Polymorphic Deletion of Three Intracellular Acidic Residues of the α2B-Adrenergic Receptor Decreases G Protein-coupled Receptor Kinase-mediated Phosphorylation and Desensitization*

A polymorphic variant of the human α2B-adrenergic receptor (α2BAR), which consists of a deletion of three glutamic acids (residues 301–303) in the third intracellular loop was found to be common in Caucasians (31%) and to a lesser extent in African-Americans (12%). The consequences of this deletion were assessed by expressing wild-type and the Del301–303 receptors in Chinese hamster ovary and COS cells. Ligand binding was not affected, although a small decrease in coupling efficiency to the inhibition of adenylyl cyclase was observed with the mutant. The deletion occurs within a stretch of acidic residues that is thought to establish the milieu for agonist-promoted phosphorylation and desensitization of the receptor by G protein-coupled receptor kinases (GRKs). Agonist-promoted phosphorylation studies carried out in cells coexpressing the α2BARs and GRK2 revealed that the Del301–303 receptor displayed ∼56% of wild-type phosphorylation. Furthermore, the depressed phosphorylation imposed by the deletion was found to result in a complete loss of short term agonist-promoted receptor desensitization. Thus the major phenotype of the Del301–303 α2BAR is one of impaired phosphorylation and desensitization by GRKs, and thus the polymorphisms renders the receptor incapable of modulation by this key mechanism of dynamic regulation.

␣ 2 -Adrenergic receptors (␣ 2 AR) 1 are cell surface receptors for catecholamines that bind to the G i /G o family of G proteins, coupling to multiple effector systems including inhibition of adenylyl cyclase activity (1). ␣ 2 AR are widely expressed within the central and peripheral nervous system (2)(3)(4) and participate in a broad spectrum of physiologic functions such as regulation of blood pressure both centrally and within the vasculature, sedation, analgesia, regulation of insulin release, renal function, and cognitive and behavioral functions (5)(6)(7)(8)(9)(10)(11)(12). Three human ␣ 2 AR subtypes have been cloned and characterized (␣ 2A , ␣ 2B , and ␣ 2C ). The ␣ 2B AR has a distinct pattern of expres-sion within the brain, liver, lung, and kidney, and recent studies using gene knockouts in mice have shown that disruption of this receptor effects mouse viability (13), blood pressure responses to ␣ 2 AR agonists (13), and the hypertensive response to salt loading (14).
Like the ␣ 2A AR subtype (15,16), the ␣ 2B AR undergoes short term agonist promoted desensitization (17). This desensitization is due to phosphorylation of the receptor, which evokes a partial uncoupling of the receptor from functional interaction with G i /G o (18,19). Such phosphorylation appears to be due to G protein-coupled receptor kinases (GRKs), a family of serine/ threonine kinases that phosphorylate the agonist-occupied conformations of many G protein-coupled receptors (20). The process serves to finely regulate receptor function providing for rapid adaptation of the cell to its environment. Desensitization may also limit the therapeutic effectiveness of administered agonists. For the ␣ 2B AR, phosphorylation of serines/threonines in the third intracellular loop of the receptor is dependent on the presence of a stretch of acidic residues in the loop that appears to establish the milieu for GRK function (18). In this report we delineate the phenotype of a common polymorphism of the ␣ 2B AR (21,22), which consists of a deletion of three glutamic acid residues in this region; such a variation has a pronounced effect on receptor phosphorylation leading to a loss of agonist-promoted desensitization.

