G Protein βγ Subunits Inhibit Nongenomic Progesterone-induced Signaling and Maturation in Xenopus laevis Oocytes

Progesterone-induced maturation ofXenopus oocytes is a well known example of nongenomic signaling by steroids; however, little is known about the early signaling events involved in this process. Previous work has suggested that G proteins and G protein-coupled receptors may be involved in progesterone-mediated oocyte maturation as well as in other nongenomic steroid-induced signaling events. To investigate the role of G proteins in nongenomic signaling by progesterone, the effects of modulating Gα and Gβγ levels in Xenopus oocytes on progesterone-induced signaling and maturation were examined. Our results demonstrate that Gβγ subunits, rather than Gα, are the principal mediators of progesterone action in this system. We show that overexpression of Gβγ inhibits both progesterone-induced maturation and activation of the MAPK pathway, whereas sequestration of endogenous Gβγ subunits enhances progesterone-mediated signaling and maturation. These data are consistent with a model whereby endogenous free Xenopus Gβγ subunits constitutively inhibit oocyte maturation. Progesterone may induce maturation by antagonizing this inhibition and therefore allowing cell cycle progression to occur. These studies offer new insight into the early signaling events mediated by progesterone and may be useful in characterizing and identifying the membrane progesterone receptor in oocytes.

The phenomenon of progesterone-induced maturation of Xenopus oocytes serves as a useful experimental model for study-ing nongenomic steroid signaling (16, 19 -21). The maturation of an oocyte refers to the meiotic stage at which an oocyte rests. "Immature" oocytes are arrested in prophase of meiosis I. Before ovulation, oocytes are induced to re-enter the cell cycle, finally resting in metaphase II. These "mature" oocytes are then competent for ovulation and subsequent fertilization, after which the final stages of meiosis are completed (16).
Evidence suggests that progesterone-induced maturation of Xenopus oocytes is mediated by cell surface rather than nuclear receptors. First, maturation is unaffected by the transcriptional inhibitor actinomycin D (21). Second, progesterone covalently attached to either polymers or bovine serum albumin and therefore unable to diffuse through the oocyte membrane still mediates maturation (16,22). Third, progesterone injection directly into oocytes does not induce maturation (16,21,23). Finally, progesterone appears to bind specifically and with relatively high affinity to oocyte membranes (24,25). At this point, however, no progesterone binding proteins have been identified in Xenopus oocyte membranes.
Progesterone signaling may be coupled to G proteins. Progesterone treatment of oocytes results in a rapid decrease in intracellular cAMP, perhaps through attenuation of adenylyl cyclase activity (26 -29). This suggests that G␣ i , and therefore a G␣ i -coupled receptor, may be involved in progesterone-mediated signaling. In addition, progesterone activates the mitogenactivated protein kinase (MAPK) 1 cascade in oocytes (30 -34), which could be mediated by G␤␥ subunits. Finally, 1-methyladenine induces maturation of starfish oocytes in a pertussis toxin-sensitive fashion, suggesting that G␣ i mediates oocyte maturation in a comparable system (35)(36)(37). Interestingly, progesterone binds to the G protein-coupled oxytocin receptor, thereby partially blocking oxytocin binding and oxytocin-mediated signaling (38). One hypothesis is that the high amounts of progesterone present during pregnancy bind to the oxytocin receptor and prevent the induction of labor. When progesterone levels decrease at the end of pregnancy, this "release of inhibition" may allow oxytocin-mediated contractions, and therefore labor, to begin. Again, these data are consistent with progesterone modulation of G protein-coupled receptor signaling.
The role of G proteins in progesterone-mediated Xenopus oocyte maturation was examined by systematically depleting or overexpressing G protein subunits in Xenopus oocytes. Our data suggest that, in fact, G␣ i is not important for this process. Rather, G␤␥ subunits appear to be the principal mediators of progesterone action in Xenopus oocytes, acting as inhibitors of oocyte maturation. Progesterone may therefore induce matu-ration by antagonizing constitutive G␤␥-mediated inhibition of cell cycle progression.

