Ovulation Involves the Luteinizing Hormone-Dependent Activation of G_q/11 in Granulosa Cells

Abstract

The LH receptor (LHR) activates several families of heterotrimeric G proteins, but only the activation of Gs and subsequent generation of cAMP are universally accepted as important mediators of LH actions. To examine the involvement of the G_q/11 family on the actions of LH, we crossed Cyp19Cre and Gα_q^f/f;Gα₁₁^−/− mice to generate mice with a granulosa cell-specific deletion of Gα_q in the context of a global deletion of Gα₁₁. Granulosa cells from Gα_q^f/f;Gα₁₁^−/−;Cre⁺ mice have barely detectable levels of Gα_q/11, have a normal complement of LHR, and respond to LHR activation with a transient increase in cAMP accumulation, but they fail to respond with increased inositol phosphate accumulation, an index of the activation of Gα_q/11. The LHR-provoked resumption of meiosis, cumulus expansion, and luteinization are normal. However, the Gα_q^f/f;Gα₁₁^−/−;Cre⁺ mice display severe subfertility because many of the oocytes destined for ovulation become entrapped in preovulatory follicles or corpora lutea. Because follicular rupture is known to be dependent on the expression of the progesterone receptor (Pgr), we examined the LHR-induced expression of Pgr and 4 of its target genes (Adamts-1, Ctsl1, Edn2, and Prkg2). These actions of the LHR were impaired in the ovaries of the Gα_q^f/f;Gα₁₁^−/−;Cre⁺ mice. We conclude that the defect in follicular rupture is secondary to the failure of the LHR to fully induce the expression of the Pgr. This is the first conclusive evidence for the physiological importance of the activation of G_q/11 by the LHR and for the involvement of Gα_q/11 in ovulation.

Ovulation is a complex process that can be subdivided into at least 3 sequential steps: the resumption of meiosis of the oocyte, the expansion of cumulus granulosa cells, and finally the rupture of the follicular wall that allows for the release of cumulus/oocyte complexes. These events are triggered by the binding of LH to the LH receptor (LHR) in mural granulosa cells, a signal that is propagated to LHR-negative cumulus cells in a paracrine fashion through the epidermal growth factor (EGF) network and by gap junction-mediated diffusion of intracellular cGMP to LHR-negative oocytes.

Oocytes are maintained in meiotic arrest by the high level of intraoocyte cAMP generated by endogenous G_s-coupled receptors (1, 2), but a cGMP-mediated inhibition of phosphodiesterase 3A is essential for the maintenance of high intraoocyte cAMP and thus meiotic arrest. The required cGMP is derived from intracellular cGMP generated in the somatic cells that diffuses into the oocyte through gap junctions. The generation of cGMP in the somatic compartment is in turn controlled by the paracrine actions of atrial natriuretic peptides generated in mural granulosa cells that engage the atrial natriuretic peptide receptors present in mural and cumulus cells (3–5).

The binding of LH to the LHR in mural granulosa cells activates G_s and increases cAMP accumulation (6, 7). An important consequence of the activation of this signaling pathway is the increased expression of 3 EGF-like growth factors (amphiregulin, epiregulin, and betacellulin) that are released in a soluble form and transactivate the EGF receptor (EGFR) in mural and cumulus cells (6). This results in the activation of the ERK1/2 cascade, a process that is essential for the expression of several genes, including Has2, Ptgs2, and Tnfaip6, which are expressed in mural and cumulus cells and are necessary for cumulus expansion (8, 9). Acting through the EGF network (but perhaps not exclusively), the LHR also decreases cGMP levels in mural and cumulus granulosa cells, and an ERK1/2-dependent closure of gap junctions between mural and cumulus cells ultimately lowers cGMP levels in the somatic compartment and in the oocyte. The decrease in intraoocyte cGMP activates phosphodiesterase 3A and reduces intraoocyte cAMP levels, thus allowing meiosis to resume (3, 4, 10–12).

The rupture of the follicular wall, the last step in the process of ovulation, is also an LH-dependent event, and it requires increased expression and activation of the progesterone receptor (PGR) in mural granulosa cells (13–15). The activated PGR in turn enhances the expression of a complex gene network including other nuclear receptors, transcription factors, and extracellular proteases that degrade the follicular wall (14, 15). Pgr null mice are anovulatory because of a defect in follicular rupture (16). Some of the ovulatory actions of the PGR are mediated by A disintegrin and metalloproteinase with thrombospondin-like repeats 1 (ADAMTS-1), an extracellular protease synthesized and secreted by mural granulosa cells in response to PGR target activation (17), because Adamts-1 null mice have a partial defect in follicular rupture (18).