MATERIALS AND METHODS
Polymorphism Detection-The sequence encoding the third intracellular loop of the human ␣ 2B AR (GenBank TM accession number AF005900) was examined for polymorphic variation by performing polymerase chain reactions (PCRs) to amplify this portion of the cDNA from genomic DNA derived from blood samples. In this paper the adenine of the initiator ATG codon of the open reading frame of the receptor is designated as nucleotide 1, and amino acid 1 is the encoded methionine. The human receptor consists of 450 amino acids. For initial examination, DNA from 39 normal individuals was utilized. Two overlapping fragments encompassing the third intracellular loop region were generated using the following primer pairs: fragment 1 (534 bp), 5Ј-GCTCATCATCCCTTTCTCGCT (sense) and 5Ј-AAAGCCCCACCAT-GGTCGGGT (antisense) and fragment 2 (588 bp), 5Ј-CTGATCGC-CAAACGAGCAAC (sense) and 5Ј-AAAAACGCCAATGACCACAG (antisense). The 5Ј end of each sense and antisense primer also contained sequences corresponding to the M13 forward (5Ј-TGTAAAACGACGGC-CAGT) and M13 reverse (5Ј-CAGGAAACAGCTATGACC) universal sequencing primers, respectively. The PCR reactions consisted of ϳ100 ng of genomic DNA, 5 pmol of each primer, 0.8 mM dNTPs, 10% Me 2 SO, 2.5 units of Platinum Taq DNA polymerase (Life Technologies, Inc.), 20 l of 5ϫ buffer J (Invitrogen) in a 100-l reaction volume. Reactions were started by an initial incubation at 94°C for 4 min, followed by 35 cycles of 94°C for 30 s, 58°C (fragment 1) or 60°C (fragment 2) for 30 s, and 72°C for 1 min, followed by a final extension at 72°C for 7 min. PCR reactions were purified using the QIAquick PCR purification system (Qiagen), and automated sequencing of both strands of each PCR product was performed using Applied Biosystems 370 sequencer using dye primer methods. As discussed below, a 9 bp in-frame deletion beginning at nucleotide 901 was detected that resulted in a loss of three glutamic acid residues at amino acid positions 301-303. Thus, this polymorphism was denoted Del301-303. No other nonsynonymous or synonymous polymorphisms were identified. PCR amplification of 209-and 200-bp fragments encompassing this polymorphic region allowed screening of additional DNA samples whose genotypes were distinguished by size when run on 4% Nuseive agarose gels. PCR conditions were the same as described above except that buffer F was used with the following primers: 5Ј-AGAAGGAGGGTGTTTGTGGGG (sense) and 5Ј-ACCTATAG-CACCCACGCCCCT (antisense).
Constructs and Cell Transfection-To create the polymorphic ␣ 2B AR construct, a 1585-bp PCR product encompassing the ␣ 2B AR gene was amplified from a homozygous deletion individual using the following primers: 5Ј-GGCCGACGCTCTTGTCTAGCC (sense) and 5Ј-CAAGGG-GTTCCTAAGATGAG. This fragment was digested and subcloned into the XcmI and BamHI sites of the wild-type ␣ 2B AR sequence in the expression vector pBC12BI (17). The integrity of the construct was verified by sequencing. Chinese hamster ovary cells (CHO-K1) were stably transfected by a calcium phosphate precipitation technique as described previously using 30 g of each receptor construct and 0.5 g of pSV 2 neo to provide for G418 resistance (23). Selection of positive clones was carried out in 1.0 mg/ml G418, and expression of the ␣ 2B AR from individual clonal lines was determined by radioligand binding as described below. Several clonal lines with matched expression levels between 500 and 1000 fmol/mg were utilized as indicated. Cells were grown in monolayers in Ham's F-12 medium supplemented with 10% fetal calf serum, 100 units/ml penicillin, 100 g/ml streptomycin, and 80 g/ml G418 (to maintain selection pressure) at 37°C in a 5% CO 2 atmosphere. For phosphorylation experiments, receptors were epitope tagged with the influenza hemagglutinin nonopeptide (YPYDVPDYA) at the amino terminus. This was accomplished by constructing vectors using the above constructs with insertions of in-frame sequence encoding the peptide using PCRs essentially as described previously (24). Tagged receptors were expressed at ϳ15 pmol/mg, along with GRK2 (␤ARK1), in COS-7 cells using a DEAE Dextran technique as described (16).
