Bidirectional communication between oocytes and ovarian follicular somatic cells is required for meiotic arrest of mammalian oocytes

Significance

The fertility of mammalian females depends upon the coordinated development of ovarian follicles and the developing eggs, called oocytes, contained within them. It has been known for many years that the somatic cells of follicles arrest the progression of meiosis in the oocyte until both the follicle and the oocyte are fully developed and ready for ovulation and fertilization. Here we show that, although the somatic compartment of ovarian follicles clearly plays an essential role in the maintenance of oocyte meiotic arrest, this function of the somatic cells is regulated by signals from the oocyte itself.

Abstract

Coordinated regulation of oocyte and ovarian follicular development is essential for fertility. In particular, the progression of meiosis, a germ cell-specific cell division that reduces the number of chromosomes from diploid to haploid, must be arrested until just before ovulation. Follicular somatic cells are well-known to impose this arrest, which is essential for oocyte–follicle developmental synchrony. Follicular somatic cells sustain meiotic arrest via the natriuretic peptide C/natriuretic peptide receptor 2 (NPPC/NPR2) system, and possibly also via high levels of the purine hypoxanthine in the follicular fluid. Upon activation by the ligand NPPC, NPR2, the predominant guanylyl cyclase in follicular somatic cells, produces cyclic guanosine monophosphate (cGMP), which maintains meiotic arrest after transfer to the oocyte via gap junctions. Here we report that both the NPPC/NPR2 system and hypoxanthine require the activity of inosine monophosphate dehydrogenase (IMPDH), the rate-limiting enzyme required for the production of guanylyl metabolites and cGMP. Furthermore, oocyte-derived paracrine factors, particularly the growth differentiation factor 9–bone morphogenetic protein 15 heterodimer, promote expression of Impdh and Npr2 and elevate cGMP levels in cumulus cells. Thus, although the somatic compartment of ovarian follicles plays an essential role in the maintenance of oocyte meiotic arrest, as has been known for many years, this function of the somatic cells is surprisingly regulated by signals from the oocyte itself.

Fertility in mammals depends on the coordinated development of ovarian follicles and the oocytes contained within them. The ovulation of oocytes by Graafian follicles must coincide with oocyte meiotic progression. Meiosis is a germ cell-specific cell division that reduces the number of chromosomes from diploid to haploid. In mammals, oocytes are arrested at the diplotene stage (diploid) of meiosis until the surge of luteinizing hormone (LH) from the pituitary gland initiates the ovulatory process. The LH surge initiates the resumption of meiosis and its progression to the metaphase of the second meiotic division, and these haploid oocytes (eggs) are ovulated into the oviduct to await fertilization. Ever since the classic experiments of Pincus and Enzmann (1) in the 1930s, it has been widely accepted that follicular somatic cells of Graafian follicles maintain the fully grown oocytes in meiotic arrest until the LH surge: Removing them from follicles for culture in a supportive medium before the surge results in an LH-independent resumption of meiosis. Thus, isolated oocytes resume meiosis simply because they have been separated from the inhibitory effect of follicular somatic cells. The mechanisms by which these cells maintain oocyte meiotic arrest have been the subject of numerous studies over the last 75 y.

Oocytes arrested at the diplotene stage are referred to as germinal vesicle (GV)-stage oocytes, and the dissolution of the oocyte nuclear envelope, often referred to as germinal vesicle breakdown (GVB), is the first morphological manifestation of the resumption of meiosis. GVB is followed by condensation of chromosomes and separation of chromosomal homologs to produce a haploid egg that is arrested at metaphase of the second meiotic division.

