GENETIC VARIATION IN THE β₂-ADRENERGIC RECEPTOR: IMPACT ON INTERMEDIATE CARDIOVASCULAR PHENOTYPES

Abstract

Genetic variation in drug targets (e.g. receptors) can have pronounced effects on clinical responses to endogenous and exogenous agonists. Polymorphisms in the gene encoding the β₂-adrenergic receptor (β₂-AR) have been associated with altered expression, down-regulation, and altered cell signaling in vitro. Because β₂-ARs play a crucial role in the regulation of the cardiovascular system, the functional importance of genetic variation in the β₂-AR on cardiovascular responses to physiological or pharmacological stimuli has gained widespread attention. The objective of this review is to characterize these intermediate cardiovascular phenotypes and their influence on cardiovascular disease and adrenergic drug responses.

Two common single nucleotide polymorphisms, encoded at codon 46 (Gly¹⁶Arg) and 79 (Gln²⁷Glu) of the β₂-AR gene, have been studied intensively. They have been shown to be associated with altered vasodilator responses to regional and systemic administration of β₂-agonists, altered cardiovascular responses to sympathoexcitatory maneuvers, and altered myocardial function. Importantly, these intermediate physiological patterns may influence the development of and the outcomes associated with hypertension and other cardiovascular diseases. As recently reported, β₂-AR gene variation can risk-stratify patients receiving β-blocker therapy and may predict β-blocker efficacy in patients post acute coronary syndrome or in patients with heart failure.

Further studies will advance our understanding of the link between β₂-AR genotypes, intermediate cardiovascular phenotypes, and clinical phenotypes. In the long term, reassessment of the benefits of β-blocker-therapy within genotype groups should be carried out with the ultimate goal to design the optimal therapeutic regimen for the individual patient.

INTRODUCTION

It is a common clinical problem that the same drug leads to variable responses in different patients. Modern clinical trials often average the treatment effect for thousands of individuals, summarized by a single statistical value; however, important individual differences are lost, and the drug effect may even be dangerous in some patients [Kent, D. and Hayward, R. 2007]. Although many non-genetic factors – including age, organ function, concomitant therapy / drug interactions, and the nature of the disease – may influence drug effects, there is increasing evidence that inter-individual differences in drug response may also be caused by sequence variants in genes encoding drug-metabolizing enzymes, drug transporters, or the drug targets themselves [Evans, W. E. and McLeod, H. L. 2003]. It is estimated that genetic variation can account for 20 to 95 percent of the variability in drug disposition and effects [Kalow, W. et al. 1998].

The ultimate promise of pharmacogenetics is the possibility that knowledge of a patient’s DNA might be used to maximize drug efficacy, to give drugs only to those patients that are likely to respond, and to predict risks and thus avoid adverse drug reactions [Weinshilboum, R. and Wang, L. 2004]. Pharmacogenetic research began with a focus on drug metabolism, and after years of fundamental research based on understanding the functional consequences of polymorphisms in genes encoding drug-metabolizing enzymes, individualized therapy using genetic analyses is entering clinical practice. Examples include testing for lack of the enzyme thiopurine S-methyltransferase (TPMT) to prevent severe toxicity to thiopurine drugs [Weinshilboum, R. and Wang, L. 2004], or testing for CYP2D6 deficiency to determine tamoxifen’s effectiveness in adjuvant breast cancer therapy [Borges, S. et al. 2006].

Interest in drug target genetic variation and the potential for pronounced effects on endogenous and exogenous agonist responsiveness [Evans, W. E. and McLeod, H. L. 2003] is relatively recent when viewed within the longer history of pharmacogenetic research. Because of the pivotal role in the regulation of cardiovascular function, genetic variation in the β₂-adrenergic receptor (β₂-AR) has been studied intensively during the last decade. From the initial differentiation of adrenergic receptors in 1948 to the description of the β₂-AR gene sequence in 1987, the “pre-genomic” era was likely under the assumption that β₂-AR’s are similar in all individuals (Fig. 1). In 1993, Green et al. published the first in-vitro functional characterization of a naturally occurring variant in the human β₂-AR, showing that it alters ligand binding and the functional properties of the receptor [Green, S. A. et al. 1993]. In these formative stages of the “genomic era,” candidate gene approaches determined that the β₂-AR gene may be associated with hypertension [Kotanko, P. et al. 1997], but over time incrimination of specific polymorphisms has been inconclusive probably due to the multifactorial and polygenic nature of hypertension. Nevertheless, the relationship between β₂-AR gene variation and β₂-AR function became a fundamental physiologic question, and “physiological bridge building” strategies emerged as a means to determine whether variation in the β₂-AR gene influences intermediate physiological traits relevant to the pathogenesis of cardiovascular disease.

Figure 1.

Historical highlights in β₂-adrenergic receptor research.

The objective of this review is to summarize and discuss studies that focus on characterizing these intermediate phenotypes and on exploring their influence on cardiovascular disease and adrenergic drug response. Growing evidence suggests that β₂-AR gene variation influences blood pressure regulation, adrenergic drug response, and clinical outcome, and the implications from these findings may be important in achieving the ultimate goal to individualize drug therapy based on patient’s genotype.