Despite the importance of follicular rupture to the ovulatory response, the second-messenger pathways that mediate the LHR-dependent increase in the expression of the PGR have not been defined. The most relevant information on this pathway comes from studies done in primary cultures of granulosa cell showing that optimal induction of Pgr expression can be achieved only by simultaneous pharmacological activation of the protein kinase A and protein kinase C pathways (19). This is an interesting finding because, in addition to activating adenylyl cyclase, engagement of the LHR in granulosa cells (20–23) or in heterologous cells expressing the recombinant LHR (24–26) also results in the activation of phospholipase C and the generation of inositol phosphates and diacylglycerol, which activate protein kinase C. These findings suggest that induction of the Pgr and ovulation involve the LHR-dependent activation of G protein families distinct from Gs.

To investigate the role of the Gq/11 family of G proteins on ovulation and female fertility, we used a Cre/LoxP approach to generate mice with a granulosa cell-specific deletion of Gα_q in the context of a global deletion of Gα₁₁. The reproductive phenotype of Gα_q^f/f;Gα₁₁^−/−;Cre⁺ mice resembles that of the Pgr^−/− (16) or Adamts1^−/− mice (18) in which many of the oocytes destined for ovulation become entrapped in preovulatory follicles or corpora lutea. Moreover, the LHR-induced expression of Pgr and 4 of its target genes (Adamts-1, Ctsl1, Edn2, and Prkg2) is reduced in the ovaries of the Gα_q^f/f;Gα₁₁^−/−;Cre⁺ mice.

Materials and Methods

Mice

Gα_q^f/f;Gα_q^−/− (27–29) and Cyp19Cre transgenic mice (9, 30) are in a C57BL/6 background. A colony of Gα_q^f/f;Gα_q^−/−;Cre⁺ mice was generated and subsequently maintained by crossing Gα_q^f/f;Gα_q^−/−;Cre⁻ females with Gα_q^f/f;Gα_q^−/−;Cre⁺ males. The resulting Gα_q^f/f;Gα_q^−/−;Cre⁺ and Gα_q^f/f;Gα_q^−/−;Cre⁻ females were used as experimental and control animals, respectively. Genotyping was done using tail genomic DNA and PCR amplification as described (9, 27–30). The segregation of the Cre allele in the female pups closely approximated Mendelian inheritance (47% Cre⁺ and 53% Cre⁻).

Most experiments were done using immature female mice (21–24 days old) that had been injected ip with 5 IU pregnant mare serum gonadotropin (PMSG; National Hormone and Peptide Program) or injected with 5 IU PMSG followed by an ip injection of an ovulatory dose (5 IU) of human chorionic gonadotropin (hCG; Sigma Chemicals). Ovaries were collected at different times after the injection of hCG as indicated in the figure legends.

For the experiments using primary cultures of differentiated granulosa cells, the mice were injected sc with 4.5 μg of diethylstilbestrol (Sigma Chemicals) and 15 hours later with 5 IU of PMSG. Differentiated granulosa cells were isolated 44–48 hours after the injection of PMSG. All animal procedures were approved by the Institutional Animal Care and Use Committee for the University of Iowa.

Granulosa cell cultures

Differentiated granulosa cells were isolated (31) and plated in 24-well plates precoated with bovine fibronectin in medium (DMEM/F12 with 10 mM HEPES and 50 μg/mL gentamicin) containing 1 mg/mL BSA (23). We usually plated 3 wells (2–5 × 10⁵ cells/well) from each mouse in a total volume of 500 μL. The cells were allowed to attach for 4–5 hours and collected immediately for the Western blots of Gα_q/11 or the medium was replaced. The cells were then incubated with buffer only or with hCG (1 μg/mL) prior to the collection for cAMP determination or for Western blots of phosphorylated (P)-ERK1/2 or P-CCAAT/enhancer-binding protein/β.

For the hCG binding and inositol phosphate assays, the cells were plated in culture medium that was also supplemented with insulin (1 μg/mL), transferrin (1 μg/mL), and selenium (1 ng/mL) as well as 15 ng/mL of T (Sigma Chemicals) and 100 ng/mL or recombinant human FSH (EMD Serono) (23). After 24 hours the medium was replaced and the cells were cultured for an additional 24 hours prior to measuring ¹²⁵I-hCG binding as described earlier (23). For the inositol phosphate assays, the medium was replaced with the same medium supplemented with 4 μCi/mL of ³H-myoinositol, and the cells were cultured for an additional 24 hours. Finally, the labeling medium was aspirated and the cells were washed and used to measure the accumulation of ³H-inositol phosphates during 1 hour incubation in medium containing 20 mM LiCl without or with 1 μg/mL recombinant hCG as described earlier (23).