Maintenance of elevated cyclic adenosine monophosphate (cAMP) in oocytes is essential for sustaining meiotic arrest (2–6). After the LH surge, activation of an oocyte cAMP phosphodiesterase (PDE3A) reduces cAMP levels within the oocytes and activates the pathways culminating in GVB (7–9). Before the LH surge, PDE3A activity is inhibited in oocytes by the influx of cyclic guanosine monophosphate (cGMP) from the cumulus cells, which are the follicular somatic cells immediately adjacent to the oocyte, via membrane channels called gap junctions that metabolically couple the two cell types (10, 11). cGMP is produced by both the cumulus cells and the mural granulosa cells, the population of follicular somatic cells located inside the follicular wall (12, 13).

cGMP, which inhibits PDE3A in oocytes, is produced from GTP in granulosa cells by the activity of guanylyl cyclases, mainly natriuretic peptide receptor 2 (NPR2) (12, 13) (Fig. 1). GTP is produced from inosine monophosphate (IMP) by a series of enzymatic reactions beginning with the activity of inosine monophosphate dehydrogenase (IMPDH), the rate-limiting enzyme in the biosynthesis of guanylyl metabolites. The metabolite IMP is at the convergence of the de novo purine metabolic pathway and the HPRT salvage pathway, which rescues hypoxanthine (HX) from demolition and excretion and converts it back into IMP.

Fig. 1.

Abbreviated pathway to illustrate cGMP production and its role in oocyte meiotic arrest. The action of IMPDH, which converts IMP to GMP, is blocked by mizoribine. However, the inhibitory effect of mizoribine can be bypassed by addition of exogenous guanosine downstream of IMPDH inhibition. NPR2 is a guanylyl cyclase activated by the ligand NPPC and produces cGMP from GTP. cGMP enters the oocyte via gap junctions, represented by the heavy line at the point of entry of cGMP to the oocyte. Within the oocyte, cGMP inhibits PDE3A, a cAMP phosphodiesterase, which, if uninhibited, would degrade oocyte cAMP and cause the resumption of meiosis. IMP can be produced either by de novo pathways or by salvage of hypoxanthine. However, as indicated in the text, little salvaged hypoxanthine is converted to guanylyl metabolites in COCs (22, 23) but rather may promote the de novo synthesis of IMP or prevent its loss from availability for conversion to GMP by IMPDH.

HX was measured in millimolar concentrations in preparations of mouse and pig follicular fluids, and at those high concentrations HX maintained mouse oocytes in meiotic arrest in vitro (14–16). Because ovarian follicles are not vascularized except in the thecal cell layers, which are located outside the follicular wall, the follicular interior granulosa cells and oocyte environment are probably relatively hypoxic (17, 18), a condition that promotes generation of HX (19, 20). However, whether the high levels of HX measured in preparations of follicular fluid are physiological in healthy follicles is uncertain because levels of HX become rapidly elevated in postmortem tissues (21). Nevertheless, significant HX levels may exist in follicles. Specific inhibitors of IMPDH, such as mizoribine, reverse the meiotic arresting action of HX in vitro, indicating that the arresting action of HX is mediated by IMPDH (16). However, salvage of HX and conversion to guanylyl metabolites is minimal and the relevance of this salvage, if any, to the physiological mechanisms maintaining meiotic arrest is unclear (22).

The main source of IMP serving as IMPDH substrate in mouse cumulus–oocyte complexes (COCs) is de novo synthesis (22, 23). The importance of IMPDH in meiotic arresting mechanisms is buttressed by the findings that IMPDH inhibitors mizoribine (also known as bredinin) and mycophenolic acid induce the rapid precocious gonadotropin-independent resumption of meiosis by competent oocytes in vivo (24). This disrupts coordination of meiotic progression and ovulation and compromises oocyte quality and fertility (25). These findings show that IMPDH is crucial for maintaining meiotic arrest, presumably by providing guanylyl substrate for the generation of cGMP by guanylyl cyclase. Therefore, this study focuses on the regulation of Impdh and its function in cumulus cells for maintaining meiotic arrest.

The predominant guanylyl cyclase present in granulosa cells is NPR2 (13), a receptor whose activity is stimulated by the ligand natriuretic peptide type C (NPPC, also known as CNP). Treatment of isolated COCs with NPPC promotes elevation of cGMP levels (12, 26). Moreover, mutations in either the Npr2 or Nppc genes in mice result in failure to maintain meiotic arrest and precocious resumption of meiosis (12, 27, 28). Thus, the NPPC/NPR2 system for generating cGMP in cumulus cells is crucial for the maintenance of meiotic arrest and the coordination of oocyte meiosis with follicular ovulatory mechanisms.

