Estrogens are key mediators of increased uterine contractility at labor. We sought to determine whether membrane-associated estrogen receptors, such as the recently described seven-transmembrane receptor G protein-coupled receptor 30 (GPR30), mediated some of this effect. Using human myometrium obtained at term cesarean section before or after the onset of labor, we demonstrated the presence of GPR30 mRNA and protein using quantitative RT-PCR and Western blotting. GPR30 receptor was localized to the cell membrane and often colocalized with calveolin-1. Using the specific estrogen membrane receptor agonist G-1 and myometrial explants, we showed that membrane receptor activation led to phosphorylation of MAPK and the actin-modifying small heat shock protein 27. Using myometrial strips incubated with G-1 or vehicle we demonstrated that estrogen membrane receptor activation increased the myometrial contractile response to oxytocin. These data suggest that activation of the plasma membrane estrogen receptor GPR30 likely participates in the physiology of the human myometrium during pregnancy and identifies it as a potential target to modify uterine activity.
Labor in the majority of mammals, including women, is associated with increased production of estrogens (1–3). Estrogens have been shown to play a critical role in reorganizing the myometrium at the end of pregnancy to allow the onset of labor by stimulating the expression of a group of genes described as contractility-associated proteins, including connexin 43 and cyclooxygenase 2 (4–6). The estrogens have been assumed to act via the classical nuclear estrogen receptors. In human myometrium the preponderant nuclear estrogen receptor is the estrogen receptor α (ERα) (7, 8). Studies using quantitative RT-PCR have identified increased ERα in myometrium at the time of labor (4, 8).
Recently, G protein-coupled receptor 30 (GPR30) (9–12), a novel ER associated with cell membranes, has been identified, that has many different properties to the classical nuclear ER. The mRNA and protein of GPR30 are expressed in many cancer cells (13–17) and in a variety of sites including brain (18) and ovary (19) and in vascular (20) and other tissues (21, 22), but its presence in the pregnant uterus has not been examined. GPR30 is a typical G protein-coupled receptor (9–12) that can induce rapid nongenomic actions, which include activation of adenylate cyclase, phosphatidylinositol 3-kinase, p42/44 MAPK, and intracellular calcium mobilization. Notably, each of these pathways has also been linked to myometrial activation at the time of labor (23–25). It has been previously shown that GPR30 is expressed in human arterial smooth muscle where it affects smooth muscle function through a nongenomic mechanism (20). Recently, it has been reported that the effects previously ascribed to GRP30 may instead be due to activation of a shortened ER, a variant known as ERα36 (26). This variant lacks the activation domain of the classical genomic ER and has altered ligand specificity responding to the G-1 ligand (27) that is inactive at classical receptors but activates GPR30. However, ERα36 is not expressed in human myometrium (28).
We have previously demonstrated that actin polymerization at the time of human labor is associated with the phosphorylation of the small heat shock protein 27 (HSP27) (29).
We therefore hypothesized that the effects of estrogen in stimulating the onset of labor are mediated, at least in part, via activation of a GPR30-signaling cascade using MAPK that leads to phosphorylation of HSP27 and increased myometrial contractility.
Materials and Methods
Experimental subjects, myometrial sample collection, and cell culture
All experimental procedures performed in this study were approved by the University of Newcastle Ethics Committee in accordance with the institutional guidelines of the John Hunter Hospital, Australia. Informed consent was obtained from all women. Women who were delivered by elective cesarean section at term and showed no signs of labor (uterine contractions and/ or cervical changes) formed the nonlaboring group. Women who entered spontaneous and established labor (at least 2 h) but required emergency cesarean section formed the laboring group. Clinical indications for emergency cesarean section included breech, fetal distress, prolonged labor, and previous cesarean section. All subjects were between 37 and 40 wk gestational age. Women exhibiting clinical signs of infection were excluded. Myometrial tissue strips (5 × 10 mm) were collected from the upper margins of the lower uterine segment, freed of connective tissue, and then divided; aliquots were immediately snap frozen in liquid nitrogen and stored at −80 C for subsequent mRNA and protein analysis or fixed in 10% neutral buffered saline followed by paraffin embedding for immunohistochemical analysis, or prepared for explants or bioassay. Normal term placenta was also collected at deliveries to act as a positive control for GPR30 detection. The breast cancer cell line MCF-7, which has previously been shown to express GPR30 (12), was cultured in DMEM (Invitrogen, Mulgrave, Australia) with 2 mml-glutamine, 1 mm pyruvate, 1 × antibiotic antimycotic (Anti-Anti, Invitrogen) and 10% fetal calf serum (Invitrogen) in a humidified environment at 37 C and in 5% CO2 and 95% air overnight, and cells were extracted for mRNA and protein for RT-PCR and Western analysis, respectively.
