Current Topics in Developmental Biology, Volume 65 (Current Topics in Developmental Biology)

Contents of Previous Volumes 261 7. Adult Bone Marrow-Derived Hemangioblasts, Endothelial Cell Progenitors, and EPCs ...

Author: Gerald P. Schatten

40 downloads 951 Views 4MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Contents of Previous Volumes

261

7. Adult Bone Marrow-Derived Hemangioblasts, Endothelial Cell Progenitors, and EPCs Gina C. Schatteman

8. Synthetic Extracellular Matrices for Tissue Engineering and Regeneration Eduardo A. Silva and David J. Mooney

9. Integrins and Angiogenesis D. G. Stupack and D. A. Cheresh

Volume 65 1. Tales of Cannibalism, Suicide, and Murder: Programmed Cell Death in C. elegans Jason M. Kinchen and Michael O. Hengartner

2. From Guts to Brains: Using Zebrafish Genetics to Understand the Innards of Organogenesis Carsten Stuckenholz, Paul E. Ulanch, and Nathan Bahary

3. Synaptic Vesicle Docking: A Putative Role for the Munc18/Sec1 Protein Family Robby M. Weimer and Janet E. Richmond

4. ATP-Dependent Chromatin Remodeling Corey L. Smith and Craig L. Peterson

5. Self-Destruct Programs in the Processes of Developing Neurons David Shepherd and V. Hugh Perry

6. Multiple Roles of Vascular Endothelial Growth Factor (VEGF) in Skeletal Development, Growth, and Repair Elazar Zelzer and Bjorn R. Olsen

7. G-Protein Coupled Receptors and Calcium Signaling in Development Geoffrey E. Woodard and Juan A. Rosado

8. Differential Functions of 14-3-3 Isoforms in Vertebrate Development Anthony J. Muslin and Jeffrey M. C. Lau

9. Zebrafish Notochordal Basement Membrane: Signaling and Structure Annabelle Scott and Derek L. Stemple

10. Sonic Hedgehog Signaling and the Developing Tooth Martyn T. Cobourne and Paul T. Sharpe

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.

Nathan Bahary (47), Department of Molecular Genetics and Biochemistry and Department of Medicine, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania 15261 Martyn T. Cobourne (255), Department of Craniofacial Development and Orthodontics, GKT Dental Institute, King’s College London, London SE1 9RT, United Kingdom Michael O. Hengartner (1), Institute of Molecular Biology, University of Zurich, 8057 Zurich, Switzerland Jason M. Kinchen (1), Department of Molecular Genetics and Microbiology, Stony Brook University, Stony Brook, New York 11743 and Institute of Molecular Biology, University of Zurich, 8057 Zurich, Switzerland Jeffrey M. C. Lau (211), Center for Cardiovascular Research, Department of Medicine and Department of Cell Biology and Physiology, Washington University School of Medicine, St. Louis, Missouri 63110 Anthony J. Muslin (211), Center for Cardiovascular Research, Department of Medicine and Department of Cell Biology and Physiology, Washington University School of Medicine, St. Louis, Missouri 63110 Bjorn R. Olsen (169), Harvard Medical School, Department of Cell Biology, Boston, Massachusetts 02115 V. Hugh Perry (149), School of Biological Sciences, University of Southampton, Southampton SO16 7PX, United Kingdom Craig L. Peterson (115), Program in Molecular Medicine, University of Massachusetts Medical School, Worcester, Massachusetts 01605 Janet E. Richmond (83), Department of Biological Sciences University of Illinois-Chicago, Chicago, Illinois 60607 Juan A. Rosado (189), Department of Physiology, University of Extremadura, Ca´ceres, Spain 10071 Annabelle Scott (229), Wellcome Trust Sanger Institute, Wellcome Trust Genome Campus, Hinxton, Cambridge CB10 1SA, United Kingdom

xi

xii

Contributors

Paul T. Sharpe (255), Department of Craniofacial Development, GKT Dental Institute, King’s College London, London SE1 9RT, United Kingdom David Shepherd (149), School of Biological Sciences, University of Southampton, Southampton SO16 7PX, United Kingdom Corey L. Smith (115), Program in Molecular Medicine, University of Massachusetts Medical School, Worcester, Massachusetts 01605 Derek L. Stemple (229), Wellcome Trust Sanger Institute, Wellcome Trust Genome Campus, Hinxton, Cambridge CB10 1SA, United Kingdom Carsten Stuckenholz (47), Department of Molecular Genetics and Biochemistry, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania 15261 Paul E. Ulanch (47), Department of Molecular Genetics and Biochemistry, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania 15261 Robby M. Weimer (83), Howard Hughes Medical Institute, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York 11724 Geoffrey E. Woodard (189), Metabolic Diseases Branch, National Institute of Diabetes, Digestive and Kidney Diseases, Bethesda, Maryland 20892 Elazar Zelzer (169), Department of Molecular Genetics, Weizmann Institute of Science, Rehovot 76100, Israel

Preface This volume of Current Topics in Developmental Biology explores an incredibly broad array of fundamental processes, from apoptosis to DNA replication and repair to protein signaling. While deftly elucidating these complex processes, the authors of the outstanding chapters in this volume also present the intriguing unanswered questions that remain, compelling challenges for the new generation of developmental biologists now in training. Tales of Cannibalism, Suicide, and Murder: Programmed Cell Death in C. elegans by Jason Kinchen and Michael Hengartner of Stony Brook University reviews this model of human apoptosis, which is becoming particularly robust as genetic screens reveal the genes responsible for apoptotic dynamics. Modeling shows these dynamics to be surprisingly nonlinear: a gene might participate in several actions at diVerent ends of the cell-death and removal spectrum, perhaps illuminating some of the complicated dynamics of human autoimmune diseases, the aberrant proteins of Alzheimer’s, and other neurodegenerative processes. From Guts to Brains: Using Zebrafish Genetics to Understand the Innards of Organogenesis by Carsten Stuckenholz, Paul Ulanch, and Nathan Bahary of the University of Pittsburgh comprehensively reviews current techniques in and discoveries from genetic screens in this relatively new model of development. Understanding the development of organs helps to reveal the mechanisms of human disease that resemble growth and diVerentiation processes gone awry, like cancer. Notably, zebrafish are now being used for studies of the central nervous system, as their development resembles that of higher vertebrates. Synaptic Vesicle Docking: A Putative Role for the Munc18/Sec1 Protein Family by Robby Weimer of the Cold Spring Harbor Lab and Janet Richmond of the University of Chicago explores the current understanding of how UNC-18 proteins promote vesicle/plasma membrane interaction, crucial for enabling release of neurotransmitter into the synaptic cleft, and proposes several pathways for this mechanism. The authors conclude by challenging our current morphological characterization that defines vesicle docking by proximity to the plasma membrane for a molecular one based on exo- or endocytosis. In ATP-Dependent Chromatin Remodeling by Corey Smith and Craig Peterson of the University of Massachusetts, the authors provide a comprehensive primer on the enzyme-driven chromatin dynamics that permit access to DNA for replication and repair and so are the subject of a great deal of biomedical research. New graduate students take note: the xiii

xiv

Preface

authors call for a better understanding of the complexity of these enzymes in vivo, which they assure will reveal the fascinating diVerences among them. Self-Destruct Programs in the Processes of Developing Neurons by David Shepherd and V. Hugh Perry of the University of Southampton explores axon and synapse degeneration after cell body removal following programmed cell death and asks: How are these withdrawn during development of the central nervous system? Can axons and dendrites be removed independent of the cell body? The authors review the available models and conclude – contrary to the classical notion of neuronal death at the cell soma – that indeed the loss of axons, dendrites, and synapses can be initiated by events at their locale. Multiple Roles of Vascular Endothelial Growth Factor in Skeletal Development, Growth, and Repair by Elazar Zelzer of the Weizmann Institute and Bjorn Olsen of Harvard reviews the role of VEGF in bone formation, including the exciting notion that VEGF participates in bone morphogenetic processes quite diVerent from its known role in vascular development. Again, students designing research projects take note: the authors pose four important questions in need of answers, and the translational implications are clear. G-Protein Coupled Receptors and Calcium Signaling in Development by GeoVrey Woodard of the National Institutes of Health and Juan Rosado of the University of Extremadura is a sweeping review of a family of proteins implicated in a diverse array of developmental processes. As this chapter shows, calcium’s ubiquity in the modulation of development calls for continued and careful attention to its storage and activation. The protein story continues in DiVerential Functions of 14-3-3 Isoforms in Vertebrate Development by Anthony Muslin and JeVrey Lau of Washington University. While the processes overseen by this family of proteins are understood in lower organisms, their role in the development of vertebrates is still coming to light, as described in this chapter. Interestingly, these proteins have diVerent roles in diVerent organisms, an intriguing problem in development that we look forward to seeing solved. In Zebrafish Notochordal Basement Membrane: Signaling and Structure by Annabelle Scott and Derek Stemple of the Wellcome Trust Sanger Institute, the authors explore how the basement membrane, known to be crucial to notochord development, influences nearby tissues. Again, as its sequence comes to light, the zebrafish will prove to be a valuable model for investigating these and other exciting issues in vertebrate development. Finally, Sonic Hedgehog Signaling and the Developing Tooth by Martyn Cobourne and Paul Sharpe of King’s College London review the role of the Shh gene in odontogenesis. As shown in this chapter, the tooth germ is a valuable model for the way that proteins of the Hedgehog family influence populations of developing cells. But once again, there are intriguing

Preface

xv

questions remaining to be answered, in particular about tooth and dental axis development. This volume has 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. The members of the Editorial Board are thanked for their suggestions of topics and authors. I also thank Leah KauVman for her fabulous editorial insight and Anna Vacca for her exemplary administrative support. Finally, we are grateful to everyone at the Pittsburgh Development Center of Magee-Women’s Research Institute here at the University of Pittsburgh School of Medicine for providing intellectual and infrastructural support for Current Topics in Developmental Biology. Jerry Schatten Pittsburgh Development Center, Pennsylvania

1

____________________________________________________________________________

Tales of Cannibalism, Suicide, and Murder: Programmed Cell Death in C. elegans Jason M. Kinchen*,{ and Michael O. Hengartner{ *Department of Molecular Genetics and Microbiology Stony Brook University Stony Brook, New York 11743 { Institute of Molecular Biology University of Zurich 8057 Zurich, Switzerland

I. Developmental Cell Death Occurs in C. elegans According to a Strict Cell Lineage II. Four Genes Identify a Conserved Genetic Pathway for the Induction of Cell Death A. The CED-3 Caspase B. The Apaf-1 Orthologue CED-4 C. The Bcl-2 Family Member CED-9 D. The EGL-I BH3-only Domain Protein E. Summary III. IV. V. VI.

CED-8 Plays an Unknown Role in Programmed Cell Death CED-3 Independent Cell Death—Role of icd-1 Similarities and Distinctions in the Apoptotic Machinery of C. elegans and Mammals Cell-Type Specific Regulation of Developmental Cell Death A. Regulation of Cell Death in the HSN (Hermaphrodite Specific Neurons): Involvement of the Sex Determination Machinery B. Regulation of Cell Death in the NSM Sisters—Involvement of ces-1 and ces-2

VII. C. elegans as a Model for Neuronal Degeneration VIII. Programmed Cell Death During DiVerentiation of the Hermaphrodite Germ Line A. Structure of the Hermaphrodite Germ Line B. Regulation of Germ Cell Apoptosis IX. Germ Cell Apoptosis in Response to Genotoxic Stress A. Genotoxic Stress Induces Cell Cycle Arrest and Apoptosis in the Hermaphrodite Germ Line B. Double-Strand Breaks that Arise from Failed Meiosis Activate an Apoptotic Checkpoint, but Do Not Induce Cell Cycle Arrest X. Cannibalism and More—Phagocytosis and Degradation of the Apoptotic Cell XI. Recognition of the Apoptotic Cell—The Varied Eat-Me Signal A. Phosphatidylserine (PS) Exposure and Corpse Recognition B. Annexin I Involvement in Engulfment C. Opsonizing Agents D. Nonengulfment Signals XII. Seven Genes Function in the Engulfing Cell to Mediate Phagocytosis of Apoptotic Cell Corpses in C. elegans A. CED-1, CED-6, and CED-7 B. CED-2, CED-5, CED-10, and CED-12 Current Topics in Developmental Biology, Vol. 65 Copyright 2005, Elsevier Inc. All rights reserved. 0070-2153/05 $35.00

1

2

Kinchen and Hengartner

XIII. Moving the Plasma Membrane—The Small GTPases Rac and RhoG A. CED-10, MIG-2, and RAC-2—Redundant Rac Proteins? B. MIG-2 and RhoG—Multiple Mechanisms of Engaging the CED-2, -5, -10, -12 Pathway? C. Signaling Downstream of Rac—Novel EVectors? XIV. ced-2, -5, -10, and -12 and Distal Tip Cell (DTC) Migration XV. Suicide vs Murder—Suggestions of Phagocyte-Mediated Cell Killing A. The Case of Linker Cell Death B. Evidence from Weak ced-3 Loss-of-Function Genetics XVI. Control of DNA Degradation in C. elegans—NUC-1, CPS-6/endoG, and WAH-1/AIF XVII. Conclusion Acknowledgments References

‘‘Life is pleasant. Death is peaceful. It’s the transition that’s troublesome,’’ said Isaac Asimov. Indeed, much scientific work over the last hundred years centered around attempts either to stave oV or to induce the onset of death, at both the organismal and the cellular levels. In this quest, the nematode C. elegans has proven an invaluable tool, first, in the articulation of the genetic pathway by which programmed cell death proceeds, and also as a continuing source of inspiration. It is our purpose in this Chapter to familiarize the reader with the topic of programmed cell death in C. elegans and its relevance to current research in the fields of apoptosis and cell corpse clearance. C 2005, Elsevier Inc.

I. Developmental Cell Death Occurs in C. elegans According to a Strict Cell Lineage The life cycle of C. elegans consists of a series of developmental stages (embryogenesis and four larval stages) that finally generate a fertile adult male or hermaphrodite (Brenner, 1974). During this period, exactly 131 cells in the hermaphrodite die as a result of a developmental cell death program (Kimble and Hirsh, 1979; Sulston and Horvitz, 1977; Sulston et al., 1983). The majority of embryonic cell death occurs between 250 and 500 min of embryogenesis; by the time embryos have reached the 3-fold stage (500 min), most of these doomed 131 cells have died and been rapidly phagocytosed (engulfed) by their neighbors. In worms defective in the engulfment apparatus, apoptotic cell corpses, which appear as raised, refractile disks under Nomarski optics (Fig. 3b), persist into early larval stages. A genetic pathway for programmed cell death has been elucidated through screens for mutations that aVect the number, pattern, and nature of developmental cell death. Epistasis analysis has placed these mutations into a (mostly) linear pathway by which programmed cell death is thought to

1. Programmed Cell Death in C. elegans

3

proceed (Fig. 1; Metzstein et al., 1998). This pathway can be further subdivided into four components: induction of programmed cell death, execution of apoptosis, engulfment of the apoptotic cell corpse, and degradation of the corpse within the phagocyte (Table I).

II. Four Genes Identify a Conserved Genetic Pathway for the Induction of Cell Death In the 1970s, it was discovered while analyzing the strict cell lineage of the worm that certain cells would ‘‘die’’—condense soon after cell division and be eliminated by nearby cells—in the normal course of development (Sulston and Horvitz, 1977; Sulston et al., 1983). Some of the earliest screens done in the worm searched for mutations that either aVected the clearance of these apoptotic cells or looked for worms where cells did not undergo cell death, and, instead, survived through development and the normal life of the animal (Ellis and Horvitz, 1986; Hedgecock et al., 1983). These (and later) screens identified mutations in the core apoptotic machinery: lossof-function mutations in ced-3, ced-4, and egl-1 (and gain-of-function alleles of ced-9) result in increased cell survival, whereas loss-of-function alleles of ced-9 (and gain-of-function mutations in egl-1) result in increased cell death (Fig. 1). A. The CED-3 Caspase CED-3 was the first caspase family member with a demonstrated role in apoptosis (Yuan et al., 1993); at the time, the role of caspases in apoptosis was not yet elucidated. Caspases are cysteine aspartyl proteases that, when activated, will proteolytically cleave substrate proteins (Hengartner, 2000). Caspases are made as inactive pro-enzymes—activation occurs either by proteolytic cleavage or dimerization, resulting in an active enzyme (Boatright and Salvesen, 2003). While several hundred caspase substrates have been identified in mammals (Fischer et al., 2003), almost nothing is known about the in vivo substrate(s) of activated CED-3. It has been published that activated CED-3 cleaves FEM-3, a factor involved in sex determination (Chan et al., 2000), but the meaning of this cleavage is unclear; one theory is that it may function as a feedback loop to reinforce sex-specific killing. CED-3 has also been suggested to cleave CED-9, resulting in more mature CED-3 production in a vicious circle that ultimately kills the cell (Goodwin and Ellis, 2002; Xue and Horvitz, 1997). However, the significance of this cleavage has been called into question, as caspase cleavage of target proteins are not important for the proper regulation of the

Figure 1 Programmed cell death in C. elegans proceeds via a conserved genetic pathway. The genetic pathway of programmed cell death in C. elegans has traditionally been broken down into three discrete stages—induction of death, phagocytosis (or engulfment) of the apoptotic cell, and degradation of the apoptotic corpse. Induction of cell death happens diVerently in diVerent tissues, but the core machinery that kills the cell (egl-1, ced-9, ced-4, and ced3) and the machinery that results in removal and degradation of the apoptotic cell (ced-1, ced-2, ced-5, ced-6, ced-7, ced-10, and ced-12) function independently of cell type.

Table I

List of Genes Involved in Programmed Cell Death in C. elegans and Role, Where Known

Protein name Factors involved in apoptotic death CED-3 CED-4

Homologue/Function

Reference

Caspase protease Apaf1 homologue, functions to mediate autocatalytic cleavage of CED-3

(Yuan et al., 1993) (Irmler et al., 1997; Seshagiri and Miller, 1997; Shaham and Horvitz, 1996; Yuan and Horvitz, 1992) (Chen et al., 2000; Hengartner and Horvitz, 1994b; Hengartner et al., 1992; Spector et al., 1997; Wu et al., 1997; Xue and Horvitz, 1997) (Chen et al., 2000; Conradt and Horvitz, 1998, 1999; del Peso et al., 1998, 2000) (Bloss et al., 2003)

CED-9

Bcl-2 orthologue, localizes CED-4 to mitochondria, preventing CED-3 activation

EGL-1

BH3-only domain containing protein, displaces CED-4 from CED-9, inducing apoptosis NAC homologue, associated with mitochondrial membrane, induces ced-3 independent death

ICD-1 Factors involved in transcriptional regulation of cell-specific death CES-1 CES-2 HLH-2, HLH-3 EOR-1, EOR-2 EGL-41 TRA-1

Negatively regulates transcription of egl-1, antagonizes HLH-2/HLH-3 binding Negatively regulates transcription of ces-1 by binding to elements in its promoter Positively regulates expression of EGL-1 in the NSM sisters Putative transcription factors, eor(lf) results in inappropriate HSN survival egl-41(gf) results in inappropriate death of the HSNs binds egl-1 locus to repress transcription in hermaphrodite HSN neurons

(Ellis and Horvitz, 1991; Metzstein and Horvitz, 1999; Thellmann et al., 2003) (Ellis and Horvitz, 1991; Metzstein and Horvitz, 1999; Metzstein et al., 1996) (Thellmann et al., 2003) (Howard and Sundaram, 2002; Rocheleau et al., 2002) (Desai and Horvitz, 1989) (Conradt and Horvitz, 1999)

(Continued )

Table I

Continued

Protein name Factors involved in physiological germ cell death CGH-1 DAZ-1 Factors involved in DNA damageinduced apoptosis HUS-1 MRT-2 RAD-5/CLK-2 CEP-1 APE-1 Factors involved in DNA degradation NUC-1 WAH-1 CPS-6 CRN-1 Factors involved in programmed cell death of unknown function CED-8

Homologue/Function

Reference

RNA helicase RNA-binding protein required for exit from pachytene of meiosis I

(Navarro et al., 2001) (Karashima et al., 2000)

Forms PCNA-like structure with MRT-2, localizes to distinct puncta on irradiated chromosomes Forms PCNA-like structure with HUS-1, mortal germ line Involved in DNA damage response as well as a key regulator of lifespan C. elegans p53 homologue Directly interacts with cep-1, negative regulator of cep-1-induced apoptosis

(Hofmann et al., 2002)

DNase II homologue, required for resolution of TUNEL(+) ends AIF EndoG FEN-1 homologue

XK homologue, ced-8(lf) causes delayed corpse appearance and clearance

(Ahmed and Hodgkin, 2000) (Ahmed et al., 2001; Benard et al., 2001; Lim et al., 2001) (Derry et al., 2001; Schumacher et al., 2001) (Bergamaschi et al., 2003)

(Wu et al., 2000) (Wang et al., 2002) (Parrish et al., 2001) (Parrish and Xue, 2003; Parrish et al., 2003)

(Hoeppner et al., 2001; Stanfield and Horvitz, 2000)


7

apoptotic cascade in a reconstituted heterologous system (Chris Hawkins, personal communication).

B. The Apaf-1 Orthologue CED-4 When CED-4 was cloned, it was found to be a novel protein (Yuan and Horvitz, 1992); only later, with the identification and subsequent characterization of its mammalian homologue, Apaf1 (Zou et al., 1997a), was it shown that CED-4 is required for activation of the ced-3 caspase homologue (Seshagiri and Miller, 1997). Through alternative splicing, the ced-4 locus gives rise to a major isoform, CED-4s, and a minor isoform, CED-4L, diVering only in a portion of a single exon. Interestingly, these two proteins have antagonistic eVects: when overexpressed, CED-4L has anti-apoptotic eVects, while CED-4S promotes cell death (Shaham and Horvitz, 1996). However, the exact mechanics of involvement of these two splice variants in the ‘‘apoptosome’’ remains uncharacterized. Whereas worm genetics has been essential in delineating the core apoptotic machinery, it has largely been the mammalian cell biologists and biochemists who have elucidated mechanistic details. Using protein expressed in mammalian cell systems, it has been shown that CED-4 likely functions as an oligomer whose function is to promote the autocatalytic cleavage of proCED-3 (Yang et al., 1998). Normally localized to the mitochondria in the C. elegans embryo, CED-4 translocates to the perinuclear region upon induction of apoptosis (Chen et al., 2000). In ced-3(lf) mutants (that lack a functional CED-3 caspase), translocation still occurs, suggesting CED-3 plays no role in perinuclear recruitment; however, whether perinuclear localization is important for CED-4 function is still unknown, as is the subcellular localization of both pro-CED-3 and activated CED-3.

C. The Bcl-2 Family Member CED-9 ced-9 was the first (and, for a long time, the only) anti-apoptotic factor isolated in C. elegans (Hengartner and Horvitz, 1994b; Hengartner et al., 1992). The pro-survival function of ced-9 is confirmed by its mutant phenotypes—loss-of-function mutations result in extensive death of cells that normally survive, whereas a rare gain-of-function allele results in increased cell survival (Hengartner and Horvitz, 1994a,b). The mammalian orthologue of ced-9, bcl-2, also plays a protective role in cell death (Adams and Cory, 1998). CED-9 is thought to function by directly binding to CED-4 and sequestering it in a monomeric state, rendering it unable to activate the CED-3 caspase (Chen et al., 2000; Spector et al., 1997). While attractive

8


because of its simplicity, this model does not explain why CED-9 is associated with mitochondria. Data from cell culture studies suggest that Bcl-2 family members might regulate apoptosis via diVerent mechanisms, as Bcl-2 likely protects from apoptosis by binding to Bax and preventing Bax-mediated release of cytochrome c and other proteins from the intermembrane space (Degli and Dive, 2003). Moreover, while there is no Bax homologue in C. elegans, genetic studies have suggested that CED-9 possesses both anti- and pro-apoptotic functions (Hengartner and Horvitz, 1994a; Xue and Horvitz, 1997). Thus, an additional, more ‘‘traditional’’ Bax-like function cannot be excluded for CED-9. Indeed, several reports have suggested that at least some intermembrane space proteins, such as AIF/WAH-1 and endoG/CPS-6, can exit mitochondria upon induction of apoptosis (Parrish et al., 2001; Wang et al., 2002). At least some conservation of molecular function is also suggested by the observation that Bcl-2 can prevent apoptosis when overexpressed in C. elegans (Vaux et al., 1992) and can at least partially substitute for CED-9 (Hengartner and Horvitz, 1994a). The crystal structure of the central core of CED-9 has been solved and has been found to share great similarity with BCL-XL, BAX, and BCL-W. CED9 consists of an unstructured region (aa 1–67), the BH domains (aa 68–242), a flexible linker region (aa 243–351), and a transmembrane region (aa 252–280) (Woo et al., 2003). One interesting diVerence between CED-9 and other Bcl-2-like proteins is that the C-terminal region of CED-9 immediately before the transmembrane segment interacts with the BH3 domain, which may regulate CED-9’s ability to interact with either CED-4 or EGL-1. Another diVerence is in the shape of the BH3 binding groove, which is startlingly diVerent from that found in other BH3-domain-containing proteins (Woo et al., 2003). Indeed, the interaction between EGL-1 and CED-9 is much more extensive, and much more stable, than equivalent interactions reported so far in mammals. How both of these diVerences impact function of CED-9 in programmed cell death remains to be determined.

D. The EGL-1 BH3-only Domain Protein BH3-only proteins are divergent Bcl-2 family members that interfere with the function of pro-survival Bcl-2 homologues and induce apoptosis (Yang et al., 1995). The only common feature of proteins in this group is a short 12 aa sequence, known as the Bcl-2 Homology 3 (BH3) domain, which can directly bind to the hydrophobic groove found in ‘‘full-fledged’’ Bcl-2 family members (Petros et al., 2004). In C. elegans, the BH3 domain protein EGL-1 functions to induce cell death during somatic development and following DNA damage in C. elegans (Conradt and Horvitz, 1998, 1999; Gartner et al., 2000; Thellmann et al., 2003). EGL-1 interacts with CED-9 via its BH3


9

domain (Bcl-2 Homology 3 domain); binding of EGL-1 to CED-9 results in a conformational change in CED-9 (Woo et al., 2003) and the untethering of CED-4 from CED-9, followed by onset of apoptosis (Conradt and Horvitz, 1998; del Peso et al., 1998, 2000). Expression of EGL-1 during embryogenesis results in translocation of CED-4 to the perinuclear region and induction of apoptosis (Chen et al., 2000), confirming this model. Many cells appear to control the induction of programmed cell death by controlling the expression of EGL-1 at the transcriptional level (Conradt and Horvitz, 1999). Specific examples of cell-type specific regulation of egl-1 expression are discussed in the following text.

E. Summary Activation of the C. elegans caspase homologue, CED-3, is tightly regulated by the actions of CED-4/Apaf-1, CED-9/Bcl-2, and the EGL-1/BH3-only domain protein. CED-3 must interact with CED-4 in order to become active; thus, ced-4(lf) mutants have the same phenotype as ced-3(lf) mutants. Loss of function alleles of ced-9 result in increased cell death, suggesting that CED-9 has a protective role in apoptosis (sequestering CED-4 to the mitochondria); a gain-of-function mutation in ced-9 has also been shown to decrease cell death. EGL-1, the most upstream member of the core machinery, is proapoptotic (Conradt and Horvitz, 1998, 1999); biochemical data suggest that EGL-1 binding to CED-9 releases CED-4 and thereby allows caspase activation (del Peso et al., 2000).

III. CED-8 Plays an Unknown Role in Programmed Cell Death ced-8 appears to be primarily involved in the kinetics of cell death—ced-8(lf) mutants show a slower appearance (and removal) of cell corpses (Hoeppner et al., 2001; Stanfield and Horvitz, 2000) during development. Why corpse appearance is delayed is unknown; it may be a result of either a delay in the onset of apoptosis or in the visual appearance of a corpse due, for example, to defects in corpse condensation. Interestingly, ced-8 mutations can enhance cell survival in weak ced-3(lf) mutants (similar to mutations in both DNA degradation and engulfment), suggesting a role either in cell killing or in cell engulfment (to be discussed further). The CED-8 protein is an integral membrane protein similar to human XK, which is mutated in patients with McLeod’s Syndrome (Stanfield and Horvitz, 2000), a disorder characterized by acanthocytic (spiny) erythrocytes and progressive dementia. The significance of this sequence similarity is presently unknown.

10


IV. CED-3 Independent Cell Death—Role of icd-1 While ced-3 is clearly a key eVector of apoptosis in C. elegans, not all programmed cell deaths appear to require functional caspase. For example, the death of the linker cell in males was reported to occur 50% of the time in ced-3 mutants (Ellis and Horvitz, 1986). In 2003, a more general pathway for CED-3-independent cell death was described by Rothman and colleagues, who identified the icd-1 gene as an anti-apoptotic factor in a reverse genetic screen looking for genes that regulate early embryogenesis (Bloss et al., 2003). RNAi-mediated knockdown of icd-1 resulted in a greatly increased number of cell corpselike structures in a ced-4-dependent, ced-3-independent fashion. The ced-9(gf) mutation reduces its ability to interact with the proapoptotic BH3-only domain containing protein EGL-1; this mutation had no eVect on icd-1-induced apoptosis, suggesting that this death occurs independently of egl-1 induction. Interestingly, icd-1 (RNAi)-induced death is also independent of the strict lineage system—cells that would normally live, such as those that make up the adult intestine, for example, undergo apoptosis and are eYciently engulfed, suggesting this gene plays a role in keeping the apoptotic machinery quiescent in cells that are not programmed to die. ICD-1 appears to be associated with the mitochondria, and is similar to the mammalian protein NAC, which is thought to play a role in regulation of protein localization during translocation to the mitochondria (Bloss et al., 2003). Thus, ICD-1 may normally function to localize CED-9 to the mitochondria. How ICD-1 can protect from ced-3-independent cell death is unknown. ICD-1 may contain a putative CARD (caspase activation and recruitment) domain, which has been shown to mediate caspase interactions; thus, one possibility is that icd-1 knockdown activates one of the other caspases present in C. elegans (Shaham, 2003).

V. Similarities and Distinctions in the Apoptotic Machinery of C. elegans and Mammals While the genetic pathway of programmed cell death is conserved between C. elegans and higher metazoans, the biochemical mechanism by which caspases are activated, as well as the subcellular localization of certain proteins, is quite distinct. While CED-9 and Bcl-2 are both localized to the mitochondria (Chen et al., 2000; Hockenbery et al., 1990), the mammalian CED-4 homologue, Apaf-1, does not associate with Bcl-2 and is not localized to the mitochondrion (Hausmann et al., 2000). In worms, transcriptional regulation of the BH3-only protein EGL-1 appears to be the major mechanism of control of developmental cell death, with EGL-1 displacing CED-4 from CED-9; CED-4 then mediates activation of the CED-3 caspase


11

and induction of cell death. In contrast, in mammals it is release of cytochrome c from the mitochondria that favors activation of the APAF1/Caspase complex (Zou et al., 1997b). There is no known role for mitochondrial permeabilization and release of cytochrome c in the activation of CED-4 activity; indeed, CED-4 diVers from APAF1 in that it does not have the WD40 repeat known to act as the cytochrome c interaction motif (Danial and Korsmeyer, 2004). In addition, there is little evidence for the involvement of multiple caspases in the induction of apoptotic cell death, although there are other caspase homologues in the C. elegans genome (Shaham, 1998). Perhaps these proteases are used in other biological processes, similar to mammalian ICE, which is a caspase homologue that plays a role in generation of IL-1 (Li et al., 1995). There are also two IAP homologues in C. elegans, but neither has been shown to play a role in apoptosis. The surviving homologue bir-1, like its mammalian and yeast homologues, is required for chromosome segregation and cytokinesis (Fraser et al., 1999; Speliotes et al., 2000). No function has yet been ascribed to the two-BIR domain-containing protein, BIR-2. In summary, it appears that while the proteins that mediate cell death have been conserved during evolution, they have also developed new abilities, and the control of cell death induction has also become more complex in vertebrates.

VI. Cell-Type Specific Regulation of Developmental Cell Death During mammalian development, a large number of cells are born, then removed in a process known as tissue sculpting. This is vital to the formation of many organs, as well as nonvital structures such as fingers and toes. How apoptosis is induced (and restricted to) a particular population of cells is a subject of particular interest. In C. elegans, because of the strict embryonic lineage system, it is simple to study cell-specific induction of programmed cell death. It is known when and where a particular cell will be born, as well as the approximate time that death will ensue. Thus, apoptosis of particular cells can be studied in a highly reproducible setting.

A. Regulation of Cell Death in the HSN (Hermaphrodite Specific Neurons): Involvement of the Sex Determination Machinery In C. elegans, the decision of several cells to die is based on the sex of the animal. The two bilaterally symmetrical serotonergic motor neurons, HSNR and HSNL, are born in the tail region close to midway through embryonic

12


development. In the hermaphrodite, the newly born HSNs migrate anteriorly until they take up position at approximately the middle of the worm, where they will eventually innervate the egg-laying muscles, as the HSNs are responsible for inducing vulval muscle contractions associated with egglaying (Desai et al., 1988). In males, the HSNs are not needed and are thus eliminated via apoptosis soon after birth (Kimble and Hirsh, 1979; Sulston et al., 1980). Likewise, the hermaphrodite does not need neurons that are involved in the male’s complex mating behavior, such as the CEMs (Sulston et al., 1980); these are also eliminated shortly after birth, this time specifically in hermaphrodites. A key sex-determination protein, TRA-1A, appears to be the main method of determining life or death of the HSN. The TRA-1A transcription factor has been shown to bind to an element 6 kb downstream of the egl-1 transcriptional unit in hermaphrodites (Conradt and Horvitz, 1999), resulting in repression of transcription and cell survival. In males, however, TRA-1A levels are much lower, and egl-1 is expressed, resulting in cell death. Mutations that disrupt the TRA-1A binding site in egl-1, such as n1084gf, allow egl-1 expression in the HSNs of both sexes, resulting in inappropriate death of the HSN in hermaphrodites. Screens for mutations that suppress the inappropriate death of the HSN in egl-1(n1084gf) mutants have been performed, with some degree of success. Two genes found, eor-1 and eor-2, may function in regulation of programmed cell death in the HSNs; mutations in these genes result in enhanced survival of the HSN (D. Hoeppner, M. Spector, and MOH, manuscript in preparation). eor-1 and eor-2 were isolated in multiple screens as suppressors of egl-1, enhancers of lin-45/raf, as well as having defects in phasmid neuron dye uptake. It has been shown that eor-1 and eor-2 function downstream of let-60/Ras in a number of tissues (Howard and Sundaram, 2002; Rocheleau et al., 2002), potentially by promoting (or repressing) transcription downstream of Ras signaling. eor-1 and eor-2 mutants show a significant embryonic lethality associated with inappropriate function of the excretory cell, which causes the worm to die as fluid-filled ‘‘rods’’ (Rogalski et al., 1982). This phenotype is also common to genes that result in ras(lf); indeed, double mutant analysis suggests that the two eor genes function downstream or in parallel to mpk-1, the C. elegans ERK homologue during vulval morphogenesis (Rocheleau et al., 2002). The eor genes also appear to function downstream of wingless (Wnt) signaling, as mutations in eor-1 and eor-2 suppress vulval phenotypes associated with pry-1/Axin; thus, the eors may be required for multiple signaling pathways (Howard and Sundaram, 2002). Both eor-1 and eor-2 are cloned: eor-1 codes for a Zn-finger containing transcription factor most similar to PLZF, and eor-2 for a ‘‘novel’’ but conserved protein. Both proteins are localized primarily to the nucleus, as


13

determined by translational fusions, consistent with a role in transcriptional regulation. eor-1 and eor-2 have been shown to regulate transcription of egl17 in vulval tissues (Howard and Sundaram, 2002), further emphasizing a possible function in transcriptional activation/repression. The mode of eor function in HSN cell death is currently unknown; while transcriptional regulation of egl-1 is one attractive hypothesis, it is also possible that the eors may function in the HSNs as they do in the vulva, downstream of an uncharacterized Ras signaling module.

B. Regulation of Cell Death in the NSM Sisters—Involvement of ces-1 and ces-2 The NSM sister cells are, as their name implies, sisters of the two bilaterally symmetrical pharyngeal NSM neurons. NSMs and their sisters are generated by the division of a common mother cell [Abaraap(a/p)paa] during embryonic development (Sulston et al., 1983). However, unlike the HSN neurons, these sister cells are induced to undergo programmed cell death, independent of the sex of the worm, shortly after birth. The ces genes (cell death specification) were isolated as mutations that prevent cell death in a limited number of neuronal cells that are generated and subsequently killed during normal C. elegans development. This is in contrast to the genes that make up the core apoptotic machinery, such as egl-1, which play a role in all developmental cell deaths. Loss-of-function mutations in ces-2, and gain-of-function mutations in ces-1, result in inappropriate survival of the NSM sisters. Epistasis analysis using ces-1(lf), where death of the NSMs is normal, revealed that ces-2 acts as a negative regulator of ces-1 (Ellis and Horvitz, 1991). Both CES-1 and CES-2 are potential transcription factors, of the Snail family and HLF(bZIP) family, respectively (Metzstein and Horvitz, 1999; Metzstein et al., 1996). Antagonism of CES-1 function by CES-2 is likely to occur at the transcriptional level, with CES-2 regulating CES-1 expression (Metzstein et al., 1996). Consistent with this, mutations that result in gain of ces-1 function localized not to the protein coding sequence but rather to a putative 50 regulatory element (Metzstein and Horvitz, 1999). Interestingly, this element contains a bZIP family consensus-binding motif that was able to interact with CES-2 in a gel shift assay (EMSA—electrophoretic mobility shift assay). These results suggest that CES-2 may antagonize CES-1 function via transcriptional repression. How does CES-1 antagonize NSM sister cell death? A prime candidate for the CES-1 transcription factor is the BH3 domain protein EGL-1. Recent work has shown that ces-1 activity is required for expression of an egl-1 transcriptional reporter construct in the NSM sisters. Further analysis of the

14


egl-1 locus has identified a region that has several Snail-binding sites, which CES-1 is able to bind to in vitro (Thellmann et al., 2003). The egl-1 locus also contains a distinct E-box motif in the midst of the CES–1 Snail interaction motifs, which is commonly bound to by bHLH transcription factors, which serve as transcriptional activators. Using reverse genetics, two bHLH transcription factors, hlh-2 and hlh-3, were identified that bind to the E-box motif; RNA interference-mediated knockdown of these genes increased NSM sister cell survival. Epistasis analysis using ces-2(lf) and ces-1(lf) suggests that HLH-2 and HLH-3 function downstream of the ces genes to induce NSM apoptosis (Thellmann et al., 2003). Together, these results suggest an elegant mechanism for regulation of the induction of apoptosis in the NSM sisters, in which apoptosis is induced by HLH-2/HLH-3mediated upregulation of egl-1 transcription, with CES-1 antagonizing this activity by competing with HLH-1/HLH-3 for binding to the EGL-1 enhancer element. This remains only a theory, however, as the mechanics of CES-1 blockade have yet to be described, either in vivo or in vitro. How each of these transcriptional activities is developmentally regulated to correctly specify the death of the NSM sisters (and likely of other cells as well) remains to be elucidated.

VII. C. elegans as a Model for Neuronal Degeneration Neuronal degeneration is a major cause of age-related death and dementia in the elderly. Due to research accomplished in the past two decades, C. elegans is fast becoming a major model system for these diseases. There appear to be two types of neurodegenerative death in worms—one that occurs due to overactive ion channels and is independent of CED-3 caspase activity (Hall et al., 1997), and one due to overexpression of poly-Glucontaining proteins that is dependent on the CED-3 caspase (Faber et al., 1999, 2002). Several genes in C. elegans, collectively referred to as degenerins (Driscoll and Chalfie, 1991; Faber et al., 1999, 2002), have been identified in screens for mutants suVering from uncoordinated movement. Rare dominant gainof-function (gf) alleles in several degenerins induce necroticlike death of cells upon overactivation; these deaths are distinct from apoptotic deaths both genetically (they are independent of both ced-3 and ced-4) and morphologically. Developmental apoptosis is characterized by the condensation of apoptotic cells; in neurons aVected by mec-4(d) and deg-1(d) alleles, however, cells appear to swell, with vacuoles accumulating around the nucleus, which eventually dissolves (Hall et al., 1997). This cell expansion has proved to be the key to determining the function of the degenerins, which encode overactivated ion channels. In degenerin(gf ) mutants, the cell must take in


15

fluid to alleviate the osmotic imbalance generated by the overactive ion channels, and eventually dies because of it, which also explains why these deaths are CED-3 independent. Further work has determined that, while degeneration due to mec-4(d) and deg-1(d) mutations appears similar, the molecular basis is quite divergent. Loss-of-function mutations in the small chaperone calreticulin (crt-1) were isolated as suppressors of mec-4(d)-mediated degeneration and also of degeneration associated with an activated Gs subunit (Xu et al., 2001). Ca2þ release from the endoplasmic reticulum (ER) was determined to be essential for mec-4(d)-mediated degeneration by a variety of genetic and pharmacological assays. Interestingly, this does not seem to be the case for deg1(d)-mediated degeneration, as crt-1(lf) does not reduce neurodegenerative death in these mutant worms. In addition to degeneration induced by activated ion channels, neurodegeneration can also be generated by overexpression of certain proteins, similar to what is seen in mammalian cells upon accumulation of proteins in Huntington’s disease and Alzheimer’s disease (Faber et al., 2002; Goedert, 2003; Nass et al., 2001). Interestingly, overexpression of poly-Glu-containing proteins results in degenerative cell death that is at least partially dependent on the core apoptotic machinery (Faber et al., 1999), suggesting that there may be a developmental pathway that has evolved to remove cells that inadvertently overexpress certain proteins. The purpose of this removal is as yet unknown, but perhaps it is to prevent a degenerative form of death that would damage the processes of adjacent neurons.

VIII. Programmed Cell Death During Differentiation of the Hermaphrodite Germ Line A. Structure of the Hermaphrodite Germ Line The hermaphrodite gonads in C. elegans consist of a pair of bilaterally symmetrical tubes containing germ cells; these tubes are surrounded by the somatic sheath cells and joined together through a common uterus (Fig. 2A) (Sulston, 1988). The gonad of the worm begins to develop in the early larval stages, when four gonadal progenitor cells, Z(1–4), start dividing. Z2 and Z3 will give rise to the germ cells, while Z1 and Z4 generate the somatic gonad, which surrounds the germ cells and isolates them from the rest of the worm (Kimble and Hirsh, 1979). The final shape of the adult gonad is determined by the migration path of the Distal Tip Cells, or DTCs. These somatic gonadal cells migrate along the body wall muscle in response to cues secreted into the extracellular matrix (Nishiwaki, 1999). This migration has been shown to be dependent on a number of diVerent proteins, including the

16


Figure 2 Germ cell apoptosis in the C. elegans Hermaphrodite Gonad. (A) Schematic representation of the adult hermaphrodite gonad. The hermaphrodite gonad is a bilobed, U-shaped tube that is connected at a central uterus. The shape of the mature gonad is determined by the migration of the distal tip cell (DTC, arrow). Cells are generated in the mitotic zone (a), then progress through the transition zone and enter meiosis I (b). When cells exit from the pachytene stage (c), they become competent to undergo apoptosis; apoptotic cells cellularize away from the common syncytial cytoplasm (arrowhead) and are eYciently engulfed by the somatic sheath cells, which surround the gonad; surviving cells enlarge and arrest at the diakinesis stage (d). Upon activation, mature oocytes progress through the spermatheca (e), where they are fertilized, and into the uterus (f ), where they wait to be laid through the vulva. (B) Apoptosis during oogenesis. In wild-type worms, two to three apoptotic cells are visible in the adult hermaphrodite gonad at any given time, dependent on the age of the animal (wild-type, arrow). In ced-9(n1653ts) mutants, germ cells undergo extensive apoptosis due to constitutive induction of programmed cell death (arrows). Size bar, 10 m. (C ) DNA damage induces


17

Rac pathway members ced-2, -5, -12, and ced-10/Rac1 (Lundquist et al., 2001). As is the case in the reproductive tissues of higher organisms, including humans, cells in the germ line of C. elegans undergo extensive programmed cell death (Fig. 2B). In contrast to the deaths that occur in the soma during development, deaths in the germ line do not follow a set pattern (Gumienny et al., 1999); there is no fixed lineage specifying which cells will live or die. In addition, cell death in the germ line increases over the life of the adult animal, such that older adults have more germ cell corpses than do young adults. The cause of this increase is not clear. It is possible that, over time, replicative errors accumulate in the mitotic germ cell nuclei, and that the increased cell death is caused by an increase in damaged germ cells (Gumienny et al., 1999). A simpler explanation is that, as the worm ages, it continues to grow; thus, the germ line is larger in adults and would contain a larger number of germ cells that can die. Characterization of germ cell death has been significantly facilitated through the use of the vital dyes SYTO 12 and acridine orange (AO). AO is a DNA (and RNA) intercalating agent, which selectively stains apoptotic corpses in the germ line (Abrams et al., 1993; Gumienny et al., 1999); acridine orange seems to stain only a subset of apoptotic cells, probably at a particular stage of programmed cell death. Interestingly, the Ras/MAP kinase pathway is required both for proper progression of germ cells through meiosis and for germ line apoptosis (Gumienny et al., 1999). Activation of this pathway appears to be a permissive (rather than instructive) event for apoptosis; Ras-MAPK-dependent progression through pachytene of meiosis I appears to be essential to establish a diVerentiation state where the germ cells are receptive to an apoptotic stimulus (Gumienny et al., 1999).

B. Regulation of Germ Cell Apoptosis The cell death machinery is largely conserved between developmental cell death and germ line cell death. A major diVerence, however, is that egl-1 is not required for physiological germ cell death to occur; egl-1(lf) mutants show wild-type levels of apoptotic death in the hermaphrodite germ line. In addition, a ced-9(gf) mutation that reduces the aYnity of CED-9 for EGL-1 (Woo et al., 2003) also has little eVect on germ cell death (Gumienny et al., multiple responses in germ cells. Following irradiation, mitotic germ cells undergo a transient cell cycle arrest (black arrowheads) and become significantly larger than cycling cells (white arrowheads). In the meiotic zone, DNA damage induces extensive germ cell apoptosis (irradiated, white arrows), potentially due to irreparably damaged DNA. Size bar, 10 m.

18


1999). This would suggest that there must be a novel mechanism for controlling the progression of programmed cell death in the hermaphrodite germ line independent of the BH3-only domain protein EGL-1. Several genes have already been discovered that appear to specifically aVect programmed cell death in the germ line. One of these, cgh-1, is a predicted RNA helicase that appears to function specifically in the germ line. Knockdown of cgh-1 by RNA interference (RNAi) results in increased germ cell death (Navarro et al., 2001). This is intriguing, as it suggests that control of cell death in the germ line may occur post-transcriptionally, rather than at the transcriptional level, as is the case during somatic development. daz-1, another RNA-binding protein, also shows increased apoptosis in the germ line in loss-of-function mutants (Karashima et al., 2000). This protein, a homologue of mammalian DAZ (Deleted in Azoospermia), is required for proper spermatogenesis in humans; mutations of this gene cause Azoospermia, which is characterized by diverse defects in sperm development. C. elegans daz-1(lf), however, is required for proper oogenesis, but is not involved in spermatogenesis. Surprisingly, daz-1(lf) germ cells show an arrest in pachytene of meiosis I (Karashima et al., 2000). Since progression through the pachytene checkpoint (and activation of MAPK) is required for induction of apoptosis in the germ line (Gumienny et al., 1999), presumably daz-1 mutant germ cells arrest after this and enter apoptosis because they are blocked from further meiotic progression. Interestingly, two other genes, nos-1 and nos-2, are also involved in germ line development and, when depleted by RNAi, have increased numbers of apoptotic cells (Subramaniam and Seydoux, 1999). Whether this increased cell death is further indicative of the importance in post-transcriptional control of apoptosis, or of a checkpoint that is activated by abnormal meiotic events, is as yet undetermined.

IX. Germ Cell Apoptosis in Response to Genotoxic Stress A. Genotoxic Stress Induces Cell Cycle Arrest and Apoptosis in the Hermaphrodite Germ Line Maintenance of genome integrity is of crucial importance to the life and proper functioning of the organism; indeed, defects in genes that mediate the response to DNA damage, such as p53, have been implicated in numerous diseases (e.g., cancer). Recent work has suggested that the C. elegans hermaphrodite gonad provides a good experimental system for the study of this process (Hofmann et al., 2000; Stergiou and Hengartner, 2004). There are two obvious phenotypic responses to a genotoxic insult in the adult hermaphrodite germ line (Fig. 2C). First, mitotic cells stop proliferating, presumably to repair damaged DNA; interestingly, cell growth is not


19

blocked, such that arrested cells will eventually be much larger than cycling mitotic germ cells (Fig. 2c). In the meiotic zone, nuclei that are too damaged to survive undergo programmed cell death and are eYciently removed by the engulfment apparatus (Fig. 2c) (Gartner et al., 2000). The molecular mechanisms that mediate the response to a genotoxic insult are just beginning to be dissected in C. elegans. Work from other organisms, primarily yeast and mammals, suggests damage such as double-strand breaks activate a kinase cascade, which ultimately activates p53 and induces cell cycle arrest and/or apoptosis, depending on cell type and severity of the damage (Zhou and Elledge, 2000). This generality appears to hold true in C. elegans as well, with a few key diVerences. Several proteins have been found to be essential for the germ line response to DNA damage in C. elegans. HUS-1 and MRT-2/RAD1 are both homologous to yeast proteins that form a PCNA-like clamp that has been implicated in recognition of double strand breaks (Ahmed and Hodgkin, 2000; Hofmann et al., 2002). Mutations in these genes result in worms that have neither cell cycle arrest nor apoptosis following induction of DNA damage. Following detection of the double-strand break, it is thought that a signal transduction cascade occurs that requires the ATM and ATR/L kinases to either induce apoptosis (via p53 phosphorylation) or cell cycle arrest and repair (Morgan and Kastan, 1997). Following this, factors are recruited for the repair of the damaged site; many of these factors are also involved in the repair of DNA strand breaks associated with meiotic recombination. Thus, for example, mutations in RAD-51, which binds to singlestranded DNA following strand resection around the break generated by the SPO-11 endonuclease, prevent both the completion of meiosis and repair, leading to the activation of the DNA damage response pathway and a massive increase in germ cell apoptosis (Alpi et al., 2003). Recently, a worm homologue of the mammalian p53 transcriptional regulator has been described (Derry et al., 2001; Schumacher et al., 2001). Interestingly, cep-1 (C. elegans p53 homologue) is only involved in the apoptotic response to DNA damage and not in mitotic cell-cycle arrest, suggesting that there are at least two divergent signaling arms downstream of recognition of the DNA lesion (Derry et al., 2001; Schumacher et al., 2001). However, nothing is currently known about how the cep-1 protein is regulated. Mammalian p53 appears to be involved in a tug-of-war between the actions of MDM2, which mediates export from the nucleus and degradation of p53, and ARF, which antagonizes MDM2 action to stabilize p53 and retain it in the nucleus by sequestering MDM2 to the nucleolus (Zhang and Xiong, 2001). However, worms do not have clear ARF or MDM2 homologues, and CEP-1 is constitutively nuclear, suggesting that function is probably controlled by post-translational modifications. One apparent negative regulator of CEP-1 function has been identified, ape-1, a homologue of the

20


iASPP protein in mammals (Bergamaschi et al., 2003). RNAi-mediated knockdown of ape-1 results in increased apoptotic germ cell corpses that is suppressed in cep-1(lf), suggesting that in ape-1(RNAi) mutants, CEP-1 is constitutively active and signaling for death. The precise mechanism of how ape-1 mediates inhibition of cep-1 remains to be determined. Apoptosis of meiotic germ cells following genotoxic stress is induced in much the same manner as during development, namely, by transcriptional regulation of the BH-3-only domain protein EGL-1 (Hofmann et al., 2002). Transcription of egl-1 is dependent on functional cep-1 (Hofmann et al., 2002); in other systems, p53 has been reported to directly regulate transcription of BH3-only domain containing proteins; it will be interesting to see if cep-1 bind to the egl-1 locus to regulate transcription following induction of the DNA damage response. Unlike the apoptotic response, the mechanism by which cell-cycle arrest is initiated is largely unknown, as is the stage at which the cell cycle arrests. In mammalian cells, p53 mediates cell cycle arrest at the G1/S transition (Bates and Vousden, 1996); however, as has been mentioned, cep-1 does not appear to function in cell cycle arrest in worms. RAD-5 was originally identified as a gene that, when mutated, reduced embryonic survival following irradiation (Hartman and Herman, 1982); rad-5 was later determined to be allelic to clk-2, a gene reported to control life span (Ahmed et al., 2001). RAD-5 encodes a homologue of the S. cerevisiae protein TEL2p, which is responsible for maintaining genome integrity by regulating telomere length. Confoundingly, no consensus exists as to whether RAD-5 functions in telomere maintenance, as diVerent groups have reported no eVect, decrease in telomere length, and increase in telomere length in rad-5 mutants. Unlike CEP-1, the RAD-5 protein participates both in cell cycle arrest and in apoptotic response following irradiation, suggesting that it may function upstream of cep-1, before the repair/death decision. The exact role of RAD-5 in damage response has yet to be determined. Double mutant analysis with hus-1 and mrt-2 alleles suggests that rad-5 may either function in a separate pathway or play a role in an additional checkpoint separate from hus-1 or mrt-2. Interestingly, the human homologue of rad-5/clk-2 (denoted as hCLK-2) has also been shown to play roles in cell cycle progression and apoptotic response, as well as in telomere length regulation (Jiang et al., 2003).

B. Double-Strand Breaks that Arise from Failed Meiosis Activate an Apoptotic Checkpoint, but Do Not Induce Cell Cycle Arrest To generate a mature oocyte, a mitotic germ cell must divide to generate a daughter cell, which will subsequently progress through the germ line and enter meiosis (Austin and Kimble, 1987). Chromosomes must then pair and


21

undergo a crossover event in order to be properly segregated; these crossover events occur under the aegis of a variety of conserved proteins that are currently the focus of much research in the worm. Following chromosome pairing, the worm homologue of the yeast SPO11 nuclease, also called SPO-11, makes a double-strand break in the chromosome (Dernburg et al., 1998). Following this, a group of proteins encoded by mre-11 and rad-50 (eponymous with S. cerevisiae homologues) catalyze strand resection, leading to strand invasion and eventually resulting in Holliday junction formation (Chin and Villeneuve, 2001). Intriguingly, mutants in genes that are involved in the resolution of SPO-11 induced double strand breaks (and breaks induced by DNA damage) show an increased number of apoptotic germ cell corpses, suggesting that aberrant meiotic intermediates can activate a checkpoint in the meiotic zone that triggers apoptosis (Alpi et al., 2003; Boulton et al., 2002; Chin and Villeneuve, 2001; Gartner et al., 2000). It has also been suggested that defects in chromosome segregation may activate this checkpoint; however, spo-11(lf) worms do not have an increased number of corpses but do have increased mis-segregation of chromosomes, suggesting that it is double-strand breaks that are sensed. In addition, mutations in him-5, which has increased mis-segregation of the X chromosome, do not show increased cell death (J. M. Kinchen, unpublished observations). Cell cycle arrest is not induced in the mitotic region in these meiosis mutants, likely due to the fact that the damage is specific to meiotic germ cells, but it does suggest that the two activities, cell cycle arrest and germ cell apoptosis, are separately inducible. Recently, one group has reported that the C. elegans homologues of BRCA-1 and BARD-1 function in meiotic DNA repair in C. elegans, suggesting that the role of these proteins is evolutionarily conserved (Boulton et al., 2004). Consistent with defects in resolution of meiotic breaks, brc-1 (RNAi) worms have increased apoptosis during normal germ cell development as well as following irradiation. In addition, other mutations in repair genes, such as RAD-51, RAD-50, and MRE-11, also show increased apoptosis following DNA damage (Alpi et al., 2003; Chin and Villeneuve, 2001; Gartner et al., 2000), suggesting that screens for increased germ line apoptosis may identify genes that are required for DNA repair as well as apoptosis.

X. Cannibalism and More—Phagocytosis and Degradation of the Apoptotic Cell Engulfment is the process by which cells that have died, either by apoptotic or necrotic means, are cleared, usually by professional phagocytes or bystander cells (Kerr et al., 1972). Apoptotic cells in C. elegans, which have no

22


professional phagocytes, are engulfed by cells immediately adjacent to the apoptotic cell, usually within an hour of the onset of apoptosis (Robertson and Thomson, 1982). In mammals, necrotic lysis of cells can result in local inflammation and exposure of auto-antigens to immune cells. Indeed, in some cases of lupus, anti-DNA antibodies have been discovered that would be diYcult to explain without the inappropriate lysis of a cell (Mevorach, 1999; Scott et al., 2001; Tan, 1994). Mice mutant in the c-Mer tyrosine kinase, which has been suggested to function in engulfment, also accumulate anti-self antibodies and become diseased (Scott et al., 2001). The ability to remove apoptotic cells and prevent exposure of self-epitopes would thus appear to be essential for averting inappropriate immune responses and diseased states in the animal. In C. elegans, most apoptotic cells progress through several diVerent morphological stages during their degradation (Hoeppner et al., 2001; Robertson and Thomson, 1982), resulting almost invariably in the formation of the familiar buttonlike corpse, which corresponds to the compacted apoptotic cell (Fig. 3A). Condensation (as observed by DIC optics) of

Figure 3 Engulfment of apoptotic cells and migration of the DTCs are defective in ced-2, -5, -10, and -12 mutants. Condensed apoptotic cells (A, B, arrows) are eYciently phagocytosed by an adjacent cell within a short time after onset of apoptosis in wild-type worms (A). In worms mutant in components of the engulfment machinery, such as ced-12(k149), corpses persist late into development (B, arrows). Size bar, 10 m. In addition to defects in corpse removal, ced-2, -5, -10, and -12 also have defects in certain cell migrations (C–F). Distal Tip Cell migration route determines the shape of the adult hermaphrodite gonad. In wild-type animals (C, E) the DTC’s migration creates a U-shaped tube. In ced-12(k149) mutants, for example, the DTCs do not migrate properly and can make extra turns (D, F), resulting in a malformed gonad. Size bar, 40 m.


23

the apoptotic cell begins at the cell periphery, resulting in an erythrocytelike morphology (refractile edges giving the cell a concave appearance); following this, the cell becomes fully refractile. Later, following engulfment, onset of degradation results in the appearance of less refractile areas of the corpse, as if it were deflating, until the corpse is no longer distinguishable from the engulfing cell (Hoeppner et al., 2001; Robertson and Thomson, 1982).

XI. Recognition of the Apoptotic Cell—The Varied Eat-Me Signal A. Phosphatidylserine (PS) Exposure and Corpse Recognition Following induction of apoptosis, various membrane events occur, leading to redistribution of phospholipids on the membrane surface (Fadok et al., 1998). One of these, phosphatidylserine (PS), is normally only present on the inner leaflet of the plasma membrane, and is actively removed from the outer membrane by a translocase activity (Schlegel et al., 1996). During apoptosis, translocase activity is inhibited and PS is actively transported to the outer leaflet by the activity of a transmembrane scramblase (Fadok et al., 2001; Schlegel et al., 1996). PS exposure is a conserved feature of apoptosis, and has been observed from mammals to flies (van den Eijnde et al., 2001); whether PS exposure also occurs in apoptotic cells in C. elegans is likely but has yet to be tested. Many have suggested that PS or a similar phospholipid acts as an ‘‘eat-me signal,’’ that is, a signal that tells an engulfing cell that an apoptotic cell is nearby and needs to be cleared. Indeed, the discovery of a specific receptor for PS, aptly named the phosphatidylserine receptor (or PSR), has brought this signal to the forefront of candidates (Fadok et al., 2000). Studies published in 2003 suggest that PSR may play a key role in engulfment in mice (Li et al., 2003); however, mutant worms defective in psr-1, the C. elegans orthologue of PSR, show only a very weak engulfment defect (Wang et al., 2003). Interestingly, PSR appears to be predominantly localized to the nucleus (via 5 NLS motifs) and cannot be visualized on the plasma membrane via either immunostaining or overexpression of GFPtagged constructs, confusing the interpretation of these experiments (Cui et al., 2004). It is formally possible that PSR shuttles between the nucleus and the plasma membrane; it will be interesting to see whether PSR serves a diVerent function in the nucleus (e.g., to make mouse macrophages competent to engulf) or whether it is a transcriptional regulator of the true PS receptor.

24


Figure 4 Genes required for engulfment are conserved between mammals and C. elegans. The factors that mediate engulfment in worms (A) and mammals (B) are closely related. While several potential ‘‘eat-me’’ signals are known in mammals (possibly restricted by tissue type),


25

B. Annexin I Involvement in Engulfment Annexin V has long been used to detect exposure of PS on the surface of apoptotic cells. However, research into the role of annexins in the recognition of apoptotic cells has long been neglected. It has been reported (Arur et al., 2003) that, upon induction of apoptosis, Annexin I is cleaved in a caspase-dependent manner to generate a protein that is then translocated to the outside of the cell, where it binds to PS (Fig. 4). Annexin exposure then promotes eYcient engulfment of the apoptotic cell by neighboring phagocytes. RNA interference-mediated knockdown of nex-1, a C. elegans homologue of annexin I, also results in clearance defects, suggesting that annexin I-mediated engulfment is an evolutionarily conserved event (Arur et al., 2003). However, deletion of the nex-1 locus does not show a defect in the engulfment of apoptotic cells (L. Neukomm and MOH, unpublished observations), suggesting that RNAi-mediated knockdown of nex-1 may have cross-reacted with another RNA or with multiple members of the annexin gene family in the worm. Thus, the role of nex-1 in programmed cell death in C. elegans is uncertain.

C. Opsonizing Agents Recognition of apoptotic cells in mammals appears to be quite complex, potentially utilizing diVerent molecules in diVerent tissues (Fig. 4). One of these molecules, MFG-E8, has been shown to opsonize the apoptotic cell and promote engulfment (Hanayama et al., 2002) (Fig. 4). Likewise, C1q, a component of complement, has been shown to attach to apoptotic cells in greater density than to nonapoptotic cells and enhance uptake by macrophages (Korb and Ahearn, 1997; Navratil et al., 2001; Ogden et al., 2001). MFG-E8 normally coats milk vesicles, and has been found to be associated with exosomes (Oshima et al., 2002); MFG-E8 is thus normally found in the extracellular space bound to vesicles rather than as a free protein. Whether MFG-E8 coats the apoptotic cell itself, or coats an exosome containing a component required for eYcient engulfment or immune maturation, remains to be determined.

only one protein, NEX-1, has been implicated in C. elegans. Similarly, a wide variety of apoptotic receptors have been identified in mammalian cells, with only two, CED-1 and PSR-1, potentially functioning in recognition of the apoptotic cell in the worm. Interestingly, while integrin receptors appear to play a role upstream of Dock180 in mammals, homologues of these proteins do not play a primary role in engulfment in C. elegans (Gumienny et al., 2001).

26


Gas6 has been shown in vitro to bind to PS and stimulate uptake of apoptotic cells (Ishimoto et al., 2000); Gas6 is a known ligand for the Mer tyrosine kinase. Interestingly, Merkd knockout mice, which express a truncated form of Mer that lacks the cytoplasmic domain, have defects in the engulfment of apoptotic cells (Scott et al., 2001). Since the allele used in these experiments still had a wild-type extracellular domain, the function of Mer as a receptor for apoptotic cells was not determined; indeed, macrophages from Merkd mutant mice were able to bind apoptotic cells with the same eYciency as wild-type mice. Unfortunately, C. elegans does not have a clear homologue of the Mer tyrosine kinase, so evolutionarily conserved function of this protein in engulfment cannot be assayed. Whether Mer functions as a receptor for apoptotic cells or as a player in signaling following recognition of the apoptotic cell needs to be examined further.

D. Nonengulfment Signals Studies in mammalian cells have also identified receptors that need to be disengaged in order to promote eYcient engulfment. One of these, CD31, was found to be involved in a homophilic interaction in wild-type cells (resulting in detachment of viable cells from phagocytes); this interaction is disengaged upon induction of apoptosis (Brown et al., 2002). This may be a signal to keep phagocytic cells from engulfing cells that expose a marker of cell death but are not actually dying (such as exposure of PS by aging erythrocytes or activated platelets), providing another level of signaling to prevent accidental cell killing. Mutations that result in ‘‘gain-of-function’’ of engulfment, where healthy cells are inappropriately engulfed, have never been isolated in the worm. Whether such ‘‘don’t eat me’’ signals also exist in C. elegans thus remains to be determined.

XII. Seven Genes Function in the Engulfing Cell to Mediate Phagocytosis of Apoptotic Cell Corpses in C. elegans That specific genes are required for engulfment of apoptotic cells in C. elegans was originally discovered in a series of Nomarski screens looking for changes in the pattern of normal programmed cell death (Ellis et al., 1991; Hedgecock et al., 1983). Loss-of-function mutations in genes that function in engulfment result in the accumulation of persistent apoptotic cell corpses (Table II). These mutations define two gene groups that can be placed into distinct genetic ‘‘pathways’’ based on additivity of the persistent cell corpse phenotype: loss-of-function alleles in genes of either pathway

27

1. Programmed Cell Death in C. elegans Table II Genes with a Known Role in Engulfment in C. elegans Protein name CED-1 CED-6 CED-7

CED-2 CED-5 CED-10

Homologue/Function Transmembrane receptor, similar to SREC, CD91/LRP scavenger receptors Adaptor protein, interacts with CED-1, may homodimerize when activated ABC transporter, required for membrane localization of CED-1 around apoptotic cells CrkII homologue, adaptor protein Dock 180 homologue, large ‘‘docking’’ protein, RacGEF Rac orthologue

CED-12

ELMO homologue, function in mediating CED-10 activation

NEX-1

Homologue of Annexin I, knockdown results in persistent cell corpses RhoG orthologue, may regulate CED-12 mediated Rac activation PSR homologue, knockdown results in a weak engulfment defect

MIG-2 PSR-1

Reference (Zhou et al., 2001b) (Liu and Hengartner, 1998; Su et al., 2000a, 2002) (Wu and Horvitz, 1998a; Zhou et al., 2001b) (Reddien and Horvitz, 2000) (Brugnera et al., 2002; Wu and Horvitz, 1998b) (Lundquist et al., 2001; Reddien and Horvitz, 2000) (Brugnera et al., 2002; Gumienny et al., 2001; Katoh and Negishi, 2003; Wu et al., 2001; Zhou et al., 2001a) (Arur et al., 2003) (Katoh and Negishi, 2003; Lundquist et al., 2001) (Wang et al., 2003)

show only a partial defect in engulfment; double mutants defective in both pathways show a more severe, persistent cell corpse phenotype, while the defect in double mutants between genes within the same pathway is only as severe as in the stronger single mutant (Ellis et al., 1991; Gumienny et al., 2001). Interestingly, even after inactivation of both pathways, persistent cell corpses slowly disappear, suggesting that there may be a third salvage pathway that ineYciently removes dead cells throughout the life of the worm.

A. CED-1, CED-6, and CED-7 As has been mentioned, the known engulfment genes in C. elegans fall into two partially redundant gene groups (Fig. 1, Table II). The first engulfment group contains CED-1, CED-6, and CED-7, which mediate early events in the activation of a signal transduction cascade. CED-1 is an integral membrane protein, which potentially functions as a receptor that recognizes

28


dying cells. CED-1 has been suggested to be homologous to the SREC (Zhou et al., 2001b) and the CD91/LRP (Su et al., 2002) scavenger receptors, which have been previously implicated in the engulfment of apoptotic cells in cell culture systems (Ogden et al., 2001); antibody-mediated cross-linking of the CD91/LRP receptor will stimulate uptake of bound erythrocytes (Ogden et al., 2001). Similarly, the only known receptor for engulfment in Drosophila, croquemort/CD36, is a scavenger receptor-family molecule (Franc et al., 1996), suggesting a conserved mechanism for the recognition of apoptotic cells by scavenger receptor family members. Unfortunately, the homology between the extracellular domain of CED-1 and that of SREC and CD91/ LRP is not convincing, and the nature of the CED-1 ligand remains unpredictable. Domain analysis of CED-6 has identified a PTB (phosphotyrosine binding) domain and a PxxP motif (involved in interactions with SH3 domains), supporting a function as an adapter protein. Both CED-1 and CD91/LRP have NPXY motifs that interact with CED-6 and its mouse orthologue, GULP, respectively, suggesting that CED-1 uses the CED-6 adaptor protein to transmit a recognition signal into the engulfing cell (Su et al., 2002). Consistent with this, overexpression of CED-6 can partially rescue the engulfment defect of ced-1 mutants (Liu and Hengartner, 1998). The small adaptor protein CED-6 is thought to homodimerize after activation via a leucine zipper motif to induce engulfment (Su et al., 2000a) by an as yet uncharacterized mechanism. Surprisingly, recent genetic data suggest that ced-1 and ced-6 both act upstream of the CED-10 Rac GTPase (JM Kinchen and MOH, unpublished observations). It is possible that CED-6 signals to downstream proteins via its PxxP motif; however, the interacting partner that utilizes this motif has yet to be identified. CED-7, like CED-1, is also a plasma membrane protein that may play a role in recognition of apoptotic cell corpses. Unlike the six other engulfment genes, which all function in the engulfing cell, ced-7 function is required in both the engulfing and the dying cell for eYcient engulfment to occur (Wu and Horvitz, 1998a). ABCA1, a mammalian CED-7 homologue, has been shown to be involved in the rearrangement of plasma membrane phospholipids (Hamon et al., 2000). ABCA1 has been proposed to promote PS exposure in both the engulfing and the dying cell, mediating an as yet uncharacterized recognition event. Interestingly, ABCA1 is mutated in Tangier’s Disease (Remaley et al., 1999), which is characterized by a defect in cellular eZux of lipids and cholesterol onto nascent high-density lipoproteins. ABCA1 is expressed on both the cell membrane and the membrane of the Golgi complex, where it mediates apo-AI associated export of cholesterol and phospholipids from the cell, and is regulated by cholesterol flux. Mutations in ABCA1 also result in abnormal cavaeolae, suggesting a role for ABCA1 in the generation


29

or maintenance of specific membrane domains, such as lipid rafts (Hamon et al., 2002). Work from the worm supports a role for CED-7 in the maintenance of lipid subdomains; studies show that CED-1::GFP normally localizes to distinct domains on the plasma membrane around the engulfing cell (Zhou et al., 2001b). In a ced-7(n1996) background, which is a genetic null, CED-1::GFP is no longer localized around corpses, suggesting a role for CED-7 in the promotion of CED-1::GFP relocalization around an adjacent apoptotic cell corpse. Indeed, recent studies suggest that PS exposure still occurs in ced-7(lf) mutants (S. Zuellig and MOH, unpublished observations), confirming that CED-7 does not play a primary role in PS exposure in the worm.

B. CED-2, CED-5, CED-10, and CED-12 The second pathway (Table II) in the worm that functions in engulfment is composed of four genes: ced-2/crkII, ced-5/dock180, ced-10/rac1, and ced-12/ elmo. These genes make up a signal transduction pathway that likely leads to the activation of CED-10/Rac1 and the reorganization of the actin cytoskeleton around the dying cell in a largely uncharacterized manner (Gumienny et al., 2001; Reddien and Horvitz, 2000; Wu and Horvitz, 1998b; Wu et al., 2001; Zhou et al., 2001a). The most upstream member of this pathway is CED-2/CrkII (Reddien and Horvitz, 2000). CrkII was first discovered in mammalian cells as a viral oncogene, v-Crk, that induces cell migration and increases metastatic potential (Matsuda and Kurata, 1996). CrkII is essentially an adapter molecule composed of one SH2 domain and two SH3 domains. In mammalian systems, CrkII associates with a protein called p130CAS (Crk-associated substrate) (Sakai et al., 1994), which is highly phosphorylated when activated. p130CAS functions with CrkII as a ‘‘molecular switch’’ (Klemke et al., 1998), transducing signals from a putative transmembrane receptor via interaction of Cas with the CrkII SH2 motif (Hamasaki et al., 1996). No p130CAS homologue has yet been characterized in the C. elegans genome, however, and signaling upstream of CED-2 may be divergent (see following text). The N-terminal CrkII SH3 domain has been shown to bind to Dock180 (Matsuda et al., 1996), through which a signal is transduced to Rac, leading to reorganization of the actin cytoskeleton (Kiyokawa et al., 1998). The order of this pathway has been confirmed in C. elegans, with ced-2/CrkII and ced-5/Dock180 functioning upstream of ced-10/Rac1 (Reddien and Horvitz, 2000; Wu and Horvitz, 1998b). Functional conservation of this signaling complex in engulfment in mammalian cells has also been shown (Albert et al., 2000; Gumienny et al., 2001).

