Current Topics in Developmenta1 Biology
Volume 32
Series Editors Roger A. Pedersen
and
Laboratory of Radiobiology ...
53 downloads
1503 Views
14MB 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
Current Topics in Developmenta1 Biology
Volume 32
Series Editors Roger A. Pedersen
and
Laboratory of Radiobiology and Environmental Health University of California, San Francisco San Francisco, California 94143
Gerald P. Schatten Department of Zoology University of Wisconsin Madison, Wisconsin 53706
Editorial Board Peter Gruss Max-Planck-Institute of Biophysical Chemistry D-37077 Gottingen, Germany
Philip lngham Imperial Cancer Research Fund Oxford, 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
Virginia Walbot Stanford University, California
Mitsuki Yoneda Kyoto University, Japan
Founding Editors A. A. Moscona Alberto Monroy
Current Topics in Developmental Biology Volume 32 Edited by
Roger A. Pedersen Laboratory of Radiobiology and Environmental Health University of California, San Francisco San Francisco, California
Gerald P. Schatten Department of Zoology University of Wisconsin Madison, Wisconsin
Academic Press San Diego
New York
Boston London Sydney Tokyo Toronto
Front cover photograph: Scanning electron micrograph of an E 1 I .5 mouse embryo just after amputation of the hindlimb bud and just prior to culture in a roller bottle. (See Chapter 6, Figure 6 for more details.)
This book is printed on acid-free paper.
@
Copyright 0 1996 by ACADEMIC PRESS, INC. 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.
Academic Press, Inc. A Division of Harcourt Brace & Company 525 B Street, Suite 1900, San Diego, California 921 01-4495 United Kingdom Edition published by Academic Press Limited 24-28 Oval Road. London NWI 7DX
International Standard Serial Number: 0070-2 153 International Standard Book Number: 0-12-1 53 132-5 PRINTED IN THE UNITED STATES OF AMERICA 96 97 9 8 9 9 00 01 EB 9 8 7 6 5
4
3
2
1
Contents
Contributors Preface xi
ix
1 The Role of SRY in Cellular Events Underlying Mammalian Sex Determination Blanche Cape1 Introduction I Classic Views of Sex Determination and the Isolation of Sly 4 Origin, Differentiation, and Cell Types of the Gonad Expression of Sry 13 Structure of the Genital Ridge Transcript and the SRY Protein V1. Future Directions 28 References 29 1. 11. 111. IV. V.
2
21
2 Molecular Mechanisms of Gamete Recognition in Sea Urchin Fertilization Kay Ohlendieck and William ). Lennarz I. Introduction
39
40 11. Multistep Recognition Process in Fertilization 41 111. Chernoattraction and Activation of Sperm 42 IV. Gamete Interactions at the Egg Plasma Membrane
V. Mechanisms to Prevent Polyspermy 54 VI. Egg Activation in Sea Urchins 55 VII. Prospects References 55
52
V
vi
Contents
3 Fertilization and Development in Humans Alan Trounson and Ariff Bongso
I. Introduction 59 11. Oocyte Maturation 61 111. Sperm Capacitation, the Acrosome Reaction, and the Sperm Maturation IV. Gamete Interactions 67 V. Fertilization 69 VI. Fertilization Abnormalities 72 VII. Micromanipulative Fertilization Techniques 74 76 VIII. Embryonic Cleavage and Developmental Anomalies IX. Determination of Genetic Errors in Gametes and Embryos 82 X. Cryopreservation of Oocytes and Embryos 84 XI. Embryo Metabolism and Viability 88 XII. Conclusions 91 References 92
4 Determination of Xenopus Cell lineage by Maternal Factors and Cell Interactions Sally Moody, Daniel V. Bauer, Alexandra M . Hainski, and Sen Huang
I. How Cell Lineages Are Studied
103
11. Why Study Xenopus? 105 111. Cell Fate Mapping in Xenopus
108 IV. Are There Early Progenitors for Specific Tissues, Organs, or Cell Types? 109 V. Does Position in the Mitotic Pattern Determine Cell Fate? 110 VI. Does Inheritance of a Maternal Cytoplasmic Factor Determine 1 13 Cell Fate? VII. Cell-Cell Signaling in Fate Determination 123 VIII. Conclusions 129 References 131
5 Mechanisms of Programmed Cell Death in Caenorbabdifis elegans and Vertebrates Masayuki Miura and lunying Yuan 1. Introduction 139 11. Programmed Cell Death in the Nematode C. elegans 140 111. Genetic Control of Programmed Cell Death in C. elegans 141
64
vii
Contents IV. Molecular Mechanisms of Programmed Cell Death in Vertebrates V. Do All Cells Have a Suicide Program? 163 VI. Future Prospects 164 References 164
149
6 Mechanisms of Wound Healing in the Embryo and Fetus Paul Martin
I . Overview and Introduction
175 176 History of Embryonic Wound Healing Studies 180 Reepithelialization of an Embryonic Wound Appears to Be Driven by 182 Contraction of an Actin Purse String in the Marginal Epidermal Cells 184 Assembly of the Purse String and Mechanism of Purse-String Closure Some Natural Morphogenetic Movements May Be Driven by the Same 188 Contractile Purse String That Closes an Embryonic Wound Contraction of the Embryonic Wound Mesenchyme 189 190 Role of Cell Proliferation during Embryonic Wound Healing 190 Early Signals for Initiating Tissue Movements of Wound Closure 191 Inflammation Does Not Occur Following Wounding in the Embryo Fetal Wound Healing Environment - Extracellular Matrix and Growth Factors 193 196 Adult Skin in the Fetal Environment and Vice Versa Is There a Critical Transition Phase in Late Fetal Development When Healing Becomes Adult-like? 197 197 Healing of Tissues Other Than the Skin (Not All Fetal Healing Is Perfect) Operating on the Human Fetus: Perfect Repair of Embryonic DefectsRealistic Dream or Fantasy? 198 References 199
11. Adult Wound Healing Review
111. IV. V. VI. VII. VIII. IX. X. XI.
XII. XIII.
XIV. XV.
7 Biphasic Intestinal Development in Amphibians: Embryogenesis and Remodeling during Metamorphosis Yun-Bo Shi and Atsuko Ishizuya-Oka I. Introduction
205
207 111. Intestinal Remodeling during Metamorphosis IV. Summary and Prospects 227 References 229 11. Embryogenesis of Amphibian Intestine
Index
237
212
This Page Intentionally Left Blank
Contributors
Numbers
in parentheses
indrmte the puges on which the uuthors' contributions begin.
Daniel V. Bauer Department of Anatomy and Neuroscience Program, The George Washington University Medical Center, Washington, District of Columbia 20037, and Department of Anatomy and Cell Biology, The University of Virginia Health Sciences Center, Charlottesville, Virginia 22908 (1 03) Ariff Bongso Department of Obstetrics & Gynecology, National University of Singapore, Singapore (59) Blanche Cape1 Department of Cell Biology, Duke University Medical Center, Durham, North Carolina 27710 (1) Alexandra M. Hainski Department of Anatomy and Neuroscience Program, The George Washington University Medical Center, Washington, District of Columbia 20037, and Department of Anatomy and Cell Biology, The University of Virginia Health Sciences Center, Charlottesville, Virginia 22908 (103) Sen Huang Department of Anatomy and Neuroscience Program, The George Washington University Medical Center, Washington, District of Columbia 20037, and Department of Anatomy and Cell Biology, The University of Virginia Health Sciences Center, Charlottesville, Virginia 22908 ( 103) Atsuko Ishizuya-Oka Department of Anatomy, Dokkyo University School of Medicine, Mibu, Tochiigi 321-02, Japan (205) William J. Lennarz Department of Biochemistry and Cell Biology, State University of New York at Stony Brook, Stony Brook, New York 11794 (39) Paul Martin Departments of Anatomy and Developmental Biology, and Plastic Surgery, University College London, London, United Kingdom (175) Masayuki Miura Department of Molecular Neurobiology, Institute of Medical Sciences, University of Tsukuba, Tsukuba, Ibaraki, Japan (139) Sally A. Moody Department of Anatomy and Neuroscience Program, The George Washington University Medical Center, Washington, District of Columbia 20037 and Department of Anatomy and Cell Biology, The University of Virginia Health Sciences Center, Charlottesville, Virginia 22908 (103) Kay Ohlendieck Department of Pharmacology, University College Dublin, Belfield, Dublin, Ireland, (39) ix
X
Contributors
Yun-Bo Shi Laboratory of Molecular Embryology, National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, Maryland 20892 (205) Alan Trounson Institute of Reproduction and Development, Monash University, Clayton, 3168, Australia (59) Junying Yuan Cardiovascular Research Center, Massachusetts General Hospital-East, Charleston, Massachusetts 02 129 (139)
Preface
This volume continues the recent tradition of this Series in addressing developmental mechanisms in a variety of experimental systems. The conceptual sequence of topics begins with sexual determination, continues with gamete recognition and fertilization, embryonic cell lineage determination, programmed cell death, and embryonic and fetal wound healing, and ends with amphibian metamorphosis. The chapter by Capel integrates classic studies of mammalian sex determination and testis morphogenesis with current molecular concepts of Sry function. Ohlendieck and Lennarz summarize current knowledge of the molecular mechanisms of species-specific sperm-egg recognition in sea urchins, especially the sperm receptor in Strongylocentrotus purpuratus eggs. Trounson and Bongso concentrate on the events of human fertilization, such as gamete maturation and sperm-egg interaction, including the major aspects of current in vitro fertilization technologies, such as sperm injection and blastomere biopsy. In their chapter, Moody and co-authors examine the relative roles of maternally inherited components and epigenetic interactions in cell fate determination in Xenopus laevis embryos. Miura and Yuan compare the mechanisms of programmed cell death in development of the nematode Caenorhabditis eleguns and in vertebrate organisms, especially mammals. The chapter by Martin presents vertebrate embryonic wound healing as a model for tissue movements during morphogenesis and as a counterpart to adult wound healing. Finally, the chapter by Shi and Ishizuya-Oka focuses on intestinal remodeling in X . luevis during metamorphosis from tadpole to adult as a model for embryonic organogenesis. Together with 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 vertebrate and invertebrate development, as well as to students and other professionals who want an introduction to current topics in cellular and molecular approaches to developmental biology. This volume in particular will be essential reading for anyone interested in fertilization, the role of growth factors in amphibian development, cell death, wound healing, and amphibian metamorphosis. 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. We thank the members of the Editorial Board for their suggestions of xi
xii
Preface
topics and authors. We thank Liana Hartanto and Diana Myers for their exemplary administrative support. Finally, we salute the Laboratory of Radiobiology and Environmental Health for hosting the Series office for the past five years and the Pedersen laboratory for the past 25 years. Roger A . Pedersen Gerald P. Schatten
1 The Role of Sry in Cellular Events Underlying Mammalian Sex Determination Blanche Capel Department of Cell Biology Duke University Medical Center Durham. North Carolina 27710
I . Introduction 11. Classic Views of Sex Determination and the Isolation of Sry
A. Testis-Determining Gene Triggers Male Development B. Isolation of Testis-Determining Gene Sry 111. Origin, Differentiation, and Cell Types of the Gonad A. The Gonad Arises within the Urogenital Tract B. Hormonal Propagation of the Signal C. Cell Q p e s of the Gonad D. Early Gene Expression in the Urogenital Ridge IV. Expression of Sry A. Expression of Sry in the Fetal Gonad and the Initiation of Cellular Differentiation B. Timing of Sry Expression versus Development of the Urogenital Ridge C. Level of Expression of Sry D. Other Sites of Expression of Sry V. Structure of the Genital Ridge Transcript and the SRY Protein A. Structure of the Genital Ridge Transcript B. Predicted Structure of the SRY Protein, the HMG Box Domain, and Binding Studies C. Interspecies Comparisons D. The Hypothesis that Sry Expression Is in Balance with a Gene on the X Chromosome E. Other Possibilities for the Action Sry VI. Future Directions References
1. Introduction Unlike many genes isolated over the past several years, Sry is a gene of known function, at least at the level of the organism. Some 50 years of microscopy, genetic, and endocrine research have produced a sound theoretical framework that predicted how the sex-determining gene must operate during development. Research conducted over the past 5 years on Sry, a gene isolated from the sexdetermining region of the Y chromosome, has shown that Sry satisfies many of these predictions. Sry acts as a classic genetic switch in development, initiating Currenr Topicr in Developinenml Blolo~v,Vo/. 32 Copyright 0 1996 by Academic Press, Inc. All rlghts of reproductiun in any form reserved.
1
2
Blanche Capel
development of a testis rather than an ovary from the indifferent gonad. Sry encodes a DNA binding protein, and presumably acts to regulate transcription of downstream genes within the male gonad that lead to bifurcation of cellular pathways of male versus female development. It affords an accessible model of how the expression of a single gene directs a new pathway of development from the point of view of both the initiation of differentiation within a cell and the initiation of new cellular interactions within a local cell community in the embryo. Our present task is to understand how this genetic switch operates in molecular detail. Nothing is known about the genes expressed immediately downstream of Sry; consequently a major aim of current research is to isolate some of the molecular players in this developmental pathway. Some of these genes are likely to be involved in cellular interactions that lead to morphogenesis of a testis. Cascades of gene expression will influence and be influenced by the process of morphogenesis as cells come into new signaling environments, and under the paracrine and juxtaprine influence of new neighbors. Other genes acting downstream of Sry must be involved in the hormonal pathways that export the maledetermining signal from the gonad. A second principle aim of current research is to understand the regulation of Sry itself at the transcriptional and post-transcriptional levels. Evidence is accumulating that the regulation of this gene is unusual and may occur at multiple levels between the control of transcription and the production of a functional protein. This review will focus primarily upon classic predictions about sex determination and testis morphogenesis, acquired from many areas of research, and their integration with the current molecular characterization of the modus operandi of sry.
II. Classic Views of Sex Determination and the Isolation of Sry A. Testis-Determining Gene Triggers Male Development
Male and female embryos are morphologically indistinguishable throughout the early stages of development in mammals. The embryonic gonad arises as an indifferent tissue whose cells normally follow one of two courses of development. If the Y chromosome is present, and the sex-determining gene is expressed at the appropriate time in development, testis organogenesis is initiated: the cells of the gonad, which were formally identical between male and female embryos, now begin to organize into cords and simultaneously acquire testis-specific characteristics at both the ultrastructural and the molecular levels. If this gene is not
1. Role of Sry in Mammalian Sex Determination
3
expressed, the cells of the gonad do not initiate organization until several days later when ovarian follicular structures begin to appear (Eicher and Washburn, 1986; McLaren, 1988, 199I). Subsequent masculinization of feminization of the embryo depends on hormonal secretions of a testis or ovary. In 1947 Alfred Jost showed that removal of a gonad from a male rabbit embryo resulted in feminization of the embryo, whereas removal of a gonad from a female had no effect on female development (Jost, 1947). This demonstration that male development requires the presence of a testis led to the hypothesis that the female pathway is the default pathway in development, and the male pathway requires the intervention of a testis-determining gene.
6. isolation of the Testis-Determining Gene Sry
When the Y chromosome was first observed cytologically, it was thought to provide a stable and heritable mechanism for determining male and female offspring in equal proportions, free of external influences (Wilson, 1909). However, in Drosophila melanogaster, the first Y-bearing animal in which the mechanism of sex determination was worked out, the presence or absence of a Y chromosome proved to be irrelevant: Bridges established that sex determination in Drosophila depends on the X chromosome-to-autosome ratio (Bridges, 1921). The segregation of the Y chromosome does indirectly control the number of X chromosomes since the natural sexes are XX (female) and XY (male). However, XO individuals are also male, and in some closely related species such as D . annulimana, and in most nematodes, there is no Y chromosome (White, 1973). For many years it was assumed that the X-to-autosome ratio would operate in a similar way to determine sex in mammals, but this proved not to be the case. By 1959, it became clear that the presence of a Y chromosome determines maleness in mammals, regardless of the presence of supernumerary X chromosomes. This finding localized the theoretical sex-determining gene to the Y chromosome in mammals (Ford et a l . , 1959; Welshons and Russell, 1959). The gene was termed Tdy in mouse and TDF in humans. In the intervening years between 1959 and 1990, the chromosomal location of TDF was narrowed to successively smaller regions of the human Y chromosome by cytological and genetic analysis (for review see Goodfellow and Darling, 1988). In 1966 the human gene was localized to Yp (Jacobs and Ross, 1966); by 1986, it was known to lie proximal to the pseudoautosomal region (Affara et al., 1986; Muller et al., 1986); by 1989 it was narrowed to within 200 kilobases (kb) of the pseudoautosomal boundary (Page et al., 1987). In 1990, its location was determined to be within 35 kb of this boundary by the analysis of XX males who carried small segments of the Y extending from the pseudoautosomal boundary, presumably transferred to the X in rare crossover events that did not resolve
4
Blanche Capel
within the pseudoautosomal region. The gene Sry was isolated in a chromosomal walk from the pseudoautosomal boundary through this region of the Y chromosome, and shown to be conserved on the Y chromosome among all mammals so far tested (Sinclair et al., 1990; Gubbay et al., 1990; Foster et al., 1992). The presence or absence of mouse Sry or human SRY was shown to correlate exactly with male versus female sex determination in a number of mutants available for analysis. For example, an XXSxrb mouse, carrying the smallest region of the mouse Y chromosome known to have the property of sex reversal (Cattanach et al., 197 l), was tested for the presence of Sry, and shown to carry this gene. Coordinately, Sry was found to be absent in an XY mouse that develops as a fertile female (Gubbay et al., 1990; Lovell-Badge and Robertson, 1990). Later Sry was shown to map to the center of a small 11-kb deletion in this Y chromosome (Gubbay et al., 1992). Further genetic support for the functional role of SRY in sex determination has accumulated from the study of XY humans that develop as females. A large number of mutations have been found within the highly conserved DNA binding domain of SRY in these individuals (Berta et al., 1990; Jager et al., 1991 Hawkins et al., 1992a,b; Hawkins, 1995), as well as changes in the 5' region of the gene (McElreavy et al., 1992). These experiments demonstrated that the deletion of Sry, or any one of numerous mutations within or 5' to the human SRY conserved domain, was associated with male-to-female sex reversal. The microinjection of 14 kb of genomic DNA containing the Sry locus into XX mouse embryos resulted in the development of male mice bearing testes, male secondary sex characteristics, and male mating behavior (Koopman et al., 1991). XXSry transgenics are sterile because of the presence of two X chromosomes that lead to an arrest in male meiosis (Burgoyne et al., 1992), and because of the absence of other loci on the Y chromosome that are critical for spermatogenesis (Conway et al., 1993). The Sry transgenic mouse demonstrated that the addition of a functional copy of Sry to an otherwise XX genome results in female-to-male sex reversal, and had the important implication that all genes that operate upstream or downstream of Sry in the male sex-determining pathway are present on the X or autosomes.
111. Origin, Differentiation, and Cell Types of the Gonad A. The Gonad Arises within the Urogenital Tract
1. The Origin of the Urogenital System
The urogenital system arises from the intermediate mesoderm, which lies between the somites and the lateral plate. In mouse, at about 9 days postcoitum (dpc), this mesoderm appears as a pair of mounds on the coelomic surface of the
1. Role of Sry in Mammalian Sex Determination
5
dorsal side of the embryo, lying on either side of the neural tube and dorsal aorta, and running the full length of the body cavity. The most anterior region of this mesodermal field, called the pronephros, disappears soon after it arises in vertebrates. The mid section, the mesonephros, gives rise to the gonad itself. The most posterior section, the metanephros, gives rise to the definitive kidney. The mesonephric duct, called the Wolffian duct later in development, first appears within the mesonephros in segments at the anterior end, eventually extending as a continuous duct, and ending in the cloaca at the posterior end of the animal. S-shaped tubules coalesce within the mesonephros on the side of the duct toward the dorsal aorta and grow to fuse with the mesonephric duct. These tubules contact capillary beds extending from the dorsal aorta, and, in some vertebrates, this system is believed to function as a primitive kidney during embryogenesis. The differentiation of the kidney is initiated by the growth of the ureteric bud from the mesonephric duct where it has joined the cloaca back into the metanephric region. Growth of this epithelial duct induces the organization of kidney tubules from mesenchymal cells in this region (Grobstein, 1956; Wartiovaara ef al., 1974; Herzlinger et al., 1993). Within the mesonephric region, a second pair of ducts, the Miillerian ducts, arise as an invagination of the surface epithelium, extending parallel to the Wolffian duct, but fused at their most posterior ends to the other duct of the pair. Both complete ductal systems coexist in both male and female embryos at 11.5 dpc (Fig. 1).
2. Possible Origins of the Cells of the Gonad The gonad begins to condense as a distinct tissue at about 10.5 dpc in the mouse and is first visible in scanning electron micrographs as a region that can be distinguished from the mesenchymal tissue on the inside of the mesonephros (Capel and Lovell-Badge, 1993) (Fig. 1). The origin of the cells that contribute to the gonadal tissue has long been a controversial issue. Some workers have suggested that cells from the mesonephric tubules contribute to the population of cells that constitute the gonad, and seed epithelialization in this region (Merchant-Larios, 1979; Wartenberg, 1982; Satoh, 1985; Kanai et al., 1989; Wartenberg et al., 1991). By analogy to extensive work on the induction of kidney tubules, this is a reasonable suggestion. Recent whole-mount staining with an antibody against laminin, present in the basement membrane of the mesonephric ducts and tubules, revealed cellular bridges extending from the mesonephric tubules to the gonadal primordium at early stages of gonad formation from 10.5 to 12.0 dpc in both male and female embryos. This phase of cell migration does not appear to be sex specific (Karl and Capel, 1995). Others have presented evidence that cells from the coelomic epithelium may also invaginate into the interior of the condensing gonad (Smith and MacKay, 1991). At least in chick, where dye labeling has been used to mark cells of the coelomic epithelium, this seems to occur in association with migrating germ cells
Fig. 1 (a) Diagrammatic representation of 11.5 urogenital ridge. The gonad arises bilaterally as part of the urogenital tract. The gonad condenses on the inside of the mesonephric region of the intermediate mesoderm, and is visible as a distinct tissue by I I dpc that is morphologically indistinguishable between male and female embryos. Two complete ductal systems form in both male and female embryos consisting of a mesonephric (Wolffian) duct and a Miillerian duct. The kidney arises as the result of an inductive interaction between the ureteric bud (which grows back from the region where the mesonephric duct joins the cloaca) and the metanephric region of the intermediate mesoderm. (B) Diagrammatic representation of sections through a developing male gonad over the period of cord formation from 9.5 to 12.5 dpc. 9.5-10.0 dpc: 'lhhules condense within the mesonephros and connect with the mesonephric duct, extending toward the dorsal aorta. Primordial germ cells (PGC) migrate into the region via the gut mesentery. 10.5 dpc: PGCs enter the gonadal primordia. A Miillerian duct forms by invagination of the coelomic epithelium in the mesonephros. 115 1 2 . 0 dpc: Sry is expressed in cells of the gonadal blastema. The cell population in this region expands by cell migration andlor proliferation. 12.5 dpc: Cords organize around germ cells. The three somatic cell types of the male gonad are morphologically distinguishable: myoid cells border the testis cords, Sertoli cells lie inside the cords surrounding the germ cells, and Leydig cells (not shown) lie in interstitial space. The Miillerian ducts disappear in response to AMH produced by the differentiating Sertoli cells. Mesonephric tubules aggregate near the center of the gonad and, together with the Wolffian duct, form the efferent ductal system of the male gonad.
1. Role of Sry in Mammalian Sex Determination
7
that enter the region at this time (Rodemer-Lenz, 1989). Scanning electron micrographs of mouse embryos from 10.5 dpc reveal pores from the surface epithelium into the interior of the gonad through which cells could be invaginating (Capel and Lovell-Badge, 1993). Labeling the cells lining the coelom with a Iive dye may reveal whether epithelial cells enter the interior of the developing gonad. Primordial germ cells are completing their migratory pilgrimage from the base of the allantois through the gut mesentery and are arriving in the gonads during this period of development (Ginsburg et a l . , 1990). In mouse, primordial germ cells are thought to enter the gonad through the mesonephros. A double stain with SSEA- 1, an antibody that specifically labels germ cells, should help to clarify the issue of whether any germ cells are associated with the epithelial surface of the gonad in mouse. In any case, it is clear that the presence of germ cells in the fetal testis is not a requisite for cord formation in the testis. In sterile mutants such as dominant white spotting (We) or steel (Sl?, which have no germ cells, testis cord formation proceeds normally (Mintz and Russell, 1957). This is not the case in female development, where the presence of germ cells seems to be required to seed follicular organization (McLaren, 1985). In vitro organ culture systems have demonstrated that cells from the mesonephric portion of the urogenital ridge (ugr) must enter the gonad between 11.5 and 12.5 dpc in development in order for cord formation to occur. Experiments with marked cells suggest that cells migrating from the mesonephros contribute to the myoid and (perhaps) Leydig cell populations in the differentiating testis (Buehr et a l . , 1992; Merchant-Larios et a l . , 1993; Moreno-Mendoza et a l . , 1995). Recently we have shown this wave of cell migration depends on the presence of an XY gonad: in organ culture experiments in which the gonad comes from an XY embryo, whether the mesonephros is XX or XY, cells migrate into the gonad; if the gonad is from an XX embryo, there is little cell migration, regardless of the source of the mesonephros (Capel et a l . , in preparation). From a molecular point of view, this means that there is a phase of cell migration that is signaled downstream of the expression of Sry and is critical for the process of cell organization that leads to testis cord formation. Two other developmental mechanisms that are known to show differences between male and female gonads at this stage are the rate of proliferation and the extent of vascularization. Rates of proliferation of male versus female cells in the mesonephros and gonad during the period of gonadogenesis have also been investigated and show to be higher in mesonephros than in gonadal tissue, and higher in male than in female tissue at the earliest stages of gonadal development (Merchant-Larios, 1979; Merchant-Lanos and Taketo, 1991). Since extensive vasculature is evident earlier in the male than the female gonad, it is thought that this may be another pathway through which Sry is signaling to induce cellular organization of the testis. Some of the cells migrating from the mesonephros appear to be endothelial cells (Buehr et a l . , 1992; Merchant-Larios et a l . , 1993; Capel et a l . , in preparation).
8
Blanche Capel
B. Hormonal Propagation of the Signal In mammals there is more than one process controlling sexual differentiation. The initial step that is classically considered the sex-determining step is the expression of Sry , which controls a molecular switch in development. However, downstream pathways of sexual differentiation that are controlled by endocrine mechanisms are rapidly activated in the differentiating gonad, and these pathways clearly affect (and can even reverse) the initial determination to develop as a male or female. In flies and nematodes, sex determination is directly linked to X-chromosome dosage compensation, and both decisions are accomplished simultaneously in all cells. As Hodgkin has pointed out, this mechanism has the effect of unifying all cells of the organism since cells that do not succeed in controlling dosage in a manner appropriate to their genotype die (Hodgkin, 1992). In mammals, where sex determination occurs only in cells of the gonad, the goal of unifying the development of all cells of the organism is accomplished by hormones. The hormonal system has the important advantage that it allows for fluctuation of sexual differentiation throughout the life of the animal, for example, in estrus or pregnancy, and consequently builds a great deal of flexibility into the system. In mammals, once the fate of the gonad is determined by the expression or nonexpression of Sry, the sex-specific organization of the cells of the gonad follows, and all secondary sex characteristics derive from the hormonal secretions of the testis or ovary when differentiation of the cell types of the gonad is under way. One of the earliest known markers of testis development is antiMullerian hormone, AMH (also called Mullerian inhibiting substance, MIS). This “hormone,” which has now been cloned in a number of mammals (Cate et al., 1986; Picard et al., 1986; Munsterberg and Lovell-Badge, 1991), is actually a TGF-P-like growth factor produced by Sertoli cells. Under the influence of this factor, the Mullerian ducts disappear in male embryos. Testosterone, which is an early product of the Leydig cells of the testis, is important for the maintenance and elaboration of the Wolffian ducts, which will constitute the male efferent ductal system, including the vas deferens and epididymis. In the female where neither AMH nor testosterone is expressed, the Wolffian ducts are lost, and the Mullerian ducts are elaborated to form the oviduct and uterus and part of the vagina.
C. Cell Types of the Gonad
The cells of the indifferent gonad are thought to be bipotential, able to differentiate as the cell types of the ovary or testis. In addition to germ cells that are present in males and females, three somatic cell lineages are defined in the adult ovary and testis: supporting cells, which differentiate as Sertoli cells in males, or follicle cells in ovaries; steroidogenic cells, which may be Leydig cells in males or Theca cells in females; and connective tissue cells, which constitute the blood
1 . Role of Sry in Mammalian Sex Determination
9
vessels and tunica of both organs, and give rise to myoid cells in males, and stromal cells in females. The origin of these different cell types among the cells of the undifferentiated genital ridge is unknown. No markers exist before cells are morphologically distinguishable once cords or follicles begin to organize. Numerous experiments have illuminated the flexibility (and reversibility) of the cells of the gonad to follow male or female differentiation pathways. In ovaries where germ cells are absent or have disappeared, the tissue often bears more structural similarity to a testes than an ovary, containing cells that bear a strong resemblance to Sertoli cells (for review see Burgoyne and Palmer, 1993). This is also true of fetal ovaries that have been transplanted beneath the kidney capsule of adult male mice (Taketo and Merchant-Larios, 1985). One of the best studied examples of this conversion phenomenon is the freemartin. In calf twins, who share a common blood system through placentas which are anastomosed, a male can masculinize a female sibling by the presence of circulating AMH (Jost et ul., 1975). This phenomena has been explored more directly by culturing rat ovaries with AMH in vitru (Vigier et aE., 1987), and by the construction of XX mice expressing a transgene for AMH (Behringer et al., 1990). In either case, a conversion to testis-like organization is observed. These experiments have led to the hypothesis that AMH has a direct role in testis determination; however, there are men with persistent Mullerian ducts, who bear a documented mutation in their AMH gene, yet have fully developed testes (Josso et al., 1991). This shows that AMH is not necessary for the development of a testis, although it has an important role in propagating some aspects of male development downstream of Sry, and seems to be sufficient to redirect development along the male pathway in some abnormal circumstances. The ability of secreted hormones and growth factors downstream of Sry to induce a global conversion of tissue complicates the task of assigning direct roles in sex determination to genes or proteins that have sex-reversing properties. The system seems to be finely balanced between the male and the female pathways. The cells of the gonad function as a community: development is strongly canalized such that once a pathway is determined, the entire population of cells is normally recruited into that pathway. Perhaps for this reason, true hermaphrodites are rare among mammals, and hermaphrodites that do arise more frequently have a testis on one side and an ovary on the other rather than bilateral ovotestes. However, this is not true of the European mole Talpu uccidentulis, in which fertiIe females typically develop bilateral ovaries containing regions of testicular tissue (JimCnez, 1993). Perhaps this exception can help us to better understand the general rule.
D. Early Gene Expression in the Urogenital Ridge
From the point of view of molecular data, expression of a growing number of genes has been observed in the developing ugr. A representative sampling includes a family of homeobox genes, Hox 0 . 2 , 0.3, 0 . 4 ; the paired box gene,
I
- I-E m
I
4 I I
1. Role of Sry in
Mammalian Sex Determination
11
Pax-2; Lim-I; the Wilm’s tumor gene, WT-I; two steroid receptors, SF-I and Dux-I; and at least two Sox genes, Sox-3 and Sox-8. After Sry is expressed, AMH, other hormones and steroids, and their associated synthetic enzymes such as testosterone and aromatase have been characterized in developing testis (Fig. 2). Other genes known to be involved in primordial germ cell migration and proliferation in the ridge, such as c-kit, Lif, and S1, have also been characterized, but do not appear to be critical for gonad development, and will not be considered here. Three members of the HoxD family are expressed in the ugr beginning before 10.5 dpc, and continuing into fetal gonad stages (Izpisua-Belmonte et al., 1990, 1991). HoxC.3 has also been seen within the 12.5 dpc kidney and developing testis (Gaunt el al., 1990). These genes are most closely related to the Abd-B homeobox gene in Drosophila. Abd-B and abd-A have been shown to play critical roles in specifying the position of the development of the gonad along the body axis in Drosophila. Overexpression of abd-A and Abd-B beyond their normal expression boundaries causes development of a gonad in a more diffuse region (Delorenzi and Bienz, 1990; Karch et al., 1990; Cumberledge et al., 1992). Mutants in abd-A show little evidence of gonadal differentiation, and Abd-B mutants show abnormal, diffuse gonad formation (Brookman er al., 1992). In mammals there is no information yet on the specification of the region along the length of the intermediate mesoderm that will give rise to the ugr and the gonadal blastema. The three members of the HoxD family show subtle, quantitative differences between genes, but no sexual dimorphism in the mouse embryo, and they continue to be expressed in developing testis cords and condensing follicles of the ovary. Lim-I is known to be expressed in the intermediate mesoderm, and may have a role in establishing the competence of this tissue, but the specific expression profile during ugr development is not documented (Barnes et al., 1994). Mice with no Lim-1 expression form no gonads (Shawlot and Behringer, 1995). The paired box gene Pax-2 and the Wilm’s tumor gene W T - I have come under
Fig. 2 Time course of expression of known molecular markers in the urogenital ridge. Expression of three members of the HoxD family has been characterized in the ugr from the earliest stages of intermediate mesoderm formation through 13.0 dpc (Izpisua-Belmonte et al., 1990, 1991). Expression of WT-1 has also been seen in mesenchymal tissue of the ugr from 9.5 through adult life (Pelletier er al., 1991b). Pax-2 expression is seen in all mesonephric ducts and condensing mesenchyme of the kidney between 10 and -13 dpc (Dressler et al., 1990). The expression patterns of HoxD, W T - I , and Par-2 genes appear to be identical in males and females. Sry is expressed in male embryos, specifically between 10.5 and 12.5 dpc. Amh also shows a sexually dimorphic expression pattern, appearing between 12 and 12.5 dpc in male embryos, declining postnatally, and first appearing in female embryos about the time of birth (Miinsterberg and Lovell-Badge, 1991). SF-I is expressed identically in male and female embryos at the earliest stages of gonadogenesis, then is shut off selectively in female embryos for the period between 12.5 and -18 dpc (Ikeda et al., 1994). In male gonads SF-I expression is continuous until about the time of birth.
12 Blanche Cape1 extensive investigation in kidney development. Both of these genes are expressed from the early stages of ugr development, but show no sexual dimorphism. Because of the common origin of the gonad and the kidney, it is not surprising that disorders which affect the development of the kidney often also affect gonad development (Pelletier et al., 1991a; Phelps and Dressler, 1993). WT-I is expressed in mesenchymal tissue of the mesonephros from its time of origin at 9.5 dpc. Expression is limited to somatic cell types of the testis and ovary in adult life (Armstrong et al., 1992; Pelletier et al., 1991b). Pax-2 expression has been documented from about 10 dpc, at the time of the elaboration of the mesonephric ductal system. Expression of Pax-2 is limited to the epithelialized ducts, tubules, ureteric bud, and the condensing mesenchyme at the tips of these tubules. It is thought to be involved in the mesenchyme-to-epithelial transition. Expression declines once condensation of the mesenchyme has been established (Dressler et al., 1990; Dressler and Douglas, 1992). The expression of SF-1 begins at about 9.5 dpc in the ugr in both male and female embryos. It has been shown by in situ hybridization that this gene develops a sexually dimorphic pattern later in development. In males this gene continues to be expressed during testis organogenesis, declining about Day 18.5. In females, SF-I expression is thought to be down-regulated between 12.5 and 18.5 dpc. Up-regulation would then coincide with the onset of steroidogenesis in the ovary (Ikeda et al., 1994). This hiatus of expression in the female suggests that at least one of the downstream effects of Sry expression is to sustain expression of enzymes in the steroid pathway in the Leydig cells of the male gonad. An SF-1 knock-out mouse shows complete adrenal and gonadal agenesis in male and female animals, suggesting an early role for this gene in development of the ugr (Luo et al., 1994). Dux-I, Sox-3, and Sox-9 are all known to be expressed in the ugr, but their specific expression profiles are not yet known (A. Swain, A. Hacker, personal communications; Wright et al., 1995). SOX-9 has been linked to the sex determination pathway in humans by virtue of the fact that it was isolated from the chromosomal breakpoint of a campomelic dysplasia patient (Foster et al., 1994; Wagner et al., 1994). This disease is known to be associated with male to female sex reversal. Since SOX9females develop normal ovaries, this gene is speculated to act somewhere in the male pathway. The earliest expression of Amh has been observed between 1 1 and 11.5 dpc in male gonads (Hacker et al., 1995). For many years AMH was the earliest known marker of Sertoli cell development. It seemed a likely direct target for Sry. Although a binding sites for Sry can be identified in the region 5’ of Amh, mutation of this site in the human gene does not affect the activation of Amh in vitro (Haqq et al., 194). These data suggest that Arnh is more likely an indirect target of Sry. After 12.5 dpc many male specific hormones and receptors are found in the developing testis.
I . Role of Sry in Mammalian Sex Determination
13
IV. Expression of Sry A. Expression of Sry in the Fetal Gonad and the Initiation of Cellular Differentiation Sry is expressed specifically during the narrow window in development when condensation of the fetal gonad and the initiation of cord formation occur in male embryos (Koopman et al., 1990). This correlation is striking in terms of its timing and spatial restriction. Reverse transcription polymerase chain reaction (RT-PCR) expression studies and RNase protection studies (Hacker et al., 1995) detect Sry expression specifically in the ugr at 10.5 dpc, with a peak of expression occurring at 11.5 dpc, and a rapid decline thereafter, such that at 12.5 dpc, expression is nearly undetectable. Several lines of evidence suggest that differentiation of the Sertoli cell requires the cell autonomous expression of Sry. First, the expression of Sry has been localized by in situ hybridization at l l .5 dpc to the region of the ugr where Sertoli cells arise as the first morphologically distinguishable cell type (Koopman et al., 1990). Second, in a genetic mosaic animal with contributions of XX and XY cells, 95% of Sertoli cells in the mature testis are XY, whereas other somatic cell types of the testis, Leydig and myoid cells, can be XX or XY with equal probability (Burgoyne et al., 1988; Palmer and Burgoyne, 1991a). These data suggested that although the Y chromosome was not essential for the differentiation of myoid or Leydig cells, it was required for the initial differentiation of Sertoli cells. This point has been further refined with studies of chimeras between XYSry+ c-, XYSry- (XV), where cells from the XYSry- component rarely contribute to the Sertoli cell population (Burgoyne and Palmer, 1993). Data from organ culture experiments also support this view. When the gonad and mesonephros from 1 1.5 dpc are explanted and cultured for several days in v i m , as long as the gonadal portion of the explant comes from an XY embryo, cords form whether the mesonephros is contributed by an XX or an XY embryo (Buehr et al., 1992). The finding that a Y chromosome is required only in the gonadal portion of the explant is consistent with the observation that expression of Sry is limited to the gonadal portion of the ridge by in situ hybridization. Other mosaic studies where the contribution of XY cells is very low have indicated that if at least 25% of the cells of the gonad are XY, the organ will develop as a testis (McLaren, 1984; Palmer and Burgoyne, 1991b). These results are generally believed to mean that the initiation of Sertoli cell differentiation requires the cell autonomous expression of Sry. Once 25% or more of the cell population begins Sertoli cell differentiation, XX cells can be recruited to this population, presumably by cellular interactions during the process of testis cord formation. Since cell autonomous Sry expression is not required for the differentiation of
14
Blanche Capel
myoid and Leydig cells some early signaling pathway must exist between the Sertoli cell and its neighbors, which initiates differentiation of these cell types. This step is critical for both structuring the architecture of the testis and establishing the synthesis and secretion of the steroid hormones that masculinize the embryo. No early markers are known for either myoid or Leydig cells before they acquire their typical morphology, surrounding the organizing tubules of the testis. The relationship between the changing morphology of the gonad as it develops and the process of molecular differentiation as measured by changes in gene expression is unknown. In other systems, for example, the developing mammary gland, the process of epithelialization and contact with extracellular matrix elements is known to affect gene expression through laminins, integrins, and other proteins that connect the cells with their local environment (Strueli et al., 1991, 1993). It is important to bear in mind that gene expression cascades will both influence and be influenced by structural changes as cords form in the developing testis. A number of experimental systems indicate that cell-cell interactions are critical for early steps of testis development. Sertoli and myoid cell populations purified from adult testis have been shown to influence positively each other's survival in culture (Skinner et al., 1985; Tung and Fritz, 1980, 1986). Mixed populations of cells also have been shown to influence each other's expression patterns in short-term explant culture (ring and Fritz, 1980), as have interactions with the extracellular matrix (Dym et al., 1991; Suarez-Quian et al., 1985). Organ culture experiments indicate that the mixing of cells from mesonephros and gonad is critical for cord formation (Buehr et al., 1992; Merchant-Lanos et al., 1993). Nothing is known about the molecular nature of the cellular interactions in the fetal gonad through which differentiation is propagated and tubule formation is effected. This problem is complicated by a lack of appropriate markers for different cell types at the earliest stages, and because there is a shortage of material for analysis since the window in development when this organization occurs is very narrow, and the fetal gonad is very small. It is hoped that immortalized cell explants from the ugr at this stage of development will provide adequate material for a biochemical approach to the identification of these early signaling pathways (Capel et al., submitted).
B. Timing of Sry Expression versus Development of the Urogenital Ridge An additional important finding from organ culture experiments is that the XY genital ridge must reach a particular stage of development before it is competent to initiate cord formation in vitro. Intact XY genital ridges explanted before 11.O dpc did not initiate cord formation (Buehr et al., 1992; Taketo and Koide, 1981) even though expression of Sry appears in vivo by 10.5 dpc. This finding suggests
1. Role of Sry in Mammalian Sex Determination
15
that some signals or cellular interactions are required between the genital ridge and its in vivo embryonic context before 11 dpc in order for the competence to form cords to be established within the genital ridge itself. This notion of a narrow window of development in which the events of sex determination are precisely orchestrated is not new. Genetic analysis conducted on the Y chromosome from a strain of mice isolated in Poschiavinus, Switzerland (YP)), has also suggested such a pattern. Within the Poschiavinus strain this Y chromosome behaves normally, but when crossed on certain genetic backgrounds YPOSgives rise to a high percentage of intersex or sex-reversed offspring (Eicher ef al., 1982). It has been suggested that this is a timing defect, which results from a mismatch between the expression of the sex-determining gene on YpOs and the morphogenesis of the gonad in some inbred strains. Analysis of this genetic case was largely responsible for a classic model proposing that the testisdetermining gene acts within a narrow window of development to deflect development from the ovarian (default) pathway. If the level or timing of Sry expression is not appropriate, the ovarian program is initiated, and the opportunity for initiation of testis development is past (Eicher and Washburn, 1986; Palmer and Burgoyne, 1991~).Two other cases of genetic loci that affect the frequency of sex-reversal are the dominant white spotting locus, W , and T hairpin fail.It is not apparent how these mutations interfere with the process of sex determination, but one possibility is that among the aspects of morphogenesis that these mutations affect, there is one that renders the timing of gonadogenesis out of synchrony with the timing of Sry expression. Since both of these mutations exhibit their sexreversing effects of the C57BL/6 background (as does YPos) it has been suggested that this background has some predisposing factor such that a slight delay in the onset of the testis-determining pathway fails to preempt the ovarian pathway (Burgoyne and Palmer, 1991).
C. level of Expression of Sry Experiments in several different systems indicate that the level of expression of S l y at 11.5 dpc is critical for sex determination. First, among transgenics carrying 14 kb of DNA covering the Sry locus, there was an initial finding that only about 30% of XXSry embryos develop as males (Koopman e l al., 1991). It seemed a reasonable hypothesis that this effect resulted from differences in the expression of the transgene from different integration sites. Subsequent analysis of carrier lines with variable rates of sex reversal is contributing more information about this phenomenon (N. Vivian, personal communication). Second, a series of Y-chromosome deletion mutants that showed heritable sex reversal were generated. In one of these lines that fails to initiate male development, expression of Sry has been shown to be reduced or absent at 11.5 dpc by RT-PCR analysis (Cape1 e f al., 1993a).
16
Blanche Cape1
1. Transgenic Studies The transgenic experiment indicated that Sry alone, acting on an otherwise XX genetic background, is sufficient to trigger male development; all other genes that are necessary in the pathway are present on the X chromosome or autosomes. However only 30% of transgenics were sex-reversed in this stringent functional test for a transgene. Since all F1 transgenics have an identical genetic background, this result suggested that either the copy number or the integration site might be critical for determining whether the transgene functioned to initiate male development. In subsequent experiments, lines were established from nonsex-reversed, fertile carriers (XXSry). Some of these lines produced no sexreversed offspring and some produced a variable frequency of XxTry males and intersexes. Homozygosity or high copy number of the transgene might have been expected to correlate with a high frequency of sex reversal, but this correlation has proved to be inconsistent (N. Vivian, personal communication). It is difficult to analyze the level of Sly expression relative to sex reversal because the time at which Sry is expressed in the embryo precedes the time at which sexual phenotype can be determined. In other words, if an embryo is killed to determine its level of Sry expression, one cannot also know whether it would have developed as a male. Interpretation of these results in later generations is complicated by the fact that establishing lines from the original Fl(C57BL/6 X CBA) transgenics requires inter- and back-crosses that result in segregation of the genetic background in the F2 and later generations. It has become obvious over the course of these breeding experiments that such background effects must be influencing the rate of sex reversal since, within a line, the sites of integration and copy number are constant, yet the rate of sex reversal is variable. Whether background effects are influencing the level, timing, or duration of Sry expression, or some other aspect of the pathway, is not clear. Matings are under way to cross a single transgenic line onto several different inbred and random bred backgrounds to determine whether the frequency of sex reversal is affected. Autosomal loci such as Tas, Tda-I, and Tda-2 have been postulated to interact with the testisdetermining gene during the orchestrated events of sex determination (Eicher, 1988). It has been suggested that such genes may have coevolved within a strain to act in a coordinated manner (Washburn and Eicher, 1989; Burgoyne and Palmer, 1991). If Sry’s ability to initiate male sex determination is at a threshold level in some transgenics, these breeding experiments may uncover other genes that interact in some critical way with this developmental pathway. Such interactive loci might be expected to affect sex determination at any of a number of different steps. For example, the timing or level of Sry expression might be affected at either the transcriptional or the post-transcriptional levels. Nothing is known about the molecular steps involved in the activation of Sry itself except that since it is activated on an XX background in the case of X m r y transgenics, all necessary upstream factors must be present. Second, as discussed
1. Rolc of Sry in Mammalian Sex Determination
17 above, interactive loci may affect the developmental rate of the genital ridge itself, determining the window of time during which expression of Sry is critical to trigger male development. Third, some of these genes may act coordinately with the SRY protein either as cofactors or competitive binding proteins for the control of transcription of downstream genes in the sex-determining pathway. 2. Deletion Mutants A second set of experiments has revealed more information about the relationship between the level of expression and chromosomal location of Sry. Normally the unique region of the Y chromosome has no pairing partner during male meiosis, which has made it difficult to map the order of genes in this region. Cattanach designed a set of breeding experiments placing the Sxr region opposite the Y chromosome (XSjcr Y) (Capel et al., 1993a). This experiment was designed to map the order of loci in this region by standard recombination methods, and also was expected to lead to anomalous recombination events in the unique region of the Y that might generate changes in Sry leading to sex-reversed phenotypes. In these experiments, the X chromosomes were marked with a coat color gene such that offspring carrying one X were distinguishable from offspring carrying two X chromosomes. This system provides a rapid screen for offspring whose coat color is not in accordance with sexual phenotype. These crosses did generate offspring that developed as females, but carried a Y chromosome, termed Yd. Strains carrying these Yd chromosomes were established (usually from the more fertile XXYd females), and the phenotype provided to be heritable. Originally it was expected that many of these sex-reversing recombination events would have generated changes in and around the Sry gene. When these strains were analyzed at the molecular level, it turned out that the Sry gene itself and more than 14 kb surrounding the gene were intact. However, deletions affecting a series of repeats between Sry and the centromere had occurred in all cases. Although the order of genes in the sex-determining region of the Y is not known for certain, these experiments strongly suggest that Sry lies most proximal. Figure 3 depicts the probable gene order in the unique region of the mouse Y chromosome (Mitchell and Bishop, 1992; Capel et al., 1993a). Expression of Sry at 11.5 dpc is greatly reduced or absent in these deletion mutants, leading to the hypothesis that the shift in the chromosomal location of Sry relative to the centromere has affected the timing or level of expression. Perhaps Sry has come under the influence of heterochromatin in this region and is quite sensitive to this effect. This explanation has been offered before to account for cases when translocation of this region of the Y to the inactive X has resulted in the selective loss of testis-determining function when other genes interspersed in this region, for example, the male specific antigen (Hya),remain active (Evans et al., 1982; McLaren and Monk, 1982; McLaren et al., 1984). The mechanism by which such an effect on expression is propagated is unknown, and might
Blanche Capel
18 YpTel
Zf 1.86
k
2.6C
S x l D (1.8) 30 kb .(
>goo kb
217 kb Sxl C (2.6)
YqTel Fig. 3 The map order of the Y chromosome short arm (Yp) and probable location of Sry. The H-Y antigen (Hyu), the spermatogenesis gene (Spy), and the ubiquitin activating enzyme homologue (Ubely-I),as well as molecular probes Zf 1.8B,2.6c, and the S x l D repeats, are all known to map between the zinc finger genes Zfy-I and Zfy-2, although the order of Hya and Spy is unknown. This entire contig is greater than 900 kb, and may lie in either orientation relative to the centromere. Sry and the Sxl C repeats are known to lie outside this contig, but may map on either side, near the telomere or near the centromere (Mitchell and Bishop, 1992). The deletion mutant study implies that the Sxl C repeats lie between the centromere and Sry, and places Sry proximal to the 900-kb contig (reprinted from Capel et al., 1993a with permission).
involve specific remote elements that control accessibility of the chromatin, as have been defined at the globin locus (Philipsen er al., 1993), or a more general effect of heterochrominization on expression, for example, affecting the timing of replication of this locus.
1 . Role of Sry in Mammalian Sex Determination
19
D. Other Sites of Expression of Sry Sry expression has been reported to occur in preimplantation mouse embryos at the two-cell and blastocyst stages (Zwingman et al., 1993). RT-PCR experiments using high cycle numbers have also revealed expression of Sry in other fetal tissues such as brain and spleen, but no expression in these adult tissues (Koopman et al., 1990; Jeske et al., 1995). It is not clear what significance these sites of expression might have. XY preimplantation embryos have been shown to grow faster than XX embryos, and a scheme for sex determination based on an accelerated male growth rate has been proposed (Tsunoda et al., 1985; Mittwoch, 1986). The situation has become more complicated with the discovery that there are two phases of growth-rate differences between male and female embryos, a preimplantation and a postimplantation effect. The preimplantation difference in growth rate has proved to be inconsistent and depends on a strain-specific effect of the Y chromosome and on the absence of a paternal X chromosome in male embryos (Burgoyne and Thornhill 1993; Thornhill and Burgoyne, 1993). The postimplantation effect of the Y chromosome does not map to Sry. Embryos bearing a Y chromosome with a deletion of 1 1 kb surrounding and including the Sry locus (Y) still show a growth advantage (but do not develop as males). Furthermore, embryos carrying Sxr on their X chromosomes (and thus Sry) show no postimplantation growth advantage in the absence of a Y chromosome (P. Burgoyne, personal communication). At present it is difficult to see how early growth differences might be related to the bifurcation of the male and female pathways of gonadal development. Although Sry is also expressed in adult testis, probably in round spermatids, the transcript is very different from the transcript in the ugr. The testis transcript exists as a circular RNA molecule, probably formed by splicing from a longer primary transcript. The start site and termination site of this primary transcript are still unmapped, but known to extend more than 500 base pairs (bp) upstream and 4 kb downstream of the coding region (Fig. 4) (Hacker et al., 1995). Since all known genes on the Y chromosome are expressed in adult testis, it has been suggested that transcription from the Y is ubiquitous in this tissue, and may run throughout the region from cryptic promoters far upstream. The mechanism by which the circular transcript forms is under investigation. The circular transcript does not sediment with actively translating polysome fractions, and seems unlikely to be translated (Capel et al., 1993b). Expression from the 14-kb genomic construct in heterologous cell types has also failed to produce protein (A Swain, personal communication). In addition, genetic analysis has so far revealed no function for Sry in the adult testis. Expression in the adult testis depends on the presence of germ cells, unlike the case in genital ridge where the presence of germ cells is not required for Sry expression (Koopman ef al., 1990). It is thought from cell separation studies that
20
Blanche Capel
GENITAL RIDGE TRANSCRIPT 8027
13000
SA
GENOMIC LOCUS
SD
v
--->
--7480
8201
9432
10219
ADULT TESTIS TRANSCRIPT
0 9432
5 "I
I I
Fig. 4 The mouse transcript for Sty is linear in urogenital ridge and exists as a circle in adult testis. The 14-kb region encompassing the genomic locus has been sequenced (GenBank Accession No. X67204) and shown to encode a conserved DNA binding domain (hatched) within 2.8 kb of unique sequence at the center of a large inverted repeat (Gubbay et al., 1992). There is an open reading frame extending from just 5' of the box 17. Two stop codons are indicated V.The initiation and stop sites of the primary transcript in adult testis are unknown, but extend into the inverted repeat upstream and downstream of the unique region. This could generate a transcript with a stem loop structure which may facilitate the use of splice donor (SD) and splice acceptor (SA) sites in the circularization of the transcript in adult testis (Capel et al., 1993b). Three specific start sites have been mapped in the ugr transcript around 8027, and the major termination site is near bp 13,000, giving rise to a ugr transcript -5 kb in length (Hacker et al., 1995).
Sry is expressed in round spermatids, although lower levels of expression in
additional cell types of the testis has not been excluded. XYY mice carrying a Y chromosome deleted for 11 kb surrounding and including the Sry locus can produce functional sperm carrying Y , and mosaics of the genotype XY t,X Y also transmit the Y chromosome (Lovell-Badge and Robertson, 1990). These cases demonstrate that Sry is not required cell autonomously for sperm development, but do not rule out the possibility that expression of Sry from the normal Y
1. Role of Sry in Mammalian Sex Determination
21
chromosome in the same testis can compensate for the absence of Sry in adjacent sperm. Nonetheless, the unusual structure of this transcript argues against a function for Sry in the adult testis, and brings the structure of the transcript at other sites of expression into question. Conclusive information about the significance of Sry expression at sites other than the genital ridge will await a good method for detecting the presence of the protein, or, better yet, some direct functional effect that can be demonstrated genetically.
V. Structure of the Genital Ridge Transcript and the SRY Protein A. Structure of the Genital Ridge Transcript
In the case of the genital ridge, the form of the transcript has been shown to be different from the circle found in adult testis (Cape1 et al., 1993b). Since the expression level of Sry is very low in the ugr, it has been impossible to determine the size of the transcript from Northern analysis. Precise start sites have been mapped at the 5' end of the ridge transcript by RNase protection and 5' rapid amplification of cDNA ends, and recently, the 3' termination site has been determined far downstream, defining a transcript nearly 5 kb in length (Hacker et al., 1995). The mouse gene consists of a single exon, encoding a protein that is 395 amino acids long, including a region of glutamine, histidine repeats at the C-terminal end of the DNA binding domain that is completely absent in the human gene. The extensive 3' untranslated region of the ugr transcript suggests that it could be important for post-transcriptional control of this gene. Taken together with the precise window of Sry expression seen in the fetal gonad, and the circle formation in adult testis, these results all suggest a criticaI control of the timing and level of Sry expression coupled with a tight control of translation of the protein. B. Predicted Structure of the SRY Protein, the HMG Box Domain, and Binding Studies
SRY is a member of a family of DNA binding proteins (Gubbay et al., 1990; Harley et al., 1992). It is thought to act as a transcription factor, regulating a set of target genes required to trigger testis development (Koopman, et al., 1990; Lovell-Badge, 1992a). A growing number of Sry-related genes, called Sox genes, have now been isolated in diverse species from human to Drosophila (Denny et al., 1992a,b; Stevenovic et al., 1993; Collignon et al., submitted). All of these genes encode proteins that are related through a conserved DNA binding domain, the HMG box. The HMG box was originally characterized in non-
22
Blanche Cape1
histone, high mobility group proteins associated with chromatin. Some members of this family of proteins recognize and bind cruciform structural elements in DNA and show no sequence specificity. A second class of HMG-type proteins, to which Sry and the Sox genes are more closely related, show a sequence specific binding affinity, although this class of HMG-type proteins will also bind cruciform DNA (for review see Lilley, 1992). It has been shown that Sry (and other HMG-type proteins which have been tested) induce a 120" bend in DNA when bound (Giese er al., 1991, 1992; Bianchi et al., 1992; Ferrari et al., 1992). This is a tantalizing observation, but, so far, it is not clear how to interpret it. Affinity of SRY for bent DNA is higher than that for a linear substrate. However, since the amount of SRY protein that might be present in a cell is orders of magnitude lower than the more ubiquitous HMG chromatin-binding proteins that recognize a similar motif, SRY is not expected to compete effectively for any common structural site (Lovell-Badge, 1992b). SRY, the other SOX proteins, and other members of the family such as TCF- 1, a T-cell receptor transcription factor that also belongs to this family, all show a high affinity for the same specific sequence in in vitro binding studies (Oosterwegel et al., 1991;van de Wetering et al., 1991;Ferrari er al., 1992), despite the fact that when all of these proteins are aligned there is not a single amino acid in the DNA binding domain that is universally conserved. It is not clear how specificity is built into this system in vivo. In vitro binding studies have usually been done with protein from the HMG box region produced in bacteria. It may be that regions of these proteins outside the conserved DNA binding domain lend specificity in vivo. In many cases members of this family are expressed in different cell types, but this is not always true. Of course, other cofactors that interact with these proteins or with the binding site may be involved in lending site specificity in vivo. It is not known how the binding affinity of SRY is affected for a recognition sequence lying within a structural motif. Perhaps the in vivo binding site bears elements of both structure and sequence, which in combination afford specificity. It has been suggested that the bending of DNA, which SRY either induces or recognizes when it binds, might bring two distant elements together or otherwise be involved in the assembly of the transcriptional complex, as has been shown for other family members (Giese er al., 1992; Kamachi et al., 1995). The effect of this interaction might be either to activate or repress a target locus (Gubbay and Lovel-Badge, 1994).
C. lnterspecies Comparisons
Transgenic mouse studies conducted in several laboratories using the human gene showed no sex reversal despite the fact that the human SRY transgene was shown to be transcribed in the genital ridge at 11.5 dpc (R. Behringer, P.
1. Role of Sry in Mammalian Sex Determination
23
Koopman, R. Lovell-Badge, R. Palmiter, personal communications). Given the high degree of homology between mouse and human genes within the DNA binding domain (Fig. 5 ) , this was a somewhat surprising result and led to several conclusions. The human gene appears to have been transcriptionally activated appropriately on a mouse XX background. This means that upstream regulatory pathways are functionally similar between mouse and humans. On the other hand, human SRY did not function downstream to trigger the male-determining pathway in mouse. This result could mean that some post-transcriptional step prevented the production of a functional human protein, or it could mean that differences in the amino acid sequence between the human and the mouse genes render human SRY dysfunctional in mouse. With respect to the second possibility, there are many differences between the mouse and the human proteins outside the DNA binding domain. The most striking difference is a long CAG repeat at the 3‘ end of the mouse gene that gives rise to the repeating histidine, glutamine domain that is completely absent in the human gene (also in the marsupial gene, Fig. 5). Any of these differences might be critical for interactions with other factors, or it is equally possible that one of the differences within the HMG box domain is significant for DNA binding interactions. Domain swap experiments are under way to produce transgenics carrying different regions of the genes from other species substituted for regions of the mouse gene. It is hoped that these experiments will help to define the elements of the gene that are critical to its function. In the process of sequencing and comparing Sry genes from different mammals, it became apparent that there are many differences outside the DNA binding domain between all species, suggesting a rapid evolution of sequences outside the conserved HMG box (see Fig. 5). In fact it was found that more of the changes in sequence lead to a change in amino acid than do not (Whitfield et al., 1993; Tucker and Lundrigan, 1993). A ratio of “nonsynonymous” to “synonymous’’ changes greater than 1 is thought to reflect positive selection for changes in the protein (Li et al., 1985). This result suggests that there might be selective pressure for a divergence of the SRY protein among species. Although this is an interesting idea to consider in terms of molecular mechanisms of speciation, the interpretation of these results is difficult. As a whole, the Y chromosome is thought to evolve rapidly (Tucker et al., 1989). The unique region of the Y chromosome has the unusual property of being unpaired in meiosis, and there is little basis for understanding how rates of evolutionary change might operate on genes that have no allelic partners. However, an overall comparison of mouse Zfy-Iwith human ZFY does not reveal a similar high rate of nonsynonymous substitution even though this gene is also located in the unique region of the Y (Whitfield et al., 1993). On the other hand, the possibility that regions outside the conserved HMG box in Sry evolve rapidly because they have little functional significance has not been ruled out. It has been pointed out that, among the sex determination systems that have
(in
120
Mm MI Mp
L L Q W D R N P H A
I T Y R Q D W S R A
. . . . . .
...
Mh
. . .
Ha SI
Re Sm Hs
. . . . . . . . . . T . .T
c . . . . . .
. . . . . . . . . . . . . . .M . . .T . . . . . N . L . T
- I K T Q H L K E D EV L . N R L Y R
. . . . . . . . . .
..T.. .P...E.A..
. . . . . . . A.T
SI
...
. -
-
. . S . . . . . . R S . . . . K.
-
. .
P E I - S F Y C A E H Q L O H L P P
1
Q 1
QQQQQQQQQQ
2w
190
180
170
Q F H D H H Q Q K Q
QFHDHHQQQQ
Q Q Q Q F Y D H H Q
A A A A A A A P A A V P A A A P A A P V R . . . Q Q Q Q Q H
AAAP...... .FH.HQ....
.............................. .............................. ..............................
F W - G L F T N E T
E T C R S L l H G Q
M D C N E C R S 1 9
M8
Ha
. .
. . . S . . . . . . S . . . . . . . . Q S - H -
150
GHLQQQQQQQ - - P . L R S - G. . . ..p .I . . . .H WAPTTAAAAA - I - - - - - - S H W V P T T T A A P A -1.- I.MN-. . . . - L L . W . . R . L . . . . . Yp E S T Y L - N W M N L P P E - E N T E N A A S S P Q QR D R Y S HWTKL @ . . . . . . .
Q Q Q Q F H N H H Q
MP Mh
1 0
F Y W Q P V D I P T
. . . . . . . . . L
. ..j..VRA.. . . , . f . . A . . S ~ T I Y E N T M K C D D C T K A T H S R M
160
Mm
IW A H L Y S K N Q Q S
ppp.......
Re
Sm
Ha
210
Mm Q F H D H H H H H Q MS MP Mh Ha SI
.......... ........................................ .......... .......... ........................................ ........................................
Re
Sm HS Mm MX
MP Mh Ha $1
2w
D H H H H Q Q Q Q Q
210
F H D H Q Q Q Q Q Q
280
F H D H Q Q Q Q H Q
.................... .......... .......... .......... .................... ....................
H D H P Q Q K Q 290 Q F H D H P Q Q Q Q 3Qw
................... ................... ...................
Re
Sm
HS 310
Mm F H D H H H Q Q Q Q MS MP Mh Ha SI
Re Sm HE
Mh MP
Ha 9
Re Sm Ha
334
K Q Q F H D H H Q Q
3 0
K Q Q F H D H H Q Q
350
QQQFHDHHQQ
.......... ........................................ .......... .......... ........................................ ........................................ 3m
Mm MS
320
K Q Q F H D H H Q Q
QQQQQQQQQQ
370
Q F H D Q Q L T Y L
.......... ..... .......... .......... ......
.EAN. ...*..AAN. P...
380
L T A D I T G E H T
390
P Y Q E H L S T A L
395
WLAVSP
. . " S @
..DS@
. . . . . p..L.
. Q
. . . . . K..
. @
25 been elucidated recently, there is no conservation at the molecular level of any of the specific players (Erickson and Cline, 1993; Hodgkin, 1990, 1992). This observation stands in sharp contrast to systems of embryonic axis pattern formation, for example, where the Hox genes seem to have been recruited to this purpose in such widely divergent systems as arthropods, cephalochordates, and vertebrates (Holland et al., 1992, 1994). This suggests that sex has evolved independently a number of times and must be one of the most rapidly evolving systems in the organism. 1. Role of Sry in Mammalian Sex Determination
D. The Hypothesis that Sry Expression I s in Balance with a Gene on the X Chromosome
Mammalian sex determination depends on the presence of a Y chromosome, unlike sex determination in Drosophila melanogaster or Caenorhabditis elegans, which depend on the ratio of X chromosomes to autosomes. In flies and worms this ratio determines dosage from the X chromosome and sex determination in all cells of the organism in a linked system. Many of the molecular details of this process are worked out in both organisms. In Drosophila both processes are mediated by the genes Sex-lethal ( S x l ) , doublesex, and transformer, and controlled at the post-transcriptional level (Baker, 1989; Mattox and Baker, 1991; Gorman et al., 1993). Female sex-specific splicing of Sxl occurs initially as a result of the X-to-autosome ratio. Only this splice variant is translated and able to autoregulate the continued productive splicing of its own transcript. The Sxl protein is able to initiate female development of the soma as well as female dosage compensation (Keyes et al., 1992). In the nematode, regulation appears to occur at both the transcriptional and the post-transcriptional levels. The X-toautosome ratio regulates the level of xol-I. The product of xol-I acts to regulate negatively expression of the sdc genes, which control both somatic sex determination and dosage compensation (Villeneuve and Meyer, 1990; DeLong et al.,
Fig. 5 Alignment of SRY sequences from 6 species of Old World mice and rat, marsupial, and human. The Mus musculus (Mm) Sry sequence is given from the initial methionine (1). The conserved DNA binding domain ends at 8 1, where Sry sequences rapidly diverge among species. Mus spretus (MS), Mus pahari ( M P ) , Mastomys hildebrantii (Mh),Hylomyscus alleni (Ha), Stochyomys lonicaudarus ( S l ) , R a m s exulans (Rr) (Tucker and Lundrigan, 1993), Sminthopsis macroura (Sm) (Foster et al., 1992), and Homo sapiens ( H s ) (Sinclair et al., 1990). Dashes indicate sequence identity. Dots indicate alignment gaps. Conserved amino acids are shaded, and in most cases are conserved among rabbit and primate Sry genes sequenced to date as well (Whitfield et a / . , 1993). Although no data exist for amino acid positions 144-395 for the species Ms,Mp, and Re by PCR analysis, the length of the repeat is similar to Mm in Ms and Mp, but shorter in Re. The repeat is absent in Sm, H s , and other primates. Mh, Ha, and S1 cannot be unambiguously aligned at the 3' end because of insertions and variability in the 3' repeating sequence. @ designates a stop codon.
26
Blanche Cape1
1993). Splicing variants of at least one gene in this pathway, tra-1, have been reported (Zarkower and Hodgkin, 1993). In mammals, the two processes of dosage compensation and sex determination appear to be unlinked. Dosage compensation must be accomplished in all cells of the organism, whereas sex determination occurs specifically in the gonadal cells of mammals (for review see Hodgkin, 1992). In mammals dosage compensation is controlled by inactivating one of the two X chromosomes in females. It seems likely that X-inactivation is controlled by the Xist locus, which has now been cloned in mouse (Brockdofi et af., 1992; Kay et al., 1993) and humans (Ballabio and Willard, 1992; Brown et al., 1992), and appears to act in cis to inactivate the chromosome from which it is expressed. The molecular basis for X-inactivation is still unknown. In rare cases of polyploid cells, more than one X can remain active, which suggests that dosage in mammals may also respond to the X-to-autosome ratio (Migeon et af., 1979; Jacobs and Migeon, 1989). Normally any gene located on the X can be kept at a constant single copy expression level by the overriding process of X-inactivation. In general the only way to get an elevated level of expression of a gene on the X chromosome is by releasing it from dosage control via duplication (or translocation to an autosome). It is possible that the presence of a Y chromosome could normally exert its dominant effect on sex determination by a tightly regulated expression of Sry sufficient to override the single-dose X signal (Fig. 6).
1. Role of Sry in Mammalian Sex Determination
27
There have been a number of suggestions that Sry might act as a repressor of genes critical to female development (McElreavey et al., 1993; Gubbay and Lovell-Badge, 1994). This sort of molecular action is consistent with the model of a genetic switch that acts to divert the ovarian “default” pathway. Cases of X-linked sex reversal have been reported in a number of mammals. In the domestic horse an X-linked recessive or autosomal sex-limited dominant gene has been postulated to account for heritable XY sex reversal (Kent el al., 1986). In wood lemmings a cytologically identifiable X* leads to female development in the presence of a normal Y chromosome (Fregda et al., 1976). Cases have also been reported in humans (Bernstein et al., 1980; Simpson, 1989; Ogata et al., 1992). Many of these cases map to a region of Xp that, when duplicated, can lead to male to female sex reversal. The identification of regions of the X that can act in a dominant way to cause sex reversal suggests one possible model whereby increasing the expression of some gene(s) on the X might put them out of a critical balance with the expression of Sry, ovemding the repressive effect of Sry on the female developmental pathway. This is an appealing explanation in that it links dosage compensation with sex determination in mammals and bears striking resemblance to molecular mechanisms that have been shown to operate in the fly and the nematode. One of the Sox genes, Sox-3, is an X-linked gene in mouse and humans (Stevanovic er al., 1993, and Collignon et al., in preparation). Of all the Sox genes isolated so far, it is the most closely related to Sry, bearing a 89.97% similarity to the Sry HMG box (Gubbay et al., 1990). Sox3 is expressed in the genital ridge in mouse as revealed by RNase protection studies (Collignon et al., in preparation; A . Hacker, personal communication). Given that Sry and all of the Sox genes recognize the same binding site in vitro, suspicions were aroused that some dose-dependent interaction exists between Sry and Sox-3, perhaps to compete for a binding site in vivo. However, two pieces of information from studies in humans dampen enthusiasm for the hypothesis that such an interaction might be important for sex determination. First, the region of the X chromosome that is involved in X-linked sex reversal is not close to the region where SOX-3 has been accurately mapped. Second, human male patients with a deletion of SOX-3 show mental retardation and early testicular failure, but the initial events of testis determination evidently proceeded normally (Rousseau et al., 1991). Females with larger deletions in this region of the human X show premature ovarian failure, but at present there is no good evidence to link this effect with SOX-3 (Stevanovic et al., 1993). A second X-linked gene, DAX 1, has been isolated from the dosage-sensitive region of Xp and shown to encode an orphan steroid receptor (Bardoni et al., 1994; Muscatelli et al., 1994; Zanaria et al., 1994). Although these data are tantalizing, other genes are present in this region of the X, and, as yet, there is no clear evidence that DAXl is the gene in Xp responsible for sex reversal. Experiments are under way to test the effect of an extra copy of this gene in transgenic mice.
28
Blanche Capel
Many other explanations are possible to explain cases of X-linked sex reversal. The most likely of these are mechanisms that involve the regulation of hormone and hormone receptor pathways downstream of Sry. We have seen that, given the global effects of hormone regulation on the organism, and the ability of many hormones to reverse sexual differentiation, it is dangerous to assign a primary role in sex determination to an effect that is defined at the genetic level. E. Other Possibilities for the Action of Sry
It is also possible that Sry acts in more than one way. For example, Sry may directly activate genes in the male pathway in addition to repressing genes in the female pathway. Familial cases of female to male sex reversal in XX individuals that are negative for Sry have been interpreted as evidence that Sry might act to repress an autosomal recessive gene that normally represses male development (McElreavey el af., 1993). Individuals homozygous for a defect in such a gene would be expected to initiate male development in the absence of Sty. Such interpretations of inherited patterns of sex determination have usually been based on ideas about how dominant, recessive, autosomal, or sex-linked genes would behave. Although the genetic information that has accumulated over the years is quite useful, it can often be interpreted in different ways. One problem is relating the language associated with a phenotypic analysis of classic genetic crosses to molecular biology. Concepts like “recessive,” “dominant,” “partial penetrance,” or “expressivity” do not have a simple meaning at the molecular level. Current research is concerned with isolating some of the downstream players in the pathway and illuminating some of the molecular detail to provide molecular mechanisms for the genetic data.
VI. Future Directions We are still a long way from understanding how Sry plays its sex-determining role. What are the molecular targets for Sry? How is Sry itself regulated? What is the significance of its association with DNA bending? What cofactors are involved and how is it specifically achieved at its site(s) of action? Is there a tight species restriction involved in the function of this gene? It is hoped that the study of this classic switch in development will lead to a better understanding of how genetic cascades are propagated and how they are integrated with external signals to control differentiation and morphogenesis. What are the initial cell structural changes that are triggered by Sry expression? How do these changes lead to the differentiation of the three somatic cell types of the testis? How do these changes effect the reorganization of the cells of the gonad into cords? This is a case of nature versus nurture at the cellular level. It is
1 . Role of Sry in Mammalian Sex Determination
29
clear that as a result of the “nature” or genetic content of XY cells, an intrinsic expression cascade that is normally required for differentiation along the male pathway is activated. It is also apparent, however, that the external cellular environment plays a critical role in this decision and can propagate or reverse the intracellular signal. As the testis differentiates, cells reorganize their cytoskeletal structure as they epithelialize and form tight junctions with their neighbors. Changes at the level of cellular structure have been shown to have profound effects on gene expression. In addition, cells reorganize into new communities in testis cords, and consequently come under the influence of new signals from their neighbors. It is important to understand how internal genetic cascades are interwoven with cell structural changes and changes in external signals that constitute the “nurture” component of the cell’s environment during the processes of morphogenesis in mammals. We hope that an understanding of Sry and its downstream mode of action will help to define and characterize key molecular players in the cell-cell and cellsubstrate interactions that drive organogenesis, and consequently to derive a better understanding of how genes direct normal morphogenesis of the embryo.
Acknowledgments I am grateful to Robin Lovell-Badge for countless discussions and helpful comments on this manuscript, to Peter Goodfellow and Paul Burgoyne for sharing a vast knowledge of the field, and to many of my colleagues who have contributed information and experimental results that are not yet published at the time of this writing.
References Affara, N. A , , Ferguson, S. M . , Tolrnie, J., Kwok, K., Mitchell, M . , Jamieson, D., Cooke, A,, and Florentin, L. (1986). Variable transfer of Y-specific sequences in XX males. Nucteic Acids Res. 14, 5375-5387. Armstrong, J., Pritchard-Jones, K., Bickmore, W., Hastie, N., and Bard, J. (1992). The expression of the Wilms’ tumour gene, W T I , in the developing mammalian embryo. Mechanisms Dev. 40, 85-97. Baker, B. S. (1989). Sex in flies: The splice of life. Nature (London) 340, 521-524. Ballabio, A,, and Willard, H. (1992). Mammalian X-chromosome inactivation and the XIST gene. Curr. Opin. Genet. Dev. 2, 439-447. Bardoni, B . , Zanaria, E., Guioli, S., Floridia, G., Worley, K. C . , Tonini, G . , Ferrante, E., Chiumello, G., McCabe, E. R. B., Fraccaro, M . , Zuffardi, O., and Camerino, G . (1994). A dosage sensitive locus at chromosome Xp21 is involved in male to female sex reversal. Nature Genet. 7 , 497-501. Barnes, J. D., Crosby, J. L., Jones, M., Wright, C. V. E., and Hogan B. L. M. (1994). Embryonic expression of LimI, the mouse homolog of Xenopus XLirn-1, suggests a role in lateral mesoderm differentiation and neurogenesis. Dev. B i d . 161, 168-178. Behringer, R. R., Cate, R. L., Froelick, C. I., Palmiter, R. D., and Brinster, R. L. (1990). Ab-
30
Blanche Capel
normal sexual development in transgenic mice chronically expressing Miillerian inhibiting substance. Nature (London) 345, 167-170. Bernstein, R., Koo, G., and Wachtel, S. (1980). Abnormality of the X chromosome in human 46, XY female siblings with dysgenetic ovaries. Science 207, 768-769. Berta, P., Hawkins, J. R., Sinclair, A. H., Taylor, A., Griffiths, B. L., Goodfellow, P. N., and Fellous, M. (1990). Genetic evidence equating SRY and the male sex determining gene. Nature (London) 348, 248-250. Bianchi, M. E., Falciola, L., Ferrari, S . , and Lilley, D. (1992). The DNA binding site of HMGl protein is composed of two similar segments (HMG boxes), both of which have counterparts in other eurkaryotic regulatory proteins. EMBO J 11, 1055-1063. Bridges, C. (1921). Triploid intersexes in Drosophilu melanogaster. Science 54, 252-254. Brockdorff, N., Ashworth, A., Kay, G . , McCabe, V., Nonis, D., Cooper, P., and Rastan, S. (1992). The product of the mouse Xist gene is a 15 kb inactive X-specific transcript containing no conserved ORF and located in the nucleus. Cell 71, 515-526. Brookman, J. I., Toosy, A. T., Shashidhara, L. S., and White, R. A. (1992). The 412 retrotransposon and the development of gonadal mesoderm in Drosophila. Development 116, 1185- 1 192. Brown, C., Hendrich, B . , Rupert, J., Lafreniere, R., Xing, V.,Lawrence, J., and Willard, H , (1992). The human XIST gene: Analysis of a 17kb inactive X-specific RNA that contains conserved repeats and is highly localized within the nucleus. Cell 71, 527-542. Buehr, M., Gu, S . , and McLaren, A. (1992). Mesonephric contribution to testis differentiation in the fetal mouse. Development 117, 273-281. Burgoyne, P. S . , Buehr, M., Koopman, P., Rossant, J., and McLaren, A. (1988). Cell autonomous action of the testis-determining gene: Sertoli cells are exclusively XY in XX tf XY chimaeric mouse testes. Development 102, 443-450. Burgoyne, P. S., Mahadevaiah, S. K., Sutcliffe, M. J., and Palmer, S. J. (1992). Fertility in mice requires X-Y pairing and a Y-chromosomal “spermiogenesis” gene mapping to the long arm. Cell 71, 391-398. Burgoyne, P. S . , and Palmer, S. J. (1991). The genetics of XY sex reversal in the mouse and other animals. Semin. Dev. Biol. 2, 277-284. Burgoyne, P., and Palmer, S . (1993). Cellular basis of sex determination and sex reversal in mammals. In “Gonadal Development and Function” (S. G . Hillier, Ed.), pp. 17-29. Raven Press, New York. Burgoyne, P., and Thornhill, A. (1993). The genetic basis of XX-XY differences present before gonadal sex differentiation in mice. In “Sex Chromosomes and Sex-Determining Genes” (K. C. Reed and J. A. M. Graves, Eds.), pp. 373-376. Harwood Academic Publishers. Chur, Switzerland. Capel, B . , and Lovell-Badge, R. (1993). The Sry gene and sex determination in mammals. Adv. Dev. Biol. 2 , 1-35. Capel, B., Raspberry, C., Dyson, J., Bishop, C., Simpson, E., Vivian, N., Lovell-Badge, R., Rastan, S . , and Cattanach, B. (1993a). Deletion of Y chrosomal sequences located outside the testis determining region can influence Sry expression and cause XY female sex reversal. Nature Genet. 5 , 301-307. Capel, B., Swain, A., Nicolis, S . , Hacker, A., Walter, M., Koopman, P., Goodfellow, P., and Lovell-Badge, R. (1993b). Circular transcripts of the testis-determining gene Sry in adult mouse testis. Cell (Cambridge, Mass.) 73, 1019-1030. Cate, R. L., Mattaliano, J., Hession, C., Tizard, R., Farber, N. M., Cheung, A , , Ninfa, E. G . , Frey, A. Z., Gash, D. J., Chow, E. P., et al. (1986). Isolation of the bovine and human genes for Miillerian inhibiting substance and expression of the human gene in animal cells. Cell (Cambridge, Mass.) 45, 685-698. Cattanach, B. M., Pollard, C. E., and Hawkes, S. G . (1971). Sex reversed mice: XX and XO males. Cytogenetics 10, 318-337.
