当前位置: 首页 > 医学版 > 期刊论文 > 内科学 > 内分泌学杂志 > 2005年 > 第4期 > 正文
编号:11168413
Bone Morphogenetic Proteins (BMP) -4, -6, and -7 Potently Suppress Basal and Luteinizing Hormone-Induced Androgen Production by Bovine Theca
     School of Animal and Microbial Sciences, The University of Reading, Whiteknights, Reading RG6 6AJ, United Kingdom

    Address all correspondence and requests for reprints to: Prof. P. G. Knight, School of Animal and Microbial Sciences, The University of Reading, Whiteknights, Reading RG6 6AJ, United Kingdom. E-mail: p.g.knight@reading.ac.uk.

    Abstract

    We reported recently that bovine theca interna cells in primary culture express several type-I and type-II receptors for bone morphogenetic proteins (BMPs). The same cells express at least two potential ligands for these receptors (BMP-4 and -7), whereas bovine granulosa cells and oocytes express BMP-6. Therefore, BMPs of intrafollicular origin may exert autocrine/paracrine actions to modulate theca cell function. Here we report that BMP-4, -6, and -7 potently suppress both basal (P < 0.0001; respective IC50 values, 0.78, 0.30, and 1.50 ng/ml) and LH-induced (P < 0.0001; respective IC50 values, 5.00, 0.55, and 4.55 ng/ml) androgen production by bovine theca cells while having only a moderate effect on progesterone production and cell number. Semiquantitative RT-PCR showed that all three BMPs markedly reduced steady-state levels of mRNA for P450c17. Levels of mRNA encoding steroidogenic acute regulatory protein, P450scc, and 3?-hydroxy- steroid dehydrogenase were also reduced but to a much lesser extent. Immunocytochemistry confirmed a marked reduction in cellular content of P450c17 protein after BMP treatment (P < 0.001). Exposure to BMPs led to cellular accumulation of phosphorylated Smad1, but not Smad2, confirming that the receptors signal via a Smad1 pathway. The specificity of the BMP response was further explored by coincubating cells with BMPs and several potential BMP antagonists, chordin, gremlin, and follistatin. Gremlin and chordin were found to be effective antagonists of BMP-4 and -7, respectively, and the observation that both antagonists enhanced (P < 0.01) androgen production in the absence of exogenous BMP suggests an autocrine/paracrine role for theca-derived BMP-4 and -7 in modulating androgen production. Collectively, these data indicate that an intrafollicular BMP signaling pathway contributes to the negative regulation of thecal androgen production and that ovarian hyperandrogenic dysfunction could be a result of a defective autoregulatory pathway involving thecal BMP signaling.

    Introduction

    RECENT EVIDENCE INDICATES that bone morphogenetic proteins (BMPs) are expressed in a cell-specific manner in the mammalian ovary and have important roles in the regulation of folliculogenesis, ovulation, and corpus luteum function (1, 2). In several species, including rat and bovine, BMP-4 and -7 are expressed by theca cells (3, 4, 5), BMP-2 and -6 are expressed by granulosa cells (4, 5), and BMP-15 and -6 are expressed by oocytes (4, 5, 6, 7, 8). In common with other members of the TGF-? superfamily, BMPs are dimeric proteins that interact with target cells by forming heterooligomeric complexes with two types of serine/threonine kinase receptor (type-I and type-II) located on the cell surface (9, 10, 11). Transphosphorylation of type-I receptor by complexed type-II receptor, leads to phosphorylation of intracellular signaling intermediaries called Smads 1, 5, or 8, which associate with co-Smad4 before translocating to the nucleus to alter gene transcription. Consistent with autocrine/paracrine BMP signaling within the ovary, theca cells, granulosa cells, oocytes, and luteal cells express a range of BMP-responsive type-I and -II receptors (3, 4, 5). These include three type-I receptors (also referred to as activin receptor-like kinases, ALK) called BMPRIA (ALK3), BMPRIB (ALK6), and ActRIA (ALK2) and three type-II receptors called BMPRII, ActRIIA, and ActRIIB.

    Such evidence for a functional intraovarian BMP system is supported by the in vivo finding that sheep with naturally occurring mutations in either the BMPRIB or BMP-15 gene display aberrant ovarian follicle development (12, 13, 14). Moreover, in vitro experiments on isolated granulosa cells have revealed modulatory effects of BMPs on steroidogenesis (3, 5, 15, 16, 17), inhibin/activin secretion (5, 17), follistatin secretion (5), and cell proliferation (5, 16). For instance, with rat granulosa cells, BMP-4 and -7 enhanced FSH-induced estradiol secretion and cell proliferation while suppressing FSH-induced progesterone (P4) secretion (3, 15). With bovine granulosa cells, BMP-4, -6, and -7 enhanced both basal and IGF-stimulated secretion of estradiol, inhibin-A, activin-A, and follistatin, while suppressing P4 secretion (5).

    The synthesis of androgens by thecal cells is a key aspect of follicle function because, in accordance with the two-cell, two-steroid model (18), the ability of granulosa cells in developing antral follicles to synthesize increasing amounts of estrogen requires an adequate supply of androgen substrate for P450 aromatase. Moreover, hypersecretion of thecal androgens is a characteristic of polycystic ovarian syndrome (PCOS), a widespread although poorly understood condition that compromises human fertility and has other pathophysiological consequences (19, 20). Despite this, there is a relative paucity of information on the local regulatory factors affecting theca cell function. Given the recent discovery of the intraovarian BMP system and the demonstration that theca cells express BMP receptors, the possibility that BMP ligands of thecal, granulosa, or oocyte origin exert autocrine/paracrine actions to modulate basal and gonadotropin-dependent theca cell function warrants detailed evaluation. An earlier study (21), involving a human ovarian theca-like tumor cell model, showed that BMP-4, like activin, reduced forskolin-induced androgen secretion and P450c17 expression. However, we are not aware of any studies to date investigating the actions of BMPs on primary theca cell cultures from any species.