30


CED-5/DOCK180 seems to serve as a scaVold on which many proteins assemble, forming a signaling module that can be recruited to the cell surface to induce ‘‘migration’’ around the apoptotic cell. Dock180 interacts with Rac1 via a basic region at the C-terminus (Kobayashi et al., 2001), and with CED-12/ELMO via a SH3 domain at the N-terminus (Gumienny et al., 2001; Wu et al., 2001; Zhou et al., 2001a). In 2002, a particular domain at the C-terminus of Dock180 was identified, named the DOCKER domain, which specifically binds to nucleotide-free Rac1 and loads GTP onto Rac1 to activate it (Brugnera et al., 2002). A new member of this pathway has been identified in C. elegans (and further characterized in mammalian cells) named ced-12/Elmo (Gumienny et al., 2001; Wu et al., 2001; Zhou et al., 2001a). CED-12/Elmo interacts with CED-5/DOCK180 via a PxxP motif; ELMO co-localizes at the cell membrane with CrkII, DOCK180, and Rac upon induction of membrane ruZing. The molecular function of CED-12/ELMO is still largely unknown; it is a novel protein with no significant homology to any other known protein class; however, it has been suggested that ELMO may promote Rac GTP exchange in conjunction with Dock180 (Brugnera et al., 2002) by stabilizing the Dock180::nt-free Rac transition state (Lu et al., 2004). Cross-species transgenic rescue experiments with CrkII, Dock180, and ELMO have shown that the function of these proteins in cell migration, but not engulfment, is conserved (Gumienny et al., 2001; Wu and Horvitz, 1998b; Wu et al., 2001; Zhou et al., 2001a). Interestingly, another CED-5 homologue, Dock4, is functional in the engulfment of apoptotic cells in worms, but has lost the ability to mediate long-term cell migrations (Yajnik et al., 2003). CED-5, Dock180, and Dock4 are part of a larger family of CDM (ced-5, dock180, and myoblast city) proteins that can be classified into distinct subtypes based on presence or absence of CrkII/ELMO interaction motifs (Cote and Vuori, 2002). CED-5 is an early branch oV the DockA/B lineage, with Dock180 a DockA member and Dock4 a DockB member; thus, perhaps there was an evolutionary divergence, sometime after C. elegans split from the mammalian lineage, separating ability to function in engulfment from function in long-term cell migrations. The receptor that signals to the CED-2, -5, -10, and -12 signaling complex has yet to be discovered. In mammalian cells, integrin receptors (v 3 and v 5) act upstream of Rac activation, with p130Cas bridging the activated integrin receptor and the CrkII adaptor protein (Fig. 4) (Albert et al., 2000); this does not seem to be the case in C. elegans, however, as there is no clear worm Cas homologue. Moreover, C. elegans integrin mutants show little defect in engulfment of apoptotic corpses (Gumienny et al., 2001). Integrin receptor function in engulfment in mammalian tissue culture systems may be a case where increased adhesion to cells also increases the


31

number of cells that are engulfed; further study to ascertain this diVerence is needed.

XIII. Moving the Plasma Membrane—The Small GTPases Rac and RhoG The small GTPase Rac has been implicated in rearrangement of the actin cytoskeleton and remodeling of the plasma membrane in response to extracellular stimuli, a process known as ‘‘membrane ruZing’’ (Ridley et al., 1992). Rac, like other G proteins, cycles between a GDP-bound, inactive conformation and a GTP-bound, active conformation based on molecules that promote GTP hydrolysis (GAPs or GTPase activating proteins), proteins that promote GTP/GDP exchange (GEFs or guanine nucleotide exchange factors), and proteins that keep GDP from dissociating (GDIs or guanine nucleotide dissociation inhibitors) (Schmitz et al., 2000; Scita et al., 2000). The recent discovery that Dock180 (the CED-5 homologue) is a candidate RacGEF (Brugnera et al., 2002) suggests that CED-5 (together with CED-12) acts as the GEF for Rac during engulfment. No GAP or GDI has yet been identified that functions in engulfment of apoptotic cells either in C. elegans or in mammalian cells. However, loss-of-function mutations in GAPs/GDIs may be diYcult to identify since ‘‘overactivation’’ of the engulfment machinery (by ced-10 overexpression, for example) does not show a persistent cell corpse defect, rendering such mutations invisible to traditional screening techniques.

A. CED-10, MIG-2, and RAC-2—Redundant Rac Proteins? While there are three Rac proteins in C. elegans, CED-10/Rac1 appears to be the only one specifically required for engulfment. However, when the engulfment machinery is partially compromised (‘‘sensitized’’ backgrounds), mig2(lf) can increase the number of apoptotic cell corpses (see following text) (Lundquist et al., 2001). ced-10(lf) mutants also have several pleiotropic defects resulting from inappropriate cell migration, such as mismigration of the distal tip cells (DTCs). While it has previously been published that CED-10 functions redundantly with other Rac orthologues during ventral enclosure and certain cell migrations, the isolation of stronger ced-10 alleles suggests a primary role for ced-10 in these processes (Lundquist et al., 2001). ced-10(null) alleles show maternal-eVect embryonic lethality due to defects in enclosure of the embryo, resulting in a Gex (gut on the exterior) phenotype (Soto et al.,

32


2002). Thus, a re-evaluation of the relationship between CED-10 and the other two Rac orthologues in the worm would seem to be necessary. B. MIG-2 and RhoG—Multiple Mechanisms of Engaging the CED-2, -5, -10, -12 Pathway? RhoG is a small Rho-family GTPase that appears to activate Rac and Cdc42 signaling [based on overexpression data utilizing activated RhoG(G12V), which causes phenotypes associated with both Rac and Cdc42 overexpression (Gauthier-Rouviere et al., 1998)]. Normally localized to the perinuclear region, RhoG is transported to the membrane dependent on the microtubule network, and in the presence of nocodazole RhoG is localized to the perinuclear region only, suggesting that upon activation, RhoG actively moves to the cell membrane (Gauthier-Rouviere et al., 1998). The Trio GEF has also been shown to play a role in RhoG activation, inducing similar phenotypes upon overexpression, and is also dependent on microtubule network-mediated locomotion (Blangy et al., 2000). Recently, it has been suggested that the mechanism by which RhoG aVects Rac activation is via a regulatory interaction with CED-12/ELMO (Katoh and Negishi, 2003). Activated RhoG, when co-expressed with ELMO and Dock180, resulted in 2-fold more GTP loading of Rac than ELMO/Dock180 co-expression alone. The authors further suggested that activated RhoG interacts with an ELMO/Dock180 complex, targeting the complex to the membrane downstream of integrin receptor activation. In C. elegans, the MIG-2 Rac orthologue has also been shown to utilize a TrioGEF, unc-73, to mediate eVects on cell migration similar to RhoG, suggesting it may be a candidate regulator of ELMO activity (Lundquist et al., 2001; Steven et al., 1998). mig-2(lf) animals, similar to ced-10(lf), have incompletely penetrant defects in the migration of the distal tip cells (DTCs). When mig-2(lf) mutants are placed into weak engulfment mutant backgrounds, they can enhance the engulfment defect (Lundquist et al., 2001); mig-2(lf), however, does not have any engulfment defect outside of these sensitized backgrounds, and may play a greater role in activation of the CED-5/CED-12/ CED-10 signaling module during migration of the distal tip cells than during the engulfment of apoptotic cells. C. Signaling Downstream of Rac—Novel Effectors? No Rac eVectors have yet been shown to function in engulfment in C. elegans, though several (GEX-2 and GEX-3, UNC-115) have been shown to be involved in migration of certain cell types in a CED-10/Rac–dependent


33

manner (Gitai et al., 2003; Sawa et al., 2003; Soto et al., 2002; StruckhoV and Lundquist, 2003). It is interesting that, though signaling pathways tend to converge at CED-10/Rac, diVerent processes appear to use divergent eVector pathways. Further work will hopefully elucidate what genes are responsible for signaling downstream of CED-10 during engulfment, and the interplay between downstream eVectors of CED-10 involved in generating diverse phenotypes.

XIV. ced-2, -5, -10, and -12 and Distal Tip Cell (DTC) Migration As has been mentioned, ced-2, -5, -10, and -12, unlike ced-1, -6, and -7, have pleiotropic defects in cell migration. Indeed, null alleles of ced-10 have a penetrant embryonic lethality due to defects in cell migration (Lundquist et al., 2001; Soto et al., 2002), though ced-2, ced-5, and ced-12 have only a partial defect in this process. CED-10-mediated migration has been most studied in the Distal Tip Cells (DTCs), largely because their migration path determines the shape of the hermaphrodite gonad, and mismigration results in an obvious change in overall gonad structure (Fig. 3C–F) (Gumienny et al., 2001; Lundquist et al., 2001; Reddien and Horvitz, 2000; Wu and Horvitz, 1998b; Wu et al., 2001; Zhou et al., 2001a). Interestingly, CED-10/ Rac1 is not required for migration of the DTCs per se, but is rather required for appropriate directional migrations in response to extracellular signals, most likely including the netrin family of secreted molecules (Merz et al., 2001; Su et al., 2000b).

XV. Suicide vs Murder—Suggestions of Phagocyte-Mediated Cell Killing Work from several sources suggests that, in worms, cell death is not always solely determined by the core apoptotic machinery (CED-3, -4, -9, and EGL-1), and can instead also be mediated through direct engulfment by neighboring cells.

A. The Case of Linker Cell Death The linker cell determines the shape of the male gonad in much the same way that the DTC determines hermaphrodite gonad morphology (Kimble and White, 1981). However, whereas the DTC persists throughout the lifetime of the hermaphrodite, the linker cell is eYciently removed following its

34


migration, in order to open a path to the male mating apparatus so the sperm can exit (Kimble and Hirsh, 1979). However, it does not appear that the cell undergoes normal programmed cell death, since the linker cell is still removed in ced-3(lf) mutants half of the time (Ellis and Horvitz, 1986). Thus, it seems that a process of cell murder, rather than the ‘‘normal’’ method of cell suicide, may remove the linker cell.

B. Evidence from Weak ced-3 Loss-of-Function Genetics Observations from our laboratory and others have shown that the programmed cell death pathway is not as linear as our models would have us believe (Fig. 1). In weak ced-3 mutants, cells that are not visibly apoptotic can be ‘‘eaten alive’’ in a manner that is dependent on the engulfment machinery (Hoeppner et al., 2001; Reddien et al., 2001). It has also been shown that engulfment mutants can weakly suppress non-apoptotic cell death caused by mutations in let-33 and let-24, suggesting an active role of the phagocyte in cell removal (Kim, 1994). This close interplay between genes involved in apoptosis and genes involved in engulfment suggests that the linear pathway we conveniently use to think about programmed cell death is not entirely correct, and that there is some overlap between what has previously been thought of as two distinct processes. This could be due to early processing of the eat-me signal, such that negative regulatory machinery may be able to stop apoptosis of the cell, but not the exposure of the eat-me signal (Hoeppner et al., 2001). The engulfment machinery would then remove these cells, without the semi-apoptotic cell undergoing the phenotypic stages of apoptosis. Alternately, the eat-me signal may be generated in a caspase-independent manner; thus, while ced-3 activation may be downregulated, the cell may still be eYciently removed. Further characterization of engulfment-mediated cell death should provide a clearer answer.

XVI. Control of DNA Degradation in C. elegans—NUC-1, CPS-6/endoG, and WAH-1/AIF The gene encoding the NUC-1 DNase was first isolated as a mutation that caused ingested bacterial DNA to persist undegraded in the intestine. In this mutant, apoptosis and engulfment proceed normally, but the genomic DNA of engulfed cells is not degraded, and persists even though the corpse is no longer refractile. Wu and colleagues adapted the TUNEL staining technique to visualize the generation of DNA fragments during C. elegans


35

apoptosis (Wu et al., 2000). Normally, this is a very transient state: in wildtype 1.5-fold embryos, only 12% of apoptotic cells are TUNELpositive. Mutations in the core apoptotic machinery eliminate staining, confirming that TUNEL specifically labels apoptotic cells. In nuc-1 (lf) embryos, the number of TUNEL-positive cells is greatly increased over wild-type, as undegraded DNA from both engulfed corpses and unengulfed corpses can be stained. Mutations in nuc-1 have been reported to result in increased cell survival, suggesting that NUC-1 may actively promote death of cells fated to die. This also appears to be true of some of the more recently isolated genes that function in apoptotic DNA degradation (Parrish and Xue, 2003; Parrish et al., 2001, 2003; Wang et al., 2002). Since engulfment mutants contain large numbers of Feulgen-reactive unengulfed corpses (Hedgecock et al., 1983), it has been suggested that the bulk of DNA degradation occurs within the phagosome following engulfment. However, generation and resolution of TUNEL-positive ends occurs with wild-type kinetics in most engulfment mutants, with the exception of ced-1 and ced-7 deficient animals. In these two mutants, there are fewer TUNELpositive apoptotic cells, suggesting that engagement of the CED-1 (and CED-7) engulfment receptor is required for generation of TUNEL-positive ends in an outside-in signaling event. This is very interesting, as it also suggests that the linear programmed cell death pathway is very artificial, as well as showing another safeguard in the regulation of apoptosis. If a cell is to suddenly decide that it doesn’t want to die, it is useful to limit the degradation of DNA to a later timepoint, say, after an engulfment signal is generated, giving the cell a greater amount of time to make the life/death decision. A worm homologue of mammalian endoG was discovered in a screen for mutations that suppress activated caspase-mediated apoptotic cell death; AIF was identified by reverse genetics (Parrish et al., 2001; Wang et al., 2002). These genes both encode mitochondrial proteins that are involved in DNA degradation; interestingly, loss-of-function in these genes has been shown to delay apoptotic cell death, suggesting that some sort of feedback mechanism exists that can limit the onset of apoptosis if the DNA degradation machinery is impaired. The worm AIF homologue, wah-1, has been shown to be localized primarily to mitochondria; in apoptotic cells WAH-1 is released from the mitochondria along with CED-4, allowing it to participate in DNA degradation (Wang et al., 2002). CPS-6/endoG is also localized to the mitochondria, but contains a nuclear localization sequence (NLS) (Parrish et al., 2001). Reverse genetics has been very useful in identifying candidate genes in the corpse degradation machinery. Using the sequenced and annotated C. elegans genome, proteases and nucleases were tested for a role in generating TUNEL-positive bodies during embryonic development (Parrish and Xue,

36


2003). Immunoprecipitation experiments were then used to identify which genes act together to degrade apoptotic DNA, identifying multiple nuclease complexes that degrade DNA following onset of apoptosis.

XVII. Conclusion Analyses of large chromosomal deletions (collectively removing 75% of the genome) have shown that a significant number of genes involved in programmed cell death have yet to be discovered (Sugimoto et al., 2001). Indeed, the pathways that we currently know of are missing key molecules both to receive and transduce signals, as well as the elusive signal produced by the dying cell that induces phagocytosis. The need for new screens to identify these genes is great; one of the problems with such analyses is the proportion of false positives—apparently, genes that are involved in specification of tissue fate result in engulfment defects, suggesting that appropriate tissue specification, or perhaps the cell–cell interactions that these tissues initiate during development, are essential for engulfment of developmental cell corpses (Kodama et al., 2002; Soto et al., 2002). Currently, several labs are using forward and reverse genetic means to isolate and characterize these unknown players. The role of these as yet uncharacterized genes in engulfment and cell migration should prove enlightening. Finally, it has been shown that the linearity of the canonical genetic pathway for apoptosis is actually quite convoluted, with both genes that are responsible for DNA degradation and genes that are essential for engulfment of apoptotic corpses also playing a role in cell killing. The mechanism by which this occurs, and the mechanism by which cells can recover from early phenotypic stages of apoptosis (cell condensation as evidenced by increased refractility under DIC optics, for example), should be of great interest to biomedical science, as ways to regulate apoptosis associated with everything from Alzheimer’s disease to autoimmune disorders would be of great therapeutic interest. Ultimately, genetics, whether forward or reverse, should provide interesting answers to the how, when, where, and why of the mechanism of programmed cell death.

Acknowledgments The authors thank Kodimangalam S. Ravichandran, Linda van Aelst, Greg Hannon, and members of the Hengartner lab, especially Gari Stergiou, for thoughtful discussions. This work was supported by grants from the Swiss National Science Foundation, The Ernst Hadorn Foundation, and the European Union (FP5 project APOCLEAR) to MOH.


37

References Abrams, J. A., White, K., Fessler, L. I., and Steller, H. (1993). Programmed cell death during Drosophila development. Development 117, 29–43. Adams, J. M., and Cory, S. (1998). The Bcl-2 protein family: Arbiters of cell survival. Science 281, 1322–1326. Ahmed, S., Alpi, A., Hengartner, M. O., and Gartner, A. (2001). C. elegans RAD-5/CLK2 defines a new DNA damage checkpoint protein. Curr. Biol. 11, 1934–1944. Ahmed, S., and Hodgkin, J. (2000). MRT-2 checkpoint protein is required for germline immortality and telomere replication in C. elegans. Nature 403, 159–164. Albert, M. L., Kim, J.-I., and Birge, R. B. (2000). Alphav-beta5 integrin recruits the CrkII– Dock180–Rac1 complex for phagocytosis of apoptotic cells. Nat. Cell. Biol. 2, 899–905. Alpi, A., Pasierbek, P., Gartner, A., and Loidl, J. (2003). Genetic and cytological characterization of the recombination protein RAD-51 in Caenorhabditis elegans. Chromosoma 112, 6–16. Arur, S., Uche, U. E., Rezaul, K., Fong, M., Scranton, V., Cowan, A. E., Mohler, W., and Han, D. K. (2003). Annexin I is an endogenous ligand that mediates apoptotic cell engulfment. Dev. Cell 4, 587–598. Austin, J. A., and Kimble, J. E. (1987). glp-1 is required in the germ line for regulation of the decision between mitosis and meiosis in C. elegans. Cell 51, 589–599. Bates, S., and Vousden, K. H. (1996). p53 in signaling checkpoint arrest or apoptosis. Curr. Opin. Genet. Dev. 6, 12–18. Benard, C., McCright, B., Zhang, Y., Felkai, S., Lakowski, B., and Hekimi, S. (2001). The C. elegans maternal-eVect gene clk-2 is essential for embryonic development, encodes a protein homologous to yeast Tel2p, and aVects telomere length. Development 128, 4045–4055. Bergamaschi, D., Samuels, Y., O’Neil, N. J., Trigiante, G., Crook, T., Hsieh, J. K., O’Connor, D. J., Zhong, S., Campargue, I., Tomlinson, M. L., Kuwabara, P. E., and Lu, X. (2003). iASPP oncoprotein is a key inhibitor of p53 conserved from worm to human. Nat. Genet. 33, 162–167. Blangy, A., Vignal, E., Schmidt, S., Debant, A., Gauthier-Rouviere, C., and Fort, P. (2000). TrioGEF1 controls Rac- and Cdc42-dependent cell structures through the direct activation of rhoG. J. Cell Sci. 113(Pt. 4), 729–739. Bloss, T. A., Witze, E. S., and Rothman, J. H. (2003). Suppression of CED-3-independent apoptosis by mitochondrial betaNAC in Caenorhabditis elegans. Nature 424, 1066–1071. Boatright, K. M., and Salvesen, G. S. (2003). Mechanisms of caspase activation. Curr. Opin. Cell Biol. 15, 725–731. Boulton, S. J., Gartner, A., Reboul, J., Vaglio, P., Dyson, N., Hill, D. E., and Vidal, M. (2002). Combined functional genomic maps of the C. elegans DNA damage response. Science 295, 127–131. Boulton, S. J., Martin, J. S., Polanowska, J., Hill, D. E., Gartner, A., and Vidal, M. (2004). BRCA1/BARD1 orthologs required for DNA repair in Caenorhabditis elegans. Curr. Biol. 14, 33–39. Brenner, S. (1974). The genetics of Caenorhabditis elegans. Genetics 77, 71–94. Brown, S., Heinisch, I., Ross, E., Shaw, K., Buckley, C. D., and Savill, J. (2002). Apoptosis disables CD31-mediated cell detachment from phagocytes promoting binding and engulfment. Nature 418, 200–203. Brugnera, E., Haney, L., Grimsley, C., Lu, M., Walk, S. F., Tosello-Trampont, A. C., Macara, I. G., Madhani, H., Fink, G. R., and Ravichandran, K. S. (2002). Unconventional Rac-GEF activity is mediated through the Dock180-ELMO complex. Nat. Cell Biol. 4, 574–582.

38


Chan, S. L., Yee, K. S., Tan, K. M., and Yu, V. C. (2000). The Caenorhabditis elegans sex determination protein FEM-1 is a CED-3 substrate that associates with CED-4 and mediates apoptosis in mammalian cells. J. Biol. Chem. 275, 17925–17928. Chen, F., Hersh, B. M., Conradt, B., Zhou, Z., Riemer, D., Gruenbaum, Y., and Horvitz, H. R. (2000). Translocation of C. elegans CED-4 to nuclear membranes during programmed cell death. Science 287, 1485–1489. Chin, G. M., and Villeneuve, A. M. (2001). C. elegans mre-11 is required for meiotic recombination and DNA repair but is dispensable for the meiotic G(2) DNA damage checkpoint. Genes Dev. 15, 522–534. Conradt, B., and Horvitz, H. R. (1998). The C. elegans protein EGL-1 is required for programmed cell death and interacts with the Bcl-2-like protein CED-9. Cell 93, 519–529. Conradt, B., and Horvitz, H. R. (1999). The TRA-1A sex determination protein of C. elegans regulates sexually dimorphic cell deaths by repressing the egl-1 cell death activator gene. Cell 98, 317–327. Cote, J. F., and Vuori, K. (2002). Identification of an evolutionarily conserved superfamily of DOCK180-related proteins with guanine nucleotide exchange activity. J. Cell Sci. 115, 4901–4913. Cui, P., Qin, B., Liu, N., Pan, G., and Pei, D. (2004). Nuclear localization of the phosphatidylserine receptor protein via multiple nuclear localization signals. Exp. Cell Res. 293, 154–163. Danial, N. N., and Korsmeyer, S. J. (2004). Cell death: Critical control points. Cell 116, 205–219. Degli Esposti, M., and Dive, C. (2003). Mitochondrial membrane permeabilisation by Bax/Bak. Biochem. Biophys. Res. Commun. 304, 455–461. del Peso, L., Gonzalez, V. M., Inohara, N., Ellis, R. E., and Nunez, G. (2000). Disruption of the CED-9.CED-4 complex by EGL-1 is a critical step for programmed cell death in Caenorhabditis elegans. J. Biol. Chem. 275, 27205–27211. del Peso, L., Gonzalez, V. M., and Nunez, G. (1998). Caenorhabditis elegans EGL-1 disrupts the interaction of CED-9 with CED-4 and promotes CED-3 activation. J. Biol. Chem. 273, 33495–33500. Dernburg, A. F., McDonald, K., Moulder, G., Barstead, R., Dresser, M., and Villeneuve, A. M. (1998). Meiotic recombination in C. elegans initiates by a conserved mechanism and is dispensable for homologous chromosome synapsis. Cell 94, 387–398. Derry, W. B., Putzke, A. P., and Rothman, J. H. (2001). Caenorhabditis elegans p53: Role in apoptosis, meiosis, and stress resistance. Science 294, 591–595. Desai, C., Garriga, G., McIntire, S. L., and Horvitz, H. R. (1988). A genetic pathway for the development of the Caenorhabditis elegans HSN motor neurons. Nature 336, 638–646. Desai, C., and Horvitz, H. R. (1989). Caenorhabditis elegans mutants defective in the functioning of the motor neurons responsible for egg laying. Genetics 121, 703–721. Driscoll, M., and Chalfie, M. (1991). The mec-4 gene is a member of a family of Caenorhabditis elegans genes that can mutate to induce neuronal degeneration. Nature 349, 588–593. Ellis, H. M., and Horvitz, H. R. (1986). Genetic control of programmed cell death in the nematode C. elegans. Cell 44, 817–829. Ellis, R. E., and Horvitz, H. R. (1991). Two C. elegans genes control the programmed deaths of specific cells in the pharynx. Development 112, 591–603. Ellis, R. E., Jacobson, D. M., and Horvitz, H. R. (1991). Genes required for the engulfment of cell corpses during programmed cell death in Caenorhabditis elegans. Genetics 129, 79–94. Faber, P. W., Alter, J. R., MacDonald, M. E., and Hart, A. C. (1999). Polyglutamine-mediated dysfunction and apoptotic death of a Caenorhabditis elegans sensory neuron. Proc. Natl. Acad. Sci. USA 96, 179–184.


39

Faber, P. W., Voisine, C., King, D. C., Bates, E. A., and Hart, A. C. (2002). Glutamine/prolinerich PQE-1 proteins protect Caenorhabditis elegans neurons from huntingtin polyglutamine neurotoxicity. Proc. Natl. Acad. Sci. USA 99, 17131–17136. Fadok, V. A., Bratton, D. L., Frasch, S. C., Warner, M. L., and Henson, P. M. (1998). The role of phosphatidylserine in recognition of apoptotic cells by phagocytes. Cell Death DiVer. 5, 551–562. Fadok, V. A., Bratton, D. L., Rose, D. M., Pearson, A., Ezekewitz, R. A., and Henson, P. M. (2000). A receptor for phosphatidylserine-specific clearance of apoptotic cells. Nature 405, 85–90. Fadok, V. A., de Cathelineau, A., Daleke, D. L., Henson, P. M., and Bratton, D. L. (2001). Loss of phospholipid asymmetry and surface exposure of phosphatidylserine is required for phagocytosis of apoptotic cells by macrophages and fibroblasts. J. Biol. Chem. 276, 1071–1077. Fischer, U., Janicke, R. U., and Schulze-OsthoV, K. (2003). Many cuts to ruin: A comprehensive update of caspase substrates. Cell Death DiVer. 10, 76–100. Franc, N. C., Dimarcq, J. L., Lagueux, M., HoVmann, J., and Ezekowitz, R. A. (1996). Croquemort, a novel Drosophila hemocyte/macrophage receptor that recognizes apoptotic cells. Immunity 4, 431–443. Fraser, A. G., James, C., Evan, G. I., and Hengartner, M. O. (1999). Caenorhabditis elegans inhibitor of apoptosis (IAP) homologue BIR-1 plays a conserved role in cytokinesis. Curr. Biol. 9, 292–301. Gartner, A., Milstein, S., Ahmed, S., Hodgkin, J., and Hengartner, M. O. (2000). A conserved checkpoint pathway mediates DNA damage-induced apoptosis and cell cycle arrest in C. elegans. Mol. Cell 5, 435–443. Gauthier-Rouviere, C., Vignal, E., Meriane, M., Roux, P., Montcourier, P., and Fort, P. (1998). RhoG GTPase controls a pathway that independently activates Rac1 and Cdc42Hs. Mol. Biol. Cell 9, 1379–1394. Gitai, Z., Yu, T. W., Lundquist, E. A., Tessier-Lavigne, M., and Bargmann, C. I. (2003). The netrin receptor UNC-40/DCC stimulates axon attraction and outgrowth through enabled and, in parallel, Rac and UNC-115/AbLIM. Neuron 37, 53–65. Goedert, M. (2003). Neurodegenerative tauopathy in the worm. Proc. Natl. Acad. Sci. USA 100, 9653–9655. Goodwin, E. B., and Ellis, R. E. (2002). Turning clustering loops: Sex determination in Caenorhabditis elegans. Curr. Biol. 12, R111–R120. Gumienny, T. L., Brugnera, E., Tosello-Trampont, A. C., Kinchen, J. M., Haney, L. B., Nishiwaki, K., Walk, S. F., Nemergut, M. E., Macara, I. G., Francis, R., Schedl, T., Qin, Y., Van Aelst, L., Hengartner, M. O., and Ravichandran, K. S. (2001). CED-12/ELMO, a novel member of the CrkII/Dock180/Rac pathway, is required for phagocytosis and cell migration. Cell 107, 27–41. Gumienny, T. L., Lambie, E., Hartwieg, E., Horvitz, H. R., and Hengartner, M. O. (1999). Genetic control of programmed cell death in the Caenorhabditis elegans hermaphrodite germline. Development 126, 1011–1022. Hall, D. H., Gu, G., Garcia-Anoveros, J., Gong, L., Chalfie, M., and Driscoll, M. (1997). Neuropathology of degenerative cell death in Caenorhabditis elegans. J. Neurosci. 17, 1033–1045. Hamasaki, K., Mimura, T., Morino, N., Furuya, H., Nakamoto, T., Aizawa, S., Morimoto, C., Yazaki, Y., Hirai, H., and Nojima, Y. (1996). Src kinase plays an essential role in integrinmediated tyrosine phosphorylation of Crk-associated substrate p130Cas. Biochem. Biophys. Res. Commun. 222, 338–343. Hamon, Y., Broccardo, C., Chambenoit, O., Luciani, M. F., Toti, F., Chaslin, S., Freyssinet, J. M., Devaux, P. F., McNeish, J., Marguet, D., and Chimini, G. (2000). ABC1 promotes

40


engulfment of apoptotic cells and transbilayer redistribution of phosphatidylserine. Nat. Cell Biol. 2, 399–406. Hamon, Y., Chambenoit, O., and Chimini, G. (2002). ABCA1 and the engulfment of apoptotic cells. Biochim. Biophys. Acta 1585, 64–71. Hanayama, R., Tanaka, M., Miwa, K., Shinohara, A., Iwamatsu, A., and Nagata, S. (2002). Identification of a factor that links apoptotic cells to phagocytes. Nature 417, 182–187. Hartman, P. S., and Herman, R. K. (1982). Radiation-sensitive mutants of C. elegans. Genetics 102, 159–178. Hausmann, G., O’Reilly, L. A., van Driel, R., Beaumont, J. G., Strasser, A., Adams, J. M., and Huang, D. C. (2000). Pro-apoptotic apoptosis protease-activating factor 1 (Apaf-1) has a cytoplasmic localization distinct from Bcl-2 or Bcl-x(L). J. Cell Biol. 149, 623–634. Hedgecock, E. M., Sulston, J. E., and Thomson, J. N. (1983). Mutations aVecting programmed cell deaths in the nematode Caenorhabditis elegans. Science 220, 1277–1279. Hengartner, M. O. (2000). The biochemistry of apoptosis. Nature 407, 770–776. Hengartner, M. O., Ellis, R. E., and Horvitz, H. R. (1992). C. elegans gene ced-9 protects cells from programmed cell death. Nature 356, 494–499. Hengartner, M. O., and Horvitz, H. R. (1994a). Activation of C. elegans cell death protein CED-9 by an amino-acid substitution in a domain conserved in Bcl-2. Nature 369, 318–320. Hengartner, M. O., and Horvitz, H. R. (1994b). C. elegans cell death gene ced-9 encodes a functional homolog of mammalian proto-oncogene bcl-2. Cell 76, 665–676. Hockenbery, D., Nun˜ ez, G., Milliman, C., Schreiber, R. D., and Korsmeyer, S. J. (1990). Bcl2 is an inner mitochondrial membrane protein that blocks programmed cell death. Nature 348, 334–336. Hoeppner, D. J., Hengartner, M. O., and Schnabel, R. (2001). Engulfment genes cooperate with ced-3 to promote cell death in Caenorhabditis elegans. Nature 412, 202–206. Hofmann, E. R., Milstein, S., Boulton, S. J., Ye, M., Hofmann, J. J., Stergiou, L., Gartner, A., Vidal, M., and Hengartner, M. O. (2002). Caenorhabditis elegans HUS-1 is a DNA damage checkpoint protein required for genome stability and EGL-1-mediated apoptosis. Curr. Biol. 12, 1908–1918. Hofmann, E. R., Milstein, S., and Hengartner, M. O. (2000). DNA-damage-induced checkpoint pathways in the nematode Caenorhabditis elegans. Cold Spring Harb. Symp. Quant. Biol. 65, 467–473. Howard, R. M., and Sundaram, M. V. (2002). C. elegans EOR-1/PLZF and EOR-2 positively regulate Ras and Wnt signaling and function redundantly with LIN-25 and the SUR-2 Mediator component. Genes Dev. 16, 1815–1827. Irmler, M., Hofmann, K., Vaux, D., and Tschopp, J. (1997). Direct physical interaction between the Caenorhabditis elegans ‘‘death proteins’’ CED-3 and CED-4. FEBS Lett. 406, 189–190. Ishimoto, Y., Ohashi, K., Mizuno, K., and Nakano, T. (2000). Promotion of the uptake of PS liposomes and apoptotic cells by a product of growth arrest-specific gene, gas6. J. Biochem. (Tokyo) 127, 411–417. Jiang, N., Benard, C. Y., Kebir, H., Shoubridge, E. A., and Hekimi, S. (2003). Human CLK2 links cell cycle progression, apoptosis, and telomere length regulation. J. Biol. Chem. 278, 21678–21684. Karashima, T., Sugimoto, A., and Yamamoto, M. (2000). Caenorhabditis elegans homologue of the human azoospermia factor DAZ is required for oogenesis but not for spermatogenesis. Development 127, 1069–1079. Katoh, H., and Negishi, M. (2003). RhoG activates Rac1 by direct interaction with the Dock180-binding protein Elmo. Nature 424, 461–464. Kerr, J. F., Wyllie, A. H., and Currie, A. R. (1972). Apoptosis: A basic biological phenomenon with wide-ranging implications in tissue kinetics. Br. J. Cancer 26, 239–257.


41

Kim, S. (1994). Two C. elegans genes that can mutate to cause degenerative cell death. In ‘‘Department of Biology,’’ PhD Thesis, MIT. Cambridge, MA. Kimble, J., and Hirsh, D. (1979). The postembryonic cell lineages of the hermaphrodite and male gonads in Caenorhabditis elegans. Dev. Biol. 70, 396–417. Kimble, J. E., and White, J. G. (1981). On the control of germ cell development in Caenorhabditis elegans. Dev. Biol. 81, 208–219. Kiyokawa, E., Hashimoto, Y., Kobayashi, S., Sugimura, H., Kurata, T., and Matsuda, M. (1998). Activation of Rac1 by a Crk SH3-binding protein, DOCK180. Genes Dev. 12, 3331–3336. Klemke, R. L., Leng, J., Molander, R., Brooks, P. C., Vuori, K., and Cheresh, D. A. (1998). CAS/Crk coupling serves as a ‘‘molecular switch’’ for induction of cell migration. J. Cell Biol. 140, 961–972. Kobayashi, S., Shirai, T., Kiyokawa, E., Mochizuki, N., Matsuda, M., and Fukui, Y. (2001). Membrane recruitment of DOCK180 by binding to PtdIns(3,4,5)P3. Biochem. J. 354, 73–78. Kodama, Y., Rothman, J. H., Sugimoto, A., and Yamamoto, M. (2002). The stem-loop binding protein CDL-1 is required for chromosome condensation, progression of cell death, and morphogenesis in Caenorhabditis elegans. Development 129, 187–196. Korb, L. C., and Ahearn, J. M. (1997). C1q binds directly and specifically to surface blebs of apoptotic human keratinocytes: Complement deficiency and systemic lupus erythematosus revisited. J. Immunol. 158, 4525–4528. Li, M. O., Sarkisian, M. R., Mehal, W. Z., Rakic, P., and Flavell, R. A. (2003). Phosphatidylserine receptor is required for clearance of apoptotic cells. Science 302, 1560–1563. Li, P., Allen, H., Banerjee, S., Franklin, S., Herzog, L., Johnston, C., McDowell, J., Paskind, M., Rodman, L., Salfeld, J., et al. (1995). Mice deficient in IL-1 beta-converting enzyme are defective in production of mature IL-1 beta and resistant to endotoxic shock. Cell 80, 401–411. Lim, C. S., Mian, I. S., Dernburg, A. F., and Campisi, J. (2001). C. elegans clk-2, a gene that limits life span, encodes a telomere length regulator similar to yeast telomere binding protein Tel2p. Curr. Biol. 11, 1706–1710. Liu, Q. A., and Hengartner, M. O. (1998). Candidate adaptor protein CED-6 promotes the engulfment of apoptotic cells in C. elegans. Cell 93, 961–972. Lu, M., Kinchen, J. M., Rossman, K., Grimsley, C., Brugnera, E., Haney, L. B., deBakker, C., Klingele, D., Sondek, J., Hengartner, M., and Ravichandran, K. S. (2004). PH domain of ELMO functions in trans to regulate Rac activation via Dock180. Nat. Struct. Mol. Biol. 8, 756–762. Lundquist, E. A., Reddien, P. W., Hartwieg, E., Horvitz, H. R., and Bargmann, C. I. (2001). Three C. elegans Rac proteins and several alternative Rac regulators control axon guidance, cell migration, and apoptotic cell phagocytosis. Development 128, 4475–4488. Matsuda, M., and Kurata, T. (1996). Emerging components of the Crk oncogene product: The first identified adaptor protein. Cell Signal 8, 335–340. Matsuda, M., Ota, S., Tanimura, R., Nakamura, H., Matuoka, K., Takenawa, T., Nagashima, K., and Kurata, T. (1996). Interaction between the amino-terminal SH3 domain of CRK and its natural target proteins. J. Biol. Chem. 271, 14468–14472. Merz, D. C., Zheng, H., Killeen, M. T., Krizus, A., and Culotti, J. G. (2001). Multiple signaling mechanisms of the UNC-6/netrin receptors UNC-5 and UNC-40/DCC in vivo. Genetics 158, 1071–1080. Metzstein, M. M., Hengartner, M. O., Tsung, N., Ellis, R. E., and Horvitz, H. R. (1996). Transcriptional regulator of programmed cell death encoded by Caenorhabiditis elegans gene ces-2. Nature 382, 545–547.

42


Metzstein, M. M., and Horvitz, H. R. (1999). The C. elegans cell death specification gene ces-1 encodes a snail family zinc finger protein. Mol. Cell 4, 309–319. Metzstein, M. M., Stanfield, G. M., and Horvitz, H. R. (1998). Genetics of programmed cell death in C. elegans: Past, present, and future. Trends Genet. 14, 410–416. Mevorach, D. (1999). The immune response to apoptotic cells. Ann. N. Y. Acad. Sci. 887, 191–198. Morgan, S. E., and Kastan, M. B. (1997). p53 and ATM: Cell cycle, cell death, and cancer. Adv. Cancer Res. 71, 1–25. Nass, R., Miller, D. M., and Blakely, R. D. (2001). C. elegans: A novel pharmacogenetic model to study Parkinson’s disease 7, pp. 185–191. Navarro, R. E., Shim, E. Y., Kohara, Y., Singson, A., and Blackwell, T. K. (2001). cgh-1, a conserved predicted RNA helicase required for gametogenesis and protection from physiological germline apoptosis in C. elegans. Development 128, 3221–3232. Navratil, J. S., Watkins, S. C., Wisnieski, J. J., and Ahearn, J. M. (2001). The globular heads of C1q specifically recognize surface blebs of apoptotic vascular endothelial cells. J. Immunol. 166, 3231–3239. Nishiwaki, K. (1999). Mutations aVecting symmetrical migration of distal tip cells in Caenorhabditis elegans. Genetics 152, 985–997. Ogden, C. A., deCathelineau, A., HoVmann, P. R., Bratton, D., Ghebrehiwet, B., Fadok, V. A., and Henson, P. M. (2001). C1q and mannose binding lectin engagement of cell surface calreticulin and CD91 initiates macropinocytosis and uptake of apoptotic cells. J. Exp. Med. 194, 781–795. Oshima, K., Aoki, N., Kato, T., Kitajima, K., and Matsuda, T. (2002). Secretion of a peripheral membrane protein, MFG-E8, as a complex with membrane vesicles. Eur. J. Biochem. 269, 1209–1218. Parrish, J., Li, L., Klotz, K., Ledwich, D., Wang, X., and Xue, D. (2001). Mitochondrial endonuclease G is important for apoptosis in C. elegans. Nature 412, 90–94. Parrish, J. Z., and Xue, D. (2003). Functional genomic analysis of apoptotic DNA degradation in C. elegans. Mol. Cell 11, 987–996. Parrish, J. Z., Yang, C., Shen, B., and Xue, D. (2003). CRN-1, a Caenorhabditis elegans FEN-1 homologue, cooperates with CPS-6/EndoG to promote apoptotic DNA degradation. EMBO J. 22, 3451–3460. Petros, A. M., Olejniczak, E. T., and Fesik, S. W. (2004). Structural biology of the Bcl-2 family of proteins. Biochim. Biophys. Acta 1644, 83–94. Reddien, P. W., Cameron, S., and Horvitz, H. R. (2001). Phagocytosis promotes programmed cell death in C. elegans. Nature 412, 198–202. Reddien, P. W., and Horvitz, H. R. (2000). CED-2/CrkII and CED-10/Rac control phagocytosis and cell migration in Caenorhabditis elegans. Nat. Cell Biol. 2, 131–136. Remaley, A. T., Rust, S., Rosier, M., Knapper, C., Naudin, L., Broccardo, C., Peterson, K. M., Koch, C., Arnould, I., Prades, C., Duverger, N., Funke, H., Assman, G., Dinger, M., Dean, M., Chimini, G., Santamarina-Fojo, S., Fredrickson, D. S., Denefle, P., and Brewer, H. B., Jr. (1999). Human ATP-binding cassette transporter 1 (ABC1): Genomic organization and identification of the genetic defect in the original tangier disease kindred. Proc. Natl. Acad. Sci. USA 96, 12685–12690. Ridley, A. J., Paterson, H. F., Johnston, C. L., Diekmann, D., and Hall, A. (1992). The small GTPbinding protein rac regulates growth factor-induced membrane ruZing. Cell 70, 401–410. Robertson, A., and Thomson, N. (1982). Morphology of programmed cell death in the ventral nerve cord of Caenorhabditis elegans larvae. J. Embryol. Exp. Morph. 67, 89–100. Rocheleau, C. E., Howard, R. M., Goldman, A. P., Volk, M. L., Girard, L. J., and Sundaram, M. V. (2002). A lin-45 raf enhancer screen identifies eor-1, eor-2, and unusual alleles of Ras pathway genes in Caenorhabditis elegans. Genetics 161, 121–131.


43

Rogalski, T. M., Moerman, D. G., and Baillie, D. L. (1982). Essential genes and deficiencies in the unc-22 IV region of Caenorhabditis elegans. Genetics 102, 725–736. Sakai, R., Iwamatsu, A., Hirano, N., Ogawa, S., Tanaka, T., Mano, H., Yazaki, Y., and Hirai, H. (1994). A novel signaling molecule, p130, forms stable complexes in vivo with v-Crk and v-Src in a tyrosine phosphorylation-dependent manner. EMBO J. 13, 3748–3756. Sawa, M., Suetsugu, S., Sugimoto, A., Miki, H., Yamamoto, M., and Takenawa, T. (2003). Essential role of the C. elegans Arp2/3 complex in cell migration during ventral enclosure. J. Cell Sci. 116, 1505–1518. Schlegel, R. A., Callahan, M., Krahling, S., Pradhan, D., and Williamson, P. (1996). Mechanisms for recognition and phagocytosis of apoptotic lymphocytes by macrophages. Adv. Exp. Med. Biol. 406, 21–28. Schmitz, A. A., Govek, E. E., Bottner, B., and Van Aelst, L. (2000). Rho GTPases: Signaling, migration, and invasion. Exp. Cell Res. 261, 1–12. Schumacher, B., Hofmann, K., Boulton, S., and Gartner, A. (2001). The C. elegans homolog of the p53 tumor suppressor is required for DNA damage-induced apoptosis. Curr. Biol. 11, 1722–1727. Scita, G., Tenca, P., Frittoli, E., Tocchetti, A., Innocenti, M., Giardina, G., and Di Fiore, P. P. (2000). Signaling from Ras to Rac and beyond: Not just a matter of GEFs. EMBO J. 19, 2393–2398. Scott, R. S., McMahon, E. J., Pop, S. M., Reap, E. A., Caricchio, R., Cohen, P. L., Earp, H. S., and Matsushima, G. K. (2001). Phagocytosis and clearance of apoptotic cells is mediated by MER. Nature 411, 207–211. Seshagiri, S., and Miller, L. K. (1997). Caenorhabditis elegans CED-4 stimulates CED-3 processing and CED-3-induced apoptosis. Curr. Biol. 7, 455–460. Shaham, S. (1998). Identification of multiple Caenorhabditis elegans caspases and their potential roles in proteolytic cascades. J. Biol. Chem. 273, 35109–35117. Shaham, S. (2003). Apoptosis. A process with a (beta)NAC for complexity. Cell 114, 659–661. Shaham, S., and Horvitz, H. R. (1996). An alternatively spliced C. elegans ced-4 RNA encodes a novel cell death inhibitor. Cell 86, 201–208. Soto, M. C., Qadota, H., Kasuya, K., Inoue, M., Tsuboi, D., Mello, C. C., and Kaibuchi, K. (2002). The GEX-2 and GEX-3 proteins are required for tissue morphogenesis and cell migrations in C. elegans. Genes Dev. 16, 620–632. Spector, M. S., Desnoyers, S., Hoeppner, D. J., and Hengartner, M. O. (1997). Interaction between the C. elegans cell-death regulators CED-9 and CED-4. Nature 385, 653–656. Speliotes, E. K., Uren, A., Vaux, D., and Horvitz, H. R. (2000). The survivin-like C. elegans BIR-1 protein acts with the Aurora-like kinase AIR-2 to aVect chromosomes and the spindle midzone. Mol. Cell 6, 211–223. Stanfield, G. M., and Horvitz, H. R. (2000). The ced-8 gene controls the timing of programmed cell deaths in C. elegans. Mol. Cell 5, 423–433. Stergiou, L., and Hengartner, M. O. (2004). Death and more: DNA damage response pathways in the nematode C. elegans. Cell Death DiVer. 11, 21–28. Steven, R., Kubiseski, T. J., Zheng, H., Kulkarni, S., Mancillas, J., Ruiz Morales, A., Hogue, C. W., Pawson, T., and Culotti, J. (1998). UNC-73 activates the Rac GTPase and is required for cell and growth cone migrations in C. elegans. Cell 92, 785–795. StruckhoV, E. C., and Lundquist, E. A. (2003). The actin-binding protein UNC-115 is an eVector of Rac signaling during axon pathfinding in C. elegans. Development 130, 693–704. Su, H. P., Brugnera, E., Van Criekinge, W., Smits, E., Hengartner, M., Bogaert, T., and Ravichandran, K. S. (2000a). Identification and characterization of a dimerization domain in CED-6, an adapter protein involved in engulfment of apoptotic cells. J. Biol. Chem. 275, 9542–9549.

44


Su, H. P., Nakada-Tsukui, K., Tosello-Trampont, A. C., Li, Y., Bu, G., Henson, P. M., and Ravichandran, K. S. (2002). Interaction of CED-6/GULP, an adapter protein involved in engulfment of apoptotic cells with CED-1 and CD91/low density lipoprotein receptor-related protein (LRP). J. Biol. Chem. 277, 11772–11779. Su, M., Merz, D. C., Killeen, M. T., Zhou, Y., Zheng, H., Kramer, J. M., Hedgecock, E. M., and Culotti, J. G. (2000b). Regulation of the UNC-5 netrin receptor initiates the first reorientation of migrating distal tip cells in Caenorhabditis elegans. Development 127, 585–594. Subramaniam, K., and Seydoux, G. (1999). nos-1 and nos-2, two genes related to Drosophila nanos, regulate primordial germ cell development and survival in Caenorhabditis elegans. Development 126, 4861–4871. Sugimoto, A., Kusano, A., Hozak, R. R., Derry, W. B., Zhu, J., and Rothman, J. H. (2001). Many genomic regions are required for normal embryonic programmed cell death in Caenorhabditis elegans. Genetics 158, 237–252. Sulston, J. (1988). Cell lineage. In ‘‘The Nematode Caenorhabditis elegans’’ (W. B. Wood, Ed.), pp. 123–155. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Sulston, J. E., Albertson, D. G., and Thomson, J. N. (1980). The Caenorhabditis elegans male: Postembryonic development of nongonadal structures. Dev. Biol. 78, 542–576. Sulston, J. E., and Horvitz, H. R. (1977). Post-embryonic 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. Tan, E. M. (1994). Meaning of autoantibodies in lupus-like syndromes. Lupus 3, 483–485. Thellmann, M., Hatzold, J., and Conradt, B. (2003). The snail-like CES-1 protein of C. elegans can block the expression of the BH3-only cell-death activator gene egl-1 by antagonizing the function of bHLH proteins. Development 130, 4057–4071. van den Eijnde, S. M., van den HoV, M. J., Reutelingsperger, C. P., van Heerde, W. L., Henfling, M. E., Vermeij-Keers, C., Schutte, B., Borgers, M., and Ramaekers, F. C. (2001). Transient expression of phosphatidylserine at cell–cell contact areas is required for myotube formation. J. Cell Sci. 114, 3631–3642. Vaux, D. L., Weissman, I. L., and Kim, S. K. (1992). Prevention of programmed cell death in Caenorhabditis elegans by human bcl-2. Science 258, 1955–1957. Wang, X., Wu, Y. C., Fadok, V. A., Lee, M. C., Gengyo-Ando, K., cheng, L. C., Ledwich, D., Hsu, P. K., Chen, J. Y., Chou, B. K., Henson, P., Mitani, S., and Xue, D. (2003). Cell corpse engulfment mediated by C. elegans phosphatidylserine receptor through CED-5 and CED12. Science 302, 1563–1566. Wang, X., Yang, C., Chai, J., Shi, Y., and Xue, D. (2002). Mechanisms of AIF-mediated apoptotic DNA degradation in Caenorhabditis elegans. Science 298, 1587–1592. Woo, J. S., Jung, J. S., Ha, N. C., Shin, J., Kim, K. H., Lee, W., and Oh, B. H. (2003). Unique structural features of a BCL-2 family protein CED-9 and biophysical characterization of CED-9/EGL-1 interactions. Cell Death DiVer. 10, 1310–1319. Wu, D., Wallen, H. D., Inohara, N., and Nunez, G. (1997). Interaction and regulation of the Caenorhabditis elegans death protease CED-3 by CED-4 and CED-9. J. Biol. Chem. 21449–21454. Wu, Y.-C., and Horvitz, H. R. (1998a). The C. elegans cell-corpse engulfment gene ced-7 encodes a protein similar to ABC transporters. Cell 93, 951–960. Wu, Y.-C., and Horvitz, H. R. (1998b). C. elegans phagocytosis and cell-migration protein CED-5 is similar to human DOCK180. Nature 392, 501–504. Wu, Y. C., Stanfield, G. M., and Horvitz, H. R. (2000). NUC-1, a Caenorhabditis elegans DNase II homolog, functions in an intermediate step of DNA degradation during apoptosis. Genes Dev. 14, 536–548.


45

Wu, Y. C., Tsai, M. C., Cheng, L. C., Chou, C. J., and Weng, N. Y. (2001). C. elegans CED-12 acts in the conserved crkII/DOCK180/Rac pathway to control cell migration and cell corpse engulfment. Dev. Cell 1, 491–502. Xu, K., Tavernarakis, N., and Driscoll, M. (2001). Necrotic cell death in C. elegans requires the function of calreticulin and regulators of Ca(2+) release from the endoplasmic reticulum. Neuron 31, 957–971. Xue, D., and Horvitz, H. R. (1997). Caenorhabditis elegans CED-9 protein is a bifunctional celldeath inhibitor. Nature 390, 305–308. Yajnik, V., Paulding, C., Sordella, R., McClatchey, A. I., Saito, M., Wahrer, D. C., Reynolds, P., Bell, D. W., Lake, R., van den Heuvel, S., Settleman, J., and Haber, D. A. (2003). DOCK4, a GTPase activator, is disrupted during tumorigenesis. Cell 112, 673–684. Yang, E., Zha, J., Jockel, J., Boise, L. H., Thompson, C. B., and Korsmeyer, S. J. (1995). Bad, a heterodimeric partner for Bcl-xL and Bcl-2, displaces Bax and promotes cell death. Cell 80, 285–291. Yang, X., Chang, H. Y., and Baltimore, D. (1998). Essential role of CED-4 oligomerization in CED-3 activation and apoptosis. Science 281, 1355–1357. Yuan, J., and Horvitz, H. R. (1992). The Caenorhabditis elegans cell death gene ced-4 encodes a novel protein and is expressed during the period of extensive programmed cell death. Development 116, 309–320. Yuan, J., Shaham, S., Ledoux, S., Ellis, H. M., and Horvitz, H. R. (1993). The C. elegans cell death gene ced-3 encodes a protein similar to mammalian interleukin-1beta converting enzyme. Cell 75, 641–652. Zhang, Y., and Xiong, Y. (2001). Control of p53 ubiquitination and nuclear export by MDM2 and ARF. Cell Growth DiVer. 12, 175–186. Zhou, B. B., and Elledge, S. J. (2000). The DNA damage response: Putting checkpoints in perspective. Nature 408, 433–439. Zhou, Z., Caron, E., Hartwieg, E., Hall, A., and Horvitz, H. R. (2001a). The C. elegans PH domain protein CED-12 regulates cytoskeletal reorganization via a Rho/Rac GTPase signaling pathway. Dev. Cell 1, 477–489. Zhou, Z., Hartwieg, E., and Horvitz, H. R. (2001b). CED-1 is a transmembrane receptor that mediates cell corpse engulfment in C. elegans. Cell 104, 43–56. Zou, H., Henzel, W. J., Liu, X., Lutschg, A., and Wang, X. (1997a). Apaf-1, a human protein homologous to C. elegans CED-4, participates in cytochrome c-dependent activation of caspase-3. Cell 90, 405–413. Zou, H., Henzel, W. J., Liu, X., Lutschg, A., and Wang, X. (1997b). Apaf-1, a human protein homologous to C. elegans CED-4, participates in cytochrome c-dependent activation of caspase-3. Cell 90, 405–413.

2

____________________________________________________________________________

From Guts to Brains: Using Zebrafish Genetics to Understand the Innards of Organogenesis Carsten Stuckenholz,* Paul E. Ulanch,* and Nathan Bahary *,{ *Department of Molecular Genetics and Biochemistry {

Department of Medicine University of Pittsburgh School of Medicine Pittsburgh, Pennsylvania 15261

I. Introduction A. Forward and Reverse Genetics B. Zebrafish as a Vertebrate Model of Organogenesis II. Forward and Reverse Genetics in Zebrafish A. Forward Genetic Screens in Zebrafish B. Reverse Genetics: Controlling Gene Expression in Zebrafish C. Conclusions III. Zebrafish as a Tool to Investigate Neurodevelopment A. Neural Tube Formation B. Neural Induction C. Conclusions IV. Gastrointestinal Development in Zebrafish A. Endoderm Specification B. Embryonic Gastrointestinal Development in the Zebrafish V. Conclusions and Future Directions Acknowledgments References

Overview Zebrafish have been used to analyze embryonic development and organogenesis. In particular, large-scale forward genetic screens demonstrated that an entire vertebrate genome can be analyzed using random mutagenesis. The isolated mutants have already significantly contributed to our understanding of early vertebrate development and organogenesis. As new and innovative screens are carried out and many of the previously isolated mutants are further analyzed, a comprehensive picture of development and organogenesis will emerge. Zebrafish are uniquely capable of being utilized to also study genes using a host of reverse genetic approaches. Gene overexpression and knockdown techniques are routinely used. The gastrointestinal system and Current Topics in Developmental Biology, Vol. 65 Copyright 2005, Elsevier Inc. All rights reserved. 0070-2153/05 $35.00

47

48

Stuckenholz et al.

the brain are used to illustrate the principles of forward and reverse genetics and the ease with which the zebrafish can be manipulated. Because of the many features that make the zebrafish so attractive to study organogenesis, its importance in unraveling pathways and genes involved in organ development will undoubtedly grow.

I. Introduction As several complete animal genomes have been sequenced in their entirety, with more soon to come, the task of assigning function to the thousands of identified genes and uncovering their role in development, organogenesis in particular, has only grown in importance. There is increasing appreciation that developmental pathways involved in organogenesis are often misregulated in human disease and cancer. For examples, MEF2A has been implicated in coronary artery disease and myocardial infarction (Wang et al., 2003), but is also important for muscle development (reviewed in Black and Olson, 1998). The EGF pathway, one of the key pathways regulating growth and diVerentiation in developing animals, is often misregulated in cancers (Holbro and Hynes, 2004; Yarden, 2001). Analysis of the medulloblastoma transcriptome revealed that the most aggressive cancers have a transcriptional profile reminiscent of early developing brains (Kho et al., 2004). Unraveling the functional pathways operating during organogenesis and repair is therefore vital to subsequently providing targets for the clinical investigation into novel treatments for human diseases.

A. Forward and Reverse Genetics Forward genetic screens have been successfully employed in the past to unravel complex developmental processes. The genetically tractable organisms Caenorhabditis elegans, Drosophila melanogaster, and Arabidopsis thaliana provide examples of using genetics to learn about genes and pathways involved in pattern and axis formation (Ferguson and Horvitz, 1985; Mayer et al., 1991; Nu¨ sslein-Volhard and Wieschaus, 1980). Forward genetic screens identify genes solely by their mutant phenotype, requiring no genespecific or molecular information for their isolation (Fig. 1). Once a mutant animal is identified, the underlying molecular lesion is cloned and the isolated gene is studied further through reverse genetics, including biochemical and molecular methods (Fig. 1). In the past, unicellular, invertebrate, animal, and plant models have been most amenable to forward genetic screens because these models have small genomes, are mutable to saturation, and are small and cost-eVective to maintain in large populations. It was

2. Modeling Organogenesis in the Zebrafish

49

Figure 1 Relationship between Forward and Reverse Genetics. Approaches connecting phenotypic mutations with their molecular lesion through forward genetics are illustrated. Conversely, reverse genetics identifies the phenotype associated with a particular molecular change in previously isolated genes.

unanticipated how findings from invertebrates were directly transferable to vertebrate and mammalian organisms, even though they share little outward resemblance. While this might be expected for cellular processes, such as cell cycle control (Lee and Nurse, 1987), pattern and axis formation also use homologous genes from Drosophila to mouse and man (Capecchi, 1997; Hirth and Reichert, 1999; Panganiban and Rubenstein, 2002; Veraksa et al., 2000). In contrast, reverse genetics seeks to address the question of what phenotype arises from a specific molecular alteration (Fig. 1). Molecular information about a gene is used to introduce specific changes into an animal, and subsequently to assay its phenotype. Reverse genetics has been an invaluable tool in analyzing gene function. In particular, targeted gene knockout technology in the mouse (reviewed in van der Weyden et al., 2002) has illustrated the power of this approach. However, in order to apply reverse genetics, a gene has to first be identified. In the past, many mouse genes were identified by their homology to invertebrate genes, by their ability to interact with other proteins, such as isolation from two-hybrid screens or biochemical purification schemes. But any of these approaches requires some molecular information upon which further analysis rests. Because much of mammalian

50

Stuckenholz et al.

genetics is limited to reverse genetics, researchers have typically examined genes that might be predicted a priori to play a role in a physiological process. We therefore do not have a comprehensive picture of the molecules involved in the biological processes specific to vertebrates, such as organogenesis. Technological advances are beginning to open forward genetics to vertebrate systems. For example, forward genetic screens have now been carried out in human cell lines using RNAi (Berns et al., 2004; Paddison et al., 2004). These types of screens are primarily suited to studying cellular processes, and are diYcult to apply toward organogenesis with its many interactions between growing and diVerentiating cells in the context of the developing organism. A few large-scale forward genetic screens in whole organisms have been reported and are ongoing in the mouse, but because they are so resource intensive, they have identified dominant mutant phenotypes scorable after birth (Hrabe de Angelis et al., 2000; Nolan et al., 2000) or were restricted to phenotypes caused by mutations in specific portions of the genome (Kile et al., 2003; Shedlovsky et al., 1988). Organogenesis is one area that may benefit substantially from applying forward genetic screens in a vertebrate model, since it is often unique to vertebrate physiology. Screening for defects in organogenesis in the mouse would ideally involve a genome-wide search for dominant and recessive mutations during development in utero, which is logistically diYcult and costly.

B. Zebrafish as a Vertebrate Model of Organogenesis In 1981, George Streisinger, who pioneered the use of the zebrafish (Danio rerio) as a model organism, published a landmark paper demonstrating that zebrafish could be used as a genetically tractable vertebrate model organism (1981). In December 1996, labs in Boston and Tu¨ bingen published the results of large-scale forward genetic screens that added the zebrafish as a second, genetically tractable vertebrate model system (Driever et al., 1996; HaVter et al., 1996). This multilab eVort described several thousand zebrafish mutants, which spanned almost every organ system, and included even some mutants that resemble human diseases. For instance, the HoltOram syndrome in humans with its congenital heart and limb defects (Holt and Oram, 1960) has a counterpart in the zebrafish mutant heartstrings, which causes congenital heart and pectoral fin defects. In both cases, the defect was found to lie in the tbx5 gene, a T-box transcription factor (Garrity et al., 2002). Thus, the zebrafish mutant replicates a human disease and these fish can be utilized to further understand the disease pathology. Positional cloning of weissherbst, an anemic mutant, led to the discovery of the iron transporter ferroportin in the gastrointestinal tract (Donovan et al., 2000). Ferroportin is now a target gene for the rational molecular design of


51

medicines to treat hemochromatosis. Remarkably, screens setting out to identify mutants with defects in development have already resulted in the unintended, fortuitous identification of several new models for human diseases. These and other disease models have already put the zebrafish on the map as a bona fide medical model organism, a position that will only be bolstered as the advantages of this organism are further exploited. Zebrafish are easy to raise with a short generation time of three months. Because they are small, they can be cheaply and eYciently kept in large numbers. Every cell within a developing embryo can be visualized using a microscope, since the embryos are transparent and develop outside of the mother. Embryos grow rapidly; they have a beating heart and visible erythrocytes by 24 hours (Fig. 2). A zebrafish female can lay hundreds of eggs at weekly intervals. The organism maintains the diploid state, an important diVerence to other fish that can be triploid or tetraploid and thus are not well suited to genetic analysis. Furthermore, zebrafish can be eYciently mutated by a variety of chemical, biological, and radiological means. On a molecular level, zebrafish embryos are easily processed by in situ hybridization, antibody staining, and other standard biochemical and molecular techniques. Gene expression can be aVected through either the injection of RNA to increase translation of a particular gene or by injection of modified antisense oligonucleotides that reduce and, in some cases, even eliminate protein expression of the target gene. Transgenic animals are simple to create and aVord control of gene expression throughout the life of the animal. Finally, the Wellcome Trust/Sanger Center is currently sequencing the entire zebrafish genome and is scheduled to complete it within two years. A large portion of the genome is already publicly available on its website (http://www.sanger.ac.uk/Projects/D_rerio/). This resource will aid the rapid identification and isolation of many genes involved in organogenesis, facilitating the translation of these pathways into the realm of human disease modeling. Even though fish do not share the outward physical characteristics of either mice or humans, there is both structural and functional conservation of many organ systems between the zebrafish and mammals. Conservation between mammalian and teleost brains and gut is pronounced (see Sections III and IV), but other organs are highly conserved as well (reviewed in Shin and Fishman, 2002). Because zebrafish can be eYciently manipulated without diYculty during development, including organogenesis, they will likely contribute to our understanding of organ development. If the prior experience gained from the study of pattern and axis formation in invertebrates is an indication, these findings will be directly transferable to higher vertebrates, including humans.

52

Stuckenholz et al.

Figure 2 Embryonic Development in the Zebrafish. These panels illustrate rapid organogenesis in the zebrafish embryos during early development. (A) 14 somite (s) embryos, approximately 15 h post fertilization (hpf), demonstrate a prominent CNS and eye. (B) At 28 hpf, the heart beats and red blood cells circulate through the developing vasculature. (C–E) Organogenesis continues rapidly over the next 3 days. By 4 days post fertilization (dpf), the swimbladder inflates; liver, pancreas, and intestines are functioning. Zebrafish embryos begin to feed independently by 5 dpf.

II. Forward and Reverse Genetics in Zebrafish Genetic screens are an eYcient means to isolate genes involved in biological processes based solely on a mutant phenotype. This aVords an unbiased, comprehensive survey of the genes required for a given process. After mutagenesis, animals are phenotypically scored and bred to perpetuate the genetic lesion. Once mutants are isolated, they are placed in genetic relationships to each other, the underlying molecular lesion is cloned, and further experiments are conducted to understand how this lesion results in the phenotype.


53

Despite their value, genetic screens do have limitations. As new mutants may be rare, it is necessary to quickly and reliably screen thousands of mutagenized animals to find the few with the desired, heritable phenotype. Just about any biological process that can be assayed can form the basis for a genetic screen. Genes that do not produce a phenotype when mutated, even if they are involved in the underlying process, will not be isolated in a screen. The two major classes of genes unlikely to produce a detectable phenotype are noncoding RNAs because of their small size (Lagos-Quintana et al., 2001; Lim et al., 2003) or redundancy within the gene itself (Caparros et al., 2002; Stuckenholz et al., 2003; Wutz et al., 2002), as well as genes that share at least some redundant function between them (Solomon and Fritz, 2002). These kinds of genes may play an essential role in development, and likely in organogenesis as well (Solomon and Fritz, 2002; Wienholds et al., 2003a; Yekta et al., 2004), but they require modified genetic screens in order to be identified. However, the majority of genes playing a role in organogenesis will be amenable to properly designed forward genetic screens and fill in the many gaps existing today. A. Forward Genetic Screens in Zebrafish 1. Mutagenizing the Zebrafish Genome a. Point Mutagens. The majority of zebrafish screens to date were carried out using N-ethyl-N-nitrosourea (ENU) (for example, Driever et al., 1996; HaVter et al., 1996), an alkylating agent mostly causing single nucleotide point mutations in zebrafish (Knapik, 2000; Mullins et al., 1994; SolnicaKrezel et al., 1994; Wienholds et al., 2002, 2003b). Because the lesions will most likely aVect just a single nucleotide, resulting mutations tend to be partial loss-of-function rather than complete null alleles. The ease with which even subtle changes in phenotype can be assayed, however, will determine to a large extent the kinds of mutations that are successfully isolated. Standard ENU mutagenesis schemes treat adult male fish repeatedly with ENU (Mullins et al., 1994; Riley and Grunwald, 1995; Solnica-Krezel et al., 1994; but also see Grunwald and Streisinger, 1992b), resulting in mutagenesis of germ cells at all stages of spermatogenesis. Since ENU modifies only one strand of DNA, extra rounds of replication are necessary to restore complementarity between the two strands. In pre-meiotically mutated sperm, the replication occurs during spermatogenesis and mature sperm will either be mutant or wild type on both DNA strands at any given locus. If, however, post-meiotic sperm are mutated, the mutations will be fixed during the first rounds of replication in the embryo. Therefore, the embryo will be mosaic because it contains cells with normal alleles and cells with mutant alleles. Since only a variable portion of

54

Stuckenholz et al.

developing embryos are mutant, they are more diYcult to analyze and their mutations are more diYcult to isolate and keep, because the germ line is mosaic as well (Riley and Grunwald, 1995). These disadvantages, however, are oVset by a roughly 10-fold larger mutational load tolerated in mosaic mutants, making it possible to screen many more mutations in an individual animal (Riley and Grunwald, 1995). Because of a need to screen many genomes as unambiguously and quickly as possible and to simplify mutant recovery, most ENU screens rely on pre-meiotic mutation of the germ line, giving rise to nonmosaic oVspring (Driever et al., 1996; HaVter et al., 1996). b. Deletion Mutagens. Gamma and UV irradiation are eVective mutagens in zebrafish (Chakrabarti et al., 1983; Fritz et al., 1996; Grunwald and Streisinger, 1992a; Walker, 1999; Walker and Streisinger, 1983). In contrast to ENU mutagenesis, -radiation induces mostly large deficiencies (Chakrabarti et al., 1983; Grunwald and Streisinger, 1992a), although lesions aVecting only a few bases can be generated (Schulte-Merker et al., 1994). Thus, the resulting complex and pleiotropic phenotype may be more diYcult to study. Gamma-ray mutagenesis screens have been used successfully to identify genes in functions as diverse as axial patterning (Fisher et al., 1997) and sensory placode development (Solomon and Fritz, 2002). Using a polymerase chain reaction (PCR)-based screen, Solomon and Fritz (2002) identified a -ray mutant with deletions of the two partially redundant and closely linked genes dlx3 and dlx7, resulting in a lack of both otic and olfactory placodes. Because ENU induces mostly single point mutations, recovering this particular phenotype from an ENU screen is highly unlikely. Thus, in certain cases, the ability to cause larger genetic lesions may help uncover phenotypes refractory to single gene analysis. The mutagen trimethylpsoralen has been reported to produce smaller deletions than -rays do in C. elegans and has been employed in a small screen in zebrafish, making it a possible alternative to chemical and -ray mutagenesis (Ando and Mishina, 1998; Yandell et al., 1994). Since trimethylpsoralen-induced deletions are smaller in size (on average 2.8 kb; Yandell et al., 1994), they are more likely to aVect just a single gene and perhaps be easier to identify molecularly than single nucleotide changes induced by ENU. c. Biological Mutagenesis Agents. In other genetically tractable model systems, from yeast to D. melanogaster, transposons have been successfully used in mutagenesis screens (Bessereau et al., 2001; Chun and Goebl, 1996; Cooley et al., 1988; Parinov and Sundaresan, 2000). They integrate into the host genome semi-randomly and frequently inactivate nearby genes, which may result in an observable phenotype. They can also be used for expression screening. Reporter genes, such as lacZ or green fluorescent protein (GFP),


55

subcloned into these elements often acquire the transcriptional profile of a nearby gene. Examination of the reporter expression pattern may identify genes at that insertion site that are good candidates for further study. Since the sequence of the inserted element is known, it provides a molecular tag, allowing straightforward and rapid cloning of the insertion site and therefore identification of the genes in this region (Bellen et al., 1989; Bier et al., 1989). Insertional mutagenesis provides a major advantage over other protocols, which require the labor-intensive process of identifying lesions by positional cloning. Additionally, transposons can also be used for the creation of small local deletions resulting from imprecise excision, potentially providing a tool to create many mutant derivatives for further study and analysis (Bessereau et al., 2001; Geyer et al., 1993). Zebrafish harbor a number of repetitive DNA elements and remnants of inactivated transposons (reviewed in Ivics et al., 1999). It should therefore be possible to extend mutagenesis using biological agents to the zebrafish as well. Indeed, transformation of zebrafish has been reported with both retroviruses (Allende et al., 1996; Amsterdam et al., 1999; Burns et al., 1993; Gaiano et al., 1996a,b; Golling et al., 2002; Kawakami et al., 2000) and transposons (Davidson et al., 2003a; Fadool et al., 1998; Ivics et al., 1997, 1999; Izsvak et al., 2000; Raz et al., 1998). Currently, the rate of transposon integration seems too low to support a screen using insertional mutagenesis (Davidson et al., 2003a; Raz et al., 1998). In a tour de force eVort, retroviruses have been used successfully to create a library of insertional mutants (Amsterdam et al., 1999; Gaiano et al., 1996a,b; Golling et al., 2002). These mutants display a range of phenotypes similar to that obtained in chemical mutagenesis screens. However, fish displaying broad defects and fish without overt defects were necessarily discarded in ENU screens, but kept for further analysis in the retroviral screen (Driever et al., 1996; Golling et al., 2002; HaVter et al., 1996). It is possible that future focused analyses will find subtle defects in these fish as well. Recent advances in the synthetic transposons Sleeping Beauty (Davidson et al., 2003a; Ivics et al., 1997; Zayed et al., 2004) and Frog Prince (Miskey et al., 2003) hold the promise that easy, fast, and eYcient transduction of the zebrafish germline is feasible. Transposons and retroviruses could be a significant addition to the zebrafish genetic tool chest. 2. Screening for Recessive Mutations Most newly induced mutations are recessive and dominant mutations are rare. To screen for recessive mutations, however, the recessive allele has to be bred to homozygosity to unmask its phenotype. Males are treated with the mutagen of choice and bred with non-mutagenized females (Fig. 3). Their oVspring, the founder fish generation F1, inherit half of their genome from the mutagenized male; this half potentially carries detectable mutations.

56

Stuckenholz et al.

Figure 3 Standard Three–Generation Crossing Scheme. In the parental generation, wild-type males are treated with a mutagen of choice (e.g., ENU, radiation, biological agents) and crossed with untreated females producing F1 heterozygous progeny (+/m) that contain mutated chromosomes. Progeny are then pairwise crossed to produce families of F2 fish, half of which will contain a mutated chromosome of interest. Subsequent sibling matings will uncover recessive mutations (m/m), schematically shown as gray embryos, aVecting 25% of a clutch.

