molecular and cellular endocrinology of the testis
TRANSCRIPT
Ernst Schering Research Foundation Workshop
Supplement 1 Molecular and Cellular Endocrinology of the Testis
Ernst Schering Research Foundation Workshop
Editors: Gunter Stock Ursula-F. Habenicht
Vol. 5 Sex Steroids and the Cardiovascular System Editors: P. Ramwell, G. Rubanyi, E. Schillinger
Vol. 6 Transgenic Animals as Model Systems for Human Diseases Editors: E. F. Wagner, F. Theuring
Vol. 7 Basic Mechanisms Controlling Term and Preterm Birth Editors: K. Chwalisz, R. E. Garfield
Vol. 8 Health Care 2010 Editors: C. Bezold, K. Knabner
Vol. 9 Sex Steroids and Bone Editors: R. Ziegler, J. Pfeilschifter, M. Brautigam
Vol. 10 Non-Genotoxic Carcinogenesis Editors: A. Cockburn, L. Smith
Vol. 11 Cell Culture in Pharmaceutical Research Editors: N. E. Fusenig, H. Graf
Supplement 1 Molecular and Cellular Endocrinology of the Testis Editors: G. Verhoeven, U.-F. Habenicht
Ernst Schering Research Foundation Workshop, Supplement 1
Molecular and Cellular Endocrinology of the Testis
G. Verhoeven, U.-F. Habenicht Editors
With 55 Figures
Springer-Verlag Berlin Heidelberg GmbH
ISBN 978-3-662-22191-4 ISBN 978-3-662-22189-1 (eBook) DOI 10.1007/978-3-662-22189-1
CIP data applied for
This work is subject to copyright. All rights are reserved, whether the whole or part ofthe material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other way, and storage in data banks. Duplication of this publication or parts thereof is permitted only under the provisions oftheGerman Copyright Law of September 9, 1965, in its current version, and permission for use must always be obtained from Springer-Verlag Berlin Heidelberg. Violations are liable for prosecution under the German Copyright Law.
©Springer-Verlag Berlin Heidelberg 1994 Originally published by Springer-Verlag Berlin Heidelberg New York in 1994 Softcoverreprint of the hardcover Ist edition 1994
The use of generat descriptive names, registered names, trademarks, etc. in this publication does not impty, even in the absence of a specific Statement, that such names are exempt from the relevant proteelive taws and regutations and therefore free for generat use.
Product liability: The publishers cannot guarantee the accuracy of any information about dosage and application contained in this book. In every individual case the user must check such information by consulting the relevant titerature.
Typesetting: RTS, Wiesenbach
SPIN: lOt 33 978 21/3130-5 4 3 2 1 0- Printedon acid-free paper
Preface
The European Workshops on Molecular and Cellular Endocrinology of the Testis have become by now a well-established tradition. Thanks to their special format, the quality of the main lectures and miniposters, and the vivid discussions, they enjoy the ever-increasing interest and active participation of all European scientists working in the field. Moreover, since the very beginning they have attracted investigators from all over the world. The most recent "Testis Workshop" was held in De Panne, Belgium, from 27-3\ March, 1994. As always, the framework of the workshop was provided by a series of lectures delivered by a panel of internationally recognized authorities. These lectures are gathered in the present volume of the Ernst Schering Foundation Workshop series. Together with the Miniposter book they constitute an excellent written account of the Proceedings of the 8th European Testis Workshop.
The testis undoubtedly represents one of the most complex and intriguing tissues in the body. Both its endocrine function, the secretion of male sex hormones, and its exocrine role, the production of mature spermatozoa, continue to raise startling questions to clinicians, physiologists, endocrinologists, and scientists involved in fundamental research. Few organs maintain and support a differentiation process as complicated as spermatogenesis; few tissues continually display both mitotic and meiotic cell cyles in such a stringently controlled fashion or a comparable need for coordinated endocrine and local control. As a consequence, the testis has become a preferred model for basic research
VI Preface
on cell differentiation, cell cycle control, environmental and paracrine modulation of cell function, hormone action, signal transduction, and regulation of gene expression. Recent advances in cell isolation cell culture techniques, molecular biology, and genetic techniques such as development of transgenic animals, have found eager applications in research on molecular and cellular endocrinology of the testis. The present volume offers an elective, but representative, sample of the progress that has been made. A final chapter summarizes the impact that this type of research has had and will have with respect to the clinical diagnosis and treatment of male reproductive dysfunction.
The authors deserve our special thanks not only for delivering excellent state-of-the-art lectures, but also for providing nicely prepared manuscripts well before the meeting.
We are also glad to acknowledge the continuous and enthusiastic support and assistance of Focko Rommerts, coorganizer of the meeting, of the Permanent Scientific Committee of the Testis Workshops, and of the members of the Local Organinzing Committee as listed below.
C. Verhoeven
Permanent Scientific Committee of the European Workshops on Molecular and Cellular Endocrinology of the Testis Brian Cooke, London; Vidar Hansson, Oslo; lIpo Huhtaniemi, Turku; Eberhard Nieschlag, Munster; Martin Ritzen, Stockholm; Focko F.G. Rommerts, Rotterdam; Jose Saez, Lyon; Mario Stefanini, Rome; Guido Verhoeven, Leuven. Organizing Committee of the 8th European Workshop on Molecular and Cellular Endocrinology of the Testis Guido Verhoeven, Leuven; Walter Heyns. Leuven; Dirk Vanderschueren, Leuven; Ursula-F. Habenicht, Berlin; Focko F.G. Rommerts, Rotterdam.
Table of Contents
Hormonal Receptors in the Genital Tract A. Chauchereau. A. Mantel, K. Delabre, M. Misrahi, P. Lescop, M. Perrat-Applanat, H. Loosfelt, M.T. Vu Hai, N. Chinea, C. Meduri, J.-F. Savauret, and E. Milgram . ............... 1
2 Transplantation of Male and Female Germ and Somatic Cells R.C. Casden . ...................................... 15
3 Proliferation and Differentiation of Testicular Interstitial Cells: Aspects of Leydig Cell Development in the (Pre)Pubertal and Adult Testis K.J. Tcerds, MB. Veldhui::en-Tsoerkan, F.F.G. Rommerts, D.C. de Rooij, and JH. Dorrington ..... 37
4 Regulation of the Acute Production of Steroids in Steroidogenic Cells D.M. Stocco and B.J. Clark . .......................... 67
5 Extracellular Matrix Elements, Cell Adhesion Molecules, and Signal Transduction in the Control of Sertoli Cell Function M.D~ ........................................... ~
VIII Table of Contents
6 Interactions Between Androgens, Sertoli Cells and Genn Cells in the Control of Spennatogenesis R.M. Sharpe, C. McKinnell, T. McLaren, M.Millar, T.P. West, S. Maguire, I. Gaughan, V. Syed, B. legou, lB. Kerr, and P.TK Saunders ............................... 115
7 Cell Cycle Checkpoints in Male and Female Genn Cells D.F. Albertini . .................................... 143
8 Signal Transduction in Mammalian Spennatozoa G.S. Kopf, P. Kalab, P. Leclerc, X.P. Ning, D. Pan, and P. Visconti . ............................ 153
9 Diversity and Regulation of cAMP-Dependent Protein Kinases K. Tasken, B.S. Skdlhegg, K.A. Tasken, R. Solberg, F.O. Levy, T. Lea, T. lahnsen, and V. Hansson . .......... 185
10 The Nuclear Respone to cAMP During Spermatogenesis: The Key Role of Transcription Factor CREM P. Sassone-Corsi . ................................. 219
II Transgenic Animals and the Study of Gonadal Function M.M. Matzuk ..................................... 253
12 Clinical Relevance and Irrelevance of Molecular and Cellular Research on the Testis E. Nieschlag ...................................... 273
List of Contributors
Albertini, D.F. Department of Anatomy and Cellular Biology, Tufts University, 136 Harrison Avenue, Boston, MA 02111, USA
Chauchereau, A. Hormones et Reproduction, INSERM Unite 135, 78, rue du General Leclerc, 94275 Le Kremlin Bicetre, France
Clark, B . .T. Department of Biochemistry and Molecular Biology, Texas Tech University Health Sciences Center, Lubbock, Texas 79430, USA
Delabre. K. Hormones et Reproduction, INSERM Unite 135, 78, rue du General Leclerc, 94275 Le Kremlin Bicetre, France
de Raaij, D.C. Department of Cell Biology, Section of Cell Proliferation and Differentiation, Medical School, Utrecht University, P.O. Box 80.157, 3508 TD Utrecht, The Netherlands
Dorringtoll,l.H. Banting and Best Department of Medical Research, University of Toronto, C.H. Best Institute, 112 College Street, Toronto, Ontario, M5G IL6 Canada
Dvm. M. Department of Cell Biology, Georgetown University Medical Center, :1900 Reservoir Road, NW, Washington, DC 20007, USA
x List of Contributors
Gaughan,l. MRC Reproductive Biology Unit, Centre for Reproductive Biology, 37 Chalmers Street, Edinburgh EH3 9EW, Scotland, UK
Ghinea,N. Hormones et Reproduction, INSERM Unite 135, 78, rue du General Leclerc, 94275 Le Kremlin Bicetre, France
Gosden, R.G. Department of Physiology, University Medical College, Teviot Place, Edinburgh EH8 9AG, UK
Hansson, V. Institute of Medical Biochemistry, University of Oslo, P.O. Box 1112, 0317 Oslo, Norway
lahnsen, T. Institute of Medical Biochemistry, University of Oslo, P.O. Box 1112, 0317 Oslo, Norway
liigou, B. G.E.R.M., Universite de Rennes I, INSERM CJF 91-04, Campus de Baulieu, 35042 Rennes Cedex, France
Kalab, P. Division of Reproductive Biology, Department of Obstetrics and Gynecology, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104-6080, USA
Kerr, lB. Department of Anatomy, Monash University, Clayton, Victoria 3168, Australia
Kopf, G.S. Division of Reproductive Biology, Department of Obstetrics and Gynecology, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104-6080, USA
Lea, T. Institute of Immunology and Rheumatology, The National Hospital, 0027 Oslo, Norway
List of Contributors XI
Leclerc. P. Division of Reproductive Biology, Department of Obstetrics and Gynecology, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104-6080, USA
Lescop, P. Hormones et Reproduction, INSERM Unite 135, n, rue du General Leclerc, 94275 Le Kremlin Bicetre, France
Levy, F.O. Institute of Medical Biochemistry, University of Oslo, P.O. Box 1112, 0317 Oslo, Norway
LoosjClI, H. Hormones et Reproduction, INSERM Unite 135, 78, rue du General Leclerc, 94275 Le Kremlin Bicetre, France
Maguire, S. MRC Reproductive Biology Unit, Centre for Reproductive Biology, 37 Chalmers Street, Edinburgh EH3 9EW, Scotland, UK
Mantel, A.
Hormones et Reproduction, INSERM Unite 135, 78, rue du General Leclerc, 94275 Le Kremlin Bicetre, France
Mal:uk, M.M. Department of Pathology and Institute of Molecular Genetics, Baylor College of Medicine, Houston, Texas 77030, USA
McKinnell, C. MRC Reproductive Biology Unit, Centre for Reproductive Biology, 37 Chalmers Street, Edinburgh EH3 9EW, Scotland, UK
McLaren, T. MRC Reproductive Biology Unit, Centre for Reproductive Biology, 37 Chalmers Street, Edinburgh EH3 9EW, Scotland, UK
Meduri, C. Hormones et Reproduction, INSERM Unite 135, 78, rue du General Leclerc, 94275 Le Kremlin Bicetre, France
XII List of Contributors
Milgram, E. Honnones et Reproduction, INSERM Unite 135, 78, rue du General Leclerc, 94275 Le Kremlin Bicetre, France
Millar,M. MRC Reproductive Biology Unit, Centre for Reproductive Biology, 37 Chalmers Street, Edinburgh EH3 9EW, Scotland, UK
Misrahi,M. Honnones et Reproduction, INSERM Unite 135, 78, rue du General Leclerc, 94275 Le Kremlin Bicetre, France
Nieschlag, E. Institute of Reproductive Medicine of the University (WHO Collaborating Center for Research on Human Reproduction), Steinfurter Str. 107, 48149 MUnster, Gennany
Ning,X.P. Division of Reproductive Biology, Department of Obstetrics and Gynecology, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104-6080, USA
Pan,D. Division of Reproductive Biology, Department of Obstetrics and Gynecology, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104-6080, USA
Perrot-Applanat, M. Hormones et Reproduction, INSERM Unite 135, 78, rue du General Leclerc, 94275 Le Kremlin Bicetre, France
Rommerts, F.F.G. Department of Endocrinology and Reproduction, Medical School, Erasmus University Rotterdam, P.O.Box 1738, 3000 DR Rotterdam, The Netherlands
Sassone-Corsi, P. Laboratoire de Genetique Moleculaire des Eucaryotes, CNRS, Ul84 INSERM, Institut de Chimie Biologique, Faculte de Medecine, II, rue Humann, 67085 Strasbourg, France
List of Contributors
Saullders. P.T.K.
MRC Reproductive Biology Unit, Centre for Reproductive Biology, 37 Chalmers Street, Edinburgh EH3 9EW, Scotland, UK
Sal'lIuret. J.-F.
Hormones et Reproduction, INSERM Unite l35, 7'1'" rue du General Leclerc, 94275 Le Kremlin Bicetre, France
Sharp('. R.M. MRC Reproductive Biology Unit, Centre for Reproductive Biology, 37 Chalmers Street, Edinburgh EH3 9EW, Scotland, UK
Skdlhegg. B.S. Institute of Medical Biochemistry. University of Oslo, P.O. Box 1112, 0317 Oslo, Norway
Solherg. R. Institute of Medical Biochemistry, University of Oslo, P.O. Box 1112, 03 I 7 Oslo, Norway
Sto(,co. D.M.
Department of Biochemistry and Molecular Biology,
XIII
Texas Tech University Health Sciences Center, Lubbock, Texas 79430, USA
Syed. V. G.E.R.M .. Universite de Rennes I, INSERM ClF 91-04, Campus de Baulieu. 35042 Rennes Cedex, France
'{askhl. K. Institute of Medical Biochemistry, University of Oslo, P.O. Box 1112, 03 17 Oslo, Norway
Taskcn, K.A. Institute of Medical Biochemistry, University of Oslo, P.O. Box 1112, 03 17 Oslo, Norway
'{('ads. K..T.
Department of Cell Biology and Histology, Faculty of Veterinary Medicine. Utrecht University, P.O.Box 80.157,3508 TD Utrecht, The Netherlands
XIV List of Contributors
Veldhuizen-Tsoerkan, M.B. Department of Cell Biology and Histology, Faculty of Veterinary Medicine, Utrecht University, P.O.Box 80.157, 3508 TD Utrecht, The Netherlands
Visconti, P. Division of Reproductive Biology, Department of Obstetrics and Gynecology, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104-6080, USA
VuHai,M.T. Hormones et Reproduction, INSERM Unite 135, 78, rue du General Leclerc, 94275 Le Kremlin Bicetre, France
West, T.P. MRC Reproductive Biology Unit, Centre for Reproductive Biology, 37 Chalmers Street, Edinburgh EH3 9EW, Scotland, UK
1 Hormonal Receptors in the Genital Tract
A. Chauchereau, A. Mantel, K. Delabre, M. Misrahi, P. Lescop, M. Perrot-Applanat, H. Loosfelt, M.T. Vu Hai, N. Ghinea, G. Meduri, J.-F. Savouret, and E. Milgrom
I . I Sex Steroid Receptors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 I. 1.1 Molecular Mechanisms of Transactivation . . . . . . . . . . . . . . . . . . .. 2 1.1.2 Off-DNA Regulations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.1.3 Posttranscriptional Modifications. . . . . . . . . . . . . . . . . . . . . . . . . .. 4 1.104 Macromolecular Structure of Unactivated Complexes. . . . . . . . . .. 5 I. 1.5 Cellular Traffic of Nuclear Receptors. . . . . . . . . . . . . . . . . . . . . . .. 5 I. 1.6 Antihormones.......................................... 5 I .2 Gonadotropin Receptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 1.2.1 Molecular Structure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 1.2.2 Chromosomal Localization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 9 1.2.3 Immunocytochemical Studies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 10
Honnonal receptors are rare and fragile molecules. Their study was difficult and often yielded contradictory results until immunological and molecular probes became available. Sex steroid receptor cDNAs were cloned in 1986-1987. Monoclonal antibodies had previously been prepared and for this reason the understanding of their structure and function is already very advanced to date. In contrast, gonadotropin receptors were not cloned until 1989-1990 and well-characterized antibodies still remain a problem in many cases and thus our knowledge of their mere basic structural and functional features is still preliminary in many cases.
2 A. Chauchereau et al.
1.1 Sex Steroid Receptors
1.1.1 Molecular Mechanisms of Transactivation
Many studies are now focused on understanding the molecular interactions which lead to the regulation of gene transcription. The usual model involves binding of receptor dimers to palindromic structures (hormone responsive elements) upstream from genes. Perfect palindromes are scarce and very often the responsive elements are comprised of repeats of degenerated or incomplete palindromes. This interaction is followed by contacts with the transcriptional machinery. The exact nature of the transcription factors involved and the possible role of intermediary proteins are still a matter of debate. Several studies have shown that transcription factor lIB (TFIIB), which seems to play an important role in general transactivation mechanisms (Lin and Green 1991; Roberts et al. 1993), could also be the target ofthe members of the steroid-thyroid receptors superfamily (lng et al. 1992). Alternative models are also being described. In some cases, the structure of the chromatin plays an important role. For instance in the case of the hormone responsive elements of mouse mammary tumor virus (MMTV) it has been proposed that the role of the receptor is to bind to DNA and to knock out a nucleosome. This allows the transcription factor NF-l to gain access to its binding site on DNA (Cordingley et al. 1987; Pina et al. 1990). In this case, the role of the receptor is only indirect, another factor directly contacting the transcription machinery.
1.1.2 Off·DNA Regulations
In several receptor-mediated systems, off-DNA regulations have been described (J onat et al. 1990; Doucas et al. 1991; Savouret et al. 1991). In the case of the progesterone receptor (PR) gene, a single intragenic estrogen responsive element (ERE) is responsible for the estrogen inducibility as well as the progestin-mediated downregulation of its transcription, although PR does not bind this ERE (Savouret et al. 1991). Several estrogen and progesterone responsive elements are scattered throughout the upstream region of the promoter of the PR gene, but they do not seem to be involved in the physiological regulations, at least in
Hormonal Receptors in the Genital Tract 3
100
oF. 80
z 60 Q ~ 0 40 ::J CI Z
20
0 -12 -11 -10 -9 -8 -7
A ESTRADIOL (log M)
if< ~ 0 t-e""-Z Q • ~ 20
\ c:t ...J ::J 40 C) W 0:: 60 I Z • ~ 80 \. I , 0 CI •
100 I I I I I .,
-12 -11 -10 -9 -8 -7 B R 5020 (logM)
Fig. 1 A, B. Estrogen inducibility and progestin-mediated downregulation of the intragenic estrogen responsive element (ERE) in the progesterone receptor (PR) gene. A T 47-0 cells were transfected with a chloramphenicol acetyl transferase (CAT) expression vector containing either a restriction fragment (+23/+788) (-) or a 37-bp synthetic oligonucleotide (-) encompassing the ERE at (+698/+723) and an estrogen receptor (ER) expression vector. Cells were treated with increasing amounts of estradiol for 40 h. B Cos-7 cells were transfee ted with a CAT expression vector bearing the (+23/+788) restriction fragment and the ER and PR expression vectors. Cells were treated with 10 nM estradiol and challenged with increasing amounts of the synthetic progestin R 5020. The curve shows the decrease of estradiol induction elicited by R 5020, expressed in percentage of downregulation
4 A. Chauchereau 8t al.
breast cancer cells. These sites may be active in other tissues, depending on the state of differentiation. Figure 1 shows the dose-dependent kinetics of these regulations imparted on chloramphenicol acetyl transferase (CAT) expression vectors by a restriction fragment of the gene or a synthetic oligonucleotide encompassing the PR gene ERE.
The functional analysis of PR deletion mutants revealed that the steroid binding region and the DNA binding domain are required in the process of transcriptional downregulation. It seems likely that down-regulation is not a tissue-specific process but involves ubiquitous members of the transcriptional machinery since it could be reproduced by transfection of receptor expression vectors in nontarget cells.
These regulatory events often correspond to a cross talk between various transduction systems operating inside the cell. For instance, growth factors activate the AP-l system which eventually regulates the transcription of several genes. Different steroid hormone receptors inhibit the action of fos/jun proteins (AP-1) without binding to the DNA target. Direct interactions between the receptors and fos/jun dimers have been observed in some cases, but it is not known if they explain all such effects.
The inhibitory activity exists also in the case of hormone action: in several cases the AP-1 proteins have been shown to decrease the biological effects of steroids (Yang -Y en et al. 1990; J onat et al. 1990; Schiile et al. 1990).
1.1.3 Posttranscriptional Modifications
The role of the post translational modifications of receptors and especially the role of phosphorylation still remains unclear. Steroid hormone receptors are phosphoproteins which become hyperphosphorylated under the effect of the hormone (Logeat et al. 1985; Van Laar et al. 1991). There are several serines (and in some cases threonines) which have been shown to be modified (Denner et al. 1990; Chauchereau et al. 1991). It has been proposed that the receptor needs to be hyperphosphorylated in order to act on gene transcription (Weigel et al. 1992). Furthermore, enhancement of transcription of target genes by cAMP or dopamine analogs in presence of steroid receptors has been described. Receptor hyperphosphorylation has been proposed as one of
Hormonal Receptors in the Genital Tract 5
the mechanisms responsible for such ligand-independent receptor activation (Power et al. 1991). However, the role of receptor hyperphosphorylation remains debated (Chauchereau et a1. 1991).
l.l.4 Macromolecular Structure of Unactivated Complexes
Interactions of nonactivated receptors with various heat shock proteins (hsp90, hsp70) have also been the subject of great interest (Bresnick et al. 1989; Schowalter et al. 1991). Most remarkable has been the recent identification of one of these proteins, an immunophilin, which has a molecular weight of 59 kDa (Callebaut et a1. 1992).
l.l.5 Cellular Traffic of Nuclear Receptors
Progress has been made recently in the study of the cellular traffic of steroid hormone receptors. Site-directed mutagenesis showed that PR contains two nuclear localization signals (NLS): A stretch of five basic amino acids (638-642) is homologous to the NLS in the large T antigen of SV 40 and acts constitutively. The second NLS is less well defined and spans the second half of the DNA binding domain (593-640). As the two functions lie in close apposition, one may speculate that they act in vivo as a larger single unit (Guiochon-Mantel et al. 1989). PR shuttles towards the nucleus in a monomeric form in the absence of hormone, and ligand binding induces oligomerization. Energy deprivation experiments in cultured cells showed that the receptor continuously shuttles between the nucleus and the cytoplasm. This shuttle mechanism was also demonstrated in heterokaryon experiments, as shown in Fig. 2. Entry into the nucleus requires energy, while the cytoplasmic-bound diffusion does not (Guiochon-Mantel et al. 1991).
1.1.6 Antihormones
The mechanism of action of antisteroids has been extensively studied; most of them seem to provoke binding of receptors to hormone responsive elements but fail to elicit activation of transcription. Some hormone
6 A. Chauchereau et al.
Fig. 2 A-D. Transfer of receptor from mouse to human nuclei in heterokaryons. A, 8 Mouse L cells containing wild-type progesterone receptor (PR) were fused with 293 human cells devoid of receptor. Heterokaryons were observed. Cycloheximide was administered to prevent PR neosynthesis. PR was labeled by immunofluorescence 12 h after the fusion. Human (A, arrows) and mouse nuclei could be distinguished by fixation of Hoechst 33258 dye (8) and also to some extent by their size. After the fusion, PR appears in 293 human cell nuclei. C, D Control experiment in which cells were not fused. PR was labeled by immunofluorescence (C). Hoechst 33258 dye allows the two different cell types to be distinguished (D). The 293 human cells are devoid of PR (arrow). (FromGuiochon-ManteletaI.1991)
antagonists exert a partial agonistic activity. It has been suggested in the case of the antiestrogen tamoxifen and the antiprogestin RU 38486 that such effects are correlated with the constitutive transactivating function present in the N -terminal domain of the corresponding receptors (Berry et al. 1990; Meyer et al. 1990).
In the case of antiprogestins, initial studies with RU 38486 demonstrated that it acted on the final steps of hormonal stimulation of gene expression: RU 38486 allowed the PR to bind DNA at its cognate binding site but unproductively. A new molecule, ZK 98299, has recently been proposed as the prototype for a new class of antiprogestins
Hormonal Receptors in the Genital Tract 7
acting through the disruption of PR binding to DNA. This conclusion was based on in vitro experiments. We have devised methods to study in vivo the effects of these compounds on PRo We found that both types of antiprogestins had similar abilities in terms of receptor activation, dimerization, and DNA binding. This latter characteristic was analyzed through the ability of antagonist-bound PR to inhibit the transcriptional stimulation by the constitutive PR mutant (~ 663-930) of a reporter gene construct bearing the PR binding sites of the MMTV -CAT in cotransfection experiments. In all the types of experiments mentioned above, we have always observed a ten-fold difference between the concentrations of RU 38486 and ZK 98299 necessary to obtain the half-maximal effect on the various receptor functions (RU 38486 acting at the lower concentration). It is possible that the previously reported differences in the action of RU 38486 and ZK 98299 may be explained by the lower affinity of ZK 98299 for PR (Delabre et al. 1993).
1.2 Gonadotropin Receptors
1.2.1 Molecular Structure
The cDNA for the luteinizing hormone (LH) receptor was first to be cloned (Loosfelt et al. 1989; McFarland et al. 1989), followed by the other members of the family [thyroid stimulating hormone, TSH, and follicle-stimulating hormone, FSH (Libert et al. 1989; Misrahi et al. 1990; Sprengel et al. 1990; Minegish et al. 1991) I. All these receptors display the characteristic pattern of seven transmembrane spans specific of G-protein coupled receptors. They also have a large extracellular domain, the binding site of the hormone, which consists of repeats of a leucine-rich motif similar to those found in proteins which interact with amphipathic surfaces. Figure 3 shows the comparison of the domain organization of the three human receptors.
Besides displaying the full-length receptor, cloning also revealed the presence of several splice variants (Loosfelt et al. 1989). In the case or the pig LH receptor they are devoid of the transmembrane region and consist either only of the extracellular domain or of both the extracellular and the intracellular domains. These variants have been shown to be secreted from the cells and to bind the hormone (Vu Hai et al. 1992).
8
hTSHR
hLHR/ hCGR
hFSHR
El
108
105
103
A. Chauchereau et al.
E2 E3 E4E5 11 12
289 385403416 662708 764
32%
282 33 1 350 363 629656 695
Fig. 3. Comparison of the structure of human luteinizing honnone/chorionic gonadotrophic honnone (LH/CG; Minegish et al. 1990) human follicle-stimulating hormone (FSH; Minegish et al. 1991) and human thyroid stimulating hormone (TSH; Misrahi et al. 1990) receptors. Receptors are divided into regions according to the extent of homology (marked by the shaded areas on the figure, representing the FSH and TSH receptors). Dashed regions represent the putative signal peptide and the seven membrane spans. E1-5 are the putative extracellular and Il-2 intracellular domains. Amino acid numbering is shown above and below the figure. hTSHR, human thyroid stimulating honnone receptor; hLHR, human luteinizing hormone receptor; hCGR, human chorionic gonadotrophic hormone receptor; hFSHR, human follicle-stimulating honnone receptor. (From Misrahi et al. 1993)
Their presence in the circulation and the eventuality of a biological and/or physiopathological role still remain unresolved questions.
Immunoblotting showed the presence in relatively large amounts of another protein species: a high-mannose precursor of the receptor which tends to accumulate inside the cell (Vu Hai et al. 1992). Such an accumulation suggests the possibility of regulations involving recruitment of this inactive pool of receptor under specific physiological circumstances.
Hormonal Receptors in the Genital Tract 9
1.2.2 Chromosomal Localization
The human genes for both LH and FSH receptor are located on chromosome 2p21 (Rousseau-Merck et al. I 990a, 1992). Their structure is similar, with a unique ex on encoding a short part of the extracellular domain and the whole of the transmembrane and of the intracellular domains, whereas the rest of the extracellular domain is split into nine (FSH receptor gene) or ten (LH receptor gene) exons (Tsai-Morris et al. 1991). In contrast, the TSH receptor gene has been assigned to chromosome 14q31 (Rousseau-Merck et a1. 1990b). It is composed of ten exons, which is similar to the FSH receptor gene (Gross et a1. 1991). Considering their high homology, LH, FSH, and TSH receptors genes may have evolved by duplication of a common ancestor: LH and FSH receptor genes remained in close vicinity while the TSH receptor gene was scattered in the genome during evolution.
1.2.3 Immunocytochemical Studies
Monoclonal antibodies against LH receptors have been used to study the intracellular traffic of LH receptors by ultrastructural immunocytochemistry (Ghinea et al. 1992). The receptors were shown to be internalized into lysozomes via coated pits and vesicles. There was a constitutive level of internalization which was increased about II-fold by administration of hormone. Recycling was very limited, as most of the internalized receptor underwent degradation. This mechanism explains the important downregulation induced by ligand binding. Double labeling with anti receptor antibody and hCG bound to gold particles of different sizes showed that hormone and receptor follow the same intracellular route.
Immunocytochemical light microscope studies showed in pig ovaries some new aspects of a functional zonation of LH receptor distribution in follicles and cyclic corpora lutea (Meduri et a1. 1992). In preovulatory follicles, the receptor was detected in the theca interna. The region ofthe theca interna closer to the lamina basalis (approximatively one third of the cells) appeared devoid of receptors, whereas the most external region was strongly labeled. Granulosa cells were also labeled at a later stage in large antral follicles. In cyclic corpora lutea, only the
10 A. Chauchereau et al.
most external cells (probably of thecal origin) were stained while the most abundant cells (of granulosa origin) were not labeled. These data indicate that both the theca interna and the corpus luteum are heterogenous structures composed of two zones, of which only one is directly sensitive to LH. The physiological significance of this zonation and the possibility of an indirect regulation of the receptor-devoid zone through factors released by the LH-stimulated external cell layer remain unresolved questions. Few immunochemical or immunocytochemical studies have been performed on human gonadotropin receptors, probably due to the lack of sufficiently effective antibodies.
References
Berry M, Metzger D, Chambon P (1990) Role of the two activating domains of the estrogen receptor in the cell-type and promoter-context dependent agonistic activity of the anti-oestrogen 4-hydroxytamoxifen. EMBO J 9:2811-2818
Bresnick EH, Dalman FC, Sanchez ER, Pratt WB (1989) Evidence that the 90-kDa heat shock protein is necessary for the steroid binding conformation of the L cell glucocorticoid receptor. J BioI Chern 264:4992-4997
Callebaut J, Renoir JM, Lebeau MC, Massol N, Bumy A, Baulieu EE, Momon JP (1992) An immunophilin that binds Mr 90,000 heat shock protein: main structural features of a mammalian p59 protein. Proc Nat! Acad Sci USA 89:6270-6274
Chauchereau A, Loosfelt H, Milgrom E (1991) Phosphorylation of transfected wild type and mutated progesterone receptors. J BioI Chern 266: 18280-18286
Cording ley MG, Riegel AT, Hager GL (1987) Steroid-dependent interaction of transcription factors with the inducible promoter of mouse mammary tumor virus in vivo. Cell 48:261-270
Delabre K, Guiochon-Mantel A, Milgrom E (1993) In vivo evidence against the existence of antiprogestins disrupting receptor binding to DNA. Proc Nat! Acad Sci USA 90:4421-4425
Denner LA, Schrader WT, O'Malley BW, Weigel NL (1990) Hormonal regulation and identification of chicken progesterone receptor phosphorylation sites. J BioI Chern 265:16548-16555
Doucas V, Spyrou G, Yaniv M (1991) Unregulated expression of c-Jun or cFos proteins but not Jun D inhibits oestrogen receptor activity in human breast cancer derived cells. EMBO J 10:2237-2245
Hormonal Receptors in the Genital Tract 11
Ghinea N, Vu Hai MT, Groyer-Picard MT, Houllier A, Schoevaert D, Milgrom E (1992) Pathways of internalization of the hCG/LH receptor: immunoelectron microscopic studies in leydig cells and transfected L-cells. 1 Cell Bioi 118:1347-1358
Gross B, Misrahi M, Sar S, Milgrom E (1991) Composite structure of the human thyrotropin receptor gene. Biochem Biophys Res Commun 177:679-687
Guiochon-Mantel A, Loosfelt H, Lescop P, Sar S, Atger M, Perrot-AppIan at M, Milgrom E (1989) Mechanisms of nuclear localization of the progesterone receptor: evidence for interaction between monomers. Cell 57:1147-1154
Guiochon-Mantel A, Lescop P, Christin-Maitre S, Loosfelt H, Perrot-AppIan at M, Milgrom E (1991) Nucleocytoplasmic shuttling of the progesterone receptor. EMBO 1 10:3851-3859
Ing NH, Beekman JM, Tsai SY, Tsai Ml, O'Malley BW (1992) Members of the steroid hormone receptor superfamily interact with TFIIB (S300-II). 1 BiolChem267:17617-17623
lonat C, Rahmsdorf HJ, Park KK, Cato ACB, Gebel S, Ponta H, Herrlich P (1990) Antitumor promotion and anti inflammation: down-modulation of API (Fos/Jun) activity by glucocorticoid hormone. Cell 62: 1189-1204
Libert F, Parmentier M, Lefort A, Dinsart C, Van Sande 1, Maenhaut C, Simons Ml, Dumont IE, Vassart G (1989) Selective amplification and cloning of four new members of the G protein-coupled receptor family. Science 244:569-572
Lin Y -S, Green MR (1991) Mechanism of action of an acidic transcriptional activator in vitro. Cell 64:971-981
Logeat F, Le Cunff M, Pamphile R, Milgrom E (1985) The nuclear-bound form of the progesterone receptor is generated through a hormone dependent phosphorylation. Biochem Biophys Res Commun 131 :421-427
Loosfelt H, Misrahi M, Atger M, Salesse R, Vu Hai MT, Jolivet A, GuiochonMantel A, Sar S, lallal B, Garnier J, Milgrom E (1989) Cloning and scquencing of porcine LH/hCG receptor. Variants lacking transmembrane domain. Science 245:525-528
McFarland KC, Sprengel R, Phillips HS, Kohler M, Rosemblit N, Nikolics K, Segal off DL, Seeburg PH (1989) Lutropin-choriogonadotropin rcceptor: an unusual member of the G protcin-coupled receptor family. Sciencc 245:494-499
Meduri G, Vu Hai MT, lolivet A, Milgrom E (1992) New functional zonation in the ovary as shown by immunohistochemistry of LH receptor. Endocrinology 131 :366-373
12 A. Chauchereau et al.
Meyer ME, Pornon A, Ji J, Bocquel MT, Chambon P, Gronemeyer H (1990) Agonistic and antagonistic activities of RU486 on the functions of the human progesterone receptor. EMBO J 9:3923-3932
Minegish T, Nakamura K, Takakura Y, Miyamoto K, Hasegawa Y, Ibuki Y, Igarashi M (1990) Cloning and sequencing of human LH/hCG receptors cDNA. Biochem Biophys Res Commun 172: 1049-1054
Minegish T, Nakamura K, Takakura Y, Ibuki Y, Igarashi M (1991) Cloning and sequencing of human FSH receptor cDNA. Biochem Biophys Res Commun 175: 1125-1130
Misrahi M, Loosfelt H, Atger M, Guiochon-Mantel A, Milgrom E (1990) Cloning, sequencing and expression of human TSH receptor. Biochem Biophys Res Commun 166:394--403
Misrahi M, Vu Hai MT, Ghinea N, Loosfelt H, Meduri G, Atger M, Jolivet A, Gross B, Savouret JF, Dessen P, Milgrom E (1993) Molecular and cellular biology of gonadotropin receptors. In: Adashi EY, Leung PC (eds) The ovary. Raven, New York, pp 57-92
Pin a B, Brliggemeier U, Beato M (1990) Nucleosome positioning modulates accessibility of regulatory proteins to the mouse mammary tumor virus promoter. Cell 60:719-731
Power RF, Mani SK, Codina J, Conneely OM, O'Malley BW (1991) Dopaminergic and ligand-independent activation of steroid hormone receptors. Science 254: 1636-1639
Roberts SGE, Ha I, Maldonado E, Reinberg D, Green MR (1993) Interaction between an acidic activator and transcription factor TFIIB is required for transcriptional activation. Nature 363:741-744
Rousseau-Merck MF, Misrahi M, Atger M, Loosfelt H, Milgrom E, Berger R (1990a) Localization of the human luteinizing hormone choriogonadotropin receptor gene (LHCGR) to chromosome 2p21. Cytogenet Cell Genet 54:77-79
Rousseau-Merck MF, Misrahi M, Loosfelt H, Atger M, Milgrom E, Berger R (I 990b ) Assignment of the human thyroid stimulating hormone receptor (TSHR) gene to chromosome 14q31. Genomics 8:233-236
Rousseau-Merck MF, Atger M, Loosfelt H, Milgrom E, Berger R (1992) The chromosomal localization of the human follicle-stimulating hormone receptor gene (FSHR) on 2p21-p16 is similar to that of the luteinizing hormone receptor gene. Genomics 15:222-224
Savouret JF, Bailly A, Misrahi M, Rauch C, Redeuilh G, Chauchereau A, Milgrom E (1991) Characterization of the hormone responsive element involved in the regulation of the progesterone receptor gene. EMBO J 10: 1875-1883
Hormonal Receptors in the Genital Tract 13
Schowalter DB, Sullivan WP, Maihle NJ, Dobson ADW, Conneely OM, O'Malley BW, Toft DO (1991) Characterization of progesterone receptor binding to the 90- and 70-kDa heat shock proteins. J Bioi Chem 266:21165-21173
Schiile R, Rangararajan P, Kliever S, Ransone LJ, Bolado J, Yang N, Verma 1M, Evans RM (1990) Functional antagonism between oncoprotein c-Jun and the glucocorticoid receptor. Cell 62: 1217-1226
Sprengcl R, Braun T, Nikolics K, Segaloff DL, Seeburg PH (1990) The testicular receptor for follicle stimulating hormone: structure and functional cxpression of cloned cDNA. Mol Endocrinol 4:525-530
Tsai-Morris CH, Buczko E, Wang W, Xie XZ, Dufau ML (1991) Structural organization of the rat luteinizing hormone (LH) receptor gene. J Bioi Chem 266:11355-11359
Van Laar JH, Berrevoets CA, Trapman J, Zegers ND, Brinkmann AO (1991) Hormone-dependent androgen receptor phosphorylation is accompanied by receptor transformation in human lymph node carcinoma of the prostate cells. J Bioi Chern 266:3734-3738
Vu Hai MT, Misrahi M, Houllier A, Jolivet A, Milgrom E (1992) Variant forms of the pig lutropin/choriogonadotropin receptor. Biochemistry 31 :8377-83X3
Weigel NL. Carter TH, Schrader WT, O'Malley BW (1992) Chicken progesterone receptor is phosphorylated by a DNA-dependent protein kinase during in vitro transcription assays. Mol Endocrinol 6:8-14
Yang-Yen HF, Chambard JC, Sun YL, Smeai T, Schmidt TJ, Drouin J, Karin M (J 990) Transcriptional interference between c-Jun and the glucocorticoid receptor: mutual inhibition of DNA binding due to direct protein-protein interaction. Cell 62: J 205-12 J 5
2 Transplantation of Male and Female Germ and Somatic Cells
R.G. Gosden
2.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 15 2.1.1 Varieties of Gonadal Transplantation. . . . . . . . . . . . . . . . . . . . . . .. 15 2.1.2 Potential Applications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 17 2.2 A Brief History of Gonadal Transplantation. . . . . . . . . . . . . . . . . .. 19 2.3 Ovarian Transplantation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 21 2.3.1 Autografts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 21 2.3.2 Allografts.............................................. 23 2.3.3 Xenografts............................................. 24 2.4 Testicular Transplantation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 25 2.4.1 Whole Organ. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 25 2.4.2 Stem Cells. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 26 2.5 Survival of Allografts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 27 2.6 Prospects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 29 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 30
2.1 Introduction
2.1.1 Varieties of Gonadal Transplantation
Gonadal transplantation simply involves the transfer to or replacement of gonadal tissue, but there are many variations on this theme. In most circumstances the host is sterile, although grafts can function even when the host's own gonads are intact. The gender of the donor organ and of
16 R.G. Gosden
the host is normally the same, but not necessarily so. Grafts of gonadal tissue, like those of other organs, are often labelled according to their origin and genetic affinity with the host because the degree of antigenic similarity often predicts their survival. Isografts are derived from different, but genetically identical individuals such as monozygotic twins and inbred strains of animals. Autografts are genetically comparable to isografts, but specifically derived from the host itself. When the donor is genetically different, the graft is either called an allograft or a xenograft, depending on whether the donor and host are from the same or a different species, respectively. All these combinations of graft and host have been tested with gonadal tissues for various purposes at one time or another.
The site of attachment of grafts is not always the anatomically normal, or orthotopic, site. Foreign, or heterotopic, sites can be very successful because the endocrine functions resume as soon as the glands have re-established a vascular link. There are some circumstances, however, when a heterotopic site does not restore a normal secretory pattern, as may occur for instance when ovarian secretions drain into the hepatic portal circulation (Biskind et al. 1950) or when the guinea-pig ovary is remote from luteolytic influences of the neighbouring uterus (Bland and Donovan 1968). The gametogenic function of the gonads, on the other hand, obviously requires a connection or proximity to the reproductive tract in order to promote conception unassisted. Heterotopic gonadal grafts may seem perverse, but they have sometimes played important roles in experimental endocrinology, such as providing an immunologically privileged environment for allografts. The kidney is not so privileged and is suitable only for isografts and autografts; nevertheless, it carries the important advantages of being well vascularized and having a capsule for securing an implant (Felicio et al. 1983). Other sites are sometimes used because they provide better experimental access to the organ for repeated blood sampling, e.g. anastomosis to vessels in the neck (Goding et al. 1967).
Over the past century, hundreds of papers have been published describing the techniques and results of gonadal transplantation in a number of mammalian species, including humans. The majority of them have served experimental science well and the techniques continue to contribute to understanding reproductive endocrinology (Krohn 1977; Gosden 1992a; Gosden and Murray 1993). At the beginning, gonadal
Transplantation of Male and Female Germ and Somatic Cells 17
transplantation was conceived as a method for overcoming hypogonadism in either sex, and this review is an opportunity to revisit this theme and re-evaluate the practical potential.
2.1.2 Potential Applications
2.1.2.1 The Ovary The major application for gonadal transplantation in reproductive medicine is to overcome sterility or reduce the risks of precocious hypogonadism (Table 1). All women who live to their sixth decade will sooner or later undergo menopause, but the probability of having a full menstrual lifespan of about 36 years depends on the numbers and dynamics of primordial follicles. This store is formed before birth and progressively depleted thereafter until very few remain at menopausal age (Block 1952; Faddy et al. 1992). Premature loss of follicles resulting from surgical removal of ovarian tissue or from cytotoxic damage is expected to hasten menopause. More rarely, ovarian failure occurs congenitally because the follicle store is not formed, such gonadal dysgenesis being frequently caused by a chromosomal anomaly (e.g. 45,XO). In all these cases, hormone replacement therapy (HRT) is indicated, and in some cases assisted reproduction is desirable.
Pregnancies can sometimes be established after premature ovarian failure by ovum donation and hormonal priming to make a uterine environment receptive for embryo implantation and early pregnancy (Craft et al. 1987; Edwards et al. 1991; Sauer et al. 1993). Unfortunately, this treatment is costly and limited by the availability of donors. Ovarian transplantation therefore deserves serious consideration as an alternative with these advantages: (I) a single operation, (2) a possibility of conceiving naturally afterwards, (3) production of physiological hormone levels during pregnancy, and (4) re-establishing menstrual cycles, thus delaying the need for HRT. While there is a good case for giving a graft to patients with premature ovarian failure, I believe that an even stronger one can be made for those with gonadal dysgenesis and "streak" gonads. Nevertheless, ovarian allografts raise ethical questions and transplantation immunity is likely to be a problem.
Transsexual grafting is a potentially more controversial issue and unlikely to win much public approval. From a purely theoretical point of
18 R.G. Gosden
Table 1. Potential medical applications of gonadal cell transplants
I. Reversal of premature hypogonadism 2. Treatment of gonadal dysgenesis 3. Restoration of sex hormone production 4. Replacement of germ cells in patients transmitting severe genetic disease 5. Transsexual grafting 6. Fundamental investigations of gonadal cell biology and pathology
view, an ovarian graft would probably produce ovulatory-type cycles in a castrated man because the mechanism for generating an ovulatory surge of gonadotrophins is not sexually differentiated in primates. This conclusion has been borne out by experiments in which ovaries established hormonal cycles after transplantation to immunosuppressed and castrated male macaque monkeys (Norman and Spies 1986).
In view of anticipated difficulties with ovarian allografts, autografts are much more appealing. Their immediate application is to restore ovarian function after low temperature storage of tissue while cancer patients undergo chemotherapy and/or radiotherapy. Provided the tissues do not harbour neoplastic cells, grafts containing large numbers of primordial follicles could offer a practicable alternative to the frozen storage of embryos or secondary oocytes, especially since neither of these two strategies is appropriate for prepubertal children. Experience with other cell types suggests that follicular cells deposited in liquid nitrogen could be stored satisfactorily for many years.
2.1.2.2 The Testis In contrast to the early disappearance of oogonia in the ovary, spermatogonia are normally present throughout life, though they do not produce spermatozoal progeny until puberty. Spermatogenic cells are also vulnerable to damage from cytotoxic chemicals, ionizing radiation, infection, ischaemia and heat, although cellular repopulation may occur from stem cells. In the ovary the loss of gametes and sex hormones occur pari passu, but this need not be the case in the testis since the interstitial cells and the seminiferous tubules are separate (though intercommunicating) compartments. Since testosterone levels may be well maintained even in azoospermic men and because hormone replacement therapy is an ef-
Transplantation of Male and Female Germ and Somatic Cells 19
fective remedy for hypoandrogenism, there is a stronger case a priori for transferring spermatogonial stem cells than for Leydig cells.
The potential benefits of spermatogenic cell transfer have, to some extent, been overtaken by microinsemination of oocytes in vitro for oligospermic men, on the one hand, and artificial insemination by donor for azoospermia, on the other. Nevertheless, some young agonadal men would probably wish to have a transplant, if available, and those who are likely to become hypogonadal as a result of cancer treatment would undoubtedly wish to protect their organs, if possible.
The following sections summarize a large and dispersed literature on gonadal transplantation and draw attention to some recent developments. But it is with the early history of the subject that I begin because that helps to explain why gonadal transplantation has been neglected for many years.
2.2 A Brief History of Gonadal Transplantation
The effects of castration on the male phenotype and behaviour have been recognized for the more than two millenia since Aristotle, but there are no records of attempts to overcome the castration syndrome by transplantation until the late eighteenth century. In the words of the Scottish surgeon-anatomist John Hunter, who performed the first grafting operation: "I had formerly transplanted the testicles of a cock into the abdomen of a hen, and they had sometimes taken root there, but not frequently, and then had never come to perfection" (Palmer 1837). Some of his original specimens survive to this day in the Hunterian Museum of the Royal College of Surgeons in London.
The Gottingen biologist Berthold (1849) was evidently more successful because he avoided the allograft reaction by grafting testicles back into the abdominal cavity of the same capon. His study was a major milestone in the history of endocrinology because, in showing that the comb, plumage and courting behaviour were maintained, it was apparent that the gonadal influence was blood borne rather than nervous. The full implications of Berthold's work, like those of his contemporary, Gregor Mendel, were largely overlooked until the end of the nineteenth century when a number of clinicians and biologists attempted to trans-
20 R.G. Gosden
plant both ovaries and testes. Another 30 years were to pass before testosterone was isolated and synthesized.
Robert Morris, a leading New York surgeon, pioneered the clinical transplantation of ovaries before successful animal experiments had been performed. He reported that a sterile woman "received an ovarian graft in the fundus of her uterus from the ovary of another patient". And in another case, "a small piece of her own diseased ovary was transferred to the interior of the stump of one oviduct" (Morris 1895). The first patient menstruated 2 months later and the second became pregnant, although miscarried after only 3 months. Encouraged by this progress, he continued with the operations and some years later reported a live birth using these methods (Morris 1906). His success stimulated others to test the operation on their patients (see Woodruff 1960). During this period, ovarian autotransplantation was also being used to bypass occluded fallopian tubes, but this was rarely successful and was eventually superceded by tubal microsurgery and in vitro fertilization (Adams 1979; Biggers 1984). It is impossible to be sure whether Morris's remarkable claims were justified since his evidence no longer exists and the more mundane possibility of residual follicles in the "sterile" ovaries cannot be ruled out. Many unsubstantiated claims were being made for transplantation at the time and authors were ignorant of the allograft reaction.
The first attempts to transplant ovaries in laboratory mammals were made by the distinguished French scientist Paul Bert (1865) but were unsuccessful. It was some 30 years later that Knauer (1896) obtained convincing results by autografting rabbit ovaries to the broad ligament, showing that ovarian function and uterine weight were maintained. Shortly afterwards, Foa (1900) made the important discovery that immature ovaries function precociously after grafting into an adult environment. These experiments were successful, unlike many others at the time, because they used either autografts or donors and hosts that were closely related (inbred strains of rodents were not available). Marshall and Jolly (1908), working in my department in the first decade of this century, had already realized that ovarian allografts in rats and monkeys seldom survived: "Homoplastic transplantation of ovaries is very considerably easier to perform successfully than heteroplastic transplantation, (but the latter) is apparently easier to perform successfully when ... near relatives of each other". Had others heeded their warning,
Transplantation of Male and Female Germ and Somatic Cells 21
some enterprising doctors may not have gone down one of the most notorious blind alleys of clinical science, namely, rejuvenation therapy.
This episode originated from the misleading but influential speculation that the signs and symptoms of old age are secondary to sex honnone deficiency. Its author, Brown-Sequard (1889), reasoned that by replacing the deficiency of "invigorating substances" (the word "honnone" had not been coined) with injections of testicular extracts he could ward off the ills of old age and rejuvenate the body. After many disappointing trials, organ transplantation was considered to be a more permanent remedy. Morris was one of those interested in these ideas, but it was Lydston (1916), Stanley (1922) and Thorek (1924) in the United States and Voronoff (1923) in France who put them into practice (Hamilton 1986). Their widely publicized claims were discredited in the 1930s by better experiments and the availability of pure sex honnones. Testicular transplantation as a rejuvenation therapy fell into disgrace and, although rather less publicity surrounded ovarian grafting, the latter suffered a similar fate (Pettinari 1928). The subject of gonadal transplantation became tainted in the eyes of respectable scientists and was not regarded as fit for serious investigation until many years later.
The modern era of gonadal transplantation opened in the 1950s when inbred strains of laboratory rodents became available. When the potential problems of the allograft reaction had been recognized and overcome, biologists were able to concentrate on improving surgical techniques and applying them to important biological questions. Summarizing the evidence accumulating up to 1965, Krohn demonstrated that ovarian transplantation could be remarkably efficient in animals when due care was taken. Transplantation of testes attracted less investigation because of greater technical difficulty and less obvious applications. But, some 30 years on, the prospects for gonadal transplantation in a range of circumstances are greater than ever before.
2.3 Ovarian Transplantation
2.3.1 Autografts
Vascular anastomosis carries the major advantage of re-establishing a circulation within minutes or, at most, hours after removing the organ
22 R.G. Gosden
and thus minimizes necrotic changes resulting from ischaemia. The anticipated benefits of this procedure in terms of follicle survival have not, however, been assessed quantitatively yet (Dempster 1954; Goding et al. 1967; Betteridge 1970; Winston and McClure Browne 1974). The first autografts of human ovaries mostly involved simple implantation and/or transposition of the ovary to overcome tubal obstruction. However, there have been a few isolated cases in recent years in which the ovary has been moved to another site of the body (e.g. axilla) using vascular surgery to avoid the damaging effects of cancer treatment (e.g. irradiation for Wilm's tumour). It is, however, too early to properly assess the benefits of this strategy.
Until microsurgical skills were developed, most ovarian grafts were simply attached to the vacant ovarian pedicle or to another site in the body. This procedure demands much less skill than microsurgery and is more suitable when cryopreservation of the ovary is required because frozen storage of whole organs is impracticable (except perhaps for small laboratory animals and fetuses). Parrot (1960) stored the ovaries of mice at -79°C using glycerol as a cryoprotectant and succeeded in restoring fertility to the same animals after thawing and grafting. In principle, however, it is desirable to minimize the mass of tissue in order to control chemical and thermal equilibration during cooling. When primordial follicles were disaggregated enzymatically from mouse ovaries and stored at -196°C in medium containing dimethylsulfoxide, they retained their fertility on being returned to the body (Carroll and Gosden 1993). It is doubtful whether such procedures can be adapted successfully for larger ovaries, which require a long incubation period in proteolytic enzymes, and even then yield very few viable follicles (Roy and Treacy 1993).
Since the majority of small follicles are peripheral, a thin slice of ovarian cortex is a compromise which has proven successful in the sheep (Gosden et al. 1994). Nine months after grafting, II of 12 autografts (half of which were frozen for 3 weeks) still contained follicles and two of the six sheep conceived and carried their pregnancy to full term. Since the size, structure and composition of sheep and human ovaries are not appreciably different, this method may one day find a place for protecting ovaries during cancer treatment. In view of the release of free radicals and possible tissue damage during reperfusion of organs following ischaemia (Sugino et al. 1993), the impact of any
Transplantation of Male and Female Germ and Somatic Celis 23
transplantation operation on the quality of the gametes requires investigation.
2.3.2 Allografts
Simple implantation of a whole or part of an ovary of one animal into the evacuated ovarian capsule of another sometimes gives impressive results. For example, Krohn (1965) obtained 17 litters and 79 offspring from a single mouse after transplantation, despite the probability of losing half the primordial follicle population during ischaemia. In rodents, the functional lifespan of grafts may be limited by ageing of the neuroendocrine mechanisms responsible for generating ovulatory gonadotrophin surges because these mechanisms are impaired by long-term oestrogen exposure (Finch et al. 1984; Brawer et al. 1993). Accordingly, grafts may function for longer when animals have been chronically ovariectomized (Aschheim 1965; Felicio et al. 1983). By contrast, the equivalent mechanisms in humans are potentially operative after the menopause and, there is therefore no apparent upper age limit for ovarian graft function.
Experiments have not yet defined the lower age limit for germ cell donation. Fetuses possess the largest number of germ cells of any age, and in humans these amount to several million at mid-gestation (Baker 1963). Germ cells in the ovaries of fetal rodents continue developing and form follicles after grafting to adult hosts, where they can generate fertility (Russell and Gower 1950; Hashimoto et al. 1992; Fig. I). In view of these results, there is at least the technical possibility that ovaries from human abortuses would function in sterile women (Gosden 1992a,b).
The main advantages of isolating follicles before transplantation are to control the numbers and quality and improve cryopreservation. Primordial and small, growing follicles isolated from the delicate ovaries of immature mice with collagenase can be transferred to host ovaries that are either sterilized by X-irradiation or by precocious ageing (e.g. CBA strain). Since mouse ovaries are small, injecting suspended cells into the stroma is unsatisfactory and reaggregation in a fibrin or collagen gel vehicle has been adopted. The scrambled mixture of small follicles and other ovarian cells become reorganized to restore normal ovarian
24 R.G. Gosden
Fig 1. Graft of a fetal ovary (day 16 post-coitum) into an X-ray-sterilized ovary of an adult, syngeneic mouse. Three weeks after surgery, the germ cells have formed follicles which have grown to reach Graafian sizes. The lower diminutive ovary is the ungrafted control from the contralateral side of the same host. Haematoxylin and eosin (scale bar = 350 Ilm)
morphology and physiology, including the ovulation of fertile oocytes (Gosden 1990; Telfer et al. 1990; Carroll and Gosden 1993).
2.3.3 Xenografts
Xenografts are more than experimental curiosities for, although there is no intention of using them for producing oocytes to assist fertility, they serve useful roles for experimental purposes. Small laboratory animals are used as hosts because they are genetically well defined and can be rendered immunologically tolerant by, for example, (1) thymectomy with irradiation, (2) the mutantion nude (deficient in T lymphocytes) and (3) the mutation scm (severe combined immunodeficiency) (lack ofB and T lymphocytes) (Bosma et al. 1983). Such models already have a place in the study of growth and metastasis of ovarian tumours (Kleine 1986), and other applications are now emerging.
Recently, SCID mice have been used to host cortical slices of sheep and human ovaries for assessing the survival of either fresh or frozen-
Transplantation of Male and Female Germ and Somatic Cells 25
Fig 2. Cortical slice of a sheep ovary 6 months after grafting under the renal capsule of a SCID mouse. Follicles range from primordial (arrow) to Graafian sizes (G). Haematoxylin and eosin (scale hal' = 350 11m)
stored autografts (Boulton et al. 1993; Wade and Gosden 1994). The method involves inserting a small piece of tissue under the renal capsules of hosts, whose own ovaries need not be removed. Shortly after grafting, any developing follicles die but in the following weeks surviving primordial follicles give rise to a new population offollicles, some of which reach 4 mm in diameter (Fig. 2). Whether the final stages of development fail in the xenograft because of inappropriate gonadotrophic stimulation or for other reasons is not known.
2.4 Testicular Transplantation
2.4.1 Whole Organ
Operations such as testicular transplantation are indicated in cases of accidental castration or agonadism because they (re)establish the dual functions of the gland. Since the testis is a bulky organ with major functional compartments in the medulla, organ slices are never appropriate and the operation requires microsurgical anastomosis of testicular
26 R.G. Gosden
blood vessels and the excurrent ducts or vas deferens. This is therefore a more demanding procedure than orchiopexy to transpose maldescended testes. Re-establishing a circulation must also be carried out without delay because the seminiferous epithelium is very sensitive to ischaemia and necrosis begins within 2--4 h even in chilled organs. Nevertheless, with due care and skill surprisingly successful results have been obtained in animals (Auaran et al. 1966; Lee et al. 1971; Gittes et al. 1972).
While human testicular transplantation was frequently attempted in an earlier era, these operations usually involved organ slices and were not intended to overcome infertility. It was only recently that the first successful transplant to an anorchic man was carried out (Silber 1978). This case was exceptional, however, because the donor was lucky enough to have a genetically identical twin brother who was prepared to donate one organ after completing his family size. The single testicle secreted testosterone and produced sperm after the operation, and the patient later impregnated his wife. Such cases will always be very rare and, if there is any future for the operation, whole testicle transplantation will depend on inducing specific tolerance of hosts to allografts.
2.4.2 Stem Cells
Deficiency of either or both Leydig cells and germ cells can theoretically be reversed by transferring cells from a donor or by returning the patient's own cells which were removed for the protection of cold storage during chemotherapy/radiotherapy. The prospect of such procedures is still speCUlative in humans and lacks much experimental foundation, but the main principles are worth considering even though the task of restoring the production of millions of spermatozoa per day is obviously much harder than of one ovulation per month in the grafted ovary.
Despite the presence of a developing tubule system, the postnatal testis still has some cellular plasticity, -which is encouraging. For example, when the immature testis is completely disaggregated into a single cell preparation and then reag~d and transplanted to an adult host, the Sertoli cells reorganize to-form tubules of normal diameter. Most of the tubules are sterile, as expected, but some spermatogonial stem cells are found within them and undergo meiotic transformati-
Transplantation of Male and Female Germ and Somatic Cells 27
on to fonn spennatocytes (Fig. 3). Interestingly, these reconstituted testes are deficient in interstitial cells.
Although there are no published records of attempts to isolate spermatogonia for reintroduction into sterile tubules, there have been attempts to produce interstitial cell grafts for specifically restoring testosterone secretion. These grafts only produced mild androgenization and incomplete restoration of androgen-dependent target organs; furthermore, the sexual behaviour of the castrated hosts failed to return to normal (Fox et al. 1973; Boyle et al. 1976; Tai et al. 1989; van Dam et al. 1989). Since the grafts grew well (Fig. 4), the problem may lie in the abnormal tissue architecture and/or the absence of paracrine signals from cells in the seminiferous tubules, which may be necessary for nonnal cellular differentiation (Tahka 1986).
2.5 Survival of Allografts
In the past, there were many studies purporting to show that the gonads carry a degree of immunological privilege. The distinguished transplant surgeon Sir Michael Woodruff, reviewing the evidence, concluded that "It is almost impossible ... after reading critically the early literature on the subject, to escape the conclusion that ovarian homotransplants may survive for weeks or even months in various different species" (Woodruff 1960). Nevertheless, recent experience shows that where graft and host differ at histocompatibility loci, ovaries are rapidly rejected unless drugs or antibodies are administered to eliminate peripheral immunocompetent cells (Cornier et al. 1985; Gosden and Murray, unpublished). Furthennore, the immaturity of fetal cells has been overestimated as they evidently express transplantation antigens and enjoy no peculiar privilege as allografts (Seigler and Metzgar 1970). It is only when grafts are placed in immunologically privileged sites, such as the anterior chamber of the eye (Dameron 1951), that allografts are routinely tolerated. Such results should be anticipated since gonadal cells, at least in the female, express antigens of the major histocompatibility complex (MHC) (Table 2), and gene expression is amplified further when these cells are exposed to y-interferon released by immunocompetent cells (Hill et al. 1990; Gosden and Murray, unpublished).
28 R.G. Gosden
Fig 3. Cross-section of a seminiferous tubule obtained after disaggregation of an infant mouse testis followed by reaggregation in a fibrin clot and transplantation to the spermatic cord of a castrated host. Although most of the tubules were sterile and contained only Sertoli cells, this segment has spermatocytes undergoing meiosis (arrow). Haematoxylin and eosin (scale bar = 30 f.lm)
Fig 4. Interstitial cell graft obtained from isolated-reaggregated cells from immature mouse testes and attached to the spermatic cord of a castrated host. Haematoxylin and eosin (scale bar = 130 f.lm). From Gosden (1992a)
Transplantation of Male and Female Germ and Somatic Cells 29
Table 2. Expression of antigens of the major histocompatibility complex by granulosa cells
Species
Human Mouse
Class I
++ +/-
Class IT
The results were based on immunohistochemistry and flow cytometry and scored semi-quantitatively on a scale ranging from - (undetectable) to +++ (intense signal). (R.G. Gosden and A.A. Murray, unpublished)
The possibility that follicular oocytes and spennatogenic cells also express MHC antigens remains unresolved (Goldbard et al. 1985; Dohr et al. 1987), but it may not be the most critical factor for the afferent arm of graft destruction. Indeed, the locations of these cells within the follicle wall or the blood-testis barrier would probably afford them protection from immunosurveillance as allografts. Whether it is feasible to overcome sterility by transferring donor genn cells into the protective environment of the host's own somatic supporting cells is a matter of much greater doubt.
2.6 Prospects
For much of their early history, gonadal grafting techniques were controversial and not very successful. In recent years, the range of techniques in experimental transplantation has broadened and potential applications have expanded as cryobiology has become more sophisticated (Fig. 5). Whole organ transplantation using vascular anastomosis, whether as autografts or allografts, is most attractive since it causes less tissue damage than simple implantation, which leaves follicle survival to chance. Microsurgery is, however, unnecessary in many experimental circumstances and is incompatible with freeze-storage. Ovarian cortical slices appear to offer a better prospect of clinical application than either whole organ grafting or the injection of primordial follicles recovered from human ovaries. Cellular repopulation of testes that are azoospermic or deficient in Leydig cells is theoretically possible but faces major technical hurdles.
30 R.G. Gosden
Fig S. Schematic diagram of potential strategies for transplanting gonadal cells
While gonadal autografts offer the most tractable path for immediate progress, the potential for allografting should not be neglected. Allografting presents two major problems. The first is the source of donated cells because, although egg donation and artificial insemination are already widely accepted, new ethical and legal questions will have to be settled, particularly if cells from aborted fetuses are to be used. The second is the problem of the allograft reaction because, although this can be abrogated by immunosuppression, these drugs carry risks which are unacceptable for patients in whom there is only a reproductive disorder. Fortunately, neither of these problems is encountered with the autotransplantation, where most progress is expected and the needs for testing the quality of gametes and improving the results are most urgent.
Acknowledgements. I gratefully acknowledge financial support from The Wellcome Trust and The Galton Institute (London).
References
Adams CE (1979) Consequences of accelerated ovum transport, including a re-evaluation of Estes' operation. J Reprod Fertil 55:239-246
Transplantation of Male and Female Germ and Somatic Cells 31
Aschheim P (1965) Resultats fournis par la gref'fe heterochrnne dcs ovaircs dans I'etude de la regulation hypothalamo-hypophyso-ovarienne de la ratte senile. Gerontologia 10:65-75
Attaran SE, Hodges CV. Crary LS Jr, Vangalder GC, Lawson RK. Ellis LR (1966) Homotransplants of the testis. J Urnl 95:387-389
Baker TG (1963) A quantitative and cytological study of germ cells in human ovaries. Proc R Soc B 158:417--433
Bert P (1865) Sur la gref"fe animale. C R Soc Bioi (Paris) 61 :587 Berthold AA (1849) Transplantation der Hoden. Arch Anat Physiol Med
16:42--46 Betteridge KJ (1970) Homotransplantation of ovaries with vascular anastomo
ses in rabbits: response of transplants to human chorionic gonadotrophin. J EndocrinoI47:451--461
Biggers JD (1984) In vitro fertilization and embryo transfer in historical perspective. In: Trounson A, Wood C (cds) In vitro fertilization and embryo transfer. Churchill-Livingstone, Edinburgh, pp 3-15
Biskind GR, Kordan B. Biskind MS (1950) Ovary transplanted to spleen in rats: the effect of unilateral castration. pregnancy, and subsequent castration. Cancer Res 10:309-318
Bland KP. Donovan BT (1968) The effect of autotransplantation of the ovaries to the kidneys or uterus on the oestrous cycle of the guinea pig. J Endocr 41:95-103
Block E (1952) Quantitative morphological investigations of the follicular system in women. Variations at different ages. Acta Anat 14: I 08-123.
Bosma GC, Custer RP. Bosma MJ (1983) A severe combined immunodeficiency mutation in the mouse. Nature 301:527-530
Boulton M, Gosden RG. Webb R (1993) The potential of the ovariectomized SCID mouse model to investigate genotypic differences in follicular growth from lambs possessing the Booroola fecundity (Fec 13 ) gene. J Reprod Fert (Abstr Ser) 12:67
Boyle PF. Fox M. Slater D (1976) Transplantation of interstitial cells of the testis: effect of implant site. graft mass and ischaemia. Br J UroI47:891-898
Brawer JR. Beaudet A, Desjardins Gc, Schipper HM (1993) Pathologic effect of estradiol on the hypothalamus. BioI Reprod 49:647-652
Brown-Scquard CE (1889) Des effets produits chez l'homme par des injections SOLlscLltanees d'un liquide retire des testicules frais de cobayc et de chien. Comptes Rendus Hebdomaidaires de Seances et Memoire de la Societe de Biologie I (9th ser): 415--419
Carroll J, Gosden RG (1993) Transplantation of frozen-thawed mouse primordial follicles. Hum Reprod 8: 1163-1167
32 R.G. Gosden
Cornier E, Sibella P, Chatelet F (1985) Etudes histologiques et devenir fonctionnel des greffes de trompe et d' ovaire chez la ratte (isogreffes et allogreffes traittees par cyclosporine A). J Gynecol Obstet BioI Reprod 14:567-573
Craft I, Brinsden P, Simons E, Lewis P, Moye MA, Hansen J, Robertson D, Serhal P (1987) Successful births after ovum donation. Lancet i:916-917
Dameron IT (1951) The anterior chamber of the eye for investigative purposes. A site for transplantation of fetal endocrine tissues and cancer, and for the study of tissue reaction. Surgery 30:787-799
Dempster WJ (1954) A technique for the study of the autotransplanted kidney, adrenals and ovary of the dog. J Physiol (Lond) 124:XV-XVI
Dohr GA, Motter W, Leitinger S, Desoye G, Urdl W, Winter R, Wilders-Truschnig MM, Uchanska-Ziegler B, Ziegler A (1987) Lack of expression of histocompatibility leukocyte antigen class I and class II molecules on the human oocyte. J Immunol 138:3766-3770
Edwards RG, Morcos S, MacNamee M, Balmaceda JP, Walters DE, Asch R (1991) High fecundity of amenorrhoeic women in embryo-transfer programmes. Lancet 338:292-294
Faddy MJ, Gosden RG, Gougeon A, Richardson SJ, Nelson JF (1992) Accelerated disappearance of ovarian follicles in mid-life-implications for forecasting menopause. Hum Reprod 7: 1342-1346
Felicio LS, Nelson IF, Gosden RG, Finch CE (1983) Restoration of ovulatory cycles by young ovarian grafts in aging mice: potentiation by long-term ovariectomy decreases with age. Proc Nat! Acad Sci USA 80:6076-6080
Finch CE, Felicio LS, Mobbs CV, Nelson JF (1984) Ovarian and steroidal influences on neuroendocrine aging processes in female rodents. Endocrine Rev 5:467-497
Foa C (1900) La greffe des ovaires en relation avec quelques questions de biologie generale. Arch Ital BioI 34:43-73
Fox M, Boyle PF, Hammonds JC (1973) Transplantation of interstitial cells of the testis. Br I UroI45:696-70l
Gittes RF, Altwein JE, Yen SSC, Lee S (1972) Testicular transplantation in the rat: long-term gonadotropin and testosterone radioimmunoassays. Surgery 72:187-192
Goding JR, McCracken JA, Baird DT (1967) The study of ovarian function in the ewe by means of a vascular autotransplantation technique. J Endocr 39:37-52
Goldbard SB, Gollnick SO, Warner CM (1985) Synthesis of H-2 antigens by preimplantation mouse embryos. BioI Reprod 33:30-36
Gosden RG (1990) Restitution of fertility in sterilized mice by transferring primordial ovarian follicles. Hum Reprod 5:499-504
Transplantation of Male and Female Germ and Somatic Cells 33
Gosden RG (1992a) Transplantation of ovaries and testes. In: Edwards RG (ed) Fetal transplants in medicine. Cambridge University Press, Cambridge. pp 253-279
Gosden RG (1992b) Transplantation of fetal germ cells. J Assist Reprod Genet 9:118-123
Gosden RG, Murray A (1993) Transplantation of germ and follicle cells. In: Hillier SG (cd) IXth workshop on development and function of the reproductive organs, Ares-Serono symposium. Raven, New York, pp237-249
Gosden RG, Baird DT, Wade JC, Webb R (1994) Restoration of fertility to oophorectomized sheep by ovarian autografts stored at -I 96"C. Hum Reprod (in press)
Hamilton D (1986) The monkey gland affair. Chatto and Windus, London Hashimoto K. Noguchi M. Nakatsuji N (1992) Mouse offspring derived from
fetal ovaries or reaggregates which were cultured and transplanted into adult females. Dev Growth Differ 34:233-238
Hill JA, Welch WR Faris HMP. Anderson DJ (1990) Induction of class II major histocompatibility complex antigen expression in human granulosa cells by interferon gamma. A potential mechanism contributing to autoimmune failure. Am J Obstet Gynecol 162:534-540
Kleine W (1986) Prognostic significance of growth characteristics of xenotransplanted ovarian carcinomas into nude mice. Gyneeol Oncol 25:65-72
Knauer E ( 18(6) Einigc Versuche tiber Ovarientransplantation bei Kaninchen. Zentralbl Gynakolfur 20:524-528
Krohn PL (1965) Transplantation of endocrine organs. With special reference to the ovary. Br Med Bull 21: 157-162
Krohn PL (1977) Tr<illsplantation of the ovary In: Zuckerman Lord, Weir BJ (eds) The ovary, vol 2: physiology. Academic, New York, pp 101-128
Lee S. Tung KS, Orloff MJ (197 I) Testicular transplantation in the rat. Transplantation Proc 3:586-590
Lydston GF (1916) Sex gland implantation. Additional cases and conclusions to date. J Am Med Assoc 66: 1540-1543
Marshall FHA, Jolly W A (1908) On the results of heteroplastic ovarian transplantation as compared with those produced by transplantation in the same individual. Q J Exp Physiol 1: 115-120
Morris RT (1895) The ovarian graft. N Y Med J 62:436 Morris RT (1906) A case of heteroplastic ovarian grafting, followed by
pregnancy. and the delivery of a living child. Medical Record 69:697-698 Norman RL, Spies HG (1986) Cyclic ovarian function in a male macaque: ad
ditional evidence for a lack of sexual differentiation in tbe physiological
34 R.G. Gosden
mechanisms that regulate the cyclic release of gonadotropins in primates. Endocrinology 118 :2608-2610
Palmer IF (1837) The works of John Hunter, FRS, with notes, vol III. Green and Longman, London, p 273
Parrot DMV (1960) The fertility of mice with orthotopic ovarian grafts derived from frozen tissue. J Reprod Fertil 1:230-241
Pettinari V (1928) Greffe ovarienne et action endocrine de I'ovaire. Doin, Pans
Roy SK, Treacy BJ (1993) Isolation and long-term culture of human preantral follicles. Fertil Steril 59:783-790
Russell WL, Gower JS (1950) Offspring from transplanted ovaries of fetal mice homozygous for a lethal gene (Sp) that kills before birth. Genetics 35: 133
Sauer MV, Paulson RJ, Lobo R (1993) Pregnancy after age 50: application of oocyte donation to women after natural menopause. Lancet 341 :321-323
Seigler HF, Metzgar RS (1970) Embryonic development of human transplantation antigens. Transplantation 9:478-486
Silber SJ (1978) Transplantation of a human testis for anorchia. Fertil Steril 30:181-187
Stanley LL (1922) An analysis of one thousand testicular substance implantations. Endocrinology 6:787-794
Sugino N, Nakamura Y, Ok uno N, Ishimatu M, Teyama T, Kato H (1993) Effects of ovarian ischemia-reperfusion on luteal function in pregnant rats. Bioi Reprod 49:354-358
Tahka KM (1986) Current aspects of Leydig cell function and its regulation. J Reprod Fertil 78:367-380
Tai J, Johnson HW, Tze WJ (1989) Successful transplantation of Leydig cells in castrated inbred rats. Transplantation 47:1087-1089
Telfer E, Torrance C, Gosden RG (1990) Morphological study of cultured preantral ovarian follicles of mice after transplantation under the kidney capsule. J Reprod Fertil 89:565-571
Thorek M (1924) Experimental investigations of the role of the Leydig, seminiferous and Sertoli cells and effects of testicular transplantation. Endocrinology 8:61-90
Voronoff S (1923) Greffes testiculaires. Doin, Paris Van Dam JH, Teerds KI, Rommerts FFG (1989) Transplantation and sub
sequent recovery of small amounts of isolated Leydig cells. Arch Androl 22:123-129
Wade JC, Gosden RG (1994) Assessment of oocyte survival in ovarian cortical grafts after frozen storage and xenografting. In: Schoemaker J, Schats R
Transplantation of Male and Female Germ and Somatic Cells 35
(eds) Ovarian endocrinopathies (the 8th Reinier de Graaf symposium). Parthenon, Carnforth, UK (in press)
Winston RML, McClure Browne JC (1974) Pregnancy following autograft transplantation of fallopian tube and ovary in the rabbit. Lancet ii:494-495
Woodruff MFA (1960) The transplantation of tissues and organs. Thomas, Springfield
3 Proliferation and Differentiation of Testicular Interstitial Cells: Aspects of Leydig Cell Development in the (Pre)Pubertal and Adult Testis
K.J. Teerds, M.B. Veldhuizen-Tsoerkan, F.F.G. Rommerts, D.G. de Rooij, and J.H. Dorrington
3.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 37 3.2 Proliferation and Differentiation of Precursor Cells
into Immature (Adult-Type) Leydig Cells. . . . . . . . . . . . . . . . . . .. 40 3.2.1 The (PrelPubertal Testis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 40 3.2.2 The Adult Testis .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 43 3.3 Proliferation of Immature (Adult-Type) Leydig Cells. . . . . . . . . .. 47 3.3.1 The (Pre)Pubertal Testis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 47 3.3.2 The Adult Testis Following hCG or EDS Administration. . . . . . .. 52 3.4 Differentiation of Immature (Adult-Type) Leydig Cells ......... 53 3.5 Dynamics of the Leydig Cell Population
in the Adult Testis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 55 3.6 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 56 References .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 57
3.1 Introduction
It has been known for several decades that the gonadotrophins luteinizing hormone (LH) and follicle-stimulating hormone (FSH) are required for the normal growth and differentiation of Leydig cells and Sertoli cells, respectively (Greep et al. 1936; Christensen 1975). The actions of
38 K.J. Teerds et al.
these gonadotrophins on the somatic cells of the testis result in the establishment of the appropriate microenvironment in the seminiferous tubule that is conducive to the process of spermatogenesis (Dorrington and Armstrong 1979). One of the essential factors for the progression of the process of spermatogenesis is testosterone. The gonadotrophin LH acts on Leydig cells to stimulate the synthesis and secretion of this androgen, which subsequently acts as a paracrine factor within the seminiferous tubules (Ahmad et al. 1973; Sharpe 1987).