Histology

Ovaries were fixed in Bouin's fixative, embedded in paraffin, sectioned, and stained with hematoxylin and eosin using standard conditions. The sections were used to count the number of entrapped oocytes or assess germinal vesicle breakdown.

Entrapped oocytes were counted in every 10th serial section (5 μm) of an entire ovary 24 hours after an injection of an ovulatory dose of hCG to PMSG-primed mice. Depending on the genotype of the mice and the structure being scored (antral follicles or corpora lutea), this resulted in the counting of 0–30 entrapped oocytes. The counting was performed by 2 different individuals who were blinded to the genotypes of the mice. The averages were used.

Germinal vesicle breakdown, a measurement of the resumption of meiosis, was assessed 4 hours after an injection of an ovulatory dose of hCG to PMSG-primed mice using serial sections of the entire ovary. It resulted in the scoring of 15–35 oocytes/ovary, and it was done by a single individual who was blinded to the genotypes of the mice.

Fertilization and embryo transfer

The percent fertilization of ovulated oocytes was measured using immature female mice that had been injected with PMSG and hCG as described above and mated with Gα_q^f/f;Gα_q^−/−;Cre⁻ male mice of known fertility at the time of the hCG injection (in the afternoon). Females with a vaginal plug the following morning were euthanized 2 days later and the uterine horns were flushed with isotonic saline. Embryos and unfertilized oocytes were counted and the fertilization rate was calculated (32).

For the embryo transfer experiments, mature females (80–120 days old) were mated with vasectomized males in the afternoon. Females with a vaginal plug on the next day were used 3 days later as recipients of wt blastocysts. Blastocysts were obtained 3–4 days after the detection of a vaginal plug from wt immature females (21–26 days old) that were superovulated (injected with 5 IU PMSG followed by 5 IU hCG 44 hours later) and mated with males of known fertility in the afternoon. Twelve to 14 healthy blastocysts were implanted in each uterine horn and the implanted embryos were counted 3 days later (32).

Real-time PCR

Ovarian RNA was prepared from 2 ovaries homogenized in 2 mL of Trizol using a Polytron homogenizer (23). RNA from granulosa cell cultures was prepared from individual wells (see above) using the QIAGEN RNeasy minikit. Equal amounts of purified RNA were reversed transcribed using dN6 random primers (Applied Biosystems) and Moloney murine leukemia virus reverse transcriptase (Promega) as described elsewhere (33). Real-time PCR reactions were performed in a 25 μL volume using 300 nM of each primer and 1× iQ SYBR Green Super Mix (Bio-Rad Laboratories), and fluorescence was detected on the CFX96 real-time PCR detection system (Bio-Rad Laboratories). The conditions for the quantitative PCR (qPCR) were optimized for each gene, and the target gene expression was normalized to an internal control, glyceraldehyde-3-phosphate dehydrogenase (Gapdh). The primers used for Ereg, Areg, Btc, and Gapdh were previously described (34, 35). The primers for Adamts-1, Edn2, and Prkg2 were those described by Kim et al (36), and those for Cyp19a1 and the Lhcgr have been described by us (23). Those used for Ptgs2 are: forward, AAAGGTTCTTCTACGGAGAGAGTTCA, reverse, TGGGCAAAGAATGCAAACATC; Tnfaip6: forward, GCAGCTAGAGGCAGCCAGAA, reverse, ACTCTACCCTTGGCCATCCA; and Has2: forward, GGGAACTCAGACGACGACCTT, and reverse, GATGTACGTGGCCGATTTGTC. All data were analyzed using the method of Pfaffl (37).

