Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Effects and mechanisms of growth differentiation factor 9 on activin A-regulated inhibin B and progesterone… Shi, Fengtao 2010

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2010_fall_shi_fengtao.pdf [ 1.89MB ]
Metadata
JSON: 24-1.0071346.json
JSON-LD: 24-1.0071346-ld.json
RDF/XML (Pretty): 24-1.0071346-rdf.xml
RDF/JSON: 24-1.0071346-rdf.json
Turtle: 24-1.0071346-turtle.txt
N-Triples: 24-1.0071346-rdf-ntriples.txt
Original Record: 24-1.0071346-source.json
Full Text
24-1.0071346-fulltext.txt
Citation
24-1.0071346.ris

Full Text

EFFECTS AND MECHANISMS OF GROWTH DIFFERENTIATION FACTOR 9 ON ACTIVIN AREGULATED INHIBIN B AND PROGESTERONE PRODUCTION IN HUMAN GRANULOSA CELLS by  Fengtao Shi B.Sc., Sun Yat-Sen University, 2000 M.Sc., Sun Yat-Sen University, 2003  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Reproductive and Developmental Sciences) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) October 2010 © Fengtao Shi, 2010  ABSTRACT Activin A (homodimer of inhibin βA-subunit) is known to increase inhibin βB-subunit and inhibin B (heterodimer of inhibin α- and βB-subunit) levels and decrease progesterone accumulation in human granulosa cells. Growth differentiation factor 9 (GDF9) is a potent paracrine regulator of ovarian function, but its overall effects, particularly relating to activin A actions, are unknown. We examined the potential crosstalk between activin A and GDF9 in primary cultures of human granulosa-lutein (hGL) cells. Pretreatment of hGL cells with GDF9 for 24 h resulted in an increased expression of activin receptors and Smad2/3, and decreased inhibitory Smad7 activity. These effects were attenuated by BMP type II receptor ectodomain (BMPR2 ECD), a GDF9 antagonist. These GDF9-induced changes, in turn, increased the cellular response to activin A stimulation and resulted in significantly greater production of βB-subunit mRNA and inhibin B compared to activin A treatment alone. Interestingly, endogenous GDF9 mRNA and protein were detected in hGL cells. Reduction of endogenous GDF9 by GDF9 siRNA resulted in decreased levels of activin receptors and Smad2/3/4, but increased expression of Smad7. Consequently, GDF9 siRNA treatment significantly attenuated the stimulation of activin A on βB-subunit mRNA and inhibin B levels. Additionally, GDF9 suppressed the expression of follistatin (FST) and follistatin-like 3 (FSTL3), which are extracellular inhibitors of activin A. These effects were attenuated by BMPR2 ECD and GDF9 siRNA. Treatment with FST or FSTL3 siRNA augmented activin A-induced βB-subunit mRNA levels. Conversely, GDF9 enhancement of activin A-induced βB-subunit mRNA was attenuated by FST. ii  Activin A decreased expression of StAR but not P450scc and 3βHSD, this effect lead to reduced basal and FSH-induced progesterone accumulation. GDF9 reversed these effects of activin A on StAR and progesterone; these GDF9 effects were attenuated by inhibin α-subunit siRNA. Together, these findings support a novel hypothesis that GDF9 exerts both paracrine and autocrine control of key components in the activin receptor-signaling pathway and the extracellular inhibition of activin A in hGL cells. As a result, GDF9 may serve to enhance activin A-induced accumulation of inhibin B, which in turn acts to reverse activin A suppression of progesterone accumulation during granulosa cell luteinization.  iii  PREFACE Refereed papers: z  Shi FT, Cheung AP, Huang HF, Leung PC 2010 Growth differentiation factor 9 (GDF9) suppresses follistatin and follistatin-like 3 production in human granulosa-lutein cells. Endocrionolgy, under review. This work is located at chapter 4.  z  Proportion of Contribution  Shi FT  Cheung AP  Huang HF  Leung PC  Identification and design of the research program  80%  5%  5%  10%  Analysis of the research data  85%  10%  0%  5%  Performance of the various parts of the research  100%  0%  0%  0%  Preparation of manuscripts  75%  20%  0%  5%  Shi FT, Cheung AP, Klausen C, Huang HF, Leung PC 2010 Growth differentiation factor 9 (GDF9) reverses activin A suppression of steroidogenic acute regulatory protein (StAR) expression and progesterone production in human granulosa-lutein cells. The Journal of clinical endocrinology and metabolism, accepted. This work is located at chapter 5. Proportion of Contribution  z  Shi FT Cheung AP Klausen C Huang HF Leung PC  Identification and design of the research program  75%  5%  5%  5%  10%  Analysis of the research data  80%  10%  5%  0%  5%  Performance of the various parts of the research  100%  0%  0%  0%  0%  Preparation of manuscripts  65%  30%  0%  0%  5%  Shi FT, Cheung AP, Huang HF, Leung PC 2009 Effects of endogenous growth differentiation factor 9 on activin A-induced inhibin B production in human granulosa-lutein cells. The Journal of clinical endocrinology and metabolism 94:5108-5116 (PMID: 19846738). This work is located at chapter 3. Proportion of Contribution  Shi FT  Cheung AP  Huang HF  Leung PC  Identification and design of the research program  80%  5%  5%  10%  Analysis of the research data  85%  10%  0%  5%  Performance of the various parts of the research  100%  0%  0%  0%  Preparation of manuscripts  75%  20%  0%  5%  iv  z  Shi FT, Cheung AP, Leung PC 2009 Growth differentiation factor 9 enhances activin A-induced inhibin B production in human granulosa cells. Endocrinology 150:3540-3546 (PMID: 19423755). This work is located at chapter 2. Proportion of Contribution  Shi FT  Cheung AP  Huang HF  Leung PC  Identification and design of the research program  80%  5%  5%  10%  Analysis of the research data  85%  10%  0%  5%  Performance of the various parts of the research  100%  0%  0%  0%  Preparation of manuscripts  70%  25%  0%  5%  Abstracts and Presentations: z  Shi FT, Cheung AP, Huang HF, Leung PC 2010 Growth differentiation factor 9 suppresses follistatin and follistatin-like 3 production in human granulosa-lutein cells. The 92nd Annual Meeting of the Endocrine (ENDO 2010), San Diego, CA, United States (Jun. 19-22, 2010) – Poster presentation  z  Shi FT, Cheung AP, Leung PC 2010 Mechanisms of growth differentiation factor 9 in enhancing activin A-induced inhibin B and subunits in human granulosa cells. CIHR-Institute of Human Development, Child and Youth Health Scientific Forum on Global Health, Vancouver, Canada (May 4, 2010) – Poster presentation  z  Shi FT, Cheung AP, Klausen C, Huang HF, Leung PC 2010 Effects of growth differentiation factor 9 (GDF9) and activin A on steroidogenic acute regulatory protein (StAR) and progesterone production in human granulosa-lutein cells. 14th International Congress of Endocrinology, Kyoto, Japan (Mar. 26-30, 2010) – Oral presentation  z  Shi FT, Cheung AP, Leung PC 2010 Mechanisms of growth differentiation factor 9 in enhancing activin A-induced inhibin B and subunits in human granulosa cells. The 91st Annual Meeting of the Endocrine (ENDO 2009), Washington, DC, United States (Jun. 10-13, 2009) – Poster presentation  z  Shi FT, Cheung AP, Leung PC 2009 Mechanisms of growth differentiation factor 9 in enhancing activin A-induced inhibin B production in human granulosa cells. OB/GYN Academic day, University of British Columbia, Vancouver, Canada (Mar.12, 2009) – Oral presentation v  z  Shi FT, Cheung AP, Leung PC 2008 Effects of growth differentiation factor 9 on activin A-induced inhibin βB-subunit expression in human granulosa cells. CFRI Student Research Forum, University of British Columbia, Canada (Jun.19, 2008) – Poster presentation  z  Shi FT, Cheung AP, Leung PC 2007 Growth differentiation factor 9 (GDF9) and activin A independently and synergistically stimulate inhibin subunits transcription and inhibin B secretion in human granulosa-luteal cells. The Auersperg Symposium on the Etiology of Ovarian Cancers, Vancouver, Canada (Jul. 15, 2007) – Poster presentation  Awards z  International Society of Endocrinology Travel Grant, Kyoto, Japan (2010)  z  Child and Family Research Institute (CFRI) Trainee Travel Grant, Vancouver, Canada (2010)  z  Interdisciplinary Women’s Reproductive Health Research Training Program (IWRH) Trainee Travel Grant, Vancouver, Canada (2010)  z  Endocrine Society's (ENDO) Trainee Award, Washington, United States (2009)  z  Interdisciplinary Women’s Reproductive Health Research Training Program (IWRH) Trainee Travel Grant, Vancouver, Canada (2009)  z  University of British Columbia Graduate Entrance Scholarship, Vancouver, Canada (2004)  Research Ethics Board z  UBC Children’s and Women’s Health Research Ethics Board Certificate Number: H90-70337  vi  TABLE OF CONTENTS Abstract ..................................................................................................................................................... ii Preface...................................................................................................................................................... iv Table of Contents .................................................................................................................................... vii List of Tables............................................................................................................................................ ix List of Figures ............................................................................................................................................x List of Abbreviations............................................................................................................................... xii Acknowledgements ................................................................................................................................ xix 1.  Literature Review...................................................................................................................... 1 1.1. 1.2. 1.3. 1.4. 1.5. 1.6. 1.7. 1.8.  2.  GDF9 Enhances Activin A-Induced Inhibin B accumulation in Human Granulosa-lutein Cells ................................................................................................................................................ 51 2.1. 2.2. 2.3. 2.4.  3.  Introduction............................................................................................................... 51 Materials and Methods.............................................................................................. 51 Results....................................................................................................................... 56 Discussion................................................................................................................. 58  Effects of Endogenous GDF9 on Activin A-Induced Inhibin B Accumulation in Human Granulosa-Lutein Cells ........................................................................................................... 69 3.1. 3.2. 3.3. 3.4.  4.  Introduction................................................................................................................. 1 Folliculogenesis .......................................................................................................... 2 Steroidogenesis in Ovary ............................................................................................ 3 Overview of TGFβ Superfamily ............................................................................... 10 Inhibin and Activin ................................................................................................... 23 FST and FSTL3 ........................................................................................................ 27 GDF9 and BMP15 .................................................................................................... 32 Objectives ................................................................................................................. 39  Introduction............................................................................................................... 69 Materials and Methods.............................................................................................. 70 Results....................................................................................................................... 72 Discussion................................................................................................................. 74  GDF9 Suppresses Follistatin and Follistatin -Like 3 Protein Accumulation in Human Granulosa-Lutein Cells ........................................................................................................... 86 4.1. 4.2.  Introduction............................................................................................................... 86 Materials and Methods.............................................................................................. 87  vii  4.3. 4.4. 5.  GDF9 Reverses Activin A Suppression of Steroidogenic Acute Regulatory Protein (StAR) Expression and Progesterone Accumulation in Human Granulosa-Lutein Cells.................. 105 5.1. 5.2. 5.3. 5.4.  6.  Results....................................................................................................................... 89 Discussion................................................................................................................. 92  Introduction............................................................................................................. 105 Materials and Methods............................................................................................ 106 Results..................................................................................................................... 108 Discussion............................................................................................................... 111  Conclusion and Recommendations for Future Work............................................................ 125 6.1. 6.2.  Conclusion .............................................................................................................. 125 Recommendations for Future Work........................................................................ 131  References ............................................................................................................................................. 135 Appendices ............................................................................................................................................ 165  viii  LIST OF TABLES Table 1.1 The type II receptors of the TGFβ superfamily and the ligands known to bind the receptor . 41 Table 1.2 The type I receptors of the TGFβ superfamily and the ligands known to bind the receptor... 41 Table 1.3 Combinational interaction of TGFβ superfamily receptors and Smads .................................. 42 Table 1.4 Smad-interacting proteins ....................................................................................................... 42 Table 1.5 Non-signalling binding proteins.............................................................................................. 44 Table S1 Nucleotide sequences of primers used for quantitative real-time PCR................................. 165  ix  LIST OF FIGURES Fig. 1.1  Members of the TGFβ superfamily play important roles in the bi-directional communication between oocyte and granulosa cells, and granulosa and theca cells. ...................................... 45  Fig. 1.2  The protein structure of TGFβ superfamily members............................................................. 46  Fig. 1.3  Diagrammatic representation of the TGFβ superfamily signaling pathway ........................... 47  Fig. 1.4  The protein structures of the three subfamilies of Smads ....................................................... 48  Fig. 1.5  Schematic diagram of different isoforms of inhibin and activin............................................. 49  Fig. 1.6  Levels of follicle-stimulating hormone (FSH), luteinizing hormone (LH), inhibin A, inhibin B and activin across the menstrual cycle .................................................................................... 50  Fig. 2.1  Concentration-dependent effect of activin A (A, C, E) or GDF9 (B, D, F) on inhibin subunit mRNA level in hGL cells........................................................................................................ 63  Fig. 2.2  Time-dependent effect of activin A (A, C, E) or GDF9 (B, D, F) on inhibin subunit mRNA level in hGL cells .................................................................................................................... 64  Fig. 2.3  GDF9 pretreatment enhanced activin A-induced βA- (A, B) and βB- (C, D) mRNA level in hGL cells, effects attenuated in the presence of BMPR2 ECD, a GDF9 antagonist....................... 65  Fig. 2.4  GDF9 pretreatment enhanced activin A-induced inhibin B accumulation in culture media .. 66  Fig. 2.5  GDF9 pretreatment for 24 h increased cell response to activin A by regulating mRNA levels of ACVR2B/1B (panels B and C), Smad2/3 (panels D and E), and Smad7 (panels G).............. 67  Fig. 2.6  GDF9 pretreatment for 24 h transiently enhanced phosphorylation of Smad3 (Ser423/425) and Smad2 (Ser465/467) induced by activin A, effects that were neutralized by BMPR2 ECD....... 68  Fig. 3.1  GDF9-targeting siRNA reduced the GDF9 expression in hGL cells...................................... 79  Fig. 3.2  GDF9-targeting siRNA diminished the effects of activin A on levels of inhibin βB-subunit mRNA in hGL cells ................................................................................................................ 81  Fig. 3.3  GDF9-targeting siRNA diminished the effects of activin A on levels of inhibin βB-subunit mRNA in hGL cells, and the effects of siRNA were reversed by exogenous GDF9 treatment ................................................................................................................................................ 82  Fig. 3.4  GDF9-targeting siRNA reduced activin A-induced inhibin B accumulations in culture media ................................................................................................................................................ 83  Fig. 3.5  GDF9-targeting siRNA decreased cell response to activin A by regulating mRNA levels for ACVR2B/1B, Smad2/3, and Smad7 in the activin signaling pathway ................................... 84  Fig. 3.6  GDF9-targeting siRNA decreased phosphorylation of Smad3 (Ser423/425) and Smad2 (Ser465/467) in hGL cells induced by activin A (25 ng/ml)......................................................................... 85  B  B  B  x  Fig. 4.1  Time- (A and B) and concentration- (C and D) dependent effect of GDF9 on FST and FSTL3 mRNA levels in hGL cells and corresponding concentration-response in protein levels (E and F) ............................................................................................................................................. 97  Fig. 4.2  Comparison of FST and FSTL3 protein levels in culture media of hGL cells ....................... 98  Fig. 4.3  GDF9 reversed activin A-induced FST and FSTL3 mRNA (A) and protein (B) levels, effects attenuated by BMPR2 ECD (“ECD”) ..................................................................................... 99  Fig. 4.4  GDF9-targeting siRNA increased mRNA (A) and protein (B) levels of FST and FSTL3, effects that were reversed by adding exogenous GDF9 ................................................................... 100  Fig. 4.5  GDF9-targeting siRNA increased activin A-induced mRNA (A) and protein (B) levels of FST and FSTL3, effects that were reversed by adding exogenous GDF9.................................... 101  Fig. 4.6  GDF9-targeting siRNA transfection attenuated while FST or FSTL3-targeting siRNA transfection enhanced activin A-induced inhibin βB-subunit mRNA levels ......................... 102 B  Fig. 4.7  GDF9 pretreatment enhanced activin A-induced inhibin βB-subunit mRNA levels in hGL cells, effects that were attenuated by FST or FSTL3 in a concentration-dependent manner ......... 104  Fig. 5.1  GDF9 pretreatment reduced activin A suppression of StAR mRNA levels in hGL cells, effects that were attenuated by BMPR2 ECD (“ECD”), a GDF9 antagonist ................................... 115  Fig. 5.2  The effects of GDF9 and activin A on P450scc and 3βHSD mRNA levels in hGL cells .... 116  Fig. 5.3  GDF9 pretreatment reduced activin A suppression of basal and FSH-induced StAR protein levels in hGL cells, effects that were attenuated by BMPR2 ECD....................................... 117  Fig. 5.4  The effects of GDF9 and activin A on P450scc and 3βHSD protein levels in hGL cells..... 118  B  Fig. 5.5 GDF9 pretreatment reduced activin A-suppressed progesterone accumulations in culture media, effects that were attenuated by BMPR2 ECD....................................................................... 119 Fig. 5.6  After transfection of hGL cells with inhibin α-subunit siRNA, accumulations of activin A increased; accumulations of inhibin A and inhibin B decreased following treatment with activin A alone or activin A with GDF9 pretreatment .......................................................... 120  Fig. 5.7  GDF9 reduced the suppressive effects of activin A on StAR expression and progesterone accumulation, effects that were attenuated when cells were transfected with inhibin α-subunit siRNA ................................................................................................................................... 121  Fig. 5.8  The role of inhibin α-subunit siRNA in the interaction of GDF9 and activin A on P450scc and 3βHSD expression and progesterone accumulation in hGL cells ......................................... 122  Fig. 5.9  Transfection of hGL cells with GDF9 siRNA enhanced the suppressive effects of activin A on StAR expression and progesterone accumulation................................................................. 123  Fig. 5.10 The role of GDF9 siRNA in the interaction of GDF9 and activin A on P450scc and 3βHSD expression and progesterone accumulation in hGL cells ...................................................... 124 Fig. 6.1  A model, suggested by the results, for the interactions between GDF9 and activin A in the regulation of inhibin B and progesterone accumulation by human granulosa-lutein cells ... 134  xi  LIST OF ABBREVIATIONS 3βHSD  3β-Hydroxysteroid dehydrogenase/∆5-∆4 isomerases  aa  Amino acid  ACVR  Activin receptor  ALK  Activin receptor-like kinase  AMH  Anti-müllerian hormone  AMHR  AMH receptor  ANOVA  Analysis of variance  ATF2  Activating transcription factor 2  bHLH  Basic helix-loop-helix  BAMBI  BMP and activin membrane-bound inhibitor  BMP  Bone morphogenetic protein  BMPR  BMP receptor  BMPR2 ECD  BMP type II receptor extracellular domain  bp  Base pair  Ca2+  Calcium  CamKII  Ca2+/Calmodulin-dependent protein kinase II  cAMP  Cyclic adenosine monophosphate  cDNA  Complementary deoxyribonucleic acid homolog  CL  Corpus luteum  Co-Smad  Common Smad  COC  Cumulusoocyte complexe  COX2  Cyclo-oxygenase 2, prostaglandin-endoperoxide synthase 2 xii  CR-1  Cripto-1  CRE  cAMP-response elements  CREB  CRE-binding protein  Dab2  Disabled2  dNTP  Deoxynucleoside triphosphate  DMEM  Dulbecco's Modified Eagle's Medium  DNA  Deoxynucleic acid  DNase  Deoxyribonuclease  E2  17β-Estradiol  ECD  Extracellular domain  EDTA  Ethylene diaminetetraacetic acid  EGF  Epidermal growth factor  EGF-CFC  Epidermal growth factor–Cripto/FRL-1/Cryptic  ELISA  Enzyme-lined immunosorbant assay  EMT  Epithelial to mesenchymal transition  ER  Endoplasmic reticulum  Erk  Extracellular signal-regulated kinase  FBS  Fetal bovine serum  FGF  Fibroblast growth factor  FHA  Forkhead-associated  Fig-α  Factor in the germline alpha  FLRG  FST-related gene  FSD  FST domain  xiii  FSH  Follicle stimulating hormone  FSHR  Follicle stimulating hormone receptor  FSTL  FST-like proteins  FST  Follistatin  g  Acceleration of gravity  GAPDH  Glyceraldehyde-3-phosphate dehydrogenase  GCNF  Germ cell nuclear factor  GCT  Granulosa cell tumor  GDF  Growth differentiation factor  GDNF  Glial cell-derived neurotrophic factor  GH  Growth hormone  GIPC  GAIP-interacting protein C terminus  GnRH  Gonadotropin-releasing hormone  GnRH-II  Gonadotropin-releasing hormone type II  GnRHR  Gonadotropin-releasing hormone receptor  GS-domain  Glycine-serine rich domain  h  Hour  HAS2  Hyaluronan synthase 2  hCG  Human chorionic gonadotropin  HECT  Homologous to the E6-accessory protein C-terminus  hGL  Human granulosa-lutein  IGF  Insulin-like growth factor  IGFBP  Insulin-like growth factor binding protein  xiv  IL  Interleukin  I-Smad  Inhibitory Smad  IU  International unit  IVF  In vitro fertilization  JNK  c-Jun N-terminal kinase  Kb  Kilobase  kDa  Kilodaltons  KL  Kit ligand  LH  Luteinizing hormone  LHR  Follicle stimulating hormone receptor  LRH-1  liver receptor homolog-1  μ  Micro  MAP  Mitogen-activated protein  MAPK  MAP kinases  MADH  Mad-homologues  MEKK1  MAPK/Erk kinase kinase 1  MH  Mad-homology  ml  Mililiters  min  Minutes  MMP  Matrix metalloproteinase  mRNA  Messenger ribonucleic acid  MW  Molecular weight  n (as in nM)  Nono  xv  NES  Nuclear export signal  NFκB  Nuclear factor kappa B  NLS  Nuclear localization sequence  OOX  Cumulus-oocyte complexe  P (as in pM)  Pico  P450scc  P450 cholesterol side-chain cleavage enzyme  P450c17α  cytochrome P450 17α-hydroxylase/17,20-lyase  PAGE  Polyacrylamide gel electrophoresis  PBS  Phosphate-buffered solution  PCO  Polycystic ovaries  PCOS  Polycystic ovary syndrome  PCR  Polymerase chain reaction  PGC  Primordial germ cell  PGE2  Prostaglandin E2  PGF2α  Prostaglandin F2α  PKA  Protein kinase A  PKC  Protein kinase C  PI3K  Phosphatidylinositol-3-kinase  PR  Progesterone receptor  RNA  Ribonucleic acid  R-Smad  Receptor-activated Smad  RT-PCR  Reverse transcription polymerase chain reaction  SAD  Smad activation domain  xvi  SAPK  Stress-activated protein kinase  SARA  Smad anchor for receptor activation  SBE  Smad binding element  SCF  Stem cell factor  SD  Standard deviation  SDS  Sodium dodecyl sulphate  Sec  Seconds  Ser  Serine  SF-1  Steroidogenic factor 1  siRNA  Small interfering RNA  Smad  Son of mothers against decapentaplegia  Smurf  Smad ubiquitination-related factor  StAR  Steroidogenic acute regulatory protein  STAT  Signal transducer and activator of transcription  TβR  TGFβ receptor  Taq  Thermus acuaticus, source of a DNA polymerase  TE  Tris-EDTA  TEMED  N,N,N’,N’-tetramethylethlenediamine  TGF  Transforming growth factor  Thr  Threonine  TNFα  Tumor necrosis factor α  Tris  Tris(hydroxyl methyl) aminomethane  YY1  Transcription factor Yin Yang 1  xvii  uPA  Urokinase plasminogen activator  VEGF  Vascular endothelial growth factor  xviii  ACKNOWLEDGEMENTS This thesis was conducted at the Department of Obstetrics and Gynecology in the Child and Family Research Institute and in the Program of Reproductive and Developmental Sciences at the University of British Columbia from 2004 to 2010. First, I would like to express my deepest gratitude to my supervisor, Dr. Peter C.K. Leung, for his continuous support and great supervision throughout my PhD study. His advice, guidance and encouragement have proven to be valuable to my research and future career pursuits. Second, I need to thank to my co-mentor of IWRH program, Dr. Anthony P. Cheung, who has always offered the most valuable comments on the most critical issues of my research. I appreciate his professional advice and support throughout these studies. I would also like to thank other members of my supervisory committee, Dr. Young.S. Moon, Dr. Rajadurai Rajamahendran and Dr. Colleen C. Nelson, for providing invaluable guidance and suggestions on this thesis. I wish to express my appreciation to Dr. Christian Klausen for his valuable advice, comments, friendship and encouragement. My warm thanks also go to Dr. Aaron J.W. Hsueh of Stanford University who has given friendly comments and constructive criticism on my work and has also provided fascinating ideas. I would like to take this opportunity to express my sincere gratitude to my past and present colleagues from the lab and friends for their constant support, friendship and for sharing all of the good and the bad times during this project. I am grateful to Ms. Roshni Nair for her excellent administrative and secretarial help. I would also like to thank the Canadian Institutes of Health Research, the Child and Family Research Institute and the Interdisciplinary Women’s Reproductive Health Research Training Program for supporting this research. I gratefully acknowledge the staff of the Women’s Health Center for their time and energy in supplying the cell samples used in this thesis. I am also very grateful to my mother, father and younger brother for their understanding, continuous support and encouragement. In the past six years, no matter what difficulty I encountered, a great woman was always there to provide endless love and support without complaint. I hereby dedicate this thesis and express the deep gratitude to you, my dear wife. You mean the world to me and my heart belongs to you. xix  1. LITERATURE REVIEW 1.1.  Introduction In the past, an incredible amount of effort has been made to understand the development of  ovary, a part of the reproductive system of all female mammals. Ovary is not essential to individual life but very important to the perpetuation of species. The major function of the ovary is to produce a well-differentiated and fertilizable oocyte for successful propagation. In addition, the ovary secretes steroid hormones into the reproductive tract and establishes a supportive environment for fertilization and pregnancy to occur. Both activities are included in the continuous cycling through follicle growth, ovulation and corpus luteum formation and degeneration. The ovary is consisted of numerous follicles which are at the different developmental stages. An oocyte is centered in the follicle and enclosed by inner layers of granulosa cells and outer layers of theca cells (1). Numerous neural, neuroendocrine, paracrine and endocrine cell-cell communication pathways are involved in a complicated system that controls follicular development. The roles of the gonadotropin-releasing hormone (GnRH), luteinizing hormone (LH), follicle stimulating hormone (FSH) and gonadal steroids such as progesterone and estrogen are well-known, but there is increasing evidence that various cytokines and growth factors are also involved in this process (review in (1)). These include activin, inhibin, growth differentiation factor (GDF) 9 and bone morphogenetic protein (BMP) 15, which are to the members of the transforming growth factor β (TGFβ) superfamily, and follistatin (FST), which is a structurally distinct molecule that is functionally related to the members of the TGFβ superfamily (Fig. 1.1) (2-4). At the different stages of folliculogenesis, not only many other molecules regulate the 1  signaling of these molecules but also they can interact with each other. The absence or genetic alterations of these factors have been found to have profound effects on fertility (4). The first part of this thesis, the Literature Review section, focuses on the TGFβ signaling pathway molecules and signaling mechanisms. The courses of ovarian folliculogenesis and steroidogenesis are briefly reviewed, and the biological roles of activins/inhibins, follistatin, follistatin-like (FSTL) 3 and the oocyte-secreted factors (GDF9 and BMP15) are described. The second part, including four chapters of research results, focuses on the interactions between activin A and GDF9 on inhibin B and progesterone accumulation in human granulosa-lutein (hGL) cells. The last section discusses the outcome of the study and offers some suggestions for further study.  1.2.  Folliculogenesis The primary function of the follicle is to support the oocyte. In human beings, the ovaries  contain about one million immature, primordial follicles, which contain an oocyte suspended in prophase of meiosis surrounded by one layer of pregranulosa cells that is enveloped by a basement membrane. After the first menstruation in puberty, a group of follicles begin folliculogenesis, initiating a growth pattern that ends in atresia or ovulation. Folliculogenesis is a long process and only occurs during the reproductive lifespan. It takes about 375 days for a primordial follicle to develop and grow to the ovulatory stage, which includes two periods. The first period is called preantral stage, where the follicle grows and differentiates independently from FSH and LH but is driven by growth factors. In the second period, the antral stage, a few of the follicles stimulated by gonadotropins and growth factors grow large and become dominant antral follicles, finally reaching the preovulatory stage, while most other follicles  2  undergo atresia. During each reproductive cycle, the dominant antral follicle enlarges due to the effects of preovulatory FSH and LH, surges and ruptures, releasing the mature oocyte for fertilization, while the corpus luteum is established by the remaining granulosa and theca cells.  1.3.  Steroidogenesis in Ovary  1.3.1. Overview of steroidogenesis Steroidogenesis involves the synthesis of steroids from cholesterol and their subsequent transformation into alternative types of steroids. It is the basis of essential physiologic functions such as reproduction, metabolism and the immune response. The steroid hormones are generated from subtle modifications in the four-fused ring of the sterol skeleton (three 6-carbon rings and one 5-carbon ring) and the side chain. Cholesterol is the basic building block of steroidogenesis. During steroidogenesis, the number of carbon atoms in cholesterol or any other steroid molecule can be reduced but never increased. Steroid production starts with the transformation of cholesterol (27 carbons) to pregnenolone (21 carbons) by the cytochrome P450 cholesterol side-chain cleavage (P450scc) enzyme (5). The transformation of pregnenolone to progesterone (21 carbons) is catalyzed by 3β-hydroxysteroid dehydrogenase (3βHSD) (6). In some tissues, progesterone is hydroxylated, and the side chain is subsequently cleaved by cytochrome P450 17α-hydroxylase/17, 20-lyase (P450c17) to form androstenedione (19 carbons). Androstenedione can be reduced to testosterone (19 carbons) by 17β-hydroxysteroid dehydrogenase (17βHSD). Both androstenedione and testosterone are converted to estrogens (estrone and estradiol, all 18 carbons) by an aromatization reaction that is mediated by cytochrome P450 aromatase (P450arom) in the endoplasmic reticulum (7). The steroidogenic acute regulatory protein (StAR) principally regulated the rate-limiting 3  step of steroid production, that means the translocation of cholesterol from the outer to the inner mitochondrial membrane where fully active cytochrome P450scc locates (8). In the ovary, StAR expression is primarily present in steroid-producing cells, including the luteal and theca cells. The StAR mRNA and protein is rapidly stimulated by hormones and catalyzes the intermembrane transfer of cholesterol to P450scc to initiate steroidogenesis. In the ovary, granulosa cells are the cellular source of estradiol and progesterone, which are the most important steroids involved in ovarian function. Progesterone production only needs granulosa cells, but estrogen synthesis requires collaboration between granulosa and theca cells. The cooperation of these two cell types and of the two gonadotropins (FSH and LH) in estrogen synthesis can be logically explained by the concept of the two cell/two gonadotropin system, first proposed by Falck in 1959 (9). In human preantral and antral follicles, the LH receptor (LHR) is only expressed on the theca cells and the FSH receptor (FSHR) only on the granulosa cells. Androgens produced by LH-stimulated theca cells are the main substrates for estrogen synthesis by FSH-stimulated granulosa cells. FSH increases granulosa cell aromatase activity, which is absent in theca cells. The interaction between the granulosa and theca compartments causes accelerated estrogen production. At the later follicular phase, the LH surge that occurs during ovulation causes a remarkable stimulation of progesterone production and a sharp decline in estradiol production by the remaining luteinizing granulosa and theca cells (10). Together with terminal differentiation and luteinization, progesterone production of granulosa cells depends on an increase in the number of LHR just before the LH surge. The dramatic rise in LH leads to StAR expression in granulosa cells, thereby endowing granulosa cells with the ability to make progesterone from cholesterol (11).  4  After ovulation, it is believed the two cell types continue to function as a two-cell system; theca-lutein cells produce androgens for aromatization into estrogens by granulosa-lutein cells. The newly formed corpus luteum produces progesterone predominantly to prepare the uterine endometrium for implantation and to maintain gestation. Luteal cells express high levels of StAR, P450scc and 3βHSD to produce large amounts of progesterone (11). If implantation occurs, the lifespan of the corpus luteum can be prolonged by the emergence of a new, rapidly propagated hormone, human chorionic gonadotropin (hCG), which maintains luteal function until placental steroidogenesis is well established. Otherwise, the corpus luteum would regress, allowing for another reproductive cycle. The demise of the corpus luteum may involve luteolytic activity against its own estrogen production, which is regulated by prostaglandin and endothelin-1.  1.3.2. Regulation of the StAR gene Many agents have been shown to be able to modulate StAR expression positively or/and negatively by acting on its promoter. Studies have been performed to investigate which transcription factors and other proteins interact with the regulatory elements in the StAR promoter. Intracellular cyclic adenosine monophosphate (cAMP) is thought to be a positive modulator that can cause a rapid increase in StAR expression (12). LH, hCG, insulin and growth hormone (GH) stimulate StAR transcription in the theca, luteinized granulosa and luteal cells. Several locally produced growth factors, such as insulin, insulin-like growth factor (IGF)-I and IGF-II, also directly induce or modify gonadotropin-stimulated steroidogenesis by regulating the mRNA expression of StAR. In pig granulosa cells, IGF-I induces StAR mRNA  5  production and enhances the effects of FSH, LH and cAMP in steroid production (13-15). Low concentrations (10 ng/ml) of leptin stimulate, while high concentrations (1000 ng/ml) reduce StAR mRNA in porcine granulosa cells (16). In luteal cells, StAR production is stimulated by the luteotroph prostaglandin E2 (PGE2) (17). Various peptides secreted by the oocyte, granulosa or theca cells suppress the basal or gonadotropin-stimulated  StAR  expression  and  ovarian  steroidogenesis  through  autocrine/paracrine pathways. These negative regulators include TGFβ in the theca cells (18), leptin (16), TGFβ (19), BMP2 (20), BMP4 (20, 21), BMP5 (22), BMP6 (20), BMP7 (20), BMP15 (23) and epidermal growth factor (EGF) (15) in the granulosa cells, tumor necrosis factor α (TNFα) (24) in the luteal cells and cholesterol sulfate in a granulosa tumor cell line (25). Oocyte-secreted GDF9 modulates StAR expression in a species-specific manner, increasing basal StAR expression in cultured mice granulosa cells but not human granulosa-lutein cells (26). GDF9 further reduces cAMP-induced StAR expression in human granulosa-lutein cells (26). This protein also inhibits basal StAR protein levels and steroidogenesis in human theca cells. StAR promoter activation can be regulated by numerous transcription factors, including steroid factor-1 (SF-1), GATA4, GATA6, C/EBP and CREB. SF-1 controls the basal and cAMP-induced transcription of the human StAR (27). GATA4 and GATA6 play important roles in StAR transcription. Overexpression of both GATA4 and GATA6 increase the activity of StAR promoter in porcine granulosa cells (28). The StAR promoter is stimulated by the concerted effects of C/EBP and GATA4 in porcine granulosa cells (29). Basal and cAMP-induced murine StAR promoter activation can be increased by overexpression of CREB  6  in MA-10 and Y1 cells (30).  