A Gain-of-function Polymorphism in a G-protein Coupling Domain of the Human β1-Adrenergic Receptor*

The β1-adrenergic receptor (β1AR) is a key cell surface signaling protein expressed in the heart and other organs that mediates the actions of catecholamines of the sympathetic nervous system. A polymorphism in the intracellular cytoplasmic tail near the seventh transmembrane-spanning segment of the human β1AR has been identified in a cohort of normal individuals. At amino acid position 389, Gly or Arg can be found (allele frequencies 0.26 and 0.74, respectively), the former previously considered as the human wild-type β1AR. Using site-directed mutagenesis to mimic the two variants, CHW-1102 cells were permanently transfected to express the Gly-389 and Arg-389 receptors. In functional studies with matched expression, the Arg-389 receptors had slightly higher basal levels of adenylyl cyclase activities (10.7 ± 1.2 versus 6.1 ± 0.4 pmol/min/mg). However, maximal isoproterenol-stimulated levels weremarkedly higher for the Arg-389 as compared to the Gly-389 receptor (63.3 ± 6.1 versus 20.9 ± 2.0 pmol/min/mg). Agonist-promoted [35S]guanosine 5′-O-(thiotriphosphate) binding was also increased with the Arg-389 receptor consistent with enhanced coupling to Gsand increased adenylyl cyclase activation. In agonist competition studies carried out in the absence of guanosine 5′-(β,γ-imido)triphosphate, high affinity binding could not be resolved with the Gly-389 receptor, whereas Arg-389 displayed an accumulation of the agonist high affinity receptor complex (R H = 26%). Taken together, these data indicate that this polymorphic variation of the human β1AR results in alterations of receptor-Gs interaction with functional signal transduction consequences, consistent with its localization in a putative G-protein binding domain. The genetic variation of β1AR at this locus may be the basis of interindividual differences in pathophysiologic characteristics or in the response to therapeutic βAR agonists and antagonists in cardiovascular and other diseases.

The ␤ 1 -adrenergic receptor (␤ 1 AR) is a key cell surface signaling protein expressed in the heart and other organs that mediates the actions of catecholamines of the sympathetic nervous system. A polymorphism in the intracellular cytoplasmic tail near the seventh transmembrane-spanning segment of the human ␤ 1 AR has been identified in a cohort of normal individuals. At amino acid position 389, Gly or Arg can be found (allele frequencies 0.26 and 0.74, respectively), the former previously considered as the human wild-type ␤ 1 AR. Using site-directed mutagenesis to mimic the two variants, CHW-1102 cells were permanently transfected to express the Gly-389 and Arg-389 receptors. In functional studies with matched expression, the Arg-389 receptors had slightly higher basal levels of adenylyl cyclase activities (10.7 ؎ 1.2 versus 6.1 ؎ 0.4 pmol/min/mg). However, maximal isoproterenol-stimulated levels were markedly higher for the Arg-389 as compared to the Gly-389 receptor (63.3 ؎ 6.1 versus 20.9 ؎ 2.0 pmol/min/ mg). Agonist-promoted [ 35 S]guanosine 5-O-(thiotriphosphate) binding was also increased with the Arg-389 receptor consistent with enhanced coupling to G s and increased adenylyl cyclase activation. In agonist competition studies carried out in the absence of guanosine 5-(␤,␥-imido)triphosphate, high affinity binding could not be resolved with the Gly-389 receptor, whereas Arg-389 displayed an accumulation of the agonist high affinity receptor complex (R H ‫؍‬ 26%). Taken together, these data indicate that this polymorphic variation of the human ␤ 1 AR results in alterations of receptor-G s interaction with functional signal transduction consequences, consistent with its localization in a putative G-protein binding domain. The genetic variation of ␤ 1 AR at this locus may be the basis of interindividual differences in pathophysiologic characteristics or in the response to therapeutic ␤AR agonists and antagonists in cardiovascular and other diseases.
