Phosphorylation of the type 1A angiotensin II receptor by G protein-coupled receptor kinases and protein kinase C.

The type 1A angiotensin II receptor (AT1A-R), which mediates cardiovascular effects of angiotensin II, has been shown to undergo rapid agonist-induced desensitization. We investigated the potential role of second messenger-activated kinases and G protein-coupled receptor kinases (GRKs) in the regulation of this receptor. In 293 cells transfected with the AT1A-R, a 3-min challenge with angiotensin II engendered a 46% decrease in subsequent angiotensin II-stimulated phosphoinositide hydrolysis in intact cells. This agonist-induced desensitization correlated temporally and dose-dependently with the phosphorylation of the receptor to a stoichiometry of 1 mol of phosphate/mol of receptor, as assessed by immunoprecipitation of receptors from cells metabolically labeled with 32Pi. Agonist-induced receptor phosphorylation was reduced by 40-50% by either overexpression of a dominant negative K220R mutant GRK2 or treatment of the cells with the protein kinase C (PKC) inhibitor staurosporine, in a virtually additive fashion. Cellular overexpression of GRK2K220R not only inhibited agonist-induced AT1A-R phosphorylation, but also prevented receptor desensitization, as assessed by angiotensin II-stimulated GTPase activity in membranes prepared from agonist-treated and control cells. In contrast, PKC inhibition by staurosporine did not affect homologous desensitization of the AT1A-R. Overexpression of GRKs 2, 3, or 5 significantly augmented the agonist-induced AT1A-R phosphorylation 1.5- to 1.7-fold (p < 0.001). These findings suggest a role for receptor phosphorylation by one or several GRKs in the rapid agonist-induced desensitization of the AT1A-R.

The type 1A angiotensin II receptor (AT 1A -R), which mediates cardiovascular effects of angiotensin II, has been shown to undergo rapid agonist-induced desensitization. We investigated the potential role of second messenger-activated kinases and G protein-coupled receptor kinases (GRKs) in the regulation of this receptor. In 293 cells transfected with the AT 1A -R, a 3-min challenge with angiotensin II engendered a 46% decrease in subsequent angiotensin II-stimulated phosphoinositide hydrolysis in intact cells. This agonist-induced desensitization correlated temporally and dose-dependently with the phosphorylation of the receptor to a stoichiometry of 1 mol of phosphate/mol of receptor, as assessed by immunoprecipitation of receptors from cells metabolically labeled with 32 P i . Agonist-induced receptor phosphorylation was reduced by 40 -50% by either overexpression of a dominant negative K220R mutant GRK2 or treatment of the cells with the protein kinase C (PKC) inhibitor staurosporine, in a virtually additive fashion. Cellular overexpression of GRK2 K220R not only inhibited agonist-induced AT 1A -R phosphorylation, but also prevented receptor desensitization, as assessed by angiotensin II-stimulated GTPase activity in membranes prepared from agonist-treated and control cells. In contrast, PKC inhibition by staurosporine did not affect homologous desensitization of the AT 1A -R. Overexpression of GRKs 2, 3, or 5 significantly augmented the agonist-induced AT 1A -R phosphorylation 1.5-to 1.7-fold (p < 0.001). These findings suggest a role for receptor phosphorylation by one or several GRKs in the rapid agonistinduced desensitization of the AT 1A -R.
Agonist binding to a G protein-coupled receptor activates catalytic cascades of intracellular mediators which greatly amplify the response to an extracellular stimulus. In many instances, this same event triggers counter-regulatory pathways which attenuate receptor signaling (1). Several distinct, yet highly coordinated mechanisms act together to bring about receptor desensitization. Rapid agonist-specific or homologous desensitization is due to the functional uncoupling of the heterotrimeric G protein from the receptor. On a molecular level, phosphorylation of agonist-occupied receptors by G protein-coupled receptor kinases (GRKs) 1 or by second messengeractivated kinases have been shown to attenuate receptor interaction with G proteins (2). While these mechanisms underlying receptor desensitization have been studied most extensively for the rhodopsin and ␤ 2 -adrenergic receptor systems, they may also be operative in the regulation of several other G proteincoupled receptors.
