Regulation of Ovarian Progestin Production by Epidermal Growth Factor in Cultured Rat Granulosa Cells*

The modulation of ovarian steroidogenesis by epider- mal growth factor (EGF) was investigated in cultured rat granulosa cells. Granulosa cells, obtained from ova- ries of immature, hypophysectomized, estrogen-treated rats, were incubated for 2 days with EGF, follicle-stim- ulating hormone (FSH), or EGF plus FSH. Treatment with EGF did not affect estrogen production, but stim- ulated progestin (i.e. progesterone and 20a-hydroxy-pregn-4-en-3-one) production in a dose-dependent man- ner. Stimulation of progestin production by EGF appears to be the result of an increase in pregnenolone biosynthesis as well as increases in the activities of 20a-hydroxysteroid dehydrogenase and 3jLhydroxy- steroid dehydrogenase/isomerase. Treatment with FSH increased both estrogen and progestin production by cultured granulosa cells. When cells were treated concomitantly with EGF, FSH-stimulated estrogen pro- duction was inhibited, while progestin production was further enhanced. The EGF enhancement of FSH-stim- ulated progestin production appears to be the result of synergistic increases in pregnenolone biosynthesis and 20a-hydroxysteroid dehydrogenase activity, resulting in substantial increases in 20a-hydroxypregn-4-en-3- one but not progesterone production. The effects of EGF were shown to be time-dependent.

and Schomberg (5) reported that EGF inhibits the FSHinduced increase in luteinizing hormone receptor content of cultured rat granulosa cells. Similarly, we have shown that EGF inhibits the human chorionic gonadotropin-induced increase in testosterone production by cultured rat Leydig cells (41, while Ascoli (6) has reported that EGF decreases luteinizing hormone receptor content in a clonal strain of murine Leydig tumor cells. Thus, EGF modulates steroidogenesis and gonadotropin receptor content in cultured gonadal cells.
The present investigation extends our earlier studies on the EGF modulation of granulosa cell functions. Specifically, we have characterized the granulosa cell EGF receptor and examined the effect of EGF on progestin and estrogen biosyntheses.
Granulosa Cell Cultures-Immature female Sprague-Dawley rats (21-23 days old) were hypophysectomized by Curtis Johnson Laboratories, Chicago, IL, and delivered on the third postoperative day. Silastic capsules (10 mm) containing diethylstilbestrol were implanted at the time of surgery. Hypophysectomized animals were given a mixture of bread, milk, tap water, and physiological saline (0.9% NaCl solution) ad libitum.
Four to six days after surgery, granulosa cells were obtained from the hypophysectomized rats as described previously (7). These cells were cultured in Falcon tissue culture dishes (35 X 10 m m ) (2-4 X lo5 viable cells/dish) in 1 ml of McCoy's 5a medium supplemented with 2 mM L-glutamine, 100 units/ml of penicillin, 100 p g / d of strepto-11268 mycin sulfate, and 100 n~ androstenedione. Androstenedione was included in the culture medium to serve as the aromatase substrate and to synergize with FSH in stimulating progesterone production (8,9,10). FSH and EGF were diluted in sterile culture medium and added in 50-pl aliquots. Cells were cultured at 37 "C in a humidified 95% air, 5% CO, incubator. At the end of the incubation, media were collected and stored at -20 "C until analyzed for progesterone, 20a-OH-P, and estrogen contents by radioimmunoassays.
To examine pregnenolone production, granulosa cells were incubated for 2 days with the appropriate hormones. Media were then replaced with fresh medium containing M cyanoketone, an inhibitor of 3P-hydroxysteroid dehydrogenase ( l l ) , and the cells were incubated for 25 min. After the incubation, the appropriate hormones were added back to the cultures and the cells were reincubated for 4 h. Media were analyzed for pregnenolone contents by radioimmunoassay.