NPR2 could not function to produce meiosis-inhibiting levels of cGMP without generation of sufficient substrate via IMPDH. We therefore investigated the requirement for IMPDH in the ability of NPPC/NPR2 and HX to maintain meiotic arrest. Although we previously found that oocyte-derived paracrine factors (ODPFs) promoted the expression of Npr2 mRNA in cumulus cells (12), this action would be futile without also promoting expression of IMPDH to provide substrate needed by NPR2 for cGMP production. Here we report that oocytes regulate Impdh expression and control the production of the cGMP in their companion cumulus cells that is necessary for the action of both NPPC/NPR2- and HX-mediated maintenance of meiotic arrest. Thus, the follicular system for maintaining oocyte meiotic arrest is not unidirectional from the somatic compartment to oocytes but bidirectional, requiring orchestration by the oocyte itself.

Results

Maintenance of Meiotic Arrest in Vitro by NPPC and HX.

Isolated COCs were cultured in medium containing NPPC (0–25 nM) and HX (0–1 mM) for 24 h and the percentage of oocytes that had undergone GVB was determined (Fig. 2). There was a dose-dependent maintenance of meiotic arrest when either NPPC or HX was used alone, although the highest concentration of HX used alone (1 mM) maintained only 20% of the oocytes in meiotic arrest. In contrast, 25 nM NPPC maintained about 90% of the oocytes in meiotic arrest. HX augmented the effect of 1 and 10 nM NPPC.

Fig. 2.

Effect of NPPC and hypoxanthine on oocyte maturation in vitro. COCs were cultured for 24 h in medium containing NPPC (0–25 nM) and/or HX (0–1 mM), and the percentage of oocytes that underwent GVB was determined. All media contained 100 µM E2 to maintain NPR2 receptors throughout the culture period (35). Bars indicate the mean ± SEM of seven experiments. Bars without letters in common were considered significantly different (P < 0.05).

Inhibition of IMPDH Activity Reverses NPPC- and HX-Mediated Meiotic Arrest and cGMP Elevation.

Mizoribine is a specific inhibitor of IMPDH activity (29, 30) that reverses the inhibition of GVB mediated by 4 mM HX (16) (Fig. 3), indicating that IMPDH activity participates in HX-mediated meiotic arrest in vitro. Mizoribine (200 µM) also reverses inhibition of GVB by 10 nM NPPC and by a combination of HX and NPPC (Fig. 3). To confirm that mizoribine acts by inhibiting the IMPDH production of guanylyl metabolites, we bypassed the IMPDH inhibition by treatment of COCs with guanosine, which enters the guanylyl metabolite pathway downstream of the IMPDH inhibitor (Fig. 1). At 50 µM, guanosine alone does not affect GVB (92.3 ± 1.6%), but it does abrogate the action of mizoribine, demonstrating that IMPDH activity in producing guanylyl metabolites is essential for potentiating the action of both NPPC and HX in maintaining meiotic arrest.

Fig. 3.

Effect of NPPC and/or HX on maintenance of meiotic arrest (Left) and cGMP levels (Right). COCs were cultured for 24 h in medium containing neither NPPC nor HX (control), or 10 nM NPPC, 4 mM HX, or both (blue bars). The COCs cultured with NPPC or HX, or both, were also treated with 200 µM mizoribine (red bars) or mizoribine plus 50 µM guanosine (yellow bars). Mizoribine treatment prevented the GVB-inhibiting effects of NPPC, HX, or both, and this inhibition was circumvented by guanosine (Left). NPPC and NPPC+HX promoted dramatic increases in the levels of cGMP in the COCs (Right). These increases were blocked by mizoribine, and the effect of mizoribine was bypassed by exogenous guanosine. Interestingly, HX did not raise the level of cGMP above that in control COCs. However, this level was decreased by mizoribine treatment. Thus, although HX maintained meiotic arrest (Left), an increase in cGMP was not detected (Right). Bars indicate the mean ± SEM of four experiments. Bars without letters in common were considered significantly different (P < 0.05).

