T Vänttinen 1, J Liu 2, C Hydén-Granskog 3 and R Voutilainen 1,2

Transcription

1 557 Biphasic regulation of activin A secretion by gonadotropins in cultured human ovarian granulosa-luteal cells leads to decreasing activin:inhibin ratios during continuing gonadotropin stimulation T Vänttinen 1, J Liu 2, C Hydén-Granskog 3 and R Voutilainen 1,2 1 Department of Pediatrics, Kuopio University and University Hospital, Kuopio, Finland 2 Department of Pathology, Haartman Institute, University of Helsinki, Helsinki, Finland 3 Department of Obstetrics and Gynecology, Helsinki University Central Hospital, Helsinki, Finland (Requests for offprints should be addressed to R Voutilainen, Department of Pediatrics, Kuopio University Hospital, PO Box 1777, FIN Kuopio, Finland; Raimo.Voutilainen@uku.fi) Abstract Pituitary gonadotropins mediate part of their effects on ovarian function via local hormones and growth factors produced by granulosa cells. Activins and inhibins are among these factors, and they have often opposite effects on various components of the reproductive system. The purpose of this study was to investigate the regulation of ovarian activin A secretion using cultured human ovarian granulosa-luteal cells as a model. The granulosa-luteal cells, obtained from women taking part in an in vitro fertilization program, were cultured and treated with FSH, LH, 8-bromo camp (8-BrcAMP, a protein kinase A activator) and 12-O-tetradecanoyl phorbol-13-acetate (TPA, a protein kinase C activator). Conditioned cell culture media were analyzed for activin A, inhibin A and progesterone concentrations with specific enzyme immunoassays. FSH and LH (1 100 IU/l) increased activin A secretion with 24 h of treatment (to 132% and 253% of control respectively; P<0 05 for both), but their effects were inhibitory in 48-h treatments (26% and 16% decreases respectively; P<0 05 for both). In the same experiments, FSH and LH increased inhibin A and progesterone secretion after both 24 and 48 h of treatment. 8-BrcAMP ( µm) increased activin A in 24- and 48-h experiments (to 206% and 148% of control respectively; P<0 01 for both). Inhibin A and progesterone secretion were stimulated by 8-BrcAMP time- and dose-dependently. TPA increased activin A secretion dose-dependently ( ng/ml) in both 24- and 48-h experiments. At 100 ng/ml concentration, it increased activin A up to 61-fold and inhibin A up to 16-fold of control in 24-h experiments. We conclude that gonadotropins regulate immunoreactive activin A secretion biphasically in cultured human granulosa-luteal cells: initial stimulation is followed by inhibition. In contrast, gonadotropins increase inhibin A and progesterone secretion continuously. Consequently, continuing gonadotropin stimulation leads to a decreasing activin:inhibin ratio, which may have a significant role in the local fine-tuning of ovarian steroidogenesis. Introduction Activins and inhibins are structurally related dimeric glycoproteins consisting of subunits (α, βa or βb), each coded by a separate, differentially regulated gene (reviewed in Ying 1988, Hillier 1991, Knight & Glister 2001). Activins are dimers of β-subunits (βa:βa, βa:βb and βb:βb for activin A, activin AB and activin B respectively). Inhibins have a common α-subunit connected with either of the β-subunits (α:βa and α:βb for inhibin A and inhibin B respectively). In the human ovary, the synthesis of activins and inhibins occurs mainly in the granulosa cells, where the genes for the activin and inhibin subunits are expressed throughout the menstrual cycle (Roberts et al. 1993). In preovulatory and early luteal granulosa cells, the α- and βa-subunits are highly expressed, whereas the expression of the βb-subunit is absent, which suggests that activin A and inhibin A are the most important activin/inhibin forms during these stages. Gonadotropins up-regulate α- and βa-subunit expression (Erämaa et al. 1994) and activin A and inhibin A peptide secretion in human granulosa-luteal cells (Muttukrishna et al. 1997, Vänttinen et al. 2000). Activins and inhibins have often opposite effects on ovarian function. Hypotheses as to the para/autocrine role of activins and inhibins in human ovaries are based on in vitro studies. In cultured human follicular cells, activin A modulates steroidogenesis by inhibiting progesterone secretion and aromatase activity (Rabinovici et al. 1992) /02/ Society for Endocrinology Printed in Great Britain Online version via