EXPERIMENTAL PROCEDURES
Oocyte Preparation-Oocytes were harvested from female Xenopus laevis (Nasco) and treated as described previously (39,40). Briefly, follicular cells were removed by incubation of the oocytes for 3-4 h at room temperature with 1 mg/ml collagenase A (Roche Molecular Biochemicals) in modified Barth's solution (MSBH) without Ca 2ϩ . Oocytes were then washed extensively and incubated overnight at 16°C in MSBH II (MSBH with 1 mg/ml bovine serum albumin, 1 mg/ml Ficoll, 100 units/ml penicillin, and 100 g/ml streptomycin). Stage V-VI oocytes were selected, and maturation assays were performed on each preparation to determine its sensitivity to progesterone-induced maturation.
Progesterone Maturation Assay-Injected or uninjected stage V-VI ooctyes were washed extensively with MSBH to remove the bovine serum albumin from the storage buffer. They were then incubated with various doses of progesterone (Sigma) for the times indicated in the figure legends. The ethanol concentration was always kept constant at 0.1% regardless of the progesterone concentration. Maturation was detected by visualizing germinal vesicle breakdown, which manifests itself as a white spot on the dark animal pole of the oocyte (16).
Oocyte Membrane Preparation-Oocyte membranes were prepared from Stage V-VI or Stage I-IIII oocytes as described (25). In summary, oocytes were homogenized with a Dounce homogenizer (Kontes, pestle A, 15 strokes) in membrane buffer (83 mM NaCl, 1 mM MgCl 2 , 1 mM phenylmethylsulfonyl fluoride, 10 mM Hepes, pH 7.6) at 4°C. The homogenate was centrifuged at 800 ϫ g for 10 min, and the supernatant was removed and centrifuged again at 800 ϫ g. After a third centrifugation at 800 ϫ g, the supernatant was centrifuged at 20,000 ϫ g for 25 min, and the membrane pellet was resuspended in membrane buffer, taking caution to avoid disturbing the black melanin component of the pellet. Membranes were centrifuged two more times at 20,000 ϫ g and finally resuspended in MSBH at a protein concentration of 5 mg/ml. Samples were then frozen at Ϫ80°C until needed.
Binding Assay-1,2,6,7-3 H(N)]-progesterone (PerkinElmer Life Sciences) was diluted in ethanol on ice to 100ϫ the final concentrations indicated in the figure legends. 5 l of each of the 100ϫ solutions was added to individual 1.5-ml polypropylene microcentrifuge tubes with either 1 l of unlabeled progesterone stock in ethanol (concentration indicated in the figure legends) or 1 l of ethanol alone. Each condition (i.e. concentration of radiolabeled progesterone) was tested in triplicate for every experiment. Membranes were diluted to the concentrations indicated in the figure legends in MSBH at 4°C, and 500 l of this mixture was added to each tube containing radiolabeled steroid. The tubes were then rocked at 4°C for 1 h. 25 l was collected from each tube, diluted in scintillation fluid, and counted to determine the concentration of free progesterone at the end of the assay. Glass microfiber filters (Millipore) that had been preincubated in MSBH at 4°C were placed on a prechilled vacuum filter holder (Millipore), and 400 l of each sample was applied to individual filters. The filters were washed with 20 ml of cold MSBH, removed, and placed in 5 ml of scintillation fluid (Budget-Solve, Research Products International Corp.) for counting on a Beckman LS1801 scintillation counter.
Oligonucleotide Design, Plasmid Construction, cRNA Synthesis, and cRNA Injection-Sense and antisense oligonucleotides directed against sequences near the start codons of Xenopus G␣ i 1 (accession number AF086606), G␣ i 2 (accession number X56089), and G␣ i 3 (accession number X56090) (41) were synthesized by Life Technologies, Inc. These sequences were as follows: ATCGCTGCCATGGGCTGCACCTCGA-GCGCC (G␣ i 1), GCTGAACGGAGAACCGTCGCCATGGGATGTACT (G␣ i 2), and CGGGAGGCGCCGAGCGAAGTAAAATGATC (G␣ i 3). 50.6 nl of 2 mg/ml solutions of oligonucleotides in 10 mM Hepes were injected into each oocyte as indicated in the figure legend.
