to vertebrate steroidogenic enzymes). DAF‐9 catalyzes the synthesis of dafachronic acid (Held et al., 2006; Motola et al., 2006). This lipophilic hormone binds DAF‐12, the downstream nuclear receptor that acts in the dauer decision and in overall developmental timing. Liganded DAF‐12 represses dauer development, whereas unliganded DAF‐12 promotes dauer development.

189

5. Environmental Responses in C. elegans A External sensory cues Low pheromone LOW T⬚ Food

Non-cell Neuroendocrine autonomous signals action Sensory neurons ASI ADF ASJ G-coupled receptors cGMP ASJ

DAF-7/TGF-b production in ASI

DAF-9 cyt-P450

Unliganded DAF-12 nuclear receptor

Dauer formation

Dafachronic acid Liganded DAF-12 nuclear receptor

DAF-28 insulin(s) in ASI, ASJ

B

Non-dauer

Dauer

Figure 3 Dauer development. (A) Simplified representation of the signal transduction cascade involved in dauer formation. (B) DiVerential expression of DAF‐28/insulin in amphid neurons of non‐dauer and dauer larvae. A daf‐28::GFP reporter reveals transcriptional repression of daf‐28 in ASI and ASJ sensory neurons (arrows) of individuals adopting the dauer developmental pathway (right panel); the corresponding DIC images are shown at the lower right corner (images from Li et al., 2003).

In summary, sensory cues such as pheromone, food, and temperature act through sensation of the amphid neurons. The combination of these cues is processed via a small neuronal network that results, under favorable growth conditions, in the activation of TGF‐ and insulin neuroendocrine signaling and the production of a liganded form of the nuclear hormone receptor DAF‐12. In unfavorable growth conditions, DAF‐12 is unliganded and promotes a cell‐specific dauer developmental program (Fig. 3A).

D. Evolution of Dauer Formation Dauer formation represents a dramatic developmental response to adjust to variable environments. We thus might expect evolutionary diVerences in the extent of this response among genotypes occupying diVerent ecological conditions. Consistent with this scenario, diVerent C. elegans wild isolates may diVer greatly in their propensity to generate dauer larvae. Viney et al. (2003) showed that dauer induction response to varying concentrations of dauer

190

Braendle et al.

pheromone displays wide genetic variation. For example, the laboratory strain N2 enters the dauer stage already at low pheromone concentrations. In contrast, at the same pheromone dose, other isolates produce few (DR1350) or no dauer larvae at all (CB4932). It is currently unknown which genetic factors have evolved to mediate this diVerential response; however, analysis of recombinant inbred lines derived from parents with opposing pheromone response indicates the presence of multiple loci involved in this diVerence (Viney et al., 2003). Considering the complexity of the genetic network involved in dauer formation, there are many potential scenarios of how such genetic diVerences may evolve. In addition, diVerent genotypes may also have evolved to respond to diVerent types or combinations of environmental cues. The detailed understanding of the molecular mechanisms governing dauer formation in N2 should allow future dissection of the molecular genetic changes underlying this evolutionary diVerence. Evolutionary diVerences in dauer formation have also been observed at higher taxonomic levels; for example, the C. briggsae isolate AF16 (India) appears to be more sensitive to certain dauer‐inducing conditions than many C. elegans isolates, including N2 (Fodor et al., 1983; Golden and Riddle, 1984a). Genome sequencing of C. briggsae (Stein et al., 2003) and additional Caenorhabditis species will facilitate molecular interspecific comparisons in dauer regulatory pathways. Moreover, based on several phenotypic similarities, the C. elegans dauer developmental stage is thought to correspond to the infective stage found in many parasitic nematodes (Bird and Opperman, 1998; Blaxter, 2003). The production of the infectious developmental stage may be influenced by the environment, and may involve the same neuronal perception mechanisms as in C. elegans dauer formation (Tissenbaum et al., 2000). Some downstream signaling cascades and eVectors may also be conserved, for example, the role of insulin signaling (Brand and Hawdon, 2004). However, recent molecular genetic and genomic analyses suggest that molecular regulation of dauer and parasite formation also exhibit considerable diVerences, for example, in the role of TGF‐ signaling (Viney et al., 2005).

VI. Environmental Robustness of C. elegans Vulva Formation The vulva of the C. elegans hermaphrodite is the organ used for egg laying and mating with males. Many defects in this reproductive structure can lead to a decrease in oVspring production (or impede outcrossing), which may cause negative fitness eVects. As a consequence, we expect vulval development to have evolved to generate an invariant and reproducible phenotype. This seems to be the case as the vulval phenotype exhibits no apparent variation in its underlying cell pattern development and the phenotype appears to be insensitive to variation in many environmental factors. How

5. Environmental Responses in C. elegans

191

does an organism maintain such phenotypic stability when exposed to variable and unfavorable environmental conditions? We address this question by studying C. elegans vulval development in diVerent laboratory environments. Here we discuss how the environment aVects the vulva phenotype, the underlying developmental process, and which molecular and developmental features may contribute to maintain a precise vulval phenotype in diVerent environments. We further present evidence that environmental eVects on vulval development vary among diVerent C. elegans wild isolates, indicating evolutionary variation in the robustness of this trait.

A. Vulva Development The C. elegans vulva is formed by a subset of ventral epidermal blast cells, the Pn.p cells (posterior descendants of Pn cells), and is connected to the ventral uterus of the gonad by the anchor cell. Twelve Pn.p cells, labeled P1.p–P12.p, are generated during the L1 stage. Six central cells, P3.p–P8.p, express the Hox gene lin‐39 and remain unfused and competent to form vulval tissue (while five of the remaining cells fuse with the surrounding syncytial epidermis). P(3–8).p form the vulva competence group and are called the vulva precursor cells (Clark et al., 1993; Maloof and Kenyon, 1998) (note that P3.p may also fuse with the epidermis and is thus not always competent; Fig. 4A). During the L3 stage, these vulval precursor cells adopt alternative cell fates on the basis of activation of conserved signaling pathways (Fig. 4). Although all six cells are capable of forming all vulval cell types, only three of them (P5.p–P7.p) will form the vulva (Sternberg and Horvitz, 1986; Sulston and White, 1980). The central cell, P6.p, adopts the 1 (or inner vulval) fate while its neighboring cells, P5.p and P7.p, adopt the 2 (or outer vulval) fate. Although competent to adopt a vulval fate, for example, on ablation of the central cells, P3.p, P4.p, and P8.p usually adopt a non‐vulval 3 fate, and their progeny finally fuse to the epidermal syncytium (Horvitz and Sternberg, 1991). The pattern of three cell fates is induced via a LIN‐3/EGF signal released from the anchor cell in the gonad overlaying the vulval precursor cells. This signal activates a RAS/MAPK signaling cascade via the LET‐23/ EGFR receptor (Aroian et al., 1994; Sternberg, 2005). The central cell P6.p receives the highest dose of LIN‐3/EGF signal, which activates RAS/ MAPK signaling at a high level and causes it to adopt a 1 vulval fate (Katz et al., 1995). RAS/MAPK signaling in P6.p also activates transcription of DELTA‐like molecules. This lateral signal results in LIN‐12/NOTCH activation in P5.p and P7.p, which induces the 2 fate and prevents them from adopting the 1 vulval fate (Sternberg, 1988; Yoo et al., 2004). P3.p, P4.p, and P8.p receive insuYcient levels of either signal and thus adopt the 3 vulval fate. The RAS/MAPK signaling cascade involves an array of negative

192

Braendle et al.

5. Environmental Responses in C. elegans

193

and positive regulators (Fig. 4B). Components of a third signaling pathway (WNT) regulate LIN‐39/HoxB5 expression and maintain competence of vulval precursor cells during the L2 stage (Eisenmann et al., 1998). In addition, WNT signaling may play a later role in inducing vulval cell fates, since hyperactivation of this pathway is suYcient for vulva induction when RAS/MAPK signaling is compromised (Gleason et al., 2002). During the L3 stage, this patterning process causes invariant fate‐specific cell divisions, resulting in a total of 22 vulval cells (Fig. 4A). Vulval morphogenesis takes place during the L4 stage and the complete vulval organ is formed by the final moult to the adult.

B. Environmental Robustness of the Final Vulva Phenotype C. elegans vulval development has been mainly studied using the N2 isolate in standard laboratory conditions. In these conditions, deviations from the canonical vulval cell fate pattern are generally rare (Delattre and Fe´lix, 2001). This result indicates that the final vulval phenotype of N2 is robust to stochastic variation within the standard environment. We have further examined how the final vulval phenotype responds to variation in the laboratory environment (C. B. and M.‐A. F., unpublished data). Tested environments included standard conditions at diVerent temperatures (15–25  C), starvation in the L2 stage, liquid culture, and conditions in which individuals passed through the dauer stage. In all conditions, defects leading to an abnormal vulva (e.g., hypoinduction or cell lineage errors) were rarely more frequent than in standard conditions. Thus, the process of vulval formation appears robust to a variety of environmental conditions. However, we found additional deviations from the wild‐type fate patterns adopted by P3.p–P8.p that did not lead to defects. The frequency of such variants diVered between environmental conditions. For example, individuals in the starvation environment showed a weak, yet consistent propensity to center their vulva on P5.p instead of P6.p. The occurrence of an anterior shift in vulval centering was largely limited to this environment and was never observed in standard environments. This result indicates that specific environmental conditions may elicit specific

Figure 4 Vulval development. (A) Vulval precursor cell fate specification. (1) Establishment and maintenance of the vulval competence group in the L1 and L2 stages. (2) Specification of the pattern of three vulval precursor cell fates during the early L3 stage. (3) Cell lineages. AC, anchor cell; T, transverse (left–right) division; L, longitudinal (anteroposterior) division; U, undivided; S, fusion to the epidermal syncytium (non‐vulval fate). (B) Feedback loops and cross talk between EGF/RAS/MAPK and NOTCH pathways during vulval cell fate patterning (see text). WNT signaling is not shown here, but may be involved in the specification of vulval cell fates.

194

Braendle et al.

developmental changes while maintaining the production of a robust phenotype (as vulval centering on P5.p does not result in an apparent loss of functionality).

C. Environmental Sensitivity of Vulva Developmental Processes While the final vulval phenotype is robust and shows little environmental variability, it is currently unclear to what extent underlying molecular processes may vary in diVerent environmental conditions. However, several lines of evidence suggest that vulval development is sensitive to environmental inputs. In a screen for mutants showing defects in vulval formation, Ferguson and Horvitz (1985) reported variation in the penetrance of several loss‐of‐function mutations in genes involved in the EGF/RAS/MAPK signaling cascade. Such mutations cause vulval defects due to hypoinduction, leading to an induction of fewer than three vulval precursor cells. Starvation conditions and passage through the dauer stage resulted in a drastic suppression of some of these mutant phenotypes relative to normal growth conditions. Furthermore, starvation and dauer conditions aVected the penetrance of a diVerent set of mutations, indicating that the two conditions may have diVerent eVects on vulval development. Two recent studies indicate in more detail how environmental and physiological inputs may be transduced to modify mechanisms underlying vulval cell fate patterning. Battu et al. (2003) showed that starvation conditions might modulate vulval induction via the G‐coupled receptor (SRA‐13) and the G‐protein GPA‐5, which are expressed in body wall muscles and sensory neurons. In their experimental assay, starvation stimuli caused a decrease in vulval induction as inferred from mutant analyses, indicating that the EGF/ RAS/MAPK cascade may be modified by environmental inputs. However, where this input acts to aVect vulval induction is not known. Another study (Moghal et al., 2003) reveals how the environment may cause physiological changes to modify pathway activities during vulval patterning. Culture in liquid medium led to suppression of loss‐of‐function mutants of the EGF/ RAS/MAPK cascade, suggesting that liquid growth may increase vulval induction levels. The observed suppression requires the heterotrimeric Gq protein (EGL‐30, expressed in neurons) and the voltage‐gated calcium channel (EGL‐19, expressed in muscle cells). This result shows that environmental modification of vulval induction could be mediated by neuronal and muscle cells. The environmental eVect transduced by EGL‐30 acts in parallel or downstream of the EGF/RAS/MAPK pathway. As the eVect requires WNT signaling via BAR‐1/ ‐catenin, the signals involved may act through the WNT pathway. This is consistent with the hypothesis that partial redundancy of WNT and EGF/RAS/MAPK pathways during vulval induction

5. Environmental Responses in C. elegans

195

may be important to maintain robustness of the final cell fate output pattern in varying environments. Hence, vulval development is sensitive to a number of environmental conditions. Specific environmental factors may modify the activities of signaling pathways underlying this developmental process. However, it is currently unclear to what extent such modifications occur in the wild‐type situation as they were only described in sensitized (mutant) genetic backgrounds. In addition, it is unknown how such environmental modifications relate to the apparent robustness of the final vulval phenotype in the corresponding environmental conditions. D. Developmental and Molecular Features Causing Robustness to Variation in the Environment The current data suggest that the vulval phenotype is robust to a number of environmental conditions. In this section, we discuss some of the molecular and developmental features that are likely to cause tolerance to changes in developmental parameters induced by the environment (see also Fe´lix and Wagner, in press). 1. Cellular Redundancy As mentioned earlier, the wild‐type vulva is derived from three vulval precursors, P5.p–P7.p, whereas a total of six cells can potentially adopt vulval fates (P3.p–P8.p). Experimental elimination of one or several of the central cells allows replacement among competent cells, which ensures the formation of a complete vulva (Sulston and White, 1980). In addition, the extent of the competence group allows correct vulval formation if the anchor cell is centered on cells other than P6.p. Thus, cellular replacement within the vulval competence group can maintain the complete and correct vulval pattern (2 1 2 ) when the cell fate patterning is disturbed. 2. Pathway Redundancy Both an intermediate level of LIN‐3/EGF signal and activation of the LIN‐12/ NOTCH pathway may generate a 2 vulval fate (Katz et al., 1995; Koga and Ohshima, 1995; Simske and Kim, 1995). While the contribution of LIN‐3/ EGF signal to 2 fate induction is not essential, the two signaling pathways show partial functional redundancy in specifying the 2 vulva cell fate. As mentioned earlier, vulval induction may occur in the absence of LET‐60/ RAS signaling if the WNT signaling pathway is overactivated (Gleason et al., 2002). Thus, the two pathways can show functional redundancy in vulval induction, indicating that WNT signal may compensate for loss in LET‐60/RAS signal.

196

Braendle et al.

3. Pathway Properties The switch‐like behavior of the EGF/RAS/MAPK cascade causes insensitivity to small EGF fluctuation: low doses have no eVect, while for any dose above a certain threshold the pathway will show a much higher level of activity (Huang and Ferrell, 1996). In C. elegans, such switch behavior may be reinforced by numerous positive feedback loops, for example, through EPS‐8 activation (Stetak et al., 2006) or DEP‐1 inhibition (Berset et al., 2005) (Fig. 4B). Negative feedback loops, such as the SEM‐5/ARK‐1 feedback loop, which inhibits EGFR activation (Hopper et al., 2000), might contribute to the buVering against noise (Becskei and Serrano, 2000) (Fig. 4B). These features are likely to limit not only the eVects of stochastic variation, but also the eVects of environmental and genetic variation. The RAS/MAPK signaling cascade actually shows a large number of regulators that are functionally redundant in terms of their eVect on the cell fate pattern: inactivation of one of them does not alter the vulva pattern output, except in sensitized genetic backgrounds (Fig. 4B). Such silent regulators include both positive (e.g., EPS‐8, KSR‐1, KSR‐2) and negative (e.g., ARK‐1, GAP‐1, LIP‐1, UNC‐101, SLI‐1) inputs to the RAS pathway (Berset et al., 2001; Hajnal et al., 1997; Hopper et al., 2000; Jongeward et al., 1995; Lee et al., 1994; Ohmachi et al., 2002; Stetak et al., 2006). This observation demonstrates buVering of LET‐60/RAS activity as mutation of these regulators seem to change LET‐60/RAS activity, yet not suYciently to cause an abnormal vulval phenotype. Furthermore, it is also possible that diVerent regulators per se perform buVering functions to adjust LET‐60/RAS signaling when development is perturbed, for example, in diVerent environmental conditions. As for the RAS/MAPK pathway, functional redundancy has been observed among genetic components in NOTCH (Chen and Greenwald, 2004) and WNT (Gleason et al., 2006) pathways during vulval cell fate specification. Further properties indicating robustness to pathway activity changes are provided by the fact that mutations in lin‐3/egf and other Ras pathway genes are haplo‐suYcient and that mild overexpression of lin‐3/egf still produces a wild‐type vulval phenotype (J. M. and M.‐A. F., unpublished data).

4. Pathway Cross Talk The interplay between the Ras and NOTCH pathways results in stable discrimination between the 1 and 2 fates (Berset et al., 2001; Yoo et al., 2004). High Ras pathway activity in P6.p causes the adoption of the 1 fate and triggers the degradation of the NOTCH receptor, preventing the adoption of 2 fate (Shaye and Greenwald, 2002). At the same time, release of DELTA‐like proteins toward P5.p and P7.p achieves the activation of a 2 vulval fate in these cells and prevent their adoption of a 1 fate through

5. Environmental Responses in C. elegans

197

the activation of negative regulators of the Ras pathway (Chen and Greenwald, 2004). As shown by modeling approaches, lateral inhibition ensures much stronger fate segregation between adjacent cells compared to fate segregation induced by a morphogen gradient alone (Giurumescu et al., 2006).

5. Gene Redundancy In addition to these network properties, functionally redundant genes arising from gene duplication seem to play a minor role in conferring robustness of vulval formation to stochastic and environmental variation. For example, in the RAS pathway, there is one known case of partial functional redundancy between two positive regulators (KSR‐1 and KSR‐2) (Ohmachi et al., 2002).

E. Evolution of Vulval Development Despite intraspecific conservation of the final vulval phenotype, analysis of distinct C. elegans wild isolates suggests evolution of the underlying molecular mechanisms. Introgression of vulval mutations (isolated in N2) into diVerent wild isolates reveals that a given mutational eVect is highly variable depending on the wild genetic background (J. M. and M.‐A. F., unpublished results). Moreover, C. elegans wild isolates show diVerences in how vulval formation responds to environmental variation. After starvation, vulval (mis)centering on P5.p is frequent in N2; in another isolate (JU258), vulval centering on P5.p is very rare but other deviations from the wild‐type pattern, such as fusion of vulval precursor cells are very frequent (C. B. and M.‐A. F., unpublished results). The vulval phenotype (the fate pattern and the corresponding cell division patterns) of C. elegans is conserved in all known Caenorhabditis species. Such extensive evolutionary conservation of a phenotype is typical for a robust trait under stabilizing selection. Recent experiments show that this morphological stasis underlies considerable evolutionary divergence in underlying developmental mechanisms. Anchor cell ablation experiments at diVerent developmental time points and mild EGF overexpression in diVerent Caenorhabditis species reveal many diVerences in the regulation of vulval development (Fe´lix, 2007). Thus, intra‐ and interspecific features of vulval development indicate that the underlying genetic architecture of a robust and evolutionarily invariant phenotype can evolve. However, the genetic changes responsible for this evolution have not yet been identified. In addition, it remains a major challenge to determine to what extent these evolutionary diVerences might result as a consequence of the robustness of the developmental system.

198

Braendle et al.

Figure 5 Genotype–phenotype maps illustrating scenarios of phenotypic plasticity and environmental robustness. In both cases, a given genotype interacts with two environments (E1 and E2) and these interactions cause variation in the intermediate phenotype (e.g., developmental, physiological, metabolic phenotype). In the case of phenotypic plasticity, variation in the intermediate phenotype causes variation in the final phenotype of the process (e.g., cell fate, morphology, life history trait) (confer to the example of phenotypic plasticity of dauer formation). In the case of environmental robustness, variation in the intermediate phenotype does not cause variation in the final phenotype (confer to the example of environmental robustness of vulva formation). Note that the extent of both environmental and stochastic (within environment) variation may change between intermediate and final phenotypes.

VII. Conclusion Research on C. elegans illustrates the multitude of interactions between organism and environment. These interactions occur at all levels of biological organization and evolve readily. To gain a complete understanding of how a genotype maps into a phenotype and how this mapping process evolves, it is particularly important to study the developmental role of the environment (Fig. 5).

Acknowledgments We thank Michael Ailion, Jean‐Louis Bessereau, Jonathan Ewbank, Thomas Flatt, Simon Harvey Anne‐Franc¸oise Ruaud, and Mark Viney for helpful comments on a previous version of this review. C. B. thanks Thomas Flatt for insightful and extensive discussions on GxE and G&T. C. B. is holder of a Marie Curie individual fellowship (EIF) and acknowledges previous financial support by the Swiss National Science Foundation. J. M. is supported by a Ph.D. fellowship from the Ministry of Research (France). Work in our laboratory is funded by the Centre National de la Recherche Scientifique and grants from the Association pour la Recherche sur le Cancer (no. 3749) and from the Agence Nationale de la Recherche (ANR‐05‐BLAN‐0231‐01).

5. Environmental Responses in C. elegans

199

References Ailion, M., and Thomas, J. H. (2000). Dauer formation induced by high temperatures in Caenorhabditis elegans. Genetics 156, 1047–1067. Ailion, M., and Thomas, J. H. (2003). Isolation and characterization of high‐temperature‐ induced Dauer formation mutants in Caenorhabditis elegans. Genetics 165, 127–144. Altun, Z. F., and Hall, D. H. (2005). Hermaphrodite anatomy. In ‘‘WormAtlas’’ (Z.F. Altun and D.H. Hall, Eds.) http://www.wormatlas.org/handbook/excretory.htm. Ancel Meyers, L., and Bull, J. J. (2002). Fighting change with change: Adaptive variation in an uncertain world. Tree 17, 551–557. Anderson, G. L. (1978). Responses of dauer larvae of Caenorhabditis elegans (Nematoda: Rhabditidae) to thermal stress and oxygen deprivation. Can. J. Zool. 56, 1786–1791. Ao, W., Gaudet, J., Kent, W., Muttumu, S., and Mango, S. (2004). Environmentally induced foregut remodeling by PHA‐4/FoxA and DAF‐12/NHR. Science 305, 1743–1746. Aroian, R. V., Lesa, G. M., and Sternberg, P. W. (1994). Mutations in the Caenorhabditis elegans let‐23 EGFR‐like gene define elements important for cell‐type specificity and function. EMBO J. 13, 360–366. Ayyadevara, S., Ayyadevara, R., Hou, S., Thaden, J. J., and Shmookler, R. J. (2001). Genetic mapping of quantitative trait loci governing longevity of Caenorhabditis elegans in recombinant‐inbred progeny of a Bergerac BO X RC301 interstrain cross. Genetics 157, 655–666. Ayyadevara, S., Ayyadevara, R., Vertino, A., Galecki, A., Thaden, J., and Reis, R. (2003). Genetic loci modulating fitness and life span in Caenorhabditis elegans: Categorical trait interval mapping in CL2a x Bergerac‐BO recombinant‐inbred worms. Genetics 163, 557–570. Azevedo, R. B. R., Keightley, P. D., Laure´n‐Ma¨a¨tta¨, C., Vassilieva, L. L., Lynch, M., and Leroi, A. M. (2002). Spontaneous mutational variation for body size in Caenorhabditis elegans. Genetics 162, 755–765. Badyaev, A. V. (2005). Stress‐induced variation in evolution: From behavioural plasticity to genetic assimilation. Proc. Biol. Sci. 272, 877–886. Bargmann, C. (2006). Chemosensation in C. elegans. In ‘‘The C. elegans Research Community’’ (WormBook, Ed.), doi/10.1895/wormbook.1.7.1, http://www.wormbook.org. Bargmann, C. I., and Horvitz, H. R. (1991). Control of larval development by chemosensory neurons in Caenorhabditis elegans. Science 251, 1243–1246. Barrie`re, A., and Fe´lix, M.‐A. (2005a). High local genetic diversity and low outcrossing rate in Caenorhabditis elegans natural populations. Curr. Biol. 15, 1176–1184. Barrie`re, A., and Fe´lix, M.‐A. (2005b). Natural variation and population genetics of C. elegans. In ‘‘The C. elegans Research Community’’ (WormBook, Ed.), doi/10.1895/wormbook.1.7.1, http://www.wormbook.org. Barrie`re, A., and Fe´lix, M.‐A. (2007). Temporal dynamics and linkage disequilibrium in natural C. elegans populations. Genetics 176, 999–1011. Battu, G., Hoier, E. F., and Hajnal, A. (2003). The C. elegans G‐protein‐coupled receptor SRA‐13 inhibits RAS/MAPK signalling during olfaction and vulval development. Development 130, 2567–2577. Baugh, L. R., and Sternberg, P. W. (2006). DAF‐16/FOXO regulates transcription of cki‐1/Cip/ Kip and repression of lin‐4 during C. elegans L1 arrest. Curr. Biol. 16, 780–785. Becskei, A., and Serrano, L. (2000). Engineering stability in gene networks by autoregulation. Nature 405, 590–593. Berset, T., Hoier, E. F., Battu, G., Canevascini, S., and Hajnal, A. (2001). Notch inhibition of RAS signaling through MAP kinase phosphatase LIP‐1 during C. elegans vulval development. Science 291, 1055–1058.

200

Braendle et al.

Berset, T. A., Hoier, E. F., and Hajnal, A. (2005). The C. elegans homolog of the mammalian tumor suppressor Dep‐1/Scc1 inhibits EGFR signaling to regulate binary cell fate decisions. Genes Dev. 19, 1328–1340. Bird, D., and Opperman, C. (1998). Caenorhabditis elegans: A genetic guide to parasitic nematode biology. J. Nematol. 30, 299–308. Blaxter, M. (2003). Nematoda: Genes, genomes and the evolution of parasitism. Adv. Parasitol. 54, 101–195. Brand, A., and Hawdon, J. M. (2004). Phosphoinositide‐3‐OH‐kinase inhibitor LY294002 prevents activation of Ancylostoma caninum and Ancylostoma ceylanicum third‐stage infective larvae. Int. J. Parasitol. 34, 909–914. Candido, E. P. (2002). The small heat shock proteins of the nematode Caenorhabditis elegans: Structure, regulation and biology. Prog. Mol. Subcell. Biol. 28, 61–78. Carvalho, S., Barriere, A., and Pires‐Dasilva, A. (2006). The world of a worm: A framework for Caenorhabditis evolution. Workshop on the study of evolutionary biology with Caenorhabditis elegans and closely related species. EMBO Rep. 7, 981–984. Cassada, R. C., and Russell, R. L. (1975). The dauer larva, a post‐embryonic developmental variant of the nematode Caenorhabditis elegans. Dev. Biol. 46, 326–342. Caswell‐Chen, E. P., Chen, J., Lewis, E. E., Douhan, G. W., Nadler, S. A., and Carey, J. R. (2005). Revising the standard wisdom of C. elegans natural history: Ecology of longevity. Sci. Aging Knowl. Environ. 40, pe30. Chen, N., and Greenwald, I. (2004). The lateral signal for LIN‐12/Notch in C. elegans vulval development comprises redundant secreted and transmembrane DSL proteins. Dev. Cell 6, 183–192. Cherkasova, V., Ayyadevara, S., Egilmez, N., and Reis, R. (2000). Diverse Caenorhabditis elegans genes that are upregulated in dauer larvae also show elevated transcript levels in long‐lived, aged, or starved adults. J. Mol. Biol. 300, 433–448. Cheung, B. H., Arellano‐Carbajal, F., Rybicki, I., and de Bono, M. (2004). Soluble guanylate cyclases act in neurons exposed to the body fluid to promote C. elegans aggregation behavior. Curr. Biol. 14, 1105–1111. Clark, S. G., Chisholm, A. D., and Horvitz, H. R. (1993). Control of cell fates in the central body region of C. elegans by the homeobox gene. lin‐39. Cell 74, 43–55. Couillault, C., Pujol, N., Reboul, J., Sabatier, L., Guichou, J. F., Kohara, Y., and Ewbank, J. (2004). TLR‐independent control of innate immunity in Caenorhabditis elegans by the TIR domain adaptor protein TIR‐1, an ortholog of human SARM. Nat. Immunol. 5, 488–494. Cutter, A. D. (2006). Nucleotide polymorphism and linkage disequilibrium in wild populations of the partial selfer Caenorhabditis elegans. Genetics 172, 171–184. Dalley, B. K., and Golomb, M. (1992). Gene expression in the Caenorhabditis elegans dauer larva: Developmental regulation of Hsp90 and other genes. Dev. Biol. 151, 80–90. Darby, C. (2006). Interactions with microbial pathogens. In ‘‘The C. elegans Research Community’’ (WormBook, Ed.), doi/10.1895/wormbook.1.7.1, http://www.wormbook.org. Davies, A. G., Bettinger, J. C., Thiele, T. R., Judy, M. E., and McIntire, S. L. (2004). Natural variation in the npr‐1 gene modifies ethanol responses of wild strains of C. elegans. Neuron 42, 731–743. de Bono, M., and Bargmann, C. I. (1998). Natural variation in a neuropeptide Y receptor homolog modifies social behavior and food response in C. elegans. Cell 94, 679–689. de Visser, J. A., Hermisson, J., Wagner, G. P., Ancel Meyers, L., Bagheri‐Chaichian, H., Blanchard, J. L., Chao, L., Cheverud, J. M., Elena, S. F., Fontana, W., Gibson, G., Hansen, T. F., et al. (2003). Perspective: Evolution and detection of genetic robustness. Evol. Int. J. Org. Evol. 57, 1959–1972. Delattre, M., and Fe´lix, M.‐A. (2001). Polymorphism and evolution of vulval precursor cell lineages within two nematode genera, Caenorhabditis and Oscheius. Curr. Biol. 1, 631–643.

5. Environmental Responses in C. elegans

201

Dusenbery, D. B. (1996). ‘‘Life at Small Scale—The Behavior of Microbes.’’ HPHLP, New York. Ebert, R. H., Shammas, M. A., Sohal, B. H., Sohal, R. S., Egilmez, N. K., Ruggles, S., and Shmookler, R. R. (1996). Defining genes that govern longevity in Caenorhabditis elegans. Dev. Genet. 18, 131–143. Emmons, S. (2005). Sexual behavior of the Caenorhabditis elegans male. Int. Rev. Neurobiol. 69, 99–123. Eisenmann, D. M., Maloof, J. N., Simske, J. S., Kenyon, C., and Kim, S. K. (1998). The ‐catenin homolog BAR‐1 and LET‐60 Ras coordinately regulate the Hox gene lin‐39 during Caenorhabditis elegans vulval development. Development 125, 3667–3680. Ewbank, J. J. (2006). Signaling in the immune response. In ‘‘The C. elegans Research Community’’ ( WormBook, Ed.), doi/10.1895/wormbook.1.7.1, http://www.wormbook.org. Fe´lix, M. A. (2007). Cryptic quantitative evolution of the vulva intercellular signaling network in Caenorhabditis. Curr. Biol. 17, 103–114. Fe´lix, M.‐A., and Wagner, A. Robustness and evolution: Concepts, insights and challenges from a developmental model system. Heredity, in press. Ferguson, E., and Horvitz, H. R. (1985). Identification and characterization of 22 genes that aVect the vulval cell lineages of Caenorhabditis elegans. Genetics 110, 17–72. Flatt, T. (2005). The evolutionary genetics of canalization. Q. Rev. Biol. 80, 287–316. Flemming, A., Shen, Z.‐Z., Cunha, A., Emmons, S. W., and Leroi, A. M. (2000). Somatic polyploidization and cellular proliferation drive body size evolution in nematodes. Proc. Natl. Acad. Sci. USA 97, 5285–5290. Fodor, A., Riddle, D. R., Nelson, F. K., and Golden, J. W. (1983). Comparison of a new wild‐ type Caenorhabditis briggsae with laboratory strains of C. briggsae and C. elegans. Nematologica 29, 203–217. Friedenberg, N. (2003). Experimental evolution of dispersal in spatiotemporally variable microcosms. Ecol. Lett. 6, 953–959. Fujiwara, M., Sengupta, P., and McIntire, S. (2002). Regulation of body size and behavioral state of C. elegans by sensory perception and the EGL‐4 cGMP‐dependent protein kinase. Neuron 36, 1091–1102. Fukuyama, M., Rougvie, A., and Rothman, J. (2006). C. elegans DAF‐18/PTEN mediates nutrient‐dependent arrest of cell cycle and growth in the germline. Curr. Biol. 16, 773–779. Gems, D., and Riddle, D. L. (2000). Genetic, behavioral and environmental determinants of male longevity in Caenorhabditis elegans. Genetics 154, 1597–1610. Gibson, G., and Dworkin, I. (2004). Uncovering cryptic genetic variation. Nat. Rev. Genet. 5, 681–90. Giles, A., Rose, J., and Rankin, C. (2005). Investigations of learning and memory in Caenorhabditis elegans. Int. Rev. Neurobiol. 69, 37–71. Giurumescu, C. A., Sternberg, P. W., and Asthagiri, A. R. (2006). Intercellular coupling amplifies fate segregation during Caenorhabditis elegans vulval development. Proc. Natl. Acad. Sci. USA 103, 1331–1336. Gleason, J. E., Korswagen, H. C., and Eisenmann, D. M. (2002). Activation of Wnt signaling bypasses the requirement for RTK/Ras signaling during C. elegans vulval induction. Genes Dev. 16, 1281–1290. Gleason, J. E., Szyleyko, E. A., and Eisenmann, D. M. (2006). Multiple redundant Wnt signaling components function in two processes during C. elegans vulval development. Dev. Biol. 298, 442–457. Golden, J. W., and Riddle, D. L. (1984a). A pheromone‐induced developmental switch in Caenorhabditis elegans. Temperature‐sensitive mutants reveal a wild‐type temperature‐ dependent process. Proc. Natl. Acad. Sci. USA 81, 819–823.

202

Braendle et al.