1.3.3. Regulation of the P450scc gene The level of P450scc is mainly regulated by two pituitary-secreted gonadotropins, LH and FSH, in the granulosa, theca and luteal cells. The regulation of P450scc in folliculogenesis has similarities with that of StAR. Compared with StAR, P450scc mRNA expression is first detected in the granulosa cells of the preovulatory follicles of most species and it increases following follicular development (31-33). Similar to StAR, P450scc levels are stimulated by LH in the theca cells of developing follicles (34). Before ovulation, P450scc expression is elevated by gonadotropin surges and is retained at high levels in luteal cells (33). Consistent with the requirement of LH to induce luteal P450scc mRNA expression in vivo, P450scc is induced by FSH, LH and hCG in cultured rat, porcine and human granulosa cells (33, 35-38). The continued expression of P450scc during the luteal phase is species-specific: it either becomes constitutive, as observed in rats, or remains reliant on LH, as observed in primates (33, 39). In bovine luteal cells, P450scc mRNA accumulation is also increased by estradiol (E2) (17). The corpus luteum rescued by hCG in humans (40) or prolactin in rodents (41) during gestation keeps the expression of P450scc continuing. The stimulating effects of gonadotropin on P450scc expression can be further enhanced by growth factors, such as IGF-I (42) in rats, IGF-II (43) in humans and EGF (15) in porcine granulosa cells. P450scc expression can also be increased by estradiol (33), progesterone (44), GH (45), insulin (37, 46) in granulosa cells, amphiregulin (47) in cumulus-oocyte complexes and insulin (46) in theca cells. The effect of TGFβ on mRNA expression is species dependent: it induces P450scc protein synthesis in rat granulosa cells (48) but suppresses P450scc mRNA in cultured bovine  7  granulosa cells (19). Several BMP members, such as BMP4, BMP6 and BMP15, inhibit FSH-induced P450scc mRNA or protein levels in granulosa cells by impairing FSH signaling. BMP4/6/7 also reduce basal P450scc mRNA in theca cells (49). Other negative regulators of P450scc expression include PGF2α (50) and TNFα (51) in granulosa cells and PGF2α (52) in luteal cells. P450scc promoter activation involves the same transcription factors as StAR, but also includes its own regulatory factors. FSH or forskolin stimulates P450scc gene transcription via transcription factors, such as SF-1 (53) and GATA4 (54) in rat granulosa cells. P450scc expression in human granulosa cells is regulated by liver receptor homolog-1 (LRH-1) (55). The transcription of P450scc can be regulated by the activating protein 2 (AP-2) in the human placenta (56). The activation of the human P450scc promoter, which is induced by LRH-1, is inhibited remarkably by the nuclear receptor DAX-1 in human granulosa cells (55). Unlike StAR, the roles of C/EBP and NR4A members in the regulation of P450scc transcription remain unclear.  1.3.4. Regulation of 3βHSD genes There are two isoforms of the 3βHSD gene in humans: type I (3βHSD1) and type II (3βHSD2). 3βHSD2 is expressed in the adrenal gland, ovary and testis whereas 3βHSD1 is expressed in other tissues, such as the prostate, mammary gland and placenta. The expression of 3βHSD is stimulated by the FSH and LH surges that occur before ovulation and 3βHSD is expressed in porcine, bovine and sheep luteal cells (57, 58). In vitro, FSH, LH and hCG increase 3βHSD expression in the hypophysectomized rat ovary and monkey corpus luteum (39, 59). Growth factors, such as IGF-I and insulin, also induce 3βHSD  8  expression in granulosa cells, but their enhanced effects on FSH-stimulated 3βHSD mRNA are species-specific: IGF-I cannot work synergistically with FSH to increase 3βHSD mRNA expression in rat granulosa cells (42), while insulin is able to cooperate with FSH to augment its mRNA levels in human granulosa cells (60). IGF-I can also increase 3βHSD mRNA levels in cultured theca cells from hypophysectomized immature rats (61). EGF enhances the action of FSH on 3βHSD expression, and this interaction is further augmented by TGFβ in cultured granulosa cells isolated from the ovaries of immature rats primed with diethylstilbestrol (62). cAMP-stimulated 3βHSD activity is enhanced by fibroblast growth factor (FGF) in cultured human theca and granulosa-lutein cells (63). Similar to P450scc, BMP2 (64), BMP4 (21), BMP6 (64) and BMP15 (23) can impair the FSH signaling pathway, thereby decreasing the FSH-augmented 3βHSD expression in granulosa cells. BMP4, BMP6 and BMP7 also reduce basal 3βHSD mRNA in bovine theca cells (49). TGFβ negatively affects 3βHSD mRNA in cultured bovine granulosa cells (19). Other inhibitors of 3βHSD expression include PGF2α in cultured porcine granulosa and luteal cells (50) and prolactin in the rat ovary (59, 65). Like StAR and P450scc, human 3βHSD promoter activity can be driven by SF-1 or LRH-1 and inhibited by DAX-1 in human granulosa tumor cells (66); additionally, the co-expression of GATA4 or GATA6 can interact with SF-1 and LRH-1 to synergistically activate the promoter of human 3βHSD in both heterologous cells (CV-1 fibroblasts) and steroidogenic cells (MA-10 and H295R) (67). One difference of 3βHSD as compared to StAR and P450scc in granulosa cells is the reaction of the promoter to Nur77, which also regulates StAR and P450scc. Exogenous expression of Nur77 augments the 3βHSD promoter in human granulosa tumor cells but is not synergistic with SF-1 (68). Another difference is that FXR in human  9  adrenocortical cells (69) and Stat5 in Hela cells (70) modulate 3βHSD transcription. For more details about the regulation of StAR, P450scc and 3βHSD genes, please refer to review (71).  1.4.  Overview of TGFβ Superfamily  1.4.1. Ligands The first member of the TGFβ superfamily, TGFβ1, was discovered in 1981 (72). Today, there are more than 45 members found in this superfamily, which includes the TGFβ subfamily (comprising TGFβ1, TGFβ2, TGFβ3), the activin/inhibin subfamily (including activin A, AB, B, C, D, E and inhibin A, B), the GDF subfamily (with at least 9 members), the BMP subfamily (with more than 20 members), the glial cell-derived neurotrophic factor (GDNF) subfamily (including GDNF, artemin and neuturin) and some other members, such as nodal, lefty and anti-Müllerian hormone (AMH; also known as Müllerian-inhibiting substance, MIS) (1). These peptides are multifunctional growth factors that control proliferation, differentiation, motility, migration, apoptosis, adhesion, matrix synthesis and other functions in many cell types. They express in complicated temporal and tissue-specific patterns and consequently play an important role in the development of almost all tissues in many organisms, from insect to humans. Collectively, these factors account for a substantial portion of the intercellular signaling pathway controlling cell fate. There are several distinctive features shared by the TGFβ superfamily members. Firstly, the pre-proprotein of the TGFβ superfamily members consist of a short N-terminal signal peptide, a propeptide that consists of 200-300 amino acids (aa) pro-region and a C-terminal 10  mature region (Fig. 1.2A) (73). The signal peptide directs the protein to the endoplasmic reticulum (ER) and through the secretory pathway. The variations in the pro-regions may relate to the folding and dimerization of proteins and ultimately determine the different structures of different TGFβ superfamily members (74). Six to nine evolutionarily conserved cysteine residues are found in the mature region of the C-terminus. Six cysteines engage in an intramolecular cystine knot configuration that consists of three covalent disulfide bonds (Fig. 1.2B) (75). In addition, this knot fastens the base of several β-sheets together and it seems to help to drive molecular dimerization and increase the stability of the TGFβ superfamily members with butterfly-shaped structures. Most proteins from the TGFβ superfamily are only functional as disulfide bond-linked homodimers or heterodimers, consisting of two polypeptide chains, known as monomers. The seventh conserved cysteine residue (the fourth consecutive cysteine from the N-terminus) present in most TGFβ superfamily members is involved in forming this disulfide bond (76). Nevertheless, some superfamily members, such as GDF3, GDF9 and BMP15, cannot form covalent dimers due to a lack of this cysteine residue (77-81). Instead, these ligand monomers may be linked by noncovalent bonds.  1.4.2. The signaling pathway Except inhibin and the GDNF subfamily, most ligands of the TGFβ superfamily firstly bind to the transmembrane components of both type I and type II receptor and then induce the formation of a ligand-receptors complex. This complex initiates the autophosphorylation of the type I receptor and then actives the second message system by phosphorylating cytoplasmic Smads (Sons of mothers against decapentaplegia), which go to the nucleus and bind to the  11  specific DNA sequence to modulate the transcription of the targeted gene.  1.4.2.1. Signaling receptors Based on their structures and functions, the receptors of the TGFβ superfamily can be separated into two subfamilies: type I and type II receptors. There are five type II receptors (TβR2, ACVR2A, ACVR2B, BMPR2 and AMHR2) and seven type I receptors [activin receptor-like kinase (ALK) 1-7] recognized in mammals (see Table 1.1 and Table 1.2) (82, 83). Proper signaling of the TGFβ superfamily ligands requires the involvement of both type I and type II receptor. These receptors are similar in structure, but their kinase domains are different in function. The receptor peptide includes an N-terminal extracellular domain, a transmembrane domain and a C-terminal intracellular serine/threonine (Ser/Thr) kinase domain.  1.4.2.2. The interaction between ligand and receptors The activative peptides of the TGFβ superfamily are homo- or heterodimers. Their structures suggest that they function to bring the dimers of the type I and type II receptors together, forming heterotetrameric receptor complexes. The TGFβ superfamily members bind differently to their receptors, and there are two binding patterns that have been observed for these interactions. TGFβ and activin do not interact with the isolated type I receptors and have high affinities for type II receptors (84). In contrast, some BMP ligands, such as BMP2 and BMP7, bind first to type I receptors, then recruit the type II receptor to the complex of ligand and type I receptor (84). This ligand-type I-type II receptor combination causes the type II receptor to phosphorylate specific serine residues in the 30-amino glycine-serine-rich domain (GS-domain) located upstream of the type I receptor kinase domain. The phosphorylated type I 12  receptor then further activates a second messenger system, Smads, by phosphorylation, to regulate the downstream signaling pathway (Fig. 1.3).  1.4.2.3. The Smads family Smads were initially recognized as the products of the Drosophila Mad and C. elegans Sma genes, and their name is a result of the combination of the Sma and Mad genes (85, 86). In vertebrates, eight Smad proteins are identified: Smad 1 through 8. They are universally expressed in all adult tissues and play diverse roles in intracellular signaling pathways (87, 88). Based on the structural and functional features, there are three groups in the Smads family (Fig. 1.4): receptor-activated Smads (R-Smads: Smad1, Smad2, Smad3, Smad5 and Smad8), which can be phosphorylated by the type I receptors; the common Smad (Co-Smad: Smad4), which can oligomerize with the dimer of phosphorylated R-Smads; and inhibitory Smads (I-Smads: Smad6 and Smad7), which can be stimulated by the TGFβ superfamily ligands. I-Smads can compete with R-Smad receptor binding or co-operate with Smurfs (Smad ubiquitination-related factor) to stimulate receptor degradation via ubiquitination and consequently inhibit the signaling pathway (Fig. 1.3). In addition, Smurfs can mediate R-Smads degradation through the ubiquitination. The differences in the structures of the ligands lead to diverse combinations of type I and type II receptors and the activation of various R-Smads. Phosphorylated TGFβ/activin/ nodal/myostatin (also called GDF8) type I receptors (ALK4, ALK5 and ALK7) can activate Smad2 and Smad3, while BMP type I receptors (ALK2, ALK3 and ALK6) can activate Smad1, Smad5 and Smad8 (Table 1.3). Phosphorylated R-Smads combine with Smad4 before entering the nucleus and at there, they cooperate with diverse DNA-binding proteins, transcription  13  activators and transcription repressors, such as FoxH1 (89), AML (90) and p300/CBP (91) and bind to specific DNA binding sequences to modulate target gene transcription in a cell type-specific manner. Different transcription factors cooperate with Smads, and cross-talking with other signaling pathways can cause various responses to ligands in different cells. Smad6 inhibits the TGFβ superfamily signaling by competing with phosphorylated R-Smads to bind to Smad4. In contrast, Smad7 can bind to the type I receptors to block the signaling. Smad6 can specifically block the BMP signaling pathway, while Smad7 can disrupt both the TGFβ/activin and BMP signaling pathways by preventing R-Smads from phosphorylation (review in (92)). In the unstimulated cells, R-Smads remain in the cytoplasm, mainly as monomers that do not form complexes with Smad4, which shuttles between the cytoplasm and nucleus, while I-Smads are localized to the nucleus. Cytoskeletal proteins such as filamin also play roles in the localization and signaling of Smads (93).  1.4.2.3.1. The conserved structure of Smad family There are two kinds of conserved domains in the Smad family: the N-terminal MH1 (Mad homology 1) domain and the C-terminal MH2 (Mad homology 2) domain. In contrast with MH1, the similarity of MH2 domain sequence in all Smads is high. The linker segment between these two domains is variable in sequence and length (Fig. 1.4 and Table 1.4) (review in (94)). Smad4 and R-Smads have both MH1 and MH2 domains. I-Smads only contain the MH2 domain, but N-terminal sequences of them are mildly similar to the MH1 domain. The MH1 domains of Smad4 and R-Smads are able to bind to a specific DNA sequence (Smad binding element, SBE) of the target gene. However, Smad2 has an extra residue in exon 3, leading to a loss in the ability to bind to the DNA sequence (Fig. 1.4) (95). MH1 and MH2 domains can co-operate with the transcription factors, transcription co-activators and 14  co-repressors to modulate specific gene transcription (Table 1.4). The linker region is a multifunctional region, which includes the sites for regulatory phosphorylation, cytoplasmic adaptors (filamin), ubiquitination adaptors (Smurf1 and Smurf2) and other transcription regulation proteins (Table 1.4).  1.4.2.3.2. The phosphorylation of Smad family Activation of the Smad signaling pathway is achieved through the phosphorylation of its SSXS motif in the C-terminus by phosphorylated type I receptors. The structures of L3 loop in the MH2 domain and the type I receptor of R-Smads determine the specificity of the receptor substrate, R-Smads (Fig. 1.4) (94). TGFβ/activin/nodal/GDF8 signaling activates Smad2 and Smad3 by phosphorylation, while BMPs/GDFs signaling activates Smad1, Smad5 and Smad8. In addition, other signaling pathways such as Ca2+/Calmodulin-dependent protein kinase II (CamKII) (96), extracellular signal-regulated kinase (Erk)-family MAP (mitogen-activated protein) kinases (97) and protein kinase C (PKC) (98) can interact with the Smad signaling pathway and then phosphorylate Smads (review in (94)).  1.4.2.3.3. Smad-independent signaling pathway In addition to the general Smad-dependent signaling pathway, the TGFβ superfamily ligand-induced cellular signaling can be modulated by Smad crosstalk with other signaling pathways, adding further complexity to TGFβ superfamily signaling. The crosstalk can be achieved through other signal molecules that directly alter the function of the Smad by alternative phosphorylation, or the Smad can modify the action of other signal molecules. The interaction between these signaling pathways enables the reduction or magnification of other growth factor effects. 15  MAPK/Erk kinase kinase 1 (MEKK1) activation can phosphorylate the MAP kinase sites in the Smad2 linker region and hence stimulate the interaction of TGFβ-induced Smad2 with Smad4 and the nuclear translocation and transcription of Smad2 in cultured bovine aortic endothelial cells (99). The C-terminal SSXS motif of Smad2 is not required in this step, although it is the site of TGFβ type I receptor-mediated phosphorylation (99). In addition, the overexpression of Smad7 can attenuate the MEKK1-stimulated transcriptional activity of Smad2 (99). The MKK3/p38 pathway, activated by TGFβ, can phosphorylate the activating transcription factor 2 (ATF2, also called CRE-BP1), enhancing ATF2-Smad4 complex formation to regulate target gene transcription (100). PKC directly phosphorylates Smad3, abrogating its DNA binding ability, causing the disruption of the transcriptional regulation and finally leading to the down-regulation of TGFβ-induced growth inhibition and apoptosis (98). CamKII activation induces Smad2 and Smad3 phosphorylation to a lesser extent, blocking the import of Smad2 into the nucleus, thereby stimulating the heteromerization of Smad2-Smad4 and inhibiting TGFβ-induced Smad2-dependent transcription (96). TGFβ can induce the process of epithelial to mesenchymal transition (EMT), although it can also inhibit tumorigenesis. The Rho GTPase signaling pathways play an important role in this process. In epithelial cells, TGFβ induces stress fiber formation and mesenchymal characteristics by rapidly activating the RhoA-dependent signaling pathways, while the expression of the dominant-negative mutants blocks RhoA signaling, inhibiting TGFβ-induced EMT (101). The studies has shown that the long-term treatment of human prostate carcinoma cells with TGFβ can cause the assembly of stress fibers via a collaboration between the Smad and Rho GTPase signaling pathways (102). The PI3K/Akt pathway is another Smad-independent pathway. PI3K involves in  16  TGFβ-regulated the proliferation of fibroblast. This TGFβ-induced response occurs through the p21-activated kinase-2 kinase, but it is not Smad2 or Smad3 dependent (103). In liver cells, the PI3K/Akt pathway activated by insulin interferes with TGFβ-induced apoptosis by blocking TGFβ-induced caspase-3 activity rather than through the suppression of the heteromerization of Smad2-Smad4 or nuclear translocation (104). Compared to the phosphorylation of the receptor-regulated Smads, the phosphorylation of Smad7, which is independent of TGFβ stimulation, cannot affect TGFβ signaling while regulating the transcription caused by TGFβ-independent Smad7 activation (105).  1.4.2.4. Extracellular regulation of TGFβ superfamily signaling Several cell membrane-bound proteins, such as betaglycan, cripto, endoglin and BAMBI (BMP and activin membrane-bound inhibitor), can interact with the TGFβ superfamily ligands and facilitate ligand-receptor binding (see Table 1.5). Betaglycan is reported to be the first such molecule and is identified as a TGFβ type III receptor (106), while it is now also recognized to be a co-receptor for inhibin signaling (107). Betaglycan is a membrane-anchored, glycosylated, transmembrane protein that can regulate the signaling pathway of TGFβ, inhibin and activin. It can restore the autocrine function of TGFβ through facilitating TGFβ binding to its receptors in human breast cancer cells (108). In contrast, betaglycan inhibits the tumor-promoting activity of TGFβ and therefore suppresses the malignant properties of human carcinoma cells (109). Betaglycan also inhibits activin binding to its own receptors (ACVRs) by promoting betaglycan/inhibin/ACVRs complex formation in human embryonic kidney (HEK) 293 cells (107). In addition, betaglycan enables inhibin to compete with BMPs in the human liver carcinoma cell line, HepG2, for binding to  17  the BMP type II receptor (BMPR2), which does not bind inhibin in the absence of betaglycan, and thus blocks BMP signaling (110). In human granulosa-lutein cells, the expression of betaglycan mRNA is stimulated by FSH, LH and PGE2, probably through the PKA pathway (111). Human cripto (CR-1) is a member of EGF-Cripto/FRL-1/Cryptic (EGF-CFC) family which has only been identified in vertebrates, and the main members of this family include cripto in humans, cryptic in mice and FRL1 in Xenopus (112, 113). The study already shows that CR-1 is an important signaling co-receptor for nodal during early embryonic development (114). CR-1 interacts with ALK4 via its CFC motif and nodal via its EGF-like motif to form a complex consisting of ACVR2B-ALK4-nodal-CR-1, which allows nodal to induce the activation of Smad2 by phosphorylation (115). In contrast to its enhancing effect on nodal, CR-1 can block activin signaling by combining with activin and ACVR2A/2B, thereby inhibiting ALK4 phosphorylation and activation (116). CR-1 is also found to be critical in cancers and can stimulate the growth and spread of tumors. The stable overexpression of human CR-1 in the mammary glands of mice can induce the formation of mammary hyperplasias and papillary adenocarcinomas (117). Endoglin is a membrane glycoprotein primarily expressed in human vascular endothelial cells and is shown to bind TGFβ by forming a stable ligand-receptor complex (118). Endoglin cannot bind ligands on its own but facilitates TGFβ1 and TGFβ3 binding to their receptors through an association with TβR2 (119). Endoglin specifically enhances the Smad 1/5/8 phosphorylation induced by TGFβ1 and suppresses the migration of endothelial cells dependent on the crosstalk between endoglin and the scaffolding protein GIPC (GAIP-interacting protein C terminus) (120). Endoglin shares corresponding regions of  18  sequence identity (71% amino acid sequence similarly with 63% identity) with betaglycan (121). The complex formation between endoglin and betaglycan may be a important modulator in TGFβ signaling regulation in chondrocytes (122). With the exception of TGFβ itself, endoglin can bind members of the TGFβ superfamily, such as activin A, BMP2 and BMP7 (123). BAMBI is a transmembrane glycoprotein that is evolutionarily conserved in vertebrates and close to the TGFβ superfamily type I receptors in the extracellular domain (124). It can act as a general antagonist of TGFβ/BMP/activin signaling. BAMBI is a pseudo receptor and has a shorter intracellular domain that does not encode the Ser/Thr kinase domain required for signaling. It can form a constant association with type I and type II receptors independent of the ligand to block TGFβ superfamily ligand signaling (124). In addition, BAMBI cooperates with Smad7 to form a ternary association with ALK5 to inhibit the interaction of ALK5 with Smad3, thus impairing Smad3 activation and consequently blocking TGFβ signaling (125). BAMBI transcription can be activated by BMP4 (126) and Wnt/β-catenin signaling (127). TGFβ signaling also can directly increase BAMBI expression through the three tandem repeats of 13 bp sequences containing SBE in the BAMBI promoter (128). Other proteins, including follistatin (FST), follistatin-like 3 (FSTL3), inhibin, latency-associated protein (LAP), noggin, chordin and the related factors caronte, cerberus and gremlin, are recognized to antagonize TGFβ superfamily signaling by binding the TGFβ superfamily ligands and preventing them from binding to the receptors. FST, a glycosylated single-chain protein, is functionally linked to but structurally different from ligands of the TGFβ superfamily. It is characterized by its inhibition of pituitary cells FSH production (129). FST has been shown to be able to bind with activin and prevent activin  19  binding to its receptor, neutralizing most but not all of the actions of activin in the rat ovary (130). In addition, FST also binds and regulates the actions of other ligands of the TGFβ superfamily, such as BMP2, BMP4, BMP7 and BMP15, by forming a trimeric complex (131). FSTL3 is also a member of the follistatin-related protein family and contains a highly conserved follistatin domain which is cysteine-rich. Like FST, FSTL3 binds activin with high affinity and prevents activin from binding to its receptors, neutralizing its biological activities (132). Inhibin is a member of the TGFβ superfamily and a naturally occurring antagonist of activin. Inhibin antagonizes activin action in Chinese hamster ovary cells by combining with the activin type II receptors ACVR2A and ACVR2B, thereby blocking activin binding (133). Inhibin has also been shown to suppress cellular responses to various BMP family members in several BMP-responsive cell types (110). Mature TGFβ is proteolytically derived from the C terminus of its propeptide. Unlike most other hormones, the mature TGFβ remains associated with its propeptide after secretion. LAP, the N-terminal remnant of the TGFβ prepeptide, is able to interact with all isoforms of TGFβ (TGFβ1, TGFβ2, and TGFβ3), binding and neutralizing their activities (134). Noggin and chordin are secreted proteins expressed in the Spermann's organizer (SO) that induce the ventral mesoderm to become the lateral mesoderm. Noggin is a small glycoprotein while Chordin is a large protein. Although Noggin and Chordin do not have structural similarities, neither of them can bind to activin and TGFβ, but they can specifically bind to BMPs, such as BMP2, BMP4 and BMP7, suppressing BMPs signaling by antagonizing BMPs interactions with their receptors (135-138).  20  1.4.2.5. Intracellular regulation of TGFβ superfamily signaling Smads are the main intracellular regulators of TGFβ superfamily signaling, and they are regulated by many factors, such as their access to receptors, protein concentrations, phosphorylation by activated type I receptors, formation of receptors-Smads complexes and nuclear accumulation. The binding of R-Smads to activated TGFβ superfamily type I receptors can be assisted by the SARA protein (139). Smad2 and Smad3 can bind to its centurial FYVE domain and adjacent Smad binding domain, while Smad1 cannot. Overexpression of SARA in COS-1 cells leads to the clustering of Smad2/3 and increases Smad2/3 phosphorylation as mediated by the activated type I receptor (139). The regulation of SARA protein expression or interactions with Smad2/3 could regulate the signaling of the TGFβ superfamily ligands. The loss of the TGFβ response by the interruption of SARA function could cause tumor formation. The intracellular concentration of Smads can be modulated by Smurf1/2 through protein ubiquitination and consequent proteasomal degradation. Smurf1/2 are E3 ubiquitin ligases that can promote the degradation of R-Smads by binding to and blocking TGFβ superfamily signaling (140, 141). In addition to Smurfs, other ubiquitin ligases also play a role in the degradation of Smads (142). The activated type I receptor phosphorylating R-Smads is a vital process in TGFβ superfamily signaling. After activation, phosphorylated R-Smad forms an oligomeric complex with Smad4. Except for type I receptors, which are activated by TGFβ superfamily ligands, R-Smads can also be phosphorylated by other signaling molecules as mentioned above, such as MAPK, PKC and CamKII. In contrast, Smad-mediated signaling can be blocked by specific phosphatases through dephosphorylating TGFβ-activated Smad2 and Smad3 and the 21  promoting their nuclear export (143). Therefore, Smads activity is not only induced but also regulated by phosphorylation, providing mechanisms for the interaction of Smad signaling with other signaling pathways. The combining of receptors and Smads can be interrupted by some mechanisms, and inhibitory Smads and Smurfs are involved in these processes (review in (92)). The combination of Smad7 with Smurf1/2 can bind to the TGFβ type I receptors, resulting in accelerated receptor turnover through the polyubiquitination pathway (144). In contrast with the R-Smads, the expressions of both Smad6 and Smad7 are induced by TGFβ, activin and BMPs and then are used as negative feedback to block further signaling (145, 146). Smad7 expression can also be stimulated by the pathways that inhibit TGFβ signaling, such as interferon-γ, through the JAK1 tyrosine kinase and STAT1 transcription factor in immune cells (147) and the proinflammatory cytokines TNFα and interleukin-1β through actions of the NFκB/RelA transcription factor (148). The phosphorylated R-Smads combine with Smad4 and enter the nucleus to regulate the transcription of targeted gene. The nuclear accumulation of Smads can be modulated by other signaling pathways. Ras signaling can inhibit ligand-induced Smad1/2/3 nuclear accumulation, directly interfering with Smad-dependent responses, while this effect is regulated through the Smad1/2/3 phosphorylation by Erk1 and Erk2 protein kinases which are activated by Ras (149). Ubiquitin/proteasome-mediated degradation is involved in the decrease in the nuclear concentration of phosphorylated Smad2, which is activated by TGFβ (150). The exact function of the ubiquitin-dependent degradation of Smad2 remains unclear. It may rapidly eliminate the signal of TGFβ or remove extra phosphorylated Smads from the nucleus by targeting Smads that do not bind to the DNA sequence of the targeted gene.  22  1.4.3. The roles of TGFβ superfamily members in folliculogenesis In addition to the pituitary gonadotropins (FSH and LH) and GH, steroids, growth factors and other cytokines act in on the autocrine/paracrine systems and play important roles in ovarian follicle growth (review in (92)). Before the small antral stage, the development of the follicle is thought to be gonadotropin-independent. At these early stages, signals from the oocyte and the surrounding granulosa and theca cells are considered to promote the progression of follicular development. Five different classes of growth factors within human ovarian follicles have been described: IGF, TGFβ, TGFα, FGF and cytokines. The role of TGFβ superfamily members in ovarian organogenesis and folliculogenesis has been studied widely in animals. These studies show that the oocyte, granulosa and theca cells express various TGFβ superfamily ligands in a developmental-stage related manner, and these ligands play important roles in folliculogenesis, including the initiation of the primordial follicles, the proliferation/atresia of the somatic cells, the expression of the gonadotropin receptors, the production of steroids, oocyte maturation, ovulation, luteinization and the formation of the corpus luteum (reviewed in (1, 151)).  1.5.  Inhibin and Activin  1.5.1. The gene and protein structures of inhibin and activin Inhibins are heterodimers of two inhibin subunits (α- and either βA- or βB-subunit), from which two isoforms are associated: inhibin A (αβA) and inhibin B (αβB) (Fig. 1.5). The homology of amino acid sequence between the α subunit and β-subunits is 23 to 27%, while the homology between β subunits is 64% (74). Activins are homodimers of the β-subunit. The primary isoforms of activin are activin A 23  (βAβA), AB (βAβB) and B (βBβB). There are other isoforms of activin (C, D, E) have also been B  recognized, however their biological significance remains unclear (152). Activin C is highly expressed in the mouse liver (153) and human prostate and liver (154), while the βC-subunit is disable to dimerize with the α subunit to make inhibin C (αβC ) (154). Activins and inhibins are primarily identified as they can increase or decrease pituitary cells FSH production in vitro. The effect of activin is thought to be antagonized by inhibin. Whether inhibin A shows qualitative and quantitative similarities with inhibin B in biological activities still remains unclear. Inhibin A and B exhibit similar bioactivities, inhibiting rat pituitary FSH secretion in vitro (155), while inhibin B activity shows 15 to 20% similarity with that of inhibin A in ovine pituitary cell cultures (156). In addition, activin A suppresses EGF-stimulated DNA synthesis in rat hepatocyte primary cultures, while activin B is inactive (157). It is uncertain whether the activins or inhibins exhibit similar behavior in humans.  1.5.2. Signaling pathways The activin signaling pathway is well characterized, while the inhibin signaling pathway remains unclear. Firstly activin combines with the its type II receptors (ACVR2A and ACVR2B) specifically. After combining, the ligand-type II receptor dimer binds type I receptors (ACVR1A or ACVR1B) to create a complex. This union promotes the type I receptor auto-phosphorylation, which then activates the downstream Smads system. Activin causes Smad 2/3 to associate with Smad 4 and then go to the nucleus together. Interacting with other transcription factors, the complex binds to specific DNA sequence and leads to gene stimulation or suppression. Compared to activin, it remains unclear about inhibin signaling pathway. There are no  24  specific inhibin receptor identified in genome searches (158). Based on the inhibin-induced suppression of the activin stimulatory effects on FSH production in pituitary and in other cells where the action of activin is antagonized by inhibin, that the authors speculate that inhibin may not act through a particular inhibin receptor but through binding to activin receptor and then blocking activin binding to its receptor (159). High dose of inhibin can bind to ACVR2A/B and block activin signaling (133). The study has been shown that betaglycan can binds with inhibin and promotes the binding of inhibin to ACVR2A/B and thus it is a very potent antagonist of activin signaling pathway (107). However, while betaglycan appears to play an important role in inhibin signaling, additional findings propose other mechanisms may be involved. The inhibin-binding protein (InhBP/p120) does not bind inhibin, but it facilitates the inhibin-induced suppression of activin signaling (160). It is yet to be established whether inhibin can bind to its particular receptor or whether it competes with other TGFβ superfamily members. Both inhibin and activin are shown to have a similar action in some systems and this is hard to clarify based on the theory that inhibin inhibits activin actions. However, inhibin has recently been shown to also be able to antagonize BMP signaling, suggesting that it has a wider mode of action (110). This observation may explain the activin-independent stimulation of inhibin mentioned above, and inhibin may regulate BMP signaling. Other binding proteins, such as follistatin and FSTL3, may also play important roles in this process.  1.5.3. Expression profiles of inhibins and activin A in the menstrual cycle In the human menstrual cycle, inhibin A, inhibin B and activin A have different expression profiles. Inhibin A begins to increase in the late follicular phase and has one peak at ovulation  25  and another in the middle luteal phase, while inhibin B increases in the later luteal phase, peaking in the middle follicular phase and one day after LH surge (Fig. 1.6) (review in (161)). Activin exists in both a free form and a follistatin-bound form, in which it is nearly irreversibly bound (162). In the human menstrual cycle, free activin A levels do not fluctuate (163), whereas total activin A shows small peaks before the beginning of menses, in ovulation and in the later luteal phase (164).  1.5.4. Regulation of inhibins The productions of inhibin A and inhibin B are under different regulation. Several studies have already shown that FSH and cAMP induce inhibin A, but not inhibin B, inhibin A levels increase while inhibin B levels remain constant in follicular fluid following follicle maturation (165-168). In contrast to the promoters of the inhibin α- and βA-subunits (169-171), the promoter of the inhibin βB-subunit lacks cAMP-response element (169, 172); thus, inhibin B secretion cannot be induced by FSH and cAMP in granulosa cells. Other studies have shown that activin, TGFβ and BMP2 can regulate inhibin βB-subunit expression (173-175) and that inhibin B secretion requires IGF-I, BMP2, GDF9 and/or the stimulation of the PKC pathways (168, 175-177).  1.5.5. The functions of inhibins and activin A in the ovary The regulation of FSH levels is very important for follicular development. Inhibins have been reported to suppresses FSH secretion in pituitary culture (178); however, the relative function of inhibin A and inhibin B in humans is still being established. In the late luteal phase, following the corpus luteum demise, estradiol levels decrease and  26  consequently FSH levels increase because of the removal of the negative feedback. Inhibin A decreases in a similar manner and it suggests that inhibin A may play a role in this process. Some studies provide indirect evidence that inhibin B functions in FSH modulation because FSH increases selectively following the inhibin B decline in the early follicular phase (167). Serum activin A levels show a small rise following the increase of FSH in the menstrual cycle, but in physiologic situations, activin A levels do not fluctuate along with largely varied FSH levels (179, 180). Therefore, activin A may not play an endocrine-related role in regulating the secretion of FSH. In addition to their potential endocrine-related roles in FSH control, some studies have showed that inhibin and activin play vital roles in regulating follicle development (181).  1.6.  FST and FSTL3 FST is a single chain protein that was first isolated from the follicular fluids of both  porcine (182) and bovine (183) and found to possess pituitary FSH release-inhibitory activity. FST has a similar biological activity as inhibin, but its structure is totally different, because it is a monomeric protein, while inhibin is a dimeric protein. FST neutralizes most activin activities by binding activin with a very high affinity (Kd=50-500 pM) and preventing it from binding to its receptors (184, 185). These activin activities include promoting pituitary FSH release (186), enhancing FSH-stimulated ovarian granulosa cell differentiation (187) and inducing the formation of Xenopus embryo mesodermal tissue (188). It also binds other ligands of the TGFβ superfamily such as BMP2, BMP4, BMP7, BMP15 and GDF11, but its affinity is lower than activin (130, 131, 189-191). FSTL3 is a secreted glycoprotein that was first identified from a B-cell leukemia line; because of its major sequence homology to FST, it was firstly called  27  FST-related gene (FLRG) (192). FSTL3 has been detected in the cultured media of HeLa, JAR and LOVO cells (192). In addition, a recombined mouse FSTL3 protein is shown to bind activin and antagonize activin-regulated gene transcription in vitro (132). Compared to FST, FSTL3 cannot bind to BMP2, BMP4 and BMP7 (193).  