Abstract
In the hen ovary, selection of a follicle into the preovulatory hierarchy occurs from a small cohort of prehierarchal (6–8 mm) follicles. Prior to follicle selection the granulosa layer remains in an undifferentiated state despite elevated follicle-stimulating hormone receptor (FSHR) expression. The present studies describe a role for bone morphogenetic protein 4 (BMP4) in supporting FSHR mRNA expression in granulosa cells from prehierarchal follicles and promoting differentiation at follicle selection. Culture of undifferentiated granulosa cells in culture medium alone resulted in a significant decline in levels of FSHR mRNA (by ∼80% compared to freshly collected cells). By comparison, granulosa cultured with BMP4 (10–100 ng/ml) maintained FSHR and expression at approximately in vivo levels. Because both granulosa and theca tissues from prehierarchal follicles express BMP4, it is suggested that BMP4 acts in a paracrine and/or autocrine fashion to support elevated FSHR expression prior to follicle selection. Granulosa cells cultured with BMP4 for 24 h also initiated FSH-induced cAMP production and indirectly initiated anti-Mullerian hormone (AMH), CYP11A, and STAR expression plus progesterone production. However, pretreatment with the BMP antagonist NOGGIN or the mitogen-activated protein kinase (MAPK) agonist transforming growth factor alpha attenuated or blocked each action promoted by BMP4. We conclude that prior to and immediately after selection, BMP4 serves to support FSHR expression within the granulosa layer, yet prior to selection, multiple factors (including inhibitory MAPK signaling, AMH, and BMP antagonists) can modulate FSHR expression and suppress FSH-mediated cell signaling to prevent granulosa cell differentiation prior to follicle selection.
Introduction
In the hen ovary, the daily selection of a follicle into the preovulatory hierarchy occurs from a small cohort of prehierarchal (6–8 mm) follicles. It has previously been established that the highest levels of follicle-stimulating hormone receptor (FSHR) expression during follicle development occur within the granulosa layer of such prehierarchal follicles [1, 2]. Nevertheless, FSH-induced signaling via cAMP at this stage of development is essentially absent [3]. Consequently, prior to selection, the granulosa layer remains in an undifferentiated state as shown by nondetectable to very low levels of CYP11A [4] and STAR mRNA [5]. This results in the inability of granulosa to produce progesterone (P4; the primary steroid produced by the granulosa layer) following a 3-h challenge of freshly collected cells with ovine FSH [3]. Luteinizing hormone (LH) receptor (LHR) mRNA expression within the granulosa layer is detected only after follicle selection and is initiated by cAMP signaling in response to FSH and possibly vasoactive intestinal peptide [6–8]. By comparison, the adjacent theca layer from prehierarchal follicles expresses readily detectable levels of both FSHR [1] and LHR mRNA [7], and responds to either FSH or LH treatment in vitro, with increased steroid production [9, 10]. Accordingly, it is evident that the LH- and FSH-mediated steroid production previously reported from incubated whole prehierarchal follicles [11, 12] is attributed to the theca layer.
Current studies in our laboratory are focused on identifying: 1) mechanisms that maintain the granulosa layer from follicles within the prehierarchal cohort in an undifferentiated state until follicle selection; and 2) factors that initiate FSH responsiveness and the consequent differentiation of the granulosa layer at follicle selection. In the hen ovary, several factors within the transforming growth factor beta (TGFβ) superfamily have been reported to directly promote FSHR mRNA expression in cultured granulosa from prehierarchal follicles. These factors include TGFβ [2, 13] and activin A [2, 13, 14]. In addition, bone morphogenetic protein 6 (BMP6) was found to promote FSHR expression following 24 h of culture, and this resulted in the acquisition of FSH responsiveness plus the initiation of granulosa cell differentiation [15]. Treatment with vitamin D was also found to increase FSHR expression; however, that effect may be indirectly attributed to an inhibition of anti-Mullerian hormone (AMH) activity [16].
There is also much evidence that cell signaling via the mitogen-activated protein kinase (MAPK) pathway maintains the granulosa layer from prehierarchal follicles in an undifferentiated state [17, 18]. In this regard, BMP2 was found to suppress TGFβ1-induced FSHR expression, and this inhibitory effect was attributed to enhanced epidermal growth factor receptor ligand (EGFRL) expression and inhibitory MAPK signaling [19]. Moreover, AMH has been implicated in restricting early follicle development and FSHR expression [20, 21], yet perhaps paradoxically, BMP6 was found to enhance AMH expression [15]. Accordingly, it has been proposed that prior to selection, the maintenance of prehierarchal follicles requires a balance between multiple differentiation-inducing versus -inhibiting factors.