Radioimmunoassays-Medium progesterone and estrogen contents were measured using specific antisera (12). Medium 20a-OH-P contents were measured using specific antisera supplied by Ralph Schwall, University of California, San Diego. The 20a-OH-P antiserum 316 cross-reacts < 0.1% with progesterone, CIS, CIH, and other steroids. Medium pregnenolone contents were measured with specific antisera supplied by Dr. P. Chrousos, National Institutes of Health, Bethesda, MD; this antiserum cross-reacts <0.4% with progesterone, 20a-OH-P, 17a-hydroxyprogesterone, lla-hydroxyprogesterone, estradiol, and testosterone. High concentrations M) of diethylstilbestrol, androstenedione, or cyanoketone in the culture medium do not interfere with the radioimmunoassays.
Enzyme Assays-The assay of 2Oa-hydroxysteroid dehydrogenase (EC 1.1.1.149) activity was based on a procedure developed by Eckstein et al. (13) and modified by us (14) in which 20a-hydroxysteroid dehydrogenase activity was measured as the rate of conversion of The assay of 3P-hydroxysteroid dehydrogenase/steroid A-isomerase (EC 1.1.1.51/EC 5.3.3.1) activity was based on a procedure developed by Murono and Payne (15) in which 3P-hydroxysteroid dehydrogenase activity was measured as the rate of conversion of pregnenolone to progesterone. Our modification of this procedure has been described elsewhere (16). E G F Receptor Assay-EGF was iodinated using a method developed by Carpenter and Cohen (17). The reaction mixture contained 5 pg of EGF in 20 pl of distilled water, 1 mCi of carrier-free Na"'I in 30 pl of 0.5 M sodium phosphate buffer (pH 7.51, and 20 pg of chloramine-?' in 10 p1 of distilled water. The reaction was stopped in 40 s by the addition of 40 pg of sodium metabisulfite in 10 p1 of distilled water. After adding 500 pl of the elution buffer (0.05 M sodium phosphate buffer with 0.075 M sodium chloride, pH 7.5), the mixture was passed through a Sephadex G-25 column to separate Iz5I-EGF from free iodine. The labeled EGF was stored at -20 "C in the presence of 0.1% bovine serum albumin. The specific activity of Iz5I-EGF, measured by self-displacement with specific EGF antiserum, was 143 f 17 pCi/pg ( N = 3), or -130,000 to 190,000 cpm/ng. The maximum binding capacity of the tracer in the presence of excess EGF receptors was determined to be 30 f 2% (N = 3), as measured by the radioreceptor assay using increasing amounts of rat liver homogenate (18).
In uiuo FSH treatment increases ""I-EGF binding capacity in granulosa cells by -2-fold (see "Results"). Thus, granulosa cells were obtained from the ovaries of immature, hypophysectomized, diethylz D r " D L stilbestrol-treated rats which had been injected subcutaneously twice daily for 2 days with FSH (6.7 pg/O.1 ml of saline). In addition, samples of liver, kidney, spleen, and cerebellum were obtained to examine the tissue specificity of '"I-EGF binding. The granulosa cells or minced tissues were briefly homogenized (-60 s) in ice-cold Dulbecco's phosphate-buffered saline (pH 7.2) using a glass-Teflon homogenizer. Three aliquots of cell suspension or tissue homogenate were incubated in tubes coated with bovine serum albumin for 16 h at 22 "C with '*'I-EGF (-180,000 cpm) with or without excess (50 ng) unlabeled EGF. At the end of the incubation, the reaction mixtures were diluted with 2.5 ml of ice-cold buffer and centrifuged at 1,500 X g for 30 min at 4 "C. After the pellet was resuspended with the buffer and recentrifuged, radioactivity in the pellet was determined by counting in a y-spectrometer. After correcting for the radioactivity bound in the presence of excess unlabeled EGF, the amount of specifically bound EGF was determined. Less than 0.1% of the tracer bound to tubes in the absence of tissue.
To determine the dissociation constant ( K d ) of the ovarian EGF receptors, the radioreceptor assay was performed as described above. Ovarian granulosa cells were incubated with -90,OOO cpm of '2sI-EGF in the presence or absence of varying amounts of unlabeled EGF. After determination of specific binding, the data were plotted in a Scatchard plot (18). The slope and intercept of the resultant line were calculated using a linear regression program.
Protein and DNA Determinations-The Bio-Rad protein assay (Bio-Rad Co.) was used to measure protein content (19) using bovine y-globulin as the standard. DNA content was determined according to the method of Burton (20).