Quantitative RT-PCR
Total RNA was extracted and purified from tissue specimens and MCF-7 cells using the acid-phenol method (30), treated with deoxyribonuclease to remove any contaminating genomic DNA, and quantified, and 1.5 μg were reverse transcribed with random primers using Superscript II reverse transcriptase (Life Technologies, Inc., Melbourne, Australia). The sequences of primers for GPR30 are: forward, 5′-CGTCCTGTGCACCTTCATGT-3′; and reverse, 5′-AGCTCATCCAGGTGAGGAAGAA-3′. Primers for 18s rRNA (PN4319413E) were obtained from Applied Biosystems (Foster City, CA). Assays were validated for all primer sets by confirming that single amplicons of appropriate size and sequence were generated according to predictions. Quantitative PCR was performed in the presence of SYBR Green (Applied Biosystems), and amplicon yield was monitored during cycling in an ABI PRISM 7700 Sequence Detector (Applied Biosystems) that continually measured fluorescence caused by the binding of the dye to double-stranded DNA. The cycling conditions were 50 C for 2 min, 95 C for 10 min, and then 40 cycles of 95 C for 15 sec, 60 C for 1 min. The cycle at which the fluorescence reached a set threshold (cycle threshold) was used for quantitative analyses. The cycle threshold in each assay was set at a level at which the exponential increase in amplicon abundance was approximately parallel between all samples. All mRNA abundance data were expressed relative to the abundance of the constitutively expressed 18s rRNA.
Western blotting
Samples of myometrium and additional samples of placenta were crushed under liquid nitrogen and dissolved in two-dimensional extraction buffer (29). MCF-7 cells were scraped and dissolved in two-dimensional extraction buffer. The amount of protein was measured using the two-dimensional Quant kit (Amersham Biosciences, Piscataway, NJ). Myometrial, placental, and MCF-7 proteins (30–40 μg) were loaded on precast 10% NuPAGE acrylamide gels (Invitrogen) and separated electrophoretically using the Xcell SureLock Mini-Cell system at 190 V (constant) until the bromophenol blue dye migrated to the bottom of the gel. Proteins were transferred to an activated polyvinylidene difluoride membrane (Amersham Biosciences) by using a transfer system (Bio-Rad Laboratories, Inc., Hercules, CA) under constant 100 V for 75 min in a cold room (4 C). Membranes were blocked with 5% skim milk in Tris-buffered saline supplemented with 0.01% Tween 20 (TBST) for 2 h and then incubated with antibodies for GPR30 (1:2500; Novus Biologicals, Littleton, CO), total 42/44-MAPK (1:1000, Cell Signaling Technology, Inc., Danvers, MA), phospho-42/44 MAPK (1:1000, Cell Signaling Technology), total HSP27 (1:2500), pHSP27-Ser15 (1:1000), pHSP27-Ser78 (1:2000), and pHSP27-Ser82 (1:2000) in 5% skim milk/TBST for 12 h at 4 C. HSP27 antibodies were purchased from Upstate Biotechnology, Inc. (Lake Placid, NY). Membranes were then washed five times (5 min per wash) in TBST, and the appropriate secondary antibody (1:2000) was added for 1 h at room temperature. Immunoreactive protein was detected with enhanced chemiluminescence (GE Healthcare, Piscataway, NJ) using an Intelligent Dark Box LAS-3000 Imager (Fuji Photo Film, Tokyo, Japan).
