INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.

The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.

In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book.

Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.

A Bell & Howell Information Company 300 North Zeeb Road. Ann Arbor. Ml 48106-1346 USA 313/761-4700 800/521-0600 Order Number 9517090

Partial characterization and purification of steroidogenic factors in thymic epithelial cell culture-conditioned medium

Uzumcu, Mehmet, Ph.D.

The Ohio State University, 1994

UMI 300 N. ZeebRd. Ann Arbor, Ml 48106 PARTIAL CHARACTERIZATION AND PURIFICATION OF STEROIDOGENIC FACTORS IN THYMIC EPITHELIAL CELL CULTURE- CONDITIONED MEDIUM

Dissertation

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of the Ohio State University

By

Mehmet Uzumcii, D.V.M., M.S.

The Ohio State University

1994

Dissertation Committee:

Young C. Lin, DVM, Ph.D. Approved by

Robert L. Hamlin, DVM, Ph.D.

Joseph S. Ottobre, Ph.D. Advisor Yasuko Rikihisa, Ph.D. Dept. Vet. Physio/Pharm. To Meryem ACKNOWLEDGEMENTS

In the name o f Allah, the Most Merciful, the Most Beneficent.

Limitless praise belongs to Allah by whose mercy this study has been completed.

Then, I express my sincere appreciation to Dr. Young C. Lin, for his support, guidance and insight throughout this study. My appreciation also goes to the other members of my advisory committee Drs. Joseph S. Ottobre, Yasuko Rikihisa and Robert L. Hamlin for their invaluable help. I also thank Dr. David Brigstock for the help that he provided during my study.

All kinds of assistance that were provided to me by the staff, especially Ms. Margie

Maxwell and Mrs. Arlene Myers, and my colleagues, Drs. Shigeo Akira, Yan Gu, Hiroshi

Ohmura, George Chang, Serdar Coskun, Yun Fu Hu, Sam Kulp, Falah Shidaifat, William

Chang, Halit Canatan, Moustafa El-Banna, and Muhammad Mushtaq is gratefully appreciated.

I appreciate the financial support that was provided for the early part of this study by

The Scientific and Technical Research Council of Turkey (TUBlTAK).

I would also like to express my gratitude to my parents for everything that they provided for me. Finally and above all, I offer sincere thanks to my wife, Selma, for her continuous support, understanding and encouragement during my study. I thank my children, VITA

March 31, 1964...... Bom-Karapinar, Konya, Turkey.

1986 ...... Doctor of Veterinary Medicine (D.V.M.), University of Ankara, College of Veterinary Medicine, Turkey.

1986-198 8 ...... Doctoral Student, Department of Reproduction and Artificial Insemination, University of Ankara, College of Veterinary Medicine.

1987- 1988 ...... Veterinarian, Ministry of Agriculture, Ankara, Turkey.

199 0...... M.S. in Veterinary Physiology and Pharmacology, The Ohio State University.

1991 -present Graduate Research Associate, The Ohio State University.

PUBLICATIONS:

Uziimcii M and Lin YC (1990) Factor(s) from rat thymic cell culture conditioned medium (TCM) stimulate(s) progesterone secretion (P) in cultured rat granulosa cells. Biol Reprod 42: Suppl. # 1, pl63.

Uztimcii M. and Lin YC, (1991) Evidence for thymic factors (TF) in the regulation of steroidogenesis in cultured rat granulosa cells (RGC). Program and Abstract of the 73rd Annual Meeting of Endocrine Society, p368.

Moh PP, Uziimcii M. Chang GCJ, Hu YF, Brewer ML, and Lin YC, (1991) Effects of estrogen on covalent binding of gossypol to microsomal proteins in rats. Biol Reprod 45: Suppl # 1, pl31. Uziimcii M. Akira S, and Lin YC, (1992) Stimulatory effect of thymic factor(s) on steroidogenesis in cultured rat granulosa cells. Life Sciences 51:1217-1228.

Uziimcii M. and Lin YC, (1992) Gossypol (GP) inhibits thymic factors (TF)-stimulated steroidogenesis in cultured rat granulosa cells (GC). Biol Reprod 46: Suppl. #1, p i68.

Ohmura H, Chang WY, Uziimcii M. Coskun S, Akira S, Araki T, and Lin YC, (1993) Transforming growth factor B1 stimulates progesterone production in cultured granulosa cells from gonadotropin-primed adult rats. In: Leung PCK, Hsueh AJW, Friesen HG (eds), Molecular Basis of Reproductive Endocrinology, Springer-Verlag, New York, 1993; pp 215- 221.

Uziimcii M. Ohmura H, Chang CJG, Araki T, and Lin YC, (1993) Factor(s) in thymic cell culture conditioned medium (TCM) stimulate(s) steroidogenesis in cultured porcine granulosa cells. FASEB J 7: A20.

Fehim EM, Lin YC, Golder JM, Bruggemeier RW, and Uziimcii M. (1993) Effect of an aromatase inhibitor on steroid hormone secretion in cultured rat granulosa cells (RGC). FASEB J 7: A617.

Ghosh PK, Uziimcii M. York JP, and Lin YC, (1993) Effects of galanin on lactic acid and estradiol production by Sertoli cells isolated from immature rats. Program and Abstracts of the 75th Annual Meeting of Endocrine Society, pl02.

Akira S, Ohmura S, Uziimcii M. Araki T, Lin YC, (1994) Gossypol inhibits aromatase activity in cultured porcine granulosa cells. Theriogenology 41:1489-1497.

Uziimcii M. and Lin YC, (1994) Characterization of stimulatory action of thymic factor(s) on basal and gonadotropin-induced steroidogenesis in cultured rat granulosa cells. Mol Cel Endocrinol 105:209-216.

Kulp S, Uziimcii M. Mushtaq M, Rikihisa Y, and Lin YC, (1994) Inhibition of rat Sertoli cell function in vitro by a larval tapeworm excretory/secretory product. FASEB J 8: A576.

Uziimcii M. Brigstock DR and Lin YC (1994) Rat thymic epithelial cell culture conditioned medium (TCM) contains two distinct steroidogenic factors. Biol Reprod 50: Suppl # 1, p74.

Coskun S, Uziimcii M. Lin YC, Friedman Cl and Alak B (1994) Porcine cumulus cell steroidogenesis is regulated by oocytes. Biol Reprod 50: Suppl # 1, pl44.

v FIELD OF STUDY

Major Field: Veterinary Physiology and Pharmacology Studies in Reproductive Endocrinology and Immunology. TABLE OF CONTENTS

DEDICATION...... ii

ACKNOWLEDGEMENTS...... iii

VITA...... iv

TABLE CONTENTS...... vii

LIST OF FIGURES...... ix

LIST OF PLATES...... xii

LIST OF TABLES...... xiii

INTRODUCTION...... 1

CHAPTER

I. THYMIC EPITHELIAL CELL CULTURE CONDITIONED-MEDIA (TCM) STIMULATES STEROIDOGENESIS IN CULTURED RAT GRANULOSA CELLS...... 21

Introduction ...... 21 Materials and methods ...... 23 Results...... 30 Discussion ...... 33 Summary...... 37 Figures ...... 39 Plates...... 64

vii CHAPTER

H. CHARACTERIZATION OF THE STIMULATORY ACTIONS OF THYMIC FACTORS ON BASAL AND GONADOTROPIN-INDUCED STEROIDOGENESIS IN CULTURED RAT GRANULOSA CELLS...... 77

Introduction ...... 77 Materials and methods ...... 79 Results...... 85 Discussion ...... 88 Summary...... 93 Figures ...... 95

CHAPTER

HI TWO DISTINCT STEROID-MODULATING FACTORS WERE IDENTIFIED IN RAT THYMIC EPITHELIAL CELLS CULTURE-CONDITIONED MEDIA (TCM)...... 112

Introduction ...... 112 Materials and methods ...... 114 Results...... 121 Discussion ...... 124 Summary...... 130 Figures ...... 132

CONCLUDING REMARKS...... 152

LIST OF REFERENCES...... 157

viii LIST OF FIGURES

FIGURE PAGE

1. Effect of different doses of TCM, HCM or ME on the basal progesterone secretion in cultured rat granulosa cells (24 hours) ...... 40

2. Lack of any effect of different conditioned media on intracellular cAMP formation in cultured rat granulosa cells ...... 42

3. Effect of 10% TCM on basal progesterone secretion in cultured rat granulosa cells in time course study ...... 44

4. Effect of 10% TCM on basal estradiol secretion in cultured rat granulosa cells in time course study ...... 47

5. Effect of different doses of TCM, HCM or ME on basal progesterone secretion in cultured rat granulosa cells ...... 50

6. Effect of different doses of TCM, HCM or ME on basal estradiol secretion in cultured rat granulosa cells ...... 52

7. Effect of various physicochemical treatments on progesterone- stimulating activity of TCM ...... 54

8. Effects of various doses of thymosin a 1 and thymulin and combination of both peptides on basal progesterone secretion in cultured rat granulosa cells for 24 hours ...... 56

9. Effects of various doses of thymosin a 1 and thymulin and combination of both peptides on basal estradiol secretion in cultured rat granulosa cells for 24 hours ...... 58

ix 10. Effects of various doses of thymosin a 1 and thymulin and combination of both peptides on basal progesterone secretion in cultured rat granulosa cells for 48 hours ...... 60

11. Effects of various doses of thymosin a 1 and thymulin and combination of both peptides on basal estradiol secretion in cultured rat granulosa cells for 48 hours ...... 62

12. Effect of various doses of FSH (0-1000 ng/ml) on progesterone and estradiol secretions in cultured rat granulosa cells ...... 96

13. Effects of different conditioned media on basal and FSH-induced progesterone secretion in cultured rat granulosa cells ...... 98

14. Effects of different conditioned media on basal and FSH-induced 20a- hydroxy-progesterone secretion in cultured rat granulosa cells...... 100

15. Effects of different conditioned media on basal and FSH-induced estradiol secretion in cultured rat granulosa cells ...... 102

16. Effects of different conditioned media on basal and FSH-induced and aromatase enzyme activity, as measured with [3H]-H20 release method, in cultured rat granulosa cells ...... 104

17. Effects of different conditioned media on total cell protein content in cultured rat granulosa cells in the absence and presence of 100 ng/ml FSH ...... 106

18. Effects of different conditioned media on DNA synthesis in cultured rat granulosa cells, in the absence of FSH ...... 108

19. Effects of different conditioned media on DNA synthesis in cultured rat granulosa cells, in the presence of 100 ng/ml FSH 110

20. Effect of TCM gel filtration fractions on progesterone and estradiol secretion ...... 133

21. Effect of TCM gel filtration fractions on 20-a hydroxy- progesterone and aromatase enzyme activity as measured by [3H]- HjO release method ...... 135

22. Morphology of the granulosa cell following treatments ...... 137

x 23. Synergistic action of 1 and 22 kDa factors on progesterone secretion ...... 142

24. Antagonistic effect of 22 kDa on 1 kDa-induced estradiol ...... 144

25. Effect of various doses of FSH (0-2000 ng/ml) on progesterone and estradiol secretions in cultured rat granulosa cells ...... 146

26. Dose response study for 1 and 22 kDa factors for basal and FSH-induced progesterone secretions ...... 148

27. Dose response study for 1 and 22 kDa factors for basal and FSH- induced estradiol secretions ...... 150

xi LIST OF PLATES

PLATE PAGE

I. A thymus gland section from 21 days old female rat (H&E stained) ...... 65 n. An electron micrograph of thymus gland from 21 days old female rats...... 67

HI. Phase-contrast micrograph of thymic cell culture ...... 69

IV Phase-contrast micrograph of cardiac cell culture ...... 71

V. Phase-contrast micrograph of rat granulosa cell culture ...... 73

VI. Cultured thymic cells that were stained immunohistochemically for keratin ...... 75 LIST OF TABLES

TABLE PAGE

1. Comparative potency of unfractionated and fractionated TCM on progesterone and estradiol secretion in cultured rat granulosa cells...... 125

xiii INTRODUCTION

The concept that the thymus gland influences the development and maintenance of ovarian function is only recently realized. Although its influence on ovarian function has been gradually accepted, the exact mechanism(s) is not clear. In this section, currently proposed regulatory mechanisms for the influence of thymus on ovaries will be discussed and relevant evidences from the literature will be presented. Before presenting this evidence, brief information about the thymus; the cells that constitute thymus and their secretory products will be given.

THE THYMUS GLAND

The adult thymus is a bilobate structure that lies above the heart. Each lobe is surrounded by a connective tissue capsule that at intervals also pushes deep into the tissue to the cortical medullary junction. At the histological level, three main thymic areas are defined: the subcapsular zone, the cortex and the medulla.

The thymus develops from the third and fourth pharyngeal pouches in mammals.

Three elements are critical to the normal development of the thymus: ectoderm of the branchial clefts, endoderm of the pharyngeal pouch and mesenchyme from the pharyngeal arch. If any of the components is missing, the thymus fails to develop.

1 The major role of the fully formed and functional thymus is the generation of the T- cells for immune reactivity in the periphery. The T-cells repertoire is formed before puberty when exposure to new antigens is maximal. Opinions vary as to the importance of the thymus thereafter. Generally, the thymus gland is at its the largest size, relative to the body, after birth. The organ begins to involute after puberty (Goldstein and Mackay, 1969). Fatty infiltration commences under the capsule and spreads towards to the medulla, thereby reducing the cortex and number of thymocytes. The medulla is spared from the fatty acid infiltration and, even in old people it contains many cells immunoreactive to antithymulin antibodies (Kendall et al., 1991). Thus, the thymus still produces thymic factors that may play a role in different organ systems late in life (Kendall and Clarke, 1994).

THE CELLS OF THE THYMUS GLAND

The thymus gland of most animals contains morphologically similar cells, although the cell sizes and relative numbers vary greatly from species to species (Plates I & II).

Thymic epithelial cells

The stroma of the gland is composed of epithelial cells derived from endoderm.

These cells are to be distinguished from mesenchymal or mesenchymal reticular cells by the presence of tonofilaments and desmosomes in the former. At the beginning, function of the stromal cell was considered to be merely supportive, but it has now become very clear that stromal cells are the source of the thymic factors and hormones and are therefore the prime regulator of the intra-thymic environment (Kendall, 1981). Macrophages

Macrophages are found scattered throughout all thymic areas, but differ in their morphology and presumably in function. In general, phagocytosis is regarded as the function of the mesenchymal reticular cell or macrophage. Macrophages are found in the capsule, septa and perivascular space, as well as in the medulla and the cortex. The residence period and the fate of thymic macrophages largely remain unknown. Their number is variable and very high in the glands during the time of involution, and increases with age (Kendall, 1981).

Interdigitating reticulum cells

Another important cell population of the thymus, also associated with T-cell maturation and differentiation, is the interdigitating reticulum cells. They are located in the thymic medulla. In isolated thymus glands in culture, these cells appear to have a close relationship with lymphocytes. The exact identity of this cell type has not been established

(Kendall 1981).

Mvoid cells

Myoid cells have an irregular distribution throughout the thymus; they are most frequently found in clusters in the medulla and are often closely associated with a Hassal's corpuscle (concentric whorls of epithelial cells). These cells are usually oval/round in shape and, by electron microscopy, can be seen to contain skeletal muscle-like striations. They show strong immunostaining with antibodies to actin and myosin. (Ritter and Crispe, 1992). Lymphoid cells

Since the work of Miller (1961), there have been numerous demonstrations that the presence of the thymus is essential for the full development and maintenance of cell- mediated immunity. In cellular immunity, the prime function is accomplished by T-cells.

The thymus gland accepts pre-thymic, T-restricted progenitor cells from adult bone marrow or embryonic liver and yolk sac. These cells, in the microenvironment created by the epithelial reticular cells and interdigitating reticulum cells, proliferate, differentiate and acquire the immunological markers of mature T cells. During this entire process these lymphocytes are named thymocyte or intrathymic T precursor cells. Approximately 5% of the lymphoid cells reside in the subcapsular region. The majority of the thymocytes (80-

85%) are found in the cortex. The remaining 10% of the thymocytes are located in the medulla. After maturation, these cells emigrate from the thymus to lymph nodes and T- dependent areas of the body. At this time these cells are called post-thymic T-precursors

(Ritter and Crispe, 1992).

Apoptotic cells

Ninety to 95% of thymocytes are eliminated during T cell receptor repertoire selection. Cortical thymocytes are also highly sensitive to external stress such as acute infection. Death occurs by apoptosis, a programmed cell death mechanism whereby endonuclease cleaves nuclear DNA into many smaller pieces. Small apoptotic fragments are commonly seen within thymic macrophages (Ritter and Crispe, 1992). THYMIC FACTORS AND HORMONES

The growth of interest in the thymus and thymic hormones has provided evidence that variety of fractions and preparations from thymic tissue, blood, and other tissues may be eligible for designation as thymic hormones.

The most critical criterion for acceptance of any putative endocrine product as a hormone is its ability to replace specific functions of the extirpated or absent gland in experimental conditions. In order to classify an active factor as a thymic hormone, it should exhibit activity in one or more of the following biological models:

(1) improvement of immunological impairment in the neonatally thymectomized animal; adult thymectomized, immunosuppressed animal; and the nude mouse; and

(2) other biological criteria: e.g. enhancement of immunological responses evaluated in various in vivo and in vitro assays reflecting the activities of T-cells and their effects on B- cells and macrophages (Mihich and Fefer, 1983).

Thvmosin fraction 5 and its composite polypeptides

Thymosin fraction 5 is one of the first thymic preparations to be isolated. Thymosin fraction 5 was prepared from calf thymus, as described by Hooper et al. (1975). Analytical isoelectric focusing of thymosin fraction 5 has revealed the presence of a number of components ranging in size from 1-15 kDa in the preparation. A nomenclature based on the isoelectric focusing pattern of thymosin fraction 5 has been described by Goldstein et al.

(1977). The separated polypeptides are divided into 3 regions: The region a consists of polypeptides with isoelectric points below 5.0, the 6 region 5.0-7.0, and the y region above 7.0. The subscript numbers a l, a2, Bl, B2, etc. are used for identification of the polypeptide from that region as they are isolated. The purified thymosin polypeptides are the following:

of fraction 5 has been termed thymosin a l. The yield of thymosin a l from fraction 5 is about 0.6%. Thymosin a, is a polypeptide consisting of 28 amino acid residues with a MW of 3108 and has been chemically synthesized by Wang et al. (1978). This synthetic peptide was tested in some biologic assays and appears to have activity similar to that of the natural form.

To evaluate the species variation of thymosin polypeptide, Low and Goldstein

(1978) prepared thymosin fraction 5 from thymus tissues of different species including human, pig, sheep, cattle, chinchilla, and mouse. Human, porcine and ovine thymosin a 1 appear to have an identical sequence to bovine a l.

Thymosin a5 and a7: Both of these purified peptides were isolated from thymosin fraction 5 by ion-exchange chromatography on Carboxymethyl-Cellulose (CMC), Diethyl

Amino Ethyl-Cellulose (DEAEC) and gel filtration on Sephadex G-75. They are highly acidic; Thymosin a5 has a MW of 3000 and thymosin a l of 2200.

Polypeptide B1: The amino acid sequence of Bl, composed of 74 amino acid residues, has a MW of 8451 and isoelectric point of 6.7. This polypeptide has no known biological activity.

Thymosin B3 and B4: These are partially purified preparations, isolated from fraction

5 by chromatography on DEAE cellulose and gel infiltration on Sephadex G-75. Thymosin

63 has an isoelectric point of 5.2 and a MW of approximately 5500; thymosin B4 has an isoelectric point of 5.1 and a MW o f4982.

Thymulin

Another hormone from thymus gland is thymulin, formerly called FTS (facteur thymique serique). This hormone was initially isolated from porcine serum (Dardenne et al.,

1977) but has also been purified from human serum. Its amino acid sequence was determined to be [Glu-Ala-Lys-Ser-Gln-Gly-Gly-Ser-Asn] (Pleau et al, 1977). It has a MW of 900, and a neutral isoelectric point.

Direct evidence for the presence of thymulin in the thymus has been obtained by several approaches including its isolation from thymic extracts (Dardenne et al., 1981) and the demonstration of its presence in thymic epithelial cells by immunochemistry, using xenoantisera (Monier et al., 1980; Jambon et al., 1981) and specific monoclonal antibodies

(Savino et al., 1982; Auger et al., 1982). There is no apparent species specificity, as the amino acid analysis of thymulin from calf and human was identical to that of porcine thymulin (Lacovara and Utermohlen, 1983).

Other Thvmic Peptides

Some of other thymic hormones by name are thymopoietin I and n, thymostimulin, porcine thymic hormone, thymic humoral factor, and hemostatic thymic hormone. Biological activities of thymic factors and hormones

Many effects have been attributed to thymic hormones. These fall into four main categories: phenotypic modulation of progenitor T-cells; functional effect on peripheral immune system such as enhancement or suppression of peripheral T cell response and restoration of T-cell function in thymus-deprived animals (Ritter and Crispe, 1992); modulation of hypothalamic-pituitary axis function such as stimulation of gonadotropin releasing hormone (GnRH) release from hypothalamus (Rebar et al., 1981c); and modulatory roles on ovarian development and functions such as elevation of germ cell number in fetal ovaries co-cultured with thymus gland (Prepin, 1991)

STEROIDOGENESIS IN THE GRANULOSA CELLS

Granulosa cells are the cellular source of the two most important ovarian steroids, estradiol and progesterone. Although granulosa cells and their luteinized counterparts are capable of producing progesterone independent of other cell types, the biosynthesis of estrogens requires the interaction of granulosa cells and neighboring theca cells. This situation is explained through the "two-cell, two-gonadotropin hypothesis" (Fortune and

Armstrong 1978, Dorrington et al., 1978).

According to this hypothesis, LH stimulates the biosynthesis of androgens from cholesterol in the theca interna compartment (Tang et al., 1980; Fortune and Armstrong,

1977). Androgens diffuse across the lamina basalis and are converted to estrogens by aromatase enzyme complex, the action of which is stimulated by FSH in granulosa cells

(Dorrington et al., 1975). Since granulosa cells secrete progesterone in response to 9 gonadotropins, it is also possible that granulosa cell progesterone may diffuse into theca cells to serve as a substrate for androgen biosynthesis (Bjersing, 1967).

Although aromatase is mainly stimulated by FSH, in vitro studies with FSH-primed rat granulosa cells showed that LH also directly stimulates estrogen secretion (Wang et al.,

1981).

Progestin Biosynthesis

The granulosa cells contain all the necessary enzymes for the de novo biosynthesis of progesterone via cholesterol-pregnenolone biosynthetic pathway. The cellular cholesterol may be derived from one of three possible sources:

(1) cholesterol taken up from the blood in the form of circulating lipoproteins;

(2) cholesterol stored within the cell, either as free cholesterol, a constituent of cell membranes, or liberated from cholesterol esters stored within cytoplasmic lipid droplets;

(3) cholesterol synthesized de novo in the granulosa cell from 2-carbon components

(primarily acetate), derived from the metabolism of carbohydrates, fats or proteins within the cell. The last process is dependent on the activities of the rate-limiting 3-hydroxy 3- methylglutaryl coenzyme reductase (HMG-CoA reductase) (Gore-Langton and Armstrong,

1988).

Cholesterol (C27-Sterol) -Sidejghain cleavage

The rate-limiting step in progestin biosynthesis is side-chain cleavage of cholesterol to pregnenolone. This step involves the cleavage of C-20,22 bond resulting in the C21 compound pregnenolone, and a 6-carbon fragment, isocaproic aldehyde. The enzyme system is localized on the matrix side of the inner mitochondrial membranes (Farkash et al., 1986).