Statistical Analyses-A four-parameter logistic curve-fitting program was used to obtain dose-response curves (21). A linear regression program was used to approximate the Scatchard plot. Other statistical analyses were performed using the Student's t test. Comparisons with p 2 0.05 were not considered significant.

Effects of EGF and FSH on
Progestin and Estrogen Production by Cultured Granulosa CeZls-To examine the effect of EGF on ovarian steroidogenesis, granulosa cells were incubated in medium alone (control), or with increasing concentrations of EGF, FSH (10 ng/ml), or EGF plus FSH. The medium concentration of 20a-OH-P was low in control cells while FSH treatment stimulated 20a-OH-P production from <0.3 ng/ml to 24.9 ng/ml (Fig. lA). Similarly, EGF treatment stimulated 20a-OH-P production in a dose-dependent manner with 10 ng of EGF/ml stimulating 2Oa-OH-P production to 14.3 ng/ml. Furthermore, concomitant treatment with FSH and increasing concentrations of EGF resulted in a synergistic stimulation of 2Oa-OH-P production. The synergistic effect of EGF was dose-dependent with an EDSo value of -0.4 ng of EGF/ml.
In the same cultures, medium progesterone was low (<0.2 ng/ml) in control cells, while FSH treatment stimulated progesterone production by >80-fold (Fig. 1B). In  and with a maximal stimulation of -3 ng/ml of progesterone. In contrast to the synergistic action of EGF on FSH-stimulated 20a-OH-P production, treatment with increasing concentrations of EGF did not affect FSH-stimulated progesterone production. FSH treatment increased estrogen production from -0.09 ng/ml to 2.54 ng/ml (Fig. IC) while treatment with EGF alone did not affect estrogen production. Concomitant treatment with EGF inhibited the stimulatory effect of FSH in a dosedependent manner and the EDs0 value was determined to be 0.206 ng of EGF/ml. High concentrations of EGF inhibited FSH-stimulated estrogen production by -70%.
Time Course of Action of EGF and FSH upon Steroidogenesis by Cultured Granulosa Cells-To examine the time course of action of EGF and FSH, granulosa cells were incubated with medium alone (control), EGF (3 ng/ml), FSH (10 ng/ml), or EGF plus FSH. Media from sets of quadruplicate cultures were removed throughout a 48-h incubation period to study the time-dependent changes in the accumulation of 20a-OH-P, progesterone, and estrogen in the media. As shown in Fig. 2.4, treatment with either FSH or EGF for 48 h stimulated 20a-OH-P production to 33.5 or 10.1 ng/ml, respectively. The synergistic stimulation of 2Oa-OH-P production by concomitant FSH and EGF treatment was evident after 34 h of treatment.
In the same cultures, FSH treatment stimulated progesterone production from 2.11 ng/ml at 10 h of incubation to 36.4 ng/ml at 48 h. In a similar temporal pattern, treatment with EGF alone resulted in an increase in progesterone production to 2.65 ng/ml after 48 h. In contrast, concomitant treatment with EGF did not further augment FSH action.
As shown in Fig. 2C, the concentration of medium estrogen in FSH-treated cells increased from t0.1 ng/ml after 10 h of incubation to 3.85 ng/ml after 48 h of incubation. In contrast, an EGF inhibition of FSH-stimulated estrogen production was observed after 24 h of incubation, resulting in an 84% decrease in estrogen production at 48 h. Treatment with EGF alone did not affect estrogen biosynthesis.