2 558 TVA}NTTINEN and others Activin A in granulosa cells Activins inhibit androgen production in cultured human theca cells, whereas inhibins stimulate it (Hillier et al. 1991a,b). The mechanisms of the antagonistic actions of inhibins and activins have not been understood until recently, when inhibins were shown to be able to block activin receptors via the transforming growth factor-β type III receptor, betaglycan (Lewis et al. 2000, Matzuk 2000). The purpose of the present study was to examine the dynamics of gonadotropin-regulated activin A secretion and to compare it with that of inhibin A and progesterone secretion, to gain a better understanding of the endocrine regulation of the activin/inhibin system in the human ovary. We found that gonadotropins had a biphasic effect on activin A secretion initial stimulation was followed by inhibition whereas they increased inhibin A and progesterone secretion continuously. Materials and Methods Ethical considerations The study was approved by the Research Ethics Committees of the Kuopio and Helsinki University Hospitals, and the patients gave informed written consent to participate. Cell cultures Human granulosa cells were obtained from women undergoing an in vitro fertilization (IVF) program in Helsinki University Hospital as described previously (Vänttinen et al. 2000, Liu et al. 2001). Before follicular aspiration, the patients were treated with a gonadotropin-releasing hormone analog to suppress endogenous gonadotropin secretion. Follicle-stimulating hormone (FSH) was used to induce the development of several follicles. Follicular aspirations were performed after h of administration of a single dose of human chorionic gonadotropin (hcg). After follicular aspiration, the cumulus oocyte complexes were separated for IVF and the granulosa cells from all follicles of several women were pooled, pelleted and transported immediately to the laboratory. The granulosa-luteal cells were purified and cultured as described previously (Vänttinen et al. 2000). In brief, the cells were dispersed and separated from red blood cells, washed and plated with a medium containing DMEM Ham s F12 (1:1) supplemented with 10% fetal calf serum (Gibco or Bioclear UK Ltd, Calne, Wilts, UK), 2 mm L-glutamine, 100 IU/ml penicillin, and 100 µg/ml streptomycin sulfate (Gibco) at a density of cells/ well on 35-mm six-well dishes. The cells were grown at 37 C in a 95% air 5% CO 2 humidified environment, and the culture media were changed every 2 3 days. All hormonal and other treatments were performed during the 7th to 10th days of culture, when the cells are the most responsive in this culture system (Erämaa et al. 1994, Voutilainen et al. 1986). At this stage the cultures were subconfluent. The functional viability of the cells was assessed by measuring progesterone secretion by the cells. Recombinant human FSH and luteinizing hormone (LH) were gifts from Serono (Vantaa, Finland). 8-Bromo camp (8-BrcAMP) and 12-O-tetradecanoyl phorbol-13-acetate (TPA) were purchased from Sigma (St Louis, MO, USA). Hormone assays The concentrations of total (free and follistatin-bound) activin A in culture media were measured by an ultrasensitive ELISA intended for use with various human and animal fluids, including culture media. The method has been described in detail previously (Knight et al. 1996). The activin A assay kit (product code MCA 1426 KZZ) was purchased from Serotec (Oxford, UK). The detection limit of the assay was 50 pg/ml. The intra- and interassay coefficients of variation were 6 1 and 6 5% respectively. All samples were assayed in duplicate. According to the manufacturer, there is no detectable cross-reaction with other forms of activins or inhibins. Inhibin A was measured with an ultrasensitive two-site immunosorbent assay by a technique similar to that used for activin A. The method has been described in detail previously (Groome et al. 1994, Muttukrishna et al. 1994). The inhibin A assay kit (product code MCA 950 KZZ) was purchased from Serotec. The limit of detection of the assay was 2 pg/ml. The intra- and interassay coefficients of variation were 2 4 and 7 6% respectively. There should not be any detectable cross-reaction with inhibin B or activins. Progesterone was measured by a competitive enzyme immunoassay purchased from Diagnostic Systems Laboratories, Inc. (Webster, Texas, USA; product code DSL ). The limit of detection of the assay was 1 nmol/l. The intra- and interassay coefficients of variation were reported to be 7 5% and 9 4% respectively. Statistical analyses Each experiment was repeated at least three times with different cell batches. Single experiments consisted of several hormonal manipulations each performed in two or three parallel wells. The data are presented as arithmetic means S.E.M., with the control mean adjusted to 100%. The statistical significances were estimated by the nonparametric Mann Whitney test or the Kruskal Wallis test in the multiple comparisons. Results Basal activin A, inhibin A and progesterone secretion by cultured granulosa-luteal cells Cultured granulosa-luteal cells secreted activin A, inhibin A and progesterone with different patterns, depending on

