Estrogen receptor beta is required for optimal cAMP production in mouse granulosa cells

Abstract

Granulosa cells of preovulatory follicles differentiate in response to FSH, and this differentiation is augmented by estradiol. We have previously shown that FSH-mediated granulosa cell differentiation requires functional estrogen receptor-beta (ERbeta) by demonstrating that the granulosa cells of ERbeta(-/-) FSH-treated mice are unable to maximally induce expression of the LH receptor (an indicator of granulosa cell differentiation) compared with ERbeta(+/+) controls. As a result, FSH-primed ERbeta(-/-) granulosa cells exhibit a reduced response to a subsequent ovulatory dose of LH. In this study, we further characterized the attenuated response of ERbeta(-/-) granulosa cells to stimulation by LH and FSH using isolated mouse granulosa cells and primary granulosa cell cultures. We observed a 50% reduction in cAMP levels in cultured ERbeta(-/-) granulosa cells exposed to LH compared with ERbeta(+/+) controls. We also observed an attenuated genomic response in granulosa cells isolated from FSH-primed ERbeta(-/-) mice compared with ERbeta(+/+) controls. Our data indicate that this attenuated response may result from inadequate levels of cAMP, because cAMP levels in cultured ERbeta(-/-) granulosa cells exposed to forskolin were approximately 50% lower than in ERbeta(+/+) granulosa cells. Phosphorylation of cAMP regulatory element binding protein, an indicator of protein kinase A activity, was also reduced in FSH-treated ERbeta(-/-) granulosa cells compared with ERbeta(+/+) controls. These are the first data to indicate that ERbeta plays a role in the induction of the cAMP pathway in mouse granulosa cells and that disruption of proper ERbeta signaling associated with this pathway may cause negative effects on ovulation and fertility.

Figure 1

Granulosa cell expression of Lhcgr after PMSG treatment of ERβ^+/+, ERβ^+/−, and ERβ^−/− mice. A, Immature mice were treated with saline or PMSG (3.25 IU) for 48 h, and ovaries were collected and processed for in situ hybridization to detect Lhcgr gene expression. GC, Granulosa cells; TC, thecal cells. B, Immature mice were treated with □ saline or ▪ PMSG (3.25 IU) for 48 h, granulosa cells were isolated and pooled, and total RNA was isolated for Lhcgr gene expression analysis by qRT-PCR compared with an Rpl7 control (±sem of three independent experiments). Analysis by ANOVA found a significant interaction (P = 0.002) between PMSG treatment and genotype, indicating that the response to PMSG correlates with the dosage of the ERβ allele. A Bonferroni post test was also used to compare replicate means by genotype: a, P < 0.001. C, Granulosa cells were isolated from immature mice, cultured for 2 h as for all other in vitro assays (see Materials and Methods), and total RNA isolated for Fshr gene expression analysis by qRT-PCR compared with an Rpl7 control (±sem of four independent experiments). veh, Vehicle.

Figure 2

Effect of LH and forskolin treatment on gene expression and cAMP levels in FSH-primed primary granulosa cells. Granulosa cells were isolated from untreated, immature mice, pooled, and then cultured in FSH (50 ng/ml) + T (5 ng/ml) for 48 h, or FSH (50 ng/ml) + T (5 ng/ml) followed by LH (100 ng/ml) or forskolin (10 μm) for 4 h. Total RNA was isolated and analyzed by qRT-PCR for (A) Ptgs2 and (B) Pgr expression. Data are expressed as the average ± sem of four independent experiments. Analysis by ANOVA found a significant interaction between LH treatment and genotype for Pgr (P = 0.002), but not for Ptgs2 (P = 0.1), indicating that the magnitude of increased Pgr expression in response to LH is dependent on the genotype, i.e. on the dosage of the ERβ allele. No significant interaction between forskolin treatment and genotype was observed for either Pgr or Ptgs2. However, the effect of both LH and forskolin treatment on individual genotypes compared with vehicle was significant for Ptgs2 (LH, P = 0.02; forskolin, P = 0.002) and for Pgr (forskolin, P = 0.001). A Bonferroni post test was also used to compare replicate means by genotype: a, P < 0.05; b, P < 0.001; c, P < 0.01. C, Granulosa cells were isolated from untreated, immature mice, pooled, and then cultured in FSH (50 ng/ml) + T (5 ng/ml) for 48 h, or FSH + T (5 ng/ml) followed by LH (100 ng/ml, left panel) or forskolin (10 μm, right panel) for 0, 0.5, 1, or 2 h. Cells were immediately frozen on dry ice at indicated times and stored at −80 C until cAMP levels were determined by ELISA. Data are expressed as average cAMP (picomoles/100,000 cells) ± sem of three independent experiments. a, P = 0.002; b, P < 0.05. No statistical significance was observed in cAMP levels between forskolin-treated ERβ^+/+ and ERβ^−/− granulosa cells at any time point.