This study focuses on the vascular type 1 angiotensin II receptor (AT 1 -R), which has long been known to undergo rapid desensitization (3). Most of the known effects of angiotensin II, the major effector molecule of the renin-angiotensin system, are mediated via this receptor (reviewed in Ref. 4). Two different, highly homologous subtypes (AT 1A -R and AT 1B -R) of this receptor were identified in rat and mouse by cloning experiments. Although functionally and pharmacologically indistinguishable from the AT 1B -R, the AT 1A -R is the predominant form in most tissues including the kidney, heart, liver, and aorta (5). Modulation of the AT 1A -R gene in mice by gene targeting experiments has lent further support to the notion that angiotensin II exerts its hemodynamic effects mainly by the type 1A angiotensin II receptor (6). Two recent studies have provided evidence that rapid desensitization of the AT 1 -R involves G protein uncoupling and receptor phosphorylation. First, short-term desensitization of the AT 1 -R in bovine adrenal glomerulosa cells was found to correspond to a shift in receptor affinity for agonist from high (G protein-coupled) to low (uncoupled) (7). As GTP␥S mimicked this effect, it was concluded that agonist-induced desensitization results from the uncoupling of AT 1 -R from its G protein. Second, the AT 1 -R in rat vascular smooth muscle cells was demonstrated to be phosphorylated in response to agonist stimulation (8). However, neither the kinases which phosphorylate the receptor nor the functional consequences of receptor phosphorylation were directly addressed.
The present study was designed to determine whether G protein-coupled receptor kinases and/or second messenger-activated kinases phosphorylate and desensitize the AT 1A -R in intact cells. To this end, effects of kinase inhibitors on receptor phosphorylation and desensitization were studied. Finally, we asked whether individual GRKs can phosphorylate the AT 1A -R. Three members of the G protein-coupled receptor kinase family, i.e. GRK2 (␤ARK1), GRK3 (␤ARK2), and GRK5, were chosen for this study on the basis of their widespread tissue expression (9). Tissue culture media and fetal bovine serum were from  Life Technologies, Inc.; 293 cells were from the American Type Culture  Collection; protein A-Sepharose CL4B was from Pharmacia Biotech  Inc.; okadaic acid and staurosporine were from Calbiochem; A23187,  angiotensin II, dibutyryl-cAMP, forskolin, 3-isobutyl-1-methylxanthine, phorbol 12-myristate 13-acetate, and protease inhibitors  Plasmids-All recombinant DNA procedures were carried out following standard protocols. A nonapeptide sequence (YPYDVPDYA) recognized by the monoclonal antibody 12CA5 was inserted after the aminoterminal initiator methionine of rat AT 1A -R by the polymerase chain reaction method using Hot Tub polymerase (Amersham) and pKSCa18b as template (10). The 5Ј (sense strand) primer was: 5Ј-CGCGGGGGA-TCCACCATGTACCCATACGACGTCCCAGACTACGCCGCCCTTG-ACTCTTCTGCTGAAGATGGTATC-3Ј. The start codon and 29 bases corresponding to amino acids 2 to 11 are underlined. A point mutation (in bold), previously shown to preserve agonist binding (11), changes Asn 3 to Asp to facilitate immunoprecipitation. The 3Ј-oligonucleotide primer was: 5Ј-CGCGGGGCGGCCGCTTACTCCACCTCAAAACAAG-ACGCA-3Ј, with nucleotides 1327 to 1348 of the native sequence underlined. The resulting oligonucleotide was digested with BamHI and NotI (Promega) and ligated into pcDNA I (Invitrogen). Constructs were verified by dideoxy sequencing. cDNA constructs for bovine GRK2, bovine GRK3, bovine GRK5, and the GRK2 K220R dominant negative mutant have been described (12,13).