However, these fish also contain one copy of the non-mutagenized maternal genome, which masks the phenotype of recessive mutations. Therefore, founder fish need to be bred to create oVspring homozygous for induced mutations. Zebrafish oVer several ways of making homozygous mutants,


57

including some powerful methods to create haploid and diploid embryos containing genetic contribution from only one parent. a. Standard Three-Generation Crossing Scheme. The crossing scheme employed in most mutagenesis screens (for examples, see Driever et al., 1996; HaVter et al., 1996) follows the standard scheme devised by Haldane (1956) (Fig. 3). In this scheme, individual F1 founder fish, each carrying one mutagenized parental genome, are outcrossed to wild-type fish, resulting in F2 families that share one mutant genome among the heterozygous siblings. These siblings are crossed, in turn, to obtain homozygous oVspring. Because the F2 fish carry several mutations scattered across their genome and only portions of the genome become homozygous in a single cross, many F2 incrosses need to be set up from each F2 family to recover as many mutants as is practical (van Eeden et al., 1999). Only 25% of embryos in crosses with two heterozygote parents are expected to display a phenotype, but these can nevertheless be easily visualized, given the large size of zebrafish clutches. b. Haploid Zebrafish Embryos. Although the zebrafish is normally a diploid organism, it is fairly easy to create both haploid and diploid embryos containing genetic contributions from only one parent (Fig. 4). Since mutagenized DNA may represent the only allele at a given locus, these embryos allow the direct assessment of recessive mutations. Screens employing such techniques minimize the workload, since fewer generations are required and the F1 generation is more eYciently screened (Cheng and Moore, 1997). Haploid embryos are created by in vitro fertilization of either wild-type zebrafish eggs with UV-irradiated sperm, whose DNA is inactivated and therefore does not contribute to the future embryo (gynogenetic haploids; Streisinger et al., 1981), or by using wild-type sperm and UV-inactivated eggs (androgenetic haploids; Corley-Smith et al., 1996). The latter procedure, however, is technically more demanding and rarely used. Since haploid embryos are hemizygous at all loci, recessive mutations will display a phenotype. Haploid screens therefore allow the immediate assessment of mutagenesis and reduce the number of generations in a screen by one. Haploid embryos are inviable and usually die after embryogenesis by day four. During embryogenesis, they exhibit several defects, including small cell size; abnormal brain, eye, and blood morphology; circulation problems; vacuolated body cavities; and a short and stocky stature (Corley-Smith et al., 1996, 1999; Ho¨ rstgen-Schwark, 1993; Kane and Kimmel, 1993; Streisinger et al., 1981; Walker, 1999). Despite these defects, they develop many structures normally and can be used in screens to detect mutants aVecting early development. Once a mutant has been identified, the parent is outcrossed and F2 oVspring assayed for the mutation, either by creating additional haploid oVspring or by creating normal diploid embryos from sibling

58

Stuckenholz et al.

Figure 4 Schema for Generation of Haploids and Gynogenetic Diploids in the Zebrafish. For either type of embryo, a female is squeezed, her eggs collected and in vitro fertilized with UV inactivated sperm, which cannot contribute to the developing embryo. If the oVspring are subjected to high pressure shortly after fertilization, meiosis II is inhibited. This results in retention of the sister chromatid, producing viable diploid oVspring, whose genetic contribution is solely from the female (gynogenetic diploid). If instead, the fertilized oocytes are left to develop without intervention, haploid embryos will result because of a reduction in their DNA content from 2n to n in meiosis II. Haploid embryos die by day 4 after fertilization secondary to broad defects in later development. Application of heat shock inhibits the first mitotic division of the haploid embryo, restoring the embryo to diploidy.

incrosses. For example, screening haploid embryos has been used to isolate a gene involved in neuronal diVerentiation (Moens et al., 1996), several genes implicated in heart induction and patterning (Alexander et al., 1998), and assisted in the PCR-mediated screening of -radiation mutants (Fritz et al., 1996). If suitable, haploid embryos can vastly speed up and simplify a mutagenesis screen. c. Uniparental Diploid Embryos. Haploid embryos can also be made diploid through the application of heat shock (HS) during the first mitotic division of an egg fertilized with UV-irradiated sperm (Streisinger et al., 1981) or by application of hydrostatic pressure after fertilization (Beattie et al., 1999; Gestl et al., 1997; Streisinger et al., 1981; Fig. 4). Even though either androgenetic or gynogenetic haploid embryos can be restored to


59

diploidy, these techniques are most commonly applied to gynogenetic embryos. A few uniparental diploid embryos display reproducible epigenetic defects, such as cyclopia, reduction in anterior structures, and notochord malformations; yet, many survive to adulthood and are fertile (Beattie et al., 1999). Unless a screen specifically seeks mutations in these structures, uniparental diploid embryos can be employed, because true mutant phenotypes can be reliably distinguished from these epigenetic defects. Haploid embryos may be rendered diploid through the application of a brief heat shock treatment, which prevents cytokinesis during the first cell cycle of the new embryo, thus restoring diploidy to the embryo. These embryos are homozygous at all loci, but are rarely used because their viability is low (Streisinger et al., 1981). Alternatively, early hydrostatic pressure (EP) may be applied to the oocyte soon after fertilization, before the second meiotic division is completed and the second polar body is extruded. Pressure in excess of 6000 psi causes microtubules to disassociate and the spindle to break down, and therefore retention of the sister chromatid in the second polar body (Beattie et al., 1999; Gestl et al., 1997; Streisinger et al., 1981). Because early pressure disrupts meiosis II, the resulting embryo is a half-tetrad and will therefore only be homozygous at all loci proximal to recombination events. Because of high chiasmata interference in zebrafish, double recombinants are very rare, even over whole chromosome arms (Streisinger et al., 1986). Thus, the farther a mutation is away from the centromere, a smaller percentage of homozygous mutants is present in a clutch, making scoring the mutation increasingly diYcult. In practice, however, even at distal loci a significant number of embryos are still homozygous for the mutation (Beattie et al., 1999). Within a clutch, all homozygous mutant embryos display a common phenotype. Since an egglay is large, these mutations can be distinguished from embryos displaying nonspecific, epigenetic defects characteristic of gynogenetic diploid embryos (Beattie et al., 1999; Henion et al., 1996). In general, 50 to 80% of eggs subjected to EP treatment develop into fertile adults, a number suYcient for large-scale screens (Beattie et al., 1999; Gestl et al., 1997; Streisinger et al., 1981). EP screens have been successfully used in the past to identify mutants in many diverse biological processes, including the development of neural crest cells (Henion et al., 1996), neuronal pathfinding (Beattie et al., 2000), and mutants with defects in retinal and gut development (Mohideen et al., 2003), among others. 3. Screening Techniques a. Screening for Morphologic Change. The two large-scale screens reported in the special issue of Development (1996, Vol. 123; reviewed in Currie, 1996; Eisen, 1996; Grunwald, 1996; Holder and McMahon, 1996)

60

Stuckenholz et al.

and many subsequent screens isolated mutants by looking for obvious, apparent defects in early development under a dissecting scope. While this approach is clearly applicable to identify mutants in a wide variety of biological processes (as validated in the many mutants covering most, if not all, early biological processes), it may miss mutations causing subtle phenotypes. Mohideen et al. (2003) demonstrated that even subtle defects in organ formation can be identified by screening histological sections of parthenogenetic diploid embryos. This approach resulted in the isolation of seven mutations, aVecting eye, gut, and general organ formation with phenotypes diYcult, if not impossible, to identify using a dissecting scope (Mohideen et al., 2003). b. Screening with Molecular Markers. A screen aimed primarily at one specific biological process may employ a more specialized assay to identify mutants in this process. Because zebrafish are easily processed by whole mount in situ hybridization or antibody staining, it is feasible to identify mutants based on an abnormal staining of a particular marker. Henion et al. (1996) were first to isolate zebrafish mutants based on altered distribution of a molecular marker, using monoclonal antibodies against Hu as a marker for sensory neurons of the dorsal root ganglion. They successfully isolated several genes aVecting development of the neural crest. As another example, valentino was identified as a gene with an important role in hindbrain formation because of misexpression of krox20, an early rhombomere marker (Moens et al., 1996). A screen to identify mutants in heart induction and patterning employed a combination of in situ hybridization and immunohistochemistry to isolate several new mutants defective in this process (Alexander et al., 1998). This ability to combine molecular probes and forward genetics in vertebrate organisms significantly enhances the detail by which organogenesis can be described. It is also possible to screen zebrafish embryos for changes in expression of reporter genes, such as GFP. In such cases, it is not even necessary to carry out in situ hybridization, because the expression pattern can be observed in real time, in live embryos without a staining procedure. A growing number of laboratories are utilizing GFP to simplify screening for defects in organogenesis, such as vessel formation, among others (for example, see Lawson et al., 2003). Because the embryo does not have to be sacrificed in the process but can be grown to sexual maturity if the mutation does not induce sterility or lethality, the mutation can be directly recovered in its oVspring. Numerous GFP transgenic fish have already been created (for a list, see Udvadia and Linney, 2003) and recent successes with diVerentially colored fluorescent proteins indicate that it will be possible to screen for expression of several fluorescent genes at the same time (Gong et al., 2003; Lawson and Weinstein, 2002; Wan et al., 2002), further increasing the eYciency of screening for novel mutants.


61

Expanding on the large-scale mutagenesis screen in Tu¨ bingen, injection of fluorescent dyes into larval retinas was used to assay for its ability to project neurons from the retina to the contralateral tectum (Baier et al., 1996). Mutations with defects in both axon pathfinding and mapping to the correct part of the tectum were identified (Karlstrom et al., 1996; Trowe et al., 1996), indicating that screening in the zebrafish can accommodate assays to address very detailed questions in thousands of embryos, if the assay is well designed. c. Functional Screens. Farber et al. (2001) devised a screen to monitor organ function. Zebrafish were screened for defects in lipid metabolism by following uptake of quenched fluorescent phospholipids through their digestive system. Upon ingestion of the lipid, processing of the phospholipids caused removal of the quenching moiety and strong fluorescent staining in the gall bladder. Since the embryo is optically clear, it is easy to distinguish glowing, normal embryos from those lacking fluorescence because of defects in lipid uptake or processing (Farber et al., 2001). Depending upon the physiology of the organ screened, other functional assays may be extended to provide an easy read-out to scan thousands of zebrafish easily and eYciently. 4. Cloning Mutations Once a mutation of interest has been identified, standard positional cloning techniques are available to identify point and deletion mutations. Because of the large clutch size in zebrafish, an ample number of meioses are easily obtained and analyzed. The use of haploid and gynogenetic diploid fish reduces the genetic complexity and simplifies the cloning process even further. These uniparental fish along with detailed genetic and physical maps make the zebrafish an extremely robust genetic system for the characterization and cloning of mutant phenotypes. Additionally, zebrafish and human chromosomes are surprisingly syntenic (Barbazuk et al., 2000; Gates et al., 1999; Postlethwait et al., 1998) and comparative genomics can often be used to identify a candidate gene. Once a few genes close to the mutation have been identified, the syntenic region in the human or any other genome can be searched for genes with features and expression patterns that might be expected of the mutated gene. The zebrafish orthologue of such a gene then becomes a candidate gene for the mutant allele. There are cases in which the mutation was cloned purely by synteny (for an example, see Karlstrom et al., 1999). More often, information gleaned from analysis of the syntenic region of chromosomes in humans and other species has aided cloning of numerous zebrafish genes, such as kugelig and weissherbst (Davidson et al., 2003b; Donovan et al., 2000). Despite the large size of the zebrafish genome, positional cloning has

62

Stuckenholz et al.

identified the genes involved in numerous mutations and advances in the genome project will undoubtedly aid in the identification of many more. Alternatively, biological mutagens serve both as mutagen and as molecular tag to allow rapid identification of their insertion site through molecular techniques such as inverse-PCR. Demonstrating the strength of insertional mutagenesis, Golling et al. (2002) successfully identified candidate genes far more rapidly than is feasible for ENU screens. The limiting factor for ascertaining candidate genes was identification of the insertion site in the zebrafish sequence. As the zebrafish genome project progresses toward completion, the percentage of insertion sites that can be placed on the zebrafish map will continue to increase.

B. Reverse Genetics: Controlling Gene Expression in Zebrafish Once an interesting mutation has been identified and the corresponding gene has been cloned, further experiments are needed to elucidate its function. Zebrafish can be utilized to delineate organogenesis using a wide variety of standard biochemical and molecular biology techniques to probe for gene structure and interacting proteins. Use of the zebrafish also facilitates study of organogenesis by reverse genetic approaches, and some of the more frequently used techniques are outlined below. 1. Knockdown and Knockout Techniques Reverse genetics often aims to selectively decrease or entirely eliminate expression of one specific gene. Morpholinos have been demonstrated to be particularly suited to this end (reviewed in Ekker and Larson, 2001; Heasman et al., 2000). They are chemically modified short oligonucleotides designed to be complementary to a target RNA that are injected into the developing embryo. Once a morpholino binds to the translation initiation region of an RNA, it can block translation almost completely, producing a severe, often null phenotype (Nasevicius and Ekker, 2000). Alternatively, if a morpholino is directed toward a splice site, the resulting phenotype can be partial (Draper et al., 2001). Morpholinos have been used successfully to create models of human disease solely through their injection. For example, targeting uroporphyrinogen decarboxylase with morpholinos created embryos mimicking the disease hepatoerythropoietic porphyria (Nasevicius and Ekker, 2000). Because early development of the embryo is so rapid and because transgenes need not be established, morpholinos are a very fast, powerful method to investigate early gene function. Other modified oligonucleotides to selectively inhibit RNA functionality are currently under evaluation for increased specificity and stability (Urtishak et al., 2003).


63

Altering the zebrafish germline to permanently remove expression of a single gene by targeted mutagenesis is currently the subject of intense research. Homologous recombination in germline-competent cell lines is required for targeted mutagenesis. Separately, homologous recombination and germline competency in zebrafish cell lines have been observed. As germline-competent cell lines can now be propagated long enough to isolate homologous recombination events, targeted mutagenesis zebrafish may soon be a reality (Fan et al., 2003; Hagmann et al., 1998; Ma et al., 2001). Additionally, the zebrafish has recently joined the list of cloned animals using cells that had been cultured for months (Huang et al., 2003; Lee et al., 2002). Presumably, DNA of these cells could be modified by homologous recombination while the cells are in culture. The technique of targeting induced local lesions in genomes (TILLING) has allowed the fast and eYcient isolation of mutations from a library of ENU mutants in a specific gene of interest (Wienholds et al., 2003b), for example, rag1 and dicer (Wienholds et al., 2002, 2003a). While this technique does not permit designing specific mutations, several alleles for each gene are usually isolated that can often be arranged in an allelic series, which can be an invaluable tool in analyzing gene function. 2. Overexpression Strategies Overexpressing genes in the zebrafish is just as important as selectively reducing gene function. The easiest and most straightforward approach to gene overexpression in zebrafish is the injection of in vitro-transcribed RNA into embryos, where it is translated into protein. Because RNA injections do not require the establishment of a transgenic line, this approach represents a fast method to increase protein levels during early embryogenesis. Injecting the wild-type RNA of a mutated gene and reverting the phenotype is a particularly powerful method to ensure that the identified mutation causes the observed phenotype. Dominant-negative RNA constructs may also be employed to reproduce a phenotype (Bauer et al., 2001). Constitutively active RNA constructs can help clarify the position of a gene in a genetic pathway (Bauer et al., 2001). Using chemical modifications of the RNA, some temporal and spatial expression control is possible (Ando and Okamoto, 2003; Ando et al., 2001). If expression past the first few days of embryogenesis or a stock stably overexpressing the gene is desired, zebrafish are easily made transgenic via injection of a transposon, retrovirus, or linearized DNA, including complete BACs, (for reviews, see Linney et al., 1999; Linney and Udvadia, 2004). Such transgenic constructs allow the expression of a wild-type gene in a mutant background to rescue the phenotype. In the absence of targeted knock-in mutations in fish, expression of engineered transgenes in mutants is a powerful way to ask questions regarding structure and function of the gene in question.

64

Stuckenholz et al.

To control transgene expression, the Gal4-UAS system has been shown to work in zebrafish (Scheer and Campos-Ortega, 1999; Scheer et al., 2001). In this two-component system, a driver strain expressing Gal4 in a specific tissue at a specific time is crossed to another strain which has a gene of interest transcriptionally regulated by the upstream activating sequences (UAS). In embryos containing both components, Gal4 binds to the UAS, thereby activating expression of the gene of interest in Gal4-positive cells. The combinatorial nature of the Gal4-UAS system allows reusing of driver and target lines to reduce the number of transgenic lines necessary to express diVerent genes in diVerent structures throughout development. It also allows establishment of lines bearing toxic transgenes.

C. Conclusions Zebrafish are unique among vertebrate model organisms. First, as has been demonstrated by the increasing number of published screens, zebrafish are exceedingly amenable to forward genetics. Their transparent extramaternal development allows the design of innovative screens to analyze detailed aspects of development not easily accessible in other vertebrate organisms. The creation of haploid and viable uniparental diploid organisms is distinctive among vertebrate model systems. Second, a wide variety of molecular techniques exist to facilitate using reverse genetics to study gene function. There is currently no other model organism available that combines such powerful tools to decipher pathways of organ development.

III. Zebrafish as a Tool to Investigate Neurodevelopment The central nervous system (CNS) is easily visualized during early zebrafish development. As a result, a large number of laboratories have utilized a variety of tools to investigate and model brain organogenesis in the zebrafish (Development (1996), Vol. 123). Well before the embryo is even a day old, the brain and eyes are easily recognizable structures (Fig. 2). The following sections provide examples of how zebrafish embryology and genetics are powerful methods to study CNS organogenesis.

A. Neural Tube Formation Similar to what has been described in other organisms, at the onset of gastrulation, zebrafish neural ectoderm is organized as a monostratified epithelium called the neural plate (Papan and Campos-Ortega, 1994).


65

Through additional cell movements, the neural plate transforms into a neural tube, with the axial-most neural plate cells forming the ventral neural tube, the mediolateral cells forming the dorsal neural tube, and the most lateral cells forming neural crest and epidermis. The planar structure of the neural plate is important for the establishment of morphogenic gradients and cell–cell interactions that later influence the organogenesis of the nervous system. Various mutagenesis screens have identified mutants with defects in key regulatory genes aVecting neural development. Swirl, snailhouse, and somitabun mutants were isolated from the large-scale screens as mutants having a dorsalized phenotype (Mullins et al., 1996), and were later found to have defects in bmp2b, bmp7, and smad5 respectively (Dick et al., 2000; Hild et al., 1999; Kishimoto et al., 1997; Nguyen et al., 1998; Schmid et al., 2000). Interestingly, these mutants also have CNS defects, revealing that the nervous system is patterned, in part, through laterally expressed Bone Morphogenic Proteins (BMPs), forming a lateral-to-medial gradient in the neural plate (Nguyen et al., 2000). Further investigations in the zebrafish have helped demonstrate that high levels of ventral BMP signaling in the neural plate promote epidermal diVerentiation, while lower BMP activity levels specify neural crest cells (reviewed in Kelsh and Raible, 2002) and CNS neurons that will later reside in the dorsal neural tube. These molecular events are similar to those observed in chick and mouse, even though particular cellular events are slightly diVerent. The floor plate forms from the ventral-most cells of the neural tube caudal to the forebrain. Floor plate cells are in contact with the axial mesoderm (the future notochord) and molecular signaling from both regulates ventral neural tube formation. Similar to other vertebrates, zebrafish Hedgehog (Hh) molecules (sonic hedgehog, shh; tiggywinkle hedgehog; and echidna hedgehog) are expressed in the floor plate and notochord. The zebrafish sonic-you mutant (syu), defective for shh and isolated from the large-scale screens, lacks lateral floor plate cells, although the medial floor plate and motoneurons that surround them are present (Odenthal et al., 2000; Schauerte et al., 1998). Since three Hh molecules are expressed in the notochord and/or floor plate in zebrafish, these genes most likely have redundant functions. Reducing the activity of only a single hedgehog gene therefore results in an incomplete phenotype, as seen in syu mutants. Simultaneous morpholino-based reduction of all three zebrafish Hedgehogs demonstrates their functional redundancy. Interestingly, reducing the expression of all Hh genes eliminates the lateral floor plate cells and the majority of motoneurons, but the medial floor plate cells still remain (Lewis and Eisen, 2001). This contrasts with other vertebrates, as in vitro studies in chick (Roelink et al., 1995) and targeted deletion of shh in mouse (Chiang et al., 1996) result in a loss of all floor plate and motoneuron cells. These

66

Stuckenholz et al.

results are suggestive that Hh signaling is suYcient and necessary for the specification of ventral spinal cord fates in these animals. However, the experimental evidence from zebrafish suggests that only the lateral floor plate and motoneurons require Hh signaling. Reduction of all Hh signaling is seen in the zebrafish slow muscle omitted mutant, which has a defect in the smoothened gene, a key Hh signaling component. This mutant also has reduced numbers of motoneurons, but still retains medial floor plate cells. This suggests that in zebrafish the Hh pathway is necessary for forming ventro-lateral neural tissue, but not for the medial floor plate, which most likely requires additional signaling mechanisms (Chen et al., 2001; Varga et al., 2001). After domain specificity is conferred to the neural tube, certain cell fates are determined that initiate the first cellular specification events.

B. Neural Induction In the neural tube, certain neural cell types are located within specific bilateral longitudinal domains. The domains in which these neurons reside result from earlier signaling events, such as from the BMP, Hh, and Nodal signaling pathways that pattern the embryonic axis. Clusters of proneural cells are established within these domains. Although each cell within a cluster has the potential of gaining a neural fate, only one of these cells will do so. The dynamic regulation of proneural genes, both through autoregulation and lateral inhibition, ensures that the proper number of neuroectodermal cells attain the neural fate. Delta-Notch signaling, a pathway first characterized in Drosophila and present in all vertebrates, plays a significant role in regulating the lateral inhibition process (Appel et al., 2001). In the nervous system, all cells within a proneural cluster express at least one Delta and Notch receptor. Delta ligand reduces Delta expression on nearby cells through activating Notch receptors, which in turn reduces the expression of proneural genes. The cell in a cluster expressing the highest level of proneural genes and Delta is destined to become a neuron. Loss-of-function studies on the Delta-Notch lateral inhibition pathway result in an increase in terminally diVerentiated neurons, as well as a reduction in proliferative nonneural cells. The DeltaNotch pathway has been similarly characterized in zebrafish. For example, overexpression of a dominant negative form of DeltaA through mRNA injection into early zebrafish embryos results in an increased number of neuronal cells by blocking Notch receptors (Appel and Eisen, 1998). Similarly, the zebrafish deltaA mutant has an increased numbers of neurons (Appel et al., 2001; Cornell and Eisen, 2000), but at the loss of oligodendrocytes (Park and Appel, 2003) and floorplate cells (Appel et al., 1999).


67

The mind bomb (mib) mutation reveals a previously unknown and important element in Delta-Notch signaling. Mib was first identified in the large-scale mutagenesis screens and subsequently positionally cloned and found to encode a RING ubiquitin ligase (Itoh et al., 2003). It appears that Mib is required to enhance Delta’s ability to activate Notch, perhaps through targeted protein degradation. The mouse mib homologue, called Dip1, was found to regulate the expression of the death-associated protein kinase (DAPK; Jin et al., 2002). Thus, mib not only regulates Delta-Notch signaling, but most likely other molecular pathways as well. The zebrafish deltaD promoter has been isolated and characterized, and transgenic fish have been created, which express GFP only in the subset of cells that would normally express deltaD (Hans and Campos-Ortega, 2002). This promoter will be a valuable tool for misexpressing genes to better understand the Delta-Notch signaling pathway.

C. Conclusions These and other studies clearly demonstrate how the tools available in the zebrafish to study organogenesis are being utilized to help unravel the complexities of vertebrate CNS development. Not only is development of the zebrafish CNS nearly identical to that of higher vertebrates, but also the many advantages of the zebrafish system oVer forward and reverse genetic approaches to examine the function of newly identified genes. Additionally, previously identified genes and pathways can be further characterized. Taken together, the zebrafish system is helping to define aspects of brain development that will ultimately help provide insights into human neurological diseases.

IV. Gastrointestinal Development in Zebrafish Another example of using zebrafish to define organogenesis can be found in the patterning, specification, growth, and diVerentiation of embryonic endoderm into the gastrointestinal tract (Grapin-Botton and Melton, 2000; Gualdi et al., 1996; Jung et al., 1999; Keng et al., 1998; Ober et al., 2003; Rawdon, 2001; Roberts, 2000) (Fig. 5). The gastrointestinal tract includes the stomach and intestines, as well as associated organs, such as liver, kidney, bladder, pancreas, thyroid, and thymus. On a molecular level, gastrointestinal formation across vertebrates demonstrates a remarkable degree of conservation and thus insights gained in the zebrafish can be transferred to mammalian gastrointestinal formation (Pack et al., 1996). As is true for other organs in the zebrafish, a fully functional gastrointestinal

68

Stuckenholz et al.

Figure 5 Schema of Zebrafish Gastrointestinal Organogenesis. Several genes, including fau (gata5), bon (mixer), cas, the nodal pathway genes oep, sqt, and cyc, as well as mezzo, appear to function in endoderm specification. Once specified, the gastrointestinal organs are derived from endoderm through the action of genes that include pdx1 in the pancreas and hhex in the liver. Most of the pathways and genes involved in further growth and diVerentiation of the gastrointestinal tract remain to be identified.

tract develops quite rapidly. Within four days after fertilization, movements of the endodermal progenitors lead to the proliferation and expansion of the various gastrointestinal organs from endoderm (Field et al., 2003b; Ober et al., 2003; Pack et al., 1996; Tam et al., 2003; Wallace and Pack, 2003). The fast, external development of the zebrafish embryo coupled with its optical clarity permit the application of the techniques described earlier to delineate the pathways involved in gastrointestinal organogenesis. A. Endoderm Specification Since the gastrointestinal tract is derived from embryonic endoderm, a complete understanding of gastrointestinal organogenesis begins with the specification of the endoderm. A number of molecular pathways and transcription factors required for the specification of endoderm are conserved from zebrafish to mammals. They include nodal and BMP signaling pathways, SOX proteins, and a number of transcription factors from the GATA, MIX, paired, and forkhead domain families (Stainier, 2002). One of


69

the earliest gene families involved in the formation of the embryonic endoderm across all species studied to date is the nodal pathway, which has also been implicated in brain development. By forward genetic screening in the zebrafish, two nodal family members, squint (sqt) and cyclops (cyc), have been shown to function in a partially redundant manner as critical mediators of endoderm development. Each single mutant has near normal endodermal development, while the double mutant demonstrates a near complete loss of the endoderm (Chen and Schier, 2001; Dougan et al., 2003; Zhang et al., 1998). Aided by the transparency and extrauterine development of zebrafish embryos, it was noted that embryos lacking both maternal and zygotic contributions of one-eyed pinhead (oep) fail to form endoderm and most mesoderm, similar to the phenotype of the sqt/cyc double mutant (Zhang et al., 1998). The zebrafish oep mutant encodes an EGF-CFC family member, implicating this family as a necessary co-receptor for proper nodal signaling, and by extension, proper brain and gastrointestinal development (Gritsman et al., 1999). A number of genes have been implicated to act downstream of the nodal pathway in endodermal development. Sox17 encodes a high mobility group DNA binding domain protein with an SRY homeobox and has been implicated in endodermal development in Xenopus, zebrafish, and mouse (Kikuchi et al., 2000, 2001; Stainier, 2002). In the zebrafish, the novel Sox protein casanova (cas) has been identified as an important endodermal determinant (Alexander et al., 1999). The zebrafish mutant cas lacks all endodermal precursors, but its defects are restricted to the endoderm; the mesoderm develops normally, unlike oep and cyc/sqt double mutants. To date, no murine homologue of cas has been described, suggesting that another gene, perhaps another sox family member, mediates the functions attributed to cas in the zebrafish (Ober et al., 2003; Stainier, 2002). The GATA family of transcription factors are zinc finger proteins which recognize the GATA motif. In mouse, six diVerent GATA transcription factors have been identified. Gatal through Gata3 are predominantly expressed and function within the hematopoietic system, while Gata4, Gata5, and Gata6 are expressed in endoderm and other tissues (Patient and McGhee, 2002; Weiss and Orkin, 1995). The zebrafish mutant faust (fau) encodes the zebrafish homolog of gata5. Its phenotype includes not only loss of the gut tube, but also defects in the development of the other endodermal organs, including the thyroid, thymus, liver, and pancreas, implicating a significant role for zebrafish gata5 in GI development (Reiter et al., 1999, 2001; Weber et al., 2000). The zebrafish mutant bonnie and clyde (bon) is defective for a mix-type homeodomain protein whose loss of function causes specific endodermal defects (Kikuchi et al., 2000). Indeed, bon mutants have a profound reduction in the number of sox17 expressing endodermal precursors and

70

Stuckenholz et al.

thus lack most gut tissue. More recently, a second mix family member, mezzo, was identified, which acts downstream of nodal signals and is expressed in the mesendoderm during gastrulation (Poulain and Lepage, 2002). Numerous experiments have analyzed the relationship between fau (gata5), bon (mixer), cas, sox17, the nodal pathway genes, and mezzo (Poulain and Lepage, 2002). These experiments have utilized several tools available in the zebrafish, including mutant analysis, morpholino knockdown, RNA overexpression, and genetic epistasis studies. Overall, these experiments suggest a model in which fau, bon, cas, and mezzo are downstream of nodal signaling and upstream of sox17 (Ober et al., 2003; Reiter et al., 1999; Stainier, 2002). While we do not understand endodermal specification in the zebrafish—or in other vertebrate model systems—in fine detail, a number of important genes have been identified, aided by the unique features of the zebrafish model system, including transparent embryos and their extrauterine development. Although some zebrafish genes, such as cas, have no obvious mammalian orthologue, the hierarchy and interactions of the various pathways involved in both mammalian and zebrafish endoderm development are conserved.

B. Embryonic Gastrointestinal Development in the Zebrafish The molecular similarities in endoderm and gut diVerentiation among zebrafish, mouse, and human extend to conservation at the histological levels of detail as well (Fig. 5). The lumen of the zebrafish gastrointestinal tract is visible at about 36 to 48 hours, with the rapid development of polarized epithelial and goblet cells visible between 2 and 5 days post fertilization (dpf) by standard staining techniques. Periodic Acid SchiV (PAS) staining and electron microscopy demonstrate the presence of a prominent microvillous brush border, tight junctions, and muscular layers similar to those seen in mice or humans (Pack et al., 1996). The liver resembles the mammalian liver in cell types, but lacks a lobular architecture/portal tract triad. By 2 days of development, a liver rudiment is visible (Farber et al., 2001; Pack et al., 1996), and by 4 days, a motile gut is seen (personal observations). In situ hybridization of the developing zebrafish gastrointestinal tract demonstrates that the zebrafish orthologues of many human genes are expressed similarly during zebrafish and mammalian embryonic development. For instance, the insulin gene, expressed in the mammalian endocrine pancreas, marks the developing zebrafish pancreas (Pack et al., 1996). Intestinal fatty acid binding protein (IFABP), expressed in human and mouse enterocytes, is similarly expressed in the zebrafish (Andre et al., 2000; Pierce et al., 2000).


71

The ability to directly visualize organogenesis in the zebrafish embryos has enabled investigators to take full advantage of transgenic zebrafish expressing fluorescent markers under the control of tissue-specific promoters. Various groups have separately established transgenic lines expressing GFP under the control of organ- or cell type-specific promoters: in the pancreas under the control of the insulin promoter (Huang et al., 2001), in the liver using the liver fatty acid binding protein promoter (LFABP; Her et al., 2003), and both green and red fluorescent proteins in the enterocytes of the developing gut using ifabp (Her et al., 2004). While GFP expression is restricted to specific organs or cell types in the previously mentioned transgenic lines, Tobias Roeser created a GFP transgenic line in Herwig Baier’s lab through the fusion of the constitutive EF-1 promoter to the GFP. Details of its expression pattern have been previously reported (Field et al., 2003b; Ober et al., 2003). Initially, there is widespread expression of the fluorescent transgene, which becomes restricted to the developing gastrointestinal tract by 24 hours (Field et al., 2003b; Ober et al., 2003). Using this transgenic line, Field et al. (2003a,b). have shown that zebrafish liver and pancreas develop similarly to the human and mouse organs (Fig. 5). This transgenic line and others like it are being actively utilized in forward genetic screens to define new pathways and genes involved in gastrointestinal organogenesis. Reverse genetics has also been employed to dissect gastrointestinal organogenesis. A pdx1 zebrafish morphant (a morpholino-injected embryo) demonstrates pancreas budding deficiency, analogous to the loss of PDX1 in humans and mice (Yee et al., 2001). In the mouse, targeted disruption of the homeobox gene Hex results in death at approximately E10.5 without substantial liver development (Keng et al., 2000). Similarly, an antisense morpholino targeted to the zebrafish hhex gene causes deficiencies in embryonic zebrafish liver formation (Wallace et al., 2001) (Fig. 5). These results not only illustrate the developmental similarity between zebrafish and mammals, but also highlight the far more rapid assessment of gene function that can be accomplished using the zebrafish compared with traditional murine knockout technology. From these studies, we can see that embryonic gastrointestinal development in the zebrafish closely resembles gastrointestinal development in other vertebrates, implying a strong conservation of the genes and pathways between humans and the zebrafish to form the embryonic gastrointestinal tract. Despite what is known about embryonic gastrointestinal development, much more remains to be discovered. As a result of the parallels between human and zebrafish gastrointestinal development, we expect that new genes and pathways uncovered in future zebrafish studies will provide unique insights into normal gastrointestinal development as well as distinct human pathological conditions.

72

Stuckenholz et al.

V. Conclusions and Future Directions In this chapter, we have shown that zebrafish have been successfully used to analyze embryonic development and organogenesis. In particular, largescale genetic screens, which have propelled the zebrafish to scientific prominence, demonstrated for the first time that an entire vertebrate genome can be analyzed in detail using random mutagenesis. The isolated mutants have already significantly contributed to our understanding of early vertebrate development and organogenesis. As new and innovative screens are carried out and many of the previously isolated mutants are further analyzed, a comprehensive picture of development and organogenesis will emerge. Zebrafish are uniquely capable of being utilized to carry out forward genetic screens and to study genes using a host of reverse genetic approaches. Gene overexpression and knockdown techniques are routinely used and have greatly added to our current understanding of molecular mechanisms. In addition, it is already possible to select fish with mutations in particular genes. With targeted mutagenesis expected to be available soon, the zebrafish genome will be amenable to specific, designed manipulation, allowing the researcher an unprecedented level of control over the biology of a vertebrate model organism. As connections between development and human disease become more apparent, increased emphasis will be placed on understanding the basic biology of organ development. Because of the many features that make the zebrafish so attractive to study organogenesis, its importance in unraveling pathways and genes involved in organ development will undoubtedly grow. As unlikely as it may seem at first glance, use of the zebrafish as a model organism will lead to the identification of novel targets for therapeutic intervention and treatments for human disease.

Acknowledgments We thank Mr. Christopher Schafer and Ms. Betsy Johnson for assistance with zebrafish husbandry, sectioning, and photography. This work was, in part, supported by grants from the Crohn’s and Colitis Foundation of America and the March of Dimes.

References Alexander, J., Rothenberg, M., Henry, G. L., and Stainier, D. Y. (1999). Casanova plays an early and essential role in endoderm formation in zebrafish. Dev. Biol. 215, 343–357. Alexander, J., Stainier, D. Y., and Yelon, D. (1998). Screening mosaic F1 females for mutations aVecting zebrafish heart induction and patterning. Dev. Genet. 22, 288–299.


73

Allende, M. L., Amsterdam, A., Becker, T., Kawakami, K., Gaiano, N., and Hopkins, N. (1996). Insertional mutagenesis in zebrafish identifies two novel genes, pescadillo and dead eye, essential for embryonic development. Genes Dev. 10, 3141–3155. Amsterdam, A., Burgess, S., Golling, G., Chen, W., Sun, Z., Townsend, K., Farrington, S., Haldi, M., and Hopkins, N. (1999). A large-scale insertional mutagenesis screen in zebrafish. Genes Dev. 13, 2713–2724. Ando, H., Furuta, T., Tsien, R. Y., and Okamoto, H. (2001). Photo-mediated gene activation using caged RNA/DNA in zebrafish embryos. Nat. Genet. 28, 317–325. Ando, H., and Mishina, M. (1998). EYcient mutagenesis of zebrafish by a DNA cross-linking agent. Neurosci. Lett. 244, 81–84. Ando, H., and Okamoto, H. (2003). Practical procedures for ectopic induction of gene expression in zebrafish embryos using Bhc-diazo-caged mRNA. Methods Cell Sci. 25, 25–31. Andre, M., Ando, S., Ballagny, C., Durliat, M., Poupard, G., Briancon, C., and Babin, P. J. (2000). Intestinal fatty acid binding protein gene expression reveals the cephalocaudal patterning during zebrafish gut morphogenesis. Int. J. Dev. Biol. 44, 249–252. Appel, B., and Eisen, J. S. (1998). Regulation of neuronal specification in the zebrafish spinal cord by Delta function. Development 125, 371–380. Appel, B., Fritz, A., Westerfield, M., Grunwald, D. J., Eisen, J. S., and Riley, B. B. (1999). Delta-mediated specification of midline cell fates in zebrafish embryos. Curr. Biol. 9, 247–256. Appel, B., Givan, L. A., and Eisen, J. S. (2001). Delta-Notch signaling and lateral inhibition in zebrafish spinal cord development. BMC Dev. Biol. 1, 13. Baier, H., Klostermann, S., Trowe, T., Karlstrom, R. O., Nu¨ sslein-Volhard, C., and BonhoeVer, F. (1996). Genetic dissection of the retinotectal projection. Development 123, 415–425. Barbazuk, W. B., Korf, I., Kadavi, C., Heyen, J., Tate, S., Wun, E., Bedell, J. A., McPherson, J. D., and Johnson, S. L. (2000). The syntenic relationship of the zebrafish and human genomes. Genome Res. 10, 1351–1358. Bauer, H., Lele, Z., Rauch, G. J., Geisler, R., and Hammerschmidt, M. (2001). The type I serine/threonine kinase receptor Alk8/Lost-a-fin is required for Bmp2b/7 signal transduction during dorsoventral patterning of the zebrafish embryo. Development 128, 849–858. Beattie, C. E., Melancon, E., and Eisen, J. S. (2000). Mutations in the stumpy gene reveal intermediate targets for zebrafish motor axons. Development 127, 2653–2662. Beattie, C. E., Raible, D. W., Henion, P. D., and Eisen, J. S. (1999). Early pressure screens. Methods Cell Biol. 60, 71–86. Bellen, H. J., O’Kane, C. J., Wilson, C., Grossniklaus, U., Pearson, R. K., and Gehring, W. J. (1989). P-element-mediated enhancer detection: A versatile method to study development in Drosophila. Genes Dev. 3, 1288–1300. Berns, K., Hijmans, E. M., Mullenders, J., Brummelkamp, T. R., Velds, A., Heimerikx, M., Kerkhoven, R. M., Madiredjo, M., Nijkamp, W., Weigelt, B., Agami, R., Ge, W., Cavet, G., Linsley, P. S., Beijersbergen, R. L., and Bernards, R. (2004). A large-scale RNAi screen in human cells identifies new components of the p53 pathway. Nature 428, 431–437. Bessereau, J. L., Wright, A., Williams, D. C., Schuske, K., Davis, M. W., and Jorgensen, E. M. (2001). Mobilization of a Drosophila transposon in the Caenorhabditis elegans germ line. Nature 413, 70–74. Bier, E., Vaessin, H., Shepherd, S., Lee, K., McCall, K., Barbel, S., Ackerman, L., Carretto, R., Uemura, T., Grell, E., Jan, L. Y., and Jan, Y. N. (1989). Searching for pattern and mutation in the Drosophila genome with a P-lacZ vector. Genes Dev. 3, 1273–1287. Black, B. L., and Olson, E. N. (1998). Transcriptional control of muscle development by myocyte enhancer factor-2 (MEF2) proteins. Annu. Rev. Cell Dev. Biol. 14, 167–196.

74

Stuckenholz et al.

Burns, J. C., Friedmann, T., Driever, W., Burrascano, M., and Yee, J. K. (1993). Vesicular stomatitis virus G glycoprotein pseudotyped retroviral vectors: Concentration to very high titer and eYcient gene transfer into mammalian and nonmammalian cells. Proc. Natl. Acad. Sci. USA 90, 8033–8037. Caparros, M. L., Alexiou, M., Webster, Z., and BrockdorV, N. (2002). Functional analysis of the highly conserved exon IV of XIST RNA. Cytogenet. Genome Res. 99, 99–105. Capecchi, M. R. (1997). Hox genes and mammalian development. Cold Spring Harb. Symp. Quant. Biol. 62, 273–281. Chakrabarti, S., Streisinger, G., Singer, F., and Walker, C. (1983). Frequency of gamma-ray induced specific locus and recessive lethal mutations in mature germ cells of the zebrafish, Brachydanio rerio. Genetics 103, 109–123. Chen, W., Burgess, S., and Hopkins, N. (2001). Analysis of the zebrafish smoothened mutant reveals conserved and divergent functions of hedgehog activity. Development 128, 2385–2396. Chen, Y., and Schier, A. F. (2001). The zebrafish Nodal signal Squint functions as a morphogen. Nature 411, 607–610. Cheng, K. C., and Moore, J. L. (1997). Genetic dissection of vertebrate processes in the zebrafish: A comparison of uniparental and two-generation screens. Biochem. Cell Biol. 75, 525–533. Chiang, C., Litingtung, Y., Lee, E., Young, K. E., Corden, J. L., Westphal, H., and Beachy, P. A. (1996). Cyclopia and defective axial patterning in mice lacking Sonic hedgehog gene function. Nature 383, 407–413. Chun, K. T., and Goebl, M. G. (1996). The identification of transposon-tagged mutations in essential genes that aVect cell morphology in Saccharomyces cerevisiae. Genetics 142, 39–50. Cooley, L., Kelley, R., and Spradling, A. (1988). Insertional mutagenesis of the Drosophila genome with single P elements. Science 239, 1121–1128. Corley-Smith, G. E., Brandhorst, B. P., Walker, C., and Postlethwait, J. H. (1999). Production of haploid and diploid androgenetic zebrafish (including methodology for delayed in vitro fertilization). Methods Cell Biol. 59, 45–60. Corley-Smith, G. E., Lim, C. J., and Brandhorst, B. P. (1996). Production of androgenetic zebrafish (Danio rerio). Genetics 142, 1265–1276. Cornell, R. A., and Eisen, J. S. (2000). Delta signaling mediates segregation of neural crest and spinal sensory neurons from zebrafish lateral neural plate. Development 127, 2873–2882. Currie, P. D. (1996). Zebrafish genetics: Mutant cornucopia. Curr. Biol. 6, 1548–1552. Davidson, A. E., Balciunas, D., Mohn, D., ShaVer, J., Hermanson, S., Sivasubbu, S., CliV, M. P., Hackett, P. B., and Ekker, S. C. (2003a). EYcient gene delivery and gene expression in zebrafish using the Sleeping Beauty transposon. Dev. Biol. 263, 191–202. Davidson, A. J., Ernst, P., Wang, Y., Dekens, M. P., Kingsley, P. D., Palis, J., Korsmeyer, S. J., Daley, G. Q., and Zon, L. I. (2003b). cdx4 mutants fail to specify blood progenitors and can be rescued by multiple hox genes. Nature 425, 300–306. Dick, A., Hild, M., Bauer, H., Imai, Y., Maifeld, H., Schier, A. F., Talbot, W. S., Bouwmeester, T., and Hammerschmidt, M. (2000). Essential role of Bmp7 (snailhouse) and its prodomain in dorsoventral patterning of the zebrafish embryo. Development 127, 343–354. Donovan, A., Brownlie, A., Zhou, Y., Shepard, J., Pratt, S. J., Moynihan, J., Paw, B. H., Drejer, A., Barut, B., Zapata, A., Law, T. C., Brugnara, C., Lux, S. E., Pinkus, G. S., Pinkus, J. L., Kingsley, P. D., Palis, J., Fleming, M. D., Andrews, N. C., and Zon, L. I. (2000). Positional cloning of zebrafish ferroportin1 identifies a conserved vertebrate iron exporter. Nature 403, 776–781. Dougan, S. T., Warga, R. M., Kane, D. A., Schier, A. F., and Talbot, W. S. (2003). The role of the zebrafish nodal-related genes squint and cyclops in patterning of mesendoderm. Development 130, 1837–1851.


75

Draper, B. W., Morcos, P. A., and Kimmel, C. B. (2001). Inhibition of zebrafish fgf8 premRNA splicing with morpholino oligos: A quantifiable method for gene knockdown. Genesis 30, 154–156. 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. Eisen, J. S. (1996). Zebrafish make a big splash. Cell 87, 969–977. Ekker, S. C., and Larson, J. D. (2001). Morphant technology in model developmental systems. Genesis 30, 89–93. Fadool, J. M., Hartl, D. L., and Dowling, J. E. (1998). Transposition of the mariner element from Drosophila mauritiana in zebrafish. Proc. Natl. Acad. Sci. USA 95, 5182–5186. Fan, L., Alestrom, A., Alestrom, P., and Collodi, P. (2003). Development of cell cultures with competency for contributing to the zebrafish germ line. Crit. Rev. Eukaryot. Gene Expr. 14, 43–52. Farber, S. A., Pack, M., Ho, S. Y., Johnson, I. D., Wagner, D. S., Dosch, R., Mullins, M. C., Hendrickson, H. S., Hendrickson, E. K., and Halpern, M. E. (2001). Genetic analysis of digestive physiology using fluorescent phospholipid reporters. Science 292, 1385–1388. Ferguson, E. L., and Horvitz, H. R. (1985). Identification and characterization of 22 genes that aVect the vulval cell lineages of the nematode Caenorhabditis elegans. Genetics 110, 17–72. Field, H. A., Dong, P. D., Beis, D., and Stainier, D. Y. (2003a). Formation of the digestive system in zebrafish. II. Pancreas morphogenesis. Dev. Biol. 261, 197–208. Field, H. A., Ober, E. A., Roeser, T., and Stainier, D. Y. (2003b). Formation of the digestive system in zebrafish. I. Liver morphogenesis. Dev. Biol. 253, 279–290. Fisher, S., Amacher, S. L., and Halpern, M. E. (1997). Loss of cerebum function ventralizes the zebrafish embryo. Development 124, 1301–1311. Fritz, A., Rozowski, M., Walker, C., and Westerfield, M. (1996). Identification of selected gamma-ray induced deficiencies in zebrafish using multiplex polymerase chain reaction. Genetics 144, 1735–1745. Gaiano, N., Allende, M., Amsterdam, A., Kawakami, K., and Hopkins, N. (1996a). Highly eYcient germ-line transmission of proviral insertions in zebrafish. Proc. Natl. Acad. Sci. USA 93, 7777–7782. Gaiano, N., Amsterdam, A., Kawakami, K., Allende, M., Becker, T., and Hopkins, N. (1996b). Insertional mutagenesis and rapid cloning of essential genes in zebrafish. Nature 383, 829–832. Garrity, D. M., Childs, S., and Fishman, M. C. (2002). The heartstrings mutation in zebrafish causes heart/fin Tbx5 deficiency syndrome. Development 129, 4635–4645. Gates, M. A., Kim, L., Egan, E. S., Cardozo, T., Sirotkin, H. I., Dougan, S. T., Lashkari, D., Abagyan, R., Schier, A. F., and Talbot, W. S. (1999). A genetic linkage map for zebrafish: Comparative analysis and localization of genes and expressed sequences. Genome Res. 9, 334–347. Gestl, E. E., KauVman, E. J., Moore, J. L., and Cheng, K. C. (1997). New conditions for generation of gynogenetic half-tetrad embryos in the zebrafish (Danio rerio). J. Hered. 88, 76–79. Geyer, P. K., Patton, J. S., Rodesch, C., and Nagoshi, R. N. (1993). Genetic and molecular characterization of P element-induced mutations reveals that the Drosophila ovarian tumor gene has maternal activity and a variable null phenotype. Genetics 133, 265–278. Golling, G., Amsterdam, A., Sun, Z., Antonelli, M., Maldonado, E., Chen, W., Burgess, S., Haldi, M., Artzt, K., Farrington, S., Lin, S. Y., Nissen, R. M., and Hopkins, N. (2002). Insertional mutagenesis in zebrafish rapidly identifies genes essential for early vertebrate development. Nat. Genet. 31, 135–140.

76

Stuckenholz et al.

Gong, Z., Wan, H., Tay, T. L., Wang, H., Chen, M., and Yan, T. (2003). Development of transgenic fish for ornamental and bioreactor by strong expression of fluorescent proteins in the skeletal muscle. Biochem. Biophys. Res. Commun. 308, 58–63. Grapin-Botton, A., and Melton, D. A. (2000). Endoderm development: From patterning to organogenesis. Trends Genet. 16, 124–130. Gritsman, K., Zhang, J., Cheng, S., Heckscher, E., Talbot, W. S., and Schier, A. F. (1999). The EGF-CFC protein one-eyed pinhead is essential for nodal signaling. Cell 97, 121–132. Grunwald, D. J. (1996). A fin-de-siecle achievement: Charting new waters in vertebrate biology. Science 274, 1634–1635. Grunwald, D. J., and Streisinger, G. (1992a). Induction of mutations in the zebrafish with ultraviolet light. Genet. Res. 59, 93–101. Grunwald, D. J., and Streisinger, G. (1992b). Induction of recessive lethal and specific locus mutations in the zebrafish with ethyl nitrosourea. Genet. Res. 59, 103–116. Gualdi, R., Bossard, P., Zheng, M., Hamada, Y., Coleman, J. R., and Zaret, K. S. (1996). Hepatic specification of the gut endoderm in vitro: Cell signaling and transcriptional control. Genes Dev. 10, 1670–1682. 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., Vogelsang, E., Beuchle, D., Schach, U., Fabian, C., and Nu¨ sslein-Volhard, C. (1996). The identification of genes with unique and essential functions in the development of the zebrafish, Danio rerio. Development 123, 1–36. Hagmann, M., Bruggmann, R., Xue, L., Georgiev, O., SchaVner, W., Rungger, D., Spaniol, P., and Gerster, T. (1998). Homologous recombination and DNA-end joining reactions in zygotes and early embryos of zebrafish (Danio rerio) and Drosophila melanogaster. Biol. Chem. 379, 673–681. Haldane, J. B. S. (1956). The detection of autosomal lethals in mice induced by mutagenic agents. J. Genet. 54, 327–342. Hans, S., and Campos-Ortega, J. A. (2002). On the organization of the regulatory region of the zebrafish deltaD gene. Development 129, 4773–4784. Heasman, J., Kofron, M., and Wylie, C. (2000). Beta-catenin signaling activity dissected in the early Xenopus embryo: A novel antisense approach. Dev. Biol. 222, 124–134. Henion, P. D., Raible, D. W., Beattie, C. E., Stoesser, K. L., Weston, J. A., and Eisen, J. S. (1996). Screen for mutations aVecting development of zebrafish neural crest. Dev. Genet. 18, 11–17. Her, G. M., Chiang, C. C., and Wu, J. L. (2004). Zebrafish intestinal fatty acid binding protein (I-FABP) gene promoter drives gut-specific expression in stable transgenic fish. Genesis 38, 26–31. Her, G. M., Yeh, Y. H., and Wu, J. L. (2003). 435-bp liver regulatory sequence in the liver fatty acid binding protein (L-FABP) gene is suYcient to modulate liver regional expression in transgenic zebrafish. Dev. Dyn. 227, 347–356. Hild, M., Dick, A., Rauch, G. J., Meier, A., Bouwmeester, T., HaVter, P., and Hammerschmidt, M. (1999). The smad5 mutation somitabun blocks Bmp2b signaling during early dorsoventral patterning of the zebrafish embryo. Development 126, 2149–2159. Hirth, F., and Reichert, H. (1999). Conserved genetic programs in insect and mammalian brain development. Bioessays 21, 677–684. Holbro, T., and Hynes, N. E. (2004). ErbB receptors: Directing key signaling networks throughout life. Annu. Rev. Pharmacol. Toxicol. 44, 195–217. Holder, N., and McMahon, A. (1996). Genes from zebrafish screens. Nature 384, 515–516. Holt, M., and Oram, S. (1960). Familial heart disease with skeletal malformations. Brit. Heart J. 22, 236–242.


77

Ho¨ rstgen-Schwark, G. (1993). Production of homozygous diploid zebrafish (Brachydanio rerio). Aquaculture 112, 25–37. Hrabe de Angelis, M. H., Flaswinkel, H., Fuchs, H., Rathkolb, B., Soewarto, D., Marschall, S., HeVner, S., Pargent, W., Wuensch, K., Jung, M., Reis, A., Richter, T., Alessandrini, F., Jakob, T., Fuchs, E., Kolb, H., Kremmer, E., Schaeble, K., Rollinski, B., Roscher, A., Peters, C., Meitinger, T., Strom, T., Steckler, T., Holsboer, F., Klopstock, T., Gekeler, F., Schindewolf, C., Jung, T., Avraham, K., Behrendt, H., Ring, J., Zimmer, A., Schughart, K., PfeVer, K., Wolf, E., and Balling, R. (2000). Genome-wide, large-scale production of mutant mice by ENU mutagenesis. Nat. Genet. 25, 444–447. Huang, H., Ju, B., Lee, K. Y., and Lin, S. (2003). Protocol for nuclear transfer in zebrafish. Cloning Stem Cells 5, 333–337. Huang, H., Vogel, S. S., Liu, N., Melton, D. A., and Lin, S. (2001). Analysis of pancreatic development in living transgenic zebrafish embryos. Mol. Cell Endocrinol. 177, 117–124. Itoh, M., Kim, C. H., Palardy, G., Oda, T., Jiang, Y. J., Maust, D., Yeo, S. Y., Lorick, K., Wright, G. J., Ariza-McNaughton, L., Weissman, A. M., Lewis, J., Chandrasekharappa, S. C., and Chitnis, A. B. (2003). Mind bomb is a ubiquitin ligase that is essential for eYcient activation of Notch signaling by Delta. Dev. Cell 4, 67–82. Ivics, Z., Hackett, P. B., Plasterk, R. H., and Izsva´ k, Z. (1997). Molecular reconstruction of Sleeping Beauty, a Tc1-like transposon from fish, and its transposition in human cells. Cell 91, 501–510. Ivics, Z., Izsva´ k, Z., and Hackett, P. B. (1999). Genetic applications of transposons and other repetitive elements in zebrafish. Methods Cell Biol. 60, 99–131. Izsva´ k, Z., Ivics, Z., and Plasterk, R. H. (2000). Sleeping Beauty, a wide host-range transposon vector for genetic transformation in vertebrates. J. Mol. Biol. 302, 93–102. Jin, Y., Blue, E. K., Dixon, S., Shao, Z., and Gallagher, P. J. (2002). A death-associated protein kinase (DAPK)-interacting protein, DIP-1, is an E3 ubiquitin ligase that promotes tumor necrosis factor-induced apoptosis and regulates the cellular levels of DAPK. J. Biol. Chem. 277, 46980–46986. Jung, J., Zheng, M., Goldfarb, M., and Zaret, K. S. (1999). Initiation of mammalian liver development from endoderm by fibroblast growth factors. Science 284, 1998–2003. Kane, D. A., and Kimmel, C. B. (1993). The zebrafish midblastula transition. Development 119, 447–456. Karlstrom, R. O., Talbot, W. S., and Schier, A. F. (1999). Comparative synteny cloning of zebrafish you-too: Mutations in the Hedgehog target gli2 aVect ventral forebrain patterning. Genes Dev. 13, 388–393. Karlstrom, R. O., Trowe, T., Klostermann, S., Baier, H., Brand, M., Crawford, A. D., Grunewald, B., HaVter, P., HoVmann, H., Meyer, S. U., Muller, B. K., Richter, S., van Eeden, F. J., Nu¨ sslein-Volhard, C., and BonhoeVer, F. (1996). Zebrafish mutations aVecting retinotectal axon pathfinding. Development 123, 427–438. Kawakami, K., Amsterdam, A., Shimoda, N., Becker, T., Mugg, J., Shima, A., and Hopkins, N. (2000). Proviral insertions in the zebrafish hagoromo gene, encoding an F-box/WD40-repeat protein, cause stripe pattern anomalies. Curr. Biol. 10, 463–466. Kelsh, R. N., and Raible, D. W. (2002). Specification of zebrafish neural crest. Results Probl. Cell DiVer. 40, 216–236. Keng, V. W., Fujimori, K. E., Myint, Z., Tamamaki, N., Nojyo, Y., and Noguchi, T. (1998). Expression of Hex mRNA in early murine postimplantation embryo development. FEBS Lett. 426, 183–186. Keng, V. W., Yagi, H., Ikawa, M., Nagano, T., Myint, Z., Yamada, K., Tanaka, T., Sato, A., Muramatsu, I., Okabe, M., Sato, M., and Noguchi, T. (2000). Homeobox gene Hex is essential for onset of mouse embryonic liver development and diVerentiation of the monocyte lineage. Biochem. Biophys. Res. Commun. 276, 1155–1161.

78

Stuckenholz et al.

Kho, A. T., Zhao, Q., Cai, Z., Butte, A. J., Kim, J. Y., Pomeroy, S. L., Rowitch, D. H., and Kohane, I. S. (2004). Conserved mechanisms across development and tumorigenesis revealed by a mouse development perspective of human cancers. Genes Dev. 18, 629–640. Kikuchi, Y., Agathon, A., Alexander, J., Thisse, C., Waldron, S., Yelon, D., Thisse, B., and Stainier, D. Y. (2001). casanova encodes a novel Sox-related protein necessary and suYcient for early endoderm formation in zebrafish. Genes Dev. 15, 1493–1505. Kikuchi, Y., Trinh, L. A., Reiter, J. F., Alexander, J., Yelon, D., and Stainier, D. Y. (2000). The zebrafish bonnie and clyde gene encodes a Mix family homeodomain protein that regulates the generation of endodermal precursors. Genes Dev. 14, 1279–1289. Kile, B. T., Hentges, K. E., Clark, A. T., Nakamura, H., Salinger, A. P., Liu, B., Box, N., Stockton, D. W., Johnson, R. L., Behringer, R. R., Bradley, A., and Justice, M. J. (2003). Functional genetic analysis of mouse chromosome 11. Nature 425, 81–86. Kishimoto, Y., Lee, K. H., Zon, L., Hammerschmidt, M., and Schulte-Merker, S. (1997). The molecular nature of zebrafish swirl: BMP2 function is essential during early dorsoventral patterning. Development 124, 4457–4466. Knapik, E. W. (2000). ENU mutagenesis in zebrafish—From genes to complex diseases. Mamm. Genome 11, 511–519. Lagos-Quintana, M., Rauhut, R., Lendeckel, W., and Tuschl, T. (2001). Identification of novel genes coding for small expressed RNAs. Science 294, 853–858. 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. Lawson, N. D., and Weinstein, B. M. (2002). In vivo imaging of embryonic vascular development using transgenic zebrafish. Dev. Biol. 248, 307–318. Lee, K. Y., Huang, H., Ju, B., Yang, Z., and Lin, S. (2002). Cloned zebrafish by nuclear transfer from long-term-cultured cells. Nat. Biotechnol. 20, 795–799. Lee, M. G., and Nurse, P. (1987). Complementation used to clone a human homologue of the fission yeast cell cycle control gene cdc2. Nature 327, 31–35. Lewis, K. E., and Eisen, J. S. (2001). Hedgehog signaling is required for primary motoneuron induction in zebrafish. Development 128, 3485–3495. Lim, L. P., Glasner, M. E., Yekta, S., Burge, C. B., and Bartel, D. P. (2003). Vertebrate microRNA genes. Science 299, 1540. Linney, E., Hardison, N. L., Lonze, B. E., Lyons, S., and DiNapoli, L. (1999). Transgene expression in zebrafish: A comparison of retroviral-vector and DNA-injection approaches. Dev. Biol. 213, 207–216. Linney, E., and Udvadia, A. J. (2004). Construction and detection of fluorescent, germline transgenic zebrafish. Methods Mol. Biol. 254, 271–288. Ma, C., Fan, L., Ganassin, R., Bols, N., and Collodi, P. (2001). Production of zebrafish germ-line chimeras from embryo cell cultures. Proc. Natl. Acad. Sci. USA 98, 2461–2466. Mayer, U., Torres Ruiz, R. A., Berleth, T., Mise´ ra, S., and Ju¨ rgens, G. (1991). Mutations aVecting body organization in the Arabidopsis embryo. Nature 353, 402–407. Miskey, C., Izsva´ k, Z., Plasterk, R. H., and Ivics, Z. (2003). The Frog Prince: A reconstructed transposon from Rana pipiens with high transpositional activity in vertebrate cells. Nucleic Acids Res. 31, 6873–6881. Moens, C. B., Yan, Y. L., Appel, B., Force, A. G., and Kimmel, C. B. (1996). valentino: A zebrafish gene required for normal hindbrain segmentation. Development 122, 3981–3990. Mohideen, M. A., Beckwith, L. G., Tsao-Wu, G. S., Moore, J. L., Wong, A. C., Chinoy, M. R., and Cheng, K. C. (2003). Histology-based screen for zebrafish mutants with abnormal cell diVerentiation. Dev. Dyn. 228, 414–423.


79

Mullins, M. C., Hammerschmidt, M., HaVter, P., and Nu¨ sslein-Volhard, C. (1994). Large-scale mutagenesis in the zebrafish: In search of genes controlling development in a vertebrate. Curr. Biol. 4, 189–202. Mullins, M. C., Hammerschmidt, M., Kane, D. A., Odenthal, J., Brand, M., van Eeden, F. J. M., Furutani-Seiki, M., Granato, M., HaVter, P., Heisenberg, C.-P., Jiang, Y.-J., Kelsh, R. N., and Nu¨ sslein-Volhard, C. (1996). Genes establishing dorsoventral pattern formation in the zebrafish embryo: The ventral specifying genes. Development 123, 81–93. Nasevicius, A., and Ekker, S. C. (2000). EVective targeted gene ‘‘knockdown’’ in zebrafish. Nat. Genet. 26, 216–220. Nguyen, V. H., Schmid, B., Trout, J., Connors, S. A., Ekker, M., and Mullins, M. C. (1998). Ventral and lateral regions of the zebrafish gastrula, including the neural crest progenitors, are established by a bmp2b/swirl pathway of genes. Dev. Biol. 199, 93–110. Nguyen, V. H., Trout, J., Connors, S. A., Andermann, P., Weinberg, E., and Mullins, M. C. (2000). Dorsal and intermediate neuronal cell types of the spinal cord are established by a BMP signaling pathway. Development 127, 1209–1220. Nolan, P. M., Peters, J., Strivens, M., Rogers, D., Hagan, J., Spurr, N., Gray, I. C., Vizor, L., Brooker, D., Whitehill, E., Washbourne, R., Hough, T., Greenaway, S., Hewitt, M., Liu, X., McCormack, S., Pickford, K., Selley, R., Wells, C., Tymowska-Lalanne, Z., Roby, P., Glenister, P., Thornton, C., Thaung, C., Stevenson, J. A., Arkell, R., Mburu, P., Hardisty, R., Kiernan, A., Erven, A., Steel, K. P., Voegeling, S., Guenet, J. L., Nickols, C., Sadri, R., Nasse, M., Isaacs, A., Davies, K., Browne, M., Fisher, E. M., Martin, J., Rastan, S., Brown, S. D., and Hunter, J. (2000). A systematic, genome-wide, phenotype-driven mutagenesis programme for gene function studies in the mouse. Nat. Genet. 25, 440–443. Nu¨ sslein-Volhard, C., and Wieschaus, E. (1980). Mutations aVecting segment number and polarity in Drosophila. Nature 287, 795–801. Ober, E. A., Field, H. A., and Stainier, D. Y. (2003). From endoderm formation to liver and pancreas development in zebrafish. Mech. Dev. 120, 5–18. Odenthal, J., van Eeden, F. J., HaVter, P., Ingham, P. W., and Nu¨ sslein-Volhard, C. (2000). Two distinct cell populations in the floor plate of the zebrafish are induced by diVerent pathways. Dev. Biol. 219, 350–363. Pack, M., Solnica-Krezel, L., Malicki, J., Neuhauss, S. C., Schier, A. F., Stemple, D. L., Driever, W., and Fishman, M. C. (1996). Mutations aVecting development of zebrafish digestive organs. Development 123, 321–328. Paddison, P. J., Silva, J. M., Conklin, D. S., Schlabach, M., Li, M., Aruleba, S., Balija, V., O’Shaughnessy, A., Gnoj, L., Scobie, K., Chang, K., Westbrook, T., Cleary, M., Sachidanandam, R., McCombie, W. R., Elledge, S. J., and Hannon, G. J. (2004). A resource for large-scale RNA-interference-based screens in mammals. Nature 428, 427–431. Panganiban, G., and Rubenstein, J. L. (2002). Developmental functions of the Distalless/Dlx homeobox genes. Development 129, 4371–4386. Papan, C., and Campos-Ortega, J. A. (1994). On the formation of the neural keel and neural tube in the zebrafish Danio (Brachydanio) rerio. Roux’s Archives of Developmental Biology 203, 178–186. Parinov, S., and Sundaresan, V. (2000). Functional genomics in Arabidopsis: Large-scale insertional mutagenesis complements the genome sequencing project. Curr. Opin. Biotechnol. 11, 157–161. Park, H. C., and Appel, B. (2003). Delta-Notch signaling regulates oligodendrocyte specification. Development 130, 3747–3755. Patient, R. K., and McGhee, J. D. (2002). The GATA family (vertebrates and invertebrates). Curr. Opin. Genet. Dev. 12, 416–422.

80

Stuckenholz et al.

Pierce, M., Wang, Y., Denovan-Wright, E. M., and Wright, J. M. (2000). Nucleotide sequence of a cDNA clone coding for an intestinal-type fatty acid binding protein and its tissuespecific expression in zebrafish (Danio rerio). Biochim. Biophys. Acta 1490, 175–183. Postlethwait, J. H., Yan, Y. L., Gates, M. A., Horne, S., Amores, A., Brownlie, A., Donovan, A., Egan, E. S., Force, A., Gong, Z., Goutel, C., Fritz, A., Kelsh, R., Knapik, E., Liao, E., Paw, B., Ransom, D., Singer, A., Thomson, M., Abduljabbar, T. S., Yelick, P., Beier, D., Joly, J. S., Larhammar, D., Rosa, F., et al. (1998). Vertebrate genome evolution and the zebrafish gene map. Nat. Genet. 18, 345–349. Poulain, M., and Lepage, T. (2002). Mezzo, a paired-like homeobox protein is an immediate target of Nodal signalling and regulates endoderm specification in zebrafish. Development 129, 4901–4914. Rawdon, B. B. (2001). Early development of the gut: New light on an old hypothesis. Cell Biol. Int. 25, 9–15. Raz, E., van Luenen, H. G., Schaerringer, B., Plasterk, R. H., and Driever, W. (1998). Transposition of the nematode Caenorhabditis elegans Tc3 element in the zebrafish Danio rerio. Curr. Biol. 8, 82–88. Reiter, J. F., Alexander, J., Rodaway, A., Yelon, D., Patient, R., Holder, N., and Stainier, D. Y. (1999). Gata5 is required for the development of the heart and endoderm in zebrafish. Genes. Dev. 13, 2983–2995. Reiter, J. F., Kikuchi, Y., and Stainier, D. Y. (2001). Multiple roles for Gata5 in zebrafish endoderm formation. Development 128, 125–135. Riley, B. B., and Grunwald, D. J. (1995). EYcient induction of point mutations allowing recovery of specific locus mutations in zebrafish. Proc. Natl. Acad. Sci. USA 92, 5997–6001. Roberts, D. J. (2000). Molecular mechanisms of development of the gastrointestinal tract. Dev. Dyn. 219, 109–120. Roelink, H., Porter, J. A., Chiang, C., Tanabe, Y., Chang, D. T., Beachy, P. A., and Jessell, T. M. (1995). Floor plate and motor neuron induction by diVerent concentrations of the aminoterminal cleavage product of sonic hedgehog autoproteolysis. Cell 81, 445–455. Schauerte, H. E., van Eeden, F. J. M., Fricke, C., Odenthal, J., Stra¨ hle, U., and HaVter, P. (1998). Sonic hedgehog is not required for the induction of medial floor plate cells in the zebrafish. Development 125, 2983–2993. Scheer, N., and Campos-Ortega, J. A. (1999). Use of the Gal4-UAS technique for targeted gene expression in the zebrafish. Mech. Dev. 80, 153–158. Scheer, N., Groth, A., Hans, S., and Campos-Ortega, J. A. (2001). An instructive function for Notch in promoting gliogenesis in the zebrafish retina. Development 128, 1099–1107. Schmid, B., Furthauer, M., Connors, S. A., Trout, J., Thisse, B., Thisse, C., and Mullins, M. C. (2000). Equivalent genetic roles for bmp7/snailhouse and bmp2b/swirl in dorsoventral pattern formation. Development 127, 957–967. Schulte-Merker, S., van Eeden, F. J., Halpern, M. E., Kimmel, C. B., and Nu¨ sslein-Volhard, C. (1994). no tail (ntl) is the zebrafish homologue of the mouse T (Brachyury) gene. Development 120, 1009–1015. Shedlovsky, A., King, T. R., and Dove, W. F. (1988). Saturation germ line mutagenesis of the murine t region including a lethal allele at the quaking locus. Proc. Natl. Acad. Sci. USA 85, 180–184. Shin, J. T., and Fishman, M. C. (2002). From Zebrafish to human: Modular medical models. Annu. Rev. Genomics Hum. Genet. 3, 311–340. Solnica-Krezel, L., Schier, A. F., and Driever, W. (1994). EYcient recovery of ENU-induced mutations from the zebrafish germline. Genetics 136, 1401–1420. Solomon, K. S., and Fritz, A. (2002). Concerted action of two dlx paralogs in sensory placode formation. Development 129, 3127–3136.


81

Stainier, D. Y. (2002). A glimpse into the molecular entrails of endoderm formation. Genes. Dev. 16, 893–907. Streisinger, G., Singer, F., Walker, C., Knauber, D., and Dower, N. (1986). Segregation analyses and gene-centromere distances in zebrafish. Genetics 112, 311–319. Streisinger, G., Walker, C., Dower, N., Knauber, D., and Singer, F. (1981). Production of clones of homozygous diploid zebra fish (Brachydanio rerio). Nature 291, 293–296. Stuckenholz, C., Meller, V. H., and Kuroda, M. I. (2003). Functional redundancy within roX1, a noncoding RNA involved in dosage compensation in Drosophila melanogaster. Genetics 164, 1003–1014. Tam, P. P., Kanai-Azuma, M., and Kanai, Y. (2003). Early endoderm development in vertebrates: Lineage diVerentiation and morphogenetic function. Curr. Opin. Genet. Dev. 13, 393–400. Trowe, T., Klostermann, S., Baier, H., Granato, M., Crawford, A. D., Grunewald, B., HoVmann, H., Karlstrom, R. O., Meyer, S. U., Muller, B., Richter, S., Nu¨ sslein-Volhard, C., and BonhoeVer, F. (1996). Mutations disrupting the ordering and topographic mapping of axons in the retinotectal projection of the zebrafish, Danio rerio. Development 123, 439–450. Udvadia, A. J., and Linney, E. (2003). Windows into development: Historic, current, and future perspectives on transgenic zebrafish. Dev. Biol. 256, 1–17. Urtishak, K. A., Choob, M., Tian, X., Sternheim, N., Talbot, W. S., Wickstrom, E., and Farber, S. A. (2003). Targeted gene knockdown in zebrafish using negatively charged peptide nucleic acid mimics. Dev. Dyn. 228, 405–413. van der Weyden, L., Adams, D. J., and Bradley, A. (2002). Tools for targeted manipulation of the mouse genome. Physiol. Genomics 11, 133–164. van Eeden, F. J., Granato, M., Odenthal, J., and HaVter, P. (1999). Developmental mutant screens in the zebrafish. Methods. Cell Biol. 60, 21–41. Varga, Z. M., Amores, A., Lewis, K. E., Yan, Y. L., Postlethwait, J. H., Eisen, J. S., and Westerfield, M. (2001). Zebrafish smoothened functions in ventral neural tube specification and axon tract formation. Development 128, 3497–3509. Veraksa, A., Del Campo, M., and McGinnis, W. (2000). Developmental patterning genes and their conserved functions: From model organisms to humans. Mol. Genet. Metab. 69, 85–100. Walker, C. (1999). Haploid screens and gamma-ray mutagenesis. Methods Cell Biol. 60, 43–70. Walker, C., and Streisinger, G. (1983). Induction of mutations by gamma-rays in pregonial germ cells of zebrafish embryos. Genetics 103, 125–136. Wallace, K. N., and Pack, M. (2003). Unique and conserved aspects of gut development in zebrafish. Dev. Biol. 255, 12–29. Wallace, K. N., YusuV, S., Sonntag, J. M., Chin, A. J., and Pack, M. (2001). Zebrafish hhex regulates liver development and digestive organ chirality. Genesis 30, 141–143. Wan, H., He, J., Ju, B., Yan, T., Lam, T. J., and Gong, Z. (2002). Generation of two-color transgenic zebrafish using the green and red fluorescent protein reporter genes gfp and rfp. Mar. Biotechnol. (NY) 4, 146–154. Wang, L., Fan, C., Topol, S. E., Topol, E. J., and Wang, Q. (2003). Mutation of MEF2A in an inherited disorder with features of coronary artery disease. Science 302, 1578–1581. Weber, H., Symes, C. E., Walmsley, M. E., Rodaway, A. R., and Patient, R. K. (2000). A role for GATA5 in Xenopus endoderm specification. Development 127, 4345–4360. Weiss, M. J., and Orkin, S. H. (1995). GATA transcription factors: Key regulators of hematopoiesis. Exp. Hematol. 23, 99–107. Wienholds, E., Koudijs, M. J., van Eeden, F. J., Cuppen, E., and Plasterk, R. H. (2003a). The microRNA-producing enzyme Dicer1 is essential for zebrafish development. Nat. Genet. 35, 217–218.

