Cyclic GMP from the surrounding somatic cells regulates cyclic AMP and meiosis in the mouse oocyte

Abstract

Mammalian oocytes are arrested in meiotic prophase by an inhibitory signal from the surrounding somatic cells in the ovarian follicle. In response to luteinizing hormone (LH), which binds to receptors on the somatic cells, the oocyte proceeds to second metaphase, where it can be fertilized. Here we investigate how the somatic cells regulate the prophase-to-metaphase transition in the oocyte, and show that the inhibitory signal from the somatic cells is cGMP. Using FRET-based cyclic nucleotide sensors in follicle-enclosed mouse oocytes, we find that cGMP passes through gap junctions into the oocyte, where it inhibits the hydrolysis of cAMP by the phosphodiesterase PDE3A. This inhibition maintains a high concentration of cAMP and thus blocks meiotic progression. LH reverses the inhibitory signal by lowering cGMP levels in the somatic cells (from approximately 2 microM to approximately 80 nM at 1 hour after LH stimulation) and by closing gap junctions between the somatic cells. The resulting decrease in oocyte cGMP (from approximately 1 microM to approximately 40 nM) relieves the inhibition of PDE3A, increasing its activity by approximately 5-fold. This causes a decrease in oocyte cAMP (from approximately 700 nM to approximately 140 nM), leading to the resumption of meiosis.

Fig. 1.

In vitro characterization of the Epac2-camps300 sensor for cAMP. (A) Diagram of the sensor, showing cyan fluorescent protein (CFP, turquoise) and yellow fluorescent protein (YFP, yellow), linked by the EPAC2 (RAPGEF4) cAMP-binding domain (red) [constructed based on the crystal structures of EPAC2 (Protein Data Bank accession number 107F) and green fluorescent protein (Protein Data Bank accession number 1GFL)]. The location of the K405E point mutation introduced into the EPAC2 sequence of Epac2-camps to generate Epac2-camps300 is shown. Binding of cAMP decreases Förster resonance energy transfer (FRET) between CFP and YFP. (B) Concentration-response curves for Epac2-camps and Epac2-camps300 for cAMP. The EC₅₀ (50% of the maximum change in the YFP/CFP emission ratio) values are 820±130 nM (mean±s.e.m., n=7) for Epac2-camps, and 320±18 nM (n=6) for Epac2-camps300. (C) Concentration-response curves for Epac2-camps300 for cAMP and cGMP. The EC₅₀ value for cGMP is 14±4 μM (n=4).

Fig. 2.

Confocal images of an antral follicle-enclosed oocyte containing 5 μM Epac2-camps300. The oil drop is present as a result of microinjection. Upon excitation of CFP, the ratio of YFP to CFP emission was measured as an indicator of the concentration of cAMP in the oocyte.

Fig. 3.

The concentration of cAMP in follicle-enclosed oocytes decreases in response to luteinizing hormone. (A-C) The YFP/CFP emission ratio from follicle-enclosed mouse oocytes containing Epac2-camps300, before and after injecting 1 mM cAMP, for no treatment (basal) (A) and for ∼1 hour (B) or ∼5 hours (C) after applying luteinizing hormone (LH). (D) Percentage change in the YFP/CFP emission ratio in response to injection of 1 mM cAMP for follicle-enclosed oocytes with or without LH exposure. The percentage change was 15.1±1.8 (n=15) for no LH, 25.6±0.9 (n=18) for 1-1.4 hours LH, and 31.9±1.4 (n=5) for 5 hours LH.

Fig. 4.