3 Activin A in granulosa cells TVA}NTTINEN and others 559 Figure 1 Concentrations (logarithmic scale) of activin A (black bars), inhibin A (white bars) and progesterone (shaded bars) in conditioned culture media of human granulosa-luteal cells in relation to the period of culture. The cells were cultured for 12 days. Culture media were replaced every third day and analyzed for activin A, inhibin A and progesterone concentrations. The bars represent mean+s.e.m. hormone concentrations (ng/ml per 72 h) from four separate experiments. All hormone concentrations changed significantly during the 12-day culture period (P<0 01; Kruskal Wallis test). the culture age (Fig. 1). Activin A secretion followed a U-shaped pattern, increasing clearly between days 7 12 in culture. The patterns of secretion of inhibin A and progesterone were similar to each other, decreasing gradually after 6 days of culture. The activin A:inhibin A ratio increased 10-fold from days 7 9 to of culture (P<0 05; Mann Whitney test), whereas the sum of activin A and inhibin A secretion did not change at this stage. Effect of gonadotropins on activin A, inhibin A and progesterone secretion FSH and LH had similar effects on the secretion of activin A, inhibin A and progesterone. In preliminary experiments a dose-dependent stimulatory effect of these gonadotropins (1 100 IU/l) on activin A secretion was measurable after 6 h, reaching a maximum after approximately 24 h, and diminishing rapidly after 36 h in culture (data not shown). On the basis of the preliminary experiments, gonadotropin concentrations of 100 IU/l were used in the following experiments. Incubation of granulosa-luteal cells with FSH (Fig. 2A) increased activin A secretion in 24-h experiments ( % of control; P<0 05), but in 48-h experiments the effect of FSH was inhibitory ( % of control; P<0 05). Inhibin A and progesterone secretion were increased in the presence of FSH after 24 h ( % and % of control respectively; P<0 01 for both) and 48 h ( % and % of control respectively; P<0 01 for both) of incubation. LH (Fig. 2B) increased activin A secretion in 24-h experiments ( % of control; P<0 05), but inhibited it in 48-h experiments ( % of control; P<0 05). Inhibin A and progesterone secretion were increased in the presence of LH after 24 h ( % and % of control respectively; P<0 01 for both). After 48 h, the effect of LH on inhibin A and progesterone secretion was less pronounced: inhibin A was increased up to % of control (P=0 663) and progesterone up to % of control (P<0 01). The inhibitory effect of gonadotropins on activin A secretion was confirmed by performing five 48-h experiments with different doses of FSH (1 100 IU/l) and LH (1 100 IU/l) (Fig. 2C and D). FSH proved to have more significant effects than LH, but the pattern of their dose-dependent effects was similar. Low concentrations of FSH (1 IU/l) increased activin A secretion but higher concentrations ( IU/l) tended to inhibit it (P=0 08; Kruskal Wallis test), whereas inhibin A secretion was increased dose-dependently (P<0 05). Low concentrations of LH (1 10 IU/l) stimulated activin A secretion, but high concentrations (100 IU/l) were inhibitory (P<0 05). Effect of 8-BrcAMP and TPA on activin/inhibin A secretion In preliminary experiments, 8-BrcAMP (an activator of protein kinase A) increased the secretion of activin A, inhibin A and progesterone dose dependently ( µm) in both 24 and 48 h treatments (dose response data not shown). 8-BrcAMP (100 µm; Fig. 3) stimulated activin A secretion in the 24-h experiments ( % of control; P<0 01) and in the 48-h experiments, but in the latter the stimulation was less ( % of control; P<0 01). The secretion of inhibin A and progesterone was stimulated in both 24-h experiments ( % and % of control respectively; P<0 01 for both) and 48-h experiments ( % and % of control respectively; P<0 01 for both). The protein kinase C activator, TPA ( ng/ml, dose response data not shown), stimulated activin A secretion dose dependently, with the maximal effect at the concentration of 100 ng/ml in 24-h treatments. TPA (100 ng/ml) increased activin A and inhibin A secretion in 24-h treatments ( % and % of control respectively; P<0 05) and 48-h treatments