Deglycosylation of GPR30
Myometrial protein (40 μg) was digested separately with N-glycosidase F (PNGase F) (New England Biolabs, Beverly, MA; catalog no. P0704) and protein deglycosylation mix (New England Biolabs, catalog no. P6039) for 12 h following the manufacturer's instructions. After digestion, proteins were subjected to Western blotting for GPR30 as described in Western blotting.
Immunohistochemical localization
Myometrial tissue sections (6 μm) were prepared for immunohistochemistry according to standard protocols. Sections were treated with 3% hydrogen peroxide in methanol for 30 min. The sections were washed with 70% alcohol. Sections were boiled with antigen retrieval solution (ImmunoSolution PTY LTD, Everton Park QLD, Australia) at 95 C for 10 min. Sections were blocked with blocking solution (0.5% BSA, 0.05% saponins, 0.05% sodium azide) for 30 min. Tissue sections were incubated with primary antibodies for GPR30 (1:50), Caveolin-1 (1:25, Novus Biologicals) and protein disulfide isomerase (1:25, Abcam, Cambridge, MA) in blocking solution overnight. Tissue sections were washed with TBST and then incubated with secondary antibody conjugated with fluorescence probes for 2 h. Each slide was stained with 4′,6-diamidino-2-phenylindole and mounted with Prolong*Gold antifade reagent (Invitrogen). The slides were visualized using an Olympus confocal fluorescence microscope (Olympus IX81 with fluorview FV1000 software, Mt. Waverley, Australia).
Bioassay for contractile activity in myometrial strips
Tissues were stored in saline at 4 C for up to 16 h before bioassay. Tissues were washed several times with saline, cut into small strips (2 × 8 mm), and incubated in serum-free and phenol red-free DMEM with 2 mml-glutamine, 1 mm pyruvate, 1 × antibiotic antimycotic in a humidified environment at 37 C and in 5% CO2 and 95% air overnight. On the experimental day, tissues were washed with PBS and supplemented with fresh media. Tissues were then treated with vehicle (dimethylsulfoxide-containing media), 17β-estradiol (E2) (10−6m) and the GPR30 agonist G-1 (10−6m) (27) for 12–14 h. To assess effects on contractility, myometrial strips were suspended in a 15 ml organ bath containing a Krebs' buffer plus investigational agents equilibrated with 95% O2/5% CO2 at 37 C. Force generation by the strips was measured with a Grass FTO3c force transducer interfaced with a Grass Polygraph (Grass Instruments, Quincy, MA). Data were acquired with a Maclab 8 (ADInstruments, Melbourne, Australia). Strips were progressively stretched to a resting tension of 1 mN over a period of 60–90 min, accompanied by replacement of the buffer in the organ bath every 15 min. Strips were then allowed to develop spontaneous rhythmic contractile activity (1–2 h). After rhythmic contractions were established, strips were treated with stimulatory doses of oxytocin (0.01 nm to 1000 nm). All tissues maintained their viability and contractile nature after 2 d of culture in phenol red-free and serum-free DMEM. Dose responses were fitted with a sigmoid curve using PRISM5 software (GraphPad Software, Inc., La Jolla, CA).
Explant cultures of myometrium
Myometrial samples were washed several times with saline. Tissues were cut into small pieces (1 × 1 mm), and five to six pieces were incubated in serum-free and phenol red-free DMEM with 2 mml-glutamine, 1 mm pyruvate, 1 × Anti-Anti in a humidified environment at 37 C and in 5% CO2 and 95% air for 14 h. Explants were then washed, supplemented with fresh media, and treated for 15 min with different doses of the GPR30 agonist, G1(10−9m, 10−8m, 10−7m, 10−6m). Tissues were then frozen in liquid nitrogen for subsequent Western blot analysis of second messenger pathways.