Because of the importance of testosterone in the progression of spermatogenesis, it is understandable that in the past 20 years the main interest in the study of testicular interstitial cells has been focused on the regulation of testosterone secretion by the Leydig cells. However, since the level of testosterone that can be attained at any stage depends on the numbers of Leydig cells in the interstitium, it is also important to understand how the Leydig cell popUlation is formed and which factors playa role in the regulation of this developmental process.
It is generally accepted that Leydig cells in the rat testis undergo two clearly defined periods of proliferation and differentiation, the first during fetal life and the second during the (pre)pubertal period (Hardy et al. 1989; Mendis-Handagama et al. 1987). The morphology and profile of secreted steroids by the Leydig cell popUlation in the fetal testis are unique to this phase of testicular development (de Kretser and Kerr 1988). After birth the number of fetal-type Leydig cells per testis decreases, although part of the fetal-type Leydig cell popUlation persists even in the adult testis (Kerr and Knell 1988).
The second generation of Leydig cells, the so-called adult-type Leydig cells are derived from the differentiation of mesenchymal-like precursor cells and from the proliferation of the newly formed Leydig cells (Hardy et al. 1989; Vergouwen et al. 1991). In the rat, this wave of adult-type Leydig cell proliferation is initiated between days 14 and 21 of postnatal life. Peak values are found around day 28, whereafter the proliferative activity of these cells ceases and becomes negligible around day 70 of age (Hardy et al. 1989; Teerds et al. 1988, 1 990a). In vivo studies have indicated that the differentiation of precursor cells into Leydig cells and proliferation and differentiation of newly formed Leydig cells is LH-dependent (Ketelslegers et al. 1978; Chemes et al. 1985, 1992; Teerds et al. 1989a). Acute administration of the LH analog human chorionic gonadotrophin (hCG) to prepubertal boys, whose te-
Proliferation and Differentiation
immature precursor cells 1---,,-------1~ Leydig cells I-----.-----~ I
differentiation differentiation
LH (+) LH (+) LH (+)
39
mature Leydig cells
Fig. 1. Model for the dcvclopmcnt of Lcydig cells. Differentiation of precursor cells into immature Leydig cells, and proliferation (circle with arrowhead) and differentiation of these immature cells into mature Leydig cells are luteinizing hormonc!human chorionic gonadotrophin (LHlhCG)-dependent processes
stes are devoid of morphologically recognizable Leydig cells, results in the differentiation of Leydig cell precursors (Chemes et al. 1985, 1(92). In hypophysectomized immature rats Leydig cells do not proliferate, but proliferation can be stimulated by LH treatment (Teerds et al. 1989a). Taken together, these data indicate that the generation of the full complement of adult-type Leydig cells that are steroidogenically active is dependent on the presence of LH in vivo (Fig. 1).
In the past few years more information has become available about the process of adult-type Leydig cell development and the regulatory role of LH and paracrine/autocrine factors, both in the immature and the adult animal. Most of this information has been gained using the rat model and this will be discussed in the following paragraphs. The description of the developmental pathway starts with the proliferation and differentiation of Leydig cell precursors to the stage at which they become morphologically recognizable Leydig cells (Sect. 2). In Sect. 3 the factors that affect the proliferation of these newly formed immature Leydig cells are discussed, while in Sect. 4 the differentiation process which these cells undergo after they cease to proliferate is described. When adulthood has been reached the Leydig cell population becomes quiescent, although under some experimental conditions it is still possible to stimulate proliferation and differentiation of precursor cells and proliferation of Leydig cells. This is discussed in Sect. 5. From the data summarized in Sect. 6 we propose a current model for the regulation of Leydig cell development.
40 K.J. Teerds et al.
3.2 Proliferation and Differentiation of Precursor Cells into Immature (Adult-Type) Leydig Cells
3.2.1 The (Pre)Pubertal Testis
As indicated in the introduction, the general opinion is that the adult population of Leydig cells (so-called adult-type Leydig cells) originates from differentiation of non-Leydig cell precursors in the (pre)pubertal testis (de Kretser and Kerr 1988). In the past it has also been postulated that adult-type Leydig cells may originate from the first generation of Leydig cells, the fetal-type Leydig cells that are formed during prenatal development. Fetal-type Leydig cells were thought to regress after birth and morphologically change into fibroblastlike cells which are then reactivated during testicular development and acquire adult-type characteristics (Gondos 1977). However, recent evidence suggests that, although fetal-type Leydig cells do undergo some morphological changes and cease to proliferate after birth, they can still be identified as fetaltype cells and persist even in the adult testis (Orth 1980; Kerr and Knell 1988; Kuopio et al. 1989). Definite evidence for the origin of the Leydig cell precursor has been provided recently by Hardy and coworkers who demonstrated in a pulse-chase experiment that labeled mesenchymallike interstitial cells differentiated into adult-type Leydig cells (Hardy et al. 1989).
In the immature rat the proliferative activity ofthe mesenchymal-like Leydig cell precursors is high until day 14 after birth and drops sharply between days 14 and 21 postpartum (Hardy et al. 1989). At the same time the mesenchymal-like precursor cells also start to differentiate into immature (adult-type) Leydig cells (henceforward called immature Leydig cells) (Hardy et al. 1989). It is not clear which factors in collaboration with LH are involved in stimulating the proliferative activity and inducing the differentiation of these Leydig cell precursors. To obtain more information about these processes, a few groups have investigated whether it is possible to induce precursor cell development in vitro.
Hardy et al. (1990, 1991) were the first to show that it is indeed possible to induce differentiation of precursor cells into immature Leydig cells in vitro. Precursor cell differentiation was dependent upon the presence of both LH and the Sa-reduced steroid dihydrotestosterone (DHT), LH or DHT alone having no effect (Hardy et al. 1990). The
Proliferation and Differentiation 41
observation that the fonnation of Leydig cells in vitro requires both LH and DHT was intriguing. We, therefore, investigated the role of LH in the process of precursor cell differentiation in vitro in more detail. Furthem10re, we considered that the age of the rats at the time of isolation may be important for the capacity of these cells to develop into Leydig cells in vitro. Hence, we have isolated interstitial cells from rats of different ages: ~-day-old rats (undifferentiated precursor cells), 13-day-old rats (start of differentiation phase of precursor cells), and 18-day-old rats (peak of differentiation phase of precursor cells). Leydig cell fonnation was measured by staining the cultured cells for the presence of 3~-hydroxysteroid dehydrogenase (3~-HSD), a marker enzyme for Leydig cells (Klinefelter et al. 1987). These experiments show that LH alone can stimulate the conversion of precursor cells into Leydig cells in vitro (Fig. 2, 3). The dose ofLH necessary to induce the differentiation of precursor cells is dependent upon the age of the rats at the time of interstitial cell isolation and the duration of the culture period (Fig. 3). For example in cultures of interstitial cells isolated from 13-day-old rat testes a LH dose of 0.1 ng/ml stimulates the conversion of precursor cells into 3~-HSD positive immature Leydig cells within 3 days, whereas this takes ~ days with cells isolated from testes of 8-day-old rats. For all three age groups we found that concomitantly with the rise of 3~-HSD positive cells, the binding of I 25I-labeled hCG and the production of cAMP also increases (Teerds, de Rooij and Rommerts, unpublished data). Our data obtained with interstitial cells isolated from testes of 1 ~-day-old rats confinn the results acquired by Hardy et al. (1990). These authors cultured cells isolated from 21-dayold rats for 3 days in the presence of LH, also without observing an increase in the percentage of 3~-HSD-positive cells. Thus within a culture period of 3 days, LH is indeed unable to induce the conversion of precursor cells into immature Leydig cells. However, when these cells were cultured for 8 days, we found that LH increased the percentage of 3~-HSD-positive cells twofold, indicating that LH induction of ditlerentiation of precursor cells isolated from 18- to 21-day-old rat testes takes more than 3 days. Obviously these Leydig cell precursors must possess functional LH receptors in order to respond to LH. This has recently been confinned by Shan and Hardy (1992), who showed that precursor cells express the LH receptor gene and can bind 125I-labeled hCG.
42 K.J. Teerds et al.
A
Fig. 2 A-D. Interstitial cells isolated from 13-day-old rat testes were cultured for I, 3, or 8 days in the absence or presence of luteinizing hormone (LH; 0.1 ng/ml). Cells were isolated using the procedures described by Khan et al. (1992a) with some modifications. Leydig cells are identified by enzyme histochemical staining for the presence of 3~-hydroxysteroid dehydrogenase. A, B Cells cultured for 1 or 3 days, respectively, without any additions. C, D Cells cultured for 3 or 8 days, respectively, in the presence of 0.1 ng/ml of LH. Original magnification, x525
Proliferation and Differentiation 43
With the establishment of this in vitro system for LH-dependent differentiation of precursor cells into immature Leydig cells, we are able to study the mechanism of LH-stimulated differentiation and the possible role of other factors in this process in more detail.
3.2.2 The Adult Testis
In the adult testis, proliferation and differentiation of Leydig cell precursors and the subsequent formation of new Leydig cells can be induced by daily injections with supraphysiological doses ofhCG or by administration of a single injection of the Leydig cell toxicant ethane dimethane sulphonate (EDS) (e.g., Christensen and Peacock 1980; Kerr et al. 1985; Teerds et al. 1988, 1992). In this section the process of precursor proliferation and differentiation up to the formation of morphologically recognizable Leydig cells in the adult testis is described.
Daily treatment with high doses of hCG stimulates the proliferation of mesenchymal-like cells, which are presumed to be the precursors of the Leydig cells both in the immature and the adult testis (Teerds et al. 1988; Hardy et al. 1989). hCG also stimulates the differentiation of these precursor cells into immature (adult-type) Leydig cells. Two hCGdependent waves of differentiation can be recognized; the first rapid wave of differentiation takes place within 2 days after the start of hCG treatment, whereas the second one becomes apparent approximately 8 days after the start of hCG treatment. This second wave is merely the result of differentiation of precursor cells that have first proliferated after elevation of plasma hCG levels (Teerds et al. 1988, 1992). The observation that hCG treatment can induce two separate waves of precursor cell differentiation may indicate that in the mature rat testis precursor cells are present in different stages of development. Elevation of plasma LH/hCG levels may complete the differentiation process of the more advanced precursor cells into recognizable Leydig cells rapidly, whereas the less-differentiated precursor cells may require a longer period before they acquire Leydig cell properties. This would indicate that the precursor cell population in the adult testis is heterogenous. Another explanation is that the rapidly differentiating cells are actually inactive Leydig cells which are not identified morphologically as Leydig cells; an increase in plasma LH/hCG levels induces the rapid activa-
44 K.J. Teerds et al.
am 0.1 10
LH in nglml
30r-------------------------------, • day a (C)
o day 3 25 OJ day 8
.!!! a; ~ 20
.~ ·in
8. o 15 til ::r
"" M
;I! 10
5
~. ::::::-:-.-.
LH in nglml
Fig. 3 A-C. Effects of different concentrations of luteinizing hormone (LH) on the formation of 3~-hydroxysteroid dehydrogenase (HSD)-positive Leydig cells in vitro. Interstitial cells were isolated from 8 (A), 13 (8), and 18 (C) day old rat testes, respectively, and cultured for 0 (C, control), 3 or 8 days
Proliferation and Differentiation
30,-------------------------------~
• day 0 (C)
o day3 25 0 day 8
c o 0.01 0 .1 10
LH In nQlml
45
tion of these cells into recognizable Leydig cells. It is clearly important to obtain more information about the precursor cell population to determine whether different types of Leydig cell precursors exist and to examine the way in which heG regulates precursor cell differentiation.
Another compound which can induce proliferation and differentiation of interstitial cells in the adult rat testis is the alkylating agent EDS. Following in vivo administration, this drug specifically destroys Leydig cells, a process followed by regeneration of a new Leydig cell population (Kerr et al. 1985, 1986; Molenaar et al. 1985; Bartlett et a!. 1986; Jackson et a!. 1986; Kerr and Donachie 1986; Morris et a!. 1986; O'Leary et al. 1986; Teerds et al. 1988, 1990a; Gaytan et a!. 1992). Within the first few days after EDS administration, when Leydig cells are degenerating, a wave of proliferation of mesenchymal-like cells occurs (Teerds et al. 1988, 1990a). Pulse-chase experiments have shown that these labeled cells are precursor cells which develop into the first new Leydig cells between 7 and 14 days after EDS administration (Teerds et a!. 1990a).
In EDS-treated rats, this regeneration process takes place in the presence of high plasma levels of LH due to the absence of the negative
46 K.J. Teerds et al.
feedback control of testosterone (Molenaar et al. 1985, 1986; Bartlett et al. 1986). Since hCG stimulates the overall developmental process in the adult testis (Teerds et al. 1988), it was investigated which developmental steps can take place without hCG/LH and which steps require gonadotrophin stimulation. In adult hypophysectomized, EDS-treated rats, proliferation of Leydig cell precursors and the first part of the differentiation process of these cells along the Leydig cell lineage can take place in the absence of LH and other pituitary hormones (Teerds et al. 1989b). To date it is not known which (growth) factors allow these processes to proceed and whether these factors are locally produced by, for instance, Sertoli cells or testicular macrophages.
Further development of the partially differentiated precursor cells in hypophysectomized, EDS-treated rats into Leydig cells occurred only after hCG administration (Teerds et al. 1989b). This indicates that gonadotrophins are only essential for the final part of this developmental process, i.e., the formation of new Leydig cells (Teerds et al. 1989b).
Moreover, since the precursor cells in hypophysectomized EDStreated rats respond to hCG treatment, these cells must possess LH receptors. Indeed, northern blot analysis of Poly(A)+ RNA isolated from testes of these rats has revealed that Leydig cell precursors in the testes of these EDS-treated rats contain the four different LH receptor transcripts that have been described previously in the (pre)pubertal and the adult rat testis (Fig. 4) (McFarland et al. 1989; Pakarinen et al. 1990; LaPolt et al. 1991; Veldhuizen-Tsoerkan, Ivell and Teerds, in preparation). The amounts of the 7.0, 4.2 and 2.5 kb transcripts in the hypophysectomized, EDS-treated testes are considerably lower while the amount of the 1.8 kb transcript is slightly higher, than in intact adult control testes (Fig. 4). Daily hCG injections result in the formation of new steroid producing Leydig cells (Teerds et al. 1989b), an increase in the amounts of all four transcripts but not in a change in the relative intensities of the bands (Fig. 4; Teerds et a\. 1989b; Veldhuizen-Tsoerkan et a\., in preparation). The same four transcripts of the LH receptor mRNA have been reported to be present in precursor cells isolated from (pre )pubertal rats (Shan and Hardy 1992). The regulation of this LH/hCG-dependent phase of precusor cell differentiation both in the immature and in the adult testis is presently under investigation.
Proliferation and Differentiation
sor------------------------------,
40
.~ 3C C .. " (j
Q) > 'jij 20 ~
10
o
• intact fa hypox o 7 days hCG
7.0 kb 4 .2 kb 2.S kb 1.8 kb
47
Fig. 4. Northern blot analysis of testicular mRNA for luteinizing hormone (LH) receptor. Poly (At mRNA was isolated from testes of intact adult rats (solid bars), of hypophysectomized ethane dimethane sulphonate (EDS)-treated adult rats (hypox; halched bars), and of hypophysectomi zed EDS-treated adult rats which received daily injections of 100 IU human chorionic gonadotrophin (hCG) for 7 days before being sacrificed (7 days hCG; slippled /Jars). The oligonucleotide probe for the LH receptor was hybridized with northern blots of rat testicular poly (At mRNA, and hybridization only occurred with RNA species of correct molecular sizes, according to the literature (McFarland et al. 1989). The integrated optical densities of the four di fferent transcript signals were determined by !BAS image processing of the autoradiograms (IBAS image analysis system; Zeiss/Kontron , Eching, Germany); the integrated optical densities of the LH receptor mRNA were normalized to those of d-glyceraldehyde-3 -phosphate dchydrogenase (GAPDH) and plotted as percentage of relative quantities
3.3 Proliferation of Immature (Adult-Type) Leydig Cells
3.3.1 The (Pre)Pubertal Testis
During the (pre)pubertal period from day 14 after birth onwards the Leydig cell population grows rapidly, not only by differentiation of precursor cells, as has been described in the previous section, but also by
48 K.J. Teerds et al.
proliferation of the newly formed, immature Leydig cells (Hardy et al. 1989). We have investigated factors or combinations of factors that stimulate newly formed immature Leydig cells to enter the S phase of the cell cycle and to initiate DNA synthesis in vitro. It is well established that at multiple sites growth factors are essential for the progression through G1 phase of the cell cycle leading to DNA synthesis and celI division (Pledger 1985; Fig. 5). We have focused on the identification of these growth factors that will promote DNA synthesis in the immature Leydig cells of the (pre)pubertal rat testis and have examined their interactions with LH (Teerds et al. 1989a; Khan et al. 1992a).
Locally produced peptide factors that affect the metabolic activities of testicular celI types are: insulin-like growth factor-I (IGF-I) (Handelsman et al. 1985; Cailleau et al. 1990; Lin et al. 1990; Moore and Morris 1993), transforming growth factor-a (TGF; Holmes et al. 1986; Skinner et al. 1989), TGF-~ (Skinner and Moses 1989; A valIet et al. 1988,1990), interleukin-l (Calkins et al. 1988; Khan et al. 1987, 1988; Verhoeven et al. 1988; Hales 1992; Mauduit et al. 1992) and basic fibroblast growth factor (bFGF) (Fauser et al. 1988; Murono and Washburn 1990; Sordoillet et al. 1988).
TGF-a and IGF-I have been identified as factors that promote DNA synthesis in immature Leydig cells isolated from 21-day-old rats. In contrast to the clear positive effects of LH treatment on Leydig cell proliferation in vivo (Teerds et al. 1989a), administration of LH to Leydig cells in vitro has a negligible stimulatory effect on DNA synthesis. However, when the cells are incubated concomitantly with LH and the growth factors TGF-a and IGF-I or insulin, a clear synergistic effect on the stimulation of DNA synthesis becomes apparent, suggesting that the ability of LH to stimulate DNA synthesis in vitro depends on the presence of these growth factors (Fig. 6; Khan et al. 1992a).
To determine whether TGF-a is present in the interstitium during the (pre )pubertal period when the Leydig cells undergo their wave of proliferation, immunohistochemical techniques were used. Approximately 50% of the Leydig cells present in the testis at 21 days of age contained TGF-a (Teerds et al. 1990b). As the wave of proliferation proceded, all Leydig cells aquired intense staining for TGF-a. The increase in the appearance of TGF-a-containing cells occurred during the period in which LH is essential for growth promotion, while in the absence ofLH in prepubertal, hypophysectomized rats Leydig cell proliferation did not
Proliferation and Differentiation
Commitment Point
G, -phase progression (dependent on progression factors)
Growth Factors Growth Factors
49
I Competent state' I __ ~--""-_+VI __ ~--"'----I.~ S-phase . . RNA and Protein Protein Synthesis '-----'---- --'
Synthesis
Competence Factors
Go arrest (quiescent cells
DNA Synthesis
~ ~-M---p-h-as-e--~I~
Cell Division
Fig. S. The mammalian cell cycle. Under the control of growth factors cells may be stimulated to enter the cell cycle
occur (Teerds et al. 1989a). IGF-I is also present in Leydig cells during the (pre)pubertal period (Cailleau et al. 1990). Therefore. we postulate that LH can only increase the number of Leydig cells in the (pre)pubertal testis in the presence of IGF-I and TGF-a. In addition to their growth promoting action, TGF-a and IGF-I have positive effects on androgen production (Lin et a!. 1986; Verhoeven and Cailleau 1986). Hence it is possible that in the rat the rise in testicular androgens at puberty (Tapanaien et al. 1984) also results from interactions between LH, TGF-a and IGF-I.
Whereas LH and the intragonadal factors TGF-a and IGF-I stimulate DNA synthesis in Leydig cells, TGF-p is the only factor we have investigated that can attenuate the response of the Leydig cells to these stimulatory agents (Fig. 7) (Khan et al. 1994). Immunolocalization studies ofTGF-p in the rat testis have shown that all Leydig cells contained TGF-PI at 7 days of age, but the percentage of positively stained Leydig cells declined thereafter, so that by 2l days of age only approximately
50 K.J. Teerds et al.
8,------------------------------,
• no treatment ~ pretreatment with 2ng LH/ml
6
4
2
c LH TGFC/. I+TGFC/.
Fig. 6. [3Hlthymidine incorporation into DNA by rat Leydig cells isolated from 21-day-old rats, and in culture treated with insulin and transforming growth factor (TGF)-cx. During the 48 h of culture from plating to the time of treatment, the Leydig cells were cultured in the absence (solid bars) or in the presence (hatched bars) of luteinizing hormone (LH; 2 ng/ml). The cells were then treated for the subsequent 18 h with LH (100 ng/ml), insulin (/; I Ilg/ml) and/or TGF-cx (10 ng/ml). Control (C) cultures were not treated during this 18-h period. eHlthymidine incorporation per microgram of protein was determined after a 4-h incubation with 0.5 IlCi eHlthymidine/0.5 ml culture medium. Data are means ± SO of three determinations. From Khan et al. (I992a) with minor modifications
50% of the cells contained TGF-pi (Teerds and Dorrington 1993). As the wave of Leydig cell proliferation progressed, all these cells ceased to express TGF-PI, so that by 35 days of age no TGF-p immunoreactivity could be detected in the interstitial space (Teerds and Dorrington 1993). Since TGF-p inhibited the actions of LH, TGF-cx and IGF-I on DNA synthesis in immature cells (Khan et al. 1994), we propose that this factor plays a physiological role to hold the growth of immature Leydig cells in abeyance until the (pre)pubertal stage, even when the tissue is exposed to intragonadal stimulators (e.g., TGF-cx and IGF-I). In addition, TGF-p inhibits androgen production by immature Leydig cells (Lin et al. 1987; Morera et al. 1988), and thus its presence in the interstitial
Proliferation and Differentiation 51
4 • ·TGFI} 12 +TGFIl
c-O§! e a. Ol ~
"I ~ )(
E a. :E.
'" c '5 'E >-J: t-;-
.... I
0 C TGFo. i+TGFo.
Fig. 7. ,3Hlthymidine incorporation into DNA by rat Leydig cells isolated from 21 -day-old rats cultured in the continuous presence of luteinizing honnone (LH; 2 ng/mll. Effects of insulin (I; 0.5 IJ-g/mll, transforming growth factor (TGF)-a (10 ng/ml) and a combination of insulin and TGF-a on Leydig cell DNA synthesis in the presence or absence of TGF-~ (10 ng/mll. For further ex peri mental detai Is see legend to Fig. 6. C, control. From Khan et al. (1994) with minor modifications
compartment prior to puberty may suppress steroidogenesis as well as proliferation, thereby avoiding precocious puberty. The decline in TGF~ levels during the (pre)pubertal period and its subsequent loss from the interstitial compartment would allow LH together with TGF-a and IGF-I to produce a full complement of steroid producing Leydig cells.
As indicated above, most studies investigating the role of growth factors in Leydig cell development have focused on TGF-a, TGF-~ IGF-I and their interaction with LH. Recently it has become clear that other locally produced factors may also be involved in the regulation of this process. One of these factors is interleukin-I, since high levels of an interleukin-l-like factor have been found in human and rat testis (Khan et aL 1987, 1988). This factor stimulated the proliferation of a variety of cells and interleukin-I receptors have been localized predominantly in the interstitial compartment (Takao et aL 1990). Indeed Khan et aL (1992b) have shown pronounced stimulatory effects of interleukin-l ~
52 K.J. Teerds et al.
on DNA synthesis of Leydig cells isolated from immature rat testes, while the effects of LH, TGF-ex, and insulin on Leydig cell growth were significantly enhanced by interleukin-1~. Although the interleukin-1 ex and interleukin-1 ~ genes have been reported to be expressed in adult Leydig cells and macrophages (Lin et al. 1993), it remains to be investigated whether the interleukin-1 ex and interleukin-l ~ proteins are also synthesized in the (pre)pubertal testis to act as autocrine and/or paracrine regulators of Leydig cell proliferation.
There are several other factors, e.g., seminiferous growth factor (SGF), Sertoli cell secreted growth factor (SCSGF), nerve growth factor (NGF) and bFGF, that are also locally produced within the testis and which may play a local role during testicular development (see, for example, reviews by Bellve and Zheng 1989; Spiteri-Grech and Nieschlag 1993). Although some of these factors regulate functional properties of Leydig cells, no information is available on the possible role of these (growth) factors on Leydig cell proliferation during the (pre )pubertal growth phase.
3.3.2 The Adult Testis Following hCG or EDS Administration
Continuous hCG administration stimulates the proliferation of Leydig cells from day 4 of treatment onwards, reaching maximal values after 7 days (Teerds et al. 1988). Pulse-chase experiments suggest that part of the dividing Leydig cells are newly formed immature cells, although it cannot be excluded that the existing mature Leydig cells also contribute to the pool of proliferating cells following daily hCG treatment (Teerds et al. 1988; Teerds, de Rooij, Rommerts, Wensing, unpublished data). Further aspects of the proliferation of the existing Leydig cells will be discussed in Sect. 5. Taken together, these data indicate that elevated levels of LH/hCG can stimulate Leydig cell proliferation not only during (pre )puberty but also in mature rats.
As has been indicated earlier, following EDS administration the first new Leydig cells develop between 7 and 14 days after EDS (Kerr et al. 1985; Molenaar et al. 1985; Bartlett et al. 1986; Jackson et al. 1986; Teerds et al. 1990a). These newly formed Leydig cells, which have characteristics in common with the immature Leydig cells in the (pre)pubertal testis (Vreeburg et al. 1988; Myers and Abney 1990;
Proliferation and Differentiation 53
O'Shaughnessy and Murphy 1991), are proliferating actively. Peak values in Leydig cells labeled in S-phase and in M-phase of the cell cycle are found at days 21 and 22 after EDS, respectively. This is followed by a gradual decline (Teerds et al. 1990a; Myers and Abney 1991; Gaytan et al. 1992).
To date the factors involved in the regulation of the increased proliferative activity of the newly formed Leydig cells after EDS are unknown. Plasma LH levels are elevated up to day 14 after EDS due to the absence of testosterone-producing Leydig cells. Between days 14 and 21 after EDS, however, the testosterone produced by the newly formed Leydig cells has an effect on the pituitary via a feedback mechanism, resulting in a reduction of plasma LH levels. By 21 days after EDS administration plasma LH levels have returned to the control range (Jackson et al. 1986; Molenaar et al. 1986). Since LH levels are not elevated anymore at the time when Leydig cells are actively proliferating, it appears that other factors, possibly together with LH, may be responsible for the stimulation of Leydig cell proliferation. This situation may be more or less comparable to that of the (pre)pubertal rat, where in vivo immature Leydig cell proliferation is also maximally stimulated when LH levels are not significantly elevated (Odell and Swerdloff 1976). As has been indicated earlier, in vitro studies have shown that in order to stimulate proliferation of immature Leydig cells, LH requires the presence of growth factors like TGF-a and IGF-I (Khan et al. 1992a, 1994). Since these growth factors are locally produced in the testis in both the immature and the adult rat, it is tempting to speculate that they also playa role in the stimulation of Leydig cell proliferation in vivo (Handelsman et al. 1985; Skinner et al. 1989; Teerds et al. 1990b).
3.4 Differentiation oflmmature (Adult-Type) Leydig Cells
The proliferative activity of the newly formed, immature Leydig cells continues until 54-70 days of age in the rat (Teerds et al. 1988; Hardy et al. 1989). At the same time differentiation of these newly fonned Leydig cells is initiated, as becomes apparent by changes in the activities of steroidogenic enzymes. In the (pre)pubertal testis testosterone is not the primary androgen produced by the Leydig cells. The primary androgens
54 K.J. Teerds et al.
produced by these testes are androstenedione and the 5a-reduced androgens DHT and 5a-androstenediol (Wiebe 1976). Around 40 days of age, a developmental change takes place in the activity of several steroidogenic enzymes; the activities of 7a-hydroxylase (Rosness et al. 1977; Rommerts and van der Molen 1989) and 11 ~-hydroxysteroid dehydrogenase (Neumann et al. 1993) start to increase, while at the same time the activity of 5a-reductase declines rapidly, leaving testosterone as the major testicular androgen in the adult (Podesta and Rivarola 1974; van der Molen et al. 1975; Moger 1977).
During (pre)pubertal development in the rat also an increase in serum FSH levels occurs, which promotes the formation of gonadotrophin receptors in the testis (Odell and Swerdloff 1976). Treatment of immature hypophysectomized rats with FSH results in a rise in LH receptor numbers (Closset and Hennen 1989), leading to an increased sensitivity of Leydig cells for LH (Ojeda and Urbanski 1988). Since Leydig cells do not have FSH receptors (Orth and Christensen 1977), the effects of FSH on Leydig cells are presumably mediated by the secretion of some unknown factor(s) by the Sertoli cells. The precise regulation of the differentiation of immature adult-type Leydig cells in the (pre)pubertal testis to mature cells in the adult testis and the role of LH and Sertoli cell secreted factors in this differentiation process are not yet known.
The newly formed Leydig cells after EDS administration to adult rats have many characteristics in common with the differentiating immature Leydig cells in the (pre)pubertal testis. Both cell types for instance undergo a similar phase of proliferation (Hardy et al. 1989; Teerds et al. 1990a; Myers and Abney 1991; Gaytan et al. 1992; Sect. 3.2). Furthermore, not only the morphology of the regenerating Leydig cells after EDS, but also the changes that occur in the pattern of steroids secreted during differentiation show clear similarities with differentiating immature Leydig cells in the (pre)pubertal testis (Vreeburg et al. 1988; Myers and Abney 1990; O'Shaughnessy and Murphy 1991). These data and the fact that both cell types are insensitive to (another) injection of EDS (Morris 1985; Kerr and Knell 1987; Rommerts et al. 1988; Edwards et al. 1989; KeIce et al. 1991) have led to the hypothesis that the newly formed Leydig cells after EDS administration are of the adult type and undergo differentiation along a similar pathway as the immature Leydig cells in the (pre )pubertal testis.
Proliferation and Differentiation
3.5 Dynamics of the Leydig Cell Population in the Adult Testis
55
Although the growth stimulating factors (LH, TGF-a, and IGF-I) remain present after the rat testis has fully matured around day 70 postpartum, the processes of growth and differentiation have been completed to generate the mature population (Christensen 1975; Teerds et al. 19X8). A recent study of the capacity of cell renewal in the interstitial tissue of the normal adult rat testis has shown that the Leydig cell population in the adult does undergo turnover. This process takes place at a slow rate, the turnover time being between 242 days and the maximal life span of the animal (Teerds et al. 1989c). The fact that the turnover rate is rather slow suggests that in the adult animal either most Leydig cells have undergone terminal differentiation or an unidentified growth inhibiting factor is produced which specifically blocks proliferation of Leydig cells.
As has been indicated earlier, it is possible to stimulate the proliferation of both newly formed immature and existing mature Leydig cells in adult rats by daily administration of supraphysiological doses of hCG (Christensen and Peacock 1980; Teerds et al. 1988; Teerds, de Rooij, Rommerts, Wensing, unpublished observations). The stimulatory effects of hCG on the proliferation of immature and mature Leydig cells are surprising when one takes into account that the doses of hCG used cause a rapid loss of available LH/hCG receptors, a decrease in LH receptor mRNA levels, and desensitization of adenylate cyclase (Sharpe 1976; Haour and Saez 1977; Hsueh et al. 1977; Calvo et al. 1984; Nozu et al. 19X I; Pakarinen et al. 1990; LaPolt et al. 1991). Moreover, this also shows that not all mature Leydig cells have undergone terminal differentiation, although under normal conditions they are inhibited in their proliferative activity. The dramatic rise in plasma LH/hCG which occurs as a result of daily hCG treatment is presumably able to overcome the inhibited condition and could presumably, in collaboration with other (growth) factors, stimulate the proliferative capacity of these existing Leydig cells.
Taken together, these data show that under normal conditions the mature Leydig cell popUlation in the adult testis is a stable population of cells with very limited turnover. Only supraphysiological doses ofhCG, which induce signs of desensitization, can stimulate cell dynamics.
56 K.J. Teerds et al.
3.6 Conclusions
The data presented in the previous sections are summarized in a model (Fig. 8). Briefly, Leydig cell precursors in both (pre)pubertal and adult animals originate from a heterogenous mesenchymal-like precursor cell popUlation. Proliferation and/or differentiation of these cells can be stimulated, and this is possibly mediated and modulated by growth factors and by LH. The presence of functional LH receptors has been confirmed by northern blot analysis with a specific probe that recognizes LH receptor mRNA and by LH binding studies. The final phase of precursor cell differentiation up to the stage that they are converted into Leydig cells is highly dependent on LH.
The newly formed, immature Leydig cells have the capacity to undergo at least one cycle of cell replication. In vivo studies have shown that a gonadotrophin stimulus is essential for Leydig cell proliferation. However, in vitro studies indicate that LH's effectiveness in stimulating Leydig cell proliferation is dependent on the relative amounts ofTGF-a and TGF-~. LH requires TGF-a together with IGF-I to stimulate growth, while other factors such as interleukin-l ~ may also be involved. The actions of these factors are abolished by TGF-~. The in vivo observation in the (pre)pubertal rat that TGF-~ disappears from the Leydig cells at the onset of the wave of cell proliferation and the generation of a population of Leydig cells enriched in TGF-a is consistent with this hypothesis. Since TGF-a and IGF-I are present in the mature Leydig cells in the adult testis, it is tempting to speculate that these factors also playa role in the stimulation of Leydig cell proliferation in vivo in the adult rat.
Immature Leydig cells continue to proliferate for several weeks both in the (pre)pubertal and in the adult testis following EDS administration. At the same time differentiation of these cells is initiated, as becomes apparent by changes in the activities of steroidogenic enzymes such as Sa-reductase. However, the precise role of LH and possibly other locally produced factors in the differentiation of immature Leydig cells into mature cells are not yet known.
The cessation of Leydig cell growth in the adult testis, in the presence of growth stimulatory factors, is not understood. However, a dramatic rise in plasma LH/hCG levels can overcome this inhibition. Hence, the Leydig cell population in the adult testis consists of a semi-stable
Proliferation and Differentiation
proliferation
precursor cells
proliferation LH
Leyd.g cells (immalUre)
Sa- reductase ""
57
~t~oHn(+) r--~-=---' steroidogenic actIvity (+)
Leydig cells (mature)
Sa·reductase t
testosterone
Fig. 8. Current model for the development of the mature population of Leydig cells. The model is discussed in Sect. 6
population of cells which under normal physiological conditions does not undergo many changes after adulthood has been reached. However, experimental conditions can affect the dynamics of this cell population and stimulate both cell proliferation and steroidogenic activity.
Acknowledgements. Katja J. Teerds is recipient of a fellowship of the Royal Netherlands Academy of Arts and Sciences (KNA W fellow programme) .
References
Ahmad N, Haltmeyer GC, Eik-Nes KB (1973) Maintenance of spermatogenesis in rats with intratesticular implants containing testosterone or dihydrotestosterone (DHT). Bioi Reprod g:4II-419
Avallet 0 , Perrar-Sapori MH, Vigier M, Skalli M, Saez JM (InS) Comparison of the effects of transforming growth factor-~ and porcine Sertoli ccll conditioned medium on porcine Leydig cell function. In: Cooke BA, Shar
pe RM (cds) Molecular and cellular endocrinology of the testis , vol 50. Raven, New York, pp 319- 324
58 K.J. Teerds et al.
Avallet 0, Vigier M, Albalado V, de Cesaris P, Saez JM (1990) Transforming growth factor ~ gene expression in cultured porcine Sertoli cells and Leydig cells: effects of hormone and growth factors. Proceedings of the 6th European testis workshop on molecular and cellular endocrinology of the testis, abstract 013
Bartlett JMS, Kerr JB, Sharpe RM (1986) The effect of selective destruction and regeneration of rat Leydig cells on the intratesticular distribution of testosterone and morphology of the seminiferous epithelium. J Androl 7:240-253
Bellve AR, Zheng W (1989) Growth factors as autocrine and paracrine modulators of male gonadal functions. J Reprod Ferti!85:771-793
Cailleau J, Vermeire S, Verhoeven G (1990) Independent control of the production of insulin-like growth factor I and its binding protein by cultured testicular cells. Mol Cell Endocrinol 69:79-89
Calkins JH, Sigel MM, Nankin HR, Lin T (1988) Interleukin -I inhibits Leydig cell steroidogenesis in primary culture. Endocrinology 123: 1605-161 0
Calvo JC, Radicella JP, Pignataro OP, Charreau EH (1984) Effect of a second injection of human chorionic gonadotropin on the desensitized Leydig cells. Mol Cell EndocrinoI34:31-38
Chemes HE, Gottlieb SE, Pasqualini T, Domenichini E, Rivarola MA, Bergada C (1985) Response to acute hCG stimulation and steroidogenic potential of Leydig cell fibroblastic precursors in humans. J Androl 6: 102-112
Chemes H, Cigorraga S, Bergada C, Schteingart H, Rey R, Pellizzari E (1992) Isolation of human Leydig cell mesenchymal precursors from patients with the androgen insensitivity syndrome: testosterone production and response to human chorionic gonadotropin stimulation in culture. Bioi Reprod 46:793-801
Christensen AK (1975) Leydig cells. In:. Hamilton OW, Greep RO (eds) Handbook of physiology, vol 5. American Physiological Society, Washington, pp 57-94
Christensen AK, Peacock KC ( 1980) Increase in Leydig cell number in the testis of adult rats treated chronically with an excess of human chorionic gonadotropin. Bioi Reprod 22:383-391
C10sset J, Hennen G (1989) Biopotency of highly purified porcine FSH and human LH on gonadal function. J Endocrinol 120:89-96
De Kretser OM, Kerr JB (1988) The cytology of the testis. In: Knobi! E, Neill JD (eds) The physiology of reproduction. Raven, New York, pp 837-932
Dorrington JH, Armstrong DT (1979) Effects of FSH on gonadal functions. Rec Prog Horm Res 35:301-342
Edwards G, Lendon R, Morris ID (1989) Leydig cell recovery following a second challenge with ethane-l,2-dimethanesulphonate is enhanced by short
Proliferation and Differentiation 59
intervals between cytotoxin administration to rats. J Endocrinol 123: 197-2m
Fauser BCJM, Baird A, Hsueh AJW (1988) Fibroblast growth factor inhibits luteinizing hormone-stimulated androgen production by cultured rat testicular cells. Endocrinology 123:2935-2941
Gaytan F, Aceitero J, Lucena C, Aguilar E, Pinilla L, Garnelo P, Bellido C (1992) Simultaneous proliferation and differentiation of mast cells and Leydig cells in the rat testis. Are common regulatory factors involved? J Androl 13:387-397
Gondos B (1977) Testicular development. In: Johnson AD, Gomes WR (eds) The testis. Advances in physiology, biochemistry and function 4. Academic, New York, pp 1-37
Greep RO, Fevold HL, Hisaw FL (1936) Effect of two hypophyseal gonadotropic hormones on the reproductive system of the male rat. Anat Rec 65:261-271
Hales DB (1992) Interleukin-l inhibits Leydig cell steroidogenesis primarily by decreasing 17a-hydroxylase/C 17-20 lyase cytochrome P450 expression. Endocrinology 13 I :2 I 65-2172
Handelsman OJ, Spaliviero JA, Scott CD, Baxter RC (1985) Identification of insulin-like growth factor-I and its receptors in the rat testis. Acta Endocrinol (Copenh) 109:543-549
Haour F, Saez JM (1977) hCG-dependent regulation of gonadotropin receptor sites: negative control in testicular Leydig cells. Mol Cell Endocrinol 7: 17-24
Hardy MP, Zirkin BR, Ewing LL (1989) Kinetic studies on the development of the adult population of Leydig cells in testis of the pubertal ral. Endocrinology 124:762-770
Hardy MP, Kelce WR, Klinefelter GR, Ewing LL (1990) Differentiation of Leydig cell precursors in vitro: a role for androgen. Endocrinology 127:488-490
Hardy MP, Gelber SJ, Zhou Z, Penning TM, Riciliano JW, Ganjam VK, Nonneman 0, Ewing LL (1991) Hormonal control of Leydig cell differentiation. In: Robaire B (ed) The male germ cell: spermatogonium to fertilization. Ann N Y Acad Sci 637:152-163
Holmes SO, Spotts G, Smith RG (1986) Rat Sertoli cells secrete a growth factor that blocks epidermal growth factor binding to its receptor. J BioI Chern 261 :4076-4080
Hsueh AJW, Durau M, Call KJ (1977) Gonadotropin-induced regulation of luteinizing hormone receptors and desensitization of testicular 3':5'-cyclic AMP and testosterone responses. Proc Natl Acad Sci USA 74:592-595
60 K.J. Teerds et al.
Jackson AE, O'Leary PC, Ayers MM, de Kretser DM (1986) The effects of ethylene dimethane sulphonate (EDS) on rat Leydig cells: evidence to support a connective tissue origin of Leydig cells. BioI Reprod 35:425-437
Kelce WR, Zirkin BR, Ewing LL (1991) Immature rat Leydig cells are intrinsically less sensitive than adult Leydig cells to ethane dimethanesulfonate. Toxicol Appl Pharmacollll:189-200
Kerr 18, Donachie K (1986) Regeneration of Leydig cells in unilateral cryptorchid rats: evidence for stimulation by local testicular factors. Cell Tissue Res 245:649-655
Kerr 18, Knell CM (1987) The regenerative capacity of testicular interstitial tissue. A recurrent pattern of experimental destruction and regeneration of rat Leydig cells. Rouxs Arch Dev Bioi 196:467-471
Kerr 18, Knell CM (1988) The fate of fetal Leydig cells during the development of the fetal and postnatal rat testis. Development 103:535-544
Kerr 18, Donachie K, Rommerts FFG (1985) Selective destruction and regeneration of rat Leydig cells in vivo. Cell Tissue Res 242: 145-156
Kerr 18, Bartlett JMS, Donachie K (1986) Acute response of testicular interstitial tissue to the cytotoxic drug ethane dimethyl sulphonate. An ultrastructural and hormonal assay study. Cell Tissue Res 243:405-411
Ketelslegers JM, Hetzel WD, Sherins RJ, Catt KJ (1978) Developmental changes in testicular gonadotropin receptors: plasma gonadotropins and plasma testosterone in the rat. Endocrinology 103: 1910-1912
Khan SA, Soder 0, Syed Y, Gustafsson K, Lindh M, Ritzen EM (1987) The rat testis produces large amounts of an interleukin-l-Iike factor. lnt J AndrollO:495-503
Khan SA, Schmidt K, Hallin P, Di Pauli R, De Geyter C, Nieschlag E (1988) Human testis cytosol and ovarian follicular fluid contain high amounts of interleukin-l-Iike factor(s). Mol Cell EndocrinoI58:221-230
Khan S, Teerds K, Dorrington J (1992a) Growth factor requirements for DNA synthesis by Leydig cells from the immature rat. BioI Reprod 46:335-341
Khan SA, Khan SJ, Dorrington JH (1992b) Interleukin-I stimulates deoxyribonucleic acid synthesis in immature rat Leydig cells in vitro. Endocrinology 131: 1853-1857
Khan SA, Teerds K, Dorrington J (1994) Regulation of DNA synthesis in Leydig cells. Proceedings of the XIIth North American testis workshop (in press)
Klinefelter GR, Hall PF, Ewing LL (1987) Effect of luteinizing hormone deprivation in situ on steroidogenesis of rat Leydig cells purified by a multistep procedure. BioI Reprod 36:769-783
Proliferation and Differentiation 61
Kuopio T, Tapanainen J, Pclliniemi LJ, Huhtaniemi I (1989) Developmental stages of fetal-type Lcydig cells in prepubertal rats. Development 107:213-220
LaPolt PS, Jia XC, Sincich C, Hsueh AJW (1991) Ligand-induced down-regulation of testicular and ovarian luteinizing hormone (LH) receptors is preceded by tissue-specific inhibition of alternatively processed LH receptor transcripts. Mol Endocrinol 5:397-403
Lin T, Haskall J, Vinson N, Terracio L (1986) Characterization of insulin and insulin-like growth factor-I receptors of purified Leydig cells and their role on steroidogenesis in primary culture: a comparative study. Endocrinology 119:1641-1647
Lin T, Blaisdell J, Haskell JF (1987) Transforming growth factor-~ inhibits Leydig cell steroidogenesis in primary culture. Biochem Biophys Res Commun 146:387-394
Lin T, Wang 0, Calkins JH, Guo H, Chi R, Hausley PR (1990) Regulation of insulin-like growth factor-I messenger ribonucleic acid expression in Leydig cells. Mol Cell Endocrinol 73: 147-152
Lin T, Wang 0, Nagpal ML (1993) Human chorionic gonadotropin induces interleukin-l gene expression in rat Leydig cells in vivo. Mol Cell EndocrinoI95:139-145
Mauduit C, Chauvin MA, Hartmann OJ, Revol A, Morera M, Benahmed M (1992) Interleukin-I a as a potent inhibitor of gonadotropin action in porcine Leydig cells: site(s) of action. Bioi Reprod 46: 1119-1126
McFarland KC, Sprengel R, Phillips HS, Kohler M, Rosemblit N, Nikolics K, Segal off DL, Seeburg PH (1989) Lutropin-choriogonadotropin receptor: an unusual member of thc G protcin-coupled receptor family. Science 245:494-499
Mendis-Handagama SMLC, Risbridger GP, de Kretser OM (1987) Morphometric analysis of the components of the neonatal and the adult testis intcrstitium. Int J Androl 10:525-534
Mogcr WH (1977) Serum 5a-androstane-3a, 17~-diol androsterone and testosterone concentrations in the male rat. Influence of age and gonadotropin stimulation. Endocrinology 100: I 027-1 032
Molenaar R, de Rooij DG, Rommerts FFG, Reuvers PJ, van der Molen HJ (1985) Specific destruction of Leydig cells in mature rats after in vivo administration of ethane dimethyl sulfonate. BioI Reprod 33: 1213-1222
Molenaar R, de Rooij DG, Rommerts FFG, van der Molen HJ (1986) Repopulation of Leydig cells in mature rats after selective destruction of the existent Leydig cell population with ethylene dimethane sulfonate (EDS) is dependent on LH and not FSH. Endocrinology I 18:2546-2554
62 K.J. Teerds et al.
Moore A, Morris ID (1993) The involvement of insulin-like growth factor-I in local control of steroidogenesis and DNA synthesis of Leydig and nonLeydig cells in rat testicular interstitium. J Endocrinol 138: 107-114
Morera AM, Cochet C, Keramidas K, Chauvin MA, de Peretti E, Benahmed M (1988) Direct regulating effects of transforming growth factor ~ on the Leydig cell steroidogenesis in primary culture. J Steroid Biochem 30:443-447
Morris ID (1985) Leydig cell resistance to the cytotoxic effect of ethylene 7 dimethanesulphonate in the adult rat testis. J Endocrinol 105:311-316
Morris ID Phillips DM, Bardin CW (1986) Ethylene dimethane sulphonate destroys Leydig cells in the rat testis. Endocrinology 118:709-719
Murono EP, Washburn AL (1990) Fibroblast growth factor inhibits 5a-reductase activity in cultured immature Leydig cells. Mol Cell Endocrinol 68:RI9-R23
Myers RB, Abney TO (1990) Testosterone and androstanediol production by regenerating Leydig cells in the ethylene dimethane sulphonate-treated maturerat. IntI AndroI13:4-16
Myers RB, Abney TO (1991) Interstitial cell proliferation in the testis of the ethylene dimethane sulfonate-treated rat. Steroids 56:91-96
Neumann A, Haider SG, Hilscher B (1993) Temporal coincidence of the appearance of elongated spermatids and histochemical reaction of II P-hydroxysteroid dehydrogenase in rat Leydig cells. Andrologia 25:263-269
Nozu K, Matsuura S, Catt KJ, Dufau ML (1981) Modulation of Leydig cell androgen biosynthesis and cytochrome P-450 levels during estrogen treatment and human chorionic gonadotropin-induced desensitization. 1 Bioi Chem 256: 10012-100 I 7
Odell WD, Swerdloff RS (1976) Etiologies of sexual maturation: a model system based on the sexually maturing rat. Rec Prog Horm Res 32:245-288
Ojeda SR, Urbanski HF (1988) Puberty in the rat. In: Knobil E, Neill JD (eds) The physiology of reproduction. Raven, New York, pp 1699-1737
O'Leary PO, Jackson AE, Averill S, de Kretser DM (1986) The effects of ethane dimethyl sulphonate (EDS) on bilaterally cryptorchid rat testes. Mol Cell Endocrinol 45: 183-190
Orth 1M (1980) Proliferation of Sertoli cells in fetal and postnatal rats: a quantitative autoradiographic study. Anat Rec 203:485-492
Orth J, Christensen AK (1977) Localization of 12SI_labelied FSH in the testis of hypophysectomized rats by autoradiography at the light and electron microscope levels. Endocrinology 10 I :262-278
O'Shaughnessy PJ, Murphy L (1991) Steroidogenic enzyme activity in the rat testis following Leydig cell destruction by ethylene-I,2-dimethancsul-
Proliferation and Differentiation 63
phonate and during subsequent Leydig eell regeneration. J Endocrinol 131:451-457
Pakarinen P. Vihko KK. Voutilainen R. Huhtaniemi I (1990) Differential rcsponsc of luteinizing hormone receptor and steroidogenic enzyme genc expression to human chorionic gonadotropin stimulation in the neonatal and adult rat testis. Endocrinology 127:2469-2474
Pledger WJ (1985) Regulation of cell proliferation: serum growth factors control an ordered scrics of GI events. In: Vcneziale CM (cd) Control of cell growth and proliferation. Von Nostrand Rcinhold, New York, pp 108-13 I
Podesta EJ, Rivarola MA (1974) Concentration of androgens in whole testes. seminiferous tubulcs and interstitial tissue of rats at different stages of development. Endocrinology 95:455-461
Rommerts FFG, van der Molen HJ (1989) Testicular steroidogenesis. In: Burger H, de Kretser D (eds) The testis. Raven, New York, pp 303-328
Rommerts FFG, Teerds KJ, Hoogerbrugge JW (1988) In vitro effects of ethylenedimethane sulphonate (EDS) on Leydig cells: inhibition of steroid production and cytotoxic effects arc dcpendent on species and age of rat. Mol Cell Endocrinol 55:87-94
Rosness PA. Sunde A, Eik-Ness KB (1977) Production and effects of 7a-hydroxytestosterone on testosterone and dihydrotestosterone metabolism in rat testis. Biochim Biophys Acta 488:55-68
Shan LX. Hardy MP (1992) Developmental changes in levels of luteinizing hormone receptor and androgen receptor in rat Leydig cells. Endocrinology 131:1107-1114
Sharpe RM (1976) HCG-induced decrease in availability of rat testis receptors. Nature 264:644-646
Sharpe RM (1987) Testosterone and spermatogenesis. J Endocrinol 113: 1-3 Skinner MK, Moses HL (1989) Transforming growth factor-~ gene expression
and action in the seminiferous tubule: peritubular cell-Sertoli cell interaction. Mol Endocrinol 3:625-634
Skinner MK, Takacs K, Coffey RJ (1989) Transforming growth factor-a expression and action in the seminiferous tubule: peritubular cell Sertoli cell interactions. Endocrinology 124:845-854
Sordoillet C, Chauvin MA, Revol A, Morera AM, Benahmed M (1988) Fibroblast growth factor is a regulator of testosterone secretion in cultured immature Leydig cells. Mol Cell Endocrinol 58:283-286
Spiteri-Grech J, Nieschlag E (1993) Paracrine factors relevant to the regulation of spermatogenesis - a review. J Reprod Fertil 98: 1-14
Takao T. Mitchell WM, Tracey DE, deSouza EB (1990) Identification of intcrleukin-I receptors in the mouse testis. Endocrinology 127:251-258
64 K.J. Teerds et al.
Tapanaien J, Kuopio T, Pelliniemi LJ, Huhtaniemi I (1984) Rat testicular endogenous steroids and number of Leydig cells between the fetal period and sexual maturity. BioI Reprod 31: 1027-1035
Teerds KJ, Dorrington JH (1993) Localization of transforming growth factor ~1 and ~2 during testicular development in the rat. BioI Reprod 48:40--45
Teerds KJ, de Rooij DG, Rommerts FFG, Wen sing CJG (1988) The regulation of the proliferation and differentiation of rat Leydig cell precursor cells after EDS administration or daily hCG treatment. J Androl 9:343-351
Teerds KJ, Closset J, Rommerts FFG, de Rooij DG, Stocco DM, Colenbrander B, Wensing CJG, Hennen G (1989a) Effects of pure FSH and LH preparations on the number and function of Leydig cells in immature hypophysectomized rats. J EndocrinoI120:97-106
Teerds KJ, de Rooij DG, Rommerts FFG, van den Hurk R, Wensing CJG (1989b) Stimulation of the proliferation and differentiation of Leydig cell precursors after the destruction of existing Leydig cells with ethane dimethyl sulphonate (EDS) can take place in the absence of LH. J Androl 10:472--477
Teerds KJ, de Rooij DG, Rommerts FFG, van der Tweel, Wensing CJG (1989c) Turnover time of Leydig cells and other interstitial cells in testes of adult rats. Arch Androl 23: 1 05-111
Teerds KJ, de Rooij DG, Rommerts FFG, Wen sing CJG (1990a) Development of a new Leydig cell population after destruction of existing Leydig cells by ethane dimethane sulphonate in rats: an autoradiographic study. J Endocrinol 126:229-236
Teerds KJ, Rommerts FFG, Dorrington JH (1990b) Immunohistochemical detection of transforming growth factor-a in Leydig cells during development of the rat testis. Mol Cell Endocrinol 69:R I-R6
Teerds KJ, de Rooij DG, Wensing CJG, Rommerts FFG (1992) Hormone-induced resistance of rat Leydig cells to the cytotoxic effects of ethane-I,2-dimethane sulphonate. J Endocrinol 134:85-90
van der Molen HJ, Grootegoed JA, Greef-Bijleveld MJ de, Rommerts FFG, van der Vusse GJ (1975) Distribution of steroids, steroid production and steroid metabolizing enzymes in rat testis. In: French FS, Hansson V, Ritzen EM, Nayfeh SN (eds) Hormonal regulation of spermatogenesis. Plenum, New York, pp 665-676
Vergouwen RPFA, Jacobs SGPM, Huiskamp R, Davids JAG, de Rooij DG (1991) Proliferative activity of gonocytes, Sertoli cells and interstitial cells during testicular development in mice. J Reprod Ferti! 93:233-234
Verhoeven G, Cailleau J (1986) Stimulatory effect of epidermal growth factor on steroidogenesis in Leydig cells. Mol Cell Endocrinol 47:99-106
Proliferation and Differentiation 65
Verhoeven G, Cailleau J, Van Damme J, Billiau A (1988) Interleukin-I stimulates steroidogenesis in cultured rat Leydig cells. Mol Cell Endoerinol 57:51-60
Vreeburg JTM, Qoms MP, Rommerts FFG, Teerds KJ (1988) Functional properties of developing rat Leydig cells after treatment with ethylene dimethanesulphonate (EDS). J Reprod Ferti! 84:63--69
Wiebe JP (1976) Steroidogenesis in rat Leydig cells: changes in activity of 5-ane and 5-ene 3~ hydroxysteroid dehydrogenases during sexual maturation. Endocrinology 98:505-513
4 Regulation of the Acute Production of Steroids in Steroidogenic Cells
D.M. Stoeeo and B.J. Clark
4.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 67 4.2 Proposed Mechanisms Involved in the Acute Regulation
of Steroidogenesis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 70 4.2.1 Cellular Architecture. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 70 4.2.2 Sterol Carrier Protein 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 71 4.2.3 Steroidogenesis Activator Peptide. . . . . . . . . . . . . . . . . . . . . . . . . .. 72 4.2.4 Peripheral Benzodiazepine Receptor and 8.2-kDa Protein. . . . . . .. 73 4.2.5 The 30-kDa Mitochondrial Proteins. . . . . . . . . . . . . . . . . . . . . . . .. 76 4.3 Role of Transcription. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 77 4.4 Role of Phosphorylation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 79 4.5 Correlations Between Steroidogenesis and the 30-kDa Proteins. .. 81 4.6 Molecular Characterization of the 30-kDa Proteins. . . . . . . . . . . .. 83 4.6.1 Purification.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 83 4.6.2 Amino Acid Sequence Determination. . . . . . . . . . . . . . . . . . . . . . .. 85 4.6.3 Cloning and Sequencing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 86 4.7 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 87 References .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 87
4.1 Introduction
The mechanism regulating the production of steroids in response to trophic hormone stimulation has been the subject of investigation for over three decades. When considering the effects of trophic hormones
68 D.M. Steeee and B.J. Clark
on the steroidogenic process it is necessary to first distinguish between acute effects and chronic effects. Acute effects are those which result in the very rapid (within minutes) synthesis and secretion of steroids in response to hormone stimulation and involve the rapid translocation of intracellular cholesterol to the site of its cleavage, as will be discussed later. Chronic effects are those which occur on the order of hours to tens of hours and involve increased gene transcription and translation of the proteins involved in the biosynthesis of steroids. This chapter will focus on studies designed to elucidate the mechanisms involved in the acute regulation of steroid production in response to hormone stimulation. Overviews of the effects of chronic stimulation on steroidogenic enzymes have appeared in several excellent review articles (Simpson and Waterman 1983; Miller 1988; Hanukoglu 1992).
Cholesterol transport within steroidogenic cells can be thought of as occurring in two separate processes. The first part of the process is the mobilization of cholesterol from cellular stores to the outer mitochondrial membrane while the second part consists of the transfer of cholesterol from the outer to the inner mitochondrial membrane (Liscum and Dahl 1992). It has also been reported that in MA-lO mouse Leydig tumor cells the immediate source of cholesterol for transfer to the inner mitochondrial membrane is the plasma membrane (Freeman 1989). Although the rate-limiting enzymatic step in steroidogenesis is the conversion of the substrate cholesterol to pregnenolone by the cholesterol side-chain cleavage enzyme (CSCC; Stone and Hechter 1954; Karaboyas and Koritz 1965; Garren et al. 1971), the true rate-limiting step in this process is the delivery of the substrate to the CSCC, which is located in the inner mitochondrial membrane of steroidogenic cells (Crivello and Jefcoate 1980; Jefcoate et al. 1987; Privalle et al. 1983a,b; Lambeth et al. 1987). Thus, it is the second part of this two-part process which is considered rate limiting. The CSCC complex is comprised of three proteins, an iron sulfur protein, adrenodoxin; a flavoprotein, adrenodoxin reductase; and lastly, cytochrome P450 side-chain cleavage (P450scc) which cleaves cholesterol to pregnenolone (Lambeth et al. 1979; Hanukoglu and Hanukoglu 1986; Tuls et al. 1987). The major barrier to be overcome in the translocation of cholesterol to the P450scc is the aqueous space between the outer and inner mitochondrial membranes through which this relatively hydrophobic compound must pass. Since the aqueous diffusion of cholesterol is extremely slow (Phillips et
Regulation of the Acute Production of Steroids 69
a!. 1987; Rennert et a!. 1993; Schroeder et a!. 1991) and could not provide sufficient substrate to account for the rapid and large increase in steroid production observed in steroidogenic cells, then, the successful stimulation of steroidogenesis would appear to require the presence of a mechanism which rapidly transports cholesterol across this barrier. Thus, in simple terms, the regulation of steroidogenesis is controlled by a factor(s) which facilitates the translocation of cholesterol from cellular stores across the aqueous intermembrane space of the mitochondria to the inner membrane.
Early studies attempting to elucidate the mechanism responsible for the acute regulation of steroidogenesis demonstrated that steroid production in response to hormone stimulation had an absolute requirement for the synthesis of new proteins. The first of such studies demonstrated that stimulation of corticoid synthesis in adrenal glands by adrenocorticotropic hormone (ACTH) was sensitive to the protein synthesis inhibitor puromycin (Ferguson 1963). Subsequent studies (Garren et al. 1965) indicated that adrenocorticoid production in response to ACTH could be totally blocked by inhibition of translation but not transcription. Later, Simpson et a!. (1979) determined that the cycloheximide-sensitive step in this process was located in the mitochondria, but, importantly, it was also noted that protein synthesis inhibitors had no effect on the activity of the CSCC complex itself (Arthur and Boyd 1976). These earlier observations have been elegantly confirmed recently in both mouse Y-I adrenal tumor cells and MA-IO mouse Leydig tumor cells which demonstrated that while cholesterol could be delivered to a "pre-steroidogenic pool" in the presence of cycloheximide, pregnenolone production did not take place until the inhibitor was removed and the cells subsequently stimulated with hormone (Stevens et a!. 1993). As a result of these and other observations, the model arose that the acute production of steroids was dependent on a rapidly synthesized, cycloheximide-sensitive and highly labile protein which appeared in response to trophic hormone treatment. A graphic illustration of the components of the system which depicts the problem involved in the transport of cholesterol is shown in Fig. I. Several of the proposed mechanisms by which the acute transfer of cholesterol to the mitochondria and eventually to its inner membrane are controlled are summarized below.
70
OUTER
CSCC Complex
1 adrenodoxin 2 adx red 3 P450scc
cholesterol
D.M. Stoeeo and B.J. Clark
INNER MEMBRANE CRISTAE
pregneno lone
Fig. 1. The rate-limiting step in acute steroidogenesis. This figure illustrates the basic problem involved in the acute regulation of steroidogenesis. Some. though not all, of the components of the system are shown in the illustration and include the mitochondria with its various compartments and the cholesterol side-chain cleavage (CSCC) complex which resides on the inner mitochondrial membrane. The main problem to be overcome in the hormone-regulated production of steroids is the transfer of the substrate cholesterol to the inner mitochondrial membrane. The difficulty lies in the fact that the intermembrane space is aqueous and cholesterol diffuses through water very slowly. Thus, the factor(s) responsible for the acute synthesis of steroids must somehow mediate the rapid transfer of cholesterol to the inner membrane and the CSCC. adx red, adrenodoxin reductase; P450scc, cytochrome P450 side-chain cleavage
4.2 Proposed Mechanisms Involved in the Acute Regulation of Steroidogenesis
4.2.1 Cellular Architecture
A number of studies have indicated that cellular architecture plays an important role in the delivery of cholesterol to the mitochondria for subsequent steroidogenesis. The transfer of cholesterol to the outer mitochondrial membrane in response to trophic hormone stimulation was found to occur in the absence of de novo protein synthesis (Crivello and lefcoate 1980; Privalle et al. 1983a,b) but was inhibited by com-
Regulation of the Acute Production of Steroids 71
pounds which disrupted microtubules and microfilaments (Mrotek and Hall 1977; Hall et a!. 1979; Crivello and Jefcoate 1978,1980; Hall 1984a,b; Nagy and Freeman 1990a,b). Thus, the cytoskeleton is believed to have an important function in steroidogenesis (Hall I 984a,b, 1985, 1988; Hall and Almahbobi 1992). It has also been shown that trophic hormone stimulation of human granulosa cells results in a morphological rounding of the cells which causes a clustering of steroidogenic organelles potentially bringing mitochondria in closer contact to pools of cholesterol (Soto et a1. 1986). More recently, several studies have demonstrated the importance of the 10-nm intermediate filaments in the movement of cholesterol within Y -I mouse adrenal tumor cells and primary cultures of bovine adrenal and rat Leydig cells (Almahbobi and Hall 1990; Almahbobi et al. I 992a,b, 1993). These studies demonstrated that both lipid droplets and mitochondria are colocalized on intermediate filaments and hypothesized that this colocalization was the means by which cholesterol was delivered to the mitochondria from the lipid droplet. Thus, it appears that subcellular structures such as microtubules, microfilaments, and intermediate filaments are instrumental in the delivery of cholesterol to the mitochondria. However, there is no convincing evidence that any of these structures plays a role in the transfer of cholesterol from cellular stores or the outer mitochondrial membrane to the inner mitochondrial membrane where cleavage takes place.
4.2.2 Sterol Carrier Protein 2
Sterol carrier protein 2 (SCP2) is a 13-kDa protein which has been found in high abundance in liver as well as in various steroidogenic tissues. Also known as nonspecific lipid transfer protein, the role of this protein in liver and adrenal cells has been reviewed by Vahouny et al. (1985) and Vahouny et al. (1987). SCP2 has been demonstrated to transfer cholesterol from lipid droplets to mitochondria in a I: I ratio (Chanderbhan et al. 1982). It has also been shown to be capable of stimulating steroid production in isolated adrenal mitochondria (Chanderbhan et al. 1982; Vahouny et al. 1983). SCP2 arises in the cell from a larger precursor, and both the precursor and the mature forms reside primarily in peroxisomes (Keller et al. 1989; van Amerongen et al. 1989a; Fujiki
72 D.M. Stoeeo and B.J. Clark
1989a; Fujiki et al. 1989; Mendis-Handagama et al. 1990). While the synthesis of SCP2 has been shown to be under the regulation of ACTH (Trzeciak et al. 1987), this regulation only occurred after many hours and in fact, acute stimulation of adrenal cells with ACTH had no effect on SCP21evels (van Amerongen et al. J989b). However, in rat testicular Leydig cells, acute stimulation with luteinizing hormone (LH) appeared to result in a redistribution of SCP2 within the cell (Van Noort et al. 1988). Another indication of the role of SCP2 in steroidogenesis can be found in the observation that treatment of adrenal cells with anti-SCP2 antibody resulted in an inhibition of steroid production (Chanderbhan et al. 1986). Also, cotransfection of COS (green monkey kidney cells) cells with CSCC and SCP2 resulted in a threefold increase in steroid production over cells transfected with CSCC only (Yamamoto et al. 1991). However, evidence that SCP2 is able to effectively transfer cholesterol to the inner mitochondrial membrane and the P450scc in amounts adequate to support steroidogenesis in response to hormone stimulation is not apparent. This, coupled to the observation that SCP2 levels do not rapidly change in response to either acute stimulation or to treatment with cycloheximide, makes it unlikely that this protein has a major role in the acute regulation of steroid production and rather functions in maintaining sterol movement within the cell in support of steroidogenesis. Thus, in steroidogenic cells, although a significant number of studies have been performed, the role of this protein has yet to be completely elucidated.
4.2.3 Steroidogenesis Activator Peptide
Another peptide thought to playa role in the acute regulation of steroidogenesis is steroidogenesis activator peptide (SAP). Originally purified as a 2.2-kDa peptide from rat adrenal cells (Pedersen and Brownie 1983; Pedersen 1984), SAP was later determined to be a 30 amino acid (3.2-kDa) peptide when it was purified from rat Leydig tumor cells (Pedersen and Brownie 1987). This peptide was found to be present only in steroidogenic cells; its levels could be acutely increased by trophic hormone stimulation and this increase was prevented by cycloheximide (Pedersen 1987; Mertz and Pedersen 1989a; Frustaci et al. 1989). Addition of SAP to isolated mitochondria was able to increase
Regulation of the Acute Production of Steroids 73
indicating that SAP may playa role in cholesterol transfer within this organelle (Pedersen and Brownie 1983,1987). SAP was found to be nearly completely identical to the carboxy tenninal of a minor heat shock protein known as glucose-regulated protein 78 (Mertz and Pedersen 1989b; Li et al. 1989). While a mechanism whereby SAP can transport cholesterol to the inner mitochondrial membrane has not yet been satisfactorily demonstrated, this small peptide must still be considered when discussing intramitochondrial cholesterol transport.
4.2.4 Peripheral Benzodiazepine Receptor and 8.2-kDa Protein
Recently, a great deal of attention has been given to the possible role of the peripheral benzodiazepine receptor (PBR; also called the mitochondrial benzodiazepine receptor, MBR) in the acute regulation of steroidogenesis. PBR are 18-kOa proteins present in most cell types and originally found to be present in high concentrations in the outer mitochondrial membrane (Anholt et al. 1986; Calvo et al. 1990). Excellent reviews of both the characteristics of the PBR and their potential role in steroidogenesis have recently appeared (Venna and Snyder 1989; Krueger et al. 1991; Krueger and Papadopoulos 1992; Papadopoulos 1993). PBR were found to be present in especially high concentrations in the outer mitochondrial membranes of steroidogenic tissues (Verma and Snyder 1989; Amsterdam and Suh 1991; Mukhin et al. 1989; Papadopoulos et al. 1990). Treatment of either MA-10 mouse Leydig tumor cells or Y -I mouse adrenal tumor cells with PBR agonists resulted in significant increases in steroid production (Mukhin et al. 1989; Papadopoulos et al. 1990; Krueger and Papadopoulos 1990; Papadopoulos et al. 1991a; Papadopoulos 1993; Gamier et al. 1993). Stimulation of steroid production in response to PBR ligands was also observed in bovine primary adrenal cell cultures (Yanagibashi et al. 1989), human placental tissue (Bamea et al. 1989), and ovarian granulosa cells (Amsterdam and Suh 1991). It was also demonstrated that the endogenous ligand for the PBR was an approximately 10-kOa peptide known as diazepam-binding inhibitor (OBI; Costa and Guidotti 1991). Subsequent studies showed that both OBI and OBI processing products were able to stimulate steroid production in rat adrenal primary cultures, Y-I adrenal, and MA-IO Leydig cells (Cavallaro et al. 1992; Papado-
74 D.M. Stoeeo and B.J. Clark
poulos et al. 1991b,c). Furthennore, addition of the PBR ligand endozepine to a CSCC reconstituted enzyme system was able to stimulate the conversion of cholesterol to pregnenolone (Brown and Hall 1991). Also, a DBI antagonist was shown to inhibit the trophic honnone stimulated steroid production in both Y-l adrenal and MA-lO Leydig cells (Papadopoulos et al. 1991a). Further studies indicated that DBI as well as another PBR ligand, 4' -chlorodiazepam, were both able to stimulate pregnenolone production in the C6-2B glioma cell line (Papadopoulos et al. 1992; Guarneri et al. 1992). A particularly striking observation which illustrates the ability of DBIJPBR to provide cholesterol for steroidogenesis was the demonstration that removal of DBI from MA-10 Leydig cells using cholesterol-linked phosphorothioate antisense oligonucleotides to DBI resulted in a loss of trophic honnone stimulated steroid production in these cells (Boujrad et al. 1993). Strengthening the role of DBIJPBR in steroidogenic regulation were the studies of Yanagibashi et al. (1988) in which they demonstrated that an 8.2-kDa peptide purified from bovine adrenal cells was able to stimulate pregnenolone production in isolated mitochondria. The significance of this observation became apparent when the peptide was identified as des-(Gly-Ile)-endozepine (Besman et al. 1989). Thus, the 8.2-kDa peptide and DBI were virtually identical.
Snyder and colleagues (McEnery et al. 1992) have recently demonstrated that the PBR can be isolated from mitochondria as a complex of three proteins consisting of the 18-kDa PBR, the 32-kDa voltage-dependent anion carrier (VDAC) and the 30-kDa adenine nucleotide carrier (ANC). They also illustrated that ligands to the PBR were able to link this receptor to inner mitochondrial membrane ion channels at nanomolar concentrations (Kinnally et al. 1993). The significance of these findings is that the VDAC is associated with areas of the mitochondrial membrane known as "contact sites," where the inner and outer membranes come in close apposition to each other (Ohlendieck et al. 1986; Sandri et al. 1988; Moran et al. 1990). The role of contact sites in the intramitochondrial transfer of cholesterol has recently been reviewed (Jefcoate et al. 1992) and it appears that the fonnation of such contacts may be a necessary component for sterol transfer. Thus, PBRligand interactions may enhance the fonnation of contact sites through its association with the VDAC, thus allowing cholesterol to transfer to the inner membrane.
