A direct role for arrestins in desensitization of the luteinizing hormone/choriogonadotropin receptor in porcine ovarian follicular membranes

Abstract

The luteinizing hormone/choriogonadotropin (LH/CG) receptor (R) is a heptahelical R that, upon agonist binding, activates the stimulatory guanine nucleotide-binding protein (Gs) and the downstream effector adenylyl cyclase (AC). Like other G protein-coupled Rs, the LH/CG R subsequently exhibits reduced agonist-dependent effector activity, or desensitization, in response to saturating agonist. Unlike desensitization of many other G protein-coupled Rs, the in vivo desensitization response of LH/CG R-stimulated AC activity of ovarian follicles to the preovulatory surge of LH can be mimicked under cell-free conditions. Based on evidence that porcine ovarian follicular membranes unexpectedly contained beta-arrestin-1, the role of arrestins in desensitization of the LH/CG R was investigated. Results showed that neutralizing arrestin antibodies blocked the development of desensitization and that desensitization was rescued with a synthetic peptide corresponding to the antibody-binding epitope on beta-arrestin-1. These results suggest that endogenous beta-arrestin-1 participates in agonist-dependent desensitization of the LH/CG R. Addition of recombinant purified beta-arrestin-1 mimicked human chorionic gonadotrophin to promote desensitization of human chorionic gonadotrophin-stimulated AC activity, in the presence of the ATP phosphorylation antagonist adenylyl-imidodiphosphate, with an ED50 of approximately 0.1 nM. Increased levels of an 87-kDa protein reactive with glycoprotein hormone R-reactive antibody, consistent with the LH/CG R, coimmunoprecipitated with follicular membrane beta-arrestin-1 in response to LH/CG R activation compared with unactivated R. Taken together, these results show that ovarian follicles contain membrane-associated beta-arrestin-1, that beta-arrestin-1 participates in agonist-dependent desensitization of the LH/CG R, and that the trigger for beta-arrestin-1 binding to the LH/CG R appears to be R activation.

Figure 1

Porcine ovarian follicular membranes contain β-arrestin-1. Membrane proteins (100 μg) boiled in 3× SDS sample buffer were separated on a 10.5% SDS-polyacrylamide gel and transferred to Immobilon, and immunoblot analysis was performed on the blot by using F4C1 mAb to visual arrestin (7) (Left) and an mAb (Transduction Laboratories) generated against rat β-arrestin-1 (Right). Molecular weight markers are indicated at the left. Results for each blot are representative of at least duplicate experiments.

Figure 2

Effect of pretreatment of porcine follicular membranes with neutralizing arrestin antibodies on desensitization of hCG-stimulated AC activity. (A) Membranes (≈30 μg) were subjected to a preincubation (15 min at room temperature; 1 h at 4°C) with indicated final concentrations (in 50 μl reaction volume) of arrestin antibody A9C6 or NIS, followed by the two-stage desensitization reaction at 30°C, which consisted of a 40-min stage 1 desensitization incubation followed by a 5-min AC assay. Figure shows basal (BSA/BSA), full hCG-stimulated (BSA/hCG), and hormone-desensitized hCG-sensitive AC activity (hCG/hCG). Results are means ± SEM of quadruplicate determinations of a single assay and are representative of four separate experiments. Percentage of desensitization (% D) represents the reduction in hCG-stimulated AC activity when incubations contained hCG in stage 1 compared with activity measured with BSA in stage 1 over basal (BSA/BSA) AC activities. (B) Composite effect of different concentrations of arrestin antibody A9C6 or NIS at indicated final concentrations on percentage of desensitization of hCG-stimulated AC activity. Results are means ± SEM of four separate experiments. Remaining details are as in A. ∗, Differences between percentage desensitization values with arrestin antibody and NIS are significantly differently (P < 0.05). (C) Membranes were preincubated with indicated final concentrations of NIS, A9C6, F4C1, or both A9C6 and F4C1, as in A. Results are means ± SEM of quadruplicate determinations in a single assay.

Figure 3

Effect of addition of a synthetic peptide corresponding to arrestin antibody-binding site on hCG-stimulated desensitization of porcine follicular membrane AC activity. (A) Membranes (≈30 μg) were preincubated with indicated amounts of synthetic β-arrestin-1 peptide or water and then subjected to the two-stage AC reaction. Results are mean ± SEM of quadruplicate determinations of a single experiment and are representative of two experiments. Full hCG-stimulated (solid bars) and desensitized hCG-stimulated (hatched bars) AC activities are significantly different (P < 0.05) for all treatments. (B) Membranes were preincubated with NIS or arrestin antibody A9C6 (1:50 dilution in final 50-μl reaction volume), 6.5 μM synthetic β-arrestin-1 peptide, or A9C6 or NIS (1:50 dilution in final 50-μl reaction volume) plus 6.5 μM synthetic arrestin peptide, as indicated, followed by the two-stage desensitization reaction. Results are means ± SEM of quadruplicate determinations of a single experiment and are representative of five experiments. Full hCG-stimulated AC activities (solid bars) for each treatment are not significantly different (P > 0.05). Desensitized hCG-stimulated AC activities (hatched bars) are significantly different (P < 0.05) from full hCG-stimulated AC activities (solid bars) for all treatments except for preincubation with arrestin antibody, where full and desensitized hCG-stimulated AC activities are not significantly different (P > 0.05). (C) Composite effect of synthetic β-arrestin-1 peptide to negate effect of arrestin antibody on desensitization of hCG-stimulated AC activity. Results are means ± SEM of five separate experiments. ∗, Significantly different from other values (P < 0.05).

Figure 4

Effect of addition of arrestins on desensitization of hCG-stimulated AC activity. (A) Membranes (≈30 μg protein) were preincubated (15 min at room temperature; 1 h at 4°C) with water or 0.4 μM visual arrestin that had been boiled (+boil) or not been boiled (−boil), as detailed in the legend to Fig. 2, then subjected to the two-stage AC desensitization incubation. Results are means ± SEM of quadruplicate determinations and are representative of two separate experiments. Desensitized hCG-sensitive AC activity (hatched bars) is significantly different (P < 0.05) from full hCG-stimulated AC activity (solid bars) in membranes incubated with water and boiled arrestin; these values are not significantly different (P > 0.05) for membranes incubated with arrestin that had not been boiled. (B) Membranes (≈30 μg) were preincubated, as in A, with indicated concentrations of purified visual arrestin or rec β-arrestins or water, and then subjected to two-stage desensitization incubation. Solid squares represent BSA in stage 1 and hCG in stage 2. Open bars represent BSA in stages 1 and 2 for water controls (i.e., basal AC activity). Desensitization incubation contained 50 μM AMP-PNP, and AC assay contained 1 mM AMP-PNP. Results are means ± SEM of quadruplicate determinations and are representative of two independent experiments. Dotted lines reflect ED₅₀ values.

Figure 5

Immunoprecipitation of β-arrestin-1. Follicular membranes (800 μg protein) were subjected to stage 1 desensitization incubation in the presence of BSA or hCG, as indicated (in the absence of [³H]cAMP); membrane proteins were solubilized, β-arrestin-1 was immunoprecipitated with β-arrestin-1 antibody, and immunoprecipitated proteins were separated by SDS/PAGE and transferred to Immobilon, then probed with glycoprotein hormone receptor-specific antibody mAb A7. Molecular weight markers are on the right. Molecular weight of indicated band was calculated by linear regression of the migration position of protein standards (30–97 kDa). Results are representative of three separate experiments.