A previous study by Onagbesan et al. [22] documented the expression of BMP receptors (BMPR-IA, -IB, and II) together with BMP4 in hen granulosa and theca layers. In addition, both Onagbesan et al. [22] and Elis et al. [23] examined the actions of BMP4 on cell proliferation, STAR protein expression, and P4 production in differentiated granulosa tissue from preovulatory (F1, F2, and F3 plus F4) follicles, but those studies did not evaluate potential effects on gonadotropin receptor expression. Moreover, to our knowledge, the effects of BMP4 in undifferentiated granulosa have not been examined. In the present studies, we focused on a role for BMP4 in supporting FSHR expression and promoting differentiation (e.g., CYP11A and STAR expression plus P4 production) within the granulosa layer at selection. In addition, we evaluated the potential for a BMP antagonist, NOGGIN, or inhibitory MAPK signaling to modulate the actions of BMP4 within the granulosa layer from prehierarchal follicles.
Materials and Methods
Animals and Reagents
Single-comb white Leghorn hens 34–55 weeks of age and laying sequences of five or more eggs were used in all studies described. Hens were housed individually in laying batteries, with continuous access to feed and water, under a controlled photoperiod of 15L:9D (lights-on at 0130 h). Hens were euthanized by cervical dislocation 5–8 h after a mid-sequence ovulation, and the ovary was immediately removed and placed in ice-cold 1% NaCl solution until granulosa and theca tissues were collected. All procedures described herein were reviewed and approved by the Pennsylvania State University Institutional Animal Care and Use Committees and were performed in accordance with The Guiding Principles for the Care and Use of Laboratory Animals.
Recombinant human BMP4 was purchased from PeproTech and consists of the biologically active 113 amino acid homodimeric glycoprotein corresponding to residues 296–408 of the full-length human BMP4 precursor. The corresponding amino acid sequence for Gallus gallus BMP4 is predicted to be 94% identical (95% similar) to the recombinant human BMP4 sequence, and all seven Cys residues are conserved between the predicted G. gallus BMP4 sequence and the recombinant human BMP4. Recombinant human NOGGIN and TGFα proteins were purchased from PeproTech, while recombinant human FSH was provided by the National Hormone and Peptide Program.
Granulosa Cell Cultures
Ovarian follicles were grouped by stage of maturation, and granulosa cell layers were collected and dispersed for culture as previously described [2, 3]. Theca tissues were immediately frozen and later processed for total RNA. In some instances, an aliquot of collagenase-dispersed granulosa cells was immediately frozen at −20°C (T0). Granulosa cells (approximately 500 000 cells per well) from 6- to 8-mm follicles or the second largest (F2) follicle were preplated for 3–4 h at 40°C in an atmosphere of 95% air: 5% CO2 in 12-well polystyrene culture plates (Beckton Dickinson) containing 1 ml of complete medium, which consisted of Dulbecco modified Eagle medium/high glucose (DMEM; Thermo Scientific HyClone) containing 2.5% FBS (PAA Laboratories), 0.1 mM nonessential amino acids (Invitrogen), plus 1% antibiotic-antimycotic mixture (Invitrogen). In initial experiments, granulosa cells were cultured with complete medium in the absence or presence of BMP4 for 23 h and then challenged (after replacement with fresh medium) with 10 ng FSH/ml [15] for 1 h (for cAMP) or 3 h (for P4). In follow-up studies, NOGGIN (100 ng/ml) [15] or TGFα (10 ng/ml) was added 1 h prior to BMP4. In addition, 1-h incubations were conducted with freshly collected granulosa cells from 6- to 8-mm follicles to measure FSH-induced cAMP. Approximately 5 × 105 cells in 1 ml of complete medium were placed in 12- × 75-mm polypropylene tubes and incubated in a shaking water bath at 40°C in ambient air. In all instances, the cell content of cAMP was measured from granulosa cells cultured in the presence of 10 mM 3-isobutyl-1-methylxanthine (IBMX; Sigma-Aldrich).