Calibration of the cAMP concentrations in follicle-enclosed oocytes with or without LH treatment. (A,B) Determination of the percentage change in the YFP/CFP emission ratio for Epac2-camps300 in going from a minimum to a maximum cAMP concentration in vivo. (A) Record from a follicle-enclosed mouse oocyte that was injected with 5 μM Epac2-camps300 plus 100 μg/ml PDE3A catalytic domain (to lower cAMP to a minimum level), and then ∼1 hour later injected with 1 mM cAMP. (B) Percentage change in YFP/CFP emission ratio in follicle-enclosed oocytes injected with PDE3A, or BSA control, and then ∼1 hour later injected with 1 mM cAMP. Injection of 100-200 μg/ml PDE3A lowered the YFP/CFP emission ratio by 35.5±1.0% (n=8). (C) cAMP concentrations in follicle-enclosed oocytes with or without LH treatment, as calculated using the in vitro concentration-response curve for Epac2-camps300 (replotted here from Fig. 1B), the 35.5% value for the maximum change in YFP/CFP emission ratio for Epac2-camps300 in vivo (A,B), and the data presented in Fig. 3. The following example illustrates how these calculations were made. For the measurement shown in Fig. 3A, the change in YFP/CFP emission ratio in going from the baseline level to that after injecting 1 mM cAMP was 14%. Using the percentage change in YFP/CFP ratio over the dynamic range of the sensor in vivo, shown above to be 35.5%, we calculated the percentage change in YFP/CFP ratio corresponding to a change from ∼0 mM cAMP to the baseline cAMP level in Fig. 3A: 35.5-14=21.5%. 21.5/35.5=61% of the maximum change in YFP/CFP ratio over the dynamic range of the sensor. 61% on the y-axis of Fig. 4C corresponds to 540 mM cAMP on the x-axis, as determined using Origin software. The horizontal bars in C represent the mean±s.e.m. for the cAMP concentrations calculated for the sets of measurements summarized in Fig. 3D. Basal, 660±110 nM (n=15); 1-1.4 hours, 140±18 nM (n=18); 5 hours, 43±17 nM (n=5). The cAMP concentrations at 1-1.4 and 5 hours after LH are significantly different from the basal value (P<0.0001 and P=0.005, respectively), and from each other (P<0.01).

Fig. 5.

The concentration of cGMP in follicle-enclosed oocytes decreases in response to LH. (A,B) The YFP/CFP emission ratio from follicle-enclosed mouse oocytes containing cGi500, before and after injecting 1 mM cGMP, for no treatment (basal) (A) and for ∼1 hour after applying LH (B). (C) Percentage change in the YFP/CFP emission ratio for cGi500 in response to injection of 1 mM cGMP, for follicle-enclosed oocytes with or without LH treatment. The percentage change in YFP/CFP emission ratio was 13.9±1.4 (n=21) for no LH, 30.1±0.6 (n=25) for 1-1.3 hours LH, and 30.6±0.5 (n=9) for 5 hours LH.

Fig. 6.

Calibration of cGMP concentrations in follicle-enclosed oocytes with or without LH treatment. (A,B) Determination of the percentage change in the YFP/CFP emission ratio for cGi500 in going from a minimum to a maximum cGMP concentration in vivo. (A) Record from a follicle-enclosed mouse oocyte containing cGi500 that was injected with 360 μg/ml PDE9A catalytic domain (to lower cGMP to a minimum level) and ∼1 hour later injected with 1 mM cGMP. (B) Percentage change in YFP/CFP emission ratio in follicle-enclosed oocytes injected with PDE9A, or BSA control, and then ∼1 hour later injected with 1 mM cGMP. Injection of 180-360 μg/ml PDE9A lowered the YFP/CFP emission ratio by 35.0±0.9% (n=16). (C) cGMP concentrations in follicle-enclosed oocytes with or without LH treatment, as calculated by the procedure described in the legend to Fig. 4C, using the in vitro concentration-response curve for cGi500 (replotted here from Fig. S1 in the supplementary material), the 35.0% value for the maximum change in YFP/CFP emission ratio for cGi500 in vivo (A,B), and the data presented in Fig. 5. The horizontal bars in C represent the mean±s.e.m. for the cGMP concentrations calculated for the set of measurements summarized in Fig. 5C. Basal, 890±150 nM (n=21); 1-1.3 hours, 39±7 nM (n=25); 5 hours, 30±5 nM (n=9); for clarity, only the basal and 1-1.3 hour values are shown on the graph. The cGMP concentrations at 1-1.3 and 5 hours after LH are significantly different from the basal value (P<0.0001 and P=0.001, respectively), but not from each other (P=0.5).