Regulation of the Acute Production of Steroids 75
However, several inconsistencies arise in studies dealing with the role of OBI/PBR in the acute regulation of steroidogenesis in steroidogenic cells. It was first reported (Hall et al. 19X8) that the acute synthesis of des-(Gly-Ile)-OBI was rapidly increased by ACTH in adrenocortical cells and that the half-life of this protein was very short in the presence of cycloheximide, both characteristics for a potential steroidogenic regulatory protein as described by Ferguson (1963), Garren et al. (1965), and Stevens et al. (1993). However, more recent studies have shown that OBI is not acutely regulated by trophic hormone treatment of either adrenocortical or Leydig cells and that the half-life of this protein is greater than 3 h (Brown et al. 1992). This makes it unlikely that hormonal regulation of OBI could be responsible for the acute production of steroids. In another study it was demonstrated that rat adrenal OBI and PBR levels were reduced dramatically following 9 days of hypophysectomy. ACTH administration to these hypophysectomized rats resulted in an increase in steroidogenesis which peaked within I h while both PBR and OBI mRNA and protein levels showed no increase for approximately 12 h, leading the authors to conclude that steroidogenesis and PBR/OBllevels were not temporally related to the acute steroidogenesis induced by ACTH and may reflect its long-term trophic action on adrenocortical cells (Cavallaro et a1. 1993). It was also argued that the outer mitochondrial location of the PBR provided strong evidence for its specific role in the regulation of cholesterol transfer and steroidogenesis (Papadopoulos et al. 1990). However, PBR, while heavily concentrated on the outer mitochondrial membrane, have recently been shown to be present on the cell surface in adrenal cortex cells as well (Oke et al. 1992). Further, it has been reported that OBI is identical to acyl-CoA ester-binding protein and as such it has been argued that this protein plays no role as a ligand in signal transduction pathways (Knudsen and Nielsen 1990). In addition, Snyder et al. (1987) caution that the affinity of the PBR for some porphyrins is 1000 times greater than its affinity for endozepine and thus the endogenous ligands may, in fact, be porphyrins and that the endozepinelike activity found in extracts may be due to the presence of small amounts of porphyrin.
Lastly, while noted above that most investigations using ligands of the PBR have shown that such treatment is stimulatory, a number of investigations have demonstrated that treatment of steroidogenic tissues with PBR ligands can also be inhibitory to steroidogenesis (Shibata et
76 D.M. Stocco and B.J. Clark
al. 1986; Holloway et al. 1989; Thomson et al. 1992; Python et al. 1993).
4.2.5 The 30-kDa Mitochondrial Proteins
Another group of proteins which have been speculated to be involved in the acute regulation of steroid production in steroidogenic tissues are a family of mitochondrial proteins and phosphoproteins which have been described and characterized during the past decade. These proteins were perhaps first observed as being synthesized in response to ACTH stimulation of Y -1 adrenal tumor cells (Nakamura et al. 1978). Subsequent studies from Onne-Johnsons' laboratory using radiolabeling and twodimensional polyacrylamide gel electrophoresis (2D PAGE) demonstrated that these proteins consisted of a family of approximately 30-kDa proteins which rapidly appeared in response to corticotropic honnone stimulation of rat adrenal cells and were cycloheximide sensitive (Krueger and Onne-Johnson 1983). Later studies indicated that similar proteins were also synthesized in response to trophic honnone treatment in the corpus luteum (Pon and Onne-Johnson 1986) as well as in primary cultures of mouse testicular Leydig cells (Pon et al. 1986a; Epstein and Onne-Johnson 199Ia). These proteins were then determined to be phosphoproteins in both adrenal and corpus luteum cells (Pon et al. 1986b; Pon and Onne-Johnson 1988). This group then used several amino acid analogs to demonstrate that steroidogenesis could be inhibited under conditions only partially inhibitory to protein synthesis, further indicating the necessity for the production of a labile protein in the acute regulation of steroidogenesis (Krueger and Onne-Johnson 1988). Subcellular fractionation illustrated that these rapidly synthesized proteins were localized in the mitochondria of honnone-stimulated adrenal cells (Alberta et al. 1989; Epstein et al. 1989). Recent observations have indicated that in ACTH-stimulated rat adrenal cortex cells the 30-kDa mitochondrial proteins arise as a result of the processing of two larger precursor fonns having molecular weights of 37 kDa and 32 kDa, respectively (Epstein and Onne-Johnson 1991b). Much of the work on the acute regulation of steroidogenesis in adrenal cells was recently reviewed by Onne-Johnson (1990). Lastly, it should be noted that in at least two additional laboratories, mitochondrial proteins similar or identical in molecular weights and isoelectric points have been described in
Regulation of the Acute Production of Steroids 77
honnone-stimulated, small luteal cells and adrenal glomerulosa cells (Mittre et al. 1990; Elliot et al. 1993).
In our laboratory the MA-1O mouse Leydig tumor cell line is employed as a model sytem in which to study the acute regulation of steroid production in response to hormone stimulation. These cells were derived from the M5480P tumor and have been shown to have functional LHjCG (chorionic gonadotrophin) receptors and produce large amounts of progesterone in response to stimulation (Ascoli 1981). Observations made in our studies indicate that MA-I0 cells also synthesize a family of 37-kDa, 32-kDa, and 30-kDa mitochondrial proteins in response to trophic honnone and cAMP analog treatment and that the appearance and quantity of these proteins are highly correlative with the temporal appearance and quantitative levels of steroids produced (Stocco and Kilgore 1988; Stocco and Chaudhary 1990; Stocco and Chen 1991; Chaudhary and Stocco 1991; Stocco and Sodeman 1991; Stocco 1992; Stocco and Ascoli 1993; Stocco and Clark 1993). The nature of these correlations will be discussed more fully in a later section of this chapter. 2D PAGE gels demonstrating the appearance of the 30-kDa mitochondrial proteins in response to honnone stimulation in MA-1O cells are shown in Fig. 2.
4.3 Role of Transcription
A number of studies have been perfonned to detennine the role of de novo transcription in the acute regulation of steroidogenesis. While one such study (Mendelson et al. 1975) indicated that new transcription was required for the acute production of steroids in rat testicular cells in response to honnone treatment, the majority of opinion indicates that acute steroidogenesis can occur independently of new RNA synthesis (Ferguson and Morita 1964; Ney et al. 1966; Schulster 1974; Nakamura et al. 1978; Cooke et al. 1979). It has been argued that the rapidity with which steroid synthesis begins in response to honnone stimulation would make it most difficult to account for both the transcription and translation of a new protein (Schulster et a1. 1974). Thus, mechanisms were proposed in which either an existing, inactive protein required for steroidogenesis was rapidly converted to an active protein by the action of trophic honnone or that a new protein was rapidly synthesized using
78
A
30-
1
30-
D.M. Stoeeo and B.J. Clark
~~ 2 3
cont rol
~ t
1 2 3
dbcAMP
4
Fig. 2 A, B. The effects of honnone stimulation on the synthesis of mitochondrial proteins. The two-dimensional polyacrylamide gel electrophoresis (2D PAGE) profiles of mitochondrial proteins shown were obtained following incubation of control and hormone stimulated MA-lO cells in the presence of [35S]methionine for a period of 6 h. Cells were harvested, mitochondria isolated, and proteins solubilized. Proteins were separated on 2D gels and subjected to fluorography. Fluorograms were developed and only those areas of the gels depicting the 30-kDa proteins were photographed and are shown here. A Control cells; B 1 mM bibutyryl cAMP treated cells. The arrows numbered 1-4 point to the different forms of the 30-kDa mitochondrial proteins synthesized in response to hormone treatment
Regulation of the Acute Production of Steroids 79
a preexisting and stable mRNA (Schulster et al. 1974). To date a satisfactory conclusion as to whether one of these two or perhaps some other mechanism is operational has not been made. It would appear to be clear, however, based on a large volume of data using protein synthesis inhibitors (Ferguson 1963; Garren et al. 1965; Schulster et al. 1974; Crivello and Jefcoate 1978; Krueger and Orme-Johnson 1988) that a newly synthesized protein is required for cholesterol transfer. Thus, any mechanism proposed for the activation of a preexisting mRNA with trophic hormone and resulting in the synthesis of a new protein would be consistent with this data. In summary, it appears that while de novo protein synthesis is required for the acute regulation of steroid production in response to hormone stimulation, de novo transcription is not.
4.4 Role of Phosphorylation
Steroidogenic cells respond to trophic hormones with the rapid production of steroids. This response is mediated through the cAMP second messenger system (Rommerts et al. 1974; Hsueh et al. 1984; Cooke et al. 1976; Marsh 1975; Sala et al. 1979), which in tum activates cAMPdependent protein kinase (protein kinase A). Activation of protein kinase A normally results in the phosphorylation of proteins on either threonine or serine residues (Edelman et al. 1987). Thus, steroidogenesis would appear to be dependent upon the phosphorylation of a protein(s) substrate on serine or threonine residues in response to hormone stimulation. Using amino acid analogs which were incapable of being phosphorylated, it has been demonstrated that in both rat adrenal cortex cells (Griffin Green and Orme-Johnson 1991) and MA-lO Leydig tumor cells (Stocco and Clark 1993), the phosphorylation of a protein on a threonine residue is required for steroidogenesis. The presence of newly synthesized phosphoproteins have been observed in stimulated adrenal cells (Pon et al. 1986b; Pon and Orme-Johnson 1988; Alberta et al. 1989; Epstein and Orme-Johnson 1991 b), in primary cultures of mouse Leydig cells (Epstein and Orme-Johnson 1991a), as well as in stimulated MA-lO cells (Chaudhary and Stocco 1991), but a cause and effect relationship between these phosphoproteins and steroid production cannot be made at this time. A diagram depicting the possible mechanisms whereby phosphorylation may playa role in steroidogenesis is shown in Fig. 3.
80
p
ex isting protein
D.M. Stoeeo and B.J. Clark
mitochondr ia
Fig. 3. Possible mechanisms for the role of phosphorylation in the regulation of steroidogenesis. This illustration depicts several points at which phosphorylation events may regulate steroid production in steroidogenic cells. The first pathway (1) consists of the rapid activation of an existing protein transcription factor by phosphorylation which results in the transcription and ultimately the translation of an mRNA for a protein which is involved in the transfer of cholesterol to the inner mitochondrial membrane. The second possibility (2) is that a phosphorylation event results in the activation of an existing mRNA, which in turn is translated into a protein which functions in cholesterol transfer. The third possible control point (3) consists of the activation of the "cholesterol transport protein" by phosphorylation after it has been synthesized de novo. Lastly, it is also possible that the protein responsible for the transfer of cholesterol to the inner mitochondrial membrane is constitutively present in the cell and is activated by phosphorylation (4). Also, as shown in this model, it is possible that two phosphorylations may be required, one for the de novo synthesis of the protein through the action of a transcription factor and one for the activation of this protein following translation. The data currently available would favor pathway number two as being the most likely mechanism involved in acute regulation. LH, luteinizing hormone; LHR, luteinizing hormone receptor; Gs, stimulatory G protein; AC, adenylate cyclase; ATP, adenosine triphosphate; cAMP, 3'-S'cyclic adenosine monophosphate; pKA, protein kinase A; P, phosphate groups. (Reprinted from Stocco and Clark 1993)
Regulation of the Acute Production of Steroids
4.5 Correlations Between Steroidogenesis and 30-kDa Proteins
81
While a cause and effect relationship between any of the putative steroidogenic regulatory factors listed above and steroidogenesis has not been unequivocally demonstrated, there are a number of strong correlations which exist between the 30-kOa mitochondrial proteins and steroid production. These relationships have been found to exist in adrenal fasciculata cells, adrenal glomerulosa cells, primary cultures of luteal cells and Leydig cells, and MA-lO tumor Leydig cells. More specifically, a list of these correlations is as follows:
1. Presence of similar (probably identical) proteins in response to hormone stimulation in primary cultures of adrenal fasciculata (Elliot et al. 1993; Krueger and Orme-Johnson 1983,1988; Pon and OrmeJohnson 1986; Pon et al. 1986b; Alberta et al. 1989; Epstein et al. 1989; Epstein and Orme-Johnson 1991b), adrenal glomerulosa (Elliot et al. 1993), luteal (Pon and Orme-Johnson 1988; Mittre et al. 1990), primary Leydig cell cultures (Pon et al. 1986a; Stocco and Kilgore 1988; Epstein and Orme-Johnson 1991a), and MA-lO Leydig tumor cells (Stocco and Kilgore 1988; Stocco and Chaudhary 1990; Stocco and Chen 1991; Chaudhary and Stocco 1991; Stocco and Sodeman 1991; Stocco 1992; Stocco and Asco1i 1993; Stocco and Clark 1993).
2. Appearance of the proteins following hormone stimulation in a time- and dose-dependent manner in rat adrenal (Krueger and Orme-Johnson 1983) and MA-lO mouse Leydig tumor cells (Stocco and Sodeman 1991).
3. Finding that the synthesis of the 30-kOa proteins is sensitive to cycloheximide in rat adrenal (Krueger and Orme-Johnson 1983), mouse Leydig cells (Epstein and Orme-Johnson 199Ia), corpus luteum cells (Pon and Orme-Johnson 1988) and MA-J 0 cells (Stocco, unpublished observations).
4. Constitutive presence of the 30-kOa mitochondrial proteins in the constitutively steroid producing R2C rat Leydig tumor cell line (Stocco and Chen 1991).
5. Finding that the metal chelator orthophenanthroline, which inhibits the mitochondrial matrix protease thus blocking the processing of
82 D.M. Stoeeo and B.J. Clark
the 37 -kDa protein to the 30-kDa proteins, results in the disappearance of the 30-kDa proteins and the total inhibition of steroidogenesis in MA-lO cells (Stocco 1992). However, in R2C cells steroidogenesis is not completely inhibited by this compound and significant amounts of the 30-kDa proteins can be found in the mitochondria (Stocco 1992).
6. Inhibition of both steroidogenesis and the appearance of these proteins in the mitochondria in MA-lO cells in response to carbonyl cyanide m-chlorophenylhydrazone (mCCCP), a compound which disrupts the electrochemical potential across the inner mitochondrial membrane and prevents translocation of mitochondrial proteins (Stocco 1992).
7. Loss of both constitutive steroid-producing capacity and the mitochondrial proteins in a "revertant" of the R2C cell line (Stocco 1992).
8. Correlation between steroid production and the mitochondrial 30-kDa proteins in two stable constructs of the MA-lO cell line, one of which produces a mutant regulatory subunit of protein kinase A and the other which constitutively overproduces a cAMP phosphodiesterase (Stocco and Ascoli 1993).
9. Correlation between steroid production and the mitochondrial 30-kDa proteins in MA-lO cells in response to treatment of stimulated MA-lO cells with threonine and serine analogs (Stocco and Clark 1993), hydrogen peroxide (Stocco et al. 1993), and dimethyl sulfoxide (Stocco et ai., unpublished observations).
While these correlations do not provide absolute proof for the role of the 30-kDa family of mitochondrial proteins in the regulation of steroidogenesis, they do allow us to propose a model wherein the synthesis and processing of these proteins may result in cholesterol transfer to the inner mitochondrial membrane. We propose that in response to hormone stimulation the 37-kDa precursor form of the protein is rapidly synthesized in the cytosol probably using a preexisting mRNA. The 37-kDa becomes immediately associated with the mitochondria by virtue of its signal sequence recognizing a specific receptor. As the transfer of this protein to the inner mitochondrial compartment begins, contact sites between the inner and outer membranes are formed and the signal sequence is removed by a matrix protease forming the 32-kDa form of
Regulation of the Acute Production of Steroids 83
the protein. Further processing of the protein removes the targeting sequence and the mature 30-kOa protein remains associated with the inner mitochondrial compartment. We propose that it is during the processing of the protein with the accompanying formation of contact sites that cholesterol is able to transfer from the outer to the inner mitochondrial membrane (Reinhart 1990) and, hence, be available to the CSCC complex. Following processing of the 37-kDa protein to the 30-kDa proteins, the membranes once again separate, and it is the continued synthesis and processing of precursor proteins which allow for the continued transport of cholesterol to the inner membrane. Since the half-life of the 37 -kDa and 32-kDa precursors of the 30-kDa mitochondrial proteins have been shown to be very short (Epstein and Orme-Johnson 1991 b), this would explain the observation that steroidogenesis decays very quickly in the absence of new protein synthesis. In support of this model, we and others have demonstrated the presence of such precursor proteins (Stocco and Sodeman 1991; Epstein and OrmeJohnson 1991 b) and have shown that the 30-kDa proteins are found associated with the inner mitochondrial compartment (Stocco and Sodeman 1991). These observations, coupled to those indicating that the transport of mitochondrial proteins across the membranes occurs at contact sites (Schwaiger et al. 1987; Vestweber and Schatz 1988; Rassow et al. 1989; Pon et al 1989; Pfanner et al. 1990; Hwang et al. 1991), make this a viable model at this point in time. A summary of this model is shown in Fig. 4.
4.6 Molecular Characterization of 30-kDa Proteins
4.6.1 Purification
In order to more fully characterize the 30-kOa mitochondrial proteins it was necessary to purify them to homogeneity and to obtain amino acid sequence information for subsequent studies. Since the 30-kDa proteins are found only in the mitochondrial fraction of hormone-stimulated cells, we set out to purify them from dbcAMP-stimulated MA- \0 cells. One of the major difficulties encountered in the purification of the 30-kOa proteins was the lack of a bioassay. However, the 30-kDa proteins are readily detectable after silver staining of mitochondrial
84
~AAAAA
• ~-ri bosome
+ =
cycle starts again
D.M. Stocco and B.J. Clark
0",,' m.m'''~." . mom"'"
? 32 kDa
cholesterol transfer and cleavage of 37 kDa protein
~
~ ~~. ? 30 kDa
refo lding of prot ein second cleavage
Fig. 4. Proposed mechanism for the acute regulation of steroidogenesis. Upon stimulation of the Leydig cell with trophic hormone, a 37-kDa protein is rapidly synthesized in the cytoplasm. This protein is accompanied by chaperone proteins which prohibit folding of the 37-kDa protein, a condition which would make transfer of the protein into the mitochondrion impossible. The 37-kDa protein is then transported to the mitochondria where it interacts with a specific receptor protein on the outer mitochondrial membrane. At this time the insertion process begins and contact sites between the outer and inner mitochondrial membranes are formed concomitant with the first cleavage event. The cleavage of the first signal precursor by the matrix protease results in the formation of a 32-kDa intermediate form of his protein. It is during this time that cholesterol is able to transfer from the outer to the inner mitochondrial membrane. Lastly, the inner and outer membrane separate and the 32-kDa protein is cleaved a second time to give the 30-kDa product which is no longer able to function in the further transfer of cholesterol. TS, targeting sequence; SS, signal sequence
proteins isolated from dbcAMP-stimulated MA-lO cells and separated by 2D sodium dodecyl sulfate (SDS)-PAGE. Therefore, the recovery of the 30-kDa proteins following each step of the purification was monitored by 2D-SDS PAGE. Quantitation of the 30-kDa protein recovery was assessed using the Visage 2000 computer-assisted image analysis sy-
Regulation of the Acute Production of Steroids 85
stem (Biolmage, Ann Arbor, MI, USA) to automatically capture and determine the integrated intensities (II) of all silver-stained spots on the gel, including the 30-kDa proteins. A brief summary of the purification of the 30-kDa proteins is as follows. MA-lO cells were stimulated with dbcAMP for 6 h, the cells harvested, and the mitochondria isolated as previously described (Stocco and Kilgore 1988; Stocco and Chen 1991). Mitoplasts (mitochondria stripped of outer membrane) were prepared by hypotonic swelling of the mitochondria, followed by homogenization and centrifugation (Ardail et al. 1990). The mitoplasts were then solubilized with 3-[ (3-cholamidopropyl) dimethylammonio) I I-propanesulfonate (CHAPS) detergent [I: 1 (wtwt) protein:detergent ratio I and the CHAPS insoluble proteins were removed by centrifugation. The CHAPS soluble proteins were separated by preparative SDS-PAGE, the 32- to 28-kDa section of the gel excised and the proteins in this section recovered by electroelution from the polyacrylamide gel. The eluted sample was concentrated and treated with alkaline phosphatase (Escherichia coli) to convert the phosphorylated 30-kDa forms 3 and 4 into the nonphosphorylated forms 1 and 2 (Stocco and Clark 1993). The proteins were then resolved by 2D-SDS PAGE, electrophoretically transferred to nitrocellulose paper, and detected by Ponceau S stain. The specific 30-kDa protein spots 1 and 2 were cut out, washed briefly with HPLCgrade H20, and stored at -20°C. The 30-kDa protein spots 1 and 2 represent 40% and 60% of the total 30-kDa proteins, respectively. This purification enriched the 30-kDa proteins 13-fold and made possible the isolation of sufficient amounts of protein by 2D-SDS PAGE for in situ tryptic digestion and amino acid sequence determination.
4.6.2 Amino Acid Sequence Determination
In situ digestion, peptide separation, and peptide sequencing was performed at the Harvard Microchemical Facility, Cambridge MA, USA. A minimum of 100 pmol protein (3 flg for a 30-kDa protein) is required for microsequence analysis. Therefore, several 2D-SDS PAGE gels were necessary to obtain a sufficient amount of protein for amino acid determination (approximately 24 flg of 30 kDa protein 2 was collected from 8, 2D-SDS PAGE gels). Even though 30-kDa proteins 1 and 2 are most likely identical (Stocco and Chen 1991), the protein spots were not
86 D.M. Stoeeo and B.J. Clark
combined and only 30-kDa protein 2 was analyzed to ensure that one homogeneous protein was used for microsequencing. Three peptides from the resulting tryptic map were selected for amino acid microsequencing. Laser desorption time of flight (TOF) mass spectrofotometry was used to ensure that each peptide to be sequenced was homogeneous (Farrow and Rakestraw 1992). The peptide sequences which were determined with high confidence are 12, 14, and 19 amino acids in length. The sequences obtained from 30-kDa 2 were compared to known protein and DNA sequences in the OENEMBL data base (OCO software package, University of Wisconsin-Madison). The data base search indicates that 30-kDa 2 represents a novel protein.
4,6.3 Cloning and Sequencing
Total RNA was isolated from MA-IO mouse Leydig tumor cells that had been stimulated with dbcAMP for 4 h. PolyA+ RNA was twice selected over a gravity-flow oligo-dT column and used to construct a cDNA library. Degenerative oligonucleotides based on the peptide amino acid sequences were designed to be used as primers for the polymerase chain reaction in order to generate longer nondegenerative probes for screening the cDNA library (Saiki et al. 1988). One positive clone was obtained from an initial screen of 500 000 clones. The cloned cDNA is approximately 1400 base pairs in length and sequencing of this cDNA, while nearly complete, is still in progress. However, preliminary DNA sequence data indicates an open reading frame that encodes 300 amino acids and includes all three peptide sequences which were obtained from the initial microsequencing of the selected tryptic peptides. At the time of this writing the end of the open reading frame has not been determined, but this information is imminent. Secondary structure analysis indicates the presence of an amphipathic (X-helix for the first 25 amino acids, which would be consistent with a mitochondrial targeting sequence. The function of the protein encoded by the cDNA remains to be determined. To verify that the cDNA encodes the 30-kDa proteins, the protein will be expressed in MA-l 0 cells and the size and subcellular localization determined by SDS-PAOE. In addition, the effect of expression of this protein on steroidogenesis in nonhormone-stimulated MA-lO cells will be determined. We also plan to determine the effects
Regulation of the Acute Production of Steroids 87
of the expression of this protein on steroidogenesis in COS cells which have been cotransfected with CSCc.
4.7 Conclusions
The factor(s) responsible for the movement of cholesterol from the outer mitochondrial membrane or possibly some other cellular location such as the plasma membrane, as proposed by Freeman (1989), to the inner
mitochondrial membrane and thus those necessary for regulating the acute production of steroids have not yet unequivocally been identified. While a number of promising candidates exist, further investigations are required to identify this long-sought putative regulator protein. While we favor the basic concepts outlined in the model shown in Fig. 4, it is extremely possible that a combination of models is operational. Thus, perhaps, one could envisage a multi protein complex at the mitochondrial membrane consisting of one or more of the proteins PBR, OBI, VOAC, ANC, SAP, and 30-kOa, resulting in the formation of contact sites and thus cholesterol transfer. However, whatever the mechanism responsible for the rapid transfer of cholesterol to the inner mitochondrial membrane for cleavage by the CSCC, the most important and interesting question remaining to be answered in this area is: "What is the factor(s) responsible for triggering the mechanism involved in this transfer'?"
References
Alberta JA. Epstein LF. Pon LA. Orme-Johnson NR (1989) Mitochondrial localization of a phosphoprotein that rapidly accumulates in adrenal cortex cells exposed to adrenocorticotropic hormone or to cAMP. J Bioi Chem 264:2368-2372
Almahbobi G, Hall PF (1990) The role of intermediate filaments in adrenal steroidogenesis. J Cell Sci 97:679-687
Almahbobi G, Williams LJ, Hall PF (1992a) Attachment of steroidogenic lipid droplets to intermediate filaments in adrenal cells. J Cell Sci 101 :383-393
Almahbobi G, Williams LJ, Hall PF (1992b) Attachment of mitochondria to intermediate filaments in adrenarcells:relcvance to the regulation of steroid synthesis. Exp Cell Res 200:361-369
88 D.M. Stoeeo and B.J. Clark
Almahbobi G, Williams LJ, Han X-G, Hall PF (1993) Binding of lipid droplets and mitochondria to intermediate filaments in rat Leydig cells. J Reprod FertiI98:209-217
Amsterdam A, Suh BS (1991) An inducible functional peripheral benzodiazepine receptor in mitochondria of steroidogenic granulosa cells. Endocrinology 129:503-510
Anholt RRH, Pedersen PO, Souza EB, Snyder SH (1986) The peripheral-type benzodiazepine receptor. J Bioi Chern 261 :576--583
Ardail D, Privat JP, Egret-Charlier M, Levrat C, Lerme F, Louisot P (1990) Mitochondrial contact sites: lipid composition and dynamics. J Bioi Chern 265: 18797-18802
Arthur JR, Boyd GS (1976) The effect of inhibitors of protein synthesis on cholesterol side-chain cleavage in the mitochondria of luteinized rat ovaries. Eur J Biochem 49: 117-127
Ascoli M (1981) Characterization of several clonal lines of cultured Leydig tumor cells:gonadotropin receptors and steroidogenic responses. Endocrinology 108:88-95
Barnea ER, Fares F, Gavish M (1989) Modulatory action of benzodiazepines on human term placental steroidogenesis in vitro. Mol Cell Endocrinol 64:155-159
Besman MJ, Yanagibashi K, Lee TD, Kawamura M, Hall PF, Shively JE (1989) Identification of des-(Gly-Ile)-endozepine as an effector of corticotropin-dependent adrenal steroidogenesis: stimulation of cholesterol delivery is mediated by the peripheral benzodiazepine receptor. Proc Nat! Acad Sci USA 86:4897--4901
Boujrad N, Hudson JR, Papadopoulos V (1993) Inhibition of hormone-stimulated steroidogenesis in cultured Leydig tumor cells by a cholesterol-linked phosphorothioate oligodeoxynucleotide antisense to diazepam-binding inhibitor. Proc Nat! Acad Sci USA 90:5728-5731
Brown AS, Hall PF (1991) Stimulation by endozepine of the side-chain cleavage of cholesterol in a reconstituted enzyme system. Biochem Biophys Res Commun 180:609-614
Brown AS, Hall PF, Shoyab M, Papadopoulos V (1992) Endozepine/diazepam binding inhibitor in adrenocortical and Leydig cell lines: absence of hormonal regulation. Mol Cell EndocrinoI83:1-9
Calvo DJ, Ritta MN, Calandra RX, Medina JH (1990) Peripheral-type benzodiazepine receptors are highly concentrated in mitochondrial membranes of rat testicular interstitial cells. Neuroendocrinology 52:350-353
Cavallaro S, Komeyev A, Guidotti A, Costa E (1992) Diazepam-binding inhibitor (DBI)-processing products, acting at the mitochondrial DB! receptor,
Regulation of the Acute Production of Steroids 89
mediate adrenocorticotropic hormone-induced steroidogenesis in rat adrenal gland. Proc Nat! Acad Sci USA X9: 1 0598-1 0602
Cavallaro S, Pani L, Guidotti A. Costa E (1993) ACTH-induced mitochondrial OBI receptor (MOR) and diazepam binding inhibitor (OBI) expression in adrenals of hypophysectomized rats is not cause-effect related to its immediate steroidogenic action. Life Sci 53: I 137-1 147
Chanderbhan R, Noland BJ, Scallen n, Yahouny GY (19X2) Sterol carrier protein2: delivery of cholesterol from adrenal lipid droplets to mitochondria for pregnenolone synthesis. J Bioi Chern 257:8928-8934
Chanderbhan RF, Kharroubi AT, Noland BJ, Scallen n, Yahouny GY (1986) SCP2: further evidence for its role in adrenal steroidogenesis. Endocr Res 12:351-360
Chaudhary LR, Stocco OM (1991) Effect of different steroidogenic stimuli on protein phosphorylation in MA-IO mouse Leydig tumor cells. Biochim Biophys Acta 1094:175-184
Cooke BA, Lindh LM, Janszen FHA (1976) Correlation of protein kinase activation and testosterone production after stimulation of Leydig cells with luteinizing hormone. Biochem J 160:439-446
Cooke BA. Janszen FHA. van Oriel MJA, van der Molen HJ (1979) Evidence for the involvement of lutropin-independent RNA synthesis in Leydig cell steroidogenesis. Mol Cell Endocrinol 14: 181-189
Costa E, Guidotti A (1991) Diazepam binding inhibitor (OBI): a peptide with multiple biological actions. Life Sci 49:325-344
Crivello JF, Jefcoate CR (197X) Mechanism of corticotropin action in rat adrenal cells. I. Effect of inhibitors of protein synthesis and of microfilament formation on corticosterone synthesis. Biochim Biophys Acta 542:3 I 5-329
Crivello JF, Jefcoate CR (19S0) Intracellular movement of cholesterol in rat adrenal cells. Kinetics and effects of inhibitors. J BioI Chern 255:X 144-SI51
Edelman AM, Blumenthal OK, Krebs EG (1987) Protein serine/threonine kinases. Annu Rev Biochem 56:567-613
Elliott ME, Goodfriend TL, Jefcoate CR (1993) Bovine adrenal glomerulosa and fasciculata cells exhibit 2X.5-kilodalton proteins sensitive to angiotensin, other agonists, and atrial natriuretic peptide. Endocrinology 133: 1669-1677
Epstein LF, Orme-Johnson NR (1991 a) Acute action of luteinizing hormone on mouse Leydig cells: accumulation of mitochondrial phosphoproteins and stimulation of testosterone biosynthesis. Mol Cell Endocrinol 81: 113-126
Epstein LF, Orme-Johnson NR ( 1991 b) Regulation of steroid hormone biosynthesis: identification of precursors of a phosphoprotein targeted to the mito-
90 D.M. Stoeeo and B.J. Clark
chondrion in stimulated rat adrenal cortex cells. J Bioi Chern 266: 19739-19745
Epstein LA, Alberta JA, Pon LA, Onne-Johnson NR (1989) Subcellular localization of a protein produced in adrenal cortex cells in response to ACTH. Endocr Res 15: 117-127
Farrow RL, Rakestraw DJ (1992) Detection of trace molecular species using degenerate four-wave mixing. Science 257:1894-1900
Ferguson JJ (1963) Protein synthesis and adrenocorticotropin responsiveness. J Bioi Chern 238:2754-2759
Ferguson JJ, Morita Y (1964) RNA synthesis and adrenocorticotropin responsiveness. Biochim Biophys Acta 87:348-350
Freeman DA (1989) Plasma membrane cholesterol: removal and insertion into the membrane and utilization as substrate for steroidogenesis. Endocrinology 124:2527-2534
Frustaci J, Mertz LM, Pedersen RC (1989) Steroidogenesis activator polypeptide (SAP) in the guinea pig adrenal cortex. Mol Cell Endocrinol 64: 137-143
Fujiki Y, Tsuneoka M, Tashiro Y (1989) Biosynthesis of nonspecific lipid transfer protein (sterol carrier protein 2) on free polyribosomes as a larger precursor in rat liver. J Biochem 106: 1126-1131
Garnier M, Boujrad N, Oke BO, Brown AS, Riond J, Ferrara P, Shoyab M, Suarez-Quian CA, Papadopoulos V (1993) Diazepam binding inhibitor is a paracrine-autocrine regulator of Leydig cell proliferation and steroidogenesis. Action via peripheral-type benzodiazepine receptor and independent mechanisms. Endocrinology 132:444-458
Garren LD, Ney RL, Davis WW (1965) Studies on the role of protein synthesis in the regulation of corticosterone production by ACTH in vivo. Proc Nat! Acad Sci USA 53:1443-1450
Garren LD, Gill GN, Masui H, Walton GM (1971) On the mechanism of action of ACTH. Rec Prog Horm Res 27:433-478
Griffin Green E, Onne-Johnson NR (1991) Inhibition of steroidogenesis in rat adrenal cortex cells by a threonine analog. J Steroid Biochem Mol Bioi 40:421-429
Guarneri P, Papadopoulos V, Costa E (1992) Regulation of pregnenolone synthesis in C6 glioma cells by 4' -chlorodiazepam. Proc Nat1 Acad Sci USA 89:5118-5122
Hall PF (1984a) The role of the cytoskeleton in hormone action. Can J Biochern Cell Bioi 62:653-665
Hall PF (I 984b ) Cellular organization for steroidogenesis. Int Rev Cytol 86:53-95.
Regulation of the Acute Production of Steroids 91
Hall PF (1985) The role of the cytoskeleton in the supply of cholesterol for steroidogenesis. In: Strauss JF, Menon KMJ (eds) Lipoprotein and cholesterol metabolism in steroidogenic tissues. Strickley, Philadelphia, pp 207-217
Hall PF (1988) Testicular steroid synthesis:organization and regulation. In: Knobil E, Neill J (cds) The physiology of reproduction. Raven, New York, pp 975-998
Hall PF, Almahbobi G (1992) The role of the cytoskeleton in the regulation of steroidogenesis. J Steroid Biochem Mol Bioi 43:769-777
Hall PF, Charponnier C, Nakamura M, Gabbiani G ( 1979) The role of microfilaments in the response of adrenal tumor cells to ACTH. J BioI Chern 254:90S0-90S4
Hall PF, Papadopoulos V, Yanagibashi K (l9S8) On the mechanism of action of ACTH. In: Imura H, Shizume K, Yoshida S (eds) Progress in endocrinology. Elsevier Science, Amsterdam, pp 253-258
Hanukoglu I (1992) Steroidogenic enzymes: structure function, and role in the regulation of steroid hormone biosynthesis. J Steroid Biochem Mol BioI 43:779-804
Hanukoglu I, Hanukoglu Z ( 1986) Stoichiometry of mitochondrials, cytochromes P450, adrenodoxin and adrenodoxin reductase in adrenal cortex and corpus luteum. Eur J Biochem 157:27-31
Holloway CD, Kenyon CJ, Dowie LJ, Corrie JET, Gray CE, Fraser R (l9S9) Effect of the benzodiazepines diazepam, des-n-methyldiazepam and midazolam on corticosteroid biosynthesis in bovine adrenocortical cells in vitro; location of site of action. J Steroid Biochem 33:219-225
Hsueh AJW, Adashi EY, Jones PBC, Welsh TH (1984) Hormonal regulation of the differentiation of cultured ovarian granulosa cells. Endocr Rev 5:76-127
Hwang ST, Wachter C, Schatz G (1991) Protein import into the yeast mitochondrial matrix: a new translocation intermediate between the two mitochondrial membranes. J Bioi Chern 266:21083-21089
Jefcoate CR, DiBartolomeos MJ, Williams CA, McNamara BC (I9S7) ACTH regulation of cholesterol movement in isolated adrenal cells. J Steroid Biochern 27:721-729
Jefcoate CR, McNamara BC, Artemenko I, Yamazaki T (1992) Regulation of cholesterol movement to mitochondrial cytochrome P450scc in steroid hormone synthesis. J Steroid Biochem Mol Bioi 43:751-767
Karaboyas GC, Koritz SB (1965) Identity of the site of action of cAMP and ACTH in corticosteroidogenesis in rat adrenal and beef adrenal cortex slices. Biochemistry 4:462-46S
92 D.M. Staeea and B.J. Clark
Keller GA, Scallen TJ, Clarke 0, Maher PA, Krisans SK, Singer SJ (1989) Subcellular localization of sterol carrier protein-2 in rat hepatocytes: its primary localization to peroxisomes. J Cell Bioi 108: 1353-1361
Kinnally KW, Zorov DB, Antonenko YN, Snyder SH, McEnery MW, Tedesschi H (1993) Mitochondrial benzodiazepine receptor linked to inner membrane ion channels by nanomolar actions of ligands. Proc Nat! Acad Sci USA 90:1374-1378
Knudsen J, Nielsen M (1990) Diazepam-binding inhibitor: a neuropeptide and/or an acylco A ester binding protein? Biochem J 265:927-928
Krueger RJ, Orme-Johnson NR (1983) Acute adrenocorticotropic hormone stimulation of adrenal corticosteroidogenesis. J Bioi Chern 258: 10 159-10 167
Krueger RJ, Orme-Johnson NR (1988) Evidence for the involvement of a labile protein in stimulation of adrenal steroidogenesis under conditions not inhibitory to protein synthesis. Endocrinology 122: 1869-1875
Krueger KE, Papadopoulos V (1990) Peripheral-type benzodiazepine receptors mediate translocation of cholesterol from outer to inner mitochondrial membranes in adrenocortical cells. J Bioi Chern 265: 15015-15022
Krueger KE, Papadopoulos V (1992) Mitochondrial benzodiazepine receptors and the regulation of steroid biosynthesis. Ann Rev Pharmacol Toxicol 32:211-237
Krueger KE, Whalin ME, Papadopoulos V (1991) Regulation of steroid synthesis in the adrenals by mitochondrial benzodiazepine receptors. In: Costa E, Paul SM (eds) Neurosteroids and brain function, vol 8. Thieme Medical, New York, pp 155-160
Lambeth JD, Seybert OW, Kamin H (1979) Ionic effects on adrenal steroidogenic electron transport. The role of adrenodoxin as an electron shuttle. J Bioi Chern 254:7255-7264
Lambeth JD, Xu XX, Glover M (1987) Cholesterol sulfate inhibits adrenal mitochondrial cholesterol side-chain cleavage at a site distinct from cytochrome P450scc. Evidence for an intramitochondrial cholesterol trans locator. ] Bioi Chern 262:9181-9188
Li X, Warren OW, Gregoire J, Pedersen RC, Lee AS (1989) The rat 78,000 dalton glucose-regulated protein (GRP78) as a precursor of the rat steroidogenesis-activator polypeptide (SAP): the SAP coding sequence is homologous with the terminal end ofGRP78. Mol EndocrinoI3:1944-1952
Liscum L, Dahl NK (1992) Intracellular cholesterol transport. J Lipid Res 33: 1239-1254
Marsh JM (1975) The role of cyclic AMP in gonadal function. Adv Cycl Nucl Res 6:137-199
McEnery MW, Snowman AM, Trifiletti RR, Snyder SH (1992) Isolation of the mitochondrial benzodiazepine receptor:association with the voltage-de-
Regulation of the Acute Production of Steroids 93
pendent anion channel and the adenine nucleotide carrier. Proc Natl Acad Sci USA 89:3170--3174
Mendelson C, Dufau ML, Catt KJ (1975) Dependence of gonadotropin-induced steroidogenesis on RNA and protein synthesis in the interstitial cells of the rat testis. Bioehim Biophys Acta 411:222-230
Mendis-Handagama SMLC, Zirkin BR, Seallen TJ, Ewing LL (1990) Studies on peroxisomes of the adult rat Leydig cell. J Androl II :270-27X
Mertz LM, Pedersen RC (1989a) Steroidogenesis activator polypeptide may be a product of glucose regulated protein 78 (GRP78) Endocr Res 15:101-115
Mertz LM, Pedersen RC (\ 989b) The kinetics of steroidogenesis activator polypeptide in the rat adrenal cortex. J Bioi Chern 264: 15274-15279
Miller WL (1988) Molecular biology of steroid hormone synthesis. Endoer Rev 9:295-318
Mittre H, Aunai P, Benhaim A. Leymarie P (1990) Acute stimulation by lutropin of mitochondrial protein synthesis in small luteal cells. Eur J Biochem 187:721-726
Moran 0, Sandri G, Panfili E, Stuhmer W, Sorgato MC (1990) Electrophysiological characterization of contact sites in brain mitochondria. J Bioi Chern 265:908-913
Mrotek J, Hall PF (1977) Response of adrenal tumor cells to ACTH: site of inhibition by cytochalasin B. Biochemistry 16:3177-3181
Mukhin AG, Papadopoulos V, Costa E, Krueger KE (1989) Mitochondrial benzodiazepine receptors regulate steroid biosynthesis. Proc Natl Acad Sci USA 86:9813-9816
Nagy L, Freeman DA (1990a) Cholesterol movement between the plasma membrane and the cholesteryl ester droplets of cultured Leydig tumor cells. Biochem J 271 :809-814
Nagy L, Freeman DA (1990b) Effect of cholesterol transport inhibitors on steroidogenesis and plasma membrane cholesterol transport in cultured MA-1O Leydig tumor cells. Endocrinology 126:2267-2276
Nakamura M, Watanuki M, Hall PF (1978) On the role of protein synthesis in the response of adrenal tumor cells to ACTH. Mol Cell Endocrinol 12:209-219
Ney RL, Davis WW, Garren LD (1966) Heterogeneity of template RNA in adrenal glands. Science 153:896--897
Ohlendieck K, Riesinger I, Adams V, Krause J, Dieter B (1986) Enrichment and biochemical characterization of boundary membrane contact sites from rat liver mitochondria. Biochim Biophys Acta 860:672-689
94 D.M. Stoeeo and B.J. Clark
Oke BO, Suarez-Qui an CA, Riond J, Ferrara P, Papadopoulos Y (1992) Cell surface localization of the peripheral-type benzodiazepine receptor in adrenal cortex. Mol Cell Endocrinol 87:R I-R6
Orme-Johnson NR (1990) Distinctive properties of adrenal cortex mitochondria. Biochim Biophys Acta 1020:213-231
Papadopoulos Y (1993) Peripheral-type benzodiazepine/diazepam binding inhibitor receptor: biological role in steroidogenic cell function. Endocr Rev 14:222-240
Papadopoulos Y, Mukhin AG, Costa E, Krueger KE (1990) The peripheraltype benzodiazepine receptor is functionally-linked to Leydig cell steroidogenesis. J Bioi Chern 265:3772-3779
Papadopoulos Y, Nowzari FB, Krueger KE (l991a) Hormone-stimulated steroidogenesis is coupled to mitochondrial benzodiazepine receptors. Trophic hormone action on steroid biosynthesis is inhibited by flunitrazepam. J Bioi Chern 266:3682-3687
Papadopoulos Y, Berkovich A, Krueger KE, Costa E, Guidotti A (1991 b) Diazepam binding inhibitor and its processing products stimulate mitochondrial steroid biosynthesis via an interaction with mitochondrial benzodiazepine receptors. Endocrinology 129: 1481-1488
Papadopoulos Y, Berkovich A, Krueger KE (1991 c) The role of diazepam binding inhibitor and its processing products at mitochondrial benzodiazepine receptors: regulation of steroid biosynthesis. Neuropharmacology 30:1417-1423
Papadopoulos Y, Guarneri P, Krueger KE, Guidotti A, Costa E (1992) Pregnenolone biosynthesis in C6-2B glioma cell mitochondria: regulation by a mitochondrial diazepam binding inhibitor receptor. Proc Natl Acad Sci USA 89:5113-5117
Pedersen RC (1984) Polypeptide activators of cholesterol side-chain cleavage. Endocr Res 10:533-561
Pedersen RC (1987) Steroidogenesis activator polypeptide (SAP) in the rat ovary and testis. J Steroid Biochem 27:731-735
Pedersen RC, Brownie AC (1983) Cholesterol side-chain cleavage in the rat adrenal cortex: isolation of a cycloheximide-sensitive activator peptide. Proc Nat! Acad Sci USA 80: 1882-1886
Pedersen RC, Brownie AC (1987) Steroidogenesis-activator polypeptide isolated from a rat Leydig cell tumor. Science 236: 188-190
Pfanner N, Rassow J, Wienhues U, Hergersberg C, Sollner T, Becker K, Neupert W (1990) Contact sites between inner and outer membranes: structure and role in protein translocation into the mitochondria. Biochim Biophys Acta 1018:239-242
Regulation of the Acute Production of Steroids 95
Phillips Me. Johnson W J, Rothblat GH ( 1(87) Mechanisms and consequences of cellular cholcstcrol exchange and transfer. Biochim Biophys Acta 906:223-276
Pon LA, Orme-Johnson NR (1986) Acute stimulation of steroidogenesis in corpus luteum and adrenal cortcx by peptide hormones. J Bioi Chcm 261 :6594-6599
Pon LA, Orme-Johnson NR (1988) Acute stimulation of corpus luteum cells by gonadotropin or adenosine 3' 5' -monophosphate causes accumulation of a phosphoprotein concurrent with acccleration of steroid synthesis. Endocrinology 123:1942-1948
Pon LA, Epstein LF, Ormc-Johnson NR (1986a) Acute cAMP stimulation in Leydig cells: rapid accumulation of a protein silimar to that detected in adrenal cortex and corpus luteum. Endocr Res 12:429-446
Pon LA, Hartigan JA, Orme-Johnson NR (1986b) Acute ACTH regulation of adrenal corticosteroid biosynthesis: rapid accumulation of a phosphoprotein. J Bioi Chem 261: 13309-13316
Pon L. Moll T, Yestweber D, Marshallsay B, Schatz G (1989) Protein import into mitochondria: ATP-dependcnt protcin translocation activity in a submitochondrial fraction enriched in mcmbrane contact sites and specific proteins. J Cell Bioi 109:2603-2616
Privalle CT, Crivello JF, Jcfcoate CR (1983a) Regulation of intramitochondrial cholesterol transfer to side-chain cleavage cytochromc P450,cc in rat adrenal gland. Proc Natl Acad Sci USA 80:702-706
Privalle CT, McNamara BC, Dhariwal MS, Jefcoatc CR (1983b) ACTH control of cholesterol side-chain cleavage at adrenal mitochondrial cytochrome P45(kcc. Regulation of intramitochondrial cholesterol transfer. Mol Cell EndocrinoI53:87-101
Python CP. Rossicr MF. Yallotton MB, Capponi AM (1993) Peripheral-type benzodiazepines inhibit calcium channels and aldosterone production in adrenal glomerulosa cells. Endocrinology 132: 1489-1495
Rassow J, Guiard B, Weinhucs U, Herzog y, Hartl FU, Ncupert W (1989) Translocation arrest by reversible folding of a precursor protein imported into mitochondria. A means to quantitate translocation contact sites. J Cell Bioi 109:1421-1428
Rcinhart MP (1990) Intracellular sterol trafficking. Expcrientia 46:59<}-61 0 Rennert H, Chang YJ, Strauss JF (1993) Intraccllular cholesterol dynamics in
stcroidogenic cells: a contemporary view. In: Adashi EY, Leung PK (eds) The ovary. Ravcn, New York, pp 147-164
Rommcrts FFG, Cooke BA, van del' Molen HJ (1974) The role of cAMP in the regulation of steroid biosynthesis in testis tissue. J Steroid Biochem 5:27<}-285
96 D.M. Stoeeo and B.J. Clark
Saiki RK, Gelfand DH, Stoffel S, Scharf SJ, Higuchi R, Horn GT, Mullis KB, Erlich HA (1988) Primer-directed enzymatic amplification of DNA with a thermostable DNA polymerase. Science 239:487-491
Sala GB, Hayashi K, Catt KJ, Dufau ML (1979) Adrenocorticotropin action in isolated adrenal cells. J BioI Chern 254:3861-3875
Sandri G, Siagri M, Panfili E (1988) Influence of Ca-2+ on the isolation from rat brain mitochondria of a fraction enriched of boundary membrane contact sites. Cell Calcium 9: 159-165
Schroeder F, Jefferson JR, Kier AB, Knittel J, Scallen TJ, Wood WG, Hapala I (199 I) Membrane cholesterol dynamics:cholesterol domains and kinetic pools. Proc Soc Exp BioI Med 196:235-252
Schulster D (1974) Corticosteroid and ribonucleic acid synthesis in isolated adrenal cells: inhibition by actinomycin D. Mol Cell Endocrinol 1:55-64
Schulster D, Richardson MC, Palfreyman JW (1974) The role of protein synthesis in adrenocorticotrophin action: effects of cycloheximide and puromycin on the steroidogenic response of isolated adrenocortical cells. Mol Cell Endocrinol 2: 17-29
Schwaiger M, Herzog V, Neupert W (1987) Characterization of translocation contact sites involved in the import of mitochondrial proteins. J Cell BioI 105:235-246
Shibata H, Kojima I, Ogata E (J 986) Diazepam inhibits potassium-induced aldosterone secretion in adrenal glomerulosa cell. Biochem Biophys Res Commun 135:994-999
Simpson ER, Waterman MR (1983) Regulation by ACTH of steroid hormone biosynthesis in the adrenal cortex. Can J Biochem Cell BioI 61:692-707
Simpson ER, McCarthy JL, Peterson JA (J 979) Evidence that the cycloheximide-sensitive site of ACTH action is in the mitochondrion. J BioI Chern 253:3135-3139
Snyder SH, Verma A, Trifiletti RR (1987) The "peripheral-type" benzodiazepine receptor: a protein of mitochondrial outer membrane utilizing porphyrins as endogenous ligands. FASEB J 1:282-288
Soto EA, Kliman HJ, Strauss JF, Paavola LG (1986) Gonadotropins and cyclic adenosine 3',5'-monophosphate (cAMP) alter the morphology of cultured human granulosa cells. BioI Reprod 34:559-569
Stevens VL, Xu T, Lambeth JD (1993) Cholesterol trafficking in steroidogenic cells: reversible cycloheximide-dependent accumulation of cholesterol in a pre-steroidogenic pool. Eur J Biochem 216:557-563
Stocco DM (1992) Further evidence that the mitochondrial proteins induced by hormone stimulation in MA-1O mouse Leydig tumor cells are involved in the acute regulation of steroidogenesis. J Steroid Biochem Molec BioI 43:319-333
Regulation of the Acute Production of Steroids 97
Stocco OM, Ascoli M (1993) The use of genetic manipulation of MA-LO Leydig tumor cells to demonstrate the role of mitochondrial proteins in thc acute regulation of steroidogenesis. Endocrinology 132:959-967
Stocco OM, Chaudhary LR (1990) Evidence for the functional coupling of cAMP in MA-IO mouse Leydig tumor cells. Cell Signal 2: 161-170
Stocco OM, Chen W (1991) Presence of identical mitochondrial proteins in unstimulated constitutive steroid-producing R2C rat Leydig tumor and stimulated nonconstitutive steroid-producing MA- \0 mouse Leydig tumor cells. Endocrinology 128: 1918-1926
Stocco OM, Clark BJ (1993) The requirement of phosphorylation on a threonine residue in the acute regulation of steroidogenesis in MA-I 0 mouse Leydig tumor cells. J Steroid Biochem Mol Bioi 46:337-347
Stocco OM, Kilgore MW (1988) Induction of mitochondrial proteins in MA-10 Leydig tumor cells with human choriogonadotropin. Biochem J 249:95-103
Stocco OM, Sodeman TC (1991) The 30-kOa mitochondrial proteins induced by hormone stimulation in MA-I 0 mouse Leydig tumor cells are proccssed from larger precursors. J BioI Chern 266: 19731-19738
Stocco OM, Wells J, Clark BJ (1993) Effects of hydrogen peroxide on steroidogenesis in mouse Leydig tumor cells. Endocrinology 133:2827-2832
Stone 0, Hechter 0 (1954) Studies on ACTH action in perfused bovine adrenals: site of action of ACTH in corticosteroidogenesis. Arch Biochem Biophys 51 :457-469
Thomson T, Fraser R, Kenyon CJ (1992) Inhibition of bovine adrenocortical steroidogenesis by benzodiazepines: a direct effect on microsomal hydroxylation or an inhibition of calcium uptake. J EndocrinoI135:361-362
Trzeciak WH, Simpson ER, Scallen TJ, Vahouny GV, Waterman MR (1987) Studies on the synthesis of SCP-2 in rat adrenal cortical cells in monolayer culture. J Bioi Chern 262:3713-3720
Tuls J, Geren L, Lambeth JO, Millet F (1987) The use of a specific fluorescence probe to study thc interaction of adrenodoxin reductase and cytochromc P450scc. J Bioi Chern 262:10020-10025
Vahouny GV, Chanderbhan R, Noland BJ, Irwin 0, Dennis P, Lambeth JO, Scali en TJ (1983) Sterol carrier protcin2: identification of adrenal sterol carrier protein2 and site of action for mitochondrial cholesterol utilization. J Bioi Chern 258: 11731-11737
Vahouny GV, Chanderbhan R, Noland BJ, Scali en TJ (1985) Cholesterol ester hydrolase and sterol carrier proteins. Endocr Res 10:473-505
Vahouny GV, Chanderbhan R, Kharroubi A, Noland BJ, Pastuszyn A, Scali en TJ (\ 987) Sterol carrier and lipid transfer proteins. Adv Lipid Res 22:83-113
98 D.M. Stocco and B.J. Clark: Regulation of the Acute Production
van Amerongen A, van Noort M, van Beckhoven JRCM, Rommerts FFG, Orly J, Wirtz KW A (1989a) The subcellular distribution of the nonspecific lipid transfer protein (sterol carrier protein 2) in rat liver and adrenal gland. Biochim Biophys Acta 1001:243-248
van Amerongen A, Demel RA, Westerman J, Wirtz KW A (I 989b ) Transfer of cholesterol and oxysterol derivatives by the non-specific lipid transfer protein (sterol carrier protein 2): a study on its mode of action. Biochim Biophys Acta 1004:36-43
van Noort M, Rommerts FFG, van Amerongen A, Wirtz KW A (1988) Regulation of SCP2 in the soluble fraction of rat Leydig cells. Kinetics and possible role of calcium influx. Mol Cell Endocrinol 56: 133-138
Verma A, Snyder SH (1989) Peripheral type benzodiazepine receptors. Annu Rev Pharmacol Toxicol 29:307-322
Vestweber D, Schatz G (1988) A chimeric mitochondrial precursor protein with internal disulphide bridges blocks import of authentic precursors into mitochondria. A means to quantitate translocation contact sites. J Cell Bioi 107:2037-2043
Yamamoto R, Kallen CB, Babalola GO, Rennert H, Billheimer JT, Strauss JF (1991) Cloning and expression of a eDNA encoding human sterol carrier protein 2. Proc Nat! Acad Sci USA 88:463-467
Yanagibashi K, Ohno Y, Kawamura M, Hall PF (1988) The regulation of intracellular transport of cholesterol in bovine adrenal cells: purification of a novel protein. Endocrinology 123:2075-2082
Yanagibashi K, Ohno Y, Nakamichi N, Matsui T, Hyashida K, Takamura M, Yamada K, Tou S, Kawamura M (1989) Peripheral-type benzodiazepine receptors are involved in the regulation of cholesterol side chain cleavage in adrenocortical mitochondria. J Biochem 106: 1026-1029
5 Extracellular Matrix Elements, Cell Adhesion Molecules, and Signal Transduction in the Control of Sertoli Cell Function
M.Oym
5.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 99 5.2 The Tunica Propria of the Rat Seminiferous Tubule. . . . . . . . . .. I () I 5.3 Cell Adhesion Molecules ................................ 103 5.4 Signal Transduction from the ECM ........................ 105 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. I 12
5.1 Introduction
Most cells in intact higher organisms are in contact with an extracellular matrix (ECM) composed of both protein and carbohydrate in the form of glycoproteins and proteoglycans (Hay 1981). Decades ago it was postulated that the ECM plays an important role in the development and maintenance of the differentiated state and it is now generally accepted that the ECM is fundamental in the regulation of development and in differentiation (Martin and Timpl 1987; Adams and Watt 1993). One specialized form of the ECM is the basement membrane. Basement membranes are acellular, amorphous layers or sheets of material, generally 0.1-0.2 !lm thick that surround nerves, muscle cells, fat cells, and underlie all endothelial cells and epithelial cells (Fig. I; Martin and Timpl 1987). The basement membrane is composed of laminin, type IV
100
S .. StflmBnt mBmbrlNH1
<dI> ConnecU ••
tissue
M.Oym
-"":;."c;..-~~-B",",menr msmbrlfne
Fig. 1. The distribution of the basement membrane in a number of tissues. The basement membrane underlies all epithelial (epidermis shown) and endothelial tissues in a polarized manner. Smooth muscle cells and fat cells (not shown) are completely surrounded by the basement membrane. It also surrounds nerves and parts of the brain. Modified from Leblond and Inoue (1989)
collagen, heparan sulfate proteoglycan, and nidogen/entactin. Laminin (Mr = 800 000), a cell-binding molecule of the basement membrane, is composed of three distinct chains held together by disulfide bonds (A, Mr = 400 000; Bl, Mr = 2lO 000; B2, Mr = 200 000; Fig. 2). Laminin has been shown to possess a great deal of biological activity (Kleinman et al. 1993). For example, laminin stimulates attachment, migration, and tissue-specific differentiation in many cell types, including epithelial, muscle, and neuronal cells. Various active domains and sequences within the laminin molecule have been described which mediate these cellular responses (Kleinman et al. 1993). Active sites on the B 1 chain for cell attachment include the YIGSR (Y, tyrosine; I, isoleucine; G, glycine; S, serine; R, arginine) sequence. Two active sequences on the A chain are the RGD (R, arginine; G, glycine; D, aspartic acid) and the SIKV A V (S, serine; r, isoleucine; K, lysine; V, valine; A, alanine; V, valine) sites. Given the multiple biological activities and several distinct
Extracellular Matrix Elements, Cell Adhesion Molecules 101
Fig. 2. The laminin molecule, a trimeric structure composed of a central A chain (A) and two similar and smallcr B chains (Bl and B2). Note the YIGSR sequence on the B I chain and the RGD and SIKVAV sequences on the A chain. (Drawing courtesy of Dr. Derrick Grant, National Institute of Dental Research, NIH)
active sites, it is not surprising that there are several cell surface receptors for laminin (see below).
5.2 The Tunica Propria of the Rat Seminiferous Tubule
In the adult rat testis, the seminiferous tubules are separated from the interstitium by a series of noncellular and cellular layers collectively referred to as the tunica propria (Dym 1983; (Fig. 3). Immediately deep to the plasma membranes of the Sertoli cells and the spermatogonia is a thin, clear zone referred to as the lamina lucida, next to which is a darker region called the lamina den sa. Both the lamina lucida and the lamina densa are considered to be part of the basement membrane. Outside the basement membrane is a clear zone containing type I collagen fibrils in varying degrees of orientation and peripheral to this collagen zone is a layer of flattened cells, the peritubular myoid cells, followed by a layer of lymphatic endothelial cells. The exact appearance and composition of
102 M.Oym
Fig. 3. This electron micrograph demonstrates the structures of the tunica propria at the base of the seminiferous epithelium. The asterisks indicate the lamina densa of the basement membrane. Note the thin clear lamina lucida between the basal plasma membranes of the seminiferous epithelium and the lamina densa. Both the lamina densa and the lamina lucida are considered to be part of the basement membrane. External to the lamina densa is a zone of type I collagen fibrils (arrows) seen in cross section, external to which is a myoid cell. The myoid cell also has basement membrane material on its surface. A lymphatic endothelial cell is present external to the myoid cell
the basement membrane varies in different tissues, but there are a number of common features.
The distribution of laminin, type IV collagen, and heparan sulfate proteoglycan was examined in the rat testicular tunica propria by electron microscopic immunocytochemistry (Hadley and Dym 1987). Distinct patterns were observed for each antigen within the ECM layers of the tunica propria. Laminin, type IV collagen, and heparan sulfate proteoglycan all localized to the seminiferous tubule basement membrane. Type IV collagen and heparan sulfate proteoglycan, but not laminin, localized to the seminiferous tubule side of the peritubular myoid cells. All four of the antigens were localized between the peritubular cells and the lymphatic endothelial cells. Failure to localize fibronectin in the basement membrane underlying the seminiferous epithelium supports
Extracellular Matrix Elements, Cell Adhesion Molecules 103
- SDIINIFEROUS EPITHELIUI1
LAMININ
TYPE IV COLL.>.GEN
HEPARAN SULFATE PROTEOGLYCAN
NlDOGEN/ENTACTIN
Fig. 4. The chemical composition or the basement membrane is depicted in this light micrograph. Immunocytochemical studies have demonstrated the presence of laminin, type IV collagen. heparan sulfate proteoglycan, and nidogen/entactin in the basement membrane. From Hadley and Dym (1987) and Lian et al. (1992)
the notion that adult Sertoli cells do not produce fibronectin . More recently , Enders and colleagues (Lian et al. 19(2) have localized nidogen/entactin in the basement membrane of the seminiferous tubule and we have also confirmed this result (unpublished observations). Figure 4 summarizes the chemical composition of the ECM underlying the semi niferous epithelium in the rat.
5.3 Cell Adhesion Molecules
The first cell surface binding protein for laminin that was identified was the M r ::: 67 000 protein purified from murine and human neoplastic cell lines and bovine myoblasts (Martin and Timpl 1(87). This protein binds to the YIGSR peptide of the laminin B 1 chain. The DNA sequence of
104 M.Oym
the 67 000 laminin binding protein actually predicts a protein that is 32 000. Indeed, the 67 000, 32 000, and another protein at 45 000 were found to cross-react with each other when antisera to a fusion protein from the sequence of the full length clone for the 32 000 laminin binding protein were used. Laminin binding proteins in the range of the 32000--36000 have also been identified and the 32 000 laminin binding protein has been postulated to be a precursor of the 67 000 laminin receptor.
We have identified and characterized a number of Sertoli cell surface molecules which interact with laminin (Davis et al. 1991). Using laminin-Sepharose affinity chromatography and 125I-Iabeled laminin binding to Sertoli cell plasma membranes, binding proteins have been identified with the Mr 110 000, 67 000, 55 000, 45 000, 36 000, and 25 000. In addition, the Mr 110000 and 67 000 laminin-binding proteins were phosphorylated. The 67 000, 45 000, and 36 000 proteins react with antibodies to the previously characterized laminin receptor and these antibodies stain the basolateral membranes of Sertoli cells and possibly spermatogonia in vivo. We were surprised to note that the laminin receptor staining was not uniformly distributed at the base of the seminiferous tubule; we attribute this to the fact that large amounts of laminin in the area may inhibit binding of the antibody to the laminin. In addition, the laminin-binding proteins appeared to be present between the Sertoli cells and the spermatogonia, but always beneath the area demarcated by the Sertoli-Sertoli cell tight junctions. This work represents the first identification and characterization of ECM-binding proteins in an endocrine organ and suggests that they have an important role in matrix-induced biological function.
Integrins are a family of cell surface receptors that mediate the action of the ECM on cells (Hynes 1992). The integrins, several of which bind laminin, are noncovalently linked glycoprotein a/~ heterodimers with at least 12a and 8 ~ subunits. There is increasing evidence to suggest that the integrins act as true receptors and that they are capable of transducing signals that are involved in the regulation of a number of cell functions. Furthermore, in a recent publication, Jones and Watt (1993) presented evidence to suggest that in a renewing system such as the skin, stem cells which rest on the basement membrane possess the highest levels of ~ 1 integrins. The reduction in the number of ~ I integrins is a stimulus for the terminal differentiation of the cells. If this hypothesis is
Extracellular Matrix Elements, Cell Adhesion Molecules 105
correct, it will be interesting to examine the levels of integrins in the spermatogonial population in order to determine whether the A stem cells possess higher levels of integrins than the type A I to A4, intennediate, or type B spermatogonia. In our own results in 20-day-old rats, we localized ~ I integrins to the tunica propria of the seminiferous tubules and within the epithelium as well, as discrete staining (Davis et al. 1991). Palombi and colleagues (1992) demonstrated a more precise localization of the ~ 1 integrins in the seminiferous epithelium, around late spermatid heads and above spermatogonia at stages I-VIII. However, they did not find integrins at the base of the spermatogonia and Sertoli cells, i.e., at the region of cell contact with the basement membrane.
5.4 Signal Transduction from the ECM
The prevailing view on the mechanism of the signal that is transduced from the matrix receptor to the intracellular machinery suggests that the integrins or other matrix-binding proteins transmit signals by reorganizing the cytoskeleton. Indeed, a number of actin-binding proteins such as tal in, vinculin, and a-actinin have been shown to colocalize both with the matrix receptors and with the actin cytoskeleton (Burridge et al. 1988). The cytoskeletal changes then would somehow induce subsequent changes in gene expression and thus lead to altered function. One report demonstrated that PC12 cell neurite elongation in response to ECM could occur in the absence of microtubules, suggesting that shape changes may not be a prerequisite for the response to the ECM (Lamoureux et al. 1990). In more recent work, Bissell and her colleagues used a suspension culture system for mammary epithelial cells and concluded that the basement membrane induces tissue-specific gene expression in the absence of cell-to-cell interactions and morphological polarity (Streuli et al. 1991). Thus, an alternate paradigm has recently been proposed in which the matrix-binding proteins in fact act as true receptors capable of giving rise to biochemical signals directly without invoking cytoskeletal changes or shape changes (Juliano and Haskill 1993). Several reports have shown increased tyrosine phosphorylation of a number or proteins of ~ 120-130 kDa after integrin clustering with anti-integrin antibodies (Kornberg et al. 1991). This phosphorylation
106 M. Dym
was induced quite specifically by the integrins since clustering of other cell surface proteins did not lead to changes in phosphorylation patterns. One particular protein of 125 kDa that is phosphorylated is present at focal contacts and has been termed the pp 125 focal adhesion kinase (pp 1 25fak; Juliano and Haskill 1993; Burridge et al. 1992). Signal transduction by integrins has also been shown to induce collagenase and stromelysin gene expression (Werb et al. 1989). Kleinman and her colleagues demonstrated that cell surface, non-integrin laminin receptors are phosphorylated during laminin-mediated process formation in neuronal cells (Weeks et al. 1990). Other work demonstrated that protein kinase C is involved in laminin-mediated neurite outgrowth (Bixby and Jhabvala 1990). During heart development, it was shown that Ca2+
and protein kinase C may playa role in signal transduction and it was suggested that a regulatory GTP-binding protein (G protein) may be involved (Runyan et al. 1990).
The signal transduction mechanisms by which the basement membrane modulates cell function is an area of research that is still in its infancy. Our results in the testis demonstrate that the transducers and second messengers that are instrumental in the ECM-mediated events leading to changes in cell function may not be very different from the transducers and second messengers involved in hormone-mediated signal transduction. We examined the signaling events that occur when basement membrane components bind to Sertoli cells. On several basement membrane substrates, Sertoli cells in culture assume a phenotype generally more similar to that of the in vivo differentiated cells. They are tall and columnar and contain organelles in a polarized location. In addition, spreading in culture is markedly enhanced when Sertoli cells are grown on various basement membrane substrates.
In an initial experiment, Sertoli cells from lO-day-old rats were cultured on plastic and on different ECMs, including laminin, Matrigel (a reconstituted basement membrane), and a synthetic laminin peptide containing the RGD peptide sequence. The objective of these studies was to investigate the effects of the ECM on follicle-stimulating hormone (FSH) responsiveness. Laminin and Matrigel enhanced the cyclic AMP (cAMP) response to FSH and cholera toxin, indicating modifications at the G protein levels (Dym et al. 1991). Sertoli cells grown on either of these two substrates responded to physiological levels of FSH (25-50 ng/ml) whereas pharmacological levels (500 ng/ml) were requi-
Extracellular Matrix Elements, Cell Adhesion Molecules 107
red for the cells grown on plastic or on the RGD containing laminin peptide (Fig. 5). Furthermore, immunoblotting of Sertoli cell plasma membranes with antibodies directed against the a subunit of the stimulatory G protein (Gsa) of adenylyl cyclase indicated that Sertoti cell culture on either laminin or Matrigel increased the amounts of Gs. These results were further confirmed by immunoprecipitating the Gsa protein from the particulate fraction of 13SS]methionine metabolically labeled Sertoti cells (Fig. 6). These data suggest that culture of epithelial Sertoli cells on basement membrane substrates enhances the Gs complex of adenylyl cyclase and the cAMP response to FSH, consistent with the more differentiated morphology and function of the cells. In a similar type of study, Ikeda and colleagues (1993) examined the effects of the ECM on the signal transduction by calciotropic hormones in osteoblastlike cells. They demonstrated that type J collagen increases calcitonin and parathyroid hormone receptor-mediated signal transduction in UMR 106-06 osteoblast-like cells. Their results demonstrate that an ECM component, in this case collagen J, may playa role in signal transduction affecting both receptors and G proteins in bone cells.
In order to evaluate whether calcium acts as a second messenger in laminin-mediated Sertoli cell spreading, we quantitated intracellular free calcium ICa2+]i in Sertoli cells by utilizing the fluorescent probe Fura-2AM (Ravindranath et a1. 1993). The average values for I Ca2+Ji increased two- to threefold over basal levels 1-2 h after plating cells on laminin, suggesting a role for calcium in Sertoti cell spreading. Addition of soluble laminin to Sertoli cells that were already spread did not result in changes in ICa2+]i. We next localized calcium reservoirs in Sertoli cells using the oxalate-pyroantimonate cytochemical technique described by Pillai and colleagues (1993). Calcium-containing precipitates localized primarily within the cytosol, mitochondria, and nucleus. Numerous large, round precipitates appeared in the euchromatin of the nucleus. The nucleolus and heterochromatin were devoid of any precipitates. In comparison to the nuclear precipitates, small, round precipitates were observed densely distributed in mitochondria and appeared to be associated solely with the inner mitochondrial membrane. An interesting observation was the presence of a single large precipitate within vesicular structures present in the cytoplasm. Some of these vesicles appeared to be closely associated with the endoplasmic reticulum; others were associated with the Golgi apparatus. The calcium localizati-
108 M.Dym
200
180
160 \ 0:: MATRIGEL 'f 0 M 140 ..... !!2 Qj <.)
'" 0 120
'" <1> "0 E 100 a. 0: ::2 « <.) 80
60
40
20
0 10" 104 0
Fig. 5. Effect of increasing concentrations of follicle stimulating hormone (FSH; abscissa) on Sertoli cell cyclic AMP content (ordinate). Ten-day old rat Sertoli cells were grown on top of different extracellular matrix substrates (Matrigel, laminin, RGD, plastic). At day 4 of the culture, cells were treated with the indicated concentrations of FSH in the presence of the phosphodiesterase inhibitor 3-isobutyl-l-methyl-xanthine (MIX) (0.25 mM). Thirty minutes after the addition of FSH, intracellular cAMP was extracted and measured by radioimmunoassay. The highest levels of cAMP were detected when the cells were grown on Matrigel or laminin. From Dym et al. (1991)
Extracellular Matrix Elements, Cell Adhesion Molecules 109
A B c o
-48000
-42000
Fig. 6. Immunoprecipitation of l.1SS]methionine membrane preparations from Sertoli cells cultured on plastic (lane 8), laminin (lane C), and Matrigel substrates (lane 0). The membrane preparations were immunoprecipitated with antibody to Gs protein (lanes 8 , C, 0) or with normal rabbit serum as a control (lan(' A). Densitometry of the gel revealed a 390% increase in the synthesis of Gs when cells grown on Matrigel (lane 0) were compared to cells grown on plastic (lan(' 8). From Dym et al. (1991)
on within the nucleus suggests that calcium may act as a signal transducer for gene transcription leading to synthesis of proteins by the cell. The vesicular structures which contained large precipitates could be calciosomes and may represent the exchangeable pool of calcium present in Sertoli cells. In another study in human endothelial cells, it was demonstrated that the fibronectin- and vitronectin-induced cellular spreading in culture triggers an elevation of intracellular free calcium (Schwartz 1993).
110 M.Oym
Certain protooncogenes such as c-fos and c-jun are believed to play regulatory roles in cell proliferation and/or differentiation (Muller et al. 1984). In NIH 3T3 cells stimulated by growth factors, there is a dramatic increase in c-fos mRNA within minutes (Greenberg and Ziff 1984). Conti and colleagues demonstrated that FSH, bovine basic fibroblast growth factor (FGF), and Sertoli cell FGF each induces a rapid increase in c-fos mRNA in cultured Sertoli cells (Smith et al. 1989). In a more recent publication, Norton and Skinner (1992) demonstrated that c-fos may act as a transcription factor in the P-Mod-S induction of Sertoli cell differentiation. In another report using PC12 cells stimulated by laminin, Yamada and his colleagues found a correlation between c-fos expression and the basement membrane induced cell division (Kubota et al. 1992). Thus, we examined the modulation of Sertoli cell c-fos in response to basement membrane components. Since in the Sertoli cells the basement membrane components have biological activity affecting differentiation, we postulated that the early response oncogenes may be turned on by the ECM. It is possible that this early response oncogene acts as a transcription factor to influence the basement membrane-induced differentiation of Sertoli cells by stimulating subsequent gene expression.
In an initial experiment, we confirmed the previous work of Conti and his colleagues and demonstrated that in our hands FSH also induced a marked and immediate increase in c-fos mRNA in Sertoli cells. In another experiment, we demonstrated that c-fos mRNA is modulated by laminin and by several of the biologically active laminin peptides. Two of these peptides (YIGSR, SIKV A V) induced Sertoli cell spreading within minutes of plating. In one experiment, Sertoli cells were incubated for 60 min on plastic dishes or on dishes coated with laminin or one of the laminin-derived biologically active peptides in the absence or presence of FSH. There was a dramatic increase in c-fos message when Sertoli cells were cultured on the peptides without FSH (Fig. 7). In the presence of FSH, the SIKV A V peptide enhanced the gene expression of c-fos quite dramatically. As an internal control, the same blot was probed for transferrin mRNA, and this transcript was also induced by several of the laminin-derived peptides although in a manner that is clearly different from the c-fos induction. Note that in this experiment, laminin appears to down regulate c-fos mRNA both in the presence or absence of FSH. These results demonstrating an immediate increase in
Extracellular Matrix Elements, Cell Adhesion Molecules 111
+ + + + +
L y s R c Fig. 7. The c-fos mRNA levels in Sertoli cells after I h of culture on various basement membrane components. The lop panel is a photograph of the ethidium bromide stained gel. The middle panel shows c-fos mRNA levels in Sertoli cells after culture on plastic (C); RGO (R); SIKY A Y (S); YIGSR (y); laminin (L) either in the presence (+) or absence (-) of FSH. Note that in the absence of FSH, the three laminin-derived peptides (RGO, SIKYAY, YIGSR) all induced significant levels of c-fos mRNA. In the presence of follicle-stimulating hormone (FSH), the SIKY A Y peptide stimulated the Sertoli cells to produce very large amounts of c-fos mRNA. As an internal control, the same blot was probed for transferrin mRNA (lower panel). Note that each of the peptides induced transferrin mRNA in the absence of FSH. FSH seemed to downregulate the basement membrane-induced increases in c-fos mRNA except in the presence of SIKYAY. The RNA was collected using a "mini" method (Chomczynski and Sacchi 19R7). From Oym (1994)
112 M.Dym
c-fos mRNA levels in Sertoli cells in response to basement membrane components suggest that this early response oncogene may serve as a transcription factor that ultimately leads to altered structural and functional changes.
