Citation
Mechanisms that control follicular dominance in cattle

Material Information

Title:
Mechanisms that control follicular dominance in cattle
Creator:
Diaz Zambrano, Thais del Valle, 1957-
Publication Date:
Language:
English
Physical Description:
x, 228 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Cattle ( jstor )
Cultured cells ( jstor )
Dosage ( jstor )
Estrus cycle ( jstor )
Follicular fluid ( jstor )
Granulosa cells ( jstor )
Heifers ( jstor )
Ovaries ( jstor )
Secretion ( jstor )
Theca cells ( jstor )
Animal Science thesis, Ph. D ( lcsh )
Cattle -- Reproduction -- Endocrine aspects ( lcsh )
Dissertations, Academic -- Animal Science -- UF ( lcsh )
Ovulation -- Induction ( lcsh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1998.
Bibliography:
Includes bibliographical references (leaves 203-227).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Thais del Valle Diaz Zambrano.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
029225181 ( ALEPH )
41186756 ( OCLC )

Downloads

This item has the following downloads:


Full Text









MECHANISMS THAT CONTROL FOLLICULAR DOMINANCE IN CATTLE


By

THAIS DEL VALLE DIAZ ZAMBRANO















A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA


1998
























To my parents, Francisco and Maria Antonia (Tofia), my brother, Francisco Javier, my sisters, Ana Maria, Flor Alba, Yuritza and Elsa, and my late uncle, Juan.












ACKNOWLEDGMENTS


My endless gratitude goes to Dr. William W. Thatcher, chairman of the supervisory committee, for his guidance, insight, perseverance, trust and overall excellent graduate training. Appreciation is extended to Dr. Maarten Drost for his involvement in my project, for sharing of valuable insight, clinical expertise, and constant good spirit and for helping me keep my Veterinary spirit. Important contributions were made by additional members of the supervisory committee, Dr. Peter J. Hansen, Dr. Frank Simmen, for teaching me how to be humble, and Dr. Naser Chegini.

I want to express my eternal gratitude to Dr. R. Luzbel de la Sota, Dr. Eric JP. Schmitt, Mr. Mario Binelli, Dr. J. Divakar Ambrose and Dr. Rafael M. Roman for their friendship, scientific contributions, endless support and for sharing the joys and hardships of graduate school.

Special thanks and endless gratitude go to Mrs. Marie-Joelle Thatcher for her invaluable support during the long days of granulosa cell culture, and for sharing the ups and downs of the culture. Merci beaucoup, Marie-Joelle.

I am grateful to the people at the Dairy Research Unit of the University of Florida for their endless help, especially Mrs. Mary Russell, Mr. Winslow Eddie Fredriksson, Ms. Charlene Roomes, Mrs. Elese Griffin, Mr. James Lindsey and








especially Mr. Dale 'Valentine" Hissem, for his friendship and endless support during the long days of farm work.

Graduate school is less difficult because of the support of other graduate students and postdoctoral fellows. Special thanks go to Dr. Alfredo N. Garcia, Mr. Tombs Ignacio Belloso, Mr. Andrbs A. Kowalski, Dr. Carlos Ar6chiga, Dr. Alice de Moraes, Ms. Susan Gottshall, Ms. Rocio Rivera, Dr. Ricardo Mattos, Mrs. Jennifer Phillips Trout, Dr. Amelia Luengo, Dr. Carlos A. Vargas, Dr. Guenahel DanetDesnoyer, Dr. Florence Ndikum Moffor, Mr. Daniel Arnold, Dr. Fabiola Paula Lopes, Dr. Frederico Moreira, Dr. Lannett Edwards, Dr. Brian Cleaver, Ms. Heather Greaves, Mr. Morgan Peltier, Mr. Aydin Guzeloglu, Mr. Marvin Hoekema, Mrs. Mika Robinson, Mrs. Peggy Briggs, and Mr. Jesse W. Johnson.

Special thanks go to Mrs. Mary Ellen Hissem, for her endless smile and support, for her sincere friendship, for those lunches during the beautiful days of the fall, to Mrs. Joyce Hayen and Mrs. Helen Hester, for their constant support, friendship and sharing of the good and not so good moments. They will always have a very special place in my heart and my eternal gratitude. I want to thank Dr. H. Herbert Head, Graduate Coordinator, for his friendship and the chatting times shared, and Dr. Daniel C. Sharp for his friendship.

I want to acknowledge the Universidad Central de Venezuela, my Alma Mater, and Consejo Nacional de Investigaciones Cientificas y Tecnol6gicas for their financial support. Special thanks to my fellows at the University in Venezuela: Drs. Juan F. Troc6niz, Pedro S. Bastidas, Magaly R. Manzo, Luis A. Vasquez, Nora








Guerrero, Oswaldo A. Silva and Beatriz Quintero, for their support and understanding during these years of my absence. Mil gracias a todos.

Also, life in Gainesville was easier because of the support of very special friends Teresita, Tombs and Matias de la Sota, Carlos Ren, Carmen, Jorge Juan and Caroly Beltrdn, Elsa, Abraham, Manuel e Indira Garcia, Dr. Maria Eugenia Cadario Gonzclez-Pola, Dr. Michael Byron Porter (thank you for teaching me to enjoy the "magnificient" Colombian music), Francisco, Giordana, Johanna, Pedro and Fabianna Ovalles, Diego and Maria del Carmen Rochinotti, and Tomas Ignacio Belloso (thank you for being there during my last months in Gainesville) and thanks for making me feel part of your lives and families, and for having warm words when they were necessary. My gratitude will be forever and there is a very special place for all of you in my heart.

I have always thought that family helps you to have courage to work towards your goals. Without my family's support it would have been impossible to pursue my goals. Thank you for your faith, your love, your patience, encouragement and moral support.













TABLE OF CONTENTS

ACKNOW LEDGMENTS ......................................... iii

ABSTRACT ...... ............................................. ix

CHAPTERS

1 INTRODUCTION ......................................... 1

2 LITERATURE REVIEW .................................... 4

Ovarian Follicular Growth and Development ..................... 4
Early Stages of Follicular Development .................... 5
Ovarian Follicular Dynamics During the Bovine Estrous Cycle . . 9
Structure and Function of the Antral Follicle ..................... 21
The Granulosa Cell Compartment ....................... 23
The Theca Cell Compartment ......................... 29
Two-Cell, Two-Gonadotropin Theory ..................... 31
Molecular Events Occurring in Theca and Granulosa Cells .... 34
Hormonal Control of Ovarian Follicular Dynamics ................ 38
Steroid Hormones .................................. 38
Gonadotropic Hormones ............................. 40
Other Factors Involved in Control of Ovarian Follicular
Dynam ics .... ..................................... 43
Inhibin ........... .. .. ......................... 45
A ctivin ..... ....................................... 50
Follistatin ................. ......................... 55
Transforming Growth Factor 0 ......................... 57
Epidermal Growth Factor/Transforming Growth factor-a ...... 60
Insulin-Like Growth Factors (IGFs) and IGF-Binding Proteins
(IG FBPs) .... ..................................... 63
Manipulation of the Follicular Dynamics in Cattle ................. 66
Hormonal Manipulation: Human Chorionic Gonadotropin and
Gonadotropin Releasing Hormone ....................... 68
Ablation of the Dominant Follicle ........................ 69
Superstimulation ................................... 70








3 HUMAN CHORIONIC GONADOTROPIN-INDUCED ALTERATIONS IN
OVARIAN FOLLICULAR DYNAMICS DURING THE ESTROUS CYCLE OF H EIFERS ...... ......................................... 73


Introduction ... ........................
Material and Methods ...................
Statistical Analyses ...............
Results ... ...........................
Copora Lutea and Plasma Progesterone Follicular Dynamics ...............
Estradiol-170 Concentrations .........
Duration of Estrous Cycle ...........
Discussion .... ........................
Im plications ... ........................


4 EFFECTS OF FSH-P ON FOLLICULAR DYNAMICS AND OVARIAN
RESPONSE TO A SUPEROVULATORY TREATMENT FOLLOWING
ASPIRATION OF A FIRST WAVE PERSISTENT DOMINANT
FO LLICLE ..... ......................................... 92


Introduction ..... ..................................
Materials and Methods ..............................
Experim ent1 .................................
Experim ent2 .................................
Statistical Analyses .... .............................
R esults ..... .....................................
Follicular Dynam ics ... ........................
Plasma Progesterone and Estradiol-1713 Concentrations Ovarian Responses to Superovulation Treatment ..... D iscussion ..... ...................................
Im plications ..... ..................................


92 S94 S94
S98 S99
101 101 105 106 107 113


5 IN VITRO SECRETION OF ESTRADIOL BY BOVINE
ANTRAL GRANULOSA CELLS ......................

Introduction ..... .................................
Materials and Methods .............................
Granulosa Cell Isolation .......................
Granulosa Cell Culture ........................
Granulosa Cell Number .......................
Radioimmunoassays .........................
Gel Electrophoresis ..........................
Statistical Analyses ..........................
R esults ..... ....................................
Time Course of Basal Hormone Secretion ..........


...... 115


115 118 118
120 121 123 125 126 126 126


..................
. . . . . . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . ... .. . .......
..................
...........-.......
. . . . . . . . . . . .








Estradiol ..............
Progesterone ...........
Effect of FSH-Stimulated Secretion
Estradiol ..............
Progesterone ...........
Effect of FSH on Cell Number ....
Effect of FSH on Cell Morphology
Gel Electrophoresis ...........
Discussion ........................
Im plications .......................


6 EFFECT OF BOVINE FOLLICULAR FLUID FROM DAY 5 DOMINANT
AND DAY 12 ATRETIC DOMINANT FOLLICLES ON


IN VITRO ESTRADIOL-1713 SECRETION BY BOVINE
ANTRAL GRANULOSA CELLS ....................

Introduction .. .................................
Material and Methods ...........................
Inhibin Immunoaffinity Column ................
Pools of Bovine Follicular Fluid ...............
Electrophoresis and Western Blotting Procedures .
Granulosa Cell Culture System ...............
Statistical Analyses ..............................
R esults .. ....................................
Validation of Dextran-Coated Charcoal Treatment Western Blot Analysis of Inhibin in Follicular Fluid .
Effect of Bovine Follicular Fluid on FSH-Stimulated Estradiol-1713 Secretion .....................
Estradiol-17130 ........................
Progesterone ........................
Estradiol:Progesterone ratio ............
Discussion .. ..................................
Im plications .. .................................

7 GENERAL DISCUSSION AND CONCLUSIONS .......


REFERENCES .. ...................................

BIOGRAPHICAL SKETCH .............................


147

147 151 151 152 155 158 161 163 163 167

171 171 176 177 181 184

186


203


......... 228


126 129 130 130
134 137 138
140 141 145


......................
......................
. . . . . . . . . . . . . . . . . . . . .
......................
......................
......................
......................
......................
......................
......................












Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy


MECHANISMS THAT CONTROL FOLLICULAR DOMINANCE IN CATTLE

By

Thais del Valle Diaz Zambrano

May 1998

Chairperson: William W. Thatcher
Major Department: Animal Science

A series of experiments examined potential mechanisms that control follicular dominance in cattle. On day 5 of the estrous cycle, ablation of the first wave dominant follicle with an injection of human Chorionic Gonadotropin (hCG; 3,000 IU) resulted in ovulation of the follicle, formation of an accessory corpus luteum and higher concentrations of plasma progesterone in treated heifers. The second wave dominant follicle emerged earlier in hCG-treated heifers and all treated heifers had three follicular waves. Treatment with hCG initiated turnover of the dominant follicle and altered subsequent follicular dynamics. A second experiment tested the effect of follicle-stimulating hormone-pituitary (FSH-P) injections beginning at day 12 on folliculogenesis and ovulatory responses in the presence or absence of a first wave persistent dominant follicle. Follicle recruitment and induction of class 2 (6 to 9 mm)








and class 3 (>9 mm) follicles were delayed during the superinduction period in the presence of a persistent dominant follicle. However, no differences were detected between groups in the final number of ovulatory follicles, corpus luteum and recovered embryos. Inhibitory effects of a persistent dominant follicle were overridden with continuous injections of FSH. A granulosa cell culture system was developed to test the effects of intra- and interovarian regulators (inhibin forms) on granulosa cell secretion of estradiol. Purified porcine FSH (CY-FSH) stimulated estradiol secretion of granulosa cells between 72 to 96 h of culture. Also evaluated were effects of different fractions of follicular fluid (follicular fluid complete, follicular fluid fraction without inhibins and an inhibin-enriched fraction) from day 5 dominant and day 12 atretic dominant follicles on estradiol secretion of granulosa cells. A 10 ng/ml concentration of CY-FSH stimulated estradiol secretion compared to 0.5 ng/ml. Fractions of follicular fluid without inhibins and enriched in inhibins from day 5 dominant follicles decreased estradiol secretion. In contrast, follicular fluid complete and follicular fluid without inhibins from day 12 atretic dominant follicles inhibited estradiol secretion, but the enriched fraction of inhibins stimulated granulosa cells to secrete estradiol. Collectively, the series of in vivo and in vitro experiments documented that the dominant follicle exerts both interovarian and intrafollicular effects on follicle function in cattle.













CHAPTER 1
INTRODUCTION



On the basis of gross and histological studies of ovaries, it was proposed in 1960 that two waves of follicular activity occurred during the bovine estrous cycle (Rajakoski, 1960). After measuring follicles and quantifying steroids in blood and follicular fluid, Ireland and Roche (1983) concluded that there were three follicular waves during the estrous cycle, and each wave resulted in the formation of a dominant follicle. In 1984, Pierson and Ginther, characterized bovine ovarian folliculogenesis, for antral follicles > 3 mm by repeated daily ultrasound measurement and concluded that heifers had estrous cycles with two follicular waves. However, others (Fortune et al., 1988; Savio et al., 1988, Sirois and Fortune, 1988) reported three waves of follicular development during the estrous cycle. Research in this area is motivated by the desire to solve the long-time mystery involving the mechanisms underlying recruitment, selection and development of follicular dominance that is characteristic of monovulatory species (Ginther et al., 1996).

In general two processes lead to development of the normal, species-specific number of ovulatory follicles. First, follicle recruitment results in the development of a pool or cohort of follicles from which the preovulatory follicle emerges as a







2

result of follicle selection. The follicle becomes dominant and continues development towards ovulation, while others regress (Fortune et al., 1991). This pattern of dominant follicle development also occurs during the luteal phase of the estrous cycle in which dominant follicles undergo atresia instead of ovulation. The phenomenon of dominance is clearly of interest because mechanisms of follicular dominance must be subverted to successfully superovulate domestic animals or humans. Furthermore, an understanding of dominant follicle development and dominance is important to the development of reproductive control system to synchronize ovulation in dairy cattle.

At present, little is known about mechanisms by which a follicle attains and maintains dominance (Fortune, 1993). To study follicular dominance, differences between morphological and functional dominance have to be established. Morphological dominance is defined as the largest developing follicle on either ovary. The functionally dominant follicle appears to have the ability to inhibit growth of smaller follicles and can ovulate under appropriate hormonal conditions (Fortune, 1993).

Basic research is needed to elucidate the mechanisms controlling dominance during follicular development in cattle to optimize the techniques to control reproductive cycles in cattle. Objectives of the present research were: a) to characterize alterations in ovarian follicular and corpus luteum dynamics after injection of hCG on day 5 of the estrous cycle; b) to evaluate the effect of folliclestimulating hormorne-pituitary (FSH-P) on superinduction of follicles and ovulatory responses in the presence or absence of an active-first wave persistent dominant







3

follicle, and to elucidate the role of inhibins as intraovarian regulator of estradiol secretion during different periods of follicular dominance. The last objective require development of an in vitro system to examine regulation of estradiol secretion by bovine antral granulosa cells.













CHAPTER 2
LITERATURE REVIEW


Ovarian Follicular Growth and Development


Reproductive cyclicity in the female is maintained by endocrine, paracrine, and autocrine interactions. The central role of pituitary follicle-stimulating hormone (FSH) and luteinizing hormone (LH) is to direct ovarian follicular recruitment, growth, atresia, and ovulation (Richards and Hedin, 1988). The ovary has two main functions: production of the oocyte, the female gamete, and to provide the hormonal environment to sustain reproductive function through the production of steroid hormones, such as progesterone (P4), estradiol-1713 (E2), and other regulatory factors. Gonadotropin secretion influences, among other things, development of follicles in the ovary. The factor or factors that determine the early stages of follicular growth in the ovary at a pre-determined time are unknown. The fate of a follicle could be to remain quiescent, to begin development but later become atretic, or to mature and ovulate. In mammals the number of primordial follicles present in the ovaries is fixed at birth or in the immediate post-natal period depending on the species. In the cow, between 40,000 to 80,000 quiescent primordial follicles can be found at birth, but this number is reduced to approximately 3,000 by 15 to 20 years of age (Erickson, 1966). Since 99% of the primordial follicles fail to ovulate during









the lifespan of an animal, development of an ovulatory follicle is a rare biological event and folliculogenesis is a complex process (Ireland, 1987).


Early Stages of Follicular Development


The initial stages of folliculogenesis occur independently of gonadotropic hormones. Antral follicles initially become responsive to FSH and then are dependent on FSH for sustained growth. Inhibins, activin, insulin-like growth factor I (IGF-1) and their binding proteins have direct and indirect effects on granulosa and theca cells that can modulate follicular development and steroidogenesis (Roche, 1996).

Ginsburg et al. (1990) demonstrated that primordial germ cells in the mouse appear by day 7 post coitum in the extra embryonic mesoderm posterior to the primitive streak. These cells migrate to the gonad by day 12 post coitum. Transforming growth factor-1 (TGF- 1) produced by the gonad appears to be one of the factors responsible for the chemotropic attraction of primordial germ cells (Godin and Wylie, 1991).

According to Hilscher (1991), female germ cells undergo only one proliferative wave of oogonial divisions, whereas gametogenesis in the male involves two proliferative waves. In the bovine ovary, onset of meiosis occurs at approximately day 70 of embryonic life (Erickson, 1966). However, meiosis is not completed in the embryonic ovary, and germ cells are arrested in the diplotene stage of prophase I. Conversion of oogonia into primary oocytes depends on contact with cells derived from the rete ovarii. The first germ cells to enter meiosis







6

are located at the inner border of the ovary establishing a two-way communication between germ and somatic cells. The first follicles are formed around day 70 of pregnancy in the ewe and day 130 of pregnancy in the cow (Mariana et al., 1991).

The majority of mammals restrict oogonial proliferation to the prenatal period of development (e.g., pigs and ruminants) or to shortly after birth during the early neonatal period (e.g., rodents and rabbits; Van den Hurk et al., 1997). Oogonia are transformed into primary oocytes characterized by a prolonged meiotic prophase and surrounded by a layer of flattened pregranulosa cells. When a full layer of cuboidal granulosa cells is acquired around the oocyte, the follicle becomes an intermediary and then a primary follicle. These primordial, intermediary and primary follicles constitute the resting stockpile of nongrowing follicles located at the periphery of the ovarian cortex and form the bulk of follicles contained in the ovary (Driancourt et al., 1993b). These follicles are progressively depleted during the reproductive life span. Initiation of ovarian function in the cow occurs during fetal development. Primordial follicles start to grow in response to an unknown trigger (Webb et al., 1992) when they continue development until ovulation or atresia. The primordial follicle possesses an oocyte surrounded by a single granulosa layer consisting of 14-29 flattened granulosa cells in cattle, which is in turn encompassed by a basement membrane (Van den Hurk et al., 1997). It has been reported that initiation of follicle growth is not dependent on gonadotropins. It has been estimated that it takes approximately 6 months for a primordial follicle to develop into a large dominant follicle in sheep (Cahill and Maul6on, 1980). In the cow, the estimated







7

time for a follicle to grow from 0.13 to 8.56 mm is approximately 41.5 days, the equivalent of two estrous cycle (Lussier et al., 1987).

One of the most critical steps in folliculogenesis is the transformation of primordial follicles into primary follicles. The growth of a follicle appears to begin with enlargement of the primary oocyte, the proliferation of surrounding granulosa cells, and the organization of thecal cells external to the basement membrane (Paton and Collins, 1992). The nature of the signal to convert flattened pregranulosa cells into a cuboidal epithelium is not clear. However, the oocyte provides cues to this conversion (Greenwald and Roy, 1994). By a series of mitotic divisions, the unilaminar primary follicle is converted into a multilayer preantral secondary follicle. Each primordial follicle contains a single layer of pregranulosa cells resting on a basal lamina and radially arranged around an oocyte which is arrested in diplotene stage of meiosis (Paton and Collins, 1992). Interstitial cells surround the follicle in immediate proximity to the basal laminae and start to differentiate into theca cells when follicular growth starts. During the early stage of a secondary follicle, connective tissue fibers are arranged parallel to the basement membrane underneath the granulosa to form a thecal layer. At the same time , a glycoprotein coat, the zona pellucida, is formed between the growing oocyte and the innermost granulosa layer. At the end of this stage, hormone producing large epithelioid cells and a capillary network additionally constitute the theca (Van den Hurk et al., 1997). With the appearance of an antral cavity, the secondary follicle is converted into a tertiary follicle. It is frequently stated that gonadotropic hormones are not necessary for early growth of follicles and that FSH and LH








8

become indispensable for further development only at the transformation of the secondary to the tertiary follicle. The relative constant number of preantral follicles throughout the estrous cycle of the mouse, is evidence that cyclic changes in gonadotropins do not affect the pool of preantral follicles. However, in the rat, rhesus monkey, and hamster, the number and responsiveness of preantral follicles to gonadotropins do change during the estrous cycle. In the human ovary, longterm follicular growth is severely impaired in the absence of gonadotropins (Greenwald and Roy, 1994). In the hypophysectomized ewe follicular development is arrested at a stage between 2 to 3 mm in diameter (McNatty et al., 1990).

Schematically, antral follicular development involves two phases. In the first phase, early antral follicles grow slowly (up to 2 mm in sheep and 3 - 4 mm in the cow) and follicular growth rate is related to the proliferation rate of granulosa cells, and is not really dependent on gonadotropin supply. The second phase or growth is dependent on gonadotropin secretion. This phase is characterized by rapid follicular growth, mainly due to enlargement of the antrum, and corresponds to terminal development of antral follicles up to the preovulatory stage. During this phase the steroidogenic capacity and the responsiveness of granulosa cells to FSH and LH increases (Monniaux et al., 1997). It has been shown (Adams et al., 1992b) in domestic ruminants that terminal follicular growth is associated with fluctuations of FSH concentrations, which suggests a stimulating effect of FSH on emergence of each follicular wave.

Goodman and Hodgen (1983) suggested the terms of recruitment, selection and dominance to describe the development of antral follicles. Recruitment is a







9

gonadotropin-dependent event during which a group of follicles gain the ability to respond to gonadotropin and develop dependance on them for continued growth. Selection is a process whereby only a few of the recruited follicles are selected to escape atresia and survive to ovulate. In cattle, selection is defined as the time when an estrogen-active follicle promotes its growth and inhibits the growth of other follicles (Sunderland et al., 1994). Factors other than size and E2 concentrations, such as P4 may be important for establishing which follicle becomes dominant during the selection phase (Sunderland et al., 1994). Xu et al. (1995b) suggest that selection of the dominant follicle may be a passive process in which the first follicle that acquires LH receptors in its granulosa cells is selected to become the dominant follicle. The acquisition of LH receptors in granulosa cells will enable these follicles to respond to LH in addition to FSH, whose concentrations in blood have declined to basal levels at the time of selection (Adams et al., 1992b). Dominance is the mechanism that an ovulatory follicle uses to escape atresia and inhibits recruitment of a new cohort of follicles. Expression of LH receptor in granulosa cells is associated with dominance (Xu et al., 1995a; Yuan et al., 1998).


Ovarian Follicular Dynamics During the Bovine Estrous Cycle


Antral follicle development was considered originally to be a continuous state of turnover without a distinct pattern of follicular growth, regression and atresia (Marion et al., 1968). A classical study of Rajakoski (1960) along with studies by Matton et al. (1981) indicated that at least two periods of turnover of antral follicles occur during the estrous cycle of the cow. Later during the 1980's daily ultrasound







10

scanning of ovaries confirmed that generally two, but sometimes only one nonovulatory follicle wave occurs in the luteal phase of the cycle before final development of the ovulatory dominant follicle after luteal regression (Pierson and Ginther, 1984; Fortune et al., 1988; Savio et al., 1988; Sirois and Fortune, 1988; Ginther et al., 1989a; Lucy et al., 1992; Sunderland et al., 1994; Roche, 1996).
Unlike primates (Goodman and Hodgen, 1983), non-ovulatory follicles in cattle develop during the early- and mid-luteal phase of the estrous cycle (Rajakoski, 1960; Ireland, 1987). This is reflected by the increase in E2 concentrations in blood a few days after ovulation and during the mid-luteal phase. Each follicular wave is characterized by a simultaneous emergence of a group of 5 to 7 follicles (> 5 mm in diameter) from the pool of small follicles. One from this group rapidly emerges and grows larger than the others in the cohort and is considered the "dominant" (Fortune, 1993), while the others become atretic and regress. The dominant follicle normally reaches a maximum size of 10 to 15 mm in diameter and remains dominant for a period of 5 to 7 days, until it becomes atretic and regresses in size. The regressing dominant follicle is replaced with a new dominant follicle grown from the next wave of follicles. If luteal regression occurs during the growth phase or early period of dominance, then the dominant follicle, free from the restrictive hormonal milieu imposed by P4 secretion from the corpus luteum upon the hypothalamus and pituitary gland, will continue to develop to preovulatory size and trigger the hormonal cascade leading to ovulation (Webb et al., 1992).







11

Follicle growth is a process during which the follicle progressively acquires a number of properties, each of which is an essential prerequisite for further development. Failure to acquire these properties at the correct time and in an exact sequence will lead to failure of the process and to deterioration of the follicle through atresia (Campbell et al., 1995). Antrum formation occurs at a follicular diameter (fixed ovaries) of 0.14-0.28 mm in cattle, and it takes 40 days for follicles to reach ovulatory size (Lussier et al., 1987). In cattle there is a marked hierarchy in the antral follicle population with a large number (20-30 follicles of 3-4 mm in diameter) of gonadotropin-responsive follicles (follicles that can respond to gonadotropin stimulation, but do not need gonadotropins), a few (1-4 follicles of > 4-5 mm) gonadotropin-dependent follicles (follicles that need gonadotropins to further develop) and one ovulatory follicle (Campbell et al., 1995).

Follicular dynamics can be defined as the process of continual growth and regression of antral follicles that leads to the development of the preovulatory follicle (Lucy et al., 1992). Ireland (1987) hypothesized that the turnover of dominant follicles during the estrous cycle is regulated by the differential response of selected and unselected follicles in the cohort to alterations in patterns of secretion of gonadotropins, which in turn, result in a differential production of intrafollicular stimulatory or inhibitory factors that control selection, dominance and atresia. Analysis of patterns of development of large follicles in mammalian species shows that large follicles do not develop at random, but their development occurs only during particular reproductive states and/or during particular times of the reproductive cycle (Fortune, 1994). This pattern of follicle development is







12

associated with follicular changes in expression of mRNA encoding gonadotropin receptors (Xu et al., 1995a) and steroidogenic enzymes (Xu et al., 1995b) that allow selected follicles, when exposed to the requisite hormonal environment, to develop and ovulate in response to the preovulatory gonadotropin surge. Endocrine signals such as gonadotropins, inhibin, and steroids, as well as locally produced growth factors, such as insulin-like growth factor I (IGF-I), transforming growth factor a (TGF-a), transforming growth factor B (TGF-3), epidermal growth factor (EGF), and other peptide hormones such as activin, and follistatin, are responsible for the control and coordination of these process (Armstrong and Webb, 1997).

Early during the estrous cycle a cohort of follicles is recruited out of the pool of smaller antral follicles (2 to 4 mm). Recruitment is not a random or isolated phenomenon. Follicles seem to be recruited as groups or cohorts, suggesting that they have received a signal that allows them to continue growth and development rather than regress (Fortune, 1994). The mechanism that controls recruitment of these small follicles and determines which follicles are recruited is unknown. The signal that stimulates recruitment appears to be a slight elevation in plasma FSH (Fortune, 1994). Xu et al. (1995a) hypothesized that changes in expression of mRNA for FSH and LH receptors may be important for recruitment of a cohort of follicles and selection and atresia of the dominant follicle in cattle. However, the steady state of FSH receptor mRNA level in healthy follicles did not correlate with follicle size, nor did the level of FSH receptor mRNA change with stage of the first follicular wave. Adams et al. (1992b) reported that 2 to 4 days before a new wave of follicle development there is an increase in FSH, which suggests that increases







13

in circulating concentrations of FSH initiate the emergence phase for dominant follicle growth. The number of follicles recruited is usually greater than the typical number of ovulatory follicles for a given species. However, only a species-specific number of ovulatory follicles continues to grow for more than a few days and reaches ovulatory size (Fortune, 1994).

After 2 to 4 days of recruitment (days 2 to 4 of the estrous cycle), several medium-sized follicles (6 to 9 mm) can be detected by ultrasonography. This is the phase of selection in which a single follicle emerges from the pool of recruited follicles and continues to grow, whereas other recruited follicles decrease in size (Lucy et al., 1992). Sunderland et al. (1994), using ultrasound analysis and the ratio of E2:P4 concentrations in follicular fluid, reported that days 1 to 3 of the estrous cycle are the selection phase for development of the early diestrous first wave dominant nonovulatory follicle. Ginther et al. (1996) defined time of deviation (or selection) as the beginning of the greatest difference in growth between the two largest follicles in the ovary. They indicate that the term deviation of follicles is a major event in the selection process, and the terms deviation and selection are synonymous. Moreover, deviation was characterized by a slower rate of growth for 1 to 4 days before the subordinate follicle attained maximum diameter and by the immediate cessation of growth of the largest subordinate follicle (Ginther et al., 1996; 1997).

The completion of the selection phase is defined in the cow as the time when an estrogen-active follicle promotes its own growth and inhibits the growth of other follicles (Sunderland et al., 1994). Adams et al. (1993b) showed that the decrease








14

in circulating concentrations of FSH is an integral component of the selection mechanism. Exogenous FSH given for 2 days at the expected time of selection delayed selection as indicated by a 1.5 day delay in significant divergence of the growth profiles of the dominant and first subordinate follicles. There was greater growth and delayed regression of the first and second subordinate follicles. The same treatment regimen given after selection did not alter follicular development and atresia. However, it is not known how the FSH decline exerts its effect on selection. Perhaps selection involves competition among the cohort of follicles for utilization of FSH whereby the most successful follicle becomes dominant. This competition becomes acute when FSH levels decline. Results from Xu et al. (1995a; 1995b) suggest that selection of the dominant follicle may be a passive process in which the first follicle that acquires LH receptors in its granulosa cells is selected to become the dominant follicle. The acquisition of LH receptors in granulosa cells will enable these follicles to respond to LH in addition to FSH. Indeed concentrations of FSH in blood have declined to basal levels at the time of selection (Adams et al., 1992b).

On approximately day 5, growth of usually only one follicle is sustained while growth rate of other follicles declines. This first wave dominant follicle remains dominant from day 5 to = day 9 (Driancourt et al., 1991a), and during this period of follicular dominance no new follicles > 5 mm are detected within the ovaries (Lucy et al., 1992; Fortune, 1993). Therefore, a dominant follicle is defined as a large ovarian follicle (> 10 mm) that is recruited and selected during a follicular wave, and regulates growth of other follicles on the ovary (Lucy et al., 1992). The dominant







15

follicle would somehow cause demise of follicles of the same cohort and would also suppress a new wave of follicular development. At present little is known about the mechanisms by which a follicle attains and maintains dominance (Fortune, 1993). Perhaps a follicle is selected for dominance because it is in the right stage of development and is better able to respond to the slight elevation of and the subsequent decline in FSH to continue its growth (Fortune, 1994). The ability of the dominant follicle to continue growth and development in an environment of lower levels of FSH may be due to increased blood flow and/or to the acquisition of LH receptors by the granulosa cells (Zeleznick, 1993). Xu et al. (1995b) showed that theca interna cells of dominant follicles collected on day 4 after initiation of the first follicular wave had the highest levels of mRNAs for P450, and P450c1l7, 20 lyase, which ensure that the theca interna cells of these follicles are capable of producing large amounts of androgen substrates for E2-170 biosynthesis in the granulosa cells. The acquisition of LH receptors in granulosa cells may be critical to the establishment and maintenance of follicular dominance, whereas FSH receptors may only play a permissive role (Xu et al., 1995a). Large antral follicles can transfer their gonadotropic requirements from FSH to LH. This transition in gonadotropic requirement from FSH to LH is the probable mechanism whereby the preovulatory follicle can withstand the fall in FSH that occurs at the onset of the follicular phase following luteal regression (Campbell et al., 1995).

The follicle selected for dominance from each wave not only continues to grow, but also differentiates functionally in ways that prepare it for ovulation. The secretion of increased quantities of E2 by the selected follicle appears to be of







16
primary importance and sets it apart from its sister subordinate follicles. Feedback regulators, such as E2 and inhibin produced by the dominant follicle (or perhaps by the whole cohort of follicles during the first few days after recruitment), cause a decrease in FSH concentrations that will not even support the further growth of subordinate follicles (Fortune, 1994).

It is important to distinguish between morphological and functional dominance. Morphological dominance is characterized by size of the follicle, with the largest follicle present on a pair of ovaries being defined as morphologically dominant. Functional dominance appears to have two aspects: ability of the dominant follicle to inhibit the growth of smaller follicles and the capability to ovulate under appropriate hormonal conditions (Fortune, 1993). The largest follicle of the first and second wave of cows with three waves is by definition morphologically dominant (Fortune et al., 1991; Fortune, 1993). The first wave dominant follicle at day 12, despite being the largest follicle, had lost its functional dominance as early as day 10 based on emergence of a new wave of follicular development (Sunderland et al., 1994). These results and others (Savio et al., 1990a) show that morphological dominance of a follicle persists much longer than functional dominance. Driancourt et al. (1991) reported that dominance is characterized by the large difference in diameter between the largest and the second largest follicles and the decrease in the number of follicles smaller than 8 mm.

Two hypotheses have been postulated to explain how the dominant follicle exerts dominance. One hypothesis states that the dominant follicle secretes a product that directly impairs further growth and development of subordinate follicles.







17

In monotocous species, like cattle, such a factor would have to be endocrine in nature, since it would induce regression of subordinate follicles on both ovaries. The other hypothesis states that the dominant follicle could cause the regression of subordinate follicles indirectly, via negative feedback mechanisms in which E2 and inhibin would cause a decrease in FSH concentrations that would not support further growth of subordinate follicles (Fortune, 1994). A third hypothesis, which explains dominance from an intraovarian point of view, states that production of local ovarian factors such as the insulin-like growth factor (IGF) system, components of the transforming growth factor 13 superfamily, fibroblast growth factor, and the epidermal growth factor/transforming growth factor a family can inhibit the development of subordinate follicles directly (Campbell et al., 1995; Armstrong and Webb, 1997).

Sunderland et al. (1994) indicated that the period of days 1 to 3 of the estrous cycle is the selection phase for development of the early diestrous dominant nonovulatory follicle; whereas days 10 to 12 is boyh the selection phase for development of the next dominant follicle and the period when the first dominant follicle ceases to function or loses its dominance, becomes estrogen-inactive and begins to undergo atresia. In two-wave cycles, maturation of the second dominant follicle coincides with spontaneous regression of the corpus luteum, and the follicle ovulates after luteolysis. Alternatively, the second wave dominant follicle may become atretic and a third follicular wave will be initiated (Lucy et al., 1992) and becomes the ovulatory follicle.







18
Sirois and Fortune (1988) reported that follicular waves during the estrous cycle occur at approximately 7 day intervals. For three-wave cycles the first, second and third waves start on days 2, 9 and 16, respectively, for two-wave cycles the second wave started about 2 days later than the average for animals with three waves (day 11; Fortune et al., 1991). Waves of follicular development in cattle occur regularly during both the estrous cycle and pregnancy. During the estrous cycle, the preovulatory gonadotropin surge perturbs the pattern of regular waves by triggering an abrupt ovulation of the follicle that is functionally dominant at the time of estrus.

More than 99% of ovarian follicles undergo the degenerative process of atresia during reproductive life (Hsueh et al., 1994). If the ovulatory signal is absent, the mature follicle undergoes degeneration. It has been shown that the dominant follicle of the first wave, which normally undergoes atresia, can ovulate provided that the ovulatory stimulus is given before it reaches the advanced regression stage of development (Savio et al., 1990a; Sirois and Fortune, 1990).

In antral bovine follicles > 1 mm in diameter, the earliest and most prominent feature of atresia is death of granulosa cells (presence of pycnotic nuclei) that leads to almost total destruction of the granulosa cell layer lining the inner follicle wall (Rajakoski, 1960). Atresia of the dominant nonovulatory follicle is characterized by a significant decrease in the number of granulosa cells, a decrease in both LH and FSH receptors (Ireland and Roche, 1983), and a diminished capacity to produce E2 between days 7 and 13 of the estrous cycle (Badinga et al., 1992; de la Sota, 1995). The factors involved in regulation of follicular atresia are not clear, but it has been








19

demonstrated that decreasing the LH pulse frequency to a luteal phase secretory pattern results in faster atresia of the dominant follicle (Sirois and Fortune, 1990). Thus the dominant follicle may avoid atresia when exposed to a higher LH pulse frequency during the follicular phase. Induction of a low progesterone environment (<2 ng/ml) results in increased growth or size of the dominant follicle, higher plasma concentrations of E2 and a greater lifespan or persistence (Savio et al., 1993a). In addition to sustained follicular growth, functional dominance also was observed as there was a suppression in the number of other follicles > 4 mm in diameter (Savio et al., 1993a). Stock and Fortune (1993) reported that slight increases in LH pulse frequency promoted prolonged follicular growth and dominance associated with increased plasma E2. They suggested that the demise of nonovulatory dominant follicles during normal estrous cycles occurs through feedback effects of luteal progesterone, which maintains a low LH pulse frequency and E2 production.

Based mainly on morphological criteria, atresia of antral follicles can be divided into several stages (Hsueh et al., 1994). Stage I is characterized by a small number of granulosa cells (< 10%) with pyknotic nuclei close to the follicular cavity; however, some of the granulosa cells are still undergoing mitosis. Stage II shows many pyknotic granulosa cells (10-30%), few cells in mitosis, and cell debris in the follicular cavity. The basement membrane loses its integrity, and there is a leukocytic infiltration into granulosa cell layers. In the oocyte, meiotic-like changes are evident. Stage III is characterized by a reduction in granulosa cell number, an absence of mitosis, and collapse of the follicle. The oocyte also undergoes germinal vesical breakdown (Hsueh et al., 1994).







20

Fate of the thecal cell layer during atresia varies among different species. In the human, rat, and rabbit, theca cells undergo extensive hypertrophy during follicular atresia. In contrast, hamster follicles do not exhibit marked morphological changes despite sharply reduced follicular vascularity and total collapse of the granulosa layer. In the sheep, theca cells undergo nuclear condensation and degeneration similar to those observed in granulosa cells (Hsueh et al., 1994).

Biochemical changes also occur during atresia. These changes include reduced DNA synthesis of granulosa cells, suppressed expression of connexin, a gap junction protein, decreased gonadotropin binding, decreased expression of mRNAs for aromatase and gonadotropins receptors, and decreased E2 synthesis concomitant with increased progesterone production. The latter change appears to be due to both a decrease in C17, 20-lyase activity, that leads to a decrease in androgen substrate for granulosa cell aromatization, and a loss of aromatase activity (Hsueh et al., 1994).

During atresia, there is an increase in expression of several genes, including the IGF-binding proteins (IGFBP), which is related to a deprivation of endogenous IGF-I that is essential for survival of follicle cells (Hsueh et al., 1994). Recent biochemical evidence has demonstrated apoptosis of granulosa cells during follicular atresia in bovine ovaries (Jolly et al., 1994). These authors suggested that apoptotic death of granulosa cells in the cow may also occur in healthy follicles during the luteal phase of the estrous cycle and/or occur early in the atresia process before other morphological or biochemical signs of degeneration or dysfunction are








21

evident. Thus a variable incidence of apoptotic cell death may occur at different stages of follicular development.


Structure and Function of the Antral Follicle


The primary function of the mammalian ovarian follicle is the release of an oocyte capable of being fertilized by sperm. This involves growth and maturation of the follicle as well as enlargement, ovulation, and meiotic division of the egg (Richards, 1980). The identification of the ovarian follicle occurred in 1672 by Regnier de Graaf, and mature preovulatory follicles are named Graafian follicles. It was not until 1827 that Karl Ernst von Baer discovered that the mammalian egg was a tiny part of the entire follicular structure. In the early 1900's the gonads as well as the pituitary were recognized as organs of internal secretion (Richards, 1980).

A single follicle consists of a diplotene oocyte and associated somatic cells surrounded with a basal lamina. Small follicles have a single layer of follicular cells and form the largest number in the ovary. The ovarian follicle consists of outer layers of theca cells which encircle inner layers of granulosa cells, which are separated from blood vessels and theca cells by a basement membrane lining the follicle. Inner layers of granulosa cells surrounding the oocyte are designated as the oocyte-cumulus cell complex (Amsterdam et al., 1989). Follicular growth ends with ovulation or atresia. The growth of the follicle is ensured by the growth of the oocyte, follicular cell proliferation, and enlargement of the antrum (Mariana et al., 1991). Follicles appearto begin growth under all physiological conditions, including







22

pregnancy, ovulation, or periods of anovulation (Richards, 1980; Amsterdam and Rotmensch, 1987). The number of follicles beginning to grow each day changes throughout life and appears related primarily to the number of follicles in the nongrowing or resting pool. The larger the pool, the greater the number of follicles beginning to grow. With age, as follicles continuously leave the pool of nongrowing follicles, the number of follicles in the resting pool becomes drastically reduced and ultimately is exhausted. This reduction is associated with a decrease in the number of follicles that begin to grow and eventually ovulate (Richards, 1980).

The first sign of growth is the resumption of cell proliferation by the squamous granulosa cells, an increase in the size of the oocyte and a change in shape of the granulosa cells (Hirshfield, 1991). In the adult ovary, folliculogenesis starts when follicles leave the pool of resting follicles to enter the growth phase. From there, the early growing follicle undergoes a developmental process including a dramatic course of cellular proliferation and differentiation (Gougeon, 1996). The oocyte and the follicle grow in a biphasic manner. During the first phase, which ends shortly before antrum formation, growth of oocyte and granulosa cells is linearly and positively correlated. The second phase consists of enhanced mitotic activities of granulosa and thecal cells, and an increase in volume of follicular fluid (Greenwald and Roy, 1994). The oocyte enlarges rapidly during early follicular growth and reaches full size early in the developmental process (Hirshfield, 1991). Later in development, the growth of the oocyte is slow compared to growth of the follicle. Growth of the oocyte involves accumulation of proteins either through direct synthesis within the oocyte or by transfer from the antrum. During the active phase







23
of oocyte growth, content of proteins is increased 100-fold and RNA content enhanced 20-fold. RNA synthesis is increased in the nucleus of the growing oocyte compared to that found in the primordial oocyte. Synthetic activity decreases steadily in oocytes from follicles with an antrum (Mariana et al., 1991).

The follicle acquires several distinctive morphological features during growth: theca interna (steroidogenic cells) and theca externa (connective tissue cells) which form the outer layers of the follicle, a fluid-filled antral cavity, a capillary network, and the thick acellular zona pellucida surrounding the oocyte (Hirshfield, 1991). As the follicle enters its final phase of growth, fluid-filled spaces appear between the granulosa cells which soon coalesce into a large single fluid-filled antral cavity. The accumulating fluid appears to be formed by filtration of thecal blood. The composition of follicular fluid differs considerably from plasma, with lower glucose and lipid concentrations in follicular fluid. There are also differences in amino acid concentrations (Hirshfield, 1991).


The Granulosa Cell Compartment


The ovarian follicle has the unique capacity to change in structure and function to influence the cyclic processes of female reproduction. A finely tuned process of differentiation occurs in all constituents of the follicle during folliculogenesis. Granulosa cells play an obligatory role in creating the conditions necessary to resume oogenesis, ovulation, fertilization and implantation (Paton and Collins, 1992). Granulosa cells display a high degree of structural change and play a key role in the functional maturation of the entire follicle.








24

The early development of the female gonad is characterized by the migration of a mitotically active cohort of germ cells from the region of the embryonic hindgut and endoderm of the yolk sac to the gonadal ridge. As germ cells enter meiosis, granulosa cells seem to arise from the rete ovari or from mesonephric tubules or mesenchymal cells in the gonadal anlage and they organize around the primary oocytes (Amsterdam and Rotmensch, 1987). The development of the steroidogenic potential in differentiating granulosa cells consists of early and late events (Amsterdam et al., 1989). The early events are related to mobilization of cholesterol, reorganization of endosomes, lysosomes, lipid droplets and particularly mitochondria and smooth endoplasmic reticulum, which are the carriers of membrane-bound enzymes converting cholesterol to a variety of steroid hormones. Late events consist of stable changes in cell shape, intercellular communication through gap junction formation and development of organelles associated with steroidogenesis (Amsterdam et al., 1989).

Granulosa cells from the various layers within the same follicle display morphological heterogeneity suggesting the existence of a differentiation gradient in the follicle. Cells adjacent to the follicular basement membrane contain a higher density of LH receptors than cells located in the inner layers of the pre-ovulatory follicle (Amsterdam et al., 1989). The granulosa cell layer surrounding the oocyte, called the corona radiata, produces elongated cytoplasmic processes which traverse the zona pellucida towards the oocyte and form gap junctions at the area of contact. Due to the nature of a nonvascular environment of granulosa cells, they need a way of intercellular communication between neighboring cells. In addition








25

to a complex paracrine system of membrane-soluble hormones, there is also a direct cell-to-cell communication via gap junctions (Amsterdam and Rotmensch, 1987). Gap junctions are cellular structures between adjacent cells, wherein the apposed cellular membranes are separated by an apparent gap of approximately 3 nm. Gap junctions are formed by connexins. They allow communication or electrical coupling between adjacent cells that can be open or closed. Gap junctions are composed of two symmetrical structures (connexons) that create an intracellular channel that allows passage of ions and small molecules from cell to cell (Grazul-Bilska et al., 1997). Also adherence junctions have been described in granulosa cells. They are characterized by association of filamentous dense material on the intracellular aspect of the plasma membrane (Amsterdam and Rotmensch, 1987). Gap junctions allow ions and small, water soluble molecules to pass directly from the cytoplasm of one cell to the cytoplasm of another, thereby coupling the cells both electrically and metabolically (Amsterdam et al., 1989; Paton and Collins, 1992). Gap junctions are absent in the primordial follicle, where desmosomes and adherence junctions are the dominant structures comprising the junctional complex (Amsterdam and Rotmensch, 1987; Amsterdam et al., 1989). Gap junctions appear during differentiation of the antral follicle, and parallel the onset of gonadotrophic action on the follicle (Amsterdam and Rotmensch, 1987; Amsterdam et al., 1989; Paton and Collins, 1992). Molecules of 1200 daltons and smaller (e.g., inorganic ions, amino acids, nucleotides, and sugars) can pass through the junctional channels, and move from cell to cell. For example, cyclic AMP (cAMP) has been identified as a possible mediator of biological function







26

through gap junctions. The functional significance of gap junctions between granulosa cells may lie in propagating cAMP-mediated events such as the response to FSH or LH, thereby allowing the signal to reach the interior of the avascular preovulatory follicle (Amsterdam and Rotmensch, 1987). Changes in the degree of dispersion of granulosa cells during follicular maturation seem to be related to quantitative changes in intercellular junctions. Biochemical differentiation of granulosa cells is associated with gap junction formation in response to FSH stimulation. Morphological studies suggest that cAMP might be associated with changes in cell shape and organization of the cytoskeleton (Ben-Ze'ev and Amsterdam, 1987).

Granulosa cells display a high degree of structural change during the differentiation process. There are alterations in cellular shape and size, in plasma membranes, and in internal membranes of the endoplasmic reticulum, mitochondrial cristae, intercellular junctions, and cytoskeleton. Granulosa cells are flat and epithelioid in the earliest stage of development. When proliferation starts, the flattened layer of granulosa cells becomes a multilayered stratified epithelium consisting of cuboidal cells (Paton and Collins, 1992). There is also a change in the type of intracellular organelles. Immature granulosa cells contain mainly a rough endoplasmic reticulum, which reflects a high level of protein synthesis required for growth and proliferation. Smooth endoplasmic reticulum increases its development when steroidogenic potential increases (Paton and Collins, 1992). Complex tubular mitochondrial cristae is typical of FSH-stimulated cells, whereas a lamellar type is observed in immature cells (Amsterdam and Rotmensch, 1987).








27

The cytoskeleton mediates many of the structural alterations that coincide with differentiation-dependent changes in the metabolic activity of granulosa cells (Paton and Collins, 1992). Cytoskeletal elements are believed to be involved in aggregation and internalization of ligand-bound gonadotropin receptors (Amsterdam and Rotmensch, 1987). These elements provide the machinery for moving organelles from one place to another when granulosa cells are exposed to gonadotropins and movement of substrate among organelles also needs to be facilitated to enhance steroid synthesis (Paton and Collins, 1992). The regulation of granulosa cell differentiation at the level of the cytoskeleton seems likely to prove as complex as other intraovarian regulatory mechanisms. One important mechanism by which organization of cytoskeleton can be regulated is by the activity of transmembrane proteins which link cytoskeletal proteins to the extracellular matrix. The extracellular matrix is an important regulator of differentiation of granulosa cells (Paton and Collins, 1992).

Granulosa cells are circumscribed by two layers of insoluble extracellular material. The peripheral cells border the basal membrane of the follicle and corona radiata cells attach to the zona pellucida (Amsterdam and Rotmensch, 1987). Zona pellucida glycoproteins are synthesized by the growing oocyte (Driancourt et al., 1993b). It is unclear to what extent granulosa cells contribute to the formation of these layers; however, their interaction with the extracellular matrix components appears to affect differentiation events resulting in varying biochemical and structural characteristics.







28
A gradient of differentiation exists in granulosa cells in relation to their distance from the basal membrane of the follicle. Subpopulations of granulosa cells exist in ovarian follicles (Amsterdam et al., 1989). In the rat, there exists evidence of morphological and functional differences between antral cells in close proximity to the antral cavity and mural cells in close proximity to the basement membrane. They differ relative to number of gap junctions (Wiesen and Midgley, 1993), FSH receptors (Monniaux and De Reviers, 1989) and in the mitotic index (Hirshfield, 1992). Likewise, basal steroid production is lower in mural than in antral bovine granulosa cells cultured in the absence of FSH (Roberts and Echternkamp, 1994). Although morphological similarities were noticed between bovine antral and mural granulosa cells in response to FSH stimulation, the steroidogenic responses to FSH of antral granulosa cells in terms of E2, P4, and testosterone production are considerably higher than those of mural granulosa cells (Rouillier et al., 1996). There is a higher content of LH receptors, mitochondrial P-450 side chain cleavage enzyme and 38-hydroxysteroid dehydrogenase in the peripheral granulosa cell layers (in apposition to basal membrane) than in the inner layers. Also cells close to the basal membrane exhibit a decreased proliferative activity, which suggests an active role for the basement membrane in the induction and maintenance of granulosa cell differentiation (Amsterdam and Rotmensch, 1987).

Extracellular matrix is associated with changes in cell shape and contact, which are accompanied by simultaneous alterations in organization and expression of cytoskeletal proteins. Gonadotropins and extracellular matrix may induce granulosa cell differentiation in a coordinated fashion by their effect on cell shape,







29

cell contact, intercellular communication, and expression of cytoskeletal elements (Amsterdam and Rotmensch, 1987).

Fibronectin is present in the follicle as a component of the basal lamina and as a soluble fraction of follicular fluid which increases with increased follicle size (Luck, 1994). Fibronectin production and steroidogenesis represent two distinct states ofgranulosa cell differentiation, and it is proposed that fibronectin production is high in those tissues where extracellular matrix support is needed, like primordial follicles, atretic follicles, and early corpus luteum (Luck, 1994). In bovine granulosa cells, fibronectin secretion is greater in confluent cultures of cells from large follicles (15 mm), and is lower in sparse cultures of cells from small follicles (4-5 mm; Savion and Gospodarowicz, 1980).


The Theca Cell Compartment


Theca cells are the other cell type involved in follicular steroidogenesis. They are LH-responsive secretory cells which comprise the follicular envelope. The stem cell origin of theca cells is less clear than the origin of granulosa cells (Hirshfield, 1991). Theca cells appear to share a common ancestor with fibroblast and the stromal/connective tissue elements of the ovary (Gore-Langton and Armstrong, 1988). However, it has been suggested that granulosa and theca cells share a common ancestor (Hirshfield, 1991). Since the theca layer is not present in primary follicles but differentiates as follicles grow and mature, it is evident that theca cells arise continually throughout reproductive life. The aggregation and ultimate differentiation of the theca folliculi is assumed to occur under the influence of growth







30

factors emanating from the oocyte-granulosa complex, but there is no direct evidence to substantiate a casual relationship (Greenwald and Roy, 1994). The theca interna of sheep follicles, less than 3 mm in diameter, already consists of 8 to 12 layers of flattened cells and capillaries; 20% of the cells have the ultrastructural features of steroidogenically active cells; others are fibroblasts, and the majority are undifferentiated cells. In follicles 3-6 mm, approximately 40% of the theca interna possess tubular endoplasmic reticulum. In all species, the theca interna is separated from the granulosa cells by the basal lamina or basement membrane (Greenwald and Roy, 1994).

The time of appearance of a well-differentiated theca interna relative to development of granulosa cells varies from species to species. For example in sheep, differentiation of the theca appears in follicles of 2-3 mm in diameter with maximal differentiation occurring in late estrus, when numerous lipid droplets accumulate in theca interna cells. Two cell populations emerge in the bovine theca interna at 3 to 4 days before ovulation. One population consists of large epithelial cells with round nuclei increasing in area and another group of fibroblast-like cells in which the nuclear area does not increase. In large (10 - 12 mm) porcine follicles, theca interna cells form a discrete layer, contain many more lipid droplets than granulosa cells, and have a very high 31-hydroxysteroid dehydrogenase activity (Greenwald and Roy, 1994).

The theca interna layer develops a vascular supply concomitant to its differentiation as a steroidogenic tissue (Greenwald and Roy, 1994). In the rat, primordial follicles do not have an independent blood supply until multilaminar







31

granulosa and theca cell layers have formed. The vascular bed develops in the inner portion of the theca interna adjacent to the basal membrane and membrana granulosa. In turn, the theca interna vasculature is linked to an outer series of arterioles and venules in the theca externa layer (Greenwald and Roy, 1994). It is possible that follicular growth is angiogenisis-dependent. The increase in granulosal and thecal cells must be accompanied and/or preceded by an increase in new capillaries that grow toward and within the theca layer of growing follicles (Greenwald and Roy, 1994).


Two-Cell, Two-Gonadotropin Theory


In all mammalian species there are two types of cells involved in follicular steroidogenesis: LH-responsive secretory cells comprised of the theca interna cells of the follicular envelope and interstitial cells of the ovarian stroma, and FSHresponsive granulosa cells, that acquire the ability to respond to LH during the later stage of follicular maturation. These two cell types fulfill distinct roles in the steroidogenic process by virtue of their different regulatory hormones and their dissimilar expression of steroidogenic enzymes (Gore-Langton and Armstrong, 1988). The studies of Falck (1959; cited by Richards, 1980) provided the first evidence that ovarian estrogen biosynthesis required the interplay of at least two ovarian cell types. He isolated granulosa, theca, interstitial, and luteal cells and autotransplanted these cell types either alone or in combination to the eye chamber of ovariectomized female rats. Results from his experiments indicated that estrogen was produced only when theca or interstitial cells were cotransplanted with







32

granulosa cells or luteal cells or when intact follicles were autotransplanted. (Richards, 1980).

The steroidogenic pathways present in granulosa and theca cells of developing follicles are responsible for the local as well as the systemic sex-steroid milieu. Three major classes of sex-steroids are produced by mammalian ovarian follicles: progestins, androgens, and estrogens, which enterthe systemic circulation and exert important actions both within and outside the hypothalamo-pituitary gonadal axis (Urban and Veldhuis, 1992). Synergistic interactions between granulosa and theca cells have been proposed for some time as a mechanism by which ovarian follicles can produce the three classes of sex-steroids (Hsueh et al., 1984). This two-cell, two-gonadotropin theory of steroidogenesis can be summarized as follows.

Luteinizing hormone from the pituitary stimulates androgen production by theca cells (Fortune and Armstrong, 1977), which also appear to possess little or no ability to aromatize androgens to E2 (Fortune and Armstrong, 1978). The locally produced androgens diffuse through the basement membrane to the granulosa cells. Follicle stimulating hormone stimulates aromatase enzyme activity in granulosa cells to allow conversion of androgens to estrogens (Fortune and Armstrong, 1977). Granulosa cells seem to be incapable of synthesizing androgens (Fortune and Armstrong, 1977) but produce progesterone in vitro. Granulosa cells appear to lack the 17a-hydroxylase, 17-20 desmolase enzyme necessary to continue the steroidogenic pathway to androgens (Fortune and Quirk, 1988). In







33

granulosa cells from immature hypophysectomized rats, aromatase enzyme is not present but can be induced by treatment with FSH (Dorrington et al., 1975).

Experiments with bovine theca and granulosa cells isolated from proestrous follicles have provided evidence that both theca and granulosa cells are necessary for follicular E2 production in cattle (Fortune and Quirk, 1988). Androstenedione is the primary form of androgen secreted by bovine granulosa cells when cultured in vitro and stimulated by LH (Fortune, 1986). Luteinizing hormone and FSH increase production of P4 by granulosa cells. Nevertheless, granulosa cells do not convert progestins to E2 (Fortune, 1984). These observations led to the conclusion that the same two-cell, two-gonadotropin theory proposed for rats can be applied to the bovine follicle (Fortune and Quirk, 1988; Figure 2-1: model for regulation of E2 production by bovine preovulatory follicles). Theca cells are stimulated by LH to synthesize androgens, but they are unable to convert androgens to E2. Granulosa cells cannot synthesize androgens de novo, but they can convert exogenous androgens to E2. It has been shown that granulosa cells can secrete E2 in a longterm cell culture (Saumande, 1991; Rouillier et al., 1996; Diaz et al., 1997).

Fortune (1986) postulated that granulosa cells provide additional pregnenolone precursor (e.g., cholesterol) for thecal androgen production, which in turn enhances the ability of the granulosa cells to synthesize pregnenolone. This implies that cattle preferentially use the As pathway to synthesize androgens (Fortune and Quirk, 1988). Estrogens secreted by granulosa cells act as a mitogen and, in conjunction with FSH, induces appearance of LH receptors on granulosa cells. In bovine follicles, E2 exerts a positive feedback to stimulate its own







34

production and has a negative effect on P4 secretion due to inhibition of the 311hydroxysteroid dehydrogenase in granulosa cells. Based on these observations, Fortune and Quirk (1988) hypothesized that as the bovine follicle develops capability for E2production, E2 will inhibit conversion of pregnenolone to P4 in both theca and granulosa cells. This increases the amount of pregnenolone available for conversion to androgens via the A5 pathway in theca cells. Differential inhibitory effects of E2 on P4 versus androstenedione production could provide a mechanism by which E2 initially exerts a positive feedback on its own production to increase androgen synthesis, but eventually inhibits its own production by blocking thecal cell conversion of dehydroepiandrosterone to androstenedione through inhibition of 318hydrxoysteroid dehydrogenase (Fortune, 1986). Molecular Events Occurring in Theca and Granulosa Cells


Luteinizing hormone acts on theca cells via cAMP to regulate both C27 side change cleavage P-450 multienzyme complex (cholesterol 22-hydroxylase, cholesterol 22a-hydroxylase and C20-22-lyase), C21 side chain cleavage P-450 (17ahydroxylase, C17, 20-lyase) activation to increase androstenedione biosynthesis. In granulosa cells, androstenedione (from thecal origin) is converted to testosterone by 1713-hydroxysteorid dehydrogenase increasing the capacity of granulosa cells to synthesize pregnenolone (Fortune, 1986). Estradiol then is synthesized from testosterone by the aromatase P-450 enzyme system. Estradiol, in turn, binds to E2 receptors present in granulosa cells and may regulate gene expression (e.g., inhibin a- and IB-subunit mRNA; Turner et al., 1989). However, FSH must be







35

present to enhance intrafollicular E2 effects (Richards, 1980; Richards and Hedin, 1988: Figure 2-1).

There is substantial evidence that the cAMP/protein kinase A (PKA) pathway is one of the most important regulatory mechanisms in theca cells. Actions of LH on steroidogenesis of theca cells are mediated through the cAMP/PKA signal transduction pathway which stimulates the expression of steroidogenic enzyme genes required for androgen synthesis (P-450 side chain cleavage, 3i-HSD and P450 17a-hydroxylase 17, 20-lyase). Once theca cells have expressed the steroidogenic enzymes, activation of the cAMP/PKA pathway also stimulates the rate-limiting step in steroid hormone biosynthesis which is the transfer of cholesterol across mitochondrial membranes (Magoffin and Erickson, 1994). However, a number of observations has indicated that the activity and amount of of the P-450 side chain cleavage enzyme are unaffected by trophic hormone stimulation (Stocco, 1997). Recently it has been thought that the rate-limiting step in steroidogenesis is the de novo synthesis of a protein whose function is to facilitate the transfer of the substrate cholesterol from the outer mitochondrial membrane to the inner mitochondrial membrane and the P-450 side chain cleavage enzyme. The synthesis of this protein is dependent upon a functional cAMP-dependent protein kinase A signaling pathway that is linked to the steroidogenic potential of the cell.









R SH + FUH


LH go RLH


Figure 2-1. Schematic representation of the molecular events coordinating estradiol, FSH, and LH action in theca and granulosa cells of preovulatory follicles (Adapted from Richards and Hedin, 1988; Gore-Langton and Armstrong, 1988; Fortune and Quirk, 1988).


The protein was named the Steroidogenic Acute Regulatory (StAR) protein (Stocco, 1997). Interaction of StAR with the mitochondria may cause the formation of a protein complex consisting of P-450 side chain cleavage and 31l-HSD, enzymes required for the first two steps in steroidogenesis (Stocco, 1997). It appears that cAMP in theca cells increases the content of the regulatory (R) subunit of cAMPdependent protein kinase type II (RI151), P-450 side chain cleavage enzyme, and 17a-hydroxylase 17, 20-lyase.







37
receptors have little or no effect on proliferation of granulosa cells unless FSH acts via the FSH receptor to increase intracellular concentrations of cAMP (Richards, 1980). On the other hand, E2 is required for FSH to induce LH receptors in granulosa cells. The effects of E2 on granulosa cell function are mediated via the translocation of the cytosol E2-receptor complex to the nucleus. Binding of the E2receptor complex to nuclear acceptor sites results in altered genomic expression, synthesis of new messenger RNA, and a change in any one of a number of the components of the FSH receptor system. It is also possible that E2 enhances the ability of FSH to stimulate the adenylate cyclase system. Estradiol may increase the content of a specific cAMP-binding protein that is distinct from the regulatory subunit of the protein kinase or is the regulatory subunit of a specific protein kinase. Estradiol could also induce the synthesis of a specific protein that when phosphorylated in the presence of active kinase might enter the nucleus to affect gene transcription (Richards, 1980).