STAR Promoter Activity
Chicken STAR promoter-luciferase construct [24] was generously provided by Dr. T. Yoshimura (Nagoya University, Japan). Briefly, the STAR promoter-luciferase construct was cotransfected with Renilla luciferase (as an internal control for transfection efficiency) into undifferentiated granulosa tissue from 6- to 8-mm prehierarchal follicles. Cell transfections were accomplished by using Lipofectamine 2000 transfection reagent, as described by the manufacturer (Invitrogen). Briefly, 200 000 cells were seeded into 12-well plates and cultured overnight in complete medium. The next morning, cells were washed with Hanks balanced salt solution (Thermo Scientific HyClone), which was then replaced with 300 μl of serum- and antibiotic-free transfection medium. For each transfection, 1 μl of Lipofectamine and 1 μg of STAR promoter-luciferase in pGL3-Basic vector DNA with 10 ng of pRL-SV40 was incubated in 200 μl of the transfection medium for 20 min. Following this incubation, the entire 200-μl amount was added to each well. Cells were cultured for 5 h, followed by the addition of 500 μl of cell culture medium containing 5% FBS and then cultured for an additional 21 h without or with 10 ng of BMP4/ml. BMP4 cultured cells were subsequently treated without or with 10 ng TGFα/ml for 20 min and then 10 ng of FSH/ml for 3 h. Luciferase activity was measured using the dual-luciferase reporter assay system according to the manufacturer's protocol (catalog no. E1910; Promega) and a FLUOstar Omega plate reader (BMG Labtech).
PCR Analyses
Total RNA was isolated from freshly collected and cultured granulosa cells at different stages of follicle development, using Trizol reagent (Invitrogen). Purity and concentration of RNA were determined using a NanoDrop spectrophotometer (Thermo Scientific). Prior to RT reaction, isolated RNAs (1.5 μg per reaction) were treated with DNase I (Promega) to eliminate contaminating genomic DNA. RT and PCR analysis were conducted as previously described [15]. Briefly, randomly primed, reverse transcribed cDNA synthesis reactions were performed using the RT system (Promega). For negative RT samples, all components of cDNA synthesis were used but lacked the reverse transcriptase enzyme to ensure the absence of genomic DNA contamination. For water control samples, all components of the RT-PCR or real-time PCR reactions were added, but water was substituted for the template to ensure the lack of primer contamination. For PCR analysis, the primers, dNTP mixture (Promega), Taq polymerase (New England Biolabs), and 100 ng of template cDNA were added to a 30-μl reaction volume. Forward and reverse primers directed toward G. gallus BMP4, FSHR, and CYP11A mRNA and 18S rRNA are described in Table 1. Final concentrations of the sense and antisense primers were determined for each primer pair based upon optimal amplification efficiency. Reactions were conducted with the following conditions: 2 min at 94°C, followed by 40 cycles each for 15 sec at 94°C, 30 sec at 56°C, and 60 sec at 72°C, followed by 7 min at 72°C and completed using the 7800 Fast thermocycler (Applied Biosystems). All primer pairs were validated for target specificity by evaluating the melting curve. In addition, the amplified product was visualized by ethidium bromide staining after running PCR products on a 2% agarose gel (Invitrogen). The product was extracted from the gel, and the identity was verified by sequencing.
Primer pairs for G. gallus BMP4, FSHR, CYP11A, AMH, and 18S rRNA.
| Target | Accession no. | Sequence | Position |
|---|---|---|---|
| BMP4 | NM_205237 | 5′-AGACGTTGGTCAACTCCGTCAACT-3′ | Fwd |
| 5′-TCCACCACCATCTCCTGGTAGTTT-3′ | Rev | ||
| FSHR | NM_205079 | 5′-TTAATTCCTGTGCTAACCCTTTCC-3′ | Fwd |
| 5′-CCAAACTTGCTCAACAGAATGAAG-3′ | Rev | ||
| CYP11A | NM_001001756 | 5′-ACTTCAAGGGACTGAGCTTTGGGT-3′ | Fwd |
| 5′-AGTTCTCCAGGATGTGCATGAGGA-3′ | Rev | ||
| AMH | NM_205030 | 5′-CCCCTCTGTCCCTCATGGA-3′ | Fwd |
| 5′-CGTCATCCTGGTGAAACACTTC-3′ | Rev | ||
| 18S rRNA | AF173612 | 5′-TTAAGTCCCTGCCCTTTGTACAC-3′ | Fwd |
| 5′-CGATCCGAGGAACCTCACTAAAC-3′ | Rev |
| Target | Accession no. | Sequence | Position |
|---|---|---|---|
| BMP4 | NM_205237 | 5′-AGACGTTGGTCAACTCCGTCAACT-3′ | Fwd |
| 5′-TCCACCACCATCTCCTGGTAGTTT-3′ | Rev | ||
| FSHR | NM_205079 | 5′-TTAATTCCTGTGCTAACCCTTTCC-3′ | Fwd |
| 5′-CCAAACTTGCTCAACAGAATGAAG-3′ | Rev | ||
| CYP11A | NM_001001756 | 5′-ACTTCAAGGGACTGAGCTTTGGGT-3′ | Fwd |
| 5′-AGTTCTCCAGGATGTGCATGAGGA-3′ | Rev | ||
| AMH | NM_205030 | 5′-CCCCTCTGTCCCTCATGGA-3′ | Fwd |
| 5′-CGTCATCCTGGTGAAACACTTC-3′ | Rev | ||
| 18S rRNA | AF173612 | 5′-TTAAGTCCCTGCCCTTTGTACAC-3′ | Fwd |
| 5′-CGATCCGAGGAACCTCACTAAAC-3′ | Rev |
Primer pairs for G. gallus BMP4, FSHR, CYP11A, AMH, and 18S rRNA.