The ␤ 1 -adrenergic receptor (␤ 1 AR) 1 is a member of the adrenergic family of G-protein-coupled receptors, with epinephrine and norepinephrine being endogenous agonists. Like other members of the G-protein-coupled receptor superfamily, the amino terminus is extracellular, the protein is predicted to traverse the cell membrane seven times, and the carboxyl terminus is intracellular. In the adrenergic receptor family, agonists bind in a pocket formed by the transmembrane-spanning domains, and G-protein binding and activation occur at intracellular domains of the loops and tail, typically near the membrane (1). ␤ 1 ARs couple to the stimulatory G-protein, G s , activating adenylyl cyclase, as well as to non-cAMP pathways such as the activation of ion channels (2). ␤ 1 ARs are expressed on a number of cell types including cardiomyocytes where they serve to increase cardiac inotropy and chronotropy, adipocytes where they mediate lypolysis, and juxtaglomerular cells of the kidney where they regulate renin secretion. It has been known for decades that these responses, as well as those of the ␤ 2 AR, are somewhat variable in the human population (3,4). Indeed, we have recently found and characterized several ␤ 2 AR polymorphisms that have altered signaling phenotypes compared with wild-type receptor (5)(6)(7). This finding has prompted us to examine the ␤ 1 AR coding sequence for variability in the human population. Here we report a common single nucleotide polymorphism resulting in a Gly to Arg switch at intracellular amino acid 389, within a region important for G-protein coupling. The resulting phenotype of the Arg-389 receptor is one of enhanced receptor-G s interaction, functionally manifested as increased activation of the adenylyl cyclase effector.

MATERIALS AND METHODS
Nomenclature-The wild-type sequence of the intronless human ␤ 1 AR gene is considered that reported by Frielle et al. (8) (GenBank TM accession number J03019). In the current report the adenine of the initiator ATG of the coding block is designated as nucleotide 1, and amino acid 1 is the encoded methionine. The receptor consists of 477 amino acids.
Polymorphism Detection-Genomic DNA was extracted from blood or banked tissue from normal individuals by standard techniques (9). The study was approved by the University of Cincinnati Institutional Review Board. PCR reactions were established to produce overlapping products spanning 1131 base pairs of the coding region starting from nucleotide 300 of the second transmembrane-spanning domain past the stop codon into the 3Ј-untranslated region. A consistent deviation from the published sequence was noted at nucleotide 951 where guanine was found in all subjects instead of adenine and likely represents a sequencing error in the original report. Both codons encode for Leu at amino acid 317. At nucleotide 1165, a guanine ("wild-type") or cytosine was found (Fig. 1A), altering the encoded amino acid from Gly to Arg at amino acid position 389. The primers used to amplify the 488-base pair fragment where this polymorphism at position 1165 was detected were: 5Ј-CAGGAAACAGCTATGACCACTGGAGCCGCCTCTTCGTCTTCTT-CAACTG-3Ј (sense) and 5Ј-TGGGCTTCGAGTTCACCTGCTATC-3Ј (antisense). The sense primer contains a 5Ј M13 forward primer sequence, so that dye primer chemistry using the universal primer could be utilized in the sequencing reactions. PCRs consisted of ϳ500 ng of DNA, each dNTP at 62.5 nM, 100 pmol of each primer, 0.5 l of Pwo polymerase (Roche Molecular Biochemicals), 10% of the supplied buffer, and 5% Me 2 SO in a final volume of 50 l. Cycling conditions started at * This work was supported by National Institutes of Health Grants HL52318 and HL41496. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Constructs and Cell Transfection-Site-directed mutagenesis was carried out by methods previously described (5) to mimic the Arg-389 polymorphism. Arg-389 and Gly-389 ␤ 1 AR cDNAs were subcloned into the mammalian expression vector pBC12BI (10). CHW-1102 fibroblasts (CHW cells) were permanently transfected with the Gly-389 and Arg-389 ␤ 1 AR constructs by calcium phosphate precipitation as described previously (11). Positive clones were selected based upon resistance to 300 g/ml G418. Cultures were maintained at 37°C in a 5% CO 2 atmosphere in Dulbecco's modified Eagle's medium with 10% fetal calf serum, 100 units/ml penicillin, 100 g/ml streptomycin, and 80 g/ml G418. COS-7 cells were transiently transfected with the ␤ 1 AR constructs and rat G␣ s in pCDNA1 for [ 35 S]GTP␥S binding studies. These transient transfections were performed by the DEAE-dextran/chloroquine method as described previously (12).