Gonadotropic hormone control of granulosa cell steroidogenesis occurs primarily by way of cAMP-dependent pathways, but evidence exists that additional intracellular regulatory signals can modify this process. Two such secondmessenger systems involve intracellular hydrolysis of phosphatidylinositol to produce inositol trisphosphate (IP3) and diacylglycerol. Intracellular Ca++ is increased in response to IP3 which in turn affects activity of protein kinase C and activation of the calcium-calmodulin system (Urban and Veldhuis, 1992).









Hormonal Control of Ovarian Follicular Dynamics


During each estrous cycle, bovine ovaries synthesize and secrete E2-1711 and P4, which coordinate function of the female reproductive system. Each estrous cycle is comprised of follicular and luteal phases. The follicular phase is characterized by development of the preovulatory follicle and its secretion of E2, whereas the luteal phase is characterized by secretion of P4 by the corpus luteum, which is formed after ovulation of the preovulatory follicle. At the end of the luteal phase, the corpus luteum regresses by the action of prostaglandin F2a(PGF2a), and final development of the next preovulatory follicle occurs. Gonadotropin secretion plays a central role in control of the estrous cycle. The developing preovulatory follicle produces a critical level of E2 that stimulates the hypothalamus to increase the frequency and amplitude of gonadotropin-releasing hormone pulses(GnRH). In cattle this is inferred based on an increased frequency and amplitude of LH pulses. Increased LH pulses amplify E2 secretion, complete follicular development, induce estrous behaviour and trigger the preovulatory surge of LH. Ovulation occurs a 30 h after the preovulatory LH surge (Chenault et al., 1976; Robinson and Shelton, 1991; Driancourt et al., 1993a).


Steroid Hormones


Steroid hormones such as progestins, androgens and estrogens are produced by the ovarian follicle during its development. Each wave of development of a dominant follicle goes through a selection, dominance and atresia or ovulation








39

phase. During the selection phase, E2 production by each ovary is similar (Ireland and Roche, 1987). At the end of this phase, one follicle becomes larger than all other follicles and is responsible for most of the E2 production, and contributes to a transient increase in concentration of E2 in the peripheral circulation. If luteal regression occurs coincident with the dominance phase in development of a dominant follicle, then the follicle ovulates. If luteal regression does not occur during a dominance phase, the dominant follicle undergoes atresia and a new selection phase begins (Ireland and Roche, 1987). In cows, the preovulatory surge of LH is induced by increasing concentrations of E2 after demise of the corpus luteum (Chenault et al., 1975, 1976; Kesner et al., 1981). Estradiol initiates the preovulatory surge of LH by acting on the hypothalamus to increase secretion of GnRH at the pituitary level, and the sensitivity of the gonadotropes to GnRH. A preovulatory surge of FSH occurs concomitantly with the surge of LH (Walters and Schallenberger, 1984). The preovulatory follicle is the source of this increase in E2. Plasma E2 concentrations are high before the LH surge and decline during the surge. Follicular fluid concentrations of androgens and progestins are much lower than E2 before the LH surge. Following the surge, androgens decline in follicular fluid whereas P4 increases. In contrast, atretic follicles differ from healthy follicles in that they have much lower concentrations of E2 than healthy follicles (Fortune and Hansel, 1985; Badinga et al., 1992; de la Sota, 1995).

Demise of the first dominant follicle at midcycle is due to the negative feedback effect of progesterone from the corpus luteum on LH secretion (Savio et al., 1993a). In the absence of a normal corpus luteum and in an environment with







40

low progestin, the first dominant follicle continues to grow and suppresses growth of other follicles for more than 20 days (Savio et al., 1993a). This sustained growth and functional dominance is due to increased LH pulse frequency. The presence of high levels of P4 during the luteal phase assures that the period of functional dominance of any follicle is limited due to induction of a low LH pulse frequency that is insufficient to maintain follicular function and leads to atresia. In contrast, maintenance of P4 concentrations between 1 and 2 ng/ml prolongs both morphological and functional follicular dominance (Sirois and Fortune, 1990; Savio et al., 1993b). In cases where the estrous cycle is lengthened by maintaining normal luteal concentrations of P4, either artificially or naturally as during pregnancy (Ginther et al., 1989c), continuous follicular waves occur. However, occurrence of follicle waves in pregnancy is more prevalent on the ovary contralateral to the uterine horn bearing the conceptus.


Gonadotropic Hormones


Regulation of steroid production by bovine theca and granulosa cells seems to depend on gonadotropins and on paracrine regulation by ovarian hormones. Theca cells have LH, but not FSH receptors (Ireland and Roche, 1983). Binding of LH to theca cells increases as the preovulatory follicle develops. In contrast, granulosa cells can bind both LH and FSH. As the preovulatory follicle develops, its granulosa cells increase their specific binding of LH, but exhibit decreased binding of FSH (Ireland and Roche, 1983). Both LH and FSH are essential for follicular development and steroidogenesis in vivo (Fortune et al., 1988).







41

Follicle stimulating hormone is the key hormone stimulating emergence of follicular waves and declines in FSH are associated with selection of a dominant follicle which becomes dependent on LH for its final fate (Roche, 1996). Adams et al. (1992b) concluded that there is a temporal sequence between surges of FSH and subsequent emergence of follicular waves. Time of follicle selection coincides with the first decrease in FSH secretion. Turzillo and Fortune (1990) reported that the secondary FSH surge (e.g., at 24 h after the FSH preovulatory surge) is important for the initiation of follicular development early in the bovine estrous cycle. Adams et al. (1992b) and Sunderland et al. (1994) reported that there is a cyclic pattern of FSH secretion during the estrous cycle of cattle with enhanced FSH secretion responsible for emergence of each follicle wave. Circulating concentrations of FSH decrease once selection is initiated as indicated by the presence of one estrogen-active follicle. This suggests that the dominant follicle secretes some inhibitory substance(s) to decrease FSH. The dominant follicle continues to grow demonstrating that large amounts of FSH are not necessary to sustain follicular dominance.

Terminal stages of antral follicular development and follicular hierarchy in monovular species may be regulated at the ovarian level by the presence of an active dominant follicle. Turnover of dominant follicles which permits development of new follicular waves appears to be determined by the inability of the dominant follicle to continue growing and suppressing growth of other follicles. There are physiological conditions which supportthe concept that LH exerts a regulatory effect on growth and dominance of dominant non-ovulatory and ovulatory follicles in cattle.







42
Such situations include the early postpartum period in dairy and beef cows, in which development of follicles < 8 mm occurs (Savio et al., 1990b). In beef cows, during the early postpartum period two to three non-ovulatory follicles develop before the first postpartum ovulatory follicle develops (Murphy et al., 1990). During these two situations LH concentrations are at basal levels. Also, growth and ovulation of follicles in prepubertal heifers can be achieved by repeated injections of LH (Tortonese et al., 1990).

In summary, turnover of ovarian follicles during the estrous cycle in cattle is regulated by the concentration of P4 in plasma acting via a negative feedback effect on LH secretion. Low frequency of LH pulses during the luteal phase is not sufficient to maintain continued growth and function of the dominant follicle. It is probable that a low androgen secretion by theca cells limits subsequent function of granulosa cells. Functional granulosa cells are required for terminal follicular development and sustained dominance of the follicle. When granulosa cell functional is compromised, the dominant follicle no longer suppresses growth of other follicles and this leads to recruitment of a new follicular wave (Savio et al., 1993a). Enhanced FSH secretion is also responsible for development of the next follicular wave. Adams et al. (1992b) showed that a surge in circulating FSH occurred during the plateau phase in maximal diameter of an anovulatory dominant follicle, and that the FSH surge was related to detectable emergence ( 4 or 5 mm follicles) of a follicular wave.









Other Factors Involved in Control of Ovarian Follicular Dynamics


Growth and development of a tissue requires the local production and integrated actions of specific growth factors. These growth factors mediate critical cell-cell interactions that control cell proliferation and organ development. Gonadal development also requires growth factor-mediated cell-cell interactions as a general mechanism to control cellular proliferation. Ovarian physiology requires rapid and continuous growth regulation associated with the process of folliculogenesis (Skinner, 1992). There is increasing evidence that growth factors modulate folliculogenesis but their precise role in the processes of follicular growth, differentiation and atresia is still unknown (Monget and Monniaux, 1995). Growth factor-mediated interactions between theca cells, granulosa cells, and the oocyte are required for the maintenance of ovarian function and the process of oogenesis. In most large animals, follicle size increases from millimeter to centimeter (e.g., cow, mare). Granulosa and theca cells are responsible for this follicular expansion. In addition to cell proliferation required during follicle development, follicles of various stages of development become atretic and cell growth is arrested. Regulation of cell proliferation in the follicle requires stimulatory and inhibitory growth factors (Skinner, 1992).

Ovarian follicular dynamics are controlled not only by the interaction of endocrine signals (steroid and gonadotropic hormones) but by locally produced ovarian peptide hormones and growth factors (Monget and Monniaux, 1995). The response of the two major follicular cell types, granulosa and theca cells, to







44

gonadotropins is regulated by the local production of growth factors and peptides. Ovarian follicles are known to produce a range of locally acting peptide/protein growth factors that can interact directly with the same cell type from which they are produced in an autocrine manner or with other cell types via a paracrine action to attenuate or stimulate the cellular response to gonadotropins. Also these factors could act in a juxtacrine fashion, activating receptors on adjacent cells, or stimulate the cell in which the factor is secreted without prior secretion of the factor from the cell (intracrine action; Armstrong and Webb, 1997).

Inhibins, activin, IGF-I, IGF-Il, IGF binding proteins, TGF-a, EGF, and TGF-G and other facors have direct and indirect effects on granulosa and theca cells that can modulate follicular development and steroidogenesis. Inhibins have both autocrine and paracrine effects. They increase LH-induced androgen synthesis in theca cells (Hillier et al., 1991b; Wrathall and Knight, 1995), and inhibin production is stimulated by steroids and FSH (Wrathall and Knight, 1993). This is indicative of a local feedback loop within individual follicles involving a sequential change of inhibins, activins and their binding proteins, which determine the different fates of the selected and unselected follicles that develop in the same systemic environment of gonadotropins and growth factors (Roche, 1996). There is evidence, mainly from studies in vitro that the inhibin-related peptides have actions on all the functional stages of folliculogenesis including ovulation (Findlay, 1993).









Inhibin


The term "inhibin" was proposed originally by McCullagh in 1932 to denote the activity of an aqueous extract of the testis that had the capacity to suppress formation of castration cells (cells that appear in the pituitary following damage of the seminiferous tubules) in the anterior pituitary gland. This was before the time when gonadotropins, FSH and LH, were defined as separate hormones and before testosterone was isolated as a pure substance. The term inhibin was used to designate a non-steroidal gonadal product with the capacity to specifically suppress FSH secretion. However, isolation of this molecule did not occur until 53 years after the original postulate (de Kretser and Robertson, 1989). Inhibin is a slow acting but powerful inhibitor of pituitary FSH biosynthesis and secretion by the anterior pituitary gland (Gaddy-Kurten et al., 1995).

Inhibin is a heterodimeric glycoprotein composed of an a subunit (relative molecular mass Mr = 18 kDa) linked by a disulfide bridge to one of two highly homologous B-subunits (approximate M, = 14 kDa) to form either inhibin A (a-BA) or inhibin B (a-IB; Findlay et al., 1993; Rose and Gaines-Das, 1996; Halvorson and DeCherney, 1996). By sequence analysis of cDNA clones, it has been shown that the three subunits (a, IA and kB) are coded by different genes, and the mature structures of the a, BA, and BIB subunits are present at the C-terminus of much larger precursor protein. The a subunit precursor is composed of four segments: a signal peptide, prosequence (Pro), N-terminal (aN) and C-terminal (aC) peptides in addition to the a component. The 31-kDa inhibin has been postulated as the







46

mature form of inhibin in the circulation (Findlay et al., 1993). Robertson et al. (1995, 1996), using a fractionation procedure, reported the identification of a range of bioactive and immunoactive forms of inhibin in human plasma. These forms have molecular masses of 28 to 128 kDa attributed to differences in glycosylation of the a subunit and differential processing of the a and I subunits. The inhibin I sequences also exhibit close homology with the TGF-lI superfamily of structurally related proteins with wide-spread biologic activities (Massague, 1987). Granulosa cells are the site of expression of mRNA encoding the different inhibin subunits in bovine antral follicles (Torney et al., 1989; Findlay et al., 1993).

Inhibin has a negative feedback effect on secretion of FSH (De Jong, 1988; Glencross et al., 1994). Inhibin is likely a chemical signal of the number of growing follicles in the ovary (Taya et al., 1996), and through its systemic action inhibin probably has an important role in determining species-specific ovulation rates (Taya et al., 1996; e.g., ovulation of only one follicle in cattle; Scanlon et al., 1993). The initial isolation of inhibin was achieved from bovine follicular fluid (bFF) as a 58 kDa glycoprotein consisting of two disulphide-linked subunits of apparent molecular masses of 43 kDa and 15 kDa. The larger subunit has been termed a, and the smaller as the B-subunit (Robertson et al., 1985). Ling et al. (1985) isolated two forms of inhibin, termed inhibin A and inhibin B, which differs by the NH2-terminal amino acid sequences of their B-subunits which are now termed SA and 1B. More recently, Good et al. (1995) isolated nine different biologically and immunologically active molecular variants of inhibin (29, 34, 49, 53, 58, 77, 88, 110, and > 160 kDa) from bovine follicular fluid. Quantitative immunoblot analysis has shown alterations







47
in the nine different molecular variants in bovine follicular fluid associated with functional status of dominant follicles. The predominant inhibin forms are free a subunits and large (> 34 kDa) inhibin forms rather than the fully processed inhibin form (Ireland et al., 1994). According to Good et al. (1995) each form of inhibin, except the 29-kDa form, inhibited basal FSH secretion and enhanced GnRHinduced LH secretion in pituitary bioassays. The dimeric forms of inhibin retained biological activity after isolation from bovine follicular fluid.

Granulosa cells of the follicle and luteal cells in primates produce inhibin, and its production can be regulated via endocrine (FSH, LH), paracrine (EGF, TGF-a, interferon-y, androstenedione) and autocrine (IGF-1, TGF-B, activin, follistatin) controls (Findlay, 1993). Follicle stimulating hormone, LH (at low doses), IGF-I, TGF-g, and activin can stimulate inhibin production by granulosa cells. Epidermal growth factor, TGF-a, follistatin, interferon-y, and high doses of LH negatively regulate inhibin production in the presence of FSH and other agents that increase cAMP concentrations (Findaly et al. 1993). Secretion of inhibin by rat, bovine (Henderson and Franchimont, 1983), human, and nonhuman primate granulosa cells is regulated by gonadotropins and sex steroids in vitro (Hillier and Turner, 1991). The expression of the inhibin subunit genes is regulated developmentally and inducible by FSH in the rat (Woodruff et al., 1987; 1987). In cultured rat granulosa cells, expression of a- and PB-subunit mRNA is enhanced by E2. This suggests that there is a mechanism whereby locally produced estrogen could influence relative rates of inhibin and activin synthesis by granulosa cells during follicular development in vivo (Hillier and Turner, 1991).







48
Ireland and Ireland (1994) found a high correlation of a and fA mRNA with intrafollicular changes in E2. P4, E2:P4 ratio, and total inhibin immunoactivity during growth of nonovulatory follicles. Ireland and Ireland (1994) reported that amounts of follicular inhibin a and A mRNAs increased coincident with increases in size of nonovulatory follicles, and expression was reduced markedly in atretic nonovulatory follicles compared with estrogen-active follicles. It was also shown that amounts of I mRNA remained unchanged during follicular development.

De la Sota (1995) reported that there are two very distinct patterns of change in the amounts of immunoblot bovine inhibin in follicular fluid of dominant and subordinate follicles on day 5, 8 and 12 of the estrous cycle. Absolute amounts of four forms (>160, 160-110, 77 and 49 kDa) were high in dominant follicles of day 5 and 8, but lower in atretic dominant follicle of day 12. Conversely, absolute amounts of the 31 kDa form of bovine inhibin in follicular fluid increased with atresia of the dominant follicle. In heifers, intrafollicular concentrations of inhibin decrease during growth of dominant ovulatory follicles, but increase during growth of dominant non-ovulatory follicle (Martin et al., 1991). Thus inhibins may have an important role in regulation of growth, differentiation and ovulatory quota of dominant follicles during the estrous cycle in heifers. It has been shown (Scanlon et al., 1993) that immunization of beef heifers with a (1-26 Gly-Tyr) subunit of 32 kDa bovine inhibin conjugated to human serum albumin increased ovulation rate in 32% of estrous cycles of heifers. Five injections of steroid-free bovine follicular fluid (which contained inhibin), given every 12 h beginning 12 h after the onset of estrus, suppressed completely the secondary surge of FSH without affecting LH







49
concentrations. This treatment delayed the appearance of the first wave of follicular development (Turzillo and Fortune, 1990).

Sunderland et al. (1996) demonstrated that the gonadotropic surge causes changes in levels or proportions of certain forms of inhibins. There was no change of inhibin forms during the period of most active E2 synthesis (preovulatory phase). After the gonadotropic surge, the 110-kDa form decreased while the 29-kDa inhibin form and total inhibin immunoactivity increased. In addition to gonadotropic regulation of inhibin production, paracrine and autocrine factors are also involved. First, steroids have the ability to change the secretion of inhibin in vitro. In rat granulosa cell cultures, FSH-induced immunoreactive inhibin production was increased by androstenedione and E2, whereas in the absence of FSH neither steroid was effective (Findlay et al., 1993). Various growth factors have also been demonstrated to influence inhibin production. Both IGF-I and TGF-I1 increase basal and FSH-induced inhibin production by rat granulosa cells. In contrast, TGF-l inhibited basal and FSH-stimulated inhibin production by pig granulosa cells (Michel et al., 1991). Epidermal growth factor appears to be a negative regulator of both inhibin protein (Michel et al., 1991) and a and RA subunit mRNA (LaPolt et al., 1990). Interferon-y is also a negative regulator of inhibin production by cultured immature rat granulosa cells, but only in the presence of FSH or other agents that increase cAMP levels (Xiao and Findlay, 1990).

In sheep, Campbell et al. (1996) reported that granulosa cells from large follicles secreted E2 in response to FSH and secreted inhibin and P4 regardless of the presence or absence of FSH. Insulin and LR3 IGF-I (human recombinant Long







50

R3 IGF-1) had a significant interaction in the stimulation of hormone production by granulosa cells from small and large follicles. On the other hand, insulin alone stimulated E2 and inhibin secretion.

A receptor with specific affinity for inhibin has not been demonstrated. However, inhibin has been shown to bind to activin type II receptors, although with lower affinity than activin. Therefore, rather than acting through a unique receptor, inhibin may exert its effect through competition with activin for activin receptor sites (Gaddy-Kurten et al., 1995; Halvorson and DeCherney, 1996).


Activin


Inhibin and activin are closely related peptides. Activin is composed of homodimers or heterodimers of the same Il-subunits linked by interchain disulfide bonds, resulting in activin-A (GA-IIA), activin-AB (IIA-IB), or activin-B (1IB-1IB; Halvorson and DeCherney, 1996). The molecular weight of activin is 24 to 28 kDa (Findlay et al., 1993). The l-subunit mRNAs are translated into pre-pro-I forms, which are proteolytically processed to the mature l-subunits. Activin A has been isolated from bovine follicular fluid, but activin B has not been isolated from mammalian sources so far (Findlay et al., 1993). Robertson et al. (1992) showed that the lA subunit monomer is present in bovine follicular fluid at a level 25% to 60% that of the &A subunit dimer (activin A), and its effect on in vitro responses are similar to those of the dimer. However, the monomer is less immunologically and biologically active (18% to 45%) than the dimer. It is unclear if dimerization of the monomer is a necessary prerequisite for biologic activity.







51

Activin is a product of the granulosa cells, as determined from expression of the B subunit mRNA in these cells and the isolation of activin from follicular fluid (Findlay, 1993; Findlay et al., 1993). Activin has the ability to stimulate FSH production and decrease inhibin biopotency (Robertson et al., 1988). The activin/inhibin IB subunit is expressed within rat gonadotropic cells, is locally secreted as 1BB B activin and may function as an autocrine modulator of basal FSH secretion and expression. Inhibin and activin are part of a larger superfamily that includes transforming growth factor (TGF)-B, mllerian inhibiting substance, the decapentaplegic gene complex of Drosophila, the bone morphogenic proteins, and the vegetal growth factor gene of Xenopus (Mather et al., 1992).

It is not known why different isoforms of inhibin and activin exist. It has been proposed that they may have different potencies or serve divergent functions. Also regulatory factors that determine the preferential formation of a-1l (inhibin) versus 1-11 (activin) dimers in cells that produce both subunits are poorly understood. Differential expression of the two subunits probably serves as one regulatory mechanism, in which an excess of a-subunit shifts production toward inhibin (Li et al., 1995; Halvorson and DeCherney, 1996). Activin is recognized as a regulator of a variety of actions, including neuronal differentiation, neuronal lifespan,and mesoderm induction in Xenopus (Mather et al., 1992).

Activin affects the differentiation of granulosa cells in an autocrine manner, and its action depends on the stage of differentiation of the cells. It promotes differentiation of cells during the preantral and early antral stages of folliculogenesis and prevents premature luteinization of cells from the later stages of antral







52

development. The coordinated effects lead to a promotion and maintenance of folliculogenesis (Findlay, 1993). Data from rat and bovine granulosa cell models support the hypothesis that activin regulates granulosa cell differentiation in an autocrine fashion, and this action of activin is related to the stage of granulosa cell differentiation (Findlay, 1993).

Activin is not only a mitogen for granulosa cells, but seems to play an important role in granulosa cell differentiation and ovarian follicular morphogenesis. Hulshof et al. (1997) reported that activin-A and activin receptor were co-localized in oocytes and granulosa cells of preantral follicles, suggesting an autocrine action of activin on preantral follicles. Activin stimulated both an increase in diameter of preantral follicles and proliferation of the granulosa cells. Li et al. (1995) showed that activin induced reaggregation of granulosa cells into a follicle-like structure. Activin (> 100 ng/ml) enhanced FSH-induced aromatase activity (Hutchinson et al., 1987), LH binding sites, P4 production, FSH-induced inhibin levels and, both a and I subunit mRNA (Xiao et al., 1990). The mechanisms of activin action are likely to involve stimulation of FSH receptor formation and cAMP production (Findlay et al., 1993). In addition, activin increased the responsiveness of granulosa cells to both gonadotropins (Xiao and Findlay, 1991). These in vitro observations are consistent with activin acting as an autocrine regulator to promote folliculogenesis during the preantral or early antral stages. Mainly, activin induces FSH receptors as a mechanism whereby preantral follicles may become responsive to FSH (Findlay et al., 1993). This proposition assumes that the granulosa cells either constitutively express or acquire receptors for activin prior to this important step in differentiation







53
(Findlay, 1993). It is proposed that activin, produced by granulosa cells of preantral follicles lacking FSH receptors, acts on these same cells to induce FSH receptors. The FSH then acts on the cells to stimulate production of more activin and the activin receptor, and FSH and activin together further enhance FSH receptor number in partially differentiated cells. FSH and activin also induce follistatin production, which modulates the action of activin and may have an effect on the granulosa cell, independent of its activin binding properties. This model also implies that some differentiating actions of FSH on granulosa cells may be mediated by activin, which would explain why follistatin inhibits FSH-induced differentiation of the cells in vitro (Findlay, 1993).

Activin action on differentiated granulosa cells is consistent with a role in preventing premature luteinization (Findlay, 1993). In bovine granulosa cell cultures, recombinant human activin-A caused a time- and dose-dependent inhibitory effect on LH-induced production of P4 and oxytocin (Shukovski et al., 1991). Bovine activin A prevented the spontaneous luteinization of fully differentiated bovine granulosa cells in vitro (Shukovski and Findlay, 1990). Activin is also a potential autocrine modulator of aromatization in granulosa cells (Hillier, 1991). It has been shown that activin is an atretogenic agent (Woodruff et al., 1990). Exposure of rat and human theca cells to activin results in a dosedependent reduction in LH-induced androstenedione production (Hsueh et al., 1987). This would constitute a paracrine action of activin in the ovarian follicle (granulosa-thecal cells regulation) and could have important implications in the establishment of dominance in follicles (Findlay, 1993). These results suggest that







54

activin plays an important role in regulating the functional integrity and development of granulosa cells in the ovarian follicle and these effects may be mediated directly through activin receptors on granulosa cells.

Activin receptors can be divided into two classes: type I (actRI) and type II (actRII). ActRI binds activin with high affinity; however, binding only occurs in the presence of the signaling peptide, actRIl. Conversely, actRll activity depends on the formation of a noncovalent, heteromeric complex with actRI (Mathews, 1994). This complex functions as a serine-threonine kinase (Halvorson and DeCherney, 1996). Multiple isoforms have been detected within the activin type I and type II receptor classes. Each isoform demonstrates distinct ligand-binding affinities and cytoplasmic domain structure. This observation raises the possibility that dose- and cell-specific responses could be attained through the expression of various combinations of activin receptor isoforms and number (Halvorson and DeCherney, 1996). Type I receptors are expressed and reach the cell membrane even in the absence of the type II receptors. Type I receptors do not bind ligand without the coexpression of a type II receptor. Both classes are required for any response (Gaddy-Kurten et al., 1995).

Activin and inhibin share a common I subunit, and both bind to a2macroglobulins and to follistatin, which is a specific activin B subunit-bindini protein (Gaddy-Kurten et al., 1995). Although the physiological relevance of the inhibin-a2macroglobulin and activin-a2-macroglobulin interaction is unknown, both complexes are biologically active to modulate FSH secretion. Although most activin appears to be complexed with a2-macroglobulin, complex formation probably does not affect







55
activin function. The a2-macroglobulin presumably binds activin with lower affinity than follistatin or cell surface receptors, which could account for the lack of effect on activin bioactivity; alternatively, activin may be able to interact with its receptor while complexed to a2-macroglobulins. The physiological function of this complex formation is not known, but a2-macroglobulins may serve roles in storage, delivery and clearance of activin (Mathews, 1994). In contrast, activin-follistatin complex is biologically inactive and it is likely that this binding protein plays an important role in limiting exposure of cells to activin (Gaddy-Kurten et al., 1995).


Follistatin


Follistatin is a single peptide chain (molecular weight 32,000 - 35,000 Daltons) that was first isolated from porcine follicular fluid, and is distinct from inhibin and activin. This cysteine-rich protein can inhibit the release but not the synthesis of FSH and has no effect on LH in cultures of pituitary cells (Tonetta and diZerega, 1989). Follistatin is structurally related to the activin-inhibin subunits. The primary structure of follistatin across species is highly conserved (> 97%). Follistatin is coded by a single gene organized into three homologous domains, each with > 50% homology. Alternative mRNA splicing results in two mature mRNA forms: the longer encoding FS-315 and the shorter encoding FS-288. The existence of two isoforms of follistatin provides cells with a mechanism to control follistatin activity through alterations in post-transcriptional processing (Halvorson and DeChemey, 1996). Follistatin is expressed in a number of different tissues, but







56

a major site of production is the ovarian granulosa cells under FSH regulation (Findlay et al., 1993).

It has been demonstrated that follistatin mRNA and protein expression change in association with folliculogenesis and atresia (Findlay et al., 1993). Follistatin mRNA is first detected in granulosa cells of secondary follicles and then becomes more abundant as the follicle forms an antrum, with uniform expression in the whole granulosa cell layer. The strongest signal is found in preovulatory follicles and newly formed corpora lutea. Follistatin protein is confined to healthy dominant preovulatory follicles and a subpopulation of tertiary follicles in animals entering estrus (Shukovski et al., 1992). Atretic follicles show no immunoreactive follistatin in granulosa cells. The bovine follistatin mRNA signal increases as folliculogenesis progresses, with the strongest signal being observed in preovualtory bovine follicles (Shukovski et al., 1992).

Findlay (1993) and Findlay et al. (1993) hypothesized that follistatin modulates granulosa cell function in an autocrine fashion and that follistatin action is through binding and neutralization of activin (Mathews, 1994). Thus, follistatin is likely to favor the process of follicular luteinization or atresia. However, follistatin may also have direct actions on granulosa cells independent of its activin-binding activity. Xiao et al. (1990) reported that follistatin enhanced the stimulatory action of forskolin on P4 production by rat granulosa cells but did not influence the effect of forskolin on either aromatase activity or inhibin production. In contrast, activin enhanced all three responses stimulated by forskolin. The binding of activin to follistatin is through two binding sites present in activin-A (1A-B1A), whereas inhibin-A







57

(a-BA) has only one binding site for follistatin. Shimonaka et al. (1991) hypothesized that follistatin binds activin and inhibin through the common B subunit.

There is strong evidence for a short loop feedback system involving activin and follistatin in the ovarian follicle. Follicle-stimulating hormone can stimulate production of both activin and follistatin. Follistatin can neutralize activin action and possibly some actions of FSH that might be mediated by activin (Findlay et al., 1993). The production of the activin-binding protein, follistatin, in the same tissue may further limit the bioavailaibility of activin and inhibit activin-induced cell growth and formation of follicles, as reported by Li et al. (1995). Alternatively, because the affinity of activin for follistatin is comparable to that of activin for its receptors, follistatin could directly compete with activin receptors of lower affinity for activin. However, higher affinity activin receptors could still bind ligand and elicit signal transduction (Mathews, 1994).


Transforming Growth Factor 1B


Transforming growth factor-B is a family of multifunctional growth factors originally named for their ability to induce normal rat kidney fibroblasts to grow in soft agar in the presence of epidermal growth factor (EGF; Mulheron and Schomberg, 1993). The TGF-1 superfamily is made up of a number of proteins with the potential to act as intraovarian regulators of ovarian function (Armstrong and Webb, 1997). Also, mullerian inhibiting substance is included among the members of the TGF-B family (Mulheron and Schomberg, 1993). Transforming growth factors-B are ubiquitous peptides and have been shown to affect nearly every cell







58

in the body. There are five TGF-fl subtypes, three of which (TGF-31, TGF-12, and TGF-113)are expressed by cells of the mammalian ovary (Hernandez et al., 1990; Mulheron and Schomberg, 1990). They are products of separate genes, but few cellular responses to TGF-1 can be ascribed solely to one isoform. Transforming growth factor-f is synthesized as an inactive precursor protein that must be cleaved for the 25-kDa homodimer to elicit its effects.

The precise mechanism of TGF-I activation in vivo is unknown, but is thought to occur via induction of proteases such as plasmin or perhaps by processing in an acidic environment. In vitro TGF-I can be activated by exposure to heat, detergent, or acid. Once activated, TGF-I is thought to affect target cells through TGF-9-specific membrane-bound receptors, but the precise cellular mechanisms of TGF-I actions have not been elucidated (Mulheron and Schomberg, 1993). The most likely mediators of TGF-I signal transduction are three cell surface proteins, which were identified based on their ability to bind and be chemically cross-linked to radiolabeled TGF-I (Massague and Like, 1985). These three molecules have been designated type I, II, and III receptors and have apparent molecular weights of 55, 80, and 280 kDa, respectively. Data from a variety of studies indicate that receptors types I and II are responsible primarily for signal transduction. The type III receptor, a membrane-anchored proteoglycan, also called betaglycan, that may help modulate cellular access to TGF-B by sequestering it near the cell membrane (Mulheron and Schomberg, 1993). The type 11 receptor has a functional cytoplasmic serine/threonine kinase domain, which suggests that serine/threonine phosphorylation may be an important component in the intracellular







59

signal transduction of TGF-R. The type II receptor is structurally related to the activin receptor. Regulation of TGF-I activity is controlled by the activity of proteases that release mature TGF-11 from the extracellular matrix-associated latent complex. The conversion of plasminogen to plasmin within the follicle and the presence of plasminogen activator inhibitors would be expected to be critical in the control of TGF-13 activity during folliculogenesis. This mechanism would be expected to influence development of dominance since TGF-Is act synergistically with gonadotropins to control the differentiation of follicular cells. Specific extracellular proteases play a central role in the regulation of TGF-B3 bioactivity during follicle growth and development of dominance (Armstrong and Webb, 1997).

Given that TGF-B is a multifunctional growth factor, it can act as both a growth stimulator and an inhibitor depending on cell type and culture conditions. Transforming growth factor-B is primarily an inhibitor of follicle growth and therefore facilitates follicle cell differentiation by modulating granulosa cell expression of functional gonadotropin receptors (Mulheron and Schomberg, 1993). Ovarian theca cells have been shown to express and produce TGF-I (Skinner et al., 1987a). Transforming growth factor-11 has been shown to stimulate a number of granulosa cell functions, including FSH induction of LH receptors, EGF actions, FSH induction of aromatase activity, production of IGF-I and inhibin. Transforming growth factor-1B also can influence theca cell function and steroidogenesis, and oocyte maturation (Skinner, 1992). Transforming growth factor B inhibits growth of bovine and porcine granulosa cells (Skinner et al., 1987a) and theca cells (Roberts and Skinner, 1991) induced by TGF-a/EGF. The influence of TGF-B on cell function







60
may be mediated indirectly through inhibition of cellular proliferation. Growth inhibition may be important to prevent premature cell growth of the preantral follicle, arrest cell growth during atresia, and control cell growth during follicle cell expansion (Skinner, 1992). Overall, these data indicate that TGF-Ifs are local regulators of ovarian function (Armstrong and Webb, 1997). In cows, TGF-fls inhibit granulosa and theca cell proliferation while enhancing gonadotropin-stimulated steroidogenesis (Roberts and Skinner, 1991).


Epidermal Growth Factor/Transforming Growth Factor-a


Other growth factors known to affect ovarian function include TGF-a, a structural analogue of EGF. Transforming growth factor-a is expressed in bovine theca cells and has been localized in the thecal cell (Skinner and Coffey, 1988) layer during follicular growth. However, there is a decrease in immunoreactive TGFa in preovulatory follicles that is correlated with a decrease in mitotic activity of granulosa cells (Lobb and Dorrington, 1992). Epidermal growth factor and TGF-a are products of separate genes, however, they are structurally similar and both elicit their effects through the EGF receptor (Mulheron and Schomberg, 1993). Epidermal growth factor and TGF-a are believed to be produced by ovarian theca cells (Skinner et al. 1987b).

Epidermal growth factor is a small (53 amino acid, 6043 molecular weight), single-chain polypeptide containing six half-cysteine residues that form three intrachain disulfide bonds. It is synthesized as a large molecular weight precursor that can be proteoliticaly cleaved to generate the biologically active form. Within







61
species, two to six EGF isoforms have been elucidated that have variable degrees of receptor binding and bioactivity (Mulheron and Schomberg, 1993).

Transforming growth factor-a is synthesized as a glycosylated precursor protein of 159 or 160 amino acids with the mature 50-amino acid form being generated by proteolytic cleavage. The precursor form contains a signal peptide sequence and a hydrophobic transmembrane domain, suggesting that the precursor form of TGF-a is an integral membrane glycoprotein (Mulheron and Schomberg, 1993). Membrane-anchored TGF-a binds to EGF receptors on adjacent cells and elicits a response. Such a mechanism would allow for activation of adjacent cells by membrane-anchored TGF-a in the absence of proteases. In the follicle, this could provide a direct means of intracellular communication and consequently an additional mechanism for controlling follicle cell growth. Epidermal growth factor/TGF-a acts on target cells via interaction with specific 170-kDa transmembrane receptors which are tyrosine kinases. The phosphorylated Cterminus is thought to unmask the catalytic domain of the receptor. A number of additional intracellular signaling pathways are activated within minutes of EGF binding to its receptor: activation of phospholipases, increase in intracellular calcium, activation of protein kinase C, and activation of serine/threonine kinases. Ultimately these individual changes result in altered cellular growth and differentiation (Mulheron and Schomberg, 1993).

Epidermal growth factor/TGF-a may be a primary mitogen of granulosa cells in most species. (Lobb and Dorrington, 1992). Transforming growth factor-a/EGF has been also shown to stimulate maturation of granulosa cells (Tonetta and







62

DiZerega, 1989) but inhibits follicular E2 secretion in sheep (Webb et al., 1994). Epidermal growth factor can inhibit production of both inhibin and progesterone by bovine granulosa cells, but stimulates DNA and protein synthesis (Tonetta and DiZerega, 1989). Epidermal growth factor binding is stimulated by FSH, and EGF in turn enhances FSH binding (Urban and Veldhuis, 1992). EGF inhibits FSHinduced aromatase and the formation of LH receptors in the rat and pig (Hsueh et al., 1984). Prior exposure of granulosa cells to FSH has been reported to diminish the inhibitory effects of EGF on cell differentiation (Paton and Collins, 1992). In humans, EGF and TGF-a are negative modulators of FSH-induced synthesis of E2 and positive modulators of granulosa cell proliferation. Their production by human follicles seems to decrease as follicular diameter enlarges (Gougeon, 1996). In bovine granulosa cells in vitro, EGF, TGF-a and bFGF stimulate proliferation but inhibit steroidogenesis (FSH-induced E2 secretion; Rouillier et al., 1997) and inhibin production (Skinner and Coffey, 1988; Vernon and Spicer, 1994), suggesting that they might act in vivo by enhancing growth and delaying terminal maturation of follicles (Monniaux et al., 1997). Apoptotic cell death in granulosa cells of follicles selected for ovulation is prevented by the paracrine actions of EGFITGF-a produced by thecal-interstitial cells or the autocrine and paracrine actions of bFGF synthesized by granulosa cells (Hsueh et al., 1994).

Lobb and Dorrington (1992) proposed a model for the role of transforming growth factors in ovarian physiology and follicular development. In primordial follicles, TGF-a of theca origin diffuses into the granulosa cell compartment and positively stimulates granulosa cell mitosis. Granulosa cells in these follicles are not







63

fully differentiated and do not synthesize estrogens. While promoting granulosa cell growth, TGF-a may also maintain the granulosa cell in an undifferentiated state. At a latter stage of development (large antral follicles), granulosa cell mitosis declines and the cells begin to synthesize estrogens. TGF-P negatively affects granulosa cell growth while promoting gonadotrophin-induced steroidogenesis in these follicles. The interplay of the two growth factors leads at first to increase cell number and latter the augmentation of differentiate function, that culminates into a mature estrogen-active preovulatory follicle. Insulin-like Growth Factors (IGFs) and IGF-binding Proteins (IGFBPs)


Insulin-like growth factors have direct effects on cultured ovarian cells. These effects include stimulation of granulosa cell mitogenesis, granulosa and luteal progesterone production, and thecal cell androgen production that appear similar among species (Spicer and Echternkamp, 1995). Administration of growth hormone stimulates growth of small antral follicles probably by an indirect mechanism involving IGF-I of endocrine origin. Terminal follicular growth is a strictly gonadotropin dependent process corresponding to initiation of follicular waves, selection of dominant follicles and terminal maturation of preovulatory follicles. The IGF family of growth factors may control the growth of small antral follicles in conjunction with FSH in the initiation of follicular waves. They are likely to be part of the players in the process of selection of the dominant follicle (Monniaux et al., 1997). The IGF system is composed of different elements: two ligands (IGF-I and IGF-Il), two receptors, and six or more IGF-binding proteins. IGFs function as







64

modulators of gonadotrophin action at the cellular level and stimulate granulosa and thecal cell proliferation and differentiation (Armstrong and Webb, 1997). The type I receptor mediates most of the somatomedin-like actions of both IGF-I and -I1. It is an a2132 tetramer structurally and functionally related to the insulin receptor. The affinity of this receptor for IGF-I is slightly higher than for IGF-Il and much higher than for insulin. The type II receptor binds IGF-II and molecules that possess a mannose-6-phosphate residue such as lysosomal enzymes. The IGF-II receptor does not bind insulin, and binds IGF-I with very low affinity (Monget and Monniaux, 1995).

The six IGFBPs bind IGF-I and II with high affinity. They are classified into two groups: IGFBP-1, -2, -4, -5 and -6 called 'small complex', which are present in serum and other fluids. Their molecular masses range between approximately 24 kDa and 35 kDa. Their concentrations in serum are either negatively (IGFBP-1 and

-2) or not regulated by growth hormone. IGFBP-3 is the most predominant IGFBP in serum. It is present in a 150 kDa form or 'large complex'. The concentration of IGFBP-3 is positively regulated by growth hormone and IGF-I. IGFBPs increase the half life of IGFs, and can both inhibit or potentiate IGF action at target cells. Their affinity for IGF-I and -II is of the same order of magnitude as the affinity of the type I receptor. Actions of IGF-I on follicular cells are mediated on granulosa and thecal cells by type I receptors (Giudice, 1992). In cows, expression of the IGF-I receptor increases in small antral follicles (Wandji et al., 1992) whereas dominant follicles express abundant IGF-I and IGF-Il mRNA (Yuan et al., 1998) but failed to express IGFBP-2 which is a binding protein associated with atresia (Monget and Monniaux,







65

1995). In sheep as in rats, pigs and humans, IGF-I stimulates both proliferation and differentiation of granulosa cells from follicles of 1-3 mm in diameter (Monget and Monniaux, 1995) and thecal cell steroidogenesis (Giudice, 1992; Paton and Collins, 1992). On follicles > 5 mm, IGF-I stimulates secretion of P4 by granulosa cells (Monniaux and Pisselet, 1992). The mRNA encoding IGF-Il has been detected in thecal tissue of bovine ovarian follicles (Armstrong and Webb, 1997), and localized in both granulosa and thecal cells of sheep (Leeuwenberg et al., 1995). The functional significance of these differences in the expression of mRNA encoding IGFs between species in unknown. It may reflect changes in the relative roles of IGF-I and -II that have evolved to fit the particular pattern of follicular development (Armstrong and Webb, 1997). As with the IGFs, the spatial expression of the binding proteins within ovarian follicles is species-specific. In cows (Armstrong et al., 1996; Yuan et al., 1998) and sheep (Besnard et al., 1996a), expression of mRNA encoding IGFBP-4 and -2 is restricted to theca and granulosa cells, respectively.

The release of IGFs from the IGFBPs is controlled by the action of specific IGFBP proteases. Specific IGFBP proteases have been detected in follicular fluid from ewes (Besnard et al., 1996b) and the amounts and activity of these enzymes change during folliculogenesis. Interaction between IGFs and other peptides (e.g., EGF, TGF-a) may be significant in regulating not only the proliferation but also the survival of granulosa cells (Paton and Collins, 1992). In cattle, doses of 100 ng/ml of IGF-I either had no effect or inhibited basal and FSH-induced E2 production by granulosa cells collected from small (< 5 mm) follicles, and was a weak stimulator







66

of FSH-induced E2 production by granulosa cells from large follicles (> 8 mm). Insulin, IGF-I and IGF-II stimulate progesterone production by bovine granulosa cells and, concomitant with FSH treatment, enhances the stimulatory effect of IGF-I. Insulin and IGF-I synergize with LH to promote thecal cell androgen production, and bovine thecal cells appear much more sensitive to insulin (Spicer et al., 1993).

No changes in follicular fluid concentrations of IGF-I were detected in first wave dominant (Badinga et al., 1992, de la Sota et al., 1996) and subordinate (de la Sota et al., 1996) follicles collected on days 5, 8 and 12 of the estrous cycle. In contrast, Echternkamp et al. (1994) reported higher concentrations of IGF-I in follicular fluid of large E2 active follicles compared to E2 inactive and small follicles. De la Sota et al. (1996) and Echternkamp et al. (1994) reported changes in the absolute amounts and proportions of IGFBP-2, -4 and -5. They were at low levels in E2 active dominant follicles but are at high levels in atretic dominant follicles, which may reduce the concentrations of free IGF-I and -II available in follicular fluid to exert their trophic activity on granulosa cells.


Manipulation of The Follicular Dynamics in Cattle


The failure of hormonal treatments to control follicular development and superovulation in women and domestic animals in a very precise manner gives evidence of significant gaps in the understanding of the mechanisms that regulate development and function of ovulatory follicles (Fortune et al., 1988). Furthermore, the mechanisms that control the ovulatory quota in different domestic animals are not completely understood. It is clear that the development of medium- sized and







67

large follicles in cattle occurs in waves. The regularity of the patterns allows experimental manipulations designed to determine how these patterns are controlled (Fortune et al., 1991).

Control of follicular development in cattle involves the removal of the suppressive effect of the dominant follicle to allow emergence of a new follicular wave at a specific time after treatment. Different approaches have been followed to control follicular dynamics in cattle. Hormonal approaches have been directed to cause luteinization or atresia of the follicles present at the time of treatment (Bo et al., 1994). This has been accomplished by using hCG (Rajamahendran and Sianangama, 1992; Diaz et al., 1993) or GnRH (Macmillan and Thatcher, 1991; Schmitt et al., 1996b) to induce follicle luteinization or ovulation, or by progestogens and E2 to cause atresia of the dominant follicle (Bo et al., 1994).

A multitude of experimental approaches have been used to control follicle dynamics in cattle, including treatments with hCG (Diaz et al., 1993; Schmitt et al., 1996b; Sianangama and Rajamahendran, 1996), GnRH analogs (Macmillan and Thatcher, 1991; Rusbridge et al., 1992; Schmitt et al., 1996a), electrocautery of the dominant follicle during its growing phase (Ko et al., 1991; Adams et al., 1993b), ultrasound-guided follicle aspiration (Bergfelt et al., 1994), pretreatment with FSH given early in the estrous cycle to increase the number of embryos recovered after a superovulatory treatment (Goulding et al., 1991; Gray et al., 1992), and treatment with high doses of P4 during the growing phase of the dominant follicle to suppress the dominant follicle and hasten the emergence of the next wave (Adams et al., 1992a; Burke et al., 1994).











Hormonal Manipulation: Human Chorionic Gonadotropin and Gonadotropin Releasing Hormone


It has been demonstrated that the first wave dominant follicle can be ovulated after administration of Gonadotropin Releasing Hormone (GnRH; Rusbridge et al., 1992: Schmitt et al.. 1996a) or human Chorionic Gonadotropin (hCG). Human Chorionic Gonadotropin is a glycoprotein with LH-like activity that promotes ovulation of the first wave dominant follicle and formation of an accessory corpus luteum (CL) when administered on d 4 (Breuel et al., 1989; Price and Webb, 1989), day 5 (Walton et al., 1990; Schmitt et al., 1996a, 1996b), day 6 (Fricke et al., 1993). day 7 (Breuel et al., 1989; Rajamahendran and Sianangama, 1992; Sianangama and Rajamahendran, 1992; 1996). day 10 (Breuel et al., 1989), or day 14 to 16 (Price and Webb. 1989) of the estrous cycle.

Both GnRH and hCG have the ability to induce an accessory CL during the early luteal phase of the estrous cycle in cattle, which could have some practical applications. Establishment and maintenance of pregnancy as well as embryo survival in cattle are related to the ability of the CL to secrete P4 (Thatcher et al., 1994b). Subluteal levels of P4 could be a factor that contribute to embryo losses (Thatcher et al., 1994b). Also, the rate of increase in plasma P4 during the early luteal phase could play a role in establishment of pregnancy (Shelton et al., 1990). Progesterone supplementation after insemination in heat stressed cows may correct deleterious effects of impaired luteal function on pregnancy rates. Inducing formation of an additional CL during the luteal phase of the estrous cycle with an







69
injection of either GnRH or hCG could be a strategy to increase concentrations of P4 in plasma during the period when maternal recognition of pregnancy occurs (Schmitt, 1995).


Ablation of the Dominant Follicle


There is a potential for artificial control of follicular dynamics using ablation procedures. Two types of ablation procedures have been tried: electrocauterization (Ko et al., 1991; Adams et al., 1993a) and ultrasound-guided transvaginal follicle aspiration (Bergfelt et al., 1994; Stubbings and Walton, 1995). The first one is an invasive method and not repeatable on a within cow basis, whereas follicle aspiration is a noninvasive and repeatable procedure that can be used for the retrieval of oocytes for in vitro fertilization (Pieterse et al., 1988, 1991a, 1991b).

It has been shown that electrocauterization (Ko et al., 1991; Adams et al., 1993a) of the dominant follicle during its growing phase hastened emergence of the next follicular wave. Ko et al. (1991) showed that cauterization of first wave dominant follicle on days caused an early emergence of wave 2, and increased the incidence of 3-wave interovulatory intervals.

Follicle ablation via ultrasound-guided follicle aspiration at random (all follicles < 5 mm) stages of the estrous cycle synchronized subsequent follicular wave emergence and resulted in a high degree of synchronized ovulations in heifers following PGF2 induced luteolysis (Bergfelt et al., 1994).









Superstimulation


The objective of ovarian superstimulatory treatment in cattle is to obtain the maximum number of viable embryos by stimulating growth and subsequent ovulation of competent antral follicles with exogenous gonadotropins (Adams, 1994). However, the variable and unpredictable superovulatory response of the donor animal has remained one of the most limiting factors to successful embryo transfer (Armstrong, 1993). To date, all methods of inducing superovulation in cows have involved administration of exogenous gonadotropic hormones: crude or purified hypophyseal extracts, such as porcine follicle-stimulating hormone and luteinizing hormone, equine chorionic gonadotropin and human menopausal gonadotropin (Mapletoft et al., 1994).

Many reports have been published on dosage regimes and types of gonadotropin preparations for ovarian superstimulation (Elsden et al., 1978; Murphy et al., 1984; McGowan et al., 1985; Lerner et al., 1986; Pawlyshyn et al., 1986; Armstrong, 1993) but timing of treatment with respect to monitored follicular wave development has received little attention (Adams et al., 1993b). Animal variability is important as a determinant of superovulatory response. The ovarian status of the donor at the time of hormone treatment appears to be a major determinant of the superovulatory response (Monniaux et al., 1983). In recent years, research has been focused on making responses more predictable by administering superstimulation treatments with specific regard to the status of follicular development at initiation of treatment (reviewed by Armstrong, 1993). Results from







71

different studies (Pierson and Ginther, 1988; Grasso et al., 1989; Guilbault et al., 1991) suggested that the superovulatory response is affected by the presence of a large dominant follicle (Pierson and Ginther, 1988; Guilbault et al., 1991; Adams et al., 1993b; Wolfsdorf et al., 1997). Superovulation may be induced with equal efficacy when treatment is initiated during the time of the first or second follicular waves when a dominant follicle is present in the ovaries, and the superovulatory response is enhanced if treatment is initiated at the time of wave emergence, before the time of follicle selection (Adams, 1994). It has been shown (Wehrman et al., 1996) that FSH treatment in presence of a persistent dominant follicle resulted in follicles that were unable to respond to the preovulatory surge of LH.

Priming doses of FSH given at the beginning of the cycle have been shown to increase the superovulatory response to FSH treatments (Rajamahendran et al., 1987; Ware et al., 1987; Touati et al., 1991). Others have failed to show the beneficial effect of early priming (Grasso et al., 1989; Gray et al., 1992). The idea of the priming is to increase the number of follicles in the FSH-responsive pool. It is likely that priming with other hormones, such as PMSG, which possesses both FSH and LH activity, could increase the number of FSH-responsive follicles (Monniaux et al., 1984).

Passive immunization against inhibitors of gonadotropin secretion may also be a means of increasing ovulatory rates by short-term elevation of levels of endogenous follicle-stimulating hormone (Mapletoft et al., 1994). It has been shown that passive immunization of ewes against steroid free-follicular fluid or against synthetic inhibin peptides increased ovulation rates two- to four-fold (O'Shea et al.,







72

1991; Schanbacher et al., 1991). Ovulation rate and prolificacy in the cow are more difficult to manipulate by inhibin vaccination (O'Shea et al., 1994). However, there are reports that immunization of heifers or cows against steroid-free follicular fluid increases ovulation rate (Price et al., 1987; Glencross et al., 1992; Morris et al., 1993; Scanlon et al., 1993) and number of transferable embryos (Alvarez et al., 1997).

In summary, there are several factors involved in the control of ovarian follicular dynamics in cattle, and some of the mechanisms that control ovarian follicular growth have been elucidated, but still there are more that are not completely understood; as an example, factor(s) and mechanisms that control dominance during a follicular wave. A set of experiments was designed and performed with the aim of this study to identify the factor (s) and mechanisms that control follicular dominance in cattle,. Understanding these mechanisms will be important for the design of programs for synchronization of the estrous cycle and ovulation, and for superovulation and embryo transfer.












CHAPTER 3
HUMAN CHORIONIC GONADOTROPIN-INDUCED ALTERATIONS IN
OVARIAN FOLLICULAR DYNAMICS DURING THE ESTROUS CYCLE OF HEIFERS


Introduction



The understanding of factors controlling ovarian folliculogenesis should lead to development of new and more precise methods of controlling reproductive cycles in livestock. One aspect of follicular dynamics is the process of continual growth and regression of antral follicles that leads to development of the preovulatory follicle (Lucy et al., 1992). One to four waves of follicular growth and development occur during a single estrous cycle in heifers and cows, and the preovulatory follicle is derived from the last wave. During each wave, development of antral follicles > 2 mm proceeds through stages of follicular recruitment, selection, and dominance (Savio et al., 1988; 1990; Lucy et al., 1992; Fortune, 1994; Ginther et al. 1996). Development of the first dominant follicle (DF) is the most consistent event related to follicular dynamics during the estrous cycle in heifers and cows. The first dominant follicle can be identified retrospectively by ultrasonography between day 2 and 4 of the estrous cycle (Savio et al., 1988; 1990; Sirois and Fortune, 1988; Driancourt et al., 1991; Ginther et al., 1996).

Human chorionic gonadotropin (hCG) is a glycoprotein with LH-like activity







74

that promotes ovulation of the first wave DF (FWDF) and formation of an accessory corpus luteum (CL) when administered on day 4 (Breuel et al., 1989; Price and Webb, 1989), day 5 (Walton et al., 1990; Schmitt et al., 1996a, 1996b), day 6 (Fricke et al., 1993), day 7 (Breuel et al., 1989; Rajamahendran and Sianangama, 1992; Sianangama and Rajamahendran, 1992; 1996), day 10 (Breuel et al., 1989), or day 14 to 16 (Price and Webb, 1989) of the estrous cycle. With formation of an accessory CL, there is a subsequent increase in concentrations of plasma progesterone (P4; Breuel et al., 1989; Fricke et al., 1993; Schmitt et al., 1996a). Only one study has characterized ovarian follicular dynamics during this period (Sianangama and Rajamahendran, 1996). The objective of the present experiment was to characterize hCG-induced alterations in ovarian follicular and CL dynamics in heifers when hCG was injected on day 5 of the estrous cycle.


Materials and Methods


Seventeen sexually mature Holstein heifers were selected from the dairy herd at the University of Florida. Heifers were kept in a lot with minimal pasture and fed peanut hay and corn silage. Estrous cycles were synchronized following removal of a norgestomet ear implant (6 mg norgestomet implant without injection of the norgestomet/estradiol valerate solution; Synchromate-B; Sanofi Animal Health, Inc., Overland Park, KS) that had been in place for 7 days and injection of prostaglandin (PG) F2a (25 mg i.m., Lutalyse, Pharmacia & Upjohn Co., Kalamazoo, MI) on day 6 or 1 day before withdrawal of the implant. Brightly colored enamelbased paint (Impervo; Benjamin Moore, Montvale, NJ) was applied to the tailhead







75
of each heifer at the time of PGF2a administration. When the ear implant was removed, the paint strip was covered with a contrasting color of chalk (All-WeatherO Painstick�, Lake Chem. Co., Chicago, IL; Van Cleeff et al., 1996). The extent in cover of tail paint and chalk (TPC) was scored based on a modification of the method of Macmillan et al. (1988) using a scale of 5 (no signs of estrus and full presence of paint and chalk) to 0 (standing estrus and absence of paint and chalk; Van Cleeff et al., 1996). Scoring of the TPC and visual detection of estrus were performed twice daily (0600 to 0700 and 1800 to 1900) for 3 days after removal of the implant. On day 5 of the estrous cycle (day 0= day of estrus), heifers (n=6) were treated with hCG (1,000 IU given i.v.; 2,000 IU given i.m.; Steris Laboratories, Phoenix, AZ; n=6; Schmitt et al., 1996a) or i.m. injection of saline (control group; n=7). A real time ultrasound scanner equipped with a linear array 7.5 MHz transrectal transducer (Equisonic LS 1000, Tokyo Keiki Co., LTD, Tokyo, Japan) was used to monitor follicular development daily from day 3 of the estrous cycle until the next ovulation (first cycle). During the following estrous cycle (second cycle), follicular development and CL growth were monitored every other day until estrus. Ovarian structures (follicles > 2 mm and CL) were measured, and their relative positions recorded on ovarian maps drawn during the examination. Blood samples were taken daily during the first and second estrous cycles. The dominant follicle for each follicular wave was defined as the largest newly emerged ovarian follicle that was 2 mm greater than the second largest follicle and grew linearly (Sirois and Fortune, 1990).

Blood samples (10 ml) were collected daily via venipuncture from the jugular







76
vein into heparinized tubes. Blood samples were stored in an ice bath, plasma separated by centrifugation (1,800 x g for 15 minutes), and stored at -20 oC until assayed for P4 and estradiol (E2). Concentrations of P4 in plasma were measured by radioimmunoassay (Knickerbocker et al., 1986). Sensitivity of the assay was 0.15 ng/ml. Intra- and inter-assay coefficients of variation were 4.6 and 13.2%, respectively. Concentrations of E2 from day 6 before ovulation until day 7 of the second cycle were measured by a radioimmunoassay described by Tortonese et al. (1990) and modified by Badinga et al. (1992). Sensitivity of the assay was 1.0 pg/mi and the intra- and inter-assay coefficients of variation were 6.9 and 19.9%, respectively.


Statistical Analyses


Data were analyzed by least squares analysis of variance using the General Linear Models procedure of the Statistical Analysis System (SAS, 1988). The following definitions were used to characterize each follicular wave: a) day of emergence of DF was the day when the DF first appeared in the ovary and was at least 5 mm in size, b) day of selection of DF was the day of the cycle when the difference between DF and second largest follicle (subordinate) was > 2 mm, c) day of maximum recruitment was the day when maximum number of class 2 follicles (69 mm) was present in the ovary, and d) duration of a follicular wave was the difference between day of emergence of the DF of two consecutive waves.

The mathematical models used to analyze size of dominant and subordinate follicles and number of follicles within discrete size classes (Class 1, 2 to 5 mm;







77

Class 2, 6-9 mm; Class 3, > 10 mm) included effects of treatment, cow(treatment), day of the wave (wave II, days 1 to 18 [days relative to day of emergence of second wave DF] and wave III, days 1 to 13 [days relative to day of emergence of third wave DF]), and the interaction of day of the wave by treatment. The main effect of treatment was tested using the mean square of cow (treatment) as the error term. Tests for homogeneity of regression equations were performed on equations for follicular diameter versus day of wave for dominant and subordinate follicle of the second and third follicular waves. These analyses were performed as described by Wilcox et al. (1990).

The same analytical approach was used to analyze plasma concentrations of P4 during the first and second estrous cycles for hCG and control heifers, plasma concentrations of E2 during the peri-ovulatory period, sizes of the original and induced CL, and patterns of growth of original CL in hCG-treated and control heifers. However, day of cycle (P4), days from ovulation (E2), and days after ovulation (original vs. induced CL, and comparison of original CL growth for both groups) were used for these analyzes. The mathematical model used to analyze the follicular dynamics and plasma concentrations of P4 between heifers with two and three follicular waves during the first and second estrous cycles did not include hCG-treated heifers during the first estrous cycle. The model included effects of wave, cycle, cow(wave-cycle), day, and interactions wave by day, cycle by day, and wave by cycle by day. The main effects of wave and cycle were tested using the mean square of cow(wave-cycle) as the error term.

Duration of the second and third follicular waves in all heifers with three-wave







78

cycles were analyzed using a mathematical model that included the effects of treatment, wave(second or third wave), and the interaction of treatment by wave. The frequencies of accessory CL and estrous cycles of two versus three follicular waves between treatments were compared using Chi-square analyses.


Results


Corpora lutea and plasma progesterone


Treatment with hCG at day 5 induced ovulation and formation of an accessory CL in all six heifers of the hCG group, while none of the control heifers ovulated the first wave dominant follicle (X2= 13, P < 0.01). Progesterone concentrations increased from 0.9 ng/ml on day 1 to 6.4 ng/ml on day 6 and did not differ (P > 0.10) between heifers in both treatments. However, by day 9, P4 concentrations were greater (P < 0.01; Figure 3-1A) for heifers of the hCG group (15.3 ng/ml) than for control heifers (11.1 ng/ml) and remained elevated until day 17. Maximal concentrations of P4 were observed on day 15 (hCG group, 20.4 vs. control group, 12.9 ng/ml). On day 19.5 and 20.3 of the estrous cycle, there was a 50% decrease in P4 concentrations for hCG and control groups, respectively, indicating that luteolysis had been initiated; P4 concentrations were < 1 ng/ml on day 21 for both groups. Concentrations of P4 during the following estrous cycle did not differ between groups (P > 0.10). The profile was the same in both groups, reaching maximal concentrations on day 13 (11.3 vs. 10.6 ng/ml, for hCG and control groups, respectively; Figure 3-1 B).















E 3(3000iu
15- crsalire 15C C o 0
010- 100
o 0
. a .

%- I- I7


0 5 10 15 20 25 0 5 10 15 20 25
framOacn DefrCmOaIM

Figure 3-1. Plasma concentrations of progesterone (LSM � SEM) during the first (A) and second
(B) estrous cycle in control (0) and hCG-treated (0) heifers. Asterisk (*) indicated significant differences (P < .01) between mean concentrations within day.




There was a difference (P < 0.01; Figure 3-2) in concentrations of P4 between heifers with two (13 heifer-cycle cells) and three follicular waves (8 heifercycle cells). This comparison was made on all heifers except hCG-treated heifers of the first experimental cycle. Concentrations of P4 from day 13 to 20 were lower in heifers having two-wave cycles compared with three-wave cycles. There was a 50% decrease in P4 concentrations by day 17.7 in heifers with two waves and day 21.3 in heifers with three follicular waves, which is indicative that the luteolytic mechanism from the uterus was active earlier in heifers with two follicular waves (Figure 3-2).

Dynamics of CL development differed (P < 0.001) between groups due to







80

15- **** ***


12

E
o9


O p

3





0 5 10 15 20 25
Days from Ovulation

Figure 3-2. Plasma concentrations of progesterone (LSM + SEM) for heifers with two (m) and three (0-) follicular waves during the first and second estrous cycles. Asterisk (*) indicates differences (P < .01)
between mean concentrations within day.


induction of accessory CL in the hCG-treated heifers (Figure 3-3). In hCG-treated heifers, growth of the original and induced CL did not differ (P > 0.10) from day 3 to day 8 after ovulation. The accessory CL was smaller (P < 0.01) than the original CL and regressed concurrently with the original CL in hCG-treated heifers. There was no difference (P > 0.10) in average size between original CL of hCG-treated and control heifers (Figure 3-3A and 3-3B), but there was a treatment by day interaction (P < 0.001) for the pattern of growth of the original CL in hCG-treated and control heifers. Between day 3 and 11 of the estrous cycle, original CL of hCGtreated heifers grew to a larger size following injection of hCG.
0)
2 Q.
3


0
0 5 10 15 20 25 Days from Ovulation

Figure 3-2. Plasma concentrations of progesterone (LSM �SEM) for heifers with two (U) and three (0J) follicular waves during the first and second estrous cycles. Asterisk (*) indicates differences (P < .01)
between mean concentrations within day.


induction of accessory CL in the hCG-treated heifers (Figure 3-3). In hOG-treated heifers, growth of the original and induced CL did not differ (P > 0. 10) from day 3 to day 8 after ovulation. The accessory CL was smaller (P < 0.01) than the original CL and regressed concurrently with the original CL in hCG-treated heifers. There was no difference (P > 0.10) in average size between original CL of hCG-treated and control heifers (Figure 3-3A and 3-313), but there was a treatment by day interaction (P < 0.001) for the pattern of growth of the original CL in hCG-treated and control heifers. Between day 3 and 11 of the estrous cycle, original CL of hOGtreated heifers grew to a larger size following injection of hOG.

