1.6.1. The gene and protein structures of FST and FSTL3 FST is present in the follicular fluid in several forms with molecular weights ranging from 31 to 39 kDa, according to several protein studies (182, 183). The amino acid homology of FST protein is high among different species. The FST gene consists of six exons, and its length is about 6 kb. There are two major preproteins of FST that are generated at an alternative splicing site. One preprotein is full-length and has 344 aa, and the another has a shortened C-terminus, missing exon 6, and has 317 aa (194). The mature peptides of FST are evolutionarily conserved in structure. They consist of a signal peptide that is followed by an N-terminal domain, three successive follistatin domains (FS domains) and a C-terminal domain (195). Although encoded by separate exons, each FS domain consists of 73-77 aa and is distinguished by 10 conserved cysteine residues. With the exception of FST, a number of secreted proteins are also found to have these FS domains, including follistatin-like (FSTL) proteins such as FSTL3 (192), SPARC (for secreted protein acidic and rich in cysteine (196), agrin (197) and matrix glycoprotein SC1 (198). Apart from follistatin and FSTL3, which are major regulators of activin action, most members do not have the activity like follistatin. The three main domains of FST have different functions. Residing in the N-terminal domain of FST, two specific tryptophan residues at positions 4 and 36 have been suggested to be the important sites accountable for the binding of FST with activin (199). FS domains are  28  presumed to serve as growth factor binding motifs (200). In addition, residues on the second FS domain are found to be important for activin binding (201). The number and sequence of FS domains are essential for full FST activation because the ability of FST to bind activin can be attenuated or destroyed by the replacement or rearrangement of the FS domains (201). Two FST molecules largely surround one activin dimer (202) and completely block the binding sites for both activin type I and type II receptors. Furthermore, the C-terminal domain is able to stabilize the FST-activin complex, as a study has shown that one FST molecule appears to contact the N-terminus of the other FST molecule, so that they become fastened in place (202), providing a potential explanation for the almost irreversible dynamics mentioned above (162). The removal of the signal peptide from the preprotein reveals the mature peptides, and three main isoforms of follistatin (FST315, 315 aa; FST300, 300 aa; FST288, 288 aa) form (review in (203)). After different proteolytic processing, the C-terminal sequences of these isoforms are different. Less than 5% of follistatin mRNA produces FST288, while the longer isoform FST315 is the main product (204). The ability to bind heparin-sulfated proteoglycans located in the cellular surface is the main difference between these isoforms functions. The main binding sites for heparin are located in residues 72–86 of the first FS domain (182). The studies have already shown that the peptide fragments from this region have a high heparin binding affinity (205). This structural characteristic allows the shortest isoform, FST288, to have a higher affinity for heparin, while the longest isoform, FST315, has a reduced heparin affinity possibly caused by its C-terminal acidic tail that covers the sites for heparin binding, which FST288 does not possess (205, 206). This difference in heparin binding allows FST288 to bind to the cell surface, bringing activin for degradation and consequently blocking its autocrine-, paracrine- and endocrine-related actions (207), while FST315 localizes mainly in  29  circulation, consistent with its small or absent affinity to heparin (208). FST315 can act as storage for follistatin in the circulation, carry activin to target cells and inhibit activin from binding to FST288 for degradation. FST303, which is produced by the proteolytically cleavage of the C-terminal acidic tail, is distributed predominantly in the fluids and extracts of the gonadotrope (209). FSTL3 has several structural characteristics in common with FST. Like FST, FSTL3 can also bind to activin A with a high affinity to form a complex. Structural studies of the complex revealed that two FSTL3 molecules envelop one activin molecule, while FSTL3 has a more general contact with activin than FST, promoting its activin binding affinity despite its lack of the third FS domain (210). The specificity of ligand binding may be affected by this extensive activin binding pattern (210) and also may be the potential explanation for the difference in the binding of BMP to FST and FSTL3, as previously mentioned (193). Like FST315, FSTL3 also does not have heparin binding sites, and therefore, under normal conditions, it does not have the ability to bind cell-surface proteoglycans. These biochemical distinctions also affect their biosynthesis and secretion. FST315 is the most rapidly produced and released, while FST288 is released more slowly and FSTL3 is the slowest to be released, with newly produced peptides being both released and translocated to the nucleus (211). Taken together, these results indicate that FSTL3 and different FST isoforms have different biological actions in vivo.  1.6.2. The actions of FST and FSTL3 When FST was first purified from follicular fluid, there was little information about its signaling mechanism, but later the concept of FST as an activin binding protein allowed for a breakthrough in the research (130). FST can bind activin with a high affinity, and the Kd can be  30  up to 500 pM. In addition because the binding of activin to FST is almost irreversible (162), FST can be looked at as another kind of activin receptor that can strongly inhibit activin actions. However, the ratio of the neutralization ability of FST to activin is varied, from about equal to sevenfold in different cell systems, which may be caused by the amounts and types of activin receptors expressed in various cell and the different binding affinities to FST. Generally, two FST molecules contact one activin dimer to form the complex, and thus one β-subunit of activin binds to one FST molecule (212). Based on this characteristic, activin B or activin AB can bind FST with similar affinities as that observed with activin A (146), while inhibin binds FST with a markedly lower affinity because it only has one β-subunit (212). It remains unclear whether other activins such as activin C, activin D and activin E can interact with FST. How does FST antagonize activin actions at the cellular level? Some studies already provide some answers to this question. As mentioned above, FST288 binds to activin, and this complex is able to bind to the heparin-sulfated proteoglycans and be endocytosed quickly by lysosomal enzymes (207). Although FST315 binds activin with a similar affinity to FST288, it has a lower affinity for heparin, and the binding of activin to FST315 can allow it to escape degradation (213). In certain situations, activin can be released from its bond with FST315 and interact with its receptor to activate downstream signaling. The role of fate in the FST315-activin complex is not well understood and need to be further investigated. In addition to the main function of FST, which is to antagonize the actions of activin, compelling evidence suggests that it is not exclusively an activin binding protein and can also bind to other TGFβ superfamily members, particularly BMPs (131). The affinity of FST for these proteins remains unclear, but it is supposed to be lower than that for activin, and the  31  physiological significance of these findings has not yet been described in detail. Even though the interaction between FST and GDF9 or BMP6 was not reported, FST can bind BMP15 and inhibit BMP15-induced the proliferation of rat granulosa cells and -reduced FSH receptor expression in rat granulosa cells (191). These results suggest the roles of FST in the ovary are wide. Because BMP15 is the homologue to GDF9, these results also suggest that FST is possible to bind to GDF9 and modulates its function, but this hypothesis needs to be investigated further. FST is also shown that it can bind with BMP2, BMP4, BMP7 and BMP4/BMP7 heterodimers and neutralize their actions in Xenopus embryos, however no reports are obtainable in mammalian (189). Like FST, FSTL3 can also bind to GDF8 (also called myostatin) (193). However, unlike FST, BMP2, BMP4, BMP6 and BMP7 did not compete with activin to bind to FSTL3 (193). The relative abilities of the FST isoforms or FSTL3 to bind cell-surface proteoglycans are correlated with neutralization of exogenous GDF8 or BMPs bioactivity. These results indicate that the differential biological actions among the FST isoforms and FSTL3 are principally reliant on their relative cell-surface binding abilities and ligand specificities.  1.7.  GDF9 and BMP15  1.7.1. The gene and protein structure of GDF9 and BMP15 GDF9 was first recognized as a TGFβ superfamily member from mouse genomic DNA in 1993 (77). Human GDF9 was obtained from a complementary DNA library prepared from mRNA that was isolated from human adult ovaries (214). Five years after the discovery of GDF9, a close homolog named BMP15 (or GDF9B) was discovered simultaneously by two research groups using homology-based cloning approaches (78, 79). 32  GDF9 and BMP15 have some common traits, and their mature protein regions are quite small. Both of them are created as precursor proteins, with the mature proteins locating in their C-terminus. Human GDF9 and BMP15 precursor protein consists of 454 and 392 aa respectively. The human GDF9 precursor protein contains a hydrophobic stretch of 24 aa at the N-terminus as a signal peptide for secretion, followed by a 295 aa propeptide with has a putative RXXR cleavage site and a 135 aa C-terminal mature protein that is high homologous to other TGFβ superfamily members (214), while the BMP15 precursor protein consists of a 18 aa signal peptide, 249 aa propeptide and a 125 aa mature protein (79). The sequence of the mature GDF9 protein is relatively conserved among different species, while that of the BMP15 mature protein varies a lot. In the C-terminus of the mature protein, the identity between GDF9 and BMP15 is 52.4% in human, 53.2% in sheep, 47.6% in mouse and 44.8% in rat. The human GDF9 propeptide contains six potential N-glycosylation sites at 106, 163, 236, 255, 268 and 338. Unlike GDF9, there are only five potential N-glycosylation sites at 87, 147, 237, 277 and 373 for BMP15. Most members of the TGFβ superfamily can form homodimers or heterodimers with other members of TGFβ superfamily. In general, TGFβ superfamily member mature proteins contain seven cystines. Six of them form a specific cystine knot and the remaining one involves in the formation of a dimer that is linked by a disulfide bond (74). In contrast, GDF9 and BMP15 lack the cystine involved in dimer formation and the missing cysteine residue is replaced by a serine residue (76, 78, 79, 215). Therefore, they may only form dimers through noncovalent interactions. Although the biological activities of other members of TGFβ superfamily require the dimer formation, whether it is also important for GDF9 and BMP15 remains unclear. However, recent results show that it is possible for the formation of biologically active heterodimers of GDF9/BMP15 because when GDF9 and  33  BMP15 are co-expressed in transfected HEK293T cells, both heterodimers and homodimers can be formed (216). Because combining conditioned media containing GDF9 and BMP15 and adding it to the granulosa cell culture is enough to observe their co-operative effects, it is thought that these proteins eliciting their effects may not need to be produced by the same cell (217). Therefore, these growth factors could be present as homodimers, heterodimers or even as monomers to affect granulosa cell development.  1.7.2. The expression and regulation of GDF9 and BMP15 The GDF9 mRNA and protein are expressed in the growing oocytes at all stages of follicular development but not in primordial follicles (218-221). Nevertheless, GDF9 mRNA can be detected in primordial follicles of bovine and ovine ovaries (222). GDF9 mRNA continues to be expressed in mouse oocytes through ovulation until 1.5 days after fertilization (214). BMP15 mRNA is also expressed in the mouse oocyte from the primary follicle stage until fertilization (78). GDF9 is originally thought to be an oocyte-specific factor; however, GDF9 is found recently to express in monkey granulosa cells (223), goat (224) and pig (225) ovaries and granulosa and cumulus cells from human ovaries (226, 227). GDF9 mRNA has also been found in the human uterus, placenta and testis, in non-reproductive tissues such as the bone marrow, adrenal gland, pituitary gland and thymus (228), in rodent testis and hypothalamus (228), in ovine cortical slices (229) and in the brushtail possum pituitary gland (230). In addition to the oocyte, which is the main site of expression, BMP15 and its receptors BMPR1B and BMPR2 are also detected during the in vitro culturing of ovine cortical slices (229). In summary, these data propose that the action of GDF9 and BMP15 may be not exclusive in the  34  ovary. The function of this low-level expression of GDF9 in other tissues outside of the ovary is unclear, but the characteristic high-level expression of GDF9 in the ovary and the regulatory elements in its promoter are of interest to researchers. Using transgenic mice, in which regions of the GDF9 locus are fused to the reporter genes, Yan et al. report that a conserved E-box sequence (CAGCTG) is a vital regulatory sequence for GDF9 expression in the ovary (231). However, the factors and mechanisms involved in the commencement of GDF9 expression in the oocyte have still not been identified. Cho et al. report that the mRNA levels of the Kit ligand and Kit itself, but not GDF9 and BMP15, are induced in a culture of cumulus-oocyte complexes and granulosa cells by FGF7, which may play an important role in regulating the stimulation of oocyte growth (232). The germ cell nuclear factor (GCNF) which is a transcription factor, can repress the transcription of GDF9 and BMP15 by binding to a AGGTCA repeat with 0 base pair (bp) spacing between the half sites (DR0) in their promoter (233). This repression directly regulates oocyte-somatic cell paracrine communication and hence affects the fertility of female mice. The results of oocyte-specific GCNF knockout female mice has also shown that the absence of GCNF repression may cause increased expression of GDF9 and BMP15 and consequently the formation of abnormal double-oocyte follicles (233).  1.7.3. GDF9 and BMP15 Signaling pathway Recent studies have been shown that GDF9 signaling is regulated by the TGFβ type I receptor (TβRI) (234) and that its type II receptor is a BMP type II receptor (BMPR2) (235); therefore, it mediates signaling by phosphorylated Smad2 and Smad3 (176, 236). This is a  35  particular type I-type II receptors complex and is the first report of a TGFβ superfamily member utilizing both a TGFβ-type receptor (TβRI) and a BMP-type receptor (BMPR2) for its signaling. BMP15 probably utilizes the same type II receptor as GDF9 (i.e. the BMPR2) (235), while its type I receptor is suggested to be ALK6 (BMP1B) (79); thus, it can activate Smad1, Smad3 and Smad5 to regulate downstream signaling (237). In addition, mouse BMP15 is reported to cooperate with GDF9 via the BMPR2 and ACVR1B/TGFβR1/ ACVR1C-mediated pathways to stimulate rat granulosa cell proliferation (238).  1.7.4. Biological functions of GDF9 and BMP15 in follicle development GDF9 and BMP15 are reported to be the important regulators of follicular development, including the initiation of the primordial follicles and the further development of follicles. Both GDF9 and BMP15 are considered to be related to follicle formation because the primordial follicle population is not altered in GDF9 knock-out mice, and the follicle number in sheep is not changed in sheep missing BMP15 (219). These results are consistent with the lack of GDF9 expression in murine follicles and the lack of BMP15 expression in ovine follicles until the follicles start growing. The follicular development of the GDF9 null mutant female mice are arrested at the primary stage and are infertile, while male mice are fertile, demonstrating that GDF9 plays an important role in stimulating early follicle growth (219, 239). The results of Vitt et al. show that GDF9 in vivo treatment promotes the primary-preantral follicle transition (240). Furthermore, GDF9 stimulates the survival and development of human primordial follicles (241). The in vitro exposure of rodent ovarian tissues to GDF9 has been shown to stimulate the growth of the primary follicle (242). In contrast with GDF9 knockout  36  mice, null BMP15 mutations have little effect on folliculogenesis and fertility, and the mice show weak ovarian phenotypes (243). Taken together, GDF9, but not BMP15, may be critical to the beginning of the growth of primordial follicle. Following the follicle growth, the oocyte within the antral follicle continues to produce GDF9, BMP15 and BMP6 to affect the cell development of the surrounding granulosa cells and consequently regulates antral follicle growth (244, 245). The immunoneutralization of GDF9 and BMP15 by their corresponding peptides in sheep causes anovulation, and these results suggest that they are necessary for follicle development before ovulation (246). Following the preovulatory LH surge, cumulus cells lose intercellular connections with the oocyte, but they undergo cumulus expansion, which is regulated by some proteins, such as cyclo-oxygenase 2 (COX2), hyaluronan synthase 2 (HAS2), urokinase plasminogen activator (uPA), tumor necrosis factor-induced protein 6 and pentraxin 3. Studies on mice have shown that recombinant GDF9 or BMP15 can mimic the actions of oocytes, and they are critical in regulating the proteins involved in cumulus expansion (26, 247). In addition, GDF9 can regulate StAR and LHR transcription in granulosa cells from antral follicles, which are also related to cumulus expansion in mouse (26).  1.7.5. Roles of GDF9 and BMP15 in granulosa cell functions Both GDF9 and BMP15 have been reported to be vital in regulating granulosa cell functions. First, GDF9 (248) and BMP15 (249) can stimulate granulosa cell mitosis and proliferation. Also, GDF9 can increase inhibin α-subunit mRNA levels in explants of rat neonatal ovaries (220) and inhibin A and B production in cultured rat granulosa cells (250). The effects of GDF9 obtained from different species on inhibin production in ovine granulosa  37  cells are different: mouse GDF9 inhibits while ovine GDF9 stimulates inhibin production (251). In contrast, BMP15 does not have any effect on the inhibin production in bovine or ovine granulosa cells (251). In addition, GDF9 (252) and BMP15 (249) can be modulators of steroidogenesis in granulosa cells. GDF9 and BMP15 can also attenuate the stimulated effects of FSH in cAMP, estradiol and progesterone production and LHR expression in rat granulosa cells, possibly through inhibiting FSHR expression and/or FSHR binding to the Gs protein (248). Additionally, the expression of the Kit ligand, which is a critical regulator in female reproduction, can be modulated by GDF9 and BMP15. Kit ligand expression is reduced by GDF9 in mouse granulosa cells (253) and induced by BMP15 in rat granulosa cells (254). This difference may be caused by the species difference between mouse and rat.  38  1.8.  Objectives As mentioned above, activins, inhibins and GDF9 have been shown to play important roles  in follicle development. During the human menstrual cycle, plasma inhibin B levels fluctuate in a typical, cyclical pattern distinct from that of inhibin A (255, 256). Gonadotropins stimulate steady-state inhibin α- and βA-subunits mRNA levels in hGL cell cultures (257), but do not stimulate βB-subunit mRNA levels which are up-regulated by TGFβ (174), activin A (173) and B  BMP2 instead (177). Treatment with either activin A (173) or GDF9 (176) alone has been shown to increase inhibin βB-subunit mRNA and inhibin B levels in hGL cells. However, previous studies have only focused on the effects of activin A or GDF9 alone, while the interactions between activin A and GDF9 on inhibin B accumulation remain unknown. Furthermore, activin A is a well-known inhibitor of luteinization but whether GDF9 plays a role in the intracellular and extracellular regulation of activin A is unclear. The objectives of this thesis were to investigate the following: z  To investigate the interaction between GDF9 and activin A on inhibin subunit (α, βA, βB) mRNA levels, inhibin A and B accumulation, and related intracellular regulating signaling mechanisms in hGL cells;  z  To determine if GDF9 is expressed in hGL cells, and if it so, whether it plays a role in regulating activin A-induced inhibin subunit (α, βA, βB) mRNA levels, inhibin A and B accumulation, and the mechanisms involved;  z  To characterize the role of GDF9 in the expression of follistatin and follistatin-like 3 protein, well-known extracellular inhibitors of activin A, and whether they are involved in regulating GDF9 and hence, its interaction with activin A;  39  z  To study whether the progesterone production, an important function of hGL cells, is affected by any interactions between GDF9 and activin A. To accomplish this, we used human granulosa-lutein cells obtained from women  undergoing in vitro fertilization (IVF) as our research model. Because these cells had been exposed to pharmacological doses of exogenous gonadotropins and in the process of luteinization from hCG stimulation, we recognize the limitations of extrapolating our results to normal ovarian physiology. However, compared to other in vitro and in vivo research models, our hGL cell primary culture model does provide a homogenous population of granulosa cells which can be used to study the cell-type specific spatio-temporal roles of, and interactions between, selected growth factors for research by eliminating the impacts from others. In addition, our in vitro culture model allows for detailed studies involving multiple manipulations and endpoints, which would be difficult or impossible to do in vivo. Thus, in the absence of granulosa cells from the unstimulated, normal human ovaries that are easily accessible and of sufficient amount for research, findings from our cell model do provide interesting hypotheses for further exploration particularly during the follicular-luteal transition.  40  1.9.  Tables  Table 1.1. The type II receptors of the TGFβ superfamily and the ligands known to bind with the type II receptor. Type II receptors TβR2 ACVR2A ACVR2B  Alternative names  Ligands binding to the receptor  AAT3, FAA3, LDS1B, LDS2B,  TGFβ  MFS2, RIIC, TAAD2, MFS2 ACVR2, ActRII, ActRIIA,  Activin A, Inhibin A/B, GDF5/6/7, GDF8 (Myostatin),  ActR-IIA  GDF11 (BMP11), BMP2, BMP3, BMP6/7, BMP10, BMP15  ActRIIB, ActR-IIB,  Activin A, Inhibin A/B, Nodal, GDF5, GDF8 (Myostatin),  MGC116908  GDF1, GDF11, (BMP11), BMP2, BMP6/7  BMPR-2, BMPR-IIBRK-3, BMPR2  FLJ41585, FLJ76945, PPH1,  Inhibin A (with TβR3) , GDF5/6, GDF9, BMP2/4, BMP6/7, BMP15  T-ALK AMHR2  AMHR, MISR2, MISRII, MRII  AMH (MIS)  Table 1.2. The type I receptors of the TGFβ superfamily and the ligands known to bind with the type I receptor. Type I receptors ALK1 ALK2 ALK3 ALK4 ALK5 ALK6 ALK7  Alternative names  Ligands binding to the receptor  ACVRL1, TSR1, SKR3, HHT, HHT2,  TGFβ1, Activin A  ORW2, ACVRLK1 ACVR1A, ActRIA, TSK7L, SKR1,  TGFβ, Activin A, AMH (MIS), BMP6/7  FOP, TSR1, ACVRL2 BMPR1A, BMPR1, BRK1, Tfrl1,  BMP2/4, BMP6/7, BMP10, GDF5/6/7, GDF8  ACVRLK3, CD292, SKR5  (Myostatin), AMH Activin A, activin B, GDF1 and Nodal (with Cripto),  ACVR1B, ActRIB, SKR2, ACVRLK4  BMP3, GDF8 (Myostatin), GDF11 (BMP11)  TβR1, SKR4, AAT5, LDS1A, LDS2A,  TGFβ, GDF8 (Myostatin), GDF9  ACVRLK5 BMPR1B, BRK2, CDw293, ACVRLK6  BMP2/4, BMP6/7, BMP10, GDF5/6/7, BMP15, AMH  ACVR1C, ACTR-IC, ACVRLK7  Nodal, Activin B  41  Table 1.3. Combinational interaction of TGFβ superfamily receptors and Smads  Type II receptors  Type I receptors  Activated Smads  TβR2  ALK5, ALK1 ALK2  Smad2, Smad3, Smad1, Smad5  ACVR2A, ACVR2B  ALK4  Smad2  ACVR2B  ALK7  Smad2  BMPR2  ALK2, ALK3, ALK6  Smad1, Smad5, Smad8  AMHR  ALK3, ALK2, ALK6  Smad1, Smad5  Table 1.4. The proteins interacted with Smads linker  MH1 Function  Regulatory phosphorylation  Nuclear import  Ubiquitination  MH2 Oligomerisation  Cytoplasmic anchoring  Cytoplasmic anchoring  DNA-binding transcription  Transcription  CamKII (–) (S2)  CamKII (–) (S2)  PKC (–) (S2, S3)  Erk (–) (S1-3)  Type I receptors (+) (S1-3, S5, S8)  Receptors  ALK1-7  Oligomerisation  R-Smads, Co-Smad  Cytoplasmic  Calmodulin (S1-4) Filamin  adaptors-effectors  (S1-6)  Filamin (S1-6)  Axin, Axil (S2, S3)  Importin-b 1 (S3) Dab2 (S2, S3) SARA, Hrs/Hgs (S2, S3) ARIP (S3)  STRAP (S2, S3, S6, S7)  β-catenin (S4) Microtubules (S2-4) TAK1 (S6) Ubiquitination  HEF1 (N-ter) (S3)  Smurf1 (S1, S5, S7)  HEF1 (C-ter) (S3)  Smurf2 (S2, S3, S7)  SCF subunits (S3)  adaptors-substrates APC subunits (S3) Transcription factors  ATF2 (S3, 4)  AR (S3)  Jun, JunB, JunD (S3, S4)  BF-1 (S1-4)  Lef1/Tcf (S2, S3)  E1A (S1-3)  42  linker  MH1 Transcription factors  MH2  Sp1, Sp3 (S2-4)  ERα (S2-4)  TFE3 (m E3) (S3, S4)  Evi-1 (S3)  VDR (S3)  FAST (FoXH1) (S2, S3)  YY1 (S1, S3, S4)  Fos (S3) GR (S3) Lef1/Tcf (S2, S3) Menin (S2, S3) Milk (S2) Mixer (S2) Gli3 D C-ter (S1-4)  OAZ (S1, S4)  HNF4 (S3)  Runx/CBFα  p52 (NFκB) (S3) Transcriptional co-activators  pX HBV (S4)  MSG1 (S4) p300/CBP (S1-4) P/CAF (S1-4) Swift (S1, S2)  Transcriptional co-repressors  HDAC (?) (S3)  Hoxc-8 (S1)  Hoxc-8 (S1)  SIP1 (S1-3, S5) Ski (S2-4) SnoN (S2-4) TGIF (S2) Tob (S1, S4, S5, S8)  SNIP1 (S1, S2, S4)  A simplified diagram of the three Smad domains is followed by a table of the Smad post-translational modifications and protein-protein interactions known to occur in each domain. The symbols (+ and –) indicate regulatory phosphorylation of Smads that results in functional activation or inhibition, respectively. Entries in more than one domain indicate interactions with or modifications by the same factor at multiple domains. The specific Smad members that are known to exhibit the listed modifications or interactions are shown in parenthesis and are abbreviated as S1-S8 for Smad1 to Smad8, respectively. Proteins, for which the specific Smad domain that they interact with is not yet determined, are listed in the centre in stippled boxes. A question mark (?) indicates that HDAC activity but not physical protein interaction has been found to associate with the MH1 domain of Smad3. The names of factors not discussed in the text are: TAK1 (TGF-activated kinase 1), pX HBV (pX oncoprotein of hepatitis B virus), Swift (Xenopus BRCA1 C-terminal domain nuclear protein), MSG1 (melanocyte specific gene 1, transcriptional co-activator), Hoxc-8 (homeobox c-8 transcriptional repressor), SNIP1 (Smad nuclear interacting protein 1, Smad- and p300-associating transcriptional corepressor), SIP1 (Smad interacting protein 1, zinc-finger/homeodomain repressor), Tob (transducer of ErbB2, APRO/Btg family of anti-proliferative factors), ATF2 (activating transcription factor 2), Lef1/TCF (lymphoid enhancer-binding factor 1/T cell-specific transcription factor 1), Sp1, Sp3 (Specificity protein 1, zinc finger transcription factor), TFE3 (transcription factor recognising the immunoglobulin enhancer motif E3), VDR (vitamin D receptor, nuclear hormone receptor), YY1 (yin yang 1, zinc finger transcription factor), AR (androgen receptor, nuclear hormone receptor), BF-1 (brain factor 1 oncoprotein), E1A (early region of adenovirus binding transcription factor 1A), ER (estrogen receptor), Evi-1 (Evi-1 oncoprotein), FAST (Forkhead activin signal transducer), GR (glucocorticoid receptor, nuclear hormone receptor), Menin (multiple endocrine neoplasia-type 1 tumour suppressor protein), Milk (Mix 1-related homeobox transcription factor), Mixer (homeobox transcription factor), OAZ (olfactory factor O/E-1-associated zinc finger protein), Runx (runt  43  domain transcription factor), Gli3 C-ter (glioblastoma Kruppel zinc finger transcription factor-3 with deletion of the C-terminal domain), HNF4 (hepatocyte nuclear factor 4, nuclear hormone receptor), NFκB (B cell-specific nuclear factor binding to the intronic light chain enhancer). For references see review (94). (Moustakas A, Souchelnytskyi S, Heldin CH 2001. Journal of cell science 114:4359-4369. Adapted with permission)  Table 1.5. Non-signalling binding proteins of TGFβ superfamily Name  Alternative names  Ligands binding to the receptor  Betaglycan  TβR3, BGCAN  TGFβ1-3, inhibin A (with ActRIIA)  BAMBI  NMA  BMPs, activins  Cripto  FRL-1  Nodal  Endoglin  CD105, END, FLJ41744, HHT1, ORW, ORW1  TGFβ1/3, activin A, BMP2/7  Inhibin coreceptor, InhBP (Inhibin binding protein)  IGCD1, IGDC1, IgSF1, KIAA0364, MGC75490, p120, PGSF2  44  Inhibin A/B (with ALK4)  1.10.  Figures  FIG. 1.1 Members of the TGFβ superfamily play important roles in the bi-directional communication between oocyte and granulosa cells, and granulosa and theca cells. Both autocrine (thick grey arrows) and paracrine (thick black arrows) signalling events are likely, depending on the expression of appropriate combinations of type-I and type-II receptors on the cell surface. Abbreviations: AMH, Anti-müllerian hormone; BMP, Bone morphogenetic protein; GDF, Growth differentiation factor; TGF, Transforming growth factor. (Knight PG, Glister C 2006. Reproduction 132:191-206. (c) Society for Reproduction and Fertility (2010). Reproduced with permission)  45  A Signal peptide  Pro-region  Mature-region COOH  NH3  I II III * IV C CXGXC CC  RXXR  V VI CXC  B CV  CII X  X CI  CIV CVI  G  C  *  CIII  X  FIG. 1.2 The protein structure of TGFβ superfamily members. (A) A pre-proprotein of TGFβ (Transforming growth factor) superfamily members is composed of a signal peptide, pro-region and a mature region. Cleavage at a dibasic site (RXXR motifs) releases a mature region that contains seven highly conserved cysteine residues, forming three intramonomeric disulphide bonds (CI–CIV, CII–CV and CIII–CVI). NH3, amino-terminus; COOH, carboxyterminus; black line, intermolecular disulfide bond; “C” labelled with “*”, the fourth cysteine residue lacking in GDF9 (Growth differentiation factor 9) and BMP15 (Bone morphogenetic protein 15). (B) The cystine knot motif arises from the three intra-monomeric disulphide bonds by two (CII–CV and CIII–CVI) of them building an eight-membered ring structure, through which the third (CI–CIV) passes. The remaining cysteine (labelled with *) forms an inter-monomeric disulphide bond with that of the other monomer into a dimmer. (Lin SY, Morrison JR, Phillips DJ, de Kretser DM 2003. Reproduction 126: 133–148. (c) Society for Reproduction and Fertility (2010). Adapted with permission)  46  Ligands  P Cytoplasm  II  I  R-Smad  II I  P  Smad6/7 Smurf1/2  R-Smad Smad4 Smurf1/2  P  R-Smad  R-Smad R-Smad  P  Smad4  Nucleus R-Sm P ad R-Sm P ad Smad 4  Co-activator  SBE  Co-respressor Transcription factor  Target gene transcription  FIG. 1.3 Diagrammatic representation of the TGFβ superfamily signaling pathway. Abbreviations: Smad, Son of mothers against decapentaplegia; R-Smad, Receptor-activated Smad; SBE, Smad binding element; Smurf, Smad ubiquitination-related factor;  47  P  : Phosphorylation.  R-Smads (Smad1, Smad2, Smad3, Smad5, Smad8) H2 NLS  β-hairpin ex3  SS*XS* NES L3  PY  Co-Smad (Smad4) H2 β-hairpin NLS  NES  SAD  H3/4 L3 loop  I-Smads (Smad6, Smad7) NLS?  PY  NES? L3  FIG. 1.4 The protein structures of the three subfamilies of Smads. The protein diagrams are arbitrarily aligned relative to their C-terminus. The MH1 (Mad homology 1) domain is coloured in grey and the MH2 (Mad homology 2) domain in wave pattern. Selected domains and sequence motifs are indicated as follows: -helix H2, L3 and H3/4 loops, β-hairpin, the unique exon 3 of Smad2 (ex3), NLS (nuclear localization signal) and NES (nuclear export signal) motifs or putative (?) such motifs, the proline-tyrosine (PY) motif of the linker that is recognised by the Hect domain of Smurfs (Smad ubiquitination-related factors), the unique SAD (Smad activation domain) domain of Smad4 and the SSXS motif of R-Smads with asterisks indicating the phosphorylated serine residues. Abbreviations: Smad, Son of mothers against decapentaplegia; R-Smad, Receptor-activated Smad; Co-Smad, Common Smad; I-Smad, Inhibitory Smad. (Moustakas A, Souchelnytskyi S, Heldin CH 2001. Journal of Cell Science 114:4359-4369. Adapted with permission)  48  Inhibin A, B *  Activin A, AB and B *  α  Inhibin A  Activin A  βA Activin AB *  Inhibin B  *  α βB  Activin B  FIG. 1.5 Schematic diagram of different isoforms of inhibin and activin.  49  βA βA βA βB βB βB  40 100  LH (IU/L)  FSH (IU/L)  150  20  0 150  10  100 5 50  Inhibin B (pg/mL)  0 Inhibin A (IU/mL)  Activin (10xng/mL)  50  0  0 0  7 14 21 Days of Menstrual Cycle  28  FIG. 1.6 Levels of follicle-stimulating hormone (FSH), luteinizing hormone (LH), inhibin A, inhibin B and activin across the menstrual cycle. Total activin A levels only at the time of menses, the midcycle, and the midluteal phase were measured while other times in the menstrual cycle were not measured.  50  2. GDF9 ENHANCES ACTIVIN A-INDUCED INHIBIN B ACCUMULATION IN HUMAN GRANULOSA-LUTEIN CELLS 1 2.1.  Introduction Activin A or GDF9 alone has been shown to increase inhibin βB-subunit mRNA levels in  hGL cells (173, 176), but little is known about their interactions in regulating inhibin subunits and, if so, the underlying cellular mechanisms involved. The objectives of the present work were to further examine GDF9 on basal and activin A-induced inhibin subunit mRNA levels and inhibin B accumulations and activation of relevant receptors and Smads in the activin signaling pathway in hGL cells.  2.2.  Materials and Methods  We first conducted concentration-response studies of activin A (1-50 ng/ml) or GDF9 (1-200 ng/ml) alone on α-, βA-, and βB-subunit mRNA levels in cultured hGL cells to identify the respective concentration for corresponding time-dependent and cotreatment experiments. Free plasma activin A levels have been estimated to be 30-45 ng/ml across the menstrual cycle (161). In our concentration-response experiments, activin A induced increases in βB-subunit mRNA levels in a concentration-related manner from 10 to 50 ng/ml. At 25 ng/ml, βB-subunit mRNA levels were significantly greater than levels for control (0 ng/ml) and the 1 ng/ml concentration; at 50 ng/ml, levels were significantly greater than all lower concentrations of activin A (Fig. 2.1E). In our GDF9 concentration-response experiments, an increasing trend in  1  A version of this chapter has been published. Shi FT, Cheung AP, Leung PC 2009 Growth differentiation factor 9 enhances  activin A-induced inhibin B production in human granulosa cells. Endocrinology 150:3540-3546  51  βB-subunit mRNA levels from baseline (0 ng/ml) or the 10 ng/ml concentration was first observed at the 100 ng/ml concentration (a common concentration chosen in reported studies) but reached statistical significance only at the 200 ng/ml concentration. Guided by these findings, the concentration of 25 ng/ml of activin A and 100 ng/ml of GDF9 were selected to assess the corresponding time-dependent changes and on whether cotreatment of recombinant GDF9 and activin A could affect levels of inhibin subunits, inhibin A and B, and activin A receptors and Smads involved in downstream cell signaling.  Preparation of hGL cells The study was approved by the Research Ethics Board of the University of British Columbia. hGL cells were obtained from women undergoing IVF treatment. For each patient, cells from multiple follicles and consequently follicular fluid were pooled respectively. The extraction procedure of granulosa cells from each patient were modified from that described previously (258). Follicular fluid from each subject was divided equally into 15-ml disposable, sterile tubes and centrifuged at 400 × g for 10 min. After removing the supernatant, the layers of hGL cells with the red blood cell pellet were re-suspended in 2 ml of Hanks' solution (GIBCO BRL Life Technologies, Grand Island, NY) containing 50 µg/ml deoxyribonuclease I (Worthington Biochemical Inc., Freehold, NJ), 0.1% hyaluronidase (Sigma, St. Louis, MO), 50 U/ml Heparin (Sigma), and 0.1 U/ml Blendzyme 3 (Roche, Inc., Indianapolis, IN) in a sterile 50-ml centrifuge tube. The cell suspensions were shaken at 200 r.p.m. for 20 min at 37 C and then layered on 8.0 ml Ficoll-Paque (Amersham Biosciences, Piscataway, NJ) in 15-ml sterile tubes and centrifuged at 600 × g for 20 min. The cell layer was removed from each Ficoll-Paque column and washed three time with 10 ml of DMEM/nutrient mixture F-12 Ham (DMEM/F-12; Sigma) supplemented with 10% fetal bovine serum (FBS; HyClone 52  Laboratories, Logan, UT), 100 U/ml penicillin (GIBCO), 100 µg/ml streptomycin sulphate (GIBCO) and 1 × GlutaMAXTM (GIBCO), and the cells were suspended in 5 ml of medium, counted on a hemocytometer, and brought to a final concentration of 2 × 105 cells/ml. Cell viability was determined by 0.04% Trypan Blue dye (GIBCO), and 1 × 105 viable cells were seeded per well in 24-well culture plates and cultured in a humidified atmosphere of 5% CO2-95% air at 37 C. After 48 h, the above medium containing only 0.5% FBS (“low-serum medium”) was added to each well, and the cell culture was now designated as time 0 for all subsequent experiments described below. Culture media were collected after treatment and stored frozen until assays for inhibin A and B.  Activin A and GDF9 experiments hGL cells were seeded to each well of a 24-well plate for mRNA study and inhibin A and B assays (1 × 105 cells/well) and to each well of a 12-well plate for Smad experiments (2 × 105 cells/well). Cells were stimulated with 1-50 ng/ml activin A (Sigma), or 1-200 ng/ml GDF9 (Peprotech Inc., Rocky Hill, NJ) for 24 h in concentration-response studies. For time-dependent experiments, cells were treated with the concentration of 25 ng/ml of activin A or 100 ng/ml of GDF9 as mentioned earlier, and for 3, 12, and 24 h. For experiments with both GDF9 and activin A, cells were preincubated with 100 ng/ml of GDF9 in low-serum medium for 24 h before stimulation with 25 ng/ml of activin A. In neutralization experiments to render GDF9 inactive, 2 µg/ml of recombinant extracellular domain (ECD) fused to the Fc region of human IgG (receptor-ECD/Fc chimera) of human BMPR2 (BMPR2 ECD; R&D Systems, Minneapolis, MN) and 100 ng/ml of GDF9 were preincubated in low-serum medium for 30 min before adding to cultured hGL cells.  53  RNA extraction and Real-time RT-PCR At the end of the treatment period, medium was removed from the culture plate and RNA was extracted using TRIzol (Invitrogen, Carlsbad, CA). Total RNA (800 ng) was reverse-transcribed into first-strand cDNA according to the protocol of the ThermoScript amplification System (Invitrogen). The primers used for SYBR Green real-time PCR were designed using the Primer Express Software (Applied Biosystems, Foster City, CA) (See Table S1 at Appendices) and tested with the intron-spanning assay. Real-time PCR was performed on the ABI PRISM 7300 sequence detection system according to the manufacturer’s protocol (Applied Biosystems). Amplification specificity using the melting curve and analysis and quantification of the relative mRNA levels using the comparative cycle threshold method were carried out on the ABI Prism 7300 Sequence Detection Software version 1.3 (Applied Biosystems). Relative expression levels were quantified using the comparative Ct method with normalization to human glyceraldehyde-3-phosphate dehydrogenase (GAPDH), and changes following treatments were recorded as fold differences from values in untreated controls at each time point as appropriate.  