Materials-
Cell Culture and Transfection-293 cells were grown at 37°C in a humidified atmosphere of 5% CO 2 , 95% air in minimum essential medium supplemented with 10% fetal bovine serum, penicillin (100 units/ ml), and streptomycin (100 g/ml). The cells were transiently transfected using the calcium phosphate precipitation technique with pcDNA I-AT 1A R (10 g/100-mm dish) together with the empty vector pcDNA I or pcDNA I expression vectors for GRK2, GRK3, GRK5, or a GRK2 dominant negative mutant (GRK2 K220R ) (5 g/dish). The total amount of DNA added per dish was kept constant within each experiment by adding up with empty pcDNA I vector. Sixteen hours after transfection, cells were glycerol-shocked and split into six-well plates at 1-1.5 ϫ 10 6 cells per well. Assays were performed 64 h after transfection. Transfection efficiencies as judged by flow cytometry ranged from 45 to 65%, receptor expression levels as determined by ligand binding assays using [ 125 I-Sar 1 -Ile 8 ]angiotensin II (DuPont NEN) (10) were typically 600 fmol/mg of cellular protein.
Inositol Phosphate Assay-Transfected 293 cells were metabolically labeled for 16 -24 h with myo-[ 3 H]inositol at 2 Ci/ml in culture medium supplemented with 5% fetal bovine serum. Cells were then washed briefly with Dulbecco's PBS, preincubated for 10 min in PBS containing 20 mM LiCl, and then stimulated with agonist for 10 min at 37°C in the presence of 0.2% protease-free bovine serum albumin (Boehringer Mannheim). The reaction was stopped with 0.4 M perchloric acid. Total inositol phosphates were extracted and separated by anion exchange chromatography as described (16).
Immunofluorescence and Flow Cytometry-Immunofluorescence and flow cytometry were used to quantitate AT 1A -R expression and agonistinduced receptor sequestration essentially as described previously (14). Transfected cells were incubated (1 h/4°C) with a 100-fold dilution of the 12CA5 ascites (65 g of IgG/ml) in PBS, 1% gelatin, 20 mM NaN 3 , washed with the same buffer by centrifugation, and labeled (1 h/4°C) with a 1:100 dilution of a goat anti-mouse F(abЈ) 2 fluorescein isothiocyanate conjugate (Jackson ImmunoResearch Laboratories). Cells were analyzed using a Becton-Dickinson flow cytometer by setting quadrants which resulted in less than 5% staining of untransfected control cells. The relative expression of AT 1A -R by cells used in a single assay was determined by multiplying the fraction of cells staining positive for AT 1A -R by the mean fluorescence of positive cells. Receptor internalization was determined by the same procedure after incubation of cells with or without 100 nM angiotensin II for 3 min at 37°C. Incubation was stopped by placing the cells on ice and rapidly washing them once with ice-cold PBS and 50 mM glycine, 150 mM NaCl, pH 3.0, respectively.
Intact Cell Phosphorylation-Three days after transfection, 293 cells in 6-well dishes were washed once with phosphate-free Dulbecco's modified Eagle's medium and incubated at 37°C for 60 to 90 min in the same medium containing 32 P i (150 Ci/ml). Okadaic acid (1 M) was added 10 min prior to cellular stimulation. Following the addition of different stimuli as indicated, cells were washed twice with ice-cold Dulbecco's PBS and scraped in lysis buffer (0.8 ml per well) (50 mM Tris-HCl, pH 8, 150 mM NaCl, 5 mM EDTA, 1% (v/v) Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 10 mM sodium fluoride, 10 mM disodium pyrophosphate plus protease inhibitors). Insoluble material was removed by centrifugation at 200,000 ϫ g for 20 min. After removal of aliquots for protein determinations, equal volumes of cell lysates were precleared (1 h/4°C) with 1% (v/v) protein A-Sepharose in the presence of 0.2% bovine serum albumin. Epitope-tagged AT 1A -R were immunoprecipitated by incubating (2 h/4°C) lysates with 13 g of 12CA5 IgG and 1% protein A-Sepharose. After five washes with ice-cold lysis buffer, immune complexes were desorbed from protein A-Sepharose beads by incubation with SDS sample buffer for 45 to 90 min at 37°C. Samples were loaded onto 10% SDS-polyacrylamide gels with equivalent amounts of receptors in each lane, determined as follows. The receptor expression for each cell line as assessed by flow cytometry was multiplied by the protein content of each immunoprecipitation tube. The resulting products were normalized and used to adjust the volumes of SDS sample buffer loaded onto the gel. Gels were stained with Coomassie Brilliant Blue R250 and dried before autoradiography using Biomax MR films (Kodak) and an intensifying screen at Ϫ80°C for 3 to 7 days. The same gels were subsequently analyzed with a Molecular Dynamics PhosphorImager.