| Target | Accession no. | Sequence | Position |
|---|---|---|---|
| BMP4 | NM_205237 | 5′-AGACGTTGGTCAACTCCGTCAACT-3′ | Fwd |
| 5′-TCCACCACCATCTCCTGGTAGTTT-3′ | Rev | ||
| FSHR | NM_205079 | 5′-TTAATTCCTGTGCTAACCCTTTCC-3′ | Fwd |
| 5′-CCAAACTTGCTCAACAGAATGAAG-3′ | Rev | ||
| CYP11A | NM_001001756 | 5′-ACTTCAAGGGACTGAGCTTTGGGT-3′ | Fwd |
| 5′-AGTTCTCCAGGATGTGCATGAGGA-3′ | Rev | ||
| AMH | NM_205030 | 5′-CCCCTCTGTCCCTCATGGA-3′ | Fwd |
| 5′-CGTCATCCTGGTGAAACACTTC-3′ | Rev | ||
| 18S rRNA | AF173612 | 5′-TTAAGTCCCTGCCCTTTGTACAC-3′ | Fwd |
| 5′-CGATCCGAGGAACCTCACTAAAC-3′ | Rev |
| Target | Accession no. | Sequence | Position |
|---|---|---|---|
| BMP4 | NM_205237 | 5′-AGACGTTGGTCAACTCCGTCAACT-3′ | Fwd |
| 5′-TCCACCACCATCTCCTGGTAGTTT-3′ | Rev | ||
| FSHR | NM_205079 | 5′-TTAATTCCTGTGCTAACCCTTTCC-3′ | Fwd |
| 5′-CCAAACTTGCTCAACAGAATGAAG-3′ | Rev | ||
| CYP11A | NM_001001756 | 5′-ACTTCAAGGGACTGAGCTTTGGGT-3′ | Fwd |
| 5′-AGTTCTCCAGGATGTGCATGAGGA-3′ | Rev | ||
| AMH | NM_205030 | 5′-CCCCTCTGTCCCTCATGGA-3′ | Fwd |
| 5′-CGTCATCCTGGTGAAACACTTC-3′ | Rev | ||
| 18S rRNA | AF173612 | 5′-TTAAGTCCCTGCCCTTTGTACAC-3′ | Fwd |
| 5′-CGATCCGAGGAACCTCACTAAAC-3′ | Rev |
For real-time PCR, primers and 50 ng of cDNA template were added to the 10-μl total reaction volume, using the reagents provided in PerfeCTa SYBR Green FastMix Low ROX (Quanta Biosciences, Inc.). Reactions were completed with the 7500 Fast real-time PCR system (Applied Biosystems) under the following conditions: 30 sec at 95°C, followed by 40 cycles each for 3 sec at 95°C, 30 sec at 56°C, and 30 sec at 72°C. Melting curves were evaluated for each run. The cycle number at which the fluorescence exceeded a threshold level (Ct) was determined for each reaction (run in triplicate) by using the 7500 unit software (version 2.0.4). In the initial study, Ct values were used to roughly compare BMP4 and CYP11A levels in granulosa tissue and theca tissue. Quantification was accomplished by standardizing reactions to 18S rRNA and then using the ΔΔCt method [25].
cAMP Enzyme Immunoassay
Intracellular cAMP accumulation was measured by enzyme immunoassay (EIA; catalog no. 581001, Cayman Chemical Co.). Briefly, incubated cells or cell cultures were pretreated with 10 mM IBMX for 15 min and then treated or not with FSH (10 ng/ml) for 1 h. The medium was removed, and cells were lysed in 0.1 M HCl by repeated pipetting and then incubated for 20 min at room temperature. After centrifugation at 1000 × g for 10 min, the supernatant was collected for the assay, and protein was quantified (Bio-Rad). The percent bound (B:B0) for all samples fell between 20% and 80%, while the mean sample intra- and interassay coefficients of variation were consistently less than 15%. Values for cAMP were standardized to total protein in each sample (fmol/μg protein).