4 560 TVA}NTTINEN and others Activin A in granulosa cells Figure 2 Time-dependent (A, B) and dose-dependent (C, D) effects of FSH and LH on activin A, inhibin A and progesterone secretion by cultured human granulosa-luteal cells. The cells were incubated for 24 or 48 h in the presence of FSH (100 IU/l) (A) or LH (100 IU/l) (B). The bars represent the relative activin A, inhibin A and progesterone secretions per 24 and 48 h (mean+s.e.m.) when the respective control is adjusted to 100%. N, number of experiments performed. In dose response experiments (N=5), FSH (1 100 IU/l) (C) decreased activin A secretion at high concentrations, whereas it increased inhibin A secretion dose dependently during a 48-h treatment period. LH treatment (D) caused a similar trend in activin A and inhibin A secretion. *P<0 05, **P<0 01 (Mann Whitney test). ( % and % of control respectively; P<0 05), but it had no significant effect on progesterone secretion in either 24-h ( % of control; P=0 48) or 48-h treatments ( % of control; P=0 25) (Fig. 4). Discussion Our data show that activin A secretion in cultured granulosa-luteal cells is regulated by gonadotropins in a biphasic manner. Activin A secretion is increased when the treatment period is short or the gonadotropin dose is low, but long-term treatment or high gonadotropin concentrations lead to its inhibition. As the secretion of inhibin A is meanwhile stimulated by increasing gonadotropin concentrations and incubation time, the result is a clear decrease in the activin A:inhibin A secretion ratio. In addition, without gonadotropin stimulation, cultured granulosa-luteal cells exhibit the opposite tendency, to increase the activin:inhibin secretion ratio spontaneously.

5 Activin A in granulosa cells TVA}NTTINEN and others 561 Figure 3 Time-dependent effect of 8-BrcAMP (protein kinase A activator) on activin A, inhibin A and progesterone secretion by cultured human granulosa-luteal cells. The cells were incubated for 24 or 48 h in a culture medium containing 8-BrcAMP (100 μm) or culture medium only (control). The bars represent the relative secretion of activin A, inhibin A and progesterone per 24 and 48 h (mean+s.e.m.) by the cells when the respective control is adjusted to 100%. N, number of experiments performed. **P<0 01 (Mann Whitney test). Figure 4 Time-dependent effect of 12-O-tetradecanoyl phorbol- 13-acetate (TPA, protein kinase C activator) on activin A, inhibin A and progesterone secretion (logarithmic scale) by cultured human granulosa-luteal cells. The cells were incubated for 24 or 48 h in a culture medium containing TPA (100 ng/ml) or no TPA (control). The bars represent the relative secretion of activin A, inhibin A and progesterone per 24 and 48 h (mean+s.e.m.) when the respective control is adjusted to 100%. N, number of experiments performed. *P<0 05 (Mann Whitney test). The biphasic gonadotropin regulation of activin A secretion can be explained by previous mrna expression data from similar granulosa-luteal cell models. Gonadotropin (Erämaa et al. 1994) and 8-BrcAMP (Tuuri et al. 1996) stimulation increased the expression of βasubunit rapidly and transiently, whereas they increased the expression of α-subunit for at least 48 h. Our data suggest that the rapid stimulation of βa-subunit expression leads to increased activin A secretion, whereas the later dominance of the α-subunit leads to increased inhibin A secretion during continuing gonadotropin stimulation. 8-BrcAMP also induced βa-subunit mrna expression at 48 h (Tuuri et al. 1996), which explains the increase in activin A secretion after both 24 and 48 h of incubation. Tuuri and co-workers (1996) showed that TPA increases the expression of the βa-subunit gene without any effect on the α-subunit. Our data suggest that most of this increased βa-subunit mrna is translated and processed to activin A. The increase in inhibin A after TPA treatment shows that granulosa-luteal cells form sufficient α-subunit for inhibin production. However, we do not know the significance of these regulatory mechanisms in vivo, as the physiological activators of the protein kinase C signal transduction pathway in granulosa cells are poorly known. The observation that the ratio of