Stoichiometry of Receptor Phosphorylation-The extent of receptor phosphorylation in moles of phosphate per mol of receptor was determined essentially as described (15). In brief, 100-mm dishes of 293 cells transfected with pcDNA I-AT 1A -R were incubated either for 48 h prior to stimulation in normal cell culture medium with 100 Ci [ 35 S]methionine/[ 35 S]cysteine per ml, or for 6 h prior to stimulation in the same medium with 0.5 mCi/ml 32 P i . 32 P i -labeled cells were incubated for 3 min at 37°C with or without 200 nM angiotensin II. Cell lysis and immunoprecipitation of the AT 1A -R were performed as described above. Following autoradiography, the AT 1A -R bands were excised and liquid scintillation counting was performed as described (15). The specific activity of 35 S in total cellular proteins was calculated by solubilization of cells on a duplicate dish with 0.1 N NaOH (15). Untransfected 293 cells were treated in parallel and were used to determine background activity in gel slices which corresponded to the AT 1A -R.
Membrane Preparation and GTPase Assay-293 cells expressing AT 1A -R were plated in 100-mm dishes and were grown to 75% confluency. Okadaic acid (1 M) was added 10 min prior to cellular stimulation. Cells were washed once with warm minimal essential medium and incubated for 3 min at 37°C with or without 400 nM angiotensin II in medium. After one rinse with 50 mM glycine, 150 mM NaCl, pH 3.0 for 30 s, cells were washed with PBS and scraped into ice-cold lysis buffer (10 mM Tris-HCl, pH 7.4, 2 mM EDTA plus protease inhibitors as above). Cells were homogenized with 10 strokes of a Teflon pestle followed by centrifugation at 30,000 ϫ g for 15 min. The resulting pellet was resuspended in 10 mM triethanolamine, pH 7.4, at a protein concentration of 300 g/ml. The angiotensin II-stimulated hydrolysis of [␥-32 P]GTP in 15-g aliquots of these membrane preparations was determined following a modification of the protocol originally developed by Cassel and Selinger (17) as described by Gierschik et al. (18). Reaction rates were linear for 15 min and were routinely stopped at 10 min.
Data Analysis-Experimental procedures were performed at least three times. Results were analyzed for statistical significance by an unpaired Student's t-test. Saturation isotherms were analyzed by computer-assisted curve-fitting (FigP for Windows 2.0, Biosoft). RESULTS We first demonstrated that the AT 1A -R in transiently transfected 293 cells undergoes rapid agonist-induced desensitiza-tion. Pretreatment for 3 min with 100 nM angiotensin II reduced the subsequent maximal angiotensin II-stimulated phosphoinositide hydrolysis by 46% as compared to vehiclepretreated control cells (Fig. 1, top panel). Attenuation of inositol phosphate generation was maximal when cells were pretreated with 10 to 30 nM angiotensin II and half-maximal effects were observed after pretreatment with approximately 2 nM angiotensin II (Fig. 1, Bottom Panel). By varying the time of pre-exposure to the agonist from 30 s to 12 min, we found that angiotensin II rapidly induced receptor desensitization with maximal effects observed by pretreatment for 1 min (data not shown).
Agonist-induced internalization of the AT 1A -R is a well established phenomenon (19). However, receptor internalization does not appear to account for the diminished inositol phosphate response in cells desensitized in these experiments. As determined by flow cytometry using a monoclonal antibody (12CA5) which is specific for the amino-terminal epitope-tag of our AT 1A -R construct, incubation of AT 1A -R-expressing cells with 100 nM angiotensin II for 3 min resulted in the loss of only 15 Ϯ 2% of plasma membrane receptors. In the range of receptor expression used in these experiments, such a small decrease in receptor number, modeled by transfecting cells with varied amounts of plasmid, has no effect on angiotensin IIstimulated phosphoinositide hydrolysis (data not shown).