Progesterone Enzyme Immunoassays
Progesterone in medium samples from prehierarchal follicles was quantified by EIA (catalog no. 582601; Cayman Chemical Co.) as previously described [15]. Medium samples from preovulatory follicle granulosa were assayed by EIA (catalog no. EA74; Oxford Biomedical Research). P4 antiserum is reported to cross-react with pregnenolone (2.5%), androstenedione (1.0%), and 17-hydroxy P4 (0.4%) and less than 0.5% for other progestins, androgens, or estrogens. Serial dilutions of medium samples over a 5-fold range produced a line parallel to the standard curve. Data are means ± SEM for the replicated experiments. The mean within-assay coefficient of variation for samples was less than 15%.
Immunoblot Analysis
Western blot analysis of STAR, phospho-SMAD1, and total SMAD1 protein was conducted as previously described [19]. Briefly, cells were homogenized in a protein lysis buffer (radio-immunoprecipitation assay [RIPA]; Santa Cruz Biotechnology) containing a cocktail of enzyme (including phosphatase) inhibitors (Sigma-Aldrich). The STAR antiserum was generously provided by Dr. Buck Hales (Southern Illinois University School of Medicine) and was used at a dilution of 1:5000 [26]. The rabbit phospho-SMAD1 (Ser463/465)/Smad5 (Ser463/465)/Smad8 (Ser426/428) polyclonal antibody (catalog no. 9511; Cell Signaling) was used at a dilution of 1:1000, while a rabbit polyclonal anti-SMAD1/5/8 antibody (catalog no. PA-41238; 1:6000 dilution; Thermo Scientific) was used for standardization. Incubations with primary antibodies were conducted overnight at 4°C with gentle agitation. The horseradish peroxidase-conjugated anti-rabbit IgG secondary antibody (diluted 1:10,000; Pierce) was incubated for 1 h at room temperature. Blots were incubated with Pierce ECL2 Western blotting substrate (Amersham Biosciences). Chemiluminescent signals were detected using the Storm 860 optical scanner (Amersham Biosciences), and the signals were analyzed using Image Quant TL software (Amersham Biosciences). The chemiluminescence signal intensity of each band was calculated using a local average background correction.
Data Analysis
Experiments were independently replicated a minimum of three times unless otherwise stated. Real-time PCR results were expressed as fold differences compared with an appropriate control tissue or treatment. The data from replicate experiments were analyzed by t-test or by one-way ANOVA followed by the Tukey or Duncan multiple range test. Medium P4 and STAR promoter activity from the combined replicate experiments were analyzed by one-way ANOVA, while cellular cAMP between treatments was compared by t-test.
Results
Analysis of tissues collected from follicles prior to selection (3–5 and 6–8 mm), immediately after selection (9–12 mm), and from a preovulatory follicle undergoing rapid growth (F5 follicle) established the fact that BMP4 levels progressively increase with stage of development in the granulosa layer (mean Ct values range from 25.1 in 3–5 mm follicles to 22.6 in the F5 follicle), but remain relatively unchanged within the theca layer (mean Ct varies from 25.3 in 3- to 5-mm follicles to 23.9 in the 9- to 12-mm follicle) (Fig. 1). CYP11A levels increase from very low in granulosa from 6–8 mm follicles (mean Ct, 25.0) to dramatically elevated levels shortly after selection (e.g., 9- to 12-mm follicle; mean Ct, 19.5). CYP11A levels in theca were readily detectable (mean Ct ranged from 22.0 in 3- to 5-mm follicle theca to 20.4 in the F5 theca) and essentially unchanged at the stages of development investigated.
BMP4 and CYP11A mRNA expression in granulosa and theca layers from prehierarchal (3–5 mm, 6–8 mm) and the two most recently selected (9–12 mm and F5 preovulatory) follicles. Data are fold differences ± SEM compared to 3–5 mm follicles. A,B,CP < 0.05; N.S. = not significant: P > 0.05.