    Here we report the use of a primary bovine theca interna cell culture model to investigate the effects of BMP-4, -6, and -7 on 1) basal and LH-induced secretion of androstenedione (A4) and P4 and cell proliferation/survival, 2) accumulation of phosphorylated (p)Smad1, and 3) expression of key regulatory proteins and enzymes involved in the steroidogenic pathway. Having demonstrated a potent suppressive effect of each BMP on A4 secretion, P450c17 expression and pSmad1 accumulation, we 4) evaluated the ability of a range of potential BMP antagonists to neutralize these BMP actions. This approach then allowed us to test the hypothesis that endogenous BMP-4 and/or -7 produced by theca cells serve to attenuate androgen secretion in an autocrine/paracrine fashion.

    Materials and Methods

    Isolation and culture of theca interna cells

    Ovaries were obtained from cattle slaughtered at random stages of the estrous cycle at an abattoir. Follicles (4–6 mm diameter) were dissected from ovaries, aspirated, and hemisected, and a plastic inoculation loop was used to dislodge the granulosa cell layer as previously described (22, 23). Follicle halves were then shaken vigorously to remove any remaining granulosa cells, and the medium was changed three times. Follicle halves were examined under a dissecting microscope, and the theca interna layer was gently peeled away from the basement membrane. Pooled theca layers were incubated with collagenase (type IV, 1 mg/ml; Sigma Ltd., Poole, UK) and trypsin inhibitor (100 μg/ml; Sigma) in a shaking water bath at 37 C. After 30 min, the cell layers were triturated with a Pasteur pipette and returned to the water bath for another 15 min. Finally, any remaining undigested material was allowed to settle, and the resulting theca-cell-rich supernatant decanted. Cells were centrifuged (800 x g for 10 min) and the pellet resuspended in PBS before being subjected to an osmotic shock treatment (22) to lyse any red blood cells present. After washing again, cells were resuspended in culture medium [McCoy’s 5A modified medium supplemented with 1% (vol/vol) antibiotic-antimycotic solution, 10 ng/ml bovine insulin, 2 mM L-glutamine, 10 mM HEPES, 5 μg/ml apotransferrin, 5 ng/ml sodium selenite, and 0.1% BSA; all purchased from Sigma). Cells were counted using a hemocytometer, and viability, assessed using trypan blue, ranged from 80–90%. Cells (75 x 103 viable cells/50 μl) were dispensed into wells of 96-well plates or 16-well chamber slides containing preequilibrated medium and incubated at 38.5 C in 5% CO2 and 95% air for 48 h. Conditioned media were then removed (discarded), and cells were incubated for two successive 48-h periods with/without treatments. Conditioned media samples were stored at –20 C for steroid immunoassays. At the end of culture period, viable cell number was determined by neutral red dye uptake as described elsewhere (22, 24).

    Steroid measurements

    Concentrations of A4 were determined by RIA (25) with a detection limit of 50 pg/ml and intra- and interassay coefficients of variation of 8 and 10%, respectively. P4 was measured by ELISA (26) with a detection limit of 20 pg/ml and intra- and interassay coefficients of variation of 7 and 11%, respectively.

    Experiment 1: dose-dependent effects of LH on secretion of A4 and P4

    Theca cells were cultured for 4 d with a wide range of LH dose levels [0 and 25–3200 pg/ml; NIDDK-oLH-S26, National Hormone and Pituitary Program (NHPP), Torrance, CA]. Conditioned media were collected for measurement of A4 and P4, and viable cell number was determined at the end of the culture period.

    Experiment 2: effect of BMPs on steroid production and cell number under basal and LH-stimulated conditions

    Theca cells were cultured for 4 d with recombinant human BMP-4, -6, and -7 (R&D Systems, Abingdon, UK) at dose levels of 0, 0.08, 0.4, 2, 10, and 50 ng/ml both in the absence and presence of LH (0, 100, and 1600 pg/ml). As above, conditioned media were collected and viable cell number determined at the end of culture.

    Experiment 3: effect of BMPs on expression of mRNAs encoding LH receptor (LHR), steroidogenic acute regulatory protein (StAR), P450scc, 3?-hydroxysteroid dehydrogenase (3?-HSD), and P450c17.

    For mRNA expression studies, theca cells were cultured with/without BMPs (2 ng/ml) at the same density as above but in 25-cm2 flasks (5 x 106 cells/10 ml per flask). At the end of the culture period, cells were dislodged using a cell scraper and transferred with conditioned medium to a 20-ml tube. After centrifugation (10 min at 800 x g), medium was removed for steroid assay, the cell pellet was washed with culture medium, and total RNA was extracted using 1 ml of Trizol reagent (Sigma) according to the manufacturer’s recommendations. To exclude DNA contamination, isolated RNA samples were treated with RNase-free DNase (Promega, Southampton, UK) and the Trizol reagent extraction process was repeated; integrity and purity of final RNA extracts were assessed by agarose gel electrophoresis and spectrophotometry (A260/A280 ratio). First-strand cDNA was synthesized from total RNA using a PowerScript reverse transcriptase kit (Clontech, Palo Alto, CA) according to the manufacturer’s recommendations. The cDNA product was diluted 1:10 with nuclease-free water and stored at –20 C. PCR primer pairs (see Table 1) were designed to amplify partial cDNAs for bovine LHR, StAR, P450scc, 3?-HSD, P450c17, and the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH). PCRs were carried out in a total volume of 25 μl using Extensor Hi Fidelity ReddyMix PCR Master Mix (Abgene, Epsom, UK). All PCR included a denaturing step (94 C for 60 sec), annealing step (62–72 C for 30 sec; see Table 1) and extension step (68 C for 60 sec). For negative controls, cDNA template was substituted with nuclease-free water. PCR products were analyzed by agarose gel electrophoresis. Gels were stained with ethidium bromide and photographed under UV illumination, and images were subsequently analyzed using an imaging densitometer with Quantity One software (Bio-Rad, Hemel Hempstead, UK). Expression values for each mRNA were normalized to GAPDH expression.