In summary, the basement membrane underlying epithelial cells has been shown to be instrumental in biological function, specifically maintaining polarity and the differentiated state. In culture, Sertoli cell polarity and spreading are both enhanced by basement membrane components. Integrins and nonintegrin matrix receptors have been localized to the basal regions of the seminiferous epithelium and it is likely that these matrix binding proteins are instrumental in the basement membrane-induced signal transduction in Sertoli cells. G proteins, cAMP, and Ca2+ may be involved in the biochemical message transmitted to the nucleus and protooncogenes such as c-fos may be a nuclear transcription factor in the cellular response.
References
Adams JC, Watt FM (1993) Regulation of development and differentiation by the extracellular matrix. Development 117: 1183-1198
Bixby JL, Jhabvala P (1990) Extracellular matrix molecules and cell adhesion molecules induce neurites through different mechanisms. J Cell BioI III :2725-2732
Burridge K, Fath K, Kelly T, Nuckolls G, Turner C (1988) Focal adhesions: the extracellular matrix and the cytoskeleton. Annu Rev Cell Bioi 4:487-525
Burridge K, Turner CE, Romer LH (1992) Tyrosine phosphorylation of paxillin and pp125FAK accompanies cell adhesion to extracellular matrix: a role in cytoskeletal assembly. J Cell Bioi 119:893-903
Chomczynski P, Sacchi N (1987) Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 162:156-159
Davis CM, Papadopoulos V, Jia MC, Yamada Y, Kleinman HK, Dym M (1991) Identification and partial characterization of laminin binding proteins in immature rat Sertoli cells. Exp Cell Res 193:262-273
Dym M (1983) The male reproductive system. In: Weiss L (ed) Histology: cell and tissue biology. Elsevier Biomedical, New York, pp 1000-1053
Extracellular Matrix Elements, Cell Adhesion Molecules 113
Oym M (1994) Basement membrane regulation of Sertoli cells. Endocr Rev 15:102-115
Oym M, Lamsam-Casalotti S, Jia MC, Kleinman HK, Papadopoulos V (1991) Basement membrane increases G-protein levels and follicle-stimulating hormone responsiveness of Sertoli cell adenylyl cyclase activity. Endocrinology 128: 1167-1176
Greenberg ME, Ziff EB (1984) Stimulation of 3T3 cells induces transcription of the c-fos proto-oncogene. Nature 311 :433-438
Hadley MA, Oym M (1987) Immunocytochemistry of extracellular matrix in the lamina propria of the rat testis: electron microscopic localization. Bioi Reprod 37: 1283-1289
Hay EO (1981) Collagen and embryonic development. In: Hay EO (ed) Cell biology of extracellular matrix. Plenum, New York, pp 379-409
Hynes RO (1992) Integrins: versatility, modulation and signaling in cell adhesion. Cell 69: 11-25
Ikeda K, Michelangeli VP, Martin n, Findlay OM (1993) Type I col\agen substrate increases calcitonin and parathyroid hormone receptor-mediated signal transduction in UMR 106-06 osteoblast-like cel\s. J Cell Physiol 156:130-137
Jones PH, Watt FM (1993) Separation of human epidermal stem cells from transit amplifying cel\s on the basis of differences in integrin function and expression. Cel\ 73:713-724
Juliano RL, Haskill S (1993) Signal transduction from the extracellular matrix. J Cel\ Bioi 120:577-585
Kleinman HK, Weeks BS, Schnaper HW, Kibbey MC, Yamamura K, Grant OS (1993) The laminins: a family of basement membrane glycoproteins important in cell differentiation and tumor metastases. Vitam Horm 47:161-186
Kornberg LJ, Earp HS, Turner CE, Prockop C, Juliano RL (1991) Signal transduction by integrins: increased protein tyrosine phosphorylation caused by clustering of B 1 integrins. Proc Nat! Acad Sci USA 88:8392-8396
Kubota S, Tashiro K, Yamada Y (1992) Signaling site of laminin with mitogenic activity. J Bioi Chern 267:4285-4288
Lamoureux P, Steel VL, Regal C, Adgate L, Buxbaum RE, Heidemann SR (1990) Extracel\ular matrix allows PC 12 neurite elongation in the absence of microtubules. J Cell Bioi 110:71-79
Leblond CP, Inoue S (1989) Structure, composition, and assembly of basement membranes. Am J Anat 185:367-390
Lian G, Mil\er KA, Enders GC (1992) Localization and synthesis of entactin in seminiferous tubules of the mouse. Bioi Reprod 47:316-325
114 M.Dym: Extracellular Matrix Elements
Martin G, Timpl R (1987) Laminin and other basement membrane components. Annu Rev Cell Bioi 3:57-85
Muller R, Bravo R, Burckhardt J, Curran T (1984) Induction of c-fos gene and protein by growth factors precedes activation of c-myc. Nature 312:716-720
Norton IN, Skinner MK (1992) Regulation of Sertoli cell differentiation by the testicular paracrine factor PModS: potential role of immediate-early genes. Mol EndocrinoI6:2018-2026
Palombi F, Salanova M, Tarone G, Farini D, Stefanini M (1992) Distribution of I integrin subunit in rat seminiferous epithelium. Bioi Reprod 47: 1173-1182
Pillai S, Menon GK, Bikle DD, Elias PM (1993) Localization and quantitation of calcium pools and calcium binding sites in cultured human keratinocytes. J Cell PhysioI154:101-1 12
Ravindranath N, Vomberger W, Zitzmann D, Dym M (\993) Laminin-mediated Sertoti cell spreading. A possible role for calcium. J Cell Bioi 567:34 (abstract)
Runyan RB, Potts JD, Sharma RV, Loeber CP, Chiang JL, Bhalla RC (1990) Signal transduction of a tissue interaction during embryonic heart development. Cell Regul 1:30 I -313
Schwartz MA (\993) Spreading of human endothelial cells on fibronectin or vitronectin triggers elevation of intracellular free calcium. J Cell Bioi 120:1003-1010
Smith EP, Hall SH, Monaco L, French FS, Wilson EM, Conti M (1989) A rat Sertoli cell factor similar to basic fibroblast growth factor increases c-fos messenger ribonucleic acid in cultured Sertoti cells. Mol Endocrinol 3:954-961
Streuli CH, Bailey N, Bissell MJ (1991) Control of mammary epithelial differentiation: basement membrane induces tissue-specific gene expression in the absence of cell-cell interaction and morphological polarity. J Cell Bioi 115:1383-1395
Weeks BS, DiSalvo J, Kleinman HK (1990) Laminin-mediated process formation in neuronal cells involves protein dephosphorylation. J Neurosci Res 27:418-426
Werb Z, Tremble PM, Behrendtsen 0, Crowley E, Damsky CH (\989) Signal transduction through the fibronectin receptor induces collagenase and stromelysin gene expression. J Cell Bioi 109:880-885
6 Interactions Between Androgens, Sertoli Cells and Germ Cells in the Control of Spermatogenesis
R.M. Sharpe, C. McKinnell, T. McLaren, M. Millar, T.P. West, S. Maguire, J. Gaughan, V. Syed, B. Jegou, J.B. Kerr, and P.T.K. Saunders
6. I Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. I 15 6.2 Subtraction Hybridization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. I It) 6.2.1 Methodology and Approach. . . . . . . . . . . . . . . . . . . . . . . . . . . . .. I It)
6.2.2 Identification of Subtracted Clones. . . . . . . . . . . . . . . . . . . . . . . .. 122 6.3 Effects of Testosterone Withdrawal on Protein Secretion
by Isolated Seminiferous Tubules. . . . . . . . . . . . . . . . . . . . . . . . .. 125 6.3.1 Methodology and Approach. . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 125 6.3.2 Androgcn-Dependent Changes in Total Protein Secretion. . . . . .. 126 6.3.3 Analysis of the Pattern of Protein Secretion. . . . . . . . . . . . . . . . .. 132 6.4 Discussion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 135 Rcferences ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 140
6.1 Introduction
One of the major unanswered questions in andrology is 'How does testosterone control spermatogenesis'?' As testosterone is all-important for maintenance of spermatogenesis in all species of mammals that have been studied, answering this question is likely to give us major insight into the basic mechanisms involved in the control of spermatogenesis. Four years ago we initiated a major programme of work which sought to determine to the molecular level how testosterone controlled spermato-
116 R.M. Sharpe et al.
genesis. This paper summarizes the results of these studies including the surprises, the disappointments and the achievements. The first important decision we took was deciding how to answer the question. A survey of the literature (reviewed in Sharpe 1994) made it plain that testosterone appeared to exert little, if any, effect on the function of isolated Sertoli cells, irrespective of whether these were isolated from immature (15-21 days of age) or adult (hypophysectomized) rats. We chose to interpret this lack of information as evidence that these studies had used inappropriate approaches to answering the question, a decision which has proved to be 100% correct (not all of our thinking has turned out to be so accurate!). The available data shows clearly that Sertoli cells from immature rats are primarily follicle-stimulating hormone (FSH)- rather than testosterone-responsive, so the absence of any well-described effects of testosterone in this situation was predictable (Sharpe 1994). Where Sertoli cells had been isolated from adult rats, the latter had been hypophysectomized 3 or more weeks earlier so as to facilitate Sertoli cell isolation (by removal of most of the germ cells) as well as inducing major testosterone-withdrawal. In light of the growing evidence for germ cell modulation of Sertoli cell function (reviewed in Jegou 1993 and Sharpe 1993), we concluded that perhaps a normal germ cell complement was required for testosterone to act on the Sertoli cells -another conclusion that we have since proved to be correct.
So what positive information did we have to go on? In fact, there was really only one positive lead and that was the observation that, following testosterone withdrawal, seminiferous tubules (ST) at stage VII (and perhaps VIII) of the spermatogenic cycle were the first to show abnormal morphological changes, namely the appearance of a few degenerating germ cells at 3-4 days post-treatment. These observations had been made originally in 1977-1981 (Russell and Clermont 1977; Russell et al. 1981) and then ignored. We had confirmed these findings using the new method of ethane dimethane sulphonate (EDS)-induced Leydig cell destruction as the method of inducing testosterone withdrawal (Bartlett et al. 1986; Sharpe et al. 1988, 1990). We went on to characterize the time-frame of germ cell degeneration at stages VII -VIII and to quantify the number of germ cells affected (Sharpe et al. 1992; Kerr et al. 1993). These findings proved to be somewhat surprising (Fig. 1). They showed clearly that stages VII-VIII were the first stages affected but that germ cell degeneration only increased moderately from 3-8 days after EDS-
Interactions Between Androgens, Sertoli Celis
None observed up to Day B
I I I
Day 3
Day Day 4 4-5
117
Fir Day d 5-6
I
51 time of appearance 01
e~rl~~r~~~i~~~e"J"S estosterone withdrawal
@ ®. ~:@ c@:@{:(ij),:(j):ilii~I~~·~ @@Iiti) ~1~1®1@j)1(jJ 1@ ~I®~ ~ &aJ On onl In 9 1 ~el Pil "I tl I ~z z P
A. ~\ C@Al @A, i2)AJ@Aj(@AJ@AJ@01~2@A~@A2~A3@A3 , • I 1 ~] _ ' , ] t! J
II III IV V VI VII VIII IX X XI XII XIII XIV
Stage of the cycle
o Maximum degree of germ cell degeneration observed up to Day 8 after EDS·treatment
Fig. L Diagrammatic summary of the temporal changes in stage-dependent germ cell degeneration which occurs following ethane dimethane sulphonate (EDS)-induced testosterone withdrawal in the adult rat. The time at which degenerating germ cells first appear at a particular stage is indicated by the verti· cal columns and the degree of degeneration/loss of individual germ cell types at day 8 after EDS treatment is indicated by the degree of" shading of the germ cells . Note that stages VlJ-VIII are the first to show germ cell degeneration after EDS treatment but that later germ cell degeneration at stages X-XII is far more pronounced than at stages VII-VIII. Roman numerals indicate stages. Based on data in Kerr et al. (1993) and unpublished data
induced testosterone withdrawal. In contrast, stages X-XII, which showed little or no genn cell degeneration up to 5 days after EDS-treatment showed catastrophic loss of genn cells by day 8 with 80% or more of spennatids missing or degenerative (Figs. 1,2).
Our interpretation of these findings was that genn cells which passed through stages VII-VIII in the absence of testosterone failed to undergo
118 R.M. Sharpe et al.
Fig. 2. Representative changes in the cross-sectional appearance of seminiferous tubules (ST) at stage VII from a control rat (A), a rat treated 4 days earlier with ethane dimethane sulphonate (EDS) (8) or with EDS + testosterone (EDS + TE; C). Also shown (D) is the appearance of a ST at stage XII in a rat treated 8 days earlier with EDS to illustrate the almost complete loss/degeneration of step 12 spermatids. Arrows indicate degenerating germ cells, which are not evident in control rats or rats treated with EDS + TE
Interactions Between Androgens, Sertoli Cells 119
certain testosterone-dependent changes which were prereqUlsltes for further changes when these cells had progressed to stages X-XII (Sharpe 1994). As yet we have not produced any evidence which contradicts this interpretation (see below), but there remains a possibility that testosterone withdrawal could have delayed effects on the function of Sertoli (or germ) cells at stages X-XII. However, based on these findings, we decided to concentrate our effort on stages VI-VIII at 4-6 days after EDS-induced testosterone withdrawal. This was a period when we surmised that maximum 'switch-off' of androgen-regulated genes would have occurred at stages VI-VIII with minimum secondary changes resulting from germ cell loss, as the germ cell complement was 95% normal at these times. Two separate but complementary approaches were used: (l) subtraction hybridization to identify androgen-regulated genes and (2) two-dimensional sodium dodecyl sulphate-polyacrylamide gel electrophoresis (2D SDS-PAGE) to identify androgen-regulated proteins secreted by Sertoli cells. The results obtained from these two approaches are described separately below, but it is emphasized that they were operated in parallel with the intention that one or other (or both) would lead us to the androgen-regulated genes which we presumed were switched on specifically at stages VI-VIII.
6.2 Subtraction Hybridization
6.2.1 Methodology and Approach
The basic strategy employed for subtraction hybridization is shown as a flow diagram in Fig. 3 and full details of the methodology are described elsewhere (Saunders et al. 1993). There were three separate phases of 'subtraction' which employed different starting materials (phase I, whole testes; phase II, isolated ST at stages VI-VIII; phase III, isolated cells enriched in Sertoli cells) but all utilized the same principles. The underlying presumption was that testosterone controlled spermatogenesis by regulating the transcription of a small number of key genes, analagous to how androgens control prostatic cell growth and function (Parker et al. 1980). Thus, if the effects of testosterone on spermatogenesis involved switching certain genes' on' or 'off', then this might lead to major differences in the number of messenger RNA (mRNA) transcripts for
120 R.M. Sharpe et al.
Strategy for identifying Androgen-regulated genes in the adult rat testis
* * * Control EDS EDS + TE
~ ~ or Control
+ Isolate whole testes (Phase I) seminiferous tubules at stages VI - VIII of the spermatogenic cycle (Phase II) or isolated Sertoli cells (Phase III) and extract mRNA
make cDNA library
\
make Sing!-stranded radiolabelled cDNA
~ subtract cDNA from EDS + TE or
Control (Phases I and II) or cRNA (Phase III) with 10-fold excess
of mRNA from EDS-treated group
+ amplify using PCR
Screen library using ........ 1-_____ ---11 subtracted probes
+ Isolate positive plaques
+ Re-screen
+ Recover into plasm ids and purify
+ Sequence
Fig. 3. Flow diagram illustrating the basic principles and procedures of the three phases of subtraction hybridization which were used in attempts to isolate androgen-regulated genes. peR, polymerase chain reaction; EDS, ethane dimethane sulphonate; EDS + TE, EDS + testosterone; eDNA, complementary DNA; eRNA, complementary RNA
Interactions Between Androgens, Sertoli Cells 121
such genes when comparing the normal' androgens on' situation with an experimental 'androgens off' situation. It has been described above how the period of 4-6 days after EDS treatment was identified as the likely time when androgen-regulated genes would be maximally 'switched off', but when secondary changes in transcription of other genes, because of germ cell loss, would have been minimal. To perform subtraction hybridization, single-stranded, radiolabelled, complementary DNA (cDNA)was synthesized from mRN As isolated from control tissue. This pool of cDNAs should include representatives of all genes which are being actively transcribed, including any androgen-regulated genes. These cDNAs were then 'subtracted' with a tenfold excess of pooled biotinylated mRNAs isolated from tissue from rats at 4-6 days after EDS treatment. The latter should contain mRNAs for all the non-androgen-regulated genes which should hybridize to their cDNA from control tissue, forming duplexes which are then removed together with the excess single-stranded mRNA. In theory, the remaining single-stranded cDNA will be enriched for androgen-regulated genes. Little cDNA remains after subtraction and once made double-stranded this was amplified by 'lone-linker' polymerase chain reaction (peR) and radiolabelled and used to screen a reference cDNA library. (Fig. 3). In practice, the actual 'subtraction' of cDNA with excess mRNA is usually performed twice in order to increase the chances of removing all of the highly expressed transcripts which are common to the 'on' and 'off' situations.
There have been three phases to our subtraction hybridization strategy. The first involved generating cDNA from mRNA isolated from the whole testes of rats treated with EDS+ testosterone esters (TE) on day 6 and then subtracting this with excess mRNA isolated from whole testes of rats treated 6 days earlier with EDS alone (Fig. 3). The second involved generation of cDNA from mRNA isolated from ST at stages VI -VIII from control rats and subtracting this with a tenfold excess of mRNA isolated from tubules at the same stages from rats treated 6 days earlier with EDS. Around 20 subtracted clones were generated from phases I and II and then evaluated using northern and in situ hybridization with samples from three treatment groups, namely control, EDS+6 days and EDS+ TE+6 days. Northern analysis was found to be far less informative than in situ hybridization, though this requires correctly-fixed tissue for optimum results (Millar et al. 1993). After in situ hybridization of individual cloned cDNAs, microscopic analysis
122 R.M. Sharpe et al.
allows determination of (a) the cellular site(s) of expression, (b) the stages of the spermatogenic cycle at which the gene is expressed, and (c) whether expression is androgen dependent. However, to be able to make definitive conclusions regarding these three endpoints it is essential to use testicular sections from animals which have been perfusion-fixed with Bouin's (Millar et al. 1993), as other preparative methods are vastly inferior and can yield misleading results.
6.2.2 Identification of Subtracted Clones
Of the 20 clones generated in phases I and II, no expression could be detected for three whilst six showed no signs of stage specificity or androgen dependence and were considered to be housekeeping genes; they were therefore not studied further. Ten of the clones recognized a transcript with an identical size on northern blots and with a similar pattern of expression on tissue sections (Fig. 4) which turned out to be mitochondrial cytochrome oxidase-II (COX-II). This gene was expressed very strongly near the base of the seminiferous epithelium and was particularly pronounced at stages VII-XII. Closer analysis, including cell depletion studies, indicated that COX-II was expressed very strongly in pachytene spermatocytes (Saunders et al. 1993). However, expression of COX-II was not androgen dependent and it was concluded that this clone was isolated because of its high level of expression. In this regard, it is noteworthy that subtraction hybridization approaches for many tissues have reported isolation of the mitochondrial genes for cytochrome oxidase.
One subtracted clone did show a marked stage-dependent pattern of expression by in situ hybridization (Fig. 4) with expression commencing at stage VII and terminating at stage XIII. However, expression of this gene was clearly in round/elongating spermatids and not in Sertoli cells (Fig. 4). Sequence analysis confirmed that this transcript had homology to transition protein-2 (TP-2), which plays an intermediary role during the replacement of histones by protamines in the chromatin during condensation of the spermatid nucleus. We were interested in this gene for two reasons. First, it was switched on at stage VII (i.e. the androgen-dependent phase; Saunders et al. 1992) and, second, when spermatogenesis fails following testosterone withdrawal, this failure
Interactions Between Androgens, Sertoli Cells 123
Fig. 4 A-D. Examples of in situ hybridization using digoxigenin-Iabelled ribonucleotide probes to two of the genes which were isolated by subtraction hybridization. A, B Mitochondrial cytochrome oxidase-II (COX-II). C, D Transition protein-2 (TP-2). Photomicrographs taken using low magnification (A, C) and at higher power magnification (8, D). Sites of hybridization are indicated by the dark staining
124 R.M. Sharpe et al.
stems primarily from the massive degeneration of the condensing spermatids at stages X-XII which are expressing TP-2 (see Fig. 1; Kerr et al. 1993). However, detailed analysis by both northern and in situ hybridization failed to demonstrate any androgen dependence for the expression of TP-2 (Saunders et al. 1992), and our more recent studies have shown that the same is true for TP-l.
Therefore, irrespective of whether whole testes (phase I) or isolated ST at stages VI-VIII (phase II) were used as the source of mRNA for subtraction hybridization, it was clear that the subtracted clones which were generated were primarily directed at COX-II, housekeeping genes or heavily expressed germ celI transcripts (TP-2). No specific Sertoli cell transcripts were isolated using this approach. Therefore, as our conviction was that regulation of spermatogenesis by testosterone at the transcriptional level had to occur via effects on the Sertoli cell, we instigated a phase III subtraction hybridization which aimed to avoid cloning of heavily expressed germ cell transcripts. Phase III was centred around directional cloning of cDNAs synthesized from mRNA extracted from a Sertoli cell-enriched, isolated cell preparation (West et al. 1994) obtained from control rats and from rats treated 6 days previously with EDS. Single-stranded cDNAs from controls were then subtracted with an excess of biotinylated cRNA generated from the directionally cloned cDNAs from Sertoli cells from EDS-treated rats. Thereafter, the basic protocol utilized similar procedures to those used for phases I and II (Fig. 3). To date, the phase III subtraction has yielded more than 20 cloned genes. About 30% of the clones have a very low level of expression when analysed by northern analysis, whereas a further 30% appear to be directed to a single abundant transcript (approx 1.5 kDa). The remaining clones are still under evaluation but appear still to include germ cell as well as Sertoli cell transcripts.
Viewed in retrospect, the results of the subtraction hybridization can be interpreted in two rather different ways. The first is that the subtraction procedures are fundamentally incapable of effectively subtracting out strongly expressed genes such as COX-II and TP-2 and that, as a result, there is only a remote chance that androgen-regulated Sertoli cell genes will be identified against this background 'noise'. The second interpretation is more positive and concludes that there are no highly expressed, androgen-regulated Sertoli cell transcripts to be detected and hence all that will be detected will be the background 'noise'. These two
Interactions Between Androgens, Sertoli Cells 125
possible conclusions are not mutually exclusive but at present there is insufficient evidence to say which is more likely to be correct. However, results from the study of androgen-regulated proteins (described below) could be interpreted as favouring the second of these two possibilities, so it is appropriate to now consider these data.
6.3 Effects of Testosterone Withdrawal on Protein Secretion by Isolated Seminiferous Tubules
6.3.1 Methodology and Approach
The basic approach used for these studies was to isolate from control and treated animals, ST in long lengths (> I cm) at three different stage groupings, namely II-V, VI-VIII and IX-XII. This choice was based partly on practical expedience, in that these three stage groupings are the easiest to distinguish from each other, whilst the omission of stages XIII-I aided in ensuring that stages hefore and after the presumptive androgen-dependent stages (VI-VIII) were separated unequivocally. However, an equally important reason for choosing these stage groupings was to ensure that long lengths of ST could be isolated for culture. Most of the previous work on stage-dependent changes in ST function have utilized small fragments of a few millimetres in length (see Parvinen et al. 1986). Our earlier studies on the secretion of immunoactive inhibin by isolated, unstaged ST had shown unequivocally that only ST isolated in > I cm lengths allowed faithful reflection of the in vivo situation and maintenance of responsiveness to FSH (Allenby et al. 1991 a,b). In addition, we reasoned that isolation of long lengths of ST might enable better maintenance of the nonnal internal milieu of the ST for at least a proportion of its length. Especial care was taken to avoid stretching or distortion of the ST being isolated for culture so as to minimize artefactual changes resulting from mechanical damage.
In all of the studies described below, a standard experimental protocol was used, and this has been described in detail elsewhere (Sharpe et al. 1992). In brief, a total of 10 cm ST at each of the stage groupings was isolated from individual animals and cultured at 34°C for 22 h in a simple, methionine-free, defined medium supplemented with l50/.lCi/ml [3SS Jmethionine. After removing exfoliated germ cells by
126 R.M. Sharpe et al.
centrifugation, the culture medium was then assessed for its total concentration of [35S1methionine-Iabelled proteins (= total secreted proteins) and the pattern of secretion of individual proteins was analysed by 2D SDS-P AGE. It had always been our intention to focus on secreted proteins as we had reasoned that communication between the Sertoli and germ cells, and vice versa, was likely to occur mainly via secreted proteins. This view has been supported by our analysis of incorporation of radiolabelled methionine into intracellular proteins (i.e. mainly nonsecreted proteins) in the experiments described below, which has failed to detect any significant change in any of the experimental situations described.
6.3.2 Androgen-Dependent Changes in Total Protein Secretion
The first major, and entirely unexpected, finding was that isolated ST at stages VI-VIII secreted about twice as much total protein as did stages II-V or IX-XII, based on the incorporation of [35Sjmethionine (Fig. 5). This finding has proved to be remarkably repeatable and, according to our current thinking, is probably the most important of our findings (discussed below). It is equally clear that this stage-dependent increase in secretion of total proteins is absolutely testosterone dependent, as it is abolished completely within 4 days of EDS administration (Fig. 5). Indeed, even by day 3 after EDS treatment there is a substantial diminution of this increase (see Sharpe et al. 1992) and this is the earliest point at which any morphological signs oftestosterone withdrawal are evident at stage VII (see Russell and Clermont 1977; Bartlett et al. 1986; Sharpe et al. 1990; Kerr et al. 1993). The fact that only a minor proportion « 1 %) of germ cells at stages VII -VIII are degenerati ve at 4 days after EDS-induced testosterone withdrawal argues strongly that abolition of the normal increase in protein secretion at stages VI-VIII at this time is a primary consequence of testosterone withdrawal and not a secondary consequence of major germ cell loss (see also below). Administration of long-acting TE to EDS-treated rats from day 0, in a dose known to be capable of maintaining quantitatively normal spermatogenesis and fertility (Sharpe et al. 1988), maintained the normal stage-dependent pattern of ST protein secretion (Fig. 5) as well as preventing the appearance of any degenerating germ cells (Fig. 2). Comparable changes to those
Interactions Between Androgens, Sertoli Cells 127
induced by EDS are also observed after immunoneutralization of luteinizing hormone (LH), demonstrating that the effects observed are due to testosterone withdrawal and not to some obscure toxic effect of the EDS (Sharpe et al. 1992).
6.3.2.1 Effect of Germ Cell Depletion The increase in total protein secretion by ST at stages VI-VIII was puzzling because there is no evidence for any major differences in androgen receptor content of Sertoli cells at these stages (Parvinen 1993); nor is there any good evidence that testosterone levels in or around ST at these stages are any different (Sharpe 1994). So, other than testosterone, what dictates whether the increase in protein secretion at stages VI-VIII occurs? We surmised that it must be the change in germ cell complement, as this was the only factor that was clearly different at stages VI-VIII when compared with earlier and later stages. This conclusion was reinforced by the growing evidence that stage-dependent changes in Sertoli cell function all appear to be regulated by the changing germ cell complement (Parvinen 1993; Sharpe 1993). We therefore assessed whether specific depletion of either pachytene spermatocytes, round (steps 6-8) spermatids or elongate (steps 18-19) spermatids a/one from ST at stages VI-VIII would attenuate the normal stage-dependent increase in protein secretion. For these studies we utilized the chemical methoxyaceticacid (MAA) in a single oral dose of 650 mgikg (see Bartlett et al. 1988). At this dose it results in rapid degeneration of 80%-100% of pachytene and later spermatocytes at most stages of the spermatogenic cycle with few, if any, other discernible effects. As the kinetics of spermatogenesis are not affected by MAA treatment, it is possible to pick times for ST isolation when either pachytene spermatocytes (MAA + 4 days), round spermatids (MAA + 18 days) or elongate spermatids (MAA + 30 days) are depleted at stages VI-VlII but the remaining germ cell complement is normal (see McKinnell and Sharpe 1992; Sharpe et al. 1993a).
To our surprise, depletion of anyone of the adluminal germ cells from ST at stages VI-VIII virtually abolished the normal increase in total protein secretion that should have occurred at these stages. Indeed, these treatments were nearly as effective as EDS-induced testosterone withdrawal itself (Fig. 5), despite the fact that there is no measurable change in testosterone levels at any point after MAA treatment (Bartlett
~
E " a
4()
()()
()
3()(
)()(
)
2()
()()
()
E
1()()
()()
a. "
It) a :::
Q
) c ·c
o :c
Q) E
en ~ o c o ~ ~ 8 E
A 0
5000
0
4()
()()
()
3()(
)()(
)
2()(
)()(
)
1()(
)()(
)
c 0
1
T
Elle
et
of
pre
tre
atm
en
t w
ith
E
OS
or
EO
S+T
E
a
II-V
V
I-V
III
D C
OIlI
,ol
o E
DS
•
ED
S+T
E
IX-X
II
Sta
ges
of t
he s
perm
atog
enic
cyc
le
Pre
tre
atm
en
t w
ith
NB
a r-fi
D
C""
"ol
ON
B
* -
t--=-
l II
~ I -
I
rr II-
V
VI-
VIII
IX
-XII
Sta
ges
of
the
sper
mat
ogen
ic e
ycle
8()(
)()(
)
6000
0
4()(
)()(
)
2()(
)()(
)
Elle
et
of
ge
rm c
ell
de
ple
tio
n
wit
h I
wit
ho
ut
EO
S-t
rea
lme
nl
ST
AG
ES
VI-
VIII
B
o! I
I)
),
4()
()()
()
3()(
)()(
)
2()(
)()(
)
1()(
)()(
)
o 0
C""
lro
i E
DS
·P
S
·AS
·E
S
·PS
·P
S
·AS
E
DS
Tre
atm
ent a
nd
lor g
erm
ce
lilyp
e d
eple
ted
Pre
tre
atm
en
t w
ith
m-O
NB
a T
1
D C
onlro
l D
mD
NB
* T
T
1
1 T
1
T
1
Ii-V
V
I-V
III
IX-X
li
Sta
ges
01 t
he s
perm
atog
enic
cyc
le
4()
()()
()
Pre
tre
atm
en
t w
ith
he
al
3()(
)()(
)
a D
Con
lrol
,. ...
OH
oal
*
ill T
rl1 1
2()(
)()(
)
1()(
)()(
)
E 0
II-
V
VI-
VIII
IX
-XII
Sia
ges
of t
he s
perm
atog
enic
cyc
le
I\.)
(X
l
;n s:: (f
) :::
T OJ
-a (!) ~ ~
Interactions Between Androgens, Sertoli Cells 129
et al. 1988; Allenby et al. 1991 b). It is equally clearthat the reduction in total protein secretion by these germ cell-depleted ST could not be simply the consequence of removing the protein secretory contribution of the germ cell type which had been depleted, as this would amount to far more than the total protein secreted by intact ST at stages VI-VIII (see Fig. 5). This conclusion was reinforced by three pieces of data. First, it was demonstrated that depletion of both pachytene spermatocytes and round spermatids from ST at stages VI-VIII caused only a marginally larger decrease in total protein secretion than did depletion of either genn cell type alone (Fig. 5) (see Sharpe et al. 193a). Second, depletion of pachytene spermatocytes combined with the administration of EDS 4 days earlier had no greater effect on protein secretion by ST at stages VI-VIII than did administration of EDS alone (Fig. 5). The third piece of information concerned direct comparison of the secretion of [35S [methionine-labelled proteins by isolated round spermatids and pachytene spermatocytes. These cells were isolated from the testes of
Fig. SA-E. Effect of variou~ experimental manipulations on the stage-dependent pallern of secretion of :bS-labellcd proteins by isolated seminiferous tubules (ST) cultured for 22 h in vitro in the presence of ,35Sjmethionine. Adult rats were treated in vivo with either ethane dime thane sulphonate (EDS) ± testosterone (TE) replacement and ST isolated 4 days later (A), or were treated with methoxyaceticacid and ST at stages VI-VIII isolated at various time points after treatment when either pachytene spermatocytes (-PS), round spermatids (-RS) or elongate spermatids (-ES) were depleted selectively (8). In other instances, two types of germ cells were co-depleted (-PS, -RS) or EDS-treatment (+ 4 days) and pachytene spermatocyte depletion were combined (EDS, -PS; 8). In further experiments (C-E), ST at the different stages were isolated from rats pretreated 24 h earlier with either nitrobenzene (NS; 300 mg/kg; C) or meta-dinitrobenzene (m-DNS; 50 mg/kg; D) or which had been exposed 24 h earlier to local testicular heating (43°C for 30 min; E). In all instances the data have been expressed as the cpm [35S]methionine-Iabelled protein secreted into the culture medium by a total of 10 cm ST at the indicated stages. Data are the mean ± SD for ST from three rats per group and representative experiments are illustrated. a Indicates a significantly (p 0.01 or p 0.00 I ) higher incorporation of radiolabel into proteins secreted by ST at stages V 1-Vlll than those at II-V or IX -XII from the same group. * Indicates a significant (ji 0.01 or p 0.(01) reduction in incorporation of radio label into STsecreted proteins when compared with the appropriate control group. The illustrated data has been adapted from a number of studies cited in the text
130 R.M. Sharpe et al.
adult animals (Meistrich et al. 1981; Onoda and Djakiew 1991) and the number of these cells assessed to be present in 10 cm ST (Wing and Christensen 1982) were then cultured for 22 h with radiolabelled methionine under conditions directly comparable to those used for the isolated ST, except that lactate and pyruvate were added to the culture medium. Under these conditions, the isolated round spermatids (7.5x106 cells per millilitre) secreted 10-20 times as much radiolabelled protein as did the pachytene spermatocytes (2.25x106 cells per ml); both cell types exhibited 93% viability after culture. Although these isolated germ cells would not have derived exclusively from ST at stages VIVIII, it seems reasonable to conclude that pachytene spermatocytes and round spermatids at stages VI-VIII are likely to differ considerably in their level of protein secretion, yet depletion of either cell type from ST at stages VI-VIII caused a comparable reduction in protein secretion (Fig. 5). From these studies we had to conclude that, in addition to normal levels of testosterone, afull and complete germ cell complement is a prerequisite for the normal androgen-dependent increase in ST protein secretion at stages VI-VIII.
6.3.2.2 Effect of Treatments Which Impair Spermatogenesis A number of other studies have served to highlight the potential importance of the androgen-driven, stage-dependent increase in ST protein secretion at stages VI-VIII. First, pretreatment of adult rats with either of two known Sertoli cell toxicants, meta-dinitrobenzene (mDNB) or nitrobenzene (NB), results within 24 h in abolition of the normal increase in protein secretion at stages VI-VIII (Fig. 5; McLaren et al. 1993a). This effect occurs in the absence of any detectable change in testosterone levels (AlIenby 1990) and is far more rapid than the change in protein secretion observed after testosterone withdrawal (Sharpe et al. 1992). Moreover, addition of m-DNB or NB during culture to ST isolated from untreated control rats is also able to reduce protein secretion by ST at stages VI-VIII (McLaren et al. 1993a). The effects of m-DNB and NB administration in vivo are associated at 24 h post-treatment with the degeneration/loss of pachytene spermatocytes (Allenby 1990; McLaren et al. 1993a), raising the possibility that it is the loss of these cells which prevents the normal increase in protein secretion by ST at stages VIVIII. Though this possibility cannot be excluded, studies at 12 h after treatment with m-DNB or NB have demonstrated that protein secretion
Interactions Between Androgens, Sertoli Cells 131
by ST at stages VI-VIII is already reduced (but not maximally), whereas degenerating pachytene spermatocytes are not yet evident (McLaren et al. 1993a).
Local testicular heating (43°C for 30 min) is known to cause major disruption of spermatogenesis, and though the mechanism of this effect is really not known (see Jegou et al. 1984), it does not appear to involve any major change in testosterone levels (Jegou et al. 1984; Bartlett and Sharpe 1987). We assessed whether local heating of the testis might affect ST protein secretion and found that at 24 h after treatment, the normal stage-dependent increase in protein secretion by ST at stages VI-VIJI was abolished completely (Fig. 5) and that most pachytene spennatocytes at stages VII-XIJI were degenerating or absent (McLaren et al. 1994). Remarkably, when ST were isolated at 4 h after local testicular heating, instead of observing attenuation of the increase in protein secretion by ST at stages VI-VIII, it was observed that the normal increase was douh/cd compared with controls (McLaren et al. 1994). No changes in ST protein secretion were observed at stages II-V or IX-XII at either 4 or 24 h after heat treatment. We do not have a detai led explanation for these observations but our current interpretation is that local testicular heating amplifies basic metabolic and enzyme activity and that the enhancement of protein secretion at 4 h at stages VI-VIII is evidence that these processes are activated already, presumably by testosterone. The heat-enhanced metabolic activity at stages VI-VIII cannot, however, be sustained and the subsequent reduction to baseline levels of protein secretion (i.e. levels observed under control conditions at stages II-V and IX-XIII) at 24 h after heat-treatment (Fig. 5) is a reflection of this. These observations could explain some of the adverse effects of local testicular heating on spermatogenesis.
From the studies described above it is not entirely clear as to whether abolition by m-DNB, NB and local testicular heating of the normal stage-dependent increase in protein secretion at stages VI-VIII is a cause or a consequence of the loss of pachytene spermatocytes (see Fig. 5). Time-course studies favour the former of these possibilities and, if this is correct, then it would have to be concluded that one or more events in the pathway of testosterone action on spermatogenesis are extremely susceptible to perturbation.
132 A.M. Sharpe et al.
6.3.3 Analysis of the Pattern of Protein Secretion
For the analysis, a constant amount of ST-secreted, [35S]methionine-labelled protein (approx. 300 000 cpm) was subjected to 2D SDS-PAGE using procedures described elsewhere in detail (Sharpe et al. 1992), Because comparable counts per minute of radio labelled protein were run on each gel, straightforward comparison of the autoradiographs does not take account of the major differences in overall level of protein secretion between groups, as shown in Fig. 5. Ideally, quantitative data for each protein of interest is required but there are a number of technical limitations which have made this difficult, except for a small number of proteins.
The main limitation in use of 2D SDS-PAGE cited by most researchers is its reproducibility. We have spent considerable effort in optimizing and standardizing procedures for 2D SDS-PAGE so as to maximize reproducibility. In addition, in describing the main protein changes observed, two other factors should be kept in mind. First, only a limited number of proteins have been selected for comparison. This selection was based on major reproducible changes being observed in three or more gels using material from different animals. Second, selection of putative androgen-regulated proteins (ARP) was also based on stage-dependent differences in their secretion, i.e. they were secreted more abundantly at stages VI-VIII than at earlier or later stages (see Sharpe et al. 1992),
Secretion of the three main products of the Sertoli cell, sulphated glycoproteins -1 and -2 (SGP-l and SGP-2) and cyclic protein-2 (CP-2), remained more or less constant through all of the treatment regimens described below, though there are one or two specific exceptions which have been detailed elsewhere (Sharpe et al. 1993b), EDS-induced testosterone withdrawal resulted in major decreases in the relative secretion of six proteins (ARP-l to ARP-4, ARP-6 and -7) and the appearance of one protein (ARP-5) which is thought to be a low pI isoform of SGP-2 (Fig. 6). Supplementation of EDS-treated rats with TE reversed all of these changes (Fig. 6), Another finding of considerable significance was that secretion of these ARPs was absent or barely detectable when (unstaged) ST were isolated from 28-day-old rats (Fig. 6), whereas they were clearly detectable when unstaged ST were isolated from adult rats and put into culture (McLaren et al. 1993b). These findings suggest that
Interactions Between Androgens, Sertoli Cells 133
secretion of the ARPs is age dependent as well as being androgen and stage dependent.
When each of the main germ cell types was depleted from ST at stages VI-VIII using MAA, major changes were observed in the secretion of some of the ARP's (McKinneli and Sharpe 1992). The most dramatic change was the virtual disappearance of ARP-2 and a more variable decrease in abundance of ARP-I when round spermatids were depleted (Fig. 7), whereas the relative abundance of these two proteins was unaffected by the depletion of pachytene spermatocytes or elongate spermatids (McKinneli and Sharpe 1992). This was interpreted as probably indicating that absence of round spermatids had altered secretion of ARP-I and ARP-2 by the Sertoli cell, although there was the less likely possibility that the two proteins actually derived from round spermatids. Surprisingly, the latter proved to be the case as analysis of radiolabelled proteins secreted by isolated round spermatids in culture (see above) showed unequivocally that ARP-I and ARP-2 were normal secretory products of round spermatids, though probably not of pachytene spermatocytes (Fig. 7). This finding was unexpected because the prevailing opinion is that testosterone cannot act directly on germ cells but only via the Sertoli (or peritubular) cells (reviewed in Sharpe 1994). As we still do not know the identity of ARP-I and ARP-2, it is not possible to assess whether their secretion is primarily androgen-regulated (and if so, how) or whether the reduction in their secretion following EDS-treatment is secondary to some earlier change in Sertoli cell secretory function.
ARP-3 and ARP-4 provide a marked contrast with ARP-I and ARP-2. Neither of these proteins were major secretory products of isolated pachytene spermatocytes or round spermatids (Fig. 7). Depletion of pachytene spermatocytes from ST at stages VI-VIII using MAApretreatment did cause a decrease in the relative abundance of these two proteins, although the magnitude of this decrease was rather variable between experiments. In contrast, depletion of round spermatids from ST at stages VI-VIII consistently increased the relative abundance of ARP-3 and ARP-4 on 2-D tluorograms (Fig. 7) (McKinnell and Sharpe 1992). Our current belief is that these two proteins are secreted by the Sertoli cells, though their identity is still unknown.
ARP-6 and ARP-7 correspond in terms of their molecular weight and pi to the A and B forms of P-Mod-S, an androgen-regulated protein
~ u ~ ~ ~ M ~ U U 4.7 5.1 5.8 6.1 6.4 6.9 7.2 7.4 I I I I I I I I I I , I I I I I
94 - SGP1 -+ . SGP'''' _
67 -
43 - . SGP2
30 -
• , • • 20_1-
• ~ 14.4-
_ Slages VI-VIII Control A Siages VI-VIII EOS+4d. .!!.. I I I I I 1
j ::i 94 -
67 -
30 -
20_' ·
14.4 ·
.' -~
Stages VI-VIII EOS+ TE+4d c
SGP 1
-d SGP2
Ur;slaged 28<1. old
, [
CP 2 ;, -. .. ~
,
Fig. 6. Analysis by two-dimensional sodium dodecyl sulfate polyacrylamide gel electrophoresis (20 SOS-PAGE) of 35S-labelled proteins secreted by seminiferous tubules (ST) isolated at stages VI-VIII from control adult rats (A), adult rats treated with ethane dimethane sulphonate (EDS; B) or EOS + testosterone (TE; C) 4 days earlier or unstaged ST isolated from immature rats aged 28 days (D). Each gel was loaded with equal cpm radiolabelled protein and the illustrated fluorograms were obtained by exposure of the gels to photographic film for 21 days. On each gel the three main secretory products of the Sertoli cells (SGP-I, SGP-2 and CP-2) are indicated together with seven putative androgen-regulated proteins (ARPs) numbered 1-7. Note that the ARPs are either absent or are barely detectable as secretory products of unstaged ST isolated from rats aged 28 days, whereas all are clearly detectable secretory products of un staged ST isolated from adult rats (data not shown)
cJ o
Interactions Between Androgens, Sertoli Cells 135
secreted by isolated peritubular cells in culture (Skinner et al. 1988). Depletion of either pachytene spermatocytes or round spermatids from ST at stages VI-VIII caused consistent decreases in the secretion of ARP-6 and ARP-7 (McKinnell and Sharpe 1992) and both were perhaps just detectable as minor secretory products of isolated round spermatids (Fig. 7). The real identity of ARP-6 and ARP-7 remains to be established but if they do correspond to P-Mod-S, then our findings suggest that this protein could emanate from germ cells as well as from peritubular cells.
Pretreatment ofrats with m-DNB or NB (McLaren et al. 1993a) also causes major changes in the secretion of some of the ARP's by ST at stages VI-VIII. ARP-I, ARP-2, ARP-3 and ARP-4 are all reduced consistently in relative abundance, a finding that is highly reproducible (McLaren et al. 1993a). These decreases occur in the absence of any change in intratesticular testosterone levels though the androgen-dependent increase in total protein secretion by ST at stages VI-VIII is also abolished, as described above. In contrast, local testicular heating has no consistent, reproducible effects on secretion of the ARP's (McLaren et al. 1994), despite the fact that this treatment also abolishes the androgen-dependent increase in total protein secretion by ST at stages VIVIII (Fig. 5).
6.4 Discussion
When we initiated this programme of work our anticipation was that, via subtraction hybridization or via microsequencing of androgen-regulated proteins, we would identify one or more of the androgen-regulated genes which we thought would mediate the supportive effects of testosterone on spermatogenesis. Despite an immense effort we have failed in this principal objective, though in failing we have advanced our understanding considerably. Indeed, as will be argued below it could well be that our primary objective was guaranteed not to succeed because testosterone does not regulate spermatogenesis in the straightforward way that we had presumed. But before engaging in this speculation it is worthwhile looking back over our work of the past 4 years and asking 'what has it taught us?'.
136 R.M. Sharpe et al.
pH 4.7 5.1 5.8 6.1 6.4 6.9 7.2 7.4 I I I I I I I I
4.7 5.1 5.8 6.1 6.4 6.9 7.2 7.4 I I I I I I I I
SGP 1 • '"
•
a 20.1 -
g 14.4 · • 0. Slagcs VI· VIII Conlrol
" I I I
3: 94 . ::i
67 ·
5 43 ·0
30 ·
20.1·
14.4 ·
, •
lsolaled PS
Fig. 7. Analysis by two-dimensional sodium dodecyl sulfate polyacrylamide gel electrophoresis (20 SOS-PAGE) of 35S-labelled proteins secreted by seminiferous tubules (ST) at stages VI-VIII isolated from control rats (A) or from rats treated 18 days previously with methoxyaceticacid to deplete round spermatids (-RS) at stages VI-VIII (8). Also shown are comparable gels for radiolabelled proteins secreted by isolated pachytene spermatocytes (PS; C) or round spermatids (RS;D) which were isolated from whole testes from untreated adult rats and cultured in vitro in numbers, and under conditions, which were comparable to those for the isolated ST. Other details are as given in Fig. 6
Interactions Between Androgens, Sertoli Cells 137
It is most appropriate to begin with testicular morphology as it is this hiological endpoint which we have used as our starting point for launching the molecular and biochemical studies. We remain unshaken in our conviction that this was the correct starting point and all of our studies on androgen-regulated protein secretion bear this out. If our studies had been based around the use of isolated Sertoli cells from immature rats, we would have found none of the major changes in general and specific protein secretion which are testosterone regulated, as these are only switched on during puberty or adulthood and even then are conditional on the presence of an appropriate germ cell complement. But other than just enabling the other studies to progress, the morphological studies have also given us real insight into how testosterone controls spermatogenesis and how this process fails following testosterone withdrawal, i.e. via the delayed degeneration of germ cells, notably condensing/elongating spermatids at stages X-XIII (see Fig. I; Kerr et al. 1993; Sharpe 1994). As we know already that all gene transcription in spermatids finishes by stage VIII because of the restructuring/condensation of DNA that occurs in the spermatids at stages IX-XIII, it is perhaps logical that one or more of the transcripts required during this period are switched on by testosterone during stages VII-VIII.
Conclusive proof that testosterone action at stage VII is a prerequisite for later steps in germ cell developmelJt at stages IX-XIII is, however, not available. For example, demonstration that at stages VIIVIII, testosterone switched on transcription of the gene for one of the proteins involved in nuclear condensation, would provide such evidence. As yet we have been unable to demonstrate any such change, as the example of TP-2 described above illustrates. Indeed, all our attempts to identify androgen-regulated genes at stages VI-VIII using subtraction hybridization have so far failed to identify any such gene. As discussed earlier, this may be simply a retlection of inherent inadequacies in the subtraction hybridization approach. Alternatively, this glaring negative could be telling us about inadequacies in our thinking.
Perhaps testosterone does not work in the way that we had envisaged, i.e. switching on a small number of key genes during stages VI-VIII. What positive evidence do we have that would support this line of thinking? In fact we have several pieces of such evidence. The most powerful is the effect of testosterone in 'upregulating' overall protein
138 R.M. Sharpe et al.
secretion at stages VI-VIII. This effect does not involve the selective upregulation of a few key proteins (though the ARPs are a minor exception) but as far as we are able to tell involves a doubling in secretion of virtually all ST-secreted proteins, probably including those that emanate from the germ cells. We know from other studies that the production of seminiferous tubule fluid at stages VIJ-VIII is also doubled (reviewed in Sharpe 1994), though this change does not simply account for the increased secretion of proteins at these stages (Sharpe et al. 1994). Clearly, testosterone cannot be altering transcription of all the genes which code for these proteins. Instead, it must be upregulating some aspect of the translation or stability of mRNA, protein synthesis and/or packaging or the processes involved in protein secretion. The absence of any corresponding increase in synthesis of intracellular proteins at stages VI-VIII perhaps argues against effects on translational/synthetic machinery, though at 4 h after exposure of rats to local testicular heating, when the normal increase in protein secretion by ST at stages VI-VIII is exacerbated, there is also a corresponding increase in total intracellular protein (McLaren et al. 1994). The fact that at 24 h after local testicular heating or treatment with the Sertoli cell toxicants m-DNB and NB the increase in protein secretion at stages VI-VIII was prevented completely, without interfering with testosterone levels, again argues for an effect at some point in the protein synthesis/secretion pathway. The fact that all three of these treatments result in major impairment of spermatogenesis also testifies further to the importance of the normal increase in protein secretion at stages VI-VIII.
The observation that depletion of anyone of the adluminal germ cell types at stages VI-VIII is able to prevent the normal increase in total protein secretion at these stages is extremely puzzling, but again suggests that post-transcriptional rather than transcriptional events are involved. The fact that these 'unknown' events are dually regulated by both testosterone and the germ cells and that neither of these triggers on their own are sufficient to induce the overall increase in protein secretion provides a remarkable example of how stage-specific changes in ST function are controlled.
Identification of seven ARPs does not, at first sight, support the above line of thinking - rather it supports our initial idea that testosterone acts to regulate the transcription of several specific genes. Or does it? The fact that at least two of the identified ARPs derive from germ cells
Interactions Between Androgens, Sertoli Cells 139
(round spermatids) is difficult to explain when considering that such cells are thought not to possess androgen receptors (Sharpe 1994). Does this finding mean that testosterone acts directly on the germ cells? Even if the germ cells do not possess androgen receptors other pathways of testosterone action are possible, e.g. interaction with androgen-binding protein (ABP) which then binds to and regulates the germ cells, there is at least preliminary supporting evidence for such mechanisms (reviewed in Sharpe 1994) and it is noteworthy that ST at stages VII-VIII secrete far more ABP than do ST at other stages (Ritzen et al. 1982). Indeed there are a number of powerful arguments that testosterone cannot be regulating spermatogenesis via androgen receptor-coupled pathways, notably the disparity between the levels of testosterone required for maintenance of spermatogenesis and the 40-fold lower levels required for complete occupancy of all the androgen receptors (Sharpe 1994). Moreover, recent studies of our own which have investigated the hormonal control of expression of the genes for CREM and CREB (proteins which modulate the transcriptional effects of cyclic AMP), have shown that both testosterone and FSH can exert major effects on their expression in germ cells and Sertoli cells at stages VII-VIII (West et al. 1994).
So how does testosterone control spermatogenesis? Unfortunately we still have not found the answer, though we have surely taken the first steps towards doing so. These steps have involved surprises and disappointments but point the way for our next research onslaught. What will this show? Could it be that ABP secreted from the Sertoli cell interacts with testosterone to trigger germ cells to inform the Sertoli cell that stages VI-VIII 'have arrived' and that this dual trigger then induces increased protein secretion by the Sertoli cells (? and the germ cells)? Hopefully we shall know within a few years, though there are probably more surprises in store along the way.
Acknrm1edgenlen/s. Those who have participated in this research were supported by the Medical Research council (UK), INSERM (France) and the National Health & Medical Research Council (Australia).
140 R.M. Sharpe et al.
References
Allenby G (1990) Chemical disruption of spermatogenesis. PhD thesis, University of Edinburgh
Allenby G, Foster PMD, Sharpe RM (1991a) Evaluation of changes in the secretion of immunoactive inhibin by adult rat seminiferous tubules in-vitro as an indicator of early toxicant action on spermatogenesis. Fund Appl Toxicol 16:710-724
Allenby G, Foster PMD, Sharpe RM (1991b) Evidence that secretion of immunoactive inhibin by seminiferous tubules from the adult rat testis is regulated by specific germ cell types: correlation between in-vivo and in vitro studies. Endocrinology 128:467-476
Bartlett JMS, Sharpe RM (1987) Effect of local heating of the testis on the levels in interstitial fluid of a putative paracrine regulator of the Leydig cells and its relationship to changes in Sertoli cell secretory function. J Reprod FertiI80:279-287
Bartlett JMS, Kerr JB, Sharpe RM (1986) The effect of selective destruction and regeneration of rat Leydig cells on the intratesticular distribution of testosterone and morphology of the seminiferous epithelium. J Androl 7:240-253
Bartlett JMS, Kerr JB, Sharpe RM (1988) The selective removal of pachytene spermatocytes using methoxyaceticacid as an approach to the study in vivo of paracrine interactions in the testis. J Androl 9:31-40
Jegou B (1993) The Sertoli-germ cell communication network in mammals. Int Rev Cytol 147:25-96
Jegou B, Laws AO, de Kretser OM (1984) Changes in testicular function induced by short -term exposure of the rat testis to heat: further evidence for interaction of germ cells, Sertoli cells and Leydig cells. Int J Androl 7:244-257
Kerr JB, Millar M, Maddocks S, Sharpe RM (1993) Stage-dependent changes in spermatogenesis and Sertoli cells in relation to the onset of spermatogenic failure following withdrawal of testosterone. Anat Rec 235:547-559
McKinnell C, Sharpe RM (1992) The role of specific germ cell types in modulation of the secretion of androgen-regulated proteins (ARP's) by stage VJVIII seminiferous tubules from the adult rat. Mol Cell Endocrinol 83:219-231
McLaren TT, Foster PMD, Sharpe RM (1993a) Identification of stage-specific changes in protein secretion by isolated seminiferous tubules from the rat following exposure to either m-dinitrobenzene or nitrobenzene. Fund Appl Toxieol 21 :384-392
Interactions Between Androgens, Sertoli Cells 141
McLaren IT, Foster PMD, Sharpe RM (I 993b) Effect of age on seminiferous tubule protein secretion and the adverse effects of testicular toxicants in the rat. Int J Andro116:370-379
McLaren TT, Foster PMD, Sharpe RM (1994) Effect of local testicular heating on stage-dependent protein secretion by isolated seminiferous tubules from the adult rat. J Reprod Fertil (submitted)
Meistrich ML, Longtin J, Brock WA, Grimes SR, Mace ML (1981) Purification of rat spermatogenic cells and preliminary biochemical analysis of these cells. Bioi Reprod 25: 1065-1077
Millar MR, Sharpe RM, Maguire SM, Saunders PTK (1993) Cellular localization of messenger RNAs in rat testis: application of digoxigenin-Iabelled ribonucleotide probes to embedded tissue. Cell Tissue Res 273:269-277
Onoda M, Djakiew D (1991) Pachytene spermatocyte protein(s) stimulate Sertoli cells grown in bicameral chambers: dose-dependent secretion of ceruloplasmin, sulfated glycoprotein-I, sulfated glycoprotein-2 and transferrin. In Vitro Cell Dev Bioi 27 A:215-222
Parker MG, White R, Williams JG (1980). Cloning and characterization of androgen-dependent mRNA from rat ventral prostate. J Bioi Chern 255:6996-7001
Parvinen M (1993) Cyclic function of Sertoli cells. In: Russell LD, Griswold MD (eds) The Sertoli cell. Cache River Press, Clearwater, Florida, pp 331-347
Parvinen M, Vihko KK, Toppari J (1986) Cell interactions during the seminiferous epithelial cycle. Int Rev Cytol 104: 115-129
Ritzen EM, Boitani C, Parvinen M, French FS, Feldman M (1982) Stage-dependent secretion of ABP by rat seminiferous tubules. Mol Cell Endocrinol 25:25-33
Russell LD, Clermont Y (1977) Degeneration of germ cells in normal, hypophysectomized and hormone treated hypophysectomized rats. Anat Rec 187:347-366
Russell LD, Malone JP, Karpas SL (1981) Morphological pattern elicited by agents affecting spermatogenesis by disruption of its hormonal stimulation. Tissue Cell 13:369-380
Saunders PTK, Millar MR, Maguire SM, Sharpe RM (1992) Stage-specific expression of rat transition protein-2 mRNA and possible localization to the chromatoid body of step 7 spermatids by in-situ hybridization using a nonradioactive riboprobe. Mol Reprod Dev 33:385-391
Saunders PTK, Millar MR, West AP, Sharpe RM (1993) Mitochondrial cytochrome C Oxidase II messenger ribonucleic acid is expressed in pachytene spermatocytes at high levels and in a stage-dependent manner during spermatogenesis in the rat. Bioi Reprod 48:57-67
142 R.M. Sharpe et al.: Interactions
Sharpe RM (1993) Experimental evidence for Sertoli-germ cell and SertoliLeydig cell interactions. In: Russell LD, Griswold MD (eds) The Sertoli cell. Cache River Press, Clearwater, Florida, pp 391-418
Sharpe RM (1994) Regulation of spermatogenesis. In: Knobil E, Neill 10 (eds) The physiology of reproduction, 2nd edn. Raven, New York, pp 1363-1424
Sharpe RM, Fraser HM, Ratnasooriya WD (1988) Assessment of the role of Leydig cell products other than testosterone in spermatogenesis and fertility in adult rats. Int J Androl II :507-523
Sharpe RM, Maddocks S, Kerr 18 (1990) Cell-cell interactions in the control of spermatogenesis as studied using Leydig cell destruction and testosterone replacement. Am J Anat 188:3-20
Sharpe RM, Maddocks S, Millar M, Saunders PTK, Kerr 18, McKinnell C (1992) Testosterone and spermatogenesis:identification of stage-dependent, androgen-regulated proteins secreted by adult rat seminiferous tubules. J AndroI13:172-184
Sharpe RM, Millar M, McKinnell C (1993a) Relative roles of testosterone and the germ cell complement in determining stage-dependent changes in protein secretion by isolated rat seminiferous tubules. Int J Androl 16:71-81
Sharpe RM, McKinnell C, Millar M, Maguire S, Saunders PTK (1993b) Leydig cell-Sertoli cell-germ cell interactions. In: Whitcomb RW, Zirkin BR (eds) Understanding male infertility: basic and clinical approaches. Raven, New York, pp 143-153
Sharpe RM, Kerr 18, McKinnell C, Millar M (1994) Temporal relationship between androgen-dependent changes in the volume of seminiferous tubule fluid, lumen size and seminiferous tubule protein secretion in the rat. J Reprod Fertil (in press)
Skinner MK, Fetterolf PM, Anthony CT (1988) Purification of a paracrine factor, P-Mod-S, produced by testicular peritubular cells that modulates Sertoli cell function. J Bioi Chern 263:2884-2890
West AP, Sharpe RM, Saunders PTK (1994) Differential regulation of cyclic adenosine 3',5'-monophosphate (cAMP) response element-binding protein and cAMP response element modulator messenger ribonucleic acid transcripts by follicle-stimulating hormone and androgen in the adult rat testis. Bioi Reprod 50:869-881
Wing TY, Christensen AK (1982) Morphometric studies on rat seminiferous tubules. Am J Anat 165:13-25
7 Cell Cycle Checkpoints in Male and Female Germ Cells
D.F. Albertini
7.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 143 7.2 Cell Cycle Checkpoints. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 144 7.2.1 Mitosis Regulation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 144 7.2.2 Spatial and Temporal Features of Mitosis. . . . . . . . . . . . . . . . . . .. 146 7.2.3 Meiosis.............. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 146 7.3 Cell Cycle Checkpoints in Female Gametes. . . . . . . . . . . . . . . . .. 14X 7.4 Cell Cycle Checkpoints in Male Gametes. . . . . . . . . . . . . . . . . . .. 149 7.S Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ISO References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 151
7.1 Introduction
Tissues exhibit distinct differences in their abilities to replenish their cellular mass _. either in the case of normal cellular turnover or in response to cell loss due to injury. In all cases, however, the replacement of cells derives from a change in the cell cycle state of precursor or stem cells. Checkpoints in the cell cycle thus become the conceptual equivalent of signals that serve to initiate or stop progression of the cell cycle at discrete points along the pathway culminating in DNA synthesis, the S phase, or mitosis (or meiosis), the M phase. Why cells have evolved complex molecular controls for regulating or checking cell cycle status is a fascinating question and one being actively pursued in view of its imp0l1ance in understanding the problem of growth control in nonnal and abnonnal somatic tissues (Murray and Hunt 1993). This problem
144 D.F. Albertini
takes on an even more complex character when viewed in the context of germ cells. Are germ cells subjected to cell cycle checkpoints that differ from those imposed on somatic cells? What special modifications in the cell cycle of germ cells exist in relation to the control of meiosis and reductive cell divisions? How do male and female germ cells differ in their cell cycle checkpoints and what sex-specific factors contribute to these differences? While incomplete, tentative answers to these questions form the basis of this review and will hopefully steer future work in this area in a promising direction.
The importance of cell cycle checkpoints in the germ cells of mammals owes directly to the crucial role of meiosis in sexual reproduction (McLaren 1984). Not only are reductive divisions essential to the generation of haploid gametes, but meiotic checkpoints serve to regulate DNA repair and the timing of gamete interaction during the reproductive cycle of the organism. As will be discussed, the extremely dimorphic character of cell cycle checkpoints in male and female germ cells must reflect the extremely different reproductive strategies assumed by each sex. Why these differences may have evolved from asexually reproducing forms, in relation to the better-understood mitotic control mechanisms, will be considered.
7.2 Cell Cycle Checkpoints
7.2.1 Mitosis Regulation
Current views of cell cycle regulation must invoke an explanation for the orderly sequence of events that support growth first, and subsequently the separation or division of a cell into two daughter cells (Murray and Hunt 1993). At its most basic level, growth and division of cells involves chromosome replication and segregation, respectively. These two events define the two most crucial phases of the eukaryotic cell cycle: the S phase, during which DNA replication occurs, and the M phase, during which chromosome segregation occurs. Typically, a gap phase is interposed between the end of mitosis (M) and DNA synthesis (S), known as Gland G2, and refers to the gap between the end of S and the next M phase. From the perspective of intrinsic regulation, most studies have focused on the role of the p34cdc2 kinase as a universal
Cell Cycle Checkpoints in Male and Female Germ Cells 145
factor involved with the initiation of either S or M phases of the cell cycle (Albertini et al. 1993). This kinase appears to be expressed throughout the cell cycle; however its enzymatic activity is most evident at the onset of the M phase. And, the p34cdc2 kinase is now known to constitute the major catalytic component of M phase-promoting factor or MPF, a cytoplasmic activity known to induce mitosis when injected into nondividing cells (Fulka et al. 1992). Cyclins comprise the regulatory component of active MPF since these proteins change in abundance during the cell cycle and directly participate in the entry into and exit from M phase. How this regulation is accomplished is based on the interaction of cyclins with p34cdc2 so that critical phosphorylation changes in the latter cause the conversion of inactive pre-MPF into active MPF (Rattner et al. 1990). Active MPF then phosphorylates proteins essential for assembly of the mitotic spindle and chromosome alignment. As a driving force for these early events of M phase, MPF has been viewed as a cell cycle engine (Murray and Hunt 1993). At anaphase of mitosis, the engine is shut off due to the selective proteolysis of cyclins that results in inactivation of the p34 kinase and exit from M phase. Thus, MPF is the key biochemical regulator of mitotic cell cycles because its intrinsic kinase activity can be precisely and acutely modulated through changes in the cyclin subunits.
While cyclins serve as intrinsic molecular regulators of cell cycle progression, other external factors must be involved in starting and stopping the cell cycle engine (Albertini 1992). These factors are thought to act on a feedback control system that would ensure that mitosis occurs only if DNA replication (or repair) has been completed (Murray and Hunt 1993). From studies on a variety of organisms, it is now clear that at least three major checkpoints exist to regulate the cell cycle: Start, the entry of cells from an arrested state into the cell cycle; mitotic entry; and mitotic exit. Examples of how external factors would influence these checkpoints include: the stimulation of vertebrate cells to divide in response to growth factors; the influence of cell size and nutrient availability on mitotic entry in fission yeast; and the completion of meiosis in mammalian oocytes as triggered by sperm. In each ofthese cases, stoppage of the cell cycle is reversed by signals received from the environment. In addition to checkpoints, however, mitotic somatic cells exhibit features that further distinguish them from germ cells (Choi et al. 1991).
146 D.F. Albertini
7.2.2 Spatial and Temporal Features of Mitosis
Exit from mitosis, initiated by chromosome segregation of anaphase, encompasses the physical process of cytokinesis itself. In most mitotically dividing cells, cytokinesis provides the forces needed to create two equally sized and chromosomally equivalent daughter cells. Not surprisingly, key components of the contractile machinery of cytoplasm are among the substrates for the M phase kinase that must function coordinately with the spindle apparatus to achieve a symmetric cell division. Daughter cells must also inherit cytoplasmic equivalents of various organelles such as the Golgi complex, mitochondria, and Iysosomes. What spatially determines the equal and complete separation of genome and cytoplasm during mitosis is not understood and various features of this process appear to have been modified in germ cells (Ookata et a1. 1993).
Finally, the timing of cell cycle progression is an additional factor for which much variation exists in somatic cells. The regular cadence of the repeated cell divisions seen in early embryonic development sharply contrasts to the patterns of cell division seen in differentiated tissues undergoing homeostatic cell replacement. Considering further the heightened proliferative rates of tumors on tissues undergoing vascularization, one is struck with the array of proliferation responses exhibited by somatic cells. Understanding what distinguishes germ cells from somatic cells in terms of cell cycle regulation, spatial ordering, and timing rests ultimately in an analysis of the processes that distinguish meiosis from mitosis (Wickramasinghe and Albertini 1993).
7.2.3 Meiosis
The meiotic cell cycle is unique in exhibiting two consecutive M phases following a single S phase (Albertini 1992; McLaren 1984). In multicellular organisms, this type of cell division yields the haploid gametes needed for sexual reproduction. But meiosis also involves unique alterations in other aspects of the cell cycle including spatial-temporal regulation, chromosome behavior, and the imposition of checkpoints. We review general elements of these features of meiosis in this section
Cell Cycle Checkpoints in Male and Female Germ Cells 147
before moving to a detailed comparison between male and female gametes.
At the level of chromosome organization, two features distinguish meiosis from mitosis. First, genetic recombination of homologous chromosomes occurs shortly after DNA replication through as yet undefined mechanisms involving synaptonemal complexes (McLaren 1984). While synaptonemal complexes disappear before entry into meiosis I, paired homologs remain conjoined by chiasmata until the first metaphase-anaphase transition. The second distinguishing feature pertains to the behavior of sister kinetochores (De Kretser and Kerr 1988). Unlike mitosis or meiosis 11 where sister kinetochores split at anaphase and each chromatid moves to opposite poles of the spindle, in meiosis I sister kinetochores do not split, ensuring that intact homologs migrate to opposite poles. Studies on the primary spermatocytes of Drosophila indicate that this phenomenon is due to the failure of the kinetochore to split until after anaphase of meiosis 1.
With respect to meiotic cell cycle regulation, several features of MPF activation appear to be altered when compared to mitosis (Choi et al. 1991; Fulka et al. 1992). First, whereas MPF activity decreases at anaphase of meiosis I, chromatin remains condensed and nuclear envelope reassembly fails to occur, suggesting that other kinases may operate to maintain an M phase state. Second, meiosis I appears to be initiated through the activation of a preexisting pool of pre-MPF and does not require new protein synthesis, whereas meiosis II appears to depend on the synthesis of new MPF components. Since most of the studies illustrating these differences in cell cycle regulation have been performed on vertebrate oocytcs, some caution should be exercised in extrapolating this work to the male gamete. However, one common attribute to male and female gametes that may be related to meiotic cell cycle control concerns the serine kinase proto-oncogene product c-mos; cmos is only expressed in meiotic germ cells and may playa key role in biochemical modifications affecting MPF (O'Keefe et al. 1989; Paules et al. 1989). At this point, consideration will be placed on the features peculiar to male and female gametes in terms of meiotic cell cycle regulation.
148 D.F. Albertini
7.3 Cell Cycle Checkpoints in Female Gametes
By far the better understood of the two sexes, female gametes have proved to be optimal for studies of cell cycle regulation in providing an abundance of experimentally manipulable materials (Murray and Hunt 1993). Female gametes or oocytes display a meiotic cell cycle punctuated by a series of checkpoints (Albertini 1992). The first of these is imposed at birth in most mammals when oocytes become arrested in meiotic prophase and remain in this state for months or even years. The meiotic cell cycle is not reinitiated formally until gonadotrophins stimulate meiotic maturation coincident with ovulation; the maturation process entails progression through meiosis J, emission of the first polar body, and arrest at metaphase of meiosis II. Three checkpoints are imposed during meiotic maturation. The first is executed during the growth phase of oogenesis when oocytes acquire the ability to resume meiosis (Wickramasinghe and Albertini 1993). This checkpoint is based on negative feedback control since inhibitory signals generated by follicular somatic cells are transmitted to the oocyte where MPF activation is held in abeyance, ensuring maintenance of prophase arrest. This checkpoint is apparently overridden at the time of ovulation when gonadotropin stimulation induces meiotic resumption, presumably as a result of the transfer of a cell cycle-stimulating factor that allows activation of preexisting stores of MPF. The nature of this signal remains obscure although it is likely that whatever metabolic change is elicited in the oocyte, it would involve a change in the phosphorylation state of the p34cdc2 kinase (Choi et al. 1991; Ookata et al. 1993). Unlike the first two checkpoints which derive from extrinsic somatic cues, the third checkpoint imposed during meiotic maturation involves an intrinsic mechanism for arresting the cell cycle of metaphase of meiosis II. The completion of meiosis is normally triggered by fertilization, and experimental manipulations resulting in either elevation of intracellular calcium or inhibition of protein synthesis have also been shown to release oocytes from this point of cell cycle arrest (Fulka et al. 1992). Work on various oocyte model systems has begun to clarify the biochemical basis for metaphase II arrest and its release.
These studies emanate from the classical work on Masui using frog eggs in which a metaphase arresting factor was identified in mature oocytes (Murray and Hunt 1993). This factor is called cytostatic factor
Cell Cycle Checkpoints in Male and Female Germ Cells 149
(CSF). Several lines of evidence indicate that the production and maintenance of CSF is regulated by the c-mos serine kinase, a proto-oncogene product uniquely expressed in male and female germ cells (Paules et al. 1989). This c-mos synthesis is required for the activation of MPF that drives the meiotic cell cycle in oocytes from meiosis I to meiosis II and for inducing metaphase arrest. Elegant experiments using mouse oocytes injected with antisense to c-mos showed that after the completion of meiosis I, injected oocytes failed to engage in meiosis II and instead proceeded into a zygotic interphaselike state (O'Keefe et al. 1989). Antibodies to c-mos can deplete oocyte extracts of CSF activity and the activation of a calpain protease that degrades CSF has been shown to occur following fertilization (reviewed in Wickramasinghe and Albertini 1993). Collectively, this work suggests that the metaphase II cell cycle arrest is due to c-mos/CSF acting to stabilize the M phase states of oocytes until signaled by fertilization to complete meiosis.
From this discussion, three general principles about cell cycle control in female gametes emerge. First, the meiotic process is punctuated by checkpoints executed by extragametic cells (follicle cells or sperm). Second, at each checkpoint, MPF activation or inactivation are subject to "acute" regulation. And third, a cell cycle-modulating factor unique to meiosis, c-mos/CSF, functions to check the timing of the completion of meiosis. The extent to which male germ cells exhibit comparable attributes of cell cycle control is reviewed in the next section.
7.4 Cell Cycle Checkpoints in Male Gametes
In sharp contrast to the situation in the female, male germ cells progress through successive mitotic and meiotic cell cycles in a continuous, rather than discontinuous, fashion. Hormonal support by both gonadotrophins (follicle-stimulating hormone, FSH) and androgens are believed to play important roles in both the initiation of meiosis at puberty and in the maintenance of spermatogenesis throughout adult life (Courot et al. 1970). Sertoli cells mediate meiosis and differentiation of male germ cells by virtue of the fact that they express receptors for FSH and androgens and maintain intimate and dynamic interactions with germ cells within the seminiferous epithelium. Notably, the somatic cell dependence of germ cell cycle progression in the testis is further evidenced
150 D.F. Albertini
by the lack of receptors for FSH or androgens in the germ cells. How Sertoli cells convey information to germ cells that alters cell cycle progression is not at all clear. However, some features of mitotic and meiotic male germ cell behavior do suggest that somatic cells can impart checkpoint control during the process of spermatogenesis.
The fact that sexually mature males undergo continual germ cell production implies that once initiated, mitotic and meiotic cell cycles proceed uninterrupted in their course (Courot et al. 1970). It has been suggested that a meiosis-arresting factor exists in the testis of prepubertal animals and thus a negative feedback control checkpoint may operate prior to the initiation of the spermatogenic cycle (McLaren 1984). The timing of transitions through specific meiotic stages appears to be species specific but some generalities have been noted. For example, primary spermatocytes reside in meiotic prophase for long periods (days), whereas the M phases during both meiosis I and meiosis II are rapid (minutes) and thus more reminiscent of the behavior exhibited by somatic cells. Other similarities with somatic cells exist in that meiotic spindles in male gametes exhibit tapered poles, centro somes containing centrioles, and show symmetric cytokinesis. However, cytokinesis during premeiotic and meiotic divisions is incomplete due to the persistence of cytoplasmic bridges. Thus, the syncytial character produced by this cellular arrangement of pre- through postmeiotic stages would encourage randomization of cytoplasmic regulating factors in order to synchronize cell cycle progression between "daughter" cell cohorts. If syncytialization were not in evidence, autonomous or intrinsic control to each primary or secondary spermatocyte would have to be invoked to explain the synchronized cell cycle progression of spermatogenesis. Given the likelihood of this situation, it is tempting to speculate that meiotic cell cycle control in male gametes lacks the key checkpoints exhibited by female gametes and therefore more closely resembles the type of regulation observed in somatic mitotic cells.
