Hierarchy of Polymorphic Variation and Desensitization Permutations Relative to β1- and β2-Adrenergic Receptor Signaling

Abstract

Agonist-promoted desensitization of G-protein-coupled receptors results in partial uncoupling of receptor from cognate G-protein, a process that provides for rapid adaptation to the signaling environment. This property plays important roles in physiologic and pathologic processes as well as therapeutic efficacy. However, coupling is also influenced by polymorphic variation, but therelative impact of these two mechanisms on signal transduction is not known. To determine this we utilized recombinant cells expressing the human β1-adrenergic receptor (β1AR) or a gain-of-function polymorphic variant (β1AR-Arg389), and the β2-adrenergic receptor (β2AR) or a loss-of-function polymorphic receptor (β2AR-Ile164). Adenylyl cyclase activities were determined with multiple permutations of the possible states of the receptor: genotype, basal, or agonist stimulated and with or without agonist pre-exposure. For the β1AR, the enhanced function of the Arg389 receptor underwent less agonist-promoted desensitization compared with its allelic counterpart. Indeed, the effect of polymorphic variation on absolute adenylyl cyclase activities was such that desensitized β1AR-Arg389 signaling was equivalent to non-desensitized wild-type β1AR; that is, the genetic component had as much impact as desensitization on receptor coupling. In contrast, the enhanced signaling of wild-type β2AR underwent less desensitization compared with β2AR-Ile164, thus the heterogeneity in absolute signaling was markedly broadened by this polymorphism. Inverse agonist function was not affected by polymorphisms of either subtype. A general model is proposed whereby up to 10 levels of signaling by G-protein-coupled receptors can be present based on the influences of desensitization and genetic variation on coupling. Agonist-promoted desensitization of G-protein-coupled receptors results in partial uncoupling of receptor from cognate G-protein, a process that provides for rapid adaptation to the signaling environment. This property plays important roles in physiologic and pathologic processes as well as therapeutic efficacy. However, coupling is also influenced by polymorphic variation, but therelative impact of these two mechanisms on signal transduction is not known. To determine this we utilized recombinant cells expressing the human β1-adrenergic receptor (β1AR) or a gain-of-function polymorphic variant (β1AR-Arg389), and the β2-adrenergic receptor (β2AR) or a loss-of-function polymorphic receptor (β2AR-Ile164). Adenylyl cyclase activities were determined with multiple permutations of the possible states of the receptor: genotype, basal, or agonist stimulated and with or without agonist pre-exposure. For the β1AR, the enhanced function of the Arg389 receptor underwent less agonist-promoted desensitization compared with its allelic counterpart. Indeed, the effect of polymorphic variation on absolute adenylyl cyclase activities was such that desensitized β1AR-Arg389 signaling was equivalent to non-desensitized wild-type β1AR; that is, the genetic component had as much impact as desensitization on receptor coupling. In contrast, the enhanced signaling of wild-type β2AR underwent less desensitization compared with β2AR-Ile164, thus the heterogeneity in absolute signaling was markedly broadened by this polymorphism. Inverse agonist function was not affected by polymorphisms of either subtype. A general model is proposed whereby up to 10 levels of signaling by G-protein-coupled receptors can be present based on the influences of desensitization and genetic variation on coupling. β1- and β2-adrenergic receptors, respectively stimulatory guanine nucleotide-binding protein control conditions desensitized conditions basal state of adenylyl cyclase activation isoproterenol stimulated state of adenylyl cyclase activation analysis of variance Like a number of other G-protein-coupled receptors, the β1- and β2-adrenergic receptors (β1AR and β2AR,1respectively) undergo desensitization during continuous exposure to agonist. Such desensitization occurs maximally after several minutes of agonist exposure and is due to decreased interaction with Gs, which is evoked by receptor phosphorylation (1Liggett S.B. Lefkowitz R.J. Sibley D. Houslay M. Regulation of Cellular Signal Transduction Pathways by Desensitization and Amplification. John Wiley & Sons, London1993: 71-97Google Scholar). Thus the signal transduction of these receptors can be characterized as one of two potential conditions or states, defined here as “control” (no recent exposure to agonist) and “desensitized.” However, we have recently shown that an alteration in receptor-Gscoupling can also be imposed by genetic mechanisms. A single nucleotide polymorphism found in the β1AR gene in the human population (2Mason D.A. Moore J.D. Green S.A. Liggett S.B. J. Biol. Chem. 1999; 274: 12670-12674Abstract Full Text Full Text PDF PubMed Scopus (569) Google Scholar, 3Moore J.D. Mason D.A. Green S.A. Hsu J. Liggett S.B. Hum. Mutat. 