4- fi15
E
o E E 30- A..,30- B $ hOG
( (3000 iu E al E25- -10I25- . Ca5
E E

-J-J

0.15- 150.

8 815
10-I I I 10- I I I I
0 5 10 15 20 25 0 5 10 15 20 2
ay d Eslrs Oyde Day ~cEMsQde

Figure 3-3. Growth of the original (M) and induced (0) corpora lutea in hCG-treated heifers (A) and original (A) corpus luteum in control (B) heifers during the first estrous cycle. Asterisk (*) indicates differences (P < .01) between least squares means within day.


Follicular Dynamics


All hCG-treated heifers had three follicular waves during the first experimental estrous cycle (Figure 3-4). In the control group, in contrast, three heifers had three follicular waves (Figure 3-5A) and four had two follicular waves (Figure 3-5B). The frequency of three follicular waves was greater (X2= 4.95; P <

0.05) in the hCG group.

The second wave dominant follicle emerged earlier in hCG-treated heifers (day 7.3; Figure 3-4) compared with the control group (day 10.4; P < 0.01; Figure 3-5A and 3-5B), and reached 10 mm in diameter on day 10 of the estrous cycle for hCG-treated heifers compared with 13.6 day for heifers in control group with either










16- hCG
3 000 iu
1412
E
Iv
E 8

0


2
0
0 5 10 15 20 25 Day of Estrous Cycle

Figure 3-4. Growth of the first (0), second (0), and third (A) dominant follicles in hCG-treated heifers (n=6) during the first
estrous cycle.

two or three follicular waves. In both groups, the second wave dominant follicle was selected on day 3.4 of the wave, and day of maximum recruitment was day 3.7. A treatment by wave interaction was detected (P < 0.01) for duration of waves for heifers with three follicular waves during the first cycle (Figure 3-4 and 3-5A). In the control group, the second and third follicular waves lasted 8.3 � 0.9 days and 5.3 � 0.9 days, respectively. For the hCG group, duration of the second follicular wave was shorter (6.3 � 0.7 days), possibly due to greater P4 concentrations whereas duration of the third wave was longer (9.2 � 0.7 days) than in the control group. Maximum diameter of the second wave dominant follicle was smaller (P < 0.01) in hCG treated-heifers (12.8 � 0.8 mm; Figure 3-4) than in control heifers (15.6 � 0.8 mm; Figure 3-5A) with three-wave cycles. Associated with the differences in duration of third wave follicles (hCG, 9.2 days > control, 5.3 days; P < 0.01) were differences in follicular dominance as measured by growth patterns of the largest subordinate follicle. The subordinate follicle of the hCG group decreased in size










18- A 18- B

16- 1614- 14S12- 12


E E
m6- 8-6z
4- 4

2 2 0 0
0 3 6 9 12 15 18 21 24 27 0 3 6 9 12 15 18 21 24 2 eWd~juzsode DdEa usOde Figure 3-5. Growth of the first (U), second (0), and third (A) dominant follicles in control heifers with three (n= 4; A) and two (n= 3; B) follicular waves during the first
estrous cycle.

during the 6 days of the preovulatory period (b= -0.007 mm/day). In contrast, size of the subordinate follicle of the control group increased during this period (b= 0.47 mm/day; P < 0.01). This may be a reflection of a more developed and active preovulatory dominant follicle in hCG-treated heifers that suppressed growth of the subordinate follicle. The third dominant follicle emerged on day 14.7 and 18 of the estrous cycle for heifers of the hCG and control groups, respectively, selection occurred on day 3.7 of the third wave for both groups.

During the second estrous cycle, one hCG-treated heifer had three follicular waves and five had two follicular waves, while three saline-treated heifers had three follicular waves and four heifers had two follicular waves. A total comparison of twoand three-wave cycles, excluding first cycles for hCG-treated heifers, was made







84

16- A 16- B

14- 1412- 12EE
10- 10E8- E86- 6
.2
6 4- 4L. LL.
2- 2

0 1I I I I I I 0 I I I 1 I I I I
0 3 6 9 12 15 18 21 24 27 0 3 6 9 12 15 18 21 24 2
Dayof Estrous Oyde tay of Estus Ode

Figure 3-6. Growth of the first (U), second (0), and third (A) dominant follicles in heifers with two (n=13; A), and three (n=8; B) follicular waves during the first and second estrous
cycle. Heifers treated with hCG during first estrous cycle are not included.

utilizing first and second estrous cycles. The second wave dominant follicle in twowave cycle heifers had a longer life span (11 days) than the second dominant follicle of three-wave cycle heifers (6 days; Figure 3-6A and 3-6B). Similarly, there was a significant wave by day interaction (P < 0.01) for size of the ovulatory follicle in heifers with two and three follicular waves. The ovulatory follicle of heifers with two-wave cycles entered a plateau phase for 3 days before ovulation; whereas, growth of the ovulatory follicle of heifers with three-wave cycles was linear until ovulation.


Estradiol-1711 Concentrations


Plasma concentrations of E2 from day 6 before ovulation until 7 days post ovulation did not differ (P > 0.10; Figure 3-7A) between treatments. Heifers treated









22-N 22- B
2)" 2018- 1816- 16~14- 1412- 6.122 .2
0 V
6- 6
*~42- 21

-6 -4 -2 0 2 4 6 8 -6 -4 -2 0 2 4 6


Figure 3-7. Plasma concentrations of estradiol-171rS (LSM + SEM) in hCG-treated (0) and control (0) heifers (A), and control heifers with two follicular waves (M; n= 4; B) and three
follicular waves (O; n=3; B) during the peri-ovulatory period.

with hCG tended to have greater E2 concentrations (15.6 pg/ml) than heifers of the control group (8.2 pg/ml) on the day before ovulation. This trend was due to one heifer that had very high concentrations of E2. Following ovulation, plasma concentrations of E2 increased to 8.7 pg/ml on day 4 for hCG-treated heifers and to 5.6 pg/ml on day 5 for the control group. This increase in E2 did not differ between treatments and coincided with growth of the first wave dominant follicle before the luteal phase increase in concentrations of P4. Plasma concentrations of E2 tended (P < 0.10) to be greater from day -3 until day of ovulation (day 0; Figure 3-7B) in control heifers with two follicular waves (n=2) than in control heifers with three follicular waves (n=3). This coincides with the fact that the second dominant







86

follicle was present longer on the ovary. The ovulatory follicle in control heifers with two follicular waves reached a size> 10 mm on day 13.5 of the estrous cycle versus day 22.3 for control heifers with three follicular waves (Figure 3-5A and 3-5B).


Duration of Estrous Cycles


Duration of the first estrous cycle for hCG group and control group was 22.9 S0.9 and 22.1 � 0.9 days (P > 0.10), and 21.8 � 1.0 and 22.7 � 1.0 days (P > 0.10) for the second estrous cycle. Analysis of the first and second experimental cycles, excluding the first estrous cycle of hCG-treated heifers, indicated that estrous cycles with two waves of follicular development were shorter (20.5 � 0.5 days; P > 0.01) than those of three-wave estrous cycles (24.6 � 0.7 days).


Discussion


Cows and heifers exhibit two or three successive waves of follicular development during the estrous cycle (Savio et al., 1988; Sirois and Fortune, 1988; Ginther et al., 1989a; 1989b). In this study, the FWDF present at day 5 ovulated in response to hCG and formed a CL which is in agreement with others (Breuel et al., 1989; Price and Webb, 1989; Walton et al., 1990; Rajamahendran and Sianangama, 1992; Sianangama and Rajamahendran, 1996; Schmitt et al., 1996a, 1996b). Previous work indicated that the response was affected by stage of the estrous cycle when hCG was given. Induction of an accessory CL was greater on days 4 to 7 and days 14 to 16 than between days 8 to 13 of the estrous cycle (Price and Webb, 1989). This is due likely to transitional stages in follicular status within







87

the first (Badinga et., 1992; Thatcher et al., 1996) and second follicular waves that are either responsive or nonresponsive to hCG. Schmitt et al. (1996a) reported that the greater growth of original CL in hCG-treated compared to control heifers could be due to a luteotropic effect of hCG on development of the original CL after day 5.

In the present study, hCG acutely altered follicular dynamics by ovulating the first wave dominant follicle on day 5. All hCG-treated heifers had three follicular waves during the experimental estrous cycle. Elimination of the first wave dominant follicle via hCG induction of ovulation leads to an earlier recruitment and emergence of the second-wave DF. Badinga et al. (1992) reported that removing the ovary bearing the DF increased FSH leading to follicular recruitment, thus, removal of the FWDF by hCG-induced ovulation may have had the same effect in this study. Ko et al. (1991) reported that destruction of the FWDF later during the wave is followed by an earlier recruitment of the next wave. A comparable effect appears to occur with elimination of the DF via induction of ovulation with hCG.

Although an earlier emergence of the second wave follicle occurred in hCGtreated heifers, maximum size of the second wave follicle was reduced and had a shorter lifespan. This was attributed to the greater P4 concentrations associated with induction of an accessory CL in this group. Indeed, greater concentrations of P4 reduce growth of the DF and induce follicle turnover (Savio et., 1993; Burke et al., 1994; Thatcher et al., 1994; Kinder et al., 1996) due to reduction in LH secretion (Kinder et al., 1996). Functional duration of the third wave follicle was increased because of a shorter life span of the second-wave follicle in hCG-treated heifers,. Indeed, a greater decrease in size of the subordinate follicle during the preovulatory







88

period for the third wave dominant follicle of the hCG group was detected compared to that of the preovulatory subordinate follicle of three-wave cycles in the control group. Regardless of day of emergence or duration of a wave, dominant follicles were selected between day 3 and 4 of the respective wave. Ginther et al. (1996) reported that between day 0 and 4 of a wave, the dominant and largest subordinate follicles diverge gradually in diameter.

Administration of hCG during the luteal phase of the estrous cycle increased P4 concentrations in blood and in some cases extended duration of the estrous cycle (Seguin et al., 1977; Breuel et al., 1989; Walton et al., 1990; Rajamahendran and Sianangama, 1992; Fricke et al., 1993; Schmitt et al., 1996a). This effect appears to be somewhat dependent on presence of a functional CL because hCG administered on day 1 (early luteal phase) or day 17 (late luteal phase) of the estrous cycle had no effect on serum P4 concentrations (Seguin et al., 1977; Breuel et al., 1989). In this experiment, administration of hCG on day 5 increased plasma P4 concentrations from day 9 to 17. The increase in P4 concentrations was due most likely to additional P4 secretion by accessory CL (Schmitt et al., 1996b) at this stage of the estrous cycle as opposed to a stimulation of the original CL (Fricke et al., 1993). Based upon diameter of the induced CL and the temporal increase in plasma P4 concentrations, induced CL appear to be functional but smaller than the original CL. Sianangama and Rajamahendran (1996) demonstrated that the CL induced by hCG given on day 7 of the bovine estrous cycle is functional but appears to be smaller and secretes less P4 than the spontaneous CL of similar age.

Because concentrations of plasma P4 during the second cycle were similar








89

between groups, luteotrophic effects and alterations in dynamics of follicular growth observed in the first cycle after hCG administration failed to elicit any carry-over effects that influenced CL development and function of the second cycle. Similarly, Breuel et al. (1989) did not observe any effect of hCG on CL formed during subsequent cycles.

Greater concentrations of plasma P4 in heifers with three spontaneous follicular waves during the cycle could contribute to a greater turnover of the second wave dominant follicle as a consequence of decreased LH pulsatility (Kinder et al., 1996). Estrous cycles were significantly longer for three-wave cycles and growth characteristics of the ovulatory follicle were different between three-wave versus two-wave follicle cycles. The duration of the luteal phase appears to determine, at least in part, the number of follicular waves during an estrous cycle (Fortune, 1994). Estrous cycles with three follicular waves have slightly but significantly longer luteal phases than cycles with two waves (Ginther et al., 1989b; Fortune, 1993). We choose to interpret that two-wave follicular cycles induce an earlier CL regression compared to three-wave cycles. Contributing to this is presence of an active estrogenic follicle when CL regression is initiated. This could be determined partially by high plasma P4 concentration leading to a turnover of the second wave follicle versus persistence of a potentially estrogenic second wave follicle in luteal phase with lower progesterone concentrations.

Third wave ovulatory follicles underwent a linear growth until ovulation versus a distinct plateau phase for the second wave ovulatory follicle. Whether fertility differs to insemination following ovulation of two versus three wave ovulatory







90
follicles, needs to be examined. Any differences in development of the third wave dominant follicle between hCG and control heifers were not reflected in differences of plasma E2. Howard etal. (1990) reported that hCG (10,000 IU, i.m.) administered on day 10 caused a transient increase in E2 with normal concentrations observed during the subsequent pre-estrual/estrual period. The tendency of higher concentrations of E2 in control heifers with two waves versus three waves could reflect the presence of a more persistent estrogenic dominant follicle that is less fertile than a less estrogenic third wave dominant ovulatory follicle.

An additional factor that may regulate fertility is the occurrence of two- versus three-wave cycles after insemination. Heifers with two-wave cycles produced a potentially estrogenic preovulatory follicle (L> 10 mm) as early as day 13.5 versus day 17.7 for three-wave cycles. Since the signal mechanism induced by the conceptus to extend the CL lifespan occurs between 15 to 17 days, it would appear that cows with three-waves cycles would be less likely to potentially antagonize the antiluteolytic mechanism. Injection of estradiol stimulates uterus secretion of PGF2a (Knickerbocker et al., 1982) and cauterization of follicles extends CL lifespan (Fogwell et al., 1985). Thus heifers with three-wave cycles may conceivably have a higher rate of embryo survival. Reproductive management systems to optimize ovarian follicular development in either the preovulatory period or following insemination to improve conception rate and embryo survival warrant further investigation.




Full Text

PAGE 1

MECHANISMS THAT CONTROL FOLLICULAR DOMINANCE IN CATTLE By THAIS DEL VALLE DIAZ ZAMBRANO ' A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1998

PAGE 2

To my parents, Francisco and Maria Antonia (Tona), my brother, Francisco Javier, my sisters, Ana Maria, Flor Alba, Yuritza and Eisa, and my late uncle, Juan.

PAGE 3

especially Mr. Dale "Valentine" Hissem, for his friendship and endless support during the long days of farm work. Graduate school is less difficult because of the support of other graduate students and postdoctoral fellows. Special thanks go to Dr. Alfredo N. Garcia, Mr. Tomas Ignacio Belloso, Mr. Andres A. Kowalski, Dr. Carlos Arechiga, Dr. Alice de Moraes, Ms. Susan Gottshall, Ms. Rocio Rivera, Dr. Ricardo Mattos, Mrs. Jennifer Phillips Trout, Dr. Amelia Luengo, Dr. Carlos A. Vargas, Dr. Guenahel DanetDesnoyer, Dr. Florence Ndikum Moffor, Mr. Daniel Arnold, Dr. Fabiola Paula Lopes, Dr. Frederico Moreira, Dr. Lannett Edwards, Dr. Brian Cleaver, Ms. Heather Greaves, Mr. Morgan Peltier, Mr. Aydin Guzeloglu, Mr. Marvin Hoekema, Mrs. Mika Robinson, Mrs. Peggy Briggs, and Mr. Jesse W. Johnson. Special thanks go to Mrs. Mary Ellen Hissem, for her endless smile and support, for her sincere friendship, for those lunches during the beautiful days of the fall, to Mrs. Joyce Hayen and Mrs. Helen Hester, for their constant support, friendship and sharing of the good and not so good moments. They will always have a very special place in my heart and my eternal gratitude. I want to thank Dr. H. Herbert Head, Graduate Coordinator, for his friendship and the chatting times shared, and Dr. Daniel C. Sharp for his friendship. I want to acknowledge the Universidad Central de Venezuela, my Alma Mater, and Consejo Nacional de InvestigacionesCientificasyTecnologicas for their financial support. Special thanks to my fellows at the University in Venezuela: Drs. Juan F. Troconiz, Pedro S. Bastidas, Magaly R. Manzo, Luis A. Vasquez, Nora iv

PAGE 4

Guerrero, Oswaldo A. Silva and Beatriz Quintero, for their support and understanding during these years of my absence. Mil gracias a todos. Also, life in Gainesville was easier because of the support of very special friends Teresita, Tomas and Matias de la Sota, Carlos Rene, Carmen, Jorge Juan and Caroly Beltran, Elsa, Abraham, Manuel e Indira Garcia, Dr. Maria Eugenia Cadario Gonzalez-Pola, Dr. Michael Byron Porter (thank you for teaching me to enjoy the "magnificient" Colombian music), Francisco, Giordana, Johanna, Pedro and Fabianna Ovalles, Diego and Maria del Carmen Rochinotti, and Tomas Ignacio Belloso (thank you for being there during my last months in Gainesville) and thanks for making me feel part of your lives and families, and for having warm words when they were necessary. My gratitude will be forever and there is a very special place for all of you in my heart. I have always thought that family helps you to have courage to work towards your goals. Without my family's support it would have been impossible to pursue my goals. Thank you for your faith, your love, your patience, encouragement and moral support. V

PAGE 5

TABLE OF CONTENTS ACKNOWLEDGMENTS iii ABSTRACT ix CHAPTERS 1 INTRODUCTION 1 2 LITERATURE REVIEW 4 Ovarian Follicular Growth and Development 4 Early Stages of Follicular Development 5 Ovarian Follicular Dynamics During the Bovine Estrous Cycle . . 9 Structure and Function of the Antral Follicle 21 The Granulosa Cell Compartment 23 The Theca Cell Compartment 29 Two-Cell, Two-Gonadotropin Theory 31 Molecular Events Occurring in Theca and Granulosa Cells .... 34 Hormonal Control of Ovarian Follicular Dynamics 38 Steroid Hormones 38 Gonadotropic Hormones 40 Other Factors Involved in Control of Ovarian Follicular Dynamics 43 Inhibin 45 Activin 50 Follistatin . 55 Transforming Growth Factor p 57 Epidermal Growth Factor/Transforming Growth factor-a 60 Insulin-Like Growth Factors (IGFs) and IGF-Binding Proteins (IGFBPs) 63 Manipulation of the Follicular Dynamics in Cattle 66 Hormonal Manipulation: Human Chorionic Gonadotropin and Gonadotropin Releasing Hormone 68 Ablation of the Dominant Follicle 69 Superstimulation 70 vi

PAGE 6

3 HUMAN CHORIONIC GONADOTROPIN-INDUCED ALTERATIONS IN OVARIAN FOLLICULAR DYNAMICS DURING THE ESTROUS CYCLE OF HEIFERS 73 Introduction 73 Material and Methods 74 Statistical Analyses 76 Results 78 Copora Lutea and Plasma Progesterone 78 Follicular Dynamics 81 Estradiol-1 7(3 Concentrations 84 Duration of Estrous Cycle 86 Discussion 86 Implications 91 4 EFFECTS OF FSH-P ON FOLLICULAR DYNAMICS AND OVARIAN RESPONSE TO A SUPEROVULATORY TREATMENT FOLLOWING ASPIRATION OF A FIRST WAVE PERSISTENT DOMINANT FOLLICLE 92 Introduction 92 Materials and Methods 94 Experiment 1 94 Experiment 2 98 Statistical Analyses 99 Results 101 Follicular Dynamics 101 Plasma Progesterone and Estradiol-1 7p Concentrations .... 105 Ovarian Responses to Superovulation Treatment 106 Discussion 107 Implications 113 5 IN VITRO SECRETION OF ESTRADIOL BY BOVINE ANTRAL GRANULOSA CELLS 115 Introduction 115 Materials and Methods 118 Granulosa Cell Isolation 118 Granulosa Cell Culture 120 Granulosa Cell Number 121 Radioimmunoassays 123 Gel Electrophoresis 125 Statistical Analyses 126 Results 126 Time Course of Basal Hormone Secretion 126 vii

PAGE 7

Estradiol 126 Progesterone 129 Effect of FSH-Stimulated Secretion 130 Estradiol 130 Progesterone 134 Effect of FSH on Cell Number 137 Effect of FSH on Cell Morphology 1 38 Gel Electrophoresis 140 Discussion 141 Implications 145 6 EFFECT OF BOVINE FOLLICULAR FLUID FROM DAY 5 DOMINANT AND DAY 12 ATRETIC DOMINANT FOLLICLES ON IN VITRO ESTRADIOL-1713 SECRETION BY BOVINE ANTRAL GRANULOSA CELLS 147 Introduction 147 Material and Methods 151 Inhibin Immunoaffinity Column 151 Pools of Bovine Follicular Fluid 152 Electrophoresis and Western Blotting Procedures 155 Granulosa Cell Culture System 158 Statistical Analyses 161 Results 163 Validation of Dextran-Coated Charcoal Treatment 163 Western Blot Analysis of Inhibin in Follicular Fluid 167 Effect of Bovine Follicular Fluid on FSH-Stimulated Estradiol-1 7(3 Secretion 171 Estradiol-17p 171 Progesterone 176 EstradiolProgesterone ratio 177 Discussion 181 Implications 184 7 GENERAL DISCUSSION AND CONCLUSIONS 186 REFERENCES 203 BIOGRAPHICAL SKETCH 228 viii

PAGE 8

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy MECHANISMS THAT CONTROL FOLLICULAR DOMINANCE IN CATTLE By Thais del Valle Diaz Zambrano May 1998 Chairperson: William W. Thatcher Major Department: Animal Science . A series of experiments examined potential mechanisms that control follicular dominance in cattle. On day 5 of the estrous cycle, ablation of the first wave dominant follicle with an injection of human Chorionic Gonadotropin (hCG; 3,000 lU) resulted in ovulafion of the follicle, formation of an accessory corpus luteum and higher concentrations of plasma progesterone in treated heifers. The second wave dominant follicle emerged earlier in hCG-treated heifers and all treated heifers had three follicular waves. Treatment with hCG initiated turnover of the dominant follicle and altered subsequent follicular dynamics. A second experiment tested the effect of follicle-stimulating hormone-pituitary (FSH-P) injections beginning at day 12 on folliculogenesis and ovulatory responses in the presence or absence of a first wave persistent dominant follicle. Follicle recruitment and induction of class 2 (6 to 9 mm) ix

PAGE 9

and class 3 (>9 mm) follicles were delayed during the superlnduction period in the presence of a persistent dominant follicle. However, no differences were detected between groups in the final number of ovulatory follicles, corpus luteum and recovered embryos. Inhibitory effects of a persistent dominant follicle were overridden with continuous injections of FSH. A granulosa cell culture system was developed to test the effects of intraand interovarian regulators (inhibin forms) on granulosa cell secretion of estradiol. Purified porcine FSH (CY-FSH) stimulated estradiol secretion of granulosa cells between 72 to 96 h of culture. Also evaluated were effects of different fractions of follicular fluid (follicular fluid complete, follicular fluid fraction without inhibins and an inhibin-enriched fraction) from day 5 dominant and day 12 atretic dominant follicles on estradiol secretion of granulosa cells. A 10 ng/ml concentration of CY-FSH stimulated estradiol secretion compared to 0.5 ng/ml. Fractions of follicular fluid without inhibins and enriched in inhibins from day 5 dominant follicles decreased estradiol secretion. In contrast, follicular fluid complete and follicular fluid without inhibins from day 12 atretic dominant follicles inhibited estradiol secretion, but the enriched fraction of inhibins stimulated granulosa cells to secrete estradiol. Collectively, the series of in vivo and in vitro experiments documented that the dominant follicle exerts both interovarian and intrafollicular effects on follicle function in cattle. X

PAGE 10

CHAPTER 1 INTRODUCTION On the basis of gross and histological studies of ovaries, it was proposed in 1960 that two waves of follicular activity occurred during the bovine estrous cycle (Rajakoski, 1960). After measuring follicles and quantifying steroids in blood and follicular fluid, Ireland and Roche (1983) concluded that there were three follicular waves during the estrous cycle, and each wave resulted in the formation of a dominant follicle. In 1984, Pierson and Ginther, characterized bovine ovarian folliculogenesis, for antral follicles > 3 mm by repeated daily ultrasound measurement and concluded that heifers had estrous cycles with two follicular waves. However, others (Fortune et al., 1988; Savio et al., 1988, Sirois and Fortune, 1988) reported three waves of follicular development during the estrous cycle. Research in this area is motivated by the desire to solve the long-time mystery involving the mechanisms underlying recruitment, selection and development of follicular dominance that is characteristic of monovulatory species (Ginther etal., 1996). In general two processes lead to development of the normal, species-specific number of ovulatory follicles. First, follicle recruitment results in the development of a pool or cohort of follicles from which the preovulatory follicle emerges as a 1

PAGE 11

2 result of follicle selection. The follicle becomes dominant and continues development towards ovulation, while others regress (Fortune et al., 1991). This pattern of dominant follicle development also occurs during the luteal phase of the estrous cycle in which dominant follicles undergo atresia instead of ovulation. The phenomenon of dominance is clearly of interest because mechanisms of follicular dominance must be subverted to successfully superovulate domestic animals or humans. Furthermore, an understanding of dominant follicle development and dominance is important to the development of reproductive control system to synchronize ovulation in dairy cattle. At present, little is known about mechanisms by which a follicle attains and maintains dominance (Fortune, 1993). To study follicular dominance, differences between morphological and functional dominance have to be established. Morphological dominance is defined as the largest developing follicle on either ovary. The functionally dominant follicle appears to have the ability to inhibit growth of smaller follicles and can ovulate under appropriate hormonal conditions (Fortune, 1993). Basic research is needed to elucidate the mechanisms controlling dominance during follicular development in cattle to optimize the techniques to control reproductive cycles in cattle. Objectives of the present research were: a) to characterize alterations in ovarian follicular and corpus luteum dynamics after injection of hCG on day 5 of the estrous cycle; b) to evaluate the effect of folliclestimulating hormorne-pituitary (FSH-P) on superinduction of follicles and ovulatory responses in the presence or absence of an active-first wave persistent dominant

PAGE 12

3 follicle, and to elucidate the role of inhibins as intraovarian regulator of estradiol secretion during different periods of follicular dominance. The last objective require development of an in vitro system to examine regulation of estradiol secretion by bovine antral granulosa cells.

PAGE 13

CHAPTER 2 LITERATURE REVIEW Ovarian Follicular Growth and Development Reproductive cyclicity in the female is maintained by endocrine, paracrine, and autocrine interactions. The central role of pituitary follicle-stimulating hormone (FSH) and luteinizing hormone (LH) is to direct ovarian follicular recruitment, growth, atresia, and ovulation (Richards and Hedin, 1988). The ovary has two main functions: production of the oocyte, the female gamete, and to provide the hormonal environment to sustain reproductive function through the production of steroid hormones, such as progesterone (P4), estradiol-17R (Ej), and other regulatory factors. Gonadotropin secretion influences, among other things, development of follicles in the ovary. The factor or factors that determine the early stages of follicular growth in the ovary at a pre-determined time are unknown. The fate of a follicle could be to remain quiescent, to begin development but later become atretic, or to mature and ovulate. In mammals the number of primordial follicles present in the ovaries is fixed at birth or in the immediate post-natal period depending on the species. In the cow, between 40,000 to 80,000 quiescent primordial follicles can be found at birth, but this number is reduced to approximately 3,000 by 15 to 20 years of age (Erickson, 1 966). Since 99% of the primordial follicles fail to ovulate during

PAGE 14

the lifespan of an animal, development of an ovulatory follicle is a rare biological event and folliculogenesis is a complex process (Ireland, 1987). Early Stages of Follicular Development The initial stages of folliculogenesis occur independently of gonadotropic hormones. Antral follicles initially become responsive to FSH and then are dependent on FSH for sustained growth. Inhibins, activin, insulin-like growth factor I (IGF-I) and their binding proteins have direct and indirect effects on granulosa and theca cells that can modulate follicular development and steroidogenesis (Roche, 1996). Ginsburg et al. (1990) demonstrated that primordial germ cells in the mouse appear by day 7 post coitum in the extra embryonic mesoderm posterior to the primitive streak. These cells migrate to the gonad by day 12 post coitum. Transforming growth factor-B, (TGF-R,) produced by the gonad appears to be one of the factors responsible for the chemotropic attraction of primordial germ cells (Godin and Wylie, 1991). According to Hilscher (1991), female germ cells undergo only one proliferative wave of oogonial divisions, whereas gametogenesis in the male involves two proliferative waves. In the bovine ovary, onset of meiosis occurs at approximately day 70 of embryonic life (Erickson, 1966). However, meiosis is not completed in the embryonic ovary, and germ cells are arrested in the diplotene stage of prophase I. Conversion of oogonia into primary oocytes depends on contact with cells derived from the rete ovarii. The first germ cells to enter meiosis

PAGE 15

6 are located at the inner border of the ovary establishing a two-way communication between germ and somatic cells. The first follicles are formed around day 70 of pregnancy in the ewe and day 130 of pregnancy in the cow (Mariana et al., 1991). The majority of mammals restrict oogonial proliferation to the prenatal period of development (e.g., pigs and ruminants) or to shortly after birth during the early neonatal period (e.g., rodents and rabbits; Van den Hurketal., 1997). Oogoniaare transformed into primary oocytes characterized by a prolonged meiotic prophase and surrounded by a layer of flattened pregranulosa cells. When a full layer of cuboidal granulosa cells is acquired around the oocyte, the follicle becomes an intermediary and then a primary follicle. These primordial, intermediary and primary follicles constitute the resting stockpile of nongrowing follicles located at the periphery of the ovarian cortex and form the bulk of follicles contained in the ovary (Driancourt et al., 1993b). These follicles are progressively depleted during the reproductive life span. Initiation of ovarian function in the cow occurs during fetal development. Primordial follicles start to grow in response to an unknown trigger (Webb et al., 1 992) when they continue development until ovulation or atresia. The primordial follicle possesses an oocyte surrounded by a single granulosa layer consisting of 14-29 flattened granulosa cells in cattle, which is in turn encompassed by a basement membrane (Van den Hurk et al., 1997). It has been reported that initiation of follicle growth is not dependent on gonadotropins. It has been estimated that it takes approximately 6 months for a primordial follicle to develop into a large dominant follicle in sheep (Cahill and Mauleon, 1980). In the cow, the estimated

PAGE 16

7 time for a follicle to grow from 0.13 to 8.56 mm is approximately 41.5 days, the equivalent of two estrous cycle (Lussier et al., 1987). One of the most critical steps in folliculogenesis is the transformation of primordial follicles into primary follicles. The growth of a follicle appears to begin with enlargement of the primary oocyte, the proliferation of surrounding granulosa cells, and the organization of thecal cells external to the basement membrane (Paton and Collins, 1992). The nature of the signal to convert flattened pregranulosa cells into a cuboidal epithelium is not clear. However, the oocyte provides cues to this conversion (Greenwald and Roy, 1994). By a series of mitotic divisions, the unilaminar primary follicle is converted into a multilayer preantra! secondary follicle. Each primordial follicle contains a single layer of pregranulosa cells resting on a basal lamina and radially arranged around an oocyte which is arrested in diplotene stage of meiosis (Paton and Collins, 1992). Interstitial cells surround the follicle in immediate proximity to the basal laminae and start to differentiate into theca cells when follicular growth starts. During the early stage of a secondary follicle, connective tissue fibers are arranged parallel to the basement membrane underneath the granulosa to form a thecal layer. At the same time , a glycoprotein coat, the zona pellucida, is formed between the growing oocyte and the innermost granulosa layer. At the end of this stage, hormone producing large epithelioid cells and a capillary network additionally constitute the theca (Van den Hurk et al., 1997). With the appearance of an antral cavity, the secondary follicle is converted into a tertiary follicle. It is frequently stated that gonadotropic hormones are not necessary for early growth of follicles and that FSH and LH

PAGE 17

8 become indispensable for further development only at the transformation of the secondary to the tertiary follicle. The relative constant number of preantral follicles throughout the estrous cycle of the mouse, is evidence that cyclic changes in gonadotropins do not affect the pool of preantral follicles. However, in the rat, rhesus monkey, and hamster, the number and responsiveness of preantral follicles to gonadotropins do change during the estrous cycle. In the human ovary, longterm follicular growth is severely impaired in the absence of gonadotropins (Greenwald and Roy, 1 994). In the hypophysectomized ewe follicular development is arrested at a stage between 2 to 3 mm in diameter (McNatty et al., 1990). Schematically, antral follicular development involves two phases. In the first phase, early antral follicles grow slowly (up to 2 mm in sheep and 3-4 mm in the cow) and follicular growth rate is related to the proliferation rate of granulosa cells, and is not really dependent on gonadotropin supply. The second phase or growth is dependent on gonadotropin secretion. This phase is characterized by rapid follicular growth, mainly due to enlargement of the antrum, and corresponds to terminal development of antral follicles up to the preovulatory stage. During this phase the steroidogenic capacity and the responsiveness of granulosa cells to FSH and LH increases (Monniauxetal., 1997). It has been shown (Adams etal., 1992b) in domestic ruminants that terminal follicular growth is associated with fluctuations of FSH concentrations, which suggests a stimulating effect of FSH on emergence of each follicular wave. Goodman and Hodgen (1983) suggested the terms of recruitment, selection and dominance to describe the development of antral follicles. Recruitment is a

PAGE 18

9 gonadotropin-dependent event during which a group of follicles gain the ability to respond to gonadotropin and develop dependance on them for continued growth. Selection is a process whereby only a few of the recruited follicles are selected to escape atresia and survive to ovulate. In cattle, selection is defined as the time when an estrogen-active follicle promotes its growth and inhibits the growth of other follicles (Sunderland et al., 1994). Factors other than size and Ej concentrations, such as P4 may be important for establishing which follicle becomes dominant during the selection phase (Sunderland etal., 1994). Xuetal. (1995b) suggest that selection of the dominant follicle may be a passive process in which the first follicle that acquires LH receptors in its granulosa cells is selected to become the dominant follicle. The acquisition of LH receptors in granulosa cells will enable these follicles to respond to LH in addition to FSH, whose concentrations in blood have declined to basal levels at the time of selection (Adams et al., 1992b). Dominance is the mechanism that an ovulatory follicle uses to escape atresia and inhibits recruitment of a new cohort of follicles. Expression of LH receptor in granulosa cells is associated with dominance (Xu et al., 1995a; Yuan et al., 1998). Ovarian Follicular Dynamics During the Bovine Estrous Cycle Antral follicle development was considered originally to be a continuous state of turnover without a distinct pattern of follicular growth, regression and atresia (Marion et al., 1968). A classical study of Rajakoski (1960) along with studies by Matton et al. (1 981 ) indicated that at least two periods of turnover of antral follicles occur during the estrous cycle of the cow. Later during the 1980's daily ultrasound

PAGE 19

10 scanning of ovaries confirmed that generally two, but sometimes only one nonovulatory follicle wave occurs in the luteal phase of the cycle before final development of the ovulatory dominant follicle after luteal regression (Pierson and Ginther, 1984; Fortune et al., 1988; Savio et al., 1988; Sirois and Fortune, 1988; Gintheretal., 1989a; Lucy etal., 1992; Sunderland etal., 1994; Roche, 1996). Unlike primates (Goodman and Hodgen, 1983), non-ovulatory follicles in cattle develop during the earlyand mid-luteal phase of the estrous cycle (Rajakoski, 1960; Ireland, 1987). This is reflected by the increase in E2 concentrations in blood a few days after ovulation and during the mid-luteal phase. Each follicular wave is characterized by a simultaneous emergence of a group of 5 to 7 follicles (> 5 mm in diameter) from the pool of small follicles. One from this group rapidly emerges and grows larger than the others in the cohort and is considered the "dominant" (Fortune, 1993), while the others become atretic and regress. The dominant follicle normally reaches a maximum size of 10 to 15 mm in diameter and remains dominant for a period of 5 to 7 days, until it becomes atretic and regresses in size. The regressing dominant follicle is replaced with a new dominant follicle grown from the next wave of follicles. If luteal regression occurs during the growth phase or early period of dominance, then the dominant follicle, free from the restrictive hormonal milieu imposed by P4 secretion from the corpus luteum upon the hypothalamus and pituitary gland, will continue to develop to preovulatory size and trigger the hormonal cascade leading to ovulation (Webb etal., 1992).

PAGE 20

11 Follicle growth is a process during which the follicle progressively acquires a number of properties, each of which is an essential prerequisite for further development. Failure to acquire these properties at the correct time and in an exact sequence will lead to failure of the process and to deterioration of the follicle through atresia (Campbell et al., 1995). Antrum formation occurs at a follicular diameter (fixed ovaries) of 0.14-0.28 mm in cattle, and it takes 40 days for follicles to reach ovulatory size (Lussier et al., 1987). In cattle there is a marked hierarchy in the antral follicle population with a large number (20-30 follicles of 3-4 mm in diameter) of gonadotropin-responsive follicles (follicles that can respond to gonadotropin stimulation, but do not need gonadotropins), a few (1-4 follicles of > 4-5 mm) gonadotropin-dependent follicles (follicles that need gonadotropins to further develop) and one ovulatory follicle (Campbell et al., 1995). Follicular dynamics can be defined as the process of continual growth and regression of antral follicles that leads to the development of the preovulatory follicle (Lucy et al., 1992). Ireland (1987) hypothesized that the turnover of dominant follicles during the estrous cycle is regulated by the differential response of selected and unselected follicles in the cohort to alterations in patterns of secretion of gonadotropins, which in turn, result in a differential production of intrafollicular stimulatory or inhibitory factors that control selection, dominance and atresia. Analysis of patterns of development of large follicles in mammalian species shows that large follicles do not develop at random, but their development occurs only during particular reproductive states and/or during particular times of the reproductive cycle (Fortune, 1994). This pattern of follicle development is

PAGE 21

12 associated with follicular changes in expression of mRNA encoding gonadotropin receptors (Xu etal., 1995a) and steroidogenic enzymes (Xu etal., 1995b) that allow selected follicles, when exposed to the requisite hormonal environment, to develop and ovulate in response to the preovulatory gonadotropin surge. Endocrine signals such as gonadotropins, inhibin, and steroids, as well as locally produced growth factors, such as insulin-like growth factor I (IGF-I), transforming growth factor a (TGF-a), transforming growth factor B (TGF-B), epidermal grov\/th factor (EOF), and other peptide hormones such as activin, and follistatin, are responsible for the control and coordination of these process (Armstrong and Webb, 1997). Early during the estrous cycle a cohort of follicles is recruited out of the pool of smaller antral follicles (2 to 4 mm). Recruitment is not a random or isolated phenomenon. Follicles seem to be recruited as groups or cohorts, suggesting that they have received a signal that allows them to continue growth and development rather than regress (Fortune, 1994). The mechanism that controls recruitment of these small follicles and determines which follicles are recruited is unknown. The signal that stimulates recruitment appears to be a slight elevation in plasma FSH (Fortune, 1994). Xu et al. (1995a) hypothesized that changes in expression of mRNA for FSH and LH receptors may be important for recruitment of a cohort of follicles and selection and atresia of the dominant follicle in cattle. However, the steady state of FSH receptor mRNA level in healthy follicles did not correlate with follicle size, nor did the level of FSH receptor mRNA change with stage of the first follicular wave. Adams et al. (1992b) reported that 2 to 4 days before a new wave of follicle development there is an increase in FSH, which suggests that increases

PAGE 22

13 in circulating concentrations of FSH initiate the emergence phase for dominant follicle growth. The number of follicles recruited is usually greater than the typical number of ovulatory follicles for a given species. However, only a species-specific number of ovulatory follicles continues to grow for more than a few days and reaches ovulatory size (Fortune, 1994). After 2 to 4 days of recruitment (days 2 to 4 of the estrous cycle), several medium-sized follicles (6 to 9 mm) can be detected by ultrasonography. This is the phase of selection in which a single follicle emerges from the pool of recruited follicles and continues to grow, whereas other recruited follicles decrease in size (Lucyetal., 1992). Sunderland etal. (1994), using ultrasound analysis and the ratio of £2^4 concentrations in follicular fluid, reported that days 1 to 3 of the estrous cycle are the selection phase for development of the early diestrous first wave dominant nonovulatory follicle. Ginther et al. (1996) defined time of deviation (or selection) as the beginning of the greatest difference in growth between the two largest follicles in the ovary. They indicate that the term deviation of follicles is a major event in the selection process, and the terms deviation and selection are synonymous. Moreover, deviation was characterized by a slower rate of growth for 1 to 4 days before the subordinate follicle attained maximum diameter and by the immediate cessation of growth of the largest subordinate follicle (Ginther et al., 1996; 1997). The completion of the selection phase is defined in the cow as the time when an estrogen-active follicle promotes its own growth and inhibits the growth of other follicles (Sunderland et al., 1994). Adams et al. (1993b) showed that the decrease

PAGE 23

14 in circulating concentrations of FSH is an integral component of the selection mechanism. Exogenous FSH given for 2 days at the expected time of selection delayed selection as indicated by a 1.5 day delay in significant divergence of the growth profiles of the dominant and first subordinate follicles. There was greater growth and delayed regression of the first and second subordinate follicles. The same treatment regimen given after selection did not alter follicular development and atresia. However, it is not known how the FSH decline exerts its effect on selection. Perhaps selection involves competition among the cohort of follicles for utilization of FSH whereby the most successful follicle becomes dominant. This competition becomes acute when FSH levels decline. Results from Xu et al. (1995a; 1995b) suggest that selection of the dominant follicle may be a passive process in which the first follicle that acquires LH receptors in its granulosa cells is selected to become the dominant follicle. The acquisition of LH receptors in granulosa cells will enable these follicles to respond to LH in addition to FSH. Indeed concentrations of FSH in blood have declined to basal levels at the time of selection (Adams et al., 1992b). On approximately day 5, growth of usually only one follicle is sustained while growth rate of other follicles declines. This first wave dominant follicle remains dominant from day 5 to ^ day 9 (Driancourt et al., 1991a), and during this period of follicular dominance no new follicles > 5 mm are detected within the ovaries (Lucy et al., 1992; Fortune, 1993). Therefore, a dominant follicle is defined as a large ovarian follicle (> 10 mm) that is recruited and selected during a follicular wave, and regulates growth of other follicles on the ovary (Lucy et al., 1992). The dominant

PAGE 24

15 follicle would somehow cause demise of follicles of the same cohort and would also suppress a new wave of follicular development. At present little is known about the mechanisms by which a follicle attains and maintains dominance (Fortune, 1993). Perhaps a follicle is selected for dominance because it is in the right stage of development and is better able to respond to the slight elevation of and the subsequent decline in FSH to continue its growth (Fortune, 1 994). The ability of the dominant follicle to continue growth and development in an environment of lower levels of FSH may be due to increased blood flow and/or to the acquisition of LH receptors by the granulosa cells (Zeleznick, 1993). Xu et al. (1995b) showed that theca interna cells of dominant follicles collected on day 4 after initiation of the first follicular wave had the highest levels of mRNAs for P450scc and P450c1 7, 20 lyase, which ensure that the theca interna cells of these follicles are capable of producing large amounts of androgen substrates for E2-I73 biosynthesis in the granulosa cells. The acquisition of LH receptors in granulosa cells may be critical to the establishment and maintenance of follicular dominance, whereas FSH receptors may only play a permissive role (Xu et al., 1 995a). Large antral follicles can transfer their gonadotropic requirements from FSH to LH. This transition in gonadotropic requirement from FSH to LH is the probable mechanism whereby the preovulatory follicle can withstand the fall in FSH that occurs at the onset of the follicular phase following luteal regression (Campbell et al., 1995). The follicle selected for dominance from each wave not only continues to grow, but also differentiates functionally in ways that prepare it for ovulation. The secretion of increased quantities of E2 by the selected follicle appears to be of

PAGE 25

16 primary importance and sets it apart from its sister subordinate follicles. Feedback regulators, such as and inhibin produced by the dominant follicle (or perhaps by the whole cohort of follicles during the first few days after recruitment), cause a decrease in FSH concentrations that will not even support the further growth of subordinate follicles (Fortune, 1994). It is important to distinguish between morphological and functional dominance. Morphological dominance is characterized by size of the follicle, with the largest follicle present on a pair of ovaries being defined as morphologically dominant. Functional dominance appears to have two aspects: ability of the dominant follicle to inhibit the growth of smaller follicles and the capability to ovulate under appropriate hormonal conditions (Fortune, 1993). The largest follicle of the first and second wave of cows with three waves is by definition morphologically dominant (Fortune et al., 1991; Fortune, 1993). The first wave dominant follicle at day 12, despite being the largest follicle, had lost its functional dominance as early as day 10 based on emergence of a new wave of follicular development (Sunderland et al., 1994). These results and others (Savio et al., 1990a) show that morphological dominance of a follicle persists much longer than functional dominance. Driancourt et al. (1991) reported that dominance is characterized by the large difference in diameter between the largest and the second largest follicles and the decrease in the number of follicles smaller than 8 mm. Two hypotheses have been postulated to explain how the dominant follicle exerts dominance. One hypothesis states that the dominant follicle secretes a product that directly impairs further growth and development of subordinate follicles.

PAGE 26

17 In monotocous species, like cattle, such a factor would have to be endocrine in nature, since it would induce regression of subordinate follicles on both ovaries. The other hypothesis states that the dominant follicle could cause the regression of subordinate follicles indirectly, via negative feedback mechanisms in which and inhibin would cause a decrease in FSH concentrations that would not support further growth of subordinate follicles (Fortune, 1994). A third hypothesis, which explains dominance from an intraovarian point of view, states that production of local ovarian factors such as the insulin-like growth factor (IGF) system, components of the transforming growth factor P superfamily, fibroblast growth factor, and the epidermal growth factor/transforming growth factor a family can inhibit the development of subordinate follicles directly (Campbell et al., 1995; Armstrong and Webb, 1997). Sunderland et al. (1994) indicated that the period of days 1 to 3 of the estrous cycle is the selection phase for development of the early diestrous dominant nonovulatory follicle; whereas days 10 to 12 is boyh the selection phase for development of the next dominant follicle and the period when the first dominant follicle ceases to function or loses its dominance, becomes estrogen-inactive and begins to undergo atresia. In two-wave cycles, maturation of the second dominant follicle coincides with spontaneous regression of the corpus luteum, and the follicle ovulates after luteolysis. Alternatively, the second wave dominant follicle may become atretic and a third follicular wave will be initiated (Lucy et al., 1992) and becomes the ovulatory follicle.

PAGE 27

18 Sirois and Fortune (1988) reported that follicular waves during the estrous cycle occur at approximately 7 day intervals. For three-wave cycles the first, second and third waves start on days 2, 9 and 16, respectively, for two-wave cycles the second wave started about 2 days later than the average for animals with three waves (day 11; Fortune et al., 1991). Waves of follicular development in cattle occur regularly during both the estrous cycle and pregnancy. During the estrous cycle, the preovulatory gonadotropin surge perturbs the pattern of regular waves by triggering an abrupt ovulation of the follicle that is functionally dominant at the time of estrus. More than 99% of ovarian follicles undergo the degenerative process of atresia during reproductive life (Hsueh et al., 1994). If the ovulatory signal is absent, the mature follicle undergoes degeneration. It has been shown that the dominant follicle of the first wave, which normally undergoes atresia, can ovulate provided that the ovulatory stimulus is given before it reaches the advanced regression stage of development (Savio et al., 1990a; Sirois and Fortune, 1990). In antral bovine follicles > 1 mm in diameter, the earliest and most prominent feature of atresia is death of granulosa cells (presence of pycnotic nuclei) that leads to almost total destruction of the granulosa cell layer lining the inner follicle wall (Rajakoski, 1960). Atresia of the dominant nonovulatory follicle is characterized by a significant decrease in the number of granulosa cells, a decrease in both LH and FSH receptors (Ireland and Roche, 1983), and a diminished capacity to produce between days 7 and 1 3 of the estrous cycle (Badinga et al., 1 992; de la Sota, 1 995). The factors involved in regulation of follicular atresia are not clear, but it has been

PAGE 28

19 demonstrated that decreasing the LH pulse frequency to a luteal phase secretory pattern results in faster atresia of the dominant follicle (Sirois and Fortune, 1990). Thus the dominant follicle may avoid atresia when exposed to a higher LH pulse frequency during the follicular phase. Induction of a low progesterone environment (< 2 ng/ml) results in increased growth or size of the dominant follicle, higher plasma concentrations of Ej and a greater lifespan or persistence (Savio et al., 1 993a). In addition to sustained follicular growth, functional dominance also was observed as there was a suppression in the number of other follicles > 4 mm in diameter (Savio et al., 1 993a). Stock and Fortune (1 993) reported that slight increases in LH pulse frequency promoted prolonged follicular growth and dominance associated with increased plasma E2. They suggested that the demise of nonovulatory dominant follicles during normal estrous cycles occurs through feedback effects of luteal progesterone, which maintains a low LH pulse frequency and Ej production. Based mainly on morphological criteria, atresia of antral follicles can be divided into several stages (Hsueh et al., 1994). Stage I is characterized by a small number of granulosa cells (< 1 0%) with pyknotic nuclei close to the follicular cavity; however, some of the granulosa cells are still undergoing mitosis. Stage II shows many pyknotic granulosa cells (10-30%), few cells in mitosis, and cell debris in the follicular cavity. The basement membrane loses its integrity, and there is a leukocytic infiltration into granulosa cell layers. In the oocyte, meiotic-like changes are evident. Stage III is characterized by a reduction in granulosa cell number, an absence of mitosis, and collapse of the follicle. The oocyte also undergoes germinal vesical breakdown (Hsueh et al., 1994).

PAGE 29

20 Fate of the thecal cell layer during atresia varies among different species. In the human, rat, and rabbit, theca cells undergo extensive hypertrophy during follicular atresia. In contrast, hamster follicles do not exhibit marked morphological changes despite sharply reduced follicular vascularity and total collapse of the granulosa layer. In the sheep, theca cells undergo nuclear condensation and degeneration similar to those observed in granulosa cells (Hsueh et al., 1994). Biochemical changes also occur during atresia. These changes include reduced DNA synthesis of granulosa cells, suppressed expression of connexin, a gap junction protein, decreased gonadotropin binding, decreased expression of mRNAs for aromatase and gonadotropins receptors, and decreased synthesis concomitant with increased progesterone production. The latter change appears to be due to both a decrease in C^y 20-lyase activity, that leads to a decrease in androgen substrate for granulosa cell aromatization, and a loss of aromatase activity (Hsueh et al., 1994). During atresia, there is an increase in expression of several genes, including the IGF-binding proteins (IGFBP), which is related to a deprivation of endogenous IGF-I that is essential for survival of follicle cells (Hsueh et al., 1994). Recent biochemical evidence has demonstrated apoptosis of granulosa cells during follicular atresia in bovine ovaries (Jolly et al., 1 994). These authors suggested that apoptotic death of granulosa cells in the cow may also occur in healthy follicles during the luteal phase of the estrous cycle and/or occur early in the atresia process before other morphological or biochemical signs of degeneration or dysfunction are

PAGE 30

21 evident. Thus a variable incidence of apoptotic cell death may occur at different stages of follicular development. Structure and Function of the Antral Follicle The primary function of the mammalian ovarian follicle is the release of an oocyte capable of being fertilized by sperm. This involves growth and maturation of the follicle as well as enlargement, ovulation, and meiotic division of the egg (Richards, 1980). The identification of the ovarian follicle occurred in 1672 by Regnier de Graaf, and mature preovulatory follicles are named Graafian follicles. It was not until 1827 that Karl Ernst von Baer discovered that the mammalian egg was a tiny part of the entire follicular structure. In the early 1900's the gonads as well as the pituitary were recognized as organs of internal secretion (Richards, 1980). A single follicle consists of a diplotene oocyte and associated somatic cells surrounded with a basal lamina. Small follicles have a single layer of follicular cells and form the largest number in the ovary. The ovarian follicle consists of outer layers of theca cells which encircle inner layers of granulosa cells, which are separated from blood vessels and theca cells by a basement membrane lining the follicle. Inner layers of granulosa cells surrounding the oocyte are designated as the oocyte-cumulus cell complex (Amsterdam et al., 1 989). Follicular growth ends with ovulation or atresia. The growth of the follicle is ensured by the growth of the oocyte, follicular cell proliferation, and enlargement of the antrum (Mariana et al., 1991). Follicles appear to begin growth under all physiological conditions, including

PAGE 31

22 pregnancy, ovulation, or periods of anovulation (Richards, 1980; Amsterdam and Rotmensch, 1987). The number of follicles beginning to grow each day changes throughout life and appears related primarily to the number of follicles in the nongrowing or resting pool. The larger the pool, the greater the number of follicles beginning to grow. With age, as follicles continuously leave the pool of nongrowing follicles, the number of follicles in the resting pool becomes drastically reduced and ultimately is exhausted. This reduction is associated with a decrease in the number of follicles that begin to grow and eventually ovulate (Richards, 1980). The first sign of growth is the resumption of cell proliferation by the squamous granulosa cells, an increase in the size of the oocyte and a change in shape of the granulosa cells (Hirshfield, 1991). In the adult ovary, folliculogenesis starts when follicles leave the pool of resting follicles to enter the growth phase. From there, the early growing follicle undergoes a developmental process including a dramatic course ofcellular proliferation and differentiation (Gougeon, 1996). The oocyte and the follicle grow in a biphasic manner. During the first phase, which ends shortly before antrum formation, growth of oocyte and granulosa cells is linearly and positively correlated. The second phase consists of enhanced mitotic activities of granulosa and thecal cells, and an increase in volume of follicular fluid (Greenwald and Roy, 1994). The oocyte enlarges rapidly during early follicular growth and reaches full size early in the developmental process (Hirshfield, 1991). Later in development, the growth of the oocyte is slow compared to growth of the follicle. Growth of the oocyte involves accumulation of proteins either through direct synthesis within the oocyte or by transfer from the antrum. During the active phase

PAGE 32

23 of oocyte growth, content of proteins is increased 100-fold and RNA content enhanced 20-fold. RNA synthesis is increased in the nucleus of the growing oocyte compared to that found in the primordial oocyte. Synthetic activity decreases steadily in oocytes from follicles with an antrum (Mariana et al., 1991). The follicle acquires several distinctive morphological features during growth: theca interna (steroidogenic cells) and theca externa (connective tissue cells) which form the outer layers of the follicle, a fluid-filled antral cavity, a capillary network, and the thick acellularzona pellucida surrounding the oocyte (Hirshfield, 1991). As the follicle enters its final phase of growth, fluid-filled spaces appear between the granulosa cells which soon coalesce into a large single fluid-filled antral cavity. The accumulating fluid appears to be formed by filtration of thecal blood. The composition of follicular fluid differs considerably from plasma, with lower glucose and lipid concentrations in follicular fluid. There are also differences in amino acid concentrations (Hirshfield, 1991). The Granulosa Cell Compartment The ovarian follicle has the unique capacity to change in structure and function to influence the cyclic processes of female reproduction. A finely tuned process of differentiation occurs in all constituents of the follicle during folliculogenesis. Granulosa cells play an obligatory role in creating the conditions necessary to resume oogenesis, ovulation, fertilization and implantation (Paton and Collins, 1992). Granulosa cells display a high degree of structural change and play a key role in the functional maturation of the entire follicle.

PAGE 33

24 The early development of the female gonad is characterized by the migration of a mitotically active cohort of germ cells from the region of the embryonic hindgut and endoderm of the yolk sac to the gonadal ridge. As germ cells enter meiosis, granulosa cells seem to arise from the rete ovarii or from mesonephric tubules or mesenchymal cells in the gonadal aniage and they organize around the primary oocytes (Amsterdam and Rotmensch, 1987). The development of the steroidogenic potential in differentiating granulosa cells consists of early and late events (Amsterdam et al., 1989). The early events are related to mobilization of cholesterol, reorganization of endosomes, lysosomes, lipid droplets and particularly mitochondria and smooth endoplasmic reticulum, which are the carriers of membrane-bound enzymes converting cholesterol to a variety of steroid hormones. Late events consist of stable changes in cell shape, intercellular communication through gap junction formation and development of organelles associated with steroidogenesis (Amsterdam et al., 1989). Granulosa cells from the various layers within the same follicle display morphological heterogeneity suggesting the existence of a differentiation gradient in the follicle. Cells adjacent to the follicular basement membrane contain a higher density of LH receptors than cells located in the inner layers of the pre-ovulatory follicle (Amsterdam et al., 1989). The granulosa cell layer surrounding the oocyte, called the corona radiata, produces elongated cytoplasmic processes which traverse the zona pellucida towards the oocyte and form gap junctions at the area of contact. Due to the nature of a nonvascular environment of granulosa cells, they need a way of intercellular communication between neighboring cells. In addition

PAGE 34

25 to a complex paracrine system of membrane-soluble hormones, there is also a direct cell-to-cell communication via gap junctions (Amsterdam and Rotmensch, 1987). Gap junctions are cellular structures between adjacent cells, wherein the apposed cellular membranes are separated by an apparent gap of approximately 3 nm. Gap junctions are formed by connexins. They allow communication or electrical coupling between adjacent cells that can be open or closed. Gap junctions are composed of two symmetrical structures (connexons) that create an intracellular channel that allows passage of ions and small molecules from cell to cell (Grazul-Bilska et al., 1997). Also adherence junctions have been described in granulosa cells. They are characterized by association of filamentous dense material on the intracellular aspect of the plasma membrane (Amsterdam and Rotmensch, 1 987). Gap junctions allow ions and small, water soluble molecules to pass directly from the cytoplasm of one cell to the cytoplasm of another, thereby coupling the cells both electrically and metabolically (Amsterdam et al. , 1 989; Paton and Collins, 1992). Gap junctions are absent in the primordial follicle, where desmosomes and adherence junctions are the dominant structures comprising the junctional complex (Amsterdam and Rotmensch, 1987; Amsterdam et al., 1989). Gap junctions appear during differentiation of the antral follicle, and parallel the onset of gonadotrophic action on the follicle (Amsterdam and Rotmensch, 1987; Amsterdam et al., 1989; Paton and Collins, 1992). Molecules of 1200 daltons and smaller (e.g., inorganic ions, amino acids, nucleotides, and sugars) can pass through the junctional channels, and move from cell to cell. For example, cyclic AMP (cAMP) has been identified as a possible mediator of biological function

PAGE 35

26 through gap junctions. The functional significance of gap junctions between granulosa cells may lie in propagating cAMP-mediated events such as the response to FSH or LH, thereby allowing the signal to reach the interior of the avascular preovulatory follicle (Amsterdam and Rotmensch, 1987). Changes in the degree of dispersion of granulosa cells during follicular maturation seem to be related to quantitative changes in intercellular junctions. Biochemical differentiation of granulosa cells is associated with gap junction formation in response to FSH stimulation. Morphological studies suggest that cAMP might be associated with changes in cell shape and organization of the cytoskeleton (Ben-Ze'ev and Amsterdam, 1987). Granulosa cells display a high degree of structural change during the differentiation process. There are alterations in cellular shape and size, in plasma membranes, and in internal membranes of the endoplasmic reticulum, mitochondrial cristae, intercellular junctions, and cytoskeleton. Granulosa cells are flat and epithelioid in the earliest stage of development. When proliferation starts, the flattened layer of granulosa cells becomes a multilayered stratified epithelium consisting of cuboidal cells (Paton and Collins, 1 992). There is also a change in the type of intracellular organelles. Immature granulosa cells contain mainly a rough endoplasmic reticulum, which reflects a high level of protein synthesis required for growth and proliferation. Smooth endoplasmic reticulum increases its development when steroidogenic potential increases (Paton and Collins, 1 992). Complex tubular mitochondrial cristae is typical of FSH-stimulated cells, whereas a lamellar type is observed in immature cells (Amsterdam and Rotmensch, 1987).

PAGE 36

27 The cytoskeleton mediates many of the structural alterations that coincide with differentiation-dependent changes in the metabolic activity of granulosa cells (Paton and Collins, 1992). Cytoskeletal elements are believed to be involved in aggregation and internalization of ligand-bound gonadotropin receptors (Amsterdam and Rotmensch, 1987). These elements provide the machinery for moving organelles from one place to another when granulosa cells are exposed to gonadotropins and movement of substrate among organelles also needs to be facilitated to enhance steroid synthesis (Paton and Collins, 1992). The regulation of granulosa cell differentiation at the level of the cytoskeleton seems likely to prove as complex as other intraovarian regulatory mechanisms. One important mechanism by which organization of cytoskeleton can be regulated is by the activity of transmembrane proteins which link cytoskeletal proteins to the extracellular matrix. The extracellular matrix is an important regulator of differentiation of granulosa cells (Paton and Collins, 1992). Granulosa cells are circumscribed by two layers of insoluble extracellular material. The peripheral cells border the basal membrane of the follicle and corona radiata cells attach to the zona pellucida (Amsterdam and Rotmensch, 1 987). Zona pellucida glycoproteins are synthesized by the growing oocyte (Driancourt et al., 1993b). It is unclear to what extent granulosa cells contribute to the formation of these layers; however, their interaction with the extracellular matrix components appears to affect differentiation events resulting in varying biochemical and structural characteristics.