A 1-h challenge of freshly collected granulosa cells from 6- to 8-mm follicles with 10 ng FSH/ml failed to induce a significant increase in intracellular levels of cAMP (Fig. 2, top panel). Similarly, an initial study with cells precultured for 23 h under control conditions (DMEM medium containing 2.5% FBS) established that a subsequent treatment with FSH for 1 h had no stimulatory effect on cAMP formation or FSHR expression compared to untreated (Con) cells (Fig. 2, bottom panel). Furthermore, pretreatment for 23 h with 10 ng BMP4/ml by itself failed to significantly increase cAMP formation, however, this dose of BMP4 promoted FSHR expression and facilitated FSH-induced cAMP accumulation.
Top) FSH (10 ng/ml) failed to induce significant cAMP formation after a 1-h incubation of freshly collected granulosa cells from prehierarchal follicles compared to untreated (Con) cells. Bottom) Effects of BMP4 (10 ng/ml) without and with FSH (10 ng/ml) on FSHR mRNA expression and intracellular cAMP in undifferentiated granulosa cells from prehierarchal follicles. Means ± SEM; n = 5 replicate cultures. A,BP < 0.05.
Culture of undifferentiated granulosa cells for 24 h with complete medium resulted in a decline in FSHR levels of approximately 80% compared to levels in freshly collected (T0) cells (Fig. 3). FSH treatment (10 ng/ml) during the last 3 h of culture failed to alter the extent of this decline in FSHR compared to control cultured cells. By comparison, treatment with 10–100 ng BMP4/ml throughout the culture period maintained FSHR levels at or slightly above preculture (T0) levels.
FSHR mRNA levels decline by ∼80% in undifferentiated granulosa cells cultured in DMEM plus 2.5% FBS for 24 h compared to freshly collected (T0) cells. By comparison, preculture with BMP4 (10–100 ng/ml) maintained levels of FSHR at or above T0 levels. Where indicated, FSH (10 ng/ml) was added during the last 3 h of culture. Data are fold differences versus T0 cells. n = 4 replicate experiments. In a planned comparison, T0 was compared to Con cultured cells by t-test. A,B,CP < 0.05; means ± SEM from four replicated cultures.
Moreover, FSH treatment alone during the last 3 h of culture did not increase levels of CYP11A expression or P4 production (Fig. 4) compared to true control cells. By comparison, cells precultured for 21 h with increasing doses of BMP4 produced an apparent biphasic effect on CYP11A expression following a 3-h challenge with 10 ng FSH/ml. Highest levels of expression were induced by 10 and 25 ng BMP4/ml. Similar to the pattern for CYP11A expression, P4 production following a 3-h treatment with FSH was significantly increased in cells precultured with 10 and 25 ng BMP4/ml.
BMP4 facilitated FSH-induced CYP11A expression (top) and progesterone production (bottom) in undifferentiated granulosa cells from prehierarchal follicles. Cells were cultured with BMP4 (10 ng/ml) for 21 h, followed by a 3-h challenge with FSH (10 ng/ml). CYP11A data are fold difference versus Con (untreated) cells. A,BP < 0.05; means ± SEM from four replicated cultures.
While preculture with 10 ng BMP4/ml followed by a 3-h challenge with FSH resulted in enhanced FSHR, CYP11A , and AMH expression (Fig. 5) plus initiated STAR expression and P4 production (Fig. 6), a 1-h pretreatment with either TGFα (10 ng/ml) or NOGGIN (100 ng/ml) prevented each of these responses. Pretreatment with NOGGIN blocked BMP4-induced SMAD1 phosphorylation (Fig. 7), while TGFα attenuated BMP4 and FSH-induced STAR promoter activity (Fig. 8, bottom panel). Moreover, following treatment with 10 ng BMP4/ml for 21 h, pretreatment with TGFα for as little as 20 min decreased FSH-induced cAMP accumulation by 35% (P = 0.022) (Fig. 8, top panel).
Pretreatment for 1 h with either the MAPK activator TGFα (10 ng/ml) or the BMP antagonist NOGGIN (100 ng/ml) prevented BMP4 (10 ng/ml for 23 h) induced FSHR, CYP11A, and AMH mRNA expression. Data are fold differences versus untreated cells (not shown). A,BP < 0.05; means ± SEM from n = 5–7 replicates.
Pretreatment of undifferentiated granulosa cells with TGFα (10 ng/ml) or NOGGIN (100 ng/ml) suppresses BMP4 plus FSH-induced STAR expression (top) and progesterone production (bottom). Western blotting was replicated once with similar results; progesterone, A,BP < 0.005; means ± SEM, n = 3 (TGFα) or 4 (NOGGIN) replicates.
Replicated Western blotting provided evidence that a 30-min pretreatment with NOGGIN (100 ng/ml), but not with TGFα (10 ng/ml), inhibits SMAD1 phosphorylation induced by a subsequent 30-min treatment with BMP4 (10 ng/ml).