Effects of EGF and FSH on Pregnenolone Production by Cultured Granulosa
Cells-The observation that concomitant treatment with EGF and FSH results in an increase in 20a-OH-P production with no alteration in progesterone production suggested that EGF may increase pregnenolone biosynthesis. To examine this possibility, granulosa cells were incubated in medium alone (control), with increasing concentrations of EGF, FSH (10 ng/ml), or EGF plus FSH for 2 days. Media were replaced and cells reincubated for 4 h in fresh medium containing appropriate hormones plus cyanoketone to inhibit 3P-hydroxysteroid dehydrogenase activity. As shown in Fig. 3, FSH treatment stimulated pregnenolone production by >120-fold. Treatment with EGF alone also resulted in a dose-dependent increase in pregnenolone production with an EDm value of 0.9 ng of EGF/ml and a maximal stimulation of -30 ng of pregnenolone/ml. Concomitant treatment with FSH and increasing concentrations of EGF resulted in a synergistic stimulation of pregnenolone production. This effect was dose-dependent with an EDm value of -0.3 ng of EGF/ml. (16), we demonstrated that FSH treatment of cultured granulosa cells increases the activity of 3P-hydroxysteroid dehydrogenase, which converts pregnenolone to progesterone. To examine the effect of EGF on 3P-hydroxysteroid dehydrogenase activity, granulosa cells were incubated for 2 days in medium alone (control), with FSH (10 ng/ml), increasing concentrations of EGF, or FSH plus EGF. The enzyme activities of the treated cells were determined as described under "Experimental Procedures." As shown in Fig. 4, FSH treatment increased enzyme activity by -4-fold. Likewise, treatment with increasing concentrations of EGF resulted in a dose-dependent increase in 3P-hydroxysteroid dehydrogenase activity with an ED5o of 0.843 ng of EGF/ml and a maximal stimulation of 4.7-fold as compared to the untreated controls. Furthermore, treatment with increasing concentrations of EGF in the presence of FSH resulted in a maximal 30% increase in enzyme activity as compared to the FSH-treated cells.

Effects of EGF and FSH on 2Oa-Hydroxysteroid Dehydrogenase Activity of Cultured Granulosa
Cells-The observation that concomitant treatment with EGF and FSH results in synergistic increases of both pregnenolone and 20a-OH-P biosynthesis with no alteration in progesterone production suggested that EGF treatment may increase the activity of 20a-hydroxysteroid dehydrogenase, which converts progesterone to 20a-OH-P. To examine this possibility, granulosa cells were incubated for 2 days with medium alone (control), FSH (10 ng/ml), increasing concentrations of EGF, or FSH plus EGF. The enzyme activities of the treated cells were determined as described under "Experimental Procedures." AS shown in Fig. 5, FSH treatment resulted in a slight increase in enzyme activity. In contrast, treatment with increasing concentrations of EGF alone resulted in a dose-dependent increase in 20a-hydroxysteroid dehydrogenase activity with an EDm value of 0.75 ng of EGF/ml. When cultures were treated with FSH and high concentrations of EGF (1 to 10 ng of EGF/ml), further synergistic increases in 20a-hydroxysteroid dehydrogenase activity were detected. CONCtNTRATlON Of EGf Inytmll   FIG. 3 (left). Effects of EGF and FSH upon pregnenolone production. Granulosa cells (2 X lo5 viable cells/dish) were cultured in medium alone (control), with FSH (10 ng/ml), increasing concentrations of EGF, or FSH plus EGF. Pregnenolone production was evaluated as described under "Experimental Procedures." Data points represent mean & S.E. of four replicate cultures.

Effect of EGF and FSH on Cellular Protein and DNA Content of Cultured Granulosa Cells-To examine the possible mitogenic effects of EGF and FSH on granulosa cells, cells were incubated for 2 days in medium alone (control), or
FIG . 4 (center). Effects of EGF and FSH on 3p-hydroxysteroid dehydrogenase activity of cultured granulosa cells. Granulosa cells (4 X lo5 viable cells/dish) were cultured for 2 days in medium alone (control), with FSH (10 ng/ml), increasing concentrations of EGF, or FSH plus EGF. Enzyme activities were determined as Effects of EGF and FSH on protein and DNA contents of cultured granulosa cells. Granulosa cells were incubated in medium alone (control), with EGF (5 ng/ml), FSH (10 ng/ml), or EGF plus FSH. After 2 days, media were removed, cells were scraped from the dishes, and protein and DNA contents were determined as described under "Experimental Procedures." Left, data points represent mean -t S.E. for four samples (-2 X lo5 viable cells/dish). Right, data points represent mean k S.E. for three cultures (-lo6 viable cells/dish).