Smad activation experiments and Western blot analysis After treatment, cells were washed twice with ice-cold PBS and lysed in ice-cold lysis buffer (20 mM Tris-HCl, 150 mM NaCl, 1 mM Na2EDTA, 1 mM EGTA, 1% Triton X-100, 2.5 mM sodium pyrophosphate, 1 mM β-glycerophosphate, 1 mM Na3VO4, 1 µg/ml leupeptin) containing protease inhibitors (Sigma). The extract was centrifuged at 14,000 r.p.m for 15 min at 4 C to remove cellular debris; protein concentrations were determined by the Bradford method (Bio-Rad Laboratories, Inc., Hercules, CA). Twenty micrograms of protein samples  54  were heated to 95 C for 5 min, run in 12% SDS-PAGE gels, and transferred onto nitrocellulose membranes (Bio-Rad Laboratories). The membranes were blocked for 1 h in Tris-buffered saline containing 0.01% Tween 20 with 5% non-fat dried milk, and incubated overnight at 4 C with  the  relevant  antibodies  (Cell  Signaling  Technology,  Inc.,  Beverly,  MA):  anti-phospho-Smad2 (Ser245/250/255), anti-phospho-Smad2 (Ser465/467), anti-phospho-Smad3 (Ser423/425), anti-Smad2, and anti-Smad3. After washing, the membranes were incubated with the secondary antibody, a peroxidase-conjugated anti-IgG (Bio-Rad Laboratories) for 1 h. Immunoreactive proteins were detected using enhanced chemiluminescence reagents (Amersham Bioscience) followed by exposure to Thermo CL-X Posure film (Thermo Fisher Scientific Inc., Waltham, MA). Antiserum to total β-actin (Santa Cruz Biotechnology Inc., Santa Cruz, CA) was used as the internal control.  Inhibin A and B assays Dimeric inhibin A and B levels in the spent culture media were quantified with specific ELISA kits (Diagnostic Systems Laboratories, Webster, TX). The lowest limits of detection for human inhibin A and B were 1 pg/ml and 7 pg/ml, respectively. Corresponding intrassay coefficients of variation were 3.0 and 3.5%; interassay coefficients of variation were 6.1% and 6.2%. All hormone measurements were performed in duplicate. Secreted hormone levels were normalized to total cellular protein content.  Statistical analysis Each set of experiments consisted of granulosa cells from at least three sets of experiments (each from a separate patient). Real-time PCR samples were assayed in triplicate and ELISA samples were assayed in duplicate. Data were analyzed by one-way ANOVA, followed by 55  Tukey’s multiple comparison tests if the overall P values were significant using the computer software PRISM (GraphPad Software Inc., San Diego, CA). Results were reported as means ± SEM and were considered significantly different from each other at P < 0.05.  2.3.  Results  Dose- and time-dependent effects of activin A or GDF9 on inhibin subunit mRNA levels Activin A, but not GDF9, increased inhibin α-subunit mRNA levels which reached statistical significance at the 50 ng/ml concentration (Fig. 2.1, A and B). Activin A could also induce inhibin βA-mRNA levels, but the small increases (Fig. 2.1C) were not significantly greater than control (0 ng/ml) value; whereas GDF9, at 200 ng/ml, increased inhibin βA-subunit mRNA to levels that were significantly greater than those for all lower concentrations (P < 0.05)  (Fig.  2.1D).  Activin  A  induced  inhibin  βB-subunit  mRNA  levels  in  a  concentration-dependent manner with the maximal level 58-fold of control at the 50 ng/ml concentration (P < 0.001) (Fig. 2.1E); in contrast, GDF9 increased inhibin βB-subunit mRNA level only at 200 ng/ml and to only 1.5-fold of control (P < 0.05) (Fig. 2.1F). For activin A, there were no significant time-dependent changes in inhibin α- and βA-subunit mRNA levels (Fig. 2.2, A and C) but statistically significant time-dependent increases in inhibin βB-subunit mRNA levels (Fig. 2.2E). For GDF9, significant increases were not observed for inhibin α- and βA-subunit mRNA (Fig. 2.2, B and D) but were observed for inhibin βB-subunit mRNA at 24 h (P < 0.05) (Fig. 2.2F).  Effects of GDF9 pretreatment on activin A-induced inhibin βA- and βB-subunit mRNA B  levels and dimeric inhibin A and B secretion Neither concurrent treatment with activin A and GDF9 nor pretreatment with activin A for 56  24 h before adding both activin A and GDF9 enhanced inhibin βA- or βB-subunit mRNAs at 3, 6, 12, and 24 h (data not shown). In contrast, GDF9 pretreatment for 24 h amplified activin A-induced inhibin βA- and βB-subunit mRNAs in a time-dependent manner, reaching a peak at 6 and 12 h for inhibin βA- and βB-subunit mRNA, respectively, before leveling off (Fig. 2.3, A and C, P values all < 0.05). BMPR2 is the type 2 receptor for GDF9 and the ECD of BMPR2 is known to antagonize recombinant GDF9 bioactivity (235, 259) and the mitogenic effects of GDF9 on granulosa cells (260). When GDF9 preincubated with BMPR2 ECD for 30 min was added to the cell culture before activin A treatment, the enhancing effects of GDF9 on activin A-induced inhibin βA- and βB-subunit mRNA levels were attenuated at both 12 and 24 h (Fig. 2.3, B and D, P values all < 0.05). In contrast, there were no significant changes with respect to control when BMPR2 ECD alone was added (Fig. 2.3, B and D). Protein concentrations of inhibin A and B in the corresponding culture media showed that activin A induced only a small rise in inhibin A in cells pretreated with GDF9 at 24 h, and as a result, no significant decrease was observed when GDF9 was first neutralized with BMPR2 ECD. In contrast, Activin A increased inhibin B secretion by 2-fold of control value (P < 0.01) and 4-fold in cells pretreated with GDF9 (P < 0.001), and the latter was attenuated when GDF9 was first neutralized with BMPR2 ECD (Fig. 2.4).  Effects of GDF9 treatment on activin receptors and Smads and cell response to activin A GDF9 enhanced the activin receptors ACVR2B/1B and Smad2/3 mRNAs and reduced Smad7 mRNAs in a time-dependent manner but had no effect on ACVR2A and Smad4 mRNAs (Fig. 2.5). (Baseline ACVR1A mRNA level was too low for further study). Corresponding cell lysates analyzed 5, 15, 30, and 45 min after activin A stimulation showed 57  that GDF9 pretreatment enhanced activin A-induced Smad3 (Fig. 2.6A) and Smad2 (Ser465/467) phosphorylation (Fig. 2.6B), effects that were attenuated by BMPR2 ECD. In contrast, there was little change from control in Smad2 (Ser245/250/255) phosphorylation irrespective of GDF9 treatment and consequently BMPR2 ECD (Fig. 2.6C). The stimulatory effects of activin A, with and without GDF9 pretreatment, on Smad phosphorylation were evident by 30 min after adding activin A because there was little change at 5 and 15 min (results not shown). Thus, GDF9 enhances cell response to activin A by increasing mRNA levels of facilitating components and simultaneously reducing that of inhibitory components of the activin signaling pathway.  2.4.  Discussion  As reported by some investigators (173), our study confirmed that activin A had only limited stimulatory effects on inhibin α- and βA-subunit mRNA levels but marked stimulatory effects on inhibin βB-subunit mRNA levels in hGL cells. We also confirmed that GDF9 alone had no effects on inhibin α-subunit mRNA levels and modestly induced inhibin βA-subunit mRNA levels in these cells only at the higher concentration of 200 ng/ml. Whereas GDF9 increased inhibin βB-subunit mRNA and corresponding inhibin B levels as observed in a B  previous study (176), the effects were much lower than those induced by activin A. Differences in species and follicle development are important consideration in studying the effects of GDF9. For example, GDF9 can stimulate not only inhibin βA- and βB-subunit mRNA levels, but also inhibin α-subunit promoter activity in granulosa cells obtained from small antral follicles of estrogen-treated immature rats (250). In contrast to the effects of GDF9 alone, our study demonstrates for the first time that GDF9 enhances activin A-induced inhibin βB-subunit  58  mRNAs and inhibin B secretion in hGL cells and the underlying mechanisms involved. Pretreatment with GDF9 before activin A stimulation is required to elicit the enhancing effects of GDF9 on activin A-induced inhibin βB-subunit mRNA and inhibin B levels from hGL cells because concurrent activin A and GDF9 cotreatment or activin A pretreatment followed by GDF9 stimulation have revealed no enhancing actions. Using a number of approaches, we further characterized that these enhancing effects were specifically related to GDF9 actions. First, the enhancing effects of GDF9 are neutralized by BMPR2 ECD, a well known GDF9 antagonist (Fig. 2.3 and 2.4); hence, inhibin βA- and βB-subunit mRNA level and inhibin B secretion were not significantly different from those for activin A alone when GDF9 was preincubated with BMPR2 ECD. Second, mRNA levels of activin receptors and Smads in the activin signaling pathway are altered by GDF9. GDF9 increased mRNA level of the receptors, ACVR2B and ACVR1B, and activating Smads, Smad2/3, but decreased mRNA level of the inhibitory Smad, Smad7. Furthermore, GDF9 enhanced activin A-induced Smad2/3 phosphorylation, effects that were attenuated by BMPR2 ECD. Smad2 has two different phosphorylation sites (Ser465/467 and Ser245/250/255). Our results showed that activin A and GDF9 had effect on Ser465/467, located in the carboxy-terminal SSXS sequence of Smad2 but not Ser245/250/255, located in the linker region of Smad2. Phosphorylation of Smad2 on Ser465 and Ser467 is required for oligomerization with Smad4 and translocation into the nucleus in mammalian cells (261). In contrast, Ser245/250/255 serves as phosphorylation sites for proline-directed protein kinases including ERK (262). Thus, whereas our data show GDF9 can activate Ser465/467, other ovarian regulators are required to activate Ser245/250/255. Inhibin A and inhibin B exhibit distinct patterns of secretion across the menstrual cycle.  59  Serum inhibin B levels increase across the luteal-follicular transition, reaching a peak in the midfollicular phase and a second peak on the day after the LH surge. In contrast, inhibin A levels begin to rise in the late follicular phase, reaching a peak in the midcycle and another peak in the midluteal phase (161). Whereas activin A may be present as a free form, it is almost irreversibly bound to follistatin, and there is little variation in free activin A levels (30-45 ng/ml in plasma) across the menstrual cycle (161). hGL cells likely have the capacity to express the various inhibin subunits, but the expression of a specific subunit, hence secretion of a particular inhibin, is influenced, among other factors, by the prevailing facilitating and/or inhibiting autocrine or paracrine factors. Of particular interest is that GDF9 mRNA is expressed in hGL cells and together with activin A, can significantly influence the inhibin βB-subunit mRNA levels and consequently inhibin B secretion. We cannot rule out the possibility that increased levels of inhibin βB-subunit mRNA lead to an increased synthesis of activin B (βBβB) or activin AB (βAβB), but specific activin B and B  activin AB assays are currently unavailable. Although GDF9, activin A, or both have limited effects on inhibin α-subunit mRNA levels, it is expressed in theca interna in addition to granulosa cells, and its levels are higher than those of inhibin βB-subunit mRNA in the ovary (263-265). Because the inhibin α-subunit is available in abundance, a selective rise in inhibin βB-subunit mRNA levels is sufficient to cause a clear increase in the amount of inhibin B produced by hGL cells when stimulated by GDF9, activin A or both. In the Gdf9 null female mouse model (266), inhibin α- and βA-subunits are expressed in Gdf9 null ovaries at similar levels to controls, whereas inhibin βB-subunit is dramatically decreased. However, this may also be related to a lack of antral follicles in Gdf9 null mice. Pan et al. (267) found that mouse GDF9 mRNA was markedly up-regulated between the primordial  60  and primary follicle stage and reached a maximum in the oocyte of the secondary follicle, but its level was still higher in the large antral follicle than in the primordial follicle. Correspondingly, inhibin βB-subunit mRNA was up-regulated during follicular development and reached a peak in the large antral follicle. Teixeira et al. (268) reported that oocyte GDF9 mRNA expression increased progressively during follicle development of the human ovary, with near maximum amounts observed in oocytes of fully grown secondary follicles. In addition, GDF9 mRNA expression remained high in oocytes of healthy small Graafian follicles. These studies and our findings suggest that inhibin βB-subunit mRNA levels in both the human and the mouse ovary are related to those of GDF9. We acknowledge the limitations of extrapolating our results from human granulosa-lutein cells obtained from women undergoing IVF treatment to normal ovarian physiology as these cells have been exposed to pharmacological concentrations of exogenous gonadotropins and in the process of luteinization from hCG stimulation. Nevertheless, in the absence of granulosa cells from the unstimulated, normal ovaries that are easily accessible for research, findings from our cell culture model do provide interesting hypotheses for further evaluation of the role of GDF9 and related mechanisms involved in regulating inhibin subunit and protein production during the periovulatory transition. This notwithstanding, our study suggests that GDF9 increases hGL cell response to activin A by acting on its receptors, BMPR2/TβR1, which then activates facilitating Smad2/3 downstream to form complexes with Smad4. These complexes then activate transcription factors in the nucleus to induce target genes that increase ACVR2B/1B and Smad2/3 and reduce Smad7 activities. These changes, in turn, increase cellular response to activin A stimulation. It is tempting to speculate that increasing GDF9 expression (268) during  61  folliculogenesis enhances human granulosa cell response to activin A, which leads to rising inhibin B levels in the follicular phase. With release of the oocyte after ovulation, a main source of GDF9 is removed; hence, cell response to activin A with respect to inhibin βB-subunit is withdrawn, which may explain the decline in inhibin B secretion after ovulation. However, further studies will be required to characterize the roles of endogenous GDF9 activities in human granulosa cells and its interactions with activin A in modulating levels of inhibin and its subunits.  62  βB subunit mRNA  E  1.5 1.0  a  a  a  a  b  0.5 0.0  0  1 10 25 50  α subunit mRNA  2.0  2.0  a a  1.5 1.0  a  a  a  0.5 0.0  0  1 10 25 50  βA subunit mRNA  D  F 80  c  60  b  40  ab  20 0  a  a  0 1 10 25 50 Activin A (ng/ml)  βB subunit mRNA  βA subunit mRNA  C  B  (fold change relative to control) (fold change relative to control) (fold change relative to control)  α subunit mRNA  A  (fold change relative to control) (fold change relative to control) (fold change relative to control)  2.5 Figures  2.0 1.5 1.0  a  a  a  0  10 100 200  a  0.5 0.0  2.0  b  1.5 1.0  a  a  a  0  10 100 200  0.5 0.0  2.5 2.0 1.5 1.0 0.5 0.0  b ab a  0  a  10 100 200 GDF9 (ng/ml)  FIG. 2.1. Concentration-dependent effect of activin A (A, C, E) or GDF9 (B, D, F) on inhibin subunit mRNA level in hGL cells. After 48 h preculture, hGL cells were cultured in low-serum media (containing only 0.5% FBS) and treated with different concentration of activin A (0-50 ng/ml) or GDF9 (0-200 ng/ml) for up to 24 h. RNA of hGL cells were isolated and mRNA contents were assessed by real-time PCR. Results were the means ± SEM from at least three sets of experiments (each from a separate patient), and in each set, measurements were made in triplicate. Means without a common letter are significantly different (P < 0.05).  63  1.0  a  a  a  0.5 0.0  0  3  12  24  α subunit mRNA  1.5  a  2.0 1.5 1.0  a  a  a  a  0.5 0.0  0  3  12  24  β A subunit mRNA  D  60  c  40 20 0  a 0  ab 3  b 12  24  β B subunit mRNA  F  Time (h) of activin A treatment  (fold change relative to control) (fold change relative to control)  2.0  (fold change relative to control)  α subunit mRNA β A subunit mRNA  β B subunit mRNA  E  (fold change relative to control) (fold change relative to control)  C  B  (fold change relative to control)  A  2.0 1.5 1.0  a  a  a  a  0  3  12  24  a  a  a  a  0  3  12  24  a  a  0  3  0.5 0.0  2.0 1.5 1.0 0.5 0.0  2.0 1.5 1.0  ab  b  0.5 0.0  12  24  Time (h) of GDF9 treatment  FIG. 2.2. Time-dependent effect of activin A (A, C, E) or GDF9 (B, D, F) on inhibin subunit mRNA level in hGL cells. After 48 h preculture, hGL cells were cultured in low-serum media (containing only 0.5% FBS) and treated with 25 ng/ml activin A or 100 ng/ml GDF9 for up to 24 h. RNA of hGL cells were isolated and mRNA contents were assessed by Real-time PCR. Results were the means ± SEM from at least three sets of experiments (each from a separate patient), and in each set, measurements were made in triplicate. Means without a common letter are significantly different (P < 0.05).  64  3  b c  2 b a a a 1 a a a  a a a  b ab a  β B subunit mRNA  b  0 0 3 6 9 12 15 18 21 24 27 Time (h)  2  1  0  c  ab abab  b  d bc cd aba aab  12  aba  24  βB subunit mRNA  D 3  (fold change relative to control)  C 4  (fold change relative to control)  (fold change relative to control)  βA subunit mRNA β A subunit mRNA  B  (fold change relative to control)  A  86  c  c  66  Control GDF9 Activin A GDF9+Activin A  b  46  b  c  b b 6 b 2 a a 1 a a  26  0  a  a  a  a  0 3 6 9 12 15 18 21 24 27 Time (h) 80  d  d b  60  bc  bc  Control GDF9 Activin A GDF9+Activin A GDF9+ECD+Activin A GDF9+ECD ECD  b  40 20 2 aa 1 0  aa aa 12  aa 24  Time (h)  Time (h)  FIG. 2.3. GDF9 pretreatment enhanced activin A-induced βA- (A, B) and βB- (C, D) subunit mRNA B  level in hGL cells, effects attenuated in the presence of BMPR2 ECD, a GDF9 antagonist. After 48 h preculture, hGL cells were cultured in low-serum media (containing only 0.5% FBS). hGL cells were preincubated with 100 ng/ml of GDF9 for 24 h in low-serum media before stimulation with 25 ng/ml of activin A. In neutralization experiments to render GDF9 inactive, 2 µg/ml of BMPR2 ECD and 100 ng/ml of GDF9 were preincubated in low-serum media for 30 min before adding to cultured hGL cells. RNA of hGL cells were then isolated and mRNA contents were assessed by real-time PCR. Results were the means ± SEM from at least three sets of experiments (each from a separate patient), and in each set, measurements were made in triplicate. At each time point, means without a common letter are significantly different (P < 0.05).  65  C Inhibin B (pg/mL)  500 400 300 200 100 0 0  6  12 18 Time (h)  24  500 400 300 200  bcd d cd ab a 100 a abc 0  Control GDF9 Activin A GDF9+Activin A  400 300  b b a a  100 0 0  b  b a  200  a a  a a 6  12 18 Time (h)  D  B Inhibin A (pg/mL)  c  500  30  Inhibin B (pg/mL)  Inhibin A (pg/mL)  A  24  30  c  500 400  b  300  b  200 100 a  a  a  a  Control GDF9 Activin A GDF9+Activin A GDF9+ECD+Activin A GDF9+ECD ECD  0 24 h Treatment  24 h Treatment  FIG. 2.4. GDF9 pretreatment enhanced activin A-induced inhibin B accumulation in culture media. The culture conditions were similar to those in Fig. 2.3. After treatments, media were collected and assayed for inhibin A (panels A and B) and B (panels C and D) levels by ELISA. Results were the means ± SEM from at least three sets of experiments (each from a separate patient), and in each set, measurements were made in duplicate. At each time point, means without a common letter are significantly different (P < 0.05).  66  ACVR2A  a  0  a  a  a  ACVR2B mRNA  2.5 2.0 1.5 1.0 0.5 0.0  3 12 24 Time (h)  2.5 2.0 1.5 1.0 0.5 0.0  ACVR1B  a  0  a  a  b  Smad2 mRNA  D  3 12 24 Time (h)  2.5 2.0 1.5 1.0 0.5 0.0  Smad3  c  b a  0  a  Smad4 mRNA  F  3 12 24 Time (h)  (fold change relative to control) (fold change relative to control)  B  (fold change relative to control)  ACVR2A mRNA Smad7 mRNA  G  (fold change relative to control) (fold change relative to control)  ACVR1B mRNA Smad3 mRNA  E  (fold change relative to control)  C  (fold change relative to control)  A  2.5 2.0 1.5 1.0 0.5 0.0  2.5 2.0 1.5 1.0 0.5 0.0  2.5 2.0 1.5 1.0 0.5 0.0  ACVR2B  a  0  a  b  b  3 12 24 Time (h)  Smad2  a  0  a  ab  b  3 12 24 Time (h)  Smad4  a  0  a  a  a  3 12 24 Time (h)  Smad7  2.0 1.5 1.0  a  a  ab  b  0.5 0.0  0  3 12 24 Time (h)  FIG. 2.5. GDF9 pretreatment for 24 h increased cell response to activin A by regulating mRNA levels of ACVR2B/1B (B and C), Smad2/3 (D and E), and Smad7 (G). After 48 h preculture, hGL cells were cultured in low-serum media (containing only 0.5% FBS) and treated with 100 ng/ml GDF9 for up to 24 h. RNA of hGL cells were isolated and activin receptors and Smad mRNA levels were assessed by real-time PCR. Results were the means ± SEM from at least three sets of experiments (each from a separate patient), and in each set, measurements were made in triplicate. Means without a common letter are significantly different (P < 0.05).  67  30 min Control GDF9 BMPR2 ECD Activin A A P-Smad3 (Ser423/425) Smad3  +  + +  45 min +  + +  + + +  + -  + + -  + -  β-actin  B  P-Smad2 (Ser465/467) Smad2 β-actin  C  P-Smad2 (Ser245/250/255) Smad2 β-actin  FIG. 2.6. GDF9 pretreatment for 24 h transiently enhanced phosphorylation of Smad3 (Ser423/425) and Smad2 (Ser465/467) induced by activin A, effects that were neutralized by BMPR2 ECD. The culture and treatment conditions were similar to those in Fig. 2.3 except for the time scale of activin A treatment. Cell lysates were collected and assessed by Western blot analysis for the expression of phosphorylated Smad3 (Ser423/425) (A), phosphorylated Smad2 (Ser465/467) (B), and phosphorylated Smad2 (Ser245/250/255) (C). β-actin was used as the internal control. Results are representative of at least three sets of experiments (each from a separate patient).  68  3. EFFECTS  OF  ENDOGENOUS  GDF9  ON  ACTIVIN  A-INDUCED INHIBIN B ACCUMULATION IN HUMAN GRANULOSA-LUTEIN CELLS 2 3.1.  Introduction GDF9 was first identified in the early 1990s as a member of the TGFβ family (77). In mice,  GDF9 is expressed mainly in the oocytes of ovarian follicles (214) but is also expressed in testicular and hypophyseal tissues (228). Female mice deficient in GDF9 are infertile and fail to demonstrate any normal follicles beyond the primary one-layer follicle stage, whereas GDF9-deficient male mice are fertile (219). Recent studies have reported GDF9 mRNA expression in human cumulus (227, 269) and granulosa cells (226, 270) and not just in the oocyte, suggesting that GDF9 may have autocrine effects on granulosa cells as well. Treatment with activin A or GDF9 alone has been shown to induce βB subunit mRNA levels and inhibin B accumulation in hGL cells (173, 176). We have demonstrated recently that GDF9 pretreatment for 24 h can significantly enhance activin A-induced βB mRNA levels, inhibin B accumulation and Smad2/3 phosphorylation and can induce ACVR2B/1B and Smad2/3 but reduce Smad7 (an inhibitory Smad) mRNA levels (271). However, it was unknown whether endogenous GDF9 could exert similar autocrine effects. To characterize the effects of endogenous GDF9, we first confirmed the presence of GDF9 mRNA and protein expression in cultured hGL cells in addition to follicular fluid by real-time PCR and Western blot analysis. We then transfected hGL cells with GDF9-targeting small 2  A version of this chapter has been published. Shi FT, Cheung AP, Huang HF, Leung PC 2009 Effects of endogenous growth differentiation factor 9 on activin A-induced inhibin B production in human granulosa-lutein cells. The Journal of Clinical Endocrinology and Metabolism 94:5108-5116  69  interfering RNA (siRNA) to reduce endogenous GDF9 expression and evaluated corresponding changes, with and without activin A treatment, in inhibin βB-subunit mRNA and inhibin B levels and key components of the activin signaling pathway. Our study showed for the first time that reduced expression of endogenous GDF9 by specific siRNA in hGL cells can attenuate the stimulatory effects of activin A on inhibin βB-subunit mRNA levels and inhibin B secretion by modulating key components of the activin signaling pathway. Hence, GDF9 may be a critical autocrine regulator of inhibin subunit mRNA levels and inhibin B accumulation by the granulosa cells during the menstrual cycle.  3.2.  Materials and Methods  Preparation of hGL cells hGL cells from each patient were extracted as described in Section 2.2. Two milliliters of the corresponding follicular fluid supernatant were stored frozen for GDF9 protein detection with Western blot analysis; 2x105 viable cells were seeded per well in 12-well culture plates and cultured in the same conditions as described in Section 2.2. After 48 h, the culture media were removed, and the hGL cells were collected for RNA or protein extraction. Human GDF9 mRNA and protein levels were assessed by real-time PCR and Western blot analysis.  Knockdown analysis for human GDF9 We performed transient knockdown assays using 80 nM of GDF9-targeting siRNA or control nontargeting siRNA (ON-TARGETplus SMARTpool, Dharmacon Research, Inc., Lafayette, CO). After preculture of hGL cells for 48 h, the media were replaced by fresh culture media but without antibiotics and GDF9-targeting siRNA with Lipofectamine  70  RNAiMAX (Invitrogen, Carlsbad, CA), non-targeting siRNA (“control siRNA”) with Lipofectamine RNAiMAX, or RNAiMAX alone (“RNAiMAX”) was then added to the culture media. The cell culture, immediately after adding transfection reagents, was now designated as “Time 0 h” for all subsequent experiments described below. After 24 h (“Time 24 h”), RNAiMAX with or without siRNA in the spent media were replaced with fresh antibiotic-free culture media, and the hGL cells were cultured for another 24 hours (“Time 48 h”). The spent media were then replaced with the original culture media but containing only 0.5% FBS (“low-serum media”), and hGL cells were incubated for an additional 24 h (“Time 72 h”) to starve the cells under conditions identical to those in Section 2.2. mRNA levels of GDF9, and receptors and Smads of the activin signaling pathway in harvested cells were quantified with real-time PCR at 48 and 72 h after adding the transfection reagents. Corresponding protein levels were quantified with Western blotting at 72 h. In separate experiments, 25 ng/ml activin A (Sigma) was added to the media 72 h after transfection, and hGL cells were then collected to assess inhibin βB-subunit mRNA levels at 6, 12, and 24 h, or Smad2 and Smad3 phosphorylation at 30 and 45 min.  RNA extraction and Real-time PCR Total RNA extraction, first-strand cDNA synthesis and real-time PCR were performed as described in Section 2.2 using primers listed in Table S1 of the Appendices.  Western blot analysis Western blot analysis was performed as described in Section 2.2 with the following antibodies: antiphospho-Smad2 (Ser245/250/255), antiphospho-Smad2 (Ser465/467), antiphosphoSmad3 (Ser423/425), anti-Smad2, and anti-Smad3 (Cell Signaling Technology); anti-ACVR2B 71  and anti-ACVR1B (R&D Systems); anti-GDF9, anti-Smad4, anti-Smad7 and anti-β-actin (Santa Cruz Biotechnology Inc.). Particularly, we used the SuperSignal Femto West (Pierce Chemical Co., Rockford, IL) reagent which is more sensitive than normal ECL reagent to detect the signal of antibody bound to GDF9. In addition, Scion Image Analysis software (Scion Co., Frederick, MD) was used to determine protein density levels.  Inhibin B assays In experiments in which hGL cells were cultured for an additional 24 hours with or without 25 ng/ml of activin A, 72 h after siRNA transfection, the culture media were collected at 12 and 24 h and stored frozen until assay for inhibin B as described in Section 2.2.  Statistical analysis The methods used for data analysis and presentation of results were as described in Section 2.2.  3.3.  Results  GDF9 mRNA and protein levels and effects of GDF9-targeting siRNA transfection To determine whether cultured hGL cells were capable of expressing GDF9 mRNA, we used a set of real-time PCR primers to verify the identity of the amplicon by sequencing. The PCR primers spanned the single 1577-bp intron in the GDF9 gene and yielded a single PCR fragment of the appropriate size and sequence (data not shown). GDF9 mRNA and protein were detected in cultured hGL cells from all 11 patients tested (Fig. 3.1A, top and middle panels). GDF9 protein was also detected in all follicular fluid samples on Western blot analysis (Fig. 3.1A, bottom panel). Transfection with siRNA specific for human GDF9 significantly 72  decreased GDF9 mRNA levels at 24 h and 48 h (Fig. 3.1B) and GDF9 protein at 72 h (Fig. 3.1C) in hGL cells; as expected, transfection with control siRNA or RNAiMAX (transfection reagent only) showed no changes.  Effects of GDF9-targeting siRNA transfection on activin A-induced inhibin βB-subunit mRNA levels and inhibin B accumulations in hGL cells After GDF9-targeting siRNA transfection, inhibin α-, and βA-subunit mRNA levels showed no changes with RNAiMAX or control siRNA treatment at all time points (Fig. 3.2, A and B). Although activin A could induce inhibin α- and βA-subunit mRNA levels, changes reached statistical significance only at 24 h, and the increases were small. Although these levels were lower when cells were first transfected with GDF9-targeting siRNA before activin A treatment, the observed changes were not significantly different. In contrast, activin A induced significant increases in inhibin βB-subunit mRNA levels at 6, 12 and 24 h, effects that were significantly attenuated at all time points (P < 0.01; Fig. 3.2) when cells were first transfected with GDF9-targeting siRNA. Furthermore, the effects of GDF9-targeting siRNA were reversed by 100 ng/ml exogenous GDF9 (Fig. 3.3). As a comparison, RNAiMAX, control siRNA, or GDF9 siRNA alone did not induce inhibin βB-subunit mRNA levels. Correspondingly, activin A increased inhibin B protein concentration in the culture media by 3-fold of control value (P < 0.001), and its effect was attenuated significantly (by 32% at 24 h) with GDF9-targeting siRNA transfection (Fig. 3.4).  Effects of GDF9-targeting siRNA transfection on mRNA and protein levels of components involved in the activin signaling pathway in hGL cells When endogenous GDF9 activities were reduced following GDF9-targeting siRNA  73  transfection, levels of mRNA for ACVR2B/1B and Smad2/3 decreased, whereas those for Smad7 increased, with changes significantly different from cells treated with RNAiMAX or control siRNA at 72 h. A small decrease was also noted in Smad4 mRNA level which reached statistical significance at 72 h. These data affirm that endogenous GDF9 enhances cell response to activin A by increasing mRNA for facilitating components and simultaneously reducing that for inhibitory components of the activin A pathway (Fig. 3.5A) as demonstrated by our results from experiments with exogenous GDF9 (271). Corresponding changes in protein levels of these components were similarly observed (Fig. 3.5B).  Effects of GDF9-targeting siRNA transfection on activation of activin A-induced Smads Transfection with GDF9-targeting siRNA reduced the levels of phosphorylated Smad3 (Ser423/425) and Smad2 (Ser465/467) following treatment with activin A for 30 or 45 minutes (Fig. 3.6, A and B). However, because GDF9 siRNA also reduced total Smad3 and total Smad2 levels, the ratios of phosphorylated Smad to total Smad were either lower as for Smad3 (Ser423/425) or remained unchanged as for Smad2 (Ser465/467) (Fig. 3.6, A and B). In contrast, there were no changes in phosphorylation of Smad2 (Ser245/250/255) irrespective of activin A treatment and/or GDF9-targeting siRNA transfection (data not shown).  3.4.  Discussion  GDF9 was originally thought to be an oocyte-specific growth factor. However, recent studies have shown that GDF9 is present in granulosa cells from monkey (223), goat (224) and pig (225) ovaries, and granulosa and cumulus cells from human ovaries (226, 227, 269, 270). Our study confirmed that GDF9 mRNA and protein were indeed present in hGL cells from all 11 patients examined (Fig. 3.1A, top and middle panels). The availability of more sensitive 74  real-time PCR and Western blot analysis in current studies might explain why previous studies failed to identify its presence in granulosa cells (218, 228). In addition, GDF9 protein was also detected in all our follicular fluid samples by Western blot analysis (Fig. 3.1A, bottom panel), findings consistent with those recently reported by Huang et al. (270). Taken together, these findings raise the possibility of important paracrine and autocrine roles for GDF9 in regulating ovarian functions. Indeed, our study provided evidence for the first time of an autocrine role for endogenous GDF9 in regulating inhibin βB-subunit mRNA expression, and hence, inhibin B accumulation by human granulosa-lutein cells. When endogenous GDF9 activities in these cells were reduced following GDF9-targeting siRNA transfection, the mRNA level of GDF9 decreased significantly at 24 h and decreased further at 48 h (Fig. 3.1B), and protein level was reduced significantly at 72 h (Fig. 3.1C). Correspondingly, the effects of activin A on inhibin βB-subunit mRNA levels and inhibin B accumulations were diminished (Fig. 3.2-3.4). The lower inhibin βB-subunit mRNA levels and inhibin B accumulations induced by activin A in our present experiments relative to those observed in experiments with exogenous GDF9 pretreatment in our previous study (271) were caused by the additional transfection time of 48 h before activin A treatment: ~19.3-fold vs. ~54.8-fold increase, respectively at 24 h for inhibin βB-subunit; ~131.9 pg/ml vs. ~207 pg/ml, respectively, at 24 h for inhibin B. We have further characterized that heightened cell response to activin A in regulating inhibin βB-subunit mRNA expression is directly related to endogenous GDF9 mRNA expression. Thus, decreased endogenous GDF9 mRNA levels were associated with decreased ACVR2B/1B and Smad2/3/4 and increased Smad7 mRNA and protein levels in the activin signaling pathway (Fig. 3.5). Furthermore, decreased endogenous GDF9 activity was  75  associated with decreased activin A-induced phosphorylation of Smad3 (Ser423/425) and Smad2 (Ser465/467) (Fig. 3.6) but not Smad2 (Ser245/250/255) in hGL cells. Our findings with endogenous GDF9 mirror those reported in our previous study with exogenous GDF9 (271) in which GDF9 pretreatment for 24 h significantly enhanced activin A-induced inhibin βB-subunit mRNA levels, inhibin B accumulation, and Smad2/3 phosphorylation [effects attenuated by BMPR2 ECD, a well-known GDF9 antagonist (235, 259, 260)] and induced ACVR2B/1B and Smad2/3 but reduced Smad7 mRNA levels. The results of our current work demonstrate similar roles for endogenous GDF9 in regulating inhibin βB-subunit mRNA level and inhibin B accumulation and the cell signaling pathway involved and further affirm that GDF9 can activate phosphorylation of Smad2 on Ser465 and Ser467, while other ovarian regulators are required to activate Ser245/250/255. The synthesis and secretion of inhibin and activin dimers in the human ovary is dependent on the regulation of the three known inhibin/activin subunits that are controlled not only by endocrine hormones but also by local factors. Although circulating inhibin B profiles (256) and inhibin B concentrations in human follicular fluid have been determined (272), the regulation of inhibin B secretion in human granulosa cells at different stages in the menstrual cycle is unclear. Some studies have shown that TGFβ family ligands such as TGFβ, activin A, and BMP2 (173, 174, 177, 273) can stimulate mRNA levels of inhibin βB-subunit in cultured hGL cells but not inhibin α- and βA-subunits, whereas gonadotropins can up-regulate those of inhibin α- and βA-subunits (257). Indeed, our study shows that GDF9 can only influence activin A-induced inhibin βB-subunit mRNA levels and inhibin B accumulations but not inhibin α- and βA-subunit mRNA levels. Our study further affirms that Smad-dependent signaling pathway is involved in the regulation of inhibin βB-subunit mRNA levels in hGL  76  cells, which is consistent with the results of Bondestam et al. (274) and ours on the roles of exogenous GDF9 (271). Our current study with endogenous GDF9 further supports our proposal that increasing GDF9 expression during folliculogenesis (268) enhances human granulosa cell response to activin A, which leads to rising inhibin B levels in the follicular phase. Although it is tempting to suggest that with release of the oocyte after ovulation, a main source of GDF9 is removed (hence, the cell response to activin A with respect to inhibin βB-subunit mRNA is withdrawn) to explain the decline in inhibin B secretion after ovulation, the contribution of GDF9 in granulosa cells in this regard remains unknown. Because GDF9 mRNA levels in granulosa cells have been positively correlated with the number of dominant follicles observed but not the number of oocytes retrieved during ovarian stimulation for IVF treatment (270), the authors suggest that GDF9 in granulosa cells may play a role in FSH-dependent follicle maturation. This may explain, in part, the higher serum inhibin B levels observed during gonadotropin treatment in women with polycystic ovary syndrome (PCOS) than women with normal ovulatory function despite similar basal inhibin B levels (Cheung AP, University of British Columbia, Vancouver, Canada; unpublished data) because women with PCOS have a predilection for developing a large number of codominant follicles during ovarian stimulation. A previous study has found similar inhibin B levels in size-matched small antral follicles from unstimulated ovaries in women with PCOS and normal ovulatory women (275). In contrast, higher circulating basal inhibin B levels, but similar inhibin B levels during monofollicular ovarian response to FSH stimulation, in women with PCOS than normal ovulatory women have been reported (276). Thus, these temporal associations are far from clear. A recent study has found decreased GDF9 mRNA levels in developing oocytes from women with PCOS or  77  PCO (polycystic ovaries) compared with normal ovaries; the decreased levels are evident throughout folliculogenesis, particularly at the primary and secondary stages (268). During IVF/intracytoplasmic sperm injection treatment, women with PCOS have lower GDF9 mRNA levels in cumulus cells than women with normal menstrual cycles (227). However, changes in endogenous GDF9 activities in granulosa cells and its interactions with activin A and FSH in regulating inhibin βB-subunits and inhibin B at different stages of follicle development have not been characterized. To what extent results from granulosa-lutein cells obtained from women undergoing IVF treatment reflect normal ovarian physiology needs to be further evaluated because these cells have been exposed to pharmacological concentrations of exogenous gonadotropins and are in the process of luteinization from hCG stimulation. Further studies will be needed to determine more precisely when granulosa cell GDF9 expression begins, what regulates its expression, and what are the relative autocrine/paracrine contributions of granulosa cell vs. oocyte GDF9 from unstimulated ovaries. Finally, GDF9 mRNA has been detected in the human testis, uterus, and placenta and non-reproductive tissues such as bone marrow, adrenal gland, pituitary gland, and thymus (228), the rodent testis and hypothalamus (228), and the brushtail possum pituitary gland (230), suggesting that GDF9 may have actions not exclusive to the ovary. In summary, our work indicates that the endogenous GDF9 can regulate the mRNA and protein levels of activin receptors ACVR2B/1B and downstream signaling molecules Smad 2/3/4/7. These changes, in turn, increase cellular response to activin A stimulation in inhibin βB-subunit mRNA levels and inhibin B accumulations.  