Figure 3

Effect of FSH, forskolin treatment, and cAMP analog treatment on Lhcgr expression in primary granulosa cells isolated from ERβ^+/+ and ERβ^−/− mice. Granulosa cells were isolated from untreated, immature mice, pooled, and then cultured in A: FSH (50 ng/ml) + T (5 ng/ml) or forskolin (10 μm) + T (5 ng/ml) for 48 h, or B: the cAMP analog, 8CPT (1 mm) + T (5 ng/ml) for 48 h. Total RNA was prepared and Lhcgr expression analyzed by qRT-PCR. Data are expressed as average Lhcgr expression compared with an Rpl7 control ± sem of four independent experiments. a, P < 0.01 comparing forskolin-treated wild-type and ERβ^−/− cells. veh, Vehicle.

Figure 4

Effect of FSH and forskolin on cAMP levels and phosphodiesterase expression in primary granulosa cells isolated from ERβ^+/+ and ERβ^−/− mice. Granulosa cells were isolated from untreated, immature mice, pooled, and then cultured in the absence or presence of the phosphodiesterase inhibitor, IBMX, for the times indicated: A and E, FSH (50 ng/ml) + T (5 ng/ml); B, forskolin (10 μm) + T (5 ng/ml); C, FSH + T (5 ng/ml) + IBMX; or D, forskolin (10 μm) + T (5 ng/ml) + IBMX for 0, 0.5, 1.0, and 2.0 h. Cells were immediately frozen on dry ice at indicated times and stored at −80 C until cAMP levels were determined by ELISA. E, Box plot of data in panel A indicating that the median cAMP levels are higher in wild-type cells than in ERβ^−/− cells after 0.5 and 1.0 h treatment with FSH + T in vitro. In a box plot, the box contains the middle 50% of the data. The upper and lower edges of the box indicate the 75th and the 25th percentiles of the dataset, respectively. The line in the box indicates the median value of the data, whereas the ends of the vertical lines indicate the minimum and maximum data values. F, Box plot (see panel E for explanation of a box plot) showing cAMP levels in untreated granulosa cells. Data are expressed as cAMP levels (picomoles/100,000 cells) ± sem of five (A, B, E, and F) or three (C and D) independent experiments. G and H, Granulosa cells were isolated from untreated, immature mice, pooled, and then cultured in serum-free medium for 2 h before harvest to mimic the culture conditions used for all in vitro studies before gonadotropin or other treatment. Total RNA was prepared and analyzed by qRT-PCR for Pde1c (G) and Pde1b (H) mRNA levels. Data are expressed as average gene expression compared with an Rpl7 control ± sem of four independent experiments. a, P < 0.02; b, P < 0.07; c, P < 0.03; *, P = 0.05.

Figure 5

pCREB levels are lower in FSH-stimulated ERβ^−/− granulosa cells than in ERβ^+/+ granulosa cells. A, Granulosa cells were isolated from untreated, immature mice, pooled, and then cultured in serum-free medium for 2 h before treatment with FSH (50 ng/ml) + T (5 ng/ml) for 0, 15, 30, or 60 min. Whole-cell protein extracts were prepared and analyzed for pCREB and CREB protein levels by Western blot analysis. B, Densitometric analysis of Western blot data indicating pCREB:CREB ratios at each time-point.

Figure 6

ERβ is required for maximal expression of many FSH-regulated genes identified by microarray analysis. Immature mice were treated with saline (□) or PMSG (▪) (3.25 IU) for 48 h, and granulosa cells were isolated and pooled, and total RNA was isolated for gene expression analysis by quantitative RT-PCR. Data are expressed as the average ± sem of three independent experiments. Analysis by ANOVA found a significant interaction between PMSG treatment and genotype for Comp (P = 0.002), Lrp11 (P = 0.0002), Car14 (P = 0.04), and Mro (P = 0.03), indicating that the magnitude of increased Comp, Lrp11, Car14, and Mro expression in response to PMSG is dependent on the genotype, i.e. on the presence of the ERβ allele. For Hsd3b1, Inhba, and Prkarb, no significant interaction between PMSG treatment and genotype was observed. However, the effect of PMSG treatment on individual genotypes compared with vehicle was considered significant for Hsd3b1 (P = 0.003), Inhba (P = 0.002), and Prkarb (P < 0.0001). A Bonferroni post test was also used to compare replicate means by genotype: a, P < 0.001; b, P < 0.01; c, P < 0.05.