We then studied whether the AT 1A -R is phosphorylated upon agonist stimulation. 293 Cells expressing the AT 1A -R were equilibrated with 32 P i to label their ATP pool and were stimulated with increasing concentrations of agonist. In a dosedependent fashion, angiotensin II induced within 5 min the phosphorylation of a 50 -90-kDa band which was absent in untransfected 32 P i -labeled control cells (Fig. 2). This protein co-migrated with the photoaffinity-labeled AT 1A -R from transfected 293 cell membranes. Phosphorylation of the AT 1A -R was induced by as little as 0.1 nM angiotensin II and reached a maximum at 100 nM angiotensin II; the EC 50 value was 4.5 nM. By labeling cells to equilibrium with 32 P i and [ 35 S]methionine, the amount of phosphate incorporated in the AT 1A -R was determined to be 1.0 Ϯ 0.4 (mean Ϯ S.D.) mol of phosphate per mol of receptor in cells stimulated with 200 nM angiotensin II. In unstimulated cells, the AT 1A -R contained 0.2 Ϯ 0.02 mol of phosphate per mol of receptor.
To investigate which kinases effect AT 1A -R phosphorylation, we challenged cells with both specific activators and inhibitors of the second messenger-dependent kinases, protein kinase C (PKC) or cAMP-dependent protein kinase (PKA). Whereas the phorbol ester PMA significantly increased AT 1A -R phosphorylation above basal values, neither the PKA activators forskolin or dibutyryl cAMP, nor the calcium ionophore A23187, had an effect on receptor phosphorylation (Fig. 3). The amount of 32 P incorporated into the receptor upon stimulation of PKC was only 40% of that achieved by angiotensin II stimulation. Furthermore, staurosporine, which at this concentration (500 nM) blocks both PKA and PKC activity, inhibited the agonistinduced AT 1A -R phosphorylation by only 42%. These results imply that both PKC, the kinase activated by signaling pathways downstream of the AT 1A -R, and one or several protein kinases which are not affected by staurosporine, participate in the agonist-induced phosphorylation of this receptor.
To test the hypothesis that agonist-induced AT 1A -R phosphorylation is effected by GRKs endogenously expressed in 293 cells, cells were co-transfected with a GRK2 dominant negative mutant (GRK2 K220R ). This mutant GRK retains the ability to bind the agonist-occupied receptor, but is devoid of catalytic activity and therefore acts as a competitive inhibitor of GRK activity. Overexpression of this mutant at an excess over endogenous GRKs has been found to inhibit GRK-mediated receptor phosphorylation (13) and desensitization (20). As shown in Fig. 4, the overexpression of GRK2 K220R reduced receptor phosphorylation consequent to angiotensin II stimulation for 5 min by 50%, as compared to control cells, but did not affect agonist-independent receptor phosphorylation through the PKC pathway. The combined inhibition of PKC by staurosporine and GRKs by GRK2 K220R overexpression additively reduced agonist-induced AT 1A -R phosphorylation, resulting in an inhibition of receptor phosphorylation by 85% as compared to control cells. These findings imply that agonist-induced AT 1A -R phosphorylation is mediated by two different types of kinases: by one or more GRKs, which are directly activated by the agonist-occupied receptor, and by PKC, which is activated by second messengers generated consequent to receptor stimulation.
Since both GRK and PKC activity appear important in agonist-induced AT 1A -R phosphorylation, we sought to characterize the kinetics of kinase activities. Previously, it had been demonstrated that GRK and cAMP-dependent protein kinase activity on the ␤ 2 -adrenergic receptor proceed by distinct kinetics (21,22). The time course of angiotensin II-induced receptor phosphorylation revealed the existence of a biphasic process, characterized by a rapid (t1 ⁄2 Ϸ30 s) and a somewhat slower (t1 ⁄2 Ϸ3 min) time course (Fig. 5). Inhibition of PKC by staurosporine eliminated any agonist-induced receptor phosphorylation occurring after 3 min, but did not affect the rapid phase of AT 1A -R phosphorylation. Conversely, PKC-mediated receptor phosphorylation stimulated by PMA proceeded with a t1 ⁄2 of approximately 3 min. These observations support the notion that rapid agonist-induced AT 1A -R phosphorylation derives principally from GRK activity. Receptor phosphorylation seen with prolonged (over 3 min) agonist exposure, however, may derive significantly from PKC.