It is a complex composed of three components: cytochrome P-450 side-chain cleavage

(SCC), a flavin adenine dinucleotide (FAD)-containing flavoprotein and the sulfur- containing heme protein luteodoxin or adrenodoxin. The reaction utilizes nicotine adenine dinucleotide phosphate generated within the mitochondria. Evidence summarized by

Liberman et al. (1984) suggests that the overall reaction in vivo probably does not involve stable, free hydroxylated intermediates. Instead, the cholesterol and the intermediates remain bound to the P-450 SCC until pregnenolone is formed and released.

36- hydroxvsteroid dehydrogenase (3B-HSD)

Pregnenolone is the key steroidogenic intermediate common to all classes of steroid hormones produced by the follicles, as well as by other steroidogenic tissues (Koritz and

Kumar, 1970; DuBois et al., 1981). Pregnenolone is converted to progesterone by a microsomal enzyme complex; A5-3B-hydroxysteroid dehydrogenase: A^-isomerase. The enzyme utilizes NAD+ as an electron acceptor and the reaction is irreversible under physiologic conditions. Similar, but not identical enzymes, convert 17a-OH-pregnenolone and dehydroepiandrosterone (DHEA) to 17a-OH-progesterone and androstenedione, respectively (Gower, 1984). 11

20tt-hydroxysteroid dehydrogenase

The secretion of progesterone by granulosa cells may be modulated by changes in the conversion of progesterone to its metabolites. The main route of progesterone breakdown is mediated by 20a-hydroxysteroid dehydrogenase which reversibly converts progesterone to its inactive metabolite, 20a-hydroxy-pregn-4-en-3-one (20a-OH-P) (Hsueh et al., 1984).

Unlike progesterone, 20a-OH-P is not capable of maintaining pregnancy (Gower, 1984)

5 a-Reductase

In the granulosa cells, 5a-reductase may also play a minor role in metabolism of progesterone and 20a-OH-P to 5a-reduced products. This reductase pathway is more active in immature rats, while the enzyme activity decreases following the first ovulation (Mizutani et al., 1979)

Estrogen biosynthesis

Ovarian estrone and estradiol-17B are synthesized from their androgen precursors, androstenedione and testosterone, respectively, by a microsomal enzyme complex, referred to as aromatase because of the aromatic structure of the products. Aromatization involves removal of the methyl group at C-19 as a formic acid, and rearrangement of ring A to an aromatic structure. These reactions require three moles of 0 2 and NADPH per mole of estrogen formed. In addition, these reactions require the participation of both cytochrome

P-450 and NADPH-cytochrome C-reductase enzymes. Estradiol-17B is the major ovarian estrogen even though androstenedione is the most abundant androgen in the ovary. The conversion of androstenedione to testosterone or estrone to estradiol is mediated by 176- hydroxysteroid dehydrogenase. The participation of aromatase and 176-hydroxysteroid dehydrogenase provides optimal estradiol biosynthesis in the granulosa cells (Hsueh et al.,

1984)

EVIDENCES FOR THE ROLE OF THE THYMUS GLAND ON

OVARIAN FUNCTION

The existence of a direct relationship between the thymus gland and the gonads was first suspected by Calzolari (1898), noting hypertrophy of the thymus in castrated male rabbits. Subsequent studies have confirmed that thymic enlargement occurs after gonadectomy of either sex in almost every species of domestic and experimental animal as well as man (Henderson, 1904; Chiodi, 1940). Conversely, androgens and estrogens have been shown to induce atrophy of the thymus when administered to intact animals

(Dougherty, 1952). More recently, sex steroid receptors have been characterized for thymocytes (Grossman et al.,1979; McCruden and Stimson, 1981; Sakabe et al., 1986).

Recent experimental evidence suggests that there is an important regulatory relationship between the thymus gland and reproductive function. Nishizuka and Sakakura

(1969) reported that removal of the thymus in the mouse between 2-4 days of age results in high frequency of ovarian dysgenesis, a condition characterized by lymphocyte infiltration of follicles, a decline in oocyte number, a subsequent decrease in the number of follicles and corpora lutea, and interstitial cell hypertrophy. The timing of the thymectomy appears to be critical since thymectomy before or after this age does not cause typical ovarian dysgenesis. 13

Before this, Flanagan (1966) had already reported a low fertility rate in the female mice of a new strain of congenitally athymic nude mice. Since these reports, Besedowsky and Sorkin

(1974) have reported that the females of the congenitally athymic mice also have a delay in vaginal opening and in first ovulation. Additionally, these females show accelerated follicular atresia and subsequent premature ovarian failure (Alten and Groscurth, 1975;

Lintem-Moore and Pantelouris, 1975).

Extensive histologic studies, in both neonatally thymectomized mice (Nishizuka and

Sakakura, 1969, and 1971) and congenitally athymic nude mice (Lintem-Moore and

Pantelouris, 1975 and 1976a), have shown that at birth the ovarian morphologies of congenitally athymic nude mice and neonatally thymectomized mice are very similar to those of normal mice. Starting from 24 days of age, a progressive dysgenesis was observed in the ovaries of the thymectomized mice and nude mice characterized by a slight reduction in the number of healthy-looking follicles and presence of lymphocytes in and around medium and large-sized follicles. At 30 days of age, there is a more pronounced decline of oocyte numbers and an increase in lymphocytic infiltration. The destruction of the follicular apparatus is usually complete by 60 days of age and is accompanied by a preponderance of hypertrophied interstitial-like cells and degenerating corpora lutea. From 60-120 days of age, further atrophy occurs depicted by near dominance of interstitial-like cells (Marmor and

Michael, 1984).

Additionally, endocrine imbalances accompanying ovarian dysgenesis have been found in both congenitally athymic and neonatally thymectomized mice. These animals show: 14

1) Variable levels of estrogens;

2) A reduction in levels of serum progesterone, serum corticosterone, luteinizing hormone

(LH), follicle-stimulating hormone (FSH), and growth hormone (GH) prior to puberty in neonatally thymectomized mice (Michael et al., 1980 and 1981);

3) Higher levels of testosterone at 60 days of age with increasing levels to 120 days of age

(Marmor and Michael, 1984);

4) Higher circulating levels of thymosin a l after 7 days of age (Michael et al., 1981);

5) A reduction in gonadotropin secretion has also been reported in both sexes of the congenitally athymic mouse (Rebar et al., 1981b and 1982b).

CURRENTLY SUGGESTED MECHANISMS TO EXPLAIN THE ROLE OF THE

THYMUS GLAND IROVARIANJUNCHQN

Immunologic mechanism

Findings related to this mechanism suggest that removal of the thymus at an early age (i.e. 2-4 days of age) disrupts or does not provide a chance for establishment of the delicate balance which gives the immune system the ability of differentiating 'self from

'non-self. Studies using indirect immunofluorescence and horse-radish peroxidase-labeled antibody techniques have demonstrated that neonatally thymectomized mice produce circulating autoantibodies against the ooplasm, indicating an autoimmune etiology for ovarian failure. The circulating antibodies and oophoritis are found only after thymectomy at the 3rd day of age but not after thymectomy at 0 and 7 days of age (Taguchi et al., 1980).

The autoimmune nature of the oophoritis is further demonstrated by the following: 15

1) Successful passive transfer of ovarian dysfunction into infant mice by an i.p. injection of spleen cells from adult mice with ovarian dysgenesis;

2) failure of passive transfer by injection of spleen cells from the same source, but pre-treated with antibodies directed against T-cell surface markers which renders T cells non-functional;

3) failure of passive transfer by injection of spleen cells from thymectomized and ovariectomized animals;

4) dysgenesis of an ovary grafted into a thymectomized animal with ovarian dysgenesis;

5) no dysgenesis if the above host had been previously ovariectomized (Taguchi and

Nishizuka, 1980; Sagaguchi et al., 1982a, b),

6) requirement for the specific population of T cells (defined by Lyt 1 + Qa-Ia -surface markers) taken from intact mice for prevention of the ovarian dysgenesis in thymectomized animals (Sagaguchi et al., 1982a and b).

As mentioned earlier thymosin ex 1 levels are higher in the neonatally thymectomized animals than in controls. This high level of hormone was considered to be the reason for high helper T cell population which may induce autoantibody production by B cells against ooplasm. Some studies have evaluated theability effect of antisera to thymosin a l in to inhibit lymphocyte differentiation and, in turn, overcome ovarian dysgenesis. However, these trials were unsuccessful in preventing ovarian dysgenesis and the high level of autoantibody to ooplasm, but the treatment was able to return the levels of estrogen and testosterone to those found in intact animals, suggesting the existence of a sensitive balance between the thymus and the ovary which may not be related to changes in only a single thymic hormone

(De Angelo and Michael, 1987). Further, De Angelo et al. (1989) assessed the aromatase 16 activity in neonatally thymectomized mice. They found that serum levels of estradiol were lower starting from 20-30 days of age. Levels of this steroid were increased at 30 days of age, rose to a higher level by 60 days of age, then decreased to levels similar to those of intact animals. Levels of testosterone were elevated at 60 days of age and continued to rise through 120 days of age. Elevation of testosterone was probable due to replacement of follicular structures with interstitial tissue. In the ia situ culture of ovaries, basal estrogen and progesterone production were very similar among the control and neonatally thymectomized mice, but in the presence of testosterone and FSH, secretion of these hormones into the medium was higher in thymectomized mice. These findings provide some evidence for the existence of autoantibodies to the ovary in thymectomized mice which causes damage of the ovary and abnormal steroid production (De Angelo et al., 1989). More recently, following a thorough investigation of the correlation between ovarian dysgenesis and the appearance of autoimmune antibodies against the oocytes, Kosiewicz and Michael

(1990) reported that anti-oocyte antibodies were not detectable until 30 days of age, although there was a dramatic decrease in the primordial follicle population at 10 days of age and the growing follicle population at 20 days of age in thymectomized mice. Therefore, there is no evidence to indicate that lower number of primordial and growing follicles at this age is due to an autoimmune reaction. However, by 30 days of age, their study found a correlation between follicular atresia and anti-oocyte antibody. Thus, Kosiewicz and

Michael (1990) suggested that, before 30 days of age the endocrine aspect of the thymus, rather than its immune aspect, could be responsible for ovarian dysgenesis. They also suggested that the symptoms of the ovarian dysgenesis after 30 days of age was a result of 17 tissue-specific autoimmune response.

Thymic-hypothlamo-pituitary-ovarian axis mechanism

Histologic observations of the ovaries of rodents, challenged with anti-gonadotropin, were very similar to those of both the congenitally athymic mice and the neonatally thymectomized mice (Hardy et al., 1974). Subsequently, Lintem-Moore and Pantelouris

(1976a, 1976b) postulated that the histologic changes in these mice might be the result of insufficient gonadotropin stimulation and demonstrated that administrations of pregnant mare serum gonadotropin (PMSG) to the athymic mice eliminates the histologic differences between the ovaries of athymic mice and their heterozygous littermates. Rebar et al. (1981a) later noted decreased levels of both LH and FSH in the pituitary gland and in the circulation of immature athymic mice and decreased levels of circulating estrogens in adult athymic mice.

Additionally, Rebar et al. (1981b) observed reduced concentration of GnRH in hypothalamic extract of athymic nude mice at ages of 10,20 and 30 days in comparison with their heterozygous littermates. Then, they tested the ability of hypothalamic extract from 20- day-old athymic and heterozygous mice to release LH by injecting the extract intravenously into rats. The extract from athymic mice consistently released less rat LH than that from the heterozygous mice. At the same time, they discovered that the LH response of the athymic mice to exogenous GnRH administered in vivo and the maximal LH and FSH response of the dispersed pituitary cells to the GnRH in vitro were identical to those of the heterozygote

(Rebar etal., 1981b). To rule out the possibility of a direct effect of the thymus gland on the gonads, they examined the maximum binding of purified human chorionic gonadotropin (hCG) to crude membrane fractions of ovarian tissue from the athymic and normal heterozygous females of various ages. Their preliminary data showed that there was no decrease in maximum binding of hCG to its receptor in the ovary of athymic nude mice relative to those of heterozygotes.

Thus, these data did not provide any evidence of an inherent ovarian defect in athymic mice.

Rebar et al. (1981c) postulated that the thymus gland might exert its effect through the hypothalamic-pituitary unit. They were able to demonstrate that thymosin 84, but not thymosin a l, stimulated secretion of GnRH from medial basal hypothalami from normal cycling female rats, superfused in vitro. They also observed a release of LH from pituitary glands superfused together with hypothalami, but no release of LH in response to thymosins when pituitaries were superfused alone. Afterward, in confirmation of the observations of

Rebar et al. (1981c), Hall et al. (1982) noted that thymosin 84, but not thymosin al, increased serum LH levels in the adult female when administered by intraventricular injection.

Thymosin fraction 5 was also shown to stimulate secretion of GnRH from hypothalami of adult, castrated and prepubertal male and female rats and mice, when incubated in vitro (Rebar et al., 1982a, 1983). However, the direct effect of the thymic factors on either basal or GnRH-stimulated LH and FSH secretion from cultured anterior pituitary cells of ovariectomized rats was not demonstrated. In a similar experiment, Rebar et al.

(1982a, 1983) also failed to show any direct effect of the thymic peptides on estrogen and androgen productions from granulosa and theca cells, respectively (Rebar et al., 1982a, 19

1983). Strich et al. (1985) demonstrated that in vivo administration of a crude thymic extract

(containing several thymic peptides) and serum thymic factors can decrease hypothalamic concentration of the GnRH (due to an acceleration of secretion) in normal male heterozygous mice and may increase pituitary concentrations of gonadotropins in prepubertal mice in general. Based upon their studies in athymic mice with natural and synthetic thymic peptides, Rebar (1984) suggested another mechanism for the role of the thymus gland on ovaries. This mechanism is dependent on the hormonal interaction between the hypothalamic-pituitary unit and thymus at the early age of development. According to this hypothesis, in the absence of the thymus at an early age, the development of the hypothalamus is incomplete. Thus, hypothalamus cannot release sufficient GnRH, which in turn, causes gonadotropin deficiency and ovarian dysgenesis.

Evidence for the existence of a direct effect of thymic peptides on ovaries

Both of these suggested mechanisms for ovarian dysgenesis, one of which depends on autoimmune disorder, and the other on the hormonal interactions between the hypothalamic-pituitary unit and thymus, have tended not to give much chance to the existence of a direct hormonal influence of the thymic peptides on the ovary. However, a considerable body of recent experimental evidence suggests that there is a very close influence of immune cell secretory products on ovarian function (Adashi et al., 1994;

Kendall and Clarke, 1994). Among such very recent experimental evidence, the most interesting observations related to our current study are:

1) Enhancement of gonadotropin-induced progesterone secretion from cultured porcine 20

granulosa cells by serum thymic factor (STF). In this study, STF significantly

enhanced both FSH and LH stimulated progesterone secretion over 2-4 days of

incubation, even though basal secretion was not altered by STF (Ledwitz-Rigby and

Scheid, 1990). This study suggested that STF, a nanopeptide produced by the thymus

gland, could play an important role in granulosa cell maturation.

2) Observation of a higher number of germ cells in fetal ovaries cultured in medium

supplemented with thymulin or co-cultured with fragments of fetal thymus as

compared to the control ovaries (Prepin, 1991).

3) Stimulation of basal and inhibition of FSH-induced progesterone and estradiol secretion

by thymosin fraction 5 (TF-5), which is a collection of a number of thymic peptides

including thymosin a l and thymosin 64, in cultured rat granulosa cells (Gorospe et

al., 1993).

4) Detection of a high level of prothymosin-a mRNA (Oikawa et al., 1990), thymosin- 04

and P 10 mRNA and protein in rat ovaries. More interestingly, PMSG, hCG and

PGF2o modulated the differential expression of these hormones (Hall et al. 1991a, b).

Objectives of this study

Our current study attempts to demonstrate the presence of thymic factors in thymic epithelial cell culture-conditioned medium (TCM) that modulate steroidogenic activity and morphology of cultured rat granulosa cells. In addition, this study attempts to characterize these thymic factors and their action on cultured rat granulosa cells. CHAPTER I

THYMIC EPITHELIAL CELL CULTURE CONDITIONED-MEDIA (TCM)

STIMULATES STEROIDOGENESIS IN CULTURED RAT GRANULOSA

CELLS

INTRODUCTION

In the last few years, a considerable amount of evidence has accumulated supporting the fact that the immune system has an important regulatory role in ovarian function (Marchetti,

1989; Adashi et al., 1990). The normal physiological development of reproductive function depends on the presence of an intact immune system, since immunologically suppressed or incompetent animals show numerous reproductive disorders. For example, congenitally athymic nude mice show delays in vaginal opening and in first ovulation (Besedovsky and

Sorkin, 1974). In addition, these females show accelerated follicular atresia and premature ovarian failure (Lintem-Moore and Pantelouris, 1975). Similarly, neonatally thymectomized female mice show ovarian dysgenesis (Nishizuka and Sakakura , 1969), which can be prevented by thymic or splenic grafts, or by injections of cell suspension from thymus, spleen or lymph nodes (Kojima etal., 1973; Sakakura and Nishizuka, 1972). More recently, it has been reported that testosterone levels of adult male rats become undetectable following splenectomy. However, the administration of spleen dialysate restored the testosterone levels

21 22 to the values observed in the controls (Pawlikowski et al., 1990).

Other recent evidence has unequivocally shown that immune system secretory products

(cytokines) regulate ovarian granulosa cell function in vitro. For example, interleukin-1 (IL-

1), a macrophage secretory product, inhibits steroidogenesis and differentiation of rat

(Gottschall et al., 1987, 1988, and 1989) and porcine (Fukuoka et al., 1988, 1989a and

1989b) granulosa cells in vitro. In addition, another macrophage secretory product, tumor necrosis factor-a (TNF-a), has been shown to have regulatory roles on follicular steroidogenesis in vitro (Emota and Baird, 1988, Roby and Terranova, 1988 and 1990).

Furthermore, novel factors derived from rat spleen cell cultures have been shown to modulate basal and gonadotropin-stimulated steroid secretion by cultured rat granulosa cells

(Gorospe and Kasson, 1988; Hughes et al., 1991). A similar type of result was reported in the pig, when the effect of spleen cell conditioned medium was tested on porcine granulosa cell steroidogenesis (Hughes et al., 1990).

It is a well-recognized fact that thymic epithelial cells are the main source of thymic hormones, and through these secretory factors, thymus epithelium regulates T-cell differentiation and maturation in vivo (Lewis et al., 1978). In addition, the thymic epithelial cells were shown to produce the thymic hormones (i.e. thymulin) in vitro, and this in vitro produced thymulin was shown to be active in stimulating lymphocyte cell proliferation

(Dardenne et al., 1987 and 1989). Interestingly, it was shown that serum thymic factor

(STF), a nanopeptide produced by the thymus gland, could enhance gonadotropin-induced progesterone (P4) secretion in cultured porcine granulosa cells (Ledwitz-Rigby and Scheid,

1990). Hence, to further explore the potential effects of thymic factors on ovarian granulosa 23 cell function, conditioned medium from thymic epithelial cell culture (TCM) might be used as a potential source for the thymic factors. In this study, thymic epithelial cells were isolated and cultured for TCM production. Following its production, the effect of TCM on steroidogenesis in cultured rat granulosa cells was examined. In addition, some physicochemical treatments (i.e., temperature, acetone treatment, acidification and activated charcoal sedimentation) were employed to characterize the nature of the factor(s).

MATERIALS AND METHODS

Animals

For the thymic cell culture, 21- to 23-day old Sprague-Dawley female rats were used.

After animals were sacrificed, thymi were aseptically removed for cell culture preparation.

For the preparation of the granulosa cell culture, immature Sprague-Dawley female rats were used. When they reached 21-23 days of age, they were injected with 250 pg DES/day in 100 pi com oil s.c. for five days (5 mg DES/kg body weigth/day) to induce granulosa cell mitogenic activity. Eight hours after the last DES injection, the rats were sacrificed and ovaries were removed for granulosa cell preparations.

Chemicals and reagents

Fetal Bovine Serum (FBS) was obtained from Hyclone Laboratories, Inc. (Logan, Utah,

U.S.A.). The antibiotic-antimycotic mixture (penicillin, streptomycin, and amphotericin B) and trypsin were purchased from Gibco Laboratories (Grand Island, NY, U.S.A.). Anti­ keratin rabbit antiserum (lot # 901676) was purchased from Calbiochem (San Diego, CA, 24

U.S.A.), Dextran-T 70 was obtained from Pharmacia (Gaithersburg, MD, U.S.A.), and

[l,2,5,7,21-3H(N)]-progesterone and [2,4,6,7-3H(N)]-estradiol were purchased from New

England Nuclear (Boston, MA, U.S.A.). Progesterone and estradiol antisera were obtained from Endocrine Sciences (Tarzana, CA, U.S.A.). All other chemicals were purchased from

Sigma Chemical Co. (St. Louis, MO, U.S.A.), Aldrich Chemical Co. (Milwaukee,WI,

U.S.A.) or Fisher Scientific (Pittsburgh, PA, U.S.A.).

PREPARATION OF CONDITIONED MEDIA

Preparation of the Thymic cell culture Conditioned Medium (TCM)

Thymic cells were isolated using the technique previously described (Sun et al., 1984).

Briefly, thymi were removed from 4-5 female rats which were 26- to 28-days old. The thymi were minced thoroughly with scissors and the tissue fragments were rinsed and centrifuged

(50 x g; for 1 min) five times in 40 ml Hank's Calcium-Magnesium Free Balanced Salt

Solution (HCMF) to remove most of the lymphocytes. The remaining tissue fragments were digested with a mixture of trypsin (0.125%) and EDTA (0.01%) in HCMF at 37°C with stirring for 45 min. The released thymic cells were washed, counted and plated in 25 cm2 culture flasks at 2 x 106 cells/ml density. (Only relatively larger epithelial-like cells were counted.) The culture medium used was Minimum Essential Medium D-Valine modification(MEM) supplemented with 10% FBS, 100 U/ml penicillin, 100 |ig/ml streptomycin sulfate and 250 pg/L amphotericin B. The average viability and total yield of the epithelial- like cells were 90% and 19 x 105/rat, respectively. After 24 hours in culture

(95% air, 5% C 02, 100% humidity, 37°C), the medium was replaced and unattached cells 25

(mostly lymphocytes) were removed. The attached cells were allowed to grow for an additional three days and the medium was replaced with fresh medium. The cells were cultured for an additional five days and then the medium was collected, centrifuged and stored at -20°C until used for evaluation (Plate III).

Preparation of the Heart cell Conditioned Medium (HCM)

Hearts were aseptically removed from female rats (21- to 23-days old) after sacrificing the animals. The hearts were sliced and minced into small pieces with scissors. The tissue particles were enzymatically digested with 0.1% collagenase type-IV in HCMF for 1 hour at 37°C. At 10 minute intervals, free single cells were removed and fresh buffer added for volume replacement. After two washes, single cells were suspended in MEM supplemented with 10% FBS and incubated the same way as thymic cell culture. Then, HCM was evaluated in the granulosa cell culture as a control medium (Plate IV). We selected heart conditioned medium as a control medium in previous studies it has been used as control medium by others (Gorospe and Kasson 1988), and incidentally, the heart is located in close proximity to the thymus gland.