    TABLE 1. Details of PCR primers employed, including their annealing temperatures and expected size of each amplicon

    Experiment 4: effects of BMP-4, -6, and -7 on cellular expression of P450c17 protein

    Cells were cultured in chamber slides and incubated with/without BMP-4, -6, or -7 (each at 2 ng/ml) for 4 d. After fixing for 30 min in 4% paraformaldehyde, cells were immunostained with rabbit antiserum against P450c17 (1/500 dilution; gift from Professor S. Kominami, Hiroshima University, Hiroshima, Japan). Normal rabbit serum (1/500 dilution) was used as a control. The second antibody used was Alexa 488-conjugated goat antirabbit IgG (10 μg/ml; Molecular Probes, Leiden, The Netherlands). Images were captured using a confocal microscope (Leica TCS-NT, Leica Lasertechnik GmBH, Heidelberg, Germany), and the relative intensity of immunostaining achieved under each experimental condition was quantified using NIH Image 1.63 software by measuring the fluorescence intensity of 20 randomly selected areas of each image (each containing approximately 10–15 cells). Care was taken to use fixed settings on the microscope (i.e. laser intensity, photomultiplier voltage, and contrast) while capturing each series of related images to facilitate subsequent quantitative comparison of relative fluorescence intensities.

    Experiment 5: effects of BMP-4, -6, and -7 and activin-A on cellular accumulation of pSmad1 and -2

    After culture with/without LH (100 pg/ml) for 4 d, cells were exposed to BMP-4, -6, or -7 or activin-A (50 ng/ml) for 4 h before fixing (30 min in 4% paraformaldehyde) and immunostaining with affinity-purified rabbit polyclonal antibodies against pSmad1 and pSmad2 (5 μg/ml; Upstate, Botolph Claydon, UK) as described previously for granulosa cells (5). Normal rabbit IgG (5 μg/ml) was used as a control. The second antibody used was Alexa 488-conjugated goat antirabbit (10 μg/ml; Molecular Probes). Images were captured and fluorescence intensities quantified as described above for experiment 4.

    Experiment 6: effects of putative BMP antagonists on steroid production in the presence and absence of exogenous BMPs

    Theca cells were cultured under both basal and LH-stimulated conditions as described for experiments 1 and 2. Cells were exposed to different combinations of BMP-4, -6, or -7 (at 0 and 2 ng/ml) with/without three putative BMP antagonists, recombinant mouse chordin (0 and 1.25 μg/ml; R&D Systems), recombinant gremlin (0 and 1.25 μg/ml; R&D Systems), and recombinant human follistatin (0 and 0.14 μg/ml; NHPP). The concentrations of chordin and gremlin used were selected on the basis of potency estimates provided by the supplier.

    Statistical analysis

    Unless stated otherwise, values are presented as arithmetic means ± SEM and P values < 0.05 were considered to be statistically significant. ANOVA of log-transformed data from four or more independent cultures were used to evaluate the effect of LH, BMPs, and BMP antagonists on A4 or P4 secretion and on viable cell number. Post hoc Fisher’s protected least significant difference tests were used for pair-wise comparisons providing ANOVA yielded a significant F ratio. Comparison of the relative fluorescence intensities of cells immunostained with P450c17 or pSmad1/2 antibodies were made by ANOVA and Fisher’s protected least significant difference test with results based on three independent cell cultures. To facilitate evaluation of semiquantitative RT-PCR data, mRNA expression values (normalized to GAPDH expression) from three independent cultures were plotted as means ± 95% confidence intervals.

    Results

    Effect of LH on steroid secretion and cell number

    Treatment of theca cells with LH had a marked dose-dependent stimulatory effect on steroid secretion (Fig. 1). LH-induced A4 secretion showed a biphasic response (P < 0.0001) with a maximum 6-fold increase over basal levels seen in cells treated with 100 pg/ml LH; higher dose levels of LH gave suboptimal A4 secretory responses. In contrast to A4 secretion, the P4 secretory response (P < 0.0001) was monophasic, and the maximal response was elicited by the highest dose level of LH tested (3200 pg/ml). Treatment with LH promoted a modest although significant decrease in viable cell number (2-fold; P < 0.0001) determined at the end of the 144-h culture period.

    FIG. 1. Dose-dependent effect of LH on secretion of androstenedione (A) and progesterone (B) by bovine theca interna cells in primary culture. C, Viable cell number at the end of the culture period. Values are means ± SEM (n = 8 independent cultures), and ANOVA results are indicated.

    Effects of BMP-4, -6, and -7 on basal and LH-induced steroid secretion and cell number

    The effects of BMP-4, -6, and -7 were tested at a range of dose levels (0, 0.08, 0.4, 2, 10, and 50 ng/ml) in the presence and absence of LH (0, 100, or 1600 pg/ml). As shown in Fig. 2A, when cells were cultured under basal conditions (i.e. in the absence of LH stimulation), all three BMPs tested caused a significant dose-dependent decrease in A4 secretion (P < 0.0001). Conversely, a significant (P < 0.0001) dose-dependent increase in both P4 secretion and cell number was seen. Two-way ANOVA revealed a statistically significant difference (P < 0.05) between the effects of BMP-4, -6, and -7 on basal A4 secretion with IC50 values of 0.8, 0.3, and 1.5 ng/ml, respectively (Table 2).

    FIG. 2. Effects of BMP-4, -6, and -7 alone (A) and in the presence of 100 pg/ml LH (B) or 1600 pg/ml LH (C) on the secretion of androstenedione and progesterone and on viable cell number at the end of the culture period. Values are means ± SEM (n = 4 independent cultures), and results of two-way ANOVA are summarized.

    TABLE 2. Mean (± SEM) IC50 values for BMP-induced suppression of basal and LH-dependent androstenedione secretion by bovine theca cells

    When theca cells were cotreated with LH at a dose level of 100 pg/ml (Fig. 2B) or 1600 pg/ml (Fig. 2C), all three BMPs caused a marked dose-dependent suppression (P < 0.0001) in A4 secretion. Two-way ANOVA revealed a significant difference between the BMPs (P < 0.0001) when cocultured with 100 pg/ml LH, with BMP-6 (IC50 value 0.55 ng/ml) being significantly more potent than BMP-4 and -7 (IC50 values, 5.00 and 4.55 ng/ml, respectively). Conversely, all three BMPs enhanced P4 secretion in the absence of LH, although not statistically so in its presence.