1999; 14: 271Crossref PubMed Scopus (102) Google Scholar) results in either Gly or Arg being encoded at amino acid position 389 of the proximal portion of the cytoplasmic tail. In studies using transfected cells with equivalent expression of the two receptors, the β1AR-Arg389 displays an increase in Gs coupling compared with β1AR-Gly389 (2Mason D.A. Moore J.D. Green S.A. Liggett S.B. J. Biol. Chem. 1999; 274: 12670-12674Abstract Full Text Full Text PDF PubMed Scopus (569) Google Scholar). So, one can consider that the human β1AR can exist in four agonist-stimulated states: Gly389 control and desensitized, and Arg389 control and desensitized. And, since basal (non-agonist) activity is also affected by these genetic and desensitization processes, eight states can be considered. For the β2AR, a polymorphism that results in a substitution of Ile for Thr at amino acid 164 in the fourth transmembrane-spanning domain results in a decrease in Gs coupling (4Green S.A. Cole G. Jacinto M. Innis M. Liggett S.B. J. Biol. Chem. 1993; 268: 23116-23121Abstract Full Text PDF PubMed Google Scholar). Thus, eight such states can be considered for the β2ARs as well, again based on genotype and desensitization status. As opposed to their allelic variants, only the β1AR-Gly389and the β2AR-Thr164 (the receptors often designated as “wild-type”) have been studied in regards to desensitization in recombinantly expressing cells (5Freedman N.J. Liggett S.B. Drachman D.E. Pei G. Caron M.G. Lefkowitz R.J. J. Biol. Chem. 1995; 270: 17953-17961Abstract Full Text Full Text PDF PubMed Scopus (307) Google Scholar, 6Hausdorff W.P. Bouvier M. O'Dowd B.F. Irons G.P. Caron M.G. Lefkowitz R.J. J. Biol. Chem. 1989; 264: 12657-12665Abstract Full Text PDF PubMed Google Scholar). Based on the significant impact that both desensitization and polymorphic variation have on coupling, we considered that control signaling with one variant might even be equivalent to the desensitized signaling of the other. Since inverse agonists act to lower the frequency of spontaneous activation of βARs, there is the potential for coupling polymorphisms to influence this response as well. Knowing the hierarchy of these states facilitates understanding the molecular basis of receptor response to therapeutic agents and receptor dysfunction that can occur in pathologic states, where both genetic and post-translational modifications occur together. Such studies also provide for a general model that depicts the interaction of genetic and desensitization mechanisms in G-protein-coupled receptor signaling. To investigate this, we expressed these four receptors in Chinese hamster fibroblasts at equivalent levels and studied the relative effects of these genetic modifications and those of short term agonist-promoted desensitization on receptor function. Site-directed mutagenesis was performed on the wild-type cDNA templates as described previously so as to mimic the human Arg389 β1AR and Ile164 β2AR variants (2Mason D.A. Moore J.D. Green S.A. Liggett S.B. J. Biol. Chem. 1999; 274: 12670-12674Abstract Full Text Full Text PDF PubMed Scopus (569) Google Scholar, 4Green S.A. Cole G. Jacinto M. Innis M. Liggett S.B. J. Biol. Chem. 1993; 268: 23116-23121Abstract Full Text PDF PubMed Google Scholar). Wild-type and polymorphic cDNAs were cloned into the mammalian expression vector pBC12B1. CHW-1102 cells were stably transfected by calcium phosphate precipitation. Positive clones were selected based on resistance to 300 μg/ml G418. Cultures were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum at 37 °C, 5% CO2, in 100 μg/ml streptomycin, 100 units/ml penicillin, and 80 μg/ml G418. COS-7 cells were transiently transfected and maintained as described (7McGraw D.W. Forbes S.L. Kramer L.A. Liggett S.B. J. Clin. Invest. 1998; 102: 1927-1932Crossref PubMed Scopus (262) Google Scholar). Confluent layers of CHW cells were washed three times with cold phosphate-buffered saline, lysed in hypotonic buffer (5 mm Tris, 2 mm EDTA, pH 7.4), and mechanically detached with a rubber policeman in a small volume. The particulates were homogenized with a polytron and then centrifuged at 42,000 × g for 10 min. Pellets were then resuspended in 75 mm Tris, 5 mmMgCl2, 2 mm EDTA, pH 7.4. Expression levels were determined in saturation binding assays. Membranes were incubated with 400 pm125I-cyanopindolol and 100 μm GTP for 2 h at room temperature with nonspecific binding determined in the presence of 1 μm propranolol. Reactions were stopped by dilution and vacuum filtration over Whatmann glass fiber filters. The percentage of the receptor pool that is expressed at the cell surface was determined exactly as described previously (8Suzuki T. Nguyen C.T. Nantel F. Bonin H. Valiquette M. Frielle T. Bouvier M. Mol. Pharmacol. 