Preparation of the Mock Extract (ME1)

To alleviate the potential nonspecific effects of fetal bovine serum components, ME was prepared in an identical fashion as TCM and HCM, but in cell-free culture dishes. ME was then evaluated in the granulosa cells culture as a control medium. 26

Physicochemical treatments of TCM

To characterize the stimulatory factor(s) in TCM, TCM was exposed to the various physicochemical treatments and then added to the granulosa cell cultures at the level of 25%.

In this experiment After 48 hours of treatment, the medium was collected for progesterone radioimmunoassay (RIA).

Temperature: Two separate sets of aliquots of TCM were placed in 100° and 56CC water baths for 20 and 30 min, respectively. After these incubations, the aliquots were centrifuged (500 x g; 5 min) to remove the precipitated material and the supernatant was allowed to cool and then immediately tested in the granulosa cell culture.

Acetone treatment: Acetone was added to TCM (2:1; acetone:TCM) and the mixture was vortexed for 2 min at room temperature. After centrifugation (500 x g; 5 min), the supernatant was removed and acetone was evaporated under a steady stream of N2. The remaining medium was used for the granulosa cell culture.

Activated charcoal treatment: 2 mg of activated charcoal/ml of TCM and 0.2 mg

Dextran-T 70/ml of TCM were added to aliquots of TCM. The aliquots were then incubated at 4°C for 10 min and centrifuged (600 x g, for 10 min, 4°C). The supernatant was removed and tested in the granulosa cell culture.

pH: TCM was acidified from pH 7 to pH 3 by the addition of 1 N HC1. After centrifugation, the supernatant was neutralized with 5 N NaOH and tested in the granulosa cell culture. 27

Preparation of granulosa cell culture

Rat granulosa cells were isolated using the nonenzymatic needle puncture method previously described (Hsueh et al., 1980). Briefly, ovaries were aseptically removed, cleaned of all fat and connective tissue, and placed in Dulbecco's Modified Eagle's Medium and

Ham's Nutrient Mixture F-12 (DME/F-12). Granulosa cells were freed from follicles by gentle and repeated puncture of the ovaries with a sterile bundle of beading needles (cat #

17860; William Prym, Inc. Dayville, CT, U.S.A.). The isolated granulosa cells were collected and washed twice by centrifugation (250 x g for 3 min, 4°C) in DME/F-12. The average cell viability was 44% and the average viable cell yield from one rat was 2.4 x 106.

The granulosa cells were plated at a density of 4 x 105 viable cells/well in 12-well tissue culture plates (Corning Inc, Oneonta, NY, U.S.A.) containing 1 ml DME/F-12 medium supplemented with 5 pg/ml insulin, 5 pg/ml transferrin, 10 ng/ml selenium, 10 ng/ml

Epidermal Growth Factor (EGF), 100 U/ml penicillin, 100 pg/ml streptomycin sulfate and

250 pg/L amphotericin B. The cells were first cultured for 24 hours at 37C° under a water- saturated atmosphere, and gassed with 95% air and 5% C 02 (Plate V).

In a pilot study, following 24 hours initial incubation under the above conditions, cultured granulosa cells were exposed to TCM, HCM or ME at the doses of 2.5-40%. The cells were exposed to the conditioned media for 24 hours. Then, culture media were collected for progesterone RIA and the granulosa cells were collected for protein and intracellular cAMP content determination. The protein content of cultured rat granulosa cells was determined by the dye binding method of Bradford (1976) and cAMP content was determined by RIA as previously described (Steiner et al., 1972). 28

In a time course study, the cells were exposed to 10% TCM, HCM and ME. In a dose- response study, the granulosa cells were incubated in increasing amounts of TCM, HCM and

ME for 48 hours. Then, the medium was collected for progesterone and estradiol RIA. The progesterone and estradiol levels were determined by radioimmunoassay (RIA) as previously described (Stouffer et al., 1976). The intra- and interassay coefficients of variation in RIA were 3.9% and 13.2%, and 3.2% and 11.3% for progesterone and estradiol-17 p, respectively.

The cross-reactivities of progesterone antiserum were 15% with pregnenolone and less than

1% with either estradiol-17p and/or testosterone using the 50% inhibition of binding method, as described by Abraham (1969).

The steroid hormone data points were expressed as ng or pg/ml medium. We did not find it appropriate to express the steroid hormone secretion as a function of viable granulosa cell number plated at the beginning of culture because of the possible increase of cell numbers during incubation due to cell division.

Effect of thymosin «1 and/or thvmulin on steroidogenesis in cultured rat granulosa cells

To determine whether the known thymic factors, namely thymosin a, and thymulin, have an effect on the granulosa cells similar to that of TCM, the granulosa cells were treated with 5,25,125 and 250 ng/ml thymosin a 1 and/or thymulin for 24 or 48 hrs. We selected these two thymic factors, because they were widely used and their biological activities have been well described in various in vitro conditions. After incubation, the basal progesterone and estradiol secretion into the medium was determined. In preliminary experiments following the above experiment, we tested a wider range of doses of thymulin and thymosin 29 a 1 (1-2500 ng/ml) for their stimulation of progesterone or estradiol secretion in cultured rat granulosa cells. Thymulin and/or thymosin a 1 did not show any stimulatory activity for the for progesterone and estradiol secretion in cultured rat granulosa cells, at any doses tested

(data not shown)

Immunofluorescent staining for keratin in cultured rat thymic cells

It has been shown that keratin as a subunit of tonofilaments is present in all epithelial cells, but not other cell types (Sun et al., 1984). We wanted to find out the ratio of epithelial cells to other cell types in thymic cell culture. The thymic cells which were grown on the surface of sterile cover slides were stained using indirect immunofluorescent staining method as previously described (Gibson-D'ambrossio et al., 1987).

Lung cells, which were prepared from young female rat lungs in the same fashion as thymic cells, and hepatocytes of adult female rats, which were isolated by in situ perfusion, were also stained by immunofluorescent staining as negative and positive controls, respectively.

Data analysis

Data points are shown as the mean ± S.D. of the quadruplicate culture dishes from a representative experiment. Each experiment was repeated twice. The result from these two experiments demonstrated essentially identical trends. The difference between the effect of different treatments were evaluated by two sample J; test. MINITAB data analysis software

(MINITAB, Inc.) was used. The value of P < 0.05 was considered as significant. 30

RESULTS

Pilot Study

In this study, a dose-dependent stimulation of TCM on granulosa cell progesterone secretion was observed (Figure 1) The stimulatory action of TCM was maximal at the dose of 10%. Therefore, this dose of TCM was used in the time-course studies. In the same set of experiments intracellular cAMP was also measured. No parallel increase in intracellular cAMP was observed (Figure 2).

Time course study

A time course study was designed to determine at which incubation time, 10% TCM would maximally stimulate progesterone secretion. The granulosa cells were incubated with

10% (100 pl/ml medium) TCM, HCM, or ME for 0, 15, 24, 36,48, and 60 hours. In this experiment (Figure 3A), even at the shortest incubation time (15 hours), TCM caused a significantly higher progesterone secretion as compared to HCM and ME (P<0.05). With time, progesterone levels increased gradually to a plateau and the stimulatory action of TCM became more profound with time. The results of this experiment were also expressed in terms of the hourly production of progesterone (ng progesterone /ml medium/hr). Expressed in this way, the results indicate that the elevation in progesterone secretion, which was stimulated by 10% TCM was independent of the progesterone that accumulated during the longer incubation intervals (Figure 3B). The maximum stimulation of progesterone per hour by 10% TCM occured at 48 hours of incubation. Both progesterone and estradiol levels in the media were measured. Estradiol secretion in TCM-treated granulosa cells was 31 significantly higher than that in the controls at 36 hours incubation (P<0.05; TCM vs. HCM and ME). This difference became more readily apparent with longer incubation times

(Figure 4A). Although the estradiol concentration in the medium increased with incubation time in all treatment groups, the effect of 10% TCM could be differentiated from the other treatments indicating that TCM has a specific stimulatory action on estradiol secretion in cultured rat granulosa cells which is independent of the incubation periods (Figure 4B).

Dose response study

After it became evident that maximum progesterone and estradiol secretion occurred at the 48-hour incubation period, it was decided to generate a dose-response curve using this optimal incubation time. The granulosa cells were incubated with increasing amounts of conditioned media ranging from 1.5 to 48% (15-480 pl/ml medium). Three percent TCM stimulated progesterone secretion above that in cells treated with the same amounts of HCM and ME. The difference in progesterone secretion continued to increase up to the highest concentration tested (48%) (P<0.05 TCM vs HCM and ME; Figure 5). In contrast to progesterone secretion, TCM could stimulate significantly higher levels of estradiol secretion at only 12, 24, and 48% concentrations (Figure 6). The HCM- and ME-treated groups were not significantly different with respect to progesterone and estradiol secretion (P > 0.05).

The progesterone and estradiol concentrations were increased in HCM- and ME-treated granulosa cell culture media, as well as TCM-treated granulosa cell culture medium. Since this gradual increase in concentrations of progesterone and estradiol was observed in the ME- treated group as well, the observed increase is probably due to the non-specific stimulatory 32 effect of increasing levels of FBS present in the all of the conditioned media. The important point here is the significantly higher increase in progesterone and estradiol concentration in the culture medium of the granulosa cells that were exposed to TCM.

Effect of TCM on the progesterone secretion after the physicochemical treatments

To gain a greater understanding about the nature of factor(s) in TCM which is/are responsible for the stimulation of steroidogenesis in cultured rat granulosa cells, several physicochemical characteristics were assessed. Because the most prominent effect of TCM is on progesterone secretion, we tested the stability of this stimulatory activity by subjecting

TCM to various physicochemical treatments.

Boiling TCM at 100°C for 20 min before addition to the granulosa cell culture totally eliminated the stimulatory activity of TCM, whereas exposure of TCM to 56°C had minimal effect. Furthermore, addition of acetone and acidification to pH 3 followed by neutralization also eliminated the stimulatory activity. In contrast, activated charcoal treatment had little effect on TCM's stimulatory activity (Figure 7).

Effect of thymosin a 1 and/or thymulin on steroidogenesis in cultured rat granulosa cells

No statistically significant effects of thymulin and/or thymosin a l observed on either progesterone or estradiol secretion in cultured rat granulosa cells during the 24 or 48 hours treatment periods (Figure 8,9,10 and 11). 33

Immunofluorescent staining for keratin in rat thymic cells

The acetone-fixed cultured rat thymic cells were stained using the indirect immunofluorescent method for the presence of keratin, a subunit of tonofilaments in epithelial cells. After staining, cells were examined by fluorescence microscopy (Nikon

Microphot-FX). Our results showed that 91±5% of the cultured thymic cell population were epithelial cells (Plate VI).

DISCUSSION

In the present study, we used a primary culture of immature, DES-treated rat granulosa cells to demonstrate that secretory product(s) present in the thymic cell culture conditioned medium (TCM) is/are capable of stimulating steroidogenesis in cultured rat granulosa cells.

Under the described conditions, granulosa cells produced a measurable amount of basal progesterone. Treatment of the cultured granulosa cells with TCM resulted in a dose- and time-dependent increase in both progesterone and estradiol secretion. TCM had its maximum stimulatory effect at 48 hours of incubation at a concentration of 10%. The effect of TCM on progesterone secretion was more prominent than its effect on estradiol secretion.

Although in incubation periods shorter than 36 hours the stimulatory action of TCM on estradiol secretion was minimal, this action of TCM became more apparent when the incubation period was extended beyond 36 hours. Moreover, neither thymosin a 1 and thymulin nor HCM and ME mimicked the stimulatory action of TCM under the same experimental conditions. Thus, the lack of stimulation by the former suggests that the stimulatory action of TCM is the result of some thymic product(s) other than these known 34 thymic factors. In addition, the lack of stimulation by the latter suggests that this stimulatory action is specific to the product(s) secreted by thymic cells in culture rather than a non­ specific entity which may potentially be present in any type of cell culture or serum.

To increase synthesis and secretion of steroid hormones in granulosa cells, it is necessary to increase intracellular cAMP (Marsh, 1975). Thus, the failure to observe parallel increases in the intracellular cAMP content and progesterone secretion of the granulosa cells deserves special attention. A possible explanation for this finding is the following: There are reports indicating that alternative second messenger systems (e.g. inositol lipid metabolism and calcium signalling), in addition to the cAMP-mediated system, may be involved in steroid hormone biosynthesis in granulosa cells (for review, see Leung and Wang, 1989). The active stimulatory factor(s) in TCM may increase progesterone secretion through these potential alternative systems. In addition, it is well known that the secretion of progesterone by granulosa cells may be modulated by changes in its conversion to its major metabolite,

20a hydroxyprogesterone (Hsueh et al, 1984). The increase in progesterone secretion into culture medium may be due to inhibition by TCM of the conversion of progesterone to its metabolites rather than stimulation of its production. Thus, this possibility should be tested by measuring 20-hydroxyprogesterone along with progesterone.

Previously, Rebar et al. (Rebar et al., 1981) were able to demonstrate that thymosin- 64 but not thymosin-al, stimulated GnRH secretion from medial basal hypothalami from normal cycling female rats that were superfused in vitro. Rebar et al. (1982, 1983) also reported that thymosin fraction 5 stimulated secretion of GnRH from hypothalami of adult, castrated and prepubertal male and female rats and mice incubated in vitro. However, Rebar 35 et al. (1982, 1983) failed to show any direct effect of the thymic peptides on estrogen and androstenedione production from rat granulosa and theca cells, respectively. Rebar (1984), thus, proposed that the thymus gland might exert its effect through the hypothalamic-pituitary unit, but not directly on the ovaries. However, our current study suggests that, before rejecting the possibility of the existence of such a direct effect of the thymus gland, additional in vivo experiments should be done with the TCM. In agreement with the study by Rebar et al. (1982,1983), we were also unable to show any stimulatory action of thymosin a 1 and/or thymulin on basal progesterone and estradiol secretion from cultured rat granulosa cells. However, it was reported that serum thymic factor (STF; thymulin) enhanced gonadotrophin-induced progesterone secretion from cultured porcine granulosa cells

(Ledwitz-Rigby and Scheid, 1990). Since the experimental conditions are not completely identical in all of the aforementioned studies, including ours, the discrepancies among them may not necessarily be contradictory. Considering our own study, however, we can conclude that the stimulatory action of TCM on progesterone secretion was not mimicked by either thymosin a, or thymulin when the granulosa cells were incubated with increasing doses of these two peptides separately or in combination for 24 and 48 hours.

Because the main source of thymic hormones in the thymus gland is thymic epithelial cells (Dardenne et al., 1987 and 1989), we suspected that our initial observation of TCM's stimulatory action (Uziimcii and Lin., 1990) could be due to a thymic epithelial cell factor which was present in the TCM. It is known that in conventional cell growth medium, epithelial cells are outgrown by fibroblasts and macrophages (Bearsley et al., 1983). If the

L-valine in the conventional cell culture medium is substituted with D-valine which inhibits 36 fibroblast proliferation, we might be able to produce more thymic factor and, in turn, obtain more prominent stimulatory action by TCM (Nieburgs et al., 1985). Based upon these facts,

MEM culture medium containing D-valine was employed in the thymic cell culture.

To assess the above expectations, indirect immunofluorescent staining of thymic cells was employed to detect keratin, a structural component of epithelial cells. The results of this study showed that a vast majority (91 ± 5%) of the cultured thymic cell population used for our study was epithelial cells. Thus, from this result we can speculate that the factor(s) in

TCM affecting steroidogenesis in cultured rat granulosa cells are produced by the thymic epithelial cells.

The results from the physicochemical treatment study demonstrated that the TCM- stimulatory activity is heat, acid and acetone labile, but cannot be sedimented by activated charcoal. Therefore, it is reasonable for us to assume that the factor(s) possessing the biological activity is, by nature, a protein. Additional studies are required to further determine the nature of this active factor(s).

Since thymic cells did not grow well in serum-free culture medium, supplementation of

10% FBS to the culture medium was essential. Therefore, TCM, HCM and ME contain 10%

FBS. However, during the physicochemical treatments of TCM, charcoal treatment was used to remove some serum components from TCM. Still, the charcoal-stripped TCM was able to stimulate progesterone secretion to a level comparable to that of the control TCM.

Although the levels were comparable, the difference was still statistically significant.

Therefore, additional studies with a more purified product of TCM are required before a definite conclusion may be drawn. 37

In conclusion, our data provide evidence for the existence of factor(s) in TCM which stimulate progesterone and estradiol secretion from cultured granulosa cells of DES-treated immature rats. In addition, our results provide evidence for the possibility of a direct effect of thymic factor(s) on ovarian granulosa cell function.

SUMMARY

Thymic cells from immature female rats were isolated and used for production of thymic cell culture conditioned medium (TCM). Granulosa cells were obtained from immature diethylstilbestrol (DES)-treated rats. TCM stimulated basal progesterone and estradiol secretion from the granulosa cells in a dose and time dependent manner. Maximal stimulation occurred at 48 hours of incubation, during which period TCM caused approximately 5 times more progesterone secretion than heart cell conditioned medium

(HCM) and mock extract (ME). The maximum stimulatory dose for progesterone secretion was 48% TCM which caused 7 times more progesterone secretion than controls. Under the same maximum stimulatory conditions, however, TCM only approximately doubled estradiol secretion compared to levels secreted in the presence of HCM or ME. Thus, the effect of

TCM on progesterone secretion was more prominent than its effect on estradiol secretion, and these effects were not mimicked by HCM, thymosin a, or thymulin. The stimulatory action of TCM on steroidogenesis did not appear to be mediated by the cAMP system. The stimulatory factor(s) in TCM were heat, acid and acetone labile, but could not be sedimented by activated charcoal. Thus, the present study demonstrates that the secretory product(s) of thymic epithelial cells can stimulate steroidogenesis in cultured rat granulosa cells. Our data imply that thymic factor(s) may have a direct effect on ovarian function. FIGURES Figure 1. Effect of different doses of TCM (thymic cell culture medium), HCM (heart cell culture medium) and ME (mock extract) on basal progesterone secretion from cultured rat granulosa cells. The cells were incubated with the conditioned media for an additional 24 hours after the initial 24 hour incubation. Data points are shown as the mean ± S.D. of quadruplicate culture wells from a representative experiment. The experiment was repeated twice under similar experimental conditions and identical results were obtained. *= significantly different from the control (HCM and ME). The difference between the effects of 2.5 % TCM and 10% TCM on progesterone secretion is statistically significant.

40 Progesterone secretion (pg/ml/24 hr) 50 - 2500 7500 00 - 5000 0 Percent volumes of conditioned media conditioned of volumes Percent 05 .0 0 < P M C H Figure1. = ± D. .D S ± n a e M = T 4 = N 41 Figure 2. Lack of any significant effect of different doses of TCM, HCM and ME on intracellular cAMP formation in cultured rat granulosa cells. The cells were incubated with the conditioned media for an additional 24 hour after the initial 24 hour incubation. Data points are shown as the mean ± S.D. of quadruplicate culture wells from a representative experiment. The experiment was repeated twice under similar experimental conditions and identical results were obtained.

42 cAMP (pmoles/pg protein/24 hr) 300 - 450 0 - 600 750 5 - 150 0 Percent volumes of conditioned media conditioned of volumes Percent .5 2 E M M C H M C T = ± D. .D S ± n a e M = T 4 = N Figure 2. 10

20 0 4 43 Figure 3. The effect of 10% TCM on basal progesterone secretion from cultured rat granulosa cells in time-course study. The cells were incubated with 10% TCM, HCM or ME up to 60 hours after the initial 24 hour incubation. The increase in progesterone secretion is time-dependent (A). Data points are shown as the mean ± S.D. of quadruplicate culture wells from a representative experiment. The experiment was repeated twice under similar experimental conditions and identical results were obtained. *= significantly different from the control. The above data were also expressed the hourly production of progesterone by cultured rat granulosa cells (B). This figure shows that maximum progesterone production was obtained at 48 hours of incubation.

44 Progesterone secretion (ng/ml/) 0 - 40 0 - 30 50 20 10 0 - - P<0. 5 .0 0 < P = ± S. . .D S ± n a e M M C T % 0 1 ME E M % 0 1 M C H % 0 1 Incubation time (hours) time Incubation 6 3 4 2 Figure3. 8 60 48 45 Progesterone secretion (ng/ml/hour) Figure 3 (continued). 3 Figure 0.6 0.8 0.0 0.2 - 0.4 1.0 - - - = 4 = N = Mean D. .D S ± n a e M = T B =P<0. 5 .0 0 < P = * 5 0 6 8 4 6 3 4 2 15 M C H % 0 1 E M % 0 1 M C T % 0 1 Incubation time (hours) time Incubation Figure 4. Effect of 10% TCM on basal estradiol secretion from cultured rat granulosa cells in time-course study. The cells were incubated with 10% TCM, HCM or ME up to 60 hours after the initial 24 hour incubation. The increase in estradiol secretion is time-dependent (A). Data points are shown as the mean ± S.D. of quadruplicate culture wells from a representative experiment. The experiment was repeated twice under similar experimental conditions and identical results were obtained. * = significantly different from the control. The above data were also expressed as the hourly production of estradiol by cultured rat g ranulosa cells (B). This figure shows that the difference in estradiol production between TCM-treated and control (i. e. HCM and ME-treated) groups was the highest 48 hours of incubation.

47 Estradiol secretion (pg/ml/) 0 0 2 120 6 - 160 0 - 40 0 - 80 0 - = ± D. .D S ± n a e M = T 4 = N A P<0. 5 .0 0 < P = * 0 E M % 0 1 M C H % 0 1 M C T % 0 1 Incubation time (hours) time Incubation 5 6 3 4 2 15 Figure 4. 8 60 48 Estradiol secretion (pg/ml/hour) Figure 4 (continued). 4 Figure 0 - 4 6 8 - = ± D. .D S ± n a e M = T 4 = N B P<0. 5 .0 0 < P = * 15 E M % 0 1 M C H % 0 1 M C T % 0 1 Incubation time (hours) time Incubation 4 2 6 3 8 4 60 49 Figure 5. Effect of different doses of TCM, HCM and ME on basal progesterone secretion from cultured rat granulosa cells. The cells were incubated with the conditioned media for 48 hours after the initial 24 hour incubation. The increase in progesterone secretion is dose- dependent. Data points are shown as the mean ± S.D. of quadruplicate culture wells from a representative experiment. The experiment was repeated twice under similar experimental conditions and identical results were obtained. * = significantly different from the control.

50 Progesterone secretion (ng/ml/48 hr) 20 30 10 0 = ± S. . .D S ± n a e M = T Percent volumes of conditioned media conditioned of volumes Percent = 4 = N = 05 .0 0 < P = * . 3. 0 12 .0 6 .0 3 1.5 E M M C H M C T Figure5. 4 2 * 48 51 Figured. Effect of different doses of TCM, HCM and ME on basal estradiol secretion from cultured rat granulosa cells. The cells were incubated with the conditioned media for 48 hours after the initial 24 hour incubation. The increase in estradiol secretion is dose- dependent. Data points are shown as the mean ± S.D. of quadruplicate culture wells from a representative experiment. The experiment was repeated twice under similar experimental conditions and identical results were obtained. * = significantly different from the control.

52 Estradiol secretion (pg/ml/48 hr) 400 4 - 240 2 - 320 6 - 160 0 - 80 0 Percent volumes of conditioned media conditioned of volumes Percent . 3. 0 12 .0 6 .0 3 1.5 = Mean D. .D S ± n a e M = T 4 = N =P<0. 5 .0 0 < P = * Figure 6. 53 Figure 7. Effect of 25% TCM after various physicochemical treatments on progesterone secretion in cultured rat granulosa cells. After the initial 24 hour incubation, the granulosa cells were incubated with 25% treated TCM for 48 hours. Data points are shown as the mean ± S.D. of eight culture wells from a representative experiment. The experiment was repeated twice under similar experimental conditions and identical results were obtained. *=significantly different from TCM (control).