Golden, J. W., and Riddle, D. L. (1984b). The Caenorhabditis elegans dauer larva: Developmental eVects of pheromone, food, and temperature. Dev. Biol. 102, 368–378. Goldschmidt, R. (1938). ‘‘Physiological Genetics.’’ McGraw‐Hill, New York. Goranson, N., Ebersole, J., and Brault, S. (2005). Resolving an adaptive conundrum: Reproduction in Caenorhabditis elegans is not sperm‐limited when food is scarce. Evol. Ecol. Res. 7, 325–333. Gray, J. M., Karow, D. S., Lu, H., Chang, A. J., Chang, J. S., Ellis, R. E., Marletta, M. A., and Bargmann, C. I. (2004). Oxygen sensation and social feeding mediated by a C. elegans guanylate cyclase homologue. Nature 430, 317–322. Grewal, P. S. (1991). Influence of bacteria and temperature on the reproduction of Caenorhabditis elegans (Nematoda: Rhabditidae) infesting mushrooms (Agaricus bisporus). Nematologica 37, 72–82. Gutteling, E. W., Riksen, J. A., Bakker, J., and Kammenga, J. E. (2006). Mapping phenotypic plasticity and genotype‐environment interactions aVecting life‐history traits in Caenorhabditis elegans. Heredity 98, 28–37. Haag, E. S., Chamberlin, H., Coghlan, A., Fitch, D. H., Peters, A. D., and Schulenburg, H. (2007). Caenorhabditis evolution: If they all look alike, you aren’t looking hard enough. Trends Genet. 23, 101–104. Haber, M., Schu¨ngel, M., Putz, A., Mu¨ller, S., Hasert, B., and Schulenburg, H. (2005). Evolutionary history of Caenorhabditis elegans inferred from microsatellites: Evidence for spatial and temporal genetic diVerentiation and the occurrence of outbreeding. Mol. Biol. Evol. 22, 160–173. Hajnal, A., Whitfield, C. W., and Kim, S. K. (1997). Inhibition of Caenorhabditis elegans vulval induction by gap‐1 and by let‐23 receptor tyrosine kinase. Genes Dev. 1, 2715–2728. Harvey, S. C., and Viney, M. E. (2007). Thermal variation reveals natural variation between isolates of Caenorhabditis elegans. J. Exp. Zool. B Mol. Dev. Evol., in press. Held, J. M., White, M. P., Fisher, A. L., Gibson, B. W., Lithgow, G. J., and Gill, M. S. (2006). DAF‐12‐dependent rescue of dauer formation in Caenorhabditis elegans by (25S)‐cholestenoic acid. Aging Cell 5, 283–291. Herndon, L. A., Schmeissner, P. J., Dudaronek, J. M., Brown, P. A., Listner, K. M., Sakano, Y., Paupard, M. C., Hall, D. H., and Driscoll, M. (2002). Stochastic and genetic factors influence tissue‐specific decline in ageing C. elegans. Nature 419, 808–814. Hobert, O. (2005). Specification of the nervous system. In ‘‘The C. elegans Research Community’’ (WormBook, Ed.), doi/10.1895/wormbook.1.7.1, http://www.wormbook.org. Hodgkin, J., and Doniach, T. (1997). Natural variation and copulatory plug formation in Caenorhabditis elegans. Genetics 146, 149–164. Holt, S. J., and Riddle, D. L. (2003). SAGE surveys C. elegans carbohydrate metabolism: Evidence for an anaerobic shift in the long‐lived dauer larva. Mech. Ageing Dev. 124, 779–800. Hope, I. A. (1999). C. elegans—A practical approach. In ‘‘The Practical Approach Series’’ (B. D. Hames, Ed.), p. 281. Oxford University Press, Oxford, NY. Hopper, N. A., Lee, J., and Sternberg, P. W. (2000). ARK‐1 inhibits EGFR signaling in C. elegans. Mol. Cell 6, 65–75. Horvitz, H., and Sternberg, P. (1991). Multiple intercellular signaling systems control the development of the Caenorhabditis elegans vulva. Nature 351, 535–541. Horvitz, H., Chalfie, M., Trent, C., Sulston, J. E., and Evans, P. D. (1982). Serotonin and octopamine in the nematode C. elegans. Science 216, 1012–1014. Huang, C.‐Y. F., and Ferrell, J. E., Jr. (1996). Ultrasensitivity in the mitogen‐activated protein kinase cascade. Proc. Natl. Acad. Sci. USA 93, 10078–10083. Inoue, T., Takamura, K., Yamae, H., Ise, N., Kawakami, M., Tabuse, Y., Miwa, J., and Yamaguchi, Y. (2003). Caenorhabditis elegans DAF‐21 (HSP90) is characteristically and predominantly expressed in germline cells: Spatial and temporal analysis. Dev. Growth DiVer. 45, 369–376.

5. Environmental Responses in C. elegans

203

Jablonka, E., and Lamb, E. J. (1995). ‘‘Epigenetic Inheritance and Evolution—The Lamarckian Dimension.’’ Oxford University Press, Oxford, NY. Jeong, P. Y., Jung, M., Yim, Y. H., Kim, H., Park, M., Hong, E., Lee, W., Kim, Y. H., Kim, K., and Paik, Y. K. (2005). Chemical structure and biological activity of the Caenorhabditis elegans dauer‐inducing pheromone. Nature 433, 541–545. Jiang, H., Guo, R., and Powell‐CoVman, J. A. (2001). The Caenorhabditis elegans hif‐1 gene encodes a bHLH‐PAS protein that is required for adaptation to hypoxia. Proc. Natl. Acad. Sci. USA 98, 7916–7921. Johnson, T. E., and Hutchinson, E. W. (1993). Absence of strong heterosis for life‐span and other life‐history traits in Caenorhabditis elegans. Genetics 134, 465–474. Jones, S. J., Riddle, D. L., Pouzyrev, A. T., Velculescu, V. E., Hillier, L., Eddy, S. R., Stricklin, S. L., Baillie, D. L., Waterston, R., and Marra, M. A. (2001). Changes in gene expression associated with developmental arrest and longevity in Caenorhabditis elegans. Genome Res. 11, 1346–1352. Jongeward, G. D., Clandinin, T. R., and Sternberg, P. W. (1995). sli‐1, a negative regulator of let‐23‐mediated signaling in C. elegans. Genetics 139, 1553–1566. Jovelin, R., Ajie, B. C., and Phillips, P. C. (2003). Molecular evolution and quantitative variation for chemosensory behaviour in the nematode genus Caenorhabditis. Mol. Ecol. 12, 1325–1337. Katz, W., Hill, R., Clandinin, T., and Sternberg, P. (1995). DiVerent levels of the C. elegans growth factor LIN‐3 promote distinct vulval precursor fates. Cell 82, 297–307. Kessin, R. H., Gundersen, G. G., Zaydfudim, V., Grimson, M., and Blanton, R. L. (1996). How cellular slime molds evade nematodes. Proc. Natl. Acad. Sci. USA 93, 4857–4861. Kimble, J., and Hirsh, D. (1979). Post‐embryonic cell lineages of the hermaphrodite and male gonads in C. elegans. Dev. Biol. 70, 396–417. Kiontke, K., and Sudhaus, W. (2006). Ecology of Caenorhabditis species. In ‘‘The C. elegans Research Community’’ (WormBook, Ed.), doi/10.1895/wormbook.1.7.1, http://www.wormbook.org. Klass, M., and Hirsh, D. (1976). Non‐ageing developmental variant of Caenorhabditis elegans. Nature 260, 523–525. Knight, C. G., Azevedo, R. B., and Leroi, A. M. (2001). Testing life‐history pleiotropy in Caenorhabditis elegans. Evolution Int. J. Org. Evolution 55, 1795–1804. Koga, M., and Ohshima, Y. (1995). Mosaic analysis of the let‐23 gene function in vulval induction of Caenorhabditis elegans. Development 121, 2655–2666. Koga, M., Zwaal, R., Guan, K.‐L., Avery, L., and Ohshima, Y. (2000). A Caenorhabditis elegans MAP kinase kinase, MEK‐1, is involved in stress responses. EMBO J. 19, 5148–5156. Lamitina, S. T., Morrison, R., Moeckel, G. W., and Strange, K. (2004). Adaptation of the nematode Caenorhabditis elegans to extreme osmotic stress. Am. J. Physiol., Cell Physiol. 286, C785–C791. Lee, J., Jongeward, G. D., and Sternberg, P. W. (1994). unc‐101, a gene required for many aspects of Caenorhabditis elegans development and behavior, encodes a clathrin‐associated protein. Genes Dev. 8, 60–73. Lee, Q., and Widden, P. (1996). Folsomia candida, a ‘‘fungivorous’’ collembolan, feeds preferentially on nematodes rather than soil fungi Soil Biol. Biochem. 28, 689–690. Lehtinen, M. K., Yuan, Z., Boag, P. R., Yang, Y., Villen, J., Becker, E. B., DiBacco, S., de la Iglesia, N., Gygi, S., Blackwell, T. K., and Bonni, A. (2006). A conserved MST‐ FOXO signaling pathway mediates oxidative‐stress responses and extends life span. Cell 125, 987–1001. Lewontin, R. C. (1974). The analysis of variance and the analysis of causes. Am. J. Hum. Genet. 26, 400–411. Li, W., Kennedy, S. G., and Ruvkun, G. (2003). daf‐28 encodes a C. elegans insulin superfamily member that is regulated by environmental cues and acts in the DAF‐2 signaling pathway. Genes Dev. 17, 844–858.

204

Braendle et al.

Li, Y., Alvarez, O. A., Gutteling, E. W., Tijsterman, M., Fu, J., Riksen, J. A. G., Hazendonk, E., Prins, P., Plasterk, R. H. A., Jansen, R. C., Breitling, R., and Kammenga, J. E. (2006). Mapping determinants of gene expression plasticity by genetical genomics in C. elegans. PLoS Genet. 2, 2155–2161. Lu, R., Maduro, M., Li, F., Li, H. W., Broitman‐Maduro, G., Li, W. X., and Ding, S. W. (2005). Animal virus replication and RNAi‐mediated antiviral silencing in Caenorhabditis elegans. Nature 436, 1040–1043. Maloof, J. N., and Kenyon, C. (1998). The Hox gene lin‐39 is required during C. elegans vulval induction to select the outcome of Ras signaling. Development 125, 181–190. Maupas, E. (1900). Modes et formes de reproduction des ne´matodes. Arch. Zool. Exp. 8, 463–624. McCulloch, D., and Gems, D. (2003). Evolution of male longevity bias in nematodes. Aging Cell 2, 165–173. Meiklejohn, C. D., and Hartl, D. L. (2002). A single mode of canalization. Trends Ecol. Evol. 17, 468–473. Menzel, R., Rodel, M., Kulas, J., and Steinberg, C. E. (2005). CYP35: Xenobiotically induced gene expression in the nematode Caenorhabditis elegans. Arch. Biochem. Biophys. 438, 93–102. Moghal, N., Garcia, L. R., Khan, L. A., Iwasaki, K., and Sternberg, P. W. (2003). Modulation of EGF receptor‐mediated vulva development by the heterotrimeric G‐protein G(alpha)q and excitable cells in C. elegans. Development 130, 4553–4566. Mohri, A., Kodama, E., Kimura, K. D., Koike, M., Mizuno, T., and Mori, I. (2005). Genetic control of temperature preference in the nematode Caenorhabditis elegans. Genetics 169, 1437–1450. Morck, C., and Pilon, M. (2006). C. elegans feeding defective mutants have shorter body lengths and increased autophagy. BMC Dev. Biol. 6, 39. Motola, D. L., Cummins, C. L., Rottiers, V., Sharma, K. K., Li, T., Li, Y., Suino‐Powell, K., Xu, H. E., Auchus, R. J., Antebi, A., and Mangelsdorf, D. J. (2006). Identification of ligands for DAF‐12 that govern dauer formation and reproduction in C. elegans. Cell 124, 1209–1223. Murphy, C. T. (2006). The search for DAF‐16/FOXO transcriptional targets: Approaches and discoveries. Exp. Gerontol. 41, 910–921. Nicholas, H. R., and Hodgkin, J. (2004). The ERK MAP kinase cascade mediates tail swelling and a protective response to rectal infection in C. elegans. Curr. Biol. 14, 1256–1261. Nigon, V. (1949). Modalite´s de la reproduction et de´terminisme du sexe chez quelques ne´matodes libres. Ann. Sci. Nat., Zool. Biol. Anim. 11, 1–132. Nijhout, H. F. (1999). Control mechanisms of polyphenic development. Bioscience 49, 181–192. Nijhout, H. F. (2003). Development and evolution of adaptive polyphenisms. Evol. Dev. 5, 9–18. Ohmachi, M., Rocheleau, C. E., Church, D., Lambie, E., Schedl, T., and Sundaram, M. V. (2002). C. elegans ksr‐1 and ksr‐2 have both unique and redundant functions and are required for MPK‐1 ERK phosphorylation. Curr. Biol. 12, 427–433. Olsen, A., Vantipalli, M. C., and Lithgow, G. J. (2006). Using Caenorhabditis elegans as a model for aging and age‐related diseases. Ann. NY Acad. Sci. 1067, 120–128. O’rourke, D., Baban, D., Demidova, M., Mott, R., and Hodgkin, J. (2006). Genomic clusters, putative pathogen recognition molecules, and antimicrobial genes are induced by infection of C. elegans with M. nematophilum. Genome Res. 16, 1005–1016. Pigliucci, M. (2001). ‘‘Phenotypic Plasticity: Beyond Nature and Nurture.’’ Johns Hopkins University Press, Baltimore, MD. Pradel, E., Zhang, Y., Pujol, N., Matsuyama, T., Bargmann, C. I., and Ewbank, J. J. (2007). Detection and avoidance of a natural product from the pathogenic bacterium Serratia marescens by Caenorhabditis elegans. Proc. Natl. Acad. Sci. USA 104, 2295–2300.

5. Environmental Responses in C. elegans

205

Pujol, N., and Ewbank, J. (2005). Pathogen avoidance using toll signaling in C. elegans. In ‘‘Toll and Toll‐Like Receptors: An Immunologic Perspective’’ (T. Rich, Ed.), pp. 162–167. Kluwer, Academic/Plenum Publications, New York. Rankin, C. H. (2000). Context conditioning in habituation in the nematode Caenorhabditis elegans. Behav. Neurosci. 114, 496–505. Reichert, K., and Menzel, R. (2005). Expression profiling of five diVerent xenobiotics using a Caenorhabditis elegans whole genome microarray. Chemosphere 61, 229–237. Ren, P., Lim, C.‐S., Johnsen, R., Albert, P. S., Pilgrim, D., and Riddle, D. L. (1996). Control of C. elegans larval development by neuronal expression of a TGF‐Beta homolog. Science 274, 1389–1391. Riddle, D. L. (1988). The dauer larva. In ‘‘The Nematode Caenorhabditis elegans’’ (W. B. Wood, Ed.), pp. 393–412. Cold Spring Harbor Press, Cold Spring Harbor, NY. Riddle, D. L., and Albert, P. S. (1997). Genetic and environmental regulation of dauer larva development. In ‘‘C. elegans’’ II (D. L. Riddle, T. Blumenthal, B. J. Meyer, and J. R. Priess, Eds.), pp 739–768. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Robertson, H. M. (2000). The large srh family of chemoreceptor genes in Caenorhabditis nematodes reveals processes of genome evolution involving large duplications and deletions and intron gains and losses. Genome Res. 10, 192–203. Rutherford, S. L., and Lindquist, S. (1998). Hsp90 as a capacitor for morphological evolution. Nature 396, 336–342. Savage‐Dunn, C., Maduzia, L. L., Zimmerman, C. M., Roberts, A. F., Cohen, S., Tokarz, R., and Padgett, R. W. (2003). Genetic screen for small body size mutants in C. elegans reveals many TGF beta pathway components. Genesis 35, 239–247. Schackwitz, W. S., Inoue, T., and Thomas, J. H. (1996). Chemosensory neurons function in parallel to mediate a pheromone response in C. elegans. Neuron 17, 719–728. Schafer, W. R. (2006). Egg‐laying. In ‘‘The C. elegans Research Community’’ (WormBook, Ed.), doi/10.1895/wormbook.1.7.1, http://www.wormbook.org. Schlichting, C. D., and Pigliucci, M. (1998). ‘‘Phenotypic Evolution: A Reaction Norm Perspective.’’ Sinauer, Sunderland, MA. Schulenburg, H., and Ewbank, J. (2004). Diversity and specificity in the interaction between Caenorhabditis elegans and the pathogen Serratia marcescens. BMC Evol. Biol. 4, 49. Schulenburg, H., and Muller, S. (2004). Natural variation in the response of Caenorhabditis elegans towards Bacillus thuringiensis. Parasitology 128, 433–443. Shapira, M., Hamlin, B. J., Rong, J., Chen, K., Ronen, M., and Tan, M. W. (2006). A conserved role for a GATA transcription factor in regulating epithelial innate immune responses. Proc. Natl. Acad. Sci. USA 103, 14086–14091. Shaye, D. D., and Greenwald, I. (2002). Endocytosis‐mediated downregulation of LIN‐12/ Notch upon Ras activation in Caenorhabditis elegans. Nature 420, 686–690. Shen, C., Nettleton, D., Jiang, M., Kim, S. K., and Powell‐CoVman, J. A. (2005). Roles of the HIF‐1 hypoxia‐inducible factor during hypoxia response in Caenorhabditis elegans. J. Biol. Chem. 280, 20580–20588. Shook, D. R., and Johnson, T. E. (1999). Quantitative trait loci aVecting survival and fertility‐ related traits in Caenorhabditis elegans show genotype‐environment interactions, pleiotropy and epistasis. Genetics 153, 1233–1243. Shook, D. R., Brooks, A., and Johnson, T. E. (1996). Mapping quantitative trait loci aVecting life history traits in the nematode Caenorhabditis elegans. Genetics 142, 801–817. Simske, J. S., and Kim, S. K. (1995). Sequential signalling during Caenorhabditis elegans vulval induction. Nature 375, 142–146. Sivasundar, A., and Hey, J. (2005). Sampling from natural populations with RNAi reveals high outcrossing and population structure in Caenorhabditis elegans. Curr. Biol. 15, 1598–1602.

206

Braendle et al.

Staniland, L. N. (1957). The swarming of rhabditid eelworms in mushroom houses. Plant Pathol. 6, 61–62. Stearns, S. C. (1992). ‘‘The Evolution of Life Histories.’’ Oxford University Press, Oxford. Stein, L. D., Bao, Z., Blasiar, D., Blumenthal, T., Brent, M. R., Chen, N., Chinwalla, A., Clarke, L., Clee, C., Coghlan, A., Coulson, A., D’Eustachio, P., et al. (2003). The genome sequence of Caenorhabditis briggsae: A platform for comparative genomics. PLoS Biol. 1, 166–192. Sternberg, P. W. (1988). Lateral inhibition during vulval induction in Caenorhabditis elegans. Nature 335, 551–554. Sternberg, P. W. (2005). Vulval development. In ‘‘The C. elegans Research Community’’ (WormBook, Ed.), doi/10.1895/wormbook.1.7.1, http://www.wormbook.org. Sternberg, P. W., and Horvitz, H. R. (1986). Pattern formation during vulval development in Caenorhabditis elegans. Cell 44, 761–772. Stetak, A., Hoier, E. F., Croce, A., Cassata, G., Di, F. P. P., and Hajnal, A. (2006). Cell fate‐ specific regulation of EGF receptor traYcking during Caenorhabditis elegans vulval development. EMBO J. 25, 2347–2357. Stewart, M. K., Clark, N. L., Merrihew, G., Galloway, E. M., and Thomas, J. H. (2005). High genetic diversity in the chemoreceptor superfamily of Caenorhabditis elegans. Genetics 169, 1985–1996. Sulston, J. E., and Horvitz, H. R. (1977). Postembryonic cell lineages of the nematode. Caenorhabditis elegans. Dev. Biol. 56, 110–156. Sulston, J. E., Schierenberg, E., White, J. G., and Thomson, J. N. (1983). The embryonic cell lineage of the nematode. Caenorhabditis elegans. Dev. Biol. 100, 64–119. Sulston, J. E., and White, J. G. (1980). Regulation and cell autonomy during postembryonic development of C. elegans. Dev. Biol. 78, 577–597. Suo, S., Kimura, Y., and Van Tol, H. H. (2006). Starvation induces cAMP response element‐ binding protein‐dependent gene expression through octopamine‐Gq signaling in Caenorhabditis elegans. J. Neurosci. 26, 10082–10090. Szewczyk, N. J., Kozak, E., and Conley, C. A. (2003). Chemically defined medium and Caenorhabditis elegans. BMC Biotechnol. 3, 19. Szewczyk, N., Udranszky, I., Kozak, E., Sunga, J., Kim, S., Jacobson, L., and Conley, C. (2006). Delayed development and lifespan extension as features of metabolic lifestyle alteration in C. elegans under dietary restriction. J. Exp. Biol. 209, 4129–4139. Tenaillon, O., Denamur, E., and Matic, I. (2004). Evolutionary significance of stress‐induced mutagenesis in bacteria. Trends Microbiol. 12, 264–270. Teotonio, H., Manoel, D., and Phillips, P. C. (2006). Genetic variation for outcrossing among Caenorhabditis elegans isolates. Evolution Int. J. Org. Evolution 60, 1300–1305. Thomas, J. H. (2006). Concerted evolution of two novel protein families in Caenorhabditis species. Genetics 172, 2269–2281. Thomas, J. H., Kelley, J. L., Robertson, H. M., Ly, K., and Swanson, W. J. (2005). Adaptive evolution in the SRZ chemoreceptor families of Caenorhabditis elegans and Caenorhabditis briggsae. Proc. Natl. Acad. Sci. USA 102, 4476–4481. Tissenbaum, H. A., Hawdon, J., Perregaux, M., Hotez, P., Guarente, L., and Ruvkun, G. (2000). A common muscarinic pathway for diapause recovery in the distantly related nematode species Caenorhabditis elegans and Ancylostoma caninum. Proc. Natl. Acad. Sci. USA 97, 460–465. Troemel, E. R., Kimmel, B. E., and Bargmann, C. I. (1997). Reprogramming chemotaxis responses: Sensory neurons define olfactory preferences in C. elegans. Cell 91, 161–169. Van Voorhies, W. A., Fuchs, J., and Thomas, S. (2005). The longevity of Caenorhabditis elegans in soil. Biol. Lett. 1, 247–249.

5. Environmental Responses in C. elegans

207

Van Voorhies, W. A., and Ward, S. (2000). Broad oxygen tolerance in the nematode Caenorhabditis elegans. J. Exp. Biol. 203, 2467–2478. Viney, M. E., Gardner, M. P., and Jackson, J. A. (2003). Variation in Caenorhabditis elegans dauer larva formation. Dev. Growth DiVer. 45, 389–396. Viney, M. E., Thompson, F. J., and Crook, M. (2005). TGF‐beta and the evolution of nematode parasitism. Int. J. Parasitol. 35, 1473–1475. Waddington, C. H. (1942). Canalization of development and the inheritance of acquired characters. Nature 150, 563–565. Waddington, C. H. (1956). Genetic assimilation of the bithorax phenotype. Evol. Int. J. Org. Evol. 10, 1–13. Walker, G., White, T., McColl, G., Jenkins, N., Babich, S., Candido, E. P., Johnson, T. E., and Lithgow, G. J. (2001). Heat shock protein accumulation is upregulated in a long‐lived mutant of Caenorhabditis elegans. J. Gerontol. A Biol. Sci. Med. Sci. 56, B281–B287. Walker, G. A., Thompson, F. J., Brawley, A., Scanlon, T., and Devaney, E. (2003). Heat shock factor functions at the convergence of the stress response and developmental pathways in Caenorhabditis elegans. FASEB J. 17, 1960–1962. Wang, J., and Kim, S. K. (2003). Global analysis of dauer gene expression in Caenorhabditis elegans. Development 130, 1621–1634. Wang, J., Tokarz, R., and Savage‐Dunn, C. (2002). The expression of TGF beta signal transducers in the hypodermis regulates body size in C. elegans. Development 129, 4989–4998. Wes, P. D., and Bargmann, C. I. (2001). C. elegans odour discrimination requires asymmetric diversity in olfactory neurons. Nature 410, 698–701. West‐Eberhard, M. J. (2003). ‘‘Developmental Plasticity and Evolution.’’ Oxford University Press, Oxford, NY. Wilkins, C., Dishongh, R., Moore, S. C., Whitt, M. A., Chow, M., and Machaca, K. (2005). RNA interference is an antiviral defence mechanism in Caenorhabditis elegans. Nature 436, 1044–1047. Wood, W. B. (1988). ‘‘The Nematode Caenorhabditis elegans.’’ Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. Yoo, A. S., Bais, C., and Greenwald, I. (2004). Crosstalk between the EGFR and LIN‐12/Notch pathways in C. elegans vulval development. Science 303, 663–666. Zhang, Y., Lu, H., and Bargmann, C. I. (2005). Pathogenic bacteria induce aversive olfactory learning in Caenorhabditis elegans. Nature 438, 179–184.

6

____________________________________________________________________________

Molluscan Shell Proteins: Primary Structure, Origin, and Evolution Fre´de´ric Marin,* Gilles Luquet,* Benjamin Marie,* and Davorin Medakovic{ *UMR CNRS 5561 ‘Bioge´osciences,’ Universite´ de Bourgogne 6 Boulevard Gabriel, 21000 DIJON, France { Center for Marine Research Rovinj, Ruder Boskovic Institute 5 Giordano Paliaga, 52210 ROVINJ, Croatia

I. Introduction: The Shell, a Biologically Controlled Mineralization II. Molluscan Shell Formation: Developmental Aspects A. The Larval Shell B. The Juvenile and Adult Shell C. Transient Amorphous Calcium Carbonate III. The Topographic Models of Shell Mineralization A. Early Nacre Descriptions and Models B. Recent Nacre Models and Evolving Views C. Prism Models IV. Molluscan Shell Proteins: Characterization of Their Primary Structure A. Extremely Acidic Shell Proteins B. Moderately Acidic Shell Proteins C. Basic Shell Proteins D. Partially Characterized Shell Proteins E. Other Molluscan Proteins: The Extrapallial Fluid and the Mantle F. Remarks on Molluscan Shell Proteins V. Origin and Evolution of Molluscan Shell Proteins A. The Cambrian Origin of Mollusk Shell Mineralization B. The ‘‘Ancient Heritage’’ Scenario C. The ‘‘Recent Heritage and Fast Evolution’’ Scenario D. Long‐Term Evolution of Shell Matrices and Microstructures: The Bivalve Example VI. Concluding Remarks Acknowledgments References

In the last few years, the field of molluscan biomineralization has known a tremendous mutation, regarding fundamental concepts on biomineralization regulation as well as regarding the methods of investigation. The most recent advances deal more particularly with the structure of shell biominerals at nanoscale and the identification of an increasing number of shell matrix protein components. Although the matrix is quantitatively a minor constituent in the Current Topics in Developmental Biology, Vol. 80 Copyright 2008, Elsevier Inc. All rights reserved.

209

0070-2153/08 $35.00 DOI: 10.1016/S0070-2153(07)80006-8

210

Marin et al.

shell of mollusks (less than 5% w/w), it is, however, the major component that controls diVerent aspects of the shell formation processes: synthesis of transient amorphous minerals and evolution to crystalline phases, choice of the calcium carbonate polymorph (calcite vs aragonite), organization of crystallites in complex shell textures (microstructures). Until recently, the classical paradigm in molluscan shell biomineralization was to consider that the control of shell synthesis was performed primarily by two antagonistic mechanisms: crystal nucleation and growth inhibition. New concepts and emerging models try now to translate a more complex reality, which is remarkably illustrated by the wide variety of shell proteins, characterized since the mid‐1990s, and described in this chapter. These proteins cover a broad spectrum of pI, from very acidic to very basic. The primary structure of a number of them is composed of diVerent modules, suggesting that these proteins are multifunctional. Some of them exhibit enzymatic activities. Others may be involved in cell signaling. The oldness of shell proteins is discussed, in relation with the Cambrian appearance of the mollusks as a mineralizing phylum and with the Phanerozoic evolution of this group. Nowadays, the extracellular calcifying shell matrix appears as a whole integrated system, which regulates protein–mineral and protein–protein interactions as well as feedback interactions between the biominerals and the calcifying epithelium that synthesized them. Consequently, the molluscan shell matrix may be a source of bioactive molecules that would oVer interesting perspectives in biomaterials and biomedical fields. ß 2008, Elsevier Inc.

I. Introduction: The Shell, a Biologically Controlled Mineralization Biomineralization refers to the dynamic physiological process by which a living organism elaborates a mineralized structure. Biomineralization refers also to the final product, the mineralized structure, whatever it is, a rigid skeleton or a nonskeletal mineralization (Lowenstam and Weiner, 1989). In living systems, biominerals display a wide range of functions: tissues support, UV protection, shelter against predation, nutrition, reproduction, gravity, light or magnetic field perceptions, storage of mineral ions (Simkiss and Wilbur, 1989). In the metazoan world, calcium carbonate skeletons are the most commonly encountered biomineralizations, and the most abundant, from diploblastic animals, sponges, and corals to deuterostomes, echinoderms, and chordates. Among mollusks, calcium carbonate biomineralization exhibits a huge diversity of morphologies (Lowenstam and Weiner, 1989): epithelial spicules, scales and plates, operculum, intracellular detoxifying granules, egg capsules, love dart, pearls, statoconia, and statoliths, but the most well‐known molluscan calcium carbonate biomineralization is the shell, the primary function of which is to

6. Molluscan Shell Proteins

211

support these soft‐bodied organisms and protect them from predation and desiccation. The molluscan shell is an organo‐mineral composite, where the dominant mineral—aragonite, or calcite, or in particular cases, vaterite—is intimately associated to an organic matrix, which accounts only for 0.1–5% of the shell weight. This matrix represents amalgamate of proteins, glycoproteins, chitin and acidic polysaccharides, secreted by the calcifying tissues during skeletogenesis. This mixture is consequently sealed within the skeleton during its growth. At macroscopic level, the adjunction of organic components to a mineralized structure enhances the mechanical properties to the whole organo‐mineral assembly. At molecular level, the matrix plays a key role in the mineralization process. According to the terminology introduced by Stephen Mann (1983), the construction of the shell is the archetype of a biologically controlled mineralization. This concept can be summarized by five identification criteria. (1) The process requires specialized cellular machinery, which means that the minerals formed are not just by‐products of the metabolic activity but correspond to a specialized metabolic pathway. (2) The mineral synthesis is an active process, that is, the minerals are synthesized far from the equilibrium with the environment. (3) The formed minerals are diVerent in their shape and size from their inorganically formed counterpart. (4) The minerals are not formed in direct contact with the environment, that is, the organism has developed a strategy for delimiting the space where the minerals are synthesized. (5) The biomineralization process is mediated by an extracellular organic matrix. The molluscan shell complies with all these criteria. During decades, the molluscan shell matrix was considered as a single entity and there were considerable eVorts to propose hypotheses on its putative functions (Bevelander and Nakahara, 1969; Krampitz et al., 1976; Lowenstam and Weiner, 1989; Simkiss and Wilbur, 1989; for a review, see Marin and Luquet, 2004). Although these functions are generally accepted, they have been mainly deduced from detailed micro‐ and ultrastructural observations of the final product (SEM, TEM), from physical measurements (XRD), from biochemical characterizations, and/or from in vitro tests that poorly mimic the real conditions. Until now, a large part of the shell biomineralization process, that is, the mysterious transition from precursor fluids to the final product, the solid shell, still escapes our comprehension, and the self‐assembling capacity of shell matrices remains a ‘‘black box.’’ In spite of our ignorance in knowing each step of the secretory sequence that leads to the shell, several putative functions are attributed to the associated matrix, as listed as follows: the shell matrix presumably concentrates locally the precursors ions; it constitutes a tridimensional framework, acts as a template for crystals, and allows the nucleation of crystals only where appropriate; it selects the calcium carbonate polymorph; it controls crystal elongation in privileged crystallographic axes and inhibits crystal growth by poisoning

212

Marin et al.

their faces; it determines the spatial arrangement of crystal units at diVerent scales, from nanometer to millimeter. Beside these physicochemical aspects of matrix–mineral interactions, the organic matrix is likely to display enzymatic functions and to be involved in cell signaling. Proteins and glycoproteins represent essential components of the shell matrix, and there are at least three good reasons for studying these macromolecular components from a fundamental point of view. First, obtaining the structure of shell proteins may allow a better understanding of their respective functions and may help to refine the biomineralization model. This in turn may provide solid cues for studying the phylogenetic relationships of the diVerent mineralized tissues and for understanding how these systems were formed. At last, because the shell is a closed system, which sooner or later becomes an integral part of the fossil record, knowledge on the biochemical properties of shell proteins may help to trace their diagenetic evolution during burial and fossilization and, more generally, may provide a strong basis for analyzing the diagenesis of skeletal carbonates. Other reasons for studying molluscan shell proteins lie in the fascinating and challenging applied perspectives that these proteins oVer (Mann, 2001; Marin et al., 2007). For example, shell proteins may be employed in nanotechnologies, for micromanipulating nanocrystals and semiconductors. They can be used for biomimicry purposes, that is, the synthesis at room temperature of composite materials, which exhibit high‐mechanical properties. Molluscan shell proteins may also be used as natural bioactive factors, in particular in bone surgery. The best example is provided by the osteoinductive/osteogenic properties of nacre matrix and nacre implants. Another domain, which may benefit from advances in the knowledge on shell proteins, is the pearl industry, a major economical activity in the South Pacific area. At last, molluscan shell proteins may be used as natural biodegradable antiscaling and antifouling agents. The aim of this chapter is to review our present knowledge on molluscan shell proteins and to resituate them in an evolutionary framework. However, in order to bring a dynamic view of the system, we will first describe some developmental aspects of the shell and present evolving views on the models of molluscan mineralization.