82

Stuckenholz et al.

Wienholds, E., Schulte-Merker, S., Walderich, B., and Plasterk, R. H. (2002). Target-selected inactivation of the zebrafish rag1 gene. Science 297, 99–102. Wienholds, E., van Eeden, F., Kosters, M., Mudde, J., Plasterk, R. H., and Cuppen, E. (2003b). EYcient target-selected mutagenesis in zebrafish. Genome. Res. 13, 2700–2707. Wutz, A., Rasmussen, T. P., and Jaenisch, R. (2002). Chromosomal silencing and localization are mediated by diVerent domains of Xist RNA. Nat. Genet. 30, 167–174. Yandell, M. D., Edgar, L. G., and Wood, W. B. (1994). Trimethylpsoralen induces small deletion mutations in Caenorhabditis elegans. Proc. Natl. Acad. Sci. USA 91, 1381–1385. Yarden, Y. (2001). The EGFR family and its ligands in human cancer: Signalling mechanisms and therapeutic opportunities. Eur. J. Cancer 37(Suppl 4), S3–8. Yee, N. S., YusuV, S., and Pack, M. (2001). Zebrafish pdx1 morphant displays defects in pancreas development and digestive organ chirality, and potentially identifies a multipotent pancreas progenitor cell. Genesis 30, 137–140. Yekta, S., Shih, I. H., and Bartel, D. P. (2004). MicroRNA-directed cleavage of HOXB8 mRNA. Science 304, 594–596. Zayed, H., Izsvak, Z., Walisko, O., and Ivics, Z. (2004). Development of hyperactive sleeping beauty transposon vectors by mutational analysis. Mol. Ther. 9, 292–304. Zhang, J., Talbot, W. S., and Schier, A. F. (1998). Positional cloning identifies zebrafish oneeyed pinhead as a permissive EGF-related ligand required during gastrulation. Cell 92, 241–251.

3

____________________________________________________________________________

Synaptic Vesicle Docking: A Putative Role for the Munc18/Sec1 Protein Family Robby M. Weimer * and Janet E. Richmond { *Howard Hughes Medical Institute Cold Spring Harbor Laboratory Cold Spring Harbor, New York 11724 { Department of Biological Sciences University of Illinois-Chicago Chicago, Illinois 60607

I. Introduction II. Synaptic Transmission and the Synaptic Vesicle Cycle III. The Molecules and Mechanisms Driving Synaptic Vesicle Exocytosis A. Fusion B. Calcium Sensing C. Priming D. Docking IV. The Sec1p/Munc-18 (SM) Protein Family V. SM Mutant Phenotypes A. SM Mutants in Yeast B. SM Mutants and the Synapse VI. SM Protein Interactions: Putative Links That Anchor Synaptic Vesicles A. Plasma Membrane-Associated Proteins That Interact with SM Proteins B. Vesicle-Associated Proteins That Interact with SM Proteins VII. SM Proteins Promote Vesicle Docking Indirectly: An Alternative Hypothesis A. SNARE Sorting at the Synapse B. SNARE TraYcking to the Synapse VIII. Summary and Future Directions References

I. Introduction At presynaptic nerve terminals, neurotransmitter is stored in small membranebound organelles, termed ‘‘synaptic vesicles.’’ In order to release their contents into the synaptic cleft via membrane fusion, synaptic vesicles must first associate, or ‘‘dock,’’ with the plasma membrane. Recent data suggest that the cytoplasmic protein UNC-18 (a.k.a. nSec1/Munc18-1/rbsec-1 or Rop) promotes synaptic vesicle docking through an as yet uncharacterized pathway. Here, we present an overview of synaptic vesicle exocytosis, review the Current Topics in Developmental Biology, Vol. 65 Copyright 2005, Elsevier Inc. All rights reserved. 0070-2153/05 $35.00

83

84

Weimer and Richmond

current data regarding UNC-18 and the Sec1/Munc18 (SM) protein family in membrane traYcking, and propose several models which could describe the molecular mechanism by which UNC-18 promotes synaptic vesicle docking.

II. Synaptic Transmission and the Synaptic Vesicle Cycle Synaptic transmission is the process by which information is transferred from a neuron to an adjacent cell via a specialized site of cell–cell contact termed the synapse. At chemical synapses, synaptic transmission is mediated by the release of chemical neurotransmitter from the presynaptic cell, which diVuses across the synaptic cleft and activates cognate postsynaptic receptors. Thus, the eYcacy of chemical synaptic transmission relies, at least in part, upon the regulated release of chemical neurotransmitter from presynaptic nerve terminals—a responsibility bestowed upon a specialized membrane traYcking pathway: the synaptic vesicle cycle. The synaptic vesicle cycle resembles general membrane traYcking1 and can be divided into three phases: synaptic vesicle filling, exocytosis, and recycling (Fig. 1). During filling, neurotransmitter synthesized in the cytoplasm is loaded into synaptic vesicles via vesicular neurotransmitter transporters (reviewed in Schuldiner et al., 1995). The energy used by vesicular transporters to concentrate neurotransmitter in vesicles comes from an electrochemical gradient established by the vacuolar-type proton ATPase (Anderson et al., 1982; Hell et al., 1990; Moriyama et al., 1990). Exocytosis of neurotransmitter-filled synaptic vesicles is a multistep pathway (Fig. 2). First, vesicles associate, or dock, with the plasma membrane. Second, vesicles undergo a maturation step, termed ‘‘priming,’’ in which vesicles become fusion competent. Third, increases in cytoplasmic calcium levels resulting from calcium channel activation are sensed, triggering the final step, fusion, in which synaptic vesicle and plasma membranes fuse and allow neurotransmitter release. After fusion, synaptic vesicle membrane is incorporated into 1

The basic mechanisms that operate during membrane traYcking events in eukaryotic cells resemble synaptic vesicle exocytosis both morphologically and molecularly. For example, traYcking of secreted proteins through the secretory pathway requires budding of transport vesicles containing secreted proteins from the endoplasmic reticulum, which then fuse with golgi membrane; repetitive fission/fusion events shuttle these proteins through the golgi network until they are packaged into a vesicle bound to fuse with the plasma membrane. At the molecular level, each membrane traYcking pathway appears to require members of the conserved SNARE, Rab, and SM protein families (Jahn et al., 2003)., suggesting that the molecular mechanism of membrane traYcking is highly conserved. Thus, insight gained from the study of all membrane traYcking pathways has advanced the understanding of synaptic vesicle exocytosis and vice versa.

3. UNC-18 Promotes Synaptic Vesicle Docking

85

Figure 1 The synaptic vesicle cycle. During filling, neurotransmitters are synthesized in the cytoplasm and transported into the lumen of the synaptic vesicle. During synaptic vesicle exocytosis, vesicles dock with the plasma membrane near a release site, undergo a maturation step, termed priming, and upon a calcium signal, fuse with the plasma membrane to release neurotransmitter. Synaptic vesicle membrane and membrane proteins are recycled from the plasma membrane via endocytosis.

Figure 2 The four steps of synaptic vesicle exocytosis: (A) docking, in which a synaptic vesicle becomes tethered to the plasma membrane; (B) priming, a maturation step in which a docked vesicle becomes competent to fuse with the plasma membrane; (C) calcium signaling, in which the calcium influx due to the activation of voltage-gated calcium channels is sensed; (D) fusion, lipid mixing which permits neurotransmitter release. Abbreviations: snt, synaptotagmin; snb, synaptobrevin; syx, syntaxin; s25, SNAP25; VGCC, voltage-gated calcium channel.

86

Weimer and Richmond

the plasma membrane. Thus synaptic vesicle components must be recycled from the plasma membrane for continued neurotransmitter release (reviewed in Morgan et al., 2002). Advances over the past 20 years have begun to shed light upon the molecular mechanisms driving synaptic vesicle cycling, in particular, synaptic vesicle exocytosis.

III. The Molecules and Mechanisms Driving Synaptic Vesicle Exocytosis One may imagine the synaptic vesicle exocytosis pathway beginning with the association, or docking, of a synaptic vesicle with the plasma membrane. However, when discussing the molecular events which contribute to exocytosis, it is often easier to begin the discussion at the end point, that is, fusion. A. Fusion The final step of synaptic vesicle exocytosis, the fusion of the synaptic vesicle membrane with the plasma membrane, occurs rapidly, within milliseconds of calcium entry (Llinas et al., 1981; Sabatini and Regehr, 1996). Vesicle fusion is thought to progress through an intermediary structure, termed the ‘‘fusion pore’’ (reviewed in Jahn et al., 2003), and is mediated by a conserved family of proteins referred to as SNAREs (soluble NSF attachment protein receptors). SNAREs are 18 to 32 kDa membrane-associated proteins that interact to form SNARE complexes (reviewed in Chen and Scheller, 2001). SNARE complexes at the synapse contain a bundle of four -helices in which one -helix is donated by the vesicle SNARE (v-SNARE, a.k.a. R-SNARE, named for a conserved arginine (R) in the helix) synaptobrevin and three -helices are donated by the target SNAREs (t-SNAREs, a.k.a. Q-SNAREs, named for a conserved glutamine (Q) residue in the helix) SNAP-25 and syntaxin (Bennett and Scheller, 1993; Hanson et al., 1997; Lonart and Sudhof, 2000; Sollner et al., 1993b; Sutton et al., 1998)—these -helical regions are referred to as SNARE domains. In the predicted crystal structure of the four -helical SNARE bundle, the -helices are aligned in parallel, a configuration which in trans would result in close apposition of the synaptic vesicle and plasma membranes upon SNARE complex formation (Sutton et al., 1998). Several lines of experimental data suggest that the SNARE proteins alone can mediate membrane fusion. First, both clostridial neurotoxins, which act by directly cleaving SNARE proteins, and antibodies directed against SNAREs inhibit fusion, even when applied only moments before exocytosis (Chen et al., 1999, 2001; Xu et al., 1999). Second, in the absence of the SNARE protein syntaxin, vesicles fail to fuse with the plasma membrane


87

(Aravamudan et al., 1999). Third, reconstituted vesicles containing only the SNARE proteins undergo membrane fusion in vitro (Weber et al., 1998). These data together suggest that the SNARE proteins represent the minimal membrane fusion machinery. Membrane fusion progresses through a fusion pore intermediary (Jahn et al., 2003). Recent data implicate the transmembrane domain of syntaxin as a constituent of this structure. Specifically, tryptophan-scanning experiments identified three residues in the -helical transmembrane domain of syntaxin which, when mutated to tryptophans, result, specifically, in decreased fusion pore conductance (Han et al., 2004). Fusion pore conductance can be measured by amperometric recordings of neuroendocrine release (Chow et al., 1992; Han et al., 2004)—a regulated exocytosis pathway which utilizes much of the same machinery as synaptic vesicle exocytosis (Zucker, 1996). The three syntaxin residues are predicted to align along the same face of the syntaxin -helical transmembrane domain, which suggests that this region of syntaxin constitutes, at least in part, the lining of the fusion pore (Han et al., 2004). Conductance-based pore size estimates indicate that a fusion pore could contain as many as 5 to 8 syntaxin molecules contributing their transmembrane domain to a barrel-shaped pore (Han et al., 2004). Perhaps similar structures exist in the vesicle membrane, that is, the transmembrane domain of several synaptobrevin molecules could align and contribute the vesicle portion of the fusion pore (Han et al., 2004). From these data and ideas, one can begin to paint a detailed picture of membrane fusion: the zippering together of the SNARE domains in transSNARE complexes brings the synaptic vesicle and plasma membranes into close proximity while aligning the transmembrane domains of syntaxin and synaptobrevin. This configuration then contributes to an intermediary fusion pore that catalyzes lipid mixing and membrane fusion. One must point out, however, that the rates of SNARE-only mediated fusion in vitro are low in comparison to the rate of vesicle fusion at the synapse (Goda and Stevens, 1994; Parlati et al., 1999; Weber et al., 1998). Addition of the putative calcium sensor synaptotagmin (discussed in the following text) increases, in a calcium-dependent manner, the rate of fusion in vitro (Tucker et al., 2004), but not to a rate comparable to those observed in vivo. This suggests that in vivo other factors contribute, during or after SNARE complex formation, to mediate fusion.

B. Calcium Sensing The latency between calcium entry and neurotransmitter release, which can be as little as 60 microseconds (Llinas et al., 1981; Sabatini and Regehr, 1996), suggests that the calcium sensor which links membrane fusion to

88

Weimer and Richmond

calcium entry is intimately associated with the fusion machinery: the SNAREs. Several lines of evidence implicate the synaptic vesicle protein synaptotagmin as the calcium sensor. First, synaptotagmins are not found in all eukaryotes. This suggests that synaptotagmin is not a component of the core membrane traYcking machinery, such as the SNAREs, but is rather an accessory protein which may regulate neurotransmitter release. Second, synaptotagmin contains two C2 calcium-binding motifs that have been shown to bind both calcium (Shao et al., 1996; Ubach et al., 1998) and phospholipids (Brose et al., 1992; Davletov and Sudhof, 1993; Zhang et al., 1998). Third, synaptotagmin is required for calcium-dependent neurotransmitter release in vivo (Geppert et al., 1994b; Littleton et al., 1993, 1994), and dramatically increases the rate of SNARE-mediated membrane fusion in vitro in a calcium-dependent manner (Tucker et al., 2004). Fourth, mutations that result in incremental changes in the aYnity of synaptotagmin for calcium incrementally change the calcium dependence of release (Fernandez-Chacon et al., 2001). Currently, the mechanism by which synaptotagmin promotes membrane fusion is under debate. Does synaptotagmin promote membrane fusion through interactions with the SNARE proteins, through interactions with phospholipids independent of SNARE interactions, or both? Synaptotagmin interacts with the SNARE proteins syntaxin (Chapman et al., 1995) and SNAP-25 (Schiavo et al., 1997). Calcium increases the aYnity of synaptotagmin for the SNARE proteins (Bai et al., 2004; Tucker et al., 2004), and mutations in synaptotagmin that impair these interactions adversely aVect fusion (Bai et al., 2004; Tucker et al., 2004). Together, these data suggest that synaptotagmin promotes calcium-dependent fusion via interactions with the SNAREs—likely, by promoting the complete zippering together of the SNARE complex (Brose et al., 1992; Ernst and Brunger, 2003; Kee and Scheller, 1996; Pevsner et al., 1994; Schiavo et al., 1997; Wu et al., 1999). However, if Sr2+ is used as a ‘‘calcium agonist’’ to probe regulated release, synaptotagmin, under these conditions, can promote membrane fusion independent of SNARE interactions (Shin et al., 2003)—presumably through interactions with phospholipids (Brose et al., 1992; Davletov and Sudhof, 1993; Shin et al., 2003; Zhang et al., 1998). Thus, it is plausible that synaptotagmin may act in parallel with SNAREs to facilitate fusion. It is interesting to note that, in addition to promoting calcium-dependent release, synaptotagmin can also inhibit membrane fusion in vitro when calcium is absent (Tucker et al., 2004). This apparent duality suggests that in vivo synaptotagmin is positioned to act as the switch that dictates when a vesicle poised to fuse can undergo exocytosis. The implications of which will be discussed, in detail, shortly. In addition to synaptotagmin, a number of other proteins have been implicated in calcium sensing during regulated release, primarily due to the


89

fact they contain C2 calcium-binding motifs. This list includes the presynaptic proteins UNC-13 (Maruyama and Brenner, 1991), Rim (Wang et al., 1997), Doc2 (Orita et al., 1995), and Rabphillin (Shirataki et al., 1993); current functional data fail to support a role for these proteins acting as the calcium sensor during fast-regulated release. However, calcium may modulate synaptic activity via these proteins. Functional data does support a role for the small transmembrane protein complexin in a late step of regulated release—either in calcium sensing or fusion (Reim et al., 2001; Tokumaru et al., 2001). Specifically, neurons in complexin knockout mice exhibit reduced neurotransmitter release eYciency due to reduced calcium sensitivity (Reim et al., 2001). Complexin interacts with SNARE complexes and promotes oligomerization of multiple SNARE complexes in vitro (Hu et al., 2002; Tokumaru et al., 2001); such a function may be required in vivo to align SNARE complexes for eYcient fusion—perhaps to promote a syntaxin/synaptobrevin-lined fusion pore, as has been discussed.

C. Priming The short latency between calcium entry into the presynaptic nerve terminal and neurotransmitter release implies that there is a population of vesicles poised to fuse with the plasma membrane upon calcium entry. In fact, a requirement for such a maturation step was originally inferred from the observation that synaptic vesicle fusion is inhibited before the pool of docked vesicles is depleted (Wickelgren et al., 1985). Presumably, the remaining docked vesicles have yet to undergo priming and are therefore fusion incompetent. With respect to synaptic vesicle exocytosis, this maturation step is referred to as ‘‘priming’’ (Robinson and Martin, 1998). At the molecular level, priming is thought to require interactions between SNAREs in trans (Broadie et al., 1995; Chen et al., 2001; Hanson et al., 1997; Lonart and Sudhof, 2000; Sollner et al., 1993b). Specifically, plasma membrane-associated syntaxin and SNAP-25 interact in trans with vesicle-bound synaptobrevin. Two lines of evidence support a role for the SNARE proteins in the priming step of exocytosis. First, SNARE protein interactions can be detected prior to membrane fusion (Hua and Charlton, 1999; Lonart and Sudhof, 2000). Second, increases in trans-SNARE complex interactions result in an increase in the number of primed, or fusion competent, synaptic vesicles (Lonart and Sudhof, 2000). The implication that SNARE complex assembly primes synaptic vesicles brings up the following quandary: If SNARE complex formation is suYcient to drive membrane fusion, what prevents primed synaptic vesicles from fusing with the plasma membrane instantaneously? One possibility is that complete SNARE complex formation—that is, the complete zippering together of the

90

Weimer and Richmond

four -helical SNARE domains—does not occur until the arrival of a calcium signal. In fact, several experiments indicate that SNARE complexes do not fully assemble before fusion (Chen et al., 1999; Xu et al., 1999). Perhaps, during priming, a semi-zippered structure forms that is unable, or prevented, from zippering together until the calcium signal arrives. Synaptotagmin, the putative calcium sensor previously discussed, is capable of inhibiting, at least in vitro, SNARE-mediated membrane fusion in the absence of calcium (Tucker et al., 2004). In the presence of calcium, however, synaptotagmin dramatically increases SNARE complex formation (Littleton et al., 2001) and the rate of fusion (Tucker et al., 2004). Thus, synaptotagmin may coordinate the arrival of calcium with the complete zippering of SNARE complexes and, consequently, membrane fusion. Alternatively, or in addition, other factors may be required in vivo for membrane fusion. Since neurotransmitter release is tightly regulated, synaptic vesicle priming is also likely regulated. Several studies centered on the t-SNARE syntaxin suggest that priming is regulated via the SNARE proteins themselves. Syntaxin is a multidomain protein with a globular amino terminal domain, a SNARE domain, and a carboxy terminal transmembrane domain (Bennett et al., 1992; Fernandez et al., 1998; Kee et al., 1995). In solution, syntaxin adopts a default-closed configuration in which the amino terminus folds over the SNARE domain within syntaxin (Dulubova et al., 1999). The amino terminus of syntaxin is capable of inhibiting SNARE complex formation in vitro (Calakos et al., 1994). Therefore, the closed state of syntaxin is incompatible with SNARE complex formation and priming. Because closed syntaxin cannot interact with other SNAREs, proteins that regulate the transition between the ‘‘closed’’ and ‘‘open’’ configurations of syntaxin likely regulate priming. UNC-13 proteins are absolutely required for synaptic vesicle priming (Aravamudan et al., 1999; Augustin et al., 1999; Richmond et al., 1999) and bind to the amino terminus of syntaxin (Betz et al., 1997). Expression of a constitutively open form of syntaxin bypasses the requirement for UNC-13 in vesicle priming (Richmond et al., 2001), suggesting that UNC-13 promotes priming by promoting the opening of syntaxin. UNC-13 interacts with the presynaptic protein Rim (Betz et al., 2001), which is also implicated in synaptic vesicle priming (Koushika et al., 2001). Similarly, the requirement for Rim is also bypassed by open syntaxin (Koushika et al., 2001). These data suggest that UNC-13 and Rim function together to promote the opening of syntaxin and thus vesicle priming. Vesicle priming also appears to be regulated even after the conversion of syntaxin from the closed to the open state. Tomosyn is a 130kDa protein that contains a large amino-terminus with WD40 repeats and a carboxyterminal SNARE motif (Fujita et al., 1998; Masuda et al., 1998). The SNARE domain of tomosyn interacts with the SNARE domains of syntaxin and SNAP-25 to form a quasi SNARE complex (Hatsuzawa et al., 2003).


91

In vitro, tomosyn is capable of inhibiting synaptobrevin-SNAP-25-syntaxin SNARE complex formation, and overexpression of tomosyn in vivo inhibits vesicle exocytosis (Fujita et al., 1998; Hatsuzawa et al., 2003; Masuda et al., 1998; Yizhar et al., 2004). Together, these data suggest that vesicle priming is regulated by multiple mechanisms, all of which likely contribute to the specificity of regulated release.

D. Docking Of the exocytic steps outlined, the least understood is synaptic vesicle docking—the process by which synaptic vesicles associate with the plasma membrane. One reason for this lack of understanding is the diYculty in assaying synaptic vesicle docking. Observation of single synaptic vesicles, via FM dye loading, reveals that synaptic vesicles approach the plasma membrane, become associated with the membrane for an interval of time, and then detach and drift away (Murthy and Stevens, 1999). Occasionally, vesicles become ‘‘captured,’’ that is, associated with the membrane and these resident vesicles subsequently undergo fusion (Zenisek et al., 2000). Thus, docking may be an easily reversible step prior to vesicle priming, at which time a stable interaction with the plasma membrane is established. Synaptic vesicle docking can be quantified via electron microscopy. At the ultrastructural level, synaptic vesicles appear to associate with the plasma membrane near putative neurotransmitter release sites (see Fig. 3; reviewed in Zucker, 1996; Robinson and Martin, 1998). Specifically, synaptic vesicles cluster around electron-dense structures, termed ‘‘presynaptic specializations,’’ with a proportion of those vesicles abutting the plasma membrane. One may argue that these plasma membrane-associated vesicles represent docked vesicles; however, ultrastructural analysis, while unmatched in both breadth and resolution, is static, meaning the final fate of these vesicles cannot be traced. While a proportion of the morphologically ‘‘docked’’ vesicles represent truly docked vesicles, by ultrastructural criteria, the docked pool will also contain primed vesicles ready to undergo exocytosis and vesicles which are being endocytosed. The fraction of vesicles morphologically docked at the plasma membrane appears to be in equilibrium with the number of vesicles in the reserve pool, as assayed in mutants that decrease (Hunt et al., 1994; Koenig et al., 1989; Li and Schwarz, 1999; Reist et al., 1998) or increase the vesicle pool (Richmond et al., 1999). An increase in docked vesicles can also be observed in the wild type: larger synapses have more docked vesicles (Harris and Sultan, 1995; Murthy et al., 2001). Thus, there does not appear to be a fixed number of docking sites per active zone but rather there is a low aYnity interaction that ensures that a proportion of the vesicles are associated with the plasma membrane.

92

Weimer and Richmond

Figure 3 Synaptic vesicles dock with the plasma membrane near sites of neurotransmitter release. Individual synaptic vesicles can be resolved in electron micrographs. Shown here is a micrograph of a C. elegans neuromuscular junction in duplicate. At these synapses, synaptic vesicles cluster around presynaptic specializations (asterisk) at active zones. In this micrograph, several vesicles are in close proximity to the plasma membrane (arrows) and would be considered as morphologically docked vesicles. However, it is likely that at least some of the vesicles are not functionally docked vesicles but rather undergoing endocytosis or later steps in exocytosis. For example, the membrane of the synaptic vesicle indicated with the black arrow appears to be in a hemifusion state with the plasma membrane, and thus likely represents a vesicle undergoing fusion rather than docking.

The spatial specificity of docking, that is, the clustering of synaptic vesicles around the presynaptic specialization, implies that a recognition mechanism mediates the attachment of a synaptic vesicle to the target membrane; however, the molecular identity of the docking apparatus remains a mystery. It was originally proposed that interactions between the SNARE proteins provided the spatial specificity of docking (Sollner et al., 1993b). SNARE family members are localized to distinct subcellular compartments within the cell—in essence, representing a unique molecular tag at each compartment. Thus, formation of SNARE complexes between pathway-specific SNARE proteins may act as a mechanism to ensure the correct anchoring of vesicles to target compartments in the cell. However, subsequent biochemical and genetic characterization of SNARE family members has, thus far, failed to support such a role for the SNARE proteins at the synapse. In Drosophila, SNARE family members exhibit promiscuity. For example, the nonsynaptic t-SNARE SNAP-24 can interact with syntaxin and synaptobrevin to form SNARE complexes in vitro (Niemeyer and Schwarz, 2000) and can substitute for SNAP-25 in vivo to promote neurotransmitter release (Vilinsky et al., 2002). Such promiscuity suggests that SNARE interactions alone would be incapable of providing specificity for membrane targeting. Furthermore, perturbation of the SNAREs in squid and Drosophila fails to block the docking of synaptic vesicles at the synapse (Broadie et al., 1995;


93

Hunt et al., 1994), consistent with the idea that SNAREs mediate priming and fusion steps subsequent to vesicle docking, as has been discussed. Recently, members of the Sec-1p/Munc-18 (SM) family of proteins have been implicated in docking (Voets et al., 2001a; Weimer et al., 2003). In the remainder of the chapter, we explore the evidence implicating SM proteins in vesicle docking and describe several SM molecular interactions that may act in concert to promote vesicle docking.

IV. The Sec1p/Munc-18 (SM) Protein Family The Sec1p/Munc-18 (SM) protein family is a group of conserved cytosolic proteins of 60 to 80 kD, which appear to be important for all membrane traYcking pathways in eukaryotic cells (Toonen and Verhage, 2003). The founding member of the SM protein family, UNC-18, was initially identified by Sydney Brenner in a screen for C. elegans mutants exhibiting uncoordinated locomotion (Brenner, 1974). Based on sequence similarity, five additional SM proteins appear to be encoded by the C. elegans genome (T07A9.10, C44C1.4, B0303.9, FA3D9.3, C56C10.1—all have yet to be characterized, thus are referred to by the predicted open reading frame designation), mammals have seven SM proteins (Munc18-1,-2, and -c, sly-1, Vps45, and Vps33a and b), Drosophila have four (ROP, Sly1, Vps45, and Vps33), as do yeast (Sec-1p, Sly-1p, Vps45p, Vps33p) (Toonen and Verhage, 2003). Members of this SM family have been studied extensively in several organisms using both genetic and biochemical approaches, yet a unifying role for SM proteins in membrane traYcking remains elusive.

V. SM Mutant Phenotypes SM mutants have been isolated in yeast, flies, worms, and mice. On an organismal level, the severity of phenotypes associated with the loss of these orthologous SM proteins varies (described in the following text). Yet, these mutants share a common attribute: the disruption of specific membrane traYcking pathways.

A. SM Mutants in Yeast Of the four identified yeast SM proteins, Sec1p is most similar to UNC-18 by sequence homology. Sec1 mutants were originally isolated in a genetic screen for abnormally dense cells (Novick and Schekman, 1979). Sec1 mutants exhibit defects in plasma membrane exocytosis, leading to the accumulation

94

Weimer and Richmond

of secreted proteins—such as, invertase and acid phosphatase—and secretory vesicles, arrested cell growth, and division (Novick and Schekman, 1979). The combination of arrested cell growth and accumulation of secreted proteins ultimately results in an increase in cell density. Consistent with a role in secretion, Sec1p localizes to secretion sites in yeast cells such as the emerging daughter bud, the bud tip, and the mother–daughter junctions during cytokinesis (Carr et al., 1999). Although models of Sec1p function suggest a very late role in vesicle fusion, the precise stage of vesicle traYcking aVected in Sec1 mutants is still undefined. Mutants of the related yeast SM proteins Vps33p and Vps45p accumulate vesicles destined for the vacuole (Peterson and Emr, 2001; Piper et al., 1994). Vps33p is localized to vacuolar membrane (Seals et al., 2000), while Vps45p is associated with golgi membrane (Piper et al., 1994). Both of these SM proteins are known to assemble into large multiprotein complexes implicated in vesicle tethering, a process akin to docking of synaptic vesicles (Cowles et al., 1994; Rieder and Emr, 1997). Mutants of the fourth yeast SM protein, Sly1p, exhibit defects in ER to golgi transport. Specifically, Sly1 mutants accumulate unfused vesicles even though membrane targeting appears normal (Cao et al., 1998). These data suggest that Sly1p acts after vesicle tethering to promote vesicle fusion. Thus, the mutant analysis of SM proteins in yeast presents a somewhat inconsistent role for yeast SM protein function. Vps33p and Vps45p appear to be required for vesicle tethering while Sly1p appears to function after tethering to promote membrane fusion.

B. SM Mutants and the Synapse To date, SM protein function at the synapse has been studied in three species by mutation analysis; those are Drosophila Rop, C. elegans unc-18, and mouse Munc18-1 mutants. 1. Rop In Drosophila, the SM protein required for synaptic vesicle exocytosis Rop (for Ras opposite, based on the position and direction of gene transcription relative to the neighboring Ras gene) is highly expressed in the nervous system as well as other secretory cells. Rop mutants die as embryos and exhibit morphological defects including impaired cuticle secretion and muscle developmental abnormalities (Harrison et al., 1994), suggesting that Rop is required for membrane traYcking pathways in the fly. The severity of the morphological defects in Rop null mutants have precluded detailed analysis of the null mutant phenotype at the synapse.


95

However, a number of Rop missense mutations have been characterized and indicate that Rop regulates synaptic function. In Rop, temperature-sensitive mutants’ retinal synaptic recordings are reversibly disrupted at restrictive temperatures consistent with a neurotransmission defect (Harrison et al., 1994). Neuromuscular synaptic recordings from heterozygous Rop null mutants also show decreased evoked and endogenous synaptic activity, demonstrating a rate-limiting role for Rop in neurotransmitter release (Wu et al., 1998). Two isolated Rop missense mutants (G11 and A19) cause reduced synaptic release, while two others (F3 and G17) result in increased release relative to wild-type levels (Wu et al., 1998). These observations, coupled with the finding that overexpression of Rop partially inhibits release, suggest that Rop may have both permissive and inhibitory roles (Wu et al., 1998). This duality of function appears to be conserved between Drosophila and C. elegans. 2. unc-18 The C. elegans SM protein UNC-18 is orthologous to Rop. UNC-18 is expressed in all neurons. unc-18 mutants exhibit severe defects in locomotion, accumulate acetylcholine, and exhibit resistance to aldicarb, an inhibitor of acetylcholinesterase (Hosono and Kamiya, 1991; Hosono et al., 1987). These characteristics are all consistent with a defect in neurotransmitter release in unc-18 mutants. Unlike Drosophila Rop null mutants, unc-18 null mutants are viable as adults, as are many exocytosis-defective mutants in C. elegans. unc-18 mutants are, therefore, accessible to detailed electrophysiological analysis (Weimer et al., 2003). Evoked synaptic responses at the neuromuscular junction of unc-18 mutant worms are greatly reduced, although not completely abolished (20% of wild-type amplitudes). Minis are still evident in C. elegans unc-18 mutants, although at reduced frequencies. The postsynaptic receptors are properly clustered, respond normally to exogenous neurotransmitter, and mini amplitudes are not reduced, indicating that the synaptic defect is presynaptic. The unc-18 mutant defect does not appear to be in calcium-sensing as the remaining evoked release has normal Ca2+sensitivity. However, the primed vesicle pool is reduced in unc-18 mutants, suggesting that either vesicle priming and/or upstream stages of the vesicle cycle are defective. The latter of these alternatives has emerged as the leading contender for the role of UNC-18, based on the following evidence. At the ultrastructural level, neuromuscular junctions in C. elegans unc-18 mutants contain more synaptic vesicles than the wild type. However, the proportion of vesicles docked near release sites is reduced in unc-18 mutants. These data suggest that in C. elegans, UNC-18 likely promotes neurotransmitter release by supporting vesicle docking.

96

Weimer and Richmond

Our unpublished observations also suggest that UNC-18 may have an additional negative regulatory role in synaptic transmission. First, a missense mutation at position R39C equivalent to the Rop mutation (F3 at position R50C) causes enhanced release compared to wild-type C. elegans UNC-18. This result is similar to the augmented release observed at Rop(F3) synapses (Wu et al., 1998). Second, overexpression of wild-type UNC-18 causes a reduction in evoked release at C. elegans neuromuscular junctions similar to the eVects of overexpressing Rop at Drosophila synapses (Wu et al., 1998). These data suggest that both Rop and UNC-18 exhibit dual functions in the regulation of neurotransmitter release: probably facilitating release by promoting docking and inhibiting priming by binding to closed syntaxin. 3. Munc18-1( / ) Unlike C. elegans unc-18 mutants, in mice, the elimination of the ortholog Munc18-1 completely abolishes neurotransmission (Verhage et al., 2000). These mutant mice die at birth; therefore, the analysis of synaptic vesicle exocytosis has been restricted to embryonically derived tissues. Neocortical neurons in slice show a complete absence of miniature synaptic events (Verhage et al., 2000). A similar result was observed at Munc18-1 mutant neuromuscular synapses; in this case, both endogenous minis and -latrotoxin-induced vesicle fusion were eliminated (Verhage et al., 2000). Furthermore, Munc18-1 null synapses do not appear to have a reduction in morphologically docked vesicles based on ultrastructural analysis of the neocortical synapses (Verhage et al., 2000). A somewhat diVerent picture emerges from analysis of chromaYn cells derived from the same Munc18-1 knockout mice. These cells show a 10-fold reduction in both depolarization-induced and Ca2+-induced dense-core vesicle fusion (Voets et al., 2001): A deficit that could not be overcome by increasing Ca2+-levels, indicating that Ca2+-sensitivity is unaVected. Endogenous fusion events are also still detectable by amperometry and have normal spike kinetics and amplitude. To test whether the reduction in secretion in Munc18-1 knockout chromaYn cells is due to a change in the number or distribution of vesicles, an ultrastructural analysis was performed. In wildtype chromaYn cells at this embryonic stage, the majority of dense core vesicles are aligned with the plasma membrane within a distance of 25nm. In Munc18-1 nulls, a striking 10-fold reduction in the number of morphologically docked granules is observed. The paucity of docked vesicles is not due to changes in the overall number of dense core vesicles, ruling out a role for Munc18-1 in vesicle biogenesis or translocation from the golgi. A similar docking defect is observed in pituitary somatotrophs derived from Munc18-1 knockout mice (Verhage and Korteweg, personal communication). In most


97

respects, the phenotype observed in these peptidergic cells resembles those observed at C. elegans unc-18 mutant neuromuscular junctions. The vesicle docking defect observed at C. elegans unc-18 neuromuscular junctions and mouse Munc18-1( / ) peptidergic cells is hard to reconcile with the lack of a docking defect at mouse Munc18-1( / ) central synapses. The inevitable degeneration of central neurons in Munc18-1 knockout mice raises a potential concern about the health of these neurons at the time of analysis, although care was taken to examine the neurons before the onset of apoptosis. Why this neuronal degeneration occurs in mouse but not C. elegans neurons is currently unknown. Possibly, there are redundant SM proteins in C. elegans (e.g., T07A9.10) that provide residual SM protein function. The variability in docking phenotypes is even more perplexing. If docking is the central role performed by SM proteins, one might expect to see a correlation between the severity of the secretory and docking defects. In the case of Munc18-1 null synapses, this correlation clearly does not exist. In general, synapses of even severe fusion defective mutants in mice do not accumulate synaptic vesicles, which might be expected if vesicles were continually made but not exocytosed. Homologous C. elegans mutants, however, tend to accumulate vesicles if exocytosis is impaired (Richmond et al., 1999; Weimer et al., 2003). Perhaps diVerences in the turnover of vesicles obscure defects in the potentially very transient docking step. The strongest support for a docking role for SM proteins is that observed in Munc18-1 null chromaYn cells, in which granule turnover is likely to be very diVerent from that at synapses. In summary, SM mutant phenotypes from a variety of organisms suggest that all membrane traYcking in eukaryotes requires SM proteins. For many SM mutants, the precise stage of the membrane traYcking event that is aVected has yet to be defined. However, during regulated release, such as neurotransmitter or neuroendocrine release, there is increasing evidence that SM proteins participate in vesicle docking. How might the SM proteins promote synaptic vesicle docking?

VI. SM Protein Interactions: Putative Links That Anchor Synaptic Vesicles In the simplest of models, SM proteins would promote synaptic vesicle docking directly by linking the vesicle to the plasma membrane. Since no lipid binding activities have been described for the SM proteins, such a link would likely require interactions with proteins associated with both the synaptic vesicle and plasma membrane. Several possible interactions have been identified on the basis of biochemical or genetic evidence (Fig. 4).

98

Weimer and Richmond

Figure 4 SM protein interactions which may promote synaptic vesicle docking. Several interactions between SM proteins and synaptic proteins have been described, black arrows indicate physical interactions, the dashed line indicates an implied genetic interaction, and the blue indicates an indirect interaction. SM proteins may promote synaptic vesicle docking, either directly or indirectly, through these interactions.

A. Plasma Membrane-Associated Proteins That Interact with SM Proteins 1. Syntaxin All SM proteins bind with high aYnity to members of the syntaxin protein family. However, the mode of interaction between SM proteins and syntaxin is surprisingly variable. To date, there are at least four binding modes described: (1) SM binding to a closed conformation of syntaxin, (2) SM binding to the N-terminal of open syntaxin, (3) SM binding to syntaxin assembled in a SNARE complex, and (4) SM in a preassembled complex binding indirectly to syntaxin (reviewed in Toonen and Verhage, 2003). In addition to these diVerent modes of SM–syntaxin interaction, the syntaxin binding interface of SM proteins is also variable, despite the fact that, overall, the structure of the SM proteins and syntaxin homologs appears to be highly conserved (Dulubova et al., 2003). For example, while syntaxin1 in its closed configuration binds to the central cavity of the Munc18-1 horseshoe (Misura et al., 2000), the N-terminus of the yeast syntaxin sed5p binds to the outer surface of the horseshoe formed by the SM protein Sly1p (Bracher and Weissenhorn, 2002). This may explain the apparent diVerences in ascribed SM protein function observed at diVerent vesicle traYcking steps. For example, there appears to be a distinct and conserved binding mode for SM proteins involved in synaptic release that


99

diVers from SM proteins that function in constitutive release. Specifically, all evidence suggests that UNC-18 and its homologs involved in regulated secretion only interact with syntaxin in its default closed state, whereas other SMs interact with either open syntaxin or syntaxin in SNARE complexes. These diVerences suggest that binding of SM to closed syntaxin is an evolutionary specialization of SM function for fast, regulated fusion. Is the SM-syntaxin interaction required for docking? If the SM-syntaxin interaction were involved in the proposed UNC-18 docking role, one would predict that mutants that eliminate binding between these two proteins would specifically disrupt docking. One approach to assess the relevance of the SM-syntaxin interaction in vesicle docking is to examine docking in syntaxin null mutants. To date, this has only been examined in Drosophila syntaxin null mutants, which die as embryos (Broadie et al., 1995). Synapses in syntaxin mutant embryos have no synaptic transmission, yet actually have a significant increase in the number of morphologically docked vesicles (Broadie et al., 1995). These data suggest that SNARE proteins are not required for vesicle docking. EVorts to disrupt syntaxin function have also been performed at squid synapses using botulinum C (Marsal et al., 1997; O’Connor et al., 1997). In two independent studies, synaptic release was markedly inhibited by toxin cleavage of syntaxin; however, no docking defect was observed. In a diVerent experiment, introduction of the H3 domain of syntaxin was shown to inhibit release without aVecting docking, acting instead to perturb SNARE complex formation (O’Connor et al., 1997). Intriguingly, and in contrast to these studies, introduction of the syntaxin 1 H3 domain into pancreatic beta cells specifically inhibits docking of insulin containing granules (Torii et al., 2004). Another approach to study the eVects of the SM-syntaxin interaction on docking is to study the consequences of overexpressing these proteins. Rop and UNC-18 overexpression leads to partial inhibition of synaptic vesicle release in Drosophila and C. elegans, respectively. Since Munc 18-1 is known to bind and stabilize the closed conformation of syntaxin 1a (Dulubova et al., 1999), the inhibition of release at these SM overexpressing synapses is likely to be due to the inability of the occluded syntaxin SNARE motif to participate in SNARE complex formation. If the interaction between SM proteins and membrane-bound closed syntaxin is required as a platform for vesicle docking, it is entirely possible that the number of docked vesicles may actually be enhanced when Rop or UNC-18 are overexpressed, at the same time that vesicle fusion is inhibited. The ultrastructural examination of docked vesicles at synapses overexpressing Rop or UNC-18 proteins has yet to be examined. In summary, the current evidence addressing the relevance of the SMsyntaxin interaction for vesicle docking suggests that syntaxin is not required

100

Weimer and Richmond

for docking. Future studies will likely shed light on the relevance of this interaction.

2. MINTS Munc18-1 also interacts with a protein family, termed MINTs, for Munc18interacting proteins, which are localized to the plasma membrane. MINTS are multidomain proteins that have divergent N-terminals and a common C-terminal containing PTB and PDZ domains. The N-terminal of mammalian MINT-1 binds to Munc18-1 (Okamoto and Sudhof, 1997). MINTs have also been shown to form heteromeric complexes with CASK, a membrane-associated guanylate kinase and Veli(mLin-7/MALS) (Borg et al., 1998; Butz et al., 1998). This heteromeric complex associates with Munc-18/syntaxin dimers and also interacts with the cytoplasmic PDZ domain of membrane-bound neurexin (Biederer and Sudhof, 2000; Hata et al., 1996). Neurexin has been shown to be localized to growth cones and presynaptic terminals, and can induce vesicle clustering, an action that is dependent on the presence of the neurexin PDZ domain (Dean et al., 2003). Thus, the interaction with neurexin could be involved in specifically targeting SM proteins in association with the MINT/CASK/Veli complex to the plasma membrane at sites of exocytosis (Biederer and Sudhof, 2000; Dean et al., 2003). Neurexin triple knockouts exhibit severe exocytosis defects that have been attributed to disorganization of release machinery, specifically, presynaptic calcium channels. The ultrastructure of the synapses appeared normal, although the number of morphologically docked vesicles was not reported (Missler et al., 2003). The role of MINTs has been explored in C. elegans and mouse mutants. The first member of the MINT family, LIN-10, was identified as a C. elegans neuronal protein (Whitfield et al., 1999). In the case of the C. elegans MINT mutant lin-10, the phenotypes thus far elucidated include vulval developmental defects due to mistargeting of EGF receptors (Kaech et al., 1998; Simske et al., 1996), and AMPA receptor mislocalization (Rongo et al., 1998). C. elegans lin-10 mutants show no obvious locomotion phenotypes that would indicate a major role for lin-10 in exocytosis. Similarly, a recent mouse MINT-1 knockout exhibits only a very mild synaptic phenotype. Specifically, the MINT-1 null mutant has essentially normal release with the exception of enhanced paired pulse depression in GABAergic neurons, where MINT-1 is preferentially expressed (Ho et al., 2003). This result was interpreted to mean that loss of MINT-1 increases release probability, a result that would argue against a requirement for MINT-1 in vesicle docking.


101

B. Vesicle-Associated Proteins That Interact with SM Proteins 1. Doc2 Munc-18 also physically interacts with the vesicle-associated protein Doc2 (Verhage et al., 1997). Doc2 resides on synaptic vesicles and associates with both Munc18 and Munc13, rendering Doc2 an attractive candidate for a vesicle docking and/or priming component. Doc2 proteins, named for their double C2 domains, translocate to the plasma membrane (Duncan et al., 1999) when cells are exposed to phorbol ester, a synthetic diacyl glycerol that binds to Munc13. This translocation depends on an interaction between the amino-terminus of Doc2 and the active zone protein Munc13. Interference with this interaction dramatically reduces evoked release, possibly by disrupting the priming function of Munc13 (Mochida et al., 1998). The relevance of the Munc-18/Doc2 interaction, however, has yet to be elucidated. The first C2 domain of Doc2 is known to interact with Munc18. Furthermore, Doc2 competes with syntaxin for Munc18 binding, and this competition is enhanced by the presence of SNAP-25 (Verhage et al., 1997). One possible interpretation of these data is that Doc-2 may remove Munc-18 at sites of release, allowing syntaxin to open and participate in SNARE complexes. Whether the interaction between Doc2 and Munc-18 also acts in a docking capacity is unknown. In vivo experiments to test the importance of the Doc2/Munc18 interaction have yet to be reported. If Doc2 is involved in vesicle docking via an interaction with Munc-18, one might expect to see enhanced docking when Doc2 is overexpressed. Consistent with this notion, Doc2A overexpression in PC12 cells leads to enhanced peptide release (Orita et al., 1996). Unfortunately, ultrastructural analyses of these cells have not been performed; therefore, it is unclear whether the increased release reflects increased granule docking via a Munc-18-dependent interaction. Alternatively, Doc2 overexpression may increase priming either by displacing Munc18 from syntaxin or via a Munc-13 interaction. In mice, Doc2A and Doc2B are both expressed in the nervous system (Korteweg et al., 2000). To date, only the Doc2A isoform has been knocked out. The Doc2A mutant mice exhibit a very mild synaptic phenotype, showing a slight increase in fatigue rates during 5 Hz stimulations and slightly impaired LTP (Sakaguchi et al., 1999). While this result would be consistent with a docking defect, ultrastuctural data to support such a role have yet to be reported. Furthermore, compensatory mechanisms due to upregulation of the additional Doc2 genes make the interpretation of this phenotype diYcult. Future double and triple knockouts will undoubtedly provide a more definitive answer as to whether Doc2 has a role in docking.

102

Weimer and Richmond

Doc2 homologs have not been identified in either the C. elegans or Drosophila genomes, suggesting that Doc2 is either a specialized vertebrate protein or that related C2 domain proteins, such as members of the rabphilin or synaptotagmin families, are Doc2 orthologs in invertebrates. In summary, Doc-2 remains an attractive candidate for a docking function; however, there is as yet no ultrastructural evidence to support such a role. Furthermore, the significance of the Doc2–Munc18-l interaction in any putative docking capacity has not been explored. 2. Rabs Several lines of evidence implicate members of the Rab protein family of small protein GTPases as mediators of vesicle docking. First, both Rab3 and possibly Rab27 are synaptic vesicle-associated (Fischer von Mollard et al., 1990; Nonet et al., 1997), placing Rabs in a location conducive to function in vesicle docking. Second, eliminating Rab3 function alters synaptic activity (Geppert et al., 1994a; Nonet et al., 1997), implicating Rab3 in synaptic transmission. Third, overexpression of Rab3 in PC12 cells or Rab27 in pancreatic beta cells increases the number of dense-core vesicles morphologically docked at the plasma membrane (Martelli et al., 2000). Since densecore vesicle traYcking uses many of the same proteins as synaptic vesicle cycling (Zucker, 1996), Rabs may function to promote synaptic vesicle docking as well. Fourth, stimulation-dependent recruitment of docked vesicles to active zones in brain-derived synaptosomes is impaired in rab3a knockout mice (Leenders et al., 2001). Finally, in C. elegans rab-3 mutants, synaptic vesicles are diVusely distributed at presynaptic nerve terminals with fewer vesicles clustered near active zones (Nonet et al., 1997), a phenotype consistent with a docking defect. Who are the molecular binding partners of Rab? When bound to GTP, Rab3 directly associates with the presynaptic proteins Rabphilin (Stahl et al., 1996) and Rim (Wang et al., 1997) and only indirectly with SM proteins. Rabphilin is a cytoplasmic protein that associates with synaptic vesicles in a Rab3dependent manner (Stahl et al., 1996), whereas Rim is an active zone protein (Wang et al., 1997). Thus, one could imagine that the interaction between Rab3 and Rim anchors vesicles at active zones, while Rabphilin modulates the Rab3Rim interaction (Sudhof, 1995; Sudhof and Scheller, 2001). However, recent data suggest that this model is inaccurate. First, synaptic vesicles dock with the plasma membrane at wild-type levels in C. elegans Rim-null mutants, suggesting that Rim is not required for vesicle docking. Second, detailed characterization of Rabphilin null mutants suggests that Rabphilin is likely required after docking to facilitate synaptic transmission by a Rab3-independent mechanism (Schluter et al., 1999; Staunton et al., 2001). Thus, synaptic vesicle docking appears to be a Rim- and Rabphilin-independent phenomenon.


103

Whether SM proteins represent the Rab eVectors required for vesicle docking is unclear. There are several reports of genetic interactions between Rabs and SM proteins from yeast. First, a specific mutation in the yeast SM protein Sly1p (Sly1-20) partially suppresses a null mutant in the Rab protein Ypt1p (Dascher et al., 1991). Interestingly, the site of the mutation in Sly1-20 is on the outer face of Sly1p on the opposite arm of the sly1p horseshoe structure which binds the yeast syntaxin Sed5p (Bracher and Weissenhorn, 2002; Dascher et al., 1991), possibly reflecting the ability of SM proteins to simultaneously interact with vesicle-linked and plasma membrane-associated proteins. Second, a mutation in the yeast SM Sec1, when combined with a mutant of the Rab Sec4, results in synthetic lethality (Finger and Novick, 2000) and overexpression of Sec4p can partially rescue the Sec1 mutant phenotype (Salminen and Novick, 1987). Third, a genetic interaction has been observed between the SM mutant Vps45 and the Rab protein Vps21, which both function in golgi to endosome traYcking (Tall et al., 1999). Together, these data suggest that Rab and SM proteins may function in the same pathway. While there is no evidence of a direct physical interaction between SM and Rab proteins, multimeric protein complexes containing SM proteins interact biochemically with Rab proteins. For example, the yeast SM protein Vps33p is a component of the HOPs complex (for homotypic fusion and vacuole protein sorting) that interacts with the Rab protein Ypt7p in pull-down experiments (Seals et al., 2000). The HOPS complex is actually composed of two protein complexes: (1) the Vam2/Vam6 protein complex; Vam6 is a Ypt7-GEF (Wurmser et al., 2000), and (2) the VpsC complex, which consists of four proteins Vps11, Vps16, Vps18, and Vps33 (Rieder and Emr, 1997). During vacuolar tethering, which is analogous to synaptic vesicle docking (discussed further in Robinson and Martin, 1998), the HOPS complex associates with Ypt7 (Price et al., 2000), facilitates the exchange of GDP bound to Ypt7 for GTP (Eitzen et al., 2000; Wurmser et al., 2000), and forms a stable interaction with the GTP-bound form of Ypt7 (Eitzen et al., 2000; Haas et al., 1995; Price et al., 2000; Seals et al., 2000; Ungermann et al., 1998). Since, the HOPS complex is localized to the vacuolar membrane (Rieder and Emr, 1997), and Ypt7 is found on the membrane of the incoming vesicle (Haas et al., 1995), the HOPS/Ypt7 interaction is thought to tether the vesicle to the vacuole. Consistent with this hypothesis, perturbation of Ypt7 or the HOPS complex blocks vacuolar traYcking at the tethering step (Sato et al., 2000; Seals et al., 2000; Ungermann et al., 1998). Therefore, a homolog of UNC-18, Vps33, promotes vacuolar tethering through an association with Ypt7, a Rab3 homolog. Similarly, human SM protein Vps45 interacts with endosomal Rab5 via the Rab5 eVector rabenosyn (Nielsen et al., 2000). These complexes may act as molecular bridges linking vesicles to target membranes.

104

Weimer and Richmond

VII. SM Proteins Promote Vesicle Docking Indirectly: An Alternative Hypothesis While several SM-interacting proteins have been implicated in docking, it is entirely possible that the observed SM-docking defects reflect an indirect role of SM proteins in vesicle docking. Several models have been proposed in which SM proteins promote synaptic vesicle docking in an indirect manner.

A. SNARE Sorting at the Synapse SM protein function may be required at the synapse for the proper sorting of SNARE proteins after synaptic vesicle exocytosis. Upon vesicle fusion, trans-SNARE complexes (SNARE complexes that link vesicles with the plasma membrane) become cis-SNARE complexes (all three SNAREs are in the same membrane after the vesicle merges with the plasma membrane). These cis-SNARE complexes are then disassembled by the ATPase NSF (N-ethylmaleimide-sensitive f actor) (Littleton et al., 1998; Sollner et al., 1993a; Wilson et al., 1992). After disassembly, SM protein binding to the closed state of syntaxin may be important to prevent inappropriate syntaxin interactions with proteins destined for recycling; synaptic vesicle-specific proteins are then selectively removed from the plasma membrane by endocytosis. According to this model, inappropriate syntaxin interactions could titrate out a factor required for docking. Current data fail to support a role for the vesicle SNARE, synaptobrevin in vesicle docking (Broadie et al., 1995) but other vesicle-associated proteins may be impacted. Alternatively, plasma membrane-associated proteins may be inappropriately incorporated onto recycled vesicles in the absence of UNC-18, and interfere with subsequent vesicle docking.

B. SNARE Trafficking to the Synapse In another indirect model, SM proteins are not required at the synapse per se, but are required to traYc a factor required for vesicle docking from the cell body to the synapse. Again, the SNARE protein syntaxin may constitute this factor. Overexpression experiments in heterologous systems suggest that SM proteins are required for the transport of syntaxin to the plasma membrane. The mislocalization of syntaxin, or an as yet unidentified factor, could result in the apparent docking defect observed at sites of regulated release in SM mutants.


105

VIII. Summary and Future Directions Regulated release, like general membrane traYcking, requires that membranes destined to fuse associate with one another. With respect to neuroendocrine and synaptic release, this process is referred to as docking. While it appears that the core exocytic machinery, i.e., the SNAREs, function in a similar capacity in all membrane-traYcking pathways, it remains unclear if other traYcking proteins, such as the SM proteins, have a conserved function. Current data from the study of constitutive membrane-traYcking pathways suggest that SM proteins are required for a post docking step in membrane traYcking, perhaps at the membrane fusion step. However, a divergent role for the SM proteins has emerged from the study of regulated release: SM proteins in some of these pathways appear to promote the docking of vesicles with the plasma membrane through an unknown molecular mechanism. Delineating this mechanism is of utmost importance for understanding both SM protein function and synaptic vesicle exocytosis, and will require a concerted eVort from multiple fields. First and foremost, a better understanding of synaptic vesicle docking is required, which begins by better defining vesicle docking. Currently, we define vesicle docking using morphological criteria: a vesicle is docked if it is associated with the plasma membrane near sites of neurotransmitter release. While true in principle—by necessity, a vesicle destined to undergo exocytosis must associate, or dock, with the plasma membrane—this definition is crippled by its broadness. The pool of synaptic vesicles associated with the presynaptic plasma membrane will inevitably consist of vesicles undergoing exocytosis, both docked and primed, as well as vesicles being retrieved from the membrane by endocytosis. Distinguishing between these populations will aid in the identification of the relevant pool of vesicles which are truly docked at the plasma membrane. A full understanding of vesicle docking will also require further characterization of SM proteins as well as other synaptic proteins. This entails the cataloging of all proteins and protein interactions present at the synapse, detailed morphometric analysis of synaptic mutants, and scrutiny of each protein interaction with respect to protein localization and vesicle docking. While daunting, such approaches will unscramble the riddle of synaptic vesicle docking and will shed light upon the molecular mechanisms driving synaptic vesicle exocytosis.

References Anderson, D. C., King, S. C., and Parsons, S. M. (1982). Proton gradient linkage to active uptake of [3H]acetylcholine by Torpedo electric organ synaptic vesicles. Biochemistry 21, 3037–3043.

106

Weimer and Richmond

Aravamudan, B., Fergestad, T., Davis, W. S., Rodesch, C. K., and Broadie, K. (1999). Drosophila UNC-13 is essential for synaptic transmission. Nat. Neurosci. 2, 965–971. Augustin, I., Rosenmund, C., Sudhof, T. C., and Brose, N. (1999). Munc13-1 is essential for fusion competence of glutamatergic synaptic vesicles. Nature 400, 457–461. Bai, J., Wang, C. T., Richards, D. A., Jackson, M. B., and Chapman, E. R. (2004). Fusion pore dynamics are regulated by synaptotagmin*t-SNARE interactions. Neuron 41, 929–942. Bennett, M. K., Calakos, N., and Scheller, R. H. (1992). Syntaxin: A synaptic protein implicated in docking of synaptic vesicles at presynaptic active zones. Science 257, 255–259. Bennett, M. K., and Scheller, R. H. (1993). The molecular machinery for secretion is conserved from yeast to neurons. Proc. Natl. Acad. Sci. USA 90, 2559–2563. Betz, A., Okamoto, M., Benseler, F., and Brose, N. (1997). Direct interaction of the rat unc-13 homologue Munc13-1 with the N terminus of syntaxin. J. Biol. Chem. 272, 2520–2526. Betz, A., Thakur, P., Junge, H. J., Ashery, U., Rhee, J. S., Scheuss, V., Rosenmund, C., Rettig, J., and Brose, N. (2001). Functional interaction of the active zone proteins Munc13-1 and RIM1 in synaptic vesicle priming. Neuron 30, 183–196. Biederer, T., and Sudhof, T. C. (2000). Mints as adaptors. Direct binding to neurexins and recruitment of munc18. J. Biol. Chem. 275, 39803–39806. Borg, J. P., Straight, S. W., Kaech, S. M., de Taddeo-Borg, M., Kroon, D. E., Karnak, D., Turner, R. S., Kim, S. K., and Margolis, B. (1998). Identification of an evolutionarily conserved heterotrimeric protein complex involved in protein targeting. J. Biol. Chem. 273, 31633–31636. Bracher, A., and Weissenhorn, W. (2002). Structural basis for the Golgi membrane recruitment of Sly1p by Sed5p. EMBO J. 21, 6114–6124. Brenner, S. (1974). The genetics of Caenorhabditis elegans. Genetics 77, 71–94. Broadie, K., Prokop, A., Bellen, H. J., O’Kane, C. J., Schulze, K. L., and Sweeney, S. T. (1995). Syntaxin and synaptobrevin function downstream of vesicle docking in Drosophila. Neuron 15, 663–673. Brose, N., Petrenko, A. G., Sudhof, T. C., and Jahn, R. (1992). Synaptotagmin: A calcium sensor on the synaptic vesicle surface. Science 256, 1021–1025. Butz, S., Okamoto, M., and Sudhof, T. C. (1998). A tripartite protein complex with the potential to couple synaptic vesicle exocytosis to cell adhesion in brain. Cell 94, 773–782. Calakos, N., Bennett, M. K., Peterson, K. E., and Scheller, R. H. (1994). Protein–protein interactions contributing to the specificity of intracellular vesicular traYcking. Science 263, 1146–1149. Cao, X., Ballew, N., and Barlowe, C. (1998). Initial docking of ER-derived vesicles requires Uso1p and Ypt1p but is independent of SNARE proteins. EMBO J. 17, 2156–2165. Carr, C. M., Grote, E., Munson, M., Hughson, F. M., and Novick, P. J. (1999). Sec1p binds to SNARE complexes and concentrates at sites of secretion. J. Cell Biol. 146, 333–344. Chapman, E. R., Hanson, P. I., An, S., and Jahn, R. (1995). Ca2+ regulates the interaction between synaptotagmin and syntaxin 1. J. Biol. Chem. 270, 23667–23671. Chen, Y. A., Scales, S. J., Patel, S. M., Doung, Y. C., and Scheller, R. H. (1999). SNARE complex formation is triggered by Ca2+ and drives membrane fusion. Cell 97, 165–174. Chen, Y. A., Scales, S. J., and Scheller, R. H. (2001). Sequential SNARE assembly underlies priming and triggering of exocytosis. Neuron 30, 161–170. Chen, Y. A., and Scheller, R. H. (2001). SNARE-mediated membrane fusion. Nat. Rev. Mol. Cell Biol. 2, 98–106. Chow, R. H., von Ruden, L., and Neher, E. (1992). Delay in vesicle fusion revealed by electrochemical monitoring of single secretory events in adrenal chromaYn cells. Nature 356, 60–63.


107

Cowles, C. R., Emr, S. D., and Horazdovsky, B. F. (1994). Mutations in the VPS45 gene, a SEC1 homologue, result in vacuolar protein sorting defects and accumulation of membrane vesicles. J. Cell Sci. 107 (Pt 12), 3449–3459. Dascher, C., Ossig, R., Gallwitz, D., and Schmitt, H. D. (1991). Identification and structure of four yeast genes (SLY) that are able to suppress the functional loss of YPT1, a member of the RAS superfamily. Mol. Cell Biol. 11, 872–885. Davletov, B. A., and Sudhof, T. C. (1993). A single C2 domain from synaptotagmin I is suYcient for high aYnity Ca2+/phospholipid binding. J. Biol. Chem. 268, 26386–26390. Dean, C., Scholl, F. G., Choih, J., DeMaria, S., Berger, J., IsacoV, E., and ScheiVele, P. (2003). Neurexin mediates the assembly of presynaptic terminals. Nat. Neurosci. 6, 708–716. Dulubova, I., Sugita, S., Hill, S., Hosaka, M., Fernandez, I., Sudhof, T. C., and Rizo, J. (1999). A conformational switch in syntaxin during exocytosis: Role of munc18. EMBO J. 18, 4372–4382. Dulubova, I., Yamaguchi, T., Arac, D., Li, H., Huryeva, I., Min, S. W., Rizo, J., and Sudhof, T. C. (2003). Convergence and divergence in the mechanism of SNARE binding by Sec1/Munc18-like proteins. Proc. Natl. Acad. Sci. USA 100, 32–37. Duncan, R. R., Betz, A., Shipston, M. J., Brose, N., and Chow, R. H. (1999). Transient, phorbol ester-induced DOC2-Munc13 interactions in vivo. J. Biol. Chem. 274, 27347–27350. Eitzen, G., Will, E., Gallwitz, D., Haas, A., and Wickner, W. (2000). Sequential action of two GTPases to promote vacuole docking and fusion. EMBO J. 19, 6713–6720. Ernst, J. A., and Brunger, A. T. (2003). High resolution structure, stability, and synaptotagmin binding of a truncated neuronal SNARE complex. J. Biol. Chem. 278, 8630–8636. Fernandez, I., Ubach, J., Dulubova, I., Zhang, X., Sudhof, T. C., and Rizo, J. (1998). Threedimensional structure of an evolutionarily conserved N-terminal domain of syntaxin 1A. Cell 94, 841–849. Fernandez-Chacon, R., Konigstorfer, A., Gerber, S. H., Garcia, J., Matos, M. F., Stevens, C. F., Brose, N., Rizo, J., Rosenmund, C., and Sudhof, T. C. (2001). Synaptotagmin I functions as a calcium regulator of release probability. Nature 410, 41–49. Finger, F. P., and Novick, P. (2000). Synthetic interactions of the post-Golgi sec mutations of Saccharomyces cerevisiae. Genetics 156, 943–951. Fischer von Mollard, G., Mignery, G. A., Baumert, M., Perin, M. S., Hanson, T. J., Burger, P. M., Jahn, R., and Sudhof, T. C. (1990). rab3 is a small GTP-binding protein exclusively localized to synaptic vesicles. Proc. Natl. Acad. Sci. USA 87, 1988–1992. Fujita, Y., Shirataki, H., Sakisaka, T., Asakura, T., Ohya, T., Kotani, H., Yokoyama, S., Nishioka, H., Matsuura, Y., Mizoguchi, A., et al. (1998). Tomosyn: A syntaxin-1-binding protein that forms a novel complex in the neurotransmitter release process. Neuron 20, 905–915. Geppert, M., Bolshakov, V. Y., Siegelbaum, S. A., Takei, K., De Camilli, P., Hammer, R. E., and Sudhof, T. C. (1994a). The role of Rab3A in neurotransmitter release. Nature 369, 493–497. Geppert, M., Goda, Y., Hammer, R. E., Li, C., Rosahl, T. W., Stevens, C. F., and Sudhof, T. C. (1994b). Synaptotagmin I: A major Ca2+ sensor for transmitter release at a central synapse. Cell 79, 717–727. Goda, Y., and Stevens, C. F. (1994). Two components of transmitter release at a central synapse. Proc. Natl. Acad. Sci. USA 91, 12942–12946. Haas, A., Scheglmann, D., Lazar, T., Gallwitz, D., and Wickner, W. (1995). The GTPase Ypt7p of Saccharomyces cerevisiae is required on both partner vacuoles for the homotypic fusion step of vacuole inheritance. EMBO J. 14, 5258–5270. Han, X., Wang, C. T., Bai, J., Chapman, E. R., and Jackson, M. B. (2004). Transmembrane segments of syntaxin line the fusion pore of Ca2+-triggered exocytosis. Science 304, 289–292.

108

Weimer and Richmond

Hanson, P. I., Roth, R., Morisaki, H., Jahn, R., and Heuser, J. E. (1997). Structure and conformational changes in NSF and its membrane receptor complexes visualized by quick-freeze/deep-etch electron microscopy. Cell 90, 523–535. Harris, K. M., and Sultan, P. (1995). Variation in the number, location, and size of synaptic vesicles provides an anatomical basis for the nonuniform probability of release at hippocampal CA1 synapses. Neuropharmacology 34, 1387–1395. Harrison, S. D., Broadie, K., van de Goor, J., and Rubin, G. M. (1994). Mutations in the Drosophila Rop gene suggest a function in general secretion and synaptic transmission. Neuron 13, 555–566. Hata, Y., Butz, S., and Sudhof, T. C. (1996). CASK: A novel dlg/PSD95 homolog with an N-terminal calmodulin-dependent protein kinase domain identified by interaction with neurexins. J. Neurosci. 16, 2488–2494. Hatsuzawa, K., Lang, T., Fasshauer, D., Bruns, D., and Jahn, R. (2003). The R-SNARE motif of tomosyn forms SNARE core complexes with syntaxin 1 and SNAP-25 and downregulates exocytosis. J. Biol. Chem. 278, 31159–31166. Hell, J. W., Maycox, P. R., and Jahn, R. (1990). Energy dependence and functional reconstitution of the gamma-aminobutyric acid carrier from synaptic vesicles. J. Biol. Chem. 265, 2111–2117. Ho, A., Morishita, W., Hammer, R. E., Malenka, R. C., and Sudhof, T. C. (2003). A role for Mints in transmitter release: Mint 1 knockout mice exhibit impaired GABAergic synaptic transmission. Proc. Natl. Acad. Sci. USA 100, 1409–1414. Hosono, R., and Kamiya, Y. (1991). Additional genes which result in an elevation of acetylcholine levels by mutations in Caenorhabditis elegans. Neurosci. Lett. 128, 243–244. Hosono, R., Sassa, T., and Kuno, S. (1987). Mutations aVecting acetylcholine levels in the nematode Caenorhabditis elegans. J. Neurochem. 49, 1820–1823. Hu, K., Carroll, J., Rickman, C., and Davletov, B. (2002). Action of complexin on SNARE complex. J. Biol. Chem. 277, 41652–41656. Hua, S. Y., and Charlton, M. P. (1999). Activity-dependent changes in partial VAMP complexes during neurotransmitter release. Nat. Neurosci. 2, 1078–1083. Hunt, J. M., Bommert, K., Charlton, M. P., Kistner, A., Habermann, E., Augustine, G. J., and Betz, H. (1994). A post-docking role for synaptobrevin in synaptic vesicle fusion. Neuron 12, 1269–1279. Jahn, R., Lang, T., and Sudhof, T. C. (2003). Membrane fusion. Cell 112, 519–533. Kaech, S. M., Whitfield, C. W., and Kim, S. K. (1998). The LIN-2/LIN-7/LIN-10 complex mediates basolateral membrane localization of the C. elegans EGF receptor LET-23 in vulval epithelial cells. Cell 94, 761–771. Kee, Y., Lin, R. C., Hsu, S. C., and Scheller, R. H. (1995). Distinct domains of syntaxin are required for synaptic vesicle fusion complex formation and dissociation. Neuron 14, 991–998. Kee, Y., and Scheller, R. H. (1996). Localization of synaptotagmin-binding domains on syntaxin. J. Neurosci. 16, 1975–1981. Koenig, J. H., Kosaka, T., and Ikeda, K. (1989). The relationship between the number of synaptic vesicles and the amount of transmitter released. J. Neurosci. 9, 1937–1942. Korteweg, N., Denekamp, F. A., Verhage, M., and Burbach, J. P. (2000). DiVerent spatiotemporal expression of DOC2 genes in the developing rat brain argues for an additional, nonsynaptic role of DOC2B in early development. Eur. J. Neurosci. 12, 165–171. Koushika, S. P., Richmond, J. E., Hadwiger, G., Weimer, R. M., Jorgensen, E. M., and Nonet, M. L. (2001). A post-docking role for active zone protein Rim. Nat. Neurosci. 4, 997–1005. Leenders, A. G., Lopes da Silva, F. H., Ghijsen, W. E., and Verhage, M. (2001). Rab3a is involved in transport of synaptic vesicles to the active zone in mouse brain nerve terminals. Mol. Biol. Cell 12, 3095–3102.