The Xenopus G␣ i 1, G␣ i 2, and G␣ i 3 cDNAs were cloned by polymerase chain reaction from a Xenopus oocyte cDNA library created from purified Xenopus oocyte mRNA (Fast-Track, Invitrogen) using standard techniques. The modified G␣ i 2 cDNA (G␣ i 2-Mod) contains the following 5Ј sequence: CCAGCCATGGGTTGCACA. The G␣ i 2-Mod cDNA will match with only 12 of 33 nucleotides of the G␣ i 2 antisense oligonucleotide. The rat G␣ i -WT and G␣ i -Q203L cDNAs were a gift from A. Gilman (UTSW). All of these cDNAs were cloned into the Xenopus oocyte expression vector pGEM HE (a gift from L. Jan, University of California at San Francisco) and were sequenced using the dideoxy method. The bovine G␤ 1 cDNA in pGEM HE and bovine G␥ 2 cDNA in pFROG were gifts from L. Jan (University of California at San Francisco). The bovine transducin G␣ cDNA in pFROG was a gift from S. Coughlin (University of California at San Francisco), and the GRK-minigene in pGEM HE was a gift from E. Reuveny (Tel Aviv). All constructs were linearized and transcribed in vitro with either T7 or SP6 (Promega transcription kit). Stage V-VI oocytes were injected with 50.6 nl of cRNA at a concentration of approximately 200 ng/l using a Drummond automatic injector, and injected oocytes were then incubated 30 -48 h in MSBH II before the maturation or MAPK assays were performed.
MAPK Assay-Injected oocytes were washed with MSBH and incubated with 50 nM progesterone as described in the experimental procedures under "Progesterone Maturation Assay." Oocytes were lysed at the indicated times with lysis buffer (150 mM NaCl, 2 mM EDTA, 0.5 mM sodium vanadate, 2 mM NaF, 1% Tx-100, 20 mM Tris, pH 7.6) at 4°C. 20 l of lysis buffer was used per oocyte. Lysates were microcentrifuged at full speed for 10 min, and the supernatant was removed and diluted in 2ϫ Laemmli sample buffer. Samples were resolved by electrophoresis on 10% polyacrylamide gels, and proteins were transferred to Immobilon membranes (Millipore). Membranes were blocked with 5% nonfat dry milk in TBST (100 mM NaCl, 0.1% Tween-20, 50 mM Tris, pH 7.4) for 1 h, incubated overnight at 4°C with a rabbit anti-phospho-p44/42 MAPK antibody (New England Biolabs), washed four times with TBST, incubated for 1 h at room temperature with horseradish peroxidaseconjugated goat anti-rabbit antibody (1:4000, Bio-Rad), and washed another four times with TBST. Blots were then treated with ECL-Plus (Amersham Pharmacia Biotech) to visualize the proteins. Next, blots were stripped in 100 mM 2-mercaptoethanol, 2% SDS, 62.5 mM Tris, pH 6.7 at 55°C for 30 min, washed four times with TBST, and reblotted as above using a rabbit anti-p44/42 MAPK polyclonal antibody (New England Biolabs) that will bind all p44/p42 regardless of its phosphorylation status. 50 for maturation in the range of 75-150 nM, depending upon the oocyte preparation (Fig. 1A). The sensitivity of the in vitro maturation response to low concentrations of progesterone suggests that it could be relevant in vivo, where physiologic concentrations of progesterone in the ovary could easily be in the 100 nM range (38). The maturation response was also relatively specific to progesterone, because much higher concentrations of pregnenolone (EC 50 ϭ ϳ1 M), 17-OH progesterone, corticosterone, and aldosterone (all with EC 50 values of ϳ0.5-1 M) were needed to induce maturation (data not shown).

The EC 50 of Progesterone-induced Maturation Correlated with Specific High Affinity Binding of Progesterone to Maturation-competent Oocyte Membranes-Maturation experiments using increasing concentrations of progesterone revealed an EC
Progesterone binding studies were performed on Xenopus oocyte membranes to compare the affinity of progesterone binding to the EC 50 for progesterone-induced maturation. Scatchard analysis revealed two different sets of progesterone binding sites on the oocyte membranes: a set of lower affinity sites (K d ϭ 140 nM), which most likely represented nonspecific binding, and a set of high affinity sites (K d ϭ 3 nM), which likely represented the specific progesterone binding sites (Fig. 1B). The density of high affinity binding sites in this experiment was approximately 1 ϫ 10 Ϫ13 mol/mg (ϳ3 ϫ 10 8 molecules/oocyte).