II. Molluscan Shell Formation: Developmental Aspects A. The Larval Shell A review on the diVerent modes of embryonic and postembryonic mollusk development is far beyond the scope of this chapter, and we advise the reader to refer to the very detailed review of Nielsen (2004) for the early

213

6. Molluscan Shell Proteins A

B

60 µm

70 µm

Figure 1 SEM picture of veliger larvae of M. galloprovincialis. The soft tissues were removed and the shells were cleaned with dilute NaOCl (0.026% active chlorine, 1 hour). (A) Larval shell in the later « D » stage. White arrows indicate prodissoconch I/II boundary, which marks the moment when the valves hermetically enclose larval body. The valves are equal in convexity and dimension of prodissoconch I is on the order of 80–90 mm. The surface of this layer is characterized by nearly smooth to small ‘‘pitted–furrowed’’ zones with faint radial striations. Prodissoconch II is distinctly co‐marginally striated. On the ventral side of the well‐calcified left shell valve (black arrow), linear prodissoconch I hinge is visible. (B) The inner part of prodissoconch I hinge is simple and still not interlocking the valves. At the start of prodissoconch II development, the central part of the hinge is disconnected. The valves fit together by tiny rugosities, starting denticulations on the edge of hinge (black arrows). From this primordial structure during further larval development, hinge teeth and ligament, umbo of the shell will be formed. Calcified portion contains tiny unequal granules, which become smaller toward the prodissoconch I/II boundary.

developmental stages. Let us however recall few general considerations about mollusk development and the phylogenetic position of the group. Mollusks are triploblastic protostomial (the blastopore gives the mouth of the adult) schizocoelomates (the coelomic cavity is produced by the splitting of the mesoderm). Within protostomes, mollusks belong to the lophotrochozoan superphylum (Aguinaldo et al., 1998; Halanych et al., 1995), together with brachiopods, bryozoans, annelids, platyhelminthes, acanthocephala, and some minor phyla. They are eutrochozoans, the characteristic of which is to produce a swimming‐ciliated larva, the trochophore larva (Lecointre and Le Guyader, 2001). In the early embryonic stages, the mollusk development exhibits several similarities with that of annelids, a trait that brings these two phyla in a close phylogenetic relationship (Nielsen, 2004). Contrarily to vertebrate or sea urchin eggs, the molluscan egg undergoes a determinate spiral cleavage (except for cephalopods), which knows several variations from species to species. The first cleavages are, for a majority of species, unequal, and one of the cells produced after the second division, the D cell, is bigger than the three others. The subsequent unequal divisions of the D cell produce the 2d micromere, which will give the cells that produce

214

Marin et al.

the shell. This general scheme knows some exceptions: for the gastropod Patella, the shell gland develops mainly from 2a and 2c micromeres (Dictus and Damen, 1997). It is striking to notice that, in gastropods, the sense of the third cleavage (the first spiral cleavage, right‐ or left‐handed) determines the sense of the shell coiling, but the link between the two remains obscure. The shell coiling seems under the control of a recessive gene called sinistral (Schilthuizen and Davison, 2005). In mollusks, two types of postembryonic development are observed (Martoja, 1995). First, an indirect development is shared by most of molluscan classes (in particular bivalves and gastropods) and characterized by a transition from a ciliated trochophore to a veliger larva and a metamorphosis from the veliger to a juvenile. The first transition implies the acquisition of a velum used for swimming. The veliger larval stage is typical of mollusks. In some cases, like in the freshwater unionid bivalves, the veliger larva, called glochidium, adopts a parasitic life mode on fish gills. In gastropods, the veliger stage corresponds to the larval torsion, which twists the head and foot by 180
relative to the shell, mantle, and visceral mass. The main metamorphosis occurs when the pelagic veliger larva settles down for a benthic existence. This transformation is profound and corresponds to the disappearance of the velum, the development of the foot, and the organization of the digestive gland and of the reproductive organs (Bonar, 1976). The second mode of development is direct, without larval stages neither metamorphosis, which implies that juveniles look like adults in reduction. This most derived developmental mode is particular of cephalopods and is characterized for most of them, except Nautilus, by an internalization of the shell. The following description is mainly applied to bivalves and gastropods, the two main classes by their number of species. The first steps of the larval shell formation occur during the trochophore stage. The shell—of ectodermic origin—is produced by a group of cells located on the posterior side of the larva. These cells define the shell field (Kniprath, 1981). The shell field is diVerentiated at the end of the gastrulation stage, by the thickening of the median portion of the ectoderm in the post‐ trochal dorsal region (Moor, 1983). Cells of the shell field invaginate, according to various pathways described by Kniprath (1981), and this invagination produces the transitory shell gland, also called preconchylian gland. According to Kniprath (1977, 1980), the invagination is functionally required, for allowing the cells at the periphery of the shell gland (cells that are not internalized during invagination) to produce the early organic membrane, which will be the first template for calcium carbonate minerals deposition. This organic lamella is the future periostracum. Let us remind that the periostracum is the leathery outer layer of the shell, particularly visible in species like the edible mussel, Mytilus edulis. In the next step, the shell gland gradually flattens and/or evaginates and spreads by mitotic divisions, while

6. Molluscan Shell Proteins

215

transforming into the larval mantle epithelium. In the meantime, the early periostracum expands. Between the periostracum and the cells of the shell field, the primary mineralization takes place. In bivalves, the first shell, secreted during the trochophore stage, is called prodissoconch I and is characterized by a granular aspect (Mao Che et al., 2001). During the transition from trochophore to veliger stage, the prodissoconch II shell is secreted and marks a change in the secretory regime (Fig. 1). The prodissoconch II stage is indeed characterized by the appearance of growth lines on the valves. It is suggested that the cells that secrete the prodissoconch II shell are diVerent from those which produce the prodissoconch I (Mao Che et al., 2001). Following metamorphosis of the veliger larva, the dissoconch shell is formed. Again, this change is marked on the outer surface of the shell by an accentuated growth line. As mentioned by Jablonski and Lutz (1980), the terminology used for describing the successive shell stages in gastropods is diVerent: the first shell formed at late trochophore stage is then called protoconch I, the second shell secreted during the veliger larval stage, protoconch II, and the postmetamorphosis shell, teleoconch. In comparison to other phyla, in particular echinoderms and arthropods, the connection between the physiology of the larval shell development and the underlying genetic machinery, which controls and patterns the process, is supported by a limited number of studies in mollusks. Of outstanding interest are the works of Moshel et al. (1998), Jacobs et al. (2000), Wanninger and Haszprunar (2001), Klerkx et al. (2001), Nederbragt et al. (2002), and Hinman et al. (2003). In particular, the first four cited papers underline the key role played by engrailed (En) in molluscan shell development. The En class encodes homeodomain‐containing DNA‐binding proteins involved in major steps of metazoan development (Hidalgo, 1996). These multifunctional transcription factors are, among others, involved in the patterning of the nervous system (neurogenesis); in the body segmentation in annelids, arthropods, and vertebrates; and in several other derived functions, such as the specification of the ventral compartment in vertebrate limbs, or the patterning of the mid‐hindbrain boundary (Gibert, 2002). Besides working at transcriptional level, En also modulates translation and seems to be able to act as morphogens (Morgan, 2006). En has been identified in most metazoan lineages, including mollusks (Wray et al., 1995). In the marine mud snail (Ilyanassa) embryo, the expression of En is localized only in the shell gland (Moshel et al., 1998). In the trochophore larva of the tusk shell Antalis, En is expressed in shell‐secreting cells at the border of the protoconch. However, after metamorphosis, En expression was not observed in the cells that produce the adult shell, the teleoconch (Wanninger and Haszprunar, 2001). In the chiton, En is expressed in region that bound skeletal plates, and in the clam, En expression surrounds each developing valve and the hinge (Jacobs et al., 2000). According to these authors, En would be directly involved in

216

Marin et al.

skeletogenesis by marking the skeletal boundaries. They also claim that En would display the same function in other calcifying bilaterian metazoans. This suggests that the acquisition of a calcified skeleton may be a unique event across metazoan phylogeny, which would explain the sudden emergence of calcified skeletons at the dawn of the Cambrian times. This view on the direct role of En in skeletogenesis was contested by Nederbragt et al. (2002), for whom En is primarily involved in delimitating compartment boundaries between cells of the shell gland and the other ectodermic cells. If so, the contribution of En to shell formation is undirect. The paper of Hinman et al. (2003) underlines the major role of Hox1 and Hox4 during the development of the abalone larva. In the trochophore larva, Hox1 is expressed in a ring of cells corresponding to the outer mantle edge. Hox4 is expressed in the mantle, but at later stage, after the larval shell is fully formed. Hinman et al. suggest that both Hox genes may have been co‐opted into a role in patterning shell. At last, some developmental genes may also indirectly contribute to the shell formation, by their absence of expression in the shell‐forming cells. This is the case for E32, a gene encoding a putative RNA‐binding protein, not expressed in the shell gland of Patella, but expressed in the cells, which are maintained in an undiVerentiated state (Klerkx et al., 2001). E32 would block the cell diVerentiation process. Clearly, further studies, as the ones described here, are required for mollusk phylum before a complete picture of the role of developmental genes in shell formation can emerge. Another important aspect in the formation of the embryonic shell deals with the enzymatic activity that occurs during the whole process of larval development. This aspect has, however, been widely neglected. In the freshwater snail, Lymanea, the old study of Timmermans (1969) showed that the expression of alkaline phosphatase (ALP) was the highest during the evagination process, while the expressions of DOPA‐oxidase (tyrosinase) and peroxidase were maximal at the borders of the shell gland, after evagination. This zone corresponds to the zone where the periostracum is secreted. In the edible mussel, Mytilus, the level of carbonic anhydrase was recorded during the whole developmental process. In larvae, high expressions of carbonic anhydrase were found to precede the formation of the shell field in the gastrula stage, the formation of the shell gland and periostracum in the trochophore stage, and the mineral deposition in the prodissoconch I and prodissoconch II stages (Medakovic, 2000). In the embryo of the freshwater snail Biomphalaria glabrata, the temporal and spatial activities of ALP, peroxidase, and acid phosphatase were analyzed by histochemical staining (Marxen et al., 2003b,c). An ALP activity was observed in trochophore larva in the invaginated shell field (shell gland), prior secretion of any shell material. A peroxidase activity was found in small vesicles of cells involved in the secretion of the periostracum. Acid phosphatase was localized in the shell

6. Molluscan Shell Proteins

217

field and around the shell field invagination. At last, a recent key paper of Weiss et al. (2006) underlined the importance of chitin synthase, a transmembrane glycosyltransferase, which synthesizes chitin. In situ hybridization experiments performed on larvae of the mussel Mytilus galloprovincialis showed that the chitin synthase transcript was present in early and late veliger stages, in the cells in close contact with the larval shell. A striking finding is the presence of a myosin motor head domain in the intracellular N‐terminus of the identified chitin synthase. This clearly suggests, for the first time, that the cytoskeleton plays a crucial, although poorly understood, role in chitin formation.

B. The Juvenile and Adult Shell Once the metamorphosis of the veliger has occurred, the resulting juvenile mollusk calcifies rapidly, and its shell growth approximately follows a von BertalanVy law (Seed, 1980), during the whole life of the animal. Classically, the physiology of molluscan shell calcification can be described as a succession of compartments (Wilbur and Saleuddin, 1983), where the central element is the mantle, the thin organ, which coats the inner surface of the shell, the other compartments being the extrapallial space and the shell. The mantle is a polarized tissue, and comprises an inner epithelium, in contact with the ambient medium (e.g., seawater), the mantle interior, which comprises pallial muscles, connective tissues, nerve fibers, and finally the outer epithelium. As shown in Fig. 2, the outer epithelium is the epithelium, which mineralizes the shell. Whether this epithelium is in direct contact with the shell is still debated (see below). The extremity of the mantle is characterized by a succession of folds, three among bivalves two among gastropods. The ridge between the outer and median folds defines the periostracal groove, which secretes the periostracum. As described in Section II.A, the primary role of the periostracum is to provide a support for the mineralization. Its second important role is to delimitate and seal the space where the mineralization takes place. Actually, the invention of the periostracum corresponds to an old strategy that mollusks have set up for mineralizing in a confined space, the extrapallial space. The periostracum is secreted as a liquid film of tyrosine‐rich precursors, which rapidly becomes insoluble and sclerotized by a quinone‐tanning process (Waite, 1983). Precursor ions—calcium and bicarbonate—are taken up from the ambient medium, through the inner epithelium, or from the gill; both ions can also originate from the mollusk metabolism (food and fluids). They transit in the connective tissues of the mantle via the hemolymph—the interstitial fluid that bathes the mollusk cells—and are directed toward the outer mantle epithelium. Calcium and bicarbonate ions can be actively extruded from

218

Marin et al.

Prismatic layer

Nacreous layer

Extrapallial space Periostracum Outer epithelium

Pallial muscle Periostracal groove

Mucocyte

Inner epithelium

Figure 2 Physiology of the shell calcification in a nacro‐prismatic bivalve, redrawn from Saleuddin and Petit (1983). The calcification takes place at the distal border of the shell in the extrapallial space. In the outer epithelium, the cells responsible for the deposition of nacre are not localized in the same area as the ones responsible for the secretion of prism precursors. The prisms and nacre tablets are not drawn to scale.

the cytosol to the second compartment involved in shell formation, the extrapallial space (Fig. 2). However, in several cases, calcium can be temporarily stored as intracellular or extracellular amorphous granules. This has been particularly well studied for bivalves (Fournie´ and Che´tail, 1982; Istin, 1970; Istin and Girard, 1970) for which a colocalization of granules and carbonic anhydrase was observed. Amorphous granules can be used by the cells for detoxifying the cytosol from an excess of calcium or heavy metals (Simkiss, 1977, 1993). Granules are also a source of calcium, which is rapidly available, in particular, for a rapid shell mineralization and repair. The possibility of outer epithelial cells to release intracellular calcium granules by exocytosis is not documented, although suspected. The second—already mentioned—compartment is the extrapallial space (Fig. 2). This space is supposed to concentrate the precursor mineral ions, owing to calcium and bicarbonate pumps. However, these pumps have not

6. Molluscan Shell Proteins

219

been localized on the outer epithelium, mainly because they have been poorly characterized at the molecular level (Endo et al., 2003). The extrapallial space is supposed to be the place where the ‘‘mysterious’’ transition from the liquid precursors to the solid biominerals occurs. The concept of self‐assembling is currently put forward for explaining how the shell matrix and the mineral ions interact in a controlled manner to produce a finely structured organo‐ mineral material. Chemical analyses of the extrapallial fluid have shown that it is supersaturated with respect to calcium carbonate (Misogianes and Chasteen, 1979; Moura et al., 2000). This fluid is also enriched in mineral/ organic mineralization inhibitors, without which calcium carbonate would precipitate anarchically. This obviously never happens. The extrapallial space is also the place where the shell matrix is secreted and where the subtle transition from liquid to solid operates. As said in the introduction, and detailed later, this matrix is a mixture of proteins, glycoproteins, acidic polysaccharides, chitin, and presumably lipids. The matrix interacts with the mineral ions and controls the shape of the produced crystals. It is consumed by the system, that is, ‘‘entrapped’’ within the construction in progress. This means that it has to be brought to the mineralization front when required during the whole mineralization period. It is also likely that several proteins, which are present in the extrapallial fluid, are not incorporated in the shell. Another unknown parameter is the precise temporal sequence of secretion of the shell matrix. On the other hand, we know that the secretory regime is diVerent depending on the position of the cells involved in this process on the outer epithelium. This has been demonstrated, both at transcriptional and protein levels, in particular for species, which exhibit a bitextured shell (an outer prismatic layer, an inner nacreous layer; Fig. 2). The outer epithelial cells that secrete the matrix involved in controlling the prism formation occupy a more distal position (from the shell hinge) than the ones that secrete the matrix involved in the nacre deposition (Jolly et al., 2004; Sudo et al., 1997). A paper from Takeuchi and Endo (2005) confirms the existence of two zones in the outer mantle epithelium, with transcripts specific of the prism zone, transcripts specific of the nacre zone, and transcripts present in both. Attempts to introduce an epithelial cell typology have been made (Sud, 2002). At last, because the deposition of calcium carbonate is accompanied by the release of protons, these latter have to be reabsorbed by the calcifying epithelium, for precluding acidification of the extrapallial fluid, and possible resolubilization of the newly formed minerals. It has been suggested that proton pumps are involved in extruding protons from the extrapallial fluid toward the cytosol. However, this process is poorly documented for mollusks (Coimbra et al., 1988). It has also been suggested that proton pumps work, in some cases, in the reverse sense, inducing then acidosis in the extrapallial space (Moura et al., 2003).

220

Marin et al.

The classical view described here above seems a priori ‘‘well established.’’ However, several ‘‘evidences,’’ like the transit of precursor ions, are indirect or deduced from the general metabolism of the calcifying mantle. Several physiological aspects are left in the dark or bring unsatisfactory answer. For example, our description attributes a key role to the extrapallial fluid. This suggests that there are no direct contacts between the secreting epithelium and the location where mineralization occurs. This view of a ‘‘remote control of the mineralization by the epithelium’’ is questionable (Addadi et al., 2006). Some authors consider that the mantle cells have to keep close contact with the mineralization front in order to receive feedback molecular signals from the newly formed biominerals. Another point, which remains obscure, is the function of hemocytes—free circulating cells of the hemolymph—in the shell construction process. Hemocytes play a major role in the immune defense of mollusks (Glinski and Jarosz, 1997) and in tissue repair (Serpentini et al., 2000). Their role in shell repair processes has already been underlined (Bubel et al., 1977; Watabe, 1983), but their contribution to the normal shell formation may be widely underestimated. A milestone paper from Mount et al. (2004) attributes to hemocytes (of the granulocyte type) of the oyster Crassostrea virginica, an important function in the normal shell construction process, by releasing calcite crystals—not amorphous granules—which can be remodeled at the mineralization site. So far, it is diYcult to evaluate whether this process is particular to the studied model or whether it is a general metabolic pathway that mollusks use for mineralizing their shell.

C. Transient Amorphous Calcium Carbonate One aspect, which has for a long time been passed largely unnoticed, but which needs to be urgently reevaluated, is the key role played by amorphous calcium carbonate (ACC) in molluscan shell formation and, more generally, in all calcium carbonate biomineralizations (Weiner et al., 2003). A mineral phase is considered to be amorphous when it lacks the long‐range order, that is, when it does not have the regular repeating lattice structure that provides the basis for so many of the features of a crystal (Simkiss, 1993). Several amorphous minerals can exhibit a short‐range order, which is not conserved at a longer scale. This implies that the mineral does not give an X‐ray diVraction pattern characterized by spots materializing the diVerent diVraction plans. Amorphous minerals display several advantages in comparison to their crystalline equivalent: their formation requires less energy; they are more easily solubilized, which means that their constitutive ions are more easily available; they incorporate ionic impurities with a higher tolerance than the crystalline form; they exhibit percolation channel that

6. Molluscan Shell Proteins

221

allows ion diVusion, that is, they can be more easily remodeled (Simkiss, 1991). Recent X‐ray diVraction data indicate that the early larval shell mineralization (prodissoconch I) may be a transient amorphous phase, a fact that earlier works on Tridacna squamosa (LaBarbera, 1974) or on Crassostrea gigas (Lee, 1990) had not detected. The presence of ACC has been demonstrated for the freshwater snail, B. glabrata (Hasse et al., 2000; Marxen et al., 2003a), for which the transformation of ACC into aragonite could be recorded at diVerent developmental stages. In the mussel, M. edulis, the first mineral formed is also ACC, but its quantity dramatically drops only 40 hours after fertilization (Medakovic, 2000), in profit of the aragonitic phase. The early developmental stages of the oyster, Ostrea edulis, are also marked by an extremely broad X‐ray diVraction ‘‘peak’’ that may be interpreted as ACC (Medakovic et al., 1997). In the clam Mercenaria mercenaria, the prodissoconch I contains ACC, which transforms after several days in aragonite. In the oyster C. gigas, the prodissoconch I contains ACC and aragonite (Weiss et al., 2002). By extrapolating these data to adult shells, and by keeping in mind the presence of amorphous granules in the molluscan mantle tissues, it is conceivable that the whole adult shell formation occurs via a transient ACC phase. As we mention in the next section, recent findings have shown that interfacial amorphous phases exist also in the ‘‘finished’’ shell, which suggests a stabilization mechanism of ACC.

III. The Topographic Models of Shell Mineralization One key issue in research on molluscan shell biomineralization is the understanding of the relationships between the organic matrix and the mineral phase at ultrastructural level. This question is central to current hypotheses on biologically controlled mineralizations, but is still extremely debated. From the late 1960s when the early topographic models of molluscan mineralization emerged to now, there has been a considerable evolution of the concepts, partly because of the evolving methodologies used for observing biominerals. In the early years, scanning electron microscopy (SEM) was the unique investigation tool. The use of surface etching treatments, partial decalcification and fixation (Mutvei, 1979), enzymatic degradation brought additional information by revealing extremely fine substructures (Cuif et al., 1983). At high magnification, transmission electron microscope (TEM) and, more recently, cryo‐TEM and atomic force microscope (AFM) brought surprising ultrastructural informations such as the presence of crystal nanodomains. Finally, in the last years, XANES (X‐ray absorption near edge spectroscopy) and NanoSIMS brought additional structural informations. SEM combined with immunogold proved to be a promising technique (Marin

222

Marin et al.

et al., 2007) for visualizing single‐protein components within biominerals. In the next section paragraph, we will discuss some recent outcomes on nacre and prisms, the two most familiar molluscan shell textures.

A. Early Nacre Descriptions and Models For several reasons such as its economical interest or its mechanical properties, nacre, also called mother‐of‐pearl, is still one of the most‐studied shell textures. Many consider it as the reference model because of its apparent geometrical simplicity and because of its universality among mollusks. Nacre is indeed a widespread molluscan shell texture, and is, with one exception, a bryozoan (Weedon and Taylor, 1995), typical of this phylum. It is represented within the three main classes of mollusks, bivalves, gastropods, and cephalopods. The ‘‘nacre’’ terminology refers to a well‐defined type of microstructure, described as follows: small flat tablets of aragonite, about half a micron thick, which are tightly packed together by organic cement. The tablets can be rectangular, hexagonal, or rounded, and look like monocrystals. They form superimposed layers of uniform thickness. Basically, there are two broad types of nacre, depending on the disposition of tablets (Erben, 1972; Nakahara, 1991): ‘‘brick wall’’ nacre, which is classical among bivalves, ‘‘columnar’’ nacre, found in gastropods (Fig. 3). In the first type, in cross section, crystals are positioned in staggered rows, just like bricks in a wall (Checa and Rodriguez‐Navarro, 2005; Oaki and Imai, 2005). Bivalve nacre tablets have their a, b, and c axes co‐oriented, with the c axis perpendicular to the nacre surface, and the b axis parallel to the local growth direction of the shell margin (Checa et al., 2006). In the second type, flat tablets are aligned on top of each other, and thus, form piles (or towers) of crystals (Lin and Meyers, 2005). Tablets of the same pile are co‐oriented (c axis along the axis of the pile), but from pile to pile, the a and b axes are not ordered. TEM studies have shown that a thin layer of organic matrix, the interlamellar matrix, delimitates the lower and upper tablet surfaces (Fig. 3). The thickness of this matrix is about 20 nm. Within a same lamella, an organic matrix [the intercrystalline matrix of Bevelander and Nakahara (1969)] separates adjacent tablets. Early amino acid analyses showed that the matrix around the tablets was enriched in Ala and Gly residues, a composition, which conferred to the matrix hydrophobic properties, similar to that of worm silk. Additional ultrastructural studies showed that nacre tablets were not homogeneous. In particular, Crenshaw and Ristedt (1975) evidenced that sulfated polysaccharides were localized in the central part of nacre tablets. These organic compounds were supposed to act as crystal nucleators. Mutvei (1979), by etching nacre

Figure 3 Structure of the two main molluscan nacre textures. (A and B) SEM pictures of the nacre of the freshwater bivalve U. pictorum. (C and D) SEM pictures of the nacre of the gastropod Haliotis tuberculata (bar scales ¼ 10 mm). (E) The ‘‘brick‐wall’’ model of bivalvian nacre. (F) The ‘‘columnar’’ model of gastropod nacre. These simplified models, adapted from Nakahara (1991), do not take in account the existence of pores in the interlamellar organic matrix, the substructures of nacre tablets, and the existence of a thin ACC layer around the tablets. The constituents are: E, the secreting mantle epithelium; S, the organic sheets; SS, the newly formed surface sheets; Cr, the aragonite crystals; T, the top of the newly formed crystals.

224

Marin et al.

tablets with a glutaraldehyde‐acetic acid solution, observed extremely complex structures such as twinning patterns or concentric growth lamellae. One of the first ‘‘modern’’ nacre models was proposed by Bevelander and Nakahara (1969). The ‘‘compartment model’’ supposed that the aragonite nacre tablets grow in a preformed mold made from the interlamellar matrix. The main drawback of this model was to ignore all the organic ingredients of the matrix, namely chitin, hydrophobic ‘‘silk‐fibroin‐like’’ proteins and above all, acidic Asp‐rich proteins. The next model, published in the early 1980s by Weiner and coworkers, integrated all the recent findings of that time: an acidic template for nucleating crystals (Weiner and Hood, 1975), acidic macromolecules for inhibiting the crystal growth (Wheeler et al., 1981), specific amino acid sequences (Asp‐rich) for chelating calcium ions (Weiner, 1979, 1983), sulfated polysaccharides for attracting calcium ions (Addadi et al., 1987). In this model, the insoluble framework was constituted by a chitin core taken in sandwich between two hydrophobic silk fibroin‐like sole, on top of which lay a
‐sheet of Asp‐rich soluble proteins. The soluble polyanionic sheet was supposed to function as a template by nucleating aragonite crystals, while covalently bound acidic polysaccharides were supposed to concentrate calcium ions at the vicinity of the template. The growth of the crystal was stopped by the addition of an inhibiting layer of acidic macromolecules on top of the newly formed tablets. The successive steps of nucleation and inhibition were explaining the regularity and repetitiveness of nacre. Because the aragonite tablets nucleated and grew on an organic template, this crystal growth model was assimilated to heteroepitaxy.

B. Recent Nacre Models and Evolving Views The heteroepitactic model for nacre achieved a frank success for more than a decade, until AFM and TEM observations of the columnar nacre of the abalone showed that holes, of diameter comprised between 5 and 50 nm, were present in the interlamellar matrix (Scha¨Ver et al., 1997; Song et al., 2003). This finding suggested that nacre platelet grows in continuity with the underlying ones, through mineral bridges, and not by heteroepitaxy. Song et al. (2003) calculated that each platelet exhibits about 1400–1900 holes. Holes had been observed previously, in particular by Mutvei (1969) and Nakahara (1991), in gastropod and bivalve nacre. At that time, it was suggested that the interlamellar holes facilitate the passage of mineral precursors for filling the empty compartments. So far, we do not know whether the presence of holes is a general feature of nacre or represents particular cases. Furthermore, it is still unclear whether they are mineral bridges or just holes of a sieve, for allowing diVusion of the organic and mineral precursors to the site of mineralization.

6. Molluscan Shell Proteins

225

Another drastic evolution of the topography of the model occurred about 6 years ago, when Levi‐Kalisman et al. (2001) observed the nacre of the bivalve Atrina with cryo‐TEM in the hydrated state. The changes go as follows:
‐chitin is the highly ordered polymer, which gives the framework that organizes the orientation of the crystals; the silk fibroin‐like proteins, which are highly insoluble in their final state, are secreted as a disordered gel; minerals grow in this gel and push it aside when they laterally extend; the gel comprises also clusters of acidic macromolecules that are involved in nucleating crystals. This model was nicely reshaped and integrated in a dynamic perspective (Addadi et al., 2006), which presents the formation of nacre in four stages: (1) assembly of the matrix (chitin, then silk gel); (2) formation of the first mineral, ACC; (3) nucleation of aragonitic tablets via polyanionic polymers; and (4) growth of the tablets, first in thickness (until reaching the top interlamella) then laterally. Although there is a general consensus on the fact that nacre tablets grow from their center and expand laterally until reaching the confluence with neighboring tablets, the ultrastructure of single‐nacre tablets remains unclear. Histochemical observations of Nautilus nacre by Nudelman et al. (2006) confirmed the old finding of Crenshaw and Ristedt (1976), that is, the concentration of reactive groups (carboxylate), presumably involved in nucleating aragonite, in the center of single tablets. In addition, a zonation was observed, which consisted of, from the tablet center to the periphery, a central ring‐shaped area rich in sulfates, an intermediate zone rich in carboxylate, and finally a tablet‐surrounding matrix rich in carboxylates and sulfates. Another recent paper (Nassif et al., 2005) showed that the nacre tablets of the abalone were coated by an extremely thin layer (3‐ to 5‐nm thick) of ACC. This layer may be a stabilized remnant of the transient ACC phase, described by Addadi et al. (2006). A possible scenario for a single‐tablet growth suggests a lateral tablet expansion and the subsequent expulsion of the gel. The process is driven by hydrophobic interactions. By doing so, the organic ‘‘impurities’’ progressively concentrate in a front at the interface between the mineral and the gel. During this growth phase, the transient ACC is replaced by aragonite. When the centrifugal front meets a similar front of the neighboring tablet, the degree of impurities becomes so high that ACC is stabilized, which prevents further crystallization of aragonite. At higher magnification, single‐nacre tablets exhibit a remarkable hierarchical architecture, which is somehow diYcult to conciliate with what has been described above. Nacre tablets have fractal properties in the sense that they exhibit diVerent levels of substructures that reproduce the same motif: these are, for example, flat nanobuilding blocks (Oaki and Imai, 2005), or ‘‘nanotablets’’ of 30‐ to 180‐nm long and less than 100‐nm thick, that self‐assemble and self‐orientate. The laminated structure of single‐nacre tablets has also been observed independently by Rousseau et al. (2005).

226

Marin et al.

AFM studies by the same authors suggest that the nanograins that constitute each platelet are encapsulated in a continuous network of an organic intracrystalline phase (Rousseau et al., 2005). This phase looks like a foam, which suggests that the early steps of tablet formation are performed in an emulsion. Reticulate circular imprints of an organic framework at triple junctions between mature platelets in bivalve nacre have also been observed (Rousseau et al., 2005). They are supposed to localize the spot where the new tablets grow. Clearly, since the beginning of the twenty‐first century, the nacre model knows a complete revolution and requires the integration of diVerent levels of observation, from micrometric to nanometric scales. The future topographic models will have to consider the fine architecture of the matrix, the sequence of the secretory events, as well as purely crystallographic and geometrical considerations, such as crystal competition.

C. Prism Models Beside the well‐studied nacre, prisms constitute another key model and an important shell texture found most frequently, but not exclusively, in molluscan outer shell layer, in particular, in gastropods, cephalopods, and bivalves. Like nacre, prisms are supposed to be an archaic type of shell texture. Because prisms are often associated to nacre, it has been proposed that nacre evolved through simple horizontal partitioning of vertical prisms (Carter and Clark, 1985; Taylor, 1973). This appealing idea, based on simple geometric considerations, needs to be reevaluated with accurate crystallographic and biochemical criteria. This may help to understand the transition from one microstructure to the other and to reconstitute primitive shell textures. It is interesting to notice that prism‐like or palisade‐like minerals, with growth axis perpendicular to the growth plan, represent an extremely common and fast strategy found by diVerent biomineralizing systems (brachiopods, mollusks, eggshell) for filling a space with minerals. In a first approximation, prismatic textures exhibit many similarities with purely chemical crystal growth. However, as we briefly show here, this view is probably oversimplified, and the deposition of prisms, similarly to nacre ones, is finely regulated over diVerent scales. Prisms are calcitic or aragonitic needles of various lengths and diameters, from the thin oblique calcitic prisms of the edible mussel, M. edulis, to the large‐sized calcitic prisms (‘‘simple’’ prism type), developed perpendicularly to the shell surface, among the fan mussel Pinna nobilis, or the aragonite prisms of the freshwater mussel, Unio pictorum. Prisms of the outer shell layer are secreted on the inner surface of the periostracum, at the growing shell edge. They grow inward by the accretion of crystal units. They are

6. Molluscan Shell Proteins

227

enveloped by an organic insoluble and hydrophobic sheath, which forms a honeycomb‐like framework. This sheath is not homogeneous, but composed of at least three layers (Dauphin, 2002; Marin et al., 2007). It maintains all the prisms together and allows a certain flexibility of the structure. When isolated from their organic envelopes, prisms comprise an intracrystalline organic fraction. In calcitic prisms, this matrix is particularly acidic (Marin et al., 2005). Like for nacre, the formation of prisms is far from being elucidated. From Grigor’ev (1965), it is well known that prism‐like crystals can be obtained by purely chemical ‘‘competition for space’’ crystal growth. The starting point includes nucleation spots, more or less uniformly spread on a surface, from which spherulites grow concentrically. When the spherulites come into contact, their growth is constrained in one direction, perpendicular to the surface. This happens in natural environments without the need of a sophisticated organic template. Competition for space can be easily simulated (Ubukata, 1994, 1997). By certain aspects, the prism growth in mollusks looks like crystal competition. In a similar way, the early step of molluscan prism construction is the formation of spherulites in the inner surface of the periostracum. This has been clearly shown for the bivalves, Pinna nobilis (Cuif et al., 1983) and Lamprotula sp. (Checa and Rodriguez‐Navarro, 2001). However, a competition for space phenomenon might occur only in the early steps of prism formation (disappearance of minute prisms just below the periostracum; see Checa et al., 2005), but may not describe accurately the subsequent steps of prism growth, mainly for two reasons: the sheaths are formed before the prism mineral infilling and the growth is constrained by the organic sheaths around the prisms. Another aspect that renders a simple ‘‘competition for space’’ model inappropriate is the multiscale structure of prisms and their striking complexity. A well‐known example is that of the ‘‘simple type’’ calcitic prisms of Pinna nobilis. Optically, each prism of P. nobilis behaves like a monocrystal, with a single extinction when observed with polarized‐analyzed light (Cuif et al., 1983). However, enzymatic treatment of the prism preparation shows that each prism is constituted of a pile of flat crystal units, which can be entirely dissociated after pyrolysis. These crystallites are the growth units. They are separated from each other by an organic intracrystalline template. Within a pile, they are perfectly positioned according to their a, b, and c axes. In spite of looking homogeneous, these crystallites are composed of subdomains, emphasized by enzymatic treatments (Cuif et al., 1983) or immunostaining (Marin, unpublished data). These subdomains might as well be composed of nanocrystal aggregates. Similarly to P. nobilis, ultrastructural observation of the prisms of Cristaris plicate (Tong et al., 2002) revealed a complex lacelike framework of intracrystalline matrix. Checa et al. (2005) hypothesized that the interprismatic ‘‘honeycomb‐like’’ sheaths are formed

228

Marin et al.

by interfacial tensions that occur in a precursor liquid–liquid emulsion. This elegant hypothesis needs further testing. Is there a unique prism model? Nothing is less certain. We describe similar objects by using a single terminology. However, many ultrastructural studies show that significant diVerences occur, even in closely related taxa, as shown in Fig. 4. The best example is that of Pinna nobilis and Pinctada margaritifera (Cuif et al., 1991; Dauphin, 2003). In cross sections (perpendicular to the growth axis), the prisms of Pinna nobilis look homogeneous and behave like monocrystals. On the contrary, those of Pinctada margaritifera exhibit sinuous intraprismatic membranes (particularly well visible by SEM, after etching) that separate the section in domains, and these domains do not have the same crystallographic orientation. Another case is the prismatic outer layer of Unio, which is absolutely diVerent from the two types cited above: the prisms of Unio are composed of single‐crystal fibers radiating from spherulites (Checa and Rodriguez‐Navarro, 2001; Cuif et al., 1983). Clearly, important eVorts need to be put in the elucidation of the prism growth and to relate it to the biochemical properties of the associated matrix.

Figure 4 Prismatic microstructures among bivalve shells. (A) Unio pictorum. (B) Anodonta sp. For A and B, the aragonitic prisms (above) are in contact with the nacreous layer. (C) Oblique thin calcitic prisms of the edible mussel Mytilus edulis. (D) Calcitic prisms of Atrina rigida.

