265 122 2MB
English Pages 1-160 [201] Year 2003
Current Topics in Developmental Biology Volume 58
This Page Intentionally Left Blank
Current Topics in Developmental Biology Volume 58 Edited by Gerald P. Schatten Director, PITTSBURGH DEVELOPMENT CENTER Deputy Director, Magee-Womens Research Institute Professor and Vice-Chair of Ob-Gyn-Reproductive Sci. & Cell Biol.-Physiology University of Pittsburgh School of Medicine Pittsburgh, PA 15213
Amsterdam Boston Heidelberg London New York Oxford Paris San Diego San Francisco Singapore Sydney Tokyo
Elsevier Academic Press 525 B Street, Suite 1900, San Diego, California 92101-4495, USA 84 Theobald s Road, London WC1X 8RR, UK This book is printed on acid-free paper. Copyright ß 2003, Elsevier 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. The appearance of the code at the bottom of the first page of a chapter in this book indicates the Publisher s consent that copies of the chapter may be made for personal or internal use of specific clients. This consent is given on the condition, however, that the copier pay the stated per copy fee through the Copyright Clearance Center, Inc. (www.copyright.com), for copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for creating new collective works, or for resale. Copy fees for pre-2002 chapters are as shown on the title pages. If no fee code appears on the title page, the copy fee is the same as for current chapters. 0070-2153/2003 $35.00 Permissions may be sought directly from Elsevier s Science & Technology Rights Department in Oxford, UK: phone: (þ44) 1865 843830, fax: (þ44) 1865 853333, e-mail: [email protected]. You may also complete your request on-line via the Elsevier homepage (http://elsevier.com), by selecting Customer Support and then Obtaining Permissions. For all information on all Academic Press publications visit our Web site at www.academicpress.com ISBN: 0-12-153158-9 PRINTED IN THE UNITED STATES OF AMERICA 03 04 05 06 07 08 9 8 7 6 5 4
3
2
1
Contents
Contributors Preface xi
ix
1 A Role for Endogenous Electric Fields in Wound Healing Richard Nuccitelli
I. II. III. IV.
Introduction 2 Endogenous Electric Fields Measured near Wounds 4 Epithelial Cells Exhibiting Galvanotaxis 7 How Do Motile Cells Sense and Respond to the Imposed Electric Field? 11 V. Can Imposed Electric Fields be used to Stimulate Wound Healing? 14 VI. Clinical Trials Using Electric Fields to Stimulate Wound Healing 19 VII. Summary 20 References 21
2 The Role of Mitotic Checkpoint in Maintaining Genomic Stability Song-Tao Liu, Jan M. van Deursen, and Tim J. Yen
I. II. III. IV. V.
Introduction 27 The Kinetochore is a Complex Structure for Cell Division 29 The Mitotic Checkpoint 31 Mutations in Mitotic Checkpoint Proteins and Tumorigenesis 41 Conclusions and Future Directions 45 Acknowledgments 48 References 48
v
vi
Contents
3 The Regulation of Oocyte Maturation Ekaterina Voronina and Gary M. Wessel
I. II. III. IV. V. VI.
Concept of Oocyte Maturation 54 Life and Death of an Oocyte 56 Maturation Promoting Factor 70 Initiation of Oocyte Maturation 73 Cytoplasmic Control of Maturation 84 Concluding Remarks 98 Acknowledgments 99 References 99
4 Stem Cells: A Promising Source of Pancreatic Islets for Transplantation in Type 1 Diabetes Cale N. Street, Ray V. Rajotte, and Gregory S. Korbutt
I. II. III. IV.
Introduction 112 Islet Transplantation: Success with Limitations 113 Alternative Sources of Transplantable Islets 114 Differentiation of Embryonic Stem Cells into Insulin-producing Cells 115 V. Identification of the Elusive Pancreatic Stem Cell 119 VI. Existence of Nonpancreatic Islet Progenitor Cells 126 VII. Summary 128 References 129
5 Differentiation Potential of Adipose Derived Adult Stem (ADAS) Cells Jeffrey M. Gimble and Farshid Guilak
I. II. III. IV. V.
Introduction 137 Adipose Tissue—Physiology 138 Adipose Tissue—Pathology 142 Adipose Derived Adult Stem (ADAS) 144 ADAS Cell Differentiation Potential and Potential Therapeutic Applications 146
vii
Contents
VI. Functional Tissue Engineering Considerations VII. Conclusions and Future Directions 153 Acknowledgments 153 References 154 Index 161 Contents of Previous Volumes
167
151
This Page Intentionally Left Blank
Contributors
Numbers in parantheses indicate the pages on which the authors’ contributions begin.
Jan M. van Deursen (27), Department of Pediatric and Adolescent Medicine, Mayo Clinic, Rochester, MN 55905, USA Jeffrey M. Gimble (137), Pennington Biomedical Research Center, Louisiana State University, 6400 Perkins Road, Baton Rouge, Louisiana 70808, USA Farshid Guilak (137), Department of Surgery, Division of Orthopedic Surgery, 375 MSRB, Box 3093, Duke University Medical Center, Durham, North Carolina 27710, USA Gregory S. Korbutt (111), Surgical-Medical Research Institute, Rm. 1074 Dentistry/Pharmacy Bldg., University of Alberta Edmonton, AB, Canada T6G 2N8; Department of Surgery, Rm. 1074 Dentistry/Pharmacy Bldg., University of Alberta Edmonton, AB, Canada T6G 2N8; Stem Cell Network of Canada, Rm. 1074 Dentistry/Pharmacy Bldg., University of Alberta Edmonton, AB, Canada T6G 2N8 Song-Tao Liu (27), Fox Chase Cancer Center, 7701 Burholme Avenue, Philadelphia, PA 19111, USA Richard Nuccitelli (1), RPN Research 144, Carroll St., New Britain, CT 06053, USA Ray V. Rajotte (111), Surgical-Medical Research Institute, Rm. 1074 Dentistry/Pharmacy Bldg., University of Alberta Edmonton, AB, Canada T6G 2N8; Department of Surgery, Rm. 1074 Dentistry/Pharmacy Bldg., University of Alberta Edmonton, AB, Canada T6G 2N8; Department of Medicine, Surgical Medical Research Institute, Rm. 1074 Dentistry/ Pharmacy Bldg., University of Alberta Edmonton, AB, Canada T6G 2N8 Cale N. Street (111), Surgical-Medical Research Institute, Rm. 1074 Dentistry/Pharmacy Bldg., University of Alberta Edmonton, AB, Canada T6G 2N8 Ekaterina Voronina (53), Department of Molecular and Cell Biology and Biochemistry, Brown University, 69 Brown St, Providence, RI 02912, USA
ix
x
Contributors
Gary M. Wessel (53), Department of Molecular and Cell Biology and Biochemistry, Brown University, 69 Brown St, Providence, RI 02912, USA Tim J. Yen (27), Fox Chase Cancer Center, 7701 Burholme Avenue, Philadelphia, PA 19111, USA
Preface CTDB Volume 58
This volume of Current Topics in Developmental Biology showcases some of the most exciting research being conducted in the enormously important fields of tissue repair, regenerative medicine and especially the cell cycle regulation and developmental plasticity of stem cells and their precursors – oocytes and embryos. As developmental biology shows us the details of embryogenesis and organogenesis, we are learning how to generate properly functioning cells and tissues in vitro and in vivo, techniques that will eventually provide clinicians with new knowledge about devastating injuries, illnesses, and developmental diseases. A Role for Endogenous Electric Fields in Wound Healing by Richard Nuccitelli of RPN Research describes electric fields in the epithelia of vertebrates, which is measurable as a ‘‘leakage current’’ when epithelial layers are disrupted, as happens in wounds. This acts as a signal to epidermal epithelial cells, which migrate to the wound and initiate repair. As described in detail here, epithelial cells respond similarly to an exogenous electric field, with important clinical implications for the acceleration of wound repair. The Role of Mitotic Checkpoint in Maintaining Genomic Stability by Song-Tao Liu and Tim Yen of Fox Chase Cancer Center and Jan van Deursen of the Mayo Clinic examines the kinetochore and its role as a molecular tugboat, helping to guide chromosomes and other structures to their proper places – a signal pathway essential for normal cell division, and essential information for researchers working on cancer and developmental issues regarding aneuploidy. The Regulation of Oocyte Maturation by Ekaterina Voronina and Gary Wessel of Brown University considers the molecular signals that drive the final stage in the development of ‘‘fertilization competence,’’ the cascade of chemical signals that enable an oocyte to accept a sperm, become activated, and develop into an embryo. Surprisingly, although development is similar among disparate species, the oocyte maturation signaling network is widely disparate among even closely-related species, which raises questions about divergent evolution and selection pressure, and highlights the risks in extrapolating milestones in development from one species to another.
xi
xii
Preface
Stem Cells: A Promising Source of Pancreatic Islets for Transplantation in Type 1 Diabetes by Cale Street, Ray Rajotte and Gregory Korbutt of the University of Alberta, offers hope that transplantation of insulin-dependent diabetics might soon gain independence from daily injections with the autologous transplant of new pancreatic islets derived from adult stem cells. While efforts to coax embryonic stem cells to differentiate into pancreatic islets have proven inefficient, there is increasing evidence both that the pancreas contains its own precursor cells, and that adult stem cells can be coaxed to differentiate into pancreatic islets. Differentiation Potential of Adipose Derived Adult Stem (ADAS) Cells by Jeffrey Gimble and Farshid Guilak of Louisiana State University comprehensively examines the potential clinical applications of adult stem cells derived from adipose tissue, based on what is known about how adipose develops and differentiates. Researchers and clinicians will no doubt find this work a seminal contribution to the new field of regenerative medicine. Together with other volumes in this Series, this volume provides a comprehensive survey of major issues at the forefront of modern developmental biology and developmental medicine. These chapters should be valuable to clinical and fundamental researchers in the fields of development biology and developmental medicine, as well as students and other professionals who want an introduction to current topics in cellular and molecular approaches to developmental biology and clinical problems of aberrant development. This volume in particular will be essential reading for anyone interested in wound repair, signaling and genetic stability, the development of oocytes, and stem cells. This volume has benefited from the ongoing cooperation of a team of participants who are jointly responsible for the content and quality of its material. The authors deserve the full credit for their success in covering their subjects in depth yet with clarity, and for challenging the reader to think about these topics in new ways. The members of the Editorial Board are thanked for their suggestions of topics and authors. I also thank Laura Hewitson and Leah Kauffman for their fabulous scientific insights and Rhonda Genes for her exemplary administrative support. Finally, we are grateful to everyone at the Pittsburgh Development Center of MageeWomens Research Institute here at the University of Pittsburgh School of Medicine for providing intellectual and infrastructural support for Current Topics in Developmental Biology. Jerry Schatten Pittsburgh Development Center Pennsylvania
1 A Role for Endogenous Electric Fields in Wound Healing Richard Nuccitelli* RPN Research 144, Carroll St., New Britain, CT 06053, USA
I. Introduction II. Endogenous Electric Fields Measured Near Wounds III. Epithelial Cells Exhibiting Galvanotaxis A. Embryonic Cells B. Epidermal Cells C. Corneal Epithelial Cells D. Lens Epithelial Cells IV. V. VI. VII.
How Do Motile Cells Sense and Respond to the Imposed Electric Field? Can Imposed Electric Fields be used to Stimulate Wound Healing? Clinical Trials Using Electric Fields to Stimulate Wound Healing Summary References
This review focuses on the experimental evidence supporting a role for endogenous electric fields in wound healing in vertebrates. Most wounds involve the disruption of epithelial layers composing the epidermis or surrounding organs in the body. These epithelia generate a steady voltage across themselves that will drive an injury current out of the wounded region, generating a lateral electric field that has been measured in four different cases to be 40–200 mV/mm. Many epithelial cells, including human keratinocytes, have the ability to detect electric fields of this magnitude and respond with directed migration. Their response typically requires Ca2 þ influx, the presence of specific growth factors and intracellular kinase activity. Protein kinase C is required by neural crest cells and cAMP-dependent protein kinase is used in keratinocytes while mitogen-activated protein kinase is required by corneal epithelial cells. Several recent experiments support a role for electric fields in the stimulation of wound healing in the developing frog neurula, adult newt skin and adult mammalian cornea. Some experiments indicate that when the electric field is removed the wound healing rate is 25% slower. In addition, nearly every clinical trial using electric fields to stimulate healing in mammalian wounds reports a significant increase in the rate of healing from 13 to 50%. However, these trials have utilized many different field strengths and polarities, so much work is needed to optimize this approach for the treatment of mammalian wounds.
*To whom correspondence should be addressed. E-mail: [email protected]. Current Topics in Developmental Biology, Vol. 58 Copyright ß 2003, Elsevier, Inc. All rights reserved. 0070-2153/03 $35.00
1
2
Nuccitelli
I. Introduction Most wounds involve the disruption of the epithelial layers composing the epidermis or surrounding organs in the body. They pose a serious threat to the well being of the organism by allowing both the invasion of microorganisms and the leakage of internal body fluids. Therefore, wound healing is one of the most important regenerative processes that most organisms exhibit. It usually involves the migration and proliferation of epidermal cells to reseal the epidermal layer. A recent review described the process: ‘‘The initial inflammatory response leads to the influx of macrophages and neutrophils, which release cytokines, growth factors, and nitric oxide, and induce nearby keratinocytes to migrate across the wounded epithelium’’ (Hackam and Ford, 2002). While cytokines and growth factors can indeed stimulate keratinocyte migration, a much earlier stimulus of directed migration that is triggered by wounding is not even mentioned by those authors. I am referring to the electrical field generated by the flow of current out of the wound. As discussed in detail below, this wound-field is generated by the transepithelial potential driving current out of the low resistance pathway at the wound and comes into being immediately upon wounding. This wound field-stimulated migration is known as galvanotaxis and has been studied extensively in human keratinocytes and mammalian corneal epithelial cells. Here, I will review the evidence for the involvement of endogenous electric fields in vertebrate wound healing. There are earlier reviews of the role of electric fields in regeneration and wound healing that may be of interest to the reader (Borgens, 1982; Borgens et al., 1989; Cho, 2002; Jaffe and Vanable, 1984; McCaig and Zhao, 1997; McCaig et al., 2002; Nuccitelli, 1984, 1988; Robinson, 1985; Vanable, 1989) and due to the availability of this earlier work, I will concentrate mainly on more recent studies conducted for the most part during the past 10 years. Most biologists learn early in their studies that all cells generate a voltage across their plasma membranes called the membrane potential. This voltage is used in many different ways, including aiding the transport of molecules across the membrane through coupled transport, rapid signaling to the entire cell that an event has occurred somewhere on its surface, and for relaying signals long distances as in the case of neurons. When we shift our attention from the single cell to the multicellular domain, we find a striking parallel. Just as the cell is surrounded by a plasma membrane, all of our organs are bound by an outer epithelium and indeed the largest organ in our body, our skin, is a multi-layered epithelium. However, in contrast to the extensive coverage given to the electrical properties of the plasma membrane, the electrical properties of epithelia are hardly ever mentioned
1. A Role for Endogenous Electric Fields in Wound Healing
3
in textbooks. Yet these properties are very important for organ function. The epithelium is to the organ as the plasma membrane is to the cell. Just as the plasma membrane forms a boundary that controls what goes in and out of the cell, the bounding epithelium determines what goes in and out of the organ that it encapsulates (with additional contributions from the circulatory system). The cells composing each epithelial layer are coupled with gap junctions so that they can be thought of electrically, to be one continuous entity, much as we think of the plasma membrane. Analogous to the plasma membrane, all epithelia generate a voltage across this epithelial layer that has many different uses. The polarity of this voltage is usually inside positive, which is opposite in sign to the plasma membrane potential that is generally inside negative. How is this transepithelial potential (TEP) generated? The TEP is due for the most part to the polarized distribution of ion channels in the epithelial cells. Most Na þ channels are localized to the apical membrane and most K þ channels are found in the basolateral membranes along with the Na þ /K þ -ATPase. Since the Na þ /K þ -ATPase maintains high internal K þ and low internal Na þ , this localization of channels leads to Na þ influx across the apical membrane and K þ efflux across the basolateral membrane. This flow of positive charge into the apical end and out of the basal end of the epithelial cells constitutes a transepithelial ion flow that must complete the current loop by flowing
Figure 1 Diagram of a typical epithelial cell in a monolayer with Na þ channels localized on the apical plasma membrane and K þ channels localized on the basolateral membranes along with the Na þ /K þ -ATPase. This asymmetric distribution of ion channels generates a transcellular flow of positive current that must flow back between the cells through the paracellular pathway (Ipara). This current flow generates a transepithelial potential that is positive on the basolateral side of the monolayer.
4
Nuccitelli
back through a paracellular route to the apical side of the epithelium (Fig. 1). The transepithelial potential will be proportional to the resistance of this paracellular pathway but typical values for this TEP range from 15 to 60 mV, inside positive in our bodies. It is this TEP that is the driving force for most endogenous electric fields in embryos and adults. This voltage across epithelia will drive current out of regions of low resistance where there has been a break in the epithelium (wounds) or where tight junction resistance is low, such as along the primitive streak (Jaffe and Stern, 1979; Winkel and Nuccitelli, 1989) or the posterior intestinal portal (Hotary and Robinson, 1990) in chick or mouse embryos or at the forming limb bud in amphibian, chick and mouse embryos (Altizer et al., 2001; Borgens et al., 1983; Robinson, 1983). This ‘‘leakage current’’ will in turn generate a lateral electric field along its path that will be proportional to the resistivity in that region. This electric field results from Ohm’s Law in a conductive medium, E ¼ J, where J is current density and is the local resistivity. Typical values for such fields will be discussed next.
II. Endogenous Electric Fields Measured Near Wounds The earliest measurements of the electrical phenomena associated with wounds did not measure the electric field itself, but rather the current flowing out of the wound. DuBois-Reymond (1843) used a unique galvanometer that he built with more than two miles of wire and measured about 1 A flowing out of a cut in one of his fingers. This was confirmed in 1849 and 1910 by other investigators and the history of these measurements is presented in a scholarly review by Vanable (1991). More modern techniques have also been used to study this wound current. The ‘‘leakage current’’ that is driven out of epithelia in low resistance regions has been measured using the vibrating probe technique (Jaffe and Nuccitelli, 1974) in several systems. One of the earliest such measurements was a current as large as 100 A/cm2 leaving the stumps of regenerating newt limbs (Borgens et al., 1977). Similar measurements have also been made on fingertip amputation currents in humans (Illingworth and Barker, 1980) where up to 30 A/cm2 was detected leaving the accidentally amputated stump for about three weeks. These currents will certainly generate electric fields just beneath the epidermis that will be proportional to the resistivity encountered in the tissue. The range of human tissue resistivity spans 200–1000 ohm-cm (Faes et al., 1999) so these currents would be expected to generate an electric field within the tissue of about 10–100 mV/mm.
5
1. A Role for Endogenous Electric Fields in Wound Healing Table I Endogenous Electric Fields Measured Near Wounds Wound type
E field (mV/mm)
Species
Tissue
Bovine
Cornea
Cut
42
Notophthalmus viridescens (newt)
Digit
Digit tip amputation
40
Notophthalmus viridescens (newt) Guinea pig
Limb stump Skin
Limb amputation Small cut
7–50 100–200
Reference Chiang et al. (1992); Sta Iglesia and Vanable (1998) Chiang et al. (1989); McGinnis and Vanable (1986); Sta Iglesia et al. (1996) McGinnis and Vanable (1986) Barker et al. (1982)
However, since this tissue resistivity can vary substantially as a function of cell density and tissue anatomy, it is always more reliable to measure these fields directly in the tissue rather than estimating them based on the transembryonic current density. This has been accomplished in four different wound types in skin and cornea. The classic approach to these measurements is to use KCl-filled glass microelectrodes to penetrate the outer epithelium and measure the voltage just beneath it in several positions along a line leading away from the wound. However, for skin measurements, another approach is more common. That is to measure the potential gradient just beneath the stratum corneum on the surface of the epidermis either with surface electrodes or by other means. The field generated by the current flowing between the upper surface of the epidermis and the stratum corneum is often larger than that generated below the epidermis due to the higher resistivity of that upper region. The range of field strengths measured in the four cases in the literature is surprisingly small, between 40 and 200 mV/mm (Table I). The field direction is a function of position. Beneath the epidermis the field polarity has the negative pole at the wound center and above the epidermis the woundcurrent is flowing in the opposite direction so that the positive pole is at the wound (Fig. 2). These wound fields have some useful properties for signaling. First, they appear immediately upon wounding since the TEP is continuously present to drive current out of any low-resistance region as soon as it is formed. Second, the lateral electric field illustrated in Fig. 2 that is generated by the wound current will persist until the resistance increases as the wound heals. Thus we have a signal that is immediate and persistent. These are ideal properties for a physiological signal to stimulate wound healing. If the epithelial cells forming the epidermis were able to detect such electric fields, they would be able to initiate wound healing
6
Nuccitelli
A
B
Figure 2 Generation of skin wound electric fields. (A) Unbroken skin maintains a ‘‘skin battery’’ or transepithelial potential, generated by the apical influx of Na þ and basolateral efflux of K þ . (B) When wounded, the potential drives current flow through the newly formed low resistance pathway generating an electric field whose negative vector points toward the wound center at the lower portion of the epidermis and away from the wound on the upper portion below the stratum corneum.
1. A Role for Endogenous Electric Fields in Wound Healing
7
immediately upon wounding. This is in fact the case, as I will discuss in the next section.
III. Epithelial Cells Exhibiting Galvanotaxis The phenomenon of galvanotaxis was first observed over 100 years ago in leukocytes (Dineur, 1891). Since then more than 15 cell types have been found to exhibit this ability to detect external electric fields (Table II) and to migrate along field lines (Nuccitelli, 1988; Robinson, 1985) but not all cells exhibit this behavior (Grahn et al., 2003; Sillman et al., 2003). It is not clear why some of those cells such as amoebae and slime molds have developed this capability since they may not encounter electric fields in their native environment. However, epithelial cells will all be exposed to electric fields generated by leakage currents driven by the TEP and it is not at all surprising that these cells have evolved the ability to sense and respond to small electric fields.
A. Embryonic Cells Embryonic cells that are known to migrate long distances within embryos were among the first to be found to exhibit galvanotaxis. Neural crest cells from both avians and amphibians (Cooper and Keller, 1984; Nuccitelli and Erickson, 1983; Stump and Robinson, 1983b) as well as fibroblasts from avians, amphibians and mammals (Erickson and Nuccitelli, 1984; Luther et al., 1983; Yang et al., 1984) respond to fields as low as 10 mV/mm by exhibiting enhanced migratory activity in the direction of the negative pole. As the field is increased, the degree of directed migration increases, reaching a maximal directed orientation of 0.8 on a scale of 0 to 1 at 100 mV/mm. At this field strength, approximately 98% of the cells actively migrate toward the negative pole.
B. Epidermal Cells Epidermal cells from amphibians, fish, and human skin also respond very strongly to physiological electric fields. Fish scale keratocytes are very fast movers and migrate toward the cathode in an imposed field (Cooper and Schliwa, 1986a,b). Neither the spontaneous locomotion nor the electrically guided motility were found to be microtubule dependent. The lamellipodial extension and locomotion of keratocytes are reversibly inhibited by a variety
8
Nuccitelli
of calcium channel antagonists, whereas their motility is unaffected by hyperpolarizing and depolarizing (low and high K þ ) media. Rivkah Isseroff’s group at UC Davis has collaborated with me to study primary human keratinocytes in culture. These cells migrate well on glass coverslips coated with collagen and their motility is highly sensitive to imposed dc electric fields. In fields as low as 10 mV/mm, more cells migrate toward the cathode than toward the anode (Fig. 3). The optimal field strength for this response is 100 mV/mm at which 98% of the cells migrate toward the cathode, exhibiting an average cosine of 0.8 on a scale of 0 to 1. They exhibit a response half-time of about 10 min and their directed response requires Ca2 þ influx (Fang et al., 1998; Trollinger et al., 2002) and epidermal growth factor (EGF) receptor phosphorylation (Fang et al., 1999). Inhibitor studies implicate a role for PKA in galvanotaxis, but not PKC or myosin light chain kinase (Pullar et al., 2001). The presence of a growth factor (EFG) is required for galvanotaxis in these cells, as well as in the corneal epithelial cell system described next.
C. Corneal Epithelial Cells Corneal epithelial cells from bovine, rabbit, and human sources have been extensively investigated (Table II). Many of these studies have been conducted in vitro and a few were done in situ. In every case, these cells migrate toward the negative pole of the imposed field and responded to fields on the order of 100 mV/mm. There has not been much work using lower field strengths so the threshold field is not accurately known. However, the endogenous field near the bovine cornea wound of 42 mV/mm (Chiang et al., 1992) would suggest that these corneal epithelial cells should be able to respond to fields of this magnitude. Some very interesting results have emerged recently from in situ studies on the rat cornea system that will be discussed below.
D. Lens Epithelial Cells McCaig’s group has also studied the galvanotaxis of lens epithelial cells from central and peripheral regions (Wang et al., 2003a,b). Field-directed cell migration required serum, or growth factors. Cells cultured in serumfree medium are blinded to the electric field, but their ability to respond was restored partially by the addition of basic fibroblast growth factor. The direction of cell migration depended on both the field strength and the origin of the lens epithelial cells. Both central and peripheral lens epithelial cells moved anodally at 150–250 mV/mm, but surprisingly the peripheral cells
1. A Role for Endogenous Electric Fields in Wound Healing
9
Figure 3 galvanotaxis of human keratinocytes. Every point represents the position of a single cell after 2 h in the indicated field strength after starting at the origin. Top: No field control exhibits average cosine ’ ¼ 0.07. Middle: 10 mV/mm exhibits significant migration toward the negative pole. Bottom: 100 mV/mm stimulates significant galvanotaxis in 98% of cells. (Reproduced with permission of Nishimura et al., 1996.)
10
Nuccitelli
Table II Vertebrate Cells Exhibiting Galvanotaxis
Cell type
Response direction/ threshold (mV/mm)
Neural crest cells Quail
Cathode/10
Xenopus Ambystoma Fibroblasts Quail somite
Cathode/10 Cathode/nda
Mouse C3H/10T1/2 Mouse NIH 3T3 and SV101 Cornea Rat epithelial Rabbit epithelial Rabbit endothelial Rabbit stromal fibroblast Bovine Human Lens Bovine Retina Human pigment epithelial Vascular endothelium Bovine Human granulocyte Human leukocyte
Cathode/nda Cathode/nda
Human macrophage Rabbit osteoclasts Rabbit osteoblasts Bovine chondrocyte Rat prostate cancer cell line. MAT-LyLu Epidermal cells Xenopus embryo Fish scale Human skin a
Cathode/10
Reference Nuccitelli and Erickson (1983) Nuccitelli et al. (1993) Stump and Robinson (1983a) Cooper and Keller (1984) Nuccitelli and Erickson (1983); Erickson and Nuccitelli (1984) Yang et al. (1984) Brown and Loew (1994)
Cathode/nda Cathode/400 Anode/200 Anode/600 Cathode/100 Cathode/100 Cathode/50 Anode/150–200 Cathode/600
Song et al. (2002a) Soong et al. (1990a) Chang et al. (1996) Soong et al. (1990b) Zhao et al. (1996) Farboud et al. (2000) Wang et al. (2003a)
Cathode/100
Li and Kolega (2002)
Anode/100 Both anode and cathode/nda Anode/nda Anode/nda Cathode/nda Cathode/80 Cathode/10
Rapp et al. (1988) Dineur (1891); Fukushima et al. (1953) Cho et al. (2002); Orida and Feldman (1982) Ferrier et al. (1986) Ferrier et al. (1986) Chao et al. (2000) Djamgoz et al. (2001)
Cathode/nda Cathode/50 Cathode/10
Luther et al. (1983) Cooper and Schliwa (1986b) Nishimura et al. (1996)
Sulik et al. (1992)
Not determined.
migrated in the opposite direction at 50 mV/mm. This bidirectional migration depending on field strength is not very common and could prove to be quite useful for studies of the mechanism used by these cells to detect the field. The only previous reports of a bi-directional response involved the polarization of algal eggs in which some batches exhibited the opposite
1. A Role for Endogenous Electric Fields in Wound Healing
11
polarization response to the same field (Peng and Jaffe, 1976) and some exhibited a bipolar response as a function of field strength. Neuronal growth can exhibit a bidirectional response to the same field if the substrate charge is modified (Rajnicek et al., 1998).
IV. How Do Motile Cells Sense and Respond to the Imposed Electric Field? This question has been the focus of most studies of galvanotaxis over the past few years. Two clues to the mechanism seem to be widely applicable: (1) Ca2 þ influx appears to be required for many of the cells and they are unable to detect the field when Ca2 þ influx is reduced by removing extracellular Ca2 þ or by adding Ca2 þ channel blockers (Cooper and Schliwa, 1986a; Fang et al., 1998; Nuccitelli and Smart, 1989, 1991; Trollinger et al., 2002); (2) Growth factors are required along with protein kinase activity. It was found earlier on that specific growth factors in the culture media such as EGF and FGF were absolutely required for galvanotaxis of human keratinocytes (Fang et al., 1998), corneal epithelial cells (Zhao et al., 1996), and lens epithelial cells (Wang et al., 2003b). Subsequent work indicated that EGF receptor phosphorylation was required, and inhibitors of EGF receptor phosphorylation would make human keratinocytes blind to the imposed electric field (Fig. 4) while reducing their migration rate very slightly. The phosphorylated receptor is rapidly localized to the cathodal side of these cells (Fang et al., 1999; Zhao et al., 2002), with significant asymmetry being detected within 5 min of field application. The two mechanisms that have been proposed to accomplish this localization are localized secretion and lateral electrophoresis (Jaffe, 1977), discussed in detail below. However, more experimental work is needed to determine the exact mechanism generating this receptor asymmetry. Receptor phosphorylation requires protein kinase activity, and several kinases are known to be involved (Table III). Protein kinase C was the first found to be needed for the galvanotaxis response in neural crest cells (Nuccitelli et al., 1993), but it does not appear to be required for the response of human keratinocytes (Pullar et al., 2001) where the activity of cAMP-dependent protein kinase (PKA) appears to be more critical. Corneal epithelial cells utilize the mitogen-activated protein (MAP) kinase, ERK1/2 (Zhao et al., 2002) as do lens epithelial cells (Wang et al., 2003b). In the corneal cells, a cathodal accumulation of lipids, EGF receptors, and ERK1/ 2 induces but does not depend on a cathodal redistribution of F-actin. Inhibitors of MAP kinase signaling inhibit the directed cell migration and cathodal asymmetry of ERK and F-actin. In lens epithelial cells Wang
12
Nuccitelli
Figure 4 Response of human keratinocytes to a 100 mV/mm electric field in the presence of various tyrosine kinase inhibitors. PD158780 is the most specific inhibitor of EGFR phosphorylation and completely blinds the cells to the field while only inhibiting their rate of movement by 40%. (Reproduced with permission of Fang et al., 1999.)
Table III Signal Transduction Molecules Involved in Galvanotaxis Species
Tissue
Required molecule
Reference
Quail Bovine Bovine
Neural crest Lens Cornea
PKC, PKA TGF-, FGF, ERK1/2, MAPK HGF, EGF, EGFR, MAPK
Human
Skin
EGFR, EGF, PKA, Ca2 þ
Nuccitelli et al. (1993) Wang et al. (2003a,b) McBain et al. (2003); Zhao et al. (1996, 2002) Fang et al. (1998, 1999); Pullar et al. (2001); Trollinger et al. (2002)
et al. observed an asymmetric activation of ERK, with much weaker activity in cathode-facing wounds than in those facing the anode. This correlates well with the wound healing response in which anode-facing wounds healed faster than cathode-facing ones (Wang et al., 2003b). Given these clues, there have been several speculations regarding the mechanism by which cells can sense the presence of external electrical fields. The most direct of these is via the redistribution of lipid or protein molecules in the plasma membrane (Jaffe, 1977). Electrical fields can exert force on charged molecules to move them directly by electrophoresis or indirectly via electroosmosis. Electrophoresis is the movement of a charged molecule in an electric field that is a function of its net charge and molecular weight, whereas electroosmosis refers to the fluid flow through the medium that results from the movement of the water of hydration that surround all charges. The bulk water movement of electroosmosis can actually ‘‘drag’’
1. A Role for Endogenous Electric Fields in Wound Healing
13
membrane lipids or proteins in the direction opposite to that in which electrophoresis would move them. Most membrane proteins and lipids have a net negative charge and would be electrophoresed towards the positive pole of an electric field. However, many cells exhibit the opposite protein redistribution toward the positive pole! How does this happen? Each of the negative charges on the membrane proteins and lipids must be counter-balanced by a positive charge in solution, and each of those charges is surrounded by waters of hydration. The external electric field will move the positive counter-charges towards the negative pole and the waters of hydration will move with the positive charges. This causes a bulk water movement toward the negative pole that can actually drag membrane proteins along with it, even if those proteins have a net negative charge. The strength of electroosmosis will be highly dependent on the net surface charge and reducing this charge by removing sugar groups, for example, can actually change the direction of movement of membrane proteins (McLaughlin and Poo, 1981). Once the redistribution of proteins and lipids occurs, asymmetry is established and the redistributed proteins can influence cell motility via the signaling cascades discussed above. The second speculation regarding the mechanism by which cells could sense the field involves localized changes in the membrane potential. All cells generate a voltage across their plasma membrane that is about 70 mV inside negative in animal cells and often much larger in plant cells. When placed in an electrical field, the voltage across the plasma membrane will be modified most in regions that are perpendicular to the field lines. The ends of the cell that face the two poles of the field will experience the largest effect. The voltage drop across the cell is determined by the field strength multiplied by the length of the cell along the field lines. For a 100 m-long cell in a 100 mV/mm field, this voltage drop would be 10 mV. Since current passing through a cell will encounter the bulk of the resistance at the plasma membrane and very little across the cytoplasm, approximately half of this voltage drop will occur across each membrane facing the poles. That means that the plasma membrane at the end of the cell facing the positive pole will have 5 mV more voltage across it and the membrane facing the negative pole will have 5 mV less across it. (The cytoplasm will also have a small voltage across it, typically in the microvolt range, due to its much lower resistivity.) So the question becomes, can this 7% change in the membrane potential in a localized region influence cell motility? Voltage-gated ion channels typically open upon a depolarization of about 10–20 mV, so this small change is not likely to trigger channel opening. However, it will certainly bias the force driving ions through any open channels. Positive ions will experience a larger force driving them into the cell at the membrane region facing the positive pole of the field and a slightly lower force driving
14
Nuccitelli
influx at the end of the cell facing the negative pole. This could contribute to the generation of an intracellular ion concentration asymmetry that might influence the direction of cell motility. The obvious candidate-ion to consider is Ca2 þ , since many cells require Ca2 þ influx for galvanotaxis. Ca2 þ will be driven into the cell by both the negative membrane potential and the concentration gradient since the extracellular [Ca2 þ ] is typically 10,000 times greater than the intracellular [Ca2 þ ]. The best way to determine the total driving force on Ca2 þ is to calculate its equilibrium potential from this concentration gradient. The equilibrium potential is the voltage at which as much Ca2 þ would be forced out per unit time by the voltage as would leak in per unit time due to the high concentration outside. It is given by the Nernst equation, and for most mammalian cells with [Ca2 þ ]o of 1.2 mM and [Ca2 þ ]i of 10 4 mM, this equilibrium potential would be 114 mV. Since most cells have a resting potential near 70 mV, Ca2 þ is 184 mV away from this equilibrium. A 5 mV change in the membrane potential will only increase the total driving force by 3.8% at the positive pole and reduce it by 3.8% at the negative pole. However, over time, the consequent asymmetry in flux could result in an intracellular ion concentration gradient that might influence cell motility.
V. Can Imposed Electric Fields be used to Stimulate Wound Healing? Now that we know that endogenous electric fields in the range of 40–200 mV/mm are naturally present near wounds and that skin cells respond to fields of this magnitude with directed motility, the possibility arises that the electric field may be playing a role in the stimulation of wound healing. There have only been a few well-controlled experiments designed to test this hypothesis to date, along with many more clinical attempts that have been less satisfactory. I will briefly review these here. Robinson’s group carried out the first controlled experiment designed to investigate the role of the electric field in wound healing, utilizing a very simple system, the neurula-stage frog embryo (Rajnicek et al., 1988; Stump and Robinson, 1986). Transected frog embryos will heal completely within 7 h in a pond water medium that begins with a rapid purse string-like contraction requiring microfilaments but not Na þ . This is followed by a slower phase that is blocked by either Na þ -free medium, or the addition of amiloride, benzamil, or ouabain, drugs that inhibit Na þ flux through the epithelium as indicated by a rapid reduction in the transepithelial potential. This indicates that the slow phase of wound healing requires the endogenous
1. A Role for Endogenous Electric Fields in Wound Healing
15
Figure 5 Diagram of experiment designed to modify electric field within newt digit during wound healing. (A) The newt was placed in a three-compartment chamber with only the feet immersed in pond water. (B) View of one current-supplying electrode and the two microelectrodes used for voltage measurements. (Reproduced with permission from Chiang et al., 1991.)
Na þ -carried electric current and certainly supports a role for the electric field in wound healing. The second well-controlled experiment to test the role of electric fields in wound healing was conducted on the newt by Vanable’s group (Chiang et al., 1991). They made a small skin wound in one hindlimb digit on both the right and left foot of Notophthalamus viridescens and monitored the healing rate while changing the lateral electric field near the wounds by passing current through one digit, across the body and out the contralateral digit (Fig. 5). The amount of current passed was adjusted so that the lateral field of one wound was zero while the contralateral wound had an enhanced field. They observed that the wounds with the enhanced field healed more rapidly than the wounds with the zero field. When digits on one side were treated with 30 M benzamil in an artificial pond water so that their wound fields were reduced to approximately zero, and the contralateral wounds were kept in artificial pond water without benzamil so that they had normal wound fields, there was significantly less epithelization of the benzamil-treated wounds than of the control wounds. This effect on wound healing was reversed by adding currents that restored the normal wound fields, but not by adding currents that reversed the wound fields to the opposite polarity. When currents were added to reverse the wound fields on one side of the animal, leaving the contralateral wounds free of added currents, the wounds with the
16
Nuccitelli
Figure 6 Epithelialization after 4 h of healing vs lateral field. When the lateral electric field is reduced to 0 the rate of healing is reduced by about 25%. (Reproduced with permission of Chiang et al., 1991.)
reversed fields healed more poorly than the contralateral wounds with normal fields. These results are consistent with the hypothesis that the intrinsic lateral electric fields in the vicinity of wounds promote epithelization of these wounds. These experiments are the most elegant ones to date on this question and the overall conclusion is that in the absence of a lateral electric field, the rate of wound healing is reduced by about 25% (Fig. 6). Vanable’s group went on to ask if applying a larger electric field could increase the rate of wound healing. The endogenous lateral field near a wound is typically 40 mV/mm. Augmenting this to 80 and 100 mV/mm reduced the rate of healing so they concluded that the newt epithelialization rate was nearly maximal at the normal field strength (Sta Iglesia et al., 1996). Vanable’s group has also contributed the third well controlled experiment to determine the involvement of electric fields in wound healing (Sta Iglesia and Vanable, 1998). Here they used bovine corneal lesions with a 1.5 mm circular wound. A decrease in the field strength by submersion of the lesions or by treating the lesions with the Na þ -channel blocker, benzamil, significantly retarded healing. An increase in the field strength of lesions treated with Na þ -depleted Hanks’ solution, by the addition of direct current, increased epithelialization. Epithelialization was fastest in wounds with field strengths raised to 80 mV/mm, more than twice the normal field strength present in wounds maintained in Hanks’ solution alone.