The EC 50 for progesterone-induced maturation correlated with the equilibrium constants calculated for both sets of binding sites, suggesting that progesterone binding to either set of receptors could be mediating the maturation response. To confirm that the high affinity sites were mainly responsible for progesterone-mediated effects, membranes from maturationcompetent (Stage V-VI) and incompetent (Stage I-III) oocytes were tested for progesterone binding. A radiolabeled progesterone concentration of 10 nM was used to favor detection of the high affinity binding sites. Although stage V-VI oocytes contained specific high affinity binding sites for [ 3 H]progesterone, the earlier stage oocytes (Stages I-III) did not (Fig. 1C). The presence of the high affinity progesterone binding sites almost exclusively on maturation-competent oocytes suggests that the specific binding of progesterone to the high affinity sites may indeed be linked to the maturation process. Competition studies with other steroids revealed that progesterone binding to the stage V-VI oocyte membranes was specific, because over 100-fold more corticosterone, testosterone, and aldosterone were needed to compete for the high affinity progesterone binding sites (data not shown).
Progesterone Binding to Its Membrane Receptor Was Unaffected by the State of G␣ Activation-The affinity of a GPCR agonist for its receptor is generally decreased when associated G␣ subunits are in the GTP-bound, activated state (42). For example, agonist affinity for the ␤-adrenergic receptor is markedly decreased in the presence of GTP (43). If progesterone is acting as a GPCR agonist to mediate oocyte maturation, then, accordingly, its affinity for this receptor might be expected to be FIG. 1. Progesterone-mediated oocyte maturation and progesterone binding to oocyte membranes. A, the EC 50 for progesteronemediated maturation was determined by treating oocytes with the indicated concentrations of progesterone (x axis) for 18 h. Maturation, which was detected by looking for germinal vesicle breakdown, is displayed on the y axis as a percentage of the total number of oocytes/concentration of progesterone (n ϭ 20). This experiment was performed over 10 times with a range of EC 50 from 50 to 150 nM. B, binding studies on stage V-VI oocyte membranes using a wide range of radiolabeled progesterone concentrations was performed. 300 g of membranes were used for each sample. Data represent the means of three samples at each concentration of progesterone and are expressed as a Scatchard plot. This experiment was performed three times with three different membrane preparations with nearly identical results. C, binding assays were performed on membranes from stage I-III and stage V-VI oocytes. 50 g of membranes were incubated with 10 nM radiolabeled progesterone Ϯ 100 nM unlabeled progesterone for each sample. Data represent the means Ϯ S.D. (n ϭ 3) of the total counts after subtracting background, with background defined as the number of counts in samples incubated with labeled plus excess unlabeled progesterone. The total counts bound in Stage V-VI membranes was approximately 2-fold above background. D, binding studies on stage V-VI oocyte membranes (300 g/point) were done using the indicated radiolabeled progesterone concentrations. Experiments were performed as described in the experimental procedures only the MSBH contained 5 mM MgCl 2 , and half of the membranes were preincubated with 200 M GTP␥S in MSBH for 30 min at 4°C (open squares) before being added to the radiolabeled progesterone. Data represent the means of three samples at each concentration. This experiment was performed three times with nearly identical results. lower when G␣ subunits are in the activated state. Xenopus oocyte membranes were incubated both with and without the nonhydroyzable GTP analogue GTP␥S. As demonstrated in Fig. 1D, incubation with GTP␥S had no effect on progesterone binding to the oocyte membranes, suggesting that progesterone is not binding to its receptor in a fashion typical for a GPCR agonist.
Progesterone-induced Maturation of Xenopus Oocytes Was Affected by Changes in G␣ i Expression-The role of G␣ i in progesterone-induced maturation was examined by modulating its expression in oocytes. Depletion of endogenous G␣ i was attained by injection of antisense oligonucleotides directed against mRNAs encoding three known Xenopus G␣ i proteins into oocytes. Antisense experiments work well in oocytes because the oligonucleotides can be injected directly into the cell, thus avoiding the usual problems of degradation and inconsistent uptake seen with cultured cells (44). Note that progesterone concentrations used in this and subsequent experiments were just below the EC 50 calculated from Fig. 1A. This allowed more subtle changes in progesterone sensitivity to be detected. Given the slight variability in EC 50 between oocyte preparations, however, all experiments were performed multiple times to confirm the results. Injection of antisense oligonucleotides against mRNA encoding each individual G␣ i subunit (G␣ i 1, G␣ i 2, and G␣ i 3) had minimal effect on progesterone-induced maturation when compared with injection of matching sense oligonucleotides ( Fig. 2A). In contrast, simultaneous injection of antisense oligonucleotides against mRNAs encoding all three G␣ i subunits significantly attenuated progesterone-mediated maturation when compared with oocytes injected with matching sense oligonucleotides ( Fig. 2A).