Treatment with NPPC, or NPPC plus HX, dramatically increased cGMP levels in COCs, and mizoribine prevented this increase (Fig. 3). As with the GVB end point, the action of mizoribine on cGMP levels was abrogated by guanosine. In an apparent paradox, treatment with 4 mM HX did not increase cGMP levels (Fig. 3), even though this concentration of HX effectively maintained oocyte meiotic arrest (Fig. 3). Nevertheless, mizoribine treatment decreased cGMP levels in HX-treated COCs below basal (control) levels and prevented the ability of HX to maintain meiotic arrest (i.e., mizoribine induced GVB in the presence of 4 mM HX).

Oocyte Regulation of Impdh Expression.

Our previous study indicated that ODPFs elevated the steady-state expression of Npr2 mRNA (12). Effects on NPR2 protein levels were not assessed because a reliable antibody was not available. Nevertheless, elevation of Npr2 expression would be futile in the absence of adequate substrate for NPR2 guanylyl cyclase activity for producing cGMP (Fig. 1). We therefore assessed the effect of ODPFs on Impdh expression. There are two isoforms of Impdh, Impdh1 and Impdh2, that are encoded by different genes, but the products encoded by these isoforms carry out indistinguishable functions (31). Microsurgical removal of oocytes (oocytectomy; OOX) reduced the steady-state levels of mRNAs encoding both the Impdh1 and Impdh2 isoforms in cumulus cells by 24-h culture. Coculture of OOX cumulus cells with GV-stage denuded oocytes sustained expression of these mRNAs at the levels of intact COCs (Fig. 4A). For comparison, in these same samples, levels of Lhcgr were increased by OOX and decreased by oocyte coculture (32). Likewise, OOX reduced IMPDH protein levels in cumulus cells, and these were sustained by coculture with oocytes (Fig. 4B). Thus, ODPFs maintain elevated expression of Impdh, both mRNA and protein, in OOX cumulus cells.

Fig. 4.

Effect of ODPFs on levels of Impdh1 and Impdh2 mRNA and IMPDH2 protein in cumulus cells. Oocytectomized cumulus cells (OOX) were cultured alone or cocultured with oocytes (0.5–2.0 oocytes per µL) and levels of Impdh1 and Impdh2 transcripts (A) and IMPDH protein (B) were assessed after a 24-h culture. Intact COCs were also cultured for comparison. OOX caused a dramatic reduction in mRNA and protein levels. ODPFs produced by 0.5 oocytes per µL were sufficient to sustain both transcript and protein at the same or above the levels of intact COCs. Bars indicate the mean ± SEM of four experiments. Bars without letters in common were considered significantly different (P < 0.05).

Oocyte Regulation of cGMP Levels.

Oocyte control of Impdh and Npr2 in cumulus cells would be irrelevant to maintenance of meiotic arrest if cGMP levels were not affected by ODPFs. Therefore, we assessed the effect of OOX and OOX plus coculture with oocytes as in the experiments described above. OOX dramatically reduced the levels of cGMP in cumulus cells, but the levels were maintained by coculture of OOX cumulus cells with GV-stage oocytes (Fig. 5). To determine which ODPFs are responsible for increased cumulus cell cGMP, OOX cumulus cells were cultured for 24 h in medium supplemented with mouse recombinant ligands growth differentiation factor 9 (GDF9) (30 ng/mL), bone morphogenetic protein 15 (BMP15) (30 ng/mL), GDF9+BMP15 (30 ng/mL each), or the highly active GDF9–BMP15 heterodimer (3 ng/mL) (33). BMP15 alone had no effect on cGMP levels. GDF9 alone stimulated cGMP levels (∼30-fold), but not to the same extent as the two homodimers together (∼90-fold) or a low level of the GDF9–BMP15 heterodimer (∼100-fold) (Fig. 5). Thus, ODPFs promote elevated cGMP levels in the oocytes’ companion cumulus cells.

