Current Topics Developmental Biology (Current Topics in Developmental Biology)

Current Topics in Developmental BioIogy Volume 41

Series Editors Roger A. Pedersen

and

Reproductive Genetics Division Department of Obstetrics, Gynecology, and Reproductive Sciences University of California San Francisco, California 94143

Gerald P. Schatten Departments of Obstetrics-Gynecology and Cell and Developmental Biology Oregon Regional Primate Research Center Oregon Health Sciences University Beaverton, Oregon 97006-3499

Editorial Board Peter Gruss Max Planck Institute of Biophysical Chemistry, Gottingen, Germany

Philip lngham 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

Current Topics in Developmental Biology Volume 41 Edited by

Roger A. Pedersen Reproductive Genetics Division Department of Obstetrics, Gynecology, and Reproductive Sciences University of California San Francisco, California

Gerald P. Schatten Departments of Obstetrics-Gynecology and Cell and Developmental Biology Oregon Regional Primate Research Center Oregon Health Sciences University Bea verton, Oregon

Academic Press San Diego

London Boston

New York

Sydney Tokyo Toronto

Front cover photograph: Expression patterns of Tbx5 (left) and Tbx4 (right) in stage 23 chick embryo forelimbs and hindlimbs, respectively.

This book is printed

on acid-free paper.

@

Copyright 0 1999 by ACADEMIC PRESS All Rights Reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the Publisher. The appearance of the code at the bottom of the first page of a chapter in this book indicates the Publisher's consent that copies of the chapter may be made for personal or internal use of specific clients. This consent is given on the condition, however, that the copier pay the stated per copy fee through the Copyright Clearance Center, Inc. (222 Rosewood Drive, Danvers, Massachusetts 01923). for copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for creating new collective works, or for resale. Copy fees for pre-1999 chapters are as shown on the title pages. If no fee code appears on the title page, the copy fee is the same as for current chapters. 0070-2153/99 $25.00

Academic Press a division of Harcourt Brace & Company 525 B Street, Suite 1900, San Diego, California 92101-4495, USA http://www.apnet.com Academic Press 24-28 Oval Road, London NWI 7DX, UK http://www.hbuk.co.uk/ap/ International Standard Book Number: 0-12-153141-4 PRINTED IN THE UNITED STATES OF AMERICA 98 99 0 0 0 1 02 0 3 B B 9 8 7 6

5

4

3 2 1

Contents

Contributors Preface xi

ix

1 Pattern Formation in Zebrafish-Fruitful and Genetics

liaisons between Embryology

L ilianna Solnica -Krezel

I. Establishment of the Dorsoventral Polarity during Zebrafish Cleavage Stages 2 11. Establishment of the Dorsal Blastula Organizer (Nieuwkoop Center) 111. Induction of the Gastrula Organizer by the Blastula Organizing Center IV. Structure and Function of the Dorsal Gastrula Organizer in Zebrafish 22 V. Coordination of Gastrulation Movements VI. Conclusions 28 References 29

4

7 9

2 Molecular and Cellular Basis of Pattern Formation during Vertebrate limb Development Jennifer K. Ng, Koji Tamura, Dirk Buscher, and Juan Carlos Izpislja-Belmonte

I. Introduction 38 11. The Proximal-Distal Axis 39 111. The Anterior-Posterior Axis 46 52 IV. The Dorsal-Ventral Axis 59 V. Conclusions References 60

V

vi

Contents

3 Wise, Winsome, or Weird? Mechanisms of Sperm Storage in Female Animals Deborah M. Neubaum and Mariana F. Wolfner

I. 11. 111. IV. V. VI. VII.

Introduction 68 Mechanisms of Sperm Storage 73 80 The Fate of Unstored Sperm and Secretions Sperm inside the Storage Organs 80 85 Molecules Important for Sperm Storage 88 The Adaptive Significance of Sperm Storage 89 Conclusions References 90

4 Developmental Genetics of Caenorhabdifis eregans Sex Determination Patricia E. Ku wabara I. 11. 111. IV. V. VI. VII. VIII. IX. X. XI.

Introduction 100 104 The Role of the X:A Ratio Genetic Analysis of Sex Determination 108 1 11 Molecular Analysis of Sex Determination TRA-I Targets and the Conservation of Sex-Determining Mechanisms How to Count Chromosomes: The X:A Ratio Revisited 117 119 Analysis of Germ-Line Sex Determination 120 The Hermaphrodite Sperm-Oocyte Decision Phylogenetic Comparisons and the Evolution of Sex-Determining Genes Unresolved Questions 124 Future Perspectives 126 References 127

L

5 Petal and Stamen Development Vivian F. Irish

I. 11. 111. IV. V. VI. VII.

Introduction 133 Petal and Stamen Ontogeny 135 Genes Controlling the Specification of Petal and Stamen Identities 143 Differentiation of Petals Differentiation of Stamens 148 Coordination of Gene Expression and Tissue Differentiation 152 Summary 153 References 154

138

116

123

vii

Contents

6 Gonadotropin-Induced Resumption of Oocyte Meiosis and MeiosisActivating Sterols Claus Yding Andersen, Mogens Baltsen, and Anne Grete Byskov

I. Introduction

163

165 111. Possible Signal Transduction Pathways Involved in Resumption of Meiosis 169 IV. Hypothesis of a Role for MAS in Gonadotropin-Induced Resumption 175 of Meiosis 178 V. Possible Implications for Fertility References 179 11. Gonadotropin-Induced Resumption of Meiosis

Index

187

Contents of Previous Volumes

193

This Page Intentionally Left Blank

Contributors

Numbers in puretitheses indii.ute the pugrs

011

which the authors' contribrctions begin.

Mogens Baltsen (163), Laboratory of Reproductive Biology, The Juliane Marie Centre for Children, Women, and Reproduction, University Hospital of Copenhagen, DK-2 100 Copenhagen, Denmark Dirk Biischer (37), Gene Expression Laboratory, The Salk Institute, La Jolla, California 92037 Anne Grete Byskov (1 63), Laboratory of Reproductive Biology, The Juliane Marie Centre for Children, Women, and Reproduction, University Hospital of Copenhagen, DK-2 100 Copenhagen, Denmark Vivian F. Irish (133), Department of Molecular, Cellular and Developmental Biology, Yale University, New Haven, Connecticut 06520 Juan Carlos Izpisua-Belmonte (37), Gene Expression Laboratory, The Salk Institute, La Jolla, California 92037 Patricia E. Kuwabara (99), Medical Research Council, Laboratory of Molecular Biology, Cambridge CB2 2QH, United Kingdom Deborah M. Neubaum" (67), Section of Genetics and Development, Cornell University, Ithaca, New York 14853 Jennifer K. Ng (37), Gene Expression Laboratory, The Salk Institute, La Jolla, California 92037 Lilianna Solnica-Krezel ( 1 ), Department of Molecular Biology, Vanderbilt University, Nashville, Tennessee 37232 Koji Tamura (37), Gene Expression Laboratory, The Salk Institute, La Jolla, California 92037 Mariana F. Wolfner (67), Section of Genetics and Development, Cornell University, Ithaca, New York 14853 Claw Yding Andersen (163), Laboratory of Reproductive Biology, The Juliane Marie Centre for Children, Women, and Reproduction, University Hospital of Copenhagen, DK-2 100 Copenhagen, Denmark *Current address: Cardiovascular Research Center, Massachusetts General Hospital/Harvard Medical School, Charlestown, Massachusetts 02 129.