βa- to α-subunit mrnas increases during the period of culture (Erämaa et al. 1994) is also in agreement with our data, as the activin A:inhibin A ratio increased in relation to the time in culture. On the basis of the findings of the present and previous studies, it seems that the activin A:inhibin A secretion ratio can be predicted well from the mrna expression data of the activin:inhibin subunits. Muttukrishna and co-workers (1997) were the first to show that the secretion of activin A in cultured human granulosa-luteal cells is stimulated by gonadotropins. Their finding that LH stimulates activin A secretion only after 24 h, but not after 48 h, is in accordance with our data. However, their report that FSH stimulates activin A secretion weakly after 48 h but not after 24 h of incubation is in contrast to our results. A possible reason for this difference is that they used a markedly different model, with granulosa-luteal cells cultured for only 24 or 48 h, in contrast to our cells, which were cultured for 1 week or more before hormonal treatments. It is possible that their granulosa-luteal cells were desensitized to FSH at the experimental period as a result of in vivo ovarian hyperstimulation. This is supported by the fact that FSH did not affect the steroid secretion of their cells, in contrast to ours, which exhibited significantly increased progesterone secretion in the presence of FSH. FSH and LH did not have completely similar effects in our study either. In particular, the sensitivity to LH-treatment varied, indicating that granulosa-luteal cells harvested from a heterogenous group of patients do not always behave exactly similarly. Nevertheless, the inhibin:activin secretion ratio increased uniformly during prolonged stimulation with both FSH and LH. It has been hypothesized that the follicular environment changes from activin- to inhibin-dominant during folliculogenesis, because granulosa cells exhibit an increase in gene expression of the α-subunit towards the end of the follicular phase (Hillier 1991, Knight & Glister 2001,

6 562 TVA}NTTINEN and others Activin A in granulosa cells Roberts et al. 1993). This shift was demonstrated at the protein level when Schneyer and co-workers (2000) showed a decrease in follicular fluid activin:inhibin ratio in relation to follicular growth. The findings of our study suggest that the decrease in activin:inhibin ratio could result from increasing gonadotropin stimulation during folliculogenesis as the granulosa cells start to express first FSH and then LH/hCG receptors at the preovulatory stage (Minegishi et al. 1997a,b). We speculate that increasing gonadotropin stimulation would then allow granulosa cells to produce increasing amounts of inhibin, whereas the production of activin A would be less than during the earlier stages of the menstrual cycle. It is also possible that the increase in granulosa cell secretion of progesterone is partly a consequence of a diminution of the para/autocrine inhibitory effect exerted on it by activin A towards the luteal phase. This hypothesis is supported by the finding of increased serum activin A concentrations in patients with hypothalamic hypogonadotropism (Petraglia et al. 1998). Conclusion The secretion of immunoreactive activin A is regulated by gonadotropins via the protein kinase A signal transduction pathway in human granulosa-luteal cells. Activin A secretion is initially increased by gonadotropin stimulation, but later diminishes rapidly, whereas inhibin A and progesterone secretion are increased time- and dose-dependently. This differential regulation of activin A and inhibin A secretion, leading to a decreasing activin:inhibin ratio during increasing gonadotropin stimulation, may be an important component in the local fine-tuning of ovarian function. Acknowledgements The laboratory technicians Mss Merja Haukka and Minna Heiskanen are thanked for their skillful assistance. Recombinant human FSH and LH for in vitro experiments were generously provided by Serono-Nordic (Vantaa, Finland). This study was financially supported by Academy of Finland, the Emil Aaltonen Foundation, the Foundation for Pediatric Research, the Jalmari and Rauha Ahokas Foundation, Novo Nordisk Foundation, Sigrid Juselius Foundation, and Kuopio University Hospital. References Erämaa M, Tuuri T, Hildén K& Ritvos O 1994 Regulation of inhibin α- and βa-subunit messenger ribonucleic acid levels by chorionic gonadotropin and recombinant follicle stimulating hormone in cultured human