Adenylyl Cyclase Activities-␣ 2 AR inhibition of adenylyl cyclase was determined in membrane preparations from CHO cells stably expressing the two receptors using methods similar to those described previously (25). Briefly, cell membranes (ϳ20 g) were incubated with 27 M phosphoenolpyruvate, 0.6 M GTP, 0.1 mM cAMP, 0.12 mM ATP, 50 g/ml myokinase, 0.05 mM ascorbic acid, and 2 Ci of [␣-32 P]ATP in a buffer containing 40 mM HEPES, pH 7.4, 1.6 mM MgCl 2 , and 0.8 mM EDTA for 30 min at 37°C. Activities were measured in the presence of water (basal), 5 M forskolin, and 5 M forskolin with the indicated concentrations of agonists. Reactions were terminated by the addition of a stop solution containing excess ATP and cAMP and ϳ100,000 dpm of [ 3 H]cAMP. Labeled cAMP was isolated by gravity chromatography over alumina columns with [ 3 H]cAMP used to quantitate column recovery. Results are expressed as percentages of inhibition of forskolinstimulated activity. For desensitization experiments, cells were pretreated for 30 min at 37°C with medium alone or with medium containing 10 M norepinephrine, placed on ice, and washed five times with cold phosphate-buffered saline prior to membrane preparation. Desensitization of ␣ 2B AR is manifested by a shift to the right in the dose-response curve for the inhibition of adenylyl cyclase (i.e. increase in EC 50 ) without a significant change in the maximal response (17)(18)(19). To quantitate the magnitude of this desensitization, the inhibitory response under control conditions at a submaximal concentration of agonist (the EC 50 ) in the assay was determined from the curve and compared with the response to this same concentration from membranes derived from cells exposed to norepinephrine. This method has been previously validated (26) in several G protein-coupled receptor systems.
Radioligand Binding-Expression of mutant and wild-type ␣ 2B AR was determined using saturation binding assays as described (25,27) with Intact Cell Receptor Phosphorylation-Transiently transfected COS-7 cells expressing equivalent levels of each receptor were grown to confluence and incubated with [ 32 P]orthophosphate (ϳ4 mCi/100-mm plate) for 2 h at 37°C in 5% CO 2 atmosphere. Cells were then incubated in the presence or absence of 100 M norepinephrine for 15 min, washed five times with ice-cold phosphate-buffered saline, and solubilized in 1 ml of a buffer containing 1% Triton X-100, 0.05% SDS, 1 mM EDTA, and 1 mM EGTA in phosphate-buffered saline, by rotation in a microcentrifuge tube for 2 h at 4°C. This and all subsequent steps included the protease inhibitors benzamidine (10 g/ml), soybean trypsin inhibitor (10 g/ml), aprotinin (10 mg/ml), and leupeptin (5 g/ml) and the phosphatase inhibitors sodium fluoride (10 mM) and sodium pyrophosphate (10 mM). (A separate flask was scraped in 5 mM Tris/2 mM EDTA, and the membranes were prepared for radioligand binding as described above.) Unsolubilized material was removed by centrifugation at 40,000 ϫ g at 4°C for 10 min. The hemagglutinin epitope tagged ␣ 2B ARs were immunoprecipitated using an anti-hemagglutinin high affinity monoclonal antibody (Roche Molecular Biochemicals) as described previously (24). Briefly, solubilized material was preincubated for 2 h at 4°C with protein G-Sepharose beads to remove nonspecific binding. The supernatant was then incubated with protein G-Sepharose beads and a 1:200 dilution of antibody for 18 h at 4°C. Following immunoprecipitation, the beads were washed three times by centrifugation and resuspension and then incubated at 37°C for 1 h in SDS sample buffer. Proteins in the supernatant were then fractionated on a 10% SDS-polyacrylamide gel with equal amounts of receptor (based on radioligand binding) loaded in each lane. Signals were visualized and quantitated using a Molecular Dynamics PhosphorImager with Image-Quant Software.