Radioligand Binding-Confluent monolayers of CHW cells were washed three times with phosphate-buffered saline, lysed in a hypotonic buffer (5 mM Tris, 2 mM EDTA, pH 7.4), detached by scraping with a rubber policeman, and centrifuged at 42,000 ϫ g for 10 min. Crude membranes were resuspended in buffer (75 mM Tris, 12.5 mM MgCl 2 , 2 mM EDTA, pH 7.4) and saturation binding experiments performed with 125 I-CYP as described (12), with nonspecific binding defined by coincubation with 1 M propranolol. Reactions for these and other radioligand experiments were terminated by dilution and vacuum filtration over glass fiber filters. One-site agonist competition binding studies were carried out with 40 pM 125 I-CYP, 100 M GTP, and varying concentrations of isoproterenol, epinephrine, or norepinephrine for 2 h at 25°C. To assess high and low affinity agonist receptor binding, membranes were prepared as described above, except two additional centrifugations were included before the addition of reaction buffer to assure the removal of endogenous GTP. Membranes were incubated with 40 pM 125 I-CYP and 18 concentrations of isoproterenol in the presence or absence of the nonhydrolyzable GTP analog Gpp(NH)p at 100 M for 1 h at 37°C. Competition data were fit to one-site and two-site models by an iterative least squares technique as described previously (13). A two-site model was considered valid if by F-test the fit was statistically better (p Ͻ 0.05) than that obtained with a one-site model.
Adenylyl Miscellaneous-Protein concentrations were determined by the copper bicinchoninic acid method (15). Curve fitting was carried out using PRISM software (GraphPad, San Diego, CA). Statistical comparisons were made by paired or unpaired t tests, as appropriate, with significance considered when p Ͻ 0.05. Data are presented as mean Ϯ S.E. of the indicated number of independent experiments.

RESULTS
Sequencing results from DNA derived from 30 individuals from a defined healthy cohort revealed variance in the examined ␤ 1 AR region (encompassing the coding block from the second transmembrane domain to the last amino acid in the carboxyl terminus of the receptor) in only one location (Fig. 1), at nucleotide 1165, where guanine or cytosine could be found, resulting in a Gly or Arg at amino acid 389. Using BsmF1 restriction digests, additional genotyping at this locus was carried out for a total of 50 samples. While the wild-type human ␤ 1 AR had previously been reported (8) as Gly at amino acid 389, we found a higher frequency of Arg in our normal population. The allele frequencies were 0.74 for Arg and 0.26 for Gly.
To ascertain whether the Arg in this position conferred a distinct phenotype compared with the Gly-389 receptor (which we (11) and others (16) had previously characterized), sitedirected mutagenesis was carried out to mimic Arg-389, and the two cDNAs were subcloned into mammalian expression vectors. The final constructs were identical except for the single nucleotide difference. These constructs were used to establish permanent lines of transfected CHW cells, which endogenously  1. Genetic variance of the human ␤ 1 AR at nucleotide 1165, amino acid 389. A, PCR fragments generated from genomic DNA were sequenced as described under "Materials and Methods." The G in the indicated position of codon 389 results in a Gly (left-hand sequence), whereas a C encodes for Arg (right-hand sequence). Shown are representative chromatograms from two individuals homozygous for the two alleles. B, PCR fragments were digested with the restriction endonuclease BsmF1, which fails to digest the Arg-389 product because of the G to C transition. Homo., homozygous; control, no enzyme in the reaction.