Data analysis
GPR30 mRNA expression in laboring and nonlaboring myometrium was analyzed using the Wilcoxon rank-sum nonparametric test. Data analysis of the bioassay results was performed by nonlinear regression using GraphPad PRISM5 software, and the data were expressed as sigmoid dose-response curves, providing results for the oxytocin concentration inducing half-maximal contraction (EC50) and maximum-fold increase in contraction (Bmax). Stata 11.0 software (StataCorp, College Station, TX) was used for additional statistical analysis of the myometrial contractile responsiveness data. The association of contraction force using E2 or G-1 compared with control under increasing doses of oxytocin was modeled by random-effects generalized least squares regression. Contractile force and oxytocin concentration were both expressed as log10 transformed. The model contained a random intercept term for subject to adjust for repeated measurements. Results are reported with P values, regression coefficients, and 95% confidence intervals (CI). Statistical significance of HSP27 phosphorylation data was determined by one-way ANOVA (Dunnett's test). A two-tailed significance level of 5% was used throughout.
Results
Expression of GPR30 in myometrium
Real time RT-PCR data showed that mRNA for GPR30 was readily detectable in nonlaboring and laboring term human myometrium, although there was no significant difference in the pattern of expression between these groups. The GPR30 mRNA in term non-laboring myometrium, and spontaneous laboring myometrium with respect to 18s rRNA expression and a calibrator sample (MCF-7) were 3.26 ± 0.37 (n = 9), and 3.94 ± 0.54 (n = 6), respectively. As a positive expression control, term human placenta also expressed GPR30. Western blotting using antibodies to GPR30 and myometrial tissue obtained at term from women both before and in labor detected a band at 38 kDa (Fig. 1A), consistent with a nascent nonglycosylated form of GPR30 as previously described in human placenta (11) and rat central nervous system (31). Furthermore, a strong band at approximately 120 kDa, consistent with a mature glycosylated form, was also detected in myometrium and placenta. GPR30 protein expression was also found in MCF-7 cells predominantly in a nonglycosylated form at approximately 38 kDa and a glycosylated form at about 65 kDa. All bands corresponding to nonglycosylated and glycosolated forms were abolished using the specific blocking peptide (Fig. 1B). Glycosylation of GPR30 was demonstrated using an N-glycosidase enzyme PNGase F. The 120-kDa mature GPR30 was reduced to an 80-kDa form with PNGaseF digestion for 12 h (Supplemental Fig. 1 published on The Endocrine Society's Journals Online web site at http://endo.endojournals.org). When a deglycosylation mix that contained both N-glycosidase and O-glycosidase was used, a 50-kDa (indicated by a double star in Fig. 1C) band was observed in addition to the 80-kDa band (indicated by single star in Fig. 1C).
GPR30 protein expression in human myometrium. A, Membrane protein was extracted from term nonlaboring myometrium (Myo1, Myo 2), placenta, and MCF-7 cells. Western blot analysis performed on the extracts revealed nonglycosylated and glycosylated forms of GPR30 in this representative blot. HSP27 was used as a loading control. More than 10 different myometria have been analyzed with similar results. B, GPR30 immunodetection was completely abolished by addition of antibody-specific blocking peptide. C, Enzymatic digestion with a deglycosylation mix, containing both N-glycosidase and O-glycosidase, removed glycosylation from glycosylated GPR30 shifting it to two lower molecular mass bands (*, 80 kDa; and **, 50 kDa). HSP27 was used as a loading control.
GPR30 localization in human myometrium
Immunohistochemistry and confocal microscopy demonstrated that, in term nonlaboring and laboring myometrial tissues, GPR30 is found localized to the plasma membrane and in some areas colocalized with caveolae in myometrial smooth muscle cells (Fig. 2, A–D). The GPR30 did not localize to the endoplasmic reticulum in the myometrial smooth muscle cells as shown using protein disulfide isomerase as an endoplasmic reticulum marker (Fig. 2, E–H).
GPR30 and caveolin-1 colocalization in human myometrium. Dual labeled fluorescence immunohistochemistry and confocal microscopy demonstrating GPR30 detection (red, A, C, E, and G), cellular periphery caveolin-1 detection (green, B and C) and the endoplasmic reticulum marker PD-1 (green, F and G) in NL myometrial sections. Yellow labeling (C) indicates areas of colocalization and demonstrates regions in which GPR30 localizes to caveolae structures. The lack of green labeling in panel C indicates that almost all caveolin-1 is colocalized with GPR30. However, not all GPR30 is colocalized with caveolin-1. GPR30 is not localized with the endoplasmic reticulum in term myometrium (G). Blue labeling indicates the nucleus, whereas the scale bar represents 5 μm. Differential interference contrast (DIC) image of tissue section C is presented in D and that of G in H, demonstrating the myometrial morphology. Experiments were repeated in at least three individual term myometria.