PAGE 37

28 A gradient of differentiation exists in granulosa cells in relation to their distance from the basal membrane of the follicle. Subpopulations of granulosa cells exist in ovarian follicles (Amsterdam et al., 1989). In the rat, there exists evidence of morphological and functional differences between antral cells in close proximity to the antral cavity and mural cells in close proximity to the basement membrane. They differ relative to number of gap junctions (Wiesen and Midgley, 1993), FSH receptors (Monniaux and De Reviers, 1989) and in the mitotic index (Hirshfield, 1992). Likewise, basal steroid production is lower in mural than in antral bovine granulosa cells cultured in the absence of FSH (Roberts and Echternkamp, 1994). Although morphological similarities were noticed between bovine antral and mural granulosa cells in response to FSH stimulation, the steroidogenic responses to FSH of antral granulosa cells in terms of Eg, P4, and testosterone production are considerably higher than those of mural granulosa cells (Rouillier et al., 1996). There is a higher content of LH receptors, mitochondrial P-450 side chain cleavage enzyme and 3(i-hydroxysteroid dehydrogenase in the peripheral granulosa cell layers (in apposition to basal membrane) than in the inner layers. Also cells close to the basal membrane exhibit a decreased proliferative activity, which suggests an active role for the basement membrane in the induction and maintenance of granulosa cell differentiation (Amsterdam and Rotmensch, 1987). Extracellular matrix is associated with changes in cell shape and contact, which are accompanied by simultaneous alterations in organization and expression of cytoskeletal proteins. Gonadotropins and extracellular matrix may induce granulosa cell differentiation in a coordinated fashion by their effect on cell shape,

PAGE 38

29 cell contact, intercellular communication, and expression of cytoskeletal elements (Amsterdam and Rotmensch, 1987). Fibronectin is present in the follicle as a component of the basal lamina and as a soluble fraction of follicular fluid which increases with increased follicle size (Luck, 1994). Fibronectin production and steroidogenesis represent two distinct states of granulosa cell differentiation, and it is proposed that fibronectin production is high in those tissues where extracellular matrix support is needed, like primordial follicles, atretic follicles, and early corpus luteum (Luck, 1994). In bovine granulosa cells, fibronectin secretion is greater in confluent cultures of cells from large follicles (1 5 mm), and is lower in sparse cultures of cells from small follicles (4-5 mm; Savion and Gospodarowicz, 1980). The Theca Cell Compartment Theca cells are the other cell type involved in follicular steroidogenesis. They are LH-responsive secretory cells which comprise the follicular envelope. The stem cell origin of theca cells is less clear than the origin of granulosa cells (Hirshfield, 1991). Theca cells appear to share a common ancestor with fibroblast and the stromal/connective tissue elements of the ovary (Gore-Langton and Armstrong, 1988). However, it has been suggested that granulosa and theca cells share a common ancestor (Hirshfield, 1991). Since the theca layer is not present in primary follicles but differentiates as follicles grow and mature, it is evident that theca cells arise continually throughout reproductive life. The aggregation and ultimate differentiation of the theca folliculi is assumed to occur under the influence of growth

PAGE 39

30 factors emanating from the oocyte-granulosa complex, but there is no direct evidence to substantiate a casual relationship (Greenwald and Roy, 1994). The theca interna of sheep follicles, less than 3 mm in diameter, already consists of 8 to 12 layers of flattened cells and capillaries; 20% of the cells have the ultrastructural features of steroidogenically active cells; others are fibroblasts, and the majority are undifferentiated cells. In follicles 3-6 mm, approximately 40% of the theca interna possess tubular endoplasmic reticulum. In all species, the theca interna is separated from the granulosa cells by the basal lamina or basement membrane (Greenwald and Roy, 1994). The time of appearance of a well-differentiated theca interna relative to development of granulosa cells varies from species to species. For example in sheep, differentiation of the theca appears in follicles of 2-3 mm in diameter with maximal differentiation occurring in late estrus, when numerous lipid droplets accumulate in theca interna cells. Two cell populations emerge in the bovine theca interna at 3 to 4 days before ovulation. One population consists of large epithelial cells with round nuclei increasing in area and another group of fibroblast-like cells in which the nuclear area does not increase. In large (10-12 mm) porcine follicles, theca interna cells form a discrete layer, contain many more lipid droplets than granulosa cells, and have a very high 36-hydroxysteroid dehydrogenase activity (Greenwald and Roy, 1994). The theca interna layer develops a vascular supply concomitant to its differentiation as a steroidogenic tissue (Greenwald and Roy, 1994). In the rat, primordial follicles do not have an independent blood supply until multilaminar

PAGE 40

31 granulosa and theca cell layers have formed. The vascular bed develops in the inner portion of the theca interna adjacent to the basal membrane and membrana granulosa. In turn, the theca interna vasculature is linked to an outer series of arterioles and venules in the theca externa layer (Greenwald and Roy, 1994). It is possible that follicular growth is angiogenisis-dependent. The increase in granulosal and thecal cells must be accompanied and/or preceded by an increase in new capillaries that grow toward and within the theca layer of growing follicles (Greenwald and Roy, 1994). Two-Cell. Two-Gonadotropin Theory In all mammalian species there are two types of cells involved in follicular steroidogenesis: LH-responsive secretory cells comprised of the theca interna cells of the follicular envelope and interstitial cells of the ovarian stroma, and FSHresponsive granulosa cells, that acquire the ability to respond to LH during the later stage of follicular maturation. These two cell types fulfill distinct roles in the steroidogenic process by virtue of their different regulatory hormones and their dissimilar expression of steroidogenic enzymes (Gore-Langton and Armstrong, 1988). The studies of Faick (1959; cited by Richards, 1980) provided the first evidence that ovarian estrogen biosynthesis required the interplay of at least two ovarian cell types. He isolated granulosa, theca, interstitial, and luteal cells and autotransplanted these cell types either alone or in combination to the eye chamber of ovariectomized female rats. Results from his experiments indicated that estrogen was produced only when theca or interstitial cells were cotransplanted with

PAGE 41

32 granulosa cells or luteal cells or when intact follicles were autotransplanted. (Richards, 1980). The steroidogenic pathways present in granulosa and theca cells of developing follicles are responsible for the local as well as the systemic sex-steroid milieu. Three major classes of sex-steroids are produced by mammalian ovarian follicles: progestins, androgens, and estrogens, which enter the systemic circulation and exert important actions both within and outside the hypothalamo-pituitary gonadal axis (Urban and Veldhuis, 1992). Synergistic interactions between granulosa and theca cells have been proposed for some time as a mechanism by which ovarian follicles can produce the three classes of sex-steroids (Hsueh et al., 1984). This two-cell, two-gonadotropin theory of steroidogenesis can be summarized as follows. Luteinizing hormone from the pituitary stimulates androgen production by theca cells (Fortune and Armstrong, 1977), which also appear to possess little or no ability to aromatize androgens to (Fortune and Armstrong, 1 978). The locally produced androgens diffuse through the basement membrane to the granulosa cells. Follicle stimulating hormone stimulates aromatase enzyme activity in granulosa cells to allow conversion of androgens to estrogens (Fortune and Armstrong, 1 977). Granulosa cells seem to be incapable of synthesizing androgens (Fortune and Armstrong, 1977) but produce progesterone in vitro. Granulosa cells appear to lack the 17a-hydroxylase, 17-20 desmolase enzyme necessary to continue the steroidogenic pathway to androgens (Fortune and Quirk, 1988). In

PAGE 42

33 granulosa cells from immature hypophysectomized rats, aromatase enzyme is not present but can be induced by treatment with FSH (Dorrington et al., 1975). Experiments with bovine theca and granulosa cells isolated from proestrous follicles have provided evidence that both theca and granulosa cells are necessary for follicular production in cattle (Fortune and Quirk, 1988). Androstenedione is the primary form of androgen secreted by bovine granulosa cells when cultured in vitro and stimulated by LH (Fortune, 1986). Luteinizing hormone and FSH increase production of P4 by granulosa cells. Nevertheless, granulosa cells do not convert progestins to Ej (Fortune, 1 984). These observations led to the conclusion that the same two-cell, two-gonadotropin theory proposed for rats can be applied to the bovine follicle (Fortune and Quirk, 1988; Figure 2-1: model for regulation of E2 production by bovine preovulatory follicles). Theca cells are stimulated by LH to synthesize androgens, but they are unable to convert androgens to Ej. Granulosa cells cannot synthesize androgens de novo, but they can convert exogenous androgens to Ej. It has been shown that granulosa cells can secrete Ej in a longterm cell culture (Saumande, 1991; Rouillieret al., 1996; Diaz et al., 1997). Fortune (1986) postulated that granulosa cells provide additional pregnenolone precursor (e.g., cholesterol) for thecal androgen production, which in turn enhances the ability of the granulosa cells to synthesize pregnenolone. This implies that cattle preferentially use the pathway to synthesize androgens (Fortune and Quirk, 1988). Estrogens secreted by granulosa cells act as a mitogen and, in conjunction with FSH, induces appearance of LH receptors on granulosa cells. In bovine follicles, Ej exerts a positive feedback to stimulate its own

PAGE 43

34 production and has a negative effect on P4 secretion due to inhibition of the 3Shydroxysteroid dehydrogenase in granulosa cells. Based on these observations, Fortune and Quirk (1988) hypothesized that as the bovine follicle develops capability for E2 production, Ej will inhibit conversion of pregnenolone to P4 in both theca and granulosa cells. This increases the amount of pregnenolone available for conversion to androgens via the pathway in theca cells. Differential inhibitory effects of E2 on P4 versus androstenedione production could provide a mechanism by which Ej initially exerts a positive feedback on its own production to increase androgen synthesis, but eventually inhibits its own production by blocking thecal cell conversion of dehydroepiandrosterone to androstenedione through inhibition of 3Bhydrxoysteroid dehydrogenase (Fortune, 1986). Molecular Events Occurring in Theca and Granulosa Cells Luteinizing hormone acts on theca cells via cAMP to regulate both C^i side change cleavage P-450 multienzyme complex (cholesterol 22-hydroxylase, cholesterol 22a-hydroxylase and C2o.22-lyase), C21 side chain cleavage P-450 (17ahydroxylase, 0,7 20-lyase) activation to increase androstenedione biosynthesis. In granulosa cells, androstenedione (from thecal origin) is converted to testosterone by 17P-hydroxysteorid dehydrogenase increasing the capacity of granulosa cells to synthesize pregnenolone (Fortune, 1986). Estradiol then is synthesized from testosterone by the aromatase P-450 enzyme system. Estradiol, in turn, binds to E2 receptors present in granulosa cells and may regulate gene expression (e.g., inhibin aand Bg-subunit mRNA; Turner et al., 1989). However, FSH must be

PAGE 44

35 present to enhance intrafollicular effects (Richards, 1980; Richards and Hedin, 1988: Figure 2-1). There is substantial evidence that the cAMP/protein kinase A (PKA) pathway is one of the most important regulatory mechanisms in theca cells. Actions of LH on steroidogenesis of theca cells are mediated through the cAMP/PKA signal transduction pathway which stimulates the expression of steroidogenic enzyme genes required for androgen synthesis (P-450 side chain cleavage, 3S-HSD and P450 17a-hydroxylase 17, 20-lyase). Once theca cells have expressed the steroidogenic enzymes, activation of the cAMP/PKA pathway also stimulates the rate-limiting step in steroid hormone biosynthesis which is the transfer of cholesterol across mitochondrial membranes (Magoffin and Erickson, 1994). However, a number of observations has indicated that the activity and amount of of the P-450 side chain cleavage enzyme are unaffected by trophic hormone stimulation (Stocco, 1997). Recently it has been thought that the rate-limiting step in steroidogenesis is the de novo synthesis of a protein whose function is to facilitate the transfer of the substrate cholesterol from the outer mitochondrial membrane to the inner mitochondrial membrane and the P-450 side chain cleavage enzyme. The synthesis of this protein is dependent upon a functional cAMP-dependent protein kinase A signaling pathway that is linked to the steroidogenic potential of the cell.

PAGE 45

36 LH c * LH ATP CfiNP 'fBH 4 — fSH i (HlHilbHJL ^ PFBTeCLOE 4 CjiSccP^so 1.17a-lvixKlase A®-3p-HSD r A*-3P-«D| AMKJBIUCOOE ' ThecaCM * % TBST Cs i mu rc 17p QanJosaCM LH Figure 2-1. Schematic representation of the molecular events coordinating estradiol, FSH, and LH action in theca and granulosa cells of preovulatory follicles (Adapted from Richards and Hedin, 1988; Gore-Langton and Armstrong, 1988; Fortune and Quirk, 1988). The protein was named the Steroidogenic Acute Regulatory (StAR) protein (Stocco, 1 997). Interaction of StAR with the mitochondria may cause the formation of a protein complex consisting of P-450 side chain cleavage and 3S-HSD, enzymes required for the first two steps in steroidogenesis (Stocco, 1997). It appears that cAMP in theca cells increases the content of the regulatory (R) subunit of cAMPdependent protein kinase type II (RII51), P-450 side chain cleavage enzyme, and 17a-hydroxylase 17, 20-lyase.

PAGE 46

37 receptors have little or no effect on proliferation of granulosa cells unless FSH acts via the FSH receptor to increase intracellular concentrations of cAMP (Richards, 1980). On the other hand, is required for FSH to induce LH receptors in granulosa cells. The effects of £3 on granulosa cell function are mediated via the translocation of the cytosol Ej-receptor complex to the nucleus. Binding of the Ejreceptor complex to nuclear acceptor sites results in altered genomic expression, synthesis of new messenger RNA, and a change in any one of a number of the components of the FSH receptor system. It is also possible that Ej enhances the ability of FSH to stimulate the adenylate cyclase system. Estradiol may increase the content of a specific cAMP-binding protein that is distinct from the regulatory subunit of the protein kinase or is the regulatory subunit of a specific protein kinase. Estradiol could also induce the synthesis of a specific protein that when phosphorylated in the presence of active kinase might enter the nucleus to affect gene transcription (Richards, 1980). Gonadotropic hormone control of granulosa cell steroidogenesis occurs primarily by way of cAMP-dependent pathways, but evidence exists that additional intracellular regulatory signals can modify this process. Two such secondmessenger systems involve intracellular hydrolysis of phosphatidylinositol to produce inositol trisphosphate (IP3) and diacylglycerol. Intracellular Ca** is increased in response to IP3 which in turn affects activity of protein kinase C and activation of the calcium-calmodulin system (Urban and Veldhuis, 1992).

PAGE 47

Hormonal Control of Ovarian Follicular Dynamics 38 During each estrous cycle, bovine ovaries synthesize and secrete Ej-1 7(i and P4, which coordinate function of the female reproductive system. Each estrous cycle is comprised of follicular and luteal phases. The follicular phase is characterized by development of the preovulatory follicle and its secretion of Ej, whereas the luteal phase is characterized by secretion of P4 by the corpus luteum, which is formed after ovulation of the preovulatory follicle. At the end of the luteal phase, the corpus luteum regresses by the action of prostaglandin F2o,(PGF2j, and final development of the next preovulatory follicle occurs. Gonadotropin secretion plays a central role in control of the estrous cycle. The developing preovulatory follicle produces a critical level of Ej that stimulates the hypothalamus to increase the frequency and amplitude of gonadotropin-releasing hormone pulses(GnRH). In cattle this is inferred based on an increased frequency and amplitude of LH pulses. Increased LH pulses amplify E2 secretion, complete follicular development, induce estrous behaviour and trigger the preovulatory surge of LH. Ovulation occurs ^ 30 h after the preovulatory LH surge (Chenaultetal., 1976; Robinson and Shelton, 1991; Driancourt et al., 1993a). Steroid Hormones Steroid hormones such as progestins, androgens and estrogens are produced by the ovarian follicle during its development. Each wave of development of a dominant follicle goes through a selection, dominance and atresia or ovulation

PAGE 48

39 phase. During the selection phase, production by each ovary is similar (Ireland and Roche, 1987). At the end of this phase, one follicle becomes larger than all other follicles and is responsible for most of the production, and contributes to a transient increase in concentration of in the peripheral circulation. If luteal regression occurs coincident with the dominance phase in development of a dominant follicle, then the follicle ovulates. If luteal regression does not occur during a dominance phase, the dominant follicle undergoes atresia and a new selection phase begins (Ireland and Roche, 1987). In cows, the preovulatory surge of LH is induced by increasing concentrations of Ej after demise of the corpus luteum (Chenault et al., 1975, 1976; Kesner et al., 1981). Estradiol initiates the preovulatory surge of LH by acting on the hypothalamus to increase secretion of GnRH at the pituitary level, and the sensitivity of the gonadotropes to GnRH. A preovulatory surge of FSH occurs concomitantly with the surge of LH (Walters and Schallenberger, 1 984). The preovulatory follicle is the source of this increase in Ej. Plasma Ej concentrations are high before the LH surge and decline during the surge. Follicular fluid concentrations of androgens and progestins are much lower than E2 before the LH surge. Following the surge, androgens decline in follicular fluid whereas P4 increases. In contrast, atretic follicles differ from healthy follicles in that they have much lower concentrations of E2 than healthy follicles (Fortune and Hansel, 1985; Badinga et al., 1992; de la Sota, 1995). Demise of the first dominant follicle at midcycle is due to the negative feedback effect of progesterone from the corpus luteum on LH secretion (Savio et al., 1993a). In the absence of a normal corpus luteum and in an environment with

PAGE 49

40 low progestin, the first dominant follicle continues to grow and suppresses growth of other follicles for more than 20 days (Savio et al., 1 993a). This sustained growth and functional dominance is due to increased LH pulse frequency. The presence of high levels of P4 during the luteal phase assures that the period of functional dominance of any follicle is limited due to induction of a low LH pulse frequency that is insufficient to maintain follicular function and leads to atresia. In contrast, maintenance of P4 concentrations between 1 and 2 ng/ml prolongs both morphological and functional follicular dominance (Sirois and Fortune, 1 990; Savio et al., 1993b). In cases where the estrous cycle is lengthened by maintaining normal luteal concentrations of P4, either artificially or naturally as during pregnancy (Ginther et al., 1 989c), continuous follicular waves occur. However, occurrence of follicle waves in pregnancy is more prevalent on the ovary contralateral to the uterine horn bearing the conceptus. Gonadotropic Hormones Regulation of steroid production by bovine theca and granulosa cells seems to depend on gonadotropins and on paracrine regulation by ovarian hormones. Theca cells have LH, but not FSH receptors (Ireland and Roche, 1983). Binding of LH to theca cells increases as the preovulatory follicle develops. In contrast, granulosa cells can bind both LH and FSH. As the preovulatory follicle develops, its granulosa cells increase their specific binding of LH, but exhibit decreased binding of FSH (Ireland and Roche, 1983). Both LH and FSH are essential for follicular development and steroidogenesis in vivo (Fortune et al., 1988).

PAGE 50

41 Follicle stimulating hormone is the key hormone stimulating emergence of follicular waves and declines in FSH are associated with selection of a dominant follicle which becomes dependent on LH for its final fate (Roche, 1996). Adams et al. (1992b) concluded that there is a temporal sequence between surges of FSH and subsequent emergence of follicular waves. Time of follicle selection coincides with the first decrease In FSH secretion. Turzillo and Fortune (1990) reported that the secondary FSH surge (e.g., at 24 h after the FSH preovulatory surge) is important for the initiation of follicular development early in the bovine estrous cycle. Adams et al. (1992b) and Sunderland et al. (1994) reported that there is a cyclic pattern of FSH secretion during the estrous cycle of cattle with enhanced FSH secretion responsible for emergence of each follicle wave. Circulating concentrations of FSH decrease once selection is initiated as indicated by the presence of one estrogen-active follicle. This suggests that the dominant follicle secretes some inhibitory substance(s) to decrease FSH. The dominant follicle continues to grow demonstrating that large amounts of FSH are not necessary to sustain follicular dominance. Terminal stages of antral follicular development and follicular hierarchy in monovular species may be regulated at the ovarian level by the presence of an active dominant follicle. Turnover of dominant follicles which permits development of new follicular waves appears to be determined by the inability of the dominant follicle to continue growing and suppressing growth of other follicles. There are physiological conditions which support the concept that LH exerts a regulatory effect on growth and dominance of dominant non-ovulatory and ovulatory follicles in cattle.

PAGE 51

42 Such situations include the early postpartum period in dairy and beef cows, in which development of follicles < 8 mm occurs (Savio et al., 1990b). In beef cows, during the early postpartum period two to three non-ovulatory follicles develop before the first postpartum ovulatory follicle develops (Murphy etal., 1990). Duhng these two situations LH concentrations are at basal levels. Also, growth and ovulation of follicles in prepubertal heifers can be achieved by repeated injections of LH (Tortonese et al., 1990). In summary, turnover of ovarian follicles during the estrous cycle in cattle is regulated by the concentration of in plasma acting via a negative feedback effect on LH secretion. Low frequency of LH pulses during the luteal phase is not sufficient to maintain continued growth and function of the dominant follicle. It is probable that a low androgen secretion by theca cells limits subsequent function of granulosa cells. Functional granulosa cells are required for terminal follicular development and sustained dominance of the follicle. When granulosa cell functional is compromised, the dominant follicle no longer suppresses growth of other follicles and this leads to recruitment of a new follicular wave (Savio et al., 1993a). Enhanced FSH secretion is also responsible for development of the next follicular wave. Adams et al. (1992b) showed that a surge in circulating FSH occurred during the plateau phase in maximal diameter of an anovulatory dominant follicle, and that the FSH surge was related to detectable emergence ( 4 or 5 mm follicles) of a follicular wave.

PAGE 52

43 Other Factors Involved in Control of Ovarian Follicular Dynamics Growth and development of a tissue requires the local production and integrated actions of specific growth factors. These growth factors mediate critical cell-ceil interactions that control cell proliferation and organ development. Gonadal development also requires growth factor-mediated cell-cell interactions as a general mechanism to control cellular proliferation. Ovarian physiology requires rapid and continuous growth regulation associated with the process of folliculogenesis (Skinner, 1992). There is increasing evidence that growth factors modulate folliculogenesis but their precise role in the processes of follicular growth, differentiation and atresia is still unknown (Monget and Monniaux, 1995). Growth factor-mediated interactions between theca cells, granulosa cells, and the oocyte are required for the maintenance of ovarian function and the process of oogenesis. In most large animals, follicle size increases from millimeter to centimeter (e.g., cow, mare). Granulosa and theca cells are responsible for this follicular expansion. In addition to cell proliferation required during follicle development, follicles of vanous stages of development become atretic and cell growth is arrested. Regulation of cell proliferation in the follicle requires stimulatory and inhibitory growth factors (Skinner, 1992). Ovarian follicular dynamics are controlled not only by the interaction of endocrine signals (steroid and gonadotropic hormones) but by locally produced ovarian peptide hormones and growth factors (Monget and Monniaux, 1995). The response of the two major follicular cell types, granulosa and theca cells, to

PAGE 53

44 gonadotropins is regulated by the local production of growth factors and peptides. Ovarian follicles are known to produce a range of locally acting peptide/protein growth factors that can interact directly with the same cell type from which they are produced in an autocrine manner or with other cell types via a paracnne action to attenuate or stimulate the cellular response to gonadotropins. Also these factors could act in a juxtacrine fashion, activating receptors on adjacent cells, or stimulate the cell in which the factor is secreted without prior secretion of the factor from the cell (intracrine action; Armstrong and Webb, 1997). Inhibins, activin, IGF-I, IGF-II, IGF binding proteins, TGF-a , EGF, and TGF-S and other facors have direct and indirect effects on granulosa and theca cells that can modulate follicular development and steroidogenesis. Inhibins have both autocrine and paracnne effects. They increase LH-induced androgen synthesis in theca cells (Hillieretal., 1991b; Wrathall and Knight, 1995), and inhibin production is stimulated by steroids and FSH (Wrathall and Knight, 1993). This is indicative of a local feedback loop within individual follicles involving a sequential change of inhibins, activins and their binding proteins, which determine the different fates of the selected and unselected follicles that develop in the same systemic environment of gonadotropins and growth factors (Roche, 1 996). There is evidence, mainly from studies in vitro that the inhibin-related peptides have actions on all the functional stages of folllculogenesis including ovulation (Findlay, 1993).

PAGE 54

45 Inhibin The term "inhibin" was proposed originally by McCullagh in 1932 to denote the activity of an aqueous extract of the testis that had the capacity to suppress formation of castration cells (cells that appear in the pituitary following damage of the seminiferous tubules) in the anterior pituitary gland. This was before the time when gonadotropins, FSH and LH, were defined as separate hormones and before testosterone was isolated as a pure substance. The term inhibin was used to designate a non-steroidal gonadal product with the capacity to specifically suppress FSH secretion. However, isolation of this molecule did not occur until 53 years after the original postulate (de Kretser and Robertson, 1989). Inhibin is a slow acting but powerful inhibitor of pituitary FSH biosynthesis and secretion by the anterior pituitary gland (Gaddy-Kurten et al., 1995). Inhibin is a heterodimeric glycoprotein composed of an a subunit (relative molecular mass = 18 kDa) linked by a disulfide bridge to one of two highly homologous 6-subunits (approximate M, = 14 kDa) to form either inhibin A (a-SA) or inhibin B (a-SB; Findlay et al., 1993; Rose and Gaines-Das, 1996; Halvorson and DeCherney, 1996). By sequence analysis of cDNA clones, it has been shown that the three subunits (a, and Be) are coded by different genes, and the mature structures of the a, BA, and SB subunits are present at the C-terminus of much larger precursor protein. The a subunit precursor is composed of four segments: a signal peptide, prosequence (Pro), N-terminal (aN) and C-terminal (aC) peptides in addition to the a component. The 31 -kDa inhibin has been postulated as the

PAGE 55

46 mature form of inhibin in the circulation (Findlay et al., 1993). Robertson et al. (1995, 1996), using a fractionation procedure, reported the identification of a range of bioactive and immunoactive forms of inhibin in human plasma. These forms have molecular masses of 28 to 128 kDa attributed to differences in glycosylation of the a subunit and differential processing of the a and S subunits. The inhibin B sequences also exhibit close homology with the TGF-R superfamily of structurally related proteins with wide-spread biologic activities (Massague, 1987). Granulosa cells are the site of expression of mRNA encoding the different inhibin subunits in bovine antral follicles (Torney et al., 1989; Findlay et al., 1993). Inhibin has a negative feedback effect on secretion of FSH (De Jong, 1988; Glencross etal., 1994). Inhibin is likely a chemical signal of the number of growing follicles in the ovary (Taya et al., 1996), and through its systemic action inhibin probably has an important role in determining species-specific ovulation rates (Taya et al., 1996; e.g., ovulation of only one follicle in cattle; Scanlon et al., 1993). The initial isolation of inhibin was achieved from bovine follicular fluid (bFF) as a 58 kDa glycoprotein consisting of two disulphide-linked subunits of apparent molecular masses of 43 kDa and 15 kDa. The larger subunit has been termed a, and the smaller as the S-subunit (Robertson et al., 1985). Ling et al. (1985) isolated two forms of inhibin, termed inhibin A and inhibin B, which differs by the NHj-terminal amino acid sequences of their B-subunits which are now termed and Cq. More recently. Good et ai. (1995) isolated nine different biologically and immunologically active molecular variants of inhibin (29, 34, 49, 53, 58, 77, 88, 110, and > 160 kDa) from bovine follicular fluid. Quantitative immunoblot analysis has shown alterations

PAGE 56

47 in the nine different molecular variants in bovine follicular fluid associated with functional status of dominant follicles. The predominant inhibin forms are free a subunits and large (> 34 kDa) inhibin forms rather than the fully processed inhibin form (Ireland et al., 1994). According to Good et al. (1995) each form of inhibin, except the 29-kDa form, inhibited basal FSH secretion and enhanced GnRHinduced LH secretion in pituitary bioassays. The dimeric forms of inhibin retained biological activity after isolation from bovine follicular fluid. Granulosa cells of the follicle and luteal cells in primates produce inhibin, and its production can be regulated via endocnne (FSH, LH), paracrine (EGF, TGF-a, interferon-Y, androstenedione) and autocrine (IGF-I, TGF-B, activin, follistatin) controls (Findlay, 1993). Follicle stimulating hormone, LH (at low doses), IGF-I, TGF-R, and activin can stimulate inhibin production by granulosa cells. Epidermal growth factor, TGF-a, follistatin, interferon-y, and high doses of LH negatively regulate inhibin production in the presence of FSH and other agents that increase cAMP concentrations (Findaly et al. 1993). Secretion of inhibin by rat, bovine (Henderson and Franchimont, 1983), human, and nonhuman primate granulosa cells is regulated by gonadotropins and sex steroids in vitro (Hillier and Turner, 1991). The expression of the inhibin subunit genes is regulated developmentally and inducible by FSH In the rat (Woodruff et al., 1987; 1987). In cultured rat granulosa cells, expression of aand Pg-subunit mRNA is enhanced by E^. This suggests that there is a mechanism whereby locally produced estrogen could influence relative rates of inhibin and activin synthesis by granulosa cells during follicular development in vivo (Hillier and Turner, 1991).

PAGE 57

48 Ireland and Ireland (1994) found a high correlation of a and mRNA with intrafollicular changes in P4, E2:P4 ratio, and total inhibin immunoactivity during growth of nonovulatory follicles. Ireland and Ireland (1994) reported that amounts of follicular inhibin a and mRNAs increased coincident with increases in size of nonovulatory follicles, and expression was reduced markedly in atretic nonovulatory follicles compared with estrogen-active follicles. It was also shown that amounts of Sq mRNA remained unchanged duhng follicular development. De la Sota (1 995) reported that there are two very distinct patterns of change in the amounts of immunoblot bovine inhibin in follicular fluid of dominant and subordinate follicles on day 5, 8 and 12 of the estrous cycle. Absolute amounts of four forms (>160, 160-110, 77 and 49 kDa) were high in dominant follicles of day 5 and 8, but lower in atretic dominant follicle of day 12. Conversely, absolute amounts of the 31 kDa form of bovine inhibin in follicular fluid increased with atresia of the dominant follicle. In heifers, intrafollicular concentrations of inhibin decrease during growth of dominant ovulatory follicles, but increase during growth of dominant non-ovulatory follicle (Martin et al., 1991). Thus inhibins may have an important role in regulation of growth, differentiation and ovulatory quota of dominant follicles duhng the estrous cycle in heifers. It has been shown (Scanlon et al., 1993) that immunization of beef heifers with a (1-26 Gly-Tyr) subunit of 32 kDa bovine inhibin conjugated to human serum albumin increased ovulation rate in 32% of estrous cycles of heifers. Five injections of steroid-free bovine follicular fluid (which contained inhibin), given every 12 h beginning 12 h after the onset of estrus, suppressed completely the secondary surge of FSH without affecting LH

PAGE 58

49 concentrations. This treatment delayed the appearance of the first wave of follicular development (Turzillo and Fortune, 1990). Sunderland et al. (1996) demonstrated that the gonadotropic surge causes changes in levels or proportions of certain forms of inhibins. There was no change of inhibin forms dunng the period of most active synthesis (preovulatory phase). After the gonadotropic surge, the 1 1 0-kDa form decreased while the 29-kDa inhibin form and total inhibin immunoactivity increased. In addition to gonadotropic regulation of inhibin production, paracrine and autocrine factors are also involved. First, steroids have the ability to change the secretion of inhibin in vitro. In rat granulosa cell cultures, FSH-induced immunoreactive inhibin production was increased by androstenedione and Ej, whereas in the absence of FSH neither steroid was effective (Findlay et al., 1993). Various growth factors have also been demonstrated to influence inhibin production. Both IGF-I and TGF-S increase basal and FSH-induced inhibin production by rat granulosa cells. In contrast, TGF-B inhibited basal and FSH-stimulated inhibin production by pig granulosa cells (Michel et al., 1991). Epidermal growth factor appears to be a negative regulator of both inhibin protein (Michel et al., 1991) and a and SA subunit mRNA (LaPolt et al., 1990). Interferon-Y is also a negative regulator of inhibin production by cultured immature rat granulosa cells, but only in the presence of FSH or other agents that increase cAMP levels (Xiao and Findlay, 1990). In sheep, Campbell et al. (1996) reported that granulosa cells from large follicles secreted Ej in response to FSH and secreted inhibin and P4 regardless of the presence or absence of FSH. Insulin and LR3 IGF-I (human recombinant Long

PAGE 59

50 R3 IGF-I) had a significant interaction in the stimulation of hormone production by granulosa cells from small and large follicles. On the other hand, insulin alone stimulated and inhibin secretion. A receptor with specific affinity for inhibin has not been demonstrated. However, inhibin has been shown to bind to activin type II receptors, although with lower affinity than activin. Therefore, rather than acting through a unique receptor, inhibin may exert its effect through competition with activin for activin receptor sites (Gaddy-Kurten et al., 1995; Halvorson and DeCherney, 1996). Activin Inhibin and activin are closely related peptides. Activin is composed of homodimers or heterodimers of the same R-subunits linked by interchain disulfide bonds, resulting in activin-A (RA-SA), activin-AB (SA-SB), or activin-B (SB-BB; Halvorson and DeCherney, 1 996). The molecular weight of activin is 24 to 28 kDa (Findlay et al., 1993). The B-subunit mRNAs are translated into pre-pro-S forms, which are proteolytically processed to the mature B-subunits. Activin A has been isolated from bovine follicular fluid, but activin B has not been isolated from mammalian sources so far (Findlay et al., 1993). Robertson et al. (1992) showed that the BA subunit monomer is present in bovine follicular fluid at a level 25% to 60% that of the BA subunit dimer (activin A), and its effect on in vitro responses are similar to those of the dimer. However, the monomer is less immunologically and biologically active (18% to 45%) than the dimer. It is unclear if dimerization of the monomer is a necessary prerequisite for biologic activity.

PAGE 60

51 Activin is a product of the granulosa cells, as determined from expression of the Q> subunit mRNA in these cells and the isolation of activin from follicular fluid (Findlay, 1993; Findlay et al., 1993). Activin has the ability to stimulate FSH production and decrease inhibin biopotency (Robertson et al., 1988). The activin/inhibin SB subunit Is expressed within rat gonadotropic cells, is locally secreted as SBBB activin and may function as an autocrine modulator of basal FSH secretion and expression. Inhibin and activin are part of a larger superfamily that includes transforming growth factor (TGF)-S, muilerian inhibiting substance, the decapentaplegic gene complex of Drosophila, the bone morphogenic proteins, and the vegetal growth factor gene of Xenopus (Mather et al., 1992). It is not known why different isoforms of inhibin and activin exist. It has been proposed that they may have different potencies or serve divergent functions. Also regulatory factors that determine the preferential formation of a-R (inhibin) versus R-fi (activin) dimers in cells that produce both subunits are poorly understood. Differential expression of the two subunits probably serves as one regulatory mechanism, in which an excess of a-subunit shifts production toward inhibin (Li et al., 1995; Halvorson and DeCherney, 1996). Activin is recognized as a regulator of a variety of actions, including neuronal differentiation, neuronal lifespan,and mesoderm induction in Xenopus (Mather et al., 1992). Activin affects the differentiation of granulosa cells in an autocrine manner, and its action depends on the stage of differentiation of the cells. It promotes differentiation of cells during the preantral and early antral stages of folliculogenesis and prevents premature luteinization of cells from the later stages of antral

PAGE 61

52 development. The coordinated effects lead to a promotion and maintenance of folliculogenesis (Findlay, 1993). Data from rat and bovine granulosa ceil models support the hypothesis that activin regulates granulosa cell differentiation in an autocrine fashion, and this action of activin is related to the stage of granulosa cell differentiation (Findlay, 1993). Activin is not only a mitogen for granulosa cells, but seems to play an important role in granulosa cell differentiation and ovarian follicular morphogenesis. Hulshof et al. (1997) reported that activin-A and activin receptor were co-localized in oocytes and granulosa cells of preantral follicles, suggesting an autocrine action of activin on preantral follicles. Activin stimulated both an increase in diameter of preantral follicles and proliferation of the granulosa cells. Li et al. (1995) showed that activin induced reaggregation of granulosa cells into a follicle-like structure. Activin (> 100 ng/ml) enhanced FSH-induced aromatase activity (Hutchinson etal., 1987), LH binding sites, production, FSH-induced inhibin levels and, both a and Q> subunit mRNA (Xiao et al., 1990). The mechanisms of activin action are likely to involve stimulation of FSH receptor formation and cAMP production (Findlay et al., 1993). In addition, activin increased the responsiveness of granulosa cells to both gonadotropins (Xiao and Findlay, 1991). These in vitro observations are consistent with activin acting as an autocrine regulator to promote folliculogenesis during the preantral or early antral stages. Mainly, activin induces FSH receptors as a mechanism whereby preantral follicles may become responsive to FSH (Findlay et al., 1993). This proposition assumes that the granulosa cells either constitutively express or acquire receptors for activin prior to this important step in differentiation

PAGE 62

53 (Findlay, 1 993). It is proposed that activin, produced by granulosa cells of preantral follicles lacking FSH receptors, acts on these same cells to induce FSH receptors. The FSH then acts on the cells to stimulate production of more activin and the activin receptor, and FSH and activin together further enhance FSH receptor number in partially differentiated cells. FSH and activin also induce follistatin production, which modulates the action of activin and may have an effect on the granulosa cell, independent of its activin binding properties. This model also implies that some differentiating actions of FSH on granulosa cells may be mediated by activin, which would explain why follistatin inhibits FSH-induced differentiation of the cells in vitro (Findlay, 1993). Activin action on differentiated granulosa cells is consistent with a role in preventing premature luteinization (Findlay, 1993). In bovine granulosa cell cultures, recombinant human activin-A caused a timeand dose-dependent inhibitory effect on LH-induced production of and oxytocin (Shukovski et al., 1991). Bovine activin A prevented the spontaneous luteinization of fully differentiated bovine granulosa cells in vitro (Shukovski and Findlay, 1990). Activin is also a potential autocrine modulator of aromatization in granulosa cells (Hillier, 1991). It has been shown that activin is an atretogenic agent (Woodruff et al., 1990). Exposure of rat and human theca cells to activin results in a dosedependent reduction in LH-induced androstenedione production (Hsueh et al., 1987). This would constitute a paracrine action of activin in the ovarian follicle (granulosa-thecal cells regulation) and could have important implications in the establishment of dominance in follicles (Findlay, 1993). These results suggest that

PAGE 63

54 activin plays an important role in regulating the functional integrity and development of granulosa cells In the ovarian follicle and these effects may be mediated directly through activin receptors on granulosa cells. Activin receptors can be divided into two classes: type I (actRI) and type II (actRII). ActRI binds activin with high affinity; however, binding only occurs in the presence of the signaling peptide, actRII. Conversely, actRII activity depends on the formation of a noncovalent, heteromeric complex with actRI (Mathews, 1994). This complex functions as a serine-threonine kinase (Halvorson and DeCherney, 1996). Multiple isoforms have been detected within the activin type I and type II receptor classes. Each isoform demonstrates distinct ligand-binding affinities and cytoplasmic domain structure. This observation raises the possibility that doseand cell-specific responses could be attained through the expression of various combinations of activin receptor isoforms and number (Halvorson and DeCherney, 1996). Type I receptors are expressed and reach the cell membrane even in the absence of the type II receptors. Type I receptors do not bind ligand without the coexpression of a type II receptor. Both classes are required for any response (Gaddy-Kurten et al., 1995). Activin and inhibin share a common Q> subunit, and both bind to Qjmacroglobulins and to follistatin, which is a specific activin S subunit-binding protein (Gaddy-Kurten et al., 1 995). Although the physiological relevance of the inhibin-Osmacroglobulin and activin-a2-macroglobulin interaction is unknown, both complexes are biologically active to modulate FSH secretion. Although most activin appears to be complexed with Os-macroglobulin, complex formation probably does not affect

PAGE 64

•••«9r^'55 activin function. The a2-macroglobulin presumably binds activin with lower affinity than follistatin or cell surface receptors, which could account for the lack of effect on activin bioactivity; alternatively, activin may be able to interact with its receptor while complexed to Qj-macroglobulins. The physiological function of this complex formation is not known, but aj-macroglobulins may serve roles in storage, delivery and clearance of activin (Mathews, 1994). In contrast, activin-follistatin complex is biologically inactive and it is likely that this binding protein plays an important role in limiting exposure of cells to activin (Gaddy-Kurten et al., 1995). Follistatin Follistatin is a single peptide chain (molecular weight 32,000 35,000 Daltons) that was first isolated from porcine follicular fluid, and is distinct from inhibin and activin. This cysteine-rich protein can inhibit the release but not the synthesis of FSH and has no effect on LH in cultures of pituitary cells (Tonetta and diZerega, 1989). Follistatin is structurally related to the activin-inhibin subunits. The primary structure of follistatin across species is highly conserved (> 97%). Follistatin is coded by a single gene organized into three homologous domains, each with > 50% homology. Alternative mRNA splicing results in two mature mRNA forms: the longer encoding FS-315 and the shorter encoding FS-288. The existence of two isoforms of follistatin provides cells with a mechanism to control follistatin activity through alterations in post-transcriptional processing (Halvorson and DeCherney, 1996). Follistatin is expressed in a number of different tissues, but

PAGE 65

56 a major site of production is the ovarian granulosa cells under FSH regulation (Findlay et al., 1993). It has been demonstrated that follistatin mRNA and protein expression change in association with folliculogenesis and atresia (Findlay et al., 1993). Follistatin mRNA is first detected in granulosa cells of secondary follicles and then becomes more abundant as the follicle forms an antrum, with uniform expression in the whole granulosa cell layer. The strongest signal is found in preovulatory follicles and newly formed corpora lutea. Follistatin protein is confined to healthy dominant preovulatory follicles and a subpopulation of tertiary follicles in animals entering estrus (Shukovski et al., 1992). Atretic follicles show no immunoreactive follistatin in granulosa cells. The bovine follistatin mRNA signal increases as folliculogenesis progresses, with the strongest signal being observed in preovualtory bovine follicles (Shukovski et al., 1992). Findlay (1993) and Findlay et al. (1993) hypothesized that follistatin modulates granulosa cell function in an autocrine fashion and that follistatin action is through binding and neutralization of activin (Mathews, 1994). Thus, follistatin is likely to favor the process of follicular luteinization or atresia. However, follistatin may also have direct actions on granulosa cells independent of its activin-binding activity. Xiao et al. (1990) reported that follistatin enhanced the stimulatory action of forskolin on P4 production by rat granulosa cells but did not influence the effect of forskolin on either aromatase activity or inhibin production. In contrast, activin enhanced all three responses stimulated by forskolin. The binding of activin to follistatin is through two binding sites present in activin-A (Sa-^a) . whereas inhibin-A

PAGE 66

57 (a-li^) has only one binding site for follistatin. Shimonaka et al. (1 991 ) hypothesized that follistatin binds activin and inhibin through the common S subunit. There is strong evidence for a short loop feedback system involving activin and follistatin in the ovarian follicle. Follicle-stimulating hormone can stimulate production of both activin and follistatin. Follistatin can neutralize activin action and possibly some actions of FSH that might be mediated by activin (Findlay et al., 1993). The production of the activin-binding protein, follistatin, in the same tissue may further limit the bioavailaibility of activin and inhibit activin-induced cell growth and formation of follicles, as reported by Li et al. (1995). Alternatively, because the affinity of activin for follistatin is comparable to that of activin for its receptors, follistatin could directly compete with activin receptors of lower affinity for activin. However, higher affinity activin receptors could still bind ligand and elicit signal transduction (Mathews, 1994). Transforming Growth Factor R Transforming growth factor-S is a family of multifunctional growth factors originally named for their ability to induce normal rat kidney fibroblasts to grow in soft agar in the presence of epidermal growth factor (EGF; Mulheron and Schomberg, 1 993). The TGF-S superfamily is made up of a number of proteins with the potential to act as intraovarian regulators of ovarian function (Armstrong and Webb, 1997). Also, mullerian inhibiting substance is included among the members of the TGF-S family (Mulheron and Schomberg, 1993). Transforming growth factors-S are ubiquitous peptides and have been shown to affect nearly every cell

PAGE 67

58 in the body. There are five TGF-fi subtypes, three of which (TGF-lil , TGF-B2. and TGF-I3.3)are expressed by cells of the mammalian ovary (Hernandez et al., 1990; Mulheron and Schomberg, 1990). They are products of separate genes, but few cellular responses to TGF-G. can be ascribed solely to one isoform. Transforming growth factor-B is synthesized as an inactive precursor protein that must be cleaved for the 25-kDa homodimer to elicit its effects. The precise mechanism of TGF-li activation in vivo is unknown, but is thought to occur via induction of proteases such as plasmin or perhaps by processing in an acidic environment. In vitro TGF-S can be activated by exposure to heat, detergent, or acid. Once activated, TGF-B is thought to affect target ceils through TGF-li-specific membrane-bound receptors, but the precise cellular mechanisms of TGF-R actions have not been elucidated (Mulheron and Schomberg, 1993). The most likely mediators of TGF-B signal transduction are three cell surface proteins, which were identified based on their ability to bind and be chemically cross-linked to radiolabeled TGF-B (Massague and Like, 1985). These three molecules have been designated type I, II, and III receptors and have apparent molecular weights of 55, 80, and 280 kDa, respectively. Data from a variety of studies indicate that receptors types I and II are responsible primarily for signal transduction. The type III receptor, a membrane-anchored proteoglycan, also called betaglycan, that may help modulate cellular access to TGF-B by sequestering it near the cell membrane (Mulheron and Schomberg, 1993). The type II receptor has a functional cytoplasmic serine/threonine kinase domain, which suggests that serine/threonine phosphorylation may be an important component in the intracellular

PAGE 68

59 signal transduction of TGF-R. The type II receptor is structurally related to the activin receptor. Regulation of TGF-fJ activity is controlled by the activity of proteases that release mature TGF-li from the extracellular matrix-associated latent complex. The conversion of plasminogen to plasmin within the follicle and the presence of plasminogen activator inhibitors would be expected to be critical in the control of TGF-B activity dunng folliculogenesis. This mechanism would be expected to influence development of dominance since TGF-Bs act synergistically with gonadotropins to control the differentiation of follicular cells. Specific extracellular proteases play a central role in the regulation of TGF-R bioactivity during follicle growth and development of dominance (Armstrong and Webb, 1 997). Given that TGF-6 is a multifunctional growth factor, it can act as both a growth stimulator and an inhibitor depending on cell type and culture conditions. Transforming growth factor-S is primarily an inhibitor of follicle growth and therefore facilitates follicle cell differentiation by modulating granulosa cell expression of functional gonadotropin receptors (Mulheron and Schomberg, 1993). Ovarian theca cells have been shown to express and produce TGF-B (Skinner et al., 1 987a). Transforming growth factor-6 has been shown to stimulate a number of granulosa cell functions, including FSH induction of LH receptors, EGF actions, FSH induction of aromatase activity, production of IGF-I and inhibin. Transforming growth factor-S also can influence theca cell function and steroidogenesis, and oocyte maturation (Skinner, 1992). Transforming growth factor S inhibits growth of bovine and porcine granulosa cells (Skinner et al., 1987a) and theca cells (Roberts and Skinner, 1991) induced by TGF-a/EGF. The influence of TGF-S on cell function

PAGE 69

60 may be mediated indirectly through inhibition of cellular proliferation. Growth inhibition may be important to prevent premature cell growth of the preantral follicle, arrest cell growth during atresia, and control cell growth during follicle cell expansion (Skinner, 1992). Overall, these data indicate that TGF-Rs are local regulators of ovarian function (Armstrong and Webb, 1997). In cows, TGF-Bs inhibit granulosa and theca cell proliferation while enhancing gonadotropin-stimulated steroidogenesis (Roberts and Skinner, 1991). Epidermal Grov^th Factor/Transforming Growth Factor-g Other growth factors known to affect ovarian function include TGF-a, a structural analogue of EGF. Transforming growth factor-a is expressed in bovine theca cells and has been localized in the thecal cell (Skinner and Coffey, 1988) layer during follicular growth. However, there is a decrease in immunoreactive TGFa in preovulatory follicles that is correlated with a decrease in mitotic activity of granulosa cells (Lobb and Dorrington, 1992). Epidermal growth factor and TGF-a are products of separate genes, however, they are structurally similar and both elicit their effects through the EGF receptor (Mulheron and Schomberg, 1993). Epidermal growth factor and TGF-a are believed to be produced by ovarian theca cells (Skinner et al. 1987b). Epidermal growth factor is a small (53 amino acid, 6043 molecular weight), single-chain polypeptide containing six half-cysteine residues that form three intrachain disulfide bonds. It is synthesized as a large molecular weight precursor that can be proteoliticaly cleaved to generate the biologically active form. Within

PAGE 70

61 species, two to six EGF isoforms have been elucidated that have variable degrees of receptor binding and bioactivity (Mulheron and Schomberg, 1993). Transforming growth factor-a is synthesized as a glycosylated precursor protein of 159 or 160 amino acids with the mature 50-amino acid form being generated by proteolytic cleavage. The precursor form contains a signal peptide sequence and a hydrophobic transmembrane domain , suggesting that the precursor form of TGF-a is an integral membrane glycoprotein (Mulheron and Schomberg, 1993). Membrane-anchored TGF-a binds to EGF receptors on adjacent cells and elicits a response. Such a mechanism would allow for activation of adjacent cells by membrane-anchored TGF-a in the absence of proteases. In the follicle, this could provide a direct means of intracellular communication and consequently an additional mechanism for controlling follicle cell growth. Epidermal growth factor/TGF-a acts on target cells via interaction with specific 170-kDa transmembrane receptors which are tyrosine kinases. The phosphorylated Cterminus is thought to unmask the catalytic domain of the receptor. A number of additional intracellular signaling pathways are activated within minutes of EGF binding to its receptor: activation of phospholipases, increase in intracellular calcium, activation of protein kinase C, and activation of serine/threonine kinases. Ultimately these individual changes result in altered cellular growth and differentiation (Mulheron and Schomberg, 1993). Epidermal growth factor/TGF-a may be a primary mitogen of granulosa cells in most species. (Lobb and Dorrington, 1992). Transforming growth factor-a/EGF has been also shown to stimulate maturation of granulosa cells (Tonetta and

PAGE 71

62 DiZerega, 1989) but inhibits follicular secretion in sheep (Webb et al., 1994). Epidermal growth factor can inhibit production of both inhibin and progesterone by bovine granulosa cells, but stimulates DNA and protein synthesis (Tonetta and DiZerega, 1989). Epidermal growth factor binding is stimulated by FSH, and EGF in turn enhances FSH binding (Urban and Veldhuis, 1992). EGF inhibits FSHinduced aromatase and the formation of LH receptors in the rat and pig (Hsueh et al., 1984). Prior exposure of granulosa cells to FSH has been reported to diminish the inhibitory effects of EGF on cell differentiation (Paton and Collins, 1992). In humans, EGF and TGF-a are negative modulators of FSH-induced synthesis of Ej and positive modulators of granulosa cell proliferation. Their production by human follicles seems to decrease as follicular diameter enlarges (Gougeon, 1996). In bovine granulosa cells in vitro, EGF, TGF-a and bFGF stimulate proliferation but inhibit steroidogenesis (FSH-induced Ej secretion; Rouillier et al., 1 997) and inhibin production (Skinner and Coffey, 1988; Vernon and Spicer, 1994), suggesting that they might act in vivo by enhancing growth and delaying terminal maturation of follicles (Monniaux et al., 1997). Apoptotic cell death in granulosa cells of follicles selected for ovulation is prevented by the paracrine actions of EGF/TGF-a produced by thecal-interstitial cells or the autocrine and paracrine actions of bFGF synthesized by granulosa cells (Hsueh et al., 1994). Lobb and Dorrington (1992) proposed a model for the role of transforming growth factors in ovarian physiology and follicular development. In primordial follicles, TGF-a of theca origin diffuses into the granulosa cell compartment and positively stimulates granulosa cell mitosis. Granulosa cells in these follicles are not

PAGE 72

63 fully differentiated and do not synthesize estrogens. While promoting granulosa cell growth, TGF-a may also maintain the granulosa cell in an undifferentiated state. At a latter stage of development (large antral follicles), granulosa cell mitosis declines and the cells begin to synthesize estrogens. TGF-P negatively affects granulosa cell growth while promoting gonadotrophin-induced steroidogenesis in these follicles. The interplay of the two growth factors leads at first to increase cell number and latter the augmentation of differentiate function, that culminates into a mature estrogen-active preovulatory follicle. Insulin-like Growth Factors (IGFs) and IGF-bindinq Proteins (IGFBPs) Insulin-like growth factors have direct effects on cultured ovarian cells. These effects include stimulation of granulosa cell mitogenesis, granulosa and luteal progesterone production, and thecal cell androgen production that appear similar among species (Spicer and Echternkamp, 1995). Administration of growth hormone stimulates growth of small antral follicles probably by an indirect mechanism involving IGF-I of endocrine origin. Terminal follicular growth Is a strictly gonadotropin dependent process corresponding to initiation of follicular waves, selection of dominant follicles and terminal maturation of preovulatory follicles. The IGF family of growth factors may control the growth of small antral follicles in conjunction with FSH in the initiation of follicular waves. They are likely to be part of the players in the process of selection of the dominant follicle (Monniaux et al., 1997). The IGF system is composed of different elements: two ligands (IGF-I and IGF-II), two receptors, and six or more IGF-binding proteins. IGFs function as

PAGE 73

64 modulators of gonadotrophin action at the cellular level and stimulate granulosa and thecal cell proliferation and differentiation (Armstrong and Webb, 1997). The type I receptor mediates most of the somatomedin-like actions of both IGF-I and -II. It is an a232 tetramer structurally and functionally related to the insulin receptor. The affinity of this receptor for IGF-I is slightly higher than for IGF-II and much higher than for insulin. The type II receptor binds IGF-II and molecules that possess a mannose-6-phosphate residue such as lysosomal enzymes. The IGF-II receptor does not bind insulin, and binds IGF-I with very low affinity (Monget and Monniaux, 1995). The six IGFBPs bind IGF-I and II with high affinity. They are classified into two groups: IGFBP-1, -2, -4, -5 and -6 called 'small complex', which are present in serum and other fluids. Their molecular masses range between approximately 24 kDaand35kDa. Their concentrations in serum are either negatively (IGFBP-1 and -2) or not regulated by growth hormone. IGFBP-3 is the most predominant IGFBP in serum. It is present in a 1 50 kDa form or 'large complex'. The concentration of IGFBP-3 is positively regulated by growth hormone and IGF-I. IGFBPs increase the half life of IGFs, and can both inhibit or potentiate IGF action at target cells. Their affinity for IGF-I and -II is of the same order of magnitude as the affinity of the type I receptor. Actions of IGF-I on follicular cells are mediated on granulosa and thecal cells by type I receptors (Giudice, 1992). In cows, expression of the IGF-I receptor increases in small antral follicles (Wandji et al., 1992) whereas dominant follicles express abundant IGF-I and IGF-II mRNA (Yuan et al., 1998) but failed to express IGFBP-2 which is a binding protein associated with atresia (Monget and Monniaux,

PAGE 74

65 1 995). In sheep as in rats, pigs and humans, IGF-I stimulates both proliferation and differentiation of granulosa cells from follicles of 1-3 mm in diameter (Monget and Monniaux, 1 995) and thecal cell steroidogenesis (Giudice, 1 992; Paton and Collins, 1992). On follicles > 5 mm, IGF-I stimulates secretion of P4 by granulosa cells (Monniaux and Pisselet, 1992). The mRNA encoding IGF-II has been detected in thecal tissue of bovine ovarian follicles (Armstrong and Webb, 1997), and localized in both granulosa and thecal cells of sheep (Leeuwenberg et al., 1995). The functional significance of these differences in the expression of mRNA encoding IGFs between species in unknown. It may reflect changes in the relative roles of IGF-I and -II that have evolved to fit the particular pattern of follicular development (Armstrong and Webb, 1997). As with the IGFs, the spatial expression of the binding proteins within ovarian follicles is species-specific. In cows (Armstrong et al., 1996; Yuan et al., 1998) and sheep (Besnard et al., 1996a), expression of mRNA encoding IGFBP-4 and -2 is restricted to theca and granulosa cells, respectively. The release of IGFs from the IGFBPs is controlled by the action of specific IGFBP proteases. Specific IGFBP proteases have been detected in follicular fluid from ewes (Besnard et al., 1996b) and the amounts and activity of these enzymes change during folliculogenesis. Interaction between IGFs and other peptides (e.g., EGF, TGF-a) may be significant in regulating not only the proliferation but also the survival of granulosa cells (Paton and Collins, 1992). In cattle, doses of 100 ng/ml of IGF-I either had no effect or inhibited basal and FSH-induced E2 production by granulosa cells collected from small (< 5 mm) follicles, and was a weak stimulator

PAGE 75

66 of FSH-induced production by granulosa cells from large follicles (> 8 mm). Insulin, IGF-I and IGF-II stimulate progesterone production by bovine granulosa cells and, concomitant with FSH treatment, enhances the stimulatory effect of IGF-I . Insulin and IGF-I synergize with LH to promote thecal cell androgen production, and bovine thecal cells appear much more sensitive to insulin (Spicer et al., 1993). No changes in follicular fluid concentrations of IGF-I were detected in first wave dominant (Badinga et al., 1992, de la Sota et al., 1996) and subordinate (de la Sota et al., 1996) follicles collected on days 5, 8 and 12 of the estrous cycle. In contrast, Echternkamp et al. (1994) reported higher concentrations of IGF-I in follicular fluid of large active follicles compared to E2 inactive and small follicles. De la Sota et al. (1996) and Echternkamp et al. (1994) reported changes in the absolute amounts and proportions of IGFBP-2, -4 and -5. They were at low levels in E2 active dominant follicles but are at high levels in atretic dominant follicles, which may reduce the concentrations of free IGF-I and -II available in follicular fluid to exert their trophic activity on granulosa cells. Manipulation of Th e Follicular Dynamics in Cattle The failure of hormonal treatments to control follicular development and superovulation in women and domestic animals in a very precise manner gives evidence of significant gaps in the understanding of the mechanisms that regulate development and function of ovulatory follicles (Fortune etal., 1988). Furthermore, the mechanisms that control the ovulatory quota in different domestic animals are not completely understood. It is clear that the development of mediumsized and

PAGE 76

67 large follicles in cattle occurs in waves. The regularity of the patterns allows experimental manipulations designed to determine how these patterns are controlled (Fortune et al., 1991). Control of follicular development in cattle involves the removal of the suppressive effect of the dominant follicle to allow emergence of a new follicular wave at a specific time after treatment. Different approaches have been followed to control follicular dynamics in cattle. Hormonal approaches have been directed to cause luteinization or atresia of the follicles present at the time of treatment (Bo et al., 1994). This has been accomplished by using hCG (Rajamahendran and Sianangama, 1992; Diaz et al., 1993) or GnRH (Macmillan and Thatcher, 1991; Schmitt et al. , 1 996b) to induce follicle luteinization or ovulation, or by progestogens and Ej to cause atresia of the dominant follicle (Bo et al., 1994). A multitude of experimental approaches have been used to control follicle dynamics in cattle, including treatments with hCG (Diaz et al., 1993; Schmitt et al., 1996b; Sianangama and Rajamahendran, 1996), GnRH analogs (Macmillan and Thatcher, 1991; Rusbridgeetal., 1992; Schmitt etal., 1996a), electrocautery of the dominant follicle during its growing phase (Ko et al., 1991; Adams et al., 1993b), ultrasound-guided follicle aspiration (Bergfelt et al., 1994), pretreatment with FSH given early in the estrous cycle to increase the number of embryos recovered after a superovulatory treatment (Goulding etal., 1991; Gray etal., 1992), andtreatment with high doses of during the growing phase of the dominant follicle to suppress the dominant follicle and hasten the emergence of the next wave (Adams et al., 1992a; Burke etal., 1994).

PAGE 77

68 Hormonal Manipulation: Human Chorionic Gonadotropin and Gonadotropin Releasing Hormone It has been demonstrated that the first wave dominant follicle can be ovulated after administration of Gonadotropin Releasing Hormone (GnRH; Rusbridge et al., 1992; Schmitt et al.. 1996a) or human Chorionic Gonadotropin (hCG). Human Chorionic Gonadotropin is a glycoprotein with LH-like activity that promotes ovulation of the first wave dominant follicle and formation of an accessory corpus luteum (CL) when administered on d 4 (Breueletal., 1989; Price and Webb, 1 989), day 5 (Walton et al. , 1 990; Schmitt et al , 1 996a, 1 996b), day 6 (Fricke et al., 1993), day 7 (Breuel et al , 1989; Rajamahendran and Sianangama, 1992; Sianangama and Rajamahendran, 1992; 1996). day 10 (Breueletal., 1989), or day 14 to 16 (Pnce and Webb, 1989) of the estrous cycle. Both GnRH and hCG have the ability to induce an accessory CL during the early luteal phase of the estrous cycle in cattle, which could have some practical applications. Establishment and maintenance of pregnancy as well as embryo survival in cattle are related to the ability of the CL to secrete (Thatcher et al., 1994b). Subluteal levels of could be a factor that contribute to embryo losses (Thatcher et al., 1994b). Also, the rate of increase in plasma P4 during the early luteal phase could play a role in establishment of pregnancy (Shelton et al., 1990). Progesterone supplementation after insemination in heat stressed cows may correct deleterious effects of impaired luteal function on pregnancy rates. Inducing formation of an additional CL during the luteal phase of the estrous cycle with an

PAGE 78

69 injection of either GnRH or hCG could be a strategy to increase concentrations of P4 in plasma during the period when maternal recognition of pregnancy occurs (Schmitt, 1995). Ablation of the Dominant Follicle There is a potential for artificial control of follicular dynamics using ablation procedures. Two types of ablation procedures have been tried: electrocauterization (Ko et al., 1991; Adams et al., 1993a) and ultrasound-guided transvaginal follicle aspiration (Bergfelt et al., 1994; Stubbings and Walton, 1995). The first one is an invasive method and not repeatable on a within cow basis, whereas follicle aspiration is a noninvasive and repeatable procedure that can be used for the retrieval of oocytes for in vitro fertilization (Pieterse et al., 1988, 1991a, 1991b). It has been shown that electrocauterization (Ko et al., 1991; Adams et a!., 1 993a) of the dominant follicle during its growing phase hastened emergence of the next follicular wave. Ko et al. (1991) showed that cauterization of first wave dominant follicle on days caused an early emergence of wave 2, and increased the incidence of 3-wave interovulatory intervals. Follicle ablation via ultrasound-guided follicle aspiration at random (all follicles < 5 mm) stages of the estrous cycle synchronized subsequent follicular wave emergence and resulted in a high degree of synchronized ovulations in heifers following POFs^ induced luteolysis (Bergfelt et al., 1994).

PAGE 79

70 Superstimulation The objective of ovarian superstimulatory treatment in cattle is to obtain the maximum number of viable embryos by stimulating growth and subsequent ovulation of competent antral follicles with exogenous gonadotropins (Adams, 1994). However, the variable and unpredictable superovulatory response of the donor animal has remained one of the most limiting factors to successful embryo transfer (Armstrong, 1993). To date, all methods of inducing superovulation in cows have Involved administration of exogenous gonadotropic hormones: crude or purified hypophyseal extracts, such as porcine follicle-stimulating hormone and luteinizing hormone, equine chorionic gonadotropin and human menopausal gonadotropin (Mapletoft et al., 1994). Many reports have been published on dosage regimes and types of gonadotropin preparations for ovarian superstimulation (Elsden etal., 1978; Murphy et al., 1984; McGowan et al., 1985; Lerner et al., 1986; Pawlyshyn et al., 1986; Armstrong, 1993) but timing of treatment with respect to monitored follicular wave development has received little attention (Adams et al., 1993b). Animal variability is important as a determinant of superovulatory response. The ovarian status of the donor at the time of hormone treatment appears to be a major determinant of the superovulatory response (Monniaux et al., 1983). In recent years, research has been focused on making responses more predictable by administering superstimulation treatments with specific regard to the status of follicular development at initiation of treatment (reviewed by Armstrong, 1 993). Results from

PAGE 80

71 different studies (Pierson and Ginther, 1988; Grasso et a!., 1989; Guilbault et a!., 1991) suggested that the superovulatory response is affected by the presence of a large dominant follicle (Pierson and Ginther, 1 988; Guilbault et al., 1 991 ; Adams et al., 1993b; Wolfsdorf et al., 1997). Superovulation may be induced with equal efficacy when treatment is initiated during the time of the first or second follicular waves when a dominant follicle is present in the ovaries, and the superovulatory response is enhanced if treatment is initiated at the time of wave emergence, before the time of follicle selection (Adams, 1994). It has been shown (Wehrman et al., 1996) that FSH treatment in presence of a persistent dominant follicle resulted in follicles that were unable to respond to the preovulatory surge of LH. Priming doses of FSH given at the beginning of the cycle have been shown to increase the superovulatory response to FSH treatments (Rajamahendran et al., 1987; Ware et al., 1987; Touati et al., 1991). Others have failed to show the beneficial effect of early phming (Grasso et al., 1989; Gray et al., 1992). The idea of the phming is to increase the number of follicles in the FSH-responsive pool. It is likely that priming with other hormones, such as PMSG, which possesses both FSH and LH activity, could increase the number of FSH-responsive follicles (Monniaux et al., 1984). Passive immunization against inhibitors of gonadotropin secretion may also be a means of increasing ovulatory rates by short-term elevation of levels of endogenous follicle-stimulating hormone (Mapletoft et al. , 1 994). It has been shown that passive immunization of ewes against steroid free-follicular fluid or against synthetic inhibin peptides increased ovulation rates twoto four-fold (O'Shea et al.,

PAGE 81

72 1991; Schanbacheretal., 1991). Ovulation rate and prolificacy in the cow are more difficult to manipulate by inhibin vaccination (O'Shea et al., 1994). Hov/ever, there are reports that immunization of heifers or cows against steroid-free follicular fluid increases ovulation rate (Price et al., 1987; Glencross et al., 1992; Morris et al., 1993; Scanlon et al., 1993) and number of transferable embryos (Alvarez et al., 1997). In summary, there are several factors involved in the control of ovarian follicular dynamics in cattle, and some of the mechanisms that control ovarian follicular growth have been elucidated, but still there are more that are not completely understood; as an example, factor(s) and mechanisms that control dominance during a follicular wave. A set of experiments was designed and performed with the aim of this study to identify the factor (s) and mechanisms that control follicular dominance in cattle,. Understanding these mechanisms will be important for the design of programs for synchronization of the estrous cycle and ovulation, and for superovulation and embryo transfer.