109

Li, J., and Schwarz, T. L. (1999). Genetic evidence for an equilibrium between docked and undocked vesicles. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 354, 299–306. Littleton, J. T., Bai, J., Vyas, B., Desai, R., Baltus, A. E., Garment, M. B., Carlson, S. D., Ganetzky, B., and Chapman, E. R. (2001). Synaptotagmin mutants reveal essential functions for the C2B domain in Ca2+-triggered fusion and recycling of synaptic vesicles in vivo. J. Neurosci. 21, 1421–1433. Littleton, J. T., Chapman, E. R., Kreber, R., Garment, M. B., Carlson, S. D., and Ganetzky, B. (1998). Temperature-sensitive paralytic mutations demonstrate that synaptic exocytosis requires SNARE complex assembly and disassembly. Neuron 21, 401–413. Littleton, J. T., Stern, M., Perin, M., and Bellen, H. J. (1994). Calcium dependence of neurotransmitter release and rate of spontaneous vesicle fusions are altered in Drosophila synaptotagmin mutants. Proc. Natl. Acad. Sci. USA 91, 10888–10892. Littleton, J. T., Stern, M., Schulze, K., Perin, M., and Bellen, H. J. (1993). Mutational analysis of Drosophila synaptotagmin demonstrates its essential role in Ca(2+)-activated neurotransmitter release. Cell 74, 1125–1134. Llinas, R., Steinberg, I. Z., and Walton, K. (1981). Relationship between presynaptic calcium current and postsynaptic potential in squid giant synapse. Biophys. J. 33, 323–351. Lonart, G., and Sudhof, T. C. (2000). Assembly of SNARE core complexes prior to neurotransmitter release sets the readily releasable pool of synaptic vesicles. J. Biol. Chem. 275, 27703–27707. Marsal, J., Ruiz-Montasell, B., Blasi, J., Moreira, J. E., Contreras, D., Sugimori, M., and Llinas, R. (1997). Block of transmitter release by botulinum C1 action on syntaxin at the squid giant synapse. Proc. Natl. Acad. Sci. USA 94, 14871–14876. Martelli, A. M., Baldini, G., Tabellini, G., Koticha, D., and Bareggi, R. (2000). Rab3A and Rab3D control the total granule number and the fraction of granules docked at the plasma membrane in PC12 cells. TraYc 1, 976–986. Maruyama, I. N., and Brenner, S. (1991). A phorbol ester/diacylglycerol-binding protein encoded by the unc-13 gene of Caenorhabditis elegans. Proc. Natl. Acad. Sci. USA 88, 5729–5733. Masuda, E. S., Huang, B. C., Fisher, J. M., Luo, Y., and Scheller, R. H. (1998). Tomosyn binds t-SNARE proteins via a VAMP-like coiled coil. Neuron 21, 479–480. Missler, M., Zhang, W., Rohlmann, A., Kattenstroth, G., Hammer, R. E., Gottmann, K., and Sudhof, T. C. (2003). Alpha-neurexins couple Ca2+ channels to synaptic vesicle exocytosis. Nature 423, 939–948. Misura, K. M., Scheller, R. H., and Weis, W. I. (2000). Three-dimensional structure of the neuronal-Sec1-syntaxin 1a complex. Nature 404, 355–362. Mochida, S., Orita, S., Sakaguchi, G., Sasaki, T., and Takai, Y. (1998). Role of the Doc2 alphaMunc13-1 interaction in the neurotransmitter release process. Proc. Natl. Acad. Sci. USA 95, 11418–11422. Morgan, J. R., Augustine, G. J., and Lafer, E. M. (2002). Synaptic vesicle endocytosis: The races, places, and molecular faces. Neuromolecular Med. 2, 101–114. Moriyama, Y., Maeda, M., and Futai, M. (1990). Energy coupling of L-glutamate transport and vacuolar H(+)-ATPase in brain synaptic vesicles. J. Biochem. (Tokyo) 108, 689–693. Murthy, V. N., Schikorski, T., Stevens, C. F., and Zhu, Y. (2001). Inactivity produces increases in neurotransmitter release and synapse size. Neuron 32, 673–682. Murthy, V. N., and Stevens, C. F. (1999). Reversal of synaptic vesicle docking at central synapses. Nat. Neurosci. 2, 503–507. Nielsen, E., Christoforidis, S., Uttenweiler-Joseph, S., Miaczynska, M., Dewitte, F., Wilm, M., Hoflack, B., and Zerial, M. (2000). Rabenosyn-5, a novel Rab5 eVector, is complexed with hVPS45 and recruited to endosomes through a FYVE finger domain. J. Cell Biol. 151, 601–612.

110

Weimer and Richmond

Niemeyer, B. A., and Schwarz, T. L. (2000). SNAP-24, a Drosophila SNAP-25 homologue on granule membranes, is a putative mediator of secretion and granule–granule fusion in salivary glands. J. Cell. Sci. 113, 4055–4064. Nonet, M. L., Staunton, J. E., Kilgard, M. P., Fergestad, T., Hartwieg, E., Horvitz, H. R., Jorgensen, E. M., and Meyer, B. J. (1997). Caenorhabditis elegans rab-3 mutant synapses exhibit impaired function and are partially depleted of vesicles. J. Neurosci. 17, 8061–8073. Novick, P., and Schekman, R. (1979). Secretion and cell-surface growth are blocked in a temperature-sensitive mutant of Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA 76, 1858–1862. O’Connor, V., Heuss, C., De Bello, W. M., Dresbach, T., Charlton, M. P., Hunt, J. H., Pellegrini, L. L., Hodel, A., Burger, M. M., Betz, H., et al. (1997). Disruption of syntaxinmediated protein interactions blocks neurotransmitter secretion. Proc. Natl. Acad. Sci. USA 94, 12186–12191. Okamoto, M., and Sudhof, T. C. (1997). Mints, Munc18-interacting proteins in synaptic vesicle exocytosis. J. Biol. Chem. 272, 31459–31464. Orita, S., Sasaki, T., Komuro, R., Sakaguchi, G., Maeda, M., Igarashi, H., and Takai, Y. (1996). Doc2 enhances Ca2+-dependent exocytosis from PC12 cells. J. Biol. Chem. 271, 7257–7260. Orita, S., Sasaki, T., Naito, A., Komuro, R., Ohtsuka, T., Maeda, M., Suzuki, H., Igarashi, H., and Takai, Y. (1995). Doc2: A novel brain protein having two repeated C2-like domains. Biochem. Biophys. Res. Commun. 206, 439–448. Parlati, F., Weber, T., McNew, J. A., Westermann, B., Sollner, T. H., and Rothman, J. E. (1999). Rapid and eYcient fusion of phospholipid vesicles by the alpha-helical core of a SNARE complex in the absence of an N-terminal regulatory domain. Proc. Natl. Acad. Sci. USA 96, 12565–12570. Peterson, M. R., and Emr, S. D. (2001). The class c vps complex functions at multiple stages of the vacuolar transport pathway. TraYc 2, 476–486. Pevsner, J., Hsu, S. C., Braun, J. E., Calakos, N., Ting, A. E., Bennett, M. K., and Scheller, R. H. (1994). Specificity and regulation of a synaptic vesicle docking complex. Neuron 13, 353–361. Piper, R. C., Whitters, E. A., and Stevens, T. H. (1994). Yeast Vps45p is a Sec1p-like protein required for the consumption of vacuole-targeted, post-Golgi transport vesicles. Eur. J. Cell. Biol. 65, 305–318. Price, A., Seals, D., Wickner, W., and Ungermann, C. (2000). The docking stage of yeast vacuole fusion requires the transfer of proteins from a cis-SNARE complex to a Rab/Ypt protein. J. Cell. Biol. 148, 1231–1238. Reim, K., Mansour, M., Varoqueaux, F., McMahon, H. T., Sudhof, T. C., Brose, N., and Rosenmund, C. (2001). Complexins regulate a late step in Ca2+-dependent neurotransmitter release. Cell 104, 71–81. Reist, N. E., Buchanan, J., Li, J., DiAntonio, A., Buxton, E. M., and Schwarz, T. L. (1998). Morphologically docked synaptic vesicles are reduced in synaptotagmin mutants of Drosophila. J. Neurosci. 18, 7662–7673. Richmond, J. E., Davis, W. S., and Jorgensen, E. M. (1999). UNC-13 is required for synaptic vesicle fusion in C. elegans. Nat. Neurosci. 2, 959–964. Richmond, J. E., Weimer, R. M., and Jorgensen, E. M. (2001). An open form of syntaxin bypasses the requirement for UNC-13 in vesicle priming. Nature 412, 338–341. Rieder, S. E., and Emr, S. D. (1997). A novel RING finger protein complex essential for a late step in protein transport to the yeast vacuole. Mol. Biol. Cell 8, 2307–2327. Robinson, L. J., and Martin, T. F. (1998). Docking and fusion in neurosecretion. Curr. Opin. Cell. Biol. 10, 483–492.


111

Rongo, C., Whitfield, C. W., Rodal, A., Kim, S. K., and Kaplan, J. M. (1998). LIN-10 is a shared component of the polarized protein localization pathways in neurons and epithelia. Cell 94, 751–759. Sabatini, B. L., and Regehr, W. G. (1996). Timing of neurotransmission at fast synapses in the mammalian brain. Nature 384, 170–172. Sakaguchi, G., Manabe, T., Kobayashi, K., Orita, S., Sasaki, T., Naito, A., Maeda, M., Igarashi, H., Katsuura, G., Nishioka, H., et al. (1999). Doc2alpha is an activity-dependent modulator of excitatory synaptic transmission. Eur. J. Neurosci. 11, 4262–4268. Salminen, A., and Novick, P. J. (1987). A ras-like protein is required for a post-Golgi event in yeast secretion. Cell 49, 527–538. Sato, T. K., Rehling, P., Peterson, M. R., and Emr, S. D. (2000). Class C Vps protein complex regulates vacuolar SNARE pairing and is required for vesicle docking/fusion. Mol. Cell 6, 661–671. Schiavo, G., Stenbeck, G., Rothman, J. E., and Sollner, T. H. (1997). Binding of the synaptic vesicle v-SNARE, synaptotagmin, to the plasma membrane t-SNARE, SNAP-25, can explain docked vesicles at neurotoxin-treated synapses. Proc. Natl. Acad. Sci. USA 94, 997–1001. Schluter, O. M., Schnell, E., Verhage, M., Tzonopoulos, T., Nicoll, R. A., Janz, R., Malenka, R. C., Geppert, M., and Sudhof, T. C. (1999). Rabphilin knock-out mice reveal that rabphilin is not required for rab3 function in regulating neurotransmitter release. J. Neurosci. 19, 5834–5846. Schuldiner, S., Shirvan, A., and Linial, M. (1995). Vesicular neurotransmitter transporters: From bacteria to humans. Physiol. Rev. 75, 369–392. Seals, D. F., Eitzen, G., Margolis, N., Wickner, W. T., and Price, A. (2000). A Ypt/Rab eVector complex containing the Sec1 homolog Vps33p is required for homotypic vacuole fusion. Proc. Natl. Acad. Sci. USA 97, 9402–9407. Shao, X., Davletov, B. A., Sutton, R. B., Sudhof, T. C., and Rizo, J. (1996). Bipartite Ca2+binding motif in C2 domains of synaptotagmin and protein kinase C. Science 273, 248–251. Shin, O. H., Rhee, J. S., Tang, J., Sugita, S., Rosenmund, C., and Sudhof, T. C. (2003). Sr(2+) Binding to the Ca(2+) binding site of the synaptotagmin 1 C(2)B domain triggers fast exocytosis without stimulating SNARE interactions. Neuron 37, 99–108. Shirataki, H., Kaibuchi, K., Sakoda, T., Kishida, S., Yamaguchi, T., Wada, K., Miyazaki, M., and Takai, Y. (1993). Rabphilin-3A, a putative target protein for smg p25A/rab3A p25 small GTP-binding protein related to synaptotagmin. Mol. Cell Biol. 13, 2061–2068. Simske, J. S., Kaech, S. M., Harp, S. A., and Kim, S. K. (1996). LET-23 receptor localization by the cell junction protein LIN-7 during C. elegans vulval induction. Cell 85, 195–204. Sollner, T., Bennett, M. K., Whiteheart, S. W., Scheller, R. H., and Rothman, J. E. (1993a). A protein assembly–disassembly pathway in vitro that may correspond to sequential steps of synaptic vesicle docking, activation, and fusion. Cell 75, 409–418. Sollner, T., Whiteheart, S. W., Brunner, M., Erdjument-Bromage, H., Geromanos, S., Tempst, P., and Rothman, J. E. (1993b). SNAP receptors implicated in vesicle targeting and fusion. Nature 362, 318–324. Stahl, B., Chou, J. H., Li, C., Sudhof, T. C., and Jahn, R. (1996). Rab3 reversibly recruits rabphilin to synaptic vesicles by a mechanism analogous to raf recruitment by ras. EMBO J. 15, 1799–1809. Staunton, J., Ganetzky, B., and Nonet, M. L. (2001). Rabphilin potentiates soluble N-ethylmaleimide sensitive factor attachment protein receptor function independently of rab3. J. Neurosci. 21, 9255–9264. Sudhof, T. C. (1995). The synaptic vesicle cycle: A cascade of protein–protein interactions. Nature 375, 645–653.

112

Weimer and Richmond

Sudhof, T. C., and Scheller, R. H. (2001). Mechanism and regulation of neurotransmitter release. Synapses 177–216. Sutton, R. B., Fasshauer, D., Jahn, R., and Brunger, A. T. (1998). Crystal structure of a SNARE complex involved in synaptic exocytosis at 2.4 A resolution. Nature 395, 347–353. Tall, G. G., Hama, H., DeWald, D. B., and Horazdovsky, B. F. (1999). The phosphatidylinositol 3-phosphate binding protein Vac1p interacts with a Rab GTPase and a Sec1p homologue to facilitate vesicle-mediated vacuolar protein sorting. Mol. Biol. Cell 10, 1873–1889. Tokumaru, H., Umayahara, K., Pellegrini, L. L., Ishizuka, T., Saisu, H., Betz, H., Augustine, G. J., and Abe, T. (2001). SNARE complex oligomerization by synaphin/complexin is essential for synaptic vesicle exocytosis. Cell 104, 421–432. Toonen, R. F., and Verhage, M. (2003). Vesicle traYcking: Pleasure and pain from SM genes. Trends Cell Biol. 13, 177–186. Torii, S., Takeuchi, T., Nagamatsu, S., and Izumi, T. (2004). Rab27 eVector granuphilin promotes the plasma membrane targeting of insulin granules via interaction with syntaxin 1a. J. Biol. Chem. 279, 22532–22538. Tucker, W. C., Weber, T., and Chapman, E. R. (2004). Reconstitution of Ca2+-regulated membrane fusion by synaptotagmin and SNAREs. Science 304, 435–438. Ubach, J., Zhang, X., Shao, X., Sudhof, T. C., and Rizo, J. (1998). Ca2+ binding to synaptotagmin: How many Ca2+ ions bind to the tip of a C2-domain? EMBO J. 17, 3921–3930. Ungermann, C., Sato, K., and Wickner, W. (1998). Defining the functions of trans-SNARE pairs. Nature 396, 543–548. Verhage, M., de Vries, K. J., Roshol, H., Burbach, J. P., Gispen, W. H., and Sudhof, T. C. (1997). DOC2 proteins in rat brain: Complementary distribution and proposed function as vesicular adapter proteins in early stages of secretion. Neuron 18, 453–461. Verhage, M., Maia, A. S., Plomp, J. J., Brussaard, A. B., Heeroma, J. H., Vermeer, H., Toonen, R. F., Hammer, R. E., van den Berg, T. K., Missler, M., et al. (2000). Synaptic assembly of the brain in the absence of neurotransmitter secretion. Science 287, 864–869. Vilinsky, I., Stewart, B. A., Drummond, J., Robinson, I., and Deitcher, D. L. (2002). A Drosophila SNAP-25 null mutant reveals context-dependent redundancy with SNAP-24 in neurotransmission. Genetics 162, 259–271. Voets, T., Toonen, R. F., Brian, E. C., de Wit, H., Moser, T., Rettig, J., Sudhof, T. C., Neher, E., and Verhage, M. (2001). Munc18-1 promotes large dense-core vesicle docking. Neuron 31, 581–591. Wang, Y., Okamoto, M., Schmitz, F., Hofmann, K., and Sudhof, T. C. (1997). Rim is a putative Rab3 eVector in regulating synaptic-vesicle fusion. Nature 388, 593–598. Weber, T., Zemelman, B. V., McNew, J. A., Westermann, B., Gmachl, M., Parlati, F., Sollner, T. H., and Rothman, J. E. (1998). SNAREpins: Minimal machinery for membrane fusion. Cell 92, 759–772. Weimer, R. M., Richmond, J. E., Davis, W. S., Hadwiger, G., Nonet, M. L., and Jorgensen, E. M. (2003). Defects in synaptic vesicle docking in unc-18 mutants. Nat. Neurosci. 6, 1023–1030. Whitfield, C. W., Benard, C., Barnes, T., Hekimi, S., and Kim, S. K. (1999). Basolateral localization of the Caenorhabditis elegans epidermal growth factor receptor in epithelial cells by the PDZ protein LIN-10. Mol. Biol. Cell 10, 2087–2100. Wickelgren, W. O., Leonard, J. P., Grimes, M. J., and Clark, R. D. (1985). Ultrastructural correlates of transmitter release in presynaptic areas of lamprey reticulospinal axons. J. Neurosci. 5, 1188–1201. Wilson, D. W., Whiteheart, S. W., Wiedmann, M., Brunner, M., and Rothman, J. E. (1992). A multisubunit particle implicated in membrane fusion. J. Cell Biol. 117, 531–538.


113

Wu, M. N., Fergestad, T., Lloyd, T. E., He, Y., Broadie, K., and Bellen, H. J. (1999). Syntaxin 1A interacts with multiple exocytic proteins to regulate neurotransmitter release in vivo. Neuron 23, 593–605. Wu, M. N., Littleton, J. T., Bhat, M. A., Prokop, A., and Bellen, H. J. (1998). ROP, the Drosophila Sec1 homolog, interacts with syntaxin and regulates neurotransmitter release in a dosage-dependent manner. EMBO J. 17, 127–139. Wurmser, A. E., Sato, T. K., and Emr, S. D. (2000). New component of the vacuolar class C-Vps complex couples nucleotide exchange on the Ypt7 GTPase to SNARE-dependent docking and fusion. J. Cell Biol. 151, 551–562. Xu, T., Rammner, B., Margittai, M., Artalejo, A. R., Neher, E., and Jahn, R. (1999). Inhibition of SNARE complex assembly diVerentially aVects kinetic components of exocytosis. Cell 99, 713–722. Yizhar, O., Matti, U., Melamed, R., Hagalili, Y., Bruns, D., Rettig, J., and Ashery, U. (2004). Tomosyn inhibits priming of large dense-core vesicles in a calcium-dependent manner. Proc. Natl. Acad. Sci. USA 101, 2578–2583. Zenisek, D., Steyer, J. A., and Almers, W. (2000). Transport, capture, and exocytosis of single synaptic vesicles at active zones. Nature 406, 849–854. Zhang, X., Rizo, J., and Sudhof, T. C. (1998). Mechanism of phospholipid binding by the C2Adomain of synaptotagmin I. Biochemistry 37, 12395–12403. Zucker, R. S. (1996). Exocytosis: A molecular and physiological perspective. Neuron 17, 1049–1055.

4

____________________________________________________________________________

ATP-Dependent Chromatin Remodeling Corey L. Smith and Craig L. Peterson Program in Molecular Medicine University of Massachusetts Medical School Worcester, Massachusetts 01605

I. Chromatin Structure: A Short Primer II. ATP-Dependent Chromatin Remodeling Enzymes III. ATP-Dependent Chromatin Remodeling Enzymes are Involved in the Control of Numerous Cellular Processes A. The SWI2/SNF2 Complexes: Transcriptional Regulators B. ISWI Complexes: Sliding into Transcription Regulation and Chromatin Assembly C. Mi-2 Complexes: General Repressors D. Other Subfamilies: Repair and Establishing Chromatin Domains IV. Understanding the Molecular Mechanism of ATP-Dependent Chromatin Remodeling A. The SWI2/SNF2-Like ATPase Subunit is the Master Switch B. Chromatin Remodeling Enzymes are Able to Introduce Helical Torsion into DNA and Nucleosomal Substrates C. The Chromatin Remodeling Enzyme–Nucleosome Interface D. How is ATP Hydrolysis Coupled to the Generation of Remodeled Chromatin? E. Nucleosome Accessibility and Mobilization by Chromatin Remodeling Enzymes F. Disruption of Nucleosome Structure: Moving Dimers Around G. Determining How Remodeling Works In Vivo H. Remodeling at the Fiber Level V. Concluding Remarks Acknowledgments References

The study of chromatin and how this dynamic structure modulates events in the eukaryotic nucleus has become an increasingly important topic in biomedical research. A large number of enzymes have been discovered that are responsible for modifying and altering chromatin structure, either globally or specifically at particular gene promoters or regions of the chromosome. This chapter provides an introduction to the structure of chromatin and then describes how special classes of enzymes modulate chromatin structure to allow access to DNA. C 2005, Elsevier Inc.

I. Chromatin Structure: A Short Primer The structure and function of chromatin is inherently dynamic. During mitosis, individual chromatids become highly compact, align on the metaphase plate, Current Topics in Developmental Biology, Vol. 65 Copyright 2005, Elsevier Inc. All rights reserved. 0070-2153/05 $35.00

115

116

Smith and Peterson

separate into mother and daughter cells, and then decondense after anaphase. In interphase, local chromatin structure at gene promoters must be perturbed to allow the binding of a multitude of factors necessary for proper gene activation. Likewise, detection and repair of DNA damage must take place in the context of a chromatin environment. Furthermore, chromatin structure must also be maintained and propagated during replication. Chromatin structure is an interesting paradox. How do you maintain a compacted genome that will fit in the eukaryotic nucleus while still maintaining a DNA template that is readily accessible for replication, transcription, and DNA damage repair? The answer lies in the fact that chromatin is a dynamic structure with many layers of complexity and regulation. The basic unit of chromatin, the nucleosome core particle, consists of two copies of each of the four core histones, H2A, H2B, H3, and H4. The histones H3 and H4 fold together to form a tetramer to which two H2A-H2B dimers bind, resulting in the canonical histone octamer. Around this roughly cylindrical octamer are wrapped 147 base pairs of DNA (see Fig. 2; Luger et al., 1997). The individual histones have short N-terminal (15–40 amino acids in length), and in some cases, C-terminal domains, which radiate out from the core nucleosome structure. Amino acid residues in these ‘‘tails’’ have been found to be the targets of numerous post-translational modifications. The contacts made between DNA and histones are important for the stability and organization of the nucleosome core particle. It has been demonstrated that the mechanical forces needed to disrupt DNA-histone contacts at the entry/exit sites of DNA around the nucleosome are lower than at the central, dyad axis (Brower-Toland et al., 2002). It has also been shown, in vitro, that there is a slow intrinsic, yet spontaneous, accessibility of DNA in the absence of nucleosome movement (Anderson et al., 2002). This nucleosome breathing could, in vivo, allow protein binding at the edges of nucleosomes. Once bound, these proteins might recruit other chromatin modifying enzymes, which might then disrupt the stronger histone-DNA interactions located near the nucleosome dyad. At the next layer of complexity, nucleosomes are arranged into long linear arrays with a width of 10 nm (see Fig. 1). These linear arrays are further compacted by intra- and inter-nucleosomal interactions into compacted chromatin fibers. Such interactions are mediated and stabilized by association of the histone N-terminal domains with neighboring nucleosomes. As such, removal of the histone tails eliminates the folding of model nucleosomal arrays in vitro (Carruthers and Hansen, 2000). The incorporation of linker histones into the nucleosomal arrays also stabilizes array folding. These histones are diVerent from the canonical histones in that they are not found within the nucleosome core particle. Linker histones (e.g., H1 and H5) bind at the DNA entry/exit point on the nucleosome with a stoichiometry of one linker histone per nucleosome in vivo (Hansen, 2002). The binding

4. ATP-Dependent Chromatin Remodeling

117

Figure 1 Chromatin fiber condensation. The various levels of compaction of the chromatid fiber are illustrated here. A number of chromatin-associated proteins are involved in organizing chromatin folding from the simple ‘‘bead on a string’’ array to the fully condensed G1 chromatid.

of linker histones stabilizes an additional 20 bp of DNA with the nucleosome (167 bp total), forming a particle called a chromatosome. While abundant in vivo, linker histones alone are not suYcient for folding; the histone N-termini are still necessary (Carruthers and Hansen, 2000; Garcia-Ramirez et al., 1992).

118

Smith and Peterson

As chromatin compaction increases further, long-range interactions, either direct or mediated through other nonhistone proteins, increase the level of folding and condensation (WolVe, 1998). Several groups have used electron microscopy (EM) or atomic force microscopy (AFM) to investigate the structure of nucleosomal arrays. These experiments have used both in vitrogenerated arrays as well as chromatin fibers isolated from chicken erythrocytes (Bednar et al., 1995; Zlatanova and Leuba, 2003). At low ionic strengths, the extended linear arrays of nucleosomes appear as beads on a string. At higher ionic strengths, these arrays fold into a thick, sausagelike fiber that is commonly known as the 30 nm fiber (Carruthers et al., 1998; Hansen et al., 1989). Electron microscopy and analytical ultracentrifugation of both folded and unfolded arrays have elucidated some of the characteristics and shapes that chromatin arrays can assume (see Fig. 1). The addition of linker histones into nucleosomal arrays creates a zigzag repeating pattern in the arrays seen in chromatin fibers purified from chicken erythrocytes (Bednar et al., 1998). This has led to two models for folding into the 30 nm fiber. At this level of compaction, the chromatin arrays could fold either into a solenoid structure, with the nucleosomes forming a helical arrangement, or into a zigzag pattern, where nucleosomes are positioned such that the entry/exit sides are buried inside of the chromatin fiber (Hansen, 2002). These theories are still being investigated to determine exactly how chromatin folds in vivo. Numerous variants exist for some of the histones, which are incorporated into nucleosomes in diVerent regions of chromatin and play specific roles in the organization of chromatin and the establishment of specific domains with diVerent folding characteristics (Horn and Peterson, 2002). The basic histone fold remains highly conserved from yeast to humans but the composition of nucleosomes at diVerent regions within chromatin can change. For instance, at centromeres, the histone H3 variant centromere protein A (CENP-A) replaces the major form of H3 (Ahmad and HenikoV, 2001; Lo et al., 2001). While evidence supports that H3 is replaced, all the other histones—H2A, H2B, and H4—are still present. Centromeres are known areas of heterochromatic composition with higher than average levels of compaction, and CENP-A might play a key role in the maintenance of this condensation. Disruption of CENP-A gene expression in yeast, flies, and worms has illustrated that deposition of this histone variant is important for proper generation of new centromeres (Smith, 2002). Interestingly, it was found that CENP-A is deposited in a replication-independent manner, which seems to be true of all the histone variants (Ahmad and HenikoV, 2001). Other histone variants are known to exist, such as H2AZ, H2AX, macroH2A, and H3.3 (Ahmad and HenikoV, 2002; Ladurner, 2003; Redon et al., 2002). While it seems that many histone variants are involved in the regulation of specialized chromatin structures such as telomeres and centromeres, there is


119

a well-characterized histone variant, H2AZ, which has been linked to activation and repression of gene activation (Horn and Peterson, 2002). This variant is found in numerous regions of both the Drosophila and yeast genomes, and in Drosophila H2AZ is essential and cannot be substituted for by canonical H2A (Leach et al., 2000; Santisteban et al., 2000). In vitro, nucleosomal arrays that contain H2AZ are defective for intramolecular folding, suggesting that chromatin domains containing H2AZ might not be compacted to the same degree as chromatin regions containing only H2A (Fan et al., 2002). In general, the incorporation of diVerent histone variants within chromatin fibers is likely to create specialized domains with distinct properties.

II. ATP-Dependent Chromatin Remodeling Enzymes Among the major contributors to the dynamic nature of chromatin are the chromatin modification and remodeling enzymes. The first class of enzymes, the histone modifying enzymes, directly add or remove posttranslational modifications to amino acids in the various histone proteins. As seen in Fig. 2, numerous histone modifications have been found to date (Khorasanizadeh, 2004). Lysine residues can be the targets of acetylation, methylation (mono-, di-, or trimethylated), or ubiquitination. Serines and threonines are phosphorylated and arginines can either be mono- or dimethylated. These modifications are likely to alter the structure or function of chromatin fibers. Indeed, diVerent modifications are associated with distinct chromatin-mediated events such as transcriptional activation, silencing, and histone deposition. For example, histone hyperacetylation usually correlates with transcriptionally active regions, whereas methylation of H3 at lysine 9 correlates well with transcriptional repression (Hake et al., 2004). Much of the current research into chromatin biology has focused on these histone modifications and their role in the regulation of chromatin-mediated events (Fischle et al., 2003; Hake et al., 2004; Jenuwein and Allis, 2001; Khorasanizadeh, 2004). The second class of chromatin remodeling enzymes contains those enzymes that alter chromatin fiber structure by disrupting or mobilizing nucleosomes in an energy-dependent manner (ATP-dependent chromatin remodeling enzymes). For the purpose of this work, we focus on the roles that ATP-dependent chromatin remodeling enzymes play in altering nucleosomal and chromatin structure. ATP-dependent chromatin remodeling enzymes use the free energy derived from the hydrolysis of hundreds of molecules of ATP per minute to disrupt chromatin structure. These enzymes range from a single catalytic subunit to multi-subunit complexes that can exceed 1 MDa in mass. At the heart of each ATP-dependent chromatin

120

Smith and Peterson

Figure 2 Structural features of the nucleosome core particle. (a) Structure of the nucleosome with histones H2A (red), H2B (light blue), H3 (green), and H4 (yellow) represented based on the ˚ crystal structure (Luger et al., 1997). The path of the DNA is represented by the curve 1.9A around the histone octamer. The various shades represent the wraps of DNA. (b) Features of the N- and C-terminal tail regions of core histones. The colored shapes represent known modifications of the amino acid residues in the tails. The histone tails can be methylated (blue circles) at lysines and arginines, phosphorylated (yellow crosses) at serines, ubiquitylated (orange stars) at lysines, and acetylated (green triangles) at lysines.

remodeling enzyme is a helicase-like subunit of the SWI2/SNF2 family of SF2 helicases. This class of ATPases has been further subdivided into at least three major subfamilies: the SWI2/SNF2, Mi-2/CHD, and ISWI families, as


121

well as a potentially new family of Ino80-like complexes. These family assignments are based primarily on sequence homology within the catalytic subunit as well as the peculiarities of their remodeling activities (Boyer et al., 2000a; Eisen et al., 1995; Shen et al., 2000). The helicase-containing subunits in these enzymes are large multi-domain proteins that contain additional domains, including bromodomains, plecton homology domains (PHD), chromodomains, SANT domains, and AT hook regions (see Fig. 3). These other domains might play a role in stabilizing interactions with histones and/or nucleosomal DNA. For instance, bromodomains interact with acetylated lysines, AT hook regions interact with the minor groove of AT-rich regions of DNA, and SANT domains are believed to interact with histone tails (Aravind and Landsman, 1998; Boyer et al., 2000a; Goodwin and Nicolas, 2001). The hallmark of ATP-dependent chromatin remodeling enzymes is the ability to remodel chromatin by altering the DNA-histone contacts within individual nucleosome, resulting in either localized disruption of the histone– DNA contacts or mobilization of the nucleosomes on the chromatin fiber. In the rest of this chapter, we focus on the roles of these enzymes in vivo as well as how they utilize ATP hydrolysis to remodel chromatin structure. As will be discussed, members of each subfamily appear to play unique roles in vivo.

III. ATP-Dependent Chromatin Remodeling Enzymes are Involved in the Control of Numerous Cellular Processes The discovery of the first chromatin remodeling enzyme, the S. cerevisiae SWI/SNF complex, was the result of several early genetic studies of two yeast genes, HO and SUC2. The SWI1, SWI2, and SWI3 genes were originally found to act as positive regulators of HO transcription, the endonuclease involved in mating type switching, hence, the name SWItch genes (Stern et al., 1984). At the same period in time, three other genes, SNF2, SNF5, and SNF6 (Sucrose Non Fermentors), were found to be positive regulators of SUC2, encoding invertase, which is needed for yeast to utilize sucrose as a carbon source (Neigeborn and Carlson, 1984). Subsequent analysis showed that SWI2 and SNF2 were, in fact, the same gene and that all five of these gene products functioned together in a complex as positive regulators of transcription (Peterson and Herskowitz, 1992; Peterson et al., 1994). Genetic interactions between SWI and SNF genes and genes encoding components of chromatin were also observed. A mutant screen for SWI-Independent, or SIN, genes that could alleviate the eVects of swi mutations, identified two chromatin proteins encoded by the SIN1 and SIN2 genes (Kruger and Herskowitz, 1991; Peterson et al., 1991). The SIN1 gene encodes a nonhistone protein with homology to HMG1/2 proteins and the SIN2 gene

122

Smith and Peterson

Figure 3 Classes of ATP-dependent chromatin remodeling enzymes. This chart depicts representative features of each subfamily of chromatin remodeling enzymes. The tables list a number of key chromatin remodeling enzymes with their approximate sizes, ATPase subunit, total subunit compositions, and the organism in which they are found.


123

was found to encode histone H3. These genetic studies suggested that this large megadalton complex regulated transcription by antagonizing chromatin structure. The identification of SWI/SNF then led to the subsequent identification of numerous, related ATP-dependent complexes similarly involved in the alteration of chromatin (Vignali et al., 2000).

A. The SWI2/SNF2 Complexes: Transcriptional Regulators The yeast SWI/SNF complex comprises eleven subunits (encoded by the SWI1, SWI2/SNF2, SWI3, SNF5, SNF6, SNF11, SWP82, SWP73, SWP29, ARP7, and ARP9 genes) (Cairns et al., 1996a,b, 1998; Peterson et al., 1994, 1998). The catalytic subunit of the SWI/SNF complex, Swi2p, contains a bromodomain and an AT-hook region as well as the helicase-like ATPase domain. Bromodomains are believed to be involved in interactions with histone tails and the AT-hook region appears to be involved in binding to AT-rich regions of DNA: these domains might be important for targeting the ATPase subunit to chromatin (Aravind and Landsman, 1998; Boyer et al., 2004). Swi2 homologs are found in all eukaryotes, including Drosophila (Brahma) and humans (hBRM and BRG1 complexes) (Martens and Winston, 2003). Sth1p, a very close yeast paralog of Swi2p, is found in the RSC chromatin remodeling complex (Cairns et al., 1996c; Du et al., 1998; Tsuchiya et al., 1998). SWI/SNF is important in the activation of a subset of highly inducible genes in yeast, including genes involved in metabolism (HIS3, SUC2, INO1, and PHO8) and mating-type switching (HO) (Winston and Carlson, 1992). Although SWI/SNF activity is only required for 5% of constitutively expressed yeast genes, an important subset of highly inducible genes requires SWI/SNF (Holstege et al., 1998; Krebs et al., 2000; Sudarsanam et al., 2000). Likewise, gene expression at the end of mitosis also appears to require global SWI/SNF activity (Krebs et al., 2000). Interestingly the Caenorhabditis elegans homolog of SWI/SNF has been shown to be required in late mitosis for the asymmetric division of T cells (Sawa et al., 2000). A global role for SWI/SNF during mitosis is consistent with the view that SWI/SNF might regulate higher order chromatin structure in vivo (Horn et al., 2002; Krebs and Peterson, 2000). The Brahma complex (BRM), a homolog of SWI/SNF in Drosophila, is required for maintenance of homeotic gene expression and E2F-dependent transcription (Kennison and Tamkun, 1988; Staehling-Hampton et al., 1999). In adult flies loss of Brahma causes defects in the peripheral nervous system as well as a more general decrease in cell viability at the larval stages (Elfring et al., 1998). The BRM protein has been found in two separable Drosophila complexes, BAP (Brahma-associated proteins) and

124

Smith and Peterson

PBAP (polybromo Brahma-associated proteins), which contain diVerences in only a few subunits (Mohrmann et al., 2004). Both of these complexes colocalize to regions of hyperacetylated chromatin on polytene chromosomes in a distinct yet overlapping pattern. The BAP complex also appears to be involved in repression of the wingless target genes nubbin, Distal-less, and decapentaplegic, which are all crucial for proper patterning of numerous structures during fly development (Collins and Treisman, 2000). In mammals, homologs of SWI/SNF subunits (BRG1 and INI1/SNF5) are essential for early mouse development. A highly conserved subunit of SWI/SNF, INI1/SNF5, has been found to be crucial in early steps of fetal development and loss of heterozygosity of INI1/SNF5 occurs during tumor formation in mice (Guidi et al., 2001; Klochendler-Yeivin et al., 2000; Roberts et al., 2000). Likewise, in human cells, the various SWI/SNF chromatin remodeling enzymes have been found to play major roles in cell diVerentiation, early development, and tumor suppression (Huang et al., 2003; Muller and Leutz, 2001; Neely and Workman, 2002). For instance, induction of muscle and adipocyte cell diVerentiation requires the hSWI2/SNF2 homologs BRG1 and BRM (de la Serna et al., 2001; Salma et al., 2004). BRG1 has also been found to activate a subset of IFN- inducible genes in humans, linking SWI/SNF remodeling to regulation of cytokine-mediated gene expression (Huang et al., 2002). Many links to human disease exist for the human SWI/SNF complexes. Evidence suggests that the ATPase subunits (BRG1 and hBRM) of the hSWI/SNF complexes can act as tumor suppressors in their own right. BRG1 and BRM mutations have been seen in primary lung cancers and gastric carcinomas (Reisman et al., 2003; Sentani et al., 2001). A number of diVerent mutations have been discovered in the BRG1 gene in numerous other cancers including breast, lung, prostate, and pancreatic cancers (Wong et al., 2000). Mutations within hBRM as well as BRG1, BAF155/SWI3, BAF180, and BAF250 genes have also been found to be mutated in cancer cell lines (Decristofaro et al., 2001). INI1/hSNF5 has also been linked to cancer in humans. Mutations in INI1/hSNF5 have been found in pediatric malignant rhabdoid tumors (Versteege et al., 1998). The link between chromatin remodeling enzymes and disease is currently experiencing a rapid growth as more direct links are being found with SWI/SNF-like enzymes.

B. ISWI Complexes: Sliding into Transcription Regulation and Chromatin Assembly Another class of energy-dependent chromatin remodeling enzymes is the ISWI family. The ISWI family contains multiple complexes found in yeast (ISW1a, ISW1b, and ISW2), flies (NURF, CHRAC, and ACF), and


125

higher eukaryotes (RSF, ACF, WCRF, and CHRAC, reviewed in Langst and Becker, 2001b). In contrast to the SWI2/SNF2 family members, ISWI ATPases contain SANT and SLIDE domains (Boyer et al., 2004; Grune et al., 2003). These domains share homology to the c-myb DNA binding module and have been proposed to play a role in histone tail (SANT domain) and nucleosomal DNA (SLIDE) interactions (Boyer et al., 2004). Whereas the SWI/SNF complexes appear dedicated to transcriptional control, the ISWI family members appear to participate in a variety of nuclear processes. In Drosophila, ISWI genes are essential for viability and have been genetically associated with numerous nuclear processes (Corona and Tamkun, 2004). Like SWI/SNF, one ISWI-containing complex, NURF, activates transcription via chromatin remodeling at the Drosophila hsp70 and ftz promoters (Okada and Hirose, 1998; Tsukiyama et al., 1994). ISWI complexes also play roles in repression. For example, Tamkun and colleagues saw no evidence for colocalization of ISWI and RNA polymerase on Drosophila polytene chromosomes, suggesting that ISWI does not play a role in transcriptional activation like the SWI2/SNF2 family of chromatin remodeling complexes (Deuring et al., 2000). Instead, ISWI complexes play a repressive role at specific genes in larval developmental stages. Studies in Drosophila lacking ISWI have also found that the male fly larvae have a high level of global decondensation in the X chromosome, suggesting a global role in the maintenance of chromosome structure (Deuring et al., 2000). The developmental role for ISWI complexes can also be illustrated in mammals, since SNF2h, the murine homolog of ISWI, has been found to be essential for early embryonic development. In Snf2h/ mice, the embryo never progresses from the pre-implantation stage (Stopka and Skoultchi, 2003). ISWI complexes have also been found to play a global role in chromatin remodeling and reprogramming of chromatin when somatic nuclei are transplanted into unfertilized eggs (Kikyo et al., 2000). While ISWI is essential in Drosophila and mice, there are two redundant copies of ISWI (ISW1 and ISW2) in yeast which are not essential (Tsukiyama et al., 1999). The Isw2p and Itc1p proteins are components of the yeast Isw2 complex, which is required for the transcriptional repression of a set of meiotic genes in conjunction with the Sin3/Rpd3 HDAC complex (Goldmark et al., 2000). Isw1p, on the other hand, is found in two diVerent complexes: Isw1a (Isw1p and Ioc3p) and Isw1b (Isw1p, Ioc2p, and Ioc4p) (Vary et al., 2003). An interesting role for the two Isw1 complexes in transcriptional elongation and termination by RNA polymerase II (RNAPII) has been found. Both of these complexes have been found to associate with RNAPII during transcription, with the Isw1a complex associated with RNAPII at the promoter prior to gene activation keeping gene transcription in an oV state by ordering nucleosomes over the promoter

126

Smith and Peterson

region. Upon activation, Isw1a becomes dissociated and Isw1b becomes the RNAPII-associated Isw1 complex (Morillon et al., 2003). It appears that Isw1b coordinates elongation, termination, and mRNA processing. These data suggest that the interplay between these two complexes is necessary for proper transcript processing. The ISWI complexes do not just play a role in transcription; ISWIcontaining complexes have also been implicated in chromatin assembly and nucleosome spacing. The ACF complex in both flies and mammals is composed of the two subunits: ISWI and Acf1. In vitro, ACF is able to assemble, space, and mobilize nucleosomes in a cell-free system (Ito et al., 1997). ACF also appears to play a role in replication-coupled histone deposition in vivo (Mello and Almouzni, 2001). ACF, as well as a related complex, WSTF, has been found to co-localize with replication foci and is required for replication through heterochromatic regions (Bozhenok et al., 2002; Collins et al., 2002). It is still not clear if these complexes are functionally redundant or if they couple replication and histone deposition in distinct ways. A number of related ISWI complexes (CHRAC, WSTF, and NURF) all contain subunits similar to ACF and are able to catalyze the mobilization of nucleosomes on arrays in vitro (Becker and Horz, 2002). The NURF complex also contains NURF-55, a subunit of CAF-1, a histone chaperone conserved in many eukaryotes (Martinez-Balbas et al., 1998). NURF, unlike CHRAC and ACF, does not space nucleosomes but rather is believed to be involved in the randomization of spaced nucleosomal arrays. So it appears that the ISWI family of ATP-dependent remodeling enzymes is involved in a number of processes, including transcriptional activation, replication-coupled histone deposition, and the creation of regions of silenced chromatin at specific promoter regions.

C. Mi-2 Complexes: General Repressors The Mi-2 (CHD) family of chromatin remodeling enzymes all contain ATPase subunits with one or more chromodomains. The chromodomains in Mi-2 appear to be responsible for binding nucleosomal DNA in a histone tail independent manner (Bouazoune et al., 2002). These enzyme complexes appear to play roles in transcriptional repression since several Mi-2 complexes have been found to contain histone deacetylase (HDAC) subunits (Kehle et al., 1998; Tong et al., 1998; Wade et al., 1998a,b). Along with HDACs, methyl-CpG binding proteins have also been found to be part of the Xenopus Mi-2 complex (Wade et al., 1999). The discovery of methylated-DNA binding proteins as part of a chromatin remodeling complex suggests that Mi-2 functions to coordinate histone deacetylation with DNA methylation in order to silence chromatin (Wade et al., 1999).


127

As with other chromatin remodeling complexes, Mi-2 complexes have been found to play important roles in development. In Arabidopsis thaliana the gene PICKLE (PKL) encodes a Mi-2 family ATPase necessary for the transition from seed to seedling (Ogas et al., 1999). In pkl mutants, the silencing of a number of embryonic genes is lost and these genes are then expressed post germination (Ogas et al., 1997; Rider et al., 2004). Mutations in Drosophila Mi-2 (dMi-2) are embryonically lethal, and in C. elegans LET418/Mi-2 is required for the maintenance of somatic cell diVerentiation, a crucial event in early embryonic development (Khattak et al., 2002; Unhavaithaya et al., 2002). Mi-2 complexes have also been found to play a role in lymphocyte cell diVerentiation. A class of Zn2þ-finger DNA-binding proteins in the Ikaros gene family interacts directly with the Mi-2 complex NURD (orthologous to dMi-2), and in murine T cells a fraction of Ikaros and Aiolos (another member of the Ikaros gene family) were found to be stably associated with the NURD complex (Kim et al., 1999). Upon T cell activation, a fraction of Ikaros, Aiolos, and NURD all become associated with heterochromatin with similar kinetics (Kim et al., 1999). Ikaros also appears to interact with SWI/SNF in T cells, and this interaction appears to be exclusive of the Ikaros-NURD complex. This link between Ikaros, Mi-2, and SWI2/SNF2 family chromatin remodeling complexes has also been observed in adult erythroid cells (O’Neill et al., 2000). Overall, it was suggested that the interplay between Ikaros-like DNA binding proteins and chromatin remodeling complexes is responsible for the silencing of the

-globin locus and - to -globin locus switching. Consistent with this proposal, evidence supports a role for Ikaros as a potentiator of chromatin remodeling at heterochromatic sites but not as a traditional activator, especially at pericentric regions of heterochromatin (Koipally et al., 2002).

D. Other Subfamilies: Repair and Establishing Chromatin Domains Recently, more proteins have been found to have the hallmarks of chromatin remodeling enzymes, including those involved in the repair of DNA damage. Rad54p, a member of the RAD52 epistasis group, plays an essential role in several steps of homologous recombinational repair of DNA double strand breaks (DSBs) (Peterson and Cote, 2004). Rad54 is a member of the SWI2/ SNF2 family of ATPases and, since 2000, it has been discovered that the Rad54 ATPase has all the in vitro characteristics of an ATP-dependent chromatin remodeling enzyme (Alexeev et al., 2003; Alexiadis and Kadonaga, 2002; JaskelioV et al., 2003). It was proposed that Rad54 functions in recombinational repair of DNA DSBs by altering and/or moving nucleosomes that might interfere with joint molecule formation or migration of heteroduplex DNA (JaskelioV et al., 2003; Peterson and Cote, 2004).

128

Smith and Peterson

Cockayne syndrome (CS) is another human disorder that involves a SWI2/SNF2 family member. This autosomal recessive disease is associated with mental retardation, cachectic dwarfism, neural degeneration, and hypersensitivity to UV light. Two diVerent genetic complementation groups exist for CS, CSA and CSB. CSB cells have defects in their ability to perform transcription-coupled repair. CSB homologs have been found in yeast (RAD26) and humans (ERCC6), both of which have been found to be homologs of the Swi2/Swi2p ATPase (Licht et al., 2003). It has been shown that mutations in the putative helicase motifs in CSB lead to abrogation of the genetic function of CSB in RNA synthesis and survival after UV treatment (Muftuoglu et al., 2002). CSB has also been found to exhibit ATPdependent chromatin remodeling activity, as illustrated by the ability to bind to and alter histone–DNA contacts in mononucleases as assayed by DNase I accessibility (Citterio et al., 2000). The precise role of CSB proteins in damage repair is still under investigation. Mutations in another SWI2/SNF2-like ATPase gene, ATRX (alphathlassemia X-linked mental retardation) occur in patients with severe X-linked mental retardation (Gibbons et al., 1995). ATRX has been found to be associated with pericentromeric heterochromatin, PML bodies, and the heterochromatin-associated protein HP1 (McDowell et al., 1999). A truncated isoform of ATRX, ATRXt, has been discovered to associate with pericentric heterochromatin regions, but not with PML bodies, suggesting a role in regulation of chromatin structure at specific chromatin regions (Garrick et al., 2004).

IV. Understanding the Molecular Mechanism of ATP-Dependent Chromatin Remodeling One of the major avenues of research centers on determining how chromatin becomes remodeled by these ATP-dependent chromatin remodeling enzymes. What exactly is the process by which nucleosomes are disrupted? Do all of these complexes use the same basic mechanism to alter nucleosomes? In this section, we outline what is known about these enzymes and their actions on nucleosomes.

A. The SWI2/SNF2-Like ATPase Subunit is the Master Switch It is increasingly obvious that these large multi-subunit complexes play a number of diverse roles in the nucleus. Do all of these ATP-dependent chromatin remodeling enzymes share the same basic mechanism or do they


129

each act in subtle yet diVerent ways? The one thing all these enzymes share is a highly conserved helicase-like ATPase domain similar to that of yeast SWI2/SNF2. The ATPase subunit of the various ATP-dependent remodeling enzymes all contain a domain with homology to helicase-like proteins of the SF1 and SF2 superfamiles. This ATPase domain contains all of the seven common helicase motifs of the SF2 superfamily (numbered Motif I, Ia, and II–VI). In helicases, these motifs function together to convert ATP hydrolysis to the strand separation activities of the enzyme (Caruthers and McKay, 2002). The SWI2/SNF2-like ATPases contain most of the consensus residues of these motifs. The canonical helicase domain is a bipartite structure with Motifs I–III on one side (subdomain I) and Motifs IV–VI to the other side (subdomain II) of the ATP-binding cleft (Caruthers and McKay, 2002). The ATPase domain of remodeling enzymes is required for remodeling activity. Single point mutations in the highly conserved ATPase/helicase motifs cause loss of function for these enzymes in vitro and in vivo (Cote et al., 1994; Khavari et al., 1993; Peterson et al., 1994; Richmond and Peterson, 1996). Similar to traditional helicases, all ATP-dependent chromatin remodeling enzymes have been found to have DNA and/or nucleosomestimulated ATPase activity, although the preferred cofactor diVers among the diVerent classes. The SWI2/SNF2 family of enzymes have similar ATPase activity in the presence of either DNA or nucleosomes, while the ISWI and Mi-2 class of enzymes display a higher ATPase activity with a nucleosomal template (Brehm et al., 2000; Cairns et al., 1996c; Corona et al., 1999; Cote et al., 1994; Guschin et al., 2000a). These diVerences may result from subtle diVerences in mechanism or in how each enzyme binds its substrate. A key diVerence between canonical helicases and ATP-dependent chromatin remodeling enzymes is the lack of actual DNA duplex strand separation. After the identification of the helicase-related motifs in Swi2/Snf2 protein, the yeast SWI/SNF complex was tested for the ability to act as a DNA helicase, but it failed to induce duplex unwinding with a variety of substrates. Furthermore, SWI/SNF action does not lead to enhanced sensitivity of nucleosomal DNA to potassium permanganate, indicating a lack of transient duplex unwinding (Cote et al., 1998). Similarly, all other chromatin remodeling enzymes tested have yet to produce any evidence of helicase-like duplex unwinding. It should be noted that the SWI2/SNF2 family of ATPases diVers from canonical helicase domains with the addition of a large 100 amino acid insertion between helicase subdomains I and II (Eisen et al., 1995). This insertion might be responsible for the lack of DNA duplex unwinding and it may play a role in the mechanism of chromatin remodeling. Since these enzymes don’t catalyze strand separation, it had been thought that maybe the SWI2/SNF2-like subunit catalyzes the translocation of the

130

Smith and Peterson

complex along DNA. Indeed, a number of these enzymes, including SWI/ SNF, RSC, and ISWI, can track along DNA as revealed by their ability to remove a short oligonucleotide incorporated into a triple helix (JaskelioV et al., 2003; Saha et al., 2002; Whitehouse et al., 2003). This assay was first used to look at and measure the rates at which type I restriction enzymes track along DNA (Firman and Szczelkun, 2000). Although this activity might suggest movement of the enzyme along DNA, it seems more likely that the enzyme will remain anchored to the histone octamer and DNA will be moved relative to the enzyme. In this case, ATP hydrolysis might catalyze the twisting or pushing of DNA across the surface of the histone octamer. The movement of the DNA then could yield a remodeled state where the octamer has moved relative to its initial position on the DNA. This model will be addressed in more detail in the following sections.

B. Chromatin Remodeling Enzymes are Able to Introduce Helical Torsion into DNA and Nucleosomal Substrates Other evidence of helicase-like behavior is seen in ATP-dependent chromatin remodeling enzymes. Havas and colleagues have tested various chromatin remodeling enzymes for the ability to introduce superhelical torsion into DNA and chromatin substrates (Havas et al., 2000). In these studies, an assay measures extrusion of a cruciform from a DNA construct containing an inverted [AT]34 repeat. If an enzyme creates superhelical torsion on DNA, the [AT]34 repeat forms a cruciform that is cleaved by the junction-resolving enzyme, T4 Endonuclease VII. Results showed that ySWI/SNF, Xenopus Mi-2 complex, recombinant ISWI, and recombinant BRG1 were all able to generate superhelical torsion in an ATP-dependent manner. Both BRG1 and SWI/SNF were able to generate torsion on both DNA and chromatin templates while Mi-2 and ISWI only functioned on nucleosomal templates. These results showed a similarity to the beforementioned DNA/nucleosome-stimulated ATPase activities of these enzymes; enzymes that are more stimulated by nucleosomes for ATPase activity also showed a need for nucleosomal substrates in the cruciform extrusion assay. The generation of torsion on the DNA duplex could either be a consequence of remodeling or it might be how DNA–histone contacts are disrupted.

C. The Chromatin Remodeling Enzyme–Nucleosome Interface Understanding how chromatin remodeling enzymes interact with the nucleosome has been a major goal in the chromatin field for some time. A few groups are starting to use single molecule methods to look at the structures


131

of these complexes as well as at the forces and dynamics involved in chromatin remodeling. Atomic force microscopy (AFM) and electron microscopy (EM) studies have shown that the interaction of SWI/SNF with nucleosomal arrays leads to formation of large DNA loops (Bazett-Jones et al., 1999; Schnitzler et al., 2001). The AFM studies also showed that clustering of nucleosomes occurs on short (dodecameric) nucleosomal arrays, leaving long stretches of unoccupied DNA, thus again confirming the mobilization of nucleosomes in vitro. Furthermore, both studies showed a potential loss of DNA constrained by nucleosomes remodeled by SWI/ SNF, most likely at the entry/exit positions. Early this decade, the S. cerevisiae RSC and SWI/SNF complexes were imaged using 3D electron microscopy reconstructions (Asturias et al., 2002; Smith et al., 2003). Both enzymes show central cavities of approximately the same size and dimensions of a single nucleosome. The yeast SWI/SNF structure was also of the same approximate dimensions as that imaged by both Bazett-Jones and Schnitzler (Smith et al., 2003). How do these large megadalton complexes bind to nucleosomes? One of the earliest observations for SWI/SNF interaction with DNA showed a high aYnity for four-way junction (4WJ) DNA similar to that displayed by HMG-box domain proteins (Quinn et al., 1996). This 4WJ binding aYnity suggests that SWI/SNF and related complexes may bind to the entry and exit segments of the nucleosome. Site-specific DNA photoaYnity labeling has also been used to look at the interactions between both the ySWI/SNF and RSC complexes with nucleosome (Sengupta et al., 1999, 2001). These studies show that multiple subunits are in close contact with nucleosomal DNA, and that there does not appear to be a preferential binding site at the nucleosomal edge. In the ISWI complexes, binding to substrate seems to be mediated by the interface between linker DNA regions and the actual nucleosome core, since ISWI-containing complexes act only on substrates that have at least 20 bp of DNA adjacent to the nucleosome (Kagalwala et al., 2004). This binding requires other subunit(s) in the ISWI complex; for instance, Acf1p is necessary for high-aYnity nucleosome sliding by ISWI in both ACF and CHRAC complexes (Eberharter et al., 2001).

D. How is ATP Hydrolysis Coupled to the Generation of Remodeled Chromatin? Once a remodeling enzyme has bound to a nucleosomal substrate, what does DNA translocation and/or torsion generation do? Two major models have been debated for the last couple of years. The first model was the ‘‘twist diVusion’’ model discussed by van Holde and Yager in 1985 and readdressed

132

Smith and Peterson

recently (van Holde and Yager, 2003). The twist diVusion model theorizes that migration of DNA around the nucleosome is propagated by the introduction of small twist defects that cause underwinding of the DNA helix, which are then diVused around the nucleosome (see Fig. 4). If the defect collapses back on itself, the result would be no movement. On the other

Figure 4 Generation of novel nucleosomal structures during octamer mobilization. This figure illustrates some of the phenotypes of cis and trans octamer mobilization seen in vitro in various chromatin remodeling assays. Nucleosomes (a) can move on stretches of DNA, resulting in translocation (b), which in turn can result in the exposed end of the DNA fragment making novel contacts with the same nucleosome (c) or another remodeled nucleosome, thus creating a dinucleosome (d). Nucleosomes can also move via octamer transfer in trans. The histone octamer can become disassociated from the first template and then become incorporated into a second nucleosomal array (e).


133

hand, if it is propagated forward, this model allows for small slipping steps to occur, where the DNA strand is continuously pumped across the face of the nucleosome (van Holde and Yager, 2003). This model is consistent with the ability of chromatin remodeling enzymes to generate superhelical torsion (Gavin et al., 2001; Havas et al., 2000). There is evidence which puts this model into question, however, as ISWIand SWI2/SNF2-containing complexes can still mobilize nucleosomes on DNA substrates containing nicks, hairpins, or gaps (Aoyagi and Hayes, 2002; Langst and Becker, 2001a; Saha et al., 2002). These experiments argue against a simple twist diVusion model since introduction of single base pair nicks and/or addition of bulky DNA branches could prevent the propagation of twist that initiates outside the nucleosome. Nicks in the DNA might dissipate the accumulation of torsional stress while the branched DNA might interfere with the actual rotation of the DNA relative to the nucleosome. The second, related model is the reptation or bulge migration model. This model suggests that a wave of DNA is released from the histone octamer and is propagated along the surface of the nucleosome, allowing accessibility to DNA binding factors with or without generating movement. The best evidence for the creation of loops during remodeling comes from experiments conducted using cross-linked nucleosomes (Aoyagi et al., 2002). In these studies, the H2B histone was first cross-linked to DNA, and the ability of these mononucleosomes to be remodeled was scored by nuclease accessibility. Interestingly, hSWI/SNF could still enhance DNase I accessibility of nucleosomal DNA even in the absence of nucleosome movement. However, increased accessibility to restriction enzymes was lost. Thus, in the context of the bulge migration model, it appears that hSWI/SNF might create loops accessible to some factors (DNase), but other factors require actual movement of the octamer. Alternatively, this data might also be consistent simply with changes in rotational positioning of the DNA helix that would result in changes in the DNase I cleavage pattern of DNA. Surprisingly, remodeling with recombinant ISWI actually seems to be stimulated by nicks in the DNA at the entry/exit sites (Langst and Becker, 2001a). Furthermore, another study by the Kingston group has shown diVerences between the remodeling intermediates for ISWI family members (SNF2h) and SWI2/SNF2 (BRG1 and human SWI/SNF) family members (Fan et al., 2003). From these experiments, it has been suggested that BRG1 and hSWI/SNF may allow access to DNA occluded in the nucleosome without drastically mobilizing the nucleosome, while SNF2h seems to preferentially move nucleosomes without creation of stable remodeling intermediates. Van Holde and Yager have argued that both the reptation model and the results of remodeling on nicked substrates can be explained in the context of the twist defect model (van Holde and Yager, 2003). They argue

134

Smith and Peterson

that DNA writhing and bulging could be the summation of a number of twist defects in a constrained system. That is, an accumulation of twist defects could lead to the generation of a whole writhe of DNA becoming dissociated from the nucleosome if the duplex is not allowed to freely translocate relative to the histone octamer. They also argue that the nicked DNA substrate experiments do not rule out the possibility of the accumulation of twist defects. Base-stacking of the DNA could be maintained in the context of the nucleosome since the histone octamer could stabilize the DNA duplex and any incorporated nicks. Van Holde and Yager argue that nicks might even aid the torsional process, thus leading to the increased rate of remodeling seen in the nicked substrate experiments (Langst and Becker, 2001a). Branches and hairpins in the DNA duplex might also be remodeled by a twist diVusion mechanism if accumulation of twist defects leads to a writhe of DNA becoming disassociated from the nucleosome surface, accommodating the bulky DNA formation (van Holde and Yager, 2003). Alternatively, if the step-size of the remodeling reaction is small enough, nicks and bulky DNA formations might not even be a factor. That is, if twist defects were rapidly created and diVused, resulting in slippage of the DNA duplex rather than permanent rotation of the helix, it could be possible that these DNA defects might not aVect the ability to translocate DNA.

E. Nucleosome Accessibility and Mobilization by Chromatin Remodeling Enzymes One common feature of all ATP-dependent chromatin remodeling enzymes is the ability to enhance the accessibility of nucleosomal DNA to nucleases and/or transcription factors. In most cases, this activity of remodeling enzymes can be explained by the ATP-dependent movement of nucleosomes in cis along a DNA fragment (Fan et al., 2003; Langst and Becker, 2001b; Langst et al., 1999; Logie and Peterson, 1997; Schnitzler et al., 1998; Varga-Weisz et al., 1997; Whitehouse et al., 1999). Various chromatin remodeling enzymes, from diVerent subfamilies, can mobilize mononucleosomes on short stretches of DNA (146–208bp), either to the end (SWI2/SNF2 family and recombinant ISWI) or to the center (CHRAC and dMi-2) of a DNA template (Brehm et al., 2000; Flaus and Owen-Hughes, 2003; Guschin et al., 2000b; Kassabov et al., 2002, 2003; Whitehouse et al., 1999). Some of the ISWI family of remodeling enzyme seem to have a preference for shifting mononucleosomes to a central position on the DNA template, while others seem to randomize nucleosome positioning (Fan et al., 2003; Hamiche et al., 2001; Langst and Becker, 2001b). The mechanistic reason for the diVerent directionality of nucleosome movements is still unknown. Flaus and Owen-Hughes used


135

mononucleosome constructs containing additional DNA extensions flanking the nucleosome to investigate the ability of recombinant ISWI and SWI/ SNF class chromatin remodeling enzymes (ySWI/SNF and RSC) to mobilize nucleosomes. They showed that mobilization by ISWI correlated with the thermally preferred positioning of nucleosomes on the DNA template. In contrast, the SWI/SNF and RSC complexes were shown to move nucleosomes to the ends of the DNA fragment, away from the thermally preferred position (Flaus and Owen-Hughes, 2003). In fact, the SWI2/SNF2 family of complexes could shift the nucleosome oV the end of the DNA fragment, leaving the dyad axis of the nucleosome only 22 base pairs from one end. The ability of SWI/SNF to mobilize octamers oV the ends of DNA fragments may explain several novel features of SWI/SNF remodeled nucleosomes. First, SWI2/SNF2 complexes have been found to generate dinucleosome structures during remodeling of mononucleosomes (Lorch et al., 1998, 2001; Phelan et al., 2000; Schnitzler et al., 1998). In addition, a few studies have shown that chromatin remodeling enzymes can shift nucleosomes in trans from one template to another (Lorch et al., 1999; Phelan et al., 2000). One possibility is that as the nucleosome is pushed oV the end of a fragment of DNA, it is then able to be ‘‘transferred’’ to another fragment of DNA or captured by another remodeled nucleosome. Figure 5 illustrates a number of possible outcomes for remodeling on mononucleosome templates. A study of the Saccharomyces cerevisiae Isw2 complex has provided evidence that nucleosome mobilization occurs in vivo. In this study, the researchers used a galactose-inducible allele of ISW2 to study changes in chromatin structure at the promoters of a pair of test genes. The data suggested that the changes were unidirectional and localized to only a few nucleosomes (Fazzio and Tsukiyama, 2003). However, since transcriptional repression was not measured in this study, it is still not clear if nucleosome mobilization directly correlates with the biological function of Isw2.

F. Disruption of Nucleosome Structure: Moving Dimers Around An important and debated question in the field of chromatin remodeling is whether the histone octamer is disrupted during chromatin remodeling. Ten years ago, it was put forth that remodeling by SWI/SNF and other chromatin remodeling enzymes might involve dissociation of the H2A–H2B dimers and/or alteration of the core histone folds (Cote et al., 1994; Peterson and Tamkun, 1995). Histone–histone cross-linking studies have since shown that it is not absolutely necessary to disrupt nucleosome structure in order to allow restriction enzyme access and nucleosome mobility (Boyer et al., 2000b). However, several recent results suggest that disruption of the H2A–H2B dimer can be catalyzed by some chromatin remodeling enzymes.

136

Smith and Peterson

Figure 5 Torsional model for nucleosomal DNA–histone contact disruption. Generation of helical torsion on nucleosomal DNA either by twisting or pushing the DNA by a chromatin remodeling enzyme results in the generation of a twist defect. This twist defect is propagated along the duplex, which could result in either a wave or smaller slippage of DNA moving along the surface of the histone octamer. Relief of the twist defect in the forward direction could result in the movement of the DNA relative to the octamer. (a) In a simple slipping model the DNA would appear to twist along the surface of the nucleosome, resulting in conversion of nucleosome I to nucleosome III without a detectable bulge. A writhing mechanism would result in conversion of nucleosome I first to II, then to III. A writhing mechanism also could result in intermediates being trapped at the nucleosome II stage. (b) Top-down view to illustrate diVerences between slippage and writhing models.

Bruno and colleagues tested a number of ATP-dependent chromatin remodeling enzymes (ySWI/SNF, RSC, dISWI, ISw1a, and ISw1b) for the ability to exchange H2A–H2B dimers in vitro. Using fluorescently labeled histones, they measured the ability of various remodeling enzymes to catalyze the exchange of histones from one chromatin substrate to another in an ATPdependent fashion. SWI/SNF, RSC, and ISw1b were able, to some degree,


137

to transfer H2A–H2B dimers from a mononucleosome substrate to H3–H4 tetramers (Bruno et al., 2003). Dimers could also be exchanged from a circular nucleosomal array, but this reaction seems less eYcient. Although it is not yet clear whether dimer exchange is relevant to the in vivo function of most remodeling enzymes, several studies indicate that members of the INO80 family do indeed catalyze dimer exchange in vivo. Yeast contain two members of the INO80 family, Ino80 and SWR1. Each of these ATPases is a subunit of large multi-subunit complexes. Unlike SWI/ SNF and ISWI complexes, both Ino80 and SWR1 complexes contain histones as stoichiometric subunits (Mizuguchi et al., 2004). Three groups have published findings that the H2AZ variant histone copurifies with SWR1 complex and that this complex is required for the proper recruitment of Htz1 (yeast H2AZ) into chromatin in vivo (Kobor et al., 2004; Krogan et al., 2003; Mizuguchi et al., 2004). Remarkably, the SWR1 complex is able to swap H2AZ–H2B dimers for H2A–H2B dimers incorporated into in vitro nucleosomal arrays (Mizuguchi et al., 2004). In vivo deposition of Htz1 is also SWR1 complex-dependent at specific heterochromatic regions. It has been proposed that these remodeling complexes play a role in establishing boundaries for the spreading of heterochromatin (Owen-Hughes and Bruno, 2004). In this case, the incorporation of H2AZ–H2B dimers might prevent the binding of SIR proteins at the boundary of heterochromatin, which are involved in the maintenance of silencing at telomeres. It is not yet clear how dimer-exchange relates to the other outcomes of chromatin remodeling (i.e., generation of torsion, DNA translocation, etc.).

G. Determining How Remodeling Works In Vivo The biggest questions still unanswered are the questions that are the most interesting: what do chromatin remodeling enzymes do in vivo? As has been discussed, the SWR1 complex does appear to exchange histone H2AZ–H2B dimers in vivo and this seems to be functionally important. But what role does ATP-dependent nucleosome mobilization or formation of DNA loops play in vivo? Most in vivo studies of chromatin remodeling have used restriction enzymes or other nucleases, like MNase, to probe chromatin structure. However, these reagents only show that DNA accessibility has been altered; the mechanism is not clear. Furthermore, it is important to note that none of the nuclease reagents is able to detect changes in higher-order chromatin folding. Thus, a potential role for remodeling enzymes cannot yet be assayed in vivo. Thus, what seems to be lacking in the in vivo analysis is actually development of better methodologies for studying chromatin in cells. Even with the application of nuclease digestion methods, a number of labs have found that chromatin remodeling appears to be distinct at diVerent

138

Smith and Peterson

gene loci. At some genes, chromatin remodeling appears to involve the mobilization or removal of multiple nucleosomes near the promoter, as illustrated earlier by the yeast POT1 and REC104 promoter regions after remodeling by Isw2 (Fazzio and Tsukiyama, 2003). In contrast, at the mouse mammary tumor virus (MMTV) promoter, mobilization or removal of nucleosomes is not necessary for chromatin remodeling. Indeed, in this case, it appears that a nuclear hormone receptor (GR or PR) recruits a SWI/SNFlike enzyme, which then enhances the accessibility of a promoter-bound nucleosome without induced nucleosome loss (Truss et al., 1995). At the MMTV promoter, a nucleosome, referred to as nucleosome B, spans an important DNA sequence for hormone receptor binding sites known as the hormone responsive region (HRR) (Richard-Foy and Hager, 1987). Changes in DNase I sensitivity at nucleosome B suggest that chromatin remodeling is necessary for remodeling at this particular nucleosome to allow nuclear receptors and other factors to occupy the HRR (Truss et al., 1995). More recently, a requirement for an ATP-dependent chromatin remodeling activity has been shown on MMTV promoters in Drosophila extracts (Di Croce et al., 1999). While these and other studies begin to shed light on the diVerences in chromatin remodeling at diVerent genomic loci, they only begin to address how chromatin remodeling works in vivo. In the case of MMTV, this may be a candidate for ATP-dependent DNA loop formation or perhaps a role for ATP-dependent dimer loss.

H. Remodeling at the Fiber Level In vivo chromatin exists as large 100 to 400 nm fibers that are not only highly condensed but contain many nonhistone proteins as well. Indeed, incorporation of a linker histone into a nucleosomal array substrate blocks the in vitro remodeling activities of ATP-dependent remodeling enzymes (Hill and Imbalzano, 2000). In vivo, however, H1 is present at nearly every nucleosome (Hansen, 2002). How do these enzymes modulate chromatin at the fiber level? Are loops of DNA removed from the surface of the nucleosome in vivo? Are minor histone–DNA contact disruptions enough to facilitate the processivity of large enzymes like RNAP II holoenzyme? Some chromatin remodeling enzymes might be acting at the fiber level as well as the nucleosomal level. As mentioned earlier, one of the first links between SWI/SNF and chromatin came from the genetic observations that Sin mutants restore transcription in Swi and Snf mutants. One prediction of these genetic studies was that Sin chromatin might mimic the SWI/ SNF remodeled state (Wechser et al., 1997). One in vitro study demonstrated that Sin versions of histone H4 eliminate the cation-dependent intramolecular folding of nucleosomal arrays (Horn et al., 2002). Thus, these data


139

suggest that chromatin remodeling may not only act at the nucleosomal array level but might also aVect chromatin at the fiber level.

V. Concluding Remarks The study of chromatin biology has come a long way but there are still many unanswered questions. We now have a better feeling for the dynamics involved at the simplest levels of chromatin structure (nucleosomal arrays) and how chromatin remodeling enzymes recognize and act on these substrates. The work of numerous groups has attempted to elucidate the function of chromatin-modifying enzymes and their eVect on chromatin. We also have a basic understanding of the function of the ATPase subunits of these enzymes but we are just beginning to scratch the surface of understanding why some of these complexes have numerous subunits. In the future, we need to better understand how these enzymes behave in vivo. We are beginning to understand how ATP-dependent enzymes behave on short chromatin fragments but chromatids are more complex than those modeled in the laboratory. It is at the level of the compact chromatin fiber that we believe the true diVerences among diVerent enzymes will be revealed.

Acknowledgments The authors thank Peter Horn and Marc-Andre´ Laniel for critical reading of this manuscript. Research in the Peterson laboratory is funded by grants from the NIH and NCI.

References Ahmad, K., and HenikoV, S. (2001). Centromeres are specialized replication domains in heterochromatin. J. Cell Biol. 153, 101–110. Ahmad, K., and HenikoV, S. (2002). Histone H3 variants specify modes of chromatin assembly. Proc. Natl. Acad. Sci. USA 99(Suppl. 4), 16477–16484. Alexeev, A., Mazin, A., and Kowalczykowski, S. C. (2003). Rad54 protein possesses chromatinremodeling activity stimulated by the Rad51-ssDNA nucleoprotein filament. Nat. Struct. Biol. 10, 182–186. Alexiadis, V., and Kadonaga, J. T. (2002). Strand pairing by Rad54 and Rad51 is enhanced by chromatin. Genes Dev. 16, 2767–2771. Anderson, J. D., Thastrom, A., and Widom, J. (2002). Spontaneous access of proteins to buried nucleosomal DNA target sites occurs via a mechanism that is distinct from nucleosome translocation. Mol. Cell. Biol. 22, 7147–7157. Aoyagi, S., and Hayes, J. J. (2002). hSWI/SNF-catalyzed nucleosome sliding does not occur solely via a twist-diVusion mechanism. Mol. Cell. Biol. 22, 7484–7490.