6. Molluscan Shell Proteins

229

IV. Molluscan Shell Proteins: Characterization of Their Primary Structure In parallel to the structural studies on the diVerent shell textures, a considerable eVort was realized, in the last three decades, for identifying the diVerent macromolecules that constitute the shell matrix and for obtaining information on their primary structures. In most of the cases, these macromolecules were analyzed after the dissolution of the mineral phase. Until now, the most commonly used reagents are EDTA, a calcium‐chelating agent, which is eVective at neutral pH, weak dilute acids, acetic or formic acids, or rarely dilute hydrochloric acid. Our preference goes to cold dilute acetic acid, progressively added to the cleaned shell powder suspended in Milli‐Q water, until reaching pH 4, the decalcification process being performed at 4
C (Marin, 2003). We assume that this procedure minimizes protein degradation and precludes the formation of macromolecular aggregates, as EDTA does. Other extraction processes include a soft—but long—demineralization of the powder on a cation‐exchange resin (Albeck et al., 1996), or extraction with water (Pereira‐Mourie`s et al., 2002). However, in this latter case, the most strongly mineral‐linked macromolecules are not extracted. The decalcification procedure yields two organic fractions, one soluble in the decalcifying solution, the other one strongly insoluble. The ratio between the two fractions can considerably vary: while the soluble fraction represents between 0.03 and 0.5 wt %, the insoluble fraction varies in greater proportions: from 0.01% (in some crossed‐lamellar neogastropods for instance) to 4–5 wt % of the shell of the abalone or of the nautilus! Usually, the second fraction is discarded by centrifugation. In some cases, the insoluble fraction may be partially dissolved by using strong denaturing agent (urea) and/or by heating. The soluble extract can be cleared from decalcification salts by ultrafiltration or dialysis. It can be further fractionated according to standard biochemical techniques, electrophoresis, or chromatography (gel permeation, ion exchange, aYnity). However, because molluscan shell proteins have a nonstandard behavior due to polydispersity, multiple anionic charges, posttranslational modifications, classical fractionations usually fail in resolving the soluble matrix in discrete macromolecules. This technical obstacle, found also with several other calcified tissues, pestered for more than two decades the life of researchers involved in biomineralization studies! This explains in particular why the first partial amino acid sequence from a mollusk shell was obtained in the early 1990s (Rusenko et al., 1991), and the first full‐length sequence only in 1996 (Miyamoto et al., 1996). The search for the primary structure of molluscan shell proteins benefited from the major technical advances in molecular biology, in particular, from the possibility to use degenerate oligonucleotide probes deduced from short

230

Marin et al.

partial N‐terminal or internal amino acid sequences. This allowed fishing out the corresponding transcript by RT‐PCR or by cDNA library oligoscreening. Another strategy successfully developed by us was to use polyclonal antibodies raised against shell matrix for screening expression cDNA libraries (Marin et al., 2003a). On the 43 fully known sequences, 39 were obtained via molecular biology and only 4 by direct protein sequencing. It is predictable that the number of shell proteins will ‘‘explode’’ in the near future, owing to the increasing number of fully sequenced genomes (Livingston et al., 2006), and in the absence of genomic data, owing to EST technique applied on shell ‘‘secretome’’ (Jackson et al., 2006). In this chapter, we give a brief description of the diVerent shell proteins, one after the other. Complementary information can be retrieved in Marin and Luquet (2004), in Matsushiro and Miyashita (2004), and in the review of Zhang and Zhang (2006). What was possible few years ago, when the number of identified and named proteins was still reasonable, would be fastidious and redundant. We deliberately choose to present the known molluscan shell proteins in three groups, according to their theoretical isoelectric point (Fig. 5). The first group comprises proteins, the pI of which is below 4.5, 160 18

140

MW

80

3 9 8

14

1

40

16 2 4

20 0

20 ⫻ 21

10 13

60

5 6

0

2

4

7

12 15 17 11 6 pI

8

19

22 10

12

Figure 5 Graphical representation of the distribution of the molecular weights (MW) of all known molluscan shell proteins versus their isoelectric point (pI ). The theoretical MW and pI were computed (http://www.expasy.ch/tools/pi_tool.html) after identification and removal of the signal peptide (http://www.cbs.dtu.dk/services/SignalP/). □ ¼ proteins associated with aragonite; ◆ ¼ proteins associated with calcite; ( ¼ protein associated with both aragonite and calcite (1 ¼ aspein; 2 ¼ Asp‐rich proteins; 3 ¼ MSP‐1; 4 ¼ MSP‐2; 5 ¼ MSI31; 6 ¼ prismalin‐14; 7 ¼ N‐14/N16/pearlin/perline proteins masking AP7 and AP24; 8 ¼ MSI60; 9 ¼ mucoperlin; 10 ¼ nacrein from P. fucata; 11 ¼ MSI7; 12 ¼ dermatopontin; 13 ¼ tyrosinase‐like1; 14 ¼ nacrein from T. marmoratus; 15 ¼ perlucin; 16 ¼ shematrin proteins; 17 ¼ perlustrin, 18 ¼ lustrin A; 19 ¼ perlwapin; 20 ¼ N‐66; 21 ¼ tyrosine‐like2; 22 ¼ KRMPs).

6. Molluscan Shell Proteins

231

the second group comprises proteins with a pI in the range 4.5–7, and the third group comprises proteins with a pI above 7. We are fully aware that this grouping is artificial and arbitrary but practical in the absence of clearly identified protein families.

A. Extremely Acidic Shell Proteins This first group, the most homogeneous one, comprises the most acidic shell proteins (Marin and Luquet, 2007). The existence of such proteins is known since the pioneering work of Weiner and Hood (1975), followed by extremely precise chromatographic characterization (Weiner, 1979, 1983). However, for the reasons described above, they were the most diYcult to purify. In particular, they do not stain correctly on SDS‐PAGE (Marin et al., 2001). Because of their negative charge, they are even suspected to diVuse readily out of the electrophoresis gel, and they subsequently require a double fixation (Gotliv et al., 2003). The two first full‐length sequences of very acidic proteins deduced from their transcript were published in 1997 (MSI31; Sudo et al., 1997) and in 2001 (MSP‐1; Sarashina and Endo, 2001). Today, this group comprises only six proteins (Table I). One striking feature of these proteins is their association with calcitic biominerals rather than aragonite. MSI31, aspein, prismalin 14, and Asp‐rich were retrieved from calcitic prism textures, and MSP‐1 and MSP‐2 (SP‐S) from foliated calcite. The finding that acidic proteins are preferentially associated with calcite in mollusk shell is not new: Hare (1963) already noticed that ‘‘the organic matrices from the calcite layers have a consistently higher ratio of acidic to basic amino acids than the aragonitic shell units.’’ The reason of this selection is intriguing, but remains unknown. Another feature associated with very acidic molluscan shell proteins is that they are all singularly enriched in Asp residues. The ‘‘choice’’ for this amino acid, rather than Glu, is remarkable, although poorly understood. It may meet stereochemical requirements, the Asp side chain being shorter than that of Glu. Because of their high amount of Asp residues and the side chains are negatively charged under physiological conditions, these proteins are supposed to easily bind calcium ions. They consequently belong to the group of low‐aYnity, high‐capacity calcium‐binding proteins (Maurer et al., 1996), which implies that they do not exhibit the typical ‘‘high‐aYnity, low‐capacity’’ canonical calcium‐binding domains, such as EF‐hand (Kretsinger, 1976). Their moderate aYnity for calcium is compatible with a reversible binding of calcium ions. MSI31 is a Gly‐rich protein of the insoluble matrix, and supposed to be primarily a ‘‘framework’’ protein (Sudo et al., 1997), because of the 10 short poly‐Gly blocks, distributed mainly in the N‐terminal domain. It exhibits

Table I

Unusually Acidic Molluscan Shell Proteins (with a Theoretical Isoelectric Point Below 4.5)a

Protein Name BIVALVIA

Aspein MSI31 Prismalin‐14 MSP‐1 MSP‐2/SP‐S Asp‐rich protein 1 Asp‐rich protein 2 Asp‐rich protein 3 Asp‐rich protein 4 Asp‐rich protein 5 Asp‐rich protein 6 Asp‐rich protein 7 Asp‐rich protein 8 Asp‐rich protein 9 Asp‐rich protein 10

Species Pinctada fucata Pinctada fucata Pinctada fucata Patinopecten yessoensis Patinopecten yessoensis Atrina rigida Atrina rigida Atrina rigida Atrina rigida Atrina rigida Atrina rigida Atrina rigida Atrina rigida Atrina rigida Atrina rigida

Microstructure (polymorph)

MW (kDa)

pI (% Asp þ Glu)

Swiss‐Prot/ TrEMBL Accession Number

Prisms (calcite)

39.3/41.2

1.67 (61.9)

Q76K52

Prisms (calcite)

32.85/31

3.8 (14)

O02401

Tsukamoto et al., 2004 Sudo et al., 1997

Prisms (calcite)

11.9/13.5

4.24 (10.5)

Q6F4C6

Suzuki et al., 2004

Foliated (calcite) Foliated (calcite) Prisms (calcite) Prisms (calcite) Prisms (calcite) Prisms (calcite) Prisms (calcite) Prisms (calcite) Prisms (calcite) Prisms (calcite) Prisms (calcite) Prisms (calcite)

74.6/76.4

3.34 (22.8)

Q95YF6

27.9/29.8

3.48 (22.3)

Q6BC34

6.6/8.5 15/17 16.5/18.4 18/19.9 17.4/19.3 18.2/20 25.8/23.9 25.3/27.2 18.2/20 20/21.8

3.34 (50.8) 2.89 (52.8) 2.75 (60) 2.73 (56.2) 2.76 (57.5) 2.72 (59.2) 2.54 (66.2) 2.53 (65.1) 2.72 (59.2) 2.68 (60)

Q5Y821 Q5Y822 Q5Y823 Q5Y824 Q5Y825 Q5Y826 Q5Y827 Q5Y828 Q5Y829 Q5Y830

Sarashina and Endo, 1998, 2001 Hasegawa and Uchiyama, 2005 Gotliv et al., 2005 Gotliv et al., 2005 Gotliv et al., 2005 Gotliv et al., 2005 Gotliv et al., 2005 Gotliv et al., 2005 Gotliv et al., 2005 Gotliv et al., 2005 Gotliv et al., 2005 Gotliv et al., 2005

References

a The sequences of all these proteins were deduced from their respective transcript sequence. Prismalin‐14 was also biochemically characterized. In the MW column, the first number corresponds to the molecular weight of the protein without its signal peptide, and the second one to the unprocessed protein.

6. Molluscan Shell Proteins

233

acidic C‐terminal motifs (6 XSEEDY, where X is D or E, and Y is M or T). The acidic domain may be involved in nucleating crystals, or binding calcium, but this has not been tested. The hydrophobic N‐terminus may adopt a
‐sheet conformation. MSP‐1, the second unusually acidic protein of the Japanese scallop Patinopecten yessoensis, is enriched in Ser, Asp, and Gly residues and exhibits a modular structure, with a short‐basic domain, close to the N‐terminus and two GS domains that alternate with D‐rich domains (Sarashina and Endo, 1998, 2001). The two Asp‐rich domains fit with the initial model of Weiner and coworkers since they exhibit DGS and DS motifs. They also present numerous DD repeats. All these motifs are suspected to bind calcium ions or to interact with calcite crystals. In addition, several serine residues are putatively phosphorylated or glycosylated. MSP‐1 exhibits homologies with dentin phosphophoryns, very acidic proteins of the teeth. Recently was found MSP‐2, also called SP‐S, another shell protein of the Japanese scallop (Hasegawa and Uchiyama, 2005). MSP‐2 is a Ser‐Gly‐Asp‐rich protein of the scallop, which exhibits 91% identity with MSP‐1. It may represent a shortened variant of MSP‐1. The third unusually acidic protein is aspein, a protein of the pearl oyster Pinctada fucata. The composition of aspein is remarkable since Asp residues represent 60.4% of the whole protein, and its theoretical pI is 1.67, which would make it the most acidic protein found to date! The two other abundant amino acids are Gly (16%) and Ser (13%). The main body of aspein is composed of 58 poly‐Asp blocks (of 2–10 Asp residues long) interspersed by SG dipeptides. Some Ser residues may be phosphorylated. Aspein exhibits some similarities with aspolin, phosphophoryn, and bone sialoprotein‐binding protein. The primary structure of aspein suggests that it is a high‐capacity, low‐aYnity calcium‐binding protein. Asp‐rich is a family of 10 related proteins, composed of the following domains, from N‐ to C‐terminus: hydrophobic (signal peptide), short basic, acidic 1, variable acidic, DEAD repeats, and acidic 2. Interestingly, the acidic 1 domain has a high homology with calsequestrins, calcium‐binding proteins from cardiac and skeletal muscles, and may consequently bind calcium. The variable acidic domain exhibits long stretches of poly‐Asp. The DEAD motif is also found in helicases, enzymes that separate the two DNA strands. Asp‐rich and aspein are closely related proteins since they share 48% homology. The last member is prismalin‐14, the single one characterized both at protein and at transcript levels (Suzuki et al., 2004). Prismalin was extracted from the insoluble hydrophobic framework of the prismatic layer of Pinctada fucata. It is a G/Y‐rich protein, representing, respectively, 27.6% and 20% of the amino acid composition. It exhibits PIYR repeats, a G/Y‐rich region and N‐ and C‐terminal D‐rich calcium‐binding regions. It inhibits the precipitation of calcium carbonate in vitro and induces morphological changes of

234

Marin et al.

calcite crystals. Prismalin 14 as well as aspein and MSI31 are specifically expressed in the mantle, in the region, which secretes the prism matrix (Takeuchi and Endo, 2005). In addition, the expression levels of these three proteins are correlated, which suggests that they are secreted at the same time.

B. Moderately Acidic Shell Proteins As shown in Table II, the second group corresponds to acidic proteins, with pIs between 4.5 and 7. In this disparate group, one finds gastropod and bivalve nacreins, MSI60, MSI7, the N14/pearlin/N16 family, mucoperlin, AP7, AP24, a tyrosinase‐like protein, perline, and a snail dermatopontin. First of this list is nacrein, by many aspects the most‐studied molluscan shell protein family (Matsushiro and Miyashita, 2004; Miyamoto et al., 1996, 2003, 2005; Miyashita et al., 2002; Takeuchi and Endo, 2005). Nacrein, the first protein whose primary structure was deciphered, was initially found as a 50‐kDa EDTA‐soluble protein of the nacreous layer of the Japanese pearl oyster Pinctada fucata. Later on, a similar protein (nacrein), with an identical N‐terminus, was found in association with the prismatic layer (Miyashita et al., 2002). In situ hybridization studies (Miyamoto et al., 2005; Takeuchi and Endo, 2005) shows that nacrein is ubiquitous and displays probably the same functions in the two layers. Nacrein is also the first protein, which was proved to work as an enzyme. Nacrein exhibits several GXN repeats (where X is frequently D, N, or E), flanked by two carbonic anhydrase‐like subdomains. Strikingly, these two subdomains have a relatively high homology with human carbonic anhydrase II (CA). In addition, the first nacrein CA‐ like domain exhibits the three histidine residues involved in zinc binding, typical of CA. A full‐length recombinant nacrein inhibits the in vitro precipitation of calcium carbonate. Interestingly, the recombinant nacrein, which lacks the central GXN repeat domain, does not exhibit this property, whereas the repeat domain, tested alone, has a strong inhibiting ability (Miyamoto et al., 2005). Another nacrein was found in the shell of the gastropod Turbo marmoratus (Miyamoto et al., 2003). This moderately acidic protein has a longer repeat domain, constituted of GN motifs. At last, N66, retrieved from the Australian pearl oyster Pinctada maxima, belongs also to the family (Kono et al., 2000). It exhibits two carbonic anhydrase subdomains, and a longer central repeat domain constituted by 46 GXN motifs interspersed by 12 GN motifs. The repeat domain of N66 is much less acidic than the one of nacrein, which implies that N66 has a basic pI. Among the acidic shell proteins, MSI60 is an insoluble framework protein retrieved initially from the nacreous layer. It exhibits 11 poly‐Ala blocks and 39 poly‐Gly blocks dispersed throughout the sequence. The poly‐Ala blocks confer to MSI60 some homologies with spider silk fibroins. The MSI60

Table II

Moderately Acidic Molluscan Shell Proteins (with a Theoretical Isoelectric Point 4.5  pI  7)a

Species

Microstructure (Polymorph)

N14 Nacrein N16/Pearlin

Pinctada maxima Pinctada fucata Pinctada fucata

Nacre (aragonite) Nacre (aragonite) Nacre (aragonite)

13.7/16.4 48.2/50.1 12.8/15.4

4.8 (15.8) 6.85 5.14 (16.9)

Q9NL39 Q27908 O97048

MSI60 MSI7 Tyrosinase‐like protein 1 Perline

Pinctada fucata Pinctada fucata Pinctada fucata

Nacre (aragonite) Prisms (calcite) Prisms (calcite)

61.7/60 7.3/9.3 56.3/58.3

4.8 (5.9) 5.98 (2.6) 6.5 (9.3)

O02402 Q7YWA5 A1IHF0

Nacre (aragonite)

13.6/16.2

4.7 (15.8)

Q14WA6

Mucoperlin AP7

Pinctada margaritifera Pinna nobilis Haliotis rufescens

Nacre (aragonite) Nacre (aragonite)

65.4/66.7 7.6/9.9

Q9BKM3 Q9BP37

AP24

Haliotis rufescens

Nacre (aragonite)

17/19.6

Nacrein

Turbo marmoratus Biomphalaria glabrata

Nacre (aragonite)

56/57.6

Crossed‐lamellar (aragonite)

16.6 (no s.p.)

4.87 (9.5) 5.43 (12.1) 5.53 (13.6) 5.76 (10.7) 6.33 (10.8)

Protein Name BIVALVIA

GASTROP

Dermatopontin

a

MW (kDa)

pI (% Asp þ Glu)

Swiss‐Prot/ TrEMBL Accession Number

References Kono et al., 2000 Miyamoto et al., 1996 Samata et al., 1999; Miyashita et al., 2000 Sudo et al., 1997 Zhang et al., 2003a Nagai et al., 2007

Q8N0R6

Montagnani et al., 2006 Marin et al., 2000 Michenfelder et al., 2003 Michenfelder et al., 2003 Miyamoto et al., 2003

P83553

Marxen et al., 2003b

Q9BP38

The sequence of dermatopontin was obtained by direct protein sequencing. All the other proteins were retrieved from their transcript sequences. In the MW column, the first number corresponds to the molecular weight of the protein without its signal peptide, and the second one, to the unprocessed protein. s.p. ¼ signal peptide.

236

Marin et al.

N‐terminus contains two Asp‐rich domains and four Cys residues, and the C‐terminus contains one short Asp‐rich domain and one Cys residue. In situ hybridization shows that MSI60 is specifically associated with the secretion of nacre (Takeuchi and Endo, 2005). In a paper (Asakura et al., 2006), the first Asp‐rich domain (16 residues among which 11 are acidic) of MSI60 was introduced between diVerent Ala/Gly‐rich domains derived from silk fibroins, and the conformation of the resulting peptides, studied by NMR spectroscopy. It was shown that the calcium‐binding ability of the acidic domain was the most eVective when the flanking domains had a
‐sheet conformation. Another protein family is represented by a series of low molecular and moderately acidic proteins of the nacre of the pearl oyster. This protein family is studied by two independent Japanese groups, a reason that explains why this family is called either N14/N16 (Kim et al., 2004; Kono et al., 2000; Samata et al., 1999, 2003) or pearlin (Miyashita et al., 2000, 2003; Matsushiro and Miyashita, 2004; Matsushiro et al., 2003a,b) All the members of this family diVer by few amino acids. They are enriched in Gly, Tyr, and Asn residues and exhibit Gly‐Asn repeat sequences in addition to four short acidic domains (3–12 residues). They exhibit a putative phosphorylation site, in addition to a heparin‐binding domain. Pearlin, as defined by Miyashita, is a calcium‐binding protein, but this ability is conveyed by a covalently bound sulfated polysaccharide (Miyashita et al., 2003). Samata et al. (1999) observed that N16 in solution inhibits the in vitro crystal growth but induces the formation of aragonite when fixed on the water‐ insoluble matrix. By using a diVerent experimental device, Matsushiro et al. (2003a,b) observed that pearlin was able to make protein complex with pearl keratin. Only in the presence of CaCO3 saturated solution containing Mg2þ, the complex induced the formation of aragonite. The dissociated complex lost this ability, while the reconstituted complex recovered this property, suggesting that the polymorph selection is determined at supramolecular level. The N16/pearlin proteins are specific of nacre matrix, as it was shown by diVerent techniques, Northern blot (Kono et al., 2000; Samata et al., 1999) and in situ hybridization (Takeuchi and Endo, 2005). In addition, quantitative RT‐PCR showed that the levels of expression of N16 and nacrein are correlated (Takeuchi and Endo, 2005). Recently, perline was obtained from the Polynesian pearl oyster Pinctada margaritifera (Montagnani et al., submitted for publication). Perline has 93% homology with N14. Mucoperlin, a protein retrieved by immunoscreening of the bivalve Pinna nobilis cDNA library (Marin et al., 2000, 2003a), belongs to a completely diVerent protein family. This protein is composed of three domains: a short N‐terminus, a central region made of 13 tandem repeats of 31 amino acid residues each, and a C‐terminal part enriched in serine. The presence of central tandem repeats, the presence of Pro and Ser residues in the repeat

6. Molluscan Shell Proteins

237

domain, the numerous putative O‐glycosylation sites, and the demonstration that mucoperlin is glycosylated are criteria that aYliate mucoperlin to the mucin family. Mucins are ubiquitous proteins associated with epithelial tissues. They exhibit several functions in connection with their ability to form gels: they are involved in epithelial lubrication, act as eYcient barriers against chemical aggressions, but they also play a role in cell signaling. Mucoperlin was the first protein that was directly localized in the shell, owing to a polyclonal antibody raised against a recombinant mucoperlin. Mucoperlin is only present in the nacreous layer. Classical immunohistological staining (Marin et al., 2000) as well as immunogold technique (Marin, unpublished data) showed that the protein is concentrated around the nacreous tablets, in particular on the lateral sides rather than on the top/bottom layers. Mucoperlin may be one of the constituents of the gel‐like matrix described by Addadi et al. (2006), which is pushed aside when nacre tablet grows laterally. Two other proteins, named AP7 and AP24, were isolated from the EDTA‐ soluble matrix of the nacreous layer of the abalone Haliotis rufescens (Michenfelder et al., 2003, 2004). They are soluble and moderately acidic (pI around 5.4–5.5). They aVect the growth of calcite crystals in vitro. The calcium carbonate mineral interaction domain of AP7 and AP24 is localized in the first 30 amino acid residues of their N‐termini, as deduced from conformation studies by NMR spectroscopy and CD spectrometry (Kim et al., 2004; Wustman et al., 2004). AP7 and AP24 are considered to act as crystal‐modulating proteins. Another protein, MSI7, is a moderately acidic small protein (pI 5.98) of the Japanese oyster Pinctada fucata (Zhang et al., 2003a). Its N‐terminus is highly homologous to the N‐terminal domain of the much more acidic MSI31. In particular, MSI7 harbors the Gly‐rich sequence, which may be involved in calcium binding. The expression of the MSI7 transcript suggests that this protein is involved in the formation of the nacreous and prismatic layers. MSI7 aVects the in vitro growth of calcite crystals. In the same organism was recently retrieved a tyrosinase called Pfty1, from the outer prismatic layer (Nagai et al., 2007). Tyrosinase, a particular phenoloxydase, is a copper‐containing enzyme that binds oxygen. It is involved in the oxidation of phenol groups of tyrosine residues, which results in the formation of melanin. Pfty1 exhibits a conserved copper‐binding site, which suggests that its oxidative function is active. Pfty1 seems to be involved in the pigmentation of the prismatic layer and may be included in the prismatic layer. It is possible that Pfty1 plays a role in the defense of the pearl oyster against parasites. Another protein was directly purified and sequenced from the HCl‐soluble shell matrix of the freshwater snail B. glabrata, a model that has also been studied for developmental purposes (see above, Section II). Interestingly, the

238

Marin et al.

protein sequence represents the first one found in association with the crossed‐ lamellar structure of a modern gastropod (Marxen and Becker, 1997; Marxen et al., 2003b). Called dermatopontin, this protein has a striking homology with vertebrates and invertebrate dermatopontins, a family of extracellular matrix proteins involved in binding decorins and TGF‐
. It was suggested that the snail dermatopontin has a role in organizing spatially the shell matrix. Whether this protein has also a role in cell signaling is unknown. Another interesting feature about dermatopontin is the presence of a unique N‐linked saccharide in the first third of the sequence. The structure of this pentasaccharide has been determined, but its exact function is unknown.

C. Basic Shell Proteins Contrarily to the most acidic proteins, proteins with a basic pI were rather unexpected as components of the molluscan shell matrix (Table III). However, their existence could have been predicted by comparison with the sea urchin spicule model, for which several basic proteins were discovered one decade ago (Killian and Wilt, 1996; Wilt et al., 2003). The first found basic protein is also the most complex in it its primary structure and the most ‘‘popular,’’ that is, the most often cited as a ‘‘model’’ protein for biomineralization. This is lustrin A (Shen et al., 1997), an insoluble protein retrieved from the nacreous layer of the abalone H. rufescens. From all the molluscan shell proteins, lustrin A is the single protein, which established a clear link between a primary structure and the overall mechanical property of nacre. Lustrin A has been indeed the subject of diVerent structure–function studies (B.L. Smith et al., 1999; Wustman et al., 2002, 2003a,b; Zhang et al., 2002). The sequence of lustrin A comprises, from its N‐ to its C‐terminus, nine Cys‐rich modules (79–89 amino acid residues) interspersed by eight Pro‐rich modules (19–30 amino acid residues), followed by a long GS domain, a short D‐rich domain, a Cys‐rich modules, a short basic domain, and a protease inhibitor‐like C‐terminus. Interestingly, the first Pro‐rich domain exhibits homology (53%) with a collagen I‐ chain fragment. AFM pulling studies have shown that the interlamellar matrix, which is supposed to contain lustrin A, exhibits a typical sawtooth force‐extension curve with hysteretic recovery (B.L. Smith et al., 1999; Zhang et al., 2002). This mechanical behavior is explained by the successive stretching of springs (the Cys‐rich modules), separated by spacers (the Pro‐rich modules), when an increasing stretching force is applied to the molecule. In addition to the elastomeric properties of lustrin A, the GS domain (GS loop) provides further a highly flexible domain, while the basic C‐terminal domain (RKSY) may interact with other macromolecules, and the short acidic one (D4) may be a mineral‐binding region (Wustman et al., 2003a). Clearly, lustrin A is a multifunctional protein.

Table III Basic Molluscan Shell Proteins (with a Theoretical Isoelectric Point > 7)a

Protein Name BIVALVIA

GASTROP

a

Species

Microstructure (Polymorph)

MW (kDa)

pI

Swiss‐Prot/ TrEMBL Accession Number

References

59.8/62.4

8.66

Q9NL38

Kono et al., 2000

Pinctada fucata Pinctada fucata Pinctada fucata Pinctada fucata Pinctada fucata Pinctada fucata Pinctada fucata Pinctada fucata Pinctada fucata Pinctada fucata Pinctada fucata

Nacre (aragonite) þ Prisms (calcite) Prisms (calcite) Prisms (calcite) Prisms (calcite) Prisms (calcite) Prisms (calcite) Prisms (calcite) Prisms (calcite) Prisms (calcite) Prisms (calcite) Prisms (calcite) Prisms (calcite)

30.3/31.9 33.4/35.3 29.7/31.4 28.2/30.2 28.0/30.2 24.8/26.5 26.8/28.4 9.5/11.5 9.8/11.8 9.8/11.8 54.4/56.5

9.04 9.37 9.41 9.16 7.69 9.65 10.3 9.6 9.4 9.4 9.

Q1MW96 Q1MW95 Q1MW94 Q1MW93 Q1MW92 Q1MW91 Q1MW90 Q1AGW0 Q1AGV9 Q1AGV8 A1IHF1

Yano et al., 2006 Yano et al., 2006 Yano et al., 2006 Yano et al., 2006 Yano et al., 2006 Yano et al., 2006 Yano et al., 2006 Zhang et al., 2006c Zhang et al., 2006c Zhang et al., 2006c Nagai et al., 2007

Haliotis rufescens Haliotis laevigata Haliotis laevigata Haliotis laevigata Haliotis laevigata

Nacre (aragonite) Nacre (aragonite) Nacre (aragonite) Nacre (aragonite) Nacre (aragonite)

140/142.2 9.3 (no s. p.)b 18.2 (no s. p.) 14.5 (no s. p.) 4.79 (no s. p.)

8.13 8.02 7.15 8.62 8.26

O44341 P82595 P82596 P84811 P85035

Shen et al., 1997 Weiss et al., 2001 Mann et al., 2000 Treccani et al., 2006 Mann et al., 2007

N66

Pinctada maxima

Shematrin‐1 Shematrin‐2 Shematrin‐3 Shematrin‐4 Shematrin‐5 Shematrin‐6 Shematrin‐7 KRMP 1 KRMP 2 KRMP 3 Tyrosinase‐ like prot. 2 Lustrin A Perlustrin Perlucin Perlwapin Perlinhibin

The sequences of perlustrin, perlucin, perlwapin, and perlinhibin were obtained by direct protein sequencing. The other proteins were retrieved from their transcript sequences. In the MW column, the first number corresponds to the molecular weight of the protein without its signal peptide, and the second one to the unprocessed protein. b s.p. ¼ signal peptide.

240

Marin et al.

Two other basic proteins were characterized, one from the Australian pearl oyster, Pinctada margaritifera, the other one from the Japanese one, Pinctada fucata. The first one is N66 (Kono et al., 2000), which belongs to the nacrein family. N66 is described in Section IV.D. The second one is Pfty2, the second tyrosinase‐like protein (Nagai et al., 2007). Pfty2 exhibits high homology with Pfty1 (54% in 493 residues overlap). The homology is maximal in the central region. Similarly to Pfty1, Pfty2 contains a conserved copper‐ binding site. Its presumed function is the oxidation of Tyr residues, for pigmenting the shell (melanogenesis). Four proteins were purified from the nacreous layer of the abalone, Haliotis laevigata, and directly sequenced. These proteins are perlustrin (Weiss et al., 2000, 2001), perlucin (Mann et al., 2000; Weiss et al., 2000), perlwapin (Treccani et al., 2006), and perlinhibin (Mann et al., 2007). Perlustrin is a small protein, the sequence of which exhibits similarities to vertebrate insulin‐ like growth factor‐binding protein (IGF‐BP) sequences (40% homology), in particular a pattern of 12 Cys residues, spread along the sequence. The IGF‐ BPs represent a family of proteins, which bind growth factors of the insulin type. In vitro tests showed that perlustrin binds IGFs with a good aYnity and insulin with a low aYnity. Perlucin is an N‐glycosylated protein (via one asparagine residue), the sequence of which exhibits similarities with calcium‐ dependent lectins (C‐type). Perlucin has some sequence similarities with asialoglycoprotein receptors. Functional tests showed that the carbohydrate recognition domain (CRD) of perlucin has a broad specificity, but binds particularly mannose and galactose. In vitro studies showed that perlucin promotes the nucleation of CaCO3 crystals. Further AFM investigations (Blank et al., 2003) confirmed that perlucin is able to accelerate the nucleation of CaCO3 layers on top of calcite surfaces. This protein is incorporated as an intracrystalline component of the neosynthesized crystals. The third abalone shell protein, perlwapin, contains 3 repeats of 40 amino acid residues, which are very similar to the whey acidic proteins (WAP), a family of proteins characterized by a conserved pattern of 8 characteristically spaced Cys residues. These residues are involved in disulfide bond formation. Perlwapin exhibits also a high homology with the C‐terminus of lustrin A. Perlwapin has a polymorphism in its sequence (three variants). AFM studies on calcite surfaces indicated that it is a potent inhibitor of calcium carbonate precipitation by binding selectively to distinct step edges, thus preventing the crystal layer from growing further. It is suggested that perlwapin inhibits the growth of certain crystallographic planes. At last, the fourth protein perlinhibin inhibits the growth of calcite and induces the formation of aragonite (Mann et al., 2007). The functional relationships between perlucin, perlustrin, and perlwapin, perlinhibin and the insoluble matrix are unclear: it was recently shown that this latter is able to induce alone growth of flat oriented tablets that mimic nacre (Heinemann et al., 2006). Another protein, perlbikunin, has been characterized from the nacre layer of the abalone, but no sequence data are available yet.

6. Molluscan Shell Proteins

241

Recently, a set of diVerent proteins was identified in a cDNA library constructed from mantle tissues of the Japanese pearl oyster Pinctada fucata. With one exception, all these newly found proteins have a pI above 9, constituting thus the most basic proteins found to date, in association with the molluscan shell. They are distributed in two families: K‐rich matrix proteins (KRMPs) and shematrin. KRMPs represents a group of three small proteins with a molecular weight of 10 kDa (Zhang et al., 2006c). These three proteins, of 98–101 residue long, diVer only by few amino acids. They are Lys‐Gly‐Tyr rich. Apart the signal peptide, their primary structure is composed of a Lys‐ rich domain (40 amino acid residues long), which comprises also all the Cys and Trp residues, and a C‐terminus (39 amino acid long) enriched in Gly and Tyr, which comprises also a short acidic motif. The Gly‐Tyr‐rich region exhibits some homologies with few quinone‐tanned proteins, which suggests that the Tyr residues may be oxidized in DOPA in the mature protein. The Lys‐rich domain may interact with negatively charged ions (bicarbonate) or acidic matrix proteins. The protein is expressed only in the mantle edge, corresponding to the secretion of the prisms. The structure of KRMPs suggests that their function is to link the acidic soluble proteins to the hydrophobic framework of the prisms. The second family of basic proteins, shematrin, comprises seven members, of molecular weights between 25 and 33 kDa (Yano et al., 2006). With one exception (shematrin‐5, pI ¼ 7.7), all these proteins have a pI between 9 and 10.3. They all exhibit Gly‐rich domains, constituted of short motifs of the type XGnX (with 2  n  6 and X ¼ L/Y/A/ V/I/M). The C‐terminus of all shematrins ends with an RKKKY, RRKKY, or RRRKY motif. Surprisingly, the Gly‐rich domain of shematrin‐2 is almost identical to that of the acidic MSI31 (98% homology in a 227 residue overlap), but their respective C‐terminus diVers completely. On the other hand, the beginning of the C‐terminal half of shematrins exhibits a high homology (above 60% on 26 residues) with the C‐terminal Gly‐rich region of KRMPs. Shematrin‐5 is the single protein of that family to contain an acidic domain, which has homology with aspein. All shematrin transcripts are expressed in the mantle edge of P. fucata, which indicates that this protein family is expressed as components of the prism matrix. Protein sequencing of a urea extract of the water‐insoluble prism matrix showed that shematrins belong to this fraction. It is suggested that shematrins play a role as framework proteins.