54 Progesterone secretion (ng/ml/48 hr) 5 - 45 0 - 30 60 5 - 15 o N = 8 8 = N = ± S. . .D S ± n a e M = T =P<0. 5 .0 0 < P = * Figure 7. Figure * T ..... ' == TCM control TCM Basal progesterone 100°C heating TCM acetone TCM pH=3TCM C 56°TCM heating charcoalTCM control ME t. t m i Kill 1 1 Bfmlj * * 55 Figure 8. Effects of various doses of thymosin a l and thymulin alone and in combination, on basal progesterone secretion in cultured rat granulosa cells. After the initial 24-hour incubation, the cells were treated with these thymic hormones for an additional 24 hours. The treatments consisted of (1) thymosin a 1 only and (2) thymulin only, each given at the doses of 5,25,125 and 250 ng/ml; and (3) the combination of thymosin a l and thymulin in which the final concentrations of each hormone in each treatment was 5,25, 125, and 250 ng/ml. Data points are shown as the mean ± S.D. of triplicate culture wells from a representative experiment. The experiment was repeated twice under similar experimental conditions and identical results were obtained. There are no statistically significant differences between treatment groups and control.

56 Progesterone secretion (pg/ml/24 hr) 1000 0 - 800 0 - 400 200 - 600 0 - Mean D. .D S ± n a e M = T 3 = N Thymosin ocl and/or thymulin (ng/ml) thymulin and/or ocl Thymosin 0 f n in s o m y h T n a + hymulin u m y th + 1 a in s o m y h T lin u m y h T Figure8. a 1 0 5 2 5 2 1 5 2 57 Figure 9. Effects of various doses of thymosin a 1 and thymulin alone and in combination, on basal estradiol secretion in cultured rat granulosa cells. After the initial 24-hour incubation, the cells were treated with these thymic hormones for an additional 24 hours. The treatments consisted of (1) thymosin a 1 only and (2) thymulin only, each given at the doses of 5,25,125 and 250 ng/ml; and (3) the combination of thymosin a l and thymulin in which the final concentrations of each hormone in each treatment was 5,25,125, and 250 ng/ml. Data points are shown as the mean ± S.D. of triplicate culture wells from a representative experiment. The experiment was repeated twice under similar experimental conditions and identical results were obtained. There are no statistically significant differences between treatment groups and control.

58 Estradiol secretion (pg/ml/24 hr) 20 10 - 15 0 - 5 - Thymosin Thymosin = ± S. . .D S ± n a e M = T 3 = N i 0 T n in s o m y h T n in s o m y h T lin u m y h T a 1 and/or thymulin (ng/ml) thymulin 1 and/or Figure 9. a a 1 t lin u m y th + 1 1 5 0 5 2 25 1 5 2 59 Figure 10. Effects of various doses of thymosin a 1 and thymulin alone and in combination, on basal progesterone secretion in cultured rat granulosa cells. After the initial 24-hour incubation, the cells were treated with these thymic hormones for an additional 48 hours. The treatments consisted of (1) thymosin a 1 only and (2) thymulin only, each given at the doses of 5,25, 125 and 250 ng/ml; and (3) the combination of thymosin a l and thymulin in which the final concentrations of each hormone in each treatment was 5,25, 125, and 250 ng/ml. Data points are shown as the mean ± S.D. of triplicate culture wells from a representative experiment. The experiment was repeated twice under similar experimental conditions and identical results were obtained. There are no statistically significant differences between treatment groups and control.

60 Progesterone secretion (pg/ml/48 hr) 1000 1250 750 250 - 500 0 - = 3 T = ± S. . .D S ± n a e M = T 3; = N Thymosin Thymosin 0 f n in s o m y h T lin u m y h T n in s o m y h T a 1 and/or thymulin (ng/ml) thymulin 1 and/or a a Figure10. 1 1 hymuln lin u m y th + 1 5 2 0 5 2 5 2 1 61 Figure 11. Effects of various doses of thymosin a 1 and thymulin alone and in combination, on basal estradiol secretion in cultured rat granulosa cells. After the initial 24-hour incubation, the cells were treated with these thymic hormones for an additional 48 hours. The treatments consisted of (1) thymosin a 1 only and (2) thymulin only, each given at the doses of 5,25, 125 and 250 ng/ml; and (3) the combination of thymosin a 1 and thymulin in which the final concentrations of each hormone in each treatment was 5,25, 125, and 250 ng/ml. Data points are shown as the mean ± S.D. of triplicate culture wells from a representative experiment. The experiment was repeated twice under similar experimental conditions and identical results were obtained. There are no statistically significant differences between treatment groups and control.

62 Estradiol secretion (pg/ml/48 hr) 20 12 - 16 - 4 8 0 - - Thymosin a 1 and/or thymulin (ng/ml) thymulin 1 and/or a Thymosin = Mean D. .D S ± n a e M = T 3 = N 0 n in s o m y h T lin u m y h T in s o m y h T Figure11. a a 1 t lin u m y th + 1 1 0 5 2 5 2 1 5 2 63 PLATES

64 Plate I. A thymus gland section from a 21-day old female rat. The tissue was fixed, processed, sectioned and stained with Hematoxylin and Eosin staining (H&E). Note the cortex (C) consists of densely-packed small lymphoid cells. In the medulla (M), the cells are larger and sparsely distributed. Magnification : 400x.

65 66

Plate I. Plate II. An electron micrograph of a thymus gland from a 21 day old female rat. Note the epithelioid cells (E) have numerous mitochondria in a relatively larger cytoplasm. Magnification: 6000x.

67 68

Plate II. Plate III. Phase-contrast micrograph of thymic cell culture. Thymic cells were prepared from 21-23 days old female rats. The cells were at the 6th day of culture. Magnification: 350x.

69 70

Plate ID. Plate IV. Phase-contrast micrograph of cardiac cell culture. Cardiac cells were prepared from 21-23 day old female rats. The cells were at the 6th day of the culture. Magnification: 350x.

71 72

Plate IV. Plate V. Phase-contrast micrograph of rat granulosa cell culture. The cells were cultured in DME/F-12/S for 24 hours. Following this initial culture period, the cells were treated as described in the M aterials and Methods section. Magnification 350x.

73 Plate V. Plate VI. Cultured thymic cells that were stained immunohistochemically for keratin. On the 9th day of culture, the cells were stained as described in the Materials and Methods section. The stained cells appear green under fluorescence microscopy. This green staining appears white in this black and white micrograph. Magnification 350x.

75 Plate VI. CHAPTER II

CHARACTERIZATION OF THE STIMULATORY ACTIONS OF THYMIC

FACTORS ON BASAL AND GONADOTROPIN-INDUCED

STEROIDOGENESIS IN CULTURED RAT GRANULOSA CELLS

INTRODUCTION

Recent experimental evidence has clearly shown that immune system secretory products

(collectively called cytokines) play an indisputable role in ovarian function (Adashi, 1990).

For example, interleukin-1 (IL-1), a macrophage secretory product, inhibits steroidogenesis and differentiation of rat (Gottschall et al., 1987,1988, and 1989) and porcine granulosa cells

(Fukuoka et al., 1988,1989a,b) in vitro. In addition, IL-1 may also play a central role in the promotion of the ovulatory cascade in the rat (Hurwitz et al., 1992,1993; Kokia et al., 1992;

Brannstrom et al., 1993a). Like IL-1, another macrophage secretory product, tumor necrosis factor-a (TNF-a), also may contribute to the ovulatory process (Brannstrom et al., 1993b).

Possible regulatory roles of TNF-a in follicular steroidogenesis (Emota and Baird, 1988;

Roby & Terranova, 1990; Andreani et al., 1991) and luteolysis (Bagavandoss et al., 1990;

Fairchild Benyo & Pate, 1992) also have been suggested. In addition, it was reported that this cytokine might be produced locally and exert its action through protein kinase C in rat ovaries (Sancho-Tello et al., 1991,1992). Furthermore, novel factors derived from rat spleen

77 78 cell cultures have been shown to modulate basal and gonadotropin-stimulated steroid secretion in cultured rat granulosa cells (Gorospe & Kasson 1988; Hughes et al., 1991). A similar type of result was reported in the pig when the effect of spleen cell-conditioned medium was tested on porcine granulosa cell steroidogenesis (Hughes et al., 1990).

Although a close relationship between the thymus and the ovary has long been suggested

(Michael, 1983; Hall et al., 1992), the nature of this relationship has not been totally explained. At this point, it appears that normal development of ovarian function may depend upon the presence of an intact thymus gland. For example, congenitally athymic mice show delays in vaginal opening and in first ovulation (Besedowsky and Sorkin, 1974) and have an increased occurrence of follicular atresia and premature ovarian failure (Lintem-Moore and

Pantelouris, 1975). Earlier studies have noted a similar kind of morphologic disturbance in the ovaries of neonatally thymectomized female mice (Nishizuka and Sakakura, 1969) which could be prevented by thymic or splenic grafts, or by injections of cell suspensions from the thymus, spleen or lymph nodes (Kojima et al., 1973; and Sakakura and Nishizuka, 1972).

Recently, Gorospe et al. (1993) have shown that thymosin fraction 5 (TF-5), which is a collection of a number of thymic peptides including thymosin a l and thymosin B4, stimulates basal, but inhibits FSH-induced progesterone and estradiol. Previously, our laboratory has shown that conditioned media from cultured thymic epithelial cells (TCM) stimulates basal progesterone and, to a lesser extent, estradiol production (Uzumcii et al.,

1992). In light of our previous study, the objective of this study is to further explore the steroidogenic action of thymic factors in TCM. To this end, in this study, we have sought to accomplish the following aims: 1) To examine whether TCM may elevate progesterone 79 secretion simply by inhibiting its metabolism, the major metabolite of progesterone, namely

20a-hydroxy-progesterone, was measured in cultured rat granulosa cells following their treatment with TCM; 2) To examine whether the elevated estradiol secretion in the cells treated with TCM reflects a parallel elevation in the activity of aromatase, this enzyme's activity was measured by the method of [3H]-H20 release; 3) To examine potential mitogenic actions of TFs which are present in TCM, [3H]-thymidine incorporation assay was employed following exposure of rat granulosa cells to the conditioned medium; 4) Since any actions of TFs which are present in TCM on gonadotropin-induced steroidogenesis is more relevant to the physiological condition in the body, the effects of TFs on FSH-induced steroidogenesis in cultured rat granulosa cells were examined, as well as their action on basal steroidogenesis.

MATERIALS AND METHODS

Animals:

For the thymic cell culture, 21- to 23-day-old Sprague-Dawley female rats were used.

After animals were sacrificed, thymi were aseptically removed for cell culture preparation.

For the preparation of the granulosa cell culture, immature Sprague-Dawley female rats were used. At 21-23 days of age, the female rats were injected s.c. with 250 pg DES suspended in 100 pi com oil once per day for five days (5 mg DES/kg body weight). Eight hours after the last DES injection, the rats were sacrificed and ovaries were removed for granulosa cell preparations. 80

Chemicals and reagents:

Fetal Bovine Serum (FBS) was obtained from Hyclone Laboratories, Inc. (Logan, UT).

[l,2,5,7,21-3H(N)]-progesterone (109.50 Ci/mmol), [2,4,6,7-3H(N)]-estradiol (100.50

Ci/mmol), 20 a [ 1,2-3H(N)]-hydroxy-progesterone (51.20 Ci/mmol), [ 1P -3H]-androstenedione

(24.50 Ci/mmol) and [methyl 3H]-thymidine (6.70 Ci/mmol) were purchased from New

England Nuclear (Boston, MA). Follicle-Stimulating Hormone (FSH; equivalent to 50 mg

Armour standard) was purchased from Reheis Chemical Co. (Chicago, IL). The Bio-Rad protein assay dye reagent was purchased from Bio-Rad Laboratories (Richmond, CA).

Progesterone, estradiol and 20a-hydroxy-progesterone antisera were obtained from

Endocrine Sciences (Tarzana, CA). All other chemicals were purchased from Sigma

Chemical Co. (St. Louis, MO), Aldrich Chemical Co. (Milwaukee,WI) or Fisher Scientific

(Pittsburgh, PA).

PREPARATION OF CONDITIONED MEDIA

Preparation of the Thvmic Cell Culture Conditioned Medium (TCM)

Thymic cells were isolated using a slight modification of the techniques previously described (Uzumcii et al., 1992). Briefly, thymi were removed from 21-23 day-old female rats. The thymi were minced with scissors and the tissue fragments centrifuged (50 x g; for

1 min) five times in 40 ml Hank's Calcium-Magnesium Free Balanced Salt Solution (HCMF) in order to remove most of the lymphocytes. The remaining tissue fragments were digested with a mixture of trypsin (0.125%; w/v) and EDTA (0.01%; w/v) in HCMF at 37°C with stirring for 45 min. 20 x 106 thymic cells were plated in 75 cm2 culture flasks containing 81

Minimum Essential Medium D-Valine modification (MEM) supplemented with 10% (v/v)

FBS. After 24 hours in culture, the medium was replaced and unattached cells were removed. The attached cells were allowed to grow for three consecutive five-day culture periods. During the first five-day culture, the cells reached approximately 80% confluence.

Previously, on the 9th day of culture, using immunofluorescent staining for keratin, over 90% of the attached cells were shown to be epithelial in origin (Uziimcii et al., 1992). The media from the last two five-day periods were collected and used for evaluation. For the present study, four batches of TCM were prepared. After their steroidogenic activities were tested and confirmed, all batches were pooled and stored at -70°C as 4 ml aliquots until used.

Preparation of the Heart Cell Conditioned Medium (HCM)

In order to examine cell source specificity of factor(s) in TCM, heart cell conditioned medium was prepared. Hearts from 21- to 23-day-old female rats were aseptically removed.

The hearts were sliced and minced into small pieces with scissors. The tissue particles were enzymatically digested with 0.1% (w/v) collagenase type-TV in HCMF for 1 hour at 37 °C.

At 10 minute intervals, free single cells were removed and fresh buffer was added for volume replacement. After two washes, single cells were suspended in MEM supplemented with

10% (v/v) FBS and incubated the same way as the thymic cell culture. Then, HCM was evaluated in granulosa cell culture as a control medium. 82

Preparation of the Mock Extract (ME)

To alleviate the potential nonspecific effects of FBS components, ME was prepared in an identical fashion as TCM and HCM, but in cell-free culture dishes. ME was then evaluated in granulosa cell culture as a control medium.

Preparation of granulosa cell culture

Rat granulosa cells were isolated using the nonenzymatic needle puncture method previously described (Hsueh et al., 1980; Uzumcii et al., 1992). Briefly, ovaries were aseptically removed and placed in Dulbecco's Modified Eagle's Medium and Ham's Nutrient

Mixture F-12 (DME/F-12,1:1 mixture). Granulosa cells were freed from follicles (300-600 pm in diameter) by gentle and repeated puncture of the ovaries with a sterile bundle of beading needles. Isolated granulosa cells were collected and washed twice by centrifugation

(250 x g for 3 min, 4°C) in DME/F-12. The granulosa cells were plated at a density of 4 x

105 viable cells/well in 24-well tissue culture plates (Coming Inc, Oneonta, NY) containing

1 ml DME/F-12 medium supplemented with 5 pg/ml insulin, 5 pg/ml transferrin, 10 ng/ml selenium, 10 ng/ml epidermal growth factor (EGF), 100 U/ml penicillin, 100 pg/ml streptomycin sulfate and 250 pg/L amphotericin B (DME/F-12/S). The cells were first cultured for 24 hours at 37°C under a water-saturated, 95% air/5% C 02 atmosphere.

After 24 hours incubation under the above conditions, cultured granulosa cells were exposed to treatments (control, 10% (v/v) ME, HCM or TCM) in DME/F-12/S containing

100 nM androstenedione ±100 ng/ml FSH for 48 hours. 83

Dose-response study for FSH in cultured rat granulosa cells

To determine a "sub-maximal" stimulatory dose of FSH for progesterone and estradiol secretion in cultured rat granulosa cells, the cells were treated with various doses of FSH

(0-1000 ng/ml) for 48 hours. Following the treatment, media was used for steroid RIA and cells were used for protein determination.

Aromatase activity assay

Aromatase activities of granulosa cells were determined by the [3H]-H20 release method which was previously described by Gore-Langton et al. (1980) and subsequently performed with minor modifications in our laboratory (Akira et al., 1994). Briefly, following the removal of the conditioned media, granulosa cells were cultured for 6 hours (aromatization period) with fresh medium, which contained 2 pCi [1 P-3H] androstenedione in 2 pi ethanol for aromatase activity determination. At the end of the aromatization period, media were collected for aromatase activity measurements, and the cells were collected in 0.5 ml of 0.1

N NaOH for total cell protein measurements. The media were extracted with chloroform

(1:4; medium:chloroform). After centrifugation at 500 x g, 4°C for 15 min, two aliquots of supernatant were transferred to 12 x 75 mm tubes and incubated with 1% (w/v) dextran- coated activated charcoal for 20 min at 4CC. The suspension was centrifuged at 600 x g for

20 min. The supernatant was transferred to scintillation vials and mixed with scintillation cocktail. The disintegrations per minute (dpm) of released [3H]-H20 was determined by a

Beckman LS5801 scintillation counter. The protein content of the granulosa cells was determined by the dye binding method of Bradford (Bradford, 1976). Aromatase activity is 84 expressed as [lB-3H]-androstenedione consumed (pmole/mg protein).

Steroid radioimmunoassay

In unextracted aliquots of the culture medium which were collected from treated granulosa cells, 20a-hydroxy-progesterone, progesterone and estradiol concentrations were determined by RIA in a manner similar to that reported by Stouffer et al. (1976). The intra- and interassay coefficients of variation were 3.7% and 9.1%, 3.4% and 10.2%, and 2.9% and

11.0% for progesterone, estradiol-17p, and 20a-hydroxy-progesterone, respectively. The cross-reactivities of progesterone antiserum were 15% with pregnenolone, 0.5% with 20a- hydroxy-progesterone, less than 0.1% with estradiol-17 P, and testosterone. The cross­ reactivities of estradiol-17 p antiserum were 6.5% with estrone and less than 1% with estriol.

The cross-reactivity of 20a-hydroxy-progesterone antiserum was about 0.25% with progesterone. Since steroids in our assay were not separated, the values expressed in our results may be higher than the actual steroid concentrations. The cross-reactivities of the antisera were determined by using the 50% inhibition of binding method, as described by

Abraham (1969).

r3Hl-thvmidine incorporation assay

In a separate set of experiments, 1 x 105 granulosa cells were incubated under the same conditions mentioned above and treated with DME/F-12/S only, or 10% (v/v) ME, HCM, or TCM in DME/F-12/S for 48 hours. Upon removal of conditioned media, the cells were pulsed with 5 pCi of [3H]-thymidine in DME/F-12/S for 12, 24, or 48 hours for [3H]- 85 thymidine incorporation assay as previously described (Hu et al., 1993). In this experiment, the cell number was reduced to 1/4 of that used in previous experiments to ensure that cells did not reach confluence which might have caused cessation of cell division due to contact inhibition during incubation periods.

Data analysis

Data points are shown as the mean ± S.D. from quadruplicate culture wells in a representative experiment. Each experiment was repeated three times except [3H]-thymidine incorporation which was repeated four times. The differences between the effects of different treatments were evaluated by one-way ANOVA followed by Tukey's Studentized

Range Test (Snedecor & Cochran, 1980). SAS GLM procedures (SAS Institute Inc., 1985) were used for statistical analysis. A value of P<0.01 was considered as significant.

RESULTS

Dose-response study for FSH in cultured rat granulosa cells

FSH at the doses of 0-1000 ng/ml gave a typical dose-response curve. 100 ng/ml of

FSH was selected to be a "sub-maximal" stimulatory dose of FSH (Figure 12). At this dose,

FSH stimulated both progesterone and estradiol significantly. For the rest of the experiment,

100 ng/ml FSH was used for gonadotropin-stimulated steroid secretion. The rationale for using a "sub-maximal" stimulatory dose was to leave room for possible stimulation by the thymic factors. 86

Effect of TCM on basal and FSH-induced progesterone and 20cc-hvdroxy-progesterone secretion

The granulosa cells were exposed to 10% conditioned media in DME/F-12/S and

DME/F-12/S alone for 48 hours. In this experiment, TCM caused about an 80-fold increase in basal progesterone. FSH (100 ng/ml) caused about a 10-fold increase in progesterone production. TCM was able to further enhance secretion of this steroid, causing cells in TCM to produce 17 times more progesterone than the cells treated with FSH alone in unconditioned media (Figure 13). There were no significant differences among the other treatment groups. Stimulation of both basal and FSH-stimulated 20a-hydroxy-progesterone followed a similar trend as progesterone. TCM augmented basal 20a-hydroxy-progesterone approximately 40-fold, whereas it caused only about a 10-fold increase in FSH-induced 20a- hydroxy-progesterone secretion. Again, FSH had produced about a 7-fold increase in secretion of this steroid (Figure 14). Responses of both of these steroids to FSH in each tested media were statistically significant (P<0.01; not shown).

It should be recognized that in cells treated with TCM, the progesterone response to FSH is reduced from 10- to 2-fold, while the stimulation of 20a-hydroxy-progesterone secretion is similarly reduced from 7- to 1.5-fold. Furthermore, the ratio of 20a-hydroxy- progesterone:progesterone is reduced from 8:1 in cells treated with control media to approximately 4:1 in cells treated with 10% (v/v) TCM. 87

Effect of TCM on basal and FSH-induced estradiol secretion and aromatase enzvme activity

In the same set of experiments in which the effect of TCM on progestins was measured, the effects of TCM on estradiol and aromatase activity were also measured in the absence or presence of FSH (100 ng/ml). Under the stimulatory actions of TFs which are present in

TCM, rat granulosa cells were able to produce about 4 times more estradiol than cells treated with control media. Significant stimulatory action of TCM was also observed in the presence of 100 ng/ml FSH (about 2.5-fold). FSH alone was also able to produce significant stimulation (about 5-fold, Figure 15). The magnitude of stimulation by TCM was greater for basal aromatase activity than for basal estradiol production (4 vs. 15 times; estradiol vs. aromatase); however, its stimulatory effect on FSH-induced aromatase activity was of similar magnitude (about 3 times) as FSH-induced estradiol production (Figure 16).

Responses of both estradiol secretion and aromatase activity to FSH in each tested media were statistically significant (P<0.01; not shown).

Effect of TCM on total cell protein content

As depicted in Figure 17, the measurements of total cell protein contents show that there is no significant difference among the groups except for the cells in DME/F-12/S. The cell protein content in granulosa cells which were cultured in DME/F-12/S was significantly lower than that of other groups. 88

Effect of TCM on DNA synthesis of the granulosacells

To assess mitogenic activity of TFs which are present in TCM, [3H]-thymidine incorporation assay was used. The results show that while the mitotic activity of cells in

TCM did not differ significantly from that of cells exposed to DME/F-12/S, the rate of [3H]- thymidine incorporation was elevated in ME and HCM control media (Figure 18 and 19).