    Viable cell number was modestly (2-fold; P < 0.0001) increased by all three BMPs both in the absence and presence of LH at 100 or 1600 pg/ml.

    Effect of BMP-4, -6, and -7 on the expression of mRNAs encoding LHR, StAR, P450scc, 3?-HSD, and P450c17

    As shown in Fig. 3, semiquantitative RT-PCR revealed that all three BMPs dramatically reduced steady-state levels of mRNA for P450c17 relative to GAPDH expression. The three BMPs also reduced levels of mRNA encoding P450scc, 3?-HSD, and StAR but to a lesser degree.

    FIG. 3. Semiquantitative RT-PCR analysis of RNA isolated from bovine theca cells cultured for 4 d in the absence (control) or presence of BMP-4, -6, or -7 (each at 2 ng/ml). Bars show relative mRNA expression (normalized to GAPDH expression) for P450c17 (A), P450scc (B), 3?-HSD (C), StAR (D), and LHR (E). Insets show representative amplicon bands for each target mRNA (upper band) and the corresponding GAPDH mRNA (lower band). Values are means ± 95% confidence intervals based on three independent cell culture experiments.

    Immunodetection of P450c17 protein after treatment with BMP-4, -6, and -7

    Figure 4 shows representative images of theca cells stained with P450c17 antibody after culturing with/without BMP treatment. Positive cytoplasmic immunostaining for P450c17 was clearly evident in untreated cells but was barely detectable in cells exposed to BMP-4, -6, or -7. Quantification of the relative immunofluorescence signals (see histograms in Fig. 4) indicated that the P450c17 content of cells treated with BMPs was reduced by 85–95% compared with untreated cells (P < 0.001).

    FIG. 4. Representative confocal micrographs of bovine theca cells immunostained (green) with rabbit antiserum against P450c17 after culturing in the presence and absence of BMP-4, -6, and -7. Controls were stained with normal rabbit serum in place of rabbit anti-P450c17. Nuclei are counterstained (red) with propidium iodide. The histograms show the results of quantitative assessment of relative fluorescence intensity of cells stained with anti-P450c17. Values are means ± SD (n = 3 independent cultures), and letters denote significant differences (P < 0.01). Scale bar, 10 μm.

    Accumulation of phosphorylated Smad1 and Smad2 in theca cells after stimulation with BMP-4, -6, and -7 and activin-A

    Representative images of cultured theca cells immunostained with pSmad1 and pSmad2 antibodies are shown in Fig. 5. Increased pSmad 1 immunoreactivity was detected when cells were exposed to either BMP-4, -6, or -7 but not to activin-A. In contrast, increased pSmad2 immunoreactivity was detected after exposure to activin-A, but not to BMP-4, -6, or -7. Quantification of the relative fluorescent signals (see histograms in Fig. 5) confirmed that BMPs selectively induced accumulation of pSmad1 (P < 0.001), whereas activin-A selectively induced accumulation of pSmad2 (P < 0.001). Interestingly, it was also noted that BMP treatment significantly decreased the amount of pSmad2 accumulation in cells relative to that observed in untreated cells (P < 0.05).

    FIG. 5. Representative confocal micrographs showing that 4-h exposure to BMP-4, -6, and -7 selectively induces nuclear accumulation of pSmad1 in bovine theca cells whereas activin-A selectively induces accumulation of pSmad2. Controls were stained with normal rabbit IgG in place of rabbit anti-pSmad1 IgG. The histograms aligned to each row show the results of quantitative assessment of relative fluorescence. Values are means ± SD (n = 3 independent cultures), and means without a common letter are significantly different (P < 0.05). Scale bar, 10 μm.

    Reversal of BMP-induced suppression of androgen secretion by chordin, gremlin, and follistatin

    The specificity of the BMP response was further explored by coincubating cells with BMP-4, -6, or -7 and several potential BMP antagonists. Under both basal (Fig. 6A) and LH-stimulated (Fig. 6B) conditions, BMP-4, -6, and -7 alone suppressed A4 secretion. This suppression of androgen production was reversed to a variable extent depending on the BMP ligand/antagonist combination. Coincubation with chordin selectively reversed BMP-7-induced suppression of A4 secretion, whereas gremlin selectively reversed BMP-4-induced suppression of A4 secretion under both basal and LH-stimulated conditions. At the dose level tested, follistatin failed to reverse the effects of BMPs on A4 secretion. Indeed, under LH-stimulated conditions, follistatin actually enhanced (P < 0.05) the suppressive effect of both BMP-4 and -7 on A4 secretion, whereas the effect of BMP-6 was not significantly altered. Similarly, treatment with chordin enhanced the suppressive effect of BMP-4 on A4 secretion.

    FIG. 6. Interactions between BMP-4, -6, and -7 and three potential BMP antagonists (chordin, gremlin, and follistatin) in modulating thecal androstenedione secretion by bovine theca cells cultured under basal (A) and LH-stimulated (B) conditions. Values are means ± SEM based on data from four independent culture experiments. Within each set of bars, means without a common letter are significantly different (P < 0.05).

    Chordin, gremlin, and follistatin enhance thecal androgen secretion in the absence of exogenous BMPs

    Figure 7 shows that when chordin, gremlin, and follistatin were added to cells cultured without LH stimulation and without addition of exogenous BMPs, secretion of A4 increased by 45% (P < 0.05), 140% (P < 0.001), and 67% (P < 0.05), respectively. Gremlin also promoted a small increase (35%; P < 0.05) in A4 secretion under optimal LH-stimulated conditions, whereas chordin and follistatin did not.

    FIG. 7. Stimulatory effect of three potential BMP antagonists (chordin, gremlin, and follistatin) on androstenedione secretion by bovine theca cells cultured without exogenous BMPs. Values are means ± SEM based on data from four independent culture experiments. *, P < 0.05; **, P < 0.01; ***, P < 0.001 compared with controls.