7.S Conclusions
Male and female germ cells share certain general features of cell cycle control but differ markedly in many respects. Common properties include (I) the imposition of negative checkpoint controls through the action
Cell Cycle Checkpoints in Male and Female Germ Cells 151
of somatic cells on the gametes; (2) the role of honnones in initiating meiotic M phase entry; and (3) the timing of meiotic divisions evidenced in a protracted prophase of meiosis I and rapid transition from meiosis [ to II. However, among the many specific differences are features that suggest that checkpoint regulation in male gametes is more typical of that exhibited by somatic mitotic cells - symmetric cytokinesis, short M phase transit times, activation of premeiotic stem cell mitoses. Thus, while many of the details remain to be elucidated regarding the molecular and cellular basis of cell cycle regulation in gametes, it seems likely that meiosis in female gametes has evolved a precise fonn of checkpoint regulation designed to optimize reproductive success. The evolutionary pressures in mammals that may have directed such changes in cell cycle control are likely to be based in sex differences of "supply and demand" at the level of gametes and the coupling of reproductive cyclicity in females with the timing of ovulation.
Acknowledgements. This work has been supported by grants from the NIH (HD 5006g), Markey Charitable Trust Foundation, and the Center for Environmental Management of Tufts University.
References
Albertini OF (1992) Regulation of meiotic maturation in the mammalian oocyte: interplay between exogenous cues and the microtubule cytoskeleton. Bioessays 14:97-103
Albertini OF, Mattson BA, Messinger SM, Wickramasinghe D, Plancha CE (1993) Nuclear and cytoplasmic changes during oocyte maturation. In: Bavister BD (cd) Preimplantation mammalian development. Springer, Berlin Heidelberg New York, pp 5-21
Choi T, Aoki F, Mori M, Yamashita M. Nagakawa Y, Kohmoto K (1991) Activation of p34cdc2 protein kinase in meiotic and mitotic cell cycles in mouse oocytes and embryos. Development 113:789-796
Courot M. Hochereau-de Reviers M-T, Ortavant R (1970) Spermatogenesis. In: Johnson AD, Gomes WR, Vandermark NL (eds) The testis. Academic. New York, pp 339-432
De Kretser OM. Kerr JB (19gg) The cytology of the testis. In: Knobil E, Neill J (ed) The physiology of reproduction. Raven, New York, pp 837-932
152 D.F. Albertini: Cell Cycle Checkpoints
Fulka J, Jung T, Moor RM (1992) The fall of biological maturation promoting factor (MPF) and histone HI kinase activity during anaphase and telophase in mouse oocytes. Mol Reprod Dev 32:378-382
McLaren A (1984) Meiosis and differentiation of mouse germ cells. In: Evans CW, Dickinson HG (eds) Controlling events in meiosis. Company of Biologists, Cambridge, pp 7-23
Murray A, Hunt T (1993) The cell cycle. Freeman, New York O'Keefe S, Wolfes H, Kiessling AA, Cooper GM (1989) Microinjection of an
tisense c-mos oligonucleotides prevents meiosis II in the mouse oocyte. Proc Natl Acad Sci USA 86:7038-7042
Ookata K, Hisanaga S-i, Okumura E, Kishimoto T (1993) Association of p34cdc2/cyclin B complex with microtubules in starfish oocytes. J Cell Sci 105:873-881
Paules RS, Buccione K, Moschel RC, Vande Woude GF, Eppig JJ (1989) Mouse mos proto-oncogene product is present and functions during oogenesis. Proc Natl Acad Sci USA 86:5395-5399
Rattner JB, Lew J, Wang JH (1990) p34cdc2 kinase is localized to distinct domains within the mitotic apparatus. Cell Motil Cytoskel 17:227-235
Wickramasinghe D, Albertini DF (1993) Cell cycle control during mammalian oogenesis. In: Pedersen RA (ed) Current topics in developmental biology. Academic, San Diego, pp 125-153
8 Signal Transduction in Mammalian Spermatozoa
G.S. Kopf, P. Kalab, P. Leclerc, X.P. Ning, D. Pan, and P. Visconti
X.I Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 153 X.2 Regulation of Mammalian Sperm Function at a Distance
by Chemokinetic and Chemotactic Factors. . . . . . . . . . . . . . . . . .. 157 X.3 Regulation of Mammalian Sperm Function at the Site
of Fertilization: Binding of Sperm to the Zona Pellucida and the Induction of Sperm Acrosomal Exocytosis . . . . . . . . . . .. 161
X.3.1 Properties of the Zona Pellucida . . . . . . . . . . . . . . . . . . . . . . . . . .. 161 X.3.2 Interaction of Sperm with the Zona Pellucida. . . . . . . . . . . . . . . .. 164 X.3.3 Identity of Sperm-Associated Receptors for the Zona Pellucida .. 166 X.3A Zona Pcllucida-Mediated Signal Transduction. . . . . . . . . . . . . . .. 170 X.3.S Intracellular Effectors Regulating
thc Zona Pellucida-Induced Acrosome Reaction. . . . . . . . . . . . .. 174 XA Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 176 References .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 177
8.1 Introduction
Intercellular communication between gametes is essential to the unique event in the life cycle of an organism called fertilization. Achievement of successful fertilization results from requisite and reciprocal cell-induced sperm and egg activation events mediated by unique cellular and environmental cues associated with either the gametes or the reproductive tract/environment. In the case of the sperm, the interaction of this
154 G.S. Kepf et al.
highly motile cell with the female reproductive tract/environment, as well as with the egg both at a distance and in close proximity, represent a series of integrated processes designed to deliver sperm with optimal fertilizing potential to the site of fertilization. Recent studies have revealed that many aspects of gamete activation prior and subsequent to fertilization have similarities to intercellular and intracellular signaling systems utilized by somatic cells. Cell surface receptors or binding proteins on sperm for egg products have been identified in some species, while their identity in other species remains controversial. The occupancy of these receptors/binding proteins by egg products results in transmembrane signaling and stimulation of intracellular effector systems, leading to subsequent sperm activation. Such activation events may include changes in motility, chemotaxis, and induction of acrosomal exocytosis, all of which may be essential prerequisites to successful fertilization in various species. This review will summarize what is known about these intercellular communicative events, with an emphasis on the mechanisms by which mammalian sperm process egg-associated signals via signal transduction pathways to effect changes in cellular function. The nature of these signaling systems is now being elucidated at the molecular level and has revealed some unique aspects of communication and transmembrane signaling between gametes. An understanding of signal transduction in mammalian sperm will ultimately yield information about the nature of the receptors to which these signal transduction pathways are coupled, as well as the intracellular effectors that ultimately regulate sperm function. Moreover, an understanding of these regulatory pathways will be essential for the future development of clinical approaches designed to enhance or preclude fertilization.
Fertilization in all species depends upon the interaction of fully mature and fertilization-competent male and female gametes. As stated above, sperm function is regulated by specific cues that arise from the extracellular milieu and from the egg/female reproductive tract. These interactions are critical in defining the optimal conditions for successful fertilization. The coordinated interaction of such cues with sperm result in a hierarchy of interactions between both gametes and their associated structures (Fig. 1), some of which will be considered in this review. The types of interactions and the molecules involved in such interactions are dictated in a species-specific manner.
Signal Transduction in Mammalian Spermatozoa
1 2
Chemokinesis Chemotaxis
Selective transport
Specific adhesion
6 Membrane
recognition and egg activation
Penetration of egg-associated structures
5
155
3 Induction of the
acrosome reaction
Motil io/ activation
Fig. 1. The major stages of sperm-egg interaction. Shown is an egg surrounded by its extracellular coats. Acrosome-intact sperm are selectively transported towards the egg in response to chemokinetic and/or chemotactic signals (1). The first stage of interaction between the sperm and the egg occurs at the level of the extracellular coat where specific sperm-egg adhesion occurs (2). Following adhesion, the sperm undergo acrosomal exocytosis in response to signals from this coat (3). Sperm that have undergone the acrosome reaction and are still adherent to the extracellular coat sometime display changes in motility which may aid in the subsequent penetration of the coat (4). Following penetration of the extracellular coat (5) the plasma membranes of both the sperm and egg interact and then fuse (6), resulting in the activation of the egg. As discussed in the text many of these processes may be receptor mediated. Adapted from Ward and Kopf (1993)
156 G.S. Kopf et al.
For example, sperm-egg interaction in many marine invertebrates displaying external fertilization is influenced greatly by the environment into which they are released. If gamete spawning occurs in tide pools or in deeper waters, gamete interaction is limited by the vast dilution effects of sea water. Such harsh environmental factors are compensated for in a number of ways, including spawning in beds, spawning huge numbers of gametes, and spawning synchronous populations of gametes that possess full potential for fertilization. However, additional compensatory mechanisms to ensure the successful union of the male and female gametes are provided by biological signals associated with or produced by the egg that modulate sperm function by intercellular communication at both a distance and in close proximity (see review by Ward and Kopf 1993).
In contrast, the interaction of sperm and eggs in internal fertilization in, for example, mammals is governed by a different set of requirements that necessitates the adaptation and development of seemingly alternative modes of gamete-reproductive tract interactions to ensure the interaction of fully mature and functional male and female gametes. Although the environments of both the male and female reproductive tracts support the maturation of their respective gametes, they also provide a hostile environment for the survival of gametes that do not display optimal fertilizing potential. The numbers of gametes produced by mammals, generally speaking, are significantly less than those of external fertilizers, and the sperm are less synchronous as a population with regard to morphology, motility characteristics, and the ability to respond to egg products to undergo a regulated acrosome reaction. The fertilization potential of these cells, therefore, is very much dependent on all of these aforementioned parameters. Once these cells have acquired the ability for sustained forward motility in the male reproductive tract, they must then acquire the capacity to fertilize an egg upon residence in the female reproductive tract through a poorly understood process known as capacitation. As sperm traverse the female reproductive tract, there are a number of species-specific anatomical and environmental selection processes that limit the delivery of these cells to the site of fertilization, e.g., the ampullary region of the oviduct. Sperm arriving at this region encounter additional selection processes in the form of the selective barriers provided by the cellular investments and extracellular matrix (zona pellucida, ZP) associated with the ovulated egg. These
Signal Transduction in Mammalian Spermatozoa 157
selection processes serve to ensure that sperm-egg interaction occurs between gametes that are most competent to fuse in a successful manner. The response of sperm to these aforementioned selection pressures forms the basis for studies, suggesting that communication between sperm and egg and between gametes and the reproductive tract occurs both in species displaying internal and external fertilization. Such communication occurs in close proximity. but there is increasing evidence that communication also occurs at a distance. It should not be surprising. therefore, that many of these communicative processes are conserved, and that such similarities (as well as differences) could be exploited in forming the groundwork for future studies.
8.2 Regulation of Mammalian Sperm Function at a Distance by Chemokinetic and Chemotactic Factors
A question essential to the understanding of fertilization is whether this process is achieved by random collision between the sperm and the egg or whether gametes communicate prior to their union. As described above, sperm selection in the female reproductive tract could account for the initial selection of subpopulations of cells that ultimately become fertilization competent. However, signals emanating from the reproductive tract or the egg itself could constitute a second line of selection to direct or guide sperm towards the site of fertilization (see Fig. I). Such signals might accomplish this function in a number of ways. First, these signals could function through an ability to stimulate sperm motility and metabolism through a process known as chemokinesis. These signals could also function by directing the movement of sperm towards the egg by the process of chemotaxis.
Chemotaxis is a common mechanism used to direct the movement of many cell types including bacteria, eosinophils, neutrophils, and slime molds (Devreotes and Zigmond 1986). Sperm chemotaxis has been reported to occur in many plant and lower animal species in response to chemoattractants thought to be released from eggs (Miller 1985), and is considered in detail in the review by Ward and Kopf (1993). It must be emphasized that true chemotaxis must be carefully differentiated from chemokinesis which would represent the activation of sperm motility (and possibly metabolism). To demonstrate a true chemotactic response,
158 G.S. Kopf et al.
the cell must change direction and move toward an increasing concentration of the chemoattractant. Both chemokinetic and chemotactic activities may be present and ultimately modulate sperm-egg interaction. Chemokinetic substances may activate sperm, while chemotactic factors may promote the net movement of sperm toward the egg. These biochemical signals may not be necessarily separate entities; it is possible that one substance may contain both activities. The interpretation of chemotaxis has certainly been a problem in past studies designed to elucidate whether this process occurs in mammals.
In contrast to many marine invertebrates, where communication between gametes through the elaboration and perception of chemokinetic and chemotactic factors would be a logical mechanism to contend with the vast environment in which fertilization occurs, communication between gametes in mammals appears to be inherently less important. However, intercellular communication between mammalian eggs and sperm prior to fertilization may be important to successful gamete interaction. For example, in cattle, sheep, hamsters, rabbits and pigs, data indicate that most of the ejaculated sperm may be stored in the caudal isthmus of the oviduct in a state of reduced motility (Harper 1973; Flechon and Hunter 1981; Hunter and Wilmot 1984; Hunter and Nichol 1983, 1986). Sperm in these storage sites are found to resume motility and reach the ampullary region of the oviduct, the site of fertilization, within minutes following ovulation (Harper 1973; Overstreet and Cooper 1979; Flechon and Hunter 1981). In contrast, ejaculated sperm appear to be stored in the cervix in women (Zinaman et al. 1989). These observations have led to the concept that a factor(s) is released from the eggs or follicle to stimulate sperm motility and direct sperm toward the ovulated egg. Such a factor may serve to increase the chance of collision between gametes since the sperm to egg ratio at the site of fertilization is low, being anywhere from 1: 1-\0: I (Zamboni 1972; Cummins and Yanagimachi 1982; Kopf and Gerton 1990). This form of gamete communication might be important in the possible selection of sperm with optimal fertilization capacity and thus serve to protect the ovulated egg from fertilization by defective sperm. Unlike invertebrates, in which sperm are released in a fairly homogenous state and appear to possess fertilizing ability at spawning, mammals release sperm as a heterogeneously matured population that require the process of capacitation in order to gain the ability to fertilize; capacitation is
Signal Transduction in Mammalian Spermatozoa 159
most likely achieved in subpopulations of spenn at different rates. These substances could, conceivably, participate in the spenn maturation process itself, thereby ensuring the presence offully mature sperm at the site of fertilization. Alternatively, such factors could function to direct the optimal, matured sperm toward the egg. The functional consequence of such actions in toto would be to extend the fertilizable lifespan of the ejaculate by bringing subpopulations of the ejaculate into a mature, fertilizable state at different times.
At this time, little experimental evidence exists for the presence, identity and role of chemokinetic and/or chemotactic factors in mammalian sperm-egg interaction. In vitro experiments designed to unequivocally demonstrate chemokinesis/chemotaxis of mammalian sperm by egg factors are difficult to perfonn without artifactual responses, and such studies must be carefully controlled (Eisenbach and Ralt 1(92). Moreover, because spermatozoa undergo maturational processes (e.g., capacitation) in the changing environment of the female reproductive tract, it may be difficult to distinguish between the effects of such substances on chemokinesis/chemotaxis from those on sperm maturation. This is especially critical with regard to capacitation since one endpoint of this maturational process is a change in sperm motility to the "hyperactive" state (Honnan and Babcock 1(90).
A variety of molecules have been proposed as promoting mammalian sperm chemokinesis and/or chemotaxis. N-fonnyl-Met-Leu-Phe, a synthetic peptide chemoattractant for neutrophils, has been postulated to function as a chemoattractant toward bull (Iqbal et al. 1(80) and human (Gnessi et al. 1(85) sperm. However, these studies suffer from potential artifactual problems (Miller 19X2), and the physiological role that such peptides would play in normal mammalian fertilization is obscure since these classes of peptides are normally derived from bacteria. The cells of the cumulus oophorus have also been postulated to secrete substances that affect sperm function (Bronson and Hamada 1(77). These studies demonstrated that sperm incubated in the presence of unfertilized cumulus cell-enclosed eggs changed motility to more erratic patterns. In contrast, sperm swam with straighter trajectories when incubated in the presence of unfertilized eggs devoid of the cumulus oophorus and these trajectories were similar to those observed in the absence of eggs. It could be argued, however, that such changes in motility patterns could result from sperm adhering to the glass microcapillary tubes used in the
160 G.S. Kopf et al.
experiments, an observation made by these authors. Bradley and Garbers (1983) subsequently demonstrated that bovine sperm incubated with cumulus cell-enclosed eggs displayed an increased forward motility index. This effect on motility was due to the cells of the cumulus oophorus, and the effect was increased when the cumulus cells were incubated with follicle-stimulating hormone (FSH). It was also demonstrated that culture medium from isolated cumulus cells had the same effect on motility, leading the authors to conclude that the cumulus oophorus secreted a substance responsible for the changes in sperm motility. It should be noted that although these experiments were performed under conditions that would minimize sperm adherence to glass surfaces, it is not clear whether such a problem can be totally alleviated.
Human follicular fluid obtained from women undergoing in vitro fertilization has been examined for effects on sperm and it has been suggested that this material contains chemoattractant and/or chemokinetic activities. One study demonstrated that human sperm in an agarose matrix migrated toward wells containing follicular fluid (VillanuevaDiaz et al. 1990). Although the use of an agarose matrix may represent a technical problem with these studies (Eisenbach and Ra1t 1992), these data suggest that some component of the follicular fluid could be a chemoattractant. Ralt et al. (1991) demonstrated that diluted human follicular fluid could alter the swimming pattems of human sperm and could cause the net accumulation of sperm towards the follicular fluid compartment in microchemotaxis chambers. The chemokinetic/chemotactic activity(ies) of these follicular fluids varied greatly on an interand intra-patient basis, and the fluids maintained their activity for about 2 weeks at -20°C to -70°C. Interestingly, the sperm response to the fluid was variable with only the population of sperm recovered from swim-up techniques demonstrating accumulation. There also appeared to be a positive correlation between the follicular fluids that displayed chemokinetic/chemotactic activity and the ability of the eggs obtained from the follicles containing these fluids to be fertilized. These data are provocative and argue for some factor of follicular fluid origin in follicles of fertilizable eggs being diluted following ovulation into the oviduct and exerting its chemokinetic/chemotactic effects on sperm. The identity of such a factor(s) is not known and a careful consideration of the relative contribution of a sperm chemokinetic activity versus a sperm chemotactic activity must be firmly established. The capacitation state of the
Signal Transduction in Mammalian Spermatozoa 161
sperm used in all of the above studies must be also considered when interpreting these data, since sperm were prepared under both capacitating and noncapacitating conditions in the different studies. Certainly, additional studies will be needed to confirm the existence of these activities in follicular fluid, and purification of the activities is essential at this point.
Although it is premature to propose the existence of mammalian sperm cell surface receptors that may mediate chemokinetic and/or chemotactic responses, recent experiments have demonstrated the expression of olfactory receptor gene homologues in mammalian germ cells (Parmentier et a1. 1992). Although, these investigators have not established whether the respective proteins encoded by such genes are expressed, it is tempting to speculate that such gene products could be involved in mediating sperm responses to chemokinetic and or chemotactic factors.
8.3 Regulation of Mammalian Sperm Function at the Site of Fertilization: Binding of Sperm to the Zona Pellucida and the Induction of Sperm Acrosomal Exocytosis
8.3.1 Properties of the Zona Pellucida
As shown in Fig. I, a second major sperm activation event that occurs in mammalian sperm is the induction of the acrosome reaction, which is an exocytotic event. Since the acrosome reaction is an absolute prerequisite to successful fertilization in mammals, the regulation of this event is critical and there is ample evidence supporting the notion that this reaction proceeds in response to extracellular signals emanating from the egg, its associated cellular and acellular structures, and the female reproductive tract. The acrosome reaction is essential for fertilization in mammals since it is required for the penetration of the matrices surrounding the egg (e.g., the ZP in mammals). There is a substantial literature dealing with the identity of acrosome reaction inducing molecules in the mammal, much of which has been controversial through the years. This controversy has arisen, in part, by arguments pertaining to the site of the physiologically relevant acrosome reaction, that of the fertilizing sperm, since conclusions of the in vivo site for this exocytotic reaction have
162 G.S. Kepf et al.
been made based on experimental manipulations made in vitro. Such in vitro studies, while invaluable to an understanding of acrosomal exocytosis, may not necessarily reflect the actual conditions that the sperm encounter at the different levels of the male and female reproductive tracts prior to and following their interaction with the egg and its investments. Caution, therefore, must be exercised when interpreting such studies. However, there is a consensus from both in vivo and in vitro studies in numerous mammalian species that the ZP is a primary signal in mediating acrosome reactions that are associated with the fertilizing sperm (Kopf and Gerton 1990). This exocytotic reaction occurs subsequent to the species-specific binding of sperm via the ZP (see Fig. 1). Presently, only the ZP of the mouse egg has been examined in sufficient detail to permit an analysis of the component(s) that possess acrosome reaction-inducing activity, and this extracellular matrix will be considered in detail below.
Species-specific sperm-egg recognition and interaction, sperm activation (i.e., acrosomal exocytosis), and an egg-induced block to polyspermy in the mouse all appear to be mediated by the ZP, an extracellular matrix surrounding the egg (Wassarman 1988, 1990; Fig. 1). The ZP of the mouse egg is composed of three sulfated glycoproteins designated as ZP1, ZP2, and ZP3, and is truly an egg-associated product since it is synthesized and secreted throughout the period of oocyte growth (Bleil and Wassarman 1980a,b; Shimizu et aI. 1983). The ZP isolated from all other mammals studied to date are composed of two to four glycoproteins, the number of which appears to be species dependent (Wassarman 1988; Kopf and Gerton 1990; Dunbar et aI. 1991). Considerable charge heterogeneity of the ZP glycoproteins exists in different species, which is presumably due to the degree of glycosylation and/or sulfation of the individual polypeptide chains. In the mouse, ZPl (Mr = 200 000) is a dimer connected by intermolecular disulfide bonds and appears to function to maintain the three-dimensional structure of the ZP by crosslinking filaments composed of repeating structures of ZP2/ZP3 heterodimers. ZP2 (Mr = 120000 under nonreducing and reducing conditions) may mediate the binding of acrosome-reacted sperm to the ZP (Wassarman 1988; Bleil et al. 1988). Upon fertilization, the egg effects a modification of ZP2 to a form called ZP2f (Wassarman 1988, 1990) which is brought about by the action of a protease most likely secreted from the egg as a consequence of cortical granule exocytosis (Moller
Signal Transduction in Mammalian Spermatozoa 163
and Wassannan 1989). ZP2r has a Mr = 120000 under nonreducing condition that shifts under reducing conditions to Mr = 90 000, suggesting that the proteolysis of the ZP2 molecule results in the generation of fragments that are held together by disulfide bonds. The biological consequence of the conversion of ZP2 to ZP2f is that ZP2f no longer will bind to acrosome-reacted sperm (Bleil et a!. 1988). ZP3 (Mr = 83 000) accounts for both the spenn binding and the acrosome reaction-inducing activities of the ZP of unfertilized eggs (Wassarman 1988, 1990). The sperm-binding activity appears to be conferred by O-linked carbohydrate moieties and not by its polypeptide chain (Flonnan et a!. 1984; Flonnan and Wassannan 1985). a-Linked tenninal galactose residues at the nonreducing tennini of these O-linked oligosaccharide chains playa critical role in this binding activity (Bleil and Wassannan 1988). Further characterization of ZP3 glycopeptides generated by limited protease digestion has demonstrated that the sperm-binding activity of ZP3 is most likely associated with the carboxy-tenninal half of the ZP3 polypeptide, a region that is suggested to be highly glycosylated (Rosiere and Wassarman 1992). The acrosome reaction-inducing activity ofZP3, in contrast, appears to be conferred by both the carbohydrate and protein portions of the molecule, although the exact nature of the interaction between the protein and carbohydrate required for biological activity is not clear at this time (Flonnan et a!. 1984). Fertilization is associated with a loss of both the sperm-binding and acrosome reaction-inducing activities of the ZP3 molecule (Wassannan 1988, 1990). The loss of these two important biological activities is associated with a minor biochemical modification of the ZP3 molecule since the electrophoretic mobility of ZP3 from fertilized eggs is similar to that of ZP3 from unfertilized eggs. It is probable that the loss of carbohydrate from ZP3 as a consequence of a specific cortical granule glycosidase(s) released following the cortical reaction accounts for this loss of activity.
Although the ZP of a number of other species have been characterized to various degrees at the biochemical level (e.g., pig, bull, rabbit, human, rhesus monkey, rat, cat, dog), little is known about the biological activities of the individual glycoproteins comprising these ZP (Dunbar et a!. 1991). Some of these species, however, possess homologues to ZP3, consistent with the potential conservation of ZP3 and its function. For example, the human ZP3 homologue has been cloned and displays a high degree of sequence homology to the mouse (Chamberlin and Dean
164 G.S. Kopf et al.
1990). Ringuette et al. (1986) have demonstrated conserved sequences in the genomic DNA of human, cow, rat, rabbit, and pig using ZP3-specific probes, and genomic cloning of the hamster ZP3 gene reveals remarkable similarity in gene organization and in polypeptide structure to the mouse ZP3 gene and gene product, respectively (Kinloch et al. 1990). Characterization of the hamster ZP has demonstrated the presence of three major glycoproteins (Moller et al. 1990). Hamster ZP3 (Mf = 56 000), and not ZP1 and ZP2, possesses sperm-binding activity, and the intact hamster ZP possesses the ability to induce the acrosome reaction of hamster sperm (Moller et al. 1990). It is likely, therefore, that similar biological activities are associated with the homologous ZP1, ZP2, and ZP3 glycoproteins in other mammals.
8.3.2 Interaction of Sperm with the Zona Pellucida
The properties of ZP1, ZP2, and ZP3 from both unfertilized and fertilized eggs provide a framework with which to formulate a model to explain the interaction of sperm with the intact ZP. Since only acrosome-intact mouse sperm bind to the ZP (Saling et al. 1979), the existence of a specific receptor(s) for ZP3 on the plasma membrane overlying the sperm acrosome that mediates sperm binding and the induction of acrosomal exocytosis has been proposed. Since acrosome-reacted sperm do not interact with ZP3, secondary interactions of these sperm with the ZP would then occur through the interaction of a putative receptor(s) for ZP2 on the sperm inner acrosomal membrane. Upon penetration of the ZP by acrosome-reacted sperm, these cells then traverse the perivitelline space and then bind and fuse with the plasma membrane of the egg. Subsequent to sperm-egg fusion, the egg undergoes the cortical granule reaction, which results in the release of cortical granule-associated enzymes (Wassarman 1988). These enzymes convert ZP2 to ZP2f and modify ZP3, such that acrosome-intact sperm no longer bind to the ZP (via ZP3) and acrosome-reacted sperm that are bound to the ZP (via ZP2) no longer interact and penetrate the ZP since they are unable to establish secondary binding interactions with ZP2f. Such egg-induced modifications of this extracellular matrix constitute the ZP block to polyspermy. Inherent in such a model is the highly specific and coordinated nature of the interactions of acrosome-intact and acrosome-reacted sperm with
Signal Transduction in Mammalian Spermatozoa 165
ZP3 and ZP2, respectively. Such interactions would presumably be mediated via specific sperm-associated receptors for these extracellular matrix glycoproteins.
The species specificity of gamete adhesion, a prerequisite to acrosomal exocytosis, strongly implicates the existence of molecules on the sperm surface with novel domains that must interact with acrosome reaction-inducing ligands of the ZP in a specific and potent manner to elicit biological responses. Presently, putative sperm-associated receptors for known molecules that induce the acrosome reaction have not been unequivocally identified in any species, although there appears to be more available information in the mouse than in other species. Identification of such receptors has been hampered by the lack of information pertaining to the precise identity of the active sperm adhesion and acrosome reaction-inducing moieties of the ZP, as well as the seemingly complex nature of the interactions of such molecules with the sperm surface. Although some candidates have been proposed, no one candidate has satisfied all of the properties that one would expect for such a receptor.
There are several lines of evidence which support that idea that sperm-ZP interaction in the mouse may occur through specific receptormediated events. ZP3 possesses a number of properties which make it ideally suited as a ligand to mediate the initial steps of sperm-egg interaction proper (e.g., sperm binding) and subsequent sperm activation (e.g., induction of the acrosome reaction). Although conserved at the genomic level in a variety of species (Ringuette et al. 1988), ZP3 subserves very specific functions as a component of the egg-associated extracellular matrix (Wassarman 1988, 1990). (I) ZP3 is synthesized only by the growing oocyte. (2) Little apparent amino acid sequencc homology exists between ZP3 and any other known proteins or glycoproteins thus far examined (Ringuette et al. 1988). (3) The ordered crosslinking of ZP2/ZP3 heterodimers by ZP I gives rise to structural domains that ensures that the ZP3 ligand is immobilized and functions only at short distances. (4) Both the sperm-binding and acrosome reaction-inducing activities of ZP3 are observed in the nanomolar range (Florman and Wassarman 1985; Bleil and Wassarman 1986). (5) Mouse sperm appear to possess complementary binding sites (receptors?) for ZP3 that are localized over the acrosomal cap region and are present in numbers (10000-50 000 binding sites/cell) similar to that observed for
166 G.S. Kopf et al.
receptor numbers in many hormonally responsive cells (Bleil and Wassarman 1986; Vazquez et al. 1989; Mortillo and Wassarman 1991). In addition, specific binding of 125I-labeled ZP2 to the inner acrosomal membrane of acrosome-reacted, but not acrosome-intact, mouse sperm has been demonstrated (Wassarman 1988; Bleil et al. 1988), suggesting that this particular membrane may possess specific receptors for this ZP glycoprotein.
These aforementioned properties of purified ZP3 are consistent with its role as the physiologically relevant ligand that accounts for both the sperm-binding and acrosomal exocytosis-inducing properties of the structurally intact ZP. However, it should be emphasized that care must be exercised in construing the properties and mode of action of solubilized and purified ZP3 as being identical to those of an immobilized and crosslinked ZP3 associated with a three-dimensional extracellular matrix. This word of caution should apply to all studies using solubilized and purified ZP components, since the importance of spatial constraints in mediating these effects on sperm are not fully appreciated. Such constraints could certainly influence the interpretation of data pertaining to sperm-associated receptors for ZP3.
8.3.3 Identity of Sperm-Associated Receptors for the Zona Pellucida
Putative sperm-associated receptors for ZP3 have been described using two experimental approaches. The first, and more direct, approach involves the identification of components of the sperm surface that interact directly with ZP3. The binding of 125I-Iabeled ZP3 and gold-labeled ZP3 to mouse sperm by whole mount autoradiography and transmission electron microscopy, respectively, has demonstrated that binding is associated with the plasma membranes overlying the acrosomal and postacrosomal regions of the sperm head of acrosome-intact sperm, as well as the postacrosomal region of the sperm head of acrosome-reacted sperm (Bleil and Wassarman 1986; Mortillo and Wassarman 1991 ).125I-Iabeled ZP3 binding is competed by unlabeled ZP3, but not ZP2 (Bleil and Wassarman 1986), and ZP3 binding does not occur on somatic cells (Bleil and Wassarman 1986; Mortillo and Wassarman
Signal Transduction in Mammalian Spermatozoa 167
1991). These results suggest that specific sperm-associated binding sites for ZP3 exist in discrete and appropriate cellular domains of this highly differentiated cell. More recently, purified ZP3 or glycopeptides of ZP3 possessing sperm-binding activity were shown to specifically crosslink to a Mr = 56 000 protein of acrosome-intact mouse sperm (Bleil and Wassarman 1990). This protein interacts specifically with ZP3-, but not ZP2-, affinity columns, and the binding of the protein to the ZP3 column has characteristics of high-affinity interactions. This experimental approach is promising with regard to establishing the molecular identity of putative receptors for ZP3 on the sperm surface.
Ley ton and Saling (1989a) demonstrated that antiphosphotyrosine antibodies react with mouse sperm plasma membrane proteins of Mr = 52 000, 75 000, and 95 000. Indirect immunofluorescence using these antibodies demonstrated positive immunoreactivity associated with the acrosomal region of the sperm head. 1251-labeled ZP3 binds to a Mr = 95 000 sperm protein on nitrocellulose blots following electrophoretic transfer, and these investigators conclude that this protein is the same protein that reacts with the antiphosphotyrosine antibodies. Recently, these investigators have demonstrated that the Mr = 95 000 protein, when electroeluted from sodium dodecyl sulfate (SDS) gels and incubated with [y -32P1ATP, becomes phosphorylated (Ley ton et al. 1992). The phosphorylation of this protein is increased upon addition of solubilized ZP, and this phosphorylation is reduced upon the addition of tyrphostin RG-50864, an inhibitor of tyrosine kinases. This particular inhibitor was also demonstrated to inhibit the ZP-induced acrosome reaction. It is suggested that the Mr = 95 000 protein may be a receptor for ZP3 that possesses tyrosine kinase activity.
Recent studies by Kalab et a!. (1994), however, call into question the identity and function of this Mr = 9S 000 phosphotyrosine-containing protein. These investigators purified this protein and then subjected it to limited tryptic digestion and subsequent amino acid analysis. Three sequenced peptides revealed 100% amino acid identity to a mouse hepatoma hexokinase (Arora et al. 1(90). The purified protein, which migrated at Mr = 116 000 under reducing conditions (p95/116). reacted with an antiserum to the purified rat brain hexokinase, type I, and comigrated on SDS polyacrylamide gel electrophoresis (PAGE) with the purified rat brain enzyme under both nonreducing and reducing conditions. Unlike p95/116, the rat brain enzyme was not a phosphoty-
168 G.S. Kepf et al.
rosine-containing protein. The p95/116 protein could be immunoprecipitated with the hexokinase antiserum or an O-phosphotyrosine antibody. Limited tryptic digestion of the purified p95/116 and the rat brain enzyme generated subsets of identical peptides that reacted with the hexokinase antiserum. However, p95/116 also contained phosphotyrosine-containing peptides that were not present in the rat brain hexokinase. When different mouse tissues were probed with the hexokinase antiserum all tissues, with the exception of liver, contained immunoreactive protein. In contrast, only sperm and testis possessed a phosphotyrosine-containing form of hexokinase. These data suggest that the germ cell component of the testis possesses a unique tyrosine-phosphorylated form of hexokinase. The role of this protein as a putative ZP3 receptor must, therefore, be carefully reexamined.
A sperm surface ~-galactosyltransferase (Gal-transferase) activity has been postulated to mediate the binding of mouse sperm to the ZP by binding oligosaccharide residues on ZP3. This enzyme has been implicated in mediating cell--cell and cell--extracellular matrix interactions in other cells. Many of the experiments supporting the role of this enzyme in sperm-ZP interaction are based on indirect studies where inhibitors of, and antibodies to, the enzyme were shown to block sperm-ZP binding (Shur and Hall 1982; Lopez et al. 1985). Recently, however, Miller et al. (1992) have demonstrated that the mouse sperm Gal-transferase can galactosylate ZP3, but not ZPl and ZP2, in vitro. Following the acrosome reaction, the enzyme loses its ability to utilize ZP3 as a substrate, and the ZP from fertilized eggs loses its ability to serve as a substrate for galactosylation.
A second indirect approach to identifying ZP3 receptors involves the delineation of sperm-associated binding sites for the ZP glycoproteins by examining the ability of specific agents to interfere with sperm-ZP interaction. Many of these studies utilized whole ZP and not the purified glycoprotein components. Three sites have been differentiated, all based on assays of enzymatic activities, which themselves do not appear to be involved in the sperm-ZP binding process. Sperm surface-associated fucosyl transferase activity, mannosidase activity, and protease inhibitor-sensitive sites have been implicated in the binding of sperm to the ZP, presumably through ZP3. However, the relationship of these sites to one another and to those sites defined by ZP3 binding, as described above, is not clear at the present time. A detailed discussion of these
Signal Transduction in Mammalian Spermatozoa 169
sites is presented elsewhere (Kopf and Gerton 1990; Storey and Kopf 1991 ).
At first glance, the presence of multiple ZP3 receptor candidates may be confusing since there is presently no overwhelming evidence to support one candidate over another. It is possible, however, that multiple receptors might function to mediate the dual biological functions of ZP3 (e.g., sperm binding and induction of the acrosome reaction), and that the interaction of these components with ZP3 , as well as with one another, must occur in an ordered fashion on the sperm surface to first establish binding of sperm to ZP3 that then permits signal transduction to occur in order to initiate acrosomal exocytosis (see Fig. 2). The concept of multiple, interactive protein domains giving rise to ligand binding and signal transduction is not a novel concept in cell biology. In addition, these domains need not necessarily be identical. In the case of sperm-ZP3 interaction this possibility has originated from data pertaining to the nature of the interaction between the ZP (or ZP3) and the sperm surface to mediate both sperm binding and the induction of the acrosome reaction. Several laboratories have established the fact that sperm binding to the ZP, and the ZP-induced acrosome reaction are two independent processes and that the ZP3-induced acrosome reaction in the mouse appears to consist of discrete and independently regulated events (Kopf and Gerton 1990; Storey and Kopf 1991). Moreover, studies from a number of laboratories using different approaches have provided evidence that the interaction of ZP3 with the sperm surface may occur in a multivalent or cooperative fashion, and that multiple interactions followed by possible receptor aggregation may ultimately lead to signal transduction and acrosomal exocytosis (Bleil and Wassarman 1983; Kopf et al. 1989; Ley ton and Saling 1989b; Kligman et al. 1991). These observations would certainly be consistent with a model evoking multiple sperm protein domains that form a functional ZP3 receptor signal transduction complex that is capable of transducing intracellular signals to regulate the acrosome reaction (Fig. 2). It is possible that the ZP, itself an extracellular matrix, possesses the ability to aggregate receptors on the sperm surface, as this has been demonstrated in other systems (Godfrey et al. 1988).
170 G.S. Kopt et al.
~ -----,cmu--Sequential Binding PLASMA ofZP3 Ligands MEMBRANE
with Sperm Receptor Proteins ~
1, bi&~1 Ligand-Receptor
Induced
PLASMA MEMBRANE
Aggregation of ~ Receptor Proteins -----------"'-------------
1 Formation ofa
Sig/Ull Transducing Receptor Complex
and Transmembrane Sig/Ulling
ZP 3
t
PLASMA MEMBRANE
cAMP Ca +;Na +, K +, pH 1P3 ,DAG, AA, lyso·PC, choline, PA
ACROSOME REACTION
8.3.4 Zona Pellucida-Mediated Signal Transduction
The mechanism by which specific regulators of sperm function effect signal transduction, or informational flow across the sperm plasma membrane, to modulate intracellular second messenger systems leading to appropriate cellular responses is only starting to be understood. It is predicted that signal transduction processes activated through Iigandreceptor interactions are similar to those observed in somatic cells. This
Signal Transduction in Mammalian Spermatozoa 171
has certainly been shown to be the case in marine invertebrates (Ward and Kopf 1993).
The mouse remains the model system in which most is known about the signal transduction mechanisms modulating sperm function in response to specific ligands. As previously stated, the unique structure of ZP3, its biological potency, and the probable existence of complementary ZP3 receptor(s) on the sperm surface satisfy a number of criteria required to control specific cell-cell recognition events in a receptor-mediated fashion. As in many somatic cells, sperm guanine nucleotide-binding regulatory proteins (G proteins) playa critical role as signal-transducing elements in mediating ZP3-mediated acrosomal exocytosis. Sperm from all species studied thus far (invertebrates; mammals) possess G proteins, as assessed by numerous criteria (Bentley et al. 1986; Kopf et a!. 1986; Glassner et al. 1991; Kamik et a!. 1992). Mammalian sperm contain G proteins of the Gi class, and mouse sperm contain all three subtypes of Gi, namely, Gil, Gi2 and Gi3 (Glassner et al. 1991). These assignments have been made based on the ability of the ex subu-
Fig. 2. Model for the interaction of the zona pellucida glycoprotein, ZP3, with the plasma membrane overlying the acrosome of mouse spermatozoa to mediate sperm binding and acrosomal exocytosis. In this model, the ZP3 molecule is composcd of multiple "functional ligands" ( "'rl..Jtr ) which interact with complemcntary cell surface receptors/binding proteins present in the sperm plasma mcmbranc. These ZP3-associated ligands arc shown as being different from one another in this model, but this does not necessarily have to be the case. Moreover, three ZP3-associated ligands are shown although the actual numbers of ligands involved are not known and are shown in this fashion for illustrative purposes only. The proper interaction of ZP3 with the sperm surface requires the sequential binding of these ligands with sperm-associated receptors. Once these interactions are established. signal transduction is effected by the formation of a functional signal transducing complex which forms in rcsponse to ligand receptor-induced aggregation of receptor proteins. Acrosomal cxocytosis is then initiated in response to changes in second messengers/ionic conductance that are regulated in response to receptor-mediated signal transduction. In this model a variety of effector systems are shown to be targets for the ex and/or ~y subunits of heterotrimeric G proteins, as indicated by the dashed arrows. cAMP, adenosine-3',5' -cyclic monophosphate; [P3, inositol I ,4,5-trisphosphate; AA, arachidonic acid; Iyso-PC, Iysophosphatidylcholine; PA, phosphatidic acid
172 G.S. Kopt et al.
nits of these heterotrimeric proteins to serve as substrates for pertussis toxin (PTX)-catalyzed ADP-ribosylation, the molecular weights of the ex subunits, Cleveland digests of the ex subunits, and immunoreactivity with antipeptide antisera generated against conserved Gia domains (Kopf et al. 1986; Glassner et al. 1991). Mouse spenn also contain G proteins that are not PTX substrates, namely Gz (Glassner et al. 1991) and Gq (Visconti and Kopf, unpublished observations), but the functions of these proteins have not been examined. Unlike somatic cells, spenn do not appear to contain a G protein with properties similar to Gs
(Kopf and Gerton 1990). Immunocytochemical studies have demonstrated that G proteins in bovine (Garty et al. 1988), as well as mouse and guinea pig (Glassner et al. 1991), spenn are present in the acrosomal region of these cells; these G proteins appear to be of the Gi class.
The physiological role of the spenn Gi proteins in the ZP-mediated acrosome reaction has been examined in the mouse, bull, and human (Endo et al. 1987, 1988; Flonnan et al. 1989; Lee et al. 1992). Functional inactivation of the Gi proteins by PTX treatment of the cells does not affect the ability of the spenn to bind and to interact with their homologous ZP, but inhibits the bound spenn from undergoing acrosomal exocytosis. The inhibitory effect of PTX on this event is strictly confined to acrosomal exocytosis induced by ZP3 (or ZP in the human and cow), whereas exocytosis that occurs spontaneously in a small population of cells, in response to a nonspecific agent such as the divalent cation ionophore A23187 or in response to progesterone is insensitive to this treatment (Endo et al. 1988; Lee et al. 1992; Tesarik et al. 1992). These data suggest that spenn Gi proteins may act as critical signal-transducing elements downstream from initial spenn-ZP binding events to mediate acrosomal exocytosis by an extracellular matrix component, ZP3.
If the ZP (ZP3)-induced acrosome reaction is an example of stimulus-secretion-coupling that occurs in a receptor-mediated fashion, it would be predicted that receptor-G protein interaction subsequently leads to the generation of intracellular second messengers and/or the modulation of ionic changes within the spenn. Two criteria should be met in order to establish the signal-transducing function of a G protein in such a system: First, occupation of putative ZP3 receptors should result in G protein activation in a manner described for other ligandreceptor-G protein interactions; second, resultant G protein activation
Signal Transduction in Mammalian Spermatozoa 173
should then modulate the intracellular effector systems that are required for acrosomal exocytosis. There is experimental evidence that supports both of these criteria.
Recently, it has been demonstrated that solubilized mouse egg ZP stimulate high-affinity GTPase activity and specific GTPy35S binding in both permeabilized, capacitated mouse sperm (Wilde et al. 1992) and partially purified membranes obtained from these cells (Ward et al. 1992). These endpoints of G protein activation are stimulated by the ZP in a concentration dependent manner that is blocked by PTX pretreatment, suggesting that sperm Gj proteins are being activated under these conditions. The component of the intact ZP that stimulates Gj is ZP3, and not ZPI or ZP2. Moreover, ZP from fertilized eggs have lost their ability to stimulate Gj, consistent with the loss of the biological activities of ZP3 following fertilization. Incubation of ZP with sperm membranes results in the preferential acti vation of Gi 1 and Gi2, as assessed by a reduction in PTX-catalyzed l32p]ADP ribosylation in immunoprecipitates obtained following incubation of the membranes with Gja sUbtypespecific antisera (Ward, Storey and Kopf, unpublished observations). These data suggest that the ZP-mediated stimulation of Gj involves the preferential stimulation of Gj I and Gj2. More recently, it has been demonstrated that a high speed (100 000 x g) supernatant fraction obtained following detergent extraction of sperm membranes displays ZP-mediated G protein activation with properties similar to that seen in the intact membranes (Ning and Kopf, unpublished observations). These data suggest that a functional ZP-G protein signaling complex can be isolated in a soluble form, thus making it now possible to isolate and identify the components of such a complex. This complex should include the receptor(s) for ZP3.
As previously discussed, there are a number of candidates that have been suggested as putative ZP3 receptors. The activation of G proteins in sperm by ZP3 suggests the presence of functional G protein receptors that modulate ZP3 function. If one accepts the hypothesis that multiple interactive proteins could undergo aggregation and form a signaling complex to mediate the responses of sperm to ZP3, it is possible that a heretofore unidentified member of the G protein class of cell surface receptors may be involved in some aspect of ZP3-mediated signal transduction. A superfamily of such receptors has been identified which contains seven transmembrane-spanning domains and includes the mu-
174 G.S. Kopf et al.
scarinic acetylcholine receptor, the (X- and p-adrenergic receptors, and the dopamine receptor (Dohlman et al. 1991). In two recent reports, mRNAs corresponding to cDNAs encoding novel, putative G proteincoupled receptors, which may be part of this family, have been found in pachytene spermatocytes and round spermatids (Meyerhof et al. 1991; Parmentier et al. 1992). To date, however, no ligand for these putative receptors has been identified, nor has it been established that the corresponding proteins are expressed. As mentioned above, work is ongoing in this laboratory to identify the signal transduction complex that forms in response to ZP3-sperm binding. Such studies will go a long way to identify such a putative G protein-coupled ZP3 receptor.
Alternatively, some of the receptor candidates described above might couple in some manner to G proteins. The amino acid sequence for the M r = 56 000 protein (Bleil and Wassarman 1990) has not yet been published so it is difficult to assess whether this protein might be a candidate receptor that could couple to G proteins. There is considerable evidence to suggest that receptors possessing tyrosine kinase activity can interact either directly (Liang and Garrison 1991; Yang et al. 1991) or indirectly (Imamura and Kufe 1988; Huang and Ives 1989; Luttrell et al. 1990), so it is possible that putative receptor tyrosine kinases could couple to G proteins to mediate signal transduction in sperm. Although the Mr = 95000 phosphotyrosine-containing protein has been identified as a unique form of hexokinase (Kalab et al. 1994), its ability to function as a potential ZP3 receptor tyrosine kinase must be carefully reevaluated.
8.3.5 Intracellular Effectors Regulating the Zona Pellucida-Induced Acrosome Reaction
Since there is substantial evidence that acrosomal exocytosis is a highly regulated process, initiated by egg/reproductive tract-specific ligands whose effects are likely mediated by cell surface receptors and signaltransducing G proteins (at least in mammals), it is likely that intracellular regulation of acrosomal exocytosis is similar to other receptor-mediated exocytotic events. Such intracellular signals include changes in ionic conductance, changes in cyclic nucleotide metabolism and changes in phospholipid metabolism (Fig. 2). It would also be predicted that
Signal Transduction in Mammalian Spermatozoa 175
these intracellular effector systems would be modulated in a receptordependent fashion. Although there have been numerous reports describing ionic and/or second messenger systems purported to playa role in the induction of the acrosome reaction, there is little information linking such effectors to ligands in a receptor-dependent fashion due to the paucity of knowledge of the receptors and the signal transducers. Furthermore, a description of effector systems correlated with the induction of acrosomal exocytosis cannot necessarily be equated with cause-and-effect. A consideration of what is currently known regarding the nature of sperm intracellular signaling during the acrosome reaction in response to the aforementioned acrosome reaction-inducing molecules will be considered below.
The nature of the intracellular signal pathways activated in response to sperm-ZP (or ZP3) interaction are only starting to be investigated. Studies in both mouse and bull sperm have demonstrated that elevations in intracellular Ca2+ and intracellular pH represent some of the earliest responses of sperm incubated with ZP or ZP3 (Lee and Storey 1985, 1989; Endo et al. 1988; Florman et al. 1989; Kligman et al. 1991; Storey et al. 1992). Such studies were performed with Ca2+ and pH indicator dyes and have reported localized changes to the acrosomal region. In bovine sperm, the ZP-induced Ca2+ entry, as well as sperm membrane potential and the acrosome reaction, are dependent upon membrane depolarization, and the Ca2+ uptake and acrosome reaction are inhibited by antagonists of voltage-dependent Ca2+ channels (Florman et al. 1992). These data suggest that mammalian sperm, like marine invertebrate sperm, contain voltage-dependent Ca2+ channels that are activated in response to biological effectors of the acrosome reaction. PTX inhibits the ZP (and ZP3)-induced pH changes in mouse sperm (Endo et al. 1988), as well as the ZP-induced pH and Ca2+ changes in bull sperm (Florman et a1. 1989), indicating that sperm Gi proteins may regulate such ionic changes. Incubation of bovine sperm under depolarizing conditions, which would activate such voltage-dependent Ca2+ channels, bypasses the inhibitory effects of PTX on the acrosome reaction, suggesting that Gi proteins might regulate such Ca2+ channels indirectly (Florman et a1. 1992).
Alterations in phospholipid metabolism and/or cyclic nucleotide metabolism have also been suggested to play important intermediary roles in the sperm acrosome reaction (Kopf and Gerton 1990). Biologically
176 G.S. Kopf et al.
active phorbol diesters and diacylglycerols alter the kinetics of the ZP-mediated acrosome reaction in mouse sperm, thus suggesting that this exocytotic event could be regulated in some manner by protein kinase C (Lee et al. 1987). Neither the products of phospholipase C turnover (e.g., inositol 1,4,5-trisphosphate, IP3. and sn-l,2 diacylglycerol) nor the role of other phospholipases (A2 or D) have been examined yet in sperm challenged with ZP or ZP3. Noland et al. (1988) have reported that solubilized ZP from mouse eggs cause transient elevations in sperm cyclic AMP (cAMP) concentrations that are dependent on the presence of extracellular Ca2+. These cAMP elevations precede and are correlated with the induction of the acrosome reaction by the ZP, suggesting that cAMP may be a potential participant in the signaling pathway leading to acrosomal exocytosis. These ZP-induced cAMP changes appear to be mediated by the activation of the sperm adenylyl cyclase (Leclerc and Kopf 1993). It will be of interest to determine whether such intracellular signaling systems are coupled to sperm OJ proteins, since these second messenger systems are coupled in a receptor-mediated fashion to 0 proteins in other cell types.
8.4 Conclusions
We are only starting to understand and to appreciate some of the complexities involved in species-specific intercellular communication, gamete recognition, and gamete activation. There are still significant gaps in our knowledge with regard to the identity of the biological ligands that regulate sperm function, the identity and mode of action of the receptors for these regulatory ligands, and the mechanism by which ligand receptor occupancy leads to intracellular signal transduction and the appropriate biological response. Since these events appear to be receptor mediated, it is clear that substantial strides toward an understanding of these processes will only occur following the unequivocal identification of these receptors and receptor complexes. This task should be aided by the identification of the specific domains of the biologically active ligands that bind to such receptors as well as an understanding of the signal transduction mechanisms mediating such interactions. Understanding the nature of these receptors will provide
Signal Transduction in Mammalian Spermatozoa 177
the basis for experimental approaches designed to determine receptor dynamics (e.g., cooperative binding of ligands; receptor aggregation; ligand binding to multiple receptor subunits) that playa role in initiating signaling events that modulate sperm functions. Other areas of experimental focus must include the determination of the mechanism by which these ligands effect information flow across the plasma membrane to activate specific effector systems and the establishment of causeand-effect relationships between effectors and biological responses.
Acknowlcdgmcnts. I would like to acknowledge the members of my laboratory, whose hard work and dedication are greatly appreciated. This work is supported by the NIH (HD06274, HD22732; HD28514).
References
Arora KK, Fanciulli M, Pederson PL (1990) Glucose phosphorylation in tumor cells: cloning, sequencing, and overexpression in active form of a fulllength cDNA encoding a mitochondrial bindable form of hexokinase. J Bioi Chern 265:6481-6488
Bentley JK, Garbers DL, Domino SE, Noland TD, VanDop C (1986) Spermatozoa contain a guanine nucleotide binding protein ADP-ribosylated by pertussis toxin. Biochem Biophys Res Commun 138:728-734
Bleil JD, Wassarman PM (1980a) Structure and function of the zona pellucida: identification and characterization of the proteins of the mouse oocyte's zona pellucida. Dev Bioi 76: 185-202
Bleil JD, Wassarman PM (1980b) Synthesis of zona peUucida proteins by denuded and follicle-enclosed mouse oocytes during culture in vitro. Proc Natl Acad Sci USA 77:1029-1033
Bleil JD, Wassarman PM (1983) Sperm-egg interactions in the mouse: sequence or events and induction of the acrosome reaction by a zona pellucida glycoprotein. Dev Bioi 95:317-324
Bleil JD, Wassarman PM (1986) Autoradiographic visualization of the mouse egg's sperm receptor bound to sperm. J Cell Bioi 102:1363-1371
Bleil JD, Wassarman PM (1988) Galactose at the non-reducing terminus of 0-linked oligosaccharides of mouse egg zona pellucida glycoprotein ZP3 is essential for the glycoprotein's sperm receptor activity. Proc Natl Acad Sci USA 85:6778-6782
178 G.S. Kopf et al.
Bleil JD, Wassarman PM (1990) Identification of a ZP3-binding protein on acrosome-intact mouse sperm by photoaffinity crosslinking. Proc Natl Acad Sci USA 87:5563-5567
Bleil JD, Greve JM, Wassarman PM (1988) Identification of a secondary sperm receptor in the mouse egg zona pellucida: role in maintenance of binding of acrosome-reacted sperm to eggs. Dev BioI 128:376-385
Bradley MP, Garbers DL (1983) The stimulation of bovine caudal epididymal sperm forward motility by bovine cumulus-egg complexes in vitro. Biochern Biophys Res Commun 115:777-787
Bronson R, Hamada Y (1977) Gamete interactions in vitro. Fertil Steril 28:570-576
Chamberlin ME, Dean J (1990) Human homolog of the mouse sperm receptor. Proc Natl Acad Sci USA 87:6014-6018
Cummins JM, Yanagimachi R (1982) Sperm-egg ratios and the site of the acrosome reaction during in vivo fertilization in the hamster. Gamete Res 5:239-256
Devreotes PN, Zigmond SH (1986) Chemotaxis in eukaryotic cells: a focus on leukocytes and Dictyostelium. Annu Rev Cell Bioi 4:649-686
Dohlman HG, Thorner J, Caron MG, Lefkowitz RJ (1991) Model systems for the study of seven-transmembrane-segment receptors. Annu Rev Biochem 60:653-688
Dunbar BS, Prasad SV, Timmons TM (1991) Comparative structure and function of mammalian zonae pellucidae. In: Dunbar BS, O'Rand MG (eds) A comparative overview of mammalian fertilization. Plenum, New York, pp 7-114
Eisenbach M, Ralt D (1992) Precontact mammalian sperm-egg communication and role in fertilization. Am J Physiol 262 (Cell Physiol 31 ):C 1095-C II 01
Endo Y, Lee MA, Kopf GS (1987) Evidence for the role of a guanine nucleotide-binding regulatory protein in the zona pellucida-induced mouse sperm acrosome reaction. Dev BioI 119:210-216
Endo Y, Lee MA, Kopf GS (1988) Characterization of an islet activating protein-sensitive site in mouse sperm that is involved in the zona pellucida induced acrosome reaction. Dev BioI 129: 12-24
Flechon JE, Hunter RHF (1981) Distribution of spermatozoa in the utero-tubal junction and isthmus of pigs, and their relationship with the luminal epithelium after mating: a scanning electron microscope study. Tissue Cell 13:127-139
Florman HM, Babcock DF (1990) Progress towards understanding the molecular basis of capacitation. In: Wassarman PM (ed) Elements of mammalian fertilization 1. Basic concepts. CRC uniscience series. CRC Press, Boca Raton, pp 105-132
Signal Transduction in Mammalian Spermatozoa 179
Florman HM, Wassarman PM (1985) O-linked oligosaccharides of mouse egg ZP3 account for its sperm receptor activity. Cell 41 :313-324
Florman HM, Bechtol KB, Wassarman PM (1984) Enzymatic digestion of the functions of the mouse egg's receptor for sperm. Dev Bioi 106:243-255
Florman HM, Tombes RM, First NL, Babcock OF (1989) An adhesion-associated agonist from the zona pellucida activates G protein-promoted elevations of internal Ca2+ and pH that mediate mammalian sperm acrosomal exocytosis. Dev Bioi 135:133-146
Florman, HM, Corron ME, Kim TO, Babcock OF (1992) Activation of voltage-dependent calcium channels of mammalian sperm is required for zona pellucida-induced acrosomal exocytosis. Dev BioI 152:304-314
Garty NB, Galiani 0, Aharonheim A, Ho YK, Phillips OM, Dekel N, Salomon Y (1988) G-proteins in mammalian gametes: an immunocytochemical study. J Cell Sci 91:21-31
Glassner M, Jones J, Kligman I. Woolkalis MJ, Gerton GL, Kopf GS (1991) Immunocytochemical and biochemical characterization of guanine nucleotide-binding regulatory proteins in mammalian spermatozoa. Dev Bioi 146:438-450
Gnessi L, Ruff MR, Fraioli F, Pert CB (1985) Demonstration of receptor-mediated chemotaxis by human spermatozoa: a novel quantitative bioassay. Exp Cell Res 161:219-230
Godfrey EW, Dietz ME, Morstad AL, Wallskog PA, Yorde DE (1988) Acetylcholine receptor-aggregating proteins are associated with the extracellular matrix of many tissues in Torpedo. J Cell BioI 106: 1263-1272
Harpcr MJK (1973) Stimulation of sperm movement from the isthmus to the site of fertilization in the rabbit oviduct. Bioi Reprod 8:369-377
Huang CL, Ives HE (1989) Guanosine 5' -O-(3-thiotriphosphate) potentiates both thrombin- and platelet derived growth factor-induced inositol phosphate release in permeabilized vascular smooth muscle cells. J Bioi Chern 264:4391-4397
Hunter RHF, Nichol R (1983) Transport of spermatozoa in the sheep oviduct: preovulatory sequestering or cells in the caudal isthmus. J Exp Zool 228: 121-128
Hunter RHF, Nichol R (1986) A preovulatory temperature gradient between the isthmus and ampulla of pig oviducts during the phase of sperm storage. J Reprod Fertil 77:599--606
Hunter RHF, Wilmut I ( 1984) Sperm transport in the cow: periovulatory redistribution of viable cells within the oviduct. Reprod Nutr Dev 24:597-608
Imamura K, Kufe 0 (1988) Colony stimulating factor I-induced Na+ flux into human monocytes involves activation of a pertussis toxin-sensitive GTP binding protein. J Bioi Chern 263: 14093-14098
180 G.S. Kopf et al.
Iqbal M, Shivaji S, Vijayasarathy S, Balaram P (1980) Synthetic peptides as chemoattractants for bull spermatozoa structure activity correlations. Biochern Biophys Res Commun 96:235-242
Kalab P, Visconti P, Leclerc P, KopfGS (1994) p95, the major phosphotyrosine-containing protein in mouse spermatozoa, is a form of hexokinase with unique properties. J BioI Chern (in press)
Kamik NS, Newman S, Kopf GS, Gerton GL (1992) Developmental expression of G protein subunits in mouse spermatogenic cells: evidence that Gai is associated with the developing acrosome. Dev BioI 152:393-402
Kinloch RA, Ruiz-Seiler B, Wassarman PM (1990) Genomic organization and polypeptide primary structure of zona pellucida glycoprotein hZP3, the hamster sperm receptor. Dev BioI 142:414-421
Kligman I, Glassner M, Storey BT, Kopf GS (1991) Zona pellucida-mediated acrosomal exocytosis in mouse spermatozoa: characterization of an intermediate stage prior to the completion of the acrosome reaction. Dev BioI 145:344-355
Kopf GS, Gerton GL (1990) The mammalian sperm acrosome and the acrosome reaction. In: Wassarman PM (ed) Elements of mammalian fertilization I: basic concepts. CRC Press, Boca Raton, pp 153-203 (CRC uniscience series)
Kopf GS, Woolkalis MJ, Gerton GL (1986) Evidence for a guanine nucleotide-binding regulatory protein in invertebrate and mammalian sperm: identification by islet-activating protein-catalyzed ADP-ribosylation and immunochemical methods. J BioI Chern 261:7327-7331
Kopf GS, Endo Y, Mattei P, Kurasawa S, Schultz RM (1989) Egg-induced modifications of the murine zona pellucida. In: Nuccitelli RL, Clark WH, Cherr GN (eds) Mechanisms of egg activation. Plenum, New York, pp 249-272
Leclerc P, Kopf GS (1993) Regulation of mouse sperm adenylyl cyclase by solubilized zonae pellucidae and guanine nucleotides. BioI Reprod 48 [SuppllJ:105
Lee MA, Storey BT (1985) Evidence for plasma membrane impermeability to small ions in acrosome-intact mouse spermatozoa bound to mouse zonae pellucidae, using an aminoacridine fluorescent probe: time course of the zona-induced acrosome reaction monitored by both chlortetracycline and pH probe fluorescence. BioI Reprod 33:235-246
Lee MA, Storey BT (1989) Endpoint of first stage of zona pellucida-induced acrosome reaction in mouse spermatozoa characterized by acrosomal H+ and ci+ permeability: population and single cell kinetics. Gamete Res 24:303-326
Signal Transduction in Mammalian Spermatozoa 181
Lec MA, Kopf GS, Storey BT (1987) Effects of phorbol esters and a diacylglycerol on the mouse sperm acrosome reaction induced by the zona pellucida. Bioi Reprod 36:617-627
Lee MA, Check JR, Kopf GS (1992) A guanine nucleotide-binding regulatory protein in human sperm mediates acrosomal exocytosis induced by the human zona pellucida. Mol Reprod Dev 31 :78-86
Ley ton L, Saling PM (1989a) 95 kD sperm proteins bind ZP3 and serve as tyrosine kinase substrates in response to zona binding. Cell 57: 123-130
Ley ton L, Saling P (l989b) Evidence that aggregation of mouse sperm receptors by ZP3 triggers the acrosome reaction. J Cell Bioi 108:2163-2168
Ley ton L, LeGuen P, Bunch D, Saling PM (1992) Regulation of mouse gamete interaction by a sperm tyrosine kinase. Proc Natl Acad Sci USA 89:11692-11695
Liang M, Garrison JC (1991) The epidcrmal growth factor receptor is coupled to a pertussis toxin-sensitive guanine nucleotide regulatory protein in rat hepatocytes. J Bioi Chem 266: 13342-13349
Lopez LC, Bayna EM, Litoff D, Shaper NL, Shaper JR, Shur BD (1985) Receptor function of mouse sperm surface galactosyltransferase during fertilization. J Cell BioI 101:1501-1510
Luttrell L, Kilgour E, Lamer J, Romero G (1990) A pertussis toxin-sensitive G protein mediates some aspects of insulin action in BC3R-1 murine myocytes. J BioI Chem 265: 16873-16879
Mcyerhof W, Muller-Brechlin R, Richter D ( 1991) Molecular cloning of a novel putativc G protein-coupled receptor expressed during rat spermiogenesis. FEBS Lett 284: 155-160
Miller DJ, Macek MB, Shur BD (1992) Complementarity between sperm surface ~-I ,4-galactosyltransferase and egg-coat ZP3 mediates sperm-egg binding. Nature 357:589-593
Millcr RL (1982) Synthetic peptides are not chemoattractants for bull sperm. Gamete Res 5:395-401
Mi Ilcr RL (1985) Sperm chemo-orientation in the metazoa. In: Metz CB, Monroy A (cds) Biology of fertilization, vol 2. Biology of the sperm. Academic, New York, pp 275-337
Moller CC, Wassarman PM (1989) Characterization of a proteinase that cleaves zona pellucida glycoprotcin ZP2 following activation of mouse eggs. Dev Bioi 132:103-112
Moller CC, Bleil JD, Kinloch RA. Wassarman PM (1990) Structural and functional relationships between mouse and hamster zona pellucida glycoproteins. Dev BioI 137:276-286
182 G.S. Kopf et al.
Mortillo S, Wassarman PM (1991) Differential binding of gold-labeled zona pellucida glycoproteins mZP2 and mZP3 to mouse sperm membrane compartments. Development 113:141-149
Noland TD, Garbers DL, Kopf GS (1988) An elevation in cyclic AMP concentration precedes the zona pellucida-induced acrosome reaction of mouse spermatozoa. Bioi Reprod 38[Suppl]:94
Overstreet JW, Cooper GW (1979) Effect of ovulation and sperm motility on the migration of rabbit spermatozoa to the site of fertilization. J Reprod Fertil 55:53-59
Parmentier M, Libert F, Schurmans S, Schiffman S, Lefort A, Eggerickx D, Ledent C, Mollereau C, Gerard C, Perret J, Grootegoed A, Vassart G (1992) Expression of members of the putative olfactory receptor gene family in mammalian germ cells. Nature 355:453-455
Ralt D, Goldenberg M, Fetterolf P, Thompson D, Dor J, Mashiach S, Garbers DL, Eisenbach M (1991) Sperm attraction to a follicular factor(s) correlates with human egg fertilizability. Proc Nat! Acad Sci USA 88:2840-2844
Ringuette MJ, Sobieski DA, Chamow SM, Dean J (1986) Oocyte-specific gene expression: molecular characterization of a cDNA encoding for ZP3, the sperm receptor of the mouse zona pellucida. Proc Natl Acad Sci USA 83:4341-4345
Ringuette MJ, Chamberlin ME, Baur AW, Sobieski DA, Dean J (1988) Molecular analysis of cDNA coding for ZP3, a sperm binding protein of the mouse zona pellucida. Dev BioI 127:287-295
Rosiere TK, Wassarman PM (1992) Identification of a region of mouse zona pellucida glycoprotein mZP3 that possesses sperm receptor activity. Dev BioI 154:309-317
Saling PM, Sowinski J, Storey BT (1979) An ultrastructural study of epididymal mouse spermatozoa binding to zonae pellucidae in vitro: sequential relationship to the acrosome reaction. J Exp ZooI209:229-238
Shimizu S, Tsuji M, Dean J (1983) In vitro biosynthesis of three sulfated glycoproteins of murine zonae pellucidae by oocytes grown in follicle culture. J BioI Chern 258:5858-5863
Shur BD, Hall NG (1982) A role for mouse sperm surface galactosyltransferase in sperm binding to the egg zona pellucida. J Cell BioI 95:574--579
Storey BT, Kopf GS (1991) Fertilization in the mouse II. Spermatozoa. In: Dunbar BS, O'Rand MG (eds) A comparative overview of mammalian fertilization. Plenum, New York, pp 167-216
Storey BT, Hourani CL, Kim JB (1992) A transient rise in intracellular ci+ is a precursor reaction to the zona pellucida-induced acrosome reaction in mouse sperm and is blocked by the induced acrosome reaction inhibitor, 3-quinuclidinyl benzilate. Mol Reprod Dev 32:41-50
Signal Transduction in Mammalian Spermatozoa 183
Tesarik J, Carreras A, Mendoza C (1992) Differential sensitivity of progesterone- and zona pellucida-induced acrosome reactions to pertussis toxin. Mol Reprod Dev 34: 183-189
Vazquez MH, Phillips DM, Wassarman PM (1989) Interaction of mouse spcrm with purificd sperm receptors covalently linked to silica beads. J Cell Sci 92:713-722
Villanueva-Diaz C, Vadillo-Ortega F, Kably-Ambe A, Diaz-Percz M, Krivitzky SK (1990) Evidence that human follicular fluid contains a chemoattractant for spermatozoa. Fertil Stcril 54: 1180-1182
Ward CR, Kopf GS (1993) Molecular events mediating sperm activation. Dcv Bioi 158: 1-26
Ward CR, Storey BT, Kopf GS (1992) Activation of a Gi protein in cell-free membrane preparations of mouse sperm by the zona pellucida and ZP3, components of the egg's extracellular matrix. J Bioi Chern 267: 14061-14067
Wassarman PM (1988) Zona pellucida glycoproteins. Annu Rev Biochem 57:415--442
Wassarman PM (1990) Profile of a mammalian sperm receptor. Development 108:1-17
Wilde MW, Ward CR, Kopf GS (1992) Activation of a G protein in mouse sperm by the zona pellucida, an egg-associated extracellular matrix. Mol Reprod Dev 31:297-306
Yang L, BatTy G, Rhee SG, Manning D, Hansen CA, Williamson JR (1991) Pertussis toxin-sensitive Gi protein involvement in epidermal growth factor-induced activation of phospholipase C-y in rat hepatocytes. J Bioi Chern 266:22451-22458
Zamboni L (1972) Fertilization in the mouse. In: Moghissi KS, Hafez ESE (eds) Biology of mammalian fertilization and implantation. Thomas, Springfield, pp 213-262
Zinaman M, Drobins EZ, Morales P, Brazil C, Keil M, Cross NL, Hanson FW, Overstreet JW (1989) The physiology of sperm recovered from human cervix: acrosomal status and response to inducers of the acrosome reaction. Bioi Reprod 41 :791-797
9 Diversity and Regulation of cAMP-Dependent Protein Kinases
K. Tasken, B.S. Skalhegg, K.A. Tasken, R. Solberg, F.O. Levy, T. Lea, T. Jahnsen, and V. Hansson
9.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 185 9.2 Structural Features of cAK . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 186 9.2.1 Distinct Genes Encode Seven Subunits of cAK . . . . . . . . . . . . . .. 187 9.2.2 Domain Structure and Isozyme Composition. . . . . . . . . . . . . . . .. 189 9.3 Regulation of cAK Subunits. . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 192 9.3.1 Differential Expression. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 193 9.3.2 Hormonal Regulation of cAK Subunits . . . . . . . . . . . . . . . . . . . .. 196 9.3.3 Transcriptional Regulation of the RII~ Gene. . . . . . . . . . . . . . . .. 198 9.4 Subcellular Localization and Isozyme-Specific Effects of cAK. .. 199 9.4.1 cAKI: Isozyme-Specific Effects and Receptor Colocalization
in Lymphocytes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 199 9.4.2 cAKIl: Subcellular Localization and Specific Anchoring Proteins 206 9.5 Summary and Perspectives. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 207 References .................................................. 208
9.1 Introduction
Reversible protein phosphorylation is a key regulatory mechanism in eukaryotic cells. Protein phosphorylation was first demonstrated to regulate the activity of glycogen phosphorylase in response to glucagon (Fischer and Krebs 1955; Sutherland and Wosilait 1955). A heat-stable factor mediating the effect of glucagon on the phosphorylation status of glycogen phosphorylase was next identified as 3' ,5' -cyclic adenosine
186 K. Tasken et al.
monophosphate (cAMP; Sutherland and RaIl 1958), and the concept of cAMP as an intracellular second messenger to a wide range of hormones, neurotransmitters, and other signaling substances was developed (Robinson et al. 1971). The target for cAMP was purified and identified as a cAMP-regulated protein kinase (Walsh et al. 1968), termed cAMPdependent protein kinase (cAK; EC 2.7.1.37).
At present, the cAMP signaling pathway is known to involve hormone receptors that upon binding of ligand, transduce their signal over the cell membrane by coupling to G-proteins that interact with adenylyl cyclase on the inner membrane surface either to activate or to inhibit the production of cAMP (Fig. 1). cAMP has been implicated in a number of cellular processes such as metabolism (Krebs and Beavo 1979), gene regulation (Roesler et al. 1988), cell growth and division (Boynton and Whitfield 1983), cell differentiation (Liu 1982; Schwartz and Rubin 1983), and sperm motility (Tash et al. 1984), as well as ion channel conductivity (Li et al. 1993). With the exception of certain ion channels directly regulated by cAMP (Nakamura and Gold 1987; DiFrancesco and Tortora 1991), all known effects of cAMP in eukaryotic cells are mediated by cAK.
9.2 Structural Features of cAK
In the absence of cAMP, the dormant cAK holoenzyme is a tetramer consisting of two catalytic subunits (C) bound to a regulatory subunit (R) dimer. cAMP binds cooperatively to two sites on each R protomer (for review, see Beebe and Corbin 1986; D0skeland et al. 1993). Upon binding of four molecules of cAMP, the enzyme dissociates into an R subunit dimer with four molecules of cAMP-bound and two free, active catalytic subunits that phosphorylate serine and threonine residues on specific substrate proteins.
Initially, two different isozymes of cAK, termed type I and II (cAKI and cAKII, respectively), were identified based on their pattern of elution from diethylaminoethy (DEAE)-cellulose columns (Reimann et al. 1971; Corbin et al. 1975). The cAKI and cAKII isozymes, eluting at salt concentrations between 25 and 50 mM and 150 and 200 mM NaCl, respectively, were shown to contain different R subunits termed RI and
Diversity and Regulation of cAMP-Dependent Protein Kinases 187
Honnon.
TRANSDUCBR
Gprolein
PR01l!IN KINASE A
I!fFECfOR Adenylyl cycl~
SECOND MESSENGER
Fig. l. Thc 3' -5" cyclic adenosine monophosphate (cAMP) signal transduction cascadc. P, phosphate groups; C, catalytic subunits; R, regulatory subunits
RII (Beebe and Corbin 1986). Molecular cloning techniques have, however, revealed a great heterogeneity in both Rand C subunits.
9.2.1 Distinct Genes Encode Seven Subunits of cAK
Cloning of cDNAs for regulatory subunits have identified two RI subunits termed Ria (Lee et al. 1983; Sandberg et al. 1987) and RI~ (Clegg et al. 1988; Solberg et aI. 1991) and two RII subunits termed RIIa (Scott et al. 1987; 0yen et al. 1989) and RII~ (Jahnsen et al. 1986; Levy et al. 1988) as separate gene products. The RIa and R[~ subunits reveal high homology (81 % identity at the amino acid level) as do the RIIa and RlI~ subunits (68% identity at the amino acid level). The distinct R subunits are conserved in higher eukaryotes and reveal high interspecies homology (87%-98% at the amino acid level).