G-1 and E2 increase the contractile responsiveness of myometrium to oxytocin
Myometrial strips treated with the estrogen membrane receptor agonist, G-1 (10−6m) and E2 (10−6m), for 12 h showed an increased response to oxytocin compared with vehicle controls (Fig. 3). With increasing doses of oxytocin, myometrial strips showed sigmoidal contractile responses. LogEC50 (Log M) values (with 95% CI) are −8.528 (−9.284, −7.772), −8.604 (−9.156, −8.052), and −8.753 (−9.243, −8.263) for control, G-1, and E2, respectively. Maximum responses (Bmax) are 256.6 (220.2, 293.0), 336.8 (296.2, 377.5), and 348.7 (311.4, 385.9) for control, G-1, and E2, respectively. The Bmax (% of basal value) at 95% confidence intervals for G-1 and E2 did not overlap with the confidence interval for controls, indicating significant differences, whereas the 95% CI values for G-1 and E2 overlapped. When contractile force was modeled using random-effects linear regression (with force transformed using log10 to satisfy the regression assumptions), treatment with either G-1 or E2 was associated with a significant increase in contractile force compared with controls [G-1: P = 0.001, coefficient (95% CI) = 0.083 (0.033, 0.133); E2: P < 0.001, coefficient (95% CI) = 0.127 (0.079, 0.174)] and contractile force increased significantly with increasing oxytocin dose [P < 0.001, coefficient (95% CI) = 0.061 (0.051, 0.071)]. Contractile force with E2 was not significantly greater than that with G-1 [P = 0.104, coefficient (95% CI) = 0.044 (−0.009, 0.096)]. The direct addition of G-1 and E2 into the strips suspended in organ bath did not produce any immediate effect on the spontaneous activity of the myometrium (Supplemental Fig. 2).
Effect of G-1 and E2 treatments on the myometrial response to oxytocin. Contraction responses of human term nonlaboring myometrial strips (n = 5 individual term myometria) to oxytocin treatment was measured after 12 h treatment with 10−6m G-1, 10−6m E2 and vehicle control. Contraction responses (area under curve) were compared as a percentage of basal spontaneous activity before treatment. Dose responses were fitted with sigmoid curves. LogEC50 (Log M) values ± se of the mean (sem) are −8.528 ± 0.379, −8.604 ± 0.275, and −8.753 ± 0.245 for control, G-1, and E2, respectively. Maximum responses (Bmax) (% of base value) ± sem were 256.60 ± 18.25, 336.80 ± 20.28, and 348.70 ± 18.65 for control, G-1, and E2, respectively. Random-effects generalized least squares regression and nonlinear regression both indicated that G-1 and E2 increased the response of the tissue compared with control, and no difference was detectable between the effects of G-1 and E2. Con., Control.
G-1 increases the phosphorylation of HSP27
In myometrial explant cultures, 15 min treatment with the GPR30 agonist G-1 significantly (Dunnett's test) increased the phosphorylation of the Ser-78 site of HSP27 3-fold (Fig. 4, A and B) and also increased MAPK phosphorylation (Fig. 4, A and C). Furthermore, this action of G-1 led to phosphorylation of HSP27 at two additional serine sites (15 and 82) (Fig. 5, A–D).
G-1 induces phosphorylation of HSP27 and MAPK in human myometrium. A, Treatment with G-1 induced a significant, dose-dependent increase in HSP27-ser78 phosphorylation in human myometrial explants. Phosphorylation of p44/42 MAPK also increased. Densitometric analysis of HSP27 phosphorylation (B) and p44/42 MAPK (C) from three independent experiments. Statistical significance was determined by one-way ANOVA (Dunnett's test).