ix

This Page Intentionally Left Blank

Preface

This volume continues the customs of Current Topics in Developmental Biology in addressing developmental mechanisms in a variety of experimental systems and in having a theme-the molecular basis of pattern formation. In the first chapter, Liliana Solnica-Krezel from Vanderbilt University considers pattern formation in zebrafish. Then, Jennifer K. Ng, Koji Tamura, Dirk Buscher, and Juan Carlos Izpis6a-Belmonte from the Salk Institute discuss vertebrate limb development. In Chapter 5 , Vivian F. Irish from Yale University presents petal and stamen development, another molecular challenge in the broad field of pattern formation. Sex is never far from the minds of developmental biologists and Patricia E. Kuwabara from the MRC in Cambridge considers the developmental genetics of sex determination in Caenorhabditis elegans in Chapter 4. The molecules that trigger the resumption of meiosis in mammals are still being discovered, and Claus Yding Andersen, Mogens Baltsen, and Anne Grete Byskov from the University Hospital of Copenhagen review their novel results on meiosis-activating sterols during gonadotropin-induced resumption of oocyte meiosis in mammals. Weird sex is also considered in Chapter 3 by Deborah M. Neubaum and Manana F. Wolfner from Cornell University in their discussion of the mechanisms of sperm storage in female animals. Together with the other volumes in this series, this volume provides a comprehensive survey of major issues at the forefront of modem developmental biology. These chapters should be valuable to researchers in the fields of plant, invertebrate, and vertebrate development, as well as to students and other professionals who want an introduction to current topics in cellular, molecular, and genetic approaches to both developmental and reproductive biology. This volume in particular will be essential reading for anyone interested in gene regulation of pattern formation, sex determination, genetic controls of development, signaling molecules, cell cycle arrest checkpoints and resumptions, and gamete preservation and viability. 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 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. We thank the members of the Editorial Board for their suggestions of xi

topics and authors, and Michelle Emme for her exemplary administrative and editorial support. We are grateful for the unwavering support of Craig Panner and Michele Bidwell at Academic Press in San Diego and for the assistance of Kathy Nida. We are also grateful to the scientists who prepared chapters for this volume and to their funding agencies for supporting their research. Gerald P. Schatten Roger A. Pedersen

1 Pattern Formation in Zebrafish-Fruitful Liaisons between Embryology and Genetics L ilia nna Soln ica- Krezel Department of Molecular Biology Vanderbilt University Nashville, Tennessee 37232

I. Establishment of the Dorsoventral Polarity during Zebrafish Cleavage Stages 11. Establishment of the Dorsal Blastula Organizer (Nieuwkoop Center) 111. Induction of the Gastrula Organizer by the Blastula Organizing Center

I v. Structure and Function of the Dorsal Gastrula Organizer in Zebrafish A. Is the Embryonic Shield Equivalent to the Dorsal Gastrula Organizer? B. Molecular Genetics of the Inductive Functions of the Organizer V. Coordination of Gastrulation Movements A. Relationships between Dorsoventral Patterning and Gastrulation Movements VI. Conclusions References

Vertebrate embryos, despite quite diverse early morphologies, appear to employ similar cellular strategies and conserved biochemical pathways in their development (Eyal-Giladi, 1997). In the past decade, a small tropical teleost, zebrafish (Darzio rerio), became an important model system in which to study development (Streisinger er al., 1981). By combining embryology with molecular and classical genetic methods, our understanding of early inductive and morphogenetic events during vertebrate embryogenesis significantly advanced. In zebrafish, dorsal-ventral polarity is established during early cleavage and is dependent on microtubular transport of determinants from the vegetal pole to the blastomeres positioned on top of the yolk cell. The syncytium forming from these marginal blastomeres in the early blastula exhibits dorsal-ventral asymmetry with p-catenin localized to the nuclei on the presumptive dorsal side of the syncytium. The yolk cell is a source of signals that induce and pattern overlying blastoderm. Therefore, the dorsal yolk syncytial layer is equivalent to the Nieuwkoop center of the amphibian embryo. The embryonic shield, a thickening of the dorsal blastoderm margin, exhibits properties similar to the amphibian Spemann organizer. However, certain inductive and patterning signals from the organizer might be produced before the shield forms or might originate outside of the shield. Similar to the amphibian embryo, the key patterning functions of the fish dorsal organizer (Le., dorsalization of mesoderm, ectoderm, and coordination of gastrulation movements) are performed by secreted molecules that antagonize the ventralizing activity of the swirl (zbrnp-2) and zbmp-4 gene products expressed on the ventral side of the embryo. These functions of the dorsal organizer require the activity of the chordino gene (a zebrafish homologue of chordin), bozozok, mercedes and ogon loci. Copyright 0 1999 by Academic Press.

Currenr Topics in Developmenral Sio/ogy, Vol. 41 Copyright 0 1999 by Academic Press. All rights of reproduction in any form reserved. 0070.2153/99 $25.00

1

2

Lilianna Solnica-Krezel

1. Establishment of the Dorsoventral Polarity during Zebrafish Cleavage Stages The specification of embryonic polarity is initiated at early stages of zebrafish development, leading to the establishment of a dorsal blastula signaling center, an equivalent of the Nieuwkoop center, by the midblastula stage. In contrast to Xenopus, fertilization of the zebrafish egg is restricted spatially and occurs at a single sperm entry point located at the future animal pole of the embryo (Hart et al., 1992; Hart and Donovan, 1983). Fertilization is followed by a separation of cytoplasm from the yolk such that the zygote exhibits a cytoplasmic blastodisk at the animal pole of a large yolk sphere (Fig. 1A) [for an overview of stages of zebrafish development, see Kimmel et al. (1995)l. Subsequent rapid and synchronous cellular divisions occur in a stereotyped fashion. The planes of the first divisions, however, are not correlated with the future dorsoventral axis of the embryo (Abdelilah et al., 1994; Helde et al., 1994; Wacker et al., 1994). The initial cleavages are meroblastic, so that all early blastomeres maintain large cytoplasmic connections with a thin cytoplasmic layer of the yolk cell. Starting at the 16-cell stage, only the marginal blastomeres remain connected with the yolk cell. The synchrony of the early divisions is lost at the midblastula stage (9-10th division), which marks initiation of the zygotic transcription and cell motility (Kane and Kimmel, 1993). At or soon before the midblastula transition, the marginal blastomeres fuse completely with the yolk cell to form the yolk syncytial layer (YSL) (Kimmel and Law, 1985). Concurrently, the most superficial Fig. 1 Key steps in the establishment of dorsoventral polarity during zebrafish development. (A) In the zebrafish zygote, cytoplasmic streaming leads to separation of the cytoplasmic island located on top of a large yolk cell. Dorsal determinants are thought to reside at the vegetal pole of the teleost zygote. (B) Before the first cleavage, an array of parallel microtubules forms in the vegetal cytoplasmic layer of the yolk cell. The polarity of microtubules anticipates the future DV axis of the embryo. Furthermore, transport of cytoplasmic particles (dorsal determinants) by this microtubule array is necessary for axis formation. (C) During cleavage, a second array of microtubules extends from the marginal blastomeres into the yolk cytoplasmic layer. These microtubules are thought to mediate transport of dorsal determinants to the dorsal marginal blastomeres. (D) Around the midblastula transition, nuclear accumulation of p-catenin is first detected in the prospective dorsal yolk syncytial layer. This distribution of p-catenin, together with the inductive potential of the yolk cell, suggests that this region most likely corresponds to the Nieuwkoop center of zebrafish. Yellow arrows indicate putative dorsalizing signal(s) emanating from the dorsal YSL, and green arrows indicate putative mesoderm-inducing signal emanating from lateral and ventral YSL. (E) At the sphere stage, nuclear localization of p-catenin is maintained in the dorsal YSL, while it can also be detected in the dorsal blastomeres. The functional significance of this aspect of p-catenin distribution is not understood. (F) At the early stages of gastrulation, ingressing mesendoderm can be seen as a germ ring along the circumference of the embryo viewed from the animal pole. Embryonic shield is a dorsal thickening of the germ ring. Expression of zbmp-2/4 is shown schematically in the ventral and lateral regions of the gastrula; expression of the chordino locus is indicated in the dorsal region of the gastrula. Expression and functional requirements for other genes in the dorsal and ventral signaling centers are also indicated.