1. A Role for Endogenous Electric Fields in Wound Healing
17
Figure 7 The transcorneal potential can be manipulated with different drugs and its effect on wound healing, cell division, and nerve sprout orientation determined. A. Diagram of corneal preparation indicating position of wound edge. B. Plot of transcorneal potential as a function of distance from the wound edge and the presence of the indicated drugs. (Reproduced with permission from Song et al., 2002) (see color plate).
Epithelization decreased, however, when the field strengths were increased to 120 mV/mm. A similar pattern was also observed when the field’s polarity was reversed. By manipulating and monitoring the field strengths, they demonstrated for the first time that increased wound field strengths enhance corneal wound epithelialization, and that field strengths with reversed polarity also enhance this epithelialization. The final well-controlled study was also conducted on a cornea preparation in situ. McCaig’s group used the rat cornea to study wound healing in response to a similar circular wound (Song et al., 2002). They manipulated the endogenous lateral electric field near the wound by using drugs with differing actions (Fig. 7). They also found that the rate of wound closure was highly sensitive to the field strength. In addition to influencing the rate of wound closure, the wound-induced field influenced the orientation of cell division. Most epithelial cells divided
18
Nuccitelli
Figure 8 Corneal epithelial wound healing rate is influenced by the electric field. The rate is greatest when PGE2 or aminophylline are present and these trigger the largest transcorneal potential as seen in Fig. 7. Ouabain reduces this potential and generates the lowest epithelial wound healing rate. (Reproduced with permission from Song et al., 2002.)
with a cleavage plane parallel to the wound edge and perpendicular to the field vector. Increasing or decreasing the field pharmacologically, respectively increased or decreased the extent of oriented cell division. In addition, cells closest to the wound edge, where the field was highest, were oriented most strongly by the field. The frequency of cell division was also enhanced by the endogenous electric field. Because the endogenous field also influenced the wound-healing rate (Fig. 8), it appears to be one force that can stimulate both cell migration and cell division during healing. One very important additional observation made by McCaig’s group was the effect of these endogenous fields on nearby nerve growth (McCaig et al., 2002). The endogenous electric field near the wound has a very strong orienting effect on the direction of sensory nerve sprouting and growth. Between 16 and 20 h after wounding a large number of nerve sprouts project directly towards the cut wound edge in a whole-mount rat cornea (Fig. 9). Reducing the wound field with ouabain randomizes nerve fiber orientation, suggesting that the electric field is the main orienting influence for these nerve sprouts. It has been known for decades that nerve growth can be oriented by imposed electric fields (Jaffe and Poo, 1979; Sisken and Smith, 1975), but this is one of only a few well documented examples in which a naturally occurring electric field has been found to exert a strong influence over neuronal growth. In addition to nerve growth, the axis of cell division was also strongly influenced by this endogenous electric field. Treatments that enhanced the field resulted in a greater degree of mitotic spindle alignment perpendicular to the field lines (Fig. 9).
1. A Role for Endogenous Electric Fields in Wound Healing
19
Figure 9 Observations made near a wound in rat cornea. (A) Neuron outgrowths are strongly aligned by the endogenous electric field and exhibit a more random orientation when the field is reduced by ouabain addition. (B) The rate of corneal wound healing is reduced in the presence of ouabain. (C) The frequency and orientation of division planes is influenced by the field strength. Aminophylline increases the transcorneal potential and stimulates an increase in the rate of cell division and orientation of the axis of division perpendicular to the field. (Reproduced with permission from McCaig et al., 2002) (see color plate).
VI. Clinical Trials Using Electric Fields to Stimulate Wound Healing While these well-controlled experiments are fairly recent, there is a large number of earlier clinical attempts to improve wound healing with electrical stimulation. There is much interest in finding improved methods for wound management and in the past five years alone, there are at least eight reviews of the clinical trials using electrical stimulation (Akai and Hayashi, 2002; Bogie et al., 2000; Braddock et al., 1999; Cullum et al., 2001; Devine, 1998; Evans et al., 2001; Fleischli and Laughlin, 1997; Lampe, 1998). All of these reviews conclude that electro-therapy usually improves the rate of wound healing by 13–50% and sometimes stimulates the healing of chronic wounds for which no other conventional therapy had been successful. The most extensive review of the clinical literature remains Vanable’s book chapter (1989) in which he describes in detail the three negative reports in the literature (Carey and Lepley, 1962; Steckel et al., 1984; Wu et al., 1967)
20
Nuccitelli
followed by 11 positive reports grouped by methodology. Some investigators placed the cathode in the wound for the duration of the experiment (Assimacopoulos, 1968a,b; Konikoff, 1976), some used the anode at the wound throughout (Alvarez et al., 1983) and others alternate polarity with the cathode in the wound first followed by the anode (Carley and Wainapel 1985; Gault and Gatens, 1976; Page and Gault 1975; Rowley et al., 1974a,b; Wheeler et al., 1971; Wolcott et al., 1969). The three negative reports are not convincing. Carey and Lepley (1962) studied a small number of rabbit back skin wounds using very high current densities of around 400 A/cm2 and observed much necrosis at the anode. Wu et al. (1967) ignored skin healing and Steckel et al. (1984) studied horse skin wounds in which those with electrodes had purulent inflammation. The most popular approach for stimulating wound healing in humans uses both polarities of imposed fields sequentially, with the cathode in the wound first, followed by the anode. The rationale for using the cathode first is that this renders the wound free from infection (Rowley, 1972; Rowley et al., 1974a). There are reports of significant stimulation of healing in over 300 ulcers with this approach, but the control group was always a small fraction of the size of the experimental group. Nevertheless, these studies indicate that electric fields promote the healing of chronic wounds. It is also clear that much work is needed to determine the optimal protocol for electric field application since so many different treatments have been used by these investigators. This will require further studies on model systems such as the mammalian cornea and skin as well as much more work applying fields to human skin wounds.
VII. Summary 1. The outer layers of our skin compose a polarized, multilayered epithelium that generates a transepithelial potential of 20–50 mV across itself. Wounds in this organ generate a low-resistance pathway through which current will flow. This flow of current from all regions around the wound generates a lateral electric field that points toward the wound from every direction around it. The magnitude of this lateral field ranges between 40 and 200 mV/mm in mammalian wounds but there are still no reliable measurements of the lateral electric field near human skin wounds. 2. These wound fields have useful properties for signaling because they appear immediately upon wounding and persist until the wound heals. 3. Many cells have the ability to detect electric fields of this magnitude, and 13 of these are discussed here. Most of these cells migrate toward the
1. A Role for Endogenous Electric Fields in Wound Healing
4.
5.
6.
7.
21
cathode of an imposed electric field, with the optimal response occurring in fields on the order of 100 mV/mm. The mechanisms used by these cells to sense the electric field often require Ca2 þ influx, the presence of growth factors, and protein kinase activity. PKC is required by neural crest cells and PKA is used in keratinocytes, while MAP kinase is required by corneal epithelial cells. The immediate target of these fields is likely to be the plasma membrane. Due to its presence at the outer boundary of the cell and its large resistance, it will have the most interaction with the electric field. The field could act to redistribute charged lipid and protein molecules within the plasma membrane, or modify the membrane potential at the ends of the cell facing the poles of the electric field. Several well-controlled experiments support a role for electric fields in the stimulation of wound healing in frog neurula, newt skin, and mammalian cornea. The endogenous field near a wound in the rat cornea influences the rate of wound healing, the orientation of cell division, and the orientation of nerve sprouting. Several clinical trials have reported that the healing of mammalian wounds (including humans) can be promoted by electric fields. However, many different field strengths and polarities have been used, so much work is needed to optimize this approach for the treatment of mammalian wounds. Most importantly, the lateral electric field near human skin wounds must be measured to guide the informed design of fields to stimulate wound healing.
References Akai, M., and Hayashi, K. (2002). Effect of electrical stimulation on musculoskeletal systems; a meta-analysis of controlled clinical trials. Bioelectromagnetics Feb., 23(2), 132–43. Altizer, A. M., Moriarty, L. J., Bell, S. M., Schreiner, C. M., Scott, W. J., and Borgens, R. B. (2001). Endogenous electric current is associated with normal development of the vertebrate limb. Dev. Dyn. 221, 391–401. Alvarez, O. M., Mertz, P. M., Smerbeck, R. V., and Eaglstein, W. H. (1983). The healing of superficial skin wounds is stimulated by external electrical current. J. Invest Dermatol. 81, 144–148. Assimacopoulos, D. (1968a). Low intensity negative electric current in the treatment of ulcers of the leg due to chronic venous insufficiency. Preliminary report of three cases. Am. J. Surg. 115, 683–687. Assimacopoulos, D. (1968b). Wound healing promotion by the use of negative electric current. Am. Surg. 34, 423–431. Barker, A. T., Jaffe, L. F., and Vanable, J. W., Jr. (1982). The glabrous epidermis of cavies contains a powerful battery. Am. J. Physiol. 242, R358–R366. Bogie, K. M., Reger, S. I., Levine, S. P., and Sahgal, V. (2000). Electrical stimulation for pressure sore prevention and wound healing. Assist. Technol. 12.(1), 50–66.
22
Nuccitelli
Borgens, R. B. (1982). What is the role of naturally produced electric current in vertebrate regeneration and healing?. Int. Rev. Cytol. 76, 245–298. Borgens, R. B., Robinson, K. R., Vanable, J. W., Jr., McGinnis, M. E., and McCaig, C. D. Electric Fields in Vertebrate Repair. New York: Alan R. Liss, Inc. Borgens, R. B., Rouleau, M. F., and DeLanney, L. E. (1983). A steady efflux of ionic current predicts hind limb development. J. Exp. Zool. 228, 491–503. Borgens, R. B., Vanable, J. W., Jr., and Jaffe, L. F. (1977). Bioelectricity and regeneration: Large currents leave the stumps of regenerating newt limbs. Proc. Natl. Acad. Sci. USA 74, 4528–4532. Braddock, M., Campbell, C. J., and Zuder, D. (1999). Current therapies for wound healing: electrical stimulation, biological therapeutics, and the potential for gene therapy. Int. J. Dermatol. 38, 808–817. Brown, M. J. and Loew, L. M. (1994). Electric field-directed fibroblast locomotion involves cell surface molecular reorganization and is calcium independent. J. Cell Biol. 127, 117–128. Carey, L. C., Jr. and Lepley, D. Jr. (1962). Effect of continuous direct current on healing wounds. Surg. Forum 13, 33–35. Carley, P. J. and Wainapel, S. F. (1985). Electrotherapy for acceleration of wound healing: low intensity direct current. Arch. Phys. Med. Rehabil. 66, 795–801. Chang, P. C., Sulik, G. I., Soong, H. K., and Parkinson, W. C. (1996). Galvanotropic and galvanotaxic responses of corneal endothelial cells. J. Formos. Med. Assoc. 95, 623–627. Chao, P. H., Roy, R., Mauck, R. L., Liu, W., Valhmu, W. B., and Hung, C. T. (2000). Chondrocyte translocation response to direct current electric fields. J. Biomech. Eng. 122, 261–267. Chiang, M., Cragoe, E. J., Jr., and Vanable, J. W., Jr. (1989). Electrical fields in the vicinity of small wounds in Notophthalmus viridescens skin. Biol. Bull. 176(S), 179–183. Chiang, M., Cragoe, E. J., Jr., and Vanable, J. W. (1991). Intrinsic electric fields promote epithelialization of wounds in the newt, Notophthalmus viridescens. Dev. Biol. 146, 377–385. Chiang, M., Robinson, K. R., and Vanable, J. W., Jr. (1992). Electrical fields in the vicinity of epithelial wounds in the isolated bovine eye. Exp. Eye. Res. 54, 999–1003. Cho, M. R. (2002). A review of electrocoupling mechanisms mediating facilitated wound healing. IEEE Transactions on Plasma Science 30, 1504–1515. Cho, M. R., Marler, J. P., Thatte, H. S., and Golan, D.E. (2002). Control of calcium entry in human fibroblasts by frequency-dependent electrical stimulation. Front Biosci. a1–a8. Cooper, M. S. and Keller, R. E. (1984). Perpendicular orientation and directional migration of amphibian neural crest cells in DC electrical fields. Proc. Natl. Acad. Sci. USA 81, 160–164. Cooper, M. S. and Schliwa, M. (1986a). Transmembrane Ca2 þ fluxes in the forward and reversed galvanotaxis of fish epidermal cells. Prog. Clin. Biol. Res. 210, 311–318. Cooper, M. S. and Schliwa, M. (1986b). Motility of cultured fish epidermal cells in the presence and absence of direct current electric fields. J. Cell Biol. 102, 1384–1399. Cullum, N., Nelson, E. A., Flemming, K., and Sheldon, T. (2001). Systematic reviews of wound care management: (5) beds; (6) compression; (7) laser therapy, therapeutic ultrasound, electrotherapy and electromagnetic therapy. Health Technol. Assess. 5, 1–221. Devine, P. (1998). Electric stimulation and wound healing. J. Wound. Ostomy. Continence. Nurs 25, 291–295. Dineur, E. (1891). Note sur la sensibilite des leucocytes a l’electricite. Bulletin Seances Soc. Belge Microscopie (Bruxelles) 18, 113–118. Djamgoz, M. B. A., Mycielska, M., Madeja, Z., Fraser, S. P., and Korohoda, W. (2001). Directional movement of rat prostate cancer cells in direct-current electric field: involvement of voltage-gated Na þ channel activity. J. Cell Sci. 114, 2697–2705. DuBois-Reymond, E. (1843). Vorlaufiger abrifs einer untersuchung uber den sogenannten froschstrom und die electomotorischen fische. Ann. Phys. U. Chem. 58, 1.
1. A Role for Endogenous Electric Fields in Wound Healing
23
Erickson, C. A. and Nuccitelli, R. (1984). Embryonic fibroblast motility and orientation can be influenced by physiological electric fields. J. Cell Biol. 98, 296–307. Evans, R. D., Foltz, D., and Foltz, K. (2001). Electrical stimulation with bone and wound healing. Clin. Podiatr. Med. Surg. 18, 79–95. Faes, T. J., van der Meij, H. A., de Munck, J. C., and Heethaar, R. M. (1999). The electric resistivity of human tissues (100 Hz–10 MHz): a meta-analysis of review studies. Physiol. Meas. 20, R1–10. Fang, K. S., Farboud, B., Nuccitelli, R., and Isseroff, R. R. (1998). Migration of human keratinocytes in electric fields requires growth factors and extracellular calcium. J. Invest. Dermatol. 111, 751–756. Fang, K. S., Ionides, E., Oster, G., Nuccitelli, R., and Isseroff, R. R. (1999). Epidermal growth factor receptor relocalization and kinase activity are necessary for directional migration of keratinocytes in DC electric fields. J. Cell Sci. 112(Pt 12), 1967–1978. Farboud, B., Nuccitelli, R., Schwab, I. R., and Isseroff, R. R. (2000). DC electric fields induce rapid directional migration in cultured human corneal epithelial cells. Exp. Eye Res. 70, 667–673. Ferrier, J., Ross, S. M., Kenehisa, J., and Aubon, J. E. (1986). Osteoclasts and osteoblasts migrate in opposite direction in response to a constant electrical field. J. Cell Physiol. 129, 283–288. Fleischli, J. G. and Laughlin, T. J. (1997). Electrical stimulation in wound healing. J. Foot Ankle Surg. 36, 457–461. Fukushima, K., Senda, N., Inui, H., Miura, H., Tamai, Y., and Murakami, Y. (1953). Studies on galvanotaxis of leukocytes. I. Galvanotaxis of human neutrophilic leukocytes and methods of measurement. Med. J. Osaka Univ. 4, 195–208. Gault, W. R., Jr. and Gatens, P. F. (1976). Use of low intensity direct current in management of ischemic skin ulcers. Phys. Ther. 56, 265–269. Grahn, J. C., Reilly, D. A., Nuccitelli, R. L., and Isseroff, R. R. (2003). Melanocytes do not migrate directionally in physiological DC electric fields. Wound Repair Regen. 11, 64–70. Hackam, D. J., and Ford, H. R. (2002). Cellular, biochemical, and clinical aspects of wound healing. Surg. Infect. 3 Suppl 1, S23–S35. Hotary, K. B. and Robinson, K. R. (1990). Endogenous electrical currents and the resultant voltage gradients in the chick embryo. Dev. Biol. 140, 149–160. Illingworth, C. M. and Barker, A. T. (1980). Measurement of electrical currents emerging during the regeneration of amputated fingertips in children. Clin. Phys. Physiol. Meas. 1, 87–89. Jaffe, L. F. (1977). Electrophoresis along cell membranes. Nature (London) 265, 600–602. Jaffe, L. F. and Nuccitelli, R. (1974). An ultrasensitive vibrating probe for measuring extracellular currents. J. Cell Biol. 63, 614–628. Jaffe, L. F. and Poo, M.-M. (1979). Neurites grow faster towards the cathode than the anode in a steady field. J. Exp. Zool. 209, 115–128. Jaffe, L. F. and Stern, C. D. (1979). Strong electrical currents leave the primitive streak of chick embryos. Science (Wash., D.C.) 206, 569–571. Jaffe, L. F. and Vanable, J. W. (1984). Electric fields and wound healing. Clin. Dermatol. 2, 34–44. Konikoff, J. J. (1976). Electrical promotion of soft tissue repairs. Ann. Biomed. Eng 4, 1–5. Lampe, K. E. (1998). Electrotherapy in tissue repair. J. Hand Ther. 11, 131–139. Li, X. and Kolega, J. (2002). Effects of direct current electric fields on cell migration and actin filament distribution in bovine vascular endothelial cells. J. Vasc. Res. 39, 391–404. Luther, P. W., Peng, H. B., and Lin, J. J. C. (1983). Changes in cell shape and actin distribution induced by constant electric fields. Nature (London) 303, 61–64.
24
Nuccitelli
McBain, V. A., Forrester, J. V., and McCaig, C. D. (2003). HGF, MAPK, and a small physiological electric field interact during corneal epithelial cell migration. Invest. Ophthalmol. Vis. Sci. 44: 540–547. McCaig, C. D., Rajnicek, A. M., Song, B., and Zhao, M. (2002). Has electrical growth cone guidance found its potential? Trends Neurosci. 25, 354–359. McCaig, C. D. and Zhao, M. (1997). Physiological electrical fields modify cell behaviour. BioEssays 19, 819–826. McGinnis, M. E. and Vanable, J. W. Jr. (1986). Electrical fields in Notophthalmus viridescens limb stumps. Dev. Biol. 116, 184–193. McLaughlin, S. and Poo, M.-M. (1981). The role of electro-osmosis in the electric field-induced movement of charged macromolecules on the surfaces of cells. Biophys. J. 34, 85–93. Nishimura, K. Y., Isseroff, R. R., and Nuccitelli, R. (1996). Human keratinocytes migrate to the negative pole in direct current electric fields comparable to those measured in mammalian wounds. J. Cell Sci. 109, 199–207. Nuccitelli, R. (1984). The involvement of transcellular ion currents and electric fields in pattern formation, in Pattern Formation: A Primer in Developmental Biology, edited by G. M. Malacinski, and S. V. Bryant. New York, London: MacMillan Publishing Co., Collier MacMillan Publishers, pp. 23–46. Nuccitelli, R. (1988). Physiological electric fields can influence cell motility, growth, and polarity. Adv. Cell Biol. 2, 213–233. Nuccitelli, R. and Erickson, C. A. (1983). Embryonic cell motility can be guided by physiological electric fields. Exp. Cell Res. 147, 195–201. Nuccitelli, R. and Smart, T. (1989). Extracellular calcium levels strongly influence neural crest cell galvanotaxis. Biol. Bull. 176(S), 130–135. Nuccitelli, R. and Smart, T. (1991). Intracellular Ca2 þ gradients and protein kinase activity are required for the directed translocation of neural crest cells, in Basic Neurobiology: Half a Century and Future, edited by H. Ohmori, and S. Ebashi. Tokyo: Biomedical Research Foundation, pp. 189–194. Nuccitelli, R., Smart, T., and Ferguson, J. (1993). Protein kinases are required for embryonic neural crest cell galvanotaxis. Cell Motil. Cytoskeleton 24, 54–66. Orida, N. and Feldman, J. D. (1982). Directional protrusive pseudopodia activity and motility in macrophages induced by extracellular electric fields. Cell Motility 2, 243–256. Page, C. F. and Gault, W. R. (1975). Managing ischemic skin ulcers. Am. Fam. Physician 11, 108–114. Peng, H. B. and Jaffe, L. F. (1976). Polarization of fucoid eggs by steady electrical fields. Dev. Biol. 53, 277–284. Pullar, C. E., Isseroff, R. R., and Nuccitelli, R. (2001). Cyclic AMP-dependent protein kinase A plays a role in the directed migration of human keratinocytes in a DC electric field. Cell Motil. Cytoskeleton 50, 207–217. Rajnicek, A. M., Robinson, K. R., and McCaig, C. D. (1998). The direction of neurite growth in a weak DC electric field depends on the substratum: contributions of adhesivity and net surface charge. Dev. Biol. 203, 412–423. Rajnicek, A. M., Stump, R. F., and Robinson, K. R. (1988). An endogenous sodium current may mediate wound healing in Xenopus neurulae. Dev. Biol. 128, 290–299. Rapp, B., Boisfleury-Chevance, A., and Gruler, H. (1988). Galvanotaxis of human granulocytes. Dose–response curve. Eur. Biophys. J. 16, 313–319. Robinson, K. R. (1983). Endogenous electrical current leaves the limb and prelimb region of the Xenopus embryo. Dev. Biol. 97, 203–211. Robinson, K. R. (1985). The responses of cells to electrical fields: A review. J. Cell Biol. 101, 2023–2027.
1. A Role for Endogenous Electric Fields in Wound Healing
25
Rowley, B. A. (1972). Electrical current effects on E. coli growth rates. Proc. Soc. Exp. Biol. Med. 139, 929–934. Rowley, B. A., McKenna, J. M., and Chase, G. R. (1974a). The influence of electrical current on an infecting microorganism in wounds. Ann. N.Y. Acad. Sci. 238, 543–550. Rowley, B. A., McKenna, J. M., and Wolcott, L. E. (1974b). Proceedings: The use of low level electrical current for enhancement of tissue healing. Biomed. Sci. Instrum. 10(111–4), 111–114. Sillman, A.L., Quang, D. M., Farboud, B., Fang, K. S., Nuccitelli, R., and Isseroff, R.R. (2003). Human dermal fibroblasts do not exhibit directional migration on collagen I in direct current electric fields of physiological strength. Exp. Dermatol. In press. Sisken, B. F. and Smith, S. D. (1975). The effects of minute direct electrical currents on cultured chick embryo trigeminal ganglia. J. Embryol. Exp. Morphol. 33, 29–41. Song, B., Zhao, M., Forrester, J. V., and McCaig, C. D. (2002a). Electrical cues regulate the orientation and frequency of cell division and the rate of wound healing in vivo. Proc. Natl. Acad. Sci. USA. 99, 13577–13582. Soong, H. K., Parkinson, W. C., Bafna, S., Sulik, G. L., and Huang, S. C. M. (1990a). Movements of cultured corneal epithelial cells and stromal fibroblasts in electric fields. Invest. Ophthalmol. Vis. Sci. 31, 2278–2282. Soong, H. K., Parkinson, W. C., Sulik, G. L., and Bafna, S. (1990b). Effects of electric fields on cytoskeleton of corneal stromal fibroblasts. Curr. Eye Res. 9, 893–901. Sta Iglesia, D. D., Cragoe, E. J., Jr., and Vanable, J. W., Jr. (1996). Electric field strength and epithelization in the newt (Notophthalmus viridescens). J. Exp. Zool. 274, 56–62. Sta Iglesia, D. D. and Vanable, J. W., Jr. (1998). Endogenous lateral electric fields around bovine corneal lesions are necessary for and can enhance normal rates of wound healing. Wound Repair Regen. 6, 531–542. Steckel, R. R., Page, E. H., Geddes, L. A., and Van Vleet, J. F. (1984). Electrical stimulation on skin wound healing in the horse: Preliminary studies. Am. J. Vet. Res. 45, 800–803. Stump, R. F. and Robinson, K. R. (1983a). Xenopus neural crest cell migration in an applied electrical field. J. Cell Biol. 971, 1226–1233. Stump, R. F. and Robinson, K. R. (1986). Ionic currents in Xenopus embryos during neurulation and wound healing. Prog. Clin. Biol. Res. 210, 223–230. Sulik, G. L., Soong, H. K., Chang, P. C., Parkinson, W. C., Elner, S. G., and Elner, V. M. (1992). Effects of steady electric fields on human retinal pigment epithelial cell orientation and migration in culture. Acta Ophthalmol. (Copenh) 70, 115–122. Trollinger, D. R., Isseroff, R. R., and Nuccitelli, R. (2002). Calcium channel blockers inhibit galvanotaxis in human keratinocytes. J. Cell Physiol 193, 1–9. Vanable, J. W. (1989). Integumentary potentials and wound healing, in Electric Fields in Vertebrate Repair, edited by R. B. Borgens, K. R. Robinson, J. W. Vanable, Jr. M. E. McGinnis. New York: Alan R. Liss, Inc, pp. 171–224. Vanable, J. W. Jr. (1991). A history of bioelectricity in development and regeneration, in A History of Regeneration Research, edited by C. E. Dinsmore. Cambridge: Cambridge University Press, pp. 151–177. Wang, E., Zhao, M., Forrester, J. V., and McCaig, C. D. (2003a). Bi-directional migration of lens epithelial cells in a physiological electrical field. Exp. Eye Res. 76, 29–37. Wang, E., Zhao, M., Forrester, J. V., and McCaig, C. D. (2003b). Electric fields and MAP kinase signaling can regulate early wound healing in lens epithelium. Invest Ophthalmol. Vis. Sci. 44, 244–249. Wheeler, P. C., Wolcott, L. E., Morris, J. L., and Spangler, M. R. (1971). Neural considerations in the healing of ulcerated tissue by clinical electrotherapeutic application of weak direct current: Findings and theory, in Neuroelectric Research, edited by D. V. Reynolds, and A. E. Sjoberg. Springfield, IL: CC Thomas, pp. 83–99.
26
Nuccitelli
Winkel, G. K. and Nuccitelli, R. (1989). Large ionic currents leave the primitive streak of the 7.5-day mouse embryo. Biol. Bull. 176, 110–117. Wolcott, L. E., Wheeler, P. C., Hardwicke, H. M., and Rowley, B. A. (1969). Accelerated healing of skin ulcer by electrotherapy: preliminary clinical results. South Med. J. 62, 795–801. Wu, D. T., Go, N., Dennis, C., Enquist, L., and Sawyer, P. N. (1967). Effects of electrical currents and interfacial potentials on wound healing. J. Surg. Res. 7, 122–128. Yang, W.-P., Onuma, E. K., and Hui, S.-W. (1984). Response of C3H/10T1/2 fibroblasts to an external steady electric field stimulation. Exp. Cell Res. 155, 92–104. Zhao, M., Agius-Fernandez, A., Forrester, J. V., and McCaig, C. D. (1996). Orientation and directed migration of cultured corneal epithelial cells in small electric fields are serum dependent. J. Cell Sci. 109, 1405–1414. Zhao, M., Pu, J., Forrester, J. V., and McCaig, C. D. (2002). Membrane lipids, EGF receptors, and intracellular signals colocalize and are polarized in epithelial cells moving directionally in a physiological electric field. FASEB J. 16, 857–859.
2 The Role of Mitotic Checkpoint in Maintaining Genomic Stability Song-Tao Liu1, Jan M. van Deursen2 and Tim J. Yen1 1 Fox Chase Cancer Center, 7701 Burholme Avenue, Philadelphia, PA 19111, USA 2
Department of Pediatric and Adolescent Medicine, Mayo Clinic, Rochester, MN 55905, USA
I. Introduction II. The Kinetochore is a Complex Structure for Cell Division A. The Structure and Function of Kinetochores B. CENP-E: a Kinetochore Associated Kinesin-like Protein III. The Mitotic Checkpoint A. Overview B. Tension and Microtubule Occupancy C. CENP-E and the Mitotic Checkpoint D. Monitoring Microtubule Occupancy E. Checkpoint Inhibition of the APC/C IV. Mutations in Mitotic Checkpoint Proteins and Tumorigenesis A. Mad2, Bub3 and Rae1 are Haplo-Insufficient for Tumor Suppression B. Synergy between Mitotic Checkpoint Genes in Cancer Evolution C. Centromere-associated Protein Knockout Mice V. Conclusions and Future Directions References
I. Introduction ‘‘Considering that 2.5 108 cells are dividing in the human body at any given time, even if few errors occur, many genetically abnormal cells will be produced during the lifetime of an organism.’’ Conly Rieder and Alexey Khodjakov wrote in a recent review (Rieder and Khodjakov, 2003). This is not an overstatement as faithful transmission of chromosomes is a tremendously challenging task during mitosis or meiosis. For somatic cells, defects in this process generate aneuploid cells that may turn into cancer. Aneuploid germ cells on the other hand can be the cause of infertility or birth defects such as Down Syndrome (Jallepalli and Lengauer, 2001; Nicklas, 1997).
Current Topics in Developmental Biology, Vol. 58 Copyright ß 2003, Elsevier, Inc. All rights reserved. 0070-2153/03 $35.00
27
28
Liu et al.
In a typical somatic cell cycle, M phase is usually the shortest when compared to S phase and the two gap phases, G1 and G2. Nevertheless, a mitotic cell undergoes a complex sequence of highly ordered morphological changes, which have amazed biologists for more than 100 years since Walther Flemming first described ‘‘mitosis’’ in 1882. In animal cells, the distinct changes of M phase starts with chromatin condensation in prophase. Duplicated centrosomes move away from each other and the interphase microtubule array disassembles to form a bipolar spindle. After nuclear envelope breakdown, chromosomes are spilled into cytoplasm where they are captured by spindle microtubules that probe the inner space of the cell by rapid cycles of growth and shrinkage. These dynamic microtubules become stabilized if they are attached to kinetochores, a macromolecular complex that is situated on opposite sides of the centromere of chromosomes. These attachments are essential for aligning the chromosomes at the spindle equator. Only when all the chromosomes in a cell are aligned (metaphase) will the signal to release the cell into anaphase be given. Although it is clear that accurate chromosome segregation is dependent on highly complex mechanical processes that physically align and separate chromosomes (McIntosh et al., 2002), these processes are also monitored by checkpoint regulatory systems (Pines and Rieder, 2001; Rieder and Salmon, 1998). The need for checkpoints is evident because attachment of chromosomes to the spindle is a stochastic process. Chromosomes do not synchronously attach to the spindle, but are independent events where individual chromosomes are captured by microtubules through chance encounters. This explains why not all chromosomes in a cell achieve alignment at the same time. This also underscores the need for a checkpoint system that ensures that cells do not prematurely exit mitosis until all of their chromosomes are aligned. The spindle assembly checkpoint is an evolutionarily conserved activity that monitors the kinetochore-microtubule interactions and prevents cells with even a single unattached kinetochore from exiting mitosis (Amon, 1999; Musacchio and Hardwick, 2002). Genetic and biochemical studies have revealed the target of spindle assembly checkpoint is Anaphase Promoting Complex/Cyclosome (APC/C), a multisubunit E3 ubiquitin ligase whose substrates such as securin and cyclin B must be degraded to allow disjunction of chromosomes and exit from mitosis (Harper et al., 2002; Peters, 2002; Zachariae and Nasmyth, 1999). The kinetochore is one of the most important structures established for cell division (Maney et al., 2000; Rieder and Salmon, 1998). It not only provides the binding sites for spindle microtubules, but also harbors many microtubule motors and associated proteins to power and regulate the congression and separation of chromosomes. More interestingly, many
2. The Role of Mitotic Checkpoint in Maintaining Genomic Stability
29
spindle assembly checkpoint proteins are found transiently localized at kinetochores, and kinetochores have been regarded as the major source of inhibitory checkpoint signals for the metaphase-anaphase transition. In this chapter, we will discuss the most current information about the structure, composition, and assembly of kinetochores. We will focus on the role of checkpoint proteins in monitoring kinetochore:microtubule interactions and models of the signal transduction pathways that link unattached kinetochores to the APC/C. The following part of this review will deal with the roles of checkpoint proteins in normal development and in tumorigenesis. Finally, we will briefly talk about the important questions still awaiting for answer. We want to stress that although we will mainly talk about research on mitosis, basic principles also apply to the regulation of meiosis.
II. The Kinetochore is a Complex Structure for Cell Division A. The Structure and Function of Kinetochores The kinetochore is a macromolecular complex that is assembled from centromeres. Readers are referred to several recent excellent reviews that discuss centromere structure (Cleveland et al., 2003; Mellone and Allshire, 2003). Electron microscopic images of kinetochores from animal cells reveal disk-shaped structure with four morphologically distinct domains (Rieder, 1982; Rieder and Salmon, 1998). Juxtaposed to the centromeric heterochromatin is an electron dense inner plate, sometimes hard to distinguish from chromatin. An electron-dense outer plate of about 35–40 nm thickness is separated from inner plate by an electron-lucent middle layer (or central zone, interzone) of 15–35 nm thickness. However, under a different fixation condition that employed high-pressure freezing and freeze substitution, this interzone was not discernible (McEwen et al., 1998). Emanating from the surface of the outer plate is the ‘‘fibrous corona’’ which is a loose meshwork of fibrillar projections. It is likely that this structure contributes to the functional organization of the kinetochore that includes microtubule attachment, force generation, and a checkpoint system that monitors these activities.
B. CENP-E: a Kinetochore Associated Kinesin-like Protein CENP-E (Centromere Protein E) was the first kinesin-like protein that localized specifically to kinetochores in mitotic cells (Yen et al., 1991, 1992).
30
Liu et al.
Immuno-EM studies localized CENP-E at the fibrous corona on the surface of the outer kinetochore plate (Cooke et al., 1997). CENP-E is attached to the kinetochore through a domain at its carboxyl terminus (Chan et al., 1998). This configuration orients its amino-terminal motor-like domain away from the chromosome in order to maximize its interactions with microtubules. CENP-E contains a second microtubule binding domain at its extreme carboxyl-terminus but its importance to function remains to be examined. Functional analysis of human CENP-E by antibody injection, dominant negative mutants and anti-sense oligonucleotides revealed that it is only essential for monopolar chromosomes that normally exist transiently to convert to bipolar attachments (Chan et al., 1998; McEwen et al., 2001; Schaar et al., 1997; Yao et al., 2000). High resolution time-lapse studies showed that CENP-E was essential for chromosomes that are positioned near a pole at the onset of mitosis to establish bipolar attachments (McEwen et al., 2001). In the absence of CENP-E, these chromosomes maintain a monopolar attachment because their sister kinetochores are unable to capture the rare microtubule that originates from the opposite pole. For chromosomes that are positioned near the center of the spindle, the high frequency of encounters with microtubule can compensate for the loss of CENP-E as other (yet to be determined) kinetochore components are able to establish bipolar attachments. Quantitative EM analysis show that kinetochores lacking CENP-E are able to establish near normal ( 73%) the number of microtubule connections. Nevertheless, these connections are defective as tension between the sister kinetochore is never developed. This observation suggests that CENP-E is responsible for generating kinetochore tension and must therefore contribute towards poleward force generation. The caveat to these functional studies is that the methods used to inhibit CENP-E expression or function may not reflect the null state. Recent studies of cells derived from CENP-E null mouse embryos showed that they exhibited similar chromosome defects as reported for Hela cells (Putkey et al., 2002). Likewise, in vitro studies using Xenopus egg extracts depleted of CENP-E also showed an accumulation of monopolar chromosomes (Abrieu et al., 2000). The combined data suggests that CENP-E becomes critically important in areas of low microtubule density where kinetochores must efficiently capture microtubules to achieve biorientation. On the other hand, CENP-E is dispensable for attachment if kinetochores encounter microtubules at high frequencies. There must be other kinetochore components that are responsible for these connections but their identities remain to be determined. In addition to CENP-E, kinetochores also contain other molecular motors such as dynein and MCAK (a kinesin-related protein that does not behave as a motor but is a microtubule destabilizing enzyme)
2. The Role of Mitotic Checkpoint in Maintaining Genomic Stability
31
(Pfarr et al., 1990; Steuer et al., 1990; Wordeman and Mitchison, 1995). Kinetochores also contain microtubule binding proteins CLIP170 and Orbit/Mast (Dujardin et al., 1998; Maiato et al., 2002). It is unclear if any of these microtubule interacting proteins are responsible for the microtubule attachments that are seen when CENP-E is depleted from kinetochores. This question will be addressed by depleting different combinations of these proteins.
III. The Mitotic Checkpoint A. Overview The spindle assembly checkpoint can be viewed as a signal transduction cascade whereby a localized signal generated from an unattached kinetochore is amplified to inhibit the cellular targets that are required for initiating the transition from metaphase to anaphase (Fig. 1). Genetic and biochemical analysis have shown that the target of the checkpoint is the Anaphase Promoting Complex/Cyclosome (APC/C), a multisubunit E3 ubiquitin ligase that specifies the degradation of specific proteins in order to drive cells out of mitosis (Harper et al., 2002; Peters, 2002). The models for how APC is inhibited by checkpoint proteins will be discussed below. The molecular components of the mitotic checkpoint are specified by seven evolutionarily conserved genes that were first identified in budding yeast (Hoyt et al., 1991; Li and Murray, 1991; Weiss and Winey, 1996). Homologs of BUB1, BUB3, MAD1, MAD2, MAD3, and MPS1 have been shown to be essential for establishing the checkpoint response in all eukaryotes examined to date (Abrieu et al., 2001; Basu et al., 1999; Cahill et al., 1998; Chan et al., 1999; Chen et al., 1996; Kitagawa and Rose, 1999; Li and Benezra, 1996; Liu et al., 2003a; Luo et al., 2002; Taylor and McKeon, 1997). In addition, the mammalian ortholog of the yeast nuclear export factor, Rae1, has been shown to be also important for the spindle checkpoint (Babu et al., 2003). Recent studies also documented that ZW10 and ROD, two proteins that have no counterparts in S. cerevisiae but are conservative amongst metazoans, are also essential for the checkpoint (Chan et al., 2000; Scaerou et al., 2001). The appearance of ZW10 and ROD in metazoans may reflect the need for additional checkpoint proteins to monitor increasingly complex activities that are associated with the kinetochore. Where along the checkpoint pathway these proteins act remain an active area of investigation. The finding that all of these proteins preferentially bind to unattached kinetochores suggested that they participate in monitoring kinetochore microtubule interactions (Chan et al., 1999, 2000; Chen et al., 1996; Jablonski et al., 1998). However, some of these
32
Liu et al.