    Discussion

    Since the existence of an intraovarian BMP system was discovered some 5 yr ago (3), the focus of the majority of studies has been on the granulosa cell as a potential target for BMP action within the ovary. Modulatory effects of BMPs on granulosa FSH receptor expression (16); aromatase expression (15); estradiol and P4 secretion (3, 5, 15, 17); inhibin, activin, and follistatin secretion (5, 17); and cell proliferation/survival (5, 15) have been reported. To our knowledge, the present study is the first, in any species, to examine the effects of BMPs on basal and LH-induced steroidogenesis by primary cultures of theca interna cells. The finding that BMP-4, -6, and -7 are potent (IC50 values in sub-nanomolar range) modulators of basal and LH-induced steroidogenesis and cell proliferation/survival is significant within the context of local intrafollicular regulatory mechanisms because all three molecules are expressed in a cell-dependent manner within ovarian follicles, and theca cells have been shown to express a range of BMP-responsive receptors including BMPRIB, BMPRII, ActRIIA, and ActRIIB (rat, Refs. 3, 4 ; bovine, Ref. 5). As reported previously, BMP-6 is expressed by granulosa cells and oocyte (rat, Ref. 3 ; bovine, Ref. 5) and thus has the potential to function as an intrafollicular paracrine regulator of theca cell function. Likewise, BMP-4 and -7 are expressed by theca cells (rat, Refs. 3, 4 ; bovine, Ref. 5) and may exert autocrine/paracrine actions on thecal steroidogenesis and proliferation/survival. Our findings that all three BMPs tested (BMP-4, -6, and -7) dose-dependently suppressed both basal and LH-induced A4 secretion by bovine theca interna cells accord well with a previous report (21) that BMP-4 reduced forskolin-induced A4 secretion by a human ovarian theca-like tumor cell line. In contrast to the present study, however, basal A4 secretion was not affected by BMP treatment in this tumor cell model. Moreover, the tumor cells are unresponsive to gonadotropins, necessitating the use of forskolin or cAMP analogs to activate the protein kinase A-mediated signaling pathway.

    In the present serum-free culture model, the theca interna cells proliferate some 4-fold over 6 d and aggregate into small cellular clumps that attach firmly to the culture surface; they secrete readily measurable amounts of A4 and P4 under both nonstimulated basal conditions and in response to LH stimulation. Although responsiveness to LH in terms of A4 secretion was biphasic (optimal LH dose level of 100 pg/ml), P4 secretion increased monophasically and was maximal at the highest LH dose level tested (3200 pg/ml). This differential A4/P4 responsiveness to LH in vitro appears to reflect functional luteinization of the cells in response to high LH exposure, as occurs in vivo after the preovulatory LH surge. However, it should be noted that LH had a small although highly consistent dose-dependent inhibitory effect on cell proliferation/survival, whereas luteinization is characterized by enhanced cell proliferation.

    Our results clearly show that the three BMPs tested (BMP-4, -6, and -7) exerted a similar effect on each aspect of theca cell function examined, although BMP-6 was significantly more potent than BMP-4 and -7 in suppressing basal and LH-induced A4 secretion. Whether these differences reflect true potency differences among the BMPs or arise from differential association with endogenous binding proteins produced by the cultured cells is unknown at this stage.

    It is now well established that the interaction of BMPs with type-I and -II receptors on the cell surface leads to the activation of an intracellular signaling pathway involving Smad1, Smad5, and Smad8. In contrast, activin signaling involves activation of a Smad2/3 pathway (9, 10, 11). The present observations that exposure of bovine theca cells to BMP-4, -6, or -7 led to nuclear accumulation of pSmad1 (but not pSmad2) whereas, conversely, exposure to activin-A promoted nuclear accumulation of pSmad2 (but not pSmad1) are consistent with this and confirm that functional BMP ligand-receptor interactions were occurring. Despite this, it is not known which particular combinations of type-I and -II receptors are recruited by each BMP to elicit a biological response in terms of Smad activation, altered steroidogenesis, and cell proliferation/survival. As we reported recently (5), bovine theca cells express at least two type-I receptors (BMPRIB and ActRIA) and three type-II receptors (BMPRII, ActRIIA, and ActRIIB) that can potentially bind BMPs. Unexpectedly, we noted that pSmad2 accumulation in BMP-treated cells was significantly lower than in untreated control cells. One possible explanation for this is that the exogenous BMP is competing with an endogenous ligand (activin or perhaps TGF-?) for binding to ActRIIA or -IIB leading to reduced activin/TGF-? signaling via the Smad2/3 pathway.

    In an effort to determine the mechanism(s) through which BMP-4, -6, and -7 inhibit androgen production and raise P4 production, semiquantitative RT-PCR was used to compare steady-state levels of mRNA expression for key regulatory proteins and enzymes involved in the steroidogenic pathway including LHR, StAR, P450scc, 3?-HSD, and P450c17. The observation that all three BMPs promoted a profound suppression in P450c17 mRNA expression, coupled with our immunocytochemical finding of a parallel BMP-induced fall in cellular content of P450c17 protein, identifies this enzymatic step in the steroidogenic pathway as a key target of BMP action. Down-regulation of P450c17 expression by BMPs would block the conversion of C21 steroids to C19 steroids, an action consistent with the inverse changes in A4 and P4 secretion observed here. A previous study using a human ovarian theca-like tumor cell culture model also revealed a suppressive effect of BMP-4 on P450c17 expression and androgen secretion (21). Likewise, activin, another TGF-? superfamily member, was shown to inhibit P450c17 expression and androgen production in this model, whereas inhibin enhanced P450c17 expression and androgen production (27). Similar opposing effects of activin and inhibin on thecal androgen production have been reported for primary cultures of rat (28), human (29), and bovine (25) thecal cells and have been confirmed using the present bovine theca interna culture model (data not shown).