PAGE 82

CHAPTER 3 HUMAN CHORIONIC GONADOTROPIN-INDUCED ALTERATIONS IN OVARIAN FOLLICULAR DYNAMICS DURING THE ESTROUS CYCLE OF HEIFERS Introduction The understanding of factors controlling ovarian folliculogenesis should lead to development of new and more precise methods of controlling reproductive cycles in livestock. One aspect of follicular dynamics is the process of continual growth and regression of antral follicles that leads to development of the preovulatory follicle (Lucy et al., 1992). One to four waves of follicular growth and development occur during a single estrous cycle in heifers and cows, and the preovulatory follicle is derived from the last wave. During each wave, development of antral follicles > 2 mm proceeds through stages of follicular recruitment, selection, and dominance (Savio et al., 1988; 1990; Lucy et al., 1992; Fortune, 1994; Ginther et al. 1996). Development of the first dominant follicle (DF) is the most consistent event related to follicular dynamics during the estrous cycle in heifers and cows. The first dominant follicle can be identified retrospectively by ultrasonography between day 2 and 4 of the estrous cycle (Savio et al., 1988; 1990; Sirois and Fortune, 1988; Driancourtetal., 1991; Ginther et al., 1996). Human chorionic gonadotropin (hCG) is a glycoprotein with LH-like activity 73

PAGE 83

74 that promotes ovulation of the first wave DF (FWDF) and formation of an accessory corpus luteum (CL) when administered on day 4 (Breuel et al., 1989; Price and Webb, 1989), day 5 (Walton et al., 1990; Schmitt et al., 1996a, 1996b), day 6 (Fricke et al., 1993), day 7 (Breuel et al., 1989; Rajamahendran and Sianangama, 1992; Sianangama and Rajamahendran, 1992; 1996), day 10 (Breuel etal., 1989), or day 14 to 16 (Price and Webb, 1989) of the estrous cycle. With formation of an accessory CL, there is a subsequent Increase in concentrations of plasma progesterone (P^; Breuel et al., 1989; Fricke et al., 1993; Schmitt et al., 1996a). Only one study has characterized ovarian follicular dynamics during this period (Sianangama and Rajamahendran, 1996). The objective ofthe present experiment was to characterize hCG-induced alterations in ovarian follicular and CL dynamics in heifers when hCG was injected on day 5 of the estrous cycle. Materials and Methods Seventeen sexually mature Holstein heifers were selected from the dairy herd at the University of Florida. Heifers were kept in a lot with minimal pasture and fed peanut hay and corn silage. Estrous cycles were synchronized following removal of a norgestomet ear implant (6 mg norgestomet implant without injection of the norgestomet/estradiol valerate solution; Synch romate-B; Sanofi Animal Health, Inc., Overland Park, KS) that had been in place for 7 days and injection of prostaglandin (PG) F2a (25 mg i.m., Lutalyse, Pharmacia & Upjohn Co., Kalamazoo, Ml) on day 6 or 1 day before withdrawal ofthe implant. Brightly colored enamelbased paint (Impervo; Benjamin Moore, Montvale, NJ) was applied to the tailhead

PAGE 84

75 of each heifer at the time of PGF2a administration. When the ear implant was removed, the paint strip was covered with a contrasting color of chalk (All-Weather® Painstick®, Lake Chem. Co., Chicago, IL; Van Cleeff et al., 1996). The extent in cover of tail paint and chalk (TPC) was scored based on a modification of the method of Macmillan et al. (1988) using a scale of 5 (no signs of estrus and full presence of paint and chalk) to 0 (standing estrus and absence of paint and chalk; Van Cleeff et al., 1996). Scoring of the TPC and visual detection of estrus were performed twice daily (0600 to 0700 and 1800 to 1900) for 3 days after removal of the implant. On day 5 of the estrous cycle (day 0= day of estrus), heifers (n=6) were treated with hCG (1,000 lU given i.v.; 2,000 lU given i.m.; Steris Laboratories, Phoenix, AZ; n=6; Schmitt et al., 1996a) or i.m. injection of saline (control group; n=7). A real time ultrasound scanner equipped with a linear array 7.5 MHz transrectal transducer (Equisonic LS 1000, Tokyo Keiki Co., LTD, Tokyo, Japan) was used to monitorfolliculardevelopment daily from day 3 of the estrous cycle until the next ovulation (first cycle). During the following estrous cycle (second cycle), follicular development and CL growth were monitored every other day until estrus. Ovarian structures (follicles > 2 mm and CL) were measured, and their relative positions recorded on ovarian maps drawn during the examination. Blood samples were taken daily during the first and second estrous cycles. The dominant follicle for each follicular wave was defined as the largest newly emerged ovarian follicle that was 2 mm greater than the second largest follicle and grew linearly (Sirois and Fortune, 1990). Blood samples (10 ml) were collected daily via venipuncture from the jugular

PAGE 85

76 vein into heparinized tubes. Blood samples were stored in an ice bath, plasma separated by centrifugation (1,800 x g for 15 minutes), and stored at -20 °C until assayed for and estradiol (Ej). Concentrations of P4 in plasma were measured by radioimmunoassay (Knickerbocker et al., 1986). Sensitivity of the assay was 0.15 ng/ml. Intraand inter-assay coefficients of variation were 4.6 and 13.2%, respectively. Concentrations of Ej from day 6 before ovulation until day 7 of the second cycle were measured by a radioimmunoassay described by Tortonese et al. (1990) and modified by Badinga et al. (1992). Sensitivity of the assay was 1.0 pg/ml and the intraand inter-assay coefficients of variation were 6.9 and 19.9%, respectively. Statistical Analvses Data were analyzed by least squares analysis of variance using the General Linear Models procedure of the Statistical Analysis System (SAS, 1988). The following definitions were used to characterize each follicular wave: a) day of emergence of DF was the day when the DF first appeared in the ovary and was at least 5 mm in size, b) day of selection of DF was the day of the cycle when the difference between DF and second largest follicle (subordinate) was > 2 mm, c) day of maximum recruitment was the day when maximum number of class 2 follicles (69 mm) was present in the ovary, and d) duration of a follicular wave was the difference between day of emergence of the DF of two consecutive waves. The mathematical models used to analyze size of dominant and subordinate follicles and number of follicles within discrete size classes (Class 1 , 2 to 5 mm;

PAGE 86

77 Class 2, 6-9 mm; Class 3, > 10 mm) included effects of treatment, cow(treatment), day of the wave (wave II, days 1 to 18 [days relative to day of emergence of second wave DF] and wave III, days 1 to 13 [days relative to day of emergence of third wave DF]), and the interaction of day of the wave by treatment. The main effect of treatment was tested using the mean square of cow (treatment) as the error temri. Tests for homogeneity of regression equations were performed on equations for follicular diameter versus day of wave for dominant and subordinate follicle of the second and third follicular waves. These analyses were performed as described by Wilcox etal. (1990). The same analytical approach was used to analyze plasma concentrations of P4 during the first and second estrous cycles for hCG and control heifers, plasma concentrations of E2 during the peri-ovulatory period, sizes of the original and induced CL, and patterns of growth of original CL in hCG-treated and control heifers. However, day of cycle (PJ, days from ovulation (E2), and days after ovulation (original vs. induced CL, and comparison of original CL growth for both groups) were used for these analyzes. The mathematical model used to analyze the follicular dynamics and plasma concentrations of between heifers with two and three follicular waves during the first and second estrous cycles did not include hCG-treated heifers during the first estrous cycle. The model included effects of wave, cycle, cow(wave-cycle), day, and interactions wave by day, cycle by day. and wave by cycle by day. The main effects of wave and cycle were tested using the mean square of cow(wave-cycle) as the error term. Duration of the second and third follicular waves in all heifers with three-wave

PAGE 87

78 cycles were analyzed using a mathematical model that included the effects of treatment, wave(second or third wave), and the interaction of treatment by wave. The frequencies of accessory CL and estrous cycles of two versus three follicular waves between treatments were compared using Chi-square analyses. Results Corpora lutea and plasma progesterone Treatment with hCG at day 5 induced ovulation and formation of an accessory CL in all six heifers of the hCG group, while none of the control heifers ovulated the first wave dominant follicle (x^= 13, P < 0.01). Progesterone concentrations increased from 0.9 ng/ml on day 1 to 6.4 ng/ml on day 6 and did not differ (P > 0.10) between heifers in both treatments. However, by day 9, P4 concentrations were greater (P < 0.01; Figure 3-1 A) for heifers of the hCG group (15.3 ng/ml) than for control heifers (11.1 ng/ml) and remained elevated until day 17. Maximal concentrations of P4 were observed on day 15 (hCG group, 20.4 vs. control group, 12.9 ng/ml). On day 19.5 and 20.3 of the estrous cycle, there was a 50% decrease in P4 concentrations for hCG and control groups, respectively, indicating that luteolysis had been initiated; P4 concentrations were < 1 ng/ml on day 21 for both groups. Concentrations of P4 during the following estrous cycle did not differ between groups (P > 0.10). The profile was the same in both groups, reaching maximal concentrations on day 13 (11.3 vs. 10.6 ng/ml, for hCG and control groups, respectively; Figure 3-1 B).

PAGE 88

79 5 10 15 2D C^frcmCUJ^cn 25 5 10 15 2D C^lTcmCXU^cn 25 Figure 3-1 Plasma concentrations of progesterone (LSM ± SEIVI) during the first (A) and second (B) estrous cycle in control () and hCG-treated () heifers. Asterisk (*) indicated significant differences (P < .01) between mean concentrations within day. There was a difference (P < 0.01; Figure 3-2) in concentrations of between heifers with two (13 heifer-cycle cells) and three follicular waves (8 heifercycle cells). This comparison was made on all heifers except hCG-treated heifers of the first experimental cycle. Concentrations of P4 from day 13 to 20 were lower in heifers having two-wave cycles compared with three-wave cycles. There was a 50% decrease in P4 concentrations by day 17.7 in heifers with two waves and day 21.3 in heifers with three follicular waves, which is indicative that the luteolytic mechanism from the uterus was active earlier in heifers with two follicular waves (Figure 3-2). Dynamics of CL development differed (P < 0.001) between groups due to

PAGE 89

80 0 5 10 15 20 25 Days from Ovulation Figure 3-2. Plasma concentrations of progesterone (LSM ± SEM) for heifers with two () and three () follicular waves during the first and second estrous cycles. Astensl< (*) indicates differences (P < .01) between mean concentrations within day. induction of accessory CL in the hCG-treated heifers (Figure 3-3). In hCG-treated heifers, growth of the original and induced CL did not differ (P > 0.10) from day 3 to day 8 after ovulation. The accessory CL was smaller (P < 0.01) than the original CL and regressed concurrently with the original CL in hCG-treated heifers. There was no difference (P > 0.10) in average size between original CL of hCG-treated and control heifers (Figure 3-3A and 3-3B), but there was a treatment by day interaction (P < 0.001) for the pattern of growth of the original CL in hCG-treated and control heifers. Between day 3 and 1 1 of the estrous cycle, original CL of hCGtreated heifers grew to a larger size following injection of hCG.

PAGE 90

81 1 i T \ I 1U1 \ \ 1 \ 1 0 5 10 15 23 25 0 5 10 15 23 2 C^cfEstrajsCyde Cfeycf BtrosCyde Figure 3-3. Growth of the original () and induced () corpora lutea in hCG-treated heifers (A) and original () corpus luteum in control (B) heifers during the first estrous cycle. Asterisk (*) indicates differences (P < .01) between least squares means within day. Follicular Dynamics All hCG-treated heifers had three follicular waves during the first experimental estrous cycle (Figure 3-4). In the control group, in contrast, three heifers had three follicular waves (Figure 3-5A) and four had two follicular waves (Figure 3-5B). The frequency of three follicular waves was greater (x^= 4.95; P < 0.05) in the hCG group. The second wave dominant follicle emerged earlier in hCG-treated heifers (day 7.3; Figure 3-4) compared with the control group (day 10.4; P < 0.01; Figure 3-5A and 3-5B), and reached 10 mm in diameter on day 10 of the estrous cycle for hCG-treated heifers compared with 13.6 day for heifers in control group with either

PAGE 91

82 o 4o hCG (3,000 iu) 20 0 5 10 15 Day of Estrous Cycle 20 25 Figure 3-4. Growth of the first (), second (), and third () dominant follicles in hCG-treated heifers (n=6) during the first estrous cycle. two or three follicular waves. In both groups, the second wave dominant follicle was selected on day 3.4 of the wave, and day of maximum recruitment was day 3.7. A treatment by wave interaction was detected (P < 0.01) for duration of waves for heifers with three follicular waves during the first cycle (Figure 3-4 and 3-5A). In the control group, the second and third follicular waves lasted 8.3 ± 0.9 days and 5.3 ± 0.9 days, respectively. For the hCG group, duration of the second follicular wave was shorter (6.3 ± 0.7 days), possibly due to greater P4 concentrations whereas duration of the third wave was longer (9.2 ± 0.7 days) than in the control group. Maximum diameter of the second wave dominant follicle was smaller (P < 0.01) in hCG treated-heifers (12.8 ± 0.8 mm; Figure 3-4) than in control heifers (15.6 ± 0.8 mm; Figure 3-5A) with three-wave cycles. Associated with the differences in duration of third wave follicles (hCG, 9.2 days > control, 5.3 days; P < 0.01) were differences in follicular dominance as measured by growth patterns of the largest subordinate follicle. The subordinate follicle of the hCG group decreased in size

PAGE 92

83 18 1614 |i(H E CD Q. 5 ^ 0 IL 18 1614h |i(H E 0 1 6H 0 4i 0 3 6 9 12 15 18 21 24 27 C^cfBbojBQfde 0 3 6 9 12 15 18 21 24 2 C^cfEsbasCVde Figure 3-5. Growth of the first (), second (). and third () dominant follicles in control heifers with three (n= 4; A) and two (n= 3; B) follicular waves during the first estrous cycle. during the 6 days of the preovulatory period (b= -0.007 mm/day). In contrast, size of the subordinate follicle of the control group increased during this period (b= 0.47 mm/day; P < 0.01). This may be a reflection of a more developed and active preovulatory dominant follicle in hCG-treated heifers that suppressed growth of the subordinate follicle. The third dominant follicle emerged on day 14.7 and 18 of the estrous cycle for heifers of the hCG and control groups, respectively, selection occurred on day 3.7 of the third wave for both groups. During the second estrous cycle, one hCG-treated heifer had three follicular waves and five had two follicular waves, while three saline-treated heifers had three follicular waves and four heifers had two follicular waves. A total comparison of twoand three-wave cycles, excluding first cycles for hCG-treated heifers, was made

PAGE 93

84 I I I — \ — \ — \ — I — I oH — I — I — \ — I — \ — I — I — \ — I 0 3 6 9 12 15 18 21 24 27 0 3 6 9 12 15 18 21 24 2 CfeyrfEstrousCyde DayofEstrousCyde Figure 3-6. Growth of the first (), second (), and third (A) dominant follicles in heifers with two (n=13; A), and three {n=8; B) follicular waves during the first and second estrous cycle. Heifers treated with hCG during first estrous cycle are not included. Utilizing first and second estrous cycles. The second wave dominant follicle in twowave cycle heifers had a longer life span (11 days) than the second dominant follicle of three-wave cycle heifers (6 days; Figure 3-6A and 3-6B). Similarly, there was a significant wave by day interaction (P < 0.01) for size of the ovulatory follicle in heifers with two and three follicular waves. The ovulatory follicle of heifers with two-wave cycles entered a plateau phase for 3 days before ovulation; whereas, growth of the ovulatory follicle of heifers with three-wave cycles was linear until ovulation. Estradiol-17a Concentratlnns Plasma concentrations of from day 6 before ovulation until 7 days post ovulation did not differ (P > 0. 1 0; Figure 3-7A) between treatments. Heifers treated

PAGE 94

85 -6^-2 02468 -64-2 0246 Figure 3-7. Plasma concentrations of estradiol-1711 (LSIVI ± SEM) in hCG-treated () and control () heifers (A), and control heifers with two follicular waves (; n= 4; B) and three follicular waves (; n=3; B) during the peri-ovulatory period. with hCG tended to have greater concentrations (15.6 pg/ml) than heifers of the control group (8.2 pg/ml) on the day before ovulation. This trend was due to one heifer that had very high concentrations of E2. Following ovulation, plasma concentrations of increased to 8.7 pg/ml on day 4 for hCG-treated heifers and to 5.6 pg/ml on day 5 for the control group. This increase in did not differ between treatments and coincided with growth of the first wave dominant follicle before the luteal phase increase in concentrations of P^. Plasma concentrations of E2 tended (P < 0. 1 0) to be greater from day -3 until day of ovulation (day 0; Figure 3-7B) in control heifers with two follicular waves (n=2) than in control heifers with three follicular waves (n=3). This coincides with the fact that the second dominant

PAGE 95

86 follicle was present longer on the ovary. The ovulatory follicle in control heifers with two follicular waves reached a size > 1 0 mm on day 1 3.5 of the estrous cycle versus day 22.3 for control heifers with three follicular waves (Figure 3-5A and 3-5B). Duration of Estrous Cycles Duration of the first estrous cycle for hCG group and control group was 22.9 ± 0.9 and 22.1 ± 0.9 days (P > 0.10), and 21 .8 ± 1 .0 and 22.7 ± 1 .0 days (P > 0.1 0) for the second estrous cycle. Analysis of the first and second expehmental cycles, excluding the first estrous cycle of hCG-treated heifers, indicated that estrous cycles with two waves of follicular development were shorter (20.5 ± 0.5 days; P > 0.01) than those of three-wave estrous cycles (24.6 ± 0.7 days). Discussion Cows and heifers exhibit two or three successive waves of follicular development during the estrous cycle (Savioetal., 1988; Siroisand Fortune, 1988; Gintheretal., 1989a; 1989b). In this study, the FWDF present at day 5 ovulated in response to hCG and formed a CL which is in agreement with others (Breuel et al., 1989; Price and Webb, 1989; Walton et al., 1990; Rajamahendran and Sianangama, 1992; Sianangama and Rajamahendran, 1996; Schmittetal., 1996a, 1996b). Previous work indicated that the response was affected by stage of the estrous cycle when hCG was given. Induction of an accessory CL was greater on days 4 to 7 and days 14 to 16 than between days 8 to 1 3 of the estrous cycle (Price and Webb, 1 989). This is due likely to transitional stages in follicular status within

PAGE 96

87 the first (Badinga et., 1992; Thatcher et al., 1996) and second follicular waves that are either responsive or nonresponsive to hCG. Schmitt et al. (1 996a) reported that the greater growth of ohginal CL in hCG-treated compared to control heifers could be due to a luteotropic effect of hCG on development of the original CL after day 5. In the present study, hCG acutely altered follicular dynamics by ovulating the first wave dominant follicle on day 5. All hCG-treated heifers had three follicular waves during the experimental estrous cycle. Elimination of the first wave dominant follicle via hCG induction of ovulation leads to an earlier recruitment and emergence of the second-wave DF. Badinga et al. (1992) reported that removing the ovary bearing the DF increased FSH leading to follicular recruitment, thus, removal of the FWDF by hCG-induced ovulation may have had the same effect in this study. Ko etal. (1991) reported that destruction of the FWDF later during the wave is followed by an earlier recruitment of the next wave. A comparable effect appears to occur with elimination of the DF via induction of ovulation with hCG. Although an earlier emergence of the second wave follicle occurred in hCGtreated heifers, maximum size of the second wave follicle was reduced and had a shorter lifespan. This was attributed to the greater P4 concentrations associated with induction of an accessory CL in this group. Indeed, greater concentrations of P4 reduce growth of the DF and induce follicle turnover (Savio et.,1993; Burke et al., 1994; Thatcher et al., 1994; Kinder et al., 1996) due to reduction in LH secretion (Kinder et al., 1996). Functional duration of the third wave follicle was increased because of a shorter life span of the second-wave follicle in hCG-treated heifers,. Indeed, a greater decrease in size of the subordinate follicle during the preovulatory

PAGE 97

88 period for the third wave dominant follicle of the hCG group was detected compared to that of the preovulatory subordinate follicle of three-wave cycles in the control group. Regardless of day of emergence or duration of a wave, dominant follicles were selected between day 3 and 4 of the respective wave. Ginther et al. (1996) reported that between day 0 and 4 of a wave, the dominant and largest subordinate follicles diverge gradually in diameter. Administration of hCG during the luteal phase of the estrous cycle increased P4 concentrations in blood and in some cases extended duration of the estrous cycle (Seguin etal., 1977; Breueletal., 1989; Walton etal., 1990; Rajamahendran and Sianangama, 1992; Fricke et al., 1993; Schmitt et al., 1996a). This effect appears to be somewhat dependent on presence of a functional CL because hCG administered on day 1 (early luteal phase) or day 17 (late luteal phase) of the estrous cycle had no effect on serum P4 concentrations (Seguin et al., 1 977; Breuel et al., 1989). In this experiment, administration of hCG on day 5 increased plasma P4 concentrations from day 9 to 17. The increase in P4 concentrations was due most likely to additional P4 secretion by accessory CL (Schmitt et al., 1 996b) at this stage of the estrous cycle as opposed to a stimulation of the original CL (Fricke et a!., 1993). Based upon diameter of the induced CL and the temporal increase in plasma P4 concentrations, induced CL appear to be functional but smaller than the original CL. Sianangama and Rajamahendran (1996) demonstrated that the CL induced by hCG given on day 7 of the bovine estrous cycle is functional but appears to be smaller and secretes less P4 than the spontaneous CL of similar age. Because concentrations of plasma P4 during the second cycle were similar

PAGE 98

84 1614E (D E 8i a) D o 1 4H 2-1 1614•12(I) b 0) 6o 1 4-1 ~l I I I I I I I I 0 3 6 9 12 15 18 21 24 27 2-1 DE^ofEstrousCVde -| — I — I — I — I — \ — I — I — I 0 3 6 9 12 15 18 21 24 2 D^ofEstrajsCyde Figure 3-6. Growth of the first (), second (), and third () dominant follicles in heifers with two (n=13; A), and three {n=8; B) follicular waves during the first and second estrous cycle. Heifers treated with hCG during first estrous cycle are not included. Utilizing first and second estrous cycles. The second wave dominant follicle in twowave cycle heifers had a longer life span (11 days) than the second dominant follicle of three-wave cycle heifers (6 days; Figure 3-6A and 3-6B). Similarly, there was a significant wave by day interaction (P < 0.01) for size of the ovulatory follicle in heifers with two and three follicular waves. The ovulatory follicle of heifers with two-wave cycles entered a plateau phase for 3 days before ovulation; whereas, growth of the ovulatory follicle of heifers with three-wave cycles was linear until ovulation. Estradiol-17S Concentrations Plasma concentrations of from day 6 before ovulation until 7 days post ovulation did not differ (P > 0. 1 0; Figure 3-7A) between treatments. Heifers treated

PAGE 99

-6^-2 02468 -64-2 0246 C^kmCklsticn H^ftanOAJsticn Figure 3-7. Plasma concentrations of estracliol-17(l (LSM ± SEM) in hCG-treated () and control () heifers (A), and control heifers with two follicular waves (; n= 4; B) and three follicular waves (; n=3; B) during the pen-ovulatory period. with hCG tended to have greater concentrations (15.6 pg/ml) than heifers of the control group (8.2 pg/ml) on the day before ovulation. This trend was due to one heifer that had very high concentrations of E2. Following ovulation, plasma concentrations of increased to 8.7 pg/ml on day 4 for hCG-treated heifers and to 5.6 pg/ml on day 5 for the control group. This increase in Eg did not differ between treatments and coincided with growth of the first wave dominant follicle before the luteal phase increase in concentrations of P4. Plasma concentrations of Ej tended (P < 0.10) to be greater from day -3 until day of ovulation (day 0; Figure 3-7B) in control heifers with two follicular waves (n=2) than in control heifers with three follicular waves (n=3). This coincides with the fact that the second dominant

PAGE 100

86 follicle was present longer on the ovary. The ovulatory follicle in control heifers with two follicular waves reached a size > 1 0 mm on day 1 3.5 of the estrous cycle versus day 22.3 for control heifers with three follicular waves (Figure 3-5A and 3-5B). Duration of Estrous Cycles Duration of the first estrous cycle for hCG group and control group was 22.9 ± 0.9 and 22.1 ± 0.9 days (P > 0.10), and 21 .8 ± 1 .0 and 22.7 ± 1 .0 days (P > 0.10) for the second estrous cycle. Analysis of the first and second expehmental cycles, excluding the first estrous cycle of hCG-treated heifers, indicated that estrous cycles with two waves of follicular development were shorter (20.5 ± 0.5 days; P > 0.01) than those of three-wave estrous cycles (24.6 ± 0.7 days). Discussion Cows and heifers exhibit two or three successive waves of follicular development during the estrous cycle (Savio etal., 1988; Siroisand Fortune, 1988; GInther et al., 1 989a; 1 989b). In this study, the FWDF present at day 5 ovulated in response to hCG and formed a CL which is in agreement with others (Breuel et al., 1989; Price and Webb, 1989; Walton et al., 1990; Rajamahendran and Sianangama, 1992; Sianangama and Rajamahendran, 1996; Schmittetal., 1996a, 1996b). Previous work indicated that the response was affected by stage of the estrous cycle when hCG was given. Induction of an accessory CL was greater on days 4 to 7 and days 1 4 to 1 6 than between days 8 to 1 3 of the estrous cycle (Price and Webb, 1989). This is due likely to transitional stages in follicular status within

PAGE 101

87 the first (Badinga et., 1992; Thatcher et al., 1996) and second follicular waves that are either responsive or nonresponsive to hCG. Schmitt et al. (1 996a) reported that the greater growth of original CL in hCG-treated compared to control heifers could be due to a luteotropic effect of hCG on development of the original CL after day 5. In the present study, hCG acutely altered follicular dynamics by ovulating the first wave dominant follicle on day 5. All hCG-treated heifers had three follicular waves during the experimental estrous cycle. Elimination of the first wave dominant follicle via hCG induction of ovulation leads to an earlier recruitment and emergence of the second-wave DF. Badinga et al. (1992) reported that removing the ovary bearing the DF increased FSH leading to follicular recruitment, thus, removal of the FWDF by hCG-induced ovulation may have had the same effect in this study. Ko et al. (1991) reported that destruction of the FWDF later during the wave is followed by an earlier recruitment of the next wave. A comparable effect appears to occur with elimination of the DF via induction of ovulation with hCG. Although an earlier emergence of the second wave follicle occurred in hCGtreated heifers, maximum size of the second wave follicle was reduced and had a shorter lifespan. This was attributed to the greater P4 concentrations associated with induction of an accessory CL in this group. Indeed, greater concentrations of P4 reduce growth of the DF and induce follicle turnover (Savio et.,1 993; Burke et al., 1994; Thatcher et a!., 1994; Kinder et al., 1996) due to reduction in LH secretion (Kinder et al., 1996). Functional duration of the third wave follicle was increased because of a shorter life span of the second-wave follicle in hCG-treated heifers,. Indeed, a greater decrease in size of the subordinate follicle during the preovulatory

PAGE 102

88 period for the third wave dominant follicle of the hCG group was detected compared to that of the preovulatory subordinate follicle of three-wave cycles in the control group. Regardless of day of emergence or duration of a wave, dominant follicles were selected between day 3 and 4 of the respective wave. Ginther et al. (1996) reported that between day 0 and 4 of a wave, the dominant and largest subordinate follicles diverge gradually in diameter. Administration of hCG during the luteal phase of the estrous cycle increased P4 concentrations in blood and in some cases extended duration of the estrous cycle (Seguin etal., 1977; Breueletal., 1989; Walton etal., 1990; Rajamahendran and Sianangama, 1992; Fricke et al., 1993; Schmitt et al., 1996a). This effect appears to be somewhat dependent on presence of a functional CL because hCG administered on day 1 (early luteal phase) or day 17 (late luteal phase) of the estrous cycle had no effect on serum P4 concentrations (Seguin et al., 1977; Breuel et al., 1989). In this experiment, administration of hCG on day 5 increased plasma P4 concentrations from day 9 to 17. The increase in concentrations was due most likely to additional P4 secretion by accessory CL (Schmitt et al., 1 996b) at this stage of the estrous cycle as opposed to a stimulation of the original CL (Fricke et al., 1993). Based upon diameter of the induced CL and the temporal increase in plasma P4 concentrations, induced CL appear to be functional but smaller than the original CL. Sianangama and Rajamahendran (1996) demonstrated that the CL induced by hCG given on day 7 of the bovine estrous cycle is functional but appears to be smaller and secretes less P^ than the spontaneous CL of similar age. Because concentrations of plasma P4 during the second cycle were similar

PAGE 103

89 between groups, luteotrophic effects and alterations in dynamics of follicular growth observed in the first cycle after hCG administration failed to elicit any carry-over effects that influenced CL development and function of the second cycle. Similarly, Breuel et al. (1989) did not observe any effect of hCG on CL formed during subsequent cycles. Greater concentrations of plasma P4 in heifers with three spontaneous follicular waves during the cycle could contribute to a greater turnover of the second wave dominant follicle as a consequence of decreased LH pulsatility (Kinder et a!., 1996). Estrous cycles were significantly longer for three-wave cycles and growth characteristics of the ovulatory follicle were different between three-wave versus two-wave follicle cycles. The duration of the luteal phase appears to determine, at least in part, the number of follicular waves during an estrous cycle (Fortune, 1994). Estrous cycles with three follicular waves have slightly but significantly longer luteal phases than cycles with two waves (Ginther et al., 1989b; Fortune, 1993). We choose to interpret that two-wave follicular cycles induce an earlier CL regression compared to three-wave cycles. Contributing to this is presence of an active estrogenic follicle when CL regression is initiated. This could be determined partially by high plasma P4 concentration leading to a turnover of the second wave follicle versus persistence of a potentially estrogenic second wave follicle in luteal phase with lower progesterone concentrations. Third wave ovulatory follicles underwent a linear growth until ovulation versus a distinct plateau phase for the second wave ovulatory follicle. Whether fertility differs to insemination following ovulation of two versus three wave ovulatory

PAGE 104

95 to FSH superstimulation protocol. A total of eight cyclic dairy cows was synchronized with an injection of GnRH (Buserelin; 8 pg i.m.; Receptal® HoechstRoussel AgriVet, Sommerville, NJ) followed 7 days later by an Injection of prostaglandin (PGFja; 25 mg i.m.; Lutalyse, Pharmacia & Upjohn, Kalamazoo, Ml; Figure 4-1). Brightly colored enamel-based paint (Impervo, Benjamin Moore, Montvale, NJ) was applied to the tailhead of each cow one day before injection of PGF 2aOVARIECTOMY EXPERIMENT 1 SYNCHRONIZATION PERIOD 1 SMB In + PGF2a« GnRH PGF2a ASPIRATION OF DF +/EMBRYO RECOVERY 1 SMB out PGF2a»« ESTRUS £j Al« IT 0 ESTRUS A 4 • Day6: 25 mg • • Day6.5: 15 mg • •» 2 Al: 10 hs apart 5 6 7 10 12 13 14 15 17 f f SUPEROVULATION FSH (mg) AM: 6 5 3 2 PM:6 5 3 2 23 ULTRASOUND BLOOD: P4, E2 Figure 4-1 Experimental sequence for Experiment 1 and Experiment 2 On day of PGFj^ injection, the paint strip was covered with a contrasting color of chalk (All-Weather® Paintstick®, Lake Chem. Co., Chicago, IL; Van Cleeff et al., 1996). The extent of tail paint and chalk (TPC) removal was scored based on a modification of the method of Macmillan et al. (1988) using a scale of 5 (full

PAGE 105

96 presence of paint and chalk, not in estrus) to 0 (absence of paint and chalk, standing estrus; Van Cieeff et al., 1996). Scoring of TPC and visual detection of estrus were performed twice daily (0600 to 0700 h and 1800 to 1900 h) for 3 days after injection of PGFja. On day 6 of the estrous cycle (day 0 = estrus), all cows received a 6 mg Norgestomet ear implant without injection of the norgestomet/estradiol valerate solution (Syncromate-B; Sanofi Animal Health, Inc., Overland Park, KS) and two injections of PGF^a (25 mg AM and 15 mg PM, i.m.) to induce luteolysis and create a low progestin environment for development of an active-FWPDF (Savio et al., 1993b). The implant was left in place for 9 days. Cows were assigned randomly to either a nonaspiration group (control group; n=4) or the treatment group in which the induced active-FWPDF was removed by transvaginal aspiration (treatment group; n=4). The first wave DF was removed via transvaginal ultrasound guided aspiration on day 10 (e.g., 4 days after insertion of Norgestomet implant), utilizing an Aloka Echo-Camera SSD-500 unit (Aloka Co., LTD, Japan) equipped with a 5.0 MHz transvaginal convex array transducer and needle guide. Epidural anesthesia was induced with 5 ml of 2% lidocaine and the perineal region was scrubbed and disinfected. The lubricated transducer with needle guide was inserted deep into the vagina. A 17-gauge, 60 cm, single channel sterile needle, fitted with a 19-gauge, 2.5 cm disposable tip was used in the needle guide to puncture the vaginal wall and peritoneum. By manipulation of the reproductive tract per rectum, ovaries were positioned and held firmly in front of the transducer in close apposition to the vaginal

PAGE 106

97 wall. After aligning the on-screen needle travel indicator line with the follicle, the needle was advanced through the wall of the vagina, into ovarian stroma and through the wall of the follicle. As soon as the tip of the needle was seen to enter the follicle, follicular contents were aspirated into a sterile plastic tube using a suction unit (Pioneer Medical, Inc., Madison, CT) preset to a flow rate of 20 ml/min at a preset pressure of 100 mnfi Hg. Collection of follicular fluid into tube and collapse of the DF on screen were regarded as a successful aspiration of the follicle. On the basis that an immediate rise in concentrations of FSH occurs after cauterization of the DF (Adams et al., 1992b) or after removal of the ovary beahng the DF in which an increase in FSH precedes recruitment of a new follicular wave by 2 days (Badinga et al., 1992), follicular superinduction was initiated 2 days after aspiration of the active-FWPDF. For follicular superinduction, a total of 32 mg Armour units of FSH-P (FSH-P™; Schering-Plough, Animal Health Kenilworth, NJ) was given per animal, with injections initiated on day 12 of the experimental estrous cycle (e.g., 2 days after follicular aspiration) at 0700 h and continued at 12 h intervals in a decreasing regime over a 4-day period (A.M./P.M.= 6/6, 5/5, 3/3 and 2/2 mg per injection). A real time ultrasound scanner (Aloka Echo-Camera SSD500) equipped with a 7.5 MHz transrectal linear array transducer was used to monitor size and number of ovarian follicles > 2 mm in diameter from day 5 until ovariectomy on day 1 5 of the experimental estrous cycle. Ovarian maps with the relative position of follicles and corpora lutea were drawn during each examination. According to the diameter, follicles were grouped into the following size classes:

PAGE 107

98 Class 1, < 5 mm; Class 2, 6-9 mm; Class 3, > 10 mm. Accumulated follicle diameter was obtained by summing the diameter of all follicle sizes within a class. Experiment 2 The objective was to evaluate folliculogenesis, ovulatory responses and embryo yields to a superovulatory protocol with FSH-P administered in the presence or absence of an active-FWPDF. Eleven cycling dairy cows were used in the experiment. Synchronization of estrus, induction of an active-FWPDF, follicle aspiration and induction of superovulation procedures were the same as in experiment 1 (Figure 4-1). However, ultrasonography was extended until detection of estrus on day 17. Ultrasonography was used to confirm ovulation 2 days after estrus and to count number of corpora lutea on day of embryo recovery. Cows were inseminated twice at 12-h intervals starting 10 hours after detection of estrus. Estrus detection was done in the same manner as described in experiment 1. Embryos were recovered nonsurgically on day 6 after insemination, following a procedure described by Putney et al. (1988). At the time of embryo collection, number of embryos retrieved per cow was recorded and the embryos were classified by stage of development and quality according to the criteria of the International Embryo Transfer Society (1 990). Embryos were grouped into six stages: unfertilized embryos (UFO), 2-12 cell, early morula, morula, early blastocyst, and blastocyst. Embryos were given a quality classification score of 1 = excellent, perfect embryo; 2= good, trivial imperfections; 3= fair, definite but no severe problems such as extruded cells or a small amount of degeneration; and 4=

PAGE 108

99 poor, partly degenerated or vesiculated cells (Putney et al., 1988). Potential embryos were classified as UFO plus all embryos. Freezable and transferable embryos included only quality 1 and 2 embryos that were in the morula or blastocyst stages of development. Blood samples (10 ml) were collected daily from a tail vessel into heparinized tubes, from day 0 (first estrus) until the day of ovariectomy (experiment 1) or until the day of embryo recovery (day 6 of next cycle, experiment 2). Blood samples were stored in an ice bath and plasma separated by centrifugation (1,800 g for 15 min) and stored at -20°C until assayed for P^ and estradiol (Ej). Concentrations of P4 were measured by radioimmunoassay (RIA, Knickerbocker et al., 1986) in all samples. Sensitivity of the P4 assay was 0.28 ng/ml. Intraand inter-assay coefficients were 8.5% and 13.1%, respectively. Concentrations of in plasma from day 0 to day 15 (day of ovariectomy, experiment 1) or until day of estrus (experiment 2) were measured by RIA as described by Tortonese et al. (1990) and modified by Badinga et al. (1992). Sensitivity of the assay was 0.8 pg/ml, and the intraand inter-assay coefficients of variation were 8.5% and 22.8%, respectively. Statistical Analvses Data were analyzed by least squares analysis of variance using the General Linear Models procedure of the Statistical Analysis System (SAS, 1 988). Data from both experiments were analyzed together since inclusion of replicate in the mathematical model was not significant. The mathematical model used to analyze the number of follicles and the accumulated follicle diameter within discrete size

PAGE 109

100 classes, size of the dominant follicle, and plasma concentrations of P4 and E2 included effects of treatment, cow(treatment), day of the cycle, and treatment by day interaction. The main effect of treatment was tested using the mean square of cow(treatment) as the error term. When interaction treatment by day was significant, data were analyzed using orthogonal contrasts to compare hormonal concentration between groups. Plasma concentrations of E2 were analyzed in two periods from day 0 to day 1 0 and from day 1 0 to day 16. Plasma concentrations of P4 were analyzed from day 0 to day of embryo recovery. To correct for heterogeneity of variance, plasma concentrations of P4 were analyzed after logarithmic transformation (natural log) of P4 concentration + 5, from day 16 to day of embryo recovery. A one way analysis of vanance was used to analyze discrete responses to superovulation treatment, such as: follicles > 9 mm on the day of estrus; number of ovulatory follicles (difference between the number of follicles > 9 mm present in the ovary on the day of estrus and the number of nonovulatory follicles 2 days after ovulation); percentage of ovulatory follicles (number of ovulatory follicles as a percentage of follicles > 9 mm present in the ovary on the day of estrus); number of nonovulatory follicles (number of follicles > 9 mm remaining in the ovary two days after ovulation); number of corpora lutea (number of CL present in the ovary the day of embryo recovery as determined by ultrasonography); number of UFO and embryos (sum of UFO plus all embryos); recovery rate relative to ovulatory follicles (rate between number of UFO plus embryos and number of ovulatory follicles); recovery rate relative to number of CL (number of UFO plus embryos as a

PAGE 110

101 percentage of CL number); number of embryos; fertilization rate (ratio between number of embryos and total number of UFO and embryos); number of transferable embryos; percentage of transferable embryos relative to number of UFO and embryos (number of transferable embryos as a percentage of total number of UFO and embryos); percentage of transferable embryos relative to number of embryos (number of transferable embryos as a percentage of embryos); percentage of embryos relative to number of CL (number of embryos as a percentage of total CL number) and percentage of transferable embryos relative to number of CL (number of transferable embryos as a percentage of total CL number). Results Follicular Dvnamics The first wave dominant follicle was identified in all cows at the ultrasound examination on day 5 of the cycle. Under a low progesterone environment, created by insertion of the norgestomet ear implant and PGFja injections on day 6, the FWPDF increased its size from 1 1 .0 ± 0.6 mm on day 5 to 15.8 ± 0.6 mm on day 10 in the control group and from 10.3 ± 0.6 mm on day 5 to 15.1 ± 0.6 mm on day 1 0, for the treatment group (Figure 4-2). In the control group, the FWPDF continued growing and reached 19.9 ± 0.7 mm on day 16 of estrous cycle. Follicular aspiration in the treatment group was associated with disappearance of the activeFWPDF. Number of class 1 follicles (Figure 4-3A) did not differ between groups from

PAGE 111

102 day 5 until day 10, and from day 15 until day 17. Number and accumulated diameter (Figure 4-3B) for class 1 follicles decreased (P < 0.05) in the treatment group from day 1 1 until day 1 5. Number of class 1 follicles on day 1 0 (34. 1 ±3.1; Figure 4-3A) decreased to 3.2 ± 3.9 on day 16 and accumulated class 1 follicle diameter decreased from 94.9 ± 9.4 mm on day 10 to 13.8 ± 12.1 mm on day 16 (Figure 4-3B). In contrast, for the control group, number of class 1 follicles increased from day 10 (30.7 ± 2.9) until day 12 (41.4 ± 3.1) and accumulated class 1 follicle diameter increased from day 10 (97.4 ± 8.9 mm) until day 13 (126.5 ± 9.5 mm; Figure 4-3B) before decreasing to the same levels at day 16. FSH ends FSH SMBout Day of Estrous Cycle Figure 4-2. Pattern of growth of the first wave dominant follicle from day 5 to day 16 of the estrous cycle in control () and treatment (•) cows.

PAGE 112

103 Figure 4-3. Number (LSM ± SE; A) and accumulated follicle diameter (mm; LSM ± SE; B) for class 1 follicles from day 5 to day 17 in control () and treatment (•) cows. Asterisk (*) indicates differences {P < .001) between means within day. E E Day of Estrous Cycle Day of Estrous Cycle Figure 4-4. Number (LSM ± SE; A) and accumulated follicle diameter (mm; LSM ± SE; B) for class 2 follicles from day 5 to day 17 in control () and treatment (•) cows. Asterisk (*) indicates differences (P < .001) between means within day.

PAGE 113

104 The immediate reduction in class 1 follicles, in the treatment group, was associated with increased recruitment (3.3 > 0.6 ± 1.2; P < 0.1) and accumulated follicle diameter (20.5 > 4.2 ± 8.6 mm; P < 0. 1 ) of class 2 follicles (Figure 4-4, A and B) on day 12 just prior to FSH-P injections and on day 13 (1 day after FSH-P injections began) to 8.2 > 1.4 ± 1.2 and 52.0 > 9.1 ± 9.2 mm for number and accumulated diameter of class 2 follicles, respectively (P < 0.001). The decline in number of class 1 follicles of the control group (Figure 4-3A) did not occur until day 14 (30.9 ± 3.1) at which time there was an increase in number of class 2 follicles (9.6 ± 1 .3; Figure 4-4A). This increase continued until day 1 5 for the control group (18.9 ± 1 .3) when number of class 2 follicles was significantly less and declining in the treatment group (9.3 ± 1 .4; P < 0.001 ; Figure 4-4A). Accumulated diameter of class 2 follicles followed the same pattern as number of class 2 follicles (Figure 44B). The decline in number of class 2 follicles of the treatment group was associated with an earlier increase and greater numbers of class 3 follicles on days 14(5.9± 1.2VS 1.0 ± 1.1; P< 0.001; Figure 4-5A), 15 (15.9 ± 1.2 vs5.6± 1.1; P< 0.001; Figure 4-5A) and 16 (15.6 ± 1.5 vs 9.7 ± 1.3; P < 0.001) compared to the control group. Presence of an active-FWPDF delayed the FSH-induced increase in class 3 follicles by 1 to 2 days. Number of class 3 follicles was not different significantly between groups at day 17 (23.7 ± 2.3 vs 19.6 ± 2.3). However, accumulated diameter of class 3 follicles was significantly greater in the treatment group (319.0 ± 29.5 mm vs 220.0 ± 29.5 mm; P < 0.01; Figure 4-58) due to an earlier appearance in class 3 follicles on the ovary contributing to a greater

PAGE 114

105 accumulated diameter. These differences in dynamics of follicle change reflect the inhibitory effect of the active-FWPDF that can be over-ridden with continuous treatment of exogenous FSH-P. Plasma Progesterone and Estradiol-17R Concentrations Concentrations of P4 in plasma did not differ between the two groups (Figure 4-6A). Concentrations of plasma P4 increased from day 1 (0.6 ± 0.6 ng/ml) until day 6 (5.6 ± 1 .9 ng/ml) at which time two injections of PGFja were administered. After injection of PGFja, plasma concentrations of P4 decreased to basal levels and remained low (< 1 ng/ml) until day 19 when they increased to 3.0 ± 3.5 ng/ml in the control group, and 4.7 ± 3.8 ng/ml in the treatment group. Day 19 was approximately 1 to 2 days after ovulation. On the day of embryo recovery, plasma concentrations of P^ were the same for both groups. The tendency of higher P4 concentrations in the control group is due to one cow which had 108.4 ng/ml of P4 and 19 CL. Concentrations of plasma were higher (P < .001 ) in the control group than in the treatment group on day 8 (9. 1 ± 0.7 pg/ml vs 6. 1 ± 0.8 pg/ml), day 9 (9.6 ± 0.7 pg/ml vs 5.9 ± 0.8 pg/ml), day 10 (11.5 ± 0.7 pg/ml vs 5.9 ± 0.7 pg/ml), and after follicular aspiration on day 1 1 (13.5 ± 0.8 pg/ml vs 2.7 ± 0.9 pg/ml). day 12 (14.8 ± 0.8 pg/ml vs 3.1 ± 0.9 pg/ml), and day 13 (17.1 ± 0.9 pg/ml vs 7.1 ± 0.9 pg/ml; Figure 4-6B) of the estrous cycle.

PAGE 115

106 DaysofEstrousCyde Days of Estrous Cyde Figure 4-6. Concentrations of progesterone (LSM ± SE; A) from day of estrus (day 0) to day of embryo recovery and estradiol-17(J (LSM ± SE; B) from day of estrus until day 16 of the estrous cycle in control () and treatment (•) cows. Asterisk (*) indicates differences (P < 001) between means within day. Ovarian Responses to Superovulation Treatment On the day of detected estrus, number of follicles > 9 mm, number ofovulatory follicles, percentage of ovulatory follicles, and number of nonovulatory follicles did not differ between treatments (Table 4-1). The array of other ovarian and embryo responses (Table 4-2) measured on the day of embryo recovery did not differ between groups. Table 4-1. Ovulatory responses (mean ± SEM) to a superovulatory treatment measured the day of detected estrus Follicular Aspiration + (Treatment) (Control) Number of animals 5 6 Total of follicles > 9 mm 18.4 ±4.4 13.0 ± 3.9 Number of ovulatory follicles 12.0 ± 5.4 7.2 ±4.9

PAGE 116

107 Table 4-1. continued Follicular Aspiration + (Treatment) (Control) Ovulatory follicles (%) 58.7 50.7 Number of nonovulatory follicles 6.4 ± 3.2 5.8 ± 2.9 Table 4-2. Ovulatory response (mean ± SEM) to superovulatory treatment measured the day of embryo recovery Follicular Aspiration + (Treatment) (Control) Number of animals 5 6 Number of corpora lutea (US)* 10.2 ±4.6 7.5 ±4.2 Number of ova & embryos 5.0 ±3.8 6.2 ± 3.5 Recovery rate of ova & embryos/* of CL (%) 35.2 44.2 Recovery rate of ova & embryos/* ovulatory follicles (%) 32.6 46.1 Number of embryos 2.6 ± 1.9 2.7 ± 1.7 Fertilization rate (%) 60.4 34.5 Number of transferable embryos 1.0± 1.6 2.3 ± 1.5 Transferable embryos/total number of ova & embryos (%) 43.3 16.8 Transferable embryos/total number of embryos (%) 50.0 35.1 Number of embryos/number of CL (%) 24.6 20.9 Transferable embryos/number of CL (%) 14.9 3.4 * US= ultrasound Discussion Growth of the first wave DF is one of the most predictable events in ovarian follicular dynamics during the estrous cycle in cattle (Badinga et al., 1992; Fortune, 1993). Demise of the first wave DF at midcycle is due to the negative feedback effect of P4 from the corpus luteum on LH secretion (Savio et al., 1993a; Kinder et

PAGE 117

108 al., 1996). LH secretion during the estrous cycle in cattle is characterized by LH pulses of high frequency and low amplitude during proestrus and metestrus and low frequency, high amplitude pulses during diestrus (Rahe et al., 1 980). Progesterone and E2 play an important role in regulation of pituitary LH secretion. Under a low progestin treatment in the absence of a CL, the pattern of LH secretion is similar to that during the follicular phase of the estrous cycle in the cow (Kinder et al., 1996). In such a low progestin environment and an increased LH pulse frequency, the first wave DF continues to grow and suppresses growth of other follicles (Savio et al., 1993a; 1993b). In the present study, induction of a low progestin environment resulted in persistence and increased size of the DF from days 6 to 16, which is in agreement with others (Sirois and Fortune, 1990; Savio et al., 1993b). Prolonging the period of follicular dominance is an alternative model to study whether a DF can alter induced follicular development in response to exogenous FSH. The phenomenon of dominance is central to understanding folliculogenesis since it suggests that some follicles survive in a milieu suppressive to growth of other follicles. In addition, dominance is related to the success of superovulation, both in humans and domestic animals (Fortune, 1993). There are numerous reports of superovulation programs in cattle with vanable results (Elsden et al., 1978; Murphy et al., 1984; McGowan et al., 1985; Lerneret al., 1986; Pawlyshyn et al., 1986). The ovarian response to a superovulation program is determined by many factors such as ovarian responsiveness of donors, fertilization rate, embryo viability, factors related to physiological status of the animal such as pregnancy, and

PAGE 118

109 management (convenience and cost effectiveness of superovulation protocols; Armstrong, 1993). Ovarian status at the time of superovulatory treatment has been postulated as a major factor determining ovarian response (Monniaux et al., 1983). The increase in number of class 2 follicles at day 2 after aspiration of the active-FWPDF of the present study indicates that the FWPDF suppressed growth of follicles. This suppressive influence may have been through secretion of some factors, such as estradiol and/or inhibin, that deprived follicles of gonadotropin support critical for their further development (Fortune, 1993). Recruitment was attenuated in the presence of the active-FWPDF as shown by a lower number of class 2 follicles in the control group until day 14 which was 2 days after initiation of FSH treatment. Similar results were reported by Guilbault et al. (1991) when they superovulated heifers in the presence of a DF. At day 14, recruitment was evident in the control group, as the number of class 2 follicles increased in association with a decrease in number of class 1 follicles. This process occurred 2 days earlier in the treatment group (day 12) in which the active-FWPDF was aspirated on day 10. Likewise number of class 3 follicles increased earlier in the treatment group without an activeFWPDF. However, by day 1 7 the number of class 3 follicles was the same for the two groups indicating that exogenous FSH eventually overrides the inhibitory influence of the active-FWPDF on the ovaries of the control group. Nevertheless, accumulated follicle diameter was higher for the aspirated group than for the control group with the active-FWPDF. This is due likely to earlier occurrence and longer period for growth in the treatment group.

PAGE 119

110 Initiation of the superovulatory regime in the absence of a DF has been shown to improve the ovarian response to FSH (Grasso et ai. , 1 989; Guilbault et al., 1991; Savio et al., 1991; Huhtinen et al., 1992; Bungartz and Niemann, 1994; Wolfsdorf et al., 1997). Hov\/ever, Gray et al. (1992) found no difference in yield of transferable or total embryos when FSH treatment began during the period of morphological regression of the DF versus a control group in which superovulatory treatment began on day 1 0 when DF was morphologically dominant. Ovulation rate decreased when FSH was given in the presence of a DF resulting in fewer embryos recovered when compared with initiation of treatment when a DF is not present (Guilbault etal., 1991; Huhtinen et al., 1992). This beneficial effect was not shown in the present experiment where none of the variables measured on the day of estrus or the day of embryo recovery were different. Likewise follicular development, ovulation rate, and embryo recovery rate did not differ when first wave DF was ovulated with hCG and superovulation treatment started in the absence of a DF (Rajamahendran and Calder, 1993). Stock et al. (1993) reported no differences in number of ovulatory-size follicles, ovulation rates, or yield of total or transferable embryos between heifers superovulated in the absence or the presence of a DF on day 6. However, in this latter experiment, 8 of the 1 0 animals in the group with the DF present ovulated during FSH treatment. In the present experiment, the active-FWPDF was present in the ovary of all control cows during FSH-P treatment and all FWPDFs ovulated after estrus. In both the present study and that of Wehrman et al. (1996), presence and duration of a DF were controlled experimentally to evaluate effects of a DF during FSH treatment.

PAGE 120

111 Transvaginal aspiration of the active-FWPDF demonstrated that factors from the follicle exert inhibitory effects on follicular growth at the ovarian level. Immediately upon aspiration of the active-FWPDF, changes in number of class 1 and class 2 follicles occurred that were indicative of follicular recruitment prior to initiation of FSH treatment. Presence of an active-FWPDF during FSH-P treatment decreased the response to exogenous FSH-P during the first days of treatment compared to the treatment group with no FWPDF. With continuous administration (4 days) of FSH-P, the inhibitory effect of the active-FWPDF was overridden resulting in the same number of class 3 follicles at estrus as well as the same ovulatory and embryo responses. Wehrman et al. (1996) reported a reduced number of CL, total ova and transferable embryos when the FSH treatment was given 5 days after insertion of a 0.5 PRID, but not when FSH was started 2 or 8 days after insertion. In the present experiment, the norgestomet implant was inserted on day 6 and FSH treatment started 4 days later. It has been shown that passive immunization of ewes against steroid freefollicular fluid or against synthetic inhibin peptides increased ovulation rates twoto four-fold (O'Shea et al., 1991; Schanbacher et al., 1991). Ovulation rate and prolificacy In the cow are more difficult to manipulate by inhibin vaccination (O'Shea et al., 1994). However, there are reports that immunization of heifers or cows against steroid-free follicular fluid increases ovulation rate (Price et al., 1987; Glencross et al., 1992; Morris et al., 1993; Scanlon et al., 1993) and number of transferable embryos (Alvarez et al., 1997). Thus immunoneutralization of inhibin increases ovarian follicular development.

PAGE 121

112 Follicular dominance appears to be controlled by a number of mechanisms acting in concert. These include alterations in peripheral FSH concentrations in response to Ej and inhibin secreted by the DF, as well as the possible production of local ovahan factors that inhibit development of subordinate follicles (Campbell et al., 1995; Armstrong and Webb, 1997). Two hypotheses have been advanced to explain how the DF exerts dominance. The DF could cause regression of subordinate follicles indirectly, via a negative feedback on FSH secretion (Fortune, 1 994). Alternatively, the DF may secrete a factor that directly impairs further growth and development of subordinate follicles (Ko et al., 1991; Adams et al., 1992a; Fortune, 1994). In monotocous species, such a factor would clearly have to be endochne in nature, since it would need to inhibit recruitment and induce regression of subordinate follicles on both ovaries. Besides the peripheral action of inhibin on FSH secretion, there is strong evidence supporting regulatory actions of inhibin-like molecules within the ovary (Hsueh et al., 1987; Hillier et a!., 1991a; 1991b; Wood ruff eta I., 1990; Schneyeret al., 1991). Satoetal. (1982) suggested that an inhibin-like substance inhibits FSH action at the ovarian level, through binding to the FSH receptor on the granulosa cell. More recently, a report by Schneyer et al. (1991) indicated that inhibin a-subunits bind to FSH receptor sites and inhibit FSH bioactivity in granulosa cell cultures. Neither number of CL nor plasma on the day of embryo recovery differed between groups. Higher concentrations of in plasma during days 8, 9 and 10 of the estrous cycle before aspiration of the FWPDF were due to variability among cows of the control group being greater than cows of the treatment group prior to

PAGE 122

113 aspiration of the FWPDF on those days (P < 0.01). This was due to eight of nine cows in the control group having an increase in plasma of > 5 pg/ml, whereas only five of eight cows in the treatment group had this increase. On days 11, 12, and 1 3, the difference between treatments in plasma was due to presence of an active-FWPDF in the control group. Larger class 3 follicles reflected by a higher accumulated follicle diameter for this follicle class in the treatment group on days 16 and 17 was not associated with higher concentrations of Ej in plasma. Implications Initiation of a superovulation treatment in the absence of an active-FWPDF increased the number of class 2 follicles (6 to 9 mm) present on the ovary one day after beginning FSH-P injections. This response was associated with an increase in number of class 3 follicles (> 10 mm) on the last day of FSH-P injection. Increased follicle recruitment following removal of the FWPDF is an indication that presence of an active-FWPDF had an inhibitory effect on follicular responses to exogenous FSH. However, this enhanced ovarian stimulatory effect was not present at the time of estrus or embryo collection. Thus inhibition exerted at the ovarian level appears to be over-ridden by continuous exogenous exposure to FSHP. Additional research is required to investigate the effects of inhibin or inhibin forms present in follicular fluid of dominant follicles on granulosa cells in vitro to elucidate the role of inhibins at the local level within the follicle. To study these effects, it is necessary to develop an estrogenic and FSH responsive test system

PAGE 123

114 comprised of bovine granulosa cells in culture. Development of such a system is the focus of Chapter 5.

PAGE 124

CHAPTER 5 IN VITRO SECRETION OF ESTRADIOL BY BOVINE ANTRAL GRANULOSA CELLS Introduction Follicle stimulating hormone (FSH) and luteinizing hormone (LH) are essential for development and maintenance of granulosa (GC) and theca cells (TC) in ovarian follicles in mammals. Secretion of FSH and LH is regulated in a positive manner by gonadotropin releasing hormone or inhibited by ovanan hormones, such as steroids and inhibin (Taya et al., 1996). Ovarian follicle GC are epithelial cells that produce E2, upon proliferation account for the majority of follicle cell expansion, and support the developing oocyte (Skinner and Coffey, 1988). Theca cells are mesenchymal cells that surround the follicle, which contribute to its expansion and also produce steroid hormones such as androgens (Skinner and Coffey, 1988). Granulosa cells have an integral role in the maintenance and control of ovarian function through the biosynthesis of estrogens and progesterone. Therefore, regulation of GC function influences both local ovarian function and the endocrine status of the female (Skinner and Osteen, 1988). In most mammalian species, TC synthesize androgens from progestins but lack aromatase activity, whereas GC cannot convert progestins to androgens but can aromatize androgens to estrogens (Berndtson et al., 1995). In the bovine follicle, the major steroid 115

PAGE 125

116 product of GC is pregnenolone (Greenwald and Roy, 1994), which is apparently exported to the TC for ultinriate conversion to androgens (Fortune, 1986; Fortune and Quirk, 1 988; Roberts and Skinner, 1 990), and the androgens are utilized by the GC to produce (Fortune, 1986; Fortune and Quirk, 1988; Roberts and Skinner, 1990). Results from Fortune (1986) and Roberts and Skinner (1990) support a potential role for Ej produced by GC, in the enhancement of thecal pregnenolone and androstenedione production, but inhibits P4 secretion, by inhibiting the conversion of pregnenolone to P4 by 3B-hydroxysteroid dehydrogenase (Fortune 1986). Estradiol thereby increases the use of the pathway (Fortune, 1986). A large variety of membranous and intracellular receptors have been identified in GC at different stages of differentiation. For example, peptide hormones, neurotransmitters, steroid hormones, growth factors, and extracellular matrix components interact in a complex manner to control biochemical and structural differentiation of GC (Amsterdam and Rotmensch, 1987). Cultures of mammalian GC have been demonstrated by several investigators (rat: Dorrington and Armstrong, 1979; pig: Baranao and Hammond, 1985; Barano and Hammond, 1985; Hylka etal., 1989; and human: Schipper et al., 1993; sheep: Campbell etal., 1996; cattle: Saumande, 1991; Spicer and Alpizar, 1994; Berndtson et al., 1995; Rouillier et al., 1996; Gutierrez et al., 1997). Reports regarding effects of FSH on bovine GC cultured in vitro are controversial with results depending upon type of GC cultured (antral versus mural GC), conditioned medium utilized (serum-free media versus media containing fetal calf serum); dose of FSH utilized for cell stimulation (low vs high doses); type of

PAGE 126

117 follicle (small vs large) as a source of GC, treatment used to obtain the follicles (use of FSH-P vs PMSG as stimulatory hormones for follicular growth), and time in culture (short term vs long term cultures). Saumande (1991) reported that low doses of FSH (2 ng/ml) were able to stimulate the production of E2 by bovine antral GC in a long-term culture with a defined serum-free medium. Berndtson et al. (1995) using serum-free and serum-containing media showed that mural GC were able to respond to very low doses of FSH (1 and 2 ng/ml) and stimulate E2 production, whereas higher doses ( 8 1 28 ng/ml) accentuated the natural tendency of GC to luteinize in vitro and secrete progesterone (PJ. Rouiilier et al. (1996) demonstrated that mural and antral bovine GC cultured in a completely defined medium responded morphologically and functionally to FSH in a long-term culture. However, FSHand LH-induced production of and P4 productions was considerably lower in mural than in antral bovine GC. It has been demonstrated that FSH (Kuran et al., 1995), LH and equine chorionic gonadotropin (eCG; Kuran et al., 1996) stimulated differentiation and P4 production by mural (Kuran et al., 1995) and antral GC (Kuran et al., 1996) cultured in serum-free medium. More recently, Gutierrez et al. (1 997) showed that a serum-free culture system is suitable for mural GC culture. Production of Ej can be induced and/or maintained with cells remaining responsive to FSH at physiological concentrations to increase proliferation and estrogen secretion. The objective of this experiment was to study the effect of low doses of FSH on induction of Ej secretion by antral bovine GC in long term culture.