1. Role of Sly in Mammalian Sex Determination
31
Collignon, J., and Lovell-Badge, R. (1995). Sry-related genes Sox-1 and Sox-2 and inductive interactions in the nervous system. Manuscript submitted. Conway, S . J., Mahadevaiah, S . K., Darling, S . M . , Capel, B., Rattigan, A. M., and Burgoyne, P. S . (1994). Y353/B: a candidate multiple-copy spermiogenesis gene on the mouse Y chromosome. Mammalian Genome 5, 203-210. Cumberledge, S., Szabad, J., and Sakonju, S. (1992). Gonad formulation and development requires the abd-A domain of the bithorax complex in Drosophila melungaster. Development 115, 395-402. DeLong, L., Plenefisch, J., Klein, R., and Meyer, B. (1993). Feedback control of sex determination by dosage compensation revealed through Caenorhabditis elegans scd-3 mutation. Genetics 133, 875-896. Delorenzi, M . , and Bienz, M. (1990). Expression of Abdominal-B homeoproteins in Drosophila melanogaster. Development 108, 323-329. Denny, P., Swift, S . , Brand, N., Dabhade, N., Barton, P., and Ashworth, A. (1992a). A conserved family of genes related to the testis determining gene, SRY. Nucleic Acids Res. 20, 2887. Denny, P., Swift, S., Connor, F., and Ashworth, A. (1992b). A testis specific gene related to SRY encodes a sequence-specific DNA binding protein. EMBO J . 10, 3705-3712. Dressler, G . R., Deutsch, U., Chowdhury, K., Nornes, H. O., and Gruss, P. (1990). Par-2, a new murine paired-box-containing gene and its expression in the developing excretory system. Development 109, 787-795. Dressler, G . R., and Douglass, E. C. (1992). Pax-2 is a DNA-binding protein expressed in embryonic kidney. Proc. Natl. Acad. Sci. USA 89, 1179-1183. Dym, M., Lamsam, C. S . , Jia, M. C . , Kleinman, H. K., and Papadopoulos, V. (1991). Basement membrane increases G-protein levels and follicle stimulating hormone responsiveness of Sertoli cell adenylyl cyclase activity. Endocrinology 128, 1167-1 176. Eicher, E. M. (1988). Autosomal genes involved in mammalian primary sex determination. Phil. Trans. R. Soc. London B. 322, 109-1 18. Eicher, E. M., and Washburn, L. L. (1986). Genetic control of primary sex determination in mice. Annu. Rev. Genet. 20, 327-360. Eicher, E. M., Washburn, L. L., Whitney, I. J., and Morrow, K. E. (1982). Mus poschiavinus Y chromosome in the C57BL16J murine genome causes sex reversal. Science 217, 535-537. Erickson, J., and Cline, T. (1993). A bZIP protein, Sisterless-a, collaborates with bHLH transcription factors early in Drosophila development to determine sex. Genes Dev. 7, 1688-1702. Evans, E. P.. Burtenshaw, M. D., and Cattanach, B . M . (1982). Meiotic crossing over between the X and Y chromosomes of male mice carrying the sex-reversing (Sxr) factor. Nature (London) 300, 443-445. Ferrari, S., Harley, V. R., Pontiggia, A,, Goodfellow, P. N., Lovell, B. R., and Bianchi, M. E. (1992). SRY, like HMGl, recognizes sharp angles in DNA. EMBO J. 11, 4497-4506. Ford, C. E.. Jones, K. W., Polani, P. E., de, A. I., and Briggs, J. H. (1959). A sexchromosome anomaly in a case of gonadal dysgenesis (Turner’s syndrome). Lancer 1, 711713. Foster, I., Brennan, F., Hampikian, G . , Goodfellow, P., Sinclair, A,, Lovell-Badge, R., Selwood, L., Renfree, M . , Cooper, D., and Graves, J. (1992). Evolution of sex determination and the Y chromosome: SRY related sequences in marsupials. Nature (London) 359, 531-533. Foster, J. W., Dominguez-Steglich, M. A,, Guioli, S . , Kwok, C., Weller, P. A,, Stevanovic, M., Weissenbach, J., Mansour, S . . Young, I . D., Goodfellow, P. N., Brook, J. D., and Schafer, A. 3. (1994). Campomelic dysplasia and autosomal sex reversal caused by mutations in an SRY-related gene. Nature (London) 372, 525-529. Fregda, K . , Gropp, A , , Winking, H., and Frank, F. (1976). Fertile XX- and XY-type females in the wood lemming, Myopus schisticolor. Nature (London) 261, 225-227. Gaunt, S . J.. Coletta, P. L., Pravtcheva, D., and Sharpe, P. T.(1990). Mouse HOX-3.4:homeo-
32
Blanche Capel
box sequence and embryonic expression patterns compared with other members of the Hox gene network. Development 109, 329-339. Giese, K.,Amsterdam, A., and Grosschedl, R. (1991). DNA-binding properties of the HMG domain of the lymphoid-specific transcriptional regulator LEF-l . Genes Dev. 5 , 2567-2578. Giese, K., Cox, J., and Grosschedl, R. (1992). The HMG domain of lymphoid enhancer factor 1 bends DNA and facilitates assembly of functional nucleoprotein structures. Cell 69, 1-20, Ginsburg, M . , Snow, M., and McLaren, A. (1990). Primordial germ cells in the mouse embryo during gastrulation. Development (Cambridge, U K ) 110, 521-528. Goodfellow, P. N., and Darling, S. M. (1988) Genetics of sex determination in man and mouse. Development (Cambridge, UK) 102, 25 1-258. Gorman, M . , Kuroda, M., and Baker, B . (1993). Regulation of the sex-specific binding of the maleness dosage compensation protein to the male X chromosome in Drosophila. Cell 72, 39-49. Grobstein, C. (1956). Induction of tubules in mouse metanephrogenic mesenchyme. Exp. Cell Res. 10, 424-440. Gubbay, J . , Collignon, J., Koopman, P., Capel, B., Economou, A,, Munsterberg, A., Vivian, N., Goodfellow, P., and Lovell-Badge, R. (1990). A gene mapping to the sex-determining region of the mouse Y chromosome is a member of a novel family of embryonically expressed genes. Nature (London) 346, 245-250. Gubbay, J . , and Lovell-Badge, R. (1994). The mouse Y chromosome. In “Molecular Genetics of Sex Determination,” pp. 43-67. Academic Press, London. Gubbay, J., Vivian, N., Economou, A., Jackson, D., Goodfellow, P., and Lovell-Badge, R. (1992). Inverted repeat structure of the Sry locus in mice. Proc. Nutl. Acad. Sci. USA 89, 7953-7957. Hacker, A,, Capel, B., Goodfellow, P., and Lovell-Badge, R. (1995). Expression of S r y , the mouse sex determining gene. Development 121, 1603-1614. Haqq, C. M . , King, C-Y., Ukiyama, E., Falsafi, S . , Haqq, T. N., Donahoe, P. K., and Weiss, M. A. (1994). Molecular basis of mammalian sexual determination: Activation of Mullerian inhibiting substance gene expression by SRY. Science 266, 1494- 1500. Harley, V. R., Jackson, D. I., Hextal, P. J., Hawkins, J. R., Berkovitz, G. D., Sockanathan, S . , Lovell, B. R., and Goodfellow, P. N. (1992). DNA binding activity of recombinant SRY from normal males and XY females. Science 225,453-456. Hawkins, J. R. Genetics of XY sex reversal. J . Endocrinol., in press. Hawkins, J. R., Taylor, A , , Berta, P. Levilliers, J., Van, D. A. B., and Goodfellow, P. N. (1992a). Mutational analysis of SRY: Nonsense and missence mutations in XY sex reversal. Hum. Genet. 88, 471-474. Hawkins, J. R., Taylor, A,, Goodfellow, P. N., Migeon, C. J., Smith, K. D., and Berkovitz, G . D. (1992b). Evidence for increased prevalence of SRY mutations in XY females with complete rather than partial gonadal dysgenesis. Amer. J. Hum. Genet. 51, 979-984. Herzlinger, D., Abramson, R., and Cohen, D. (1993). Phenotypic conversions in renal development. J. Cell Sci. 17, 61-64. Hodgkin, J. ( 1990). Sex determination compared in Drosophila and Caenorhabditis. Nature (London) 344, 721-728. Hodgkin, J. (1992). Genetic sex determination, mechanisms and evolution. Bioessays 14, 253-261. Holland, P., Garcia-Femandez, J., Shaman, A,, and Williams, N. (1994). The origins of vertebrate development. In “BSBD Evolution of Development,” Abstract 37. Holland, P., Holland, L., William, N.. and Holland, N. (1992). An Amphioxus homeobox gene, sequence conservation, spatial expression during development and insights into vertebrate evolution. Development 116, 653-661. Ikeda, Y., Shen, W.-H., Ingraham, H., and Parker, K. (1994). Developmental expression of mouse steroidogenic factor 1, an essential regulator of the steroid hydroxylases. Mol. Endocrinol. 8, 664-662. Izpisua-Belmonte, J. C., Dolle, P., Renucci, A., Zappavigna, V., Falkenstein, H., and Duboule,
1. Role of Sry in Mammalian Sex Determination
33
D. (1990). Primary structure and embryonic expression pattern of the mouse Hox-4.3 homeobox gene. Development (Cambridge, U K ) 110, 733-745. Izpisua-Belmonte, J. C., Falkenstein, H., Dolle, P., Renucci, A , , and Duboule, D. (1991). Murine genes related to the Drosophila AbdB homeotic gene are sequentially expressed during development of the posterior part of the body. EMEO J. 10, 2279-2289. Jacobs, P., and Migeon, B. (1989). Studies of X-chromosome inactivation in trisomies. Cytogenet. Cell. Genet. 50, 75-77. Jacobs, P. A., and Ross, A. (1966). Structural abnormalities of the Y chromosome in man. Nature (London) 210, 352-354. Jager, R. J.. Pfeiffer, R. A , , and Scherer, G. (1991). An amino acid substitution in SRY leads to conditional XY gonadal dysgenesis. Int. Congr. Hum. Genet. 8, 12-13. Jeske, Y. W. A , , Bowles, J., Greenfield, A,, and Koopman, P. (1995). Expression of a linear SRY transcript in the mouse genital ridge. Nature Genet. 10, 480-482. Jiminez, R., Burgos, M., Sinchez, A,, Sinclair, A. H., Alarch, F. J., Marin, J. J., Ortega, E., and Diaz de la Guardia, R. (1993). Fertile females of the mole Talpa occidentalis are phenotypic intersexes with ovotestes. Development 118, 1303- 13I 1. Josso, N., Boussin, L., Knebelmann, B., Nihoulfekete, C., and Picard, J. (1991). Anti-Mullerian hormone and intersex states. Trends Endocrinol. Metab. 2, 227-232. Jost, A. (1947). Recherches sur la diffkrenciation sexuelle de I’embryon de lapin. Arch. Anat. Microsc. Morphol. Exp. 36, 271-315. Jost, A., Perchellet, J., Prepin, J., and Vigier, B. (1975). The prenatal development of bovine freemartins. In “Symposium on Intersexuality” (R. Rheinboth, Ed.), pp. 392-406. Springer Verlag, Berlin. Kamachi, V., Suckanathan, S., Liu, Q., Breitman, M., Lovell-Badge, R., and Kondoh, H. (1995). Involvement of SOX proteins in lens-specific activation of crystallin genes. EMEO J . 14, 3510-3519. Kanai, Y., Kurohmaru, M . , Hayashi, Y., and Nishida. T. (1989). Formation of male and female sex cords in gonadal development of C57B116 mouse. Japan. J. Vet. Sci. 51, 7-16. Karch, F., Bender, W., and Weiffenback, B. (1990). abd-A expression in Drosophila embryos. Genes Dev. 4, 1573-1587. Karl, J., and Capel, B. (1995). Three-dimensional structure of the developing mouse genital ridge. Phil. Trans. R . SOC.London E , in press. Kay, G., Penny, G., Patel, D., Ashworth, A., BrockdorlT, N . , and Rastan, S. (1993). Expression of Xist during mouse development suggests a role in the initiation of X chromosome inactivation. Cell 72, 171-182. Kent, M. G., Schoffner, R . N., Buoen, L., and Weber, A . F. (1986). XY sex-reversal syndrome in the domestic horse. Cytogenet. Cell Genet. 42, 8-18. Keyes, L., Cline, T., and Schedl, P. (1992). The primary sex determination signal of Drosophila acts at the level of transcription. Cell 68, 933-943. Koopman, P., Gubbay, J., Vivian, N., Goodfellow, P., and Lovell-Badge, R. (1991). Male development of chromosomally female mice transgenic for Sly. Nature (London) 351, 117-121. Koopman, P., Munsterberg, A,, Capel, B., Vivian, N . , and Lovell-Badge, R. (1990). Expression of a candidate sex-determining gene during mouse testis differentiation. Nature (London) 348, 450-452. Li, W-H, Wu, C-I, and Lui C-C (1985). A new method for estimating synonymous and nonsynonymous rates of nucleotide substitution considering the relative likelihood of nucleotide and codon changes. Mol. Eiol. Evol. 2, 150-174. Lilley, D. (1992). The HMGl domain has DNA wrapped up. Nature (London) 357, 282-283. Lovell-Badge, R. (l992a). Sex-determining gene expression during embryogenesis. Phil. Trans. R . Soc. London E 335, 159-164. Lovell-Badge, R . (1992b). Testis determination: soft talk and kinky sex. Curr. Opinion Genet. Development 2, 596-601.
34
Blanche Cape1
Lovell-Badge, R., and Robertson, E. (1990). XY female mice resulting from a heritable mutation in the murine primary testis determining gene, Tdy. Development 109, 635-646. Luo, X., Ikeda, Y., and Parker, K. L. (1994). A cell-specific nuclear receptor is essential for adrenal and gonadal development and sexual differentiation. Cell 77, 48 1-490. Mattox, W., and Baker, B. (1991). Autoregulation of the splicing of transcripts from the transformer-2 gene of Drosophila. Genes Dev. 5 , 786-796. McElreavy, K., Vilain, E., Abbas, N., Costa, J. M., Souleyreau, N., Kucheria, K., Boucekkine, C., Thibaud, E., Brauner, R., Flamant. F., et al. (1992). XY sex reversal associated with a deletion 5’ to the SRY “HMG box” in the testis-determining region. Proc. Natl. Acad. Sci. USA 89, 11016-1 1020. McElreavey, K., Vilain, E., Abbas, N., Herskowitz, I., and Fellous, M. (1993). A regulatory cascade hypothesis for mammalian sex determination: SRY represses a negative regulator of male development. Proc. Natl. Acad. Sci. USA 90, 3368-3372. McLaren, A. (1 984). Chimeras and sexual differentiation. “Chimeras in Developmental Biology” (N. Le Douarin and A., McLaren, Eds.), pp. 381-399. Academic Press, New York/London. McLaren, A. (1985). Relating of germ cell sex to gonadal differentiation. I n “The Origin and Evolution of Sex” (H. 0. Halverson and A. Monroy, Eds.), pp. 289-300. Academic Press, New York. McLaren, A. (1988). Sex determination in mammals. Trends Genet. 4, 153-157. McLaren, A. (1991). Development of the mammalian gonad: The fate of the supporting cell lineage. Bioessays 13, 151-156. McLaren, A,, and Monk, M. (1982). Fertile females produced by inactivation from X chromosome of ‘sex-reversed’ mice. Nature (London) 300, 446-468. McLaren, A., Simpson, E., Tomonari, K., Chandler, P., and Hogg, H. (1984). Male sexual differentiation in mice lacking H-Y antigen. Nature (London) 312, 552-555. Merchant-Larios, H. (1979). Origin of the somatic cells in the rat gonad: An autoradiographic approach. Ann. Eiol. Anim. Eiochim. Biophys. 19, 1219-1229. Merchant-Larios, H., Moreno-Mendoza, N., and Buehr, M. (1993). The role of the mesonephros in cell differentiation of morphogenesis of the mouse fetal testis. fnt. J. Dev. Biol. 37, 407415. Merchant-Larios, H., and Taketo, T. (199 I). Testicular differentiation in mammals under normal and experimental conditions. J. Electron Microsc. Technique 19, 158-171. Migeon, B., Sprenkle, J., and Do, T. (1979). Stability of the “Two Active X ’ phenotype in triploid somatic cells. Cell 18, 637-641. Mintz, B., and Russell, E. S. (1957). Gene-induced embryological modifications of primordial germ cells in the mouse. J. Exp. Zool. 134, 207-237. Mitchell, M. J., and Bishop, C. E. (1992). A structural analysis of the Sxr region of the mouse Y chromsome. Genomics 12, 26-34. Mittwoch, U . (1986). Males, females and hermaphrodites. Ann. Hum. Genet. 50, 103-121. Moreno-Mendoza, N., Herrera-Munoz, J., and Merchant-Larios, H. (1995). Limb bud mesenchyme permits seminiferous cord formation in the mouse fetal testis but subsequent testosterone output is markedly affected by the sex of the donor stromal tissue. Dev. Eiol. 169, 51-56. Muller, U., Donlon, T., Schmid, M., Fitch, N., Richer, C.-L., er al. (1986). Deletion mapping of the testis determining locus with DNA probes in 46, XX males and in 46, XY and 46, X,dic(Y) females. NAR 14, 6489-6505. Miinsterberg, A,, and Lovell-Badge, R. (1991). Expression of the mouse anti-Mullerian hormone gene suggests a role in both male and female sexual differentiation. Development 113, 613624. Muscatelli, F., Strom, T. M., Walker, A. P., Zanaria, E., Recan, D., Meindl, A,, Bardoni, B., Guioli, S., Zehetner, G., Rabl, W., Schwarz, H. P., Kaplan, J-C., Camerino, G., Meitinger,
1 . Role of Sry in Mammalian Sex Determination
35
T., and Monaco, A. P. (1994). Mutations in the DAX-I gene give rise to both X-linked adrenal hypoplasia congenita and hypogonadotropic hypogonadism. Nature (London) 372, 672-676. Ogata, T., Hawkins, J., Taylor, A., Matsuo, N., Hata, J-I, and Goodfellow, P. (1992). Sex reversal in a child with a 46,X,Yp+ karytype: Support for the existence of a gene(s), located in distal Xp, involved in testis formation. J . Med. Genet. 29, 226-230. Oostenvegel, M. A., Holstege, F., Prosser, H. M., Owen, M. J., and Clevers, H. C. (1991). TCF- 1, a T cell-specific transcription factor of the HMG box family, interacts with sequence motifs in the TCR-beta and TCR-delta enhancers. Int. lmmunol. 3, 1189-1 192. Page, D. C., Mosher, R., Simpson, E. M . , Fisher, E., Mardon, G., Pollack, J., McGillivray, B., de la Chapelle, A., and Brown, L. G. (1987). The sex-determining region of the human Y chromosome encodes a finger protein. Cell 51, 1091-1094. Palmer, S . , and Burgoyne, P. S. (1991a). In situ analysis of fetal, prepuberal and adult XX tj XY chimaeric mouse testes: Sertoli cells are predominantly, but not exclusively, XY. Development 112, 265-268. Palmer, S., and Burgoyne, P. S. (1991b). XY follicle cells in the ovaries of XO/XY and XO/XY/XYY mosaic mice. Development 111, 1017-1020. Palmer, S . J . , and Burgoyne, P. S . (1991~).The Mus musculus domesticus Tdy allele acts later than the Mus musculus musculus Tdy allele: A basis for XY sex reversal in C57BL/6-YPos mice. Development 113, 709-7 14. Pelletier, J . , Bruening, W., Li, F. P., Haber, D. A,, Glasser, T., and Housman, D. E. (1991a). WTI mutations contribute to abnormal genital system development and hereditary Wilms’ tumour. Nature (London) 353, 43 1-433. Pelletier, J., Schalling, M . , Buckler, A. J., Rogers, A., Haber, D. A., and Housman, D. (1991b). Expression of the Wilms’ tumour gene WTl in the murine urogenital system. Genes Dev. 5 , 1345-1356. Phelps, D. E., and Dressler, G. R. (1993). Aberrant expression of Pax-2 in Danforth’s short tail ( S d ) mice. Dev. Biol. 157, 251-258. Philipsen, S . , Pruzina, S . , and Grosveld, F. (1993). The minimal requirements for activity in transgenic mice of hypersensitive site 3 of the beta globin locus control region. EMBO J. 12, 1077- 1085. Picard, J. Y., Benarous, R . , Guerrier, D., Josso, N., and Kahn, A. (1986). Cloning and expression of cDNA for anti-Mullerian hormone. Proc. Natl. Acad. Sci. LISA 83, 5464-5468. Rodemer-Lenz, I. (1989). On cell contribution to gonadal soma formation in quail-chick chimeras during the indifferent stage of gonadal development. Anat. Embryol. 179, 237-242. Rousseau, F., Vincent, A . , Rivella, S . , Heitz, D., Triboli, C., Maestrini, E., Warren, S . , Suthers, G . , Goodfellow, P., Mandel, J . , Toniolo, D., and Oberle, I. (1991). Four chromosomal breakpoints and four new probes mark out 10-cM region encompassing the fragile-)< (FRAXA) locus. Amer. J . Hum. Genet. 48, 108-1 16. Satoh, M. (1985). The histogenesis of the gonad in rat embryos. J . Anat. 143, 17-37. Shawlot, W., and Behringer, R. (1995). Requirement for Liml in head-organizer function. Nature (London) 374, 425-430. Simpson, J. (1989). Genetic heterogeneity in XY sex reversal: potential pitfalls in isolating the testis determining factor (TDF). In “Evolutionary Mechanisms in Sex Determinants” ( S . S . Wachtel, Ed.), pp. 265-277. CRC Press, Boca Raton, FL. Sinclair, A. H., Berta, P., Palmer, M. S . , Hawkins, I. R., Griffiths, B. L., Smith, M. J . , Foster, J. W., Frischauf, A. M . , Lovell-Badge, R., and Goodfellow, P. N. (1990). A gene from the human sex-determining region encodes a protein with homology to a conserved DNA-binding motif. Nature (London) 346, 240-244. Skinner, M. K., Tung, P. S., and Fritz, I . B. (1985). Cooperativity between Sertoli cells and testicular pertibular cells in the production and deposition of extracellular matrix components. J . Cell. Biol. 100, 1941-1947.
36
Blanche Cape1
Smith, C., and MacKay, S . (1991). Morphological development and fate of the mouse mesonephros. J. Anut. 174, 171-184. Stevanovic, M., Lovell, B. R., Collignon, J., and Goodfellow, P. (1993). SOX3 is an X-linked gene related to Sry, Hum. Mol. Genet. 2, 2013-2018. Strueli, C., Bailey, N., and Bissell, M. (1991). Control of mammary epithelial differentiation: Basement membrane induces tissue-specific gene expression in the absence of cell-cell interaction and morphological polarity. J . Cell Biol. 115, 1383-1395. Strueli, C., Schmidhauser, C., Kobrin, M., Bissell, M., and Derynck, R. (1993). Extracellular matrix regulates expression of the TGF-PI gene. J . Cell Biol. 120, 253-260. Suarez-Quian, C., Hadley, M . , and Dym, M. (1985). Effect of substrate on the shape of Sertoli cells in vitro. Ann. N.Y. Acad. Sci. 438, 413-434. Taketo. T., and Koide, S . S . (1981). In vitro development of testis and ovary from indifferent fetal mouse gonads. Dev. Biol. 84, 61-66. Taketo, H. T., and Merchant-Larios, H. (1985). Testicular cell differentiation in fetal mouse ovaries following transplantation into adult male mice. J . Exp. Zool. 236, 229-237. Thornhill, A,, and Burgoyne, P. (1993). A paternally imprinted X chromosome retards the development of the early mouse embryo. Development 118, 173-174. Tsunoda, Y., Tokunaga, T., and Sugie, T. (1984). Altered sex-ratio of live young after transfer of fast-developing and slow-developing mouse embryos. Guntete Res. 12, 301-304. Tucker, P. K., Lee, B. K., and Eicher, E. M. (1989). Y chromosome evolution in the subgenus Mus (genus Mus).Genetics 122, 169-179. Thcker, P., and Lundrigan, B. (1993). Rapid evolution of the sex determining locus in Old World mice and rats. Nature (London) 364, 715-717. Tung, P. S . , and Fritz, 1. B. (1980). Interactions of sertoli cells with myoid cells in vitro. Biol. Reprod. 23, 207-217. Tung, P. S . , and Fritz, I. B. (1986). Extracellular matnx components and testicular peritubular cells influence the rate and pattern of Sertoli cell migration in vitro. Dev. Biol. 113, 119-134. van de Wetering, M., Oosterwegel, M., Dooijes, D., and Clevers, H. (1991). Identification and cloning of TCF- I , a T lymphocyte-specific transcription factor containing a sequence-specific HMG box. EMBO J . 10, 123-132. Vigier, B., Watrin, F., Magre, S., Tran, D., and Josso, N. (1987). Purified bovine AMH induces a characteristic freemartin effect in fetal rat prospective ovaries exposed to it in virro. Deuelopment 100, 43-55. Villeneuve, A., and Meyer, B. (1990). The regulatory hierarchy controlling sex determination and dosage compensation in Caenorhubditis elegans. Adv. Genet. 27, 117-188. Wagner, T., Wirth, J., Meyer, J., Zabel, B., Held, M . , Zimmer, J., Pasantes, J., Bricarelli, F. D., Keutel, I., Hustert, E., Wolf, U . , Tommerup, N., Schempp, W., and Scherer, G. (1994). Autosomal sex reversal and campomelic dysplasia are caused by mutations in and around the SRY-related gene SOX-9. Cell 79, 1 1 11-1 120. Wartenberg, H. (1982). Development of the early human ovary and the role of the mesonephros in the differentiation of the cortex. Anar. Embryol. 165, 253-280. Wartenberg, H., Kinsky, I., Viebahn, C., and Schmolke, C. (1991). Fine structural characteristics of testicular cord formation in the developing rabbit gonad. J. Electron Microsc. Technol. 19, 133- 157. Wartiovaara, J., Nordling, S., Lehtonen, E., and Saxen, L. (1974). Transfilter induction of kidney tubules: Correlation with cytoplasmic penetration into nucleopore filters. J . Embryol. Exp. Morphol. 31, 667-682. Washburn, L., and Eicher, E. (1989). Normal testis determination in the mouse depends on genetic interaction of a locus on chromosome 17 and the Y chromosome. Genetics 123, 173179.
1. Role of Sry in Mammalian Sex Determination
37
Welshons, W. J., and Russell, L. B. (1959). The Y chromosome as the bearer of male determining factors in the mouse. Proc. Natl. Acad. Sci. USA 45, 560-566. White, M. (1973). “Animal Cytology and Evolution.” Cambridge Univ. Press, Cambridge. Whitfield, S . , Lovell-Badge, R., and Goodfellow, P. N. (1993). Rapid sequence evolution of the sex determining gene SRY. Nature (London) 364, 713-715. Wilson, E. (1909). Quoted In “The Cell in Heredity and Development” (1925). 3rd edition. Macmillan, New York. Wright, E., Hargrave, M. R . , Christiansen, J., Cooper, L., Kum, J., Evans, T., Gangadharan, U., Greenfield, A., and Koopman, P. (1995). The Sty-related gene Sox9 is expressed during chrondrogenesis in mouse embryos. Nature Genet. 9, 15-20. Zanaria, E., Muscatelli, F., Bardoni, B., Strom, T. M., Guioli, S . , Guo, W., Lalli, E., Moser, C., Walker, A. P., McCabe, E. R. B., Meitinger, T., Monaco, A. P., Sassone-Corsi, P., and Camenno, G. (1994). An unusual member of the nuclear hormone receptor superfamily responsible for X-linked adrenal hypoplasia congenita. Nature (London) 372, 635-641 Zarkower, D., and Hodgkin, J. (1993). Zinc fingers in sex determination: Only one of the two C.elegans Tra-I proteins binds DNA in vitro. Nucleic Acids Rex 21, 3691-3698. Zwingman, T., Erickson, R . , Boyer, T., and Ao, A. (1993). Transcription of the sex-determining region genes Sry and’2 in the mouse preimplantation embryo. Proc. Narl. Acad. Sci. USA 90, 814-817.
This Page Intentionally Left Blank
7 Molecular Mechanisms of Gamete Recognition in Sea Urchin Fertilization Kay Ohlendieck Department of Pharmacology University College Dublin Belfield, Dublin, Ireland William 1. Lennarz Department of Biochemistry and Cell Biology State University of New York at Stony Brook Stony Brook, New York 11794-5215
I. 11. 111. IV.
Introduction Multistep Recognition Process in Fertilization Chernoattraction and Activation of Sperm Gamete Interactions at the Egg Plasma Membrane A. Role of the Sperm Ligand Bindin in Gamete Adhesion B. Molecular Profile of the Egg Receptor for Sperm C. Role of the Sperm Receptor Carbohydrate D. Developmental Expression of the Sperm Receptor E. Fate of the Sperm Receptor Following Fertilization V. Mechanisms to Prevent Polyspermy VI. Egg Activation in Sea Urchins VII. Prospects References
1. Introduction In free-spawning marine organisms, the successful fusion of gametes during fertilization is secured by a series of consecutive intercellular recognition steps. Sea urchins are common marine organisms that can be held in captivity under simple conditions and produce gametes for periods ranging from weeks to many months, depending on the species. For these reasons, the sea urchin is the most highly studied model system for research on fertilization. Mature gametes can be obtained in large quantities and eggs can be fertilized in vitro under well-defined conditions in artificial sea water. Since intracoelomic KCl injection may release lo7 eggs or 10'2 spermatozoa during a single spawning (Trimmer and Vaquier, 1986), sea urchin gametes are an ideal system to characterize biochemically complementary cell recognition molecules in fertilization. Current Toprcs tn Developmental Biology, Vol. 32 Copynght 0 1996 by Academic Press, Inc. All rights of reproduction in any form reserved
39
40
Kay Ohlendieck and William J. Lennarz
This review summarizes our current understanding of the molecular mechanisms involved in the echelon of species-specific recognition steps between sea urchin gametes. Since our laboratory is actively involved in the elucidation of the structure/function relationship in integral sperm receptors from echinoderms, we concentrate in this review on the implications of this novel class of cell recognition molecules for the overall fertilization process. Currently, published information on the primary structure and the subunit composition of the biologically active sperm receptor exists only for the receptor from Strongylocentrotus purpurarus, a sea urchin species present on the west coast of North America, which will be the main focus of this article. Other species-specific recognition steps in sea urchin fertilization, and aspects of egg activation, will be discussed in less detail; the reader is directed in individual sections to recent excellent reviews on these aspects of fertilization.
II. Multistep Recognition Process in Fertilization In contrast to other cell-cell recognition processes, fertilization does not represent a single interaction but comprises a series of complex cellular events. The flow chart in Fig. 1 lists these events in the multistep fertilization process in sea urchins. The chances for successful fertilization depend on the synchronization of gamete release from female and male sea urchins, which may depend on the length of daylight, water temperature, and water current. Spawning of marine invertebrates might also be directly coupled to phytoplankton blooms (Starr et al., 1990). Following spawning, peptides released from eggs cause kinetic and directional changes in sperm motility (Hardy et al., 1994). Chemotactic attraction significantly increases the chances for direct contact between spermatozoa and the egg jelly coat, which in turn induces the acrosomal reaction (Trimmer and Vaquier, 1986). During the formation of the acrosomal process, motile sperm penetrate the jelly coat. When activated sperm interact directly with the egg plasma membrane, binding preferentially occurs with protrusions of the surface membrane known as the microvilli. It is now well established that this species-specific binding process is mediated by complementary cell surface molecules, the abundant sperm protein bindin (Vaquier and Moy, 1977) and its respective egg surface receptor, a multimeric glycoprotein complex (Foltz and Lennarz, 1993; Lennarz, 1994). Sperm binding to the egg plasma membrane triggers a membrane potential change, rapid and synchronized exocytosis of cortical granules, release of intracellular calcium ions, increase in intracellular pH, and oxygen consumption, as well as complex metabolic changes. Fusion of gamete nuclei is then followed by the initiation of DNA synthesis and the first cell division. In the following sections, cell-cell or cell-ligand interactions that culminate in gamete fusion are discussed in somewhat more detail.
2. Sea Urchin Fertilization: Gamete Recognition
41
Chemotactic Attraction of Sperm to Egg Interaction w'ith Jelly Coat and Induction of Acrosomal Reaction in Sperm Sperm Receptor / Bindin Interaction at Egg Plasma Membrane Gamete Membrane Fusion
t t
Cortical Vesicle Exocytosis
Initiation of Egg Activation Cascade
Fusion of Genetic Material
Mitosis / Cell Cleavage
Early Developmental Pathway Fig. I Flow chart of the species-specific recognition steps during sea urchin fertilization that eventually lead to initiation of the early developmental pathway.
111. Chemoattraction and Activation of Sperm The diffusible, egg jelly-associated peptides speract and resact, which are released from S. purpuratus and Arbacia punctulata, respectively, initiate many specific responses in sperm, which include stimulation of sperm respiration and motility, as well as chemoattraction (as reviewed by Hardy et al., 1994). Spermactivating peptides bind to receptors localized to the sperm plasma membrane; the activation of these receptors appears to induce the increased synthesis of cyclic GMP and changes in protein phosphorylation that influence kinetic and directional changes in sperm motility. The speract receptor (from S. purpuratus sperm) is a membrane protein of 77 kDa, comprising a small intracellular tail, a single transmembrane domain, and a cysteine-rich extracellular domain (Bentley et al., 1988; Dangott et al., 1989). In contrast, in the case of the resact receptor (from A . punctulata sperm), the receptor is a membrane-bound guanylyl cyclase of 160 kDa (Suzuki et al., 1984; Shimomura et al., 1986). The reason the two
42
Kay Ohlendieck and William J. Lennarz
identified sperm-activating peptides crosslink to such different receptor molecules in two sea urchin species remains to be determined. In addition, the molecular mechanism of the signaling process of sperm-activating factors is not well understood. Recently, it was shown using fluorescein-conjugated speract that speract receptors are enriched on sea urchin sperm flagella. This finding has led to the suggestion that the receptor plays a direct regulatory role in flagellar motility (Cardullo et al., 1994), which appears to be controlled via cytosolic calcium fluxes (Cook et al., 1994). Once motile sperm cells reach the jelly coat of the egg surface via chemotaxis, induction of the acrosomal reaction results in the fusion of the acrosomal vesicle with the sperm surface membrane, which triggers the exocytosis of its vesicular contents. This exocytotic event exposes the abundant sperm head protein bindin to the exterior surface of the tip of the sperm (Trimmer and Vaquier, 1986; Ward and Kopf, 1993). The molecule(s) responsible for the initiation of the acrosomal reaction is believed to be associated with the egg jelly coat, which comprises fucose sulfate polymers, peptides, and glycoproteins (Suzuki, 1990). In S. purpurarus these molecules appear to bind to a 210-kDa sperm glycoprotein that is involved in calcium influxes necessary for the acrosomal reaction (Trimmer et al., 1986; Trimmer and Vaquier, 1986). Earlier studies identified a fucose sulfate glycoconjugate as the acrosome reaction-inducing egg component (SeGall and Lennarz, 1981). However, recent findings by Keller and Vaquier (1994) suggest the involvement of two glycoproteins of 82 and 138 kDa. Bonnell et al. (1994) argue that it is possible that earlier studies on the fractionation of egg jelly utilized conditions where fucose sulfate polymers and the acrosome reaction-inducing glycoproteins were still bound to each other, forming a biologically active complex similar to that seen in native egg jelly. This issue will only be resolved by more refined experiments using isolated egg jelly components to determine precisely what molecular components of the jelly coat are responsible for the induction of the acrosomal reaction in sea urchin sperm.
IV. Gamete Interactions at the Egg Plasma Membrane Following chemoattraction and increase in sperm motility by diffusible peptides, and the induction of the acrosomal reaction by jelly coat components, a major species-specific recognition step in sea urchin fertilization occurs at the egg plasma membrane. In the sections below we discuss current information on the molecular properties of the complementary gamete surface molecules responsible for this interaction. In addition, we will review recent findings on the developmental expression of the sperm receptor during oogenesis and the fate of this novel class of cell recognition molecules following fertilization.
2. Sea Urchin Fertilization: Gamete Recognition
43
A. Role of the Sperm Ligand Bindin in Gamete Adhesion
Our current understanding of how gamete recognition is mediated at the egg plasma membrane is based on the identification of the abundant sperm protein bindin and its complementary receptor, a homo-multimeric, integral glycoprotein complex on the egg surface. Bindin, a 30.5-kDa protein that is stored in the acrosomal vesicle, becomes exposed on the acrosomal process upon sperm activation (Moy and Vaquier, 1979). Although originally identified 18 years ago by Vaquier and Moy (1977), the physical properties of bindin are still not well understood, and it is not known how this ligand is presented on the sperm plasma membrane following the acrosomal reaction. Molecular cloning and sequencing revealed that bindin molecules from various sea urchin species (Gao et al., 1986; Minor et al., 1989) exhibit homology in a highly conserved central domain flanked by more divergent N- and C-terminal domains (reviewed by Hofmann and Glabe, 1994). Bindin is not a glycoprotein, nor does it exhibit any transmembrane domains. Although the molecular process of how bindin molecules interact with the sperm receptor is not known, the participation of the conserved central domain of bindin is implicated in membrane fusion and sulfate fucan binding (Kennedy et al., 1989; DeAngelis and Glabe, 1987). Despite the uncertainties stated above, about the molecular properties of bindin, several experimental findings clearly demonstrate that this sperm protein is at least one of the ligands involved in gamete adhesion at the egg plasma membrane. Particulate bindin aggregates eggs species-preferentially (Glabe and Lennarz, 1979; Glabe et al., 1981), and, most importantly, it exhibits specific binding to an extracellular 70-kDa fragment of the sperm receptor (Foltz and Lennarz, 1990). In addition, bindin binding was shown to occur to a recombinant protein representing the extracellular sperm binding domain (Foltz et al., 1993), and immunocytochemical techniques clearly localized bindin to the region of interaction of the two gamete surfaces (Moy and Vaquier, 1979). Thus, bindin is indeed one of the ligands, or the sole ligand, for the integral egg receptor for sperm in sea urchin fertilization. The question whether other sperm and/or egg surface molecules are involved in additional adhesion processes or downstream events in egg activation following sperm binding is currently under investigation. The fact that bindin agglutinates unfertilized eggs, but does not activate them (Glabe et al., 1981), would argue that other molecules participate in the actual fusion of gamete membranes and perhaps in the initiation of the egg activation cascade. However, since isolated bindin molecules could have been altered during purification, the simple scenario that the ligand-receptor interaction between bindin and its complementary glycoprotein complex is involved in gamete binding, membrane fusion, and egg activation cannot be excluded.
44
Kay Ohlendieck and William J. Lennarz
B. Molecular Profile of the Egg Receptor for Sperm 1. Identification of the Sea Urchin Sperm Receptor The purification and detailed analysis of the sea urchin egg receptor for sperm lagged behind the identification and purification of its putative ligand, bindin. Early studies identified and partially characterized a membrane-associated, highmolecular-weight glycoconjugate as a candidate for the sperm receptor (Schmell et al., 1977; Rossignol et al., 1981). The possible involvement of the carbohydrate chains versus the polypeptide backbone in the adhesion process deduced from these and similar studies is discussed in detail in the next section. Although the preliminary documentation of a membrane-associated sperm receptor was useful in the elucidation of some of the properties of the putative receptor, these studies were unsuccessful in yielding a homogeneous protein (Kinsey and Lennarz, 1981; Rossignol et d., 1984). The problems in these earlier studies, based on the lack of a suitable probe or antibody for the identification of the sperm receptor during purification, were overcome by an alternative approach. Proteolytic fragments of the extracellular domain of the sperm receptor were generated and characterized using established fertilization bioassays (Ruiz-Bravo and Lennarz, 1986, 1989). Following the initial characterization of heterogeneous tryptic fragments of the receptor, use of lysylendoproteinase C to cleave proteins from the egg cell surface was found to release a large fragment of 70 kDa (Foltz and Lennarz, 1990). The purified 70-kDa fragment inhibited fertilization speciesspecifically, and was shown to bind to acrosome-reacted sperm and isolated bindin particles, thereby demonstrating that this structurally defined fragment was indeed derived from the extracellular domain of the sperm receptor. A polyclonal antibody generated against the receptor fragment reacted with a single protein, of apparent 350 kDa, using Western blot analysis, and was shown by immunofluorescence microscopy to be distributed evenly over the entire egg surface (Foltz and Lennarz, 1992). 2. Molecular Characterization of the Sperm Receptor Molecular cloning and sequencing of the sperm receptor using antireceptor IgG to screen an expression library made from immature sea urchin ovary mRNAs revealed further insights into the structural features of the receptor molecule (Folz et al., 1993). The sea urchin egg receptor not only represents a novel class of cell recognition molecules, but is also the first identified member of a new species of integral sperm receptors. The earlier-discovered glycoprotein ZP-3, which was found to be responsible for induction of the acrosomal reaction and sperm binding in mammalian fertilization, is not related to the integral sea urchin sperm receptor, and represents a sperm binding protein in the zona pellucida that clearly lacks a transmembrane domain (reviewed by Wassarman, 1993). In con-
2. Sea Urchin Fertilization: Gamete Recognition
45
trast to the situation in mammals, in sea urchins, these two functions are divided between different egg surface molecules. The acrosome reaction is induced by components of the jelly coat, whereas sperm binding occurs at the plasma membrane via an integral receptor complex. The sea urchin sperm receptor exhibits a short cytoplasmic tail, a single transmembrane domain, and a large extracellular domain that exhibits similarity to portions of the hsp 70 family of proteins (Foltz ef al., 1993). As can be seen in Fig. 2, the sperm receptor also contains a cysteine-rich domain at the extreme N-terminus, which overlaps the hsp 70-like domain. the receptor exhibits numerous potential sites for 0- and N-glycosylation. That this molecule is indeed the egg receptor for sperm was demonstrated by the fact that a recombinant protein representing a portion of the extracellular receptor domain was species-specific both in binding to either acrosome-reacted sperm or bindin particles and in inhibiting fertilization (Foltz et a f . , 1993). Initial Southern blot analysis using genomic DNA isolated from related sea urchin species indicated that the sperm receptor contains species-specific and
Fig. 2 Model of the primary structure of the sea urchin egg receptor for sperm. The recent biochemical characterization of the intact sperm receptor suggests that the biologically active form of the receptor is represented by a homo-tetrameric complex of 350-kDa subunits. No other major egg surface components have been detected in the sperm binding process.
46
Kay Ohlendieck and William J. Lennarz
conserved domains. Little crosshybridization between species was observed with the cDNA corresponding to the extracellular sperm binding domain. In contrast, cDNA fragments representing the more C-terminal portion of the sperm receptor did exhibit crosshybridization with genomic DNA samples from other sea urchin species, suggesting that the cytoplasmic domain might be a conserved region of the molecule (Foltz et al., 1993). These findings are consistent with the idea that the intracellular tail of the receptor is responsible for a signal transduction event downstream of sperm adhesion, and that the extracellular N-terminal domain of the receptor contains a highly variable sperm binding domain responsible for a species-specific gamete interaction. However, a recent and more detailed Southern blot analysis of receptor domains using genomic DNA isolated from various sea urchin species and higher organisms suggests that both the intracellular and the extracellular receptor domains are relatively conserved between certain sea urchin species. The results in Fig. 3 illustrate that both cDNA probe 45A to the extracellular receptor domain and cDNA probe 7.1- 1C to the intracellular receptor domain cross-hybridize,
Flg. 3 Southern blot analysis of the sea urchin egg receptor for sperm. Genomic DNA was isolated, electrophoresed, and blotted onto nylon membranes by standard molecular biological methods (Moore, 1992). Hybridization using radioactively labeled cDNA probes 7.1-IC (a) and 45A (b [Foltz et al., 19931) was performed under stringent conditions (0. I X SSC, 0. I ?h SDS, 65°C). Under these conditions, cDNA probes to both the cytoplasmic and the extracellular domain of the Srrongylocentrotus purpurutus receptor (c) hybridized to related sea urchin species, Srrongylocenrrorus drobachiensis and Lyrechinus variegatus. With lower-stringency conditions (1 X SSC, 0.1% SDS, 50"C), more convincing hybridization was observed to genomic DNA from Lyrechinus pictus and starfish, but not to DNA from higher organisms such as fly, frog, rabbit, or human (K. Ohlendieck and W. J. Lennarz, unpublished results).
2. Sea Urchin Fertilization: Gamete Recognition
47
under relatively stringent conditions, to genomic DNA from S. purpuratus, S. drobachiensis, and Lytechinus variegatus.Only a low level of crosshybridization was found with genomic DNA purified from L. pictus and starfish. In the case of fly, frog, rabbit, or human tissues, even under lower-stringency conditions (see legend to Fig. 3), no hybridization was observed. The fact that the extracellular domain of the sperm receptor appears to be sufficiently homologus between certain sea urchin species to allow crosshybridization with the 45A cDNA probe suggests that only a relatively small region of the N-terminal domain is highly variable. The sperm binding domain(s) responsible for the species-specific binding of its ligand bindin and/or other sperm proteins appear(s) therefore to be restricted to a small stretch of amino acids. Alternatively, the sperm binding domain(s) may not be contained in a contiguous stretch of amino acids but may be due to tertiary structures not readily interpreted in terms of the primary sequence of the receptor. Experiments currently in progress (R. Stears and W. J. Lennarz, unpublished studies) using a series of deletion constructs should answer these questions and define the minimum sequence needed for receptor-ligand binding if the primary structure defines the binding site.
3. Subunit Composition of the Sperm Receptor Since detailed structure/function studies require knowledge of the mature, intact molecule, purification of the highly glycosylated sperm receptor to homogeneity was undertaken and accomplished using lectin and ion exchange chromatography (Ohlendieck et al., 1993). Following isolation, the intact receptor was found to retain biological activity. Thus, homogeneous receptor preparations inhibited sea urchin fertilization in a species-specific and dose-dependent manner. However, these experiments did not establish that the purified receptor exhibits adhesive properties. To show directly sperm binding to the intact receptor, various homogeneous preparations of the native or recombinant receptor were immobilized on polystyrene microspheres and tested for their ability to bind to sperm. It was found that these coated microbeads bound species-specifically to acrosomereacted sperm (Foltz et al., 1993; Ohlendieck et al., 1993), thus directly demonstrating that the receptor has adhesive properties. Since many cell surface molecules, such as the insulin receptor or the EGF receptor, exist in their biologically active form as oligomeric complexes (Ullrich and Schlessinger, 1990), analysis of the subunit composition of the native sperm receptor was of interest. Analytical SDS-PAGE analysis under reducing and nonreducing conditions in combination with crosslinking experiments revealed that the native sperm receptor exists as a homo-multimeric complex, estimated to be a tetramer (Ohlendieck et al., 1994a; Ohlendieck and Lennarz, 1995). Two lines of evidence demonstrate that the homotetrameric complex is the biologically active configuration of the sea urchin sperm receptor. First, whereas the nonreduced, native receptor complex inhibits fertilization in a dose-dependent
48
Kay Ohlendieck and William J. Lennarz
manner, the reduced and alkylated receptor is inactive. Second, microspheres coated with the native receptor complex bind to acrosome-reacted sperm, whereas beads with reduced and alkylated receptors do not interact with sperm (Ohlendieck et al., 1994a). That the receptor is a disulfide-bonded multimeric complex agrees well with the fact that the N-terminus of the extracellular receptor domain is rich in cysteine residues (Foltz et al., 1993). Crosslinking studies utilizing agents of various length and solubility failed to reveal the association of other major egg surface proteins with the sperm receptor. Clearly, this negative result does not exclude other proteins that interact with the receptor. However, the simplest interpretation is that the native sperm receptor exists in its biologically active form as a multimer of 350-kDa subunits that are linked via disulfide bonds to produce an integral surface membrane complex that is functional in binding acrosome-reacted sperm.
C. Role of the Sperm Receptor Carbohydrate
Early studies on sea urchin fertilization, as described above, implicated a highmolecular-weight glycoconjugate as an egg receptor for sperm (Schmell et al., 1977). Characterization of a partially purified receptor preparation showed that the receptor bound to sperm and inhibited fertilization in a species-specific manner (Rossignol e? al., 1981). Pronase digestion of this receptor preparation resulted in a carbohydrate-rich fragment that bound to acrosome-reacted sperm. Since this binding was no longer species-specific, it was concluded that the carbohydrate chains are the adhesive elements of the sperm receptor, and that the polypeptide backbone refers species-specificity to the overall binding process (Rossignol et al., 1984; Ruiz-Bravo and Lennarz, 1986, 1989). However, following the cloning and sequencing of the sperm receptor, it was demonstrated that a recombinant protein representing the sperm binding domain was speciesspecific both in binding to acrosome-reacted sperm and in inhibiting fertilization (Foltz et al., 1993). This new finding was not consistent with the earlier hypothesis that the receptor polypeptide is responsible for the species-specificityin sperm adhesion, whereas the oligosaccharide chains are the adhesive element in gamete binding. It now seems that both the polypeptide backbone and the carbohydrate chains of the receptor are involved in the binding process. Studies by Dhume and Lennarz (1 995) established that a subfraction of the total receptor oligosaccharide chains exhibits a low-affinity ionic interaction with sperm. Carbohydrate chains isolated from homogeneous sperm receptor preparations were subfractionated using lectin and ion exchange chromatography. Competitive bioassays, combined with sperm binding studies, revealed that the most highly charged, sulfated, 0-linked carbohydrate chains bind to sperm and exhibit the highest inhibitory activity in sea urchin fertilization. However, the bioactive oligosac-
49
2. Sea Urchin Fertilization: Gamete Recognition
charide lacked species-specificity, unlike the intact receptor and the recombinant aglyco protein representing the extracellular receptor domain (Dhume and Lennarz, 1995). Based on these findings, and the information accumulated about the species-specific binding of sperm to the intact receptor (reviewed in Ohlendieck and Lennarz, 1995), the model shown in Fig. 4 can be proposed. In the first step, an ionic interaction occurs between the sulfated 0-linked oligosaccharide chains of the receptor and ligands on the acrosomal process of the sperm. In the second step, this low-affinity ionic interaction, which is not species-specific, is followed by a high-affinity, species-specific interaction of the sperm ligand with one or more binding sites on the polypeptide backbone of the receptor (Dhume and Lennarz, 1995).
-
SDerrn Eaa ReceDtor Interactions:
SDerm - Eaa ReCeRtOr Interactions:
- Between negatively charged 0-linked
- Between polypeptide chain and
oligosaccharides and sperm ligands.
- Low affinity - ionic interactions.
sperm ligands.
- High affinity - species-specific interactions.
Fig. 4 Two-step model of the interactions between sperm and the egg plasma membrane receptor for sperm. See the text for a detailed description of the proposed dual interaction of sperm ligands with the 0-linked polysaccharide chains and the polypeptide binding dornain(s) of the integral sperm receptor.
50
Kay Ohlendieck and William J. Lennarz
D. Developmental Expression of the Sperm Receptor
To prepare for proper development following fertilization, stores of maternal proteins and mRNAs, as well as metabolites and specialized organelles, are assembled in the developing oocyte (Brandhorst, 1985). Extensive information on the major morphological changes during sea urchin oogenesis is available (reviewed by Giudice, 1973). With respect to the biochemical changes occurring during oocyte maturation, it is known that extracellular-matrix-containing vesicles and the endoplasmic reticulum become differentially positioned within the developing oocyte (Alliegro and McClay, 1988; Henson er al., 1990), and it is well established that changes occur in the activity of ion channels during plasma membrane restructuring (Dale and DeSantis, 1981). In contrast, remarkably little is known about the developmental expression of cell adhesion molecules during oocyte maturation, although it is clear these molecules are essential for cell-cell interactions during fertilization and development. We recently examined maturing ovaries morphologically and with respect to content of receptor mRNA, as well as to the content and subcellular distribution of the receptor glycoprotein (Ohlendieck et al., 1994b). Although in early oocyte stages neither mRNA encoding for the receptor nor receptor glycoprotein was detectable, in the last two stages of development the level of mRNA and glycoprotein accumulation increased dramatically. Sperm receptor was first detected in the oocyte by immunocytochemistry at the time of appearance of two welldefined organelles, cortical granules, and yolk platelets. Receptor synthesis correlated temporally and spatially with the formation of cortical granules. Goldtagged receptor was only detected within the oocyte and could not be found elsewhere in the ovary. It therefore appears that the sea urchin egg receptor for sperm, unlike the yolk protein (Shyu er al., 19861, is synthesized in the oocyte. Previous studies on the subcellular localization of the sperm receptor by RuizBravo et al. (1989), and our more recent developmental analysis of receptor expression during oocyte maturation, suggest that two distinct receptor populations exist, one in cortical granules and a second at the cell surface, that may be formed via secretory vesicles. In the last stage of development, receptor is equally distributed between cortical granules and the egg surface, where it is localized predominantly to the microvilli (Ohlendieck et al., 1994b). The finding that the receptor is only expressed on the egg surface late in oogenesis is in good agreement with the biological function of the receptor, since functional surface molecule-mediated egg-sperm interactions occur only between mature gametes.
E. Fate of the Sperm Receptor Following Fertilization
Binding of ligands such as growth factors or peptide hormones to surface receptors on their specific target cells usually results in receptor-mediated endocytosis
2 . Sea Urchin Fertilization: Gamete Recognition
51
and, subsequently, either degradation or recycling of the receptor to the plasma membrane (Mangelsdorf-Sonderquist and Carpenter, 1986). Receptors, like the multi-subunit molecule responsible for insulin binding, are dynamic macromolecules that move through various cellular compartments (Knutson, 199I). Very recent studies on the sperm receptor showed that this novel cell recognition molecule does not become internalized and recycled, but undergoes rapid degradation upon sperm binding. In addition, a soluble isoform of the receptor appears to be exocytosed into the perivitelline space during the cortical reaction (K. Ohlendieck, J. Partin, and W. J. Lennarz, unpublished results). Thirty seconds after sperm binding, the total amount of receptor in cortical granules and the plasma membrane decreases to 30% of its value. Immunoblot analysis of a receptor-enriched glycoprotein fraction isolated from fertilized eggs at different time points revealed that only 3% of the sperm receptor remains membranebound after 30 s. The apparent discrepancy between these two values can be explained by the fact that these two techniques measure a different subpopulation of the receptor. Since soluble receptor isoforms lacking the transmembrane domain would no longer be present in the membrane preparation, the immunoblotting technique exclusively determines the fraction of the receptor still membranebound. In contrast, immunogold labeling recognizes the total amount of receptor remaining following fertilization, i.e., the membrane-bound receptor fraction as well as the soluble receptor isoform. In either case, both techniques clearly established that the disappearance of the receptor is very rapid. Degradation of the sperm receptor protein could also be induced in the absence of sperm binding by incubation of unfertilized eggs with the calcium ionophore A23187. Thus, receptor degradation appears to result from a calcium-induced proteolytic process initiated in the egg itself, rather than a protease introduced by the sperm. It was previously shown by Carroll and Epel (1975) and Lois el al. (1986) that a sperm hydroiase exists in sea urchin eggs. However, it is not known whether this protease is responsible for the degradation of the sperm receptor following fertilization. In Fig. 5, the fate of the sperm receptor following fertilization is illustrated diagrammatically. After successful sperm adhesion to the egg surface, a specific protease appears to become activated and is secreted. Presumably, prior to fertilization this enzyme exists in an inactive form, probably stored in the cortical granules of the unfertilized egg. A fraction of the receptor in cortical granules is deposited into the perivitelline space, and the membrane-associated receptor becomes rapidly degraded. Thus, unlike other cell surface receptors, the sea urchin egg receptor for sperm does not recycle upon ligand binding but is rapidly destroyed following sperm adhesion. These findings agree with a single role postulated for this high-molecular-weight cell adhesion molecule (Ohlendieck and Lennarz, 1995). If the receptor served another adhesion function in other events in early development of the embryo, such as adhesion processes involved in the restructuring of the zygote surface in preparation for cell division, a rapid
Kay Ohlendieck and William J. Lennarz
52 MouffsflROBad Egg
ff@GU~OUZ~ Egg
ortical Granules Cortical Granule
(LEGEND
Plasma Membrane Sperm Receptor Cortical Granule Sperm Receptor
Fig. 5 Diagrammatic representation of the fate of the sperm receptor following fertilization. Recent findings on the dynamics and fate of the sea urchin egg receptor for sperm indicate that this cell recognition molecule is rapidly degraded following sperm adhesion. In addition, immunoelectron microscopical investigations on the sperm receptor in eggs following fertilization demonstrated the transient presence of a secreted isoform of the sperm receptor. See the text for a detailed discussion of the proteolytic degradation of the sperm receptor following gamete binding.
degradation of the membrane-bound receptor would not be expected. In fact, one could even imagine an increase in receptor density in certain plasma membrane domains if the surface receptor had a broader role as an adhesion molecule and/or anchor for membrane cytoskeletal or extracellular components involved in development. Thus, the sperm receptor appears to have a highly specialized role in gamete adhesion and does not function in developmental processes downstream of egg-sperm binding. However, the receptor in the cortical granule that is deposited into the perivitelline space might serve as an intermediate block to polyspermy (see next section).
V. Mechanisms to Prevent Polyspermy Polyspermy, the entry of excess sperm into an egg following successful fertilization by a sperm, results in abnormal development and ultimately in embryo death. To eliminate fusion of additional sperm cells and thereby avoid disastrous
2. Sea Urchin Fertilization: Gamete Recognition
53
genetic consequences, the egg plasma membrane immediately becomes depolarized following fertilization (Jaffe, 1976). However, this fast-acting electrical block to polyspermy (1-3 seconds following sperm adhesion) only transiently prevents further entry of sperm (Longo et al., 1986). At about 1 min, the late block to polyspermy, the hardening of the fertilization envelope (Larabell and Chandler, 1991), sets in, and thereby prevents any further interaction between gametes (see Shen, 1995, for review). In addition to these two mechanisms to prevent polyspermy, we have preliminary evidence that modulation in the availability of the sperm receptor might act as an intermediate block to polyspermy. The studies discussed above on the fate of the sperm receptor following fertilization revealed that the receptor becomes rapidly degraded following sperm adhesion, and at the same time a receptor isoform, stored in the cortical granules in the unfertilized egg, is secreted by exocytosis into the perivitelline space. The
Fig. 6 Possible mechanisms to prevent polyspermy in sea urchin fertilization. The flow diagram summarizes the molecular mechanisms that could serve as fast, intermediate, and slow blocks to polyspermy. Depolarization of the egg plasma membrane is a highly effective and quick, but only transient, block to polyspermy which lasts about 1 min. We propose that the rapid degradation of the sperm receptor following fertilization, and the secretion of an isoform of the receptor into the perivitelline space, might act as an intermediate block to polyspermy. This block might bridge the time until the slow block to polyspenny, the hardening of the fertilization envelope, sets in and prevents any further gamete interactions.
54
Kay Ohlendieck and William J. Lennarz
consecutive action of these different mechanisms to prevent polyspermy is illustrated in the flow chart of Fig. 6. Following membrane depolarization and the cortical reaction, excess sperm could bind to the secreted receptor isoform and thereby prevent additional sperm from reaching the egg plasma membrane. In addition, the discharge of excess sperm at the egg surface by membrane depolarization could be complemented by the rapid decrease in sperm receptor molecules.
VI. Egg Activation in Sea Urchins Although many of the molecular components in egg activation are identified, the exact mechanism by which sperm binding induces signal transduction and triggers egg activation is still not well understood. In contrast, it is now well established that, downstream from the sperm-egg adhesion event, redundant systems of Ca2+ mobilization are activated in the fertilized egg. It was demonstrated by Galione et al. (1993) that both inositol triphosphate and ryanodine receptor channels contribute to redundant mechanisms of CaZ+-induced Ca2+ release that are responsible for the Ca2+ waves during sea urchin fertilization. Furthermore, studies by Lee et al. (1993) established that in addition to inositol triphosphate, a novel secondary messenger, cyclic adenosine diphosphate-ribose, participates in mobilizing Ca*+ in sea urchin fertilization. Galione, in a recent review on calcium signaling (1994), discusses the different lines of evidence that point to the role of cyclic adenosine diphosphate-ribose as an endogenous activator of Caz+-induced Ca2+ release by nonskeletal muscle ryanodine receptors. However, 15 s pass following successful sperm binding before the transient increase in intracellular calcium concentration is observed in sea urchin eggs in the form of a propagating wave. The molecular events in these critical first seconds following sperm adhesion are still unknown, and it is not clear whether a G-protein-linked receptor molecule is involved in a signal transduction event or whether, possibly, a diffusible activating factor is introduced by the sperm into the egg, triggering the egg activation cascade. A review by Whitaker and Swann (1993) summarizes the possible molecular scenarios involved in the initiation and propagation of the calcium wave that initiates the early developmental pathway. With respect to the egg receptor for sperm, this molecule might be directly or indirectly involved in this process. It is possible that the sperm receptor is a species-specific anchor for bindin and induces gamete membrane fusion, and this might indirectly trigger the introduction of a second messenger into the egg cytoplasm. In a more complex scenario, the sperm receptor not only would be responsible for sperm binding and gamete membrane fusion, but also would be involved directly in a receptor-mediated signal transduction event. It is well established that, upon sperm adhesion, a rapid change in phosphorylation occurs on tyrosine residues of a variety of egg proteins, and that the phosphoinositide
2. Sea Urchin Fertilization: Gamete Recognition
55
messenger system is activated (Ciapa and Epel, 1991; Ciapa ef al., 1992). Recent studies by Moore and Kinsey (1994) identified a tyrosine kinase that is active at fertilization. Since the egg receptor for sperm becomes phosphorylated on tyrosine residues at fertilization (Abassi and Foltz, 1994; Foltz, 1994), it is quite possible that phosphorylation plays an important role in early developmental events downstream from egg-sperm binding.
VII. Prospects With respect to the molecular mechanisms that mediate the species-specific interaction between gamete surface membranes in sea urchin fertilization, we now have a good understanding of the structure of several of the complementary molecules involved in this cell-cell adhesion process. The exact molecular mechanisms that underlie the consecutive adhesion events, how these processes are coordinated, and what combination of egg/sperm components is responsible for the initiation of the egg activation cascade remain to be determined.
Acknowledgments Research from the authors’ laboratory was supported by Grant HD 18590 to W.J.L. from the National Institutes of Health. We thank the members of our laboratory for many helpful discussions on the molecular mechanisms of gamete recognition in fertilization. We thank L. Conroy for preparation of the manuscript.
References Abassi, Y. A , , and Foltz, K. R. (1994). Tyrosine phosphorylation of the sperm receptor at fertilization. Dev. Eiol. 164, 430-443. Alliegro, M. C., and McClay, D. R. (1988). Storage and mobilization of extracellular matrix proteins during sea urchin development. Dev. Eiof. 125, 208-216. Bentley, J. K., Khatra, A. S . , and Garbers. D. L. (1988). Receptor-mediated activation of detergent-solubilized guanylate cyclase. B i d . Reprod. 39, 639-647. Bonnell, B. S . , Keller, S. H., Vaquier, V. D., and Chandler, D. E. (1994). The sea urchin egg jelly coat consists of globular glycoproteins bound to a fibrous fucan superstructure. Dev. B i d . 162, 313-324. Brandhorst, B. P. (1985). Information content of the echinoderm egg. In “Developmental Biology. A Comprehensive Synthesis (L. W. Browder, ed.), Vol. I , pp. 525-576. Plenum, New York . Cardullo, R. A , , Hemck, S . B., Peterson, M . J . , and Dangott, L. J. (1994). Speract receptors are localized on sea urchin sperm flagella using a fluorescent peptide analogue. Dev. B i d . 162, 600-607. Carroll, E. J., and Epel, D. (1975). Isolation and biochemical activity of the protease released by sea urchin eggs following fertilization. Dev. B i d . 44, 222-232.
56
Kay Ohlendieck and William J. Lennarz
Ciapa, B., and Epel, D. (1991). A rapid change in phosphorylation on tyrosine accompanies fertilization of sea urchin eggs. FEES Lett. 295, 167-170. Ciapa, B., Borg, B., and Whitaker. M. J. (1992). Phosphoinositide metabolism during the fertilization wave in sea urchin eggs. Developmenr (Cambridge, U K ) 115, 187-195. Cook, S. P., Brokaw, C. J., Muller, C. H., and Babcock, D. F. (1994). Sperm chemotaxis: Egg peptides control cytosolic calcium to regulate flagellar response. Dev. Eiol. 165, 10- 19. Dale, B., and DeSantis, A. (1981). Maturation and fertilization of the sea urchin oocyte: An electrophysiological study. Dev. Eiot. 85, 474-484. Dangott, L. J., Jordan, J. E., Bellet, R. A , , and Garbers, D. L. (1989). Cloning of the mRNA for the protein that crosslinks to the egg peptide speract. Proc. Natl. Acad. Sci. U.S.A. 86, 2128-2132. DeAngelis, P. L., and Glabe, C. G. (1987). Polysaccharide structural features that are critical for the binding of sulfated fucans to bindin, the adhesive protein from sea urchin sperm. J. Eiol. Chem. 262, 13946-13952. Dhume, S. T., and Lennarz, W. J. (1995). The involvement of 0-linked oligosaccharide chains of the sea urchin egg receptor for sperm in fertilization. Glycobiology 5, 11-17. Foltz, K. F. (1994). The sea urchin egg receptor for sperm. Semin. Dev. Eiol. 5, 243-253. Foltz, K. R., and Lennarz, W. J. (1990). Purification and characterization of an extracellular fragment of the sea urchin egg receptor for sperm. J. Cell Eiol. 111, 2951-2959. Foltz, K. R., and Lennarz, W. J. (1992). Identification of the sea urchin egg receptor for sperm using an antiserum raised against a fragment of its extracellular domain. J . Cell Eiol. 116, 647-658. Foltz, K. R., and Lennarz, W. J. (1993). The molecular basis of gamete interactions at the egg plasma membrane. Dev. Biol. 158, 46-61. Foltz, K. R., Partin, J. S . , and Lennarz, W. J. (1993). Sea urchin egg receptor for sperm: Sequence similarity of binding domain and hsp70. Science 259, 1421-1425. Galione, A. (1994). Cyclic ADP-ribose, the ADP-ribosyl cyclase pathway and calcium signalling. Mol. Cell. Endocrinol. 98, 125-131. Galione, A,, McDougall, A., Busa, W. B., Willmott. N., Gillot, I . , and Whitaker, M. (1993). Redundant mechanisms of calcium-induced calcium release underlying calcium waves during fertilization of sea urchin eggs. Science 261, 348-352. Gao, B., Klein, L. E., Britten, R . J., and Davidson, E. H. (1986). Sequence of mRNA coding for bindin, a species-specific sea urchin sperm protein required for fertilization. Proc. Natl. Acad. Sci. U.S.A. 83, 8634-8638. Giudice, G. (1973). Oogenesis. In “Developmental Biology of the Sea Urchin Embryo” (G. Giudice, ed.), pp. 43-54. Academic Press, New York. Glabe, C. G., and Lennarz, W. J. (1979). Species-specific sperm adhesion in sea urchins: A quantitative investigation of bindin-mediated egg agglutination. J. Cell Biol. 83, 595-604. Glabe, C. G., Buchalter, M., and Lennarz, W. J. (1981). Studies on the interactions of sperm with the surface of the sea urchin egg. Dev. Biot. 84, 397-406. Hardy, D. M., Harurni, T., and Garbers, D. L. (1994). Sea urchin receptors for egg peptides. Semin. Dev. Eiol. 5, 217-224. Henson, J. H., Beaulien, S. M., Kaminer, B., and Begg, D. A. (1990). Differentiation of a calsequestrin-containing endoplasmic reticulum during sea urchin oogenesis. Dev. Eiol. 142, 255-269. Hofmann, A., and Glabe, C. (1994). Bindin, a multifunctional sperm ligand, and the evolution of new species. Semin. Dev. Eiol. 5, 233-242. Jaffe, L. A. (1976). Fast block to polyspermy in sea urchin eggs is electrically mediated. Narure (London) 261, 68-71. Keller, S . H., and Vaquier, V. D. (1994). The isolation of acrosome-reaction-inducing glycoproteins from sea urchin egg jelly. Dev. Eiol. 162, 304-312.
2. Sea Urchin Fertilization: Gamete Recognition
57
Kennedy, L., DeAngelis, P. L., and Glabe, C. G. (1989). Analysis of the membrane-interacting domain of the sea urchin adhesive protein bindin. Biochemistry 28, 9153-9158. Kinsey, W. H., and Lennarz, W. J. (1981). Isolation of a glycopeptide fraction from the surface of the sea urchin egg that inhibits sperm-egg binding and fertilization. J. Cell Biol. 91, 325331. Knutson, V. P. (1991). Cellular trafficking and processing of the insulin receptor. FASEB J . 5 , 21 30-2138. Larabell, C., and Chandler, D. E. (1991). Fertilization-induced changes in the vitelline envelope of echinoderm and amphibian eggs: Self-assembly of an extracellular matrix. J . Electron Microsc. Tech. 17, 294-318. Lee, H. C., Aarhus, R., and Walseth, T. F. (1993). Calcium mobilization by dual receptors during fertilization of sea urchin eggs. Science 261, 352-355. Lennarz, W. J. (1994). Fertilisation in sea urchins: How many different molecules are involved in gamete interaction and fusion? Zygote 2, 1-4. Lois, A. F., Lackey, D. A., and Carroll, E. J. (1986). Partial purification and characterization of sperm receptor hydrolase, a cortical granule proteoesterase, from eggs of the sea urchin Strongylocentrotus purpuratus. Gamete Res. 14, 397-32 1. Longo, A. C . , Lynn, J. W., McCulloh, D. H., and Chambers, E. L. (1986). Correlative ultra structural and electrophysiological studies of sperm-egg interactions of the sea urchin, Lyrechinus variegatus. Dev. Biol. 118, 155- 166. Mangelsdorf-Sonderquist, A,, and Carpenter, G. (1986). Biosynthesis and metabolic degradation of receptors for epidermal growth factor. J. Membr. Biol. 90, 97-105. Minor, J. E., Gao, B., and Davidson, E. H. (1989). The molecular biology of bindin. In “The Molecular Biology of Fertilization” (H. Schatten and G. Schatten, eds.), pp. 773-788. Academic Press, San Diego, CA. Moore, D. D. (1992). Preparation and analysis of DNA. In “Short Protocols in Molecular Biology” (F. M. Ausubel, R. Brent, R. E. Kingston, D. D. Moore, J. G. Seidman, J. A. Smith, and K. Struhl, eds.), 2nd ed., pp. 2.1-2.38. Wiley, New York. Moore, K. L., and Kinsey, W. H. (1994). Identification of an abl-related protein tyrosine kinase in the cortex of the sea urchin egg. Dev. Biol. 164, 430-443. Moy, G. W., and Vacquier, V. D. (1979). Immunoperoxidase localization of bindin during sea urchin fertilization Curt-. Top. Dev. Biol. 13, 31-44. Ohlendieck, K., and Lennarz, W. J. (1995). Role of the sea urchin egg receptor for sperm during gamete interactions. Trends Biochem. Sci. 20, 29-33. Ohlendieck, K., Dhume, S. T.,Partin, J. S., and Lennarz, W. J. (1993). The sea urchin egg receptor for sperm: Isolation and characterization of the intact, biologically active receptor. J . Cell Biol. 122, 887-895. Ohlendieck, K., Partin, J. S . , and Lennarz, W. J. (1994a). The biologically active form of the sea urchin egg receptor for sperm is a disulfide-bonded homo-multimer. J. Cell Biol. 125, 817-824. Ohlendieck, K., Partin, J., Stears, R., and Lennarz, W. J. (1994b). Developmental expression of the sea urchin egg receptor for sperm. Dev. Biol. 16.5, 53-62. Rossignol, D. P., Roschelle, A. J., and Lennarz, W. J. (1981). Sperm-egg binding: Identification of a species-specific sperm receptor from eggs of Strongylocentrotus purpuratus. J. Supramol. Struct. Cell Biochem. 15, 347-358. Rossignol, D. P., Earles, B. J., Decker, G. L., and Lennarz, W. J. (1984). Characterization of the sperm receptor on the surface of eggs of Strongylocentrotus purpuratus. Dev. Biol. 104, 308-321. Ruiz-Bravo, N., and Lennarz, W. J. (1986). Isolation and characterization of proteolytic fragments of the sea urchin sperm receptor that retain species specificity. Dev. Biol. 118, 202208.
58
Kay Ohlendieck and William J. Lennarz
Ruiz-Bravo, N., and Lennarz, W. J. (1989). Receptors and membrane interactions during fertilization. In “The Molecular Biology of Fertilization” (H. Schatten and G. Schatten, eds.), pp. 21-36. Academic Press, San Diego, CA. Ruiz-Bravo, N., Janak, D. J., and Lennarz, W. J. (1989). Immunolocalization of the sea urchin sperm receptor in eggs and maturing ovaries. B i d . Reprod. 41, 323-334. Schmell, E., Earles, B. J., Breauz, C . , and Lennarz, W. J. (1977). Identification of a sperm receptor on the surface of the eggs of the sea urchin. Arbuchiu punctuluiu. J. Cell Biol. 72, 3546. SeGall, G. K., and Lennarz, W. J. (1981). Jelly coat and induction of the acrosome reaction in echinoid sperm. Dev.Biol. 71, 33-48. Shen, S. S. (1995). Mechanisms of calcium regulation in sea urchin eggs and their activities during fertilization. Curr. Top. Dev. Biol. 30, 63-101, Shimomura, H., Dangott, L. J., and Garbers, D. L. (1986). Covalent coupling of a resact analogue to guanylate cyclase. J. Biol. Chem. 261, 15778-15782. Shyu, A. B., Raff, R. A., and Blumenthal, T. (1986). Expression of the vittelogenic gene in female and male sea urchins. Proc. Nutl. Acud. Sci. U.S.A. 83, 3865-3869. Starr, M . , Himmelmann, J. H., and Theniault, J. C. (1990). Direct coupling of marine invertebrate spawning with phytoplankton blooms. Science 247, 1070- 1074. Suzuki, N. (1990). Sperm-activating peptides from sea urchin egg jelly. Bioorg. Mar. Chem. 3, 47-70. Suzuki, N., Shimomura, H., Radany, E. W., Ramarao, C. S . , Ward, G. E., Bentley, J. K., and Garbers, D. L. (1984). A peptide associated with eggs causes a mobility shift in a major plasma membrane protein of spermatozoa. J. B i d . Chem. 259, 14874-14879. Trimmer, J. S., and Vaquier, V. D. (1986). Activation of sea urchin gametes. Annu. Rev. Cell Biol. 2, 1-26. Trimmer, J. S., Schlackman, R. W., and Vaquier, V. D. (1986). Monoclonal antibodies increase intracellular CaZ+ in sea urchin spermatozoa. Proc. Nutl. Acud. Sci. U.S.A. 83, 9055-9059. Ullrich, A , , and Schlessinger, J. (1990). Signal transduction by receptors with tyrosine kinase activity. Cell (Cambridge, Muss.) 61, 203-212. Vacquier, V. D., and Moy, G. W. (1977). Isolation of bindin: The protein responsible for adhesion of sperm to sea urchin eggs. Pmc. Nutl. Acud. Sci. U.S.A. 74, 2456-2460. Ward, C. R., and Kopf, G. S . (1993). Molecular events mediating sperm activation. Dev. Biol. 158, 9-34. Wassarman, P. M. (1993). Mammalian eggs, sperm and fertilization: Dissimilar cells with a common goal. Semin. Dev. Biol. 4, 189-197. Whitaker, M. J., and Swann, K. (1993). Lighting the fuse at fertilization. Development (Cumbridge, U K ) 117, 1-12.
3 Fertilization and Development in Humans Alan Trounson Institute of Reproduction and Development Monash Medical Centre Clayton, 3 168 Australia
Ariff Bongso Department of Obstetrics & Gynaecology National University of Singapore Singapore
I. Introduction Oocyte Maturation Sperm Capacitation, the Acrosome Reaction, and Sperm Maturation Gamete Interactions Fertilization Fertilization Abnormalities Micromanipulative Fertilization Techniques Embryonic Cleavage and Developmental Anomalies A. Cleavage of Human Embryos B . Morphology and Scoring of Human Embryos C. Cell Number D. Developmental Anomalies IX. Determination of Genetic Errors in Gametes and Embryos X. Cryopreservation of Oocytes and Embryos XI. Embryo Metabolism and Viability XII. Conclusions References
11. 111. IV. V. VI . VII. VIII.
1. Introduction The possibilities for exploring the physiology of human conception and early embryonic development were initiated with the publications of Edwards and colleagues (1970; Steptoe et al., 1971) showing that in vitro fertilized human oocytes develop to blastocysts in relatively simple culture media. These studies, and the complementary stages of human development (Fig. 1) published by Trounson et al. (1982) over a decade later, when superovulatory techniques (Trounson et al., 1981) provided the basis for the clinical application of in vitro fertilization (IVF), have provided the opportunity to study the fascinating process involved in converting the disparate gametes of men and women into an embryo Current 70pic~in Devclopmenral Biology. Vul. 32 Copyright 0 1996 by Academic Press, Inc. All ngha of reprcducuon in any lbrm reserved
59
60
Alan Trounson and Ariff Bongso
Fig. 1 Embryonic development in the human. (A) pronuclear, (B) 2-cell. (C) 4-cell, (D) 8-cell, reproduced with permission of the Journal of Reproduction 9 Fertility; (E) morula, and (F) blastocyst stages (reproduced with permission from Trounson et a / . , 1982).
with potential for development of all the tissues and organs of the human body. Despite a relatively brief time of observation and research, the data on human embryology are extensive and are revealing new and interesting aspects of mammalian fertilization and development.
3. Fertilization and Development in Humans
61
In this review of human fertilization and preimplantation embryo development we describe and interpret the rapidly expanding literature being published in this area. Remarkable achievements in this area of medicine reveal a diversity of basic research and clinical application. The capability of human sperm to fertilize oocytes in vitro; the growth and development to blastocysts in vitro; cryopreservation of oocytes and embryos of all developmental stages, cell sampling of embryos by biopsy and the identification of chromosomal and genetic errors; fertilization by direct insertion of a sperm into the mature ovum; and the maturation of immature oocytes and their development to blastocysts in vitro are among the major developments that have established unprecedented knowledge and options for clinical use in human reproductive medicine.
II. Oocyte Maturation Oocytes recovered after superovulation for IVF and gamete intrafallopian tube transfer (GIFT) are obtained by ultrasound-guided aspiration of preovulatory follicles at 36 to 38 h after administration of human chorionic gonadotropin (hCG) to induce oocyte maturation within the follicles in vivo. The majority (63%) are mature, metaphase I1 oocytes at recovery, and there is a progressive increase in oocytes completing maturation in the 4 to 6 h of culture prior to insemination, to a maximum of around 85% (Fig. 2). The delay in insemination to enable the completion of oocyte maturation increases fertilization rates and the developmental capacity of oocytes (Trounson et al., 1982; Osborn, 1993). There is a progressive decline in immature germinal vesicle stage oocytes from 16.1% at recovery to 3.1 % by 6 to 8 hours of culture, most entering maturation during this period spontaneously (Fig. 2). Human oocytes, obtained after hyperstimulation for IVF, which remain at immature germinal vesicle stages despite the administration of hCG to induce maturation of preovulatory follicles, will mature in vitro in culture medium without gonadotropins [follicle-stimulating hormone (FSH) and luteinizing hormone (LH)] within 24 to 30 h (Veeck et al., 1983; Osborn, 1993). The timing of oocyte maturation in vivo resembles this interval, with the germinal vesicle still present 8 h after hCG injection, the metaphase I stage at 20 h after hCG, and the mature metaphase I1 stage observed at 35 h after hCG (Bomsel-Helmreich et al., 1987). The maturation in vitro of immature, germinal vesicle-stage oocytes, recovered early in the follicular phase from untreated polycystic ovarian patients, appears to be retarded when compared with oocytes in vivo or when recovered 36 h after hCG injection from superovulated IVF patients (Trounson et al., 1994). When immature oocytes were recovered from patients early in the follicular phase, or from anovulatory polycystic ovarian syndrome patients, 17% were at metaphase I1 by 23 to 25 h, 60% were at metaphase I1 by 43 to 47 h, and 81% and matured to metaphase I1 by 48 to 54 h of culture in medium containing gonadotrophins, estrogen, and fetal calf serum (Fig. 3; Trounson et al., 1994).