Two distinct C subunits were initially identified by molecular cloning, and were designated Ca (Uhler et al. 1986a) and C~ (Uhler et al. 1986b; Showers and Maurer 1986). The cloning of the Ca and C~
188 K. Tasken et al.
subunits from human testis by low homology screening revealed an additional gene, encoding a distinct C subunit designated Cy, that has so far only been demonstrated in human testis (Beebe et al. 1990). The Cy subunit is homologous to both Ca. and C~ (83% and 79% identity at the amino acid level, respectively), but the homology between Ca. and C~ is higher (93% identity at the amino acid level), indicating that Cy may have specialized functions or different substrate specificity from Ca. and C~. Interestingly, homology between Cy and Ca./C~ is higher at the nucleotide level than at the amino acid level, and Cy is genetically more closely related to Ca. than to C~ (Beebe et al. 1990). Recent work has also revealed the existence of splice variants of Ca. (Ca.2) and C~ (C~2). The Ca.2 mRNA encodes a truncated 224 amino acid protein resulting from a stop codon in intron G which is not spliced out in Ca.2 (Thomis et al. 1992), whereas the C~2 mRNA encodes a protein with an additional amino terminal 47 amino acids (Wiemann et al. 1991). The functional implications of the alternatively spliced proteins are not known.
Thus, in man, molecular cloning has revealed a total of seven distinct genes encoding subunits of cAK (Sandberg et al. 1987; Solberg et al. 1991; 0yen et al. 1989; Levy et al. 1988; Beebe et al. 1990). The chromosomal assignment of these genes shows that the distinct genes localize to various chromosomes with no evidence of clustering (Table 1). At present, no hereditary diseases have been assigned to any of the cAK genes. It is possible that the closely related isomeric forms of RI, RII, and C subunits serve to rescue function in case of mutations, whereas multiple gene function knockouts may be lethal. Future gene knockout experiments will probably enlighten this topic.
Upstream regulatory sequences have been reported for the genes encoding RIa. (Nowak et al. 1987), RI~ (Rogers et al. 1992), RII~ (Kurten et al. 1992; Singh et al. 1991), Ca. (Chrivia et al. 1988), and C~ (Chrivia et al. 1988). All these genes have GC-rich and TAT A-less promoters, which are characteristics of highly regulated genes expressed at a low level. Furthermore, the cloning of a processed human pseudogene for RIa. with an alternate 5' -nontranslated area originating from an upstream exon of the RIa. gene indicates that RIa. may have alternate transcription initiation at two different promoters (Solberg et al. 1993a). This further enhances the possibilities of regulation.
Diversity and Regulation of cAMP-Dependent Protein Kinases 189
Table 1. Chromosomal localization of human cAK subunits
Subunit Locus namc Chromosome Reference
Ca PRKACA 19 (Tasken et al. 1993c) C~ PRKACB Ip31-32 (Simard et al. 1992;
Van Roy et al. 1993, R, Solberg, unpublishcd)
Cy PRKACG 9ql3 (Foss et al. 1992) RIa PRKARIA 17423-24 (Boshart et al. 1991;
Jones et al. 1991; Solberg et al. 1993a)
RIa pseudogenc PRKARIAP Ip21-3l (Solberg et al. 1993a) RI~ PRKARIB 7p22-pter (Solberg et al. 1992) RIIa PRKAR2A 3p (Task en et al. 1993c) RII~ PRKAR2B 7q22 (Wainwright et al. 19X7;
Solberg et al. 1992)
9.2.2 Domain Structure and Isozyme Composition
9.2.2.1 Structure of the C Subunits All the C subunits (Ca, C~, Cy) have catalytic core motifs that are common to all protein kinases (Hanks et aL 1988; Taylor et aL 1992) and involve a MgATP binding site as well as a peptide binding site (Fig. 2). The crystal structure of the murine Ca subunit has recently been reported and is the first protein kinase crystal structure available (Knighton et aL 1991 b). The catalytic subunit is a nearly globular protein with two lobes. The small, amino terminal lobe is involved in MgATP binding, whereas the larger carboxy terminal lobe is involved in peptide binding and catalysis. Both MgA TP and the peptide come together for catalysis in the cleft between the two lobes. The binding of the protein kinase inhibitor (PKI) involves additional sites apart from the peptide binding site (Knighton et aL 199Ia). Interestingly, the Cy subunit, which is inhibited by the R subunit, is not inhibited by PKI (Beebe et aL 1992). With the exception of C~2, all the C subunits are myristylated in the amino terminus. This probably increases the structural stability of the C subunits (Steinberg 1991; Yonemoto et al. 1993). The amino terminus of C~2, which has a highly hydrophobic region, may serve the same function (Wiemann et aL 1991).
190
C subunits:
RI subunits:
RII subunits:
Myr
Catalytic core
Mg2'ATP binding
I
Substrate binding
cAMP binding
A B
K. T asken et al.
COOH
Ac ~ I ~COOH Dime~iza7\
Hinge region Interaction with C subunit
Anchoring protein interaction cAMP binding
f A B
Ac ~ I ~ICOOH Dim~riza7\
Hinge region Interaction with C subunit
o 100 200
Amino acids
300 400
Fig. 2. Domain structure of the catalytic (C) and regulatory (R) subunits of 3'-5'cycIic adenosine monophosphate (cAMP)-dependent protein kinases. P, phosphate groUps; Myr, myristyl group; Ac; acetyl group A and 8, cAMP binding sites A and 8 ; *, pseudophosphorylation site
9.2.2.2 Structure of the R Subunits The RI and RIr subunits contain an amino terminal dimerization domain, a region responsible for interaction with the C subunit, and in the carboxy terminus two tandem cAMP binding sites, termed site A and B (Fig. 2; Corbin et al. 1978; D¢skeland 1978). Dimerization was initially discovered by the fact that proteolytic cleavage in the hinge region of the molecule would produce a monomeric R subunit with cAMP bind-
Diversity and Regulation of cAMP-Dependent Protein Kinases 191
ing activity (Potter et al. 197X). For the RI subunits, dimerization involves two disulfide bridges (Cys 16 and Cys37) giving the two RI protomers an antiparallel alignment (Bubis et al. 1987). Dimerization of the RII subunit does not involve cysteines, but the domain responsible for dimerization resides in the amino terminal part of the protein (amino acids 1-30; Scott et al. 1990). Additional residues in the amino terminus (amino acids 1-82) of the RII dimer interact with anchoring proteins (Scott et al. 1990). The hinge region of the molecule, which has a site sensitive to proteolysis, is involved in binding to the substrate binding site of the C subunit. The RII subunits serve as true substrates and are autophosphorylated by the C subunit. In contrast, the RI subunits are not phosphorylated and bind as pseudosubstrates. Of the two tandem cAMP binding sites, only site B is exposed in the inactive tetrameric cAK complex (reviewed in D¢skeland et al. 1993). Binding of cAMP to this site enhances binding of cAMP to the A site in a positively cooperative fashion, as a result of a conformational change in the molecule. The characteristics of the two cAMP binding sites have been described in detail elsewhere (reviewed in Beebe and Corbin 1986; Scott 1991; D¢skeland et al. 1993), as have the relative affinities and site selectivities of a wide array of chemically modified cAMP analogs (0greid et al. 1989). No crystal structure of the R subunit is presently available. However, the crystallization of a monomeric RI deletion mutant (,;'d-
91) was recently reported (Su et al. 1993), and crystals that diffracted to a resolution of 3.0 A. Furthermore, a model of the cAMP binding site has been proposed based on the structure of the catabolite gene activator protein in Escherichia coli (Weber et al. 1987).
9.2.2.3 Isozyme Composition and Characteristics It is generally assumed that the catalytic subunits associate freely with homodimers of all the R subunits. However, cAKI holoenzymes are more readily dissociated by cAMP in vitro than cAKII holoenzymes (Beebe and Corbin 1986; Dostmann et al. 1990). Furthermore, when RII is overexpressed in 3T3 cells, the C subunit will preferably be bound to RH, whereas RI will be present as free dimer (Otten and McKnight 1989). This indicates that cAKII holoenzyme forms preferentially compared to cAKI under physiological conditions either due to lower sensitivity to cAMP or due to kinetics of association/dissociation intluenced by salt and MgATP (reviewed in D0skeland et al. 1993). The cAKI
192 K. T asken et al.
(Rla2C2 and RI132C2) and cAKII (RIIa2C2 and RII132C2) holoenzymes have also been reported to have distinct biochemical properties. RI13 holoenzymes are two- to sevenfold more sensitive to cyclic nucleotides than RIa holoenzymes (Cadd et al. 1990; Solberg et al. 1993c). RIIa and RII13 holoenzymes elute from DEAE-cellulose columns at different positions in the cAKII area, and RIIa expressed at high levels will compete with RII13 in binding the C subunit, indicating either a higher affinity for the C subunit or a higher threshold for cAMP-induced dissociation (Otten et al. 1991).
The recent characterization of a cell line almost completely devoid of cAKII revealed the presence of an isozyme consisting of an Rla-RI13 heterodimer with associated phosphotransferase activity. This isozyme elutes in the position of cAKII by DEAE-cellulose chromatography (Tasken et al. 1993e). Formation of Rla-RI13 heterodimeric complexes was also demonstrated in vitro by coimmunoprecipitation using recombinant proteins (Fig. 3). Heterodimers of RI (RIa-RI13) increase the number of potential isozymes within a single cell and add to the possibilities for diversification of cAMP-mediated signals.
9.3 Regulation of cAK Subunits
Most cells, with some exceptions (Landmark et al. 1993; Tasken et al. 1993e), have both cAKI and cAKII isozymes. However, the proportion of cAKI to cAKII and the level of expression of the various subunits differ greatly between different cells and tissues (Corbin et al. 1975). Furthermore, expression of cAK subunits is altered as a consequence of proliferation (Ekanger et al. 1989), differentiation (Liu 1982; Lorimer and Sanwal1989), and hormonal regulation (Jahnsen et al. 1985b; 0yen et al. 1988; Landmark et al. 1991; Ratoosh et al. 1987). In addition, activation and dissociation of the holoenzyme complexes greatly alter the stability of the Rand C subunits. The free RI and C subunits are more rapidly degraded than the inactive holoenzyme (Houge et al. 1990; Steinberg and Agard 1981; Lorimer and Sanwa11989; Richardson et al. 1990; Tasken et al. 1993b), and this degradation probably serves to prevent overshoot of catalytic activity in response to cAMP-mediated signaling. In addition, the intracellular activity of cAK is modulated by the heat-stable PKI (Walsh et al. 1990), which exist in several different
Diversity and Regulation of cAMP-Dependent Protein Kinases 193
2 3 4
aRia aRllI aRllI aRia
Rill Rill Ria Ria
Ria Ria Rill Rill
a
5 6
aRia aRllI
- -Ria Ria
7
aRllI
-
Rill
8
aRia
-Rill
antibody
unlabeled subunit [32 P)-8-azido-cAMP labeled subunit
I bovine Rial human RI~
L-____________________________ --J
1 2 3 4 5 6 7 8
Fig. 3. In vitro formation of rcgulatory subunits Rla,-RI~ heterodimers of 3'-5'cyclic adenosine monophosphate (cAMP)-dependent protein kinases (Tasken et al. 1993e). Recombinant RIa, (lanes I and 2. lanes 5 and 6) or RI~ (lanes 3 and 4, lanes 7 and 8) were photoaffinity labelled by 8-azido-I,2PjcAMP and subsequently allowed to dimerize in the presence (lanes 1-4) or absence (lanes 5--8) of a fivefold molar excess of the opposite unlabeled subunit. Immunoprecipitation with antibodies directed against the labeled subunit (lanes I, 3, 5 and 7) revealcd strong bands, reprcsenting RIa, or RI~. When antibodies directed against the opposite subunit were employed (lanes 2, 4, 6, and 8), coimmunoprecipitation of labeled RIa, by the RI~ antibody was demonstrated in the presence of unlabeled RI~ (lane 2), whereas in the absence of RI~, no coimmunoprecipitation was observed (/ane 6). Vice versa, coimmunoprccipitation of labeled RI~ by the RIa, antibody was only observed in the presence of unlabelled RIa, (lane 4), whcreas the control experiment showed no coimmunoprccipiation (/al1e 8)
isoforms (PKIa, PKI~ I, PKI~2) that show tissue- and developmentalspecific expression (van Patten et al. 1992; Scarpetta and Uhler 1993).
9.3.1 Differential Expression
The a subunits (Ca, RIa, RIIa) of cAK are ubiquitously expressed in almost all cells and tissues (Uhler et al. 1986b; Lee et al. 1983; 0yen et al. 1987; Scott et al. 1987). The ~ subunits (C~, RI~, RII~) are predominantly expressed in brain and gonadal tissues (Uhler et al. 1986b; Clegg et al. 1988; lahnsen et al. 1985a; Levy et al. 1988), although a
194 K. Tasken et al.
low-level expression can be shown in a wide range of human tissues (Solberg et al. 1991; K. Tasken, unpublished observation). Expression of the Cy subunit has so far only been demonstrated in the human testis (Beebe et al. 1990).
Gonadal tissues have a high level of both a and ~ subunits of cAK, and the rat testis has proved to be a good model system for studies on differential regulation of the various cAK subunits. Age studies of whole rat testes revealed distinct developmental changes in the expression of cAK subunits 0yen et al. 1987, 1990; Landmark et al. 1993). At prepubertal stages, 10--15 days of age, the presence of RIa (2.8 and 3.2 kb), RIIa (6.0 kb), RII~ (3.2 kb), and Ca (2.4 kb) mRNAs was detected. These are the mRNA species seen in somatic cells. During early puberty, 15-25 days of age, genn cells increase exponentially and the first haploid cells are observed between 21 and 24 days of age. At later stages the large number of genn cells dominate the testis and dilute signals from somatic cells in whole testis studies. During this time period, the large RIa, RIIa, and RII~ mRNAs declined concomitantly with the appearance of low-molecular-weight mRNAs of RIa (1.7 kb), RIIa (2.2 kb), and RII~ (1.6 kb). These shorter messages result from genn cell-specific use of alternative polyadenylation site signals (0yen et al. 1990). Low-molecular-weight mRNAs for RIa and RIIa were observed between 20 and 30 days of age and after day 40, respectively. Together with the appearance of the short RIa mRNA, expression of RI~ was also detected and the levels of Ca mRNA (2.4 kb) increased. Figure 4 clearly shows the differential expression of cAK subunits in various genn cell fractions. The short message of RIa (1.7 kb) as well as RI~ (2.4 kb) are present in pachytene spennatocytes (PS) and round spennatids (RST) both at 32 and 44 days of age, whereas a lower level of expression that can be accounted for by contamination from the RST fraction is observed in elongating spennatids (ES). In contrast, mRNAs for the RII subunits are not detected at 32 days, whereas a high level expression of RIIa mRNA (2.2 kb) can be detected at 44 days when the ES fraction can be purified. RII~ (1.6 kb) is also detected at 44 days, but appears stronger in the RST fraction. The Ca message (2.4 kb) is observed at high levels both in the PS and RST fractions. Studies of mRNA expression in isolated seminiferous tubules at different stages and in situ hybridization revealed a similar pattern of expression (L6nnerberg et al. 1992). Similar regulation of RIa, RIIa, RII~, and Ca
Diversity and Regulation of cAMP-Dependent Protein Kinases 195
3.2 kb-2.9 kb-
1.7 kb_
6.0 kb-
3.1 kb-
2.2 kb-
Ria RI ~
- 2.6 kb
PS RST PS RST ES 32 44
Rlla RII (3
- 3.2 kb
-1.9 kb -1.6 kb
PS RST PS RST ES ~ 44
PS RST PS RST ES 32 44
-2.4 kb -Germ cell fr.: PS RST PS RST ES Age (days): ~ 44
Fig. 4. 3'-S'cyclic adenosine monophosphate-dependent protein kinases (cAK) subunits in testicular germ cell fractions (0yen et al. 1990). Northern blot showing mRNA levels for the different cAK subunits present in gcrm cells. RNA was extracted from germinal cell fractions obtained by Sta-Put unit-gravity sedimentation of testicular cells from rats 32 and 44 days of age. Twenty micrograms of total RNA was loaded in each lane, subjected to electrophoresis, blotted onto a nylon membrane, and subsequently hybridized with 32P_Iahelccl cDNA probes against regulatory subunits RIa, RI~, RlIa, RII~ , and catalytic subunit Ca, PS, pachytene spermatocytes; RST, round spermatids; ES, elongating spermatids
196 K. Tasken et al.
protein has been reported recently (Landmark et al. 1993). Taken together, the RIa, RI~, and Ca subunits in germ cells are induced at premeiotic and meiotic stages, whereas the RII subunits are induced only during spermatid elongation. The C~ mRNA was detected in peritubular cells and Leydig tumor cells but not in Sertoli cells or germ cells (0yen et al. 1990).
9.3.2 Hormonal Regulation of cAK Subunits
Levels of expression of the different cAK subunits are subject to regulation by hormones acting through G protein-coupled receptors (Jahnsen et al. 1985b; Landmark et al. 1993; 0yen et al. 1988) as well as by steroid hormones (Levy et al. 1989). Regulation of cAK by hormones acting through cAMP may serve as an autologous sensitization/-desensitization mechanism of the cAMP effector system. cAMP-mediated regulation of cAK subunits acts through gene transcription (Tasken et al. 1991, 1993b) and mRNA stability (Knutsen et al. 1991) as well as altered stability of the Rand C proteins after dissociation of the holoenzyme by cAMP (Houge et al. 1990; Tasken et al. 1993b). Protein kinase C (PKC) represents another major signaling pathway in cells, and crosstalk between these two signaling systems is seen beyond cAMP at the level of cAK (Tasken et al. 1990, 1992).
Rat Sertoli cells serve as a good model system for studies of hormone responsiveness in general and of cAK regulation in particular. Follicle stimulating hormone and cAMP induce messenger RNA for RIa, RIIa, RII~, and Ca with similar kinetics. However, the responses differ greatly in magnitude. Whereas cAMP-dependent stimulation of RIa, RIIa, and Ca mRNAs increases two- to fourfold, the increase in RII~ mRNA is approximately 50-fold (0yen et al. 1988; Tasken et al. 1991). The upregulation of RIa, RII~, and Ca mRNAs after treatment by cAMP is, at least partly, due to an increased transcriptional activity (Fig. 5; Tasken et al. 1991), and in the case of RII~ also involves increased stability of the mRNA (Knutsen et al. 1991). In Sertoli cells, similar regulatory changes are observed in RIa, RIIa and RII~ protein (Landmark et al. 1991).
Different mechanisms are involved in the regulation of the RII~ and RIa genes. Whereas transcriptional activity of the RIa gene (maximal at
Diversity and Regulation of cAMP-Dependent Protein Kinases 197
Time (min)
Treatment
o basal
30
60
120
120
8-CPTcAMP
8-CPTcAMP
8-CPTcAMP
8-CPTcAMP + actinomycin 0
c - fos RIot. Cot. RII 13 pBR322
•
Fig. S. Transcriptional activation of c-fos, regulatory subunit RIa, catalytic subunit Ca, and RII~ by 3 ' -5'cyclic adenosine monophosphate (cAMP; Tasken et al. 1991). Primary cultures of Sertoli cells were incubated with 8-( 4-chlorophenylthio)cAMP (8-CPTcAMP) for the time periods indicated. Subsequently, nuclear run-on analyses were performed on isolated nuclei. 32P-Iabeled RNA was isolated after DNAse I treatment, and equal amounts of radioactivity (5x I 06 cpm) were added to each filter strip for hybridization followed by autoradiography. Filterstrips contained 10 j.!g of linearized DNA for plasm ids pBR322 and v-fos/pBR322 or 3 j.!g of linear cDNA for RIa, Ca, and R"~
30 min, Fig. 5) is induced with similar kinetics as that of the c~tos gene, the induction of the RIIp gene is increasing throughout the observation period (120 min, Fig. 5). Furthermore, the RIo, gene is superinduced by combined treatment with cAMP and a protein synthesis inhibitor (cycloheximide). In contrast, inhibition of protein synthesis almost completely blocks the cAMP-mediated induction of the RIIP gene (Tasken et al. 1991). Regulation of the RIIa gene appears to be qualitatively similar to that of RIIP, but is quantitatively less pronounced.
The RIo, and RlIP genes are also subject to regulation by PKC (Tasken et al. 1992). Again, the mechanisms of regulation appear to be different. PKC-dependent activation of RIo, is unaffected by cycloheximide whereas induction of RIIP is dependent on ongoing protein syn-
198 K. Tasken et al.
thesis (Tasken et al. 1992). cAMP and 12-0-tetradecanoyl-phorbol-13-acetate (TPA) have additive effects on the regulation of the RIa message, whereas TPA inhibits the cAMP-mediated induction of the RII~ gene.
Thus, there is extensive evidence showing differential mechanisms of regulation of the R subunit genes. The RIa gene seems to be regulated by cAMP with similar characteristics as the cAMP response element (CRE) regulated c-fos gene. The 5' -flanking sequence of the Ria gene also contains a consensus CRE that is conserved between pig (Nowak et al. 1987) and man (Solberg et al. 1993b). However, the cloning of a processed pseudogene from human (Solberg et al. 1993a) indicates an alternate initiation at two different promoters that are differentially regulated (Tasken et al. 1993f). In contrast, the RII~ gene has a regulation by cAMP distinct from that of RIa and c-fos and belongs to a group of genes which respond to cAMP with slower kinetics and have cAMPresponsive regions distinct from the classical CRE, TPA-responsive element (TRE), and AP-2 elements (Lund et al. 1990; Richardson et al. 1990; Kagawa and Waterman 1990). It has therefore been of great interest to study the transcriptional regulation of the RII~ gene.
9.3.3 Transcriptional Regulation of the RII~ Gene
RII~ was first isolated and cloned from rat granulosa cells (Jahnsen et al. 1986) where a six- to tenfold induction of its mRNA by cAMP is seen (Ratoosh et al. 1987). Studies of the 5' -flanking region of the rat RII~ gene in ovarian granulosa cells revealed that the cAMP responsiveness resided within a distinct region (-395 to -293) upstream of the translation initiation codon (Kurten et al. 1992).
For transfection in Sertoli cells, 5' -deletions of the RII~ flanking region were inserted in front of a chloramphenicol acyl transferase (CAT) reporter gene (Fig. 6). Basal activity directed from the different constructs are depicted in Fig. 6A. As can be seen, the CAT activity was reduced to approximately 50% when the region -723 to -395 was included. The same region conferred a fourfold cAMP responsiveness to the CAT reporter gene (Fig. 6B). In contrast, transfections of the same constructs into rat testis peri tubular cells revealed that the cAMP responsiveness and the inhibition of basal activity that resided within
Diversity and Regulation of cAMP-Dependent Protein Kinases 199
the region -723 to -395 was specific to Sertoli cells. Mapping of the cAMP-responsive region by gel retardation and DNAse I footprinting experiments identified several protected regions that are candidates for novel cAMP responsive elements (Tasken et al. 1993g).
9.4 Subcellular Localization and Isozyme-Specific Effects ofcAK
The heterogeneity in Rand C subunits of cAK and the possibility that all the C subunits may associate freely with different R subunit homo- or heterodimers gives potential for a large number of isozymes. Several lines of evidence mentioned above are compatible with the notion that specific functions can be assigned to the various isozymes: the various isozymes have distinct biochemical properties, and the Rand C subunits are differentially regulated by hormones and reveal cell-specific expression. These observations are, however, only indicative to the idea of isozyme-specific effects.
In addition to this, distinct subcellular localization of cAK subunits and demonstration of compartmentalized effects of cAMP strongly support the concept of isozyme-specific effects. The above-mentioned, detailed studies of chemically modified cAMP analogs and their selectivity for either site A or site B of cAKI and cAKII (0greid et al. 19R9; Doskeland et al. 1991) represent an important contribution and have made it possible to study activation of either of these isozymes.
9.4.1 cAKI: Isozyme-Specific Effects and Receptor Colocalization in Lymphocytes
In general, the cAKI isozymes (Rlcx2C2, RI~2C2) are soluble and freely distributed in the cytoplasm (Meinkoth et al. 1990). Upon activation by cAMP, cAKI holoenzyme dissociates and releases C to be translocated to the nucleus whereas RI remains in the cytoplasm. cAMP also inhibits the growth of both normal and neoplastic human lymphocytes through cAK (Muraguchi et al. 1984; Blomhoff et al. 1987; Skttlhegg et al. 1992). In studies of cAKI, lymphocytes that contain a limited number or isozymes have been useful.
200
RIIB5'-flankingregion -123 CAT gene -4540 ----------"""TTT"' I I L
-4540 LI __________ -'
-1749 CI ===::J -1395 CI ::===::J
-1 005 c:==::J
-723 c=J
-395 D
-293 0
-3697 !rrJt.::*rrr-----------' -7235?
I -
K. Tasken et al.
o 20 40 60 80 100 120 A % Basal activity +/- SEM
RIIB 5'-flanking region -123 CAT gene -4540 ----------'L,r· .... j·---"---'I
-4540 LI __________ -'
-1749 .... 1 ___ -'
-1395 IC:=::=::::J
-1 005 c:==::J
-723 c=J
-395 D
-2930
o 2 3 4 5 6 B Relative cAMP response +/- SEM
Fig. 6. Basal (A) and 3'-5'cyclic adenosine monophosphate (cAMP)-responsive activity (B) of the rat RII~ promoter and 5'-flanking sequence (Tasken et al. 1993g). Various constructs containing 5' -deletions of the upstream sequence of the rat regulatory subunit RII~ gene (Kurten et al. 1992) inserted in front of a chloramphenicol acetyl transferase (CAT) reporter gene were transfected into rat Sertoli cells. Left panels: 5' -deletion fragments of the rat RlI~ gene inserted upstream of a CAT reporter gene. Numbers represent 5' -RII positons of
Diversity and Regulation of cAMP-Dependent Protein Kinases 201
9.4.1.1 Manipulation of Growth by Transfected Subunits of cAK A neoplastic B cell line (Reh) is practically devoid of cAKIl (Tasken et al. 1993e) and cell proliferation is inhibited by cAMP in these cells (Blomhoff et al. 1987). Stably transfected Reh cell lines overexpressing RIa or Ca were prepared (Tasken et al. 1993a). The expression of these subunits is directed by the metallothionine promoter which is responsive to divalent cations such as Zn2+. Figure 7 shows that in the presence of a small trigger dose of forskolin. there is a distinct zinc-dependent inhibition of r3H Ithymidine incorporation in cells transfected with Ca (panels A and B). In cells transfected with RIa or a dominant mutant of RIa, which does not bind cAMP, there is no effect of zinc on cell proliferation (panels C and D). Untransfected or control transfected cells do not show any zinc-dependent inhibition of cell proliferation (panels E and F). These results testify to the role of the C subunit in mediating cAMP-dependent inhibition of cell proliferation in lymphoid cells. Furthermore, since this cell line contains almost exclusively cAKI, there are strong indications that the inhibitory effect of cAMP on cell proliferation is mediated via this isozyme.
9.4.1.2 1sozyme-Specific Effect of cAK in T-Cell Proliferation In contrast to the neoplastic B lymphoid Reh cells, human peripheral blood T lymphocytes contain both cAKI (RIa2C~2) and cAKIl (RIla2C~2) in a proportion of 3: I (Sk1llhegg et al. 1992). In resting T cells the cAKI is 7Y,/o soluble whereas 95% of the cAKIl is particulate. Quiescent T lymphocytes can be activated to proliferate by cross-linking the T-cell receptor-CD3 complex (TCR/CD3). T-cell proliferation induced through the TCR is sensitive to inhibition by cAMP. Figure 8
the various RlI~ fragments relative to the first nucleotide of the RII~ initiation codon. All constructs ended at position -123. Fragments labeled with an as/erisk were inserted in the reverse orientation. A (Right panel:) Basal CAT activity of the various constructs. Levels of CAT activity (mean ± SEM; n = 5-1 I) were normalized to the secreted amounts of hGH directed from the cotransfeeted plasmid pXGH5. B (Righ/ panel) Relative stimulation of CAT activity (mean ± SEM, n = 9-21) in response to cAMP [3x I 0-5 M 8-( 4-ehlorophenylthio)cAMP. X-CPTcAMP. 30 h] for the various CAT constructs depicted in the left pWlel
202 K. Tasken et al.
110
~ 100 j....,..
90 t--i---i. ..
80 '\ ::E ' \ \
70 \
~ \ \ \ \
CI) 60 \ \
+l A Ca D2 ) B CaBI '!
50 C o t t 1 0 .-...... ~ 110
~ I-< ~ 0 100 --0.. 90 I-< 'k'l 0 U 80 C 70 ...... 41) 60 C
50 C RlaAI D Rla •• A4 .-"0 o 1 f 1 .-S ;>-..
110 ~ ~ E-< 100
~ , " I ,......, 90
~ M 80 '-'
~ 70
60
50 E pMEPD4 F REH
o 1 f 1
0 25 50 75 90 0 25 50 75 90
ZnS0 4 ( j..1M)
-- -F --- +51JM F
Fig. 7 A- F. Inhibition of cell proliferation by zinc-dependent induction of Ca in the presence of a small trigger dose of forskolin (Tasken et al. 1993a). Transfected cell clones and normal Reh cells were treated with increasing concentrations of ZnS0 4 in the presence (dashed line) and absence (so lid l ine) of a small trigger dose of forskolin (5 lIM). Cell cultures were incubated for 72 h, pulsed with [3Hlthymidine for the last 20 h, and thymidine uptake was determined by [3-scintillation counting. Data represent mean ± SEM (n = 18; Repeated Measures Anova test)
Diversity and Regulation of cAMP-Dependent Protein Kinases 203
'" b ~ 40 X
E 0. £ c 30 o ~ o 0.20 o o c Q)
C 10 i5 -E "£ 0 l-
I '2....
A B
.-e-....... ..-- a-AHA
.6.-.6. .\ \ . .6.Sj /.6.
8-AHA + 8-pip "-L :.&._--..-.. o -6 -5 -4 o -6 -5 -4 -3
cAMP analog concentration (log M)
Fig. 8 A, B. Synergistic activation of :I' -5' cyclic adenosine monophosphate (cAMP)-dependent protein kinase I (cAKI) using site-selective cAMP analogs in intact T lymphocytes (SkiUhegg et a!. 1992). Site-selective cAMP analogs that would complement each other in activating either cAKI or cAKIl were tested for their ability to synergize in the inhibition of T lymphocyte proliferation. Inhibition of anti-CD3 stimulated ['HJthymidine incorporation by 8-AHA in the absence (circles) or presence (triangles) of a subinhibitory concentration (7% of ICso, 90 /-lM) of 8-pip (A) or by 8-CPT in the absence (circles) or presence (triangles) of a subinhibitory concentration (7 0/0 of ICso. 30 /-lM) of N('
Bnz (B)
demonstrates that this is a cAKI mediated effect. The combination of 8-piperidino-cAMP (8-pip) and 8-aminohexylamino-cAMP (8-AHA) synergized in inhibiting incorporation of [3Hjthymidine in proliferating T cells when compared to the effect of 8-AHA alone. No such synergism was observed when inhibition by 8-(4-chlorophenylthio)cAMP (8-CPT) was examined in the absence and presence of a small priming dose of N6-benzoyl-cAMP (N6_Bnz) that by itself had no effect on T-cell proliferation. The combination 8-pip/8-AHA synergizes in the activation of cAKI since 8-pip reveals high affinity for the A site of RI whereas 8-AHA binds to the B site of both RI and RII with equal affinity. This is in contrast to activation of cAKII where both 8-pip and 8-AHA compete for binding to the B site. In contrast, the combination ofN6-Bnz and 8-CPT tends primarily to activate cAKII. This is because
204 K. Tasken et al.
8-CPT binds to the B site of RII with much higher affinity than to the cAKI B site and N6-Bnz binds to the A site of both RI and RH. Thus, inhibition ofT-cell proliferation by cAMP appears to be a cAKI-mediated effect. This was surprising since cAKI, in contrast to cAKII, is primarily soluble and may therefore not be expected to preferentially phosphorylate specific compartmentalized proteins. However, cAKIspecific effects have previously been shown for apoptosis of myeloid leukemia cells (LanoUe et a1. 1991) and for androgen production in Leydig cells (Moger 1991).
9.4.1.3 cAKI Redistributes to and Co localizes with the T-Cell Receptor Complex During Lymphocyte Activation and Capping The subcellular localization of cAKI and cAKII was examined during T-cell activation (Sk1Hhegg et a1. 1994b). Figure 9 shows that in quiescent T cells (panels A to C) the TCR/CD3 complex is widely distributed on the cell surface. In this situation, RIa is almost homogenously distributed within the cell (panel B) whereas RIIa is localized to one distinct spot (panel C) in close proximity to the nucleus. This is in agreement with our previous observation showing that cAKI is primarily soluble, whereas cAKII is particulate in resting T lymphocytes (SkiHhegg et al. 1992). When capping is induced (panels D to F), the TCR/CD3 complex is capped at one pole of the cells (panel D). Double immunoflourescence staining of the same cell by an RI antibody clearly shows that RIa redistributes to and colocalizes with the TCR/CD3 complex during activation and capping. In contrast, no effect of TCR/CD3 capping is observed in the subcellular distribution of RIIa. Immunoprecipitations of TCR/CD3 complex from capped and uncapped cells further demonstrated that 70%-75% of cAKI phosphotransferase activity, cAMP binding activity, and immunoreactive RIa and C was redistributed to and coimmunoprecipitaed with the TCR/CD3 complex only after capping and activation of T lymphocytes. Cross-linking of MHC class I, which does not involve T-cell activation, did not induce any redistribution and cocapping of cAKI (Skalhegg et a1. 1994b).
Colocalization of cAKI with the TCR/CD3 complex during capping and activation of T cells strongly supports our observation that inhibition of T-cell proliferation by cAMP is mediated via cAKI. The localization of cAKI in close proximity to the TCR/CD3 complex probably
Diversity and Regulation of cAMP-Dependent Protein Kinases 205
"0 Q) 0. 0. ro o c ::J
"0 Q) 0. 0. ro o
Anti-CD3 Anti-Ria Anti-Rlla
Fig. 9 A-F. Subcellular localization of 3'-5'cyclic adenosine monophosphate (cAMP)-dependent protein kinases (cAK) I and II in quiescent and activated T cells (Sk<1lhegg et al. I 994b ). Localization of the TCR/CD3 complex, Ria and Rlla in uncapped (A-C) or TCR-CD3 capped (D-F) T cells were examined after immunofluorescence labeling in a confocal immunofluorescence microscopc. A TCR/CD3 complex in uncapped T cells visualized by anti-CD3 antibody and Rhodamine-isothiocyanate (RITC)-conjungated antibody to IgGJ in the second layer. Note: the TCR/CD3 is scattered on the cell surface on uncapped T cells and a few small patches are seen. B The same cell as shown in A, after incubation (overnight) of permeabilized cells with anti-Ria and tluorescence-isothiocyanate (FITC)-conjugated antibody to IgG2A in the second layer. Note: the homogeneous distribution of Ria in uncapped T cells. C FlTC fluorescence of permeabilized cells incubated with anti-Rlla (arrowheads) and counterstained with 7-amino-actinomycin 0 to visualize the nucleus. Note: Rlla is localized to a distinct spot in close proximity to the nucleus. D RITC fluorescence of anti-CD3 capped T cells. Note: The TCR/CD3 complex is capped. E The same cell as shown in D is incubated with anti-RIa and FlTC-labeled antibody to IgG in the second layer. Note: distinct capping of RIa F FITC tluorescence in capped T cells after incubation with anti-RIIa (arrowheads) and counterstaining of the nucleus with 7-amino-actinomycin D. Note: the subcellular distribution of RIIa is not intluenced by TCR/CD3 activation and capping
206 K. Tasken et al.
serves to establish an inhibitory pathway that uncouples the TCR/CD3 complex from its intracellular signaling system. In addition, T-cell activation by stimulating the CD28 cell surface molecule in conjunction with the PKC activator TPA is not inhibited by cAMP, whereas activation by CD28 in conjunction with TCR/CD3 is sensitive to inhibition by cAMP (SkAlhegg et al. 1994a). This is compatible with the notion that the target for cAKI is a component of the TCR/CD3 complex or associated proteins that reside distal to the site of TCR stimulation but proximal to the site of PKC activation. Protein tyrosine kinases or proteins mediating the effect of TCR on phospholipase CYI are possible targets for cAMP effects mediated via cAKI (RIa2C2).
B lymphoid cells activated by cross-linking of immunoglobulin (Ig) receptors are also sensitive to cAMP-mediated inhibition of cell proliferation (Blomhoff et al. 1987). The observation that cAKI, but not cAKII, colocalizes with the Ig-receptor complex during capping and activation (F.O. Levy, unpublished results) suggests that cAKI cocapping with certain receptors is a general phenomenon in lymphoid cells.
9.4.2 cAKII: Subcellular Localization and Specific Anchoring Proteins
In many tissues, cAKII isozymes (RIIa2C2, RIIP2C2) are primarily associated with the particulate fraction. However, the ratio of soluble to particulate cAKII varies and is probably dependent on the level of expression of RII and the level of specific proteins binding RII (see below). Isozyme-specific effects of cAKII have been shown in cAMPdependent activation of lipolysis and glycerol release from adipocytes (Beebe et al. 1984).
Both RIIa and RIIP have been reported to localize to the Golgi centrosomal area of different cell types (Boshart et al. 1991; Keryer et al. 1993b). Centrosomallocalization is in agreement with the observations in T cells shown in Fig. 9 and suggests involvement of cAKII in cell cycle control and formation of the spindle apparatus. Colocalization and coimmunoprecipitation of RIIa of cAKII with p34cdc2 kinase has been reported (Zheng et al. 1991), whereas RIIP has recently been shown to serve as a substrate for cdc2 kinase in vitro (Keryer et al. 1993a).
Diversity and Regulation of cAMP-Dependent Protein Kinases 207
However, a specific function of cAKII that can be ascribed to this localization remains to be shown.
A number of different RII anchoring proteins have been identified and cloned (reviewed in Scott and Carr 1992). These proteins reveal specific, high-affinity interaction with dimers of either RIIa or RII~. Anchoring proteins showing cell-specific expression and distinct subcellular localization have been reported (Carr et al. I 992a,b ). Thus, such anchoring proteins may, due to their distinct localization and cell-specific expression, serve to target cAKII towards specific substrates at defined subcellular loci.
9.S Summary and Perspectives
A large number of hormones, neurotransmitters, and other signaling substances that bind to G protein-coupled cell-surface receptors converge their signals at one sole second messenger, cAMP. The question of how specificity can be maintained in a signal transduction system in which many extracellular signals that lead to a vast array of intracellular responses all are mediated through one second messenger system has been subject to thorough investigation and a great deal of speculation. An increasing number of cAK isozymes consisting of homo- or heterodimers of R subunits (RIa, RI~, RIIa, RII~) with associated catalytic subunits (Ca, C~, Cy) may contribute to the answer to this problem.
The various cAK isozymes display distinct biochemical properties and the heterogeneous subunits of cAK reveal cell-specific expression and differential regulation at the level of gene transcription, mRNA stability, and protein stability in response to a wide range of hormones and other signaling substances. The existence of a number of anchoring proteins specific to either Riia or RII~ that localize cAKII isozymes towards distinct substrates at defined subcellular loci strongly supports the idea that specific functions can be assigned to the various cAK isozymes. The demonstration that selective activation of cAKI is necessary and sufficient for cAMP-mediated inhibition of T-cell proliferation, and the observation that T-cell activation is associated with redistribution and colocalization of cAKI to the T-cell receptor is also compatible with the notion of isozyme-specific effects. It is further intriguing to speculate that colocalization with specific isozymes of
208 K. Tasken et al.
cAK may be the case for other receptors as well. The molecular mechanisms for cAKI redistribution, which probably involve specific cAKI anchoring proteins, remain to be elucidated.
References
Beebe SJ, Corbin JD (1986) Cyclic nucleotide-dependent protein kinases. Enzymes 17:43-111
Beebe SJ, Holloway R, Rannels SR, Corbin JD (1984) Two classes of cAMP analogs which are selective for the two different camp-binding sites of type II protein kinase demonstrate synergism when added together to intact adipocytes. J BioI Chern 259:3539-3547
Beebe SJ, 0yen 0, Sandberg M, Fr0ysa A, Hansson V, Jahnsen T (1990) Molecular cloning of a tissuespecific protein kinase (Cy) from human testis -representing a third isoform for the catalytic subunit of cAMP-dependent protein kinase. Mol EndocrinoI4:465-475
Beebe SJ, Salomonsky P, Jahnsen T, Li Y (1992) The Cy subunit is a unique isozyme of the cAMP-dependent protein kinase. J Bioi Chern 267:25505-25512
Blomhoff HK, Smeland EB, Beiske K, Blomhoff R, Ruud E, Bj0ro T, PfeiferOhlsson S, Watt R, Funderud S, Godal T, Ohlsson R (1987) Cyclic AMPmediated suppression of normal and neoplastic B cell proliferation is associated with regulation of myc and Ha-ras protooncogenes. J Cell Physiol 131:426-433
Boshart M, Weih F, Nichols M, Schlitz G (1991) The tissue specific extinguisher locus TSEI encodes a regulatory subunit of cAMP-dependent protein kinase. Cell 66:849-859
Boynton AL, Whitfield JF (1983) The role of cAMP in cell proliferation: a critical assessment of the evidence. Adv Cyclic Nucl Res 15: 193-294
Bubis J, Vedvick TS, Taylor SS (1987) Antiparal\el alignment of the two protomers of the regulatory subunit dimer of cAMP-dependent protein kinase I. J Bioi Chern 262:14961-149667
Cadd GG, Uhler MD, Mcknight GS (1990) Holoenzymes of cAMP-dependent protein kinase containing the neural form of type I regulatory subunit have an increased sensitivity to cyclic nucleotides. J Bioi Chern 265: 19502-19506
Carr DW, Hausken ZE, Fraser ID, Stofko-Hahn RE, Scott JD (l992a) Association of the type II cAMP-dependent protein kinase with a human thyroid RIl-anchoring protein. Cloning and characterization of the RII-binding domain. J Bioi Chern 267:13376-13382
Diversity and Regulation of cAMP-Dependent Protein Kinases 209
Carr DW, Stofko-Hahn RE, Fraser ID, Cone RD, Scott JD (1992b) Localization of the cAMP-dependent protein kinase to the postsynaptic densities by A-kinase anchoring proteins. Characterization of AKAP 79. J BioI Chern 267:16816-16~m
Chrivia JC, Uhler MD, McKnight GS (1988) Characterization of genomic clones coding for the Ca and C~ subunits of mouse cAMP-dcpendent protein kinase. J Bioi Chern 263:5739-5744
Clegg CH, Cadd GG, McKnight GS (1988) Gcnetic characterization of a brainspecific form of the type I regulatory subunit of cAMP-dependent protein kinasc. Proc Natl Acad Sci USA 85:3703-3707
Corbin JD, Keely SL, Park CR (1975) The distribution and dissociation of cyclic adenosine 3',5' -monophosphate-dependent protcin kinases in adipose, cardiac, and other tissues. J BioI Chern 250:218-225
Corbin JD, Sugden PH, West L, Flockhardt DA, Lincoln TM, McCarthy D (1978) Studies on the properties and mode of action of purified regulatory subunit of bovine heart adenosine 3' ,5' -monophosphate-dependent protein kinase. J BioI Chern 253:3997-4003
DiFrancesco D, Tortora P (1991) Direct activation of cardiac pacemaker channels by intracellular cyclic AMP. Nature 351: 145-147
D¢skeland SO (1978) Evidence that rabbit muscle protein kinase has two kinetically distinct binding sites for adenosine 3' ,5' -cyclic monophosphate. Biochem Biophys Res Commun 83:542-549
D¢skeland SO, B0e R, Bruland T, Vintermyr OK, Jastorff B, Lanotte M (1991) Criteria used to judge that a cellular response is mediated by cAMP. In: Reid E, Cook GMW. Luzio JP (eds) Cell signalling: experimental strategies. Royal Society or Chemistry, Cambridge, UK. pp 103-114
D0skeland SO. Maronde E. Gjertsen BT (1993) The genetic SUbtypes of cAMP-dependent protein kinase functionally different or redundant? Biochim Biophys Acta I 178:249-258
Dostmann WR, Taylor SS. Genieser HG, Jastorff B, D¢skeland SO, 0greid D (1990) Probing the cyclic nucleotide binding sites of cAMP-dependent protein kinases I and II with analogs of adenosine 3',5' -cyclic phosphorothioates. J BioI Chern 265: 10484-10491
Ekanger R, Vintermyr OK. Houge G, Sand TE, Scott JD. Krebs EG, Eikhom TS, Christoffersen T. 0greid D, D¢skeland SO (1989) The expression of cAMP-dependent protein kinase subunits is differentially regulated during liver regeneration. J Bioi Chern 264:4374-4382
Fischer EH, Krehs EG (1955) Conversion of phosphorylase h to phosphorylase a in muscle extracts. J BioI Chern 216:121-132
Foss KB, Simard J, Beruhe D. Beebe SJ, Sandberg M, Grzeschik KH, Gagne R. Hansson V. Jahnsen T (1992) Localization of the catalytic subunit Cyof
210 K. T asken et al.
the cAMP-dependent protein kinase gene (PRKACG) to human chromosome region 9q 13. Cytogenet Cell Genet 60:22-25
Hanks SK, Quinn AM, Hunter T (1988) The protein kinase family: conserved features and deduced phylogeny of the catalytic domains. Science 241 :42-52
Houge G, Vintermyr OK, D\ilskeland SO (1990) The expression of cAMP-dependent protein kinase subunits in primary rat hepatocyte cultures. Cyclic AMP downregulates its own effector system by decreasing the amount of catalytic subunit and increasing the mRNAs for the inhibitory (R) subunits of cAMP-dependent protein kinase. Mol Endocrinol 4:481-488
lahnsen T, Hedin L, Lohmann SM, Walter U, Richards IS (1985a) The neural type II regulatory subunit of cAMP-dependent protein kinase is present and regulated by hormones in the rat ovary. 1 Bioi Chern 261 :6637-6639
lahnsen T, Lohmann SM, Walter U, Hedin L, Richards IS (1985b) Purification and characterization of hormone-regulated isoforms of the regulatory subunit of type II cAMP-dependent protein kinase from rat ovaries. 1 Bioi Chern 260: 15980-15987
lahnsen T, Hedin L, Kidd VI, Beattie WG, Lohmann SM, Walter U, Durica 1, Schulz TZ, Schiltz E, Browner M, Lawrence CB, Goldman D, Ratoosh SL, Richards IS (1986) Molecular cloning, cDNA structure and regulation of the regulatory subunit of type II cAMP-dependent protein kinase from rat ovarian granulosa cells. J Bioi Chern 261: 12352-12361
10nes KW, Shapero MH, Chevette M, Fournier REK (1991) Subtractive hybridization cloning of a tissue-specific extinguisher: TSEI encodes a regulatory subunit of protein kinase A. Cell 66:861-872
Kagawa N, Waterman MR (1990) cAMP-dependent transcription of the human CYP21B (P450C21) gene requires a cis-regulatory element distinct from the consensus cAMP-regulatory element. 1 Bioi Chern 265: 11299-11305
Keryer G, Luo Z, Cavadore lC, Erlichman 1, Bornens M (l993a) Phosphorylation of the regulatory subunit of type II~ cAMP-dependent protein kinase by cyclin B/p34cdc2 kinase impairs its binding to microtubule-associated protein 2. Proc Nat! Acad Sci USA 90:5418-5422
Keryer G, Rios RM, Landmark BF, SkiUhegg BS, Lohmann SM, Bornens M (1993b) A high-affinity binding protein for the regulatory subunit of cAMP-dependent protein kinase II in the centrosome of human cells. Exp Cell Res 204:230-240
Knighton DR, Xuong NH, Taylor SS, Sowadski 1M (1991a) Crystallization studies of cAMP-dependent protein kinase. Cocrystals of the catalytic subunit with a 20 amino acid residue peptide inhibitor al)d MgATP diffract to 3.0 A resolution. 1 Mol BioI 220:217-220
Diversity and Regulation of cAMP-Dependent Protein Kinases 211
Knighton DR, Zheng lH, Ten Eyck LF, Ashford V A, Xuong NH, Taylor SS, Sowadski 1M (1991 b) Crystal structure of the catalytic subunit of cyclic adenosinc monophosphate-dependent protein kinase. Science 253:407-414
Knutsen HK, Tasken KA, Eskild W, lahnsen T, Hansson V (1991) Adenosine 3' ,5' -monophosphate-dependent stabilization of messenger ribonucleic acids (mRNAs) for protein kinase-A (PKA) subunits in rat Sertoli cells: rapid degradation of mRNAs for PKA subunits is dependent on ongoing RN A and protein synthcsis. Endocrinology 129:2496--2502
Krebs EG, Beavo lA (1979) Phosphorylation-dephosphorylation of enzymes. Annu Rev Biochem 4X:923-959
Kurten RC, Lcvy LO, Shey 1, Durica 1M, Richards lS (1992) Identification and characterization of the GC-rich and cyclic adenosine 3' ,5' -monophosphate (cAMP)-inducible promoter of the type II~ cAMP-dependent protein kinase regulatory subunit gene. Mol Endocrinol 6:536-550
Landmark BF, Fauske B, Eskild W, Sk[tlhegg B, Lohmann SM, Hansson V, 1ahnsen T, Beebe Sl (1991) Identification, characterization, and hormonal regulation of 3',5'cyclic adenosine monophosphate-dependent protein kinases in rat Sertoli cells. Endocrinology 129:2345-2354
Landmark BF, 0yen 0, Sk[tlhegg BS, Fauske E, 1ahnsen T, Hansson V (1993) Cellular localization and age-dependent changes of the regulatory subunits of cAMP-dependent protein kinase in rat testis. 1 Reprod Fertil 99:323-334
Lanotte M, Riviere lB, Hermouet S, Houge G, Vintermyr OK, Gjertsen BT, D0skeland SO ( 1991) Programmed cell death (apoptosis) is induced rapidly and with positive cooperativity by activation of cyclic adenosine monophosphate-kinase I in a myeloid leukemia cell line. ] Cell Physiol 146:73-80
Lee DC, Carmichael OF, Krebs EG, McKnight GS (1983) Isolation of a cDNA clone for the type I regulatory subunit of bovine cAMP-dependent protein kinase. Proc Natl Acad Sci USA XO:3608-3612
Levy FO, 0yen 0, Sandberg M, Taskcn K, Eskild W, Hansson V, lahnsen T (19XX) Molecular cloning, complementary deoxyribonucleic acid structure and predicted full-length amino acid sequence of the hormone-inducible regulatory subunit of 3' ,5' -cyclic adenosine monophosphate-dependent protein kinase from human testis. Mol Endocrinol 2: 1364-1373
Levy FO, Ree AH, Eikvar L, Govindan MV, 1ahnsen T, Hansson V (1989) Glucocorticoid receptors and glucocorticoid effects in rat sertoli cells. Endocrinology 124:430-436
Li M, West JW, Numann R, Murphy B1, Scheuer T, Catterall WA (1993) Convergent regulation of sodium channcls by protein kinase C and cAMP-dependent protein kinase. Science 261: 1439-1442
212 K. Tasken et al.
Liu AYC (1982) Differentiation-specific increase of cAMP-dependent protein kinase in 3T3-Ll cells. J Bioi Chem 257:298-306
Li.innerberg P, Parvinen M, Jahnsen T, Hansson V, Persson H (1992) Stageand cell-specific expression of cyclic adenosine 3',5'-monophosphate-dependent protein kinases in rat seminiferous epithelium. Bioi Reprod 46:1057-1068
Lorimer IAJ, Sanwal BD (1989) Regulation of cyclic AMP-dependent protein kinase levels during skeletal myogenesis. Biochem J 264:305-308
Lund J, Ahlgren R, Wu DH, Kagimoto M, Simpson ER, Waterman MR (1990) Transcriptional regulation of the bovine CYP 17 (P-450I7a) gene. Identification of two cAMP regulatory regions lacking the consensus cAMP-responsive element (CRE). J Bioi Chem 265:3304-3312
Meinkoth JL, Ji Y, Taylor SS, Feramisco JR (1990) Dynamics of the distribution of cyclic AMP-dependent protein kinase in living cells. Proc Natl Acad Sci USA 87:9595-9599
Moger WH (1991) Evidence for compartmentalization of adenosine 3',5'-monophosphate (cAMP)-dependent protein kinases in rat Leydig cells using site-selective cAMP analogs. Endocrinology 128:1414-1418
Muraguchi A, Miyazaki K, Kehrl JH, Fauci AS (1984) Inhibition of human B cell activation by diterpene forskolin: interference with B cell growth factor-induced Gj to S transition of the B cell cycle. J Immunol 133: 1283-1287
Nakamura T, Gold GH (1987) A cyclic nucleotide-gated conductance in olfactory receptor cilia. Nature 325:442-444
Nowak I, Seipel K, Schwarz M, Jans DA, Hemmings BA (1987) Isolation of a cDNA and characterization of the 5' flanking region of the gene encoding the type I regulatory subunit of the cAMP-dependent protein kinase. Eur J Biochem 167:27-33
0greid D, Ekanger R, Suva RH, Miller JP, D0skeland SO (1989) Comparison of the two classes of binding sites (A and B) of type I and type II cyclicAMP-dependent protein kinases by using cyclic nucleotide analogs. Eur J Biochem 181:19-31
Otten AD, McKnight GS (1989) Overexpression of the type II regulatory subunit of the cAMP-dependent protein kinase eliminates the type I holoenzyme in mouse cells. J Bioi Chem 264:20255-20260
Otten AD, Parenteau LA, D0skeland SO, McKnight GS (1991) Hormonal activation of gene transcription in ras-transformed NIH3T3 cells overexpressing RIIa and RII13 subunits of the cAMP-dependent protein kinase. J BioI Chem 266:23074-23082
0yen 0, Fr0ysa A, Sandberg M, Eskild W, Joseph D, Hansson V, Jahnsen T (1987) Cellular localization and age-dependent changes in mRNA for cy-
Diversity and Regulation of cAMP-Dependent Protein Kinases 213
clic adenosine 3' ,5' -monophosphate-dependent protein kinases in rat testis. Bioi Reprod 37:947-956
0yen 0, Sandberg M, Eskild W, Levy FO, Knutsen G, Beebe S, Hansson Y, Jahnsen T (1988) Differential regulation of messenger ribonucleic acids for specific subunits of cyclic adenosine 3',S'-monophosphate (cAMP)-dependent protein kinase by cAMP in rat Sertoli cells. Endocrinology 122:2658-2666
0yen 0, Myklebust F, Scott JD, Hansson Y, Jahnsen T (1989) Human testis eDNA for the regulatory subunit RIl" of cAMP-dependent protein kinase encodes an alternate amino-terminal region. FEBS Lett 246:57-64
0yen 0, Myklebust F, Scott JD. Cadd GG, McKnight GS, Hansson Y, Jahnsen T (1990) Subunits of cyclic adenosine 3',5' -monophosphate-dependent protein kinase show differential and distinct expression patterns during germ cell differentiation: alternative polyadenylation in germ cells gives rise to unique smallersized mRNA species. Bioi Reprod 43:46--54
Potter RL, Stafford PH, Taylor SS (1978) Regulatory subunit of cAMP-dependent protein kinase I from porcine skeletal muscle: purification and proteolysis. Arch Biochem Biophys 190: 174-180
Ratoosh SL, Lifka J, Hedin L, Jahnsen T, Richards JS (1987) Hormonal regulation of the synthesis and mRNA content of the regulatory subunit of cyclic AMP-dependent protein kinase type II in cultured rat ovarian granulosa cells. J Bioi Chern 262:7306-7313
Reimann EM, Walsh DA, Krebs EG (1971) Purification and properties of rabbit skeletal muscle adenosine 3',5' -monophosphate-dependent protein kinases. J Bioi Chern 246: 1986--1995
Richardson JM, Howard P, Massa JS, Maurer RA (1990) Posllranseriptional regulation of cAMP-dependent protein kinase activity by cAMP in GH3 pituitary tumor cells. Evidence for increased degradation of catalytic subunit in the presence of cAMP. J Bioi Chern 265: 13635-13640
Robinson GA, Butcher RW, Sutherland EW (1971) Cyclic AMP. Academic, New York
Roesler WJ, Vandenbark GR, Hanson RW (1988) Cyclic AMP and the induction of eukaryotic gene transcription. J BioI Chern 263:9063-9066
Rogers K Y, Boring LF, McKnight GS, Clegg CH (1992) Promoter for the regulatory type [~ subunit of the 3',S'cyclic adenosine monophosphate-dependent protein kinase directs trans gene expression in the central nervous system. Mol Endocrinol 6: 1756-1765
Sandberg M, Tasken K, 0yen 0, Hansson V, Jahnsen T (1987) Molecular cloning, cDNA structure and deduced amino acid sequence for a type I regulatory subunit of cAMP-dependent protein kinase from human testis. Biochem Biophys Res Commun 149:939-945
214 K. Tasken et al.
Scarpetta MA, Uhler MD (1993) Evidence for two additional isoforms of the endogenous protein kinase inhibitor of cAMP-dependent protein kinase in mouse. J BioI Chern 268: 10927-10931
Schwartz DA, Rubin CS (1983) Regulation of cAMP-dependent protein kinase subunit levels in Friend erythroleukemic cells. J BioI Chern 258:777-784
Scott JD (199 I) Cyclic nucleotide-dependent protein kinases. Pharmacol Ther 50:123-145
Scott JD, Carr DW (1992) Subcellular localization of the type II cAMP-dependent protein kinase. News Physiol Sci 7: 143-148
Scott JD, Glaccum MB, Zoller MJ, Uhler MD, Helfman DM, McKnight GS, Krebs EG (1987) The molecular cloning of a type II regulatory subunit of the cAMP-dependent protein kinase from rat skeletal muscle and mouse brain. Proc Nat! Acad Sci USA 84:5192-5196
Scott JD, Stofko RE, McDonald JR, Comer JD, Vitalis EA, Mangili JA ( 1990) Type II regulatory subunit dimerization determines the subcellular localization of the cAMP-dependent protein kinase. J BioI Chern 265:21561-21566
Showers MO, Maurer RA (1986) A cloned bovine cDNA encodes an alternate form of the catalytic subunit of cAMP-dependent protein kinase. J BioI Chern 261: 16288-16291
Simard J, Berube D, Sandberg M, Grzeschik KH, Gagne R, Hansson V, Jahnsen T (1992) Assignment of the gene encoding the catalytic subunit C~ of camp-dependent protein kinase to the p36 band on chromosome I. Human Genet 88:653-657
Singh IS, Luo ZJ, Eng A, Erlichman J (1991) Molecular cloning and characterization of the promoter region of the mouse regulatory subunit RII~ of type II cAMP-dependent protein kinase. Biochem Biophys Res Commun 178:221-226
SHllhegg BS, Landmark BF, D0skeland SO, Hansson V, Lea T, Jahnsen T (1992) Cyclic AMP-dependent protein kinase type I mediates the inhibitory effects of 3',5' -cyclic adenosine monophosphate on cell replication in human T lymphocytes. J BioI Chern 267:15707-15714
Skiilhegg BS, Rasmussen AM, Tasken K, Hansson V, Jahnsen T, Lea T (J 994a) T-cell activation through the TCR/CD3 complex sensitizes the CD28 pathway to the inhibitory effects of cAMP-dependent protein kinase. Scand J Immunol (in press)
Skiilhegg BS, Tasken K, Hansson V, Huitfeldt HS, Jahnsen T, Lea T (I 994b) Cyclic AMP-dependent protein kinase type I localizes to and interacts with the TCR/CD3 complex during T-cell activation. Science 263:84-87
Solberg R, Tasken K, Keiserud A, Jahnsen T (1991) Molecular cloning, cDNA structure and tissue-specific expression of the human regulatory subunit
Diversity and Regulation of cAMP-Dependent Protein Kinases 215
RI~ of cAMP-dependent protein kinases. Biochem Biophys Res Commun 176: 166--172
Solberg R, Sistonen P, Triiskelin AL, Berube D, Simard I, Krajci p, lahnsen T, de la Chapelle A (1992) Mapping of the regulatory subunits RI~ and RII~ of cAMP-dcpendent protein kinase gcncs on human chromosome 7. Genomics 14:636-983
Solberg R, Sandberg M, Spurkland A, lahnsen T (I 993a) Isolation and characterization of a human pseudogene for the regulatory subunit RIa of cAMPdependent protein kinases and its sublocalization on chromosome I. Genomics 15:591-597
Solberg R, Taskcn K, Sandberg M, Natarajan V, lahnsen T ( 1993b) Molecular cloning, exonintron organization, and 5' -upstream sequence of the human regulatory subunit Ria of cAMP-dependent protein kinases (submitted)
Solberg R, Tasken K, Wen W, Coghlan VM, Meinkoth lL, Scott ID, lahnsen T, Taylor SS (1993c) Heterologous expression, characterization and subcellular localization of the human regulatory subunit RI~ of cAMP-dependent protein kinases (submitted)
Steinberg RA (1991) A kinase-negative mutant of S49 mouse lymphoma cells is defective in posltranslational maturation of catalytic subunit of cyclic AMP-dependent protein kinase. Mol Cell Bioi 11:705-712
Steinberg RA, Agard DA (1981) Turnover of regulatory subunit of cyclic AMP-dependent protein kinase in S49 mouse lymphoma cells. Regulation hy catalytic subunit and analogs of cyclic AMP, 1 Bioi Chern 256: IOn 1-IOn4
Su y, Taylor SS, Dostrnann WRG, Xuong NH, Varughese KI (1993) Crystallization of a deletion mutant of the R-subunit of cAMP dependent protein kinase. 1 Mol Bioi no: 1091-1093
Sutherland EW, Rall TW (1958) Fractionation and characterization of a cyclic adenosine ribonucleotide formed by tissue particles. I Bioi Chern 232:1077-1091
Sutherland EW. Wosilait WD (1955) Inactivation and activation of liver phosphorylase. Nature 175:169-170
Tash IS, Kakar SS, Means AR (1984) Flagellar motility requires the cAMPdependent phosphorylation of a heat-stable NP-40 soluble 56 kDa protein, axokinin. Cell 38:551-559
Tasken K, Kvale D, Hansson V, lahnsen T (1990) Protein kinase C activation selectively increases mRNA levels for one of the regulatory suhunits (Ria) of cAMP-dependent protein kinases in HT29 cells, Biochem Biophys Res Comrnun 172:409--414
Tasken KA, Knutsen HK, Attramadal H, Tasken K, lahnsen T, Hansson V, Eskild W (1991) Different mechanisms are involved in cAMP-mediated in-
216 K. Tasken et al.
duction of mRNAs for subunits of cAMP-dependent protein kinases. Mol EndocrinoI5:21-28
Tasken KA, Knutsen HK, Eikvar L, Tasken K, Eskild W, Jahnsen T, Hansson V (1992) Protein kinase C activation by 12-0-tetradecanoylphorbol 13-acetate modulates messenger ribonucleic acid levels for two of the regulatory subunits of 3',5' -cyclic adenosine monophosphate-dependent protein kinases (RII~ and RIa) via multiple and distinct mechanisms. Endocrinology 130:1271-1280
Tasken K, Andersson KB, Erikstein BK, Hansson V, Jahnsen T, BlomhoffHK (l993a) Regulation of growth in a neoplastic B cell line (Reh) by transfected subunits of cAMP-dependent protein kinase (submitted)
Tasken K, Andersson KB, Skalhegg BS, Tasken KA, Hansson V, 1ahnsen T, Blomhoff HK (l993b) Reciprokal regulation of mRNA and protein for subunits of cAMP-dependent protein kinase (RIa and Ca) by cAMP in a neoplastic B cell line (Reh). J Bioi Chern 268:23483-23489
Tasken K, Jahnsen T, Solberg R (I 993c) The human genes for the cAMP-dependent protein kinase subunits, RIIa (locus PRKAR2A) and Ca (locus PRKACA), localizes to chromosome 3 and 19, respectively (submitted)
Tasken K, SkiHhegg BS, Solberg R, Andersson KB, Taylor SS, Lea T, Blomhoff HK, Jahnsen T, Hansson V (l993e) Novel isozymes of cAMP-dependent protein kinase exist in human cells due to formation of RIa-RI~ heterodimeric complexes. J Bioi Chern 268:21276-21283
Tasken K, Solberg R, Hansson V, Jahnsen T (l993f) Alternately spliced mRNA species of the regulatory subunit, RI, of cAMP-dependent protein kinases are differentially regulated by cAMP (submitted)
Tasken KA, Tasken K, Eskild W, Knutsen HK, Richards JS, Kurten RC, Jahnsen T, Hansson V, Guerin SL (1993g) Identification of a cAMP-responsive region within the 5' flanking sequence of the rat gene for the hormone sensitive regulatory subunit, RII~, of cAMP-dependent protein kinases in rat Sertoli cells (submitted)
Taylor SS, Knighton DR, Zheng J, Ten Eyck LF, Sowadski 1M (1992) Structural framework for the protein kinase family. Annu Rev Cell Bioi 8:429-462
Thomis DC, Floyd-Smith G, Samuel CE (1992) Mechanism of interferon action. cDNA structure and regulation of a novel splice-site variant of the catalytic subunit of human protein kinase A from interferon-treated human cells. J Bioi Chern 267: 10723-10728
Uhler MD, Carmichael DF, Lee DC, Chrivia JC, Krebs EG, McKnight GS (1986a) Isolation of cDNA clones coding for the catalytic subunit of mouse cAMP-dependent protein kinase. Proc Natl Acad Sci USA 83:1300-1304
Diversity and Regulation of cAMP-Dependent Protein Kinases 217
Uhler MD, Chrivia IC, McKnight GS (1986b) Evidence for a second isoform of thc catalytic subunit of cAMP-dependent protein kinase. J Bioi Chern 261: 15360-15363
van Patten SM, Howard P, Walsh DA, Maurer RA (1992) The u- and ~-isoforms of the inhibitor protein of the 3',5' -cyclic adenosine monophosphate-dependent protein kinase: characteristics and tissue- and devclopmentalspccific expression. Mol Endocrino16:2114-2122
Van Roy N, Laureys G, Versteeg R, Opdenakker G, Spelem,m F (1993) Highresolution fluorescence mapping of 46 DNA markers to the short arm of human chromosome 1. Genomics 18:71-78
Wainwright B, Lench N, Davies K, Scambler P, Kruyer H, Williamson R, Jahnsen T, Farrall M (l9X7) A human regulatory subunit of type II cAMPdependent protein kinase localized by its linkage relationship to several cloned chromosome 7q markers. Cytogenet Cell Genet 45:925-932
Walsh DA, Perkins JP, Krebs EG (1968) An adenosine 3',5' -monophosphatedependent protein kinase from rabbit skeletal muscle. J Bioi Chern 243:2867-2X73
Walsh DA, Angelos KL, van Patten SM, Glass DB, Garetto LP (1990) The inhibitor protein of the cAMP-dependent protein kinase. In: Kemp BE (ed) Peptides and protein phosphorylation. CRC Press, Boca Raton, pp 43-X4
Weber [T, Steitz T, Bubis J, Taylor SS ([ 987) Predicted structures of the cAMP binding domains of type I and type II regulatory subunits of cAMPdependent protein kinase. Biochemistry 26:343-35 [
Wiemann S, Kinzel V, Pyerin W (1991) Isoform C~2, an unusual form of the bovine catalytic subunit of cAMP-dependent protein kinase. J Bioi Chern 266:5140-5 [46
Yonemoto W, McGlone ML, Taylor SS (1993) N-myristylation of the catalytic subunit of cAMP-dependent protein kinase conveys structural stability. J Bioi Chern 268:2348--2352
Zheng JH, Knighton DR, Parello J, Taylor SS, Sowadski JM (1991) Crystallization of catalytic subunit of adenosine cyclic monophosphate-dependent protein kinase. Methods Enzymol 200:50X-521
10 The Nuclear Response to cAMP During Spermatogenesis: The Key Role of Transcription Factor CREM
P. Sassone-Corsi
I D.I Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 22D 10.2 The Adcnylyl Cyclase Pathway. . . . . . . . . . . . . . . . . . . . . . . . . . .. 22D I D.3 Structural Features of cAMP-Inducible Genes. . . . . . . . . . . . . . .. 221 I D.3.1 The cAMP-Responsive Element. . . . . . . . . . . . . . . . . . . . . . . . . .. 221 10.3.2 Testis Genes Responsive to cAMP. . . . . . . . . . . . . . . . . . . . . . . .. 223 1004 A Family of Activators. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 225 1004.1 DNA Binding. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 225 1004.2 Activation by Phosphorylation. . . . . . . . . . . . . . . . . . . . . . . . . . .. 226 1004.3 The Role of the Glutamine-Rich Domains. . . . . . . . . . . . . . . . . .. 229 IO.S The CREM Gene. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 230 10.5.1 A Remarkable Genomic Structure and Cell-Specific Expression . 23D IO.S.2 Different Functions of the MUltiple CREM Products. . . . . . . . . .. 231 I D.5.3 10.504 I D.6 I D.7 10.7.1
10.7.2 I D.7.3 I D.8
Antagonists and Activators from the Same Gene ............. . CREM: A Gene with Specialized Neuroendocrine Functions ... . Physiological Roles of CRE-Binding Proteins ............... . CREM and Spermatogenesis ............................. . Hormonal Control in Testis: Induction of the cAMP Signal Transduction Pathway ......... . Nuclear Effectors of the PKA Pathway in Testis ............. . The CREM Gene: Regulation and Function in Germ Cells ..... . Inducible cAMP Early Repressor: An Alternative CREM Product Giving a Ncw Dimension
233 234 237 23X
23X 239 24D
to the Neuroendocrine cAMP Pathway. . . . . . . . . . . . . . . . . . . . .. 243 10.9 Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 244 References .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 245
220 P. Sassone-Corsi
10.1 Introduction
The regulation of gene expression by specific signal transduction pathways is closely connected to the cell phenotype, and the response elicited by a given transduction pathway will vary according to the cell type. The finding that most of the known nuclear oncogenes encode proteins involved in the regulation of gene expression inspired the concept that the aberrant expression of some key genes could cause cellular transformation or altered proliferation (Lewin 1991). The study, and ultimately the understanding, of these processes will hopefully help us to unravel the profound changes that cause cancer and, by the same token, the physiology of normal growth.
An important step towards the comprehension of how the function of transcription factors can be modulated has been the discovery that many of these factors constitute final targets of specific signal transduction pathways, activated intracellularly by various signals at the cell surface. The two major signal transduction systems are those including cAMP and diacylglycerol (DAG) as secondary messengers (Nishizuka 1986; Berridge 1987). Each pathway is also characterized by specific protein kinases (protein kinase A and protein kinase C, respectively) and its ultimate target DNA control element, the cAMP-responsive element (CRE) and the 12-0-tetradecanoyl phorbo1 l3-acetate (TPA)-responsive element (TRE). Although initially characterized as distinct systems, accumulating evidence points towards extensive cross-talk between these two pathways in the cytoplasm (Cambier et al. 1987; Yoshimasa et al. 1987) and in the nucleus (Sassone-Corsi et al. 1990; Benbrook and Jones 1990; Auwerx and Sassone-Corsi 1991; Masquilier and SassoneCorsi 1992).
This review will focus primarily on the targets of the cAMP-mediated transduction response and their function during the developmental process of the male germ cells.
10.2 The Adenylyl Cyclase Pathway
Intracellular levels of cAMP are regulated primarily by adenylyl cyclase. This enzyme is in tum modulated by various extracellular stimuli mediated by receptors and their interaction with G proteins (Gilman
The Nuclear Response to cAMP During Spermatogenesis 221
1987; McKnight et al. 1988; Borrelli et al. 1992). The binding of a specific ligand to a receptor results in the activation or inhibition of the cAMP-dependent pathway, ultimately affecting the transcriptional regulation of various genes through distinct promoter-responsive sites (Montmayeur and Borrelli 1991). Increased cAMP levels directly affect the function of the tetrameric protein kinase A (PKA) complex (Krebs and Beavo 1979). Binding of cAMP to two PKA regulatory subunits releases the catalytic subunits, enabling them to phosphorylate target proteins. An important fraction of the catalytic subunit molecules migrates into the nucleus (see Fig. I). A number of isoforms for both the regulatory and catalytic subunits have been identified, suggesting a further level of complexity in this response (McKnight et al. 1988). Interestingly, some specific isoforms of both the adenylyl cyclase and cAMP-dependent kinase have been found in testis (Pariset et al. 1989; Oyen et al. J 988; 1990; Lonnerberg et al. 1992). In the nucleus, the phosphorylation state of transcription factors and related proteins appears to directly modulate their function and thus the expression of cAMPinducible genes (Fig. 1).
The analysis of promoter sequences of several genes allowed the identification of promoter elements which could mediate the transcriptional response to increased levels of intracellular cAMP (Roesler et al. 1988; Borrelli et al. 1992). A number of sequences have been identified, of which the best characterized is the CRE. The CRE is recognized by a multitude of nuclear factors; their characteristics are described in the next sections.
10.3 Structural Features of cAMP-Inducible Genes
10.3.1 The cAMP-Responsive Element
The CRE promoter element mediates the response to increased levels of intracellular cAMP (Comb et al. 1986; Andrisani et al. 1987; Oelegeane et al. 1987; Sassone-Corsi 1988). A consensus CRE site constitutes an 8-bp palindromic sequence (TGACGTCA). Several genes which are regulated by a variety of endocrinological stimuli contain similar sequences in their promoter regions, although at different positions. A comparison of the CRE sequences identified to date shows that thc 5'
222
i ••
CYTOPLASM
G~=
NUCLEUS
P. Sassone-Corsi
o 0 Ligands
e 0
::-~ cAMP ATP
= "~===~ 8 J
Activators p-.@d ~~~t
/ A ............... r:= '1Y'~
P cAMP-inducible genes
Fig. 1. The 3' -5' cyclic adenosine monophosphate (cAMP) signal transduction pathway operating from the cell membrane, through the cytoplasm and into the nucleus. Ligands interacting with transmembrane receptors (R) stimulate the enzyme adenylyl cyclase (AC) via interactions with G proteins (G). The subsequent rise in intracellular cAMP concentration results in the dissociation of the regulatory and catalytic subunits of protein kinase A (PKA) and the translocation of active catalytic subunits into the nucleus. PKA phosphorylates and thereby stimulates transcriptional activators binding to cAMP-responsive elements [CRE; activators, e.g., CRE-binding protein, CREB, cAMP-responsive element modulator, CREM't and ATF-I (Gonzalez et al. 1989; Rehfuss et al. 1991; de Groot et al. 1993a)] which induce transcription from the promoters of cAMP-responsive genes. These factors activate transcription as dimers (de Groot and Sassone-Corsi 1993). P, phosphate groups
half of the palindrome, TGACG, is the best conserved, whereas the 3' TCA motif is less constant (Borrelli et al. 1992). The binding site specificity appears to require 18-20 bp, since the five or so bases nanking the core consensus have been shown to dictate, in some cases, the permissivity of transcriptional activation (Deutsch et al. 1988). In many
The Nuclear Response to cAMP During Spermatogenesis 223
genes the CRE sequence is located in the first 200 bp upstream from the cap site. In most cases there is only one CRE element per promoter, although there are notable exceptions. The promoter of the a-chorionic gonadotropin gene, for instance, contains two identical, canonical CREs in tandem between positions -117/-142 (Delegeane et al. 1987). The promoter of the pituitary-specific transcription factor GHF-I /Pit-I, in contrast, contains two different CREs between positions -200/-150 which are separated by a 40-bp spacer (McConnick et al. 1990). The proto-oncogene c~t()S contains a powerful CRE at position -60 (SassoneCorsi et al. 1988), but other CRE-like sequences are also present within the gene regulatory region (Berkowitz et al. 1989); however, their precise function has yet to be detennined.