Other methods

For Western blots, cells or ovaries were lysed in a buffer (150 mM NaCl; 50 mM Tris; 1 mM EDTA; 1% Nonidet P-40; 0.5% sodium deoxycholate; 0.1% sodium dodecyl sulfate; 1 mM Na₃VO₄; and 1 mM NaF, pH 7.4) supplemented with a commercial mixture of protease inhibitors (Roche). Homogenates were kept on ice for 30 minutes, with occasional mechanical disruption using a pipette, followed by a centrifugation at 13 000 × g for 10 minutes. Supernatants were collected and assayed for protein content using a bicinchoninic assay protein assay kit from Bio-Rad Laboratories. Cellular lysates were used immediately or stored at −80°C until used. Western blots were done as described earlier (34, 38, 39) using 2–20 μg of lysate protein. The solutions used to block and wash the membranes and perform the primary and secondary antibody incubations contained either 5% milk or 1%–5% BSA, depending on the antigen being detected. The length (1–18 hours) and temperature (4°C or ambient) of the incubation and the dilutions of the primary antibodies also varied, depending on the antigen being detected. The secondary antibody dilution (1:3000) and length and temperature of incubation (1 hour at room temperature) were constant, however. The secondary antimouse or antirabbit antibodies coupled to horseradish peroxidase were from Bio-Rad Laboratories (catalog number 170–6515). The immune complexes in the Western blots were eventually visualized using enhanced chemiluminescence (Pierce Chemical, Rockford, Illinois) and exposed to film. The source of antibodies was as follows: Gα_q/Gα₁₁ was from Santa Cruz Biotechnology (item number sc392). P-EGFR (item number 3377), P-ERK1/2 (item number 9122), P-CCAAT/enhancer-binding protein/β (item number 3084), AKT (item number 9272), and GAPDH (item number 2118) were from Cell Signaling, and Cyp11a1 (item number AB1244) was from Chemicon. The antibodies to the precursor and mature forms of ADAMTS-1 were generously provided by Dr JoAnne Richards (the Baylor College of Medicine) (17). The antibody to steroidogenic acute regulatory protein (StAR) was kindly provided by Dr Doug Stocco (Texas Tech University).

Progesterone assays were done using a commercially available kit from Cayman Chemicals.

Results

Deletion of Gα_q/11 in mouse granulosa cells impairs ovulation and fertility

A conditional deletion of Gα_q/11 in granulosa cells was accomplished by crossing Gα_q^f/f;Gα₁₁^−/− mice (27–29) with Cyp19Cre transgenic mice (30). The latter allows for Cre expression in granulosa cells and has been extensively used for the conditional modification of several genes in this cell type (9, 30, 40–42).

Granulosa cells from Gα_q^f/f;Gα₁₁^−/−,Cre⁺ mice have barely detectable levels of Gα_q/11 (Figure 1A), and they fail to respond to hCG with an increase in inositol phosphate accumulation, a readout of the activation of the G_q/11 family of G proteins (Figure 1B). However, they express normal levels of LHR (Figure 1C), and the rapid increase and subsequent decrease in intracellular cAMP accumulation elicited by hCG are unaffected (Figure 1D).

Figure 1.

Deletion of Gα_q/11 in granulosa cells selectively impairs the LHR-provoked inositol phosphate response. A, Western blots of differentiated granulosa cells were developed with an antibody that recognizes both Gα_q and Gα_11. AKT is shown as a loading control. The results of a representative blot are shown. B, Inositol phosphate accumulation in differentiated granulosa cells. Each bar is the mean ± SEM of 3 independent cultures. Asterisks indicate statistically significant differences (P < .05, t test). C, Binding of ¹²⁵I-hCG to differentiated granulosa cells. Each bar is the mean ± SEM of 3 independent cultures. D, Intracellular cAMP accumulation in differentiated granulosa cells. Each bar is the mean ± SEM of 3 independent cultures.

When paired with Gα_q^f/f;Gα₁₁^−/−,Cre⁻ males, the number of pups produced by Gα_q^f/f;Gα₁₁^−/−,Cre⁺ females during a 6-month period was reduced by approximately 85% (Figure 2A). The number of mice per litter (10.7 ± 2 vs 1.8 ± 4) and the number of litters per month (1 ± 0 vs 0.4 ± 0.08) were also reduced in the Gα_q^f/f;Gα₁₁^−/−,Cre⁺ females. The number of oocytes that were destined for ovulation and could be released by puncturing the ovaries 10 hours after the injection of an ovulatory dose of hCG was the same in both genotypes (Figure 2B). The actual number of oocytes released into the oviducts 15 or 20 hours after the injection of hCG, however, was reduced approximately 50% in the Gα_q^f/f;Gα₁₁^−/−,Cre⁺ females (Figure 2C).

Figure 2.