lack ␤AR expression. As indicated, studies were carried out on several clonal lines from each transfection at matched levels of receptor expression. Radioligand binding studies (Table I) revealed essentially identical dissociation binding constants for 125 I-CYP. Competition binding studies with isoproterenol, epinephrine, and norepinephrine in the presence of GTP showed no differences between the two receptors in binding affinities for these agonists.
Basal, isoproterenol-, and forskolin-stimulated adenylyl cyclase activities were determined using membranes from lines expressing Gly-389 at 159 Ϯ 14 fmol/mg and Arg-389 at 151 Ϯ 13 fmol/mg ( Fig. 2A). As shown, basal activities were slightly higher with the Arg-389 receptor than with the Gly-389 receptor (10.7 Ϯ 1.2 versus 6.1 Ϯ 0.4 pmol/min/mg, n ϭ 4, p Ͻ 0.02). The most dramatic differences, however, were seen with agonist stimulation. Maximal isoproterenol-stimulated activities were higher with the Arg-389 receptor than with the Gly-389 receptor (63.3 Ϯ 6.1 versus 20.9 Ϯ 2.0 pmol/min/mg, n ϭ 4, p Ͻ 0.01). Similarly, expressing the responses as -fold increases over basal levels, the Arg-389 receptor displayed a greater stimulation of adenylyl cyclase than the Gly-389 receptor (6.1-Ϯ 0.3-fold versus 3.3-Ϯ 0.1-fold). In contrast, stimulation of adenylyl cyclase by forskolin was not consistently different between the lines (104 Ϯ 11 versus 120 Ϯ 10 pmol/min/mg). Thus, normalizing the data to the extent of stimulation of adenylyl cyclase by forskolin still revealed a significantly greater agonist stimulation by the Arg-389 variant than the Gly-389 variant (Fig. 2B). The EC 50 values for isoproterenol stimulation of adenylyl cyclase for Gly-389 and Arg-389 receptors were not different (115 Ϯ 5 versus 132 Ϯ 5 nM). Results from an additional set of studies carried out on other lines at higher expression levels are shown in Fig. 2, C and D (expression was 246 Ϯ 32 and 255 Ϯ 11 fmol/mg for Gly-389 and Arg-389, respectively). Again, the maximal extent of stimulation by agonist was greater with the Arg-389 variant. These differences in adenylyl cyclase stimulation between the two polymorphic receptors were also observed in responses to epinephrine and norepinephrine (data not shown).
To investigate the basis of these differences between the two receptors, studies were undertaken to assess interactions be-tween agonist, receptor, and G s with [ 35 S]GTP␥S binding experiments and agonist competition studies in the absence and presence of Gpp(NH)p. For [ 35 S]GTP␥S binding, we found that adequate signals were reproducibly obtained with high levels of transient receptor expression (ϳ10 pmol/mg) in COS-7 cells, which were co-transfected with G␣ s . Results of these studies are shown in Fig. 3. Maximal isoproterenol-stimulated [ 35 S]GTP␥S binding was found to be greater in membranes bearing the Arg-389 versus the Gly-389 receptor, consistent with the adenylyl cyclase studies, which also showed enhanced agonist-stimulated function with the Arg-389 receptor.