140

Smith and Peterson

Aoyagi, S., Narlikar, G., Zheng, C., Sif, S., Kingston, R. E., and Hayes, J. J. (2002). Nucleosome remodeling by the human SWI/SNF complex requires transient global disruption of histone-DNA interactions. Mol. Cell. Biol. 22, 3653–3662. Aravind, L., and Landsman, D. (1998). AT-hook motifs identified in a wide variety of DNAbinding proteins. Nucleic Acids Res. 26, 4413–4421. Asturias, F. J., Chung, W. H., Kornberg, R. D., and Lorch, Y. (2002). Structural analysis of the RSC chromatin-remodeling complex. Proc. Natl. Acad. Sci. USA 99, 13477–13480. Bazett-Jones, D. P., Cote, J., Landel, C. C., Peterson, C. L., and Workman, J. L. (1999). The SWI/SNF complex creates loop domains in DNA and polynucleosome arrays and can disrupt DNA-histone contacts within these domains. Mol. Cell. Biol. 19, 1470–1478. Becker, P. B., and Horz, W. (2002). ATP-dependent nucleosome remodeling. Annu. Rev. Biochem. 71, 247–273. Bednar, J., Horowitz, R. A., Dubochet, J., and Woodcock, C. L. (1995). Chromatin conformation and salt-induced compaction: Three-dimensional structural information from cryoelectron microscopy. J. Cell Biol. 131, 1365–1376. Bednar, J., Horowitz, R. A., Grigoryev, S. A., Carruthers, L. M., Hansen, J. C., Koster, A. J., and Woodcock, C. L. (1998). Nucleosomes, linker DNA, and linker histone form a unique structural motif that directs the higher-order folding and compaction of chromatin. Proc. Natl. Acad. Sci. USA 95, 14173–14178. Bouazoune, K., Mitterweger, A., Langst, G., Imhof, A., Akhtar, A., Becker, P. B., and Brehm, A. (2002). The dMi-2 chromodomains are DNA binding modules important for ATP-dependent nucleosome mobilization. EMBO J. 21, 2430–2440. Boyer, L. A., Latek, R. R., and Peterson, C. L. (2004). The SANT domain: A unique histone-tail-binding module? Nat. Rev. Mol. Cell. Biol. 5, 158–163. Boyer, L. A., Logie, C., Bonte, E., Becker, P. B., Wade, P. A., WolVe, A. P., Wu, C., Imbalzano, A. N., and Peterson, C. L. (2000a). Functional delineation of three groups of the ATP-dependent family of chromatin remodeling enzymes. J. Biol. Chem. 275, 18864–18870. Boyer, L. A., Shao, X., Ebright, R. H., and Peterson, C. L. (2000b). Roles of the histone H2AH2B dimers and the (H3/H4)2 tetramer in nucleosome remodeling by the SWI-SNF complex. J. Biol. Chem. 275, 11545–11552. Bozhenok, L., Wade, P. A., and Varga-Weisz, P. (2002). WSTF-ISWI chromatin remodeling complex targets heterochromatic replication foci. EMBO J. 21, 2231–2241. Brehm, A., Langst, G., Kehle, J., Clapier, C. R., Imhof, A., Eberharter, A., Muller, J., and Becker, P. B. (2000). dMi-2 and ISWI chromatin remodelling factors have distinct nucleosome binding and mobilization properties. EMBO J. 19, 4332–4341. Brower-Toland, B. D., Smith, C. L., Yeh, R. C., Lis, J. T., Peterson, C. L., and Wang, M. D. (2002). Mechanical disruption of individual nucleosomes reveals a reversible multistage release of DNA. Proc. Natl. Acad. Sci. USA 99, 1960–1965. Bruno, M., Flaus, A., Stockdale, C., Rencurel, C., Ferreira, H., and Owen-Hughes, T. (2003). Histone H2A/H2B dimer exchange by ATP-dependent chromatin remodeling activities. Mol. Cell 12, 1599–1606. Cairns, B. R., Erdjument-Bromage, H., Tempst, P., Winston, F., and Kornberg, R. D. (1998). Two actin-related proteins are shared functional components of the chromatin-remodeling complexes RSC and SWI/SNF. Mol. Cell 2, 639–651. Cairns, B. R., Henry, N. L., and Kornberg, R. D. (1996a). TFG/TAF30/ANC1, a component of the yeast SWI/SNF complex that is similar to the leukemogenic proteins ENL and AF-9. Mol. Cell. Biol. 16, 3308–3316. Cairns, B. R., Levinson, R. S., Yamamoto, K. R., and Kornberg, R. D. (1996b). Essential role of Swp73p in the function of yeast Swi/Snf complex. Genes Dev. 10, 2131–2144.


141

Cairns, B. R., Lorch, Y., Li, Y., Zhang, M., Lacomis, L., Erdjument-Bromage, H., Tempst, P., Du, J., Laurent, B., and Kornberg, R. D. (1996c). RSC, an essential, abundant chromatinremodeling complex. Cell 87, 1249–1260. Carruthers, L. M., Bednar, J., Woodcock, C. L., and Hansen, J. C. (1998). Linker histones stabilize the intrinsic salt-dependent folding of nucleosomal arrays: Mechanistic ramifications for higher-order chromatin folding. Biochemistry 37, 14776–14787. Carruthers, L. M., and Hansen, J. C. (2000). The core histone N termini function independently of linker histones during chromatin condensation. J. Biol. Chem. 275, 37285–37290. Caruthers, J. M., and McKay, D. B. (2002). Helicase structure and mechanism. Curr. Opin. Struct. Biol. 12, 123–133. Citterio, E., Van Den Boom, V., Schnitzler, G., Kanaar, R., Bonte, E., Kingston, R. E., Hoeijmakers, J. H., and Vermeulen, W. (2000). ATP-dependent chromatin remodeling by the Cockayne syndrome B DNA repair-transcription-coupling factor. Mol. Cell. Biol. 20, 7643–7653. Collins, N., Poot, R. A., Kukimoto, I., Garcia-Jimenez, C., Dellaire, G., and Varga-Weisz, P. D. (2002). An ACF1-ISWI chromatin-remodeling complex is required for DNA replication through heterochromatin. Nat. Genet. 32, 627–632. Collins, R. T., and Treisman, J. E. (2000). Osa-containing Brahma chromatin remodeling complexes are required for the repression of wingless target genes. Genes Dev. 14, 3140–3152. Corona, D. F., Langst, G., Clapier, C. R., Bonte, E. J., Ferrari, S., Tamkun, J. W., and Becker, P. B. (1999). ISWI is an ATP-dependent nucleosome remodeling factor. Mol. Cell 3, 239–245. Corona, D. F., and Tamkun, J. W. (2004). Multiple roles for ISWI in transcription, chromosome organization and DNA replication. Biochim. Biophys. Acta 1677, 113–119. Cote, J., Peterson, C. L., and Workman, J. L. (1998). Perturbation of nucleosome core structure by the SWI/SNF complex persists after its detachment, enhancing subsequent transcription factor binding. Proc. Natl. Acad. Sci. USA 95, 4947–4952. Cote, J., Quinn, J., Workman, J. L., and Peterson, C. L. (1994). Stimulation of GAL4 derivative binding to nucleosomal DNA by the yeast SWI/SNF complex. Science 265, 53–60. de la Serna, I. L., Carlson, K. A., and Imbalzano, A. N. (2001). Mammalian SWI/SNF complexes promote MyoD-mediated muscle diVerentiation. Nat. Genet. 27, 187–190. Decristofaro, M. F., Betz, B. L., Rorie, C. J., Reisman, D. N., Wang, W., and Weissman, B. E. (2001). Characterization of SWI/SNF protein expression in human breast cancer cell lines and other malignancies. J. Cell Physiol. 186, 136–145. Deuring, R., Fanti, L., Armstrong, J. A., Sarte, M., Papoulas, O., Prestel, M., Daubresse, G., Verardo, M., Moseley, S. L., Berloco, M., Tsukiyama, T., Wu, C., Pimpinelli, S., and Tamkun, J. W. (2000). The ISWI chromatin-remodeling protein is required for gene expression and the maintenance of higher order chromatin structure in vivo. Molec. Cell 5, 355–365. Di Croce, L., Koop, R., Venditti, P., Westphal, H. M., Nightingale, K. P., Corona, D. F., Becker, P. B., and Beato, M. (1999). Two-step synergism between the progesterone receptor and the DNA-binding domain of nuclear factor 1 on MMTV minichromosomes. Mol. Cell 4, 45–54. Du, J., Nasir, I., Benton, B. K., Kladde, M. P., and Laurent, B. C. (1998). Sth1p, a Saccharomyces cerevisiae Snf2p/Swi2p homolog, is an essential ATPase in RSC and diVers from Snf/Swi in its interactions with histones and chromatin-associated proteins. Genetics 150, 987–1005. Eberharter, A., Ferrari, S., Langst, G., Straub, T., Imhof, A., Varga-Weisz, P., Wilm, M., and Becker, P. B. (2001). Acf1, the largest subunit of CHRAC, regulates ISWI-induced nucleosome remodelling. EMBO J. 20, 3781–3788.

142

Smith and Peterson

Eisen, J. A., Sweder, K. S., and Hanawalt, P. C. (1995). Evolution of the SNF2 family of proteins: Subfamilies with distinct sequences and functions. Nucleic Acids Res. 23, 2715–2723. Elfring, L. K., Daniel, C., Papoulas, O., Deuring, R., Sarte, M., Moseley, S., Beek, S. J., Waldrip, W. R., Daubresse, G., DePace, A., Kennison, J. A., and Tamkun, J. W. (1998). Genetic analysis of brahma: The Drosophila homolog of the yeast chromatin remodeling factor SWI2/SNF2. Genetics 148, 251–265. Fan, H. Y., He, X., Kingston, R. E., and Narlikar, G. J. (2003). Distinct strategies to make nucleosomal DNA accessible. Mol. Cell 11, 1311–1322. Fan, J. Y., Gordon, F., Luger, K., Hansen, J. C., and Tremethick, D. J. (2002). The essential histone variant H2A.Z regulates the equilibrium between diVerent chromatin conformational states. Nat. Struct. Biol. 9, 172–176. Fazzio, T. G., and Tsukiyama, T. (2003). Chromatin remodeling in vivo: Evidence for a nucleosome sliding mechanism. Mol. Cell 12, 1333–1340. Firman, K., and Szczelkun, M. D. (2000). Measuring motion on DNA by the type I restriction endonuclease EcoR124I using triplex displacement. EMBO J. 19, 2094–2102. Fischle, W., Wang, Y., and Allis, C. D. (2003). Histone and chromatin cross-talk. Curr. Opin. Cell Biol. 15, 172–183. Flaus, A., and Owen-Hughes, T. (2003). Dynamic properties of nucleosomes during thermal and ATP-driven mobilization. Mol. Cell. Biol. 23, 7767–7779. Garcia-Ramirez, M., Dong, F., and Ausio, J. (1992). Role of the histone tails in the folding of oligonucleosomes depleted of histone H1. J. Biol. Chem. 267, 19587–19595. Garrick, D., Samara, V., McDowell, T. L., Smith, A. J., Dobbie, L., Higgs, D. R., and Gibbons, R. J. (2004). A conserved truncated isoform of the ATR-X syndrome protein lacking the SWI/SNF-homology domain. Gene 326, 23–34. Gavin, I., Horn, P. J., and Peterson, C. L. (2001). SWI/SNF chromatin remodeling requires changes in DNA topology. Mol. Cell 7, 97–104. Gibbons, R. J., Picketts, D. J., Villard, L., and Higgs, D. R. (1995). Mutations in a putative global transcriptional regulator cause X-linked mental retardation with alpha-thalassemia (ATR-X syndrome). Cell 80, 837–845. Goldmark, J. P., Fazzio, T. G., Estep, P. W., Church, G. M., and Tsukiyama, T. (2000). The Isw2 chromatin remodeling complex represses early meiotic genes upon recruitment by Ume6p. Cell 103, 423–433. Goodwin, G. H., and Nicolas, R. H. (2001). The BAH domain, polybromo and the RSC chromatin remodeling complex. Gene. 268, 1–7. Grune, T., Brzeski, J., Eberharter, A., Clapier, C. R., Corona, D. F., Becker, P. B., and Muller, C. W. (2003). Crystal structure and functional analysis of a nucleosome recognition module of the remodeling factor ISWI. Mol. Cell 12, 449–460. Guidi, C. J., Sands, A. T., Zambrowicz, B. P., Turner, T. K., Demers, D. A., Webster, W., Smith, T. W., Imbalzano, A. N., and Jones, S. N. (2001). Disruption of Ini1 leads to periimplantation lethality and tumorigenesis in mice. Mol. Cell. Biol. 21, 3598–3603. Guschin, D., Geiman, T. M., Kikyo, N., Tremethick, D. J., WolVe, A. P., and Wade, P. A. (2000a). Multiple ISWI ATPase complexes from xenopus laevis. FUNCTIONAL CONSERVATION OF AN ACF/CHRAC HOMOLOG [In Process Citation]. J. Biol. Chem. 275, 35248–35255. Guschin, D., Wade, P. A., Kikyo, N., and WolVe, A. P. (2000b). ATP-dependent histone octamer mobilization and histone deacetylation mediated by the Mi-2 chromatin remodeling complex. Biochemistry 39, 5238–5245. Hake, S. B., Xiao, A., and Allis, C. D. (2004). Linking the epigenetic ‘language’ of covalent histone modifications to cancer. Br. J. Cancer 90, 761–769.


143

Hamiche, A., Kang, J. G., Dennis, C., Xiao, H., and Wu, C. (2001). Histone tails modulate nucleosome mobility and regulate ATP-dependent nucleosome sliding by NURF. Proc. Natl. Acad. Sci. USA 98, 14316–14321. Hansen, J. C. (2002). Conformational dynamics of the chromatin fiber in solution: Determinants, mechanisms, and functions. Annu. Rev. Biophys. Biomol. Struct. 31, 361–392. Hansen, J. C., Ausio, J., Stanik, V. H., and van Holde, K. E. (1989). Homogeneous reconstituted oligonucleosomes, evidence for salt-dependent folding in the absence of histone H1. Biochemistry 28, 9129–9136. Havas, K., Flaus, A., Phelan, M., Kingston, R., Wade, P. A., Lilley, D. M. J., and OwenHughes, T. (2000). Generation of superhelical torsion by ATP-dependent chromatin remodeling activities. Cell 103, 1133–1142. Hill, D. A., and Imbalzano, A. N. (2000). Human SWI/SNF Nucleosome Remodeling Activity is Partially Inhibited by Linker Histone H1. Biochemistry 39, 11649–11656. Holstege, F. C., Jennings, E. G., Wyrick, J. J., Lee, T. I., Hengartner, C. J., Green, M. R., Golub, T. R., Lander, E. S., and Young, R. A. (1998). Dissecting the regulatory circuitry of a eukaryotic genome. Cell 95, 717–728. Horn, P. J., Crowley, K. A., Carruthers, L. M., Hansen, J. C., and Peterson, C. L. (2002). The SIN domain of the histone octamer is essential for intramolecular folding of nucleosomal arrays. Nat. Struct. Biol. 9, 167–171. Horn, P. J., and Peterson, C. L. (2002). Molecular biology. Chromatin higher order folding– wrapping up transcription. Science 297, 1824–1827. Huang, C., Sloan, E. A., and Boerkoel, C. F. (2003). Chromatin remodeling and human disease. Curr. Opin. Genet. Dev. 13, 246–252. Huang, M., Qian, F., Hu, Y., Ang, C., Li, Z., and Wen, Z. (2002). Chromatin-remodelling factor BRG1 selectively activates a subset of interferon-alpha-inducible genes. Nat. Cell Biol. 4, 774–781. Ito, T., Bulger, M., Pazin, M. J., Kobayashi, R., and Kadonaga, J. T. (1997). ACF, an ISWIcontaining and ATP-utilizing chromatin assembly and remodeling factor. Cell 90, 145–155. JaskelioV, M., Van Komen, S., Krebs, J. E., Sung, P., and Peterson, C. L. (2003). Rad54p is a chromatin remodeling enzyme required for heteroduplex DNA joint formation with chromatin. J. Biol. Chem. 278, 9212–9218. Jenuwein, T., and Allis, C. D. (2001). Translating the histone code. Science 293, 1074–1080. Kagalwala, M. N., Glaus, B. J., Dang, W., Zofall, M., and Bartholomew, B. (2004). Topography of the ISW2-nucleosome complex: Insights into nucleosome spacing and chromatin remodeling. EMBO J. 23, 2092–2104. Kassabov, S. R., Henry, N. M., Zofall, M., Tsukiyama, T., and Bartholomew, B. (2002). Highresolution mapping of changes in histone-DNA contacts of nucleosomes remodeled by ISW2. Mol. Cell. Biol. 22, 7524–7534. Kassabov, S. R., Zhang, B., Persinger, J., and Bartholomew, B. (2003). SWI/SNF unwraps, slides, and rewraps the nucleosome. Mol. Cell 11, 391–403. Kehle, J., Beuchle, D., Treuheit, S., Christen, B., Kennison, J. A., Bienz, M., and Muller, J. (1998). dMi-2, a hunchback-interacting protein that functions in polycomb repression. Science 282, 1897–1900. Kennison, J. A., and Tamkun, J. W. (1988). Dosage-dependent modifiers of polycomb and antennapedia mutations in Drosophila. Proc. Natl. Acad. Sci. USA 85, 8136–8140. Khattak, S., Lee, B. R., Cho, S. H., Ahnn, J., and Spoerel, N. A. (2002). Genetic characterization of Drosophila Mi-2 ATPase. Gene 293, 107–114. Khavari, P. A., Peterson, C. L., Tamkun, J. W., Mendel, D. B., and Crabtree, G. R. (1993). BRG1 contains a conserved domain of the SWI2/SNF2 family necessary for normal mitotic growth and transcription. Nature 366, 170–174.

144

Smith and Peterson

Khorasanizadeh, S. (2004). The nucleosome: From genomic organization to genomic regulation. Cell 116, 259–272. Kikyo, N., Wade, P. A., Guschin, D., Ge, H., and WolVe, A. P. (2000). Active remodeling of somatic nuclei in egg cytoplasm by the nucleosomal ATPase ISWI. Science 289, 2360–2362. Kim, J., Sif, S., Jones, B., Jackson, A., Koipally, J., Heller, E., Winandy, S., Viel, A., Sawyer, A., Ikeda, T., Kingston, R., and Georgopoulos, K. (1999). Ikaros DNA-binding proteins direct formation of chromatin remodeling complexes in lymphocytes. Immunity 10, 345–355. Klochendler-Yeivin, A., Fiette, L., Barra, J., Muchardt, C., Babinet, C., and Yaniv, M. (2000). The murine SNF5/INI1 chromatin remodeling factor is essential for embryonic development and tumor suppression. EMBO Rep. 1, 500. Kobor, M. S., Venkatasubrahmanyam, S., Meneghini, M. D., Gin, J. W., Jennings, J. L., Link, A. J., Madhani, H. D., and Rine, J. (2004). A Protein Complex Containing the Conserved Swi2/Snf2-Related ATPase Swr1p Deposits Histone Variant H2A.Z into Euchromatin. PLoS. Biol. 2, E131. Koipally, J., Heller, E. J., Seavitt, J. R., and Georgopoulos, K. (2002). Unconventional potentiation of gene expression by Ikaros. J. Biol. Chem. 277, 13007–13015. Krebs, J. E., Fry, C. J., Samuels, M., and Peterson, C. L. (2000). Global role for chromatin remodeling enzymes in mitotic gene expression. Cell 102, 587–598. Krebs, J. E., and Peterson, C. L. (2000). Understanding ‘‘active’’ chromatin: A historical perspective of chromatin remodeling. Crit. Rev. Eukaryot. Gene Expr. 10, 1–12. Krogan, N. J., Keogh, M. C., Datta, N., Sawa, C., Ryan, O. W., Ding, H., Haw, R. A., Pootoolal, J., Tong, A., Canadien, V., Richards, D. P., Wu, X., Emili, A., Hughes, T. R., Buratowski, S., and Greenblatt, J. F. (2003). A Snf2 family ATPase complex required for recruitment of the histone H2A variant Htz1. Mol. Cell 12, 1565–1576. Kruger, W., and Herskowitz, I. (1991). A negative regulator of HO transcription, SIN1 (SPT2), is a nonspecific DNA-binding protein related to HMG1. Mol. Cell. Biol. 11, 4135–4146. Ladurner, A. G. (2003). Inactivating chromosomes: A macro domain that minimizes transcription. Mol. Cell 12, 1–3. Langst, G., and Becker, P. B. (2001a). ISWI induces nucleosome sliding on nicked DNA. Mol. Cell 8, 1085–1092. Langst, G., and Becker, P. B. (2001b). Nucleosome mobilization and positioning by ISWIcontaining chromatin-remodeling factors. J. Cell Sci. 114, 2561–2568. Langst, G., Bonte, E. J., Corona, D. F., and Becker, P. B. (1999). Nucleosome movement by CHRAC and ISWI without disruption or trans-displacement of the histone octamer. Cell 97, 843–852. Leach, T. J., Mazzeo, M., Chotkowski, H. L., Madigan, J. P., Wotring, M. G., and Glaser, R. L. (2000). Histone H2A.Z is widely but nonrandomly distributed in chromosomes of Drosophila melanogaster. J. Biol. Chem. 275, 23267–23272. Licht, C. L., Stevnsner, T., and Bohr, V. A. (2003). Cockayne syndrome group B cellular and biochemical functions. Am. J. Hum. Genet. 73, 1217–1239. Lo, A. W., Craig, J. M., SaVery, R., Kalitsis, P., Irvine, D. V., Earle, E., Magliano, D. J., and Choo, K. H. (2001). A 330 kb CENP-A binding domain and altered replication timing at a human neocentromere. EMBO J. 20, 2087–2096. Logie, C., and Peterson, C. L. (1997). Catalytic activity of the yeast SWI/SNF complex on reconstituted nucleosome arrays. EMBO J. 16, 6772–6782. Lorch, Y., Cairns, B. R., Zhang, M., and Kornberg, R. D. (1998). Activated RSC-nucleosome complex and persistently altered form of the nucleosome. Cell 94, 29–34. Lorch, Y., Zhang, M., and Kornberg, R. D. (1999). Histone octamer transfer by a chromatinremodeling complex. Cell 96, 389–392. Lorch, Y., Zhang, M., and Kornberg, R. D. (2001). RSC unravels the nucleosome. Mol. Cell 7, 89–95.


145

Luger, K., Mader, A. W., Richmond, R. K., Sargent, D. F., and Richmond, T. J. (1997). Crystal structure of the nucleosome core particle at 2.8 A resolution. Nature 389, 251–260. Martens, J. A., and Winston, F. (2003). Recent advances in understanding chromatin remodeling by Swi/Snf complexes. Curr. Opin. Genet. Dev. 13, 136–142. Martinez-Balbas, M. A., Tsukiyama, T., Gdula, D., and Wu, C. (1998). Drosophila NURF-55, a WD repeat protein involved in histone metabolism. Proc. Natl. Acad. Sci. USA 95, 132–137. McDowell, T. L., Gibbons, R. J., Sutherland, H., O’Rourke, D. M., Bickmore, W. A., Pombo, A., Turley, H., Gatter, K., Picketts, D. J., Buckle, V. J., Chapman, L., Rhodes, D., and Higgs, D. R. (1999). Localization of a putative transcriptional regulator (ATRX) at pericentromeric heterochromatin and the short arms of acrocentric chromosomes. Proc. Natl. Acad. Sci. USA 96, 13983–13988. Mello, J. A., and Almouzni, G. (2001). The ins and outs of nucleosome assembly. Curr. Opin. Genet. Dev. 11, 136–141. Mizuguchi, G., Shen, X., Landry, J., Wu, W. H., Sen, S., and Wu, C. (2004). ATP-driven exchange of histone H2AZ variant catalyzed by SWR1 chromatin remodeling complex. Science 303, 343–348. Mohrmann, L., Langenberg, K., Krijgsveld, J., Kal, A. J., Heck, A. J., and Verrijzer, C. P. (2004). DiVerential targeting of two distinct SWI/SNF-related Drosophila chromatinremodeling complexes. Mol. Cell. Biol. 24, 3077–3088. Morillon, A., Karabetsou, N., O’Sullivan, J., Kent, N., Proudfoot, N., and Mellor, J. (2003). Isw1 chromatin remodeling ATPase coordinates transcription elongation and termination by RNA polymerase II. Cell 115, 425–435. Muftuoglu, M., Selzer, R., Tuo, J., Brosh, R. M., Jr., and Bohr, V. A. (2002). Phenotypic consequences of mutations in the conserved motifs of the putative helicase domain of the human Cockayne syndrome group B gene. Gene 283, 27–40. Muller, C., and Leutz, A. (2001). Chromatin remodeling in development and diVerentiation. Curr. Opin. Genet. Dev. 11, 167–174. Neely, K. E., and Workman, J. L. (2002). The complexity of chromatin remodeling and its links to cancer. Biochim. Biophys. Acta 1603, 19–29. Neigeborn, L., and Carlson, M. (1984). Genes aVecting the regulation of SUC2 gene expression by glucose repression in Saccharomyces cerevisiae. Genetics 108, 845–858. Ogas, J., Cheng, J. C., Sung, Z. R., and Somerville, C. (1997). Cellular diVerentiation regulated by gibberellin in the Arabidopsis thaliana pickle mutant. Science 277, 91–94. Ogas, J., Kaufmann, S., Henderson, J., and Somerville, C. (1999). PICKLE is a CHD3 chromatin-remodeling factor that regulates the transition from embryonic to vegetative development in Arabidopsis. Proc. Natl. Acad. Sci. USA 96, 13839–13844. Okada, M., and Hirose, S. (1998). Chromatin remodeling mediated by Drosophila GAGA factor and ISWI activates fushi tarazu gene transcription in vitro. Mol. Cell. Biol. 18, 2455–2461. O’Neill, D. W., Schoetz, S. S., Lopez, R. A., Castle, M., Rabinowitz, L., Shor, E., Krawchuk, D., Goll, M. G., Renz, M., Seelig, H. P., Han, S., Seong, R. H., Park, S. D., Agalioti, T., Munshi, N., Thanos, D., Erdjument-Bromage, H., Tempst, P., and Bank, A. (2000). An ikaros-containing chromatin-remodeling complex in adult-type erythroid cells. Mol. Cell. Biol. 20, 7572–7582. Owen-Hughes, T., and Bruno, M. (2004). Molecular biology. Breaking the silence. Science 303, 324–325. Peterson, C. L., and Cote, J. (2004). Cellular machineries for chromosomal DNA repair. Genes Dev. 18, 602–616. Peterson, C. L., Dingwall, A., and Scott, M. P. (1994). Five SWI/SNF gene products are components of a large multi-subunit complex required for transcriptional enhancement. Proc. Natl. Acad. Sci. USA 91, 2905–2908.

146

Smith and Peterson

Peterson, C. L., and Herskowitz, I. (1992). Characterization of the yeast SWI1, SWI2, and SWI3 genes, which encode a global activator of transcription. Cell 68, 573–583. Peterson, C. L., Kruger, W., and Herskowitz, I. (1991). A functional interaction between the Cterminal domain of RNA polymerase II and the negative regulator SIN1. Cell 64, 1135–1143. Peterson, C. L., and Tamkun, J. W. (1995). The SWI-SNF complex: A chromatin remodeling machine? Trends Biochem. Sci. 20, 143–146. Peterson, C. L., Zhao, Y., and Chait, B. T. (1998). Subunits of the yeast SWI/SNF complex are members of the actin-related protein (ARP) family. J. Biol. Chem. 273, 23641–23644. Phelan, M. L., Schnitzler, G. R., and Kingston, R. E. (2000). Octamer transfer and creation of stably remodeled nucleosomes by human SWI-SNF and its isolated ATPases. Mol. Cell. Biol. 20, 6380–6389. Quinn, J., Fyrberg, A. M., Ganster, R. W., Schmidt, M. C., and Peterson, C. L. (1996). DNAbinding properties of the yeast SWI/SNF complex. Nature 379, 844–847. Redon, C., Pilch, D., Rogakou, E., Sedelnikova, O., Newrock, K., and Bonner, W. (2002). Histone H2A variants H2AX and H2AZ. Curr. Opin. Genet. Dev. 12, 162–169. Reisman, D. N., Sciarrotta, J., Wang, W., Funkhouser, W. K., and Weissman, B. E. (2003). Loss of BRG1/BRM in human lung cancer cell lines and primary lung cancers: Correlation with poor prognosis. Cancer Res. 63, 560–566. Richard-Foy, H., and Hager, G. L. (1987). Sequence-specific positioning of nucleosomes over the steroid-inducible MMTV promoter. EMBO J. 6, 2321–2328. Richmond, E., and Peterson, C. L. (1996). Functional analysis of the DNA-stimulated ATPase domain of yeast SWI2/SNF2. Nucleic Acids Res. 24, 3685–3692. Rider, S. D., Jr., Hemm, M. R., Hostetler, H. A., Li, H. C., Chapple, C., and Ogas, J. (2004). Metabolic profiling of the Arabidopsis pkl mutant reveals selective derepression of embryonic traits. Planta. 219, 489–499. Roberts, C. W. M., Galusha, S. A., McMenamin, M. E., Fletcher, C. D. M., and Orkin, S. H. (2000). HaploinsuYciency of Snf5 (integrase interactor 1) predisposes to malignant rhabdoid tumors in mice. Proc. Natl. Acad. Sci. USA 97, 13796–13800. Saha, A., Wittmeyer, J., and Cairns, B. R. (2002). Chromatin remodeling by RSC involves ATP-dependent DNA translocation. Genes Dev. 16, 2120–2134. Salma, N., Xiao, H., Mueller, E., and Imbalzano, A. N. (2004). Temporal recruitment of transcription factors and SWI/SNF chromatin-remodeling enzymes during adipogenic induction of the peroxisome proliferator-activated receptor gamma nuclear hormone receptor. Mol. Cell. Biol. 24, 4651–4663. Santisteban, M. S., Kalashnikova, T., and Smith, M. M. (2000). Histone H2A.Z regulates transcription and is partially redundant with nucleosome remodeling complexes. Cell 103, 411–422. Sawa, H., Kouike, H., and Okano, H. (2000). Components of the SWI/SNF complex are required for asymmetric cell division in C. elegans. Mol. Cell 6, 617–624. Schnitzler, G., Sif, S., and Kingston, R. E. (1998). Human SWI/SNF interconverts a nucleosome between its base state and a stable remodeled state. Cell 94, 17–27. Schnitzler, G. R., Cheung, C. L., Hafner, J. H., Saurin, A. J., Kingston, R. E., and Lieber, C. M. (2001). Direct imaging of human SWI/SNF-remodeled mono- and polynucleosomes by atomic force microscopy employing carbon nanotube tips. Mol. Cell. Biol. 21, 8504–8511. Sengupta, S. M., Persinger, J., Bartholomew, B., and Peterson, C. L. (1999). Use of DNA photoaYnity labeling to study nucleosome remodeling by SWI/SNF. Methods 19, 434–446. Sengupta, S. M., VanKanegan, M., Persinger, J., Logie, C., Cairns, B. R., Peterson, C. L., and Bartholomew, B. (2001). The interactions of yeast SWI/SNF and RSC with the nucleosome before and after chromatin remodeling. J. Biol. Chem. 276, 12636–12644.


147

Sentani, K., Oue, N., Kondo, H., Kuraoka, K., Motoshita, J., Ito, R., Yokozaki, H., and Yasui, W. (2001). Increased expression but not genetic alteration of BRG1, a component of the SWI/SNF complex, is associated with the advanced stage of human gastric carcinomas. Pathobiology 69, 315–320. Shen, X., Mizuguchi, G., Hamiche, A., and Wu, C. (2000). A chromatin remodeling complex involved in transcription and DNA processing. Nature 406, 541–544. Smith, C. L., Horowitz-Scherer, R., Flanagan, J. F., Woodcock, C. L., and Peterson, C. L. (2003). Structural analysis of the yeast SWI/SNF chromatin remodeling complex. Nat. Struct. Biol. 10, 141–145. Smith, M. M. (2002). Centromeres and variant histones: What, where, when and why? Curr. Opin. Cell Biol. 14, 279–285. Staehling-Hampton, K., Ciampa, P. J., Brook, A., and Dyson, N. (1999). A genetic screen for modifiers of E2F in Drosophila melanogaster. Genetics 153, 275–287. Stern, M., Jensen, R., and Herskowitz, I. (1984). Five SWI/SNF genes are required for expression of the HO gene in yeast. J. Mol. Biol. 178, 853–868. Stopka, T., and Skoultchi, A. I. (2003). The ISWI ATPase Snf2h is required for early mouse development. Proc. Natl. Acad. Sci. USA 100, 14097–14102. Sudarsanam, P., Iyer, V. R., Brown, P. O., and Winston, F. (2000). Whole-genome expression analysis of snf/swi mutants of Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA 97, 3364–3369. Tong, J. K., Hassig, C. A., Schnitzler, G. R., Kingston, R. E., and Schreiber, S. L. (1998). Chromatin deacetylation by an ATP-dependent nucleosome remodelling complex. Nature 395, 917–921. Truss, M., Bartsch, J., Schelbert, A., Hache, R. J., and Beato, M. (1995). Hormone induces binding of receptors and transcription factors to a rearranged nucleosome on the MMTV promoter in vivo. EMBO J. 14, 1737–1751. Tsuchiya, E., Hosotani, T., and Miyakawa, T. (1998). A mutation in NPS1/STH1, an essential gene encoding a component of a novel chromatin-remodeling complex RSC, alters the chromatin structure of Saccharomyces cerevisiae centromeres. Nucleic Acids Res. 26, 3286–3292. Tsukiyama, T., Becker, P. B., and Wu, C. (1994). ATP-dependent nucleosome disruption at a heat-shock promoter mediated by binding of GAGA transcription factor [see comments]. Nature 367, 525–532. Tsukiyama, T., Palmer, J., Landel, C. C., Shiloach, J., and Wu, C. (1999). Characterization of the imitation switch subfamily of ATP-dependent chromatin-remodeling factors in Saccharomyces cerevisiae. Genes Dev. 13, 686–697. Unhavaithaya, Y., Shin, T. H., Miliaras, N., Lee, J., Oyama, T., and Mello, C. C. (2002). MEP1 and a homolog of the NURD complex component Mi-2 act together to maintain germlinesoma distinctions in C. elegans. Cell 111, 991–1002. van Holde, K., and Yager, T. (2003). Models for chromatin remodeling: a critical comparison. Biochem. Cell Biol. 81, 169–172. Varga-Weisz, P. D., Wilm, M., Bonte, E., Dumas, K., Mann, M., and Becker, P. B. (1997). Chromatin-remodelling factor CHRAC contains the ATPases ISWI and topoisomerase II [published erratum appears in Nature 1997 Oct 30;389(6654):1003]. Nature 388, 598–602. Vary, J. C., Jr., Gangaraju, V. K., Qin, J., Landel, C. C., Kooperberg, C., Bartholomew, B., and Tsukiyama, T. (2003). Yeast Isw1p forms two separable complexes in vivo. Mol. Cell. Biol. 23, 80–91. Versteege, I., Sevenet, N., Lange, J., Rousseau-Merck, M. F., Ambros, P., Handgretinger, R., Aurias, A., and Delattre, O. (1998). Truncating mutations of hSNF5/Ini1 in aggressive paediatric cancer. Nature 394, 203–206.

148

Smith and Peterson

Vignali, M., Hassan, A. H., Neely, K. E., and Workman, J. L. (2000). ATP-dependent chromatin-remodeling complexes. Mol. Cell. Biol. 20, 1899–1910. Wade, P. A., Gegonne, A., Jones, P. L., Ballestar, E., Aubry, F., and WolVe, A. P. (1999). Mi-2 complex couples DNA methylation to chromatin remodelling and histone deacetylation. Nat. Genet. 23, 62–66. Wade, P. A., Jones, P. L., Vermaak, D., Veenstra, G. J., Imhof, A., Sera, T., Tse, C., Ge, H., Shi, Y. B., Hansen, J. C., and WolVe, A. P. (1998a). Histone deacetylase directs the dominant silencing of transcription in chromatin: Association with MeCP2 and the Mi-2 chromodomain SWI/SNF ATPase. Cold Spring Harb. Symp. Quant. Biol. 63, 435–445. Wade, P. A., Jones, P. L., Vermaak, D., and WolVe, A. P. (1998b). A multiple subunit Mi-2 histone deacetylase from Xenopus laevis cofractionates with an associated Snf2 superfamily ATPase. Curr. Biol. 8, 843–846. Wechser, M. A., Kladde, M. P., Alfieri, J. A., and Peterson, C. L. (1997). EVects of Sin-versions of histone H4 on yeast chromatin structure and function. EMBO J. 16, 2086–2095. Whitehouse, I., Flaus, A., Cairns, B. R., White, M. F., Workman, J. L., and Owen-Hughes, T. (1999). Nucleosome mobilization catalysed by the yeast SWI/SNF complex. Nature 400, 784–787. Whitehouse, I., Stockdale, C., Flaus, A., Szczelkun, M. D., and Owen-Hughes, T. (2003). Evidence for DNA translocation by the ISWI chromatin-remodeling enzyme. Mol. Cell. Biol. 23, 1935–1945. Winston, F., and Carlson, M. (1992). Yeast SNF/SWI transcriptional activators and the SPT/ SIN chromatin connection. Trends Genet. 8, 387–391. WolVe, A. (1998). ‘‘Chromatin: Structure and Function,’’ pp. xiv 447 p. Academic Press, San Diego. Wong, A. K., Shanahan, F., Chen, Y., Lian, L., Ha, P., Hendricks, K., GhaVari, S., Iliev, D., Penn, B., Woodland, A. M., Smith, R., Salada, G., Carillo, A., Laity, K., Gupte, J., Swedlund, B., Tavtigian, S. V., Teng, D. H., and Lees, E. (2000). BRG1, a component of the SWI-SNF complex, is mutated in multiple human tumor cell lines. Cancer Res. 60, 6171–6177. Zlatanova, J., and Leuba, S. H. (2003). Chromatin fibers, one-at-a-time. J. Mol. Biol. 331, 1–19.

5

____________________________________________________________________________

Self-Destruct Programs in the Processes of Developing Neurons David Shepherd and V. Hugh Perry School of Biological Sciences University of Southampton Southampton SO16 7PX, United Kingdom

I. Introduction II. Active Axon and Synapse Degeneration III. Lessons from Drosophila A. Neuron Remodeling During Metamorphosis in Drosophila Provides an Ideal Opportunity to Study Process Elimination and Growth B. Mechanisms that Control the Development of Neuronal Processes C. The Drosophila Neuromuscular Junction and Retrograde Signaling D. Compartmentalized Process Elimination: Implications for Neurodegenerative Disease IV. Conclusions References

I. Introduction The death of neurons and glia is a significant feature of the development of both vertebrate and invertebrate nervous systems. The death of the neuronal cell body occurs by a programmed-cell-death (PCD) process and commonly has the features of apoptotic cell death. In vertebrates, the numbers of neurons and glia dying during the course of development varies across the neuraxis, but available figures indicate that at least 50% of the neurons and glia generated will undergo PCD (Burek and Oppenheim, 1996; RaV et al., 1993). Some of the rules that govern which cells will survive and which will degenerate have been extensively reviewed elsewhere, with the competition for neurotrophic factors a major component (Lewin and Barde, 1996). The molecular events leading to apoptosis and the demise of the cell soma are known to involve multiple cascades and the complexity of the intracellular pathways is well recognized (Yuan and Yankner, 2000). However, a neglected component of developmental neuronal death concerns what happens to the processes of neurons in which the cell soma has undergone or is going to undergo PCD. It has been shown that the majority of relay neurons that undergo PCD have already made connections with their target Current Topics in Developmental Biology, Vol. 65 Copyright 2005, Elsevier Inc. All rights reserved. 0070-2153/05 $35.00

149

150

Shepherd and Perry

structure prior to degeneration of the cell soma (Perry et al., 1983), and thus the axon, its synaptic terminals, and dendrites must also degenerate. Other populations of neurons develop axons or axon collaterals that will not survive into adulthood; for example, developing cortical neurons have intracortical or callosal axons that will not persist into adulthood (Innocenti, 1995). Although the presence of these ‘‘exuberant’’ axons and their synaptic terminals have been known for decades, the mechanisms that underlie their disappearance has been largely ignored. In most circumstances, it is unclear whether these processes passively degenerate, are actively retracted, actively degenerate, or are actively removed by glia. It is also not known whether diVerent neuronal populations utilize diVerent mechanisms. Similarly, the loss of dendrites, another compartment of the neuron, is not well understood. One possible reason for the apparent lack of interest in this issue may center on the view that the degeneration of axons and synapses is a passive process that simply follows the loss or degeneration of the cell body, in the same way that degeneration of the distal axon follows an axon injury. However, this scenario cannot explain how axon collaterals and synapses are remodeled after the period of cell death, as is the case at the neuromuscular junction (Sanes and Lichtman, 1999), or as part of the refinement of cortical connections (Innocenti, 1995). Furthermore, there is now good evidence to show that in the adult nervous system, the degeneration of the cell body, axon, and synapses of a neuron is a compartmentalized process as well as an active autodestructive process akin to PCD (Coleman and Perry, 2002; Gillingwater and Ribchester, 2001). Several studies highlight the importance of compartmentalized autodestruction in neuronal processes and oVer tractable models in which to investigate the molecular pathways involved.

II. Active Axon and Synapse Degeneration In mammals, transection of an axon leads to rapid degeneration of the isolated distal segment of the axon. This form of degeneration, Wallerian degeneration, is characterized by a sequence of well-documented ultrastructural changes, including the final stage of granular disintegration of the axoplasm. The morphological features of this degeneration are common across neuron populations and species (GriYn et al., 1996). It was believed that this form of axon degeneration was a passive withering of the axon due to the distal portion of the neuron being separated from the soma, the site of protein synthesis (Finn et al., 2000). Other studies suggested that an influx of calcium-activated proteases, the calpains, led to the degradation of the axon cytoskeleton and subsequent degeneration of the axon (Schlaepfer and Hasler, 1979). Furthermore, manipulating the levels of calcium can alter the rate of axon degeneration and delay it for a few hours. However, in the

5. Self-Destruct Programs in Developing Neurons s

151

mutant Wld mouse, the axon segment distal to a transection in either the PNS or CNS will survive for several weeks following transection (Lunn et al., 1989; Perry et al., 1991). Remarkably, this distal segment not only retains its morphology but also the ability to conduct a compound action potential when electrically stimulated. In Wlds mice synapses at the neuromuscular junction, which normally undergo degeneration in 1 to 3 days after axon transection in wild-type animals, may survive for 10 days or longer (Ribchester et al., 1995). These simple observations tell us that the process of axon and synapse degeneration in wild-type mice must be an active autodestructive process akin to PCD. Molecular genetic studies have isolated the autosomal dominant mutation to chromosome 4 in the mouse and shown that it has arisen as a consequence of a tandem triplication of 85 kb (Coleman et al., 1998). At the junctions of the triplication, a chimeric gene Wlds is formed that has been shown in transgenic mice to be axon protective in a dose-dependent fashion (Mack et al., 2001). The Wlds gene is formed from the first 70 amino acids of the ubiquitin ligase UbE4b and the whole open-reading frame of the enzyme nicotinamide mononucleotide adenlyltransferase (Nmnat) responsible for the synthesis of NAD. Precisely how this chimeric protein gives rise to the protection of the axon is currently a matter of intense investigation but strongly suggests a role of the ubiquitin proteasome system (UPS) in axon degeneration (Coleman and Perry, 2002). The role of the UPS in axon degeneration supported by findings in the gad mouse in which there is spontaneous degeneration of axons as the animals mature. This mouse carries a mutation in the ubiquitin hydrolase UCHL-1 (Saigoh et al., 1999). The mutation is believed to lead to a deficiency of free ubiquitin in the axon. Other genetic studies reveal that mutations in the UPS lead to degeneration of neurons, including Angelman’s syndrome (Guerrini et al., 2003), with a mutation in an E2 ubiquitin ligase leading to mental retardation, and a mutation in a ubiquitin ligase leads to early degeneration of dopaminergic cells in familial Parkinson’s disease (Shimura et al., 2000). The pathways of PCD in axons have not been extensively studied but appear to be distinct from processes operating at the cell soma. Neurotrophin withdrawal that leads to apoptosis of the cell soma of sympathetic neurons in vitro similarly leads to apoptosis of Wlds sympathetic neuron cell bodies in vitro but leaves the neurites isolated and intact for several days (Deckwerth and Johnson, 1994). The role of caspases in PCD of the cell soma is well established but Wallerian degeneration of the axon in vitro takes place in the absence of caspase activation (Finn et al., 2000; Sievers et al., 2003). The morphological features of PCD at the cell soma involve the appearance of phosphatidylserine residues on the external leaflet of the plasma membrane and these residues can be detected by Annexin V staining, followed by blebbing of the plasma membrane. A similar sequence of events

152

Shepherd and Perry

has been documented in transected neurites in vitro. Of interest is the observation that the Annexin V staining appears in a proximodistal sequence, as does the loss of the mitochondrial membrane potential, suggestive of the passage of a signal along the neurites during the degeneration process (Sievers et al., 2003). In vitro studies of Wlds neurons show that isolated neurites only acquire the slow degeneration phenotype after being in culture for several days, suggesting that some component of the protection process is being transported into the processes (Buckmaster et al., 1995). All of these observations on PCD in axons indicate a process distinct from apoptosis at the cell soma and in vivo studies on transgenic mice have provided further evidence for a distinct process. Mice overexpressing Bcl-2 show a greater resistance to injury-induced apoptosis of the cell soma but Wallerian degeneration is not delayed (Burne et al., 1996), although in contrast Bax-deficient mice do appear to show delayed axon degeneration after transection (Dong et al., 2003). Electron microscopy studies of fiber tracts in developing animals, at times when it is known that large numbers of axons are being lost, have not reported large-scale axon degeneration (Perry et al., 1983), although occasional observations of macrophages phagocytosing axon debris can be found (Innocenti et al., 1983). It is perhaps unclear how axons undergoing an autodestructive process might appear morphologically. The loss of synapses is also a major component of the modeling on the CNS and has been described in numerous systems (Purves and Lichtman, 1980). The molecular events, however, which underlie this synaptic loss are not well understood, but again, studies from Wlds strongly support the idea that the molecular machinery in the synaptic compartment may act independently of the axon and cell soma: indeed, this notion is clear from the many elegant studies that have been carried out on the NMJ in mammals. The development of the mono-innervation of a muscle fiber arises from a situation where muscle fibers are poly-innervated and this involves the selective withdrawal of axon terminals. This process has been extensively studied from the morphological perspective and it has been shown that the axon terminal which is withdrawn first gets thinner and then lifts from the surface of the muscle fiber to form a retraction bulb (Purves and Litchman, 1980). The initiation of this signal depends on the signals from the postsynaptic site since injury to a muscle fiber, or inhibition of protein synthesis, will precipitate withdrawal. This process is morphologically distinct from the degeneration of the synapse that typically follows axon injury and may involve the UPS since transection of a peripheral nerve in young Wlds mice leads to delayed asynchronous withdrawal of NMJs in a manner similar to that seen in developmental withdrawal (Gillingwater et al., 2003). Synaptic withdrawal with the formation of retraction bulbs has been documented within the climbing fiber population in the developing cerebellum and when the target

5. Self-Destruct Programs in Developing Neurons

153

field of axons is destroyed (see Bernstein and Lichtman, 1999). More subtle forms of synapse withdrawal may also be observed in the adult CNS when axon terminals withdraw their synapses from injured motorneurons (Blinzinger and Kreutzberg, 1968). The process of ‘‘synaptic apoptosis’’ has been studied in synaptosomes from the mature brain and involves activation of caspases and some of the molecular machinery associated with apoptosis at the cell soma (Mattson, 2000). Whether these same processes are involved in synaptic withdrawal in development and whether they precede the loss of the cell soma and axon is not known. Of interest is the evidence from the olfactory system showing that destruction of the olfactory bulb leads to caspase activation in olfactory receptor neurons first in the axon terminals, then in the axon, and finally in the cell soma (Cowan et al., 2001). A caspase-dependent cleavage product of amyloid precursor protein-2 shows the same spatiotemporal distribution, indicating the retrograde propagation of a pro-apoptotic signal to the cell soma from the axon terminal. It is not known whether the importins, axoplasmic proteins containing a nuclear localization signal (Hanz et al., 2003), are involved in this retrograde signaling, which could precipitate apoptosis at the cell soma. There are large gaps in our understanding of how synapses and axons are withdrawn or lost during development of the CNS. However, studies on more tractable systems than those of mammals oVer exciting new findings.

III. Lessons from Drosophila In the past decades, experiments on insects have had a huge impact on our understanding of the basic mechanisms of nervous system development. Early ideas suggested that invertebrate neurons were fundamentally diVerent from mammalian neurons in that their axonal and dendritic morphology was assembled according to a tightly regulated genetic program that did not involve the elimination of supernumerary processes. It is now, however, abundantly clear that invertebrate neurons demonstrate considerable structural plasticity during development. During embryonic development, neurons develop exuberant processes that are eliminated during the latter stages of their development (Shepherd and Laurent, 1992). During metamorphosis, the dendrites and axons of many neurons undergo a spectacular remodeling (Tissot and Stocker, 2000). Insects also demonstrate considerable synaptic remodeling in response to changes in the number of competing neurons (Shepherd and Murphey, 1986), changes in activity (Barth et al., 1997), or during normal development (Chiba et al., 1988). Despite the abundant evidence of such developmental plasticity in insects, it is surprising that

154

Shepherd and Perry

molecular genetic tools available in Drosophila have not been used to explore the mechanism underlying these crucial processes. The reason for this has been due mostly to the small size of Drosophila and the inaccessibility of its neurons. In recent years, however, this has changed and the development of new methodologies has meant that it is now possible to visualize identified neurons in the peripheral and central nervous systems of partially and completely intact animals. With this breakthrough, the fly is beginning to have a major impact on our understanding of how neuronal processes are eliminated in development and to shape our views of its role and the mechanisms controlling it.

A. Neuron Remodeling During Metamorphosis in Drosophila Provides an Ideal Opportunity to Study Process Elimination and Growth One of the definitive features of the development of the nervous system in Drosophila is the remodeling of the nervous system that takes place during the metamorphosis of the larva into the adult. This period of development is characterized by massive changes in the body of the fly and its underlying neuronal machinery. During metamorphosis, many neurons undergo a highly regulated and stereotyped pruning and remodeling. This remodeling is characterized by the loss of axons, dendrites, and synapses, often in the absence of cell death. There are many specific examples but all show a synchronized and tightly regulated temporal sequence of changes that makes them ideal for systematic studies. Unlike in the mammalian nervous system, many of these changes can be observed in situ to provide an excellent experimental system for studying the mechanisms that shape the development of the diVerent neuronal compartments. This includes the mechanisms that control both the elaboration of neuronal processes and their degeneration. These studies have shown that axonal degeneration and dendrite remodeling are regulated by a self-destruction program that is precisely controlled, both spatially and temporally. Studies show elegant examples of neurons in which there are both degeneration and growth of synapses and dendrites within diVerent regions of the same neuron. While much of the emphasis of this work has been on the hormonal and signal transduction mechanisms that control the temporal aspects of the remodeling processes, these observations illustrate a remarkable spatial precision for the actions of growth and degeneration within diVerent compartments of neurons. This demonstrates that the elimination of dendritic and axonal processes during metamorphosis is an active process controlled in diVerent ways in diVerent compartments and that it is not a global development program that aVects the entire cell.


155

B. Mechanisms that Control the Development of Neuronal Processes 1. Dendritic Outgrowth in Sensory Neurons—Intrinsic Control of Dendrite Morphology The body wall of Drosophila larvae is innervated by a stereotyped array of sensory neurons with dendritic trees that cover the surface of the body wall. These neurons, called dendritic arborization (da) neurons, are readily accessible to experimental analysis and have been subjected to considerable study to identify the mechanisms that shape the development of their dendrites. It has been shown that the development of da neuron dendritic trees is controlled by a number of cell intrinsic factors that regulate the outgrowth of dendritic branches. This work has identified genes that control the pattern and size of the dendritic trees. For example, the gene hamlet, which encodes a transcription factor, determines the morphological diVerences between 2 subclasses of sensory neurons (Moore et al., 2002). One class of neuron has a single dendrite (es neurons) whereas the second class (da neurons) has the complex da dendritic tree. ham is normally expressed in the es precursor and nascent es neurons. The same es neurons in ham mutants, however, acquire a da-like morphology and express da-specific gene markers. In the converse experiment, misexpression of ham in da neurons results in da neurons with reduced branching but not a single dendrite nor es-type gene expression. Nevertheless, the results show that the ham plays a role in establishing the basic pattern of dendrites produced. In a similar vein, the gene cut is also a regulator of dendrite pattern (Grueber et al., 2003). Within the da neurons, there are 4 identified subclasses based on the complexity of their dendritic trees and the level of dendrite complexity is directly related to the level of cut expression, with simpler neurons showing the lowest cut expression and the highest cut expression occurring in the most complex. Overexpression of cut in simpler neurons produces greater dendritic complexity. Although both these examples are informative, the modes of action of these transcription factors still remain to be determined as well as the mechanisms by which dendritic trees are assembled. 2. External Factors Controlling Dendrite Development Although there is evidence of genetic factors controlling the morphology of da dendrites, there is also variability even among identified homologues. While such variations in structure might be expected and advantageous, it also indicates that dendrites can be shaped by extrinsic factors. Foremost of these are interactions among the dendrites of other da neurons. Several studies have shown that there are exclusion reactions among dendrites

156

Shepherd and Perry

which limit the dendritic fields of the da neurons and ensure that there is complete but non-overlapping innervation of the body wall (Grueber et al., 2003; Sugimura et al., 2003). Thus, ablation of one da neuron can result in the invasion of the vacated territory by neighboring da neurons of the same class. A similar process has been demonstrated in the developing retina of mammals (Perry and Linden, 1982). Although these experiments clearly indicate the existence of repulsive interactions among dendrites, the molecular basis of the process remains unknown, although studies of the remodeling of da neurons during metamorphosis are beginning to shed light on the importance of branch retraction in shaping the dendritic tree.

3. Process Elimination is Important in Shaping Dendrites The importance of process elimination in shaping da neuron dendrites was first shown using time-lapse microscopy of da neuron remodeling during metamorphosis (Williams and Truman, 2004). This provided the first dynamic description of the remodeling of a single identified da neuron during metamorphosis and showed that remodeling of the dendritic tree is dependent not only on the growth of new branches but also on the controlled pruning of branches. During the earliest stages of metamorphosis (0–19 h), the da neuron loses its larval dendrites via a combination of local degeneration and active resorption of dendrites. Thus, some dendrites exhibit the classical signs of Wallerian degeneration while others disappear by drawing back in a proximal to distal manner into the body of the neuron without signs of Wallerian degeneration. The relative contribution of these two processes, however, is not yet known but does suggest that the selective loss of branches can be achieved by tightly controlled cell intrinsic mechanisms. After pruning its dendrites, the neuron extends new dendrites. During the early phases of re-growth, branch retraction continues to play an important role and dendrite growth is characterized by the production of a large number of filopodia and thin processes tipped with growth cones, which undergo extensive extension and retraction. In addition to this filopodial activity, however, the new dendritic branches are still labile and even the most robust dendritic branches can undergo retraction. Thus, a combination of dendritic retraction and growth is required to assemble the main scaVold of major dendritic branches. During the final stages of outgrowth, the major dendritic branches become stabilized and, although there is still considerable filopodial activity, branch retraction ceases. At this time, the basic dendritic scaVold undergoes a rapid expansion and elaboration to produce the second- and third-order branches of the adult dendritic field.


157

4. Extrinsic Factors in Dendrite Elimination By experimentally manipulating the development of the da neuron with juvenile hormone (JH), the authors showed that suppression of branch retraction is controlled by JH. When late-stage neurons are exposed to JH, the branch retraction that characterizes the early growth phases is maintained, and the dendritic branches do not stabilize and still retract. Evidently, the JH pulse at the onset of metamorphosis triggers the expression of genes that control branch retraction and activate the pruning of the larval dendritic tree. During re-growth, the branch retraction mechanisms are still expressed alongside the newly activated growth process and a combination of growth and retraction is required to control the lability and retraction of major branches. As the neuron enters the third phase of its development, the retraction program is repressed and retraction of branches is no longer seen, although outgrowth continues. Clearly, retraction of dendrites plays a key role in this process, but what is its significance? Is it that during the labile phase branches are extended to explore the milieu and, like the filopodia of a growth cone, seek the appropriate cues and, should the necessary cues not be found, the dendrite collapses? 5. Axon Pruning Branch retractions are not restricted to the peripheral nervous system or dendrites. Watts et al. (2003) have shown that selective pruning of axonal processes is also a part of the normal development of a population of central neurons in the mushroom body (MB) of Drosophila. During metamorphosis, a subset of MB neurons, the neurons, undergo a stereotyped restructuring, which sees the selective pruning of dendrites and axonal branches prior to extending new processes to produce the adult morphology of the neurons. This remodeling is synchronized with a high degree of spatial specificity. Unlike in the da neurons, axonal pruning occurs by a largely degenerative process rather than by retraction, with the axon segments to be pruned displaying the features of Wallerian degeneration. Given that the axon pruning has the appearance of Wallerian degeneration, the authors sought to determine whether the molecular mechanisms underlying the degeneration were comparable to those described for Wallerian degeneration in mammalian neurons. As has been described, the molecular mechanisms of Wallerian degeneration are largely unknown but do show two characteristic features. First, Wallerian degeneration is distinct from and independent of PCD at the cell soma (Deckwerth and Johnson, 1994) and, second, inhibition of caspase activity can prevent PCD of the cell soma but cannot rescue the distal segments of severed axons from Wallerian degeneration (Finn et al., 2000; Sievers et al., 2003). In Drosophila, it was

158

Shepherd and Perry

shown that blocking PCD of the cell soma in the neurons, by either genetically deleting the apoptosis activators (grim, hid, and rpr) or by expressing a p35 caspase inhibitor, does not prevent degeneration of the neuron axons (Watts et al., 2003). Thus, these studies mirror the results from mammals and support the concept that PCD pathways in axons undergoing Wallerian degeneration in Drosophila are distinct from the PCD pathways at the cell soma. 6. The UPS and Axon Pruning The dramatically delayed axon degeneration in the Wlds mutant implicates the UPS as the key mechanisms mediating Wallerian degeneration in the distal ends of severed axons (Coleman and Perry, 2002). To test the role of UPS in axon pruning in Drosophila, Watts et al. (2003) used genetics to manipulate UPS function during axon pruning. Selective expression of the yeast ubiquitin protease (UBP2), which is responsible for de-ubiquitinating proteins, eVectively reverses the eVects of the UPS. When examined before metamorphosis, UBP2 expressing neurons displayed normal morphology indicating normal axon outgrowth. When examined during axon pruning, however, the UBP2 expressing neurons failed to prune their axons. This indicates that inhibition of the UPS can potently block axonal pruning and that enforced de-ubiquitination is protective of axons. To extend this, Watts et al. (2004) analyzed the eVects of mutations in ubiquitin-activating enzymes on axon pruning. This showed that mutations in the ubiquitinactivating enzyme E1 (Uba1) also blocked axon pruning and confirmed that a normally active UPS is required for axon degeneration. Mutations in two genes encoding diVerent subunits of the 19S proteasome regulatory particle also block the pruning of the neurons. It thus appears that axon pruning during metamorphosis in Drosophila shares the molecular mechanism as injury-induced axon degeneration in mammals in that it is a distinct PCD process relying heavily on normal proteasome activity. These studies show that the mechanism driving the axon pruning is cell autonomous but it does not explain how the process is spatially regulated within the anatomically distinct compartments of the neuron. During the pruning process, only the dorsal and medial axons degenerate; the axon peduncle remains intact: this indicates that the mechanism driving the process elimination must be restricted to the more distal elements of the axons. Similar experiments to investigate the role of the proteasome in axon degeneration have been performed in the mouse; however, it should be noted that these manipulations were only able to rescue severed axons for a few hours (Zhai et al., 2003), in contrast to the prolonged survival of several days seen in the Wlds mouse. Furthermore, other studies using proteasome inhibitors on mammalian neurons growing in vitro have


159

shown that this treatment may actually cause neurite degeneration (Laser et al., 2003), which not only demonstrates the diverse role of the UPS in multiple cell processes, but raises questions about the specificity of pharmacological inhibitors of the proteasome. 7. Glial Cells and Axon Elimination The evidence suggests that axon pruning is executed via a cell-intrinsic program but there is evidence that glial cells may also play a role. It was shown a number of years ago that microglia phagocytose the cell bodies of neurons undergoing apoptosis in the mammalian brain (Perry et al., 1985) but evidence has emerged that the microglia may actively participate in the death of developing neurons (Marin-Teva et al., 2004). Detailed studies of the axon pruning in Drosophila now suggest that the degeneration of the MB axons involves glial cells that invade the regions of degenerating axons and engulf the cellular debris created by axonal degeneration (Watts et al., 2004). Interestingly, these glial cells selectively accumulate around the axons destined to degenerate before degeneration commences. They accumulate even when degeneration is blocked, showing that the accumulation of glial cells is not triggered by axon degeneration. Although these observations suggest that glia cannot initiate axon pruning, Awasaki and Ito (2004) have shown that the glial cells are necessary for degeneration. Inhibition of specific cellular functions in the glia, such as endocytosis, suppresses glial infiltration and axon elimination, thereby causing a severe delay in axon pruning. The results suggest that although axon pruning is currently thought to be mediated by a cell-autonomous genetic program, the presence of glia suggests that extrinsic cells may assist the progress of the intrinsic program. One interesting possibility is that it is the glia that provide the spatial precision of axon degeneration.

C. The Drosophila Neuromuscular Junction and Retrograde Signaling The Drosophila neuromuscular junction (NMJ) is established as one of the key systems for analyzing the development and function of synapses. The main body wall of Drosophila is served by a stereotyped, bilaterally symmetrical, and segmentally repeated array of 30 muscles. The innervation of these muscles is precise, with each muscle being innervated by only one or two identified motorneurons which establish branched multiterminal NMJs, with distinctive synaptic boutons, across the surface of the muscle: each NMJ is uniquely identifiable and it is possible to study identified synapses (Gramates and Budnik, 1999).

160

Shepherd and Perry

One interesting aspect of the Drosophila NMJ is that the size of the synapse changes significantly through development. During normal development, the NMJ grows from a synapse with approximately 20 boutons, with each bouton containing a single active zone, to a junction with about 100 boutons, each bearing up to 12 active zones (Schuster et al., 1996). This growth can be seen as being necessary to match the size of the synaptic input with the increasing size of the muscle and the mechanism of synaptic growth has attracted considerable interest. The synaptic growth is associated with rapid disassembly of synapses and growth of the synapse is, in eVect, a balance of bouton addition and retraction (Eaton et al., 2002). The retraction of synapses at the NMJ appears to be determined presynaptically. The first signs of synapse withdrawal are the loss of microtubules and microtubule-associated proteins in the presynaptic terminal, followed by the loss of the presynaptic proteins synapsin and vesicle-associated proteins. All of these presynaptic proteins are lost prior to loss of postsynaptic receptors and dissolution of the postsynaptic membrane specializations (Eaton et al., 2002). Although these data show that the withdrawal mechanism is intrinsic to the motorneuron, the initial decision to withdraw is controlled by events in the muscle as in mammals. The retraction and growth of the NMJ synapses is regulated by the bone morphogenetic protein BMP signaling pathway (Aberle et al., 2002; Marques et al., 2002). According to this hypothesis, the retrograde ligand from the muscle is a member of the BMP family encoded by the glass bottom boat (gbb) gene. gbb is expressed in muscles and, consistent with its presumed role, loss of function gbb mutants have a reduced NMJ. Furthermore, expression of Gbb in muscles can rescue the gbb phenotype. The retrograde Gbb signal is most likely transduced by the product of the wishful thinking (wit) gene, which encodes a type II BMP receptor (Aberle et al., 2002; Marques et al., 2002) and, consistent with its presumed role, loss of function wit mutations have a small NMJ. The BMP signaling pathway is highly conserved in vertebrates and invertebrates and has been well characterized. Typically, BMP activates a tetrameric BMP receptor complex which phosphorylates cytoplasmic receptor regulated Smads, which associate with common nonphosphorylated Smads (co-Smads) to form a phospho-Smad complex which translocates to the nucleus to regulate transcription (Zwijsen et al., 2003). All of the presumed candidate components for BMP signaling, including type 1 and II receptors, RSmads, and co-Smads, are found in the Drosophila genome and loss of function mutations in each of them can result in reduced NMJ phenotypes (Arora et al., 1996; McCabe et al., 2004). Furthermore, Gbb and Wit are expressed in neurons and the type I BMP receptor (Tkv) is detected in presynaptic boutons. Clearly, all of the evidence suggests that BMP signaling mediates NMJ growth regulation and yet Smads act at the nucleus, so the key question is how a peripherally activated signal is translated into a


161

transcription change in the nucleus. The most obvious route for this is retrograde axonal transport and, indeed, expression of a dominant negative inhibitor of the Dynein/dynactin microtubule motor prevents the accumulation of phospho-Smad in the nucleus (McCabe et al., 2003). This suggests that the retrograde transport of a BMP-activated signaling complex is responsible for regulating the growth and retraction of the NMJ in Drosophila. The UPS also plays a key role in regulating the growth of the NMJ. A mutagenic study designed to detect abnormal NMJ development uncovered a mutation that resulted in a massively overgrown synapse. The mutation was mapped to the highwire (hiw) gene, which encodes a putative E3 ubiquitin ligase (Chang et al., 2000; Wan et al., 2000). It was proposed that hiw might negatively regulate NMJ growth by tagging growth-related protein signals for proteolytic degradation. It has been shown subsequently that hiw directly associates with the Drosophila coSmad, suggesting the attractive hypothesis that hiw can set the upper limit for synapse growth by limiting the availability of coSmad for regulating transcription. The result implies that the UPS play a key role in regulating the normal development of the synapse and that eVective inhibition of ubiquitination has the eVect of promoting synaptic overgrowth and synaptic malfunction. The role of the UPS in synaptic function, and synaptic protein turnover at the postsynaptic element in particular, has been recognized in mammals (Ehlers, 2003), but its role in development and elimination of exuberant synapses remains to be explored.

D. Compartmentalized Process Elimination: Implications for Neurodegenerative Disease With the emergence of Drosophila as a model system for studying the mechanisms of neurodegenerative diseases, it has become possible to extend these developmental models to analyze the mechanisms of axon and synaptic degeneration. Drosophila has been used to model a wide range of human neurodegenerative diseases, most notably polyglutamine repeat diseases, tauopathies, Alzheimer’s disease, and Parkinson’s (for review, see Shulman et al., 2003). These studies have proven extremely valuable in understanding the potential mechanisms of disease pathogenesis, the majority of such studies have focused on neuronal death and have not considered the eVects these diseases may have on axonal, dendritic, and synaptic compartments of the aVected neurons. This emphasis on neuron death is surprising, given that the earliest stages of neuropathology in these disease models aVect distal processes long before death of the cell soma. For example, in murine models of Alzheimer’s disease (Chapman et al., 1999), prion disease (Cunningham et al., 2003), and amyotrophic lateral sclerosis (Fischer et al., 2004), synaptic

162

Shepherd and Perry

loss precedes the death of the cell soma and may underpin the earliest neurological symptoms associated with these diseases. The importance of the early loss of processes is well illustrated in a Drosophila model of tauopathies. Williams et al. (2000) showed that targeted overexpression of human tau in Drosophila sensory neurons resulted in axonal abnormalities in which axonal endings showed degeneration in the absence of neuron death. In a follow-up study, Mudher et al. (2004) showed that neuronal death in tau-expressing motorneurons is preceded by an array of cellular responses that include abnormalities in axonal transport and also neurological deficits, including abnormalities of locomotor behavior. It has also been shown that tau overexpression in motorneurons also produces abnormal NMJ morphology (McKay et al., unpublished) and failure of synaptic transmission at the neuromuscular junction (Chee et al., unpublished). The failure of synaptic transmission can, in turn, be attributed to a loss of mitochondria at the NMJ. What these studies emphasize is that the degeneration and loss of peripheral neuronal compartments begins long before the classically measured output of neuron death is evident. This illustrates the principle that in order to understand why neurons die, we must understand the mechanisms that control the death of their diVerent peripheral compartments. Given that we know that the survival and loss of these compartments may be independent and distinct from the molecular events controlling PCD at the cell soma, it is clear that we need to understand more about this aspect of neuronal cell biology. In order to achieve this, Drosophila oVers an impressive array of tools with which to analyze the mechanisms and has great potential.

IV. Conclusions The classical view of PCD during development and disease has been that the loss of neurons begins in the cell body, with activation of the classic caspasedependent apoptosis machinery which initiates the destruction of the cell body and production of the classical morphological features. The death of the axons, dendrites, and synapses is regarded as a secondary consequence of the dying cell body (Fig. 1A). The fact that the loss of axons, dendrites, and synapses is now being shown to be controlled by specific and locally acting mechanisms that are independent of PCD at the cell body raises key questions about how the neurons die. The most obvious possibility is that death of the neuron can be triggered by events at the periphery, such that signaling events at the synapse could initiate withdrawal of the synapse and trigger a controlled and progressive death of the axon by either degeneration (Fig. 1B) or resorption (Fig. 1C). Thus, rather than being triggered by events in the cell body, the neuron could die from distal to proximal and might be coupled


163

Figure 1 Possible mechanisms of neuron death. (A) The classical view of PCD. (1) Neuron death begins in the cell body with the activation of the caspase-dependent apoptosis machinery, which initiates destruction of the cell body and production of the classical morphological features. (2) The death of the axons, dendrites, and synapses is a secondary consequence of the dying cell body. (B) and (C) Alternatively, neuron death could begin at the periphery. The loss of axons, dendrites, and synapses is controlled by specific and locally acting mechanisms that are independent of PCD at the cell body. This could occur in two ways: Neuron death is triggered by events at the periphery, such that signaling events at the synapse initiate withdrawal of the synapse and trigger a controlled and progressive death of the axon by either degeneration (B,2) or resorption (C,2). In both cases, the neuron soma ultimately undergoes PCD (3), possibly associated with a retrograde signal that is relayed to the nucleus to trigger PCD. (D) Activation of glia at the terminals precipitates the degeneration of the synapses followed by the same series of events in 1B or 1C.

with a retrograde signal that is relayed to the nucleus to trigger a PCD at the soma. Another mechanism might involve the activation of glia at the terminals, which then precipitate the degeneration of the synapses (Fig. 1D)

164

Shepherd and Perry

followed by a sequence of events similar to that described in Fig. 1B or 1C. In all those circumstances where degeneration of the processes is involved, this requires the removal of large amounts of cytoplasm. It has previously been recognized that degradation of large amounts of cytoplasm involves autophagic and proteasomal pathways (Lockshin and Zakeri, 2002). Studies in the mouse and Drosophila highlight the potential role of the UPS in the sculpting and removal of neuronal processes in development. Araki et al. (2004) have demonstrated that an increase in Nmnat activity is responsible for the axon protective activity of the Wlds protein. Furthermore, they demonstrated that SIRT1, a mammalian ortholog of Sir2, is the downstream eVector of the Nmnat mediated axon protection.

References Aberle, H., Haghighi, A. P., Fetter, R. D., McCabe, B. D., Magalhaes, T. R., and Goodman, C. S. (2002). Wishful thinking encodes a BMP type II receptor that regulates synaptic growth in Drosophila. Neuron 33, 545–558. Araki, T., Sasaki, Y., and Milbrandt, J. (2004). Increased nuclear NAD biosynthesis and SIRT1 activation prevent axonal degeneration. Science 305, 1010–1013. Arora, K., Oconnor, M. B., and Warrior, R. (1996). BMP signaling in Drosophila embryogenesis. Ann. NY Acad. Sci. 785, 80–97. Awasaki, T., and Ito, K. (2004). Engulfing action of glial cells is required for programmed axon pruning during Drosophila metamorphosis. Curr. Biol. 14, 668–677. Barth, M., Hirsch, H. V. B., Meinertzhagen, I. A., and Heisenberg, M. (1997). Experiencedependent developmental plasticity in the optic lobe of Drosophila melanogaster. J. Neurosci. 17, 1493–1504. Bernstein, M., and Lichtman, J. W. (1999). Axonal atrophy: The retraction reaction. Curr. Opin. Neurobiol. 9, 364–370. Blinzinger, K., and Kreutzberg, G. (1968). Displacement of synaptic terminals from regenerating motoneurons by microglial cells. Z. Zellforsch Mikrosk. Anat. 968, 145–157. Buckmaster, E. A., Perry, V. H., and Brown, M. C. (1995). The rate of Wallerian degeneration in cultured neurons from wild type and C57BL/WldS mice depends on time in culture and may be extended in the presence of elevated K+ levels. Eur. J. Neurosci. 7, 1596–1602. Burek, M. J., and Oppenheim, R. W. (1996). Programmed cell death in the developing nervous system. Brain Pathol. 6, 427–446. Burne, J. F., Staple, J. K., and RaV, M. C. (1996). Glial cells are increased proportionally in transgenic optic nerves with increased numbers of axons. J. Neurosci. 16, 2064–2073. Chang, Q., and Balice-Gordon, R. J. (2000). Highwire, rpm-1, and futsch: Balancing synaptic growth and stability. Neuron 26, 287–290. Chapman, P. F., White, G. L., Jones, M. W., Cooper-Blacketer, D., Marshall, V. J., Irizarry, M., Younkin, L., Good, M. A., Bliss, T. V., Hyman, B. T., Younkin, S. G., and Hsiao, K. K. (1999). Impaired synaptic plasticity and learning in aged amyloid precursor protein transgenic mice. Nat. Neurosci. 2, 271–276. Chiba, A., Shepherd, D., and Murphey, R. K. (1988). Synaptic rearrangement during postembryonic development in the cricket. Science 240, 901–906. Coleman, M. P., Conforti, L., Buckmaster, E. A., Tarlton, A., Ewing, R. M., Brown, M. C., Lyon, M. F., and Perry, V. H. (1998). An 85-kb tandem triplication in the slow Wallerian degeneration (Wlds) mouse. Proc. Natl. Acad. Sci. USA 18, 9985–9990.