PAGE 127

118 Materials and Methods Granulosa Cell Isolation Cycling Holstein cows (n=3) were synchronized with a GnRH injection (Buserelin: 8 |jg i.m.; Receptal®, Hoechst-Roussel Agri Vet, Sommerville, NJ) followed 7 days later with an injection of prostaglandin (PG) Fjalpha (25 mg, i.m.; Lutalyse, Pharmacia & Upjohn Co., Kalamazoo, Ml; Figure 5-1). sy^-FOslZflalCN Fe^CD la^inH P(3F2a* Fa=2a* 5 6 1 ETRJS • C^e 25nig **Day65: 15rng /S6RRAI1CN CFDF 13\yBoil +13\«in 13\«out 1 10 12 13 14 15 SUFffBTMJLAnCN FSH(nng) PM6 5 5 3 3 Figure 5-1. Experimental sequence followed to obtain superstimulated ovaries to harvest bovine antral granulosa cells for culture. On day 6 after estrus (day 0= day of estrus) cows received a Norgestomet ear implant (6 mg norgestomet implant without the injection of the norgestomet/valerate solution; Syncromate; Sanofi Animal Health, Inc., Overland Park, KS) and were given two injections of PGFjalpha (25 mg, i.m., AM and 15 mg, Jl

PAGE 128

119 i.m., PM) to create a low progestin environment for the development of a persistent dominant follicle (Savio et a!., 1993). The implant was left in place for 6 days and then replaced by a new implant on day 12 when FSH treatment was initiated. On day 10, the persistent dominant follicle was aspirated via transvaginal ultrasound guided aspiration using the Aloka Echo-Camera SSD-500 (Aloka Co., LTD, Japan) equipped with a 5.0 MHz transvaginal convex array transducer and aspiration needle guide. Two days later, on day 12, a treatment of FSH (FSH-P™; ScheringPlough, Animal Health Kenilworth, NJ) was initiated. Twenty eight mg were given i.m. at 12 h intervals in a decreasing regime over a 3-day period (a.m./p.m. = 6/6, 5/5 and 3/3 mg). On day 15 cows were slaughtered, ovaries were collected and placed immediately in stehle saline solution (.9% NaCI; at 37°C), containing streptomycin (2 mg/ml) and fungizone (0.1 mg/ml), and transported to the laboratory. The rationale for the use of this follicle induction system is as follows: with creation of a persistent DF, growth of small follicles was suppressed; after aspiration of the persistent DF or removal of the ovary bearing the DF, there is an endogenous surge of FSH (Badinga et al., 1992) that promotes recruitment of new follicles. The FSH-P treatment was begun 2 days after aspiration of the DF to supplement the endogenous FSH recruitment of follicles and to stimulate growth to a desirable follicle size of > 9 mm. Ovaries were trimmed and washed twice with saline solution containing antibiotics and antimycotics. Ovaries were then wiped off with gauze containing 70% ethanol and placed in saline solution until GC were harvested. Antral GC were collected as described by Saumande (1 991 ) except that Menezo Bj medium (Laboratoire C. C. D., France) was used instead of INRA F,

PAGE 129

120 medium to collect GC from follicles. Consecutive follicles > 9 mm in diameter were punctured with a blunt 1 8-gauge needle attached to a 5-ml syringe and slowly rinsed 10 times with 0.5 ml of Menezo medium supplemented with heparin (100 lU/ml; Sigma, St. Louis , MO). These follicles were Ej-active based on follicular fluid concentrations (£2= 351 ng/ml). Ireland and Roche (1983) designated active follicles to have > 141 ± 35 ng/ml in follicular fluid. Granulosa cell suspension was centrifuged (300 x g for 7 min), the pellet resuspended in 0.5 ml of Menezo medium supplemented with ethylene glycol tetraacetic acid (EGTA; 2.47 mg/ml; Sigma, St. Louis, MO) and maintained in an incubator for 15 min at37°C. Following centrifugation (300 x g for 7 min), cells were resuspended in Menezo B2 medium. Cell viability (62.8%) was determined by trypan blue exclusion (Sigma, St. Louis, MO; .4% final concentration) using a hemocytometer. Cell suspension was adjusted in Menezo Bj medium at a concentration of 3x10^ cells in 50 pi. Granulosa Cell Culture Granulosa cells were cultured for 48, 72 or 96 h in 48-well culture plates (Costar, Cambridge, MA), with each one cm^ well precoated with fibronectin (3 pg in 200 Ml; Sigma, St. Louis, MO) for 30-60 min at 37°C in Ham's F-12 (Sigma, St. Louis, MO). Cells were plated at a density of 3x10^ cells/well (50 |jl), and 250 \i\ of medium added to complete a 500 pi volume/well. Ham's F-12 was buffered with NaHCOj (1 .176 mg/ml) and Hepes (20 mM; Sigma, St. Louis, MO), pH 7.4, and was supplemented with insulin (50 ng/ml; Sigma, St. Louis, MO), androstenedione (10"^ M; Steraloids. Inc., Wilton, NH), human transferrin (partially iron-loaded, 10 |jg/ml;

PAGE 130

121 Sigma, St. Louis, MO), ascorbic acid (17.6 \ig/m\; Sigma, St. Louis, MO), gentamicin (20 |jg/ml; Sigma, St. Louis, MO), nystatin (4 pg/mi; Sigma, St. Louis, MO) and no FSH or 0.5 ng/ml of respective FSH's. Medium was sterilized using a 0.22 |jm filter (Nalgene®, Nalge Nunc International, Rochester, NY). The following FSH's were evaluated: 1) porcine FSH (prepared by Dr. Y. Combarnous [CY-FSH], INRA Nouzilly, France; batch CY 1737 III; FSH activity 41 x NIH FSH PI, LH activity < .5%); 2) porcine FSH (prepared by Dr. J. F. Beckers [B-FSH], University of Liege, Belgium, FSH activity 1 0 pg of pure FSH correspond to 1 mg Armour unit or USDA SI, LH content < 1%), and 3) ovine FSH (USDA-oFSH-19 SIAFP-RP-2 prepared by Dr. A. F. Parlow [USDA-FSH], Pituitary Hormones and Antisera Center, Torrance, CA, FSH activity 94 x NIH-oFSH-SI, LH activity .025 x NIH-0LH-SI in 10 mg of BSA). The priming dose of 0.5 ng/ml was used in order to optimize the in vitro secretion of Ej (Rouillier et al., 1997) and, in conjunction with insulin, to improve the maintenance of cell viability (Saumande, 1 991 ). Cells were cultured in a humidified incubator with an atmosphere of 5% CO2 and 95% air at 37X. Culture medium was collected and replaced daily with 500 pi of fresh medium. At 24, 48 or 72 h of culture, cells were treated with different concentrations of the respective FSHs: 0, 0.5, 1, 2, 4, 6, 8, 10 and 50 ng/ml (3 replicates per dose of FSH). Granulosa Cell Number The number of viable cells at 48, 72 and 96 h was measured using the supravital dye neutral red (Sigma, St. Louis, MO) which selectively targets lysosomes in living cells and detects cell viability (Borenfreund and Puerner, 1984).

PAGE 131

122 At the end of the culture period, spent medium was replaced with 500 pl/well of culture medium containing 25 \^g of neutral red, and plates were then incubated for 3 h at 37°C. After incubation, the neutral red solution was removed from the wells by inversion and replaced with 500 pl/well of a solution of formol-calcium (4% formaldehyde, 1% calcium chloride in distilled water) to remove unincorporated neutral red and enhance attachment of cells to the substratum. After 3 min which includes addition of formol-calcium and centrifugation (300 x g), the formol-calcium was removed and 500 pl/well of acetic acid-ethanol solution (1 ml glacial acetic acid in 100 ml of 50% ethanol) added to the wells to extract the neutral red. The plates were allowed to sit at room temperature for 15 min, and then shaken briefly to distribute the dye evenly. Absorbance of the solution was determined at 540 nm using a Beckman DU-20 Spectrophotometer (Beckman Instruments, Inc., Irvine, CA). The readings were averaged and compared with a standard curve generated on the first day of culture. For a standard curve 0, 5x1 0^ 1x1 0^ 2x1 0^ 3x1 0^ 4x1 0^ , and 5x10^ cells (n= 3/cell concentration) were incubated for 3 h at 37°C in 12x75 mm borosilicate tubes in presence of 500 pi of neutral red dye. At the end of the incubation, tubes were centrifuged for 7 min at 300 x g, supernatant poured off slowly and 500 pi of formol-calcium added. During a 3 min period, tubes were centrifuged again at 300 x g, supernatant poured off slowly and 500 pi of acetic acid-ethanol added. The interval between addition of formol-calcium and completion of centrifugation must be performed within a 3 min period. Tubes were incubated at room temperature for 15 min, briefly shaken and absorbance read

PAGE 132

123 using a Beckman DU-20 Spectrophotometer. The relationship between number of cells and absorbance was linear (y= -.02348 + 3 x 1 0"^ X; where y= absorbance and X= number of cells). Regression equation was inverted (y = 0.782 x 10" + 33.33 xlO ^ X; where y = number of cells, and X = absorbance; r^= .96), and the number of cells per well in culture plates was estimated from the resulting linear regression equation. Radioimmunoassays Concentrations of E2 and P4 in unextracted cultured medium were determined with assays validated as described below. Unextracted media samples (100 |jl) were assayed in duplicate for E2 and P4. Medium (100 pi) was added to standard curve tubes to mimic the conditions of sample tubes. Estradiol-17B standards were made by serial dilutions from stock solution of 1 ,3,5(1 0)-estratrien3,17B-diol (Steraloids, Inc., Wilton, NH) diluted in benzene. Standard solutions were diluted in phosphate saline buffer with .1% of gelatin (PBSg), pH 7.5 at final concentrations of 0.039, 0.078, 0.156, 0.312, 0.625, 1.25, 2.5 and 5 ng/100 pi. A 100 Ml aliquot of estradiol, [6,7-'H(N)] (50.0 Ci/mmol; DuPont NEN, Wilmington, DE) at approximately 65,000 dpm diluted in PBSg was added to all tubes. Antiserum at 1:20,000 dilution (100 |jl; Gift from Dr. Lars Edqvist, Uppsala, Sweden) was added to give a final volume of 400 pi. After overnight incubation at 4°C, free and bound fractions were separated by 500 \i\ of a charcoal-dextran solution (250 mg/25 mg in 100 ml of PBSg; 15 min incubation). Following centrifugation (15 min) at 2,619 x g, supernatant was poured into scintillation vials, and radioactivity counted. Mean

PAGE 133

124 (± SEM) concentrations of in 50 and 1 00 [i\ of a pooled sample diluted 1:5,1:10 and non diluted were 29.9 ± 1 .3, 29.3 ± 1 .3, 29.1 ± 1 .3 ng/ml for 50 [i\ of the pool, and 27.1 ± 1.3, 26.8 ±1.3 and 24.6 ±1.3 ng/ml for 100 \i\ of the pool, respectively. Differences in concentrations of E2 (ng/ml) because of sample volume and dilution were not detected. The sensitivity of the assay was 0.39 ng/ml. Intraand interassay coefficients of variation were 9.9% and 4.9%, respectively. Progesterone standards were made by serial dilutions from a stock solution of 4-pregnen-3,20-dione (Steraloids, Inc., Wilton, NH) diluted in benzene. Standard solutions were diluted in phosphate saline buffer with . 1 % of gelatin, pH 7.5 at final concentrations of 0.31, 0.62, 1.25, 2.5, 5, 10, 15, 20 and 25 ng/100 A 100 \i\ aliquot of [1a,2a(n)-^H]Progesterone (49.0 Ci/mmol; Amersham International pic, Arlington Heights, IL) at approximately 75,000 dpm diluted in PBSg, pH 7.5 was added to all tubes. Antiserum at a 1:15,000 dilution (100 Gift from Dr. Magaly Manzo, Universidad Central de Venezuela; Knickerbocker et al., 1986) and 300 pi of PBSg were added to give a final volume of 700 pi. After overnight incubation at 4°C, free and bound fractions were separated by 500 pi of a charcoal-dextran (500 mg/50 mg in 100 ml of PBSg; 15 min incubation) solution. Following centrifugation (15 min) at 2,619 x g, supernatant was poured into scintillation vials, and counted. Mean (± SEM) concentrations of P4 in 25, 50 and 1 00 pi of a pooled sample diluted 1 :5, 1 :10 and non diluted were 95.3 ± 5.5, 101 .1 ± 5.5, and 94.0 ± 5.5 ng/ml for 25 pi of pooled sample, 117.8 ± 6.1, 105.2 ± 5.5, and 104.1 ± 5.5 ng/ml in 50 pi of pooled sample, 1 17.5 ± 5.5, 104.8 ± 5.5 and 95.3 ± 5.5 ng/ml in 1 00 pi of a pooled sample, respectively. Differences in concentrations of P4 due to sample volume and

PAGE 134

125 dilution were not detected. Sensitivity of the assay was 3.12 ng/ml. Intraand interassay coefficients of variation were 3.9% and 10.8%, respectively. Gel Electrophoresis Samples of the different FSH types (CY-FSH: 600, 700 and 800 ng/gel lane; B-FSH: 600 ng/gel lane and USDA-FSH: 500 ng/gel lane) were mixed in sample loading buffer containing 4% of sodium dodecylsulphate (SDS; BDH Laboratory Supplies, Poole, England) without (nonreducing conditions) 3-mercaptoethanol, boiled for 5 min and subjected to one dimensional 15% SDS-PAGE (Laemmli, 1970) using a Bio-Rad mini gel apparatus (Mini-Protean II, Vertical Electrophoresis Cell, Bio-Rad, Hercules, CA; 150 volts, 100 mA, 1h). After electrophoresis, the gel was soaked in 50% methanol (HPLC grade. Fisher Scientific) for 1 h. Then, the gel was stained with silver (Wray et al., 1981) for 15 min in a solution containing 20% silver nitrate, 0.36% sodium hydroxide and 14.8 M ammonium hydroxide. After staining, the gel was washed with double distilled water (2x10 min). The gel was soaked in a developing solution of 1% of citric acid and 38% formaldehyde until bands appear (usually between 10 to 15 min). To stop development, the gel was washed with water and placed in a solution of 50% methanol, 10% glacial acetic acid (Fisher Scientific) and 40% double distilled water. To determine molecular mass of FSH subunits bands, broad-range molecular mass protein standards (7.7 to 204 kDa, Bio-Rad Kaleidoscope prestained molecular weight standards) were loaded onto a different gel and subjected to 15% SDS-PAGE, but not stained with silver.

PAGE 135

126 Statistical Analyses Data were analyzed by least squares analysis of variance using the General Linear Models Procedure of the Statistical Analysis System (SAS, 1988). Data are presented as least squares means ± SEM of triplicate measurements. Basal secretion and secretion of and P4 are expressed as ng/1 0 * live cells or as ng/well at the end of the culture period. Basal secretion is defined as the secretion of steroids giving by the priming dose (0.5 ng of FSH/ml) or zero dose of FSH during the different culture periods. To analyze basal secretions of Ej and P4, the mathematical model included type of FSH (no FSH, CY-FSH, B-FSH, USDA-FSH), time in culture, and the interaction type of FSH by time. The mathematical model used to analyze the secretion of Ej and P4 after FSH treatment and the number of cells at the end of the culture included type of FSH, time in culture (48, 72 and 96 h), dose of FSH, type of FSH by time, type of FSH by dose, time by dose, type of FSH by time by dose interactions as sources of variation. Orthogonal contrasts were used to evaluate the response among the different sources of FSH. Also sequential contrasts (Littell, et al. , 1 991 ) were used to evaluate the response among doses (0 vs other doses, 0.5 vs 1 , 1 vs 2, 2 vs 4, 4 vs 6, 6 vs 8, 8 vs 1 0, and 1 0 vs 50) for the different FSH sources tested. Results Time Course of Basal Hormone Secretion Estradiol. There was a significant effect of time (P < 0.001) on basal E2 secretion

PAGE 136

127 (Figure 5-2A). Estradiol decreased from 24 h (14.0 ± 0.2 ng/well) to 48 h (5.0 ± 0.1 ng/well) to 72 h (0.9 ± 0.1 ng/well) and to 96 h (0.3 ± 0.3 ng/well) for all types of FSH. After 24 h of culture, basal secretion of was the same (P > 0.8) for cells that received no FSH (14.7 ± 0.6 ng/well) and cells primed with 0.5 ng/ml CY-FSH (14.1 ± .2 ng/well), whereas cells primed with B-FSH or USDA-FSH tended (P < 0.1) to secrete less Ej than CY-FSH and no FSH-primed cells, but Eg did not differ (P > 0.4) between the B-FSH(13.8 ± 0.2 ng/well) and USDA-FSH-primed cells (13.4 ± 0.2 ng/well; Figure 5-2A). After 48 h of culture, basal Eg secretion decreased for all FSH types, but basal secretion was still higher for no FSH and CY-FSH (5.6 ± 0.4 ng/well and 6.9 ± 0.2 ng/well, respectively) than USDA-FSH and B-FSH (4.3 ± 0.2 ng/well and 3.3 ± 0.2 ng/well, respectively; Figure 5-2A). At 72 h of culture, basal Eg secretion continued to decrease, but was higher for cells that received no FSH or CY-FSH (1 .6 ± 0.4 ng/well and 1 .2 ± 0.2 ng/well, respectively; Figure 5-2A), than for cells that received USDA-FSH or B-FSH (0.7 ± 0.2 ng/well and 0.4 ± 0.2 ng/well, respectively; Figure 5-2A). The same trend was present at 96 h (0.6 ± 0.6 ng/well; 0.4 ± 0.5 ng/well; 0.2 ± 0.7 ng/well and 0.3 ± 0.6 ng/well for no FSH, CY-FSH, USDA-FSH and B-FSH, respectively; Figure 5-2A), but the differences were not significant statistically. When basal secretion of Ej was expressed as ng of Ej/IO ^ live cells, time effect also was significant (P < 0.001) with Eg secretion decreasing from 48 h to 72 h to 96 h for all FSHs (9.4 ± 0.4 ng/10 ^ live cells > 2.2 ± 0.4 ng/10 ^ live cells > 0.6 ± 0.4 ng/1 0 * live cells, respectively). CY-FSH stimulated the highest basal secretion of E2 at 48 h (17.6 ± 0.7 ng/10 ^ live cells; P < 0.001; Figure 5-2B) followed by no

PAGE 137

128 FSH-primed cells (9.1 ± 0.8 ng/10 ^ live cells); basal production of B-FSH (6.8 ± 0.8 ng/10 ^ live cells) or USDA-FSH-primed cells (3.9 ± 1 .0 ng/10 ^ live cells) were lower and differed (P < 0.001) from each other. Time in Culture (h) Time in Culture (h) 24 48 72 96 48 72 96 Time in Culture (h) Time in Culture (h) I CY-FSH Q B-FSH Q USDA-FSH ^ No FSH Figure 5-2. Basal secretion of estradioi-17(S and progesterone expressed as ng/well (A and C, respectively) or as ng/10* live cells (B and D, respectively) by bovine antral granulosa cells primed w^ith 0.5 ng/ml of CY-FSH, B-FSH, USDA-FSH or no FSH during 96 h of culture period. Different letters within the same time indicate differences at P< 0.001.

PAGE 138

129 Estradiol basal secretion per 1 0 ^ live cells at 72 h showed the same pattern, with highest production by CY-FSH (4.9 ± 0.7 ng/10 ^ live cells; P < 0.01) and lower E2 production for no FSH-pnmed cells (1 .9 ± 0.8 ng/1 0 ^ live cells), B-FSH (1 .3 ± 0.8 ng/10 ^ live cells) and USDA-FSH (0.7 ±1.0 ng/10 ^ live cells; Figure 5-2B). During the last 24 h of culture, at 96 h, E2 secretion was low for all FSH-primed cells (1 .3 ± 0.8 ng/1 0 ^ live cells; 0.7 ± 1 .0 ng/1 0 ^ live cells; 0.3 ± 0.8 ng/1 0 ^ live cells and 0.3 ±1.0 ng/10 ' live cells, for no FSH, CY-FSH, B-FSH and USDA-FSH, respectively; Figure 5-2B) and did not differ between treatments. Progesterone . Basal P4 secretion by bovine antral GC was greater than basal E2 secretion during the 96 h of culture. There was a significant effect of time (P < 0.001) on basal P^ secretion. An FSH by time interaction (P < 0.001) was detected for P4 basal secretion per well or per 1 0 ^ cells (Figure 5-2D). Basal secretion of P4 declined with time in culture for CY-FSH-primed cells. In contrast, basal P4 secretion increased with time in culture for cells primed with B-FSH or USDA-FSH (Figure 5-2C). Basal secretion of P4 by B-FSHand USDA-FSH-phmed cells increased from 24 h (37.7 ± 1 .7 ng/well and 41 .8 ± 1 .9 ng/well, respectively; Figure 5-2C) to 48 h (41 .9 ± 1 .5 ng/well and 53.5 ± 1 .8 ng/well, respectively; Figure 5-2C) to 72 h (55.2 ± 1.6 ng/well and 82.5 ± 1.8 ng/well, respectively; Figure 5-2C). During the last 24 h of culture, basal P4 secretion induced by B-FSH or USDA-FSH decreased somewhat (36.8 ±5.1 ng/well and 66.6 ± 6.2 ng/well, respectively; Figure 5-2C). Basal P4 secretion by cells that did not receive FSH, did not differ (P > 0.1) among different times (24 h: 31.5 ± 5.0 ng/well; 48 h: 23.1 ± 3.6 ng/well; 72 h: 25.2 ± 3.6 ng/well; and 96 h: 31.0 ± 5.1 ng/well; Figure 5-2C).

PAGE 139

130 Basal P4 secretion expressed as ng of P4/IO ^ live cells followed the same pattern. At 48 h the lowest basal secretion was induced by CY-FSH (47.0 ± 17.3 ng/10 * live cells; P < 0.004), followed by cells that received no FSH (55.1 ± 14.2 ng/10 ' live cells; P < 0.1), USDA-FSH (73.3 ± 17.3 ng/10 ' live cells) and B-FSH (95.6 ± 14.2 ng/10 ^ live cells). At 72 h of culture, cells that received CY-FSH had the lowest basal secretion of P^ (29.5 ± 17.3 ng/10 ^ live cells; P < 0.01; Figure 52D) followed by cells that received no FSH (71.9 ± 14.2 ng/10 ^ live cells); B-FSH (1 90. 1 ± 14.2 ng/1 0 ^ live cells), and USDA-FSH (1 84.8 ±17.3 ng/1 0 ^ live cells) had the greatest basal secretion. By 96 h basal secretion of P4 had decreased for all types of FSH (no FSH: 30.7 ± 14.2 ng/10 ' live cells, CY-FSH: 19.4 ± 17.3 ng/10 ' live cells, B-FSH: 141 .7 ± 14.2 ng/10 ^ live cells and USDA-FSH: 54.8 ± 17.3 ng/10 ^ live cells; Figure 5-2D). However, cells primed with CY-FSH had the lowest secretion (P < 0.01). Effect of FSH-stimulated Secretion Estradiol . Overall E2 secretion by antral bovine granulosa cells decreased from 48 h to 72 h to 96 h for each FSH treatment (P < 0.01 ; Figure 5-3A). Cells treated with CY-FSH had a decrease in Ej secretion from 18.2 ± 0.3 ng/10 ^ live cells at 48 h to 8.4 ± 0.3 ng/10 ^ live cells and 3.1 ± 0.4 ng/10 ' live cells at 72 and 96 h, respectively (Figure 5-3A). Cells treated with B-FSH and USDA-FSH secreted less Ej than cells treated with CY-FSH at each time in culture. Estradiol secretion (ng/1 0 ^ live cells) at 48 h did not differ among doses of FSH for B-FSH and USDA-FSH

PAGE 140

131 (Figure 5-3B). However, cells treated with CY-FSH secreted more (P < 0.001) and 1 to 2 ng/ml of CY-FSH stimulated secretion (Figure 5-3B). FSH Dose (ng/ml) FSH Dose (ng/ml) H CY-FSH Q B-FSH [~\ USDA-FSH Figure 5-3. Secretion of estradiol-1 7(3 (ng/1 0^ live cells) by bovine antral granulosa cells over a 96 h of culture after treatment with different doses of CY-FSH, B-FSH and USDA-FSH. Main effect of time was significant at P < 0.001 (A). Estradiol-1 7li secretion during 96 h of culture (A) and in response to doses of FSH at 48 (B). 72 (C) and 96 h (D), respectively. Different letters within the same FSH source indicate differences at P < 0.001 (Figures B, C, and D). At 72 h, E2 secretion was increased by 8 and 10 ng/ml of CY-FSH (10.7 ± 1 .1 ng/1 0^ live cells and 1 2.8 ±1 .0 ng/1 0 ^ live cells; P < 0.01 ; Figure 5-3C), whereas at

PAGE 141

132 50 ng/ml of CY-FSH E2 secretion decreased to basal levels (7.4 ±1.1 ng/10* live cells; P < 0.09; Figure 5-3C). At 72 h and 96 h of culture, B-FSH and USDA-FSH did not stimulate bovine antral granulosa cells to secrete E2 (Figure 5-3, C and D). By 96 h, 6 ng/ml of CY-FSH stimulated E2 secretion (3.4 ± 1 .0 ng/10 ^ live cells; P < 0.01; Figure 5-3D) by bovine antral GC, reaching a maximum secretion at 10 ng/ml of CY-FSH (1 3.1 ± 1 .0 ng/1 0 ^ live cells). However, a dose of 50 ng/ml of CYFSH inhibited E2 secretion (1.8 ± 1.1 ng/10 ^ live cells; P < 0.01). Overall E2 secretion expressed as ng/well (Figure 5-4A) also was affected by time (P < 0.001), with secretion being higher at 48 h (5.8 ± 0.05 ng/well) than at 72 h (1.1 ± 0.05 ng/well) or 96 h (0.3 ± 0.05 ng/well). Cells treated with CY-FSH secreted more E2 at 48 h, 72 and 96 h (8.2 ± 0.1 ng/well, 1.7 ± 0.1 ng/well and 0.6 ± 0.1 ng/well, respectively; P < 0.001; Figure 5-4A) than B-FSH(4.4 ± 0.1 ng/well, 0.8 ± 0.1 ng/well and 0.3 ± 0.1 ng/well at 48 h, 72 h and 96 h, respectively; Figure 5-4A) or USDA-FSH-treated cells (4.8 ± 0.1 ng/well, 0.9 ± 0. 1 ng/well and 0.2 ± 0.1 ng/well, at 48 h 72 h and 96 h, respectively; Figure 5-4A). At 48 h there were differences (P < 0.001; Figure 5-4B) in secretion by granulosa cells stimulated by 8, 10 and 50 ng/ml of CY-FSH. In the case of B-FSH, Ej secretion increased from 3.1 ± 0.3 ng/well at a dose of 0 ng/ml of FSH to 5.1 ± 0.3 ng/well induced by a dose of 6 ng/ml of B-FSH (P < 0.06; Figure 5-4B). USDA-FSH stimulated the secretion of Ej at doses of 10 ng/ml (9.1 ± 0.3 ng/well) and 50 ng/ml (9.1 ± 0.3 ng/well; P < 0.001; Figure 5-4B). At 72 h, there was an increase in Ej secretion when GC were stimulated by 8 ng/ml of CY-FSH (2.1 ± 0.2 ng/well; Figure 5-4C), this increase was sustained when doses of 10 or 50 ng/ml of FSH were added to

PAGE 142

133 the culture (2.6 ± 0.2 ng/well and 2.5 ± 0.2 ng/well, respectively, Figure 5-4C). No differences in E2 secretory response were obtained when different doses of B-FSH were added to the culture (Figure 5-4C), whereas 1 0 ng/ml of USDA-FSH stimulated an increase of secretion by bovine antral GC (1 .8 ± 0.3 ng/well; P < .001 ; Figure 5-4C). FSH Dose (ng/ml) PSH Dose (ng/ml) m CY-FSH Q B-FSH Q USDA-FSH Figure 5-4. Secretion of estradiol (ng/well) by bovine antral granulosa cells over a 96 h of culture in response to different doses of CY-FSH, B-FSH and USDA-FSH. Main effect of time was significant at P < 0.001 (A). Estradiol-1711 secretion during 96 h of culture (A) and in response to doses of FSH at 48 (B), 72 (C) and 96 h (D), respectively. Different letters within the same FSH source indicate differences at P < 0.001 (Figures B, C, and D). At 96 h, E2 production by granulosa cells was stimulated by CY-FSH (P <

PAGE 143

134 0.001) at doses of 4 (0.4 ± 0.1 ng/well), 6 (0.6 ± 0.1 ng/well). 8 (0. 8 ± .1 ng/well), 10 (1.3 ± 0.1 ng/well) and 50 ng/ml (0.6 ± .1 ng/well; Figure 5-4D). Maximal response was detected at 10 ng/ml with a decreased response at 50 ng/ml. There were no differences among doses for B-FSH and USDA-FSH at 96 h of culture (Figure 5-4D). Progesterone . There were no significant differences in total P4 secretion (ng/1 0^ live cells) among different FSH sources (Figure 5-5A) at 48 or 96 h. However, at 72 h B-FSH induced the highest P4 secretion (P < 0.001). At 48, 72 or 96 h, no differences among doses for different FSH sources were detected (Figure 5-5B, C and D). When secretion of P^ was expressed as ng/well an interaction of FSH by dose by time interaction was detected (P < 0.001). There were differences among FSH sources (P < 0.001 ; Figure 5-6A) across times (P < 0.001) when P4 secretion (ng/well) after FSH treatment was examined. There was a decrease (P < 0.001) in P4 secretion induced by CY-FSH from 48 h (34.5 ± 1 .8 ng/well) to 72 h (23.1 ± 1 .8 ng/well), with no changes at 96 h (23.0 ± 1.8 ng/well; Figure 5-6A). For the other two FSH sources, P4 secretion increased (P < 0.001) from 48 h (B-FSH: 53.6 ± 2.2 ng/well and USDA-FSH: 69.4 ± 2.6 ng/well) to 72 h (B-FSH: 1 12.8 ± 2.2 ng/weil and USDA-FSH: 148.0 ± 2.6 ng/well), and decreased (P < 0.001) at 96 h (B-FSH: 72.9 ± 2.2 ng/well and 1 14.3 ± 2.6 ng/well; Figure 5-6A). At 48 h, P4 secretion (ng/well) was the same for all doses of CY-FSH (Figure 5-6B). Although P4 secretion was greater for B-FSH, no B-FSH dose effect was detected, whereas doses of 10 and 50 ng/ml of USDA-FSH induced a response by

PAGE 144

135 by GC (88.7 ± 7.9 ng/well and 100.6 ± 7.9 ng/well, respectively; P < 0.002; Figure 5-6B). 700w ffl600 o = 500o o §4C0 o laoo (D §2D0 (D Sioo0) 0 Q. 48 72 nrTBinCLltuB(h) 96 0 0.5 1 2 4 6 8 10 50 PSHDD8e(ng^rTl) TOOi 0 0.5 1 2 4 6 8 10 50 RSHCbse(n^rTl) 0 0.5 1 2 4 6 8 10 50 FSHCtBe(ngfrTl) CY-PSH BPSH [] USD^ran Figure 5-5. Secretion of progesterone (ng/10* live cells) by bovine antral granulosa cells over 96 h of culture in response to different doses of CY-FSH, B-FSH and USDAFSH. Progesterone secretion during 96 h of culture (A) and in response to doses of FSH at 48 (B), 72 (C) and 96 h (D), respectively.

PAGE 145

136 250 200 '150 a 100o 50 a m 48 72 Time in Culture (h) 96 250 200I P 150c c ~ o £ 100 tn (B C3) O ^ 50 I iiiiii 0 0.5 1 2 4 6 8 10 50 FSH Dose (ng/ml) 250 200 1.150 C o £ 100-1 w « 0. 50J b b X i 1 1 250 d b 200-1 ^ = f 150 c o S 100u) 0) o> o 501 III 1 1 1 1 e 1 0 0.5 1 2 4 6 8 10 50 FSH Dose (ng/ml) 0 0.5 1 2 4 6 8 10 50 FSH Dose (ng/ml) CY-FSH Q B-FSH Q USDA-FSH Figure 5-6. Secretion of progesterone (ng/well) by bovine antral granulosa cells over 96 h of culture in response to different doses of CY-FSH, B-FSH and USDA-FSH. Main effect of time was significant at P < 0.001 (A). Secretion of progesterone during 96 h of culture (A) and in response to doses of FSH at 48 (B), 72 (C) and 96 h (D), respectively. Different letters within the same FSH source indicate differences at P < .001 (Figures B, C, and D). At 72 h, there were no differences in secretion among doses of CY-FSH, except for an increase at 50 ng/ml which stimulated cells to secrete 40.6 ± 5.6 ng/well (P < 0.001; Figure 5-6C). B-FSH-stimulated cells secreted more P4 with secretion further stimulated between 6 to 50 ng/ml and highest secretion of P4 was

PAGE 146

137 stimulated at 50 ng/ml of B-FSH ( 209.5 ± 6.5 ng/well; P < 0.01; Figure 5-6C). Secretion of P^, at 72 h, by cells treated with USDA-FSH was high but no consistent dose effect was detected. At 96 h P4 secretion continued to be lowest with CY-FSH and a slight stimulatory dose effect in P4 secretion was detected between doses of 6 to 50 ng/ml of CY-FSH (Figure 5-6D). Belgium-FSH-treated granulosa cells secreted more P4, and doses of 6, 8, 10 and 50 ng/ml of B-FSH stimulated P4 secretion. Progesterone secretion was greater in granulosa cells treated with USDA-FSH and P4 secretion increased at doses 6 to 50 ng/ml when compared with all other doses (P < 0.05; Figure 5-6D). Effect of FSH on Cell Number Effects of FSH by time interaction were detected on cell number (P < 0.001 ; Figure 5-7). 140 n 1 20 g 1 00 X 80 I (D E 60 4020 0 -1— a 1 48 h 72 h 96 h Tim e in C ulture (h) CY-FSH I I B-FSH j , | USDA-FSH Figure 5-7. Number of bovine antral granulosa cells at the end of 48, 72 and 96 h of culture with CY-FSH, B-FSH and USDAFSH. Different letters indicate differences at P < 0.001 within time.

PAGE 147

138 Basically, GC number remained constant with CY-FSH treatment. However, B-FSH and USDA-FSH treatments increased number of cells maintained in culture. Clearly, number of cells maintained in culture was less than the number seeded (3 X 10^ live cells) with these losses detected at 48 h. This initial loss was attributed to cell loss with change of media at 24 h due to cells not being attached. After this period numberof cells was maintained within each FSH treatment group, and actual number of cells appeared to increase in the USDA-FSH treatment group by 96 h. This apparent proliferation did not appear to occur with the other FSHs. Effect of FSH on Cell Morpholoav After 24 h of culture in presence of 0.5 ng/ml of CY-FSH, granulosa cells maintained their original round shape, and started to form cell aggregates forming clumps (tight aggregates of spherical granulosa cells). Shape of granulosa cells varied according to the type of steroid secreted (Figure 5-8). Estrogenic granulosa cells, treated with CY-FSH at 96 h (Figure 5-8A), were round in shape, whereas non estrogenic granulosa cells, treated with USDA-FSH at 96 h (Figure 5-8B), had an enlarged fibroblast-like appearance. Clumps of granulosa cells were more frequently observed in wells that received CY-FSH and produced more E2 than those that received B-FSH or USDA-FSH. Flattened fibroblast-like cells with processes (Figure 5-8B) were more abundant in wells treated with B-FSH or USDAFSH, and this was most evident in cells treated with high doses of FSH (> 1 0 ng/ml).

PAGE 148

139 Figure 5-8. Morphological appearance of antral bovine granulosa cells after 24 h of treatment with 4 ng/ ml of CY-FSH (top panel; x400) and 1 0 ng/ml of USDA-FSH (bottom panel; x1 00). Cells were primed with 0.5 ng/ml of the respective FSH for 48 h (top panel) and 72 h (bottom panel).

PAGE 149

140 Gel Electrophoresis Figure 5-9 shows the results obtained following SDS-electrophoresis of porcine (CY-FSH and B-FSH) and ovine FSH (USDA-FSH). The gel pattern for CYFSH revealed a doublet band at the 800 ng concentration (lane 2), the upper band being darker than the bottom one. On lane 1 there is a very light band, and on lane 3 (700 ng) the band is more intense than lane 1 . For B-FSH (lane 4) there are two light bands at the same molecular weight level than for CY-FSH (23 kDa). For USDA-FSH (lane 5) the pattern is different. There is a major band at 75 kDa molecular weight which corresponds to BSA present in the preparation, and two more bands at higher (> 75 kDa) molecular weights, then at the same level than the other lanes (around 23 kDa molecular weight) there are two light bands that correspond to FSH. MW 1 2 3 4 5 Figure 5-9. One dimensional SDS-PAGE of porcine (CY-FSH and B-FSH) and ovine FSH (USDA-FSH). Lane 1 to 3 correspond to CY-FSH (600, 800 and 700 ng/gel lane, respectively), lane 4 corresponds to B-FSH (600 ng/gell lane), and lane 5 con-esponds to USDA-FSH (500 ng/gel lane).

PAGE 150

141 Discussion Luteinization of cultured bovine granulosa cells treated with high doses of FSH is a common feature (Berndtson et al., 1995; Kuran et al., 1996). In culture systems that use serum as a supplement, bovine granulosa cells also luteinize spontaneously and lose their ability to produce in response to FSH (Sirard et al., 1991). It has been shown that E2 secretion by granulosa cells is lost after the first day of culture (Skinner and Osteen, 1988). The present study demonstrated that bovine antral GCs from estrogen active follicles (E2= 351 ng/ml) can be stimulated by low doses of CY-FSH to produce Ej in a long-term culture. These results agree with other recent results (Saumande, 1991; Berndtson et al., 1995; Rouillier et al., 1996; Gutierrez et al., 1997). The present study demonstrated that the estrogenic response of bovine antral granulosa cells depended upon the source of FSH used. A quadratic dose response was obtained with CY-FSH at 96 h (Figure 5-3D), whereas the other two FSH sources were unable to stimulate an aromatase/estradiol response (Figure 53D). A 6 ng/ml dose of CY-FSH was able to stimulate E2 secretion by antral granulosa cells with maximum secretion occurring at 10 ng/ml of FSH, and a dose of 50 ng/ml inhibited E2 secretion. Saumande (1991), Berndtson et al. (1995) and Rouillier et al. (1996) reported that the maximal stimulatory dose was between 1 and 2 ng/ml, and doses equal or higher to 10 ng/ml of CY-FSH were inhibitory. Both Saumande (1991), Rouillier et al., (1996) and the present experiment

PAGE 151

142 utilized the same source of FSH (CY-FSH). Probably the difference in doseresponse among the studies was due to the type of follicles used in the experiments. In the present experiment, antral granulosa cells were harvested from follicles of cows injected with 28 mg of FSH-P during 3 days after aspiration of a persistent dominant follicle. In experiments of Saumande (1 991 ) and Rouillier et al. (1 996), granulosa cells were harvested from follicles of PMSG-treated calves. The difference in response to lower or higher doses of FSH could be attributed to differences in half life between PMSG (up to 6 days; Heap and Flint, 1 984) and FSH (about 30 min; Heap and Flint, 1984) and that PMSG may have had a carryover residual effect on the cells. The decrease in Ej secretion from 48 h to 72 h to 96 h (end of the culture period) is in agreement with previous reports (Berndtson etal., 1995; Rouillier etal., 1996) and is due likely to a reduction in activity of aromatase enzyme that has been observed previously in bovine granulosa cells (Skinner and Osteen, 1 988; Rouillier et al., 1996). The shift in the response of antral granulosa cells to different doses of FSH through 96 h of culture period could be due to a change in sensitivity of cells to higher doses of CY-FSH. At 48 h, increased secretion of E2 by bovine antral granulosa cells occurred at low doses of CY-FSH (e.g., 0.5, 1 , and 2 ng/ml). After 72 h of culture, cells were not able to respond to low doses of CY-FSH and required greater doses of FSH (e.g., 6, 8 and 10 ng/ml) to produce Ej. However, the highest dose of 50 ng/ml of CY-FSH was inhibitory, perhaps due to downregulation of the FSH receptor. It is possible that the change in sensitivity is due to a decrease in the

PAGE 152

143 turnover of FSH receptors at the cell membrane level due to internalization of receptors, which increases the requirements of FSH to elicit a response. Different types of FSH (ovine FSH: Langhoutetal., 1991, Spicer and Alpizar, 1994, Berndtson et al., 1995, Kuran et al., 1995, Spicer and Stewart, 1996, Kuran et al., 1996; bovine FSH: Gutierrez et al., 1997; porcine FSH: Saumande, 1991, Rouillier et al., 1 996) have been used to stimulate bovine granulosa cells in culture. In the present experiment, we used ovine FSH (USDA-FSH), and porcine FSH (Band CY-FSH). Only CY-FSH stimulated GC to produce during long term culture. Several explanations could account for this difference. It is possible that there are other proteins present in the preparation. For example, USDA-FSH is provided in a earner of BSA (10 pg of USDA-FSH is lyophylized with 10 mg of BSA); perhaps BSA interfered with the biological activity of FSH in culture. Alternatively, BSA may stimulate luteinization and proliferative responses due to potential presence of growth factors and fatty acids. On the other hand, both CY-FSH and B-FSH are apparently very pure preparations. Perhaps some isomeric forms of FSH present in CY-FSH are different from those in B-FSH, like other glycoprotein hormones, exists as a family of isohormones, which differ in their oligosacharide structures, including the degree of terminal sialylation and/or sulfation (Ulloa-Aguirre et a!., 1 995). Structural differences in these isoforms alter their metabolic clearance rates as well as biological and immunological potencies (Ulloa-Aguirre et al., 1995). It has been shown that the receptor-binding activity/immunoactivity ratio of the intrapituitary isoforms decreases as the pi of the corresponding isoform declines. Studies in the female hamster have demonstrated a close relationship between the

PAGE 153

144 relative sialic acid content of the isoforms and their ability to bind to the receptor (Ulloa-Aguirre et al., 1984). Perhaps these structural differences can account for differences in ability of the different FSH sources used in this experiment to induce E2 secretion by bovine antral granulosa cells. Furthermore, different preparations of FSH may be more labile to the lyophylization process. It has been shown previously that P4 secretion by granulosa cells increases with time in culture (Skinner and Osteen, 1988). This study showed that basal (Figure 5-2, C and D) and total FSH-stimulated secretion (Figure 5-5A) of P4 decreased with time when CY-FSH was used to stimulate the cells, whereas P4 secretion per well increased for Band USDA-FSH. The response of bovine antral granulosa cells to different doses of CY-FSH, in terms of P4 secretion, was not as clear as the response in terms of E2 secretion at the end of the 96 h culture period (Figures 5-3 and 5-5D). Cells treated with CY-FSH had a quadratic response to dose of FSH showing a maximal response at 10 ng/ml with a decrease in E2 secretion at a dose of 50 ng of FSH/ml. No dose effect of CY-FSH, B-FSH or USDA-FSH was detected on P4 secretion. In contrast B-FSH and USDA-FSH failed to stimulate E2 secretion. Our study suggests that the steroidogenic nature of the cells or the degree of differentiated characteristics of luteinization differed between FSH types used. The number of cells at the end of the 96 h culture period was lower for CYFSH than for B-FSH and USDA-FSH treatments. Apparently, the latter two FSH preparations induced the granulosa cells to differentiate into a cell type with a more fibroblastic appearance that secretes P4. Cells treated with USDA-FSH formed

PAGE 154

145 monolayers and assumed a fibroblastic appearance with cell processes whereas cells treated with CY-FSH were tightly packed with a round shape and grew in mutilayered aggregates, as reported previously in the rat (Knecht et al., 1981) and cow (Gutierrez et al., 1997). The dynamics of cellular proliferation in our culture were difficult to quantify since appreciable cell loss occured early (e.g., by 48 h) for all groups and cell numbers at each time were less than cells plated. Round cells typical of CY-FSH would be more prone to be lost with medium change compared to granulosa cells with fibroblastic-like appearance. Increased cell number at 96 h for the USDA-FSH may have represented a proliferative phase. Thus with the experimental approach utilized we were unable to demonstrate a stimulatory effect of FSH on proliferation of bovine granulosa cells in culture. Similar results were reported by Langhout et al. (1991), Gong et al. (1993) and Rouillier et al. (1996). On the other hand, Gutierrez et al. (1997) reported that FSH in the presence of 10 ng/ml of insulin, stimulated proliferation of cells from small, medium and large follicles as measured by cell number at the end of the culture. Similar results have been reported by Campbell et al. (1996) with ovine granulosa cells proliferation in vitro. Skinner and Osteen (1988) reported that granulosa cells isolated from large follicles had a slightly lower plating efficiency and greater loss in cell number between days 3 and 6 of culture. This could be the case in our study where the source of granulosa cells were large follicles (> 9 mm). Implications The present experimental approach demonstrated a differential response of

PAGE 155

146 bovine antral granulosa cells to different sources of FSH, and that there is a shift in the estrogenic response of bovine antral granulosa cells to doses of CY-FSH with tinne in culture. Granulosa cells from bovine ovarian follicles did not proliferate in culture in response to CY-FSH and shape of granulosa cells in culture varied according to the type of steroid secreted. Estrogenic granulosa cells fronn follicles > 9 mm exposed to CY-FSH had a round shape and grev^ in multilayered aggregates or clumps, whereas nonestrogenic granulosa cells had an enlarged fibroblastic-like appearance. Based on these results we can conclude that this bovine antral granulosa cell culture system can be used to study the regulatory factors controlling secretion of Ej. For example effects of dimeric and monomeric forms of inhibin that are present during transitory phases of follicular dominance can be evaluated as to their ability to regulate Ej secretion in granulosa cells. Based upon the results in the present chapter, it is recommended that Ej secretion be evaluated as ng of E2 per 10^ live cells since cell loss appeared to differ between types of FSH that influenced morphogenic appearance in culture. Furthermore, dose effects of CY-FSH on Ej secretion appeared to be more apparent when expressed as ng of E2 per 10^ live cells.

PAGE 156

CHAPTER 6 EFFECT OF BOVINE FOLLICULAR FLUID FROM DAY 5 DOMINANT AND DAY 12 ATRETIC DOMINANT FOLLICLES ON IN VITRO ESTRADI0L-17p SECRETION BY BOVINE ANTRAL GRANULOSA CELLS Introduction Ovarian folliculogenesis is a dynamic process characterized by a marked proliferation and differentiation of the somatic cell components of the follicle and is associated with development of a group of follicles at various stages of maturation from which a species-specific number of follicles are selected for continued growth (Armstrong and Webb, 1997). The developing follicle provides the optimal environment for the maturation of the oocyte in readiness for its subsequent ovulation and fertilization. Gonadotropins are the primary regulators of folliculogenesis. They regulate ovarian folliculogenesis via classic endocrine mechanisms. In addition, intraovarian mechanisms are involved in the control of folliculogenesis (Ireland, 1987; Tonetta and di Zerega, 1989). The concept of intraovarian control of ovarian function arose in the 1970s (Armstrong and Webb, 1997). These ovarian factors regulate both secretion of FSH and follicular responsiveness to gonadotropins (Findlay, 1993). Ovarian follicles produce an array of locally acting peptides and growth factors that interact directly with the same cell type from which they are produced 147

PAGE 157

148 or with other cell types within the developing follicle to stimulate or attenuate the cellular response to gonadotropins. The processes of recruitment, selection and dominance are controlled by the interaction of endocnne signals and locally produced ovarian growth factors that are associated with morphological and functional changes in granulosa and theca cells (Armstrong and Webb, 1997). Inhibins, activin, insulin-like growth factor I (IGF-I) and their binding proteins have direct and indirect effects on granulosa and theca cells that can modulate follicular development and steroidogenesis. Inhibin is a heterodimehc glycoprotein composed of an a subunit and a (3 subunit which are not linked covalently . The (3 subunit occurs in two different forms, termed PA and pB, giving hse to inhibin A (apA) and inhibin B (apB) on combination with the a subunit (Findlay et al., 1993; Rose and Gaines-Das, 1996). Inhibins have autocrine, paracrine and endocrine effects. Inhibin is a glycoprotein produced primarily in granulosa cells (Henderson and Franchimont, 1983), and preferentially inhibits pituitary FSH synthesis and/or secretion (De Jong, 1988; Glencross et al., 1994) by inhibiting expression of the gene encoding the p subunit of FSH (Beard et al., 1989). Inhibins through their systemic actions have an important role in determining species-specific ovulation rates (e.g., ovulation of only one follicle in cattle; Taya et al., 1996; Scanlon et al., 1993). Nine different forms of biologically and immunologically active inhibin (ranging from 29 to 160 kDa) have been isolated from bovine follicular fluid of dominant follicles (Good et al., 1995). The predominant inhibin forms are free a subunits and large (> 34 kDa) inhibin forms, rather than the fully processed inhibin

PAGE 158

149 form (Ireland etal., 1994). Inhlbin Increases in vitro LH-induced androgen synthesis of bovine theca cells (Wrathall and Knight, 1995), and Immunoactlve inhlbin secretion is stimulated by androstenedione, testosterone, and Ej (Wrathall and Knight, 1993). However, FSH had no effect on basal Immunoactlve inhlbin secretion by bovine (Wrathall and Knight, 1993) and ovine (Campbell et al., 1996) granulosa cells. Activin inhibited both LH-stimulated and Ej-stimulated androstenedione secretion by bovine theca cells (Wrathall and Knight, 1995). Furthermore, activin reversed the positive effect of inhlbin on basal LH-stimulated and Ej-stimulated androstenedione secretion but did not affect basal steroid output (Wrathall and Knight, 1995). Simultaneous addition of folllstatin and activin reversed the inhibitory effect of activin but did not modify the effects of inhlbin. These observations are indicative of opposing intrafoilicular paracrine roles for granulosa cell-derived inhlbin and activin in modulating thecal cell responses to gonadotropins and steroids in the bovine ovary. Epidermal growth factor and transforming growth factor-a stimulated marked increases in proliferation of granulosa cells from both small and large ovine follicles and a concomitant Inhibition of Ej and inhlbin synthesis (Campbell et al., 1994). Probably a local feedback loop exists within individual follicles involving a sequential change of inhlbins, activins and their binding proteins, which determines the different fates of selected and unselected follicles that develop in the same systemic environment of gonadotropins and growth factors (Roche, 1996). Intrafoilicular contents of inhlbin (34 kDa), activin and IGF binding protein 2 (IGFBP-

PAGE 159

150 2) are related inversely to Ej-producing capacity of dominant and subordinate follicles (Roche, 1996). It has been shown (Yuan et al., 1998) that dominant follicles expressed abundant IGF-I and IGF-II mRNA but failed to express IGFBP-2 which is a binding protein associated with atresia (Monget and Monniaux, 1995). On the other hand, subordinate follicles expressed lesser amounts of IGF-I and IGF-II but abundant IGFBP-2. Thus, activin, inhibin and growth factors have key local negative roles on gonadotropic action that may control processes of follicle selection and dominance (Roche, 1996). The precise physiological role of inhibin in folliculogenesis is unknown. Results of several studies support both endocrine (Scanlon et al., 1993; Glencross etal., 1994) and local (Knight etal., 1989; Guilbaultetal., 1993; Ireland and Ireland, 1994; Ireland et al., 1994) roles of inhibin in regulation of dominant follicles during the bovine estrous cycle. Sato et al. (1982) suggested that an inhibin-like substance present in bovine follicular fluid inhibits FSH action at the ovary through binding to the FSH receptor on the granulosa cell. More recently, a report by Schneyer et al. (1991) indicated that inhibin a subunits bind to FSH receptor sites and inhibit FSH bioactivity in rat granulosa cells. To understand the local role of inhibin, an experiment was designed to evaluate the effect of inhibins present in bovine follicular fluid of day 5 dominant and day 12 atretic dominant follicles on in vitro E2 secretion by bovine antral granulosa cells.

PAGE 160

Materials and Methods 151 Inhibin Immunoaffinitv Column An Inhibin immunoaffinity column was developed to obtain bovine follicular fluid (bFF) free of inhibins and to isolate the monomeric and multimeric forms of inhibin in follicular fluid (Good et al., 1995). A partially affinity purified rabbit antiserum against bovine a,.^-^^ gly-tyr (Good et al., 1995), provided by Dr. J. J. Ireland (Michigan State University, East Lansing, Michigan), was crosslinked to a HighTrap™ Protein G Sepharose® High Performance Column (Pharmacia Biotech Inc., Piscataway, NJ). The protein G column was equilibrated with three bed volumes of washing/binding buffer (10 mM sodium phosphate, pH 7.0, 150 mM sodium chloride, 10 mM ethylenediaminetetraacefic acid [EDTA]). Prior to the coupling, antiserum was centrifuged (Microcentrifuge Model 235B, Fisher) for 1 min at 13,000 X g. One ml of anti-inhibin antiserum was applied to the column and incubated for 1 h at 4 C. The column was washed with 10 ml of washing/binding buffer to remove unbound antibody, and then washed with 10 ml of 0.2 M triethanolamine (TEA), pH 8.5. To cross-link the antibody, the column was incubated for 30 min at room temperature (RT) with 10 ml of a solution containing TEA (0.2 M, pH 8.5) plus freshly made dimethylpimelimidate (10 mg/ml, pH 8.5 adjusted with NaOH). This allows optimal spatial orientation of antibodies and maximum antigen binding efficiency (Schneider et al., 1982). The reaction was stopped by washing the column withIO ml of 0.2 ethanolamine (ETA), pH 8.0, followed by incubation for 2 h at room temperature. At the end of the incubation

PAGE 161

152 period, the column was washed with 10 ml of washing/binding buffer and the column kept at 4°C until use. Pools of Bovine Follicular Fluid Pools of bovine follicular fluid (bFF) were prepared from day 5 dominant follicles (DF) and day 12 atretic DFs (ADF). Cows (n=25) were synchronized with an injection of GnRH (Buserelin: 8 \ig i.m.; Receptal®, Hoechst-Roussel Agri Vet, Sommerville, NJ) followed 7 days later with an injection of prostaglandin (PG) Fja (25 mg, i.m.; Lutalyse, Pharmacia & Upjohn Co., Kalamazoo, Ml). Atestrus cows were assigned either to the day 5 DF (n=1 3; 6.4 ml of FF) or day 1 2 ADF (n=1 0; 9.7 ml of FF) groups. On day 5 and 1 2 for the respective groups, first wave dominant follicles were aspirated via transvaginal ultrasound guided aspiration utilizing an Aloka EchoCamera SSD-500 unit (Aloka Co., LTD, Japan) equipped with a 5.0 MHz transvaginal convex array transducer and needle guide. Epidural anesthesia was induced with 5 ml of 2% lidocaine, and the perineal region was scrubbed and disinfected. The lubricated transducer with needle guide was inserted deep into the vagina. A 17-gauge, 60 cm, single channel sterile needle, fitted with a 19-gauge, 2.5 cm disposable tip was used in the needle guide to puncture the vaginal wall and peritoneum. By manipulation of the reproductive tract per rectum, ovaries were positioned and held firmly in front of the transducer in close apposition to the vaginal wall. After aligning the on-screen needle travel indicator line with the follicle, the needle was advanced through the wall of the vagina, into ovarian stroma and

PAGE 162

153 through the wall of the follicle. As soon as the tip of the needle was seen to enter the follicle, follicular contents were aspirated into a sterile plastic tube using a suction unit (Pioneer Medical, Inc., Madison, CT ) preset to a flow rate of 20 ml/minute at a preset pressure of 100 mm Hg. After collection, follicular fluid was centrifuged (2000 x g, 15 min, 4°C) to remove cellular debris and kept at 4°C throughout processing until frozen at -20°C. Dextran-coated charcoal was prepared by mixing 25 mg of dextran (65,600 molecular weight; Sigma, St. Louis, MO) and 250 mg of activated charcoal (100-400 mesh; Sigma, St. Louis, MO) in 100 ml of double distilled water for 16 h. After mixing, the solution was centrifuged at 2,619 x g for 30 min, supernatant discarded and pellets dried overnight in an oven at 35 °C. Prior to the use of follicular fluid, endogenous steroids were removed from pools of bFF by overnight treatment with dextran-coated charcoal (5 mg/ml). At the end of overnight incubation, follicular fluids were centrifuged thrice (12,365 x gfor 30 min). After charcoal treatment, both pools of bFF (day 5= 5.2 ml and day 12= 8 ml) were dialyzed (molecular weight cutoff, 3500; day 5= 5.2 ml and day 12= 7.8 ml) against Tris buffered saline (TBS, 0.1 M Tris [hydroxymethyl] Aminomethane, 1.5 M sodium chloride), pH 7.6 for 24 h, changing TBS every 6 h. To verify steroids concentrations E2 and progesterone [P4] were extracted from steroid-stripped free follicular fluid as described by Thatcher et al. (1 994b; Ej) and Knickerbocker et al. (1986; P4) and assayed using validated radioimmunoassays (Thatcher etal., 1994b). To determine if use of dextran-coated charcoal to remove steroids from bFF would not remove inhibins, follicular fluid was

PAGE 163

154 collected from ovaries retrieved from the slaughter house. Ovanes were collected and brought to the laboratory in about 2 h. They were transported at room temperature in an insulated container with sterile saline solution [0.9% NaCI] containing 1% penicillin-streptomycin [Sigma, St. Louis, MO]). Upon arrival ovaries were washed with medium for ovary collection at 37°C (Chapter 5), FF collected by aspiration and treated as follows: dextran-coated charcoal (5 mg/ml) overnight and dialyzed for 24 h; non charcoal-treated and dialyzed for 72 h, or 144 h; dialyzed for 144 h and charcoal-treated for 5 h or charcoal-treated overnight. Samples of the different treated follicular fluids (50 pg/lane) were subjected to a nonreducing onedimensional 10% SDS-PAGE (see page 9). Experimental follicular fluids aspirated from day 5 DF and day 12 ADF were pooled because the amount of follicular fluid from each individual follicle was not sufficient to accommodate the preparation of inhibin free bFF and immunoaffinity enriched inhibin in sufficient quantity for various treatments in cultures of bovine antral granulosa cell. Pools of bFF (day 5 DF and day 12 ADF) were sub-divided into two portions: one portion (day 5= 2.5 ml; day 12= 3 ml) was diluted 1 :2 in TBS buffer and treated in order to have follicular fluid without inhibin forms (FF without inh) and an inhibin enriched fraction (inh-eluate); the other portion (day 5= 2.5 ml, day 12= 4.8 ml) was left intact (FF complete [FFC]). The diluted follicular fluid (day 5= 5 ml and day 12= 6 ml) was applied to the inhibin immunoaffinity column in aliquots of 1 ml and eluted as follows: 1ml of bFF diluted 1:2 was applied to the column and incubated at 4°C for 2 h. After incubation, the column was eluted with three, 1 ml volumes of ice cold TBS buffer, pH 7.6. Each 1 ml of the eluate fraction

PAGE 164

155 was collected separately and represented bFF without inhibins. The inhibins bound to the column then were eluted with four, 1 nnl volumes of 50 mM sodium acetate, pH 2.5, and each 1 ml collected separately. Before collection, 100 ^il of 5 M NaOH were added to bullet tubes to neutralize the eluates containing 50 mM sodium acetate, pH 2.5. These fractions represent the eluate of immunoabsorbed inhibins (inh-eluate). After collection of inh-eluate, the column was washed with 2 ml of double distilled water followed by 2 ml of TBS buffer. This procedure was repeated to prepare the entire day 5 DF (5 ml at a dilution of 1 :2) and day 12 ADF FF (6 ml at a dilution of 1 :2). Eluate fractions of inhibin (day 5= 10 ml, day 12= 12 ml) were dialyzed (molecular weight cutoff, 3500) for 48 h against TBS buffer, pH 7.6, with changes in dialysis buffer every 6 h. After dialysis, protein concentration was measured in all fractions (FFC, FF without inh fractions, and inh-eluate fractions) using the bicinchinonic acid (BCA; Smith et al., 1988) assay with BSA as a standard. Based on protein concentrations, first and second ml fraction of FF without inh eluate, and second and third fraction ml of inh-eluate were pooled. Electrophoresis and Western Blotting Procedures Samples of day 5 DF and day 12 ADF fractions designated as bFFC, FF without inhibins and inh-eluate (50 \ig protein/gel lane) were mixed in 4X (12.5 pi) sample loading buffer (8% of sodium dodecyl sulphate [SDS; BDH Laboratory Supplies, Poole, England]; 4% glycerol; 250 mM Tris-HCI pH 6.8, and bromophenol

PAGE 165

156 blue solution until solution beconnes medium dark blue color [0.002%], and double distilled water to a final loading volume of 50 pi) without p-mercaptoethanol (non reducing conditions). For FF without inh, the loading of the sample was based on amount of protein equal to FFC (e.g., 50 pg protein) and by volume effraction alone equivalent to the volume of FFC necessary to have 50 [ig of protein. In the case of day 5 FF without inh, the volume was equal to 6.5 ^1 of FF diluted 1:10. The volume of day 12 FF without inh was equal to 5.8 pi of FF diluted 1:10. Samples were boiled for 5 min and subjected to one-dimensional 10% SDS-PAGE (Lammelli, 1970) using a Bio-Rad mini gel apparatus (Mini-Protean II, Vertical Electrophoresis Cell, Bio-Rad, Hercules, CA; 1 50 volts, 1 00 mA, 1 h; electrophoresis buffer: 1 92 mM glycine, 25 mM Tris, 0.1% SDS). To determine molecular mass of inhibin bands, broad-range molecular mass protein standards (7.7 to 204 kDa, Bio-Rad Kaleidoscope prestained molecular weight standards) were loaded (10 ^l) onto a single lane per gel and subjected to 10% SDS-PAGE concomitant with the bFF fractions but were not mixed in sample loading buffer and not boiled. After electrophoresis, gel and membranes were incubated in transfer buffer (25 mM Tris, 192 mM glycine, 20% methanol, pH 8.1-8.4) for 30 min. Proteins on the gel were transferred using a Trans-blot® semi-dry transfer cell (Bio-Rad; 20 volts, 400 mA, 45 min) to a Hybond™ ECL™ nitrocellulose membrane (Amersham Life Science, Little Chalfont, Buckinghamshire, England). Each membrane was blocked with 2% gelatin (Fisher Scientific, Fair Lawn, NJ) in TBS-Tween 20 (TBST; 0.1% Tween 20 plus TBS) for 2 h at RT and washed twice (1 x 15 min, 1 x 5 min)

PAGE 166

157 in TBST. The membrane was incubated for 2 h at RT with an inhibin antiserum (diluted 1:50 in 5% non-fat milk [Blotting Grade Blocker Non-Fat Dry Milk, Bio Rad, Hercules, CA]. The antiserum was generated in rabbit against bovine a gly-tyr (Good et al., 1995), and was absorbed with plasma from a 3-week old female calf. Briefly, 100 ^1 of the inhibin antiserum were diluted in 1 ml of plasma from a 3-week old female calf, incubated for 2 h at 4°C in a rotating device, centrifuged at 13,000 X g (Microcentrifuge Model 235B, Fisher) for 5 min, and the supernatant retained as absorbed anti inhibin antiserum (diluted 1:10). After incubation with the anti-bovine inhibin-a^'"^^ gly-tyr (Good et al., 1995), membranes were washed thrice in TBST (1x15 min, 2x5 min), incubated at RT for 1 h with a second antibody (peroxidase labeled goat anti-rabbit IgG antibody NIF 284, Amersham Life Science) diluted 1:12,000 in TBST containing 2% gelatin + 0.1% of goat normal serum. Following incubation, the membrane was washed in TBST (1 X 15 min, 4 x 5 min each) and protein bands identified by chemiiuminescence. The membrane then was incubated for 1 min with equal 4 ml volumes of detection reagent 1 and detection reagent 2 (ECL™ Western Blotting Analysis System, Amersham International pic, Little Chalfont, Buckinghamshire, England). Detection reagents were drained from the membrane, and the membrane wrapped in plastic saran wrap before placement into a single cassette (Kodak X-Omatic, Eastman Kodak, Rochester, NY) for 1 min with autoradiography film (Reflection™; NEN® Research Products Autoradiography film, NEN™ Life

PAGE 167

158 Science Products, Boston, MA). Developing of autoradiography film was performed in a Konica X-ray film processor QX-70 (Konica Corporation, Japan). Granulosa Cell Culture System Antral bovine granulosa cells (GC) were cultured in a defined medium using the system previously described (Chapter 5). Briefly, cycling cows (n= 6) were synchronized with an injection of GnRH (Buserelin: 8 pg i.m.; Receptal® HoechstRoussel Agri Vet, Sommerville, NJ) followed 7 days later with an injection of PGFsc (25 mg, i.m.; Lutalyse, Pharmacia & Upjohn., Kalamazoo, Ml). On day 6 of the synchronized cycle, one Synchromate B (SMB, Sanofi Animal Health, Inc., Overland Park, KS) implant was inserted and two injections of PGFja were given (25 mg at 7:00 h; 15 mg at 19:00 h). On day 10, the DF was aspirated (as described earlier), and two days later, on day 12, treatment with FSH-P (FSH-P™; Schering-Plough, Animal Health, Kenilworth, NJ) began and the SMB implant was replaced with a new SMB implant. Cows (n= 5) received 28 mg of FSH given in decreasing doses every 12 h (A.M./P.M.= 6/6, 5/5, and 3/3 mg per injection). On day 15, the SMB implant was removed and ovaries were collected at slaughter (Figure 5-1 , Chapter 5). Antral GC were harvested from all follicles > 9 mm and cultured (Chapter 5) in serum-free medium in 48-well plates (Costar, Cambridge, MA) for 48 h at 37°C gassed with 5% COj and 95% air. Wells were precoated with 3 ^g/cm^ of fibronectin (for 30 min at 37°C) before seeding of the cells (3x10^ cells/well). Cells

PAGE 168

159 were cultured in serum-free medium Ham's F-12 buffered with NaHCOj (1.176 mg/ml) and Hepes (20 mM), pH 7.4 supplemented with insulin (50 ng/ml), androstenedione (10'^ M), human transferrin (partially iron-loaded, 10 ^g/ml), ascorbic acid (17.6 fig/ml), gentamicin (20 ^g/ml), and nystatin (4 |ag/ml). To optimize the in vitro secretion of Ej, medium was supplemented with a priming dose of 0.5 ng/ml of FSH (porcine FSH [CY-FSH1= 41 x NIH-FSH-P1 , LH activity < 0.5%; gift from Dr. C. Y. Combarnous, INRA, Nouzilly, France) for the first 24 h of culture. During the second 24 h, cells were treated with 0.5 or 1 0 ng/ml of CY-FSH and each dose of the different fractions of bFF as shown on Table 6-1 (n = 2 wells per each combination of FSH dose and fraction dose). Dose 1 of day 5 DF inh-eluate was equivalent to 1 5.625 times the amount of aand dimehc forms of inhibin (2.14 |jg) in 10 ^1 of bovine follicular fluid from estrogen-active follicles (a-inhibins: 214.1 |ig/ml [0.28% of total protein] and dimeric inhibins: 0.866 ng/ml [0.0013% of total proteins]; Guilbaultetal., 1993). Dose 2 was equivalent to 244.1 times the amount of aand dimeric forms of inhibin in 10 ^1 of bovine follicular fluid from estrogenactive follicles (Guilbault et al., 1993). Dose 1 of day 12 ADF eluate inh was equivalent to 1 5.625 times the amount of aand dimeric forms of inhibin in 1 0 |al of bovine follicular fluid from estrogen-inactive follicles (a-inhibins: 90.5 i^g/ml [0.10% of total protein], dimeric inhibins: 2.96 i^g/ml [0.0034% of total proteins]; Guilbault et al., 1993). Dose 2 was equivalent to 244.1 times the amount of aand dimeric forms of inhibin in 10 ^1 of bovine follicular fluid from estrogen-inactive follicles (Guilbault et al., 1993).