granulosa-luteal cells. Journal of Clinical Endocrinology and Metabolism Groome NP, Illingworth PJ, O Brien M, Cooke I, Ganesan TS, Baird DT & McNeilly AS 1994 Detection of dimeric inhibin throughout the human menstrual cycle by two-site enzyme immunoassay. Clinical Endocrinology Hillier SG 1991 Regulatory functions for inhibin and activin in human ovaries. Journal of Endocrinology Hillier SG, Young EL, Illingworth PJ, Baird DT, Schwall RH & Mason AJ 1991a Effect of recombinant activin on androgen synthesis in cultured human thecal cells. Journal of Clinical Endocrinology and Metabolism Hillier SG, Young EL, Illingworth PJ, Baird DT, Schwall RH & Mason AJ 1991b Effect of recombinant inhibin on androgen synthesis in cultured human thecal cells. Molecular and Cellular Endocrinology 75 R1 R6. Knight PG & Glister C 2001 Potential local regulatory functions of inhibins, activins and follistatin in the ovary. Reproduction Knight PG, Muttukrishna S& Groome NP 1996 Development and application of a two-site enzyme immunoassay for the determination of total activin A concentrations in serum and follicular fluid. Journal of Endocrinology Lewis KA, Gray PC, Blount AL, MacConell LA, Wiater E, Bilezikjian LM & Vale W 2000 Betaglycan binds inhibin and can mediate functional antagonism of activin signaling. Nature Liu J, Hydén-Granskog C& Voutilainen R 2001 Gonadotrophins inhibit and activin induces expression of inhibin/activin βb subunit mrna in cultured human granulosa-luteal cells. Molecular Human Reproduction Matzuk MM 2000 In search of binding identification of inhibin receptors [Editorial]. Endocrinology Minegishi T, Tano M, Igarashi M, Rokukawa S, Abe Y, Ibuki Y & Miyamoto K 1997a Expression of follicle-stimulating hormone receptor in human ovary. European Journal of Clinical Investigation Minegishi T, Tano M, Abe Y, Nakamura K, Ibuki Y&Miyamoto K 1997b Expression of luteinizing hormone/human chorionic gonadotropin (LH/HCG) receptor mrna in the human ovary. Molecular Human Reproduction Muttukrishna S, Fowler PA, Groome NP, Mitchell GG, Robertson WR & Knight PG 1994 Serum concentrations of dimeric inhibin during the spontaneous human menstrual cycle and after treatment with exogenous gonadotrophin. Human Reproduction Muttukrishna S, Groome N& Ledger W 1997 Gonadotropic control of secretion of dimeric inhibins and activin A by human granulosa-luteal cells in vitro. Journal of Assisted Reproduction and Genetics Petraglia F, Hartmann B, Luisi S, Florio P, Kirchengast S, Santuz M, Genazzani AD & Genazzani AR 1998 Low levels of serum inhibin A and inhibin B in women with hypergonadotropic amenorrhea and evidence of high levels of activin A in women with hypothalamic amenorrhea. Fertility and Sterility Rabinovici J, Spencer SJ, Doldi N, Goldsmith PC, Schwall R& Jaffe RB 1992 Activin-A as an intraovarian modulator: actions, localization, and regulation of the intact dimer in human ovarian cells. Journal of Clinical Investigation Roberts VJ, Barth S, el-roeiy A& Yen SSC 1993 Expression of inhibin/activin subunits and follistatin messenger ribonucleic acids and proteins in ovarian follicles and the corpus luteum during the human menstrual cycle. Journal of Clinical Endocrinology and Metabolism Schneyer AL, Fujiwara T, Fox J, Welt CK, Adams J, Messerlian GM & Taylor AE 2000 Dynamic changes in the intrafollicular inhibin/activin/follistatin axis during human follicular development: relationship to circulating hormone concentrations. Journal of Clinical Endocrinology and Metabolism Tuuri T, Erämaa M, Van Schaik RHN & Ritvos O 1996 Differential regulation of inhibin/activin α- and βa-subunit and follistatin

7 Activin A in granulosa cells TVA}NTTINEN and others 563 mrnas by cyclic AMP and phorbol ester in cultured human granulosa-luteal cells. Molecular and Cellular Endocrinology Vänttinen T, Liu J, Hydén-Granskog C, Parviainen M, Penttilä I & Voutilainen R 2000 Regulation of immunoreactive inhibin A and B secretion in cultured human granulosa-luteal cells by gonadotropins, activin A, and insulin-like growth factor type-1 receptor. Journal of Endocrinology Voutilainen R, Tapanainen J, Chung B-C, Matteson KJ & Miller WL 1986 Hormonal regulation of P450 scc (20,22-desmolase) and P450c17 (17α-hydroxylase/17,20-lyase) in cultured human granulosa cells. Journal of Clinical Endocrinology and Metabolism Ying SY 1988 Inhibins, activins, and follistatins: gonadal proteins modulating the secretion of follicle-stimulating hormone. Endocrine Reviews Received 25 September 2001 Accepted 23 November