D. Partially Characterized Shell Proteins Beside fully sequenced proteins, an increasing number of shell proteins have been partially characterized, as shown in Table IV. These proteins, or protein fractions, enter four categories: the first category comprises proteins that have been partially sequenced and well characterized, in particular on gel or by

Table IV Partially Characterized Shell Proteins

A. Partial Sequence(s)

BIVALVIA

Protein Name

Species

Microstructure

Nacrein‐like protein

Pinctada fucata

Nacre (aragonite)

415 aa

WSM peptides

Pinctada margaritifera

Nacre (aragonite)

p20

Pinctada maxima Pinctada maxima Patinopecten yessoensis Patinopecten yessoensis Crassostrea nippona Crassostrea nippona Crassostrea virginica

Nacre (aragonite) Nacre (aragonite)

Repeats of Q/ KGGGI/L or Q/ KGAGI/L 21 (N‐terminal sequence) 421 aa

A0ZSF3

Nacre (aragonite)

331 aa

A0ZSF4

Nacre (aragonite)

430 aa

A0ZSF5

Nacre (aragonite)

340 aa

A0ZSF6

Nacre (aragonite)

415 aa

A0ZSF7

Foliated (calcite)

59 aa (6 internal fragments) 34 aa (6 internal fragments) 62 aa; 17 kDa (SP); putative poly‐D‐ domain

Nacrein‐like protein Nacrein‐like protein P1 Nacrein‐like protein P2 Nacrein‐like protein C1 Nacrein‐like protein C2 RP‐1 fraction

RP‐1 fraction

Adamussium colbecki

Foliated (calcite)

Caspartin

Pinna nobilis

Prisms (calcite)

Features

Swiss‐Prot Aaccession Number A0ZSF2

References Norizuki, M. (submission author) Be´douet et al., 2006

Be´douet et al., 2001 Norizuki, M. (submission author) Norizuki, M. (submission author) Norizuki, M. (submission author) Norizuki, M. (submission author) Norizuki, M. (submission author) Donachy et al., 1992; Rusenko et al., 1991 Halloran and Donachy, 1995 Marin et al., 2005, 2007

Calprismin

Pinna nobilis

Prisms (calcite)

P12, P16

Mytilus californianus

Prisms (calcite)

45‐, 21‐, and 5‐kDa proteins

Mytilus edulis

Prisms (calcite) þ nacre (aragonite)

55‐, 20‐, and 15‐kDa proteins

Atrina vexillum

Nacre (aragonite)

60‐, 32‐, and 12‐kDa proteins

Nautilus pompilius

Nacre (aragonite)

Dermatopontin2

Satsuma japonica Mandarina aureola

Crossed‐lamellar (aragonite) Crossed‐lamellar (aragonite)

Dermatopontin2

61 aa; 38 kDa (SP) D‐P‐T‐D repeats in the two proteins N‐terminal sequences: 13 aa (45 kDa); 14 aa (5 kDa) N‐terminal sequences: 30 aa (21 kDa) N‐terminal sequences: 13 aa (55 kDa); 15 aa (20 kDa); 6 aa (15 kDa) N‐terminal sequences: 13 aa (60 kDa); 10 aa (32 kDa); 19 aa (12 kDa) 51 aaa

P83631

Marin et al., 2005

Q50K83

Sarashina et al., 2006

98 aaa

Q50K84

Sarashina et al., 2006

Weiner, 1983

Keith et al., 1993

Q9TWS3

Keith et al., 1993

Zhao et al., 2003

Zhao et al., 2003

(Continued)

Table IV Continued

Protein Name GASTROPODA

Dermatopontin1 Dermatopontin2 Dermatopontin1 Dermatopontin2

Species Mandarina aureola Euhadra peliomphala Euhadra amaliae

Dermatopontin1

Euhadra herklotsi Euhadra herklotsi Euhadra brandtii

Dermatopontin2

Euhadra brandtii

Dermatopontin2

Biomphalaria glabrata Biomphalaria glabrata

Dermatopontin1

Dermatopontin1

Microstructure Crossed‐lamellar (aragonite) Crossed‐lamellar (aragonite) Crossed‐lamellar (aragonite) Crossed‐lamellar (aragonite) Crossed‐lamellar (aragonite) Crossed‐lamellar (aragonite) Crossed‐lamellar (aragonite) Crossed‐lamellar (aragonite) Crossed‐lamellar (aragonite)

Features

Swiss‐Prot Aaccession Number

References

64 aaa

Q50K85

Sarashina et al., 2006

64 aaa

Q50K86

Sarashina et al., 2006

64 aaa

Q50K87

Sarashina et al., 2006

51 aaa

Q50K88

Sarashina et al., 2006

98 aaa

Q50K89

Sarashina et al., 2006

a

64 aa

Q50K90

Sarashina et al., 2006

56 aaa

Q50K91

Sarashina et al., 2006

51 aaa

Q50K92

Sarashina et al., 2006

118 aaa

Q50K93

Sarashina et al., 2006

Dermatopontin3

Lymnea stagnalis Lymnea stagnalis Lymnea stagnalis Strombus decorus persicus

Crossed‐lamellar (aragonite) Crossed‐lamellar (aragonite) Crossed‐lamellar (aragonite) Crossed‐lamellar þ prisms (aragonite)

30‐kDa protein

Atrina rigida

Nacre (aragonite)

P10 Periostracin

Pinctada fucata Mytilus edulis

Nacre (aragonite) Periostracum

AP8

Haliotis rufescens

Nacre (aragonite)

Dermatopontin

Biomphalaria glabrata

Crossed‐lamellar (aragonite)

Dermatopontin2 Dermatopontin1 ACLS40

B. No sequence

BIVALVIA

GASTROP

a

129 aaa

Q50K94

Sarashina et al., 2006

129 aaa

Q50K95

Sarashina et al., 2006

109 aaa

Q50K96

Sarashina et al., 2006

40 kDa (SP); 25 aa (2 internal fragments) Aragonite‐ nucleating protein 10 kDa (SP) 20 kDa (SP); 55% Gly AP8‐ (8.7 kDa; SP); AP8‐
(7.8 kDa; SP) 61.2 kDa (SP); 11 aa

Pokroy et al., 2006b

Gotliv et al., 2003

Zhang et al., 2006a Waite et al., 1979 Fu et al., 2005

Marxen and Becker, 1997

Partial sequences obtained by RT‐PCR. For most of them partial sequences are available. We also indicate nonsequenced proteins (bottom), which have been only biochemically characterized. SP ¼ SDS‐PAGE, in this case, the indicated molecular weight is evaluated from the electrophoretic migration of the protein.

246

Marin et al.

chromatography. This is the case of RP‐1, p20, caspartin and calprismin, P12 and P16, and ACLS40. The second category includes proteins that are known only by a partial sequence. This is the case of the proteins of Nautilus pompilius, Atrina vexillum, M. edulis, and the 61‐kDa protein of B. glabrata. The third category includes putative proteins retrieved from partial nucleotide sequences, such as diVerent unpublished nacrein‐like proteins and homologues of dermatopontins in diVerent gastropods. The last category comprises biochemically characterized proteins, for which no sequence is available yet. In the first category, one finds RP‐1, EDTA‐soluble phosphoproteins extracted from foliated calcitic shells, the American oyster C. virginica (Donachy et al., 1992; Rusenko et al., 1991), and the Antarctic scallop Adamussium colbecki (Halloran and Donachy, 1995). These Asp‐rich proteins inhibit the in vitro precipitation of calcite, but this eVect is conveyed by phosphorylated Ser residues, and not by Asp. They exhibit similarities with phosphophoryns. RP‐1 proteins are related to very acidic proteins (MSP‐1, MSP‐2, aspein, Asp‐rich). Similarly, two Asp‐rich proteins, which were purified from the calcitic layer of the American mussel, Mytilus californianus, would also enter this protein family (Weiner, 1983). Two proteins were retrieved from the calcitic prisms of the fan mussel Pinna nobilis (Marin et al., 2005, 2007). One, caspartin, is a 17‐kDa Asp‐rich unglycosylated protein. It binds calcium ions with a low aYnity, is a strong inhibitor of calcite precipitation in vitro, and dramatically aVects the shapes of calcite crystals in interference tests. Caspartin is abundant in the prism‐soluble matrix, but is also present in the nacre‐soluble matrix, but in much lesser amount (about eight times less). Caspartin polymerizes and may form high molecular weight complexes. One polyclonal antibody raised against purified caspartin showed that this protein is localized within and around the prisms (intracrystalline and intercrystalline). Calcite crystals grown in the presence of caspartin exhibit a slight modification of their lattice parameters (Pokroy et al., 2006a), the highest variation being recorded for the c axis. A second prism‐soluble protein, calprismin, was also characterized from Pinna nobilis. Calprismin is an acidic glycoprotein of 38 kDa, for which one‐fifth of the sequence is known. It is enriched in Ala (16%), Asx (15%), Thr (12%), and Pro (12%). Although the N‐terminus is characterized by a particular 4 Cys pattern, it does not exhibit clear homology. The characterization of its glycosyl moieties is in progress. P20 is a protein extracted from the nacre of Pinctada maxima (Be´douet et al., 2001). Its N‐terminus is enriched in Tyr residues. P20 can form oligomers constituted by six monomers linked together by disulfide bridges. Recently, a 40‐kDa protein was extracted from the aragonitic crossed‐lamellar shell of the gastropod Strombus decorus persicus (Pokroy et al., 2006b). Aragonite crossed‐lamellar structure protein 40 (named ACLS40) is Glu/Asp/Ala rich (13.5%, 12.3%, and 12%, respectively). One of its internal sequences has some homology with a vertebrate

6. Molluscan Shell Proteins

247

dimethylaniline monooxygenase, but the significance of this similarity is obscure. Interestingly, ACLS40 has the ability to stabilize the thermodynamically unstable CaCO3 polymorph vaterite. At last, several short peptides of molecular weight inferior to 1 kDa were extracted from the nacre of Pinctada margaritifera (Be´douet et al., 2006). Some of these peptides are rich in Gly residues. Whether these peptides have a cell‐signaling function or whether they are degradation products of shell proteins is not known. The second category comprises proteins, which are only known by partial N‐terminal or internal sequences. One 61‐kDa protein, retrieved from the crossed‐lamellar freshwater gastropod B. glabrata (Marxen and Becker, 1997), does not have clear relationship. A similar situation is encountered with three proteins of the edible mussel, M. edulis of 45, 21, and 5 kDa, respectively (Keith et al., 1993), and of uncertain aYnities. More recently, short sequences of three proteins (60, 32, and 12 kDa, respectively) extracted from the cephalopod, N. pompilius (Zhao et al., 2003), were obtained, but do not share significant homology with other shell proteins. One of the three proteins (55, 20, and 15 kDa), extracted from the nacro‐prismatic bivalve, A. vexillum (Zhao et al., 2003), has homology with the enzyme phosphodiesterase, but this similarity may be fortuitous. The third category includes newly found proteins, the sequences of which were deduced from their incomplete transcript sequences. Diverse nacrein‐like proteins have been retrieved from the Japanese edible oyster, Crassostrea nippona, and the scallop, P. yessoensis (Norizuki, unpublished data). These fresh data confirm that nacrein is a true protein family, the members of which possess highly conserved domains. A second group of shell proteins was retrieved from diverse gastropods, by amplifying cDNAs with degenerate oligonucleotide probes encoding dermatopontin (Sarashina et al., 2006). Incomplete sequences of 13 new dermatopontins were obtained, which exhibit short conserved motifs. The analysis of the expression pattern of the transcripts showed that some dermatopontins are ubiquitous, whereas others are only expressed in the mantle tissue. Only these second ones may be real shell proteins, but this needs to be demonstrated at protein level. At last, the EST work of Jackson et al. (2006) on the abalone, Haliotis asinina, produced a huge amount of sequences encoding secreted proteins. Among these, few are suspected to encode shell proteins. The last category of partially characterized shell proteins corresponds to proteins, which have been completely purified, usually by chromatography or electrophoresis, and for which no sequence data are available. This implies that these proteins have been characterized at amino acid composition level. Several analyses performed on shell matrix ‘‘fractions’’ (HPLC fractions), which consist of mixtures of diVerent proteins (Albeck et al., 1993; Almeida et al., 2000; Pereira‐Mourie`s et al., 2002; Samata, 1990; Wheeler et al., 1988), are excluded from this category. The first purified protein is

248

Marin et al.

periostracin, a soluble precursor of the periostracal layer of the edible mussel, M. edulis (Waite et al., 1979). Periostracin is a 20‐kDa basic and hydrophobic protein with a unique amino acid composition consisting of 55% Gly, 10% Tyr, and 2.2% DOPA. Periostracin is a ‘‘transient’’ protein, which is oxidized and crossed‐linked to form the insoluble periostracum as soon as it is secreted by the periostracal groove. Since the very complete work of Waite et al., no more characterization was performed on that protein. Cariolou and Morse (1988) described two proteins, one 43‐ and one 54‐kDa polypeptides, obtained in native conditions from juvenile and adult nacre tissues of the abalone H. rufescens. Both were enriched in Asx and Gly residues, but the first was more acidic and the second was more hydrophobic. Similar amino acid compositions were obtained from the same species for proteins purified in denaturing conditions (Belcher and Gooch, 2000). Two proteins of 16 and 20 kDa were obtained. They were predominantly constituted of Gly and Asx residues. Fu et al. (2005) characterized one small acidic protein, called AP8, from the nacre of the same abalone species. AP8 has two variants of 8.7 (AP8‐) and 7.8 kDa (AP8‐
). Both are enriched in Asx (35%) and Gly (40%) residues. Interestingly, they represent the first aspartate‐rich proteins found in association with aragonite. AP8 proteins modify drastically the shape of calcite crystals grown on Kevlar. In the nacre‐soluble matrix of the abalone, they may be the most eVective crystal‐shaped modifiers since the soluble matrix depleted of AP8 has a minor eVect on calcite crystals. Other acidic proteins were extracted from the nacreous layer of the bivalve Atrina rigida (Gotliv et al., 2003). When tested in vitro together with chitin and silk (Falini et al., 1996), they induce the formation of ACC prior to its transformation into aragonite. At last, a small protein extracted from the nacre of Pinctada fucata was characterized (Zhang et al., 2006a). This hydrophobic protein, named p10, is Gly rich (37%) and contains high amounts of Leu (16%) and Ala (13%) residues. In in vitro assay, p10 accelerates the precipitation of calcium carbonate and induces the formation of aragonite needlelike crystals. In addition, p10 when tested on two cell lines (MRC‐5 and MC3T3) increases the ALP activity, suggesting that p10 may play a key role in triggering cell diVerentiation for producing bone tissues.

E. Other Molluscan Proteins: The Extrapallial Fluid and the Mantle This long list of proteins, fully or partially characterized, would not be complete without considering other proteinaceous constituents that may be, directly or indirectly, involved in the formation of the shell (Table V). These proteins are located in two diVerent compartments (Fig. 2): the extrapallial space and the mantle epithelium.

Table V Mantle Epithelium and Extrapallial Fluid (EP) Proteinsc

Protein Name

Species

Localization/ Expression

Ferritin

Pinctada fucata

Mantle

Gs Calmodulin

Pinctada fucata Pinctada fucata

Mantle and gill Mantle and gill

CaLP OT47/Tyrosinase

Pinctada fucata Pinctada fucata

Mantle Mantle

CA/Carbonic anhydrase PSKH1

Pinctada fucata

Mantle

Pinctada fucata

Mantle and gill

PFMG1a

Pinctada fucata

Mantle

Calconectin

Pinctada margaritifera

Mantle

ESTs

Haliotis asinina

Mantle

Has‐Lustrin EP

Haliotis asinina Mytilus edulis

Mantle Extrapallial fluid

a

Features 23.6 kDa (cDNA); involved in iron incorporation into the shell 44.3 kDa; G protein (‐subunit s class) 16.8 kDa (cDNA); 4 EF‐hands; involved in Ca transport and secretion 18.4 kDa (cDNA); 4 EF‐hands EC 1.14.18.1; 47 kDa (cDNA); involved in periostracum formation EC 4.2.1.1; 38 kDa (SP) Ca2þ/calmodulin‐dependent protein kinase; 47 kDa (cDNA); involved in calcium metabolism during nacre/pearl formation 15.7 kDa (cDNA); 2 EF‐hands; involved in signal transduction during nacre/ pearl formation 10.4 kDa (cDNA); 2 EF‐hands (1 complete, 1 partial domain) 530 sequences (mantle cDNA library); 25 clones present similarities with putative biomineralization genes from the L. scutum genome Fragment of 68 aa (1 of the 25 clones) 236 aa (sequence of the unprocessed protein, from the cDNA)

Swiss‐Prot Accession Number

References

Q7YW43

Zhang et al., 2003b

Q6TP31 Q6EEV2

Chen et al., 2004 Li et al., 2004

Q3BD18 Q287T6

Li et al., 2005 Zang et al., 2006b Yu et al., 2006

Q4KTY1

Dai et al., 2005

Q3YL59

Liu et al., 2007

Q1KZ60

Duplat et al., 2006

DW986183 to DW986511b

Jackson et al., 2006

A0S725 P83148

Jackson et al., 2006 Hattan et al., 2001; Yin et al., 2005

Nine other PFMG genes have been sequenced (PFMG2 to PFMG12), the function of which does not appear related to shell formation or is unknown. GenBank accession numbers (all the sequences are not yet available in the Swiss‐Prot/TrEMBL databases). c These proteins are putatively involved in shell formation, or in calcium metabolism in relation with the shell calcification. SP ¼ SDS‐PAGE, in this case, the indicated molecular weight is evaluated from the electrophoretic migration of the protein.

b

250

Marin et al.

As recalled in Section II.B, the extrapallial space, localized between the outer mantle epithelium and the growing shell, contains the extrapallial fluid where the precursors of the shell mineralization are supposed to concentrate and self‐assemble in a precise manner. This fluid has been chemically characterized in a number of cases (Misogianes and Chasteen, 1979; Moura et al., 2000). However, curiously, there are hardly any data on the protein constituents of this fluid. The single well‐characterized extrapallial fluid protein is EP of the edible mussel, M. edulis. This protein was first identified as a major component of the fluid (56% of the total fluid proteins), purified and biochemically analyzed (Hattan et al., 2001), before being characterized by molecular biology techniques (Yin et al., 2005). EP is a small acidic glycoprotein (pI 4.4; MW 14 kDa with the glycosyl moieties, 12 kDa when deglycosylated) of 213 residues, which is enriched in His (14%), Asx (12%), and Glx (13%) residues. In native conditions, EP forms 28‐kDa dimers owing to two Cys residues. EP is glycosylated via one asparagine residue in the N‐terminus. EP does not exhibit sequence homology with any known protein. EP possesses numerous short acidic motifs, suggesting that it may interact with calcium ions. The fact that histidine is the most abundant residue in the EP sequence is puzzling because none of the multiple amino acid compositions of shell matrix detected this residue as a main amino acid. On the other side, an earlier work (Hofmann et al., 1989) identified a histidine‐rich calcium‐binding protein of the sarcoplasmic reticulum of the rabbit muscle with a similar amino acid composition (His 13%; Asp 12%; Glu 19%), suggesting that the positively charged His residue may also play a functional (repulsive) role in the binding of calcium ions. Attempts to detect EP in the shell were not successful so far. More generally, this finding points out a limitation of searching only shell proteins. This implies that several proteins, which are secreted in the extrapallial fluid for controlling the shell formation process, may stay or degrade in the fluid, without being incorporated as shell matrix components. If these ‘‘silent’’ and ‘‘transient’’ proteins exist, then they will be retrieved only by overlapping EST work on secretome, as performed by Jackson et al. (2006), and shell matrix proteomics. Beside extrapallial fluid proteins, several proteins may play a very important role in the process of shell formation. For example, these are proteins specific of the molluscan mantle. The analysis of mantle proteins is far beyond the scope of this chapter and would require a review by itself. However, because we suspect mantle‐specific proteins to be key players in the physiology of calcification, we mention some of them in Table V. As illustrated, most of them have been very recently characterized. Many are not secreted because they do not exhibit signal peptides. They may play important intracellular functions, such as the regulation of intracellular calcium. In this group, one finds calconectin, a calcium‐binding protein of the mantle of the Polynesian pearl oyster (Duplat et al., 2006), calmodulin (Li et al., 2004),

6. Molluscan Shell Proteins

251

or ferritin (Zhang et al., 2003b). Concerning ferritin, it is interesting to note that this protein may be involved in iron concentration in the mantle cells and its subsequent incorporation in the shell. A tyrosinase (Zhang et al., 2006b), putatively involved in the formation of the periostracum, has also been retrieved. We can predict that the number of mantle‐specific proteins will dramatically increase soon, as illustrated by the impressive EST work of Jackson et al. (2006b) on juvenile mantle tissue of the abalone, Haliotis asinina. This work generated 530 sequences, encoding nonsecreted and secreted proteins, from which 85 encode secreted proteins. Clearly, the coming years will require accurate in silico analyses.

F. Remarks on Molluscan Shell Proteins Since 1996, when the first full sequence of nacrein was published, the number of identified proteins has grown exponentially, and we do not see any reason why this evolution would cease. However, the expanding corpus of published shell protein sequences does not hide some obscure gaps in our knowledge of these proteins. At first, a number of shell proteins have been characterized at the transcriptional level but have not been clearly identified as constituents of the shell matrix. Nacrein, N14/N16, prismalin‐14, dermatopontin, mucoperlin, AP7, and AP24 represent the few cases where the link between the transcript and the protein was established. Moreover, we dispose now of more than 40 protein sequences, but we still have a rather unprecise idea of their respective putative functions, which are mainly deduced from primary structure analysis and homology search with known proteins. In several cases, because of low homologies (for example, the N‐terminus of calprismin), this approach is ineVective. In addition, many of these proteins exhibit sites for posttranslational modifications, but we have a poor idea of the biochemical characteristics of these modifications, which can so drastically aVect the properties of the protein core. Dermatopontin is so far the single example where its sugar moieties were precisely characterized. The discrepancy between the topographic models, presented in Section III.B, and the known proteins is another problem, and a validation of the model by direct localization of each shell matrix protein on and within the shell biominerals would be extremely helpful. The last important drawback is that our knowledge is almost entirely limited to ‘‘economically interesting models,’’ in particular, the pearl oyster and the abalone, two mollusks, which exhibit nacro‐prismatic shell textures. It is very unlikely that these two studied species are representative of the huge phylogenetic and textural diversity of the phylum Mollusca, which comprises more than one hundred thousand living species. In spite of these limitations, some general characteristics can be sketched. As shown in Fig. 5, the molluscan shell proteins present indeed some

252

Marin et al.

particularities in their distribution according to their theoretical molecular weight and their pI. Along the pI axis, most of the proteins associated with calcite are either very acidic (below 4.5) or basic, the single exceptions being MSI7 and Pfty1. On the contrary, most of the proteins associated with aragonite occupy a central position, in a pI range 4.5–7.5, the exceptions being perlustrin, perlwapin, and lustrin A. Along the y axis, the protein distribution is bimodal, with a majority of ‘‘small’’ proteins (34) in the molecular weight range 6.5–40 kDa, and 9 proteins above 48 kDa. From Fig. 5 and sequence analysis, the outline of diVerent protein families can be roughly sketched. One distinguishes the N14/N16/pearlin family, which forms a very homogeneous group, the nacrein family, the shematrins, the KRMPs. In the acidic group, the Asp‐rich proteins form a group by themselves, which also include aspein as well. MSP‐1 and MSP‐2 would constitute a distinct group. One characteristic found in many of these proteins is their modular organization, each module corresponding to a functional domain. Consequently, many of these proteins are multifunctional, a common feature found in several proteins of the extracellular matrix (Engel, 1991, 1996). Some domains are clearly identified, like the carbonic anhydrase‐like domains of nacrein, the IGF‐BP domain of perlustrin, or the C‐type lectin domain of perlucin. Many domains are composed of tandemly arranged repeats (mucoperlin, MSI31) or of an alternance of two (Pro and Cys modules of lustrin A) or three (SG, D, and K modules of MSP‐1) repeats. The repeats can be extremely short: one residue (poly‐Gly, poly‐Ala, poly‐Asp, or poly‐Ser); two residues, like GS (lustrin A) or GN (nacrein); six residues (GGYGXX in shematrin‐4/5; GGGGVI in shematrin‐3; XSEEDY in MSI31); several residues, like in mucoperlin (31 residues per repeat) in lustrin A (up to 88 residues in the Cys‐rich repeats), or in MSP‐1 (95 residues in the D modules). These low complexity sequences can constitute an important part of the protein sequence. Consequently, they considerably influence the overall amino acid composition, which is dominated by few amino acids: Gly, Asp, Ser, then Asn, Tyr, Ala, Pro, Cys, and Leu. The other consequence is that these low‐complexity domains cannot be exploited for sequence comparison, and the fact that short segments of them match does not infer a phylogenetic proximity. Inside families, sequence comparisons can produce high similarities. But sequence comparison of proteins, from family to family, does not produce significant similarities and the homologies on the full‐ length sequences are generally low. Interestingly, we noticed that similar short motifs, which are not necessarily of low complexity, can be found in diVerent proteins (Marin et al., 2007), suggesting that these functional motifs are reused as ‘‘building blocks’’ by diVerent mineralizing ‘‘mosaic’’ proteins with diVerent functions. This implies that these proteins have been constructed by a genetic tinkering, which would have allowed the recombination of short nucleotide sequences, for producing novel proteins.

6. Molluscan Shell Proteins

253

A remarkable illustration of this genetic tinkering mechanism is given by the shematrin‐2/MSI31 examples. The long structural hydrophobic N‐termini (227 residues) of these two proteins have 98% homology, but their C‐termini are completely diVerent: that of shematrin is strongly basic and is supposed to work as an anchoring domain, while that of MSI31 is extremely acidic and would be involved in nucleating crystals. If the diVerence in the C‐terminal sequence is not due to a sequencing mistake, that is, a frameshift in the sequence reading, then this homology is intriguing and calls for diVerent possible mechanisms by which these proteins are constructed. Some explanations lie at the genome organization level, the other at the transcripts level. Among the first one, exon shuZing is a very likely mechanism (Gilbert, 1978), for creating new functions from old ones. Exon shuZing is very common in the genes encoding extracellular matrix proteins (Kolkman and Stemmer, 2001; Patthy, 1996, 1999, 2003). Exon shuZing is likely when the borders of an exon coincide with the beginning and end of the protein functional domain. The duplication, permutation, and rearrangement of these exons result in novel genes with new functions. Exon shuZing can occur through retrotransposition (Eickbush, 1999) or illegitimate recombination (van Rijk and Bloemendal, 2003). In the present case, this could be tested by looking in the genome of Pinctada fucata whether the 227 matching N‐terminal residues of shematrin/MSI31 correspond to an exon. Another way of explaining the partial homology of shematrin‐2 and MSI31 is the insertion/ deletion of one base in the duplicated gene, which leads to a frameshift during transcription, and the synthesis of two proteins with diVerent C‐termini. At the transcriptional level, alternative splicing can be inferred as another mode for synthesizing two variants from a single gene, if the two exons that encode the two C‐termini are placed in tandem in the gene. Alternative splicing of the mRNA may then produce either shematrin‐2 or MSI31. This hypothesis can also be tested, both at the genome and at the transcript levels. There are still simple unanswered questions: how many proteins are required for building a shell? This question may appear naive, but the answer is not simple and has a lot to do with the technique used. In a monodimensional gel, a soluble gel extract is usually characterized by the presence of few major bands, in addition to several minor ones, embedded in a smear of nondiscrete macromolecules (Marin et al., 2001). By this technique, about 10–15 proteins can be visualized, in particular when the sensitive silver staining is used or when accurate fixation method is employed (Gotliv et al., 2003). The same extract tested on a bidimensional gel brings another answer. Some protein bands, which appeared homogeneous in one‐dimension, can be constituted of very diVerent proteins (Marie et al., 2007). Furthermore, to make the problem a little bit more complicated, matrix proteins can exhibit several posttranslational modifications, that is, phosphorylations, glycosylations, sulfations. DiVerent phosphorylation patterns of a single protein will be translated on the two‐dimensional gel by series of horizontally aligned

254

Marin et al.

spots of diVerent pI. A preliminary answer can be given by assuming that the sea urchin spicule and the molluscan shell systems exhibit the same degree of complexity in their proteinaceous composition. If so, we can expect that 2D gels will visualize few tens of diVerent protein spots in the soluble shell matrix (Killian and Wilt, 1996). Working at secretome level, by analyzing transcripts that encode secreted proteins, brings a third more complete answer (Jackson et al., 2006). Such studies reveal in particular the ‘‘transient and silent’’ proteins, which are not incorporated in the shell (some extrapallial fluid proteins), as well as highly crossed‐linked shell proteins, which cannot be solubilized and analyzed by classical biochemical methods. At last, analysis of the transcriptome will reveal several more proteins of the intracellular machinery, involved in particular in the cellular traYcking, in the proper folding of extracellular matrix components.

V. Origin and Evolution of Molluscan Shell Proteins A. The Cambrian Origin of Mollusk Shell Mineralization Puzzling questions concern the origin of these shell proteins. Where do they come from? How were they recruited? Are they heavily constrained from an evolutionary point of view? These questions have to be replaced in the general context of the so‐called ‘‘Cambrian explosion.’’ Indeed, like most of the metazoan lineages, mollusks started to mineralize at the dawn of the Cambrian times, in a very short time interval, about 544 million years ago (Bengtson, 1992; Conway Morris, 2001). In the fossil record, the appearance of biologically controlled minerals among metazoans, including silica, calcium phosphate, and calcium carbonate skeletons, was the most visible aspect of the so‐called ‘‘Cambrian explosion,’’ by far the most important event in the metazoan world (Conway Morris, 1998; Knoll and Carroll, 1999; Shubin and Marshall, 2000). As shown in Fig. 6, the mollusk fossil record indicates that the main classes had representatives in the Cambrian (Lecointre and Le Guyader, 2001; Runnegar, 1996), including polyplacophores (Matthevia, Upper Cambrian), monoplacophores (Latouchella, Anabarella, Lower Cambrian), gastropods (Kobayashiella, Upper Cambrian), cephalopods (Plectronoceras, Upper Cambrian), bivalves (Pojetaia, Fordilla, Lower Cambrian). Furthermore, it seems that mollusks were able to exploit rapidly most of the design possibilities for building their skeleton [see the ‘‘Skeleton Space’’ theory of Thomas et al. (2000)]. Although there are no certainties on the Precambrian evolutionary history of mollusks, the fossil record suggests that the phylum was also one of the components of the well‐known Ediacaran fauna. The species Kimberella quadrata is generally recognized as a shell‐less mollusk (Fedonkin and Waggoner, 1997). Other Precambrian fossils, like the Chinese Circotheca

255

6. Molluscan Shell Proteins Cambrian Vendian

Lower 545 MY

Middle

Ordovician Upper

Lower 488 MY

Middle

Upper

Class

444 MY

Solenogastra Caudofoveata

Shell-less molluscs of ediacara

Polyplacophora Monoplacophora

1

2

Gastropoda Cephalopoda Bivalvia Scaphopoda

Figure 6 Origin and phylogeny of the phylum Mollusca. The phylogeny of mollusks is that proposed by Lecointre and Le Guyader (2001) in which Solenogastra and Caudofoveata occupy a basal position. Node 1 represents the subphylum Eumollusca, whereas node 2 represents the superclass Conchifera (shell‐bearing mollusks). Except scaphopods (first fossil record: Upper Ordovician, 450 million years ago), all eumolluscs emerged in the Cambrian, bivalves and monoplacophores emerging as early as the Lower Cambrian, about 540 million years ago, and taking part of the ‘‘Cambrian explosion.’’

longiconica or the American Wyattia reedensis, are supposed to be mollusks, although their taxonomic aYnities are still controversial. Another argument which suggests that mollusks were already in existence in the Precambrian comes from phylogenetic reconstructions based on molecular markers (18S rDNA). In general, all reconstructions agree with a silent ‘‘revolution,’’ somewhere in the Proterozoic, the late part of the Precambrian. This revolution led to the identification of successive radiations among the bilaterian metazoans: a first event isolating deuterostomians from protostomians, a second event splitting proteostomians in lophotrochozoans and ecdysozoans (Adoutte et al., 2000; Balavoine and Adoutte, 1998). Whatever the radiation scenario is, this means that mollusks, like several other metazoan phyla, acquired the capacity to form a mineralized exoskeleton far after their emergence as a phylum. This has direct implications on the way the ‘‘molecular tool box’’ (including skeletal proteins) required for mineralizing was recruited. To explain how calcification was implemented in nonmineralized metazoans, two ‘‘extreme’’ scenarios are possible: on one hand, calcification was inherited from ancestral functions. If so, it resulted from the recruitment and the orchestration of Precambrian functions, not related at all with mineralization (Marin et al., 2003b). The best terminology for describing this process is ‘‘exaptation,’’ a word invented by Gould and Vrba (1982), 25 years ago. The alternative scenario implies that calcification was acquired independently by

256

Marin et al.

the diVerent metazoan lineages. If so, similarities are the result of adaptive convergence. Such convergence would be explained by similar physical ‘‘constraints’’ that would ultimately drive the evolutionary process at the molecular level: secretion of an extracellular acidic template for crystal nucleation; secretion of inhibitors for controlling crystal growth and allowing crystal to grow only where needed. Let us examine these two hypotheses.

B. The ‘‘Ancient Heritage’’ Scenario By many aspects, the ‘‘ancient heritage’’ scenario is ‘‘intellectually appealing,’’ although we admit that it is supported by a thin corpus of disparate observations and laboratory data. The first argument in favor of the ‘‘ancient heritage’’ hypothesis comes from the fossil record. Detailed descriptions of shell textures of early mollusks (Feng and Sun, 2003; Kouchinsky, 2000) indicate that, although far less numerous than now, Cambrian shell textures were already complex and diversified. In particular, it is striking to observe that Cambrian prismatic or nacreous textures resemble that of living species. This clearly suggests two remarks: first, the number of textural combinations is limited and mollusks exploited rapidly all the genetic possibilities given to them for building coherent assemblages of crystals. Second, nacre and prisms, which are usually considered ‘‘primitive,’’ must be extremely ‘‘evolutionary’’ constrained. Nacre in particular is a strikingly stable and perennial texture. Intuitively, the evolutionary constraint on the texture may also apply on the macromolecules, which shape this texture. If so, then nacre matrix is necessarily built from ‘‘ancient’’ proteins, and it is tempting to establish a parallel between the molluscan nacre matrix and vertebrate bone collagen, that is, an early invention of a successful molecular tool, and its subsequent durability over the geologic ages. In 1980, in a key paper, Lowenstam and Margulis assumed that the appearance of calcium carbonate exoskeletons in the Lower Cambrian was preceded by a phase during which the regulation of intracellular calcium ions was set up. In other words, producing biominerals outside the cell requires necessarily an accurate domestication of intracellular calcium fluxes. Calcium acquired indeed a central position in the cell physiology. It is a second messenger for diverse metabolic pathways; it is essential for muscle contraction, secretion, and cell adhesion. Intracellular calcium is stored in diVerent highly structured cellular compartments (Pozzan et al., 1994), which are able to quickly deliver calcium pulses when needed. The assumption of Lowenstam and Margulis that the intracellular machinery required for handling intracellular calcium was already finely tuned when animals started to calcify suggests at least that parts of the intracellular calcifying machinery are ancient.