Effect of GPR30 activation on HSP27 phosphorylation in human myometrium. A, The GPR30 agonist G-1 phosphorylates HSP27 at serine residues 15, 78, and 82 whereas oxytocin induces phosphorylation at serine residues 15 and 78. B–D, Densitometric analyses of serine 15, 78, and 82 phosphorylation in response to G-1 are presented in panels B, C, and D, respectively.
Discussion
In this study we provide evidence for the presence of the novel seven-transmembrane receptor membrane-associated ER GPR30 in human myometrium. We show that the GPR30 receptor is localized to myometrial cell membranes. Activation of GPR30 by the agonist G-1 leads to phosphorylation of MAPK and of the small heat shock protein HSP27. We further demonstrate that activation of the G-1-mediated pathways leads to increased myometrial contractile responses to oxytocin.
GPR30 mRNA was first described in human placenta, brain, and heart (9, 10). Since then, GPR30 mRNA and protein have been demonstrated in a range of different animal tissues including rat central nervous system (31), rat Sertoli cells (21), hamster ovary (19), rodent arterial blood vessels (20), and mice pancreatic islets (32). In the human, mRNA and protein for GPR30 have been shown to be present in different cancer tissues (13–15, 17, 33), blood vessels (20), endometrium (34), and early pregnancy decidua (34). Here we extend the range of tissues in which GPR30 is known to exist to the human myometrium. Our data include both mRNA and protein demonstrating local synthesis of this receptor. Although glycosylation of GPR30 has been previously reported (35), here we show differential glycosylation patterns for GPR30 in a cancer cell-line MCF-7 compared with human myometrium and placenta. Because the glycosylation pattern of other G protein-coupled receptors is important in the targeting of the receptor to the plasma membrane, it is likely to play a similar role for GPR30 (36). We were, therefore, interested in investigating the cellular localization of GPR30 in the myometrium.
GPR30 has been described in several cellular locations in different tissues. Thomas et al. (11) reported GPR30 as a plasma membrane receptor in transfected human embryonic kidney-293 cells, whereas Revankar et al. (12) described GPR30 as an intracellular receptor that was specifically localized to the endoplasmic reticulum in transfected human embryonic kidney-293 cells. Additionally, GPR30 was expressed in the perinuclear endoplasmic reticulum of hamster ovary cells (19). Immunoelectron microscopy showed that GPR30 was expressed in the Golgi apparatus of rat neurons (31). A recent study presented evidence for nuclear localization of GPR30 in cancer-associated fibroblasts (37). In the human myometrium, the majority of GPR30 was localized to the plasma membrane where a proportion was colocalized with the calveoli. Interestingly, GPR30 did not localize to the endoplasmic reticulum. In this context, caveolin-1, which is major component of the lipid raft, is the important regulator of excitation-contraction coupling in smooth muscle (38). Caveolin-1 can interact with protein kinase C, rho-associated protein kinase, epidermal growth factor receptor, actin, and Gα-subunits (39) to act as a scaffolding molecule in smooth muscle cell. Interestingly, GPR30 was found to transactivate epidermal growth factor receptor and activate Gs-protein (11, 12). Therefore, the localization of GPR30 in the plasma membrane with calveoli-signaling compartment suggests a role in this tissue as a plasma membrane receptor for ligands.
GPR30 has been linked to the action of estrogens in several different tissues based on the use of the agonist compound G-1, which had been presumed to act via GPR30 and that is inactive at classical nuclear ER (27). However, evidence has recently been presented indicating that the effects of the G-1 compound are not specific to GPR30 and that the compound can also activate the ERα variant, known as ERα36, which is generated from a promoter located in the first intron of the ERα gene (26). These studies have shown that this variant can localize to the plasma membrane and account for actions of G-1 previously ascribed to GPR30. Further, GPR30 can activate the transcription of ERα36 via a Src/MAPK/activator protein 1 pathway. Some work suggested that GPR30 and ERα36 may associate (26). However, data in MCF-7 cells showed that G-1-activated calcium signaling was completely abolished when GPR30 was knocked down even though the MCF-7 cells expressed ERα36 in considerable amounts (40). These data confirm the ability of G-1 to increase intracellular Ca2+ independently of ERα pathways, and no ERα36 has been detected in human myometrial tissue using Western blotting and an antibody that detects this variant (28). Taken together these data suggest that GPR30 is a potential estrogen membrane receptor in the human myometrium.