10.3.2 Testis Genes Responsive to cAMP
The cAMP-responsive element modulator (CREM; see Sect. 5) activator isoform is very abundant in spermatids (Foulkes et al. 1992; Delmas et al. 1993). Thus, it is obvious to ask whether genes that are strongly activated at that stage of the spermatogenesis differentiation process may contain CRE sites in their regulatory regions. In this respect it is noteworthy that CREM proteins are able to recognize a number of different CRE motifs (Laoide et al. 1993). Indeed, we have analyzed the sequences of the promoters of genes induced in haploid round spermatids and found that several contain CREs (listed in Table I). Some promoters contain perfect consensus CRE motifs, as is the case of the transition protein I (MTP I), or quite divergent CREs, as for the promoter of thymosin. DNA-binding analyses revealed that all the oligonucleotides corresponding to CREs present in testis-specific genes are elliciently recognized by purified CREM protein (Delmas et al. 1993). Affinity for binding varies among the different CREs; for example promoters of MTPI and a testis-specific protein highly homologous to the acidic epididymal glycoprotein (tpx-I a) require a large excess of cold CRE oligonucleotide to complete binding, indicating a particularly strong affinity for CREM protein (Delmas et al. 1993). These results suggest that CREM activators may recognize a large variety of cellular targets in spenmltids. For further discussion of the role of CREM in spermatogenesis see later (Sect. 8).
Tab
le 1
. 3
'-5
' cy
clic
ade
nosi
ne m
onop
hosp
hate
(cA
MP
)-re
spon
sive
ele
men
t m
odul
ator
(C
RE
Mr)
bin
ds t
o va
riou
s C
RE
-lik
e ~
sequ
ence
s pr
esen
t in
prom
oter
s ac
tiva
ted
in r
ound
spe
rmat
ids
..,.
CR
E s
eque
nce
TT
GG
CT
GA
CG
TC
AG
AG
AG
A
AT
TG
GG
TG
AG
GT
CA
CT
TA
A
AG
CT
TC
CT
CT
TT
GA
CT
TC
AT
AA
TT
CC
T
TG
GG
CC
GA
CA
GG
TC
AC
AG
TG
GG
G
TA
TG
TA
GT
GA
CG
TC
AC
AA
GA
GA
G
AG
GA
CG
TG
CT
GG
TC
AC
CC
CC
AA
A
TT
GA
TC
TA
GA
CG
TC
AA
AA
TT
CC
T
CT
TC
CT
AA
TA
CG
TC
AC
AG
CA
TC
T
Cor
repo
ndin
g ge
ne
Som
atos
tati
n R
T7
Pro
tam
ine
1 P
rota
min
e 2
Tra
nsit
ion
prot
ein
1 T
hym
osin
T
px
-Ia
Tpx
-Ib
Ref
eren
ces
Com
b et
a1.
(198
6)
van
der
Hoo
m a
nd T
amas
ky (
1992
) Jo
hnso
n et
a.
(198
8)
John
son
et a
. (1
988)
H
eida
ran
et a
l. (1
989)
L
in a
nd M
orri
son-
Bog
orad
(19
91)
Miz
uki
et a
l. (1
992)
M
izuk
i et a
l. (1
992)
Seq
uenc
es o
f one
str
and
of t
he o
ligo
nucl
eoti
des
wer
e sy
nthe
size
d an
d us
ed in
DN
A b
indi
ng a
ssay
s (D
elm
as e
t a1.
199
3),
whi
ch c
orre
spon
d to
the
nat
ural
ly o
ccur
ring
CR
E-l
ike
site
s. T
he e
ight
bas
e C
RE
seq
uenc
e is
sho
wn
in b
old.
The
nam
e an
d th
e re
fere
nce
of t
he c
orre
spon
ding
gen
e is
giv
en.
For
the
Tpx
-I p
rom
oter
, tw
o C
RE
mot
ifs
wer
e fo
und,
Tpx
-Ia
and
Tpx
-Ib.
"U
(J) el C
J) o :J
CD
(,
o ~-
The Nuclear Response to cAMP During Spermatogenesis 225
10.4 A Family of Activators
An important step toward the understanding of cAMP-regulated gene transcription has been made with the cloning of cDNAs encoding CREbinding proteins, thereby allowing studies on the precise structure-function relationship of these factors. Several genes have been found to encode CRE-binding proteins and they constitute a family. The cDNAs for one of these, CRE-binding protein (CREB), have been cloned from human placenta and rat brain libraries (Hoeffler et al. 1988; Gonzales et al. 1989). Other members of this family have been isolated from various sources (Maekawa et al. 1989; Hai et al. 1989; Ziff 1990). These factors are ubiquitously expressed, suggesting a housekeeping role for these nuclear regulators (de Groot and Sassone-Corsi 1993).
10.4.1 DNA Binding
The CREB protein, as well as all the other members of the family, belongs to the leucine-zipper group of transcriptional regulators (bZip; Landschulz et al. 191\8); it contains a heptad repeat of four leucines in the carboxy terminus which constitutes an a-helical coiled structure (for a review see Busch and Sassone-Corsi 1990). It has been demonstrated that, as in the case of Fos, lun and C(EBP, the leucine zipper is responsible for the dimerization of the protein and that dimerization is a prerequisite for DNA binding. The model by Vinson et al. (191\9) suggests the presence of a hipartite DNA-binding domain, as dimerization ensures the correct orientation of the adjacent basic regions in order to allow their optimal contact with the recognition sequence. It has been determined that the basic region, 50% rich in lysine and arginine residues, is in fact divided into two suhdomains containing clusters of basic residues separated by a "spacer" of alanines which are conserved among all leucine zipper transcription factors. In this model, these two regions recognize the two halves of the palindromic recognition sequence. The positively charged amino acids in the basic region lie on one face of the two helices of a helix-bend-helix structure. The two positively charged a-helices lie in the major groove of the DNA helix, positioned so that the positive charges are in contact with the negative charges of the
226 P. Sassone-Corsi
phosphate backbone (Vinson et al. 1989). The structure of the DNA binding domain of CRE-binding proteins appears to be conserved.
10.4.2 Activation by Phosphorylation
An important insight into the molecular mechanisms by which the transcription of CRE-containing genes is induced came from experiments demonstrating that upon activation of the adenylyl cyclase pathway, a serine residue at position 133 of CREB is phosphorylated by PKA (Gonzalez and Montminy 1989). Phosphorylation appears indispensable for activation and the phospho serine cannot be substituted by other negatively charged residues (Lee et al. 1990). Whether phosphorylation by PKA modulates DNA binding by CREB is a slightly controversial point. Indeed, Yamamoto et al. (1988) indicate that PKA-mediated phosphorylation of CREB does not affect DNA binding. By contrast, Nichols et al. (1992) have reported that phosphorylation of CREB by PKA causes a modest increase in binding to high-affinity CRE sites but a stronger enhancement in binding to low-affinity CREs (Nichols et al. 1992). However, the major effect of phosphorylation seems rather to occur at the level of the trans-activation function of CREE. It has been proposed that this could happen by inducing a conformational change of the protein (Gonzalez et al. 1991). However, in contrast to this hypothesis, recent studies by Leonard et al. (1992) have shown that CREB can be a very potent activator in the absence of phosphorylation in the pancreatic islet cell line Tu6. The mechanism of this phosphorylation-independent activity remains to be determined. Interestingly, alternative signal transduction pathways can also induce phosphorylation of serine 133. In PC12 cells, increases in the levels of intracellular Ca2+ by membrane depolarization cause phosphorylation of serine 133 and a concomitant induction of c-ios gene expression mediated by a CRE in the clos promoter (Sassone-Corsi et al. 1988; Sheng et al. 1990,1991). CREB mutants lacking serine 133 were unable to activate clos transcription (Sheng et al. 1991). Although Ca2+ calmodulin-dependent (CAM) kinases were shown to be able to phosphorylate serine 133 in vitro (Sheng et al. 1991; Dash et al. 1991), their role in vivo remains unclear, since PKA seems to be necessary for clos in-
The Nuclear Response to cAMP During Spermatogenesis 227
duction by Ca2+ influx in PC 12 cells (Ginty et al. 1991). In addition, CREB is also phosphorylated upon stimulation by transfonning growth factor (TG F~ 1), although the target residue remains to be determined (Kramer et al. 1991). This is interesting since CREB, which has been shown also to bind to AP-l sites (Masquilier and Sassone-Corsi 1992), binds more efficiently to an AP-\ site after TGF~1 stimulation.
Experiments by Hagiwara et al. (1992) propose a mechanism to explain the attenuation of CREB activity following induction by forskolin. Their results indicate that after the initial burst of phosphorylation in response to cAMP, CREB is dephosphorylated in vivo by protein phosphatase-I (PP-I) and transcription of the somatostatin gene is correspondingly reduced. However, Nichols et al. (1992) show that both PP-l and PP-2A can dephosphorylate CREB in vitro, resulting in a decreased binding to low-affinity CRE sites in vitro. Therefore, the precise role of PP-2A in the dephosphorylation of CREB remains to be detennined.
The structure of the transcriptional activation domain of CREB includes more than the phosphoacceptor region (see Fig. 2 for a schematic representation of CREB, CREM and activating transcription factor, A TF-I). Serine 133 is located in a region of about 50 amino acids containing an abundance of phosphorylated serines and acidic residues, the phosphorylation box (P-Box) or kinase inducible domain (KID), which was shown to be essential for trans-activation by CREB (Lee et al. 1990; see Fig. 2). Although phosphorylation of serine 133 appears indispensable for activation by CREB, it is not sufficient for full activity. An acidic region just downstream of serine 133 (140-DLSSD) was shown to be important for CREB function (Lee et al. 1990; Gonzalez et al. 1991). In addition, deletion of a region called a2, containing several sites that can be phosphorylated by caseine kinase (CKII) in vitro, caused a decrease in CREB activity, although differences in the magnitude of this decrease were reported (Lee et al. 1990; Gonzalez et al. 1991). Similarly, the significance of the first glutamine-rich domain (Q I) is not completely clear, since it was reported to enhance CREB activity by Gonzalez et al. (1991), while Lee et al. (1990) failed to find an effect when they deleted this region. However, this apparent contradiction might be caused by the different CREB isoforms studied by these two groups, since Gonzalez et al. (199\) studied CREBa/34 I, while Lee et al. (1990) have used CREBL'l /327 (see below). The current
228 P. Sassone-Corsi
KlD a2
bZip
Ql CIG] PKA CIGI
I"" I I
Q2 BD LZ
CREBA I I bZip II
CREMt
CKII I'KACKIl 111 1 I I
1 I I SO LZ
~
ATF-l I I
Fig. 2. Comparison of the general structures of the activators 3'-5' cyclic adenosine monophosphate (cAMP)-responsive element binding protein (CREB), cAMP-responsive element modulator, (CREM't), and ATF-1. Schematic representation of the structural features of CREB (M327) (Gonzales et al. 1991), CREM't (Foulkes et al. 1992) and ATF-l (Rai et al. 1989). Q1 and Q2 indicate glutamine-rich activation domains; kinase-inducible domain (KID) containing the PKA site (S 133 in CREB, S 117 in CREM't and S63 in ATF-I), the a2 domain with putative phosphorylation sites for caseine kinase (CKII, S94, 97, 100, 103 and 107 in CREB, T94, S97, 100 and 105 in CREM't and S38, 41, 44,47 and 51 in ATF-I), and the highly acidic downstream CKII site (S141 in CREB, SI40 in CREM't and absent in ATF-I). BD, basic domain; LZ, leucine zipper. Note that CREM contains two bZip domains, which are alternatively spliced in the different CREM isoforms (Foulkes et al. 1991a; see Fig. 3)
model to explain the activation of CREB suggests that, upon phosphorylation of serine 133 by PKA, a conformational change is induced, which leads to exposure of the glutamine-rich activation domains (Gonzalez et al. 1991). The other regions which were identified as being important for CREB function might be involved in correctly spacing the phosphorylation site with respect to the glutamine-rich domains. Verification of this model awaits determination of the crystal structure of both unphosphorylated and phosphorylated CREE.
Interestingly, two other CRE-binding factors have been reported to be activated by PKA. ATF-I was shown to activate transcription after cotransfection of the catalytic subunit of PKA (Rehfuss et al. 1991; Flint and Jones 1991). Although ATF-l can be phosphorylated by PKA in vitro, in vivo phosphorylation has yet to be demonstrated. Since A TF-l lacks the Ql domain and the CKII site C-terrninal of the PKA site (Hai et al. 1989) (Fig. 2), these results suggest that the KID region and the Q2 domain are sufficient to mediate cAMP-induced transcription.
The Nuclear Response to cAMP During Spermatogenesis 229
The recently described activator isofonn of CREM (see Sect. 5), CREM'T (Foulkes et al. 1992), can also mediate cAMP-induced transcription (Laoide et al. 1993) (Figure 3). CREM'T is phosphorylated by PKA in vitro as well as in vivo on serine 117, the counterpart of serine 133 in CREB (de Groot et al. 1993a). The results on CREM'T are important for several reasons. It has been shown that the CREM activator is phosphorylated by at least seven different kinases: PKA, protein kinase C (PKC), casein kinase I and II, Ca2+-dependent calmodulin, glycogen-synthase-3 and p34cdc2. Multiple and cooperative phosphorylation events occur on the CREM protein, causing various effects at the functional level. While PKA, PKC, and calmodulin appear to induce the transcriptional activation potential without changing the DNA-binding activity (de Groot et al. 1993a), phosphorylation by CKI and CKII significantly enhances binding of the factor to a CRE sequence. Phosphorylation by p34cdc2, in contrast (de Groot et al. 1993b), occurs on several sites and causes a decrease in the transactivation potential. This result could be of interest with respect to the cyclic processes which are characteristic of genn cell differentiation, since p34cdc2 is a cell-cycle regulated kinase.
10.4.3 The Role of the Glutamine-Rich Domains
Flanking the P-box of both CREM and CREB there are two regions in which there are about three times more glutamine residues than in the remainder of the protein. Glutamine-rich domains have been characterized in other factors such as AP-2 and Sp I (Williams et al. 1988; Courey and Tjian 1989) as transcriptional activation domains. The current notion is that they constitute highly charged surfaces of the protein which can interact with other components of the transcriptional machinery, such as the RNA polymerase II. Little infonnation is available about their contribution to transcriptional activation in conjunction with the P-box, although recent data strongly support the notion that their presence is an absolute requirement for transactivation (Laoide et al. 1993). Interestingly, two different CREM isofonns containing either Q 1 ('Tl) or Q2 ('T2) are both transcriptional activators (Fig. 2). The Q2 domain appears to confer a slightly higher activation potential than the Q I domain (Laoide et al. 1993). These results demonstrate that the
230 P. Sassone-Corsi
Ql and Q2 regions probably function in an additive manner to generate the full activation potential of CREMr.
A possible hypothesis concerning the way how the glutamine-rich domains may contribute to the function of the protein is that transcriptional activation could be achieved by allosteric conformational changes mediated by phosphorylation of the pobox. In this model, phosphorylation triggers the generation of an acidic face on the protein by a distant conformational change. Similar intramolecular mechanisms are implicated in the function of important enzymes such as PKA.
10.5 The CREM Gene
The discovery of the CREM gene opened a new dimension in the study of transcriptional response to cAMP (Foulkes et al. 1991 a). This is due to the remarkable genomic structure of the gene, which offers clues to the understanding of the generation of functional diversity in transcription factors. Because of its dynamic and modular structure, CREM is also the first gene known to encode multiple CRE-binding proteins with both antagonistic and activator function.
10.5.1 A Remarkable Genomic Structure and Cell-Specific Expression
The CREM gene was isolated from a mouse pituitary cDNA library screened at low stringency with oligonucleotides corresponding to the leucine zipper and basic region of CREE. The logic behind this approach is that the adenylyl cyclase pathway plays an important role in the modulation of the hormonal regulation in the pituitary gland. The most striking feature about the CREM cDNA is the presence of two DNA-binding domains (Fig. 2). The first is complete and contains a leucine zipper and basic region very similar to CREB; the second is located in the 3' untranslated region of the gene, out of phase with the main coding region, and contains a half basic region and a leucine zipper more divergent from CREB. Various mRNA isoforms have been identified that are obtained by differential cell-specific splicing. Alternative usage of the two DNA-binding domains was demonstrated in various
The Nuclear Response to cAMP During Spermatogenesis 231
tissues and cell types, where quite different patterns of expression were found (Foulkes et al. 1991a). This strongly contrasts with CREB and ATF expression, which is ubiquitous (Hai et al. 1989; Habener 1990), suggesting that these proteins have a role as constitutive regulator.
CREM expression appears to be finely regulated, both transcriptionally and posttranscriptionally. In fact, not only is ccll- and tissue-specific expression observed, but also the production of isoforms with different function. Three products with antagonistic activity were the first to be described (Foulkes et al. 1991 a). These isoforms revealed alternative usage of the two DNA-binding domains (a and ~ isoforms, see Figs. 3 and 4), as well as a small deletion of 12 aminoacids (y isoform). The potential or even more complexity of CREM regulation is hinted at by the possible usage of alternative poly(A) addition sites and by the presence of ten AUUUA sequences in the 3' untranslated region, elements thought to be involved in mRNA instability (Shaw and Kamen 1986). The strict cell- and tissue-specific expression of CREM is indicative of a pivotal function in the regulation of cell-specific cAMP response. This suggests that CREM occupies a central control point in the pituitary, since it is known that the physiology of this gland is finely regulated by a multiplicity of hormones whose coupled signal transduction pathways involve adenylyl cyclase. Interestingly, other well-described examples of cell-specific splicing include the genes encoding neuronal peptides and hormones in brain and pituitary cells (Leff et al. 1986). It appears clear, thus, that cell-specific splicing is a crucial mechanism of CREM regulation, which modulates the DNA-binding specificity and activity of the final CREM products.
10.5.2 Different Functions of the Multiple CREM Products
The CREM products share a high homology with CREB, especially in the DNA-binding domains and in the P-box region (see Fig. 2; de Groot and Sassone-Corsi 1993). In a hydrophobicity plot it will appear that CREMa and CREM~ have a very similar profile, since the only difference with CREMa is in the DNA-binding domain. Comparing these CREM isoforms to CREB it appears that the resulting proteins have the same basic structure, although, strikingly, the CREM antagonists are much smaller proteins (Borrelli et al. 1992). Indeed, sequence compari-
232 P. Sassone-Corsi
KID
CREMt Activator
CREMt I Activator
CREMt 2 Activator
CREMI3 Repressor
CREMy Repressor
CREMa Repressor
S-CREM Repressor
Fig. 3. Schematic representation of 3'-5' cyclic adenosine monophosphate (cAMP)-responsive element modulator (CREM) isoforms. A family of activators and repressors of cAMP-induced transcription are generated from the CREM gene by alternative splicing (CREMt, t I, t2, a, p, Yo Foulkes et al. 1991 a, 1992; Laoide et al. 1993) as well as internal translation initiation (SCREM; Delmas et at. 1992). Isoforms t, t1, t2 activate transcription due to the presence of one or both glutamine-rich domain (Ql and Q2). Isoforms a, p, y lack the glutamine-rich domains and encode repressors of transcription (Laoide et al. 1993). S-CREM, generated from CREMt mRNA by internal translation initiation contains one of the two glutamine-rich domain but acts as a repressor of transcription probably due to the lack of the kinase-inducible domain (KID) . CREMa contains the first DNA-binding domain (DBD) while the other isoforms encode the second DBD; the most N-terminal ATG indicates the initiation codon for all the isoforms except for S-CREM, which is obtained by the use of the internal ATG; y, is a specific domain in CREM, not present in the other CRE-binding proteins
son indicates that the two glutamine-rich regions are absent in CREM, despite the perfect conservation of the P-box.
The CREM proteins specifically recognize CREs and show the same binding properties as CREB. This is not surprising, considering the high homology in the DNA-binding domains between these proteins. CREM proteins containing either DNA-binding domain I or II heterodimerize with CREB (Foulkes et ai. 1991 a; Laoide et ai. 1993), although it
The Nuclear Response to cAMP During Spermatogenesis 233
appears that CREMa-CREB heterodimer formation is more favored than CREM~-CREB. These notions suggest that CREM proteins might occupy CRE sites as CREM dimers or as CREM-CREB heterodimers, thus generating complexes with altered transcriptional functions. In fact CREM products act by impairing CRE-mediated transcription, and as such are considered as antagonists of cAMP-induced expression. In transfection experiments using CRE reporter plasmids, it was demonstrated that CREMa, ~ and y antagonist proteins block the transcriptional activation obtained by the joint action of CREB and the catalytic subunit of the cAMP-dependent PKA (Mellon et al. 1989). These observations strongly support the notion that CREM antagonist proteins negatively modulate CRE promoter elements in vivo. An important question is how CREM proteins work. The two most likely hypotheses are as follows. According to the first scenario, CREM proteins dimerize and bind to CRE sites. Downregulation is achieved by the occupation of these sites, which are unavailable for CREB. Similarly, if CREB is already bound, CREM proteins might squelch them because of their possible higher affinity for a specific site. According to the second model, CREM proteins are able to dimerize with CREB to generate nonfunctional heterodimers. Negative regulation is achieved by titrating active CREB molecules, and CREM proteins could act as activator traps. Since both CREM dimers and CREB-CREM heterodimers bind to CRE sites, both hypotheses are justified and both mechanisms may operate. However, results by Laoide et al. (1993) indicate that the production of nonfunctional heterodimers is the most likely mechanism operating to obtain CREM-mediated antagonism of cAMP-induced transcription.
10.5.3 Antagonists and Activators from the Same Gene
The first cDNA clones which were characterized from the CREM gene encode antagonists of cAMP-induced transcription (Foulkes et al. 1991 a). The central role of splicing in the regulation of this gene was already clearly hinted at by the presence of the two alternative DNAbinding domains which are used differentially in a cell-specific fashion. The CREM antagonists share extensive homology with CREB but they
234 P. Sassone-Corsi
lack two glutamine-rich domains, which have been shown to be necessary for transcriptional activation in CREB (see Sect. 4.3; Fig. 3). Interestingly, the CREM gene also encodes an activator of transcription (Foulkes et al. 1992). In the adult testis (see Sect. 8), an isoform (CREM1:) has been identified which resembles in structure one of the antagonist forms (CREM~) but includes two exons that encode two glutamine-rich domains (see Figs. 3 and 4). This form has been demonstrated to trans-activate transcription from a CRE site. In adult testis, the CREM1: isoform is expressed alone and it constitutes an abundant species in late spermatocytes and spermatids (see Sect. 8). Importantly, the CREM1: transcript in testis encodes only the full-length CREM1: protein (Delmas et al. 1993). This is of interest since it was shown that the CREM activator transcript can also generate another protein with repressor function, S-CREM (see Fig. 3), by the alternative usage of an internal initiation AUG codon (Delmas et al. 1992).
The CREM mRNA isoforms are a graphic illustration of how alternative splicing can modulate the function of a transcription factor in a tissue- and developmental-specific manner (Foulkes and Sassone-Corsi 1992).
10.5.4 CREM: A Gene with Specialized Neuroendocrine Functions
Changes in intracellular levels of cAMP constitute a major regulatory mechanism of signal transduction in the CNS. To date, several nuclear effectors of this pathway have been characterized, although their functional relevance in brain has been unclear because of their widespread distribution and their constant expression. We have reported the specific and anatomically distinct expression of the antagonist isoforms of the CREM gene in adult rat brain and the rapid induction of the a and ~ isoforms in supraoptic neurons upon physiological stimulation (Mellstrom et al. 1993). All known CREM isoforms are represented in total brain RNA after polymerase chain reaction (PCR) amplification (Foulkes et al. 1991a). However, while more quantitative techniques such as RNase protection confirmed the presence of both activator and repressor transcripts, in situ hybridization analysis shows that in neural tissues the
The Nuclear Response to cAMP During Spermatogenesis 235
antagonist isoforms have a well-defined distribution pattern (Mellstr()m et al. 1993). In contrast, the activator CREM isoforms, which include the glutamine-rich domains, in common with CREB, have a more diffuse and general distribution.
A major point is that CREM differs from the other members of the CRE/ATF family in that specific isoforms are induced upon physiological stimulation. To date, genes of the CRE/ATF class have been described as noninducible (for reviews see Habener 1990; Borrelli et al. 1992). Osmotic stimulation resulted in a differential accumulation of the two antagonist isoforms CREMu and CREM~, but no change in CREMy or the activator CREMr. Consistent with previous reports (for reviews see Habener 1990; de Groot and Sassone-Corsi 1993), no induction was observed for CREB.
The induction of several genes in the supraoptic nucleus upon osmotic stimulation has been described previously, including the early response gene c~f()s (Verma and Sassone-Corsi 1987), which has been shown to undergo a rapid and transient induction (Sherman et a!. 1986; Carter and Murphy 1990; Sharp et a!. 1991). Since CREM antagonists are able to negatively trans-regulate the activity of the cJos promoter (Foulkes et al. 1991 b), the temporal correlation between the onset of induction of CREMu and ~ and the decrease of c~t()S transcript in supraoptic neurons would suggest a role for CREM antagonists as downregulators of cJos early induction in these neurons. Although the presently available data fall short of unambiguously demonstrating such a role for CREM (Foulkes et al. 1991b), they do provide a stimulating basis for future investigations.
A remarkable aspect of the distribution of CREM antagonists in brain is the high level of expression in the anterior thalamic nuclei. This region, forming a part of the forebrain limbic system, receives input from the hippocampus and the mammillary body of the hypothalamus and projects mainly to the cingulate cortex. This anatomical circuit has been associated with memory and integration of emotions. The significance of CREM expression in the anterior thalamus is unknown, hut since this area has been reported to show no induction of early response genes after brain stimulation (Morgan et a!. 1987; Sagar et a!. 1988; Bullitt 1989), it is tempting to speculate that the high basal expression of CREM antagonists could in part account for this phenomenon. In this respect, it is noteworthy that induction of c~fos in the thalamus after
236 P. Sassone-Corsi
peripheral nociceptive or convulsive stimulation occurs in nuclei of the central, midline, and ventral thalamic complexes (Sagar et al. 1988; Bullitt 1989), which in general show a weak hybridization signal for CREM.
A second important observation is the presence of CREM antagonists in almost all the motor nuclei of the brain stem, while sensory nuclei are generally negative. Exceptions to this are the superior olive, which is associated with the auditive perceptions, and the mesencephalic trigeminal nucleus, equivalent to the dorsal root ganglia related to propioceptive sensory information, in which CREM transcripts are present. Conversely, CREM expression in motor nuclei includes the somatic motor nuclei: oculomotor, trochlear, abducens, and hypoglossal; as well as the special visceral, trigeminal and facial nuclei; and the general visceral motor nucleus of the vagus. Other positive motor nuclei are the red nucleus, the deep cerebellar nuclei, and the pontine nucleus (Mellstrom et al. 1993).
Expression of CREM antagonist isoforms in several hypothalamic nuclei associated with homeostatic regulation has also been noted. Such is the case for magnocellular neurons in the supraoptic hypothalamic nuclei which respond to osmotic stimulation by the differential temporal induction of two of the antagonist isoforms, CREMa and CREM~. Other functionally related brain areas where the CREM antagonists are expressed are nuclei involved in visual processing: the suprachiasmatic nucleus, the dorsolateral geniculate nucleus, the lateroposterior thalamic nucleus, and the medial terminal nucleus of the accessory optic tract. The latter is supposedly involved in entrainment of endocrine rhythms by light and fine adjustment of head-eye coordination. The presence of CREM in the pineal gland also points to a possible role for CREM in the processing of visual information and the establishment of circadian rhythms (Sect. 9).
These findings are a further demonstration of the physiological importance of the CREM gene among the CRE/ A TF family of factors. The discovery of the anatomically specific pattern of expression of distinct CREM isoforms, together with their potential for inducibility, sheds new light on the mechanisms whereby cAMP regulates gene expression in the brain.
The Nuclear Response to cAMP During Spermatogenesis 237
10.6 Physiological Roles of eRE-Binding Proteins
Although the results described above clearly demonstrate that CRE binding factors are important for cAMP-mediated transcriptional regulation in cultured cells, not much is known about the specific physiological roles for these proteins. This is of importance because the crucial role played by the variations in cAMP levels in neuroendocrine regulations.
An interesting first clue for a physiological function of CREB came from experiments using transgenic mice that expressed a CREB mutant which cannot be phosphorylated by PKA (Struthers et al. 1991). Since cAMP serves as a mitogenic signal for the somatotroph cells of the anterior pituitary, the mutant eDNA was placed under the control of the somatotroph-specific promoter of the growth hormone gene. The pituitary glands of transgenic mice expressing this construct were atrophied and were deficient in somatotroph cells. Moreover, the transgenic mice exhibited a dwarf phenotype. No other cell type in the pituitary was influenced by expression of the transgene. These effects might arise from repression of genes involved in proliferation and pituitary-specific gene expression, such as c-fos and GHFI/Pit-l, although the expression of these genes was not analyzed in the transgenic animals (Struthers et al. 1991). It is noteworthy that the block of CREB function by the dominant repressor generated a transgenic phenotype equivalent to the one obtained by targeted cell death of the somatomammotrophs (Borrelli et al. 1989). This could be an indication that CRE-binding proteins are likely to have pivotal functions in the normal pituitary development.
A number of reports indicate a role for CRE-binding proteins in spermatogenesis. This was not unexpected, since the metabolism of Sertoli and Leydig cells, the somatic cell types that direct the maturation of germinal cells, is regulated by the pituitary gonadotropins follicle-stimulating hormone (FSH) and luteinizing hormone (LH), which in turn activate the adenylyl cyclase pathway. The most striking example of differential regulation of CRE-binding proteins during spermatogenesis comes from recent studies on CREM (Foulkes et al. 1992). By studying the expression of the CREM gene during spermatogenesis it was shown that it generates high levels of the CREMe activator isoform by coordinate insertion of two glutamine-rich domains in the repressor isofonn CREM~ (Figs. 3,4). As a consequence of these insertions, CREM is
238 P. Sassone-Corsi
switched into a powerful transcriptional activator. An abrupt developmental switch in CREM expression was observed during spermatogenesis. Premeiotic germ cells express only the repressor forms in low amounts, while from the pachytene spermatocyte stage onwards CREMt is expressed uniquely and in very high amounts (Foulkes et al. 1992). The role of CREM during spermatogenesis is discussed in detail in the next section.
10.7 CREM and Spermatogenesis
Spermatogenesis is a process of cellular differentiation in which diploid germ cell progenitors differentiate into haploid spermatozoa. This process is finely regulated by the hypothalamic-pituitary axis and requires the coordinated action of several hormones (Veldhuis 1991). In response to hormone stimulation, testicular cells initiate a cascade of events inducing changes in cellular metabolism and gene expression. The transmission of pituitary hormonal stimuli from the cell surface to the cytoplasm and ultimately to the nucleus is mediated by the cAMP signaling pathway. Thus, nuclear factors involved in the regulation of gene expression by cAMP are likely to be of crucial importance during germ cell differentiation. Findings on CREM indicate that this gene plays a pivotal role in governing cAMP-dependent gene expression during spermatogenesis.
10.7.1 Hormonal Control in Testis: Induction of the cAMP Signal Transduction Pathway
Proliferation and differentiation of germ cells is ultimately dependent on two hormones produced by the gonadotrophs of the anterior pituitary, LH and FSH (Steinberger 1971; Lostroh 1976). In the absence of these two hormones, spermatogenesis does not proceed beyond meiotic prophase. For instance, hpg mutant mice are deficient in gonadotropin-releasing hormone, and spermatogenesis is interrupted at the diplotene stage (Cattanach et al. 1977). It appears that the developing male germ cells lack receptors for LH and FSH and it is consequently believed that they receive hormonal signals via somatic cells present in the testis
The Nuclear Response to cAMP During Spermatogenesis 239
(Moore 197X; Santen 1987). LH and FSH receptors are located on the Leydig and Sertoli somatic cells, respectively (Steinberger 1971; Lostroh 1976). Stimulation of the Leydig cells by LH results in the secretion of testosterone in the interstitial compartment, which then diffuses into the seminiferous tubules where Sertoli and germ cells are located. Upon FSH and testosterone stimulation, Sertoli cells secrete peptides and other components into the seminiferous tubules which arc required for germ cell differentiation (Russell 1980; Grootegoed et al. 1986; Jegou et al. 1992).
FSH and LH hormones are ligands of distinct trans-membrane receptors which are coupled to stimulatory G proteins (Borrelli et al. 1992). Binding of these hormones to their receptors induces the activity of adenylyl cyclase which converts ATP to cAMP. The elevated intracellular level of cAMP stimulates the activity of the PKA, which subsequently phosphorylates various cellular proteins. Among these there are the CRE- binding proteins (Borrelli et al. 1992; Fig. I).
The expression of testis-specific isoforms of PKA suggests that the cAMP pathway in testis has specialized functions which may involve the phosphorylation of testis-specific targets (Pari set et al. 1989; Oyen et al. 1988, 1990; Lonnerberg et al. 1992). Here we discuss the importance of CREM. whose role during spermatogenesis appears to be crucial.
10.7.2 Nuclear Effectors ofthe PKA Pathway in Testis
The expression of two members of the CRE-binding protein family has been described in testis, namely, the CREM and CREB genes (Waeher et al. 1991; Foulkes et al. 1992; Ruppert et al. 1992). These genes share extensive regions of homology within the coding sequence and have a similar genomic structure (de Groot and Sassone-Corsi 1993). As discussed ahove. however, the CREM gene is remarkable for several reasons (see Sect. 5).
The expression of the CREM gene during spermatogenesis shows some unique features. CREM is expressed at low levels before meiosis and only repressor isoforms are observed. In postmeiotic germ cells, an alternative splicing event causes a switch in CREM function, from transcriptional repressor to activator (Foulkes et al. 1992; see Figs. 3.4).
240 P. Sassone-Corsi
We have also observed a quantitative switch in CREM expression, since from the pachytene spermatocyte stage onwards the CREM transcript accumulates at very high levels. In contrast, the CREB gene is expressed at low levels in both somatic and germ cells. The somatic cells express the full-length CREB protein while truncated isoforms are detected in germ cells (Waeber et al. 1991; Ruppert et al. 1992). These isoforms lack the C-terminal b-Zip domain required for nuclear transport and DNA binding; thus, their functional relevance remains a mystery since the CREB truncated isoforms lack the ability to directly regulate gene expression (Waeber et al. 1991; Ruppert et al. 1992). The role of CREB in spermatogenesis is further questioned by the observation that, in mice in which the CREB gene has been inactivated by homologous recombination, spermatogenesis proceeds normally (G. Schutz, personal communication).
In the mouse, spermatogenesis is a cyclic process occuring every 12 days. The seminiferous epithelial cycle has been classified into twelve stages representing different types of cellular associations found during the cycle (Oakberg 1956). The expression of CRE-binding proteins appears to be dependent on the developmental stage of the seminiferous epithelium; for example, the CREM activator protein is detected mainly in the spermatids at stage VII-VIII of the cycle, while the CREB transcript is predominantly expressed in the Sertoli cells at stage III-IV (Waeber et al. 1991; Delmas et al. 1993). These features reflect a strict regulation of gene expression and suggest that these proteins might function at restricted developmental stages of spermatogenesis.
These observations underline the importance of alternative splicing and of its regulation as part of the physiological mechanisms responsible for differentiation of germ cells.
10.7.3 The CREM Gene: Regulation and Function in Germ Cells
The developmental switch in CREM expression occuring during meiosis raises two major questions: (1) What is the physiological mechanism controlling this switch, is it only developmental or is it hormonally regulated? (2) What is the function of this abundant CREM activator in germ cells?
The Nuclear Response to cAMP During Spermatogenesis 241
10.7.3.1 FSH Regulates CREM Expression in Testis In seasonal breeders. such as the golden hamster, spermatogenesis ceases during winter. The analysis of CREM expression in the hamster revealed a seasonally regulated appearance of the transcript corresponding to the activator isoform (Foulkes et al. 1993). Seasonal fluctuations are known to require an intact hypothalamic~pituitary axis, suggesting that CREM expression in testis may involve hormonal modulation. Since spermatogenesis is controlled by pituitary hormones, the involvement of the pituitary gland in the control of CREM expression was investigated. The CREM switch is no longer observed in hypophysectomized animals, but can be restored by injection of FSH in a few hours (Foulkes et al. 1993). Since no FSH receptors are observed on the germ cells, while they are present on Sertoli cells, it is predicted that FSH stimulation is transmitted from Sertoli to the germ cells very rapidly. The action of FSH is specific since the switch is not induced by injection of other pituitary hormones, such as LH and prolactin. A molecular analysis shows that induction of the CREM activator transcript is due to the use of an alternative polyadenylation site within the CREM transcript which truncates the 3' untranslated region. Omission of destabilizer elements thereby enhances transcript stability (Foulkes et al. 1993). This observation is of interest since it clearly demonstrates that FSH regulates the function of a cAMP-responsive nuclear factor in germ cells. Importantly, these experiments constitute the first report indicating the effect of hormonal stimulation on transcription factor message stability.
10.7.3.2 CREM: A Regulator of Gene Expression in Haploid Germ Cells A first hint as to the role of CREM during spermatogenesis was indicated by its protein expression pattern. In seminiferous epithelium, CREM transcripts accumulate in spermatocytes and spermatids, but CREM protein is detected only in spermatids (Fig. 4; Delmas et al. 1993). Thus, CREM function is restricted to a specific type of germ cell, the haploid spermatid. The absence of CREM protein in spermatocytes reflects a strict translational control and indicates multiple levels of regulation of gene expression in testis. It will be extremely important to further analyze the mechanism of this translation delay and to define whether it is also hormonally dependent.
242 P. Sassone-Corsi
Fig. 4. Peroxidase staining of rat seminiferous tubules showing expression of the 3'-5' cyclic adenosine monophosphate (cAMP)-responsive element modulator (CREM) protein in spermatids. The CREM antibody used for this experiment was prepared against a bacterially produced CREMt protein. Note the differential intensity of staining in the various tubules, indicating that CREM expression is developmentally regulated
Phosphorylation by PKA activates CREM function allowing the relay of the hormonal signal from the cytoplasm to the nucleus (de Groot et al. 1993a). The CREM activator is efficiently phosphorylated by cAMP-dependent PKA activity endogenous to the spermatids, indicating that the CREM protein is a nuclear target for the cAMP pathway in haploid spermatogenic cells (Delmas et al. 1993).
Thc detection of CREM activator protein in spermatids coincides with the transcriptional activation of several genes containing aCRE motif in their promoter region (Table 1). These genes encode mainly structural proteins required for spermatozoon assembly (transition protein, protamine, RTI, etc.), suggesting a role for CREM in the activation of genes required for the late phase of spermatid differentiation. This observation implies that the transcription of some key structural genes is directly linked to hormonal control and consequently to the level of
The Nuclear Response to cAMP During Spermatogenesis 243
cAMP present in seminiferous epithelium. A demonstration of the role of CREM in the expression of one of these genes, RT7, was shown using in vitro transcription experiments. A CREM-specific antibody blocks RT7 in vitro transcription with nuclear extracts from seminiferous tubules but not with extracts from liver (Delmas et al. 1993). In conclusion, CREM might participate in testis- and developmentalspecific regulation of genes containing a CRE in their promoter region, by expressing the repressor isoforms before meiosis and high levels of the activator after meiosis.
10.8 Inducible cAMP Early Repressor: An Alternative CREM Product Giving a New Dimension to the Neuroendocrine cAMP Pathway
During studies of CREM expression in the neuroendocrine system and in neuroendocrine cell lines, an unexpected new facet emerged: namely, the transcription of the CREM gene is inducible by cAMP (Molina et al. 1993; Masquilier et al. 1993). Furthermore, the kinetics of this induction are reminiscent of genes of the early response class. This important finding further reinforces the notion that CREM products playa fulcral role in the nuclear response to cAMP since the expression of no other CRE-binding factor has been shown to be inducible to date (de Groot and Sassone-Corsi 1993). Upon detailed analysis of the induced CREM products, more surprises were in store. The promoter which directs expression of the previously characterized CREM isoforms (a, ~,y, and 1:) was not cAMP inducible; instead, an alternative promoter, lying within an intron near the 3' end of the gene directs the cAMP induced transcription of a novel truncated CREM product, termed inducible cAMP early repressor (lCER). Importantly, ICER is the smallest b-Zip factor yet described. It functions as a powerful repressor of cAMP-induced transcription and furthermore negatively autoregulates the ICER promoter (Molina et al. 1993). The expression of ICER was first described in the pineal gland where it is the subject of a dramatic circadian pattern of expression (Stehle et al. 1993). The f1uctuation in CREM expression in the pineal gland is finely regulated by clock-distal elements; rhythmic adrenergic signals generated by the suprachiasmatic nucleus in the brain are responsible for the circadian expression of
244 P. Sassone-Corsi
CREM in the pineal gland. It has been suggested that ICER may be involved in the regulation of melatonin synthesis (Takahashi 1993). Recent data implicate dynamic ICER expression as a general feature of neuroendocrine systems (N. S. Foulkes and E. Lalli, unpublished observations).
10.9 Conclusion
It has been known for many years that the Sertoli cells, through intimate contacts, communicate hormonal stimulation to the developing germ cells. The analysis of the nuclear targets of the cAMP signal transduction pathway provides important information on the molecular mechanisms by which the hormonal signal is transmitted to the nucleus. Figure 5 summarizes the sequential events explaining how FSH stimulation modulates the gene expression pattern in germ cells. Animal models with altered CRE-binding function should shed more light upon the developmental role of these transcription factors and enable identification of their specific in vivo targets. In this respect, the normal phenotype of mice in which the CREB gene has been deleted by homologous recombination (G. Schiltz, personal communication) suggests a redundancy in the nuclear response to cAMP. However, because of its physiological relevance within the differentiation of the germ cell, it is likely that disruption of the CREM gene would have significant consequences for spermatogenesis.
The CREM gene constitutes a paradigm that reveals the complexity of the cellular response to cAMP. Its modularity of function, which is mediated by alternative and cell-specific splicing events, is an example of the versatility that the cell has to accomplish in order to permit normal and regulated cell growth in response to several stimuli. The large number of CRE-binding proteins represents another level of complexity that possibly suggests the requirements of cell specificity and potential cross-talk mechanisms with the PKC pathway. Future studies will elucidate the complex interplays among these factors and how they modulate cell physiology through the regulation of gene expression.
The Nuclear Response to cAMP During Spermatogenesis 245
Sertoll cell
00 TeSlosterone
.eydlg cells /
000--· LH
FSH
\ __ PITUITARY
ST Activation of cellu lar targets required for sperm assembly
ST Induction of CR EM protein synthesIs
sc Stabilization of CREM activator transcript by FSH
SG Low expression of CREM repressors
Fig. S. Schematic represcntation of a section of a seminiferous tubule where the 3 ' -5' cyclic adenosinc monophosphate (cAMP)-responsive element modulator (CREM) expression pattern is indicated. CREM expression is regulated at multiple levels during spermatogenesis. Pre meiotic germ cells (spermatogonia, SG) express a low level of CREM repressor isoforms. During meiotic prophase, the pituitary follicle stimulating hormone (FSH) is responsible for the stabilizalion of CREM activalor transcripts in spermatocytes (SC); CREM protein, on the other hand, is detected only after meiosis in haploid spermatids (Sn. Note the strict relationships between the Sertoli and germ cells (arrows) . In the haploid spermatids, CREM proteins activate a number of cellular genes expressed specifically during spermatid maturation (Delmas et al. 1993). LH, luteinizing hormone
References
Andrisani OM, Hayes TE, Roos E, Dixon JE (l9X7) Identification of the promoter sequences involved in the cell-specific expression of the rat somatostatin gene. Nucleic Acids Res 15:5715- 5728
246 P. Sassone-Corsi
Auwerx J, Sassone-Corsi P (1991) IP-l: a dominant inhibitor of fos/jun whose activity is modulated by phosphorylation. Cell 64:983-993
Benbrook DM, Jones NC (1990) Heterodimer formation between CREB and Jun proteins. Oncogene 5:295-302
Berkowitz LA, Riabowol KT, Gilman MZ (1989) Multiple sequence elements of a single functional class are required for cyclic AMP responsiveness of the mouse c-fos promoter. Mol Cell BioI 9:4272-4281
Berridge MJ (1987) Inositol trisphosphate and diacylglycerol: two interacting second messengers. Annu Rev Biochem 56: 159-193
Borrelli E, Heyman R, Arias C, Sawchenko P, Evans RM (1989) Transgenic mice with inducible dwarfism. Nature 339:538-541
Borrelli E, Montmayeur JP, Foulkes NS, Sassone-Corsi P (1992) Signal transduction and gene control: the cAMP pathway. Crit Rev On cog 3:321-338
Bullitt E (1989) Induction of c-fos-like protein within the lumbar spinal cord and thalamus of the rat following peripheral stimulation. Brain Res 493:391-397
Busch SJ, Sassone-Corsi P (\ 990) Dimers, leucine zippers and DNA binding domains. Trends Genet 6:36-40
Cambier JC, Newell NK, Justement LB, McGuire JC, Leach KL, Chen ZZ (\ 987) Ia binding ligand and cAMP stimulates nuclear translocation of PKC in ~ lymphocytes. Nature 327:629-632
Carter DA, Murphy D (1990) Regulation of c-fos and c-jun expression in the rat supraoptic nucleus. Mol Cell Neurobiol 10:435-446
Cattanach BM, Iddon CA, Charlton HM, Chiappa SA, Fink G (1977) Gonadotrophin releasing hormone deficiency in a mutant mouse with hypogonadism. Nature 269:338-340
Comb M, Birnberg NC, Seasholtz A, Herbert E, Goodman HM (1986) A cyclic-AMP and phorbol ester-inducible DNA element. Nature 323:353-356
Courey AJ, Tjian R (1989) Analysis of sp I in vivo reveals multiple transcriptional domains, including a novel glutamine activation motif. Cell 55:887-898
Dash PK, Karl KA, Colic os MA, Prywes R, Kandel ER (1991) cAMP response element-binding protein is activated by Ci+/calmodulin-as well as cAMP-dependent protein kinase. Proc Natl Acad Sci USA 88:5061-5065
de Groot RP, Sassone-Corsi P (1993) Hormonal control of gene expression: multiplicity and versatility of cyclic adenosine 3', 5' -monophosphate-responsive nuclear regulators. Mol Endocrinol 7: 145-153
de Groot RP, den Hertog J, Vandenheede JR, Goris J, Sassone-Corsi P (1993a) Multiple and cooperative phosphorylation events regulate the CREM activator function. EMBO J 12:3903-3911
The Nuclear Response to cAMP During Spermatogenesis 247
de Groot RP. Derua R, Goris l, Sassone-Corsi P (1993b) Phosphorylation and negative regulation of the transcriptional activator CREM by p34cdc2 Mol EndocrinoI7:1495-1501
Delcgcane A. Ferland L. Mellon PL (1987) Tissue specific enhancer of the human glycoprotein hormone a-subunit gene:dependence on cyclic AMP-inducible clements. Mol Cell BioI 7:3994-4002
Delmas Y, Laoide BM, Masquilier 0, de Groot RP, Foulkes NS, Sassone-Corsi P (1992) Alternative usage of initiation codons in mRNA encoding the cAMP-responsive-element modulator (CREM) generates regulators with opposite functions. Proc Natl Acad Sci USA 89:4226-4230
Delmas y, van der Hoorn F. Mellstrlim B. legou B, Sassone-Corsi P (1993) Induction of CREM activator proteins in spermatids: down-stream targets and implications for haploid germ cell differentiation. Mol Endocrinol 7: 1502-1514
Deutsch PJ, Hoeffler JP, Jameson JL, Habener JF (1988) Cyclic AMP and phorbol ester-stimulated transcription mediated by similar DNA elements that bind distinct proteins. Proc Natl Acad Sci USA 85:7922-7926
Flint Kl, Jones NC (1991) Differential regulation of three members of the ATF(CREB family of DNA-binding proteins. Oncogene 6:2019-2026
Foulkes NS, Sassone-Corsi P (1992) More is better: activators and repressors from the same gene. Cell 68:41 1--414
Foulkes NS, Borrelli E, Sassone-Corsi P (199Ia) CREM gene: usc of alternative DNA binding domains generates multiple antagonists of cAMP-induced transcription. Cell 64:739-749
Foulkes NS, Laoide BM, Schlotter F. Sassone-Corsi P (1991 b) Transcriptional antagonist CREM down-regulates c-fos cAMP-induced expression. Proc Natl Acad Sci USA 88:5448-5452
Foulkes NS, Mellstrom B, Benusiglio E, Sassone-Corsi P (1992) Developmental switch of CREM function during spermatogenesis: from antagonist to transcriptional activator. Nature 355:80-84
Foulkes NS. Schlotter F. Pevet P, Sassone-Corsi P (1993) Pituitary hormone FSH directs the CREM functional switch during spermatogenesis. Nature 362:264-267
Gilman AG (1987) G proteins: transducers of receptor-generated signals. Annu Rev Biochem 86:615-649
Ginty DO, Glowacka 0, Bader OS, Hidaka H, Wagner JA (1991) Induction of immediate early genes by C}+ influx requires cAMP-dependent protein kinase in PCI2 cells. J Bioi Chern 266: 17454-17458
Gonzalez GA, Montminy MR (1989) Cyclic AMP stimulates somatostatin gene transcription by phosphorylation of CREB at ser 133. Cell 59:675-680
248 P. Sassone-Corsi
Gonzalez GA, Yamamoto KK, Fischer WH, Karr K, Menzel P, Briggs W III, Vale WW, Montminy MR (1989) A cluster of phosphorylation sites on the cAMP-regulated nuclear factor CREB predicted by its sequence. Nature 337:749-752
Gonzalez GA, Menzel P, Leonard J, Fischer WH, Montminy MR (1991) Characterization of motifs which are critical for activity of the cyclic AMP-responsive transcription factor CREB. Mol Cell BioI II: 1306--1312
Grootegoed JA, Oonk RB, Toebosch AMW, Jansen R (1986) Extracellular factors that contribute to the development of spennatogenic cells. In: Stefanini M, Conti M, Geremia R, Ziparo E (eds) Molecular and cellular endocrinology of the testis. Excerpta Medica, Amsterdam, pp 215-225
Habener J (1990) Cyclic AMP response element binding proteins: a cornucopia of transcription factors. Mol Endocrionol4: 1087-1 094
Hai T-Y, Liu F, Coukos WJ, Green MR (1989) Transcription factor ATF cDNA clones: an extensive family of leucine zipper proteins able to selectively form DNA binding heterodimers. Genes Dev 3:2083-2090
Hagiwara M, Alberts A, Brindle P, Meinkoth J, Feramisco J, Deng T, Karin M, Shenolikar S, Montminy M (1992) Transcriptional attenuation following cAMP induction requires PP-I-mediated dephosphorylation of CREB. Cell 70:105-113
Heidaran MA, Kozak CA, Kistler WS (1989) Nucleotide sequence of the Stp-I gene coding for rat spermatid nuclear transition protein I (TP I): homology with protamine PI and assignment of the mouse Stp-I gene to chromosome 1. Gene 75:39-46
Hoeffler JP, Meyer TE, Yun Y, Jameson JL, Habener JF (1988) Cyclic AMPresponsive DNA-binding protein: structure based on a cloned placental cDNA. Science 242:1430-1433
Jegou B, Syed V, Sourdaine P, Byers S, Gerard N, Velez de la Calle J, Pineau C, Garnier DH, Bauche F (1992) The dialogue between late spennatids and Sertoli cells in vertebrates: a century of research. In: Nieschlag E, Habenicht U-F (eds) Spermatogenesis, fertilization, contraception, molecular, cellular and endocrine events in male reproduction. Springer, Berlin Heidelberg New York, pp 56-95 (Schering Foundation workshop, vol 4)
Johnson P, Peschon JJ, Yelick PC, Palmiter RD, Hecht NB (1988) Sequence homologies in the mouse protamine I and 2 genes. Biochim Biophys Acta 950:45-53
Kramer 11M, Koornneef I, de Laat SW, van den Eijnden-van Raaij AJM (1991) TGF-~l induces phosphorylation of the cyclic AMP responsive element binding protein in ML-CCL64 cells. EMBO J 10: 1083-1089
Krebs EG, Beavo JA (1979) Phosphorylation-dephosphorylation of Enzymes. Annu Rev Biochem 48:923-959
The Nuclear Response to cAMP During Spermatogenesis 249
Landschulz WH, Johnson PF, McKnight SL (19X8) The leucine-zipper: a hypothetical structure common to a new class of DNA binding proteins. Science 240: 1759-1764
Laoide BM, Foulkes NF, Schlotter F, Sassone-Corsi P (1993) The functional versatility of CREM is determined by its modular structure. EMBO J 12:1179-1191
Lee CQ, Yun Y, Hoeftler JP, Habener JF (1990) Cyclic-AMP-responsive transcriptional activation involves interdependent phosphorylated subdomains. EMBO J 9:4455-4465
Leff SE, Rosenfeld MG, Evans RM (1986) Complex transcriptional units: diversity in gene expression by alternative RNA processing. Ann Rev Biochern 55:1091-1117
Leonard J, Serup P, Gonzalez G, Edlund T, Montminy M (1992) The LIM family transcription factor lsi-I requires cAMP response element binding protein to promote somatostatin expression in pancreatic islet cells. Proc Natl Acad Sci USA 89:6247-6251
Lewin B (1991) Oncogenic conversion by regulatory changes in transcription factors. Cell 64:303-312
Lin SC, Morrison-Bogorad M (1991) Cloning and characterization of a testisspecific thymosin ~l() cDNA. J Bioi Chern 266:23347-23353
Lonnerberg P, Parvinen M, Jahnsen T, Hansson Y, Persson H (1992) Stageand cell-specific expression of cyclic adenosine 3',5' -monophosphate-dependent protein kinases in rat seminiferous epithelium. Bioi Reprod 46: 1057-1068
Lostroh AJ (1976) Hormonal control of spermatogenesis. In: Spilman CH, Lobi TJ, Kirton KT (eds) Regulation mechanisms of male reproductive physiology. Excerpta Medica, Amsterdam, pp 13-23
Maekawa T, Sakura H, Kanei-Ishii C, Sudo T, Yoshimura T, Fujisawa J, Yoshida M, Ishii S (1989) Leucine zipper structure of the protein CRE-BP I binding to the cyclic AMP response clement in brain. EMBO J 8:2023-202X
Masquilier D, Sassone-Corsi P (1992) Transcriptional cross-talk: nuclear factors CREM and CREB bind to AP-I sites and inhibit activation by Jun. J Bioi Chern 267:22460-22466
Masquilier D, Foulkes NS, Mattei MG, Sassone-Corsi P (1993) Human CREM gene: Evolutionary conservation, chromosomal localization, and inducibility of the transcript. Cell Growth Differ 4:931-937
McCormick A, Brady H, Theill L, Karin M (1990) Regulation of the pituitaryspecific homeobox gene GHFI by cell-autonomous and environmental cues. Nature 345:X29-832
250 P. Sassone-Corsi
McKnight SG, Clegg CH, Uhler MD, Chrivia JC, Cadd GG, Correll LA, Otten AD (1988) Analysis of the cAMP-dependent protein kinase system using molecular genetic approaches. Rec Prog Horm Res 44:307-335
Mellon PL, Clegg CH, Correll LA, McKnight SG (1989) Regulation of transcription by cyclic AMP-dependent protein kinase. Proc Nat! Acad Sci USA 86:4887-4891
Mellstri:im B, Naranjo JR, Foulkes NS, Lafarga M, Sassone-Corsi P (1993) Transcriptional response to cAMP in brain: specific distribution and induction of CREM antagonists. Neuron 10:655-665
Mizuki N, Sarapata DE, Garcia-Sanz JA, Kasahara M (1992) The mouse male germ cell-specific gene Tpx-l: molecular structure, mode of expression in spermatogenesis, and sequence similarity to two non-mammalian genes. Mammalian Genome 3:274-280
Molina CA, Foulkes NS, Lalli E, Sassone-Corsi P (1993) Inducibility and negative autoregulation of CREM: an alternative promoter directs the expression of ICER, an early response repressor. Cell 75:875-886
Montmayeur JP, Borrelli E (1991) Transcription mediated by a cAMP-responsive promoter element is reduced upon activation of dopamine D2 receptors. Proc Nat! Acad Sci USA 88:3135-3139
Moore RY (1978) Neuroendocrine regulation of reproduction. In: Yen SSC, Jaffe RB (eds) Reproductive endocrinology. Saunders, Philadelphia, pp 335-378
Morgan JI, Cohen DR, Hempstead JL, Curran T (1987) Mapping patterns of cfos expression in the central nervous system after seizure. Science 237:192-197
Nichols M, Weih F, Schmid W, DeVack C, Kowenz-Leutz E, Luckow B, Boshart M, Schiitz G (1992) Phosphorylation of CREB affects its binding to high and low affinity sites: implications for cAMP induced gene transcription. EMBO J 11 :3337-3346
Nishizuka Y (1986) Studies and perspectives of protein kinase C. Science 233:305-312
Oak berg J (1956) Duration of spermatogenesis in the mouse and timing of stages of the cycle of the seminiferous epithelium. Am J Anat 99:504-516
Oyen 0, Scott JD, Cadd GG, McKnight GS, Krebs EB, Hansson V, lahnsen T (1988) A unique mRNA species for a regulatory subunit of cAMP-dependent protein kinase is specifically induced in haploid germ cells. FEBS Lett 229:391-394
Oyen 0, Myklebust F, Scott JD, Cadd GG, McKnight SG, Hansson V, Jahnsen T (1990) Subunits of cyclic adenosine 3', 5' -monophosphate-dependent protein kinase show differential and distinct expression patterns during
The Nuclear Response to cAMP During Spermatogenesis 251
germ cell differentiation: alternative polyadenylation in germ cells gives rise to unique smaller-sized mRNA species. Bioi Reprod 43:46-54
Pariset C. Feinberg J, Dacheux JL, Oyen 0, Jahnsen T, Weinman S (1989) Differential expression and subccllular localization for subunits of cAMPdependent protcin kinase during ram spermatogenesis. J Cell Bioi 109:1195-1205
Rehfuss RP. Walton KM, Loriaux MM, Goodman RH (1991) The cAMP-rcgulated enhancer-binding protein ATF-I activates transcription in response to cAMP-dependent protein kinase A. J Bioi Chern 266: 18431-18434
Roesler WJ, Vanderbark GR, Hanson RW (1988) Cyclic AMP and the induction of eukaryotic gene expression. J Bioi Chern 263:9063-9066
Ruppert S, Colc TJ, Boshart M, Schmid E, Schutz G (1992) Multiple mRNA isoforms of the transcription activator protein CREB: generation by a1ternativc splicing and specific expression in primary spermatocytes. EMBO J I I : 1503-15 I 2
Russell LD (1980) Sertoli-germ cell interrelations: a review. Gamete Res 3: 179-202
Sagar SM, Sharp FR, Curran T (1988) Expression of c-fos protein in brain: metabolic mapping at the cellular level. Science 240: 1328-1331.
Santen RJ (1987) The testis. In: Felig P, Baxter JD, Broadus AE, Frohman LA (cds) Endocrinology and metabolism. McGraw-Hili, New York, pp 821-905
Sassone-Corsi P ( 1988) Cyclic AMP induction of early adenovirus promoters involves sequences required for E I A-transactivation. Pmc Natl Acad Sci USA 85:7192-7196
Sassone-Corsi P, Visvader J, Ferland L, Mellon PL, Verma 1M (1988) Induction of proto-oncogene fos transcription through the adenylate cyclase pathway: characterization of a cAMP-responsive element. Genes Dev 2:1529-1538
Sassone-Corsi P, Ransone LJ, Verma 1M (1990) Cross-talk in signal transduction: TPA-inducible factor Jun/AP-l activates cAMP responsive enhancer clements. Oncogene 5:427-431
Sharp FR, Sagar SM, Hicks K, Lowenstein D, Hisanaga K (1991) c-fos mRNA, Fos and Fos-related antigen induction by hypertonic saline and stress. J Neurosci 11 :2321-2331
Shaw G, Kamen R (1986) A conserved AU sequence from the 3' untranslated regioll of GM-CSF mRNA mediates selective mRNA degradation. Cell 46:659-667
Sheng M, McFadden G, Greenberg ME (1990) Membrane depolarizatioll and calcium induce c-fos transcription via phosphorylation of transcription factor CREB. Neuron 4:571-582
252 P. Sassone-Corsi: The Nuclear Response
Sheng M, Thompson MA, Greenberg ME (1991) CREB: a Ca2+-regulated transcription factor phosphorylated by calmodulin-dependent kinases. Science 252:1427-1430
Sherman TG, McKelvy JF, Watson SJ (1986) Vasopressin mRNA regulation in individual hypothalamic nuclei: a northern and in situ hybridization analysis. J Neurosci 6: 1685-1694
Stehle JH, Foulkes NS, Molina CA, Simonneaux V, Pevet P, Sassone-Corsi P (1993) Adrenergic signals direct rhythmic expression of transcriptional repressor CREM in the pineal gland. Nature 365:314-320
Steinberger E (1971) Hormonal control of mammalian spermatogenesis. PhysiolRev 51:1-22
Struthers RS, Vale WW, Arias C, Sawchenko PE, Montminy MR (1991) Somatotroph hypoplasia and dwarfism in transgenic mice expressing a nonphosphorylatable CREB mutant. Nature 350:622-624
Takahashi JS (1993) Circadian clocks a la CREM. Nature 365:299-300 van der Hoorn FA, Tarnasky HA (1992) Factors involved in regulation of the
RT7 promoter in a male germ cell-derived in vitro transcription system. Proc Nat! Acad Sci USA 89:703-707
Veldhuis 10 (1991) The hypothalamic-pituitary-testicular axis. In: Yen SSC, Jaffe RB (eds) Reproductive endocrinology. Saunders, Philadelphia, pp 409-459
Verma IM, Sassone-Corsi P (1987) Proto-oncogene fos: complex but versatile regulation. Cell 51:513-514
Vinson CR, Sigler P, McKnight SL (1989) Scissor-grip model for DNA recognition by a family of leucine zipper proteins. Science 246:911-922
Waeber G, Meyer TE, LeSieur M, Hermann HL, Gerard N, Habener JF (1991) Developmental stage-specific expression of cyclic adenosine 3',5' -monophosphate response element-binding protein CREB during spermatogenesis involves alternative exon splicing. Mol EndocrinoI5:1418-l430
Williams T, Admon A, Luscher B, Tjian R (1988) Cloning and expression of AP-2, a cell-type-specific transcription factor that activates inducible enhancer elements. Genes Dev 2: 1557-1569
Yamamoto KK, Gonzales GA, Briggs WH III, Montminy MR (1988) Phosphorylation-induced binding and transcriptional efficiency of nuclear factor CREB. Nature 334:494-498
Yoshimasa T, Sibley DR, Bouvier M, Lefkowitz RJ, Caron MG (1987) Crosstalk between cellular signalling pathways suggested by phorbol ester adenylate cyclase phosphorylation. Nature 327:67-70
Ziff EB (1990) Transcription factors: a new family gathers at the cAMP response site. Trends Genet 6:69-72
11 Transgenic Animals and the Study of Gonadal Function
M.M. Matzuk
I 1.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 253 11.2 Generation of Mammalian Models
Using Gene Targeting Strategies in ES Cells. . . . . . . . . . . . . . . .. 254 I 1.3 Infertility in Transgenic Male Mice ............... . . . . . . . .. 257 I 1.4 Tumor Development in the Testes of Transgenic Mice. . . . . . . .. 260 11.4. I Tumor Development via Expression of Oncogenes. . . . . . . . . . .. 261 11.4.2 Gonadal Tumor Development in p53 Mutant Mice . . . . . . . . . . .. 262 11.4.3 Inhibin-Deficient Mice. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 263 11.4.4 MIS-Mutant Mice. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 265 11.5 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 266 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 267
11.1 Introduction
Since the early 1980s, an avalanche of knowledge about mammalian development has been generated using transgenic mouse technology. A number of approaches using transgenic mice have been applied to address the role of specific protein products in mammalian development (Palmiter and Brinster 1986; laenisch 1988; Hanahan 1989; Stewart 1993). Until recently, the three main approaches to examine the role of a specific gene product have included overexpression of the transgene, alteration of gene function by expressing a mutant copy of the gene, and retroviral (or transgene) insertion into the gene of interest to generate a mutant allele. The classic example of overexpression of a trans gene
254 M.M. Matzuk
involved increased production of rat growth hormone (rGH) under control of the mouse metallothionein I promoter (Palmiter et al. 1982). In this study, increased production of rGH outside the pituitary was used to confirm its important physiological and developmental role. SimilarIy, production of a mutant form of the murine pro-l (I) collagen gene in transgenic mice altered development by disrupting normal collagen fibril synthesis and resulted in early death of the mice (Stacey et al. 1988). Finally, retroviral insertion has been used in a number of cases to disrupt the gene of interest (Schnieke et al. 1983) and therefore mimic specific human diseases.
Since 1990, a number of mutant mice have been produced with the advent of gene targeting and embryonic stem (ES) cell technology (Thomas and Capecchi 1987; Bradley 1987). This powerful technology involves disruption of specific genes in ES cells via homologous recombination between an introduced vector, containing sequences homologous to the gene of interest, and the endogenous ES cell gene. This technology allows one to answer questions via "loss of function" experiments as opposed to "gain of function" experiments with the standard transgenic technology described above.
This review will focus on transgenic animal studies of gonadal function, mainly on studies in which testicular function is altered. The majority of the studies have been performed using standard transgenic technology; however, with the rise of gene targeting/ES cell technology, several reports have now been described in which gonadal function is altered in the absence of the gene product.
11.2 Generation of Mammalian Models Using Gene Targeting Strategies in ES Cells
The steps involved in the generation of a gene "knockout" mouse are displayed in Fig. 1. These steps will be briefly described here and readers who wish more information are referred to an excellent recent review (Bradley et al. 1992). An important step in generating these mutant mice has been the isolation of totipotent ES cell lines from the inner cell mass of blastocysts, which can contribute to all tissues -including the germline. The first isolation of ES cell lines and contribution of these totipotent cells to the germline came in the early 1980s
Transgenic Animals and the Study of Gonadal Function 255
Totipotent embryonic stem cell line isolated
from a pre-Implantation blastocyst
Gene targeting applied to ES cell lin e
Isolation of clones with mutant gene
Blastocyst injection of genetically modified
ES cells
I Implantation of embryo into
25% homozygote
mutant animals ..expected
Intercross heterozygotes
./ ES cell contribution
to the gerrollne ,-Fl heterozygote
carrying the mutant gene
...... 1.m' m.<b.,
~ Chimeric
moUSe generated
Fig.I. Brief scheme for the generation of mutant mice via gene targeting in emhryonic stem cells (E\;)(modiried from Bradley et al. \(92)
from studies in Dr. Martin Evans laboratory (Evans and Kaufman 1981: Bradley et al. 1984). Only recently have ES cell lines (AB I and AB2.1 ) been isolated from mouse 129Sv strain-derived blastocysts which consistently contribute to the gennline when grown under the "appropriate" conditions (McMahon and Bradley 1990; Soriano et al. 1991; Matzuk ct
a1. 1992). One major goal in these experiments is to maintain the totipotential nature of these ES cells by growing them in an environment which limits differentiation and genetically undesired alterations. This can be achieved by growth in the presence of fetal calf serum supplemented with leukemia inhibitory factor (LIF; Williams et al. 1988), or more easily if the ES cells arc grown on a monolayer of mitotically inactive fibroblast cells synthesizing LIF (SNL76/7 STO cell line: McMahon and Bradley 1990).
For the homologous recombination events, most gene targeting strategies utilize "replacement" vectors containing positive and negative
256 M.M. Matzuk
selectable markers (Mansour et al. 1988) with sequences homologous to the target gene flanking the positive marker. There are three commonly used positive selectable markers, the neomycin resistance gene, the hygromycin resistance gene, and the HPRT gene, and two commonly used negative selectable markers, the herpes simplex virus thymidine kinase gene and the diphtheria toxin gene. The positive selectable markers enrich for the integration of the targeting vector into the ES cell genome, the negative markers for the homologous recombination event. ES cell clones containing the targeted event are identified by either Southern blot analysis (Ramirez-Solis et al. 1992) using external probes or by the polymerase chain reaction (PCR; Kim and Smithies 1988). Cells from targeted clones are then injected into the blastocoele of 3-day-old blastocysts (usually from C57BL/6 matings) and reimplanted into pseudopregnant females to generate chimeric mice which are a mixture of the injected ES cells and the inner cell mass (ICM) of the blastocyst (see Bradley 1987 for a technical review ofthese techniques). If the ES cells contribute to the germline, heterozygotes for the mutant allele can be obtained and intercrossed to obtain homozygotes with two copies of the targeted allele. These deficient mice are then worked up biochemically, morphologically and histologically to understand the role of the protein in mammals.
So why should one make a transgenic mouse which either overexpresses a protein or which contains a targeted mutation/deletion in a gene of interest? There are three main reasons, as follows: (l) to address the role(s) of specific protein products in mammalian development and oncogenesis; (2) to further understand mammalian development, including cell lineage development; and (3) to generate mammalian models for oncogenesis and other diseases, including developmental disorders. With respect to point 3, animal models are important for understanding the disease entity and, in addition, are needed for testing new therapeutic regimens, including gene therapy approaches.
What are the advantages of making a mouse deficient in a certain protein compared to more classical methods? Other methods to "block" the production or activities of a protein have included use of an anti-sense approach and the use of antibodies to the protein. In general, overexpression of antisense mRNA has not been very successful in ablating all of the mRNA coding for the protein of interest; however, neutralizing antibodies, which could be injected into animals to block the activities
Transgenic Animals and the Study of Gonadal Function 257
of a secreted protein, may "soak up" the antigen of interest. This approach is very time- and cost-consuming approach, though, and never truly answers the question of whether all of the activity is gone. This is especially true if the secreted protein has autocrine and/or paracrine functions. In this scenario, the neutralizing antibody may not "see" the antigen before it exerts its actions. Furthermore, this approach cannot be easily adapted to address the role of a protein during embryonic development. Clearly, the generation of a mouse with a null mutation is the only absolute means to answer the question of what happens in the complete absence of the gene product. In addition, a null mouse allows one to determine the essential role of a protein in development. Many proteins may demonstrate effects on cells in vitro and in vivo, and in situ hybridization and immunohistochemical analysis can be used to demonstrate the expression of mRNA or protein in specific cell types and tissues. However, only a null mouse can determine the relevant/essential role of the protein and confirm postulated theories on these roles.
11.3 Infertility in Transgenic Male Mice
If a transgenic mouse is viable to adulthood, male or female reproductive function is easily tested by breeding with wild-type mice. Sterility in male transgenic mice can be broken down into three main causes: (I) overexpression of the transgene in another tissue results in sterility; (2) overexpression of the transgene in the reproductive tract or precursor cells results in sterility; or (3) absence of the gene product (via transgene insertion, retroviral insertion, or gene "knockout") in either the gonads or another tissue results in sterility. In addition, the development of tumors will disrupt gonadal function (see next sections).
A number of sterile transgenic mice have been generated. In one case of sterility in a female transgenic mouse, mice deficient in LIF demonstrate normal ovarian function but fail to become pregnant secondary to failure of blastocysts to implant into the uterus (Stewart et al. 19(2). Thus, LIF expression in the uterus is critical for normal female reproductive function.
The majority of the published reports on males demonstrate that spermatogenesis is affected, as shown in Table 1. However, this altered spermatogenesis could be secondary to effects of the mutation or trans-
258 M.M. Matzuk
Table 1. Sterility in Transgenic Male Mice
Transgenic Mouse Observation Reference
MT-IFNa Diminished spermatogenesis and focal atrophy leading to complete testicular atrophy Hekman et al. (1988)
MT-IFN~ Testes involution and diminished spermatogenesis Iwakura et al. (1988)
HSV-tk Abnormal elongating spermatids and retention of mature spermatids. Braun et al. (1990)
MUP-tk Reduced spermatozoa AI-Shawi et al. (1988)
MT-MIS Feminization, abnormal Wolffian duct development, undescended testes (also female) Behringer et al. (1990)
ba-DT Small testes with altered Sertoli and Leydig cells and a block in spermatogenesis Kendall et al. (1991)
MT-hGH; Reduced fertility (failure to MT-hGHv inseminate females) Bartke et al. (1992)
Translocation Abnormal spermatogenesis Gordon et al. (1989) mutanta and Leydig cell hypertrophy
Symplastic Abnormal multinucleated MacGregor et al. spermatids (sysl syncytia (symplasts) (1990)
Germ-cell Absence of germ cells in both deficient (gcd)a males and females Pellas et al. (1991)
Lacking vigorous Abnormal spermatid shapes Magram and Bishop sperm (LVS)a (1991)
Paralysis mutanta Sperm paralysis secondary to Merlino et al. (1991) anoxeme defect
MT, mouse metallothionein I promoter; IFN, interferon; HSV-tk, herpes simplex virus thymidine kinase; MUP, major urinary protein promoter; MIS, miillerian inhibitory substance; h-DT, bovine glycoprotein subunit-diphtheria toxin gene; hGHv, human growth hormone variant. a Transgene insertion mutation.
Transgenic Animals and the Study of Gonadal Function 259
gene on any component of the system. Effects on the Sertoli cells or Leydig cells could indirectly alter sperm maturation, or direct effects at a specific time point during spermatogenesis could halt this process. Distinguishing these two scenarios may often be difficult. In this section, a few representative transgenic mice (described in Table I) will be discussed to give an idea of the types of alterations which can lead to infertility. One additional point is that males appear to require at least three genes on the Y chromosome for development of the undi fferentiated gonad into a normally functioning testis. At least two genes, one on the short arm and one on the long arm (Burgoyne et al. 1992) are necessary for normal spermatogenesis. The development of testes in XX transgenic mice carrying the Y chromosome gene Sry provided conclusive proof that this was the sex-determining gene (Koopman et al. 1991 ).