A well recognized characteristic of agonist binding to receptors such as the ␤ 2 AR is the increase in high affinity binding states, indicative of receptor/G-protein interactions (17)(18)(19). Interestingly, we have previously reported (13) that in competition studies carried out in the absence of Gpp(NH)p, agonistpromoted accumulation of a ␤ 1 AR (Gly-389) high affinity fraction is not readily resolved. This result is likely because of the similar affinities of the high and low affinity sites consistent with relatively less free energy transfer during signal transduction. We wondered whether the Arg-389 receptor would in COS-7 cells were transfected as described under "Materials and Methods." Binding in the presence of 10 M isoproterenol was greater (p Ͻ 0.05) with the Arg-389 than with the Gly-389 receptor. Because of day-to-day variation in the absolute levels of binding, data are presented as a percentage of binding to the wild-type (Gly389) receptor (mean absolute values were 7.7 Ϯ 1.4 ϫ 10 5 dpm/mg for Gly-389). Basal levels of binding were not different between the two receptors. nt, nontransfected cells.
fact have enhanced agonist-promoted high affinity binding, because this receptor has increased functional coupling to G s . Isoproterenol competition studies were thus carried out in parallel, in the absence and presence of Gpp(NH)p, with washed membranes from CHW cells expressing each receptor. These results are provided in Fig. 4 and Table II. As shown, the displacement curves in the absence of a guanine nucleotide from the Gly-389 experiments were relatively steep (mean pseudo-Hill coefficient, 0.87 Ϯ 0.03) and were best resolved statistically to a one-site model. Gpp(NH)p had no effect on agonist competition with the Gly-389 receptor (Fig. 4). In contrast, the displacement curves obtained with the Arg-389 receptor in the absence of Gpp(NH)p had mean pseudo-Hill coefficients of 0.73 Ϯ 0.03, and the data were best resolved with a two-site model (composite p value Ͻ 0.001). The percentage of receptors in the high affinity state with the Arg-389 receptor was 26 Ϯ 1% ( Fig. 4 and Table II). As shown, with this receptor the presence of Gpp(NH)p in the reaction resulted in the expected rightward shift, with monophasic, steep curves consistent with a single low affinity agonist interaction. DISCUSSION These findings reveal in the normal human population the presence of two ␤ 1 AR genetic variants with significant differences in functional signaling. For purposes of consistency, and because of the fact that a number of structure/function studies have been published using the Gly-389 receptor (13,20), we prefer to continue to designate this receptor as the "wild-type" human ␤ 1 AR, although the Arg variant appears to be more common. This is analogous to the precedent that has been set with the common ␤ 2 AR polymorphisms (5,6). The site of the variability is ϳ9 amino acids from the seventh transmembrane-spanning domain, in the intracellular portion of the tail prior to the proposed palmitoylated cysteine(s) (Fig. 5). This region is sometimes referred to as the fourth intracellular loop, but in this report will be termed the proximal portion of the cytoplasmic tail. By analogy with ␤ 2 AR (21, 22), ␣ 2 AR (23), and other G-protein-coupled receptors, this region is considered important for receptor coupling to its cognate G-protein, G s . Indeed, the difference between the Arg-389 and Gly-389 receptors is in functional coupling. Receptor-promoted binding of GTP␥S to G␣ s , another indicator of agonist-initiated coupling to G s , was similarly different between the two polymorphic ␤ 1 ARs. Consistent with these findings, agonist-promoted accumulation of the high affinity state in washed membrane preparations in the absence of a guanine nucleotide was detected with the Arg-389 receptor but could not be resolved in studies with the Gly-389 receptor. Based on the multistate model (18,19) of receptor/G-protein interactions, as well as the above results, an elevated basal activity of adenylyl cyclase (i.e. spontaneous toggling to R* in the absence of agonist) should also be expected with the Arg-389 receptor if it has a greater efficiency of stabilizing the active conformation in the presence of agonist, which in fact was the case. The differences in basal activities that we observed were small but consistent, and the magnitude likely results from the relatively low ("physiologic") levels of expression utilized in the functional studies. As such, we consider the functional phenotype as shown in Fig. 2, A and C, to be indicative of signaling in cells that endogenously express the two polymorphic ␤ 1 ARs.