1992; 41: 542-548PubMed Google Scholar). Briefly, cells were homogenized as above, centrifuged at 400 × g for 10 min, and the supernatant layered over a 35% sucrose cushion and centrifuged at 150,000 ×g for 1.5 h. The 0–35% interface (light vesicular membranes) and the pellet (plasma membranes) were collected, diluted in 5 mm Tris, 2 mm EDTA, pH 7.4, and centrifuged at 200,000 × g for 1 h. Radioligand binding with125I-cyanopindolol was then carried out with each fraction as described above. Confluent monolayers of cells were washed twice with Hanks' balanced salt solution and allowed to equilibrate in fresh DMEM for 30 min at 37 °C, 5% CO2. Cells were then incubated with either a 10 μmconcentration of the indicated agonist with 100 μmascorbic acid or with ascorbic acid alone (control) for 20 min, washed five times with cold phosphate-buffered saline, detached, and membranes prepared as above. Membranes were incubated with 30 mmTris, pH 7.4, 2 mm MgCl2, 0.8 mmEDTA, 120 μm ATP, 60 μm GTP, 100 μm cAMP, 2.8 mm phosphoenolpyruvate, 2.2 μg myokinase, the indicated concentrations of agonist, and 1 μCi of [α-32P]ATP for 30 min at 37 °C as described previously (2Mason D.A. Moore J.D. Green S.A. Liggett S.B. J. Biol. Chem. 1999; 274: 12670-12674Abstract Full Text Full Text PDF PubMed Scopus (569) Google Scholar). The stop buffer contained a [3H]cAMP standard, which accounted for column recovery. [32P]cAMP was separated from [α-32P]ATP chromatographically using alumina columns. Untreated (control) cells bearing the two β1ARs are designated Arg389C and Gly389C, while those studied after treatment with a 10 μm concentration of the agonist epinephrine (desensitized) are designated Arg389D and Gly389D. Similarly, control β2AR are designated Thr164C and Ile164C, while the desensitized state has the subscript “D.” An additional qualifier, based on whether the adenylyl cyclase response is in the absence of agonist (basal, B) or in response to isoproterenol (I), provides for eight different permutations. To ascertain the effects of inverse agonists, COS-7 cells were transfected with the indicated receptors and Gαs. At confluence, cells in 24-well plates were washed and incubated with 100 μm isobutylmethylxanthine with or without varying concentrations of inverse agonists for 45 min. cAMP produced over this time was quantitated by a competitive immunoassay (AmershamBiosciences). Protein concentrations were determined by the copper bicinchoninic acid method (9Smith P.K. Krohn R.I. Hermanson G.T. Mallia A.K. Gartner F.H. Provenzano M.D. Fujimoto E.K. Goeke N.M. Olson B.J. Klenk D.C. Anal. Biochem. 1985; 150: 76-85Crossref PubMed Scopus (18713) Google Scholar). Curve fitting was carried out with PRISM software (GraphPad, San Diego, CA). Dose-response curves were compared by ANOVA with post-hoc t tests when thep value was <0.05. Other results were compared with pairedt tests as indicated. Data are presented as means ± S.E. Expression levels of the two β1AR variants in the membrane preparations utilized for the adenylyl cyclase assays were 206 ± 16 for Gly389 and 170 ± 22 fmol/mg for Arg389. Likewise, the two β2AR cell lines had similar levels of expression (Thr164 = 783 ± 88, Ile164 = 1104 ± 111 fmol/mg). Of note, signaling characteristics were compared between the two β1AR variants or between the two β2AR variants, but not between subtypes. There was no evidence for relevant intracellular accumulation of either polymorphic variant as determined by radioligand binding of light vesicular membrane and plasma membrane fractions (Table I). Our initial goal was to assess the degree of agonist-promoted desensitization for the wild-type β1AR (Gly389) and the Arg389polymorphic receptor, and the wild-type β2AR (Thr164) and its variant, Ile164. Concerning the β1ARs, we knew from previous studies that basal and agonist-stimulated adenylyl cyclase activities (in absolute values) were higher for the Arg389 β1AR compared with the Gly389 receptor. For the current work, cells in culture were exposed to vehicle or agonist for 20 min, washed, membranes prepared, and adenylyl cyclase activities determined. When desensitization is quantitated as the percent decrease of the response relative to that in the absence of agonist pretreatment, the β1AR-Gly389 underwent 21 ± 6.7% agonist-promoted desensitization (Fig.1 a, Table