Regulation of nonnal gonadal function requires an intact hypothalamic-pituitary-gonadal axis. Humans and hypogonadal mice (hpg) with deficits in the hypothalamic peptide gonadotropin-releasing hormone (GnRH; Mason et al. 1986) have small testes secondarily to decreased luteinizing honnone (LH) and follicle-stimulating hormone (FSH). In a similar manner, ablation of the pituitary gonadotrophs which produce LH and FSH in transgenic mice (b-DT mice; Kendall et al. 1991) via directed expression of diphtheria toxin also results in mice with a similar hypogonadal phenotype.
Extragonadal and intragonadal expression of a secreted protein can influence gonadal function if there are gonadal receptors for these proteins to bind. Since normal spermatogenesis requires descent of the testes into the scrotum (i.e., a temperature lower than body temperature). any process which alters the descent will result in infertility. High level expression of mUllerian inhibiting substance (MIS) in males (Behringer et al. 1990) results in failure of the testes to descend. [n addition. these mice have other defects in the reproductive tract and feminized external genitalia. It is possible that enhanced MIS expression directly affects Leydig cell function, leading to decreased testosterone synthesis which is responsible for all of the effects seen (see Sect. 4.4, MIS-mutant mice). In a similar way, expression of the secreted proteins interferon (IFN)a (Hekman et al. 1988) and IFN~ (Iwakura et al. 1988) results in sterility secondary to testicular atrophy and diminished spermatogenesis. Since IFNa and IFN~ are normally synthesized in the gonads where
260 M.M. Matzuk
they have been postulated to have important paracrine and/or autocrine functions, it is reasonable that enhanced expression of the proteins in the testes (and in the circulation) would alter testicular homeostasis.
Expression of foreign DNA in the testes and insertion of transgenes into testis-expressed genes have also been demonstrated to disrupt spermatogenesis. Expression of the herpes simplex virus type I thymidine kinase (HSV-tk) gene product in haploid spermatids (AI-Shawi et al. 1988; Braun et al. 1990) has been demonstrated to disrupt spermatogenesis. Apparently there is a cryptic promoter within the coding region of the HSV -tk gene which allows the testes-specific transcription (AI-Shawi et al. 1991) leading to the lethal effects of this enzyme.
There are now several examples of insertion of the transgene into the genome resulting in a disruption of spermatogenesis. In one scenario, insertion of a trans gene into the genome resulted in a chromosome 2: 12 translocation and infertility (Gordon et al. 1989). Analysis of the testes demonstrated Leydig cell hypertrophy and absence of spermatozoa (likely secondary to abnormal chromosome synapsis). Alternatively, trans gene insertion into a testes-expressed allele has been demonstrated to result in male sterility when the insertion is bred to homozygosity (i.e., copies of the transgene present in the same allele on both chromosomes; MacGregor et al. 1990; Merlino et al. 1991; Pellas et al. 1991) and also has been seen in the hemizygous state (Magram and Bishop 1991). These defects include: (l) absence of germ cells possibly due to a defect in primordial germ cell migration (Pellas et al. 1991); (2) abnormalities in spermatids, including development of multinucleated syncytia (symplasts) by the spermatids (MacGregor et al. 1990) and abnormal spermatid shapes (Magram and Bishop 1991); and (3) spermatozoa paralysis secondary to abnormal axoneme (microtubule motility apparatus) stability/formation (Merlino et al. 1991). Thus, extragonadal and intragonadal mutations and expression of transgenes can lead to infertility and affect spermatogenesis.
11.4 Tumor Development in the Testes of Transgenic Mice
There have been many mammalian models of oncogenesis which have been developed using transgenic mouse technology (Palmiter and Brinster 1986; Hanahan 1989). Only a few of these transgenic mice develop
Transgenic Animals and the Study of Gonadal Function 261
gonadal tumors. In this section the tumors are described which develop in transgenic mice via inappropriate expression of an oncogene in the testes or via targeted disruption of tumor suppressor genes. These studies are summarized in Table 2.
11.4.1 Tumor Development via Expression of Oncogenes
As shown in Table 2, several transgenic mouse models have been generated via oncogene expression in the testes. The MIS-SV 40 T -antigen model was the only transgenic mouse group in which the expression was intentionally directed to the testes (specifically Sertoli cells; Peschon et al. 1992). In the other three transgenic lines, the oncogenes were found to be expressed in the testes and this expression resulted in immortalization/transfonnation of the cell types which expressed these genes (Kondoh et ai. 1991, Chalifour et ai. 1992, Paquis-Flucklinger et al. 1993). Interestingly, tumors have been derived from all major cell types of the testes, namely, genn cells, Leydig cells, and Sertoli cells. In addition, the types of testicular tumors which resulted from polyomavirus large-T antigen (Py L T) expression were different depending on the promoter. When the metallothionein I promoter was used, Leydig cell hyperplasia was seen initially and Leydig cell tumors later (Chalifour et al. 1992). However, when a viral-specific enhancer-promoter was used, Sertoli cell tumors were generated (Paquis-Flucklinger et ai. 1993). Thus, PyL T is capable of immortalizing either cell type.
In general, the development of these tumors were very slow in contrast to the inhibin-deficient mice (see below). Although tumors were seen in the MIS-SY40 T-antigen transgenic mice between 13 and 28 weeks (Peschon et ai. 1992), tumors were not seen prior to 8 months in the other three models. Thus, depending on the transgene, there appeared to be a long and variable latency period prior to overt tumor development.
Besides the know ledge that has been gained about the ability of these oncogenes to cause tumor development in mice, cell lines derived from these tumors have also been generated. These tumor cell lines will be a valuable resource in the future in understanding how these cell types function in vivo.
262 M.M. Matzuk
Table 2. Testicular tumors in transgenic mice
Mouse Tumor type(s) Reference
Human papillomavirus Germ cell tumors Kondoh et at. (1991) type 16 transgenic (seminomas)
MT-PyLT Leydig cell tumors; Chalifour et at. (1992) Leydig cell hyperplasia
PyL T (A2 strain) Sertoti cell tumors Paquis-Flucklinger et at. enhancer-promoter ( 1993)
MIS-SV40 T-antigen Sertoli cell tumors Peschon et at. (1992) transgenic
u-Inhibin-deficient mice Sex cord-stromal tumors Matzuk et at. (1992)
p53 Mutant mice Teratocarcinomas types Donehower et at. (1992); (primarily); other tumor Harvey et at. ( 1993 ) types
MIS mutant mice Leydig cell tumors; Behringer et at. Leydig cell hyperplasia (submitted)
MT, mouse metallothionein I promoter; PyLT, polyomavirus large-T antigen; MIS, mUllerian inhibiting substance.
11.4.2 Gonadal Tumor Development in p53 Mutant Mice
Mutations and allele loss of the p53 tumor-suppressor gene are the most frequently observed genetic changes in human cancers (Hollstein et al. 1991). Among the numerous tumor types where p53 has been postulated to playa role is in ovarian cancer. Loss of chromosome 17q, the location of the p53 gene, has been observed in 50% of ovarian epithelial tumor specimens (Hamilton 1992) and amplification of a mutant p53 gene is observed in 50% of tumor specimens in other analyses (Marks et al. 1991). The genetics of other types of ovarian cancers and testicular cancers is less well defined.
To generate a mouse model to understand p53 function in vivo, Donehower et al. 1992 utilized gene targeting/ES cell technology to produce mice homozygous for a mutant (null) p53 allele. These mice develop multiple types of tumors, including testicular tumors (Donehower et al. 1992; Harvey et al. 1993). Among the variety of testicular
Transgenic Animals and the Study of Gonadal Function 263
tumor types which have been observed are gonadoblastoma, Leydig cell tumor, seminoma, and teratocarcinomas. Interestingly, half of the male inbred 129Sv strain mice (9/18) develop testicular tumors with the majority of these tumors (7/8) being teratocarcinomas, for which this strain is susceptible. These data suggest that development of testicular tumors may be greatly influenced by the status of the p53 allele and that the status of other genes (i.e., the genetic background) will also influence the type of testicular tumors which will develop.
11.4.3 Inhibin-Deficient Mice
Over 20 members of the transforming growth factor (TGF)-~ superfamily of cytokines have been discovered to date. These growth regulatory proteins, synthesized as prepropeptides and processed to dimeric proteins, are structurally related in their mature, active C-terminal region. Among the mammalian members of this family are two inhibins (Vale et al. 1990), three activins (Vale et al. 1990), three TGF-~s (Roberts and Sporn 1990), MIS (Cate et al. 1990), seven bone morphogenetic proteins (BMPs; these proteins also have other names; Wozney et al. 1988; Ozkaynak et al. 1990; Lyons et al. 1991; Celeste et al. 1992), three growth/differentiation factors (GDF-I, GDF-3 and GDF-9; Lee 1990; McPherron and Lee 1993), nodal (Zhou et al. 1993), V g-I related protein 2 (V gr-2; Jones et al. 1992), and dorsalin (Basler et al. 1993). Members of this family have been shown to have diverse functions in a variety of assay systems, ranging from regression of the milllerian duct (MIS; Cate et al. 1990) to mesoderm induction (activins; Vale et al. 1990) to bone formation (BMPs; Wozney et al. 1988). These actions appear to be mediated through similar serine-threonine kinase transmembrane receptors (Massague 1992). Many studies are currently addressing the roles of the cytokines via murine ES cells/gene targeting strategies. In the next two sections, mice deficient in inhibin and MIS will be discussed.
The inhibins are heterodimeric growth factors (a:~A or a:~B) which share a common subunit (~A or ~B) with the activins (see Fig. 2). The inhibins were discovered as gonadal, endocrine peptides (Sertoli cell and granulosa cell products) which fed back on the gonadotrophs of the pituitary to inhibit FSH synthesis and secretion (Vale et al. 1990). Later
264
~ .•.•••••••••.• PA ••••••••••••
~ ACTIVIN A
~ ~
ACTIVIN B
M.M. Matzuk
~ ~ ACTIVIN AB
INHIBIN A INHIBIN B
Fig. 2. Structures of the inhibins and activins
studies have suggested important paracrine and autocrine roles of the inhibins in embryonic, extraembryonic (i.e., placenta), and adult tissues including the gonads, adrenal gland, and bone marrow (Meunier et al. 1988; Vale et al. 1990; Roberts et al. 1991).
To address the role of the inhibins in mammalian reproduction and development, we generated mice heterozygous and homozygous for a deletion of the inhibin subunit gene (Matzuk et al. 1992) which would abolish all inhibin activity (see Fig. 3). Inhibin-deficient mice were viable, indicating that inhibin was not essential for embryonic development. However, the inhibin-deficient mice developed gonadal sex cord stromal tumors (99% penetrance) in both males and females. The predisposition of inhibin-deficient mice to tumors identified inhibin as a novel tumor suppressor, the first secreted protein which has this role.
Adult inhibin-deficient females are always infertile. In contrast, adult inhibin-deficient males are initially fertile - sperm are present in the epididymis and males can mate with females. However, the development of the sex cord stromal tumors in the inhibin-deficient mice abolishes all sexual function. In the testis, there is a block in spermatogenesis and tubule degeneration is seen. These effects are visualized even in the testis contralateral to the tumor, and seminal vesicle atrophy is seen with early development of the tumors. Thus, there appears to be hormonal and growth factor/cytokine alterations with the development of the tumors.
Transgenic Animals and the Study of Gonadal Function 265
ACTIVIN INHIBIN
A S AS A S
Wild-Type + + + + +
fw.l!w. + + +
IIPA/IIPA + +
IIPS/ LIPS + +
IlPA/IlPA IIPS/ LIPS
Fig. 3. Forms of the activins and inhibins present in wild-type, a subunit-deficient (/l.a//l.a), ~A subunit-deficient (/I.~A//I.~A), ~B subunit-deficient (/I.~B//I.~B) and ~A subunit/~B subunit-deficient (/I.~A//I.~A; /I.~B//I.~B) mice
In addition to the synthesis of inhibin in the gonads and placenta, inhibin subunit mRNAs and/or proteins are also detected at lower levels in the pituitary, adrenal gland, spleen, brain, and spinal cord (Meunier et al. 1988; Vale et al. 1990). To address whether inhibin could function as a tumor suppressor in any of these tissues, we monitored gonadectomized inhibin-deficient mice for the development of tumors. Interestingly, nearly 100% of the gonadectomized inhibin-deficient male and female mice have developed adrenal cortical tumors, as compared to none of the control gonadectomized mice (Matzuk et al. 1994). Thus, these studies have identified inhibin as the first tumor suppressor with high specificity and high penetrance for the adrenal cortex. In addition, these studies may have direct relevance for the development of gonadal and adrenal tumors in humans.
11.4.4 MIS-Mutant Mice
Unlike many of the other members of the TGF-~ superfamily which are expressed in multiple tissues and cell-types, MIS expression is limited to both fetal and adult Sertoli cells of the male testis and postnatal granulosa cells in the female ovaries (Cate et al. 1990). The major function of MIS (for which it is named) is to cause regression of the
266 M.M. Matzuk
miillerian duct during embryonic development in males. It is unknown what function MIS plays in the postnatal male and female gonads. Several studies have suggested, however, that MIS may act as an "antiproliferative" growth factor in the postnatal female reproductive tract. This hypothesis is derived from the finding that MIS can inhibit the growth of some reproductive tract carcinoma cell lines (including ovarian carcinoma cell lines) in vitro and in vivo (Cate et al. 1990; Chin et al. 1991). In addition, MIS has been postulated to playa role in the descent of the testes in the male (Cate et al. 1990).
To test the above hypotheses, Behringer et al. (1990) have produced mice deficient in MIS. As might be expected from the limited expression of MIS, MIS-deficient mice are viable. In addition, MIS-deficient adult females are fertile, and adult males have descended testis (disproving one hypothesis), and in vitro, motile sperm from these males are functionally normal. However, as expected, MIS-deficient males do not have regression of the miillerian duct. Thus, the miilIerian duct develops into a uterus in the males, similar to the human syndrome known as persistent miillerian duct syndrome. Thus, these males essentially have two reproductive tracts. In addition, the presence of the uterus in these males appears to result in a blockage of sperm flow, since the males can copulate with females but cannot transfer sperm (which was found to be functionally normal in vitro as described above). A further finding in these mice, which addresses the role of MIS in tumorigenesis, is the development of Leydig cell hyperplasia and tumors in the male mice. Thus, MIS appears to be a critical protein in male reproduction and testicular development, and similar to inhibin, MIS appears to play a tumor suppressor role in the gonads.
11.5 Conclusions
A number of transgenic mice have been described which have altered gonadal function. These studies have led to a clearer understanding of the functions of a number of proteins in mammalian reproduction and oncogenesis. Future experiments, such as the isolation of the "sterility" genes at transgene insertion sites and analysis of other mutant mice will help us to understand the processes more clearly.
Transgenic Animals and the Study of Gonadal Function 267
References
AI-Shawi R, Burke J, Jones CT, Simons JP, Bishop JO (1988) A MUP promoter-thymidine kinase reporter gene shows relaxed tissue-specific expression and confers male sterility upon transgenic mice. Mol Cell Bioi 8:4821-4828
AI-Shawi R, Burke J, Wallace H, Jones C, Harrison S, Buxton D, Maley S, Chandley A, Bishop JO (1991) The herpes simplex virus type I thymidine kinase is expressed in the testes of transgenic mice under the control of a cryptic promoter. Mol Cell Bioi 11:4207-4216
Bartke A, Naar EM, Johnson L, May MR, Cecim M, Yum JS, Wagncr TE (1992) Effects of expression of human or bovine growth hormone genes on sperm production and male rcproductive performancc in four lincs of transgenic mice. J Reprod Fertil 95: 109-118
Basler K, Edlund T, Jessel TM, Yamada T (1993) Control of cell pallern in the neural tube: regulation of cell differentiation by dorsal in-I , a novcl TGF-~ family member. Cell 73:687-702
Behringer RR, Catc RL, Froelick GJ, Palmiter RD, Brinstcr RL (1990) Abnormal sexual development in transgenic miee chronically expressing Mullcrian inhibiting substance. Nature 345: 167-170
Behringer RR, Finegold MJ, Cate RL Mullerian inhibiting substance is required for normal male development and fertility (manuscript submilled)
Bradley A (1987) Production and analysis of chimeric mice. In: Robinson EJ (ed) Teratocarcinomas and cmbryonic stem cclls: a practical approach. IRL, Oxford, pp 113-151
Bradley A, Evans MJ, Kaufman MH, Robertson EJ (1984) Formation of germ linc chimeras from embryo-derived teratocarcinoma cell lines. Nature 309:255-256
Bradley A, Hasty P, Davis A, Ramirez-Solis R (1992) Modifying the mousc: design and desire. Biotechnology 10:534--539
Braun RE, Lo D, Pinkert CA, Widenl G. Flavell RA. Palmiter RD, Brinster RL (1990) Infertility in male transgenic mice: disruption of sperm development by HSV -tk expression in postmeiotic germ cells. Bioi Reprod 43:684-693
Burgoyne PS, Mahadevaiah SK. Sutcliffe MJ, Palmer SJ (1992) Fertility in mice requires X-Y pairing and a Y-chromosomal "spermiogenesis" gcnc mapping to the long arm. Cell 71:391-398
Cate RL, Donahoe PK, MacLaughlin DT (1990) Mullcrian inhibiting substance. In: Sporn MB. Roberts AS (eds) Peptide growth factors and their receptors II. Springer. Berlin Heidelberg New York. pp 179-210
268 M.M. Matzuk
Celeste AJ, Taylor R, Yamaji N, Wang J, Ross J, Wozney J (1992) Molecular cloning of BMP-8: a protein present in bovine bone which is highly related to the BMP-5/6!7 subfamily of osteoinductive molecules. J Cell Biochem [Suppl] 16F: 100
Chalifour LE, Mes-Masson A-M, Gomes ML, Wang N-S (1992) Testicular adenoma and seminal vesicle engorgement in polyomavirus large-T antigen transgenic mice. Mol Carcinog 5: 178-189
Chin T, Parry RL, Donahoe PK (1991) Human mullerian inhibiting substance inhibits tumor growth in vitro and in vivo. Cancer Res 51 :2101-2106
Donehower LA, Harvey M, Slagle BL, McArthur MJ, Montgomery CA, Butel JS, Bradley A (1992) Mice deficient for p53 are developmentally normal but susceptible to spontaneous tumours. Nature 356:215-221
Evans MJ, Kaufman MH (\ 981) Establishment in culture of pluripotential cells from mouse embryos. Nature 292:154-156
Gordon JW, Pravtcheva D, Poorman PA, Moses MJ, Brock WA, Ruddle FH (1989) Association of foreign DNA sequence with male sterility and translocation in a line of transgenic mice. Somat Cell Mol Genet 15:569-578
Hamilton TC (1992) Ovarian cancer, part I: biology. CUff Probl Cancer 16:1-57
Hanahan D (1989) Transgenic mice as probes into complex systems. Science 246: 1265-1275
Harvey M, McArthur MJ, Montgomery CA, Bradley A, Donehower LA (\ 993) Genetic background alters the spectrum of tumors that develop in p53-deficient mice. FASEB J 7:938-943
Hekman ACP, Trapman J, Mulder AH, van Gaalen JLM, Zwarthoff EC (1988) Interferon expression in the testes of transgenic mice leads to sterility. J Bioi Chern 263:12151-12155
Hollstein M, Sidronsky D, Vogelstein B, Harris CC (1991) p53 mutations. Science 253:49-53
Iwakura Y, Asano M, Nishimune Y, Kawade Y (1988) Male sterility of transgenic mice carrying exogenous mouse interferon-~ gene under the control of the metallothionein enhancer-promoter. EMBO J 7:3757-362
Jaenisch R (\988) Transgenic animals. Science 240:1468-1474 Jones CM, Simon-Chazottes D, Guenet J-L, Hogan BLM (1992) Isolation of
Vgr-2, a novel member of the transforming growth factor-~-related gene family. Mol EndocrinoI6:1961-1968
Kendall SK, Saunders JL, Jin L, Lloyd RV, Glode LM, Nett JM, Kevi RA, Nilson JH, Camper SA (1991) Targeted ablation of pituitary gonadotropes in transgenic mice. Mol Endocrinol 5:2025-2036
Kim H-S, Smithies 0 (1988) Recombinant fragment assay for gene targeting based on the polymerase chain reaction. NAR 16:8887-8903
Transgenic Animals and the Study of Gonadal Function 269
Kondoh G, Murata Y, Aozasa K, Yutsudo M, Hakura A (1991) Very high incidcnce of germ cell tumorigenesis (seminomagenesis) in human papillomavirus type 16 transgenic mice. J Virol 65:3335-3339
Koopman P, Gubbay J, Vivian N, Goodfellow P, Lovell-Badge R (1991) Male development of chromosomally female mice transgenic for Srv. Nature 351:117-121
Lee SJ (1990) Identification of a novel member (GDF-I) of the transforming growth ractor-~ superfamily. Mol Endocrinol4: 1034-1040
Lyons KM, Jones CM, Hogan BLM (1991) The DVR gene family in embryonic development. Trends Genet 7:408-412
MacGregor GR, Russell LD. Van Beek MEAB, Hanten GR, Kovac MJ, Kozak CA. Meistrich ML, Overbeek PA (1990) Symplastic spermatids (sys a recessive insertional mutation in mice causing a defect in spermatogenesis. Proc Natl Acad Sci USA 87:5016-5020
Magram J, Bishop JM (1991) Dominant male sterility in mice caused by insertion of a transgene. Proc Natl Acad Sci USA 88: 10327-10331
Mansour SL, Thomas KR, Capecchi MR (1988) Disruption of the proto-oncogene int-2 in mouse embryo-derived stem cells: a general strategy for targeting mutations to non-selectable genes. Nature 336:348-353
Marks JR, Davidoff AM, Kerns BJ, Humphrey PA, Pence JC, Dodge RK, Clarke-Pearson DL, Iglefort JD, Bast RC, Berchuck A (1991) Overexpression and mutation of p53 in epithelial ovarian cancer. Cancer Res 51 :2979-2985
Mason A, Hayfick JS, Zoeller RT, Young WS, Phillips HS, Nikolics K, Seeburg PH (1986) A deletion truncating the gonadotrophin-releasing hormone gene is responsible for hypogonadism in the hpg mouse. Science 234:1366-1371
Massague J ( 1(92) Receptors for the TGF-~ family. Cell 69: 1067-1070 Matzuk MM, Finegold MJ, Su J-GJ, Hsueh AJW, Bradley A (1992) a-Inhibin
is a tumour-suppressor gene with gonadal specificity in mice. Nature 360:313-319
Matzuk MM, Finegold MJ, Mather JP, Krummen L, Lu H, Bradley A (1994) Development of cancer cachexia syndrome and adrenal tumors in inhibindeficient mice. Proc Natl Acad Sci USA (in press)
McMahon AP, Bradley A (1990) The Wnt-I (lnt-I) proto-oncogene is required for the development of a large region of the mouse brain. Cell 62: 1073-1085
McPherron AC. Lee S-J (1993) GDF-3 and GDF-9: two new members of the transforming growth factor-~ superfamily containing a novel pattern of cysteines. J Bioi Chem 268:3444--3449
270 M.M. Matzuk
Merlino GT, Stahle C, Jhappan C, Linton R, Mahon KA, Willingham MC (1991) Inactivation of a sperm motility gene by insertion of an epidermal growth factor receptor transgene whose product is overexpressed and compartmentalized during spermatogenesis. Genes Development 5: 1395-1406
Meunier H, Rivier C, Evans RM, Vale R (1988) Gonadal and extragonadal expression of inhibin a, ~A, and ~B subunits in various tissues predicts diverse functions. Proc Natl Acad Sci USA 85:247-251
Ozkaynak E, Rueger DC, Drier EA, Corbett C, Ridge RJ, Sampath TK, Opperman H (1990) OP-l cDNA encodes an osteogenic protein in the TGF-~ family. EMBO J 9:2085-2093
Palmiter RD, Brinster RL (1986) Germ-line transformation of mice. Annu Rev Genet 20:465-499
Palmiter RD, Brinster RL, Hammer RE, Trumbauer ME, Rosenfeld MG, Birnberg NC, Evans RM (1982) Dramatic growth of mice that develop from eggs microinjected with metallothionein-growth hormone fusion genes. Nature 300:611-615
Paquis-Fluckinger V, Michiels J-F, Vidal F, Alquier C, Pointis G, Bourdon V, Cuzin F, Rassoulzadegan M (1993) Expression in transgenic mice of the large T antigen of polymavirus induces Sertoli cell tumours and allows the establishment of differentiated cell lines. Oncogene 8:2087-2094
Pellas TC, Ramachandran B, Duncan M, Pan SS, Monroe M, Chada K (1991) Germ-cell deficient (gcd) an insertional mutation manifested as infertility in transgenic mice. Proc Natl Acad Sci USA 88:8787-8791
Peschon JJ, Behringer RR, Cate RL, Harwood KA, Idzerden RL, Brinster RL, Palmiter RD (1992) Directed expression of an oncogene to Sertoli cells in transgenic mice using Mullerian inhibiting substance regulatory sequences. Mol Endocrinol 6: 1403-1411
Ramirez-Solis R, Rivera-Perez J, Wallace JD, Wims M, Zheng H, Bradley A (1992) Genomic DNA microextraction: a method to screen numerous samples. Anal Biochem 201 :331-335
Roberts AB, Sporn MD (1990) The transforming growth factor-Ws. In: Sporn MB, Roberts AB (eds) Peptide growth factors and their receptors I. Springer, Berlin Heidelberg New York, pp 419-472
Roberts VJ, Sawchenko PE, Vale W (1991) Expression of inhibin/activin subunit messenger ribonucleic acids during rat embryogenesis. Endocrinology 128:3122-3129
Schnieke A, Harbers K, Jaenisch R (1983) Embryonic lethal mutation in mice induced by retrovirus insertion into the a-I (I) collagen gene. Nature 304:315-320
Soriano P, Montgomery C, Geske R, Bradley A (1991) Targeted disruption of the c-src proto-oncogene leads to osteopetrosis in mice. Cell 64:693-702
Transgenic Animals and the Study of Gonadal Function 271
Stacey A, Bateman J, Choi T, Mascara T, Cole W, Jaenisch R (1988) Perinatal lethal osteogenesis imperfecta in transgenic mice bearing an engineered mutant pro-a I (I) collagen gene. Nature 332: 131-136
Stewart CL, Kaspar P, Brunet LJ, Bhatt H, Gadi I, Kontgen F, Abbondanzo SJ (1992) Blastocyst implantation depends on maternal expression of leukaemia inhibitory-factor. Nature 359:76-79
Stewart T A (1993) Models of human endocrine disorders in transgenic rodents. Trends Endocrinol Metab 4: 136-141
Thomas KR, Capecchi MR (1987) Site-directed mutagenesis by gene targeting in mouse embryo-derived stem cells. Cell 51 :503-512
Vale W, Hsueh A, Rivier C, Yu 1 (1990) The inhibin/activin family ofhormones and growth factors. In: Sporn MB, Roberts AB (eds) Peptide growth factors and their receptors II. Springer, Berlin Heidelberg New York, pp 211-248
Williams RL, Hilton DJ, Pease S, Willson TA, Stewart CL, Gearing DP, Wagner EF, Metcalf D, Niola NA, Gough NM (1988) Myeloid leukemia inhibitory factor maintains the developmental potential of embryonic stem cells. Nature 336:684-686
Wozney 1M, Rosen V, Celeste AI, Mitsock LM, Whitters MJ, Kriz RW, Hewick RM, Wang EA (1988) Novel regulators of bone formation: molecular clones and activities. Science 242: 1528-1534
Zhou X, Sasaki H, Lowe L, Hogan BLM, Kuehn MR (1993) Nodal is a novel TGF-~-like gene expressed in the mouse node during gastrulation. Nature 361 :543-547
12 Clinical Relevance and Irrelevance of Molecular and Cellular Research on the Testis
E. Nieschlag
12.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 273 12.2 The Expectations of Clinical Andrology. . . . . . . . . . . . . . . . . . . .. 274 12.3 Examples. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 27'5 12.3.1 Idiopathic Hypogonadotropic Hypogonadism
and Kallmann's Syndrome ............................... 27'5 12.3.2 GnRH Receptor and GnRH Antagonists .................... 277 12.3.3 Gonadotropins......................................... 277 12.3.4 Gonadotropin Receptors ................................. 279 12.3.'5 Androgen Receptor. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 281 12.3.6 Regulation of Spermatogenesis. . . . . . . . . . . . . . . . . . . . . . . . . . .. 282 12.4 The Role of Molecular and Cellular Research
in Clinical Andrology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 28'5 12.4.1 Current Lack of Therapeutic Applications. . . . . . . . . . . . . . . . . .. 28'5 12.4.2 The Time Factor ....................................... 286 12.4.3 Gene Therapy as a Goal ................................. 286 12.4.4 Integrated View ........................................ 286 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 288
12.1 Introduction
At first glance, an invitation to speak on the relevance and irrelevance of molecular and cellular research of the testis may sound absurd at a time when new clues to the pathophysiology of testicular disorders arc emerg-
274 E. Nieschlag
ing from molecular and cellular research with increasing frequency. There are numerous examples illustrating the relevance of molecular and cellular biology to clinical andrology. In fact, highest expectations are placed on molecular and cellular research to solve the burning problems of the field. So, why ask for its relevance or irrelevance to clinical andrology?
Increasingly, science and scientists are not taken for granted as an integrated constituent of society, but rather it has become common practice to question their need and their activities and role in society. The primary purpose of scientific research as the pursuit of knowledge in its own right and the acceptance of scientific research as a fundamental component of human existence are often questioned these days. More and more the public expects that research and knowledge bring them rewards and that these rewards be as immediate to the public benefit as possible. Granting agencies, largely representing the public, are put under pressure to look for such rewards when funding projects. Scientists involved in peer-dependent granting decisions are exposed to the same pressures and may similarly question the relevance and irrelevance of research projects and areas. This question becomes particularly critical in times of economic recession and shortage of available funds.
It may be in this context that the organizers of the 8th European Workshop on the Molecular and Cellular Endocrinology of the Testis are considering the relevance and irrelevance of molecular and cellular research for clinical andrology and in this context an answer will be sought.
12.2 The Expectations of Clinical Andrology
Andrology is the science of male reproduction. Consequently, clinical andrology deals with the diagnosis and treatment of disorders of male reproductive functions. As such, andrology contributes to the understanding of reproductive physiology in the male, seeking to explain the pathophysiology of fertility disorders in order to diagnose and treat these disorders properly and to provide means to regulate fertility.
When considering the high incidence and the known causes of male infertility (e.g., Nieschlag and Behre 1992) and the treatment modalities available (e.g., Nieschlag 1993) it soon becomes clear that the patho-
Clinical Relevance and Irrelevance 275
physiology of most conditions is not or is only poorly understood and rational treatment is not available for the majority of fertility disturbances. In addition, despite experimental progress in recent years, no pharmacological method for male fertility regulation is available for general use (Nieschlag et al. 1992). Thus the question of the contribution of molecular and cellular research to clinical andrology may become particularly pertinent. Clinical andrologists - and patients - may therefore ask what they can expect from molecular and cellular endocrinology.
12.3 Examples
Although andrology is a multidisciplinary field with many contributing specialities, endocrinology can be considered the backbone of andrology. Endocrinology has always provided clear pathophysiological concepts and rational treatment modalities for hypogonadism and infertility associated with hypogonadism. Therefore it is not surprising that most examples demonstrating the relevance of molecular and cellular research for clinical andrology derive from endocrinology. A few of these examples shall be mentioned here as a basis for the discussion of the question raised. Naturally, these examples derive predominantly from my own research.
12.3.1 Idiopathic Hypogonadotropic Hypogonadism and Kallmann's Syndrome
Idiopathic hypogonadotropic hypogonadism (IHH) is characterized by lack of gonadotropins and all symptoms of prepuberal androgen deficiency, including eunuchoid body proportions, lack of secondary sex characteristics, infantile penis, small testes, and azoospermia. In addition to these symptoms patients with Kallmann' s syndrome lack the sense of smell as their olfactory bulbs and tracks are defective. This condition appears to be caused by a defect in the migration of gonadotropin-releasing hormone (GnRH) producing and olfactory neurons from their common origin in the upper nasal part.
276 E. Nieschlag
Following the discovery of GnRH and its pulsatile secretion from the hypothalamus that is required for regular gonadotropin secretion (Belchetz et al. 1978), the lack of GnRH secretion from the hypothalamus was recognized as the cause of the hypogonadotropic hypogonadism in these patients. While the sequence of the GnRH gene was found to be normal in these patients (Weiss et al. 1991), the deletion of a gene, the KALIG-l gene, from the distal region of the short arm of the X-chromosome could be identified as the genetic cause of this disorder, in which the regular migration of the neurons is prevented (Franco et al. 1991). In mice bearing this genetic defect it could be demonstrated that the defect can be "healed" by gene therapy (Mason et al. 1986). While such corrective germline gene therapy in humans would have to be preceded by appropriate preimplantation diagnosis and appears to be reserved for the distant future, other scientific achievements enabled effective substitution therapy in patients with IHH or Kallmann' s syndrome.
These molecular investigations show the tremendous relevance of molecular research to the understanding of this disease and to clinical andrology in general. They also show the limitations of this approach. The principle of pulsatile hormone secretion and its importance for physiology could never have been discovered by techniques of molecular biology. Those discoveries require that the system or organism under investigation be viewed as a whole. In the case described here, the availability of synthetic GnRH and portable minipumps made possible the design of a "hypothalamic prothesis" that can be worn by the patient and that delivers GnRH in a pulsatile fashion in order to stimulate the pituitary to secrete luteinizing hormone (LH) and follicle-stimulating hormone (FSH) and to stimulate the testes to produce androgens and sperm (Hoffmann and Crowley 1982). Thus in these patients endocrine insufficiency and infertility can be treated effectively.
Notwithstanding the elegance of the molecular investigations described above, the dimensions of the problem should be kept in mind. Considering the high prevalance of infertility in the male popUlation and the incidence of IHH and Kallmann's syndrome, only about 1 of 300 men in consultation for infertility may have these disorders and benefit from such treatment. For the vast majority of patients these molecular and cellular biology findings may not be relevant, unless they can be considered as a model which may lead to the identification of similar mechanisms underlying other conditions.
Clinical Relevance and Irrelevance 277
12.3.2 GnRH Receptor and GnRH Antagonists
GnRH antagonists are considered decisive tools not only for the treatment of hormone-dependent cancers, for example, of the prostate, but also in the development of a male contraceptive. Although after extensive investigations in nonhuman primates (for review Weinbauer et a1. 1993) safe and effective antagonists are now available for clinical studies (Behre et al. 1994), they lack potency so that relatively large amounts of the peptides have to be administered, and due to their short half-life, they have to be injected, which makes them impractical for long-term use (Nieschlag et al. 1992). Ideally, new antagonists should not only be more potent but also orally effective. That oral effectiveness can be achieved for releasing-hormone analogs has been shown for growth hormone releasing hormone (GHRH) agonists (Smith et al. 1993) and endothelin antagonists (Clozel et a1. 1993); however, since classical chemistry appears to have exhausted its possibilities in the synthesis of new antagonists, it is now hoped that such GnRH antagonists can be designed by computer modeling, based on the known structure of the cloned GnRH receptor (Millar et a1. 1993). Although such GnRH antagonists have not yet become available, this approach to GnRH antagonist design illustrates another area of molecular biology highly relevant to clinical andrology.
12.3.3 Gonadotropins
The use of recombinant insulin and growth hormone has become firmly established in medical therapy. The sporadic occurrence of CreutzfeldJakob disease in patients who had been treated with growth hormone extracted from human pituitaries provided striking evidence of the need for "clean" biosynthetic peptide hormones. Gonadotropins can also be produced biosyntheticallY and recombinant FSH is currently undergoing clinical testing (Mannaerts et al. 1991). The advantage of recombinant gonadotropins over those of animal or human origin is their lack of impurities and pathogenic microorganisms. In addition, in the case of FSH, the recombinant hormone preparations are (most likely) free of LH contamination so that the "true" FSH activity pattern can be
278 E. Nieschlag
discerned (e.g., Fingscheidt et al. 1990 versus Weinbauer et al. 1994). The advantages of recombinant gonadotropins in the treatment of hypogonadotropic hypogonadism remain to be elucidated, but the combination of pure FSH with LH (or for the time being human chorionic gonadotrophin, hCG) should provide a chance to find more effective treatment regimens than those currently in use with urinary FSH preparations. However, the use of pure gonadotropins may also harbor disadvantages as, for example, impurities may have served as buffers smoothing the hormone action.
The microheterogeneity of circulating gonadotropins and their variations under different physiological and pathological conditions becomes more and more evident while the significance of this phenomenon is still poorly understood (Harsch et al. 1993). Although one might be tempted to consider microheterogeneity only an epiphenomenon without great physiological significance, it may actually be a subtle means to modulate gonadotropin action. Molecular techniques producing hybrids of the different gonadotropins (LaPolt et al. 1992) or changing the pattern of glycosylation (Galway et al. 1990) may help to elucidate the physiology of microheterogeneity and may at the same time lead to the identification of modified gonadotropins that could be useful in therapy. For example, a FSH depot preparation, eagerly awaited in the treatment of hypogonadotropic hypogonadism, might be produced by this technique or more potent preparations could be synthesized.
Another useful application of molecular research is exemplified by the search for FSH agonists or antagonists with the goal of applying these analogs in infertility treatment or male contraception. At the last European Testis Workshop, Hage-van Noort et al. (1992) demonstrated their systematic search and synthesis program based on computer modeling of the receptor binding sites of the FSH molecule, which might lead to new therapeutic modalities and applications in male contraception.
Mutations of the FSH molecule may also be identified as causes of male infertility. Recently, the case of a woman with primary amenorrhea, infertility, and isolated FSH deficiency was described in whom a two-nucleotide frame shift deletion in the coding sequence of the FSH~-subunit gene could be identified. The deletion resulted in an alteration of the amino acid codons 61-86 followed by a premature termination codon. The consequent truncated ~-subunit peptide lacked
Clinical Relevance and Irrelevance 279
the regions that are important for association with the a-subunit and for binding to and activation of the FSH receptor (Matthews et al. 1993). Since isolated FSH deficiency has also been described in infertile men, such mutations in the FSH~-subunit might also be identified as causes of male infertility. Other candidates for such mutations are infertile men in whom discrepancies between the immuno- and bioactivity of FSH have been found (Jockenh6vel et al. 1989).
Similar deletions in the gonadotropin molecule have also been described for the LH~-subunit. Weiss et al. (1992) found the substitution of arginine for glutamine in amino acid position 54 of the LH~-subunit as a cause of delayed puberty and hypogonadism. The mutated LH was immunologically active and as such could be measured in elevated concentrations in the circulation, but biological activity and LH receptor binding were lacking. While at this stage the defective LH molecule cannot be repaired, it can be substituted by exogenous hCG. This is another example of how molecular research can lead the way to rational treatment.
12.3.4 Gonadotropin Receptors
Another molecular approach to the elucidation of the causes of male infertility focuses on the gonadotropin receptors in the testis. The FSH receptor gene has been localized on chromosome 2 in close proximity to the LH receptor gene (Gromoll et al. 1994; Rousseau-Merck et al. 1993). Since no known diseases associated with infertility are localized in the vicinity of these receptor genes (Mitelman et al. 1991), for the time being this finding is only of academic interest and does not provide an immediate clue to causes of male infertility.
Currently, molecular research is being conducted to investigate the FSH receptor on Sertoli cells. Abnormal FSH receptors were postulated when decreased receptor binding of FSH was found in some infertile men (Namiki et al. 1984). However, specific defects in the FSH receptor have not yet been found in infertile men. Whether the alternative splicing of the human FSH receptor described by Gromoll et al. (1992) (Fig. I) may cause disturbances in FSH binding, and thereby infertility, remains to be seen. The expression of truncated isoforms of the FSH
280
hFSHR (695 aa)
hFSHR/E (633 aa)
E. Nieschlag
II III IV V VI VII VIII IX x 854 855
668 669
Fig. 1. Model of alternative splicing of the human testicular follicle-stimulating hormone receptor (hFSHR) transcript through a cassette exon mode. Exons, numbered I-X, are represented by filled boxes. Sequences lost in the splicing process are indicated by an open box. The length of the wild-type and of isoform E of the FSH receptor is shown in amino acids (aa). Nucleotide positions at splice junctions (exon VIII-IX and IX-X) are indicated (Gromoll et al. 1992)
receptor which act as secreted hormone-binding proteins or of membrane-spanning isoforms with altered binding and receptor activation characteristics may represent a mechanism for modulating target cell responsiveness to gonadotropins. These modulatory processes may be disturbed in certain cases of impaired spermatogenesis. The likelihood of such events in spermatogenesis is enhanced by the fact that similar alterations were found in another condition, namely, an alternative splicing pattern of the vasopressin receptor which was found to be the cause of congenital nephrogenic diabetes insipidus (Rosenthal et al. 1992).
Recently another gonadotropin receptor disorder has been identified as the cause of familial male precocious puberty. A striking feature of this disease is its independence of gonadotropin secretion. Shenker et al. (1993) discovered a mutation in the LH receptor of affected patients: a substitution of glycine for aspartate had occurred at position 578 in the sixth transmembrane helix of the LH receptor. This change causes an activation of the receptor even in the absence of LH and explains precocious testosterone production in Leydig cells even without LH stimulation: another beautiful example to demonstrate the relevance of molecular research for clinical andrology.
Clinical Relevance and Irrelevance 281
12.3.5 Androgen Receptor
Research on the androgen receptor can be considered the flagship of molecular andrology, easily demonstrating the relevance of molecular endocrinology to clinical practice. In an early phase of molecular biology abnormal binding of testosterone/dihydrotestosterone to the androgen receptor of genital skin fibroblasts was identified as the cause of syndromes of androgen resistance. Quantitative differences in the binding abnormality were found by radioreceptor assays in distinctive syndromes, ranging from complete absence of binding in testicular feminization to milder forms in the Reifenstein syndrome (for review Schweikert and Romalo 1990). Once the androgen receptor had been cloned and localized on the X-chromosome (Lubahn et al. 1989), various point mutations of the receptor associated with different conditions of androgen resistance were found (for review Brinkmann et al. 1992). Surprisingly, however, a clear correlation between the clinical phenotype of the androgen resistance and the various mutations of the androgen receptor could not be established (McPhaul et al. 1993), indicating that other factors must be involved in the expression of the phenotype and apparently suggesting the limitations of molecular research. However, at this stage it is unclear whether continued molecular research or other methodology will solve this dilemma.
Concerning the role of the androgen receptor in idiopathic infertility, a puzzling picture has emerged. In an early report, androgen resistance was identified as a cause of male infertility and an incidence of up to 50% was claimed (Aiman and Griffin 1982). Clinicians were initially enthusiastic when it seemed that a cause for idiopathic infertility had finally been found, but soon became skeptical when only a very few patients showed high LH combined with high testosterone levels in serum, which is an indicator of androgen resistance. Not surprisingly, therefore, in later reports a much lower incidence of androgen resistance was found as a cause of male infertility (Bouchard et al. 1986; Morrow et al. 1987) and the relevance of this research to clinical andrology was questioned. Nevertheless, individual cases in which androgen resistance causes infertility may exist, as shown by the recent description of a normally virilized man with azoospermia and a deletion in the androgen receptor gene (Akin et al. 1991).
282 E. Nieschlag
12.3.6 Regulation of Spermatogenesis
Since the 1930s it has been established that LH and FSH are required for spermatogenesis in mammals (including humans) and since the early 1960s hCG and human menopausal gonadotropin (hMG) of human urinary origin have been available to treat hypogonadotropic states. This treatment modality derived from classical endocrine research may soon be improved by recombinant gonadotropins, but otherwise, in the eyes of the clinician, the contributions of basic research to disorders of spermatogenesis appear to be remaining rather static.
However, during the last two decades intensive molecular and cellular research has been carried out to construe further the endocrine and paracrine regulation of spermatogenesis. A great number of factors have been found in the testis and a paracrine or autocrine role has been attributed to them. Although these factors and their potential roles have been concisely described in extensive reviews (e.g., Skinner 1991, Spiteri-Grech and Nieschlag 1993), a uniform system attributing a definite role to each of these factors in the concerto grosso of regulation of spermatogenesis has failed to emerge. The clinician gazes bewildered at pictures like the one shown in Fig. 2.
Even for the longest known and best characterized paracrine factor, testosterone, many questions on its action remain open: what level of testosterone is required to maintain normal spermatogenesis and how does testosterone exercise its role on spermatogenesis? Confronted with such gaps in information it is not surprising that all attempts to use testosterone and other androgens for the treatment of disorders of spermatogenesis have failed (for review Nieschlag 1993), while the extratesticular functions of testosterone can be well replaced in patients with hypogonadism by various testosterone preparations (for review Nieschlag and Behre 1990).
Inhibin is another example of this. Already postulated in the 1930s as a testicular factor active in the feedback mechanism, in the early 1970s inhibin was identified by the WHO Task Force for the Regulation of Male Fertility as a potential agent for male fertility regulation. It was anticipated that inhibin, once it was isolated and synthesized, would suppress FSH and thereby spermatogenesis. Consequently the Task Force initiated research that ultimately provided us with the considerable amount of information we have today on inhibin (Burger 1993;
CP
·Z
s.tn
C
RH
LF
~.MG
Vil
A
/ ,RS
P
MU
llS
CH
'CU
SG
P 1
& 2
T
nll
bu
min
Fig
. 2.
Syn
opsi
s o
f fa
ctor
s su
gges
ted
to b
e ac
tive
in
the
endo
crin
e an
d pa
racr
ine
regu
lati
on o
f L
eydi
g ce
ll fu
nct
ion
and
sper
mat
ogen
esis
. T
wo
tubu
lar
segm
ents
(w
ith
peri
tubu
lar
cell
s, g
erm
cel
ls a
nd S
erto
li c
ells
) ar
e sh
own
sche
mat
ical
ly;
two
bloo
d ve
ssel
s, e
ight
Ley
dig
cell
s an
d tw
o m
acro
phag
es a
re l
ocat
ed in
bet
wee
n. (
Dra
w
ing
base
d on
the
rev
iew
by
Spi
teri
-Gre
ch a
nd N
iesc
hlag
199
3)
Q
::l n' ~
:::IJ
Cll iO
< ~
::l
C'l
Cll ~
::l
Q.
.....
CD
iO
< ~
::l @
r\)
co
w
284 E. Nieschlag
Mather and Krummen 1992). Nevertheless, the role of inhibin (and its family members) in the feedback regulation of the testis is not fully determined and nothing applicable to clinical andrology has yet emerged, not to mention a lead to male contraception. Yet all of a sudden, a completely unexpected role as tumor suppressor has been attributed to inhibin, as shown by the development of testicular tumours in mice in which the (X-inhibin gene had been deleted (Matzuk et al. 1992).
The use of animals that genetically lack certain factors has helped to explain the physiological role of these factors. For example, dwarf rats with inborn growth hormone deficiency helped to clarify the role of growth hormone in the development of spermatogenesis. As these animals pass through puberty quite normally without any defects in spermatogenesis, the role of growth hormone in the regulation of spermatogenesis appears to be rather insignificant (Bartlett et al. 1990).
Other animal models potentially useful for the elucidation of disturbances of spermatogenesis were provided more or less accidentally by molecular research when creating transgenic animals for different purposes. A strain of transgenic mice bearing the Escherichia coli gene lacZ, encoding ~-galactosidase, on chromosome 14 developed symplastic spermatids which do not develop further and bring spermatogenesis to arrest (MacGregor et al. 1990). In another strain of mice made transgenic for the hematopoietic cell kinase (HCK) protooncogene nuclear condensation of the sperm did not occur and the animals were infertile, despite normal sexual development and behavior (Magram and Bishop 1991). Mice made transgenic for the goat fetal ~-globin gene showed early embryonic loss of primordial germ cells and were consequently infertile (Pellas et al. 1991). In these examples, the insertion of the foreign gene must have inactivated the endogenous genes responsible for the affected steps in spermatogenesis. Since similar abnormalities can be observed in infertile men, these findings indicate that creation and investigation of such transgenic model animals may be of clinical relevance. To date, however, this kind of research has contributed neither to the diagnosis nor to the therapy of male infertility.
Clinical Relevance and Irrelevance
12.4 The Role of Molecular and Cellular Research in Clinical Andrology
285
These examples could be extended to provide a large review of applications of molecular and cellular research to the investigation of male reproductive functions. However, even the few examples mentioned so far should be sufficient for some general conclusion.
12.4.1 Current Lack of Therapeutic Applications
So far, molecular and cellular research has not produced any new treatments for andrological diseases. However, effective treatment is the ultimate goal the clinician expects from basic research to consider it to be relevant. Molecular research has provided elegant explanations of the pathophysiology of certain disorders, but if treatments are available, they derive from "classical" endocrinological investigations such as the treatment of hypogonadotropic hypogonadism with gonadotropins or pulsatile GnRH. Biosynthetic gonadotropins may become the first example of the application of molecular endocrinology to andrological therapy.
If relevance is to be expressed in terms of the number of patients in need of proper pathophysiological explanation of their disorder and treatment, it is striking to note that molecular and cellular research has not contributed to the large nosologic entities in andrology: patients with varicocele, maldescended testes, and idiopathic infertility constitute about 60% of men attending infertility clinics. Pathophysiological concepts expounding these conditions have not emerged from molecular research, not to mention proper therapeutic modalities. Patients with syndromes of androgen resistance, Kallmann's syndrome, and familial male precocious puberty are very rare and represent orchids in the common flower - or rather weed - bed of clinical andrology. As regards male contraception, to date molecular and cellular research have failed to make a significant contribution. Although the WHO Task Force for the Regulation of Male Fertility already mentioned as well as other funding agencies launched a concerted initiative to exploit molecular and cellular research for the purpose of male contraception (Hamilton and Waites 1987), no practical lead has emerged.
286 E. Nieschlag
12.4.2 The Time Factor
Does this lack of clinical application of knowledge gained from molecular and cellular research justify the conclusion that it is irrelevant to clinical andrology? Before jumping to such conclusions it has to be considered that molecular and cellular biology are still young disciplines which have not yet had the chance to make their full impact on clinical medicine. The translation of basic knowledge into clinical practice always requires time. This time factor may be shortened by bridging the gap between molecular and clinical sciences. The spectrum of means to achieve this goal is multifaceted and includes activities such as joint meetings, collaboration of institutions, and close interaction of biologists and clinicians in the same unit.
12.4.3 Gene Therapy as a Goal
If molecular biology should be forced to generate clinical applications and thus prove its relevance, the most natural corollary would be gene therapy. Considering the technical problems of such therapy in other areas - not to mention ethical uncertainties - (Wivel and Walters 1993) it can be anticipated that it will still take considerable time before molecular research demonstrates its ultimate relevance to clinical andrology.
12.4.4 Integrated View
Meanwhile we should try to make molecular and cellular research as relevant as possible to our understanding of male infertility and male contraception. As the example of the regulation of spermatogenesis on a molecular and cellular level demonstrates, we are currently confronted with a vast body of knowledge on this topic, yet an integrated concept has not emerged. An additional dimension is needed to channel all this information into a meaningful systematic explanation which will then elucidate the pathophysiology of spermatogenic disorders. Techniques that will force the researcher to look not only at individual factors, receptors, or cells but at the testis as a whole organ will provide the
Clinical Relevance and Irrelevance 287
proper approach. In situ hybridization, in situ polymerase chain reaction, immunohistochemistry, and other techniques will provide such tools. This emphasizes that molecular and cellular research on the testes will be lost to clinical andrology if its findings cannot be verified in whole animal (including human) studies. Until such time as the information may be seen in the relevant context of the intact living organism, it remains piecemeal to clinical andrology and irrelevant - as symbolized in Fig. 3.
Whole animal studies
Molecular and cellular studie
Fig. 3. Pictograph symbolizing thc primacy of "whole animal" studics over molecular and cellular research in andrology
288 E. Nieschlag
Acknowledgements. I am grateful to the researchers at the Institute of Reproductive Medicine, University of Munster, for valuable, relevant suggestions, in particular to Dr. Martin Brinkworth, Dr. JOrg Gromoll, Dr. Manuela Simoni, Dr. Hermann Behre, and Dr. Gerhard Weinbauer. I am also grateful to my wife, Susan Nieschlag M.A., for editing the manuscript. The current work from the Institute of Reproductive Medicine quoted in this manuscript was supported by a Research Unit Grant from the Deutsche Forschungsgemeinschaft (DFG Ni 130/11).
References
Aiman J, Griffin JE (1982) The frequency of androgen receptor deficiency in infertile men. J Clin Endocrinol Metab 54:725-732
Akin JW, Behzadian A, Tho SPT, McDonough PG (1991) Evidence for a partial deletion in the androgen receptor gene in a phenotypic male with azoospermia. Am J Obstet Gynecol 165: 1891
Bartlett JMS, Charlton HM, Robinson ICAF, Nieschlag E (1990) Pubertal development and testicular function in the male growth hormone-deficient rat. J Endocrinol 126: 193-201
Behre HM, Backers A. Schlingheider A, Nieschlag E (1994) Sustained suppression of serum LH, FSH and testosterone and increase of high-density lipoprotein cholesterol by daily injections of the GnRH antagonist cetrorelix over 8 days in normal men. Clin Endocrinol (in press)
Belchetz PE, Plant TM, Nakai Y, Keogh EG, Knobil E (1978) Hypophysial responses to continuous and intermittent delivery of hypothalamic gonadotropin-releasing homone. Science 202:631-633
Bouchard P, Wright F, Portois MC, Couzinet B, Schaison G, Mowszowicz I (1986) Androgen insensitivity in oligospermic men: a reappraisal. J Clin Endocrinol Metab 63:1242-1246
Brinkmann AO, Jenster G, Kuoper GGJM, Ris-Stalpers C, van Laar JH, Faber PW, Trapman J (1992) Structure and function of the human androgen receptor. In: Nieschlag E, Habenicht UF (eds) Spermatogenesis-fertilization-contraception: molecular, cellular and endocrine events in male reproduction. Springer, Berlin Heidelberg New York, pp 97-122
Burger HG (1993) Clinical review 46: clinical utility of inhibin measurements. J Clin Endocrinol Metab 76: 1391-1396
Clozel M, Breu V, Burri K, Cassal JM, Fischli W, Gray GA, Hirth G, Loffler BM, Muller M, Neidhart W, Ramuz H (1993) Pathophysiological role of endothelin revealed by the first orally active endothelin receptor antagonist. Nature 365:759-761
Clinical Relevance and Irrelevance 289
Fingscheidt U, Weinbauer GF, Khan SA, Nieschlag E (1990) Follicle-stimulating hormone stimulates inhibin in the serum of male monkeys (Macaca mulatta). Acta Endocrinol (Copen h) 122:96-100
Franco B, Guilo S, Pragliola A, Incerti B, Bardoni B, Tonlorenzi R, Carrozoo R, Maestrini E, Pieretti M, Taillon-Miller P, Brown CJ, Willard HF, Lawrence C, Graziella Persico M, Camerino G, Ballabio A ( 1991 ) A gene deleted in Kallmann's syndrome shares homology with neutral cell adhesion and axonal path-finding molecules. Nature 353:529-536
Galway AB, Hsueh AJW, Keene JL Yamoto M, Fauser BCJM, Boime I (1990) In vitro and in vivo bioactivity of recombinant human follicle-stimulating hormone and partially deglycosylated variants secreted by transfected eukaryotic cell lines. Endocrinology 127:93-100
Gromoll J, Gudermann T, Nieschlag E (1992) Molecular cloning of a truncated isoform of the human follicle stimulating hormone receptor. Biochem Biophys Res Commun 188: 1077-1083
Gromoll J, Gudcrmann T, Ried T, Holtgreve H, Nieschlag E (1994) Localization of the human follicle-stimulating hormone receptor to chromosome 2 p21 using a genomic probe comprising exon 10. J Mol Endocr (in press)
Hage-van Noort M, Puijk WC, Schaaper WMM, Kuperus 0, Beekman NJCM, Plasman HH, Grootegoed JA, Mc!oen RH (1992) Development of antagonists and agonists of follicle-stimulating hormone. In: Nieschlag E, Habenicht UF (cds) Spermatogenesis-fertilization-contraception: molecular, cellular and endocrine events in male reproduction. Springer, Bcrlin Heidelberg New York, pp 33-55
Hamilton OW, Waites WMH (cds) (1987) Scientific basis of fertility regulation: cellular and molecular events in spermiogenesis. Cambridge University Press, Cambridge
Harsch lA, Simoni M, Nieschlag E (1993) Molecular heterogeneity of serum follicle-stimulating hormone in hypogonadal patients before and during androgen replacement therapy and in normal men. Clin Endocrinol (Oxt) 39:173-180
Hoffmann AR, Crowley WF Jr (1982) Induction of puberty in men by longterm pulsatile administration of low-dose gonadotropin-releasing homone. N Engl J Med 3047:1237-1241
10ckenhiivel F, Khan SA, Nieschlag E (1989) Diagnostic value of bioactive FSH in male infertility. Acta Endocrinol (Copen h) 121 :802-810
LaPolt P, Nishimori K, Fares FA, Perl as E, Boime I, Hsueh AJW (1992) Enhanced stimulation of follicle maturation and ovulatory potential by long acting follicle-stimulating hormone agonists with extended carboxyl-terminal peptides. Endocrinology 131 :2514-2520
290 E. Nieschlag
Lubahn DB, Brown TR, Simental JA, Higgs HN, Migeon CJ, Wilson EM, French FS (1989) Sequence of the intron/exon junctions of the coding region of the human androgen receptor gene and identification of a point mutation in a family with complete androgen insensitivity. Proc Nat! Acad Sci USA 86:9534-9538
MacGregor GR, Russell LD, van Beek MEAB, Hanten GR, Kovac MJ, Kozak CA, Meistrich ML, Overbeek PA (1990) Symplastic spermatids (sys): a recessive insertional mutation in mice causing a defect in spermatogenesis. Proc Nat! Acad Sci USA 87:5016-5020
Magram J, Bishop JM (1991) Dominant male sterility in mice caused by insertion of a transgene. Proc Nat! Acad Sci USA 88: 10327-10331
Mannaerts B, De Leeuw R, Geelen J, van Ravestein A, van Wezenbeek P, Schuurs A, Kloosterboer H (1991) Comparative in vitro and in vivo studies on the biological characteristics of recombinant human follicle-stimulating hormone. Endocrinology 129:2623-2630
Mason AJ, Pitts SL, Nikolics K, Szonyi E, Wilcox IN, Seeburg PH, Stewart T A (1986) The hypogonadal mouse: reproductive functions restored by gene therapy. Science 234:1372-1378
Mather JP, Krummen LA (1992) Inhibin, activin and growth factors: paracrine regulators of testicular function. In: Nieschlag E, Habenicht UF (eds) Spermatogenesis-fertilization--contraception: molecular, cellular and endocrine events in male reproduction. Springer, Berlin Heidelberg New York, pp 169-200
Matthews CH, Borgato S, Beck-Peccoz P, Adams M, Tone Y, Gambino G, Casagrande S, Tedeschini G, Benedetti A, Chatterjee VKK (1993) Primary amenorrhoea and infertility due to a mutation in the ~-subunit of folliclestimulating hormone. Nature Genet 5:83-86
Matzuk MM, Finegold MJ, Su JGJ, Hsueh AJW, Bradley A (1992) a-inhibin is a tumor-suppressor gene with gonadal specificity in mice. Nature 360:313-319
McPhaul MJ, Marcelli M, Zoppi S, Griffin JE, Wilson JD (1993) The spectrum of mutations in the androgen receptor gene that causes androgen resistance. J Clin Endocrinol Metab 76: 17-23
Millar RP, Davidson JS, Flanagan C, Golembo M, Illing N, Jacobs G, Becker I, Wakefield I, Eales A, Tsutsumi M, Chi L, Zhou W, Sealfon SC (1993) Towards the direct design of GnRH analogues: insights from cloned GnRH receptors. In: Oshima H, Burger HG (eds) Current topics in andrology. Japan Society of Andrology, Tokyo, pp 2-12
Mitelman F, Kakeo Y, Trent J (1991) Report of the committee on chromosome changes in neoplasia: 11 th international workshop on human gene mapping. Cytogenet Cell Genet 58:1053-1079
Clinical Relevance and Irrelevance 291
Morrow AF, Gyorki S, Warne GL, Burger HG, Bangah ML, Outch KH, Mirovics A, Baker HWG (1987) Variable androgen receptor levels in infertile men. J Clin Endocrinol Metab 64: 11 15-1121
Namiki M, Koide T, Okuyama A, Sonoda T, Itatani H, Miyake A, Aono T, Terada N, Matsumoto K (1984) Abnormality of testicular FSH receptors in infertile men. Acta Endocrinol (Copenh) 106:548-555
Nieschlag E (1993) Current therapy: care for the infertile male. Clin Endocrinol (Oxf) 38:123-133
Nieschlag E, Behre HM (1990) Pharmacology and clinical use of testosterone. In: Nieschlag E, Behre HM (eds) Testosterone: action, deficiency, substitution. Springer, Berlin Heidelberg New York, pp 92-114
Nieschlag E, Behre HM (1992) Male infertility due to testicular dysfunction. In: Drife 10, Templeton AA (eds) Infertility. Springer, Berlin Heidelberg New York, pp 65-80
Nieschlag E, Behre HM, Weinbauer GF (1992) Hormonal male contraception: a real chance? In: Nieschlag E, Habenicht UF (eds) Spermatogenesis-fertilization-contraception: molecular, cellular and endocrine events in male reproduction. Springer, Berlin Heidelberg New York, pp 477-501
Pellas TC, Ramachandran B, Duncan M, Pan SS, Marone M, Chada K (1991) Germ-cell deficient (GCD), an insertional mutation manifested as infertility in transgenic mice. Proc Natl Acad Sci USA 86:8787-8791
Rosenthal W, Seibold A, Antaramian A, Lonergan M, Arthus MF, Hendy GN, Birnbaumer M, Bichet DG (1992) Molecular identification of the gene responsible for congenital nephrogenic diabetes insipidus. Nature 359:233-235
Rousseau-Merck, MF, Atger A, Loosfelt H, Milgrom E, Berger R (1993) The chromosomal localization of the human follicle-stimulating hormone receptor gene (FSHR) on 2 p21-p 16 is similar to that of the luteinizing hormone receptor gene. Genomics 15:222-224
Schweikert HU, Romalo G (1990) Syndromes caused by androgen resistance. In: Nieschlag E, Behre HM (eds) Testosterone: action, deficiency, substitution. Springer, Berlin Heidelberg New York, pp 72-91
Shenker A, Laue L, Kosugl S, Merendino JJ Jr, Minegishi T, Cutler GB Jr (1993) A constitutively activating mutation of the luteinizing hormone receptor in familial male precocious puberty. Nature 365:652-654
Skinner MK (1991) Cell-cell interactions in the testis. Endocr Rev 12:45-77 Smith RG, Cheng K, Schoen WR, Pong SS, Hickey G, Jacks T, Butler B,
Chan WWS, Chaung LYP, Judith F, Taylor J, Wyvratt MJ, Fisher MH (1993) A nonpeptide growth hormone secretagogue. Science 260: 1640-1643
292 E. Nieschlag: Clinical Relevance
Spiteri-Grech J, Nieschlag E (1993) Paracrine factors relevant to the regulation of spermatogenesis - a review. J Reprod Fertil 98: 1-14
Weinbauer GF, Behre HM, Nieschlag E (1993) Gonadotropin-releasing hormone analog-induced regulation of testicular function in monkeys and men. In: Caraty A, Coelingh-Bennink HJT, Pavlou SN, Bouchard R (eds) GnRH-analogs, gonadotropins and gonadal peptides. Parthenon, Carnforth, pp 211-228
Weinbauer GF, Simoni M, Hutchison JS, Nieschlag E (1994) Pharmacokinetics and pharmacodynamics of recombinant and urinary human follicle-stimulating hormone in the male monkey. J Endocr (in press)
Weiss J, Adams E, Whitcomb RW, Crowley WF Jr, Jameson JL (1991) Normal sequence of the gonadotropin-releasing hormone gene in patients with idiopathic hypogonadotropic hypogonadism. Bioi Reprod 45:743-747
Weiss J, Axelrod L, Whitcomb RW, Harris PE, Crowley WF, Jameson JL (1992) Hypogonadism caused by a single amino acid substitution in the ~subunit of luteinizing hormone. N Engl J Med 326: 179-183
Wivel NA, Walters L (1993) Germ-line modification and disease prevention: some medical and ethical perspectives. Science 262:533-537
Subject Index
ABP 139 absence of LH 46 acrosomal cap 165 acrosomal exocytosis 154, 172 acrosome reaction 156, 165, 169,
175 ACTH 69 activins 263 acyl-CoA ester binding protein 75 adenine nucleotide carrier (ANC)
74 adenylyl cyclase 176 ADP ribosylation 172 adult Leyding cell development 39 adult -type Leydig cells 40 alkaline phosphatase 85 allograft 16, 18, 20, 21, 23, 27 alternate initiation 198 alternative polyadenylation site si-
gnals 194 amenorrhea 278 amino acid sequence 83 anchoring proteins 207 androgen receptor 127, 139,281 androgen receptor gene 281 androgen-regulated genes I 19 androgen-regulated proteins 119,
132 androgens 149
antiphosphotyrosine antibodies 167
antiprogestins 6 AP-l 4 ARPs 138-139 auto grafting 20 auto grafts 16, 18, 21-22 autotransplantation 20 azoospermic 18,29
B lymphoid cells 206 bFGF 48 bioactivity of FSH 279 block to polyspermy 162
c-fos 110 c-mos 147, 149 c-mos/CSF 149 calcium 107 cAMP 186 cAMP analog 77 cAMP binding sites 190 cAMP-dependent protein kinase
186 cAMP phosphodiesterase 82 cAMP responsive elements 199 cAMP responsiveness 198 capacitation 158 cat, dog 163
294
cattle, sheep, hamsters, rabbits, pigs 158
cell cycle 143-144, ISO cell cycle control lSI cell-extracellular matrix interacti-
ons 168 cell proliferation 201 cell-specific expression 207 cell surface receptors 173 centrosomallocalization 206 cervix 158 chemoattractants 157 chemokinesis 157 chemotaxis 154 chemotherapy/radiotherapy 26 cholesterol 68 cholesterol side-chain cleavage 68 chromatin 2 chromosomal assignment 188 condensation of DNA 137 congenital nephrogenic diabetes isi-
pidus 280 corpora lutea 9,76 cortical granule exocytosis 162 cortical granule glycosidase(s) 163 cortical granule reaction 164 CREB 139 CREM 139 Creutzfeld-lakob disease 277 cyclic AMP 176 cyclic nucleotide metabolism 174 -
175 cyclic protein-2 (CP-2) 132 cyclins 145 cycloheximide 69 cytochrome oxidase-II (COX-II)
122 cytochrome P450 side-chain cleava
ge (P450scc) 68 cytokinesis 146 cytostatic factor (CSF) 148-149
Subject Index
delayed puberty 279 des-(Gly-Ile )-endozepine 74 diacylglycerols 176 diazepam binding inhibitor (DBI)
73 dictinct genes 188 dihydotestosterone (DHT) 40 dimerization 190 dimethyl sulfoxide 82 DNA synthesis 49 DNA synthesis, the S-phase 143 down-regulation 4, 9
EDS 45,54, 124 ejaculated sperm 158 electrochemical potential 82 embryonic stem (ES) cell 254 endozepine 74 ES cell 254 ES cells 256 estrogen responsive element
(ERE) 2 ethane dimethane sulphonate
(EDS) 43, 116 exocytotic reaction 161 extracellular matrix 99 extracellular matrix (zona pellucida,
ZP) 156
familial male precocious puberty 280
female gametes 148 fertilization 153 fertilization potential 156 fetal-type Leydig cells 40 follicle stimulating hormone
(FSH) 259 formation of new Leydig cells 46 fos/jun 4 FSH 160,259,277,282 FSH agonists 278 FSH receptor 279
Subject Index
FSH receptor gene 9,279 fucosyl transferase 168 functional LH receptors 41
G protein activation 172 G protein 7 galactosylation 168 gamete activation 154 gametes 153 GCG software package 86 gene therapy 276,286 gene transcription 196, 207 GENEMBL 86 genetic recombination 147 germ cell-specific 194 germ cells 144,261 GHRH agonists 277 glomerulosa cells 77 glucose-regulated protein 78, 73 GnRH 276 GnRH antagonists 277 GnRH gene 276 GnRH receptor 277 gold-labeled ZP3 166 gonadal dysgenesis 17 gonadoblastoma 262 gonadotrophins 37,46, 146-149 gonadotropin receptors 7 gonadotropin-releasing hormonc
(GnRH) 259 granulosa cell 9, 263 growth factors 48 growth hormone 277 growth hormone deficiency 284 guanine nucleotide-binding regula-
tory protein 171 guinea pig 172
Hamster ZP3 164 hCG 43, 52, 55, 282 hCG stimulates the proliferation of
43
heat chock proteins (hsp90, hsp70) 5
heterodimer 192 heterokaryon experiments 5 hexokinase 167-168, 174 high-affinity GTPase 173 hMG 282
295
homodimers 191 hormonally-responsive cells hormone replacement therapy
166 17
hormone responsive elements 2 housekeeping genes 122 hypogonadal mice (hpg) 259 hypogonadism 17,275,279,282 hypothalamic prothesis 276
Idiopathic hypogonadotropic hypogonadism 275
idiopathic infertility 281, 285 IGF-I 48 immature cells 52 immature Leydig cells 53,54 immunophilin 5 in situ hybridization 121 in vitro fertilization 160 infertility 257 inhibin 125,263,264,266,282 inhibitory pathway 206 inner acrosomal membrane 164,
166 integrins 104 intercellular communication 156 interleukin-l 48, 51 intermediate filaments 71 intracellular effector 173 intracellular pH 175 ionic changes 172 ionic conductance 174 ischaemia 22-23, 26 isolated FSH deficiency 279 isozyme-specific effects 199 isozymes 192
296
KALIG-l gene 276 Kallmann's syndrome 275 kinetochores 147
lactate 130 laminin 100 leukemia inhibitory factor LIP 255 Leydig cell 19,48,52,55, 116,
260,261 Leydig cell hyperplasia 266 Leydig cell population 55 Leydig cell precursors 46 Leydig cell tumor 261, 262 Leydig cells in vitro 41 LH 39,41,72,259,282 LH and the 5a-reduced 40 LH receptor 7, 280 LH receptor gene 9,279 LH receptor transcripts 46 LIP 255,257 ligand-independent receptor
activation 5 ligand receptor occupancy 176 lipid droplets 71 luteinizing hormone (LH) 259 lymphocytes 199
M phase 144 MA-IO mouse Leydig tumor cells
69 major histocompatibility complex
27 maldescended testes 285 male contraception 278, 285 male fertility regulation 275, 282 male germ cells 149 male infertility 274 mannosidase 168 marine invertebrates 156 mature Leydig cells 52 mCCCP 82 meiosis 144, 146
Subject Index
meiosis-arresting factor 150 meiotic maturation 148 meiotic prophase 148 menopause 17 mesenchymal-like interstitial cells
40 meta-dinitrobenzene (mDNB) 130 methoxyaceticacid (MAA) 127 MHC 29 microfilaments 71 microheterogeneity 278 microsequencing 86 microsurgery 22, 29 microtubules 71 MIS 261,263,265-266 mitoplasts 85 molecular c10nin 187 motility 154 mouse 171-172 mouse egg 162 MPF 145,147 mRNA stability 196, 207 muscarinic acetylcholine receptor
173-174 mutations of the androgen receptor
281
N-formyl-Met-Leu-Phe 159 neurotransmitters 207 new Leydig cells 45 newly formed Leydig 54 NF-l 2 nitrobenzene (NB) 130 Northern analysis 121 Northern blots 122 nuclear condensation 284 nuclear localisation signals
(NLS) 5 nucleosome 2
O-linked carbohydrate 163 olfactory receptor gene 161
Subject Index
oocyte 14X, 162 orthophenanthroline X I ovarian graft 20 ovary 16,20,22,29 oviduct 156, 158 ovulated egg 156 ovulation 15X
P-Mod-S 133 p53 262 paraerine factor 2X2 PCR 121 peri tubular cells 133, 135 peroxisomes 71 pertussis toxin In phosphoproteins 76 phorbol diesters 176 phospholipase C 176 phospholipid metabolism 174--175 phosphorylation 4,79,167, 185 porphyrins 75 post translational modifications 4 pregnenolone 6X primary spermatocytes 150 primordial follicles 17-18, 22, 25 progesterone 77, In promote DNA synthesis 4X protease inhibitor-sensitive sites
168 protein kinase C 176, 196 protcin kinase inhibitor 189 protein stability 207 pseudo gene 198 PTX 172, 175 pyruvate 130
receptor 104, 154, 164, 165, 173, 174
receptor aggregati on 169, I 77 recombinant FSH 277 recombinant gonadotrophins 27X,
282
recombinat proteins 192 Reifenstein syndrome 281 reproductive tract 153
S phase 144 SCID (severe combined
297
immounodeficiency) 24 secondary spermatocyte 150 seminiferous epithelium 26, 103 seminiferous tubules 125 seminoma 263 serine 79 Sertoli cell 26, 101, 116, 124, 149,
196,261,263,279 signal transduction 106, 154, 169,
207 sperm 153, In sperm activation 165 sperm binding 169 sperm-associated receptors for
ZP3 166 sperm-egg fusion 164 sperm-egg recognition 162 spermatogenesis 1 X, 38, 115, 2X2 spermatogenic cycle 116 spermatogonial 26 splice variants 7 Sry 259 sterility 257,266 Steroidogenesis Activator Peptide
n Sterol Carriert Protein 2 (SCP2) 71 stimulus-secretion-coupling In subcellular localization 204 subtraction hybridization 119 symplastic spermatids 284 syndromes of androgen resistance
2X1
Tcells 201 T-cell activation 206
298
T-cell receptor 204 teratocarcinomas 263 testes 29 testicular feminisation 281 testicular heating 131 testicular transplantation 21,
25-26 testicular tumours 284 testosterone 18, 20, 26, 38, 54,
115,282 TFIIB 2 TGF- 48-49 theca intema 9 threonine 79 thymide kinase (HSV -tk) 260 TP-2 137 transcription 77 transfection 198 transgenic animals 284 transition protein-2 (TP-2) 122 translation 77 transplantation 16, 18 truncated isoforms of the FSH
receptor 279-280
Subject Index
TSH receptor gene 9 Tunica Propria 101
tyrosine kinase 167, 174 tyrosine-phosphorylated 168 tyrphostin RG-50864 167
urinary FSH preparations 278
varicocele 285 vascular anastomosis 21, 29 vasopressin receptor 280 voltage dependent anion carrier
(VDAC) 74
X-chromosme 276 xenograft 16, 24
Y -1 mouse adrenal tumor cells 73
ZP (Zona pellucida) 161-165, 168-169,173