The results from these antisense experiments are difficult to interpret, because antibodies against Xenopus G␣ i proteins are not available to confirm that the antisense oligonucleotides are indeed reducing G␣ i expression. To further examine the effects of G␣ i on maturation, cRNA encoding the Xenopus G␣ i 2 protein was injected into oocytes. Injection of this cRNA resulted in marked spontaneous maturation when compared with oocytes injected with 10 mM Hepes (mock) or nonspecific cRNA (PAR1) (Fig. 2B). Interestingly, this spontaneous maturation was inhibited by nearly 80% with co-injection of the antisense oligonucleotide directed against G␣ i 2 (Fig. 2C, G␣ i 2-WT). In contrast, co-injection of the G␣ i 2 antisense oligonucleotide with a modified cRNA encoding the wild-type G␣ i 2 protein but containing 21 of 33 mismatches in the region corresponding to antisense oligonucleotide binding had less of an inhibitory effect on maturation (Fig. 2C, G␣ i 2-Mod). This suggests that at least the G␣ i 2 antisense oligonucleotide can inhibit G␣ i activity via binding to its complementary RNA sequence. Together, these results indicate that the amount of G␣ i expressed in oocytes directly correlates with the maturation response to progesterone.
G␤␥ Inhibited Progesterone-induced Maturation of Xenopus Oocytes-Two interpretations can explain the observation that decreased G␣ i attenuated, whereas excess G␣ i enhanced, progesterone-induced maturation. First, G␣ i itself may be signaling to induce maturation. Second, G␤␥ that is released by activation of G␣ i may be affecting the maturation response. To differentiate between these two possibilities, oocytes were injected with cRNAs encoding bovine G␤ 1 or G␥ 2 subunits. These cRNAs encode functional proteins, because they have been injected together into Xenopus oocytes and markedly enhanced G␤␥-dependent G protein inward rectifying potassium channel activity (45). 2 Individual expression of either G␤ 1 or G␥ 2 had no effect on progesterone-induced maturation when a progesterone concentration just below the EC 50 was used (50 nM). In contrast, simultaneous expression of both subunits in oocytes significantly attenuated progesterone-induced maturation (Fig. 3A). Incubation with higher concentrations of progesterone overcame this inhibition by G␤␥ (Fig. 3B), indicating that the processes of G␤␥-mediated inhibition and progesteronemediated induction of maturation can compete with each other. In addition, these data demonstrate that injected G␤␥ subunits are not simply damaging their host oocytes in a nonspecific fashion.
The effect of lowering endogenous Xenopus G␤␥ was examined by injecting oocytes with cRNA encoding the bovine retinal transducin G␣ subunit. Transducin G␣ itself has no apparent activity in oocytes but has been used as a G␤␥ "sink" to inhibit G␤␥-mediated signaling (46). Oocytes injected with transducin G␣ exhibited enhancement of progesterone-induced maturation (Figs. 3C and 2B). In addition, injection of a GRK minigene encoding the carboxyl portion of the GRK1 protein, which has been shown to inhibit G␤␥ function in a very specific fashion (47,48), markedly enhanced progesterone-mediated maturation at a concentration of progesterone just below the EC 50 (50 nM) (Fig. 3D). The observation that sequestration of endogenous G␤␥ enhances, while overexpression of exogenous G␤␥ attenuates, progesterone-induced maturation suggests that increased G␣ i activity is not activating maturation; instead, endogenous G␤␥ is inhibiting maturation.