The cells, following their treatments, were divided into 3 groups and pulsed with [3H]- thymidine in DME/F-12/S for a time course to test the possibility of cessation of cell division due to contact inhibition during culture periods. A continuous increase in incorporation of

[3H]-thymidine was observed over the culture periods. However, it was evident that the rate of [3H]-thymidine incorporation was decreased over the second 24 hour period relative to that over the first 24 hour period (Figure 18 and 19).

DISCUSSION

In this study, primary cultures of immature, DES-treated rat granulosa cells were used to further explore the steroidogenic action of TFs present in TCM. Earlier, we reported

(Uziimcii et al., 1992) that proteinaceous factor(s) present in TCM can stimulate basal progesterone and estradiol production from cultured granulosa cells, the stimulatory action on estradiol being less prominent. The present study clearly confirms our previous report.

In addition, the concomitant elevations of estradiol secretion and aromatase activity by TCM demonstrated in this experiment clearly implicate the presence of factors in TCM that are stimulatory for aromatase activity of rat granulosa cells in vitro. The observed discrepancy between the magnitude of stimulation of aromatase activity and of estradiol secretion by 89

TCM could be a result of the fact that the determination of estradiol concentration is a compound measure of both aromatase and 17P-hydroxysteroid dehydrogenase (17P-HSD) activities and/or that a portion of the estrogens exist in the form of estrone, the concentration of which was not determined by RIA (except for that detected due to the 6.5% cross­ reactivity of the estradiol antibody). Moreover, the differences between the magnitudes of stimulation of progestin and estradiol by TCM may be due to the presence of more than one steroidogenic factor which are distinct in their actions on steroidogenic enzymes. The validity of this suggestion is supported by preliminary purification studies of TCM from this laboratory.

It is well-known that in the granulosa cell, secretion of progesterone can be modulated by changes in conversion of this steroid into its major metabolite, 20a-hydroxy-progesterone

(Hsueh et al., 1984). Measurement of 20a-hydroxy-progesterone in culture medium showed that its production by 20a-hydroxysteroid dehydrogenase (20a-HSD) was significantly elevated by TCM as compared to those by control media and the pattern of elevation parallels that of progesterone. In addition, the changes in the progesterone:20a-hydroxy-progesterone ratios outlined in the Results section suggest that TCM, in the presence of FSH, reduces 20a-

HSD activity by about 50% and/or that the elevation of progesterone production by TCM exceeds the capacity of this enzyme. However, the concomitant elevations of progesterone and 20a-hydroxy-progesterone secretions eliminate the possibility that the increase in progesterone after TCM treatment is due to an inhibition of metabolism and support the notion that TCM stimulates enzymes that are involved in the production of progesterone (i.e. cytochrome P-450 cholesterol side chain cleavage (CSCC) and/or 3 P-hydroxysteroid 90 dehydrogenase (3 p-HSD).

The unknown cell growth stimulatory and mitogenic potentials of TFs within TCM present the possibility that their stimulatory actions on granulosa cell steroidogenesis may reflect nonspecific mitotic activities of the cells. Measurements of cell protein and DNA synthesis, however, do not support this suggestion. The mitotic activity of cells treated with

TCM did not differ significantly from that of the cells treated with DME/F-12/S, which were lower than that of the cells exposed to ME and HCM. However, the protein content of TCM- treated cells was similar to those of cells in ME and HCM which were significantly higher than that of the cells in DME/F-12/S. Since ME and HCM stimulate both total cell protein content and DNA synthesis, it seems that the former may be due to an increase in the rate of cellular proliferation induced by the FBS present in these conditioned media and/or due to remaining conditioned media proteins that could not be washed off from the surface of the cells before determination of protein content. However, TCM increases total cell protein content without affecting cell proliferation which suggests that TCM may contain TFs that have the ability to stimulate growth while negating proliferative activity of serum- supplemented MEM. Nevertheless, since all steroidogenic activity measurements were standardized per mg protein, it is most unlikely that this increase in cell growth is responsible for the elevated steroidogenesis in granulosa cells exposed to TCM. In addition, while the persistent rise in the incorporation of [3H]-thymidine shows that cell division continued for the entire culture period, the evident decrease in the rate of the incorporation over the second

24 hours as compared to the first 24 hours may indicate the start of contact inhibition.

Taken together, these observations suggest that the elevation of steroid production in TCM- 91 treated cells is due to direct and specific steroidogenic actions of its active factors rather than elevated cell growth and/or proliferation.

This study also shows that TFs in TCM can stimulate FSH-induced, as well as basal, steroid production. Since, under in vivo conditions, granulosa cells function in the presence of gonadotropins, these results attribute to TCM potential physiologic relevance. In contrast,

Gorospe et al. (1993) reported that thymosin fraction 5 (TF-5) stimulates basal, but inhibits

FSH-induced progesterone and estradiol production. These findings suggest that the stimulatory action of TCM cannot be solely due to TF-5, if it is present in TCM. Of course, these differing results could be due to differences in the experimental conditions, such as differences in the magnitude of stimulation by FSH. In addition, it is noteworthy that the same study by Gorospe et al. (1993) reported that TF-5 caused stimulation of IL-6, a cytokine with a reported inhibitory effect on FSH-induced progesterone production in rat granulosa cells (Gorospe et al., 1992), which has been suggested to be a common mediator through which various cytokines, hormones, and gonadal regulators exert their modulatory effect on ovarian steroidogenesis (Gorospe and Spangelo, 1993). In light of these reports, the role of

IL-6 in the action of TCM on rat granulosa cells would be a useful direction for future studies. Although it has previously been demonstrated that more than 90% of the cells in thymic culture are epithelial in origin, the exact composition of the remaining portion is not known (Uziimcii et al., 1992). In addition to fibroblasts, lymphocytes and macrophages may be present which can provide a potential source of cytokines that can affect TCM's actions on granulosa cells. 92

The reduction in the size of the thymus gland after puberty (Goldstein and Mackay,

1969) and the use of thymi from prepubertal animals in the current study raise questions about the role of thymus in ovarian regulation and the physiological relevance of the results reported here. It is unknown whether the steroidogenic TFs that are present in TCM are produced by the thymus of sexually mature animals and, if they are, whether they have a similar action on granulosa cell functions in mature animals. Although these questions have not yet been answered, the presence and possible roles of various other thymic factors in mature animals (Fabris and Mocchegiani, 1985; Wolfe, et al., 1989) and humans (Lewis et al., 1978) have been reported. For example, in a study using bovine females, Wolfe et al.

(1989) characterized secretion of thymosin-al and p4 during different stages of ovarian function (prepuberty, estrus with conception, estrus without conception and pregnancy) and reported that levels of both of these thymic hormones were higher in the group showing estrus during which conception occurred than in the other groups, suggesting an association of thymic hormones in successful reproductive function. More recently, Crawford et al.

(1992) suggested that there may be a possible interrelation between thymic and reproductive aging. Furthermore, Oikawa et al. (1990) were able to detect a high level of prothymosin-a mRNA in the rat ovary. Similarly, Hall et al. (1991a, b) reported that thymosin-P4 and P 10, two other thymic hormones, were expressed and biosynthesized in both immature and adult rat ovaries, and, more interestingly, that PMSG, hCG and PGF2(( modulated the differential expression of these hormones. Cautious interpretation of these results suggest that thymic hormones originating from either the thymus gland or even the ovary itself may be capable of affecting ovarian function in both immature and mature animals. Whether the TFs present 93 in TCM, which stimulates steroidogenesis in rat granulosa cells in vitro, are produced by the thymus gland and similarly affect granulosa cell function in mature animals is worth investigation. Experimental results from our laboratory suggest that TCM from immature rats stimulate steroidogenesis in rat granulosa cells from sexually mature animals. To derive more definitive conclusions of the physiologic role of the TFs, further studies using more purified preparations are underway in our laboratory.

In any event, our data provide evidence for the existence of factor(s) in TCM which stimulate basal and gonadotropin-induced progesterone, 20a-hydroxy-progesterone and estradiol secretions in cultured granulosa cells of DES-treated immature rats, and this stimulation is due, at least in part, to an increase in the activity of the following steroidogenic enzymes: (i) cytochrome p450cscc and/or 3P-HSD (reflected by an increase in progesterone);

(ii) 20a-HSD (reflected by an elevation of 20a-hydroxy-progesterone and an increase in the progesterone:20a-hydroxy-progesterone ratio); and (iii) aromatase and/or 17P-HSD

(measured as the conversion of either "cold" or tritiated androstenedione to estradiol and/or estrone). Finally, our results provide further evidence for the possible physiological role of thymic factor(s) in ovarian granulosa cell function.

SUMMARX

Recently, numerous reports heve demonstrated the importance of the thymus gland in reproductive physiology. Previously, we have reported that thymic factors (TFs) which are present in thymic cell culture-conditioned medium (TCM) could stimulate basal progesterone and estradiol production from cultured rat granulosa cells. The current study attempts to 94 characterize the stimulatory actions of TFs on both basal and FSH-induced steroidogenesis.

Thymic epithelial cells from immature female rats were isolated and used for production of

TCM. Granulosa cells were obtained from immature diethylstilbestrol (DES)-treated rats.

TFs stimulated both basal and FSH-induced progesterone secretions 80 and 17 times, respectively, as compared to the control media. The effects of TFs on basal and FSH- induced 20a-hydroxy-progesterone secretion were comparable to those on progesterone production (40times and 10 times, respectively). In addition, TCM stimulated basal and

FSH-induced estradiol secretion approximately 4 and 2.5 times, respectively, as compared to control. Stimulation of aromatase enzyme activity followed a similar trend as estradiol secretion, and TCM stimulated basal and FSH-stimulated aromatase enzyme activity approximately 15 and 3 times, respectively as compared to control. Thus, these results indicate that the observed increases in progesterone and estradiol secretions in TCM-treated rat granulosa cells are likely to be due to elevated activities of specific steroidogenic enzymes. Measurements of total cell protein and DNA synthesis indicate that enhanced steroidogenesis in TCM-treated cells is not due to increased cell growth and/or proliferation.

Rather, the enhanced steroidogenesis is probably due to an increased steroid biosynthetic capability of the cells. In addition, not only can TCM stimulate production of these steroid hormones in the absence of FSH, but also in the presence of FSH, suggesting physiological relevance of its stimulatory actions. Overall, these results show that thymic epithelial cells are a potential source of steroidogenic factors. Moreover, the thymic factor(s) may have a physiologic role in ovarian function. FIGURES

95 Figure 12. Effect of various doses of FSH on progesterone and estradiol secretions in cultured rat granulosa cells. After initial 24 hours incubation time, the cells were incubated with FSH (0-1000 ng/ml) for 48 hours. Data points are shown as the mean ± S.D. of quadruplicate culture wells from a representative experiment. The experiment was repeated three times under similar experimental conditions and identical results were obtained. Different letters above the same type of symbol show statistical difference (P<0.01).

96 Progesterone (ng/mg cell protein) 0 - 400 800 200 0 - 600 0 Mean D. .D S ± n a e M = T 4 = N 0 0 0 20 0 1000 500 200 100 50 20 ng/ l) /m g (n H S F Figure12, 40 120 100 CTQ d " S C/3 —* )— o 1— - 0 P t P tn Q o n 3* 97 A A Figure 13. The effects of different conditioned media on basal and FSH-induced progesterone secretion in cultured rat granulosa cells. Different letters above the bars with the same shading signify statistical difference (PcO.Ol). Data points are shown as the mean ± S.D. of quadruplicate culture wells from a representative experiment. The experiment was repeated three times under similar experimental conditions and identical results were obtained. (Control : DME/F-12/S alone; ME: Mock extract; HCM : Heart cell culture- conditioned media; TCM : Thymic cell culture-conditioned media).

98 Progesterone (ng/mg protein) 1000 0 - 800 400 - 600 100 0 - 50 - 75 5 - 25 0

£ r ME HCM TCM C T M C H E M l tro n o C = Mean D. .D S ± n a e M = T 4 = N 100 ng/ l) /m g n 0 0 (1 H S F H S F o N Figure13. 10% conditioned media conditioned 10% m B 99 Figure 14. The effects of different conditioned media on basal and FSH-induced 20a- hydroxy-progesterone secretion in cultured rat granulosa cells. Data points are shown as the mean ± S.D. of quadruplicate culture wells from a representative experiment. The experiment was repeated three times under similar experimental conditions and identical results were obtained. Different letters above the bars with the same shading signify statistical difference (P<0.01).

100 20 a -OH-progesterone (ng/mg protein) 2000 - 2500 -r- 3500 00 - 3000 1500 1500 7 - 270 - 360 8 - 180 0 - 90 — 0 - V

r l tro n o C N = 4 = N = ± D. .D S ± n a e M = T 100 ml) /m g n 0 0 (1 H S F H S F o N Figure14. M C T M C H E M 10% conditioned media conditioned 10% 101 Figure 15. The effects of different conditioned media on basal and FSH-induced estradiol secretion in cultured rat granulosa cells. Data points are shown as the mean ± S.D. of quadruplicate culture wells from a representative experiment. The experiment was repeated three times under similar experimental conditions and identical results were obtained. Different letters above the bars with the same shading signify statistical difference (PcO.Ol).

102 Estradiol (ng/mg protein) 18 12 6 0 - - rol E M l o tr n o C N = 4 = N Mean D. .D S ± n a e M = T B 100 ml) /m g n 0 0 (1 H S F H S F o N Figure15. toned ia d e m d e n itio d n o c % 0 1 B M C T M C H a 103 Figure 16. The effects of different conditioned media on basal and FSH-induced aromatase enzyme activity, as measured with the [3H]-H20 release method, in cultured rat granulosa cells. Data points are shown as the mean ± S.D. of quadruplicate culture wells from a representative experiment. The experiment was repeated three times under similar experimental conditions and identical results were obtained. Different letters above the bars with the same shading signify statistical difference (P<0.01).

104 [16-3H]-Androstenedione consumed (pmole/mg protein) 21 8 - 28 35 4 - 14 0 - 7 - r ME HCM TCM C T M C H E M l tro n o C N = 4 = N = Mean D. .D S ± n a e M = T 100 ml) /m g n 0 0 (1 H S F H S F o N Figure16. toned ia d e m d e n itio d n o c % 0 1 B m a 105 Figure 17. The effects of different conditioned media on total cell protein content in cultured rat granulosa cells in the absence (No FSH) and presence (FSH) of 100 ng/ml FSH. Data points are shown as the mean ± S.D. of quadruplicate culture wells from a representative experiment. The experiment was repeated three times under similar experimental conditions and identical results were obtained. Different letters indicate statistically significant differences within a group of bars (P<0.01).

106 Total cell protein (jig/well) 5 - 45 75 0 - 30 0 - 60 5 - 15 = Mean D. .D S ± n a e M = T 4 = N C o n tro l l tro n o C ME E M % 0 1 FSH S F o N Figure17. 100 ml) /m g n 0 0 (1 M C H % 0 1 M C T % 0 1 H S F 107 Figure 18. The effects of different conditioned media on DNA synthesis in cultured rat granulosa cells, in the absence of FSH. Data points are shown as the mean ± S.D. of quadruplicate culture wells from a representative experiment. The experiment was repeated three times under similar experimental conditions and identical results were obtained. The same letters above bars within a group indicates no statistically significant differences.

108 C o n tro l 1 0 % H C M 1 0 % M E 1 0 % T C M N = 4 a T = M ean ± S.D. J

12 2 4 4 8

Hours after conditioned media withdrawal

Figure 18. Figure 19. The effects of different conditioned media on DNA synthesis in cultured rat granulosa cells, in the presence of 100 ng/ml FSH. Data points are shown as the mean ± S.D. of quadruplicate culture wells from a representative experiment. The experiment was repeated three times under similar experimental conditions and identical results were obtained. The same letters above bars within a group indicates no statistically significant differences.

110 [ H]-Thymidine incorporated (d.p.m. x 103/well) 8 - 28 21 35 4 - 14 - 7 0 Mean S. . .D S ± n a e M = T 4 = N er toned a t l a w a r d ith w ia d e m d e n itio d n o c r te f a s r u o H C o n tro l l tro n o C E M % 0 1 2 8 4 4 2 12 Figure19. M C H % 0 1 M C T % 0 1 111 CHAPTER III

TWO DISTINCT STEROID-MODULATING FACTORS WERE IDENTIFIED

IN RAT THYMIC EPITHELIAL CELLS CULTURE-CONDITIONED MEDIA

(TCM)

INTRODUCTION

The role of the thymus gland on ovarian function has long been suggested (Hall et al.,

1992), however, the exact details of this role remain to be discovered. Through numerous experimental findings, it is obvious that normal development and function of ovaries depend upon the presence of an intact thymus gland. For example, Nishizuka and Sakakura (1969) reported that neonatal thymectomy of mice results in developmental arrest of the ovary; and caused sterility in the female, but not in the male. Later, Besedovsky and Sorkin (1974) reported that female thymectomized mice exhibit a delay in vaginal opening and in the first estrus. In addition, these authors reported that in these mice the ovaries and uterus are very small and display the histological pattern of prepubertal animals. Extensive morphological and histological studies in neonatally thymectomized mice (Nishizuka and Sakakura, 1969, and 1971; Sakakura et al., 1979; Michael et al., 1980; Marmor and Michael, 1984) revealed that, while the ovaries of these mice are similar to those of their littermates at birth, a progressive ovarian dysgenesis starts at the age of 20 days and functional apparatus of

112 113 ovaries (i.e. follicles and corpora lutea) are depleted and replaced with hypertrophic interstitial tissue around the age of 120 days. Similarly, ovarian dysgenesis in thymectomized rats (Lintem-Moore, 1977 and Hattori and Brandon, 1979) and primates

(Healy et al., 1985) has also been reported. In addition, endocrine imbalances along with ovarian dysgenesis have been found in thymectomized animals. For example, a high level of serum androgens was reported in thymectomized mice which was probably due to hypertrophic interstitial cells (Nishizuka et al., 1973; Marmor and Michael, 1984).

Progesterone and estrogen levels were lower in thymectomized mice than in control mice

(Michael et al., 1981). In contrast, in thymectomized rats, total progesterone level was lower than that of sham operated rats, while the levels of estradiol were not different (Hattori and

Brandon, 1979). On the other hand, circulating FSH and LH levels of thymectomized mice were found to be lower than those of control animals between days 10-11 and 9-20 days, respectively (Michael et al., 1981). However, in thymectomized rhesus monkeys, circulating

FSH and LH levels were found to be higher than in controls on day 2 of age (Healy et al.,

1985). Interestingly, it was shown that the morphological and functional disorders could be restored to normal by grafting thymic tissues in thymectomized mice (Nishizuka and

Sakakura, 1969; Sakakura and Nishizuka, 1972; Besedovsky and Sorkin, 1974).

Previously, our laboratory has shown that conditioned media from cultured thymic epithelial cells (TCM) stimulates basal progesterone and estradiol production in cultured rat granulosa cells (Uziimcu et al., 1992). However, the stimulatory action of TCM could be mimicked by neither of two known thymic hormones (i.e. thymulin and thymosin a 1) nor conditioned medium from heart cell culture. Later, concomitant elevations in basal and FSH- 114 stimulated 20a-hydroxy-progesterone and progesterone; and estradiol-176 and [3H]-H20 release from [16-3H]-androstenedione, another indication aromatase enzyme activity by

TCM, were observed. In the latter study, elevated steroidogenesis was shown to not be due to cell growth or proliferation (Uziimcii and Lin, 1994). In the present study, TCM was fractionated by gel filtration chromatography and the fractions were examined for their effect on cultured granulosa cell steroidogenesis and morphology. Subsequent to detection of two biologically active fraction ranges which correspond to approximately 1,000- and 22,000~Mr, the interactive stimulatory actions of these two fractions on steroidogenesis in cultured rat granulosa cells were investigated. In addition, dose-dependent modulatory actions of these two factors on basal and FSH-induced progesterone and estradiol secretions were evaluated.

MATERIAL AND METHODS

Animals:

For the thymic cell culture, 21- to 23-day-old Sprague-Dawley female rats were used.

After animals were sacrificed, thymi were aseptically removed for cell culture preparation.

For the preparation of the granulosa cell culture, immature Sprague-Dawley female rats were used. When they reached 21-23 days of age, the female rats were injected with 250 pg

DES/day suspended in 100 pi com oil s.c. for five days (5 mg DES/kg body weigth/day).

Eight hours after the last DES injection, the rats were sacrificed and ovaries were removed for granulosa cell preparations. 115

Chemicals and reagents:

Fetal Bovine Seram (FBS) was obtained from Hyclone Laboratories, Inc. (Logan, UT).

[l,2,5,7,21-3H(N)]-progesterone (109.50 Ci/mmol), [2,4,6,7-3H(N)]-estradiol (100.50

Ci/mmol), 20a[l,2-3H(N)]-hydroxy-progesterone (51.20 Ci/mmol), [1 p-3H]-androstenedione

(24.50 Ci/mmol) and [methyl3H]-thymidine (6.70 Ci/mmol) were purchased from New

England Nuclear (Boston, MA). Follicle-Stimulating Hormone (FSH; equivalent to 50 mg

Armour standard) was purchased from Reheis Chemical Co. (Chicago, IL). The Bio-Rad protein assay dye reagent was purchased from Bio-Rad Laboratories (Richmond, CA).

Progesterone, estradiol and 20a-hydroxy-progesterone antisera were obtained from

Endocrine Sciences (Tarzana, CA). All other chemicals were purchased from Sigma

Chemical Co. (St. Louis, MO), Aldrich Chemical Co. (Milwaukee,WI) or Fisher Scientific

(Pittsburgh, PA).

Preparation of the Thymic Cell Culture-Conditioned Medium (TCM)

Thymic cells were isolated using a slight modification of the techniques previously described (Uziimcii et al. 1992). Briefly, thymi were removed from 21-23 day-old female rats. The thymi were minced with scissors and the tissue fragments centrifuged (50 x g; for

1 min) five times in 40 ml Hank's Calcium-Magnesium Free Balanced Salt Solution (HCMF) in order to remove most of the lymphocytes. The remaining tissue fragments were digested with a mixture of trypsin (0.125%) and EDTA (0.01%) in HCMF at 37 °C with stirring for

45 min. 20 x 106 thymic cells were plated in 75 cm2 culture flasks containing Minimum

Essential Medium D-Valine modification (MEM) supplemented with 10% FBS. 116

Replacement of L-valine with it D-form is to suppress the growth of fibroblasts which lack the machinery to utilize D- forms of amino acids. After 24 hours in culture, the medium was replaced and unattached cells were removed. The attached cells were allowed to grow for three consecutive five-day cultures. During the first five-day culture, the cells reached approximately 80% confluence. Previously, on the 9th day of culture, using immunofluorescent staining for keratin, over 90% of the attached cells shown to be epithelial in origin (Uzumcii et al., 1992). The media from the last two five-day culture were collected, centrifuged and stored at -70°C until used for gel filtration chromatography.

Lvophilization of TCM

Since each batch of TCM was stored at -70°C as 4 ml aliquots, these aliquots were combined and divided into 1 ml aliquots again. These 1 ml aliquots were frozen at -70°C and lyophilized. For lyophilization, a Freeze Dryer, (FTS Systems, Inc., Stone Ridge NY) was used. The lyophilization process was carried out at about -80°C and 25 militorr of pressure for overnight. When the TCM appeared to be converted to powder form, the lyophilization process was terminated and the dried TCM was stored at -20 °C until use.