78  3.5.  Figures A MW 1 2 3 4 5 6 7 8  9 10 11 (-)  hGL cells mRNA  GDF9  hGL cells protein lysate  GDF9  Follicular fluid  GDF9  B  24 h  RNAiMAX Control siRNA GDF9 siRNA  48 h  24 h  + - - + - - + - - + - - + - - +  MW  (-)  (-) MW GAPDH  1.5  a  1.0  a  a b  0.5  b  0.0  24  48  Time (h)  + -  + -  +  GDF9 β-actin  GDF9 protein  C RNAiMAX Control siRNA GDF9 siRNA  RNAiMAX Control siRNA GDF9 siRNA  a  (fold change relative to RNAiMAX)  GDF9 mRNA  GDF9 (fold change relative to RNAiMAX)  48 h  + - - + - - + - - + - - + - - +  RNAiMAX Control siRNA GDF9 siRNA  1.5  a 1.0  a b  0.5 0.0  72 h  FIG. 3.1. GDF9-targeting siRNA reduced the GDF9 expression in hGL cells. A, GDF9 levels in hGL cells and follicular fluid. GDF9 mRNA was expressed in hGL cells from 11 patients by real-time PCR (top panel). GDF9 protein in hGL cells was represented by a major band at 51 kD on Western blot analysis, which was detected in all 11 patients tested (middle panel). GDF9 protein (51 kD) was similarly detected in follicular fluid from the same 11 patients (bottom panel). B, After preculture for 48h, hGL cells were transfected with 80 nM GDF9-targeting siRNA for 24 h and 48 h, and RNA from hGL cells was isolated, and GDF9 mRNA levels were assessed by real-time PCR. The upper panel shows the gel electrophoresis from one patient; the lower panel shows the effects of GDF9-targeting siRNA on GDF9 mRNA levels (mean ± SEM) from hGL cells of three patients. C, Forty-eight hours after adding GDF9-targeting siRNA, the culture media were replaced by low-serum media (containing 0.5% FBS), and hGL cells were cultured for one more day. hGL cell protein was isolated, and GDF9  79  protein was assessed by Western blot analysis. The left panel shows the immunoblots from one patient; the right panel shows the effects of GDF9-targeting siRNA on GDF9 protein levels (mean ± SEM) from hGL cells of the same three patients. At each time point, means without a common letter are significantly different (P < 0.05). MW, molecular weight standards; (-), no cDNA control.  80  C  1.5 b  1.0  a a a a a  βA subunit mRNA  (fold change relative to RNAiMAX)  b  a a a a a  a a a  0.5  0.0 6  B  βB subunit mRNA  2.0  12 Time (h)  24  (fold change relative to RNAiMAX)  α subunit mRNA  (fold change relative to RNAiMAX)  A  22.5  c  c b  12.5 b  c b  2.5 1 0  a a a  6  a a a  12 Time (h)  a a a  24  2.0 RNAiMAX Control siRNA GDF9 siRNA Activin A GDF9 siRNA+Activin A  1.5  1.0  a a a a a  b b  a a a  a  a  a a a  0.5  0.0 6  12 Time (h)  24  FIG. 3.2. GDF9-targeting siRNA diminished the effects of activin A on levels of inhibin βB-subunit B  mRNA in hGL cells. After preculture for 48h, hGL cells were transfected with 80 nM GDF9-targeting siRNA for 48h. The culture media were replaced by low-serum media (0.5% FBS), and hGL cells were cultured for one more day before treatment with 25 ng/ml activin A for 6, 12, and 24 h. RNA was isolated from hGL cells, and inhibin subunit mRNA levels were assessed by real-time PCR. Results are the means ± SEM from hGL cells of four patients. At each time point, means without a common letter are significantly different (P < 0.05).  81  βB subunit mRNA  (fold change relative to Control siRNA)  40 e  30  d c  20 b  10 1 0  a  a  a  a  Control siRNA Activin A GDF9 siRNA GDF9 siRNA+Activin A GDF9 GDF9+Activin A GDF9 siRNA+GDF9 GDF9 siRNA+GDF9+Activin A  24 Time (h)  FIG. 3.3. GDF9-targeting siRNA diminished the effects of activin A on levels of inhibin βB-subunit B  mRNA in hGL cells, and the effects of siRNA were reversed by exogenous GDF9 treatment. After preculture for 48 h, hGL cells were transfected with 80 nM GDF9-targeting siRNA for 48 h. The culture media were then replaced by low-serum media (0.5% FBS), and hGL cells were cultured with and without 100 ng/ml of GDF9 for another 24 h before incubating with 25 ng/ml activin A for a further 24 h period. RNA was then extracted from cells, and inhibin βB-subunit mRNA levels in the B  cells were measured by real-time PCR. Results are the means ± SEM from hGL cells of three patients. At each time point, means without a common letter are significantly different (P < 0.05).  82  Inhibin B (pg/mL)  200 c  150  b  c  100  b  50  a  a  a  a  a  a  0 12  24  Time (h) Activin A GDF9 siRNA+Activin A  RNAiMAX Control siRNA GDF9 siRNA  FIG. 3.4. GDF9-targeting siRNA reduced activin A-induced inhibin B accumulations in culture media. The culture conditions were same as those in Fig. 3.2. After treatments, media were collected and assayed for inhibin B levels by ELISA. Results are the means ± SEM from hGL cells of four patients. At each time point, means without a common letter are significantly different (P < 0.05).  83  B  b  0.5 0.0  48  72  Time (h)  1.5 1.0  a a a  a a b  0.5 0.0  48  Smad3 mRNA  Smad2  72  Time (h)  1.5 1.0  a a a  a a b  0.5 0.0  48  72  Time (h)  Smad7 mRNA  Smad4  + -  +  RNAiMAX Control siRNA GDF9 siRNA  ACVR1B  ACVR2B  1.5 1.0  a a a  a a  ACVR1B b Smad2  0.5 0.0  48  Smad3  72  Time (h)  Smad4  Smad3 1.5 1.0  Smad7 a a a  a a  b  β-actin  0.5 0.0  1.5  48  72  Time (h)  Smad7 2.5  b  2.0 1.5 1.0  a  a a  a a  0.5 0.0  Protein levels  a a  + -  (fold change relative to control)  a  ACVR1B mRNA  a a  (fold change relative to RNAiMAX)  1.0  (fold change relative to RNAiMAX)  ACVR2B 1.5  (fold change relative to RNAiMAX)  ACVR2B mRNA Smad2 mRNA  (fold change relative to RNAiMAX) (fold change relative to RNAiMAX)  Smad4 mRNA  (fold change relative to RNAiMAX)  A  b 1.0  aa  aa b  b  aa  b  a ab b a a  b  0.5  0.0  48  a a  72  ACVR2B ACVR1B Smad2  Smad3  Smad4  Smad7  Time (h)  RNAiMAX  Control siRNA  GDF9 siRNA  FIG. 3.5. GDF9-targeting siRNA decreased cell response to activin A by regulating mRNA levels for ACVR2B/1B, Smad2/3, and Smad7 in the activin signaling pathway. After preculture for 48 h, hGL cells were transfected with 80 nM GDF9-targeting siRNA for 48 h. The culture media were replaced by low-serum media (0.5% FBS), and hGL cells were cultured for one more day. A, mRNA levels of the relevant receptors and Smads of the activin signaling pathway were measured by real-time PCR after transfection with GDF9-targeting siRNA for 48 and 72 h. Results are the means ± SEM from hGL cells of four patients. B, Protein levels of the relevant receptors and Smads of the activin signaling pathway were measured by Western blot analysis after transfection with GDF9-targeting siRNA for 72 h. The upper panel shows the immunoblots from one patient; the lower panel shows the effects of GDF9-targeting siRNA on the protein levels of the relevant receptors and Smads of the activin signaling pathway. Results are the means ± SEM from hGL cells of three patients. At each time point, means without a common letter are significantly different (P < 0.05).  84  30 min Control siRNA GDF9 siRNA Activin A  A  + -  + +  45 min  + -  + +  + -  + +  + -  30 min + +  + -  + +  45 min  + -  + +  + -  + +  + -  + +  B  p-Smad3 (Ser423/425)  p-Smad2 (Ser465/467)  Smad3  Smad2 β-actin  c  2.0  b  b  1.5 1.0  a  a  a  a  a  0.5 0.0  30 min Total Smad3  1.5  1.0  45 min  a  a  a b  b  a b  b  0.5  0.0  30 min  45 min  relative to time-matched control)  2.5  Ratio of p-Smad2 to Smad2 (fold change relative to time-matched control)  Phosphorylation of Smad3 (Ser423/425)  Protein levels of Smad2 (fold change  relative to time-matched control)  Protein levels of Smad3 (fold change  Ratio of p-Smad3 to Smad3 (fold change relative to time-matched control)  β-actin  Phosphorylation of Smad2 (Ser465/467) 20  b  b  15  10 6 3  b a  b a  a  a  0  30 min Total Smad2  1.5  1.0  45 min  a  a  a b  b  a b  b  0.5  0.0  30 min  45 min  FIG. 3.6. GDF9-targeting siRNA decreased phosphorylation of Smad3 (Ser423/425) and Smad2 (Ser465/467) in hGL cells induced by activin A (25 ng/ml). After preculture for 48 h, hGL cells were transfected with 80 nM GDF9-targeting siRNA for 48 h, the culture media were replaced by low-serum media (0.5% FBS), and hGL cells were cultured for one more day before treatment with 25 ng/ml activin A for 30 and 45 min. Cell lysates were collected, and levels of Smad3 (Ser423/425) (A) and Smad2 (Ser465/467) phosphorylation were assessed by Western blot analysis (B). β-actin was used as the internal control. The top panel shows the immunoblots from one patient, and the middle panel shows the effects of GDF9-targeting siRNA on activin A-induced phosphorylated Smad3 (pSmad3) and phosphorylated Smad2 (pSmad2). Levels of pSmad3 and pSmad3 were normalized to the levels of total Smad3 (tSmad3) and total Smad2 (tSmad2), respectively, and are expressed as fold change relative to time-matched controls (control siRNA). The bottom panel shows tSmad3 and tSmad2 levels normalized to β-actin levels. Results are the means ± SEM from hGL cells of three patients. At each time point, means without a common letter are significantly different (P < 0.05).  85  4. GDF9 SUPPRESSES FOLLISTATIN AND FOLLISTATIN -LIKE  3  PROTEIN  ACCUMULATION  IN  HUMAN  GRANULOSA -LUTEIN CELLS3 4.1.  Introduction We have recently reported (271, 277) that exogenous and endogenous GDF9 can  significantly enhance activin A-induced inhibin βB-subunit mRNA levels by inducing activin receptors (ACVR2B/1B) and Smad2/3 but reducing Smad7 (an inhibitory Smad) mRNA levels in hGL cells. We have also confirmed GDF9 expression in hGL cells (277). However, its effects on extracellular regulators of activin such as FST and FSTL3 remain unknown. Activin A can promote FST production in undifferentiated and partially differentiated rat granulosa cells, but suppress FST production in fully differentiated granulosa cells (278). Activin A can cause a 1.8-fold rise in FST release in rat anterior pituitary cells, suggesting an autocrine/paracrine role of activin and FST in the pituitary (279). Activin A can also increase FST and FSTL3 mRNA and protein levels in the human hepatoma cell line HepG2 (280) but whether the same occurs in human granulosa cells is unknown. The objectives of the present work were to examine the effects of GDF9 on FST and FSTL3 expression, with and without activin A treatment, as potential mechanisms on its enhancing action on activin A-induced inhibin βB-subunit in hGL cells.  3  A version of this chapter will be submitted for publication. Shi FT, Cheung AP, Huang HF, Leung PC 2010 Growth  differentiation factor 9 (GDF9) suppresses follistatin and follistatin-like 3 protein production in human granulosa-lutein cells.  86  4.2.  Materials and Methods  Firstly, we compared FST and FSTL3 mRNA in hGL cells and protein in culture media with and without GDF9 treatment in time- and concentration-dependent experiments. Secondly, we explored the effects of GDF9 on activin A-induced FST and FSTL3 mRNA and protein levels. Thirdly, we transfected hGL cells with GDF9 siRNA to assess changes in basal and activin A-induced FST and FSTL3 mRNA and protein levels. Fourthly, we compared inhibin βB-subunit mRNA levels after activin A treatment with and without FST or FSTL3 siRNA. Finally, to further evaluate if the enhancing effect of GDF9 on activin A-induced inhibin βB-subunit mRNA is related to FST or FSTL3 expression, we measured these changes in activin A-treated hGL cells (with and without GDF9) at different concentrations of FST or FSTL3.  Preparation of hGL cells hGL cells from each patient were extracted and cultured as described in Section 2.2, and seeded as described in Section 3.2.  Activin A, GDF9, FST and FSTL3 experiments After preculture of hGL cells for 48 h, the low-serum medium containing 0.5% FBS instead of 10% FBS was added to each well and the cell culture was now designated as “Time 0 h” for all subsequent experiments described below. For time-dependent experiments, cells were treated with the concentration of 100 ng/ml of GDF9 for 12, 24 and 48 h. Cells were stimulated with 1-200 ng/ml GDF9 for 24 h in concentration-response studies. For experiments with both GDF9 and activin A, cells were preincubated with 100 ng/ml of GDF9 in low-serum media for 24 h before stimulation with 25 ng/ml of activin A. In neutralization experiments to 87  render GDF9 inactive, 2 µg/ml of BMPR2 ECD (R&D Systems) and 100 ng/ml of GDF9 were preincubated in low-serum media for 30 min before adding to cultured hGL cells. To test the effects of recombinant FST (288-amino acid FST; Peprotech) or FSTL3 (R&D Systems) on activin A-induced inhibin βB-subunit mRNA expression, different concentrations of FST or FSTL3 were preincubated with 25 ng/ml activin A at 37 C for 1 h in PBS containing 0.1% BSA before adding to hGL cells which were then cultured for 24 h.  Knockdown analysis for human GDF9, FST or FSTL3 We performed transient knockdown assays with 80 nM of GDF9, FST or FSTL3 siRNA using non-targeting siRNA (ON-TARGETplus SMARTpool; Dharmacon Research Inc.) as control. The conditions of siRNA transfection were same as described in Section 3.2. mRNA levels of FST or FSTL3 were then quantified with real-time PCR at 48 and 72 h after adding the transfection reagents. Corresponding protein levels were quantified with ELISA at 72 h after siRNA transfection. In separate experiments, the spent media at “Time 48 h” were replaced with low-serum media and hGL cells were incubated with and without 100 ng/ml GDF9 for one more day (“Time 72 h”), and FST or FSTL3 mRNA in cells and protein levels in culture media were quantified with real-time PCR and ELISA respectively.  RNA extraction and Real-time RT-PCR Total RNA extraction, first-strand cDNA synthesis and real-time PCR were performed as described in Section 2.2 using primers listed in Table S1 of the Appendices.  FST and FSTL3 assays After treatment, culture media were collected and stored frozen until assay with specific  88  ELISA kits (FST, Peprotech; FSTL3, R&D Systems); corresponding hGL cells were lysed with lysis buffer, and total cellular protein content determined as described earlier. The lowest limits of detection for FST and FSTL3 were 23 pg/ml, and 312.5 pg/ml, respectively. All hormone measurements were performed in duplicate. Secreted hormone levels were normalized to total cellular protein content.  Statistical analysis The methods used for data analysis and the descriptions of results were as described in Section 2.2.  4.3.  Results  Effects of GDF9 on FST and FSTL3 mRNA and protein levels GDF9 significantly decreased FST and FSTL3 mRNA levels in a time-dependent manner with maximum effects at 48 h (Fig. 4.1, A and B). GDF9 also reduced mRNA levels of FST and FSTL3 in a concentration-dependent manner which reached statistical significance at both the 100 and 200 ng/ml concentrations (P values all < 0.01) (Fig. 4.1, C and D). BMPR2 is the type 2 receptor for GDF9 and the ECD of BMPR2 is a well known GDF9 antagonist (235, 259, 260, 271, 277). When 100 ng/ml of GDF9 was preincubated with BMPR2 ECD for 30 min before adding to the cell culture, the inhibitory effects of GDF9 on FST and FSTL3 mRNA levels were attenuated (Fig. 4.1, C and D; P values all < 0.05). Correspondingly, GDF9 decreased FST and FSTL3 protein levels in a concentration-dependent manner and reached statistical significance at the 100 ng/ml concentration (P values all < 0.05); as a result, no significant decreases were observed when GDF9 was first neutralized with BMPR2 ECD (Fig.  89  4.1, E and F; P values all < 0.05). As expected, there were no significant changes in FST and FSTL3 mRNA levels relative to controls when BMPR2 ECD alone was added. Basal protein levels of FST in culture media were significantly higher than FSTL3 (Fig. 4.1, E and F, 4450 vs. 548 pg/ml, P < 0.001; Fig. 4.2, 4586 vs. 573 pg/ml, P < 0.001); correspondingly, the absolute decreases in FST accumulations following GDF9 treatment were greater than those of FSTL3 (Fig. 4.1, E and F, 1440 vs. 136 pg/ml at the GDF9 concentration of 200 ng/ml; P < 0.001).  Effects of GDF9 on activin A-induced FST and FSTL3 mRNA and protein levels Activin A increased both FST and FSTL3 mRNA levels (Fig. 4.3A; P values all < 0.001). In contrast, GDF9 suppressed basal and activin A-induced FST and FSTL3 mRNA, effects that were attenuated by BMPR2 ECD (Fig. 4.3A; P values all < 0.05). As noted earlier, BMPR2 ECD alone had no effects on FST and FSTL3 mRNA levels. Changes in FST and FSTL3 protein levels in culture media followed an identical pattern to changes in mRNA levels (Fig. 4.3B). However, FST mRNA peaked at 12 h while FSTL3 peaked at 24 h in response to activin A or activin A with GDF9 (Fig. 4.3A).  Effects of GDF9-targeting siRNA transfection on basal and GDF9-reduced FST and FSTL3 mRNA and protein levels When endogenous GDF9 levels decreased following GDF9 siRNA transfection (see Fig. 3.1B for details), there were significant increases in mRNA levels of FST and FSTL3 at 48 h and 72 h (Fig. 4.4A; P values all < 0.001), and corresponding proteins levels (Fig. 4.4B; P values all < 0.001). Furthermore, these effects of GDF9 siRNA were attenuated at 24 h after 100 ng/ml GDF9 was added to the culture (“Time 72 h” in Fig. 4.4; P values all < 0.001). As a 90  comparison, transfection with control siRNA showed no changes relative to transfection reagent only (“RNAiMAX”).  Effects of GDF9-targeting siRNA transfection on activin A-induced FST and FSTL3 mRNA and protein levels GDF9 siRNA transfection increased basal and activin A-induced FST mRNA levels at 12 h (P values all < 0.05) and 24 h in hGL cells (upper panel), and basal and activin A-induced FSTL3 mRNA at 12 and 24 h (lower panel; P values all < 0.05), effects that were attenuated after 100 ng/ml GDF9 was added to the culture (P values all < 0.05) (Fig. 4.5A). Corresponding changes in accumulations of FST and FSTL3 in the culture media showed a similar pattern to mRNA levels (Fig. 4.5B).  FST or FSTL3-targeting siRNA enhanced activin A-induced inhibin βB-subunit mRNA levels As expected, control siRNA, GDF9-, FST- or FSTL3-targeting siRNA in the absence of activin A had no effect on the basal inhibin βB-subunit mRNA levels. In the presence of activin A and consistent with our previous study (277), GDF9 treatment increased while GDF9 siRNA transfection decreased inhibin βB-subunit mRNA levels (Fig. 4.6B; P values all < 0.05). With reduced endogenous FST or FSTL3 expression (Fig. 4.6A; P values all < 0.001) after targeting siRNA transfection, activin A-induced inhibin βB-subunit mRNA levels increased from 15.7 to 27.7 fold for FST and 21 fold for FSTL3 (Fig. 4.6B; P values all < 0.05).  FST or FSTL3 reversed GDF9 enhanced effect in activin A-induced inhibin βB-subunit mRNA level To further confirm that GDF9 enhanced activin A-induced inhibin βB-subunit mRNA 91  levels by reducing FST or FSTL3 expression, we compared these changes in the presence of different concentrations of FST and FSTL3. We chose FST concentrations of 1, 2, 4, and 50 ng/ml and FSTL3 concentrations of 0.1, 0.2, 0.4, 4 and 50 ng/ml based on changes in FSH or FSTL3 protein levels following GDF9 treatment (Fig. 4.1E and F) and that FST or FSTL3 binds to activin in a 2:1 molar ratio (210, 212, 281, 282). Increasing concentrations of FST (1-4 ng/ml) or FSTL3 (0.1-0.4 ng/ml) attenuated activin A-induced inhibin βB-subunit mRNA levels B  with levels completely suppressed at the saturated concentration of 50 ng/ml for both (Fig. 4.7). In the presence of GDF9, inhibin βB-subunit mRNA levels decreased in a concentrationdependent manner for both FST and FSTL3 (Fig. 4.7).  4.4.  Discussion  We have recently reported that exogenous and endogenous GDF9 enhances activin A-induced expression of inhibin βB-subunit mRNA in hGL cells through modulation of activin receptors and key components of the intracellular signaling pathway (271, 277). Using the same hGL cell culture system in our current study, we have demonstrated for the first time that GDF9 can decrease not only basal but also activin A-induced mRNA and protein levels of FST and FSTL3, known extracellular inhibitors of activin (132, 186). These actions of GDF9 were supported by our experiments when GDF9 effects were neutralized by BMPR2 ECD or when endogenous GDF9 levels were reduced by targeting siRNA transfection. We therefore hypothesize that GDF9 decreases FST or FSTL3 expression which allows more activin A to bind to its receptors and hence, enhances activin A-induced inhibin βB-subunit mRNA levels. Our hypothesis is further supported by results from two additional experimental approaches. First, reduced endogenous FST or FSTL3 mRNA and protein levels after targeting siRNA  92  transfection augmented activin A-induced inhibin βB-subunit mRNA levels similar to those observed with GDF9 treatment (Fig. 4.6B). Second, the enhancing effects of GDF9 on activin A-induced inhibin βB-subunit mRNA levels were attenuated by exogenous FST or FSTL3 in a concentration-dependent manner (Fig. 4.7). Alternative precursor mRNA splicing produces two main isoforms of mature mammalian follistatin, a core protein of 315 amino acids (FST315) and a carboxy-truncated variant of 288 amino acids (FST288) (195, 204, 283). Although both isoforms have a similar binding affinity for activin (206), FST288 also has a high affinity for heparin (205, 206). In rat pituitary cells, complexes of activin and FST288 bind to cell surface proteoglycans via the heparin binding site of FST288 and is a mechanism by which activin is targeted for degradation (207). Although FST315 has the same affinity for activin as FST288, it is primarily present in the human circulation (213) and does not bind to heparin. Instead, FST315 acts as a storage for follistatin in the circulation, which delivers activin to target cells and prevents activin from binding to FST288, and hence, degradation. However, the actual function of FST315 is yet to be elucidated. We could not differentiate the roles of these isoforms because anti-human FST antiserum used in our follistatin assay was raised against a mixture of FST isoforms. The relative ratio of activin-free to activin-bound FST is essential in determining the bioavailability of activin and estimating the potential endocrine function of circulating FST but free follistatin assays are not commercially available at present. Because FST binds activins and inhibins through the common β-subunit (212), we cannot exclude the possibility that FST may also interact with inhibins. However, FST has a 500 to 1000 fold higher affinity for activins than inhibins (162). FSTL3, which shares several structural features with FST, does not have a heparin binding  93  sequence and therefore, does not bind cell-surface proteoglycans under normal conditions (284). Isoforms of FST are secreted faster than FSTL3 by stable, transfected CHO cells and the amount of newly synthesized FSTL3 localized in the nucleus is still substantial for up to 8 h, which is significantly longer than that for FST (211). This may explain partly why the protein concentration of FST in the media of our cultured hGL cells was about 8 fold higher than that of FSTL3 (Fig. 4.2, 4586 vs. 573 pg/ml, P < 0.001). Despite a similar decrease of FST or FSTL3 mRNA and protein levels 48 h after targeting siRNA transfection (Fig. 4.6A), activin A-induced inhibin βB-subunit mRNA levels were higher following FST siRNA (Fig. 4.6B; P < 0.05). Furthermore, 4 ng/ml FST had a stronger effect than the same concentration of FSTL3 in reversing activin A-induced inhibin βB-subunit mRNA in the presence of GDF9 although this difference did not reach statistical significance (Fig. 4.7). Whether this indicates that FST may play a more dominant role than FSTL3 in regulating activin A action requires further studies. FST protein levels were lower following FST siRNA transfection (from 4601 to 2544 pg/ml, Fig. 4.6A lower panel) than 100 ng/ml GDF9 treatment (from 4451 to 3266 pg/ml, Fig. 4.1E), but corresponding activin A-induced inhibin βB-subunit mRNA levels were lower after FST siRNA transfection than GDF9 pre-treatment (Fig. 4.6B). This difference in activin A-induced inhibin βB-subunit mRNA response may suggest that GDF9 acts through not just the extracellular (FST and FSTL3) but the intracellular (activin receptors and Smads) mechanisms as reported in our previous studies (271, 277). However, we cannot exclude the additional, although smaller effect of FSTL3 accounting for this difference. In addition, we also cannot rule out the possibility that GDF9 may affect other extracellular inhibitors such as the BMP and activin membrane-bound inhibitor (BAMBI). BAMBI is a transmembrane protein related to the TGFβ superfamily type I receptors. However, BAMBI lacks an intracellular kinase  94  domain and can block activin signaling by forming stable associations with activin type IB receptor (ACVR1B, also known as ALK4) but not activin type IA receptor (ACVR1A, also known as ALK2) (124). Higher FST levels observed in some women with polycystic ovary syndrome (PCOS) have led to the suggestion that altered FST function may contribute to the PCOS phenotype (285). However, an updated study on allelic variants of the follistatin gene in PCOS suggests that the contribution of the follistatin gene to the etiology of PCOS is small (286). In mouse models, over-expression of FST has also been shown to result in a PCOS-like phenotype (287). Decreased GDF9 mRNA levels have been found in developing oocytes from women with PCOS or polycystic ovaries compared to women with normal ovaries; the decreased levels are evident throughout folliculogenesis, beginning at recruitment initiation and continuing through the small, Graafian follicle stage (268). Although increased FST or FSTL3 expression with decreased endogenous GDF9 levels after targeting siRNA transfection (Fig. 4.4) may provide a mechanism by which altered GDF9 expression can affect follicle development, our granulosa cells were not specifically obtained from women with PCOS. Future studies comparing the interactions of FST or FSTL3 and GDF9 in granulosa cells from women with and without PCOS may shed new insight on the pathophysiology of this condition. While our granulosa cell culture systems had provided a convenient model to study the inter-relationships of FST or FSTL3 and GDF9, these cells were exposed to pharmacological concentrations of exogenous gonadotropins and were undergoing luteinization following hCG administration. However, in the absence of human granulosa cells from normal ovaries for research, findings from our cell culture systems do provide interesting hypotheses on the potential role of GDF9 in granulosa-lutein cell functions.  95  In summary, our previous studies suggest that GDF9 enhances activin A-induced inhibin βB-subunit mRNA levels in hGL cells by regulating receptors and crucial intracellular components of the activin signaling pathway. Our current study shows that GDF9 can decrease FST and FSTL3 expression in addition, which then allows more free activin A to bind to its receptors and activate the signaling pathway downstream. Whether this extracellular mechanism is instrumental to explain our previous findings or whether GDF9 also has an intracellular regulating role requires further research.  96  Figures  FSTL3 mRNA  1.5 c  1.0  b a  a  0.5 0.0 0  12  24  48  Time (h)  ECD  1.5 1.0  b  b b a  0.5  FSTL3 mRNA  FST mRNA  1.5 c  1.0  bc  ab  b  a  0.0 0 10 100200  0 100  GDF9 (ng/ml)  E  0.5 0.0 0  12  24  48  Time (h)  ECD  1.5 1.0  b  b  b a  b  a  0.5 0.0 0 10 100200  0 100  GDF9 (ng/ml)  F ECD  ECD  6000  800 c  4000  c  bc  FSTL3 (pg/mL)  FST (pg/ml)  a  D (fold change relative to vehicle)  C  (fold change relative to vehicle)  FST mRNA  B (fold change relative to vehicle)  A  (fold change relative to vehicle)  4.5.  c  ab a  2000  600  c  bc  c ab  c  a  400 200  0  0  0 10 100200  0 100  0 10 100200  GDF9 (ng/ml)  0 100  GDF9 (ng/ml)  FIG. 4.1. Time- (A and B) and concentration- (C and D) dependent effect of GDF9 on FST and FSTL3 mRNA levels in hGL cells and corresponding concentration-response in protein levels (E and F). After 48 h preculture, the culture media were replaced with low-serum media (0.5% FBS); hGL cells were then treated with 100 ng/ml GDF9 for 12 h (“Time 12 h”), 24 h (“Time 24 h”) and 48 h (“Time 48 h”) in time-dependent experiments (A and B), or with different concentrations of GDF9 for 24 h in concentration-dependent experiments (C, D, E and F). In neutralization experiments to render GDF9 inactive, 2 µg/ml of BMPR2 ECD (“ECD”) and 100 ng/ml of GDF9 were preincubated in low-serum media for 30 min before adding to cultured hGL cells. FST and FSTL3 mRNA levels in hGL cells and protein concentrations in culture media were assessed by real-time PCR and ELISA, respectively. Results were the means ± SEM from at least four sets of experiments (each from a separate patient), and in each set, measurements were made in triplicate for real-time PCR or duplicate for ELISA. Means without a common letter are significantly different (P < 0.05).  97  FST and FSTL3 (pg/ml)  6000  b  4000 2000 800  a  400 0 FST  FSTL3  FIG. 4.2. Comparison of FST and FSTL3 protein levels in culture media of hGL cells. After 48 h preculture, hGL cells were cultured in low-serum media (0.5% FBS) for another 24 h. Culture media were then collected and protein concentrations of FST and FSTL3 were assessed by ELISA. Results were the means ± SEM from at least six sets of experiments (each from a separate patient), and in each set, measurements were made in duplicate for ELISA. Means without a common letter are significantly different (P < 0.05).  98  3  c  c  10000  c 2  c  b b  1  a  a a  a  bb  b a  FST (pg/ml)  FST mRNA  B (fold change relative to vehicle)  A  d  8000 6000  b  b  0  0 24  24  Time (h)  Time (h) 4  c  1200  c  c  3  c 2  1  b  c b  a  a a  a  b  bb  a  0  FSTL3 (pg/ml)  FSTL3 mRNA  b  a  4000 2000  12 (fold change relative to vehicle)  d c  c b  800  ab ab  ab a 400  0 12  24  24  Time (h)  Time (h) Vehicle GDF9 Activin A GDF9+Activin A  GDF9+ECD+Activin A GDF9+ECD ECD  FIG. 4.3. GDF9 reversed activin A-induced FST and FSTL3 mRNA (A) and protein (B) levels, effects attenuated by BMPR2 ECD (“ECD”). After 48 h preculture, hGL cells were incubated in low-serum media (0.5% FBS) with and without 100 ng/ml of GDF9 for another 24 h before stimulation with 25 ng/ml of activin A for 12 h (“Time 12 h”) and 24 h (“Time 24 h”). The neutralization experiments with BMPR2 ECD and GDF9 were as described in Fig. 4.1. FST and FSTL3 mRNA levels in hGL cells and protein concentrations in culture media were assessed by real-time PCR and ELISA, respectively. Results were the means ± SEM from at least four sets of experiments (each from a separate patient), and in each set, measurements were made in triplicate for real-time PCR or duplicate for ELISA. At each time point, means without a common letter are significantly different (P < 0.05).  99  1.0  c  b a  6000  b  b a  0.5  FST (pg/ml)  1.5  0.0  c b  b  4000  a  2000  0  48 h 1.5  1000  b  a a  72 h  72 h  c ab  ab  1.0  a  0.5  0.0  FSTL3 (pg/ml)  FSTL3 mRNA  (fold change relative to RNAiMAX)  FST mRNA  B (fold change relative to RNAiMAX)  A  800 600  b  b a  400 200 0  48 h  72 h  72 h  Control siRNA  GDF9 siRNA  Control siRNA+GDF9  GDF9 siRNA+GDF9  FIG. 4.4. GDF9-targeting siRNA increased mRNA (A) and protein (B) levels of FST and FSTL3, effects that were reversed by adding exogenous GDF9. After preculture for 48 h, hGL cells were transfected with 80 nM GDF9 targeting siRNA for 48 h (“Time 48 h”). The culture media were then replaced by low-serum media (0.5% FBS), and hGL cells were cultured with and without 100 ng/ml GDF9 for another 24 h (“Time 72 h”). FST and FSTL3 mRNA levels in hGL cells and protein concentrations in culture media were assessed by real-time PCR and ELISA, respectively. Results were the means ± SEM from at least four sets of experiments (each from a separate patient), and in each set, measurements were made in triplicate for real-time PCR or duplicate for ELISA. At each time point, means without a common letter are significantly different (P < 0.05).  100  A  B + GDF9  - GDF9  - GDF9  c cd  2  bc b  ab  b ab  a  a  e  8000  bc  c  bc a  + GDF9  10000  d  1  24 h  + GDF9  a  FST (pg/ml)  3  - GDF9 e  24 h  d  cd  c  c  6000  b  b a  4000 2000 0  0  e d c  2  cd bc  c  ab  b  bc  bc 1  1500  d  ab  ab  a  a  a  FSTL3 (pg/ml)  FST mRNA  (fold change relative to RNAiMAX)  FSTL3 mRNA  3  (fold change relative to RNAiMAX)  12 h  d  1250  c  1000 750  c b  ab  b a  500  ab  250 0  0  Control siRNA GDF9 siRNA  + + - - - + +  ++ - - - ++  ++ - - - ++  ++ - - - ++  Control siRNA GDF9 siRNA  + -  + -  +  +  + -  + -  +  +  +Activin A  - Activin A  FIG. 4.5. GDF9-targeting siRNA increased activin A-induced mRNA (A) and protein (B) levels of FST and FSTL3, effects that were reversed by adding exogenous GDF9. After preculture for 48 h, hGL cells were transfected with 80 nM GDF9 targeting siRNA for 48 h. The culture media were then replaced by low-serum media (0.5% FBS), and hGL cells were cultured with and without 100 ng/ml GDF9 for another 24 h period before stimulation with 25 ng/ml of activin A for 12 h (“Time 12 h”) and 24 h (“Time 24 h”). FST and FSTL3 mRNA levels in hGL cells and protein concentrations in culture media were assessed by real-time PCR and ELISA, respectively. Results were the means ± SEM from at least four sets of experiments (each from a separate patient), and in each set, measurements were made in triplicate for real-time PCR or duplicate for ELISA. At each time point, means without a common letter are significantly different (P < 0.05).  101  1.5  b  b  b  b  FSTL3 mRNA  1.0  a a  0.5  0.0 24  48  Time (h) 6000  1.5  a 1.0  b  a  b  a a  0.5  0.0 24  48  Time (h) 800  FSTL3 (pg/ml)  b  b  FST (pg/ml)  (fold change relative to RNAiMAX)  FST mRNA  (fold change relative to RNAiMAX)  A  4000  a 2000  b  600  RNAiMAX  b  Control siRNA  400  FST siRNA  a  FSTL3 siRNA  200  0  0  48  48  Time (h)  Time (h)  - Activin A  f e  30 d  20  c b  10 2 1 0  a  a  R  FS N A T FS si R TL N A 3 si R N A C C on o n tr tr o ol ls si iR R N N A A +G G D D F9 F9 si FS RN A T FS si T L RN A 3 si RN A  a  si  F9  ls iR  N A  a  D  tr o on C  + Activin A  40  G  Inhibin β B subunit mRNA  (fold change relative to RNAiMAX)  B  FIG. 4.6. GDF9-targeting siRNA transfection attenuated while FST or FSTL3-targeting siRNA transfection enhanced activin A-induced inhibin βB-subunit mRNA levels. A, after preculture for 48 h, B  hGL cells were transfected with 80 nM of non targeting siRNA (“Control siRNA”) and FST- or FSTL3-targeting siRNA for 24 h (“Time 24 h”) and 48 h (“Time 48 h”), FST and FSTL3 mRNA levels in hGL cells (upper panel) and protein concentrations in culture media (lower panel) were assessed by real-time PCR and ELISA, respectively. Results were the means ± SEM from at least four sets of experiments (each from a separate patient), and in each set, measurements were made in triplicate for real-time PCR or duplicate for ELISA. At each time point, means without a common letter are  102  significantly different (P < 0.05). B, after preculture for 48 h, hGL cells were transfected with 80 nM of control siRNA and GDF9-, FST- or FSTL3-targeting siRNA for 48 h. The culture media were then replaced by low-serum media (0.5% FBS), and hGL cells were cultured with and without 100 ng/ml GDF9 for another 24 h before treatment with 25 ng/ml activin A for 24 h. Inhibin βB-subunit mRNA B  levels in hGL cells were measured by real-time PCR. Results were the means ± SEM from at least four sets of experiments (each from a separate patient), and in each set, measurements were made in triplicate for real-time PCR. Means without a common letter are significantly different (P < 0.05).  103  FSTL3  Inhibin β B subunit mRNA  (fold change relative to vehicle)  FST e  70  de de  60  de cd  50  c  bc bc  40 30 6 3 0  d  bc bc  Vehicle  bc  bc  1  2 4 50 FST (ng/ml)  4 50 FSTL3 (ng/ml)  0.1 0.2 0.4  Activin A  GDF9  b  aa  aa  aa  0  bc  b  GDF9+Activin A  FIG. 4.7. GDF9 pretreatment enhanced activin A-induced inhibin βB-subunit mRNA levels in hGL cells, B  effects that were attenuated by FST or FSTL3 in a concentration-dependent manner. After 48 h preculture, the culture media were replaced by low-serum media (0.5% FBS), and hGL cells were cultured with and without 100 ng/ml GDF9 for another 24 h in low-serum media. Different concentration of FST or FSTL3 were preincubated with 25 ng activin A at 37 C for 1 h in 50 μl PBS containing 0.1% BSA before adding to the cells and cultured for another 24 h. Inhibin βB-subunit B  mRNA levels in hGL cells were measured by real-time PCR. Results were the means ± SEM from at least four sets of experiments (each from a separate patient), and in each set, measurements were made in triplicate for real-time PCR. Means without a common letter are significantly different (P < 0.05).  104  5. GDF9  REVERSES  ACTIVIN  A  SUPPRESSION  OF  STEROIDOGENIC ACUTE REGULATORY PROTEIN (STAR) EXPRESSION AND PROGESTERONE ACCUMULATION IN HUMAN GRANULOSA-LUTEIN CELLS 4 5.1.  Introduction Activin A has been shown to decrease progesterone production in human granulosa cells  of follicles ≤ 5 mm from women undergoing oophorectomy (effects attenuated by inhibin A) (288) and inhibit hCG-induced progesterone secretion in hGL cells (289). In contrast, GDF9 can induce StAR expression and progesterone production in mouse mural granulosa cells with (26, 252) and without FSH (25). GDF9 can also increase basal but decrease FSH-induced progesterone in rat granulosa cells from small antral and preovulatory follicles (248). Unlike in rodents, GDF9 inhibits basal and 8-Br-cAMP-stimulated StAR and P450scc expression and progesterone release in hGL cells (226). We have reported recently (271, 277) that exogenous and endogenous GDF9 can enhance activin A-induced inhibin B secretion in hGL cells. However, the interactions of GDF9 with activin A on StAR and ovarian steroidogenic enzyme expressions and progesterone accumulation are unknown, and there is little information on the effect of inhibin B on progesterone accumulation in granulosa cells. We hypothesized that GDF9, by increasing inhibin B levels, with inhibin B acting as an activin A competitor, could reverse activin A 4  A version of this chapter has been accepted and will be published. Shi FT, Cheung AP, Christian Klausen, Huang HF, Leung  PC 2010 Growth differentiation factor 9 (GDF9) reverses activin A suppression of steroidogenic acute regulatory protein (StAR) expression and progesterone production in human granulosa-lutein cells. The Journal of Clinical Endocrinology and Metabolism. Authors (Shi FT and Cheung AP) have equal contributions at this chapter.  105  suppression of progesterone accumulation. We therefore undertook the following study to further evaluate GDF9 in this regard.  5.2.  Materials and Methods  We first compared mRNA and protein levels of StAR, P450scc, and 3βHSD in hGL cells and progesterone accumulation in culture media with and without GDF9 and/or activin A treatment. Second, we transfected hGL cells with inhibin α-subunit siRNA to assess changes in activin A, inhibin A and B levels, StAR, P450scc and 3βHSD expression, and progesterone accumulation following activin A treatment with and without GDF9. Finally, we tranfected hGL cells with GDF9 siRNA to evaluate the effects of activin A on StAR, P450scc, and 3βHSD expression and progesterone accumulation after endogenous GDF9 expression was reduced.  Preparation of hGL cells hGL cells from each patient were extracted and cultured as described in Section 2.2, and seeded as described in Section 3.2.  