GTP␥S Inhibited Progesterone-induced Maturation-One way to formally prove that G␤␥, and not G␣, is important for progesterone-induced maturation would be to examine the effect of injected GTP␥S on progesterone-induced maturation. GTP␥S would be expected to bind irreversibly to G␣ subunits, thus activating all GTP binding proteins within the oocytes. In addition, the activated G␣ subunits would no longer interact with G␤␥ subunits; thus, free G␤␥ activity would be expected to increase as well. Indeed, injection of GTP␥S has been shown to enhance signaling by G␤␥-activated G protein inward rectifying potassium channels (49,50). Injection of GTP␥S significantly blocked oocyte maturation in response to all concentrations of progesterone tested when compared with Mock-injected cells (Fig. 4A), providing further evidence that intrinsic G␣ activity is not stimulating oocyte maturation; rather, G␤␥ subunits are inhibiting maturation. (1 M; Fig. 4A). GTP␥S may therefore be inhibiting maturation simply by disabling the oocyte in a nonspecific fashion. A more elegant way to delineate the role of G␣ i and G␤␥ subunits in oocyte maturation would be to compare the effects of a constitutively active rat G␣ i protein to its matched wild-type isoform. If G␤␥ suppression is indeed the more important signaling event, then wild-type rat G␣ i should behave similarly to the Xenopus G␣ i , whereas the constitutively active G␣ i protein, which cannot bind to and sequester G␤␥, should have little effect on progesterone-induced maturation. As expected, wild-type rat G␣ i markedly enhanced progesterone-induced maturation at progesterone concentrations just below the EC 50 (50 nM). In contrast, constitutively active rat G␣ i containing a mutation that decreased the intrinsic GTPase activity of the protein (Q204L) had minimal effect on maturation (Fig. 4B). The lack of any substantial change in progesterone sensitivity in oocytes overexpressing the activated G␣ i clearly implies that G␣ i activity is not important in the progesterone-mediated maturation response. Note that expression of these two G␣ i proteins was equal as measured 2 S. R. Hammes, unpublished results.

Constitutively Activated G␣ i Had Minimal Effect on Progesterone-induced Oocyte Maturation-Injection of GTP␥S inhibited maturation in the presence of relatively high concentrations of progesterone
by Western blot analysis of extracts from injected oocytes (see Fig. 7C).
G␤␥ Effects Are Taking Place Upstream of Activation of Extracellular Signal-regulated Kinase Phosphorylation-One of the early signaling events triggered by progesterone is activation of the MAPK cascade. This process can be measured by immunoblotting for phosphorylated MAPK p42 (ERK1). Of note, Xenopus oocytes contain only the p42 isoform of ERK (51,52). Ooctyes were injected with cRNAs encoding the aforementioned G proteins to determine whether their effects were occurring proximal to this point. Overexpression of G␤␥ resulted in decreased phosphorylation of ERK1 in response to suboptimal concentration of progesterone (50 nM) over the course of 4 h (Fig. 5A), which correlated with the decrease in maturation observed over an 18-h time period (Fig. 3). In contrast, injection of the GRK minigene markedly enhanced both the rate and amount of progesterone-induced ERK phosphorylation, consistent with the GRK protein's ability to enhance progesterone-mediated maturation (Fig. 6A). Overexpression of transducin had a similar effect on ERK phosphorylation in response to progesterone (data not shown). Finally, injection of wild-type rat G␣ i enhanced both the rate and amount of pro- Approximately 70 oocytes were used for each condition, and this experiment was performed three times using two different preparations of oligonucleotides with nearly identical results. B, oocytes were injected with 10 mM Hepes (mock), nonspecific cRNA encoding the thrombin receptor (PAR1), or an equal amount of cRNA encoding the Xenopus G␣ i 2 protein (Xe-G␣ i 2). Spontaneous maturation was measured after 30 h of incubation in MSBH. 80 oocytes were used for each condition, and the y axis represents the percentage of mature oocytes per condition. This experiment was performed several times using multiple cRNA preparations with similar results. C, oocytes were co-injected with cRNA encoding either the wild-type (G␣ i 2-WT) or modified (G␣ i 2-Mod) Xe-G␣ i 2 plus either the sense or antisense oligonucleotide directed against Xe-G␣ i 2 (concentration, 2 mg/ml). Spontaneous maturation was measured after 30 h of incubation in MSBH, and the y axis represents the percentage of inhibition of maturation by the antisense oligonucleotides compared with matching sense oligonucleotides. 30 oocytes were used for each condition, and this experiment was performed twice using different cRNA preparations with similar results. gesterone-induced ERK phosphorylation, whereas expression of the constitutively active G␣ i (Q204L) had minimal overall effect on MAPK activation (Fig. 7A). Although phosphorylated ERK was detected at the 2-h time point in G␣ i -Q204L-injected oocytes, the amount of phosphorylation at 4 h remained similar to mock injected cells. This contrasted with the markedly increased phosphorylation of ERK1 after 4 h detected in G␣ i -WT expressing cells. Of note, Western blot analysis of these samples using an antibody against rat G␣ i protein revealed that both G␣ i subunits were equally expressed in the injected oocytes (Fig. 7C). In addition, equal amounts of sample were present in each lane, as confirmed by blotting for total ERK (Figs. 5B, 6B, and 7B). Together, these data indicate that G␤␥ effects are occurring early in the progesterone-mediated signaling pathway, upstream of MAPK activation.