This Page Intentionally Left Blank

1. Pattern Formation in Zebrafish

3

blastomeres become an enveloping layer (EVL). The remainder of the blastomeres located between the EVL and YSL, so-called deep cells, will give rise to all embryonic structures (Kimmel et al., 1995). This characteristic three-layered blastula of zebrafish appears radially symmetric, without any obvious dorsoventral asymmetry [reviewed in Solnica-Krezel et al. (19931. Dorsal maternal determinants in teleost embryos are thought to be present in the vegetal mass of the yolk cell soon after fertilization (Fig. 1A) (Mizuno et al., 1997). Studies demonstrate that the specification of future dorsal structures is dependent on microtubular arrays present in the yolk cell, and most likely on microtubule-dependent transport of particles from the vegetal pole to the prospective dorsal marginal blastomeres (Fig. lB,C) (Jesuthasan and Strahle, 1997; Trimble and Fluck, 1995). The zebrafish zygote transiently exhibits a dense array of parallel microtubules at the vegetal pole, whereas microtubules at the equator do not exhibit any clear organization. During cleavage stages, microtubules extend from blastomeres into the yolk cytoplasmic layer (Jesuthasan and Strahle, 1997; Solnica-Krezel and Driever, 1994). Disruption of microtubule arrays in the zygote or during cleavage, prior to the 32-cell stage, interferes with animalward movement of particles from the vegetal hemisphere into marginal blastomeres. These treatments inhibit the subsequent axis formation, as judged by the lack of expression of dorsal specific markers at the blastula stage (nuclear p-catenin) and gastrula stage (gsc, axial). The resulting embryos either do not exhibit any dorsal axis or possess very reduced axes with medially fused somites lacking notochord and head structures [Jesuthasan and Strahle (1997) and references therein]. Furthermore, in medaka fish, the direction of movement of cytoplasmic parcels on cortical microtubules anticipates the future dorsal-ventral axis of the embryo (Trimble and Fluck, 1995). These studies form the basis for an attractive hypothesis that the asymmetric transport of dorsal determinants from the vegetal pole to marginal blastomeres is one of the initial events involved in the establishment of the dorsoventral axis in teleost embryos (Jesuthasan and Strahle, 1997; Trimble and Fluck, 1995). Notably, other work indicates that in amphibians a similar microtubuledependent transport is involved in specification of the dorsoventral axis (Rowning et al., 1997). In a frog embryo, the dorsal-ventral axis is specified during the first cell cycle, when the cortex rotates relative to the cytoplasmic core along parallel microtubules associated with the vegetal core (Elinson and Rowning, 1988). Disruption of microtubules blocks the rotation and results in embryos lacking dorsal structures. On the basis of these observations, it has been thought that the rotation of the cortex relative to the core is effecting the translocation of dorsalizing components and dorsal axis induction (Elinson and Rowning, 1988). However, whereas the dorsalizing components are transported approximately 90" from the vegetal pole, the cortex rotates only 30" during cortical rotation. This apparent contradiction was clarified by the observation that endogenous organelles as well as fluorescent carboxylated beads injected into the vegetal pole are

4

Lilianna Soinica-Krezel

transported along the subcortical microtubules at least 60" toward the future dorsal side (Rowning et al., 1997). Hence, rather than transporting the dorsal determinants directly, the cortical rotation might serve to align the subcortical microtubules. The latter in turn would mediate the transport of dorsal determinants to the prospective dorsal equatorial region, a process very reminiscent of that described for teleost embryos. Therefore, despite the distinctly different organization of the frog and fish zygotes, the process of initial polarization in the two embryos might be as similar as the later patterning events seem to be, as discussed in this chapter.

II. Establishment of the Dorsal Blastula Organizer (Nieuwkoop Center) The identities of substances transported on microtubule tracks are mostly unknown. A notable exception is p-catenin, a component of the Wnt signaling pathway, which was shown to colocalize with subcortical microtubules on the dorsal side of the frog egg at the end of cortical rotation (Rowning et al., 1997). Furthermore, an important conserved step in the pathway of dorsal axis formation appears to be a nuclear accumulation of p-catenin on the dorsal sides of both Xenopus and zebrafish blastulae (Funayama et al., 1995; Schneider et al., 1996). This region is thought to constitute the dorsal blastula organizer (Nieuwkoop center), which is responsible for the subsequent induction of the dorsal gastrula (Spemann) organizer in the overlying cells. Peter Nieuwkoop first demonstrated the importance of vegetal cells (presumptive endoderm) in induction and patterning of the mesoderm in a series of recombination experiments involving fragments of amphibian blastulae (Nieuwkoop, 1969). In those experiments, mesodermal tissues, notochord, muscles, kidney, and blood, formed only when vegetal (endodermal) cells were combined with animal cap (future ectoderm) fragments, but not when vegetal or animal cap fragments were cultured alone. Moreover, the dorsoventral polarity of the induced mesoderm was determined by the particular subregion of endoderm used. Specifically, dorsal mesoderm was obtained only when dorsovegetal fragments were the inducer (Nieuwkoop, 1969). The dorsal mesoderm-inducing activity localized to the dorsalmost vegetal cells of the Xenopus blastula was named the Nieuwkoop center (Gerhart et al., 1991). p-Catenin is a multifunctional protein involved in cadherin-dependent cell adhesion as well as in the transduction of receptor-mediated intercellular signals [reviewed in Miller and Moon (1996)l. p-Catenin is a homologue of the Drosophila segment polarity gene armadillo (arm) (Peifer and Wieschaus, 1990). The activity of arm as a signaling molecule is dependent on its nuclear localization, which is regulated by the wingless (wg)signaling pathway (Riggleman et al., 1990). In this pathway, the secreted Wg glycoprotein (Wnt in vertebrates)