Fig. 5.

Effect of ODPFs on cGMP levels in OOX cumulus cells. (A) OOX cumulus cells were cultured alone or cocultured with oocytes (0.5–2.0 oocytes per µL) for 24 h and levels of cGMP were determined. OOX caused a dramatic reduction in cGMP levels, but these levels were mantained by coculture with oocytes. (B) The effects of recombinant mouse ODPFs on cGMP levels. OOX cumulus cells were cultured without ligands (OOX) or with mouse GDF9 (30 ng/mL), BMP15 (30 ng/mL), BMP15+GDF9 (both at 30 ng/mL), or GDF9–BMP15 heterodimer (3 ng/mL). These were maximally effective concentrations needed for cumulus expansion or the expression of expansion-related transcripts when used in concert with EGF in previous studies (33). The levels of cGMP were determined after a 24-h culture. Bars indicate the mean ± SEM of four experiments. Bars without letters in common were considered significantly different (P < 0.05).

Discussion

It has been known for many years that the somatic compartment of mammalian Graafian follicles maintains meiotic arrest of fully grown oocytes. Foremost in this mechanism is the production of cGMP in the somatic cells, which are metabolically coupled to the oocyte by gap junctions. The cGMP traverses the gap junctions into the oocyte cytoplasm, where it inhibits the activity of PDE3A, a cAMP phosphodiesterase, and thereby maintains the levels of cAMP in the oocyte sufficient to maintain meiotic arrest (10, 11). Here we show that this crucial function of the somatic component in maintaining meiotic arrest requires paracrine signals from the oocyte itself. These signals promote elevated levels of cGMP in the cumulus cells by stimulating the expression of IMPDH and NPR2, key enzymes required for the production of cGMP.

This study focused on the role of two systems reported to participate in the maintenance of meiotic arrest: the NPPC/NPR2 system and HX, and their possible interactions. Both of these systems require IMPDH activity. This requirement for the action of HX was demonstrated previously (16), and confirmed here, by showing that the action of HX is prevented by the IMPDH inhibitor mizoribine. The effect of mizoribine was circumvented by supplying guanosine, which enters the cGMP-generating system downstream of IMPDH (Fig. 1). Previous studies on the mechanism(s) by which HX maintains arrest show that HX recovered to IMP is only minimally converted by IMPDH into cGMP intermediaries, and it is not clear whether this salvage has any physiological role in the maintenance of meiotic arrest (22, 23). Moreover, it is shown here that HX does not promote elevation of cumulus cell cGMP above control (basal) levels. This seems paradoxical, because inhibition of IMPDH with mizoribine prevented the ability of HX to maintain meiotic arrest. Hence, arrest was maintained without increasing cGMP levels. However, mizoribine decreased cGMP levels in the HX-treated group below that of the untreated control samples, that is, below basal levels. Resolution of this apparent conundrum probably resides in previous studies showing that HX can inhibit oocyte cAMP phosphodiesterase activity (34). However, this cAMP phosphodiesterase-inhibiting activity of HX is not sufficient for maintenance of meiotic arrest because arrest is prevented by inhibition of IMPDH (22). It appears that at least a basal level of cGMP production, dependent upon de novo production of IMP and/or HX salvage, is required for HX to maintain arrest. In either case, IMPDH activity is required because HX-mediated meiotic arrest is reversed by mizoribine.

Although there are no genetic or physiological data confirming a physiological role of HX in the maintenance of meiotic arrest in vivo, other than the apparently high concentrations in preparations of follicular fluid, hypoxic conditions within the nonvascularized follicle (17, 18) may result in the production of HX at concentrations sufficient to have a minor role in the maintenance of meiotic arrest. In fact, HX augmented the action of NPPC at some concentrations (Fig. 2). Nevertheless, whether HX actually functions in vivo in the mechanisms maintaining meiotic arrest remains unclear.