Activin A and GDF9 experiments The experimental conditions were same as described in Section 2.2 except for experiments with both GDF9 and activin A, cells were preincubated with 100 ng/ml of GDF9 in low-serum media for 24 h before activin A treatment with or without 0.2 IU/ml of recombinant human FSH (Gonal-F; Industria Farmaceutica Serono SPA, Bari, Italy) for 12, 24, 36 and 48 h.  106  Knockdown analysis for human inhibin α-subunit or GDF9 We performed transient knockdown assays with 80 nM of inhibin α-subunit or GDF9 siRNA using non-targeting siRNA (ON-TARGETplus SMARTpool; Dharmacon Research Inc.) as control. The conditions of siRNA transfection were same as described in Section 3.2. mRNA levels of inhibin α-subunit or GDF9 were then quantified with real-time PCR at 24 and 48 h after adding the transfection reagents. In separate experiments, the spent media at “Time 48 h” were replaced with low serum media and hGL cells were incubated with and without 100 ng/ml GDF9 for one more day (“Time 72 h”) before incubating with 25 ng/ml activin A for 24 and 48 h. Corresponding activin A, inhibin A and inhibin B levels in culture media were quantified with ELISA at 48 h after activin A treatment.  RNA extraction and Real-time RT-PCR Total RNA extraction, first-strand cDNA synthesis and real-time PCR were performed as described in Section 2.2 using primers listed in Table S1 of the Appendices.  Western blot analysis Western blot analysis was performed as described in Section 2.2 with the following antibodies: anti-StAR, anti-P450scc, anti-3βHSD and anti-β-actin (Santa Cruz Biotechnology Inc.). Scion Image Analysis software (Scion Co., Frederick, MD) was used to determine protein density levels.  Hormone assays After treatment, culture media were collected and stored frozen until hormone assays. The ELISA kits specific for progesterone (Cayman Chemical Co., Ann Arbor, MI), activin A (R&D 107  Systems), and inhibin A and B (Diagnostic Systems Laboratories) were used to measure the corresponding hormone levels in the culture media. The assays for inhibin A and inhibin B ELISA were as described in Section 2.2. The lowest limits of detection for progesterone was 7.8 pg/ml. Corresponding intrassay and interassay coefficients of variation were 7.3% and 7.7%, respectively. All progesterone measurements were performed in duplicate. Secreted progesterone levels were normalized to total cellular protein content.  Statistical analysis The methods used for data analysis and the descriptions of results were as described in Section 2.2.  5.3.  Results  Interactions of GDF9 and activin A on StAR expression and progesterone accumulation Activin A but not GDF9 treatment alone significantly reduced mRNA of StAR in a time-dependent manner (Fig. 5.1A). Pretreatment with GDF9 for 24 h reduced these suppressive effects of activin A in a time-dependent manner with maximum reversal at 48 h – from 0.33 fold with activin A treatment alone to 0.68 fold with GDF9 pretreatment (Fig. 5.1A; P < 0.01). As expected, when GDF9 was preincubated with BMPR2 ECD, a GDF9 antagonist (235, 259, 260, 271), the reversing effects of GDF9 on activin A suppression of StAR mRNA were attenuated and were statistically significant at 36 and 48 h (Fig. 5.1B, left panel; P < 0.01). In contrast, there were no significant changes in StAR mRNA levels with BMPR2 ECD alone or GDF9 and BMPR2 ECD (Fig. 5.1B, right panel). Compared to StAR, all the above treatments had no significant effects on corresponding mRNA levels of P450scc and 3βHSD  108  (Fig. 5.2). In parallel with changes in StAR mRNA levels, GDF9 pretreatment also attenuated activin A suppression of StAR protein levels from 0.58 fold (activin A alone) to 0.86 fold (activin A with GDF9) at 48 h (P < 0.05), effects that were neutralized by BMPR2 ECD (P < 0.05) (Fig. 5.3B). The effects of GDF9 on activin A effect were also assessed in the presence of FSH (Fig. 5.3). As expected, FSH, a known enhancer of StAR expression in granulosa cells, increased basal StAR protein levels almost 2-fold (Fig. 5.3B; P < 0.001). Activin A decreased FSH-stimulated StAR protein levels (P < 0.001), and again, its effects were attenuated by GDF9 pretreatment (P < 0.001) – 1.9 fold, 1.08 fold, and 1.76 fold for FSH alone, FSH with activin A, and FSH with activin A plus GDF9, respectively (Fig. 5.3B). FSH also induced a small rise in P450scc protein levels (P < 0.05) but had no effects on 3βHSD protein levels; neither Activin A, GDF9 nor combination had any effects on P450scc and 3βHSD protein expression (Fig. 5.4). Corresponding to the changes in StAR mRNA and protein levels, activin A but not GDF9 significantly decreased basal and FSH-stimulated progesterone accumulation (Fig. 5.5; P values all < 0.001). GDF9 pretreatment reduced the suppressive effects of activin A on basal and FSH-stimulated progesterone accumulation by 0.18 fold and 0.47 fold respectively but reached statistical significance only for the latter (Fig. 5.5; P < 0.001). As expected, these GDF9 effects were neutralized by BMPR2 ECD and the effect was statistically significant when progesterone accumulation was amplified by FSH (P < 0.001).  Effects of GDF9 and activin A on StAR expression and progesterone accumulation following inhibin α-subunit siRNA Because inhibin is a heterodimer of α- and β-subunits, and activin A a homodimer of 109  β-subunits, inhibin and activin production can be affected by the ratio of α to β subunits. After inhibin α-subunit siRNA transfection, endogenous inhibin α-subunit mRNA levels decreased as expected (Fig. 5.6A; P < 0.001); activin A levels in culture media increased by 0.9 fold from 16.4 to 31.5 pg/ml (Fig. 5.6B top panel; P < 0.001); inhibin A levels, basal or induced by activin A with and without GDF9 decreased but changes were not statistically significant (Fig. 5.6B middle panel); and, inhibin B levels induced by activin A alone or activin A with GDF9 decreased (Fig. 5.6B bottom panel; P values all < 0.05). Correspondingly, FSH-stimulated StAR mRNA levels decreased (columns 2 vs. 6 at 24 h and at 48 h, Fig. 5.7A) but reached statistical significance at 48 h only (P < 0.05); activin A suppression of FSH-stimulated StAR mRNA levels was enhanced but changes were not statistically significant (columns 3 vs. 7 at 24 h and at 48 h, Fig. 5.7A); the reversing effects of GDF9 on activin A suppression of FSH-induced StAR mRNA levels were attenuated, with maximum effect at 48 h (columns 5 vs. 9 at 24 h and at 48 h, Fig. 5.7A; P < 0.01). After inhibin α-subunit siRNA transfection, corresponding changes in StAR protein levels (Fig. 5.7B) and progesterone accumulations in the culture media (Fig. 5.7C) showed a similar pattern to StAR mRNA levels. The small increase in P450scc mRNA expression induced by FSH at 48 h (Fig. 5.8A upper panel; P < 0.05) was not affected by inhibin α-subunit siRNA transfection. None of the above treatments altered 3βHSD mRNA levels (Fig. 5.8A lower panel). Changes in P450scc and 3βHSD protein levels followed the same pattern as their corresponding mRNA (Fig. 5.8B).  Effects of activin A on StAR expression and progesterone accumulation after GDF9 siRNA With reduced endogenous GDF9 expression (see Fig. 3.1B of Charpter 3) after GDF9 110  siRNA transfection, the suppressive effects of activin A on StAR mRNA and protein levels and progesterone accumulations in the culture media were enhanced and reached statistical significance at 48 h (Fig. 5.9). The reversing effects of GDF9 on these changes were also attenuated but results were not statistically significant (Fig. 5.9). GDF9 siRNA transfection had no effects on P450scc and 3βHSD mRNA and protein levels (Fig. 5.10).  5.4.  Discussion  Our results showed that GDF9 reversed activin A suppression of both basal and FSH-stimulated StAR mRNA and protein expression, and FSH-stimulated progesterone secretion. We also demonstrated increased activin A levels and decreased inhibin B levels with decreased FSH-stimulated progesterone accumulation after inhibin α-subunit siRNA transfection (Fig. 5.6B). Increased activin A levels have been likewise observed in inhibin α-subunit knockout mice (290). The effects of GDF9 in reversing activin A suppression of FSH-induced StAR expression and progesterone accumulation were also attenuated after inhibin α-subunit siRNA transfection (Fig. 5.7). Similarly, there was increased suppression of StAR expression and progesterone accumulation by activin A following decreased endogenous GDF9 activities with GDF9 siRNA transfection (Fig. 5.9). This paralleled the reduced suppression by activin A after exogenous GDF9 pretreatment described earlier, and was consistent with unchanged plasma progesterone levels throughout the estrus cycle in GDF9 mutant sheep (291). Taken together with our recent observations (271, 277) that GDF9 can enhance activin A-induced inhibin B accumulation in hGL cells, we hypothesize that inhibin B is involved in facilitating these changes. Activin A is a known inhibitor of luteinization as reported in human (292, 293), bovine  111  (294, 295), and rat (296) ovary studies. In the menstrual cycle, serum activin A levels are higher in the early follicular phase, midcycle and late luteal phase (164) while serum inhibin B levels peak in the midfollicular phase and just after ovulation (161). During ovarian stimulation for IVF, activin A levels remained unchanged after 7-8 days of FSH treatment but rose significantly with follicle maturity (297). Based on these observations and findings of our current and previous studies (271, 277), it is tempting to speculate that this increased activin A expression in the preovulatory follicle in the presence of GDF9 further enhances inhibin B expression to overcome the inhibitory effect of activin A on luteinization. To what extent this plays a physiological role remains unknown given the dominant effects of LH on luteinization. While expression of StAR was tightly coupled with changes with progesterone accumulation in all our experiments, corresponding changes were not observed in P450scc and 3βHSD activities. Despite the known effects of FSH in stimulating P450scc and 3βHSD activities (38, 60), such changes were not observed. One explanation is that P450scc and 3βHSD activities in these hGL cells obtained from women undergoing IVF treatment are already primed for progesterone production and unlike StAR protein, are not the rate-limiting steps in our study model. Indeed, these cells have been exposed to pharmacological concentrations of exogenous gonadotropins and in the process of luteinization. However, in the absence of easily accessible granulosa cells from unstimulated ovaries, they do provide a convenient model to generate hypothesis for further study during the follicular-luteal transition. Alternatively, while the StAR, P450scc and 3βHSD genes share several common promoter elements such as SF-1 (298), each also possesses unique elements (see review in (71)) that may respond to Activin A and/or GDF9 with different sensitivities and time courses. FSH and LH are known to stimulate StAR expression in the mammalian ovary  112  predominantly via PKA signaling (299), but factors such as estradiol (300) and IGF-I (301) can potentiate, while TGFβ1 (18, 302), BMP15 (23), and BMP 2/4/6/7 (20) can suppress the actions of gonadotropins. Indeed, our current study has confirmed activin A suppression of basal and FSH-induced StAR expression and progesterone accumulation as reported previously in hGL cells (288) and porcine granulosa cells (303). We have also confirmed that GDF9 alone has no effects on basal StAR expression and progesterone accumulation as reported previously in hGL cells (226) and rat granulosa cells (248). However, species differences exist as GDF9 induces basal progesterone production in cultured granulosa cells from preovulatory follicle of mice (26, 252). We cannot rule out that decreased levels of inhibin α-subunit mRNA after targeting siRNA transfection may lead to increased synthesis of activin B (βBβB) and/or activin AB (βAβB) until sensitive and specific activin B and activin AB assays are commercially available. In addition, follistatin is known to bind activin A with high affinity which prevents activin A from binding to its receptor (130). Because granulosa cells and luteal cells are the main source of follistatin in the ovary, we also cannot exclude a regulatory role for GDF9 on follistatin activities and are currently conducting studies in this regard. The regulation of granulosa cells in the follicular–luteal transition involves complex endocrine, paracrine and autocrine interactions of multiple factors which have not been completely mapped out. Although the physiological implications of our findings remain to be tested in granulosa cells from unstimulated ovaries, they, nevertheless, suggest that local interactions between GDF9, activin A, and inhibin B can influence the functional evolution of a mature follicle to a corpus luteum and provide potential targets for future studies to further characterize the physiological process of the ovarian and menstrual cycle, and treatment manipulation in reproductive medicine.  113  In conclusion, our results indicate that GDF9 has no direct effects on basal or FSH-stimulated StAR activity and progesterone accumulation in hGL cells. Rather, GDF9 reverses the suppressive effects of activin A on StAR and progesterone accumulation by enhancing the levels of inhibin B, which in turn, acts as an inhibitor of activin A. This is corroborated by an increase in activin A suppression of StAR expression and progesterone accumulation when endogenous GDF9 levels are reduced after GDF9 siRNA transfection, and a corresponding decrease in activin A-induced (with or without GDF9 treatment) inhibin B accumulations after inhibin α-subunit siRNA transfection. While we would like to hypothesize a feedback loop between activin A and inhibin B in granulosa cells, further studies are required to establish if inhibin B can indeed directly inhibit activin A accumulation and other mechanisms involved in antagonizing activin A action.  114  5.5.  Figures  StAR mRNA  (fold change relative to vehicle)  A 1.1  a  a  a  a  a  a  a  Vehicle Activin A GDF9 GDF9+Activin A GDF9+ECD+Activin A  a  0.8 a a  0.5  a a  a  a a  a  b ab ab  a  0.2 0  12  24 36 Time (h)  48  60  StAR mRNA  (fold change relative to vehicle)  B 1.5  1.0  b  c  d c a  b  b b  b  a a  a  a a  a  a a  a  a a  b  a  a  ab  a  0.5  0.0  a  c  b  a a  12  24  a a  36  48  Time (h) Vehicle Activin A GDF9  GDF9+Activin A GDF9+ECD+Activin A  12  24 36 Time (h)  48  Vehicle GDF9+ECD ECD  FIG. 5.1. GDF9 pretreatment reduced activin A suppression of StAR mRNA levels in hGL cells, effects that were attenuated by BMPR2 ECD (“ECD”), a GDF9 antagonist. After preculture for 48 h, the culture media were replaced with low-serum media (0.5% FBS) and hGL cells were preincubated with 100 ng/ml of GDF9 for 24 h in low-serum media before stimulation with 25 ng/ml of activin A for 12, 24, 36 and 48 h. In neutralization experiments to render GDF9 inactive, 2 µg/ml of BMPR2 ECD and 100 ng/ml of GDF9 were preincubated together for 30 min before adding to cultured hGL cells. RNA was isolated from hGL cells, and mRNA levels of StAR were assessed by real-time PCR. Results are the means ± SEM from hGL cells of five patients. Panel A shows the time-dependent changes of StAR mRNA levels for each treatment group. Panel B shows StAR mRNA level changes for the various treatments at each time point. In both panels, means without a common letter are significantly different (P < 0.05).  115  P450scc mRNA  (fold change relative to vehicle)  A 1.5  1.0  a a a a a  a a a a a  a a a a a  a a a a a  12  24  36  48  a  a  a  a a a  a  a a  a  a a  0.5  0.0  12  Time (h)  24 36 Time (h)  48  3βHSD mRNA  (fold change relative to vehicle)  B 1.5  1.0  a a a a a  a a a a a  12  24  a  a  a a a  a a  a a a  a a a  a a a  a  a a  a  a a  0.5  0.0  36  48  Time (h) Vehicle Activin A GDF9  GDF9+Activin A GDF9+ECD+Activin A  12  24 36 Time (h)  48  Vehicle GDF9+ECD ECD  FIG. 5.2. The effects of GDF9 and activin A on P450scc and 3βHSD mRNA levels in hGL cells. The culture and treatment conditions were same as those in Fig. 5.1. After treatment, RNA was isolated from hGL cells, and mRNA levels of P450scc (A) and 3βHSD (B) were assessed by real-time PCR. Results are the means ± SEM from hGL cells of five patients. At each time point, means without a common letter are significantly different (P < 0.05).  116  A FSH Activin A GDF9 BMPR2 ECD  -  + -  + -  + + -  + + +  + +  +  + -  + + -  + + -  + + + -  + + + +  + + +  + +  StAR β-actin  StAR Protein  (fold change relative to vehicle)  B - FSH  2.5  + FSH d  2.0  d  d  1.5 1.0  bc  b  b  bc  bc  + +  +  d  d  + +  +  c  bc  a  a  0.5 0.0  GDF9 BMPR2 ECD -  -  + -  + -  + +  -  - Activin A  -  + -  + -  + +  +Activin A  FIG. 5.3. GDF9 pretreatment reduced activin A suppression of basal and FSH-induced StAR protein levels in hGL cells, effects that were attenuated by BMPR2 ECD. The culture and treatment conditions were as described in Fig. 5.1 except that after GDF9 pretreatment, hGL cells were stimulated with activin A for 48 h with and without 0.2 IU/ml FSH. Cell lysates were then collected and assessed by Western blot analysis for protein levels of StAR. Panel A shows the immunoblots from one patient representative of similar results from other patients; panel B shows the protein levels of StAR normalized to those of β-actin. Results are the means ± SEM from hGL cells of five patients. Means without a common letter are significantly different (P < 0.05).  117  A FSH Activin A GDF9 BMPR2 ECD  -  + -  + -  + + -  + + +  + +  +  + -  + + -  + + -  + + + -  + + + +  + + +  + +  p450scc 3βHSD β-actin  P450scc Protein 3βHSD Protein  (fold change relative to vehicle)  (fold change relative to vehicle)  B - FSH  2.5  + FSH  2.0 1.5 1.0  b  b  b  b  ab  ab  ab  a  a  a  a  a  a  a  a  +  -  -  + -  + -  + +  + +  +  a  a  a  a  ab  ab  ab  a  a  a  a  a  a  -  -  + -  + -  + +  + +  0.5 0.0  2.5 2.0 1.5 1.0 0.5 0.0  GDF9 BMPR2 ECD  - Activin A  +Activin A  FIG. 5.4. The effects of GDF9 and activin A on P450scc and 3βHSD protein levels in hGL cells. The culture and treatment conditions were same as those in Fig. 5.3. After treatment, cell lysates were then collected and assessed by Western blot analysis for protein levels of P450scc and 3βHSD. Panel A shows the immunoblots from one patient representative of similar results from other patients; panel B shows the protein levels of P450scc and 3βHSD normalized to those of β-actin. Results are the means ± SEM from hGL cells of five patients. Means without a common letter are significantly different (P < 0.05).  118  Progesterone release  (fold change relative to vehicle)  2.5  - FSH  + FSH e  e  2.0  e 1.5  1.0  cd  bc  e  e  d  d  -  + + + + - - + + +  bcd cd abc  a  ab  -  + + + + - - + + +  0.5  0.0  GDF9 BMPR2 ECD  -  - Activin A  -  +Activin A  FIG. 5.5. GDF9 pretreatment reduced activin A-suppressed progesterone accumulations in culture media, effects that were attenuated by BMPR2 ECD. The culture and treatment conditions were as described in Fig. 5.3. Results are the means ± SEM from hGL cells of five patients. Means without a common letter are significantly different (P < 0.05).  119  Inhα subunit mRNA  (fold change relative to control siRNA)  A 1.5  Control siRNA Inhα siRNA  b  b  1.0  a a 0.5  0.0 24  48  Time (h)  B Activin A (pg/mL)  40  b  Control siRNA Inhα siRNA  30 20  a  10 0  Inhibin A (pg/mL)  400 300 200 100  abc  bc a  c abc  ab  0  Inhibin B (pg/mL)  400  c  300  100  b  b  200  a  a a  0  Vehicle  Activin A GDF9+Activin A  FIG. 5.6. After transfection of hGL cells with inhibin α-subunit siRNA, accumulations of activin A increased; accumulations of inhibin A and inhibin B decreased following treatment with activin A alone or activin A with GDF9 pretreatment. After preculture for 48 h, hGL cells were transfected with 80 nM inhibin α-subunit siRNA for 48 h. The culture media were replaced by low serum media (0.5% FBS), and hGL cells were cultured with and without 100 ng/ml of GDF9 for another 24 h before incubating with 25 ng/ml of activin A for a further 48 h period. Panel A shows inhibin α-subunit mRNA levels 24 and 48 h after adding the transfection reagents; panel B shows the levels of activin A, inhibin A and inhibin B in culture media 48 h after adding activin A. Results are the means ± SEM from hGL cells of four patients. Means without a common letter are significantly different (P < 0.05).  120  abc  bc bc  c  1.5 1.0  48 h c  c  ab  ab  b  c  bc  bc  StAR mRNA  24 h  2.0  a  bc  a  a  a  a  0.5  (fold change relative to vehicle)  StAR mRNA  (fold change relative to control siRNA)  A  2.0  24 h  48 h  a a  a a  1.5 1.0 0.5 0.0  0.0  Vehicle  Control siRNA + + + + + - - - - Inhα siRNA GDF9 - - - + +  - - - + + + + - - + +  + + + + + - - - - - - - + +  - - - + + + + - - + +  Control siRNA FSH  B  C + + -  + + + -  + + +  + + + +  + + -  + + + -  + + +  + + + +  Progesterone release  FSH Control siRNA + Inhα siRNA Activin A GDF9 StAR  StAR Protein  (fold change relative to control siRNA)  β-actin  2.0  e  e  1.5  de cd  1.0  bc  bc  2.0 f  a  ef  1.5 1.0  def  cde  cde  abc  bcd  ab  a  0.5 0.0  Control siRNA + Inhα siRNA GDF9 -  ab a  (fold change relative to control siRNA)  FSH+Activin A  + -  + -  + +  + +  + -  + -  + +  + +  0.5 0.0  Control siRNA + Inhα siRNA GDF9 -  + -  + -  + +  + +  + -  + -  + +  + +  FIG. 5.7. GDF9 reduced the suppressive effects of activin A on StAR expression and progesterone accumulation, effects that were attenuated when cells were transfected with inhibin α-subunit siRNA. The culture and treatment conditions were as described in Fig. 5.6 except that after GDF9 pretreatment, hGL cells were treated with activin A for 24 and 48 h with 0.2 IU/ml FSH in addition. Fig. 5.7A shows mRNA levels of StAR 24 and 48 h after adding activin A with and without FSH. The “inset” confirmed control siRNA had no effects on StAR mRNA levels when compared with transfection reagent alone (“Vehicle”). Fig. 5.7B shows the corresponding protein expression at 48 h: the upper panel shows the immunoblots from one patient representative of similar results from other patients; the lower panel shows the protein levels of StAR normalized to those of β-actin. Fig. 5.7C shows the progesterone accumulations in culture media 48 h after adding activin A. Results are the means ± SEM from hGL cells of six patients. At each time point, means without a common letter are significantly different (P < 0.05).  121  A 1.5  a a a a  1.0  b b b b  a a a a  a  B  48 h b b b b  a  FSH Control siRNA Inhα siRNA Activin A GDF9  -  + + -  + + + -  + + +  + + + +  + + -  + + + -  + + +  + + + +  p450scc 3βHSD  0.5  β-actin  1.5  1.0  a  a a a a  a a a a  a  a a a a  a a a a  0.5  P450scc Protein  0.0  0.0  - - - + + + + - - + +  + + + + + - - - - - - - + +  - - - + + + + - - + +  3βHSD Protein  Control siRNA + + + + + - - - - Inhα siRNA GDF9 - - - + +  Vehicle FSH FSH+Activin A  (fold change relative to control siRNA) (fold change relative to control siRNA)  P450scc mRNA  3βHSD mRNA  (fold change relative to control siRNA)  (fold change relative to control siRNA)  24 h  2.0 1.5 1.0  b  b  b  b  b  b  b  b  a  a  a  a  a  a  a  a  + -  + -  + +  + +  + -  + -  + +  + +  a  0.5 0.0  2.0 1.5  a  1.0 0.5 0.0  Control siRNA Inhα siRNA GDF9  + -  FIG. 5.8. The role of inhibin α-subunit siRNA in the interaction of GDF9 and activin A on P450scc and 3βHSD expression and progesterone accumulation in hGL cells. The culture and treatment conditions were as described in Fig. 5.7. Fig. 5.8A shows the corresponding mRNA levels of P450scc and 3βHSD 24 h and 48 h after activin A treatment. Fig. 5.8B shows the protein levels of P450scc and 3βHSD 48 h after activin A treatment: the upper panel shows the immunoblots from one patient representative of similar results from other patients; the lower panel shows the protein levels of P450scc and 3βHSD normalized to those of β-actin. Fig. 5.8C shows the progesterone levels in culture media 48 h after adding activin A. Results are the means ± SEM from hGL cells of six patients. At each time point, means without a common letter are significantly different (P < 0.05).  122  24 h  1.0  c  c  c  e  bc  de  - GDF9  1.5  1.0  - - - - - ++ - - - -  + GDF9  d  d  ++++ - - ++ ++++  d  cd  StAR Protein  0.0 ++ - - - ++ ++++  + -  + + -  + +  + + +  + +  + + +  StAR  a  ++ - - - ++ - - - -  + + -  β-actin  b  a  0.5  cdecd cde c  b  b  ab  (fold change relative to control siRNA)  Progesterone release  Control siRNA + GDF9 siRNA Activin A GDF9 -  1.5  Control siRNA GDF9 siRNA GDF9  C  B  48 h  (fold change relative to control siRNA)  StAR mRNA  (fold change relative to control siRNA)  A  - GDF9  1.0  e  de  de  cd  de bc  b 0.5  a  0.0  Control siRNA GDF9 siRNA  d  + GDF9  1.5  + -  + -  +  +  + -  + -  +  +  bc  b a  0.5  - Activin A 0.0  Control siRNA GDF9 siRNA  +Activin A + -  + -  +  +  + -  + -  +  +  FIG. 5.9. Transfection of hGL cells with GDF9 siRNA enhanced the suppressive effects of activin A on StAR expression and progesterone accumulation. The culture and treatment conditions were as described in Fig. 5.6. Fig. 5.9A shows mRNA levels of StAR 24 and 48 h after adding activin A. Fig. 5.9B shows the corresponding protein expression at 48 h: the upper panel shows the immunoblots from one patient representative of similar results from other patients; the lower panel shows the protein levels of StAR normalized to those of β-actin. Fig. 5.9C shows the progesterone levels in culture media 48 h after adding activin A. Results are the means ± SEM from hGL cells of six patients. At each time point, means without a common letter are significantly different (P < 0.05).  123  24 h  B  48 h  1.5  1.0  a a a a  a a a a  a a a a  a a a a  Control siRNA + GDF9 siRNA Activin A GDF9 -  + + -  + -  + + -  + +  + + +  + +  + + +  p450scc 3βHSD  0.5  β-actin  1.5  1.0  a a a a  a a a a  a a  a  a  a a a a  0.5  P450scc Protein  0.0  Control siRNA GDF9 siRNA GDF9  ++ - - - ++ - - - -  ++ - - - + + +++ +  - Activin A  - - - - - ++ - - - -  ++++ - - ++ ++++  +Activin A  3βHSD Protein  0.0  (fold change relative to control siRNA) (fold change relative to control siRNA)  P450scc mRNA  3βHSD mRNA  (fold change relative to control siRNA) (fold change relative to control siRNA)  A  - GDF9  1.5  1.0  a  + GDF9  a  a  a  a  a  a  a  a  a  a  a  a  a  a  + -  +  +  + -  + -  +  +  0.5  0.0  1.5  1.0  a  0.5  0.0  Control siRNA GDF9 siRNA  + -  FIG. 5.10. The role of GDF9 siRNA in the interaction of GDF9 and activin A on P450scc and 3βHSD expression and progesterone accumulation in hGL cells. The culture and treatment conditions were as described in Fig. 5.9. Fig. 5.10A shows the corresponding mRNA levels of P450scc and 3βHSD 24 h and 48 h after activin A treatment. Fig. 5.10B shows the protein levels of P450scc and 3βHSD 48 h after activin A treatment: the upper panel shows the immunoblots from one patient representative of similar results from other patients; the lower panel shows the protein levels of P450scc and 3βHSD normalized to those of β-actin. Fig. 5.10C shows the progesterone accumulations in culture media 48 h after adding activin A. Results are the means ± SEM from hGL cells of six patients. At each time point, means without a common letter are significantly different (P < 0.05).  124  6. CONCLUSION AND RECOMMENDATIONS FOR FUTURE WORK 6.1.  Conclusion The main objective of this thesis was to explore the relationship between GDF9 and  activin A as it relates to the expression of inhibin βB-subunit mRNA, the accumulations of inhibin B and progesterone in human granulosa-lutein cells from women undergoing IVF and the signaling mechanisms involved. Several results provide information about these objectives. First, activin A and GDF9 alone had little stimulatory effect on inhibin α- and βA-subunit mRNA levels. In contrast, GDF9 could stimulate inhibin βB-subunit levels but to a lesser degree than could activin A. GDF9 pretreatment for 24 h significantly enhanced activin A-induced inhibin βB-subunit mRNA levels, inhibin B accumulation, and Smad2/3 phosphorylation; these effects were attenuated by BMPR2 ECD. GDF9 pretreatment also induced activin receptors (ACVR2B/1B) and Smad2/3 but reduced inhibitory Smad7 mRNA levels. Second, GDF9 was detected as mRNA and protein in hGL cells. Reduced endogenous GDF9 expression following targeting siRNA transfection was associated with decreased levels of ACVR2B/1B and Smad2/3/4 but increased Smad7 mRNA and protein and consequently, reduced levels of activin A-induced inhibin βB-subunit mRNA and inhibin B. These results further confirm the effects of exogenous GDF9 treatment mentioned previously. Third, GDF9 was found to suppress basal and activin A-induced expression of FST and FSTL3, which are extracellular inhibitors of activin A. These effects were attenuated by BMPR2 ECD and GDF9 siRNA transfection. FST or FSTL3 siRNA transfection significantly 125  augmented levels of activin A-induced inhibin βB-subunit mRNA. Furthermore, the enhancing effects of GDF9 in activin A-induced inhibin βB-subunit mRNA were attenuated by FST. This finding may explain why GDF9 can enhance expression of activin A-induced inhibin βB-subunit mRNA extracellularly; this result further supports the role of GDF9 in the regulation of inhibin βB-subunit mRNA level. Fourth, GDF9 reverses the suppressive effects of activin A on StAR expression and progesterone accumulation by enhancing the expression of inhibin B, which in turn, acts as an inhibitor of activin A. This relationship is corroborated by the following. GDF9 siRNA transfection reduces endogenous GDF9 level, thus increasing acitivin A suppression of StAR expression and progesterone accumulation. Furthermore, we detected a corresponding decrease when endogenous inhibin B accumulation was reduced after inhibin α-subunit siRNA transfection. In summary, these results describe a pathway (Fig. 6.1) in which GDF9 increases hGL cell response to activin A by acting on GDF9 receptors, BMPR2/TGFβR1, which then activate Smad2/3 to form complexes with Smad4. These complexes then activate transcription factors in the nucleus to target genes that increase ACVR2B/1B and Smad2/3 expression and reduce Smad7 activity. These changes, in turn, allow more activin A to bind to its receptors and thus increase the cellular response to activin A stimulation in inhibin B accumulation intracellularly. Additionally, GDF9 inhibits FST and FSTL3 gene transcription and then FST and FSTL3 secretions into the culture media. More activin A can avoid binding with its extracellular inhibitors, FST and FSTL3, and then activate activin receptors and a downstream signaling cascade that stimulates more inhibin B accumulation. Thus, GDF9 enhances activin A-induced inhibin B accumulation via not only intracellular mechanism but also extracellular regulating  126  mechanism, however, which mechanism plays more important role in this process must be further elucidated. Increasing inhibin B is subsequently as an inhibitor of activin A to remove its suppression in the basal and FSH-induced progesterone accumulation via reducing the expression of StAR, the rate-limiting step in steroidogenesis of hGL cells. Although we would like to hypothesize a feedback loop exists between activin A and inhibin B in granulosa cells, further studies are required to establish if inhibin B can indeed directly inhibit activin A accumulation and and other mechanisms involved in antagonizing activin A action. Inhibin A and inhibin B exhibit distinct patterns of secretion throughout the menstrual cycle (166, 255, 256, 304, 305). Serum inhibin B levels increase during luteal-follicular transition and peak in the mid-follicular phase and again the day following the LH surge. In contrast, inhibin A levels begin to rise in the late follicular phase and peak in the mid-cycle and again in the mid-luteal phase. Whereas activin A may be present in an unbound form, it is almost irreversibly bound to follistatin and there is little variation in levels of free activin A throughout the menstrual cycle (163). Conversely, the level of activin A peaks before the onset of menses, in the luteal phase and mid-cycle (164). We describe for the first time the close relationship between GDF9 and activin A as it relates to inhibin B accumulation in hGL cells. One may speculate that increasing GDF9 expression (268) during folliculogenesis enhances human granulosa cell response to activin A and leads to increased inhibin B in the follicular phase. With release of the oocyte after ovulation, a main source of GDF9 is removed; hence, cell response to activin A with respect to βB-subunit is withdrawn, which may explain the decline in inhibin B secretion after ovulation. This finding is also consistent with the marked decreased in the inhibin βB-subunit whereas αand βA-subunits are expressed in Gdf9 null mouse ovaries (266).  127  PCOS is one of the most common causes of infertility in women and can affect 5-10% of women of reproductive age worldwide (306). Studies have shown that ovary dysfunction is involved in the pathology of PCOS (307), in particular the number of growing follicles in PCOS ovaries are double that of normal ovaries (308). This important finding proposes that the process of follicle development may be abnormal in PCOS patients. Results from animal studies have shown that GDF9 has a close relationship to folliculogenesis and female fertility (309-311). Many abnormalities emerge in GDF9-deficient female mice (219, 239, 266). Decreased levels of GDF9 mRNA are found in developing oocytes from women with PCOS or PCO compared to women with normal ovaries; the decreased levels are evident throughout folliculogenesis, particularly during the primary and secondary stages (268). These data suggest that GDF9 transcription is postponed and inhibited in PCOS and PCO oocytes throughout the enlargement and differentiation phase. During IVF and intracytoplasmic sperm injection treatments, women with PCOS have lower GDF9 mRNA levels in cumulus cells than do women with normal menstrual cycles (227). Additionally, FST levels are significantly higher in women with PCOS (285); the authors suggest that altered FST function may contribute to the PCOS phenotype. Important characteristics of PCOS such as induced ovarian androgen production, decreased serum FSH level, and impaired ovarian follicle development could be caused by increased expression and function of FST (286). A PCOS-like phenotype has also been reported in an FST transgenetic mouse model (287). The roles of inhibin and activin in the pathology of PCOS remain unclear. Studies have shown that serum levels of inhibin α-subunit, pro-αC, inhibin A, and inhibin B are increased in women with PCOS (276, 312-315). One study describe that compared to size-matched follicles  128  from normal ovaries, levels of inhibin α- and βA-subunit mRNA decline in PCOS follicles while inhibin βB-subunit mRNA levels do not change (316). However, several groups have also reported that inhibin B levels in follicular fluid from PCOS follicles are not different with that of nomal follicles, even if inhibin A levels are lower in PCOS follicles (275, 304, 317). It is not very clear about the significance of these results; nevertheless, we cannot exclude the role of inhibins in the irregular inhibition of pituitary FSH secrection in women with PCOS. The results from this thesis demonstrate that GDF9 can enhance activin A-induced inhibin B accumulation in hGL cells partly via suppressed expression of FST and FSTL3. These results help define connections between these important factors involved in the pathogenesis of PCOS and further assist us to develop effective methods for curing patients. Previously, GDF9 was primarily considered to be an oocyte-specific factor. However, recent studies show that GDF9 is present in granulosa cells from monkey (223), goat (224) and pig (225) ovaries, and granulosa and cumulus cells from human ovaries (226, 227, 269, 270). Our study confirmed that GDF9 was also present in hGL cells. Indeed, our study demonstrates an autocrine role of endogenous GDF9 for the first time and further supports the enhancing action of GDF9 secreted by the oocyte in activin A-induced inhibin B accumulation. Although it is tempting to suggest that with release of the oocyte after ovulation, a main source of GDF9 is removed (hence, the cell response to activin A with respect to βB-subunit is withdrawn) to explain the decline in inhibin B secretion after ovulation, the contribution of GDF9 secreted by granulosa cells in this regard remains unknown. Progesterone production is an important function of granulosa cells. Activin A is a known inhibitor of luteinization as reported in human (292, 293), bovine (294, 295), and rat (296) ovary studies. In the menstrual cycle, serum activin A levels are higher in the early follicular,  129  mid-cycle and late luteal phases (164) while serum inhibin B levels peak in the mid-follicular phase and immediately following ovulation (161). During ovarian stimulation for IVF, activin A levels remained unchanged after 7-8 days of FSH treatment but rose significantly with follicle maturity (297). Based on these observations and findings of our studies, we speculate that increased expression of activin A in the pre-ovulatory follicle in the presence of GDF9 further enhances inhibin B expression and overcomes the inhibitory effect of activin A on luteinization. This hypothesis suggests that local activin activity is detrimental to luteal function and structure; and removal of activin from the system is critical for successful luteal function. To what extent this plays a physiological role remains unknown given the dominant effects of LH on luteinization. We acknowledge the limitations of extrapolating our results from granulosa-lutein cells obtained from women undergoing IVF treatment to normal ovarian physiology because these cells have been exposed to pharmacological concentrations of exogenous gonadotropins in the process of luteinization and hCG stimulation. Nevertheless, in the absence of granulosa cells from the unstimulated, normal ovaries that are easily accessible for research, findings from our cell culture model provide interesting hypotheses for further evaluation with regards to the role of GDF9 and related mechanisms involved in regulating inhibin B production during the peri-ovulatory transition. Furthermore, because of the limitation, we cannot link our results from the in vitro model system with other clinic parameters of IVF patients to build up the information network. Whether these conclusions can be applied to normal or PCOS patients remains to be determined. IVF is an accepted clinical procedure for treatment of specific fertility problems not amenable to other forms of therapy. In summary, our results raise interesting questions about  130  the GDF9, inhibin B, FST, and progesterone expression profiles in IVF patients. If GDP9 is clearly shown to play a crucial role in human follicular development, it will introduce new targets for IVF treatment in infertile couples, potential biochemical markers for successful fertilization and embryo development and successful pregnancy.  6.2.  Recommendations for Future Work The results advance our understanding of the regulatory roles of GDF9 in activin-induced  inhibin B production in hGL cells. However, the complete relationship between GDF9 and activin A in follicular development is still far from understood. Further studies are needed to improve our knowledge with regards to the development of hGL cells. These questions include: 1) Progesterone production is one important aspect of steroidogenesis of hGL cells. Whether interaction of GDF9, activin A and inhibin B play a role in other aspects of steroidogenesis like estradiol accumulation needs to be investigated. 