with EGF (5 ng/ml), FSH (10 ng/ml), or EGF plus FSH. As shown in Fig.  6  While several other incubation temperatures were examined (4 and 37 "C; data not shown), maximal specific binding in this system occurred at 22 "C. Thus, the EGF radioreceptor assays were performed at 22 "C for 16 h. Since in vivo FSH treatment increases the granulosa cell EGF binding capacity (picograms of lZ5I-EGF bound/granulosa cell, controls = 11.9 described under "Experimental Procedures." Data pomts represent mean f S.E. of three determinations after pooling triplicate cultures. FIG . 5 (right). Effects of EGF and FSH on 20a-hydroxysteroid dehydrogenase activity of cultured granulosa cells. Granulosa cells (-4 x lo5 viable cells/dish) were cultured for 2 days in medium alone (control), with FSH (10 ng/ml), increasing concentrations of EGF, or FSH plus EGF. Enzyme activities were determined as described under "Experimental Procedures." Data points represent mean f S.E. for three determinations after pooling triplicate cultures. In several cases, the variation was less than the symbols indicated.

PROTEIN (pgltube)
FIG. 7. Determination of optimal incubation time and tissue concentration for the EGF binding assay. Granulosa cells were obtained from immature, hypophysectomized, estrogen-treated rats after 2 days of FSH treatment in uiuo. A, the cells were incubated with labeled EGF in the absence or presence of excess unlabeled EGF for various intervals at 22 "C. B, increasing numbers of cells were incubated with labeled EGF in the absence or presence of unlabeled EGF for -16 h at 22 "C, as described under "Experimental Procedures." Data points represent mean f S.E. for three samples. f 0.9; FSH-treated = 23.0 k 0.5, N = 5 ) , a l l assays used FSHtreated granulosa cells.
To determine the optimal tissue concentrations for EGF binding, various aliquots of a granulosa cell preparation were incubated with labeled EGF in the presence or absence of excess unlabeled EGF. As shown in Fig. 7B, specific EGF binding was linear from 26 to 154 pg of protein; the correlation coefficient associated with the line was calculated to be 0.996.
All experiments with granulosa cells employed tissue concentrations within this range.
Characterization of EGF Binding-To examine the tissue specificity of EGF binding, tissue homogenates were prepared and specific binding was determined as described under "Experimental Procedures." Specific EGF binding of hornogenates from ovary, liver, and kidney tissues were determined to be 34 f 4, 627 f 32, and 22 & 5 pg/mg of protein, respectively (n = 6). In contrast, homogenates of spleen and cerebellum did not bind 12'I-EGF in a specific manner.
To determine the dissociation constant ( K d ) of the granulosa cell EGF receptor, granulosa cells were incubated with  labeled EGF in the presence or absence of increasing concentrations of unlabeled EGF as described under "Experimental Procedures." Scatchard analysis (Fig. 8) of the displacement curve indicated that the Kd of the granulosa cell EGF binding site is 2.74 f 0.51 X 10"' M or 1.67 f 0.31 n g / d (n = 3) and the binding capacity was determined to be 4950 f 930 binding sites/cell (n = 3). To further examine the specificity of EGF binding, granulosa cells were incubated with labeled EGF alone or in the presence of excess unlabeled EGF or various other peptides and protein hormones. As shown in Table I, only unlabeled EGF significantly displaced "'I-EGF from the granulosa cell binding sites, whereas co-incubation with fibroblast growth factor, nerve growth factor, and other hormones did not affect 12'I-EGF binding.

DISCUSSION
The present results demonstrate that 1) EGF binds to rat granulosa cells with high affinity and specificity; 2) in contrast to the inhibitory effect of EGF on FSH-stimulated estrogen production, EGF enhances FSH-stimulated 20a-OH-P production in a synergistic manner (this stimulatory effect appears to be related to an increase in pregnenolone biosynthesis and the stimulation of 20a-hydroxysteroid dehydrogenase activity); 3) EGF treatment alone stimulates 20a-OH-P and progesterone production, as well as 2Oa-hydroxysteroid dehydrogenase and 3b-hydroxysteroid dehydrogenase activities; and 4) while EGF is not mitogenic in the cultured rat granulosa cells, EGF treatment increases the protein content of the granulosa cells.