In contrast to HX, the role of the NPPC/NPR2 system in the maintenance of meiotic arrest is firmly based on both genetic and in vitro evidence (12, 26–28, 35). This NPPC–NPR2, ligand–receptor, interaction is the principal cGMP-generating system in ovarian follicles (13). However, promoting the activity of the NPPC/NPR2 system would be futile without adequate guanylyl substrate available for conversion to cGMP. IMPDH activity is essential for providing this substrate. Inhibition of IMPDH with mizoribine prevents NPPC/NPR2 from providing the sufficient levels of cGMP needed for maintaining oocyte meiotic arrest. Providing exogenous guanosine downstream of IMPDH inhibition abrogates this activity. In the presence of mizoribine, guanosine restores both cGMP levels and meiotic arrest promoted by NPPC. Thus, the function of the NPPC/NPR2 system for the production of cGMP and maintaining meiotic arrest depends upon IMPDH activity.

ODPFs promote expression of both Impdh mRNA and protein. However, even more crucial is the final end point of cGMP production, which was also promoted by ODPFs. This final end point probably depends upon stimulation of IMPDH and NPR2, but this does not exclude other potentially important targets that are upstream of IMPDH, such as de novo IMPDH pathways, or downstream targets affecting cGMP stability or transfer to oocytes. These issues await further study. However, the final end point, maintenance of meiotic arrest, clearly is dependent upon ODPFs, specifically the GDF9 homodimer and the highly active GDF9–BMP15 heterodimer. That mouse GDF9 homodimer has activity indicates that mouse oocytes can still participate in the maintenance of meiotic arrest even in the absence of BMP15, as in Bmp15 knockout mice (36). However, the presence of both the Bmp15 and Gdf9 genes allows the production of the GDF9–BMP15 heterodimer and potentially more efficient stimulation of follicular somatic cells to maintain arrest. The participation of other ODPFs besides GDF9 and GDF9–BMP15 in the maintenance of meiotic arrest is not excluded by the results presented here. However, FGF8, which acts in concert with oocyte-derived members of the TGF-β superfamily to promote other cumulus cell metabolic functions (37) and is expressed by mouse oocytes (37–39), did not augment the actions of mouse GDF9 homodimer or the GDF9–BMP15 heterodimer in promoting the production of cGMP (results not shown). The findings herein that mouse BMP15 is inactive compared with mouse GDF9 homodimer and that GDF9–BMP15 is highly potent are parallel to the effects of these ligands on mouse cumulus cell gene expression and cumulus expansion (33).

ODPFs play diverse roles in regulating the development and function of follicular somatic cells. Oocytes orchestrate the rate of follicular development and accelerate the differentiation of both the mural and cumulus cell somatic compartments (40). ODPFs stimulate biochemical pathways in cumulus cells that are deficient in oocytes, and the products of these pathways are provided to the oocytes (37, 41). ODPFs also suppress expression of transcripts in cumulus cells that are, as a consequence of this suppression, more abundant in mural granulosa cells. For example, ODPFs including mouse GDF9 suppress expression of LH receptor in cumulus cells (42, 43). This receptor is important for endocrine functions and for responding to LH from the pituitary beginning the cascade of ovulation-associated processes. Oocytes also stimulate granulosa cell proliferation (44–46), suppress granulosa cell apoptosis (47), and affect patterns of gene expression that differ among the granulosa cell subtypes (48). Thus, oocytes not only influence metabolic processes in cumulus cells that benefit the oocytes themselves but also influence general follicular development and differential activities of the somatic follicular compartments. The action of oocytes on cGMP production probably falls into both functional categories. For example, the NPPC/NPR2 system has generalized effects on follicular development (49) as well as the specific effects described here on the maintenance of meiotic arrest. Thus, this system, which depends upon the regulation of IMPDH, is under the control of ODPFs and participates in the coordination of both general follicular and specific oocyte developmental processes.