62
Alan Trounson and Ariff Bongso
Hours after oocyte recovery Fig. 2 Proportion of germinal vesicle (GV) stage, maturing metaphase 1 (MI), and mature metaphase I1 (MII) oocytes after recovery from superovulated IVF patients (figure drawn from data quoted by Osbom, 1993).
There was no difference between ovulatory or nonovulatory polycystic ovarian patients, and in more recent experiments (F. Barnes and A. Trounson, unpublished observations), similar observations were made with immature oocytes recovered from naturally cycling, nonpolycystic ovarian patients in the early- to mid-follicular phase. These oocytes were capable of fertilization after in vitro maturation, development to blastocysts, and development to term when embryos were replaced in the patients. Although timed observations during maturation in vitro of oocytes recovered from ovariectomy specimens were not done by Cha et al. (1991), fertilization and embryo development were achieved when oocytes were inseminated after 32 to 48 h of maturation in culture. It was of interest to observe that oocytes surrounded by atretic (dying) cumulus cells and oocytes recovered with relatively few cumulus and corono radiata cells in the early- to mid-follicular phase were capable of maturation, fertilization, and embryo development in vitro (F. Barnes and A. Trounson, unpublished data). This suggests that the human ovum, like that of the sheep (Moor and Trounson, 1977), retains its developmental capacity during the initial phase of follicular
63
3 . Fertilization and Development in Humans
[7 100
-
80
-
60
-
40
-
20
-
n=31
0 -
MI
GV
n=41
II 21-22
n=63
MI1
n=54
23-25
43-47
48-54
TIME (Hours)
Fig. 3 Maturation of oocytes recovered from anovulatory polycystic ovarian syndrome patients by transvaginal ultrasound when cultured in EMEM, 10% FCS + 0.075 IU HMG, 0.5 IU HCG, and I mg E,/ml (from Trounson, A , , Wood, C., Kausche, A. Fertil. Steril. 1994; 61:353-62. Reproduced with permission of the publisher, The American Fertility Society).
atresia in the ovary, and that maturation and fertilization can be achieved independently of the presence of healthy follicular cells. One of the problems observed with the maturation of human oocytes in polyspemic fertilization after insemination. This may be due to inadequate maturation of cortical granules, resulting in a slow or inadequate cortical reaction after fusion of the ovum and sperm. This can be corrected by intracytoplasmic sperm injection (ICSI) using micromanipulation techniques (van Steirteghem et a l ., 1993a,b). The factors governing maturation of human oocytes in vitro are yet to be elucidated. The coculture of immature oocytes with cumulus/granulosa cells of mature preovulatory follicles did not affect the rates of maturation, fertilization, or development (Trounson et al., 1994). The presence or absence of estrogen had no significant effect on maturation, fertilization, and development (F. Barnes and A . Trounson, unpublished data).
64
Alan Trounson and Ariff Bongso
111. Sperm Capacitation, the Acrosome Reaction, and Sperm Maturation Mammalian spermatozoa, once released at ejaculation, are motile, but they do not have fertilizing capacity. After they are deposited within the female reproductive tract they undergo a series of physiological events during transport through the uterus and oviducts that prepares them for the fertilization process when they encounter the oocyte in the ampullary-isthmic region of the oviduct. The three major physiological events, in order of occurrence, are capacitation, the acrosome reaction, and hyperactivated motility (Yanagimachi, 1994). Capacitation occurs while sperm continue their migratory process within the uterus and oviduct before they successfully interact with the oocyte. The composition of the environment within the reproductive tract can drastically modulate the capacitation process. Free calcium (Ca2+)and sodium (Na+) are key inducers in this process. Extracellular Ca2+ is usually required for completion of capacitation and for acrosomal exocytosis, but the concentrations required for both processes differ widely. The kinetics involved in changing intracellular Ca2+ concentrations are not well understood, although the involvement of Ca2+ ATPase has been recently suggested (Fraser, 1993). Na+ has also been implicated in both capacitation and the acrosome reaction. Monovaient cation ionophores accelerate capacitation and acrosomal exocytosis in a Na+ concentration-dependent manner. It has been suggested that Na+/K+ ATPase activity may regulate Na+ concentration during capacitation, Na+/H+ exchanges playing a key role in the initiation of the next process of acrosomal exocytosis. From experiments involving enzyme inhibitors and ionophores, Fraser (1993) postulated that during capacitation there were modest rises in Ca2+ and Na+. In capacitated spermatozoa, large rises in Na2+ activated the Na+/H+ exchanger to promote an efflux of H+ and a subsequent rise in pH. This, in turn, activated dihydropyridine-sensitive calcium channels, causing a large influx of Ca2+ and triggering acrosomal exocytosis. Capacitation is a reversible process, unlike the acrosomal reaction, and there is no visual manifestation of this preliminary process. The acrosome reaction in the human is an exocytotic event involving the fusion and vesiculation of the outer acrosomal membrane and surrounding plasma membrane, leading to dispersal and release of acrosomal contents. Good, upto-date reviews of capacitation and acrosome reaction both in man and animals have been provided by Zaneveld et al. (1991) and Drobnis (1993). Recent information on the human sperm acrosome reaction leads to questions regarding some of the old dogmas. It has been now claimed that the acrosome reaction can be initiated after activation of membrane receptors, leading to protein phosphorylation and/or increase in intracellular calcium ions (Ca2+). Three signal transduction pathways that lead to protein phosphorylation and the acrosome reaction have been identified (i) the adenylate cyclase-CAMP-protein kinase A pathway; (ii) the phospholipase C-diacylglycerol-protein kinase C pathway; and (iii) the
3. Fertilization and Development in Humans
65
guanylate cyclase-cGMP-protein kinase G pathway (Zaneveld et al., 1993). Interestingly, the third pathway can be stimulated by the atrial natriuretic peptide (ANP) found in human follicular fluid. The ANP concentration in follicular fluid was shown to be associated with the outcome of in vitro fertilization (Anderson et al., 1993). It has also been recently shown that isolated human zonae pellucidae induce the human acrosome reaction and involve all three pathways (Bielfeld et al., 1994). It has not been firmly established whether tyrosine kinase activity is involved in the human acrosome reaction. An influx of Ca2+ from the extracellular environment, progesterone, arachidonic acid, and calcium ionophore (A23 187) can induce the acrosome reaction via this mechanism. Zaneveld et al. (1993) questioned the view that a large influx of Ca*+ was an absolute requirement for the acrosome reaction. ANP and activators of protein kinases A and C can induce the acrosome reaction in the nominal absence of extracellular Ca2+ (de Jonge et al., 1991; Anderson et al., 1993). Bielfeld et al. (1994) recently demonstrated that the human acrosome reaction was osmosensitive and inhibited in a hyperosmotic medium. It was therefore hypothesize that water influx may be an important aspect of'this process; whether this occurs before or after fusion of acrosomal and plasma membranes has not been definitely established. Spermatozoa appear to have protective mechanisms preventing an early acrosome reaction during transit within the reproductive tract. The acrosomal contents, mainly acrosin, which essential for the fertilization process, and usually released only after binding of sperm to the zona pellucida. It has been suggested that specific receptors that interact with oocyte ligands (the zona glycoprotein ZP3, ANP, progesterone) are present on the sperm surface before the acrosome reaction is initiated or completed in vivo (Zaneveld et al., 1993). A 74.kDa glycoprotein (ARIG), that inhibits the acrosome reaction of human sperm induced to capacitate by stimulators of protein kinases A and C has been isolated from human seminal plasma (Drisdel et al., 1992). Hyperactivated motility of sperm occurs after sperm binding to the zona. These dashing and dancing movements of the sperm have been claimed to be necessary for the sperm to penetrate the matrix of the zona. The biochemical and physiological phenomena surrounding this event are not clearly understood, although secretory products in the ampullary-isthmic oviductal milieu have been implicated. The percentage of sperm with hyperactivated motility after 6 h of incubation in a capacitated medium correlated significantly with the fertilization rate of human oocytes in 52 couples undergoing assisted reproduction. Multivariate discriminant analysis selected six sperm hyperactivated motility and acrosome reaction variables of predictive value to classify semen samples that achieve fertilization rates above and below 70% (Wang et al., 1993). Capacitation, the acrosome reaction, and hyperactivated sperm motility can be induced and simulated in vitro. It has been demonstrated that after ejaculation in the human only a few sperm lose their acrosomes, while after 24 h of incubation in Whittingham's T6 medium supplemented with 30 mgiml bovine serum albu-
66
Alan Trounson and Ariff Bongso
min (BSA), the number of reacted sperm increases to 17%. Further, the acrosome reaction could be increased significantly by a 24-h incubation in T6 medium with follicular fluid, or by a 24-h incubation in T6 medium and exposure for a very short time to an electric field, followed by a further incubation for a few hours in T6 medium containing 3.5 mM pentoxyfilline (Palermo el al., 1992a). Using computer-automated semen analysis (CASA), Bongso et al. (1993) showed that the motility of normozoospermic and oligozoospermic sperm was significantly increased when they were coincubated with monolayers of the human tubal ampullary epithelium over I-, 5-, and 24-h periods. Curvilinear velocity (km/s) and mean amplitude of lateral head displacement (mALH, pm) values increased, but linearity (%) and beat cross frequency (hertz/s) did not increase. It was also shown in this study that such tubal monolayers did not increase the human acrosome reaction through the 24-h period. Human sperm coincubated with human follicular fluid, granulosa, or cumulus cells exhibited a significantly higher ability to penetrate zona-free hamster ova for up to 48 h (Bastias et al., 1993). These authors claimed that specific secretory factors produced in the human preovulatory ovarian follicle may enhance human sperm fertilizing capacity. In search of a reliable diagnostic assay for the human sperm acrosome reaction, Aitken and Brindle (1993) explored the behavior of probes targeted at the outer acrosomal membrane or at constituents of the acrosomal vesicle after induction of the acrosome reaction with the ionophore A23187. The Arachis hypogaea lectin, used as a probe against the outer acrosomal membrane, consistently gave higher results than the Pisum sativum lectin or monoclonal antibody CRB9 directed at the acrosomal contents. These authors concluded that the outer acrosomal membrane was dispersed from reacting sperm more rapidly than certain major constituents of the acrosomal vesicle. It appears that the epididymis is essential for normal reproduction, because sperm leaving the testis are incapable of fertilizing an ovum if simply mixed with oocytes in v i m . However, it has been claimed by some workers that it is not necessary for sperm to reside in the human epididymis in order to develop into mature cells (Mooney et al., 1972; Silber, 1989). This is in contrast to the evidence available in animal studies that sperm maturation (the acquisition of the ability to fertilize oocytes) occurs at sperm traverse the various parts of the epididymis (Cooper, 1986; Amann, 1987; Blaquier et al., 1988; Lacham and Trounson, 1991). Interestingly, human epididymal and testicular sperm have the ability to fertilize human oocytes both in vivo and in v i m . It has been suggested that human sperm fertilizing capacity develops fully in the distal part of the epididymal tract, based on the higher motility and oocyte-fusing ability of sperm taken from these regions of unobstructed tissue (Cooper, 1990). Although pregnancies have resulted from sperm aspirated from the epididymis in patients with occluded vas deferens, the pathological state of the tissue precludes definitive statements about the exact function of a normal epididymis. It thus cannot be
3. Fertilization and Development in Humans
67
concluded definitely whether the human epididymis plays a role in sperm fertilizing capability. In a classic review, Amann et al. (1993) stated that the role of the epididymis in sperm maturation was to “set a series of triggers,” each capable of initiating cellular changes at emission or near or in the oocyte, and then “setting a safety” for each trigger to prevent premature occurrence of the event. Sperm motion was characterized using CASA for five regions of eight epididymides from eight prostatic carcinoma patients undergoing castration and from eight epididymal spermatocoeles located adjacent to the head of the epididymides and the testes in five patients (Yeung et al., 1993). The mean percentage motility increased from 22.9 4.8 in the efferent ducts to 68.3 -+ 7.9 in either the mid- or distal corpus epididymides and declined in the cauda epididymis. Straight line velocity also increased in the mid-corpus epididymal region. The authors claimed that their objective quantification of sperm motion documents the maturation of sperm mobility within the human epididymis, confirming that this maturation pattern is similar to that in other mammals (Yeung et al., 1993). Human sperm maturation was also studied in vitro in the presence of human epididymal epithelial primary cultures. Progressive sperm motility and its capacity to bind to salt-stored human zona pellucida was investigated after coincubation of human caput epididymal sperm for 48 h with 3-day-old epithelial cultures from the cauda epididymis. Sperm motility and sperm binding to zonae pellucidae were significantly increased compared to controls in sperm exposed to epididymal cultures. The authors concluded that their method may be valuable for improving the fertilizing capacity of human sperm retrieved from the proximal region of the excurrent ducts (Moore et al., 1992).
*
IV. Gamete Interactions Gamete interaction in vitro is distinctly different to that in vivo because of the large number of sperm in the immediate vicinity of the oocyte during IVF. Sperm hyaluronidase disassociates the cumulus matrix rapidly in vitro, contracting corona cells around the oocyte. Sperm bind to the zona pellucida in large numbers. Transmission electron miscroscopy of oocytes inseminated in v i m shows that acrosome-reacted and acrosome-intact sperm can be found between corona radiata cells and on the surface of the zona pellucida (Chen and Sathananthan, 1986). The acrosome-reacting capacity of corona radiata cells or their secretions has not been comprehensively explored, but Tesarik ( 1985) reported that exposure of sperm to the cumulus oophorus increases acrosome reactions of human sperm. Both solubilized human zonae and the human zona glycoprotein ZP3 have sperm receptor activity, a number of studies have shown that human zonae induce the sperm acrosome reaction (Shabanowitz and Rand, 1988; de Jonge et a l ., 1988; Van Kooij and te Velde, 1988). The acrosome reaction-inducing capacity of human zona glycoproteins is yet to be elucidated, but it is likely that the 57- to
68
Alan Trounson and Ariff Bongso
73-kDa ZP3 glycoprotein is responsible. The human zona also contains ZP1 (90110 kD), ZP2 (64-76 kD), and probably two isoforms of ZP3 (Shabanowitz and Rand, 1988; Van Duin et a l . , 1992). The primary structure of human ZP3 and the complementary mRNA closely resemble that of the mouse (Chamberlin and Dean, 1990), and it is likely that, as for the mouse, correctly glycosylated sugar chains are required in conjunction with the protein backbone for the acrosome reaction to occur (Kinloch et a l . , 1991). The potential to express functional recombinant ZP3 proteins or fragments of this molecule will further define the initial stages of gamete interaction in the human. Acrosome-intact and partially reacted sperm can be found entering the zona pellucida, but only acrosome-reacted sperm are found penetrating the deeper regions of the zona (Sathananthan et a l . , 1993). While it is presumed that secondary binding of the sperm to the glycoproteins ZP2 and ZP3 is important for sperm passage through the zona, it is apparent that enzymes liberated from the inner acrosomal membrane are necessary to create the digested tract made by the penetrating sperm (Sathananthan et a l . , 1993). Hyperactive sperm motility is also essential for penetration of the zona, and motility is maintained on entry into the perivitelline space. Studies involving the injection of immotile sperm into the perivitelline space (subzonal microinjection) show a very reduced incidence of fertilization, best exemplified by the experiments reported in the mouse by Lacham and Trounson (1991). Although fertilization is possible after subzonal microinjection of immotile human sperm (Bongso et al., 1989b), fertilization rates have always been higher with motile sperm. During sperm-egg fusion, phagocytic processes on the oocyte surface fuse with the plasma membrane in the mid-segment of the sperm head extending from the equatorial vestige to the anterior region of the post-acrosomal segment. Fusion of the human gamete membranes appears to involve the amino acid sequence, Arg-Gly-Asp (RGD) of the fibronectin molecule, which is known to be involved with adhesion of cells to intergrins. This suggests there may be a homologue role in the human of the guinea pig sperm PH-30 protein, which is composed of two tightly linked, integral membrane proteins, with an extracellular disintegrin domain that interacts with an integrin receptor on the oocyte plasma membrane processes (Blobel et a l . , 1992; Fusi et a l . , 1992). It has also been shown in the human that the Clq component of the C1 immune complex can interact with immune complexes through a specific receptor (ClqR) on sperm and also with fibronectin (Fusi et al., 1991). The membrane cofactor protein (CD46) has been identified in the acrosome of human sperm (Anderson et a l . , 1989) and oocytes (Roberts et a l . , 1992), and may protect the gametes from immune attack. CD46 binds to the complement component C3b, which is present on human oocytes. These gamete interactions are likely to be crucial in their recognition and in determining species-specificity of gamete interaction. It is evident that a coherent cascade of interactions is required for uniting the gametes for fertilization.
3. Fertilization and Development in Humans
69
V. Fertilization Fertilization normally occurs in the ampulla of the fallopian tube soon after ovulation. In the human, fertilization can also occur in the uterus, as demonstrated by pregnancies after transfer of gametes to the uterine cavity (Craft et al., 1982). However, the success rate of this procedure is very much lower than that of GIFT, where gametes are transferred to the fallopian tube (Asch et al., 1988). Few sperm are in the vicinity of the ovulated oocyte in vivo, and it is likely that the success of GIFT for the treatment of unexplained (idiopathic) infertility in the human is owed to the presence of large numbers of sperm around the oocytes in the fallopian tube, increasing the probability of gamete interaction and fertilization. The major barrier between the gametes is the zona pellucida. The zona has several important properties: it allows sperm penetration, but, after the cortical reaction, prevents polyspermic penetration; it maintains the cleaving blastomeres in a nested grouping, allowing compaction after the third cleavage division to produce a spherical ball of internalized and externalized cells. The zona prevents cellular immune attack of the early cleavage stage embryo and is porous to electrolytes, sugars, carbohydrates, amino acids, and large-molecular-weight proteins. However, unlike the case in some species, such as sheep (Trounson and Moore, 1974), an intact zona is not essential for survival of human cleavagestage embryos transferred to the uterus (Cohen er al., 1990). Capacitated and acrosome-reacted sperm digest a pathway through the zona aided by the accelerated fonvard-thrusting motion produced by the hyperactive beating of the flagellum. Motile sperm enter the perivitelline space and fuse with the oocyte plasma membrane usually within 1 to 2 h after insemination in vitro. The gamete fusion event initiates extrusion of cortical granule contents from the cortical surface of the oocyte into the perivilline space. Cortical granule exocytosis occurs circumferentially and is completed within a few minutes (Sathananthan and Trounson, 1982a,b; Sathananthan e l al., 1993). The cortical granules contain enzymes that interact with the inner surface of the zona, creating a denser inner zona structure that cannot be digested by sperm enzymes, thus effectively preventing supernumerary sperm from peneterating the zona, which would create the lethal condition of polyspermia, caused by entry of two or more sperm into the oocyte. When the sperm fuses with the oocyte, a process akin to phagocytosis occurs where the midpiece of the sperm is engulfed by a micro-process extended by the oocyte cortical ooplasm, the sperm envelop inflates and the inner acrosomal membrane ruptures at several points when the sperm head is incorporated within the oocyte, allowing chromatin decondensation to begin (Sathananthan et al., 1993). Levels of maturation promoting factor (MPF) activity, which increase after extrusion of the first polar body, remain high during arrest of the oocyte at metaphase 11. MPF is a protein-serine/ threonine kinase composed of phosphory-
70
Alan Trounson and Ariff Bongso
lated cdc2 and cyclin B molecules (Murray, 1992). Activation of MPF by association of cdc2 with cyclin B and phosphorylation of cdc2 is required for dissolution of sperm head membranes, decondensation of sperm chromatin, and male pronucleus formation. Inhibition of MPF activation by inhibition of protein synthesis or phosphorylation prevents formation of male but not necessarily female pronuclei (Funahashi er al., 1993). The c-mos proto-oncogene, which encodes a protein-serinekhreonine kinase (Mos), is involved in progression of oocytes from metaphase I to metaphase I1 and, as a component of cytostatic factor, is required to maintain metaphase-I1 arrest, prevent the degradation of cyclin B, and hence, maintenance of MPF activity until after fertilization (O’Keefe et al., 1991). It has been recently shown that c-mos messenger RNA is present in human oocytes and that inhibition of protein synthesis in mature human oocytes resulted in loss of MPF activity and the induction of female pronucleus formation with the loss of meiotic arrest at metaphase I1 (Pal et al., 1994). These data demonstrate that MPF activity probably has the same role in human oocyte maturation and fertilization as in other species. The oocyte is activated soon after fusion of the gametes, with anaphase I1 of the second meiotic division. The sperm is incorporated within the oocyte and the decondensing sperm head is found beneath a fertilization cone at the time of telophase 11, when the second polar body is abstricted (Sathananthan et al., 1993). It is not known whether gamete fusion and activation of second messenger systems is the normal inducer of oocyte activation or whether the release of a sperm “oscillagen” is responsible for release of calcium stores within the oocyte and for the reactivation of meiosis (Swann, 1993). It is of interest that when a
Fig. 4 Morphological development of bipronucleate zygotes.
71
3. Fertilization and Development in Humans
sperm is microinjected directly into the oocyte cytoplasm the oocyte is usually only activated if a male pronucleus forms, suggesting that sperm release an activating factor during demembranation and sperm head decondensation. Nuclear membranes form around the decondensing sperm and oocyte chroma-
D
GERMINAL VESICLE NOPOLARBODY PROMINENT NUCLEOLUS
'ACTIVATED' SINGLE PRONUCLEUS 2 POLAR BODIES
METAPHASE I NOGVORPB
ABNORMALLY FERTILIZED 1 PRONUCLEUS + CONDENSED SPERM
00
UNFERTILIZED NO PRONUCLEI SINGLE POLAR BODY
NORMALLY FERTILIZED 2 PRONUCLEI 2 POLAR BODIES
'ACTIVATED'
POLYSPERMIC
POLYCYNIC
2 PRONUCLEI 1 POLAR BODY
3 PRONUCLEI
3 PRONUCLEI
2 POLAR BODIES ( 1 FRAGMENTED)
1 POLAR BODY
: DlGYNlC
FRAGMENTED
UNFERTILIZED NO PRONUCLEI FRAGMENTED POLAR BODY
NORMALLY FERTILIZED 2 PRONUCLEI 2 POLAR BODIES (1 FRAGMENTED)
POSSIBLY FERTILIZED 2 POLAR BODIES
VARIOUS DEGREES OF IMMEDIATE CLEAVAGE: LARGE 2ND POLAR BODY CAUSED BY SPINDLE MIGRATION
Fig. 5 The normal appearance of fertilization stages and range of abnormalities seen in human oocytes during in virro fertilization (Trounson and Osborn, 1993b. Reprinted with permission from Handbook of In Vitro Fertilization, pp. 57-84, 1994. Copyright CRC Press, Boca Raton, Florida).
72
Alan Trounson and Anff Bongso
tin and the pronuclei move into close proximity in the center of the oocyte within 3 to 6 h of gamete fusion, where they remain for 20 to 24 h after insemination. Replication of DNA occurs at the pronuclear stage. The oocyte has centrosomes but no centrioles, and inherits a centriole from the sperm, although both organize mitotic spindles. The sperm centriole is always associated with the male pronucleus (Sathananthan et d . , 1991). Microtubules nucleated by the centrioles and cenetrosome appear at the time of pronuclear membrane dissolution, forming a bipolar, barrel-shaped spindle with paternal and maternal chromosomes aligned at syngamy on the metaphase plate of the first cleavage division. Within 2 to 3 h of pronuclear membrane dissolution, the oocyte cleaves into two approximately equal-sized blastomeres containing diploid chromosome complements. Dissolution of pronuclei ,is usually synchronous, although asynchrony has been observed, particularly when there are multiple pronuclei (Fig. 4). There are many abnormalities of fertilization observed in human oocytes fertilized in vitro (Fig. 5 , Trounson and Osborn, 1993b). The presence of multiple pronuclei as a result of polyspermy leads to a high frequency of cleavage and chromosomal abnormalities. Tripronuclear occytes most frequently cleave to three cells, with mosaic chromosome numbers in daughter blastomeres (Kola et al., 1987), due to the formation of a tripolar spindle because of the presence of two functional sperm centrioles (Sathananthan et al., 1991).
VI. Fertilization Abnormalities The formation of two distinct pronuclei, each containing clear nucleoli, with the extrusion of a distinct second polar body approximately 16 to 18 h after insemination, is indicative of normal fertilization. Many errors in this normal fertilization event occur in vitro; these can only be studied and documented accurately using high-magnification inverted phase-contrast and differential interference contrast optics. Fixing of the abnormally fertilized eggs, followed by staining with Giemsa, sheds important light on their chromosomal behavior and on the decondensation process of sperm that have entered into the ooplasm, both of which, eventually, will explain the pathogenesis of the abnormal fertilization event. The two most common errors of fertilization are the formation of mono- and polypronuclear embryos. Oocytes with a single pronucleus are usually parthenogenetically activated by a variety of causes. They account for about 1.6% of oocytes inseminated in IVF laboratories and for the higher rates (of up to 7%) observed when aging oocytes (42 h after oocyte collection) are inseminated (Plachot and Crozet, 1992). Increased oocyte aspiration pressures (> 100 mm Hg), ethanol, acid Tyrode’s solution, and vigorous handling with pipettes are some of the common causes of parthenogenetic activation. Chromosome analysis of such oocytes usually reveals haploid sets (n = 23) that have been classified as
3. Fertilization and Development in Humans
73
parthenogenones , gynogenones, or androgenones. Parthenogenesis has been defined as the development of a single female pronucleus with an X chromosome without the intervention of a male gamete, whereas gynogenesis also may also occur as a result of fertilization by a spontaneously or experimentally genetically inactive spermatozoon. The development of an oocyte containing only a male genome, with the female complement being completely extruded or genetically inactivated, is called androgenesis. Cytogenetic analysis on large numbers of monopronuclear oocytes has revealed that parthenogenones or gynogenones are most frequent with a haploid chromosome complement (Plachot and Crozet, 1992). Monopronuclear embryos usually have two polar bodies and can undergo normal cleavage up to the blastocyst stage. They usually appear as morphologically good-quality embryos with minimum fragments and symmetrical blastomeres. In 45% of abnormal monopronuclear oocytes examined, decondensed sperm heads and tiny nucleus-like structures were observed in the ooplasm in addition to a single nucleus. The Y chromosome was detected in 10%of the oocytes by fluorescent in situ hybridization (FISH) using a Y chromosomespecific DNA probe. These observations provided evidence that many of these monopronuclear oocytes originated from fertilized oocytes (Balakier 1993; Balakier et al., 1993). In established IVF laboratories, oocytes are examined for fertilization (presence of two pronuclei) at around 16-18 h after insemination. A chromosomal analysis of 41 embryos derived from monopronuclear oocytes revealed a haploid chromosome complement in 12.2%, triploidy in 7.3%, and a normal diploid chromosome set in 80.5% of embryos. Also, when 312 monopronuclear oocytes were examined twice, first at 16-18 h and later at 20-24 h postinsemination, 25% showed a second pronucleus in the second examination (Staessen el al., 1993). The cleavage of the latter embryos was similar to the cleavage of embryos showing two pronuclei in the first examination. The replacement of such embryos with delayed pronuclear formation has produced live births. It was therefore suggested that a second repeat observation of monopronuclear embryos should become standard practice in IVF laboratories (Staessen et al., 1993). Fresh and aged human oocytes can be activated parthenogenetically with calcium ionophore at rates of 50 to 60%, whereas ethanol was shown to be a poor activating agent (16%; Winston et al., 1996). Recently, controlled parthenogenetic activation of human oocytes was reported by De Sutter et al. (1992). A 5- to 10-h exposure of unfertilized oocytes from an IVF program to 10 Fg/ml puromycin produced the highest percentage of activation (86 to 88%), and almost all parthenogenetically activated oocytes entered or developed beyond the first cleavage mitosis. These results were confirmed by fixing the parthenogenetic oocytes for chromosome analysis 2 h after nuclear envelope breakdown. The incidence of polypronuclear embryos in human IVF ranges between 5 and 10% (Dandekar et a l . , 1990). Triploidy may result from retention of the second
74
Alan Trounson and Ariff Bongso
polar body (digyny) or fertilization with more than one sperm (diandry or polyspermy). The great majority of triploid zygotes (86%) result from diandry (Plachot et al., 1989). Polyploid embryos could cleave regularly with nice, symmetrical blastomeres to the blastocyst stage, implant, and then abort or form hydatidiform moles. There have been reports of triploid infants that have been born and subsequently died a few weeks after birth (Plachot et al., 1989). Therefore, polyploid embryos are not usually replaced: for ethical reasons and to avoid falsely raising a patient’s hope of a viable pregnancy. However, it was shown that the third pronucleus could be a pseudopronucleus (Rudak et al., 1984); through breakdown and disappearance of this third pronucleus, a normal diploid (2n = 46) embryo could be established (Kola et al., 1987). But such triploid embryos have not been followed to see whether they develop into normal, viable offspring. Attempts have been made to correct polyspermy by removal of one of the pronuclei using micromanipulation (Gordon et al., 1989; Malter and Cohen, 1989). It has been difficult, however, to confirm accurately the exact origin of the pronucleus prior to removal. The advantages of this novel approach are limited, since only 24% of tripronuclear oocytes develop into triploid embryos after the first cleavage division (Kola et al., 1987). Other abnormalities at fertilization include (i) activated oocytes with two female pronuclei, brought about by retention of the second polar body without the contribution of a sperm, (ii) oocytes with two polar bodies but with delayed appearance, or with two very faint pronuclei, and (iii) a large second polar body caused by spindle migration (Fig. 4). It is not clear whether fragmented polar bodies and fragmented oocytes are brought about by errors of the fertilization process or by intrinsically defective oocytes. Reliable cytogenetic analysis of this entire spectrum of anomalies will shed light on the exact pathogenesis of these morphologically abnormal events.
VII. Micromanipulative Fertilization Techniques In attempts to overcome the barrier of the zona pellucida for fertilization with sperm from low-quality semen samples of infertile men, a number of different techniques have been explored. Laws-King et al. (1987) showed that human oocytes can be fertilized by microinjection of a single spermatozoon into the perivitelline space (subzonal microinjection), and Cohen et al. (1989) showed that improved fertilization rates could be obtained with poor-quality sperm if the zona was opened using glass needles (partial zona dissection) before insemination. Both of these techniques were further developed and refined for use in clinical in vitro fertilization. Despite reasonable success for the treatment of otherwise intractable male factor infertility (Fishel et al., 1990; Cohen et al., 1991, 1994; Sakkas et al., 1992; O’Neill et al., 19941, monospermic fertilization rates remained less than 30% of oocytes injected. Both partial zona dissection
3. Fertilization and Development in Humans
75
and subzonal injection of multiple sperm resulted in high rates of polyspermic fertilization, showing that attempts to increase access of sperm to the oocyte cell by disrupting the zona or injecting multiple sperm into the perivitelline space increased the number of abnormally fertilized oocytes. However, these techniques had relatively little impact on normal fertilization over that achieved with subzonal injection of single sperm because the mechanism preventing polyspermy (the zona reaction) could no longer operate. These techniques offered very little for cases involving immotile or very low motility sperm samples despite an occasional report of fertilization. Some improvement in the outcome of subzonal microinjection was reported by McLachlan et al. (1994) when the microinjected oocytes were returned to the fallopian tube after injection (microinjection and intrafallopian transfer: MIFT), suggesting that the tuba1 environment may promote gamete fusion by providing complementary molecules not available in culture. An interesting set of experiments was reported by Lazendorf et al. (1988), where sperm were injected directly into human oocytes, producing fertilized oocytes. This technique was refined by Palermo et al. (1972b) to produce pregnancies. The intracytoplasmic sperm injection (ICSI) technique was compared with subzonal microinjection and shown to be more successful (51% of oocytes fertilized normally after ICSI, compared to 14% with subzonal microinjection of sperm) in the experiments reported by Van Steirteghem et al. (1993a). The same authors (Van Steirteghem et al., 1993b) reported an increase in normal fertilization to 64% and a high rate of implantation when the developing embryos were transferred to the patients. Interestingly, there was no influence of sperm quality on fertilization rate in their study, and the authors have confirmed this observation in further clinical studies. These results have established ICSI as the method of choice for cases involving severely reduced semen quality. The ICSl technique does not involve any particular preparation of sperm for microinjection. The sperm are normally immobilized before injection using the microinjection pipette to crush the midpiece or tail. However, we have shown that motile sperm can be microinjected into the cytoplasm with equal success, but sperm from completely immotile ejaculates have a significantly reduced capacity to form pronuclei (Lacham-Kaplan and Trounson, 1994). It is perhaps surprising that sperm with intact acrosomes can be converted to a male pronucleus, but this is clearly the case. It is also reassuring that there is no apparent increase in birth defects following ICSI (Van Steirteghem, 1994). These results suggest that the barriers to sperm which act to select the sperm for fertilization in vivo have no real influence on the chromosomal normality of the conceptus. Further research has shown that immature epididymal and testicular sperm can be used to obtain fertilization and pregnancy in the human by ICSI (Trounson et al., 1994; Tournaye et al., 1994; Lacham-Kaplan and Trounson, 1994). Sperm are obtained by surgical aspiration from the epididymal tubules or by needle biopsy of the testis. As a result, men with obstructive azoospermia can be treated
76
Alan Trounson and Ariff Bongso
by ICSI. Since only one sperm is required for each oocyte when using ICSI, sperm recovered by surgery or needle biopsy can be cryopreserved for multiple cycles of oocyte recovery. Fertilization and pregnancy using cryopreserved sperm have been obtained in our own studies, enabling the more productive patient recycling procedure described by Trounson (1994) to be used for patients requiring surgical recovery of epididymal sperm. The development and application of ICSI for achieving fertilization for an extremely wide range of sperm abnormalities will provide a useful research technique for exploring gamete function in a very different way from that available in the past. It is interesting that this technique has been developed in the human and is only now being used to explore gamete function in laboratory animals and other species. The application of ICSI to aid breeding of endangered species is also being explored in our own laboratory.
VIII. Embryonic Cleavage and Developmental Anomalies A. Cleavage of Human Embryos
Until the advent of assisted reproductive techniques (ART) in the human, it was virtually impossible to record reliable information on human preimplantation embryonic cleavage and development. The first descriptive information on human embryonic cleavage from the second to the fifth day in vivo was recorded for four normal and four abnormal specimens on reproductive tracts removed at hysterectomy (Hertig et al., 1954). The more recent documentation of in vivo development was made on donated ova recovered by nonsurgical uterine lavage (Buster et al., 1985). Twenty-five uterine ova were collected by lavage 5 days after the LH peak from five fertile donors having a single periovular insemination in vivo. Embryonic development ranged from the one-cell to blastocyst stages. The mean age of five blastocysts was 109.1 h, and these included early and late cavitating stages. Today, most established centers practicing IVF culture spare embryos for freezing at stages ranging from the 2-pronuclear to blastocyst stage. It is therefore possible to record accurate information on human embryonic development in vitro. Pronuclei are usually observed at around 16 to 18 h after insemination of metaphase I1 oocytes. In some patients, pronuclei could appear as early as 12 to 14 h and as late as 20 to 22 h (A. Bongso, personal observations). By 44 to 48 h postinsemination, embryos have usually reached the four-cell stage. In some patients who became pregnant after embryo transfer, two-cell stage embryos were observed at 40 to 44 h, rapidly cleaving to the four-cell stage at 48 h just before transfer. In some cases, slow cleavage has been attributed to subnormal incubator conditions such as humidity and temperature. Polyploid embryos are known to cleave slightly faster, usually to five to six cells at 48 h post insemina-
3. Fertilization and Development in Humans
77
tion (Bongso et al., 1989a). Eight cells are usually observed in normal embryos at 72 h (third day), the compacted (morula) stage at 96 h (fourth day), and blastocysts from Day 5 onward (Veeck, 1992). The mean time for the first three cleavage divisions was reported by Trounson et al. (1982) to be at 35.6, 45.7, and 54.3 h postinsemination. Human blastocyst stages can range from an early appearance of a blastocoelic cavity (early cavitating blastocyst) to an expanded blastocyst, and these stages usually appear between Day 5 to Day 7. Interestingly, blastocysts developing as early as Day 5 produce significantly higher pregnancy rates after transfer when compared to blastocysts cavitating on Day 6 or Day 7 (Frydman et al., 1993). Also, the use of helper monolayers (cocultures) appears to yield increased numbers of blastocysts cavitating as early as the fifth day (Bongso and Fong, 1993; Menezo et al., 1993; Frydman et al., 1993). Using human tuba1 epithelial cells for coculture, Bongso et al. (l992b) observed healthy, early cavitating blastocysts at 115 h postinsemination (morning of Day 5) that resulted in ongoing pregnancies after transfer. It is apparent that blastocysts which cavitate earliest have the highest viability on transfer to patients. The distinction between goodand poor-quality blastocysts is also important for obtaining viable pregnancies (Figs. 6 to 8; C.-Y. Fong and A. Bongso, unpublished data). Human blastocyst grading as an indicator of developmental potential was recently reported (Dokras et a l . , 1993). Blastocysts were graded from 1 to 3 (BGI-BG3). A good, viable blastocyst (BGl) showed early cavitation, with the laying down of a healthy trophoectoderm and a distinct, clear inner cell mass on Day 5. BG2 blastocysts also showed cavitation on Day 5 but had a transitional, vacuole-like appearance, whereas BG3 blastocysts were those formed on Day 6 with several degenerative foci. Poor morulae developed single and multiple cavities and were classified as vacuolated morulae. It is essential for better assessment of the success of implantation rates after transfer of blastocysts to patients and for cryopreservation that the correct grade of blastocyst be selected.
B. Morphology and Scoring of Human Embryos
The morphology of embryos has been shown to be positively correlated to implantation rates (Grill0 et al., 1991; Scott et al., 1991; Trounson and Osbom, 1993b). A very sensible and reliable grading system, using three parameters (regularity of blastomere size, fragmentation, and granularity of blastomere cytoplasm) that were strongly associated with pregnancy rates, was reported by Erenus et al. (1991). A simple, well-defined grading system was proposed by Bolton et al. (1989), where the best, Grade 4, embryos had regular spherical blastomeres with no fragments, and the Grade 3 embryos had regular spherical blastomeres with some fragments. The poor, Grade 2, embryos had slightly irregular blastomeres with considerable fragments, and Grade I embryos had
78
Alan Trounson and Ariff Bongso
fig. 6 Poor quality blastocyst. Human expanded blastocyst I15 h postinsemination grown in medium alone without coculture. Note trophoblast and absence of distinct inner cell mass (ICM).
barely defined blastomeres with many fragments. The same authors also claimed that the better-quality embryos (Grades 3 and 4) produced blastocysts at higher rates (23 and 18%, respectively) compared to the poor-quality embryos (Grades 1 and 2), at rates of 5 and 6%, respectively. Two scoring systems for embryo morphology in the human have been reported. An embryo development rating (EDR) based on the formula EDR = (TO/TE) X 100 (TO, observed time of cleavage; TE, expected time of cleavage based on a normogram for expected growth rate regression) was used to predict pregnancy by Cummins et al. (1986). Embryos with an EDR of 110 to 130 were more likely to implant than slower-cleaving embryos. Later, by adding to the EDR a rating for the symmetry of the blastomeres, percentage of fragments, and cytoplasmic clarity of blastomeres, a further improvement in the prediction of pregnancy was achieved. The second scoring system of "cumulative embryo
3. Fertilization and Development in Humans
79
Fig. 7 Good quality blastocyst. Human expanded blastocyst 115 h post insemination cocultured on human oviductal epithelial cells. Note distinct inner cell mass (ICM) and trophoblast.
score” (CES) was introduced by Steer ef al. (1992). They combined the score for embryo quality with the number of embryos transferred. The CES was derived from the formula CES = 2, (EG X CN), + (EG X CN), + (EG X CN), ...(EG, embryo grade, CN, cell number for n embryos). An embryo grading system similar to that of Bolton el al. (1989) was used. Pregnancy rate increased significantly when the CES value reached 42, but showed no further increase. The multiple pregnancies occurred when the CES was greater than 42. The CES was recently calculated in 602 embryo transfers and correlated with pregnancy rate, pregnancy outcome, and the incidence of multiple pregnancies. The scoring system was applied to cycles where endotoxins were either present or absent in the culture medium, to evaluate its validity in quality control. Pregnancy rates increased from 4% for scores between 1 and I0 to 35 in the 41-50 group. A score of 20 was the statistical criterion to separate patients with good and poor pregnancy prognoses. Biochemical abortions occurred significantly more frequently with scores of <20. Birth rates after transfer of embryos below and above a score of 20 (2.8 and 19.2%) differed significantly, and the scores for births yielding triplets were >40. The CES did not show a correlation between embryo quality and endotoxin levels (Visser and Le Fourie, 1993).
80
Alan Trounson and Ariff Bongso
Fig. 8 Good quality blastocyst. Human early hatching blastocyst 120 h postinsemination cocultured on human oviductal epithelial cells. Note distinct inner cell mass (ICM) and trophoblast.
C. Cell Number Information on the number of inner cell mass (ICM) cells and on the trophoectoderm (TE) or total cell number (TCN) in human blastocysts is limited, as existing in vitro conditions are not used to generate large numbers of blastocysts. Spare embryos are usually frozen before the 8-cell stage owing to concern that in vitro degenerative changes may occur in further culture in vitro. However, it has been reported that newly expanded blastocysts had a TCN of approximately 60 cells on Day 5 postinsemination, which increased to approximately 80 and 125 on Days 6 and 7, respectively (Hardy et al., 1989a; Hardy, 1993). Morphologically abnormal blastocysts had large numbers of dead cells on Day 6 and had only half the cells of normal blastocysts. Both ICM and TE can be
3. Fertilization and Development in Humans
81
separately counted by differential labeling of nuclei using polynucleotide-specific fluorochromes. The mean mitotic index gradually decreased from 4.3 to 0.7, whereas the dead cell index increased from 8.5 to 16.7 when blastocysts were cleaving from Days 5 to 7. The same authors concluded that if cell death were ignored at the 16-cell stage, there would be approximately 5 cells allocated to the ICM and 11 cells to the TE, and at the 32-cell stage there would be approximately 11 and 21 cells allocated to the ICM and TE, respectively.
D. Developmental Anomalies
Embryonic blocks in development have been recorded for many laboratory and farm animal species. In the human, a block in cleavage between the 4- and the 8-cell stages has been suggested. Changes in the pattern of polypeptide synthesis, and some of the major qualitative changes dependent on transcription, were reported to occur between the 4- and the 8-cell stages. It appears that cleavage is not sensitive to transcriptional inhibition until after the 4-cell stage (Braude et al., 1988). Using sodium dodecyl sulfate-polycrylamide gel electrophoresis and silver staining, the polypeptides in different populations of human oocytes and embryos were visualized and compared. Two polypeptide patterns differing in the 6-kDa were resolved. These patterns showed no correlation with the potential for oocytes to fertilize and develop normally to eight cells (Capmany and Bolton, 1993). It was also demonstrated that human embryos of decreasing quality (fragments and irregular blastomeres) had high proportions of anucleate and polynucleate blastomeres (Winston et al., 1991a). More recently, Hardy (1993) examined the nuclei of disaggregated blastomeres from 200 human embryos between Days 2 and 4 after insemination in vitro by vital labeling with a polynucleotide-specific fluorochrome. Interestingly, 17% of normally fertilized embryos at the 2- to 4-cell stages had at least one binucleate blastomere. At the 9to 16-cell stages, this increased to 65% when individual embryos had between one and six binucleate blastomeres. Estimates of the volume of binucleate blastomeres, based on measurements of cell diameters and comparison with mononucleated blastomeres, indicated that these multinucleated blastomeres arose from a failure of cytokinesis between the second and the fourth cleavage divisions. It was postulated that blastomeres with either binucleate or abnormal nuclei contribute to cleavage arrest in vitro. The high frequency of abnormal nucleokinesis and cytokinesis during embryonic cleavage may be a major contribution to the low viability of human embryos grown under current in vitro conditions. Such anomalies may, in turn, be related to suboptimal culture conditions for oocyte maturation and embryonic development, because a large number of nuclear and cytoplasmic abnormalities have been observed in both unfertilized and fertilized oocytes examined after insemination in vitro (Sathananthan and Trounson, 1985; Balakier and Casper, 1991).
82
Alan Trounson and Ariff Bongso
IX. Determination of Genetic Errors in Gametes and Embryos Chromosomal imbalance has been implicated as one of the contributory causes for the estimated high frequency of very early embryonic losses that occur shortly after implantation in vivo (Schlesselman, 1979) and for first-trimester abortions (BouC et al., 1975). Chromosome anomalies in oocytes and spermatozoa may contribute to fertilization failure and to the anomalies occurring in the early preimplantation embryo as well as in abortuses unable to sustain implantation before 120 days of gestation. Normal human males were shown to have a frequency of 8 to 10% of chromosomal anomalies in their sperm (Martin et al., 1983; Brandriff et al., 1984), and the human female is even more prone to meiotic chromosomal anomalies (Mattei et al., 1979). Fluorescence in situ hybridization (FISH) with repetitive-sequence DNA probes was recently performed on human interphase sperm to determine the use of this technique for aneuploidy detection. Hybridization efficiency was 98-99% for a combination of three probes, and the authors claimed that FISH was a simple, rapid, and reliable technique that provides an accurate screen for human sperm aneuploidy (Martin et al., 1993). Incidences of chromosomal abnormalities in fresh human oocytes, oocytes failing to fertilize in stimulated cycles, and fresh, uninseminated oocytes from unstimulated cycles have been documented by several workers. It appears that the mean incidence of aneuploidy in oocytes in reports with the larger series is around 2 I %, and the mean incidence of anomalies in embryos up to the four-cell stage is around 30%. There is no documentation on the incidence of chromosomal anomalies in human blastocysts. It has been suggested that the high rate of aneuploidy in human oocytes could be due, among other reasons, to superovulation protocols used in IVF (Gras et al., 1992). Evidence for this suggestion came from reports that there was an increased frequency of chromosomal anomalies in spontaneously aborted concepti arising from ovulation induction (Bout and BouC, 1973; Kola, 1988). However, Gras et al. (1992) failed to detect significant differences in the incidence of chromosomal aneuploidy in 68 uninseminated oocytes retrieved from stimulated patients (CIomidihMGihCG and Buserelin groups) compared to 20 oocytes from unstimulated patients. Unfortunately, the number of oocytes examined in their study was too small to make definite conclusions. Conflicting reports on the relationship of aneuploidy in oocytes to maternal age have been documented. Plachot et al. (1986) and Bongso er al. (1988a) showed that aneuploidy was significantly higher in oocytes failing to fertilize in IVF patients over 35 years of age, whereas Edirisinge et al. (1992) showed no relationship to maternal age. But nondisjunction leading to aneuploid embryos occurred more frequently in oocytes from older mice (Maudlin and Fraser, 1977). Interestingly, aneuploidy rates are much higher in stimulated and unstimu-
3. Fertilization and Development in Humans
83
lated human oocytes than in other species. The reasons for this are unclear, but one major difference in the reproductive physiology between the human and the animal species is that the human suffers from a range of subfertility problems, which apparently is not the case in animals. Thus, the differences may be related to the level of subfertility in human patients, because all studies reported thus far have been on subfertile patients. It is interesting to note that two independent studies on stimulated oocytes in fertile women and on oocytes retrieved from fertile patients undergoing ovarian biopsies or ovariectomies showed aneuploidy rates ranging from only 3 to 3.6%, and were therefore much lower than those observed in subfertile women (Jagiello et al., 1975; Tarin et al., 1991). In a cytogenetic study of human oocytes remaining unfertilized after IVF, Angel1 et al. (1993) showed that they could be classified into degenerating and “healthy” oocytes. The degenerating oocytes, which had different degrees of chromosome breakage, accounted for a quarter of the total. These were found in older patients with a mean age of 35 years. The healthy oocytes, which had mainly 23 chromosomes, could also be classified into two groups: one with a normal MII, 23, X-chromosome complement; and the other, in which single chromatids replaced whole chromosomes. These authors postulated that the single chromatids at the second meiotic metaphase arise by precocious division of chromosomal univalents at anaphase I, which may be the major mechanism for trisomy formation in humans, rather than the nondisjunction of whole bivalents, which is the commonly held view. Cytogenetic analysis was recently carried out on human oocytes failing to fertilize due to dysfunctional spermatozoa. Forty-seven percent of such oocytes were chromosomally abnormal, and the incidences of chromosomal anomalies in oocytes that did and did not develop pronuclei were 26.6 and 20.4%, respectively. It was concluded from statistical analysis that there was no relationship between chromosomal abnormality in the oocyte and the capacity to achieve fertilization in vitm (Almeida and Bolton, 1993a). The same authors, in a separate study, also showed that in 237 analyzable oocytes that failed to form pronuclei after insemination with normal fertile sperm, 29.5 and 58.7% showed nuclear/cytoplasmic immaturity and chromosomal abnormalities, respectively (Almeida and Bolton, 1993b). Oocyte aneuploidy is usually brought about by errors in oogenesis through the mechanisms of anaphase lagging or nondisjunction. Theoretically, the hypo- and hyperhaploidy estimates should follow 1: 1 ratios, although numerous reports document higher values for hypohaploidy. These deviations have been attributed to an artificial loss of chromosomes brought about by the fixation technique. There is a greater tendency for loss or gain of the smaller satellite chromosomes of groups D and G of the human karyotype. Interestingly, translocations in subfertile couples with histories of habitual miscamages usually involve the D and G groups as well (Bongso et al., 1991b). Polyploidy and mosaicism (approximately 30%) have been reported in good-
84
Alan Trounson and Ariff Bongso
and poor-quality human embryos (Plachot et al., 1987). A similar incidence of polyploidy and diploid/haploid and triploid/haploid mosaicism was also observed in a large series of fragmenting, poor-quality embryos that were not replaced into IVF patients (Bongso et al., 1991a). A 40% incidence of chromosome anomalies in abnormal embryos, of which 14% were mosaic, was reported by Papadopoulos et al. (1989). More recently, a cytogenetic analysis of 118 analyzable, poor-quality, fragmenting embryos revealed that 90% of them had abnormal chromosome complements ranging from aneuploidy and mosaicism to structural rearrangements. It was suggested that the replacement of such poor-quality embryos into IVF patients would be unethical (Pellestor et al., 1994). The polyploidy observed in poor-quality human embryos is usually brought about by polyspermy (polyandry) or polygyny (retention of the first polar body), and the mosaicism as a result of mitotic nondisjunction or anaphase lagging, where whole groups of chromosomes are involved. Interestingly, the incidence of mosaicism is also high in chorionic viili used for prenatal diagnosis in the first trimester of gestation. It appears that errors in mitosis occur spontaneously in the cells of late preimplantation and early postimplantation human embryos. The FISH procedure has been used on human embryos to investigate the role of chromosome numbers 18, 13, 21, X , and Y in aneuploidy. Aberrations for these chromosomes were found in 70% of abnormally developing, monospermic embryos. True mosaicism, distinguishable from technique failure, polyploidy, monosomy, and trisomy observed in previous studies, was confirmed using the FISH technique. However, aneuploidy was the main chromosome abnormality observed in normally developing, monospermic embryos (Munnt et al., 1993). Using a chromosome 18-specific probe and the FISH technique directly on human preimplantation embryos, Schrurs et al. (1993) concluded that more than a single nucleus was necessary for accurate preimplantation diagnosis of aneuploidy. Using FISH with DNA probes for chromosomes X , Y, and 19, Benkhalifa et al. (1993) analyzed human morulae and blastocysts grown by coculture with Vero cells. The proportion of embryos with more than five polyploid cells was 30.8% for morulae and 29.3% for blastocysts. The results were confirmed by conventional cytogenetic analysis, and the authors concluded that the preimplantation diagnosis of genetic errors at these embryonic stages may not be reliable, since results were usually obtained from single cells that were not representative of the whole embryo.
X. Cryopreservation of Oocytes and Embryos Human embryos may be cryopreserved using a range of techniques and cryoprotectants that preserve the cellular integrity of blastomeres. These methods have been described and their success reviewed by Trounson and Shaw (1994) and
3. Fertilization and Development in Humans
85
Shaw et al. (1993). The most frequently used technique involves the slow-cooling of embryos in phosphate- or Hepes-buffered culture media containing 1.5 M 1,2propanediol and 0.1 M sucrose. The embryos are slow-cooled from -7"C, when ice nucleation is induced, to -30°C at O.TC/min, then rapidly to -150°C (50°C/min) before storage in liquid nitrogen. Embryos are thawed by warming in air for 50 s and rapid warming in a 30°C water bath before removal of the cryoprotectant in 0.2 M sucrose solutions (Shaw et al., 1993). Embryos are frozen at the pronuclear to eight-cell stage; in good-quality embryos, more than 80%survive with most of their blastomeres intact. The implantation of thawed embryos is not significantly different than that of nonfrozen embryos, more than doubling the pregnancy success rate from a single oocyte collection (Wang et al., 1994). The slow-freezing techniques involve progressive dehydration of the cells through extracellular ice formation, concentration of salt in the unfrozen solution around the embryos, entry of the cryoprotectant into the cells, and the diffusion of water from the cells into the extracellular compartment where it is frozen. The presence of sucrose as an extracellular solute increases dehydration of cells. Slow cooling results in large rounded ice crystals and enables the slow dehydration necessary to avoid formation of lethal intracellular ice formation. Because small intracellular ice crystals may still form, the rapid warming phase during thawing is used to avoid their growth and consequent damage to cells. Other cryoprotectants such as dimethyl sulfoxide (DMSO) may also be used, but they have slightly different properties, which require modification of technique to account for their cell toxicity, diffusion rate into cells, glass-forming (vitrification) properties, and interaction with other solutes and ions. With modification of the freezing and thawing methods, similar success of cryopreservation of embryos can be achieved. The addition of inhibitors of membrane lipid peroxidation (e.g., apotransfemn and ascorbate) can also significantly increase the rate of survival and viability of cryopreserved embryos (Tarin and Trounson, 1993). Rapid-cooling techniques, which involve higher cryoprotectant concentrations, have been used to cryopreserve human embryos (Trounson and Shaw, 1992). Solutions designed for vitrification are now commonly used for embryobanking of mouse embryos. These solutions are designed for low toxicity of cells and for their capacity to vitrify rather than to crystallize during cooling and warming (Ali and Shelton, 1993b). Using high concentrations of DMSO (4.5 M ) and sucrose (0.25 M ) , Shaw et al. (1991a) showed that all stages of mouse embryos (pronuclear to blastocyst) may be cryopreserved without loss of developmental capacity. The blastocyst stage is sensitive to high concentrations of DMSO at ambient temperatures; this toxicity can be avoided by reducing the temperature of exposure to DMSO to 0°C. Glycerol- and ethylene glycol-based media, with extracellular solutes such as sucrose and Ficoll, have also been used very successfully for a range of preimplantation embryo stages and species. There has been only limited investigation of vitrification and rapid-cooling cryopreservation methods for human embryos (Trounson and Shaw, 1992).
86
Alan Trounson and Ariff Bongso
It has been reported that fast freezing can result in increased chromosomal abnormalities in mouse embryos (Bongso el al., 1988b). This is related to cryoprotectant concentrations and the ice-forming properties of the cryopreservation medium (Shaw et al., 1991b). At low concentrations (1.5 M DMSO), chromosomal abnormalities in rapidly frozen embryos are dramatically increased. At intermediate concentrations (3 M DMSO), the abnormalities are still evident, but they are not detected at high concentrations (4.5 M DMSO). Ice forms in rapidly frozen solutions during rapid cooling in intermediate to low concentrations of DMSO, but not in high concentrations (Shaw et al., 1991b). The formation of ice during cooling may cause the chromosomal damage observed. Examination of mitochondria in human embryos frozen rapidly in 3.5 M DMSO and 0.25 M sucrose showed no effect of the cryopreservation technique (Noto et al., 1993). Further research on the use of rapid-cooling and vitrification methods for the cryopreservation of human embryos is deserved, given the success of these techniques in mice and other species (e.g., sheep; Ali and Shelton, 1993a). However, the remarkable success of cryopreservation of human embryos using slow-cooling techniques (Trounson and Jones, 1993), and the similarity of pregnancy outcome to that for nonfrozen embryos [with a low incidence of congenital malformations (Wada et al., 1994)], make it difficult to introduce a new cryopreservation method when some concerns exist about the effects of rapidcooling methods on embryonic developmental capacity and chromosomal normality. The cryopreservation of unfertilized oocytes has produced very limited success in a wide range of species. In a study on cryopreservation of mouse oocytes, only 6-14% of oocytes frozen by slow-cooling techniques and 20-38% of oocytes cryopreserved by vitrification were capable of producing fetuses (Rall, 1993). Other species have been less successful, and only two births have been recorded for frozen human oocytes from a large number of attempts (Trounson, 1990). There are many problems for the unfertilized oocyte, including reduced fertilization, abnormalities of fertilization, and chromosomal abnormalities that result from cooling and freezing (Van Blerkom, 1991; Trounson and Shaw, 1994). More recently, Gook et al. (1993) examined human oocytes frozen by conventional slow-cooling procedures using 1.5 M propanediol and 0.1 M sucrose. Sixty-four percent were judged to have survived freezing and thawing, with further loss during culture for 24 h after thawing. They showed that 60% of oocytes had apparently normal, barrel-shaped spindles after thawing, compared to 81% of nonfrozen oocytes. However, they were unable to determine the chromosomal normality of human oocytes after fertilization and extrusion of the second polar body, which is essential for determination of oocyte function and normality after cryopreservation. Interestingly, only 4% of mouse oocytes survived this freezing method, and none fertilized normally (Gook et d . , 1993).
3. Fertilization and Development in Humans
a7
Further experience using this technique (Gook et al., 1994) showed that the survival after freezing of freshly collected oocytes was 51% after thawing and only 41% after insemination with sperm. Only 46% of the surviving oocytes had two pronuclei after insemination. The appearance of two pronuclei in aged oocytes (cultured overnight before freezing) was found in only around 10% of surviving oocytes. Similarly, the overall production of two pronucleate oocytes from freshly recovered and frozen oocytes was 18.6%, which is not different from other reports of human oocyte cryopreservation and is well below the success required for the clinical application of oocyte cryopreservation in human IVF, given the very successful procedures available for fertilized oocytes and embryos. If one assumes a 10% implantation rate of pronuclear oocytes or embryos, the cumulative survival rate to fetuses would be less than 2% using this particular freezing method. Only four scoreable karyotypes were analyzed for the fertilized oocytes, all of which were diploid (46 chromosomes), and no stray chromosomes were identified using DNA fluorochromes. This is a dubious test for unincorporated chromosomes, and there was no examination of chromosomal breaks or rearrangements. With so few karyotypes adequately examined, it is not possible to conclude that the freezing of unfertilized oocytes produces functional and chromosomally normal embryos. More recent studies in the mouse (George et al., 1994), using methods designed to minimize cryodamage, have shown that fertilization rate and the capacity to develop to blastocysts in vitro can be maintained; the overall cumulative survival of cryopreserved oocytes to viable fetuses was 30 to 40% less than that for nonfrozen oocytes. The relatively high implantation rate of embryos (46%) derived from frozen and thawed oocytes and the normal rate of survival of implanted embryos to fetuses provide encouragement that some of the major obstacles can be overcome by careful design of the methods used for cryopreservation. The studies reported by George et al. (1994) are substantially better than most other reports of freezing mouse oocytes. Further development of this technique of slow cooling and slow warming in 1.5 M DMSO is warranted, and the application for cryopreserving human oocytes should be explored. The known variability of replicate observations in oocyte freezing (Bernard and Fuller, 1993) also deserves systematic analysis if highly reproducible results are to be obtained for oocyte cryopreservation. In part, this variability may be due to variable ice formation during cooling and warming, resulting from the composition of cryoprotective media and the technique of freezing and thawing. The derivation of low-toxicity vitrification media for oocyte cryopreservation is likely to overcome these problems of replicate variability. Progress in the design of such vitrification media, which include tehalose and ethylene glycol, is resulting in an improving success rate of oocyte cryopreservation (Arav et al., 1993; Rayos et al., 1994). At the present time there is no apparent advantage in the freezing of immature, rather than mature, human oocytes (Toth et al., 1994).
88
Alan Trounson and Ariff Bongso
XI. Embryo Metabolism and Viability Studies on the metabolism and viability of human embryos are limited. The major reason for this is that attempts to study the metabolism of the human embryo by manipulation and exposure to suboptimal external environmental conditions before transfer back into the patient may compromise a potential pregnancy. Noninvasive metabolic methods that were used for the mouse were first applied to humans by Leese et al. (1986) in a study on pyruvate uptake by the human embryo. The mature human oocyte was shown to have very high levels of pryuvate uptake (36 pmol/embryo/h), which declined after fertilization and then slowly increased to reach peak values just prior to the morula stage (27 pmol/embryo/h) before decreasing again. Degenerating embryos were shown to have much lower pyruvate uptakes. Later, Hardy et al. (1989b) employed a noninvasive study on pyruvate and glucose uptake by 73 individual human oocytes and embryos. The pyruvate uptake increased from approximately 28 to 40 pmoliembryoih between Days 2.5 and 4.5 (Day 1 = insemination of oocytes). Similarly, glucose uptake increased from approximately 8 to 14 pmol/embryo/h between Days 2.5 and 4.5, and then increased further to 24 pmol/embryo/h on Day 5 at the blastocyst stage. Polyspermic and parthenogenetic embryos, arrested embryos, and unfertilized oocytes had lower pyruvate uptakes at later stages. Both Leese et al. (1986) and Hardy et al. (1989b) concluded that noninvasive measurement of pyruvate uptake before embryo transfer may provide valuable information on the viability of human embryos. These studies were followed later by Gott et al. (1990), who examined the consumption of pyruvate and glucose and the release of lactate by 40 individual human embryos using a noninvasive technique. In 28 normally fertilized embryos, pyruvate uptake exceeded that of glucose on Days 2 to 5 postinsemination, and glucose became the dominant substrate at the blastocyst stage on Day 6. Considerable amounts of lactate were produced throughout development, increasing from 43.6 pmol/embryo/h on Day 2.5 to 95.4 pmol/embryo/h on Day 5.5. Pyruvate and glucose uptake and lactate production levels were lower for arrested embryos. Nearly 50% of the lactate produced could be accounted for in terms of glucose uptake from the medium, this value rising to 90% in the blastocyst. The authors postulated that the remaining lactate may have been derived from endogenous sources, probably glycogen. Early human embryo metabolism and the assessment of embryo viability was recently reviewed by Gardner and Leese (1993) and Leese et al. (1993). To date, there is no reliable method of determining the viability of human preimplantation embryos. Viability may be more accurately defined as the ability of an embryo to implant and sustain implantation to term, resulting in a healthy live birth. The noninvasive approaches will provide direct evidence of viability, as embryonic development could be followed to term whereas the invasive methods will be destructive to the embryo and provide only indirect information
89
3. Fertilization and Development in Humans
on viability [e.g., dye extrusion (trypan blue); fluorescence of blastomeres (DAPI staining); and production of fluorescent metabolites (fluorscein diacetate)]. The following procedures have been used to determine human embryonic viability: (1) analysis of the detailed morphological characteristics of the embryo, (2) growth in culture to the blastocyst stage, (3) production of plateletactivating factor, and (4) measurement of nutrient uptake. A combination of the regularity of blastomeres, percentage of fragment accumulation, and cleavage intervals is the most commonly used criterion for evaluating the pregnancy potential of an embryo before replacement. It is well known that embryos with regularly cleaving, symmetrical blastomeres and few or no fragments have the highest rate of implantation. However, there are patients who produce such highquality embryos in v i m but never become pregnant. Interestingly, it has also
Early passage Epithelioid
Wash monolayer: HBSS 2 x
Detach cells: Trypsin: EDTA IO.S%: 053mM)
t3/-
- 5-10mins Centrifuge: 1500 rprn; Smins Remove Trypsin: EDTA Wash cells (i) HBSS (ii) Medicult + 10% HS
(Conc. cell suspension)
Seed: Incubate
(Dilute cell suspension)
fWW,000 cellsKL8rnl Medicull + 10% HS, 5590
Fig. 9 Preparation of human oviductal cell monolayers in four-well plastic dishes for coculture, 24 h before oocyte recovery.
Alan Trounson and Ariff Bongso
90
been observed that some patients who undergo repeated IVF attempts, producing good embryos in one cycle, will continue to produce good embryos in subsequent cycles, irrespective of the variables in culture conditions. Similarly, those patients producing poor-quality embryos under the best culture conditions continue to produce the same poor-quality embryos in following cycles. This information suggests intrinsic factors within the embryo that may control its morphology irrespective of the extrinsic conducive conditions available (A. Bongso, personal observations). The ability of a patient’s spare embryos to grow to the blastocyst stage has been correlated with the viability of the four-cell stage embryos transferred from that particular cohort. In all patients where pregnancy occurred, at least one of the spare embryos developed into a blastocyst. The corresponding figure for nonpregnant patients was 53%. The results indicated that blastocyst development could be a useful predictor for the success of IVF. A detailed analysis revealed
(*I
Insemination tubes, MedicuB only)
5:5:9O gas mixture
DI: (9 am)
M:(9 am)
Growth dish ( h e l l plastic dish) 55:m gas mixture
Change medium to Medicult (no serum) Denude and transfer 2PN embrym to coculture
Change medium, gas (Medicult)
w: (9 am)
Change gas only
w:(9 am)
Change medium, gas (Medicult)
Hg. 10 Protocol for human blastocyst coculture.
3. Fertilization and Development in Humans
91
that this was not the case when in vitro culture before replacement was limited to 2 to 3 days (Sjogren et al., 1992). The use of helper cells in vitro (cocultures) has resulted in increased blastocyst production compared to conventional IVF culture media (Bongso et al., 1989c, 1992a; Wiemer et al., 1993; Menezo et al., 1993; Frydman et al., 1993; Quinn, 1993). Additionally, implantation and ongoing pregnancy rates also improved after transfer of four-cell-stage cocultured embryos (Wiemer et al., 1993; Bongso et al., 1992b; Menezo et al., 1993). Cocultured embryos had swollen blastorneres, lesser fragments, and more regular, symmetrical blastomeres compared to embryos grown in medium alone. A sequential transfer of two to three four-cell-stage embryos on Day 2, followed by a second transfer of a single blastocyst on Day 5 , appeared to improve clinical pregnancy rates (8/19 or 42%, Bongso et al., 1992b; A. Bongso, unpublished data). It was also shown that cocultured, blastocyst-stage single transfers on Day 5 produced significantly better pregnancy rates (52.8%) than four-cell-stage transfers on Day 2 (40.4%) in repeat transfers for patients who had at least a mean of 4.96 embryo transfers. No proper prospective, randomized, and controlled study has been undertaken to evaluate delivery rates after transfer of cocultured or noncocultured blastocysts. As such, although coculture appears to improve the viability of human embryos in vitro, reliable conclusions cannot be made until it is proven unequivocally that delivery rates are significantly increased. Perhaps the expanded or swollen nature of the blastomeres of embryos observed in coculture could be included as another morphological characteristic helping to predict viability. The preparation of monolayers and the protocol for human blastocyst coculture using human oviductal epithelial cells are illustrated in Fig. 9 and 10.
XII. Conclusions Since the birth of Louise Brown, the world’s first IVF baby, in 1978, assisted reproductive technologies have made a tremendous impact in human medicine. Major advances have taken place in this field over the last 15 years, and a wealth of scientific information has been accumulated in the field of human embryology. Many of the mysteries surrounding gamete maturation, the fertilization process, and embryonic development both in vivo and in vitro have been unravelled. We have attempted in this article to review most of the recent information available in this area, with particular reference to the human. Many further advances are expected in the coming years. It is hoped that the results of oocyte maturation in unstimulated cycles will complement the results in stimulated cycles, thus avoiding the use of hormonal ovarian stimulation, which is costly and has raised debate with regards to its side effects. The unstimulated technique also offers hope for patients with polycystic ovarian disease. If stimulation regimes are going to continue, there will be a wide choice of
92
Alan Trounson and Ariff Bongso
protocols, with some being favored for specific situations. There is general consensus that the hostility of the uterus and the viability of the replaced embryo are the two major impediments to improving take-home baby rates. Major advances are expected in the biology and clinical aspects of implantation and the development of improved in vitro systems for culturing of human embryos. Coculture of embryos with a variety of cell types offers much promise, not only in its direct use in improving embryo viability but also in providing information on the requirements and behavior of the embryo in v i m . Once these factors are identified, supplying these products to the culture media may prove simpler and equally effective. Treatment of the subfertile male has advanced by leaps and bounds recently with the advent of ICSI. Only one sperm is required, and viable offspring have been reported with no significant congenital anomalies. It appears that sperm donation may be restricted to the totally azoospermic male with a genetic etiology for his spermatogenetic arrest. The use of testicular and epididyma1 sperm, together with ICSI, is expected to yield viable pregnancies in the coming years. The cryopreservation of embryos has posed some difficult ethical and legal situations, which could be avoided with oocyte cryopreservation. Breakthroughs in oocyte cryopreservation have been slow, and no significant developments have taken place. Rapid-freezing methods such as vitrification are expected to replace the expensive, machine-programmed, slow-freezing methods. Preimplantation genetic diagnosis using FISH will have a major impact on prenatal diagnosis, as results will be reliable and rapid. Multiprobe analyses, with as many probes for the well-known genetic diseases, are expected. The development of embryonic stem cells from human blastocysts, the maturation and supply of ova from fetuses, sperm maturation in vitro, and gene therapy are some of the other new areas that are expected to provide rapid advancement in the coming years. With such advances in the field of human reproductive biology, further concurrent ethical and legal issues are expected. It is hoped that the reproductive biology, further concurrent ethical and legal issues are expected. It is hoped that the reproductive scientist will recognize these issues and not abuse, but use, the advancement of these technologies, with informed patient consent, for the benefit of humankind.
References Aitken, R. J., and Brindle, J. P. (1993). Analysis of the ability of 3 probes targeting the outer acrosomal membrane or acrosomal contents to detect the acrosome reaction in human spermatozoa. Hum. Reprod. 8, 1663-1669. Ah, J., and Shelton, J. N. (1993a). Successful vitrification of day-6 sheep embryos. J. Reprod. Fertil. 99, 65-70. Ali, J., and Shelton, J. N. (1993b). Design of vitrification solutions for the cryopreservation of embryos. J . Reprod. Fertil. 99, 471-411.
3. Fertilization and Development in Humans
93
Almeida, P. A., and Bolton, V. N. (1993a). Immaturity and chromosomal abnormalities in oocytes that fail to develop pronuclei following insemination in vitro. Hum. Reprod. 8, 229-232. Almeida, P. A., and Bolton, V. N. (1993b). The relationship between chromosomal abnormalities in the human preimplantation embryo and development in iifro. J. Reprod. Ferril 12, 64 (Abstr. 118). Amann, R. P. (1987). Function of the epididymis in bulls and rams. J . Reprod. Fertil., Suppl. 34, 115-120. Amann, R . P., Hammerstedt, R. H., and Veeramachaneni, D. N. R. (1993). The epididymis and sperm maturation: A perspective. Reprod. Ferril. Dev. 5 , 361-381. Anderson, D. J., Michaelson, J. S . , and Johnson, P. M. (1989). Trophoblast/leukocyte-common antigen is expressed by human testicular germ cells and appears on the surface of acrosomereacted sperm. Biol. Reprod. 41, 285-293. Anderson, R. A., Feathergill, K. A., Drisdell, R. C., Rawlins, R. G . , Mack, S . R., and Zaneveld, L. J. D. (1993). Atrial nutriuretic peptide (ANP) as a stimulus of the human acrosome reaction: Correlation of follicular ANP content with in vitro fertilization outcome. J . Androl. 15, 61-70. Angel], R . R., Xian, J . , and Keith, J. (1993). Chromosome anomalies in human oocytes in relation to age. Hum. Reprod. 8, 1047-1054. Arav, A,, Shehu, D., and Mattioli, M. (1993). Osmotic and cytotoxic study of vitrification of immature bovine oocytes. J . Reprod. Fertil. 99, 353-358. Asch, R. H., Balmaceda, J. P., Cittadini, E., Casas, P. F., Gomel, V., Hohl, M. K., Johnston, I . , Leeton, J., Escudero, F. I. R., Noss, U., and Wong, P. C. (1988). Gamete intrafallopian transfer. International cooperative study of the first 800 cases. Ann. N.Y. Acad. Sci. 541, 723728. Balakier, H. (1993). Tripronuclear human zygotes: The first cell cycle and subsequent development. Hum. Reprod. 8, 1892-1897. Balakier, H., and Casper, R. F. (1991). A morphologic study of unfertilized oocytes and abnorma1 embryos in human in vitro fertilization. J . In Vitro Fertil. Embryo Transfer 8, 73-79. Balakier, H., Squire, J., and Casper R. F. (1993). Characterization of abnormal one pronuclear human oocytes by morphology, cytogenetics and in sifu hybridization. Hum. Reprod. 8, 402408. Bastias, M. C., Kamijo, H., and Osteen, K. G . (1993). Assessment of human sperm functional changes after in vitro coincubation with cells retrieved from the human female reproductive tract. Hum. Reprod. 8, 1670-1677. Benkhalifa, M., Janny, L . , Vye, P., Malet, P., Boucher, D., and Menezo, Y. (1993). Assessment of polyploidy in human momlae and blastocysts using coculture and fluorescent in siru hybridization. Hum. Reprod. 8, 895-902. Bernard, A., and Fuller, B. J. (1993). Leter to the Editor: Variable loss of oocytes during cryopreservation. Hum. Reprod. 8, 1524. Bielfeld, P., Faridi, A , , Zaneveld, L. J. D., and De Jonge, C. J. (1994). The zona pellucidainduced acrosome reaction of human spermatozoa is mediated by protein kinases. Fertil. Steril. 61, 536-541. Blaquier, J. A , , Cameo, M. S . , Cnasnicu, P. S . , Gonzalez Echevevria, M. F., Pineiro, L., Tezon. J. C . , and Vazquez, M. H. (1988). On the role of epididymal factors in sperm fertility. Reprod. Fertil. Dev. 28, 1209-1214. Blobel, C. P., Wolfsberg, T. G., Turck, C. W., Myles, D. G., Primakoff, P., and White, J. M. ( 1 992). A potential fusion peptide and an integrin ligand domain in a protein active in spermegg fusion. Nature (London) 356, 248-252. Bolton, V. N., Hawes, S. M., Taylor, C. T., and Parsons, J. H. (1989). Development of spare human preimplantation embryos in v i m : An analysis of the correlations among gross morphology, cleavage rates and development to the blastmyst. J . In V i m Fertil. Embryo Transfer 6, 30-35.
94
Alan Trounson and Ariff Bongso
Bomsel-Helmreich, O., Huyen, L. V. N., Darand-Gasselin, I., Salarat-Baroux, J., and Antoine, J.-M. (1987). Timing of nuclear maturation and cumulus dissociation in human oocytes stimulated with clomiphene citrate, human menopausal gonadotrophin and human Chorionic Gonadotrophin. Fertil. Steril. 48, 586-595. Bongso, A , , and Fong, C. Y. (1993). The effect of coculture on human zygote development. Curr. Opin. Obstet. Gynecol. 5 , 585-593. Bongso, A., Ng, S. C., Ratnam, S., Sathananthan, A, H., and Wong, P. C. (1988a). Chromosome anomalies in human oocytes failing to fertilize after insemination in vitro. Hum. Reprod. 3, 645-649. Bongso, A., Ng, S . C., Sathananthan, A., Lee, M.-L., Mok, H., Wong, P. C., and Ratnam, S. S. (1988b). Chromosome analysis of two-cell mouse embryos frozen by slow and ultrarapid methods using two different cryoprotectants. Fertil. Steril. 49, 908-9 12. Bongso, A,, Ng, S. C.. and Ratnam, S . S . (1989a). Laboratory IVF technology. In Recent Advances in the Management of Infertility (C. Chen, S . L. Tan, and W. C. Cheng, eds.), pp. 191-205. McGraw-Hill, Singapore. Bongso, A., Sathananthan, A. H., Wong, P. C., Ratnam, S. S., Ng, S. C., Anandakumar, C., and Ganatra, S. (1989b). Human fertilization by microinjection of immotile sperm. Hum. Reprod. 4 , 175-179. Bongso, A., Ng, S. C., Sathananthan, H., Ng, P. L., Rauff, M., and Ratnam, S. S. (1989~). Improved quality of human embryos when cocultured with human ampullary cells. Hum. Reprod. 4, 706-713. Bongso, A , , Ng, S . C., Lim, J., Fong, C.-Y., and Ratnam, S. S. (1991a). Preimplantation genetics: Chromosomes of fragmented human embryos. Fertil. Steril. 56, 66-70. Bongso, A , , Ng, S . C., and Ratnam, S . S. (1991b). Cytogenetic evaluation of human oocytes and embryos. In "Assisted Reproductive Technology" (E. S. E. Hafez, ed.), pp. 30-40. Hemisphere Publishers, New York. Bongso, A,, Fong, C. Y., Ng, S. C . , and Ratnam, S . S. (1992a). Human ampullary cocultures for blastocyst transfer in assisted reproduction. Ann. Acad. Med. Singapore 21, 571-575. Bongso, A,, Ng, S. C., Fong, C. Y., Anandakumar, C., Marshall, B., Edirisinghe, R., and Ratnam, S. S. (1992b). Improved pregnancy rate after transfer of embryos grown in human fallopian cell coculture. F e d . Steril. 58, 569-574. Bongso, A,, Ho, J., Fong, C. Y., Ng, S. C., and Ratnarn, S. S. (1993). Human sperm function after coculture with human fallopian tuba1 epithelial cell monolayers: In vitro model for studying cell interactions in early human coception. Arch. Androl. 31, 183-190. BouB, J. G., and BouC, A. (1973). Increased frequency of chromosomal abnormalities in abortions after induced ovulation. Lancet 1, 679-680. B o d , J. G., BouC, A,, and Lazar, P. (1975). Retrospective and prospective epidemiological studies of 1500 karyotyped spontaneous human abortions. Teratology 12, 11-26. Brandriff, B., Gordon, L., Ashworth, L., Watchmaker, G . , Carrano, A,, and Worybek, A. (1984). Chromosomal abnormalities in human sperm: Comparison among four healthy men. Hum. Genet. 66, 193-201. Braude, P. R., Bolton, V. N., and Moore, S. (1988). Human gene expression first occurs between 4- and 8-cell stages of preimiplantation development. Nature (London) 332, 459-461. Buster, J. E., Bustillo, M., Rodi, I. A,, Cohen, S. W., Hamilton, M., Simon, J. A,, Thorneycroft, I. H., and Marshall, J. R. (1985). Biologic and non-biologic development of donated human ova recovered by nonsurgical uterine lavage. Am. J . Obstet. Gynecol. 153, 21 1-217. Capmany, G., and Bolton, V. N. (1993). Polypeptide profiles of human oocytes and preimplantation embryos. Hum. Reprod. 8, 1901-1905. Cha, K. Y., Koo, J. J., KO, J. J., Choi, D. H., Han, S. Y., and Yoon, T. K. (1991). Pregnancy after in v i m fertilization of human follicular oocytes collected from non-stimulated cycles, their culture in vitro and their transfer in a donor oocyte program. Ferril. Steril. 55, 109-1 13.