1. Pattern Formation in Zebrafish

5

binds to its cell surface receptor thought to be encoded by the D - f r i ~ f e d -(fi) 2 gene. The ligand-receptor complex stimulates, in an unknown fashion, the activity of a phosphoprotein Dishevelled (Dsh). Activated Dsh in turn inhibits the activity of zeste white shaggy-3 kinase (the homologue of glycogen synthase kinase-3 in vertebrates), which normally destabilizes and thus inhibits arm (p-catenin) activity. In this manner, Wingless signaling leads to nuclear accumulation of p-catenin. In the nucleus, p-catenin interacts with an architectural transcription factor pangolin (LEF-1 family of proteins in vertebrates). In the vertebrate blastula, this complex is thought to activate the transcription of genes involved in axial specification. The likely central role of the Wnt-like signaling pathway in dorsal axis formation in frog embryos is underscored by the ability of several components of the pathway [i.e., wnt genes (Sokol et al., 1991), dsh (Sokol et al., 1995), dominant negative gsk-3 (Pierce and Kimelman, 1995; He et al., 1995), p-catenin (Funayama et al., 1995; Guger and Gumbiner, 19951 to induce the formation of secondary axes when expressed ectopically in the embryo. Furthermore, treatment of blastula stage embryos with lithium, which is thought to inhibit GSK-3 kinase activity (Hedgepeth et al., 1997; Klein and Melton, 1996), also leads to the induction of dorsal mesoderm in the entire marginal zone (Backstrom, 1954; Kao and Elinson, 1986, 1988). However, only p-catenin (Heasman et al., 1994) and GSK-3 (Pierce and Kimelman, 1995; He et af., 1995) have been demonstrated to be required for the formation of the endogenous axis in frog embryos. Notably, dominant negative mutant forms of Wnt (Hoppler et al., 1996) and Dsh (Sokol, 1996), while able to inhibit the formation of ectopic axes by their wildtype counterparts, do not prevent formation of the endogenous axis. Therefore, either a redundant pathway or a distinct pathway(s) leading to dorsoventral differences in GSK-3 activity and activation of p-catenin during normal development might exist (Yost et af., 1996). An involvement of the Wnt signaling pathway components, other than a p-catenin and GSK-3, in the zebrafish embryogenesis remains to be investigated (Kelly et al., 1995; Stachel et al., 1993). Important conjugation experiments in frog embryos indicate that cells in a dorsal vegetal mass, expressing p-catenin, send a “dorsal signal” to other cells at the midblastula stage, after the onset of zygotic transcription (Wylie et al., 1996). These observations are consistent with p-catenin regulating the inducing activity of the Nieuwkoop center. A systematic analysis of changes in intracellular distribution of P-catenin during Xenopus development demonstrated that p-catenin displays greater cytoplasmic accumulation on the future dorsal side of Xenopus embryo by the two-cell stage (Larabell et al., 1997). This is likely to be effected by the dorsally directed microtubular transport discussed previously (Rowning et al., 1997). The accumulation of p-catenin in dorsal nuclei was reported as early as the 16- to 32-cell stages (Larabell et af., 1997). In another report, nuclear accumulation of p-catenin was first observed shortly after stage 8 and was most prominent at stage 8.5, disappearing before the blastopore lip

6

Lilianna Solnica-Krezel

formation at the onset of gastrulation (Schneider et al., 1996). The domain of nuclei positive for p-catenin is located in the dorsal region of the blastula, occupying approximately one-third of the equator. Interestingly, this domain extends from the marginal zone into the vegetal and animal regions and is several cell layers deep (Schneider et al., 1996). If the distribution of p-cateninpositive nuclei reflects the extent of the Nieuwkoop center, it would mean that this organizing center is not located exclusively in vegetal blastomeres but also in marginal and some animal blastomeres. Alternatively, nuclear accumulation of p-catenin might extend beyond the Nieuwkoop center. In the latter case, activities of molecules other than p-catenin need to be implicated in the specification of the Nieuwkoop center. Where is the Nieuwkoop center equivalent located in a teleost embryo? This question was addressed by transplantation experiments in Salmo (Long, 1983) and in zebrafish (Mizuno et al., 1996), as well as by blastoderm isolation experiments in Fundulus (Oppenheimer, 1936b). Results of these experiments revealed the key role of the YSL in the process of mesoderm induction and patterning in teleost embryos. In the blastoderm isolation experiments, relatively normal development occurred in blastoderms deprived of their yolk cell at the 32-cell and later cleavage stages. In contrast, blastoderms isolated earlier did not gastrulate. Oppenheimer hypothesized that the blastoderms removed from the yolk before the 32-cell stage lacked some of the substance that is passed from periblast to blastoderm during or later than the 16-cell stage (Oppenheimer, 1936b). It is intriguing that treatments that inhibit the microtubular transport of particles from the vegetal hemisphere to marginal blastomeres, and thus consequently impair formation of the dorsal axis, also are effective only at or before the 32-cell stage [see Jesuthasan and Strahle (1997)l. These studies support the notion that the microtubules translocate the “axis-inducing” activity from the vegetal pole to the marginal blastomeres by the 32-cell stage. However, yolk cells from midblastula stage embryos, when transplanted to the animal pole of another zebrafish embryo, still exhibited the full potential to induce and pattern the host mesoderm (Mizuno et al., 1996). Thus, the axisinducing activity is present in the yolk cell at the midblastula stage. In zebrafish the YSL forms when the marginal blastomeres collapse onto the yolk cell at about the ninth or tenth cleavage (Kimmel and Law, 1985). Hence, the “axisinducing” material translocated from the vegetal pole to the marginal blastomeres during early cleavage most likely becomes incorporated into the YSL during a fusion of the marginal blastomeres with the yolk cell. Consistent with this view, P-catenin-positive nuclei are detected within the YSL on the prospective dorsal side of the high blastula stage zebrafish embryos about 20-30 min after the formation of the syncytial layer (Fig. 1D) (Schneider et al., 1996). Subsequently, at the sphere stage nuclear localization of p-catenin is observed in the blastoderm cells overlying dorsal YSL (Fig. 1E). The localization of p-catenin to the dorsal YSL, together with the ability of the yolk cell to induce dorsal mesoderm upon

1. Pattern Formation in Zebrafish

7

transplantation, strongly suggests that the dorsal YSL in the zebrafish blastula corresponds to the Nieuwkoop center located in the vegetal dorsal blastomeres of the frog blastula (Schneider et al., 1996). It is noteworthy that the accumulation of p-catenin in the YSL shortly precedes the onset of zygotic transcription at the midblastula transition. Therefore, the initial specification of the dorsoventral polarity and establishment of the inductive center in the dorsal YSL are dependent on the maternal program.