PAGE 169

160 Table 6-1. Doses of different fractions of bovine follicular fluid of day 5 dominant and day 12 atretic dominant follicles Day 5 Dominant Follicle Day 1 2 Atretic Dominant Follicle Fractions Dose 1 Dose 2 Dose 1 Dose 2 Follicular Fluid Complete 10 nl (763.9 ng) 20 nl (1527.8 ng) 10 Ml (864.2 ng) 20 ^l (1728.4 ng) Follicular Fluid without Inhibins 94.4 nl (763.9 ^g) 188.8 [i\ (1527.8 ^g) 96 ^l (864.2 ^g) 192 nl (1728.4 ng) Inhibins-eluate 10.15 nl (33.55 ^g) 158.66 nl (524.21 ^g) 3.44 ^l (13.9 ^g) 53.7 ^l (218.12 ^g) In order to test that is synthesized from androstenedione (A4) by bovine antral granulosa cells in culture, wells without A4 but with FSH (0.5 and 10 ng/ml FSH; n=4 per dose) also were included. A series of control incubations were conducted to examine the effects of no follicular fluid (e.g., culture media alone and effects of different amounts of TBS buffer used as a diluent for the follicular fluid fractions [e.g., FFC, FF-inh, inh-eluate]). The volumes of TBS tested on Ej secretion by granulosa cells are shown on Table 6-2. Table 6-2. Effects of TBS volume added to culture medium (500 pi total volume) on Ep secretion by bovine antral granulosa cells in culture Medium Volume of TBS used in FF fractions as diluent 500 Ml OmI 446.3 Ml 53.7 Ml (volume of dose 2 inh-eluate of day 12) 405.6 Ml 94.4 Ml (volume of dose 1 FF without inh of day 5) 404.0 Ml 96 Ml (volume of dose 1 FF without inh of day 12) 341.4 Ml 158.6 Ml (volume of dose 2 inh-eluate of day 5) 311.2 Ml 188.8 Ml (volume of dose 2 FF without inh of day 5) 308.0 Ml 192 Ml (volume of dose 2 FF without inh of day 12)

PAGE 170

161 Statistical Analyses Data were analyzed by least squares analysis of variance using the General Linear Models procedure of the Statistical Analysis System (SAS, 1988). Data are presented as means ± SEM of duplicate measurements. Secretion of and P4 are expressed as ng/10^ live cells at the end of the culture period. For purposes of data presentation, Ej and P4 concentrations will be expressed as ng/10^ live cells (Chapter 5). The mathematical models used to analyze the data are presented in Table 6-3. Because the four way interactions were significant, each day (day 5 DF and day 12 ADF) was analyzed separately to further understand the higher order interactions. The reduced mathematical models are described in Table 6-3. Table 6-3. Mathematical models used to analyze the secretion of estradiol and Full Model Reduced Model Sources d.f. Sources d.f. Dose of CY-FSH 1 Dose of CY-FSH 1 Day of DF 1 Fraction 2 Fraction of follicular fluid 2 Dose of fraction 1 Dose of fraction 1 Dose of CY-FSH x fraction 2 Dose of CY-FSH x day of DF 1 Dose of CY-FSH x dose of fraction 1 Dose of CY-FSH x fraction 2 Fraction x dose of fraction 2 Dose of CY-FSH x dose of fraction 1 Dose of CY-FSH x fraction x dose of fraction 2 Day of DF x fraction 2 Day of DF x dose of fraction 1 Fraction x dose of fraction 2

PAGE 171

162 Table 6-3.continued Full Model Reduced Model Sources d.f. Sources d.f. Dose of CY-FSH x day of DF x fraction 2 Dose of CY-FSH x day of DF x dose of fraction 1 Day of DF x fraction x dose of fraction 2 Dose of CY-FSH x fraction x dose of fraction 2 Dose of CY-FSH x day of DF x fraction x dose of fraction 2 Error 24 Error 12 Orthogonal contrasts were used to evaluate responses. The following contrasts were tested: dose of CY-FSH (0.5 ng/ml vs 1 0 ng/ml); fraction of FF (FFC vs FF without inhibins + inh-eluate, and FF without inhibins vs inh-eluate), and dose of fraction (dose 1 vs dose 2). Contrasts were also examined for the interactions that were significant. These analyses do not include cow in the model regarding the FF preparation or cow as a source of variability relative to source of granulosa cells. As described earlier, FF from individual cows at day 5 or at day 12 were pooled to provide sufficient follicular fluid for processing of fractional components for testing in culture. Thus, significant differences regarding day of DF and FF fractions represent differences between the pools of follicular fluid from day 5 dominant and day 1 2 atretic dominant follicles. A total of 1 3 cows contributed to the day 5 FF pool and 10 cows contributed to the day 12 FF pool. Statistical differences are Interpreted to represent a true physiological difference between the follicular fluid

PAGE 172

163 pools (day 5 vs day 12) which represents the average of all cows within the pool. At risk is the possibility that a few cows within a respective pool may be outliers that account for the difference in response between the follicle pools. Hopefully, such an effect is minimized by the large number of animals within a pool at a specific stage of dominant follicle development. Granulosa cells of follicles > 9 mm were pooled across five cows that underwent the follicular superinduction scheme in order to provide sufficient number of granulosa cells for the biological test system. It is not possible to utilize as many experimental treatments and replicates on granulosa cells prepared from each donor cow separately. Furthermore, it is extremely costly. Differences among cows were minimized since follicles designated for aspiration had to be > 9 mm and healthy in appearance (well vascularized and translucent). Results Validation of Dextran-Coated Charcoal Treatment An immunoblot of bFF from ovaries collected at the slaughterhouse and subjected to charcoal treatment and dialysis is shown in Figure 6-1 . There were no major differences among the different lanes for the molecular forms of bovine inhibin after the respective treatments. Dimeric and monomeric forms of immunoabsorbed reactive bovine inhibin were evident on the blots (> 160, 110, 77, 58, 49 kDa) and agrees with those reported by Good et al. (1995) and de la Sota (1995). Apparently the 34 kDa form was not present in the FF because lane 1 (FF without any treatment; Figure 6-1 ) does not show any band at that molecular weight

PAGE 173

164 size. When FF was dialyzed for 0, 72 or 144 h without dextran-coated charcoal treatment, Ej decreased from 2.8 pg/ml to 85.8 ng/ml and 0.9 ng/ml, respectively. Because concentration was considered high even after 144 h of dialysis, charcoal treatment for 1 6 h plus dialysis for 24 h was the approach chosen (lane 2). Concentrations of Ej and P4 in dextran-charcoal treated bFF after charcoal treatment for 16 h and dialyzed for 24 h, were < 0.39 ng/ml and < 0.31 ng/ml, respectively. Thus, dextran-coated charcoal effectively removed Ej and P4 from follicular fluid. mj 1 3 4 5 6 260 kDa 148 kDa 75 kDQ 43 kDa 3Z1k[Da 17.8 kEDa Figure 6-1 . Immunoblot analysis of bFF from slaughterhouse ovaries. Fifty pg of protein from each sample were subjected to a non-reducing one-dimensional 10% SDS-PAGE and immunoblot analysis using a partially affinity purified rabbit antiserum against bovine inhibin-a^'-^ gly-tyr (Good et al., 1995). Lane 1= bFF without any treatment; lane 2= bFF charcoal-treated for 16 h and dialyzed for 24 h; lane 3= bFF dialyzed for 72 h; lane 4= bFF dialyzed for 144 h, lane 5= bFF dialyzed for 144 h and charcoal-treated for 5 h; and lane 6= bFF dialyzed for 144 h and charcoal-treated for 16h.

PAGE 174

165 Concentrations of total proteins were very high (Table 6-4) in follicular fluid and did not differ appreciably between untreated and treated FF that was subjected to dextran-coated charcoal stripping and dialysis. There were no obvious changes in immunoblotting intensity among the six treatment preparations as well, indicating no preferential removal of immunoinhibin active bands among treatments. Figure 6-2 (Commassie blue staining of the gel) shows no differences in the array and intensity of proteins among the different treatments both at high and low molecular weights. For example, the 28 kDa protein was evident among all treatments. At these high total protein concentrations in FF, the amount of preferential removal of protein at various sizes appear to be minimal (Figure 6-2). Furthermore, the immunoreactive proteins appear to be uniform. Thus, prepreparations of FF for subsequent fractionation did not appear to cause a total loss of protein or differential loss of selected protein . Table 6-4. Protein concentration (mg/ml) in dextran-coated charcoal treated follicular fluid from slaughter house ovaries Treatment Protein concentration (mg/ml) Follicular fluid without any treatment 69.4 ± 1.00 Follicular fluid charcoal-treated for 16 h + dialysis for 24 h 68.3 ±0.75 Follicular fluid no charcoal-treated + dialysis for 72 h 61.8 ±0.97 Follicular fluid no charcoal-treated + dialysis for 144 h 58.5 ± 0.56 Follicular fluid dialyzed for 144 h + charcoal-treated for 5 h 65.2 ± 0.70 Follicular fluid dialyzed for 144 h + charcoal-treated for 16 h 63.5 ± 0.60

PAGE 175

166 MW 1 2 3 4 5 6 32.1 kDa 260 kDa ^ 148 kDa ^ 75 kDa 17.8 kDa 43 kDa Figure 6-2. Commassie blue stain of a nonreducing one-dimensional 10% SDS-PAGE of bovine follicular fluid (bFF) from slaughterhouse ovanes. Fifty \ig of protein from each sample were subjected to a nonreducing onedimensional 1 0% SDS-PAGE and immunoblot analysis using a partially affinity purified rabbit antiserum against bovine inhibin-a,.^ gly-ty r (Good et al. , 1 995). Lane 1= bFF without any treatment; lane 2= bFF charcoal-treated for 16 h and dialyzed for 24 h; lane 3= bFF dialyzed for 72 h, lane 4= bFF dialyzed for 144, lane 5= bFF dialyzed for 144 h and charcoal-treated for 5 h, and lane 6= bFF dialyzed for 144 h and charcoal-treated for 16 h. Protein concentrations of experimental pools of day 5 dominant and day 12 atretic dominant follicles FF, before charcoal treatment were 83.2 mg/ml and 93.8 mg/ml for day 5 and day 12, respectively. After charcoal treatment, protein concentration did not change (day 5: 81.3 mg/ml and day 12: 90.7 mg/ml). Following dialysis protein concentrations were 76.392 mg/ml and 86.429 mg/ml for day 5 DF FFC and day 12 ADF FFC, respectively. This reflects some loss of protein < 3,500 kDa. Figure 6-3 shows the elution profile for protein content of FF without inh and inh-eluate of day 5 DF (Figure 6-3A) and day 12 ADF (Figure 6-3B).

PAGE 176

167 Fractions 1 and 2 of FF without inh and fractions 2 and 3 of inh-eluates were pooled to make up the FF without inh and inh-eluate material for testing in culture. For both day 5 and day 12 follicular pools, there was not a 100% recovery of protein from the column (Table 6-5). For day 5 DF, 43.25% of the initial protein was recovered in the fractions, whereas 66.83% of protein was recovered for the column runs of day 12 FF pool. This loss in protein concentration is due to several reasons. There was always a loss of some fluid when the follicular fluid was loaded into the column. Because of the viscous nature of FF. some foaming occurred, and when air was expelled from the syringe to avoid adding air to the column several microliters were lost. Also, when FF was loaded into the column, the last drops of buffer associated with displacement of the bed volume of the column contained some of the FF protein being loaded. Protein concentration was measured in some of these fractions and contained between 2.9 to 7.5 mg/ml of protein. These various fractions contributed to an incomplete recovery of total proteins. Western Blot Analysis of Inhibin in Follicular Fluid Immunoblot analysis of the experimental fractions using an bovine inhibin-a^^" ^® gly-tyr antibody (Good et al., 1 995) indicated that some inhibin forms (49, 1 1 0 kDa forms) were removed when immunoreactlve bands of day 5 and 12 FF without inh were compared to their respective follicular fluids that were complete (bFFC). In the inh-eluate fractions there is an enrichment of the 34 and 49 kDa proteins (Figure 64).

PAGE 177

F Tactions Figure 6-3. Elution profiles for protein concentration of FF without inhibins and inhibins-eluate fractions from day 5 DF and day 12 ADF. Asterisk indicates the fractions with the highest protein concentration that were chosen to conform the pool from the respective day. Table 6-5. Protein concentration (mg/ml) and protein recovery (mg) from follicular fluid of day 5 dominant and day 1 2 atretic dominant follicles (1 ml FF diluted 1:2) when processed (1 ml) through the immunoaffinity column Day 5 dominant follicle Day 12 atretic dominant follicle Fraction PR (mg) % Recov PR (mg) % Recov FF added 42.8 41.2 FF-inh 1 10.8 ±0.130 25.25 15.4 ±3.60 37.45 FF-inh 2 1.35 ±0.100 3.15 1.8 ±0.03 4.44 FF-inh 3 0.12 ±0.020 0.28 0.47 ± 0.03 1.14 Inh-eluate 1 0.02 ± 0.001 0.05 0.07 ± 0.03 0.17 Inh-eluate 2 3.33 ± 0.430 7.77 4.88 ± 0.70 11.85 Inh-eluate 3 2.83 ±0.100 6.61 4.06 ±0.19 9.86 Inh-eluate 4 0.06 ±0.010 0.14 0.79 ± 0.22 1.92 Total 18.51 43.25 27.47 66.83

PAGE 178

169 MW 1 2 3 4 5 6 7 8 204 kDa 121 kDa 78 kDa 30.7 kDa 39.5 kDa 19.7 kDa-^ Figure 6-4. Different molecular masses (kDa) of immunoactive inhibin bFF (50 \ig of protein) for day 5 OF FF after charcoal treatment and dialysis before passage through the immunoaffinity column (lane 1). FF without inhibins (loaded by equal amount of protein as FF before passage through the column, lane 2, or loaded by equal volume to volume of FF that contains 50 ^lg of protein, lane 3). Inhibins-eluate day 5 DF FF (50 pg of protein, lane 4). Day 12 ADF FF after charcoal treatment and dialyzed before passage through the inhibin immunoaffinity column (lane 5). FF without inhibins (loaded by equal amount of protein as FF before passage through the column, lane 6, or loaded by equal volume to volume of FF that contains 50 pg of protein, lane 7. Inhibinseluate day 12 ADF FF (50 ^g of protein, lane 78). The predominant forms detected on day 5 DF FFC (lane 1) were 49, 110, and > 160 kDa, whereas a lesser amount of the 88 kDa form was present. When FF without inh was loaded as equal amount of protein to day 5 FFC (lane 2), the bands that correspond to the inhibin forms in FFC were much lighter (49, 1 1 0, > 1 60 kDa form), and when the loading of FF without inh was done by equal volume (lane 3) to volume of day 5 FFC, only one region (> 160 kDa) was present. On the day 5 inh-eluate (lane 4) band that correspond to the > 160 kDa, 49 kDa and 34 kDa forms were evident. Day 12 ADF FFC also showed bands at 49, 110 and > 160 kDa, and a less intense band at 88 kDa. When FF without inhibins

PAGE 179

170 (lane 6) was loaded by equal amount of protein to day 1 2 FFC, the same bands that were present in FFC were present but less intense. When the loading of the sample was based on volume (same volume as day 12 FFC; lane 7), the only proteins present were those > 160 kDa forms. In lane 8 (inh-eluate day 12), there were bands at > 160 kDa, 49 kDa, and a less intense band at 34 kDa. In summary, 49, 88 and 110 kDa forms were markedly removed from day 5 FFC by the immunoaffinity column, and the > 1 60 kDa forms were decreased. The 34 kDa, 49 kDa and 88 kDa forms were present on the inh-eluate. In the case of day 12 FFC, the 49, 88, 110 and > 160 kDa forms were present. All of these proteins were markedly reduced in the FF without inh fraction and were enriched in the inh-eluate. These forms of inhibin identified in the present experiment coincide with those reported by Good et al. (1995) and de la Sota (1995) using a bovine antiserum against bovine inhibin a^^'^® gly-tyr and a mink antiserum against bovine inhibin a^^'^ gly-tyr, respectively. Figure 6-5 shows the Commassie blue stain of the array of proteins present in FFC, FF without inh and FF inh-eluate of day 5 DF (lanes 1,2,3 and 4) and day 12 ADF(lanes 5, 6, 7 and 8). There are distinct differences among the lanes. Lanes 4 and 8, which represent immunoabsorbed proteins, show a markedly different array of proteins compared to FFC (lanes 1 and 5) and FF without inh (lanes 2 and 6). The selective removal of only immunoreactive inhibin forms (lanes 4 and 8; Figure 6-4) is consistent with a similar array of commassie blue stained proteins (Figure 6-5, lanes 4 and 8). Although FFC (lanes 1 and 5) and FF without

PAGE 180

171 inh (lanes 2 and 6) have some pattern of proteins, the pattern of immunoabsorbed protein is different (Figure 6-4) indicating selective removal of inhibin forms. Figure 6-5. Commassie blue stain of a non reducing one-dimensional 10% SDSPAGE of bFF from day 5 DF and day 12 ADF and their different fractions after passage through the inhibin immunoaffinity column. Day 5 DF FF after charcoal treatment and dialysis before passage through the immunoaffinity column (50 pg of protein; lane 1). FF without inhibins (loaded by equal amount of protein as FF before passage through the column, lane 2, or loaded by equal volume to volume of FF that contains 50 |jg of protein, lane 3). Inhibins-eluate day 5 DF FF (50 ^Jg of protein, lane 74). Day 12 ADF FF after charcoal treatment and dialyzed before passage through the inhibin immunoaffinity column (lane 5). FF without inhibins (loaded by equal amount of protein as FF before passage through the column, lane 6, or loaded by equal volume to volume of FF that contains 50 pg of protein, lane 7). Inhibins-eluate day 12 DF FF (50 pg of protein, lane 8). Effect of Bovine Follicular Fluid on FSH-Stimulated Estradiol-17B Secretion Estradiol-17B. Androstenedione is the precursor for estrogen synthesis. When was not added to the culture, concentration of Ej in medium conditioned by antral granulosa cells was below the sensitivity of the assay (0.39 ng/ml; Chapter 5) for either the 0.5 and 10 ng/ml dose of CY-FSH. Thus, addition of A^ substrate to

PAGE 181

172 granulosa cells is essential for secretion since thecal cell components that produce androstenedione are absent from the culture system. A dose of 1 0 ng/ml of CY-FSH produced more (1 0^ live cells) than a dose of 0.5 ng/ml of CY-FSH (P < 0.03). A volume of TBS effect was not detected (P > 0.10; Table 6-5). Table 6-6. Estradiol (ng/10^ live cells) secretion by bovine antral granulosa cells in absence of follicular fluid and in presence of different volume of TBS Volume of TBS Estradiol (ng/10° live cells) No follicular fluid 18.9 ±1.9 53.7 |jl 13.6 ±1.9 94.4 |jl 14.7 ±1.9 96.0 Ml 16.2 ±1.9 158.6 |jl 15.5 ±1.9 188.8 Mi 18.4 ±1.9 192.0 Ml 23.1 ± 2.3 There was no evidence that increased TBS diluent would adversely decrease Ej secretion via a dilution of nutrients in the culture medium (Table 6-6). Thus follicular fluid fraction effects (positive or negative) will be due to the protein components in the fraction not the TBS diluent. In general, Ej secretion at 0.5 (11.7 ng/1 00,000 live cells) and 10 ng/ml of CY-FSH (26.3 ng/1 00,000 live cells) was greater (P < 0.06) in the absence of follicular fluid proteins than in the presence of follicular fluid proteins. Thus subsequent analyses are comparisons among follicular fluid fractions which contain proteins that appear to be stimulatory or inhibitory to Ej secretion.

PAGE 182

173 Table 6-7 shows the analysis of variance for analysis of data utilizing the full model. The four way interaction (dose of CY-FSH x day of follicular fluid x fraction X dose effraction) for E2 and P4 concentrations and £2^4 ratio in ng/10^ live cells was significant. Therefore to examine this in more detail, separate analyses were done for granulosa cell incubations treated with FF from day 5 dominant (Table 6-8) and day 12 (Table 6-9) atretic dominant follicles. 3 02 5 s "s> 1 0 id T~T~ f J 2_ _! 2_ _! ZDOSE DOSE DOSE FSH 0.5 ng/ml DOSE DOSE DOSE FSH 10 ng/ml Follicular Fluid Com plete I I Follicular Fluid w/o Inhibins I I Inhibins-Eluate Figure 6-6. Secretion of estradiol (ng/10^ live cells) by bovine antral granulosa cells over 48 h of culture in response two doses of CY-FSH (0.5 and 1 0 ng/ml) and two doses of different fractions of day 5 FF (panel A) and day 12 ADF follicular fluid (panel B). Different letters within the same day indicate differences at P < 0.0001. Dotted lines represent secretion in the absence of follicular fluid.

PAGE 183

174 A significant three way Interaction (dose of CY-FSH by fraction by dose of fraction; P < 0.0001; Figure 6-6A) was detected for E2 secretion (ng/10^ live cells) by antral granulosa cells in the presence of day 5 FF fractions. There were no differences among experimental fractions and doses of experimental fractions on E2 secretion when cells were treated with 0.5 ng/ml of CY-FSH. However, E2 secretion was higher (P < 0.001) with the 10 ng/ml dose of CY-FSH. Dose 1 (16.7 ±0.8 ng/10^ live cells) and 2 (16.0 ±0.8 ng/10^ live cells) of day 5 FFC, in presence of 10 ng/ml of CY-FSH, had no effect on E2 secretion. Dose 2 of day 5 FF without inh (8.43 ± 0. 8 ng/10^ live cells) decreased E2 secretion (P < 0.001). Thus factors other than inhibin in day 5 FF decrease E2 secretion by granulosa cells. Dose 2 of inhibins-eluate from day 5 DF FF inhibited (P < 0.001 ; Figure 6-6A) E2 secretion (9.9 ± 0.8 ng/1 0^ live cells) in culture. Thus inhibin enriched fraction of FF collected from day 5 DF decreased E2 secretion. It is possible that the 49 kDa form that is enriched in this fraction inhibits secretion of E2 by granulosa cells. In the presence of follicular fluid fractions from day 12 ADF a dose of 10 ng/ml of CY-FSH stimulated a higher secretion of E2 than a dose of 0.5 ng/ml of CYFSH (P < 0.001; Figure 6-6B). In contrast to day 5 FFC, day 12 FFC at dose 2 inhibited Ej secretion, and this degree of inhibition was comparable to the day 12 ADF FF without inh fractions at either dose 1 and dose 2.Thus, removal of the inhibin fonns increased the negative effect of FF on E2 secretion. Inhibins-eluate at both doses in presence of 10 ng/ml of CY-FSH markedly stimulated granulosa cells to secrete Ej in culture, and no differences were detected in the stimulatory response between the doses of inhibins-eluates (Figure 6-6B). In day 12 FF of atretic follicles, there appear to be a form of inhibin that stimulated E2 secretion.

PAGE 184

175 However, this stimulation is not evident when mixed with other follicular components that are inhibitory in follicular fluid. Table 6-7. Analysis of vanance for E2 and P4 concentrations, and P4 ratio (ng/1 0^ live cells) in cultured bovine antral granulosa cells treated with follicular fluid from day 5 dominant and day 12 atretic dominant follicles (full model) Source df E2 P4 Dose of CY-FSH 1 0.0001 0.0001 0.0001 Day of DF 1 0.5403 0.0040 0.1251 Fraction of follicular fluid 2 0.0012 0.0001 0.0183 Dose of fraction 1 0.0005 0.0121 0.0009 Dose of CY-FSH x day of DF 1 0.5730 0.1328 0.9455 Dose of CY-FSH x fraction 2 0.0470 0.0105 0.0293 Dose of CY-FSH x dose of fraction 1 0.0011 0.0503 0.0067 Day of DF x fraction 2 0.0367 0.1370 0.0311 Day of DF x dose of fraction 1 0.3541 0.3177 0.4547 Fraction x dose of fraction 2 0.0025 0.0118 0.0417 Dose of CY-FSH x day of DF x fraction 2 0.0712 0.2414 0.0084 Dose of CY-FSH x day of DF x dose of fraction 1 0.3166 0.5029 0.3104 Day of DF X fraction x dose of fraction 2 0.0199 0.1913 0.0123 Dose of CY-FSH x fraction x dose of fraction 2 0.0047 0.8959 0.0328 Dose of CY-FSH x day of DF x fraction x dose of 2 0.0405 0.0448 0.0260 fraction Error 24

PAGE 185

176 Table 6-8. Analysis of variance for and concentrations, and E2.P4 ratio (ng/10^ live ceils) in cultured bovine antral granulosa cells treated with follicular fluid from day 5 dominant follicle (reduced model) Source df E2 P4 ^2.P. Dose of CY-FSH 1 0.0001 0.0001 0.0001 Fraction 2 0.0148 0.0071 0.0001 Dose of fraction 1 0.0011 0.3880 0.0003 Dose of CY-FSH X fraction 2 0.0002 0.1239 0.0001 Dose of CY-FSH x dose of fraction 1 0.0004 0.4795 0.0007 Fraction x dose of fraction 2 0.0404 0.0635 0.0004 Dose of CY-FSH x fraction x dose of fraction 2 0.0047 0.03494 0.0022 Error 12 Progesterone . There were significant differences (P < 0.001; Table 6-6; Figure 67A) between doses of CY-FSH in P^ secretion expressed as ng/10^ live cells for experimental fractions from day 5 dominant and day 12 atretic dominant follicles. The dose of 1 0 ng/ml of CY-FSH was inhibitory to P4 secretion by antral granulosa cells exposed to follicular fluid fractions from either stage follicle (day 5 or day 12; P < .0009). There were no differences in P4 secretion between doses of FFC and FF without inh when a dose of 0.5 ng/ml of FSH was added to the culture. However, dose 2 of inh-eluate stimulated (P < 0.06) P4 secretion (137.2 ± 10.2 ng/10^ live cells) compared to dose 1 (94.7 ± 10.2 ng/10* live cells; Figure 6-7A). At the low dose of 0.5 ng/ml of CY-FSH, inhibin enriched fractions of day 5 DF appeared to stimulate luteinization during the 24 h treatment period (24 to 48 h) in culture. There were no differences in P4 between doses among fractions of

PAGE 186

177 follicular fluid or their doses for day 5 dominant follicles at a dose of 1 0 ng/ml of CYFSH. Table 6-9. Analysis of variance for and concentrations, and E^P^ ratio (ng/1 0^ live cells) in cultured bovine antral granulosa cells treated v\/ith follicular fluid from day 12 atretic dominant follicle (reduced model) Source df E2 P4 Dose of CY-FSH 1 0.0001 0.0001 Fraction 2 0.0096 0.0001 Dose of fraction 1 0.5637 0.0001 Dose of CY-FSH x fraction 2 0.1082 0.0012 Dose of CY-FSH x dose of fraction 1 0.2978 0.0011 Fraction x dose of fraction 2 0.8779 0.0418 Dose of CY-FSH x fraction x dose of fraction 2 0.5602 0.0073 Error 12 When granulosa cells were incubated w/ith CY-FSH at a concentration of 0.5 ng/ml, dose 2 of each FF fraction (FFC, FF without inh, inh-eluate) stimulated (P < 0.0001) P4 secretion (Figure 6-7B). This was not evident at the higher dose of 10 ng/ml of CY-FSH, except that dose 2 of inh-eluate induced a greater secretion (P < 0.01) of P4 (81 .1 ± 3.5 ng/10* live cells) than dose 1 (59.2 ± 3.5 ng/10' live cells). Thus, the intrafollicular environment of a day 12 ADF appears to stimulate P4 secretion or luteinization when exposed to a low dose of CY-FSH. This was not evident for follicular fluid fractions collected from a day 5 DF that is more estrogenic. EstradiolProaesterone ratio . There was a significant (P < 0.002) three way interaction for E2:P4 ratio (ng/10^ live cells; Figure 6-8A) for day 5 dominant follicle

PAGE 187

178 FF. There were not differences in £2^4 ratio among the different fractions and their doses added to the culture in presence of a dose of 0.5 ng/mi of CY-FSH. When a dose of 10 ng/mi of FSH was added, E^.P, ratio was higher (P < 0.001) for all fractions regardless of fraction dose. Dose 2 of FF without inhibins (0.24 ± 0.02) and inh-eluate (0.20 ± 0.02) had lower (P < 0.004) E^P, ratios than their respective one doses (FF without inh: 0.40 ± 0.02; inh-eluate:0.40 ± 0.02). This further supports presence of and an enrichment in inhibitory factors to secretion in these follicular fluid components from day 5 DF. As for the day 5 dominant follicle FF pool, there was a higher £3^4 ratio (P < 0.001; Figure 6-8B) when 10 ng/ml of CY-FSH were added to the granulosa cell culture regardless of the follicular fluid fraction added from day 12 atretic dominant follicle. There were not differences in £2^4 ratio between doses among fractions in presence of 0.5 ng/ml of CY-FSH. Estradiol:P4 ratio was lower (P < 0.1) when dose 2 of FFC (0.3 ± 0.03) and inh-eluate (0.28 ± 0.03) were added to the cells than dose 1 of the same fractions (FFC= 0.38 ± 0.03; inh-eluate= 0.43 ± 0.03). The E2:P4 ratio was decreased in the FF without inh and there were no differences in E2:P4 ratio between doses of FF without inhibins.

PAGE 188

1601401200) o o o 80o o" o 60o> c 40CL 200DOSE DOSE DOSE DOSE DOSE DOSE FSH 0.5 ng/ml FSH 10 ng/ml DOSE DOSE DOSE FSH 0.5 ng/ml DOSE DOSE DOSE FSH 10 ng/ml Hi Follicular Fluid Complete Follicular Fluid w/o Inhibins I I Inhibins-Eluate Figure 6-7. Secretion of progesterone (ng/10* live cells) by bovine antral granulosa cells over 48 h of culture in response to two doses of CY-FSH (0.5 and 10 ng/ml) and two doses of different fractions of day 5 DF (panel A) and day 12 ADF follicular fluid (panel B). Different letters within the same day indicate differences at P < 0.01.

PAGE 189

180 0} o > 0.6 n 0.5 0.4o o o o" 2 0.3 CL csi 0.2 0.1 0 — a T a a T a T a T a T t L, 1 1 2 1 2 1 2 b X c i b X c X 1 1 DOSE DOSE DOSE DOSE DOSE DOSE 0.68 0.5 a> .> o o" g 0.3"5) 2 0.2 csi LU 0.1 • 0FSH 0.5 ng/ml FSH 10 ng/ml B DOSE DOSE DOSE DOSE DOSE DOSE FSH 0.5 ng/ml FSH 10 ng/ml ^1 Follicular Fluid Com plete I I Follicular Fluid w/o Inhibins Inhibins-Eluate Figure 6-8. Estradiol:progesterone ratio (ng/10* live cells) in culture medium by bovine antral granulosa cells over 48 h of culture after 24 h of treatment with two doses of CY-FSH (0,5 and 10 ng/ml) and two doses of different fractions of day 5 DF (panel A) and day 12 ADF (panel B) follicular fluid. Different letters within the same day indicate differences at P < 0.01.

PAGE 190

181 Discussion Changes in amounts and proportions of the different molecular forms of inhibin (de la Sota, 1995; Sunderland et al., 1996) and expression of inhibin/activin mRNA for a, 3^ and pg subunits (Ireland and Ireland, 1994) are related to physiological stages of follicle growth. There are distinct patterns in molecular forms of inhibin (Guilbault et al., 1993; de la Sota, 1995; Sunderland et al., 1996) in follicular fluid of dominant (e.g., day 5 DF) or atretic follicles (e.g., day 12 atretic DF). Therefore, the role of the inhibin forms may be different according to the stage of dominant follicle development. De la Sota (1995) reported that approximately 25% of inhibin forms present in FF are monomeric forms (a-inhibins) whereas 75% are dimeric forms of inhibin. At day 12, absolute amounts of the 34 kDa form in the atretic dominant follicle increased two fold compared to the day 5 DF. In the present experiment, amounts of inhibin forms in the pool of FF for day 5 DF and day 12 ADF were not quantified. However, the effects of immunoenriched forms of inhibins in FF on bovine antral granulosa cells in culture were measured between day 5 and day 1 2 dominant follicles. Day 5 DF appears to have factors that inhibit Eg secretion which are enriched in FF without inh fraction and in the inh-eluate fraction (Figure 6-6A). Dose 2 of both FF without inh and FF inheluate fractions decreased secretion. These inhibitory components were enriched in the FF without inh and inh-eluate fractions compared to FFC doses. It has been shown that other growth factors (e.g., TGF-a; Rouillier et al., 1997) and other proteins such as IGFBPs (Monget and Monniaux, 1995), tissue inhibitor of

PAGE 191

182 metalloproteinase (Roche, 1996) and hsp90 (Roche, 1996) are present in follicular fluid and have a negative effect on E2 secretion by antral granulosa cells. In the present study, inhibitory effects on secretion were not observed with a low dose of CY-FSH (0.5 ng/ml) suggesting an interaction with FSH-induced actions. Perhaps, a dose of 10 ng/ml of CY-FSH is sufficient to compete with an intra follicular inhibitcr for the FSH receptor on granulosa cells. However, at the higher doses of FF without inh and FF inh-eluate from day 5 DF, the inhibitors decrease FSH binding and reduce E2 secretion. This may be the case for an inhibin form in the inh-eluate since inhibin or an a-inhibin form has been suggested to compete for the FSH receptor (Sato et al., 1982; Schneyer et al., 1991). Furthermore, hsp90 recently has been shown to be present in FF of bovine (Driancourt, personal communication) and ovine (Driancourt and Catelli, unpublished data) follicles, and it may possibly inhibit aromatase activity. Perhaps, hsp90 may be a candidate protein that contributes to the inhibitory effect of day 5 and day 12 follicular fluid fractions that contain reduced amount of inhibin (e.g., FF without inh) that inhibits E2 secretion. In contrast to day 5, there was a stimulation in secretion (ng/10^ live cells) by inh-eluate of day 1 2 ADF. This differential stimulation was even evident with day 12 ADF follicular fluid since FF without inh was inhibitory to Ej secretion. Day 12 ADF has a higher proportion of the mature 34 kDa form of inhibin (de la Sota, 1995), and it has been hypothesized (Baird and Smith, 1993) that increased production of E2 is in part dependent on the action of inhibin at the intrafollicular level to stimulate LH-induced androgen production (Findlay, 1993; Hillier et al.,

PAGE 192

183 1 991 ; Wrathall et al., 1 995). Therefore, a higher proportion of the 34kDa form in the inh-eluate of day 12 than in FF without inh or FFC could explain this difference. However, present results would suggest an alternative stimulatory mechanism independent of androgen production since the present test system does not include theca cells and androgen substrate is provided. Perhaps a direct stimulus in aromatase activity has occurred. The lower secretion of induced by day 1 2 ADF FF without inh and FFC could be due to other products such as TGF-a, hsp90, IGFBPs, etc. Rouillier et al. (1997) reported that addition of bFF to granulosa cell culture suppressed FSH-induced production, and that immunoneutralization of TGF-a in bFF restored Ej secretion to levels obtained in the absence of bFF, but these effects are not mediated through the FSH receptor (Legault et al., 1997). Concentrations of TGF-a in the pools of FF from day 5 dominant and day 1 2 atretic dominant follicles used in this experiment were measured by radioimmunoassay (Dr. Gregory S. Schultz, Department of Obstetrics and Gynecology, University of Florida). Day 12 FF had a higher concentration (5.41 ng/ml) than day 5 FF (2.53 ng/ml). It is possible that TGF-a present in bFF adversely influenced secretion of Ej (Rouillier etal., 1997). The decrease in Ej secretion by dose 2 of day 5 DF inh-eluate could be due to higher concentrations of the 49 kDa monomeric form that is present in higher amounts in follicular fluid of day 5 DF compared to day 12 ADF (de la Sota, 1995). This difference in concentration as determined by quantification of immunoblots utilizing different antiserum (anti bovine inhibin a^^'^® gly-tyr raised in mink; de la Sota, 1995) from the present experiment (anti bovine inhibin a^.^'^^ gly-tyr raised in

PAGE 193

184 rabbit; Good etal, 1995). It has been suggested (Sato etal., 1982; Schneyeretal., 1991 ) that an a-inhibin form could bind to FSH binding sites on granulosa cells and interfere with the stimulatory action of FSH. Bovine antral granulosa cells produced less P4 in response to the higher 10 ng/ml dose of CY-FSH in the presence of different FF fractions from day 5 DF and day 12 atretic dominant follicles. It is likely that the lower secretion of P4 was associated with a higher estrogenic capacity of the cells in response to a higher dose of CY-FSH in culture. Estradiol:P4 ratio is generally regarded as an index of estrogenic capacity of granulosa cells. Higher ratios are associated with higher estrogenic potential of the cells. The increase in E2;P4 ratio or enhanced estrogenic activity when exposed to 10 ng/ml of CY-FSH occurred in incubation regardless of whether fractions were from day 5 dominant or day 12 atretic dominant follicles (Figure 6-6A and B). Evidence for autoregulation involving intrafollicular components is supported by the observation that inh-eluate of day 12 ADF stimulated E2 secretion. However, the effect of these immunoactive inhibins were not expressed when mixed with whole follicular fluid (e.g., FFC) that contained components that were inhibitory to Ej secretion (e.g., FF without inh day 12 ADF). Implications There are differences among the different fractions of FF from day 5 DF and day 1 2 ADF to regulate Ej secretion or aromatase activity in bovine antral granulosa

PAGE 194

185 cells. These differences are attributed particularly to the different forms of inhibin and their proportions in follicular fluid of at day 5 and day 12. There is an inhibinlike factor in FF of day 12 ADF that stimulates E2 secretion. This may be the dimehc 34 kDa enriched form of inhibin. There is an alternative form of inhibin (49 kDa form, perhaps) at day 5 and possible other factors (e.g., hsp90, EGFrTGF-a) that inhibit Ej secretion in response to 10 ng/ml of CY-FSH. The balance of these factors may control estrogenic activity of the dominant follicle in response to gonadotropins during different functional stages of follicle dominance.

PAGE 195

CHAPTER 7 GENERAL DISCUSSION AND CONCLUSIONS A precise understanding of the processes involved in the growth and differentiation of ovulatory follicles in cattle is a prerequisite for development of new techniques and improvement of already existing regimes for estrus synchronization, ovulation, and superovulation. In cattle, folliculogenesis can be divided into two parts according to the follicular requirements for gonadotropins: a basal folliculogenesis which can proceed in the absence of gonadotropins, and a tonic folliculogenesis which requires gonadotropins (Driancourt, 1991). Follicular development in cattle occurs in waves, usually involving either two or three waves per estrous cycle (Pierson and Ginther, 1984; Sirois and Fortune, 1988; Savio et al., 1988). Each wave consists of three phases: recruitment, selection and dominance. During this last phase one follicle grows and through the secretion of endocrine and paracrine factors inhibits the growth of subordinate follicles in the cohort. The mechanism by which a dominant follicle is selected and continues to grow, while subordinate follicles regress, is unknown. Also the mechanism(s) by which the dominant follicle regulates growth of the subordinate follicles is not clear. A series of experiments were designed to test , in vivo, the effects of ablation 186

PAGE 196

187 of the first wave dominant follicle on dynamics of follicular growth and CL development and follicle ablation on follicle development and embryo production in response to a superovulatory treatment. A bovine granulosa cell culture system was developed to test, in vitro, the effects of various inhibin forms present in follicular fluid from day 5 dominant and day 12 atretic dominant follicles on estradiol secretion of antral granulosa cells. The dominant follicle of the first wave was chosen because is one of the most predictable events in follicular dynamics during the estrous cycle in cattle (Badinga et al., 1992; Fortune, 1993). In chapters 3 and 4, the effect of the ablation of the first wave dominant follicle was examined under two different scenarios. In chapter 3, the dominant follicle ovulated after an injection of 3,000 lU of hCG. An accessory CL was formed in all heifers that received hCG, and as a consequence, P4 levels between days 9 and 17 of the estrous cycle were higher than in the control group. This difference was probably due to a greater growth of the original CL in treated heifers as a consequence of the luteotropic effect of hCG (Schmitt et al., 1996a) and additional P4 secretion by the accessory CL (Schmitt et. Al. , 1 996b). There were no carry-over effects of the hCG injection that influenced CL development during the following estrous cycle. Human Chorionic Gonadotropin acutely affected follicular dynamics by ovulating the first wave dominant follicle on day 5. All hCG-treated heifers had three follicular waves during the expenmental estrous cycle. Recruitment and emergence of the second wave dominant follicle was earlier. It has been shown that a surge of FSH precedes emergence of a follicular wave (Adams et al., 1992), and removal

PAGE 197

188 of the ovary bearing the dominant follicle triggers an increase in FSH concentrations (Badingaetal , 1992) Removal ofthe dominant follicle advances emergence of the next follicular wave (Koetal., 1991). In the present experiment, size and lifespan of the second wave dominant follicle were reduced. This may be due to higher plasma P, concentrations which will induce follicle turnover (Savio et al., 1993; Burke etal., 1994; Thatcher et al.. 1994a; Kinder etal., 1996) due to a reduction in LH secretion (Kinder et al.. 1996). As a consequence ofthe shorter lifespan ofthe second wave dominant follicle, the third wave dominant follicle emerged earlier and lasted longer in the ovary before ovulation. Dominant follicles were selected on days 3 or 4 of their respective waves, as reported earlier (Ginther et al., 1996). Heifers with three spontaneous follicular waves during the estrous cycle had greater plasma P4 concentrations and an estrous cycle with a longer duration than heifers with two spontaneous follicular waves. Higher concentrations of P4 could contribute to a higher turnover of the second dominant follicle as a consequence of negative feedback of P, on LH secretion that decreases LH pulsatility (Kinder et al., 1 996) and permits development of a third follicular wave. The duration ofthe luteal phase appears to determine the number of follicular waves during an estrous cycle (Fortune, 1994). Fertility may be regulated by the number of follicular waves during the estrous cycle. Heifers with three waves had higher P^ concentrations at the time luteolysis would occur and this may antagonize the uterine luteolytic mechanism. Reproductive management systems to optimize conception rates should take in account the possibility of increasing P4 levels or the numbers of waves within the

PAGE 198

189 estrous cycle as an approach to increase fertility. The effects of higher P4 concentrations at the time of luteolysis, or an effect of P4 on the oocyte that will be ovulated, need further investigation. Functional dominance is regulated by a combination of endocrine, paracrine and autocrine factors that permit the dominant follicle to exert interovarian, intraovanan and intrafollicular effects. Clearly treatment with hCG or LH terminated functional dominance of the first wave follicle due to its induced ovulation. Following the loss of functional dominance, a newly recruited follicular wave occurred much earlier than the spontaneous second wave follicle of the control group. Additional evidence for endocrine regulation of follicle development was the decrease in both morphological dominance, in which the second wave follicle did not grow as large, and functional dominance in which turnover of the second wave follicle and emergence of the third wave occurred earlier compared to the control group. These effects were attributed to higher feedback inhibition of progesterone on LH secretion. Ovarian status at the time of superovulatory treatment has been postulated as a major factor determining ovarian responses (Monniaux et al., 1983). The effects of the presence of a dominant follicle at the time of initiation of a superovulatory regime has been controversial. Some researchers (Grasso et al., 1989; Guilbaultetal., 1991; Savio etal., 1991; Huhtinen etal., 1992; Bungartzand Niemann, 1994; Wolfsdorf et al., 1997) found that initiation of a superovulatory regime in the absence of a dominant follicle is beneficial. However, others (Gray et al., 1992; Rajamahendran and Calder, 1993, Stock et al., 1993) did not find

PAGE 199

190 differences in yield of transferable or total number of embryos when FSH treatment began during the period of morphological regression (Gray et al., 1992) or after ovulation of the dominant follicle (Rajamahendran and Calder, 1993, Stock et al., 1993). In the expenment described in Chapter 4, there were no differences in response to FSH induction of superovulation between groups with a persistent functionally dominant follicle present and the group with a persistent functionally dominant follicle that was ablated by aspiration during treatment. Numbers of ovulatory follicles, CL, and numbers of total and transferable embryos did not differ between groups. The persistent dominant follicle model is an alternative to study the effects of a dominant follicle during follicular development (Sirois and Fortune, 1 990; Savio et al., 1993b). In this study (Chapter 4), presence and duration of the dominant follicle were experimentally controlled through induction of a persistent dominant follicle. A similar model was utilized by Wehrman et al. (1996), who found a reduced number of CL. total ova and transferable embryos when FSH treatment started 5 days after creation of a persistent dominant follicle. In the present experiment, the dominant follicle was present in the ovary until day 1 0 of the estrous cycle, and the FSH treatment began 2 days after aspiration of the dominant follicle. The persistent dominant follicle inhibited recruitment in the control group, whereas in the group without the dominant follicle recruitment occurred 2 days earlier. Likewise number of class 3 follicles increased earlier in the treated group. This difference in follicular dynamics indicates that there is a factor secreted by the dominant follicle that inhibits recruitment and growth of follicles on both ovaries.

PAGE 200

191 However, this effect can be overridden with continued injection of high doses of FSH over a 3 day penod. Secretion of inhibins may be the inhibitory component but such an effect would likely be at the interovarian level since any decrease in basal FSH concentrations would have a minimal effect compared to the large doses of FSH (3 to 5 mg) injected over a 3 day period. There is additional evidence that a factor (s) secreted by the dominant follicle influences growth of subordinate follicles in the cohort. Passive immunization of ewes (O'Shea etal., 1991; Schanbacheretal., 1991) or heifers (Price et al., 1987; Glencross et al., 1992; Morris et al., 1993; Scanlon et al., 193) against steroid freefollicular fluid or against synthetic inhibin peptides increased ovulation rates. Follicular dominance appears to be controlled by a number of mechanisms acting in concert. Decreases in spontaneous FSH concentrations in plasma is a consequence of exerting a negative feedback on FSH secretion and the action of inhibin. The decrease in FSH may decrease development of subordinate follicles (Kg et al., 1991; Adams et al., 1992a; Fortune, 1994). Secretion of local intrafollicular paracrine factors also may inhibit development of subordinate follicles (Campbell et al., 1995, Armstrong and Webb, 1997) and play a role in manifesting suppressed follicular growth of subordinate follicles during the dominance phase. It has been shown that inhibin-like molecules have regulatory actions within the ovary that regulate the synthesis of through the stimulation in synthesis of androstenedione a precursor for Ej synthesis (Hillier et al., 1991b; Woodruff et al., 1 990; Wrathall and Knight, 1 993). In addition to inhibin stimulating androstenedione production by thecal cells, perhaps inhibin may alter Ej syntheis by granulosa cells.

PAGE 201

192 In order to test the latter hypothesis, an in vitro system for the culture of granulosa cells to produce E2 was developed (Chapter 5). Granulosa cells, through the secretion of E2, influence the endocnne status of the female (Skinner and Osteen, 1988). Granulosa cells are the follicular component that synthesize and secrete Ej, by conversion of androgens into estrogens. Within the follicle, androgens are provided by theca cells to be aromatized into estrogens by granulosa cells. In the literature, there is considerable controversy regarding the ability to maintain Ej secretion in bovine granulosa cells cultured in vitro. In most cases, granulosa cells luteinize in culture when exposed to high doses of FSH (Berndtson et al., 1995; Kuran et al., 1996). The system validated in Chapter 5 indicated that antral bovine granulosa cells are estrogenic after 96 h in culture but Ej production depended upon the type of FSH used to stimulate the cells. The response of antral bovine granulosa cells depended upon the type of FSH used to stimulate the cells. A highly purified porcine FSH preparation (CY-FSH) was able to stimulate granulosa cells to produce Ej. The production of E2 decreased with time in culture in agreement with others (Berndtson et al., 1995; Rouillier et al., 1996) and most likely reflects a reduction in the activity of the aromatase enzyme (Skinner and Osteen, 1988; Rouillier et al., 1996). The response of granulosa cells to different doses of CY-FSH varied with time. At 48 h of culture, granulosa cells responded to low doses of FSH (e.g., 0.5, 1, and 2 ng/ml). This ability was lost after 72 h, and higher doses of CY-FSH (e.g., 6, 8, and 10 ng/ml) were necessary to stimulate an estradiol response. In contrast, a dose of 50 ng/ml of CY-FSH elicited an inhibitory Ej response. Perhaps this reduction

PAGE 202

193 was due to a down regulation of the FSH receptor. Cells treated at 48 h with CYFSH had a quadratic secretory dose response to CY-FSH with a maximal response at 2 ng/ml of CY-FSH. Ovine (Langhout et al., 1 991 ; Spicer an Alpizar, 1 994; Berndtson et al., 1 995; Kuran et al., 1995, 1996; Spicer and Stewart, 1996), porcine (Saumande, 1991; Rouiilier et al., 1996), and bovine (Gutierrez et al., 1997) FSH preparations have been used to stimulate granulosa cells in culture. In the present study, responses to ovine and porcine FSH were different, but also the response to the porcine CYFSH from France was greater than the porcine FSH from Belgium. These variations are possibly due to presence of other proteins (e.g., presence of BSA as a carrier in the USDA-ovine FSH preparation) that could affect the biological action of FSH on granulosa cells, or BSA itself stimulating luteinization and proliferative responses due to presence of growth factors and fatty acids in the BSA. Also FSH exists as a family of isohormones with differences in their oligosacharide structures including the degree of sialylation and/or sulfation (Ulloa-Aguirre et al., 1995). These differences alter the metabolic clearance rates, the biological and immunological potencies, and the ability of the isoforms to bind the FSH receptor (Ulloa-Aguirre et al., 1984). Perhaps differences in oligosacharide structure among the different sources of FSH utilized in this experiment (Chapter 5) can account for differences in their ability to bind the FSH receptor and elicit an estrogenic response by bovine antral granulosa cells in culture. Differences in tertiary structure may have had differential effects on sensitivity to lyophylization and subsequent biological activity. Bovine antral granulosa cells did not show a clear response to different

PAGE 203

194 doses of FSH in terms of P4 secretion. Granulosa cells tend to luteinize with time in culture (Skinner and Osteen, 1 988). However, cells treated with CY-FSH secreted decreasing amounts of P4 per cell with time in culture, whereas preparations of BFSH and USDA-FSH increased P4 production by granulosa cells. Shape and number of the granulosa cells was influenced by the different sources of FSH. Cells treated with CY-FSH had a round shape and grew in multilayered aggregates, as reported for the rat (Knecht et al., 1981) and cow (Gutierrez et al., 1 997). However, granulosa cells treated with B-FSH and USDAFSH grew in monolayers and assumed a fibroblastic appearance. The number of cells decreased with time in culture. At 48 h there was an appreciable loss in cell numbers for all FSH treatments with fewer cells at each time compared to number of cells seeded. The ability of cells to remain plated is probably related to the shape they assume in culture. Round cells have less area of contact with the plate, hence cells are more prone to be lost with changes in culture medium at each 24 h period. Skinner and Osteen (1 988) reported that granulosa cells isolated from large follicles, as is our case, had a slightly lower plating efficiency and a greater loss of cells between 3 and 6 days of culture. Although cell numbers increased with time for USDA-FSH, it was not determined whether this increase reflected true proliferation or less cell loss. In recent reports, bovine (Gutierrez et al., 1997) and ovine granulosa cells (Campbell et al., 1996) were shown to proliferate in culture. Based on the ability of bovine antral granulosa cells to secrete estradiol in response to CY-FSH in culture, it was considered that this was an appropriate in

PAGE 204

195 vitro experimental model to examine potential regulatory factors from the dominant follicle that alter E2 secretion of granulosa cells. Emergence of a follicular wave is controlled by various factors. Folliclestimulating hormone is responsible for the emergence of each follicular wave. Once the dominant follicle is selected and starts to secrete increasing amounts of Ej, circulating concentrations of FSH decrease, which suggests that the dominant follicle secretes a factor (s) that in an endocrine manner decreases secretion of FSH from the pituitary (Figure 7-1). It is known that inhibin is a slow but powerful inhibitor of pituitary FSH biosynthesis and secretion (Gaddy-Kurten et al., 1995). It also has been suggested that an a-inhibin form can bind to FSH binding sites on granulosa cells (Satoetal., 1982; Schneyeretal., 1991)and inhibit the stimulatory action of FSH. In addition, there are changes in amounts and proportions of the different molecular forms of inhibin (de la Sota, 1995; Sunderland et al., 1996) and expression of inhibin/activin mRNA for a, and Q>q subunits (Ireland and Ireland, 1994) that are related to physiological stages of follicle growth (Guilbault et al., 1993; de la Sota, 1995; Sunderland et al., 1996). Dominant follicles have different amounts of monomeric and dimeric forms of inhibin according to their stage of development. De la Sota (1995) reported differences between day 5 DF and day 12 ADF in amounts and proportions of monomeric and dimeric forms of inhibin. Based on these findings and the hypothesis that inhibin can bind to the FSH receptor of granulosa cells, the experiment in Chapter 6 examined the effect of follicular fluid complete from day 5 dominant and day 12 atretic dominant follicles

PAGE 205

196 and respective follicular fluid fractions prepared from an antiinhibin immunoaffinity column on E2 secretion by bovine antral granulosa cells. It is reasonable that regulatory factors produced by dominant follicles would be present in the follicular fluid. It is probable that large molecular weight factors such as the dimeric forms of inhibin (> 34 kDa) in follicular fluid were produced by granulosa cells since the basement membrane would separate thecal cells from granulosa cells. Thus factors within the follicular fluid would represent an array of molecules that are candidates for intrafollicular regulation of granulosa cells, and interovarian regulators if secreted from the ovary. One family of factors may be the inhibin family of proteins described in cattle (Good et al., 1995; de la Soata, 1995). Utilizing the antiinhibin immunoaffinity column prepared as described in Chapters, two regulatory components of follicular fluid could be prepared. The inhibins-eluate represents the family of inhibin proteins, and the follicular fluid eluate without inhibins represents the remaining factors in follicular fluid that may potentially regulate estradiol secretion by granulosa cells. The preparation of the follicular fluid pools prior to fractionation eliminated the presence of steroids such as estradiol and progesterone. Results from granulosa cell cultures (Chapter 6) showed that follicular fluid from day 5 DF appears to have factors that inhibit Ej secretion. These factors are present in the fraction without inhibins and in the inhibin-enriched fraction. In the follicular fluid fraction without inhibins, other factors may be present such as TGF-a (Rouillier et al., 1997), IGFBPs (Monget and Monniaux, 1995), tissue inhibitor of metalloproteinase (Roche, 1996), and hsp90 (Roche, 1996; Driancourt, personal

PAGE 206

197 communication) that may inhibit secretion. The inhibin-enriched fraction had a 49 kDa form of immuno-reactive inhibin that could be responsible for the inhibition of E2 secretion. This 49 kDa form was shown to be a monomeric a-form (Good et al., 1995; Sunderland et al., 1996) and may bind to the FSH receptor since has been shown that ainhibin can bind to FSH binding sites on rat granulosa cells (Scheneyer et al., 1991). These inhibitory effects were not observed with a dose of 0.5 ng/ml of CY-FSH. Perhaps these inhibitors in follicular fluid without inhibin and inhibins-eluates compete for the FSH receptor or attenuate components of the signal transduction system to inhibit aromatase activity and Ej secretion. In day 12 follicular fluid fractions, the inhibins-eluate fraction of the day 12 ADF had an opposite stimulatory effect on E2 secretion. However, the follicular fluid without inhibins fraction inhibited Ej secretion by the granulosa cells. Proportions of the mature 34 kDa form are high in follicular fluid of day 12 ADF (de la Sota, 1995; Sunderland et al., 1 996), and it has been hypothesized (Baird and Smith, 1 993) that increased secretion of Ej is in part dependent on the stimulation by inhibin of LHinduced androgen synthesis by theca cells. However, this explanation may not apply to the present experimental results because the culture system lacks the thecal component of the follicle. A direct stimulatory effect of the inhibins-eluate the FSH signal transduction system leading to an increase in aromatase enzyme activity. Progesterone secretion was decreased by 1 0 ng/ml of CY-FSH. The higher E2:P4 ratio in culture media from granulosa cells treated with 10 ng/ml of CY-FSH

PAGE 207

198 indicated that granulosa cells had a higher estrogenic potential. This effect was present regardless of whether the fractions were from day 5 DF or day 12 ADF. In summary, the effect of the dominant follicle is exerted at different levels. At an endocrine level, inhibin decreases synthesis and secretion of FSH from the pituitary (Figure 7-1). At the interovarian level, secretion of an endocrine factor (e.g., inhibin forms) act on both ovaries to alter secretion (Figure 7-1), and at the intraovarian level through the action of inhibin forms within the follicle on granulosa cells to alter E2 (Figure 7-2 and 7-3). These factors appear to work in an orchestrated manner in order to control follicular development. Figure 7-1 . Endocrine and iterovarian regulation of estradiol secretion Estradiol and inhibin complement each other in regulating the secretion of FSH. Both hormones, secreted in higher amounts by the dominant follicle and

PAGE 208

199 lesser amounts by subordinate follicles, exert a negative feedback on the hypothalamus (E2) and hypophysis (inhibin) to decrease FSH release (Baird and Smith, 1993). In addition to and inhibin, the dominant follicle also secretes activin and follistatin. Activin stimulates the secretion of FSH from the pituitary, whereas follistatin acts through the binding of activin to abolish the stimulatory effect of activin. Therefore, the decrease in FSH secretion compromises development of subordinate follicles. During the luteal phase the corpus luteum secretes progesterone, which exerts a negative feedback on the secretion of LH from the adenohypophysis. The decrease in concentrations of LH negatively affects the development of the dominant follicle leading to atresia (Sirois and Fortune. 1990; Savioetal, 1993b). . i. Theca cell compartment Granulosa cell compartment Figure 7-2. intrafollicuiar and interovarian effects of inhibin on estradiol secretion

PAGE 209

200 The interovarian effect of the dominant follicle appears to be through the action of inhibin in an endocrine manner. Follicle-stimulating hormone acts on the granulosa cells of the dominant follicle to induce aromatase activity, inhibin production and LH receptors. Inhibin secreted by the dominant follicle acts on subordinate follicles within the same ovary and the contralateral ovary (Figure 7-2). The action of inhibin on subordinate follicles may be change the regulation of Ej synthesis. It has been shown that inhibin stimulates the LH-induced synthesis of androgen by theca cells (Findlay, 1993; Hillier et al.. 1991; Wrathall et al., 1995). Activin prevents luteinization through a negative effect on progesterone synthesis (Findlay, 1993) by the theca cells and stimulatory effects on E2 synthesis in granulosa cells (Baird and Smith, 1993; Figure 7-2). A potential model that could explain the regulatory mechanisms at the level of the granulosa cells that control dominance in cattle is depicted in Figure 7-3. When FSH binds to its receptor on the granulosa cell, there is an activation of the cAMP/protein kinase A and other cAMP-dependent pathways. As a consequence, there is stimulation of the aromatase enzyme, that synthesizes estrogens from thecal androgens. The inhibitory factors (e.g., an inhibin form) secreted by the dominant follicle could bind to the FSH receptor on the granulosa cell and uncouple the FSH signal transduction system. This would lead to a decrease in aromatase activity due to decreased gene expression of aromatase and decreased aromatase activity in the dominant follicle. If this factor is secreted it has the ability to act on subordinates follicles in a comparable manner to decrease follicle growth and E2 secretion. There are other factors (e.g., TGF-a, IGFBPs, hsp90) that are secreted

PAGE 210

201 Figure 7-3. Regulation of steroidogenesis in thecal and granulosa cells into the follicular fluid which may affect the granulosa cell physiology. The secretion of these factors may interact with inhibin forms to alter Ej secretion. For example the stimulatory effects of inhibins-eiuate in day 12 follicular fluid on secretion was not evident when mixed with the fraction of follicular fluid without inhibins that contained inhibitory components to Eg secretion. Such a mixture was considered to be comparable to follicular fluid complete which inhibited Ej secretion. This is not to be unexpected when it is known that atretic follicles have high concentrations of IGFBP-2, -4 and -5 (de la Sota, 1995) that may sequester IGF-I and IGF-II and

PAGE 211

202 prevent their stimulatory effects on granulosa cells (e.g., positive effects on aromatase). It has been shown (Legault et al., 1997) that TGF-a inhibits FSHinduced secretion of granulosa cells in culture. This is done through the activation of protein kinase C pathway that inhibits the formation of cAMP. In summary, several factors interplay to control follicular development in cattle. These factors act in an endocrine, interovarian, intraovarian and intrafollicular fashion. One of the principal and strongest candidates are the different inhibin forms secreted by the dominant follicle. Whereas the dimeric forms are important in the control of the FSH release by the anterior pituitary, some of the a-inhibin forms have been postulated to act at the level of the granulosa cells inhibiting the binding of FSH to its receptor.