In conclusion, our results present a surprising conceptual twist to the widely accepted mechanism of somatic cell-mediated meiotic arrest of oocytes before the LH surge. They show that the action of the somatic cells depends upon paracrine signals from the oocyte itself. Thus, the system for maintaining oocyte meiotic arrest is not unidirectional from the somatic compartment to oocytes but bidirectional, requiring orchestration by oocytes. Fig. 6 puts forward a working model for this mechanism derived from the data presented here and previous studies. Although arrest is clearly dependent upon the NPPC/NPR2 system, the possibility that HX also plays a role, albeit minor, is illustrated. The NPPC/NPR2 system and IMPDH, which provides guanylyl substrate for NPR2, are controlled by ODPFs (likely GDF9 homodimers and GDF9–BMP15 heterodimers) and together raise the levels of cGMP in the cumulus cells. The cGMP diffuses into the oocyte via the gap junctions that couple the two cell types. HX, which also gains access to the oocyte from cumulus cells via gap junctions (16), could augment the action of cGMP to inhibit PDE3A in the oocyte (22, 34). Thus, the actions of both the NPPC/NPR2 system and HX depend upon IMPDH activity, which is controlled by the oocyte, the orchestrator of follicular function.

Fig. 6.

Working model of the bidirectional mechanism required for the maintenance of meiotic arrest. Although it has been widely accepted for many years that follicular granulosa cells maintain oocytes in meiotic arrest before the preovulatory surge of LH, this model illustrates that oocytes themselves orchestrate this action of granulosa cells. ODPFs control the activity of IMPDH (green arrow). This enzyme provides the substrate for NPR2, a guanylyl cyclase that converts GTP to cGMP. The levels of cGMP in cumulus cells are elevated by ODPFs (green arrow). The purine hypoxanthine also requires IMPDH activity to participate in meiotic arrest in vitro. However, if hypoxanthine participates in meiotic arrest in vivo, it may enhance the cAMP phosphodiesterase-inhibiting activity of cGMP within the oocyte. Cumulus cells take up hypoxanthine from the follicular fluid. Both cGMP and hypoxanthine diffuse into the oocyte from cumulus cells via gap junctions represented by the heavier lines at their points of entry into the oocyte. Disruption of this arresting system would result in activation of PDE3A, and the oocytes would resume meiosis.

Materials and Methods

Animals.

COCs isolated from B6SJL/J mice were used for all experiments. Mice received an i.p. injection of 5 IU equine CG 44–48 h before removal of ovaries and isolation of COCs. All animal protocols were approved by the Administrative Panel on Laboratory Care at The Jackson Laboratory, and all experiments were carried out in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Isolation and Culture of COCs.

COCs were isolated by puncturing the ovaries with 25-gauge needles in the isolation medium described below, which also contained 10 µM milrinone, a PDE3A inhibitor, to prevent oocytes from resuming meiosis until they were distributed to experimental groups, which did not contain milrinone. For experiments to test the effects of NPPC and/or HX, ∼30 COCs were included in each group for each experiment, which was independently replicated at least three times. Unless otherwise noted, all chemicals were obtained from Sigma-Aldrich. The medium used for all experiments was Eagle’s minimum essential medium α supplemented with 3% (wt/vol) crystalline lyophilized BSA, 75 µg/mL penicillin G, 50 µg/mL streptomycin sulfate, 0.23 mM pyruvate, and 100 µM 17 β-estradiol (E2). E2 was included in all media to maintain the expression and function of NPR2 (35) regardless of whether the NPR2 ligand, NPPC, was used in the experiment group. When cumulus cell-denuded oocytes were used in coculture experiments, they were maintained at the GV stage with 10 µM milrinone. In these experiments, groups without oocytes also contained milrinone for experimental balance. Purified recombinant mouse BMP15 homodimers, GDF9 homodimers, or GDF9–BMP15 heterodimers were prepared as described (33).

Oocytectomy of Isolated COCs.

Oocytes were microsurgically removed from COCs isolated from the ovaries of 22- to 24-d-old mice injected with 5 IU eCG for 44–48 h. The procedures for OOX and subsequent coculture with or without GV-stage denuded oocytes were carried out as described previously (50).