The residues in this region of the ␤ 1 AR among other species are shown in Fig. 5. The amino acid analogous to position 389 of the human, as well as the surrounding residues, are highly conserved in species sequenced to date, with the only deviation at position 389 being found in the human, where Gly was FIG. 4. Agonist competition curves from studies of the Gly-389 and Arg-389 receptors. Experiments were carried out with CHW membranes in the absence and presence of Gpp(NH)p. The Gly-389 data obtained in the absence of a guanine nucleotide was best fit to a one-site model, and the binding was unaffected by Gpp(NH)p. In contrast, studies with Arg-389 resulted in two-site fits with high and low affinity components (see text and Table II) when carried out in the absence of Gpp(NH)p, and curves were right-shifted and represented single-site binding in its presence. Results shown are the mean data from five independent experiments. Absent error bars denote that standard errors were less than the plotting symbol. (Because these experiments were performed at a higher temperature, the K i values for isoproterenol with Gpp(NH)p are ϳ4-fold higher than those obtained in the experiments shown in Table I  originally reported. Whereas we know nothing regarding genetic variability in these other species, the high degree of consistency in this region, its importance in G-protein coupling, and the nonconservative (size and charge) nature of the Gly to Arg substitution are consistent with this variation having functional consequences. As introduced earlier, ␤ 1 AR are expressed on a number of cell types in the body. In the heart, ␤ 1 AR represent the predominant ␤AR subtype and is expressed on myocytes of the atria and ventricles, where they act to increase the force and frequency of contraction in response to sympathetic stimulation. It is intriguing to consider the potential role of the ␤ 1 AR polymorphisms at position 389 in regulating cardiac function under pathologic conditions. As left ventricular failure develops, ␤ 1 AR expression and function decrease in human heart failure (24,25). This response is thought to be a protective mechanism, sparing the heart from sustained sympathetic stimulation in the face of limited metabolic reserves. In earlier phases of the disease, maintenance of ␤ 1 AR function may contribute to improved ventricular function. Given the above circumstances, the dramatic differences in function between the two ␤ 1 AR polymorphisms suggest that the pathophysiology of congestive heart failure may be influenced by the ␤ 1 AR genotype. Indeed, there is unexplained interindividual variation in the progression of heart failure as well as the cardiac response to infused ␤-agonists used therapeutically during acute decompensation (26 -28). Interestingly, ␤AR antagonists (␤-blockers) are utilized in the chronic treatment of heart failure, the presumed basis of which is minimization of the aforementioned consequences of long term sympathetic stimulation. There is, however, significant interindividual variation in the clinical response to ␤-blockers in patients with heart failure (29). Based on our current results, it might be predicted that individuals bearing the Arg-389 receptor would be most responsive to ␤-blocker therapy because they would have a genetically determined ␤ 1 AR that achieves a greater stimulation of adenylyl cyclase. A similar scenario might be present in the treatment of essential hypertension with ␤-blockers, where responders could be predicted by ␤ 1 AR genotyping. These pharmacogenetic concepts will need to be explicitly tested in human studies. While we are unaware of therapeutic agents specifically targeting adipocyte ␤ 1 ARs, the different functional properties of the two ␤ 1 ARs may be the basis of these polymorphisms acting as disease modifiers in obesity. Finally, we have yet to delineate the frequencies of these ␤ 1 AR polymorphisms in various disease groups. Given that both polymorphisms are common in the general population, however, it is unlikely that either represents the primary basis of disease. These polymorphisms may, however, represent small risk factors in common, multi-factorial diseases such as heart failure, hypertension, and obesity.
In conclusion, we have found polymorphic variation of the human ␤ 1 AR at amino acid 389, where Gly (previously considered wild-type) or Arg can be found. This variation alters receptor-G s coupling, manifested as significant differences between the two receptors in agonist-stimulated adenylyl cyclase activation. Such variation may represent a genetic basis for interindividual differences in disease susceptibility or phenotype, or the response to agents targeting the ␤ 1 AR.