Miscellaneous-Protein determinations were by the copper bicinchoninic acid method (28). Data from adenylyl cyclase and radioligand binding assays were analyzed by iterative least square techniques using Prizm software (GraphPad, San Diego, CA). Agreement between genotypes observed and those predicted by the Hardy-Weinberg equilibrium was assessed by a Chi-squared test with one degree of freedom. Genotype comparisons were by Fisher's exact test. Comparisons of results from biochemical studies were by t tests, and significance was considered when p Ͻ 0.05. Data are provided as the means Ϯ S.E.

RESULTS AND DISCUSSION
Sequence analysis of the third intracellular loop of the ␣ 2B AR gene from 78 chromosomes revealed a single sequence variant. This consisted of an in-frame 9-bp deletion (GAAGAGGAG) beginning at nucleotide 901 (Fig. 1A) that results in loss of three glutamic acid residues at amino acid positions 301-303 of the third intracellular loop of the receptor (Fig. 2). Using the rapid detection method (Fig. 1B), allele frequencies were determined in a larger population. The frequencies of the wild-type and the Del301-303 polymorphic ␣ 2B AR are shown in Table I. The deletion polymorphism is more common in Caucasians than African-Americans, with allele frequencies of 0.31 and 0.12, respectively. The distribution of homozygous and heterozygous alleles in either population was not different than that predicted from the Hardy-Weinberg equilibrium (p Ͼ 0.8).
The consequences of this polymorphism on ligand binding and receptor function were evaluated by stably expressing the human wild-type ␣ 2B AR and the Del301-303 receptor in CHO cells (Table II). Saturation radioligand binding studies revealed a small but statistically significant lower affinity for the ␣ 2 AR antagonist [ 3 H]yohimbine for Del301-303 compared with the wild-type receptor (K d ϭ 5.1 Ϯ 0.2 versus 3.8 Ϯ 0.3 nM, respectively, n ϭ 5, p Ͻ 0.05). Agonist (epinephrine) competition binding experiments carried out in the presence of quanosine 5Ј(␤-␣-imido)triphosphate revealed a small increase in the K i for the polymorphic receptor (285 Ϯ 8.7 versus 376 Ϯ 66 nM, n ϭ 5, p Ͻ 0.05). In similar studies carried out in the To address the functional consequences of the mutation, studies examining agonist-promoted inhibition of forskolinstimulated adenylyl cyclase activities were carried out in lines expressing the wild-type ␣ 2B AR and the Del301-303 receptor at densities of 626 Ϯ 54 and 520 Ϯ 82 fmol/mg (n ϭ 7, p Ͼ 0.05). The results of these studies are shown in Table III. As can be seen, the Del301-303 receptor displayed less inhibition of adenylyl cyclase (23.4 Ϯ 2.2%) compared with wild-type ␣ 2B AR (28.5 Ϯ 1.6%, p Ͻ 0.05). Furthermore, the polymorphic receptor had a greater EC 50 (19.6 Ϯ 5.5 versus 7.9 Ϯ 2.1 M, p Ͻ 0.01). Thus, the loss of the three glutamic acids in the third intracellular loop, which is known to contain regions important for G protein coupling, results in a modest decrease in agonist-mediated receptor function.