Top) Pretreatment for 20 min with TGFα suppressed FSH-induced cAMP formation in undifferentiated granulosa cultured with BMP4 (10 ng/ml) for 21 h and then treated with TGFα (10 ng/ml) for 20 min and finally challenged with FSH (10 ng/ml) for 1 h. Bottom) Pretreatment with TGFα for 20 min reduced BMP4 (10 ng/ml) plus FSH (10 ng/ml)-induced STAR promoter activity. A,B,CP < 0.05; means ± SEM, n = 3 replicated assays.
Discussion
The comparatively low levels of CYP11A in granulosa from prehierarchal follicles (Fig. 1) are consistent with a previous report in which CYP11A expression was near-undetectable by Northern blot analysis until immediately after follicle selection (in 9- to 12-mm follicles) [4, 26]. These findings reflect the undifferentiated status of hen granulosa cells prior to follicle selection, and this is due primarily to the inability of FSH to induce cAMP signaling at this stage of follicle development (Fig. 2). Follicle selection in vivo is followed by the acquisition of FSH-responsiveness (defined herein as the ability of FSH to promote cAMP formation). Consequently, it is predicted that cell signaling via cAMP results in dramatically increased CYP11A expression (Fig. 1, in the 9- to 12-mm follicle), CYP11A enzyme activity [27], and the initiation of P4 production [3, 4]. This pattern of CYP11A expression and activity within the granulosa layer contrasts with readily detectable levels of CYP11A in the theca layer (Fig. 1) and the capacity to produce androstenedione and estradiol throughout follicle development [9, 12]. As levels of BMP4 expression essentially do not change within the theca layer during the stages of development investigated, the approximate 2- to 4-fold increase in BMP4 within the granulosa layer from 9–12 mm and the F5 follicle, compared to 6–8 mm follicles, may reflect an increasing availability of BMP4 protein within the granulosa layer subsequent to follicle selection.
Two previous studies have evaluated the actions of BMP4 in hen granulosa, but only in cells from preovulatory follicles. Onagbesan et al. [22] found that both 1 and 5 ng/ml doses of recombinant human BMP4 dramatically increased FSH-induced (and LH-induced) P4 production in granulosa from F3+F4 preovulatory follicles after a 48-h culture. In contrast, Elis et al. [23] examined the actions of 5 and 50 ng/ml human recombinant BMP4 in granulosa from F2 and F3+4 follicles cultured for 32 h and reported significant inhibitory effects of each dose of BMP on both FSH- and LH-induced P4 production, plus STAR protein expression. Because those two studies cultured differentiated granulosa cells with both BMP4 and FSH throughout a prolonged culture period (32–48 h), it is not possible to directly compare results from either report to the present study. Nevertheless, the differentiation-enhancing actions of BMP4 reported herein using granulosa cells from prehierarchal follicles and the P4-inducing effects in granulosa from the F2 preovulatory follicle (Supplemental Fig. S1; available online at www.biolreprod.org) are more similar to the steroid promoting effects in preovulatory follicles described by Onagbesan et al. [22].
Consistent with a previous report [2], levels of FSHR mRNA in undifferentiated granulosa cells decreased significantly (by 80% in the present study) after 24 h of culture in medium supplemented with only 2.5% FBS (Fig. 3). Presumably, this dramatic reduction in FSHR mRNA is paralleled by a similar reduction of functional FSHR protein; unfortunately a sensitive and reliable antiserum generated against the G. gallus FSHR is not currently available to confirm this. Several members of the TGFβ superfamily have already been determined to directly induce FSHR expression in hen granulosa cells, including TGFββ1, ACTIVIN A and BMP6 [13–15]. Given that the 10, 25 and 100 ng/ml doses of BMP4 maintained levels of FSHR at or above preculture (T0) levels (Fig. 3), the results reported herein indicate that BMP4 represents an additional paracrine/autocrine factor capable of supporting FSHR expression prior and immediately subsequent to follicle selection. Interestingly, the highest dose of BMP4 utilized in the present studies (100 ng/ml) promoted less than maximal CYP11A expression and somewhat reduced P4 production (Fig. 4), yet the mechanism and the physiological implications (if any) for these observations are not readily apparent.