Figure 1 Mitotic checkpoint as a signal transduction pathway. Unattached kinetochores in monotelic chromosomes initiate ‘‘wait anaphase’’ signal transduction and activate mitotic checkpoint, leading to the inhibition of APC/C by BUB/MAD checkpoint proteins. The status monitored at kinetochores may be microtubule occupancy or tension. Syntelic chromosomes may be transformed into mototelic by Aurora B, or the lack of tension on their kinetochores may directly start the checkpoint. It is suggested merotelic attachment may not be able to activate the mitotic checkpoint. Only when all the chromosomes in a mitotic cell reach amphitelic attachment and tension develops between sister kinetochores will the mitotic checkpoint stop and APC/C catalyze the ubiquitination and degradation of securin. The released separase then cut the cohesion between sister chromatids, thus finishing the metaphaseanaphase transition (see color plate).
proteins may also be directly involved in inhibiting the APC (Fang, 2002; Sudakin et al., 2001; Tang et al., 2001). Thus, checkpoint proteins such as Mad2 and BUBR1 may act both at kinetochores and downstream of kinetochores. The ability of unattached kinetochores to inhibit mitotic exit has been long recognized. Indeed early observations suggested that unattached kinetochores may send negative signals to prevent premature anaphase (McIntosh, 1991; Zirkle, 1970). This idea received direct experimental support through a series of experiments by Conly Rieder and Bruce Nicklas’
2. The Role of Mitotic Checkpoint in Maintaining Genomic Stability
33
labs (Li and Nicklas, 1995, 1997; Nicklas et al., 1995; Rieder et al., 1994; Rieder et al., 1995). Rieder et al. documented that anaphase onset always occurred about 23 min after the last kinetochore became attached to microtubules and aligned at the metaphase plate (Rieder et al., 1994). They also observed that it may take up to 3 h for the last monooriented chromosome to achieve bipolar attachment. However, if the unattached kinetochore of the monooriented chromosomes is ablated with a pulse of laser, the cell entered anaphase about 20 min later, the same interval that was seen in a normal cell (Rieder et al., 1995). Similar results were obtained by Nicklas’s lab who showed that repeated detachment of a chromosome in grasshopper spermatocytes delayed anaphase onset indefinitely (Nicklas et al., 1995). Furthermore, if an external force is applied to a monooriented chromosome so that the sister kinetochores are under tension as in the case for a bioriented chromosome, the cell entered anaphase (Li and Nicklas, 1995). Collectively, these experiments show that unattached kinetochores emit an inhibitor of anaphase rather than the ability of aligned chromosomes to emit a positive factor to initiate anaphase.
B. Tension and Microtubule Occupancy Although it is accepted that unattached kinetochores emit the inhibitory signal that delays anaphase onset, what is the nature of the defect that activates the checkpoint? Kinetochores of properly aligned chromosomes are saturated with microtubules and tension develops between the sister kinetochores as the opposing poleward forces try to pull them apart. Results from the micromanipulation experiments conducted by Nicklas’ lab suggested that the checkpoint was sensitive to the level of tension between sister kinetochores. However, the laser ablation experiments indicated that the checkpoint might be monitoring microtubule occupancy as there is little tension between the kinetochores of a monooriented chromosome but cells are able to exit mitosis when the unattached kinetochore is destroyed (Rieder et al., 1995). These data suggest that both tension and microtubule occupancy can regulate the spindle assembly checkpoint, but preference for one over the other may depend on mitosis versus meiosis, cell types and organisms (Zhou et al., 2002b). The notion that the checkpoint is sensitive to only tension or microtubule occupancy may be inaccurate as both criteria must be fulfilled before a cell can exit mitosis. Cells exposed to microtubule inhibitors (taxol, noscapine and low dose of vinblastine) or low temperature will arrest in mitosis despite the fact that their chromosomes appear to be ‘‘aligned’’ at the spindle equator (Shannon et al., 2002; Skoufias et al., 2001; Waters et al., 1998; Zhou et al., 2002a). As these treatments suppress microtubule dynamics
34
Liu et al.
without blocking polymer formation, kinetochores are able to achieve near maximum number of microtubule attachments. Despite this, kinetochores cannot generate tension as the suppression of microtubule dynamics prevents poleward forces from being generated. Mad2 checkpoint protein which has been shown to preferentially bind to unattached kinetochores is no longer detected at kinetochores in taxol treated cells. This suggests that Mad2 is sensitive to microtubule occupancy at kinetochores but not to tension. Although in normal cells, the release of Mad2 from fully attached kinetochores signifies the onset of anaphase, taxol treated cells remain arrested in mitosis because their kinetochores lack tension. What are the checkpoint proteins that are responsible for monitoring tension remains unresolved, but Bub1 and BubR1 have been postulated to play this role as they are present at tensionless kinetochores (Skoufias et al., 2001; McEwen et al., 2001). It should be pointed out that unlike Mad2, neither Bub1 or BubR1 completely dissociate from fully attached kinetochores in normal metaphase cells that are about to enter anaphase (Jablonski et al., 1998). The mere presence of Bub1 and BubR1 is therefore not a reliable indicator of the checkpoint status of cells. If these proteins are involved in monitoring tension, it is likely that it is achieved through regulation of their kinase activities rather than localization. However, a recent study found evidence that the loss of kinetochore tension is insufficient to block anaphase onset. When tension at kinetochores in grasshopper spermatocytes was removed by micromanipulation, the number of kinetochore microtubules decreased by 60% (King and Nicklas, 2000). This is consistent with earlier studies showing that tension altered the stability of kinetochore microtubules (Nicklas and Ward, 1994; Zhai et al., 1995). Mad2 levels were reduced to approximately 17% of that found at a fully unattached kinetochore. Despite the lack of tension and reduced microtubule occupancy, the spermatocytes were only able to delay anaphase onset rather than arrest. These cells are believed to be unable to arrest because their kinetochores, perhaps due to a reduced amount of Mad2, are unable to generate sufficient amounts of ‘‘wait anaphase’’ signal to sustain a prolonged arrest. These authors postulated that the microtubule attachment determines the strength of the output of the checkpoint signal, but tension may be essential to saturate microtubule binding at kinetochores and completely turn off the checkpoint (Nicklas et al., 2001). The interplay between microtubule attachment and tension is complex and it may be difficult to dissect their individual contributions to the checkpoint. However, it now appears that the Ipl1/Aurora B kinase may link tension to microtubule attachments. Yeast mutants with unreplicated chromosomes (cdc6) or chromosomes with defective cohesion (mcd1) arrest in mitosis because their kinetochores lack tension even though they are
2. The Role of Mitotic Checkpoint in Maintaining Genomic Stability
35
attached to microtubules (Biggins and Murray, 2001). This arrest is dependent on the Ipl1 kinase even though Ipl1 is not required for cells to arrest in mitosis in the presence of microtubule inhibitors that prevent kinetochore attachments. While one interpretation is that Ipl1 is a checkpoint protein that monitors tension, another view is that Ipl1p facilitates biorientation by promoting turnover of the microtubule connections at kinetochores that lack tension (Tanaka, 2002; Tanaka et al., 2002). Ipl1 may be part of a self-correcting mechanism that prevents syntelic attachments, a situation whereby sister kinetochores become attached to the same pole. As syntelic attachments should fulfill the microtubule occupancy checkpoint, it is critical that this defect be corrected in order to prevent nondisjunction. This can be accomplished if the absence of tension activates Ipl1 kinase to release microtubule attachments. In this scenario, Ipl1p indirectly activates the checkpoint in the absence of kinetochore tension by catalyzing the release of microtubules. This idea can also account for why Ipl1p is not required for the checkpoint arrest induced by the loss of microtubule attachments. The yeast data have now been confirmed by studies in mammalian cells that show inhibition of Aurora B kinase prevents cells from arresting in mitosis in the presence of taxol (no tension) but not nocodazole (no microtubule occupancy) (Ditchfield et al., 2003; Hauf et al., 2003). The mechanism by which Aurora B kinase monitors tension is not known, but its localization between sister kinetochores suggests that it may be sensitive to centromere stretching. The link between Aurora B and the spindle checkpoint is not entirely clear as there is a discrepancy as to whether inhibition of Aurora B kinase interferes with the ability of Mad2 and BubR1 to assemble onto kinetochores (Ditchfield et al., 2003; Hauf et al., 2003). Regardless of this discrepancy, the reason for why these cells are able to arrest in nocodazole but not taxol remains to be sorted out. One explanation hinges on the assumption that unattached kinetochores lacking Aurora B are unable to generate sufficient levels of ‘‘wait anaphase’’ signal. However, the collective output from all of the unattached kinetochores in nocodazole treated cells may be sufficient to arrest cells in mitosis. In contrast, in the absence of Aurora B, the output from kinetochores with attachments but no tension might be even lower than a fully unattached kinetochore. Consequently, the collective amount of inhibitory signal generated from kinetochores of taxol treated cells may be insufficient to arrest mitosis.
C. CENP-E and the Mitotic Checkpoint Functional studies of CENP-E revealed that its activity was also monitored by the spindle assembly checkpoint as Hela cells lacking CENP-E functions
36
Liu et al.
were arrested in mitosis (McEwen et al., 2001; Schaar et al., 1997; Yao et al., 2000). A molecular connection between the checkpoint and CENP-E was revealed when it was discovered that CENP-E interacts with hBUBR1, a protein kinase that exhibited similarities with the MAD3 and BUB1 spindle checkpoint proteins in budding yeast (Chan et al., 1998, 1999). Indeed, disruption of hBUBR1 function in Hela cells abrogated their ability to arrest in mitosis in response to microtubule destabilizing drugs (Chan et al., 1999). More importantly, it was shown that the mitotic arrest induced by the disruption of CENP-E function depended on hBUBR1. Given that hBUBR1 was localized to kinetochores, the combined results supported a model where the kinetochore activities of CENP-E was monitored by hBUBR1 kinase. We envisioned hBUBR1 to behave as a mechanosensor where its checkpoint activity was regulated by the interactions between CENP-E and microtubules. Depending on these interactions, hBUBR1 kinase activity may be allosterically regulated to either initiate or silence the checkpoint signal from kinetochores (Chan and Yen, 2003). The ability of cells to arrest in mitosis in response to inhibition of CENP-E functions is not universal. Xenopus egg extracts depleted of CENP-E fail to align their chromosomes, yet they are unable to maintain a mitotic arrest when compared to extracts treated with microtubule inhibitors (Abrieu et al., 2000). This observation supports the idea that CENP-E is an integral component of the checkpoint. However, the reason for why kinetochores lacking CENP-E fail to establish a checkpoint arrest is likely due to the absence of the Mad2 checkpoint protein at these kinetochores. This contrasts with studies in Hela cells where unattached kinetochores that were depleted of CENP-E retained Mad1, Mad2 (McEwen et al., 2001). Thus, the disparity between how egg extracts and Hela cells respond to the loss of CENP-E may be attributed to how CENP-E affects the assembly of checkpoint proteins at kinetochores. The reason for the difference between egg extracts and Hela cells is unresolved, but a likely possibility is that it is due to fundamental differences in how kinetochores are assembled between embryonic and somatic cells. Unfortunately, this explanation is clouded by the recent report where hepatic cells obtained from a conditional CENP-E knockout mouse also failed to arrest in mitosis in the presence of unaligned chromosomes (Putkey et al., 2002). However, it is unknown whether the length of the mitotic delay exhibited by these cells in response to loss of CENP-E is the same as their responses to nocodazole or taxol. The apparent discrepancy between how Hela cells and the mouse hepatocytes respond to the inactivation of CENP-E may lie in inherent differences in the duration of the checkpoint arrest.
2. The Role of Mitotic Checkpoint in Maintaining Genomic Stability
37
D. Monitoring Microtubule Occupancy The ability of Mad2 to preferentially bind to kinetochores that lack microtubule attachments suggests that one of its checkpoint functions is to monitor microtubule attachments. Indeed Mad2 is essential for arresting cells in mitosis when their kinetochores lack microtubule attachments. However, two recent studies in Hela cells suggest that there may be redundant mechanisms that monitor microtubule attachments (Liu et al., 2003b; Martin-Lluesma et al., 2002). Hela cells were found to delay in mitosis for a considerable amount of time despite the lack of detectable Mad2 at their kinetochores. In previous studies, direct inhibition of Mad2 caused cells to accelerate out of mitosis within 10 min of its inhibition (Gorbsky et al., 1998). Given that Mad2 appears to act both at kinetochores where it contributes towards initiating the signaling cascade and downstream from kinetochores by directly inhibiting the APC, it was impossible to discern these two activities by directly inhibiting Mad2. What makes these two recent studies intriguing is that they selectively disrupted Mad2 functions at kinetochores by preventing its ability to assemble there. When HEC1 was prevented from assembling onto kinetochores, Mad1, Mad2, and hMPS1 also failed to bind to kinetochores (Martin-Lluesma et al., 2002). Similarly, Mad1 and Mad2 (the localization of hMPS1 was not tested) were unable to bind to kinetochores when CENP-I was prevented from assembling there (Liu et al., 2003b). Interestingly, injection of Mad2 antibodies into the mitotically delayed cells resulted in their rapid exit from mitosis. The caveat of these studies is whether the arrest is mediated by residual Mad2 that is below the limits of detection. With this in mind, these latest findings suggest that the kinetochore localization of Mad1, Mad2 and hMPS1 are not essential for cells to arrest in mitosis in response to loss of microtubule attachment. The presence of hBUBR1, hBUB1 and hZW10 at these kinetochores suggests that they are likely responsible for maintaining the arrest. Regardless, these findings increase the complexity by which checkpoint proteins monitor kinetochore attachments as it suggests the presence of redundant monitoring systems. As described for Aurora B, it is possible that no individual kinetochore lacking Mad1, Mad2, and hMPS1 is able by itelf to generate sufficient amounts of ‘‘wait anaphase’’ signal to arrest cells in mitosis. However, the collective output from many unattached kinetochores are required to achieve the threshold level to arrest mitosis.
E. Checkpoint Inhibition of the APC/C The target of the mitotic checkpoint is the APC/C. Two models have been proposed to explain how checkpoint proteins inhibit the APC/C.
38
Liu et al.
1. Anaphase Promoting Complex/Cyclosome (APC/C) APC/C is an evolutionarily conserved multisubunit E3 ubiquitin ligase that was originally identified in clam oocytes, frog egg extracts, and yeast (Irniger et al., 1995; King et al., 1995; Sudakin et al., 1995). It is composed of at least eleven subunits (Peters, 2002), with new subunits being discovered (Hall et al., 2003). Many proteins have been found ubiquitinated by APC/C and targeted to 26S proteasome for degradation (Harper et al., 2002; Peters, 2002). However the two key substrates whose degradation is critical for sister chromatid separation and mitotic exit are securin/pds1 and cyclin B, respectively. Securin binds to and inhibits separase/ESP1, a cysteine protease that is believed to degrade the Scc1/Mcd1 subunit of the cohesin complex in order to dissolve sister chromatid cohesion (Nasmyth et al., 2000). Cyclin B associates and activates Cdc2 kinase whose activity is essential for maintaining cells in mitosis. The APC/C relies on specificity factors such as CDC20 and CDH1 to recruit substrates to the APC/C. APC/ CCDC20 activity is required for the metaphase to anaphase transition as it ubiquitinates proteins such as securin and mitotic cyclins. On the other hand, APC/CCDH1 promotes progression through the late stages of mitosis by ubiquitinating cyclin B and CDC20. Readers are referred to several comprehensive reviews for more information on APC/C (Harper et al., 2002; Peters, 2002; Zachariae and Nasmyth, 1999).
2. Sequestration Model The molecular link between the mitotic checkpoint and the APC/C was established when genetic and biochemical evidence from yeast and frog egg extracts, respectively, showed that Mad2 can bind CDC20 and thus prevent its ability to recruit substrates to the APC/C (Fang et al., 1998; Hwang et al., 1998). This finding coupled with the in vivo observation that Mad2 undergoes rapid rates of exchange at kinetochores (Howell et al., 2000) led to a molecular model for how unattached kinetochores inhibit the APC/C (Shah and Cleveland, 2000; Yu, 2002). This model suggests that Mad2 undergoes a conformational change through a transient interaction with kinetochores. Upon release from kinetochores, the ‘‘activated’’ Mad2 sequesters CDC20 and thus prevents activation of the APC/C. A critical feature of this model is that all of the steps along the checkpoint pathway (initiation, amplification and target inhibition) are intimately linked through the pool of Mad2 that cycles through the kinetochore. In a recent study, PtK1 cells were injected with a Mad1 mutant that is unable to bind kinetochores but retains its ability to bind to Mad2 (Canman et al., 2002). The injected cells were found to exit mitosis despite the presence of Mad2 at their kinetochores. Furthermore, the kinetochore associated
2. The Role of Mitotic Checkpoint in Maintaining Genomic Stability
39
Mad2 exhibited the same dynamic properties as in control cells. One interpretation is that the Mad1 mutant is sequestering the Mad2 that is released from kinetochores so that it cannot bind to CDC20. However, based on the results from the HEC1 and CENP-I studies where Mad2 acts at two discrete steps along the checkpoint pathway, an alternative view is that mitotic exit is induced because the Mad1 mutant acts on the pool of Mad2 that is directly responsible for inhibiting the APC/C. The assumption here is that the Mad2 that is released from kinetochores is unimportant for inhibiting the APC/C. Indeed, there is currently no data that document the fate of the Mad2 molecules that are released from kinetochores. A similar mechanism for how hBUBR1 inhibits the APC has been postulated. In vitro binding assays and yeast two hybrid data showed that hBUBR1 can bind to CDC20 (Fang, 2002; Tang et al., 2001; Wu et al., 2000). This observation has fueled speculation that hBUBR1 may act in parallel with Mad2 to sequester CDC20 and prevent activation of the APC/C. Using an assay whereby APC/C activity is dependent upon exogenous CDC20, addition of hBUBR1 or Mad2 prevented activation of APC/C. The caveat of this experiment is that the APC/C used in these studies are from interphase cells and thus not the physiologically relevant substrate for inhibition. Indeed, when these assays were performed with APC/C purified from mitotic cells, recombinant hBUBR1 failed to inhibit its activity (Tang et al., 2001).
3. APC/C Sensitization Model An alternative to the sequestration model proposes that the checkpoint directly inhibits the APC/C. This idea originated from the discovery in Hela cells of a factor that selectively inhibited mitotically active APC/C (Sudakin et al., 2001). Lysates prepared from mitotically arrested Hela cells were fractionated to identify factors that inhibited APC/C activity. This led to the identification of the Mitotic Checkpoint Complex MCC, which consists of the checkpoint proteins hBUBR1, hBUB3, CDC20 and Mad2. The evidence that the MCC is the physiologically relevant inhibitor of the APC/C are based on the following observations: 1. APC/C that is purified from mitotically arrested cells is either free or associated with the MCC. The free APC/C exhibits ubiquitin ligase activity while the pool that is associated with the MCC is inactive. 2. All studies that showed Mad2 can inhibit the APC/C relied on the use of recombinant Mad2 protein at levels that were at least an order of magnitude higher than endogenous levels.
40
Liu et al.
3. Direct comparison of the APC/C inhibitory activity between purified MCC and recombinant Mad2 showed that the MCC was >3000-fold more potent of an inhibitor than recombinant Mad2. 4. Based on the titration experiments, the intracellular concentration of ‘‘free’’ pool of Mad2 would have to be >3000-fold higher than MCC in order for it to be an effective inhibitor. Although there is a large pool of Mad2 that is not associated with the MCC in Hela cells, this pool is no more than 25 to 50-fold higher concentration than the amount of Mad2 that is associated with MCC. Furthermore, fractions containing this pool of Mad2 exhibits no detectable APC/C inhibitory activity. 5. The concentration of MCC is in near equal stoichiometry with the APC/C. The discovery of MCC also challenged the prevailing view that the inhibitor of the APC/C is directly generated from unattached kinetochores. The MCC was isolated in mitotic Hela cells, but it was found to be present and fully active in interphase cells. As kinetochores are not even assembled during interphase, MCC formation must occur independently of kinetochores. Although MCC synthesis and activity were not subject to cell cycle regulation, it preferentially inhibits mitotic APC/C. The need for a preformed pool of inhibitor is apparent because APC/C is phosphorylated and rapidly activated at the onset of mitosis. By necessity, the inhibition of the APC/C by the MCC must also be reversible so that cells can exit mitosis. It is believed that the interaction between MCC and the APC/C is labile unless the presence of unattached kinetochores stabilizes the interaction. This is indirectly supported by reconstitution experiments that showed APC/C activity in lysates prepared from mitotically arrested Hela cells cannot remain suppressed as ubiquiting ligase activity is reproducibly reactivated after an initial lag (Sudakin et al., 2001). This lag, which represents the checkpoint inhibited APC/C activity, can be extended when chromosomes (unattached kinetochores) are added to these extracts. As neither MCC inhibitory activity or CDC20 stimulatory activity are stimulated by chromosomes, the likely target of kinetochores appears to be the APC/C. The combined data suggests a model whereby the ‘‘wait anaphase’’ signal that is generated by kinetochores does not directly inhibit the APC/C but rather sensitizes the APC/C to prolonged inhibition by the MCC. This model predicts that the checkpoint pathway can be separated into discrete steps that are acted on by different components of the checkpoint. The checkpoint proteins that are localized at kinetochores initiate a signal in response to improper microtubule attachments. This signal must be amplified, perhaps through a kinase cascade, to target
2. The Role of Mitotic Checkpoint in Maintaining Genomic Stability
41
a large population of APC/C for prolonged inhibition by the cytosolic pool of MCC. Presently, it is unclear why kinetochores and the MCC share many of the same proteins especially when the biochemical activities required to generate the ‘‘wait anaphase’’ signal and to bind and inhibit the APC/C are likely to be quite different. One possibility is that hBUBR1, hBUB3, CDC20, and Mad2 do not form the MCC when they are associated at kinetochores. For example, hBUBR1 is believed to interact with CENP-E when it is associated at kinetochores even though it is part of the MCC in the cytoplasm. By altering protein interactions, it may be possible for a single protein to adopt multiple functions. The existence of the MCC in Hela cells has now been independently confirmed by others. MCC-like complexes have also now been identified in fission, budding yeasts and frog egg extracts (Chen, 2002; Fraschini et al., 2001; Millband and Hardwick, 2002). In budding yeast, the formation of the MCC was also found to be independent of kinetochores (Fraschini et al., 2001). Whether these complexes represent physiological inhibitors of the APC/C in those species remains to be tested. The existence of an inhibitor of the APC/C whose activity is not dependent on kinetochores is consistent with the observations that Mad2 appears to act at two distinct steps along the checkpoint pathway. When Mad2 is depleted from kinetochores by disruption of HEC1 or CENP-I, cells are able to delay mitosis in a Mad2 dependent manner (Liu et al., 2003b; Martin-Lluesma et al., 2002). The requirement for Mad2 in these cases likely reflects the role of the MCC in inhibiting the APC/C. If kinetochores lacking Mad2 are only able to generate a low level of ‘‘wait anaphase’’ signal, it is unable to sensitize the APC/C to prolonged inhibition by the MCC.
IV. Mutations in Mitotic Checkpoint Proteins and Tumorigenesis A. Mad2, Bub3, and Rae1 are Haplo-Insufficient for Tumor Suppression An estimated 85% of human cancer cells possess an abnormal number of chromosomes. Thus, researchers have long been curious about the role of aneuploidy in the multi-step cancer process. Indeed, whether chromosomal instability and aneuploidy is the cause or merely a consequence of cancer remains a central question in cancer biology. Since the discovery of the spindle assembly checkpoint in yeast, researchers have speculated that loss of this checkpoint in humans would play a key role in the development of
42
Liu et al.
aneuploidy in human cancers. Several laboratories have tested this hypothesis by disrupting key components of the spindle assembly checkpoint in the mouse by gene targeting and embryonic stem cell technology. To date, Mad2, Bub3, and Rae1 knockout studies have been reported each showing that complete mitotic checkpoint protein loss results in early embryonic death (Babu et al., 2003; Dobles et al., 2000; Kalitsis et al., 2000). In all cases, null embryos are indistinguishable from wildtype embryos until the blastocyst stage (32-cell stage), but subsequently fail to expand their inner cell mass of pluripotent mitotic cells and begin to degenerate. At first, the embryonic lethality seemed to limit the usefulness of gene knockout models for studying the connection between chromosomal instability and cancer, but recent studies of heterozygous knockout mice have proven otherwise. Although Mad2, Bub3, and Rae1 heterozygous knockout mice have no overt phenotype, cells from these mice show markedly impaired mitotic checkpoint activation and mis-segregate chromosomes at higher than normal rates (Babu et al., 2003; Michel et al., 2001). Consistent with a functional role for chromosomal instability in cancer development, Mad2, Bub3, and Rae1 heterozygous knockout mice are more susceptible to formation of spontaneous and/or carcinogeninduced lung tumors than normal mice. These observations suggest that mitotic checkpoint genes function as haplo-insufficient tumor suppressors. This class of tumor-suppressor genes may be frequent targets during cell transformation processes because inactivation of only one allele or a reduction in gene expression level is sufficient to advance the multi-step process of cancer (Fero et al., 1998). One mechanism that should be very effective in reducing mitotic checkpoint gene expression levels involves hypermethylation of CpG islands, an epigenetic means of DNA modification that is common in human cancers (Laird, 2003). Indeed, a recent study from Shichiri and coworkers shows that epigenetic silencing of the mitotic checkpoint genes Bub1 and BubR1 is a frequent event in aneuploid human colon carcinomas, with 30% of the carcinomas exhibiting at least a two-fold reduction in Bub1 or BubR1 expression (Shichiri et al., 2002). Such Bub1 or BubR1 reductions are expected to predispose cells to chromosomal mis-segregation in mitosis and may have established the aneuploidy in the tumors. Another mechanism by which reduced expression of mitotic checkpoint genes could be accomplished involves the loss of whole chromosomes in mitosis. For instance, as a result of chromosomal missegregation, a cell may lose a chromosome containing a mitotic checkpoint gene. Such a mis-segregation event might be caused by genotoxic agents (Hesterberg and Barrett, 1985; Hunt et al., 2003), mitotic checkpoint gene mutations, or might simply be a fortuitous event. Knowing that the mammalian mitotic checkpoint system is extremely sensitive to underexpression of its components, it is not surprising that many studies have
2. The Role of Mitotic Checkpoint in Maintaining Genomic Stability
43
concluded that inactivating point mutations in mitotic checkpoint genes are rare events in human tumors with chromosomal instability (Cahill et al., 1998, 1999; Gemma et al., 2001; Hempen et al., 2003; Hernando et al., 2001; Imai et al., 1999; Lin et al., 2002; Mimori et al., 2001; Nakagawa et al., 2002; Nomoto et al., 1999; Olesen et al., 2001; Ouyang et al., 2002; Percy et al., 2000; Reis et al., 2001; Sato et al., 2000; Tsukasaki et al., 2001). More recent studies have identified several novel components of the mitotic checkpoint that were not included in the reported screenings of human tumors for mitotic checkpoint mutations (Babu et al., 2003; Cahill et al., 1999). Therefore, more definitive answers on the frequency of mitotic checkpoint gene mutations in human tumors with chromosomal instability will have to await the results from additional screening studies.
B. Synergy between Mitotic Checkpoint Genes in Cancer Evolution Regardless of the actual mechanism that causes the initial chromosomal mis-segregation event, it triggers a process that generates ever new and ultimately tumorigenic karyotypes (Duesberg and Rasnick, 2000; Duesberg et al., 1999). Experiments with mice in which Rae1 and Bub3 are deleted individually or in combination suggest that mitotic checkpoint genes may act to regulate chromosomal instability rates in the evolution of cancer (Babu et al., 2003). Unlike mice that are homozygous null for Rae1 or Bub3, mice that are double heterozygous for Bub3 and Rae1 are born alive. Although double heterozygotes have no overt abnormalities, cells from these mice exhibit much greater rates of premature sister chromatid separation and chromosome mis-segregation than single haplo-insufficient cells. These findings suggest that Bub3 and Rae1 act synergistically to prevent aneuploidy. It is therefore conceivable that the initial loss of a single mitotic checkpoint gene, for instance Rae1, might start a vicious cycle in which reduced expression of that checkpoint protein causes additional chromosome loss. If that loss happened to involve mouse chromosome 7 containing the Bub3 gene locus, the rate of chromosomal instability would significantly increase. Because compound Rae1/Bub3 heterozygotes seem to be more susceptible to DMBA induced lung tumor formation than the single heterozygous mice, it seems that the increased chromosomal instability accelerates tumorigenesis. However, once a tumorigenic karyotype has been established, preservation of this karyotype might provide a selective advantage. Perhaps by regaining a chromosome containing a mitotic checkpoint gene that is haplo-insufficient (for instance chromosome 7) cells might be able to reduce the chromosomal instability rate and preserve their tumorigenic karyotype. The model is summarized and presented in Fig. 2. Haplo-insufficient mouse models will prove useful in
44
Liu et al.
Figure 2 A model for how mitotic checkpoint genes may contribute to cancer evolution. Loss of a single mitotic checkpoint gene copy may drive a normal cell into a pathway of cancer (PE). Potential causes of such a loss include carcinogens that affect chromosome segregation, mitotic checkpoint gene mutations or epigenetic events that reduce mitotic checkpoint gene expression. The resulting checkpoint-defective cell may generate new karyotypes at relatively low rates. Further loss of a chromosome that contains another mitotic checkpoint gene (C1) may accelerate the mis-segregation rate and promote the formation of more tumorigenic karyotypes. Once a highly malignant karyotype has been established, preservation of this karyotype might be advantageous. One strategy for improving karyoptypic stability might be the regaining of lost chromosomes that contain mitotic checkpoint genes (see color plate).
elucidating whether the rate of chromosomal instability indeed declines at the end stages of cancer evolution.
C. Centromere-associated Protein Knockout Mice Mad2, Bub3, and Rae1 mouse studies have taught us that the majority of cell divisions of checkpoint defective cells produce daughter cells with modal chromosome numbers. Thus, the mitotic checkpoint seems to function as a backup mechanism that prevents the occasional problem of mis-segregation from occurring in mitosis. However, the mitotic checkpoint machinery is likely to take on a more central role when key components of the chromosomal segregation machinery other than those involved in the mitotic checkpoint are altered by genetic or epigenetic events. Thus far, CENP-A, CENP-B, CENP-C, and CENP-E, have been studied in the mouse by gene disruption methods. CENP-B knockout mice are viable, have a normal lifespan and display no overt phenotype (Hudson et al., 1998; Kapoor et al., 1998; Perez-Castro et al., 1998). In contrast, CENP-A
2. The Role of Mitotic Checkpoint in Maintaining Genomic Stability
45
(Howman et al., 2000), CENP-C (Kalitsis et al., 1998) and CENP-E (Putkey et al., 2002) knockout mice die during early embryogenic development. Like Mad2, Bub3, and Rae1 null embryos, CENP-A and CENP-E null embryos develop normally until the blastocyst stage but then fail to expand their inner cell mass and die. CENP-C null embryos start to degenerate a day earlier at the morula stage (16-cell stage). Probably CENP-A, CENP-C, and CENP-E knockout embryos develop normally in the earliest stages of development due to the presence of maternal gene products. However, once these products are depleted due to RNA and protein degradation processes, the consequences of the respective gene disruptions become apparent. The difference in onset of embryo degradation between CENP-C on the one hand and CENP-A and CENP-E on the other may merely result from differences in maternal product stability or protein level requirement, or both. Early embryonic death has hampered the phenotypic analysis of the CENP-A, CENP-C, and CENP-E knockout mice, but in all cases severe mitotic problems preceded the time of embryonic death. CENP-A null mice typically displayed micro- and macronuclei formation, nuclear bridging and blebbing, and chromatin fragmentation. CENP-C null mice exhibited similar features of chromosomal mis-segregation as CENP-A null mice. CENP-E loss produced metaphases with misaligned and/or centrophilic chromosomes due to unstable attachments between kinetochores and microtubules. CENP-E binds to the mitotic checkpoint protein BubR1 and its loss might inactivate the mitotic checkpoint. The high incidence of metaphases with lagging and centrophilic chromosomes suggests that CENP-E null cells are incapable of delaying mitosis despite the presence of unattached chromosomes, implying that the mitotic checkpoint is defective. It will be of interest to further test whether CENP-E deficient cells indeed will exit mitosis prematurely in the presence of spindle poisons such as nocodazole, just like cells that are insufficient for Mad2, Bub3, or Rae1. It will also be of interest to analyze whether mice that are haplo-insufficient for CENP-A, CENP-C, or CENP-E exhibit increased chromosomal instability and carcinogen-induced tumor formation.
V. Conclusions and Future Directions The goal of this review is to provide current mechanistic views of a highly complex process that ensures that cells with even a single unaligned chromosome will not prematurely exit mitosis (Fig. 3). These models are only possible because of the discovery of the molecular components of the spindle checkpoint, the kinetochore and the Anaphase Promoting Complex. Despite these advances, outstanding questions regarding all aspects of this signaling pathway remain unanswered. It remains to be determined how
46
Liu et al.
Figure 3 The mitotic checkpoint pathways in human cells. A. At least three redundant signaling pathways exist inside the unattached/tensionless kinetochore. The sensors (e.g., CENP-E) detect the absence of microtubule binding and/or tension, affect the behavior of some checkpoint proteins at the kinetochores and emit the ‘‘wait anaphase’’ signals. Several components recycling fast at kinetochores (curved arrows) may help the amplification of signals in the cytoplasm. The exact roles of several checkpoint proteins like BUB1, RAE1 in the kinetochores are still unknown. B and C are two models to explain how the signals originating from kinetochores lead to the inhibition of APC/C. B. Sequestration model. In this model, the unattached kinetochore facilitates a conformational change of MAD2 and results in its binding to CDC20. Sequestration of CDC20 from APC/C this way inhibits its ubiqutin ligase activity and prevents the anaphase onset. Several variants of this model exist now but what is common is they all prefer de novo formation of inhibitor(s) by unattached kinetochores to inhibit APC/C. C. APC/C sensitization model. In this model the inhibitory complex MCC exists throughout the cell cycle. The signals from unattached kinetochores may be amplified and lead to phosphorylation (or other modifications) of all the APC/C in the cytoplasm. This modified form of APC/C is sensitized and inhibited by MCC (see color plate).
checkpoint proteins monitor microtubule occupancy and tension at kinetochores (Fig. 3A). Although hBUBR1 is postulated to monitor CENP-E activities at kinetochores, this remains to be rigorously demonstrated. In addition, how are the microtubule binding activities of other proteins monitored? It is formally possible that each protein is assigned a different checkpoint protein. Alternatively, the different activities are relayed to a centralized detector that may be composed of multiple checkpoint proteins. In addition to monitoring microtubule attachments, the sensor is also intimately linked to the generation of the ‘‘wait anaphase’’ signal.
2. The Role of Mitotic Checkpoint in Maintaining Genomic Stability
47
The prevalent model which is based on rapid rate at which Mad2 binds and dissociates from unattached kinetochores suggests that Mad2 is the ‘‘wait anaphase’’ signal (Fig. 3B). Unattached kinetochores are believed to catalytically convert Mad2 into an altered state such that upon its release, the ‘‘activated’’ Mad2 can bind to CdC20 and sequester it from activating the APC/C. The discovery of the MCC has led to an alternative model whereby ‘‘wait anaphase’’ is not directly responsible for inhibiting the APC/C, but rather sensitizes the APC/C to prolonged inhibition by a cytosolic pool of inhibitor (Fig. 3C). Implicit in this model is that the ‘‘wait anaphase’’ signal must be amplified so that a single unattached kinetochore can target the large number of APC/C in the cell. The molecules involved in the amplification step is unknown but a reasonable prediction is a kinase cascade, perhaps mediated by some of the checkpoint kinases themselves. Finally, the mechanism by which MCC inhibits the APC/C is unknown, but a key is likely to lie in understanding how it distinguishes between mitotic vs interphase forms of the APC/C. Separate from these mechanistic issues, the role of the spindle checkpoint in development and cancer remains a priority because of its direct connection to aneuploidy. In all cases examined so far, spindle checkpoint genes are essential for viability. This is clearly distinct from some DNA damage checkpoint genes like ATM where homozygous null mutants are viable. This difference may reflect the fact that chromosome segregation is inherently an error prone process whereby the checkpoint is essential for ensuring that mistakes are corrected. Thus, the accumulation of aneuploid cells during the earliest stages of embryogenesis may result in massive cell death or severe cellular defects that cannot sustain continued development. In the future, it will be interesting to generate conditional knockout mice so that it will be possible to test how a mature animal responds to the loss of the spindle checkpoint. One expectation is that these animals will develop tumors at frequencies that are higher than that seen for the haplo-insufficient mutants. This prediction is based on fundamental differences in the biochemical status of the spindle checkpoint between a null and a heterozygote mutant. In heterozygotes, a reduced level of a checkpoint protein might lower the overall output from an unattached kinetochore so that an unaligned chromosome cannot sustain a prolonged delay and cells exit mitosis prematurely. This contrasts with a null mutant where it may not be able to delay mitosis in response to the presence of unaligned chromosomes. Consequently, the null mutants will exit mitosis with unaligned chromosomes more frequently than a heterozygote. It is not difficult to imagine with 2.5 108 cells dividing in the human body at any given time that a small increase in the frequency of aneuploidy will accelerate tumor formation during the lifespan of an individual.
48
Liu et al.
Acknowledgments The authors gratefully acknowledge support from grants NIH-CA77262-01 and CA96985-1 (to J.M.v.D) and DMAD17-01-1-0239, NIH-GM44762, CA75138, core grant CA06927, March of Dimes and an appropriation from the Commonwealth of Pennsylvania (to T. J. Y). S.T.L. is supported by the Lawrence Greenwald Fellowship.
References Abrieu, A., Kahana, J. A., Wood, K. W., and Cleveland, D. W. (2000). Cell 102, 817–826. Abrieu, A., Magnaghi-Jaulin, L., Kahana, J. A., Peter, M., Castro, A., Vigneron, S., Lorca, T., Cleveland, D. W., and Labbe, J. C. (2001). Cell 106, 83–93. Amon, A. (1999). Curr. Opin. Genet. Dev. 9, 69–75. Babu, J. R., Jeganathan, K. B., Baker, D. J., Wu, X., Kang-Decker, N., and van Deursen, J. M. (2003). J. Cell Biol. 160, 341–353. Basu, J., Bousbaa, H., Logarinho, E., Li, Z., Williams, B. C., Lopes, C., Sunkel, C. E., and Goldberg, M. L. (1999). J. Cell Biol. 146, 13–28. Biggins, S., and Murray, A. W. (2001). Genes. Dev. 15, 3118–3129. Cahill, D. P., da Costa, L. T., Carson-Walter, E. B., Kinzler, K. W., Vogelstein, B., and Lengauer, C. (1999). Genomics 58, 181–187. Cahill, D. P., Lengauer, C., Yu, J., Riggins, G. J., Willson, J. K., Markowitz, S. D., Kinzler, K. W., and Vogelstein, B. (1998). Nature 392, 300–303. Canman, J. C., Sharma, N., Straight, A., Shannon, K. B., Fang, G., and Salmon, E. D. (2002). J. Cell Sci. 115, 3787–3795. Chan, G. K., Jablonski, S. A., Starr, D. A., Goldberg, M. L., and Yen, T. J. (2000). Nat. Cell Biol. 2, 944–947. Chan, G. K., Jablonski, S. A., Sudakin, V., Hittle, J. C., and Yen, T. J. (1999). J. Cell Biol. 146, 941–954. Chan, G. K., Schaar, B. T., and Yen, T. J. (1998a). J. Cell Biol. 143, 49–63. Chan, G. K., Yen, T. J. (2003). The mitotic checkpoint: a signaling pathway that allows a single unattached kinetochore to inhibit mitotic exit, in: Progress in Cell Cycle Research, edited by L. Meijer, A. Jezequel, and M. Roberge, Life in Progress. Roscoff, France. 431–439. Chen, R. H. (2002). J. Cell Biol. 158, 487–496. Chen, R. H., Waters, J. C., Salmon, E. D., and Murray, A. W. (1996). Science 274, 242–246. Cleveland, D. W., Mao, Y., and Sullivan, K. F. (2003). Cell 112, 407–421. Cooke, C. A., Schaar, B., Yen, T. J., and Earnshaw, W. C. (1997). Chromosoma 106, 446–455. Ditchfield, C., Johnson, V. L., Tighe, A., Ellston, R., Haworth, C., Johnson, T., Mortlock, A., Keen, N., and Taylor, S. S. (2003). J. Cell Biol. 161, 267–280. Dobles, M., Liberal, V., Scott, M. L., Benezra, R., and Sorger, P. K. (2000). Cell 101, 635–645. Duesberg, P., and Rasnick, D. (2000). Cell Motil. Cytoskeleton 47, 81–107. Duesberg, P., Rasnick, D., Li, R., Winters, L., Rausch, C., and Hehlmann, R. (1999). Anticancer Res. 19, 4887–4906. Dujardin, D., Wacker, U. I., Moreau, A., Schroer, T. A., Rickard, J. E., and De Mey, J. R. (1998). J. Cell Biol. 141, 849–862. Fang, G. (2002). Mol. Biol. Cell 13, 755–766. Fang, G., Yu, H., and Kirschner, M. W. (1998). Genes Dev. 12, 1871–1883. Fero, M. L., Randel, E., Gurley, K. E., Roberts, J. M., and Kemp, C. J. (1998). Nature 396, 177–180.