111. Induction of the Gastrula Organizer by the Blastula Organizing Center A key question is what are the downstream targets of the P-catenin-LEF-1 complex in the nuclei of the Nieuwkoop center? Moreover, what is the molecular nature of the “dorsal signal” that establishes the dorsal gastrula organizer in the overlying blastomeres (Wylie et al., 1996)? One potential candidate that might function in both frog and fish embryos is Vgl. This secreted, TGF-P-related molecule was described originally in frogs as maternal mRNA localized to the vegetal blastomeres (Rebagliati et al., 1985; Weeks and Melton, 1985). Ectopic expression of mRNA encoding Vgl fails to induce axis formation. However, injections of mRNA encoding a chimeric BMP-2-Vgl fusion protein that allows the production of the mature form of the Vgl ligand in vivo can both restore axis formation in UV-ventralized embryos and induce ectopic axes. In addition, the blastomeres expressing BVgl appear to act as a Nieuwkoop center (Thomsen and Melton, 1993). Interestingly, the zebrafish vgl orthologue (ZDVR-1) mRNA is found to be distributed uniformly in the egg and later in all embryonic cells (Helde and Grunwald, 1993). ZDVR-1 mRNA fails to induce axis formation when overexpressed in fish embryo. However, the normal zDVR-1 precursor does appear to be processed to mature protein when expressed in Xenopus and can act as a potent inducer of axial mesoderm in this system. These observations form the basis for a proposal that localized posttranslational processing of Vgl precursor protein on the prospective dorsal side plays a regulatory role in the development of the dorsal axis in frog and fish embryos (Dohrmann et al., 1996). How is the proposed activation of Vgl related to the p-catenin pathway? Because mature Vgl ligand has not been detected during normal development, the timing of its presumably proteolytic activation in vivo is not clear. However, BVgl mRNA injections can rescue the effects of depleting maternally encoded p-catenin in frog embryos, suggesting that Vgl acts downstream of p-catenin and thus around the midblastula transition (Wylie et al., 1996). These experiments cannot, however, exclude the possibility that Vgl acts in a pathway parallel to p-catenin. Indeed, additional evidence from Xenopus indicates that Vgl acts synergisticallywith a maternal Wnt-like signal to specify dorsal fates in both mesoderm and endoderm (Cui ef al., 1996).

8

Lilianna Solnica-Krezel

An important finding has been that a downstream target of p-catenin in Xenopus embryos appears to be the siamois gene, which encodes a paired-class homeodomain protein (Lemaire et al., 1995). Injections of RNA encoding Siamois into frog embryos lead to the induction of complete secondary axes, but the progeny of injected cells to not participate in the secondary axes. siamois is expressed after the midblastula transition and is present most abundantly in the dorsal endoderm of early gastrulae, indicating that it might be an important component of the Nieuwkoop center (Lemaire et al., 1995). Interestingly, whereas p-catenin and other components of the Wnt pathway highly activate expression of members of the siamois, TGF-fl pathway (such as Vgl) were shown to activate siamois only weakly (Brannon and Kimelman, 1996). This observation is consistent with Vgl acting in a parallel rather than the same pathway as p-catenin (see previous discussion). The involvement of TGF-p signaling in the formation of the zebrafish gastrula organizer is further supported by the activity of nodal and TARAM-A genes. The murine nodal gene is a TGF-P-related ligand that is essential for the formation of the node, the mouse equivalent of the Spemann organizer (Zhou et al., 1993). Injections of mouse nodal mRNA into zebrafish embryos lead to ectopic expression of the organizer-specific genes gsc and Ziml and formation of ectopic axes containing notochord and somites (Toyama et al., 1995). TARAM-A is a serinethreonine kinase type I receptor related to TGF-f3 and activin receptors (Renucci et al., 1996). Whereas ectopic expression of RNA encoding wild-type TARAMA leads only to an expansion of the axial mesoderm, injections of RNA encoding a constitutively active receptor kinase, TARAM-A-D, result in a massive induction of the dorsal and panmesodermal markers, gsc and ntl, shortly after the midblastula stage as well as numerous dorsal mesoderm genes in the late blastula stage (50% epiboly). Furthermore, injections of TARAM-A-D RNA induce the formation of complete secondary axes in which cells that inherited injected RNA mostly contribute to the anterior dorsal mesoderm (hatching gland and head mesoderm) (Peyrieras et al., 1996; Renucci et al., 1996). Notably, TARAM-A mRNA is found to be distributed uniformly in the zygote and during cleavage stages and disappears at the 500-cell stage from the marginal blastomeres that are fated to form the YSL. In the mid- and late blastula, TARAM-A mRNA is detected at the blastoderm margin around the circumference of the embryo, with its expression increasing on the dorsal side starting at 40% epiboly (Renucci et al., 1996). Both the activity and the expression pattern of TARAM-A make it a good candidate for a transducer of signals from the Nieuwkoop center involved in formation of the Spemann organizer. Another link between the Nieuwkoop center, p-catenin, TGF-P pathway, and the Spemann organizer in zebrafish has been revealed by the zebrafish zygotic bozozok mutation. Homozygous bozm’68mutants are characterized by the lack of the main derivatives of the gastrula organizer, prechordal plate and notochord, and also exhibit defects in the neuroectoderm (Solnica-Krezel et al., 1996).

1. Pattern Formation in Zebrafish

9

Notably, an embryonic shield cannot be detected morphologically in boz mutants at the early gastrula stage, and expression of all tested organizer-specific genes is reduced or absent, indicating that boz function is required for the proper formation of the dorsal gastrula organizer (L. Solnica-Krezel and K. Fekany, unpublished observations). The boz phenotype is reminiscent of defects observed in embryos in which microtubules were disrupted during cleavage stages, a treatment that prevented nuclear accumulation of p-catenin and inhibited dorsal axis formation ( Jesuthasan and Strahle, 1997). Several additional observations indicate that boz acts during the blastula stage within or downstream of the Nieuwkoop center in the pathway specifying axial mesoderm. First, expression of organizer-specific genes (gsc, JEh, TARAM-A) is already reduced at the late blastula stage. Second, whereas ectopic expression of p-catenin leads to the induction of full secondary axes in wild-type embryos (Kelly et al., 1993, the axes induced in boz embryos lack notochord and other midline cell types. Thus, p-catenin is unable to suppress the boz phenotype. Finally, the lack of notochord and prechordal plate, the hallmarks of the boz phenotype, is fully suppressed by ectopically expressed mRNAs encoding mouse nodal and the activated form of TARAM-A-D (Renucci et al., 1996), but not by wild-type TARAM-A (L. Solnica-Krezel, A. Renucci, and K. Fekany, unpublished observations). These studies indicate that the boz locus is required in a TGF-p signaling pathway that establishes the dorsal gastrula organizer in zebrafish acting downstream or in parallel with dorsalizing signals like p-catenin.