I). No change in the EC50 was observed. The hyperfunctional β1AR-Arg389 underwent a greater degree of desensitization compared with its allelic variant, amounting to 34 ± 4.1% desensitization (p < 0.01 versusβ1AR-Gly389, Fig. 1 b). For the β2ARs, we also found that the two polymorphic variants differed in the extent of agonist-promoted desensitization (Fig. 1,c and d). However, in contrast to what was observed with the β1AR, the hyperfunctional β2AR-Thr164 actually underwent decreased desensitization. β2AR-Thr164 displayed 26 ± 4.0% desensitization versus 37 ± 4.6% found for Ile164-β2AR (p < 0.05).Table IAdenylyl cyclase activation under control and desensitized conditions for the polymorphic β1- and β2-adrenergic receptorsParametersβ1-Gly389β1-Arg389β2-Thr164β2-IIe164Expression, fmol/mg206 ± 16170 ± 22783 ± 881104 ± 11(% cell surface)(80 ± 2.1)(88 ± 3.5)(85 ± 1.7)(83 ± 2.1)Adenylyl cyclase, pmol/min/mgBasalC3.7 ± 0.618.2 ± 1.14.4 ± 0.892.0 ± 0.25BasalD2.5 ± 0.435.5 ± 0.743.8 ± 0.711.9 ± 0.54IsoC15.1 ± 2.221.8 ± 2.817.1 ± 0.458.9 ± 0.45IsoD11.9 ± 2.314.9 ± 2.213.1 ± 0.566.4 ± 1.29pK actpK actC2.28 ± 0.172.34 ± 0.191.38 ± 0.221.48 ± 0.09pK actD2.32 ± 0.232.50 ± 0.191.08 ± 0.141.28 ± 0.20 Open table in a new tab Although the above data examine the extent of desensitization as a percentage of the control response, the absolute levels of adenylyl cyclase activities (pmol/min/mg) establish a hierarchy of signal transduction based on genotype and desensitization. For the β1AR, these data are shown in Fig.2 a. As is seen, the influence of genetic variation was such that even after desensitization, the maximal Arg389 receptor function (ArgDI) was equivalent to the maximal non-desensitized Gly389 variant (GlyCI). The rank order of activities for the various states for the β1AR are: ArgCI = For the β2AR, since the Ile164 receptor underwent an even greater degree of desensitization the wild-type (Fig. the heterogeneity in adenylyl cyclase activities due to the various permutations was The eight states are shown in Fig. 2 The rank order was = also whether the polymorphic affected the response to inverse considered that the effects of these might the receptor so that not be decreased spontaneous by the binding of inverse were to results in CHW due to the levels of basal adenylyl cyclase activity and the expression levels in the Thus COS-7 cells were transfected with the various receptors with exposed to the and cell cAMP over the 30 min determined in the absence or presence of various concentrations of the inverse agonists (β1AR) or is the absolute levels of cAMP that are relevant a percent are shown in The basal levels of cAMP were greater for β2AR compared with β2AR as might have been from the membrane a similar between the two β1ARs, which also might be was not observed. This be spontaneous activation is less with β1AR Mol. Pharmacol. Google and in the two variants not be as the to inverse agonists for the β1ARs, exposure to in a decrease in cAMP for both The response was for with cAMP levels of ± 0.8 and ± for Arg389 and Gly389 = Similarly, for the β2ARs the inverse agonist cAMP to the levels for both receptors ± Ile164 = ± = results that the to inverse agonists are not influenced by these genetic The of signaling by G-protein-coupled receptors has been considered of the of these receptors to to in their signaling environment. such as receptor desensitization are for of during be in states, or signaling and to characteristics of states M. 1999; PubMed Scopus Google Scholar). Desensitization also the of agonists recombinantly expressed receptors, the of rapid agonist-promoted desensitization of G-protein-coupled receptors has been receptors, such as the and the or no rapid desensitization M.G. Liggett S.B. J. Biol. Chem. 1992; Full Text PDF PubMed Google Scholar, S.B. N.J. D.A. Lefkowitz R.J. 1993; PubMed Scopus Google Scholar). a of desensitization to various mechanisms phosphorylation by G-protein-coupled receptor or such as protein or protein Such studies can be in or understanding selected of the recent of polymorphic variants of receptors such as the β1- and β2AR, which have significant impact on receptor coupling, the potential interaction between desensitization and genetic variation to be considered so as to establish receptor signaling is influenced by both we have carried out studies with polymorphic β1ARs (2Mason D.A. Moore J.D. Green S.A. Liggett S.B. J. Biol. Chem. 1999; 274: 12670-12674Abstract Full Text Full Text PDF PubMed Scopus (569) Google which have either Gly or Arg at amino acid This is in the cytoplasmic portion of the receptor, a between the transmembrane-spanning domain and the T. T. H. D.C. T. M. M. PubMed Scopus Google Scholar). the of Gly and the of this for is not