The deletion polymorphism occurs in a highly acidic stretch of amino acids (EDEAEEEEEEEEEEEE) within the third intracellular loop of ␣ 2B AR (Fig. 2). The structural importance of this region has been previously assessed and shown to be critical for short term agonist-promoted receptor phosphorylation leading to desensitization (18). These data and reports by others (29) suggest that this acidic environment is necessary for receptor phosphorylation by GRKs. Therefore, to investigate the consequences of this deletion polymorphism on receptor desensitization, agonist-promoted inhibition of adenylyl cyclase activity was determined in membranes from CHO cells expressing the wild-type and Del301-303 receptor after pretreatment with norepinephrine. In these experiments, cells were incubated with medium alone or medium containing agonist (10 M norepinephrine) for 30 min at 37°C and subsequently extensively washed at 4°C; membranes were prepared, and agonist-mediated inhibition of forskolin stimulated adenylyl cyclase activity was determined. As described previously (17) and shown in Fig. 3 and Table II,  change in the EC 50 for the deletion receptor following agonist pretreatment (29.5 M versus 31.2 M). Desensitization was quantitated by examining adenylyl cyclase activities at a submaximal concentration of agonist (the EC 50 for the control condition). At this concentration, wild-type ␣ 2B AR inhibited adenylyl cyclase activity by 16.5 Ϯ 3.9%; with agonist preexposure, inhibition at this same concentration of agonist was 7.6 Ϯ 2.3% (n ϭ 4, p Ͻ 0.05, Fig. 3C), amounting to ϳ54% desensitization of receptor function. Submaximal inhibition of adenylyl cyclase for the Del301-303 receptor, however, was not different between control and agonist-treated cells (17.1 Ϯ 3.0% versus 15.9 Ϯ 1.7%, n ϭ 4, p ϭ ns). In another two cell lines with matched expression of ϳ600 fmol/mg, the same desensitization phenotypes for wild-type and the Del301-303 polymorphic receptor were observed (data not shown).
We next performed whole cell phosphorylation studies of the wild-type and polymorphic ␣ 2B AR under the same conditions used for desensitization. We hypothesized that agonist-promoted phosphorylation would be decreased in the polymorphic receptor. However, given that this receptor displays rightwardshifted dose-response curves for inhibition of adenylyl cyclase at base line, we also considered the possibility that the receptor is significantly phosphorylated in the basal state. Studies were carried out in cells cotransfected with the receptor and GRK2 (␤ARK1), a strategy that we have previously shown to be useful in identifying receptor-GRK interactions (30). The results of a representative study are shown in Fig. 4. The wild-type ␣ 2B AR underwent a 5.84 Ϯ 0.49-fold increase in phosphorylation with agonist exposure. In contrast, although the Del301-303 receptor displayed some degree of agonist-promoted phosphorylation, the extent was clearly less (3.28 Ϯ 0.24-fold, p Ͻ 0.05 compared with wild type). Basal phosphorylation was equivalent between the two receptors.
It is interesting to note that this partial loss of phosphorylation results in a receptor that fails to undergo any degree of functional desensitization. Although it might seem reasonable to assume that such phosphorylation would be associated with some degree of desensitization, several previous studies with the ␣ 2A AR and ␣ 2B AR subtypes indicate that full (i.e. wild type) phosphorylation is necessary for the desensitization process (16,18,24). For the ␣ 2A AR, we have shown that four serines in the third intracellular loop are phosphorylated after agonist exposure (16). Removal of serines by alanine substitution mutagenesis results in a proportional decrease in phosphorylation. Such partial phosphorylation (compared with wild type), however, was found to be insufficient to cause any detectable desensitization. In a previous study of the ␣ 2B AR, we deleted and substituted the entire aforementioned acidic region (18). Agonist-promoted phosphorylation was reduced by ϳ50% in this mutant, and desensitization was ablated. These results are entirely consistent with the current work, where a restricted deletion resulted in a decrease in phosphorylation and a complete loss of desensitization. Finally, we have also recently shown that a chimeric ␣ 2A /␣ 2C AR, which undergoes agonistpromoted phosphorylation, fails to exhibit desensitization (24). Taken together with our current work, these results indicate that the conformation of the third loop evoked by GRK mediated phosphorylation which provides for the binding of arrestins (which is the ultimate step that imparts uncoupling) is highly specific. Thus, a precise phosphorylation-dependent conformation is apparently required for arrestin binding to ␣ 2 AR and subsequent functional desensitization. Perturbations of the milieu can thus have significant functional consequences, as occurs with the Del301-303 polymorphic ␣ 2B AR.