    Semiquantitative RT-PCR analysis also indicated that expression of StAR, P450scc, and 3?-HSD were also reduced after BMP treatment but to a lesser extent than P450c17 expression. Thus, it appears likely that activation of the BMP signaling pathway(s) can affect thecal steroidogenesis at multiple levels. Indeed, BMP-6 and -7 have been shown to inhibit expression of StAR by cultured rat granulosa cells (15, 16), whereas BMP-7 has been shown to inhibit P450scc expression (16). Despite this, the finding that basal P4 secretion increased in response to BMP treatment suggests that, under the present culture conditions, these other potential inhibitory actions are not rate limiting with respect to thecal synthesis of C21 steroids, at least in the absence of LH stimulation. Thus, the BMP-induced rise in progesterone can be accounted for by a reduced rate of conversion to androgen that more than compensates for any decrease in the rate of progesterone synthesis arising from reduced StAR expression. It is also possible that steady-state levels of StAR mRNA may not reflect the amount of functionally active StAR protein that is known to be regulated at the posttranslational level (30, 31).

    To further substantiate our findings, we evaluated the abilities of three potential BMP antagonists, chordin (32), gremlin (33), and follistatin (5, 34, 35), to reverse the suppression of androgen secretion induced by BMP-4, -6, and -7 under basal and LH-induced conditions. It is well established that follistatin is expressed in the ovary, principally by granulosa cells (1, 2). More recently, it has been reported that granulosa cells also express gremlin (36) and a related BMP binding protein called protein related to DAN and cerebrus (PRDC) (37). The present demonstration that chordin and gremlin effectively reversed the suppressive effects of BMP-7 and BMP-4, respectively, but did not influence the response to BMP-6 offers further evidence that the actions of these exogenous treatment preparations on androgen production in our culture model are indeed specific. The apparently selective ability of chordin and gremlin to reverse the actions of exogenous BMP-7 and BMP-4, respectively, was unexpected because both are considered antagonists of these two BMPs. Differential neutralizing ability could simply reflect the heterologous nature of the preparations used; the chordin and gremlin preparations were recombinant mouse proteins, whereas the BMP preparations were recombinant human proteins. However, using recombinant human gremlin Sudo et al. (37) also observed a selective reversal of BMP-4 action, but not BMP-6 or BMP-7 action in a luciferase-reporter gene bioassay.

    Unexpectedly, follistatin did not reverse the suppressive effects of either BMP-4, 6, or -7 on androgen secretion despite being added at an approximately 60-fold molar excess. The explanation for this is unknown at present because in a previous study (5) we showed that a 25-fold molar excess of follistatin completely reversed the effect of BMP-4 and partially reversed the effect of BMP-6 on pSmad1 accumulation in bovine granulosa cells. The present observation that follistatin actually enhanced rather than antagonized the suppressive effect of BMP-4 and -7 on LH-induced androgen secretion is notable given our previous finding that follistatin enhanced the effect of BMP-7 on pSmad1 accumulation in granulosa cells (5). Similarly, Amthor et al. (35) presented evidence that follistatin may facilitate rather than inhibit the presentation of BMP-7 to BMP receptors on embryonic myoblasts. These paradoxical findings are most likely attributable to the reversible, relatively low-affinity binding between follistatin and BMPs (1, 5, 35) and the association of follistatin to cell surface proteoglycans potentially forming a BMP reservoir at the cell surface (38). A similar mechanism might explain the seemingly paradoxical observation that chordin enhanced, rather than reversed, the suppressive effect of BMP-4 on LH-induced androgen secretion. In this regard, chordin has recently been shown to associate with cell surface proteoglycans (39), and thus, like follistatin, it could potentially act as a BMP reservoir near the cell surface. Subsequent dissociation of sequestered BMP, perhaps aided by the action of specific metalloproteinase(s) on chordin (40, 41) could result in enhanced interaction of BMP with its cognate receptors. To resolve such issues, more information is needed on the intrafollicular expression patterns of the various extracellular binding proteins implicated in the regulation of BMPs and other TGF-? superfamily members.

    Because BMP-4 and -7 are expressed by theca cells (1, 2), we hypothesized that these BMPs might function in an autoregulatory capacity to modulate theca cell function. In direct support of this, when chordin, gremlin, or follistatin were added to theca cells cultured in the absence of exogenous BMPs, all three putative antagonists promoted a significant increase in basal androgen secretion; gremlin also promoted a significant increase in LH-induced androgen secretion. The capacities of the heterologous chordin (mouse), gremlin (mouse), and follistatin (human) preparations used in this study to bind and block the actions of endogenous bovine BMPs are unknown at this stage. Although they were added to cells at high concentrations, it is quite possible that other more effective antagonists would promote even greater increases in androgen output. Nonetheless, these observations are consistent with an autocrine/paracrine role of endogenous thecal-derived BMP-4 and/or -7 in suppressing androgen production, particularly under basal conditions.

    PCOS is a major cause of infertility affecting approximately 5% of women. Despite intensive research, its etiology is poorly understood (19, 20). PCOS is associated with a variety of symptoms including menstrual irregularities, hirsutism, infertility, insulin resistance, and obesity. Ovarian hyperandrogenism is a pivotal feature and reflects increased thecal androgen production and enhanced expression of P450c17 and 3?-HSD (19, 42). In addition, follicle development arrests at the small/mid-antral stage, and granulosa cells have the characteristics of prematurely luteinized cells (20, 43). We suggest that these changes in thecal and granulosa function associated with PCOS could be caused, at least in part, by a defect in an intraovarian BMP signaling system(s). Several lines of evidence support this hypothesis: 1) BMPs act on theca cells to suppress basal and LH/forskolin-induced androgen production and P450c17 expression (this study and Ref. 21); 2) BMP antagonists can promote thecal androgen production (this study); 3) BMPs enhance granulosa cell proliferation/survival and have an antiluteinization effect on these cells (3, 5); and 4) treatment of human granulosa-lutein cells with BMPs up-regulates production of inhibin-B, a marker of healthy growing follicles but not preovulatory or luteinized follicles (44). Thus, the notion of a functional deficit in BMP secretion and/or action within ovarian follicles would be compatible with both overproduction of thecal androgens and arrested follicle development caused by diminished proliferation and/or premature luteinization of granulosa cells. In addition, our evidence that BMPs reduce expression of other enzymes and proteins involved in steroidogenesis (P450scc, 3?-HSD, and StAR) in addition to P450c17 is intriguing given the finding of multiple alterations in the thecal steroidogenic pathway in PCOS, including increased expression of P450scc, 3?-HSD, and P450c17 (45).