IV. Structure and Function of the Dorsal Gastrula Organizer in Zebrafish Transplantationexperiments performed in the 1920s by Hilde Mangold and Hans Spemann identified the dorsal blastopore lip of the amphibian gastrula as a tissue that, upon transplantation to the ventral side of the gastrula, will induce formation of the secondary body axis. The organizer tissue, while contributing mostly to chordamesoderm and to a lesser extent to paraxial mesoderm and the floor plate of the neural tube of the secondary axes, was able to induce and pattern the ectopic neuraxis as well as coordinate gastrulation movements (Spemann, 1938).

A. Is the Embryonic Shield Equivalent to the Dorsal Gastrula Organizer?

All vertebrates exhibit an equivalent dorsal gastrula organizer region, as defined by its inductive properties and expression of an array of organizer-specific genes, including gsc, HNF-3p, Xlim-1, Xnot, noggin, chordin, siamois, ADMP, and others [reviewed in Lemaire and Kodjabachian (1996)l. In zebrafish, the embryonic shield, a dorsal thickening along the blastoderm margin of the early gastrula,

10

Lilianna Solnica-Krezel

is thought to correspond to the Spemann organizer (Fig. 1F). During normal development, this region gives rise to axial mesoderm, notochord and prechordal plate, and to some neural tissue, predominantly floor plate (Kimmel et al., 1990; Shih and Fraser, 1995). Transplantations of the shield to the ventral blastoderm margin lead to the formation of ectopic axes, with the donor tissue forming notochord, prechordal plate, and floor plate of the neural tube (Ho, 1992; Oppenheimer, 1936a,c; Shih and Fraser, 1996; Driever et al., 1997). Interestingly, the ectopic axes resulting from the shield transplantations are not complete and usually lack the anterior head structures (Shih and Fraser, 1996). In addition, the reciprocal experiments-extirpations of the shield-yield embryos without notochord, but with a relatively normal anterior-posterior (AP)pattern of the neural tube (Shih and Fraser, 1996). One explanation is that in these experiments the extirpated and/or transplanted tissue did not contain the entire organizer. This could be due to experimental difficulties with extracting the entire shield, leading to residual organizer activity remaining in place. Indeed, the expression of organizer-specific genes was not monitored in the operated embryos (Shih and Fraser, 1996). It is noteworthy that a different technique, glass pipet versus eyelash knife, can be used to perform the shield extirpationtransplantation experiments, allowing one to remove most of the region expressing organizer-specific genes, such as gsc andJlh (D. Stemple, National Institute for Medical Research, London, UK, personal communication; Driever et al., 1997). The secondary axes formed upon transplantations of shields removed with this technique can be complete. The shield extirpations performed with a glass capillary result in embryos with the axial mesoderm reduced or missing completely and severe deficiencies in the anterior neural structures, including cyclopia and/or eye reduction (Fig. 2B). However, neural keel still forms in these experimental embryos, indicating that a significant level of organizer activity might reside beyond the shield region in the zebrafish embryo. Alternatively, some inductive events happen before the shield stage. Consistent with this notion, the expression of a number of genes like gsc, liml, andJlh can be detected on the dorsal side of the embryo before the embryonic shield manifests itself (Stachel et al., 1993; Toyama et al., 1995; Talbot et al., 1995). Support for the notion that the organizer in fact extends beyond the shield comes from the observation that one of the key organizer-specific genes, chordino (zebrafish chordin homologue), is expressed in a broad dorsal marginal domain extending outside the shield (Miller-Bertoglio et al., 1997; Schulte-Merker et al., 1997).

B. Molecular Genetics of the inductive Functions of the Organizer

The dorsal gastrula organizer is thought to carry out a number of inductive and patterning functions: induction and patterning of the neuroectoderm, dorsalization of paraxial mesoderm, coordination of gastrulation movements, and finally

Fig. 2 Phenotypes of mutations affecting dorsoventral patterning and their phenocopies. (A) Wildtype embryo at one day of development. (B) Embryo resulting from shield extirpation. (C) bozozok. (D-G) Phenocopies resulting from injections of increasing amounts of RNA encoding BMP-4. (H) dino mutant phenotype. (I) Phenocopy obtained after injection of CSKA plasmid expressing zmbp-4. (J) snailhouse mutant phenotype. Phenocopy of snh mutants is obtained by injections with RNA encoding Noggin (K) or dominant-negative BMP-4 receptor (L). [(B) Obtained from D. Stemple, London, UK. (D, E, F, G, K, L) Reprinted from Mechanisms of Development 62; Neave, B., Holder, N., and Patient, R. A graded response to BMP-4 spatially coordinates patterning of the mesoderm and ectoderm in zebrafish. pp. 183-195 (1997), with kind permission from Elsevier Science Ireland Ltd., Bay 15K,Shannon Industrial Estate, Co. Clare, Ireland. (H) Reprinted from Developmenr, Hammerschmidt et al. (1996b). with permission from the Company of Biologists Limited. (I) From Development, Hammerschmidt et al. (1996a), with permission from the Company of Biologists Limited. (J) Reprinted from Developmenr, Mullins et a/. (1996), with permission from the Company of Biologists Limited.]

12

Lilianna Solnica-Krezel

self-differentiation into midline structures (Spemann, 1938). Discoveries in a number of animal systems, most notably expression cloning in Xenopus, identified several molecules that might fulfill these distinct functions of the gastrula organizer [reviewed in Lemaire and Kodjabachian (1996)l. Identification of numerous mutations affecting organizer function in zebrafish provides genetic confirmation of the presumed functions for some of these molecules, as well as identification of novel genetic components. Furthermore, the analysis of the zebrafish mutants reveals genetic hierarchies that govern pattern formation in the gastrula (Hammerschmidt et al., 1996b; Mullins et al., 1996; Solnica-Krezel et al., 1996).

1. The Organizer Dorsalizes Mesoderm and Ectoderm by Antagonizing the Ventral Morphogen BMP2/4 Numerous molecular biological and genetic studies support the notion that the two main activities of the dorsal gastrula organizer, dorsalization of mesoderm and neural induction (which could be viewed as dorsalization of ectoderm), are accomplished by secreted factors that antagonize ventralizing signals derived from the ventral organizing center in the gastrula [Piccolo et al., 1996; Zimmerman et al., 1996; reviewed in Hemmati-Brivanlou and Melton (1997)l. This interaction between bone morphogenetic proteins 2 and 4 (BMP-2/4), secreted ligands of the transforming growth factor+ superfamily, and their antagonists appears to be a mechanism for embryonic dorsoventral patterning that is conserved from the fruit fly to vertebrates (Hogan, 1996; Holley and Fergusson, 1997; Holley et al., 1995). The realization that the formation of ventral fates in the embryo is not a default pathway, but rather depends on specific signaling events, was suggested initially by experiments showing that ectopic expression of BMP-4 induced ventral and posterior cell types in animal cap explants and whole Xenopus embryos (Dale et al., 1992; Jones et al., 1992, 1996). Transcripts encoding BMP-4 are present in Xenopus unfertilized eggs and initially are distributed uniformly throughout frog blastula. In contrast to Xenopus (Hemmati-Brivanlou and Thomsen, 1995), transcripts for the zebrafish homologues (zbmp-2 and zbmp-4) are not found in the maternal pool of RNAs. zbmp-2 transcripts are detected first at the sphere stage and are distributed uniformly throughout the blastoderm (Nikaido et al., 1997). However, as the yolk domes at the onset of epiboly, zbmp-2 transcripts disappear from the presumptive dorsal side. At 50% epiboly, just before the onset of the involution/ingression movements that will create the germ layers, zbmp-4 transcripts become detectable. Similar to the Xenopus embryo, at the early (shield) stages of gastrulation, both zbmp-2 and -4 exhibit high levels of transcripts in the presumptive ventrolateral blastoderm in a domain extending from the margin up through the animal pole and are absent on the dorsal side. An interesting and not yet fully understood exception is a small