2) Although our knowledge of GDF9 and BMP15 has increased, many questions about the biological roles of these proteins within the ovary remain. Our studies demonstrate that GDF9 is crucial for regulation of activin A in the accumulation of inhibin B and progesterone in hGL cells. The role of its close homologue BMP15 in this system remains unknown. Additionally, whether GDF9 and BMP15 actually function as homodimers, heterodimers or both, and whether the formation of the putative heterodimer is temporarily regulated must still be studied. 3) Based on our in vitro results, further studies are needed to test the hypothesis that the GDF9/activin/inhibin/FST/FSTL3 system is closely related with the pathogenesis of PCOS.  131  Compared to normal patients, PCOS patients have different inhibin B, FST and steroid expression profiles as mentioned previously. Additionally, very recently a number of reports have suggested that variations in GDF9 may cause abnormal ovary functioning in conditions like PCOS. Human granulosa-lutein cells from healthy women and women with PCOS need to be cultured and tested. The results can be considered in conjunction with IVF clinic data to contribute to the literature on this topic to provide important insight into the roles of these autocrine and paracrine regulators in PCOS. 4) BMP3 (318), BMP5 (22) and BMP6 (319) (also secreted by oocytes) have been shown to be expressed in granulosa cells. Whether they have similar functions as GDF9 in the regulation of activin-induced inhibin B and progesterone production needs to be further examined. 5) GDF9 enhanced activin A-induced inhibin βB-subunit mRNA level and thus inhibin B accumulation as well. Yet we cannot rule out the possibility that increased levels of βB-subunit mRNA lead to increased synthesis of activin B (βBβB) or activin AB (βAβB). Additionally, we cannot dismiss the possibility that decreased levels of inhibin α-subunit mRNA after targeted siRNA transfection may lead to increased synthesis of activin B and activin AB. Unfortunately, specific and sensitive activin B and activin AB assays are commercially unavailable. Furthermore, whether activin B and activin AB can behave similarly to activin A in inhibin B and progesterone accumulation, and whether their actions can be regulated by GDF9 remains to be determined. 6) Aside from FST and FSTL3, other extracellular regulators like BAMBI (124), Cripto (116), betaglycan (107) and InhBP/p120 (160) can antagonize while endoglin (123) can facilitate activin signaling. The role of GDF9 in the regulation of these factors needs to be further  132  investigated. 7) FST isoforms (FST288, FST303 and FST315) have different locations in the cell culture. The shortest isoform (FST288) has a higher affinity for heparin-sulfated proteoglycans, allowing it to be concentrated in the outer surface of the plasma membrane (320). This localization could prevent the autocrine effects of activin as well as the paracrine or endocrine effects. Conversely, the longest isoform (FST315) is found primarily in the circulation, a finding that is consistent with its reduced affinity for heparin (208). The intermediate isoform (FST303) is found primarily in gonadal extracts and fluids (209). The real-time PCR primers and ELISA used in this thesis can only detect the mRNA and proteins of all isoforms and thus cannot distinguish the different roles of these three isoforms. More studies must be undertaken in this regard. Furthermore, because FSTL3 could be localized to the nucleus and secrete more slowly than other FST isoforms, additional studies are needed to test the hypothesis that GDF9 not only decreases the synthesis and secretion of FST isoforms and FSTL3, but also modulates their cellular location. 8) The results from this thesis are from in vitro cell models. Whether these conclusions can be applied in vivo remains to be determined.  133  6.3.  Figures  TGFβR1 BMPR2 ACVR1B ACVR2B  FST/FSTL3  Inhibins Activin A  GDF9  GDF9 regulated genes  Cytoplasm I-Smad7 R-Smad2/3 Co-Smad4  ACVR1B↑ ACVR2B↑ Smad2 ↑ Smad3 ↑ Smad7 ↓ FST FSTL3  ↓ ↓  Progesterone  Cholesterol ↓StAR  ↑Cellular response to activin A  Activin A  ↑Free activin A  ↑Inhibin βBsubunit  Nucleus ↑Inhibin B  SBE  Target gene transcription  Fig. 6.1 A model, suggested by the results, for the interactions between GDF9 and activin A in the regulation of inhibin B and progesterone accumulation by human granulosa-lutein cells. Abbreviations: ACVR, Activin receptor; BMPR, Bone morphogenetic protein receptor; FST, Follistatin; FSTL3, Follistatin-like 3; GDF9, Growth differentiation factor 9; Smad, Son of mothers against decapentaplegia; SBE, Smad binding element; Co-Smad, Common Smad; I-Smad, Inhibitory Smad; R-Smad, Receptor-activated Smad; StAR, Steroidogenic acute regulatory protein; TGFβR, Transforming growth factor β receptor.  134  REFERENCES 1.  Knight PG, Glister C 2006 TGF-beta superfamily members and ovarian follicle development. Reproduction (Cambridge, England) 132:191-206  2.  Yi SE, Daluiski A, Pederson R, Rosen V, Lyons KM 2000 The type I BMP receptor BMPRIB is required for chondrogenesis in the mouse limb. Development 127:621-630  3.  Ying Y, Qi X, Zhao GQ 2001 Induction of primordial germ cells from murine epiblasts by synergistic action of BMP4 and BMP8B signaling pathways. Proceedings of the National Academy of Sciences of the United States of America 98:7858-7862  4.  Pangas SA, Matzuk MM 2004 Genetic models for transforming growth factor beta superfamily signaling in ovarian follicle development. Molecular and cellular endocrinology 225:83-91  5.  Simpson ER 1979 Cholesterol side-chain cleavage, cytochrome P450, and the control of steroidogenesis. Molecular and cellular endocrinology 13:213-227  6.  Simard J, Ricketts ML, Gingras S, Soucy P, Feltus FA, Melner MH 2005 Molecular biology of the 3beta-hydroxysteroid dehydrogenase/delta5-delta4 isomerase gene family. Endocrine reviews 26:525-582  7.  Simpson ER, Mahendroo MS, Means GD, Kilgore MW, Hinshelwood MM, Graham-Lorence S, Amarneh B, Ito Y, Fisher CR, Michael MD, et al. 1994 Aromatase cytochrome P450, the enzyme responsible for estrogen biosynthesis. Endocrine reviews 15:342-355  8.  Stocco DM, Clark BJ 1996 Regulation of the acute production of steroids in steroidogenic cells. Endocrine reviews 17:221-244  9.  Falck B 1959 Site of production of oestrogen in rat ovary as studied in micro-transplants. Acta physiologica Scandinavica 47:1-101  10.  Richards JS 1994 Hormonal control of gene expression in the ovary. Endocrine reviews 15:725-751  11.  Richards JS 2001 Perspective: the ovarian follicle--a perspective in 2001. Endocrinology 142:2184-2193  12.  Sugawara T, Kiriakidou M, McAllister JM, Kallen CB, Strauss JF, 3rd 1997 Multiple steroidogenic factor 1 binding elements in the human steroidogenic acute regulatory protein gene 5'-flanking region are required for maximal promoter activity and cyclic AMP responsiveness. Biochemistry 36:7249-7255  135  13.  Sekar N, Lavoie HA, Veldhuis JD 2000 Concerted regulation of steroidogenic acute regulatory gene expression by luteinizing hormone and insulin (or insulin-like growth factor I) in primary cultures of porcine granulosa-luteal cells. Endocrinology 141:3983-3992  14.  Balasubramanian K, Lavoie HA, Garmey JC, Stocco DM, Veldhuis JD 1997 Regulation of porcine granulosa cell steroidogenic acute regulatory protein (StAR) by insulin-like growth factor I: synergism with follicle-stimulating hormone or protein kinase A agonist. Endocrinology 138:433-439  15.  Pescador N, Stocco DM, Murphy BD 1999 Growth factor modulation of steroidogenic acute regulatory protein and luteinization in the pig ovary. Biology of reproduction 60:1453-1461  16.  Ruiz-Cortes ZT, Martel-Kennes Y, Gevry NY, Downey BR, Palin MF, Murphy BD 2003 Biphasic effects of leptin in porcine granulosa cells. Biology of reproduction 68:789-796  17.  Rekawiecki R, Nowik M, Kotwica J 2005 Stimulatory effect of LH, PGE2 and progesterone on StAR protein, cytochrome P450 cholesterol side chain cleavage and 3beta hydroxysteroid dehydrogenase gene expression in bovine luteal cells. Prostaglandins & other lipid mediators 78:169-184  18.  Attia GR, Dooley CA, Rainey WE, Carr BR 2000 Transforming growth factor beta inhibits steroidogenic acute regulatory (StAR) protein expression in human ovarian thecal cells. Molecular and cellular endocrinology 170:123-129  19.  Zheng X, Price CA, Tremblay Y, Lussier JG, Carriere PD 2008 Role of transforming growth factor-beta1 in gene expression and activity of estradiol and progesterone-generating enzymes in FSH-stimulated bovine granulosa cells. Reproduction 136:447-457  20.  Miyoshi T, Otsuka F, Suzuki J, Takeda M, Inagaki K, Kano Y, Otani H, Mimura Y, Ogura T, Makino H 2006 Mutual regulation of follicle-stimulating hormone signaling and bone morphogenetic protein system in human granulosa cells. Biology of reproduction 74:1073-1082  21.  Pierre A, Pisselet C, Dupont J, Mandon-Pepin B, Monniaux D, Monget P, Fabre S 2004 Molecular basis of bone morphogenetic protein-4 inhibitory action on progesterone secretion by ovine granulosa cells. Journal of molecular endocrinology 33:805-817  22.  Pierre A, Pisselet C, Dupont J, Bontoux M, Monget P 2005 Bone morphogenetic protein 5 expression in the rat ovary: biological effects on granulosa cell proliferation and steroidogenesis. Biology of reproduction 73:1102-1108  23.  Otsuka F, Yamamoto S, Erickson GF, Shimasaki S 2001 Bone morphogenetic protein-15 inhibits follicle-stimulating hormone (FSH) action by suppressing FSH receptor expression. The Journal of biological chemistry 276:11387-11392  136  24.  Chen YJ, Feng Q, Liu YX 1999 Expression of the steroidogenic acute regulatory protein and luteinizing hormone receptor and their regulation by tumor necrosis factor alpha in rat corpora lutea. Biology of reproduction 60:419-427  25.  Tsutsumi R, Hiroi H, Momoeda M, Hosokawa Y, Nakazawa F, Koizumi M, Yano T, Tsutsumi O, Taketani Y 2008 Inhibitory effects of cholesterol sulfate on progesterone production in human granulosa-like tumor cell line, KGN. Endocrine journal 55:575-581  26.  Elvin JA, Clark AT, Wang P, Wolfman NM, Matzuk MM 1999 Paracrine actions of growth differentiation factor-9 in the mammalian ovary. Molecular endocrinology (Baltimore, Md 13:1035-1048  27.  Sugawara T, Holt JA, Kiriakidou M, Strauss JF, 3rd 1996 Steroidogenic factor 1-dependent promoter activity of the human steroidogenic acute regulatory protein (StAR) gene. Biochemistry 35:9052-9059  28.  Gillio-Meina C, Hui YY, LaVoie HA 2003 GATA-4 and GATA-6 transcription factors: expression, immunohistochemical localization, and possible function in the porcine ovary. Biology of reproduction 68:412-422  29.  LaVoie HA, Singh D, Hui YY 2004 Concerted regulation of the porcine steroidogenic acute regulatory protein gene promoter activity by follicle-stimulating hormone and insulin-like growth factor I in granulosa cells involves GATA-4 and CCAAT/enhancer binding protein beta. Endocrinology 145:3122-3134  30.  Manna PR, Eubank DW, Lalli E, Sassone-Corsi P, Stocco DM 2003 Transcriptional regulation of the mouse steroidogenic acute regulatory protein gene by the cAMP response-element binding protein and steroidogenic factor 1. Journal of molecular endocrinology 30:381-397  31.  Ronen-Fuhrmann T, Timberg R, King SR, Hales KH, Hales DB, Stocco DM, Orly J 1998 Spatio-temporal expression patterns of steroidogenic acute regulatory protein (StAR) during follicular development in the rat ovary. Endocrinology 139:303-315  32.  Bao B, Garverick HA 1998 Expression of steroidogenic enzyme and gonadotropin receptor genes in bovine follicles during ovarian follicular waves: a review. Journal of animal science 76:1903-1921  33.  Goldring NB, Durica JM, Lifka J, Hedin L, Ratoosh SL, Miller WL, Orly J, Richards JS 1987 Cholesterol side-chain cleavage P450 messenger ribonucleic acid: evidence for hormonal regulation in rat ovarian follicles and constitutive expression in corpora lutea. Endocrinology 120:1942-1950  137  34.  Kiriakidou M, McAllister JM, Sugawara T, Strauss JF, 3rd 1996 Expression of steroidogenic acute regulatory protein (StAR) in the human ovary. The Journal of clinical endocrinology and metabolism 81:4122-4128  35.  Silverman E, Eimerl S, Orly J 1999 CCAAT enhancer-binding protein beta and GATA-4 binding regions within the promoter of the steroidogenic acute regulatory protein (StAR) gene are required for transcription in rat ovarian cells. The Journal of biological chemistry 274:17987-17996  36.  Urban RJ, Garmey JC, Shupnik MA, Veldhuis JD 1991 Follicle-stimulating hormone increases concentrations of messenger ribonucleic acid encoding cytochrome P450 cholesterol side-chain cleavage enzyme in primary cultures of porcine granulosa cells. Endocrinology 128:2000-2007  37.  Sekar N, Garmey JC, Veldhuis JD 2000 Mechanisms underlying the steroidogenic synergy of insulin and luteinizing hormone in porcine granulosa cells: joint amplification of pivotal sterol-regulatory genes encoding the low-density lipoprotein (LDL) receptor, steroidogenic acute regulatory (StAR) protein and cytochrome P450 side-chain cleavage (P450scc) enzyme. Molecular and cellular endocrinology 159:25-35  38.  Voutilainen R, Tapanainen J, Chung BC, Matteson KJ, Miller WL 1986 Hormonal regulation of P450scc (20,22-desmolase) and P450c17 (17 alpha-hydroxylase/17,20-lyase) in cultured human granulosa cells. The Journal of clinical endocrinology and metabolism 63:202-207  39.  Ravindranath N, Little-Ihrig L, Benyo DF, Zeleznik AJ 1992 Role of luteinizing hormone in the expression of cholesterol side-chain cleavage cytochrome P450 and 3 beta-hydroxysteroid dehydrogenase, delta 5-4 isomerase messenger ribonucleic acids in the primate corpus luteum. Endocrinology 131:2065-2070  40.  Duncan WC, Cowen GM, Illingworth PJ 1999 Steroidogenic enzyme expression in human corpora lutea in the presence and absence of exogenous human chorionic gonadotrophin (HCG). Molecular human reproduction 5:291-298  41.  Stocco C, Callegari E, Gibori G 2001 Opposite effect of prolactin and prostaglandin F(2 alpha) on the expression of luteal genes as revealed by rat cDNA expression array. Endocrinology 142:4158-4161  42.  Eimerl S, Orly J 2002 Regulation of steroidogenic genes by insulin-like growth factor-1 and follicle-stimulating hormone: differential responses of cytochrome P450 side-chain cleavage, steroidogenic acute regulatory protein, and 3beta-hydroxysteroid dehydrogenase/isomerase in rat granulosa cells. Biology of reproduction 67:900-910  138  43.  Devoto L, Christenson LK, McAllister JM, Makrigiannakis A, Strauss JF, 3rd 1999 Insulin and insulin-like growth factor-I and -II modulate human granulosa-lutein cell steroidogenesis: enhancement of steroidogenic acute regulatory protein (StAR) expression. Molecular human reproduction 5:1003-1010  44.  Swan CL, Agostini MC, Bartlewski PM, Feyles V, Urban RJ, Chedrese PJ 2002 Effects of progestins on progesterone synthesis in a stable porcine granulosa cell line: control of transcriptional activity of the cytochrome p450 side-chain cleavage gene. Biology of reproduction 66:959-965  45.  Xu YP, Chedrese PJ, Thacker PA 1995 Growth hormone amplifies insulin-like growth factor I induced progesterone accumulation and P450scc mRNA expression. Molecular and cellular endocrinology 111:199-206  46.  Mamluk R, Greber Y, Meidan R 1999 Hormonal regulation of messenger ribonucleic acid expression for steroidogenic factor-1, steroidogenic acute regulatory protein, and cytochrome P450 side-chain cleavage in bovine luteal cells. Biology of reproduction 60:628-634  47.  Shimada M, Hernandez-Gonzalez I, Gonzalez-Robayna I, Richards JS 2006 Paracrine and autocrine regulation of epidermal growth factor-like factors in cumulus oocyte complexes and granulosa cells: key roles for prostaglandin synthase 2 and progesterone receptor. Molecular endocrinology (Baltimore, Md 20:1352-1365  48.  Chen YJ, Lee MT, Yao HC, Hsiao PW, Ke FC, Hwang JJ 2008 Crucial role of estrogen receptor-alpha interaction with transcription coregulators in follicle-stimulating hormone and transforming growth factor beta1 up-regulation of steroidogenesis in rat ovarian granulosa cells. Endocrinology 149:4658-4668  49.  Glister C, Richards SL, Knight PG 2005 Bone morphogenetic proteins (BMP) -4, -6, and -7 potently suppress basal and luteinizing hormone-induced androgen production by bovine theca interna cells in primary culture: could ovarian hyperandrogenic dysfunction be caused by a defect in thecal BMP signaling? Endocrinology 146:1883-1892  50.  Li XM, Juorio AV, Murphy BD 1993 Prostaglandins alter the abundance of messenger ribonucleic acid for steroidogenic enzymes in cultured porcine granulosa cells. Biology of reproduction 48:1360-1366  51.  Veldhuis JD, Garmey JC, Urban RJ, Demers LM, Aggarwal BB 1991 Ovarian actions of tumor necrosis factor-alpha (TNF alpha): pleiotropic effects of TNF alpha on differentiated functions of untransformed swine granulosa cells. Endocrinology 129:641-648  139  52.  Neuvians TP, Schams D, Berisha B, Pfaffl MW 2004 Involvement of pro-inflammatory cytokines, mediators of inflammation, and basic fibroblast growth factor in prostaglandin F2alpha-induced luteolysis in bovine corpus luteum. Biology of reproduction 70:473-480  53.  Clemens JW, Lala DS, Parker KL, Richards JS 1994 Steroidogenic factor-1 binding and transcriptional activity of the cholesterol side-chain cleavage promoter in rat granulosa cells. Endocrinology 134:1499-1508  54.  Sher N, Yivgi-Ohana N, Orly J 2007 Transcriptional regulation of the cholesterol side chain cleavage cytochrome P450 gene (CYP11A1) revisited: binding of GATA, cyclic adenosine 3',5'-monophosphate response element-binding protein and activating protein (AP)-1 proteins to a distal novel cluster of cis-regulatory elements potentiates AP-2 and steroidogenic factor-1-dependent gene expression in the rodent placenta and ovary. Molecular endocrinology (Baltimore, Md 21:948-962  55.  Kim JW, Havelock JC, Carr BR, Attia GR 2005 The orphan nuclear receptor, liver receptor homolog-1, regulates cholesterol side-chain cleavage cytochrome p450 enzyme in human granulosa cells. The Journal of clinical endocrinology and metabolism 90:1678-1685  56.  Ben-Zimra M, Koler M, Orly J 2002 Transcription of cholesterol side-chain cleavage cytochrome P450 in the placenta: activating protein-2 assumes the role of steroidogenic factor-1 by binding to an overlapping promoter element. Molecular endocrinology 16:1864-1880  57.  Conley AJ, Kaminski MA, Dubowsky SA, Jablonka-Shariff A, Redmer DA, Reynolds LP 1995 Immunohistochemical localization of 3 beta-hydroxysteroid dehydrogenase and P450 17 alpha-hydroxylase during follicular and luteal development in pigs, sheep, and cows. Biology of reproduction 52:1081-1094  58.  Yuan W, Lucy MC 1996 Messenger ribonucleic acid expression for growth hormone receptor, luteinizing hormone receptor, and steroidogenic enzymes during the estrous cycle and pregnancy in porcine and bovine corpora lutea. Domestic animal endocrinology 13:431-444  59.  Martel C, Labrie C, Dupont E, Couet J, Trudel C, Rheaume E, Simard J, Luu-The V, Pelletier G, Labrie F 1990 Regulation of 3 beta-hydroxysteroid dehydrogenase/delta 5-delta 4 isomerase expression and activity in the hypophysectomized rat ovary: interactions between the stimulatory effect of human chorionic gonadotropin and the luteolytic effect of prolactin. Endocrinology 127:2726-2737  60.  McGee E, Sawetawan C, Bird I, Rainey WE, Carr BR 1995 The effects of insulin on 3 beta-hydroxysteroid dehydrogenase expression in human luteinized granulosa cells. Journal of the Society for Gynecologic Investigation 2:535-541  140  61.  Magoffin DA, Weitsman SR 1993 Insulin-like growth factor-I stimulates the expression of 3 beta-hydroxysteroid dehydrogenase messenger ribonucleic acid in ovarian theca-interstitial cells. Biology of reproduction 48:1166-1173  62.  Bendell JJ, Dorrington JH 1990 Epidermal growth factor influences growth and differentiation of rat granulosa cells. Endocrinology 127:533-540  63.  McAllister JM, Byrd W, Simpson ER 1994 The effects of growth factors and phorbol esters on steroid biosynthesis in isolated human theca interna and granulosa-lutein cells in long term culture. The Journal of clinical endocrinology and metabolism 79:106-112  64.  Brankin V, Quinn RL, Webb R, Hunter MG 2005 Evidence for a functional bone morphogenetic protein (BMP) system in the porcine ovary. Domestic animal endocrinology 28:367-379  65.  Martel C, Gagne D, Couet J, Labrie Y, Simard J, Labrie F 1994 Rapid modulation of ovarian 3 beta-hydroxysteroid dehydrogenase/delta 5-delta 4 isomerase gene expression by prolactin and human chorionic gonadotropin in the hypophysectomized rat. Molecular and cellular endocrinology 99:63-71  66.  Peng N, Kim JW, Rainey WE, Carr BR, Attia GR 2003 The role of the orphan nuclear receptor, liver receptor homologue-1, in the regulation of human corpus luteum 3beta-hydroxysteroid dehydrogenase type II. The Journal of clinical endocrinology and metabolism 88:6020-6028  67.  Martin LJ, Taniguchi H, Robert NM, Simard J, Tremblay JJ, Viger RS 2005 GATA factors and the nuclear receptors, steroidogenic factor 1/liver receptor homolog 1, are key mutual partners in the regulation of the human 3beta-hydroxysteroid dehydrogenase type 2 promoter. Molecular endocrinology 19:2358-2370  68.  Havelock JC, Smith AL, Seely JB, Dooley CA, Rodgers RJ, Rainey WE, Carr BR 2005 The NGFI-B family of transcription factors regulates expression of 3beta-hydroxysteroid dehydrogenase type 2 in the human ovary. Molecular human reproduction 11:79-85  69.  Xing Y, Saner-Amigh K, Nakamura Y, Hinshelwood MM, Carr BR, Mason JI, Rainey WE 2009 The farnesoid X receptor regulates transcription of 3beta-hydroxysteroid dehydrogenase type 2 in human adrenal cells. Molecular and cellular endocrinology 299:153-162  70.  Feltus FA, Kovacs WJ, Nicholson W, Silva CM, Nagdas SK, Ducharme NA, Melner MH 2003 Epidermal growth factor increases cortisol production and type II 3 beta-hydroxysteroid dehydrogenase/Delta(5)-Delta(4)-isomerase expression in human adrenocortical carcinoma cells: evidence for a Stat5-dependent mechanism. Endocrinology 144:1847-1853  141  71.  Lavoie HA, King SR 2009 Transcriptional regulation of steroidogenic genes: STARD1, CYP11A1 and HSD3B. Experimental biology and medicine 234:880-907  72.  Roberts AB, Anzano MA, Lamb LC, Smith JM, Sporn MB 1981 New class of transforming growth factors potentiated by epidermal growth factor: isolation from non-neoplastic tissues. Proceedings of the National Academy of Sciences of the United States of America 78:5339-5343  73.  Lin SY, Morrison JR, Phillips DJ, de Kretser DM 2003 Regulation of ovarian function by the TGF-beta superfamily and follistatin. Reproduction 126:133-148  74.  Kingsley DM 1994 The TGF-beta superfamily: new members, new receptors, and new genetic tests of function in different organisms. Genes & development 8:133-146  75.  Sun PD, Davies DR 1995 The cystine-knot growth-factor superfamily. Annual review of biophysics and biomolecular structure 24:269-291  76.  Schlunegger MP, Grutter MG 1992 An unusual feature revealed by the crystal structure at 2.2 A resolution of human transforming growth factor-beta 2. Nature 358:430-434  77.  McPherron AC, Lee SJ 1993 GDF-3 and GDF-9: two new members of the transforming growth factor-beta superfamily containing a novel pattern of cysteines. The Journal of biological chemistry 268:3444-3449  78.  Laitinen M, Vuojolainen K, Jaatinen R, Ketola I, Aaltonen J, Lehtonen E, Heikinheimo M, Ritvos O 1998 A novel growth differentiation factor-9 (GDF-9) related factor is co-expressed with GDF-9 in mouse oocytes during folliculogenesis. Mechanisms of development 78:135-140  79.  Dube JL, Wang P, Elvin J, Lyons KM, Celeste AJ, Matzuk MM 1998 The bone morphogenetic protein 15 gene is X-linked and expressed in oocytes. Molecular endocrinology 12:1809-1817  80.  Meno C, Saijoh Y, Fujii H, Ikeda M, Yokoyama T, Yokoyama M, Toyoda Y, Hamada H 1996 Left-right asymmetric expression of the TGF beta-family member lefty in mouse embryos. Nature 381:151-155  81.  Meno C, Ito Y, Saijoh Y, Matsuda Y, Tashiro K, Kuhara S, Hamada H 1997 Two closely-related left-right asymmetrically expressed genes, lefty-1 and lefty-2: their distinct expression domains, chromosomal linkage and direct neuralizing activity in Xenopus embryos. Genes Cells 2:513-524  82.  de Caestecker M 2004 The transforming growth factor-beta superfamily of receptors. Cytokine & growth factor reviews 15:1-11  142  83.  ten Dijke P, Hill CS 2004 New insights into TGF-beta-Smad signalling. Trends in biochemical sciences 29:265-273  84.  Massague J 1998 TGF-beta signal transduction. Annual review of biochemistry 67:753-791  85.  Sekelsky JJ, Newfeld SJ, Raftery LA, Chartoff EH, Gelbart WM 1995 Genetic characterization and cloning of mothers against dpp, a gene required for decapentaplegic function in Drosophila melanogaster. Genetics 139:1347-1358  86.  Derynck R, Gelbart WM, Harland RM, Heldin CH, Kern SE, Massague J, Melton DA, Mlodzik M, Padgett RW, Roberts AB, Smith J, Thomsen GH, Vogelstein B, Wang XF 1996 Nomenclature: vertebrate mediators of TGFbeta family signals. Cell 87:173  87.  Flanders KC, Kim ES, Roberts AB 2001 Immunohistochemical expression of Smads 1-6 in the 15-day gestation mouse embryo: signaling by BMPs and TGF-betas. Dev Dyn 220:141-154  88.  Luukko K, Ylikorkala A, Makela TP 2001 Developmentally regulated expression of Smad3, Smad4, Smad6, and Smad7 involved in TGF-beta signaling. Mechanisms of development 101:209-212  89.  Chen X, Rubock MJ, Whitman M 1996 A transcriptional partner for MAD proteins in TGF-beta signalling. Nature 383:691-696  90.  Pardali E, Xie XQ, Tsapogas P, Itoh S, Arvanitidis K, Heldin CH, ten Dijke P, Grundstrom T, Sideras P 2000 Smad and AML proteins synergistically confer transforming growth factor beta1 responsiveness to human germ-line IgA genes. The Journal of biological chemistry 275:3552-3560  91.  Pouponnot C, Jayaraman L, Massague J 1998 Physical and functional interaction of SMADs and p300/CBP. The Journal of biological chemistry 273:22865-22868  92.  Kaivo-oja N, Jeffery LA, Ritvos O, Mottershead DG 2006 Smad signalling in the ovary. Reprod Biol Endocrinol 4:21  93.  Sasaki A, Masuda Y, Ohta Y, Ikeda K, Watanabe K 2001 Filamin associates with Smads and regulates transforming growth factor-beta signaling. The Journal of biological chemistry 276:17871-17877  94.  Moustakas A, Souchelnytskyi S, Heldin CH 2001 Smad regulation in TGF-beta signal transduction. Journal of cell science 114:4359-4369  95.  Dennler S, Huet S, Gauthier JM 1999 A short amino-acid sequence in MH1 domain is responsible for functional differences between Smad2 and Smad3. Oncogene 18:1643-1648  96.  Wicks SJ, Lui S, Abdel-Wahab N, Mason RM, Chantry A 2000 Inactivation of smad-transforming growth factor beta signaling by Ca(2+)-calmodulin-dependent protein kinase II. Molecular and cellular biology 20:8103-8111  143  97.  Kretzschmar M, Doody J, Massague J 1997 Opposing BMP and EGF signalling pathways converge on the TGF-beta family mediator Smad1. Nature 389:618-622  98.  Yakymovych I, Ten Dijke P, Heldin CH, Souchelnytskyi S 2001 Regulation of Smad signaling by protein kinase C. Faseb J 15:553-555  99.  Brown JD, DiChiara MR, Anderson KR, Gimbrone MA, Jr., Topper JN 1999 MEKK-1, a component of the stress (stress-activated protein kinase/c-Jun N-terminal kinase) pathway, can selectively activate Smad2-mediated transcriptional activation in endothelial cells. The Journal of biological chemistry 274:8797-8805  100.  Hanafusa H, Ninomiya-Tsuji J, Masuyama N, Nishita M, Fujisawa J, Shibuya H, Matsumoto K, Nishida E 1999 Involvement of the p38 mitogen-activated protein kinase pathway in transforming growth factor-beta-induced gene expression. The Journal of biological chemistry 274:27161-27167  101.  Bhowmick NA, Ghiassi M, Bakin A, Aakre M, Lundquist CA, Engel ME, Arteaga CL, Moses HL 2001 Transforming growth factor-beta1 mediates epithelial to mesenchymal transdifferentiation through a RhoA-dependent mechanism. Molecular biology of the cell 12:27-36  102.  Edlund S, Landstrom M, Heldin CH, Aspenstrom P 2002 Transforming growth factor-beta-induced mobilization of actin cytoskeleton requires signaling by small GTPases Cdc42 and RhoA. Molecular biology of the cell 13:902-914  103.  Wilkes MC, Mitchell H, Penheiter SG, Dore JJ, Suzuki K, Edens M, Sharma DK, Pagano RE, Leof EB 2005 Transforming growth factor-beta activation of phosphatidylinositol 3-kinase is independent of Smad2 and Smad3 and regulates fibroblast responses via p21-activated kinase-2. Cancer research 65:10431-10440  104.  Chen RH, Su YH, Chuang RL, Chang TY 1998 Suppression of transforming growth factor-beta-induced apoptosis through a phosphatidylinositol 3-kinase/Akt-dependent pathway. Oncogene 17:1959-1968  105.  Pulaski L, Landstrom M, Heldin CH, Souchelnytskyi S 2001 Phosphorylation of Smad7 at Ser-249 does not interfere with its inhibitory role in transforming growth factor-beta-dependent signaling but affects Smad7-dependent transcriptional activation. The Journal of biological chemistry 276:14344-14349  106.  Lopez-Casillas F, Cheifetz S, Doody J, Andres JL, Lane WS, Massague J 1991 Structure and expression of the membrane proteoglycan betaglycan, a component of the TGF-beta receptor system. Cell 67:785-795  144  107.  Lewis KA, Gray PC, Blount AL, MacConell LA, Wiater E, Bilezikjian LM, Vale W 2000 Betaglycan binds inhibin and can mediate functional antagonism of activin signalling. Nature 404:411-414  108.  Chen C, Wang XF, Sun L 1997 Expression of transforming growth factor beta (TGFbeta) type III receptor restores autocrine TGFbeta1 activity in human breast cancer MCF-7 cells. The Journal of biological chemistry 272:12862-12867  109.  Bandyopadhyay A, Zhu Y, Cibull ML, Bao L, Chen C, Sun L 1999 A soluble transforming growth factor beta type III receptor suppresses tumorigenicity and metastasis of human breast cancer MDA-MB-231 cells. Cancer research 59:5041-5046  110.  Wiater E, Vale W 2003 Inhibin is an antagonist of bone morphogenetic protein signaling. The Journal of biological chemistry 278:7934-7941  111.  Liu J, Kuulasmaa T, Kosma VM, Butzow R, Vanttinen T, Hyden-Granskog C, Voutilainen R 2003 Expression of betaglycan, an inhibin coreceptor, in normal human ovaries and ovarian sex cord-stromal tumors and its regulation in cultured human granulosa-luteal cells. The Journal of clinical endocrinology and metabolism 88:5002-5008  112.  Bianco C, Strizzi L, Normanno N, Khan N, Salomon DS 2005 Cripto-1: an oncofetal gene with many faces. Current topics in developmental biology 67:85-133  113.  Strizzi L, Bianco C, Normanno N, Salomon D 2005 Cripto-1: a multifunctional modulator during embryogenesis and oncogenesis. Oncogene 24:5731-5741  114.  Schier AF 2003 Nodal signaling in vertebrate development. Annual review of cell and developmental biology 19:589-621  115.  Yeo C, Whitman M 2001 Nodal signals to Smads through Cripto-dependent and Cripto-independent mechanisms. Molecular cell 7:949-957  116.  Kelber JA, Shani G, Booker EC, Vale WW, Gray PC 2008 Cripto is a noncompetitive activin antagonist that forms analogous signaling complexes with activin and nodal. The Journal of biological chemistry 283:4490-4500  117.  Strizzi L, Bianco C, Normanno N, Seno M, Wechselberger C, Wallace-Jones B, Khan NI, Hirota M, Sun Y, Sanicola M, Salomon DS 2004 Epithelial mesenchymal transition is a characteristic of hyperplasias and tumors in mammary gland from MMTV-Cripto-1 transgenic mice. Journal of cellular physiology 201:266-276  118.  Lastres P, Letamendia A, Zhang H, Rius C, Almendro N, Raab U, Lopez LA, Langa C, Fabra A, Letarte M, Bernabeu C 1996 Endoglin modulates cellular responses to TGF-beta 1. The Journal of cell biology 133:1109-1121  145  119.  Cheifetz S, Bellon T, Cales C, Vera S, Bernabeu C, Massague J, Letarte M 1992 Endoglin is a component of the transforming growth factor-beta receptor system in human endothelial cells. The Journal of biological chemistry 267:19027-19030  120.  Lee NY, Ray B, How T, Blobe GC 2008 Endoglin promotes transforming growth factor beta-mediated Smad 1/5/8 signaling and inhibits endothelial cell migration through its association with GIPC. The Journal of biological chemistry 283:32527-32533  121.  Gougos A, Letarte M 1990 Primary structure of endoglin, an RGD-containing glycoprotein of human endothelial cells. The Journal of biological chemistry 265:8361-8364  122.  Parker WL, Goldring MB, Philip A 2003 Endoglin is expressed on human chondrocytes and forms a heteromeric complex with betaglycan in a ligand and type II TGFbeta receptor independent manner. J Bone Miner Res 18:289-302  123.  Barbara NP, Wrana JL, Letarte M 1999 Endoglin is an accessory protein that interacts with the signaling receptor complex of multiple members of the transforming growth factor-beta superfamily. The Journal of biological chemistry 274:584-594  124.  Onichtchouk D, Chen YG, Dosch R, Gawantka V, Delius H, Massague J, Niehrs C 1999 Silencing of TGF-beta signalling by the pseudoreceptor BAMBI. Nature 401:480-485  125.  Yan X, Lin Z, Chen F, Zhao X, Chen H, Ning Y, Chen YG 2009 Human BAMBI cooperates with Smad7 to inhibit transforming growth factor-beta signaling. The Journal of biological chemistry 284:30097-30104  126.  Karaulanov E, Knochel W, Niehrs C 2004 Transcriptional regulation of BMP4 synexpression in transgenic Xenopus. The EMBO journal 23:844-856  127.  Sekiya T, Adachi S, Kohu K, Yamada T, Higuchi O, Furukawa Y, Nakamura Y, Nakamura T, Tashiro K, Kuhara S, Ohwada S, Akiyama T 2004 Identification of BMP and activin membrane-bound inhibitor (BAMBI), an inhibitor of transforming growth factor-beta signaling, as a target of the beta-catenin pathway in colorectal tumor cells. The Journal of biological chemistry 279:6840-6846  128.  Sekiya T, Oda T, Matsuura K, Akiyama T 2004 Transcriptional regulation of the TGF-beta pseudoreceptor BAMBI by TGF-beta signaling. Biochemical and biophysical research communications 320:680-684  129.  Esch FS, Shimasaki S, Mercado M, Cooksey K, Ling N, Ying S, Ueno N, Guillemin R 1987 Structural characterization of follistatin: a novel follicle-stimulating hormone release-inhibiting polypeptide from the gonad. Molecular endocrinology (Baltimore, Md 1:849-855  130.  Nakamura T, Takio K, Eto Y, Shibai H, Titani K, Sugino H 1990 Activin-binding protein from rat ovary is follistatin. Science 247:836-838  146  131.  Iemura S, Yamamoto TS, Takagi C, Uchiyama H, Natsume T, Shimasaki S, Sugino H, Ueno N 1998 Direct binding of follistatin to a complex of bone-morphogenetic protein and its receptor inhibits ventral and epidermal cell fates in early Xenopus embryo. Proceedings of the National Academy of Sciences of the United States of America 95:9337-9342  132.  Tsuchida K, Arai KY, Kuramoto Y, Yamakawa N, Hasegawa Y, Sugino H 2000 Identification and characterization of a novel follistatin-like protein as a binding protein for the TGF-beta family. The Journal of biological chemistry 275:40788-40796  133.  Martens JW, de Winter JP, Timmerman MA, McLuskey A, van Schaik RH, Themmen AP, de Jong FH 1997 Inhibin interferes with activin signaling at the level of the activin receptor complex in Chinese hamster ovary cells. Endocrinology 138:2928-2936  134.  De Crescenzo G, Grothe S, Zwaagstra J, Tsang M, O'Connor-McCourt MD 2001 Real-time monitoring of the interactions of transforming growth factor-beta (TGF-beta) isoforms with latency-associated protein and the ectodomains of the TGF-beta type II and III receptors reveals different kinetic models and stoichiometries of binding. The Journal of biological chemistry 276:29632-29643  135.  Marcelino J, Sciortino CM, Romero MF, Ulatowski LM, Ballock RT, Economides AN, Eimon PM, Harland RM, Warman ML 2001 Human disease-causing NOG missense mutations: effects on noggin secretion, dimer formation, and bone morphogenetic protein binding. Proceedings of the National Academy of Sciences of the United States of America 98:11353-11358  136.  Zimmerman LB, De Jesus-Escobar JM, Harland RM 1996 The Spemann organizer signal noggin binds and inactivates bone morphogenetic protein 4. Cell 86:599-606  137.  Groppe J, Greenwald J, Wiater E, Rodriguez-Leon J, Economides AN, Kwiatkowski W, Affolter M, Vale WW, Belmonte JC, Choe S 2002 Structural basis of BMP signalling inhibition by the cystine knot protein Noggin. Nature 420:636-642  138.  Piccolo S, Sasai Y, Lu B, De Robertis EM 1996 Dorsoventral patterning in Xenopus: inhibition of ventral signals by direct binding of chordin to BMP-4. Cell 86:589-598  139.  Tsukazaki T, Chiang TA, Davison AF, Attisano L, Wrana JL 1998 SARA, a FYVE domain protein that recruits Smad2 to the TGFbeta receptor. Cell 95:779-791  140.  Zhu H, Kavsak P, Abdollah S, Wrana JL, Thomsen GH 1999 A SMAD ubiquitin ligase targets the BMP pathway and affects embryonic pattern formation. Nature 400:687-693  141.  Zhang Y, Chang C, Gehling DJ, Hemmati-Brivanlou A, Derynck R 2001 Regulation of Smad degradation and activity by Smurf2, an E3 ubiquitin ligase. Proceedings of the National Academy of Sciences of the United States of America 98:974-979  147  142.  Fukuchi M, Imamura T, Chiba T, Ebisawa T, Kawabata M, Tanaka K, Miyazono K 2001 Ligand-dependent degradation of Smad3 by a ubiquitin ligase complex of ROC1 and associated proteins. Molecular biology of the cell 12:1431-1443  143.  Lin X, Duan X, Liang YY, Su Y, Wrighton KH, Long J, Hu M, Davis CM, Wang J, Brunicardi FC, Shi Y, Chen YG, Meng A, Feng XH 2006 PPM1A functions as a Smad phosphatase to terminate TGFbeta signaling. Cell 125:915-928  144.  Di Guglielmo GM, Le Roy C, Goodfellow AF, Wrana JL 2003 Distinct endocytic pathways regulate TGF-beta receptor signalling and turnover. Nature cell biology 5:410-421  145.  Nakao A, Afrakhte M, Moren A, Nakayama T, Christian JL, Heuchel R, Itoh S, Kawabata M, Heldin NE, Heldin CH, ten Dijke P 1997 Identification of Smad7, a TGFbeta-inducible antagonist of TGF-beta signalling. Nature 389:631-635  146.  Ishisaki A, Yamato K, Hashimoto S, Nakao A, Tamaki K, Nonaka K, ten Dijke P, Sugino H, Nishihara T 1999 Differential inhibition of Smad6 and Smad7 on bone morphogenetic protein- and activin-mediated growth arrest and apoptosis in B cells. The Journal of biological chemistry 274:13637-13642  147.  Ulloa L, Doody J, Massague J 1999 Inhibition of transforming growth factor-beta/SMAD signalling by the interferon-gamma/STAT pathway. Nature 397:710-713  148.  Bitzer M, von Gersdorff G, Liang D, Dominguez-Rosales A, Beg AA, Rojkind M, Bottinger EP 2000 A mechanism of suppression of TGF-beta/SMAD signaling by NF-kappa B/RelA. Genes & development 14:187-197  149.  Kretzschmar M, Doody J, Timokhina I, Massague J 1999 A mechanism of repression of TGFbeta/ Smad signaling by oncogenic Ras. Genes & development 13:804-816  150.  Lo RS, Massague J 1999 Ubiquitin-dependent degradation of TGF-beta-activated smad2. Nature cell biology 1:472-478  151.  van den Hurk R, Zhao J 2005 Formation of mammalian oocytes and their growth, differentiation and maturation within ovarian follicles. Theriogenology 63:1717-1751  152.  Fang J, Wang SQ, Smiley E, Bonadio J 1997 Genes coding for mouse activin beta C and beta E are closely linked and exhibit a liver-specific expression pattern in adult tissues. Biochemical and biophysical research communications 231:655-661  153.  Lau AL, Kumar TR, Nishimori K, Bonadio J, Matzuk MM 2000 Activin betaC and betaE genes are not essential for mouse liver growth, differentiation, and regeneration. Molecular and cellular biology 20:6127-6137  154.  Mellor SL, Cranfield M, Ries R, Pedersen J, Cancilla B, de Kretser D, Groome NP, Mason AJ, Risbridger GP 2000 Localization of activin beta(A)-, beta(B)-, and  148  beta(C)-subunits in humanprostate and evidence for formation of new activin heterodimers of beta(C)-subunit. The Journal of clinical endocrinology and metabolism 85:4851-4858 155.  