3. Fertilization and Development in Humans
95
Chamberlin, M. E., and Dean, J. (1990). Human homolog of the mouse sperm receptor. froc. Nail. Acad. Sci. U.S.A. 87, 6014-6018. Chen, C., and Sathananthan, A. H. (1986). Early penetration of human sperm through the vestments of human eggs in vitro. Arch. Androl. 16, 183-197. Cohen, J., Maker, H., Wright, G., Kort, H., Massey, J., and Mitchell, D. (1989). Partial zona dissection of human oocytes when failure of zona pellucida penetration is anticipated. Hum. Reprod. 4, 435-442. Cohen, J., Elmer, C . , Kort, H., Maker, H., Massey, J., Mayer, M. P., and Wiemer, K. (1990). Impairment of the hatching process following IVF in the human and improvement of implantation by assisted hatching using micromanipulation. Hum. Reprod. 5 , 7-1 3. Cohen, J., Alikani, M., Maker, H. E., Alder, A., Talansky, B. E., and Rosenwaks, Z. (1991). Partial zona dissection or subzonal insertion: Microfertilization alternatives based on evaluation of sperm and embryo morphology. Fertil. Steril. 56, 696-706. Cohen, J., Schattman, G., Suzman, M., Adler, A., Alikani, M., and Rosenwaks, 2. (1994). Micromanipulating human gametes. Reprod. Ferril. Dev. 6, 69-83. Cooper, T. G. (1 986). “The Epididymis, Sperm Maturation and Fertilization.” Springer-Verlag, Heidelberg. Cooper, T. G. (1990). In defense of a function for the human epididymis. Fertil. Steril. 54, 965975. Craft, I., McLeod, F., Green, S., Djahanbakha, 0.. Bernard, A , , and Twigg, H. (1982). Human pregnancy following oocyte and sperm transfer to the uterus. Lancet 1, 1031-1033. Cummins, J. M . , Breen, T. M., Harrison, K. L., Shaw, J. M., Wilson, L. M., and Hennessey, 3. F. (1986). A formula for scoring human embryo growth rates in in vitro fertilization: Its value in predicting pregnancy and in comparison with visual estimates of embryo quality. J . In Vitro Ferril. Embryo Trun$er 3 , 284-290. Dandekar, P. V., Martin, M. C . , and Glass, R . H. (1990). Polypronuclear embryos after in vitro fertilization. Fertil. Steril. 53, 510-515. de Jonge, C. J., Rawlins, R. G., and Zaneveld, L. J. D. (1988). Induction of the human sperm acrosome reaction by human oocytes. Fertil. Steril. 50, 949-53. de Jonge, C. J., Han, H. L., Mack, S. R., and Zaneveld, L. J. D. (1991). Modulation of the human sperm acrosome reaction by effectors of the adenylate cyclasekyclic AMP second messenger pathway. J. Exp. Zool. 258, 113-125. De Sutter, P., Dozortser, D., Cieslak, J., Wolf, G., Verlinsky, Y.,and Dyban, A. (1992). Parthenogenetic activation of human oocytes by puromycin. J. Assisr. Reprod. Genet. 9, 328-337. Dokras, A., Sargent, I. L., and Barlow, D. H. (1993). Human blastocyst grading: An indicator of developmental potential? Hum. Reprod. 8, 21 19-2127. Drisdel, R. C . , Mack, S. R., and Zaneveld, L. J. D. (1992). Purification of an acrosome reaction inhibiting glycoprotein (ARIG) from human semen. Biol. Reprod. 46 (Suppl. l), 54-60. Drobnis, E. 2. (1993). Capacitation and acrosome reaction. In “Reproductive Toxicology and Infertility” (A. R. Sialli and M. J. Zinaman, eds.), pp. 77-136. McGraw-Hill, New York. Edirisinghe, W. R., Murch, A. R., and Yovich, J. L. (1992). Cytogenetic analysis of human oocytes and embryos in an in virro fertilization programme. Hum. Reprod. 7 , 230-236. Edwards, R. G., Steptoe, P. C., and Purdy, J. M. (1970). Fertilization and cleavage in vitro of preovulatory human oocytes. Narure (London) 227, 1307- 1309. Erenus, M., Zouves, C., Rajamahendran, P., Leung, S., Fluker, M., and Gomel, V. (1991). The effect of embryo quality on subsequent pregnancy rates after in vitro fertilization. Ferril. Steril. 56, 707-710. Fishel, S., Jackson, P., Antinori, S., Johnson, J., Grossi, S., and Versaci, C. (1990). Subzonal insemination for the alleviation of infertility. Fertil. Steril. 54, 828-833. Fraser, L. R. (1993). Mechanisms controlling capacitation. 9th Annu. Meet. European Society of Human Reproduction and Embryology, Thessaloniki, Abstr. No. 37, pp. 19-20.
96
Alan Trounson and Ariff Bongso
Frydman, R., Freitas, S . , Lelaidier, C., and Olivennes, F. (1993). What are the indications for coculture in an IVF program: A one year experience. J . Assist. Reprod. Gener. 55 (Suppl. lo), w3-w5. Funahashi, H., Stumpf, T. T., Terlouw, S. L., and Day, B. N. (1993). Effects of electrical stimulation before or after in vifro fertilization on sperm penetration and pronuclear formation of pig oocytes. Mol. Reprod. Dev. 36, 361-367. Fusi, F. M., Bronson, R. A., Hong, Y., and Ghebrehiwet, B. (1991). Complement componenet C l q and its receptor are involved in the interaction of human sperm with zona-free hamster eggs. Mol. Reprod. Dev. 29, 180-188. Fusi, F. M., Vignali. M., Busacca, M., and Bronson, R. A (1992). Evidence for the presence of an integrin cell adhesion receptor on the oolemma of unfertilized human oocytes. Mol. Reprod. Dev. 31, 215-222. Gardner, D. K., and Leese, J. H. (1993). Assessment of embryo metabolism and viability. In “Handbook of In Vitro Fertilization” (A. 0. Trounson and D. K. Gardner, eds.), pp. 195-212. CRC Press, London. George, M. A., Johnson, M. H., and Howlett, S. K. (1994). Assessment of the developmental potential of frozen-thawed mouse oocytes. Hum. Reprod. 9, 130- 136. Cook, D. A., Osborn, S . M., and Johnston, W. I. H. (1993). Cryopreservation of mouse and human oocytes using I ,Zpropanediol and the configuration of the meiotic spindle. Hum. Reprod. 8, 1101-1 109. Cook, D. A., Osborn, S . M., Bourne, H., and Johnston, W. I. H. (1994). Fertilization of human oocytes following cryopreservation: Normal karyotypes and absence of stray chromosomes. Hum. Reprod. 9, 684-691. Gordon, J. W., Grunfield, L., Gamsi, G. J., Navot, D., and Laufer, N. (1989). Successful microsurgical removal of a pronucleus from tripronuclear human zygotes. Ferril. Steril. 52, 367. Gott, A. L., Hardy, K., Winston, R. M. L., and Leese, H. J. (1990). Non-invasive measurement of pyruvate and glucose uptake and lactate production by single human preimplantation embryos. Hum. Reprod. 5, 104-110. Gras, L., McBain, J . , Trounson, A. 0.. and Kola, I. (1992). The incidence of chromosomal aneuploidy in stimulated and unstimulated (natural) uninseminated human oocytes. Hum. Reprod. 7, 1396-1401. Grillo, J. M., Gamerre, M., Lacroix, O., Noizet, A , , and Vitry, G. (1991). Influence of the morphological aspect of embryos obtained by in vitro fertilization on their implantation rate. J . In Virro Ferril. Embryo Transfer 8, 317-321. Hardy, K. (1993). Development of human blastocysts in vitro. In “Preimplantation Embryo Development” (B. D. Bavister, ed.), Serono Symp., pp. 184-199. Springer-Verlag, New York. Hardy, K., Handyside, A. H., and Winston, R. M. L. (1989a). The human blastocyst: Cell number, death and allocation during late preimplantation development in vifro. Development (Cambridge, UK) 107, 597-604. Hardy, K., Hooper, M. A. K., Handyside, A. H., Rutherford, A. J., and Leese, H. J. (1989b). Non-invasive measurement of glucose and pyruvate uptake by individual human oocytes and early embryos. Hum. Reprod. 4, 188-195. Hertig, A. T., Rock, J., Adams, E. C., and Mulligan, W.I. (1954). On the preimplantation stages of the human ovum: A description of four normal and four abnormal specimens ranging from the second to fifth day of development. Contrib. Embryol. 35, 199-220. Jagiello, J., Ducayen, M., Fang, I., and Graffeo, J. (1975). Cytogenetic observations in mammalian oocytes. In “Chromosomes Today” (P. L. Pearson and K. R. Lewis, eds.), pp. 43-63. Wiley, New York. Kinloch, R. A., Mortillo, S . , Stewart, C. L., and Wassarman, P. M. (1991). Embryonal carcinoma cells transfected with ZP3 genes differentially glycosylate similar polypeptides and secrete active mouse sperm receptor. J. Cell. Biol. 115, 655-664.
3. Fertilization and Development in Humans
97
Kola, I. (1988). Embryo and fetal abnormalities in IVF. Birth 15, 145-147. Kola, I., Trounson, A. 0.. Dawson, G . , and Rogers, P. (1987). Tripronuclear human oocytes: Altered cleavage patterns and subsequent karyotypic analysis of embryos. Biol. Reprod. 37, 395-401. Lacham, O., and Trounson, A. 0. (1991). Fertilizing capacity of epididymal and testicular spermatozoa microinjected under the zona pellucida of mouse oocytes. Mol. Reprod. Dev. 29, 85-93. Lacham-Kaplan, O., and Trounson, A . 0. (1994). Micromanipulation assisted fertilization: Comparison of different techniques. In “Male Factor in Human Infertility” (J. Tesarik, ed.), pp. 287-304, Ares-Serono Symposia Publications, Italy. Laws-King, A., Trounson, A. O., Sathananthan, A. H., and Kola, I. (1987). Fertilization of human oocytes by microinjection of a single spermatozoon under the zona pellucida. Fertil. Steril. 48, 637-642. Lazendorf, S . E., Malony, M. K . , Veeck, L. L., Slusser, J., Hodgen, G. D., and Rosenwaks, 2. (1 988). A preclinical evaluation of pronuclear formation by microinjection of human spermatozoa into human oocytes. Fertil. Steril. 49, 835-842. Leese, H. I . , Hooper, M. A. K., Edwards, R. G . , and Ashwood-Smith, J. J. (1986). Uptake of pyruvate by early human embryos determined by a non-invasive technique. Hum. Reprod. 1, 181-186. Leese, H. J., Conaghan, J., and Martin, K. L. (1993). Early human embryo metabolism. Bio Essays 15, 259-264. Maker, H., and Cohen, J. (1989). Embryonic development after microsurgical repair of polyspermic human zygotes. Fertil. Steril. 52, 373-380. Martin, R. H., Balkan, W., Bums, K., Rademaker, A. W., Lim, C. C . , and Rudd, N. L. (1983). The chromosome constitution of 1000 human spermatozoa. Hum. Genet. 63, 305-309. Martin, R. H., KO, E., and Chan, K. (1993). Detection of aneuploidy in human interphase spermatozoa by fluorescence in situ hybridization (FISH), Cytogenet. Cell Genet. 64, 23-26. Mattei, J. F., Mattei, M. G., Ayme, S . , and Grand, F. (1979). Origin of the extra chromosome in trisomy 21. Hum. Genet. 46, 107-110. Maudlin, I., and Fraser, L. F. (1977). The effect of PMSG dose on the incidence of chromosomal anomalies in mouse embryos fertilized in vitro. J . Reprod. Fertil. 50, 275-280. McLachlan, R. I., Fuscaldo, G . , Claderon, I., Lacham-Kaplan, O., Poulos, C., Rho, H . , Trounson, A. O., and Wood, C. (1994). Microinjection: Choice of embryo transfer technique. Reprod. Fertil. Dev. 6, 51-61. Menezo, Y., Jany, L., and Khatchadourian, C. (1993). Embryo coculture and the regulation of blastocyst formation in vitro. In “In Vitro Fertilization and Embryo Transfer in Primates” (D. P. Wolf, R. L. Stouffer, and R. M. Brenner, eds), Serono Symp., pp. 279-289. SpnngerVerlag, New York. Mooney, J. K., Horan, A. H., and Lattimer, I. K. (1972). Motility of spermatozoa in the human epididymis. J . Urol. 108, 443-450. Moor, R. M., and Trounson, A. 0. (1977). Hormonal and follicular factors affecting maturation of sheep oocytes in vitro and their subsequent developmental capacity. J. Reprod. Fertil. 49, 101- 109. Moore, H. D. M . , Curry, M. R., Penfold, L. M., and Pryor, J. P. (1992). The culture of human epididymal epithelium and in vitro maturation of epididymal spermatozoa. Fertil. Steril. 58, 776-783. Munnt, S . , Weier, H. U. G . , Stein, J., Grifo, J., and Cohen, I. (1993). A fast and efficient method for simultaneous X and Y in situ hybridization of human blastomeres. J . Assist. Reprod. Genet. 10, 82-90. Murray, A. W. (1992). Creative blocks: Cell-cycle checkpoints and feedback controls. Nature (London) 359, 599-604.
98
Alan Trounson and Ariff Bongso
Noto, V., Campo, R., Roziers, P., Swinnen, K., Vercruyssen, M., and Gordts, S. (1993). Mitochondria] distribution after fast embryo freezing. Hum. Reprod. 8, 21 15-21 18. O’Keefe, S. J., Kiessling, A. A,, and Cooper, G. M. (1991). The c-mos gene product is required for cyclin b accumulation during meiosis of mouse egg. Proc. Natl. Acad. Sci. U . S . A . 88, 7869-7872. O’Neill, C. O., Ryan, J. P., Catt, J. W., Pike, I. L., and Krzyminska, U. B. (1994). Injection of multiple sperm into the perivitelline space as a treatment of male infertility. Reprod. Fertil. Dev. 6 , 51-56. Osborn, J. (1993). Oocyte retrieval and maturation. In “Handbook of In Vitro Fertilization” (A. Trounson and D. K. Gardner, eds.), pp. 17-32. CRC Press, Boca Raton, FL. Pal, S. K., Torry, D., Serta, R., Crowell, R. C., Seibel, M. M., Cooper, G. M., and Kiessling, A. A. (1994). Expression of potential function of the c-mos proto-oncogene in human eggs. Fertil. Steril. 61, 496-503. Palermo, G., Joris, H., Devroey, P., and Van Steirteghem, A. C. (1992a). Induction of acrosome reaction in human spermatozoa used for subzonal insemination. Hum. Reprod. 7, 248-254. Palermo, G., Joris, H., Devroey, P., and Van Steixteghem, A. C. (1992b). Pregnancies after intracytoplasmic injection of single spermatozoon into an oocyte. Lancet 340, 17-18. Papadopoulos, G., Templeton, A. A,, Risk, N., and Randale, J. (1989). The frequency of chromosome anomalies in human preimplantation embryos after in vifro fertilization. Hum. Reprod. 4, 91-98. Pellestor, F., Dufour, M. C., Amal, F., and Humeau, C. (1994). Direct assessment of the rate of chromosomal abnormalities in grade IV human embryos produced by in vitro fertilization procedure. Hum. Reprod. 9, 293-302. Plachot, M., and Crozet, N. (1992). Fertilization abnormalities in human in virro fertilization. Hum. Reprod. 7 (Suppl. I), 89-94. Plachot, M., Junca, A. M., Mandelbaum, J., de Grouchy, J., Salat-Baroux, J., and Cohen, J. (1986). Chromosome investigations in early life. I: Human oocytes recovered in an IVF programme. Hum. Reprod. 8, 547-551. Plachot, M., Junca, A. M., Maudelbaum, J., de Grouchy, J., Salat-Baroux, J., and Cohen, J. (1987). Chromosome investigations in early life. 11: Human preimplantation embryos. Hum. Reprod. 2 , 29-35. Plachot, M., Mandelbaum, J., lunca, A. M., de Grouchy, J., Salat-Baroux, J., and Cohen, J. (1989). Cytogenetic analysis and development capacity of normal and abnormal embryos after IVF. Hum. Reprod. 4 (Suppl.), 99-103. Quinn, P. (1993). Use of coculture with cumulus cells in insemination medium in human IVF. J. Assist. Reprod. Genet. 10 (Suppl.), 134 (Abstr. OC-133). Rall, W. F. (1993). Cryobiology of gametes and embryos from non-human primates. In “In Vitro Fertilization and Embryo Transfer in Primates” (D. P. Wolf, R. L. Stouffer, and R. M. Brenner, eds.), Serono Symp., pp. 223-245. Springer-Verlag, New York. Rayos, A. A,, Takahashi, Y., Hishinuma, M., and Kanagawa, H. (1994). Quick freezing of unfertilized mouse oocytes using ethylene glycol with sucrose or trehalose. J. Reprod. Fertil. 100, 123-129. Roberts, J. M., Taylor, C. T., Melling, G. C., Kingsland, C. R., and Johnson, P. M. (1992). Expression of the CD46 antigen, and absence of class I MHC antigen, on the human oocyte and preimplantation blastocyst. Immunology 75, 202-205. Rudak, E., Dor, J., Mashiach, S., Nebel, L., and Goldman. B. (1984). Chromosome analysis of multinuclear human oocytes fertilized in v i m . Fertil. Steril. 41, 538-545. Gianaroli, L., and Trounson, A. 0. (1992). Sub-zonal microinjection Sakkas, D., Lacham, 0.. (SUSM) in cases of severe male factor infertility and repeated IVF failure. Ferfil. Steril. 57, 1279- 1288. Sathananthan, A. H., and Trounson, A. 0. (1982a). Ultrasound observations on cortical granules in human follicular oocytes cultured in vitro. Gamete Res. 5, 191-198.
3. Fertilization and Development in Humans
99
Sathananthan, A. H., and Trounson, A. 0. (1982b). Cortical granule release and zona interaction in monospermic and polyspermic human ova fertilized in uitro. Gamete Res. 6, 225-234. Sathananthan, A . H., and Trounson, A. 0. (1985). The human pronuclear ovum: Fine structure of monospermic and polyspermic fertilization in uitro. Gamete Res. 12, 385-394. Sathananthan, A . H., Kola, I . , Osborn, J., Trounson, A. O., Ng, S. C., Bongso, A,, and Ratnam, s. s. (1991). Centrioles in the beginning of human life. Proc. Nail. Acad. Sci. U.S.A. 88,4806-48 10. Sathananthan, A. H., Ng, S.-C., Bongso, A,, Trounson, A. O., and Ratnam, S. S. (1993). “Visual Atlas of Early Human Development for Assisted Reproductive Technology.” National University of Singapore, Singapore. Schlesselman, J. J. (1979). How does one assess the risk of abnormalities from human in uitro fertilization? Am. J . Obstet. Gynecol. 135, 135-148. Schrurs, B. M., Winston, R. M. L., and Handyside, A. H. (1993). Preimplantation diagnosis of aneuploidy using fluorescent in situ hybridization: Evaluation using a chromosome 18-specific probe. Hum. Reprod. 8, 296-301. Scott, R. T., Hoffmann, G. E., Veek, L. L., Jones, H. W., and Muasher, S. J. (1991). Embryo quality and pregnancy rates in patients attempting pregnancy through in uitro fertilization. Ferril. Steril. 55, 426-432. Shabanowitz, R. B., and Rand, M. G. (1988). Characterization of the human zona pellucida from fertilized and unfertilized eggs. J. Reprod. Fertil. 82, 151-161. Shaw, J. M., Diotallevi, L., and Trounson, A . 0. (1991a). A simple rapid 4.5 M dimethylsulfoxide freezing technique for the cryopreservation of one-cell to blastocyst stage preimplantation mouse embryos. Reprod. Ferfil. Dev. 3, 621-626. Shaw, J. M., Kola, I., MacFarlane, D. R., and Trounson, A. 0. (1991b). An association between chromosomal abnormalities in rapidly frozen 2-cell mouse embryos and the ice-forming properties of the cryoprotective solution. J . Reprod. Ferril. 91, 9-18. Shaw, J. M., Oranratnachai, A,, and Trounson, A. 0. (1993). Cryopreservation of oocytes and embryos. In “Handbook of In Vitro Fertilization” (A. 0. Trounson and D. K. Gardner, eds.), pp. 213-236. CRC Press, Boca Raton, FL. Silber, S . J. (1989). Role of epididymis in sperm maturation. Urology 33, 47-53. Sjogren, A., Sjoblom, P., and Hamberger, L. (1992). Culture of human spare pre-embryos: Association between blastocyst formation and pregnancy. J. Assist. Reprod. Genet. 9, 41-44. Staessen, C . , Janssenswillen, C., Devroey, P., and Van Steirteghem, A. C. (1993). Cytogenic and morphological observations of single pronucleated human oocytes after in uitro fertilization. Hum. Reprod. 8, 221-223. Steer, C. V., Mills, C. L., Tan, S. L., Campbell, S . , and Edwards, R. G. (1992). The cumulative embryo score: A predictive embryo scoring technique to select the optimal number of embryos to transfer in an in uirro fertilization and embryo transfer programme. Hum. Reprod. 7, 117-122. Steptoe, P. C., Edwards, R. G., and Purdy. J. M . (1971). Human blatocysts grown in culture. Nature (London) 229, 132-133. Swam, K. (1993). The soluble sperm oscillogen hypothesis. Zygote 1, 273-276. Tarin, J. J., and Trounson, A. 0. (1993). Effects of stimulation and inhibition of lipid peroxidation on freezing-thawing of mouse embryos. Biol. Reprod. 49, 1362-1368. Tarin, J. J., Gomez, E., Sampaio, M., Ruiz, M., Remohi, H., and Pellicer, A. (1991). Cytogenetic analysis of human oocytes from fertile women. Hum. Reprod. 6, 1100-1 103. Tesarik, J. (1985). Comparison of acrosome reacting-inducing activities of human cumulus oophorus, follicular fluid and ionophore A231287 in human sperm population of proven fertilizing ability in uifro.J . Reprod. Fertil. 74, 383-388. Toth, L. T., Lazendorf, S. E., Sandow, B. A , , Veeck, L. L., Hassen, W. A , , Hansen, K., and Hodgen, G. D. (1994). Cryopreservation of human prophase I oocytes collected from unstimulated follicles. Fertil. Sreril. 61. 1077-1082.
100
Alan Trounson and Ariff Bongso
Tournaye, H., Devroey, P., Liu, J., Nagy, Z., Lissens, W., and Van Steirteghem, A. C. (1994). Microsurgical epididymal sperm aspiration and intracytoplasmic sperm injection: A new effective approach to infertility as a result of congenital bilateral absence of the vas deferens. Ferti/. Steril. 61, 1045-1051. Trounson, A. 0. (1990). Cryopreservation. Br. Med. Bull. 46, 695-708. Trounson, A. 0. (1994). The choice of the most appropriate microfertilization technique for human male factor infertility. Reprod. Ferril. Dev. 6, 37-43. Trounson, A. O . , and Jones, G. (1993). Freezing of embryos: Early vs late stages. J. Assit. Reprod. Genet. 10(3), 179- 181. Trounson, A. O . , and Moore, N. W. (1974). The survival and development of sheep eggs following complete or partial removal of the zona pellucida. J. Reprod. Fertil. 41, 97-105. Trounson, A. O., and Osborn, I. (1993b). In v i m fertilization and embryo development. In “Handbook of In Vitro Fertilization” (A. 0. Trounson and D. K . Gardner, eds.), pp. 57-84. CRC Press, Boca Raton, F L Trounson, A. O . , and Shaw, J. M . (1992). Embryo cryopreservation. Reprod. Med. Rev. 1, 179188. Trounson, A. O . , and Shaw, J . (1994). The cryopreservation of human eggs and embryos. In “Reproductive Medicine and Surgery” (Ed. Wallach and H. A. Zacur, eds.), pp. 860-868. Mosby, St. Louis. Trounson, A. O . , Leeton, J. F., Wood, C . , Webb, J., and Wood, J. (1981). Pregnancies in humans by fertilization in virro and embryo transfer in the controlled ovulatory cycle. Science 212, 681-682. Trounson, A. O . , Mohr, L. R., Wood, C . , and Leeton, J. F. (1982). Effect of delayed insemination on in vitro fertilization culture and transfer of human embryos. J . Reprod. Fertil. 64, 285-291. Trounson, A. O., Wood, C., and Kausche, A. (1994). In virro maturation and the fertilization and developmaental competance of oocytes recovered from untreated polycystic ovarian patients. Fertil. Sreril. 62, 353-362. Van Blerkom, J. (1991). In “Animal Application of Research in Mammalian Development” (R. A. Pedersen, N. First, and A. McLaren, eds.), Cold Spring Harbor Lab. Press, Cold Spring Harbor, NY. Van Duin, M., Polman, J. E., Verkoelen, C. C., Bunschoten, H., Meyerink, J. H., Olive, W., and Aitken, R. J. (1992). Cloning and characterization of the human sperm receptor ligand ZP3: Evidence for a second polymorphic allele with a different frequency in the Caucasian and Japanese populations. Genomics 14(4), 1064- 1070. Van Kooij, R . J., and te Velde, E. R. (1988). The presence of zonae pellucidae influences fusion rates between spermatozoa and zona-free hamster oocytes. Hum. Reprod. 3, 773-775. Van Steirteghem, A. C. (1994). IVF and micromanipulation techniques for male-factor infertility. Curr. Opin. Obstet. Gynecol. 6, 173-177. Van Steirteghem, A. C., Liu, J . , Joris, H., Nagy, Z., Janssenswillen, C., Tournaye, H., Derde, M.-P., Van Assche, E., and Devroey, P. (1993a). Higher success rate by intracytoplasmic sperm injection than by subzonal insemination. Report of a second series of 300 consecutive treatment cycles. Hun.Reprod. 8, 1055-1060. Van Steirteghem, A. C., Nagy, Z., Joris, H.. Liu, J., Staesen, C., Smitz, J . , Wisanto, A , , and Devroey, P. (1993b). High fertilization and implantation rates after intracytoplasmic sperm injection. Hum. Reprod. 8, 1061-1066. Veeck, L. L. (1992). Fertilization and early embryonic development. Curr. Opin. Ohstet. Gynecol. 4, 702-71 1. Veeck, L. L., Wortham, J. W. E., Jr., Witmyer, J., Sandow, B. A., Acosta, A. A,, Garcia, 1. E., Jones, G. S . , and Jones, H. W., Jr. (1983). Maturation and fertilization of morphologically immature human oocytes in a program of in vitro fertilization. F e d Steril. 39, 594-600.
3. Fertilization and Development in Humans
101
Visser, D. S., and Le Fourie, F. R. (1993). The applicability of the cumulative embryo score system for embryo selection and quality control in an in vitro fertilization/embryo transfer programme. Hum. Reprod. 8 , 1719-1722. Wada, I . , MacNamee, M. C . , Wick, K., Bradfield, J. M., and Brinsden, P. R. (1994). Birth characteristics and perinatal outcome of babies conceived from cryopreserved embryos. Hum. Reprod. 9 , 543-546. Wang, C., Lee, G. S., Leung, A , , Surrey, E. S . , and CI S . Y. W. (1993). Human sperm hyperactivation and acrosome reaction and their relationships to human in vifro fertilization. Fertil. Steril. 59, 1221-1227. Wang, X. J., Ledger, W., Payne, D., Jeffrey, R., and Matthews, C. D. (1994). The contribution of embryo cryopreservation to in vitro fertilization/gamete intra-fallopian transfer: 8 years experience. Hum. Reprod. 9 , 103-109. Wiemer, K. E., Hoffman, D. i . , Maxson, W. S . , Eager, S., Muhlberger, B., Fiore, I . , and Cuervo, M. (1993). Embryonic morphology and rate of implantation of human embryos following coculture of bovine oviductal epithelial cells. Hum. Reprod. 8, 97-101. Winston, N. J . , Braude, P. R., Pickering, S. I., George, M. A , , Cant, A,, Currie, J., and Johnson, M. H . (1991a). The incidence of abnormal morphology and nucleocytoplasmic ratios in 2, 3 and 5-day old human pre-embryos. Hum. Reprod. 6 , 17-21. Winston, N. J., Johnson, M. H., Pickering, S. J., and Braude, P. R . (1991b). Parthenogenetic activation and development of fresh and aged human oocytes. Fertil. Steril. 56, 904-912. Yanagimachi, R . (1994). Mammalian fertilization. In “The Physiology of Reproduction” (E. Knobil, I. D. Neill, C. L. Markert, G. S. Greenwald, and D. N. Pfaff, eds.), 2nd ed., pp. 189-317. Raven Press, New York. Yeung, C. H., Cooper, T. G . , Oberpenning, F., Schulze, H . , and Nieschlag, E. (1993). Changes in movement characteristics of human spermatozoa along the length of the epididymis. B i d . Reprod. 49, 274-280. Zaneveld, L. J. D., de Jonge, C. J., Anderson, R. A,, and Mack, S. R. (1991). Human sperm capacitation and the acrosome reaction. Hum. Reprod. 6 , 1265-1274. Zaneveld, L. I. D., Anderson, R. A., Mack, S. R., and de Jonge, C. J. (1993). Mechanism and control of the human sperm acrosome reaction. Hum. Reprod. 8, 2006-2008.
This Page Intentionally Left Blank
4 Determination of Xenopus Cell lineage by Maternal Factors and Cell Interactions Sally A. Moody, Daniel V. Bauer, Alexandra M . Hainski, and Sen Huang Department of Anatomy and Neuroscience Program The George Washington University Medical Center Washington, D.C. 20037 and Department of Anatomy and Cell Biology The University of Virginia Health Sciences Center Charlottesville, Virginia 22908
I. How Cell Lineages Are Studied A. Lineage Tracing Techniques B. Possible Mechanisms of Fate Determination 11. Why Study Xenopus? 111. Cell Fate Mapping in Xenopus IV. Are There Early Progenitors for Specific Tissues, Organs, or Cell v p e s ? V. Does Position in the Mitotic Pattern Determine Cell Fate? VI. Does Inheritance of a Maternal Cytoplasmic Factor Determine Cell Fate? A. Molecular Stratification of the Oocyte B. Putative Maternal Cytoplasmic Determinants C. Are There Determinants for the Dorsal Axis in Xenopus? D. Evidence for Dorsal Determinants in the Animal Hemisphere VII. Cell-Cell Signaling in Fate Determination A. Are All Blastomeres Competent to Respond to Dorsal Mesoderm Inductive Signaling? B. Are There Other Important Signaling Events During Cleavage Stages? VIII. Conclusion References
I. How Cell lineages Are Studied A. Lineage Tracing Techniques
The time at which embryonic cells become committed to a particular fate, the steps by which embryonic cells assume specific phenotypes, and the cellular mechanisms that direct these developmental decisions are fundamental issues in developmental biology. At the turn of the century embryologists, spurred on by evolutionary concerns of whether developmental forces recapitulated phylogenetic forces, described the cleavage patterns of a variety of invertebrate embryos Currenr Topics in Developmenral Biology. V d 32 Copyright 0 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
103
104
Sally A. Moody et al.
(Stent and Weisblat, 1985). The fate of individual blastomeres could be visually monitored in the transparent embryos, and the experimental question asked was whether blastomeres were equivalent or distinct in developmental potential. This research approach was limited to those species containing small numbers of cells that could be visualized with simple optics by virtue of very large cells or cytoplasmic pigments, and that produced differentiated phenotypes in larval stages with a minimum number of mitoses. A recent achievement was the complete documentation of every mitosis in a nematode, from the zygote to adult, by continuous observation of the transparent, living animals with video microscopy and differential interference contrast optics (e.g., Sulston et al., 1983). However, to perform this kind of detailed analysis in more complex animals has been a significant challenge. Often there are orders of magnitude more cells to follow than in nematodes, and the embryos that can be observed throughout early development (i.e., nonplacental) often are yolk-laden or somehow difficult to resolve optically at the cellular level. A significant breakthrough revived cell lineage studies in more complex, optically challenging animals about 15 years ago. Blastomeres were injected with small volumes of a high-molecular-weight tracer that could not pass through gap junctions, was inherited by all progeny, and could be histologically detected at later developmental time points (Weisblat et a l . , 1978). Commonly now horseradish peroxidase or fluorescently tagged dextrans are used for such tracing studies in a variety of animals. With this new technique the questions of how fate is determined during development have been reevaluated. By being able to precisely define clonal relationships at different developmental time points, it is possible to ask in cellular detail how and when cell fate is determined.
B. Possible Mechanisms of Fate Determination
In nearly every animal studied there has not been a unitary rule for fate determination. Different regions or tissues may have different rules. For example, whereas nearly all developmental events in nematodes are highly invariant and often lineage-dependent, some events require cell-cell inductions (Stent and Weisblat, 1985). Numerous studies in several species highlight four basic mechanisms that determine cell fate (see Davidson, 1990, for more detail). First, a single determinative event can cause one progenitor to make a clone composed of a homogeneous phenotype. In nematode and leech this can occur early in development (Stent and Weisblat, 1985), and in vertebrates it may occur late (see Raible and Eisen, 1994). Second, position in the mitotic pattern may determine fate. This has been elegantly demonstrated in systems in which mitoses of invariant progenitors can be visualized for long periods of time (e.g., nematode, White et d., 1982; fly, Doe and Goodman, 1985). Third, there is abundant evidence in many invertebrates that maternal determinants that are unequally distributed to differ-
4. Determination of Xenopus Cell Lineage
105
ent blastomeres can determine cell fate (Davidson, 1990; Nusslein-Volhard, 1991; Nishida, 1994). Finally, cell-cell interactions and inductions, via diffusible or nondiffusible ligand/membrane receptor complexes, gap junctional communication, or the extracellular matrix, play important roles at multiple steps in phenotype determination. This chapter will review efforts made in a vertebrate model, the South African clawed frog (Xenopus laevis), to determine by lineage tracing techniques the relative roles of these mechanisms in early blastomere determination.
11. Why Study Xenopus? Xenopus has become a very popular vertebrate model for early developmental studies for a variety of reasons. Because the eggs are laid external to the female, they are accessible for manipulations throughout development. The eggs are large (- 1.4 mm in diameter), and therefore relatively easy to manipulate with microdissection tools. For example, individual cells can be transplanted (Heasman er af., 1984; Gallagher ef al., 1991), cultured (Gallagher et a l . , 1991; Godsave and Slack, 1991), or deleted (Huang and Moody, 1992, 1993) manually. Xenopus embryos also pass from fertilization to tadpole stages very rapidly. It takes only about 30 h of incubation at room temperature from fertilization to the first muscle twitches and hatching from the vitelline membrane. In about 3 to 4 days the embryo becomes a feeding, continuously swimming tadpole. Because morphogenesis proceeds so rapidly, the outcomes of early genetic or cellular manipulations can be assayed within a few days. Unlike fish or mammalian embryos, which also are favorite models for cell lineage studies, the three cardinal axes of the early Xenopus embryo can be recognized by the first cleavage. The animal-vegetal axis, which will be transformed into the anterior-posterior axis at gastrulation, can be identified in unfertilized eggs. The animal hemisphere is characterized by melanin granules in the actin cortex subjacent to the plasma membrane; this gives the animal hemisphere a dark pigmentation, whereas the vegetal hemisphere appears yellowish-white. Many organelles are distributed unequally between the animal pole and the vegetal pole, including different sized yolk platelets, whose distribution probably accounts for the buoyancy of the eggs and cleavage embryos such that when they are removed from their jelly coat, but are still within the vitelline membrane, they orient with the animal pole facing away from gravity. The dorsal-ventral axis can be identified shortly after fertilization. The future dorsal region will be located within a 30" arc by the direction of rotation of the pigmented cortical cytoplasm toward the ventral animal region after sperm entry (Vincent er al., 1986; see Section VI, below). This rotation causes a pale wedgeshaped area in the equatorial region of the animal hemisphere, called the grey crescent, to become distinct in most embryos. In about 70%of different amphib-
106
SalIy A. Moody et al.
ian embryos the grey crescent is bisected by the first cleavage furrow (Jenkinson, 1909; Schechtman, 1935; Kirschner et al., 1980; Klein, 1987). In these embryos the grey crescent indicates the future dorsal side. Using lineage tracers in naturally fertilized eggs, Klein (1987) also reported that the first cleavage furrow identifies the sagittal plane even in those embryos (30%) in which the furrow is not perpendicular to the grey crescent. Although this latter finding has been disputed (Danilchik and Black, 1988), independent confirmation was provided in a study that repeated the two different methods of analysis (Masho, 1990). Injection of lineage tracer into one blastomere at the two-cell stage, regardless of the position of the grey crescent, resulted in a <30” angular separation between the first cleavage plane and the dorsal midline of the gastrula in 94% of embryos; this was a significantly better predictor of the dorsal midline than originally reported (Danilchik and Black, 1988). When embryos were analyzed as in the Klein (1987) report, Masho found that the first cleavage furrow was within 30” of the dorsal midline in greater than 90% of the embryos. Our early fate maps, which used only embryos in which the first cleavage furrow bisected the grey crescent, are consistent with Masho’s (1990) findings. Injecting the same blastomere in a large number of naturally fertilized embryos gave the same lineage pattern in over 90% of the cases (Moody, 1987a,b, 1989; Moody and Kline, 1990). These studies demonstrate that the third cardinal axis can be identified as soon as the egg begins to divide. The midsagittal plane, most accurately defined by the first cleavage furrow (Klein, 1987; Masho, 1990), separates right and left sides. By selecting embryos at the two-cell stage whose first cleavage is exactly in the middle of the egg, and bisects the gray crescent so that half of it is in the right blastomere and half is in the left, one can define the three cardinal axes in greater than 90% of embryos. Lineage studies, fate mapping, and tests of fate determination of blastomeres have been most effective in species in which all eggs divide in identical patterns. In these cases identifiable progenitors can be marked in large numbers of embryos. However, this regularity is not found in most vertebrate embryos; cleavage patterns in mouse, fish, chick, and most amphibia are notably irregular. In Xenopus clutches of over 1000 eggs, one can, however, find many that cleave in nearly identical patterns. In about 70% of embryos the first cleavage furrow bisects the gray crescent (Kirschner er al., 1980; Klein, 1987), and there are a few common cleavage patterns through the 32-cell stages (Jacobson, 1981a). By choosing embryos for lineage studies in which the three cardinal axes are accurately identified, and that conform to one of these cleavage patterns (X-type; Fig. l), we have been able to study and manipulate the invertebrate equivalent of an “identified’ cell in a vertebrate embryo. This ability has allowed us to decrease significantly variability between embryo qualitative and quantitative maps (Moody, 1987a,b, 1989), and to identify subtle changes in fate after single-cell manipulations (Gallagher et al., 1991; Huang and Moody, 1992, 1993).
Ftg. 1 Side views of embryos from the 2-cell (left) to 32-cell (right) stages. At 2 cells the cleavage furrow is in the plane of the page and only the left blastomere is seen. At 4 cells the second cleavage furrow divides each side into a dorsal (D) and ventral (V) blastomere. Embryos that cleaved according to the patterns depicted in the top row were used for lineage tracing studies. The nomenclature shown was devised by Jacobson (Hirose and Jacobson, 1979; Jacobson and Hirose, 1981). In the bottom row is shown an unusual cleavage pattern at the 8-cell stage. In these embryos the third cleavage is radial rather than equatorial, giving blastorneres an “orange slice’’ appearance. At the next cleavage (16-cell) these embryos are indistinguishable from stereotypic ones (cf. above), but the stippling indicates the position of the descendants of one such radial blastomere (D. lr). The 32-cell embryo in the bottom row is labeled with a commonly used nomenclature that is applicable only to the 32-cell stage (Nakamura and Kishiyama, 1971).
108
Sally A. Moody et al.
A final advantage of Xenopus, for the purpose of this discussion, is that zygotic activation of the genome does not occur until 6 to 7 h after fertilization, at a time called the midblastula transition (MBT; Newport and Kirschner, 1982a). No gene transcription has been detected prior to the MBT (Newport and Kirschner, 1982b); the cleavage stages utilize proteins and mRNAs that were synthesized during oogenesis and then stored. This developmental strategy also is used in the mouse, but due to the long duration of the cell cycle, new gene transcription begins much earlier between the two- and eight-cell stages (Johnson, 1981; Clegg and Piko, 1977). Because in Xenopus there is a prolonged period directed by maternal molecules, investigation into the identity, localization, and action of these molecules is expected to add significantly to our understanding of the role of inherited genetic information in important interactions that occur prior to the onset of zygotic transcription. There are several examples in invertebrate species of localized maternal molecules dictating cell fate decisions in the blastomeres that inherit the molecules (Davidson, 1990; Nusslein-Volhard, 1991; Nishida, 1994). Xenopus is an ideal vertebrate in which to search for such molecules because early embryos utilize stored maternal molecules and can be selected with nearly identical cleavage patterns, and thus any maternal localized determinants should be parcelled into separate identifiable lineages. The studies described in the following sections have been possible because of these several unique characteristics of Xenopus development.
111. Cell Fate Mapping in Xenopus Extensive fate maps, based on intracellular lineage dye injections, for different cleavage stage embryos have been reported by a number of laboratories (Jacobson and Hirose, 1981; Hirose and Jacobson, 1979; Jacobson, 1983; Moody, 1987a,b, 1989; Moody and Kline, 1990; Dale and Slack, 1987; Masho and Kubota, 1986; Masho, 1988). Our laboratory found that when the same stereotyped blastomere was injected with lineage tracer in a large number of embryos, its fate was quite predictable (Moody, 1987a,b; 1989; Moody and Kline, 1990). Consistency in clone composition was greater than 90% in each group of embryos having the identical blastomere marked. These results demonstrate that although fate is not identical across embryos, as it is in the nematode (Sulston et al., 1983) or ascidians (Nishida, 1994), it is highly predictable when using embryos in which the cardinal axes are identified at the two-cell stage and the cleavage patterns are nearly identical. This is in contrast to results from zebrafish (Kimmel and Law, 1985) or mouse (Kelly, 1977; Ziomeck et al., 1982; Johnson and Ziomeck, 1981), in which the fates of early blastomeres seem to be randomly assigned. The predictability of fates in Xenopus, and the relative ease of experimentally manipulating the embryo, allows one to test how and when fate is determined during development.
4. Determination of Xenopus Cell Lineage
109
IV. Are There Early Progenitors for Specific Tissues, Organs, or Cell Types? During the early cleavages of nematode some progenitors become restricted to a specific tissue (e.g., endodermal lineage of E blasts; Laufer et al., 1980) or cell type (e.g., germ cell lineage of P4 blasts; Wood et nl., 1983). During mouse development the lineages of the inner cell mass, which will form the embryo, and of the trophectoderm separate between the 8- and the 32-cell stages (Balakier and Pedersen, 1982; Pedersen et al., 1986), coincident with the onset of expression of two embryo-specific transcription factors (Oct-4 and Oct-6; Kessel and Gruss, 1990; Scholer, 1991). It is reasonable to predict that if developmental points are found when blast cells produce only one tissue, organ, or cell type, then at around that time a determinative event has occurred. An early objective of our fate mapping studies was to investigate whether such events occurred in the frog at cleavage stages. The fate maps made it clear that all cleavage stage blastomeres (through 32 cells) give rise to endoderm, mesoderm, and ectoderm, demonstrating that primary germ layer restriction does not occur early. This is in contrast to nematode (Sulston et al., 1983) and ascidians (Nishida, 1994), and to the early frog fate maps made with surface vital dye staining (Nakamura and Kishiyama, 1971). Although not every blastomere contributed to every organ, no organ descended from a single progenitor (Moody, 1987a,b). By mapping the fates of each blastomere in 2- to 32-cell embryos and placing these data on a true lineage map (Moody and Kline, 1990), it was possible to illustrate that some apparent restrictions in fate do occur during cleavage stages, but that no organ system was monoclonal in origin. For example, in the ectodermal lineage (Fig. 2A), cement gland, lens, olfactory placode, cranial ganglia, and otocyst are derived from animal (Dl, V1) blastomeres but not vegetal (D2, V2) blastomeres at the 8-cell stage. However all the daughters and granddaughters of Dl and Vl each contribute to all these structures. Likewise, in the CNS lineage (Fig. 2B), rostra1 CNS structures (retina, forebrain, and midbrain) are restricted to the D cell as early as the 4-cell stage, and further restricted to its anterior daughter (Dl) at the next cell division. But at the following two cell divisions all daughters and granddaughters contribute to all these structures. These studies clearly demonstrate that at early stages germ layer, organ, and tissue fate are not restricted to a single blast cell. We reasoned that perhaps if such a restriction occurred in a complex vertebrate, it would only be seen in a specific phenotype. Tbming to the nervous system, in which there are many neuronal phenotypes that can be morphologically or neurochemically identified and quantified, we have found no evidence for a clonal origin of several neuronal types. Primary motoneurons (PMN) and Rohon-Beard neurons (RBN) each descend from a large number of 32-cell blastomeres (Moody, 1989). For PMN there are five major and three minor progenitors on each side of the embryo, and
110
Sally A. Moody et al.
for RBN there are three major and seven minor progenitors. Similarly, each GABAergic nucleus along the length of the CNS derives from multiple progenitors (Moody and Kersey, 1989). Since these three neuronal types are found in all segments along the length of the spinal cord, it seemed possible that their extensively polyclonal origin in early stages was related to their extensive axial distribution. However, spatially restricted neuronal phenotypes also descend from multiple blastomeres. Dopamine, Neuropeptide Y, GABA, and glycine amacrine cells of the retina each descend from multiple 32-cell blastomeres (Huang and Moody, 1994, 1995). Even a very small dopaminergic nucleus in the hypothalamus, which is spatially restricted to a small region of the brain and contains an average of only 33 cells at early tadpole stages, descends from as many as 7 of the 32 blastomeres (both ipsilateral and contralateral; Huang and Moody, 1992). In short, we have looked for lineage restrictions in primary germ layers, tissues, organs, and specific neuronal phenotypes during the first five cleavage cycles and have found none that might be considered determinative. This finding is consistent with studies that have shown by transplantation techniques that an embryonic cell produces a single tissue only later in development. For example, Heasman et al. (1984) showed that single vegetal pole cells from cleavage and early stages can contribute to all germ layers; however, by early gastrula stages these cells are determined to contribute only to the endoderm. Similar late tissuetype restrictions occur during gastrulation in zebrafish (Kimmel and Warga, 1986).
V. Does Position in the Mitotic Pattern Determine Cell Fate? One of the key observations in nematode lineage studies was that position within the mitotic pattern itself could determine the fate of a cell (Sulston et al., 1983), and lineage-specific genes that regulate the mitotic pattern have been identified (Ambros and Horvitz, 1985; Chalfie et al., 1981). Indeed, position within the
Fig. 2 Lineage diagrams of the relative contribution of each cleavage blastomere to the non-CNS ectodermal structures (A) and the CNS structures (B).At each cleavage the percentage that each sister cell contributed to each organ was compared, using the relative numerical values of Moody and Kline (1990). If one sister contributed 85% or more of the total value of an organ at a particular cleavage, the name of the organ appears in red in the predominant contributor and is a dash in the sister. If one sister contributed 60-84% of the total value, the name of that organ appears in blue in the major contributor and in orange in the minor contributor. If the values of both sisters fall in the 40-60% range they are considered to be equal contributors and the name of the organ appears green in both. Abbreviations: cement, cement gland; crgang, cranial ganglia, h epi, head epidermis; h ncr, head neural crest; olfact, olfactory placode; t epi, trunk epidermis; t ncr, trunk neural crest; sp cord, spinal cord.
This Page Intentionally Left Blank
4. Determination of Xenopus Cell Lineage
111
mitotic pattern of neuronal precursors distinguishes one type of motoneuron from another (White et al., 1982). Does such a “lineage” mechanism exist in vertebrates? This has been a difficult question to answer because one really needs to monitor every cell division from a blast cell to a differentiated phenotype, a process that is protracted in vertebrate embryos and involves very large numbers of cells. Further, if cell cleavages are not regular, so that the cytoplasmic parcel of egg cytoplasm that defines a blastomere is identical from one embryo to the next, the one cannot isolate the lineage aspect of fate determination. The stereotyped cleavages of Xenopus, however, have allowed a direct test of the possible role of position in the mitotic pattern. Typically the first cleavage furrow separates right and left halves (Klein, 1987; Masho, 1990), the second furrow separates dorsal and ventral halves, and the third furrow usually occurs at the equator to separate animal from vegetal hemispheres (Masho, 1988; Fig. 1). In a few embryos, however, the third furrow is radial rather than equatorial, resulting in an 8-cell embryo composed of blastomeres that resemble orange slices (bottom, Fig. 1). In these cases the equatorial furrow forms at the next cell division, so that the 16-cell stage embryos in which the third cleavage furrow was radial versus equatorial are indistinguishable. This unusual pattern allows a comparison of the clones of cells that are equivalent in position in the mitotic tree (Fig. 3) but occupy different space in the embryo, and thereby contain different regions of the egg cytoplasm (Fig. 1). For example, the dorsal animal blastomere of the stereotyped 8-cell embryo (Dl) occupies the same lineage position as the dorsal midline radial blastomere (D. lr), but occupies the space equivalent to the two dorsal animal 16-cell blastomeres (Dl. 1 plus D1.2). D. l r occupies the space equivalent to D1.l plus the vegetal blastomere, D2.1. If position in the mitotic
& A
D1.l
Dj.2 021
D2.2
* m a -
A A -D2.l- D1.2 D22
D11
Fig. 3 Mitotic patterns from the 4- to 16-cell stages of the dorsal cells in embryos with stereotypic cleavages (top; cf. top of Fig. 1) and in embryos with radial cleavages (bottom; cf. Fig. 1). Note that while D1 and D. lr occupy the same place in the D lineage, they have different daughters (underlined) that occupy different positions in the 16-cell embryo.
Sally A. Moody et al.
112
pattern determines fate, then the clones of D1 and D. Ir should be equivalent, but if position is the determining factor, then the clone of D. l r should be the same as the clones of D l . 1 and D2.1 combined. Analysis of these embryos showed that most organs were populated by progeny of D.lr in a manner indistinguishable from that of the combination of D1.1 and D2.1, i.e., according to physical position. The same was true for the progeny of V.lr. To enable a more quantitative comparison, the number of spinal neuron progeny, specifically PMN and RBN, was counted in each clone. The mean number of PMN derived from D. lr (48.8) was indistinguishable from the value obtained by summing the number derived from D1.1 and from D2.1 (473, and quite different from the mean number derived from stereotypic DI (79). The mean number of RBN derived from V. lr (10.3) was indistinguishable from the value obtained by summing the number derived from VI.1 and from V2.1 (10.9), and quite different from the mean number derived from stereotypic V 1 (34.8). These data demonstrate directly that position within the mitotic pattern at cleavage stages is not the important factor in fate determination. However, this study examined only a short segment of the lineage. Does this conclusion hold for the next several days and ,weeks of development? This is a difficult question to answer because technically it is rare that one can monitor cell divisions directly and accurately in the very large vertebrate clones. Although one can label clones later in development with either intracellular dye injection (e.g., Holt et al., 1988; Wetts and Fraser, 1988) or retroviral markers (e.g., Gray et al., 1988; Turner and Cepko, 1987), these studies provide information on clone composition (fate maps), not lineal histories. Therefore they cannot address whether birth rank influences cell fate. A few recent studies are consistent with the idea that some aspect of the cell cycle or mitotic rank may influence cell fate. In an examination of the terminal divisions of neural plate cells giving rise to Xenopus primary neurons, Hartenstein (1989) demonstrated that the last division gave rise to clones of the same fate in greater than 70% of the cases. Whether this last mitosis is determinative needs to be experimentally tested. In zebrafish a direct lineage analysis of trunk dividing neural crest progenitors in situ showed that in the majority of cases type-restricted precursors (precursors that then produced a small clone of only Schwann cells or sensory neurons, etc.) were generated one to three cell cycles prior to the terminal mitosis (Raible and Eisen, 1994). These results suggest that neural crest cells become lineage-restricted late in the mitotic tree. Without experimental manipulations we do not know whether these late lineage patterns are the result of intrinsic expression of lineage genes or of position-specific environmental influences. It is clear that the issue of position versus lineage needs to be evaluated more closely in vertebrates, but the technical difficulties of doing the right experiment are formidable. '
4. Determination of Xenopus Cell Lineage
113
VI. Does Inheritance of a Maternal Cytoplasmic Factor Determine Cell Fate? Although in the preceding section we have argued that fate is not restricted to a specific germ layer, tissue, organ, or neuronal phenotype until close to gastrulation, there are several patterning events that regionalize the Xenopus embryo prior to the establishment of these more differentiated states. There is a large body of evidence indicating that a prepattern of molecules is established during oogenesis that regionalizes the cell at the molecular level, and emerging literature indicates that some of these molecules play a role in the early steps of fate determination. Before describing our studies addressing these issues, it is important to review the evidence that Xenopus embryos might be molecularly stratified, might contain maternal determinants, and might utilize these determinants to establish the dorsal axis. A. Molecular Stratification of the Oocyte
At the very earliest stages of oogenesis the Xenopus oocyte is polarized along a line that passes through the nucleus and the centrosome (reviewed in Hausen and Riebesell, 1991). The oogonium is stratified with the nucleus toward one pole, the Golgi apparatus toward the other, and a mass of mitochondria and the centrosome lying between. This stratification apparently has no relationship to gravity, ovarian structures, or follicle cells, suggesting that the polarity arises intrinsically. This polarity is maintained in the oocyte stages, during which there are two cytoplasmic reorganizations that occur prior to fertilization that further stratify this cell. By the time oocytes reach stage IV the mitochondria1 mass has migrated toward the future vegetal pole; yolk platelets become unevenly distributed, such that the largest ones are in the vegetal hemisphere (Danilchik and Gerhart, 1987); and the germinal vesicle, containing the nucleus, has moved from the center of the cell into the animal hemisphere. The cortex of the oocyte, an actin-filled cytoplasmic zone subjacent to the plasma membrane and containing pigment granules, thins in the vegetal hemisphere so that by stage IV the vegetal half is unpigmented and the animal half is darkly pigmented. At the time the oocyte matures to an egg capable of being fertilized, other cytoplasmic rearrangements that result in further animal-vegetal differences occur. The germinal vesicle, the oocyte nucleus that is swollen with clear “nucleoplasm,” becomes leaky and this material, which is rich in RNA (Imoh, 1984), forms a yolk-free zone of cytoplasm at the basal side of the nucleus (Herkovits and Ubbels, 1979). The nucleus, shrinking in size, migrates to the animal pole where it will extrude the first polar body. The microtubular system moves with the nucleus and the nucleoplasm dissipates into the surrounding cytoplasm of the
114
Sally A. Moody et al.
animal hemisphere. As the nucleus migrates to the animal pole, strands of deep vegetal yolk platelets seem to be dragged into the central core of the animal hemisphere. Thus, during oogenesis and maturation the unfertilized egg becomes highly stratified, establishing a gradient of organelles and perhaps determinants, along the animal-vegetal axis.
B. Putative Maternal Cytoplasmic Determinants A role for localized cytoplasmic molecules in the determination of cell fate has been suggested in numerous developing systems (Davidson, 1990), and demonstrated at the molecular level in establishing both the anterior-posterior and the dorsal-ventral axes in Drosophila (Nusslein-Volhard, 1991). In ascidians specific egg cytoplasm, which is reorganized and localized during oogenesis, contains factors for muscle, endoderm, and epidermal lineages (reviewed in Nishida, 1994). Are there examples of localized material that may influence cell fate in Xenopus? Like nematode (Strome, 1992) and Drosophila (Mahowald, 1992), Xenopus contains a cytoplasmic material (germ plasm) that determines the germ cell lineage (reviewed in Ressom and Dixon, 1988). This cytoplasm contains electrondense granules that are rich in RNA and protein. In the early oocyte the germ plasm collects in the mitochondrial mass and is translocated to the vegetal pole with that mass. Removal of the material depletes germ cells, and injection of extra amounts of the material results in the differentiation of extra germ cells. Transfer of germ plasm from wild-type embryos into embryos made deficient in germ cells by UV-irradiation restores the germ line. Although the complete molecular constituents of germ plasm has not yet been elucidated, the material does contain an RNA homologous to Drosophila vasa (Komiya et al., 1994), whose product in the fly is known to regulate germ line-specific genes. There are a few specific mRNAs that have been localized to either the animal or the vegetal hemispheres of Xenopus oocytes and eggs. Many of the animalspecific maternal mRNAs have been only partially characterized. Anl-3 are localized to the animal region of oocytes, eggs, and embryos (Rebagliati et al., 1985). An1 encodes for a ubiquitin-like protein (Linnen et al., 1993) whose function is not yet known. An2 encodes for the alpha subunit of mitochondria1 ATPase, and its depletion results in arrested development after gastrulation (Dagle et al., 1990). According to sequence homology, the An3 protein probably is an ATP-dependent RNA helicase (Gururajan el al., 1991), which may have a role in bringing maternal RNA to a translatable conformation. Two less wellcharacterized molecules promise to be interesting. xlan4 is localized in the animal pole region of mature oocytes, and after MBT is limited to dorsal axial structures, especially CNS (Reddy et al., 1992). This localization has suggested a role in the specification of neuronal phenotypes, but an experimental test of this
4. Determination of Xenopus Cell Lineage
115
hypothesis has not yet been reported. Finally, x121, a partial cDNA, is localized to the animal hemisphere of oocytes and is found in an animal-to-vegetal gradient during cleavage stages (Kloc et a l . , 1991). Several of the known vegetal-specific maternal mRNAs seem to play a role in mesoderm induction, posterior axis formation, or RNA translocation. Vg-1 is localized to the vegetal pole during oogenesis (Rebagliati et a l . , 1985), is released into the vegetal cytoplasm after fertilization, and is confined to the vegetal half of the embryo by cleavage furrows (Melton, 1987; Weeks and Melton, 1987). The Vg-1 protein is a member of the TGF-P superfamily with homology to activin (Weeks and Melton, 1987). Recent protein expression studies have shown that the Vg- 1 product is activated by a proteolytic cleavage; the cleavage product induces animal cap ectoderm to form dorsal mesoderm and restores a dorsal axis to UV-ventralized embryos (Thomsen and Melton, 1993; Vize and Thomsen, 1994). mRNA for Xwnt-1 1, a member of the winglesslint gene family, has a localization pattern very similar to that of Vg-1 (Ku and Melton, 1993). During morula stages it is confined to the vegetal cytoplasm but after MBT it is at highest concentration in the dorsal marginal zone, suggesting that it also has a role in mesoderm inductive signaling. This putative role is substantiated by the fact that injection of XWnt-11 RNA into dorsal deficient embryos allows them to express some dorsal tissues (Ku and Melton, 1993). Two other maternal RNAs, Xcat2 and Xcat3, are found specifically bound to the vegetal cortex cytoskeletal domain, along with Vg-1 (Elinson et a l . , 1993). By sequence homology Xcat2 is thought to be an RNA binding protein with close similarity to nanm (Mosquera et a l . , 1993), a Drosophila protein involved in determining the posterior axis (Nusslein-Volhard, 1991). Finally, Xlsirts RNA becomes associated with the mitochondria1 mass and germ plasm early in oogenesis, translocates to the vegetal cortex coincident with the germ plasm, and seems to play a role in either translocating or binding other RNAs to the vegetal cortex (Kloc et a l . , 1993; Kloc and Etkin, 1994).
C. Are There Determinants for the Dorsal Axis in Xenopus?
Although the Xenopus oocyte appears to be radially symmetric about the animal-vegetal axis, dorsal-ventral differences arise at fertilization. The grey crescent arises in a region opposite the point at which the sperm enters the egg. Recent studies suggest that the sperm aster provides an asymmetry to the assembly of microtubules in the vegetal cortex (Houliston and Elinson, 1991a), and this vegetal array of microtubules causes a rotation of the cytoplasm adjacent to the cortex of the egg with respect to the deeper cytoplasm (Elinson and Rowning, 1988; Zisckind and Elinson, 1990; Schroeder and Gard, 1992). The rotational movement is thought not to be driven by microtubule polymerization itself, but by mechanochemical enzyme “motors” such as kinesin (Houliston and Elinson,
116
Sally A. Moody et al.
1991b; Houliston, 1994). This cortico-cytoplasmic rotation appears to be neces-
sary for normal dorsal axis formation (Vincent and Gerhart, 1987; Elinson and Rowning, 1988; Wakahara, 1989; Gerhart et al., 1989). Cortical rotation can be blocked by UV irradiation of the vegetal pole during the first cleavage cycle or by treatments that depolymerize microtubules, which in turn prevent dorsal axis formation (Scharf and Gerhart, 1983; Elinson and Rowning, 1988; Vincent et al., 1986). Manually tipping eggs so that the cortical rotation occurs in response to gravity can rescue dorsal axis formation (Vincent et al., 1986; Zisckind and Elinson, 1990). Several lines of evidence suggest that cortical rotation helps to localize or activate cytoplasmic determinants on the prospective dorsal side (reviewed in Elinson and Kao, 1989; Gerhart ef al., 1989; Wakahara, 1989). For instance, the cytoplasmic rearrangement during cortical rotation consolidates the yolk-free, RNA-rich cytoplasm from the germinal vesicle in the dorsal animal hemisphere (Herkovits and Ubbels, 1979; Imoh, 1984). Cortical rotation also produces a “swirl” of deep and cortical yolk platelets that is unique to the dorsal side of the embryo (Danilchik and Denegre, 1991). Although the formation of this swirl typically occurs coincident with the rotation of the deep vegetal yolk mass (Denegre and Danilchik, 1993), when this rotation is prevented the swirl still forms (Brown et al., 1993). It has been postulated that in normal development the sperm aster initiates the dorsal cytoplasm reorganization and that the rotation of the vegetal yolk mass enhances it so that some of the animal cytoplasm will extend into dorsal tier-3 blastomeres. These studies show that dorsal “swirl” material is contained in dorsal midline blastomeres of the cleavage embryo, and several manipulations of those blastomeres suggest that they contain “dorsal determinants.” The removal of dorsal equatorial cytoplasm (Wakahara, 1986) or dorsal equatorial blastomeres (Gimlich, 1986; Takasaki, 1987; Gallagher et al., 1991) results in embryos with defective dorsal axes. Transplanting dorsal blastomeres to the ventral midline causes a secondary dorsal axis to form (Kageura and Yamana, 1986; Gimlich, 1986; Takasaki and Konishi, 1989; Kageura, 1990; Gallagher et al., 1991), and transplanting dorsal blastomeres into dorsal-deficient UV-irradiated embryos restores their dorsal axis (Gimlich and Gerhart, 1984; Gimlich, 1986). Finally, cytoplasm from a dorsal vegetal blastomere can induce a secondary dorsal axis when injected into a ventral vegetal blastomere (Yuge et al., 1990). This inducing activity moves from the vegetal pole of the oocyte to the dorsal vegetal equatorial region (equivalent of a tier-3 blastomere) coincident with cortical rotation (Fujisue et al., 1993); treatments that block cortical rotation also block the movement of this inducing activity from the vegetal pole. Holowacz and Elinson (1993) further showed that the dorsal axis “activity” is contained in cortical, rather than deep, cytoplasm and that this activity can be destroyed if the vegetal pole of prophase I oocytes is UV-irradiated. This early UV treatment has been hypothesized to destroy dorsal determinants (Elinson and Pasceri, 1989),
4. Determination of Xenopus Cell Lineage
117
whereas UV-irradiation after fertilization prevents cortical rotation (Vincent et a l . , 1986). Thus, it is thought that dorsal cleavage stage blastomeres contain informational molecules, which are maternally synthesized and activated at about the time of fertilization, whose functions are to instruct recipient blastomeres to manifest a dorsal axial fate. The molecular nature of such determinants is not known. Dorsal and ventral blastomeres synthesize a different repertoire of proteins during cleavage stages (Miyata el a l . , 1987; Klein and King, 1988), some of which are altered after treatments that change dorsal axis formation, and there is immunohistochemical evidence that a 70-kDa maternal protein is enriched on the dorsal side at cleavage stages (Suzuki et al., 1991). However, none of these proteins has been identified, and although the differential distribution of proteins in the early embryo is likely to be due to differential localization of maternal mRNAs, such transcripts have not yet been reported. Two possible candidates for the dorsal axis-inducing activity in the vegetal hemisphere are Vg- 1 and Xwnt-1 I , and both of these seem to be involved in dorsal mesoderm induction (Thomsen and Melton, 1993; Ku and Melton, 1993). Although Vg-1 mRNA is found in both dorsal and ventral vegetal regions (Melton, 1987), and the embryonically synthesized protein is present in all vegetal cells (Tannahill and Melton, 1989), it has been suggested that localized post-translational cleavage of the Vg- 1 precursor protein occurs only on the dorsal side of the embryo (Thomsen and Melton, 1993; Vize and Thomsen, 1994). If true, this would be similar to the localized proteolytic cleavage of the S p a d e protein in Drosophila that is necessary for establishing the dorsal-ventral axis (Schneider el af., 1994). One possible important consequence of the movement of dorsal axis-inducing activity from the vegetal pole to the dorsal equator (Fujisue et al., 1993) and of “swirl” cytoplasm from the animal hemisphere to the dorsal equatorial region (Brown et al., 1993) is the establishment of the so-called “Nieuwkoop center” of dorsal mesoderm-inducing activity. Classic studies by Nieuwkoop ( 1973; Sudarwati and Nieuwkoop, 1971) demonstrated that in culture, vegetal pole cells can induce animal cells (which would express only ectodermal characteristics in isolation) to express mesodermal characteristics. Many studies indicate that an activin-like molecule may induce mesoderm with dorsal identity (e.g., notochord), whereas bFGF may induce mesoderm with ventral identity (e.g., blood) (Kimelman e f al., 1992; Dawid, 1994; Kessler and Melton, 1994). A simple model is that these two signals, arising from vegetal cells (the dorsal cells constituting the Nieuwkoop center) and spreading toward the equatorial and animal cells, act independently to produce dorsal and ventral mesoderm, and that a third signal, arising from the dorsally located Nieuwkoop center and spreading ventrally, acts to induce intermediate mesoderm (e.g., somites, nephric mesoderm; Slack, 1991). Recent studies indicate that the process is more complex, and that synergy between FGF and activin-like signaling probably accounts for the diversity in mesodermal patterning (Cornell and Kimelman, 1994a,b; Wilson
118
Sally A. Moody et al.
and Melton, 1994). Nonetheless, these interactions have been shown to begin in Xenopus at about the 64-cell stage, to require about 90 min of contact, and to be essentially completed by the onset of zygotic transcription (Jones and Woodland, 1987). Thus, mesoderm induction is an important step in specifying dorsal fate, and relies on interactions of maternal molecules that will eventually affect dorsalspecific gene expression. So far the only candidate mRNAs for the activity of the Nieuwkoop center are Vg-1 and Xwnt-1 I . As discussed above, the mRNAs are not restricted to that region, but it is possible that the active forms of their proteins are. This possibility is interesting in light of the observation that the dorsal-inducing activity in the vegetal hemisphere can cause axis specification only if it is located in the dorsal equatorial region (Fujisue et al., 1993). Other members of the TGF-P superfamily, activins A and B, also play important roles in dorsal mesoderm formation, but their mRNAs are not in the right place at the right times. Activin PA mRNA is not detectable until gastrulation (Thomsen et al., 1990), and activin PB mRNA is not detectable until the midblastula transition, at which time it is found in all regions of the embryo (Dohrmann et al., 1993). However, both mRNAs are found in the follicle cells surrounding the oocyte (Dohrmann et al., 1993; Rebagliati and Dawid, 1993), and all three activin proteins (A, AB, B) are present in the early cleavage stage blastomeres (Asashima et af., 1991; Fukui et al., 1994). We do not know whether these maternal proteins are localized to particular regions of the embryo, but it has been suggested that they could be locally inactivated by complexing with an abundant maternal pool of follistatin, a protein that binds activin (Fukui et al., 1994). Identifying the exact molecular determinant that identifies the Nieuwkoop center, the maternal signaling and responsive elements that give this center its important role in dorsal mesoderm induction, and determining how these interactions lead to the formation of Spemann’s Organizer, a region that arises after MBT in the dorsal vegetal equatorial region, are important areas of continuing investigation.
D. Evidence for Dorsal Determinants in the Animal Hemisphere
Several years ago our attention was drawn to the dorsal animal, rather than vegetal, region. The fate maps and lineage diagrams we had constructed showed that a 16-cell dorsal animal blastomere (Dl. 1 of the Hirose and Jacobson, 1979 nomenclature; Fig. 1) is a major progenitor of dorsal mesoderm (notochord) and CNS (Fig. 2; Moody, 1987a, 1989), and many of its progeny move prior to dorsal lip formation to populate the Spemann’s Organizer region in the early gastrula (Bauer et al., 1994). This latter region is composed of the dorsal mesodermal cells that are responsible for the convergent-extension movements that drive the gastrulation movements of involution and elongation, and that induce the overlying ectoderm to form neural plate (Keller et ul., 1985; Keller, 1986).
4. Determination of Xenopus Cell Lineage
119
The D1.l blastomere also contains a great deal of the yolk-free, RNA-rich cytoplasm (Herkovits and Ubbels, 1979; lmoh, 1984) and the dorsal “swirl” material (Danilchik and Denegre, 1991). We reasoned that this area might contain cytoplasmic moieties involved in the initial steps of dorsal axis specification. To test our hypothesis we first investigated the autonomy of the differentiation program of D 1.1. Three different manipulations were performed at the 16-cell stage because this time point is thought to be 1 h prior to the onset of mesoderm induction (Jones and Woodland, 1987). Thus, if evidence for an intrinsic program were found it would most likely be the result of localized maternal factors. Deletion experiments suggested that the presence of D1.1 is necessary for the normal formation of the entire length of the dorsal axis (Gallagher et al., 1991). Sixty-two percent of the embryos with D 1.1 deleted bilaterally had dorsal defects (which were defined as missing segments of brain, spinal cord, and somites) compared to 8% of embryos having dorsal defects after bilateral deletion of the ventral midline animal blastomere (V1.1). However, deletion studies alone cannot prove the existence of a localized determinant of fate because the results could be due to disruptions in mesoderm induction and/or gastrulation movements. Therefore we used the classical approach of transplanting blastomeres to a novel location to test their state of determination. Although ventral animal blastomeres transplanted to a ventral vegetal site assumed the fate of their new position, dorsal animal blastomeres maintained their original fate (Gallagher et al., 1991). By injecting the transplanted blastomeres with a lineage tracer, the cellular fates of the descendants could be determined at late tailbud stages. In about 60% of the cases, the labeled cells were found in cranial ganglia, brain, spinal cord, notochord, and central somite; these all are tissues appropriate for the donor but not for the host. In the remaining 40% the clone from the transplanted Dl . 1 participated in forming a partial secondary axis composed of neural tube, somite, and notochord populated by both donor and host cells. Thus, the progeny of a dorsal animal blastomere followed their normal lineage program to differentiate into dorsal axial tissues even after transplantation to the opposite pole of the embryo. These data indicate that by the 16-cell stage this blastomere has received instructions regarding its dorsal fate, and is intrinsically capable of carrying out some degree of its developmental program. Although the observations that this blastomere contains unique cytoplasm (Herkovits and Ubbels, 1979; lmoh, 1984; Danilchik and Denegre, 1991), and the fact that the cells were transplanted prior to mesoderm induction (Jones and Woodland, 1987), suggest that this program may be the result of localized maternal determinants, it is possible that the transplanted dorsal blastomeres, having the receptors for the dorsal mesoderm signal (e.g., activin receptor; Hemmati-Brivanlou et ul., 1992), might still respond to these signals in their new ventral vegetal position. To determine whether the dorsal program is intrinsic or requires some intraembryonic signaling, D1.1 blastomeres were removed from the 16-cell embryo and cultured in
120
Sally A . Moody et al.
groups of four to six in simple salt medium until the equivalent of tailbud stages. As an indicator of dorsal axial differentiation, explants were scored for whether they elongated (a morphological indicator of dorsal mesodermal tissue that underwent convergence-extension movements; Keller et a l . , 1985; Keller, 1986), for the immunohistochemical detection of a notochord-specific marker (Zanetti et al., 1985; Smith and Watt, 1985), and for the presence of a cement gland, whose differentiation requires the presence of dorsal mesoderm (Jamrich and Sato, 1989; Sive et al., 1989). Elongation was observed in about 70% of the D1. I explants, and notochord and cement glands were detected in about half of them. These three indicators of dorsal differentiation were not observed in explants of ventral blastomeres (Gallagher et al., 1991). Since the dorsal program of the D l . 1 blastomere appeared to be independent of mesoderm-inductive signaling, and since no other early intercellular signaling had yet been described, the most likely source of the autonomous dorsal program was a maternal determinant, presumably residing in the cytoplasm. In Drosophila the existence of axial determinants was first identified by transferring cytoplasm from one pole of the embryo to another, and observing whether cell fates changed (Nusslein-Volhard, 1991). We repeated this approach in Xenopus by transferring wild-type cytoplasm from dorsal midline blastomeres (Dl . I and D2.1) to ventral midline blastomeres (V1.l and V2.1). We were not able to transfer whole cytoplasm, as others were (Yuge et al., 1990; Fujisue et al., 1993), because it clogged our injection apparatus. Instead, we injected a fractionated cytosol (plus lineage tracer) from which lipid and particulate matter had been depleted (Hainski and Moody, 1992). Although secondary axes were not observed, some small changes in fate occurred. Cytosol from D 1.1 caused V1.l to contribute more often to forebrain and hindbrain and caused V2. I to contribute more often to branchial arch mesoderm and notochord. Cytosol from D2.1 caused V 1.1 to contribute more often to hindbrain. These results were modest compared to the dramatic secondary axes that developed after larger amounts of whole D2.1 cytoplasm were injected into V2.1 (Yuge et al., 1990; Fujisue et al., 1993). The fractionated cytoplasm we used became diluted during preparation, and it is likely that this would have diluted any cytoplasmic factor present. In the Drosophila tradition, we next isolated total RNA from the various midline blastomeres, mixed it with lineage tracer, and injected it into ventral midline cells (Hainski and Moody, 1992). About half of the embryos in which D l . I RNA was injected into both V2.1 developed secondary dorsal axes. These axes were populated by cells derived from the injected blastomere, as indicated by the lineage tracer, and by unlabeled host cells. The induction of a secondary axis was dependent upon the concentration of Dl . I RNA that was injected and was specific for D1.l RNA. Furthermore, injection of D1.l RNA into UVventralized embryos rescued the dorsal axis. Interestingly, secondary axes never were detected when the same D I . 1 RNA preparation was injected into V I . 1; a similar result has been reported for the secondary axis induction with Xwnt-8
4. Determination of Xenopus Cell Lineage
121
(Smith and Harland, 1991) and noggin (Smith and Harland, 1992) mRNAs. It is possible that this animal cell cannot participate in secondary axis formation for mechanical reasons, because in normal embryos virtually none of its progeny ingress at the ventral blastopore lip (Bauer er a / . , 1994). These results demonstrate that maternal RNA molecules contained within a specific dorsal animal blastomere are important for the establishment of the dorsal axis, and support the hypothesis that maternal dorsal determinants are located in the animal hemisphere of the early embryo. Other experiments also have supported the notion of a determinant localized in the dorsal animal quadrant of the embryo prior to mesoderm induction. A developmental bias to produce dorsal structures seems to exist prior to midblastula transition (Kageura and Yamana, 1983; Gurdon er a/., 1985; Sharpe e t a / . , 1987; London et a/., 1988; Thomsen el a / ., 1990), and a prepattern of dorsal differentiative response to activin exists in dorsal animal cells (Sokol and Melton, 1991; Kinoshita et nl., 1993). Recently Lemaire and Gurdon (1994) showed that dorsal cells dissociated from the rest of the embryo prior to mesoderm induction and cultured past the midblastula transition autonomously express the dorsal mesoderm-specific gene goosecoid. But, what is the molecular nature of this dorsal bias and does the RNA isolated from D1.1 contain “biasing factors”‘? As a first step to answer these questions, it was necessary to demonstrate that mesoderm inductive signals were not involved. First, the general claim that mesoderm induction does not occur until the 64-cell stage is not entirely accurate. Although the animal pole cells may not be responsive to mesoderm inductive signals until the 64-cell stage, the vegetal cells are competent to induce as early as the 16-cell stage (Jones and Woodland, 1987). Therefore, the cells that we isolated at the 16-cell stage (late in the cell cycle so that the cytoplasmic bridges between sister cells would be gone) could have had a short exposure to dorsal mesoderm-inducing signals from their vegetal neighbors. We carried out several culture studies to address this point (Moody et a/.,1993). If the dorsal fate of dorsal animal blastomeres is determined entirely at the 16-cell stage by inherited cytoplasmic factors (Fig. 4A), then subsequent exposure to the source of mesoderm induction, its vegetal neighbor, should not affect is differentiation program. To test this possibility we placed D1.1 and D2.1 blastomeres in culture together; these explants elongated slightly more frequently (87%) than D1. I blastomeres alone (73%). Explants in which D 1 . 1 ~were cocultured with V2.1, the source of the ventral mesoderm inducing signal, elongated at the same frequency as D1.l alone (67%). These results suggested that the D1.l dorsal differentiation program was not affected by mesoderm inducing signals. However, an earlier signal may initiate the dorsal program of D l . l . When the dorsal animal 8-cell blastomere (Dl), i.e., the mother of D1.l (Fig. l), was cultured it never elongated or produced notochord. This was not due to a general lack of responsiveness, because it could show dorsal differentiation if activin was added to the medium during the first 2-3 h of culture (71% of cases), or the cellular
122
Sally A. Moody et al. ANIMAL
s 0
v)
b
I-
Fig. 4 There are two mechanisms by which a dorsal animal blastomere might pursue an autonomous dorsal program. (A) The blastomere may uniquely inherit localized dorsal determinants. (B) All animal blastomeres may contain determinants (stippling) that are “activated” (plus signs) only in those cells receiving a signal from a vegetal neighbor. The secreted products of Vg-1 and Xwnt-1 1 are candidates for such a signal.
source of the dorsal mesodermal signal (D2.1) was added to the culture (100% of cases). Kinoshita et al. ( 1993) also have shown that the D 1 cell differentiates into dorsal mesoderm after an even shorter exposure to activin. This result suggests that the dorsal differentiation of D1.1 explants was accomplished because the blastomere was in situ long enough to receive sufficient dorsal mesoderm signaling from the vegetal cells. This brief exposure could be sufficient to initiate a dorsal “bias.” Control experiments also suggested that the explanation for the autonomous dorsal differentiation of D1.l could not simply be a localized maternal factor. We had assumed that the ventral animal blastomeres (V1.1) were devoid of these factors because these cells did not form dorsal mesoderm in culture and their RNA did not induce secondary axes (Gallagher er al., 1991; Hainski and Moody, 1992). However, when V 1.1 explants were cultured with the cellular source of the dorsal mesoderm-inducing signal (D2.1) or exposed to activin during the first 2-3 h of culture, they elongated and contained differentiated notochord (72% of cases). Kinoshita et al. (1993) performed a similar experiment with 8-cell ventral blastomeres (Vl) and showed that with shorter activin exposures the explants do not contain dorsal mesoderm structures (in contrast to dorsal explants), but with longer exposures the explants express goosecoid, a dorsal mesoderm-specific gene. Both sets of studies indicate that the ability to follow the dorsal pathway in response to the activin signal, or the native vegetal signal from D2.1, is not restricted to the dorsal animal cells. Our working hypothesis to tie together these many observations is that all animal blastomeres contain a maternal RNA(s) that is brought into the translatable pool by a brief exposure to a dorsal vegetal signal, perhaps activin, Vg-1,
4. Determination of Xenopus Cell Lineage
123
and/or Xwnt-1 1 (Fig. 4B). This proposal, of course, will require much investigation to determine: What are the signaling molecules? What are the responding molecules and how are they “activated”? A first test, which is under way, is to demonstrate whether RNA isolated from these variously treated blastomere explants confirm that a dorsal mesoderm signal transforms an animal-specific maternal RNA species into an “activated” form. Of much interest is the question whether this early signaling constitutes the establishment of dorsal bias.