If GRKs are involved in the agonist-induced AT 1A -R phosphorylation, overexpression of these kinases in 293 cells might be expected to enhance the angiotensin II-mediated receptor phosphorylation. Receptor immunoprecipitation from labeled 293 cells which overexpressed either GRK2, GRK3, or GRK5 revealed agonist-induced AT 1A -R phosphorylation which was 1.5-1.7-fold greater (p Ͻ 0.001) than that observed in control cells (Fig. 6). These experiments did not show significant differences between individual GRKs in their ability to enhance receptor phosphorylation in response to agonist. The increased basal phosphorylation of the AT 1A -R seen in cells which overexpress GRK5, the kinase which is constitutively associated with the cellular membrane (12), was also observed in studies which investigated the role of GRKs in the phosphorylation of ␤1-adrenergic (13) or ␦-opioid receptors (23). As shown by immunoblotting analysis, the transfection of 293 cells with pcDNA I constructs encoding GRK2, GRK3, or GRK5 resulted in cellular expression of these kinases at levels which were at least 20-fold higher than endogenous levels of GRK2 or GRK5.
To assess the effect of GRK inhibition on agonist-induced receptor desensitization, we exposed 293 cells expressing the AT 1A -R with or without GRK2 K220R to angiotensin II for 3 min, removed the agonist with an acid wash, and prepared membranes for the determination of angiotensin II-stimulated GTPase activity (Fig. 7). Exposure of control cells to medium containing angiotensin II decreased agonist-stimulated membrane GTPase activity by 85% (mean of 3 experiments). In cells overexpressing GRK2, pretreatment with this agonist decreased angiotensin II-stimulated membrane GTPase activity by 91% (n ϭ 4). In contrast, exposure of cells overexpressing GRK2 K220R to angiotensin II failed to diminish angiotensin II-stimulated membrane GTPase activity at all. Thus, inhibition of GRK activity by GRK2 K220R eliminates agonist-induced receptor desensitization assessed with maximal agonist stimulation.
Assessed in intact cells, however, GRK2 K220R appears to behave differently. In phosphoinositide hydrolysis experiments (data not shown), we observed that GRK2 K220R overexpression decreased even maximal angiotensin II-stimulated phosphoinositide hydrolysis in intact cells by over 90%, just as wild type GRK2 did. Taken together with the GTPase findings, these data indicate that overexpressed GRKs may inhibit receptor signaling in intact cells not only by increasing receptor phosphorylation, but also by binding to the activated receptor and preventing receptor/G protein coupling. This latter function can also be served by GRK2 K220R . The dampening effect of Receptors were immunoprecipitated and analyzed by SDS-PAGE. a shows the autoradiogram from one representative experiment. b summarizes the results of five experiments performed in duplicate. Receptor phosphorylation is expressed as a percentage of that seen in unstimulated control cells. Shown are mean Ϯ S.E.; **, p Ͻ 0.001 compared with angiotensin II-stimulated control cells. c, immunoblot depicting GRK expression in 293 cells transiently transfected with pcDNA I-AT 1A R and either empty vector, pcDNA I-GRK2, pcDNA I-GRK3, or pcDNA I-GRK5. After electrophoretic separation (10% SDS-PAGE) of cellular lysates (10 g of cellular protein per lane) and electrotransfer to nitrocellulose, proteins were visualized with anti-GRK2/3 mouse monoclonal antibody C5/1 (M. Oppermann, unpublished data; 10 g/ml) or rabbit anti-GRK5 antiserum (12) at 1:1000, a 1000fold dilution of horseradish peroxidase-conjugated goat anti-mouse or goat anti-rabbit antibodies (Jackson ImmunoResearch Laboratories), and diaminobenzidine as the peroxidase substrate with NiCl 2 enhancement. GRK2 K220R on the process of signal generation is avoided by determining agonist-stimulated GTPase activity in washed membranes.