Gel Filtration Chromatography of TCM

Lyophilized TCM was reconstituted in 25% of its original volume. Samples were clarified by centrifugation, filtered using 0.45 pm membranes, and aliquots of up to 1 ml were applied to TSK-Gel G2000 SW gel-filtration column (0.75 x 30 cm; TosoHaas, PA) equipped with a TSK-GSW guard column (0.8 x 4 cm; TosoHaas, PA). A Fast Protein 117

Liquid Chromatography (FPLC; Pharmacia, NJ) system was used to maintain the flow rate at 0.5 ml/min during elution of the proteins with PBS containing 0.5 M NaCl. The column eluate was monitored at 280 nm using a in-line UV detector and collected into 200 pi fractions. Fractions were stored at -70° C prior to assay. The molecular weights of the active factors were calculated by reference to the elution position of ovalbumin (45 kDa), trypsin inhibitor (20.1 kDa), lactalbumin (14.2 kDa), and epidermal growth factor (6 kDa).

Preparation of granulosa cell culture

Rat granulosa cells were isolated using the nonenzymatic needle puncture method previously described (Uziimcu et al., 1992). Briefly, ovaries were aseptically removed, and placed in Dulbecco's Modified Eagle's Medium and Ham's Nutrient Mixture F-12 (DME/F-

12, 1:1 mixture). Granulosa cells were freed from follicles (300-600 pm in diameter) by gentle and repeated puncture of the ovaries with a sterile bundle of beading needles. Isolated granulosa cells were collected and washed twice by centrifugation (250 x g for 3 min, 4°C) in DME/F-12. The granulosa cells were plated at a density of 4 x 105 viable cells/well in 24- well tissue culture plates (Coming Inc, Oneonta, NY) containing 1 ml DME/F-12 medium supplemented with 5 pg/ml insulin, 5 pg/ml transferrin, 10 ng/ml selenium, 10 ng/ml epidermal growth factor (EGF), 100 U/ml penicillin, 100 pg/ml streptomycin sulfate and 250 pg/L amphotericin B (DME/F-12/S). The cells were first cultured for 24 hours, 37°C under a water- saturated atmosphere, and gassed with 95% air and 5% C 0 2. 118

Evaluation of steroidogenic activities of the gel filtration fractions

After 24 hour incubation under the above conditions, cultured granulosa cells were exposed to 200 pi gel filtration fractions in 800 pi DME/F-12/S for 48 hours. Culture medium was then collected for steroid hormone RIA and cells were incubated with 2pCi [16-

3H]-androstenedione for 6 hours for aromatase activity assay.

Effect of gel filtration fraction of TCM on morphology of cultured rat granulosa cells

During exposure of rat granulosa cells to gel filtration fractions, morphologies of the cells were examined with phase-contrast microscopy. Any observation of morphological changes were recorded with micrographs.

Interaction of 1 and 22 kDa fractions on steroidogenesis of cultured rat granulosa cells

For the "combination assay", granulosa cells were exposed to 45 pi of 1 and 22 kDa fractions in combination or alone, or 90 pi of each alone in DME/F-12/S for 48 hours. When

45 pi of fraction volume was used, it was brought to 90 pi by adding 45 pi column buffer.

Dose-response study for FSH in cultured rat granulosa cells

To determine "sub-maximal" stimulatory dose of FSH for progesterone and estradiol secretions in cultured rat granulosa cells, the cells were treated with 0-2000 ng/ml FSH for

48 hours following initial 24 hour incubation in DME/F-12/S 119

Effects of 1 and 22 kDa factors on basal and FSH-induced progesterone and estradiol secretion in cultured rat granulosa cells

To determine the effect of 1 and 22 kDa fractions on FSH-induced steroidogenesis, the contents of 11 fraction tubes from each of the 1 and 22 Mr regions were combined and the cells in 820 pi DME/F-12/S ± 100 ng/ml FSH (as indicated) treated with either the 1 or 22 kDa factor at the doses of 10-180 |il. When a treatment dose was smaller than 180 jul, the volume was brought to 180 |il with column buffer. Following 48 hour treatment periods, media were collected and stored at -70°C for steroid radioimmunoassay (RIA). The protein content of the granulosa cells was determined by the dye binding method of Bradford

(Bradford, 1976)

Aromatase activity measurements using r3Hl-H?Q release method;

Aromatase activities of granulosa cells were determined by the [3H]-HzO release method which was previously described by Gore-Langton et al. (1980) and subsequently performed with minor modifications in our laboratory (Akira et al., 1994). Briefly, following their treatment, granulosa cells were cultured for 6 hours (aromatization period) with fresh medium, which contained 2 pCi [lp-3H] androstenedione in 2 pi ethanol for aromatase activity determination. At the end of the aromatization period, media were collected for aromatase activity measurements, and the cells were collected in 0.5 ml of 0.1

N NaOH for total cell protein measurements. The media were extracted with chloroform

(1:4; medium:chloroform). After centrifugation at 500 x g, 4°C for 15 min, two aliquots of supernatant were transferred to 12 x 75 mm tubes and incubated with 1% (w/v) dextran- 120 coated activated charcoal for 20 min at 4°C. The suspension was centrifuged at 600 x g for

20 min. The supernatant was transferred to scintillation vials and mixed with scintillation cocktail. The disintegrations per minute (dpm) of released [3H]-H20 was determined by a

Beckman LS5801 scintillation counter. The protein content of the granulosa cells was determined by the dye binding method of Bradford (Bradford, 1976). Aromatase activity is expressed as [lB-3H]-androstenedione consumed (pmole/mg protein).

Steroid hormone radioimmunoassay

In unextracted aliquots of the culture medium which were collected from treated granulosa cells, progesterone, 20a-hydroxy-progesterone and estradiol concentrations were determined by RIA in a manner similar to that reported by Stouffer et al. (1976). The intra- and interassay coefficients of variation were 3.7% and 9.1%, and 3.4% and 10.2% progesterone and estradiol-17 p, respectively. The cross-reactivities of progesterone antiserum were 15% with pregnenolone, 0.5% with 20a-hydroxy-progesterone, less than

0.1% with estradiol-17p, and testosterone. The cross-reactivities of estradiol-17P antiserum were 4% with estrone and less than 0.1% with estriol. The cross- reactivity of 20a-hydroxy- progesterone antiserum was about 0.25% with progesterone. The cross-reactivities of the antisera were determined by using the 50% inhibition of binding method of Abraham (1969).

Data analysis

For "activity evaluation assay", data points are single points from one representative experiment. This experiment was repeated at least 5 times with similar results. For the 121

"combination assay" and dose-response study, data points are shown as the mean ± S.D. from three experiments. Means for the "combination assay" and dose-response study were compared by one-way ANOVA followed by Tukey's Studentized Range Test (Snedecor &

Cochran 1980). SAS GLM procedures (SAS Institute Inc., 1985) were used for statistical analysis. A value of P<0.01 was considered as significant.

RESULTS

Evaluation of the steroidogenic activities of gel filtration fractions in cultured rat granulosa cells

Protein profile of gel filtration fractions as measured at 280 nm, and the effects of these fractions on progesterone and estradiol secretion in cultured rat granulosa cells are depicted in Figure 20. Progesterone secretion was stimulated significantly by two separate groups of fractions which were eluted from column in 22,000- (22 kDa factor) and 1,000-M,

(1 kDa factor) regions. The magnitude of stimulation of the progesterone secretion caused by the 22 kDa and 1 kDa factors were about 13- and 21-fold of the control. On the other hand, estradiol was only stimulated by the 1 kDa factor and the magnitude of stimulation was about 8-fold. In the culture medium, 20a-hydroxy-progesterone was also measured along with progesterone. The stimulatory pattern of 20a-hydroxy-progesterone, mimicked that of progesterone (Figure 21). In addition to estradiol secretion, aromatase activity was measured by the method of [3H]-H20 release (Osawa and Spaeth, 1971). Measurement of

[3H]-H20 release demonstrated a pattern to that produced by measurements of estradiol secretion. (Figure 21). 122

Effect of TCM gel filtration fractions on morphology of rat granulosa cells:

During the culture period, the morphology of the rat granulosa cells were monitored by phase-contrast microscopy. The 1 kDa factor induced drastic morphological changes in the rat granulosa cells which closely resembled the typical changes (retraction) that were seen with the FSH-treated cells (Figure 22c, C, d and D). The granulosa cells responded to both

FSH (500 ng/ml) and 1 kDa fractions (200 pi) by becoming rounded and leaving finger-like processes attached to the substratum. On the other hand, the morphological appearance of the cells that were treated with the 22 kDa factor did not differ from that of the cells in control medium. (Figure 22a, A, b and B). The cells appear to be well-spread and polygonal. The changes that were induced by both FSH and the 1 kDa factor start to appear

20 min after treatment and reached a maximal level at 1 hr. At 12 hr, the cells started to regain their normal shape and returned to almost normal at 48 hr (not shown).

Interactive actions of 1 and 22 kDa factors on steroidogenesis in cultured rat granulosa cells

In this experiment, each fraction from the region of 1,000 and 22,000 Mr were either used alone (45 pi) or in combination (45 pi each). When the 1 and 22 kDa factors were combined, they caused significantly more progesterone secretion than the summation of progesterone secretion that was caused by each factor alone. The fractions that caused synergistic stimulation were 53-56 and 86-89. To examine the possible presence of a self- synergistic action within each region, the treatment value of fractions from these two regions was increased to 90 pi. The stimulatory actions of the factors alone on progesterone 123 secretion, even at the 90 pi level, remained significantly lower than these observed for the

45 pi fractions (Figure 23A and B).

However, for estradiol secretion, only the 1 kDa factor was stimulatory. The 22 kDa had no effect on unstimulated estradiol secretion. In addition, the stimulatory action of the

1 kDa was antagonized when the 22 kDa factor was present in the cell culture together.

(Figure 24A and B).

Dose response studies for FSH in cultured rat granulosa cells

FSH between the doses of 0-2000 ng/ml was able to stimulate progesterone and estradiol in a dose-dependent manner (Figure 25). 100 ng/ml was used as a submaximal dose of FSH in the dose-response study of the 1 and 22 kDa factors.

Effects of 1 and 22 kDa factors on basal and FSH-induced progesterone and estradiol secretion in cultured rat granulosa cells

Both 1 and 22 kDa factors stimulated basal and FSH-induced progesterone secretion in a dose-dependent and very similar manner, except at the level of 180 pi in the presence of FSH. The stimulatory action of the 1 kDa factor on basal progesterone was higher than that of 22 kDa at 180 pi ( Figure 26A and B). However, while the 1 kDa factor was able to stimulate basal estradiol in a dose-dependent manner, the 22 kDa had no effect on this steroid's secretion. On the contrary, the 22 kDa factor was inhibitory for estradiol secretion in the presence of 100 ng/ml FSH, whereas the 1 kDa factor stimulated FSH-induced estradiol in a dose dependent manner (Figure 27A and B) 124

DISCUSSION

It has been reported from this laboratory that rat thymic epithelial cell culture- conditioned medium (TCM) contains factors that are stimulatory for progesterone, 20a-OH- progesterone, and to a lesser extent estradiol secretion. (Uziimcii et al., 1992; Uziimcu and

Lin, 1994). In the present study, gel filtration chromatographic fractionation of TCM revealed that the steroidogenic factors lie in the regions that correspond to 1,000- and

22,000-Mr. While 1 kDa factor stimulates progesterone and estradiol secretions and cause morphological alterations of cultured rat granulosa cells, the 22 kDa factor only stimulates progesterone with no effect on basal estradiol. In addition, although these two factors are synergistic stimulators of progesterone secretion, the 22 kDa factor antagonizes the stimulatory action of the 1 kDa factor on estradiol. Furthermore, the 1 kDa factor stimulated not only basal, but also FSH-induced progesterone and estradiol secretion, while the 22 kDa factor mimicked the stimulatory action of the 1 kDa factor for progesterone, it inhibited FSH- stimulated estradiol production and had no effect on basal production. The reason for the difference in the actions of the 1 kDa and 22 kDa factors is currently unknown. Previously we did not find a parallel increases in the intracellular cAMP concentration when estradiol and progesterone production were elevated by unfractionated TCM in cultured rat granulosa cell (Uzumcu et al., 1992). However, intracellular concentrations of cAMP was not determined in the either the 1 kDa or 22 kDa treated cells. Therefore, whether cAMP or other second messengers are involved in the stimulation of steroidogenesis by 1 kDa or 22 kDa factor is not known. Although, the involvement of cAMP in the stimulation of steroidogenesis in the 1 kDa or 22 kDa-treated cells is not known, since their effect on the 125 granulosa cell is different, it is possible that their mechanisms of action are alsodifferent.

Since the 1 kDa and 22 kDa factors may be novel, their stimulatory potency on steroidogenesis in cultured rat granulosa cell is difficult to comprehend without having a known reference point. Therefore, the stimulatory potency of 10% unfractionated TCM, and the 22 kDa and the 1 kDa factors as compared to that of FSH, on progesterone and estradiol secretion in cultured rat granulosa cell are summarized in Table 1.

Table 1. Comparative potencies of unfractionated and fractionated TCM on progesterone and estradiol secretion in cultured rat granulosa cells. In this comparison, FSH-induced steroid production was used as a reference. Note that unfractionated TCM is less potent for stimulating estradiol secretion than the 1 kDa factor. This is probably due to the observed antagonistic effect of the 22 kDa factor on 1 kDa-induced estradiol secretion. The underlined numbers show the dose of FSH (ng/ml) which is as potent as the corresponding form of TCM for the stimulation of respective steroid hormone secretion.

Potency of TCM in comparison to FSH (ng/ml) for stimulation of steroid hormone secretion

Forms of TCM: Progesterone Estradiol Unfractionated TCM 300 5Q

(10%; 100 (il/ml)

22 kDa gel filtration fraction

(200 pl/ml) 65 0 1 kDa gel filtration fraction

(200 pl/ml) £5 5Q 126

These findings suggest that these two factors may be unrelated and have different mechanisms of action. In addition, the lack of stimulatory action of the 22 kDa factor on basal estradiol secretion and its inhibitory action on the 1 kDa-induced estradiol secretion partially explain our previous findings (Uzumcii et al., 1992; and Uziimcii and Lin, 1994) that unfractionated TCM had a minor stimulatory action on basal estradiol secretion (4 times) as compared to its action on basal progesterone secretion (80 times). Nevertheless, in the previous studies, TCM was able to stimulate estradiol secretion. In contrast, in this study the

22 kDa factor totally reversed the stimulatory action of the 1 kDa factor. The presence of other compound(s) that were possibly attenuating the inhibitory action of the 22 kDa factor and were removed by gel filtration chromatography may explain this discrepancy.

Previously, Gorospe et al (1993) reported that thymosin fraction-5 (TF-5), a collection of a number of thymic peptides, including parathymosin a, thymosin a l, all, 64,68,69, and 611, stimulates basal, but inhibits FSH-induced progesterone and estradiol secretion.

It is noteworthy that, in the same study, Gorospe et al. (1993) reported that TF-5 caused stimulation of both basal and FSH- induced interleukin-6 (IL-6) secretion, which was previously reported to be inhibitory for FSH-induced progesterone secretion (Gorospe et al.,

1992). In comparison to the study of Gorospe et al. (1993), in the present study, the 1 kDa factor stimulated both basal and FSH-induced progesterone and estradiol secretion. On the other hand, the 22 kDa factor stimulates both basal and FSH-induced progesterone secretion with no effect on basal estradiol secretion and an inhibitory effect on FSH-induced estradiol secretion. Therefore, it is less likely that the steroid-modulating effects of thymic factors in this study are due to components of TF-5 which may be present in our fractions. More 127 closely related to the present study, Aguilera and Romano (1989) reported the presence of factor(s) in thymic reticuloepithelial cell and thymus-conditioned media with an approximate molecular weight of 28 kDa. These factors inhibited hCG-induced steroidogenesis in whole dispersed ovarian cells. Although the closeness of the molecular weight of this inhibitory factor and our 22 kDa factor suggests the possibility that these two may be the same factor, the differences in their actions diminish this possibility. Further purification and identification studies of TCM will answer the question of whether the differences could be due to the fact that these are distinct factors or due to the experimental conditions.

Lawrence et al. (1979) previously reported that, following treatment with FSH, cultured rat granulosa cells underwent dramatic morphological changes; flattened epithelioid granulosa cells assumed a spherical shape while retaining cytoplasmic processes that attached to the substratum. The morphological alterations that were observed in the present study are similar to those described by Lawrence et al. (1979). Similar observations were made in cultured human granulosa cells (Soto et al., 1986) and Y-l adrenal tumor cells (Sato et al.,

1970) subsequent to exposure to their respective tropic hormones (i.e. hCG and ACTH).

A relationship between alteration of cytoskeleton and increase in steroidogenesis by hormone has been suggested (Soto et al., 1986). After exposure of granulosa cells to gonadotropins, the lysosomes and mitochondria, which are involved in steroidogenesis, concentrate in the perinuclear region (Soto et al., 1986). The alteration of the cell shape may facilitate this clustering of the organelles. This clustering may in turn facilitate the movement of substrate among organalles and thus enhance steroid synthesis and/or sectretion. Thus, the close resemblance of the morphological changes caused by the 1 kDa fraction to that caused by 128

FSH suggests the possibility that these two compounds may cause similar changes in the cytoskeleton of granulosa cells, thereby facilitating steroid synthesis and secretion. In addition, the lack of any morphological changes in the cells that were treated with the 22 kDa fraction indicates that the 22 kDa factor elevated progesterone secretion through a mechanism other than the clustering of organelles by altering the cytoskeleton; thus emphasing the point that these two factors are unrelated use different ways for their actions.

The body of evidence that suggests roles for growth factors and cytokines in ovarian function are convincing (Skinner and Parrott, 1994; Adashi et al., 1994), Therefore, the modulatory role of these thymic factors calls the question of whether or not they are related to any of the known growth factors or cytokines. Although, we have not assessed production of any known growth factors and cytokines, it seems reasonable that some of them may be produced by cultured thymic epithelial cells and may be responsible for the steroid- modulating effects of either of these "thymic factors" on the granulosa cells. Further purification studies are needed to answer this possibility with certainity.

It is becoming more clear that, in addition to their central role in immunity, thymic peptides have a modulatory role in ovarian function. For example, Allen et al. (1984) demonstrated that a single dose of TF-5 to 30-day-old mice significantly accelerated vaginal opening and increased serum estradiol concentration. Wolfe et al. (1990) reported that higher thymosin a l and B4 levels were associated with a more successful conception rate.

Furthermore, Prepin (1991) reported that the number of germ cells was significantly higher in fetal ovaries cultured in medium supplemented with thymulin or co-cultured with a 129 fragment of fetal thymus. Overall these reports suggest that the thymus has a role in ovarian development and function. Prepin et al. (1991) particularly suggest that the role of thymic peptides is directly on ovaries.

Since the size of the thymus gland relative to the rest of the body starts to decrease after puberty, its role in adulthood has been questioned. However, there is evidence that supports this as a possibility. As mentioned above, the association between higher concentrations of serum thymosin a 1 and thymosin B4 and a higher conception rate has been reported in cattle (Wolfe et al., 1989). In addition, Ford et al. (1990) reported marked fluctuation of thymosin 64 during the estrous cycle, indicating an interaction between thymic and ovarian functions in adult animals. In addition, there are reports demonstrating the expression of thymic hormones, including parathymosin a (Oikawa et al. 1990), thymosin

64 and 610, in rat ovaries (Hall et al. 1991a and b). Moreover, gonadotropins and PGF2(t modulate differential expression of these thymic hormones in ovaries (Hall et al 1991a and b). These findings can be interpreted in the following way: Despite the reduction of the relative size of the thymus gland after puberty, thymic hormones exist in adulthood and may play roles in ovarian functions. Locally produced thymic factors may play a vigilant role in ovarian functions in a paracrine or autocrine fashion as is the case for some growth factors

(e.g. insulin-like growth factrors). Although these findings and suggestions may be far from being conclusive for the physiological role of thymic hormones in ovarian function, they do provide an incentive for further investigation of the potential physiological role of this gland and its secretory products in the function of ovaries.

In conclusion, the present study further supports the notion that thymic secretory 130 products have a role in ovarian function by demonstrating that TCM contains two distinct and interactive steroid-modulating factors. Studies for the HPLC purification and amino acid sequence analysis for these factors are in progress in our laboratory. The mechanism of action(s), in vivo role(s) and interaction(s) of these steroid-modulating thymic factors await future studies.

SUMMARY

Previously, we have reported that unknown thymic factor(s) in TCM stimulates basal and follicle-stimulating hormone (FSH)-induced steroid hormone secretion and aromatase enzyme activity in cultured rat granulosa cells. The current study attempts to further characterize these thymic factors present in TCM. Thymic epithelial cells were prepared from immature female rats and used for TCM production. Rat granulosa cells were prepared from immature diethylstilbestrol-treated rats. Lyophilized aliquots of TCM were reconstituted with distilled water at 1/4 of the original volume and were applied to a gel filtration column (TSK-Gel G2000 SW; 75 mm ID x 30 cm; TosoHaas, PA) with a flow rate of 0.5 ml/min. Fractions (200 pi) were tested for their stimulation of steroidogenesis in rat granulosa cells. Active factors in TCM were eluted from the column in the 22,000- and

1,000-Mf regions. While the 22 kDa factor stimulates only basal progesterone and has no effect on basal estrogen secretion or morphology of the cultured rat granulosa cells, the 1 kDa factor not only stimulates basal progesterone and estrogen secretions but also induces drastic morphological changes in the rat granulosa cells. This disparity in activities suggests different mechanisms of actions. The morphological changes induced by the 1 kDa factor 131 closely resemble the typical changes that are seen in FSH-treated rat granulosa cells.

However, when both active fractions were added to rat granulosa cell culture, the stimulatory action on basal progesterone secretion was synergistic. On the contrary, the effects of these two factors on estradiol secretion was antagonistic. The exact reason for the differences between the actions of the 1 kDa and 22 kDa factors is currently unknown . In addition, 1 kDa stimulates estradiol and progesterone secretions while the 22 kDa factor stimulates progesterone, but inhibits estradiol secretion in the presence of 100 ng/ml FSH (sub-maximal dose). These observations underscore the hypothesis that the two factors possess different mechanisms of action. To better understand their role in ovarian physiology, further purification of the factors is needed. To this end, purification by HPLC and analysis of biological activity of these thymic factors are in progress FIGURES

132 Figure 20. Effect of TCM gel filtration fractions on progesterone (A) and estradiol (0) secretion. TCM was applied to gel filtration column. 81 fractions (fraction # 30-110) were tested for their effect on secretion of progesterone and estradiol in cultured rat granulosa cells. Protein profile of fractions (o) was measured at 280 nm wavelength. Data points are from a representative experiment of at least five repetitions.

133 Figure 20.

o Absorbance at 280 nm 0 ~i .0 2 0.0 - 0.5 . - 1.5 1.0 - y' —y 0 0 0 0 0 0 100 90 80 70 60 50 40 Fraction numbers Fraction 110 18 - 9 - - - 21 0 6 5 gf 15 12 s 3 • CTQ^ CfQ 1 < 0 h-i- . G o 3 I—I*o c-h H-t 1 ? 4^ ua o I Figure 21. Effect of TCM gel filtration fractions on 20-a hydroxy-progesterone (A) and aromatase enzyme activity (0) as measured by [3H]-H20 release method. TCM was applied to gel filtration column. 81 fractions (fraction # 30-110) were tested for their effect on secretion of 20a-hydroxy-progesterone and aromatase enzyme activity in cultured rat granulosa cells. Protein profile of fractions (o) was measured at 280 nm wavelength. Data points are from a representative experiment of at least three repetitions.