Deletion of Gα_q/11 in granulosa cells results in reduced fertility and ovulation. A, Male Gα_q^f/f;Gα₁₁^−/−,Cre⁻ mice of known fertility were housed with sexually mature, cycling Gα_q^f/f;Gα₁₁^−/−,Cre⁻ or Gα_q^f/f;Gα₁₁^−/−,Cre⁺ females (one pair/cage), and the number of pups born to each pair was recorded. Each point is the mean ± SEM of 6 pairs. Asterisks denote statistically significant differences (P < .05, t test). B, Immature PMSG-primed mice were injected with an ovulatory dose of hCG. Ovaries were collected 10 hours later and punctured repeatedly with a 27-gauge needle to release and count the oocytes present in preovulatory follicles. Each bar shows the mean ± SEM of 13–15 mice. C, Immature PMSG-primed mice were injected with an ovulatory dose of hCG. Ovulated oocytes were collected from the oviducts 15 or 20 hours later. Each bar shows the mean ± SEM of 24–29 mice. Asterisks denote statistically significant differences (P < .05, t test).

Deletion of Gα_q/11 in mouse granulosa cells does not affect the resumption of meiosis or cumulus expansion but it impairs follicular rupture

Some of the known events required for oocyte maturation leading to ovulation are dependent on the LHR-induced activation of the cAMP pathway and the ovarian EGF network (6, 7). Because of the reduction in ovulation documented above, we tested for a potential involvement of Gα_q/11 as mediators of these LHR-provoked events.

An ovulatory dose of hCG induced the expected increase in the ovarian expression of Areg, Ereg, and Btc in Gα_q^f/f;Gα₁₁^−/−,Cre⁺ mice. The expression of some of these growth factors tended to be lower in the Gα_q^f/f;Gα₁₁^−/−,Cre⁺ mice (Figure 3A), but this difference did not attain statistical significance. Moreover, downstream events such as the hCG-induced phosphorylation of the ovarian EGFR, ERK1/2, and CCAAT/enhancer-binding proteinβ, a prominent ERK1/2 target (9, 42), were not affected (Figure 3, B and C). The resumption of meiosis and the expansion of cumulus granulosa cells are 2 other events initiated by the ovulatory LH surge and dependent on the cAMP pathway and the ovarian EGF network (6, 7). The resumption of meiois (Figure 4A) and the expression of 3 genes (Ptgs2, Tnfaip6, and Has) that are routinely used as a measurement of cumulus expansion (Figure 4B) were also indistinguishable between Gα_q^f/f;Gα₁₁^−/−,Cre⁻, and Gα_q^f/f;Gα₁₁^−/−,Cre⁺ mice.

Figure 3.

Deletion of Gα_q/11 in granulosa cells does not affect the ability of the LHR to activate the ovarian EGF network. A, Immature PMSG-primed mice were injected with an ovulatory dose of hCG and the ovaries were collected 4 hours later. RNA was prepared and used to measure the expression of the 3 indicated genes by qPCR. The expression of each gene was normalized to the expression of Gapdh, and the data are expressed relative to the normalized expression of the same gene in the ovaries of Gα_q^f/f;Gα₁₁^−/−,Cre⁻ mice. The expression of each these 3 genes was at least 200-fold to 400-fold lower in mice injected with PMSG only. Each bar is the mean ± SEM of 7–10 mice. B, Ovaries of immature PMSG-primed mice were collected before or 2 hours after the injection of an ovulatory dose of hCG and were used to assess the phosphorylation of the EGFR by Western blotting. AKT was used as a loading control. The results of a representative experiment are shown. C, Differentiated granulosa cells were allowed to attach to the culture dishes for 4 hours. The cells were then incubated with or without hCG for 30 minutes and used to assess the phosphorylation of ERK1/2 or CCAAT/enhancer-binding protein/β by Western blotting. GAPDH was used as a loading control. The results of a representative experiment are shown.

Figure 4.

Deletion of Gα_q/11 in granulosa cells does not affect the resumption of meiosis or the expression of cumulus expansion genes A, Germinal vesicle breakdown (GVB), a measurement of the resumption of meiosis, was assessed microscopically in serial sections of ovaries collected 4 hours after an injection of an ovulatory dose of hCG to PMSG-primed mice. The data are presented as the percentage of total oocytes that had undergone germinal vesicle breakdown. Each bar shows the average ± SEM of 5 mice for each genotype. B, Immature PMSG-primed mice were injected with an ovulatory dose of hCG, and the ovaries were collected 3 hours later. RNA was prepared and used to measure the expression of the 3 indicated genes by qPCR. The expression of each gene was normalized to the expression of Gapdh, and the data are expressed relative to the normalized expression of the same gene in the ovaries of Gα_q^f/f;Gα₁₁^−/−,Cre⁻ mice. The expression of each of these 3 genes was 100- to 1000-fold lower in mice injected with PMSG only. Each bar is the mean ± SEM of 9–10 mice.