1. Pattern Formation in Zebrafish

13

distinct domain of zbmp-4 expression in the shield region itself, marking the precursors of the anterior axial mesoderm (Fig. IF) (Chin et al., 1997; Hammerschmidt et al., 1996b; Nikaido et al., 1997). Overexpression of zbmp-2 and zbmp-4 in zebrafish embryos supports the conclusions drawn from studies of frog embryos (Hemmati-Brivanlou and Thomsen, 1995; Jones et al., 1991; Wilson and Hemmati-Brivanlou, 1995) that these factors respecify dorsal mesoderm to ventral mesoderm and expand the prospective epidermis at the expense of neuroectoderm (Hammerschmidt et al., 1996b; Nikaido et al., 1997; Neave et al., 1997). However, the exact phenotypes reported differ in an interesting way. Neave and colleagues performed the most systematic analysis by injecting varying amounts of Xenopus bmp-4 RNA (20-220pg/ embryo) into 1-4-cell-stage zebrafish embryos (Neave et al., 1997). Progressively ventralized embryos were observed in a dose-dependent manner (Fig. 2D-G). The lowest bmp-4 dose led to “ventralized I” class embryos, which lacked head and notochord structures, whereas the somites lost their chevron shape and became fused in the midline. In the “ventralized 11” class of embryos, besides the lack of a head and notochord, the yolk extension failed to form. More ventralized embryos (classes I11 and IV) were observed only after the injection of 100 pg or more RNA. Embryos of class I11 exhibited radial symmetry; cells accumulated at two opposite poles of the round yolk cell with somitic tissue forming at one of the ends. Finally, embryos of class IV likewise were radially symmetric; however, cells accumulated only at one end, suggesting that epiboly also was affected. A similar loss of axial structures and often severe reduction of the hypoblast was observed in embryos injected with 100 pg of zbmp-2 RNA (Nikaido et al., 1997). Analysis of axial mesoderm markers (gsc and ntl) revealed that expression of these genes was normal in bmp-4 RNA-injected embryos at the onset of gastrulation and that reduction or absence of gsc and ntl expression began to be observed by 65-70% epiboly (Neave et al., 1997). These results indicate that even high doses of BMP-2/4 signaling do not prevent the initial formation of the main component of the dorsal gastrula organizer, axial mesoderm. However, exclusion of BMP-2/4 signaling from the dorsal side of the gastrula might be important for maintenance and further development of axial mesoderm, as has been reported earlier for Xenopus (Jones et al., 1996). Interestingly, a distinct phenotype was caused by ectopically expressing BMP-4 at later stages of development by using a plasmid with bmp-4 cDNA under the control of the cytoskeletal actin (CSKA) promoter from Xenopus borealis, which directs BMP-4 expression only after the midblastula transition. Whereas there was expansion of ventral markers and reduction of neuroectodermal markers in injected embryos during gastrulation, at the end of embryogenesis only the posterior part of the notochord was reduced and embryos exhibited multiple ventral fin folds and an enlarged blood island (Fig. 21) (Hammerschmidt et al., 1996~).It is possible that the phenotype obtained with DNA injections represents the mildest ventralization and thus could reflect the lowest dose of BMP-4

14

Lilianna Solnica-Krezel

signaling obtained with injections of 20 pg of RNA, as reported by Hammerschmidt et al. (1996b). Overexpression of antagonists of BMP-2/4 signaling in zebrafish embryos leads to a reduction of ventral fates and to an expansion of dorsal fates including neural tissue (Hammerschmidt et al., 1996c; Neave et al., 1997), as reported previously for frog embryos (Graff et al., 1994; Suzuki et al., 1994; Zimmerman et al., 1996). Injections of synthetic mRNA encoding a truncated, dominant negative form of the Xenopus BMP receptor (tBR) led to ectopic domains of expression of the prechordal plate marker, the gsc gene, whereas overexpression of Xenopus noggin mRNA expanded the dorsal expression domain of gsc (Neave et al., 1997). Overexpression of both types of BMP antagonists resulted in enlargement of the notochordal rudiment as judged by the midline expression domain of ntl or axial genes (Hammerschmidt et al., 1996c; Neave et al., 1997). The expansion of axial mesoderm was associated with a reduction of intermediate mesoderm, as monitored by expression of the pax2 gene in the kidney primordia (Neave et al., 1997), as well as with the reduction of ventral mesodermal markers such as gatal (Hammerschmidt et al., 1996~). Within the ectoderm, ectopic expression of both tBR and noggin mRNAs resulted in the expansion of neuroectoderm precursors expressingfkd3 at the expense of nonneural ectoderm, as marked by reduced expression of gta3, d1x3, and eve1 (Hammerschmidt et al., 1996c; Neave et al., 1997). Importantly, the ventral but not dorsoanterior expression domain of BMP-4 also was reduced or absent in these dorsalized embryos. This is consistent with positive feedback loop regulation of bmp-4 expression in the ventral region of the zebrafish embryo, which has been postulated previously in Xenopus (Jones et al., 1992). A consistent picture of dorsoventral patterning in the vertebrate gastrula emerges from the preceding observations (Neave et al., 1997; Wilson et al., 1997). A concentration gradient of a BMP-214 morphogen (as either homo- or heterodimer) provides an instructive signal determining the range of dorsoventral fates in the mesoderm and ectoderm. This ventrodorsal gradient of BMP-2/4 signaling arises as a combination of the expression of BMP-2/4 genes, which is maintained by a positive feedback loop, and the expression of BMP antagonists that bind to the BMP ligands and block signaling through their receptors. 2. Genetic Evidence for Dorsal-Ventral Patterning via a BMP-2/4 and Its Antagonists The preceding model is now supported by analysis of mutations from large-scale genetic screens for mutations affecting zebrafish embryonic development (Driever et al., 1996; Hafter et al., 1996). Mutations defining at least 15 genes affect patterning during gastrulation (Mullins et al., 1996; Solnica-Krezel et al., 1996; Fisher et al., 1997; Halpern et al., 1993; Hammerschmidt et al., 1996b; Talbot et al., 1995). An initial characterization of mutant phenotypes led to the classification of these mutations into two groups, based on their effects on the formation of