135 Figure 21.

Absorbance at 280 nm 2.0 . - 0.5 0.0 1.0 - cf - £ fs -c s Fraction numbers Fraction 0 0 0 ! 80 70 60 0 110 100 400 600 200 OQ OQ ►73 co CD O K> 60 r 48 - 24 - 36 - - 12 S' ‘S T3 era 3 vD n> u 3 Cu P > o O O P P CB nT o o 3 p tZ>P O >—»•O- CD P r—t- o 1 - 1 3 Ou I i 6 I UJ o\ u> Figure 22. Changes in morphology of granulosa cells following treatments: The effect of Control medium (DME-F-12/S) (a and A), 22 kDa factor (b and B), FSH (500 ng/ml) (c and C), 1 lcDa factor (d and D). The granulosa cells treated with control medium or 22 kDa factor were well-spread and appeared polygonal in shape. The cells treated with FSH or 1 kDa factor became rounded and exhibited finger-like processes attaching to the substratum. Magnifications of the micrographs indicated with small and capital letters are 350x and 700x, respectively.

137 138

Figure 22.

i 139

Figure 22 (continued). Figure 22 (continued).

i R l P I m B & E & i i m

«* 31 •,'5#«C* Figure 22 (continued). Figure 23. Synergistic action of 1 and 22 kDa factors on progesterone secretion. Cultured rat granulosa cells were treated with 1 and/or 22 kDa fractions for 48 hours, following the initial 24 hour incubation. Briefly, cells were treated with the following: 1) 45 pi (A) or 90 pi (B) of fractions from 1 kDa regions (A); 2) 45 pi (A) or 90 pi (B) of fractions from 22 kDa region (V); 3) 45 pi of fraction from each of the 1 and 22 kDa regions in combination (□), (A & B). Following the treatments, progesterone concentrations in culture medium were determined by RIA. Data points represent mean ± S.D of a single culture well from three experiments. Different letters above the same fraction number represent significant differences between different treatments using this particular fraction. For example if there is different letter on the symbol that represents fraction # 54/87, this means that stimulation of progesterone secretion is significantly higher when both fraction 54 and fraction 87 were added to the cell culture than when fraction 54 or fraction 87 were added alone. Lack of any letter on a fraction number signifies the absence of significant statistical differences between the treatments within that fraction. The statistical difference between the control and all other treatments was not shown in the figure.

142 Figure 23.

Progesterone (pg/pg cell protein) 240 180 120 60 0 V 22 kDa alone -V (45 pi)V k£)a i alone (45 pi) -A A o n s 1& 22 kDa ^ combinedin culture (45+45pi) Fraction Numbers Fraction ^ 22 kDa alone (90 V^V pi) 1kDa alone (90 pi) -A A H I 1 g &CHZI 22 kDa combined in culture (45+45pi) o n p

Figure 24. Antagonistic effect of the 22 kDa factor on 1 kDa-induced estradiol. Cultured rat granulosa cells were treated with 1 and/or 22 kDa fractions as described for the Figure 23. Briefly, cells were treated with the following: 1) 45 |_il (A) or 90 jal (B) of fractions from 1 kDa regions (A); 2) 45 |jl (A) or 90 pi (B) of fractions from 22 kDa region (V); 3) 45 |il of fractions from each of the 1 and 22 kDa regions in combination (□), (A & B). Following the treatments, estradiol concentrations in culture medium were determined by RIA. Data points represent mean ± S.D of a single culture well from three experiments. Different letters above the same fraction number represent significant differences between different treatments using this particular fraction. For example, if there is different letter on the symbol that represents fraction # 54/87, this means that there is significantly higher stimulation of estradiol secretion when both fraction 54 and fraction 87 were added to the cell culture than when fraction 54 or fraction 87 were added alone. Lack of any letter on a fraction number signifies absence of significant statistical differences between the treatments within that fraction. The statistical difference between the control and all other treatments was not shown in the figure.

144 1 & 22 kDa combined in culture (45445 jil) 1 & 22 kDa combined in culture (45445 gl) B A -A lk Da alone (90 jil) A-A 1 kDa alone (45 gl) V~V 22 kDa done (90 jj.1) W 22 kDa alone (45 gl) a

a 8 - » O ?-l O h

6 - CD o W) =L

O h 4

* 0 d -C/5 b 2 H w

b b b b b

0 / n i i i i i i i i i i r ~ n — i— i— i— r~ UUUUlUUOlUlUlUONO\ u \ c-n c/i Ln Ln u» ut o o*— io H* to W 4^ Ui o \ 9 v t « • 4 • tj • ^ • \s • %i • s* • ^ • «# • OOOOOOOOOOOOOOVOVOVO'OVO w^yio\vioo\ooHK)w^ oooooooooooooovo'Ovovo'OUJ4^LnO\~JOOVOO Figure 24. Fraction Numbers u»•£- Figure 25. Effect of various doses of FSH on progesterone and estradiol secretions in cultured rat granulosa cells. The granulosa cells were cultured with different doses of FSH for 48 hours, folowing the initial 24 hours incubation. Following the culture period, progesterone and estradiol were measured by RIA. 100 ng/ml was selected as "sub-maximal" dose of FSH and used for the dose-response study for the 1 and 22 kDa factors. Data points represent mean ± S.D. of the mean of measurements from quadruplicate wells of three experiments. Different letters above the same type of symbol show statistical difference (PcO.Ol).

146 Progesterone (pg/|ig cell protein) 1200 1500 0 - 900 0 - 300 - 600 - 2 5 10 0 50 10002000 500 200 100 50 20 0 ml) /m g n ( H S F Figure 25. 0 0 1 0 2 1

147 srdo (gp cl protein) cell (pg/pg Estradiol Figure 26. Dose-response study for the 1 and 22 kDa factors for basal and FSH-induced progesterone secretions. The granulosa cells were cultured with different amounts of 1 (A) and 22 (B) kDa factors in the absence (O) and presence of 100 ng/ml FSH (•) for 48 hours, following the initial 24 hours incubation. Following the culture period, progesterone was measured by RIA. Presence of the same letter on the same type of symbol indicates lack of statistically significant differences (P <0.01).

148 750 200 g • CD A A A e Cu A

bu protein) (pg/pg secretion progesteroneBasal .G. 600 - - 160 '5i) a oa AB

§ 450 - 120 a> M

C/3 Oh d 0 l r 0 I 10 20 45 90 180 0 10 20 45 90 180 Doses of 1 kD a Factor (pi) Doses of 22 kDa Factor (pi) Figure 26. VO Figure 27. Dose-response study for 1 and 22 kDa factors for basal and FSH-induced estradiol secretions. The granulosa cells were cultured with different amounts of 1 (A) and 22 (B) kDa factors in the absence (O) and presence of 100 ng/ml FSH (•) for 48 hours, following the initial 24 hours incubation. Following the culture period, estradiol was measured by RIA. Presence of the same letter on the same type of symbol indicates lack of statistically significant differences (P <0.01).

150 60 15

a protein) (pg/pg secretion estradiol-173 Basal i 3 • w * T3 c3 24 - - 6 <-i -t—>00

0 n ------1------1------1------1------1— l------1------r 0 I 0 10 20 45 90 180 0 10 20 45 90 180 Doses of 1 kD a Factor (jil) Doses of 22 kDa Factor (|il) Figure 27. CONCLUDING REMARKS

After presentation of the experimental data in the previous three chapters, the reader would probably ask the following question. What may be the identity of the 1 and 22 kDa factors? The exact answer for this question is not available yet. Nevertheless, in this section, the findings most relevant to this study regarding the effect of several known growth factors and cytokines on ovarian granulosa cell function will be presented. Then, the actions of the

1 and 22 kDa factors on cultured granulosa cells will be summarized. Making a comparison between actions and characteristics of these two thymic factors and already known factors will be left to the reader. However, since there are extensive reviews about the actions of growth factors (Skinner and Parrott 1994; Lin et al., in press, 1994) and cytokines (Adashi et al., 1994) on ovarian function, we do not attempt to review the subject here. Rather, we will select the examples that are the most relevant to our study.

EGF/TGFa: Epidermal growth factor (EGF) is a 53-amino acid single polypeptide chain that was originally isolated from the mouse submaxillary gland. TGFa is one of the structurally related peptides belonging to the EGF family. Since they have similar protein structure, these factors act at the same receptor; therefore, they act similarly. EGF has been reported to be inhibitory for FSH-induced aromatase activity and the formation of LH receptor in rat and pig granulosa cells (Hsueh et al., 1981; Mondschein and Schomberg

152 153

1981). It has also been reported that EGF potentiates P450scc mRNA accumulation and basal and FSH-induced progesterone secretion, but has no effect on cAMP formation in rat granulosa cells (Trzeciak et al., 1987). EGF is also stimulatory for basal and hCG-induced progesterone secretion in human granulosa-luteal cells (Richardson et al., 1989).

TGFP: Transforming growth factor beta (TGFP) is a 25 kDa dimeric peptide consisting of two homologous subunits. TGFP appears to stimulate most of the granulosa cell functions including FSH-induced aromatase activity and cAMP accumulation (Adashi et al., 1989), progesterone secretion (Dodson and Schomberg, 1987; Ohmura et al, 1993), granulosa cell proliferation (Dorrington et al., 1988) and FSH-dependent LH-receptor formation (Giray-Goten et al., 1993) in the rat. In contrast, TGFP is inhibitory for most of these functions in pig granulosa cells (Mondschein et al., 1988; Chang et al., 1993; Gitay-

Goren et al., 1993)

TNFa: Tumor necrosis factor a (TNFa) is a 17 kDa homotrimeric glycoprotein, primarily produced by macrophages. TNFa is locally produced in ovaries (Sancha-Tello and

Terranova, 1992). It has stimulatory action on progesterone secretion in rat theca cells (Roby and Terranova, 1990) and human granulosa cells (Zolti et al., 1990). In contrast, TNFa inhibits FSH-induced aromatase enzyme activity (Emato and Baird, 1988).

IL-1: This 15-17 kDa protein is primarily produced by macrophages. It has been shown that IL-1 is produced in the ovary locally (Hurwitz et al., 1991). EL-1 previously has been shown to be inhibitory for gonadotropin-induced granulosa cell differentiation in both rat (Gottschall et al., 1989) and porcine granulosa cells (Fukuoka et al., 1989b). 154

This current study demonstrated that proteinaceous secretory products present in the thymic epithelium cell culture-conditioned medium are capable of stimulating progesterone and estradiol secretions. The action of these factors are not mediated through the cAMP- mediated second messenger system. Parallel elevation of 20a -hydroxyprogesterone and progesterone; and estradiol secretion and [3H]-H20 release from labeled androstenedione indicate that the observed increases in progesterone and estradiol secretions in TCM-treated rat granulosa cells are likely to be due to elevated activities of specific steroidogenic enzymes. Measurements of total cell protein and DNA synthesis indicate that enhanced steroidogenesis in TCM-treated cells is not due to increased cell growth and/or proliferation.

Rather, the enhanced steroidogenesis is probably due to an increased steroid biosynthetic capability of the cells. In addition, not only can TCM stimulate production of these steroid hormones in the absence of FSH, but also in the presence of FSH, suggesting physiological relevance of its stimulatory actions.

Gel-filtration chromatography analysis revealed that active factors in TCM were eluted from the column in the 22,000- and 1,000-Mr regions. While the 22 kDa factor stimulates only basal progesterone and has no effect on basal estrogen secretion or morphology of the cultured rat granulosa cells, the 1 kDa factor not only stimulates basal progesterone and estrogen secretions but also induces drastic morphological changes in the rat granulosa cells.

This disparity in activities suggests different mechanisms of actions. The morphological changes induced by the 1 kDa factor closely resemble the typical changes that are seen with

FSH-treated rat granulosa cells. However, when both active fractions are added to rat granulosa cell culture, the stimulatory action on basal progesterone secretion is synergistic. 155

In addition, 1 kDa stimulates estradiol and progesterone secretions while 22 kDa stimulates progesterone but inhibits estradiol secretion in the presence of 100 ng/ml FSH (sub-maximal dose). Although the exact reason for the difference between the action of 1 and 22 kDa factor is not known, these observations underscore the hypothesis that the two factors possess different mechanisms of action. However, we previously did not find a parallel increases in the intracellular cAMP concentration when estradiol and progesterone production were elevated by unfractionated TCM in cultured rat granulosa cell (Uzumcu et al., 1992). However, intracellular concentrations of cAMP was not determined in the either the 1 kDa or 22 kDa treated cells. Therefore, whether cAMP or other second messengers are involved in the stimulation of steroidogenesis by 1 kDa or 22 kDa factor is not known.

Although, the involvement of cAMP in the stimulation of steroidogenesis in the 1 kDa or 22 kDa-treated cells is not known, since their effect on the granulosa cell is different, it is possible that their mechanisms of action are also different. To better understand their role in ovarian physiology, further purification of the factors is needed. To this end, following reverse-phase HPLC separation, we have sent the 1 kDa factor for both amino acid composition and peptide sequencing analysis at Case Western Reserve University,

Molecular Biology Core Laboratory (Cleveland, OH). We were able to get amino acid composition of the 1 kDa factor and the number of occurrences in the peptide as follows:

Asp, (1); GIu, (1); Gly, (2); Ala (1) and Pro (2). (Numbers in parenthesis show the occurrence of the residue in the peptide.) However, we were not able obtain information about primary sequence of the peptide. The possible reasons are: (1) insufficient amount of peptide, (2) blockage of the N-terminal and/or (3) loss of the peptide due its low molecular 156 weight during sequencing. Currently, we are in pursuit of chromatographic separation and amino acid sequencing analysis of a larger amount of 1 kDa factor.

Approaches to carry the present study further are only limited by one's imagination.

Some of these are: further purification and identification of these thymic factors; determination of the second messenger system that is involved in the mechanism of action of these factors; delineation of the molecular biological mechanism for their production from thymic cells and their stimulatory action in the granulosa cells; in vitro action of these factors; and investigation of possible roles of these factors in other organ systems (e.g. modulation of functions of immune cells). LIST OF REFERENCES

Abraham GE (1969) Solid-phase radioimmunoassay of estradiol-17 p. J Clin Endocrinol Metab 29: 866-870.

Adashi EY, Kokia E and Hurwitz A (1994) Potential relevance of cytokines to ovarian physiology. In: Molecular Biology of the Female Reproductive (Findlay,System, J. K., ed), Academic Press, San Diego, pp 83-99.

Adashi EY (1990) The potential relevance of cytokines to ovarian physiology: The emerging role of resident ovarian cells of the white blood cells series. Endocr Rev 11:454-464.

Adashi EY, Resnick CE, Hernandez ER, MayJV, PurchioAF andTwardzikDR(1989) Ovarian transforming growth factor-beta (TGF beta): cellular site(s), and mechanism(s) of action. Mol Cell Endocrinol 61:247-256

Aguilera G and Romano MC (1989) Influence of the thymus on steroidogenesis by rat ovarian cells in vitro. J. Endocrinol. 123: 367-373.

Akira S, Ohmura H, Uziimcu M, Araki T, and Lin YC (1994) Gossypol inhibits aromatase activity in cultured porcine granulosa cells. Theriogenology 41:1489-1497.

Allen LS, McClure JE, Goldstein AL, Barkley MS and Michael SD (1984) Estrogen and thymic hormone interactions in the female mouse. J Reprod Immunol 6: 25-37.

Alten H-E and Groscurth P (1975) The postnatal development of the ovary in the "nude" mouse. Anat Embryol 145:35-46.

Andreani CL, Payne DW, Packman JN, Resnick CE, Hurwitz A and Adashi EY (1991) Cytokine-mediated regulation of ovarian function. Tumor necrosis factor alpha inhibits gonadotropin-supported ovarian androgen biosynthesis. J Biol Chem 266: 6761-6766.

Auger G, Monier JC, Dardenne M, Pleau JM and Bach J-F (1982) Identification of FTS (facteur thymique serique) on thymus ultrathin section using monoclonal antibodies. Immunol Lett 5:213-216.

Bagavandoss P, Wiggins RC, Kunkel, DG, Remick, DG and Keyes PL (1990) Tumor

157 158 necrosis factor production and accumulation of inflammatory cells in the corpus luteum of pseudopregnancy and pregnancy in rabbits. Biol Reprod 42: 367-376.

Bearsley TR, Peirschbacher M, Wetzel GD, and Hays EF (1983). Induction of T-cell maturation by cloned line of thymic epithelium (TEPI). Proc Natl Acad Sci USA 80:6005- 6009.

Besedowsky HO and Sorkin E (1974) Thymus involvement in female sexual maturation. Nature (London) 249:356-358.

Bjersing L (1967) Histochemical demonstration of A5-3B-and 17B-hydroxysteroid dehydrogenase activities in porcine ovary. Histochemie 10:295-304.

Bradford M (1976) A rapid and sensitive method for quantification of microgram quantities of protein utilizing the principles of protein dye binding. Anal Biochem 7 2:248-254.

Brannstrom M, Wang L and Norman RJ (1993a) Ovulatory effect of interleukin-1 beta on the perfused rat ovary. Endocrinology 132:399-404.

Brannstrom M, Wang L and Norman RJ (1993b) Effects of cytokines on prostaglandin production and steroidogenesis of incubated preovulatory follicles of the rat. Biol. Reprod. 48:165-171.

Calzolari A (1898) Recherhes experimentales sur un rapport probable entre la function du thymus et celle des testicules. Arch Ital Biol Torino 30:71-77.

Chang WY, Ohmura H, Coskun S, and Lin YC, (1993) Inhibition of progesterone production in cultured porcine granulosa cells by transforming growth factor B1. In: Leung PCK, Hsueh AJW, Friesen HG (eds), Molecular Basis of Reproductive Endocrinology, Springer-Verlag, New Y ork, 1993; pp 234-242.

ChiodiH (1940) The relationship between the thymus and the sexual organs. Endocrinology 26:107-116.

Crawford KA, Hirshfield AN and DePaolo LV (1992) Correlation of age-associated increases in follicle stimulating hormone secretion with decreases in antral follicles: failure of progesterone-induced acyclicity to prevent these changes. Mech Age Dev 64:111-122.

DardenneM, SavinoW, Feutren G and Bach J-F (1987) Stimulatory effects of cyclosporin A on human and mouse thymic epithelial cells. Eur J Immunol 17:275-279.

DardenneM, SavinoW,GagneraultM-C, ItohTand Bach J-F, (1989) Neuroendocrine control of thymic hormonal production. I. Prolactin stimulates in vivo and in vitro the 159 production of thymulin by human and murine thymic epithelial cells. Endocrinology 125: 3-12.

Dardenne M, Pleau JM, and Bach J-F (1980) Characterization of facteur thymique serique (FTS) in thymus, n. Direct demonstration of the presence of FTS in thymosin fraction V. Clin Exp Immunol 42:477-482.

Dardenne M, Pleau JM, Man NK, and Bach J-F (1977) Structural study of circulating thymic factor: A peptide isolated from pig serum. I. Isolation and purification. J Biol Chem 252:8040-8044.

De Angelo L, and Michael SD (1987) The effect of antisera to thymosin a 1 on the course of autoimmune ovarian dysgenesis in neonatally thymectomized mice. J Reprod Immunol 11:41-53.

De Angelo L, Kaikis-Astaras A, and Michael SD (1989) Aromatase activity in cultured ovaries from neonatally thymectomized mice with ovarian dysgenesis. J Reprod Immunol 16:115-126.

Dodson WC and Schomberg DW (1987)The effect of transforming growth factor-beta on follicle-stimulating hormone-induced differentiation of cultured rat granulosa cells. Endocrinology 120:512-516

Dorrington J, Chuma AV and Bendell JJ (1988) Transforming growth factor beta and follicle-stimulating hormone promote rat granulosa cell proliferation. Endocrinology 123:353-359.

Dorrington JH, Fritz IB, and Armstrong DT (1978) Control of testicular estrogen synthesis. Biol Reprod 18:55-64.

Dorrington JH, Moon YS and Armstrong DT (1975) Estradiol-176 biosynthesis in cultured rat granulosa cells from hypophysectomized immature rats; stimulation by follicle- stimulating hormone. Endocrinology 97:1328-1331. Dougherty JF (1952) Effects of hormones on lymphatic tissue. Physiol Rev 32:379-401.

DuBois RN, Simpson ER, Kramer RE, and Waterman MR (1981) Induction of synthesis of cholesterol side chain cleavage cytochrome P450 by adrenocorticotropin in cultured bovine adrenocortical cells. J Biol Chem 13:7000-7005.

Emota N and Baird A (1988) The effect of tumor necrosis factor/cachectin on follicle-stimulating hormone-induced aromatase activity in cultured rat granulosa cells. Biochem Biophys Res Commun 153:792-798.

Fabris N and Mocchegiani E (1985) Endocrine control of thymic serum factor production 160 in young-adult and old mice. Cellular Immunol 91:325-335.

Fairchild Benyo D and Pate JL (1992) Tumor necrosis factor-a alters bovine luteal cell synthetic capacity and viability. Endocrinology 130,854-860.

Farkash Y, Timberg R, and Orly J (1986) Preparation of anti-serum to rat cytochrome P-450 cholesterol side chain cleavage, and its use of ultrastructural localization of the immunoreactive enzyme by protein A-gold technique. Endocrinology 118:1353-1365.

Flanagan SP (1966) "Nude": A new hairless gene with pleiotropic effects in the mouse. Genet Res 8:295-309.

Ford JJ, Vakharia DD, Anderson LL and Klindt J (1990) Thymosin-beta 4 concentrations during the estrous cycle and after hypophyseal stalk transection of female pigs. P.S.E.B.M. 193:185-189.

Fortune JE and Armstrong DT (1978) Hormonal control of 176-estradiol biosynthesis in proestrous rat follicles: estradiol production by isolated theca versus granulosa. Endocrinology 102:227-235.

Fortune JE, and Armstrong DT (1977) Androgen production by theca and granulosa isolated from proestrous rat follicles. Endocrinology 100:1341-1347.

Fukuoka M, Mori T, Taii S and Yasuda K (1988) Interleukin-1 inhibits luteinization of porcine granulosa cells in culture. Endocrinology 122:367-369.

Fukuoka M, Yasuda K, Taii S, Takakura K and Mori T (1989a) Interleukin-1 stimulates growth and inhibits progesterone secretion in cultures of porcine granulosa cells. Endocrinology 124:884-890.

Fukuoka M, Taii S, Yasuda K, Takakura K and Mori T (1989b) Inhibitory effects of interleukin-1 on luteinizing hormone-stimulated adenosine 3',5'-monophosphate accumulation by cultured porcine granulosa cells. Endocrinology 125:136-143.

Gibson-D'Ambrosio RE, Samuel M, Chang CC, Trosko JE, and D'Ambrosio, SM (1987) Characteristics of long-term human epithelial cell cultures derived from normal human fetal kidney. In Vitro Cell and Devi Biol 23:279-287.

Gitay-Goren H, Kim I-C, Miggans ST and Schomberg DW (1993) Transforming growth factor beta modulates gonadotropin receptor expression in porcine and rat granulosa cells differently. Biol Reprod 48:1284-1289.

Goldstein AL, Low TLK, McAdoo M, McClure J, Thurman JB, Rossio J, Lai J-Y, Chang 161

D, Wang SS, Harvey C, Ramel AH, and Meienhofer, Jr (1977) Thymosin a 1: Isolation and sequence analysis of an immunologically active polypeptide. Proc Natl Acad Sci USA 74:725-729.

Goldstein G, and Mackay IR (1969) Structure and development of the human thymus. In: The Human Thymus, Warren H. Green, Inc., St. Louis, Missouri, pp 1-36.