VII. Cell-Cell Signaling in Fate Determination There are many examples to demonstrate that an important aspect of cell fate determination is being in the right place at the right time, as elegantly shown in the determination of nematode vulva1 lineages (Ferguson et al., 1987). As described above, the best characterized and earliest known form of cell-cell signaling in Xenopus development is mesoderm induction. Our studies (Moody et al., 1993) suggest that this signaling has an effect almost as soon as the equatorial cleavage furrow separates animal from vegetal hemispheres at the 8- to 16-cell stages. Explant studies demonstrate that whereas dorsal blastomeres contain a determinative RNA, ventral animal cells can respond to mesoderm inductive signaling to differentiate according to a dorsal program. An important question to ask is whether all blastomeres are competent to respond to this signaling. Also, is the vegetal-to-animal mesoderm inductive signal the only one occurring at the early cleavage stages, or are there others whose effects are sufficiently subtle to have been missed by morphological analyses? This section will focus on studies that have addressed these two questions.
A. Are All Blastomeres Competent to Respond to Dorsal Mesoderm Inductive Signaling? We have approached this question by focusing our analysis on a small region of the CNS, the retina (Huang and Moody, 1993). As noted earlier in this chapter, the retina descends from as many as nine blastomeres of the 32-cell embryo. Counting the number of cells in the clone descended from each progenitor demonstrated that each blastomere produced a distinct proportion of the retina (Fig. 5). To test experimentally whether the blastomeres are determined with respect to their distinct clonal patterns, the major progenitor of the retina (D1 . I . 1) was deleted bilaterally; these blastomeres together normally produce 57% of the cells in one retina. In a majority of cases the eyes developed normally and attained a normal size after the deletion, indicating that the remaining blastomeres can compensate for the deletion. To define which of the remaining blastomeres regulated to restore the retinal lineages, each blastomere was in-
124
Sally A. Moody et a/. ANIMAL
Fig. 5 A summary of the potential of 32-cell blastomeres to express retinal lineages. Five blastomeres on the left side of the embryo normally contribute to the left retina, and the typical amount of retina that they each produce is depicted as a percentage. The four dorsal animal blastomeres on the right side of the embryo also normally contribute small percentages to the left retina (not shown). After bilateral deletion of the major retinal progenitor (DI, I . 1, or A l , see Fig. I), two ventral blastomeres that are not part of the normal retinal lineage change fate to produce retina (arrows). Plus signs identify two other ventral blastomeres that can produce retinal lineages when they are transplanted to the D1.1.1 position. Asterisks identify blastomeres that cannot produce retinal lineages when they are transplanted to the D I . I . 1 position. The asterisk is smaller in D2. I .2 because that transplanted cell produced retinal progeny in 25% of cases.
jected with lineage dye at the time of the ablation. None of the tier-3 blastomeres, nor the tier-2 ventral midline blastomeres gave rise to retinal cells in these cases (Fig. 9, but almost all other animal blastomeres changed fate. Surprisingly the ventral animal blastomeres significantly increased their contribution and the dorsal animal blastomeres significantly decreased their contribution. These changes suggest that when the gap in the embryo created by the deletion healed, ventral cells moved into the “retinogenic” area and dorsal cells moved out of this area. The putative “retinogenic” area is that position normally occupied by the DI. 1.1 blastomeres. To test this hypothesis, different animal blastomeres were deleted to shift their position in predictable directions, and the size of the retinal clones of remaining blastomeres was measured. The results indicated that deletions in the dorsal-animal region result in unidirectional position shifts (ventral to dorsalanimal to vegetal) of the remaining blastomeres, and that these position shifts directly affect retinal fate. Deletion of D2.1.2 resulted in D1.1.2 taking a D2.1.2 retinal fate, and deletion of D1.1.2 resulted in D1.1.1 taking a D1.1.2 retinal fate. This position shift was in the same direction as the normal movement of the clones of these blastomeres prior to the formation of the dorsal lip of the blastopore (Bauer er al., 1994); dorsal animal clones crowd toward the vegetal equatorial region and ventral clones spread dorsally over the animal cap. Deletions of
4. Determination of Xenopus Cell Lineage
125
single cells along the dorsal midline probably allows neighbors near the animal pole to move more vegetally during these pregastrula movements, and thus acquire a retinal fate according ro cell-cell interactions appropriate for their new position (e.g., see Saha and Grainger, 1992). The deletion studies suggest, but do not directly prove, that the dorsal animal midline position is the place to be in order to be a retinal progenitor. To test this hypothesis directly several different blastomeres that do not normally produce retina were transplanted to the D l . 1.1 position (Huang and Moody, 1993). Two ventral blastomeres (V 1.1.2 and V2.1.2) produced retinal clones that were qualitatively and quantitatively indistinguishable from normal D1. I . I retinal clones (Fig. 5). Furthermore, when the D1.1.1 cell was transplanted to a ventral equatorial position it no longer produced retinal cells (Gallagher et al., 1991; Huang and Moody, 1993). These results demonstrate that in order for a retinal lineage to be produced, blastomeres must occupy a dorsal-animal midline position. It is likely that cells in this position receive information regarding their retinal fate through a series of signaling events, perhaps beginning with dorsal mesoderm induction, and including the regionalization of the mesoderm that allows retinainducing capabilities (Saha and Grainger, 1992), the establishment of the competence of ectoderm to respond to neural induction, and the interactions that lead to optic cup formation. Although the retinal fate of many of the Xenopus cleavage stage progenitors is dependent upon blastomere position, not all blastomeres are competent to respond to the signaling that directs blastomere clones to express a retinal fate. The clones of a dorsal tier-3 cell (D2.1.2) transplanted into the D1.l.l position contributed to retina in only 25% of the cases, and the clones of vegetal tier blastomeres transplanted into the D1.l. 1 position never contributed to the retina (Fig. 5). It has been shown by placing vegetal pole cells into the blastocoele at different stages that the cells we transplanted are still multipotent (Heasman et al., 1984). Thus, their lack of competence to form retinal lineages is not due to a restriction to an endodermal fate, but probably is due to a lack of a receptor or a transduction molecule(s) (see Jessell and Melton, 1992) that renders their descendants incapable of responding to whatever signal(s) instructs cells in the D1.1.1 position to make retina. Understanding the signaling and receptors molecules involved in the induction of retinal lineages and determining whether the block in this determinative pathway occurs during mesoderm, neural, or optic cup induction are important avenues of future investigation.
B. Are There Other Important Signaling Events During Cleavage Stages? Cell-cell interactions have several different roles at different stages of development. In the early stages, cell-cell interactions establish the tissues from which the wide variety of cell types required by the multicellular organism can develop.
126
Sally A. Moody et al.
For example, mesoderm induction, the earliest known cell-cell interaction, establishes the germ layer from which will develop most of the musculoskeletal and hematopoietic systems (Nieuwkoop, 1973; Sudarwati and Nieuwkoop, 1971). Later, cell-cell interactions are required for the terminal differentiation of specific cell types and control of their distribution. For example, in nematode, diffusible factor/receptor-mediated interactions induce the formation of the anchor cell, which in turn signals specific lineages in vulva1 precursors (Ferguson et a l . , 1987). In Drosophila, contact-dependent signaling between the R8 photoreceptor and the R7 precursor induces R7 development in exactly one cell in each ommatidium (Banerjee and Zipursky, 1990). Cell-cell signaling is also a mechanism for regulation of cell numbers. The lateral inhibition of Drosophila neuroblast formation through the Notch pathway exemplifies the role of cell-cell signaling in limiting the numbers of cells recruited for the formation of an organ system (Hartenstein and Campos-Ortega, 1984; Hoppe and Greenspan, 1986). The cleavage and early blastula stages of Xenopus are recognized as a period of important signaling that does not rely upon gene transcription, and our understanding of the molecular events involved, although far from complete, has been greatly expanded (Dawid et a/., 1992; Kimelman et al., 1992). Although tremendous efforts have been focused on the molecular basis of mesodermal induction (Dawid, 1994; Kessler and Melton, 1994), we know virtually nothing about whether there are other important signaling events that occur during this period. That other kinds of signaling may occur prior to MBT is suggested by several studies demonstrating that deletions of blastomeres or recombination of blastomeres can result in apparently normal embryos (Kageura and Yamana, 1983, 1984, 1986). Although the remaining cells must communicate in order to reconstitute the embryonic pattern, the signaling involved has not been elucidated. We became interested in this problem from experiments showing that deleting an animal blastomere changes the number of spinal neurons that descend from the remaining blastomeres (Jacobson, 1981b; Gallagher and Moody, 1986). In the interest of determining which cells would compensate after deleting the major progenitors of two spinal neurons (PMN and RBN; Moody, 1989), we deleted tier-2 blastomeres and studied the fates of their neighbors (Moody and Best, 1988). Whereas tier- 1 and tier-2 blastomeres compensated for these deletions by increasing the number of PMN and RBN in their clones, the tier-3 blastomeres made smaller contributions to the CNS. These experiments suggest that a potential regulatory signal in the embryo is necessary for tier-3 blastomeres to express neuronal lineages. This possibility is intriguing because the Xenopus nervous system primarily derives from the three animal-most tiers of the 32-cell embryo, with very little contribution from the vegetal tier (tier-4; Dale and Slack, 1987; Moody, 1987b, 1989). Thus the tier-3 cells are on the border of the “neurogenic” region of the cleavage stage embryo, and may require interactions with more animal neighbors in order to maintain this fate. The reduced contribution of the tier-3 blastomere to the CNS following dele-
4. Determination of Xenopus Cell Lineage
127
tion of the neighboring tier-2 blastomere could occur via three cellular mechanisms: (1) the CNS lineages of the tier-3 blastomere could be inhibited by its new tier- 1 neighbors; (2) deletion of the tier-2 blastomere might result in a change in the positional value of the tier-3 blastomere, which could suppress CNS development; or (3) the loss of the tier-2 neighbor could remove a tier-2 to tier-3 signal necessary for CNS development. To distinguish between these possibilities, the calcium-dependent adhesion between cells was disrupted without changing blastomere position. Groups of embryos were dissociated within their vitelline membranes for different 1-h periods prior to MBT to define the time period during which the putative interaction occurs. The contributions of the tier-3 blastomeres to CNS structures after 1-h dissociations at any period between the 8- and the 2000-cell stage were reduced in a manner similar to that resulting from tier-2 deletion (Fig. 6; Bauer and Moody, 1992). The contributions of the tier-3 blastomeres were most severely affected in the most rostral structures to which the blastomere normally contributes. For example, the contribution of D2.1.2 was most significantly reduced in the hindbrain, whereas the contribution of its lateral neighbor was most significantly reduced in the forebrain and midbrain. The contribution of each tier-3 cell to PMN and RBN also was significantly reduced (Fig. 6 ) . Since the neighbors and the positions of the blastomeres in the dissociated embryos did not change, some signal that was contact-dependent must have been blocked. Contact-dependent signals may be carried by diffusible factors across a close intercellular space, by the interactions of membrane proteins in opposing cell membranes, or through gap junctions. Since previous studies have implicated gap junctions in dorsal axis formation (Warner el al., 1984; Nagajski et al., 1989), gap junctional communication was tested during dissociation by injection of biocytin, which passes through open gap junctions. The incidence of dye coupling in untreated embryos and in dissociated embryos was not significantly different, indicating that the calcium-dependent signal that influences tier-3 neural lineages is not mediated by gap junctions. Apparently the dissociation medium, while removing calciumdependent adhesion, does not disrupt connexons. Incubations in calcium-free media dissociate all blastomeres from their neighbors. To determine whether a specific tier-2 to tier-3 contact is necessary for tier-3 neural lineages, a barrier was inserted for I h between individual pairs of tier-2/tier-3 blastomeres. In a pattern similar to the results described above for the other manipulations, the contributions of tier-3 blastomerei to rostral CNS structures and to PMN and RBN were significantly reduced (Fig. 6). These experiments indicate the presence of a regulatory signal in the cleavage stage embryo that is required for tier-3 contributions to the CNS. This signal depends on contact between tier-2 and tier-3 blastomeres. Because it is such an early signal, occurring before neural induction and even before the completion of mesoderm induction, it seems to be a competence modifier, and its loss results in diversion of the normal course of differentiation for the progeny of tier-3 blast-
128
Sally A. Moody et (11.
Fig. 6 The mean number of primary motoneurons (top row) and Rohon-Beard neurons (bottom row) produced by tier-3 blastomeres in normal embryos (A), and after various manipulations at the 32-cell stage: (B) deletion of tier-2 neighbors; (C) 1 h incubation in calcium-free medium; (D)insertion of a barrier between individual tier-2/tier-3 pairs for 1 h. For each manipulation there is a significant decrease in these two spinal neurons. Dashes indicate that the blastomere does not normally produce this particular neuron. The “ X ”in (D) indicates that the barrier experiment was not performed for the VI . I .2/V2. I .2 blastomere pair.
omeres. Two interesting features of this signal are its animal-to-vegetal direction and its very early timing. During the cleavage stages, mesoderm induction results from a vegetal-to-animal signaling process. Following this induction, further signals occurring within the mesoderm pass from dorsal to ventral, consolidating the dorsal mesodermal identity and dorsalizing the nearest ventral mesoderm (Dawid et al., 1992; Kimelman etal. , 1992; Slack, 1991). The signal characterized by our studies travels in an animal-to-vegetal direction, and there-
4. Determination of Xenopus Cell Lineage
129
fore is distinct from mesoderm induction. Cardellini (1988) also has postulated the existence of an animal-to-vegetal signal that does not utilize gap junctional communication. Although that interaction occurs along the same cleavage furrow as that in our studies, it transiently assigns which vegetal cells will be dorsal. Our experiments do not identify a dorsal-ventral difference, but suggest that an animal-vegetal signal that occurs long before neural induction is necessary for the formation of CNS from the tier-3 blastomeres. Identifying the molecular nature of this signal is the next important goal.
VIII. Conclusions One reason we have so many different animal models for developmental questions is that each is well suited to certain questions at some developmental periods, but poorly suited to others. In fact one challenge for a developmental biologist is to test experimentally the right question at the right developmental stage. Although the elegant studies in nematode and ascidia have shown that lineage itself plays a role in cell fate determination, similar analyses have been difficult to perform in more complex vertebrates in which the precise mitotic patterns cannot be mapped throughout all developmental stages. What we have tried to achieve in Xenopus is an experimental test of the relative roles of possible mechanisms of fate determination at the stages when we can control for lineage variation. As has been shown in virtually all developing systems, in Xenopus determination of fate is a niultistep process that begins with the polarization of the egg and the establishment of the embryonic dorsal-ventral axis; is followed by a number of signaling events that regionalize the embryo, establish tissue types, and induce organ anlage; and finishes with late interactions that finalize cellular phenotype. By having the ability to study specific identified cell lineages in the early Xenopus embryo, it has been possible to demonstrate that during the period of transcriptional quiescence both the inheritance of maternal factors and cell-cell interactions play important roles in specifying the cellular phenotypes that will differentiate in these clones. Not only is there evidence for maternal mRNA localized in dorsal animal lineages (Hainski and Moody, 1992), but its products may be activated by very early signaling from a vegetal neighbor (Moody et al., 1993), perhaps by a product of another localized maternal determinant (Yuge et a l . , 1990; Fujisue et a l . , 1993; Holowacz and Elinson, 1993). We predict that this dorsal vegetal signaling has an effect on animal cells two to three cell cycles earlier than the onset of competence to respond to mesoderm induction (Kinoshita et al., 1993). This early signaling may establish the “dorsal bias” that seems to be independent of mesoderm induction (Shape et a l . , 1987; London et al., 1988; Sokoi and Melton, 1991; Lemaire and Gurdon, 1994). One future challenge is to identify the maternal molecules, both signaling and responding, and to
130
Sally A. Moody et al.
understand the difference between the phenomena of "dorsal bias" and mesoderm induction. Are they separate events or an early and late manifestation of the same signal? Dorsal animal blastomeres, through a combination of inherited maternal factors and receipt of dorsalizing signals from dorsal vegetal cells, can express a dorsal program of differentiation autonomously (Gallagher el al., 1991; Kinoshita et al., 1993). However, not all aspects of their normal fate, e.g., expression of anterior CNS phenotypes such as retina (Gallagher et al., 1991; Huang and Moody, 1993), are expressed when the blastomeres are removed from their normal position in the embryo. Some aspect of that position, presumably the cellular interactions in which cells later engage, determines at least the ability to produce retina. Does this position-dependent signaling occur prior to gastrulation or does it occur during the anteroposterior specification of axial mesoderm and neural plate? And, why are not all blastomeres competent to respond to the position-dependent signaling? Perhaps, being of vegetal origin the nonresponding cells lack the receptors or signal transducer. Perhaps they lack the receptors for neural induction or anteriorizing neural signals. It should be possible to dissect these alternatives by providing transplanted vegetal blastomeres with the appropriate receptor mRNA. Finally, there are cell contact-dependent signaling events that recruit vegetal equatorial cells into neuronal lineages. These events emanate from the animal cells, are necessary throughout cleavage stages, and probably are independent of mesoderm inductive signaling and gap junctional signaling. Although many such interactions may normally occur during development, and perhaps be responsible for the remarkable ability of Xenopus embryos to regulate after deletions and damage, they have been elucidated only after techniques for assaying subtle changes in fate became available. An important next step will be the molecular identification of the signaling and responding moieties, and to elucidate the relationship between these signals and those used for mesoderm induction. As in any cell biological system, the cell-cell interactions seem simple and straightforward at the beginning. As we learn more about the specific events occurring, more molecules line up in a cascade of characters. There has been a tendency to expect very little of importance to occur during the cleavage stages because of the rapid cell divisions and the lack of gene transcription. The regulative ability of embryos at these stages also indicated to many that all cells were virtually equivalent, and therefore no developmental decisions of importance had yet occurred. With more sensitive assays, it is clear that many early steps in the determinative pathway have begun, and that they strongly bias the later ability of their descendants to express region-specific genes and to respond to regionspecific inductions during gastrulation. Ascertaining whether these molecular events function in the same way in other vertebrates and testing the ability of cells to recapitulate these early events after genetic and environmental damage will be important future goals.
4. Determination of Xenopus Cell Lineage
131
Acknowledgments We thank the many members of the Moody Laboratory, past and present, who have contributed significantly to this work. This research has been sponsored by NIH Grants EY10096 and NS23158.
References Ambros, V., and Horvitz, H. R. (1985). Heterochronic mutants of the nematode Cuenorhabditis elegans. Science 226, 409-4 16. Asashima, M . , Nakano, H., Sugino, H., Nakamura, T., Eto, Y., Ejima, D., Nishimatsu, S., Ueno, N., and Kinoshitia, K. (1991). Presence of activin (erythoid differentiation factor) in unfertilized eggs and blastulae of Xenopus luevis. Proc. Natl. Acad. Sci. U.S.A. 88, 65116514. Balakier, H., and Pedersen, R. A. (1982). Allocation of cells to inner cell mass and trophectoderm lineages in preimplantation mouse embryos. Dew. Eiol. 90, 352-362. Banerjee, U., and Zipursky, S. L. (1990). The role of cell-cell interaction in the development of the Drosophila visual system. Neuron 4, 177-187. Bauer, D. V., and Moody, S. A. (1992). Disruption of cell-cell interactions reduces the CNS contribution of tier-3 blastomeres of the 32-cell Xenopus embryo. Mol. Eiol. Cell 3, 107a. Bauer, D. V., Huang, S . , and Moody, S . A. (1994). The cleavage stage origin of Spemann's Organizer: Analysis of the movements of blastomere clones before and during gastrulation in Xenopus. Development (Cambridge, U K ) 120, 1179- 1189. Brown, E. E., Denegre, J. M., and Danilchik, M. V. (1993). Deep cytoplasmic rearrangements in ventralized Xenopus embryos. Dev. Biol. 160, 148-156. Cardellini, P. (1988). Reversal of dorsoventral polarity in Xenopus luevis embryos by 180" rotation of the animal micromeres at the eight-cell stage. Dev. Eiol. 128, 428-434. Chalfie, M., Horvitz, H. R., and Sulston, J. E. (1981). Mutations that lead to reiterations of cell lineages of C. elegans. Cell (Cambridge. Mass.) 24, 59-69. Clegg, K . B., and Piko, L. (1977). Size and specific activity of the UTP pool and overall rates of RNA synthesis in early mouse embryos. Dev. Eiol. 58, 76-95. Cornell, R. A., and Kimelman, D. ( I 994a). Combinatorial signalling in development. EioEssays 16, 577-581. Cornell, R. A , , and Kimelman, D. ( I 994b). Activin-mediated mesoderm induction requires FGF. Development (Cambridge, U K ) 120,453-462. Dagle, J. M . , Walder, J. A., and Weeks, D. L. (1990). Targeted degradation of mRNA in Xenopus oocytes and embryos directed by modified oligonucleotides: Studies of An2 and cyclin in embryogenesis. Nucleic Acids Res. 18, 475 1-4757. Dale, L., and Slack, J. M. W. (1987). Fate map for the 32-cell stage of Xenopus laevis. Development (Cambridge, U K ) 99, 527-55 I . Danilchik, M. V., and Black, S. D. (1988). The first cleavage plane and the embryonic axis are determined by separate mechanisms in Xenopus luevis. 1. Independence in undisturbed embryos. Dev. Eiol. 128, 58-64. Danilchik, M . V., and Denegre, J. M. (1991). Deep cytoplasmic rearrangements during early development in Xenopus luevis. Development (Cambridge. U K ) 111, 845-856. Danilchik, M. V., and Gerhart, J. C. (1987). Differentiation of the animal-vegetal axis in Xenopus laevis oocytes. I. Polarized intracellular translocation of platelets establishes the yolk gradient. Dev. Eiol. 122, 101-112. Davidson, E. H. (1990). How embryos work: A comparative view of diverse modes of cell fate specification. Development (Cambridge, U K ) 108, 365-389.
132
Sally A . Moody et al.
Dawid, I . B. (1994). Intercellular signalling and gene regulation during early embryogenesis of Xenopus laevis. J. Biol. Chem. 269, 6259-6262. Dawid, I. B., Taira, M., Good, P. J., and Rebagliati, M. R. (1992). The role of growth factors in embryonic induction in Xenopus laevis. Mol. Reprod. Dev. 32, 136-144. Denegre, J. M., and Danilchik, M. V. (1993). Deep cytoplasmic rearrangements in axisrespecified Xenopus embryos. Dev. Biol. 160, 157- 164. Doe, C. Q., and Goodman, C. (1985). Early events in insect neurogenesis. 11. The role of cell interactions and cell lineage in the determination of neuronal precursor cells. Dew. Biol. 111, 206-2 19. Dohrmann, C. E., Hemmati-Brivanlou, A., Thomsen, G . H., Firlds, A,, Woolf, T. M., and Melton, D. A. (1993). Expression of activin mRNA during early development in Xenopus laevis. Deu. Biol. 157, 474-483. Elinson, R. P., and Kao, K. R . (1989). The location of dorsal information in frog early development. Dew., Growth Difer. 31, 423-430. Elinson, R. P., and Pasceti, P. (1989). Two UV-sensitive targets in dorsoanterior specification of frog embryos. Developmenf (Cambridge, U K ) 106, 5 1 1-5 18. Elinson, R. P., and Rowning, B. (1988). A transient array of parallel microtubules in frog eggs: Potential tracks for a cytoplasmic rotation that specifies the dorso-ventral axis. Dew. Biol. 128, 185- 197. Elinson, R. P., King, M. L., and Fotristall, C. (1993). Isolated vegetal cortex from Xenopus oocytes selectively retains localized mRNAs. Dew. Biol. 160, 554-562. Ferguson, E. C., Sternberg, P. W., and Horvitz, R. (1987). A genetic pathway for the specification of vulva1 cell lineages of Cuenorhabditis eleguns. Nature (London) 326, 259-267. Fujisue, M., Kobayakawa, Y., and Yamana, K. (1993). Occurrence of dorsal axis-inducing activity around the vegetal pole of an uncleaved Xenopus egg and displacement to the equatorial region by cortical rotation. Development (Cambridge, U K ) 118, 163- 170. Fukui, A., Nakamura, T., Hideho, U., Sugino, K., Sugino, H . , and Asashima, M. (1994). Identification of activins A, AB, and B and follistatin proteins in Xenopus embryos. Deu. Biol. 163, 279-28 I . Gallagher, B . C., and Moody, S. A. (1986). The influence of neighboring blastomeres on the lineage of the V1.2 blastomere of the 16-cell Xenopus embryo as determined by ablation experiments. Soc. Neurosci. Abstr. 12, 1121. Gallagher, B. C., Hainski, A. M., and Moody, S . A. (1991). Autonomous differentiation of dorsal axial structures from an animal cap cleavage stage blastomere in Xenopus. Development (Cambridge. U K ) 112, 1103-1 114. Gerhart, J., Danilchik, M., Doniach, T., Roberts, S . , Rowning, B., and Stewart, R. (1989). Cortical rotation of the Xenopus egg: Consequences for the anteroposterior pattem of embryonic dorsal development. Development (Cambridge, U K ) 107 (Suppl.), 37-52. Gimlich, R. L. (1986). Acquisition of developmental autonomy in the equatorial region of the Xenopus embryo. Dew. Biol. 115, 340-352. Gimlich, R. L., and Gerhart, J. C. (1984). Early cellular interactions promote embryonic axis formation in Xenopus luewis. Dew. Biol. 104, 117-130. Godsave, S . , and Slack, J. M. W. (1991). Single cell analysis of mesoderm formation in the Xenopus embryo. Development (Cambridge, U K ) 111, 523-530. Gray, G. E., Glover, J. C., Majors, J., and Sanes, J. R. (1988). Radial arrangement of clonally related cells in the chicken optic tectum: Lineage analysis with a recombinant retrovirus. Proc. Natl. Acad. Sci. U.S.A. 85, 7356-7360. Gurdon, J. B., Mohun, T. J., Fairman, S . , and Brennan, S. (1985). All components required for the eventual activation of muscle specific genes are localized in the subequatorial region of an uncleaved egg. Proc. Natl. Acad. Sci. U.S.A. 82, 139-143. Gururajan, R . , Perry-O’Keefe, H., Melton, D. A,, and Weeks, D. L. (1991). The Xenopus lo-
4. Determination of Xenopus Cell Lineage
133
calized messenger RNA AN3 may encode an ATP-dependent RNA helicase. Nature (London) 349,717-119. Hainski, A. M., and Moody, S. A. (1992). Xenopus maternal RNAs from a dorsal animal blastomere induce a secondary axis in host embryos. Development (Cambridge, U K ) 116, 347355. Hartenstein, V. ( 1989). Early neurogenesis in Xenopus: The spatio-temporal pattern of proliferation and cell lineages in the embryonic spinal cord. Neuron 3, 359-41 1. Hartenstein, V., and Campos-Ortega, J. A. (1984). Early neurogenesis in wild-type Drosophila melanogaster. Wilhelm Roux’s Arch. Dev. Biol. 193, 308-325. Hausen, P., and Riebesell, M. (1991). “The Early Development of Xenopus laevis.” SpringerVerlag, New York. Heasman, J., Wylie, C. C., Hausen, P., and Smith, J. C. (1984). Fates and states of determination of single vegetal pole blastomeres of X . laevis. Cell (Cambridge, Mass.) 37, 185-194. Hemmati-Brivanlou, A,, Wright, D. A , , and Melton, D. A . (1992). Embryonic expression and functional analysis of a Xenopus activin receptor. Dev. Dyn. 194, 1-1 I . Herkovits, J., and Ubbels, G. A. (1979). The ultrastructure of the dorsal yolk-free cytoplasm and the immediately surrounding cytoplasm in the symmetrized egg of Xenopus laevis. J. Embryol. Exp. Morphol. 51, 155-164. Hirose, G., and Jacobson, M. (1979). Clonal organization of the central nervous system of the frog. I . Clones stemming from individual blastomeres of the 16-cell and earlier stages. Dev. Biol. 71, 191-202. Holowacz, T., and Elinson, R. P. (1993). Cortical cytoplasm, which induces dorsal axis formation in Xenopus, is inactivated by UV irradiation of the oocyte. Development (Cambridge, UKJ 119, 277-285. Holt, C. E., Bertsch, T. W., Ellis, H. M., and Hams, W. A. (1988). Cellular determination in the Xenopus retina is independent of lineage and birth date. Neuron 1, 15-26. Hoppe, P. E., and Greenspan, R. J. (1986). Local function of the Notch gene for embryonic ectodermal pathway choice in Drosophila. Cell (Cambridge, Mass.) 46, 773-783. Houliston, E. ( 1994). Microtubule translocation and polymerisation during cortical rotation in Xenopus eggs. Development (Cambridge, U K ) 120, 1213-1220. Houliston, E., and Elinson, R. P. (1991a). Patterns of microtubule polymerization relating to cortical rotation in Xenopus laevis eggs. Development (Cambridge. UK)112, 107-1 17. Houliston, E., and Elinson, R. P. (1991b). Evidence for the involvement of microtubules, ER, and kinesin in the cortical rotation of fertilized frog eggs. J . Cell Biol. 114, 1017-1028. Huang, S., and Moody, S. A. (1992). Does lineage determine the dopamine phenotype in the tadpole hypothalamus?: A quantitative analysis. J . Neurosci. 12, 1351-1 362. Huang, S . , and Moody, S . A. (1993). The retinal fate of Xenopus cleavage stage progenitors is dependent upon blastomere position and competence: Studies of normal and regulated clones. J . Neurosci. 13, 3183-3210. Huang, S . , and Moody, S . A. (1994). Development of glycine and GABA amacrine cells; coexpression of neurotransmitters. SOC. Neurosci. Abstr. 20, 654. Huang, S . , and Moody, S . A. (1995). Asymmetrical blastomere origin and spatial domains of dopamine and Neuropeptide Y amacrine subtypes in Xenopus tadpole retina. J . Comp. Neurol. 360, 2-13. Imoh, H. (1984). Appearance and distribution of RNA-rich cytoplasms in the embryo of Xenopus laevis during early development. Dev., Growth Difler. 26, 167- 176. Jacobson, M. (1981a). Rohon-Beard neuron origin from blastomeres of the 16-cell frog embryo. J . Neurosci. 1, 918-922. Jacobson, M. (1981b). Rohon-Beard neurons arise from a substitute ancestral cell after removal of the cell from which they normally arise in the 16-cell frog embryo. J. Neurosci. 1, 923-927. Jacobson, M. (1983). Clonal organization of the central nervous system of the frog. 111. Clones
134
Sally A. Moody et al.
stemming from individual blastomeres of the 128-, 256-, and 512-cell stages. Dev. Biol. 71, 19I -202. Jacobson, M., and Hirose, G. (1981). Clonal organization of the central nervous system of the frog. 11. Clones stemming from individual blastomeres of the 32- and 64-cell stages. J. Neurosci. 1, 271-284. Jamrich, M., and Sato, S. (1989). Differential gene expression in the anterior neural plate during gastrulation of Xenopus laevis. Development (Cambridge, UK) 105, 779-786. Jenkinson, J. (1909). On the relationship between the symmetry of the egg, the symmetry of segmentation, and the symmetry of the embryo in the frog. Biomefrica 7 , 148-209. Jessell, T. M., and Melton, D. A (1992). Diffusible factors in vertebrate embryonic induction. Cell (Cumbridge. Mass.) 68, 257-270. Johnson, M. H. (1981). The molecular and cellular basis of preimplantation mouse development. Biol. Rev. Cambridge Philos. Soc. 56, 463-498. Johnson, M. H., and Ziomek, C. A. (1981). Induction of polarity in mouse 8-cell blastomeres: Specificity, geometry and stability. J . Cell B i d . 91, 303-308. Jones, E. A , , and Woodland, H. R . (1987). The development of animal cap cells in Xenopus: A measure of the staIt of animal cap competence to form mesoderm. Development (Cambridge, UK) 101, 557-563. Kageura, H. (1990). Spatial distribution of the capacity to initiate a secondary embryo in the 32cell embryo of Xenopus laevis. Dev. Biol. 142, 432-438. Kageura, H., and Yamana, K. (1983). Pattern regulation in isolated halves and blastomeres of early Xenopus laevis. J . Embryol. Exp. Morphol. 74, 221-234. Kageura, H., and Yamana, K. (1984). Pattern regulation in defect embryos of Xenopus laevis. Dev. Biol. 101, 410-415. Kageura, H., and Yamana, K. (1986). Pattern formation in 8-cell composite embryos of Xenopus laevis. J . Embryol. Exp. Morphol. 91, 79-100. Keller, R. E. (1986). The cellular basis of amphibian gastrulation. I n ”Developmental Biology: A Comprehensive Synthesis” (L. Browder, ed.), Vol. 2, pp. 241-327. Plenum, New York. Keller, R. E., Danilchik, M., Gimlich, R., and Shih, J. (1985). The function of convergent extension during gastrulation of Xenopus laevis. J . Embryol. Exp. Morphol. 89 (Suppl.), 185-209. Kelly, S. J. (1977). Studies of the developmental potential of 4- and 8-cell stage mouse blastomeres. J . Exp. Zool. 200, 365-376. Kessel, M., and Gruss, P. (1990). Murine developmental control genes. Science 249, 374-379. Kessler, D. S., and Melton, D. A. (1994). Vertebrate embryonic induction: Mesodermal and neural patterning. Science 266, 596-604. Kimelman, D., Christian, J. L., and Moon, R. T. (1992). Synergistic principles of development: Overlapping patterning systems in Xenopus mesoderm induction. Development (Cambridge, UK) 116, 1-9. Kimmel, C. B., and Law, R. D. (1985). Cell lineage of zebrafish blastomeres. 11. Clonal analyses of the blastula and gastrula stages. Dev. Biol. 108, 94-101. Kimmel, C. B., and Warga, R. M. (1986). Tissue-specific cell lineages originate in the gastrula of the zebrafish. Science 231, 365-368. Kinoshita, K., Bessho, T., and Makoto, A. (1993). Competence prepattern in the animal hemisphere of the 8-cell-stage Xenopus embryo. Dev. Biol. 160, 276-284. Kirschner, M., Gerhart, J. C., Hara, K., and Ubbels, G. A. (1980). Initiation of the cell cycle and establishment of bilateral symmetry in Xenopus eggs. Symp. SOC. Dev. Biol. 38, 187-215. Klein, S. L. (1987). The first cleavage furrow demarcates the dorsal-ventral axis in Xenopus embryos. Dev. Biol. 120, 299-304. Klein, S. L., and King, M. L. (1988). Correlations between cell fate and the distribution of proteins that are synthesized before the midblastula transition in Xenopus. Wilhelm Roux’s Arch. Dev. Biol. 197, 275-281.
4. Determination of Xenopus Cell Lineage
135
Kloc, M., and Etkin, L. D. (1994). Delocalization of Vg-1 mRNA from the vegetal cortex in Xenopus oocytes after destruction of Xlsirt RNA. Science 265, 1101-1103. Kloc, M., Reddy, B. A., Miller, M., Eastman, E., and Etkin, L. D. (1991). A localized maternal transcript in Xenopus laevis. Mol. Reprod. Dev. 28, 341-345. Kloc, M., Spohr, G., and Etkin, L. D. (1993). Translocation of repetitive RNA sequences with the germ plasm in Xenopus oocytes. Science 262, 1712-1714. Komiya, T., Itoh, H., Ikenishi, K., and Furusawa, M. (1994). Isolation and characterization of a novel gene of the dead box protein family which is specifically expressed in germ cells of Xenopus laevis. Dev. Biol. 162, 354-363. Ku, M., and Melton, D. A. (1993). Xwnt-11: A maternally expressed Xenopus wnt gene. Development (Cambridge, U K ) 119, 1161-1173. Laufer, J. S., Bazzicalupo, P., and Wood, W. B. (1980). Segregation of developmental potential in early embryos of Caenorhabdiris elegans. Cell (Cambridge, Mass.) 19, 569-577. Lemaire, P., and Gurdon, J. B. (1994). A role for cytoplasmic determinants in mesoderm patterning: Cell-autonomous activation of the goosecoid and Xwnt-8 genes along the dorsoventral axis of early Xenopus embryos. Developmenr (Cambridge, UK) 120, 1191-1 199. Linnen, J. M., Bailey, C. P., and Weeks, D. L. (1993). Two related localized mRNAs from Xenopus laevis encode ubiquitin-like fusion proteins. Gene 128, 181- 188. London, C., Akers, R., and Phillips, C. (1988). Expression of Epi 1, an epidermis-specific marker in Xenopus laevis embryos, is specified prior to gastrulation. Dev. Biol. 129, 380-389. Mahowald, A. P. (1992). Germ plasm revisited and illuminated. Science 255, 1216-1217. Masho, R. (1988). Fates of animal-dorsal blastomeres of eight-cell stage Xenopus embryos vary according to the specific patterns of the third cleavage plane. Dev., Growth Difer. 30, 347359. Masho, R. (1990). Close correlation between the first cleavage plane and the body axis in early Xenopus embryos. Dev., Growth Differ. 32, 57-64. Masho, R., and Kubota, H. Y. (1986). Developmental fates of blastomeres of eight-cell stage Xenopus embryos. Dev., Growrh Drfer. 28, 113-123. Melton, D. A. (1987). Translocation of a localized maternal mRNA to the vegetal pole of Xenopus oocytes. Nature (London) 328, 80-82. Miyata, S., Kageura, H., and Kihara, H. K. (1987). Regional differences of proteins in isolated cells of early embryos of Xenopus laevis. Cell D i f . 21, 47-52. Moody, S. A (1987a). Fates of the blastomeres of the 16-cell stage Xenopus embryo. Dev. Biol. 119, 560-578. Moody, S . A (1987b). Fates of the blastomeres of the 32-cell stage Xenopus embryo. Dev. Biol. 122, 300-319. Moody, S. A. (1989). Quantitative lineage analysis of the origin of frog primary motor and sensory neurons from cleavage stage bIastomeres. J. Neurosci. 9, 291 9-2930. Moody, S. A., and Best, D. W. (1988). Blastomere interactions are necessary for CNS lineage expression by some frog cells. Soc. Neurosci. Abstr. 14, 1132. Moody, S. A., and Kersey, K. S . (1989). Clonal analysis of GABAergic neurons in frog embryo spinal cord. SOC. Neurosci. Abstr. 15, 885. Moody, S. A,, and Kline, M. J. (1990). Segregation of fate during cleavage of frog (Xenopus laevis) blastomeres. Anar. Embryol. 182, 347-362. Moody, S. A , , Hainski, A. M., and Bauer, D. V. (1993). Determination of dorsal fates of cleavage blastomeres in frog. Mol. Biol. Cell 4, 120a. Mosquera, L., Forristall, C . , Zhou, Y., and King, M. L. (1993). A mRNA localized to the vegetal cortex of Xenopus oocyte encodes a protein with a nanos-like zinc finger domain. Development (Cambridge, UK) 117, 377-386. Nagajski, D. J., Guthrie, S. C . , Ford, C. C., and Warner, A. E. (1989). The correlation between patterns of dye transfer through gap junctions and future developmental fate in Xenopust The
136
Sally A . Moody et al.
consequences of U.V. irradiation and lithium treatment. Development (Cambridge. U K ) 105, 747-752. Nakamura, O., and Kishiyama, K . (1971). Prospective fates of blastomeres at the 32-cell stage of Xenopus laevis embryos. Proc. Jpn, Acad. 47, 407-412. Newport, J. W., and Kirschner, M. W. (1982a). A major developmental transition in early Xenopus embryos: I. Characterization and timing of cellular changes at midblastula stage. Cell (Cambridge, Mass.) 30, 675-686. Newport, J. W., and Kirschner, M. V. (1982b). A major developmental transition in early Xenopus embryos: 11. Control of the onset of transcription. Cell (Cambridge, Mass.) 30, 687-696. Nieuwkoop, P. D. (1973). The “organisation center’’ of the amphibian embryo: Its origin, spatial organisation and morphogenetic action. Adv. Morphog. 10, 1-39 Nishida, H. (1994). Localization of egg cytoplasm that promotes differentiation to epidermis in embryos of the ascidian Halocynrhia roretzi. Developmenr (Cambridge, UKj 120, 235-243. Nusslein-Volhard, C. (1991 ). Determination of the embryonic axes of Drosuphila. Develupment (Cambridge, U K ) , Suppl 1, 1-10. Pedersen, R. A , , Wu, K., and Balakier, H. (1986). Origin of the inner cell mass in mouse embryos: Cell lineage analysis by microinjection. Dew B i d . 117, 581 -595. Raible, D. W., and Eisen. J. S . (1994). Restriction of neural crest fate in the trunk of the embry onic zebrafish. Development (Cambridge. U K ) 120,495-503. Rebagliati, M. R., and Dawid. I . B. (1993). Expression of activin transcripts in follicle cells and oocytes of Xenopus laevis. Dev. Biol. 159, 574-580. Rebagliati, M. R., Weeks, D. L., Harvey, R. P., and Melton, D. A. (1985). Identification and cloning of localized maternal RNAs from Xenopus eggs. Cell (Cambridge. Mass.) 42, 769771. Reddy, B. A,, Kloc, M., and Etkin, L. D. (1992). The cloning and characterization of a localized maternal transcript in Xenopus laevis whose zygotic counterpart is detected in the CNS. Mech. Dev. 39, 143-150. Ressom, R . E., and Dixon, K . E. (1988). Relocation and reorganization of germ plasm in Xenopus embryos after fertilization. Development (Cambridge, U K ) 103, 507-5 18. Saha, M. S . , and Crainger, R. M. (1992). A labile period in the determination of the anteriorposterior axis during early neural development. Neuron 8, 1003- 1014. Scharf, S. R., and Gerhart, J. C. (1983). Axis determination in eggs of Xenopus laevis: A critical period before first cleavage, identified by common effects of cold, pressure and ultraviolet irradiation. Dev. Biol. 99, 75-87. Schechtman, A. M . ( 1935). Relation between the grey crescent and the organizer center of a urodele egg (Tritiirus torasus). Wilhelm Roux’s Arch. Etztwickliingsmech. Org. 134, 207-208. Schneider, D. S . , Jin, Y., Moristato, D., and Anderson, K. V. (1994). A processed form of Spitzle protein defines dorsal-ventral polarity in the Drosophila embryo. Developmenr (Cambridge, U K ) 120, 1243-1250. Scholer, H. R. (1991). Octamania: The POU factors in murine development. Trends Genet. 7, 323-329. Schroeder, M. M., and Card, D. L. (1992). Organization and regulation of cortical microtubules during the first cell cycle of Xenopus eggs. Development (Cambridge, U K ) 114, 699-709. Sharpe, C. R., Fritz, A,, De Robertis, E. M . , and Gurdon, J . B. (1987). A homeobox-containing marker of posterior neural differentiation shows the importance of predetermination in neural induction. Cell (Cambridge. Muss.) 50, 749-758. Sive, H. L., Hattori, K., and Weintraub, H. (1989). Progressive determination during formation of the anteroposterior axis in Xenopirs laevis. Cell (Cumbridge. Mass.) 58, 171-180. Slack, I. M. W. (1991). “From Egg to Embryo: Regional Specification in Early Development,” 2nd ed. Cambridge Univ. Press, New York. Smith, J. C . , and Watt, F. M. (1985). Biochemical speciticity of Xenopus notochord. D@erentiution (Berlin) 29, 109- 1 15.
4. Determination of Xenopus Cell Lineage
137
Smith, W. C.. and Harland, R . M. (1991 1. Injected Xwnt-8 RNA induces a complete body axis in Xenopii,s embryos. Cell (Cumbridge. Muss.) 67, 753-765. Smith, W. C.. and Harland. R . M. (1992). Expression cloning of noggin, a new dorsalizing factor localized to the Speinann organizer in Xenopus embryos. Cell (Carnbridge. Mass.) 70, 829-840. Sokol. S . , and Melton, D. A. (1991). Pre-existent pattci-n in Xenopus animal pole cells revealed by induction with activin. Narure (London) 351, 409-41 I . Stent, G. S . , and Weisblat, D. A (1985). Cell lineage in the development of invertebrate nervous systems. Annu. Rev. Neurosci. 8, 45-70. Strome, S . (1992). The germ of the issue. Nufitre (Loridon) 358, 368-369. Sudarwati. S . , and Nieuwkoop, P. D. (1971). Mesoderm formation in the anuran, Xenupus luevis (Daudin). Wilhelm R o w Arch. Entwicklitrzgsmeclr. Or-g. 166, 189-201. Sulston, J. E., Schierenberg. E . , White, J. G.. and Thomson, J . N. (1983). The embryonic cell lineage of the nematode Caenorhmbdiris elegans. Deu. Biol. 100, 64- I 19. Suzuki. A. S . . Manabe, J . , and Hirakawa. A. (1991). Dynamic distribution of region-specific protein during oogenesis and early embryogenesis of Xenopus laevis. Wilhelrn Rouw's Arch. DeL,. Biol. 200, 213-222. Takasaki. H. (1987). Fates and roles of the presumptive organizer region in the 32-cell embryo in normal development of Xenopus Iuevis. Deu., Growth Dixer. 29, 141- 152. Takasaki, H.. and Konishi, H . (1989). Dorsal blastomeres in the equatorial region of the 32-cell Xeriopus embryo autonomously produce progeny committed to the organizer. Dev., Growth Differ. 31, 147-156. Tannahill. D.. and Melton, D. A. (1989). Localized synthesis of the Vgl protein during early Xenopus development. Development (Cambridge. U K ) 106, 775-785. Thoinsen. G , H . , and Melton. D. A. (1993). Processed Vgl protein is an axial mesoderm inducer in Xenopus. Cell (Cambridge, Mass.) 74, 433-441 Thomsen, G. H . , Woolf. T.. Whitman. M . , Sokol, S . . Vaughan. J . . Vale, W., and Melton, D. A. ( 1990).Activins are expressed early in Xenopus embryogenesis and can inducc axial mesoderm and anterior structures. Cell (Cambridge. Mass.) 63, 485-493. Turner, D. L., and Cepko, C. L. (1987). A common progenitor of neurons and glia persists in rat retina late in development. Nulure (London) 328, 131 - 136. Vincent, J. P.. and Gerhart, J. C. (1987). Subcortical rotation in Xenopus eggs: An early step in embryonic axis specification. Drv. Bid. 123, 526-539. Vincent, I. P., Oster, G. F., and Gerhart. 1. C. (1986). Kinematics of gray crescent formation in Xenopus eggs: The displacement of subcortical cytoplasm relative to the egg surface. Dev. B i d 113, 484-500. Vize, P. D., and Thomsen, G . H . (1994). Vg1 and regional specification in vertebrates: A new role for an old molecule. Trends Gene?. 10, 371-376. Wakahara, M. (1986). Modification of dorsal-ventral polarity in Xeriopits luevis embryos following withdrawal of egg contents before first cleavage. Dev.. G r o ~ t hDtfer. 28, 543-554. Wakahara, M . ( 1989). Specification and establishment of dorsal-ventral polarity in eggs and embryos of Xenopus laevis. Detl., Growth Differ. 31, 197-207. Warner. A. E . . Guthrie. S . C., and Gilula, N . B (1984). Antibodies to gap junctional protein selectively disrupt junctional communication in the early amphibian embryo. Nnture (London) 311, 127-131. Weeks, D. L.. and Melton, D. A . (1987). A maternal RNA localized to the vegetal hemisphere in Xenopus eggs codes for a growth factor related to TCF-P. Cell (Cambridge. Mass.) 51, 86 1-867. Weisblat, D. A . , Sawyer, R. T., and Stent, G . S . (1978). Cell lineage analysis by intracellular inj x t i o n of a tracer enzyme. Science 202, t 295- 1298. Wetts, R . . and Fraser. S . E. (1988). Multipotent precursors can give rise to all major cell types of the frog retina. Science 239, 1142-1 145.
138
Sally A. Moody et al.
White, J. G., Horvitz, H. R., and Sulston, J. E. (1982). Neuron differentiation in cell lineage mutants of Caenorhabditis elegans. Nature (London) 297, 584-589. Wilson, P. A., and Melton, D. A. (1994). Mesoderm patterning by an inducer gradient depends on secondary cell-cell communication. Curr. Biol. 4, 676-686. Wood, W. B., Strome, S., and Laufer, J. S. ( 1983). Localization and determination in embryos of Caenorhabditis elegans. In “Time, Space, and Pattern,” (W. R. Jeffery and R. H. Raff, eds.), pp. 221-239. Alan R. Liss, New York. Yuge, M., Kobayakawa, Y., Fujisue, M., and Yamana, K. (1990). A cytoplasmic determinant for dorsal axis formation in an early embryo of Xenopus laevis. Development (Cambridge, UK) 110, 1051-1056. Zanetti, M., Ratcliffe, A , , and Watt, F. M. (1985). Two subpopulations of differentiated chondrocytes identified with a monoclonal antibody to keratan sulfate. J. Cell Biol. 101, 5359. Ziomek, C. A., Johnston, M. H., and Handyside, A . H. (1982). The developmental potential of mouse 16-cell blastomeres. J. Exp. Zool. 221, 345-355. Zisckind, N . , and Elinson, R . P. (1990). Gravity and microtubules in dorsoventral polarization of the Xenopus egg. Dev., Growth Differ. 32, 575-581.
Mechanisms of Programmed Cell Death in Caenorhabditis elegans and Vertebrates Masayuki Miura and junying Yuan Cardiovascular Research Center Massachusetts General Hospital-East Charlestown, Massachusetts 021 29 Department of Medicine Harvard Medical School Boston, Massachusetts 021 15
I. Introduction 11. Programmed Cell Death in the Nematode C. elegans 111. Genetic Control of Programmed Cell Death in C. eleguns
A. Determination of Programmed Cell Death in C. eleguns B. Execution of Programmed Cell Death in C. elegans C. Regulation of ced-3 and red-4 Activity by ced-9 D. Genes Required for Engulfment Process of Dying Cells in C . elegans E. Degradation of Engulfed Cells in C. elegans IV. Molecular Mechanisms of Programmed Cell Death in Vertebrates A. Evolutionary Conservation of Genetic Control of Programmed Cell Death from Worm to Mammal B. Mammalian Homologs of C. elegans Programmed Cell Death Genes C. Programmed Cell Death and Oncogenesis D. Antiviral Responses and Programmed Cell Death V. Do All Cells Have a Suicide Program? VI. Future Prospects References
1. Introduction Programmed cell death is a common phenomenon during animal development. In the early stages of mouse development, soon after the blastmeres differentiate into inner cell mass ( E M ) and trophoectoderm, dead cells are observed among the ICM and trophoblast cells (El-Shershaby and Hinchliffe, 1974). It was estimated that approximately 10% of blastocyst cells die in 95-h postcoitum mouse embryo. Programmed cell death plays a significant role in morphogenesis and histogenesis during animal development (Hinchliffe, 198 I). Well-known examples include cell death in chick limb development and during metamorphosis of the tadpole tail. During neural development, as much as 50% of originally generated neurons die (Cowan, 1984; Hamburger and Oppenheim, 1982; OpCurrrnr T o p ~ ,n s Devclopmmrol Biologv. Vol. 32
Copyright 0 1996 by Academic Press. Inc. All rights of reproduction in any form reserved
139
140
Masayuki Miura and Junying Yuan
penheim, 1991). In the immune system, cell death occurs constantly to eliminate cells that may react against self-antigens (Duvall and Wyllie, 1986; Cohen et al., 1992). Programmed cell death is involved in the generation of specific tissues and organs including kidney (Coles et al., 1993; Koseki et al., 1992), lens epithelial cells (Ishizaki et al., 1993), and cartilage cells (Ishizaki et al., 1994). Thus, programmed cell death during animal development may be as important as cell proliferation, growth, and differentiation. Although the phenomenon of cell death during development was established in the 1950s (Glucksmann, 1951), it attracted the attention of only a small group of embryologists. Little was known about the molecular mechanisms of programmed cell death. The important finding that programmed cell death was strictly determined in cell lineage during development of the nematode Caenorhabditis elegans brought up the idea that programmed cell death is genetically controlled as a specific cell differentiation fate (Sulston and Horvitz, 1977). Intensive molecular genetic studies of programmed cell death in C . elegans have recently led to the identification of the central players of programmed cell death (Ellis e f al., 1991b; Hengartner and Horvitz, 1994b; Yuan et al., 1993). Progress in the studies of molecular mechanisms of cell death strongly suggest that the genes controlling programmed cell death may be evolutionally conserved from worm to mammals (Hengartner and Horvitz, 1994b; Miura et al., 1993; Vaux et al., 1992b; Yuan et al., 1993; Gagliardini et al., 1994; Kumar et al., 1994; Lin et al., 1994). The objective of this chapter is to review genetic and molecular aspects of programmed cell death in C . elegans and vertebrtates for insight into common mechanisms regulating this important biological events.
II. Programmed Cell Death in the Nematode C. eregans Among the 1090 somatic cells generated during C . elegans hermaphrodite development, 131 die (Sulston and Horvitz, 1977; Sulston et al., 1983). These deaths are reproducibly observed in every animal at specific developmental stages and positions. The invariability of the cell fate suggests that the determination and execution of cell death is genetically programmed in C . elegans. A wide variety of cell types including intestinal cells, epithelial cells, muscle cells, neurons, and gonadal cells are programmed to die (Horvitz et al., 1982). The cells undergoing programmed cell death in worm tissues all appear to exhibit similar cell death morphological changes. Viewed with a Nomarski microscope, dying cells condense and acquire a refractile, disk-like shape; cell corpses are eventually engulfed by a neighboring cell. Detailed morphological changes during the programmed cell death in C . elegans have been described by ultrastructual studies (Robertson and Thomson, 1982). Characteristic features include the nuclear condensation, nuclear membrane breakdown, cytoplasm contraction, membranous whorle formation, and the appearance of autophagic vacuoles. These
5 . Cell Death in C. &guns and Vertebrates
141
Fig. 1 The genetic pathway for programmed cell death in C. elegans. Fourteen genes have been identified that affect programmed cell death in C. eleguns. The proposed regulatory interactions between adjacent genes in the pathway arc indicated: arrow, positive interaction; bar, negative interaction. Adapted from Ellis ef a / . (1991b) and Hengartner and Horvitz (1993).
features are similar to the characterisitics of apoptosis observed in mammals (Wyllie, 1981). Interestingly, some of the doomed cells are embraced by engulfment cells before the completion of cell division, suggesting that the decision of cell death is made before the birth of the doomed cells (Roberston and Thomson, 1982). The complete cell death process generally takes only an hour (Sulston and Horvitz, 1977). The morphological similarities of all programmed cell death in C. elegans suggest that there are common molecular mechanisms of programmed cell death. To date, 14 genes that specifically affect programmed cell death have been identified. Among them, 1 1 genes that affect all programmed cell death have been identified (Ellis et d . , I99 1b). Genetic analyses have placed these genes into a developmental pathway of programmed cell death (Fig. 1). This process can be divided into four distinct steps: determination of cells undergoing programmed cell death, execution of programmed cell death, engulfment of dead cells by neighboring cells, and degradation of the engulfed cells. Because the phenomenon of programmed cell death is observed from worm to mammals, what we know about programmed cell death in C. elegans may help us to understand programmed cell death in vertbrate animals.
111. Genetic Control of Programmed Cell Death in C. elegans A. Determination of Programmed Cell Death in C. eregans
Programmed cell death is an integral part of cell lineage development in C. elegans and represents an alternative fate to differentiation. Homeotic mutations
142
Masayuki Miura and Junying Yuan
that cause one body part to develop into a structure normally found elsewhere in the body often change the specification of cell death (Horvitz et al., 1983). One such example is the gene lin-22 (Horvitz er al., 1983). During postembryonic development, five embryonically generated progenitor cells (Vl, V2, V3, V4, V6) on each side of the lateral ectoderm undergo an identical cell lineage while
Wild Type
a
/in-ZZ
V1-5
WildType
DFL DQ+ unc-86 V5.paa
+ DN
I
+T3
X
T
Rg. 2 Homeotic mutations in C.
N
X
X
DN
M
elegans alter specification of cell death. (a) Mutations in !in-22 gene transform four lateral ectoblasts V1-4 to adopt the normal V5 sublineage that include a cell death. (b) Cell lineage reiterations caused by mutations in the unc-86 gene alter the patterns of cell death in V5.paa and Q sublineage. (c) Mutations in the lin-39 gene transform the P ectodermal blast cells P3-8 to adopt a sublineage of normal PI, 2, 9- 11. Adopted and modified from Horvitz et al. (1983) and Clark et al. (1993).
143
5 . Cell Death in C. elegans and Vertebrates Wild Type
C
P1,2,9-11
P3-P8
&
+4-l lin-39 P1,2,9-11
P3-P8
I
I
Fig. 2
(continued)
one cell (V5) undergoes a similar but distinct cell lineage that includes the death of a progeny (Fig. 2a). In the lin-22 mutant animals, Vl-V4 cells are transformed into V5 and express V5 cell lineage, which has an extra cell death (Fig. 2a). Another example is the gene unc-86. The unc-86 mutation causes specific neuroblasts to behave like their own parental precursors (Chalfie er a l . , 1981); i.e., they express the cell lineages that are normally expressed by their parents. As a result, some cell deaths are omitted and some added according to their lineage (Fig. 2b). The unc-86 gene encodes a homeodomain protein with extensive sequence similarity to three mammalian transcriptional factors, Pit-1 , Oct-1 , and Oct-2 (Finney e f a t . , 1988). In wild-type hermaphrodites, the six mid-body cells P(3-8).aap become the hermaphrodite-specific VC neurons and their anterior and posterior lineage homologs P( 1 , 2, 9- 12).aap undergo programmed cell death. In lin-39 mutant animals, P3-P8 cells are transformed to express cell lineages normally expressed by P(l, 2 , 9-12) cells and as a result, the six P(38).aap cells undergo programmed cell death (Fig. 2c). The lin-39 gene encodes a protein with an Antennapedia class homeodomain and is required for mid-body region specific development (Clark er al., 1993). Thus, programmed cell death in C . elegans is incorporated into sublineages to be controlled and executed together with that sublineage. Since the decision of cell death is most likely made through cell lineage, the genes involved in the decision of cell death for individual cells can be considered as cell fate determination genes. Two serotonergic motor neurons, hermaphrodite-specific neuron (HSN), innervating vulva1 muscle cells are required for egg-laying by the hermaphrodites (Desai et a t . , 1988; Sulston and Horvitz, 1977). In males, these two HSN neurons undergo programmed cell death (Sulston and Horvitz, 1977). Dominant
144
Masayuki Miura and Junying Yuan
mutations in egl-I cause the death of HSN neurons in hermaphrodite without affecting sexual development (Ellis and Horvitz, 1986; Trent et al., 1983). As a result, egl-1 mutant animals are unable to lay eggs (Ellis and Horvitz, 1986; Trent et al., 1983). The egl-l gene thus may be involved in the decision of the HSN neurons to undergo programmed cell death. There are two bilaterally symmetric serotonergic NSM motor neurons in pharynges, and the sisters of the NSM neurons undergo programmed cell death. Gain-of-function mutations in ces-1 (ces, cell death specification) or loss-offunction mutation of ces-2 prevent the death of the sisters of the NSM neuron, and as a result, there are two extra serotonin neurons in the pharynx in these mutant animals (Ellis and Horvitz, 1991). These mutations do not affect other programmed cell deaths. Genetic studies suggest that ces-2 inhibits the ces-1 function and ces-1 activates the ced-9 gene, which suppresses the programmed cell death in living cells (Hengartner and Horvitz, 1993). These two genes may control the cell fate of the NSM sisters by regulating the genes involved in programmed cell death.
B. Execution of Programmed Cell Death in C. elegans
Once the cells are determined to die, the genes that are responsible for execution of cell death are activated in these cells. Two genes, ced-3 and cell-4 (cell death abnormal), are both required for all programmed cell deaths in C . elegans (Ellis and Horvitz, 1986). If one of these genes is mutated, all programmed cell deaths are inhibited. Mutations in ced-3 and ced-4 are very cell death specific. The cell lineages of the ced-3 or ced-4 mutant animals are identical with those of wildtype animals except for the absence of all programmed cell deaths. The cell death survivors do not carry out additional cell divisions and some of them differentiate into recognizable cell types. For example, when the cell deaths of two HSN neurons caused by the dominant mutation of egl-I are prevented by mutations in either ced-3 or ced-4, the surviving HSN neurons can differentiate into serotonergic motor neurons, make synapses to the vulva1 muscles, and control egglaying normally (Ellis and Horvitz, 1986). In this case, surviving HSN neurons carry out function as the HSN neurons in wild-type animals. Another example is M4 motor neuron in pharynx, which is essential for worms to feed. The sister of M4 neuron in the wild-type animals undergo programmed cell death. When the normal M4 neuron is ablated by laser microbeam in ced-3 mutant animals, the surviving sister substitutes the function of the normal M4 neuron (Avery and Horvitz, 1987). In another example, the sisters of serotonergic NSM motor neurons that normally die in wild-type animals express serotonin and morphologically differentiate into NSM neurons in ced-3 mutant animals (Ellis and Horvitz, 1986). The overall behavior of ced-3 or ced-4 mutant animals is normal despite the
5. Cell Death in C. elrgaiis and Vertebrates
145
general absence of programmed cell death. The only significant difference is in growth rate. The ced-3 or ced-4 mutant animals reach sexual maturity about 30% slower than wild-type (M. Chalfie, personal communication; J. Yuan, unpublished results). Thus, programmed cell death may be used to increase energy efficiency in animals by removing unnecessary cells. Genetic mosaic analysis was performed to determine the sites of action of ced-3 and ced-4 gene. Such studies demonstrated that the products of both ced-3 and ced-4 most likely function within dying cells (Yuan and Horvitz, 1990). This supports the idea that programmed cell death in C. elegans is suicide, not murder. Other evidence supporting this idea include: ( u ) many dying cells are smaller than their sisters at the time of their birth, suggesting that their cell fates have already been determined; ( b ) most cells die within an hour of their birth without expressing any differentiated phenotype (Sulston et a / ., 1983), suggesting that these cells are unlikely to die as a consequence of a failure in competing for targets; (c-) ablation of engulfing or other neighboring cells with a laser microbeam has failed to rescue these dying cells (J. E. Sulston, personal communication), showing that cells are not killed by their neighbors; ( d ) some of the doomed cells are already embraced by engulfment cells before they are born (Roberston and Thomson, 1982). However, it remains possible that cell-cell interactions or hormonal or trophic signals are required for certain programmed cell deaths. There is at least one known example of ced-3- and ced-4-independent cell death. The death of the male-specific linker cell requires the presence of another cell and occurs long after their birth (Horvitz et al., 1982). This type of cell death can be termed murder, not suicide. The phenotype of red-3 and ced-4 mutant animals are quite similar; however, the dosage dependencies of these two genes are different. Some of the mutations in ced-3 behave as a semidominant suppressor of the egg-laying defect of egl-l (n487)/+ animals, whereas ced-4 mutations behave as a recessive suppressor of this defect (Ellis and Horvitz, 1986). Molecular analyses of ced-3 and ced-4 (see below) revealed the reason behind this apparent difference: all of the ced-4 muations eliminate the expression of ced-4 mRNA, whereas many of the ced-3 mutations are of missense mutations (Yuan and Horvitz, 1992; Yuan et a l . , 1993). The ced-3 gene encodes a 2.8-kb mRNA that is most abundantly transcribed during embryogenesis when most programmed cell deaths occur. The level of ced-3 mRNA is very high (comparable to that of actin I). This suggests that ced-3 may not be transcribed only in dying cells since there are usually no more than two or three cells dying at any given time during embryonic development. Analysis of the ced-3 genomic and cDNA sequence revealed that Ced-3 protein is 503 amino acids in length. The predicted Ced-3 protein is hydrophilic and does not contain any potential signal peptide or transmembrane domain. One region (amino acid 107-205) is rich in serines and is less conserved among related nematode species (C. briggsae and C. vulgaris). The amino-temiinal
146
Masayuki Miura and Junying Yuan
region (amino acid 206-503) is well conserved among the three nematode species (84% identical). All eight EMS-induced ced-3 missense mutations altered amino acids residues that are conserved among the three different nematode species and six of eight mutations alter residues within the C-terminal 100 amino acid region. No mutation is identified within the serine-rich region. These results suggest that the carboxyl half of the Ced-3 protein is important for ced-3 function and that mutations in the serine rich region might not lead to phenotypic changes, or, alternatively, are dominant lethal to animals and thus may not be easily recovered by previous screens. The non-serine-rich regions of the Ced-3 protein have significant homology with human and murine interleukin-1p converting enzyme (ICE) (Cerretti et al., 1992; Thornbeq et al., 1992). ICE, a cysteine protease, cleaves the inactive 31kDa precursor of IL- 1p between Asp- 116 and Ala- 1 17 to release the carboxyterminal 153-amino acid polypeptide from cells as the biologically active mature L I P . The overall amino acid identity between C. elegans Ced-3 protein and human ICE protein is 29%. The carboxy-terminal region of the Ced-3 protein is most similar to ICE. A stretch of 1 15 residues (amino acids 246-360 of Ced-3) is 43% identical between the Ced-3 and human ICE proteins. This region contains a conserved pentapeptide QACRG (positions 361-365 of the Ced-3 protein), which surrounds a cysteine known to be essential for ICE function. Specific modifications of this cysteine in human ICE results in complete loss of activity (Thornbery et al., 1992). The ced-3 mutation n2433 alters the conserved glycine in this pentapeptide and eliminates ced-3 function, suggesting that this glycine is important for ced-3 activity and might be an integral part of the active site of ICE. In 5 of 8 ced-3 mutations identified, single amino acids were altered that are conserved between ICE and Ced-3. Three other mutant alleles of ced-3 have altered amino acids at the very C-terminus of Ced-3 that are not conserved in ICE. Proteolytic cleavage of ICE from a precursor 45 kDa into the P20 and P10 subunits is required for enzymatic activation of ICE. Two of these cleavage sites (Asp/X dipeptide) are conserved in Ced-3. Structure of human ICE has been determined by X-ray diffraction (Wilson et al., 1994; Walker et al., 1994). The catalytic residues and the residues that form the P1 carboxylate-binding pocket are all conserved between ICE and Ced-3. The similarity between Ced-3 and ICE proteins suggest that Ced-3 may function as a cysteine protease in controlling programmed cell death by proteolytically cleaving proteins that are crucial for initiation of cell death or cell viability. Another member of the ICElced-3 gene family, Nedd-2lIch-I, has been cloned from mouse and human (Kumar et al., 1992, 1994; Wang et al., 1994). 'Expression of Nedd-21Ich-1 can induce apoptosis when overexpressed in cultured mammalian cells. Thus, unlike C. elegans ced-3, which controls all programmed cell death, multiple members of the ICE family are likely to act in vertebrates to control cell death. The ced-4 gene has been cloned by transposon tagging using the Tc4 transposable element as a probe (Yuan and Horvitz, 1992). Like ced-3, ced-4 is most
5. Cell Death in C. elegnns and Vertebrates
147
abundantly expressed during embryogenesis. The ced-4 gene is transcribed as a 2.2-kb mRNA, which contains a 549-amino acid open reading frame. The Ced-4 protein is very hydrophilic, and two regions of Ced-4 protein show some similarity to the EF-hand motif, which is known as a calcium-binding domain (Kretsinger, 1987). Although it is not known whether calcium acts in programmed cell death in C . elegans or whether Ced-4 protein binds to calcium, calcium does play a crucial role in apoptosis in mammals (Choi, 1987, 1988; Cohen and Duke 1984; Koike, et a f . , 1989; Leonard and Salpeter, 1979; Schanne et a f . , 1979).
C. Regulation of ced-3 and ced-4 Activity by ced-9
Since the products of ced-3 and ced-4 genes appear either to be cytotoxic themselves or control proteins that are cytotoxic, the activities of ced-3 and ced-4 genes must be tightly regulated during development. The ced-9 gene is involved in this process by repressing ced-3 and ced-4 activities (Hengartner et al., 1992). Normal programmed cell death in C. elegans is prevented by a gain-of-function mutation (n1950) in the ced-9 gene. Like in ced-3;egf-1 double mutants, the deaths of HSN neurons are prevented in egl-l;ced-9 (n1950) animals and egglaying behavior by these animals is normal. This suggests that prevention of HSN death by the ced-9 (n1950) mutation generates functional surviving neurons. Loss-of-function mutations in ced-9 cause ectopic cell deaths that result in embryonic lethality and are suppressed by mutations in ced-3 or ced-4, suggesting that ced-9 normally suppresses the activity of ced-3 or ced-4 to prevent cell death. Expression levels of ced-3 and ced-4 mRNA in embryos are very high (Yuan and Horvitz, 1992; Yuan et al., 1993), suggesting that their expression is not limited to dying cells. Thus, the machinery of programmed cell death is probably present not only in the cells undergoing programmed cell death but also in other living cells as well. The switch of the cell death machinery may be negatively controlled by the gene product of ced-9. The ced-9 genomic region was mapped using restriction fragment length polymorphism, and the ced-9 gene was identified by microinjecting DNA clones from the ced-9 region into ced-9(lf) mutant animals and looking for rescue activity (Hengartner and Horvitz, 1994a). Analysis of gene structure and transcripts of the ced-9 region suggested that ced-9 is in a polycistronic locus. This locus also contains the gene similar to bovine cytochrome b,,,, which is an inner mitochondrial protein involved in system I1 of the mitochondrial electron transport chain (Yu et al., 1992); however, the role of this gene for programmed cell death is unknown. The ced-9 gene encodes a 1.3-kb mRNA that is highly expressed in embryos. The ced-9 mRNA encodes a 280-amino acid protein that is 23% identical to the human Bcl-2 proto-oncogene product. The site of the ced-9 (nl950) gain-of-function mutation has been determined (Hengartner and
148
Masayuki Miura and Junying Yuan
Horvitz, 1994a). A single amino acid substitution (G169E) was found in a region highly conserved among all bcl-2lced-9 family members. However, the same alteration of this glycine in bcl-2 led to a loss-of-function phenotype, suggesting that the function of Bccl-2 is not completely equivalent to that of Ced-9 (Yin et al., 1994). Overexpression of bcl-2 has been shown to protect cells from apoptosis in a number of systems, including deaths of certain hematopoietic cell lines induced by cytokine deprivation and neuronal cell death induced by neurotrophic factor withdrawal (see below). Overexpression of wild-type ced-9 not only prevents ectopic cell death of the ced-9(If) mutant but also prevents many of the programmed cell deaths that are normally observed during development. Overexpression of human bcl-2 can partially prevent ectopic cell death of the ced-9 ( I f ) mutant as well as normal programmed cell death (Hengartner and Horvitz, 1994a; Vaux et al., 1992a). These results strongly suggest that the pathway of programmed cell death that uses products of ced-9lbcl-2 as negative regulators and ced-3IICE as positive regulators has been conserved through evolution.
D. The Genes Required for Engulfment Process of Dying Cells in C. elegans
In C. elegans the dead cells are quickly removed by engulfment cells (usually neighboring hypodermal cells). Six genes (ced-1,2,5,6,7,10) that are involved in phagocytosis for all programmed cell deaths in C. elegans have been identified (Ellis et al., 1991a; Hedgecock et al., 1983). Mutations in any of these genes decrease the efficiency of engulfment and many dead cells persist as corpses for several hours or even days. Engulfment genes may encode cellular signals on the dying cells, receptor molecules for such signals, or molecules required for phagocytosis in the engulfment cells. Mutations in the engulfment genes do not affect any other biological activity; thus, these genes affect specifically the engulfment processes of dying cells. Single mutations in any of these six genes result in partial prevention of engulfment, while the most of the dead cells are still engulfed properly. Certain double mutant combinations of these engulfment genes result in stronger defects in engulfment (Ellis et al., 1991a). Using this analysis, these six genes can be grouped into two sets of genes which are involved in two distinct but partially redundant processes. One set consists of ced-1,6,7 and the other, ced-2,5,10. Genetic studies suggest that ced-1 may encode engulfment-inducing signals by dying cells (Ellis et a f . , 1991a; Hengartner and Horvitz, 1993).
E. Degradation of Engulfed Cells in C. elegans
The nuc-1 gene (nuclease deficient) has been identified as the gene responsible for the degradation of pyknotic DNA of dead cells and bacterial DNA on which
5. Cell Death in C. elegans and Vertebrates
149
C . elegans feeds (Sulston, 1976) . In nuc-1 mutants, determination and initiation of programmed cell death and engulfment of dying cells occur normally, while the pyknotic DNA of engulfed dead cells is not digested. Nuc-1 nuclease does not require a divalent cation for its activity (Hevelone and Hartman, 1988). Fragmentation of DNA into 180-bp nucleosome size fragments is one of the characteristics of apoptosis in mammals and may be mediated by a Ca2+/Mg2+dependent nuclease (Cohen and Duke, 1984; Wyllie, 1980; Peitsch et al., 1993). However, this genomic DNA digestion may not be involved in triggering cell death mediated by cytotoxic T lymphocytes (Ucker et al., 1992). It is not known whether digestion of DNA into nucleosome size fragments is associated with programmed cell death in C . elegans or whether nuc-1 is involved. It is certain, however, that this enzyme is not essential for programmed cell death in C . elegans.
IV. Molecular Mechanisms of Programmed Cell Death in Vertebrates A. Evolutionary Conservation of Genetic Control of Programmed Cell Death from Worm to Mammal
The phenomenon of programmed cell death during development is evolutionarily conserved from invertebrate to vertebrate. The occurrence of massive cell death is not limited to any one type of cells but rather is a general phenomenon during the development of many different types of cells, tissues, and organs of vertebrate animals (Glucksmann, 1951; Saunders, 1966). In C . elegans, most cell deaths occur soon after mitosis or the birth of doomed cells (Sulston and Horvitz, 1977). It appears that the decision of which cell will live and which cell will die has been programmed in the genome of C. elegans. In vertebrates, however, many cell deaths occur long after birth (Oppenheim, 1989, 1991). Extensive neuronal cell death occurs in many regions of the central and peripheral nervous systems during ontogeny. Most neuronal cell deaths occur during the period of synapse formation (Hamberger and Levi-Montalcini, 1949; Hamburger, 1975). The number of surviving neurons seems to depend on the size of their target (Hollyday and Hamburger, 1976; Oppenheim, 1989). Such observations have led to the proposal that neurons compete for limited amounts of target-derived trophic factors and the ones that fail in such competition die (Raff 1992; Raff, et al., 1993). These cell deaths differ from C. elegaiis programmed cell death in an important way: the decision of which cells live and which cells die is reached through cell-cell interactions rather than genetic programming. However, there are examples of programmed cell death in vertebrate animals similar to the type of cell death observed in C . elegans. One such example is neuronal cell death in chick dorsal root ganglion (DRG) (Carr and Simpson, 1982). Up to 14% of degenerating cells in DRG were labeled 2 h after the injection of [rnethyPH]
150
Masayuki Miura and Junying Yuan
thymidine into 5.5-day-old chick embryos, suggesting that these neurons die immediately after they are born. Furthermore, the number of labeled dead cells are not affected by limb bud amputation, and thus these cell deaths are not likely to be caused by failure to establish synaptic connection. Different cell-cell interactions and extracelluar signals may be required for many cells to die in vertebrate animals. In C . elegans, the genes that control the determination of programmed cell death may be different in different cell lineages. Many genes may be involved in the determination of programmed cell death. Despite the difference in mechanisms of determination of programmed cell death in each cell type, morphological features of programmed cell death in C . elegans are quite similar to that of apoptosis in mammals (Wyllie, 1981; Robertson and Thomson, 1982; Rotello et a l . , 1994; Fernandez et al., 1994). This suggests that mechanisms of execution of programmed cell death might be evolutionarily conserved. If so, our knowledge about programmed cell death in C . elegans becomes a powerful tool in the study and exploration of the mechanisms of programmed cell death in vertebrate animals.
B. Mammalian Homologs of C. elegans Programmed Cell Death Genes
1. bet-2 The bcl-2 gene was first identified at the site of translocations common to many human follicular lymphomas, which resulted in overexpression of be/-2 gene (Tsujimoto and Croce, 1986; Tsujimoto et al., 1985). The oncogenic ability of bcl-2 is attributed to its activities as a repressor of programmed cell death (Reed, 1994). The amino acid sequence of C. eleguns Ced-9, which negatively regulates programmed cell death, contains similarities to the Bcl-2 protein (Hengartner and Horvitz, 1994a). Overexpression of human bcl-2 gene in C . elegans suppresses programmed cell death (Vaux et al., 1992b; Hengartner and Horvitz, 1994b). These important findings suggest that bcl-2 is a functional homolog of ced-9 and the mechanisms that control programmed cell death are evolutionarily conserved between C . elegans and mammals. Overexpression of bcl-2 in 1L-3-dependent hemopoietic cells has been shown to inhibit programmed cell death following cytokine deprivation (Vaux et al., 1988). bcl-2 can also prevent apoptosis in response to the removal of IL-4 and granulocyte-macrophage colony-stimulating factor (GM-CSF) (Nunez et al., 1990). Microinjection of a bcl-2 expression construct into nerve growth factor (NGF)-dependent sympathetic neurons (Garcia ef a / ., I992), DRG neurons (Gagliardini et a / ., 1994), and CNS-derived brain-derived neurotrophic factor (BNDF) or neurotrphin 3 (NT-3)-dependent CNS-derived sensory neurons inhibited cell death induced by removal of neurotrophic factors (Allsopp et al., 1993).