    In summary, the present findings extend our earlier observations (5) relating to the expression of BMP ligands and receptors in bovine ovarian cell types and indicate that an intrafollicular BMP signaling pathway may contribute to the negative regulation of thecal androgen production by down-regulating expression of P450c17, the key enzyme in the conversion of C21 steroids to C19 steroids. BMP-6 of granulosa and/or oocyte origin and BMP-4 and -7 of thecal origin are implicated in this regulatory process. The possibility that a defect in intrafollicular BMP signaling could be a contributory factor in the etiology of ovarian hyperandrogenism and PCOS warrants further investigation. Such a defect could arise from aberrant expression of BMP ligands, BMP binding proteins, BMP receptors, or components of the intracellular signaling pathway. Additional information regarding both local intrafollicular BMP concentrations and cell-specific pattern of expression of BMPs, their receptors, and extracellular binding proteins throughout the different stages of follicle development is required to fully evaluate the physiological relevance of these observations.

    Acknowledgments

    We are grateful to Dr. A. F. Parlow (NHPP, Torrance, CA) for providing ovine LH and recombinant human follistatin and Dr. S. Kominami (Hiroshima University, Japan) for providing the P450c17 antibody.

    References

    Shimasaki S, Moore RK, Otsuka F, Erickson GF 2004 The bone morphogenetic protein system in mammalian reproduction. Endocr Rev 25:72–101

    Knight PG, Glister C 2003 Local roles of TGF-? superfamily members in the control of ovarian follicle development. Anim Reprod Sci 78:165–183

    Shimasaki S, Zachow RJ, Li D, Kim H, Iemura S, Ueno N, Sampath K, Chang RJ, Erickson GF 1999 A functional bone morphogenetic protein system in the ovary. Proc Natl Acad Sci USA 96:7282–7287

    Erickson GF, Shimasaki S 2003 The spatiotemporal expression pattern of the bone morphogenetic protein family in rat ovary cell types during the estrous cycle. Reprod Biol Endocrinol 1:9

    Glister C, Kemp CF, Knight PG 2004 Bone morphogenetic protein (BMP) ligands and receptors in bovine ovarian follicle cells: actions of BMP-4, -6 and -7 on granulosa cells and differential modulation of Smad-1 phosphorylation by follistatin. Reproduction 127:239–254

    Bodensteiner KJ, Clay CM, Moeller CL, Sawyer HR 1999 Molecular cloning of the ovine growth/differentiation factor-9 gene and expression of growth/differentiation factor-9 in ovine and bovine ovaries. Biol Reprod 60:381–386

    Jaatinen R, Laitinen MP, Vuojolainen K, Aaltonen J, Heikinheimo K, Lehtonen E, Ritvos O 1999 Localization of growth differentiation factor-9 (GDF-9) mRNA and protein in rat ovaries and cDNA cloning of rat GDF-9 and its novel homolog GDF-9B. Mol Cell Endocrinol 156:189–193

    Elvin JA, Yan C, Matzuk MM 2000 Oocyte-expressed TGF-? superfamily members in female fertility. Mol Cell Endocrinol 159:1–5

    Miyazono K, ten Dijke P, Heldin CH 2000 TGF-? signaling by Smad proteins. Adv Immunol 75:115–157

    Massague J, Chen YG 2000 Controlling TGF-? signaling. Genes Dev 14:627–644[Free Full Text]

    Miyazawa K, Shinozaki M, Hara T, Furuya T, Miyazono 2002 Two major Smad pathways in TGF-? superfamily signaling. Genes Cells 7:1191–1204

    Galloway SM, McNatty KP, Cambridge LM, Laitinen MP, Juengel JL, Jokiranta TS, McLaren RJ, Luiro K, Dodds KG, Montgomery GW, Beattie AE, Davis GH, Ritvos O 2000 Mutations in an oocyte-derived growth factor gene (BMP15) cause increased ovulation rate and infertility in a dosage-sensitive manner. Nat Genet 25:279–283

    Wilson T, Wu XY, Juengel JL, Ross IK, Lumsden JM, Lord EA, Dodds KG, Walling GA, McEwan JC, O’Connell AR, McNatty KP, Montgomery GW 2001 Highly prolific Booroola sheep have a mutation in the intracellular kinase domain of bone morphogenetic protein IB receptor (ALK-6) that is expressed in both oocytes and granulosa cells. Biol Reprod 64:1225–1235

    McNatty KP, Juengel JL, Wilson T, Galloway SM, Davis GH, Hudson NL, Moeller CL, Cranfield M, Reader KL, Laitinen MP, Groome NP, Sawyer HR, Ritvos O 2003 Oocyte-derived growth factors and ovulation rate in sheep. Reprod Suppl 61:339–351

    Lee WS, Otsuka F, Moore RK, Shimasaki S 2001 Effect of bone morphogenetic protein-7 on folliculogenesis and ovulation in the rat. Biol Reprod 65:994–999

    Otsuka F, Moore RK, Shimasaki S 2001 Biological function and cellular mechanism of bone morphogenetic protein-6 in the ovary. J Biol Chem 276:32889–32895

    Souza CJ, Campbell BK, McNeilly AS, Baird DT 2002 Effect of bone morphogenetic protein-2 (BMP-2) on oestradiol and inhibin-A production by sheep granulosa cells, and localization of BMP receptors in the ovary by immunohistochemistry. Reproduction 123:363–369

    Greenwald GS and Roy SK 1994 Follicular development and its control. In: Knobil E, Neill JD, eds. The physiology of reproduction. 2nd ed. New York: Raven Press; 629–724

    Nelson VL, Legro RS, Strauss 3rd JF, McAllister JM 1999 Augmented androgen production is a stable steroidogenic phenotype of propagated theca cells from polycystic ovaries. Mol Endocrinol 13:946–957

    Tsilchorozidou T, Overton C, Conway GS 2004 The pathophysiology of polycystic ovary syndrome. Clin Endocrinol (Oxf) 60:1–17