Fig. 7.

Decreasing cGMP by injecting follicle-enclosed oocytes with the catalytic domain of PDE9A decreases cAMP. (A) Determination of the percentage change in the YFP/CFP emission ratio for Epac2-camps300 in a follicle-enclosed mouse oocyte that was injected with 180 μg/ml PDE9A catalytic domain and then ∼1 hour later injected with 1 mM cAMP. The cAMP injection lowered the YFP/CFP emission ratio by 31.2±0.7% (n=6). (B) cAMP concentrations after injection of PDE9A, or BSA control (200 μg/ml), determined using the calibration curve in Fig. 4C. cAMP concentrations for basal and 1-1.4 hour LH conditions, from Fig. 4C, are shown for comparison.

Fig. 8.

Decreasing cGMP by injecting follicle-enclosed oocytes with the catalytic domain of PDE9A causes PDE3-dependent meiotic resumption. (A,B) Time course of nuclear envelope breakdown (NEBD) in response to injection of various concentrations of (A) the catalytic domain of PDE9A or a catalytically inactive mutant (D293A), or (B) the catalytic domain of PDE3A or BSA control. (C) Meiotic resumption in response to injection of PDE9A is reversibly inhibited by milrinone, an inhibitor of PDE3. Follicles were preincubated for 1-2 hours with 100 μM milrinone or with a control solution containing 0.5% DMSO; then, the follicle-enclosed mouse oocytes were injected with 180 μg/ml PDE9A and observed for NEBD. Five hours later, the follicles were washed into milrinone-free medium. Each curve represents the results from injection of 7-14 follicle-enclosed oocytes.

Fig. 9.

cGMP concentrations in follicle-enclosed oocytes treated with reagents that affect gap junction closure, and in follicle-enclosed oocytes and whole follicles in response to LH. (A,B) The YFP/CFP emission ratio from follicle-enclosed mouse oocytes containing the cGMP sensor cGi500, before and after injecting 1 mM cGMP, for follicles treated with 100 μM CBX (A), or LH plus 10 μM U0126 (B). (C) cGMP concentrations in follicle-enclosed oocytes after various treatments (see text), as determined from measurements as in A,B and the calibration curve in Fig. 6C. The values for basal and 1-1.3 hour LH conditions are from Fig. 6C. Oocytes had not undergone NEBD at the time of the measurements. (D) cGMP concentrations in whole follicles with or without treatment for 1 hour with LH (mean±s.e.m. for n independent assays). Basal, 2300±310 nM (n=8); 1 hour LH, 85±14 nM (n=8).

Fig. 10.

Working hypothesis for how meiotic arrest is maintained and how LH causes meiotic resumption in follicle-enclosed mouse oocytes. Diagramatic representations of low-magnification views of a section of mouse follicle (center), and higher-magnification views of a mural granulosa cell (top) and the oocyte (bottom). (Left) No LH. The cAMP concentration in the mural granulosa cells is relatively low because the LH receptor (LHR)/G_s/adenylyl cyclase (AC) system is inactive. The cGMP concentration is high, owing to active guanylyl cyclase (GC) and inactive cGMP phosphodiesterase (cGMP PDE). In the oocyte, the concentration of cGMP is high owing to diffusion of cGMP from the somatic cells through the gap junctions. The concentration of cAMP is high owing to the constitutive activity of GPR3 and the inhibition of PDE3A by cGMP. Elevated cAMP maintains meiotic arrest. (Right) One hour after LH application. Activation of the LH receptor activates G_s and adenylyl cyclase, elevating cAMP levels in the mural granulosa cells. This initiates a signaling pathway that closes gap junctions throughout the somatic compartment (see text). LH also causes cGMP levels in the somatic cells to decrease by decreasing guanylyl cyclase activity and/or by increasing cGMP phosphodiesterase activity. As a result of the closure of gap junctions and the decrease in cGMP in the somatic cells, cGMP levels in the oocyte decrease, inhibition of PDE3A is relieved, cAMP decreases and meiosis resumes.