2. The Role of Mitotic Checkpoint in Maintaining Genomic Stability
49
Fraschini, R., Beretta, A., Sironi, L., Musacchio, A., Lucchini, G., and Piatti, S. (2001). Embo J. 20, 6648–6659. Gemma, A., Hosoya, Y., Seike, M., Uematsu, K., Kurimoto, F., Hibino, S., Yoshimura, A., Shibuya, M., Kudoh, S., and Emi, M. (2001). Lung Cancer 32, 289–295. Gorbsky, G. J., Chen, R. H., and Murray, A. W. (1998). J. Cell Biol. 141, 1193–1205. Hall, M. C., Torres, M. P., Schroeder, G. K., and Borchers, C. H. (2003). J Biol Chem. 278, 16698–705. Harper, J. W., Burton, J. L., and Solomon, M. J. (2002). Genes Dev. 16, 2179–2206. Hauf, S., Cole, R. W., LaTerra, S., Zimmer, C., Schnapp, G., Walter, R., Heckel, A., Van Meel, J., Rieder, C. L., and Peters, J. M. (2003). J Cell Biol. 161, 281–94. Hempen, P. M., Kurpad, H., Calhoun, E. S., Abraham, S., and Kern, S. E. (2003). Hum. Mutat. 21, 445. Hernando, E., Orlow, I., Liberal, V., Nohales, G., Benezra, R., and Cordon-Cardo, C. (2001). Int. J. Cancer 95, 223–227. Hesterberg, T. W., and Barrett, J. C. (1985). Carcinogenesis 6, 473–475. Howell, B. J., Hoffman, D. B., Fang, G., Murray, A. W., and Salmon, E. D. (2000). J. Cell Biol. 150, 1233–1250. Howman, E. V., Fowler, K. J., Newson, A. J., Redward, S., MacDonald, A. C., Kalitsis, P., and Choo, K. H. (2000). Proc. Natl. Acad. Sci. USA 97, 1148–1153. Hoyt, M. A., Totis, L., and Roberts, B. T. (1991). Cell 66, 507–517. Hudson, D. F., Fowler, K. J., Earle, E., Saffery, R., Kalitsis, P., Trowell, H., Hill, J., Wreford, N. G., de Kretser, D. M., Cancilla, M. R., Howman, E., Hii, L., Cutts, S. M., Irvine, D. V., and Choo, K. H. (1998). J. Cell Biol. 141, 309–319. Hunt, P. A., Koehler, K. E., Susiarjo, M., Hodges, C. A., Ilagan, A., Voigt, R. C., Thomas, S., Thomas, B. F., and Hassold, T. J. (2003). Curr. Biol. 13, 546–553. Hwang, L. H., Lau, L. F., Smith, D. L., Mistrot, C. A., Hardwick, K. G., Hwang, E. S., Amon, A., and Murray, A. W. (1998). Science 279, 1041–1044. Imai, Y., Shiratori, Y., Kato, N., Inoue, T., and Omata, M. (1999). Jpn. J. Cancer Res. 90, 837–840. Irniger, S., Piatti, S., Michaelis, C., and Nasmyth, K. (1995). Cell 81, 269–278. Jablonski, S. A., Chan, G. K., Cooke, C. A., Earnshaw, W. C., and Yen, T. J. (1998). Chromosoma 107, 386–396. Jallepalli, P. V., and Lengauer, C. (2001). Nature Rev. Cancer 1, 109–117. Kalitsis, P., Earle, E., Fowler, K. J., and Choo, K. H. (2000). Genes Dev. 14, 2277–2282. Kalitsis, P., Fowler, K. J., Earle, E., Hill, J., and Choo, K. H. (1998). Proc. Natl. Acad. Sci. USA 95, 1136–1141. Kapoor, M., Montes de Oca Luna, R., Liu, G., Lozano, G., Cummings, C., Mancini, M., Ouspenski, I., Brinkley, B. R., and May, G. S. (1998). Chromosoma 107, 570–576. King, J. M., and Nicklas, R. B. (2000). J. Cell Sci. 113 Pt. 21, 3815–3823. King, R. W., Peters, J. M., Tugendreich, S., Rolfe, M., Hieter, P., and Kirschner, M. W. (1995). Cell 81, 279–288. Kitagawa, R., and Rose, A. M. (1999). Nat. Cell Biol. 1, 514–521. Laird, P. W. (2003). Nat. Rev. Cancer 3, 253–266. Li, R., and Murray, A. W. (1991). Cell 66, 519–531. Li, X., and Nicklas, R. B. (1995). Nature 373, 630–632. Li, X., and Nicklas, R. B. (1997). J. Cell Sci. 110 Pt. 5, 537–545. Li, Y., and Benezra, R. (1996). Science 274, 246–248. Lin, S. F., Lin, P. M., Yang, M. C., Liu, T. C., Chang, J. G., Sue, Y. C., and Chen, T. P. (2002). Leuk. Lymphoma 43, 385–391. Liu, S. T., Chan, G. K., Hittle, J. C., Fujii, G., Lees, E., and Yen, T. J. (2003a). Mol. Biol. Cell 14, 1638–1651.
50
Liu et al.
Liu, S. T., Hittle, J. C., Jablonski, S. A., Campbell, M. S., Yoda, K., and Yen, T. J. (2003b). Nat. Cell Biol. 5, 341–345. Luo, X., Tang, Z., Rizo, J., and Yu, H. (2002). Mol. Cell 9, 59–71. Maiato, H., Sampaio, P., Lemos, C. L., Findlay, J., Carmena, M., Earnshaw, W. C., and Sunkel, C. E. (2002). J. Cell Biol. 157, 749–760. Maney, T., Ginkel, L. M., Hunter, A. W., and Wordeman, L. (2000). Int. Rev. Cytol. 194, 67–131. Martin-Lluesma, S., Stucke, V. M., and Nigg, E. A. (2002). Science 297, 2267–2270. McEwen, B. F., Chan, G. K., Zubrowski, B., Savoian, M. S., Sauer, M. T., and Yen, T. J. (2001). Mol. Biol. Cell 12, 2776–2789. McEwen, B. F., Hsieh, C. E., Mattheyses, A. L., and Rieder, C. L. (1998). Chromosoma 107, 366–375. McIntosh, J. R. (1991). Cold Spring Harb. Symp. Quant. Biol. 56, 613–619. McIntosh, J. R., Grishchuk, E. L., and West, R. R. (2002). Annu. Rev. Cell. Dev. Biol. 18, 193–219. Mellone, B. G., and Allshire, R. C. (2003). Curr. Opin. Genet. Dev. 13, 191–198. Michel, L. S., Liberal, V., Chatterjee, A., Kirchwegger, R., Pasche, B., Gerald, W., Dobles, M., Sorger, P. K., Murty, V. V., and Benezra, R. (2001). Nature 409, 355–359. Millband, D. N., and Hardwick, K. G. (2002). Mol. Cell Biol. 22, 2728–2742. Mimori, K., Inoue, H., Alder, H., Ueo, H., Tanaka, Y., and Mori, M. (2001). Oncol. Rep. 8, 39–42. Musacchio, A., and Hardwick, K. G. (2002). Nat. Rev. Mol. Cell Biol. 3, 731–741. Nakagawa, H., Yokozaki, H., Oue, N., Sugiyama, M., Ishikawa, T., Tahara, E., and Yasui, W. (2002). Oncol. Rep. 9, 1229–1232. Nasmyth, K., Peters, J. M., and Uhlmann, F. (2000). Science 288, 1379–1385. Nicklas, R. B. (1997). Science 275, 632–637. Nicklas, R. B., and Ward, S. C. (1994). J. Cell Biol. 126, 1241–1253. Nicklas, R. B., Ward, S. C., and Gorbsky, G. J. (1995). J. Cell Biol. 130, 929–939. Nicklas, R. B., Waters, J. C., Salmon, E. D., and Ward, S. C. (2001). J. Cell Sci. 114, 4173–4183. Nomoto, S., Haruki, N., Takahashi, T., Masuda, A., Koshikawa, T., Fujii, Y., and Osada, H. (1999). Oncogene 18, 7180–7183. Olesen, S. H., Thykjaer, T., and Orntoft, T. F. (2001). Carcinogenesis 22, 813–815. Ouyang, B., Knauf, J. A., Ain, K., Nacev, B., and Fagin, J. A. (2002). Clin. Endocrinol. (Oxf) 56, 341–350. Percy, M. J., Myrie, K. A., Neeley, C. K., Azim, J. N., Ethier, S. P., and Petty, E. M. (2000). Genes Chromosomes Cancer 29, 356–362. Perez-Castro, A. V., Shamanski, F. L., Meneses, J. J., Lovato, T. L., Vogel, K. G., Moyzis, R. K., and Pedersen, R. (1998). Dev. Biol. 201, 135–143. Peters, J. M. (2002). Mol. Cell 9, 931–943. Pfarr, C. M., Coue, M., Grissom, P. M., Hays, T. S., Porter, M. E., and McIntosh, J. R. (1990). Nature 345, 263–265. Pines, J., and Rieder, C. L. (2001). Nat. Cell Biol. 3, E3–E6. Putkey, F. R., Cramer, T., Morphew, M. K., Silk, A. D., Johnson, R. S., McIntosh, J. R., and Cleveland, D. W. (2002). Dev. Cell 3, 351–365. Reis, R. M., Nakamura, M., Masuoka, J., Watanabe, T., Colella, S., Yonekawa, Y., Kleihues, P., and Ohgaki, H. (2001). Acta. Neuropathol. (Berl) 101, 297–304. Rieder, C. L. (1982). Int. Rev. Cytol. 79, 1–58. Rieder, C. L., Cole, R. W., Khodjakov, A., and Sluder, G. (1995). J. Cell Biol. 130, 941–948. Rieder, C. L., and Khodjakov, A. (2003). Science 300, 91–96. Rieder, C. L., and Salmon, E. D. (1998). Trends Cell Biol. 8, 310–318.
2. The Role of Mitotic Checkpoint in Maintaining Genomic Stability
51
Rieder, C. L., Schultz, A., Cole, R., and Sluder, G. (1994). J. Cell Biol. 127, 1301–1310. Sato, M., Sekido, Y., Horio, Y., Takahashi, M., Saito, H., Minna, J. D., Shimokata, K., and Hasegawa, Y. (2000). Jpn. J. Cancer Res. 91, 504–509. Scaerou, F., Starr, D. A., Piano, F., Papoulas, O., Karess, R. E., and Goldberg, M. L. (2001). J. Cell Sci. 114, 3103–3114. Schaar, B. T., Chan, G. K. T., Maddox, P., Salmon, E. D., and Yen, T. J. (1997). J. Cell Biol. 139, 1373–1382. Shah, J. V., and Cleveland, D. W. (2000). Cell 103, 997–1000. Shannon, K. B., Canman, J. C., and Salmon, E. D. (2002). Mol. Biol. Cell 13, 3706–3719. Shichiri, M., Yoshinaga, K., Hisatomi, H., Sugihara, K., and Hirata, Y. (2002). Cancer Res. 62, 13–17. Skoufias, D. A., Andreassen, P. R., Lacroix, F. B., Wilson, L., and Margolis, R. L. (2001). Proc. Natl. Acad. Sci. USA 98, 4492–4497. Steuer, E. R., Wordeman, L., Schroer, T. A., and Sheetz, M. P. (1990). Nature 345, 266–268. Sudakin, V., Chan, G. K., and Yen, T. J. (2001). J. Cell Biol. 154, 925–936. Sudakin, V., Ganoth, D., Dahan, A., Heller, H., Hershko, J., Luca, F. C., Ruderman, J. V., and Hershko, A. (1995). Mol. Biol. Cell 6, 185–197. Tanaka, T. U. (2002). Curr. Opin. Cell Biol. 14, 365–371. Tanaka, T. U., Rachidi, N., Janke, C., Pereira, G., Galova, M., Schiebel, E., Stark, M. J., and Nasmyth, K. (2002). Cell 108, 317–329. Tang, Z., Bharadwaj, R., Li, B., and Yu, H. (2001). Dev. Cell 1, 227–237. Taylor, S. S., and McKeon, F. (1997). Cell 89, 727–735. Tsukasaki, K., Miller, C. W., Greenspun, E., Eshaghian, S., Kawabata, H., Fujimoto, T., Tomonaga, M., Sawyers, C., Said, J. W., Koeffler, H. P. (2001). Oncogene 20, 3301–3305. Waters, J. C., Chen, R. H., Murray, A. W., and Salmon, E. D. (1998). J. Cell Biol. 141, 1181–1191. Weiss, E., and Winey, M. (1996). J. Cell Biol. 132, 111–123. Wordeman, L., and Mitchison, T. J. (1995). J. Cell Biol. 128, 95–104. Wu, H., Lan, Z., Li, W., Wu, S., Weinstein, J., Sakamoto, K. M., and Dai, W. (2000). Oncogene 19, 4557–4562. Yao, X., Abrieu, A., Zheng, Y., Sullivan, K. F., and Cleveland, D. W. (2000). Nat. Cell Biol. 2, 484–491. Yen, T. J., Compton, D. A., Wise, D., Zinkowski, R. P., Brinkley, B. R., Earnshaw, W. C., and Cleveland, D. W. (1991). Embo J. 10, 1245–1254. Yen, T. J., Li, G., Schaar, B. T., Szilak, I., and Cleveland, D. W. (1992). Nature 359, 536–539. Yu, H. (2002). Curr. Opin. Cell Biol. 14, 706–714. Zachariae, W. and Nasmyth, K. (1999). Genes Dev. 13, 2039–2058. Zhai, Y., Kronebusch, P. J., and Borisy, G. G. (1995). J. Cell Biol. 131, 721–734. Zhou, J., Panda, D., Landen, J. W., Wilson, L., and Joshi, H. C. (2002a). J. Biol. Chem. 277, 17200–17208. Zhou, J., Yao, J., and Joshi, H. C. (2002b). J. Cell Sci. 115, 3547–3555. Zirkle, R. E. (1970). Radiat. Res. 41, 516–537.
This Page Intentionally Left Blank
3 The Regulation of Oocyte Maturation Ekaterina Voronina and Gary M. Wessel* Department of Molecular and Cell Biology and Biochemistry, Brown University, 69 Brown St, Providence, RI 02912, USA
I. Concept of Oocyte Maturation II. Life and Death of an Oocyte A. Oocyte Origins B. Oocyte Development C. Transformations of the Oocyte During Meiotic Maturation D. Death III. Maturation Promoting Factor IV. Initiation of Oocyte Maturation A. Maturation-inducing Substance (MIS) B. Receptors for Maturation-inducing Substance C. Follicle Cells and Gap Junctions D. A Whole New Can of Worms V. Cytoplasmic Control of Maturation A. Signal Transduction through G-proteins: Ignition or Parking Brake? B. Calcium Ions C. cAMP D. PI3K E. MAPK VI. Concluding Remarks Acknowledgments References
Abbreviations 1-MA AC cADPr CaMKII DHP ER ES cells FF-MAS FSH GPCR
1-methyladenine adenylate cyclase cyclic ADP-ribose calcium/calmodulin-dependent kinase 17,20-dihydro-4-pregnen-3-one endoplasmic reticulum embryonic stem cells follicular fluid meiosis-activating sterol follicle-stimulating hormone G-protein coupled receptor
*To whom correspondence should be addressed. Tel.: (401) 863-1051; Fax: (401) 863-1182; E-mail: [email protected] Current Topics in Developmental Biology, Vol. 58 Copyright ß 2003, Elsevier, Inc. All rights reserved. 0070-2153/03 $35.00
53
54 GV GVBD hCG IBMX IP3 IP3R LH MI MII MAPK MIS MPF mPR MSP nAR nPR PDE PDK1 PGC PI3K PKA PKB PKC RPTK RyR SDF1 THP
Voronina and Wessel germinal vesicle (nucleus of the oocyte) germinal vesicle breakdown human chorionic gonadotropin (LH activity) 3-isobutyl-1-methylxantine inositol 1,4,5-triphosphate inositol 1,4,5-triphosphate receptor luteinizing hormone meiotic metaphase I meiotic metaphase II mitogen-activated protein kinase maturation-inducing substance maturation promoting factor membrane progesterone/progestin receptor major sperm protein of C. elegans nuclear androgen receptor nuclear progesterone receptor phosphodiesterase phosphoinositide-dependent kinase 1 primordial germ cell phosphatidylinositol 3-kinase protein kinase A protein kinase B protein kinase C receptor protein tyrosine kinase ryanodine receptor stromal cell-derived factor 1 17,20,21-trihydro-4-pregnen-3-one
I. Concept of Oocyte Maturation Fertilization is required for sexual reproduction, and the cells participating in this process are highly specialized. Both types of metazoan gametes— oocytes and sperm—arise from germ cells, undergo extensive differentiation, and eventually unite. Growth and maturation of both male and female gametes culminates in production of fertilization-competent eggs and sperm (Masui and Clarke, 1979). The acquisition of fertilization competence of the oocyte, the focus of this review, encompasses a number of critical events. For example, the oocyte must construct and store excessive numbers of Golgi, mitochondria, and ribosomes, make specific sperm receptors, and in oviparous animals the cell must accumulate nutritional reserves sufficient for embryonic development. The oocyte must also prepare mechanisms ensuring productive fertilization since aberrations in this process result in early death of the embryo. For instance, supernumerary sperm entering the egg bring along extra genetic material and centrosomes resulting in abnormal mitosis and death of the zygote. To avoid these harmful
3. The Regulation of Oocyte Maturation
55
consequences, the eggs have evolved several strategies, the prevailing theme of which is modification of their extracellular matrix upon fertilization to make themselves nonreceptive to additional sperm. Establishment of the permanent block to polyspermy employs exocytosis of specialized dense core secretory vesicles called cortical granules upon fertilization. Other strategies include a fast but transient membrane potential change in some marine animals that blocks additional sperm–egg fusion, and degeneration of supernumerary sperm in physiologically polyspermic yolky eggs, such as in birds. The final development of fertilization competence of an oocyte— responsible for making it a fertilizable egg—usually occurs during the process of meiosis, and is referred to as oocyte maturation1. Maturation is often segregated into nuclear and cytoplasmic events to delineate specific mechanisms and functions. Cytoplasmic maturation involves the cytoplasmic changes required to prepare the cell for fertilization, activation, and embryo development. In various model organisms, it includes acquisition by the female germ cell of competence to fuse with sperm, decondense sperm chromatin, form pronuclei, and prevent polyspermy. Nuclear maturation refers to the meiotic process of chromosomal reduction to a haploid content, so as to produce a diploid organism upon fusion with sperm. Although oocytes may be fertilized, such gamete fusions are usually nonproductive (Longo and Schuetz, 1982 and references therein). While sperm have completed their meiotic divisions prior to fertilization in every species studied, eggs are diverse in this respect. The competence to be fertilized and start zygotic development, which defines the egg, develops at different stages of meiotic division in different animal species. Fertilization can occur at the prophase stage of meiosis (clams, marine worms), at metaphase I (MI; some insects, starfish), or at metaphase II (MII; most mammals) of meiosis. The changes induced by insemination are commonly referred to as egg activation. The conversion of G2-arrested oocytes into actively dividing zygotes involves two major processes: maturation and activation. Notably, sea urchins belong to a limited group of organisms, wherein an oocyte completes meiotic maturation forming a haploid cell before it becomes fertilizable. Therefore, the processes of maturation and activation are fully separate in this animal in contrast to many others (such as starfish, frog, or mouse). Understanding how an immature oocyte transforms into an egg during oocyte maturation is critical for our knowledge of fertility and reproduction. The topic of oocyte maturation has been reviewed recently from multiple 1
In this review, the term ‘‘maturation’’ is used to define the completion of meiosis by the oocytes, and production of a fertilizable egg. This is in contrast to cases where the word ‘‘maturation’’ refers to the entire process of oocyte development, or oogenesis.
56
Voronina and Wessel
perspectives (Kishimoto, 1999; Matova and Cooley, 2001). This review integrates these perspectives and emphasizes the regulation of oocyte maturation by cell signaling networks.
II. Life and Death of an Oocyte A. Oocyte Origins Gametes develop from primordial germ cells (PGC) that are set aside during early embryogenesis. The development of the egg beginning from the formation of germ cell precursors is termed oogenesis. The driving factor for specification of cells as PGCs in a variety of animals appears to be inheritance of a defined area of egg cytoplasm, called germ plasm, containing a set of localized determinants. However, it is clear that this strategy is not universal, as in certain groups of organisms including mammals and sea urchins PGCs appear to be induced de novo from other cells in the gastrulating embryo (reviewed in Matova and Cooley, 2001; McLaren, 2003; Raz, 2002; Wylie, 1999). Furthermore, ascidian embryos appear to have an ability to regenerate germ cells following removal of PGCs (Takamura et al., 2002). In animals that utilize localized cytoplasmic determinants, germ plasm can be identified morphologically by the presence of specialized organelles, which are collectively called germ granules, or specifically known as P granules in C. elegans, polar granules in Drosophila, and germinal granules in Xenopus. Determination of primordial germ cells depends on localized RNAs and proteins, especially those comprising the germ granules. Oskar, vasa, and tudor proteins were first identified in Drosophila as components of polar granules and regulators of their formation. At least one of them, vasa, appears to have a universal role and is found consistently in the germ cells of a wide variety of animals (reviewed in Matova and Cooley, 2001; Raz, 2002; Takamura et al., 2002). In animals that rely on inductive events for primordial germ cell determination, such as mouse, key regulators of this process appear to include the newly-identified interferon-inducible transmembrane protein known as fragilis, and a nuclear protein of unknown function termed stella (Saitou et al., 2002). Coordinate with the upregulation of stella, fragilis, DAZ-like, Oct-4, and (later) vasa in primordial germ cells is a downregulation of certain homeobox genes (HoxB1, HoxA1, Lim1, and Evx1) and signaling molecules like Smad1 (McLaren, 2003). Investigators are currently attempting to determine the pathways and targets for each of these candidates by using targeted gene manipulation, microarray analysis, and other techniques.
3. The Regulation of Oocyte Maturation
57
The difference between the major mechanisms for establishing the germ line in animals (i.e., cytoplasmic localized determinants vs induction) is extreme even in closely related animals (e.g., frogs and newts). Johnson et al. (2003) have proposed that the regulative mode of germ cell specification is the primitive form, from which predetermined development by localized determinants has evolved. With this interpretation it is also easier to understand how multiple independent mechanisms may have evolved in various animal lineages (McLaren, 2003). A hallmark of germ cells is an extensive migration from the place of their formation to the developing gonad, combined with elimination of mistargeted cells elsewhere. This migration is regulated by somatic-germ cell interactions, and several molecular participants of this process have been recently identified, which appear to be distinct between different groups of organisms. For example, in Drosophila, enzymes implicated in lipid metabolism, wunen and columbus, affect PGC migration by generating repulsive and attractive signals, respectively (reviewed in Matova and Cooley, 2001). Two wunen genes are homologues of mammalian phospholipid phosphatase type 2; they encode transmembrane proteins, whose catalytic activity is required for generation of the repellant effect (Starz-Gaiano et al., 2001). Columbus (or Hmgcr) is a homologue of 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase that generates the attractant for germ cells by an unidentified mechanism. Additional genes required for germ cell migration in Drosophila act by uncharacterized mechanisms and include slam (Stein et al., 2002) and scattershot (Coffman et al., 2002). In mouse, the signaling mediated by c-kit/ Steel Factor is particularly important for germ cell migration, with additional input from integrin 2 and Fgf8 (reviewed in Matova and Cooley, 2001; McLaren, 2003). Finally, in zebrafish PGC guidance is regulated through seven transmembrane G-protein coupled receptor CXCR4 [- (CXC-) chemokine receptor type 4] expressed in the PGCs and its ligand SDF-1 chemokine (stromal cell-derived factor 1) expressed along the path of PGC migration. These molecules establish a system for the guidance of fish PGCs throughout their entire migration (Doitsidou et al., 2002; Knaut et al., 2003; Kunwar and Lehmann, 2003). Once formed, the identity and developmental potential of the germline is maintained throughout the animal’s lifetime.
B. Oocyte Development Once the gonad is assembled from somatic and germ cells, populating germ cells differentiate and proliferate. Initially, PGCs have the potential to begin either spermatogenesis or oogenesis and this decision is directed by the
58
Voronina and Wessel
gonadal environment. In a female genital ridge, or in a nongonadal environment, PGCs differentiate as oocytes, while male gonadal somatic cells direct PGCs to a spermatogenic fate (reviewed in Adams and McLaren, 2002). In many organisms, oogonia divide several times in the gonad to form clusters of interconnected cells, often referred to as cysts. The cytoplasmic bridges left after each division form the connections between cells in such clusters allowing for continuous communication between cells and coordinated development. Alternatively, when the interconnected cells form ‘‘oocyte-nurse cell’’ complexes they are able to more effectively utilize biosynthetic capabilities of multiple cells to produce maternal stores of important macromolecules in the resulting egg. The best documented example of the syncytial development of oogonia is in Drosophila. In this animal, a single cyst consists of 16 cells, one of which differentiates into an oocyte, while the remaining 15 assume the role of nurse cells that support the oocyte’s growth and differentiation throughout the rest of the oocyte’s growth period. Interconnected cells are also found in C. elegans, Xenopus, zebrafish, and different mammals (reviewed in Matova and Cooley, 2001). In contrast to Drosophila, these cysts eventually break down, such that every cell in an individual cyst of vertebrate animals has the potential to become an oocyte. Proliferating germ cells are referred to as oogonia or spermatogonia, and are pluripotent stem cells normally restricted to germ line differentiation (Matova and Cooley, 2001). Oogonia differentiate into primary oocytes when they begin meiosis by replicating their DNA, and arresting in prophase of first meiosis. In the ovary, transition from a mitotic to a meiotic program is regulated by the signals from somatic cells (Chuma and Nakatsuji, 2001; Seydoux and Schedl, 2001). The prophase oocyte then may spend various periods of time in this state. In the human female for example, primary oocytes are arrested from 12 to 50 years, while in frog the arrest lasts for 3 years, and most echinoderm primary oocytes are arrested for up to a year depending on the species. Somatic cells of the gonad then stimulate the primary oocyte to continue development. In most animals, the growing oocyte contains an enlarged nucleus, called a germinal vesicle (GV), and is active in transcribing an extensive array of genes whose products are necessary for oocyte development and for sustaining early embryonic development. During periods of very active RNA synthesis, especially in animals with large oocytes, the GV contains lampbrush chromosomes, with extended DNA loops at the sites of RNA synthesis (Smith and Richter, 1985). During oogenesis, the oocyte accumulates an extensive collection of RNAs, proteins, and organelles, such as cortical granules, yolk vesicles, ribosomes, and mitochondria (Wessel et al., 2001). Female germ cells interact with the somatic gonadal cells throughout their life. In some animals, such as mammals, somatic cells of the gonad closely
3. The Regulation of Oocyte Maturation
59
Figure 1 Schematic representation of a late mammalian follicle (after Wright et al., 2001). The cell types comprising the follicle are shown (the oocyte and somatic granulosa cells). Also pointed out are the oocyte’s nucleus (germinal vesicle), extracellular matrix produced by the oocyte (zona pellucida), antrum, and the granulosar membrane of the follicle.
associate with growing oocytes to form follicles (Fig. 1), while in other animals, such as sea cucumber, individual somatic cells establish junctions with the oocyte surface without forming complete follicles. Follicles provide a specialized compartment for the growth of the oocyte, accumulate and supply nutrients for oocyte’s growth, and generate signals governing oocyte development. Formation of the follicle in many species is essential for oocyte growth and development of a mature egg, suggesting that somatic cells provide both nutritional support and developmental information for the growing gamete. Follicle cells are intimately associated with the oocyte through numerous interdigitating processes during the growth phase and the entire follicle is coupled with gap junctions (reviewed in Matova and Cooley, 2001). In the mouse ovary, gap junctions form in advanced (secondary) follicles, but are not detected in early (primordial) follicles (Wright et al., 2001). In some reptile species (e.g., the lizard Anolis carolinensis) cytoplasmic connections may arise between an oocyte and follicle cells and form intercellular bridges (Neaves, 1971). Connections between the oocyte and somatic cells allow nutrient transfer to the developing oocyte and are essential for oocyte growth in mammals (reviewed in Carabatsos et al., 2000; Eppig et al., 1996). The oocyte loses direct physical contact with associated follicle cells just prior to the onset of maturation. Bidirectional cell–cell communications coordinate the development of oocyte and follicle cells; these interactions take the form of autocrine, paracrine, and endocrine regulation in addition to cytoplasmic gap junctional contact mentioned above. The oocyte appears to be the dominant component of the follicle, determining the overall rate of follicular
60
Voronina and Wessel
development (Eppig et al., 2002; Matzuk et al., 2002). A critical role of the oocyte was established in experiments involving the exchange of somatic and germ cell components of the follicles at different stages of development. Mouse oocytes dissociated from the midgrowth secondary follicles (12 days old) were reaggregated with the somatic cells of newborn mice and surgically implanted beneath the renal capsules of host females. Such reimplanted oocytes dramatically accelerated follicular development, resulting in formation of antral follicles with differentiated follicle cells nine days after reimplantation. This was significantly faster than normal development of antral follicles in the newborn mice, which takes 18–24 days. Thus, the oocytes appear to provide instructive signal that is able to direct the development of the somatic component of the follicle. This experimental setup provides a valuable tool for dissecting the signaling mechanisms involved in the crosstalk between the oocyte and somatic cells of the follicle during oogenesis in mammals, and perhaps other animals as well. Embryonic stems cells are now known to differentiate into oocytes and this technology may greatly enhance our ability to experiment with the germ cell differentiation pathway. Recent work from Hubner et al. (2003) showed that ES cells can acquire the morphology and gene expression profile consistent with differentiation into oocytes. While the pluripotential nature of ES cells has been appreciated for some time, the possibility of germ line differentiation in vitro was unexpected. This result has massive technical and ethical consequences. Key for this observation was use of a GFP reporter system driven by a promoter specifically active in oocytes that drives the transcription factor Oct4. ES cells in culture could then be examined noninvasively and followed over the culture period. The investigators found that Oct4-GFP was expressed in a subpopulation of the ES cells after a few days in culture. In fact, several different types of markers accumulated in the oocyte-like ES cells. These included the zona pellucida proteins ZP2 and ZP3, the transcription factor of germ cells Fig, a TGF- growth factor GDF-9, and some markers of meiosis (DCM1, SCP3). Each of the markers is consistent with the ES cell becoming an oocyte, and in concert, the population of markers tested give strong support to an oocyte differentiation pathway. Coincident with this marker expression was the morphological transitions the oocyte-like cells underwent. This includes recruiting other ES cells into a follicle-like structure (able to express several markers of estrogen biosynthesis), displacement of the oocyte from the recruited cluster that resembled oocyte ovulation, and eventually even formation of structures similar to embryonic blastocysts. These findings will have great potential in dissecting mechanistic processes in oogenesis in particular, and gametogenesis in general.
3. The Regulation of Oocyte Maturation
61
C. Transformations of the Oocyte During Meiotic Maturation 1. Nuclear Changes and Chronology of Maturation The process of oocyte maturation traditionally has been described by changes in chromosome morphology during meiosis. Before maturation starts, the oocyte contains a large germinal vesicle (GV) with a large nucleolus (Fig. 2A–C). The chromosomes in the GV are mostly decondensed, dispersed, and transcriptionally active (reviewed in Smith and Richter, 1985; Wassarman, 1983). With the initiation of maturation, transcription ceases, the chromosomes begin to condense, the GV breaks down, and nucleoli disperse (Masui and Clarke, 1979). As maturation progresses, the paired homologous chromosomes align in the middle of the
Figure 2 Events of nuclear maturation, exemplified by sea urchin oocytes. A–C: Oocytes spend their growth period arrested in G2 phase of meiotic cell cycle, with replicated DNA (4N). After resumption of meiosis, oocytes of different species arrest for the second time at different stages of meiotic progression, such as MI (D–F), or MII (G), and are relieved from this second arrest by fertilization. Then they finish their meiotic divisions and progress into mitosis. Sea urchin oocytes are fertilized as haploid eggs after completing meiotic divisions (H–J). A, D, G, H: schematic of nuclear events of meiosis, ploidies of the nuclei are noted; B, E, I: brightfield images of sea urchin oocytes undergoing meiosis; C, F, J: corresponding fluorescent images of DNA stained with Hoechst. PB1—the first polar body; PB2—the second polar body; pn—egg pronucleus; MI—metaphase I stage of meiosis; MII—metaphase II stage of meiosis.
62
Voronina and Wessel
forming meiotic spindle during metaphase I (Fig. 2D–F). Separation of the paired homologous chromosomes is followed by the first polar body formation. Then the chromosomes remaining in the oocyte are again arranged on a meiotic spindle at metaphase II. With the second meiotic division, chromatids separate and the second polar body is formed. Finally, the chromatids remaining in the oocyte decondense and a pronucleus forms (Fig. 2H–J). The meiotic cells of different organisms become fertilizable at different stages along the process of meiosis. For example, the eggs of most vertebrate species arrest at the second metaphase, and await fertilization to release them into embryonic development, while sea urchin eggs proceed through meiosis to completion, and are stored as cells with haploid pronuclei. How long does oocyte maturation take? The dynamics of oocyte maturation is difficult to assess for a number of reasons. First, exact observation of cellular features during in vitro oocyte maturation may not faithfully reproduce in vivo processes. Second, the progression of maturation is often not synchronous in individual members of an oocyte population, especially when the oocytes are obtained from females in different phases of their seasonal reproductive cycle (Masui and Clarke, 1979). Therefore, the timing can only be defined statistically. Finally, inaccuracy can be introduced when determining the exact time of initiation of oocyte maturation. While in certain species, the time of oocyte exposure to a hormone is known precisely (as in Xenopus or starfish for example), the oocytes of other species are able to mature in vitro without any external stimuli. Mammals and sea urchins belong to the latter group of animals, and therefore at this time one can only rely on obvious structural changes indicating the onset of maturation. In various species, oocyte maturation proceeds at different rates, even in members of the same phylogenetic class (Masui and Clarke, 1979). For instance, oocyte maturation in starfish takes only 1 h 40 min from GVBD to completion (Kishimoto, 1999), whereas in the sea urchin, oocyte maturation from the first detected migration of GV to the cell surface and GVBD to formation of egg pronucleus takes approximately 8 h in vitro (Berg and Wessel, 1997). The extended time it takes for the sea urchin oocyte to fully mature may be explained in part by the fact that this cell needs to complete meiosis and reform a haploid pronucleus, whereas other oocytes abbreviate their maturation by arresting at MI or MII (it takes sea urchin 4 h for the first polar body to form, so MI probably occurs at 3 h after GVBD). Other representative durations of oocyte maturation are, approximately, on average 5–6 h for fish (but up to 18 h in certain species, Thomas et al., 2002), 3 h for Xenopus (Terasaki et al., 2001), 6–7 h for mouse, rat, and hamster, and 15–20 h for human (both in vivo and in vitro, Bomsel-Helmreich et al., 1987). However, it is possible that spontaneous maturation in vitro takes
3. The Regulation of Oocyte Maturation
63
longer to proceed than the hormone-induced maturation in vivo. In a study of brushtail possum oocytes (Glazier et al., 2002), the spontaneous in vitro maturation of the oocytes took more than 24 h. However, when the oocytes were allowed to mature in vivo and recovered at various times post gonadotropin treatment for observation, intermediate stages of meiotic maturation were not observed at all, suggesting that it is an extremely rapid process. 2. Changes of the Organelles The process of oocyte maturation is accompanied by fundamental changes in cellular organelles. Oocytes possess a variety of organelles typical of most cells (such as Golgi apparatus, mitochondria, endoplasmic reticulum) as well as oocyte-specific (yolk granules, cortical granules, annulate lamellae, acidic granules, and pigment granules). Some of these organelles do not undergo any transformations during maturation (mitochondria, yolk), while others (such as cytoskeletal elements, cortical granules, and endoplasmic reticulum) exhibit dramatic changes—and are hallmarks of meiotic progression. The orchestration and execution of the nuclear events of meiosis described above depends on the cell’s cytoskeleton. The growing oocyte is characterized by elongated, labile cortical microtubules that undergo dramatic changes during oocyte maturation (Boyle and Ernst, 1989; Smiley, 1990). Microtubules progressively disassemble in the cortex, and the meiotic microtubule spindle then forms in the maturing oocyte to mediate chromosome segregation during both divisions of meiosis (a separate spindle forms in each division). Meiotic microtubule asters contributing to the spindle formation are morphologically distinct from the mitotic ones, being smaller, and not reaching distal aspects of the oocyte. In the mature sea urchin egg, microtubules can still be found at the cortex, although they are shorter and less numerous than in oocytes (Boyle and Ernst, 1989). During meiosis, eggs of most species (the most notable exception being mice) lose their microtubule organizing center, or centrosome, so that they rely on fertilization and the parental centrosome to be able to execute subsequent cytokinesis (see below). This process is particularly well documented in starfish oocytes. The starfish oocyte begins meiosis with two centrosomes, as do somatic cells in a G2 phase of the cell cycle. During the first meiotic division the cell loses one of the centrosomes to the first polar body, which exactly mirrors the processes occurring during a mitotic division. The remaining centrosome does not replicate between two meiotic divisions (as it would have in mitosis). Instead, in an event unique to meiosis, the two individual centrioles of the remaining centrosome separate and organize the second meiotic
64
Voronina and Wessel
spindle. After the second meiotic division, one centriole is segregated into the second polar body, and the egg is left with a single centriole (Schatten, 1994 and references therein). It appears that these centrioles are not equal, as the second polar body centriole retains the ability to replicate during the S-phase and can support mitotic divisions, while the egg centriole is not functional (Tamura and Nemoto, 2001). This single centriole is targeted for degradation during the anaphase of the second meiotic division, as detected by the loss of focal -tubulin localization shortly after completion of meiosis (Uetake et al., 2002). In contrast, the mouse oocyte has a number of centrosomal foci (instead of well-defined centrosomes), some of which aggregate at the poles of the meiotic spindles, and others remain spread throughout the cytoplasm, and evade elimination during meiotic divisions (reviewed in Schatten and Schatten, 1987). The cortical granules accumulated in the oocytes acquire the capacity for exocytosis during maturation. Their function is to secrete their contents at fertilization and modify the extracellular matrix of the egg, such that it does not support binding and fusion of extra sperm. It is critical for the cell to rapidly exocytose its cortical granules at fertilization to block polyspermy, but also to prevent precocious exocytosis of these cortical granules, otherwise the extracellular matrix would be irreversibly modified and the cell would not be receptive to sperm. The oocyte synthesizes and accumulates large numbers of cortical granules throughout oogenesis. In oocytes where cortical granules have been quantified, they reach 8,000 in mice and 15,000 in sea urchin (reviewed in Wessel et al., 2001). During oocyte maturation, cortical granules move to the periphery in nearly all species examined. This change in distribution is linked to the acquisition of the capacity of these granules to exocytose upon fertilization. In the sea urchin, these granules are dispersed throughout the oocytes’ cytoplasm, and translocate to the periphery of the oocyte during maturation (reviewed in Wessel et al., 2001). In other animals, such as starfish and frog, the cortical granules are already in the general vicinity of the cell cortex, and are brought to the plasma membrane and rearranged at the cell surface during oocyte maturation. In rodents, oocytes form a cortical granule-free domain over the region of meiotic spindle mainly by rearrangement, with some contribution of precocious exocytosis (Deng et al., 2003). Cortical granule rearrangements each depend on de novo formation of microfilament networks at maturation. Microfilaments form the scaffold for organelle movement during oocyte maturation. In the growing echinoderm oocyte, microfilaments are predominantly found at the cortex of the cell (Boyle and Ernst, 1989; Heil-Chapdelaine and Otto, 1996). In starfish and sea urchin oocytes, dramatic polymerization of actin takes place at the onset of maturation (Heil-Chapdelaine and Otto, 1996; Wessel et al., 2002, and
3. The Regulation of Oocyte Maturation
65
references therein). Actin polymerization is detected at the cell cortex (where actin spikes appear transiently in starfish); however, most of the polymerized actin is detected in the nucleus. This reorganization of the actin cytoskeleton is not critical for germinal vesicle breakdown, as cytochalasins suppressing all actin polymerization do not inhibit GVBD (e.g., Connors et al., 1998). Actin filaments in the germinal vesicle have been seen in many different oocytes, including frog, starfish, and sea urchin (Heil-Chapdelaine and Otto, 1996; Parfenov et al., 1995; Wessel et al., 2002). Although the function of the filaments is not clear, it is intriguing for several reasons: (i) they are transient, present just before GVBD and then disappear; (ii) their presence is coincident with initiation of cortical granule translocation; (iii) they are found at a time when the germinal vesicle is moving to the animal pole; (iv) and they are correlated with a change in the germinal vesicle shape and pending GVBD (Stricker and Schatten, 1989). One hypothesis of function is that they contribute to vesiculation of the nuclear envelope (Parfenov et al., 1995). In addition, several regulators of actin polymerization are also present in the nucleus or translocate to the nucleus (Rando et al., 2000), suggesting that there is some function of actin in the nucleus that is subject to regulation. Finally, Zhao et al. (1998) find actin and actin-related proteins in transcriptional complexes of mammalian cells. This has also been seen in yeast, with genetic evidence supporting actin function both in regulating transcription and chromatin remodeling (Shen et al., 2000). In vertebrate oocytes, actin polymerization appears to be required for the translocation of the meiotic spindle from the center of the cell to an asymmetric cortical location, and as a consequence for the first polar body extrusion (Connors et al., 1998; Kim et al., 2000; Maro and Verlhac, 2002; Ryabova et al., 1986). Recent work identifies the microfilament-binding protein formin as a necessary molecular regulator of this process, as the spindle does not migrate and polar bodies are not extruded in oocytes from formin knockout mice (Leader et al., 2002). Formins, or formin-homology (FH) proteins, comprise a recently recognized protein family, which functions in regulating cell polarity by acting as effectors of Rho family small GTP-ases and remodeling actin cytoskeleton (reviewed in Alberts, 2002; Lew, 2002; Zeller et al., 1999). The other transport event that actin microfilaments are required for is translocation of cortical granules in multiple organisms, such as sea urchin, starfish, and mouse (reviewed in Wessel et al., 2001, 2002). The scaffold formed by an extensive network of intermediate filaments containing cytokeratin is present within 1 m of the plasma membrane in immature oocytes of the hamster, Xenopus, ascidians, sea urchin, and starfish (Boyle and Ernst, 1989; Schroeder and Otto, 1991, and references therein).