165

Coleman, M. P., and Perry, V. H. (2002). Axon pathology in neurological disease: A neglected therapeutic target. Trends Neurosci. 25, 532–537. Cowan, C. M., Thai, J., Krajewski, S., Reed, J. C., Nicholson, D. W., Kaufmann, S. H., and Roskams, A. J. (2001). Caspases 3 and 9 send a pro-apoptotic signal from synapse to cell body in olfactory receptor neurons. J. Neurosci. 15, 7099–7109. Cunningham, C., Deacon, R., Wells, H., Boche, D., Waters, S., Diniz, C. P., Scott, H., Rawlins, J. N., and Perry, V. H. (2003). Synaptic changes characterize early behavioural signs in the ME7 model of murine prion disease. Eur. J. Neurosci. 17, 2147–2155. Deckwerth, T. L., and Johnson, E. M., Jr. (1994). Neurites can remain viable after destruction of the neuronal soma by programmed cell death (apoptosis). Dev. Biol. 165, 63–72. Dong, H., Fazzaro, A., Xiang, C., Korsmeyer, S. J., Jacquin, M. F., and McDonald, J. W. (2003). Enhanced oligodendrocyte survival after spinal cord injury in Bax-deficient mice and mice with delayed Wallerian degeneration. J. Neurosci. 24, 8682–8691. Eaton, B. A., Fetter, R. D., and Davis, G. W. (2002). Dynactin is necessary for synapse stabilization. Neuron 34, 729–741. Ehlers, M. D. (2003). Activity level controls postsynaptic composition and signaling via the ubiquitin-proteasome system. Nat. Neurosci. 6, 231–242. Finn, J. T., Weil, M., Archer, F., Siman, R., Srinivasan, A., and RaV, M. C. (2000). Evidence that Wallerian degeneration and localized axon degeneration induced by local neurotrophin deprivation do not involve caspases. J. Neurosci. 20, 1333–1341. Fischer, L. R., Culver, D. G., Tennant, P., Davis, A. A., Wang, M., Castellano-Sanchez, A., Khan, J., Polak, M. A., and Glass, J. D. (2004). Amyotrophic lateral sclerosis is a distal axonopathy: Evidence in mice and man. Exp. Neurol. 185, 232–240. Gillingwater, T. H., Ingham, C. A., Coleman, M. P., and Ribchester, R. R. (2003). Ultrastructural correlates of synapse withdrawal at axotomized neuromuscular junctions in mutant and transgenic mice expressing the Wld gene. J. Anat. 203, 265–276. Gillingwater, T. H., and Ribchester, R. R. (2001). Compartmental neurodegeneration and synaptic plasticity in the Wld(s) mutant mouse. J. Physiol. 1, 627–639. Gramates, L. S., and Budnik, V. (1999). Assembly and maturation of the Drosophila larval neuromuscular junction. ‘‘Neuromuscular Junctions in Drosophila.’’Int. Rev. Neurobiol. 43, 93–117. GriYn, J. W., George, E. B., and Chaudhry, V. (1996). Wallerian degeneration in peripheral nerve disease. Baillieres Clin. Neurol. 5, 65–75. Grueber, W. B., Jan, L. Y., and Jan, Y. N. (2003). DiVerent levels of the homeodomain protein cut regulate distinct dendrite branching patterns of Drosophila multidendritic neurons. Cell 112, 805–818. Grueber, W. B., Ye, B., Moore, A. W., Jan, L. Y., and Jan, Y. N. (2003). Dendrites of distinct classes of Drosophila sensory neurons show diVerent capacities for homotypic repulsion. Curr. Biol. 13, 618–626. Guerrini, R., Carrozzo, R., Rinaldi, R., and Bonanni, P. (2003). Angelman syndrome: Etiology, clinical features, diagnosis, and management of symptoms. Paediatr. Drugs 5, 647–661. Hanz, S., Perlson, E., Willis, D., Zheng, J. Q., Massarwa, R., Huerta, J. J., Koltzenburg, M., Kohler, M., van-Minnen, J., Twiss, J. L., and Fainzilber, M. (2003). Axoplasmic importins enable retrograde injury signaling in lesioned nerve. Neuron 18, 1095–1104. Innocenti, G. M., Clarke, S., and Koppel, H. (1983). Transitory macrophages in the white matter of the developing visual cortex. II. Development and relations with axonal pathways. Brain Res. 313, 55–66. Innocenti, G. M. (1995). Exuberant development of connections and its possible permissive role in cortical evolution. Trends Neurosci. 18, 397–402.

166

Shepherd and Perry

Laser, H., Mack, T. G. A., Wagner, D., and Coleman, M. P. (2003). Proteasome inhibition arrests neurite outgrowth and causes ‘‘dying-back’’ degeneration in primary culture. J. Neurosci. Res. 74, 906–916. Lewin, G. R., and Barde, Y. A. (1996). Physiology of the neurotrophins. Annu. Rev. Neurosci. 19, 289–317. Lockshin, R. A., and Zakeri, Z. (2002). Caspase-independent cell deaths. Curr. Opin. Cell Biol. 14, 727–733. Lunn, E. R., Perry, V. H., Brown, M. C., Rosen, H., and Gordon, S. (1989). Absence of Wallerian degeneration does not hinder regeneration in peripheral nerve. Eur. J. Neurosci. 1, 27–33. Mack, T. G., Reiner, M., Beirowski, B., Mi, W., Emanuelli, M., Wagner, D., Thomson, D., Gillingwater, T., Court, F., Conforti, L., Fernando, F. S., Tarlton, A., Andressen, C., Addicks, K., Magni, G., Ribchester, R. R., Perry, V. H., and Coleman, M. P. (2001). Wallerian degeneration of injured axons and synapses is delayed by a Ube4b/Nmnat chimeric gene. Nat. Neurosci. 4, 1199–1206. Marin-Teva, J. L., Dusart, I., Colin, C., Gervais, A., van Rooijen, N., and Mallat, M. (2004). Microglia promote the death of developing Purkinje cells. Neuron 19, 535–547. Marques, G., Bao, H., Haerry, T. E., Shimell, M. J., Duchek, P., Zhang, B., and O’Connor, M. B. (2002). The Drosophila BMP type II receptor wishful thinking regulates neuromuscular synapse morphology and function. Neuron 33, 529–543. Mattson, M. P. (2000). Apoptotic, and santri-apoptotic synaptic signaling mechanisms. Brain Path. 10, 300–312. McCabe, B. D., Hom, S., Aberle, H., Fetter, R. D., Marques, G., Haerry, T. E., Wan, H., O’Connor, M. B., Goodman, C. S., and Haghighi, A. P. (2004). Highwire regulates presynaptic BMP signaling essential for synaptic growth. Neuron 41, 891–905. McCabe, B. D., Marques, G., Haghighi, A. P., Fetter, R. D., Crotty, M. L., Haerry, T. E., Goodman, C. S., and O’Connor, M. B. (2003). The BMP homolog Gbb provides a retrograde signal that regulates synaptic growth at the Drosophila neuromuscular junction. Neuron 39, 241–254. Moore, A. W., Jan, L. Y., and Jan, Y. N. (2002). Hamlet, a binary genetic switch between single- and multiple-dendrite neuron morphology. Science 297, 1355–1358. Mudher, A., Shepherd, D., Newman, T. A., Mildren, P., Jukes, J. P., Squire, A., Mears, A., Berg, S., MacKay, D., Asuni, A. A., Bhat, R., and Lovestone, S. (2004). GSK-3 beta inhibition reverses axonal transport defects and behavioural phenotypes in Drosophila. Mol. Psych. 9, 522–530. Perry, V. H., and Linden, R. (1982). Evidence for dendritic competition in the developing retina. Nature 297, 683–685. Perry, V. H., Henderson, Z., and Linden, R. (1983). Postnatal changes in retinal ganglion cell and optic axon populations in the pigmented rat. J. Comp. Neurol. 20, 356–368. Perry, V. H., Hume, D. A., and Gordon, S. (1985). Immunohistochemical localization of macrophages and microglia in the adult and developing mouse brain. Neuroscience 15, 313–326. Perry, V. H., Brown, M. C., and Lunn, E. R. (1991). Very slow retrograde and Wallerian degeneration in the CNS of C57BL/Ola mice. Eur. J. Neurosci. 3, 102–105. Purves, D., and Lichtman, J. W. (1980). Elimination of synapses in the developing nervous system. Science 210, 153–157. RaV, M. C., Barres, B. A., Burne, J. F., Coles, H. S., Ishizaki, Y., and Jacobson, M. D. (1993). Programmed cell death and the control of cell survival: Lessons from the nervous system. Science 29, 695–700. Ribchester, R. R., Tsao, J. W., Barry, J. A., Asgari-Jirhandeh, N., Perry, V. H., and Brown, M. C. (1995). Persistence of neuromuscular junctions after axotomy in mice with slow Wallerian degeneration (C57BL/WldS). Eur. J. Neurosci. 1, 1641–1650.


167

Saigoh, K., Wang, Y. L., Suh, J. G., Yamanishi, T., Sakai, Y., Kiyosawa, H., Harada, T., Ichihara, N., Wakana, S., Kikuchi, T., and Wada, K. (1999). Intragenic deletion in the gene encoding ubiquitin carboxy-terminal hydrolase in gad mice. Nat. Genet. 23, 47–51. Sanes, J. R., and Lichtman, J. W. (1999). Development of the vertebrate neuromuscular junction. Annu. Rev. Neurosci. 22, 389–442. Schlaepfer, W. W., and Hasler, M. B. (1979). Characterization of the calcium-induced disruption of neurofilaments in rat peripheral nerve. Brain Res. 168, 299–309. Schuster, C. M., Davis, G. W., Fetter, R. D., and Goodman, C. S. (1996). Genetic dissection of structural and functional components of synaptic plasticity. I. Fasciclin II controls synaptic stabilization and growth. Neuron 17, 641–654. Shepherd, D., and Murphey, R. K. (1986). Competition regulates the eYcacy of an identified synapse in crickets. J. Neurosci. 6, 3152–3160. Shepherd, D., and Laurent, G. (1992). Embryonic development of a population of spiking local interneurones in the locust. J. Comp. Neurol. 319, 438–453. Shimura, H., Hattori, N., Kubo, S., Mizuno, Y., Asakawa, S., Minoshima, S., Shimizu, N., Iwai, K., Chiba, T., Tanaka, K., and Suzuki, T. (2000). Familial Parkinson disease gene product, parkin, is a ubiquitin-protein ligase. Nat. Genet. 25, 302–305. Shulman, J. M., Shulman, L. M., Weiner, W. J., and Feany, M. B. (2003). From fruit fly to bedside: Translating lessons from Drosophila models of neurodegenerative disease. Curr. Opin. Neurol. 16, 443–449. Sievers, C., Platt, N., Perry, V. H., Coleman, M. P., and Conforti, L. (2003). Neurites undergoing Wallerian degeneration show an apoptotic-like process with Annexin V positive staining and loss of mitochondrial membrane potential. Neurosci. Res. 46, 161–169. Sugimura, K., Yamamoto, M., Niwa, R., Satoh, D., Goto, S., Taniguchi, M., Hayashi, S., and Uemura, T. (2003). Distinct developmental modes and lesion-induced reactions of dendrites of two classes of Drosophila sensory neurons. J. Neurosci. 23, 3752–3760. Tissot, M., and Stocker, R. F. (2000). Metamorphosis in Drosophila and other insects: The fate of neurons throughout the stages. Prog. Neurobiol. 62, 89–111. Wan, H. I., DiAntonio, A., Fetter, R. D., Bergstrom, K., Strauss, R., and Goodman, C. S. (2000). Highwire regulates synaptic growth in Drosophila. Neuron 26, 313–329. Watts, R. J., Hoopfer, E. D., and Luo, L. Q. (2003). Axon pruning during Drosophila metamorphosis: Evidence for local degeneration and requirement of the ubiquitinproteasome system. Neuron 38, 871–885. Watts, R. J., Schuldiner, O., Perrino, J., Larsen, C., and Luo, L. Q. (2004). Glia engulf degenerating axons during developmental axon pruning. Curr. Biol. 14, 678–684. Williams, D. W., and Truman, J. W. (2004). Mechanisms of dendritic elaboration of sensory neurons in Drosophila: Insights from in vivo time lapse. J. Neurosci. 24, 1541–1550. Williams, D. W., Tyrer, N. M., and Shepherd, D. (2000). Tau and tau reporters disrupt central projections of sensory neurons in Drosophila. J. Comp. Neurol. 428, 630–640. Yuan, J., and Yankner, B. A. (2000). Apoptosis in the nervous system. Nature 12, 802–829. Zhai, Q., Wang, J., Kim, A., Liu, Q., Watts, R., Hoopfer, E., Mitchison, T., Luo, L., and He, Z. (2003). Involvement of the ubiquitin-proteasome system in the early stages of wallerian degeneration. Neuron 39, 217–225. Zwijsen, A., Verschueren, K., and Huylebroeck, D. (2003). New intracellular components of bone morphogenetic protein/Smad signaling cascades. FEBS Lett. 546, 133–139.

6

____________________________________________________________________________

Multiple Roles of Vascular Endothelial Growth Factor (VEGF) in Skeletal Development, Growth, and Repair Elazar Zelzer * and Bjorn R. Olsen{ *Department of Molecular Genetics Weizmann Institute of Science Rehovot 76100, Israel { Harvard Medical School Department of Cell Biology Boston, Massachusetts 02115

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

Introduction The Role of VEGF in Regulating Vascularization of Developing Bones VEGF Regulates Osteoclast Activity VEGF is a Key Component of a Chondrocyte Survival Pathway A Role for VEGF in Control of Osteoblastic Activity VEGF is Involved in Bone Repair Future Directions and Questions Acknowledgments References

Overview Studies of bone morphogenesis have identified a large number of critical molecules and regulatory pathways. One of these molecules is vascular endothelial growth factor, VEGF. Several studies suggest that not only is this regulator of angiogenesis important in mediating interactions between the developing bone and the vasculature, but it also has a key role in regulating processes during bone development and growth which are not directly related to angiogenesis. Studies of the detailed mechanisms by which VEGF is involved in bone development promise therefore to shed new light on key steps in bone formation. This chapter summarizes findings with emphasis on the developmental roles of VEGF in skeletal morphogenesis. Furthermore, we speculate on the future directions of research in this area and describe some of the challenges in the field. Current Topics in Developmental Biology, Vol. 65 Copyright 2005, Elsevier Inc. All rights reserved. 0070-2153/05 $35.00

169

170

Zelzer and Olsen

I. Introduction In vertebrates, the skeleton is formed by mesenchymal cells that are derived from cranial neural crest, somites, and lateral plate mesoderm. At the sites of future bones, these cells condense and form the future skeletal elements. In the cranial vault, jaws, and part of the clavicle, the condensed mesenchymal cells diVerentiate into osteoblasts and generate bone in a process termed ‘‘intramembranous bone formation.’’ The remainder of the future skeleton develops by endochondral bone formation. During endochondral bone formation, the condensing mesenchymal cells diVerentiate into chondrocytes and form avascular cartilage models of the future bones (see Fig. 1). As development proceeds, chondrocytes in the centers of the cartilage models cease to proliferate and the post-mitotic cells diVerentiate to hypertrophy. The diVerentiation of chondrocytes to hypertrophy is followed by rapid invasion of blood vessels, osteoclasts, and other mesenchymal cells from the surrounding tissue (perichondrium) into the cartilage, which is progressively eroded and replaced by bone marrow and trabecular bone in the

Figure 1 Diagram of the developmental process leading to the formation of the mouse tibia. At early stages, diVerentiation of chondrocytes from mesenchymal cells results in the formation of an avascular cartilage model of the future bone. At ED14, chondrocytes in the center of the cartilage model mature to hypertrophy. One day later, the hypertrophic cartilage begins to be invaded by sprouting blood vessels, osteoclastic cells, and hematopoietic precursors; in the perichondrium, a bone collar is forming. At later stages, the marrow cavity expands by continued erosion of the hypertrophic cartilage in the growth plates at both ends (epiphyses) of the growing bone. As the hypertrophic cartilage is removed, it is replaced by trabecular bone in the regions beneath the growth plate cartilage (modified from Horton, 1990).

6. Multiple Roles of VEGF

171

primary ossification center (Karsenty and Wagner, 2002; Kronenberg, 2003; Olsen et al., 2000). The association of angiogenesis with endochondral bone formation was the trigger for initial studies of a possible involvement of VEGF in bone vascularization. As has happened many times in research, experiments aimed at answering one question opened the door to many unanticipated discoveries. In the case of VEGF, data from several laboratories now suggest that VEGF, in addition to being required for bone vascularization, has a key role in several other steps of skeletal development, including chondrocyte diVerentiation and survival, osteoclast recruitment, and osteoblast diVerentiation.

II. The Role of VEGF in Regulating Vascularization of Developing Bones Vessel invasion into the primary ossification center and continued capillary sprouting as the center expands and growth plates are formed at both ends (epiphyses) are key steps in endochondral bone formation. Three papers by Trueta et al. (Trueta and Amato, 1960; Trueta and Buhr, 1963; Trueta and Trias, 1961) about 40 years ago firmly established the concept of a coupling between cartilage vascularization and endochondral bone formation. Although these studies were conducted with 6-week-old rabbits, the conclusions of the studies also apply to embryonic bone development. By interrupting the blood supply to the growth plate, Trueta and his colleagues observed a decrease in bone mineralization and an expansion of the hypertrophic zone in the growth plate. These findings led them to conclude that blood vessels are not only supplying oxygen and nutrition to forming (growing) bones, but play an active role in bone formation. Forty years later, Gerber et al. used injection of a soluble VEGF receptor to inhibit VEGF activity in 24-day-old mice and observed impaired angiogenesis, decreased trabecular bone formation, and expansion of the hypertrophic zone in growth plates (Gerber et al., 1999). These findings were similar to those reported by Trueta et al. and represented a first step toward a detailed understanding of the mechanism that couples cartilage vascularization and endochondral bone formation. Vessel invasion into the primary ossification center of developing bones (at E14.5 in the mouse tibia) is preceded by recruitment of vessels to the surrounding perichondrium. At the time of this perichondrial recruitment (in the mouse tibia, this happens at E13.5–E14.5), VEGF is expressed by the perichondrial cells. This suggests that VEGF might be involved in the recruitment of blood vessels to the perichondrium. Support for this hypothesis has come from studies of bone development in mice that only express

172

Zelzer and Olsen

one isoform of VEGF, VEGF120 (Zelzer et al., 2002). The murine VEGF gene encodes three isoforms: VEGF120, VEGF164, and VEGF188; all are products of alternative splicing of a single gene (Ferrara et al., 1992; Shima et al., 1996). In contrast to the other two isoforms, VEGF120 does not bind the extracellular matrix component heparan sulfate (Ferrara and DavisSmyth, 1997; Park et al., 1993). Neuropilin-1 (NRP1) and neuropilin-2 (NRP2) are co-receptors for VEGF164 and can potentiate signaling through the VEGF receptor VEGFR2 (Flk-1), but they do not bind the VEGF120 isoform (Soker et al., 1998). In VEGF120 mice, the recruitment of vessels into the perichondrium is delayed. This delay suggests that the function of VEGF expressed in the perichondrium at E13.5 is to stimulate perichondrial angiogenesis (Zelzer et al., 2002). Soon after perichondrial angiogenesis, vessels invade the hypertrophic cartilage from the perichondrium (Colnot et al., 2004), and the primary ossification center is established. This is preceded by upregulation of VEGF expression in the hypertrophic cartilage (Carlevaro et al., 2000; Colnot and Helms, 2001; Zelzer et al., 2001). As development proceeds and capillaries continue to invade the hypertrophic cartilage of the growth plate (see preceding text), VEGF expression in the hypertropic zone is maintained (Gerber et al., 1999). Vessel invasion into the primary ossification center is severely delayed in both VEGF120 mice and mice in which VEGF expression in chondrocytes is abolished. Vessel sprouting within the metaphyses is reduced as well. As a result of the reduction in vessel sprouting in the metaphysis, the erosion of the cartilage growth plate is reduced, and terminally diVerentiated hypertrophic chondrocytes accumulate (Haigh et al., 2000; Maes et al., 2002; Zelzer et al., 2002, 2004). Expansion of the zone of hypertrophic chondrocytes in the growth plate is also observed following targeted inactivation of the genes encoding matrix metalloprotease-9 (MMP-9), the transcription factor Runx2, and connective tissue growth factor (Ctgf ). Interestingly, all three genes have been reported to aVect VEGF activity in the growth plate, although the suggested mechanisms by which they act are diVerent (Ivkovic et al., 2003; Vu et al., 1998; Zelzer et al., 2001). MMP-9 is a key molecule in bone formation (see Ortega et al., 2004). As has been mentioned, the growth plates of MMP-9 null mice have expanded hypertrophic zones. Moreover, MMP-9 null growth plates show a reduction in vascularization and ossification. These features are also seen in mice where VEGF activity is blocked. MMP-9, expressed by osteo(chondro)clasts in the growth plate, has been suggested to play a role in extracellular matrix degradation during vessel invasion. Since VEGF expression in MMP-9 null growth plates appears to be normal, it is not clear how the functions of these two genes are mechanistically connected (Blavier and Delaisse, 1995; Engsig et al., 2000; Vu et al., 1998). Possible mechanisms


173

Figure 2 Diagram indicating two major functions of VEGF during endochondral ossification. In hypertrophic chondrocytes, a high-level expression of VEGF, controlled by the transcription factor Runx2, is essential for vascularization of cartilage in the primary ossification center and continued capillary sprouting under the growth plates as the bone grows. Since the level of VEGF expression is reduced in Ctgf null growth plates, it is possible that Ctgf is a component of a pathway by which Runx2 stimulates VEGF expression. In epiphyseal chondrocytes, a moderate level of VEGF expression, controlled by HIF-1 and VHL, is necessary for chondrocyte survival.

will be discussed later when the role of VEGF in osteoclast biology will be described. Runx2, a member of the runt-domain transcription factor family, has an essential role in osteoblast diVerentiation (Ducy et al., 1997). Runx2 null mice have an almost perfectly patterned skeleton composed of cartilage, but no bones (Komori et al., 1997; Otto et al., 1997). In addition, chondrocyte diVerentiation to hypertrophy is impaired, except in the distal appendicular skeleton (tibia–fibula, radius–ulna), where chondrocytes do diVerentiate to hypertrophy. Moreover, no blood vessel invasion into hypertrophic cartilage is apparent in Runx2 null mice (Inada et al., 1999; Kim et al., 1999). The lack of blood vessel invasion into hypertrophic cartilage in Runx2deficient mice suggests a link between Runx2 and VEGF expression. Indeed, the upregulated expression of VEGF in hypertrophic zones in wild-type mice is absent in the hypertrophic cartilage of Runx2-deficient mice. This finding suggests that Runx2 is involved in the regulation of VEGF in hypertrophic chondrocytes (see Fig. 2). Whether Runx2 acts as a direct transcriptional regulator of the VEGF gene is not clear; however, the VEGF promoter contains Runx2 binding sites and overexpression of Runx2 in fibroblasts, transfected with a VEGF promoter-luciferase construct, results in induced expression of the luciferase reporter (Zelzer et al., 2001). The finding that Runx2, an essential regulator of osteoblastic diVerentiation, also regulates VEGF expression in hypertrophic cartilage, demonstrates the existence of a

174

Zelzer and Olsen

Figure 3 Diagram of sprouting capillaries under the zone of hypertrophic chondrocytes in the growth plate of an endochondral bone. Synthesis of VEGF by hypertrophic chondrocytes is likely to generate a concentration gradient of VEGF (indicated by a decrease in font size away from the chondrocytes) that serves to direct the migration of VEGF receptor-expressing endothelial cells. As described in the text, capillary sprouts close to the hypertrophic cells contain no basement membrane or pericytes; further away from the cartilage, basement membranes and pericytes are surrounding the endothelial tubes.

tissue-specific genetic program that couples osteogenesis and vessel invasion into cartilage during endochondral bone formation. Connective tissue growth factor (Ctgf ), a member of the CCN family, is an important regulator of extracellular matrix production (Perbal, 2004). Gene-targeting experiments also suggest a role in bone development. Endochondral Ctgf null bones show a decrease in chondrocyte proliferation and expression of extracellular matrix molecules. Moreover, the hypertrophic zone of growth plates in Ctgf null bones is expanded, suggesting a reduction in angiogenesis and matrix erosion. Expression of VEGF in Ctgf null growth plates is also decreased. This decrease is not the result of reduced Runx2 expression, suggesting that Ctgf is either downstream of or acts in parallel with Runx2 in regulating VEGF expression (Fig. 2) (Ivkovic et al., 2003). The upregulated expression of VEGF in hypertrophic chondrocytes results in the sprouting of endothelial cells in perichondrial blood vessels. Studies of the expression of VEGF receptors in these endothelial cells have


175

identified a potential mechanism for enhancing and refining the invasion of vessels into the cartilage and the continued capillary growth under the growth plates. The expression of VEGF receptors VEGFR1 (Flt-1) and VEGFR2 (Flk-1) in the perichondrial endothelium is upregulated as a result of VEGF expression in the hypertrophic cartilage (Colnot and Helms, 2001; Zelzer et al., 2001, 2002). This upregulation is maintained through all stages of bone development in the sprouting vessels under the growth plates (Gerber et al., 1999). A reduction in VEGF expression or activity in hypertrophic cartilage results in reduced VEGF receptor expression in perichondrial endothelial cells. This suggests that invasion of vessels into hypertrophic cartilage involves an active cross-talk between hypertrophic chondrocytes and endothelial cells (Fig. 3). Interestingly, sprouting growth plate capillaries contain no basement membranes or pericytes (Hunter and Arsenault, 1990). Further away from the hypertrophic cartilage, in the region where bone matrix is being synthesized, endothelial cells are surrounded by basement membranes and pericytes. It is tempting to speculate that high levels of VEGF signaling in the vicinity of hypertrophic chondrocytes maintain endothelial cells in a sprouting mode by inducing the expression of VEGF receptors and inhibiting the formation of basement membranes and pericyte recruitment (Fig. 3).

III. VEGF Regulates Osteoclast Activity Vessel invasion into cartilage is a complex process involving the coordinated activities of both endothelial and osteo(chondro)clastic cells (referred to subsequently as osteoclasts). In addition to controlling endothelial cell activities, VEGF also regulates osteoclastic diVerentiation, migration, and activity. VEGF is, therefore, a key coordinator of the entire process. Osteoclasts, derived from monocytes, play an important role during cartilage vascularization (Tondravi et al., 1997). The discovery that monocytes express VEGFR1 (Flt-1) (Barleon et al., 1997) clearly raised the possibility that VEGF signaling may aVect osteoclasts. Indeed, Niida et al. (1999) demonstrated that VEGF can stimulate osteoclastic bone resorption in vivo. Moreover, VEGF injection into osteopetrotic op/op mice (lacking the cytokine M-CSF) rescued the osteopetrotic phenotype by inducing recruitment and survival of osteoclasts. Bone marrow cell culture experiments demonstrate that VEGF can enhance osteoclastic activity by stimulating the diVerentiation of osteoclasts from monocytic precursor cells. In the process of diVerentiation of osteoclasts from monocytes, VEGF can substitute for M-CSF as a co-stimulator (with RANKL) (Niida et al., 1999). Furthermore, studies of bone resorption by mature osteoclasts suggest that VEGF is involved not only in osteoclastic recruitment and diVerentiation

176

Zelzer and Olsen

but also in enhancing osteoclastic bone resorbing activity (Nakagawa et al., 2000). A possible pathway for regulating migration of osteoclasts by VEGF involves the extracellular signal-regulated kinases 1 and 2 (ERK1/2) (Henriksen et al., 2003). VEGF is necessary for osteoclastic activity both at the stage when the primary ossification center is established and later during bone growth. In VEGF120 mice, the number of osteoclasts is dramatically reduced in the perichondrium of developing bones at sites where vessels will invade the hypertrophic cartilage (Maes et al., 2002; Zelzer et al., 2002). Injection of a soluble chimeric VEGF receptor Flt-IgG protein into the circulation of mice blocks the recruitment of osteoclasts to the growth plates (Gerber et al., 1999). Addition of chimeric VEGF receptor protein to metatarsal organ cultures blocks recruitment of osteoclasts to the primary ossification center, whereas addition of VEGF induces osteoclast migration (Engsig et al., 2000). As has been mentioned, it has been reported that a reduced expression of the osteoclastic protease MMP-9 aVects angiogenesis-inducing activity in growth plates (Vu et al., 1998). MMP-9 null growth plates share several features with growth plates of mice in which VEGF activities are blocked, such as expansion of the hypertrophic zone and reduction in both vascularization and ossification (Vu et al., 1998). Since VEGF expression in hypertrophic chondrocytes of MMP-9 null growth plates is normal (Engsig et al., 2000), it has been suggested that MMP-9 regulates the availability of VEGF in the hypertrophic zone, perhaps by degrading extracellular matrix components and releasing VEGF from the matrix (Ortega et al., 2004). In our view, the simplest interpretation of the available data is that VEGF, produced by hypertrophic chondrocytes, induces osteoclastogenesis in the perichondrium and stimulates migration of the osteoclasts into hypertrophic cartilage by activating the appropriate migration-inducing signaling pathways. Osteoclasts are the major sources of MMP-9 and they use this protease to tunnel through the hypertrophic cartilage matrix (Blavier and Delaisse, 1995; Reponen et al., 1994; Vu et al., 1998). When MMP-9 expression is eliminated, as in MMP-9 null mice, osteoclast penetration into the cartilage is reduced, resulting in reduced angiogenesis. When VEGF signaling is reduced or blocked, osteoclast diVerentiation and invasion, as well as the level of the associated MMP-9, is reduced.

IV. VEGF is a Key Component of a Chondrocyte Survival Pathway A number of studies have led to the identification of VEGF as a critical factor for survival of chondrocytes (Maes et al., 2004; Zelzer et al., 2004). One study discovered that the epiphyseal regions of some long bones in mice


177

expressing only the VEGF188 isoform contain areas of chondrocyte cell death. The cell death is first observed at E18.5 and becomes prominent at P5 (Maes et al., 2004). Since the VEGF188 isoform in these mice is ubiquitously expressed, the skeletal abnormalities are not necessarily caused by VEGF188 expression in skeletal tissues but may be indirect consequences of changes in surrounding nonskeletal tissues. However, blocking VEGF expression in chondrocytes confirms the importance of VEGF expression for chondrocyte survival. In VEGF null bones, regions of cell death are located in the central regions of skeletal elements at E16.5, starting at an articular surface, continuing through the resting to the proliferating zones of growth plate chondrocytes, and ending in a misshapen growth plate (Zelzer et al., 2004). A similar phenotype has been described in epiphyseal cartilage of mice in which the transcription factor HIF-1 is conditionally inactivated in chondrocytes (Schipani et al., 2001). This suggests that VEGF and HIF-1 are part of a chondrocyte survival pathway. HIF-1, in which HIF-1 is a subunit, regulates the transcription of a broad range of genes, including VEGF, that are involved in a variety of processes such as glucose metabolism, angiogenesis, and cell survival. The cellular level of HIF-1 protein is tightly regulated by the tumor-suppressor von Hippel-Lindau (VHL) protein, whereas the second subunit of HIF-1, HIF-1 , is constitutively expressed. Several factors, such as hypoxia, hormones, and growth factors, are known to induce stabilization of HIF-1 protein and, therefore, the HIF-1 heterodimeric transcription complex (Pugh and RatcliVe, 2003; Semenza, 2003). The existence of a chondrocyte survival pathway in the epiphysis of developing bones raises several questions regarding its regulation and the nature of the cellular survival mechanisms involved. VEGF shows a complex pattern of expression during bone development. A robust expression of VEGF is induced as chondrocytes diVerentiate to hypertrophy within cartilage models of the future bones (Carlevaro et al., 2000; Colnot and Helms, 2001; Zelzer et al., 2001). Later in development, as growth plates are established, a more moderate expression of VEGF is also detected in epiphyseal chondrocytes (Maes et al., 2004; Schipani et al., 2001; Zelzer et al., 2004). In the absence of VEGF expression, these epiphyseal chondrocytes undergo apoptosis. Interestingly, while VEGF expression in hypertrophic chondrocytes does not appear to be HIF-1 dependent, the moderate (and later) VEGF expression in epiphyseal cells depends on HIF-1 (Schipani et al., 2001; Zelzer et al., 2004). Thus, VEGF appears to be downstream of HIF-1 in the chondrocyte survival pathway (Fig. 2). VHL, a component of a ubiquitin ligase that regulates HIF-1 degradation, is likely a component of the pathway (Bruick and McKnight, 2001; Ivan et al., 2001; Jaakkola et al., 2001; Masson et al., 2001; Yu et al., 2001). Loss of VHL function in chondrocytes increases HIF-1 stabilization and leads to its accumulation. As a consequence of

178

Zelzer and Olsen

HIF-1 accumulation, the expression of HIF-1 target genes, including VEGF, are upregulated in the growth plate (Fig. 2) (Pfander et al., 2004). Understanding the detailed reasons for the time diVerence (E14.5 versus E16.5 in the distal femur/proximal tibia in the mouse) in the initial appearance of cell death in the HIF-1 and the VEGF conditional knockouts may shed light on some mechanistic aspects of the survival pathway. Several potential explanations for this time diVerence are possible. One possibility is that chondrocyte survival depends on diVerent sets of genes at diVerent stages of development. HIF-1 is known to regulate the expression of genes that have a role in cell survival (Pugh and RatcliVe, 2003; Semenza, 2003). At early stages of development, when VEGF is not expressed by epiphyseal chondrocytes, one set of genes may be needed for cellular survival; later, when developmental needs are diVerent, a diVerent set, including VEGF, may be necessary. A second possibility is that chondrocytes, in response to increased need for oxygen and nutrients as the cartilage grows, initially make adjustments to increased metabolic stress by activating pathways that are controlled by HIF-1 without the need for VEGF. Examples of such adjustments could be an increase in glycolysis and increased levels of glucose transporters in the cell membrane. As growth continues, and the need for oxygen and nutrients increases as well, a better blood supply may become necessary. VEGF could then play a key role in regulating an angiogenic response. Such a two-step mechanism could explain the time diVerence in the initial appearance of cell death between the HIF-1 and the VEGF conditional knockouts. A third possibility is that the induction of expression of HIF-1 and its target genes is VEGF-independent, but that VEGF is required for maintaining expression of HIF-1 and/or its target genes. Elucidation of these possibilities will require definitive identification of the signals that activate the chondrocyte survival pathway and insights into the nature of the downstream cellular processes that the pathway controls. The nonvascular environment in epiphyseal cartilage and the essential roles of HIF-1 and VEGF in hypoxic response mechanisms in many other tissues have led some investigators to propose that hypoxic conditions may activate the pathway. In fact, several lines of evidence favor this possibility, but there is also evidence to support the conclusion that hypoxia is not the main regulator of the pathway in chondrocytes. Among indications that hypoxia might be involved is the detection of EF-5 as a marker for bioreductive activity in the epiphyseal cartilage (Schipani et al., 2001). Further support comes from a study of mice that only express the VEGF188 isoform. In these mice, the number of blood vessels surrounding epiphyseal regions appears to be reduced, and the EF-5 signal is more broadly distributed within epiphyseal cartilage than in wild-type littermates (Maes et al., 2004). This suggests that the epiphyseal chondrocyte cell death in VEGF188 mice may well be caused by hypoxia. Moreover, chondrocytes


179

cultured under hypoxic conditions upregulate VEGF expression via HIF-1 (Pfander et al., 2003). Good arguments can also be made against hypoxia as a major regulator. The strongest argument is that chondrocytes in culture do survive under hypoxic conditions. Moreover, elimination of the expression of both VEGF and HIF-1 in such cultures does not aVect cell survival (Pfander et al., 2003; Zelzer, unpublished results). It is possible, therefore, that the HIF-1/ VEGF pathway in epiphyseal chondrocytes is an essential component of a genetic program induced by an as yet unknown regulator, turned on to support chondrocyte survival in a specific area of cartilage and at a specific developmental time. Although HIF-1 is best known as a key component in mechanisms that mediate cell responses to hypoxic stress, a growing body of evidence suggests that HIF-1 is a key factor in mediating the induction of many genes, including VEGF, in response to hormones and growth factors (Zelzer et al., 1998). The nature of the cellular events regulated by the HIF-1/VEGF pathway in epiphyseal chondrocytes is largely unknown. The reduced numbers of blood vessels surrounding epiphyseal cartilage in the VEGF188 mice (Maes et al., 2004) indicate that VEGF may aVect peri-epiphyseal angiogenesis. However, this local reduction in vessel numbers around epiphyses could also be the consequence of abnormalities in the tissue surrounding the cartilage since the expression of the VEGF188 isoform is not restricted to chondrocytes in these mice. Histomorphometric examination of vessels surrounding epiphyseal cartilage in mice with a targeted deletion of VEGF in chondrocytes should help clarify this tissue. It is also possible that VEGF has a direct eVect on chondrocytes. In VEGF188 limbs cultured under hypoxic stress, some chondrocytes in the resting zone of growth plates became apoptotic. Exogenous VEGF164 was able to rescue this cell death (Maes et al., 2004). However, to firmly establish whether VEGF has a direct eVect on chondrocytes and to understand the mechanisms of action of the HIF-1/VEGF pathway, it will be essential to identify the receptors that mediate the VEGF eVects on chondrocytes. The finding of NRP1 and NRP2 expression in epiphyseal chondrocytes makes these receptors potential candidates for mediating actions of VEGF in chondrocytes (Maes et al., 2004; Zelzer et al., 2004).

V. A Role for VEGF in Control of Osteoblastic Activity Several factors with important roles in regulating bone formation also induce the expression of VEGF by osteoblasts. Prostaglandins E1 and E2, BMP-4, BMP-6, BMP-7, FGF-2, TGF- , endothelin-1, IGF-1, and vitamin D3 can all induce VEGF expression in osteoblasts by activating a variety of signaling pathways (Akeno et al., 2002; Deckers et al., 2002; Harada et al.,

180

Zelzer and Olsen

1995; Kozawa et al., 2000, 2001; Saadeh et al., 1999; Tokuda et al., 2001, 2003a,b,c,d; Wang et al., 1996, 2002; Yeh and Lee, 1999). This raises the possibility that VEGF itself may be involved in regulating osteoblastic activity. VEGF could play a role in osteoblast biology in several diVerent ways. First, VEGF, expressed by osteoblasts, could couple angiogenesis to bone formation by adjusting the angiogenic response to osteoblastic activity. Second, VEGF could serve as a messenger in bilateral regulation. By expressing VEGF, osteoblasts could induce cells in the vicinity to express factor(s) that, in turn, regulate osteoblastic activity. Third, VEGF could be an autocrine regulator of osteoblastic diVerentiation and activity. There is current evidence in support of all three possibilities. The first study describing a role for VEGF in coupling angiogenesis to bone formation was the study of Gerber et al. (1999) in which inhibition of VEGF in 24-day-old mice resulted in impaired angiogenesis and impaired trabecular bone formation. Our study of skeletogenesis in VEGF120 mice extended this to embryonic bone development (Zelzer et al., 2002). This study also demonstrated that VEGF has a direct eVect on osteoblastic activity. The ossification of membranous bones was reduced and osteoblastic diVerentiation was altered in VEGF120 mice. Since membranous bones are formed directly in mesenchyme without cartilage intermediates and without the need for invasion of blood vessels into cartilage, as in endochondral ossification, the identification of VEGF eVects on osteoblastic activities in membranous bones of VEGF120 mice permits a distinction to be made between direct and indirect eVects. Consistent with direct eVects, VEGF is initially expressed in craniofacial mesenchyme at E13.5; at E14.5, the expression gets stronger and more restricted to the region where mesenchymal cells are diVerentiating into osteoblasts. Further support for a regulatory role of VEGF in osteoblastic diVerentiation has been obtained by cell culture and bone explant studies. Human VEGF165 protein binds to osteoblasts in culture and is capable of inducing migration and alkaline phosphatase activity (Midy and Plouet, 1994). Moreover, exogenous VEGF stimulates mineralization of osteoblast cultures. Adding VEGF or its neutralizing receptor to the medium of calvarial explant cultures enhances or blocks bone growth (Zelzer et al., 2002). Finally, VEGF receptors VEGFR-1, VEGFR-2, and neuropilin are all expressed by osteoblasts (Harper et al., 2001; Street et al., 2002).

VI. VEGF is Involved in Bone Repair Bone fractures can heal in two diVerent ways, similar to the two ways of forming bone during embryonic development. Stabilized fractures will heal by intramembranous ossification; unstable fractures undergo endochondral


181

ossification. The similarity between the embryonic bone development and repair of fractured bones, coupled with the finding that VEGF is expressed at sites of bone fracture, suggests that VEGF is involved in bone repair as it is in bone development (Ferguson et al., 1999; Le et al., 2001; Probst and Spiegel, 1997; Vortkamp et al., 1998). A study by Street et al. (2001) demonstrates that VEGF is indeed a key regulator of bone repair. Inhibition of VEGF by administration of a soluble chimeric VEGF receptor Flt-IgG protein to mice during fracture repair impairs bone regeneration. Blocking VEGF aVects cartilage formation (soft callus) and its replacement by bone (hard callus). Moreover, the healing of cortical bone defects is also aVected. Whereas the blocking of endogenous VEGF attenuates bone repair, exogenous VEGF accelerates bone repair. The role of VEGF in the healing process involves regulation of the vascularity of the healing fracture as well as a direct eVect of VEGF on osteoblastic diVerentiation and activity. As the fractured site is believed to be under hypoxic tension, the finding that osteoblastic cultures under hypoxic conditions produce VEGF suggests the possibility of an autocrine loop that regulates osteoblastic activity at the repair site. Another possible source of VEGF at the fracture site is the soft callus and fracture hematoma (Street et al., 2001). Support for the role of VEGF in the process of soft callus ossification comes from a study by Colnot et al. (2003). This study also underscores the similarity between the genetic control of bone development and regeneration following bone fracture. During endochondral ossification of MMP-9 null bones, there is a reduction in metaphyseal angiogenesis, the zone of hypertrophic chondrocytes is expanded, and there is decreased ossification. In the nonstabilized model for bone fracture, MMP-9 null bones exhibit a delay in hypertrophic callus remodeling and bone formation (Colnot et al., 2003); exogenous VEGF can rescue this delay.

VII. Future Directions and Questions Based on studies reviewed here, it is clear that VEGF is a major regulator of bone morphogenesis. In retrospect, the connection between VEGF and angiogenesis during endochondral bone formation may not seem very surprising based on the well-documented involvement of VEGF in angiogenesis generally, but the finding that VEGF is critically important for several other aspects of bone morphogenesis is surprising and exciting. In this chapter, we have described evidence for a role of VEGF in a variety of cellular processes in skeletal tissues, including migration, diVerentiation, and survival (Fig. 4).

182

Zelzer and Olsen

Figure 4 Three major functions of VEGF during bone development: it supports chondrocyte survival in epiphyseal cartilage (controlled by HIF-1); it induces vascularization of hypertrophic cartilage and invasion of osteoclasts (controlled by Runx2); it stimulates osteoblast diVerentiation and activity (controlled by unknown mechanism).

Most of the current understanding of the diVerent roles of VEGF in bone development is derived from experiments in which VEGF activity was manipulated. Since VEGF is a secreted cytokine that binds to diVerent receptors on various target cells, it is diYcult to extrapolate from these experiments to the precise mechanism of action by which VEGF regulates cellular processes in skeletal tissues. More specifically, most of the current data do not allow a distinction to be made between autonomous and nonautonomous cell functions of VEGF. The best way to address this problem in future studies will be to identify and manipulate the activity of the responsible receptors that mediate the eVects of VEGF on the target cells. This approach will help identify both the target cells and the cellular events regulated by VEGF in these cells, and should help provide answers to four important questions. How does VEGF orchestrate the processes of endothelial and osteoclastic cell migration into hypertrophic cartilage during endochondral ossification? What are the conditions/signaling events that induce the chondrocyte survival pathway, and what are the cellular responses in chondrocytes controlled by this pathway? At what timepoint in osteoblastic diVerentiation is VEGF critical, and what are the cellular activities promoted by VEGF in osteoblasts? To what extent do the diVerent isoforms of VEGF contribute to the diVerent receptor-mediated responses in target cells?

Acknowledgments We thank E. Schipani for helpful discussion and J. Helms and Z. Werb for comments. Y. Pittel provided administrative assistance. Original studies from the authors’ laboratory were supported by NIH grants AR36819 and AR36820.


183

References Akeno, N., Robins, J., Zhang, M., Czyzyk-Krzeska, M. F., and Clemens, T. L. (2002). Induction of vascular endothelial growth factor by IGF-I in osteoblast-like cells is mediated by the PI3K signaling pathway through the hypoxia-inducible factor-2alpha. Endocrinology 143, 420–425. Barleon, B., Siemeister, G., Martiny-Baron, G., Weindel, K., Herzog, C., and Marme, D. (1997). Vascular endothelial growth factor up-regulates its receptor fms-like tyrosine kinase 1 (FLT-1) and a soluble variant of FLT-1 in human vascular endothelial cells. Cancer Res. 57, 5421–5425. Blavier, L., and Delaisse, J. M. (1995). Matrix metalloproteinases are obligatory for the migration of preosteoclasts to the developing marrow cavity of primitive long bones. J. Cell Sci. 108, 3649–3659. Bruick, R. K., and McKnight, S. L. (2001). A conserved family of prolyl-4-hydroxylases that modify HIF. Science 294, 1337–1340. Carlevaro, M. F., Cermelli, S., Cancedda, R., and Descalzi Cancedda, F. (2000). Vascular endothelial growth factor (VEGF) in cartilage neovascularization and chondrocyte diVerentiation: Auto-paracrine role during endochondral bone formation. J. Cell Sci. 113, 59–69. Colnot, C., Lu, C., Hu, D., and Helms, J. A. (2004). Distinguishing the contributions of the perichondrium, cartilage, and vascular endothelium to skeletal development. Dev. Biol. 269, 55–69. Colnot, C., Thompson, Z., Miclau, T., Werb, Z., and Helms, J. A. (2003). Altered fracture repair in the absence of MMP9. Development 130, 4123–4133. Colnot, C. I., and Helms, J. A. (2001). A molecular analysis of matrix remodeling and angiogenesis during long bone development. Mech. Dev. 100, 245–250. Deckers, M. M., van Bezooijen, R. L., van der Horst, G., Hoogendam, J., van Der Bent, C., Papapoulos, S. E., and Lowik, C. W. (2002). Bone morphogenetic proteins stimulate angiogenesis through osteoblast-derived vascular endothelial growth factor A. Endocrinology 143, 1545–1553. Ducy, P., Zhang, R., GeoVroy, V., Ridall, A. L., and Karsenty, G. (1997). Osf2/Cbfa1: A transcriptional activator of osteoblast diVerentiation. Cell 89, 747–754. Engsig, M. T., Chen, Q. J., Vu, T. H., Pedersen, A. C., Therkidsen, B., Lund, L. R., Henriksen, K., Lenhard, T., Foged, N. T., Werb, Z., and Delaisse, J. M. (2000). Matrix metalloproteinase 9 and vascular endothelial growth factor are essential for osteoclast recruitment into developing long bones. J. Cell Biol. 151, 879–890. Ferguson, C., Alpern, E., Miclau, T., and Helms, J. A. (1999). Does adult fracture repair recapitulate embryonic skeletal formation? Mech. Dev. 87, 57–66. Ferrara, N., and Davis-Smyth, T. (1997). The biology of vascular endothelial growth factor. Endocr. Rev. 18, 4–25. Ferrara, N., Houck, K., Jakeman, L., and Leung, D. W. (1992). Molecular and biological properties of the vascular endothelial growth factor family of proteins. Endocr. Rev. 13, 18–32. 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. Haigh, J. J., Gerber, H. P., Ferrara, N., and Wagner, E. F. (2000). Conditional inactivation of VEGF-A in areas of collagen2a1 expression results in embryonic lethality in the heterozygous state. Development 127, 1445–1453.

184

Zelzer and Olsen

Harada, S., Rodan, S. B., and Rodan, G. A. (1995). Expression and regulation of vascular endothelial growth factor in osteoblasts. Clin. Orthop. 313, 76–80. Harper, J., Gerstenfeld, L. C., and Klagsbrun, M. (2001). Neuropilin-1 expression in osteogenic cells: Down-regulation during diVerentiation of osteoblasts into osteocytes. J. Cell Biochem. 81, 82–92. Henriksen, K., Karsdal, M., Delaisse, J. M., and Engsig, M. T. (2003). RANKL and vascular endothelial growth factor (VEGF) induce osteoclast chemotaxis through an ERK1/2dependent mechanism. J. Biol. Chem. 278, 48745–48753. Horton, W. A. (1990). The biology of bone growth. Growth Genet. Horm. 6, 1–3. Hunter, W. L., and Arsenault, A. L. (1990). Vascular invasion of the epiphyseal growth plate: Analysis of metaphyseal capillary ultrastructure and growth dynamics. Anat. Rec. 227, 223–231. Inada, M., Yasui, T., Nomura, S., Miyake, S., Deguchi, K., Himeno, M., Sato, M., Yamagiwa, H., Kimura, T., Yasui, N., Ochi, T., Endo, N., Kitamura, Y., Kishimoto, T., and Komori, T. (1999). Maturational disturbance of chondrocytes in Cbfa1-deficient mice. Dev. Dyn. 214, 279–290. Ivan, M., Kondo, K., Yang, H., Kim, W., Valiando, J., Ohh, M., Salic, A., Asara, J. M., Lane, W. S., and Kaelin, W. G., Jr. (2001). HIFalpha targeted for VHL-mediated destruction by proline hydroxylation: Implications for O2 sensing. [see comments]. Science 292, 464–468. Ivkovic, S., Yoon, B. S., PopoV, S. N., Safadi, F. F., Libuda, D. E., Stephenson, R. C., Daluiski, A., and Lyons, K. M. (2003). Connective tissue growth factor coordinates chondrogenesis and angiogenesis during skeletal development. Development 130, 2779–2791. Jaakkola, P., Mole, D. R., Tian, Y. M., Wilson, M. I., Gielbert, J., Gaskell, S. J., Kriegsheim, A., Hebestreit, H. F., Mukherji, M., Schofield, C. J., Maxwell, P. H., Pugh, C. W., and RatcliVe, P. J. (2001). Targeting of HIF-alpha to the von Hippel-Lindau ubiquitylation complex by O2-regulated prolyl hydroxylation. [see comments]. Science 292, 468–472. Karsenty, G., and Wagner, E. F. (2002). Reaching a genetic and molecular understanding of skeletal development. Dev. Cell 2, 389–406. Kim, I. S., Otto, F., Zabel, B., and Mundlos, S. (1999). Regulation of chondrocyte diVerentiation by Cbfa1. Mech. Dev. 80, 159–170. Komori, T., Yagi, H., Nomura, S., Yamaguchi, A., Sasaki, K., Deguchi, K., Shimizu, Y., Bronson, R. T., Gao, Y. H., Inada, M., Sato, M., Okamoto, R., Kitamura, Y., Yoshiki, S., and Kishimoto, T. (1997). Targeted disruption of Cbfa1 results in a complete lack of bone formation owing to maturational arrest of osteoblasts. Cell 89, 755–764. Kozawa, O., Kawamura, H., Hatakeyama, D., Matsuno, H., and Uematsu, T. (2000). Endothelin-1 induces vascular endothelial growth factor synthesis in osteoblasts: Involvement of p38 mitogen-activated protein kinase. Cell. Signal. 12, 375–380. Kozawa, O., Matsuno, H., and Uematsu, T. (2001). Involvement of p70 S6 kinase in bone morphogenetic protein signaling: Vascular endothelial growth factor synthesis by bone morphogenetic protein-4 in osteoblasts. J. Cell Biochem. 81, 430–436. Kronenberg, H. M. (2003). Developmental regulation of the growth plate. Nature 423, 332–336. Le, A. X., Miclau, T., Hu, D., and Helms, J. A. (2001). Molecular aspects of healing in stabilized and non-stabilized fractures. J. Orthop. Res. 19, 78–84. Maes, C., Carmeliet, P., Moermans, K., Stockmans, I., Collen, D., Bouillon, R., and Carmeliet, G. (2002). Impaired angiogenesis and endochondral bone formation in mice lacking the vascular endothelial growth factor isoforms VEGF164 and VEGF188. Mech. Dev. 111, 61–73. Maes, C., Stockmans, I., Moermans, K., Van Looveren, R., Smets, N., Carmeliet, P., Bouillon, R., and Carmeliet, G. (2004). Soluble VEGF isoforms are essential for establishing epiphyseal vascularization and regulating chondrocyte development and survival. J. Clin. Invest. 113, 188–199.


185

Masson, N., Willam, C., Maxwell, P. H., Pugh, C. W., and RatcliVe, P. J. (2001). Independent function of two destruction domains in hypoxia-inducible factor-alpha chains activated by prolyl hydroxylation. EMBO J. 20, 5197–5206. Midy, V., and Plouet, J. (1994). Vasculotropin/vascular endothelial growth factor induces diVerentiation in cultured osteoblasts. Biochem. Biophys. Res. Commun. 199, 380–386. Nakagawa, M., Kaneda, T., Arakawa, T., Morita, S., Sato, T., Yomada, T., Hanada, K., Kumegawa, M., and Hakeda, Y. (2000). Vascular endothelial growth factor (VEGF) directly enhances osteoclastic bone resorption and survival of mature osteoclasts. FEBS Lett. 473, 161–164. Niida, S., Kaku, M., Amano, H., Yoshida, H., Kataoka, H., Nishikawa, S., Tanne, K., Maeda, N., and Kodama, H. (1999). Vascular endothelial growth factor can substitute for macrophage colony-stimulating factor in the support of osteoclastic bone resorption. J. Exp. Med. 190, 293–298. Olsen, B. R., Reginato, A. M., and Wang, W. (2000). Bone development. Ann. Rev. Cell Dev. 16, 191–220. Ortega, N., Behonick, D. J., and Werb, Z. (2004). Matrix remodeling during endochondral ossification. Trends Cell Biol. 14, 86–93. Otto, F., Thornell, A. P., Crompton, T., Denzel, A., Gilmour, K. C., Rosewell, I. R., Stamp, G. W., Beddington, R. S., Mundlos, S., Olsen, B. R., Selby, P. B., and Owen, M. J. (1997). Cbfa1, a candidate gene for cleidocranial dysplasia syndrome, is essential for osteoblast diVerentiation and bone development. Cell 89, 765–771. Park, J. E., Keller, G. A., and Ferrara, N. (1993). The vascular endothelial growth factor (VEGF) isoforms: DiVerential deposition into the subepithelial extracellular matrix and bioactivity of extracellular matrix-bound VEGF. Mol. Biol. Cell 4, 1317–1326. Perbal, B. (2004). CCN proteins: Multifunctional signalling regulators. Lancet 363, 62–64. Pfander, D., Cramer, T., Schipani, E., and Johnson, R. S. (2003). HIF-1alpha controls extracellular matrix synthesis by epiphyseal chondrocytes. J. Cell Sci. 116, 1819–1826. Pfander, D., Kobayashi, T., Knight, M. C., Zelzer, E., Chan, D. A., Olsen, B. R., Giaccia, A. J., Johnson, R. S., Haase, V. H., and Schipani, E. (2004). Deletion of Vhlh in chondrocytes reduces cell proliferation and increases matrix deposition during growth plate development. Development 131, 2497–2508. Probst, A., and Spiegel, H. U. (1997). Cellular mechanisms of bone repair. J. Invest. Surg. 10, 77–86. Pugh, C. W., and RatcliVe, P. J. (2003). Regulation of angiogenesis by hypoxia: Role of the HIF system. Nat. Med. 9, 677–684. Reponen, P., Sahlberg, C., Munaut, C., ThesleV, I., and Tryggvason, K. (1994). High expression of 92-kD type IV collagenase (gelatinase B) in the osteoclast lineage during mouse development. J. Cell Biol. 124, 1091–1102. Saadeh, P. B., Mehrara, B. J., Steinbrech, D. S., Dudziak, M. E., Greenwald, J. A., Luchs, J. S., Spector, J. A., Ueno, H., Gittes, G. K., and Longaker, M. T. (1999). Transforming growth factor-beta1 modulates the expression of vascular endothelial growth factor by osteoblasts. Am. J. Physiol. 277, C628–C637. Schipani, E., Ryan, H. E., Didrickson, S., Kobayashi, T., Knight, M., and Johnson, R. S. (2001). Hypoxia in cartilage: HIF-1alpha is essential for chondrocyte growth arrest and survival. Genes Dev. 15, 2865–2876. Semenza, G. L. (2003). Targeting HIF-1 for cancer therapy. Nat. Rev. Cancer 3, 721–732. Shima, D. T., Kuroki, M., Deutsch, U., Ng, Y. S., Adamis, A. P., and D’Amore, P. A. (1996). The mouse gene for vascular endothelial growth factor. Genomic structure, definition of the transcriptional unit, and characterization of transcriptional and post-transcriptional regulatory sequences. J. Biol. Chem. 271, 3877–3883.

186

Zelzer and Olsen

Soker, S., Takashima, S., Miao, H. Q., Neufeld, G., and Klagsbrun, M. (1998). Neuropilin-1 is expressed by endothelial and tumor cells as an isoform-specific receptor for vascular endothelial growth factor. Cell 92, 735–745. Street, J., Bao, M., deGuzman, L., Bunting, S., Peale, F. V., Jr., Ferrara, N., Steinmetz, H., HoeVel, J., Cleland, J. L., Daugherty, A., van Bruggen, N., Redmond, H. P., Carano, R. A., and FilvaroV, E. H. (2002). Vascular endothelial growth factor stimulates bone repair by promoting angiogenesis and bone turnover. Proc. Natl. Acad. Sci. USA 99, 9656–9661. Street, J. T., Wang, J. H., Wu, Q. D., Wakai, A., McGuinness, A., and Redmond, H. P. (2001). The angiogenic response to skeletal injury is preserved in the elderly. J. Orthop. Res. 19, 1057–1066. Tokuda, H., Harada, A., Hirade, K., Matsuno, H., Ito, H., Kato, K., Oiso, Y., and Kozawa, O. (2003a). Incadronate amplifies prostaglandin F2 alpha-induced vascular endothelial growth factor synthesis in osteoblasts. Enhancement of MAPK activity. J. Biol. Chem. 278, 18930–18937. Tokuda, H., Hatakeyama, D., Akamatsu, S., Tanabe, K., Yoshida, M., Shibata, T., and Kozawa, O. (2003b). Involvement of MAP kinases in TGF-beta-stimulated vascular endothelial growth factor synthesis in osteoblasts. Arch. Biochem. Biophys. 415, 117–125. Tokuda, H., Hatakeyama, D., Shibata, T., Akamatsu, S., Oiso, Y., and Kozawa, O. (2003c). p38 MAP kinase regulates BMP-4-stimulated VEGF synthesis via p70 S6 kinase in osteoblasts. Am. J. Physiol. Endocrinol. Metab. 284, E1202–E1209. Tokuda, H., Hirade, K., Wang, X., Oiso, Y., and Kozawa, O. (2003d). Involvement of SAPK/ JNK in basic fibroblast growth factor-induced vascular endothelial growth factor release in osteoblasts. J. Endocrinol. 177, 101–107. Tokuda, H., Kozawa, O., Miwa, M., and Uematsu, T. (2001). p38 mitogen-activated protein (MAP) kinase but not p44/p42 MAP kinase is involved in prostaglandin E1-induced vascular endothelial growth factor synthesis in osteoblasts. J. Endocrinol. 170, 629–638. Tondravi, M. M., McKercher, S. R., Anderson, K., Erdmann, J. M., Quiroz, M., Maki, R., and Teitelbaum, S. L. (1997). Osteopetrosis in mice lacking haematopoietic transcription factor PU.1. Nature 386, 81–84. Trueta, J., and Amato, V. P. (1960). The vascular contribution of osteogenesis. III. Changes in the growth cartilage caused by experimentally induced ischaemia. J. Bone Joint Surg. 42B, 571–587. Trueta, J., and Buhr, A. J. (1963). The Vascular Contribution to Osteogenesis. V. The Vasculature Supplying the Epiphysial Cartilage in Rachitic Rats. J. Bone Joint Surg. Br. 45, 572–581. Trueta, J., and Trias, A. (1961). The vascular contribution to osteogenesis. IV. The eVect of pressure upon the epiphysial cartilage of the rabbit. J. Bone Joint Surg. Br. 43-B, 800–813. Vortkamp, A., Pathi, S., Peretti, G. M., Caruso, E. M., Zaleske, D. J., and Tabin, C. J. (1998). Recapitulation of signals regulating embryonic bone formation during postnatal growth and in fracture repair. Mech. Dev. 71, 65–76. Vu, T. H., Shipley, J. M., Bergers, G., Berger, J. E., Helms, J. A., Hanahan, D., Shapiro, S. D., Senior, R. M., and Werb, Z. (1998). MMP-9/gelatinase B is a key regulator of growth plate angiogenesis and apoptosis of hypertrophic chondrocytes. Cell 93, 411–422. Wang, D. S., Yamazaki, K., Nohtomi, K., Shizume, K., Ohsumi, K., Shibuya, M., Demura, H., and Sato, K. (1996). Increase of vascular endothelial growth factor mRNA expression by 1,25-dihydroxyvitamin D3 in human osteoblast-like cells. J. Bone Miner. Res. 11, 472–479. Wang, X., Tokuda, H., Hirade, K., and Kozawa, O. (2002). Stress-activated protein kinase/cJun N-terminal kinase (JNK) plays a part in endothelin-1-induced vascular endothelial growth factor synthesis in osteoblasts. J. Cell Biochem. 87, 417–423.


187

Yeh, L. C., and Lee, J. C. (1999). Osteogenic protein-1 increases gene expression of vascular endothelial growth factor in primary cultures of fetal rat calvaria cells. Mol. Cell Endocrinol. 153, 113–124. Yu, F., White, S. B., Zhao, Q., and Lee, F. S. (2001). HIF-1alpha binding to VHL is regulated by stimulus-sensitive proline hydroxylation. Proc. Natl. Acad. Sci. USA 98, 9630–9635. Zelzer, E., Glotzer, D. J., Hartmann, C., Thomas, D., Fukai, N., Shay, S., and Olsen, B. R. (2001). Tissue specific regulation of VEGF expression by Cbfa1/Runx2 during bone development. Mech. Dev. 106, 97–106. Zelzer, E., Levy, Y., Kahana, C., Shilo, B. Z., Rubinstein, M., and Cohen, B. (1998). Insulin induces transcription of target genes through the hypoxia-inducible factor HIF-1alpha/ ARNT. EMBO J. 17, 5085–5094. Zelzer, E., Mamluk, R., Ferrara, N., Johnson, R. S., Schipani, E., and Olsen, B. R. (2004). VEGFA is necessary for chondroycte survival during bone development. Development 131, 2161–2171. Zelzer, E., McLean, W., Ng, Y.-S., Fukai, N., Reginato, A. M., Lovejoy, S., D’Amore, P. A., and Olsen, B. R. (2002). Skeletal defects in VEGF120/120 mice reveal multiple roles for VEGF in skeletogenesis. Development 129, 1893–1904.

7

____________________________________________________________________________

G-Protein Coupled Receptors and Calcium Signaling in Development Geoffrey E. Woodard* and Juan A. Rosado{ *Metabolic Diseases Branch National Institute of Diabetes Digestive and Kidney Diseases Bethesda, Maryland 20892 { Department of Physiology University of Extremadura Ca´ceres, Spain 10071

I. II. III. IV.

Introduction G Protein-Coupled Receptors General Properties of G-Protein-Coupled Receptor-Dependent Ca2þ Signaling Mechanisms of Ca2þ Signaling A. Calcium Release from Intracellular Stores B. Calcium Entry C. Mechanisms for Calcium Removal from the Cytosol D. Local and Global Aspects of Ca2þ Signaling

V. Ca2þ Signaling in Development References

I. Introduction The majority of transmembrane signal transduction in response to extracellular messengers is mediated by G protein-coupled receptors (GPCR) or serpentine receptors, so named for their seven-membrane-spanning domain structure, which compose one of the largest families of signaling molecules. A large number of physiological stimuli, such as odors, light, metal ions, peptide hormones, and neurotransmitters, induce a conformational change in GPCRs that lead to the activation of heterotrimeric G-proteins and downstream eVectors, e.g., adenylate cyclase and phospholipase C, which, in turn, generate diVusible second messengers that act within the cell, triggering changes including cytosolic calcium concentration ([Ca2þ]i; see Bu¨nemann and Hosey, 1999). Cytosolic Ca2þ is a ubiquitous intracellular signal involved in the modulation of a large number of cellular functions, and dynamic changes in cytosolic Ca2þ are associated with the sequential steps that occur during development, such as the establishment of the dorsal–ventral axis, and perhaps other embryonic axes (Ault et al., 1996), organogenesis, and Current Topics in Developmental Biology, Vol. 65 Copyright 2005, Elsevier Inc. All rights reserved. 0070-2153/05 $35.00

189

190

Woodard and Rosado

morphogenetic cell motility and rearrangement (Webb and Miller, 2003). This chapter focuses on the calcium signaling induced by the activation of G protein-coupled receptors and its role in the modulation of development.

II. G Protein-Coupled Receptors The discovery of the role of guanine nucleotides in mediating hormonal stimulation of adenylate cyclase by Rodbell and Gilman in the 1970s preceded the identification of the G protein-coupled receptors (Gillman, 1987; Rodbell, 1980). Subsequently, the signal transduction of membrane receptor occupation by several ligands of diverse structural classes was found to be sensitive to guanine nucleotides. These studies led to the isolation of the heterotrimeric G-protein associated to the photon receptor, rhodopsin, in 1982 (Baehr et al., 1982) and, one year later, the opsin apoprotein of bovine rhodopsin was cloned (Nathan and Hogness, 1983). Since 2000, the sequences of the G protein-coupled receptors have been characterized. The identification and cloning of the -adrenergic receptor by Dixon and coworkers in 1986 (Dixon et al., 1986) showed that these receptors have seven hydrophobic domains, suggesting that they cross the plasma membrane seven times. The structure of G protein-coupled receptors consists of an extracellular Nterminal segment, seven transmembrane domains, three external loops, three cytoplasmic loops, and a C-terminal segment. A fourth cytoplasmic loop is created when a cysteine residue located in the C-terminal segment is palmitoylated. The number of seven transmembrane domains has been suggested to be important in forming a core which might be essential to provide a wide number of specificities and regulatory mechanisms. The N-terminal segment is involved in ligand interaction, and its variable size (7–595 amino acids) indicates its diverse specificity (Ji et al., 1998). A positive correlation between the length of the N-terminal segment and ligand size has been proposed (Ji et al., 1995), suggesting a role for this segment in ligand binding. G protein-coupled receptors have been suggested to appear in a ternary complex model, which explains the cooperative interactions between ligands, the receptor itself, and the G proteins associated. This model proposes that the receptor exists in an equilibrium between the inactive and the active state. In the absence of ligands, basal receptor activity is determined by the equilibrium between both states and the eYcacy of ligands is a result of their ability to alter the equilibrium between these two states (Gether and Kobilka, 1998). Rhodopsin is unique among the G protein-coupled receptors since its ligand, cis-retinal, is covalently bound to the receptor and, upon absorption of a photon, isomerizes to its agonist, trans-retinal, which induces a series of conformational changes in the molecule of rhodopsin that leads to

7. Calcium Signaling in Development

191

the formation of metarhodopsin II, the activated state of rhodopsin (Farahbakhsh et al., 1993). In contrast to the extensive characterization of the conformational changes of rhodopsin by photoisomerization of retinal, very little is known about the mechanism of activation of G protein-coupled receptors for diVusible ligands and several hypotheses have been proposed. The ligand induction mechanism assumes that the free energy of ligand binding is used to overcome the energy barrier that facilitates the transition to the active receptor state (Bennett and Steitz, 1978). According to the conformational selection model, the transitions between inactive and active receptor states can occur in the absence of ligand and agonists bind to the active conformation, thereby shifting the equilibrium toward the active receptor state (Koshland and Neet, 1968). A third hypothesis has been proposed based on a sequential binding and conformational stabilization (Gether and Kobilka, 1998). This model suggests that the receptor might have three diVerent states: the unliganded receptor, the active receptor, and a third transition stabilized by inverse agonists. Moreover, the unliganded receptor may undergo transition to the active state spontaneously, which might account for the high basal activity observed for some G protein-coupled receptors. Ligand binding occurs sequentially, resulting in a series of conformational intermediate states between the unliganded and the active receptor, so that the initial binding of one structural group of the ligand might lead to binding of the remaining groups in a sequential manner. The sequential interactions between the receptor and the ligand stabilizes the receptor in the active state. This model is consistent with the rapid association rate for agonists and the slow rate of conformational change observed for 2 adrenrenergic receptors (Gether et al., 1995). The interaction of the heterotrimeric G proteins with the receptor might induce its stabilization in one of the intermediate states, which might influence both agonist-binding aYnity and the kinetics of the conformational change (Gether and Kobilka, 1998). Activation of G protein-coupled receptors is associated with diVerent intracellular pathways, through its interaction with G proteins, that lead to generation of a variety of cytosolic messengers, which, in turn, activate and/or modulate a cascade of reactions involved in a complex network that regulates cell function. A variety of heterotrimeric G proteins have been described. Basically, G proteins are molecular switches that transduce the signal initiated by the binding of an extracellular messenger to its receptor at the cell surface to an intracellular eVector. The G protein-coupled receptors regulate a specific eVector by interacting with a particular G protein. The nomenclature of the G proteins is confusing since some of these proteins were functionally characterized before they were sequenced. Thus, the G protein activating adenylate cyclase is called Gs, while that inhibiting the enzyme is termed Gi.

192

Woodard and Rosado

In these cases, the structural and functional classifications coincide. In contrast, other G proteins, such as that activating the phospholipase C (PLC ), initially called Gp, has now been included as a member of the Gq family. G proteins are heterotrimers which consist of three subunits (, , and ). At present, at least twenty subunits, five , and six subunits have been described, which increase the number of potential heterotrimeric G proteins obtained by combination of subunits. The subunit is a protein of 45 kDa with GTPase activity. This subunit is the main one responsible for the biological activity of the G protein; therefore, it is subjected to the same nomenclature as the G proteins. In the inactive state, the subunit of the G protein is bound to GDP, but upon receptor–ligand binding, a conformational change occurs so that the intracellular domains of the receptor, particularly the third intracellular loop, activate the subunit of the G protein by facilitating the exchange of GDP by GTP. The subunit then dissociates and activates a specific intracellular pathway. The activation of the subunit is rapidly terminated by the hydrolysis of GTP to GDP by its intrinsic GTPase activity. The subunit bound to GDP exhibits a high aYnity for the dimer and the inactive heterotrimeric G protein is formed again. The rapid activation cycle of the G protein (from milliseconds to seconds, depending on the protein) allows this signaling molecule to prepare for another stimulus (Svoboda et al., 2004). Although historically the specificity of the interaction between G proteins and G protein-coupled receptors, as well as the biological function, was attributed solely to the subunits, a role for the subunits in these functions has been presented (Kleuss et al., 1992). The subunit is a 35 kDa protein that binds to the and subunits (Lupas et al., 1992). The subunit is a 8.4 kDa protein that is physically close to the subunit (Mumby et al., 1990). Among the subunit eVectors, the most relevant are the adenylate cyclase, involved in the generation of cyclic AMP, which might be either activated or inhibited, depending on the subunit involved, s or i, respectively (Svoboda et al., 2004), and PLC , which cleaves phosphatidylinositol 4,5-bisphosphate (PIP2), a lipid on the inner leaflet of the plasma membrane, to form the intracellular messenger inositol 1,4,5-triphosphate (IP3) and 1,2-diacylglicerol (DAG). We limit our review to the PLCactivated signaling involved in cytosolic Ca2þ mobilization. Both IP3 and DAG act as intracellular messengers. IP3 diVuses into the cytoplasm where it binds to an intracellular receptor located on the endoplasmic reticulum (ER) that functions as a ligand-gated calcium channel, thereby initiating the release of sequestered Ca2þ (Petersen et al., 1994). On the other hand, DAG activates protein kinase C (PKC), a cytosolic enzyme that is also activated by phorbol esters (De Pont and Fleuren-Jakobs, 1984). Several phosphatidylinositol-specific isoforms of PLC have been identified in mammals, including four -, two -, four -isoforms, and numerous


193

spliced variants (Rebecchi and Pentyala, 2000). Activation of PLC results in rapid increase of IP3, which reaches a peak within several seconds and then declines to a plateau. IP3 has a lifetime of a few seconds; it is rapidly metabolized by enzymes that either add or remove a phosphate group to terminate the Ca2þ-releasing potential of IP3. The addition of a phosphate group by a Ca2þ-dependent kinase results in the production of inositol 1,3,4,5-tetrakisphosphate (IP4). Although its role in cell physiology is not completely understood, IP4 binds to a specific small GTPase-activating protein (GAP) that has been shown to modulate Ca2þ release (Cullen, 1998). G protein-coupled receptors’ occupation also induces biphasic increases in the cytosolic concentration of DAG, although this eVect is more sustained than that observed for IP3 (Williams, 2001). The signal initiated by G protein-coupled receptor occupation is terminated by several mechanisms. In addition to the previously mentioned hydrolysis of GTP to GDP by the subunit of the G proteins, which induces the reassociation of the heterotrimer, ligand dissociation from the receptor results in the transition of the receptor back to the inactive state. Moreover, a variety of mechanisms have been proposed to be involved in the regulation of the signals generated by the activation of G protein-coupled receptors, including G protein-coupled receptor kinases that phosphorylate active receptors, inducing homologous desensitization (Bu¨ nemann and Hosey, 1999); uncoupling proteins of the arrestin family have been shown to induce desensitization by preventing the association of G protein-coupled receptors and G proteins (Lefkowitz, 1998), receptor internalization by endocytosis (Ferguson, 2001), and regulators of G protein signaling, which accelerate the GTP hydrolysis on subunits and therefore reduce G protein activity (Cancela, 2001; Muallem and Wilkie, 1999).