Having ruled out the involvement of Gα_q/11 on the resumption of meiosis and cumulus expansion, we concentrated on the last event, follicular rupture, required for ovulation. The ovaries of Gα_q^f/f;Gα₁₁^−/−,Cre⁺ mice have increased numbers of oocytes that remain trapped in antral follicles or corpora lutea 24 hours after injection of an ovulatory dose of hCG (Figure 5). Most of the trapped oocytes are in antral follicles rather than corpora lutea (compare the y-axis on the left and right graphs in Figure 5).

Figure 5.

Deletion of Gα_q/11 in granulosa cells increases oocyte entrapment. Immature PMSG-primed mice were injected with an ovulatory dose of hCG. Ovaries were collected 24 hours later. Each bar shows the mean ± SEM of 7–9 mice. Asterisks denote statistically significant differences (P < .05, t test). The micrographs show representative examples of oocytes trapped in antral follicles (left) or corpora lutea (right) of Gα_q^f/f;Gα₁₁^−/−,Cre⁺ mice.

Deletion of Gα_q/11 in mouse granulosa cells does not affect fertilization, luteinization, or implantation

Although the magnitude of the reduction in ovulation in the Gα_q^f/f;Gα₁₁^−/−,Cre⁺ mice is approximately 50% (Figure 2C), fertility is reduced by approximately 85% (Figure 2A). Because the assessment of ovulation was made using superovulated immature mice and the assessment of fertility was made using sexually mature cycling mice, the discrepant results could be due to the different experimental paradigms used or to an impairment of other LHR-mediated events. To address this possibility, we measured the extent of fertilization of the oocytes ovulated by the superovulated immature mice and found it to be the same between the 2 mouse genotypes (Supplemental Figure 1A, published on The Endocrine Society's Journals Online web site at http://mend.endojournals.org). Uterine receptivity was next determined by measuring the percent implantation of wt blastocysts in both mouse genotypes. These results are shown in Supplemental Figure 1B and show that this is also the same in both mouse genotypes.

Luteinization in the superovulated Gα_q^f/f;Gα₁₁^−/−,Cre⁺ mice also appeared normal as judged by the increased expression of ovarian Cyp11a1 and StAR (Supplemental Figure 2A) and serum progesterone attained after injection of an ovulatory dose of hCG (Supplemental Figure 2B). Serum progesterone was somewhat elevated in the Gα_q^f/f;Gα₁₁^−/−,Cre⁺ mice injected with PMSG only, however. Despite the increased serum progesterone in this group, other actions of PMSG were not affected, as illustrated by the similar induction of 2 PMSG-dependent genes (Cyp19a1 and Lhcgr) in the ovaries of both mouse genotypes (Supplemental Figure 3; also see Figure 1B for a functional assay of LHR expression) and the low expression of 2 luteinization markers (Cyp11a1 and StAR) in the ovaries of PMSG-injected mice (see −hCG lanes in Supplemental Figure 1A).

Deletion of Gα_q/11 in mouse granulosa cells impairs the ability of the LHR to induce expression of the PGR and PGR-dependent genes

When considered together, these data suggest that the reduced fertility of the female Gα_q^f/f;Gα₁₁^−/−,Cre⁺ mice (Figure 2A) is due to a decrease in ovulation (Figure 2C) secondary to a block in follicular rupture (Figure 5). Because follicular rupture requires an LHR-dependent increase in the expression of the ovarian Pgr (16), we examined this induction as well as the induction of Adamts-1, Ctsl1, Edn2, and Prkg2, 4 known ovarian targets of Pgr (15, 17, 18), in immature PMSG-primed mice injected with an ovulatory dose of hCG. Figure 6 shows that the hCG-induced expression of Pgr and 4 of its downstream targets is reduced in the Gα_q^f/f;Gα₁₁^−/−,Cre⁺ mice. The magnitude of this reduction (≥50%) was similar to that of the decrease in ovulation detected in immature PMSG-primed mice injected with an ovulatory dose of hCG (Figure 2C). We also examined the expression of the protein product of Adamts-1 and found that both the precursor and the mature form of ADAMTS-1 were also substantially decreased in the ovaries of Gα_q^f/f;Gα₁₁^−/−,Cre⁺ mice that had been injected with hCG (Figure 7). When compared with the ovaries of Gα_q^f/f;Gα₁₁^−/−,Cr⁻ mice, the levels of the precursor and mature form of ADAMTS-1 in the ovaries of Gα_q^f/f;Gα₁₁^−/−,Cre⁺ mice were 22% ± 9% and 21% ± 6% (n = 8), respectively.