66
Voronina and Wessel
Upon commencement of meiotic maturation (at GVBD) in these animals, the intermediate fibers disassemble, and never reform in the mature eggs (Boyle and Ernst, 1989; Schroeder and Otto, 1991). Interestingly, an association of signaling proteins (beta-gamma subunits of heterotrimeric G-proteins) with these cytokeratin networks has been detected in starfish (Chiba et al., 1995). The subcortical network of intermediate filaments might serve as a scaffold for the assembly of the signaling protein complexes, and thus be important for signal transduction in oogenesis. The release of G-protein subunits is necessary for starfish oocyte maturation to occur (see Section VA). The timing of this release is correlated with the dispersal of intermediate filaments, so the detected structural reorganization of intermediate filaments can be indicative of the release of the cell signaling mediators at maturation. The endoplasmic reticulum (ER) of many species such as sea urchin, starfish, frog, and mouse undergoes significant changes during oocyte maturation. The changes in the ER structure are of particular interest as this organelle releases calcium at fertilization mediating egg activation, and this ability to release calcium develops during oocyte maturation. Immature oocytes of all species studied possess relatively uniform threedimensional network of ER tubules with some individual cisternae and annulate lamellae (accumulations of cisternae) deep in the cytoplasm (Bobinnec et al., 2003; Jaffe and Terasaki, 1994; Mehlmann et al., 1995; Terasaki et al., 2001). The detected changes of the ER in the course of oocyte maturation include reorganization, or formation of circular structures around the yolk platelets in starfish (Henson et al., 1990; Jaffe and Terasaki, 1994; Mehlmann et al., 1995; Terasaki et al., 2001) and accumulation of ER clusters in an organized array immediately next to the plasma membrane (Henson et al., 1990; Mehlmann et al., 1995; Shiraishi et al., 1995; Terasaki et al., 2001). Changes in the structure of the ER, especially dispersion of the nuclear envelope, and fragmentation of ER tubules, are due to the cell cycle progression, and have been documented for mitosis as well (reviewed in Jaffe and Terasaki, 1994; Lippincott-Schwartz, 2002). However, several features of ER behavior are unique for meiosis. In maturing oocytes, the ER is not associated with the meiotic spindle, while such association is detected in mitotic cells (Bobinnec et al., 2003; Terasaki, 2000). Furthermore, the formation of cortical clusters is specific for oocyte maturation and is ultimately required for formation of calcium release mechanisms in the egg at fertilization. These clusters disappear some time after fertilization and are not detected in mitotic embryonic cells (FitzHarris et al., 2003). In mouse and probably other animals too, the disappearance of cortical ER clusters depends on the decrease of MPF activity (FitzHarris et al., 2003 and references therein).
3. The Regulation of Oocyte Maturation
67
3. Changes in mRNA and Protein Patterns Oocytes are actively engaged in transcription and through their growth accumulate extensive amounts of mRNA (Masui and Clarke, 1979; Smith and Richter, 1985). However, the mRNA content of the oocyte changes upon maturation. This switch in mRNA populations is brought about by both general termination of transcription at GVBD, and degradation of the select subset of transcripts (Smith and Richter, 1985). In sea urchin oocytes, the mRNAs coding for yolk and cortical granule constituents each abruptly disappear with the beginning of oocyte maturation (LaFleur et al., 1998; Laidlaw and Wessel, 1994; Wessel et al., 1998, 2000a,b). Generally, oocyte maturation does not require mRNA synthesis; the only exception found so far is sheep, where oocyte maturation is inhibited by the transcription inhibitor -amanitin (Moor and Crosby, 1986). In most cases, mature eggs are not transcriptionally active, as they are arrested in the metaphase of the second meiotic division. The sea urchin egg, however, is an exception, in that it is arrested before fertilization with a haploid pronucleus, which is transcriptionally active and selectively accumulates histone mRNA before fertilization (Venezky et al., 1981). Protein synthesis patterns change significantly during the transition from an oocyte to a mature egg. Certain mRNAs are translationally activated while others become repressed (reviewed in Hake and Richter, 1997; Smith and Richter, 1985). Translational repression results in part from the degradation of mRNAs, but also from selective mRNA deadenylation. In sea urchins for example, production of yolk (yp30, Wessel et al., 2000b) and cortical granule content (Laidlaw and Wessel, 1994) proteins is extremely active in the primary oocyte, but ceases at the beginning of oocyte maturation, due to mRNA degradation. In Xenopus and mouse, a specific class of maternal mRNAs is deadenylated and translationally repressed during oocyte maturation (Paynton and Bachvarova, 1994; Varnum and Wormington, 1990). Translational activation of select mRNAs during oocyte maturation is achieved by regulated elongation of their poly(A) tail (Hake and Richter, 1997), and normally leads to a several-fold increase in the rate of overall protein synthesis (Smith and Richter, 1985; Wasserman et al., 1986). Qualitative changes in the patterns of protein synthesis have been observed in many animals such as starfish, frog, and mouse. The proteins being made during oocyte maturation are mostly cell cycle regulators needed to advance the oocyte through meiosis. The prime example of such proteins is cyclin B, the regulatory component of maturation promoting factor; another one is c-mos required for two consecutive divisions of meiosis. Consistently, protein synthesis inhibitors block oocyte maturation in many (but not all) species. Some animals appear to have all the proteins necessary for maturation already produced and
68
Voronina and Wessel
stored in the oocyte (like the clam, Spisula, Hunt et al., 1992). Others need to synthesize cyclin proteins (mouse, fish); still others need to make different molecular cell cycle regulators, such as ringo/speedy (Xenopus, Ferby et al., 1999; Hochegger et al., 2001; Lenormand et al., 1999). The mature egg of the sea urchin is translationally quiescent, while in other species such as Xenopus or mouse the mature egg is translationally active. This may be due to the fact that these eggs are at different phases of their meiotic cycle: frog and mouse eggs are arrested at MII and are constantly replenishing their mitotic cyclin proteins, while the sea urchin egg is arrested following MII, at a ‘‘G1-like’’ state.
4. Physiological Changes Mature eggs produce the fast block to polyspermy by changes in plasma membrane potential upon fertilization (collectively termed as ‘‘fertilization potentials’’), which transiently depolarizes in ascidians and sea urchins, and hyperpolarizes in certain mammalian species (reviewed in Dale, 1994). The egg’s ability to generate this fertilization potential develops during oocyte maturation. The mature eggs of starfish as well as bovine produce fertilization currents with greater peak amplitudes and faster rise times than do immature oocytes (Miyazaki, 1979; Tosti et al., 2002). These differences reflect the changes in ionic conductance that occur during oocyte maturation, and ultimately contribute to the establishment of the fast block to polyspermy. Thus, a decrease in potassium conductance by downregulation of Na þ /K þ pumps was proposed to occur in both starfish and Xenopus oocytes upon exposure to maturationinducing hormones (reviewed in Miyazaki, 1979; Smiley, 1990; Wasserman et al., 1986). Changes in sodium currents in Xenopus oocyte ultimately lead to an increase in the intracellular pH after progesterone stimulation from 7.2 to 7.7. The pH change occurs due to the activity of Na þ /H þ pump in the plasma membrane of the cell (Wasserman et al., 1986). This increase in pH appears to be important for the progress of maturation, as stabilizing the pH below 7.0 completely blocks GVBD. This increase in pH in the frog oocyte is dependent upon downregulation of Na þ /K þ pumps, as a high cytoplasmic concentration of potassium may directly inhibit the Na þ /H þ pump.
5. Competence for Calcium Release The capacity of the mature egg to undergo activation depends on the competence to generate a calcium transient (or transients) upon fusion with
3. The Regulation of Oocyte Maturation
69
sperm, which develops during oocyte maturation. Prophase-arrested mouse oocytes do not generate the series of calcium transients at fertilization characteristic of mature eggs (Cheung et al., 2000 and references therein). In mouse and starfish oocytes, inositoltriphosphate (IP3) receptor amounts increase during maturation (Chiba et al., 1990; Iwasaki et al., 2002), the IP3 receptor is rearranged at the cortex of the oocytes (Mehlmann et al., 1996), the endoplasmic reticulum is reorganized (reviewed in Section IIC), and intracellular calcium stores increase (Tombes et al., 1992). The maturation-associated increase in IP3 receptor protein amount results in increased sensitivity to IP3. The increase in sensitivity in turn confers to the cell the ability to generate a normal pattern of calcium oscillations and undergo normal amounts of cortical granule exocytosis (Xu et al., 2003).
D. Death The oocyte is a terminally differentiated cell. If the oocyte does not transform into a mature egg, or if the egg is not fertilized, it dies. The demise of the germ cells throughout the animal kingdom occurs through apoptosis, or programmed cell death. Apoptosis is a conserved process of cell elimination, characterized by distinct morphological hallmarks, and requiring activation of specific intracellular programs. Examples of physiological cell death required for oogenesis include C. elegans oocytes and Drosophila nurse cells that die donating their cytoplasm to the surviving gametes (it is still unclear what prevents the spreading of death signal in these syncytial cells to the surviving oocytes), or oocyte attrition and follicular atresia in mammals. The environmental conditions such as starvation or toxic substance exposure are known to cause germ cell apoptosis as well (Matova and Cooley, 2001; Voronina and Wessel, 2001). Apoptosis induction during starvation may be a nutritional strategy in some animals with significant resources invested into egg production enabling the animals to make many eggs during the time of plenty, but allowing the option to reuse these nutrients in the time of pity. This is especially true for yolk-laden oocytes where the nutritional stockpile is great and reusable. In many species, mature eggs that have not been fertilized age rapidly, and apoptose (Morita and Tilly, 1999; Sasaki and Chiba, 2001; Yuce and Sadler, 2001). This could represent the ‘‘quality control’’ mechanism in the animals that ovulate their eggs arrested at the second metaphase. Such eggs could potentially damage meiotic spindles and missegregate their chromosomes upon fertilization or even activate parthenogenetically if allowed to remain in such state for a long time, and exhaust their energy supplies.
70
Voronina and Wessel
III. Maturation Promoting Factor The cytoplasmic signaling pathways initiating the G2/M transition during meiotic resumption of the oocytes converge on activation of the maturationpromoting factor, or MPF (Taieb et al., 1997). MPF was identified in and subsequently purified from mature eggs of Xenopus (Lohka et al., 1988; Masui and Markert, 1971). It consists of cyclin B (or B1 and B2 in vertebrates) and cdk1 kinase, also referred to as cdc2 (Fig. 3; Labbe et al., 1989; Kobayashi, 1991, reviewed in: Maller, 1990). This complex is a universal cell cycle regulator advancing G2/M transition in both meiotic and mitotic cell cycles. Cyclins and their catalytic kinase partners (cdks, or cyclin-dependent kinases) are well recognized as regulators of general cell cycle progression (for example, reviewed in: Pines, 1999). Cyclin B is the regulatory subunit of the complex: cdk1 kinase is structurally prevented from activation as a monomer, and needs to complex with cyclin for activity (Morgan, 1997). Cyclin-dependent kinases preferentially complex with particular cyclins, and are required only at specific times during mitotic progression. Although the levels of cdks remain constant, the levels of cyclins usually fluctuate during the cell cycle due to periodic synthesis and degradation, resulting in transient kinase activities. Cyclins are classified as G1, S, or M-phase regulators based on when their activities are required in the cell cycle (Pines, 1999). The levels of newly synthesized mitotic cyclins peak in M-phase, and mitotic cyclin/cdk1 complexes govern the transition to and progression through M-phase (reviewed in Pines, 1999). Cyclin B-cdk1 phosphorylates and regulates the function of numerous substrates, including nuclear lamins, histone H1, transcription factors, kinesin-related motors, and cytoskeletal regulators, causing structural changes in the nuclear envelope, chromatin, and cytoskeleton (reviewed in Nigg, 1995, 2001). The cdk1 itself is a subject of tight posttranslational regulation (reviewed in Nigg, 2001). In addition to the required association with a cyclin, cdk1 can be phosphorylated on multiple sites. These phosphorylation events include both stimulatory (on Thr 161) and inhibitory (on Thr 14 and Tyr 15, localized in the ATP-binding site) regulation. The kinases responsible for inhibitory phosphorylations, Wee1 and Myt1, are active during the G2 phase, and their enzymatic activity is in turn regulated by cell signaling pathways as well as by checkpoint controls (although only Myt1 is present in oocytes). The enzyme removing inhibitory phosphates from cdk1 is a dual-specificity phosphatase cdc25. The balance of cdc25 vs Wee1/Myt1 activities integrates the checkpoint and cell signaling inputs to determine the activation state on cyclin B-cdk1 complex and allow progression to the M-phase (Nigg, 2001).
3. The Regulation of Oocyte Maturation
71
Figure 3 Regulation of cdk1 activation. Activation of cdk1 depends on association with cyclin B and the phosphorylation of cdk1 protein within the cdk1-cyclin B complex by cdk activating kinase, CAK. Protein phosphatase 2C (PP2C) can reverse the CAK-mediated phosphorylation. Active cdk1-cyclin B can be inactivated by the Myt1 protein kinase present in the oocyte, whose effects are reversed by the activating phosphatase cdc25. Plx1—polo-like kinase 1, which phosphorylates and activates cdc25. Positive feedback loops are shown as dashed lines. The figure is adapted from Ferrell 2002 (see color plate).
The presence of multiple feedback mechanisms able to regulate cyclin Bcdk1 activity results in a switch-like, or as proposed by Ferrell, a ‘‘bistable’’ system (Ferrell, 2002). Such a system is characterized by the presence of two stable states, and the ability of an ‘‘all-or-none’’ switch from one state to the other in response to graded stimuli. In a bistable system, small stimuli have little effect on cyclin B-cdk1 activity, whereas stimuli above a certain threshold have a significant effect (Fig. 3). The mechanisms of mitosis and meiosis regulation are quite different, despite the same cyclins being major molecular regulators of both processes. Oocytes spend prolonged periods of time arrested in the prophase of first meiotic division (see Section IIB); consistently, the control of entry into M-phase is different between mitosis and meiosis. To proceed through meiotic divisions, the oocyte accumulates and then activates MPF in response to the mitogenic signal (Fig. 4). Meiotic division itself is a modification of a mitotic one, where the cell does not undergo DNA replication between two successive rounds of division thereby decreasing the amount of DNA to a haploid state. This is a consequence of the fact that regulation of MPF during meiosis is different than that in the mitotic cell cycle, namely, MPF activity increases rapidly after the first meiotic division, and the second M-phase starts without a preceding S-phase. To gain
72
Voronina and Wessel
Figure 4 Schematic representation of MPF activity changes during growth and maturation of oocytes compared with levels of the regulatory subunit of MPF, cyclin B (see color plate).
insight in the molecular mechanism of meiotic M-phase initiation and progression, cell cycle dynamics of cyclin A, cyclin B, and cdk1 kinase during oocyte maturation have been studied in many organisms (reviewed in Kishimoto, 1999; Yamashita, 1998), and significant differences are found between species arguing against a unified mechanism regulating MPF activity in oocytes (Taieb et al., 1997). One mechanism regulating MPF activity in oocytes is accumulation of cyclin B protein, the best example being fish (reviewed in Yamashita, 1998, Fig. 4). In this case, the newly synthesized cyclin B associates with the catalytic subunit (cdk1), which is not subjected to inhibitory phosphorylations, immediately forming active complexes. Other layers of regulation known for MPF include posttranslational modification of the catalytic subunit, as exemplified in Xenopus, mouse, and starfish (Kishimoto, 1999). In these oocytes, cyclin B is made during the growth phase of the oocytes, and newly formed cyclin B-cdk1 complexes are immediately inactivated by inhibitory phosphorylations. These are in turn removed in response to signals inducing oocyte maturation, ensuring rapid activation of the stockpile of inactive MPF. Recent data from sea urchin oocytes suggests a similar MPF activation strategy (Voronina et al., 2003). Transport of MPF
3. The Regulation of Oocyte Maturation
73
between the cytoplasm and the nucleus is important for its activation and function, and the dynamic changes in MPF localization may contribute to its switch-like activation (Takizawa and Morgan, 2000). Recently, existence of additional MPF components has been proposed based on the observation that components of the germinal vesicle of oocyte are needed in addition to the activated cyclin B-cdc2 to induce GVBD in the nonstimulated oocytes in starfish (Kishimoto, 1999).
IV. Initiation of Oocyte Maturation Hormones are the best documented stimulators of oocyte maturation. Nevertheless, stimuli that induce oocyte meiosis in various species are quite diverse, and include fertilization (in clam, Spisula solidissima) and even a secreted sperm protein factor (Caenorhabditis elegans, described further in Section IVD, Miller et al., 2001). The somatic cells of the ovary relay the inductive signal for oocyte maturation from the hormones2 released into the circulation. A number of mechanisms for triggering oocyte maturation have been proposed so far that do not necessarily exclude each other. Among those are: production of a maturation-inducing substance by follicle cells that directs the oocyte to mature (starfish, fish, frog, maybe mammals); inactivation of the follicle-produced inhibitor (mammals); and inhibition of gap junction-mediated transport to prevent transfer of a follicle-produced inhibitor (fish, mammals).
A. Maturation-inducing Substance (MIS) Steroid hormones have been found to induce oocyte maturation in a number of vertebrate species. Although traditionally such hormones are known to mediate their effects via nuclear receptors inducing transcription, it appears that signaling in the oocyte involves activation of membrane receptors, and is independent from new transcription. At this time, candidate MISs are proposed for starfish, fish, amphibians, and mammals. Mammals. The meiosis-inducing activity of human follicular fluid was identified by an ability to promote maturation in mouse oocytes dissected from the follicles. Gas chromatography mass spectrometry identified the 2 These hormones could generally be referred to as ‘‘gonadotropins’’ (as in Masui and Clarke, 1979), but the term appears confusing, since presently it is used only for vertebrate hormones (in which capacity it will be used herein). However, the ‘‘gonad-stimulating substance’’ of starfish is functionally a ‘‘gonadotropin’’ (acting on ovary) as well.
74 Table I
Voronina and Wessel Structure of the Proposed Maturation Inducing Substances (MIS) of Various Species
Species
Name of MIS
Mammals
4,4,-Dimethyl-5-cholesta-8, 14,24-trien-3-ol (FF-MAS)
Frog
Progesterone
Fish
17,20-Dihydro-4-pregnen-3-one (DHP) is shown in black; the extra hydroxyl of 17,20, 21-trihydro-4-pregnen-3-one (THP) is shown in parentheses
Starfish
1-Methyladenine
Structure of MIS
See also Fig. 5 for the structure of C. elegans MIS.
active compound as 4,4,-dimethyl-5-cholesta-8,14,24-trien-3-ol, an intermediate in cholesterol biosynthesis (Table I). It was proposed to be the ovarian stimulatory factor inducing meiotic resumption of the oocytes, and it was named FF-MAS for follicular fluid meiosis-activating sterol (Table I,
3. The Regulation of Oocyte Maturation
75
Byskov et al., 1995). This sterol was able to induce maturation in denuded or cumulus-enclosed oocytes, as well as in rat ex-vivo perfused ovaries (Grondahl et al., 2000; Hegele-Hartung et al., 2001; Ruan et al., 1998). The origin of FF-MAS has not yet been established, but evidence suggests that cumulus cells or follicle cells are responsible for production of this factor (Byskov et al., 1995, 2002). However, other experimental evidence argues against FF-MAS being the physiological MIS, as specific inhibitors of FFMAS-synthesizing enzymes fail to suppress gonadotropin-induced meiosis in vivo in rat (Tsafriri et al., 2002), so its relevance is still a matter of debate. It would be useful to assess whether such inhibitors are indeed effective before completely ruling FF-MAS out. For FF-MAS to be a physiological signal for meiosis its levels might increase before oocyte maturation; indeed, such increase is detected. In vivo, mammalian oocytes resume meiosis after a rise in gonadotropin levels in blood: follicle-stimulating hormone (FSH) followed by luteinizing hormone (LH). The amounts of FF-MAS in rabbit ovaries rapidly and significantly increase after animals were injected with FSH followed by hCG (human chorionic gonadotropin; LH activity) injection, whereas FSH on its own does not significantly increase follicular FF-MAS levels (Grondahl et al., 2003). In amphibians, experimental evidence points to androgens being the physiological MIS. Progesterone (Table I) had been long suggested to be the MIS causing oocyte maturation, although other steroids such as aldosterone and testosterone are reported to be as efficient as progesterone (Masui and Clarke, 1979). Frog oocytes effectively mature in response to progesterone, and preventing conversion of pregnenolone to progesterone in the ovary by drugs (cyanoketone and eliptin) inhibits gonadotropin-induced oocyte maturation. However, measurements of steroid content of female frogs stimulated by hCG reveal that progesterone concentrations in the serum and ovaries are lower than expected to be effective (Fortune and Tsang, 1981; Lutz et al., 2001). The major constituents of steroids produced by the ovary are the androgens testosterone and androstenedione, that are very potent inducers of oocyte maturation, suggesting that androgens may be the natural MIS. These steroids are metabolites of progesterone, so inhibiting progesterone production (as in earlier experiments) would inhibit synthesis of androgens as well. Progesterone is quickly metabolized to androstenedione by isolated frog oocytes; however, this conversion is not necessary for progesterone-induced maturation, thereby resolving disparate conclusions of the earlier experiments (Lutz et al., 2001). In fish, two progestin derivatives, 17,20-dihydro-4-pregnen-3-one (DHP) and 17,20,21-trihydro-4-pregnen-3-one (THP), were identified as major active substances produced by the follicles upon stimulation with pituitary extract in multiple species (Table I, Petrino et al., 1993; Thomas
76
Voronina and Wessel
et al., 2002, and references therein). DHP is quickly and efficiently metabolized to an array of steroid compounds by the oocytes, however, none of the DHP metabolites is as active a trigger of GVBD as DHP itself, indicating that the induction of maturation is not due to metabolic conversion into a more active form. In a number of fish species, synthesis of DHP takes place in the follicle cells next to the oocyte in response to gonadotropins, and steroid accumulation occurs in the follicular layers as well as inside the oocyte such that DHP acts locally and transiently (Nagahama et al., 1994). In starfish, the adenine derivative, 1-methyladenine (1-MA), is the natural MIS (Table I, Kanatani, 1969; Schuetz, 1971). 1-MA is produced in the follicle cells surrounding the oocyte by the transfer of a methyl group from S-adenosylmethionine onto an ATP molecule, making 1-methylATP, which is then converted into 1-methyladenine (Mita et al., 1999a,b). 1-MA is quickly metabolized in the recipient oocytes and its metabolic product is not biologically active (Toole and Schuetz, 1974). 1-MA is synthesized and secreted into the extracellular space between the follicle and the oocyte in response to a peptide released by the radial nerve, which is analogous in action to vertebrate gonadotropin (reviewed in Smiley, 1990). The gonadstimulating substance of the starfish Asterias amurensis has been purified and identified as a polypeptide, with a molecular weight of 2.1 kDa, although its exact amino acid sequence has not been determined (Kanatani, 1985). 1-MA is effective in inducing maturation in all starfish species; however, it seems selective for this particular group of echinoderms. Sea urchins, sand dollars, sea lilies, and sea cucumbers do not respond to this compound (reviewed in Smiley, 1990; and Wessel, personal communication). Furthermore, sea cucumbers produce an unidentified MIS, preliminary characterized as 2,8 disubstituted adenine (Smiley, 1990). Perhaps closely related adenine derivatives fulfill the role of 1-MA for echinoderms other than starfish.
B. Receptors for Maturation-inducing Substance Work on characterizing the mammalian receptors for FF-MAS has only recently begun. The hydrophobic nature of the sterol would allow it to accumulate in the plasma membrane, or potentially permit its diffusion through the membrane, such that it could function intracellularly. However, intracellular injection of FF-MAS is ineffective in the induction of maturation (Grondahl et al., 2003). Furthermore, a potent antagonist of FF-MAS, 22-R-cholesterol, can efficiently inhibit meiotic resumption in response to extracellular FF-MAS only when it is applied from the outside,
3. The Regulation of Oocyte Maturation
77
and not injected into the oocyte, suggesting that it competes for binding to a surface receptor. In amphibian oocytes, the controversy over the nature of the MIS receptor is still not resolved, and new developments in this area are expected considering that androgens and not progesterone are the actual MIS. Experiments attempting to distinguish whether progesterone acts on the outside or on the inside of the cells are still inconclusive, in large part due to the fact that progesterone is a hydrophobic molecule, which is able to cross biological membranes by diffusion. What appears clear, however, is that progesterone-induced as well as androgen-induced maturation does not depend on new transcription, as it is unaffected by the transcriptional inhibitor, actinomycin D (Lutz et al., 2003; Smith and Ecker, 1969). Although extracellular application of progesterone is efficient in inducing oocyte maturation, the injection of this sterol does not always lead to resumption of meiosis. Such injections were found effective by some groups (Bayaa et al., 2000; Tso et al., 1982), but not others (Godeau et al., 1978; Masui and Clarke, 1979; Smith and Ecker, 1971). Conceptually, it is possible even for the injected steroid to diffuse out of the cell, and still act on its surface. Attempts to prevent transmembrane diffusion of progesterone included covalent crosslinking to a 20 kDa polymer (polyethylene oxide), bovine serum albumin, or to polystyrene beads. These modified progesterone compounds were still active when applied to the surface of oocyte under conditions minimizing endocytic uptake (Godeau et al., 1978; Masui and Clarke, 1979), but of inconsistent activity when microinjected into the cell (Bayaa et al., 2000; Masui and Clarke, 1979). An additional confounding factor is that the preparations of crosslinked progesterone could still contain sufficient free sterol to account for the observed biological activity, which is difficult to control (Bayaa et al., 2000). In support of the plasma membrane being the site of progesterone action, progesterone appears to bind specifically and with high affinity to frog oocyte membranes (Lutz et al., 2000). These specific high affinity binding sites can be detected on Stage V–VI (maturation-competent) oocyte membranes, but not on Stage I–III (maturation incompetent) oocyte membranes (Lutz et al., 2000). What is the nature of this membrane sterol receptor? Experimental results of Lutz et al. argue against progesterone serving as a ligand for a member of the heterotrimeric Gprotein coupled receptors (GPCR). The affinity of a GPCR agonist for its receptor usually decreases upon activation of the associated G subunits. When Xenopus oocyte membranes are incubated with the nonhydrolyzable GTP analog GTPS locking all G subunits in the activated state, progesterone binding stays at the same level as in untreated membranes, suggesting that progesterone does not bind to its receptor in a manner
78
Voronina and Wessel
typical for a GPCR agonist (Lutz et al., 2000). Thus, we are left with the following alternatives: the membrane progesterone receptor (mPR) is not a GPCR or, the mPR is a GPCR, but progesterone does not bind to it as a typical agonist (and may instead function as an antagonist, see Section VA). Nuclear progesterone receptor (hormone-dependent transcription factor, nPR) may be essential for the oocyte maturation in response to progesterone (reviewed in Maller, 2001). Overexpression of nPR increased sensitivity to progesterone and accelerated the rate of progesterone-induced meiosis (Bayaa et al., 2000; Tian et al., 2000), while injection of nPR antisense oligonucleotides into the oocytes inhibited the progesterone-induced oocyte maturation after 8 days of treatment (Tian et al., 2000). The observed effects were specific for antisense but not control oligonucleotides, and could be rescued by subsequent injection of the oocyte with mRNA coding for Xenopus or human nPR. However, additional experiments are required to firmly establish the role of nPR in maturation. The antisense treatment described (Tian et al., 2000) involves culturing oocytes in vitro for 8 days without nutritional supply, which may very well change the cell’s physiology. Furthermore, even the rescue experiments could be questioned as the overexpressed nPR could initiate maturation by a nonphysiological pathway (via direct activation of a downstream kinase Rsk; Maller, 2001). Of particular concern is the fact that sterol binding specificity characteristic of conventional nuclear progesterone receptors does not match the GVBDinducing potency profile of these sterols (reviewed in Maller, 2001). Therefore, it is important to test whether the cloned Xenopus nPR exhibits an unusual sterol-binding profile. Following the suggestion that androgens may be the in vivo signal inducing meiotic maturation in frog, the role of nuclear androgen receptor (nAR) in Xenopus oocytes maturation was tested (Lutz et al., 2003). Treatment of the oocytes with dsRNA for 2.5 days was able to significantly inhibit oocyte maturation in response to androgen without altering the responsiveness to progesterone. nAR protein was found expressed throughout the Xenopus oocyte, and a fraction of the protein was associated with the plasma membrane. A number of sterol ligands demonstrated the ability to induce nAR-dependent transcription in a heterologous system. Unexpectedly, when assayed for the ability to induce oocyte maturation, two of these sterols (dihydrotestosterone and R1881) appeared to be potent antagonists of testosterone-induced, but not progesterone-induced oocyte maturation. This may cast a doubt on the nAR function as a MIS receptor; however, at the same time it afforded an opportunity to test the importance of androgen-mediated signaling in an ovarian explant. Both R1881 and dihydrotestosterone were able to inhibit maturation of oocytes in hCGtreated Xenopus ovaries, despite efficient production of testosterone by the
3. The Regulation of Oocyte Maturation
79
ovaries under these conditions, arguing that the testosterone-induced signaling is indeed the predominant in vivo pathway for induction of oocyte maturation in frog (Lutz et al., 2003). Since progesterone- and androgen-dependent maturation does not require new transcription, the mechanism of transcription-independent nuclear receptor effect may be through activation of Src family tyrosine kinases. Recently, a proline-rich motif unique to nPR was uncovered in the N-terminal domain of the protein, which appeared to mediate direct hormone-dependent binding of nPR to SH3 domains of various cytoplasmic proteins. This motif appeared necessary and sufficient for Src activation by a SH3 domain-displacement mechanism, and presence of this domain in protein overexpressed in Xenopus oocytes was required for the enhancement of the response to progesterone (Boonyaratanakornkit et al., 2001). However, overexpression of a fusion of nPR N-terminal DNA binding domain with estrogen-binding domain in Xenopus oocytes did not make these oocytes responsive to estrogen (Bayaa et al., 2000). Nevertheless, this result could be explained by a requirement for the presence of the nPR ligand-binding domain for the N-terminal part to bind to and activate Src efficiently (Boonyaratanakornkit et al., 2001). In fish the plasma membrane of the oocyte possesses specific high affinity binding sites for MIS. Maturation-inducing substance (DHP, Table I) acts at the cell surface and is ineffective in inducing maturation when injected into oocytes. Investigation of the binding affinity of plasma membrane and nuclear MIS binding sites for a range of sterols demonstrates that they are characterized by markedly different binding affinities for a number of steroid compounds. Thus, it is unlikely that a nuclear hormone receptor participates in the MIS signaling at the plasma membrane. Furthermore, the affinity of plasma membrane sites for MIS decreases with incubation of the membranes with GTPS, consistent with the hypothesis of a GPCR being an MIS receptor, since activation of the associated G subunits leads to a decrease in the affinity of a typical GPCR for its ligand (reviewed in Thomas et al., 2002). Recently, the fish MIS candidate receptor was cloned by an innovative approach. Monoclonal antibodies were produced against a partially purified, solubilized oocyte plasma membrane fraction. They were then screened for the ability to bind to the solubilized membrane receptor using a hormone receptor capture protocol, where the efficiency of immunoprecipitation of a receptor selectively binding the radiolabeled ligand is assessed. This assay was optimized in a way to maintain specific ligand binding in the solubilized preparations of the membranes, which is not always possible for GPCRs. Three monoclonal antibodies were selected, and expression screening of the cDNA library identified a protein with no similarity to known receptors (Thomas et al., 2002; Zhu et al., 2003b). Computer analysis
80
Voronina and Wessel
of the amino acid sequence of this protein predicts seven transmembrane domains, consistent with a GPCR structure, but suggestive of representing a separate subclass of these receptors. In line with this, a number of proteins homologous to the suggested fish membrane progestin receptor (mPR) have been isolated from several vertebrate species (Zhu et al., 2003a). Immunolocalizations, as well as subcellular fractionations, consistently detect mPR localized on the plasma membrane of the oocyte. Heterologously expressed mPR confers sensitivity to DHP in transfected mammalian tissue culture cells. In these cells, DHP application reduces cAMP production and activates MAP kinase; these effects are consistent with proposed signaling through this receptor in the oocyte at resumption of meiosis. The protein profile shows that mPR is low in previtellogenic oocytes, increases with oocyte growth, and disappears upon oocyte maturation, consistent with what would be expected for a receptor, whose function is specific for oocyte maturation. Depletion of the mPR in the zebrafish oocytes by morpholino antisense oligonucleotides blocks maturation in vitro in response to DHP, consistent with this protein being the MIS receptor (Zhu et al., 2003b). However, further experiments would be helpful to definitely establish the proposed role for mPR in oocyte maturation. For example, bacterially overexpressed protein displays an unexpected sterol-binding profile, which does not match the steroid specificity of maturation stimulation. Furthermore, a rescue of the antisense treatment experiments in zebrafish would be useful, accompanied by immunological detection methods of the mPR protein. Thus, it is suggested that fish oocytes utilize a GPCR-mediated induction of oocyte maturation, in general contradiction to signaling described for frog. Are these differences to be reconciled in the future? It is possible that nuclear and plasma membrane sterol receptors participate at different stages of oocyte maturation in the MIS signaling. The homologues of the fish mPR have been recently isolated in a number of vertebrates including frogs (Zhu et al., 2003a). Future experiments will test the involvement of this GPCR family in the signaling at oocyte maturation in different species, and define the contribution of nPRs, if any. In starfish, a variety of evidence indicates that 1-MA acts at the oocyte surface. For example, intracellulary injected 1-MA does not cause reinitiation of meiosis (reviewed in Kanatani, 1985). Furthermore, 1-MA specifically binds to the oocyte surface: binding of a radiolabeled 1-MA is inhibited in a dose-dependent manner by unlabeled 1-MA and biologically active analogs (1-benzyladenine and 1-ethyladenine), but not by biologically inactive analogs (1,9-dimethyladenine or 1-methylhypoxanthine) (Mita et al., 2001; Yoshikuni et al., 1988). It is hypothesized that a seven-transmembrane GPCR in the plasma membrane binds 1-MA (Kishimoto, 1999). G-protein coupled receptors
3. The Regulation of Oocyte Maturation
81
(GPCRs) are known to have different affinities for receptor agonists depending on their association with G-proteins. The binding affinity of the starfish 1-MA receptor changes consistent with it being a GPCR: the affinity of starfish oocyte cortices for 1-MA decreased after in vitro treatment with GTPS that dissociates G-proteins from their receptors (Chiba and Hoshi, 1995). Induction of oocyte maturation by treatment with reducing agents (reviewed in Kanatani, 1985) most likely does not act on the receptor, since pertussis toxin blocks the induction of maturation by 1-MA (see Section IVA), but not by dithiothreitol (Chiba et al., 1992; Chiba and Hoshi, 1995). The exact nature of the 1-MA receptor has not been determined yet, despite many efforts to biochemically characterize or clone it (e.g., Kalinowski et al., 2003).