that coupling is different between the Arg and Gly β1AR The β2AR polymorphism (4Green S.A. Cole G. Jacinto M. Innis M. Liggett S.B. J. Biol. Chem. 1993; 268: 23116-23121Abstract Full Text PDF PubMed Google Scholar) of a substitution of Ile for Thr in the fourth transmembrane-spanning and also displays coupling to Gs, due to in the agonist binding that the of the intracellular A was not whether these polymorphisms or agonist desensitization. For the signaling of β1AR-Arg389 might less to since its is for Gs coupling, be to G-protein-coupled receptor which is on the receptor being in the here that there is a significant impact of these polymorphisms on agonist-promoted desensitization. In the of β1AR the desensitized hyperfunctional variant is equivalent to that of the non-desensitized Since the basal of which spontaneous to is also the number of into basal or agonist non-desensitized or desensitized, and two polymorphic variants, for the β1AR to For the β1AR this is displayed in with the states in Fig. which is for a of the of genetic variation in G-protein-coupled receptor signaling As is there is variation in basal and agonist-stimulated activities by genotype and desensitization status. Such a of relative signaling and the between desensitization and genotype, the degree of in physiologic to agonists and that has been observed in S.B. PubMed Scopus Google Scholar). An additional of signaling can also be found when one the response to inverse agonists, which receptor and to the and thus spontaneous to As receptors a decrease in basal adenylyl cyclase on the of the the response to inverse agonist also be affected by genetic the other effects of these in inverse agonist were not observed with either the β1AR or β2AR variants, as cAMP levels were by inverse agonists to the absolute levels of are not into Fig. since were determined using a different and are not For a general model (Fig. of the potential interaction of uncoupling due to genetic variation and to desensitization, we have several to polymorphisms with various effects on receptor signaling. of agonist, receptor, and of the polymorphic (2Mason D.A. Moore J.D. Green S.A. Liggett S.B. J. Biol. Chem. 1999; 274: 12670-12674Abstract Full Text Full Text PDF PubMed Scopus (569) Google Scholar, 4Green S.A. Cole G. Jacinto M. Innis M. Liggett S.B. J. Biol. Chem. 1993; 268: 23116-23121Abstract Full Text PDF PubMed Google we have that the of the receptor is by the polymorphisms under thus the of two based on the two different receptors and is also that when desensitized the of the receptor is by phosphorylation and interaction Pharmacol. 1998; PubMed Scopus Google and these states are thus as and The signaling of a receptor, due to spontaneous to an is also considered here as as is the effect of desensitization on basal signaling. not affected with the β1AR or β2AR, the potential for a coupling polymorphism to inverse agonist activity is in the general The of each of the above at is indicated by the and the this general model 10 relevant levels of due to the various states or of a are present at Of note, other of signaling such as receptor that is not are not the of in the their has been to a of states in such as In human β1AR and β2AR have been shown to be desensitized. with receptor this response is to be in that the with physiologic and is from by the other other studies in have that of desensitization of signaling be in S.B. J. Clin. Invest. PubMed Scopus Google Scholar). studies have shown that β1AR or β2AR polymorphisms are with physiologic or pathologic in human S.B. J. Clin. Invest. 1998; 102: PubMed Scopus Google Scholar, Liggett S.B. PubMed Scopus Google Scholar, D.A. Liggett S.B. J. Full Text PDF PubMed Scopus Google Scholar, Liggett S.B. J. PubMed Scopus Google Scholar). However, to the current has not been whether there was or of desensitization by these The in presented here that both desensitization and genetic variation can to the of signaling of β1AR and Indeed, the signaling of receptors, even in the desensitized is equivalent to their non-desensitized allelic genetic variation can have an effect of the as that of desensitization. in this be important in therapeutic or in the of new agents S.B. PubMed Scopus Google Scholar, S.B. PubMed Scopus Google Scholar). As a general we present a model by which genetic variation and desensitization of G-protein-coupled receptor signaling can be considered as multiple Although the of polymorphisms the is not defined Liggett S.B. PubMed Scopus Google G-protein-coupled receptors have been to be polymorphic in their T. Pharmacol. 41: PubMed Scopus Google Scholar, D.A. Liggett S.B. J. PubMed Scopus Google such that the model be to multiple signaling by these for cell culture and for