The role of PKC-mediated AT 1A -R phosphorylation in receptor desensitization was examined in intact cells which were pretreated with staurosporine prior to a desensitizing stimulus with angiotensin II (data not shown). To maximize potential effects of PKC, cells were treated as in Fig. 1, except that the prestimulation with angiotensin II was 12 min instead of 3 min. Pretreatment with 150 nM staurosporine for 10 min prior to this angiotensin II challenge had no effect on desensitization of agonist-induced phosphoinositide hydrolysis assessed with maximal agonist challenge. Concordant findings derive from 293 cells co-transfected with the AT 1A -R and the endothelin B receptor. Stimulation of these cells with endothelin-1, which activates PKC, fails to diminish the subsequent phosphoinositide hydrolysis stimulated by a maximal challenge with angiotensin II. 2

DISCUSSION
Many G protein-coupled receptors undergo regulatory processes which result in the diminution of their functions upon repeated or prolonged stimulation. Among several distinct mechanisms that have been implicated in the rapid desensitization of G protein-coupled receptor systems, covalent modification of agonist-bound receptors by phosphorylation appears to be particularly important. A current model of homologous desensitization of G protein-coupled receptors proposes rapid, agonist-dependent translocation to the plasma membrane of a GRK, which binds and phosphorylates the agonist-occupied receptor (2). The subsequent binding of an arrestin-like protein is believed to uncouple the activated receptor from its G protein. This paradigm was mainly established by utilizing purified proteins in vitro. The significance of GRK-mediated phosphorylation for the regulation of receptor function under intact cell conditions has been demonstrated only in few receptor systems, so far (13,(23)(24)(25)(26). In the present study, which focuses on the type 1A angiotensin II receptor, we inhibited endogenous cellular GRK activity by a dominant negative GRK2 mutant. Overexpression of this GRK2 mutant was found to inhibit significantly both the agonist-induced phosphorylation and desensitization of the AT 1A -R. Furthermore, by overexpressing GRKs 2, 3, and 5, we show that each of these kinases augments agonist-induced AT 1A -R phosphorylation. These findings imply an important role for GRKs in the regulation of the AT 1A -R.
In this study, we have concentrated on a possible role for GRKs and second messenger-dependent kinases in the agonistdependent phosphorylation of the AT 1A -R. While we cannot exclude the possibility that other mechanisms participate in the desensitization of this receptor, several lines of evidence point to a central role for GRKs under the experimental conditions used in this study. Firstly, the concentration dependence of angiotensin II-induced desensitization (EC 50 Ϸ2 nM ATII) paralleled ligand binding to the receptor. AT 1A -R phosphorylation followed the same concentration dependence. The ability to phosphorylate G protein-coupled receptors only in their activated, agonist-bound states is a hallmark of GRKs. Secondly, maximal desensitization of the AT 1A -R was observed after prestimulation of cells for only 1 min. This finding is consistent with the rapid kinetics of GRK-mediated receptor phosphorylation (22), confirmed with the AT 1A -R in this study. Thirdly, GRK inhibition by a GRK2 dominant negative mutant completely prevented AT 1A -R desensitization as assessed by angiotensin II-stimulated membrane GTPase activity. In contrast, while PKC was also found to phosphorylate the AT 1A -R after stimulation with angiotensin II, the functional consequences of PKC-mediated receptor phosphorylation appear unimportant, at least when assessed at maximal levels of receptor stimulation. Recently, desensitization has been demonstrated with a truncated mutant of AT 1A -R which lacks the carboxyl-terminal 45 amino acids, encompassing 13 serine/threonine residues (27). Although this result may seem difficult to reconcile with our own findings, the Ser and Thr residues important for AT 1A -R desensitization have not yet been determined. Indeed, four serine and threonine residues in the first or second cytoplasmic loops, not affected by this deletion mutant, involve receptor regions critical for G protein coupling (28) and may represent functionally important phosphorylation sites.