6. Molluscan Shell Proteins

257

Another point to mention concerns the serological comparisons established in the last 20 years with antibodies elicited against diverse calcified matrices. Remarkable immunologic interphylum cross‐reactivities were observed in a number of cases: vertebrate to echinoderm (Veis et al., 1986), echinoderm to prochordates (Lambert and Lambert, 1996), or mollusks to brachiopods (Marin, unpublished data). Within the phylum Mollusca, we also detected striking interclass cross‐reactivities (Marin et al., 1999), which were related neither to taxonomy nor to shell microstructures. Although cross‐reactivities may be fortuitous, due to similar overall topographies of the epitopes, they may be also highly significant of the presence of short conserved epitopes across calcifying phyla. About 10 years ago, we hypothesized that epithelial mucus substances might have been precursors of the calcifying matrices among mollusks and corals (Marin et al., 1996). Our scenario called ‘‘anticalcification’’ was based on the ‘‘functional duality’’ of acidic soluble matrices, that is, their ability to inhibit crystal growth in solution, and to promote crystal nucleation, when bound to a template. We assumed that, in the Proterozoic, epithelial mucus substances were working as inhibitors of mineralization for precluding spontaneous calcium carbonate crystallization on ‘‘naked’’ ediacaran metazoans, in a context of a heavily supersaturated ocean. At the Precambrian/Cambrian transition, the same mucus inhibitors could have been recruited to keep crystallization in check. The fact that epithelial mucins (mucus proteins) are extremely ubiquitous, that they are often associated with inhibiting systems (salivary mucins, urinary mucins) or calcifying systems (gallbladder mucins), and the finding of mucoperlin, a mucin‐like protein associated with the nacre of the bivalve P. nobilis (Marin et al., 2000) emphasize the role and the putative ancestry of this protein family in metazoan calcification. Another striking example that further supports the idea of ‘‘ancient heritage’’ is the bioactivity of nacre for bone repair (Atlan et al., 1997; Lopez et al., 1992). In diVerent series of in vivo experiments, Lopez and coworkers showed that nacre implants have the ability to promote bone repair without provoking rejection. Their experiments were confirmed by Liao et al. (1997, 2000, 2002). Later on, Lopez and coworkers demonstrated that the water‐ soluble nacre matrix was the fraction that contains the bioactive molecule(s) (Almeida et al., 2001) and that this fraction induced the diVerentiation of preosteoblastic cell lines (Rousseau et al., 2003). The most likely is that nacre contains a diVusible signal transducing factor that can be recognized by membrane receptors of bone‐forming cells, osteoblasts. So far, this factor has not been identified, but putative candidates include bone morphogenetic proteins (BMPs), members of the TGF‐
superfamily. BMPs have been identified in mollusks (Lelong et al., 2000, 2001; Matsushiro and Miyashita, 2004), but it has not been yet demonstrated that these proteins are incorporated within the shell matrix. Whatever the signaling molecule is,

258

Marin et al.

the fact that a molluscan signal transducer is able to activate a vertebrate system pleads for the antiquity of the signaling function in biomineralization (Westbroek and Marin, 1998). Finally, the analysis of the primary structure of shell matrix proteins may also contribute to suggest the antiquity of some conserved domains. The most revealing examples are the carbonic anhydrase domains of nacrein and N66. Carbonic anhydrases constitute a complex family of zinc‐containing enzymes, known in the three kingdoms, Archaea, Bacteria, and Eukarya (K.S. Smith et al., 1999). They are subdivided in three classes designated ,
, and  that evolved independently and do not exhibit sequence homologies. All known metazoan carbonic anhydrases belong to the ‐class (Lindskog, 1997), while those of the kingdom Bacteria belong to the ‐,
‐, and ‐classes. Archaea possess only
‐ and ‐classes of carbonic anhydrases (Smith and Ferry, 2000). The high homology of nacrein with human carbonic anhydrase II is puzzling. Because the reversible conversion of carbon dioxide into bicarbonate is an ancestral function, because this function is primordial in calcium carbonate biomineralization (bicarbonate is one of the precursor mineral ions for calcification), and because carbonic anhydrase domains have been found in bivalve and gastropod nacres, it is hard to conceive that such a key function in shell function results from a recent recruitment. Another point to mention concerns the Asp‐rich domains. As mentioned by Tsukamoto et al. (2004) and Gotliv et al. (2005), aspein as well as Asp‐rich proteins exhibit a homology with calsequestrins, very acidic proteins of the sarcoplasmic reticulum (Beard et al., 2004). Calsequestrins are high‐capacity, low‐aYnity calcium‐binding proteins. They can bind up to 50 cations, in particular via their short poly‐D motifs. Asp‐rich as well as aspein are supposed to display a similar function. Few years ago (Marin et al., 2003b), we have suggested that Asp‐rich domains of acidic molluscan shell proteins may derive from polyanionic domains of calsequestrins. Although our assertion is speculative, it needs to be tested further. If it happened to be true, then it would give consistency to the old intuition of Lowenstam and Margulis (1980) about the prerequisite for calcification. All these combined data taken together suggest that, at least, some shell components may be ancient and recruited early in the evolution of the phylum.

C. The ‘‘Recent Heritage and Fast Evolution’’ Scenario Facing the ‘‘ancient heritage’’ scenario, four recently published papers based on unquestionable data demonstrate that, on the contrary, many metazoan skeletal proteins evolved independently and that the calcification secretome may have a much higher plasticity than expected.

6. Molluscan Shell Proteins

259

One important paper from Livingston et al. (2006) has scanned the full genome of the sea urchin Strongylocentrotus purpuratus, in search of all the putative biomineralization proteins. Several sea urchin spicule proteins are known from classical molecular biology approach (reviewed in Wilt et al., 2003). From sequence homologies, and from the fact that biomineralization protein genes are organized in clusters in the genome of S. purpuratus, additional genes were retrieved. In total, the genes encoding spicule matrix proteins constitute a small family of 16 sea urchin specific genes, which do not have homology in other deuterostomes, that is, chordates and hemichordates. Reversely, many of the vertebrate biomineralization proteins do not have their equivalent in the sea urchin genome. This suggests that the mineralizing matrices of echinoderms, on one side, and of vertebrates, on the other side, are two ‘‘independent inventions.’’ Conversely, the protein components of the two matrices may have a unique origin, which has been completely obliterated at the primary structure level, due to loose constraints and considerable genetic drift. In vertebrates, a rather similar conclusion was drawn by reviewing all the members of the secretory calcium‐binding phosphoprotein family (SCPP) (Kawasaki and Weiss, 2006). Members of this family, including three major enamel matrix proteins, five proteins necessary for dentin and bone formation, milk casein, and salivary proteins, arose from a single ancestor by tandem gene duplications (Kawasaki et al., 2004). Because of the huge variability of the primary structures of the SCPP family members, it was concluded that ‘‘while mineralized tissues are retained during vertebrate evolution, the underlying genetic basis has extensively drifted.’’ This clearly suggests that the real evolutionary constraint at the tissue level does not apply to the primary structure of extracellular matrix proteins that control the mineralization process. By extrapolating these results, we can assume that the evolutionary constraint is maybe more eVective at the secondary or tertiary structure levels, or even at the supramolecular level. In a paper, Jackson et al. (2006) screened all the transcripts expressed during shell calcification of the abalone, H. asinina, in particular the transcripts encoding secreted proteins. Surprisingly, they found out that 85% of the secreted proteins are unknown. A comparative scan between the obtained EST sequences and the genome of the patellogastropod, Lottia scutum, showed that only 19% of the secreted proteins of H. asinina have their homologue in L. scutum, which constitutes another surprise. One of the main conclusions of this study is that the shell is constructed from a rapidly evolving secretome. Another strategy was employed by Sarashina et al. (2006) to test the antiquity of dermatopontin, the shell proteins of the freshwater snail B. glabrata, characterized by Marxen et al. (2003b). They obtained the homologues of B. glabrata dermatopontin in seven other freshwater and land snails,

260

Marin et al.

with degenerate primers based on the dermatopontin amino acid sequence. The dermatopontin gene is ancient because of its widespread repartition in several metazoan lineages and because of its general function in extracellular matrix assembly. However, the reconstruction of the phylogeny of molluscan dermatopontins demonstrates that the recruitment of dermatopontin as a shell matrix protein occurred twice independently in the two tested gastropod lineages, between Carboniferous and Tertiary. Although counterintuitive and antiparsimonious, a scenario of two independent recruitments may emphasize that the evolutionary constraints do probably not work at the level of the genes encoding shell proteins but somewhere upstream. Alternatively, it is also possible that the recruitment of the dermatopontin gene for calcification was ancient, and abandoned later by many molluscan lineages (pseudogenes) and conserved in others. In the future, this hypothesis should be confirmed by investigating the presence of dermatopontin‐like proteins in the shell of several mollusks, including bivalves and cephalopods, not only modern most derived crossed‐lamellar gastropods. At last, loose evolutionary constraints may also apply to the very acidic shell proteins that are supposed to bind calcium carbonate crystals with a high aYnity. Loose constraints can be tested experimentally with combinatorial phage display libraries. These commercially available libraries consist of a population of bacteriophages, genetically engineered to carry peptides located at the end of one of the virus coat protein. For short peptides (typically seven residues), the library contains all the possible peptides combinations with the usual 20 amino acids. The library can be put in contact with mineral surfaces. By applying rinsing steps of increasing stringency, the phages that exhibit the highest aYnity with the mineral surface are selected and can be amplified and sequenced. Few rounds of adsorption‐wash‐amplification permit the selection of the most strongly bound peptides. Experiments of this nature performed on calcium carbonate surfaces gave interesting results (Belcher and Gooch, 2000). In particular, they showed that several possible peptides with very diVerent sequences could bind calcite or aragonite surfaces with a good aYnity. This suggests that the mineral‐interacting domains on shell proteins may be not particularly constrained and may be susceptible of amino acid substitutions, which do not aVect their surface‐binding ability as long as the crystal‐binding motifs of two to three amino acids are conserved.

D. Long‐Term Evolution of Shell Matrices and Microstructures: The Bivalve Example Clearly, all the data taken together suggest that the molluscan shell secretory repertoire has certain ‘‘plasticity’’and ‘‘evolvability’’ from group to group. The impression is that a subtle balance is exerted between the plasticity of the

6. Molluscan Shell Proteins

261

primary structures of the terminal products, the shell proteins, and evolutionary constraints that operate at upstream level and keep the durability of the microstructures, for example. The complex relationships between shell microstructures and their associated matrix call for further developments. In particular in bivalves, the shell displays a certain number of microstructures, organized in layers, from two to four. Shell microstructures are recognized as one of the parameters potentially used in paleontology (after dentition, ligament insertion, adductor scars, pallial line, and shell shape) for classifying fossil bivalves (Skelton, 1985). Bivalve shell microstructures were extensively described by Boggild (1930), Kobayashi (1964), Oberling (1964), Taylor et al. (1969, 1973), Carter (1980), Carter and Clark (1985), Shimamoto (1986), Carter (1990), and others. Attempts were made to sketch some evolutionary trends in bivalve microstructures, in particular by Taylor (1973), Carter (1980), Kobayashi (1980, 1991), Uozumi and Suzuki (1981), Shimamoto (1993) but the task is singularly diYcult. First, similar microstructures may appear in taxa, which are widely separated in the phylogenetic tree, while they are absent in taxa of intermediate position. The best example is the crossed‐lamellar structure found in arcoid and veneroid bivalves. A second source of confusion is the fact that the sequence by which these microstructures are associated in the shell varies in the diVerent taxa: only in the extant venerid family, 12 combinations were distinguished by Shimamoto (1986), while Uozumi and Suzuki (1981) listed 47 combinations occurring in the bivalve class as a whole. Clearly, the bivalve shell microstructures are the result of a mosaic evolution, and adaptive convergences recur at diVerent taxonomic levels (Carter and Clark, 1985; Taylor, 1973). In term of matrix plasticity, one can wonder whether similar (in their shape) microstructures can be produced by very diVerent shell matrices, or conversely, whether diVerent microstructures are produced by very similar repertoires of proteins. This question can be tackled in terms of energetic cost. Two decades ago, Palmer (1983, 1992) calculated the cost of calcification in diverse mollusks. He observed that the heaviest energetic cost for a mollusk corresponded to the synthesis of the matrix and not the deposition of the mineral phase. He concluded that mollusks that have a high content of organic matrix in their shell were evolutionarily disadvantaged in comparison to those with a low content. This remarkable finding, which had passed largely unnoticed, can be integrated in the context of the long‐term evolution of bivalve microstructures. Nacro‐prismatic textures exhibit a high content of organic matrix (up to 4–5% of the shell weight, most of which is highly crossed‐linked and insoluble) and are therefore considered as ‘‘primitive’’ (Carter, 1980; Taylor, 1973). On the other side, crossed‐lamellar/homogeneous textures have a much lower shell matrix content (below 0.5%), most of which is soluble. Although crossed‐ lamellar microstructures appeared early in bivalves (see the lucinid family),

262

Marin et al.

most of the veneroid families (‘‘modern’’ heterodont bivalves), which all exhibit combinations of crossed‐lamellar/homogeneous microstructures, appeared in the Mesozoic times, and knew a rapid post‐Jurassic expansion, becoming consequently the dominant bivalves (in terms of taxonomic diversity) in the today’s oceans (Carter, 1980). The late but rapid burst of veneroid crossed‐ lamellar bivalves suggests that they ‘‘invented’’ a new shell matrix repertoire for controlling shell biomineralization and texture, a novelty, which gave them an unquestionable evolutionary advantage. At the microstructural level, such a possibility had been considered, in particular in the dichotomic phylogeny of bivalve microstructures proposed by Uozumi and Suzuki (1981). At the matrix level, putative diVerences in the matrices associated with crossed‐lamellar and noncrossed‐lamellar bivalves were evidenced by serological comparisons performed with diVerent shell matrix antibodies (Muyzer et al., 1984; Marin, unpublished data). Clearly, this needs to be tested now, by EST techniques, which would allow overall comparisons of the secretory repertoires. We predict that shell matrices can be represented as single points in a multidimensional space, where microstructures would represent strange attractors (Sprott, 1993). Such a representation would conciliate the apparent plasticity of shell matrix proteins and the evolutionary constraints at the microstructural level.

VI. Concluding Remarks Sketching some long‐term perspectives for the future in biomineralization research is by many aspects risky and presumptuous but forces us to look back in the past and consider the evolution of the discipline over the last decades. Forty years ago, the problem of the characterization of the molluscan shell matrix was mainly tackled at amino acid level, and the literature dealing with amino acid compositions is abundant (Gre´goire, 1972). At that time, this level of analysis was discriminant enough for microstructural, phylogenetic, or environmental purposes. However, it did not give any chance to understand the dynamic of the shell formation process. In the early 1970s, the discovery of the soluble matrix opened new biochemical opportunities by investigating the shell matrix at the protein level, in spite of certain technical limitations, inherent in the shell matrix itself. In the 1990s, the introduction of molecular biology techniques reinforced the tendency to work at the protein level, as, at the same time, it opened perspectives to characterize matrix constituents at the transcriptional level. We are still in this phase, but are clearly moving toward the upper ‘‘ome’’ level of analysis, that is, secretome (Jackson et al., 2006) and proteome (Be´douet et al., 2007), which drives us to supramolecular chemistry and hierarchy in biomineralization. Interestingly, these upper levels of analysis bring us

6. Molluscan Shell Proteins

263

back to the physiology of the calcifying mantle tissue, one aspect that had been neglected for years. Does it mean that we have now acquired a better understanding on how a mollusk makes its shell? For some aspects, the answer is clearly yes. For some others, our ignorance is abyssal, considering the size of the phylum Mollusca, and the few studied species. The general impression is that the more biological models we study, the more matrix proteins we find. Enumerating precisely the proteins is a necessary required step for drawing the outline of the diVerent functional domains and protein families, for phylogenetic purposes, but, once again, an exhaustive catalog of the shell proteins will not help so much in understanding the shell formation. We are confronted to the limits of a reductionist approach, which consists in explaining one given physical property (like the polymorph selection) by the presence of one or two matrix constituents. Shell matrices, when considered as biological objects, exhibit ‘‘emerging properties’’ that are much more than the sum of their parts, matrix proteins. Future attempts to apply reverse genetics (gene knock down with morpholino oligos) on mollusk larvae, accurate quantification of the diVerent transcripts levels in distinct parts of the calcifying mantle, and finally extended genome‐to‐genome comparisons may allow filling partially the gap. In particular, these approaches should permit to relate the structure of molluscan shell proteins to their functions, to unveil the spatiotemporal expression of the secretory sequence, and at last establish the secretory repertoire and propose phylogenies of calcifying proteins. However, serious conceptual eVorts will additionally be required for explaining how these beautifully organized shell biominerals like nacre or prisms emerge from liquid systems.

Acknowledgments This chapter is a contribution to an ‘‘Aide Concerte´e Incitative Jeunes Chercheurs’’ (ACI JC 3049) awarded to F. M. by the French ‘‘Ministe`re De´le´gue´ a` la Recherche et aux Nouvelles Technologies’’ for the period 2003–2006. For the period 2007–2010, this work is supported by an ANR project (ACCRO–Earth, ref. BLAN06–2_159971, coordinator Gilles Ramstein, LSCE, Gif/Yvette). In 2004, the ‘‘Conseil Re´gional de Bourgogne’’ (Dijon, France) provided financial supports for the acquisition of new equipment in Biogeosciences research unit. F.M. thanks Claudie Josse (Laboratoire de Re´activite´ des Solides, UB, Dijon) for her help in handling SEM, and EGIDE (PAI Cogito 09084XG) for promoting the collaboration with Professor D.M. At last, but not least, F.M. thanks Alain Godon for his kind help in redrawing Figs. 2 and 3.

References Addadi, L., Moradian, J., Shay, E., Maroudas, N. G., and Weiner, S. (1987). A chemical model for the cooperation of sulfates and carboxylates in calcite crystal nucleation: Relevance to biomineralization. Proc. Natl. Acad. Sci. USA 84, 2732–2736. Addadi, L., Joester, D., Nudelman, F., and Weiner, S. (2006). Mollusk shell formation: A source of new concepts for understanding biomineralization processes. Chem. Eur. J. 12, 980–987.

264

Marin et al.

Adoutte, A., Balavoine, G., Lartillot, N., Lespinet, O., Prud’homme, B., and De Rosa, R. (2000). The new animal phylogeny: Reliability and implications. Proc. Natl. Acad. Sci. USA 97, 4453–4456. Aguinaldo, A. M. A., Turbeville, J. M., Linford, L. S., Rivera, M. C., Garey, J. R., RaV, R. A., and Lake, J. A. (1998). Evidence for a clade of nematodes, arthropods and other moulting animals. Nature 387, 489–493. Albeck, S., Aizenberg, J., Addadi, L., and Weiner, S. (1993). Interactions of various skeletal intracrystalline components with calcite crystals. J. Am. Chem. Soc. 115, 11691–11697. Albeck, S., Weiner, S., and Addadi, L. (1996). Polysaccharides of intracrystalline glycoproteins modulate calcite crystal growth in vitro. Chem. Eur. J. 2, 278–284. Almeida, M. J., Milet, C., Peduzzi, J., Pereira, L., Haigle, J., Barthe´lemy, M., and Lopez, E. (2000). Effect of water‐soluble matrix fraction extracted from the nacre of Pinctada maxima on the alkaline phosphatase activity of cultured fibroblasts. J. Exp. Zool. 288, 327–334. Almeida, M. J., Pereira, L., Milet, C., Haigle, J., Barbosa, M., and Lopez, E. (2001). Comparative eVect of nacre water‐soluble matrix and dexamethasone on the alkaline phosphatase activity of MRC‐5 fibroblasts. J. Biomed. Mater. Res. 57, 306–312. Asakura, T., Hamada, M., Ha, S. W., and Knight, D. P. (2006). Conformational study of silk‐ like peptides modified by the addition of the calcium‐binding sequence from the shell nacreous matrix protein MSI60 using 13C CP/MAS NMR spectroscopy. Biomacromolecules 6, 1996–2002. Atlan, G., Balmain, N., Berland, S., Vidal, B., and Lopez, E. (1997). Reconstruction of human maxillary defects with nacre powder: Histological evidence for bone regeneration. C. R. Acad. Sci. Paris 320, 253–258. Balavoine, G., and Adoutte, A. (1998). One or three Cambrian radiations? Science 280, 397–398. Beard, N. A., Laver, D. R., and Dulhunty, A. F. (2004). Calsequestrin and the calcium release channel of skeletal and cardiac muscle. Prog. Biophys. Mol. Biol. 85, 33–69. Be´douet, L., Schuller, M. J., Marin, F., Milet, C., Lopez, E., and Giraud, M. (2001). Soluble proteins of the nacre of the giant oyster Pinctada maxima and of the abalone Haliotis tuberculata: Extraction and partial analysis of nacre proteins. Comp. Biochem. Physiol. B 128, 389–400. Be´douet, L., Rusconi, F., Rousseau, M., Duplas, D., Marie, A., Dubost, L., Le Ny, K., Berland, S., Pe´duzzi, J., and Lopez, E. (2006). Identification of low molecular weight molecules as new components of the nacre organic matrix. Comp. Biochem. Physiol. B 144, 532–543. Be´douet, L., Marie, A., Dubost, L., Pe´duzzi, J., Duplat, D., Berland, S., Puisse´gur, M., Boulzaguet, H., Rousseau, M., Milet, C., and Lopez, E. (2007). Proteomics analysis of the nacre soluble and insoluble proteins from the oyster pinctada margaritifera. Mar. Biotechnol. In press. Belcher, A. M., and Gooch, E. E. (2000). Protein components and inorganic structure in shell nacre. In ‘‘Biomineralization: From Biology to Biotechnology and Medical Application’’ (E. Ba¨uerlein, Ed.)., pp. 221–249. Wiley‐VCH, New York. Bengtson, S. (1992). Proterozoic and earliest Cambrian skeletal metazoans. In ‘‘The Proterozoic Earth: A Multidisciplinary Study’’ (J.W. Schopf and C. Klein, Eds.), pp. 397–411. Cambridge University Press, Cambridge. Bevelander, G., and Nakahara, H. (1969). An electron microscope study of the formation of the nacreous layer in the shell of certain bivalve molluscs. Calcif. Tissue Res. 3, 84–92. Blank, S., Arnoldi, M., Khoshnavaz, S., Treccani, L., Kuntz, M., Mann, K., Grathwohl, G., and Fritz, M. (2003). The nacre protein perlucin nucleates growth of calcium carbonate crystals. J. Microsc. 212, 280–291. Boggild, O. (1930). The Shell Structure of the Mollusks. D. Kgl. Danske Videnskab. Selsk. Skrifter, Naturvidensk. Og Mathem. Afdel. 9, 231–326.

6. Molluscan Shell Proteins

265

Bonar, D. B. (1976). Molluscan metamorphosis: A study in tissue transformation. Amer. Zool. 16, 573–591. Bubel, A., Moore, M. N., and Lowe, D. (1977). Cellular responses to the shell damage in Mytilus edulis. J. Exp. Mar. Biol. Ecol. 30, 1–27. Cariolou, M. A., and Morse, D. E. (1988). Purification and characterization of calcium‐binding conchiolin shell peptides from the mollusc, Haliotis rufescens, as a function of development. J. Comp. Physiol. B 157, 717–729. Carter, J. G. (1980). Guide to bivalve shell microstructures. In ‘‘Skeletal Growth of Aquatic Organisms’’ (D.C. Rhoads and R.A. Lutz, Eds.), pp. 69–113. Plenum Press, New York. Carter, J. G. (1990). ‘‘Skeletal Biomineralization: Patterns, Processes, and Evolutionary Trends.’’ Van Nostrand Reinhold, New York. Carter, J. G., and Clark, G. R., II (1985). Classification and phylogenetic significance of mollusk shell microstructures. In ‘‘Mollusk, Note for a Short Course’’ (T. W. Broadhead, Ed.)., pp. 50–71. Studies in Geology 13, Department of Geological Sciences, University of Tennessee Press, Tennessee. Checa, A. G., and Rodriguez‐Navarro, A. B. (2001). Geometrical and crystallographic constraints determine the self‐organization of shell microstructures in Unionidae (Bivalvia: Mollusca). Proc. R. Soc. Lond. B 268, 771–778. Checa, A. G., and Rodriguez‐Navarro, A. B. (2005). Self‐organisation of nacre in the shells of Pterioida (Bivalvia: Mollusca). Biomaterials 26, 1071–1079. Checa, A. G., Rodriguez‐Navarro, A. B., and Esteban‐Delgado, F. J. (2005). The nature and formation of calcitic columnar prismatic shell layers in pteriomorphian bivalves. Biomaterials 26, 6404–6414. Checa, A. G., Okamoto, T., and Ramirez, J. (2006). Organization pattern of nacre in Pteriidae (Bivalvia: Mollusca) explained by crystal competition. Proc. R. Soc. B 273, 1329–1337. Chen, L., Xie, L., Dai, Y., Xiong, X., Fan, W., and Zhang, R. (2004). Cloning and characterization of an mRNA encoding a novel G protein ‐subunit abundant in the mantle and gill of pearl oyster Pinctada fucata. Comp. Biochem. Physiol. B 139, 669–679. Coimbra, J., Machado, J., Fernandes, P. L., Ferreira, H. G., and Ferreira, K. G. (1988). Electrophysiology of the mantle of Anodonta cygnea. J. Exp. Biol. 140, 65–88. Conway Morris, S. (1998). ‘‘The Crucible of Creation: The Burgess Shale and the Rise of Animals.’’ Oxford University Press, Oxford Conway Morris, S. (2001). Significance of early shells. In ‘‘Palaeobiology II’’ (D. E. G. Briggs and P.R. Crowther, Eds.), pp. 31–40. Blackwell Science, Oxford. Crenshaw, M. A., and Ristedt, H. (1975). Histochemical and structural study of nautiloid septal nacre. Biomineralization 8, 1–15. Crenshaw, M. A., and Ristedt, H. (1976). The histochemical localization of reactive groups in the septal nacre from Nautilus pompilius L. In ‘‘The Mechanisms of Mineralization in the Invertebrates and Plants’’ (N. Watabe and K. M. Wilbur, Eds.), pp. 335–367. University of South Carolina Press, Columbia. Cuif, J.‐P., Dauphin, Y., Denis, A., Gaspard, D., and Keller, J. P. (1983). Etude des caracte´ristiques de la phase mine´rale dans les structures prismatiques du test de quelques mollusques. Bull. Mus. Natl. Hist. Nat. Paris 4e se´r 5, 679–717. Cuif, J.‐P., Gautret, P., and Marin, F. (1991). Correlation between the size of crystals and the molecular weight of organic fractions in the soluble matrices of mollusc, coral and sponge carbonate skeletons. In ‘‘Mechanisms and Phylogeny of Mineralization in Biological Systems’’ (S. Suga and H. Nakahara, Eds.), pp. 391–395. Springer‐Verlag, Tokyo. Dai, Y., Xie, L., Xiong, X., Chen, L., Fan, W., and Zhang, R. (2005). Cloning and characterization of a homologous Ca2þ/Calmodulin‐dependent protein kinase PSKH1 from pearl oyster Pinctada fucata. Tshinghua Sci. Technol. 10, 504–511.

266

Marin et al.

Dauphin, Y. (2002). Comparison of the soluble matrices of the calcitic prismatic layer of Pinna nobilis (Mollusca, Bivalvia, Pteromorphia). Comp. Biochem. Physiol. B 132, 577–590. Dauphin, Y. (2003). Soluble organic matrices of the calcitic prismatic shell layers of two petriomorphid bivalves. J. Biol. Chem. 278, 15168–15177. Dictus, W. J. A. G., and Damen, P. (1997). Cell‐lineage and clonal‐contribution map of the trochophore larva of Patella vulgata. Mech. Dev. 62, 213–226. Donachy, J. E., Drake, B., and Sikes, C. S. (1992). Sequence and atomic‐force microscopy analysis of a matrix protein from the shell of the oyster Crassostrea virginica. Mar. Biol. 114, 423–428. Duplat, D., Puisse´gur, M., Be´douet, L., Rousseau, M., Boulzaguet, H., Milet, C., Sellos, D., Van Wormhoudt, A., and Lopez, E. (2006). Identification of calconectin, a calcium binding protein specifically expressed by the mantle of Pinctada margaritifera. FEBS Lett. 580, 2435–2441. Eickbush, T. (1999). Exon shuZing in retrospect. Science 283, 1465–1467. Endo, K., Sarashina, I., and Asami, T. (2003). Lymnea stagnalis as a model for studies of calcium carbonate biomineralization. In ‘‘Biomineralization: Formation, Diversity, Evolution and Application’’ (I. Kobayashi and H. Osawa, Eds.), pp. 172–175. Tokai University Press, Kanagawa. Engel, J. (1991). Common structural motifs in proteins of the extracellular matrix. Curr. Opin. Cell Biol. 31, 779–785. Engel, J. (1996). Domain organizations of modular extracellular matrix proteins and their evolution. Matrix Biol. 15, 295–299. Erben, H. K. (1972). Uber die Bildung und das Wachstum von Perlmutt. Biomineralization 4, 15–46. Falini, G., Albeck, S., Weiner, S., and Addadi, L. (1996). Control of aragonite or calcite polymorphism by mollusk shell macromolecules. Science 271, 67–69. Fedonkin, M. A., and Waggoner, B. M. (1997). The Late Precambrian fossil Kimberella is a mollusc‐like bilaterian organism. Nature 388, 868–871. Feng, W., and Sun, W. (2003). Phosphate replicated and replaced microstructure of molluscan shells from the earliest Cambrian of China. Acta Palaeontol. Pol. 48, 21–30. Fournie´, J., and Che´tail, M. (1982). Accumulation calcique au niveau cellulaire chez les mollusques. Malacologia 22, 265–284. Fu, G., Valiyaveetil, S., Wopenka, B., and Morse, D. E. (2005). CaCO3 biomineralization: An acidic 8‐kDa proteins isolated from aragonitic abalone shell nacre can specifically modify calcite crystal morphology. Biomacromolecules 6, 1289–1298. Gibert, J. M. (2002). The evolution of engrailed genes after duplication and speciation events. Dev. Genes Evol. 212, 307–318. Gilbert, W. (1978). Why genes in pieces? Nature 271, 501. Glinski, Z., and Jarosz, J. (1997). Molluscan immune defenses. Arch. Immunol. Ther. Exp. 45, 149–155. Gotliv, B.‐A., Addadi, L., and Weiner, S. (2003). Mollusk shell acidic proteins: In search for individual functions. Chembiochemistry 4, 522–529. Gotliv, B.‐A., Kessler, N., Sumerel, J. L., Morse, D. E., Tuross, N., Addadi, L., and Weiner, S. (2005). Asprich: A novel aspartic acid‐rich protein family from the prismatic shell matrix of the bivalve Atrina rigida. Chembiochemistry 6, 304–314. Gould, S. J., and Vrba, E. S. (1982). Exaptation: A missing term in the science of form. Paleobiology 8, 4–15. Gre´goire, C. (1972). Structure of the molluscan shell. In ‘‘Chemical Zoology, Vol. VII: Mollusca’’ (M. Florkin and B. T. Scheer, Eds.), pp. 45–102. Academic Press, New York. Grigor’ev, D. P. (1965). "Ontogeny of Minerals." Israel Program for Scientific Translation, Jerusalem.

6. Molluscan Shell Proteins

267

Halanych, K. M., Bacheller, J. D., Aguinaldo, A. M., Liva, S. M., Hillis, D. M., and Lake, J. A. (1995). Evidence from 18S ribosomal DNA that the lophophorates are protostome animals. Science 26, 1641–1643. Halloran, B. A., and Donachy, J. E. (1995). Characterization of organic matrix macromolecules from the shell of the Antarctic scallop, Adamussium colbecki. Comp. Biochem. Physiol. B 111, 221–231. Hare, P. E. (1963). Amino acids in the proteins from aragonite and calcite in the shells of Mytilus californianus. Science 139, 216–217. Hasegawa, Y., and Uchiyama, K. (2005). cDNA cloning of shell matrix proteins from scallop shell. Fish. Sci. 71, 1174–1178. Hasse, B., Ehrenberg, H., Marxen, J. C., Becker, W., and Epple, M. (2000). Calcium carbonate modifications in the mineralised shell of the freshwater snail Biomphalaria glabrata. Chem. Eur. J. 6, 3679–3685. Hattan, S. J., Laue, T. M., and Chasteen, N. D. (2001). Purification and characterization of a novel calcium‐binding protein from the extrapallial fluid of the mollusc, Mytilus edulis. J. Biol. Chem. 276, 4461–4468. Heinemann, F., Treccani, L., and Fritz, M. (2006). Abalone nacre insoluble matrix induces growth of flat and oriented aragonite crystals. Biochem. Biophys. Res. Commun. 344, 45–49. Hidalgo, A. (1996). The roles of engrailed. Trends Genet. 12, 1–4. Hinman, V. F., O’Brien, E. K., Richards, G. S., and Degnan, B. M. (2003). Expression of anterior Hox genes during larval development of the gastropod Haliotis asinina. Evol. Dev. 5, 508–521. Hofmann, S. L., Goldstein, J. L., Orth, K., Moomaw, C. R., Slaughter, C. A., and Brown, M. S. (1989). Molecular cloning of a histidine‐rich Ca2þ‐binding protein of sarcoplasmikc reticulum that contains highly conserved elements. J. Biol. Chem. 264, 18083–18090. Istin, M. (1970). Roˆle du manteau dans le me´tabolisme du calcium chez les lamellibranches. J. Biol. Chem. 264, 18083–18090. Istin, M., and Girard, J. P. (1970). Carbonic anhydrase and mobilisation of calcium reserves in the mantle of lamellibranches. Calcif. Tissue Res. 5, 247–260. Jablonski, B., and Lutz, R. A. (1980). Molluscan larval shell morphology, ecological and paleontological applications. In ‘‘Skeletal Growth of Aquatic Organisms’’ (D. C. Rhoads and R.A. Lutz, Eds.), pp. 323–377. Plenum Press, New York. Jackson, D. J., McDougall, C., Green, K., Simpson, F., Wo¨rheide, G., and Degnan, B. M. (2006). A rapidly evolving secretome builds and patterns a sea shell. BMC Biol. 4, 40–49. Jacobs, D. K., Wray, C. G., Wedeen, C. J., Kostriken, R., DeSalle, R., Staton, J. L., Gates, R. D., and Lindberg, D. R. (2000). Molluscan engrailed expression, serial organization, and shell evolution. Evol. Dev. 2, 340–347. Jolly, C., Berland, S., Milet, C., Borzeix, S., Lopez, E., and Doumenc, D. (2004). Zonal localization of shell matrix proteins in mantle of Haliotis tuberculata (Mollusca, Gastropoda). Mar. Biotechnol. 6, 541–551. Kawasaki, K., and Weiss, K. M. (2006). Evolutionary genetics of vertebrate tissue mineralization: The origin and evolution of the secretory calcium‐binding phosphoprotein family. J. Exp. Zool. 306B, 295–316. Kawasaki, K., Suzuki, T., and Weiss, K. M. (2004). Genetic basis for the evolution of vertebrate mineralized tissue. Proc. Natl. Acad. Sic. USA 101, 11356–11361. Keith, J., Stockwell, S., Ball, D., Remillard, K., Kaplan, D., Thannhauser, T., and Sherwood, R. (1993). Comparative analysis of macromolecules in mollusc shells. Comp. Biochem. Physiol. B 105, 487–496. Killian, C. E., and Wilt, F. H. (1996). Characterization of the proteins comprising the integral matrix of Strongylocentrotus purpuratus. J. Biol. Chem. 271, 9150–9159.