DISCUSSION
Progesterone induces maturation of Xenopus oocytes via a nongenomic pathway. Very little is currently understood about the early progesterone-induced signaling events, including the identity of the putative membrane progesterone binding protein. Previous work reported the presence of specific high affinity progesterone binding sites on Xenopus membranes (24,25). Our results confirm these reports, demonstrating high affinity progesterone binding sites (K d ϭ 3 nM) on Stage V-VI oocyte membranes but not on Stage I-III oocyte membranes ( Fig. 1). The presence of high affinity sites almost exclusively on maturation-competent oocytes, combined with the similarities between the K d for high affinity progesterone binding and the EC 50 for progesterone-induced maturation (ϳ100 nM; Fig.  1A), suggest that the specific binding of progesterone to these sites is linked to the maturation process. An active role for the lower affinity sites cannot be completely ruled out, however, given that their equilibrium constant (140 nM) is also close to the EC 50 value.
Earlier work has implicated G␣ i as a possible mediator of progesterone signaling in Xenopus oocytes (16, 26 -29). The inability of constitutively active rat G␣ i to significantly alter progesterone-mediated maturation and MAPK signaling, whereas wild-type rat G␣ i enhanced both maturation and MAPK activation (Figs. 4B and 7A), argues against this hypothesis. Instead, these data suggest that intrinsic G␣ i activity (for example, its ability to inhibit adenylyl cyclase) is unimportant for progesterone-induced maturation. We propose that the changes in progesterone sensitivity observed by modulating G␣ i levels in oocytes are in fact due to changes in free G␤␥ levels. If so, G␤␥ must be acting as an inhibitor of maturation: the more free G␤␥, the less sensitivity to progesterone-induced maturation. Accordingly, reduction of endogenous G␣ i , which would increase intracellular free G␤␥, inhibited progesteroneinduced maturation ( Fig. 2A). Overexpression of G␣ i , which would deplete free G␤␥, enhanced progesterone-induced maturation as well as MAPK activation (Figs. 2B, 4B, and 7A). Finally, constitutively active G␣ i , which cannot bind to G␤␥ and therefore would not alter endogenous G␤␥ levels, minimally affected both progesterone-mediated signaling and maturation (Figs. 4B and 7A).
In further support of the inhbitory role of G␤␥ in maturation, overexpression of bovine G␤␥ in oocytes markedly reduced the sensitivity of the maturation response to progesterone. More importantly, sequestration of endogenous G␤␥ by overexpression of either transducin or the more specific G␤␥-binding mini-GRK protein enhanced responses to progesterone. This suggests that endogenous G␤␥ plays a real role in attenuating Xenopus oocyte maturation in response to progesterone and perhaps in resting oocytes as well.
It is still possible that progesterone-induced maturation is mediated by a G␣ subunit other than G␣ i . An argument against this possibility is that injection of GTP␥S, which activates all G␣ subunits, still inhibited maturation (Fig. 5A). In addition, progesterone-mediated activation of G␣ s would be unlikely given the observed decrease in cAMP after progesterone treatment (26 -29). Formal exclusion of this possibility is difficult, however, and will involve examining the effects of overexpression of all of the known classes of G␣ subunits, and perhaps of their respective constitutively active forms, on progesteroneinduced maturation. To date, in addition to the G␣ subunits presented in these studies, G␣ i 2, as well as multiple mammalian and Xenopus G␣ i subunits, enhance maturation when overexpressed in oocytes (data not shown).