5 . Cell Death
in
C. rlegam and Vertebrates
151
Immature thymocytes expressing bcl-2 in bcl-2 transgenic mice were resistant to apoptosis induced by glucocorticoid, anti-CD-3 antibodies, and y-irradiation (Sentman er al., 1991; Strasser et al., 1991). bcl-2 can also inhibit cell death induced by c-myc (Bissonnette et al., 1992; Fanidi et al., 1992), p53 (Wang et al., 1993), tumor necrosis factor (TNF) (Hennet et al., 1993), Fas antigen (Itoh et a/., 1993), and various chemotherapeutic drugs (Miyashita and Reed, 1992, 1993). Overexpression of bcl-2 does not suppress all types of physiological cell death. In transgenic animals where bcl-2 is abnormally expressed in cortical thymocytes, clonal deletion of T cells that recognize endogenous superantigens still OCCUKS, suggesting that negative selection is bcl-2-independent (Sentman et al., 1991; Strasser et al., 1991). Microinjection of bcl-2 into CNS-derived ciliary nerve trophic factor (CNTF)-dependent neurons cannot prevent cell death induced by withdrawal of CNTF (Allsopp et al., 1993). bcl-2 does not protect IL-2-dependent T cell lines, an 116-dependent myeloma cell line, or a WEHI-23 I B cell line from cell death induced by factor withdrawal (Cuende et al., 1993; Nunez et al., 1990). Overexpression of bcl-2 does not prevent cytotoxic T cell killing (Vaux et al., 1992a). These results suggest that there are at least two distinct mechanisms of programmed cell death in vertebrate animals: one that is bcl-2-dependent and one that is bcl-2-independent.
a. Expression Pattern of bcl-2. bcl-2 is widely expressed in embryonic tissues including nervous system, placental trophoblast, and interdigital tissue of limb buds (Hockenbery et al., 1991; LeBrun et al., 1993; Merry et al., 1994; Veis et al., 1993a). In the mouse embryo, bcl-2 is highly expressed in the nervous system (proliferating neuroepithelial cells, postmitotic cells of the cortical plate, cerebellum, hippocampus, and spinal cord). The subplate is a transient layer of cortical plate that disappears after birth (McConell et al., 1989); bcl-2 is highly expressed in the subplate neurons until Postnatal Day 3 and then its level decreases (Merry et a l . , 1994). Such tempera1 pattern of expression is consistent with a role of bcf-2 in the survival of subplate neurons. Postnatally, bcl-2 expression is observed in the granule cells of the cerebellum and dentate gyrus of the hippocampus. The expression of bcl-2 in the CNS declines with aging. In contrast to CNS, bcl-2 expression is retained in PNS neurons and supporting cells of sympathetic and sensory ganglia throughout life. About 60-70% of motor neurons die during the embryonic stage (mostly E13E l 8 in mouse), but continued expression of bcl-2 can be observed before, during, and after the period of cell death. About half of the DRG neurons die during development and bcl-2 is expressed in these neurons throughout the period of cell death (Merry et al., 1994). Although it is not known whether bcl-2 expression is present in dying neurons, the temporal pattern of expression in spinal cord does not suggest a role of bcl-2 in preventing the death of these neurons. The expression of bcl-2 is detected in hormonally responsive epithelia under-
152
Masayuki Miura and Junying Yuan
going cycles of hyperplasia and involution, where programmed cell death is occurring periodically (Hockenbery et al., 1991). In germinal centers of the lymphoid follicle, be/-2 expression is restricted to the follicular mantle, which comprises long-lived recirculating IgM+/IgD+ B cells, and to the portions of the light zone, where selection and maintenance of plasma and memory B cells may take place (Hockenbery et a/., 1991). In constrast, bcl-2 is not expressed from the dark zone of proliferating centroblasts and from the basal portions of the light zone where centrocytes are dying by apoptosis. In the thymus, bcl-2 is present in the surviving T cells in the medulla, whereas most immature cells in the cortex, where cell deaths occur, are negative for bcl-2 expression (Hockenbery et al., 1991; Veis et a/., 1993a). The olfactory epithelium contains the unique neuronal populations where neurogenesis continues even in the adult (Bayer, 1983; Kaplan and Hinds, 1977). Like the epidermis, mature sensory neurons sequentially differentiate from the stem cells. The level of bcl-2 expression is relatively low in the stem cells and subsequently increases during the differentiation of neurons. The differentiating cells of the cortical plate express higher levels of bcl-2 than the stem cells of ventricular zone (Merry et al., 1994). In contrast, bcl-2 is expressed in immature populations including bone marrow progenitors of all lineages, epithelial progenitors in intestine, and epidermis. In the developing kidney, inductive interactions between epithelial and mesenchymal structures are important for morphogenesis and bcl-2 is strongly expressed at the sites of inductive interactions (LeBrun et a / . , 1993). The expression of bcl-2 is detected in many different tissues during many different stages, which suggests that be/-2 may have other functions in addition to preventing developmental cell death. Bcl-2 may be a multifunctional protein with its activity regulated not only transcriptionally but also at the level of posttranslation and protein-protein interactions. The subcelluar localization of Bcl-2 has been intensively studied using subcelluar fractionation, electron microscopy, and confocal laser scanning microscopy (Hockenbery et al., 1990; Jacobson et al., 1993; Krajewski e t a / ., 1993; Monaghan et a/. , 1992). Bcl-2 is an intracellular membrane-bound protein associated with inner and outer mitochondria1 membranes, nuclear envelope, and endoplasmic reticulum. One of the common features of these subcellular compartments is the production of peroxides (Cross and James, 1991). Bcl-2 protein has a stretch of 19 hydrophobic amino acids near the carboxyl terminus. The hydrophobic C-terminus region is important for insertion into membranes, and the association with membrane appears to be necessary for Bcl-2 to prevent cell death (Chen-Levy and Cleary, 1990; Hockenbery et al., 1990; Tanaka et al., 1993).
b. Proteins That Interact with Bcl-2 and the bcl-2 Gene Family. The human ras-related protein (R-ras) p23 was found to interact with Bcl-2 using the yeast two hybrid system (Fernandez-Sarabia and Bischoff, 1993). R-ras could be
5 . Cell Death in C. drgnns and Vertebrates
153
a component in the signal transduction pathway that mediates apoptosis. This could explain why overexpression of R-ras did not lead to cellular transformation as other ras proteins do (Lowe and Goeddel, 1987; Lowe et al., 1988). Another protein, Bax was identified as a 21-kDa Bcl-2-associated protein (Oltvai et al., 1993). The amino acid sequence of Bax protein is 21% identical to that of Bcl-2 and Bax can form homodimers as well as heterodimers with Bcl-2. Expression of bax accelerated apoptotic death induced by cytokine deprivation in an IL-3-dependent cell line. Oltvai et al. (1993) proposed that the ratio of Bcl-2 to Bax may be critical in determining whether a cell undergoes apoptosis. The site specific mutations of Bcl-2 that disrupt interaction with Bax, but not with Bcl-2, abrogated Bcl-2’s death-supressor activity (Yin e f al., 1994). This result suggests that Bcl-2 exerts its action through heterodimerization with Bax. Regulation of Bcl-2 function is complicated. Amino acid substitution of Bcl-2 protein (G 145E) that is equivalent to the gain-of-function mutation of ced-9 (nl950)disrupted heterodimer formation with Bax protein, but G145E mutation allowed the honiodimer formation with Bcl-2. This mutation lost the deathrepressor activity (Yin et ul., 1994). Dimerization with appropriate partner might be crucial for Bcl-2 function, and the manner of dimerzation might be different in worm and mammals. Further, genetic studies of C. elegans suggest that the Ced-9 protein might exist in two distinct forms that have opposit effects on programmed cell death (Hengartner and Horvitz, 1994b). Five other bcl-2 family members (bcl-x, BHRFI, LMWS-HL, mcl-I, AZ) have been identified, in addition to Bcl-2, Bax, and Ced-9. The bcl-x gene is expressed in a wide variety of tissues, with highest levels of mRNA in the lymphoid and central nervous systems (Boise e t a ! . , 1993). bcl-x mRNA is alternatively spliced into two distinct mRNA species. The longer mRNA, bcl-x, encodes a protein that is 56% identical to Bcl-2 in amino acid sequence. The shorter mRNA, bcl-x,, encodes a protein that is an internal 63-amino acid-deleted version of Bcl-x,. Interestingly, bcl-xL and bd-x, have opposite functions: overexpression of bcl-x, enhances the survival of growth factor-deprived cells, whereas overexpression of bcl-x, inhibits the ability of bcl-2 to prevent cell death (Boise et a l . , 1993). The Epstein-Ban- virus gene BHRFl encodes a protein that is 22% identical in amino acid sequence to Bcl-2 (Cleary et al., 1986). Expression of BHRFl can prevent cell death induced by growth factor deprivation in human and murine hematolymphoid cells (Henderson et al., 1993). The LMWS-HL gene of African swine fever virus encodes a protein that is 26% identical in amino acid sequence to Bcl-2 (Neilan et al., 1993). Another bcl-2-related gene, mcl-I, encodes a protein with 35% identity to Bcl-2 (Kozopas et al., 1993). mcl-l is induced in a human myeloid cell line ML-1 cells within 3 h of exposure to a differentiation stimulus (TPA, 12-0-tetradecanoylphorbol-13-aceetate) (Kozopas et al., 1993). A1 is a hematopoietic-specific early response gene that encodes a protein with 40% identity to Bcl-2 (Lin et al., 1993). The functions of mcl-1, LMWS-HL, and A1 are not yet known.
154
Masayuki Miura and Junying Yuan
c. The Functions of the Bcl-2 Protein. Overexpression of bcl-2 can protect cells from y-irradiation-induced cell death, where death is caused by ionizing radiation-produced hydroxyl radicals in aqueous solution by direct radiolytic attack on H 2 0 (Block and Loman, 1986; Sentman et al., 1991; Strasser et al., 1991). Hydroxyl radicals are the most reactive oxygen free radical species, capable of direct oxidative damage to macromolecules including DNA, protein, and membrane lipids. Although it is still not known whether an oxidative burst is an essential step in apoptotic cell death, an important support to this hypothesis came from the result that the overexpression of seleno-enzyme glutatione peroxidase (GSHPx) can protect murine IL-3-dependent cell line FL5.12 from death induced by IL-3 deprivation (Hockenbery et al., 1993). Glutathione (GSH) is an important antioxidant within cells where it is a substrate in the reduction of peroxide catalyzed by GSHPx (Meister and Anderson, 1983). The reaction H20, + 2GSH w F GSSG + 2H,O converts hydrogen peroxide to water with the oxidation of GSH. These results suggest that an important step in apoptosis induced by deprivation of growth factors is the increased production of reactive oxygen species inside cells and that bcl-2 acts by preventing the injury, but not the production, of reactive oxygen species. In another case, however, bcl-2 may prevent cell death by decreasing the cellular generation of reactive oxygen species. Overexpression of bcl-2 prevented the death of GT1-7 neural cells induced by glutathione depletion from treatment with buthionine sulfoximine (BSO) (Kane et al., 1993). In control GT1-7 cells, treatment with BSO results in rapid increase of reactive oxygen species and lipid peroxides, whereas cells overexpressing bcl-2 have a much smaller increase of reactive oxygen species. Thus, bcl-2 may prevent cell death by decreasing the cellular generation of reactive oxygen species. However, an alternative explanation to this result, yet to be ruled out, is that part of the reactive oxygen species may be the result, rather than the cause, of cell death, so that in the presence of excess Bcl-2 protein, cell death is prevented and thus reactive oxygen species are not generated. To study the in vivo function of Bcl-2, bcl-2 deficient mice were created by gene targeting in embryonic stem cells (Nakayama et al., 1993; Veis et al., 1993b). In constrast to ced-9 loss-of-function mutation in C . elegans, these mice complete embryonic development normally. However, the mutant mice display growth retardation and early mortality after birth. Hematopoiesis including lymphocyte differentiation is initially normal, but by 4 weeks after birth, most of their B and T lymphocytes have disappeared from bone marrow, thymus, and periphery. Thymocytes from bcl-2 deficient mice are more sensitive to apoptotic stimulation including glucocorticoid and y-irradiation than normal thymocytes. These results indicate that bcl-2 is dispensable for lymphocyte differentation but is essential for the continued production and maintenance of immune cells. The kidneys of bcl-2-deficient mice display abnormalities characteristic of polycystic kidney disease. Pigmentation of hair coats of bcl-2-deficient mice is initially normal but during the second follicle cycle, bcl-2-deficient mice become hypopigmented and their hair turns gray, suggesting a defect in redox cycles that
5. Cell Death in C. elogans and Vertebrates
155
may regulate melanin synthesis (Prota, 1980). The abnormalities in bcl-2deficient mice are consistent with a function of bcl-2 in an antioxidant pathway. But the abnormalities in melanocyte development in bcl-2-deficient mice are still possible. The exciting discovery of the possible role of bcl-2 in the antioxidant pathway suggests that programmed cell death may play a role in the aging process, since oxidation has long been implicated as a major factor in the aging process (Rusting, 1992). It is worth noting, however, that the absence of programmed cell death in C. efegans did not seem to affect the longevity of the nematode: the life spans of ced-3 and ced-4 mutant animals are normal (Ellis and Horvitz, 1986). Direct evidence of a relationship between programmed cell death and aging in mammals is yet to be obtained.
2. Interleukin-1 P-Converting Enzyme The amino acid sequence of C. elegans Ced-3 protein is homologous with human and murine ICE (Yuan et al., 1993). Crystal structure of human ICE in complex with an inhibitor has been determined (Wilson et al., 1994; Walker et al., 1994). The active site spans both P20 and P10 subunits of ICE. Holoenzyme is a homodimer of catalytic domains formed by the P20iP10 heterodimer. The catalytic residues His-237 and Cys-285, and the residues that form the P1 carboxylate-binding pocket are all conserved with Ced-3. These observations strongly support the hypothesis that ICE and Ced-3 share functional similarity. Overexpression of the murine ICE (mICE) gene or the C. elegans ced-3 gene in Rat-1 fibroblast cells caused programmed cell death (Miura ef al., 1993). Point mutations in a region homologous between mfCE and ced-3 genes eliminated the cell death activities of these genes. The cowpox virus gene crmA can specifically inhibit ICE activity (Ray er al., 1992). The expression of crmA protein can inhibit mfCE-induced programmed cell death, indicating that the protease activity of mICE is essential to its ability to kill cells. Since ced-9 suppresses the activity of ced-3 and ced-4 to protect cells from programmed cell death, ICEinduced cell death by analogy should be suppressed by bcl-2. As expected, overexpression of bcl-2 in Rat-I cells prevents cell death induced by mICE. These results indicate that cell death induced by overexpression of mfCE is likely to be caused by the activation of normal programmed cell death mechanisms in mammals (Miura et al., 1993). In C. elegans hermaphrodites, most programmed cell death is observed in neuronal lineages (105 of 131 programmed cell deaths) (Horvitz et al., 1982). Effects of mfCE expression was tested in vertebrate neuronal cells. Overexpression of mfCE in NG108- 15 neuronal cell line (M. Miura, unpublished results) or chicken DRG neurons can induce cell death (Gagliardini et ul., 1994). These suggest that mICE may act in various cell lineages (including fibroblast and neuron) to cause programmed cell death in vertebrates. The survival of cultured DRG neurons depends on trophic factors (NGF and
156
Masayuki Miura and Junying Yuan
serum). Deprivation of trophic factors from the culture of DRG neurons mimics the normal situation in the developing nervous system, where up to 50% of neurons may die because of failure in competition for trophic factors. When trophic factors are removed from the culture media, DRG neurons die in 3 days or less. Cell death can be prevented by overexpression of bcf-2 or crmA genes (Gagliardini et al., 1994). Thus, the genes in the ICElced-3 family may play a key role in neuronal cell death during normal development. These results suggest that during neuronal development, neural trophic factors may suppress cell death by increasing Bcl-2 activity, which inhibits the ICElced-3 gene family (Fig. 3). Interestingly, the level of I L I has been found to be elevated in the brain of patients with Alzheimer’s disease (Griffin et al., 1989) and the high level of IL- 1 may promote the expression of the amyloid precursor protein (APP) (Forloni et al., 1992). These findings suggest that the ICElced-3 family may contribute to the pathogenesis of Alzheimer’s disease. ICE has been identified as a substrate-specific cysteine protease that cleaves the 31 kDa pro-ILIP between Asp-1 16 and Ala-117 to produce mature 17.5 kDa I L l P (Black et al., 1989; Kostura et al., 1989). ICE activity is required for the production of mature IL-lp. I L l p is produced not only by peripheral blood monocytes but also by astrocytes (Fontana et al., 1982), microglia (Giulian et al., 1986), sympathetic neurons (Freiden et al., 1992), and monocytic leukemia cell line THP. 1 (Cerretti et al., 1992). After murine peritoneal macrophages were stimulated with lipopolysaccharide (LPS) and induced to undergo programmed cell death by exposure to extracelluar ATP, mature active IL-lp was released into culture supernatant; in contrast, when cells were injured by scraping, IL-lp was
Extracellular
lntracellular
fig. 3 Possible relationships of trophic factors, bcl-2, and ICElced-3 family of genes. Trophic factors promote cell survival by increasing EcI-2 activity, which inhibits ICElced-3 family of genes. In the absence of trophic factors, Bcl-2 may be inactivated and consequently, ICElCed-3 family of gene products become activated. The arrows (positive regulation) and bar (negative regulation) do not imply any direct interactions.
5. Cell Death in C. elegans and Vertebrates
157
released exclusively as the unprocessed inactive form (Hogoquist et al., 1991). These results suggest that ICE might be activated on induction of programmed cell death. ICE mRNA has also been detected in a variety of tissues that do not produce IL-1p including resting and activated peripheral blood T lymphocytes, placenta, and B-lymphoblastid line CB23 (Cerretti et al., 1992). The distribution of ICE suggests that ICE may mediate programmed cell death by cleaving substrates other than pro-IL-1p. A third member of the ICElced-3 family, Ich-IlNedd-2, has been identified (Kumar et al., 1992, 1994; Lin et al., 1994). Nedd-2 (NPC expressed, developmentally down-regulated gene) was cloned by subtraction screening of a mouse neural precursor cell (NPC) cDNA library. Nedd-2 gene is highly expressed in developing central nervous system and other tissues including kidney. In adult mice, expression level of Nedd-2 is lower than that in embryo, but its expression can be detected in all tissues. Ich-1 (Ice and ced-3 homolog) was isolated from human fetal cDNA library by screening with 3’-portion of Nedd-2 gene as a probe. Structual and functional analysis revealed that Nedd-2 and Ich-1 are the same gene. Ich-11Nedd-2 protein is similar to both ICE and Ced-3 (approximately 28% identity). Overexpression of Ich-lor Nedd-2 gene in mammalian cells induces apoptosis and this apoptosis is prevented by bcl-2. Ich-1-induced apoptosis can be only marginally inhibited by crmA. This suggests that CrmA is not a general inhibitor for ICEICed-3 family. Ich-I mRNA is alternatively spliced into two different forms. One mRNA species encodes a 435-amino acid sequence protein product, ICH-I,, that is homologous to both the P20 and the P10 subunits of ICE as well as the entire Ced-3. The other mRNA encodes a 3 12amino acid sequence truncated version of Ich- 1 protein, Ich- 1s , that terminates 21-amino acid residues after the pentapeptide QACRG of Ich-1,. Ich-1, and Ich- 1 have oppasite functions: overexpression of Ich-I, induces apoptosis, whereas overexpression of the Ich-I, suppresses Rat-1 cell death induced by serum deprivation. These results suggest that Ich-1 may play important roles in both positive and negative regulation of programmed cell death, as bcl-x gene does in vertebrates. The structural and functional similarity of Ced-3, ICE, and Ich-11Nedd-2 suggests that proteases may play an important role in controlling programmed cell death. In this regard, it is interesting to note that certain protease inhibitors have been found to inhibit programmed cell death. The cysteine protease inhibitors trarzs-epoxysuccininyl-~-leucylamido-(4-guanidino) butane (E-64) and leupeptin, the calpain inhibitor acetyl-leucyl-leucyl-normethional,and the serine protease inhibitors diisopropyl fluorophosphate and phenylmethylsufonyl fluoride all showed dose-dependent blocking of death of a murine T-cell hybridoma, 2B4, induced by anti-CD3 and anti-Thy-1 (Sarin et al., 1993). Among these protease inhibitors, E-64 is known not to inhibit ICE (Thornberry et al., 1992). Thus, it is likely that additional proteases may play key roles in controlling programmed cell death in vertebrate animals.
158
Masayuki Miura and Junying Yuan
Cytolytic granules of cytotixic T lymphocytes (CTL) and natural killer (NK) cells store multiple serine proteases referred to as granzymes or fragmentins (Bleackley et al., 1988; Jenne and Tschopp, 1988; L. Shi et a / . , 1992a,b). One of these serine proteases purified from a rat NK cell line RNK-16, fragmentin 2, can induce DNA fragmentation and apoptosis in YAC-I target cells in the presence of perforin (L. Shi er al., 1992a,b) and is considered the rat ortholog of granzyme B. CTL from the granzyme B knockout mice have defects in their ability to induce rapid DNA fragmentation and apoptosis in allogeneic target cells (Heusel et al., 1994). Fragmentin 2, granzyme B, and ICE share a substrate site specificity for aspartic acid. These proteases could interact with each other (for example, ICE may be activated by fragmentin 2 or granzyme B) to induce apoptosis in mammals. Proteolytic cleavage of poly(ADP-ribose) polymerase (PARP) is associated with apoptosis in various experimental systems (Kaufmann et al., 1993). Cleavage of the 116-kDa to a fragment of 85 kDa is an early event in apoptosis. The cleavage site in PARP is identical to one of the two ICE cleavage sites in proILl p, and a specific tetrapeptide ICE inhibitor efficiently inhibits the proteolytic activity of PARP in a cytoplasmic extract prepared from apoptotic cells (Lazebnik et al., 1994). This protease resembling ICE (prICE) cannot cleave proIL-lP. It is not clear, however, whether PARP cleavage is an essential step in apoptosis.
C. Programmed Cell Death and Oncogenesis
Recently, several oncogenes have been found to have the ability to affect programmed cell death. The high frequency of translocation of chromosome 14 and 18 in human follicular B cell lymphomas, which resulted in overexpression of bcl-2, suggests that prevention of programmed cell death can be oncogenic. Presumably, prevention of programmed cell death can contribute to tumor formation in two ways: first, the absence of cell death may result in accumulation of normally unwanted cells, which by itself may be tumorigenic; second, prevention of cell death may allow cells to live significantly longer than they normally do and, thus, increase the probability that they may harbor an oncogenic mutation during their lifetime and begin abnormal proliferation, which in turn results in tumor formation. In C. elegans, the absence of normal programmed cell death did not result in any additional cell division, suggesting that, at least in C. elegans, cell death is not used to prevent excess cell division. The presence of extra cells in ced-3 and ced-4 animals did not lead to formation of any abnormal structures. Thus, the presence of extra cells per se may not lead to abnormal cell proliferation. This is supported by evidence that overexpression of bcl-2 per se is not oncogenic but bcl-2 promotes the ability of c-myc to cause pre-B cell transformation (Vaux et al., 1988). Many oncogenes have the ability to affect cell death.
5 . Cell Death in C. eleguns and Vertebrates
159
1. p53
Mutations in the tumor suppressor gene encoding p53 are detected in more than 50% of all human cancers. Dominant as well as recessive mutations in p53 gene can be tumorigenic (Levine and Momand, 1990). Recently, it was found thatp53 activates transcription of a 2 1 -kDa protein (Cip 1 / WAFL) that can block the activation of Cdk (cyclin-dependent kinase) enzymes (Harper et al., 1993; ElDeiry et a / ., 1993). The association of cyclin-dependent kinases and cyclin is an essential step for cells to enter S phase (Sherr, 1993), and thus the activation of Cipl/WAFI is likely to be a key step in blockage of cell division by p53. Expression of p53 has been associated with apoptosis under certain conditions. Overexpression of the p53 tumor suppressor gene induced apoptosis in a myeloid leukemic cell line M I , which can be suppressed by IL-6 (Yonish-Rouach et al., 1991). Apoptosis under certain conditions was found to be p53-dependent. Loss of function of the retinoblastoma tumor-suppressor gene (Rb) in mice is associated with unchecked proliferation, impaired expression of differentiation markers, and ectopic apoptosis in lens fiber cells. This ectopic apoptosis is dependent on p53 (Morgenbesser e t a / . , 1994; Pan and Griep, 1944). Rb was inactivated in photoreceptor cells by targeted expression of the HPV-16 E7 gene, which inhibits Rb function. In the presence of p53, photoreceptor cells undergo apoptosis. However, in the absence of p53, mice develop retinal tumors instead of retinal degeneration (Howes et a l . , 1994). Stability of p53 expression is an important step for EIA-induced apoptosis (Lowe et a / . , 1993b). Induction of apoptosis of an 1L-3-dependent myeloid cell line 32D by withdrawal of IL-3 can be inhibited by either p53 antisense or p53 dominant negative mutants (White, 1993). Thymocytes from p53-deficient mice are resistant to apoptosis by radiation but not anti-CD3 antibody or glucocorticoids, suggesting that induction of apoptosis by radiation, but not other forms of cell death, is p53-dependent (Clarke et a / . , 1993; Lowe et al., 1993b). Radiation and chemotherapy are two major forms of cancer therapy that are designed to selectively kill cancerous cells. A major impediment in the treatment of cancer is the unresponsiveness of certain tumor types owing to appearance of resistant cell populations. Treatment of embryonic fibroblasts expressing El A with irradiation or chemotherapeutic compounds rapidly induced apoptosis, whereas there were no effects on the viability of p53deficient E l A-expressing cells (Lowe et al., 1993a). These experiments suggest that cytotoxic effects of radiation and anticancer drugs are mediated through a p53-dependent pathway. It is not clear, however, how p53 is activated by a cell death signal or how activation of p53 induces cells to die. One possiblity is that p53 is activated by DNA damage in apoptotic cells exposed to radiation, since p53 is known to be induced in cells exposed to DNA-damaging agents (Prives and Manfredi, 1993). p53-dependent apoptosis is independent of new RNA or protein synthesis, indicating that p53 induces apoptosis without specific gene activation (Caelles et a l . , 1994).p53 may repress genes necessary for cell survival or be a component of the molecular machinary for execution of apoptosis.
160
Masayuki Miura and Junying Yuan
2. c-myc
In proliferating cells, c-myc expression is induced following mitogenic stimulation. When serum is removed fom the culure media, c-myc is down-regulated and cells arrest in GI phase (Dean et af., 1986). Constitutive, unregulated c-myc expression leads to continued DNA synthesis and apoptosis under growth factordeprivation conditions (Evan et al., 1992). Myc-induced apoptosis, but not cell growth, is prevented by expression of bcf-2 (Bissonnette et ul., 1992; Fanidi et al., 1992). Myc-induced apoptosis is also inhibited by addition of IGF-1 or PDGF to the medium, suggesting that appropriate survival factors can suppress cell death mechanisms induced by c-myc (Allsopp et a f . , 1993; Amati et al., 1993). Expression of several signaling molecules including v-rufand v-ablwere found to suppress apoptosis induced by c-myc (Reed, 1994). These molecules may mimic the intracellular signaling events that occur after the signal of survival is received. Inhibition of endogenous e-myc expression by c-myc antisense oligonucleotides can prevent the activation-induced cell death of T-cell hybridoma (Y. Shi et a / ., 1992). This suggests that expression of endogenous c-myc is required for apoptosis induced by CD3/TCR stimulation. The region of c-myc required for transformation and transcription activation is also responsible for induction of apoptosis. Dimerization of Myc with its partner Max is crucial for both cell cycle progression and induction of apoptosis (Allsopp et a l . , 1993). These suggest that the functions of c-myc in cell proliferation, transformation, and apoptosis are closely related. 3. ras The three mammalian p2lras proteins (Ha-ras, Ki-ras amd N-ras) have been identified as key signal transducers (Boguski and McCormick, 1993). In addition to their well-known effects in promoting cellular transformation and oncogenesis of proliferating cells, ras has profound effects on neurons. When NGF was added to rat pheochromocytoma PC 12 cells extracellularly, PC 12 cells stopped dividing and extended neurites in response to NGF (Green and Tischler, 1976). Overexpression of the ras genes in PC12 cells also induces neurite outgrowth (Bar-Sagi and Feramisco, 1985; Guerrero et a f . , 1986; Noda et ul., 1985). Furthermore, neurite outgrowth can be inhibited via inactivating rus by microinjection of antiras antibodies (Hagag et al., 1986). These results suggest that ras can mimic the function of NGF. Survival of DRG neurons is dependent on NGF. When the oncogene protein T24-ras (or proto-oncogene product c-Ha-ras) is introduced into DRG neurons, neuronal cell deaths induced by NGF deprivation are partially prevented (Borasio et a l . , 1989). Similar effects of ras in promoting neuronal survival were observed in BDNF-responsive nodose ganglion neurons and CNTF-responsive ciliary ganglion neurons. Thus, ras may be involved in the intracellular signaling pathway for survival effects of neurotrophic factors.
5 . Cell Death in C. elegans and Vertebrates
161
One of the lesser known facts about p21ra5 is that part of its amino acid sequence has homology to cystatins (Hisawa et al., 1987). Amino acids 24 to 59 of c-Ha-ras is 61% homologous (19% identical and 42% conservative substitution) to that of rat cystatin p. Cystatin is an endogenous cysteine protease inhibitor. p2Iraqcan inhibit cathepsin L but not cathepsin H or Ca*+-activated neutral cysteine protease (CANP) (Hisawa et al., 1988). Cysteine proteasespecific inhibitors such as N-acetyl-leucyl-leucyl-norleucinal(ALLN) or N-acetyl-leucyl-leucyl-methionial(ALLM) can promote neurite outgrowth of PC12 cells (Hisawa e f a / ., 1990). Thus, it is possible that the part of the ability of ras to prevent cell death may be attributed to its cysteine protease inhibitor activity. 4. c-fos
Expression of the c-fos proto-oncogene has been associated with a variety of biological responses, including proliferation, differentiation, and neural excitation (Morgan and Curran, 1991). In most situations, however, up-regulation of c-fos is transient. Continuous c-fos expression analyzed using fos-lacZ transgenic mice was detected in the cell populations undergoing programmed cell death (Smeyne et al., 1993). fos-lacZ expression was detected in degenerating or regressing ovarian follicular cells, developing heart valves, and at the point of fusion of the nasal septum. After administration of kainic acid, many neuronal deaths were observed among CA 1 and CA3 neurons and expression of fos-lac2 was strongly detected in those regions. The weaver mutant mouse is characterized by the death of cerebellar granule cells (Smeyne and Goldowitz, 1989). Even in the heterozygotes, extensive cell death is observed in cerebellum. foslacZ mice were mated with homozygous weaver mice and F l heterozygotes were analyzed. Cells expressing the fos-lacZ fusion protein were precisely localized where granule cells degenerated in weaver (Smeyne et al., 1993). Because neuronal survival largely depends on target-derived trophic factors, the transsection of sciatic nerve induces neuronal cell death in the spinal cord and DRG. Extensive fos-lacZ expression was detected in motor neurons and DRG neurons after sciatic nerve lesion but not in control animals. Seventy-nine percent of normal rat fibroblasts transfected with c-fos expression constructs die under serum-deprived conditions compared to 7.2% of control fibroblasts that die under the same conditions. Thus, c-fos may be very similar to c-myc in that continuous expression renders cells more sensitive to apoptotic signals and the signal transduction machinery of cell death may be shared with cell proliferation and differentiation. Since c-fos and c-myc can be induced by oxidative stress (Crawford et al., 1988; Shibanuma et al., 1988), expression of these genes might reflect apoptosis mediated by oxidative stress. Transient expression of c-fus and c-jun is also important for cell death. Expression of c-fos and c-jun is rapidly induced after cytokine deprivation in 1L2- and
162
Masayuki Miura and Junying Yuan
IL-6-dependent mouse cell lines. These inductions are transient, and expressions are undetectable at 120 min after factor deprivation. When the expression of c-fos and c-jun are reduced by adding anti-sense oligonucleotides directed against c-fos and c-jun mRNAs, survival of growth factor-deprived lymphoid cells is prolonged (Colotta et al., 1992). The early transient expression of c-fos and c-jun may be involved in initiation of programmed cell death induced by cytokine deprivation. 5. c-re1
The transcription factor c-re/ is a member of the NF-KB family. Expression of v-re1 causes fetal leukemia in birds (Beug et al., 1981) and transformation of avian hematopoietic cells and fibroblasts (Lewis et al., 1981; Morrison et al., 199I). v-rel-transformed bursa1 lymphocytes are resistant to apoptosis induced by radiation, dexamethasone, and calcium ionophore (Neiman et al., 1991). Expression of c-re1 is ubiquitously detected in developing chick embryo. However, the highest level of expression is detected in cells undergoing programmed cell death (Abbadie et al., 1993). c-re1 expressing apoptotic cells were observed in the opaque patch, anterior and posterior necrotic zones, and the interdigital mesenchyme of limb bud and cephalic mesenchyme, the spleen primordia, the gonad primordia, the central nervous system, ganglia, and migrating neural crest cells. When c-re1 was overexpressed in primary avian fibroblasts, cells were transformed and had extended life spans. In contrast, overexpression of c-re1 induced apoptosis in bone marrow cells (Abbadie et al., 1993). Thus, it is possible that, like other cellular oncogenes, c-re1 can regulate both cell proliferaion and cell death. Antioxidants can block the activation of NF-KBby various inducers including tumor necrosis factor and IL-1 (Beg et al., 1993; Schreck et al., 1991; Staal er al., 1990). This suggests that various signals activate NF-KB and activation of NF-KB might lead to the production of reactive oxygen intermediates, which mediates apoptosis. Thus, reactive oxygen intermediates might be the mediators of apoptosis by the NF-KB family.
D. Antiviral Responses and Programmed Cell Death
Animals are under great selective pressure to develop effective defense mechanisms to resist viral infections, whereas viruses have evolved numerous ways to defeat host antiviral responses. Apoptosis might be an important antiviral weapon for animals. By eliminating infected cells, animals could effectively prevent the proliferation of viruses. Accordingly, viruses appear to have developed many ways to antagonize host apoptotic responses. Some of these are borrowed from host cells, as evidenced by the homology of the BHRFZ gene of the Epstein-Barr virus and LMW.5-HL gene of the African swine fever virus to bcl-2.
5. Cell Death in C. elegnns and Vertebratcs
163
Since p53 may mediate an important signal transduction pathway that may activate apoptosis on detection of damaged DNA, some of the viral proteins may prevent cell death by inactivating p53. Simian virus large T antigen and adenovirus El B 55-kDa protein can inactivate p53 by forming heterodimers with p.53 (Debbas and White, 1993; Yew and Berk, 1992). Papillomavirus E6 protein causes rapid ubiquitin-dependent degration of p53 (Scheffner et al., 1993) Adenovirus ElB 19-kDa protein does not bind to p53, but expression of the 19-kDa protein can inhibit apoptosis very efficiently (Debbas and White, 1993). The oncogenic and antiapoptotic activities of E1B 19-kDa protein and Bcl-2 protein are interchangeable. Thus, the E1B 19-kDa protein may be an viral equivalent to Bcl-2 protein. Both EIB 19-kDa and BcI-2 are membrane-associated proteins that can be detected in mitochondria, the endplasmic reticulum, and nuclear membrane (White et a / . , 1984). Both can block apoptosis induced by p.53, TNF, and Fas antigen (Debbas and White, 1993; Hashimoto et a l . , 1991; White et d . , 1992; Yew and Berk, 1992). Although the amino acid sequence of Bcl-2 and El B 19-kDa protein share very limited similarity, both may inhibit cell death through a similar mechanism. The cowpox virus adopted a different strategy to inhibit apoptosis. The cowpox virus cytokine modifier response gene crmA encodes a serpin that can specifically inhibit ICE (Ray et al., 1992). Inactivation of crmA resulted in production of white viral pocks on the chorioallantoic membranes of chick embryos because of influx of large numbers of inflammatory cells (heterophils and macrophages). Inactivation of ICE by viral CrmA may have achieved two goals for the virus by a single protein; it inhibits the release of IL-Ip, a major inflammtory response cytokine, and it prevents the death of infected cells. The insect virus baculovirus encodes two proteins that can inhibit apoptosis (Clem et al., 1991; Crook er al., 1993). The baculovirus protein p35 can inhibit mammalian neural cell death induced by glucose or serum withdrawal or treatment of calcium ionophore (Rabizadeh et al., 1993). Overexpression of the baculovirus p35 gene can prevent programmed cell death in C . elegans (Sugimot0 et nl., 1994) as well as that in Drosophila (Hay et al., 1994). Another baculovirus gene, iap, encodes a 31-kDa protein that can also inhibit cell death induced by viral infection as well as by actinomycin D (Crook et al., 1993). The amino acid sequence of p35 contains no significant homology to any known proteins, whereas inp encodes a zinc finger-like protein. These results suggest that baculovirus may inhibit cell death through a yet uncharacterized pathway.
V. Do All Cells Have a Suicide Program? In C. elegnns, loss-of-function mutations of the ced-9 gene resulted in ectopic cell deaths in every cell lineage, and morphological changes of these ectopic cell deaths are indistinguishable from normal programmed cell death; thus, it appears that machinary for programmed cell death may preexist in every cell. In verte-
164
Masayuki Miura and Junying Yuan
brates, more and more evidence suggests that cell death machinery is constitutively present in many cells. One of the most important pieces of evidence that supports this hypothesis came from work described by Raff ef al. (1993). When nuclei of GM70I or CG4 cells were removed by treatment with cytochalasin and centrifugation, and then cell deaths were induced by the protein kinase inhibitor staurosporine, these enucleated cells underwent apoptosis on the same time course with the same morphology as nucleated cells. Apoptosis of these enucleated cells was prevented by overexpression of bcl-2. These results suggest that the machinery for cell death, as well as suppression of cell death by bcl-2, does not require active gene transcription in these cells.
VI. Future Prospects The structural and functional similarities of Ced-9 and Bcl-2, Ced-3 and ICE suggest that there may be a common molecular pathway of programmed cell death in all metazoans. One may expect the identification of more mammalian programmed cell death genes that share similarities with other C. elegans cell death genes. Isolation of these mammalian genes may allow better biochemical analyses of the programmed cell death pathways, because it is often easier to perform biochemical studies in mammalian systems than in C. elegans. Such studies in turn may increase our understanding of programmed cell death in C . elegans.
Acknowledgments We are grateful to Rocco Rotello, Louise Bergeron, and Nobuhiro Morishima for comments on the manuscript. This work is supported in part by grants to J.Y. from BristollMyer-Squibb and from the National Institute of Aging. M.M. is supported by the Mochida Memorial Foundation for Medical and Pharmaceutical Research and a Fogarty International fellowship from NIH.
References Abbadie, C., Kabrun, N., Bouali, F., Smardova, J., Stehelin, D., Vandenbunder, B., and Enrietto, P. J. (1993). High levels of c-re1 expression are associated with programmed cell death in the developing avian embryo and in bone marrow cells in v i m . Cell (Cambridge, Mass.) 75, 899-912. Allsopp, T. E., Wyatt, S., Paterson, H. F., and Davis, A. M. (1993). The proto-oncogene bcl-2 can selectively rescue neurotrophic factor-dependent neurons from apoptosis. Cell (Cambridge, MUSS.) 73, 295-307. Amati, B., Littlewood, T. D., Evan, G . I., and Land, H. (1993). The c-Myc protein induces cell cycle progression and apoptosis through dimerization with Max. EMBO J. 12, 5083-5087. Avery, L., and Horvitz, H. R. (1987). A cell that dies during wild-type C . elegans development can function as a neuron in a ced-3 mutant. Cell (Cambridge, Muss.) 51, 1071-1078.
5. Cell Death in C. elegans and Vertebrates
165
Bar-Sagi, D., and Feramisco, J. R. (1985). Microinjection of the ras oncogene protein into PC12 cells induces morphological differentiation. Cell (Cambridge, Mass.) 42, 841-848. Bayer, S . A. (1983). [3H]-Thymidine radiographic studies of neurogenesis in the rat olfactory bulb. Exp. Brain Res. 50, 329-340. Beg, A. A., Finco, T. S . . Nantermet, P. V., and Baldwin, J., A. S. (1993). Tumor necrosis factor and interleukin-I lead to phosphorylation and loss of k B a : A mechanism for NF-KB activation. Mol. Cell. Biol. 13, 3301-3310. Beug, H., Muller, H., Doederlein, G . , and Graf, T. (1981). Hematopoietic cells transformed in vitro by REV-T avian reticuloendotheliosis virus express characteristics of very immature lymphoid cells. Virology 115, 295-309. Bissonnette, R. P., Exheverri, F., Mahboubi, A, , and Green, D. R. (1992). Apoptotic cell death induced by c-myc is inhibited by bcl-2. Nature (London) 359, 552-554. Black, R. A , , Kronheim, S. R., and Sleath, P. R. (1989). Activation of interleukin-lp by a coinduced protease. FEES Lett. 247, 386-390. Bleackley, R. C., Lobe, C. G., Duggan, B., Ehrman, N., Fregeau, C . , Meier, M., Letellier, M., Havele, C . , Shaw, J., and Paetkau, V. (1988). The isolation and characterization of a family of serine protease genes expressed in activated cytotoxic T-lymphocytes. Immunol. Rev. 103, 5-19. Block, J., and Loman, H. (1986). Bacteriophage DNA as a model for correlation of radical damage to DNA and biological effects. In “Mechanisms of DNA Damage and Repair” (M. G. Simic, ed.). Plenum, New York. Boguski, M. S. and McCormick, F. (1993). Proteins regulating ras and its relatives. Nature (London) 366, 643-654. Boise, L. H., Gonzalez-Garcia, M., Postema, C. E., Ding, L., Lindsten, T., Thrka, L. A., Mao, X . , Nunez, G . , and Thompson, C. B. (1993). bcl-x, a bcl-2-related gene that functions as a dominant regulator of apoptotic cell death. Cell (Cambridge, Mass) 74, 597-608. Borasio, G . D., John, J., Wittinghofer, A,, Brade, Y.-A,, Sendtner, M., and Heumann, R. (1989). ras p21 protein promote survival and fiber outgrowth of cultured embryonic neurons. Neuron 2, 1087- 1096. Caelles, C., Helmberg, A., and Karin, M. (1994). p53-dependent apoptosis in the absence of transcriptional activation of p53-target genes. Nature (London) 370, 220-223. Cam, V. M., and Simpson, S . B., Jr. (1982). Rapid appearance of labeled degenerating cells in the dorsal root ganglia after exposure of chick embryos to tritiated thymidine. Dev. Brain Res. 2, 157-162. Cerretti, D. P., Kozlosky, C. J., Mosley, B., Nelson, N., Ness, K. V., Greenstreet, T. A., March, C. J., Kronheim, S . R., Druck, T., Cannizzaro, L. A , , Huebner, K., and Black, R. A. (1992). Molecular cloning of the interleukin-lp converting enzyme. Science 256, 97-100. Chalfie, M., Horvitz, H. R. and Sulston, J. (1981). Mutations that lead to reiterations in the cell lineages of Caenoritabditis elegans. Cefl (Cambridge, Muss) 82, 59-69. Chen-Levy, S . , and Cleary, M. L. (1990). Membrane topology of the bcl-2 proto-oncogene protein demonstrated in vitro. J . Biol. Chem. 265, 4929-4933. Choi, D. W. (1987). Ionic dependence of glutamate neurotoxicity. J . Neurosci. 7, 369-379. Choi, D. W. (1988). Calcium-mediated neurotoxicity: Relationship to specific channel types and role in ischaemic damage. Trends Neurosci. 11, 465-469. Clark, S. G., Chisholm, A. D., and Horvitz, H. R. (1993). Control of cell fates in the central body region of C. elegans by the homeobox gene lin-39. Cell (Cambridge, Mass.) 74, 43-55. Clarke, A . R . , Purdie, C. A,, Harrison, D. J., Moms, R. G., Bird, C. C., Hooper, M. L., and Wyllie, A. H. (1993). Thymocyte apoptosis induced by p53 dependent pathways. Nature (London) 362, 849-852. Cleary, M. L., Smith, S. D., and Sklar, J. (1986). Cloning and structural analysis of cDNAs for bcl-2 and a hybrid bcl-2/immunoglobulin transcript resulting from the t( 14;18) translocation. Cell (Cambridge, Mass.) 47, 19-28.
166
Masayuki Miura and Junying Yuan
Clem, R. J., Fechheimer, M. and Miller, L. K. (1991). Prevention of apoptosis by a baculovirus gene during infection of insect cells. Science 254, 1388-1390. Cohen, J. J., and Duke, R. C. (1984). Glucocorticoide activation of a calcium-dependent endonuclease in thymocyte nuclei leads to cell death. J . Immunol. 132, 38-42. Cohen, J. J., Duke, R. C . , Fadock, V. A , , and Sellins, K. S . (1992). Apoptosis and programmed cell death in immunity. Annu. Rev. Immunol. 9, 243-269. Coles, H. S. R., Burne, J. F., and Raff, M. C. (1993). Large-scale normal cell death in the developing rat kidney and its reduction by epidermal growth factor. Development (Cambridge, U K ) 118, 777-784. Colotta, F., Polentarutti, N . , Sironi, M., and Mantovani, A. (1992). Expression and involvement of c-fos and c-jun protooncogenes in programmed cell death induced by growth factor deprivation in lymphoid cell lines. J . Eiol. Chem. 267, 18278-18283. Cowan, W. M. (1984). Regressive events in neurogenesis. Science 225, 1258-1265. Crawford, D., Zbinden, I . , Amstad, P., and Cerutti, P. (1988). Oxidant stress induces the protooncogenes c-fos and c-myc in mouse epidermal cells. Oncogene 3, 27-32. Crook, N. E., Clem, R. J., and Miller, L. K. (1993). An apoptosis-inhibiting baculovirus gene with a zinc finger-like motif. J . Virol. 67, 2168-2174. Cross, A. R., and James, 0. T. G. (1991). Enzymatic mechanisms of superoxide production. Eiochim. Eiophys. Acta 1057, 281-298. Cuende, E., Ales-Martinez, J. E., Ding, L., Gonzalez-Garcia, M., Martinez-A, C., and Ntiiiez, G. (1993). Programmed cell death by bcl-2-dependent and independent mechanisms in B lymphoma cells. EMEO J . 12, 1555-1560. Dean, M., Levine, R. A,, Ran, W., Kindy, M. S . , Sonenshein, G. E., and Campisi, J. (1986). Regulation of c-myc transcription and mRNA abundance by serum growth factors and cell contact. J . Eiol. Chem. 261, 9161-9166. Debbas, M., and White, E. (1993). Wild-type p53 mediates apoptosis by ElA which is inhibited by EIB. Genes Dev. 7, 546-554. 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 (Londonj 336, 638646. Duvall, E., and Wyllie, A. H. (1986). Death and the cell. Immunol. Today 7, 115-1 19. El-Deiry, W. S . , Tokino, T., Velculescu, V. E., Levy, D. B., Parsons, R., Trent, W. M., Lin, D., Mercer, W. E., Kinzler, K. W., and Vogelstein, B. (1993). WAFl, a potential mediator of p53 tumor suppression. Cell (Cambridge, Mass.) 75, 8 17-825. Ellis, H. M., and Horvitz, H . R. (1986). Genetic control of programmed cell death in the nematode C. rlegans. Cell (Cambridge, Mass.) 44,817-829. Ellis, R. E., and Horvitz, H. R. (1991). Two C.elegans genes control the programmed death of specific cells in the pharynx. Development (Cambridge, UK) 112, 591-603. Ellis, R. E., Jacobson, D., and Horvitz, H. R. (1991a). Genes required for the engulfment of cell corpses during programmed cell death in C. elegans. Genetics 112, 591-603 Ellis, R. E., Yuan, J., and Horvitz, H. R. (1991b). Mechanisms and functions of cell death. Annu. Rev. Cell Biol. 7, 663-698. El-Shershaby, A.M., and Hinchliffe, J.R. (1974). Cell redundancy in the zona-intact preimplantation mouse blastcyst: A light and electron microscope study of dead cells and their fate. J . Embryol. Exp. Morphol. 31, 643-654. Evan, G. I . , Wyllie, A. H., Gilbert, C. S . , Littlewood, T. D., Land, H., Brooks, M., Waters, C. M., Penn, L. 2.. and Hancock, D. C. (1992). Induction of apoptosis in fibroblasts by c-myc protein. Cell (Cambridge. Muss.) 69, 119- 128. Fanidi, A,, Harrington, E. A,, and Evan, G . I. (1992). Cooperative interaction between c-myc and bcl-2 proto-oncogene. Nature (London) 359, 554-556. Fernandez-Sarabia, M. J., and Bischoff, J. R. (1993). Bcl-2 associates with the ras-related protein R-ras p23. Nature (London) 366, 214-275.
5. Cell Death in C. elegans and Vertebrates
167
Fernandez, P.-A,, Rotello, R., Rangini, Z., Doupe, A., Drexler. H. C. A., and Yuan, J. (1994). Expression of a specific marker of avian programmed cell death in both apoptosis and necrosis. Proc. Natl. Acud. Sci. U . S . A . 91, 8641-8645. Finney, M., Ruvkun, G . , and Horvitz, H. R. (1988). The C . elegans cell lineage and differentiation gene unc-86 encodes a protein with a homeodomain and extended similarity to transcriptional factors. Cell (Cambridge, Mass.) 55, 757-769. Fontana, A. F., Kristensen, F.. Duds, R., Gemsa, D., and Weber, E. (1982). Production of prostagrandin E and an interleukin-I like factor by cultured astrocytes and C6 glioma cells. J . Immunol. 129, 2413-2419. Forloni, G., Demicheli, F., Bendotti, C., and Angeretti, N. (1992). Expression of amyloid precursor protein mRNAs in endothelial, neuronal and glial cells: Modulation by interleukin-1 . Mol. Brain Res. 16, 128-134. Freiden, M., Bennett, M. V. L.. and Kessler, J. A . (1992). Cultured sympathetic neurons synthesize and release the cytokine interleukin Ip. Proc. Natl. Acad. Sci. U.S.A. 89, 1044010443. Gagliardini, V., Fernandez, P.-A,, Lee, R. K . K., Drexler, H. C., Rotello, R. J . , Fishman, M. C.. and Yuan, J. (1994). Prevention of vertebrate neuronal death by the crmA gene. Science 263, 826-828. Garcia, I . , Martinou, I . , Tsujimoto, Y., and Martinou, J. C. (1992). Prevention of programmed cell death of sympathetic neurons by the bcl-2 proto-oncogene. Science 258, 302-304. Giulian, D., Baker, T. J., Shih. L. N., and Lachman, L. B. (1986). Interleukin 1 of the central nervous system is produced by ameboid microglia. J . Exp. Med. 164, 594-604. Glucksmann, A. (1951). Cell deaths in normal vertebrate ontogeny. Biol. Rev. Cambridge Philos. Soc. 26, 59-86. Green, L. A , , and Tischler, A. S . (1976). Establishment of a noradrenaergic clonal cell line of rat adrenal pheochromocytoma cells which respond to nerve growth factor. Proc. Narl. Acad. Sci. U.S.A. 73, 2424-2428. Griffin, W. S. T., Sranley, L. C., Ling, C., White, L., MacLeod, V., Perrot, L. J., White, C. L. I . , and Araoz, C. (1989). Brain interleukin 1 and S-100 immunoreactivity are elevated in Down syndrome and Alzheimer disease. Proc. Natl. Acad. Sci. U.S.A. 86, 761 1-7615. Guerrero, I . , Wong, H., Pellicer, A , , and Burstein, D. E. (1986). Activated N-ras gene induces neuronal differentiation of PC12 rat phecromocytoma. J . Cell. Physiol. 129, 71-76. Hagag, N., Halegoua, S . , and Viola, M. (1986). Inhibition of growth factor-induced differentiation of PC12 cells by microinjection of antibody to ras p21. Nature (London) 319, 680682. Hamberger, V., and Levi-Montalcini, R. (1949). Proliferation, differentiation and degeneration in the spinal ganglia of the chick embryo under normal and experimental conditions. J . Exp. Zool. 111, 457-502. Hamburger, V. (1975). Cell death in the development of the lateral motor column of the chick embryo. J . Comp. Neurol. 160, 535-546. Hamburger, V., and Oppenheim, R. W. (1982). Naturally occuring cell death in vertebrates. Neurosci. Comment 1 , 39-55. Harper, J. W., Adami, G. R.,Wei, N., Keyomarsi, K . , and Elledge, S . J. (1993). The p21 cdkinteracting protein Cipl is a potent inhibitor of GI cyclin-dependent kinases. Cell (Cambridge. M U S S . )75, 805-816. Hashimoto, S . , Ishii, A., and Yonehara, S . (1991). The E l b oncogene of adenovirus confers cellular resistance to cytotoxicity of tumor necrosis factor and monoclonal anti-Fas antibody. Int. Immunol. 3, 343-351. Hay, 3.A., Wolff, T., and Rubin, G. M. (1994). Expression of baculovirus p35 prevents cell death in Drosophila. Development (Cambridge, U K ) 120, 2121-2129. Hedgecock, E. M . , Sulston, J. E., and Thomson, J. N. (1983). Mutations affecting programmed cell deaths in nematode Caenorhabditis eleguns. Science 220, 1277- 1279.
168
Masayuki Miura and Junying Yuan
Henderson, S . , Huen, D., Rowe, M., Dawson, C . , Johnson, G., and Rickinson, A. (1993). Epstein-Barr virus-coded BHRFI protein, a viral homologue of Bcl-2, protects human B cells from programmed cell death. Proc. Nad. Acad. Sci. U.S.A. 90, 8479-8483. Hengartner. M. O., and Horvitz, H. R. (1993). Programmed cell death in the nematode Caenorhabditis elegans. Exp. Med. 11, 2255-2261. Hengartner, M. O., and Horvitz, H. R. (1994a). C. elegans cell survival gene r e d 4 encodes a functional homolog of the mammalian proto-oncogene bcl-2. Cell (Cambridge, Mass.) 76, 665-676. Hengartner, M. 0.. and Horvitz, H. R . (1994b). Activation of C. elegans cell death protein CED-9 by an amino-acid substitution in a domain conserved in Bcl-2. Nature (London) 369, 318-320. Henganner, M. 0.. Ellis, R.E., and Horvitz, H.R. (1992). Caenorhabditis elegans gene r e d 4 protects cells from programmed cell death. Nature (London) 356, 494-499. Hennet, T., Bertoni, G . , Richter, C . , and Peterhans. E. (1993). Expression of bcl-2 protein enhances the survival of mouse fibrosarcoid cells in tumor necrosis factor-mediated cytotoxicity. Cancer Res. 53, 1456-1460. Heusel, J. W., Wesselschmidt, R. L., Shreata, S . , Russell, J. H., and Ley, T. J. (1994). Cytotoxic lymphocytes require granzyme B for the rapid induction of DNA fragmentation and apoptosis in allogenetic target cells. Cell (Cambridge, Mass.J 76, 977-987. Hevelone, J., and Hartman, P. S . (1988). An endonuclease from Caenorhabditis elegans: Partial purification and characterization. Biochem. Genet. 26, 447-461, Hinchliffe, J. R. (1981). Cell death in embryogenesis. In “Cell Death in Biology and Pathology” (1. D. Brown and R. A. Lockshin, eds.), Chapman & Hall, London. Hisawa, T., Yokoyama, S . , Ha, J.-M., Noguchi, S., and Sakiyama, S . ( I 987). c - H a m s gene products are potent inhibitors of cathepsins B and L. FEES Len. 211, 23-26. Hisawa, T., Sakiyama, S . , Yokoyama, S . , Ha, J.-M., Fujita, J., Noguchi, S . , Bando, Y., Kominami, E., and Katsunuma, N. (1988). Inhibition of cathepsin L-induced degradation of epidermal growth factor receptors by c-Ha-ras gene products. Biochem. Biophys. Res. Commun. 151, 78-85. Hisawa, T., Sawada, T., and Sakiyama, S. (1990). Cysteine protease inhibitors and ras gene products share the same biological activities including transforming activity toward NIH3T3 mouse fibroblasts and the differentiation-inducing activity toward PC 12 rat pheochromocytoma. Carcinogenesis (London) 11, 75-80. Hockenbery, D. M., Nunez, G., Milliman, C., Schreiber, R. D., and Korsnieyer, S. J. (1990). Bcl-2 is an inner mitochondria1 membrane protein that blocks programmed cell death. Nature (London) 348, 334-336. Hockenbery, D. M . , Zutter, M . , Hickey, W., Nahm, M., and Korsmeyer, S. J. (1991). BCL2 protein is topographically restricted in tissues characterized by apoptotic death. Proc. Natl. Acad. Sri. U.S.A. 88, 6961-6965. Hockenbery, D. M . , Oltvai, Z. N., Yin, X.-M., Milliman, C. L., and Korsmeyer, S . J. (1993). Bcl-2 functions in an antioxidant pathway to prevent apoptosis. Cell (Cambridge, Mass.) 75, 241 -25 1. Hogoquist, K.A., Nett, M.A., Unanue, E.R., and Caplin, D.D. (1991). Interleukin 1 is processed and released during apoptosis. Proc. Narl. Acad. Sci. U.S.A. 88, 848558489, Hollyday, M., and Hamburger, V. (1976). Reduction of naturally occuring motor neuron loss by enlargement of the periphery. J . Comp. Neural. 170, 31 1-320. Horvitz, H. R.. Ellis, H. M., and Stemberg, P. W. (1982). Programmed cell death in nematode development. Neurosci. Comment 1, 56-65. Horvitz, H. R., Stenberg, P. W., Greenwald, I. S . , Fixen, W. and Ellis, H. M. (1983). Mutations that affect neural cell lineages and cell fates during the development of the nematode Caenorhabditis elegans. Cold Spring Harbor Symp. Quant. Biol. 48, 453-463.
5. Cell Death in C. elegms and Vertebrates
169
Howes, K. A., Ranson, N . , Papermaster, D. S . , Lasudry, J. G. H., Albert, D. M., and Windle, J. J. ( 1994). Apoptosis or retinoblastoma: Alternative fates of photoreceptors expressing the HPV-16 E7 gene in the presence or absence of p53. Genes Dev. 8, 1300-1310. Ishizaki, Y., Voyvodic, J. T., Bume, J. F., and Raff, M. C. (1993). Control of lens epithelial cell survival. J . Cell Biol. 121, 899-908. Ishizaki, Y., Bume, J. F.. and Raff, M. (1994). Autocrine signals enable chondrocytes to survive in culture. J . Cell Biol. 126, 1069-1077. Itoh, N., Tsujimoto, Y., and Nagata, S . (1993). Effect of bcl-2 on Fas antigen-mediated cell death. J . Immunol. 151, 621-627. Jacobson, M. D., Burne, J. F., and Raff, M. C. (1994). Programmed cell death and Bcl-2 protein in the absence of a nucleus. EMBO, J . 13, 1899-1910. Jacobson, M. D., Burne, J. F., King, M. P., Miyashita, T., Reed, J. C . , and Raff, M. C. (1993). Bcl-2 blocks apoptosis in cells lacking mitochondrial DNA. Nature (London) 361, 365-369. Jenne, D., and Tschopp, J. (1988). Granzymes, a family of serine proteases released from granules of cytolytic T-lymphocytes upon T-cell receptor stimulation. Immunol. Rev. 103, 53-71, Kane, D. .I.,Sarafian. T. A., Anton, R., Hahn, H., Gralla, E. B., Valentine, J. S . , Ord, T., and Bredesen, D. E. (1993). Bcl-2 inhibition of neural death: Decreased generation of reactive oxigen species. Science 262, 1274- 1277. Kaplan, M. S . , and Hinds, J. W.(1977). Neurogenesis in the adult rat: Electron microscopic analysis of light radioautographs. Science 197, 1092-1094. Kaufmann, S . H., Desnoyers, S . , Ottaviano, Y., Ddvidson, N. E., and Poirier, G. G. (1993). Specific proteolytic cleavage of poly(ADP-ribose) polymerase: An early marker of chemotherapy-induced apoptosis. Cancer Res. 53, 3976-3985. Koike, T., Martin, D. P., and Johnson, E. M., Jr. (1989). Role of CaZ+channels in the ability of membrane depolarization to prevent neuronal death induced by trophic-factor deprivation: Evidence that levels of internal Ca2+ determine nerve growth factor dependence of sympathetic ganglion cells. Proc. Natl. Acud. Sci. U.S.A. 86, 6421-6425. Koseki, C . , Herzlinger, D.. and AI-Awqati, Q. (1992). Apoptosis in rnetanephric development. J. CeIl B i d . 119, 1327- 1333. Kostura, M. J., Tocci, M. J., Limjuco, G . , Chin, J., Cameron, P., Hillman, A. G . , Chartrain, N. A , , and Schmidt, I. A. (1989). Identification of a monocyte specific pre-interleukin-1P convertase activity. Proc. Narl. Acad. Sci. U.S.A. 86, 5227-5231. Kozopas, K . M., T., Y . , Buchan, H. L., Zhou, P., and Craig. (1993). MCL-I, a gene expressed in programmed myeloid cell differentiation, has sequence similarity to bcl-2. Proc. Nut/. Acad. Sci. U.S.A. 90, 3515-3520. Krajewski, S . , Tanaka, S . , Takayama, S . , Schibler, M. .I., Fenton, W., and Reed, J. C. (1993). Investigations of the subcellular distribution of the BCL-2 oncoprotein: Residence in the nuclear envelope, endoplasmic reticulum, and outer mitochondrial membranes. Cancer Res. 53, 470 1-47 14. Kretsinger, R. H. (1987). Calcium coordination and the calcium fold: Divergent versus convergent evolution. Cold Spring Harbor Symp. Quant. Biol. 52, 499-5 10. Kumar. S . , Tomooka, Y., and Noda, M. (1992). Identification of a set of genes with developmentally down-regulated expression in the mouse brain. Biochem. Biophys. Res. Commun. 185, 1155-1 161. Kurnar, S.. Kinoshita, M., Noda, M., Copeland, N. G., and Jenkins, N. A. (1994). Induction of apoptosis by the mouse NeddZ gene, which encodes a protein similar to the product of the Caenorhubdiris elegans cell death gene red-3 and the mammalian IL- 1P-converting enzyme. Genes Dev 8, 1613-1626. Lazebnik, Y. A,, Kaufmann, S. H., Desnoyers, S . , Poirier, G. G., and Eamshaw, W. C. (1994). Cleavage of poly(ADP-robose) polymerase by a proteinase with properties like ICE. Nature (London) 371, 346-347.
170
Masayuki Miura and Junying Yuan
LeBrun, D. P., Warnke, R. A,, and Cleary, M. L. (1993). Expression of bcl-2 in fetal tissues suggests a role in morphogenesis. Am. J . farhol. 142, 743-753. Leonard, J. P., and Salpeter. M. M. (1979). Agonist-induced myopathy at the neuromuscular junction is mediated by calcium. J . Cell Biol. 82, 81 1-819. Levine, A. L. and Momand, J. (1990). Tbmor suppressor genes: The p53 and retinoblastoma sensitivity genes and gene products. Biochim. Biophys. Acfu 1032, 119-136. Lewis, R. B., McClure, J., Rup, B., Nissel, D. W., Carry, R. F., Hoelzer, J. D., Nazerian, K., and Bose, J., H. R. (1981). Avian reticuloendotheliosis virus: identification of the hematopoietic target cell for transformation. Cell (Cambridge, Mass.) 25, 421-431. Lin, E. Y., Orlofsky, A,, Bergar, M., and Prystowsky, M. B. (1993). Characterization of A l , a novel hemopoietic-specific early-response gene with sequence similarity to bcl-2. J . Immunol. 151, 1979-1988. Lin, W., Miura, M., Bergeron, L., Zhu, H., and Yuan, J. (1994). Ich-1. an Icelced-3-related gene, encodes both positive and negative regulators of programmed cell death. Cell (Cambridge, Mass.) 78, 739-750. Lowe, D. G . and Goeddel, D. V. (1987).Heterologous expression and characterization of the human R-ras. Mol. Cell. Biol. 7, 2845-2856. Lowe, D. G . , Rickets, M., Levinson, A. D., and Goeddel, D. V. (1988). Chimeric proteins define variable and essential regions of Ha-ras-encoded protein. Proc. Natl. Acad. Sci. U.S.A. 85, 1015-1019. Lowe, S . W., Ruley, H. E., Jacks, T., and Housman, D. E. (1993a). p53-dependent apoptosis modulates the cytotoxicity of anticancer agents. Cell (Cambridge, Muss.) 74, 957-967. Lowe, S . W., Schmitt, E. M., Smith, S. W., Osborne, B. A , , and Jacks, T. (1993b). p53 is required for radiation-induced apoptosis in mouse thymocytes. Nature (London) 362, 847-849. McConell, S . K., Ghosh, A., and Shatz, C. J. (1989). Subplate neurons pioneer the first axon pathway from the cerebral cortex. Science 245, 978-982. Meister, A., and Anderson, M. E. (1983). Glutatione. Annu. Rev. Biochem. 52, 71 1-760. Merry, D. E., Veis, D. J., Hickey, W. F., and Korsmeyer, S. J. (1994). Bcl-2 protein expression is widespread in the developing nervous system and retained in the adult PNS. Development (Cambridge, U K ) 120, 301-311. Miura, M., Zhu, H . , Rotello, R., Hartwieg, E., and Yuan, J. (1993). Induction of apoptosis in fibroblasts by LIP-converting enzyme, a mammalian homolog of the C. elegans cell death gene red-3. Cell (Cambridge, Mass.) 15, 653-660. Miyashita, T., and Reed, J. C. (1992). Bcl-2 gene transfer increases relative resistance of S49.1 and WEH17.2 lymphoid cells to cell death and DNA fragmentation induced by glucocorticoids and multiple chemotherapeutic drugs. Cancer Res. 52, 5407-541 I . Miyashita, T., and Reed, L. C. (1993). Bcl-2 oncoprotein blocks chemotherapy-induced apoptosis in a human leukemia cell line. Blood 81, 151-157. Monaghan, P., Robertson, D., Amos, A. S . , Dyer, M. J. S., Mason, D. Y . , andGreaves, M. F. (1992). Ultrastructural localization of bcl-2 protein. 1. Hisrochem. Cytochem. 40, 1819-1825. Morgan, J. I., and Curran, T. A. (1991). Stimulus-transcription coupling in the nervous system: Involvement of the inducible proto-oncogene fos and jun. Annu. Rev. Neurosci. 14, 421-451. Morgenbesser, S. D., Williams, B. O., Jacks, T., and DePinho, R. A. (1994). p53-dependent apoptosis produced by Rb-deficiency in the developing mouse lens. Nuture (London) 371, 7274. Morrison, L. E., Boehmelt, G., and Enrietto, P. J. (1991). Expression of w-re/ by a replicationcompetent virus: Transformation and biochemical characterization. Oncogene 6, 1657. Nakayama, K.-I., Nakayama, K . , Negishi, I . , Kuida, K., Shinkai, Y . , Louie, M. C., Fields, L. E., Lucas, P. J., Stewart, V., Alt, F. W., and Loh, D. Y . (1993). Disappearance of the lymphoid system in bcl-2 homozygous mutant chimeric mice. Science 261, 1584-1588. Neilan, J., Lu, Z . , Afonso, C . , Kutish, G., Sussman, M., and Rock, D. L. (1993). An african
5 . Cell Death in C. elegnns and Vertebrates
171
Swine Fever virus gene with similarity to the bcl-2 proto-oncogene and the Epstein Barr virus gene BHRF-I. J. Virol. 61, 4391-4394. Neiman, P. E., Thomas, S. J., and Loring, G. (1991). Induction of apoptosis during normal and neoplastic B-cell development in the bursa of Fabricius. Proc. Nutl. Acad. Sci. U.S.A. 88, 5857-5861. Noda, M., KO,M., Ogura, A., Liu, D., Amano, T., Takano, T., and Ikawa, Y. (1985). Sarcoma viruses carrying ras oncogenes induce dufferentiation-associated properties in a neuronal cell line. Nature (London) 318, 73-75. Nunez, G . , London, L., Hockenbery, D. M., Alexander, M., Mckeam, J. P., and Korsmeyer, S. J. ( I 990). Deregulated bcl-2 gene expression selectively prolongs survival of growth factordeprived hemopoietic cell lines. J. Immunof. 144, 3602-3610. Oltvai, Z. N., Milliman, C. L., and Korsmeyer, S . J. (1993). Bcl-2 heterodimerizes in vivo with a conserved homolog, Bax, that accelerates programmed cell death. Cell (Cambridge. Mass.) 74, 609-619. Oppenheini, R. W. (1989). The neurotrophic theory and naturally occuring motoneuron death. Trends Neurosci. 12, 252-255. Oppenheim, R. W. (1991). Cell death during development of the nervous system. Annu. Rev. Neurosci. 14, 453-501. Pan, H . , and Griep, A. E. (1994). Altered cell cycle regulation in the lens of HPV-16 E6 or E7 transgenic mice: Implications for tumor suppressor gene function in development. Genes Dev. 8, 1285-1299. Peitsch, M. C., Polzar, B., Stephan, H . , Crompton, T., MacDonald, H. R., Mannherz, H . G . , and Tschopp, J. ( 1993). Characterization of the endogenous deoxyribonuclease involved in nuclear DNA degradation during apoptosis (programmed cell death). EMBO J. 12, 371-377. Prives, C. and Manfredi, J. J. (1993). The p53 tumor suppressor protein: Meeting review. Genes Dev. I, 529-534. Prota, G. (1980). Recent advances in the chemistry of melanogenesis in mammals. J. Invest. Dermatol. 75, 122- 127. Rabizadeh, S . , LaCount, D. J., Friesen, P. D. and Bredesen, D. E. (1993). Expression of the baculovirus p35 gene inhibits mammalian neural cell death. J . Neurochem. 61, 2318-2321. Raff, M. C. (1992). Social controls on cell survival and cell death. Nature (London) 356, 397400. Raff, M. C., Barres, B . A , , Bume, 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 262, 695-700. Ray, C. A., Black, R. A,, Kronheim, S . R., Greenstreet, T. A,, Sleath, P. R., Salvesen, G. S . , and Pickup, D. J. (1992). Viral inhibition of inflamation: Cowpox virus encodes an inhibitor of the interleukin-I p converting enzyme. Cell (Cambridge. Mass.) 69, 597-604. Reed, J. C. (1994). Bcl-2 and the regulation of programmed cell death. J. Cell Eiol. 124, 1-6. Robertson, A. M. G., and Thomson, J. N. (1982). Morphology of programmed cell death in the ventral nerve cord of Caenorhabditis elegans larvae. J. Embryol. Exp. Morphol. 67, 89- 100. Rotello, R. J. Femandez, P.-A., and Yuan, J. (1994). Anti-apogens and anti-engulfens: Monoclonal antibodies reveal specific antigenes on apoptotic and engulfment cells during chicken embryonic development. Development. 120, 1421-1431. Rusting, R. L. (1992). Why do we age'? Sci. Am. 12, 130-141. Sarin, A,, Adams, D. H., and Henkdrt, P. A. (1993). Protease inhibitors selectively block T cell receptor-triggered programmed cell death in a murine T cell hybridoma and activated peripheral T cells. J. Exp. Med. 178, 1693-1700. Saunders, J. W., Jr. (1966). Death in embryonic systems. Science 154, 604-612. Schanne, F. A. X . , Kane, A. B., Young, E. E., and Farber, 1. L. (1979). Calcium dependence of toxic cell death: A final common pathway. Science 206, 700-702.
172
Masayuki Miura and Junying Yuan
Scheffner, M., Huibregtse, J. M., Vierstra, R. D., and Howley, P. M. (1993). The HPV-16 E6 and E6-AP complex functions as a ubiquitin-protein ligase in the ubiquitination of p53. Cell (Cambridge, Mass.) 75, 495-505. Schreck, R., Rieber, P., and Baeuerle, P. A. (1991). Reactive oxygen intermediates as apparently widely used messengers in the activation of the NF-KB transcription factor and HIV- I . EMBO 1. 10, 2247-2258. Sentman, C. L., Shutter, J. R., Hockenbery, D. M., Kanagawa, O., and Korsmeyer, S. J. (1991). Bcl-2 inhibits multiple forms of apoptosis but not negative selection in thymocytes. Cell (Cambridge, Mass.) 67, 879-888. Sherr, C. J. (1993). Mammalian GI cyclins. Cell (Cambridge, Mass.) 73, 1059-1065. Shi, L., Kraut, R. P., Aebersold, R . , and Greenberg, A. H. (1992a). A natural killer cell granule protein that induce DNA fragmentation and apoptosis. J. Exp. Med. 175, 553-566. Shi, L., Kam, C.-M., Powers, J. C . , Aebersold, R . , and Greenberg, A. H. (1992b). Purification of three cytotoxic lymphocyte granule serine proteases that induce apoptosis through distinct substrate and target cell interactions. J . Exp. Med. 176, 1521-1529. Shi, Y., Glynn, J. M., Guilbert, L. J., Cotter, T. G., Bissonnette, R. P., and Green, D. R. (1992). Role for c-myc in activation-induced apoptotic cell death in T cell hybridomas. Science 257, 212-214. Shibanuma, M., Kuroki, T., and Nose, K. (1988). Induction of DNA replication and expression of proto-oncogenes c-myc and c-fos in quiescent Balb/3T3 cells by xanthinelxanthine oxidase. Oncogene 3, 17-2 1. Smeyne, R. J., and Goldowitz, D. (1989). Development and death of external granular layer cells in the weaver mouse: A quantitative study. J . Neurosci. 9, 1608-1620. Smeyne, R. J . , Vendrell, M., Hayward, M., Baker, S. J., Miao, G. G., Schilling, K., Robertson, L. M., Curran, T., and Morgan, J. I. (1993). Continuous c-fos expression precedes programmed cell death in vivo. Nature (London) 363, 166- 169. Staal, F. J. T., Roederer, M., and Herzenberg, L. A. (1990). Intracellular thiols regulate activation of the nuclear factor K and transcription of the human immunodeficiency virus. Proc. Nail. Acad. Sci. U.S.A. 87, 9943-9941. Strasser, A., Harris, A. W., and Cory, S . (1991). Bcl-2 transgene inhibits T cell death and perturbs thymic self-censorship. Cell (Cambridge, Muss.) 67, 889-899. Sugimoto, A., Friesen, P. D., and Rothman, J. H. (1994). Baculovirus p35 prevents developmentally programmed cell death and rescues a ced-9 mutant in the nematode Caenorhabditis elegans. EMBO J. 13, 2023-2028. Sulston, J. E. (1976). Post-embryonic development in the ventral cord of Caenorhabditis elegans. Philos. Trans. R . Soc. London, Ser. B 275, 287-298. Sulston, J. E., and Horvitz, H. R. (1977). Post-embryonic cell lineages of the nematode Caenorhabdifis elegans. Dev. Biol. 56, 110-156. Sulston, J. E., Schierenberg, E., White, J. G., and Thornson, N. (1983). The embryonic cell lineage of the nematode Caenorhabditis elegans. Dev. Biol. 100, 64- 119. Tanaka, S., Saito, K., and Reed, J. (1993). Structure-function analysis of the apoptosissuppressing Bcl-2 oncoprotein: Substitution of a heterologous transmenbrane domain restores function to truncated Bcl-2 proteins. J. Biol. Chem. 268, 10920-10926. Thornbeny, N. A., Bull, H. G . , Calaycay, J. R., Chapman, K. T., Howard, A D., Kostura, M. J., Miller, D. K., Molineaux, S . M . , Weidner, J. R., Aunins, J., Elliston, K. 0.. Ayala, J. M., Casano, F. J., Chin, J., Ding, G. J.-F., Egger, L. A., Gaffney, E. P., Limjuco, G., Palyha, O., C., Raju, S. M., Rolando, A. M., J. P., S . , Yamin, T.-T., Lee, T. D., Shively, J. E., MacCross, M., Mumford, R. A,, Schmidt, J. A,, and Tocci, M. J. (1992). A novel heterodimeric cysteine protease is required for interleukin - I p processing in monocytes. Nature (London) 356, 768-774.
5 . Cell Death in C. eleguns and Vertebrates
173
Trent, C . , Tsung, N., and Horvitz, H. R. (1983). Egg-laying defective mutants of the nematode Caenorhabditis eleguns. Genetics 104, 6 19-647. Tsujimoto, Y . , and Croce, C. M. (1986). Analysis of the structure, transcripts, and protein products of bcl-2, the gene involved in human follicular lymphoma. Proc. Natl. Acad. Sci. U.S.A. 83, 5214-5218. Tsujimoto, Y . , Crossman, J., Jaffe, E . , and Croce, C. (1985). Involvement of the bcl-2 gene in human follicular lymphoma. Science 228, 1440- 1443. Ucker, D. S., Obermiller, P. S., Eckhart, W., Apgar, J. R., Berger, N. A,, and Meyers, 1. ( I 992). Genomic digestion is a dispensable consequence of physiological cell death mediated by cytotoxic T lymphocytes. Mol. Cell. Biol. 12, 3060-3069. Vaux, D. L.. Cory, S., and Adams, J. M. (1988). Ec/-Z gene promotes haemopoietic cell survival and cooperates with c-myc to immortalize pre-B cells. Nature (London) 335, 440-442. Vaux, D. L . , Aguila, H. L., and Weissman, I. L. (1992a). Bcl-2 prevents death of factordeprived cells but fails to prevent apoptosis in targets of cell mediated killing. f n f . fmmunol. 4, 821-824. Vaux, D. L., Weissman, I . L., and Kim, S. K. (1992b). Prevention of programmed cell death in Cuenorhabdiris elegans by human bcl-2. Science 258, 1955- 1957. Veis, D. J., Sentman, C. L., Bach, E. A,, and Korsmeyer, S. J. (1993a). Expression of the Bcl-2 protein in murine and human thymocytes and in peripheral T lymphocytes. J. fmmunol. 151, 2546-2554. Veis, D. J . , Sorenson, C. M . , Shutter, J. R., and Korsmeyer, S. J. (1993b). Bcl-2-deficient mice demonstrate fulminant lymphoid apoptosis, polycystic kidneys, and hypopigmented hair. Cell (Cambridge. Mass.) 75, 229-240. Walker, N. P. C . , Talanian, R. V., Brady, K. D., Dang, L. C., Bump, N. J., Ferenz. C. R., Franklin, S . , Ghayur, T., Hackett, M. C., Hammill, L. D., Herzog, L., Hugunin, M., Houy, W., Mankovich, J. A., McGuiness, L., Orlewicz, E., Paskind, M., Pratt, C. A , , Reis, P., Summani, A., Terranova, M., Welch, J. P., Xiong, L., Moller, A,, Tracey, D. E., Kamen, R., and Wong, W. W. (1994). Crystal structure of the cysteine protease interleukin-lPconverting enzyme: A (p20/p10), homodimer. Cell (Cambridge, Mass.) 78, 343-352. Wang, L., Miura, M . , Bergeron, L., Zhu, H., and Yuan, J. (1994). Ich-I, an fcelCed-3 related gene, encodes both positive and negative regulators of programmed cell death. Cell (Cambridge, Mass.) 78, 739-750. Wang, Y., Szekely, L., Okan, I., Klein, G.. and Wiman, K. G . (1993). Wild-type p53-triggered apoptosis is inhibited by bcl-2 in a v-myc-induced T-cell lymphoma line. Oncogene 8, 3427343 1 . White, E. (1993). Death-defying acts: A meeting review on apoptosis. Genes Dev. 7, 2277-2284. White, E., Blose, S. H., and Stillman, B. (1984). Nuclear envelope localization of an adenovirus tumor antigen maintains the integrity of cellular DNA. Mol. Cell. Biol. 4, 2865-2875. White, E., Sabbatini, P., Debbas, M . , Wold, W. S. M., Kusher, D. I . , and Gooding, L. R. (1992). The 19-kilodalton adenovirus EIB transforming protein inhibits programmed cell death and prevents cytolysis by tumor necrosis factor a . Mol. Cell. Biol. 12, 2570-2580. Wilson, K. P., Black, J.-A. F., Thomson, J. A , , Kim, E. E., Griffith, J. P., Navia, M. A ,, Murcko, M. A,, Chambers, S . P., Aldape, R. A , , Raybuck, S. A , , and Livingston, D. J. ( 1 994). Structure and mechanism of interleukin- I p converting enzyme. Nature (London) 370, 270-275. Wyllie, A. H. ( 1980). Glucocorticoid-induced thymocyte apoptosis is associated with endogenous endonuclease activation. Nature (London) 284, 555-556. Wyllie, A. H. (1981). Cell death: A new classification separating apoptosis from necrosis. I n “Cell Death in Biology and Pathology.” (1. D. Brown and R. A. Lockshin, eds.), pp. 9-34. Chapman & Hall, London.
174
Masayuki Miura and Junying Yuan
Yew, P. R., and Berk, A. J. (1992). Inhibition of p53 transactivation required for transformation by adenovirus early 1B protein. Nature (London) 357, 82-85. Yin, X . - M . , Oltvai, Z. N . , and Korsmeyer, J. (1994). BHI and BH2 domains of Bcl-2 are required for inhibition of apoptosis and heterodimerization with Bax. Nature (London) 369, 321-323. Yonish-Rouach, E . , Resnitzky. D., Lotem, J . , Sachs, L., Kimichi, A , , and Oren, M. (1 991). Wild-type p53 induces apoptosis of myeloid leukemic cells that is inhibited by interleukin-6. Nature (London) 352, 345-341. Yu, L., Wei, Y.-Y., Usui. S . , and Yu, C.-A. (1992). Cytochrome bsho (QPsl) of mitochondria1 succinate-ubiquinone reductase. J . Biol. Chem. 267, 24508-245 15. Yuan, J., and Horvitz, H. R. (1990). The Caenorhabditis elegans genes red-3 and r e d 4 act cell autonomously to cause programmed cell death. Dev. Biol. 138, 33-41. Yuan, J . , and Horvitz, H. R. (1992). The Caenorharbditis elegans cell death gene ced-4 encodes a novel protein and is expressed during the period of extensive cell programmed cell death. Development (Cambridge. UK) 116, 309-320. Yuan, J., Shaham, S., Ledoux, S . , Ellis, H. M., and Horvitz, H. R. (1993). The C. elegans ceil death gene ced-3 encodes a protein similar to mammalian interleukin- I p converting enzyme. 75, 641-652.