Thus, the major signaling phenotype of the ␣ 2B AR Del301-303 polymorphism is one of decreased agonist-promoted phosphorylation that results in a complete loss of the ability for the receptor to undergo agonist-promoted desensitization. In addition, the polymorphism imposes a small decrease in receptor coupling. The potential physiologic consequences of the polymorphism could be related to either or both of the above phenotypes. A receptor that fails to undergo desensitization would be manifested as static signaling despite continued activation of the receptor by endogenous or exogenous agonist. Such a lack of regulation by agonist may perturb the dynamic relationship between incoming signals and receptor responsiveness that maintains homeostasis under normal or pathologic conditions. Recently, Gavras and colleagues (14) have shown that ␣ 2B Ϫ/ϩ mice fail to display a hypertensive response to salt loading after subtotal nephrectomy. Thus, a polymorphic ␣ 2B AR that fails to desensitize (i.e. does not display regulatable function) may predispose to salt-sensitive hypertension. Regarding the therapeutic response to ␣ 2 AR agonists, the phenotype of the Del301-303 receptor indicates that individuals with this poly-  morphism would display little tachyphylaxis to continued administration of agonists. In addition, the initial response to agonist might also be reduced based on the somewhat de-pressed coupling of the Del301-303 receptor. Until recently, it has been difficult to differentiate ␣ 2B AR function from the other two subtypes in physiologic studies. With the development of knockout mice lacking each ␣ 2 AR subtype (5,6,13,31), certain functions can now be definitively attributed to specific subtypes. Characterization of the ␣ 2B AR knockout mouse has indicated that the ␣ 2B AR subtype is expressed on vascular smooth muscle and is responsible for the hypertensive response to ␣ 2 AR agonists (13). This indicates that vascular ␣ 2B AR contribute to overall vascular tone and thus participate in systemic blood pressure regulation. This role may be more important, however, during adaptive conditions, such as salt loading, because resting blood pressure is normal in the heterozygous ␣ 2B Ϫ/ϩ mice (14). Whether the ␣ 2B Ϫ/Ϫ mice have altered resting blood pressures has not been studied in detail because of high perinatal lethality of the homozygous knockout (13). However, neither the region of chromosome 2 near the ␣ 2B AR coding sequence nor the deletion polymorphism has been linked or associated with hypertension (21,22,32). However, no studies have assessed whether the polymorphism is associated with salt-sensitive hypertension or other phenotypes, or the response to ␣ 2 AR agonist.
In summary, we have delineated the signaling phenotype of a polymorphism of the ␣ 2B AR that results in a deletion of three glutamic acids in the third intracellular loop of the receptor. The polymorphism is prevalent in the human population, with a frequency that is ϳ2-fold greater in Caucasians as compared with African-Americans. The polymorphic receptor displays wild-type agonist binding affinity but a small decrease in function in the resting state. However, the major phenotype is a significant decrease in agonist-promoted phosphorylation by GRKs, which results in a receptor that fails to display agonistpromoted desensitization. To our knowledge this is the first polymorphism of any G protein-coupled receptor to affect GRKmediated phosphorylation.
FIG. 3. The Del301-303 ␣ 2B AR fails to undergo short term agonist-promoted desensitization. Cells in culture expressing the two receptors were exposed to vehicle or 10 M norepinephrine for 30 min at 37°C and washed extensively; membranes were prepared, and adenylyl cyclase activities were determined as described under "Experimental Procedures." In A and B, results of full dose-response studies are shown, which reveal that whereas the wild type undergoes desensitization manifested as a rightward shift in the curve, the Del301-303 mutant does not. In C, the percentage of inhibition of adenylyl cyclase at a submaximal concentration of norepinephrine in the assay (the EC 50 of the control membranes) is shown for both conditions, indicating an ϳ54% desensitization of wild-type ␣ 2B AR. The Del301-303 failed to display such desensitization. Results are from four independent experiments. See also Table III. *, p Ͻ 0.05 compared with control. WT, wild type.
FIG. 4. The Del301-303 ␣ 2B AR has impaired agonist-promoted phosphorylation. Cells coexpressing each receptor and GRK2 were incubated with [ 32 P]orthophosphate, exposed to 10 M norepinephrine for 15 min, and receptor purified by immunoprecipitation as described under "Experimental Procedures." Shown is an autoradiogram from a single experiment representative of four performed (see text for mean results). NE, norepinephrine.