The functional differentiation of PKC-mediated from GRKmediated receptor phosphorylation seems to relate to their differing time courses of action. In this regard, our findings are supported by the kinetic analysis of angiotensin II-induced translocation of PKC in vascular smooth muscle strips (29). The spatial translocation of the cytosolic kinase to the membrane, which is a prerequisite for PKC activation, was found to peak at 5 to 10 min and then gradually decline to background values. These data are in accord with the PKC-mediated AT 1A -R phosphorylation, which we determined to proceed with a t1 ⁄2 of about 3 min. The different time course of GRK-versus PKC-mediated receptor phosphorylation resembles the kinetics of GRK-and PKA-mediated phosphorylation and desensitization of the ␤ 2 -adrenergic receptor (22). In this system, the GRKs rapidly induced receptor phosphorylation and desensitization with a t1 ⁄2 of about 15 s, whereas the second messengerdependent enzyme PKA had a t1 ⁄2 of 2 to 3.5 min.
Whereas several phospholipase C-coupled receptors have been shown to undergo negative feedback inhibition via PKC (30 -32), conflicting data have been reported in the past with regard to the AT 1 -R. PKC depletion or treatment with the selective PKC inhibitor Ro 31-7519 was shown not to affect the rapid agonist-induced desensitization in neonatal cardiac myocytes (33) or in CHO cells expressing the human AT 1 -R (34). In our own experiments, PKC inhibition with staurosporine did not affect homologous desensitization of the AT 1A -R. In contrast, Pfeilschifter and co-workers (35)(36)(37)(38) have accumulated ample evidence implicating PKC in the angiotensin II-induced homologous desensitization in glomerular mesangial cells. In aggregate, these studies and ours suggest that the AT 1A -R might be regulated by PKC in a cell type-specific manner.
Using staurosporine as a PKC inhibitor and a dominant negative GRK2 K220R mutant to prevent endogenous GRKs from interacting with the receptor, each type of kinase was shown to account for 40 -50% of the receptor phosphorylation induced by agonist within 5 min. The virtually additive effect of the two kinase inhibitors identifies GRKs and PKC as the main, if not the only, kinases which cause short-term agonist-induced phosphorylation of the AT 1A -R. The minimal residual receptor phosphorylation in the presence of both staurosporine and the dominant negative GRK2 K220R mutant is probably due to incomplete efficacy of these inhibitors. Indeed, even when present in 15-fold molar excess, GRK2 K220R was shown to inhibit only 60% of GRK2 activity on the ␤2-adrenergic receptor in the presence of G protein ␤␥ subunits (20).
Phosphorylation of the AT 1A -R by PKA was reported in a previous study, which demonstrated phosphorylation of the receptor upon prolonged (15 min) stimulation with forskolin (8). This result is in contrast to our findings which, however, were obtained by incubation with forskolin for only 5 min. While we cannot definitely rule out a possible role for PKA in the regulation of the AT 1A -R, this enzyme is not an effector kinase of the G q -coupled AT 1A -R and therefore would not be expected to be involved in agonist-induced receptor phosphorylation. In the same study, evidence was provided for both constitutive and agonist-induced tyrosine phosphorylation of the AT 1A -R (8). Phosphoamino analysis of the receptor showed that stimulation with angiotensin II induces mainly serine phosphorylation, which supports our findings of GRK-and PKC-mediated receptor phosphorylation. Tyrosine phosphorylation was observed to a small extent and only after prolonged (Ն20 min) stimulation with angiotensin II. In our own studies, the tyrosine kinase inhibitor genestein had no effect on receptor phosphorylation or functional consequences of angiotensin II stimulation (data not shown). Tyrosine phosphorylation therefore does not appear to contribute to the rapid agonistinduced desensitization of the AT 1A -R.
Our study is the first to propose a role for GRKs in the agonist-induced phosphorylation and desensitization of the AT 1A -R. Each of the three tested GRK isoforms was shown to enhance receptor phosphorylation equally well. Nonetheless, depending on the varying tissue distributions and intracellular concentrations of GRKs, individual kinases may still fulfill highly specific roles in the regulation of the AT 1A -R and other G protein-coupled receptors.