Not unexpectedly, neither the 1-h nor the 3-h challenge with 10 ng FSH/ml enhanced FSHR levels in the absence of BMP treatment (Figs. 2 and 3). Moreover, in both freshly incubated and control cultures of granulosa a 1-h challenge with FSH failed to induce any significant cAMP production (Fig. 2). It is possible that this lack of FSH-responsiveness could be attributed to an inherent inability of undifferentiated granulosa to translate the available FSHR mRNA to functional FSHR protein. However, it is more likely that the absence of cAMP formation is due to FSHR desensitization as previously inferred [3]. The exact nature of this desensitization in granulosa cells prior to follicle selection is currently under investigation.
Among the earliest events resulting from follicle selection in the hen is the acquisition of FSH responsiveness and initiation of differentiation within the granulosa layer. Results presented herein indicate that paracrine and/or autocrine signaling by BMP4 is capable of maintaining levels of FSHR expression during a 24 h culture (Fig. 3), and this action, in vivo, is likely supported by several additional TGFβ family members [13–15]. Moreover, cells cultured with BMP4 for 24 h acquire FSH-responsiveness and initiate cAMP production (Fig. 2), CYP11A and STAR expression plus P4 production (Figs. 4 and 6). Note, however, that the levels of BMP4-mediated P4 production (a maximum of ∼1 ng/μg protein/3 h) (Fig. 4) are considerably less than those produced by differentiated granulosa cultured under comparable conditions (∼30 ng/μg protein; Supplemental Fig. S1). This result is consistent with the increasing levels of P4 production that are observed as a follicle matures and progresses through the preovulatory hierarchy [28].
In light of elevated FSHR mRNA levels in 6–8 mm follicles, in vivo, it is perhaps unexpected that the granulosa layer remains in an undifferentiated state until selection [29]. Nevertheless, there also exist several potential inhibitory factors can counteract the differentiation-promoting effects of BMP4 and related factors. For instance, activation of MAPK signaling by several different EGFRL, in vitro, has previously been shown to modulate the status of differentiation in cultured granulosa [2, 17]. In the present studies TGFα treatment of cultured cells blocked the stimulatory effects of BMP4 on FSHR expression and precluded CYP11A and STAR expression as well as P4 production (Figs. 5 and 6). Even a short-term treatment with TGFα attenuated cAMP formation and had an inhibitory action at the level of the STAR promoter (Fig. 8). Note that each of these targets or processes downstream of the FSHR is primarily, if not entirely, dependent upon cell signaling via the adenylyl cyclase/cAMP pathway [4, 7].
It is also well established that the biological activity of several BMPs can be modulated by one or more extracellular BMP antagonist [30–31]. We have previously reported that one such antagonist, NOGGIN, prevents the inhibitory actions of BMP2 on FSHR expression and differentiation in cultured granulosa [19]. This inhibitory action occurs by preventing BMPs from binding to their receptors. In the present study, pretreatment with NOGGIN inhibited the differentiation promoting actions of BMP4 (Figs. 5 and 6) by preventing SMAD phosphorylation, an event required to initiate cell signaling (Fig. 7). Moreover, both BMP4 (Fig. 5) and BMP6 [15] promote AMH mRNA expression in undifferentiated granulosa cells, in vitro. Levels of AMH expression within the granulosa layer, in vivo, are highest during early follicle development (1–2 mm follicles), progressively decrease up to the 6–8 mm stage of development, and remain at very low levels after follicle selection [15, 20]. AMH has been suggested to play a role in regulating FSHR expression and/or decreasing FSH-responsiveness in hen follicles [32], much like has been reported in the mouse ovary [33]. Importantly, Johnson et al. [20] also reported a significant inhibitory effect of oocyte-conditioned medium on AMH expression, which implies that one or more factors produced by the oocyte may participate in regulating FSH-responsiveness prior to and at follicle selection.
In summary, the results reported herein provide evidence that BMP4 serves to support FSHR expression within the granulosa layer of prehierarchal follicles. Our current model predicts that one or more TGFβ family member (including TGFβ, ACTIVIN A and BMP6) promotes and maintains elevated FSHR expression and, following overnight culture (with an attenuation or absence of antagonists) can initiate FSH-induced cAMP production. Nevertheless, in vivo, these stimulatory factors are balanced by one or more inhibitory mechanisms that modulate FSHR expression and preclude FSH-induced signaling via cAMP until after follicle selection. The mechanisms that eventually attenuate inhibitory MAPK signaling, NOGGIN bioactivity and/or AMH expression and activity within the granulosa layer at follicle selection remain the subject of current investigations.
References
Author notes
Supported by National Science Foundation grant IOS-0968784 and the Walther H. Ott Endowment. Presented in part at the 45th Annual Meeting of the Society for the Study of Reproduction, August 12–15, 2012, University Park, Pennsylvania.