In the myometrium MAPK has previously been shown to have actions that promote contraction (41, 42). Both endocrine signals (41) and mechanical stretch (42) phosphorylate MAPK to induce myometrial contraction. In addition, myometrial ERK activation is related to focal adhesion signaling (43) that facilitates contraction. We have previously shown that the onset of labor is associated with a fall in myometrial levels of αB-crystallin and increased phosphorylation of HSP27; these events are linked to the formation of fibrillar actin that is required for muscle contraction (29). Data suggest that MAPK is involved in the phosphorylation of the HSP27 (44). At labor the HSP27 becomes phosphorylated at residue Ser-15 (29). The change in HSP27 phosphorylation at labor is associated with a change in the subcellular localization of the HSP27 from an association with αB-crystallin to an association with fibrillar actin (29). The implication is that, in the myometrium, activation of MAPK pathways leads to actin polymerization that is essential for myometrial contractility. For these reasons we sought to determine whether activation of GPR30 was associated with activation of MAPK and with phosphorylation of HSP27 (29). Our data demonstrate an increase in phosphorylation of MAPK and of HSP27 after application of the GPR30 agonist G-1. Interestingly, HSP27 was phosphorylated at three serine residues after activation of GPR30 by the agonist G-1 in contrast to the two that are phosphorylated by the activation of oxytocin receptors. The role of different kinases and phosphatases in this differential effect remains to be determined. The potential for a differential modification of contractile responses exists.
The effect of GPR30 activation on myometrial contractility was examined using myometrial strips under tension and stimulated with oxytocin. Using this model the GPR30 agonist G-1 produced an increase in myometrial contractility that was indistinguishable from that of E2. These data suggest a role for G-1-activated membrane ER in regulating the contractility of human myometrium. It is, however, notable that concentrations of GPR30 mRNA and protein did not change across the not-in-labor to in-labor continuum. It may be that concentrations of the GPR30 had increased before term and that no further increase is required for any physiological action at term. A further possibility is that G-1 may exert actions on ERα36 in myometrial cells but that the concentrations of ERα36 are below the level of detectability by Western blotting. Alternatively, it may be that regulation of estrogen action at term is determined more by changing concentrations of estriol and E2 as term approaches. We have previously reported that, as term approaches, there is a marked increase in the ratio of estriol to E2 in maternal blood (3). In this context it is of interest that estriol has been reported to act as an antagonist at GPR30 receptors in an estrogen-negative breast cancer cell line (45).
Together these data indicate a potential new role for GPR30 in mediating the actions of estrogens in the human myometrium. The roles may include increasing myometrial contractility through actions on MAPK and HSP27 phosphorylation as well as other potential actions including increasing intracellular calcium, myometrial cell proliferation, and myometrial cell hypertrophy (46). Targeting GPR30 specifically to modify myometrial physiology represents a potential therapeutic intervention in disorders of myometrial responses to estrogen.
Acknowledgments
We thank A. Wright, T. Finnegan, and all the staff of the Obstetrics and Gynaecology Divisions at the John Hunter Hospital, Newcastle, Australia, and KK Children's and Women's Hospital, Singapore, for patient recruitment and sample collection. We also acknowledge Dr. Phoebe Jennings from School of Biomedical Sciences and Pharmacy of the University of Newcastle, New South Wales, Australia, for her advice regarding fluorescence confocal microscopy. We thank National Health and Medical Research Council, Australia for their generous economical support (grant number: 631027).
Disclosure Summary: The authors of this manuscript have nothing to declare.
Abbreviations
- E2
17β-Estradiol
- ER
estrogen receptor
- GPR30
G protein-coupled receptor 30
- HSP27
heat shock protein 27
- PNGase F
N-glycosidase F
- TBST
Tris-buffered saline supplemented with 0.01% Tween 20.