C. Follicle Cells and Gap Junctions Oocytes of many species develop in close contact with the somatic cells of the ovary. For example, in mammals, a multilayered complex of somatic cells surrounds the oocyte in a specialized structure called a follicle (Fig. 1). The follicle cells adjacent to the oocyte form numerous specialized contacts (gap junctions) with the oocyte (reviewed in Matova and Cooley, 2001). These structures are important for supplying the oocyte with nutrients, and for communication between somatic cells and the oocyte. Gap junctional communication plays an important regulatory role for oocyte maturation. In mammals, hormones targeting oocyte development exert their action on somatic cells of the follicle, which then relay the signal to the oocyte. The pituitary gland releases the gonadotropins FSH and LH. FSH causes the follicle cells to proliferate and differentiate, and LH initiates the oocyte’s progression through meiosis and ovulation. The follicle cells mediate this initiation of oocyte maturation in response to hormones. The oocyte is usually incapable of maturation during its growth period (Masui and Clarke, 1979) but once meiotic competence is achieved in a full grown oocyte, the granulosa cells become the major suppressor of meiotic progression. Mammalian and fish defolliculated oocytes undergo spontaneous maturation in culture, while they do not when they are cultured enclosed in follicles (reviewed in Conti et al., 2002; Patino and Purkiss, 1993). As follicle cells significantly influence the state of the oocyte it was hypothesized that the oocyte is maintained in the immature state by a follicular ‘‘arrester’’ produced by granulosa cells. How do the follicle cells communicate with the oocyte to keep the cell cycle blocked?
82
Voronina and Wessel
Numerous gap junctions exist between the cells of the ovarian follicle, providing connections among somatic cells, and between follicle cells and the oocyte. These intercellular membrane channels can be detected by transfer of metabolites or small fluorescent dyes. Gap junctions are essential for oocyte growth and development and were proposed to provide the way for a small molecule ‘‘arrester’’ to diffuse into the oocyte. The abundance of gap junctions is regulated by the amounts of their constituent proteins, connexins, in the cell. ‘‘Coupling efficiency’’ of existing gap junctions is further regulated by their subunit composition and posttranslational modifications of connexins by intracellular signaling mediators (reviewed in Kidder and Mhawi, 2002). Regulation of follicle gap junctions within the follicle is hypothesized to be involved in the induction of oocyte maturation. In certain animals, like the Syrian hamster, it was determined that downregulation (decrease in number and density) of follicle cells’ gap junctions is correlated with an irreversible commitment of the oocytes to maturation (Racowsky et al., 1989). On the other hand, in the mouse, the oocyte–cumulus cells coupling index remained constant throughout the GVBD period in hCG-stimulated animals (Eppig and Downs, 1988) and in the explanted cumulus cell–oocyte complexes induced to mature with gonadotropin (Downs, 2001). The molecular mechanism for downregulation of gap junction coupling efficiency is suggested to act at several levels including phosphorylation of the protein and decrease in net levels of protein and mRNA. In rats, the gap junction protein connexin-43 is rapidly phosphorylated in response to exposure to LH (Granot and Dekel, 1994). Furthermore, prolonged exposure to LH causes complete elimination of connexin-43 protein correlated with a decrease in its mRNA level. The experimental manipulation of gap junctions can also influence the meiotic state of the oocyte. For example, application of chemicals such as 1-octanol that suppresses gap junction-mediated cell–cell coupling appears to initiate oocyte maturation in follicles of the fish Fundulus heteroclitus as well as in mouse cumulus-enclosed oocytes (Cerda et al., 1993; Downs, 1995). However, Xenopus follicle-enclosed oocytes cannot be stimulated to mature in this way (Patino and Purkiss, 1993), suggesting that the input of a gap junction-delivered arrester in mediating prophase arrest of the oocyte varies in different animals. Gap junctions appear to transmit both inhibitory and stimulatory signals between follicle cells and the oocyte. If prophase I arrested, follicle-enclosed fish or mouse oocytes are treated with stimulatory ligands, GVBD is triggered at greater frequencies than those achieved by just removing the follicle cells (Cerda et al., 1993; Downs, 1995, and references therein). Furthermore, induction of meiotic maturation in cumulus-enclosed oocytes depends significantly on the presence of functional gap junctions between the oocyte and cumulus cells, as assessed by either inhibiting gap junctions
3. The Regulation of Oocyte Maturation
83
in intact follicles chemically, or by dissociating cumulus cells from the oocyte, and subsequently coculturing them together (Downs, 1995, 2001). Experiments in Xenopus oocytes produce comparable results: while oocyte maturation of the follicle-enclosed oocytes can be induced by gonadotropin treatment, gap junction-inhibiting reagents suppress this induction (Patino and Purkiss, 1993). Thus, even though in Xenopus no inhibitory signal is transmitted through gap junctions, there appears to be an activating input instead. This effect appears to result specifically from disruption of gap junctional communication, and not from detrimental effects of the 1-octanol on the follicle cells or the oocyte itself. This conclusion is based on the result that treatment with 1-octanol does not decrease the amounts of steroids produced by the follicle cells, and does not inhibit maturation of follicle-free oocytes in response to external progesterone. It is hypothesized that gap junctional coupling of follicle cells and oocyte in Xenopus either facilitates delivery of the hydrophobic sterol to the oocyte surface, or provides a cytoplasmic route of progesterone delivery to the oocyte.
D. A Whole New Can of Worms A strategy of oocyte maturation regulation unique to the worm C. elegans has been recently described. Maturation of the oocytes in this animal is dependent on the presence of sperm in the spermatheca (but not fertilization). Major sperm protein, MSP (Fig. 5), secreted by sperm was found to be the active component inducing oocyte maturation (Miller et al., 2001). This mechanism is quite unusual in that no other known meiosisinducing hormone is a peptide (Table I). As the MSP gene family in C. elegans consists of 40 genes whose protein products differ by one to four amino acids, knockout experiments do not appear feasible. However, the evidence for MSP being the maturation-inducing activity is strong, since recombinant MSP is able to induce oocyte maturation in the absence of sperm, while injection of antibodies to MSP into the uterus of the adult nematode leads to a reduction in the oocyte maturation rate (Miller et al., 2001). A screen for the receptors of MSP was carried out by a combination of microarray-based identification of oocyte-enriched genes, and RNAimediated ablation of the candidates followed by assay for MSP binding. These efforts identified a Vab-1 Eph (ephrin) receptor protein tyrosine kinase (RPTK) as the only definite candidate (Kuwabara, 2003; Miller et al., 2003). However, the phenotype of vab-1 mutants is inconsistent with this protein positively regulating oocyte maturation. Instead, Vab-1 is a suppressor of oocyte meiosis, and MSP induces maturation by antagonizing Vab-1 activity. It appears that Vab-1 expressed on the oocyte surface
84
Voronina and Wessel
Figure 5 Crystal structure of the major sperm protein (MSP) of a nematode C. elegans (from Baker et al., 2002). Note that MSP is significantly larger than any other maturation-inducing substance molecule (Table I) (see color plate).
receives a constitutive signal from the Efn-2 ephrin also expressed in the germ line to keep up the inhibitory activity. This system thus functions as a sperm-sensing checkpoint mechanism that inhibits oocyte meiotic maturation until sperm is available for fertilization. Potentially, this model of ‘‘inhibition relief’’ may be defined by the hermaphrodite reproductive strategy of C. elegans, and appears unique to this animal. However, other organisms could utilize hormone-mediated relief of constitutive inhibition of cell cycle progression as a means to induce oocyte maturation as well (see Section VA).
V. Cytoplasmic Control of Maturation The ultimate target of signal transduction pathways initiated at oocyte maturation is activation of MPF kinase and release of the oocyte from the block in meiotic progression. Among candidate second messengers linking the MIS receptor to MPF activation are cAMP and calcium; these and more upstream components of signaling machinery will be examined in this section.
3. The Regulation of Oocyte Maturation
85
A. Signal Transduction through G-proteins: Ignition or Parking Brake? How is the presence of hormones interpreted by an oocyte to achieve release of the cell from the prophase block? The most detailed analysis of an intracellular signaling pathway activated during oocyte maturation comes from the studies of starfish. Heterotrimeric G-proteins are universal signaling molecules present in most of the tissues. As their name suggests, the complex consists of three subunit types: alpha (GTPase), beta, and gamma. The inactive state is a heterotrimeric complex with the subunit in a GDP-bound form. Upon activation, the GDP of the subunit is exchanged for GTP, and dissociates from the complex. Free GTP- and free diffuse in the plane of membrane and regulate the activity of their respective effectors. Upon hydrolysis of GTP by the subunit, it returns to the inactive GDP-conformation and reassociates with , thus concluding the activity cycle. Different G subunits have distinct sets of intracellular targets and some G proteins have opposing effects on the same targets. For example, Gs and Gi both target adenylate cyclase, however Gs activates this enzyme while Gi inhibits it. Mouse oocytes can be maintained in the prophase arrest in culture indefinitely when they are enclosed in their follicles. However, when the follicle cells are removed, mouse oocytes enter maturation spontaneously. It was documented that such cumulus-free mouse oocytes could be maintained in meiotic arrest in a transient and dose-dependent manner when microinjected with GTPS, an activator of G-protein signaling (Downs et al., 1992). This data led to a model whereby spontaneous maturation of the oocyte is induced by removal of inhibitory signal normally present in the ovary and transduced via the heterotrimeric Gproteins (summarized in Fig. 6A). This model was recently supported by work of Mehlmann and colleagues (Mehlmann et al., 2002) suggesting that the inhibitory ovarian signal is mediated by the activity of Gs. Injection of anti-Gs inhibitory antibody induced oocyte maturation in follicle-enclosed oocytes, while injection of anti-Gi inhibitory antibody did not have any effect on spontaneous oocyte maturation. Furthermore, injection of cholera toxin, which hyperactivates Gs-mediated signaling, into mouse oocytes was able to significantly inhibit both spontaneous and FF-MAS-induced maturation suggesting that activation of Gs signaling is able to sustain meiotic arrest of the oocyte (Downs et al., 1992; Grondahl et al., 2000). The most likely target activated by Gs in this cell is adenylate cyclase. It produces cAMP, a second messenger that can sustain meiotic arrest of the oocyte (see Section VC). Consistently, the effect of cholera toxin on spontaneous maturation is more pronounced when the oocyte is cultured in the presence of hypoxanthine, inhibiting cAMP degradation (Downs et al., 1992).
86
Voronina and Wessel
Figure 6 G-protein signaling and oocyte maturation. A. Maintaining vertebrate oocyte prophase arrest by G-protein signaling. Depicted is a summary of experimental data in mouse and Xenopus oocytes (see text for details and references). 7TM GPCR—seven-transmembrane G-protein-coupled receptor; AdC—adenylate cyclase; GRK—G-protein coupled receptor kinase. Blue color: reagents decreasing cAMP levels in oocytes and promoting oocyte maturation, red color: reagents found or expected to increase cAMP levels in oocytes. B. Induction of starfish (and possibly fish) oocyte maturation by G-protein signaling (see text for details and references) (see color plate).
This model of G-protein function in mouse oocytes is supported in Xenopus oocytes, which can be stimulated to mature by injection of an antibody inhibiting Gs activity. Such injected oocytes proceed through GVBD with the same time course as progesterone-stimulated ones, and the resulting eggs are fertilizable (Gallo et al., 1995). Conversely, GTPS and cholera toxin antagonize the action of progesterone on Xenopus oocytes, while GDPS has no effect (Cork et al., 1990; Schorderet-Slatkine et al., 1982). Increasing the amounts of Gs in Xenopus oocytes blocks progesterone-induced oocyte maturation as well. This effect is even more pronounced if the mutant constitutively activated form of the protein is expressed (Romo et al., 2002). Furthermore, depletion of endogenous Gs by injection of antisense oligonucleotides induced early events of oocyte maturation independent of progesterone addition (Romo et al., 2002). The activation of adenylate cyclase in response to Gs signaling can be mediated by both Gs and G complex interaction with the enzyme (Simonds, 1999). In the Xenopus oocyte, subunits released upon Gs activation are also thought to play a significant role in the signaling pathway further activating adenylate cyclase. Experiments by multiple research groups demonstrate that sequestration of endogenous enhances, while overexpression of exogenous inhibits progesterone-induced as well as androgen-induced maturation in a number of studies (Lutz et al., 2000, 2003; Sheng et al., 2001; but not Gallo et al., 1995). The observation that the amount of subunits in mouse oocytes decreases with oocyte maturation is consistent with the model that a relative decline in is involved in the commitment of the oocyte to meiotic maturation (Allworth et al., 1990).
3. The Regulation of Oocyte Maturation
87
The evidence in support of a constitutively active G-protein-coupled receptor upstream of this inhibitory Gs signaling was recently reported (Wang and Liu, 2003). Reagents known to promote classic GPCR desensitization, such as G-protein receptor kinase (GRK) and -arrestin, were found to induce Xenopus oocyte maturation independent of progesterone, presumably by downregulating the Gs signaling originated from an upstream GPCR (Wang and Liu, 2003). These results are summarized in a common model of sustaining prophase arrest in a vertebrate oocyte by a combination of signaling by Gs itself and subunits generated through Gs activation (Fig. 6A). The adenylate cyclase enzymes requiring both Gs and subunits for activation have been previously described in mammals (Simonds, 1999). However, it is still unclear how the effects of Gs activation are reversed in vivo by hormone action on a mouse follicle or a Xenopus oocyte. Hormonedependent resumption of maturation in both mouse and Xenopus oocytes includes a rapid decrease of cellular cAMP. Could this effect be mediated through activation of Gi family of heterotrimeric G-proteins (which is known to antagonize Gs by inhibiting adenylate cyclase)? It is possible to experimentally detect Gi activation during mouse oocyte maturation. The amounts of Gi proteins somewhat decrease during mouse oocyte maturation as determined by western blotting (Jones and Schultz, 1990). However, dissociation of the Gi from the subunits is detected by a substantial decrease in the amounts of pertussis toxin ribosylation substrate in the mature egg in comparison to the oocyte, that cannot be accounted for by the decrease in Gi protein levels (Allworth et al., 1990; Jones and Schultz, 1990). The reduction in the PTX substrate amount occurs as well in vitro, and can be detected after 3 h of in vitro culture, coinciding with GVBD when mouse oocytes mature spontaneously (Jones and Schultz, 1990). This suggests that activation of Gi may take place upon removal of the mouse oocyte from the follicle. Involvement of Gi-mediated signaling in the response of mouse and frog oocytes to maturation-inducing hormones was tested by using its specific inhibitors (Faerge et al., 2001; Goodhardt et al., 1984; Grondahl et al., 2000; Lutz et al., 2000). Pertussis toxin inhibits certain members of the Gi family of heterotrimeric G-proteins by ADP-ribosylation and blocks their respective signaling. Gi inhibition by pertussis toxin does not consistently affect resumption of meiosis in response to MIS in both vertebrate species suggesting that suppression of Gs signaling might not occur through induction of ‘‘opposing’’ Gi signaling. The injected pertussis toxin Asubunit was found effective (Pellaz and Schorderet-Slatkine, 1989), while incubation of oocytes with the whole pertussis toxin is generally not effective (Downs et al., 1992; Faerge et al., 2001; Goodhardt et al., 1984; Kline
88
Voronina and Wessel
et al., 1991; Sadler et al., 1984). It appears that even though the Gi signaling might be activated during the oocyte maturation, it does not contribute directly to eventual resumption of meiosis. Activation of all G-proteins in the mouse oocyte by GTPS injection does not significantly interfere with FF-MAS-induced maturation, nor does it have a stimulatory effect (Grondahl et al., 2003). In contrast, when GDPS is injected, FF-MAS-induced meiosis resumption is significantly decreased. These effects are markedly different from the spontaneous maturation, which is inhibited by GTPS injection (Downs et al., 1992). It appears, that for FF-MAS to exert its action, inhibition of G-protein signaling is not required; instead, a (pertussis toxin-insensitive) G-protein must be activated. Overexpression or injection of constitutively activated Gi in frog has minimal effects on either spontaneous or progesterone-induced maturation, which indicates that Gi activity is not important for the Xenopus maturation response (Kroll et al., 1991; Lutz et al., 2000). Participation of another type of G-protein was suggested, as injection of constitutively active Go is able to stimulate Xenopus GVBD (Kroll et al., 1991). However, the Go-mediated maturation depends on PKC activation while progesterone-mediated maturation does not. Finally, expression and activation of heterologous Gi-coupled receptors does not cause Xenopus oocyte maturation (Kalinowski et al., 2003). It is conceivable then that the function of G-proteins is to maintain the oocyte’s cell cycle arrest rather than to induce oocyte maturation. The action of progesterone thus might be in relieving the Gs/ mediated inhibition. Possible mechanisms include functioning as an antagonist of a constitutively active GPCR; in this light, the fact that progesterone does not bind to the oocyte membranes in the manner typical of GPCR ligands becomes less puzzling (see Section IVB). An alternative mechanism could be hypothesized to include a novel effect of cytoplasmic progesterone receptor on Gs, , adenylate cyclase, or phosphodiesterase potentially through activation of Src (see Section IVB). Probing the signal transduction pathway leading to oocyte maturation in fish uncovered that—consistent with other vertebrates—hyperactivation of Gs by cholera toxin inhibits the oocyte’s response to MIS. The involvement of Gi signaling in MIS-induced maturation is uncertain as microinjection of activated pertussis toxin into oocytes blocks the MIS-induced maturation only in select species (Thomas et al., 2002). Recent experiments identified a candidate fish MIS receptor, which is capable to signal through Gi when expressed in mammalian tissue culture cells (Zhu et al., 2003b). In the context of other oocytes, these contradictory results need to be further evaluated, as they suggest that there may be a diversity in oocyte maturation signal transduction pathways within vertebrates.
3. The Regulation of Oocyte Maturation
89
The 1-MA signal inducing starfish oocyte maturation proceeds through a G-protein pathway (Fig. 6B). The only G-protein involved in regulation of 1-MA-induced maturation in starfish appears to be Gi, since the 1-MAinitiated signal is inhibited by pertussis toxin (Shilling et al., 1989). The Gi subunit modified by pertussis toxin was biochemically purified from starfish oocytes and subsequently cloned (Chiba and Hoshi, 1995; Tadenuma et al., 1991). Furthermore, expression of mammalian Gi-family-linked receptors, such as human adenosine A1 receptor and rat ADP receptor, causes the starfish oocyte to resume meiosis in response to corresponding agonists (Kalinowski et al., 2003). Gs protein is present in starfish oocytes as well, as detected by ADP-ribozylation of the protein by cholera toxin (Chiba and Hoshi, 1995). Nevertheless, injections of the anti-Gs inhibiting antibody in starfish oocytes neither stimulated GVBD, nor inhibited GVBD in response to 1-MA stimulus (Gallo et al., 1995). The subunits of Gi appear to be the key for the initiation of oocyte maturation. Starfish Gi activation, especially release of subunits, is necessary and sufficient to initiate maturation, while signaling through the activated G subunit appears less significant (Chiba and Hoshi, 1995; Jaffe et al., 1993). The free complexes relay the signaling further along the pathway, including phosphatidylinositol 3-kinase (PI3K) as a likely downstream component (Section VD, Sadler and Ruderman, 1998).
B. Calcium Ions Calcium is a universal signaling molecule involved in many important processes such as muscle contraction, synaptic transmission, and fertilization. Recent research suggests that calcium also governs certain transitions of cell division cycle, such as nuclear envelope breakdown and separation of chromosomes at the metaphase–anaphase transition (reviewed in Whitaker and Larman, 2001). It is therefore possible that calcium signaling contributes to the release of the oocyte from the meiotic prophase arrest. A transient increase in the intracellular calcium concentration has long been known to be required for the induction of oocyte maturation, or GVBD, as it is blocked by injection of calcium chelators into the cell (Masui and Clarke, 1979; Pesty et al., 1998). Hormone addition to mammalian, Xenopus, and starfish oocytes leads to the increase in cytoplasmic calcium concentration, under certain conditions in the form of a pronounced spike (Su and Eppig, 2002; Wasserman et al., 1986; Witchel and Steinhardt, 1990, and references therein). Mouse oocytes dissected from the follicles and spontaneously resuming meiosis in vitro display repetitive calcium transients prior to GVBD as well (Carroll et al., 1994). These maturation-associated
90
Voronina and Wessel
calcium transients are generally smaller in amplitude than fertilizationassociated calcium transients (reviewed in Schultz and Kopf, 1995). These results are comparable to calcium signaling in mitosis documented for embryonic cells (in a host of organisms: sea urchin, starfish, Xenopus, mouse), which exhibit calcium rises associated with nuclear envelope breakdown (Whitaker and Larman, 2001 and references therein). Could simply introducing calcium ions into the cell induce oocyte maturation? Most reports conclude that calcium on its own is insufficient for induction of oocyte maturation (see below), although the evidence in Xenopus is still controversial (Han and Lee, 1995; Noh and Han, 1998; Wasserman et al., 1986 and references therein). The calcium ionophore A23187 causes oocyte maturation in a very narrow group of species (some of them, like Spisula, resume meiosis in response to fertilization, and the calcium transient in this species is normally induced by egg activation). For most species, including starfish, A23187 is ineffective, and it simply kills sea urchin oocytes (reviewed in Kanatani, 1985). This suggests that the calcium release is not sufficient for the initiation of oocyte maturation, although calcium ionophore treatment is likely to cause a super-physiological increase in cytoplasmic calcium concentration, and thus cause nonphysiological effects. Still, cytoplasmic calcium transients are necessary for the execution of certain aspects of oocyte meiosis. The stages of oocyte maturation that can be regulated by calcium include nuclear envelope breakdown (in starfish, but not in mouse), extrusion of the first polar body, and cortical granule translocation to the cell surface (Santella and Kyozuka, 1994; Santella et al., 1999; Su and Eppig, 2002; Tombes et al., 1992 and references therein). Starfish oocytes injected with heparin (which intereferes with IP3-mediated calcium release) delay the GVBD in response to 1-MA, but more importantly, do not undergo normal transformations of the cytoskeleton and do not redistribute cortical granules at the plasma membrane (Santella et al., 1999). The mechanism of calcium action at meiosis is not yet fully understood. The strongest possibility, based on comparison with the active mitotic players, is action through the calcium/calmodulin-dependent kinase, CaMKII. Activation of calmodulin and CaMKII has been reported during mitosis in a number of organisms and cultured cell types (reviewed in Whitaker and Larman, 2001). The suggested role of CaMKII at the time of mitosis entry is activation of cdc25 phosphatase, an activator of maturation promoting factor (Patel et al., 1999). In support of meiotic function of this enzyme, CaMKII was found active during the response to progesterone in Xenopus oocytes (Stevens et al., 1999). Expression of a constitutively activated mutant CaMKII is sufficient for induction of partial Xenopus oocyte maturation (movement of GV, but not GVBD) independent of hormonal signal (Waldmann et al., 1990). However, the role of CaMKII in
3. The Regulation of Oocyte Maturation
91
mouse meiotic resumption is unclear, since the CaMKII inhibitors, KN-93, myristoylated AIP (autocamptide-2-related inhibitory peptide), and specific calmodulin antagonist W-7, do not prevent mouse oocytes from spontaneously resuming maturation when cultured in vitro (Su and Eppig, 2002). Instead, CaMKII appears to fulfill a later function in mouse meiosis. All the mentioned CaMKII inhibitors inhibit extrusion of the first polar body in a dose-dependent fashion, apparently blocking the metaphase to anaphase transition. Another potential downstream target of calcium is protein kinase C (PKC). Artificial activation of PKC in Xenopus oocytes can cause GVBD in the absence of hormone (Kroll et al., 1991; Stith and Maller, 1987). However, it is probably not a physiological mediator of progesterone signaling, since inhibition of PKC by injection of inhibitory peptide does not interfere with progesterone-induced maturation (Kroll et al., 1991). Furthermore, artificial PKC activation in mouse oocytes actually prevents GVBD (Bornslaeger et al., 1986; Downs et al., 2001 and references therein). Thus, PKC does not normally mediate calcium signaling in the maturing oocyte. The pathway of calcium increase during meiosis is still under investigation. This ion can be released from the internal stores via two main types of calcium channel receptors: the inositol 1,4,5-triphosphate receptor (IP3R), and the ryanodine receptor (RyR) insensitive to IP3, but stimulated by various agents such as calcium, caffeine, nicotinic acid adenine nucleotide phosphate, or cyclic ADP-ribose (cADPr). The prevailing route for calcium release during mitosis is through generation of IP3 produced by PLC (reviewed in Whitaker and Larman, 2001). IP3 is a strong candidate for the meiotic mediator of calcium release since its levels increase in Xenopus oocytes upon progesterone stimulation (Noh and Han, 1998 and references therein). It appears that the release of calcium is mostly IP3-mediated, since injection of reagents preventing IP3 generation or binding to the receptor delays (in starfish) or prevents (in Xenopus) the onset of normal maturation, while introduction of 8NH2cADPr (inactive analog of cADPr), ruthenium red, or procaine (specific inhibitors of RyR channel) does not (Han and Lee, 1995; Iwasaki et al., 2002; Noh and Han, 1998; Santella et al., 1999). However, RyR-mediated calcium release is important (at least, in starfish) since a combination of heparin and 8NH2cADPr is more efficient than either reagent on its own, and prevents GVBD of starfish oocytes completely (Santella et al., 1999). Producing IP3 artificially in Xenopus oocyte by activation of heterologously-expressed serotonin receptor does not result in induction of oocyte maturation, which is consistent with the conclusion that calcium release, by itself, is insufficient for the induction of maturation (Noh and Han, 1998). However, the amounts of IP3 produced by this artificial mechanism are less than those of
92
Voronina and Wessel
the control oocytes responding to progesterone, making the conclusions somewhat questionable.
C. cAMP Cyclic AMP is a well-known second messenger, whose regulatory targets include kinase and GTP exchange factor activities. In somatic cells, cAMP functions both as a positive and a negative regulator of the mitotic cell cycle progression (depending on cell type), exerting its effects mainly in the G1 part of the cell cycle. In contrast, progression of the oocyte through the meiotic prophase block appears to be negatively regulated by the cAMP levels (reviewed in Conti et al., 2002). cAMP has been repeatedly proposed to be the mysterious inhibitor of oocyte maturation, being made in follicle cells and transferred into the follicle-enclosed oocytes in mammals (see Section IIIC, reviewed in Webb et al., 2002). In support of this concept, a number of reports point to inhibition of oocyte maturation by high cytoplasmic cAMP levels (see below). cAMP is produced by adenylate cyclases and degraded by phosphodiesterases. In starfish, phosphodiesterase inhibitors such as IBMX (3-isobutyl-1-methylxanthine) or other xanthine derivatives have an inhibitory effect on the maturation of 1-MA-treated oocytes (Doree et al., 1976), potentially by interfering with cAMP degradation, but their effects on cAMP levels have not been tested. Furthermore, in frog, forskolin (activator of adenylate cyclase) either delays Xenopus oocyte maturation or blocks maturation completely in a concentration-dependent manner (Noh and Han, 1998). Forskolin inhibits spontaneous oocyte maturation in mouse and fish as well (Schultz et al., 1983; Thomas et al., 2002). In the sea urchin, artificially elevating levels of cAMP in isolated oocytes also maintains the G2 arrest (Wessel et al., 2002). Do cAMP levels decrease naturally upon MIS exposure? In hCGstimulated mice, oocytes exhibit a 30% decrease in cAMP, as detected by radioimmuno assay 1.5 h after gonadotropin administration (Schultz et al., 1983). However, the concentration of cAMP in the whole follicle actually increases up to and throughout GVBD (Eppig and Downs, 1988; Schultz et al., 1983). To explain these dynamics, a regionalized increase in cAMP confined to the follicle cells is proposed resulting either from blocked gap junctional communication between the oocyte and somatic cells, or from selective downregulation of gap junction permeability for cAMP while maintaining permeability for other molecules (Qu and Dahl, 2002). In explants of oocytes enclosed in cumulus cells, stimulation of cumulus cells by FSH leads to a transient increase in cAMP levels of the oocyte as detected by the in situ fluorescent indicator FlCRhR (Webb et al., 2002).
3. The Regulation of Oocyte Maturation
93
Key here is that the oocyte increase is dependent on the presence of functional gap junctions between cumulus cells and the oocyte. Consistently, oocyte maturation in FSH-treated complexes is less than in control groups early on after FSH treatment, at the time of detected cAMP increase, and is significantly induced later (by 10 h post-FSH), when cAMP levels return to baseline (Downs et al., 1988). These data suggest that the changes in cAMP levels occurring during oocyte maturation in mouse depend on the mechanism of maturation induction. The evidence in starfish oocytes is less clear. Some reports indicate that maturation-inducing 1-MA causes a 10–30% decrease in cAMP, starting within 2 min after hormone addition (Meijer and Zarutskie, 1987), but other evidence suggests that cAMP amounts do not change significantly during oocyte maturation (reviewed in Kanatani, 1985). Even though starfish oocyte maturation depends on activation of Gi protein, signaling through the Gi (adenylate cyclase-inhibiting) subunit is not required for the induction of maturation, while it is the release of complex that ultimately stimulates meiotic resumption. The cAMP levels appear to decrease in frog and fish oocytes upon stimulation of meiosis resumption. In Xenopus, the cAMP concentration decreases (10–15%) upon progesterone stimulation within 15 min of treatment (Cork et al., 1990; Schorderet-Slatkine et al., 1982). Induction of oocyte maturation in fish is correlated with a 20–60% decline in the cAMP concentration of the follicle upon treatment with DHP (reviewed in Haider, 2003; Thomas et al., 2002). Even though these are subtle changes in cAMP concentration, they could still be important for the oocytes’ meiotic transitions after being amplified via changes in the activity of the cAMPdependent kinase, and exhibit a threshold effect. The role of cAMP in the regulation of maturation appears to be mediated by its effects on cAMP-dependent protein kinase A (PKA). Microinjections of PKA inhibitors, such as the regulatory subunit of PKA (rPKA), the PKA inhibitory protein (PKI), membrane-permeable cAMP antagonist (rp-cAMP), or small molecule inhibitor H-89 are able to induce oocyte maturation (or at least accelerate hormone-induced maturation) in some species, such as Xenopus or mouse (Leonardsen et al., 2000; Noh and Han, 1998), but not others, such as starfish (reviewed in Kanatani, 1985). On the other hand, stimulating the signaling output of PKA by injection of the catalytic subunit of PKA (cPKA) inhibits the response to hormone in both Xenopus and starfish (reviewed in Kanatani, 1985; Masui and Clarke, 1979). Interestingly, the total PKA activity in Xenopus oocytes does not change after exposure to progesterone (Cicirelli et al., 1988). One proposed mechanism of PKA action is by phosphorylation and inhibition of cdc25, a cyclin B-cdk1 activating phosphatase (Duckworth et al., 2002). Unexpectedly, the kinase-inactive cPKA blocks progesterone-induced
94
Voronina and Wessel
maturation in Xenopus as effectively as wild-type cPKA (Schmitt and Nebreda, 2002). Apparently, this inhibition is not a result of sequestering known regulatory proteins that inhibit endogenous cPKA activity (such as PKI) and represents either a new signaling mechanism or sequestration of cdc25 by the excessive kinase dead cPKA. Interestingly, while microinjection of cPKA into mouse oocytes inhibited spontaneous oocyte maturation, it failed to suppress FF-MAS-induced maturation, suggesting that a decrease in cAMP levels and PKA activity is more critical for initiation of spontaneous maturation than for the response to FF-MAS (Faerge et al., 2001). Oocytes may be able to overcome the inhibitory effects of cAMP when they are stimulated with maturation-inducing substance. For example, isolated mouse oocytes cultured in the presence of the phosphodiesterase inhibitor hypoxanthine do not experience significant changes in their cAMP levels, yet they are able to efficiently respond to FF-MAS (Grondahl et al., 2003). Furthermore, cumulus cell-enclosed mouse oocytes cultured in the presence of cell permeable dbcAMP and stimulated by FSH still enter meiosis (Downs, 1995). Similarly, the inhibitory influence of xanthine derivatives on the starfish oocytes’ response to 1-MA can also be overcome by increasing 1-MA concentration (Doree et al., 1976). Furthermore, raising cAMP in starfish or mouse oocytes by applying forskolin delays GVBD in response to 1-MA or FF-MAS respectively but does not block it completely (Hegele-Hartung et al., 1999; Meijer and Zarutskie, 1987). Finally, injection of cholera toxin into starfish oocyte is effective in increasing the levels of cytoplasmic cAMP; still, it does not block 1-MA-induced maturation (reviewed in Kanatani, 1985). Artificially decreasing the levels of cAMP in the oocyte is useful to test whether this in itself would be sufficient to cause oocyte maturation. Culture of mouse oocytes outside of their follicular environment leads to a decrease in cAMP levels. For spontaneous maturation of mouse oocytes to occur, the drop in cytoplasmic cAMP concentration needs to take place to 50–60% of initial levels (Schultz et al., 1983). Another way to decrease the levels of cAMP is by stimulation of heterologous Gi-coupled serotonin receptor overexpressed in the oocyte. In Xenopus, stimulation of 5-HT1aR by serotonin lowers cytoplasmic cAMP concentration, but never induces GVBD (Noh and Han, 1998). However, this experiment is flawed in that the resting values of cAMP in HT1aR-expressing oocytes appear to be twice that of controls, and serotonin treatment only serves to bring cAMP closer to the control levels, but never below those. Finally, expression of cAMPdegrading enzymes such as mouse phosphodiesterase 3 or rat phosphodiesterase 4 in Xenopus oocytes induces GVBD to the same extent as progesterone (Andersen et al., 1998; Conti et al., 2002). Thus, it appears that a physiological decrease in cAMP is sufficient by itself to induce a reentry of oocyte into meiosis.
3. The Regulation of Oocyte Maturation
95
The physiological mechanisms regulating cAMP levels in the oocytes are currently unknown. In different animals, evidence favors cAMP regulation at the level of synthesis (by regulating adenylate cyclase), as well as at the level of degradation (by regulating phosphodiesterase, PDE). In starfish, Xenopus, and fish, a decrease in adenylate cyclase activity was proposed to be responsible for the decrease in cAMP (Meijer and Zarutskie, 1987; Schorderet-Slatkine et al., 1982; Thomas et al., 2002). The candidate regulators of adenylate cyclase are G-proteins, and generally Gs subunits activate adenylate cyclase, while Gi subunits inhibit its function. However, this simplistic view has been altered during the last decade with the identification of multiple different types of adenylate cyclases that are responsive to multiple regulators in addition to G subunits, e.g., G subunits, calcium levels, and protein phosphorylation (Simonds, 1999). Thus, characterization of adenylate cyclases present in the oocyte is necessary to gain understanding in regulation of oocyte maturation. Recently, expression of type 3 adenylate cyclase (AC3) was detected in the oocytes of rat and mouse (Horner et al., 2003). Furthermore, the AC3 knockout mice are characterized by a significant percentage of ovarian oocytes escaping meiotic arrest. AC3 is activated by Gs protein and is inhibited by an increase in intracellular calcium through phosphorylation by CaMKII (Simonds, 1999). Indeed, AC3 activity is detected in rodent oocytes, raising intracellular calcium can inhibit it, and this inhibition can be rescued by in turn inhibiting CaMKII (Horner et al., 2003). These observations suggest a possible mechanism for the contribution of calcium signaling to the regulation of meiotic resumption through its effects on cAMP levels. The degradation of cAMP is also important for the induction of oocyte maturation. Type 3 PDE appears to be active in mouse oocytes, but its regulation has not been extensively explored (Conti et al., 2002; Webb et al., 2002). This activity is required both in vitro and in vivo, as mice injected with PDE3 inhibitors ovulate GV-arrested oocytes (reviewed in Conti et al., 2002). Inhibition of oocyte maturation by phosphodiesterase inhibitors does not distinguish between the impact of constitutive and regulated PDE activity, making important the analysis of PDE3 activity profile during maturation. A significant increase in the PDE3 activity is detected in rat oocytes preceding both spontaneous and LH-induced meiosis (Richard et al., 2001). PDE activity increases as well in fish oocytes upon stimulation of oocyte maturation with DHP (Haider, 2003). However, in Xenopus, the basal oocyte level of PDE activity does not change after progesterone treatment (Sadler and Maller, 1987). It appears that animals can utilize a decrease in cAMP production, or increase in cAMP degradation, or a combination thereof to stimulate oocytes’ meiotic resumption.
96
Voronina and Wessel
Thus, cAMP is clearly not the only intracellular messenger involved in the initiation of meiotic divisions, but lowering its concentration is important for oocyte maturation to occur. The hormonal action on the oocytes might have the effect of altering the oocytes’ sensitivity to cAMP levels in addition to decreasing these levels. The data reviewed here are consistent with the hypothesis that meiotic reinitiation in response to hormonal treatment is triggered by a stimulatory signal that can be effective despite sustained inhibitory influence of cAMP, while decreasing cAMP levels is important for spontaneous oocyte maturation. Maturation may be influenced by a threshold effect of cytoplasmic cAMP concentration, such that supraphysiological levels of cAMP are inhibitory for the meiotic progression. If, however, cAMP is only slightly elevated, the hormone-induced signaling is able to overcome the inhibition and promote maturation. In this case, normally physiological conditions reflect a balance of cAMP and hormonal activation. Such a threshold effect would impart on the oocyte an all-ornothing property for meiotic progression, analogous to MPF activation.
D. PI3K Phosphatidylinositol 3-kinase (PI3K) is another proposed mediator of hormone-induced maturation. This kinase converts the plasma membrane lipid phosphatidylinositol-4,5-bisphosphate {PI(4,5)P2} to phosphatidylinositol-3,4,5-trisphosphate {PI(3,4,5)P3} (reviewed in Cantley, 2002). Signaling proteins with plekstrin-homology domains accumulate at the sites of PI3K activation by directly binding to its product PI(3,4,5)P3. Particularly important binding partners are two serine–threonine kinases, Akt/PKB and PDK1 (phosphoinositide-dependent kinase 1). Association with PI(3,4,5)P3 at the membrane brings these two kinases in proximity, and results in phosphorylation and activation of Akt/PKB by PDK1. Akt/PKB activity then leads to multiple cellular changes, including cell polarization, cell cycle entry, cell survival, and oocyte maturation (Cantley, 2002). In Xenopus, activated forms of PI3K can induce maturation, and PI3K activity increases about 2-fold in progesterone-treated oocytes (reviewed in Ferrell, 1999). Moreover, microinjection of a dominant-negative fragment of PI3K inhibits progesterone-induced maturation. Potentially PI3K could contribute to oocyte activation through the downstream Akt/PKB kinase, which stimulates PDE3. Consistent with this notion, expression of constitutively active Akt/PKB kinase is sufficient to induce Xenopus oocyte maturation through PDE3 activation (Andersen et al., 1998). However, an inhibitor of PI3K, wortmannin, does not inhibit progesterone-induced maturation in frog, so the role of PI3K in this species is not altogether clear (Ferrell, 1999). On the other hand, PI3K contribution appears more definite
3. The Regulation of Oocyte Maturation
97
in starfish oocyte maturation. It is thought to be activated by the heterotrimeric G-protein subunits, and PI3K inhibitors, wortmannin and LY294002, inhibit 1-MA-induced maturation (Sadler and Ruderman, 1998). The downstream kinase, Akt/PKB is activated in 1-MA stimulated oocytes, and this activity is necessary for 1-MA-induced maturation (Okumura et al., 2002). Accordingly, constitutively active Akt/PKB protein is able to induce 1-MA-independent oocyte maturation. The proposed target of Akt/PKB is Myt1, the kinase that phosphorylates and inhibits cyclin B-associated cdk1. Myt1 activity decreases upon phosphorylation by Akt/PKB, thereby shifting the equilibrium towards production of active cyclin B-cdk1 complexes and resumption of meiosis (Fig. 7).