6 Mechanisms of Wound Healing in the Embryo and Fetus Paul Martin Departments of Anatomy & Developmental Biology and Plastic Surgery University College London London W C l E 6BT, United Kingdom
I . Overview and Introduction 11. Adult Wound Healing Review A. Adult Skin Wounds First Seal up with a Fibrin Clot B. Neutrophils and Macrophages Are the Key Players in an Adult Inflammatory Response C. Reepithelialization of Adult Tissues Is by Lamellipodial Crawling D. Contraction of Adult Wound Connective Tissue Is by Specialist Myofibroblasts E. Adult Wound Healing Is Not Perfect and Always Leaves a Scar 111. History of Embryonic Wound Healing Studies
1V. Reepithelialization of an Embryonic Wound Appears To Be Driven by Contraction of an Actin Purse String in the Marginal Epidermal Cells V. Assembly of the Purse String and Mechanism of Purse-String Closure VI. Some Natural Morphogenetic Movements May Be Driven by the Same Contractile Purse String That Closes an Embryonic Wound V11. Contraction of the Embryonic Wound Mesenchyme V111. Role of Cell Proliferation during Embryonic Wound Closure IX. Early Signals for Initiating Tissue Movements of Wound Closure X. Inflammation Does Not Occur Following Wounding in the Embryo XI. Fetal Wound Healing Environment-Extracellular Matrix and Growth Factors A. Collagen B. Fibronectin and Tenascin C. Growth Factors XII. Adult Skin in the Fetal Environment and Vice Versa XIII. Is There a Critical Transition Phase in Late Fetal Development When Healing Becomes Adult-like? XIV. Healing of Tissues Other than the Skin (Not All Fetal Healing is Perfect) XV. Operating on the Human Fetus: Perfect Repair of Embryonic Defects-Realistic Dream or Fantasy? References
1. Overview and Introduction Developmental biologists have long marvelled over the incredible healing capacity of embryos. Within hours of performing a relatively major operation on a tadpole or a chick or mouse embryo the wound will have closed over, usually 111 Dcvri~JpnlenIalBiolog>. Vol. 32 Copyright 0 1996 by Academic Press. Inc. All righlc or reproduction in any (om rcscrvcd
Currenr I b p m
175
176
Paul Martin
leaving very little, if any, indication of where the early incisions had been made. Moreover, scarring-a major feature and problem of adult healing-does not appear to occur. For these reasons and because of the relative ease with which embryonic healing can be studied it has become a favorite paradigm of efficient tissue repair. It is hoped that lessons can be learned from embryonic models that will allow modulation of adult wound healing to make it more efficient. Studies of embryonic and fetal healing are now also of direct clinical relevance because the human fetus has become a surgical patient. A number of pioneering surgeons now operate on human fetuses to correct life-threatening disorders, such as diaphragmatic hernias. These surgeons are keen to understand the biology underlying repair of surgical wounds they make to the fetus in utero. I argue that embryonic wound healing also offers a new model for developmental biologists interested in the cellular machinery used by the embryo to undergo the natural tissue movements of morphogenesis such as gastrulation and neurulation. By studying the cascade of events initiated by wounding and leading to wound closure it should be possible to elucidate the general mechanisms that regulate natural cell and tissue movements in the embryo.
II. Adult Wound Healing Review To allow comparison of mechanisms of embryonic wound healing with wound healing in the adult I will briefly review the phases and key players of adult healing. 1 will focus on the movements of reepithelialization and contraction of the wound granulation tissue since 1 want to draw parallels with these two tissue movements in my later discussion of embryonic healing. Similarly I will discuss the adult inflammatory response since it contrasts with the lack of inflammation during the embryonic repair process.
A. Adult Skin Wounds First Seal up with a Fibrin Clot
At any adult skin wound where there has been damage to blood vessels; leakage of blood results in formation of a temporary plug of aggregated platelets that will rapidly be converted into the definitive clot. Activation of platelets by exposure to thrombin and extravascular collagen fibrils leads to degranulation and release of various factors, including the large adhesive glycoproteins, fibrinogen, fibronectin, von Willebrand factor, and thrornbospondin, which act as a glue binding the platelets to one another and to the damaged blood vessel walls (Terkeltaub and Ginsberg, 1988). An enzymatic cascade results in the conversion of fibrinogen to fibrin, reinforcing the plug and converting it into a fibrin clot, which serves not only to seal vessels but to protect the wound tissues and provide
6. Wound Healing in Embryo and Fetus
177
a provisional matrix scaffold for invasion of inflammatory cells, vascular endothelial cells, wound fibroblasts, and the migrating epidermis. Also released from the (Y granules at the time of platelet activation are a plethora of growth factors including epidermal growth factor (EGF) (Oka and Orth, 1983), the heterodimeric AB form of platelet-derived growth factor (PDGF) (Hammacher et al., 1986), and transforming growth factor P- 1 (TGFP 1) (Assoian and Sporn, 1986), which are likely to play key roles in orchestrating the cell and tissue movements of adult wound repair (Martin et al., 1992).
B. Neutrophils and Macrophages Are the Key Players in an Adult Inflammatory Response
In most adult wounds both neutrophils and monocytes will be attracted to the site of injury by a variety of chemotactic factors including fibrin, matrix degradation products, formyl methionyl peptides released by bacteria, and also by some of the growth factors described above that will have been released at the wound site by degranulating platelets. Both neutrophils and macrophages leave the circulating blood stream by first sticking to endothelial cell walls and then crawling between neighboring endothelial cells and across the basal lamina by a process known as diapedesis. The role of neutrophils is to clear the early rush of contaminating bacteria from the wound by phagocytosis and intracellular killing. If a wound is sterile it will heal quite normally even if the neutrophil invasion is blocked with antineutrophil serum (Simpson and Ross, 1972). The neutrophil infiltration ceases after a few days and expended neutrophils are phagocytosed by macrophages (Newman et al., 1982). Macrophages continue to accumulate at the wound site by recruitment from the circulating pool of blood-borne monocytes, and unlike neutrophils appear to be essential for effective and efficient wound healing. It has been shown in adult guinea pigs that if the macrophage influx is blocked by administration of antimacrophage serum and steroids, then wound healing is severely hindered (Leibovich and Ross, 1975). The roles played by macrophages at the wound site are many and varied. They phagocytose and digest any remaining pathogenic organisms and generally clear up tissue debris as well as releasing large amounts of many different growth factors that will trigger the formation and contraction of granulation tissue and enhance the migration of epidermal cells to cover over the exposed connective tissue surfaces. A study examining mRNAs from small numbers of macrophages isolated from adult mouse wound chambers suggested that activated macrophages express mRNA for transforming growth factor a (TGFa), insulin related growth factor 1 (IGFI) and the A chain of PDGF (Rappolee et al., 1988), and protein studies reveal that at least some of these growth factors are actively secreted by the activated macrophages (Assoian et d.,1987).
178
Paul Martin
C. Reepithelialization of Adult Tissues Is by Lamellipodial Crawling
Within hours of being wounded, epithelial cells lining the wound margin, and also those at the cut mouths of hair follicles in the wound bed, begin migrating forward to cover the exposed wound connective tissues. A single layer of epidermal cells sweeps forward over the defect with leading edge cells extending tongue-like lamellipodia, which they use to drag themselves, and the sheet of epithelium behind them, forward over the underlying substratum (Odland and Ross, 1968; Buck, 1979). The substratum is a provisional matrix consisting of fibrin, fibronectin, various collagens, and tenascin (Clarke et a l . , 1982; Mackie et al., 1988). The precise signals that activate proliferation and crawling of the epidermal wound edge cells are not yet clear but there is good tissue culture evidence that various growth factors normally present at a wound site can modulate epithelial proliferative and migratory activity. In particular EGF and family members have proven to be not only mitogenic for epithelial cells but will also stimulate lamellipodial extension and motility (Barrandon and Green, 1987). These growth factor influences are likely to not all be positive; for example, TGFPs have been shown to block the mitogenic effects of EGF on cultured keratinocytes (Coffey et a l . , 1988). Migrating wound epithelial cells express a specific phenotype suited to forward crawling and different from that in a normal stratified epidermis. Much of this phenotypic transformation is likely to be regulated by the types of integrin receptors expressed by the wound edge epithelial cells. It seems that a6P4 integrins, which are key components of the hemidesmosomal bonds to the basal lamina, are downregulated, whereas a5P 1 integrins, which are fibronectin receptors, are upregulated by activated wound edge cells (Grinnell, 1992).
D. Contraction of Adult Wound Connective Tissue Is by Specialist Myofibroblasts Within 2 or 3 days a small adult lesion will usually have been bridged by the crawling epidermal monolayer. During this period formation of a specialist wound connective tissue called granulation tissue will have commenced through activation, proliferation, and immigration of the normally sessile fibroblasts and blood capillaries at the wound site. Preexisting blood vessels in the wound bed produce vascular sprouts that grow upward, forming loops and coils near the wound surface; this leads to the red appearance of a healing wound. Capillary sprouting is triggered by low oxygen tension at the wound site and by angiogenic factors such as FGFs released by platelets and macrophages. At the same time as the new capillaries are forming there is an influx of fibroblasts into and beneath the plasma clot to fill the wound space underneath the migrating epidermis. This
6. Wound Healing in Embryo and Fetus
179
influx of fibroblasts may be recruited by growth factor signals at the wound site since it is known that TGFP1, PDGF, and the EGFs are all chemotactic to fibroblasts (Pierce et al., 1989; Seppa er al., 1982; Adelmann-Grill et at., I990), whereas FGFs and, to a lesser extent, other growth factors will be mitogenic for the invading fibroblasts. Some of these growth factors have also been shown to regulate expression of genes encoding extracellular matrix molecules (Ignotz et a l . , 1987) or proteolytic modifiers of the matrix (Edwards et a l . , 1987) and thus are likely to be playing key roles in the deposition of granulation tissue. Studies by Clore and colleagues (1979) showed that whereas collagen I is the major isoform in unwounded skin, it is collagen 111 synthesis that is most significantly upregulated at the wound site. During the early phases of granulation tissue formation, migrating fibroblasts exert tractional forces dragging themselves closer toward one another using collagen fibrils as guy ropes, and in this way they initiate connective tissue contraction at the wound site. Further, more powerful, contractile forces are exerted as these fibroblasts differentiate into specialist wound fibroblasts called myofibroblasts and begin to express much of the contractile apparatus otherwise only seen in smooth muscle cells including the a-smooth muscle isoform of actin, which is standardly used as a marker for myofibroblasts (Darby and Gabbiani, 1990). Myofibroblast differentiation appears to be triggered both by mechanical forces at the wound site that are resisting contraction and by growth factor stimuli present within the wound milieu. Desmouliere and colleagues ( I 993) have shown, for example, that exogenous TGFP will stimulate transformation of fibroblasts into wound myofibroblasts. Although it is not absolutely clear how myofibroblasts draw the wound margins toward one another, it seems likely that they do it through a combination of smooth muscle-like contraction of the cells themselves, and tractional remodeling of the collagen matrix utilized by the earlier nonspecialist wound fibroblasts (Grinnell, 1994).
E. Adult Wound Healing Is Not Perfect and Always leaves a Scar
As the wound space is being filled in with granulation tissue and the migrating epidermis is bridging the wound gap, tissue remodeling begins. The monolayer of epidermis begins to stratify from the old wound margin inward and the temporary matrix over which the epidermis has migrated is replaced with a true basal lamina, which is now reconstituted by the epithelial cells themselves (Bard and Sengel, 1984). As this happens the hemidesmosomal links that bond epithelial cells to the basal lamina reassemble (Gipson et u l . , 1988). Unless a cutaneous wound is very shallow then integumentary structures such as hairs and glands never regenerate at the healed site. Initially the tensile strength of the wound is low and it is still heavily vas-
180
Paul Martin
cularized but in subsequent weeks the dense network of wound capillaries regresses and the wound gets stronger by remodeling of the collagen meshwork. This remodeling involves breakdown of early fibrils with collagenases, replacement of much of the type 111 collagen by type I collagen, reorientation of collagen fibrils in response to mechanical stress, and more intermolecular bonding (McPherson and Piez, 1988). However, the end product of adult healing is never perfect regeneration of the original tissue organization and at best a pale scar level with the surrounding skin results. In histological section even a macroscopically faint scar shows aberrant dermal structure with tight bundles of collagen fibrils compared with the looser, reticular pattern of fibrils in unwounded dermis. At worst, if the wound healing process seriously overruns and fibroblast proliferation or matrix synthesis does not stop or the granulation tissue contractile machinery does not switch off promptly, then hypertrophic and keloid scarring, or pathological contracture (as after burn wounds), will result.
111. History of Embryonic Wound Healing Studies Ever since developmental biologists first performed experimental surgery on the embryo, they have made observations about how much better embryonic tissues heal than their adult counterparts. Curiously, one of the earliest papers to focus specifically on wound closure mechanisms reported that chick embryos were unable to repair wounds until feather-forming stages (EIO) and that wounds inflicted at earlier stages simply remained open until E 10 when the embryonic wound healing machinery first became established (Weiss and Matoltsy, 1959). However, all wound healing studies since that first paper show that tissue repair in the early embryo is in fact rapid and efficient and results in almost perfect regeneration of the original tissue organization, without leaving a scar. England and Cowper (1977) made small incisional lesions, 300 p n long, in the endoderm of chick embryos at primitive-streak stages. They showed that these wounds initially gaped open due to the release of tension within the endodermal sheet, but then rapidly closed within 2-3 h by the sweeping forward of endoderm over the underlying mesodermal cells. In a later scanning electron microscopic (SEM) study comparing wound healing in primitive-streak stage chick embryos with that of Xenopus neurulae, Stanisstreet ef al. (1980) reported even more rapid wound closure in frog embryos and suggested that epithelial crawling was not responsible for wound closure because they saw no sign of filopodia or lamellipodia extending from wound front epidermal cells to the adjacent exposed mesoderm. In a histological and SEM study of repair of excisional wounds on the flank of limb-bud stage (E5) chick embryos, Thevenet (1981) showed further evidence for release of epithelial tension at the time of wounding in the form of a
6. Wound Healing in Embryo and Fetus
181
cross-sectional thickening of basal epidermal and peridermal cells close to the wound edge. Moreover, all of her sections showed blunt-faced epidermal cells at the leading edge without lamellipodia, supporting the observation of Stanisstreet el al. (1980) that the wound epidermis does not actively crawl forward over the exposed wound mesenchyme. Thevenet’s (198 1) histology also revealed that mesenchymal cells in the region of the healing wound were more closely packed together than those in neighboring unwounded regions, suggesting that cells had been drawn together and that there was modest contraction of the wound connective tissues accompanying reepithelialization of an excisional wound in the embryo. Furthermore, recent evidence for a mesenchymal as well as an epithelial response to wounding embryonic tissues comes from work on the young Funddus or killifish embryo. Fink and Trinkaus (1988) showed that almost all of the motile deep cells within an 800-km radius of a small wound made in the overlying yolk sac epithelium are triggered to reorientate rapidly and begin migrating toward the wound site within 2 min of wounding. In a series of elegant biochemical perturbation experiments in the early 1980s, Stanisstreet and co-workers tried to dissect out the biochemical basis of embryonic tissue repair processes by manipulating the saline that wounded Xenopus tadpoles swam in and thus the environment bathing the wound: In the first series of experiments cytochalasin-B, a blocker of actin polymerization, prevented wound closure, but colchicine, which blocks microtubule polymerization did not (Stanisstreet and Panayi, 1980). This of course suggests that actin polymerization plays a vital role in some aspect(s) of the embryonic wound healing machinery. Addition to the tadpole saline of EDTA, which mops up divalent ions, or lanthanum, which competes for calcium channels, also inhibited wound closure suggesting that an influx of calcium into wound edge cells might also be a requirement for triggering the tissue movements of wound closure (Stanisstreet, 1982). Unfortunately, a similar series of blocking experiments in cultured neurulae-stage rat embryos failed to produce conclusive results because both cytochalasin-B and calcium chelation caused disassociation of cells of the fragile mammalian embryos at doses that might have been expected to interfere with wound closure (Smedley and Stanisstreet, 1984). The wounds described so far all involved cutting through the basal lamina on which the various embryonic epithelial sheets sit. However, if the basal lamina is not damaged by wounding, then a completely different mode of reepithelialization, more reminiscent of adult healing, is seen. Radice (1980) plucked epithelial cells from the tail of a Xenopus tadpole without damaging the basal lamina and found that basal epithelial cells at the wound edge rapidly extended lamellipodia across the denuded basal lamina and proceeded to crawl into the wound space (Radice, 1980). Interestingly, the most superficial epidermal, or peridermal, cells in these healing tadpole wounds were never seen to migrate actively, but rather rode passively on the migrating basal epidermal cells.
182
Paul Martin
IV. Reepithelialization of an Embryonic Wound Appears To Be Driven by Contraction of an Actin Purse String in the Marginal Epidermal Cells Our own early studies in chick embryos focused on the mechanisms of reepithelialization of a wound. The standard excisional wound we made on the dorsal surface of stage 23 (Hamburger and Hamilton, 1951) chick wing buds involved dissecting away a square of skin approximately 0.5 X 0.5 mm leaving an open wound with no intact basal lamina substratum. Such excisional lesions closed rapidly; by 6 h the area of exposed mesenchyme was significantly reduced and most wounds were healed within 18-24 h (Fig. 1). To get an accurate picture of the time course of healing, we followed the process in living embryos, in ovo, observing the wound margins of single lesions and recording their positions with camera lucida drawings made at various time points during the wound closure process (Fig. 2). By measurement from these drawings, we found that the embryonic wound front moved inward at approximately 10- 15 Fm/h with no observable lag phase until the wound had closed (Martin and Lewis, 1992). Scanning electron microscopy and histological sections through the healing epidermal edges of such wounds confirmed earlier reports that leading cells were not extending the tongue-like lamellipodia used by epidermal cells at adult wound fronts to adhere to substrata and drag themselves forward. Rather, the embryonic wound fronts appeared blunt-edged and appeared to be drawing an adherent basal lamina with them as they swept forward (Fig. 3). The lack of lamellipodial protrusions seemed to rule out the conventional mechanism of cell motility-that of lamellipodial crawling-for closure of most embryonic wounds. How then does the embryonic wound epidermis sweep forward to cover over the wound? We stained whole-mount wounded limb buds with fluorescently tagged phalloidin to reveal the distribution of filamentous actin and used confocal laser scanning microscopy to view thin optical sections through the various epithelial and rnesenchymal layers of the wound. Only in the basal epidermal layer was there any variation from the steady-state distribution of cortical actin. In this basal layer we found a thick cable of actin at the leading edge of the marginal epidermal cells extending almost the full circumference of the wound (Fig. 4). We have proposed that contraction of this cable, like a purse string, provides the motive force to draw the wound epidermis forward (Martin and Lewis, 1992). Since the superficial periderm layer does not form a similar purse string at its leading edge we must presume, like Radice (1980), that the superficial cells ride “piggy-back’’ on their motile basal neighbors. When we look at closure of wounds in mouse embryos, the same actin purse-string machinery appears to be in operation at the basal epidermal wound layer (McCluskey and Martin, 1995), suggesting that this is a universal mechanism for wound closure in early embryos.
Fig. 1 Scanning electron micrographs of healmg wounds o n the dorsum of E4 chick wing buds. (A) Low magnification view of a “0 hour” wound after a square of skin has been excised. (B) Higher magnification view of the cut epidermal edge from a similar “0 hour” wound. The exposed mesenchyma1 cells on the right appear to bc expressing numerous blebs on their apical surface. (C) Six hours after wounding the epidermal margins are closing in. (D) A higher magnification detail of (C) showing epidermal cells stretched out around the wound margin and exposed mesenchymal cells now flattened down. (E) Eighteen hours after wounding the defect is almost closed. (F) A higher magnification view of a similarly almost closed wound showing the epidennaf cells beginning to pile up at the focal point of closure. Scale bars: A. C, E, 500 pm; B, D, F, 100 pm.
184
Paul Martin
3
Fig. 2 Camera lucida drawings of a single living specimen, showing the time course of healing of a square excisional wound on the dorsum of the E4 chick wing bud (as in Fig. 1A). The shaded area represents the area of mesenchyme remaining exposed at each time point.
V. Assembly of the Purse String and Mechanism of Purse-String Closure Square excisional lesions of the type described above take some 5 to 10 min to create, which makes it difficult to look at the very early stages of healing (within seconds and minutes after wounding). To circumvent this problem we have more recently made incisional lesions (a simple linear cut created by a tungsten needle) on the dorsum of the stage 23 chick wing bud and observed them healing. As with the excisional wounds a cable of actin assembles at the margin of the wound in the basal epidermal cells. This cable begins to be visible within 2 min after wounding and is close to its maximal thickness within 10 min; colocalizing with it are the myosin I1 motors needed to drive contraction of the purse string (Brock et al., in preparation). The prompt formation of the cable in response to wounding suggests that it depends on redeployment of existing actin, although an induction of actin gene expression at the transcriptional or translational level might be necessary for the cable to reach its mature size. Similarly rapid reorganizations in the actin cytoskeleton of serum-starved tissue culture fibroblasts on exposure to serum factors have been shown to be mediated by the ras-related, small GTPbinding protein, rho (Ridley and Hall, 1992); it will be interesting to discover
Fig. 3 Transmission electron micrograph of a healing epidermal front 12 h after wounding the dorsal skin of an E4 chick wing. An intact basal lamina extends right up to the leading edge cell (arrow), which is blunt-faced and shows no sign of lamellipodial extensions. Scale bar: 2 pm.
Fig. 4 Confocal laser scanning micrographs of the healing chick wound 12 h after wounding. (A) Low-power view of the wound showing a cable of actin extending the full margin of the wound. (B) High-power view of the wound in B focusing on the superficial periderm layer, showing pavementlike cells but no actin specialization at the wound margin. (C) Same magnification and region of wound margin as (B), but focused deeper, in the plane of the basal ectodermal cells. A clear cable of actin (brightly stained with FITC-phalloidin) is apparent at the leading edge of the front row cells. Nuclei have also been stained using 7AAD dye. Scale bar: A, 200 pm; B, C, 20 pm.
187
6. Wound Healing in Embryo and Fetus
whether an early “wound signal” acts via rho to mediate actin purse-string assembly in the embryonic wound front epithelial cells. It seems unlikely that the primary wound signal initiating formation of the embryonic wound healing machinery is simply exposure to serum factors, since these are presumably not at limiting levels in the rapidly proliferating embryonic epithelium prior to wounding. Rather, we suspect that stretching of cells at the epithelial margin, as tension within the epithelial sheet is released and the wound gapes, might provide a mechanical wound signal. In support of this hypothesis are some experiments by Kolega (1986), where he bodily stretched cultured fish epidermal cells at right angles to their direction of movement (tangential to the leading edges of their lamellipodium) and found that they retracted their lamellipodia and reorganized their actin into a cable oriented along the major axis of stress. As an embryonic wound closes, the length of epidermal free edge around the perimeter decreases. This might be expected to correlate with a narrowing of the segment of perimeter formed by each cell, converting all marginal cells to an inward-pointing wedge shape. But in fact, the average shape of the cells around the wound margin does not appear to change significantly as the wound closes, implying that wound closure must involve changes in cell-cell contacts, and not just changes in cell shape. We propose that tension in the actin cable drives cells to rearrange their contacts and lose their place in the marginal row so that the number of cells in the marginal row gets progressively smaller as the wound closes (see Fig. 5). Our incisional wounds show this phenomenon in an extreme form: they appear to heal by “zipping-up’’ at their ends (Brock et al.,in preparation).
A
B
Fig. 5 A diagrammatic impression of how we believe cel1:cell relationships must be changing at the wound margin in order to allow the wound perimeter to get smaller and the wound to close. (A) The wound shortly after it has been made with all the epidermal cells bordering the “0 hour” wound margin shaded. (B) Several hours later the wound has closed somewhat and shaded cells that were once at the wound margin have been forced out into rows further back.
188
Paul Martin
In support of the capacity of contractile purse strings to drive cel1:cell rearrangements within the wound front epithelium are some studies of a gut epithelial cell line, Caco-2BBE. Monolayers of these cells heal their wounds not by lamellipodial crawling across the petri dish but, like in situ embryonic wounds, by purse-string closure. Actin and myosin I1 are both localized as a cable in the leading edges of wound marginal cells (Bement et al., 1993). Elegant time-lapse studies of such healing wounds directly reveal cells that were initially in the front row being extruded into rows further back as closure proceeds (Bement et ul., 1993). Clearly, for a contractile actinomyosin purse string to operate in the wound front epithelilal cells, other proteins including those involved in stabilizing actin and myosin filaments and the adherens junction proteins that serve as anchor points for the intracellular segments of actin cable must be reorganized appropriately. Interestingly, an experiment on the early Xenopus blastula suggests that other cytoskeletal elements, notably the epithelial intermediate filament keratins, might be essential also: Torpey et al. (1992) show that antisense oligonucleotide depletion of keratin 8 mRNA will completely block keratin filament formation in the Xenopus embryo. Such embryos fail to close a simple slash wound that untreated embryos close within minutes. Treated embryos also seem unable to gastrulate properly, suggesting a link between the movements of reepithelialization of a wound and those of gastrulation because of their shared requirement for keratin filaments.
VI. Some Natural Morphogenetic Movements May Be Driven by the Same Contractile Purse String That Closes an Embryonic Wound Since it is unlikely that amniote embryos would naturally be wounded, it is probable that there is no specialized machinery for wound closure in embryos. It seems more likely that the tissue movements that close an embryonic wound are simply adaptions of the normal tissue movements of morphogenesis. For example, folding of the neural plate into a tube almost certainly utilizes contractile actin filaments analogous to those discussed above that drive reepithelialization of wounds. Indeed, a number of authors have shown bundles of microfilaments in the apex of neurectoderm cells at the time when the neurectoderm is folding up (Baker and Schroeder, 1967; Burnside, 1971). Morriss-Kay and Tuckett (1985) went on to show that formation of these bundles was blocked by treating cultured rat embryos with cytochalasin-D and that these embryos subsequently failed to complete cranial neurulation. In a recent paper, Young et al. (1993) showed elegant genetic proof that contractile actinomyosin purse strings play key roles in normal morphogenesis. They described the morphogenetic movement of dorsal closure in Drosophilu,
6 . Wound Healing in Embryo and Fetus
189
whereby the dorsal epidermis closes over the amnioserosa of the embryo. In the cells at the leading edge of the epidermis an actin cable assembles, just like that seen in epidermal wound marginal cells, and is again associated with myosin 11. In this system, there is strong genetic evidence that active contraction of the actin cable is required for closure, since closure fails in the zipper mutant, which lacks zygotic myosin 11.
VII. Contraction of the Embryonic Wound Mesenchyme Although few would question that embryonic and fetal wounds are able to reepithelialize rapidly and efficiently, there has been much debate about whether they also undergo connective tissue contraction. Early studies in the rabbit suggested that open excisional wounds in the fetus did not contract (Somasunderam and Prathrap, 1970; Krummel et al., 1989). However, contraction did occur if the wound was covered by a silastic patch (Somasunderam and Prathrap, 1972). Moreover, rabbit fetal fibroblasts were shown to contract actively a collagen gel in vitro (Krummel et af., 1989), suggesting that the lack of contraction in vivo was due to some inhibitory factor in the amniotic fluid bathing the open wound. It has been suggested that this unusual trait of rabbit amniotic fluid might be explained by the presence of unusually large amounts of maternal immunoglobulins because of the size of the rabbit zona chorion during placentation (Whitby and Ferguson, 1991a; from Wild, 1965). Open wounds in late-stage sheep fetuses have been clearly shown to undergo contraction, and transmission electron microscopy studies of these wounds have revealed the presence of myofibroblasts within the contracting wound connective tissue (Longaker el al., 1991a; Estes et al., 1994). In our own in ovo studies of wound healing in chick embryos we labeled small groups of exposed mesenchymal cells at the wound margin with a lipophilic dye, DiI. This allowed us to trace mesenchymal movements during the wound closure period and revealed a significant contraction of the initially exposed mesenchyme by the time the wound had become fully reepithelialized (Martin and Lewis, 1992). Similar experiments on wounded, limb-bud stage (El 1 . 3 , mouse embryos grown in culture show that the initially exposed wound mesenchyme contracts to about 50% of its original area by the time the wound has fully closed, suggesting that reepithelialization and connective tissue contraction contribute equally to the total wound closure effort (McCluskey and Martin, 1995). However, our studies suggest that embryonic connective tissue contraction does not require conversion of fibroblasts to myofibroblasts, as we see no cells staining positive for the myofibroblast marker a-smooth muscle actin at the healing wound site. In an organ culture model of fetal rat skin wound healing lhara and Motobayashi (1992) enzymatically separated the epithelial and connective tissue layers of a full-thickness excisional wound and clearly showed significant contractile capacity of the connective
190
Paul Martin
tissue component of the wound, even in the absence of an overlying epithelium. This contraction was completely blocked in the presence of cytochalasin-B, confirming that, like reepithelialization, it was dependent on the actin cytoskeleton.
VIII. Role of Cell Proliferation during Embryonic Wound Healing Embryos are growing at a significant pace throughout the developmental period and so both epithelial and mesenchymal cells must be proliferating rapidly even in unwounded tissue. Even given the high background levels of cell proliferation, there is some evidence from wounds made to the flank of E5 chick embryos that both cell types are locally triggered to proliferate even more rapidly around the wound site, leading to the suggestion that this aids wound closure (Thevenet, 1984). One of the few studies of wound healing in Drosophilu embryogenesis shows a burst of cell proliferation by epithelial cells at the margin of wounds made to imaginal wing discs (Dale and Bownes, 1980). However, since in both of these models the tissue movements of healing begin very rapidly after wounding, it seems unlikely that the relatively slow upregulation in proliferative capacity in response to any possible wound mitogen could be directly responsible for driving these movements, at least in the early stages of wound closure. Moreover, an organ culture study of wound healing in fetal (E16) rat skin provides evidence that cell proliferation might not be necessary at all for wound closure: Ihara and Motobayashi (1992) completely block cell proliferation with the drug hydroxyurea and show that closure of a 1-mm-diameter wound is as rapid and complete in tissues that are blocked from cell division as in those that are untreated. Their sections through healing in vitro wounds show that both epithelial and connective tissues adjacent to the wound are composed of fewer cell layers in drug-treated specimens, suggesting that hydroxyurea-treated tissues are able to compensate by stretching fewer cells further.
IX. Early Signals for Initiating Tissue Movements of Wound Closure Soon after damage to the adult skin, degranulating platelets (leaking from cut vessels) and invading macrophages are believed to initiate and orchestrate the tissue movements of reepithelialization and connective tissue contraction, as I have described above. However, in early embryos there are no platelets; megakaryocytes, the progenitors of platelets, do not begin to differentiate until E13.5 in the mouse (Rugh, 1990) and, as will be discussed later, the inflammatory
6. Wound Healing in Embryo and Fetus
191
response to a wound in the early embryo is weak or nonexistent. So where do the signals that initiate the movements of wound closure in embryos come from? We have begun looking at expression of some of the immediate-early transcription factors at the wound site and find that one of these, the proto-oncogene, c-fos, is rapidly upregulated in wounded limb-bud stage (El 1.5) rat embryos (Fig. 6). We found that c-fos, which is not normally expressed by the embryonic skin, was induced in a band of cells about four to five cells wide at the epithelial wound margin and, to a lesser extent, by the superficial exposed wound mesenchyme cells (Martin and Nobes, 1992). Fos protein was apparent at the wound within 15 min of wounding, with levels peaking at about 1 h and by 4 h it was all gone. Although c-fos itself may not be the only immediate-early transcription factor upregulated at the wound site, this study does show how downstream effector genes, such as growth factors, which are probably necessary to activate the movements of wound closure, do not need to be brought to the wound site by platelets and macrophages, but might simply be transcriptionally activated in the resident epithelial and connective tissue cells present at the wound site. Fos upregulation by wound marginal cells might also play some role in the enhanced cell proliferation at embryonic wound sites discussed above. What is it about wounding that might lead to induction of immediate-early transcription factors? We have some preliminary evidence that upregulation of c-fos by wound-edge cells might be due to mechanical damage inflicted at the time of wounding and the resultant influx of Ca2+ ions (R. Errington and P. Martin, unpublished data). Our next steps will be to determine what other transcription factors are upregulated at the embryonic wound site and which of these plays a functional role in the wound closure process.
X. Inflammation Does Not Occur Following Wounding in the Embryo Inflammation is an early response of adult tissues to wounding but many studies show that different components of the inflammatory response are reduced or absent at embryonic and fetal wounds. The various inflammatory cell lineages make their first appearance outside the hemopoietic organs only at late embryonic/early fetal stages. Not until halfway through gestation (E10) in mouse embryos are macrophages first consistently seen leaving the yolk sac and entering mesenchymal tissues (Morris et al., 1991), and other studies show that differentiated neutrophils first appear even later in development (Rugh, 1990). Our studies in wounded mouse embryos and fetuses, using the macrophagespecific antibody F4/80 (Austyn and Gordon, 1981), show that even at stages when macrophages have begun patrolling embryonic tissues they are not necessarily recruited to wounds. In limb bud stage (E11.5) embryos we found that
Flg. 6 Transcription factor expression at the embryonic wound site. (A) Scanning electron micrograph of an E11.5 mouse embryo just after amputation of the hindlimb bud and just prior to culture in a roller bottle. (B) Low-power view of whole-mount Fos immunocytochemistry 1 h after wounding showing Fos protein (dark staining) in the epidermal cells at the wound margin (arrows) and to a lesser extent in the superficial mesenchymal cells. (C) High-power detail of (B) showing Fos protein in the nuclei of cells extending as far as three or four rows back from the wound edge. Scale bars: A, 1 mm; B, 200 pm; C, 50 pm.
6 . Wound Healing in Embryo and Fetus
193
macrophages were not recruited to the wound site during the wound closure period and this state of affairs persisted in wounds made at E12.5 and E13.5, but at E14.5, when the limbs are nearly fully patterned, amputation of a single toe resulted in substantive recruitment of macrophages to the wound surface within 12 h of wounding (Hopkinson-Woolley et al., 1994). We tested whether the lack of a macrophage response to wounding younger embryos was due to the naive status of embryonic macrophages or due to the absence of sufficient chemotactic cue emitted by the excisional wound, by inflicting small burn lesions causing significant localized cell death to the E l 1.5 limb bud. Embryos wounded in such a fashion showed significant macrophage recruitment to the site of necrosis (Hopkinson-Woolley et ai., 1994), prompting us to guess that a key reason for the lack of macrophage recruitment to simple excisional wounds in the young embryo might be the sparsity of cell death at the wound site (McCluskey et al., 1993). Using upregulation of immunoglobulins as another marker of inflammation, Whitby and Ferguson (1991a) found abundant endogenous IgG and IgM immunoglobulins at both neonatal and adult mouse wounds but none were detectable at the site of healing incisional lesions made to the fetal mouse (E16) lip. Studies of late-stage rabbit fetuses (E23 of the 3 1-day gestation) using enzyme histochemistry show that although, by these stages, macrophages are beginning to be recruited into subcutaneously implanted wound chambers, there is still a conspicuous absence of neutrophils at the wound site (Adzick et al., 1985). An earlier study by Dixon (1960) confirms the very late involvement of neutrophils in fetal healing, reporting them absent from fetal wounds until E19, just prior to birth in the rat, where term is 22 days.
XI. Fetal Wound Healing Environment-Extracellular and Growth Factors
Matrix
A. Collagen
To date there have been no reported studies of the involvement of extracellular matrix in wound healing in early embryos, but many studies have compared various matrix components of the later-stage fetal versus adult wound healing environments in the search for clues about why fetal wounds heal so much better than their adult equivalents. Since excess and poorly oriented collagen matrices are deemed to be the twin keys to (end products of) scar tissue formation, a number of early investigations focused on differences in collagen synthesis at the wound site. Adzick et al. (1985) placed standard Gore-Tex cylinders into small subcutaneous wounds in late-stage fetal, newborn, and adult rabbits, allowing a direct comparison of cellular infiltration and hydroxyproline accumulation (collagen synthesis) at various time points after wounding tissues at these three
194
Paul Martin
developmental stages. Their results clearly showed that both tissue ingrowth and collagen synthesis were more rapid in the fetus than in the newborn, and more rapid in the newborn than in the adult. More recently, immunohistochemical studies using antibodies raised against collagen types I , 111, IV, and VI have shown that all these isoforms are more rapidly laid down in murine fetal lip wounds than in equivalent neonate or adult wounds (Whitby and Ferguson, 1991a). Possibly more important for wound healing is the profile of these isoforms rather than total collagen at the wound site. Both fetal and adult wound connective tissues have elevated type 111 collagen content (Merkel et al., 1988), but normal fetal tissue contains a much higher proportion of type 111 collagen than does normal adult tissue (Smith et af., 1986). Clearly quantity and isoform type cannot fully explain why collagen matrix is apparently regenerated at the fetal wound site in a fashion identical to that of its neighboring unwounded tissue but very poorly replicated at an adult wound site. One possible difference that might affect orientation of collagen fibrillogenesis is tension within the wound connective tissue, and although this has not been investigated in the fetus, it is clear that scarring of an adult wound is significantly influenced by stretching (Burgess et al., 1990). Another possible influencing factor is the capacity of fibroblasts to migrate within the wound matrix. In this respect it is important to note that the high-molecular-weight glycosaminoglycan hyaluronate may provide a permissive environment for various developmental cell migrations, such as neural crest emigration from the neural tube during embryonic morphogenesis, owing to its high volume of hydration (Toole, 1981). Hyaluronate is present at relatively high amounts and persists far longer in fetal than in adult wounds (Longaker et al., 1991b).
B. Fibronectin and Tenascin
It is clear that matrix glycoproteins may play important roles in wound healing either as a scaffold for fibroblast migration within the wound connective-tissue or as a substratum over which reepithelialization can occur. It is particularly noteworthy that at an adult wound, not only is significant fibronectin released from the plasma at damaged blood vessels, but also the wound fibroblasts upregulate expression of fibronectin mRNAs spliced in a way normally only seen in the developing embryo (ffrench-Constant et al., 1989). Fibronectin is generally more abundant in fetal than neonatal or adult skin, and an early study in the rabbit suggested it might be deposited earlier at a fetal wound site than at adult wounds (Longaker et al., 1989). However, a more recent study using a fetal lip wound model showed that fibronectin was deposited at fetal and adult wound sites equally rapidly, within about an hour of wounding (Whitby and Ferguson, 1991a). To date no studies have yet addressed which of the spliced variants of
6. Wound Healing in Embryo and Fetus
195
fibronectin are expressed at the fetal wound. Tenascin, another glycoprotein implicated in the migratory aspects of tissue repair (Mackie et al., 1988), does appear to show a temporal difference in deposition when fetal and adult wounds are compared. Immunohistochemistry shows that tenascin is present at fetal wounds within the first few hours of wounding but not until 24 h after wounding in adult tissues (Whitby and Ferguson, 1991a; Whitby e t a l . , 1991). The authors of these studies speculate that rapid tenascin deposition may aid rapid reepithelialization in the fetal wound situation.
C. Growth Factors
In a study of embryonic wound healing in limb-bud stage mouse embryos we have shown rapid upregulation of TGFPl at the mRNA level, by cells at the epidermal wound margin and by superficial cells of the exposed wound mesenchyme (Martin et al., 1993). Interestingly, these are precisely the cells that previously upregulated the transcription factor, c-fos (see above), and it is well established that c-foslc-jun binding to upstream AP1 binding sites in TGFP 1’s two promoter regions can directly regulate TGFP 1 transcription (Kim et al., 1990; Roberts et al., 1991). In situ hybridization studies show TGFPl mRNA expressed at the wound site within an hour of wounding, and immunohistochemical studies show that soon afterward significant levels of TGFP 1 protein have been secreted into the wound mesenchyme. Because it is well established that exogenous TGFP 1 will stimulate tissue-culture fibroblasts to contract vigorously a collagen gel (Montesano and Orci, 1988), we have speculated that TGFPl , synthesized by the epidermal wound marginal cells and secreted into the wound mesenchyme, might be acting in a paracrine fashion to stimulate contraction of the embryonic wound connective tissue. Curiously, the high levels of TGFP 1 apparent initially at the wound site appeared to be rapidly cleared so that by 18 h after wounding, just a few hours before the wound was closed, TGFPl levels were back to normal background levels. This subtle upregulation and then clearance of a growth factor from the wound site is unique to the embryo and may partly explain why embryonic skin repair is perfect and adult wounds leave a scar. Indeed, although Whitby and Ferguson (199 1b) found PDGF protein, they could detect neither TGFPl nor basic fibroblast growth factor (bFGF) at fetal lip wound sites and have gone on to argue that reducing levels of growth factors naturally present at adult wound sites might be of therapeutic benefit in blocking scarring. Their preliminary findings testing this hypothesis look very promising with incisional wounds on the backs of adult rats apparently healing with significantly reduced scar formation following injection of a neutralizing antibody to TGFPl and -2 into the wound bed at the time of wounding (Shah et al., 1992, 1994). The converse experiment where an excess
196
Paul Martin
of exogenous TGFP is introduced into a fetal wound site, has been shown to trigger a fibrotic response in the normally nonscarring fetal tissues (Krummel et al., 1988).
XII. Adult Skin in the Fetal Environment and Vice Versa In an attempt to determine whether the fetal wound environment as outlined above or the fetal tissues themselves are responsible for efficient scar-free healing, a number of groups have tried grafting fetal and adult tissues to developmentally foreign sites prior to wounding. Longaker et al. (1990a) grafted fullthickness maternal sheep skin onto the backs of 60-day fetal lambs (term is 145 days). The adult skin was not rejected but was perfused by fetal blood and bathed in amniotic fluid. Forty days after grafting (100 days gestation), incisional wounds were made that cut through both host (fetal) and donor (adult) skin. By 14 days after wounding the fetal wounds had healed without scarring but the adult skin wounds had formed scar tissue, suggesting that even by supplying many components of the fetal environment it was not possible to encourage adult tissue to heal in a scar-free fetal fashion. The converse experiment was performed by grafting skin from aborted human fetuses to adult athymic nude mice (Lorenz et al., 1992). Human fetal skin from 15 to 22 weeks of gestation (term is 40 weeks) was grafted either cutaneously or subcutaneously to the backs of nude mice, and incisional wounds were made to the donor fetal skin 7 days after grafting. Curiously, the moist subcutaneous grafts healed in a scar-free fashion, but the cutaneous grafts exposed to air showed accelerated differentiation and healed, like adult skin, with a scar (Lorenz et al., 1992). These experiments show that neither the unique combination of factors present in amniotic fluid nor continual perfusion by fetal serum factors is essential for scar-free healing of fetal skin, but also show that fetal skin can scar if exposed to some of the adult wound site factors. However, precisely which of the adult wound site factors is responsible for triggering scarring is not yet clear. In many respects studies of wound healing in the pouch young of marsupials such as Monodelphis domestica or the South American opossum (Block, 1960; Ferguson and Howarth, 1991) have reinforced some of the conclusions drawn from the grafting experiments described above. These animals are born at relatively early stages of development compared to eutherian mammals (approximately equivalent to late-limb-bud rodent stages) and provide a model system unique in its combination of fetal tissues that are exposed to an environment outside of the uterus. The newborn opossum’s skin is keratinized, to protect from dehydration, but is very thin and highly vascularized much like the fetal mouse. The pouch young are very immature immunologically, gaining passive immunity only from antibodies delivered via their mother’s milk. Wounds made to these young opossums form a scab like an adult wound but heal without an inflamma-
6. Wound Healing in Embryo and Fetus
197
tory response and without subsequent scarring, which again demonstrates that being bathed in amniotic fluid is not the simple key to scar-free healing.
XIII. Is There a Critical Transition Phase in late Fetal Development When Healing Becomes Adult-like? It is clear that embryonic and early fetal healing differ significantly from that in the adult, and so an obvious question is when in development does the transition from fetal to adult-like healing occur? Ihara et al. (1990) using both an in vivo and a skin organ culture model suggest that a dramatic transition occurs between Gestational Days 16 and 18 in the rat (gestation is 22 days). At the earlier of these two stages both of their wound models close rapidly, not by obvious infilling of tissues, but by contraction of wound margin tissues, and differentiating hair bulbs reappear in the healed regions of the wounds. At El8 their in vitro wounds fail to contract and, instead, slowly fill in by fibroblastic invasion of the defect, whereas their in vivo wounds slowly cover over with new epidermis but no hair bulbs regenerate. Our own studies in the mouse suggest a fairly sharp cut-off point at E14.5 (gestation 21 days), beyond which one component of the typical adult inflammatory response, the macrophage, first appears to play a role in the wound healing process (see earlier). However, many other aspects of the wound healing process have not been carefully investigated. We do not know precisely when the other components of a mature inflammatory response first make their appearance at fetal wound sites and neither do we know at what stage in development the mode of reepithelialization changes from one driven by a contractile purse string to one relying on lamellipodial crawling. Careful studies at various stages of gestation in long-gestation animals, including sheep (Longaker et al., 1990b) and primates (Lorenz er al., 1993) and some clinical observation in the human fetus, confirm that the transition from scar-free healing to adult-like healing with scars occurs before birth, in the late second or early third trimester, dependent on the size of the wound. These studies, together with those of Whitby and Ferguson (1991a,b), suggest that, although individual components of an adult wound healing response may switch on at precise time points in development, the overall transition from fetal to adult-style healing is likely to be gradual, taking place over an extended gestational period.
XIV. Healing of Tissues Other Than the Skin (Not All Fetal Healing I s Perfect) Most studies of wound healing in the embryo and fetus focus largely on skin lesions but there have been some interesting studies examining repair of the skeleton and viscera before birth. If an incisional wound is made completely
198
Paul Martin
across the radius of the developing chick wing at E7, a few days before the onset of osteogenesis, then the wound heals without callus formation, although the eventually ossified long bone appears somewhat constricted at the wound site (McCullagh et al., 1990). A similar incisional wound made to the ulna after primary ossification in fetal lambs at 100 days of gestation resulted in some early osteoblast cell death at the wound site but led to perfectly healed bones 10 days later (Moellen et al., 1991). Even relatively major excisional lesions in the same fetal lamb ulna model involving removal of a length of bone three times its width are able to fill the gap with apparently normal cortical bone and marrow and after healing show no apparent histological sign of a fracture line (Moelleken el al., 1991). It is tempting to guess from the reports described above of skin and skeletal healing that all early fetal tissues are likely to repair almost perfectly without leaving a trace of earlier wound sites, but a paper reporting observations of healing after wounding the fetal diaphragm show this not to be the case. Longaker and colleagues (1991~)made bilateral incisional wounds to the 100-day fetal lamb diaphragm and left the wound open on one side to amniotic fluid while covering the contralateral wound with a skin flap. At this stage of lamb development incisional skin wounds heal in a scar-free fashion. However, both the diaphragm wound bathed in amniotic fluid and the one excluded from exposure to amniotic fluid healed with scar formation and without regeneration of muscle across the healed wound. This study suggests that the transition from fetal to adult-style healing occurs at different stages of development for different tissues and organs and may be more related to state of differentiation of the tissue than actual gestational age.
XV. Operating on the Human Fetus: Perfect Repair of Embryonic Defects-Realistic Dream or Fantasy? As eluded to in the opening paragraph of this review, it is hoped that studies of embryonic and fetal healing will not just provide clues about the mechanisms of morphogenetic movements in embryos and explanations of why wounded adult tissues scar, but will also provide support for the new clinical field of human fetal surgery. Pioneering work has been done in this area by Adzick and Harrison (199 I ) in their Fetal Treatment Programme at the University of California, San Francisco. This group successfully operates on human fetuses at midgestational ages (between 18 and 30 weeks) to correct a few carefully selected conditions that are surgically correctable and would otherwise be life-threatening at or before birth (for example, diaphragmatic hernias where the fetal intestines have escaped into the thoracic space and are blocking growth of the lungs). These operations have proven to be safe for the mother and not to jeopardize her ability to carry subsequent pregnancies. With continual advances in intrauterine diagnostic tech-
6. Wound Healing in Embryo and Fetus
199
nology and the possibility of scar-free surgery it seems likely that non-lifethreatening disorders, particularly of the limb and face will soon be considered candidates for in utero corrective surgery. In the near future it may be the case that the vast experience gained by embryologists surgically manipulating organisms, such as the chick, to establish the patterning mechanisms of organogenesis may be useful in guiding the human fetal surgeon in trying to correct faces and limbs in which this natural patterning machinery has gone awry.
Acknowledgments I would thank the ICRF, the MRC, the Wellcome Trust, and the Phoenix Appeal for funding work of mine and my colleagues described in this review. I also thank Kate Nobes and Janie McCluskey for critical reading of the manuscript and Julian Lewis for constant inspiration and guidance.
References Adelmann-Grill, B. C., Wach, F., Cully, Z., Hein, R., and Krieg, T. (1990). Chemotactic migration of normal dermal fibroblasts towards epidermal growth factor and its modulation by platelet-derived growth factor and transforming growth factor beta. Eur. J . Cell B i d . 51, 322326. Adzick, N. S . , and Harrison, M. R. (1991). The fetal surgery experience. I n “Fetal Wound Healing” (N. S. Adzick and M. T. Longaker, eds.), pp. 1-23. Elsevier, New York. Adzick, N. S . , Harrison, M. R., Glick, P. L., Beckstead, J. H., Villa, R. L., Scheuenstuhl, H., and Goodson, W. H., III (1985). Comparison of fetal, newborn, and adult wound healing by histologic, enzyme-histochemical, and hydroxyproline determinations. J . Pediatr. Surg. 20, 3 IS-3 19. Assoian, R. A., and Sporn, M. B. (1986). Type-beta transforming growth factor in human platelets: Release during platelet degranulation and action on vascular smooth muscle cells. J. Cell B i d . 102, 1712-1733. Assoian, R. K . , Fleurdelys, B. E., Stevenson, H. C., Miller, P. J., Madtes, D. K., Raines, E. W., Ross, R . , and Sporn, M. B. (1987). Expression and secretion of type p transforming growth factor by activated human macrophages. Proc. Natl. Acad. Sci. U.S.A. 84, 60206024. Austyn, J. M., and Gordon, S. (1981). F4/80,a monoclonal antibody directed specifically against the mouse macrophage. Eur. J. Immunof. 11, 805-815. Baker, P. C . , and Schroeder, T. E. (1967). Cytoplasmic filaments and morphogenetic movement in the amphibian neural tube. Dev. B i d . 15, 432-450. Bard, S . , and Sengel, P. (1984). Reconstitution of the epidermal basement membrane after enzymatic dermal-epidermal separation of embryonic mouse skin. Arch. Anat. Microsc. Morphol. Exp. 73, 239-257 Barrandon, Y., and Green, H. (1987). Cell migration is essential for sustained growth of keratinocyte colonies: The roles of transforming growth factor (Y and epidermal growth factor. Cell (Cambridge. Mass.) 50, 1131-1137. Bement, W. M., Forscher, P., and Mooseker, M. S . (1993). A novel cytoskeletal structure involved in purse-string wound closure and cell polarity maintenance. J. Cell B i d . 121, 565578.
200
Paul Martin
Block, M. (1960). Wound healing in the newbom opossum. Narure (London) 187, 340-341. Brock, J., Midwinter, K., Lewis, J., and Martin, P. Healing of incisional wounds in the embryonic chick wing bud: Characterisation of the actin purse-string. In preparation. Buck, R. C. (1979). Cell migration in repair of mouse corneal epithelium. Invest. Ophthalmol. Visual Sci. 18, 767-784. Burgess, L. P. A , , Morin, G . V., Rand, M., Vossoughi, J., and Hollinger, J. 0. (1990). Wound healing: Relationship of wound closing tension to scar width in rats. Arch. Otolaryngol., Head & Neck Surg. 116, 798-802. Bumside, B. (1971). Microtubules and microfilaments in amphibian neurulation. Am. Zool. 13, 989- 1006. Clarke, R. A,, Lanigan, J. M., DellaPelle, P., Manseau, E., Dvorak, H. F., and Colvin, R. B. (1982). Fibronectin and fibrin provide a provisional matrix for epidermal cell migration during wound re-epithelialisation. J . Invesr. Dermatol. 79, 264-269. Clore, J. N., Cohen, I . K., and Diegelmann, R. F. (1979). Quantitation of collagen types I and 111 during wound healing in rat skin. Proc. Soc. Exp. B i d . Med. 161, 337-340. Coffey, R. J., Bascom, C. C., Sipes, N. J., Graves-Deal, R., Weissman, B. E., and Moses, H. L. (1988). Selective growth inhibition of growth-related gene expression in murine keratinocytes by transforming growth factor p. Mol. Cell. B i d . 8, 3088-3093. Dale, L., and Bownes, M. (1980). Is regeneration in Drosophila the result of epimorphic regulation? Wilhelm Roux’s Arch. Dev. B i d . 189, 91-96. Darby, I., and Gabbiani, G.(1990). cu-smooth muscle actin is transiently expressed by myofibroblasts during experimental wound healing. Lab. Invest. 63, 21-29. Desmouliere, A , , Geinoz, A , , Gabbiani, F., and Gabbiani, G.(1993). Transforming growth factor-p I induces a-smooth muscle actin expression in granulation tissue myofibroblasts and in quiescent and growing cultured fibroblasts. J . Cell B i d . 122, 103- I I I . Dixon, J. B. (1960). Inflammation in the fetal and neonatal rat. The local reactions to skin bums. J . Parhol. Bacreriol. 80, 73-82. Edwards, D. R., Murphy, G., Reynolds, J. J., Whitman, S. E., Docherty, A. I. P., Angel, P., and Heath, J. K. (1987). Transforming growth factor beta modulates the expression of collagenase and metalloproteinase inhibitor. EMBO J . 6, 1899- 1904. England, M. A,, and Cowper, S . V. (1977). Wound healing in the early chick embryo studied by scanning electron microscopy. Anat. Ernbryol. 152, 1- 14. Estes, J. M., Vande Berg, J. S . , Adzick, N. S , MacGillivray, T. E., Desmouliere, A., and Gabbiani, G. (1994). Phenotypic and functional features of myofibroblasts in sheep fetal wounds. Differentiation (Berlin) 56, 173- 181. Ferguson, M. W. J., and Howarth, G.F. (1991). Marsupial models of scarless fetal wound healing. I n “Fetal Wound Healing” (N. s. Adzick and M. T. Longaker, eds.), pp. 95-124. Elsevier, New York. ffench-Constant, C., Van De Water, L., Dvorak, H. F., and Hynes, R. 0. (1989). Reappearance of an embryonic pattern of fibronectin splicing during wound healing in the adult rat. J. Ceil B i d . 109, 903-914. Fink, R. D., and Trinkaus, I. P. (1988). Fundulus deep cells: Directional migration in response to epithelial wounding. Dev. B i d . 129, 179- 190. Gipson, I. K . , Spurr-Michaud, S . J., and Tisdale, A. S. (1988). Hemidesmosomes and anchoring fibril collagen appear synchronously during development and wound healing. Dev. B i d . 126, 253-262. Grinnell, F. (1992). Wound repair, keratinocyte activation and integrin modulation. J . Cell Sci. 101, 1-5. Grinnell, F. (1994). Fibroblasts, myofibroblasts, and wound contraction. J . Cell B i d . 124, 401404. Hamburger, V., and Hamilton, H. L. (1951). A series of normal stages in the development of the chick embryo. J . Morphol. 88, 49-92.
6. Wound Healing in Embryo and Fetus
201
Hammacher, A., Hellman, U . , Johnson, A,, Ostman, A , , Gunnarsson, K., Westerman, B., Wasteson, A . , and Heldin, C. H. (1986). A major part of PDGF purified from human platelets is a heterodimer of one A chain and one B chain. J . B i d . Chem. 263, 16493-16498. Hopkinson-Woolley, J . , Hughes, D., Gordon, S., and Martin, P. (1994). Macrophage recruitment during limb development and wound healing in the embryonic and fetal mouse. J. Cell Sci. 107, 1159-1 167. Ignotz, R . , Endo, T., and Massague, J. (1987). Regulation of fibronectin and type I collagen mRNA levels by transforming growth factor+. J . Biol. Chem. 262, 6443-6446. Ihara, S . , and Motobayashi, Y. (1992). Wound closure in fetal rat skin. Development (Cumbridge, UKJ 114, 573-582. Ihara, S . , Motobayashi, Y., Nagao, E., and Kistler, A. (1990). Ontogenetic transition of wound healing pattern in rat skin occurring at the fetal stage. Development (Cambridge, U K ) 110, 67 1-680. Kim, S.-J., Angel, P., Lafyatis, R . , Hattori, K., Kim, K. Y., Sporn, M., Karin, M., and Roberts, A. B. (1990). Autoinduction of transforming growth factor D l is mediated by the AP-1 complex. Mol. Cell. Biol. 10, 1492-1497. Kolega, J. (1986). Effects of mechanical tension on protrusive activity and microfilament and intermediate filament organisation in an epidermal epithelium moving in culture. J . Cell Biol. 102, 1400-1411. Krummel, T. M., Michna, B. A., Thomas, B. L . , Spom, M. B., Nelson, J. M., Salzberg, A. M., Cohen, 1. K., and Diegelmann, R . F. (1988). Transforming growth factor beta (TGF-P) induces fibrosis in a fetal wound model. J . fediutr. Surg. 23, 647-652. Krummel, T. M., Erlich, H. P., Nelson, J. M., Michna, B . A., Thomas, B. L . , Haynes, J. M., Cohen, I . K., and Diegelmann, R . F. (1989). Fetal wounds do not contract in utero. Surg. Forum. 40, 613-614. Leibovich, S. J., and Ross, R . (1975). The role of the macrophage in wound repair. A study with hydrocortisone and antimacrophage serum. Am. J . fatho/. 78, 71-92. Longaker, M. T., Whitby, D. J . , Ferguson, M. W. J., Harrison, M. R . , Crombleholme, T. M., Langer, J. C . , Cochrum, J. C., Verrier, E. D., and Stem, R . (1989). Studies in fetal healing: 111. Early deposition of fibronectin distinguishes fetal from adult wound healing. J . fediutr. Surg. 24, 799-805. Longaker, M. T., Whitby, D. J., Jennings, R. W., Duncan, B. W., Cromblholme, T. M., Harrison, M. R . , Ferguson, M. W. J., and Adzick, N. S . (1990a). Adult skin in the fetal environment heals with scar formation. Surg. Forum. 41, 639-641. Longaker, M. T., Whitby, D. J., Adzick, N. S . , Crombleholme, T. M., Langer, J. C., Duncan, B. W., Bradley, S. M., Stern, R . , Ferguson, M. W. J., and Harrison, M. R . (1990b). Studies in fetal wound healing. VI. Second and early third trimester fetal wounds demonstrate rapid collagen deposition without scar formation. J . fediutr. Surg. 25, 13-69. Longaker, M. T., Burd, D. A. R . , Gown, A. M., Yen, T. S . B., Jennings, R . W., Duncan, B. W., Harrison, M. R . , and Adzick, N. S . (1991a). Midgestation fetal lamb wounds contract in utero. J. fediutr. Surg. 26, 942-948. Longaker, M. T., Chiu, E. S . , Adzick, N. S . , Stem, M., Harrison, M. R . , and Stern, R. (1991b). Studies in fetal healing: V. A prolonged presence of hyaluronic acid characterises fetal wound fluid. Ann. Surg. 213, 292-296. Longaker, M. T., Whitby, D. J., Jennings, R. W., Duncan, B. W., Ferguson, M. W. J., Harrison, M. R., and Adzick, N. S . (1991~).Fetal diaphragmatic wounds heal with scar formation. J. Surg. Res. 50, 375-385. Lorenz, H. P., Longaker, M. T., Perkocha, L. A , , Jennings, R . W., Harrison, M. R . , and Adzick, N. S. (1992). Scarless wound repair: A human fetal skin model. Development (Cumbridge, U K ) 114, 253-259. Lorenz, H. P., Whitby, D. J., Longaker, M. T., and Adzick, N. S. (1993). Fetal wound healing: The ontogeny of scar formation in the non-human primate. Ann. Surg. 217, 391-396.
202
Paul Martin
Mackie, E. J., Halfter, W.. and Liverani, D. (1988). Induction of tenascin in healing wounds. J . Cell Biol. 107, 275772767, Martin, P., and Lewis, J. (1992). Actin cables and epidermal movement in embryonic wound healing. Nature (London) 360, 179-183. Martin, P., and Nobes, C. D. (1992). An early molecular component of the wound healing response in rat embryos-induction of c-fos protein in cells at the epidermal wound margin. Mech. Dev. 38, 209-216. Martin, P., Hopkinson-Woolley, J., and McCluskey, J. (1992). Growth factors and cutaneous wound repair. Prog. Growth Factor Res. 4, 25-44. Martin, P., Dickson, M. C., Millan, F. A,, and Akhurst, R. J. (1993). Rapid induction and clearance of TGFPI is an early response to wounding in the mouse embryo. Do,. Genet. 14, 225238. McCluskey, J., and Martin, P. (1995). Analysis of the tissue movements of embryonic wound healing-DiI studies in the limb bud stage mouse embryo. Dev. Eiol. (in press). McCluskey, J., Hopkinson-Woolley, J., Luke, B., and Martin, P. (1993). A study of wound healing in the E11.5 mouse embryo by light and electron microscopy. Tissue Cell 25, 173-181. McCullagh, J. J., Gill, P., and Wilson, D. J. (1990). Repair of cartilaginous fractures during chick limb development. J . Orthop. Res. 8, 127- 13I . McPherson, J. M., and Piez, K. A. (1988). Collagen in dermal wound repair. In “The Molecular and Cellular Biology of Wound Repair” (R. A. F. Clark and P. M. Henson, eds.), pp. 471496. Plenum, New York. Merkel, J. R., DiPaulo, B. R., Hallock, G. G., and Rice, D. (1988). ’@px I and type I11 collagen content of healing wounds in fetal and adult rats. Proc. SOC.Exp. Biol. Med. 187, 493-497. Moelleken, B. R. W., Longaker, M. T., Cheng, J. C., and Simmons, D. J. (1991). Fetal fracture healing. In “Fetal Wound Healing” (N. S. Adzick and M. T. Longaker, eds.). Elsevier, New York . Montesano, R., and Orci, L. (1988). Transforming growth factor P stimulates collagen matrix contraction by fibroblasts: Implications for wound healing. Proc. Narl. Acad. Sci. U.S.A. 85, 4894-4897. Moms, L., Graham, C. F., and Gordon, S. (1991). Macrophages in haemopoetic and other tissues of the developing mouse detected by the monoclonal antibody F4/80. Development (Cambridge, U K ) 112, 517-526. Momss-Kay, G., and lickeft, F. (1985). The role of microfilaments in cranial neurulation in rat embryos: Effects of short term exposure to cyochalasin D. J. Embryol. Exp. Morphol. 88, 333-348. Newman, S. L., Henson, J. E., and Henson, P. M. (1982). Phagocytosis of senescent neutrophils by human monocyte-derived macrophages and rabbit inflammatory macrophages. J . Exp. Med. 156,430-442. Odland, G., and Ross, R. (1968). Human wound repair. I. Epidermal regeneration. J. Cell Biol. 39, 1135-1151. Oka, Y., and Orth, D. N. (1983). Human plasma epidermal growth factorlb-urogastrone is associated with blood platelets. J. Clin. Invest. 72, 249-259. Pierce, G . F., Mustoe, T. A,, Lingelbach, J., Masakowski, V. R., Griffen, G. L., Senior, R. M., and Deuel, T. F. (1989). Platelet derived growth factor and transforming growth factor-beta enhance tissue repair activities by unique mechanisms. J . Cell Eiol. 109, 429-440. Radice, G. P. (1980). The spreading of epithelial cells during wound closure in Xenopus larvae. Dev. Biol. 76, 26-46. Rappolee, D. A., Mark, D., Banda, M. J., and Werb, Z. (1988). Wound macrophages express TGFa and other growth factors in vivo: Analysis by mRNA phenotyping. Science 241, 708-712. Ridley, A. J., and Hall, A. (1992). The small GTP-binding protein rho regulates the assembly of focal adhesions and actin stress fibers in response to growth factors. Cell (Cambridge, Mass.) 70, 389-399.
6. Wound Healing in Embryo and Fetus
203
Roberts, A. B., Kim, S.-J., Kondaiah, P., Jakowlew, S. B., Denhez, F., Click, A. B., Geiser, A . G., Watanabe, S., Noma, T., Lechleider, R., and Sporn, M. B. (1991). Transcriptional control of expression of the TGFPs. Ann. N . Y. Acad. Sci. 593, 43-90. Rugh, R. (1990). “The Mouse: Its Reproduction and Development.” Oxford Univ. Press, Oxford. Seppa, H., Grotendorst, G. R., Seppa, S., Schiffman, E., and Martin, G.R. (1982). Plateletderived growth factor is chemotactic for fibroblasts. J . Cell Biol. 92, 584-588. Shah, M., Foreman, D. M . , and Ferguson, M. W. J. (1992). Control of scarring in adult wounds by neutralising antibody to transforming growth factor p. Lancet 339, 213-214. Shah, M., Foreman, D. M., and Ferguson, M. W. J. (1994). Neutralising antibody to TGFP1,2 reduces cutaneous scarring in adult rodents. J. Cell Sci. 107, 1137-1157. Sirnpson, D. M . , and Ross, R. (1972). The neutrophilic leukocyte in wound repair: A study with antineutrophil serum. J . Clin. Invest. 51, 2009-2023. Smedley, M. J., and Stanisstreet, M. (1984). Scanning electron microscopy of wound healing in rat embryos. J. Embryol. Exp. Morphol. 83, 109-117. Smith, L. T., Holbrook, K . A,, and Madri, J. A. (1986). Collagen types I , I11 and V in human embryonic and fetal skin. Am. J. Anar. 175, 507-521. Somasunderam, K., and Prathrap, K. (1970). Intra-uterine healing of skin wounds in rabbit foetuses. J . farho/. 100, 81-86. Somasunderam, K., and Prathrap, K. (1972). The effect of exclusion of amniotic fluid on intrauterine healing of skin in rabbit foetuses. J . farhol. 107, 127-130 Stanisstreet, M. (1982). Calcium and wound healing in Xenopus early embryos. J . Ernbryol. Exp. Morphol. 67, 195-205. Stanisstreet, M., and Panayi, M. (1980). Effects of colchicine, cytochalasin-B and papaverine on wound healing in Xenopus early embryos. Experientia 36, 1 1 10- I 1 I 1. Stanisstreet, M . , Wakely, J., and England, M. A. (1980). Scanning electron microscopy of wound healing in Xenopus and chicken embryos. J . Embryol. Exp. Morphol. 59, 341-353. Terkeltaub, R. A,, and Ginsberg, M. H. (1988). Platelets and response to injury. In “The Molecular and Cellular Biology of Wound Repair.” (R. A . F. Clark and P. M. Henson, eds.), pp. 35-48. Plenum, New York. Thevenet, A. (1981). Wound healing of the integument of the 5 day chick embryo. Arch. Anat. Microsc. Morphol. Exp. 70, 227-244. Thevenet, A . (1984). Prolifiration cellulaire au cours de la cicatrisation du tegument d’embryon de poulet de 5 jours. Arch. Anat. Microsc. Morphol. Exp. 73, 121-132. Took, B. P. (1981). Glycosaminoglycans in morphogenesis. In “Cell Biology of the Extracellular Matrix” (E. D. Hay, ed.). Plenum, New York. Torpey, N., Wylie, C. C.. and Heasman, J. (1992). Function of maternal cytokeratin in Xenopus development. Nature (London) 357, 4 13-4 15. Weiss, P., and Matoltsy, G. (1959). Wound healing in chick embryos in vivo and in vitro. Dev. Biol. 1, 302-326. Whitby, D. J., and Ferguson, M. W. J. (1991a) The extracellular matrix of lip wounds in fetal, neonatal and adult mice. Developmenr (Cambridge, UK) 112, 65 1-668. Whitby, D. J., and Ferguson, M. W. J. (1991b). Immunohistochemical localization of growth factors in fetal wound healing. Dev. Biol. 147, 207-215. Whitby, D. J., Longaker, M. T., Harrison, M. R., Adzick, N. S., and Ferguson, M. W. J. (1991). Rapid epithelialisation of fetal wounds is associated with the early deposition oftenascin. J . Cell Sci. 99, 583-586. Wild, A. E. (1965). Protein composition of the rabbit fetal fluids. Proc. R . SOC.London, Ser. B . 163, 90-1 15. Young, P. E., Richman, A. M., Ketchum A. S., and Kiehart, D. P. (1993). Morphogenesis in Drosophila requires nonmuscle myosin heavy chain function. Genes Develop. 7, 29-41,
This Page Intentionally Left Blank
7 Biphasic Intestinal Development in Amphibians: Embryogenesis and Remodeling during Metamorphosis Yun-Bo Shi Laboratory of Molecular Embryology National Institute of Child Health and Human Development National Institutes of Health Bethesda, Maryland 20892
Atsuko hhizuya-Oka Department of Anatomy Dokkyo University School of Medicine Mibu, Tochigi 321-02, Japan
I. Introduction 11. Embryogenesis of Amphibian Intestine A. Formation of the Digestive Tract B. Histogenesis and Differentiation of the Larval Small Intestine C. Gene Expression during Gut Embryogenesis 111. Intestinal Remodeling during Metamorphosis A. Morphological and Cellular Transformations B. Hormonal Control of Intestinal Remodeling C. Gene Regulation by Thyroid Hormone in the Intestine D. Cell-Cell and Cell-Extracellular Matrix Interactions in Intestinal Remodeling IV. Summary and Prospects References
1. Introduction The intestine serves as a unique m o d system for studying not only organogenesis during embryonic development but also cell proliferation and differentiation in the adult. During embryogenesis, the intestinal epithelium is gradually developed into a complex structure that provides an enormous luminal surface area for efficient food processing and absorption, the primary function of the adult organ (Glass, 1968; Segal and Petras, 1992). In higher vertebrates, this is achieved by first forming multiple circular epithelial folds (Fig. 1; Glass, 1968). Then numerous finger-like villi and crypts are formed along these folds. Finally, each villus/ crypt consists of a densely packed monolayer of columnar epithelial cells, which themselves have a large number of microvilli (the brush border) to amplify further Curnnr Topirs i,i Dewlopmenral Bio1og.y. Vol. 32 Copyright 0 1996 by Academic Press, Inc. All rights of reproduclion in any form reserved
205
Higher vertebrate differentiated adult intestine
6-
+
Monolayer Primordial Multi-layered columnar endoderm endodermal cells epithelium
:I
differentiated larval intestine
Typhlosole
fully differentiated
Metamorphosis: development of connective tissue and muscle; lawal epithelial cell-death; adult epithelial cell proliferation and differentiation
Amphibian
= Epithelial folds
Fig. 1 Comparison of intestinal development in amphibians and hlgher vertebrates. The primordial endodermal cells first form a multilayered cell mass. The endodermal cells are then converted into a monolayer of columnar epithelial cells (EP) tightly associated with the connective tissue (CT) though a basement membrane. Further development in amphibians dlverges from that in higher vertebrates. In the latter, the columnar cells develop into multiply folded epithelium surrounded by elaborate connective tissue (stlppled area) and muscles (hatched area). In amphibians, the epithelium remains as a simple tubular structure with only a single fold, the typhlosole. The differentiated epithelial cells in both cases have numerous microvilli in the brush border (bb) on the luminar surface for efficient nutrient absorption. Unlike higher vertebrates, the amphibian lntestine then undergoes a second phase of development that results in the replacement of larval epithelium with adult epithelium as well as extensive development of connective tissues and muscles.
7. Amphibian Intestinal Remodeling
207
the absorptive surface. In the adult intestine, the epithelium is constantly renewed through cell proliferation exclusively in the crypts. As cells migrate up along the crypt-villus axis, they gradually differentiate. Eventually at the tip of the villus, the old epithelial cells undergo cell death and are sloughed into the lumen. The intestine in adult amphibians resembles that in higher vertebrates (Fig. 1, Reeder, 1964; McAvoy and Dixon, 1977; Ishizuya-Oka and Shimozawa, 1987a). It has elaborate connective tissue and muscles. The epithelium also forms multiple circular folds; however, the villi and crypts are absent (although the amphibian epithelial folds have been referred as villi/crypts in some publications). Instead, the epithelial cells with numerous microvilli line the luminal surface of the folds with the proliferative cells confined toward the trough and differentiated cells toward the crest, thus generating a cell renewal system along the troughcrest axis similar to that in higher vertebrates. In contrast, the tadpole intestine has a much longer but simpler structure. It consists of a single layer of columnar epithelium surrounded by thin layers of muscles with little intervening connective tissue (Fig. 1, McAvoy and Dixon, 1977; Kordylewski, 1983; Ishizuya-Oka and Shimozawa, 1987a). There is only a single epithelial fold, the typhlosole, present in the anterior part of the intestine where larval connective tissue is abundant. These structural differences between larval and adult intestines presumably reflect changes in the physiological functions between herbivorous tadpoles and carnivorous frogs. Compared to the extensive studies on the development of avian and mammalian intestines, amphibian intestinal embryogenesis has received little attention. However, the remodeling of the intestine during metamorphosis has been investigated for decades. This is largely due to the fact that this intriguing phenomenon, which involves almost complete degeneration of the larval organ and development of an adult intestine, is controlled by a single compound, thyroid hormone (TH) (Dodd and Dodd, 1976; Yoshizato, 1989; Smith-Gill and Carver, 198 1). These investigations have accumulated an enormous amount of information on intestinal remodeling at both the cellular and the morphological levels. More recently, the molecular cloning and characterization of genes that are regulated during embryogenesis and metamorphosis of the intestine have revealed many interesting aspects of these processes. In this article, we will review these recent findings with particular emphasis on the analysis of molecular information in the context of morphological changes during intestinal development. Although this review will focus on Xenopus laevis intestine, on which most of the recent work was performed, the principles and conclusions drawn from these data should be applicable to other amphibians.
II. Embryogenesis of Amphibian Intestine Since the discovery of embryonic induction and the organizer by Spemann and Mangold (1924), special attention has been focused on the early development of
Fig. 2 Larval small intestine of Xenopus tadpoles before metamorphosis. (A) Cross section of the anterior intestine. The primary epithelium (PE) is simple columnar. The connective tissue (CT) is localized in the typhlosole (Ty). The muscle (M) is thin. L, lumen. (B) Higher magnification of the apical surface of the primary epithelium. The brush border consists of long microvilli (mv). (C) The epithelial-connective tissue interface. The connective tissue cell is poorly differentiated. The extracellular matrix (ECM) is sparse except for the region close to the thin basal lamina (bl). Scale bars, 100 p n (A) and 1 k m (B and C).
7 . Amphibian Intestinal Remodeling
209
amphibian embryos. To date, owing to newly developed molecular genetic technology, the mechanism of embryonic induction is gradually being clarified in Xenopus embryos. In contrast, the progress in the study of subsequent organogenesis, an important step in our understanding of animal development, still remains slow. Compared to early embryogenesis, organogenesis involves a larger number of cells, in addition to more complicated cell-cell interactions. In this article, we first summarize normal development of the amphibian small intestine, with special reference to cell types composing this organ. Then we will discuss some recent molecular studies on gene expression during intestinal embryogenesis.
A. Formation of the Digestive Tract Following gastrulation, the alimentary canal subdivides into a foregut, a long midgut, and a hindgut ending in the anal canal piece (Nieuwkoop and Faber, 1967; Hausen and Riebesell, 1991). Various digestive organs develop from the gut depending on the region along the anterior-posterior axis: the pharynx, esophagus, and stomach develop from the foregut; the small intestine, mainly from the midgut; and the large intestine, from the hindgut. Among the digestive organs, the small intestine is characterized by a long tube that occupies a major part of the digestive tract and continues to elongate until about the end of prometamorphosis (Marshall and Dixon, 1978a; Ishizuya-Oka and Shimozawa, 1987a). Within the narrow abdominal cavity, the long tube is double-coiled as inner and external spirals. Its wall is thin and flat except for a single fold called the “typhlosole” (Figs. 1 and 2A), which runs longitudinally in the anterior part of the small intestine during pre- and prometamorphosis in Xenopus tadpoles (Nieuwkoop and Faber, 1967; Marshall and Dixon, 1978a).
B. Histogenesis and Differentiation of the larval Small Intestine
Intestinal tissues are derived from the endoderm (the epithelium), the mesoderm (the connective tissue, muscles, and serosa), and the neural crest (the nerve). Epigenetic influences of the mesodermal derivatives on endoderm development during organogenesis have been proposed in the amphibian intestine (Holtfreter, 1939; Kemp, 1946; Okada, 1960) as well as in the avian and mammalian intestines (Le Douarin, 1964; Haffen et al., 1983; Yasugi, 1993; Louvard et al., 1992; Mamajiwalla et al., 1992; Simon-Assmann and Kedinger, 1993; Donjacour and Cunha, 1991). The epithelium initially appears as a solid cell mass containing large amounts of yolk granules, which rapidly decrease in amount after hatching (Nieuwkoop and Faber, 1967; Heusser et al., 1992). During development, the multilayered
21 0
Yun-Bo Shi and Atsuko Ishizwya-Oka
endoderm gradually differentiates into a simple columnar cell layer (the primary epithelium). This layer consists of absorptive cells, goblet cells, and endocrine cells, typical of absorptive epithelium. The absorptive cells possess the brush border consisting of long microvilli on their apical surface (Bonneville, 1963; Fox et al., 1972; Fig. 2B) and numerous enzymes for digestion (Kaltenbach et al., 1977; Hourdry et al., 1979). In fact, they function as principal cells of the absorptive epithelium during the larval period (Sugimoto eb al., 1981). The goblet cells, containing numerous mucous granules, are scattered among the absorptive cells. The endocrine cells are very few, but they have been shown in frogs to contain hormones such as serotonin and cholecystokinin (L'Hermite et al., 1988; Scalise and Vigna, 1988). In addition, lymphocytes are frequently dispersed among the epithelial cells, mainly in the basal region of the epithelial layer. Overall, the primary epithelium of the tadpole intestine is similar to the intestine of other vertebrates; however, there is one major difference. The differentiated larval epithelial cells are capable of dividing independently of their location (Marshall and Dixon, 1978b), in contrast to adult frogs or higher vertebrates where the proliferating cells are less or not at all differentiated and localized in the trough of an epithelial fold (frog) or crypt (higher vertebrates). In addition, there has been no evidence for the existence of undifferentiated stem cells in the primary tadpole epithelium similar to those in the mammalian small intestine (Bjerknes and Cheng, 1981). The cells of the remaining tissues of the intestine, the connective tissue, muscle, and serosa differentiate from the splanchnic mesodermal layer after gastrulation. Until the end of prometamorphosis, the connective tissue remains very thin in all regions except for the typhlosole (Figs. 1 and 2A). Even in the typhlosole, the number of connective tissue cells is very small, and almost all of the identifiable cells are undifferentiated slender cells. The extracellular matrix is also sparse (Ishizuya-Oka and Shimozawa, 1987b; Fig. 2C). Similarly, the muscle remains thin until the end of prometamorphosis (Fig. 3) when a slight increase in thickness is detected. During embryogenesis, undifferentiated myoblast-like cells first form a circular layer surrounding the epithelial cells and the connective tissue. Beyond this is a longitudinal layer of muscle cells. In the larval small intestine, the inner circular muscle layer is always thicker and more differentiated than the outer longitudinal layer (Kordylewski, 1983).