15 cell fates in the embryo (Table I). One class of mutants exhibits a loss of cell fates derived from the dorsal part of the gastrula, including neuroectoderm, whereas the second class of mutants has deficiencies in ventral and posterior structures, in some cases accompanied by an expansion of neuroectoderm (Hammerschmidt et al., 1996b; Mullins et al., 1996; Solnica-Krezel et al., 1996). Within the first class, mutations in six loci, including cyclops (cyc) (Hatta et al., 1991),Jloating head (Talbot et al., 1995), bozozok (boz), one-eyed pinhead (oep), and schmalspur (sur; also previously named uncle freddy or unf), result predominantly in deficiencies of mesodermal and neuroectodermal cell fates derived from the dorsal region of the gastrula (axial mesoderm and anterior-ventral neuroectoderm), without an obvious increase in ventrally derived fates (Fig. 2) (SolnicaKrezel et al., 1996; Brand et al., 1996; Schier et al., 1997; Strahle et al., 1997). 1. Pattern Formation in Zebrafish

Table I Mutations Affecting Dorsal and Ventral Organizing Centers Phenotypic class

Locus

Dorsal bozozok organizer (bflz) class Ia; dorsoanterior fates decreased one-eyed pinhead (oepl

cyclops

joating head

UW no tail

Molecule encoded

Alleles m168; i2

References Solnica-Krezel et al., 1996; Blagden et al., 1997

Phenotype Chordamesoderm and prechordal mesoderm missing or reduced; reduced anterior and ventral neural fates

m134. Hammerschmidt et Derivatives of prechoral., 1996a; dal plate missing; tz257, ICRFI, X50 Schier et al., cyclopia and deficiencies in ventral 1997; Strahle et al., 1997 aspects of CNS b16, m101, Brand et al., 1996; Reduced prechordal plate; cyclopia and m122, Hatta et al., 1991; Solnicadeficiencies in venm294, Krezel et al., tral aspects of CNS tj2 19: te262c 1996 m768; Brand et al., 1996; Reduced prechordal ty68b Solnica-Krezel plate; cyclopia and deficiencies in venet al., 1996 tral aspects of CNS n l ; b327; Masai et al., 1997; Lack of notochord and Transcription neurogenesis in epiStemple et al., rm229, factor physis 1996; Talbot et tk249, al., 1995 m614 Abnormal notochord b160: b195; Halpern et al., T-box differentiation in the 1993; Odenthal m149, transcription trunk and lack of nom550, et ul., 1996; factor tochord in the tail; Stemple et al., tb244e, tail reduction 1996 tc41, ts260 (continues)

16

Lilianna Solnica-Krezel

Table I (Continued) Phenotypic class Dorsal organizer class Ib; axial mesoderm present ventral posterior fates increased

Locus

chodino (din)

Molecule encoded Chordin

Alleles

tm84; Fisher et al., tt2.50; m52; 1997; m 70; Hammerschmidt m282; et ul., 1996b; m346; Solnica-Krezel m.586; c4, et a[., 1996 b30.5; b386;

m60

Ventral organizer class 11; decreased ventral and posterior fates in some increased neuroectod e m and somites

mercedes (mes) swirl (swr) zbmp-2

rz209; tm309 tc300, ta72

somitabun (sbn)

dct24

snailhouse (snh) piggytail (pgy)

mini jin In4

References

dty40, dti216. re22 7u, tni124a, 10206, tx223 tmllOb, mlOO tv9b,

tc263a, t1203a, tyl30u, tb24/c, lj21 lC1, tfz 1%.

tn2 17b

Phenotype Variable degrees of multiplication of caudal fin, increased ventral and posterior fates (blood, pronephros); decreased neuroectoderm and somitic mesoderm, shorter body axis

Solnica-Krezel et al., 1996 Hammerschmidt el al., 1996b Mullins et al., Variable degrees of re1996 duction of ventral and posterior fates, lack of caudal fin, progressive deletions of tail and trunk; reduction of ventral Mullins et al., and mesodermal 1996 fates, blood, and pronephros; expansion of somites and neuroectodem Mullins et al., Enlongated gastrula 1996 shape Mullins et al., 1996

Mullins el al.. 1996; SolnicaKrezel et al., 1996 Mullins et al., I996

1. Pattern Formation in Zebrafish 17 Mutations in three loci, chordino (din) (Hammerschmidt et al., 1996c; SchulteMerker et al., 1997), mercedes (mes) (Hammerschmidt et al., 1996b), and ogon (ogo) (Solnica-Krezel et al., 1996) lead to an increase in ventral fates and concomitant decrease in dorsolateral fates, mostly somites and neuroectoderm. The phenotypes of mutants in each of the preceding classes can be strikingly phenocopied either by ectopic expression of BMP-2/4 from RNA or DNA constructs or by injections of RNAs encoding BMP-2/4 antagonists, as described earlier (Fig. 2). Specifically, the phenotypes resulting from mutations at din, mes, and ogo loci are characterized by multiple ventral fin folds (Fig. 3), increased numbers of blood cells, and increased expression levels of ventral markers like eve1 (Fisher et al., 1997; HammerSchmidt et al., 1996b,c; Solnica-Krezel et al., 1996). Expansion of ventral fates is accompanied by a reduction of neuroectoderm, somitic, and chordamesoderm in din mutants. All of the preceding defects also were observed in embryos after injections with bmp-4 cDNA on a CSKA plasmid. In addition, the phenotype of the din mutation can be suppressed by ectopic expression of tBR and noggin, further supporting the notion that the defect in this mutant is in the inhibition of BMP-2/4 signaling (Hammerschmidt et al., 1996~). On the other hand, the phenotypes of the "dorsalized" class of mutants at the swirl (swr), somitabun (sbn), snailhouse (snh), lost-a-jin (laf ), piggytail (pgy), and mini j n (mfn) (Mullins et al., 1996; Solnica-Krezel et al., 1996) closely

Fig. 3 Effects of dorsalizing-ventralizing mutations and treatments on caudal fin development. (A, B) Lateral view of the caudal fin at day I of development in wild-type (A) and luf mutant (B) embryos. (C-G) Posterior view of the caudal fin in wild-type (C), mes (D), din (E), and ogo (C) mutant embryos and in embryos injected with 20 pg of bmp-4 RNA (F). [(A, B) Reprinted from Development, Solnica-Krezel et al. (1996). (C-F) Reprinted from Development, Hammerschmidt ef al. (1996b) with permission from the Company of Biologists Limited.]

18

Lilianna Solnica-Krezel

resemble defects obtained by ectopic expression of the BMP-2/4 antagonists (Neave et al., 1997) (Hammerschmidt et al., 1996~). Moreover, analysis of swrdin double mutants demonstrated that swr is epistatic to din (Hammerschmidt et al., 1996~).These observations gave support to the hypothesis that din encodes an inhibitor of swr function, with din and swr affecting the components of the morphogenetic system itself, chordin and BMP-2/4, respectively (Hammerschmidt et al., 1996b; Fisher et al., 1997; Holley and Fergusson, 1997). It is very satisfying that this hypothesis has been confirmed by evidence that din alleles affect the zebrafish chordin homologue (hence the proposed new name of the locus, chordino; Schulte-Merker et al., 1997). First, din mutant allele and the zebrafish chordin gene were shown to be closely linked (