E. MAPK Activation of the MAPK (mitogen-activated protein kinase) cascade is associated with resumption of oocyte meiosis in many different species. Mos codes for a serine/threonine kinase that functions upstream of MAP kinase, and its normal expression is confined to germ cells. Oocytes accumulate mos mRNA, which is translated into protein during meiotic maturation and causes activation of downstream MAPK. For a number of years, it was inferred that the signal inducing meiotic resumption is transduced through
Figure 7 Signaling pathways regulating the state of MPF activation in the maturing oocyte (details in text). In red: protein kinase A activity is inhibitory to MPF activation. In green: PI3K activity resulting in Akt/PKB activation leads to activation of MPF (see color plate).
98
Voronina and Wessel
MAPK and requires mos protein accumulation, based on the ability of mos and its downstream targets to induce hormone-independent maturation (reviewed in Dupre et al., 2002). Furthermore, initial experiments in Xenopus suggested that mos synthesis is required for GVBD. However, later data obtained in various animals and with many techniques such as specific MAPK inhibitors (starfish, Sadler and Ruderman, 1998), transgenic mos RNAi (mouse, Stein et al., 2003), mos gene knockouts, mos DNA antisense (mouse, O’Keefe et al., 1989), and morpholino antisense treatments (Xenopus, Dupre et al., 2002) suggest that MAPK activation is not required for the resumption of meiosis (or GVBD) in any animal. Instead, MAPK activity plays a role in acceleration of MPF activation in the oocyte at meiotic resumption (Dupre et al., 2002), and is needed for suppression of interphase (DNA replication) between two divisions of meiosis and subsequent arrest at the second metaphase of meiosis (Dupre et al., 2002; O’Keefe et al., 1989; Stein et al., 2003).
VI. Concluding Remarks In this review of the signaling networks bringing about oocyte maturation of different species, we encountered significant variability among animals at almost every level. The maturation-inducing hormones range from a nucleotide (starfish) and complex sterols (vertebrates) to a protein (nematode), or even to the uncertainty of whether such physiological inducers even exist (mouse). The receptors of these hormones are mostly unknown, but the evidence for them ranges from a receptor tyrosine kinase (nematode), to a seven-transmembrane G-protein-coupled receptor (fish and starfish), to a cytoplasmic transcription factor hormone receptor, or an unidentified membrane receptor (Xenopus). Consequently, the identity of the signaling mediators varies, and the heterotrimeric G-proteins, which are strongly implicated in the induction of oocyte maturation in certain species (starfish), appear inhibitory to maturation in other species (mouse, Xenopus), or not involved in regulation of oocyte maturation in yet other species (nematode). Significant discrepancies are encountered even in the involvement of the particular second messengers such as levels of cAMP, calcium, or phosphatidylinositol-3,4,5-trisphosphate (product of PI3K activity) in the regulation of meiotic resumption. The common thread to all these regulatory mechanisms appears to be the final downstream requirement to activate the cyclin B-cdk1 complex (MPF). To identify more common principles of oocyte maturation regulation, it would be useful to compare and contrast the regulatory mechanisms present in evolutionary close organisms, which nevertheless exhibit noticeable differences in the progression of oocyte maturation. An example of a pair
3. The Regulation of Oocyte Maturation
99
suitable for comparative studies is starfish and sea urchin. Since they are both members of the echinoderm phyla, we would expect to find similarities in their oocytes. However, several aspects are divergent. For example, the starfish stores its gametes as oocytes, and resumption of oocyte meiosis coincides (and is caused by the same hormone, 1-MA) with ovulation and release of gametes by the animal. In contrast, the sea urchin stores its gametes as mature eggs, which have already resumed and completed meiotic division within the gonad. Accordingly, fertilization occurs at different meiotic stages in these animals. Furthermore, meiotic divisions are relatively quick in starfish (90 min), while in sea urchin meiosis takes considerably longer (9 h). Sea urchin eggs have been an extremely popular model for studies of fertilization and early development. Nevertheless, the regulation of sea urchin oocyte maturation has not been thoroughly scrutinized despite providing an important advantage of ‘‘divorcing’’ oocyte maturation and egg activation. We find it very strange that something as fundamental as meiotic maturation and fertilization appears to have diverged in mechanisms so greatly between animals, when other key developmental events are far better conserved. Perhaps this divergence in germ cells reflects the positive selection pressure on the genes involved in reproduction relative to the rest of the genome (Swanson and Vacquier, 2002). Finding other highly divergent genes in the future may be suggestive of their involvement in this process. In addition, with recent advances in genomics, development of databases comparing gene expression patterns in the oocytes of various animals might help in identifying new important molecular regulators of oogenesis on the basis of evolutionarily conserved pattern of expression (Schlecht and Primig, 2003).
Acknowledgments The authors would like to thank Laurinda Jaffe for the helpful comments on this manuscript. Our research is supported by grants from the NSF and the NIH.
References Adams, I. R., and McLaren, A. (2002). Sexually dimorphic development of mouse primordial germ cells: switching from oogenesis to spermatogenesis. Development 129, 1155–1164. Alberts, A. S. (2002). Diaphanous-related Formin homology proteins. Curr. Biol. 12, R796. Allworth, A. E., Hildebrandt, J. D., and Ziomek, C. A. (1990). Differential regulation of G protein subunit expression in mouse oocytes, eggs, and early embryos. Dev. Biol. 142, 129–137.
100
Voronina and Wessel
Andersen, C. B., Roth, R. A., and Conti, M. (1998). Protein kinase B/Akt induces resumption of meiosis in Xenopus oocytes. J. Biol. Chem. 273, 18705–18708. Baker, A. M., Roberts, T. M., and Stewart, M. (2002). 2.6 A resolution crystal structure of helices of the motile major sperm protein (MSP) of Caenorhabditis elegans. J. Mol. Biol. 319, 491–499. Bayaa, M., Booth, R. A., Sheng, Y., and Liu, X. J. (2000). The classical progesterone receptor mediates Xenopus oocyte maturation through a nongenomic mechanism. Proc. Natl. Acad. Sci. USA 97, 12607–12612. Berg, L. K., and Wessel, G. M. (1997). Cortical granules of the sea urchin translocate early in oocyte maturation. Development 124, 1845–1850. Bobinnec, Y., Marcaillou, C., Morin, X., and Debec, A. (2003). Dynamics of the endoplasmic reticulum during early development of Drosophila melanogaster. Cell Motil. Cytoskeleton 54, 217–225. Bomsel-Helmreich, O., Huyen, L. V., Durand-Gasselin, I., Salat-Baroux, J., and Antoine, J. M. (1987). Timing of nuclear maturation and cumulus dissociation in human oocytes stimulated with clomiphene citrate, human menopausal gonadotropin, and human chorionic gonadotropin. Fertil. Steril. 48, 586–595. Boonyaratanakornkit, V., Scott, M. P., Ribon, V., Sherman, L., Anderson, S. M., Maller, J. L., Miller, W. T., and Edwards, D. P. (2001). Progesterone receptor contains a proline-rich motif that directly interacts with SH3 domains and activates c-Src family tyrosine kinases. Mol. Cell 8, 269–280. Bornslaeger, E. A., Poueymirou, W. T., Mattei, P., and Schultz, R. M. (1986). Effects of protein kinase C activators on germinal vesicle breakdown and polar body emission of mouse oocytes. Exp. Cell Res. 165, 507–517. Boyle, J. A., and Ernst, S. G. (1989). Sea urchin oocytes possess elaborate cortical arrays of microfilaments, microtubules, and intermediate filaments. Dev. Biol. 134, 72–84. Byskov, A. G., Andersen, C. Y., and Leonardsen, L. (2002). Role of meiosis activating sterols, MAS, in induced oocyte maturation. Mol. Cell Endocrinol. 187, 189–196. Byskov, A. G., Andersen, C. Y., Nordholm, L., Thogersen, H., Xia, G., Wassmann, O., Andersen, J. V., Guddal, E., and Roed, T. (1995). Chemical structure of sterols that activate oocyte meiosis. Nature 374, 559–562. Cantley, L. C. (2002). The phosphoinositide 3-kinase pathway. Science 296, 1655–1657. Carabatsos, M. J., Sellitto, C., Goodenough, D. A., and Albertini, D. F. (2000). Oocytegranulosa cell heterologous gap junctions are required for the coordination of nuclear and cytoplasmic meiotic competence. Dev. Biol. 226, 167–179. Carroll, J., Swann, K., Whittingham, D., and Whitaker, M. (1994). Spatiotemporal dynamics of intracellular [Ca2 þ ]i oscillations during the growth and meiotic maturation of mouse oocytes. Development 120, 3507–3517. Cerda, J. L., Petrino, T. R., and Wallace, R. A. (1993). Functional heterologous gap junctions in Fundulus ovarian follicles maintain meiotic arrest and permit hydration during oocyte maturation. Dev. Biol. 160, 228–235. Cheung, A., Swann, K., and Carroll, J. (2000). The ability to generate normal Ca(2 þ ) transients in response to spermatozoa develops during the final stages of oocyte growth and maturation. Hum. Reprod. 15, 1389–1395. Chiba, K., and Hoshi, M. (1995). G-protein-mediated signal transduction for meiosis reinitiation in starfish oocyte. Prog. Cell Cycle Res. 1, 255–263. Chiba, K., Kado, R. T., and Jaffe, L. A. (1990). Development of calcium release mechanisms during starfish oocyte maturation. Dev. Biol. 140, 300–306. Chiba, K., Longo, F. J., Kontani, K., Katada, T., and Hoshi, M. (1995). A periodic network of G protein subunit coexisting with cytokeratin filament in starfish oocytes. Dev. Biol. 169, 415–420.
3. The Regulation of Oocyte Maturation
101
Chiba, K., Tadenuma, H., Matsumoto, M., Takahashi, K., Katada, T., and Hoshi, M. (1992). The primary structure of the subunit of a starfish guanosine-nucleotide-binding regulatory protein involved in 1-methyladenine-induced oocyte maturation. Eur. J. Biochem. 207, 833–838. Chuma, S., and Nakatsuji, N. (2001). Autonomous transition into meiosis of mouse fetal germ cells in vitro and its inhibition by gp130-mediated signaling. Dev. Biol. 229, 468–479. Cicirelli, M. F., Pelech, S. L., and Krebs, E. G. (1988). Activation of multiple protein kinases during the burst in protein phosphorylation that precedes the first meiotic cell division in Xenopus oocytes. J. Biol. Chem. 263, 2009–2019. Coffman, C. R., Strohm, R. C., Oakley, F. D., Yamada, Y., Przychodzin, D., and Boswell, R. E. (2002). Identification of X-linked genes required for migration and programmed cell death of Drosophila melanogaster germ cells. Genetics 162, 273–284. Connors, S. A., Kanatsu-Shinohara, M., Schultz, R. M., and Kopf, G. S. (1998). Involvement of the cytoskeleton in the movement of cortical granules during oocyte maturation, and cortical granule anchoring in mouse eggs. Dev. Biol. 200, 103–115. Conti, M., Andersen, C. B., Richard, F., Mehats, C., Chun, S. Y., Horner, K., Jin, C., and Tsafriri, A. (2002). Role of cyclic nucleotide signaling in oocyte maturation. Mol. Cell Endocrinol. 187, 153–159. Cork, R. J., Taylor, M., Varnold, R. L., Smith, L. D., and Robinson, K. R. (1990). Microinjected GTPS inhibits progesterone-induced maturation of Xenopus oocytes. Dev. Biol. 141, 447–450. Dale, B. (1994). Oocyte activation in invertebrates and humans. Zygote 2, 373–377. Deng, M., Kishikawa, H., Yanagimachi, R., Kopf, G. S., Schultz, R. M., and Williams, C. J. (2003). Chromatin-mediated cortical granule redistribution is responsible for the formation of the cortical granule-free domain in mouse eggs. Dev. Biol. 257, 166–176. Doitsidou, M., Reichman-Fried, M., Stebler, J., Koprunner, M., Dorries, J., Meyer, D., Esguerra, C. V., Leung, T., and Raz, E. (2002). Guidance of primordial germ cell migration by the chemokine SDF-1. Cell 111, 647–659. Doree, M., Guerrier, P., and Leonard, N. J. (1976). Hormonal control of meiosis: specificity of the 1-methyladenine receptors in starfish oocytes. Proc. Natl. Acad. Sci. USA 73, 1669–1673. Downs, S. M. (1995). The influence of glucose, cumulus cells, and metabolic coupling on ATP levels and meiotic control in the isolated mouse oocyte. Dev. Biol. 167, 502–512. Downs, S. M. (2001). A gap-junction-mediated signal, rather than an external paracrine factor, predominates during meiotic induction in isolated mouse oocytes. Zygote 9, 71–82. Downs, S. M., Buccione, R., and Eppig, J. J. (1992). Modulation of meiotic arrest in mouse oocytes by guanyl nucleotides and modifiers of G-proteins. J. Exp. Zool. 262, 391–404. Downs, S. M., Cottom, J., and Hunzicker-Dunn, M. (2001). Protein kinase C and meiotic regulation in isolated mouse oocytes. Mol. Reprod. Dev. 58, 101–115. Downs, S. M., Daniel, S. A., and Eppig, J. J. (1988). Induction of maturation in cumulus cellenclosed mouse oocytes by follicle-stimulating hormone and epidermal growth factor: evidence for a positive stimulus of somatic cell origin. J. Exp. Zool. 245, 86–96. Duckworth, B. C., Weaver, J. S., and Ruderman, J. V. (2002). G2 arrest in Xenopus oocytes depends on phosphorylation of cdc25 by protein kinase A. Proc. Natl. Acad. Sci. USA 99, 16794–16799. Dupre, A., Jessus, C., Ozon, R., and Haccard, O. (2002). Mos is not required for the initiation of meiotic maturation in Xenopus oocytes. EMBO J. 21, 4026–4036. Eppig, J. J., and Downs, S. M. (1988). Gonadotropin-induced murine oocyte maturation in vivo is not associated with decreased cyclic adenosine monophosphate in the oocyte-cumulus cell complex. Gamete. Res. 20, 125–131. Eppig, J. J., O’Brien, M., and Wigglesworth, K. (1996). Mammalian oocyte growth and development in vitro. Mol. Reprod. Dev. 44, 260–273.
102
Voronina and Wessel
Eppig, J. J., Wigglesworth, K., and Pendola, F. L. (2002). The mammalian oocyte orchestrates the rate of ovarian follicular development. Proc. Natl. Acad. Sci. USA 99, 2890–2894. Faerge, I., Terry, B., Kalous, J., Wahl, P., Lessl, M., Ottesen, J. L., Hyttel, P., and Grondahl, C. (2001). Resumption of meiosis induced by meiosis-activating sterol has a different signal transduction pathway than spontaneous resumption of meiosis in denuded mouse oocytes cultured in vitro. Biol. Reprod. 65, 1751–1758. Ferby, I., Blazquez, M., Palmer, A., Eritja, R., and Nebreda, A. R. (1999). A novel p34(cdc2)binding and activating protein that is necessary and sufficient to trigger G(2)/M progression in Xenopus oocytes. Genes Dev. 13, 2177–2189. Ferrell, J. E., Jr. (1999). Xenopus oocyte maturation: new lessons from a good egg. Bioessays 21, 833–842. Ferrell, J. E., Jr. (2002). Self-perpetuating states in signal transduction: positive feedback, double-negative feedback and bistability. Curr. Opin. Cell Biol. 14, 140–148. FitzHarris, G., Marangos, P., and Carroll, J. (2003). Cell cycle-dependent regulation of structure of endoplasmic reticulum and inositol 1,4,5-trisphosphate-induced Ca(2 þ ) release in mouse oocytes and embryos. Mol. Biol. Cell 14, 288–301. Fortune, J. E., and Tsang, P. C. (1981). Production of androgen and estradiol-17 beta by Xenopus ovaries treated with gonadotropins in vitro. Gen. Comp. Endocrinol. 43, 234–242. Gallo, C. J., Hand, A. R., Jones, T. L., and Jaffe, L. A. (1995). Stimulation of Xenopus oocyte maturation by inhibition of the G-protein S subunit, a component of the plasma membrane and yolk platelet membranes. J. Cell Biol. 130, 275–284. Glazier, A. M., Mate, K. E., and Rodger, J. C. (2002). In vitro and in vivo maturation of oocytes from gonadotrophin-treated brushtail possums. Mol. Reprod. Dev. 62, 504–512. Godeau, J. F., Schorderet-Slatkine, S., Hubert, P., and Baulieu, E. E. (1978). Induction of maturation in Xenopus laevis oocytes by a steroid linked to a polymer. Proc. Natl. Acad. Sci. USA 75, 2353–2357. Goodhardt, M., Ferry, N., Buscaglia, M., Baulieu, E. E., and Hanoune, J. (1984). Does the guanine nucleotide regulatory protein Ni mediate progesterone inhibition of Xenopus oocyte adenylate cyclase? EMBO J. 3, 2653–2657. Granot, I., and Dekel, N. (1994). Phosphorylation and expression of connexin-43 ovarian gap junction protein are regulated by luteinizing hormone. J. Biol. Chem. 269, 30502–30509. Grondahl, C., Breinholt, J., Wahl, P., Murray, A., Hansen, T. H., Faerge, I., Stidsen, C. E., Raun, K., and Hegele-Hartung, C. (2003). Physiology of meiosis-activating sterol: endogenous formation and mode of action. Hum. Reprod. 18, 122–129. Grondahl, C., Lessl, M., Faerge, I., Hegele-Hartung, C., Wassermann, K., and Ottesen, J. L. (2000). Meiosis-activating sterol-mediated resumption of meiosis in mouse oocytes in vitro is influenced by protein synthesis inhibition and cholera toxin. Biol. Reprod. 62, 775–780. Haider, S. (2003). Cyclic AMP level and phosphodiesterase activity during 17,20-dihydroxy4-pregnen-3-one induction and theophylline inhibition of oocyte maturation in the catfish, Clarias batrachus. Comp. Biochem. Physiol. A Mol. Integr. Physiol. 134, 267–274. Hake, L. E., and Richter, J. D. (1997). Translational regulation of maternal mRNA. Biochem. Biophys. Acta. 1332, M31–M38. Han, J. K., and Lee, S. K. (1995). Reducing PIP2 hydrolysis, Ins(1,4,5)P3 receptor availability, or calcium gradients inhibits progesterone-stimulated Xenopus oocyte maturation. Biochem. Biophys. Res. Commun. 217, 931–939. Hegele-Hartung, C., Grutzner, M., Lessl, M., Grondahl, C., Ottesen, J., and Brannstrom, M. (2001). Activation of meiotic maturation in rat oocytes after treatment with follicular fluid meiosis-activating sterol in vitro and ex vivo. Biol. Reprod. 64, 418–424. Hegele-Hartung, C., Kuhnke, J., Lessl, M., Grondahl, C., Ottesen, J., Beier, H. M., Eisner, S., and Eichenlaub-Ritter, U. (1999). Nuclear and cytoplasmic maturation of mouse
3. The Regulation of Oocyte Maturation
103
oocytes after treatment with synthetic meiosis-activating sterol in vitro. Biol. Reprod. 61, 1362–1372. Heil-Chapdelaine, R. A., and Otto, J. J. (1996). Characterization of changes in F-actin during maturation of starfish oocytes. Dev. Biol. 177, 204–216. Henson, J. H., Beaulieu, S. M., Kaminer, B., and Begg, D. A. (1990). Differentiation of a calsequestrin-containing endoplasmic reticulum during sea urchin oogenesis. Dev. Biol. 142, 255–269. Hochegger, H., Klotzbucher, A., Kirk, J., Howell, M., le Guellec, K., Fletcher, K., Duncan, T., Sohail, M., and Hunt, T. (2001). New B-type cyclin synthesis is required between meiosis I and II during Xenopus oocyte maturation. Development 128, 3795–3807. Horner, K., Livera, G., Hinckley, M., Trinh, K., Storm, D., and Conti, M. (2003). Rodent oocytes express an active adenylyl cyclase required for meiotic arrest. Dev. Biol. 258, 385–396. Hubner, K., Fuhrmann, G., Christenson, L. K., Kehler, J., Reinbold, R., De La Fuente, R., Wood, J., Strauss III, J. F., Boiani, M., and Scholer, H. R. (2003). Derivation of oocytes from mouse embryonic stem cells. Science 300, 1251–1250. Hunt, T., Luca, F. C., and Ruderman, J. V. (1992). The requirements for protein synthesis and degradation, and the control of destruction of cyclins A and B in the meiotic and mitotic cell cycles of the clam embryo. J. Cell Biol. 116, 707–724. Iwasaki, H., Chiba, K., Uchiyama, T., Yoshikawa, F., Suzuki, F., Ikeda, M., Furuichi, T., and Mikoshiba, K. (2002). Molecular characterization of the starfish inositol 1,4,5-trisphosphate receptor and its role during oocyte maturation and fertilization. J. Biol. Chem. 277, 2763–2772. Jaffe, L. A., Gallo, C. J., Lee, R. H., Ho, Y. K., and Jones, T. L. (1993). Oocyte maturation in starfish is mediated by the -subunit complex of a G-protein. J. Cell Biol. 121, 775–783. Jaffe, L. A., and Terasaki, M. (1994). Structural changes in the endoplasmic reticulum of starfish oocytes during meiotic maturation and fertilization. Dev. Biol. 164, 579–587. Johnson, A. D., Drum, M., Bachvarova, R. F., Masi, T., White, M. E., and Crother, M. I. (2003). Evolution of predetermined germ cells in vertebrate embryos: implications for macroevolution. Evolution & Developement 5, 414–431. Jones, J., and Schultz, R. M. (1990). Pertussis toxin-catalyzed ADP-ribosylation of a G protein in mouse oocytes, eggs, and preimplantation embryos: developmental changes and possible functional roles. Dev. Biol. 139, 250–262. Kalinowski, R. R., Jaffe, L. A., Foltz, K. R., and Giusti, A. F. (2003). A receptor linked to a G(i)-family G-protein functions in initiating oocyte maturation in starfish but not frogs. Dev. Biol. 253, 139–149. Kanatani, H. (1969). Induction of spawning and oocyte maturation by L-methyl-adenine in starfishes. Exp. Cell Res. 57, 333–337. Kanatani, H. (1985). Oocyte growth and maturation in starfish, in Biology of Fertilization, Vol. 1, edited by C. B. Metz, and A. Monroy. Orlando, FL: Academic Press, pp. 119–140. Kidder, G. M., and Mhawi, A. A. (2002). Gap junctions and ovarian folliculogenesis. Reproduction 123, 613–620. Kim, N. H., Cho, S. K., Choi, S. H., Kim, E. Y., Park, S. P., and Lim, J. H. (2000). The distribution and requirements of microtubules and microfilaments in bovine oocytes during in vitro maturation. Zygote 8, 25–32. Kishimoto, T. (1999). Activation of MPF at meiosis reinitiation in starfish oocytes. Dev. Biol. 214, 1–8. Kline, D., Kopf, G. S., Muncy, L. F., and Jaffe, L. A. (1991). Evidence for the involvement of a pertussis toxin-insensitive G-protein in egg activation of the frog, Xenopus laevis. Dev. Biol. 143, 218–229.
104
Voronina and Wessel
Knaut, H., Werz, C., Geisler, R., The Tubingen 2000 Screen Consortium, and NussleinVolhard, C. (2003). A zebrafish homologue of the chemokine receptor Cxcr4 is a germ-cell guidance receptor. Nature 421, 279–282. Kobayashi, H., Minshull, J., Ford, C., Golsteyn, R., Poon, R., and Hunt, T. (1991). On the synthesis and destruction of A- and B-type cyclins during oogenesis and meitotic maturation in Xenopus laevis. J. Cell. Biol. 114, 755–765. Kroll, S. D., Omri, G., Landau, E. M., and Iyengar, R. (1991). Activated subunit of Go protein induces oocyte maturation. Proc. Natl. Acad. Sci. USA 88, 5182–5186. Kunwar, P. S., and Lehmann, R. (2003). Developmental biology: Germ-cell attraction. Nature 421, 226–227. Kuwabara, P. E. (2003). The multifaceted C. elegans major sperm protein: an ephrin signaling antagonist in oocyte maturation. Genes Dev. 17, 155–161. Labbe, J. C., Capony, J. P., Caput, D., Cavadore, J. C., Derancourt, J., Kaghad, M., Lelias, J. M., Picard, A., and Doree, M. (1989). MPF from starfish oocytes at first meiotic metaphase is a heterodimer containing one molecule of cdc2 and one molecule of cyclin B. EMBO J. 8, 3053–3058. LaFleur, G. J., Jr., Horiuchi, Y., and Wessel, G. M. (1998). Sea urchin ovoperoxidase: oocytespecific member of a heme-dependent peroxidase superfamily that functions in the block to polyspermy. Mech. Dev. 70, 77–89. Laidlaw, M., and Wessel, G. M. (1994). Cortical granule biogenesis is active throughout oogenesis in sea urchins. Development 120, 1325–1333. Leader, B., Lim, H., Carabatsos, M. J., Harrington, A., Ecsedy, J., Pellman, D., Maas, R., and Leder, P. (2002). Formin-2, polyploidy, hypofertility and positioning of the meiotic spindle in mouse oocytes. Nat. Cell Biol. 4, 921–928. Lenormand, J. L., Dellinger, R. W., Knudsen, K. E., Subramani, S., and Donoghue, D. J. (1999). Speedy: a novel cell cycle regulator of the G2/M transition. EMBO J. 18, 1869–1877. Leonardsen, L., Wiersma, A., Baltsen, M., Byskov, A. G., and Andersen, C. Y. (2000). Regulation of spontaneous and induced resumption of meiosis in mouse oocytes by different intracellular pathways. J. Reprod. Fertil. 120, 377–383. Lew, D. J. (2002). Formin’ actin filament bundles. Nat. Cell Biol. 4, E29–E30. Lippincott-Schwartz, J. (2002). Cell biology: ripping up the nuclear envelope. Nature 416, 31–32. Lohka, M. J., Hayes, M. K., and Maller, J. L. (1988). Purification of maturation-promoting factor, an intracellular regulator of early mitotic events. Proc. Natl. Acad. Sci. USA 85, 3009–3013. Longo, F., and Schuetz, A. W. (1982). Male pronuclear development in starfish oocytes treated with 1-methyladenine. Biol. Bull. 163, 453–464. Lutz, L. B., Cole, L. M., Gupta, M. K., Kwist, K. W., Auchus, R. J., and Hammes, S. R. (2001). Evidence that androgens are the primary steroids produced by Xenopus laevis ovaries and may signal through the classical androgen receptor to promote oocyte maturation. Proc. Natl. Acad. Sci. USA 98, 13728–13733. Lutz, L.B., Jamnongjit, M., Yang, W.H., Jahani, D., Gill, A., and Hammes, S.R. (2003). Selective modulation of genomic and nongenomic androgen responses by androgen receptor ligands. Mol. Endocrinol. 17, 1106–1116. Lutz, L. B., Kim, B., Jahani, D., and Hammes, S. R. (2000). G protein subunits inhibit nongenomic progesterone-induced signaling and maturation in Xenopus laevis oocytes. Evidence for a release of inhibition mechanism for cell cycle progression. J. Biol. Chem. 275, 41512–41520. Maller, J. L. (1990). MPF and cell cycle control. Adv. Second Messenger Phosphoprotein Res. 24, 323–328.
3. The Regulation of Oocyte Maturation
105
Maller, J. L. (2001). The elusive progesterone receptor in Xenopus oocytes. Proc. Natl. Acad. Sci. USA 98, 8–10. Maro, B., and Verlhac, M. H. (2002). Polar body formation: new rules for asymmetric divisions. Nat. Cell Biol. 4, E281–E283. Masui, Y., and Clarke, H. J. (1979). Oocyte maturation. Int. Rev. Cytol. 57, 185–282. Masui, Y., and Markert, C. L. (1971). Cytoplasmic control of nuclear behavior during meiotic maturation of frog oocytes. J. Exp. Zool. 177, 129–145. Matova, N., and Cooley, L. (2001). Comparative aspects of animal oogenesis. Dev. Biol. 231, 291–320. Matzuk, M. M., Burns, K. H., Viveiros, M. M., and Eppig, J. J. (2002). Intercellular communication in the mammalian ovary: oocytes carry the conversation. Science 296, 2178–2180. McLaren, A. (2003). Primordial germ cells in the mouse. Dev. Biol. in press. Mehlmann, L. M., Jones, T. L., and Jaffe, L. A. (2002). Meiotic arrest in the mouse follicle maintained by a Gs protein in the oocyte. Science 297, 1343–1345. Mehlmann, L. M., Mikoshiba, K., and Kline, D. (1996). Redistribution and increase in cortical inositol 1,4,5-trisphosphate receptors after meiotic maturation of the mouse oocyte. Dev. Biol. 180, 489–498. Mehlmann, L. M., Terasaki, M., Jaffe, L. A., and Kline, D. (1995). Reorganization of the endoplasmic reticulum during meiotic maturation of the mouse oocyte. Dev. Biol. 170, 607–615. Meijer, L., and Zarutskie, P. (1987). Starfish oocyte maturation: 1-methyladenine triggers a drop of cAMP concentration related to the hormone-dependent period. Dev. Biol. 121, 306–315. Miller, M. A., Nguyen, V. Q., Lee, M. H., Kosinski, M., Schedl, T., Caprioli, R. M., and Greenstein, D. (2001). A sperm cytoskeletal protein that signals oocyte meiotic maturation and ovulation. Science 291, 2144–2147. Miller, M. A., Ruest, P. J., Kosinski, M., Hanks, S. K., and Greenstein, D. (2003). An Eph receptor sperm-sensing control mechanism for oocyte meiotic maturation in Caenorhabditis elegans. Genes Dev. 17, 187–200. Mita, M., Saneyoshi, M., Yoshikuni, M., and Nagahama, Y. (1999a). Methyl donor for 1-methyladenine biosynthesis in starfish ovarian follicle cells. Mol. Reprod. Dev. 54, 63–68. Mita, M., Yoshikuni, M., and Nagahama, Y. (1999b). 1-Methyladenine production from ATP by starfish ovarian follicle cells. Biochim. Biophys. Acta. 1428, 13–20. Mita, M., Yoshikuni, M., Nagahama, Y., Maekawa, M., Saito, M., and Saneyoshi, M. (2001). Interaction of N1-substituted adenines with 1-methyladenine receptors of starfish oocytes in induction of maturation. Comp. Biochem. Physiol. B Biochem. Mol. Biol. 130, 427–434. Miyazaki, S. (1979). Fast polyspermy block and activation potential. Electrophysiological bases for their changes during oocyte maturation of a starfish. Dev. Biol. 70, 341–354. Moor, R. M., and Crosby, I. M. (1986). Protein requirements for germinal vesicle breakdown in ovine oocytes. J. Embryol. Exp. Morphol. 94, 207–220. Morgan, D. O. (1997). Cyclin-dependent kinases: engines, clocks, and microprocessors. Annu. Rev. Cell Dev. Biol. 13, 261–291. Morita, Y., and Tilly, J. L. (1999). Oocyte apoptosis: like sand through an hourglass. Dev. Biol. 213, 1–17. Nagahama, Y., Yoshikuni, M., Yamashita, M., and Tanaka, M. (1994). Regulation of oocyte maturation in fish, in: Molecular endocrinology of fish, Vol. XIII, edited by N. M. Sherwood, and C. L. Hew, San Diego: Academic Press, pp. 393–439. Neaves, W. B. (1971). Intercellular bridges between follicle cells and oocyte in the lizard, Anolis carolinensis. Anat. Rec. 170, 285–301.
106
Voronina and Wessel
Nigg, E. A. (1995). Cyclin-dependent protein kinases: key regulators of the eukaryotic cell cycle. Bioessays 17, 471–480. Nigg, E. A. (2001). Mitotic kinases as regulators of cell division and its checkpoints. Nat. Rev. Mol. Cell Biol. 2, 21–32. Noh, S. J., and Han, J. K. (1998). Inhibition of the adenylyl cyclase and activation of the phosphatidylinositol pathway in oocytes through expression of serotonin receptors does not induce oocyte maturation. J. Exp. Zool. 280, 45–56. O’Keefe, S. J., Wolfes, H., Kiessling, A. A., and Cooper, G. M. (1989). Microinjection of antisense c-mos oligonucleotides prevents meiosis II in the maturing mouse egg. Proc. Natl. Acad. Sci. USA 86, 7038–7042. Okumura, E., Fukuhara, T., Yoshida, H., Hanada Si, S., Kozutsumi, R., Mori, M., Tachibana, K., and Kishimoto, T. (2002). Akt inhibits Myt1 in the signalling pathway that leads to meiotic G2/M-phase transition. Nat. Cell Biol. 4, 111–116. Parfenov, V. N., Davis, D. S., Pochukalina, G. N., Sample, C. E., Bugaeva, E. A., and Murti, K. G. (1995). Nuclear actin filaments and their topological changes in frog oocytes. Exp. Cell Res. 217, 385–394. Patel, R., Holt, M., Philipova, R., Moss, S., Schulman, H., Hidaka, H., and Whitaker, M. (1999). Calcium/calmodulin-dependent phosphorylation and activation of human Cdc25-C at the G2/M phase transition in HeLa cells. J. Biol. Chem. 274, 7958–7968. Patino, R., and Purkiss, R. T. (1993). Inhibitory effects of n-alkanols on the hormonal induction of maturation in follicle-enclosed Xenopus oocytes: implications for gap junctional transport of maturation-inducing steroid. Gen. Comp. Endocrinol. 91, 189–198. Paynton, B. V., and Bachvarova, R. (1994). Polyadenylation and deadenylation of maternal mRNAs during oocyte growth and maturation in the mouse. Mol. Reprod. Dev. 37, 172–180. Pellaz, V., and Schorderet-Slatkine, S. (1989). Evidence for a pertussis toxin-sensitive G protein involved in the control of meiotic reinitiation of Xenopus laevis oocytes. Exp. Cell Res. 183, 245–250. Pesty, A., Avazeri, N., and Lefevre, B. (1998). Nuclear calcium release by InsP3-receptor channels plays a role in meiosis reinitiation in the mouse oocyte. Cell Calcium 24, 239–251. Petrino, T. R., Lin, Y. W., Netherton, J. C., Powell, D. H., and Wallace, R. A. (1993). Steroidogenesis in Fundulus heteroclitus V.: purification, characterization, and metabolism of 17 alpha,20 beta-dihydroxy-4-pregnen-3-one by intact follicles and its role in oocyte maturation. Gen. Comp. Endocrinol. 92, 1–15. Pines, J. (1999). Four-dimensional control of the cell cycle. Nat. Cell Biol. 1, E73–E79. Qu, Y., and Dahl, G. (2002). Function of the voltage gate of gap junction channels: selective exclusion of molecules. Proc. Natl. Acad. Sci. USA 99, 697–702. Racowsky, C., Baldwin, K. V., Larabell, C. A., DeMarais, A. A., and Kazilek, C. J. (1989). Down-regulation of membrana granulosa cell gap junctions is correlated with irreversible commitment to resume meiosis in golden Syrian hamster oocytes. Eur. J. Cell Biol. 49, 244–251. Rando, O. J., Zhao, K., and Crabtree, G. R. (2000). Searching for a function for nuclear actin. Trends. Cell Biol. 10, 92–97. Raz, E. (2002). Primordial germ cell development in zebrafish. Semin. Cell Dev. Biol. 13, 489–495. Richard, F. J., Tsafriri, A., and Conti, M. (2001). Role of phosphodiesterase type 3A in rat oocyte maturation. Biol. Reprod. 65, 1444–1451. Romo, X., Hinrichs, M. V., Guzman, L., and Olate, J. (2002). Gs levels regulate Xenopus laevis oocyte maturation. Mol. Reprod. Dev. 63, 104–109. Ruan, B., Watanabe, S., Eppig, J. J., Kwoh, C., Dzidic, N., Pang, J., Wilson, W. K., and Schroepfer, G. J., Jr. (1998). Sterols affecting meiosis: novel chemical syntheses and the
3. The Regulation of Oocyte Maturation
107
biological activity and spectral properties of the synthetic sterols. J. Lipid Res. 39, 2005–2020. Ryabova, L. V., Betina, M. I., and Vassetzky, S. G. (1986). Influence of cytochalasin B on oocyte maturation in Xenopus laevis. Cell Differ. 19, 89–96. Sadler, K. C., and Ruderman, J. V. (1998). Components of the signaling pathway linking the 1methyladenine receptor to MPF activation and maturation in starfish oocytes. Dev. Biol. 197, 25–38. Sadler, S. E., and Maller, J. L. (1987). In vivo regulation of cyclic AMP phosphodiesterase in Xenopus oocytes. Stimulation by insulin and insulin-like growth factor 1. J. Biol. Chem. 262, 10644–10650. Sadler, S. E., Maller, J. L., and Cooper, D. M. (1984). Progesterone inhibition of Xenopus oocyte adenylate cyclase is not mediated via the Bordetella pertussis toxin substrate. Mol. Pharmacol. 26, 526–531. Saitou, M., Barton, S. C., and Surani, M. A. (2002). A molecular programme for the specification of germ cell fate in mice. Nature 418, 293–300. Santella, L., De Riso, L., Gragnaniello, G., and Kyozuka, K. (1999). Cortical granule translocation during maturation of starfish oocytes requires cytoskeletal rearrangement triggered by InsP3-mediated Ca2 þ release. Exp. Cell Res. 248, 567–574. Santella, L., and Kyozuka, K. (1994). Reinitiation of meiosis in starfish oocytes requires an increase in nuclear Ca2 þ . Biochem. Biophys. Res. Commun. 203, 674–680. Sasaki, K., and Chiba, K. (2001). Fertilization blocks apoptosis of starfish eggs by inactivation of the MAP kinase pathway. Dev. Biol. 237, 18–28. Schatten, G. (1994). The centrosome and its mode of inheritance: the reduction of the centrosome during gametogenesis and its restoration during fertilization. Dev. Biol. 165, 299–335. Schatten, G., and Schatten, H. (1987). Cytoskeletal alterations and nuclear architectural changes during mammalian fertilization. Curr. Top. Dev. Biol. 23, 23–54. Schlecht, U., and Primig, M. (2003). Mining meiosis and gametogenesis with DNA microarrays. Reproduction 125, 447–456. Schmitt, A., and Nebreda, A. R. (2002). Inhibition of Xenopus oocyte meiotic maturation by catalytically inactive protein kinase A. Proc. Natl. Acad. Sci. USA 99, 4361–4366. Schorderet-Slatkine, S., Schorderet, M., and Baulieu, E. E. (1982). Cyclic AMP-mediated control of meiosis: effects of progesterone, cholera toxin, and membrane-active drugs in Xenopus laevis oocytes. Proc. Natl. Acad. Sci. USA 79, 850–854. Schroeder, T. E., and Otto, J. J. (1991). Snoods: a periodic network containing cytokeratin in the cortex of starfish oocytes. Dev. Biol. 144, 240–247. Schuetz, A. W. (1971). Induction of oocyte maturation in starfish by 1-methyladenosine. Role of the ovarian wall. Exp. Cell Res. 66, 5–10. Schultz, R. M., and Kopf, G. S. (1995). Molecular basis of mammalian egg activation. Curr. Top. Dev. Biol. 30, 21–62. Schultz, R. M., Montgomery, R. R., and Belanoff, J. R. (1983). Regulation of mouse oocyte meiotic maturation: implication of a decrease in oocyte cAMP and protein dephosphorylation in commitment to resume meiosis. Dev. Biol. 97, 264–273. Seydoux, G., and Schedl, T. (2001). The germline in C. elegans: origins, proliferation, and silencing. Int. Rev. Cytol. 203, 139–185. Shen, X., Mizuguchi, G., Hamiche, A., and Wu, C. (2000). A chromatin remodelling complex involved in transcription and DNA processing. Nature 406, 541–544. Sheng, Y., Tiberi, M., Booth, R. A., Ma, C., and Liu, X. J. (2001). Regulation of Xenopus oocyte meiosis arrest by G protein subunits. Curr. Biol. 11, 405–416. Shilling, F., Chiba, K., Hoshi, M., Kishimoto, T., and Jaffe, L. A. (1989). Pertussis toxin inhibits 1-methyladenine-induced maturation in starfish oocytes. Dev. Biol. 133, 605–608.