PAGE 212

REFERENCES Abeyawardene, S. A. and G. S. Pope. 1 987. The involvement of progesterone and luteinizing hormone in the termination of the post-ovulatory rise in plasma oestradiol-17B concentration in cattle. Anim. Reprod. Sci. 15:27-36. Adams, G. P. 1994. Control of ovarian follicular wave dynamics in cattle: implications for synchronization & superstimulation. Theriogenology. 41:19-24. Adams, G. P., K. Kot, C. A. Smith and O. J. Ginther. 1993a. Effect of the dominant follicle on regression of its subordinates. Can. J. Anim. Sci. 73:267-275. Adams, G. P., K. Kot, C. A. Smith and O. J. Ginther. 1993b. Selection of a dominant follicle and suppression of follicular grov\/th in heifers. Anim. Reprod. Sci. 30:259-271. Adams, G. P., R. L. Matteri and O. J. Ginther. 1992a. Effect of progesterone on ovarian follicles, emergence of follicular waves and circulating follicle-stimulating hormone in heifers. J. Reprod. Fertil. 95:627-640. Adams, G. P., R. L. Matteri, J. P. Kastelic, J. C. Ko, and O. J. Ginther. 1992b. Association between surges of follicle-stimulating hormone and the emergence of follicular waves in heifers. J. Reprod. Fertil. 94:177-188. Alvarez, R. H., J. B. P. Carvalho, A. A. Rosa e Silva, C. N. Perone, M. T. C. P. Ribela and E. B. Oliveira Filho. 1997. Endocrine profiles and ovulation rate of cows superovulated with FSH following passive immunization against steroid free-bovine follicular fluid. Theriogenology. 47:164 (Abstr.). Amsterdam, A. and S. Rotmensch. 1987. Structure-function relationships during granulosa cell differentiation. Endoc. Rev. 8:309-337. Amsterdam, A., S. Rotmensch and A. Ben-Ze'ev. 1989. Coordinated regulation of morphological and biochemical differentiation in a steroidogenic cell: the granulosa cell model. TIBS. 14:377-382. 203

PAGE 213

204 Armstrong, D. T. 1993. Recent advances in superovulation of cattle. Theriogenology. 93:7-24. Armstrong, D. G., G. Baxter, C. O. Hogg and R. Webb. 1996. Insulin-like growth factor binding protein-2 and -4 mRNA expression in bovine follicles in vivo: the effect of follicle size and atresia. J. Reprod. Fertil. Abstract Series. 18: abstract 72. Armstrong, D. G. and R. Webb. 1997. Ovarian follicular dominance: the role of intraovarian growrth factors and novel proteins. Rev. Reprod. 2:139-146. Badinga, L., M. A. Driancourt, J. D. Savio, D. Wolfenson, M. Drost, R. L. de la Sota, and W. W. Thatcher. 1 992. Endocrine and ovarian responses associated with the first wave dominant follicle in cattle. Biol. Reprod. 47:871-883. Baird, D. F. and K. B. Smith. 1993. Inhibin and related peptides in the regulation of reproduction. Oxf. Rev. Reprod. Biol. 15:191-232. Baranao, J. L. S. and J. H. Hammond. 1985. Multihormone regulation of steroidogenesis in cultured porcine granulosa cells: studies in serum-free medium. Endocrinology. 116:2143-2151. Barano, J. L. S. and J. H. Hammond. 1985. Serum-free medium enhances growth and differentiation of cultured pig granulosa cells. Endocrinology. 1 16:51-58. Beard, A. J., D. Savva, R. G. Glencross, B. J. McLeod and P. G. Knight. 1989. Treatment of ovariectomized heifers with bovine follicular fluid specifically suppresses pituitary levels of FSH-3 mRNA. J. Mol. Endocrinol. 3:85-91. Ben-Ze'ev, A. and A. Amsterdam. 1987. In vitro regulation of granulosa cell differentiation: involvement of cytoskeletal protein expression. J. Biol. Chem. 262:5366-5376. Berndtson, A. K., S. E. Vincent and J. E. Fortune. 1995. Low and high concentrations of gonadotropins differentially regulate hormone production by theca interna and granulosa cells from bovine preovulatory follicles. Biol. Reprod. 52:1334-1342. Besnard, N., C. Pisselet, D. Monniaux, A. Locatelli, F. Benne, F. Gasser, F. Hatey and P. Monget. 1996a. Expression of messenger ribonucleic acids of insulin-like growth factor binding proteins -2, -4 and -5 in the ovine ovary: localization and changes during growth and atresia of antral follicles. Bio. Reprod. 55:1356-1367. Besnard, N., C. Pisselet, J. Zapf, W. Hornbeck, D. Monniaux and P. Monget. 1996b. Proteolytic activity is involved in changes in intrafollicular insulin-like growth

PAGE 214

205 factor binding protein levels during growth and atresia of ovine follicles. Endocrinology. 137:1599-1607. Bo, G. A., G. P. Adams, R. A. Pierson, M. Caccia, H. Tribulo and R. J. Mapletoft. 1994. Follicular wave dynamics after estradiol-17p treatment of heifers with or without a progestogen implant. Theriogenology. 41:1555-1569. Borenfreund, E. and J. A. Puerner. 1984. A simple quantitative procedure using monolayer cultures for cytotoxicity assays (HTD/NR-90). J. Tissue Culture Meth. 9:7-9. Breuel, K. F., J. C. Spitzer, and D. M. Hendricks. 1989. Systemic progesterone concentration following human chorionic gonadotropin administration at various times during the estrous cycle in beef heifers. J. Anim. Sci. 67:1564-1572. Bungartz, L. and H. Niemann. 1994. Assessment of the presence of a dominant follicle and selection of dairy cows suitable for superovulation by a single ultrasound examination. J. Reprod. Fertil. 101:583-591. Burke, C. R., M. Mihm, K. L. Macmillan and J. F. Roche. 1994. Some effects of prematurely elevated concentrations of progesterone on luteal and follicular characteristics during the oestrous cycle in heifers. Anim. Reprod. Sci. 35:27-39. Cahill L.P., P. Mauleon. 1981. A study of the population and small follicles in the sheep. J. Reprod. Fertil. 61: 201-206. Campbell, B. K., R. J. Scaramuzzi and R. Webb. 1995. Control of antral follicle development and selection in sheep and cattle. J. Reprod. Fertil. Suppl. 49:335350. Campbell, B. K., R. J. Scaramuzzi and R. Webb. 1996. Induction and maintenance of oestradiol and immunoreactive inhibin production with FSH by ovine granulosa cells cultured in serum-free media. J. Reprod. Fertil. 106:7-16. Chenault, J. R., W. W. Thatcher, P. S. Kaira, R. M. Abrams and C. J. Wilcox. 1975. Transitory changes in plasma progestins, estradiol, and luteinizing hormone approaching ovulation in the bovine. J. Dairy Sci. 58:709-717. Chenault, J. R., W. W. Thatcher, P. S. KaIra, R. M. Abrams and C. J. Wilcox. 1976. Plasma progestins, estradiol, and luteinizing hormone following prostaglandin F2 alpha injection. J. Dairy Sci. 59:1342-1346. De Krester, D. M. and D. M. Robertson. 1989. The isolation and physiology of inhibin and related hormones. Biol. Reprod. 40:33-47.

PAGE 215

206 De la Sota, R. L. 1995. Endocrine and molecular changes associated with ovarian follicular dominance in cattle PhD dissertation University of Florida. De la Sota, F. A Simmen, T Diaz and W. W. Thatcher. 1996. Insulin-like growth factor system in bovine first-wave dominant and subordinate follicles. Biol. Reprod. 55:803-812 DeJong, F. H. 1988. Inhibin Physiological Rev. 68:555-607. Diaz. T., M. -J. Thatcher, J. D. Ambrose, M. Binelli, M. Drost and W. W. Thatcher. 1997. Response of bovine antral granulosa cells to different types of FSH. Biol. Reprod. 56(Suppl. 1):144 (Abstr.). Diaz, T , E. J. -P. Schmitt, R. L. de la Sota and W. W. Thatcher. 1993. hCGinduced alterations in ovarian follicular dynamics during the estrous cycle of heifers. J. Anim. Sci. 71(Suppl. 1):217 (Abstr.). Dorrington, J. H. and D. T. Armstrong. 1979. Effect of FSH on gonadal functions. Recent Prog. Horm Res 35:301-342. Dorrington. J H Y S Moon and D. T. Armstrong. 1975. Estradiol-17li biosynthesis in cultured granulosa cells from hypophysectomized immature rats; stimulation by follicle stimulating hormone. Endocrinology. 97:1328-1331. Driancourt, M. A. 1991 . Follicular dynamics in sheep and cattle. Theriogenology. 35:55-79. Driancourt, M. A . A. Gougeon, D Royere and C. Thibault. 1993a. Ovarian function. In: Reproduction in mammals. C. Thibault, M. C. Levasseurand R. H. F. Hunter, eds. Ellipses. Pans, pp 281-305 Driancourt, M. -A., D. Royere, B. Heon and M. C. Levasseur. 1993b. Oestrousand menstrual cycles. In: Reproduction in mammals. C. Thibault, M. C. Levasseurand R. H. F. Hunter, eds. Ellipses. Paris, pp. 589-603. Driancourt, M. A., W. W. Thatcher, M. Terqui and D. Adrieu. 1991. Dynamics of ovarian follicular development in cattle dunng the estrous cycle, early pregnancy and in response to PMSG. Domestic Anim. Endocrinol. 82:209-221. Echternkamp, S. E., H J Howard, A. J. Roberts, J. Grizzle and T. Wise. 1994. Relationships among concentrations of steroids, insulin-like growth factor-!, and insulin-like growth factor binding proteins in ovarian follicular fluid of beef cattle. Biol. Reprod. 51:971-981

PAGE 216

207 Elsden R. P., L. D. Nelson and G. E. Seidel. 1978. Superovulation of cows with follicle stimulating hormone and pregnant mare's serum gonadotrophin. Theriogenology. 9:17-26. Erickson, B.H.I 966. Development and senescence of the postnatal bovine ovary. J. Anim. Sci. 25:800-805. Findlay. J.K.I 993. An update on the roles of inhibin, activin, and follistatin as local regulators of folliculogenesis. Biol. Reprod. 48:15-23. Findlay, J. K., S. Xiao, L. Shukovski and U. Michel. 1993. Novel peptides in ovarian physiology: inhibin, activin and follistatin. In: E. Y. Adashi and P. C. K. Lueng, eds. The Ovary. New York, Raven Press, pp. 413-432. Fogwell, R. L., J. L. Cowley, J. A. Wortman, N. K. Ames, and J. J. Ireland. 1985. Luteal function in cows following destruction of ovarian follicles at midcycle. Theriogenology 23:389-398. Fortune, J. E. 1984. Effects of prolactin on steroidogenesis by theca and granulosa cells from bovine proestrus follicles. Biol. Reprod. 30 (SuppI 1):38 (Abstr.). Fortune, J. E. 1986. Bovine theca and granulosa cells interact to promote androgen production. Biol. Reprod. 35:292-299. Fortune, J. E. 1993. Follicular dynamics during the bovine estrous cycle: a limiting factor in improvement of fertility? Anim. Reprod. Sci. 33:111-125. Fortune, J. E. 1994. Ovarian follicular growth and development in mammals. Biol. Reprod. 50:225-232. Fortune, J. E. and D. T. Armstrong. 1977. Androgen production by theca and granulosa isolated from proestrus rat follicles. Endocrinology. 100:1341-1347. Fortune, J. E. and D. T. Annstrong. 1978. Hormonal control of 17S-estradiol biosynthesis in proestrus rat follicles: estradiol production by isolated theca versus granulosa. Endocrinology. 102:227-235. Fortune, J. E. and W. Hansel. 1985. Concentrations of steroids and gonadotropins in follicular fluid from normal heifers and heifers primed for superovulation. Biol. Reprod. 32:1069-1079. Fortune, J. E. and S. M. Quirk. 1988. Regulation of steroidogenesis in bovine preovulatory follicles. J. Anim. Sci. 66 (SuppI. 2): 1-8.

PAGE 217

208 Fortune, J. E., J. Sirois and S. M. Quirk. 1988. The growth and differentiation of ovarian follicles during the bovine estrous cycle. Thehogenology. 29:95-109. Fortune, J. E., J. Sirois, A. M. Turzillo and M. Lavoir. 1991. Follicle selection in domestic animals. J. Reprod. Anim. Suppl. 43:187-198. Fricke, P. M., L. P. Reynolds, and D. A Redmer. 1993. Effect of human chorionic gonadotropin administered early in the estrous cycle on ovulation and subsequent luteal function in cows. J. Anim. Sci. 71:1242-1246. Gaddy-Kurten, D., K. Tsuchida and W. Vale. 1 995. Actvins and the receptor serine kinase superfamily. Rec. Progress in Horm. Res. 50:109-129. Ginsburg, M., M. H. L. Snow and A. McLaren. 1990. Primordial germ cells in the mouse embryo during gastrulation. Development. 1 10:521-528. Ginther, O. J., J. P. Kastelic, and L. Knopf. 1989a. Composition and characteristics offollicular waves during the bovine estrous cycle. Anim. Reprod. Sci. 20:187-200. Ginther, O. J., L. Knopf, and J. P. Kastelic. 1989b. Temporal associations among ovarian events in cattle during oestrus cycles with two and three follicular waves. J. Reprod. Fertil. 87:223-230. Ginther, O. J., L. Knopf and J. P. Kastelic. 1989c. Ovarian follicular dynamics in heifers during early pregnancy. Biol. Reprod. 41:247-254. Ginther, O. J., M. C. Wiltbank, P. M., Fricke, J. R. Gibbons, and K. Kot. 1996. Selection of the dominant follicle in cattle. Biol. Reprod. 55:1187-1194. Ginther, O. J., K. Kot, L. J. Kulick and M. C. Wiltbank. 1997. Emergence and deviation of follicles during the development of follicular waves in cattle. Thehogenology. 48:75-87. Giudice, L. C. 1992. Insulin-like growth factors and ovarian follicular development. Endocnne Rev. 13:641-669. Glencross, R. G., E. C. L. Bleach, B. J. McLeod, A. J. Beard and P. G. Knight. 1992. Effect of active immunization of heifers against inhibin on plasma FSH concentration, ovarian follicular development and ovulation rate. J. Endocrinol. 134:11-18. Glencross, R. G., E. C. L. Bleach, S. C. Wood and P. G. Knight. 1994. Active immunization of heifers against inhibin: effects on plasma concentrations of gonadotropins, steroids and ovarian follicular dynamics during prostaglandinsynchronized cycles. J. Reprod. Fertil. 100:599-605.

PAGE 218

209 Godin, I. and C. C. Wylie. 1991. TGF 3, inhibits proliferation and has a chemotropic effect on mouse primordial germ cells in culture. Development. 113:1451-1457. Gong, J., D. McBride, T. Bramley and R. Webb. 1993. Effects of recombinant bovine somatotrophin, insulin-like growth factor-1 and insulin on the proliferation of bovine granulosa cells in-vitro. J.Endocrinol. 139:67-75. Good. T. E. M., P. S. D. Weber, J. L. H. Ireland, J. Pulaski, V. Padmanabhan, A. L. Schneyer, G. Lambert-Messerlian, B. R. Ghosh, W. L. Miller, N. Groome and J. J. Ireland. 1995. Isolation of nine different biologically and immunologically active molecular variants of bovine follicular inhibin. Biol. Reprod. 53:1478-1488. Goodman, A. I. and Hodgen, G. D. 1983. The ovarian triad of the primate menstrual cycle. Recent Prog. Horm. Res. 39:1-73. Gore-Langton, R. and D. T. Armstrong. 1988. Follicular steroidogenesis and its control. In: E. Knobil and J. D. (Eds.) The Physiology of Reproduction. Raven Press, Ltd., New York. pp. 331-385. Gougeon, A. 1 996. Regulation of ovarian follicular development in primates: facts and hypothesis. Endoc. Rev. 17:121-155. Grasso, F., L. A. Guilbault, G. L Roy, P. Matton and J. G. Lussier. 1989. The influence of the presence of a dominant follicle at the time of initiation of a superovulatory treatment on superovulatory responses in cattle. Thenogenology. 31:199 (Abstr.). Gray, B. W., R. E. Cartee, D. A. Stringfellow. M. D. Riddell, K. P. Riddell and J. C. Wright. 1 992. The effects of FSH-priming and dominant follicular regression on the superovulatory response of cattle. Theriogenology. 37:631-639. Grazul-Bilska, A. T., L. P. Reynolds and D. A Redmer. 1997. Gap junctions in the ovaries. Biol. Reprod. 57:947-957. Greenwald, G. S. and Roy, S. K. 1994. Follicular development and its control. In: The Physiology of Reproduction. E. Knobil and J. D. Neill, eds. Raven Press, Ltd.. New York. pp:629-724. Guilbault, L. A., F. Grasso, J. G. Lussier, P. Rouillier and P. Matton. 1991. Decreased superovulatory responses in heifers superovulated in the presence of a dominant follicle. J. Reprod. Fertil. 91:81-89. Guilbault, L. A., P. Rouillier, P. Matton, R. G. Glencross, A. J. Beard, P. G. Knight. 1993. Relationships between the level of atresia and inhibin contents (a subunit

PAGE 219

210 and a-p dimer) in morphologically dominant follicles during their growing and regressing phases of development in cattle. Biol. Reprod. 48:268-276. Gutierrez, C. G.. B K Campbell and R. Webb. 1997. Development of a long-term bovine granulosa cell culture system, induction and maintenance of estradiol production, response to follicle-stimulating hormone, and morphological characteristics. Biol. Reprod. 56:608-616. Halvorson, L M. and A H DeCherney. 1996. Inhibin, activin, and follistatin in reproductive medicine. Fertil. Stenlity. 65:459-469. Heap, R B and A P F. Flint. 1984. Pregnancy. In: C. R. Austin and R. V. Short (Eds.) Reproduction in mammals 3: Hormonal control of reproduction. Cambridge University Press, Cambridge, pp. 153-. Henderson, K. M. and P. Franchimont. 1983. Inhibin production by bovine ovarian tissue in vitro and its regulation by androgens. J. Reprod. Fertil. 67:291-298. Hernandez, E. R., A Hurwitz, D. W. Payne, A M. Dharmarajan, A. Purchio and E. Y.Adashi. 1990. Transforming growth factor-G.1 inhibits androgen production: gene expression, cellular localization, mechanism(s), and site(s) of action. Endocrinology. 127:2804-2811. Hillier, S. G., E. L. Yong, P. J. Illingworth, D. T. Baird, R. H. Schwall, A. J. Mason. 1991b. Effect of a recombinant inhibin on androgen synthesis in cultured human theca cells. Molecular Cell. Endoc. 75:R1-6. Hillier, S. G. 1991 Regulatory functions for inhibin and activin in human ovaries. J. Endocrinology 131:171-175. Hillier, S G., E L Yong, P J Illingworth, D. T. Baird, R. H. Schwall and A. J. Mason. 1991a. Effect of a recombinant activin on androgen synthesis in cultured humanthecacells J. Clin Endoc. Metab 72:1206-1211. Hillier, S. G. and I M. Turner. 1991. Steroid signaling in the follicular paracrine system. In: Signaling mechanisms and gene expression in the ovary. Geula Gibori, eds Springer Verlag. New York, pp 84-96. Hilscher, W 1991 . The genetic control and germ cell kinetics of the female and male germ line in mammals including man. Hum. Reprod. 6:1416-1425. Hirshfield, A N. 1991 Development of follicles in the mammalian ovary. International Rev. Cytology. 124:43-101.

PAGE 220

211 Hirshfield, A. N. 1992. Heterogeneity of cell populations that contribute to the formation of primordial follicles in rats. Biol. Reprod. 47:466-472. Howard, H. J., R. G. Scott, and J. H. Britt. 1990. Associations among progesterone, estradiol-17B, oxytocin and prostaglandin in cattle treated with hCG during diestrus to extend corpus luteum function. Prostaglandins. 40:51-70. Hsueh, A. J. W., E. Y. Adashi, P. B. C. Jones and T. H. Welsh, Jr. 1984. Hormonal regulation of the differentiation of cultured ovahan granulosa cells. Endocrine Reviews. 5:76-127. Hsueh, A. J. W., H. Billig and A. Tsafriri. 1994. Ovarian follicle atresia: a hormonally controlled apoptotic process. Endocrine Rev. 15:707-724. Hsueh, A. J., K. D. Dahl, J. V. Vaughan, E. Tucker, J. Rivier, C. Wayne Bardine, W. Vale. 1987. Heterodimers and homodimers of inhibin subunits have different paracrine action in the modulation of luteinizing hormone-stimulated androgen biosynthesis. Proc. Natl. Acad. Sci. 84:5082-5086. Huhtinen, M., B. Rainio, J. Aalto, P. Bredbaciak and A. Maki-Tanila. 1992. Increased ovarian responses in the absence of a dominant follicle in superovulated cows. Theriogenology. 37:457-463. Hulshof, S. C. J., J. R. Figueiredo, J. F. Beckers, M. M. Bevers, H. Vanderstichele and R. van den Hurk. 1997. Bovine preantral follicles and activin: immunocytochemistry for activin and activin receptor and the effect of bovine activin A in vitro. Theriogenology. 48:133-142. Hutchinson, L. A., J. K. Findlay, F. L. de Vos and D. M. Robertson. 1987. Effect of bovine inhibin, transforming growth factor-B and bovine activin-A on granulosa cell differentiation. Biochem. Biophys. Res. Commun. 146:1405-1412. Hylka, V. M., M. K. Khaki and G. S. diZirega. 1989. Steroidogenesis of porcine granulosa cells from small and medium-sized follicles: effects of follicle-stimulating hormone, forskolin, and adenosine 3'5'-cyclic. Endocrinology. 124:1204-1209. International Embryo Transfer Society. 1990. Manual of the International Embryo Transfer Society. Second edition. D. A. Stringfellow and S. M Seidel, eds. Champaign, IL. p. 79. Ireland, J. J. 1987. Control of follicular growth and development. J. Reprod. Fertil. 34 (Suppl.):39-54.

PAGE 221

212 Ireland, J. J. and J. F. Roche. 1983a. Development of nonovulatory antral follicles in heifers: changes in steroids in follicular fluid and receptors for gonadotropins. Endocnnology. 112:150-156. Ireland, J. J. and J. F. Roche. 1983b. Growth and differentiation of large antral follicles after spontaneous luteolysis In heifers: changes in concentration of hormones in follicular fluid and specific binding of gonadotropins to follicles. J. Anim. Sci. 57:157-167. Ireland, J. J. and J. F. Roche. 1987. Hypothesis regarding development of dominant follicles during a bovine estrous cycle. In: Follicular growth and ovulation rate in farm animals. J. F. Roche and D. O'Callaghan (Eds.). Martinus Hijhoff Publishers, The Hague, pp. 1-18. Ireland, J. L. H. and J. J. Ireland. 1994. Changes in expression of inhibin/activin a, 3a. and pB subunit messenger ribonucleic acids following increases in size and during different stages of differentiation or atresia of non-ovulatory follicles in cows. Biol. Reprod. 50:492-501. . . Ireland, J. L. H., T. E. M. Good, P. G. Knight and J. J. Ireland. 1994. Alterations in amounts of different forms of inhibin during follicular atresia. Biol. Reprod. 50:1265-1276. Jolly, P. D., D. J. Tisdall, D. A. Heath, S. Lun and K. P. McNatty. 1994. Apoptosis in bovine granulosa cells in relation to steroid synthesis, cAMP response to FSH and LH and follicular atresia. Biol. Reprod. 51:934-944. Kessner, J. S., E. M. Convey and C. R. Anderson. 1981. Evidence that estradiol induces the preovulatory LH surge in cattle by increasing pituitary sensitivity to LHRH and then increasing LHRH release. Endocrinology. 108:1386-1391. Kinder, J. E., F. N. Kojima, E. G. M. Bergfeld, M. E. Wehrman, and K. E. Fike. 1996. Progestin and estrogen regulation of pulsatile LH release and development of persistent ovarian follicles in cattle. J. Anim. Sci. 74:1424-1440. Knecht, M., A. Amsterdam and K. J. Catt. 1981. The regulatory role of cyclic AMP in hormone induced granulosa cell differentiation. J. Biol. Chem. 256:1062810623. Knickerbocker, J. J., W. W. Thatcher, F. W. Bazer, M. Drost, D. H. Barron, K. B. Fincher and R. M. Roberts. 1986. Proteins secreted by day 16 to 18 conceptuses extend corpus luteum functioning in cows. J. Reprod. Fertil. 77:381-391.

PAGE 222

213 Knickerbocker, J. J., W. W. Thatcher, D. B. Foster, D. Wolfenson, F. F. Bartol, and D. Caton. 1982. Uterine production of PGFja in response to estradiol-17& in cyclic cattle. J. Anim. Sci. 55 (Suppl. 1): 364 (Abstr.). Knight, P. G., A. J. Beard, J. H. M. Wrathall and R. J. Castillo. 1989. Evidence that the bovine ovary secretes large amounts of monomeric inhibin a subunit and its isolation from bovine follicular fluid. J. Mol. Endocrinol. 2:189-200. Ko, J. C. H., J. P. Kastelic, M. R. Del Campo, and O. J. Ginther. 1991. Effects of a dominant follicle on ovarian follicular dynamics during the oestrous cycle in heifers. J. Reprod. Fertil. 91:511-519. Kuran, M., J. S. Morley Hutchinson and P. J. Broadbent. 1996. The response of bovine granulosa cells to different gonadotropins in culture. Anim. Reprod. Sci. 45:1-12. Kuran, M., P. J. Broadbent and J. S. M. Hutchinson. 1995. Follicle stimulating honnone stimulated differentiation and progesterone production of bovine granulosa cells in culture. Anim. Reprod. Sci. 39:237-249. Laemmelli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage V. Nature. 227:680-685. Langhout, D. J., L. J. Spicer and R. D. Geisert. 1991. Development of a culture system for bovine granulosa cells: effects of growth hormone, estradiol, and gonadotropins on cell proliferation, steroidogenesis, and protein synthesis. J. Anim. Sci. 69:3321-3334. LaPolt, P. S., G. N. Piquette, D. Soto, C. Sincich and A. J. Hsueh. 1990. Regulation of inhibin subunit messenger hbonucleic acid levels by gonadotropins, growth factors, and gonadotropin-releasing hormone in cultured rat granulosa cells. Endocrinology. 127:823-831. Leeuwenberg, B. R., P. R. Hurst K. P McNatty. 1995. Expression of IGF-I mRNA in the ovine ovary. J. Mol Endocrinol. 15:251-258. Legault, S., J. L. Bailey, M. A. Fortier, P. Rouillier and L. A. Gullbault. 1997. Intracellular regulation of estradiol production by cultured bovine granulosa cells. Biol. Reprod. 56 (Suppl. 1):109. (Abstr.). Lerner, S. P., W. V. Thayne, R. D. Baker, T. Hensche, S. Meredith, E. K. Inskeep, R. A. Dailey, P. E. Lewis and R. L. Butcher. 1986. Age, dose of FSH and other factors affecting superovulation in Holstein cows. J. Anim. Sci. 63:176-183.

PAGE 223

214 Li, R. D. M. Phillips and J. P. Mather. 1995. Activin promotes ovarian follicle developnaent /n v/fro. Endocrinology. 136:849-855. Ling N S. Y. Ying, H, Ueno, F. Esch, L. Denorary and R. Guillemin. 1985. Isolation and partial characterization of a Mw 32,000 protein with inhibin activity from porcine follicular fluid. Proc. Natl. Acad. Sci. USA. 82:7217-7221. Littell, R. C, R. J. Freund and P. C. Spector. 1991. SAS® System for Linear Models. Third edition. Gary, NC: SAS Institute Inc. p. 329. Lobb, D. K and J. Dorrington. 1992. Intraovarian regulation of follicular development. Anim. Reprod. Sci. 28:343-354. Luck, M. R. 1994. The gonadal extracellular matrix. Oxf. Rev. Reprod. Biol. 16:33-86. Lucy, M., J. D. Savio, L. Badinga, R. L. de la Sota and W. W. Thatcher. 1992. Factors that affect ovarian follicular dynamics in cattle. J. Anim. Sci. 70:3615-3626. Lussier, J. G., Matton, P., and Dufour, J. J. 1987. Growth rates of follicles in the ovary of the cow. J. Reprod. Fertil. 81:301-307. Macmillan, K. L., V. K. Taufa, D. R. Barnes, A. M. Day and R. Henry. 1988. Detecting estrus in synchronized heifers using tailpaint and an aerosol raddle. Theriogenology. 30:1099-1114. Macmillan, K. L. and W. W. Thatcher. 1991 . Effects of an agonist of gonadotropinreleasing hormone on ovarian follicles in cattle. Biol. Reprod. 45:883-889. McNatty, K. P., D. A. Heath, N. Hudson and I. J. Clarke. 1990. Effect of longterm hypophysectomy on ovahan follicle populations and gonadotrophin-induced adenosine cyclic 3',5'-monophosphate output by follicles from Booroola ewes with or without the F gene. J. Reprod. Fertil. 90:515-522. Magoffin, D. A. and G. F. Erickson. 1994. Control systems of theca-interstitial cells. In: Molecular biology of the female reproductive system. J. K. Findlay (Ed.). Academic Press, Inc. San Diego, pp. 39-65. Mapletoft, R. J., G. A. Bo and R. A. Pierson. 1994. Recruitment of follicles for superovulation. Compend. Cont. Educ. 16:127-141. ' • Mariana, J. C, D. Monniaux, M. A. Driancourt and P. Mauleon. 1991. Folliculogenesis. In: Reproduction in domestic animals. P. T. Cupps (Ed.) 4'". Edition. Academic Press, Inc., San Diego, pp. 1 19-.

PAGE 224

215 Marion, G. B., H. T. Gierand J. B. Choudary. 1968. Micromorphology of the bovine ovarian follicular system. J. Aninn. Sci. 27:451-465. Martin, T. L., R. L. Fogwell and J. J. Ireland. 1991. Concentrations of inhibins and steroids in follicular fluid during development of dominant follicles in heifers. Biol. Reprod. 44:693-700. Masague, J. 1987. The TGF-S family of growth and differentiation factor. Cell. 49:437-438. Massague, J. and B. Like. 1985. Cellular receptors for type beta transforming growth factor. J. Biol. Chem. 260:2636-2645. Mather, J. P., T. K. Woodruff and L. A. Krummen. 1992. Paracrine regulation of reproductive function by inhibin and activin. P. S. E. B. M. 201:1-15. Mathews, L. 1994. Activin receptors and cellular signaling by the receptor serine kinase family. Endocnne Rev. 15:310-325. Matton, P., V. Adelakoun, Y. Couture, J. J. Dufour. 1981. Growth and replacement of the bovine ovanan follicles during the estrous cycle. J. Anim. Sci. 52: 813-820. McGowan, M. R., M. Braithwaite, W. Jochle and R. J. Mapletoft. 1985. Superovulation of beef cattle with Pergonal (hMG): a dose-response trial. Theriogenology. 24:173-184. Michel, U., S. Ludemann, H. Jarry and W. Wuttke. 1991. Effects of growth factors and hormones on basal and FSH-stimulated inhibin production by porcine granulosa cells in vitro. Reprod. Fertil. Dev. 3:201-213. Monget, P. and D. Monniaux. 1995. Growth factors and the control of folliculogenesis. J. Reprod. Fert. Suppl. 49:321-333. Monniaux, D. and M. M. De Reviers. 1989. Quantitative autoradiographic study of FSH binding sites in prepubertal ovaries of three strains of rats. J. Reprod. Fertil. 85:151-162. Monniaux, D., D. Chupin and J. Saumande. 1983. Superovulatory responses of cattle. Theriogenology. 19:55-81. Monniaux, D., J. C. Mariana and W. R. Gibson. 1984. Action of PMSG on follicular populations in the heifer. J. Reprod. Fertil. 70:243-253.

PAGE 225

216 Monniaux, D., P. Monget, N. Besnard, C. Huet and C. Pisselet. 1997. Growth factors and antral follicular development in domestic ruminants. Theriogenology. 47:3-12. Monniaux, D. and C. Pisselet. 1992. Control of proliferation and differentiation of ovine granulosa cells by insulin-like growth factor-l and follicle-stimulating homnone in vitro. Biol. Reprod. 46:109-119. Morris, D. G., M. G. McDermott, M. G. Diskin, C. A. Morrison, P. J. Swift and J. M Sreenan. 1993. Effect of immunization against synthetic peptide sequences of bovine inhibin a-subunit on ovulation rate and twin-calving rate in heifers. J. Reprod. Fertil. 97:255-261. Mulheron, G. W. and D. W. Schomberg. 1990. Rat granulosa cells express transforming growth factor-S type 2 messenger ribonucleic acid which is regulatable by follicle-stimulating hormone in vitro. Endochnology. 126:1777-1779. Mulheron, G. W. and D. W. Schomberg. 1993. The intraovarian transforming growth factor system. In: The Ovary. E. Y. Adashi and P. C. K. Lueng (Eds). New York, Raven Press, pp.337-361. Murphy, M. G., M. P. Boland and J. F. Roche. 1990. Pattern of follicular growth and resumption of ovarian activity in post-partum beef suckler cows. J. Reprod. Fertil. 90:523-533. Murphy, B. D., R. J. Mapletoft, J. Manns and W. D. Humphrey. 1984. Variability in gonadotropin preparations as a factor in the superovulatory response. Theriogenology. 21:117-125. O'Shea, T., C. M. Andrews, B. M. Bindon, M. A. Hillard, K. Miyamoto and M. J. Sinosich. 1991 . Immunization of merino ewes with a synthetic inhibin peptide or with preparations obtained from bovine and porcine follicular fluids by immunoaffinity chromatography result in different effects on ovulation rate and on plasma gonadotropin concentrations. Reprod. Fertil. Dev. 3:659-670. O'Shea. T., M. A. Hillard, S. T. Anderson, B. M. Bindon, J. K. Findlay, C. G. Tsonis and J. F. Wilkins. 1994. Inhibin immunization for increasing ovulation rate and superovulation. Theriogenology. 41:3-17. Paton, A. C. and W. P. Collins. 1992. Differentiation process of granulosa cells. Oxford Rev. Reprod. Biol. 14:169-223. Pawlyshyn, V., C. E. Lindsell, M. Braithwaite and R. J. Mapletoft. 1986. Superovulation of beef cows with FSH-P: a dose response trial. Thenogenology. 25:179 (Abstr.).

PAGE 226

217 Pierson, R. A. and O. J. Ginther. 1984. Ultrasonography of the bovine ovary. Theriogenology. 21:495-504. Pierson, R. A. and O. J. Ginther. 1988. Follicular population during the estrous cycle in heifers: III. Time of selection of the ovulatory follicle. Anim. Reprod. Sci. 16:81-95. Pieterse, M. C; K. A. Kappen, T. A. M. Kruip and M. A. M. Taverne. 1988. Aspiration of bovine oocytes duhng transvaginal ultrasound scanning of the ovaries. Theriogenology. 30:751-762. Pieterse, M. C; P. L. A. M. Vos, T. A. M. Kruip, A. H. Willemse and M. A. M. Taverne. 1991a. Characteristics of bovine estrous cycle during repeated transvaginal, ultrasound-guided puncturing of follicles for ovum pick-up. Theriogenology. 35:401-413. Pieterse, M. C; P. L. A. M. Vos, T. A. M. Kruip, Y. A. Wurth, T. H. van Beneden, A. H. Willemse and M. A. M. Taverne. 1991b. Transvaginal ultrasound guided follicular aspiration of bovine oocytes. Theriogenology. 35:857-862. Price, C. A., B. A. Morris, T. O'Shea and R. Webb. 1987. Active immunization of cattle against partly purified follicular fluid from sheep. J. Reprod. Fertil. 81:161168. , , , Price, C. A., and R. Webb. 1989. Ovarian response to hCG treatment during the estrous cycle in heifers. J. Reprod. Fertil. 86:303-308. Putney, D. J., M. Drost and W. W. Thatcher. 1988. Embryonic development in superovulated dairy cattle exposed to elevated ambient temperatures between days 1 to 7 post insemination. Theriogenology. 30:195-209. Rahe, C. H., R. E. Owens, H. J. Newton, J. L Fleeger and P. G. Harms. 1980. Pattern of plasma luteinizing hormone in the cyclic cow: dependence upon the period of the cycle. Endocrinology. 107:498-503. Rajakoski, E. 1960. The ovarian follicular system in sexually mature heifers with special reference to seasonal, cyclical and left right variations. Acta Endocrinologica. Suppl. 52:1-68. Rajamahendran, R., and P. C. Sianangama. 1992. Effect of human chorionic gonadotrophin on dominant follicles in cows: formation of accessory corpora lutea, progesterone production and pregnancy rates. J. Reprod. Fertil. 93:577-584.

PAGE 227

218 Rajamahendran, R., R. S. Canseco and C. J. Denbow. 1987. Effect of low dose of FSH given at the beginning of the estrous cycle and subsequent superovulatory response in Holstein cows. Theriogenology. 28:59-65. Rajamahendran, R. and M. D. Calder. 1993. Superovulatory responses in dairy cows following ovulation of the dominant follicle of the first wave. Thenogenology. 40:99-109. Richards, J. (1980). Maturation of ovarian follicles: actions and interactions of pituitary and ovarian hormones on follicular cell differentiation. Physiol. Rev. 60:51 89. Richards, J. S. and L. Hedin. 1988. Molecular aspects of hormone action in ovarian follicular development, ovulation, and luteinization. Ann. Rev. Physiol. 50:441-463. Richards, J. S. and R. L. Robker. 1997. Hormonal control of the cell cycle in ovarian cells: proliferation versus differentiation. Biol. Reprod. 56 (Suppl. 1):28 (Abstr.). Roberts, A. J. and M. K. Skinner. 1990. Estrogen regulation of thecal cell steroidogenesis and differentiation: thecal cell-granulosa cell interactions. Endocrinology. 127:2918-2929. Roberts, A. J. and M. K. Skinner. 1991. Transforming growth factors-a and -B differentially regulate growth and steroidogenesis of bovine thecal cells during antral follicle development. Endocrinology. 129:2041-2048. Roberts, A. J. and S. E. Echternkamp. 1994. In vitro production of estradiol by bovine granulosa cells: evaluation of culture condition, stage of follicular development, and location of cells within follicles. Biol. Reprod. 51:273-282. Robertson, D. M., L. M. Foulds, L. Leversha, F. J. Morgan, M. T. W. Hearn, H. G. Burger, R. E. H. Wettenhall, D. M. de Krester. 1985. Isolation of inhibin from bovine follicular fluid. Biochem. Biophys. Res. Comun. 126:220-226. Robertson, D. M., C. G. Tsonis, R. I. McLachlan, D. J. Handelsman, R. Leask, D. T. Baird, A. S. McNeilly, S. Hayward, D. L. Healy, J. K. Findlay, H. G. Burger and D. M. de Krester. 1988. Comparison of inhibin immunological and in vitro biological activities in human serum. J. Clin. Endocrinol. Metab. 67:438-448. Robertson, D. M., H. G. Burger, J. Sullivan, N. Cahir, N. Groome, E. Poncelet, P. Franchimont, T. Woodruff and J. P. Mather. 1996. Biological and immunological characterization of inhibin forms in human plasma. J. Clin. Endocrinol. Metab. 81:669-676.

PAGE 228

219 Robertson, D. M., L. M. Foulds, M. Prisk and M. P Hedger. 1992. Inhibin/activin S subunit monomer: isolation and characterisation. Endocrinology. 130:1680-1687. Robertson, D. M., J. Sullivan, M. Watson and N. Cahir. 1995. Inhibins forms in human plasma. J. Endocrinol. 144:261-269. Robinson, T. J. and J. N. Shelton. 1991. Reproduction in cattle. In: Reproduction in domestic animals. P. T. Cupps (Ed.). 4". Edition. Academic Press, Inc. San Diego, pp. 445-470. Roche, J. F. (1996). Control and regulation of folliculogenesis a symposium in perspective. Rev. Reprod. 1:19-27. Rose, M. P. and R. E. Gaines-Das. 1996. Commentary: standardization of nonsteroidal regulators of gonadotrophin secretion. J. Reprod. Fertil. 107:151-152. Rouillier, P., P. Matton, M.-A. Sirard and L. A. Guilbault. 1996. Follicle-stimulating hormone-induced estradiol and progesterone production by bovine antral and mural granulosa cells cultured in vitro in a completely defined medium. J. Anim. Sci. 74:3012-3019. Rouillier, P., M. -A. Sirard, P. Matton and L A. Guilbault. 1997. Immunoneutralization of transforming growth factor a present in bovine follicular fluid prevents the suppression of the follicle-stimulating hormone-induced production of estradiol by bovine granulosa cells cultured in vitro. Biol. Reprod. 57:341-346. Rusbridge, S. M., T. A. Bramley and R. Webb. 1992. A comparison of GnRHinduced corpora lutea and spontaneously formed CL in heifers. J. Reprod. Fertil. Abstr. Series. (9):33. (Abstr.). SAS. 1988. In: SAS/STAT™ User's guide (Release 6.03 Ed.). SAS Institute Inc. Gary, NC. Sato, E., T. Ishibashi and A. Iritani. 1982. Purification and action sites of a follicle stimulating hormone inhibitor from bovine follicular fluid. J. Anim. Sci. 55:873-877. Saumande, J. 1991. Culture of bovine granulosa cells in a chemically defined serum-free medium: the effect of insulin and fibronectin on the response to FSH. J. Steroid Biochem. Molec. Biol. 38:189-196. Savio, J. D., M. P. Boland, N. Hynes and J. F. Roche. 1990a. Resumption of follicular activity in the early post-partum dairy cov\/s. J. Reprod. Fertil. 88:569-579.

PAGE 229

220 Savio J D., M. P. Boland and J. F. Roche. 1990b. Development of dominant follicles and ovarian activity in the post-partum period of dairy cows. J. Reprod. Fertil. 88:581-591. ' ' Savio. J. D., M. P. Boland. N. Hynes. and J. F. Roche. 1990c. Resumption of follicular activity in the early post-partum dairy cows. J. Reprod. Fertil. 88:569-579. Savio J DH.Bongers.M.Drost, M.C.Lucy and W.W.Thatcher. 1991. Follicular dynamics and superovulatory response in Holstein cows treated with FSH-P in different endocrine states. Theriogenology. 35:915-929. Savio, J. D., L. Keenan, M. P. Boland and J. F. Roche. 1988. Pattern of growth of dominant follicles during the estrous cycle of heifers. J. Reprod. Fertil. 83:663-671. Savio. J. D.. W. W. Thatcher, L. Badinga, R. L. de la Sota and D. Wolfenson. 1993a. Regulation of dominant follicle turnover during the oestrous cycle in cows. J. Reprod. Fertil. 97:197-203. Savio, J. D., W. W. Thatcher, G. R. Morris, K. Entwistle, M. Drost and M. R. Mattiacci. 1 993b. Effects of induction of low plasma progesterone concentrations with a progesterone-releasing intravaginal device on follicular turnover and fertility in cattle. J. Reprod. Fertil. 98:77-84. Savion. N. and D. Gospodarowicz. 1980. Patterns of cellular peptide synthesis by cultured bovine granulosa cells. Endocrinology. 107:1798-1807. Scanlon, A. R., S. J. Sunderland, T. L. Martin, D. Goulding, D. O'Callaghan, D. H. Williams, D. R. Headon, M. P. Boland, J. J. Ireland and J. F. Roche. 1993. Active immunization of heifers against a synthetic fragment of bovine inhibin. J. Reprod. Fertil. 97:213-222. Schanbacher, B. D.. S. R. Schemm and S. M. Rhind. 1991. Gonadotrophin concentrations and ovulation rates in Suffolk ewes actively or passively immunised against inhibin alpha. J. Reprod. Fertil. 93:133-139. Schipper, I., B. C. J. M. Fausser, E. B. O. van Gaver, P. W. Zarutskie and K. D. Dahl. 1993. Development of a human granulosa cell culture model with follicle stimulating hormone responsiveness. Hum. Reprod. 8:1380-1386. Schmitt. E. J-P., C. M. Barros, P. A. Fields, M. J. Fields, T. Diaz, J. M. Kluge, and W. W. Thatcher. 1996a. A cellular and endocrine characterization of the original and induced corpus luteum after administration of a gonadotropin-releasing hormone agonist or human chorionic gonadotropin on day five of the estrous cycle. J. Anim. Sci. 74:1915-1929.

PAGE 230

221 Schmitt, E. J-P., T. Diaz, C M. Barros, R. L. de la Sota, E. W. Fredricksson, C. R. Staples, R, Thorner, and W. W. Thatcher. 1996b. Differential response of the luteal phase and fertility in cattle following ovulation of the first wave dominant follicle with human chorionic gonadotropin versus a gonadotropin-releasing hormone agonist. J. Anim. Sci. 74:1074-1083. Schmitt, E. J. -P. 1995. Pharmacological manipulation of ovarian follicular function in cattle. PhD Dissertation. University of Florida. Schneider, C, R. A. Newman, D. R. Sutherland, U. Asser and M. F. Greaves. 1982. A one-step purification membrane proteins using a high efficiency immunomatrix. J. Biol. Chem. 257:10766-10769 Schneyer, A. L., P M Sluss, R W. Whitcomb, K. A. Martin, R. Sprengel and W. F. Crowley Jr. 1991 Precursors of a-inhibin modulate follicle-stimulating hormone receptor binding and biological activity. Endocrinology. 129:1987-1999. Seguin, B. E., W. D. Oxender, and J, H. Britt. 1977. Effect of human chorionic gonadotropin and gonadotropin releasing hormone on corpus luteum function and estrous cycle duration in dairy heifers. Am. J. Vet. Res. 38:1153-1156. Shelton, K., M. F. Gayerie De Abreu, M. G. Hunter, T. J. Parkinson and G. E. Lamming. 1 990. Luteal inadequacy during the early luteal phase of subfertile cows. J. Reprod. Fertil. 90:1-10. Shimonaka, M., S. Inouye. S. Shimasaki and N. Ling. 1991. Follistatin binds to both activin and inhibin through the common beta-subunit. Endocrinology. 128:3313-3315. Shukovski, L. and J. K. Findlay. 1990. Activin-A inhibits oxytocin and progesterone production by preovulatory bovine granulosa cell in vitro. Endocrinology. 1 26:22222224. Shukovski, L., J. K Findlay and D. M. Robertson. 1991. The effect of folliclestimulating hormone-suppressing protein or follistatin on luteinizing bovine granulosa cells in vitro and its antagonistic effect on the action of activin. Endocrinology. 129:3395. ;. Shukovski, L., Z. W Zhang. U. Michel and J. K. Findlay. 1992. FSH-suppressing protein (FSP) mRNA expression in bovine ovarian tissues. J. Reprod. Fertil. 95:861. Sianangama, P. C , and R. Rajamahendran 1992. Effect of human chorionic gonadotropin administered at specific times following breeding on milk progesterone and pregnancy in cows. Theriogenology 35:85-96.

PAGE 231

222 Sianangama, P. C, and R. Rajamahendran. 1996. Effect of hCG administration on day 7 of the estrous cycle on follicular dynamics and cycle length in cows. Theriogenology 45:583-592. Sirard, M.-A., S. Bilodeau, M. A. Fortier, and A. Belanger. 1991. Bovine granulosa cells steroidogenesis in vitro, II: Response to FSH, forskolin, and IBMX. In: G. Gibori (Ed.) Signaling Mechanisms and Gene Expression in the Ovary, p 435. Springer-Verlag, New York. Sirois, J., and J. Fortune. 1988. Ovarian follicular dynamics during the estrous cycle monitored by real time ultrasonography. Biol. Reprod. 39:308-317. Sirois, J. and J. Fortune. 1990. Lengthening the bovine estrous cycle with low levels of exogenous progesterone: a model for studying ovarian follicular dominance. Endocrinology. 127:916-925. Skinner, M. K. 1992. Growth factors in gonadal development. J. Anim. Sci. 70(Suppl.2):30-41. Skinner, M. K. and R. J. Coffey. 1988. Regulation of ovarian cell grovAh through the local production of transforming growrth factor-a by theca cells. Endocrinology. 123:2632-2639. Skinner, M. K. and K. G. Osteen. 1988. Developmental and hormonal regulation of bovine granulosa cell function in the preovulatory follicle. Endocrinology. 123:1668-1675. Skinner, M. K., J. Keski-Oja, K. G. Osteen and H. L. Moses. 1987a. Ovahan thecal cells produce transforming growth factor-R which can regulate granulosa cells growth. Endocrinology. 121:786-792. Skinner, M. K., D. Lobb and J. H. Dorrington. 1987b. Ovarian thecal interstitial cells produce an epidermal growth factor-like substance. Endocrinology. 121:18921899. Smith, P. K., R. I. Krohn, G. T. Mallia, F. H. Gartner, M. D. Provenzano, E. K. Fujimoto, N. M. Goeke, B. J. Oson and D. C. Klent. 1985. Measurement of protein using bicinchinonic acid. Anal. Biochem. 150:76-83. Spicer, L. J. and E. Alpizar. 1994. Effects of cytokines an FSH-induced estradiol production by bovine granulosa cells in vitro: dependence on size of follicle. Domestic Anim. Endocrinol. 1 1 :25-34.

PAGE 232

223 Spicer, L. J. and S. E. Echternkamp. 1995. The ovarian insulin and insulin-like growth factor system with an emphasis on domestic animals. Dom. Anim. Endocrinol. 12:223-245. Spicer, L. J., A. Alpizarand S. E. Echternkamp. 1993. Effects of insulin, insulin-like growth factor I, and gonadotropins on bovine granulosa cell proliferation, progesterone production, estradiol production, and(or) insulin-like growth factor-l production in vitro. J. Anim. Sci. 71:1232-1241. Spicer, L. J. and R. E. Stewart. 1996. Interaction among bovine somatotropin, insulin, and gonadotropins on steroid production by bovine granulosa and thecal cells. J. Dairy Sci. 79:813-821. Stocco, D. M. 1997. A StAR search: implications in controlling steroidogenesis. Biol. Reprod. 56:328-336. Stock, A. E., J. E. Ellington and J. E. Fortune. 1993. Follicular dominance in cattle: superovulatory responses in the presence vs absence of a dominant follicle. Biol. Reprod. 48 (Suppl. 48):189 (Abstr.) Stock, A. E. and J. E. Fortune. 1993. Ovarian follicular dominance in cattle: Relationship between prolonged growth of the ovulatory follicle and endocrine parameters. Endocrinology. 132:1108-1114. Stubbings, R. B. and J. S. Walton. 1995. Effect of ultrasonically-guided follicle aspiration on estrous cycle and follicular dynamics in Holstein cows. Theriogenology. 43:705-712. Sunderland, S. J., M. A. Crowe, M. B. Boland, J. F. Roche and J. J. Ireland. 1994. Selection, dominance and atresia of follicles during the oestrous cycle of heifers. J. Reprod. Fertil. 101:547-555. Sunderland, S. J., P. G. Knight, M. P. Boland, J. F. Roche and J. J. Ireland. 1996. Alterations in intrafollicular levels of different molecular mass forms of inhibin during development of follicularand luteal-phase dominant follicles in heifers. 54:453462. Biol. dcfKip. Taya, K., H. Kaneko, T. Takedomi, H. Kishi and G. Watanabe. 1996. Role of inhibin in the regulation of FSH secretion and folliculogenesis in cows. Anim. Reprod. Sci. 42:563-570. Thatcher, W. W., R. L. de la Sota, E. J-P. Schmitt, T. C. Diaz, L. Badinga, F. A. Simmen, C. R. Staples, and M. Drost. 1996. Control and management of ovarian follicles in cattle to optimize fertility. Reprod. Fertil. Dev. 8:203-217.

PAGE 233

224 Thatcher, W. W., K. L. Macmillan, P. J. Hansen, and M. D. Drost. 1994a. Embryonic Losses: Cause and Prevention. In: Factors affecting Calf Crop. M. J. Fields and R. S. Sand. (Eds.) CRC Press. Boca Raton, pp. 135-153. Thatcher, W. W., C. R. Staples, G. Danet-Desnoyers, B. Oldick and E. J.-P. Schmitt. 1994b. Embryo health and mortality in sheep and cattle. J. Anim. Sci. 72:16. Thatcher, M-J. D., V. M. Shille, W. C. Buhi, I. M. Alvarez, P. W. Concannon, D. Thibeault and M. Cotton. 1994c. Canine conceptus appearance and de novo protein synthesis in relation to the time of implantation. Theriogenology. 41 :16791692. Tonetta, S. A. and G. S. diZerega. 1989. Intragonadal regulation of follicular maturation. Endochne Rev. 10:205-229. Torney. A. H., Y. M. Hodgson, R. Forage and D. M. de Krester. 1989. Cellular localization of inhibin mRNA in the bovine ovary by in-situ hybridization. J. Reprod. Fertil. 86:391-399. Tortonese, D. J., P. E. Lewis, H. Papkoff and K. E. Inskeep. 1990. Roles of the dominant follicle and the pattern of oestradiol in induction of preovulatory surges of LH and FSH in prepubertal heifers by pulsatile low doses of LH. J. Reprod. Fertil. 90:127-135. Touati, K., J. F. Beckers and F. Ectors. 1991. Hormonal control of folliculogenesis in the bovine: better superovulatory responses after pure FSH administration preceding classical treatment. Thehogenology. 35:285 (Abstr.). Turner, I. M., P. T. K. Saunders, S. Shimasaki and S. G. Hillier. 1989. Regulation of inhibin subunit gene expression by FSH and estradiol in cultured rat granulosa cells. Endocrinology. 122:741-748. Turzillo, A. M. and J. E. Fortune. 1990. Suppression of the secondary FSH surge with bovine follicular fluid is associated with delayed ovarian follicular development in heifers. J. Reprod. Fertil. 89:643-653. Ulloa-Aguirre, A., C. Miller, L. Hyland S. Chapel. 1984. Production of all folliclestimulating hormone isohormones from a purified preparation by neuraminidase digestion. Biol. Reprod. 30:382-387. Ulloa-Aguirre, A., A. Rees Midgley Jr., I. Z. Beitins and V. Padmanabhan. 1995. Follicle-stimulating Isohormones: characterization and physiological relevance. Endoc. Rev. 16:765-787.

PAGE 234

225 Urban, R. J. and J. D. Veldhuis. 1992. Endocrine control of steroidogenesis in granulosa cells. Oxford Rev. Reprod. Biol. 19:225-262. Van Cleeff, J., K. L. Macmillan, M. Drost, M. C. Lucy and W. W. Thatcher. 1996. Effects of administering progesterone at selected intervals after insemination of synchronized heifers on pregnancy rates and resynchronization of returns to service. Theriogenology. 46:1117-1130. Van den Hurk, R., M. M. Bevers and J. F. Beckers. 1997. In-vivo and in-vitro development of preantral follicles. Theriogenology. 47:73-82. Vernon, R. K. and L. J. Spicer. 1994. Effects of basic fibroblast growth factor and heparin on follicle-stimulating hormoneinduced steroidogenesis by bovine granulosa cells. J. Anim. Sci. 72:2696-2702. Walters, D. L. and E. Schallenberger. 1984. Pulsatile secretion of gonadotrophins, ovarian steroids and ovarian oxytocin during the periovulatory phase of the estrous cycle in the cow. J. Reprod. Fertil. 71:503-512. Walton, J. S., G. W. Halbert, N. A. Robinson, and K. E. Leslie. 1990. Effects of progesterone and human chorionic gonadotrophin administration five days postinsemination on plasma and milk progesterone and pregnancy rates of nomrial and repeat breeder dairy cows. Can. J. Vet. Res. 54:305-308. Wandji, S. A., G. Pelletier and M. A. Sirard. 1992. Ontogeny and cellular localization of ^^^l-labeled insulin-like growth factor-l, '^^l-labeled follicle-stimulating hormone, and ^"l-labeled human chorionic gonadotropin binding sites in ovaries from bovine fetuses and neonatal calves. Biol. Reprod. 47:814-822. Ware, C. B., D. L. Northey and N. L. First. 1987. Effect of administration of FSH at the beginning of the cycle on the subsequent response to superovulation treatment in heifers. Theriogenology. 27:292 (Abstr.). Webb, R., J. G. Gong and T. A. Bramley. 1994. Role of growth hormone and intrafollicular peptides in follicle development in cattle. Theriogenology. 41 :31-44. Webb, R., J. G. Gong, A. S. Law and S. M. Rusbridge, 1992. Control of ovarian function in cattle. J. Reprod. Fertil. Suppl. 45:141-156. Wehrman, M. E., K. E. Fike, F. N. Kojima, E. G. Bergfeld, A. S. Cupp, V. Mariscal, T. Sanchez and J. E. Kinder. 1996. Development of persistent ovarian follicles during synchronization of estrus influences the superovulation response to FSH treatment in cattle. Theriogenology. 45:593-610.

PAGE 235

226 Wiesen J F and A. R. Midgley, Jr. 1993. Changes in expression of connexin 43 gap junction messenger ribonucleic acid and protein during ovanan follicular growth. Endocrinology. 133:741-746. Wilcox C J W W. Thatcher, and F. G. Martin. 1990. Statistical analysis of repeated measurements in physiology experiments, in: Proc. Final Research Coordination Meeting of the FAO/IAEA/ARCAL III Regional Network for Improving the Reproductive Management of Meatand Milk-producing Livestock in Latin Amenca with the aid of Radioimmunoassay. Intemational Atomic Energy Agency, Vienna, Austria, p. 141-155. Wolfsdorf, K. E., T. Diaz, E. J-P. Schmitt, M. J. Thatcher, M. Drost and W. W Thatcher. 1997. The dominant follicle exerts an interovanan inhibition on FSH induced follicular development. Theriogenology. 48:435-447. Woodruff, T. K., R. J. Lyon, S. E. Hansen, G. C. Rice and J. P. Mather. 1990. Inhibin and activin locally regulate rat ovarian folliculogenesis. Endocnnology. 127:3196-3205. Woodruff, T. K., H. Meunier. P. B. Jones, A. J. W. Hsueh and K. E. Mayo. 1987. Rat inhibin: molecular cloning of aand p-subunit complementary deoxynbonucleic acids and expression in the ovary. Mol. Endocrinol. 1:561-569. Wrathall, J. H. M. and P. G. Knight. 1993. Production of immunoactive inhibin by bovine granulosa cells in serum-free culture: effects of exogenous steroids and FSH. Dom. Anim. Endocrinol. 10:289-304. Wrathall, J. H. M. and P. G. Knight. 1995. Effects of inhibin-related peptides on oestradiol on androstenedione and progesterone secretion by bovine theca cells in vitro. J. Endocrinol. 145:491-500. Wray, W., T. Boulikas, V. P. Wray and R. Hancock. 1981. Silver staining of proteins in polyacrylamide gels. Annal. Bioch. 118:197-203. Xiao, S. and J. K. Findlay. 1990. Modulation of differentiation of rat granulosa cells in vitro by interferon gamma. J. Endocrinology. 133:131-139. Xiao, s. and J. K. Findlay. 1 991 . Interactions between activin and FSH suppressing protein and their mechanisms of action on cultured rat granulosa cells. Mol. Cell Endocrinol. 79:99-107. Xiao, S., J. K. Findlay and D. M. Robertson. 1990. The effect of bovine activin and follicle-stimulating hormone (FSH) suppressing protein/follistatin on FSH-induced differentiation of rat granulosa cells in vitro. Mol. Cell. Endocrinol. 69:1-8.

PAGE 236

227 Xu, Z. Z., H. A. Garverick, G. W. Smith, M. F. Smith, S. A. Hamilton and R. S. Youngquist. 1995a. Expression of follicle-stimulating hormone and luteinizing hormone receptor messenger ribonucleic acids in bovine follicles during the first follicular wave. Biol. Reprod. 53:951-957. Xu, Z. Z., H. A. Garverick, G. W. Smith, M. F. Smith, S. A. Hamilton and R. S. Youngquist. 1995b. Expression of messenger ribonucleic acid encoding cytochrome P450 side-chain cleavage, cytochrome P450 17a-hydroxylase, and cytochrome P450 aromatase in bovine follicles during the first follicular wave. Endocnnology. 136:981-989. Yuan, W., B. Bao, H. A. Garverick, R. S. Youngquist and M. C. Lucy. 1998. Follicular dominance in cattle is associated with divergent patterns of ovahan gene expression for insulin-like growth factor (IGF)-I, IGF-II, and IGF binding protein-2 in dominant and subordinate follicles. Dom. Anim. Endocrinol. 15:55-63. Zeleznik, A. J. 1993. Dynamics of primate follicular growth. A physiological perspective. In: The Ovary. E. Y. Adashi, P. C. K. Leung, p. 41. Raven Press. New York.

PAGE 237

BIOGRAPHICAL SKETCH Thais del Valle Diaz Zambrano was born September 15, 1957, to Maria Antonia and Francisco Diaz in Lagunillas, Zulia State, Venezuela. She is one of six children. She was accepted as a student at the Veterinary Sciences College of Universidad Central de Venezuela in 1975 where she received the degree of Medico Veterinario (Venezuelan equivalent of D.V.M.) in November 1980. During the same year she started her Masters under the supervision of Dr. Magaly R. de Manzo. Three years later, November 1983, she received the degree of Magister Scientiarum from Universidad Central de Venezuela. After graduation, she started to work as Assistant Researcher at the Veterinary College of Universidad Central de Venezuela. In 1986 she was accepted at the Veterinary College as Instructor Professor in the department of Animal Production. In 1990 she became Assistant Professor, which is her current position. In August 1992, she began her doctoral program under the supervision of Dr. William W. Thatcher. After completing the degree of Doctor of Philosophy, she will return to Venezuela where she will resume her position at the Veterinary College of Universidad Central de Venezuela. 228

PAGE 238

1 certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and/is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy Thatcher Graduate Research Professor of Animal Science I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Maarten Drost Professor of Veterinary Medicine I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Frank A. Simmen Professor of Dairy and Poultry Sciences I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree ofpoctorpf Philosophy. Peter J. Mansen Professoc/of Dairy and Poultry Sciences I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor ofphilosophy. B^^negi Associate Professor of Anatomy and Cell Biology

PAGE 239

This thesis was submitted to the Graduate Faculty of the College of Agriculture and to the Graduate School and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. May, 1998


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EQ08LPX7W_Z6BLXJ INGEST_TIME 2015-04-13T20:02:43Z PACKAGE AA00029979_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES