Gravin-mediated formation of signaling complexes in beta 2-adrenergic receptor desensitization and resensitization.

Agonist-induced desensitization and resensitization of G-protein-linked receptors involve the interaction of receptors with protein kinases, phosphatases, beta-arrestin, and clathrin organized by at least one scaffold protein. The dynamic composition of the signaling complexes and the role of the scaffold protein AKAP250 (gravin) in agonist-induced attenuation and recovery of beta-adrenergic receptors were explored by co-immunoprecipitation of target elements, antisense suppression, and confocal microscopy. Gravin associated with unstimulated receptor, and the association was increased significantly after agonist stimulation for up to 60 min. Agonist stimulation also induced a robust association of the receptor-gravin complex with protein kinases A and C, G-protein-linked receptor kinase-2, beta-arrestin, and clathrin. Confocal microscopy of the green fluorescence protein-tagged beta(2)-adrenergic receptor showed that the receptor underwent sequestration after agonist stimulation. Suppression of gravin expression via antisense oligodeoxynucleotides disrupted agonist-induced association of the receptor with G-protein-linked receptor kinase-2, beta-arrestin, and clathrin as well as receptor recovery from desensitization. Gravin deficiency also inhibited agonist-induced sequestration. These data reveal that gravin-mediated formation of signaling complexes with protein kinases/phosphatases, beta-arrestin, and clathrin is essential in agonist-induced internalization and resensitization of G-protein-linked receptors.

Agonist-induced desensitization and sequestration of GPLRs are a complex process involving phosphorylation of the receptor by various protein kinases, followed by the interaction of the receptor with other proteins such as ␤-arrestin and clathrin (6,7,9,10). Recent studies have revealed that protein kinase-anchoring and scaffold proteins are essential elements in many aspects of cell signaling (11)(12)(13)(14). Anchoring proteins bind to subcellular structures and localize their complement of enzymes/adaptor proteins close to their site of action. For example, protein kinase A is targeted to its substrate by association with A kinase-anchoring proteins (AKAPs) (13). Protein kinase C is tethered to the cytoskeleton or at submembranous sites through association with a family of substrate binding proteins called receptors for activated protein kinase C (12). Scaffold proteins (e.g. Ste5p and AKAP79) simultaneously associate with several kinases/phosphatases or other components of a signaling pathway, forming an ordered module that permits sequential activation of each enzyme and the recruitment of other components (13). AKAP250 (gravin) is a unique AKAP protein that displays protein kinase/phosphatase-binding motifs, associates with protein kinases A and C and phosphatase 2B (15), and serves as a scaffold protein in signal transduction. Gravin is expressed in endothelial cells and several other adherent cell types in vitro (16), and the expression of gravin can be induced in human erythroleukemia cells by phorbol ester (15). Immunolocalization experiments show that gravin is concentrated at the cell periphery and is enriched in filopodia of erythroleukemia cells (16). Recently, we reported that gravin is expressed in human epidermoid carcinoma (as well as hamster smooth muscle cells) and is found in association with ␤ 2 -adrenergic receptors (17). Inhibition of gravin expression by antisense oligodeoxynucleotides disrupts receptor resensitization (17). In the current work, we studied the role of gravin in the formation of signaling complexes involved in agonist-induced GPLR desensitization and resensitization. Gravin was found to be essential to the organization of the signaling complexes composed of protein kinases/phosphatases, ␤-arrestin, and clathrin. Blocking the expression of gravin by antisense oligodeoxynucleotides inhibited the association of the ␤ 2 -adrenergic receptor with GRK2, ␤-arrestin, and clathrin. Gravin deficiency blocked the sequestration and resensitization of ␤ 2 -adrenergic receptors.

EXPERIMENTAL PROCEDURES
Cell Culture-Human epidermoid carcinoma cells (A431) were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum (Hyclone Laboratories, Logan, UT), 60 g/ml penicillin, and 100 g/ml streptomycin and grown in a humidified atmosphere of 5% CO 2 and 95% air at 37°C.
Radioligand Binding Studies-The number of ␤ 2 -adrenergic receptors was determined by radioligand binding. Intact A431 cells were incubated with 0.5 nM [ 125 I]iodocyanopindolol (NEN Life Science Products) in the presence or absence of 10 M propranolol at 23°C for 90 min. The incubation buffer contained 50 mM Tris-HCl, pH 7.5, 10 mM MgCl 2 , and 150 mM NaCl (2). The cells were collected on GF/C membranes at reduced pressure and washed rapidly. The radioligand bound to the washed cell mass retained by the filter was quantified by use of a ␥-counter.
Sequestration of ␤ 2 -Adrenergic Receptor-Receptor sequestration was assayed using the hydrophilic, membrane-impermeable ␤-adrenergic antagonist [ 3 H]CGP-12177 (NEN Life Science Products) (17). A431 cells were preincubated with isoproterenol (10 M) for periods up to 60 min or preincubated with isoproterenol for 30 min, followed by washing and a second, post-wash incubation for 60 min. The cells were then resuspended in Dulbecco's modified Eagle's medium containing 20 mM HEPES, pH 7.4, and 70 nM [ 3 H]CGP-12177 at 4°C for 6 h. The cells were diluted, collected on GF/C membranes, and washed rapidly. The radioligand bound to the washed cell mass retained on the filter was counted by liquid scintillation spectrometry. Nonspecific binding is defined as the radioligand binding insensitive to competition by the unlabeled ␤-adrenergic antagonist propranolol (10 M).
Desensitization and Resensitization of ␤ 2 -Adrenergic Receptor-Two days prior to the analysis, cells were seeded in 96-well plates at a density of 20,000 cells/well. Cells were washed and challenged with or without 10 M isoproterenol for periods up to 60 min at 37°C. At the end of the first challenge, cells were washed three times and incubated in 20 mM HEPES buffer, pH 7.4, containing Ro 20-1724 (0.1 mM; Calbiochem) and adenosine deaminase (0.5 unit/ml) for 5 min prior to (and included in) the second challenge with 1 M agonist. For resensitization, isoproterenol-treated cells were washed free of agonist and maintained in buffer without phosphodiesterase inhibitor for 60 min after the first challenge. Five minutes before the second challenge of the agonist, cells were incubated again in the presence of Ro 20-1724 and adenosine deaminase. The agonist-induced cyclic AMP production within 5 min of incubation in the second challenge was determined as described (2,18). The data are calculated as percent desensitization, where "100% desensitization" reflects no cyclic AMP accumulation in response to a second challenge with the agonist. "0% desensitization" indicates that a cyclic AMP response to a second challenge is equivalent to that obtained in response to the first challenge.
Suppression via Antisense Oligodeoxynucleotides-Antisense and control missense oligodeoxynucleotides with the same base composition, but in scrambled order, were synthesized and purified to cell culture-grade (Operon Technologies, Inc., Alameda, CA) as described (2). Before addition to cells, oligodeoxynucleotides were mixed at a ratio of 1:3 (w/w) with N- [1-(2,3-dioleoyloxy)propyl]-N,N,N-trimethylammonium methyl sulfate (Roche Molecular Biochemicals), a cationic diacylglycerol in liposomal form that serves as a delivery vehicle. A431 cells were treated with oligodeoxynucleotides (5 g/ml) for at least 48 -72 h prior to the analysis of the expression of the target molecule. Cells in which gravin or protein kinase C was specifically suppressed by antisense oligodeoxynucleotides were then analyzed for the association of the ␤ 2 -adrenergic receptor with arrestin, GRK2, and clathrin as well as for receptor sequestration as determined by binding of the hydrophilic, cell-impermeable ligand CGP-12177 as described (2) and by confocal microscopy of the autofluorescent emission of the green fluorescence protein (GFP)-tagged form of the ␤ 2 -adrenergic receptor.
For studies of ␤-arrestin, A431 cells were scanned first for expression of ␤-arrestin isoforms, and only ␤-arrestin isoform-2 was detected (data not shown). Suppression of ␤-arrestin required the use of S-modified oligodeoxynucleotides antisense to ␤-arrestin-2 (5Ј-CCCATAGGTGCG-GCGCCC-3Ј), using a sequence in the 5Ј-untranslated region of the gene as a target. The missense oligodeoxynucleotides of the same base composition were S-modified (5Ј-ACTCCGATGCGGGGCCCC-3Ј) and employed as indicated above. Antibodies used in studies of ␤-arrestin were a generous gift of Dr. J. L. Benovic (Kimmel Cancer Institute, Thomas Jefferson University, Philadelphia, PA).
Immunoblot Analysis-Cells were harvested and homogenized in 10 mM HEPES buffer, pH 7.4, containing 2 mM MgCl 2 , 2 mM EDTA, 10 g/ml leupeptin, 10 g/ml aprotinin, and 0.1 mM phenylmethylsulfonyl fluoride. Nuclei were removed by low speed centrifugation. Twenty or fifty micrograms of post-nuclear fraction protein/lane was subjected to 10% SDS-polyacrylamide gel electrophoresis. Separated proteins were transferred onto a nitrocellulose membrane, blocked with 10% (w/v) bovine serum albumin, and incubated with primary antibodies specific for a signal element. After washing, the blots were incubated with a second anti-mouse/rabbit IgG antibody labeled with peroxidase, and the blots were developed by the enhanced chemiluminescence method (NEN Life Science Products). Polyclonal anti-GRK2 and anti-clathrin antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Polyclonal anti-␤-arrestin antibody was a generous gift from Dr. J. L. Benovic, and polyclonal anti-gravin antibody was a generous gift from Dr. J. D. Scott (Howard Hughes Medical Institute, Vollum Institute, Portland, OR). Quantification of immunoblotting data was performed using a Bio-Rad imaging densitometer fitted with Discovery imaging software (Bio-Rad).
Immunoprecipitation-The association of the ␤ 2 -adrenergic receptor with other target signaling proteins was probed by analysis of immunoprecipitations. Cells were harvested and subjected to lysis buffer (1% Triton X-100, 0.5% Nonidet-40, 10 M dithiothreitol, 5 g/ml aprotinin, 5 g/ml leupeptin, 100 g/ml bacitracin, 100 g/ml benzamidine, 2 mM sodium orthovanadate, 150 mM NaCl, 5 mM EDTA, 50 mM NaF, 40 mM sodium pyrophosphate, 50 mM KH 2 PO 4 , 10 mM sodium molybdate, and 20 mM Tris-HCl, pH 7.4). Immunoprecipitation reactions were performed with antibodies in lysis buffer. The lysates were precleared with protein A/G-agarose for 90 min and then subjected to immunoprecipitation for 2 h with antibodies specific either for the ␤ 2 -adrenergic receptor (CM-4) or for AKAP250 (gravin). The primary antibodies were linked covalently to a protein A/G-agarose matrix. The variance of immunoprecipitation, sample loading, and/or immunoblotting in these experiments in the aggregate was established within 10%.
Protein Kinase A and C Assays-Protein kinase A and C activities were assessed using commercially available assays (Life Technologies, Inc.) following the manufacturer's protocol. Protein kinase A activity is defined as the amount of phosphate incorporated into a substrate peptide (Kemptide) in the presence of 10 M cyclic AMP minus any activity incorporated in the presence of the protein kinase A inhibitor peptide (1 M). The specific activity of protein kinase C is defined as the difference between the amount of phosphorylation of an acetylated peptide derived from myelin basic protein in the presence of 10 nM phorbol 12myristate 13-acetate and the amount of phosphorylation occurring in the presence of 20 M protein kinase C inhibitory peptide (protein kinase C␣ peptide- (19 -36)).
Protein Phosphatase 2B Assay-For protein phosphatase 2B, two complementary techniques were employed for assay: immunostaining with antibodies to the catalytic subunit of protein phosphatase 2B (C26920, Transduction Laboratories, Lexington, KY) and identification by the calmodulin overlay assay. Calmodulin binding to the renatured SDS-polyacrylamide gel (overlay) is detected using antibodies to calmodulin. The immunocomplexes were then made visible as described above. Both assays provide comparable results.
Confocal Microscopy-A431 cells transfected with the GFP-tagged ␤ 2 -adrenergic receptor or GFP-tagged gravin were treated with or without gravin or protein kinase C antisense oligodeoxynucleotides for 48 h. The cells were incubated with 10 M isoproterenol for 30 min, fixed with 3% paraformaldehyde, and washed three times with MSM/PIPES buffer (18 mM MgSO 4 , 5 mM CaCl 2 , 40 mM KCl, 24 mM NaCl, and 5 mM PIPES, pH 6.8). The cells were analyzed by confocal microscopy on an Odyssey instrument (Noran Instruments, Inc.). The construct pcDNA3-␤ 2 AR-GFP, encoding a ␤ 2 -adrenergic receptor fusion protein with GFP at its carboxyl terminus (7), was a generous gift from Dr. J. L. Benovic. The construct pEGFP-N1-gravin, encoding a gravin fusion protein with GFP at its carboxyl terminus, was a generous gift from Dr. J. D. Scott. The confocal microscopy was performed at the University Microscopy Imaging Center at Stony Brook.
Data Presentation and Analysis-The values presented are means Ϯ S.E. The autoradiograms are representative of multiple (at least three) independent experiments. In all figures, an asterisk denotes a mean value with statistical significance (p Յ 0.05) compared with the mean values of the control (time 0) or with the control group, as indicated in the figure legends.

Agonist-induced Cellular
Responses-Human epidermoid carcinoma A431 cells were employed, as this cell line has been widely used to study agonist-induced desensitization of G-protein-coupled receptors (2,17,19,20). Stimulation of A431 cells with ␤-adrenergic agonist resulted in desensitization of ␤-adrenergic receptors in response to a second stimulation (Fig. 1A). Desensitization was characterized by a reduction in the cyclic AMP response of these cells, which were challenged previously with agonist (10 M isoproterenol) for periods up to 60 min. The ␤-adrenergic receptors gradually desensitized following agonist stimulation, displaying a Ͼ50% decline in their cyclic AMP response when pre-challenged with agonist for 60 min. After washout of agonist for 60 min (w60), the functional activity of the receptors recovered to nearly control levels (Fig. 1A). In agreement with earlier observations (17), agonist stimulation catalyzed the increased association of the receptor with protein kinase A (ϳ2-fold) and protein kinase C (ϳ5-fold) (Fig. 1B), as determined in immunoprecipitates using an antibody specific for the ␤ 2 -adrenergic receptor combined with enzyme activity assays. The association of protein phosphatase 2B activity with the receptor displayed a transient decrease at 5 min, followed by a 2-fold increase in receptor association by 60 min after challenge with agonist ( Fig. 1C). GPLRs typically undergo sequestration following receptor activation and phosphorylation (5)(6)(7)21). We tested this in binding assays using the hydrophilic, cell-impermeable ␤-adrenergic antagonist ligand CGP-12177. Our data show that the number of receptors accessible to the cell-impermeable ␤-antagonist ligand CGP-12177 decreased gradually, displaying a Ͼ30% decline within 60 min of stimulation. Because CGP-12177 is hydrophilic, it can bind to these receptors only on the cell surface. The CGP-12177 binding data demonstrate that agonist stimulation catalyzes sequestration of this prototypic GPLR.
Complex Formation of ␤ 2 -Adrenergic Receptor with GRK2, ␤-Arrestin, Clathrin, and Gravin-After agonist stimulation, GPLRs undergo sequestration involving the interaction of the receptor with protein kinases/phosphatases, ␤-arrestin, and clathrin. Complex formation of the receptor with ␤-arrestin, GRK2, clathrin, and gravin in A431 cells was studied by immunoprecipitation via an antibody against the ␤ 2 -adrenergic receptor, followed by immunoblot analysis designed to identify target interacting proteins. The ␤ 2 -adrenergic receptor is a substrate for phosphorylation by GRK2 (22). The phosphorylation of the receptor catalyzed by GRK2 is important for receptor desensitization and sequestration in A431 cells (2) and other cells (23). Our results show that GRK2 displayed a robust, agonist-induced association with the ␤ 2 -adrenergic receptor, reaching a maximal level (ϳ3-fold) 30 min post-stimulation ( Fig. 2A), whereas its association with the receptor was nearly undetectable in unstimulated cells. After phosphorylation by protein kinases, the ␤ 2 -adrenergic receptor binds to ␤-arrestin and undergoes clathrin-dependent endocytosis (9 -11, 24). Consistent with other observations, ␤-arrestin was found to be associated with ␤-adrenergic receptors in unstimulated cells. Agonist stimulation gradually provoked an increase in the association of ␤-arrestin with the receptor by 4-fold after 60 min of stimulation ( Fig. 2A). In the case of clathrin, isoproterenol stimulation induced a relatively rapid increase in the level of association with the receptor (ϳ3-fold), although the association of clathrin with ␤-adrenergic receptors was not observed in the unstimulated cells. This agonist-induced complex formation of the ␤ 2 -adrenergic receptor with GRK2, ␤-arrestin, and clathrin deduced from several immunoblot analyses is displayed as a time course (Fig. 2B). The order of associations of target proteins with the ␤ 2 -adrenergic receptor following agonist challenge is GRK2, followed by clathrin and ␤-arrestin, suggesting that relatively small changes in ␤-arrestin binding may catalyze enhanced clathrin-mediated endocytosis (9 -11, 22, 24).
Our previous studies have shown that the scaffold protein AKAP250 (gravin) associates with the ␤ 2 -adrenergic receptor in unstimulated cells (17). To study the role of agonist stimulation in the association of gravin with the ␤ 2 -adrenergic receptor, cells were stimulated with isoproterenol for periods up to 60 min. The association of gravin with the receptor was determined first by immunoprecipitation with an antibody against the ␤ 2 -adrenergic receptor (or anti-gravin), followed then by immunoblot analysis of the immunoprecipitate to detect the relevant associated partner, i.e. gravin (or the ␤ 2adrenergic receptor). Agonist stimulation increased the association of gravin with the ␤ 2 -adrenergic receptor, measured using immunoprecipitation with antibodies to either partner (Fig. 3A). The association of the receptor and AKAP250 displayed a somewhat greater signal when immunoprecipitation was carried out with an antibody against gravin, followed by immunoblotting with antibodies to detect ␤ 2 -adrenergic recep-FIG. 1. Agonist stimulation regulates the cyclic AMP accumulation, sequestration, and complex formation of ␤-adrenergic receptors with protein kinases/phosphatases. A431 cells were treated with isoproterenol (10 M) for periods up to 60 or 30 min, followed by washing and incubating for 60 min (w60). Agonist-induced receptor desensitization was determined as described under "Experimental Procedures" (A). The association of the ␤ 2 -adrenergic receptor with protein kinases A (PKA) and C (PKC) (B) and phosphatase 2B (PP2B) (C) was determined by immunoprecipitation with antibodies against the ␤ 2 -adrenergic receptor (CM-4) that were covalently conjugated to protein A/G-agarose beads. The receptor-associated kinase/ phosphatase activities were measured as described under "Experimental Procedures." Receptor sequestration was assayed using the tritiated ␤-adrenergic receptor ligand CGP-12177, which is hydrophilic and measures the availability of exofacial ␤-adrenergic receptor for binding to intact cells (D). The values presented are the means Ϯ S.E. from at least three separate experiments.
tors. Following challenge of the cells with agonist for 60 min, receptor-gravin association increased ϳ4-fold compared with that measured in the unstimulated cells (Fig. 3B). The time course for agonist-induced association of gravin with the ␤ 2adrenergic receptor was rapid, reaching peak levels within 10 min. AKAP250 was found in complex with the ␤ 2 -adrenergic receptor in extracts prepared from cells that were unstimulated, arguing in favor of a significant pool of the receptorgravin complex in the unstimulated state.
Distribution of ␤ 2 -Adrenergic Receptor and Gravin-Earlier results showed that the ␤ 2 -adrenergic receptor does form complexes with gravin (17). The ␤ 2 -adrenergic receptor is known to undergo sequestration via clathrin-dependent endocytosis after agonist stimulation (5)(6)(7)21). To study the role of agonist challenge in the distribution of the ␤ 2 -adrenergic receptor and gravin, fusion proteins with the green autofluorescent protein (GFP-tagged) of the ␤ 2 -adrenergic receptor (7) and gravin were expressed stably in individual clones of A431 cells. As noted previously (7), the distribution of the ␤ 2 -adrenergic receptor and gravin was analyzed by confocal microscopy. The ␤ 2 -adre-nergic receptor was localized predominately on the cell surface in unstimulated cells (Fig. 4A, Top Plane). At the nuclear plane, GFP-␤ 2 -adrenergic receptor fusion proteins were dense in the plasma membrane, with some found in the perinuclear regions (Fig. 4B). After stimulation with isoproterenol for 30 min, in contrast, ␤ 2 -adrenergic receptors migrated from the plasma membrane (Top Plane) into the cells (Nuclear Plane), mainly localized around the nuclei (Fig. 4, C and D). In sharp contrast to the plasma membrane localization of the ␤ 2 -adrenergic receptor, gravin was found evenly distributed throughout the unstimulated cells (Fig. 4, E and F). Isoproterenol stimulation of the cells produced no apparent change in the distribution of gravin, i.e. gravin remained uniformly distributed in the cells (Fig. 4, G and H). These results confirm the earlier observations showing agonist-induced sequestration of ␤ 2 -adrenergic receptors while demonstrating further that the cellular distribution of gravin in cells is rather uniform and unaffected by agonist stimulation. The abundance of gravin may be well in excess of that of ␤ 2 -adrenergic receptors and/or the cellular complement of G-protein-coupled receptors in toto, FIG. 2. Complex formation of ␤-adrenergic receptors with GRK2, ␤-arrestin, and clathrin is increased in response to agonist stimulation. Cell lysates were prepared from A431 cells treated with isoproterenol (Iso; 10 M) for periods up to 60 or 30 min, followed by washing and additional incubating for 60 min (30w60). Lysates were incubated with antibodies to the ␤ 2 -adrenergic receptor (␤ 2 AR) (CM-4) that were covalently conjugated to protein A/G-agarose beads. The immunocomplexes were subjected to SDS-polyacrylamide gel electrophoresis; transferred to nitrocellulose; and probed with an antibody specific for GRK2, ␤-arrestin, or clathrin (A). The immunocomplexes were made visible by the enhanced chemiluminescence method. The relative densities of the band were determined with a Bio-Rad imaging densitometer (GS-700) and MultiAnalyst densitometer software (Bio-Rad) (B). The data presented are representative of at least four separate determinations, each performed with separate cell lysates. IP, immunoprecipitation; IB, immunoblotting.
providing an explanation for the ability of agonist to drive ␤ 2 -adrenergic receptor-gravin complex formation, although gravin distribution itself does not appear to change in response to agonist.

Suppression of Protein Kinase C Expression Decreases Association of ␤ 2 -Adrenergic Receptor with ␤-Arrestin and Blocks
Agonist-induced Receptor Sequestration-Protein kinase C has been shown to play an important role in the desensitization and resensitization of the ␤ 2 -adrenergic receptor (2,17,25). Previous results demonstrated that stimulation of ␤ 2 -adrenergic receptors increases the association of protein kinase C with the receptor, whereas protein kinase C deficiency disrupts the recovery of the desensitized receptor (2, 25). Further insight into the role of protein kinase C in agonist-induced signaling com-plex formation and receptor sequestration was obtained using antisense oligodeoxynucleotides, immunoprecipitation, and confocal microscopy techniques.
␤-Arrestin associated with the ␤ 2 -adrenergic receptor in unstimulated cells as defined by co-immunoprecipitation reactions ( Figs. 2A and 5A). Consistent with earlier observations, agonist stimulation increased the association of the receptor with ␤-arrestin by ϳ3-fold within 60 min of challenge. Protein kinase C deficiency, achieved by antisense oligodeoxynucleotides (2,25), led to a Ͼ80% decrease in the association of ␤-arrestin with the receptor in both unstimulated and agoniststimulated cells. In view of the observations that protein kinase C deficiency inhibited the association of ␤-arrestin with the receptor and that ␤-arrestin was important in GPLR seques- FIG. 3. Association of gravin with the ␤-adrenergic receptor: agonist stimulation of a dynamic association. A431 cells were treated with isoproterenol (Iso; 10 M) for periods up to 60 min. Whole cell lysates were incubated with antibodies against either the ␤ 2 -adrenergic receptor (␤ 2 AR) (CM-4) or gravin that were covalently conjugated to protein A/G-agarose beads. The immunocomplexes were subjected to SDS-polyacrylamide gel electrophoresis, transferred to nitrocellulose, and probed with an antibody specific for gravin or the ␤-adrenergic receptor (A). The immunocomplexes were made visible by the chemiluminescence method. The relative densities of the bands were determined with a Bio-Rad imaging densitometer (GS-700) and MultiAnalyst densitometer software (B). The data presented are representative of at least three separate determinations performed with separate cell lysates. IP, immunoprecipitation; IB, immunoblotting. tration, we explored if protein kinase C deficiency alters agonist-induced receptor desensitization and resensitization at the level of complex formation.
Protein kinase C-deficient cells displayed a more profound desensitization compared with wild-type cells (Fig. 5B). In addition, protein kinase C deficiency strongly inhibited receptor resensitization, with desensitization persisting 60 min after wash off of agonist (Fig. 5B). Agonist-induced receptor sequestration was explored both by CGP-12177 binding and by confocal microscopy of GFP-tagged ␤-adrenergic receptors. A431 cells showed an ϳ25% decrease in CGP-12177 binding 60 min after agonist stimulation and recovered to the basal level within 60 min after removal of agonist (Fig. 5B). In contrast, protein kinase C-deficient cells showed no decrease in CGP-12177 binding after agonist stimulation (Fig. 5B). Similar results were obtained if cells were treated with the protein kinase C inhibitor bisindolylmaleimide (100 nM) rather than made deficient in protein kinase C by antisense oligodeoxynucleotides (Fig. 5B).
Agonist-induced receptor sequestration was studied also by using cells stably transfected with an expression vector for the GFP-tagged ␤ 2 -adrenergic receptor used in tandem with confocal microscopy. Most GFP-tagged receptors were localized on cell surface in unstimulated cells (Figs. 4A and 5C). When examined at the optical plane of the cells, GFP-tagged ␤ 2adrenergic receptors were observed confined largely to the plasma membrane (Fig. 5C, panel a). Agonist stimulation provoked the movement of receptors to intracellular perinuclear FIG. 5. Suppression of protein kinase C inhibits the association of the ␤-adrenergic receptor with ␤-arrestin and regulates agonist-induced ␤-adrenergic receptor desensitization, resensitization, and sequestration. A431 cells were preincubated with or without antisense oligodeoxynucleotides (4 g/ml) to protein kinase C␣ (PKC) for 48 h. The cells were incubated with 10 M isoproterenol (Iso) for periods up to 60 min, and the association of ␤-arrestin with the ␤ 2 -adrenergic receptor (␤ 2 AR) was analyzed by immunoprecipitation (IP) with an antibody specific for the ␤ 2 -adrenergic receptor and immunoblotting (IB) with an antibody against ␤-arrestin (A). Agonist-induced receptor desensitization (B, panel a) and resensitization (panel b) were determined as described under "Experimental Procedures." Agonist-induced receptor sequestration was analyzed by CGP-12177 binding (B, panel c). The values presented are the means Ϯ S.E. (n ϭ 4). For confocal microscopy analysis of receptor sequestration (C), cells were transfected with the GFP-tagged ␤ 2 -adrenergic receptor and pretreated without (panels a and b) or with (panel c and d) antisense oligodeoxynucleotide to protein kinase C␣ for 48 h. The cells were stimulated with or without isoproterenol (10 M) for 30 min, and confocal microscopy was performed as described under "Experimental Procedures." The results reveal the distribution of ␤ 2 -adrenergic receptors at an optical plane at the mid-point of the nucleus (Nuclear Plane). Note that unlike wild-type cells, in which the receptor migrated from the plasma membrane to the perinuclear region (as indicated by arrows in C, panel b) in response to agonist, cells lacking protein kinase C␣ showed no apparent receptor sequestration in response to ␤ 2 -adrenergic receptor agonist. regions (Fig. 5C, panel b), as noted in Fig. 4. In contrast, protein kinase C-deficient cells showed no receptor sequestration after agonist stimulation. GFP-tagged ␤ 2 -adrenergic receptors remained confined largely to the plasma membrane in the presence of agonist (Fig. 5C, panels c and d). These results confirm the CGP-12177 ligand binding results and suggest that protein kinase C is critical for signaling complexes of GPLRs, as loss of protein kinase C activity by suppression of protein kinase C with antisense oligodeoxynucleotides or protein kinase C inhibitor abolishes several key aspects of receptor biology.

Suppression of Gravin Expression Disrupts Formation of Signaling Complexes and Agonist-induced Sequestration-Gravin
has been shown to associate with protein kinases A and C, protein phosphatase 2B, and the ␤ 2 -adrenergic receptor (Fig. 1) (15,17). Previous data show that agonist stimulation increases the association of AKAP250 with the ␤ 2 -adrenergic receptor, suggesting that gravin may organize multiple signaling components in GPLR complexes. To study the role of gravin in signaling complex formation and agonist-induced sequestration, we used antisense oligodeoxynucleotides to block specifically gravin expression in cells (17) and then studied the gravin-deficient cells by immunoprecipitation, immunoblotting, and confocal microscopy techniques. Treatment of gravin with antisense (but not missense) oligodeoxynucleotides suppresses the expression of gravin by Ͼ80% in these cells (17) .
We explored to what extent loss of gravin would alter the interaction between the ␤ 2 -adrenergic receptor and ␤-arrestin (Fig. 6, A and B). In control cells, agonist stimulation significantly increased the association of the ␤ 2 -adrenergic receptor with ␤-arrestin (Figs. 2 and 6). In gravin-deficient cells, in sharp contrast, agonist-induced association of ␤-arrestin with the receptor was totally abolished (Fig. 6A).
The increase in the association of the ␤ 2 -adrenergic receptor with both GRK2 and clathrin by agonist was also impaired in cells deficient in gravin. In gravin-deficient cells, the association of GRK2 with the receptor in response to agonist was inhibited by ϳ50% (Fig. 6B). Clathrin association with the ␤ 2 -adrenergic receptor in response to agonist was abolished in gravin-deficient cells (Fig. 6C). In the gravin-deficient cells, agonist stimulation provoked a significant loss of receptor association with clathrin (Fig. 6C). Gravin deficiency also blocked agonist-induced GPLR sequestration as determined by both CGP-12177 binding and confocal microscopy of the GFP-tagged ␤ 2 -adrenergic receptor. In the control cells, agonist stimulation decreased CGP-12177 binding in response to challenge with isoproterenol for 60 min, recovering within 60 min following wash-off of the agonist (Fig. 6D). In contrast, agonist-induced sequestration of ␤ 2 -adrenergic receptors was abolished in the gravin-deficient cells. Confocal microscopy of the GFP-tagged ␤ 2 -adrenergic receptor was employed to test the data from CGP-12177 binding. Gravin deficiency blocked agonist-induced sequestration of the ␤ 2 -adrenergic receptors. The perinuclear increase in receptor in response to agonist in control cells was abolished in the gravin-deficient cells (Fig. 6E). Taken together, the results from CGP-12177 binding analysis and confocal microscopy demonstrate that gravin deficiency disrupts the formation of GPLR signaling complexes and blocks agonistinduced GPLR sequestration.
The association of ␤-arrestin with the ␤ 2 -adrenergic receptor is critical for GRK-mediated regulation of receptor signaling (26,27). Gravin-deficient cells were shown to display a sharp decline in the level of association of the ␤ 2 -adrenergic receptor with ␤-arrestin, as shown by immunoprecipitation reactions (Fig. 6A). Based upon these observations, one might hypothesize that agonist-induced desensitization would decline. We tested the hypothesis directly by studying desensitization in cells made deficient in ␤-arrestin by treatment with S-modified antisense oligodeoxynucleotides. Treatment of ␤-arrestin with antisense (but not missense) oligodeoxynucleotides suppressed expression of ␤-arrestin by Ͼ85% under the conditions employed (Fig. 7, inset). Desensitization in response to agonist stimulation was measured in the ␤-arrestin-deficient cells and found to be enhanced in comparison with that observed in control cells and in cells treated with S-modified missense oligodeoxynucleotides (Fig. 7). We interpret these data in light of earlier observations that there exist, in A431 cells, both GRK-dependent and GRK-independent mechanisms of agonistinduced desensitization (2). In the ␤-arrestin-deficient cells, the GRK-independent mechanism of desensitization dominated, and the failure of the receptor to resensitize via recycling (Fig. 6E) led to a more rapid and profound accumulation of desensitized receptors.

DISCUSSION
The attenuation that accompanies chronic stimulation of receptors is a fundamental feature of cell signaling. In the case of GPLRs, agonists induce attenuation of the signaling within minutes of agonist exposure, a process often referred to as desensitization (1,28). Protein phosphorylation is a central feature of agonist-induced desensitization, involving the participation of protein kinase A (29,30), protein kinase C (31), and G-protein-linked receptor kinases (22,23). Agonist stimulation also promotes rapid sequestration of many GPLRs such as the ␤ 2 -adrenergic receptor, m1 muscarinic cholinergic receptor, and luteinizing hormone/human chorionic gonadotropin receptor (5,21,32). These GPLRs are believed to traffic to an intracellular endosomal pool via clathrin-dependent endocytosis (5)(6)(7)21). The sequestration appears obligate for resensitization of the receptor function (2). Agonist-induced sequestration of the ␤ 2 -adrenergic receptor involves the interaction of the receptor with ␤-arrestin and clathrin (6,7,9,10).
In recent studies, protein kinase-anchoring (AKAP) and scaffold proteins have emerged as central elements in many aspects of cell signals (11)(12)(13)(14). Scaffold proteins simultaneously associate with several kinases/phosphatases or other components of a signaling pathway, forming an ordered module that permits sequential activation of each enzyme and that recruits other components (13). Recently, we showed that gravin, a 250-kDa scaffold protein, associates with the ␤ 2 -adrenergic receptor in unstimulated cells (17). Gravin deficiency was found to disrupt receptor resensitization (17), suggesting that this AKAP250 molecule plays an important role in GPLR desensitization and resensitization.
In the current study, we present several lines of evidence demonstrating the formation of macromolecular complexes in the GPLR signaling pathway and highlighting the central role of gravin in these processes. The hypothesis is that agonist stimulation enhances the association of the receptor with gravin, which brings protein kinases/phosphatases into close association with the receptor and which recruits other components to the complexes. Indeed, agonist stimulation significantly increases the association of gravin with the ␤ 2 -adrenergic receptor, accompanied by increased association of the receptor with protein kinases A and C, GRK2, and protein phosphatase 2B (17), as has been discovered recently in other signaling cascade (13). Gravin is shown to associate with protein kinases A and C in unstimulated A431 cells. ␤ 2 -Adrenergic receptors are a substrate for both of these enzymes and GRK2 (2,17,25). Gravin catalyzes the formation of ␤ 2 -adrenergic receptor complexes with protein kinase A and GRKs, facilitating signaling attenuation upon activation. Following phosphorylation, GPLRs undergo sequestration mediated by ␤-arrestin (10,18). Arrestins are a family of proteins including visual arrestin and ␤-arrestin-1 and -2 (10). When ␤-arrestin is overexpressed in cells, not only is desensitization of ␤ 2 -adrenergic receptors augmented, but receptor sequestration is promoted as well (33)(34)(35). Expression of a "dominant-negative" mutant of ␤-arrestin (V53D) impairs receptor desensitization and sequestration (35). Our results show that agonist stimulation induces a rapid association of a ␤ 2 -adrenergic receptor-gravin complex with ␤-arrestin. This association with ␤-arrestin is obligate for clathrin-dependent endocytosis (5)(6)(7)21). Indeed, a rapid association between clathrin and the receptor-gravin signaling complex was provoked by agonist stimulation. These results demonstrate that agonist stimulation induces the formation of gravin signaling complexes involving the receptor, protein kinases/phosphatases, ␤-arrestin, and clathrin in GPLR signaling pathways.
Antisense oligodeoxynucleotides that specifically inhibit the expression of gravin were found to block agonist-induced association of the ␤ 2 -adrenergic receptor with GRK2, ␤-arrestin, and clathrin. The data suggest that complex formation of the receptor with these key proteins is orchestrated in some manner by gravin. Earlier results have shown that the association of protein kinase A with the ␤ 2 -adrenergic receptor is mediated by gravin (17). The HT31 peptide, which specifically inhibits the binding of protein kinase A to AKAPs (14,36,37), inhibits the binding of protein kinase A to the ␤ 2 -adrenergic receptor (17). Gravin binds ␤ 2 -adrenergic receptors, bringing protein kinases/phosphatases into close proximity to the receptor and recruiting ␤-arrestin to the signaling complexes to target the receptor for sequestration via clathrin-dependent endocytosis.
A unique role of protein kinase C in agonist-induced desensitization and resensitization is revealed in the current study. Protein kinase C has been shown to modulate the function of several G-protein-linked receptors, and consensus sites for phosphorylation have been identified in many of these receptors. Although a few early studies suggested that activation of protein kinase C desensitizes the ␤ 2 -adrenergic receptors (38,39), most subsequent studies reported either no effect of protein kinase C activation on desensitization (40,41) or frank potentiation of ␤ 2 -adrenergic receptor signaling rather than desensitization (42)(43)(44)(45)(46)(47)(48). Protein kinase C is associated in complex with the receptor-gravin complex in unstimulated cells. The formation of protein kinase C-gravin complexes is enhanced by agonist challenge.
In our system, suppression of protein kinase C by antisense oligodeoxynucleotides amplified rather than attenuated agonist-induced desensitization. Resensitization was blocked in protein kinase C-deficient cells. Protein kinase C deficiency also inhibited the association of ␤-arrestin with receptor-gravin complexes and agonist-induced sequestration as determined by immunoprecipitation, CGP-12177 binding, and confocal microscopy. One possible explanation is that protein kinase C modulates the binding of ␤-arrestin to the signaling complexes via phosphorylation. Indeed, it has been reported that GRK2 activity is enhanced in cells following protein kinase C activation (49). In vitro experiments with purified proteins showed that protein kinase C could directly phosphorylate GRK2 and increase its activity (49). Further studies demonstrated that activation of protein kinase C with phorbol esters caused not only an activation of cytosolic GRK2, but also a translocation of GRK2 from the cytosol to the membrane fraction (50). GRK2 phosphorylates the C-terminal cytoplasmic regions of GPLRs, permitting ␤-arrestin to bind and target the receptors for sequestration and resensitization via clathrin-dependent endocytosis (51). Protein kinase C deficiency has been shown to block the association of GRK2 with ␤-adrenergic receptors. The results suggest that protein kinase C deficiency impairs GRK2 association with the receptor-gravin complex, which impairs the binding of ␤-arrestin. Reduced binding of ␤-arrestin impairs receptor sequestration and recovery. The result of protein kinase C deficiency is accumulation of "desensitized" receptor in the cells, largely on the cell surface. Indeed, the protein kinase C inhibitor bisindolylmaleimide displays similar effects on ␤-arrestin binding and receptor sequestration.
The importance of gravin-mediated formation of GPLR complexes in receptor sequestration was revealed by suppression of gravin with antisense oligodeoxynucleotides combined with CGP-12177 binding and confocal microscopy. Gravin deficiency completely blocks agonist-induced receptor sequestration as established both by CGP-12177 binding and independently by confocal microscopy of GFP-tagged receptors. These results demonstrate that gravin-mediated signaling complex formation is essential for agonist-induced receptor sequestration. In our earlier report (17), we showed that gravin deficiency had no effect on ␤ 2 -adrenergic receptor desensitization, but strongly inhibited receptor resensitization. Gravin deficiency provoked a loss in the association of ␤-arrestin with the receptor, sug-FIG. 6. Suppression of gravin expression inhibits complex formation of the receptor with ␤-arrestin and regulates agonistinduced ␤-adrenergic receptor desensitization, resensitization, and sequestration. A431 cells were pretreated with or without oligode oxynucleotides antisense to gravin for 48 h (5 g/ml). The cells were treated with or without isoproterenol (Iso; 10 M) for periods up to 60 min. The association of the ␤-adrenergic receptor with ␤-arrestin, GRK2, and clathrin was analyzed by immunoprecipitation (IP) using an antibody against the ␤ 2 -adrenergic receptor (␤ 2 AR) and subsequent immunoblotting (IB) using antibodies specifically against arrestin (A), GRK2 (B), and clathrin (C). The relative densities of the corresponding bands were determined with a Bio-Rad imaging densitometer (GS-700) and MultiAnalyst densitometer software. Agonist-induced receptor sequestration was analyzed by CGP-12177 binding (D) or by confocal microscopy of the GFP-tagged ␤ 2 -adrenergic receptor (E) as described under "Experimental Procedures. "   FIG. 7. Suppression of ␤-arrestin expression promotes more rapid and robust agonist-induced ␤-adrenergic receptor desensitization. A431 cells were pretreated with or without S-modified oligodeoxynucleotides antisense and missense to ␤-arrestin for 72 h (5 g/ml). Levels of ␤-arrestin were suppressed Ͼ85% as measured by immunoblotting with the KEM anti-␤-arrestin-2 antibody. The cells were treated with or without isoproterenol (10 M) for periods up to 60 min. Agonist-induced receptor desensitization was performed as described under "Experimental Procedures." The data shown are the means Ϯ S.E. from three experiments performed on separate occasions. gesting that gravin plays an important role as a scaffold for these protein-protein interactions of GPLRs. The current results demonstrate that gravin-mediated receptor trafficking to clathrin-dependent endocytosis, a likely pathway for receptor resensitization, is mediated or facilitated by the gravin scaffold protein. If the function of gravin is interrupted, the desensitized receptors cannot undergo resensitization.
If the function of gravin is interrupted, the loss of ␤-arrestinreceptor interactions leads to loss of the ␤-arrestin-dependent recycling of receptors via clathrin-dependent endocytosis. In A431 cells, suppression of ␤-arrestin provoked a more rapid and robust desensitization, reflecting the continued function of the GRK-independent mechanism of desensitization in these cells (2) coupled with loss of recycling of the desensitized receptors (17). Receptors desensitized by the GRK-independent mechanism now accumulate, having been precluded from resensitization and recovery.
These studies provide strong evidence to support the hypothesis that G-protein linked receptors (such as the ␤ 2 -adrenergic receptor that activates adenylyl cyclase) participate in macromolecular complexes that are composed minimally of a GPLR, several protein kinases/phosphatases, ␤-arrestin, and clathrin organized by the scaffold protein gravin. The formed complexes target the receptor for regulation and later sequestration via clathrin-dependent endocytosis. The sequestration acts as a gateway for receptor resensitization. The formation of signaling complexes is important functionally, as disruption of specific interactions among these molecules influences agonistinduced desensitization, sequestration, and resensitization.