Ling N, Ying SY, Ueno N, Shimasaki S, Esch F, Hotta M, Guillemin R 1986 Pituitary FSH is released by a heterodimer of the beta-subunits from the two forms of inhibin. Nature 321:779-782  156.  Robertson DM, Cahir N, Findlay JK, Burger HG, Groome N 1997 The biological and immunological characterization of inhibin A and B forms in human follicular fluid and plasma. The Journal of clinical endocrinology and metabolism 82:889-896  157.  Niimi S, Horikawa M, Seki T, Ariga T, Kobayashi T, Hayakawa T 2002 Effect of activins AB and B on DNA synthesis stimulated by epidermal growth factor in primary cultured rat hepatocytes. Biological & pharmaceutical bulletin 25:437-440  158.  Pangas SA, Woodruff TK 2000 Activin signal transduction pathways. Trends in endocrinology and metabolism: TEM 11:309-314  159.  Robertson DM, Hertan R, Farnworth PG 2000 Is the action of inhibin mediated via a unique receptor? Reviews of reproduction 5:131-135  160.  Bernard DJ, Chapman SC, Woodruff TK 2002 Inhibin binding protein (InhBP/p120), betaglycan, and the continuing search for the inhibin receptor. Molecular endocrinology (Baltimore, Md 16:207-212  161.  Welt CK 2004 Regulation and function of inhibins in the normal menstrual cycle. Seminars in reproductive medicine 22:187-193  162.  Schneyer AL, Rzucidlo DA, Sluss PM, Crowley WF, Jr. 1994 Characterization of unique binding kinetics of follistatin and activin or inhibin in serum. Endocrinology 135:667-674  163.  Demura R, Suzuki T, Tajima S, Mitsuhashi S, Odagiri E, Demura H, Ling N 1993 Human plasma free activin and inhibin levels during the menstrual cycle. The Journal of clinical endocrinology and metabolism 76:1080-1082  164.  Muttukrishna S, Fowler PA, George L, Groome NP, Knight PG 1996 Changes in peripheral serum levels of total activin A during the human menstrual cycle and pregnancy. The Journal of clinical endocrinology and metabolism 81:3328-3334  165.  Schneyer AL, Fujiwara T, Fox J, Welt CK, Adams J, Messerlian GM, Taylor AE 2000 Dynamic changes in the intrafollicular inhibin/activin/follistatin axis during human follicular development: relationship to circulating hormone concentrations. The Journal of clinical endocrinology and metabolism 85:3319-3330  166.  Welt CK, Martin KA, Taylor AE, Lambert-Messerlian GM, Crowley WF, Jr., Smith JA, Schoenfeld DA, Hall JE 1997 Frequency modulation of follicle-stimulating hormone (FSH)  149  during the luteal-follicular transition: evidence for FSH control of inhibin B in normal women. The Journal of clinical endocrinology and metabolism 82:2645-2652 167.  Welt CK, Adams JM, Sluss PM, Hall JE 1999 Inhibin A and inhibin B responses to gonadotropin withdrawal depends on stage of follicle development. The Journal of clinical endocrinology and metabolism 84:2163-2169  168.  Welt CK, Smith ZA, Pauler DK, Hall JE 2001 Differential regulation of inhibin A and inhibin B by luteinizing hormone, follicle-stimulating hormone, and stage of follicle development. The Journal of clinical endocrinology and metabolism 86:2531-2537  169.  Feng ZM, Li YP, Chen CL 1989 Analysis of the 5'-flanking regions of rat inhibin alpha- and beta-B-subunit genes suggests two different regulatory mechanisms. Molecular endocrinology (Baltimore, Md 3:1914-1925  170.  Ardekani AM, Romanelli JC, Mayo KE 1998 Structure of the rat inhibin and activin betaA-subunit gene and regulation in an ovarian granulosa cell line. Endocrinology 139:3271-3279  171.  Pei L, Dodson R, Schoderbek WE, Maurer RA, Mayo KE 1991 Regulation of the alpha inhibin gene by cyclic adenosine 3',5'-monophosphate after transfection into rat granulosa cells. Molecular endocrinology (Baltimore, Md 5:521-534  172.  Dykema JC, Mayo KE 1994 Two messenger ribonucleic acids encoding the common beta B-chain of inhibin and activin have distinct 5'-initiation sites and are differentially regulated in rat granulosa cells. Endocrinology 135:702-711  173.  Eramaa M, Hilden K, Tuuri T, Ritvos O 1995 Regulation of inhibin/activin subunit messenger ribonucleic acids (mRNAs) by activin A and expression of activin receptor mRNAs in cultured human granulosa-luteal cells. Endocrinology 136:4382-4389  174.  Eramaa M, Ritvos O 1996 Transforming growth factor-beta 1 and -beta 2 induce inhibin and activin beta B-subunit messenger ribonucleic acid levels in cultured human granulosa-luteal cells. Fertility and sterility 65:954-960  175.  Vanttinen T, Liu J, Liu J, Hyden-Granskog C, Parviainen M, Penttila I, Voutilainen R 2000 Regulation of immunoreactive inhibin A and B secretion in cultured human granulosa-luteal cells by gonadotropins, activin A and insulin-like growth factor type-1 receptor. The Journal of endocrinology 167:289-294  176.  Kaivo-Oja N, Bondestam J, Kamarainen M, Koskimies J, Vitt U, Cranfield M, Vuojolainen K, Kallio JP, Olkkonen VM, Hayashi M, Moustakas A, Groome NP, ten Dijke P, Hsueh AJ, Ritvos O 2003 Growth differentiation factor-9 induces Smad2 activation  150  and inhibin B production in cultured human granulosa-luteal cells. The Journal of clinical endocrinology and metabolism 88:755-762 177.  Jaatinen R, Bondestam J, Raivio T, Hilden K, Dunkel L, Groome N, Ritvos O 2002 Activation of the bone morphogenetic protein signaling pathway induces inhibin beta(B)-subunit mRNA and secreted inhibin B levels in cultured human granulosa-luteal cells. The Journal of clinical endocrinology and metabolism 87:1254-1261  178.  Vale W, Rivier C, Hsueh A, Campen C, Meunier H, Bicsak T, Vaughan J, Corrigan A, Bardin W, Sawchenko P, et al. 1988 Chemical and biological characterization of the inhibin family of protein hormones. Recent progress in hormone research 44:1-34  179.  Loria P, Petraglia F, Concari M, Bertolotti M, Martella P, Luisi S, Grisolia C, Foresta C, Volpe A, Genazzani AR, Carulli N 1998 Influence of age and sex on serum concentrations of total dimeric activin A. European journal of endocrinology / European Federation of Endocrine Societies 139:487-492  180.  Welt CK, Lambert-Messerlian G, Zheng W, Crowley WF, Jr., Schneyer AL 1997 Presence of activin, inhibin, and follistatin in epithelial ovarian carcinoma. The Journal of clinical endocrinology and metabolism 82:3720-3727  181.  Woodruff TK, Lyon RJ, Hansen SE, Rice GC, Mather JP 1990 Inhibin and activin locally regulate rat ovarian folliculogenesis. Endocrinology 127:3196-3205  182.  Ueno N, Ling N, Ying SY, Esch F, Shimasaki S, Guillemin R 1987 Isolation and partial characterization of follistatin: a single-chain Mr 35,000 monomeric protein that inhibits the release of follicle-stimulating hormone. Proceedings of the National Academy of Sciences of the United States of America 84:8282-8286  183.  Robertson DM, Klein R, de Vos FL, McLachlan RI, Wettenhall RE, Hearn MT, Burger HG, de Kretser DM 1987 The isolation of polypeptides with FSH suppressing activity from bovine follicular fluid which are structurally different to inhibin. Biochemical and biophysical research communications 149:744-749  184.  de Winter JP, ten Dijke P, de Vries CJ, van Achterberg TA, Sugino H, de Waele P, Huylebroeck D, Verschueren K, van den Eijnden-van Raaij AJ 1996 Follistatins neutralize activin bioactivity by inhibition of activin binding to its type II receptors. Molecular and cellular endocrinology 116:105-114  185.  Nakamura T, Sugino K, Titani K, Sugino H 1991 Follistatin, an activin-binding protein, associates with heparan sulfate chains of proteoglycans on follicular granulosa cells. The Journal of biological chemistry 266:19432-19437  151  186.  Kogawa K, Nakamura T, Sugino K, Takio K, Titani K, Sugino H 1991 Activin-binding protein is present in pituitary. Endocrinology 128:1434-1440  187.  Nakamura T, Hasegawa Y, Sugino K, Kogawa K, Titani K, Sugino H 1992 Follistatin inhibits activin-induced differentiation of rat follicular granulosa cells in vitro. Biochimica et biophysica acta 1135:103-109  188.  Asashima M, Nakano H, Uchiyama H, Sugino H, Nakamura T, Eto Y, Ejima D, Nishimatsu S, Ueno N, Kinoshita K 1991 Presence of activin (erythroid differentiation factor) in unfertilized eggs and blastulae of Xenopus laevis. Proceedings of the National Academy of Sciences of the United States of America 88:6511-6514  189.  Yamashita H, ten Dijke P, Huylebroeck D, Sampath TK, Andries M, Smith JC, Heldin CH, Miyazono K 1995 Osteogenic protein-1 binds to activin type II receptors and induces certain activin-like effects. The Journal of cell biology 130:217-226  190.  Gamer LW, Cox KA, Small C, Rosen V 2001 Gdf11 is a negative regulator of chondrogenesis and myogenesis in the developing chick limb. Developmental biology 229:407-420  191.  Otsuka F, Moore RK, Iemura S, Ueno N, Shimasaki S 2001 Follistatin inhibits the function of the oocyte-derived factor BMP-15. Biochemical and biophysical research communications 289:961-966  192.  Hayette S, Gadoux M, Martel S, Bertrand S, Tigaud I, Magaud JP, Rimokh R 1998 FLRG (follistatin-related gene), a new target of chromosomal rearrangement in malignant blood disorders. Oncogene 16:2949-2954  193.  Sidis Y, Mukherjee A, Keutmann H, Delbaere A, Sadatsuki M, Schneyer A 2006 Biological activity of follistatin isoforms and follistatin-like-3 is dependent on differential cell surface binding and specificity for activin, myostatin, and bone morphogenetic proteins. Endocrinology 147:3586-3597  194.  Shimasaki S, Koga M, Esch F, Mercado M, Cooksey K, Koba A, Ling N 1988 Porcine follistatin gene structure supports two forms of mature follistatin produced by alternative splicing. Biochemical and biophysical research communications 152:717-723  195.  Shimasaki S, Koga M, Esch F, Cooksey K, Mercado M, Koba A, Ueno N, Ying SY, Ling N, Guillemin R 1988 Primary structure of the human follistatin precursor and its genomic organization. Proceedings of the National Academy of Sciences of the United States of America 85:4218-4222  196.  Lane TF, Sage EH 1994 The biology of SPARC, a protein that modulates cell-matrix interactions. Faseb J 8:163-173  152  197.  Magill C, Reist NE, Fallon JR, Nitkin RM, Wallace BG, McMahan UJ 1987 Agrin. Progress in brain research 71:391-396  198.  Maurer P, Hohenadl C, Hohenester E, Gohring W, Timpl R, Engel J 1995 The C-terminal portion of BM-40 (SPARC/osteonectin) is an autonomously folding and crystallisable domain that binds calcium and collagen IV. Journal of molecular biology 253:347-357  199.  Wang Q, Keutmann HT, Schneyer AL, Sluss PM 2000 Analysis of human follistatin structure: identification of two discontinuous N-terminal sequences coding for activin A binding and structural consequences of activin binding to native proteins. Endocrinology 141:3183-3193  200.  Okabayashi K, Shoji H, Onuma Y, Nakamura T, Nose K, Sugino H, Asashima M 1999 cDNA cloning and distribution of the Xenopus follistatin-related protein. Biochemical and biophysical research communications 254:42-48  201.  Keutmann HT, Schneyer AL, Sidis Y 2004 The role of follistatin domains in follistatin biological action. Molecular endocrinology 18:228-240  202.  Thompson TB, Lerch TF, Cook RW, Woodruff TK, Jardetzky TS 2005 The structure of the follistatin:activin complex reveals antagonism of both type I and type II receptor binding. Developmental cell 9:535-543  203.  Xia Y, Schneyer AL 2009 The biology of activin: recent advances in structure, regulation and function. The Journal of endocrinology 202:1-12  204.  Michel U, Albiston A, Findlay JK 1990 Rat follistatin: gonadal and extragonadal expression and evidence for alternative splicing. Biochemical and biophysical research communications 173:401-407  205.  Sumitomo S, Inouye S, Liu XJ, Ling N, Shimasaki S 1995 The heparin binding site of follistatin is involved in its interaction with activin. Biochemical and biophysical research communications 208:1-9  206.  Sugino K, Kurosawa N, Nakamura T, Takio K, Shimasaki S, Ling N, Titani K, Sugino H 1993 Molecular heterogeneity of follistatin, an activin-binding protein. Higher affinity of the carboxyl-terminal truncated forms for heparan sulfate proteoglycans on the ovarian granulosa cell. The Journal of biological chemistry 268:15579-15587  207.  Hashimoto O, Nakamura T, Shoji H, Shimasaki S, Hayashi Y, Sugino H 1997 A novel role of follistatin, an activin-binding protein, in the inhibition of activin action in rat pituitary cells. Endocytotic degradation of activin and its acceleration by follistatin associated with cell-surface heparan sulfate. The Journal of biological chemistry 272:13835-13842  153  208.  Schneyer AL, Wang Q, Sidis Y, Sluss PM 2004 Differential distribution of follistatin isoforms: application of a new FS315-specific immunoassay. The Journal of clinical endocrinology and metabolism 89:5067-5075  209.  Sugino H, Sugino K, Hashimoto O, Shoji H, Nakamura T 1997 Follistatin and its role as an activin-binding protein. J Med Invest 44:1-14  210.  Stamler R, Keutmann HT, Sidis Y, Kattamuri C, Schneyer A, Thompson TB 2008 The structure of FSTL3.activin A complex. Differential binding of N-terminal domains influences follistatin-type antagonist specificity. The Journal of biological chemistry 283:32831-32838  211.  Saito S, Sidis Y, Mukherjee A, Xia Y, Schneyer A 2005 Differential biosynthesis and intracellular  transport  of  follistatin  isoforms  and  follistatin-like-3.  Endocrinology  146:5052-5062 212.  Shimonaka M, Inouye S, Shimasaki S, Ling N 1991 Follistatin binds to both activin and inhibin through the common subunit. Endocrinology 128:3313-3315  213.  Schneyer AL, Hall HA, Lambert-Messerlian G, Wang QF, Sluss P, Crowley WF, Jr. 1996 Follistatin-activin complexes in human serum and follicular fluid differ immunologically and biochemically. Endocrinology 137:240-247  214.  McGrath SA, Esquela AF, Lee SJ 1995 Oocyte-specific expression of growth/differentiation factor-9. Molecular endocrinology 9:131-136  215.  Daopin S, Piez KA, Ogawa Y, Davies DR 1992 Crystal structure of transforming growth factor-beta 2: an unusual fold for the superfamily. Science (New York, NY 257:369-373  216.  Liao WX, Moore RK, Otsuka F, Shimasaki S 2003 Effect of intracellular interactions on the processing and secretion of bone morphogenetic protein-15 (BMP-15) and growth and differentiation factor-9. Implication of the aberrant ovarian phenotype of BMP-15 mutant sheep. The Journal of biological chemistry 278:3713-3719  217.  McNatty KP, Juengel JL, Reader KL, Lun S, Myllymaa S, Lawrence SB, Western A, Meerasahib MF, Mottershead DG, Groome NP, Ritvos O, Laitinen MP 2005 Bone morphogenetic protein 15 and growth differentiation factor 9 co-operate to regulate granulosa cell function. Reproduction 129:473-480  218.  Aaltonen J, Laitinen MP, Vuojolainen K, Jaatinen R, Horelli-Kuitunen N, Seppa L, Louhio H, Tuuri T, Sjoberg J, Butzow R, Hovata O, Dale L, Ritvos O 1999 Human growth differentiation factor 9 (GDF-9) and its novel homolog GDF-9B are expressed in oocytes during early folliculogenesis. The Journal of clinical endocrinology and metabolism 84:2744-2750  154  219.  Dong J, Albertini DF, Nishimori K, Kumar TR, Lu N, Matzuk MM 1996 Growth differentiation factor-9 is required during early ovarian folliculogenesis. Nature 383:531-535  220.  Hayashi M, McGee EA, Min G, Klein C, Rose UM, van Duin M, Hsueh AJ 1999 Recombinant growth differentiation factor-9 (GDF-9) enhances growth and differentiation of cultured early ovarian follicles. Endocrinology 140:1236-1244  221.  Jaatinen R, Laitinen MP, Vuojolainen K, Aaltonen J, Louhio H, Heikinheimo K, Lehtonen E, Ritvos O 1999 Localization of growth differentiation factor-9 (GDF-9) mRNA and protein in rat ovaries and cDNA cloning of rat GDF-9 and its novel homolog GDF-9B. Molecular and cellular endocrinology 156:189-193  222.  Bodensteiner KJ, Clay CM, Moeller CL, Sawyer HR 1999 Molecular cloning of the ovine Growth/Differentiation factor-9 gene and expression of growth/differentiation factor-9 in ovine and bovine ovaries. Biology of reproduction 60:381-386  223.  Duffy DM 2003 Growth differentiation factor-9 is expressed by the primate follicle throughout the periovulatory interval. Biology of reproduction 69:725-732  224.  Silva JR, van den Hurk R, van Tol HT, Roelen BA, Figueiredo JR 2005 Expression of growth differentiation factor 9 (GDF9), bone morphogenetic protein 15 (BMP15), and BMP receptors in the ovaries of goats. Molecular reproduction and development 70:11-19  225.  Paradis F, Novak S, Murdoch GK, Dyck MK, Dixon WT, Foxcroft GR 2009 Temporal regulation of BMP2, BMP6, BMP15, GDF9, BMPR1A, BMPR1B, BMPR2 and TGFBR1 mRNA expression in the oocyte, granulosa and theca cells of developing preovulatory follicles in the pig. Reproduction 138:115-129  226.  Yamamoto N, Christenson LK, McAllister JM, Strauss JF, 3rd 2002 Growth differentiation factor-9 inhibits 3'5'-adenosine monophosphate-stimulated steroidogenesis in human granulosa and theca cells. The Journal of clinical endocrinology and metabolism 87:2849-2856  227.  Zhao SY, Qiao J, Chen YJ, Liu P, Li J, Yan J 2009 Expression of growth differentiation factor-9 and bone morphogenetic protein-15 in oocytes and cumulus granulosa cells of patients with polycystic ovary syndrome. Fertility and sterility  228.  Fitzpatrick SL, Sindoni DM, Shughrue PJ, Lane MV, Merchenthaler IJ, Frail DE 1998 Expression of growth differentiation factor-9 messenger ribonucleic acid in ovarian and nonovarian rodent and human tissues. Endocrinology 139:2571-2578  229.  Mery L, Lefevre A, Benchaib M, Demirci B, Salle B, Guerin JF, Lornage J 2007 Follicular growth in vitro: detection of growth differentiation factor 9 (GDF9) and bone morphogenetic protein 15 (BMP15) during in vitro culture of ovine cortical slices. Molecular reproduction and development 74:767-774  155  230.  Eckery DC, Whale LJ, Lawrence SB, Wylde KA, McNatty KP, Juengel JL 2002 Expression of mRNA encoding growth differentiation factor 9 and bone morphogenetic protein 15 during follicular formation and growth in a marsupial, the brushtail possum (Trichosurus vulpecula). Molecular and cellular endocrinology 192:115-126  231.  Yan C, Elvin JA, Lin YN, Hadsell LA, Wang J, DeMayo FJ, Matzuk MM 2006 Regulation of growth differentiation factor 9 expression in oocytes in vivo: a key role of the E-box. Biology of reproduction 74:999-1006  232.  Cho JH, Itoh T, Sendai Y, Hoshi H 2008 Fibroblast growth factor 7 stimulates in vitro growth of oocytes originating from bovine early antral follicles. Molecular reproduction and development 75:1736-1743  233.  Lan ZJ, Gu P, Xu X, Jackson KJ, DeMayo FJ, O'Malley BW, Cooney AJ 2003 GCNF-dependent repression of BMP-15 and GDF-9 mediates gamete regulation of female fertility. The EMBO journal 22:4070-4081  234.  Mazerbourg S, Klein C, Roh J, Kaivo-Oja N, Mottershead DG, Korchynskyi O, Ritvos O, Hsueh AJ 2004 Growth differentiation factor-9 signaling is mediated by the type I receptor, activin receptor-like kinase 5. Molecular endocrinology (Baltimore, Md 18:653-665  235.  Vitt UA, Mazerbourg S, Klein C, Hsueh AJ 2002 Bone morphogenetic protein receptor type II is a receptor for growth differentiation factor-9. Biology of reproduction 67:473-480  236.  Kaivo-Oja N, Mottershead DG, Mazerbourg S, Myllymaa S, Duprat S, Gilchrist RB, Groome NP, Hsueh AJ, Ritvos O 2005 Adenoviral gene transfer allows Smad-responsive gene promoter analyses and delineation of type I receptor usage of transforming growth factor-beta family ligands in cultured human granulosa luteal cells. The Journal of clinical endocrinology and metabolism 90:271-278  237.  Li Q, Rajanahally S, Edson MA, Matzuk MM 2009 Stable expression and characterization of N-terminal tagged recombinant human bone morphogenetic protein 15. Molecular human reproduction 15:779-788  238.  McIntosh CJ, Lun S, Lawrence S, Western AH, McNatty KP, Juengel JL 2008 The proregion of mouse BMP15 regulates the cooperative interactions of BMP15 and GDF9. Biology of reproduction 79:889-896  239.  Carabatsos MJ, Elvin J, Matzuk MM, Albertini DF 1998 Characterization of oocyte and follicle development in growth differentiation factor-9-deficient mice. Developmental biology 204:373-384  156  240.  Vitt UA, McGee EA, Hayashi M, Hsueh AJ 2000 In vivo treatment with GDF-9 stimulates primordial and primary follicle progression and theca cell marker CYP17 in ovaries of immature rats. Endocrinology 141:3814-3820  241.  Hreinsson JG, Scott JE, Rasmussen C, Swahn ML, Hsueh AJ, Hovatta O 2002 Growth differentiation factor-9 promotes the growth, development, and survival of human ovarian follicles in organ culture. The Journal of clinical endocrinology and metabolism 87:316-321  242.  Nilsson EE, Skinner MK 2002 Growth and differentiation factor-9 stimulates progression of early primary but not primordial rat ovarian follicle development. Biology of reproduction 67:1018-1024  243.  Yan C, Wang P, DeMayo J, DeMayo FJ, Elvin JA, Carino C, Prasad SV, Skinner SS, Dunbar BS, Dube JL, Celeste AJ, Matzuk MM 2001 Synergistic roles of bone morphogenetic protein 15 and growth differentiation factor 9 in ovarian function. Molecular endocrinology 15:854-866  244.  Eppig JJ 2001 Oocyte control of ovarian follicular development and function in mammals. Reproduction 122:829-838  245.  Gilchrist RB, Ritter LJ, Armstrong DT 2004 Oocyte-somatic cell interactions during follicle development in mammals. Animal reproduction science 82-83:431-446  246.  Juengel JL, Hudson NL, Heath DA, Smith P, Reader KL, Lawrence SB, O'Connell AR, Laitinen MP, Cranfield M, Groome NP, Ritvos O, McNatty KP 2002 Growth differentiation factor 9 and bone morphogenetic protein 15 are essential for ovarian follicular development in sheep. Biology of reproduction 67:1777-1789  247.  Yoshino O, McMahon HE, Sharma S, Shimasaki S 2006 A unique preovulatory expression pattern plays a key role in the physiological functions of BMP-15 in the mouse. Proceedings of the National Academy of Sciences of the United States of America 103:10678-10683  248.  Vitt UA, Hayashi M, Klein C, Hsueh AJ 2000 Growth differentiation factor-9 stimulates proliferation but suppresses the follicle-stimulating hormone-induced differentiation of cultured granulosa cells from small antral and preovulatory rat follicles. Biology of reproduction 62:370-377  249.  Otsuka F, Yao Z, Lee T, Yamamoto S, Erickson GF, Shimasaki S 2000 Bone morphogenetic protein-15. Identification of target cells and biological functions. The Journal of biological chemistry 275:39523-39528  250.  Roh JS, Bondestam J, Mazerbourg S, Kaivo-Oja N, Groome N, Ritvos O, Hsueh AJ 2003 Growth differentiation factor-9 stimulates inhibin production and activates Smad2 in cultured rat granulosa cells. Endocrinology 144:172-178  157  251.  McNatty KP, Juengel JL, Reader KL, Lun S, Myllymaa S, Lawrence SB, Western A, Meerasahib MF, Mottershead DG, Groome NP, Ritvos O, Laitinen MP 2005 Bone morphogenetic protein 15 and growth differentiation factor 9 co-operate to regulate granulosa cell function in ruminants. Reproduction 129:481-487  252.  Elvin JA, Yan C, Matzuk MM 2000 Growth differentiation factor-9 stimulates progesterone synthesis in granulosa cells via a prostaglandin E2/EP2 receptor pathway. Proceedings of the National Academy of Sciences of the United States of America 97:10288-10293  253.  Joyce IM, Clark AT, Pendola FL, Eppig JJ 2000 Comparison of recombinant growth differentiation factor-9 and oocyte regulation of KIT ligand messenger ribonucleic acid expression in mouse ovarian follicles. Biology of reproduction 63:1669-1675  254.  Otsuka F, Shimasaki S 2002 A negative feedback system between oocyte bone morphogenetic protein 15 and granulosa cell kit ligand: its role in regulating granulosa cell mitosis. Proceedings of the National Academy of Sciences of the United States of America 99:8060-8065  255.  Groome NP, Illingworth PJ, O'Brien M, Cooke I, Ganesan TS, Baird DT, McNeilly AS 1994 Detection of dimeric inhibin throughout the human menstrual cycle by two-site enzyme immunoassay. Clinical endocrinology 40:717-723  256.  Groome NP, Illingworth PJ, O'Brien M, Pai R, Rodger FE, Mather JP, McNeilly AS 1996 Measurement of dimeric inhibin B throughout the human menstrual cycle. The Journal of clinical endocrinology and metabolism 81:1401-1405  257.  Eramaa M, Tuuri T, Hilden K, Ritvos O 1994 Regulation of inhibin alpha- and beta A-subunit messenger ribonucleic acid levels by chorionic gonadotropin and recombinant follicle-stimulating hormone in cultured human granulosa-luteal cells. The Journal of clinical endocrinology and metabolism 79:1670-1677  258.  Tsai SJ, Wu MH, Chuang PC, Chen HM 2001 Distinct regulation of gene expression by prostaglandin F(2alpha) (PGF(2alpha)) is associated with PGF(2alpha) resistance or susceptibility in human granulosa-luteal cells. Molecular human reproduction 7:415-423  259.  Gilchrist RB, Ritter LJ, Myllymaa S, Kaivo-Oja N, Dragovic RA, Hickey TE, Ritvos O, Mottershead DG 2006 Molecular basis of oocyte-paracrine signalling that promotes granulosa cell proliferation. Journal of cell science 119:3811-3821  260.  Gilchrist RB RL, Myllymaa S, Kaivo-Oja N, Amato F, Ritvos O, Mottershead DG. 2004 Mouse oocyte-secreted factors and GDF-9 stimulate granulosa cell proliferation via BMPR-II and activate smad2/3 pathway. Biol Reprod Special Issue:568  158  261.  Abdollah S, Macias-Silva M, Tsukazaki T, Hayashi H, Attisano L, Wrana JL 1997 TbetaRI phosphorylation of Smad2 on Ser465 and Ser467 is required for Smad2-Smad4 complex formation and signaling. The Journal of biological chemistry 272:27678-27685  262.  Davis RJ 1993 The mitogen-activated protein kinase signal transduction pathway. The Journal of biological chemistry 268:14553-14556  263.  Mason AJ, Hayflick JS, Ling N, Esch F, Ueno N, Ying SY, Guillemin R, Niall H, Seeburg PH 1985 Complementary DNA sequences of ovarian follicular fluid inhibin show precursor structure and homology with transforming growth factor-beta. Nature 318:659-663  264.  Bicsak TA, Tucker EM, Cappel S, Vaughan J, Rivier J, Vale W, Hsueh AJ 1986 Hormonal regulation of granulosa cell inhibin biosynthesis. Endocrinology 119:2711-2719  265.  Yamoto M, Minami S, Nakano R 1993 Immunohistochemical localization of inhibin subunits in polycystic ovary. The Journal of clinical endocrinology and metabolism 77:859-862  266.  Elvin JA, Yan C, Wang P, Nishimori K, Matzuk MM 1999 Molecular characterization of the follicle defects in the growth differentiation factor 9-deficient ovary. Molecular endocrinology 13:1018-1034  267.  Pan H, O'Brien M J, Wigglesworth K, Eppig JJ, Schultz RM 2005 Transcript profiling during mouse oocyte development and the effect of gonadotropin priming and development in vitro. Developmental biology 286:493-506  268.  Teixeira Filho FL, Baracat EC, Lee TH, Suh CS, Matsui M, Chang RJ, Shimasaki S, Erickson GF 2002 Aberrant expression of growth differentiation factor-9 in oocytes of women with polycystic ovary syndrome. The Journal of clinical endocrinology and metabolism 87:1337-1344  269.  Sidis Y, Fujiwara T, Leykin L, Isaacson K, Toth T, Schneyer AL 1998 Characterization of inhibin/activin subunit, activin receptor, and follistatin messenger ribonucleic acid in human and mouse oocytes: evidence for activin's paracrine signaling from granulosa cells to oocytes. Biology of reproduction 59:807-812  270.  Huang HY, Wang HS, Chan SH, Lee CL, Wang CW, Soong YK 2009 Granulosa-lutein cell growth differentiation factor-9 (GDF-9) messenger RNA and protein expression in in vitro fertilization (IVF) cycles: relation to characteristics of ovulation induction and IVF. Fertility and sterility 91:1583-1585  271.  Shi FT, Cheung AP, Leung PC 2009 Growth differentiation factor 9 enhances activin A-induced inhibin B production in human granulosa cells. Endocrinology 150:3540-3546  272.  Magoffin DA, Jakimiuk AJ 1997 Inhibin A, inhibin B and activin A in the follicular fluid of regularly cycling women. Human reproduction 12:1714-1719  159  273.  Eramaa M, Heikinheimo K, Tuuri T, Hilden K, Ritvos O 1993 Inhibin/activin subunit mRNA expression in human granulosa-luteal cells. Molecular and cellular endocrinology 92:R15-20  274.  Bondestam J, Kaivo-oja N, Kallio J, Groome N, Hyden-Granskog C, Fujii M, Moustakas A, Jalanko A, ten Dijke P, Ritvos O 2002 Engagement of activin and bone morphogenetic protein signaling pathway Smad proteins in the induction of inhibin B production in ovarian granulosa cells. Molecular and cellular endocrinology 195:79-88  275.  Magoffin DA, Jakimiuk AJ 1998 Inhibin A, inhibin B and activin A concentrations in follicular fluid from women with polycystic ovary syndrome. Human reproduction 13:2693-2698  276.  Anderson RA, Groome NP, Baird DT 1998 Inhibin A and inhibin B in women with polycystic ovarian syndrome during treatment with FSH to induce mono-ovulation. Clinical endocrinology 48:577-584  277.  Shi FT, Cheung AP, Huang HF, Leung PC 2009 Effects of endogenous growth differentiation factor 9 on activin A-induced inhibin B production in human granulosa-lutein cells. The Journal of clinical endocrinology and metabolism 94:5108-5116  278.  Shintani Y, Dyson M, Drummond AE, Findlay JK 1997 Regulation of follistatin production by rat granulosa cells in vitro. Endocrinology 138:2544-2551  279.  Liu ZH, Shintani Y, Sakamoto Y, Harada K, Zhang CY, Fujinaka Y, Abe M, Goto T, Saito S 1996 Effects of LHRH, FSH and activin A on follistatin secretion from cultured rat anterior pituitary cells. Endocrine journal 43:321-327  280.  Bartholin L, Maguer-Satta V, Hayette S, Martel S, Gadoux M, Corbo L, Magaud JP, Rimokh R 2002 Transcription activation of FLRG and follistatin by activin A, through Smad proteins, participates in a negative feedback loop to modulate activin A function. Oncogene 21:2227-2235  281.  Cataldo NA, Rabinovici J, Fujimoto VY, Jaffe RB 1994 Follistatin antagonizes the effects of activin-A on steroidogenesis in human luteinizing granulosa cells. The Journal of clinical endocrinology and metabolism 79:272-277  282.  Tano M, Minegishi T, Nakamura K, Nakamura M, Karino S, Miyamoto K, Ibuki Y 1995 Regulation of follistatin messenger ribonucleic acid in cultured rat granulosa cells. Molecular and cellular endocrinology 109:167-174  283.  Inouye S, Guo Y, DePaolo L, Shimonaka M, Ling N, Shimasaki S 1991 Recombinant expression of human follistatin with 315 and 288 amino acids: chemical and biological comparison with native porcine follistatin. Endocrinology 129:815-822  160  284.  Sidis Y, Tortoriello DV, Holmes WE, Pan Y, Keutmann HT, Schneyer AL 2002 Follistatin-related protein and follistatin differentially neutralize endogenous vs. exogenous activin. Endocrinology 143:1613-1624  285.  Eldar-Geva T, Spitz IM, Groome NP, Margalioth EJ, Homburg R 2001 Follistatin and activin A serum concentrations in obese and non-obese patients with polycystic ovary syndrome. Human reproduction 16:2552-2556  286.  Urbanek M, Wu X, Vickery KR, Kao LC, Christenson LK, Schneyer A, Legro RS, Driscoll DA, Strauss JF, 3rd, Dunaif A, Spielman RS 2000 Allelic variants of the follistatin gene in polycystic ovary syndrome. The Journal of clinical endocrinology and metabolism 85:4455-4461  287.  Guo Q, Kumar TR, Woodruff T, Hadsell LA, DeMayo FJ, Matzuk MM 1998 Overexpression of mouse follistatin causes reproductive defects in transgenic mice. Molecular endocrinology 12:96-106  288.  Alak BM, Coskun S, Friedman CI, Kennard EA, Kim MH, Seifer DB 1998 Activin A stimulates meiotic maturation of human oocytes and modulates granulosa cell steroidogenesis in vitro. Fertility and sterility 70:1126-1130  289.  Li W, Yuen BH, Leung PC 1992 Inhibition of progestin accumulation by activin-A in human granulosa cells. The Journal of clinical endocrinology and metabolism 75:285-289  290.  Matzuk MM, Finegold MJ, Mather JP, Krummen L, Lu H, Bradley A 1994 Development of cancer cachexia-like syndrome and adrenal tumors in inhibin-deficient mice. Proceedings of the National Academy of Sciences of the United States of America 91:8817-8821  291.  Nicol L, Bishop SC, Pong-Wong R, Bendixen C, Holm LE, Rhind SM, McNeilly AS 2009 Homozygosity for a single base-pair mutation in the oocyte-specific GDF9 gene results in sterility in Thoka sheep. Reproduction 138:921-933  292.  Myers M, Gay E, McNeilly AS, Fraser HM, Duncan WC 2007 In vitro evidence suggests activin-A may promote tissue remodeling associated with human luteolysis. Endocrinology 148:3730-3739  293.  Myers M, van den Driesche S, McNeilly AS, Duncan WC 2008 Activin A reduces luteinisation of human luteinised granulosa cells and has opposing effects to human chorionic gonadotropin in vitro. The Journal of endocrinology 199:201-212  294.  Hutchinson LA, Findlay JK, de Vos FL, Robertson DM 1987 Effects of bovine inhibin, transforming growth factor-beta and bovine Activin-A on granulosa cell differentiation. Biochemical and biophysical research communications 146:1405-1412  161  295.  Shukovski L, Findlay JK 1990 Activin-A inhibits oxytocin and progesterone production by preovulatory bovine granulosa cells in vitro. Endocrinology 126:2222-2224  296.  Miro F, Smyth CD, Hillier SG 1991 Development-related effects of recombinant activin on steroid synthesis in rat granulosa cells. Endocrinology 129:3388-3394  297.  Lockwood GM, Muttukrishna S, Groome NP, Knight PG, Ledger WL 1996 Circulating inhibins and activin A during GnRH-analogue down-regulation and ovarian hyperstimulation with recombinant FSH for in-vitro fertilization-embryo transfer. Clinical endocrinology 45:741-748  298.  Caron KM, Clark BJ, Ikeda Y, Parker KL 1997 Steroidogenic factor 1 acts at all levels of the reproductive axis. Steroids 62:53-56  299.  Miller WL, Strauss JF, 3rd 1999 Molecular pathology and mechanism of action of the steroidogenic acute regulatory protein, StAR. The Journal of steroid biochemistry and molecular biology 69:131-141  300.  Townson DH, Wang XJ, Keyes PL, Kostyo JL, Stocco DM 1996 Expression of the steroidogenic acute regulatory protein in the corpus luteum of the rabbit: dependence upon the luteotropic hormone, estradiol-17 beta. Biology of reproduction 55:868-874  301.  LaVoie HA, Garmey JC, Veldhuis JD 1999 Mechanisms of insulin-like growth factor I augmentation of follicle-stimulating hormone-induced porcine steroidogenic acute regulatory protein gene promoter activity in granulosa cells. Endocrinology 140:146-153  302.  Minegishi T, Tsuchiya M, Hirakawa T, Abe K, Inoue K, Mizutani T, Miyamoto K 2000 Expression of steroidogenic acute regulatory protein (StAR) in rat granulosa cells. Life sciences 67:1015-1024  303.  Ford JJ, Howard HJ 1997 Activin inhibition of estradiol and progesterone production in porcine granulosa cells. Journal of animal science 75:761-766  304.  Lambert-Messerlian GM, Hall JE, Sluss PM, Taylor AE, Martin KA, Groome NP, Crowley WF, Jr., Schneyer AL 1994 Relatively low levels of dimeric inhibin circulate in men and women with polycystic ovarian syndrome using a specific two-site enzyme-linked immunosorbent assay. The Journal of clinical endocrinology and metabolism 79:45-50  305.  Knight PG, Muttukrishna S, Groome NP 1996 Development and application of a two-site enzyme immunoassay for the determination of 'total' activin-A concentrations in serum and follicular fluid. The Journal of endocrinology 148:267-279  306.  Legro RS, Spielman R, Urbanek M, Driscoll D, Strauss JF, 3rd, Dunaif A 1998 Phenotype and genotype in polycystic ovary syndrome. Recent progress in hormone research 53:217-256  162  307.  Jacobs HS 1987 Polycystic ovaries and polycystic ovary syndrome. Gynecol Endocrinol 1:113-131  308.  Hughesdon PE 1982 Morphology and morphogenesis of the Stein-Leventhal ovary and of so-called "hyperthecosis". Obstetrical & gynecological survey 37:59-77  309.  Eppig JJ, Chesnel F, Hirao Y, O'Brien MJ, Pendola FL, Watanabe S, Wigglesworth K 1997 Oocyte control of granulosa cell development: how and why. Human reproduction 12:127-132  310.  Matzuk MM 2000 Revelations of ovarian follicle biology from gene knockout mice. Molecular and cellular endocrinology 163:61-66  311.  Erickson GF, Shimasaki S 2000 The role of the oocyte in folliculogenesis. Trends in endocrinology and metabolism: TEM 11:193-198  312.  Pigny P, Cortet-Rudelli C, Decanter C, Deroubaix D, Soudan B, Duhamel A, Dewailly D 2000 Serum levels of inhibins are differentially altered in patients with polycystic ovary syndrome: effects of being overweight and relevance to hyperandrogenism. Fertility and sterility 73:972-977  313.  Mizunuma H, Andoh K, Obara M, Yamaguchi M, Kamijo T, Hasegawa Y, Ibuki Y 1994 Serum immunoreactive inhibin levels in polycystic ovarian disease (PCOD) and hypogonadotropic amenorrhea. Endocrine journal 41:409-414  314.  Pigny P, Desailloud R, Cortet-Rudelli C, Duhamel A, Deroubaix-Allard D, Racadot A, Dewailly D 1997 Serum alpha-inhibin levels in polycystic ovary syndrome: relationship to the serum androstenedione level. The Journal of clinical endocrinology and metabolism 82:1939-1943  315.  Lockwood GM, Muttukrishna S, Groome NP, Matthews DR, Ledger WL 1998 Mid-follicular phase pulses of inhibin B are absent in polycystic ovarian syndrome and are initiated by successful laparoscopic ovarian diathermy: a possible mechanism regulating emergence of the dominant follicle. The Journal of clinical endocrinology and metabolism 83:1730-1735  316.  Fujiwara T, Sidis Y, Welt C, Lambert-Messerlian G, Fox J, Taylor A, Schneyer A 2001 Dynamics of inhibin subunit and follistatin mRNA during development of normal and polycystic ovary syndrome follicles. The Journal of clinical endocrinology and metabolism 86:4206-4215  317.  Laven JS, Imani B, Eijkemans MJ, de Jong FH, Fauser BC 2001 Absent biologically relevant associations between serum inhibin B concentrations and characteristics of polycystic  163  ovary syndrome in normogonadotrophic anovulatory infertility. Human reproduction (Oxford, England) 16:1359-1364 318.  Jaatinen R, Rosen V, Tuuri T, Ritvos O 1996 Identification of ovarian granulosa cells as a novel site of expression for bone morphogenetic protein-3 (BMP-3/osteogenin) and regulation of BMP-3 messenger ribonucleic acids by chorionic gonadotropin in cultured human granulosa-luteal cells. The Journal of clinical endocrinology and metabolism 81:3877-3882  319.  Elvin JA, Yan C, Matzuk MM 2000 Oocyte-expressed TGF-beta superfamily members in female fertility. Molecular and cellular endocrinology 159:1-5  320.  Welt C, Sidis Y, Keutmann H, Schneyer A 2002 Activins, inhibins, and follistatins: from endocrinology to signaling. A paradigm for the new millennium. Experimental biology and medicine 227:724-752  164  APPENDICES Table S1. Nucleotide sequences of primers used for quantitative real-time PCR Gene Name Inhibin α-subunit Inhibin βA-subunit Inhibin βB-subunit ACVR1A ACVR1B ACVR2A ACVR2B Smad2 Smad3 Smad4 Smad7 GDF9 FST FSTL3 StAR P450scc 3βHSD GAPDH B  Forward primer sequence 5'-GTCTCCCAAGCCATCCTTTT-3' 5'-CTCGGAGATCATCACGTTTG-3' 5'-ATCAGCTTCGCCGAGACA-3' 5'-TCATGAATTTGGCTTTTGGA-3' 5'-ATATTGGGAGATTGCTCGAAGA-3' 5'-AAAGCCCAGTTGCTTAACGA-3' 5'-TGTCAAGATCTTCCCACTCCA-3' 5'-GCCTTTACAGCTTCTCTGAACAA-3' 5'-CCCCAGCACATAATAACTTGG-3' 5'-TGGCCCAGGATCAGTAGGT-3' 5'-CGATGGATTTTCTCAAACCAA-3' 5'-CTCTTCACCCCCTGTACCC-3' 5'-TGCTCTGCCAGTTCATGG-3' 5’-CTACATCTCCTCGTGCCACA-3’ 5'-AAACTTACGTGGCTACTCAGCATC-3' 5'-CAGGAGGGGTGGACACGAC-3' 5'-GCCTTCAGACCAGAATTGAGAGA-3' 5'-ATGGAAATCCCATCACCATCTT-3'  165  Reverse primer sequence 5'-TGGCAGCTGACTTGTCCTC-3' 5'-CCTTGGAAATCTCGAAGTGC-3' 5'-GCCTTCGTTGGAGATGAAGA-3' 5'-TTTGGCAGTGTGACGCTTAC-3' 5'-GGCAGCTGATATTCTTCATGG-3' 5'-TGCCATGACTGTTTGTCCTG-3' 5'-CATGCCAGGTGTGCTGAA-3' 5'-ATGTGGCAATCCTTTTCGAT-3' 5'-AGGAGATGGAGCACCAGAAG-3' 5'-CATCAACACCAATTCCAGCA-3' 5'-ATTCGTTCCCCCTGTTTCA-3' 5'-CAGTTCCACTGATGGAAGGAT-3' 5'-CTTGACGGAGCCAGCAGT-3' 5’-TCTTCTGCAGACTCACCACCT-3’ 5'-GACCTGGTTGATGATGCTCTTG-3' 5'-AGGTTGCGTGCCATCTCATAC-3' 5'-TCCTTCAAGTACAGTCAGCTTGGT-3' 5'-CGCCCCACTTGATTTTGG-3'  166  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0071346/manifest

Comment

Related Items