Gore-Langton RE, and Armstrong DT (1988) Follicular steroidogenesis and its control. In: The Physiology of Reproduction. (E Knobil and J Neil, eds). Raven Press, Ltd. New york, pp 331-385.

Gore-Langton R, McKeracher H and Dorrington J (1980) An alternative method for the study of follicle-stimulating hormone effects on aromatase activity in Sertoli cell cultures. Endocrinology 107:464-471.

Gorospe WC, Fong YY and Spangelo BL (1993) Thymosin fraction-5 modulates steroidogenesis and interleukin-6 release by rat ovarian granulosa cells in vitro. Endocr J 1:35-39.

Gorospe WC, Hughes FM Jr and Spangelo BL (1992) Interleukin-6: effects on and production by rat granulosa cells in vitro. Endocrinology 130:1750-1752.

Gorospe WC and Kasson BG (1988) Lymphokines from concanavalin-A-stimulated lymphocytes regulate rat granulosa cell steroidogenesis in vitro. Endocrinology 123:2462- 2471.

Gorospe WC and Spangelo BL (1993) Interleukin-6: potential roles in neuroendocrine and ovarian function. Endocr J 1:3-9.

Gottschall PE, Katsuura G, Hoffman ST and Arimura A (1988)Interleukin 1: an inhibitor of luteinizing hormone receptor formation in cultured rat granulosa cells. FASEB J 2:2492- 2496.

Gottschall PE, Uehara A, Hoffman ST and Arimura A (1987) Interleukin-1 inhibits follicle stimulating hormone-induced differentiation in rat granulosa cells in vitro. Biochem Biophys Res Commun 149:502-509.

Gottschall PE, Katsuura G and Arimura A (1989) Interleukin-1 suppresses follicle-stimulating hormone-induced estradiol secretion from cultured ovarian granulosa cells. J Reprod Immunol 15:281-290.

Gower DB (1984) The role of cytochrome P-450 in steroidogenesis and properties of the steroid-transforming enzymes. In: Biochemistry of Steroid Hormones. (H U Makin, ed). 162

Blackwell Scientific, Oxford, UK, pp 230-292.

Grossman CJ, Sholiton LJ and Nathan P (1979) Rat thymic estrogen receptors. I. Preparation, location and physicochemical properties. J Steroid Biochem 11:1233-1240.

Hall NR, McGillis JP, Spangelo BL, and Goldstein AL (1982) Evidence for an interaction between thymosin peptides and the pituitary-gonadal axis. Fed Proc, Abstr 41:1267.

Hall A K, Aten RI and Behrman HR (1991a) Thymosin gene expression is modulated by pregnant mare's serum gonadotropin, human chorionic gonadotropin, and prostaglandin F2 alpha in the immature rat ovary. Endocrinology 128:951-957.

Hall NR, O'Grady MP and Menzies RA (1992) Thymic regulation of the hypothalamic-pituitary-gonadal axis. Int J Immunopharmac 14,353-359.

Hall AK, Aten RI and Behrman HR (1991b) Differential modulation of thymosin genes in the immature rat ovary by gonadotropins. Moll Cell Endocrinol 79:37-43.

Hardy B, Damon D, Eshkol A, and Lunenfeld B (1974) Ultrastructural changes in the ovaries of infant mice deprived of endogenous gonadotropins after substitution with FSH. J Reprod Fertil 36:345-352.

Hattori M and Brandon M R (1979) Thymus and the endocrine system: ovarian dysgenesis in neonatally thymectomized rats. J Endocrinol 83:101-111

Healy DL, Bacher J and Hodgen GD (1985) Thymic regulation of primate fetal ovarian-adrenal differentiation. Biol Reprod 32:1127-1133.

Henderson J (1904) On the relationship of thymus to the sexual organs. I. The influence of castration on the thymus. J Physiol 31:222-229.

Hooper JA, MacDaniel MC, Thurman GB, Cohen GH, Schulof RS, and Goldstein AL (1975) The purification and properties of bovine thymosin. Ann NY Acad Sci 249:145-153.

Hsueh AJW, Wang D, and Erickson GF (1980) Inhibitory effect of gonadotropin-releasing hormone upon follicle-stimulating hormone induction of luteinizing hormone receptor and aromatase activity in rat granulosa cells. Endocrinology 106:1697-1705.

Hsueh AJW, Adashi EY, Jones PBC, and Welsh TH Jr (1984) Hormonal regulation of the differentiation of cultured ovarian granulosa cells. Endocr Rev 5:76-127.

Hu YF, Chang CJG, Brueggemeier RW and Lin YC (1993) Gossypol inhibits basal and estrogen-stimulated DNA synthesis in human breast carcinoma cells. Life Sci 53:433-438. 163

Hughes FM Jr, Lane TA, Chen TT and Gorospe WC (1990) Effects of cytokines on porcine granulosa cell steroidogenesis in vitro. Biol Reprod 43:812-817.

Hughes FM Jr, Pringle CM and Gorospe WC (1991) Production of progestin-stimulatory factor(s) by enriched populations of rat T and B lymphocytes. Biol Reprod 44:922-929.

Hurwitz A, Ricciarelli E, Bolero L, Rohan RM, Hernandez ER and Adashi EY (1991). Endocrine and autocrine-mediated regulation of ovarian (thecal-interstitial) interleukin-16 gene expression: gonadotropin dependent preovulatory acquisition. Endocrinology 129: 4227-3429.

Hurwitz A, Dusknik M, Solomon H, Ben-Chetrit A, Finci-Yeheskel Z, Mildwidsky A, Mayer M, Adashi EY and Yagel S (1993) Cytokine-mediated regulation of rat ovarian function: interleukin-1 stimulates the accumulation of a 92-kilodalton gelatinase. Endocrinology 132:2709-2714.

Hurwitz A, Hernandez ER, Payne DW, Dharmarajan AM and Adashi EY (1992) Interleukin-1 is both morphogenic and cytotoxic to cultured rat ovarian cells: obligatory role for heterologous, contact-independent cell-cell interaction. Endocrinology 131:1643-1649.

Jambon B, Montagne P, Bene MC, Brayer MP, Faure G, and Bach J-F (1981) Immunohistologic localization of "facteur thymique serique" (FTS) in human thymic epithelium. J Immunol 127:2055-2059.

Jones PBC, Welsh Jr TH, Hsueh AJW (1982) Regulation of ovarian progestin production by epidermal growth factor in cultured rat granulosa cells. J Biol Chem 257:11268-11273.

Kendall MD (1981) The cells of the thymus gland. In: The Thymus Gland. Academic Press, London, pp 63-83.

Kendal MD, Safieh B, Sareen A, Venn G, Matheson L, and Ritter M (1991) Thymulin secreting cells in the man: Distribution, LM histochemistry and plasma thymulin levels. In: Lymphatic Tissues and In vivo Immune Responses (B Imhof, S Berrih-Aknin and S Ezine, eds) Marcel-Dekker, New York, pp 41-43.

Kendal MD and Clarke AG (1994) The female thymus and reproduction in mammals. Oxford Reviews of Reproductive Biology 16: 162-213.

Kojima A, Sakakura T, Tanaka Y, and Nishizuka Y (1973) Sterility in neonatally thymectomized female mice: Its nature and prevention by the injection of spleen cells. Biol Reprod 8:358-361.

Kokia E, Hurwitz A, Ricciarelli E, Tedeschi C, Resnick CE, Mitchell MD and Adashi EY 164

(1992) Interleukin-1 stimulates ovarian prostaglandin biosynthesis: evidence for heterologous contact-independent cell-cell interaction. Endocrinology 130:3095-3097.

Koritz SB, and Kumar AM (1970) On the mechanism of action of adrenocorticotropic hormone: the stimulation of the activity of enzymes involved in pregnenolone synthesis. J Biol Chem 245:152-159.

Kosiewicz MM, and Michael SD (1990) Humoral or cellular thymic reconstitution of ovary function in neonatally thymectomized mice. Biol Reprod 42: Suppl #1 Abstract #104.

Kosiewicz MM and Michael SD (1990) Neonatal thymectomy affects follicle population before the onset of autoimmune oophoritis in B6A mice. J Reprod Fert 88 427-440.

Lacovara J, and Utermohlen V (1983) Isolation and assay of a human plasma factor affecting human thymus-derived lymphocytes. Clin Immunol Immunopathol 27:428-432.

Lawrence T S, Ginzberg RD, Gilula NB and Beers WH (1979) Hormonally induced cell shape changes in cultured rat ovarian granulosa cells. J Cell Biology 80:21-36

Ledwitz-Rigby F, and Scheid PG (1990) Serum thymic factor enhances porcine granulosa cell responsiveness to gonadotropins in vitro. VIII Ovarian Workshop: Regulatory Processes and Gene Expression in the Ovary. July 12-14, 1990. Maryville, Tennessee. Abstr #31.

Leung PCK and Wang J (1989). The role of inisitol lipid metabolism in the ovary. Biol Reprod 40:703-708.

Lewis VM, Twomey JJ, Belamear P, Goldstein G and Good RA (1978) Age, thymic involution, and circulating thymic hormone activity. J Clin Endocrinol Metab 47:145-150.

Lin YC, Coskun S and Chang WY (1994) Growth factor regulation of oocyte maturation in mammals. Arch Androl (in press).

Liberman S, Greenfield NJ, and Wolfson A (1984) A heuristic proposal for understanding steroidogenic processes. Endocr Rev 5:127-148.

Lintem-Moore S (1977) Effect of athymia on the initiation of follicular growth in rat ovary. Biol Reprod 17:155-161.

Lintem-Moore S, and Pantelouris EM (1976b) Ovarian development in athymic mice. V. The effect of PMSG upon the numbers and growth of follicles in the early juvenile ovary. Mech Ageing Dev 5:259-265.

Lintem-Moore S and Pantelouris EM (1975) Ovarian development in athymic nude mice. 165

I. The size and the composition of the follicle population. Mech Ageing Dev 4:385-390.

Lintem-Moore S, and Pantelouris EM (1976a) Ovarian development in athymic mice. EL The effect of PMSG and estradiol upon the size and composition of the ovarian follicle population. Mech Ageing Dev 5:33-38.

Low TL, and Goldstein AL (1978) Structure and function of thymosin and other thymic factors In: The Year in Hematology (R Silber, J LoBue, and AS Gordon, eds). Marcel Decker, New York, pp 49-131.

Marchetti B (1989) Involvement of thymus in reproduction. Prog Neuroendocrineimmunol 2:64-69.

Marmor LH, and Michael SD (1984) Circulating levels of testosterone in neonatally thymectomized female mice. IRCS Med Sci 12:1022-1023.

Marsh JM (1975). The role of cyclic AMP in gonadal function. Adv Cyclic Nucleotide Res 6:137-199.

McCruden, A. B. and Stimson, W. H. (1981) Androgen binding cytosol receptors in the rat thymus: Physicochemical properties, specificity and localization. Thymus 3,105-117.

Michael SD (1983) Interactions of the thymus and the ovary. In: Factors Regulating Ovarian Functions. (GS Greenwald and PF Terranova, eds). Raven Press, New York, pp 445-464.

Michael SD, Taguchi O and Nishizuka Y (1980) Effect of neonatal thymectomy on ovarian development and plasma LH, FSH, GH and PRL in the mouse. Biol Reprod 22:343-350.

Michael SD, Taguchi O, Nishizuka Y, McClure JE Goldstein AL, and Barkley MS. (1981) The effect of neonatal thymectomy on early follicular loss and circulating levels of corticosterone progesterone, estradiol and thymosin. In: Dynamics o f Ovarian Function (NB Schwartz and M Hanzicker-Dunn, eds) Raven Press, New York, pp 279-284.

Mihich E and Fefer A (1983) Thymic factors and Hormones. In: Biological Response Modifiers; Subcommittee Report, (NCI monograph # 63), Bethesda Maryland, pp 104-137.

Miller JFAP (1961) Immunological function of thymus. Lancet 2:748-749.

Mizutani S, Akashi S, Terada N, and Matsumoto K (1979) A comparison of metabolism of progesterone to 5a-steroids in monkey, mouse, and rat ovaries. J Steroid Biochem 10: 549- 552.

Monier JC, Dardenne M, Pleau JM, Deschaux P, and Bach J-F (1980) Characterization of facteur thymique serique (FTS) in the thymus. I. Fixation of anti-FTS antibodies on thymic 166 reticuloendothelial cells. Clin Exp Immunol 42:470-476.

Mondschein JS and Schomberg DW (1981) Growth factors modulate gonadotropin receptor induction in granulosa cell cultures. Science 211:1179-1180.

Mondschein JS, Canning SF and Hammond JM (1988) Effects of transforming growth factor-beta on the production of immunoreactive insulin-like growth factor I and progesterone and on [3H]thymidine incorporation in porcine granulosa cell cultures. Endocrinology 123:1970-1976

Nieburgs AC, Picciano PT, Korn JH, McCalister T, Allred C, and Cohen S (1985) In vitro growth and maintenance of two morphologically distinct population of thymic epithelial cells. Cell Immunol 90:439-450.

Nishizuka Y, Sakakura., Tsujimura T and Matsumoto K (1973) Steroid biosynthesis in vitro by dysgenetic ovaries induced by neonatal thymectomy in mice. Endocrinology 93: 786-792.

Nishizuka Y, and Sakakura T (1969) Thymus and reproduction: sex-linked dysgenesis at the gonad after neonatal thymectomy in mice. Science 166: 753-755.

Nishizuka Y and Sakakura T (1971). Ovarian dysgenesis induced by neonatal thymectomy in the mouse. Endocrinology 89:886-893.

Ohmura H, Chang WY, Uziimcii M. Coskun S, Akira S, Araki T, and Lin YC, (1993) Transforming growth factor B, stimulates progesterone production in cultured granulosa cells from gonadotropin-primed adult rats. In: Leung PCK, Hsueh AJW, Friesen HG (eds), Molecular Basis of Reproductive Endocrinology, Springer-Verlag, New York, 1993; pp 215- 221.

Oikawa M, Dargan C, Ny T and Hsueh AJW (1990) Expression of gonadotropin-releasing hormone and prothymosin-alpha messenger ribonucleic acid in the ovary. Endocrinology, 127:2350-2356.

Osawa Y and Spaeth DG (1971) Estrogen biosynthesis. Stereospecific distribution of tritium in testosterone-1 alpha,2 alpha-t2. Biochemistry 10: 66-71.

Pawlikowski M, Kula K, and Slawikowska J (1990) Effect of splenectomy and spleen dialysate on the testicular response to human chorionic gonadotropin in rats. Med Sci Res 18:177-178.

Pleau JM, Dardenne M, Blouquit Y, and Bach J-F (1977) Structural study of circulating thymic factor: A peptide isolated from pig serum . II. Amino acid sequencing. J Biol Chem 252:8045-8047. 167

Prepin J (1991) Fetal thymus and thymulin stimulate the in vitro proliferation of oogonia in the fetal rat ovary. C R Acad Sci Paris, 313: Seri HI, 407-411.

Rebar RW, Morandini IC, Silva de Sa MF, Erickson GF and Petze JE (1981b) The importance of thymus gland for normal reproductive function in mice. In: The Dynamics of Ovarian Function (NB Schwartz and M Hunzicker-Dunn, eds). Raven Press, New York, pp 465-469.

Rebar RW, Morandini IC, Erickson GF, and Petze JE (1981a) The hormonal basis of reproductive defects in athymic mice. I. Diminished gonadotropin concentration in prepubertal females. Endocrinology 108:120-126.

Rebar RW, Morandini IC, Petze JE Erickson GF (1982b) Hormonal basis of reproductive defects in athymic mice: Reduced gonadotropins and testosterone in males. Biol Reprod 27:1267-1276.

Rebar RW, Miyake A, Erickson GF, Low TLK, and Goldstein AL (1983). The influence of thymus gland on reproductive function: A hypothalamic site of action. In: Factors Regulating Ovarian Function. (GS Greenwald and PF Terranova, eds). Raven Press New York, pp 465-469.

Rebar RW (1984) Effects of thymic peptides on hypothalamic-pituitary function. In: Thymic Hormones and Lymphokines. (AL Goldstein, ed). Plenum Press, New York, pp 325-334.

Rebar R W , Miyake A, Low TLK, and Goldstein AL( 1981c). Thymosin stimulates secretion of luteinizing hormone-releasing factor. Science 214:669-671.

Rebar RW, Latham A, and Petze J (1982a) Thymic peptides stimulate secretion of Luteinizing hormone Releasing Factor (LRF). Abstracts of the 64th Annual Meeting of Endocrine Society, Abstr #11.

Richardson MC, Gadd SC, and Masson GM (1989) Augmentation by epidermal growth factor of basal and stimulated progesterone production by human luteinized granulosa cells. J Endocrinology 121:397-402.

Roby K and Terranova PF (1988) Tumor necrosis factor alpha alters follicular steroidogenesis in vitro. Endocrinology 123:2952-2954.

Roby K and Terranova PF (1990) Effects of tumor necrosis factor-alpha in vitro on steroidogenesis of healthy and atretic follicles of the rat: theca as a target. Endocrinology 126:2711-2718.

Sagaguchi S, Takahashi T, and Nishizuka Y (1982a) Study on the cellular event in post­ thymectomy autoimmune oophoritis in mice. I. Requirement of Lyt-1 Effector cells for 168 oocytes damage after adoptive transfer. J Exp Med 156:1565-1576

Sagaguchi S, Takahashi T, and Nishizuka Y (1982b) Study on the cellular event in post­ thymectomy autoimmune oophoritis in mice. E. requirement of Lyt-1 cells in normal female mice for prevention of oophoritis. J Exp Med 156:1577-1586.

Sakabe K, Seiki K and Fujii-Hanamoto H (1986) Histochemical localization of progestin receptor cells in the rat thymus. Thymus 8,97-107.

Sakakura, T. and Nishizuka, Y. (1972). Thymic control mechanism in ovarian development: Reconstitution of ovarian dysgenesis in thymectomized mice replacement with thymic and other lymphoid tissues. Endocrinology 90:431-437.

Sakakura, T., Taguchi, O., Kojima, A. and Nishizuka, Y. (1979) Rapid oocyte death prior to and at puberty in neonatally thymectomized mice. In: Ovarian Follicular and Corpus Luteum Function. CP Channing, ( JM Marsh and WA Sadler, eds), Plenum Press, New York, pp 315-320.

Sancho-Tello M, Perez-Roger I, Imakawa K, Tilzer L and Terranova PF (1992). Expression of tumor necrosis factor-alpha in the rat ovary. Endocrinology 130:1359-1364.

Sancho-Tello M and Terranova PF (1991) Involvement of protein kinase C in regulating tumor necrosis factor alpha-stimulated progesterone production in rat preovulatory follicles in vitro. Endocrinology 128:1223-1228

SAS Institute Inc. (1985) SAS User's Guide: Statistics, Version 5 Edition. Cary, NC: SAS Institute Inc., 956 pp.

Sato G, Augusti-Tocco G, Posner M and Kelly P (1970) Hormone-secreting and hormone-responsive cell cultures. Rec Prog Horm Res 26:536-546.

Savino W, Dardenne M, Papiemik M, and Bach J-F (1979) Thymic hormone-containing cells: Characterization and localization of serum thymic factor in young mouse thymus studied by monoclonal antibodies. J Exp Med 156:628-634.

Skinner MK, and Parrott JF (1994) Growth factor-mediated cell-cell interactions in the ovary. In: Molecular Biology of the Female Reproductive System, (Findlay , J. K., ed) pp. 67-81, Academic Press, San Diego.

Snedecor GW and Cochran WG (1980) Statistical Methods, Seventh Edition, The Iowa State University Press, Ames, Iowa.

Soto EA, Kliman HJ Strauss, J F IE and Paavola LG (1986) Gonadotropins and cyclic adenosine 3',5-monophosphate (cAMP) alter the morphology of cultured human granulosa 169 cells. Biol Repro 34:559-569.

Steiner AL, Parker CW, and Kipnis DM (1972) Radioimmunoassay for cyclic nucleotides. J Biol Chem 247:1106-1114.

Stouffer RL, Nixon WE, Gulyas BJ, Johnson DK, and Hodgen GD (1976) hi vitro evaluation of corpus luteum function of cycling and pregnant rhesus monkey: progesterone production by dispersed luteal cells. Steroids 27:1106-1113.

Strich G, Petze JE, Silva de Sa MF, and Rebar RW (1985) The effects of thymus-derived peptides on hypothalamic LRF and pituitary gonadotropin content in prepubertal congenitally athymic nude mice and their normal heterozygous littermates. J Reprod Immunol 7:351-359.

Sun T-T, Bonltz P, and Bums WH (1984) Cell culture of mammalian thymic epithelial cells: Growth, structural, and antigenic properties. Cell Immunol 83:1-13.

Taguchi O, and Nishizuka Y (1980) Autoimmune oophoritis in thymectomized mice: T cell requirement in adoptive cell transfer. Clin Exp Immunol 42:324-331.

Taguchi O, Nishizuka Y, Sakakura T, and Kojima A (1980) Autoimmune oophoritis in thymectomized mice: Detection of circulating antibodies against oocytes. Clin. Exp Immunol 40:540-553.

Tang BK, Armstrong DT, and Whitfield JF (1980) Steroid biosynthesis by isolated human ovarian follicular cells in vitro. J Clin Endocrinol Metab 51: 1407-1411.

Trzeciak WH, Duda T, Waterman MR and Simpson ER (1987) Effects of epidermal growth factor on the synthesis of the cholesterol side-chain cleavage enzyme complex in rat ovarian granulosa cells in primary culture. Mol Cell Endocrinol 52:43-50.

Uziimcu M and Lin YC (1994) Characterization of the stimulatory actions of thymic factor(s) on basal and gonadotropin-induced steroidogenesis in cultured rat granulosa cells. Mol Cel Endocrinol 105:209-216.

Uzumcii M, Akira S and Lin YC (1992) Stimulatory effect of thymic factor(s) on steroidogenesis in cultured rat granulosa cells Life Sci 51:1217-1228.

Uziimcu M, and Lin YC (1990) Factor(s) from rat thymic cell culture conditioned medium (TCM) stimulate(s) progesterone secretion (P) in cultured rat granulosa cells. Biol Reprod 42: Suppl. 1,163 (Abstr.).

Uziimcii M, Brigstock DR and Lin YC (1994) Rat thymic epithelial cell culture conditioned medium (TCM) contains two distinct steroidogenic factors. Biol Reprod 50: Suppl # 1, p74. 170

Wang S-S, Kulasha ID, and Winter DP (1978) Synthesis of thymosin a 1. J Am Chem Soc 101:253-254.

Wang C, Hsueh AJW, and Erickson GF (1981) LH stimulation of estrogen secretion in cultured granulosa cells. Mol Cell Endocrinology 24:17-28.

Wiebe JP (1976) Steroidogenesis in rat leydig cell: Changes in activity of 5-ane and 5-ene 3fi-hydroxysteroid dehydrogenase during sexual maturation. Endocrinology 98:505-513.

Wolfe MW, Vakharia DD, Wise TH and Kinder JE (1989) Characterization of secretion of thymosin alpha 1 and thymosin beta 4 during prepuberty, estrus and pregnancy in the bovine female. Domestic Anim Endocrinol 6:71-78

Zolti M, Meirom R, Shemesh M, Wolach D, Mashiach S, Shore, L and Rafael ZB (1990). Granulosa cells as a source and target organ for tumor necrosis factor-alpha. FEBS Letters 261: 253-255.