108
Voronina and Wessel
Shiraishi, K., Okada, A., Shirakawa, H., Nakanishi, S., Mikoshiba, K., and Miyazaki, S. (1995). Developmental changes in the distribution of the endoplasmic reticulum and inositol 1,4,5-trisphosphate receptors and the spatial pattern of Ca2 þ release during maturation of hamster oocytes. Dev. Biol. 170, 594–606. Simonds, W. F. (1999). G protein regulation of adenylate cyclase. Trends Pharmacol. Sci. 20, 66–73. Smiley, S. (1990). A review of echinoderm oogenesis. J. Electron Microsc. Tech. 16, 93–114. Smith, L. D., and Ecker, R. E. (1969). Role of the oocyte nucleus in physiological maturation in Rana pipiens. Dev. Biol. 19, 281–309. Smith, L. D., and Ecker, R. E. (1971). The interaction of steroids with Rana pipiens oocytes in the induction of maturation. Dev. Biol. 25, 232–247. Smith, L. D., and Richter, J. D. (1985). Synthesis, accumulation, and utilization of maternal macromolecules during oogenesis and oocyte maturation, in Biology of Fertilization, Vol. 1, edited by C. B. Metz, and A. Monroy. Orlando, FL: Academic Press, pp. 141–188. Starz-Gaiano, M., Cho, N. K., Forbes, A., and Lehmann, R. (2001). Spatially restricted activity of a Drosophila lipid phosphatase guides migrating germ cells. Development 128, 983–991. Stein, J. A., Broihier, H. T., Moore, L. A., and Lehmann, R. (2002). Slow as molasses is required for polarized membrane growth and germ cell migration in Drosophila. Development 129, 3925–3934. Stein, P., Svoboda, P., and Schultz, R. M. (2003). Transgenic RNAi in mouse oocytes: a simple and fast approach to study gene function. Dev. Biol. 256, 188–194. Stevens, I., Derua, R., Rondelez, E., Waelkens, E., Merlevede, W., and Goris, J. (1999). Identification of cyk, a cyclin B2 kinase, as a novel calcium/calmodulin-dependent protein kinase II and its role during Xenopus laevis oocyte maturation. Exp. Cell Res. 252, 303–318. Stith, B. J., and Maller, J. L. (1987). Induction of meiotic maturation in Xenopus oocytes by 12O-tetradecanoylphorbol 13-acetate. Exp. Cell Res. 169, 514–523. Stricker, S. A., and Schatten, G. (1989). Nuclear envelope disassembly and nuclear lamina depolymerization during germinal vesicle breakdown in starfish. Dev. Biol. 135, 87–98. Su, Y. Q., and Eppig, J. J. (2002). Evidence that multifunctional calcium/calmodulin-dependent protein kinase II (CaM KII) participates in the meiotic maturation of mouse oocytes. Mol. Reprod. Dev. 61, 560–569. Swanson, W. J., and Vacquier, V. D. (2002). The rapid evolution of reproductive proteins. Nat. Rev. Genet. 3, 137–144. Tadenuma, H., Chiba, K., Takahashi, K., Hoshi, M., and Katada, T. (1991). Purification and characterization of a GTP-binding protein serving as pertussis toxin substrate in starfish oocytes. Arch. Biochem. Biophys. 290, 411–417. Taieb, R., Thibier, C., and Jessus, C. (1997). On cyclins, oocytes, and eggs. Mol. Reprod. Dev. 48, 397–411. Takamura, K., Fujimura, M., and Yamaguchi, Y. (2002). Primordial germ cells originate from the endodermal strand cells in the ascidian Ciona intestinalis. Dev. Genes Evol. 212, 11–18. Takizawa, C. G., and Morgan, D. O. (2000). Control of mitosis by changes in the subcellular location of cyclin-B1- Cdk1 and Cdc25C. Curr. Opin. Cell Biol. 12, 658–665. Tamura, M., and Nemoto, S. (2001). Reproductive maternal centrosomes are cast off into polar bodies during maturation division in starfish oocytes. Exp. Cell Res. 269, 130–139. Terasaki, M. (2000). Dynamics of the endoplasmic reticulum and Golgi apparatus during early sea urchin development. Mol. Biol. Cell 11, 897–914. Terasaki, M., Runft, L. L., and Hand, A. R. (2001). Changes in organization of the endoplasmic reticulum during Xenopus oocyte maturation and activation. Mol. Biol. Cell 12, 1103–1116.
3. The Regulation of Oocyte Maturation
109
Thomas, P., Zhu, Y., and Pace, M. (2002). Progestin membrane receptors involved in the meiotic maturation of teleost oocytes: a review with some new findings. Steroids 67, 511–517. Tian, J., Kim, S., Heilig, E., and Ruderman, J. V. (2000). Identification of XPR-1, a progesterone receptor required for Xenopus oocyte activation. Proc. Natl. Acad. Sci. USA 97, 14358–14363. Tombes, R. M., Simerly, C., Borisy, G. G., and Schatten, G. (1992). Meiosis, egg activation, and nuclear envelope breakdown are differentially reliant on Ca2 þ , whereas germinal vesicle breakdown is Ca2 þ independent in the mouse oocyte. J. Cell Biol. 117, 799–811. Toole, B. P., and Schuetz, A. W. (1974). Evidence for metabolism of 1-methyladenine during induction of starfish oocyte maturation in vitro. Dev. Biol. 39, 177–182. Tosti, E., Boni, R., and Cuomo, A. (2002). Fertilization and activation currents in bovine oocytes. Reproduction 124, 835–846. Tsafriri, A., Cao, X., Vaknin, K. M., and Popliker, M. (2002). Is meiosis activating sterol (MAS) an obligatory mediator of meiotic resumption in mammals. Mol. Cell Endocrinol. 187, 197–204. Tso, J., Thibier, C., Mulner, O., and Ozon, R. (1982). Microinjected progesterone reinitiates meiotic maturation of Xenopus laevis oocytes. Proc. Natl. Acad. Sci. USA 79, 5552–5556. Uetake, Y., Kato, K. H., Washitani-Nemoto, S., and Nemoto Si, S. (2002). Nonequivalence of maternal centrosomes/centrioles in starfish oocytes: selective casting-off of reproductive centrioles into polar bodies. Dev. Biol. 247, 149–164. Varnum, S. M., and Wormington, W. M. (1990). Deadenylation of maternal mRNAs during Xenopus oocyte maturation does not require specific cis-sequences: a default mechanism for translational control. Genes Dev. 4, 2278–2286. Venezky, D. L., Angerer, L. M., and Angerer, R. C. (1981). Accumulation of histone repeat transcripts in the sea urchin egg pronucleus. Cell 24, 385–391. Voronina, E., Marzluff, W. F., and Wessel, G. M. (2003). Cyclin B synthesis is required for sea urchin oocyte maturation. Dev. Biol. 256, 258–275. Voronina, E., and Wessel, G. M. (2001). Apoptosis in sea urchin oocytes, eggs, and early embryos. Mol. Reprod. Dev. 60, 553–561. Waldmann, R., Hanson, P. I., and Schulman, H. (1990). Multifunctional Ca2 þ /calmodulindependent protein kinase made Ca2 þ independent for functional studies. Biochemistry 29, 1679–1684. Wang, J., and Liu, X. J. (2003). A G protein-coupled receptor kinase induces Xenopus oocyte maturation. J. Biol. Chem. 278, 15809–15814. Wassarman, P. M. (1983). Oogenesis: synthetic events in the developing mammalian eggs, in Mechanism and control of animal fertilization, edited by J. F. Hartmann. New York, NY: Academic Press, pp. 1–54. Wasserman, W. J., Penna, M. J., and Houle, J. G. (1986). The regulation of Xenopus laevis oocyte maturation, in: Gametogenesis and the early embryo, edited by J. G. Gall. New York: Alan R. Liss, 111–130. Webb, R. J., Marshall, F., Swann, K., and Carroll, J. (2002). Follicle-stimulating hormone induces a gap junction-dependent dynamic change in [cAMP] and protein kinase A in mammalian oocytes. Dev. Biol. 246, 441–454. Wessel, G. M., Berg, L., Adelson, D. L., Cannon, G., and McClay, D. R. (1998). A molecular analysis of hyalin – a substrate for cell adhesion in the hyaline layer of the sea urchin embryo. Dev. Biol. 193, 115–126. Wessel, G. M., Brooks, J. M., Green, E., Haley, S., Voronina, E., Wong, J., Zaydfudim, V., and Conner, S. (2001). The biology of cortical granules. Int. Rev. Cytol. 209, 117–206. Wessel, G. M., Conner, S., Laidlaw, M., Harrison, J., and LaFleur, G. J., Jr. (2000a). SFE1, a constituent of the fertilization envelope in the sea urchin is made by oocytes and contains low-density lipoprotein-receptor-like repeats. Biol. Reprod. 63, 1706–1712.
110
Voronina and Wessel
Wessel, G. M., Conner, S. D., and Berg, L. (2002). Cortical granule translocation is microfilament mediated and linked to meiotic maturation in the sea urchin oocyte. Development 129, 4315–4325. Wessel, G. M., Zaydfudim, V., Hsu, Y. J., Laidlaw, M., and Brooks, J. M. (2000b). Direct molecular interaction of a conserved yolk granule protein in sea urchins. Dev. Growth Differ. 42, 507–517. Whitaker, M., and Larman, M. G. (2001). Calcium and mitosis. Semin. Cell Dev. Biol. 12, 53–58. Witchel, H. J., and Steinhardt, R. A. (1990). 1-Methyladenine can consistently induce a furadetectable transient calcium increase which is neither necessary nor sufficient for maturation in oocytes of the starfish Asterina miniata. Dev. Biol. 141, 393–398. Wright, C. S., Becker, D. L., Lin, J. S., Warner, A. E., and Hardy, K. (2001). Stage-specific and differential expression of gap junctions in the mouse ovary: connexin-specific roles in follicular regulation. Reproduction 121, 77–88. Wylie, C. (1999). Germ cells. Cell 96, 165–174. Xu, Z., Williams, C. J., Kopf, G. S., and Schultz, R. M. (2003). Maturation-associated increase in IP(3) receptor type 1: role in conferring increased IP(3) sensitivity and Ca(2 þ ) oscillatory behavior in mouse eggs. Dev. Biol. 254, 163–171. Yamashita, M. (1998). Molecular mechanisms of meiotic maturation and arrest in fish and amphibian oocytes. Semin. Cell Dev. Biol. 9, 569–579. Yoshikuni, M., Ishikawa, K., Isobe, M., Goto, T., and Nagahama, Y. (1988). Characterization of 1-methyladenine binding in starfish oocyte cortices. Proc. Natl. Acad. Sci. USA 85, 1874–1877. Yuce, O., and Sadler, K. C. (2001). Postmeiotic unfertilized starfish eggs die by apoptosis. Dev. Biol. 237, 29–44. Zeller, R., Haramis, A. G., Zuniga, A., McGuigan, C., Dono, R., Davidson, G., Chabanis, S., and Gibson, T. (1999). Formin defines a large family of morphoregulatory genes and functions in establishment of the polarising region. Cell Tissue Res. 296, 85–93. Zhao, K., Wang, W., Rando, O. J., Xue, Y., Swiderek, K., Kuo, A., and Crabtree, G. R. (1998). Rapid and phosphoinositol-dependent binding of the SWI/SNF-like BAF complex to chromatin after T lymphocyte receptor signaling. Cell 95, 625–636. Zhu, Y., Bond, J., and Thomas, P. (2003a). Identification, classification, and partial characterization of genes in humans and other vertebrates homologous to a fish membrane progestin receptor. Proc. Natl. Acad. Sci. USA 100, 2237–2242. Zhu, Y., Rice, C. D., Pang, Y., Pace, M., and Thomas, P. (2003b). Cloning, expression, and characterization of a membrane progestin receptor and evidence it is an intermediary in meiotic maturation of fish oocytes. Proc. Natl. Acad. Sci. USA 100, 2231–2236.
4 Stem Cells: A Promising Source of Pancreatic Islets for Transplantation in Type 1 Diabetes Cale N. Street,1 Ray V. Rajotte,1,2,4 and Gregory S. Korbutt1–3,* 1 Surgical-Medical Research Institute, Rm. 1074 Dentistry/Pharmacy Bldg., University of Alberta Edmonton, AB, Canada T6G 2N8 2
Department of Surgery, Rm. 1074 Dentistry/Pharmacy Bldg., University of Alberta Edmonton, AB, Canada T6G 2N8 3
Stem Cell Network of Canada, Rm. 1074 Dentistry/Pharmacy Bldg., University of Alberta Edmonton, AB, Canada T6G 2N8 4 Department of Medicine, Surgical Medical Research Institute, Rm. 1074 Dentistry/Pharmacy Bldg., University of Alberta Edmonton, AB, Canada T6G 2N8
I. II. III. IV. V. VI. VII.
Introduction Islet Transplantation: Success with Limitations Alternative Sources of Transplantable Islets Differentiation of Embryonic Stem Cells into Insulin-producing Cells Identification of the Elusive Pancreatic Stem Cell Existence of Nonpancreatic Islet Progenitor Cells Summary References
Diabetes is a disease that affects millions and causes a major burden on the health care system. Type 1 diabetes has traditionally been managed with exogenous insulin therapy, however factors such as cost, lifestyle restriction, and life threatening complications necessitate the development of a more efficient treatment alternative. Pancreas transplantation, and more recently transplant of purified pancreatic islets, has offered the potential for independence from insulin injections. Islet transplantation is gaining acceptance as it has been shown to be effective for certain patients with type 1 diabetes. One obstacle, however, is the fact that there is an inadequate supply of cadaveric human islets to implement this procedure on a widespread clinical basis. A promising source of transplantable islets in the future will come through the use of adult or embryonic stem cells. This chapter presents an overview of the advancements made in the development of a stem cell based application to islet transplantation. Advantages and limitations are discussed regarding the use of embryonic stem cells, adult pancreatic stem/progenitor
*To whom correspondence should be addressed. Tel.: (780) 492-4657; Fax: (780) 492-1627; E-mail: [email protected]. Current Topics in Developmental Biology, Vol. 58 Copyright ß 2003, Elsevier, Inc. All rights reserved. 0070-2153/03 $35.00
111
112
Street et al.
cells, and the use of nonpancreatic tissues based on current experimental models in the literature. It is concluded that stem cells offer the greatest potential for the development of an abundant source of pancreatic islets, although specific obstacles must be overcome before this can become a reality.
I. Introduction Type 1, insulin dependent diabetes is a complex and, as yet, little understood disease that affects millions worldwide. The disorder, which is characterized by juvenile onset of severe insulin deficiency due to loss of pancreatic islet cells, is believed to be caused by a combination of both genetic and environmental factors (Chowdhury et al., 1999; Kraine and Tisch, 1999) and persists throughout the life of an afflicted individual. Long term complications associated with type 1 diabetes are severe and result from fluctuating glycemia levels (Skyler, 1996). These consist of problems including hypertension (de La Sierra and Ruilope, 2000), nephropathies (Chiarelli et al., 1999; Ibrahim and Vora, 1999), neuropathies (Donaghue and Silink, 1999), and retinopathies (Lovestam-Adrian et al., 1999) and often lead to early mortality in these patients (Sochett and Daneman, 1999). The pathophysiology of type 1 diabetes has been attributed to an autoimmune disorder whereby the insulin producing -cells of the pancreas are selectively destroyed, likely due to the perceived immunogenicity of selfantigens present on the surface of these cells (Roep, 2003). Treatment of the disease thus stems from the need to replace circulating insulin that has been lost as well as prevent the development of potentially lethal ketoacidosis (Faich et al., 1983). The discovery and purification of insulin in 1921 by Banting and Best has provided a treatment that has remained the clinical standard, and daily injections of exogenous insulin have allowed patients with type 1 diabetes to live long and relatively normal lives. Results from the Diabetes Control and Complications Trial (1993) demonstrated that intensive insulin therapy and strict blood glucose monitoring can control glycemic fluctuations and limit diabetes related complications. However, the increased risk of severe hypoglycemia as well as problems with patient compliance to intensive injection regimens means that even the best case scenario of exogenous insulin therapy does not afford the glycemic control provided by a normally functioning pancreas. In an attempt to better control glycemia levels, transplantation of the whole pancreas has been performed with successful outcome. These transplants can provide stable and continuous normoglycemia (Shapira et al., 1999), and combined kidney-pancreas transplant can be effective in diabetic patients manifesting end-stage renal disease. However, the
4. Stem Cells as a Source of Pancreatic Islets
113
morbidity and mortality rates owing to the extreme invasiveness of the procedure, the need for lifelong immunosuppression, and the complicated nature of the surgery limit the usefulness of pancreas transplantation to only the most severe cases of type 1 diabetes. The advantages and disadvantages of both insulin injections and whole pancreas transplantation have led to the concept of transplantation of only the pancreatic endocrine cells within the islets of Langerhans.
II. Islet Transplantation: Success with Limitations Another therapeutic alternative for people with type 1 diabetes is the transplantation of isolated pancreatic islets to achieve insulin independence. The separation of endocrine islets from surrounding exocrine tissue was first attempted by microdissection in 1964 (Hellerstrom, 1964) and was later improved on by Lacy and Kostianovsky (1967) who used intraductal distention with collagenase followed by mechanical and enzymatic disruption to isolate rat islets. This general protocol of collagenase digestion is still in use for human islet isolation, with the added step of islet purification achieved by centrifugation on ficoll gradients (Scharp et al., 1973). Figure 1 depicts islets within intact human pancreatic biopsies (A) and in a purified clinical islet preparation. The procedure of islet transplantation, although successful initially in small animal models (Ballinger and Lacy, 1972; Reckard and Barker, 1973), proved to be difficult in humans. The main reason for this is the difficulty in obtaining islet-enriched preparations from the more fibrous human pancreas (Scharp et al., 1980). In results reported by the International Islet Transplant Registry (2001), over 267 islet transplants were performed between 1990 and 1998, however only 12.4% of these resulted in insulin independence for periods of 1 week or more.
Figure 1 Insulin immunostaining and hemotoxylin counterstain to show the presence of islets of Langerhans in human pancreatic biopsies (A) and purified clinical human islet grafts (B) (magnification: 200 ) (see color plate).
114
Street et al.
In 2000, Shapiro and coworkers in Edmonton, Canada reported a 100% success rate in achieving insulin independence through islet transplantation in 7 long term diabetic patients (Shapiro et al., 2000). Reasons for the dramatically increased success rate included a new immunosuppressive regimen excluding islet toxic glucocorticoids and the transplant of at least 10,000 islet equivalents per kilogram body weight. In recent follow up studies on the Edmonton Protocol (Ryan et al., 2001, 2002), a total of 12/15 (80%) of patients have maintained insulin independence beyond one year posttransplant with minimal side effects. However, although clinically successful, the widespread use of islet transplantation for treatment of type 1 diabetes has been limited by several major factors. First of all, the risks associated with the lifelong immunosuppression required to prevent graft rejection may outweigh those associated with daily insulin therapy in all but the most severe cases of diabetes. Secondly, the reliance on imperfect isolation techniques and limited availability of cadaveric donor organs has resulted in an extreme shortage of transplantable islets. For this reason, there is a need to find an alternative source of islets, either as a supplement to or replacement for the small number available from cadaveric donor organs.
III. Alternative Sources of Transplantable Islets An inadequate supply of islet tissue represents a major obstacle to the widespread implementation of islet transplantation. One possibility for an alternative source of islets involves a xenogeneic supply, whereby islets from another species could be used for transplantation to humans. In this regard, porcine islets are the most attractive alternative due to physiological similarities between pigs and humans as well as the fact that porcine insulin differs from human insulin by only one amino acid (Home et al., 1982). The extreme difficulty, however, of isolating viable adult porcine islets (Kirchhof et al., 1994; Ricordi et al., 1990) has hampered progress in this area. To circumvent this problem, methods have also been developed for the isolation and in vitro maintenance of neonatal porcine pancreatic cells (Korbutt et al., 1996) and evidence suggests that this immature tissue can develop into functional endocrine cells both in vivo (Korbutt et al., 1996) and in vitro (Korbutt et al., 1997). Even using this tissue, however, the problem of hyperacute rejection (Korbutt et al., 1996), continuing controversy over transfer of endogenous porcine viruses to the human genome (Blusch et al., 2002), and general public stigma over the use of animal organs for transplant have impeded the progress of this method as a clinical alternative. There also exists several cell based approaches to generate an abundant supply of islets or -cells. The use of gene therapy and advanced transfection techniques to bioengineer suitable primary cells or cell lines has been
4. Stem Cells as a Source of Pancreatic Islets
115
used to create insulin producing tissue (Cheung et al., 2000; Mashima et al., 1996; Shaw et al., 2002). These studies, however, are plagued with problems relating to clinical applicability. For example, techniques for the reliable delivery of genes are lacking and sustained expression has been elusive (Giannoukakis et al., 1999). Furthermore, it will prove exceedingly difficult to recreate all of the required biological machinery to allow for insulin synthesis, release, and glucose responsiveness in a cell not predisposed to these processes (Halban et al., 2001). For these reasons, as well as issues involving immune rejection and the tumorigenic risk of transplanting cell lines, these approaches remain far from providing an answer to the problem of islet availability. Among the most promising and most actively researched alternative islet sources is the use of embryonic or adult stem cells. The ability to isolate and expand progenitor cells that may subsequently be differentiated into pancreatic endocrine cells will represent a major advancement in the fields of islet transplantation and type 1 diabetes. Advantages to this approach include the possibility of propagating an unlimited number of cells that already possess the ability to become fully functioning endocrine tissue, as well as elimination of the aforementioned problems associated with xenotransplantation. Furthermore, the potential use of adult stem cells offers the advantage of an autologous model whereby a patient’s own cells can be used, thereby circumventing immune rejection. Similarly, embryonic stem cells (ES cells) in an undifferentiated state have been proposed to be undefined immunologically and programmable by the recipient’s own immune system as ‘‘self’’ tissue. The main drawback to the use of stem cells in islet transplantation at this time is simply that techniques to identify and subsequently differentiate stem cells to the islet endocrine phenotype are lacking. For this reason, the focus of this review is to provide a comprehensive summary of research on both ES cells and adult stem cells that is centered towards the generation of an unlimited supply of insulin producing cells. Specific issues relating to the maintenance and differentiation of insulin producing cells from ES cell lines will be addressed. In addition, progress and considerations in the identification of adult islet precursor cells will be discussed in detail with a focus on potential locations for this elusive population.
IV. Differentiation of Embryonic Stem Cells into Insulin-producing Cells The derivation of pluripotent stem cell lines from cells of the inner cell mass of a developing blastocyst has provided the potential for in vitro growth
116
Street et al.
of virtually any somatic tissue type. Indeed, it has been shown that a variety of cellular phenotypes, including heart, brain, muscle, endocrine, and hematopoietic cells can arise from ES cell cultures. Protocols for establishment and maintenance of ES cell lines were created over 20 years ago using murine cells (Evans and Kaufman, 1981; Martin, 1981), and many studies to this day still utilize mouse ES cells for the characterization of ES cell physiology and the development of differentiation protocols. In addition to significant advances with mouse ES cells, an important milestone was achieved in 1998 with the derivation of the first human ES cell lines by Thomson et al. (1998). Although differences have been defined between mouse and human ES cells with respect to surface antigen expression, morphology, and culture requirements (Hadjantonakis and Papaioannou, 2001), both are derived in similar ways and exhibit similar differentiation characteristics. Culture conditions for both typically involve the maintenance of single, undifferentiated cells (characterized by expression of markers such as Oct4) followed by the formation of cellular aggregates (embryoid bodies), before the induction of cellular differentiation (Donovan and Gearhart, 2001; Thomson et al., 1998). Initial protocols required the use of murine fibroblast feeder layers, complicating the issue of clinical feasibility in humans due to xenogeneic exposure, however newer methods require only conditioned media and defined growth factors (Xu et al., 2001), allowing the future possibility for transplantation to humans. Overall, the observation that mouse and human ES cells can be expanded indefinitely in an undifferentiated state and possess the inherent ability to develop into pancreatic islet endocrine cells suggests that these cells may be a potential source of transplantable tissue for type 1 diabetics. Furthermore, using techniques such as somatic cell nuclear transfer (Wakayama et al., 2001), ES cells could be created from a patient’s own cells eliminating the possibility of immunorejection after islet differentiation and transplant. One of the major limitations in ES cell research, however, is the inability to produce well controlled, directed differentiation into specific tissue types. When cultured in suspension, these cells form embryoid bodies that have been shown to contain partially differentiated cells of all three embryonic germ layers (mesoderm, endoderm, ectoderm) (Itskovitz-Elder et al., 2000). This heterogenous differentiation poses difficulties when attempting to create a large number of pancreatic islet cells. For this reason, specific culture conditions and growth factors have been utilized in an attempt to attain a more controlled, homogenous differentiation process. A study by Schuldiner et al. (2000) analyzed the effects of certain growth factors on ES cell differentiation. While it was found that specific growth factors such as EGF and TGF- could direct a percentage of cells towards a certain lineage, none were effective at producing homogenous cultures of a specific cellular phenotype. Importantly, it was also shown that the addition of nerve growth
4. Stem Cells as a Source of Pancreatic Islets
117
factor induced expression of some endodermal genes in these cultures, including the islet developmental transcription factor PDX-1. Although these results were promising, the problems with spontaneous differentiation were indicative of the difficulties still experienced today using ES cell lines. One study that indicated the possible development of pancreatic islets from mouse ES cells and spurred a worldwide outbreak of subsequent research was that of McKay and coworkers (2001). The protocol described in this study involved the production of an enriched cell population from murine ES embryoid bodies expressing the central nervous system precursor marker nestin through the use of serum free media, followed by expansion with basic fibroblast growth factor (bFGF). Differentiation involved withdrawal of bFGF and supplementation with B27 and nicotinamide to induce a pancreatic endocrine phenotype. Using these techniques, the authors reported the appearance of cells containing insulin as well as the other islet hormones glucagon, somatostatin, and pancreatic polypeptide. Expression of other markers of differentiated -cells, such as PDX-1, glucose transporter-2, and islet amyloid polypeptide was also seen in these islet-like structures. Functional assessment showed some degree of glucose stimulated insulin secretion, although these cells contained far less insulin than a native -cell and were not able to correct hyperglycemia when transplanted into diabetic mice. Interestingly, although a significant amount of research is now being based on this study, recent reports have questioned these findings (Rajagopal et al., 2003). In particular, as the media used during the differentiation process was supplemented with extremely high concentrations of insulin, it has been proposed that insulin immunoreactivity in these cells is related to uptake from the media and not from endogenous synthesis. This proposal is supported by: (i) lack of insulin mRNA by PCR, (ii) absence of the proinsulin cleavage product c-peptide, (iii) absence of secretory granules by electron microscopy, and (iv) lack of activity from an introduced insulin promoter-driven GFP gene in subsequent experiments using the same protocol. Although controversy now exists over the significance of results reported by the McKay group, several labs continue to use similar techniques with improved analysis methods in an attempt to create a homogenous islet population from ES cells. Since the McKay protocol for differentiation of nestin-positive mouse ES cells, a number of other groups have reported the derivation of insulincontaining cells from these cultures. A subsequent study, also utilizing murine ES cells, by Soria et al. (2000) reported the generation of a relatively homogenous population of insulin-secreting cells using a cell trapping technique. In this study, a transfected neomycin gene under control of the insulin promoter was used to positively select insulin expressing ES cells and subsequently develop a clonal population. Grown under appropriate
118
Street et al.
conditions, these cells were shown to contain significant amounts of insulin. Glucose induced insulin release, however, was abnormal and, although implantation led to transient correction of hyperglycemia in diabetic mice, this effect could not be maintained for extended periods. A study by Assady et al. (2001) was the first to report insulin production using human ES cells. Using the H9 human ES cell line, it was shown by immunostaining that a small percentage of differentiated embryoid body cells were positive for insulin and that insulin levels increased in the media compared to undifferentiated cultures. The expression of -cell specific markers such as glucokinase and glucose transporter-2 was also observed in differentiated cultures although it was not proven whether cells expressing these markers were the same cells that stained positive for insulin. In addition to these earlier studies, several other groups have recently used specific culture supplements or gene transfection strategies to induce ES cells to insulin production. Hori et al. (2002) used an inhibitor of the intracellular signaling molecule phosphoinositide 3-kinase to produce a population of insulincontaining cells from mouse ES cells. Insulin content in these cultures was reported to be about 30 times greater than that using the standard McKay protocol, although still only 10% of that seen in isolated islet preparations. Furthermore, aggregates were shown to be similar in morphology to islets, comprising predominately insulin-positive cells, some glucagon-positive cells, but no somatostatin or pancreatic polypeptide containing cells. These differentiated cells were also shown to improve the health of chemically induced diabetic mice, however their large size precluded the transplantation of a sufficient quantity to fully correct hyperglycemia. Another study by Wobus and coworkers (2003) used electroporation to transfect mouse ES cells with the islet/-cell developmental transcription factors PDX-1 and Pax4 in an attempt to induce pancreatic endocrine differentiation. They observed that Pax4 activation in nestin-positive embryoid bodies caused the induction of other transcription factors such as neurogenin 3, as well as an increase in insulin immunoreactivity (about 60% positive). Surprisingly, the authors do not address the possibility that increased insulin immunoreactivity in transfected cultures over wildtype cultures is due to increased cellular uptake from the media over controls as a result of membrane damage from electroporation. Furthermore, although the authors report rescue of experimentally induced diabetes in mice using these transfected cultures, the transplantation protocol used was questionable because animals were only followed for a relatively short time period of 14 days. Finally, a group from Belgium recently reported that the transition of mouse ES cells into insulin-containing cells using established protocols does not require the transcription factor HNF-6 (Houard et al., 2003). The authors used ES cells generated from HNF-6 knockout mice to reproduce the results obtained by McKay and others. The significance of this study is that normal
4. Stem Cells as a Source of Pancreatic Islets
119
-cell development is absolutely dependent on expression of HNF-6, showing that the mechanism of ES cell differentiation in these experiments differs from that of natural islet formation during development. Overall, the progress toward generating clinically transplantable islets from ES cells, although promising, has been hampered by fundamental biological hurdles involving culture conditions and control of differentiation. Regulatory and ethical issues, which are beyond the scope of this paper but are comprehensively reviewed in other manuscripts (McLaren, 2001; Resnick, 2002), also provide an obstacle for the clinical use of human ES cells. It should also be noted that the creation of a homogenous insulin producing, glucose-responsive cell population from ES cells may not be sufficient to provide a source for transplantation. The need to create a fully functional islet structure consisting of all four endocrine cell types should be recognized by researchers and this should be the ultimate goal of these experiments.
V. Identification of the Elusive Pancreatic Stem Cell Another possible solution to the problem of islet supply is the use of stem cells derived from adult tissues. The potential use of adult stem cells offers the advantage of an autologous model in which a patient’s own cells could be used, thereby eliminating the problem of graft rejection. The concept that stem cells exist in adult tissues was proposed years ago, and they have subsequently been identified in a diverse range of tissues, including liver, intestine, and skin. Recent advancements in adult stem cell research include the isolation and in vitro manipulation of neural (Reynolds and Weiss, 1992, 1996), hematopoietic (Hollands, 1987), and muscular (Zammit and Beauchamp, 2001) progenitor cells. The identification and exploitation of a pancreatic stem cell or precursor cell, using similar techniques, would represent a significant advancement for cell replacement therapy for type 1 diabetes. Islet neogenesis and transplantation of adult pancreatic stem cells is a concept that has attracted significant research attention. Evidence of continuous cell turnover in other organ systems (e.g., blood, intestine) throughout life suggests that the proportion of endocrine cells in the pancreas may also undergo dynamic changes in response to growth, development, and conditions such as pregnancy or obesity. A study by Finegood et al. (1995) proposed a mathematical model to estimate the dynamics of -cell turnover in the pancreas. According to this model, a balance is maintained in the pancreas between the processes of cell division, growth, and death. Furthermore, the lifespan of an average -cell was proposed to be variable depending on replication rate but most likely
120
Street et al.
from 1 to 3 months. This data would suggest that a significant amount of cell turnover in the pancreas occurs throughout life. Furthermore, studies involving experimentally induced pancreatic damage in animal models also support the idea of a stem cell population. Partial pancreatectomy in rats has been shown to result in islet regeneration and compensation for decreased circulating insulin levels (Li et al., 2001). Other modes of damage such as chemical destruction of islets using alloxan or streptozotocin (Tourrel et al., 2001; Yamamoto et al., 2000), cellophane wrapping (Swenne, 1983), ductal ligation (Wang et al., 1995), and exposure to transient hyperglycemia (Bonner-Weir et al., 1989; Lipsett and Finegood, 2002) have also been proven to result in pancreatic endocrine cell regeneration. Although evidence exists that -cells themselves may be induced to replicate under certain conditions (Duvillie et al., 2002; Maedler et al., 2002), the level of islet turnover apparent in the pancreas, as well as the fact that terminally differentiated cells do not normally undergo active proliferation, suggest that islet neogenesis from a pancreatic precursor cell plays a significant role in islet turnover. There is a substantial amount of evidence to support the hypothesis that islet neogenesis in the mature pancreas occurs via cells in or associated with the ductal epithelium. Ductal cells comprise about 5–10% of the normal pancreas and form the transport network for release of digestive enzymes into the gut. These cells are characteristically simple, undifferentiated, and hence lack specific identification markers such as those used for endocrine or exocrine cells. Despite this, ductal cells have been phenotypically defined in human and animal models through expression patterns of specific cytokeratin intermediate filament proteins. Ductal cells in both rat and human pancreas as well as pancreatic cultures have been shown to express a variety of cytokeratins; most prominently CK7 and CK19, in a reliable pattern (Bouwens et al., 1995; Real et al., 1993; Vila et al., 1994). More recently, the lectin cell surface protein Dolichus Biflorus Agglutinin (DBA) was also proposed as a marker for ductal epithelium and was used to isolate a population of these cells from a heterogenous pancreatic culture (Kobayashi et al., 2002). Evidence for elucidating the mechanisms of islet neogenesis in the adult pancreas may result from the study of embryonic and fetal pancreatic development. It is known that early pancreatic development from the endodermal bud progresses via ‘‘branching morphogenesis’’ of cytokeratin expressing ductal structures (Bouwens and DeBlay, 1996; Bouwens et al., 1997). Cells within these structures eventually lose expression of cytokeratins and develop into both the endocrine and exocrine compartments of the pancreas (Bouwens et al., 1997). In the adult pancreas, individual -cells as well as intact islets have been observed in close association with cytokeratin-positive ductal epithelium (Bertelli et al., 2001;
4. Stem Cells as a Source of Pancreatic Islets
121
Bouwens and Pipeleers, 1998). Transitional cells, expressing both insulin and CK19 have also been described in adult human pancreatic sections (Bouwens and Pipeleers, 1998). These data provide evidence that pancreatic stem cells may be ductal in nature or reside in close association with ductal epithelium. Experimental evidence, both in vivo and in vitro, exists to support the theory that islet neogenesis in the mature pancreas occurs from ductal cells. In 1991, it was shown that the cotransplantation of adult rat pancreatic epithelium and fetal-derived mesenchyme into the epididymal fat pad of rats resulted in the appearance of cells expressing both insulin and glucagon (Dudek et al., 1991). This study provided evidence for the widely held belief that factors released from surrounding mesenchymal tissues can induce pancreatic islet differentiation. As discussed earlier, several models of pancreatic damage have also shown the regeneration of islets in vivo from ductal cells. In particular, Wang et al. (1995) reported that ligation of the tail of the pancreas in rats resulted in a ductal to endocrine transition via proliferation and subsequent differentiation of cytokeratin-positive cells. Specific culture conditions and in vitro manipulation of pancreatic ductal cells has also been used to demonstrate that ductal tissue contains islet precursor cells. Bonner-Weir et al. (2000) cultured human ductal cells as a monolayer overlaid with MatrigelTM, and observed the growth of ‘‘islet buds’’ containing CK19 expressing as well as insulin-positive cells. Moreover, these cultures demonstrated increased insulin content and some degree of glucose-induced insulin secretion. In addition, a more recent study from Heimberg and coworkers (2002) showed that viral transduction of the early islet developmental transcription factor neurogenin 3 in human ductal cell cultures could initiate differentiation to -cell phenotype. Although these studies show convincing evidence that ductal cells contribute to islet neogenesis in the adult, the low proportions of differentiating cells suggest that either the methods are as yet inefficient, or that only a specific subpopulation of ductal cells are true islet progenitors. Several lines of evidence also suggest that a subpopulation of cells exist within the pancreatic ducts that may be endocrine precursors. For example, ductal cells have been shown to exhibit plasticity and are capable of expressing nonwildtype proteins under abnormal conditions. The expression of PDX-1, for example, has been documented in pancreatic ductal cells. PDX-1 is a homeobox transcription factor expressed in all pancreasdedicated cells of the endoderm during early development (Sander and German, 1997). Over the course of organogenesis, however, expression is gradually lost until only mature -cells of the adult pancreas express PDX-1 (Offield et al., 1996). It has been shown that PDX-1 is absolutely necessary for proper pancreas formation, as mice that are homozygous for the null mutation are born without a pancreas and die shortly thereafter
122
Street et al.
(Jonsson et al., 1994). PDX-1 has been demonstrated to be reexpressed in ductal cells of the mature pancreas under certain conditions. Leach and coworkers (1999) demonstrated the increased expression of PDX-1 in premalignant ductal epithelium in the pancreas of transgenic mice overexpressing transforming growth factor-alpha. Another study reported the increased expression of PDX-1 in rat ductal cells after partial pancreatectomy (Sharma et al., 1999). Furthermore, Heimberg et al. (2000) showed the presence of PDX-1 in a significant proportion of human pancreatic ducts, although phosphorylation patterns and complex formation was different than in mature -cells. Pancreatic ductal cells in culture have also been shown to reexpress PDX-1, as they do transiently during embryonic development. Nonendocrine pancreatic cultures derived from both rodent and human pancreas, consisting of primarily ductal cells, showed increased levels of PDX-1 expression (Gmyr et al., 2000; Rooman et al., 2000). Results from our lab also suggest that the proportion of ductal cells expressing PDX-1 after several days culture is donor-age dependent, with an increase seen in young donors (