Figure 6.

Deletion of Gα_q/11 in granulosa cells decreases the expression of the ovarian Pgr and Pgr-dependent genes. Immature PMSG-primed mice were used 48 hours after the injection of PMSG (labeled as −hCG) or injected with an ovulatory dose of hCG and used at the indicated times to measure the expression of the indicated genes by qPCR. The expression of all genes was normalized to the expression of Gapdh and the data are expressed relative to the normalized expression of the same gene in the ovaries of PMSG-injected Gα_q^f/f;Gα₁₁^−/−,Cre⁻ mice. Each bar is the mean ± SEM of 8–10 mice. Asterisks denote statistically significant differences (P < .05, t test).

Figure 7.

Deletion of Gα_q/11 in granulosa cells decreases the expression of ovarian ADAMTS-1. Immature PMSG-primed mice were used 48hours after the injection of PMSG (labeled −hCG) or injected with an ovulatory dose of hCG and used 12 hours later. Western blots were developed with antibodies that recognize the precursor (Pro-ADAMTS-1) or the mature (MP-ADAMTS-1) forms of ADAMTS-1. The results of a representative experiment showing 2 mice of each genotype for each hormonal treatment are shown.

Discussion

The deletion of Gα_q/11 in granulosa cells does not affect follicular development, oocyte maturation, cumulus expansion, fertilization, luteinization, or uterine receptivity, but it impairs follicular rupture, thus preventing the release of some of the oocytes that were destined for ovulation. Because follicular rupture is dependent on the increased expression of Adamts-1 by a pathway that involves the LHR-mediated induction and activation of the ovarian PGR (13–18, 43), we examined the ability of the LHR to activate this pathway. The increased expression of the Pgr and several of its downstream targets (Adamts-1, Ctsl1, Edn2, and Prkg2) induced by an ovulatory dose of hCG was impaired in the ovaries of Gα_q^f/f;Gα₁₁^−/−,Cre⁺ mice. Progesterone levels after an injection of hCG were normal, however. Based on these data and the similarity of the reproductive phenotype of female Gα_q^f/f;Gα₁₁^−/−,Cre⁺ (this paper), Pgr^−/− (16), and Adamts-1^−/− (18) mice, we conclude that the LHR-mediated activation of Gα_q/11 in granulosa cells is involved in follicular rupture because this G protein family is a mediator of the LHR-provoked induction of the granulosa cell Pgr.

We also note that the reduction in the precursor and mature forms of ADAMTS-1 is more pronounced than the reduction in Pgr or Adamts-1 expression, thus raising the possibility that G_q/11 may also affect the stability of ADAMTS-1. ADAMTS-1, however, is not the only Pgr target that mediates follicular rupture because Adamts-1^−/− mice display only a partial defect in ovulation (18). Cathepsin L, another extracellular protease, is also a Pgr target (16), but its involvement in follicular rupture has not been studied in detail. Like Adamts-1, however, ovarian Ctsl1 expression is also reduced in the Gα_q^f/f;Gα₁₁^−/−,Cre⁺ mice. Another Pgr target, Edn2 (and one of its downstream targets, Prkg2) are also decreased in the ovaries of Gα_q^f/f;Gα₁₁^−/−;Cre⁺ mice, but the involvement of Edn2 on follicular rupture and ovulation is unresolved (43–45).