    Dooley CA, Attia GR, Rainey WE, Moore DR, Carr BR 2000 Bone morphogenetic protein inhibits ovarian androgen production. J Clin Endocrinol Metab 85:3331–3337

    Glister C, Tannetta DS, Groome NP, Knight PG 2001 Interactions between follicle-stimulating hormone and growth factors in modulating secretion of steroids and inhibin-related peptides by nonluteinized bovine granulosa cells. Biol Reprod 65:1020–1028

    Glister C, Groome NP, Knight PG 2003 Oocyte-mediated suppression of follicle-stimulating hormone- and insulin-like growth factor-induced secretion of steroids and inhibin-related proteins by bovine granulosa cells in vitro: possible role of transforming growth factor . Biol Reprod 68:758–765

    Campbell BK, Scaramuzzi RJ, Webb R 1996 Induction and maintenance of oestradiol and immunoreactive inhibin production with FSH by ovine granulosa cells cultured in serum-free media. J Reprod Fertil 106:7–16

    Wrathall JH, Knight PG 1995 Effects of inhibin-related peptides and oestradiol on androstenedione and progesterone secretion by bovine theca cells in vitro. J Endocrinol 145:491–500

    Sauer MJ, Foulkes JA, Worsfold A, Morris BA 1986 Use of progesterone 11-glucuronide-alkaline phosphatase conjugate in a sensitive microtitre-plate enzyme immunoassay of progesterone in milk and its application to pregnancy testing in dairy cattle. J Reprod Fertil 76:375–391

    Sawetawan C, Carr BR, McGee E, Bird IM, Hong TL, Rainey WE 1996 Inhibin and activin differentially regulate androgen production and 17-hydroxylase expression in human ovarian thecal-like tumor cells. J Endocrinol 148:213–221

    Hsueh AJ, Dahl KD, Vaughan J, Tucker E, Rivier J, Bardin CW, Vale W 1987 Heterodimers and homodimers of inhibin subunits have different paracrine action in the modulation of luteinizing hormone-stimulated androgen biosynthesis. Proc Natl Acad Sci USA 84:5082–5086

    Hillier SG 1991 Regulatory functions for inhibin and activin in human ovaries. J Endocrinol 131:171–175

    Arakane F, King SR, Du Y, Kallen CB, Walsh LP, Watari H, Stocco DM, Strauss 3rd JF 1997 Phosphorylation of steroidogenic acute regulatory protein (StAR) modulates its steroidogenic activity. J Biol Chem 272:32656–32662

    Clark BJ, Ranganathan V, Combs R 2001 Steroidogenic acute regulatory protein expression is dependent upon post-translational effects of cAMP-dependent protein kinase A. Mol Cell Endocrinol 173:183–192

    Piccolo S, Sasai Y, Lu B, De Robertis EM 1996 Dorsoventral patterning in Xenopus: inhibition of ventral signals by direct binding of chordin to BMP-4. Cell 86:589–598

    Hsu DR, Economides AN, Wang X, Eimon PM, Harland RM 1998 The Xenopus dorsalizing factor gremlin identifies a novel family of secreted proteins that antagonize BMP activities. Mol Cell 1:673–683

    Iemura S, Yamaoto TS, Takagi C, Uchiyama H, Natsume T, Shimasaki S, Sugino H, Ueno N 1998 Direct binding of follistatin to a complex of bone-morphogenetic protein and its receptor inhibits ventral and epidermal cell fates in early Xenopus embryo. Dev Biol 95:9337–9342

    Amthor H, Christ B, Rashid-Doubell F, Kemp CF, Lang E, Patel K 2002 Follistatin regulates bone morphogenetic protein-7 (BMP-7) activity to stimulate embryonic muscle growth. Dev Biol 243:115–127

    Pangas SA, Jorgez CJ, Matzuk MM 2004 Growth differentiation factor 9 regulates expression of the bone morphogenetic protein antagonist, gremlin. J Biol Chem 279:32281–32286

    Sudo S, Avsian-Kretchmer O, Wang LS, Hsueh AJ 2004 Protein related to DAN and cerebrus is a bone morphogenetic protein antagonist that participates in ovarian paracrine regulation. J Biol Chem 279:23134–23141

    Nakamura T, Sugino K, Titani K, Sugino H 1991 Follistatin, an activin-binding protein, associates with heparin sulfate chains of proteoglycans on follicular granulosa cells. J Biol Chem 266:19432–19437

    Jasuja R, Allen BL, Pappano WN, Rapraeger AC, Greenspan DS 2004 Cell-surface heparan sulfate proteoglycans potentiate chordin antagonism of bone morphogenetic protein signaling and are necessary for cellular uptake of chordin. J Biol Chem 279:51289–51297

    Oelgeschlager M, Larrain J, Geissert D, De Robertis EM 2000 The evolutionarily conserved BMP-binding protein Twisted gastrulation promotes BMP signaling. Nature 405:757–763

    Canalis E, Economides AN, Gazzerro E 2003 Bone morphogenetic proteins, their antagonists, and the skeleton. Endocr Rev 24:218–235

    Wickenheisser JK, Quinn PG, Nelson VL, Legro RS, Strauss 3rd JF, McAllister JM 2000 Differential activity of the cytochrome P450 17-hydroxylase and steroidogenic acute regulatory protein gene promoters in normal and polycystic ovary syndrome theca cells. J Clin Endocrinol Metab 85:2304–2311

    Mason H 2000 Function of the polycystic ovary. Hum Fertil (Camb) 3:80–85

    Jaatinen R, Bondestam J, Raivio T, Hilden K, Dunkel L, Groome N, Ritvos O 2002 Activation of the bone morphogenetic protein signaling pathway induces inhibin ?(B)-subunit mRNA and secreted inhibin B levels in cultured human granulosa-luteal cells. J Clin Endocrinol Metab 87:1254–1261

    Strauss 3rd JF 2003 Some new thoughts on the pathophysiology and genetics of polycystic ovarian syndrome. Ann NY Acad Sci 997:42–48(Claire Glister, Stephenie)