RNA Isolation and Quantitative RT-PCR.

After harvesting, COCs or OOX cumulus cells were stored at −80 °C until analyzed for relative expression levels of mRNA. Total RNA was isolated from frozen samples using the RNeasy MicroRNA Isolation Kit (Qiagen) as recommended by the manufacturer. Reverse transcription was performed directly after RNA isolation using the QuantiTect Reverse Transcription Kit (Qiagen) as recommended by the manufacturer. Reverse-transcribed cDNA was diluted 1:5 with nuclease-free water and stored at −80 °C until quantitative real-time PCR was performed. The relative amount of target gene expression for each sample was conducted using an ABI 7500 real-time PCR instrument (Applied Biosystems). PCR primer sequences for Impdh1 were 5′-CCTGCCCTTCATTCCTGATTC-3′ (forward) and 5′-GCCTTCAACTGCAGCTGAGAA-3′ (reverse) and for Impdh2 were 5′-TGTCACAGAGGCTGGAATGG-3′ (forward) and 5′-CAGTTGTGGTGGATGAAACCAA-3′ (reverse). Levels of mRNA were normalized relative to the abundance of an endogenous control housekeeping transcript, ribosomal protein L19 (Rpl19). The primers for Rpl19 were reported previously (12, 51). Relative levels of target gene expression for each sample were calculated using the formula 2^−∆∆Ct as described previously (52). Sequencing the purified PCR products validated the identities of all amplicons.

Measurement of cGMP.

COCs or OOX cumulus cells were collected and immersed in 50 µL 1 N HCl and heated for 3 min at 100 °C. Thirty to 60 complexes were used for each measurement. The samples were diluted with 140 µL enzyme immunoassay (EIA) buffer, and cGMP was measured using a cGMP-EIA Kit (Cayman Chemicals), a Spectra Max 250 plate reader (Molecular Devices), and a spreadsheet tool for data analysis from the Cayman web site (www.caymanchem.com/analysis/eia) (12). cGMP data are expressed as f moles cGMP/COC or cGMP/cumulus cells of one OOX complex.

Western Blot Analysis.

Protein samples were prepared as described (53) with minor modifications. Cumulus cells of 30 COCs were harvested and washed once with PBS. Supernatants were removed and Laemmli sample buffer (Bio-Rad) containing 5% (vol/vol) β-mercaptoethanol was added to the pellets. The lysates were heated to 100 °C for 5 min. After cooling on ice for 5 min and centrifuging at 13,000 rpm for 2 min, samples were frozen at −70 °C until use. Proteins were separated on a 10% gradient polyacrylamide gel (Bio-Rad) and electroblotted onto a PVDF membrane (Millipore). Following transfer and blocking in 5% (wt/vol) nonfat dry milk in Tris-buffered saline containing 0.1% (vol/vol) Tween-20 (TBST) for 1 h, the membrane was then incubated with monoclonal anti-IMPDH antibody (sc-166551; Santa Cruz Biotechnology), which recognizes mouse IMPDH1 and IMPDH2, diluted 1:100 in blocking buffer overnight at 4 °C. After three washes, 15 min each, with TBST, membranes were incubated for 1 h at room temperature with horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG (Jackson ImmunoResearch Laboratories) diluted 1:10,000 in blocking buffer. The membranes were washed three more times in TBST, followed by detection with SuperSignal West Pico Chemiluminescent Substrate (Pierce). Membranes were stripped in Restore Plus Western Blot Stripping Buffer (Thermo Scientific), followed by detection of β-actin (ACTB) protein (positive control) using HRP-conjugated anti–β-actin antibody (ab49900; Abcam). The intensities of individual bands were analyzed using ImageJ 1.46 (National Institutes of Health) and normalized to the expression of ACTB.

Statistical Analyses.

JMP10 software was used to conduct t tests or Tukey–Kramer honestly significant difference tests for paired or multiple comparisons, respectively (SAS Institute). Data for percentages of oocytes undergoing GVB were first subjected to arcsine transformation. A P value of less than 0.05 was considered statistically significant.