When compared with the anovulatory phenotype of the Pgr^−/− mice (16), the partial decrease in ovulation and induction of the Pgr in the Gα_q^f/f;Gα₁₁^−/−,Cre⁺ mice suggests that Gα_q/11 is not the only family of G proteins that mediates the effects of the LHR on Pgr induction. The latter conclusion is supported by previous studies on the induction of the Pgr or a Pgr reporter gene in granulosa cell cultures by second-messenger analogs (19). Addition of phorbol-12-myristate-13-acetate has little or no effect on Pgr expression, whereas the addition of forskolin has a substantial stimulatory effect. The addition of forskolin and phorbol-12-myristate-13-acetate, however, acted synergistically to induce Pgr expression (19). Thus, a partial reduction in the induction of the ovarian Pgr and ovulation in the Gα_q^f/f;Gα₁₁^−/−,Cre⁺ mice injected with an ovulatory dose of hCG would be expected because the G_s/adenylyl cyclase/protein kinase A pathway is intact in these mice. Full induction of the Pgr and normal ovulation would not be possible, however, because of the genetic ablation of Gα_q/11 (Figure 8). Gα_q/11-dependent pathways could also impact ovulation in a Pgr-independent fashion by directly modulating gene transcription, protein stability, or posttranslational modifications of other proteins that participate in follicular rupture (Figure 8). Such an effect has already been demonstrated for the expression of Adamts-1 that is coordinately regulated by the LHR in Pgr-dependent and -independent fashions (46). Our data support the possibility that Gα_q/11 also impacts the expression of ADAMTS-1 at a posttranscriptional level because the reduction in the expression Adamts-1 is less pronounced than the reduction in the expression of its protein products (Figures 6 and 7). Lastly, because the expression of Cyp19Cre is induced by PMSG (30), we must also consider the possibility that variation in the level of Cre expression among different follicles could result in mosaicism in the recombination of the floxed Gα_q allele. We cannot exclude this possibility, but it appears unlikely because the loss of expression of Gα_q/11 in granulosa cells is extensive (Figure 1A) and the penetrance of the phenotype caused by the Cyp19Cre-mediated recombination of other floxed alleles in granulosa cells is complete (9, 30, 40–42).

Figure 8.

Ovulation involves the LH-dependent activation of 2 families of heterotrimeric G proteins. When engaged by an agonist, the LHR present in mural granulosa cells activates G_s and G_q/11. Gα_s-dependent and -independent pathways that result in the activation of the EGF network and the ERK1/2 cascade are involved in 2 of the 3 steps required for ovulation, oocyte maturation, and cumulus expansion. These are not impaired by the deletion of Gα_q/11in granulosa cells. Follicular rupture, the final step required for ovulation, is dependent on Gα_s and Gα_q/11 and is thus partially disrupted by the deletion of Gα_q/11 in granulosa cells. Gα_q/11-dependent pathways are involved in the induction and transactivation of the progesterone receptor, an event that is required for the induction of Adatms-1 and other genes involved in follicular rupture. Gα_q/11-dependent pathways may also act downstream of the PGR and participate in follicular rupture by directly modulating gene transcription, protein stability, or by posttranslational protein modifications as shown by the dashed arrow marked with a question mark.

The impairment in ovulation and follicular rupture in the Gα_q^f/f;Gα₁₁^−/−,Cre⁺ mice is very similar to that reported for the Adamts-1^−/− mice (18). In contrast to the Adamts-1^−/− mice, which also have impaired fertilization (18), the Gα_q^f/f;Gα₁₁^−/−,Cre⁺ mice do not (Supplemental Figure 1). This discrepancy could be explained by the finding that Adamts-1 is expressed in both mural and cumulus granulosa cells (17), whereas Cyp19Cre and Pgr are expressed only in mural granulosa cells (9, 14, 16, 30, 40–42, 47). Thus, the decrease in Adamts-1 expression in the Gα_q^f/f;Gα₁₁^−/−,Cre⁺ mice is restricted to mural granulosa cells, whereas the decrease in Adamts-1 expression in the Adamts-1^−/− is global and includes cumulus and mural granulosa cells.

There is also a discrepancy in the magnitude of the reduction in ovulation (∼50%) and fertility (∼85%) in the Gα_q^f/f;Gα₁₁^−/−,Cre⁺ mice (Figure 2). This could be due to the different experimental paradigms used (cycling vs superovulated animals), the pharmacokinetics of the hormones that induce ovulation (LH in the cycling females and hCG in the superovulation protocol), or a disruption of other physiological events required for fertility in cycling mice. We know that the granulosa cell-specific deletion of Gα_q/11 does not impair fertilization of the ovulated oocytes or luteinization in superovulated mice, and we also know that it does not affect uterine receptivity in pseudopregnant females. On the other hand, we have not yet examined the effects of the granulosa cell-specific deletion of Gα_q/11 on the estrous cycle or luteinization in cycling or pregnant mice, and thus, it is possible that changes in these physiological events also contribute to the larger magnitude of the decrease in fertility compared with the decrease in ovulation.

In summary, we have generated a novel mouse model with a granulosa cell-specific deletion of Gα_q/11, and we have shown that the females experience severe subfertility, a defect in follicular rupture, and a failure of the LHR to fully induce the expression of the Pgr. This is the first conclusive evidence demonstrating that the in vivo activation of G_q/11 by the LHR is involved in ovulation.