EGF modulates FSH action in a disparate manner in cultured granulosa cells. EGF inhibits FSH-stimulated increases in estrogen production (4) and luteinizing hormone receptor content (5), while stimulating FSH-induced 20a-OH-P production synergistically (Figs. 1 and 2). These divergent effects of EGF on FSH-stimulated steroidogenesis indicate that estrogen production and progestin production are regulated through different pathways. The notion that EGF can exert divergent effects upon endocrine cells is also supported by the observation that EGF treatment results in an increase in prolactin synthesis and a decrease in growth-hormone synthesis in a cultured pituitary tumor cell line (22).
The EGF enhancement of FSH-stimulated 20a-OH-P production is not accompanied by apparent changes in progesterone production but rather appears to be the result of alterations in several steroidogenic enzymes. This is characterized by a synergistic stimulation of pregnenolone biosynthesis and a synergistic increase in 20a-hydroxysteroid dehydrogenase activity. Furthermore, EGF treatment enhanced FSH-stimulated 3P-hydroxysteroid dehydrogenase activity by -30%.
In the present study, treatment with EGF alone was shown to stimulate progesterone and 20a-OH-P (but not estrogen) production by cultured granulosa cells. These observations strengthen the concept that the cellular mechanisms controlling estrogen and progestin production are regulated differently in rat granulosa cells. The EGF-induced increase in progestin production appears to be the result. of increased pregnenolone biosynthesis, as well as increases in the activities of 20a-hydroxysteroid dehydrogenase and 3P-hydroxysteroid dehydrogenase. EGF-stimulated pregnenolone production may be the result of increased side chain cleavage activity, although one cannot exclude the possibility that EGF stimulates enzymes involved in cholesterol biosynthesis. The observation that EGF stimulates progestin biosynthesis in granulosa cells from these pre-antral follicles places EGF in a unique category with FSH because only FSH and several CAMPinducing agents (such as cholera toxin and prostaglandin E2) are capable of enhancing progestin production in these rela-tively undifferentiated cells (23). Also, it is highly unlikely that the stimulatory effects of the EGF preparation are due to contamination with FSH since EGF induces disparate effects on granulosa cell steroidogenesis, while FSH enhances both estrogen and progestin biosynthesis. The physiological role of EGF in the differentiation of granulosa cells awaits further study.
Rat granulosa cells were shown to bind EGF specifically and with high affinity. The calculated K d value (2.74 X 10"" M) for EGF binding to the rat granulosa cells is similar to that reported by Vlodavsky et al. (24) in bovine granulosa cells (2.4 X 10"' M). Thus, EGF directly modulates steroidogenesis of rat granulosa cells by altering the activities of steroidogenic enzymes and these actions of EGF are probably mediated through the high affinity EGF binding sites. EGF may bind to granulosa cell receptors to induce changes in intracellular calcium levels (25) or a CAMP-independent phosphorylation of membrane components (26-29). Recently, a direct interaction of EGF with the cell nucleus has been reported (30, 31). The rat grandosa cell culture system provides a valuable model for further examination of the mechanism of EGF action.
Early studies reported that EGF is a mitogen for cultured bovine, porcine, and human granulosa cells, but not for cultured rat granulosa cells (32). Our results (Fig. 6) concur with the latter observation. In addition to the growth-promoting function of EGF, an endocrine role for this peptide is becoming evident. In uiuo treatment of neonatal female rats with EGF suppresses ovarian growth and follicular development (33). This is consistent with the findings that EGF inhibits FSH stimulation of aromatases (Fig. IC) and LH receptor content (5) in cultured granulosa cells. In addition to the present finding of an EGF stimulation of progestin biosynthesis, EGF treatment also stimulates progesterone production by human choriocarcinoma (34) and rat ovarian tumor cells (35) as well as increases ornithine decarboxylase activity of cultured porcine granulosa cells (36).
EGF has been found in various biological fluids (1-3). Byyny et al. (37) have reported that the concentration of EGF in the plasma of female mice is -1.2 ng/ml, a concentration which is effective in the present rat granulosa cell culture system. Although EGF acts as a mitogen for a variety of cell types, our results demonstrate that this "growth factor" regulates steroidogenesis of rat granulosa cells. Thus, EGF may play an important endocrine role in controlling ovarian development by regulating the differentiation and steroidogenesis of granulosa cells.