268

Marin et al.

Kim, I. W., Morse, D. E., and Evans, J. S. (2004). Molecular characterization of the 30‐AA N‐terminal mineral interaction domain of the biomineralization protein AP7. Langmuir 20, 11664–11673. Klerkx, A. H. E., de Boer, E., and van Loon, A. E. (2001). Spatio‐temporal expression of a gene encoding a putative RNA‐binding protein during the early development of the mollusc Patella vulgate. Dev. Genes Evol. 211, 423–427. Kniprath, E. (1977). Zur Ontogenese des Schalenfeldes von Lymnaea stagnalis. Roux’s Arch. Dev. Biol. 181, 11–30. Kniprath, E. (1980). Larval development of the shell and the shell gland in Mytilus (Bivalvia). Roux’s Arch. Dev. Biol. 188, 201–204. Kniprath, E. (1981). Ontogeny of the molluscan shell field. Zoologica Scripta 10, 61–79. Knoll, A. H., and Carroll, S. B. (1999). Early animal evolution: Emerging views from comparative biology and geology. Science 284, 2129–2136. Kobayashi, I. (1964). Introduction to the shell structure of bivalved molluscs. Earth Sci. 73, 1–12. Kobayashi, I. (1980). Various patterns of biomineralization and its phylogenetic significance in bivalve molluscs. In ‘‘The Mechanisms of Biomineralization in Animals and Plants’’ (M. Omori and N. Watabe, Eds.), pp. 145–155. Tokai University Press, Tokyo. Kobayashi, I. (1991). Evolutionary trends of shell microstructure in bivalve molluscs. In ‘‘Mechanisms and Phylogeny of Mineralization in Biological Systems’’ (S. Suga and H. Nakahara, Eds.), pp. 415–419. Springer‐Verlag, Tokyo. Kolkman, J. A., and Stemmer, W. P. (2001). Directed evolution of proteins by exon shuZing. Nat. Biotechnol. 19, 423–428. Kono, M., Hayashi, N., and Samata, T. (2000). Molecular mechanism of the nacreous layer formation in Pinctada maxima. Biochem. Biophys. Res. Commun. 269, 213–218. Kouchinsky, A. (2000). Shell microstructures in early Cambrian molluscs. Acta Palaeontol. Pol. 45, 119–150. Krampitz, G., Engels, J., and Cazaux, C. (1976). Biochemical studies on water‐soluble proteins and related components of gastropod shells. In ‘‘The Mechanisms of Mineralization in the Invertebrates and Plants’’ (N. Watabe and K. M. Wilbur, Eds.), pp. 155–173. University of South Carolina Press, Columbia. Kretsinger, R. H. (1976). Calcium‐binding proteins. Annu. Rev. Biochem. 45, 239–266. LaBarbera, M. (1974). Calcification of the first larval shell of Tridacna squamosa (Tridacnidae: Bivalvia). Mar. Biol. 25, 233–238. Lambert, G., and Lambert, C. C. (1996). Antibodies to echinoid larval spicule proteins cross react with the spicular complex in the ascidian Herdmania momus. Bull. Inst. Oce´anogr. 14(4), 253–261. Lecointre, G., and Le Guyader, H. (2001). "Classification Phyloge´ne´tique du Vivant." 2e`me e´d. Belin, Paris. Lee, D. D. (1990). The structure and mechanism of growth of calcium carbonate minerals in early stages of shells of the oyster Crassostrea virginica. J. Crystal Growth 102, 262–268. Lelong, C., Mathieu, M., and Favrel, P. (2000). Structure and expression of mGDF, a new member of the transforming growth factor‐
superfamily in the bivalve mollusc Crassostrea gigas. Eur. J. Biochem. 267, 3986–3993. Lelong, C., Mathieu, M., and Favrel, P. (2001). Identification of new bone morphogenetic protein‐related members in invertebrates. Biochimie 83, 423–426. Levi‐Kalisman, Y., Falini, G., Addadi, L., and Weiner, S. (2001). Structure of the nacreous organic matrix of a bivalve mollusk shell examined in the hydrated state using cryo‐TEM. J. Struct. Biol. 135, 8–17.

6. Molluscan Shell Proteins

269

Li, S., Xie, L., Zhang, C., Zhang, Y., Gu, M., and Zhang, R. (2004). Cloning and expression of a pivotal calcium metabolism regulator: Calmodulin involved in shell formation from pearl oyster (Pinctada fucata). Comp. Biochem. Physiol. B 138, 235–243. Li, S., Xie, L., Ma, Z., and Zhang, R. (2005). cDNA cloning and characterization of a novel calmodulin‐like protein from pearl oyster Pinctada fucata. FEBS J. 272, 4899–4910. Liao, H., Brandsten, C., Lundmark, C., Wurtz, T., and Li, J. (1997). Response of bone to titania‐ hydroxyapatite composite and nacreous implants: A preliminary comparison by in situ hybridization. J. Mater. Sci. Mater. Med. 8, 823–827. Liao, H., Mutvei, H., Sjo¨stro¨m, M., Hammarstro¨m, L., and Li, J. (2000). Tissue response to natural aragonite (Margaritifera shell) implants in vivo. Biomaterials 21, 457–468. Liao, H., Mutvei, H., Hammarstro¨m, L., Wurtz, T., and Li, J. (2002). Tissue response to nacreous implants in rat femur: An in situ hybridization and histochemical study. Biomaterials 23, 2693–2701. Lin, A., and Meyers, M. A. (2005). Growth and structure in abalone shell. Mater. Sci. Eng. 390, 27–41. Lindskog, S. (1997). Structure and mechanism of carbonic anhydrase. Pharmacol. Ther. 74, 1–20. Liu, H., Liu, S., Ge, Y., Wang, X., Xie, L., Zhang, R., and Wang, Z. (2007). Identification and characterization of a biomineralization related gene PFMG1 highly expressed in the mantle of Pinctada fucata. Biochemistry 46, 844–851. Livingston, B. T., Killian, C. E., Wilt, F., Cameron, A., Landrum, M. J., Ermolaeva, O., Sapojnikov, V., Maglott, D. R., Buchanan, A. M., and Ettensohn, C. A. (2006). A genome‐ wide analysis of biomineralization‐related proteins in the sea urchin Strongylocentrotus purpuratus. Dev. Biol. 300, 335–348. Lopez, E., Vidal, B., Berland, S., Camprasse, S., Camprasse, G., and Silve, C. (1992). Demonstration of the capacity of nacre to induce bone formation by human osteoblasts maintained in vitro. Tissue Cell 24, 667–679. Lowenstam, H. A., and Margulis, L. (1980). Evolutionary prerequisites for early Phanerozoic calcareous skeletons. Biosystems 12, 27–41. Lowenstam, H. A., and Weiner, S. (1989). "On Biomineralization." Oxford University Press, New‐York. Mann, K., Weiss, I. M., Andre´, S., Gabius, H. J., and Fritz, M. (2000). The amino acid sequence of the abalone (Haliotis laevigata) nacre protein perlucin. Eur. J. Biochem. 267, 5257–5264. Mann, K., Siedler, F., Treccani, L., Heinemann, F., and Fritz, M. (2007). Perlinhibin, a cysteine‐, histidine‐, and arginine‐rich miniprotein from abalone (Haliotis laevigata) nacre, inhibits in vitro calcium carbonate crystallization. Biophys. J. 93, 1246–1254. Mann, S. (1983). Mineralization in biological systems. Struct. Bond. 54, 125–174. Mann, S. (2001). "Biomineralization. Principles and Concepts in Bioinorganic Materials Chemistry." Oxford University Press, Oxford. Mao Che, L., Golubic, S., Le Campion‐Alsumard, T., and Payri, C. (2001). Developmental aspects of biomineralization in the polynesian pearl oyster Pinctada margaritifera var. cumingii. Oceanol. Acta 24, S37–S49. Marie, B., Luquet, G., Pais De Barros, J. P., Guichard, N., Morel, S., Alcaraz, G., Bollache, L., and Marin, F. (2007). The shell matrix of the unionid freshwater mussel Unio pictorum (Paleoheterodonta, Unionoida): Involvement of acidic polysaccharides from glycoproteins in nacre mineralization. FEBS J. 274, 2933–2945. Marin, F. (2003). Molluscan shell matrix characterization by preparative SDS‐PAGE. Scientific World J. 3, 342–347. Marin, F., De Groot, K., and Westbroek, P. (2003a). Screening molluscan cDNA expression libraries with anti‐shell matrix antibodies. Protein Expr. Purif. 30, 246–252.

270

Marin et al.

Marin, F., Layrolle, P., De Groot, K., and Westbroek, P. (2003b). The Origin of metazoan skeleton. In ‘‘Biomineralization: Formation, Diversity, Evolution and Application’’ (I. Kobayashi and H. Osawa, Eds.), pp. 50–53. Tokai University Press, Kanagawa. Marin, F., and Luquet, G. (2004). Molluscan shell proteins. C. R. Palevol. 3, 469–492. Marin, F., and Luquet, G. (2007). Unusually acidic proteins. In ‘‘Handbook of Biomineralization’’ (E. Ba¨uerlein, Ed.)., Vol. 1, pp. 273–290. Wiley‐VCH Verlag, Weinheim. Marin, F., Smith, M., Isa, Y., Muyzer, G., and Westbroek, P. (1996). Skeletal matrices, muci, and the origin of invertebrate calcification. Proc. Natl. Acad. Sci. USA 93, 1554–1559. Marin, F., Gillibert, M., Westbroek, P., Muyzer, G., and Dauphin, Y. (1999). Evolution: Disjunct degeneration of immunological determinants. Geol. Mijnb. 78, 135–139. Marin, F., Corstjens, P., De Gaulejac, B., De Vrind‐De Jong, E., and Westbroek, P. (2000). Mucins and molluscan calcification: Molecular characterization of mucoperlin, a novel mucin‐like protein from the nacreous shell layer of the fan mussel Pinna nobilis (Bivalvia, Pteriomorphia). J. Biol. Chem. 275, 20667–20675. Marin, F., Pereira, L., and Westbroek, P. (2001). Large‐scale fractionation of molluscan shell matrix. Protein Expr. Purif. 23, 175–179. Marin, F., Amons, R., Guichard, N., Stigter, M., Hecker, A., Luquet, G., Layrolle, P., Alcaraz, G., Riondet, C., and Westbroek, P. (2005). Caspartin and calprismin, two proteins of the shell calcitic prisms of the Mediterranean fan mussel Pinna nobilis. J. Biol. Chem. 280, 33895–33908. Marin, F., Pokroy, B., Luquet, G., Layrolle, P., and De Groot, K. (2007). Protein mapping of calcium carbonate biominerals by immunogold. Biomaterials 28, 2368–2377. Martoja, M. (1995). "Mollusques." Institut Oce´anographique, Paris. Marxen, J. C., and Becker, W. (1997). The organic shell matrix of the freswater snail Biomphalaria glabrata. Comp. Biochem. Physiol. B 118, 23–33. Marxen, J. C., Becker, W., Finke, D., Hasse, B., and Epple, M. (2003a). Early mineralization in Biomphalaria glabrata: Microscopic and structural results. J. Moll. Stud. 69, 113–121. Marxen, J. C., Nimtz, M., Becker, W., and Mann, K. (2003b). The major soluble 19.6 kDa protein of the organic shell matrix of the freswater snail Biomphalaria glabrata is an N‐glycosylated dermatopontin. Biochim. Biophys. Acta 1650, 92–98. Marxen, J. C., Eckhard Witten, P., Finke, D., Reelsen, O., Rezgaoui, M., and Becker, W. (2003c). A light‐ and electron‐microscopic study of enzymes in the embryonic shell‐forming tissue of the freshwater snail, Biomphalaria glabrata. Invert. Biol. 122, 313–325. Matsushiro, A., and Miyashita, T. (2004). Evolution of hard‐tissue mineralization: Comparison of the inner skeletal system and the outer shell system. J. Bone Miner. Metab. 22, 163–169. Matsushiro, A., Miyashita, T., Miyamoto, H., Morimoto, K., Tonomura, B., and Sato, K. (2003a). The presence of protein complexes formed between pearlin and pearl keratin is prerequisite for aragonite crystallization. In ‘‘Biomineralization: Formation, Diversity, Evolution and Application’’ (I. Kobayashi and H. Osawa, Eds.), pp. 137–140. Tokai University Press, Kanagawa. Matsushiro, A., Miyashita, T., Miyamoto, H., Morimoto, K., Tonomura, B., Tanaka, A., and Sato, K. (2003b). Presence of protein complex is prerequisite for aragonite crystallization in the nacreous layer. Mar. Biotechnol. 5, 37–44. Maurer, P., Hohenester, E., and Engel, J. (1996). Extracellular calcium‐binding proteins. Curr. Opin. Cell Biol. 8, 609–617. Medakovic, D. (2000). Carbonic anhydrase activity and biomineralization process in embryos, larvae and adult blue mussels Mytilus edulis L. Helgol. Mar. Res. 54, 1–6. Medakovic, D., Popovic, S., Grzeta, B., Plazonic, M., and Hrs‐Brenko, M. (1997). X‐ray diVraction study of calcification processes in embryos and larvae of the brooding oyster Ostrea edulis. Mar. Biol. 129, 615–623.

6. Molluscan Shell Proteins

271

Michenfelder, M., Fu, G., Lawrence, C., Weaver, J. C., Wustman, B. A., Taranto, L., Evans, J. S., and Morse, D. E. (2003). Characterization of two molluscan crystal‐modulating biomineralization proteins and identification of putative mineral binding domains. Biopolymers 70, 522–533; erratum in Biopolymers (2004) 73, 291. Michenfelder, M., Fu, G., Lawrence, C., Weaver, J. C., Wustman, B. A., Taranto, L., Evans, J. S., and Morse, D. E. (2004). Erratum: Characterization of two molluscan crystal‐ modulating biomineralization proteins and identification of pcetative mineral binding domains. Biopolymers 73, 291. Misogianes, M. J., and Chasteen, N. D. (1979). A chemical and spectral characterization of the extrapallial fluid of Mytilus edulis. Anal. Biochem. 100, 324–334. Miyamoto, H., Miyashita, T., Okushima, M., Nakano, S., Morita, T., and Matsushiro, A. (1996). A carbonic anhydrase from the nacreous layer in oyster pearls. Proc. Natl. Acad. Sci. USA 93, 9657–9660. Miyamoto, H., Yano, M., and Miyashita, T. (2003). Similarities in the structure of nacrein, the shell‐matrix protein, in a bivalve and a gastropod. J. Moll. Stud. 69, 87–89. Miyamoto, H., Miyoshi, F., and Kohno, J. (2005). The carbonic anhydrase domain protein nacrein is expressed in the epithelial cells of the mantle and acts as a negative regulator in calcification in the mollusc Pinctada fucata. Zool. Sci. 22, 311–315. Miyashita, T., Takagi, R., Okushima, M., Nakano, S., Miyamoto, H., Nishikawa, E., and Matsushiro, A. (2000). Complementary DNA cloning and characterization of pearlin, a new class of matrix protein in the nacreous layer of oyster pearls. Mar. Biotechnol. 2, 409–418. Miyashita, T., Takagi, R., Miyamoto, H., and Matsushiro, A. (2002). Identical carbonic anhydrase contributes to nacreous or prismatic layer formation in Pinctada fucata (Mollusca: Bivalvia). The Veliger 45, 250–255. Miyashita, T., Miyamoto, H., and Matsushiro, A. (2003). Expression of pearlin in Escherichia coli, a protein which participates in the nacreous formation of P. fucata pearls. In ‘‘Biomineralization: Formation, Diversity, Evolution and Application’’ (I. Kobayashi and H. Osawa, Eds.), pp. 176–177. Tokai University Press, Kanagawa. Montagnani, C., Fleury, E., Belliard, C., Piquemal, D., Flaven, E., and Cochennee‐Laureau, N. (2006). Characterizations of perline and calcine, two conserved shell matrix genes implicated in biomineralization processes in the pearl oyster Pinctada margaritifera. Submitted. Moor, B. (1983). Organogenesis. In ‘‘The Mollusca, Vol. 3: Development’’ (N.H. Verdonk, J.A. M. van den Biggelaar, and A.S. Tompa, Eds.), pp. 123–177. Academic Press, New York. Morgan, R. (2006). Engrailed: Complexity and economy of a multi‐functional transcription factor. FEBS Lett. 580, 2531–2533. Moshel, S. M., Levine, M., and Collier, J. R. (1998). Shell diVerentiation and engrailed expression in the Ilyanassa embryo. Dev. Genes Evol. 208, 135–141. Mount, A. M., Wheeler, A. P., Paradkar, R. P., and Snider, D. (2004). Hemocyte‐mediated mineralization in the eastern oyster. Science 304, 297–300. Moura, G., Vilarinho, L., Carvalho Santos, A., and Machado, J. (2000). Organic compounds in the extrapalliula fluid and haemolymph of Anodonta cygnea (L.) with emphasis on the seasonal biomineralization process. Comp. Biochem. Physiol. B 125, 293–306. Moura, G., Machado, J., and Coimbra, J. (2003). Insights on nacre formation in the freshwater clam Anodonta cygnea (L.): An overview. In ‘‘Biomineralization: Formation, Diversity, Evolution and Application’’ (I. Kobayashi and H. Osawa, Eds.), pp. 129–132. Tokai University Press, Kanagawa. Mutvei, H. (1969). On the micro‐ and ultrastructure of the conchiolin in the nacreous layer of some recent and fossil molluscs. Stockholm Contr. Geol. 20, 1–17.

272

Marin et al.

Mutvei, H. (1979). On the internal structures of the nacreous tablets in molluscan shells. Scann. Electron. Microsc. 11, 457–462. Muyzer, G., Westbroek, P., de Vrind, J. P. M., Tanke, J., Vrijheid, T., De Jong, E. W., Brunning, J. W., and Wehmiller, J. F. (1984). Immunology and organic geochemistry. Org. Geochem. 6, 847–855. Nagai, K., Yano, M., Morimoto, K., and Miyamoto, H. (2007). Tyrosinase localization in mollusc shells. Comp. Biochem. Physiol. 146, 207–214. Nakahara, H. (1991). Nacre formation in bivalve and gastropod mollusks. In ‘‘Mechanisms and phylogeny of Mineralization in Biological Systems’’ (S. Suga and H. Nakahara, Eds.), pp. 343–350. Springer‐Verlag, Tokyo. Nassif, N., Pinna, N., Gehrke, N., Antonietti, M., Ja¨ger, C., and Co¨lfen, H. (2005). Amorphous layer around aragonite platelets in nacre. Proc. Natl. Acad. Sci. USA 102, 12653–12655. Nederbragt, A. J., van Loon, A. E., and Dictus, W. J. A. G. (2002). Expression of Patella vulgata orthologs of engrailed and dpp‐BMP2/4 in adjacent domains during molluscan shell development suggests a conserved compartment boundary mechanism. Dev. Biol. 246, 341–355. Nielsen, C. (2004). Trochophora larvae: Cell‐lineages, ciliary bands and body regions. 1 Annelida and Mollusca. J. Exp. Biol. 302, 35–68. Nudelman, F., Gotliv, B.‐A., Addadi, L., and Weiner, S. (2006). Mollusk shell formation: Mapping the distribution of organic matrix components underlying a single aragonitic tablet in nacre. J. Struct. Biol. 153, 176–187. Oaki, Y., and Imai, H. (2005). The hierarchical architecture of nacre and its mimetic material. Angew. Chem. Int. Ed. Engl. 44, 6571–6575. Oberling, J. J. (1964). Observations on some structural features of the pelecypod shell. Mittel. Naturforschr. Gesell. Bern 20, 1–63. Palmer, A. R. (1983). Relative cost of producing skeletal organic matrix versus calcification: Evidence from marine gastropods. Mar. Biol. 75, 287–292. Palmer, A. R. (1992). Calcification in marine molluscs: How costly is it? Proc. Natl. Acad. Sci. USA 89, 1379–1382. Patthy, L. (1996). Exon shuZing and other ways of module exchange. Matrix Biol. 15, 301–310. Patthy, L. (1999). Genome evolution and the evolution of exon‐shuZing — a review. Gene 238, 103–114. Patthy, L. (2003). Modular assembly of genes and the evolution of new functions. Genetica 118, 217–231. Pereira‐Mourie`s, L., Almeida, M. J., Ribeiro, C., Peduzzi, J., Barthe´lemy, M., Milet, C., and Lopez, E. (2002). Soluble silk‐like organic matrix in the nacreous layer of the bivalve Pinctada maxima. Eur. J. Biochem. 269, 4994–5003. Pokroy, B., Fitch, A. N., Marin, F., Kapon, N., Adir, N., and Zolotoyabko, E. (2006a). Anisotropic lattice distorsion in biogenic calcite induced by intra‐crystalline organic molecules. J. Struct. Biol. 155, 96–103. Pokroy, B., Zolotoyabko, E., and Adir, N. (2006b). Purification and functional analysis of a 40‐kDa protein extracted from the Strombus decorus persicus mollusc shell. Biomacromolecules 7, 550–556. Pozzan, T., Rizzuto, R., Volpe, P., and Meldolesi, J. (1994). Molecular and cellular physiology of intracellular calcium stores. Physiol. Rev. 74, 595–636. Rousseau, M., Pereira‐Mourie`s, L., Almeida, M. J., Milet, C., and Lopez, E. (2003). The water‐ soluble matrix fraction from the nacre of Pinctada maxima produces ealier mineralization of MC3T3‐E1 mouse pre‐osteoblasts. Comp. Biochem. Physiol. B 135, 1–7. Rousseau, M., Lopez, E., Stempfle´, P., Brendle´, M., Franke, L., Guette, A., Naslain, R., and Bourrat, X. (2005). Multiscale structure of sheet nacre. Biomaterials 26, 6254–6262.

6. Molluscan Shell Proteins

273

Runnegar, B. (1996). Early evolution of the Mollusca: The fossil record. In ‘‘Origin and Evolutionary Radiation of the Mollusca’’ (J. Taylor, Ed.)., pp. 77–87. Oxford University Press, Oxford. Rusenko, K. W., Donachy, J. E., and Wheeler, A. P. (1991). Purification and characterization of a shell matrix phosphoprotein from the american oyster. In ‘‘Surface Reactive Peptides and Polymers, Discovery and Commercialization’’ (C. S. Sikes and A. P. Wheeler, Eds.), pp. 107–124. ACS Symposium Series 444, Washington, DC. Saleuddin, A. S. M., and Petit, H. P. (1983). The mode of formation and the structure of the periostracum. In ‘‘The Mollusca, Vol. 4: Physiology’’ (A. S. M. Saleuddin and K. M. Wilbur, Eds.), pp. 199–233. Academic Press, New York. Samata, T. (1990). Ca‐binding glycoproteins in molluscan shells with diVerent types of ultrastructure. Veliger 33, 190–201. Samata, T., Hayashi, N., Kono, M., Hasegawa, K., Horita, C., and Akera, S. (1999). A new matrix protein family related to the nacreous layer formation of Pinctada fucata. FEBS Lett. 462, 225–229. Samata, T., Kono, M., Hayashi, N., Oyama, S., and Suganaga, M. (2003). Molecular mechanism of the nacreous layer formation. In ‘‘Biomineralization: Formation, Diversity, Evolution and Application’’ (I. Kobayashi and H. Osawa, Eds.), pp. 141–144. Tokai University Press, Kanagawa. Sarashina, I., and Endo, K. (1998). Primary structure of a soluble matrix protein of scallop shell: Implications for calcium carbonate biomineralization. Am. Mineral. 83, 1510–1515. Sarashina, I., and Endo, K. (2001). The complete primary structure of Molluscan Shell Protein 1 (MSP‐1), an acidic glycoprotein in the shell matrix of the scallop Patinopecten yessoensis. Mar. Biotechnol. 3, 362–369. Sarashina, I., Yamaguchi, H., Haga, T., Iijima, M., Chiba, S., and Endo, K. (2006). Molecular evolution and functionally important structures of molluscan dermatopontin: Implications for the origins of molluscan shell matrix proteins. J. Mol. Evol. 62, 307–318. Scha¨Ver, T. E., Ionescu‐Zanetti, C., Proksch, R., Fritz, M., Walters, D. A., Almqvist, N., Zaremba, C. M., Belcher, A. M., Smith, B. L., Stucky, G. D., Morse, D. E., and Hansma, P. K. (1997). Does abalone nacre form by heteroepitaxial nucleation or by growth through mineral bridges? Chem. Mater. 9, 1731–1740. Schilthuizen, M., and Davison, A. (2005). The convoluted evolution of snail chirality. Naturwissenschaften 92, 504–515. Seed, R. (1980). Shell growth and form in the Bivalvia. In ‘‘Skeletal Growth of Aquatic Organisms’’ (D. C. Rhoads and R. A. Lutz, Eds.), pp. 23–67. Plenum Press, New York. Serpentini, A., Ghayir, C., Poncet, J.‐M., Hebert, V., Gale´ra, P., Pujol, J.‐P., Boucaud‐ Camou, E., and Lebel, J.‐M. (2000). Collagen study and regulation of the de novo synthesis by IGF‐1 in hemocytes from the gastropod mollusc, Haliotis tuberculata. J. Exp. Zool. 287, 275–284. Shen, X., Belcher, A. M., Hansma, P. K., Stucky, G. D., and Morse, D. E. (1997). Molecular cloning and characterization of lustrin A, a matrix protein from shell and pearl nacre of Haliotis rufescens. J. Biol. Chem. 272, 32472–32481. Shimamoto, M. (1986). Shell microstructure of the Veneridae (Bivalvia) and its phylogenetic implications. Sci. Rep., 2nd ser. Tohoku Univ. 56, 1–39. Shimamoto, M. (1993). Shell microstructure and amino acid composition of organic matrix from venerid shell. In ‘‘Structure, Formation and Evolution of Fossil Hard Tissues’’ (I. Kobayashi, H. Mutvei, and A. Sahni, Eds.), pp. 89–97. Tokai University Press, Tokyo. Shubin, N. H., and Marshall, C. R. (2000). Fossils, genes, and the origin of novelty. Paleobiology Suppl. 26(4), 324–340. Simkiss, K. (1977). Biomineralization and detoxification. Calcif. Tiss. Res. 24, 199–200.

274

Marin et al.

Simkiss, K. (1991). Amorphous minerals and theories of biomineralization. In ‘‘Mechanisms and phylogeny of Mineralization in Biological Systems’’ (S. Suga and H. Nakahara, Eds.), pp. 375–382. Springer‐Verlag, Tokyo. Simkiss, K. (1993). Amorphous minerals in biology. Bull. Inst. Oceanogr. Monaco, n special 14, 49–54. Simkiss, K., and Wilbur, K. M. (1989). ‘‘Biomineralization, Cell biology and Mineral Deposition.’’ Academic Press Inc., New York. Skelton, P. W. (1985). Bivalvia–Rostroconchia. In ‘‘Atlas of Invertebrate Macrofossils’’ (J. W. Murray, Eds.)., pp. 81–101. Longman, The Palaeontological Association, London. Smith, B. L., Scha¨Ver, T. E., Viani, M., Thompson, J. B., Frederick, N. A., Kindt, J., Belcher, A., Stucky, G. D., Morse, D. E., and Hansma, P. K. (1999). Molecular mechanistic origin of the toughness of natural adhesives, fibres and composites. Nature 399, 761–763. Smith, K. S., Jakubzick, C., Whittam, T. S., and Ferry, J. G. (1999). Carbonic anhydrase is an ancient enzyme widespread in prokaryotes. Proc. Natl. Acad. Sci. USA 96, 15184–15189. Smith, K. S., and Ferry, J. G. (2000). Prokaryotic carbonic anhydrases. FEMS Microbiol. Rev. 24, 335–366. Song, F., Soh, A. K., and Bai, Y. L. (2003). Structural and mechanical properties of the organic matrix layers of nacre. Biomaterials 24, 3623–3631. Sprott, J. C. (1993). ‘‘Strange Attactors: Creating Patterns in Chaos.’’ M&T Books, New York. Sud, D. (2002). Les cultures de cellules: Une approche expe´rimentale de l’e´tude des processus de biomine´ralisation chez les inverte´bre´s marins. Thesis, Muse´um d’Histoire Naturelle, Paris, p. 268. Sudo, S., Fujikawa, T., Nagakura, T., Ohkubo, T., Sakagushi, K., Tanaka, M., and Nakashima, K. (1997). Structures of mollusc shell framework proteins. Nature 387, 563–564. Suzuki, M., Murayama, E., Inoue, H., Ozaki, N., Tohse, H., Kogure, T., and Nagasawa, H. (2004). Characterization of Prismalin‐14, a novel matrix protein from the prismatic layer of the Japanese pearl oyster (Pinctada fucata). Biochem. J. 382, 205–213. Takeuchi, T., and Endo, K. (2005). Biphasic and dually coordinated expression of the genes encoding major shell matrix proteins in the pearl oyster Pinctada fucata. Mar. Biotech. 8, 52–61. Taylor, J. D. (1973). The structural evolution of the bivalve shell. Palaeontology 16, 519–534. Taylor, J. D., Kennedy, W. J., and Hall, A. (1969). The shell structure and mineralogy of the bivalvia. Introduction. Nuculacea‐Trigonacea. Bull. Br. Mus. (Nat. Hist.) Zool. Lond. 3(Suppl.), 1–125. Taylor, J. D., Kennedy, W. J., and Hall, A. (1973). The shell structure and mineralogy of the bivalvia. II. Lucinacea‐Clavagellacea. Conclusions. Bull. Br. Mus. (Nat. Hist.) Zool. Lond. 22, 253–294. Timmermans, L. P. M. (1969). Studies on shell formation in molluscs. Neth. J. Zool. 19, 417–523. Thomas, R. D. K., Shearman, R. M., and Stewart, G. W. (2000). Evolutionary exploitation of design option by the first animals with hard skeletons. Science 288, 1239–1242. Tong, H., Hu, J., Ma, W., Zhong, G., Yao, S., and Cao, N. (2002). In situ analysis of the organic framework in the prismatic layer of mollusc shell. Biomaterials 23, 2593–2598. Treccani, L., Mann, K., Heinemann, F., and Fritz, M. (2006). Perlwapin, an abalone nacre protein with three four‐disulfide core (whey acidic protein) domains, inhibits the growth of calcium carbonate crystals. Biophys. J. 91, 2601–2608. Tsukamoto, D., Sarashina, I., and Endo, K. (2004). Structure and expression of an unusually acidic matrix protein of pearl oyster shells. Biochem. Biophys. Res. Commun. 320, 1175–1180. Ubukata, T. (1994). Architectural constrainst on the morphogenesis of prismatic structure in Bivalvia. Palaeontology 37, 241–261.

6. Molluscan Shell Proteins

275

Ubukata, T. (1997). Microscopic growth of bivalve shells and its computer simulation. Veliger 40, 165–177. Uozumi, S., and Suzuki, S. (1981). The evolution of shell structure in the Bivalvia. In ‘‘Study of Molluscan Paleobiology,’’ (T. Habe, and M. Omori, Eds.), pp. 63–77. Prof. M. Omori Memorial Volume. van Rijk, A., and Bloemendal, H. (2003). Molecular mechanisms of exon shuZing: Illegitimate recombination. Genetica 118, 245–249. Veis, D. J., Albinger, T. M., Clohisy, J., Rahima, M., Sabsay, B., and Veis, A. (1986). Matrix proteins of the teeth of the sea urchin Lytechinus variegatus. J. Exp. Zool. 240, 35–46. Waite, H. (1983). Quinone‐tanned scleroproteins. In ‘‘The Mollusca, Vol. 4: Physiology’’ (A. S. M. Saleuddin, and K. M. Wilbur, Eds.), pp. 467–504. Academic Press, New York. Waite, J. H., Saleuddin, A. S. M., and Andersen, S. O. (1979). Periostracin ‐ A soluble precursor of sclerotized periostracum in Mytilus edulis L. J. Comp. Physiol. 130, 301–307. Wanninger, A., and Haszprunar, G. (2001). The expression of an engrailed protein during embryonic shell formation of the tusk‐shell, Antalis entalis (Mollusca, Scaphopoda). Evol. Dev. 3, 312–321. Watabe, N. (1983). Shell repair. In ‘‘The Mollusca, Vol. 4: Physiology’’ (A. S. M. Saleuddin and K. M. Wilbur, Eds.), pp. 289–316. Academic Press, New York. Weedon, M. J., and Taylor, P. D. (1995). Calcitic nacreous ultrastructures in bryozoans: Implications for comparative biomineralization of lophophorates and molluscs. Biol. Bull. 188, 281–292. Weiner, S. (1979). Aspartic acid‐rich proteins: Major components of the soluble organic matrix of mollusk shells. Calcif. Tissue Int. 29, 163–167. Weiner, S. (1983). Mollusk shell formation: Isolation of two organic matrix proteins associated with calcite deposition in the bivalve Mytilus californianus. Biochemistry 22, 4139–4145. Weiner, S., and Hood, L. (1975). Soluble proteins of the organic matrix of mollusc shells: A potential template for shell formation. Science 190, 987–989. Weiner, S., Levi‐Kalisman, Y., Raz, S., and Addadi, L. (2003). Biologically formed amorphous calcium carbonate. Connect. Tisue Res. 44, 214–218. Weiss, I. M., Kaufmann, S., Mann, K., and Fritz, M. (2000). Purification and characterization of perlucin and perlustrin, two new proteins from the shell of the mollusc Haliotis laevigata. Biochem. Biophys. Res. Commun. 267, 17–21. Weiss, I. M., Go¨hring, W., Fritz, M., and Mann, K. (2001). Perlustrin, a Haliotis laevigata (abalone) nacre protein, is homologous to the insulin‐like growth factor binding protein N‐terminal module of vertebrates. Biochem. Biophys. Res. Commun. 285, 244–249. Weiss, I. M., Tuross, N., Addadi, L., and Weiner, S. (2002). Mollusc larval shell formation: Amorphous calcium carbonate is a precursor phase for aragonite. J. Exp. Zool. 293, 478–491. Weiss, I. M., Schno¨nitzer, V., Eichner, N., and Sumper, M. (2006). The chitin synthase involved in marine bivalve mollusk shell formation contains a myosin domain. FEBS Lett. 580, 1846–1852. Westbroek, P., and Marin, F. (1998). A marriage of bone and nacre. Nature 392, 861–862. Wheeler, A. P., George, J. W., and Evans, C. A. (1981). Control of calcium carbonate nucleation and crystal growth by soluble matrix of oyster shell. Science 212, 1397–1398. Wheeler, A. P., Rusenko, K. W., Swift, D. M., and Sikes, C. S. (1988). Regulation of in vitro and in vivo CaCO3 crystallization by fractions of oyster shell organic matrix. Mar. Biol. 98, 71–80. Wilbur, K. M., and Saleuddin, A. S. M. (1983). Shell formation. In ‘‘The Mollusca, Vol. 4: Physiology’’ (A. S. M. Saleuddin and K. M. Wilbur, Eds.), pp. 235–287. Academic Press, New York. Wilt, F. H., Killian, C. E., and Livingston, B. T. (2003). Development of calcareous skeletal elements in invertebrates. DiVerentiation 71, 237–250.