Our results are strikingly different than earlier data reported using starfish oocytes. Injection of G␣ i subunits into starfish oocytes inhibits, whereas injection of G␤␥ enhances, maturation (35,37), suggesting that activation of G␣ i and release of G␤␥ are critical for starfish ooctye activation. These contrasting results are not surprising, however. First, 1-methyladenine-induced maturation of starfish oocytes is inhibited by pertussis toxin (53)(54)(55), whereas progesterone-induced maturation of Xenopus oocytes is not (26, 56 -58). 2 This supports the involvement of G␣ i in starfish oocyte maturation and argues against its importance for Xenopus oocyte maturation. Second, 1-methyladenine binding to starfish oocyte membranes decreases in the presence of GTP (53), whereas GTP␥S had no effect on high affinity progesterone binding to Xenopus oocyte membranes (Fig. 1D), suggesting that 1-methyladenine binds to its membrane receptor in a fashion typical for a GPCR agonist, whereas progesterone does not. Taken together, these results indicate that early signaling events in oocyte maturation may be quite different between starfish and Xenopus; thus, comparisons between the two should be made with caution.
Although G␤␥ promotes re-entry into the cell cycle in starfish oocytes, the homologous complex in Saccharomyces cerevisiae, Ste4p/Ste18p, arrests the cell cycle in the haploid mating response (59). The mechanism behind this cell cycle arrest appears to involve Ste4p/Ste18-induced activation of the MAPK pathway. Experiments in mammalian systems have also demonstrated that G␤␥ can activate the MAPK pathway (60,61). Although our data demonstrate that G␤␥ can mediate cell cycle arrest in Xenopus oocytes, G␤␥ does not appear to activate the MAPK cascade. In fact, overexpression of G␤␥ attenuates, whereas sequestration of G␤␥ enhances, ERK1 phosphorylation over the course of 4 h (Figs. 5-7). These results are supported by earlier work demonstrating that progesterone-induced MAPK activation in Xenopus oocytes is primarily mediated by the Mos protein and not by G␤␥ (31,44,(62)(63)(64)(65). In fact, G␤␥ may be playing a role in regulating Mos protein expression, because preliminary results indicate that sequestration of G␤␥ enhances Mos expression (data not shown).
The observation that endogenous G␤␥ inhibits maturation offers an intriguing model whereby maturation is held in check by low levels of constitutive G␤␥ signaling. Resting G␤␥ activity has been observed before in Xenopus oocytes and in other systems. Constitutive G␤␥-mediated G protein-coupled inward rectifying potassium channel activity in resting oocytes can be inhibited by sequestration of free G␤␥ (45,48,66). In addition, free G␤␥ activity has also been proposed to constitutively permit endocytosis in mammalian CV1 cells (67). Progesterone may be inducing maturation by antagonizing this constitutive inhibitory G␤␥ activity. For example, progesterone could attenuate G␤␥ activity by binding directly to G␤␥ or to another molecule that interferes with G␤␥ signaling. Of note, we observed no increase in progesterone binding to membranes from oocytes overexpressing G␤␥ subunits (data not shown). Alternatively, one tantalizing possibility is that progesterone acts as an antagonist, or even an inverse agonist, on an unknown constitutively active GPCR that signals via G␤␥ to inhibit maturation. This type of "nonagonist" interaction between progesterone and a GPCR is consistent with our binding data demonstrating no change in progesterone binding to its receptor in membranes treated with GTP␥S. Interestingly, this release of inhibition model is reminiscent of the oxytocin receptor story, where progesterone may act by competing with oxytocin for binding to its GPCR (38). Of note, oxytocin had no effect on any of the described experiments with progesterone (data not shown). Improving our understanding of early signaling events in progesterone-mediated maturation may provide us with useful clues in searching for the elusive membrane progesterone binding protein and may prove helpful in elucidating the mechanisms involved in other nongenomic steroid signaling pathways as well.
Acknowledgments-We are extremely thankful for all of the support and help from Dr. Shaun Coughlin at the University of California at San Francisco. Interest for this work began in his laboratory and his generosity and ideas were critical for this project. We also thank Alex Yi and Dr. L. Jan at the University of California at San Francisco for invaluable help, ideas, and reagents. Special thanks is also given to Drs. A. Gilman, E. Ross, and R. Auchus for critical review of this manuscript.

FIG. 7. The effects of G␣ i proteins on progesterone-induced phosphorylation of p42 MAPK.
A and B, oocytes were injected with either 10 mM Hepes (Mock) or the indicated cRNAs and treated as described in Fig. 5. C, extracts from the t ϭ 0 samples of A were resolved by SDS-polyacrylamide gel electrophoresis and blotted with an anti-G␣ i polyclonal antibody (A. Gilman, University of Texas Southwestern Medical Center). 15 oocytes were used for each sample, and these experiments were performed three times using multiple cRNA preparations with similar results.