III. General Properties of G-Protein-Coupled Receptor-Dependent Ca2þ Signaling Stimuli that bind to G protein-coupled receptors switch on the G protein turnover cycle, resulting in the activation of the subunit of the G protein. Activation of eVectors by the subunit leads to amplification of the signal generated by the extracellular stimuli and to activation or inhibition of downstream signaling molecules, depending on the nature of the G protein involved, Gs, Gi, or Gq. An important second messenger regulated by G protein-coupled receptors is cytosolic Ca2þ. A remarkable variety of physiological stimuli transduce their information by inducing an increase in [Ca2þ]i. Ca2þ signaling consists of two components, a biochemical mechanism that includes the G protein-coupled receptor and eVector proteins, Gq protein and PLC , that lead to the generation of IP3. IP3 then diVuses toward the ER and activates a specific receptor located on the membrane of the ER, thus

194

Woodard and Rosado 2þ

releasing stored Ca into the cytosol and initiating a biophysical component, which includes the mechanisms that increase [Ca2þ]i and those that return [Ca2þ]i to the resting level (Kiseljov et al., 2002). Under physiological conditions, the Ca2þ mobilization might follow diVerent patterns, from transient or sustained responses to repetitive [Ca2þ]i oscillations. A key question in cell signaling is how cells generate specific Ca2þ signals upon activation of G protein-coupled receptors. A number of studies in the last few years began to answer this question. The specificity of Ca2þ signaling is determined by multiple biochemical and biophysical mechanisms, which include localization of signaling complexes in cellular microdomains, diVerences in receptor coupling to G proteins, regulation of the G protein turnover cycle by RGS proteins, generation of multiple Ca2þ-releasing messengers such as IP3, cyclic ADP ribose, and nicotinic acid–adenine dinucleotide phosphate (NAADP) that might interact with each other, or regulation of the activity of the Ca2þ transporters (Kiseljov et al., 2002). The mechanisms involved in the regulation of cytosolic Ca2þ concentration upon the occupation of a G protein-coupled receptor are described in the following text.

IV. Mechanisms of Ca2þ Signaling As has been reported, Ca2þ is a ubiquitous intracellular messenger responsible for the modulation of a large number of cellular processes. Increases in [Ca2þ]i can initiate and modulate many diVerent events, ranging from short-term responses, such as muscle contraction (Reembold, 1992) and secretion (Brown et al., 1985) to long-term processes like cell growth (Means, 1994) or diVerentiation (Buonanno and Fields, 1999). Although the free calcium concentration in the extracellular medium is about 1 mM, [Ca2þ]i is maintained in a range of 20 to 100 nM at rest, depending on the cell type. Cellular agonists can increase [Ca2þ]i by generating Ca2þ-mobilizing molecules, including IP3 (Streb et al., 1983), cyclic ADP ribose (Galione, 1994), or NAADP (Cancela et al., 1999), that release compartmentalized Ca2þ from intracellular stores or by stimulating the entry of extracellular Ca2þ across plasma membrane channels by diVerent mechanisms that will be discussed later. When the agonist stimulation ceases, a number of mechanisms, mainly pumps and exchangers, remove Ca2þ from the cytosol to restore the resting [Ca2þ]i when stimulation is terminated by removal of the agonist. A. Calcium Release from Intracellular Stores Ca2þ release from the intracellular stores is controlled by various channels located in the membrane of the stores, including the IP3 receptor, the so-called ryanodine receptor, the NAADP receptor, and the sphingosine


195

1-phosphate receptor. The IP3 and ryanodine receptors have been extensively studied while the NAADP and sphingosine 1-phosphate receptors are still not well characterized (Berridge et al., 2000). 1. The Inositol Triphosphate (IP3) Receptor The molecule that regulates the opening of this channel is inositol 1,4,5 triphosphate, generated by the occupation of G protein-coupled receptors that stimulate PLC or by growth factor receptors coupled to PLC (Berstein et al., 1992; Kim et al., 2000). IP3 receptors have been suggested to couple with certain G protein-coupled receptors via the protein Homer, which might be important for the colocalization of these functionally coupled proteins (Tu et al., 1998). At least three IP3 receptor isoforms are known and they appear as homoand heterotetramers (Monkawa et al., 1995). IP3 receptors are structurally divided into three regions: an N-terminus cytoplasmic domain with an IP3 binding site, a modulatory central domain, and a transmembrane region near the C terminus composed of six-spanning domains (Bosanac et al., 2002; Hamada et al., 2003). The transmembrane region is important for the intermolecular interaction in the formation of a tetrameric complex (Sayers et al., 1997). The central domain has been suggested to be essential for binding of allosteric eVectors, such as Ca2þ or ATP, or for phosphorylation, which regulate(s) the opening of the channel (Mikoshiba, 1993; Sato et al., 2004). The eVect of cytosolic Ca2þ on the activity of the IP3 receptor depends on the concentration of the ion, so that low Ca2þ concentrations (100–300 nM) are stimulatory, while higher concentrations are either inhibitory or, after IP3 binding, do not aVect IP3 receptor activity (Berridge et al., 2000; Bootman and Lipp, 1999). 2. The Ryanodine Receptor Ryanodine receptors were initially described in the skeletal muscle as a tetrameric channel responsible for sarcoplasmic reticulum Ca2þ release and, subsequently, they have been recognized to have a role in Ca2þ regulation in nonmuscle cells. As for the IP3 receptors, there are also three ryanodine receptor isoforms. These channels are regulated by physiological and pharmacological molecules, such as cyclic ADP ribose; Ca2þ itself, which is the most important modulator of the ryanodine receptors; caVeine; or the neutral plant alkaloid ryanodine, which greatly aided in the identification of these receptors (McPherson and Campbell, 1993). Ryanodine, at nanomolar concentrations, has been shown to be able to lock the Ca2þ channel in an open state leading to depletion of the Ca2þ store (Fill and Coronado, 1988). In skeletal muscle, ryanodine receptors are activated by

196


direct coupling to L-type Ca channels facilitated by the proximity of the T tubules and the sarcoplasmic reticulum (McPherson and Campbell, 1993). 3. NAADP Receptor Since the discovery of Ca2þ release by NAADP in sea urchin eggs (Lee and Aarhus, 1995), there has been a growing interest in this novel Ca2þ-releasing agent. NAADP receptors are likely located on Ca2þ stores distinct from the endoplasmic reticulum, and the activation of these channels evokes complex changes in [Ca2þ]i, which involve cross-talk with the IP3 and ryanodine receptors (Patel, 2004). In several cell types, such as sea urchin and starfish eggs, there is considerable interaction between NAADP, IP3, and cyclic ADP-ribose in regulating the magnitude of Ca2þ mobilization. NAADP initiates the activation of Ca2þ-induced Ca2þ release from IP3- and cyclic ADP ribose-sensitive stores, leading to the appearance of an oscillatory pattern (Churchill and Galione, 2001; Santella et al., 2000). This phenomenon has also been observed in pancreatic acinar cells, where acetylcholinestimulated local Ca2þ responses, which are triggered via IP3 receptors, are converted into global responses by the presence of NAADP or cyclic ADP ribose, and Ca2þ mobilization stimulated by cholecystokinin, mediated by converging pathway, involving cyclid ADP ribose and NAADP, is potentiated by IP3 (Cancela et al., 1999). 4. Sphingosine 1-Phosphate Receptor Sphingosine 1-phosphate releases Ca2þ from the endoplasmic reticulum by possibly binding to a sphingolipid Ca2þ-release-mediating protein of the endoplasmic reticulum (Mao et al., 1996) or by activating ryanodine receptors, since sphingosine 1-phosphate enhances ryanodine binding to its receptor and Ca2þ release by sphingosine 1-phosphate is prevented by ryanodine receptor blockers (Dettbarn et al., 1995).

B. Calcium Entry Ca2þ release from finite internal Ca2þ stores is often insuYcient for full activation of cellular mechanisms and many cellular processes, as well as the refilling of the intracellular stores, require a sustained increase in cytosolic Ca2þ, and therefore Ca2þ entry into the cell plays an important role. In excitable cells, such as neurons, muscle, and some endocrine cells, Ca2þ entry generally occurs through voltage-operated Ca2þ channels (VOC); however, in non-excitable cells, where voltage-mediated Ca2þ entry is negligible,

7. Calcium Signaling in Development 2þ

197

Ca influx is mainly mediated by receptor-operated channels (ROC), second messenger-operated channels (SMOC), or store-operated channels (SOC). VOCs are Ca2þ permeable channels that become briefly activated during action potentials (Tsien et al., 1995). These channels are mainly found in excitable cells, such as neurons, muscle, and endocrine cells, where they open in response to membrane depolarizations to allow Ca2þ to enter the cell (McCleskey, 1994). ROC include a series of functionally diverse channels that are specially relevant in secretory cells and neurones. ROC are activated by a number of cellular agonists, usually neurotransmitters, including acetylcholine, glutamate, ATP, and ADP. Activation of ROC induces a rapid Ca2þ entry, indicative of a direct coupling between the receptor and a Ca2þ permeable channel. ROC opening might be activated directly by interaction of the ligand with the subunits of a transmembrane protein or by some coupled system, perhaps involving a heterotrimeric G protein (Sage, 1992). SMOC are Ca2þ channels activated by a diVusible intracellular molecule, such as inositol phosphate or Ca2þ itself, whose concentration was increased as a result of agonist-receptor binding (Sage, 1992). Evidence for SMOC in excitable cells is rather scarce. In non-excitable cells, a Ca2þ permeable channel activated by Ca2þ and inositol 1,3,4,5-tetrakisphosphate has been found in plasma membrane of endothelial cells (Lu¨ckhoV and Clapham, 1992). In human platelets, thrombin has been reported to activate a storeindependent (noncapacitative) Ca2þ entry, which is activated via the stimulation of PLC, the formation of DAG, and so the activation of PKC (Rosado and Sage, 2000). In addition, some transient receptor potential channels (TRPC) have been reported to be activated by DAG analogues in diVerent non-excitable cells (Ma et al., 2000). The major mechanism for Ca2þ entry in non-excitable cells is storemediated Ca2þ entry through SOC, activated by depletion of the intracellular Ca2þ stores, either by physiological agonists or by pharmacological agents. It is not yet clear how the filling state of the intracellular Ca2þ stores is communicated to the plasma membrane but a number of hypotheses have been suggested. These can be divided into those which propose a role for a diVusible messenger and those which propose a direct interaction between proteins in the ER and PM (conformational coupling). DiVusible messengers which might be involved in the activation of SMCE include small GTPbinding proteins, cyclic GMP, a product of cytochrome P450, a Ca2þ influx factor (CIF), a tyrosine phosphorylation-dependent step, and a Ca2þcalmodulin-dependent step (see Parekh and Penner, 1997). On the other hand, the conformational coupling model suggests an interaction between

198

Woodard and Rosado

the IP3 receptor (IP3R) in the membrane of the ER, and a Ca2þ permeable channel in the PM (Berridge, 1995). A diVerent hypothesis has been proposed based on the insertion of channels previously in intracellular vesicles, in the PM (Somasundaram et al., 1995). The conformational coupling has received support from studies that propose a secretion-like coupling for the activation of SMCE (Patterson et al., 1999; Redondo et al., 2003; Rosado et al., 2000a). These studies have proposed a model where Ca2þ store depletion leads to traYcking of portions of the ER toward the PM to enable a reversible coupling of the IP3Rs in the ER and a Ca2þ channel in the PM, which requires a reorganization of the actin cytoskeleleton. While inhibition of actin filament polymerization has reported conflicting results in diVerent cell types, which might be due to the variability in the actin cytoskeleton intrinsic properties, a negative role for the cortical actin filament network in the activation of SMCE is generally accepted. The cell-permeant peptide, jasplakinolide, which induces actin polymerization and stabilizes actin filaments into a thick cortical layer, reduces SMCE in several cell types, including muscle cells (Patterson et al., 1999), platelets (Rosado et al., 2000a) and pancreatic acinar cells (Redondo et al., 2003). Similar results were obtained with calyculin A, an inhibitor of protein phosphatases 1 and 2, which induces phosphorylation-dependent association of the actin filaments to the PM (Patterson et al., 1999; Rosado et al., 2000a; Xie et al., 2002). These results provide compelling evidence against the direct activation of SMCE by a diVusible messenger, since a molecule would be expected to pass across the actin barrier and activate Ca2þ entry, in the same way that IP3 generated by physiological agonists is able to reach the ER and release stored Ca2þ in jasplakinolide-treated cells (Patterson et al., 1999; Rosado et al., 2000a). The secretion-like coupling appears as an integrative model, where components of the indirect and direct coupling might be merged. Consistent with this, small GTP-binding proteins, tyrosine kinase proteins, initially considered as diVusible messengers belonging to the indirect coupling hypotheses for the activation of SMCE, are essential for actin polymerization induced by store depletion (Rosado and Sage, 2000a; Rosado et al., 2000b). In addition, reorganization of the cytoskeletal cortical barrier has been suggested to facilitate the activation of SMCE by CIF (Xie et al., 2002). The conformational coupling model for the activation of SMCE proposed a direct interaction between the IP3Rs in the ER and a Ca2þ channel in the PM. The role of the IP3Rs in SMCE has widely been confirmed (Irvine, 1990; Mikoshiba, 1997), and the role of mammalian homologues of the Drosophila TRPC is receiving support from studies that demonstrate coupling between TRPCs and IP3Rs in transfected cells (Kiseljov et al., 2002; Vazquez et al., 2001) and in cells naturally expressing TRPCs (Rosado and Sage, 2000d; Rosado et al., 2002).


199

C. Mechanisms for Calcium Removal from the Cytosol Once the Ca2þ signal has achieved its function, Ca2þ is rapidly removed from the cytosol to maintain [Ca2þ]i low in the resting state. Ca2þ removal is carried out by several Ca2þ pumps and exchangers which reintroduce Ca2þ into the internal stores or extrude it out of the cell. 1. Calcium Reuptake into Internal Stores Ca2þ uptake into the intracellular stores mostly occurs against a concentration gradient, since [Ca2þ]i is lower than intraluminal Ca2þ, which, at rest, is in a high-micromolar to low-millimolar concentration range. Released Ca2þ is pumped back to the stores by the sarcoendoplasmic reticulum Ca2þ ATPase (SERCA), which was first isolated from muscle cells (for review, see Wuytack et al., 1992). There are three diVerent SERCA genes, and additional isoform subtypes are generated by alternative splicing (Exton, 1997). In addition, several SERCA isoforms coexpress in the same cell type, which might be associated with diVerent Ca2þ pools (Cavallini et al. 1995). SERCA isoforms have a high aYnity for Ca2þ (0.1–0.4 M); therefore, SERCA is likely to be regulated by intracellular Ca2þ concentration (activated by an increase in [Ca2þ]i) and the intraluminal free Ca2þ concentration (inhibited by an increase in luminal Ca2þ; Carafoli, 1992). Several pharmacological tools have been developed to investigate the role of SERCA in Ca2þ signaling. The activity of SERCA is inhibited by agents such as thapsigargin, 2,5-di(tert-butyl)-1,4-benzohydroquinone (TBHQ), and cyplopiazonic acid. Thapsigargin binds stoichiometrically to all SERCAs and causes an essential irreversible inhibition of their activity by blocking the ATPase in the Ca2þ-free state (Wictome et al., 1992). The action of TBHQ is similar to that of thapsigargin, but it has a significantly lower potency and some SERCA isoforms seem to be insensitive to this agent (Cavallini et al., 1995). 2. Calcium Extrusion Across the Plasma Membrane Another major mechanism for the removal of cytosolic Ca2þ is the extrusion of Ca2þ to the extracellular medium against a concentration gradient. Ca2þ eZux to the external medium is carried out by diVerent transporters, the plasma membrane Ca2þ ATPase (PMCA), and the Naþ/Ca2þ exchanger. The PMCA is an ATPase characterized by the formation of a covalent phosphorylated intermediate and by its inhibition by vanadate and lanthanum (Lajas et al., 2001; Pariente et al., 1999; Pedersen and Carafolli, 1987). Several molecular biology studies revealed the expression of at least four

200

Woodard and Rosado

PMCA isoforms in humans: PMCA1, PMCA2, and PMCA3, showing an 85% amino acid identity, and PMCA4 (Pariente et al., 2003; Strehler and Zacharias, 2001). The number of isoforms is increased by the existence of alternative splice variants (Strehler and Zacharias, 2001). The structure of the PMCA consists of ten transmembrane segments and five extracellular domains, with the NH2 and COOH termini located in the cytosolic site of the membrane (Guerini, 1998; Strehler and Zacharias, 2001). The activity of PMCA has been shown to be regulated by several messenger molecules, including Ca2þ/calmodulin, protein tyrosine kinases, PIP2, protein serine/threonine kinases, such as PKA and PKC, and by proteases like calpain (Pariente et al., 2003; Strehler and Zacharias, 2001). Ca2þmobilizing agonists might either increase or inhibit the PMCA activity in several cell types by activation of these intracellular pathways (Pariente et al., 2001; Rosado and Sage, 2000b). The Naþ/Ca2þ exchanger is a bidirectional electrogenic ion transporter that couples the movement of Naþ in one direction with the transport of Ca2þ in the opposite direction. The Naþ/Ca2þ exchanger is involved in the regulation of [Ca2þ]i either by removing Ca2þ from the cytosol (forward mode) or by promoting Ca2þ entry in the cell (reverse mode). At least three diVerent Naþ/Ca2þ exchangers have been described: the Kþ-independent Naþ/Ca2þ exchanger, the Kþ-dependent Naþ/Ca2þ exchanger, both electrogenic, and an electroneutral mitochondrial Naþ/Ca2þ exchanger (Matsuda et al., 1997). The Kþ-independent Naþ/Ca2þ exchanger is thought to catalyze the countertransport of either 3 or 4 Naþ for 1 Ca2þ (Blaustein and Lederer, 1999) and three isoforms have been described (Li et al., 1994; Nicoll et al., 1990, 1996). In contrast, the Kþ-dependent Naþ/Ca2þ family of exchangers catalyzes the exhange of 4 Naþ by 1 Ca2þ and 1 Kþ (Dong et al., 2001) and is composed of six members (Cai and Lytton, 2004). 3. Role of Mitochondria in Ca2þ Signaling Mitochondria are another relevant component of the mechanisms involved in Ca2þ signaling. Localized in the vicinity of the Ca2þ-releasing channels, mitochondria sequesters Ca2þ released from the agonist-sensitive stores and then releases it back slowly when the Ca2þ signal is terminated (Berridge et al., 2000). Transport of Ca2þ across the inner mitochondrial membrane is mediated by three mechanisms. Ca2þ entry into mitochondria is controlled by a highcapacity and low-aYnity uniporter that transports Ca2þ down a electrochemical gradient created by protein extrusion to allow ATP synthesis. The low aYnity of the uniporter requires local high [Ca2þ]i to function, which suggests that mitochondria should be close to the Ca2þ-releasing channels (Berridge et al., 2000; Pariente et al., 2003). EZux of Ca2þ takes place by two diVerent

201


þ

þ

exchangers that countertransport Ca for either Na or H , or through a permeability transition pore that might be involved in Ca2þ-induced Ca2þ release from mitochondria, since it is activated by rises in [Ca2þ]i (Ichas et al., 1997). The permeability transition pore might have two diVerent states: a reversible low-conductance state, which allows mitochondria to participate in Ca2þ signaling, and an irreversible high-conductance state, which collapses the mitochondrial membrane potential, leading to the release of cytochrome c and the activation of apoptosis (see Berridge et al., 2000). Mitochondrial Ca2þ accumulation has a dual role in cell function: a universal role in the activation of mitochondrial enzymes involved in the generation of ATP, and a more specific role in the modulation of Ca2þ signaling (Brini, 2003). The role of mitochondria is important for the spatiotemporal aspects of Ca2þ signaling as well as for the regulation of the amplitude of the increments in [Ca2þ]i (Camello-Almaraz et al., 2002; Gonzalez and Salido, 2001). 4. Ca2þ-Binding Proteins There are a large number of Ca2þ-binding proteins involved in Ca2þ signaling acting as Ca2þ sensors, eVectors, or buVers of free Ca2þ ions that initiate, execute, or terminate Ca2þ-dependent cellular functions. Most of the Ca2þbinding molecules inside the cell act as buVers maintaining a low [Ca2þ]i at rest, and there is an equilibrium between cytosolic free Ca2þ and Ca2þ bound to these proteins. It is well established that, in both excitable and nonexcitable cells, aproximately 98 to 99% of Ca2þ ions in the cytosol are bound by buVer molecules, including Ca2þ-binding proteins (Mogami et al., 1999; Neher and Augustine, 1992). Ca2þ-binding buVer proteins include calretinin, calbindin, parvalbumin, predominantly in the cytosol, and calsequestrin and calreticulin and several sarcoplasmic Ca2þ-binding protein in the intracellular stores. D. Local and Global Aspects of Ca2þ Signaling Ca2þ signals are a result of the activity of the mechanism previously described and the amplitude, location, and distribution of the signals determine its physiological role. The shape of the Ca2þ signal varies, depending on the intensity of the stimulus. High concentrations of agonist lead to rapid and transient rises in [Ca2þ]i; however, when a more physiological concentration of agonist is employed, an oscillatory pattern of Ca2þ signals can be observed, consisting of spikes of Ca2þ release from intracellular stores that rapidly decrease and return toward the prestimulation level (for a review, see Williams, 2001).

202


Ca signals initiate in a discrete area of the cell that has been termed the ‘‘trigger zone.’’ In pancreatic acinar cells, where this event has been widely investigated, the trigger zone is localized at the luminal cell pole where the secretory granules are located (Toescu et al., 1992). Cell stimulation with low concentrations of agonists generates local rises in [Ca2þ]i that are usually limited to these areas (Thorn et al., 1993). In contrast, higher concentrations of agonist lead to a global Ca2þ signal that spreads across the cell as a wave (Amundson and Clapham, 1993; Thorn et al., 1993). Supporting the propagation of Ca2þ signals throughout the cytosoplasm, a mechanism of calcium-induced calcium release (CICR) has been proposed by which Ca2þ stimulates its own release from intracellular stores (Nathanson et al., 1992). This process has been shown to be cyclical, leading to the appearance of repetitive Ca2þ spikes (Berridge and Bootman, 1996). When cells are connected by gap junctions, these Ca2þ signals are propagated from one cell to the next, as an intercellular wave might coordinate the function of cell popullations (Sanderson et al., 1994).

V. Ca2þ Signaling in Development During development, a number of factors need to operate dynamically in an integrated system in the embryo in order to adapt to a changing environment without losing or misinterpreting information. Among these factors, the highly versatile Ca2þ signaling has been shown to play an important role (Webb and Miller, 2003). The involvement of Ca2þ signals in development has been described in several cell types and animals. In zebrafish, Ca2þ mobilization starts at the beginning of egg activation and continues during the major phases of pattern formation, such as cell proliferation and diVerentiation, axis determination, and the formation of primary organ systems (Webb and Miller, 2000). A similar role was found in human oocytes, where Ca2þ oscillations are required for activation (Herbert et al., 1997) and fertilization triggers Ca2þ waves that activate egg development (Liu et al., 2001). The source for these Ca2þ oscillations is the Ca2þ stored in the ER. Interestingly, it has been reported that mitochondria show Ca2þ oscillations of the same frequency as those observed in the cytosol; in addition, mitochondrial dysfunction abolished the Ca2þ waves, suggesting that mitochondria plays a key role in Ca2þ signaling that activates egg development or apoptotic cell death (Liu et al., 2001). In addition to the role of Ca2þ signaling in oocyte development, a role has been proposed for this second messenger in cell diVerentiation. Morphometric analyses of cultured Purkinje cells have provided evidence supporting that Ca2þ stored in the ER is essential for dendrite outgrowth, as demonstrated by the inhibitory eVect of


203

thapsigargin, which prevents Ca storage in the ER by SERCA (Reitstetter and Yool, 1998). Special attention should be given to Wnt/calcium signaling for its interesting role in development (Ku¨hl, 2003; Veeman et al., 2003). The Wnts act as a ligand family of proteins coded by approximately 19 genes in the WntFrizzled biochemical pathway. Glycoproteins, with lipid modification, are the products of these genes (Willert et al., 2003). Wnt proteins bind to the Frizzled family of receptors to exert their intracellular signal (Wodarz et al., 1998). The Frizzled family of receptors have seven transmembrane segments, similar to their G-protein-coupled receptor relatives. It seems likely that there are numerous signaling mechanisms for the Wnt ligands and their receptors given the large family in which they exist. This large family has been divided into two subcategories. The Wnt/beta-catenin pathway is stimulated by the Wnt-1/wingless category of ligand/receptors. It is implicated in the tranformation of C57mg mammary epithelial cells and it will cause secondary axis formation in Xenopus. However, the Wnt-5A category of ligand/receptors that includes Wnt-5A, Wnt-4, and Wint-11 seems to have an antagonistic eVect on the Wnt/beta-catenin pathway as these receptors inhibit secondary axis formation and are involved in what is termed noncanonical Wnt signaling. The TCF/LEF family of transcription factors are known to interact with the disheveled, adenomatous poliposis coli, and axin proteins after their stabilization of cytoplasmic beta-catenin in the canonical Wnt/beta-catenin pathway. Beta-catenin is only involved in the canonical Wnt/beta-catenin pathway. Whereas calcium release from intracellular stores (Ku¨ hl et al., 2000a), Rho GTPases and Jun-N-terminal kinase (Habas et al., 2003) transduce noncanonical Wnt signaling. Wnts have also been shown to control cGMP signaling (Ahumada et al., 2002). When Wnts bind to Frizzled, they activate heterotrimeric G-proteins and, in turn, cause a stimulation of phospholipase C as a result of G-proteins beta/gamma dimerization (Slusarski et al., 1997a,b). The resultant phospholipase C will cleave phosphatidylinositol 4,5-bisphosphate (PIP2) into inositol triphosphate (IP3) and 1,2-diacylglicerol (DAG) with IP3 binding to its receptor, causing Ca2þ to be released from intracellular stores. Several proteins that are sensitive to calcium will be activated upon this release of calcium from intracellular stores including protein kinase C (PKC) (Sheldahl et al., 1999), calcium-calmodulin–dependent kinase II (Ku¨ hl et al., 2000b), and/or calcineurin (Saneyoshi et al., 2002). RNA overexpression studies in zebrafish and Xenopus embryos (Ku¨hl et al., 2000b; Saneyoshi et al., 2002; Sheldahl et al., 1999; Slusarski et al., 1997a,b) were performed in trying to define the Wnt/Ca2þ pathway of intracellular signaling, yet because Ca2þ represents a fast response upon ligand/receptor interaction, a more direct model to study the signaling pathway was needed. Frizzled chimeric receptors were created that included

204

Woodard and Rosado

intracellular Frizzled domains and extracellular/transmembrane segments of the 2-adrenergic receptor (DeCostanzo et al., 2002; Ku¨ hl et al., 2000b; Liu et al., 1999). They were used to show that when you stimulate the beta2adrenergic/Rfz-2 chimeric receptor, Wnt signaling occurs, resulting in a quick release of calcium from intracellular stores (Ku¨ hl et al., 2000a). Likewise, further evidence of this Wnt/calcium pathway was shown in the presence of Wnt-11-treated medium but not in control medium, resulting in stimulation of calcium-calmodulin–dependent kinase II (Pandur et al., 2002). Further conclusive experiments would involve using protein-purified Wnt ligands to demonstrate the dependence of calcium release upon Wnt binding.

References Ahumada, A., Slusarski, D. C., Liu, X., Moon, R. T., Malbon, C. C., and Wang, H. (2002). Signaling of rat frizzled-2 through phosphodiesterase and cyclic GMP. Science 298, 2006–2010. Amundson, J., and Clapham, D. (1993). Calcium waves. Curr. Opin. Neurobiol. 3, 375–382. Ault, K. T., Durmowicz, G., Galione, A., Harger, P. L., and Busa, W. B. (1996). Modulation of Xenopus embryo mesoderm-specific gene expression and dorsoanterior patterning by receptors that activate the phosphatidylinositol cycle signal transduction pathway. Development 122, 2033–2041. Baehr, W., Morita, E. A., Swanson, R. J., and Applebury, M. L. (1982). Characterization of bovine rod outer segment G-protein. J. Biol. Chem. 257, 6452–6460. Bennett, W. S., Jr., and Steitz, T. A. (1978). Glucose-induced conformational change in yeast hexokinase. Proc. Natl. Acad. Sci. USA 75, 4848–4852. Berstein, G., Blank, J. L., Smrcka, A. V., Higashijima, T., Sternweis, P. C., Exton, J. H., and Ross, E. M. (1992). Reconstitution of agonist-stimulated phosphatidylinositol 4,5-bisphosphate hydrolysis using purified m1 muscarinic receptor, Gq/11, and phospholipase C-beta 1. J. Biol. Chem. 267, 8081–8088. Berridge, M. J. (1995). Capacitative calcium entry. Biochem. J. 312, 1–11. Berridge, M. J., and Bootman, M. D. (1996). Calcium signaling. In ‘‘Signal Transduction’’ (C.-H. Heldin and M. Stanley Thornes Purton, Eds.), pp. 205–223. Chestenham. Berridge, M. J., Lipp, P., and Bootman, M. D. (2000). The versatility and universality of calcium signalling. Nat. Rev. 1, 11–21. Blaustein, M. P., and Lederer, W. J. (1999). Sodium/calcium exchange: Its physiological implications. Physiol. Rev. 79, 763–854. Bootman, M. D., and Lipp, P. (1999). Calcium signalling: Ringing changes to the ‘‘bell-shaped curve’’. Curr. Biol. 9, 876–878. Bosanac, I., Alattia, J. R., Mal, T. K., Chan, J., Talarico, S., Tong, F. K., Tong, K. I., Yoshikawa, F., Furuichi, T., Iwai, M., Michikawa, T., Mikoshiba, K., and Ikura, M. (2002). Structure of the inositol 1,4,5-trisphosphate receptor binding core in complex with its ligand. Nature 420, 696–700. Brini, M. (2003). Ca2þ signalling in mitochondria: Mechanism and role in physiology and pathology. Cell Calcium 34, 399–405. Brown, B. L., Walker, S. W., and Tomlinson, S. (1985). Calcium calmodulin and hormone secretion. Clin. Endocrinol. 23, 201–218.


205

Bu¨ nemann, M., and Hosey, M. M. (1999). G-protein coupled receptor kinases as modulators of G-protein signalling. J. Physiol. 517, 5–23. Buonanno, A., and Fields, R. D. (1999). Gene regulation by patterned electrical activity during neural and skeletal muscle development. Curr. Opin. Neurobiol. 9, 110–120. Cai, X., and Lytton, J. (2004). Molecular cloning of a sixth member of the Kþ-dependent Naþ/ Ca2þ exchanger gene family, NCKX6. J. Biol. Chem. 279, 5867–5876. Camello-Almaraz, C., Salido, G. M., Pariente, J. A., and Camello, P. J. (2002). Role of mitochondria in Ca2þ oscillation and shape of Ca2þ signals in pancreatic acinar cells. Biochem. Pharmacol. 63, 283–292. Cancela, J. M. (2001). Specific Ca2þ signaling evoked by cholecystokinin and acetylcholine: The roles of NAADP, cADPR, and IP3. Annu. Rev. Physiol. 63, 99–117. Cancela, J. M., Churchill, G. C., and Galione, A. (1999). Coordination of agonist-induced Ca2þ-signalling patterns by NAADP in pancreatic acinar cells. Nature 398, 74–76. Cancela, J. M., Gerasimenko, O. V., Gerasimenko, J. V., Tepikin, A. V., and Petersen, O. H. (2000). Two diVerent but converging messenger pathways to intracellular Ca2þ release: The roles of nicotinic acid adenine dinucleotide phosphate, cyclic ADP-ribose, and inositol trisphosphate. EMBO J. 19, 2549–2557. Carafoli, E. (1992). Calcium pump of the plasma membrane. Physiol. Rev. 71, 283–292. Cavallini, L., Coassin, M., and Alexandre, A. (1995). Two classes of agonist-sensitive Ca2þ stores in platelets, as identified by their diVerential sensitivity to 2,5-di-(tert-butyl)-1,4benzohydroquinone and thapsigargin. Biochem. J. 310, 449–452. Churchill, G. C., and Galione, A. (2001). NAADP induces Ca2þ oscillations via a two-pool mechanism by priming IP3- and cADPR-sensitive Ca2þ stores. EMBO J. 20, 2666–2671. Clapham, D. E. (1995). Calcium signaling. Cell 80, 259–268. Cullen, P. J. (1998). Bridging the GAP in inositol 1,3,4,5-tetrakisphosphate signalling. Biochim. Biophys. Acta 1436, 35–47. DeCostanzo, A. J., Huang, X-P., Wang, H., and Malbon, C. C. (2002). The Frizzled-1/beta2adrenergic receptor chimera: Pharmacological properties of a unique G protein-linked receptor. Naunyn-Schmiedeberg’s Arch. Pharmacol. 365, 341–348. De Lean, A., Stadel, J. M., and Lefkowitz, R. J. (1980). A ternary complex model explains the agonist-specific binding properties of the adenylate cyclase-coupled beta-adrenergic receptor. J. Biol. Chem. 255, 7108–7117. De Pont, J. J. H. H. M., and Fleuren-Jakobs, A. M. M. (1984). Synergistic eVect of A23187 and a phorbol ester on amylase secretion from rabbit pancreatic acini. FEBS Lett. 170, 64–68. Dettbarn, C., Betto, R., Salviati, G., Sabbadini, R., and Palade, P. (1995). Involvement of ryanodine receptors in sphingosylphosphorylcholine-induced calcium release from brain microsomes. Brain Res. 669, 79–85. Dixon, R. A., Kobilka, B. K., Strader, D. J., Benovic, J. L., Dohlman, H. G., Frielle, T., Bolanowski, M. A., Bennett, C. D., Rands, E., Diehl, R. E., Mumford, R. A., Slater, E. E., Sigal, I. S., Caron, M. G., Lefkowitz, R. J., and Strader, C. D. (1986). Cloning of the gene and cDNA for mammalian beta-adrenergic receptor and homology with rhodopsin. Nature 321, 75–79. Dong, H., Light, P. E., French, R. J., and Lytton, J. (2001). Electrophysiological characterization and ionic stoichiometry of the rat brain Kþ-dependent Naþ-Ca2þ exchanger, NCKX2. J. Biol. Chem. 276, 25919–25928. Exton, J. H. (1997). New developments in phospholipase D. J. Biol. Chem. 272, 15579–15582. Farahbakhsh, Z. T., Hideg, K., and Hubbell, W. L. (1993). Photoactivated conformational changes in rhodopsin: A time-resolved spin label study. Science 262, 1416–1419. Ferguson, S. S. (2001). Evolving concepts in G protein-coupled receptor endocytosis: The role in receptor desensitization and signaling. Pharmacol. Rev. 53, 1–24.

206

Woodard and Rosado

Fill, M., and Coronado, R. (1988). Ryanodine receptor channel of sarcoplasmic reticulum. Trends Neurosci. 11, 453–457. Galione, A. (1994). Cyclic ADP-ribose, the ADP-ribosyl cyclase pathway and calcium signalling. Mol. Cell. Endocrinol. 98, 125–131. Gether, U., and Kobilka, B. K. (1998). G protein-coupled receptors. II. Mechanism of agonist activation. J. Biol. Chem. 273, 17979–17982. Gether, U., Lin, S., and Kobilka, B. K. (1995). Fluorescent labeling of purified 2 adrenergic receptor. Evidence for ligand-specific conformational changes. J. Biol. Chem. 270, 28268–28275. Gillman, A. G. (1987). G proteins: Transducers of receptor-generated signals. Annu. Rev. Biochem. 56, 615–649. Gonzalez, A., and Salido, G. M. (2001). Participation of mitochondria in calcium signalling in the exocrine pancreas. J. Physiol. Biochem. 57, 331–339. Guerini, D. (1998). The significance of the isoforms of plasma membrane calcium ATPase. Cell Tissue Res. 292, 191–197. Habas, R., Dawid, I. B., and He, X. (2003). Coactivation of Rac and Rho by Wnt/Frizzled signaling is required for vertebrate gastrulation. Genes Dev. 17, 295–309. Hamada, K., Terauchi, A., and Mikoshiba, K. (2003). Three-dimensional rearrangements within inositol 1,4,5-trisphosphate receptor by calcium. J. Biol. Chem. 278, 52881–52889. Herbert, M., Gillespie, J. I., and Murdoch, A. P. (1997). Development of calcium signalling mechanisms during maturation of human oocytes. Mol. Hum. Reprod. 3, 965–973. Ichas, F., Jouaville, L. S., and Mazat, J. P. (1997). Mitochondria are excitable organelles capable of generating and conveying electrical and calcium signals. Cell 89, 1145–1153. Irvine, R. F. (1990). ‘‘Quantal’’ Ca2þ release and the control of Ca2þ entry by inositol phosphates—A possible mechanism. FEBS Lett. 263, 5–9. Ji, T. H., Grossmann, M., and Ji, I. (1998). G protein-coupled receptors. I. Diversity of receptor-ligand interactions. J. Biol. Chem. 273, 17299–17302. Ji, T. H., Murdoch, W. J., and Ji, I. (1995). Activation membrane-receptors. Endocrine 3, 187–194. Kim, M. J., Kim, E., Ryu, S. H., and Suh, P. G. (2000). The mechanism of phospholipase C-gamma1 regulation. Exp. Mol. Med. 32, 101–109. Kiseljov, K., Shin, D. M., Luo, X., Ko, S. B., and Muallem, S. (2002). Ca2þ signaling in polarized exocrine cells. Adv. Exp. Med. Biol. 504, 175–183. Kleuss, C., Scherubl, H., Hescheler, J., Schultz, G., and Wittig, B. (1992). DiVerent beta-subunits determine G-protein interaction with transmembrane receptors. Nature 358, 424–426. Koshland, D. E., Jr., and Neet, K. E. (1968). The catalytic and regulatory properties of enzymes. Annu. Rev. Biochem. 37, 359–410. Ku¨ hl, M. (Ed.) (2003). In ‘‘Wnt Signaling in Development.’’ Landes Biosiences and Kluwer Academic/Plenum Publishers, Georgetown, TX. Ku¨ hl, M., Sheldahl, L. C., Park, M., Miller, J. R., and Moon, R. T. (2000a). The Wnt/Ca2þ pathway. Trends Genet. 16, 279–283. Ku¨ hl, M., Sheldahl, L. C., Malbon, C. C., and Moon, R. T. (2000b). Ca2þ/calmodulindependent protein kinase II is stimulated by Wnt and Frizzled homologs and promotes ventral cell fates in Xenopus. J. Biol. Chem. 275, 12701–12711. Lajas, A. I., Sierra, V., Camello, P. J., Salido, G. M., and Pariente, J. A. (2001). Vanadate inhibits the calcium extrusion in rat pancreatic acinar cells. Cell. Signal. 13, 451–456. Lee, H. C., and Aarhus, R. (1995). A derivative of NADP mobilizes calcium stores insensitive to inositol trisphosphate and cyclic ADP-ribose. J. Biol. Chem. 270, 2152–2157. Lefkowitz, R. J. (1998). G protein-coupled receptors. III. New roles for receptor kinases and beta-arrestins in receptor signaling and desensitization. J. Biol. Chem. 273, 18677–18680.


207

Li, Z., Matsuoka, S., Hryshko, L. V., Nicoll, D. A., Bersohn, M. M., Burke, E. P., Lifton, R. P., and Philipson, K. D. (1994). Cloning of the NCX2 isoform of the plasma membrane NaþCa2þ exchanger. J. Biol. Chem. 269, 17434–17439. Liu, L., Hammar, K., Smith, P. J., Inoue, S., and Keefe, D. L. (2001). Mitochondrial modulation of calcium signaling at the initiation of development. Cell Calcium 30, 423–433. Liu, X., Liu, T., Slusarski, D. C., Yang-Snyder, J., Malbon, C. C., Moon, R. T., and Wang, H. (1999). Activation of a Frizzled-2/beta-adrenergic receptor chimera promotes Wnt signaling and diVerentiation of mouse F9 teratocarcinoma cells via Galphao and Galphat. Proc. Natl. Acad. Sci. USA 96, 14383–14388. Lu¨ ckhoV, A., and Clapham, D. E. (1992). Inositol 1,3,4,5-tetrakisphosphate activates an endothelial Ca2þ-permeable channel. Nature 335, 356–358. Lupas, A. N., Lupas, J. M., and Stock, J. B. (1992). Do G protein subunits associate via a threestranded coiled coil? FEBS Lett. 314, 105–108. Ma, H. T., Patterson, R. L., van Rossum, D. B., Birnbaumer, L., Mikoshiba, K., and Gill, D. L. (2000). Requirement of the inositol triphosphate receptor for activation of store-operated Ca2þ channels. Science 287, 1647–1651. Mao, C., Kim, S. H., AlmenoV, J. S., Rudner, X. L., Kearney, D. M., and Kindman, L. A. (1996). Molecular cloning and characterization of SCaMPER, a sphingolipid Ca2þ release-mediating protein from endoplasmic reticulum. Proc. Natl. Acad. Sci. USA 93, 1993–1996. Matsuda, T., Takuma, K., and Baba, A. (1997). Naþ-Ca2þ exchanger: Physiology and pharmacology. Jpn. J. Pharmacol. 74, 1–20. McCleskey, E. W. (1994). Calcium channels: Cellular roles and molecular mechanisms. Curr. Opinion Neurobiol. 4, 304–312. McPherson, P. S., and Campbell, K. P. (1993). The ryanodine receptor/Ca2þ release channel. J. Biol. Chem. 268, 13765–13768. Means, A. R. (1994). Calcium, calmodulin, and cell cycle regulation. FEBS Lett. 347, 1–4. Mikoshiba, K. (1993). Inositol 1,4,5-trisphosphate receptor. Trends. Pharmacol. Sci. 14, 86–89. Mikoshiba, K. (1997). The InsP3 receptor and intracellular Ca2þ signaling. Curr. Opin. Neurobiol. 7, 339–345. Mogami, H., Gardner, J., Gerasimenko, O. V., Camello, P., Petersen, O. H., and Tepikin, A. V. (1999). Calcium binding capacity of the cytosol and endoplasmic reticulum of mouse pancreatic acinar cells. J. Physiol. 518, 463–467. Monkawa, T., Miyawaki, A., Sugiyama, T., Yoneshima, H., Yamamoto-Hino, M., Furuichi, T., Saruta, T., Hasegawa, M., and Mikoshiba, K. (1995). Heterotetrameric complex formation of inositol 1,4,5-trisphosphate receptor subunits. J. Biol. Chem. 270, 14700–14704. Muallem, S., and Wilkie, T. M. (1999). G protein-dependent Ca2þ signaling complexes in polarized cells. Cell Calcium 26, 173–180. Mumby, S. M., Casey, P. J., Gilman, A. G., Gutowski, S., and Sternweis, P. C. (1990). G protein gamma subunits contain a 20-carbon isoprenoid. Proc. Natl. Acad. Sci. USA 87, 5873–5877. Nathan, J., and Hogness, D. S. (1983). Isolation, sequence analysis, and intron–exon arrangement of the gene encoding bovine rhodopsin. Cell 34, 807–814. Nathanson, M. H., Padfield, P. J., O’Sullivan, A. J., Burgstahler, A. D., and Jamieson, J. D. (1992). Mechanism of Ca2þ wave propagation in pancreatic acinar cells. J. Biol. Chem. 267, 18118–18121. Neher, E., and Augustine, G. J. (1992). Calcium gradients and buVers in bovine chromaYn cells. J. Physiol. 450, 273–301. Nicoll, D. A., Longoni, S., and Philipson, K. D. (1990). Molecular cloning and functional expression of the cardiac sarcolemmal Naþ-Ca2þ exchanger. Science 250, 562–565.

208

Woodard and Rosado

Nicoll, D. A., Quednau, B. D., Qui, Z., Xia, Y. R., Lusis, A. J., and Philipson, K. D. (1996). Cloning of a third mammalian Naþ-Ca2þ exchanger, NCX3. J. Biol. Chem. 271, 24914–24921. Pandur, P., La¨ sche, M., Eisenberg, L. M., and Ku¨ hl, M. (2002). Wnt-11 activation of a noncanonical Wnt signaling pathway is required for cardiogenesis. Nature 418, 636–641. Parekh, A. B., and Penner, R. (1997). Store depletion and calcium influx. Physiol. Rev. 77, 901–930. Pariente, J. A., Lajas, A. I., Pozo, M. J., Camello, P. J., and Salido, G. M. (1999). Oxidizing eVects of vanadate on calcium mobilization and amylase release in rat pancreatic acinar cells. Biochem. Pharmacol. 58, 77–84. Pariente, J. A., Camello, C., Camello, P. J., and Salido, G. M. (2001). Release of calcium from mitochondrial and nonmitochondrial intracellular stores in mouse pancreatic acinar cells by hydrogen peroxide. J. Membr. Biol. 179, 27–35. Pariente, J. A., Redondo, P. C., Granados, M. P., Lajas, A. I., Gonzalez, A., Rosado, J. A., and Salido, G. M. (2003). Calcium signaling in non-excitable cells. E. C. Qua. L. 1, 29–43. Patel, S. (2004). NAADP-induced Ca2þ release—A new signalling pathway. Biol. Cell 96, 19–28. Patterson, R. L., van Rossum, D. B., and Gill, D. L. (1999). Store-operated Ca2þ entry: Evidence for a secretion-like coupling model. Cell 98, 487–499. Pedersen, P. L., and Carafoli, E. (1987). Ion motive ATPase. I. Ubiquity, properties, and significance for cell function. Trends Biochem. Sci. 12, 146–150. Petersen, O. H., Petersen, C. C. H., and Kasai, H. (1994). Calcium and hormone action. Annu. Rev. Physiol. 56, 297–319. Rebecchi, M. J., and Pentyala, S. N. (2000). Structure, function, and control of phosphoinositide-specific phospholipase C. Physiol. Rev. 80, 1291–1335. Reembold, C. M. (1992). Regulation of contraction and relaxation in arterial smooth muscle. Hypertension 20, 129–137. Redondo, P. C., Lajas, A. I., Salido, G. M., Gonzalez, A., Rosado, J. A., and Pariente, J. A. (2003). Evidence for secretion-like coupling involving pp60src in the activation and maintenance of store-mediated Ca2þ entry in mouse pancreatic acinar cells. Biochem. J. 370, 255–263. Reitstetter, R., and Yool, A. J. (1998). Morphological consequences of altered calciumdependent transmembrane signaling on the development of cultured cerebellar Purkinje neurons. Brain. Res. Dev. Brain Res. 107, 165–167. Rodbell, M. (1980). The role of hormone receptors and GTP-regulatory proteins in membrane transduction. Nature 284, 17–22. Rosado, J. A., Brownlow, S. L., and Sage, S. O. (2002). Endogenously expressed Trp1 is involved in store-mediated Ca2þ entry by conformational coupling in human platelets. J. Biol. Chem. 277, 42157–42163. Rosado, J. A., Jenner, S., and Sage, S. O. (2000a). A role for the actin cytoskeleton in the initiation and maintenance of store-mediated calcium entry in human platelets. Evidence for conformational coupling. J. Biol. Chem. 275, 7527–7533. Rosado, J. A., Graves, D., and Sage, S. O. (2000b). Tyrosine kinases activate store-mediated Ca2þ entry in human platelets through the reorganization of the actin cytoskeleton. Biochem. J. 351, 429–437. Rosado, J. A., and Sage, S. O. (2000a). Farnesylcysteine analogues inhibit store-regulated Ca2þ entry in human platelets: Evidence for involvement of small GTP-binding proteins and actin cytoskeleton. Biochem. J. 347, 183–192. Rosado, J. A., and Sage, S. O. (2000b). Regulation of plasma membrane Ca2þ-ATPase by small GTPases and phosphoinositides in human platelets. J. Biol. Chem. 275, 19529–19535.


209

Rosado, J. A., and Sage, S. O. (2000c). Protein kinase C activates non-capacitative calcium entry in human platelets. J. Physiol. 529, 159–169. Rosado, J. A., and Sage, S. O. (2000d). Coupling between inositol 1,4,5-trisphosphate receptors and human transient receptor potential channel 1 when intracellular Ca2þ stores are depleted. Biochem. J. 350, 631–635. Sage, S. O. (1992). Three routes for receptor-mediated Ca2þ entry. Curr. Biol. 2, 312–314. Sanderson, M. J., Charles, A. C., Boitano, S., and Dirksen, E. R. (1994). Mechanisms and function of intercellular calcium signaling. Mol. Cell. Endocrinol. 98, 173–187. Saneyoshi, T., Kume, S., Amasaki, Y., and Mikoshiba, K. (2002). The Wnt/calcium pathway activates NF-AT and promotes ventral cell fate in Xenopus embryos. Nature 417, 295–299. Santella, L., Kyozuka, K., Genazzani, A. A., De Riso, L., and Carafoli, E. (2000). Nicotinic acid adenine dinucleotide phosphate-induced Ca2þ release. Interactions among distinct Ca2þ mobilizing mechanisms in starfish oocytes. J. Biol. Chem. 275, 8301–8306. Sato, C., Hamada, K., Ogura, T., Miyazawa, A., Iwasaki, K., Hiroaki, Y., Tani, K., Terauchi, A., Fujiyoshi, Y., and Mikoshiba, K. (2004). Inositol 1,4,5-trisphosphate receptor contains multiple cavities and L-shaped ligand-binding domains. J. Mol. Biol. 336, 155–164. Sayers, L. G., Miyawaki, A., Muto, A., Takeshita, H., Yamamoto, A., Michikawa, T., Furuichi, T., and Mikoshiba, K. (1997). Intracellular targeting and homotetramer formation of a truncated inositol 1,4,5-trisphosphate receptor-green fluorescent protein chimera in Xenopus laevis oocytes: Evidence for the involvement of the transmembrane spanning domain in endoplasmic reticulum targeting and homotetramer complex formation. Biochem. J. 323, 273–280. Sheldahl, L. C., Park, M., Malbon, C. C., and Moon, R. T. (1999). Protein kinase C is diVerentially stimulated by Wnt and Frizzled homologs in a G-protein-dependent manner. Curr. Biol. 9, 695–698. Slusarski, D. C., Corces, V. G., and Moon, R. T. (1997a). Interaction of Wnt and a Frizzled homologue triggers G-protein-linked phosphatidylinositol signaling. Nature 390, 410–413. Slusarski, D. C., Yang-Snyder, J., Busa, W. B., and Moon, R. T. (1997b). Modulation of embryonic intracellular Ca2þ signaling by Wnt-5A. Dev. Biol. 182, 114–120. Somasundaram, B., Norman, J. C., and Mahaut-Smith, M. P. (1995). Primaquine, an inhibitor of vesicular transport, blocks the calcium-release-activated current in rat megakaryocytes. Biochem. J. 309, 725–729. Svoboda, P., Teisinger, J., Novotny, J., Bourova, L., Drmota, T., Hejnova, L., Moravcova, Z., Lisy, V. V., Rudajev, V., Stohr, J., Vokurkova, A., Svandova, I. I., and Durchankova, D. (2004). Biochemistry of transmembrane signaling mediated by trimeric G proteins. Physiol. Res. 53, 141–152. Streb, H., Irvine, R. F., Berridge, M. J., and Schulz, I. (1983). Release of Ca2þ from a nonmitochondrial intracellular store in pancreatic acinar cells by inositol-1,4,5-trisphosphate. Nature 306, 67–69. Strehler, E. E., and Zacharias, D. A. (2001). Role of alternative splicing in generating isoform diversity among plasma membrane calcium pumps. Physiol. Rev. 81, 21–50. Thorn, P., Lawrie, A. M., Smith, P. M., Gallacher, D. V., and Petersen, O. H. (1993). Local and global cytosolic Ca2þ oscillations in exocrine cells evoked by agonists and inositol trisphosphate. Cell 74, 661–668. Toescu, E. C., Lawry, A. M., Petersen, O. H., and Gallacher, D. V. (1992). Spatial and temporal distribution of agonist-evoked cytoplasmic Ca2þ signals in exocrine acinar cells analyzed by digital image microscopy. EMBO J. 11, 1623–1629. Tsien, R. W., Lipscombe, D., Madison, D., Bley, K., and Fox, A. (1995). Reflections on Ca2þchannel diversity, 1988–1994. Trends in Neurosciences 18, 52–54.

210

Woodard and Rosado

Tu, J. C., Xiao, B., Yuan, J. P., Lanahan, A. A., LeoVert, K., Li, M., Linden, D. J., and Worley, P. F. (1998). Homer binds a novel proline-rich motif and links group 1 metabotropic receptors with IP3 receptors. Neuron 21, 717–726. Vazquez, G., Lievremont, J. P., Bird, G., and Putney, J. W., Jr. (2001). Human Trp3 forms both inositol trisphosphate receptor-dependent and receptor-independent store-operated cation channels in DT40 avian B lymphocytes. Proc. Natl. Acad. Sci. USA 98, 11777–11782. Veeman, M. T., Axelrod, J. D., and Moon, R. T. (2003). A second canon: Functions and mechanisms review of beta-catenin-independent Wnt signaling. Developmental Cell 5, 1–20. Webb, S. E., and Miller, A. L. (2003). Calcium signalling during embryonic development. Nat. Rev. Mol. Cell. Biol. 4, 539–551. Webb, S. E., and Miller, A. L. (2000). Calcium signalling during zebrafish embryonic development. Bioessays 22, 113–123. Wictome, M., Henderson, I., Lee, A. G., and East, J. M. (1992). Mechanism of inhibition of the calcium pump of sarcoplasmic reticulum by thapsigargin. Biochem. J. 283, 525–529. Willert, K., Brown, J. D., Danenberg, E., Duncan, A. W., Weissman, I. L., Reya, T., Yates, J. R., and Nusse, R. (2003). Wnt proteins are lipid-modified and can act as stem cell growth factors. Nature 423, 448–452. Williams, J. A. (2001). Intracellular signaling mechanisms activated by cholecystokininregulating synthesis and secretion of digestive enzymes in pancreatic acinar cells. Annu. Rev. Physiol. 63, 77–97. Wodarz, A., and Nusse, R. (1998). Mechanisms of Wnt signaling in development. Annu. Rev. Cell Dev. Biol. 14, 59–88. Wuytack, F., Raeymaekers, L., De Smedt, H., Eggermont, J. A., Missiaen, L., Van Den Bosch, L., De Jaegere, S., Verboomen, H., Plessers, L., and Casteels, R. (1992). Ca2þ-transport ATPases and their regulation in muscle and brain. Ann. N. Y. Acad. Sci. 671, 82–91. Xie, Q., Zhang, Y., Zhai, C., and Bonanno, J. A. (2002). Calcium influx factor from cytochrome P-450 metabolism and secretion-like coupling mechanisms for capacitative calcium entry in corneal endothelial cells. J. Biol. Chem. 277, 16559–16566.

8

____________________________________________________________________________

Differential Functions of 14-3-3 Isoforms in Vertebrate Development Anthony J. Muslin and Jeffrey M. C. Lau Center for Cardiovascular Research Department of Medicine and Department of Cell Biology and Physiology Washington University School of Medicine St. Louis, Missouri 63110

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

Introduction to 14-3-3 Proteins Structure of 14-3-3 Proteins Regulation of 14-3-3 Protein Activity Molecular Interference by 14-3-3 Proteins Evolution of 14-3-3 Genes Role of 14-3-3 Proteins in Invertebrate Development Role of 14-3-3 Proteins in Amphibian Development Role of 14-3-3 Proteins in Mammalian Development Conclusions References

Overview 14-3-3 proteins are intracellular dimeric phosphoserine/threonine-binding molecules that are found in all eukaryotic organisms and that participate in developmental, signal transduction, checkpoint control, nutrient sensing, and cell survival pathways. Previous work established that 14-3-3 proteins are essential for survival in budding and fission yeast. In fungi, plants, and animals, there are multiple 14-3-3 proteins encoded by separate genes. Work has begun to uncover the unique developmental functions of individual 14-3-3 isoforms in vertebrate development. The specific roles of 14-3-3 family members in animal development will be discussed in this topical chapter. C 2005, Elsevier Inc.

I. Introduction to 14-3-3 Proteins The oddly-named 14-3-3 proteins were discovered in 1967 by Moore and Perez in a screen for abundant mammalian brain proteins (Moore et al., 1967). The discovery that 14-3-3 proteins bind to and modulate the activity Current Topics in Developmental Biology, Vol. 65 Copyright 2005, Elsevier Inc. All rights reserved. 0070-2153/05 $35.00

211

212

Muslin and Lau

of critical signaling and cell cycle proteins, including tryosine hydroxylase, protein kinase Cs (PKC), Raf-1, BCR/Abl, and Cdc25, in the late 1980s and early 1990s led to an explosion of research into this family of proteins. (Fu et al., 2000; Muslin and Xing, 2000). 14-3-3 proteins are ubiquitously expressed in eukaryotic organisms, and they are highly abundant in many— if not all—cell types. 14-3-3 monomers form constitutive homo- and heterodimers in cells (Aitken, 2002). In the budding yeast Saccharomyces cerevisiae, there are two 14-3-3 genes, BMH1 and BMH2, that are essential for survival in some—but not all— strains (Gelperin et al., 1995; van Heusden et al., 1995). Similarly, in the fission yeast Schizosaccharomyces pombe, there are two 14-3-3 genes, RAD24 and RAD25 (Ford et al., 1994). There are also two 14-3-3 genes in the worm Caenorrhabditis elegans, FTT-1 and FTT-2, and in the fruit fly Drosophila melanogaster, D14-3-3 " and D14-3-3 (Benton et al., 2002; Li et al., 1997; Morton et al., 2002; Tien et al., 1999; Wang and Shakes, 1997). In mammals and probably in most vertebrates, there are 7 diVerent 14-3-3 family member genes, called 14-3-3 , , ", , , , and (Aitken, 2002). In plants, there are even more 14-3-3 genes than in mammals, with 12 family members in Arabidopsis thaliana (Ferl, 2004). The large number of 14-3-3 family members in individual animal and plant species has led to speculation that the function of individual 14-3-3 proteins might be unique. Specific functions of 14-3-3 isoforms may be a by-product of their tissuespecific expression patterns, subcellular localization, individual regulatory mechanisms, or specific binding partners. In 1996, 14-3-3 proteins were found to be phosphoserine-binding proteins that specifically recognized the motif RSxpSxP, where pS is phosphoserine and x is any amino acid (Muslin et al., 1996). In 1997, an additional binding motif was discovered by peptide library screening: RxY/FxpSxP (YaVe et al., 1997). Subsequent work demonstrated that 14-3-3 proteins can also bind to phosphothreonine-containing motifs, and that 14-3-3 proteins can also bind, albeit with lesser aYnity, to degenerate phosphopeptides similar to the two ‘‘optimal’’ motifs (Aitken, 2002). Many intracellular proteins contain one or both of the optimal 14-3-3 binding motifs and the number of 14-3-3 binding partners is ever expanding. Indeed, well over 100 binding partners of 14-3-3 proteins have been identified to date. 14-3-3 family members bind to many diVerent protein kinases involved in signal transduction, such as PKCs, Raf-1, B-Raf, MEKK1–3, Ask1, and PCTAIRE (Fanger et al., 1998; Fantl et al., 1994; Freed et al., 1994; Fu et al., 1994; Ichimura et al., 1987; Isobe et al., 1991; Meller et al., 1996; Papin et al., 1996; Reuther et al., 1994; Sladeczek et al., 1997; Toker et al., 1990; Zhang et al., 1999). They also bind to a variety of scaVolding proteins involved in signal transduction, including KSR-1, IRS-1, Cbl, and p130 Cas (Craparo et al., 1997; Garcia-Guzman et al., 1999; Liu et al., 1996;

8. 14-3-3 Isoform Function in Vertebrate Development

213

Xing et al., 1997). Important cytoskeletal proteins bind to 14-3-3 family members, including vimentins, keratins, Tau, and Kif1C (Dorner et al., 1999; Hashiguchi et al., 2000; Liao and Omary, 1996; Tzivion et al., 2000). 14-3-3 proteins bind to cell cycle regulatory proteins, such as Cdc25 phosphatase, cyclinB1/cdc2, wee1, Chk1, and p53 (Chan et al., 1999; Conklin et al., 1995; Jiang et al., 2003; Lee et al., 2001; Wang et al., 2000; Waterman et al., 1998). Several transcription factors and associated proteins bind to 143-3 family members including NF-AT, forkhead family members, histone deacetylases, the TATA-box binding proteins TBP and TFIIB, and histone acetyl transferase 1 (Brunet et al., 1999; Chow and Davis, 2000; Grozinger and Schreiber, 2000; McKinsey et al., 2000; Pan et al., 1999). In addition, many proteins involved in the regulation of apoptosis and cell metabolism bind to 14-3-3 family members (Aitken, 2002; Fu et al., 2000; Muslin and Xing, 2000; YaVe, 2002). Several investigators have tried to determine whether individual 14-3-3 family members bind to diVerent phosphoserine-containing motifs; however, most evidence supports the notion that all 14-3-3 family members bind with similar aYnity to the two ‘‘optimal’’ motifs. Clearly, additional contact points exist between 14-3-3 isoforms and binding partners that are outside of the phosphoserine/phosphothreonine binding pocket. For example, PAR1 binds to D14-3-3 " in a phosphorylation-independent mechanism that leaves the phosphoserine/phosphothreonine binding pocket unoccupied (see Following text) (Morton et al., 2002). Therefore, although all 14-3-3 family members are phosphoserine binding proteins, they also have the ability to bind to unphosphorylated motifs via other domains that may be unique to particular isoforms.

II. Structure of 14-3-3 Proteins A crystallographic analysis of mammalian 14-3-3 and was performed several years ago and showed both isoforms to be highly helical dimeric proteins (Liu et al., 1995; Xiao et al., 1995). 14-3-3 monomers are composed of nine anti-parallel -helices with uncharacterized carboxy-terminal extensions. These monomers form homodimers with a large, deep channel in the center running the length of the dimer. Each wall of the channel contains an amphipathic groove. Dimerization is achieved by binding of the first (amino-terminal) -helix of one monomer to the second -helix of the other monomer. Co-crystal structures of 14-3-3 with a serine phosphorylated peptide showed that the peptide binds in an extended conformation within a single amphipathic groove (Petosa et al., 1998). The phosphoserine/phosphothreonine phosphate forms hydrogen and ionic bonds with three basic residues

214

Muslin and Lau

that are conserved in all 14-3-3 family members, K49, R56, and R127. These three basic residues form a basic pocket in a protein that is, in general, highly negatively charged. Each 14-3-3 dimer contains two amphipathic grooves and can therefore bind to two phosphoserine-containing peptides (Fig. 1). Comparison of diVerent mammalian 14-3-3 genes demonstrates that the residues that line the amphipathic grooves are highly conserved among all family members. On the other hand, the carboxy-terminal amino extensions of 14-3-3 family members are divergent, as are the residues on the outer surface of the binding pocket. More recently co-crystallization of 14-3-3 with N-acetyltransferase (AANAT) was performed (Obsil et al., 2001). AANAT catalyzes acetyl transfer from acetylcoenzyme A to serotonin, the penultimate step in the synthesis of melatonin. AANAT was co-crystallized with human 14-3-3 at a stoichiometry of 2:2. Each monomer of the 14-3-3 dimer binds one molecule of pAANAT. Phosphothreonine-31 of pAANAT points into the positively charged depression in the middle of the amphipathic groove of 14-3-3. 14-3-3 binding to AANAT alters the conformation of the AANAT dimer to increase its ability to bind substrate. In addition to phosphothreonine-31, there are additional residues in pAANAT that bind to 14-3-3 outside of the amphipathic groove. These additional contact points might explain how individual 14-3-3 family members might bind preferentially to specific target proteins.

Figure 1 Stick figure representation of a 14-3-3 dimer that is adapted from the work of Xiao (1995). Each circle represents one of the 9 alpha helices in the 14-3-3 monomer. The triangles are located in the docking positions of phosphoserine-containing motifs in interacting proteins.


215

III. Regulation of 14-3-3 Protein Activity Individual 14-3-3 family members diverge at their carboxy-termini, and their divergent areas may be the sites of unique regulatory mechanisms that contribute to family member-specific functions. In all of the crystal structures solved to date, the carboxyl-terminus of 14-3-3 has not been included, presumably, because it is disordered in structure. Interestingly, Aitken’s group showed that two mammalian isoforms, and , but not others, are phosphorylated at the carboxyl-terminus on threonine residues (Dubois et al., 1997). In addition, phosphorylation of threonine-233 of 14-3-3 rendered the protein unable to bind to Raf-1 kinase. One kinase that phosphorylates threonine 233 is casein kinase l (Dubois et al., 1997). The implication of this work is that two 14-3-3 family members, 14-3-3 and 14-3-3, but not others, might be regulated by kinase activity, and this may lead to specific biological activities of these proteins.

IV. Molecular Interference by 14-3-3 Proteins 14-3-3 proteins are primarily localized in the cytosol, but they may also be found at low levels in the nucleus in animal cells. In plant cells, 14-3-3 proteins are more often found in the nucleus, where they bind to, and modify the activity of, a variety of transcription factor complexes. 14-3-3 proteins do not have intrinsic enzymatic activity, nor do they have specific subcellular targeting sequences, such as nuclear localization or mitochondrial localization motifs. Based on these properties, the ‘‘molecular interference’’ model of 14-3-3 protein action was elaborated (Fig. 2) (Muslin and Xing, 2000). In this model, 14-3-3 proteins do not provide intrinsic information about localization or enzymatic activity to their binding partners. The specific consequence of a 14-3-3-to-binding partner interaction is solely determined by the localization of the phosphoserine-containing motif(s) relative to other domains in the binding partner. The ‘‘molecular interference’’ model of 14-33 protein action does not exclude the possibility that 14-3-3-encoding mRNA might be specifically localized in early development (see following text). The molecular interference model explains the action of 14-3-3 proteins on a variety of binding partners (Muslin and Xing, 2000). For example, placement of a RSxSxP or RxY/FxSxP motif near a nuclear localization signal (NLS) creates a method for regulating the localization of the partner protein that depends on its phosphorylation status, because binding of the 60 kDa acidic 14-3-3 dimer interferes with the activity of a nearby NLS (Fig. 2). This mechanism appears to regulate the subcellular localization of a variety of transcription regulatory proteins and cell cycle proteins, including

216

Muslin and Lau

Figure 2 Schematic representation of ‘‘molecular interference’’ by 14-3-3 proteins. Left panel, a hypothetical transcription factor with a nuclear localization signal (NLS) positioned adjacent to a 14-3-3 binding motif (RSxSxP). When transcription factor is unphosphorylated on the serine residue within the 14-3-3 binding motif, the NLS is exposed, and the protein translocates into the nucleus. Right panel, when the 14-3-3 binding site is phosphorylated on serine (RSxpSxP), a 14-3-3 dimer binds and interferes with the NLS. The transcription factor is unable to translocate into the nucleus, and is localized in the cytosol.

histone deacetylases (HDACs), FKHRL1 (FOXO3b), and CDC25 phosphatase (Giles et al., 2003; Graves et al., 2001; Kao et al., 2001; Muslin and Xing, 2000; Van Der Heide et al., 2004). Alternatively, placement of the RSxSxP motif near the mitochondrial localization motif of the pro-apoptotic protein BAD creates a mechanism by which phosphorylation regulates the ability of BAD to localize in the mitochondrial membrane and interact with Bcl-2 or Bcl-xL (Zha et al., 1996). Furthermore, placement of an RSxSxP near the plasma-membrane localization domain of the MAPK cascade scaVold KSR-1 creates a mechanism by which phosphorylation of KSR-1 by the serine/threonine kinase C-TAK1 leads to 14-3-3 protein binding that prevents plasma membrane localization and scaVold activity (Muller et al., 2001).


217

V. Evolution of 14-3-3 Genes Wang and Shakes performed a detailed alignment and tree construction for more than 50 14-3-3 genes (Wang and Shakes, 1996). Their analysis demonstrated that there are five highly conserved sequence blocks in all 14-3-3 genes. These conserved sequence blocks correlate with alpha helices 3, 5, 7, and 9. Interestingly, helices 3, 5, and 7 are directly involved in binding to phosphoserine/phosphothreonine-containing peptide motifs. Vertebrate 14-3-3 "’s are more similar to plant and yeast isoforms than to other vertebrate isoforms, and therefore mammalian 14-3-3 " may be more similar to the ancestral protein. Individual mammalian 14-3-3 isoforms are more similar to their corresponding isoforms from other mammals than they are to diVerent isoforms from the same species. For example, rat 14-3-3 is more similar to human 14-3-3 than it is to rat 14-3-3 or rat 14-3-3 . Divergence of the seven mammalian isoforms occurred before the divergence of mammals and may have occurred before the divergence of vertebrates (Wang and Shakes, 1996).

VI. Role of 14-3-3 Proteins in Invertebrate Development The role of 14-3-3 proteins in the development of Caenorrhabditis elegans is well studied by analysis of mutants and by use of RNAi. There are two 14-33 genes in worms, ftt-1 (par-5) and ftt2 (Morton et al., 2002; Wang and Shakes, 1997). The ftt-1 gene is highly expressed in early embryos as maternal transcript in a symmetrical pattern, but expression decreases markedly in later development. In contrast, ftt-2 is undetectable in 1-cell embryos and is first detected in 2-cell or 4-cell embryos, suggesting that it is an early zygotic transcript. The ftt-2 gene is expressed at high levels in the posterior region of post-proliferative embryos. In a search for genes involved in the asymmetric first cell division of C. elegans embryos that gives rise to cells of diVerent sizes called AB and P1, Morton identified a gene, par-5, that was identical to ftt-1 (Morton et al., 2002). Embryos with reduced ftt-1, as a consequence of mutation or RNAi injection, have a markedly abnormal asymmetric first cell division, with a pseudocleavage furrow that is markedly diVerent from wild type (Morton et al., 2002). Furthermore, embryos with reduced ftt-1 have defects in cleavage orientation and timing of the second cell division. Later in development, embryos with reduced ftt-1 fail to undergo morphogenesis. In ftt-1 mutant embryos, MEX-5 and P granules, early markers of embryonic polarity, fail to localize normally to the poles of the embryo. Also, embryos with reduced ftt-1 have mislocalized PAR-1 and PAR-2 proteins. Therefore,

218

Muslin and Lau

FTT-1 protein regulates early asymmetric cell divisions in worms, but there is no evidence that FTT-1 is asymmetrically localized itself. In contrast to FTT-1, there is no evidence that FTT-2 regulates the early asymmetric cell divisions of C. elegans. Injection of embryos with RNAi for ftt-2 did not aVect early asymmetric cell divisions and resulted in a low level of lethality (

Current Topics in Developmental Biology, Volume 65 (Current Topics in Developmental Biology)

Current Topics Developmental Biology (Current Topics in Developmental Biology)

Current Topics Developmental Biology (Current Topics in Developmental Biology)