C. Gene Expression during Gut Embryogenesis
The mechanisms that determine the spatial temporal differentiation in the digestive tract are still unknown. Recently, however, it has been shown in Xenopus that homeobox genes specify the anterior-posterior axis during early develop-
21 1
7. Amphibian Intestinal Remodeling Intestinal metamorphosis begins:Length reduction; IFABP gene repression; connective tissue and muse 1e development
Plasma T3 concentration (ArblKrary Units)
~ a r v ~ l epithelial cell death; adult epithelial cell proliferation
I
IFABP and villin gene activation Tadpole hatching
//
Stage
1
Time
0
(days)
'
I
I
I
35/36
45
51
2.1
4.1
17
\
..55, .32
I 58
61
63
66
44
48
51
58
Premetamorphosis
Climax
Prometamorphosis
Wg. 3 Chronological events during Xenopus laevis intestinal development in relation to thyroid hormone levels during metamorphosis. The plasma concentrations for thyroid hormone, T,, are from Leloup and Buscaglia (1977) and the developmental stages (time) are based on Nieuwkoop and Faber (1967). See text for additional details.
ment (Cho er al., 1991; Gurdon, 1992; McGinnis and Krumlauf, 1992). In addition, in the gut endoderm, the homeobox gene XIHbox 8 has been found to be expressed in a region-specific fashion during organogenesis (Wright et al., 1988). Thus, it seems that some homeobox genes may be involved in axis determination during gut embryogenesis. The differentiation of primary epithelium leads to the activation of epitheliumspecific genes. Two enterocyte (absorptive cell)-specific genes have been found to be expressed in prehatching embryos. The first gene, villin, is one of the five major proteins in the intestinal microvillar microfilaments (Bretscher and Weber, 1979, 1980; Tihey and Mooseker, 1971; Mamajiwalla et al., 1992). Although the gene has not been cloned in Xenopus, immunohistochemical analysis has localized villin to the brush border of the developing and mature enterocytes (Heusser et al., 1992). More importantly, villin was found to be present in the apical domain of some endodermal cells bordering the primitive cavity as early as stage 30 when the epithelium still exists as multilayered endodermal
21 2
Yun-Bo Shi and Atsuko Ishizwya-Oka
cells (Fig. 3; Heusser et al., 1992). Thus the early expression of villin in amphibians seems to resemble that in higher vertebrates (Shibayama et al., 1987). Moreover, the spatio-specific expression of villin indicates that the cells facing the cavity begin to differentiate before the others in the multilayered endodermal cell mass and that cell-cell interactions are likely to be important during this process. The second gene, intestinal fatty acid binding protein (IFABP) gene, was recently cloned in X . laevis. The frog IFABP gene is expressed primarily, if not exclusively, in the enterocytes (Shi and Hayes, 1994), similar to the expression pattern in mammals (Alpers et al., 1984; Sweetser et a f . , 1987; 1988; Green et a l . , 1992). Proliferative, undifferentiated epithelial cells in the crypts in mammals or in the troughs of the epithelial folds in adult Xenopus (Ishizuya-Oka et al., 1994) have no or very low levels of IFABP expression. Like villin, the IFABP gene was found to be expressed as early as stage 33/34 (Shi and Hayes, 1994), before extensive morphological differentiation of the epithelium (Nieuwkoop and Faber, 1967; Heusser et al., 1992). It is unclear what transcription factors trigger the expression of these epithelial genes during epithelial differentiation. In this regard, it is worth noting that at least two transcription factor genes are expressed during this early period. A homeobox protein was found to be expressed in a narrow band of the endoderm in Xenopus (Wright et a f . , 1988). Its expression in the intestine is largely confined to epithelial cells of the duodenum where IFABP gene is also highly expressed (Ishizuya-Oka et al., 1994). As mentioned above, this gene and/or other homeobox genes could play a role in the determination of the anteriorposterior axis in the intestine, and thus establishing an gradient of IFABP gene. The second transcription factor implicated in the development of the epithelium is the Xenopus GATA-4 gene (Kelley et al., 1993). GATA-4 belongs to a family of GATA-binding transcription factors involved in the activation of genes during cell differentiation (Evans et al., 1990; Orkin, 1990). GATA-4 is expressed in the gut endodermal cell during embryogenesis (Kelley et al., 1993). In the adult, the expression level of GATA-4 is high in the small intestine but low in the large intestine, similar to the IFABP expression pattern. Thus it is possible that GATA-4 is involved in the activation of intestine epithelial cell specific genes such as IFABP.
111. Intestinal Remodeling during Metamorphosis A. Morphological and Cellular Transformation
Accompanying the switch from a herbivorous to carnivorous diet, the amphibian small intestine undergoes dramatic changes during a short period of metamorphosis (Fig. 3). At the end of metamorphosis, the amphibian small intestine is basically analogous to the mammalian small intestine (Fig. 1).
7. Amphibian Intestinal Remodeling
213
1. Gross Anatomy Around the onset of metamorphic climax the long larval small intestine suddenly begins to shorten and this process continues until the end of metamorphosis (Marshall and Dixon, 1978a; Ishizuya-Oka and Shimozawa, 1987a). Following the beginning of this shortening, the morphogenesis of intestinal folds occurs. These appear as several circular folds that run longitudinally and are straight along the gut axis (Figs. 1 and 4A), gradually increasing in number and height, and finally being modified into longitudinally zigzaged folds (Fig. 4B). This morphogenetic process also occurs in the avian small intestine (Grey, 1972); however, in this case the zigzag folds continue to develop into definitive villi. In amphibians, the zigzag folds remain throughout adulthood (Fig. 1; McAvoy and Dixon, 1978a).
2. Epithelial Transformation The epithelial transition from larval to adult form of the amphibian intestine can be divided into two processes, degeneration of the primary epithelium and development of the adult (secondary) epithelium. Degenerative cellular changes occur around the onset of metamorphic climax. For example, the microvilli composing the brush border decrease in number and height, whereas lysosomes increase in number and in hydrolytic activity (Bonneville, 1963; Bonneville and Weinstock, 1970; Hourdry and Dauca, 1977). Recently, it has been shown by electron microscopy that cell death of the primary epithelium in the amphibian intestine occurs by apoptosis (programmed cell death) (Wyllie et al., 1980). These apoptotic bodies are at least partially phagocytosed by macrophages localized in the degenerating primary epithelium (Ishizuya-Oka and Shimozawa, 1992b; Figs. 5A and 5B). The macrophages are eventually extruded into the lumen while still retaining the apoptotic bodies (Fig. 5C). In addition, a sharp transition of major histocompatability complexes during metamorphosis has been reported (Flajnik et al., 1987; Du Pasquier and Flajnik, 1990), suggesting that in addition to macrophages, other immune cells may also be involved in the primary epithelial degeneration process. Just after the beginning of the primary epithelial degeneration, primordia of the secondary epithelium are detected at the epithelial-connective tissue interface as small islets consisting of undifferentiated epithelial cells. It is still unclear whether these secondary epithelial cells are derived from a pool of undifferentiated cells in the primary epithelium (Bonneville, 1963) or transformed from differentiated primary cells (Marshall and Dixon, 1978b). In any case, the primordia rapidly grow into the connective tissue through active cell proliferation and differentiate to form the secondary epithelium, replacing the degenerating primary epithelium (McAvoy and Dixon, 1977; Ishizuya-Oka and Shimozawa, 1987a; Figs. 6A and 6B). With the progression of fold formation, proliferative
214
Yun-Bo Shi and Atsuko Ishizwya-Oka
Fig. 4 Scanning electron micrographs of the luminal surface of the Xenopus small intestine. (A) Straight folds (Fo) at the early stage of metamorphic climax. (B) Zigzag folds at the end of metamorphosis. Scale bars, 10 pm.
cells of the secondary epithelium become localized in the trough of the folds like those in the mammalian crypts (Cheng and Leblond, 1974), in contrast to the primary epithelium (see above). There are also some quantitative differences
fig. 5 Degeneration of the primary epithelium (PE) of the Xenopus small intestine during metamorphosis. (A) An intraepithelial macrophage (arrowheads) including an apoptotic body (Ab) characteristic of condensed chromatin close to its nuclear membrane. L, lumen. (B) Many macrophages (arrows) are localized in the primary epithelium, but not in the islet (Is) of undifferentiated epithelial cells. CT, connective tissue. (C) An intraluminal macrophage including apoptotic bodies. Scale bars. 2 wm.
Fig. 6 Development of the secondary epithelium (SE) of the Xenopus small intestine during metamorphosis. (A) Islets (Is) growing into the connective tissue (CT) at the onset of metamorphic climax. Mitotic cells (arrows) are numerous in the islets. The connective tissue forms a thick layer. PE, primary epithelium. (B) Cross section of the small intestine at the late stage of metamorphic climax. The secondary epithelium is differentiated and covers well-developed intestinal folds (Fo). L, lumen. M, muscle. (C) Higher magnification of the apical surface of the secondary epithelium at the end of metamorphosis. The brush border consists of microvilli (mv) shorter than those of the primary epithelium in Fig. 1B. Scale bars, 50 pm (A and B) and 1 pm (C).
7. Amphibian Intestinal Remodeling
21 7
between the primary and the secondary epithelia, for example, shortening of the brush border (Bonneville, 1963; Bonneville and Weinstock, 1970; Fig. 6C), and increases in enzyme activities such as alkaline phosphatase, y-glutamyltranspeptidase, and catalase (Kaltenbach et al., 1977; Hourdry et al., 1979; Dauca et al., 198 1). In addition, it is known that after metamorphosis secondary epithelial cells migrate from the trough to the crest of intestinal folds and gradually differentiate during migration, similar to processes in the mammalian small intestine (McAvoy and Dixon, 1977, 1978b). Thus, during metamorphosis, amphibian intestinal folds seem to acquire a basic structure analogous to the cell renewal system in the mammalian small intestine (Bjerknes and Cheng, 1981).
3. Development of the Connective Tissue Growing evidence suggests a close relationship between the epithelium and the connective tissue during intestinal remodeling. When the epithelial transition from the larval to adult form begins, the connective tissue suddenly increases in mitotic activity, cell number, and thickness. The connective tissue at this time consists of various types of cells such as immature mesenchymal cells, fibroblasts, macrophages, and mast cells (Ishizuya-Oka and Shimozawa, 1987b, 1992b; Yoshizato, 1989). Furthermore, remarkable changes in the connective tissue occur close to the epithelium. When the primary epithelium begins to degenerate, the basal lamina, which is thin throughout the larval period (Fig. 2C), becomes thick in the entire region beneath the epithelium and remains thick until the primary epithelium disappears. In addition, throughout the thick basal lamina, fibroblasts that possess well-developed rough endoplamic reticulum often contact the epithelial cells (Fig. 7A). These cell contacts are most frequently observed around the primordia of the secondary epithelium when the epithelial cells most actively proliferate. These observations suggest that the thickening of the basal lamina and the cell contacts are related to the primary epithelial cell death and the secondary epithelial cell proliferation, respectively. A histochemical study indicated that only the thick basal lamina and fibroblasts close to the epithelium stain positive for lectins such as concanavalin A and wheat germ agglutinin when the epithelial transition begins, suggesting that connective tissue changes are also occurring at the molecular level (Ishizuya-Oka and Shimozawa, 1990; Fig. 7B). At later stages, the basal lamina becomes thin beneath the secondary epithelium. In addition the cell contacts and all cell types of the connective tissue, except fibroblasts, decrease in number. By the end of metamorphosis, almost all of the connective tissue cells are ordinary fibroblasts. In the trough of the epithelial folds, these fibroblasts are close to the epithelium and aligned parallel to the curvature of the epithelial basal surface (Fig. 7C). This structure is similar to the subepithelial fibroblastic sheath reported to be present in the crypt of the mammalian small intestine. This sheath has been thought to play important roles in epithelial cell proliferation and/or differentiation (Marsh and Trier, 1974).
Fig. 7 The epithelial-connective tissue interface in Xenopus small intestine at the onset of the metamorphic climax (A and B) and at the end of metamorphosis (C). (A) A cell contact (arrowheads) between a fibroblast (F)possessing well-developed rough endoplasmic reticulum (rer) and an epithelial cell of an islet (Is). The basal lamina (bl) is thick. Collagen fibrils (cf) are developed. (B) Histochemical section stained for concanavalin A. Connective tissue cells (CT, arrows) close to the epithelium and the basement membrane (bm) stain positive strongly. (C) Trough region of a zigzag fold. Fibroblasts close to the secondary epithelium (SE) are aligned parallel to the curvature of the basal epithelial surface. Scale bars, 1 pm (A and C) and 20 pm (B).
7 . Amphibian Intestinal Remodeling
21 9
4. Other Tissues
Although much work has been done on the epithelial and connective tissue, very little information exists on other intestinal tissues. One such tissue, muscle, becomes considerably thicker during metamorphosis primarily due to the increase of the inner circular muscle layer. In contrast, after metamorphosis, the thickening is mainly due to that of the outer longitudinal muscle layer. Consequently, in adult Xenopus, the thickness of each layer is almost the same (Kordylewski, 1983). Replacement of the neurons from larval to adult types in the myenteric plexus of the bullfrog intestine has been also observed during metamorphosis (Torihashi, 1990).
B. Hormonal Control of Intestinal Remodeling As early as 1912, it was found that a substance in the thyroid gland could precociously induce changes that normally occur during metamorphosis (Gudernatsch, 1912). The substance was later identified as TH (for a review, see Dodd and Dodd, 1976). During normal development, the synthesis of endogenous TH triggers the metamorphic process (Fig. 3). Blocking this synthesis inhibits metamorphosis. In contrast, simple addition of TH to tadpole rearing water can induce the process precociously, including intestinal remodeling (Ishizuya-Oka and Shimozawa, 1991; Dodd and Dodd, 1976; Dauca and Hourdry, 1985; Shi and Hayes, 1994). More importantly, the response to TH is organ-autonomous, as individual organs can be induced to metamorphose even when cultured individually. Thus, organ cultures of the Xenopus larval intestine undergo epithelial cell death as seen during normal development in the presence of thyroid hormone T,, although insulin and cortisol are required in addition to T, for secondary epithelial cell proliferation and differentiation (Ishizuya-Oka and Shimozawa, 1991; Fig. 8). These results indicate that although TH is the controlling agent of intestine remodeling, other hormones are also important to ensure complete organ transformation (White and Nicoll, 1981; Kikuyama et al., 1993). How does TH influence metamorphosis? Early investigations revealed the existence of amphibian cellular and plasma TH-binding proteins (Galton, 1983). To date, however, the functions of these proteins remains unknown. Only after the identification of high affinity nuclear TH receptors (TRs) was it suggested that TH controls metamorphosis by regulating gene expression (Galton, 1983). This idea gained further support when avian and mammalian TRs were cloned and shown to be transcription factors (Sap et al., 1986; Weinberger et al., 1986; Evans, 1988; Green and Chambon, 1988). TRs belong to the superfamily of steroid hormone receptors. These receptors share similar structural organizations. Each receptor has a unique binding domain for its cognate hormone in the carboxyl terminus and a more conserved DNA binding domain in the amino
fig. 8 Organ culture of the larval small intestine of Xrnopus tadpoles. (A) Control explant of the anterior intestine cultured in the medium deprived of hormones. The epithelium (E) is maintained in good conditions, and metamorphic changes do not occur. Both the connective tissue (CT) and the muscle (M) remain undeveloped. Go, goblet cell; bb, brush border. (B) Explant of the anterior intestine cultured with T,, insulin, and cortisol. A typical islet (Is) grows by rapid cell proliferation (arrows). The connective tissue develops. (C) Explant of the posterior intestine. No islets are formed. The number of epithelial cells is small. Connective tissue cells are few. Scale bars, 10 pm.
7. Amphibian Intestinal Remodeling
221
terminus of the protein. The DNA binding domain mediates specific recognition of the hormone response elements present in the transcription-regulatory regions of the target genes. These receptors function as dimers. In the case of TR, it can bind thyroid hormone response elements as either a homodimer or a heterodimer (Forman and Samuels, 1990). The best dimerization partners for TR are retinoid X receptors (RXR), which bind 9-cis retinoic acid (Yu et al., 1991; Heyman et a l . , 1992). In fact, in the presence of T,, TR-RXR heterodimers have been shown to regulate specifically TH response genes, suggesting that they are the active receptor complexes in vivo (Perlmann et al., 1993; Kurokawa et a l . , 1993; Zhang and Pfahl, 1993). There are two types of TRs in vertebrates, TRa and TRP. Both TRa and TRP genes have been cloned in amphibians (Brooks et al., 1989; Yaoita et al., 1990; Schneider and Galton, 1991). In X . laevis, the two TRa and two TRP genes are all expressed during metamorphosis (Yaoita and Brown, 1990; Kawahara et al., 1991). At the mRNA level, the TRa genes are highly expressed throughout tadpole development and metamorphosis. In contrast, the TRP genes are expressed at very low levels until the beginning of prometamorphosis when they are up-regulated by the increasing concentration of endogenous TH (Fig. 3). These results, coupled with the fact that no other TRs are known to exist, strongly suggest that these receptors mediate the effect of TH during metamorphosis. Furthermore, TRa and TRP may play different roles in this process.
C. Gene Regulation by Thyroid Hormone in the Intestine
To understand how TH regulates amphibian metamorphosis and more specifically intestinal remodeling, it is important to identify genes that are involved in this process. In the absence of a genetic method in amphibians, an alternative approach has been to first isolate genes expressed during the process and then study their functions. The control of metamorphosis by TH makes this approach feasible. TH presumably binds TRs, producing complexes that directly activate or repress a set of genes at the transcriptional level without a requirement for new protein synthesis. These genes are the so-called early or direct response genes. The products of these genes can then modulate either directly or indirectly the expression of downstream TH response genes. Eventually, this cascade of gene regulation leads to metamorphosis-specific changes. Tissue-specific transformations during metamorphosis are thus determined by the activation and/or repression of tissue-specific genes at any stage(s) of the regulation cascade. Clearly, the identification and characterization of the early response genes are of primary importance in the understanding of the gene regulation cascade necessary for tissue remodeling. The development of polymerase chain reaction (PCR)-based methods for subtractive differential screening has allowed genes involved in this cascade to be isolated. Many genes whose expression is altered by TH within the
222
Yun-Bo Shi and Atsuko Ishizwya-Oka
first 24 h of treatment have been isolated from different Xenopus tissues. Initial characterization of these genes suggest that they are likely involved in the metamorphic transition of these tissues (Buckbinder and Brown, 1992; Wang and Brown, 1993; Shi and Brown, 1993; Shi, 1994). In the intestine, over 20 TH response genes have been isolated (Shi and Brown, 1993). In addition, several other genes isolated from tail (Wang and Brown, 1993) and limb (Buckbinder and Brown, 1992) were also found to be regulated by TH in the intestine. These genes are not we11 characterized at this time. Therefore, this discussion will focus on the genes isolated from intestine. Not surprisingly, these genes fall into several different categories with regard to their regulation by TH, their developmental profiles of expression, and their possible functions during metamorphosis. Among them, only a single gene was found to be down-regulated by TH. The identity and function of this gene remain unknown. The vast majority of up-regulated genes are activated in the intestine of premetamorphic tadpoles within several hours after treatment with physiological concentrations of T,. This up-regulation occurs even in the presence of protein synthesis inhibitors, indicating that no new protein synthesis is required for their activation. Thus, these genes are considered to be early or direct TH response genes. In general, the up-regulated genes can be grouped into three classes according to their developmental expression profiles. Genes in the first class are expressed almost exclusively during metamorphosis. Genes in the second class are activated during metamorphosis and their expression remains high even in adult intestine. Genes in the third class are expressed at high levels immediately before and after the climax of metamorphosis but minimally at the actual climax. These results suggest that genes in different classes may exert different biological functions during intestinal remodeling. By direct sequence comparison, a few genes have been identified as homologs of previously characterized genes in other animal species (Table I). Among them are three types of transcription factors: NF- 1, TRP, and the basic-leucine-zipper motif containing transcription factors. NF- 1 belongs to the second developmental class, whereas the other transcription factors are in the first class. The activation of these genes provides strong support for the gene regulation cascade model described above, as the transcriptional factors will presumably activate and/or repress the transcription of downstream genes involved in intestinal remodeling. Another gene in the first developmental class has been identified to encode the extracellular matrix metalloproteinase stromelysin-3. This enzyme could potentially regulate cell-cell and cell-matrix interactions. This function will be described in more detail below. Most of the TH response genes are regulated by TH in many tissues. There are several genes that are expressed and/or regulated by TH only in the intestine, but no homologs can be found in the data bank (Shi and Brown, 1993). This is not too surprising because very few intestinal genes have been cloned in any species. The activation of a large number of ubiquitous genes does, however, suggest that
7. Amphibian Intestinal Remodeling
223
Table I Thyroid Hormone Response Genes in Amphibian Intestine ~~
Gene
Response to TH
Zinc finger TFa
Direct (early)
Possible function Transcriptional regulation of intermediate and/or late TH response genes
~
Reference Wang and Brown (1993)
Wang and Brown (1993); Shi and Brown (1993)
I
bZIP TF" NF- 1 '
M . Puzianowska-Kuznicki and Y.-B. Shi, unpublished observations
Stromelysin-3
Gelatinase
Extracellular matrix remodeling and degradation Indirect (late or intermediate)
Alkaline phosphate
I Brush border enzymes
1
Glutamyl transpeptidase Villin
IFABPd
5
Wang and Brown (1993); Shi and Brown (1993)
Patterton et al. (1995)
Dauca er al. (1980)
1
Brush border structural protein
Figiel et al. (1989)
Fatty acid metabolism
Shi and Hayes (1994)
"Zinc finger containing transcription factor. bBasic leucine-zipper motif containing transcription factor. 'Nuclear factor- 1. %testinal fatty acid binding protein.
the early gene regulation program induced by TH may be similar in different tissues. Furthermore, the tissue-specific transformations appear to be determined through the cooperation of the common TH response genes and those tissuespecific factors not regulated by TH. The most dramatic change during intestinal remodeling is the replacement of
224
Yun-Bo Shi and Atsuko Ishizwya-Oka
the larval epithelium by the adult form. Undoubtedly, many genes expressed in the larval epithelium will be repressed, whereas genes expressed in the adult epithelium will be activated during this process. These activated and repressed genes are expected to be further downstream in the gene regulation cascade than the early response genes described above. Indeed, a few of these genes have been characterized and found to respond slowly to TH treatment (Table I). Among the first studied are two intestinal brush border enzymes, alkaline phosphatase, and glutamyl transpeptidase (Dauca et al., 1980). The activities of these enzymes in the frog Rana catesbeiana increase after prolonged treatment (3- 10 days) with TH. Similarly, the epithelial protein villin has been shown to be regulated by TH in Alytes obstetricians (Figiel et al., 1989). The levels of villin were initially down-regulated after about 1 week of TH treatment but were subsequently reelevated upon further treatment. This agrees well with that observed during normal development (Figiel et al., 1987). As a structural protein of the brush border, villin is present in larval intestinal epithelium. Upon degeneration of the epithelium at the metamorphic climax, villin may be expected to be absent. It is expressed again when adult epithelium is formed. However, neither villin nor brush border enzyme genes have been cloned in any amphibians. It is thus unclear how the changes in the levels of these proteins are regulated by TH. Unlike villin and brush border enzymes, the IFABP gene has been shown to be regulated by TH at the mRNA level inX. laevis (Shi and Hayes, 1994). Similarly this regulation appears to be a slow process. Like villin, the IFABP gene is expressed in the epithelium of both tadpoles and frogs but repressed in tadpoles at the climax of metamorphosis. TH treatment of premetamorphic tadpoles can reproduce this developmental expression pattern. More importantly, the downregulation of IFABP gene during normal or TH-induced intestinal remodeling occurs before any visible morphological changes in the larval epithelial cells (Shi and Hayes, 1994; Ishizuya-Oka et al., 1994). This suggests that its regulation by TH, although slow and very likely indirect, precedes epithelial cell degeneration, consistent with the fact that larval epithelial degeneration occurs by a programmed cell death pathway. In addition, the activation of IFABP gene in adult epithelial cells occurs early during epithelial cell differentiation, before the development of the brush border. IFABP gene can be induced by prolonged treatment of premetamorphic tadpoles with TH. This activation, as well as the repression in larval epithelium, is most likely at the transcriptional level. The slow response to TH by IFABP, villin, and the brush border enzyme genes suggests that these genes are indeed downstream of the early response genes described above.
D. Cell-Cell and Cell-Extracellular Matrix Interactions in Intestinal Remodeling
Although the epithelium is responsible for the major biological function of the intestine, its development requires the participation of the connective tissue and
7. Amphibian Intestinal Remodeling
225
the extracellular matrix (ECM). This has been best demonstrated by experimentation in birds and mammals (for reviews, see Simon-Assmann and Kedinger, 1993; Louvard et al., 1992; Dauca et al., 1990). Many studies have shown that mesenchymal-epithelial interactions are crucial for the differentiation of the intestinal epithelium. The best evidence was obtained from experiments using recombinant organ cultures along with morphological and biochemical markers (Ishizuya-Oka and Mizuno, 1984, 1985; Haffen et al., 1982). These experiments directly demonstrated that intestinal mesenchyme is capable of inducing intestine-specific morphogenesis as well as cytodifferentiation of heterologous embryonic endoderm. The importance of mesenchyme in epithelial development is also supported by in vitro experiments demonstrating that epithelial cells can proliferate and differentiate when cocultured with fibroblasts but not when cultured alone (Kedinger et a l . , 1987; Stallmach et al., 1989). In addition, mesenchymal cells participate in the formation of extracellular matrix, thus indirectly affecting epithelial development (Weiser et al., 1990; Louvard et al., 1992; Simon-Assmann and Kedinger, 1993). The intestinal epithelium is separated from the mesenchyme by a special extracellular matrix, the basal lamina, whose major components include laminin, entactin, type IV collagen, and proteoglycan. Concurrent with epithelial morphogenesis, i.e., the formation of epithelial folds and villi/crypts, the interstitial connective tissue develops extensively. The first indication that ECM may play a role in intestinal cell interactions comes from the fact that the composition and spatial distribution of ECM molecules changes during intestinal development (Louvard et a l ., 1992; Simon-Assmann and Kedinger, 1993). Thus, it has been reported that laminin expression is activated around the stage of villus formation and that the two subunits of laminin are differentially regulated (Simo et al., 1991). Similarly, changes in the composition of the glycosaminoglycans in ECM appear to be associated with intestinal morphogenesis and/or cell differentiation (Bouziges e t a / ., 1991). Spatially, different ECM molecules, such as tenascin and fibronectin, are also shown to be distributed differently, e.g., along the cryptvillus axis, and their distribution changes during intestinal development (Aufderheide and Ekblom, 1988; Beaulieu et al., 1991; Simon-Assmann and Kedinger, 1993). The epithelial cells are intimately associated with ECM and their differentiation occurs as they migrate from the crypt to villus along the ECM. These associations suggest that the spatial and temporal regulation of the ECM will undoubtedly affect epithelial development. More direct evidence for a role of the ECM in intestinal development is obtained from experiments involving cocultures of epithelial and fibroblastic cells (Louvard et al., 1992). The differentiation of the epithelial cells is preceded by the formation of a basement membrane with ECM components synthesized by both types of cells. More importantly, antibodies to laminin can inhibit epithelial differentiation when added to the cocultures (Simo e f al., 1992). Although the aforementioned studies involve higher vertebrates, evidence is accumulating that cell-cell and cell-ECM interactions are important also during
226
Yun-Bo Shi and Atsuko Ishizwya-Oka
intestinal metamorphosis in amphibians. As discussed earlier, intestinal remodeling is TH-dependent and can be induced in organ cultures. By using fragments derived from different regions of the intestine, it has been demonstrated that the development of adult epithelium requires the presence of larval connective tissue (Ishizuya-Oka and Shimozawa, 1992a). The anterior small intestine in premetamorphic tadpoles has a single epithelial fold, the typhlosole, where the larval connective tissue is abundant, whereas other regions, including the posterior half of the small intestine, have only a thin layer of connective tissue surrounding the epithelium. When organ cultures of the anterior intestine are treated with TH, both larval epithelial cell death and adult cell proliferation and differentiation take place (Fig. 8B). In contrast, when the posterior intestine is used, only cell death is detected (Fig. 8C). More importantly, similar to higher vertebrates, when epithelium derived from any part of the small intestine is cocultured with the connective tissue from the typhlosole, adult epithelium development can be induced by TH treatment. This requirement of connective tissue has also been suggested by the temporal regulation of the developing connective tissue and adult epithelium as described above. The connective tissue is drastically remodeled during metamorphosis. Moreover, extensive cell contacts are present between the epithelium and the connective tissue. In addition, precursors of adult epithelial cells proliferate as cell nests or islets between the developing mesenchyme and degenerating larval epithelium. The origin of the adult epithelial cells is currently unknown. Since differentiated larval epithelial cells are mitotically active, it is possible that under certain conditions the larval epithelial may be induced to proliferate and give rise to adult epithelium (Marshall and Dixon, 1978b; Ishizuya-Oka and Shimozawa, 1987a). Based on these studies, it has thus been postulated that either direct mesenchymal-epithelial interaction or signals released by the mesenchymal cells may influence the fate of epithelial cells. It is likely that ECM is involved in mediating the mesenchymal-epithelial cell interaction. ECM could also directly regulate cellular events through its physical contacts with cells. Although there are no direct experimental demonstrations for these possible roles during intestinal remodeling, the basal lamina is altered during intestinal remodeling. In addition, a few ECM degrading/modifying metalloproteinase genes have been found to be expressed during this process. Most interesting among them is the Xenopus homolog of the mammalian stromelysin-3 gene, which is activated during both intestinal remodeling and tail resorption (Wang and Brown, 1993; Shi and Brown, 1993). In mammals, the gene has been found to be expressed in tissues undergoing cell death during development and in all human carcinomas examined (Lefebvre et al., 1992; Wolf et al., 1993; Basset et al., 1990). Similarly, it is expressed in the tail and intestine, two organs in which extensive cell death occurs during metamorphosis. Interestingly the gene is activated in the intestine before any epithelial cell death is visible. In addition, its expression is restricted to the fibroblastic cells of the connective tissue (Patterton et al., 1995). This is in sharp contrast to the expression of two other metal-
227
7 . Amphibian Intestinal Remodeling
loproteinase genes gelatinase A and stromelysin-1 genes (Table I; Patterton et a / ., 1995). Whereas stromelysin- 1 is expressed throughout intestinal develop-
ment and is unaffected by TH, the gelatinase A gene is activated after larval epithelial cells have begun to undergo apoptosis and thus later than stromelysin-3 gene during development. Similarly, during TH-induced precocious metamorphosis, stromelysin-3 gene is activated within hours of TH treatment, whereas the up-regulation of gelatinase A gene requires several days of treatment. These results suggest that gelatinase A may be involved in the degradation of ECM after cell death has occurred, as reflected in its delayed expression. In contrast, stromelysin-3 may be involved in the modification of ECM early during intestinal remodeling. The change in ECM could in turn influence cell fate in response to TH, i.e., cell death in the larval epithelium and proliferation of adult epithelial cells. The fibroblastic expression of stromelysin-3 reinforces the idea that mesenchymal cells are important participants of epithelial development.
IV. Summary and Prospects Intestinal development in anurans is a biphasic process. The embryogenesis of intestine resembles that in higher vertebrates. The subsequent remodeling process during metamorphosis to produce an adult organ is controlled by TH. Recent progress in studying TH action and its application to amphibian metamorphosis has provided considerable insights into the remodeling process. One possible model for the TH-induced gene regulation cascade of intestinal remodeling is presented in Fig. 9. It is assumed that TRs function as heterodimers with RXRs. In the absence of TH, the TR-RXR heterodimers can bind to TH response elements present in the target genes and repress the basal transcription of these genes (Fondell et a/., 1993; Damm et al., 1989; Sap et al., 1989; Ban-
Transcription Factors, e.g. TRP
-
b + TH
RXR-TR
Transcription repressed
Transcrlpllon activated
-
Down-stream response genes
odiiication of Others. e.g -extracellular matrix stromelysin-3 and surface. secretion of growth Direct response genes
Larval eDlthelial cell death. adult
ep’thelial proliferation celland differentiahon. etc
Regulation01 cellcell and cell-matrix interactions and signal transduction
Fig. 9 Thyroid hormone-induced gene regulation cascade that leads to intestinal remodeling. For simplicity, the model assumes that RXR-TR heterodimers mediate the effects of TH. It is possible, however, that TR homodimers and heterodimers with other receptors are also involved. In addition, downregulated, direct TH-response genes are not included in the model, as only a single gene in this class has been isolated and its identity is still unknown (Shi and Brown, 1993).
228
Yun-Bo Shi and Atsuko Ishizwya-Oka
iahmad et al., 1992; Ranjan et al., 1994). The binding of TH leads to conformational changes in the receptor complexes that in turn activate gene transcription. The products of these early response genes then participate in the activation of the remaining gene regulation cascade. Exactly how this occurs remains unknown. Interestingly, the early response genes include not only transcription factors but also other proteins such as metalloproteinases. The transcription factors could activate or repress downstream TH response genes directly. Other proteins are likely to assert their effect indirectly. For example, they could modify the ECM or cell surface. In addition, they could regulate and/or participate in signal transduction by growth factors, The cooperation between these complex intra- and extracellular processes eventually results in the degeneration of the larval organ and formation of the adult tissue. This simplified scheme immediately raises many questions. Although the mRNAs for TRs and RXRs are present in the intestine and the other tissues during metamorphosis (Yaoita and Brown, 1990; Kawahara et al., 1991; Y.-B. Shi, unpublished observations), it is unknown whether the mRNA levels reflect the protein levels. It also remains to be tested whether TR-RXR is indeed the functional complex in vivo and whether the heterodimer is responsible for the activation of the early response genes isolated to date. The majority of the early response genes are ubiquitous. Of the few intestinespecific genes, none of them have yet been identified by sequence analysis. It is of great interest to understand how the same genes expressed in tissues undergoing drastically different changes can exert their biological effects. It is likely that together with existing proteins in the intestine, these early genes regulate tissuespecific downstream genes, which in turn determine the tissue-specific transformation. An important issue is to establish the identity of these downstream genes. Unique to intestine is its simple organization in tadpoles and its more complex but morphologically well-characterized structure in adults. The strong resemblance of the amphibian intestine to the intestines of higher vertebrates in both embryogenesis and adult structure makes the amphibian intestine a good model system. The regulation of the remodeling by TH makes it possible to isolate genes involved in this process. Some of these genes may also be useful for studying early development of the intestine. Already, the villin and IFABP genes, two genes regulated indirectly by TH during remodeling, appear to be good markers for early intestinal epithelial differentiation. If any of the early TH response genes are expressed in the intestine during embryogenesis, they could exert similar functions in both the embryogenesis of the tadpole intestine and its remodeling during metamorphosis. Epithelial-mesenchymal cell interactions and the ECM appear to be important in amphibian intestinal remodeling, similar to their roles in other vertebrates. It is unclear how they affect this process. These could invoke direct cell-cell contacts, modulate the growth factors or other diffusible substances, or be involved
7 . Amphibian Intestinal Remodeling
229
in signal transduction through cell surface molecules such as integrins (see Damsky and Werb, 1992, for a review on integrins in signal transduction). Amphibian intestinal remodeling provides an excellent system for investigating these different aspects of the development. This process can be easily controlled by thyroid hormone in vivo as well as in organ cultures. In addition, amphibian intestine has been morphologically well characterized at the cytological level, thus making it easy to study at the molecular level. Most importantly, the availability of many genes involved in this process makes it possible to address these questions directly.
Acknowledgments We thank Drs. M. hzianowska-Kuznicki, M . Stolow, and J. Wong for critical reading of the manuscript and Ms. T. Vo for its preparation.
References Alpers, D. H., Strauss, A . W., Ockner, R. K., Bass, N. M., and Gordon, J. 1. (1984). Cloning of a cDNA encoding rat intestinal fatty acid binding protein. froc. Natl. Acad. Sci. U.S.A. 81, 313-317. Aufderheide, E., Ekblom, P. (1988). Tenascin during gut development: Appearance in the mesenchyme, shift in molecular forms and dependence on epithelial-mesenchymal interactions. J . Cell Eiol. 107, 2341-2349. Baniahmad, A,, Kohne, A. C., and Renkawitz, R . (1992). A transferable silencing domain is present in the thyroid hormone receptor, in the v-erbA oncogene product and in the retinoic acid receptor. EMEU J . 11, 1015-1023. Basset, P., Bellocq, J. P., Wolf, C., Stoll, I . , Hutin, P., Limacher, J. M . , Podhajcer, 0. L., Chenard, M. P., Rio, M . C . , and Chambon, P. (1990). A novel metalloproteinase gene specifically expressed in stromal cells of breast carcinomas. Nature (London) 348, 699-704. Beaulieu, J. F., Vachon, P. H., and Chartrand, S . (1991). Immunolocalization of extracellular matrix components during organogenesis in the human small intestine. Anat. Embryo/. 183, 363-369. Bjerknes, M., and Cheng, H. (1981). The stem-cell zone of the small intestinal epithelium. IV. Evidence for controls over orientation of boundaries between the stem-cell zone, proliferation zone, and the maturation zone. Am. J. Ann?. 160, 105-115. Bonneville, M. A. (1963). Fine structural changes in the intestinal epithelium of the bullfrog during metamorphosis. J . Cell Eiol. 18, 579-597. Bonneville, M. A., and Weinstock, M. (1970). Brush border development in the intestinal absorptive cells of Xenopus during metamorphosis. J. Cell Eiol. 44, 151-171. Bouziges, F., Simon-Assmann, P., Simo, P., and Kedinger, M. (1991). Changes in glycosaminoglycan expression in the rat developing intestine. Cell. Eiol. Int. Rep. 15, 97-106. Bretscher, A , , and Weber, K. (1979). Villin: The major microfilament-associated protein of the intestinal microvillus. froc. Narl. Acad. Sci. U.S.A. 76, 2321-2325. Bretscher, A,, and Weber, K. (1980). Villin is a major protein of the microvillus cytoskeleton which binds both G and F actin in a calcium-dependent manner. Cell (Cambridge, Mass.) 20, 839-847.
230
Yun-Bo Shi and Atsuko Ishizwya-Oka
Brooks, A. R., Sweeney, G., and Old, R. W. (1989). Structure and functional expression of a cloned Xenopus thyroid hormone receptors. Nucleic Acids Res. 17, 9395-9405. Buckbinder, L., and Brown, D. D. (1992). Thyroid hormone-induced gene expression changes in the developing frog limb. J. Biol. Chem. 267, 25786-25791. Cheng, H., and Leblond, C. P. (1974). Origin, differentiation and renewal of the four main epithelial cell types in the mouse small intestine. 1. Columnar cell. Am. J. Anal. 141, 503520. Cho, K. W. Y.,Morita, E. A., Wright, C. V. E., and De Robertis, E. M. (1991). Overexpression of a homeodomain protein confers axis-forming activity to uncommitted Xenopus embryonic cells. Cell (Cambridge, Mass.) 65, 55-64. Damm, K . , Thompson, C. C., and Evans, R. M. (1989). Protein encoded by v-erbA functions as a thyroid-hormone receptor antagonist. Nature (London) 339, 593-597. Damsky, C. H., and Werb, Z . (1992). Signal transduction by integrin receptors for extracellular matrix: Cooperative processing of extracellular information. Current Biol. 4, 772-78 I . Dauca, M., and Hourdry, J. (1985). Transformations in the intestinal epithelium during anuran metamorphosis. In “Metamorphosis” (M. Balls and M. Bownes, eds.), pp. 36-58. Clarendon Press, Oxford. Dauca, M., Hourdry, J . , Hugon, J. S . , and Menard, D. (1980). Amphibian intestinal brush border enzymes during thyroxine-induced metamorphosis: A biochemical and cytochemical study. Histochemistry 70, 33-42. Dauca, M., Hourdry, J., Hugon, J. S . , and Menard, D. (1981). Amphibian intestinal brush border membranes. 111. Comparison during metamorphosis of the protein, glycoprotein and enzyme patterns after gel electrophoretic separation of SDS-solubilized membranes. Comp. Biochem. Physiol. B 698, 15-22. Dauca, M., Bouziges, F., Colin, S . , Kedinger, M., Keller, J.-M., Schilt, J., Simon-Assmann, P., and Haffen, K. (1990). Development of the vertebrate small intestine and mechanisms of cell differentiation. Inr. J. Dev. Biol. 34, 205-218. Dodd, M. H. I., and Dodd, J. M. (1976). The biology of metamorphosis. In “Physiology of the Amphibia” (B. Lofts, ed.), pp. 467-599. Academic Press, New York. Donjacour, A. A . , and Cunha, G. R. (1991). Stromal regulation of epithelial function. In “Regulation Mechanisms in Breast Cancer” (M. Lippman and R. Dickson, eds.), pp 335-364. Kluwer Academic Publishers, Boston. Du Pasquier, L., and Flajnik, M. F. (1990). Expression of MHC class I1 antigens during Xenopus development. Dev. fmmunol. 1, 85-95. Evans, R. M. (1988). The steroid and thyroid hormone receptor superfamily. Science 240, 889895. Evans, T., Felsenfeld, G., and Reitman, M. (1990). Control of globin gene transcription. Annu. Rev. Cell Biol. 6, 95- 124. Figiel, A., Schilt, J., Dudouet, B., Robine, S . , and Dauca, M. (1987). Stage-specific polypeptide and villin expression during the intestinal epithelium substitution of the metamorphosing amphibian. Diferentiation (Berlin) 36, 116- 124. Figiel, A., Keller, J.-M., Schilt, J . , and Dauca, M. (1989). Stage-specific polypeptide and villin expression during thyroid-hormone-induced substitution of the amphibian intestinal epithelium. Differentiation (Berlinj 40, 166- 175, Flajnik, M. F., Hsu, E., Kaufman, I. F., and Du Pasquier, L. (1987). Changes in the immune system during metamorphosis of Xenopus. Immunol. Today 8, 58-64. Fondell, J. D., Roy, A. L., and Roeder, R. G.. (1993). Unliganded thyroid hormone receptor inhibits formation of a functional preinitiation complex: Implications for active repression. Genes Dev. 7, 1400-1410. Forman, B . M., and Sarnuels, H. H. (1990). Interaction among a subfamily of nuclear hormone receptors: The regulatory zipper model. Mol. Endocrinol. 4, 1293-1301,
7. Amphibian Intestinal Remodeling
23 1
Fox, H., Bailey, E., and Mahoney, R. (1972). Aspects of the ultrastructure of the alimentary canal and respiratory ducts in Xenopus laevis larvae. J . Morphol. 138, 387-406. Galton, V. A. (1983). Thyroid hormone action in amphibian metamorphosis. I n “Molecular Basis of Thyroid Hormone Action” (J. H. Oppenheimer and H. H. Samuels, eds.), pp 445-483. Academic Press, New York. Glass, G. B. J. (1968). “Introduction to Gastrointestinal Physiology.” Prentice-Hall, Englewood Cliffs, NJ. Green, R. P., Cohn, S. M., Sacchettini, J. C., Jackson, K. E . , and Gordon, J. I . (1992). The mouse intestinal fatty acid protein gene: Nucleotide sequence, pattern of developmetnal and regional expression, and proposed structure of its protein product. DNA Cell Biol. 11, 31-41. Green, S . , and Chambon, P. (1988). Nuclear receptors enhace our understanding of transcription regulation. Trends Genet. 4, 309-3 13. Grey, R. D. (1972). Morphogenesis of intestinal villi. I. Scanning electron microscopy of the duodenal epithelium of the developing chick embryo. J . Morphol. 137, 193-214. Gudernatsch, J. F. (1912). Feeding experiments on tadpoles. I. The influence of specific organs given as food on growth and differentiation: A contribution to the knowledge of organs with internal secretion. Arch. Enttz~icklungsrnech.Org. 35, 457-483. Gurdon, J. B. (1992). The generation of diversity and pattern in animal development. Cefl (Cambridge, Mass.) 68, 185-199. Haffen, K., Kedinger, M., Simon-Assmann, P. M., and Lacroix, B. (1982). Mesenchymedependent differentiation of intestinal brush-border enzymes in the gizzard endodem of the chick embryo. In “Embryonic Development. Part B: Cellular Aspects” (A. R. Liss, ed.), pp. 261-270. Alan R. Liss, New York. Haffen, K., Lacroix, B., Kedinger, M . , and Simon-Assmann, P. M. (1983). Inductive properties of fibroblastic cell cultures derived from rat intestinal mucosa on epithelial differentiation. Differentiation (Berlin) 23, 226-233. Hausen, P., and Riebesell, M. (1991). “The Early Development of Xenopus laevis: An Atlas of the Histology.” Springer-Verlag. Berlin. Heusser, S . , Colin, S . , Figiel, A., Huet, C., Keller, J. M., Pornet, P., Robine, S., Vandamme, J., Vandekerckhove, J . , and Dauca, M. (1992). Amphibian intestinal villin: Isolation and expression during embryonic and larval development. J. Cell Science 103, 699-708. Heyman, R. A,, Mangelsdorf, D. J., Dyck, J. A , , Stein, R. B., Eichele, G . , Evans, R. M., and Thaller, C. (1992). 9-Cis retinoic acid is a high affinity ligand for the retinoid X receptor. Cell (Cambridge, Mass.) 68, 397-406. Holtfreter, J. ( 1939). Studien zur ermittlung der gestaltungsfaktoren in der organentwicklung der amphibien. I. Dynamisches verhalten isolierter furchungszellen und entwicklungsmechanik der entodermorgane. Arch. Ennuicklungsrnech. Org. 139, 110-190. Hourdry, J., and Dauca, M. (1977). Cytological and cytochemical changes in the intestinal epithelium during anuran metamorphosis. Int. Rev. Cyfol., Suppl. 5 , 337-385. Hourdry, J., Chabot, J. G., Menard, D., and Hugon, 3. S. (1979). Intestinal brush border enzyme activities in developing amphibian Rana cafesbeiana. Cornp. Biochem. Physiol. A 63A, 121125. Ishizuya-Oka, A., and Mizuno, T. (1984). Intestinal cytodifferentiation in vitro of chick stomach endoderm induced by the duodenal mesenchyme. J . Embryol. Exp. Morph. 82, 163-176. Ishizuya-Oka, A,, and Mizuno, T. (1985). Chronological analysis of the intestinalization of chick stomach endoderm induced in vitro by duodenal mesenchyme. Roux’s Arch. Dev. Biol. 194, 301-305 Ishizuya-Oka, A . , and Shimozawa, A. (1987a). Development of the connective tissue in the digestive tract of the larval and metamorphosing Xenopus laevis. Anat. Anz. 164, 81-93. Ishizuya-Oka, A , , and Shimozawa, A. (1987b). Ultrastructural changes in the intestinal connective tissue of Xenopus laevis during metamorphosis. J . Morphol. 193, 13-22.
232
Yun-Bo Shi and Atsuko Ishizwya-Oka
Ishizuya-Oka, A,, and Shimozawa, A. (1990). Changes in lectin-binding pattern in the digestive tract of Xenopus laevis during metamorphosis. 11. Small intestine. J . Morphol. 205, 9-15. Ishizuya-Oka, A,. and Shimozawa, A. (1991). Induction of metamorphosis by thyroid hormone in anuran small intestine cultured organotypically in vitro. In V i m Cell Dev. Biol. 27A, 853857. Ishizuya-Oka, A , , and Shimozawa, A. (1992a). Connective tissue is involved in adult epithelial development of the small intestine during anuran metamorphosis in vitro. Roux’s Arch. Dev. Ei01. 201, 322-329. Ishizuya-Oka, A,, and Shimozawa, A. (1992b). Programmed cell death and heterolysis of larval epithelial cells by macrophage-like cells in the anuran small intestine in vivo and in vitro. J . Morphol. 213, 185-195. Ishizuya-Oka, A., Shimozawa, A,, Takeda, H., and Shi, Y.-B. (1994). Cell-specific and spatiotemporal expression of intestinal fatty acid-binding protein gene in the small intestine during amphibian metamorphosis. Rous’s Arch. Dev. Biol. 204, 150- 155. Kaltenbach, J. C . , Lipson, M. J., and Wang, C. H. K. (1977). Histochemical study of the amphibian digestive tract during normal and thyroxine-induced metamorphosis: Alkaline phosphatase. J . Exp. 2001.202, 103-120. Kawahara, A., Baker, B. S., and Tata, J. R. (1991). Developmental and regional expression of thyroid hormone receptor genes during Xenopus metamorphosis. Development (Cambridge, U K ) 112,933-943. Kedinger, M., Simon-Assmann, P., Alexandre, E.m and Haffen, K. (1987). Importance of a fibroblastic support for in vitro differentiation of intestinal endodermal cells and for their response to glucocorticoids. Cell Difer. 20, 171-182 Kelley, C . , Blumberg, H., Zon, L. I . , and Evans, T. (1993). GATA-4 is a novel transcription factor expressed in endocardium of the developing heart. Development (Cambridge, U K ) 118, 8 17-827. Kemp, N. E. (1946). Regulation in the endoderm of the tree frog Hyla regilla. Univ. Cali$, Eerkekley, Publ. Zool. 51, 159-183. Kikuyama, S . , Kawamura, K., Tanaka, S . , and Yamamoto, K. (1993). Aspects of amphibian metamorphosis: Hormonal control. Int. Rev. Cyrol. 145, 105- 148. Kordylewski, L. (1983). Light and electron microscopic observations of the development of intestinal musculature in Xenopus. Anat. Forsch. 97, 719-734. Kurokawa, R., Yu, V. C., Naar, A,, Kyakumoto, S . , Han, Z . , Silverman, S . , Rosenfeld, M. G., and Glass, C. K. (1993). Differential orientations of the DNA-binding domain and carboxyterminal dimerization interface regulate binding site selection by nuclear receptor heterodimers. Genes Dev. 7 , 1423-1435. Le Douarin, N. (1964). Etude expkrimentale de I’orgnogenese du tube digestif et du foie chez I’embryon de poulet. Bull. Biol. Fr. Belg. 98, 543-676. Lefebvre, O., Wolf, C., Limacher, J.-M., Hutin, P., Wendling, C., LeMeur, M., Basset, P., and Rio, M.-C. (1992). The breast cancer-associated stromelysin-3 gene is expressed during mouse mammary gland apoptosis. J . Cell. Eiol. 119, 997-1002. Leloup, J., and Buscaglia, M. (1977). La triiodothyronine: Hormone de la mktamorphose des amphibiens. C.R. Hebd. Seanes Acad. Sci. 284, 2261-2263. L‘Hermite, A . , Hourdry, J., and Ferrand, R. (1988). Investigations of endocrine cells in the gastrointestinal tract and pancreas during the metamorphosis of an anuran (Alytes obstetricans L.): Histochemical detection of APUD cells. Gen. Comp. Endocrinol. 69, 38 1-390. Louvard, D., Kedinger, M., and Hauri, H. P. (1992). The differentiating intestinal epithelial cell: Establishment and maintenance of functions through interactions between cellular structures. Ann. Rev. Cell. Biol. 8, 157-195. Mamajiwalla, S . N., Fath, K. R., and Burgess, D. R. (1992). Development of the chicken intestinal epithelium. Curr. Top. Dev. Eiol. 26, 123-143.
7. Amphibian Intestinal Remodeling
233
Marsh, M. N., and Trier. J . S . (1974). Morphology and cell proliferation of subepithelial fibroblasts in adult mouse jejunum. I. Structural features. Gastroenterology 67, 622-635. Marshall, J. A., and Dixon, K. E. (1978a). Cell specialization in the epithelium of the small intestine of feeding Xenopus laevis. J. Anat. 126, 133-144. Marshall, J. A.. and Dixon, K. E. (1978b). Cell proliferation in the intestinal epithelium of Xenopus laevis tadpoles. J. Exp. Zoo/. 203, 31-40. McAvoy, J. W., and Dixon, K. E. (1977). Cell proliferation and renewal in the small intestinal epithelium of metamorphosing and adult Xetzopus laevis. J. Exp. Zoo/. 202, 129-138. McAvoy, J. W.. and Dixon, K. E. (1978a). Cell specialization in the small intestinal epithelium of adult Xenopus laevis: Structural aspects. J . Anat. 125, 155-169. McAvoy, J. W., and Dixon, K. E. (1978b). Cell specialization in the small intestinal epithelium of adult Xenopus laevis: Functional aspects. J . Anat. 125, 237-245. McCinnis, W., and Krumlauf, R. (1992). Homeobox genes and axial patterning. Cell (Cumbridge, Mass.) 68, 283-302. Nieuwkoop, P. D., and Faber, J. (1967). “Normal Table of Xenopus laevis.” North-Holland Publ., Amsterdam. Okada, T. S . (1960). Epithelio-mesenchymal relationships in the regional differentiation of the digestive tract in the amphibian embryo. Roux’ Arch. Entwicklungsmech. Org. 152, 121. Orkin, S. H. (1990). Globin gene regulation and switching: Circa 1990. Cell (Cambridge, Mass.) 63, 665-672. Patterton, D., Hayes. W. P., and Shi, Y.-B. (1995). Transcriptional activation of the matrix metalloproteinase gene stromelysin-3 coincides with thyroid hormone-induced cell death during frog metamorphosis. Dev. Biol. 167, 252-262. Perlmann, T., Rangarajan, P. N.. Umesono, K . . and Evans R. M. (1993). Determinants for selective RAR and TR recognition of direct repeat HREs. Genes Dev. 7, 1411-1422. Ranjan, M., Wong, J., and Shi, Y.-B. (1994).Transcriptional repression of Xenopus TRP gene is mediated by a thyroid hormone response element located near the start site. J. Biol. Chem. 269, 24699-24705. Reeder, W. C. (1964). The digestive system. In ”Physiology of the Amphibia.” (J. A. Moore, ed.), pp. 99-149. Academic Press, New York. Sap, J., Munoz, A,, Damm, K., Goldberg, Y.; Ghysdael, J., Leutz, A,, Berg, H., and Vennstrom, B.. (1986). The C-erb-A protein is a high afftnity receptor for thyroid hormone. Narure (London) 324, 635-640. Sap, J., Munoz, A,, Schmitt, J., Stunnenberg, H. and Vennstrom, B. (1989). Repression of transcription mediated at a thyroid hormone response element by the v-erb-A oncogene product. Nature (London) 340, 242-244. Scalise, F. W., and Vigna, S . R. (1988). Temporal pattern of appearance and distribution of cholecystokinin-like peptides during development in Xenopus laevis. Gen. Comp. Endocrinol. 72, 303-3 I I . Schneider, M. J . , and Galton, V. A,. (1991). Regulation of c-erbA-a messenger RNA species in tadpole erythrocytes by thyroid hormone. Mol. Endocrinol. 5 , 201-208. Segal, C. H.. and Petras, R. E. (1992). Small intestine. I n “Histology for Pathologists” (S. S . Sternberg, ed.), pp. 547-571. Raven Press, New York. Shi, Y.-B. (1994). Molecular biology of amphibian metamorphosis: A new approach to an old problem. Trends Endocrinol. Metab. 5 , 14-20. Shi, Y.-B., and Brown, D. D. (1993). The earliest changes in gene expression in tadpole intestine induced by thyroid hormone. J. Biol. Chem. 268, 203 12-203 17. Shi, Y.-B., and Hayes, W. P. (1994). Thyroid hormone-dependent regulation of the intestinal fatty acid-binding protein gene during amphibian metamorphosis. Dev. Biol. 161, 48-58. Shibayama, T., Carboni, J. M., and Mooseker, M. S. (1987). Assembly of the intestinal brush
234
Yun-Bo Shi and Atsuko Ishizwya-Oka
border: Appearance and redistribution of microvillar core proteins in developing chick enterocytes. J . Cell Biol. 105, 335-344. Simo, P., Simon-Assmann, P., Bouziges, F., Leberquier, C., Kedinger, M., Ekblom, P., and Sorokin, L. (1991). Changes in the expression of laminin during intestinal development. Development (Cambridge, U K ) 112, 477-487. Simo, P., Simon-Assmann, P., Arnold, C., and Kedinger, M. ( I 992). Mesenchyme-mediated effect of dexamethasone on laminin in cocultures of embryonic gut epithelial cells and mesenchyme-derived cells. J . Cell. Sci. 101, 161-171. Simon-Assmann, P., and Kedinger, M. (1 993). Heterotypic cellular cooperation in gut morphogenesis and differentiation. Semin. Cell Biol. 4, 221-230. Smith-Gill, S . J., and Carver, V. (1981). Biochemical characterization of organ differentiation and maturation. In “Metamorphosis: A Problem in Developmental Biology” (L. I. Gilbert and E. Frieden, eds.), pp. 491-544, Plenum, New York. Spemann, H., and Mangold, H. (1924). Uber induktion von embryonal anlagen durch implantation artfremder organisatoren. Arch. Entwicklungsmech. Org. 100, 599-638. Stallmach, A,, Hahn, U., Merker, H. J., Hahn, E. G . , and Riecken, E. 0. (1989). Differentiation of rat intestinal epithelial cells is induced by organotypic mesenchymal cells in vitro. Gut 30, 959-970. Sugimoto, K., Ichikawa, Y., and Nakamura, I. (1981). Absorption of protein molecules by the small intestine of the bullfrog tadpole, Rana catesbeiana. J. Exp. Zool. 215, 53-62. Sweetser, D. A., Birkenmeier, E. H . , Klisak, I. J., Zollman, S . , Sparks, R. S . , Mohandas, T., Lusis, A. J., and Gordon, I. I. (1987). The human and rodent intestinal fatty acid binding protein genes: A comparative analysis of their structure, expression, and linkage relationships. J . Biol. Chem. 262, 16060-16071. Sweetser, D. A., Hauft, S. M., Hoppe, P. C., Birkenmeier, E. H., and Gordon, J. 1. (1988). Transgenic mice containing intestinal fatty acid-binding protein-human growth hormone fusion genes exhibit correct regional and cell-specific expression of the reporter gene in their small intestine. Proc. Narl. Acad. Sci. U.S.A. 85, 9611-9615. Tilney, L. G., and Mooseker, M. S. (1971). Actin in the brush border of epithelial cells of the chicken intestine. Proc. Narl. Acad. Sci. U.S.A. 68, 261 1-2615. Torihashi, S. (1990). Morphological changes of the myenteric plexus neurons in the bullfrog (Rana catesbeiana) duodenum during metamorphosis. J . Comp. Neurol. 302, 54-65. Wang, Z., and Brown, D.D. (1993). Thyroid hormone-induced gene expression program for amphibian tail resorption. J . Biol. Chem. 268, 16270-16278. Weinberger, C . , Thompson, C. C., Ong, E. S., Lebo, R., Gruol, D. J., and Evans, R. M. (1986). The c-erb-A gene encodes a thyroid hormone receptor. Nature (London) 324, 641646. Weiser, M. M., Sykes, D. E., and Killen, P. D. (1990). Rat intestinal basement membrane synthesis. Lab. Invest. 62, 325-330. White, B. A., and Nicoll, C. S. (1981). Hormonal control of amphibian metamorphosis. In “Metamorphosis: A Problem in Developmental Biology” (L. I. Gilbert and E. Frieden, eds.), pp. 363-396. Plenum, New York. Wolf, C., Rouyer, N., Lutz, Y., Adida, C., Loriot, M., Bellocq, J.-P., Charnbon, P., and Basset, P. (1993). Stromelysin 3 belongs to a subgroup of proteinases expressed in breast carcinoma fibroblastic cells and possibly implicated in tumore progression. Proc. Natl. Acad. Sci. U.S.A. 90, 1843-1847. Wright, C. V. E., Schnegelsberg, P., and De Robertis, E. M. (1988). XLH box 8: A novel Xenopus homeoprotein restricted to a narrow band of endodern. Developmenr (Cambridge. UK) 104, 787-794. Wyllie, A. H . , Kerr, J. F. R., and Curie, A. R. (1980). Cell death: The significance of apoptosis. In?. Rev. Gyrol. 68, 251-306.
7. Amphibian Intestinal Remodeling
235
Yaoita, Y., and Brown, D. D. (1990). A correlation of thyroid hormone receptor gene expression with amphibian metamorphosis. Genes Dev. 4, 1917- 1924. Yaoita, Y., Shi, Y.-B., and Brown, D. D. (1990). Xenopus luevis a and p thyroid hormone receptors. Proc. Natl. Acad. Sci. U.S.A. 87, 7090-7094. Yasugi, S . (1993). Role of epithelial-mesenchymal interactions in differentiation of epithelium of vertebrate digestive organs. Dev. Growth Di’er. 35, 1-9. Yoshizato, K. (1989). Biochemistry and cell biology of amphibian metamorphosis with a special emphasis on the mechanism of removal of larval organs. f n t . Rev. Cyrol. 119, 97-149. Yu, Y. C . , Delsert, C., Andersen, B., Holloway, J. M . , Devary, 0. V., Naar, A. M . , Kim, S . Y., Boutin, J. M., Glass, C. K . , and Rosenfeld, M. G . (1991). RXRP: A coregulator that enhances binding of retinoic acid, thyroid hormone, and vitamin D receptors to their cognate response elements. Celt (Cambridge, Mass.) 67, 125 1- 1266. Zhang, X., and Pfahl, M . (1993). Regulation of retinoid and thyroid hormone action through homodimeric and heterodimeric receptors. Trends Endocrinol. Metub. 4, 156-162.
This Page Intentionally Left Blank
Index
A A23 187, induced acrosorne reaction, 65-66 Abnormalities, fertilization, 72-74 Acrosome reaction human sperm, 64-67 sea urchin sperm, 44-45, 47-49 Actin cable, assembly at wound margin, 184-188 purse string, contraction, 182-183 Activin-like signaling, synergy with fibroblast growth factor, 1 17- 1 I8 Alzheimer’s disease, elevedted interleukin-l , 156 Amphibians, intestine embryogenesis, 207-2 12 remodeling during metamorphosis, 212-227 Aneuploidy rates, human oocytes, 82-84 Animal hemisphere, dorsal determinants in, 118-123 Animal-vegetal axis, early Xenopus embryo, 105 Anomalies, developmental, and embryonic cleavage, 76-81 Anterior-posterior axis, early Xenopus embryo, 105 Anti-Miillerian hormone marker of Sertoli cell development, 12 produced by Sertoli cells, 8 role in testis determination, 9 Antioxidant pathway, bcl-2 role, 154- 155 Apoptosis c-myc-induced, 160 in mammals, 150 p.53 role, 159 reactive oxygen species in, 154 Apoptotic bodies, primary epithelial, 213 Autosomal loci, interactive, 16- 17 Axes cardinal, early Xenopus embryo, 105- 106 animal-vegetal, 105 anterior-posferior, 105 dorsal-ventral, 105- 106 dorsal, determinants, 115-123
Azoospermia, obstructive , intracytoplasrnic sperm injection for, 75-76
B Basal lamina, wounding and reepithelialization, 181 Bax, interaction with Bcl-2, 153 bcl-2 gene expression, 15 1- I52 overexpression, 150-155 bcl-2 gene family, 152-153 Bcl-2 protein, 152- 155 Bindin, sperm ligand, role in gamete adhesion, 43 Blastoc ysts cell numbers, 80-81 development, 90-91 stages, 76-77 Blastomeres animal, containing maternal RNA, 122-123 disaggregated, 8 I individual fate, 104 response to dorsal mesoderm inductive signaling, 123- 125 tracer injection, 104, 106, 119-121 Xenopus, stereotyped, 111-112
C
Caenorhabditis elegans. programmed cell death, 140-149 Calcium induced acrosome reaction, 65 mobilization in sea urchin fertilization, 54-55 Capacitation, sperm, and acrosome reaction, 64-67 Carbohydrates, sperm receptor, 48-49 CD46, 68 fed-3 gene, 144-148 ced-4 gene, 144-148 ced-9 gene, 147-148
237
238 Ced-3 protein, 145-147 similarity to ICE, 155-158 Ced-4 protein, 145-147 Cell-cell signaling, in fate determination, 123-129 Cell-extracellular matrix interactions, in intestinal remodeling, 224-227 Cell fate and inheritance of maternal cytoplasmic factor, 113-123 and position in mitotic pattern, 110-112 Cell lineages, study methodology, 103- 105 Cell number, inner mass cells and blastocysts, 80-81 Cell proliferation, during embryonic wound healing, 190 Cell types, gonadal bipotentiality, 8-9 origins, 5-7 Central nervous system, bcl-2 expression, 150-151 c-fos gene, continuous and transient expression, 161-162 Chemoattraction, and activation of sea urchin sperm, 41-42 Chicken, embryo wound healing, 183-186, I 98 Chromatin, accessibility, I8 Chromosomes X dosage compensation, 8 and Sty expression, 25-28 X and Y, gene isolation, 3-4 Y, rapid evolution, 23 Cleavage embryonic, and developmental anomalies, 76-81 proteolytic, 146, 158 stages, signaling events during, 125- 129 Xenopus patterns, 106- 108 c-myc gene, induced apoptosis, 160 Cocultures D1.l with V2.1, 121-122 embryo, 91-92 Collagen in fetal wound healing, 193-194 meshwork remodeling, 180 Conservation, evolutionary, programmed cell death, 149-150 Contraction actin purse string, 182- I83
Index adult wound connective tissue, 178-179 embryonic wound mesenchyme, 189- 190 Conversion phenomenon, in gonadal cells, 9 c-re1 gene, apoptotic cells expressing, 162 Cross-hybridization, sea urchin sperm receptor cDNA, 46-47 Cryopreservation oocytes, 92 oocytes and embryos, 84-87 sperm, 76 Cumulative embryo score, 78-79 Cumulus cells, atretic, oocytes surrounded by, 62-63 Cytogenetic analysis, human oocytes, 83-84 Cytoplasm pigmented cortical, rotation, 105-106 swirl, 116-117 Cytoplasmic factor, maternal, inheritance, 113- I23
D Decondensation, sperm head, 71-72 Degradation engulfed cells in Caenorhabdiris elegans, 148-149 sperm receptor after fertilization, 51-54 Deletion studies, Xenopus embryos, 119 Determinants dorsal, in animal hemisphere, 118-123 for dorsal axis in Xenopus, 115-1 18 maternal cytoplasmic, 114-1 15 Development connective tissue, 217 epithelial, mesenchyme role, 225-227 and fertilization, human, 59-92 fetal, transition to adult-like healing, 197 male in absence of Sry, 28 triggering by testis-determining gene, 2-3 urogenital ridge, 14- 15 Differentiation cellular, initiation, 13-14 dorsal axial, 120- 123 gonadal, 4- 12 larval small intestine, 209-210 mature sensory neurons, 152 Digestive tract, amphibian, formation, 209 Dimethyl sulfoxide, and chromosomal abnormalities, 86-87 Dissociation, embryo, 127
Index DNA bent, SRY affinity, 22 genomic, sea urchin species, 45-47 Dorsal axis, determinants, 115-123 Dorsal bias independent of mesoderm induction, 129-130 molecular nature, 121-123 Dorsal mesoderm, inductive signaling, 123-125 Dorsal root ganglion neuronal cell death, 149-151 neuronal survival, 155-156 Dorsal-ventral axis, 105- 106 Dosage compensation and sex determination, 25-27 X-chromosome, 8 Drosophila, neuroblast formation, 126
239 Epidermal cells, marginal, actin purse string contraction, 182- I83 Epididymis, role in normal reproduction, 66-67 Epithelial cells tadpole intestine, 210 in villus/crypt, 205-207 wound, crawling, 180-181 Epithelium amphibian intestine, transformation, 213-217 larval, replacement by adult form, 224 Equatorial furrow, Xenopus embryo, 1 1 I Evolution, rapid, Y chromosome, 23 Evolutionary conservation, programmed cell death, 149- 150 Exocytosis, acrosomal, Na*+/sup/H+/supexchanges, 64 Extracellular matrix-cell interactions, in intestinal remodeling, 224-227
E EIB protein, 55-kDa. 163 Eggs, see also Oocytes activation, in sea urchin species, 54-55 plasma membrane, gamete interactions at, 42-52 Egg receptor, sea urchin sperm carbohydrate role, 48-49 developmental expression and fate, 50-54 molecular profile, 44-48 Elongation, in D1.l explants, 120-121 Embryo cryopreservation, 84-87 dissociation, 127 genetic errors, 82-84 human cleavage, 76-77 morphology and scoring, 77-79 metabolism and viability, 88-91 monopronuclear and polypronuclear, 73-74 wound healing mechanisms, 175- 199 wounding, inflammation occurrence, 191- 193 XY preimplantation, 19 Embryo development rating, 78 Embryogenesis, amphibian intestine, 207-2 I2 Engulfment cells associated genes, 148 embracing doomed cells, 141 Enterocytes, specific genes, 21 1-212
F Fate determination basic mechanisms, 104-105 cell-cell signaling in, 123- 129 Fate mapping, Xenopus, 108 Fertilization, see also I n vitro fertilization abnormalities, 72-74 micromanipulative techniques, 74-76 multistep recognition process, 40 and phagocytosis, 69-72 sperm receptor fate after, 50-54 Fetus gonadal Sty expression, 13-14 human, repair of embryonic defects, 198-199 transition to adult-like healing, 197 wound healing environment, 193- 196 mechanisms, 175-199 Fibrin clot, sealing adult skin wounds, 176- 177 Fibronectin, in fetal wound healing, 194-195 Flagellar motility, speract receptor, 41-42 Fluorescent in siru hybridization, 73, 82, 84, 92 Fluorochromes, polynucleotide-specific, 8 1
G Gametes genetic errors, 82-84
240
lndex I
Gametes (cont.) human, interactions in virro, 67-68 interactions at egg plasma membrane, 42-52 recognition, in sea urchin fertilization, 39-55 Gap junctions, in dorsal axis formation, 127 Gene expression early, in urogenital ridge, 9-12 during gut embryogenesis, 210-212 Genes enterocyte-specific, 21 1-212 homeobox, 9-12 regulation by thyroid hormone, 22 1-224 Genetic control, programmed cell death in Caenorhabdiris elegans. 141- 149 Genetic errors, in gametes and embryos, 82-84 Genetic mosaic analysis, ced-3 and ced-4, 145-146 Genome, zygotic activation, 108 Germ cells, migration to gonads, 7 GI ycoproteins complex on egg surface, 43 human zona ZP3, 65-68 Gonads arising within urogenital tract, 4-7 cell types, 5-7, 8-9 fetal, S l y expression, 13-14 Granulation tissue, filling wound space, 179- I80 GTP-binding protein, rho, 184-187
ICE, see Interleukin- I P-converting enzyme Inflammation, after wounding in embryo, 19I- I93 Inflammatory response, adult, neutrophil and macrophage role, 177 Inheritance, maternal cytoplasmic factor, and cell fate, 113-123 Injection, see also Intracytoplasmic sperm injection immotile sperm into perivitelline space, 68 sperm, micromanipulative techniques, 74-76 Interleukin- 1P-converting enzyme, similarity to Ced-3, 146, 155-158 Interspecies comparisons, DNA binding domain, 22-25 Intestinal fatty acid binding protein expressed in enterocytes, 212 regulated by thyroid hormone, 224 Intestine amphibian, embryogenesis, 207-212 remodeling during metamorphosis, 21 2-227 Intracytoplasmic sperm injection correction of polyspermy, 63 for obstructive azoospermia, 75-76 treatment of subfertile male, 92 In virro fertilization laboratories, 73-74 oocytes hyperstimulated for, 61 repeated attempts, 90
H
K
Hermaphrodite-specific neuron, programmed cell death, 143-144 Hernia, diaphragmatic, embryonic surgical repair, 198-199 Heterodimers, thyroid hormone receptorretinoid X receptor, 221, 227-228 Histogenesis, larval small intestine, 209-210 HMG box domain, predicted structure, 21-22 Homeobox genes, 9-12 Human fertilization and development, 59-92 fetus, repair of embryonic defects, 198-199 sperm, acrosome reaction, 64-67 Hybridization, see also Cross-hybridization fluorescent in siru, 73, 82, 84, 92
Karyotypes, scoreable, 87 Keratin filament, formation in Xenopus, 188
L Lamellipodial crawling in reepithelialization, 178 into wound space, 181 Leydig cells, and Sry expression, 14 Lineage tracing techniques, 103- 104
M Macrophages and neutrophils, in adult inflammatory response, 177 recruitment to wound surface, 193
Index Mammals homologs of nematode programmed cell death genes, 150-158 sex determination, Sr.y role, 1-29 stromelysin-3 gene, 226-227 Markers molecular, urogenital ridge, 10- 11 Sertoli cell development, 12 Marsupials, pouch wound healing, 196- 197 Maturation human oocytes, 61-63 oocyte, in unstimulated cycles, 91-92 sperm, and acrosome reaction, 64-67 Maturation promoting factor, during oocyte arrest, 69-70 Media dissociation, 127 vitrification, 87 Mesenchyme embryonic wound, contraction, 189- 190 importance in epithelial development, 225-227 Mesoderm dorsal, inductive signaling, 123- 125 induction, dorsal bias independent of, 129-130 intermediate, Lim-l expression, 11 Mesonephros and cord formation, 13-14 primordial germ cells entering gonad via, 7 Messenger RNA Anl-3, 114 ICE, 157 maternal, 129 Xwnt-11, 115, 117-118 Metabolism, embryo, 88-91 Metamorphosis, intestinal remodeling during, 212-227 Micromanipulation, fertilization techniques, 74-16 Mitotic pattern, position, and cell fate, 110-112 Morphology, human embryos, 77-79 Motor neurons primary mean number, 1 12 number increase, 126- 127 progenitors, 109- I 10 serotonergic, 143-144 Mouse bcl-2-deficient, 154-155
24 1 fetal lip wound, 194-195 oocytes, vitrification, 86-87 Poschiavinus strain, I5 SF-I knockout, 12 weaver mutant, 161 X x l i r y transgenic, 4, 16-17 Movements morphogenetic, driven by purse string, 188- 189 tissue, of wound closure, 190-191 Miillerian inhibiting substance, see AntiMiillerian hormone Muscle, thickening during metamorphosis, 219 Mutants & - A and abd-B, 11 red-3 and red-4, 144-146 murine weaver, 161 Y-chromosome deletion, 15, 17-18 Mutations gain-of-function, 144, 147-148 G145E. 153 homeotic, 141-144 interfering with sex determination, 15 loss-of-function, 163- 164 Myofibroblasts, in contraction of adult wound connective tissue, 178- 179
N Nedd-2 gene, 157 nur-Z gene, 148-149 Nematode, see Caenorhabditis elegans Neurons, see also Motor neurons; Sensory neurons dorsal root ganglion, survival, 155-156 hermaphrodite-specific, programmed cell death, 143-144 mature sensory, differentiation, 152 Rohon-Beard mean number, 112 number increase, 126-127 progenitors, 109- 110 Neutrophils absence at wound site, 193 and macrophages, in adult inflammatory response, 177 NF-&kgr;B activation, 162 Nieuwkoop center, 117-1 18 Nonsynonymous-to-synonymous ratio, 23
242 0 Obstructive azoospermia, intracytoplasmic sperm injection for, 75-76 Oncogenesis, and programmed cell death, 158-162 Oocytes, see also Eggs activation, 69-72 cryopreservation, 84-87 human aneuploidy rates, 82-84 maturation, 61-63 molecular stratification, 113-1 14 Overexpression, bcl-2, 150- 155
P p53, in apoptosis, 159 Parthenogenesis, with haploid chromosome complement, 73 Perivitelline space injection of immotile sperm, 68 sperm receptor isoform deposition, 5 1-54 Phagocytosis, during egg-sperm fusion, 68-72 Plasma membrane, egg, gamete interactions at, 42-52 Polymerase chain reaction, reverse transcription, 13, 19 Polypeptide backbone, sperm receptor, 48-49 Polypeptide patterns, human oocytes, 81 Polyspermy correction, 63 preventive mechanisms, 52-54 Pregnancy rate, and cumulative embryo score, 79 Primary germ layer, lineage restriction, 109-1 10 Progenitor cells, embryonically generated, 142-143 Programmed cell death, see also Apoptosis and antiviral responses, 162-163 in Caenorhabditis elegans, 140- 149 and oncogenesis, 158- 162 primary epithelium, 213-217 in vertebrates, 149-163 Proliferation, larval epithelial cells, 226 Pronucleus appearance after insemination, 76-77 dissolution, 72 Protease inhibitors, cysteine and serine, 157 Pseudoautosomal boundary, and Sty gene isolation, 3-4
Index Purse string actin assembly and closure, 184-188 contraction, 182-183 contractile, closing of embryonic wound, 188-189 Pyruvate, uptake by human embryo, 88
R Rapid-cooling techniques, oocytes and embryos, 85-87 ras gene, in intracellular signaling pathway, 160- 16 1 Reactive oxygen species in apoptosis, 154 and NFKB activation, 162-163 Recognition process, multistep, in sea urchin fertilization, 40 Recombination events, sex-reversing, 17 Reepithelialization adult tissues, 178 embryonic wound, 182- 183 Remodeling collagen meshwork, I80 intestinal, during metamorphosis, 212-227 Retinogenic area, ventral cells moving to, 124-125 Retinoid X receptors, and thyroid hormone receptor, heterodimers, 22 1 RNA maternal, animal blastomeres containing, 122- 123 messenger, see Messenger RNA mixed with lineage tracer, 120- 121 testis circular transcript, 19-20 Rohon-Beard neurons mean number, 1 12 number increase, 126-127 progenitors, 109-1 10 Rotation cortico-cytoplasmic, 115-1 17 pigmented cortical cytoplasm, 105- 106 R-ras, interaction with Bcl-2, 152- 153
S Scamng, in adult wound healing, 179-180 Scoring systems, human embryos, 77-79 Sea urchin, see Strongylocentrorus purpurarus Semen analysis, computer-automated, 66-67 Sensory neurons, mature, differentiation, 152
Index Sertoli cells anti-Miillerian hormone produced by, 8 differentiation, 13-14 Sex determination classic views, 2-4 and dosage compensation, 25-27 Sex reversal frequency, 15-16 interspecies comparisons, 22-25 and sox-9 gene, 12 X-linked, 27-28 Signaling activin-like, 117-1 18 cell-cell, in fate determination, 123-129 mesoderm-inductive, 120-122 Skin, adult in fetal environment, 196-197 wounds, 176-177 Slow-freezing techniques, oocytes and embryos, 85-87 Southern blot analysis, sea urchin sperm receptor, 45-47 Sox gene, 27 Speract receptor, in flagellar motility, 41-42 Sperm activation, and chemoattraction, 41-42 capacitation and maturation, 64-67 egg receptor carbohydrate role, 48-49 developmental expression and fate, 50-54 molecular profile, 44-48 injection techniques, 74-76 intracytoplasmic sperm injection technique correction of polyspermy, 63 for obstructive azoospermia, 75-76 treatment of subfertile male, 92 Sry gene, role in mammalian sex determination, 1-29 SRY protein, predicted structure, 21-22 Stratification, molecular, oocyte, 113-1 14 Stromelysin-3 gene, 226-227 Strongylocentrotus purpuratus, fertilization, gamete recognition in, 39-55 Subunit composition, sea urchin sperm receptor, 47-48 Suicide program, cellular, 163-164
T Tenascin, in fetal wound healing, 194-195 Tension actin cable, 187- 188 within wound connective tissue, 194
243 Testis adult, Sry function, 19-21 determining gene, triggering male development, 2-3 Testis cord, formation, 7, 13-14 Thyroid hormone control of intestinal remodeling, 219-221 gene regulation by, 221-224 Thyroid hormone receptor, and retinoid X receptor, heterodimers, 22 1 Tissues adult, reepithelialization and contraction, 178 connective, development, 217 fetal, healing, 197-198 gonadal, cells contributing to, 5-7 movements, of wound closure, 190-191 Tracers, blastomeres injected with, 104, 106, 119-121 Transcript genital ridge, structure, 21-28 testis circular, 19-20 Transcription factors expressed during epithelial differentiation, 212 in intestinal metamorphosis, 222-223 Transformation, cellular, amphibian intestine, 2 12-2 19 Transforming growth factor P,/inf, in fetal wound healing, 195-196 U
Urogenital ridge development, 14-15 early gene expression, 9-12 mesonephric portion, 7 Urogenital system, origin, 4-5 UV irradiation, Xenopus embryos, 114-1 17
V Vertebrates, programmed cell death, 149- 163 Vg-1, 115, 117-118 Viability, embryo, 88-91 Villin, in endodermal cell apical domain, 21 1-212 Villuslcrypt axis, amphibian, 205-207 Vitrification, oocytes, 85-87
W Wound healing adult, review, 176- 180
244 Wound healing (con!.) embryonic historical studies, 180- 18 1 role of cell proliferation, 190 fetal, environment, 193-196 Wounds, embryonic closure by purse string, 188-189 reepithelialization, 182- 183
X Xenopus laevis cell lineage determination, 103- 130
Index intermediate keratin filament, 188 small intestine larval, 208-210 during metamorphosis, 214-220 study advantages, 105- 108
Z Zona pellucida human, salt-stored, 67-68 important properties, 69-70 Zygotes, triploid, 74