276

Marin et al.

Wray, C. G., Jacobs, D. K., Kostriken, R., Vogler, A. P., Baker, R., and DeSalle, R. (1995). Homologues of the engrailed gene from five molluscan classes. FEBS Lett. 365, 71–74. Wustman, B. A., Morse, D. E., and Evans, J. S. (2002). Structural analyses of polyelectrolyte sequence domains within the adhesive elastomeric biomineralization protein lustrin A. Langmuir 18, 9901–9906. Wustman, B. A., Weaver, J. C., Morse, D. E., and Evans, J. S. (2003a). Structure‐function studies of the lustrin A polyelectrolyte domains, RKSY and D4. Connect. Tissue Res. 44, 10–15. Wustman, B. A., Weaver, J. C., Morse, D. E., and Evans, J. S. (2003b). Characterization of a Ca(II), mineral‐interactive polyelectrolyte sequence from the adhesive elastomeric biomineralization protein lustrin A. Langmuir 19, 9373–9381. Wustman, B. A., Morse, D. E., and Evans, J. S. (2004). Structural characterization of the N‐terminal mineral modification domains from the molluscan crystal‐modulating biomineralization proteins, AP7 and AP24. Biopolymers 74, 363–376. Yano, M., Nagai, K., Morimoto, K., and Miyamoto, H. (2006). Shematrin: A family of glycin‐ rich structural proteins in the shell of the pearl oyster. Comp. Biochem. Physiol. 144, 254–262. Yin, Y., Huang, J., Paine, M. L., Reinhold, V. N., and Chasteen, D. (2005). Structural characterization of the major extrapallial fluid protein of the mollusc Mytilus edulis: Implication for function. Biochemistry 44, 10720–10731. Yu, Z., Xie, L., Lee, S., and Zhang, R. (2006). A novel carbonic anhydrase from the mantle of the pearl oyster (Pinctada fucata). Comp. Biochem. Physiol. 143, 190–194. Zhang, B., Wustman, B. A., Morse, D., and Evans, J. S. (2002). Model peptide studies of sequence regions in the elastomeric biomineralization protein, lustrin A. I. The C‐domain consensus ‐PG‐, ‐NVNCT‐motif. Biopolymers 63, 358–369. Zhang, C., and Zhang, R. (2006). Matrix proteins in the outer shells of molluscs. Mar. Biotechnol. 8, 572–586. Zhang, C., Xie, L., Meng, Q., Jiang, T., Pu, R., Chen, L., and Zhang, R. (2003a). A novel matrix protein participating in the nacre framework formation of pearl oyster, Pinctada fucata. Comp. Biochem. Physiol. 133, 565–573. Zhang, C., Li, S., Huang, J., Ma, Z., and Zhang, R. (2006a). A novel matrix protein p10 from the nacre of pearl oyster (Pinctada fucata) and its eVect on both CaCO3 crystal formation and mineralogenic cells. Mar. Biotechnol. 8, 624–633. Zhang, C., Xie, L., Huang, J., Chen, L., and Zhang, R. (2006b). A novel putative tyrosinase involved in periostracum formation from the pearl oyster (Pinctada fucata). Biochem. Biophys. Res. Commun. 342, 632–639. Zhang, C., Xie, L., Huang, J., Liu, X., and Zhang, R. (2006c). A novel matrix protein family participating in the prismatic layer formation of pearl oyster, Pinctada fucata. Biochem. Biophys. Res. Commun. 344, 735–740. Zhang, Y., Meng, Q., Jiang, T., Wang, H., Xie, L., and Zhang, R. (2003b). A novel ferritin subunit involved in shell formation from the pearl oyster (Pinctada fucata). Comp. Biochem. Physiol. B 135, 43–54. Zhao, H., Samata, T., Takakura, D., Hashimoto, R., Miyazaki, Y., Nozawa, T., and Hikita, Y. (2003). Organic matrix proteins preserved in fossil molluscan shells. In ‘‘Biomineralization: Formation, Diversity, Evolution and Application’’ (I. Kobayashi and H. Osawa, Eds.), pp. 108–111. Tokai University Press, Kanagawa.

7

____________________________________________________________________________

Pathophysiology of the Blood–Brain Barrier: Animal Models and Methods Brian T. Hawkins* and Richard D. Egleton{ *Laboratory of Pharmacology and Chemistry, National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina 27709 { Department of Pharmacology, Physiology and Toxicology, Joan C. Edwards Medical School, Marshall University, Huntington, West Virginia 25755

I. The Blood–Brain Barrier A. Introduction B. Regulation of Paracellular Permeability C. Catalyzed Transport and Biotransformation D. Endocytotic Transport E. The Neurovascular Unit and the Limitations of In Vitro Models II. Animal‐Based Methods in BBB Pathophysiology A. General Considerations B. Brain Uptake Measurements C. In Vivo Imaging D. Genomic/Proteomic Approaches III. BBB Dysfunction as a Complication of Peripheral Disease A. The BBB in CNS Disease B. The BBB in Diabetes C. Inflammatory Pain and the BBB IV. The BBB in Disease Etiology V. Concluding Remarks Acknowledgments References

The specialized cerebral microvascular endothelium interacts with the cellular milieu of the brain and extracellular matrix to form a neurovascular unit, one aspect of which is a regulated interface between the blood and central nervous system (CNS). The concept of this blood–brain barrier (BBB) as a dynamically regulated system rather than a static barrier has wide‐ranging implications for pathophysiology of the CNS. While in vitro models of the BBB are useful for screening drugs targeted to the CNS and indispensable for studies of cerebral endothelial cell biology, the complex interactions of the neurovascular unit make animal‐based models and methods essential tools for understanding the pathophysiology of the BBB. BBB dysfunction is a complication of neurodegenerative disease and brain injury. Studies on animal models have shown that diseases of the periphery, Current Topics in Developmental Biology, Vol. 80 Copyright 2008, Elsevier Inc. All rights reserved.

277

0070-2153/08 $35.00 DOI: 10.1016/S0070-2153(07)80007-X

278

Hawkins and Egleton

such as diabetes and inflammatory pain, have deleterious eVects on the BBB which may contribute to neurological complications associated with these conditions. Furthermore, genetic and/or epigenetic abnormalities in constituents of the BBB may be significant contributing factors in disease etiology. Research that approaches the BBB as a dynamic system integrated with both the CNS and the periphery is therefore critical to understanding and treating diseases of the CNS. Herein, we review various methodological approaches used to study BBB function in the context of disease. These include measurement of transport between blood and brain, imaging‐based technologies, and genomic/proteomic approaches. ß 2008, Elsevier Inc.

I. The Blood–Brain Barrier A. Introduction Although capillary beds in the periphery permit relatively free exchange of solutes between blood and the surrounding tissue, the microvasculature of the vertebrate brain presents a dynamic and highly regulated interface between the blood and central nervous system (CNS). This blood–brain barrier (BBB) protects the brain from potentially neurotoxic substances, facilitates exchange of nutrients and waste products between the brain and blood, and maintains an optimal extracellular environment for neuronal function. The evolutionary necessity of selective barriers to neural tissue is perhaps best illustrated by the fact that CNS barriers are not limited to animals with closed circulatory systems (Carlson et al., 2000). Well‐characterized examples include the septate junctions between perineural glia in Drosophila that isolate neurons from the hemolymph (Carlson et al., 1997; Schwabe et al., 2005), a glial‐based barrier in the cockroach (Lane and Treherne, 1969, 1972), and a specialized metabolic barrier to nicotine in the tobacco hornworm Manduca sexta (Murray et al., 1994). Interestingly, the cuttlefish Sepia, which has a complex brain with an extensive microvasculature, also appears to have a glial‐based rather than endothelial‐based barrier (Abbott and Pichon, 1987). Whether such systems represent primitive barriers from which the endothelial‐based barriers were derived or are examples of convergent evolution, it is clear that a highly regulated extracellular environment is integral to the development and survival of complex neural systems (Abbott, 2005; Abbott et al., 1985). BBB research has focused primarily on the BBB as an obstacle to drug delivery into the CNS (Begley, 2004), and indeed, this represents the most significant problem of translational neuroscience (Pardridge, 2005a). However, the role of the BBB in disease is emerging as an important field of

7. Animal Methods in BBB Pathophysiology

279

inquiry in its own right, raising the possibility that the BBB itself may represent a primary or secondary therapeutic target in diseases aVecting the CNS. Furthermore, a greater understanding of the BBB as part of a neurovascular unit integrated with other components of the CNS (Lo et al., 2003; NINDS, 2002) points to the need for integrative, whole‐animal studies to address the involvement of the BBB in complex disease processes. In this chapter, we discuss the concept of the BBB, animal models and methods used in BBB pathophysiology, and putative roles for the BBB in diseases originating both in the CNS and in the periphery. Although the existence and physical substrate of the BBB were the subjects of much debate for nearly a century (Hawkins and Davis, 2005), the pioneering work of Reese, Karnovsky, and Brightman (Brightman and Reese, 1969; Reese and Karnovsky, 1967) established that in mammals, the primary barrier exists at the level of the cerebral capillary endothelium. The selective permeability of the BBB is maintained by several constituents of the endothelium (Fig. 1), including (1) epithelial‐like tight junctions (TJs) that limit paracellular diVusion of even small molecules; (2) carrier‐mediated transport proteins that regulate the passage of nutrients from blood to brain, maintain inorganic ion balance in brain interstitial fluid, and eZux xenobiotics and metabolites from brain to blood, often coupled with the activity of enzymes catalyzing biotransformation at the endothelium; and (3) receptor‐mediated and absorptive endocytotic mechanisms. Furthermore, the specialized functions of the endothelium are induced, maintained, and regulated via complex interactions with other cell types and the extracellular matrix of the brain, constituting a neurovascular unit.

B. Regulation of Paracellular Permeability The intercellular space of cerebral capillary endothelium is typically characterized by a continuous band of adhesive proteins mediating homophilic interactions between the cells. This junctional complex includes TJs, adherens junctions (AJs), and gap junctions. Of these, the TJs and AJs are thought to play the primary structural roles while gap junctions are limited to intracellular communication (Bazzoni and Dejana, 2004; Hawkins and Davis, 2005), though evidence suggests that functional gap junctions may be necessary for maintaining endothelial barrier function as well (Nagasawa et al., 2006). The TJ consists of the transmembrane proteins junctional adhesion molecule (JAM), occludin, and claudin, linked via accessory proteins including zonula occludens (ZO)‐1 and ‐2 to the actin cytoskeleton (Fig. 2). TJs are found in various tissues throughout the body where they serve as both impediments to paracellular diVusion and facilitators of cell polarity.

A

B

C

Diffusion Flux down gradient Energy independent Flux proportional to concentration Two types at BBB

Carrier mediated Involves interaction with carrier protein in membrane Saturable, substrate specificity, competition Can be assymetric Can be active against gradient primary (hydrolysis of ATP) or secondary (dependent on chemical gradient)

Endocytosis Invagination of plasma membrane to form internalized membrane vesicle Three types fluid phase (FPE), adsorptive mediated (AME), receptor mediated (RME). FPE, AME, and some forms of RME undergo constituitive endocytosis, i.e endocytosis in the absence or presence of ligand. In some cases RME is ligand stimulated

Paracellular small water soluble solutes TJ limits, sucrose

Transcellular Lipophilic solutes CTAP (Abbruscato et al., 1997), Diazepam (Arendt et al., 1987)

Facilitated GLUT-1 (Pardridge et al., 1990) L1, y+ amino acid transporters (Hawkins et al., 2006)

Energy dependent primary PgP MRPs BCRP (Loscher and Potschka 2005) Na/K ATPase (Johshita et al., 1994)

FPE HRP (Defazio et al., 1997) Lucifer yellow (Guillot et al., 1990)

CTAP

LY GLUT-1

AME cationized proteins/ peptides (Tamai et al., 1997) Glycoproteins (Banks et al., 1997b) Amphipathic (Dhanasakeran and Plot, 2005) viruses (Banks et al ., 2001)

Digoxin

Glucose Sucrose

Energy dependent secondary EAATs ASC (Hawkins et al., 2006) OATs (Kusuhara and Sugiyama, 2005)

gp-120

RME constitutive transferrin (Roberts et al., 1993) LRP (Deane et al., 2004a) RAGE (Deane et al., 2004b) ligand stimulated insulin (Frank et al., 1986)

Transferrin

Pgp

ATP

ADP

Glucose L-Glu Lysosome

K GLUT-1

EAAT 3Na, H

Sucrose

CTAP

Glucose

L-Glu

Figure 1 Examples of transporters and transport mechanisms at the BBB that have been targeted to improve drug distribution to the brain. (A) Both paracellular and transcellular diVusion have been targeted to improve brain delivery of drugs. Typically at the BBB there is limited paracellular transport due to the presence of the TJs. Paracellular diVusion can however be increased using hyperosmotic shock to open TJs

281

7. Animal Methods in BBB Pathophysiology Apical/luminal plasma membrane

Blood

Occludin

ZO-2 7H6

ZO-2

ZO-1

ZO-1

Actin

ZO-2 AF-6

ZO-1

Claudin 3/5 ZO-1

Actin Cingulin

Jam Brain

Figure 2 Basic molecular organization of BBB TJs. Reproduced with permission from Hawkins and Davis (2005).

TJs are dynamic structures, with decreased TJ protein expression or alteration of subcellular localization associated with alterations in paracellular permeability (Hawkins and Davis, 2005; Huber et al., 2001a). Multiple factors regulate the expression, traYcking, and protein–protein interactions of TJ proteins over the course of development and in disease (Bazzoni and Dejana, 2004; Harhaj and Antonetti, 2004). Moreover, constituents of the (Brown et al., 2004), and has been used clinically for delivery of chemotherapeutics in brain cancer treatment (Neuwelt et al., 1984). Transcellullar transport has been targeted by drug modifications to increase lipophilicity including methylation (Witt et al., 2000), and reducing H‐bonding potential (Chikhale et al., 1994). (B) Carrier mediated transport plays an important role in targeting drugs to the BBB, and also in reducing drug entry into the brain. Facilitative transporters such as the large neutral amino acid transporter have been used to transport a number of drugs including L‐dopa (Wade and Katzman, 1975). Primary energy dependent mechanisms such as the ABC transporters have a large impact on drug delivery at the BBB, with substrates as diverse as digoxin (Mayer et al., 1996) and cyclosporine A (Sakata et al., 1994). Secondary independent transporters include members of the SLC family, and also have a significant role in limiting brain drug entry of organic anions and cations (Kusuhara and Sugiyama, 2005). (C) Endocytosis is an important target for large molecule and gene delivery at the BBB. FPE is not generally a target for delivery however a number of strategies target both AME and RME. Perhaps the best characterized is the use of antibodies to the transferrin receptor, which have been used to deliver genetic material (Pardridge, 2005b), and growth factors (Song et al., 2002) to the brain.

282

Hawkins and Egleton

TJs themselves may play a role in signal transduction, communicating the state of cell–cell contacts to the nucleus and participating in the regulation of growth, diVerentiation, and gene expression (Matter et al., 2005). AJs are also likely regulators of paracellular permeability. Formed at the BBB by the transmembrane protein vascular endothelial (VE)‐cadherin linked via catenins and other accessory proteins to actin, AJs are thought to mediate cell–cell adhesion necessary for the formation of TJs and establishment of high‐resistance barrier tissues (Vincent et al., 2004), and are particularly sensitive to Ca2þ (Brown and Davis, 2002). Whether AJs are critical determinants of BBB permeability in mature tissue is not completely clear. Loss of AJ proteins often parallels that of TJ proteins, under conditions of increased permeability (Song et al., 2007), which makes determining the relative contributions of TJ and AJ to barrier function diYcult. However, it appears that TJ proteins are the main determinants of low permeability and high electrical resistance in vitro (Romero et al., 2003). Further evidence for the preeminence of the TJ at the BBB includes the finding that an antibody raised against VE‐cadherin increases paracellular permeability in cardiac muscle and lung tissue, but not in brain (Corada et al., 1999).

C. Catalyzed Transport and Biotransformation As the TJ and endothelial membranes constitute a physical barrier to most polar molecules, it is essential that water‐soluble nutrients needed for brain function are able to pass through the endothelium via facilitated diVusion or active transport. Among the best characterized transport systems at the BBB are those that mediate brain uptake of amino acids (Hawkins et al., 2006; Nalecz et al., 2004) and glucose (Qutub and Hunt, 2005). Isoforms of the monocarboxylate transporter (MCT) (Enerson and Drewes, 2003) and organic anion‐transporting polypeptide (OATP) (Kim, 2003) are also expressed at the BBB. Organic cations (e.g., thiamine and choline) are transported at the BBB by at least two transporter systems (Lockman et al., 2004), though the proteins responsible have not yet been identified (Tsuji, 2005). Beyond these critical roles in brain homeostasis and potentially brain disease (Section IV), such transporters may also be promising targets for enhancing delivery of small‐molecule (Tsuji, 2005) and peptide (Egleton and Davis, 2005) therapeutics into the brain. Another critical homeostatic function of the BBB is the maintenance of interstitial fluid ion concentrations within the necessary bounds for neuronal function. Although the interstitial fluid is continuous with the cerebrospinal fluid produced by the choroid plexus, it is estimated that interstitial fluid is secreted largely (if not primarily) by the capillary endothelium (Abbott, 2004). Ion transport proteins expressed at the brain capillary endothelium

7. Animal Methods in BBB Pathophysiology þ

þ

283

include the Na , K ATPase (Eisenberg and Suddith, 1979; Zlokovic et al., 1993); Naþ, Kþ, 2Cl cotransporter (O’Donnell et al., 1995; Sun et al., 1995); Naþ–Hþ exchanger (Hom et al., 2007); and possibly aquaporins 1 and/or 4 (Amiry‐Moghaddam et al., 2004; Dolman et al., 2005; Kobayashi et al., 2001). However, the relative contribution of each to vectoral ion transport as well as their distributions to the luminal and abluminal membranes are not completely resolved. What is known is that under normal conditions, the ionic composition of brain interstitial fluid is highly resistant to fluctuations in the ionic composition of the blood (Stummer et al., 1995). However, dysregulation of ion transporter expression or activity at the BBB by chronic insults, such as hypertension or nicotine, may predispose individuals to developing postischemic brain edema due to a diminished capacity to respond to sudden changes in interstitial fluid ion content (Abbruscato et al., 2004; Hom et al., 2007; Wang et al., 1994). The capillary endothelium also expresses members of the ATP‐binding cassette (ABC) family that mediate the active eZux of xenobiotics and metabolites from brain to blood. The best studied of these is P‐glycoprotein (Pgp, also known as MDR‐1). Its high level of expression at the luminal membrane of the brain capillary endothelium, along with its broad substrate specificity for therapeutics, indicate that Pgp is likely the most significant mediator of xenobiotic eZux at the BBB (Bauer et al., 2005). Other ABC transporters expressed at the BBB and involved in brain eZux include several of the multidrug resistance proteins (MRPs) (Dallas et al., 2006) and breast cancer resistance protein (BCRP) (Eisenblatter et al., 2003). Coordination of ABC transporters with members of the solute carrier (SLC) family, including organic anion transporters and OATPs, enables the BBB to eZux a wide variety of xenobiotics and metabolites from brain to blood (Kusuhara and Sugiyama, 2005). Transporters for the monoamine neurotransmitters, norepinepherine, and serotonin have been described in mouse cerebral microvessels, where it is speculated that they may play a role in clearance of these transmitters from synapses (Wakayama et al., 2002). This would suggest a general role for the BBB in facilitating eYcient transmission by neurotransmitters and neuromodulators, a function likely augmented by proteolytic enzymes expressed in the endothelium that degrade neuropeptides (Egleton and Davis, 2005). The BBB also contains a number of the phase I and phase II enzyme systems responsible for the biotransformation of numerous drugs. These include several cytochrome P‐450s (CYP) such as CYP4X1 (Al‐Anizy et al., 2006), CYP1A1, CYP1B1 (Dey et al., 1999; Filbrandt et al., 2004; Granberg et al., 2003), CYP3A1, CYP3A2 (Mei et al., 2004; Rosenbrock et al., 2001), CYP2E1 (Haorah et al., 2005), and CYP 2B (Chat et al., 1998). There is also a significant expression and activity of several conjugation enzymes such as glyoxylase 1 and 2 (Wu et al., 2002), glutathione‐S‐transferase (Lowndes

284

Hawkins and Egleton

et al., 1994), and uridine diphosphate glucuronosyltransferase (UGT) (Chat et al., 1998). Many of these enzymes produce substrates for eZux transporters such as Pgp, and the MRPs, and may share common regulation with the eZux transporters. For example, phenobarbital is a commonly prescribed anticonvulsant used in epilepsy. Apart from its actions as an antiepileptic, phenobarbital is also a ligand for the constitutive androstane receptor (CAR); in liver, activation of CAR leads to an induction of several phase I enzymes such as CYP2B1 and CYP2B2 (Schuetz et al., 1986), phase II enzymes such as UGT (Sugatani et al., 2001), and eZux transporters such as MRP2 (Patel et al., 2003; Xiong et al., 2002). As these enzymes and transporters are also localized at the BBB, it is possible that similar regulation occurs there.

D. Endocytotic Transport Despite the low levels of vesicles reported in cerebral endothelia, ranging from 3 to 6 vesicles per mm3 compared to 82–93 per mm3 in skeletal muscle endothelia, a range of endocytotic processes do occur at the BBB (Smith and Gumbleton, 2006). These include fluid phase endocytosis, a process in which a solute is present in close proximity to the membrane. During endocytosis, some of the extracellular fluid is caught in the lumen of the vesicle that forms at the surface and then enters the cell. Solutes which undergo fluid phase endocytosis at the BBB include horseradish peroxidase (Broadwell, 1989; Defazio et al., 1997) and Lucifer yellow (Guillot et al., 1990). Adsorptive endocytosis, a process by which a solute interacts with the cell membrane or binds nonspecifically to cell surface proteins, also occurs at the BBB. Solutes that undergo this form of endocytosis include wheat germ agglutinin (Villegas and Broadwell, 1993), basic and cationized peptides (Drin et al., 2003; Tamai et al., 1997), glycoproteins and glycopeptides (Banks et al., 1997b; Dhanasekaran and Polt, 2005), and viruses (Banks et al., 2001). This method of transport across the BBB has been exploited to deliver several classes of drugs to the CNS including opioid peptides (Deguchi et al., 2004) and doxorubicin (Rousselle et al., 2001). There are a number of receptors at the BBB that can carry ligands into and across the cell via receptor‐mediated endocytosis. Examples of this would include the transferrin receptor (JeVeries et al., 1984; Roberts et al., 1993), the low‐density lipoprotein receptor (Benchenane et al., 2005; Deane et al., 2004a; Meresse et al., 1989), and the insulin receptor (Frank et al., 1986; Pardridge et al., 1985; Yu et al., 2006). Perhaps the best characterized receptor system at the BBB is the transferrin receptor. This receptor is highly expressed at the

7. Animal Methods in BBB Pathophysiology

285

BBB and undergoes constitutive endocytosis (Smith and Gumbleton, 2006). Antibodies against this receptor can be used as vehicles to move a number of substances across the BBB. This includes pegylated immunoliposomes carrying antisense, siRNA, and genes to treat brain cancer and Parkinson’s disease (Pardridge, 2004, 2005b; Shi et al., 2001; Zhang et al., 2003). This technology has also been used to alleviate damage in stroke models by delivering growth factors (Song et al., 2002; Zhang and Pardridge, 2001). E. The Neurovascular Unit and the Limitations of In Vitro Models Although the endothelium represents the site of the physical, biochemical, and metabolic barriers between the blood and brain, its specialized phenotype is dependent on interactions between the endothelium and the surrounding tissue, as demonstrated by grafting and coculture experiments (Hawkins and Davis, 2005). Various aspects of BBB function are influenced by contact with astrocytes (Abbott et al., 2006; Janzer and RaV, 1987), pericytes (Lai and Kuo, 2005), and neurons (Kalinin et al., 2006), as well as with the extracellular matrix (del Zoppo et al., 2006). Thus, while the ‘‘phenomenon’’ of the BBB is mediated primarily by the capillary endothelium, the specialized functions thereof are induced and modulated by interactions with other constituents of the neurovascular unit (Lo et al., 2003). In practical terms, this points to a severe limitation of cell culture‐based models, particularly in addressing questions of how complex disease processes aVect and are mediated by the BBB. While cell culture models continue to serve as indispensable tools in BBB research, translational studies in BBB pathophysiology necessarily require animal models and methods.

II. Animal‐Based Methods in BBB Pathophysiology A. General Considerations A persistent problem in the field of BBB pathophysiology is the diYculty of comparing and synthesizing the findings of diVerent laboratories using diVerent techniques to measure BBB function. Moreover, the definition of ‘‘BBB function’’ is itself largely measurement dependent. For the purposes of the present discussion, we are focusing specifically on the overall integrity of the BBB to molecules not transported into the brain via specific mechanisms, that is, only on the paracellular permeability of the capillary endothelium. It is important to note, however, that many of these experimental approaches can be used to investigate the function of specific transport mechanisms as well.

286

Hawkins and Egleton

An investigator wishing to assess the state of the BBB in a disease model has an array of animal‐based techniques from which to chose, including quantitative, semiquantitative, and qualitative measurements of BBB integrity. These range from simple intravenous injections followed by macroscopic evaluation of brain uptake to in vivo imaging and include both terminal and nonterminal measurements. The most appropriate approach in a given situation must be determined by weighing considerations of sensitivity, technical diYculty, cost, and the number of animals required to generate meaningful data. Generally, measurements of BBB permeability involve the introduction of a tracer molecule into the peripheral circulation and measurement of its penetration into the brain. An ideal permeability probe would be: (1) too polar to pass through the endothelial cell membrane, (2) not actively transported by the BBB in either direction, (3) easily measured by virtue of its radioactive, fluorescent, spectroscopic, or paramagnetic properties, and (4) stable within the time frame of the experiment. Table I summarizes the most widely used permeability probes for in vivo BBB studies. These molecules are also often referred to as ‘‘vascular space markers,’’ as they are used to measure the vascular fraction of brain drug accumulation in the intact BBB. In studying the eVect of a particular pathology or insult on the overall integrity of the BBB, the choice of an appropriate probe is critical. For example, extremely large molecules [such as albumin or high‐molecular‐weight (MW) dextrans] only cross the BBB under conditions of extreme compromise. Moreover, a nonlinear size dependence of measured vascular spaces on MW of the marker has been observed in the brain, implying that drug distribution studies employing very large vascular space markers may overestimate drug accumulation in the brain due to underestimation of the vascular space (Smith et al., 1988). Accordingly, any change in BBB permeability may also be underestimated by high‐MW markers. Increased permeation of smaller, nontransported molecules can signal less‐severe, but clinically relevant, increases in BBB permeability (Starr et al., 2003). Of particular usefulness are studies where several probes of diVerent sizes are measured to make inferences about the magnitude of BBB opening (Huber et al., 2006b; Juhler et al., 1984). Many of the techniques discussed below have been reviewed extensively with regard to drug transport studies (Bickel, 2005); as such, we focus our discussion on the applications of these techniques to BBB pathophysiology. In addition to these methods, BBB function has also been measured with brain microdialysis and quantitative autoradiography. Both of these techniques are better suited to assessment of BBB characteristics in particular regions of the brain. In the case of microdialysis, measurements are limited to the immediate vicinity of the probe, whereas with quantitative autoradiography, a global analysis of tracer permeation is time consuming and labor intensive, relative to the techniques discussed below.

7. Animal Methods in BBB Pathophysiology Table I

287

Commonly Used BBB Permeability Markers

Probe

MW (Da)

Detection

Sample References

Mannitol Sucrose

182 342

Radioactive (14C or 3H) Radioactive (14C or 3H)

Fluorescein (Na salt) Gd‐DTPA

376

Fluorescence

570

Magnetic resonance

Preston et al., 1995 Hawkins et al., 2007a; Huber et al., 2006b Farkas et al., 1998; Natah et al., 2005 Barzo et al., 1996; Ewing et al., 2003; Knight et al., 2005 Huber et al., 2006b; Juhler et al., 1984 Carvey et al., 2005; Reese and Karnovsky, 1967 Huber et al., 2006b

Inulin Horseradish peroxidase Albumin

Fluorescent‐ labeled dextrans

5000

Radioactive (14C or 3H)

44,000

Electron microscopy

65,000

Absorption spectroscopy or fluorescence (Evans Blue) Radioactive (125I) Fluorescence

4000–250,000

Gaber et al., 2004; Mooradian et al., 2005; Olsson et al., 1975

B. Brain Uptake Measurements 1. Intravenous Injection The most direct way to assess BBB function is via introduction of a tracer into the peripheral vasculature via intravenous injection, followed by qualitative or quantitative assessment of tracer permeation into the brain. This is the least invasive approach and allows for tracer circulation under physiological conditions. Due to multiple passes through the cerebral microvasculature, it can be among the most sensitive ways to measure BBB function (Bickel, 2005), particularly when radioactive tracer molecules are used. Much of the earliest (and serendipitous) characterization of the BBB was performed via peripheral injection of ‘‘intravital dyes’’ bound to serum proteins (for review, see Bechmann et al., 2007). This approach has persisted for more than a century. For scenarios in which a large disruption of the BBB is known to occur or is anticipated (e.g., ischemic stroke or acute traumatic brain injury), it is generally suYcient to address questions of gross BBB damage. However, chronic and progressive diseases may have insidious eVects on the BBB that are not so easily detected (Huber et al., 2006b). Furthermore, great care must be taken in the interpretation of results based on protein or protein‐bound probes. For example, it has been shown that inadequate flushing of the brain can lead to

288

Hawkins and Egleton

false positives of BBB disruption, in which Evans blue‐labeled albumin trapped in the vasculature is misinterpreted as staining the brain parenchyma and having crossed the BBB (Ovadia et al., 2001). In this case, washout must be verified quantitatively by capillary depletion (Triguero et al., 1990) or qualitatively by microscopy. A variation on this approach that is appealing in its simplicity is postmortem immunostaining of the brain for native immunoglobulins (Hoshino et al., 1996). While the presence of immunoglobulins in the brain may indeed indicate a disruption of the BBB, the barrier to antibodies is not absolute under normal conditions (Reid et al., 1993; Yoshimi et al., 2002). Furthermore, it has been argued that invasion of the perivascular space by peripheral immune cells via the postcapillary venules is mechanistically distinct from direct damage to the BBB per se (Bechmann et al., 2007); thus, the presence of immunoglobulins in the brain parenchyma may not necessarily reflect an opening of the BBB to other molecules. Similarly, imunostaining for native albumin accumulation in postmortem or in resected brain tissue would indicate that, at some stage, there had been an opening of the BBB, allowing plasma proteins to enter the brain (van Vliet et al., 2007). However, while postmortem staining for native proteins indicates that the BBB has been compromised, it does not supply any temporal or quantitative information regarding the opening. Quantification of brain uptake can be performed on most of the tracers listed in Table I, in which the measured concentration of tracer in the brain is compared to that in the blood. The most significant caveat of this approach is that BBB permeability is only one factor that aVects brain distribution of compounds; when a tracer is allowed to circulate throughout the body, changes in plasma protein binding, metabolism, absorption, or excretion all have the potential to alter apparent brain distribution. Caution is therefore warranted in interpreting results of studies purporting to assess BBB function if no accounting for these potentially confounding variables is made. Ruling out changes in these factors can necessitate additional and time‐consuming analysis of tracer distribution to other tissues. An elegant solution to this problem is the graphical analysis method of Patlak (Patlak et al., 1983), which uses multiple time‐point sampling of the blood over the entire exposure time to correct for changes in plasma tracer concentration. The ratio of measured tracer concentration in brain (Cbr) to that in plasma (Cp) at each time T is plotted versus the ratio of the integral of the plasma concentration curve to the plasma concentration at that time. In the case of unidirectional uptake of the tracer from blood to brain, a straight line is produced and fit by linear regression to: Cbr ðTÞ ¼ Kin Cp ðTÞ

RT 0

Cp ðtÞdt þ Vi Cp ðTÞ

ð1Þ

7. Animal Methods in BBB Pathophysiology

289

The slope of the line is the unidirectional (influx) rate constant Kin, which is proportional to the BBB permeability of the tracer. Vi, or initial volume of distribution, represents the sum of the vascular volume of the brain (typically measured at 15 ml/g), the concentration of tracer within the endothelium (typically negligible in the case of vascular space markers), and the distribution of the tracer into any brain compartment in rapid equilibration with plasma (again assumed to be negligible in the case of most vascular space markers).

2. Brain Uptake Index One way to avoid the confounding eVects of peripheral metabolism on brain distribution measurements is to use a single‐pass approach: that is, introducing a tracer directly into the cerebral blood supply and not allowing it to recirculate. The brain uptake index (BUI) method, developed by Oldendorf for the rat (Oldendorf, 1970), accomplishes this via bolus injection into the common carotid artery and rapid (