β₂-AR GENE VARIATION

β₂-ARs are expressed in numerous tissues including vascular and bronchial smooth muscle cells, kidneys and the heart [Kirstein, S. L. and Insel, P. A. 2004]. The gene encoding the β₂-AR is intronless, located on chromosome 5q31-32 and contains several polymorphic sites (Fig. 2): At least 51 variants have been identified so far across the span of 5.3 kilobases [Hawkins, G. A. et al. 2006; Kirstein, S. L. and Insel, P. A. 2004]. Five of these are non-synonymous single nucleotide substitutions in the β₂-AR coding region, occurring at positions 46 (extracellular amino terminus; Gly¹⁶Arg), 79 (extracellular amino terminus; Gln²⁷Glu), 100 (1^st transmembrane-spanning domain; Val³⁴Met), 491 (4^th domain; Thr¹⁶⁴Ile), and 659 (5^th domain; Ser²²⁰Cys). The single nucleotide polymorphisms (SNPs) at position 46 (Gly¹⁶Arg) and 79 (Gln²⁷Glu) are the most common, with minor allele frequencies in the white population around 0.4 (Arg¹⁶, Glu²⁷), whereas the Thr¹⁶⁴Ile SNP is rare (minor allele frequency around 0.02) and the Val³⁴Met and Ser²²⁰Cys variants are extremely rare [Hawkins, G. A. et al. 2006; Kirstein, S. L. and Insel, P. A. 2004]. Another common SNP (Cys^-19Arg) at position -47 is within a short open reading frame, the 5’-leader cistron, which encodes a signaling peptide involved in regulation of β₂-AR translation [McGraw, D. W. et al. 1998; Scott, M. G. et al. 1999]; the minor allele frequency for this SNP is 0.44 [Hawkins, G. A. et al. 2006].

Figure 2.

Localization of polymorphisms in the human β₂-adrenergic receptor (β₂-AR). The β₂-AR is a 413 amino-acid long member of the “superfamily” of adrenergic receptors, with seven membrane-spanning regions. Common single nucleotide polymorphisms (SNPs) include a change at nucleotide position 46 which results results in the amino acid sequence of the β₂-AR containing a glycine (Gly) at position 16 instead of an arginine (Arg), and a change at nucleotide position 79 which results in glutamine (Gln) instead of glutamate (Glu) at amino acid position 27. Variation at amino acid position 164 has the most significant cardiovascular consequence; however, the isoleucine (Ile) variant is rare, and the homozygous form is not present in humans. From [McNamara, D. M. et al. 2002].

The polymorphisms are in strong linkage disequilibrium. Drysdale et al. have resequenced the β₂-AR in order to evaluate gene variation and haplotype structure [Drysdale, C. M. et al. 2000]. They selected 13 polymorphic sites in the promoter and coding region, which mathematically would produce 8192 inherited combinations in humans (see Fig. 3). However, only 12 haplotypes were found in their population. It is postulated that the reason behind so few numbers of chromosomally phased SNPs is that the human species is relatively young, and therefore combinations of individual polymorphisms are in linkage disequilibrium. For example, Glu²⁷ homozygotes are nearly uniformly homozygous for Gly¹⁶ (haplotype 2 in Fig. 3), and this haplotype is the only one that encodes the Arg variant at position -47 in the 5’-leader cistron [Drysdale 2000; Hawkins 2006]. Further, Arg¹⁶ homozygotes are nearly uniformly Gln²⁷ homozygotes (haplotype 4 in Fig. 3).

Figure 3.

Frequency of common haplotypes of the β₂-AR in Whites and African Americans. The figure shows the haplotypes defined by Drysdale et al. [Drysdale, C. M. et al. 2000] and their frequencies in Whites and African Americans as observed by Hawkins et al. [Hawkins, G. A. et al. 2006]. Only haplotypes with a frequency >0.01 are shown. Only 3 haplotypes (#2, #4, and #6) occur with frequencies >5% in the white population. Bold letters indicate tagging SNPs often used in association studies.

Significant differences in distribution of some haplotypes were noted in Caucasian, African-American, Asian, and Hispanic-Latino ethnic groups with >20-fold differences among the frequencies of the four major haplotypes [Drysdale, C. M. et al. 2000]. In a larger study by Hawkins et al., 31 polymorphisms from sequencing data from 429 whites and 240 African Americans were used to estimate haplotype frequencies, and a total of 24 haplotypes were observed [Hawkins, G. A. et al. 2006]. When analysis was limited to the polymorphisms studied by Drysdale et al. [Drysdale, C. M. et al. 2000], only haplotypes 2, 4, 6, 7, and 10 were observed in whites, whereas only haplotypes 1, 2, 4, 6, and 9, were observed in African Americans. Frequencies of haplotype 2 and 4 were higher in whites, whereas haplotype 6 was more frequent in African Americans (Fig. 3).

IN-VITRO STUDIES

In order to study the functional consequences of these polymorphisms, in vitro experiments in recombinant systems have been performed (for review see [Brodde and Leineweber 2005]). It was found that neither agonist binding nor G-protein coupling, resulting in stimulation of adenylyl cyclase activity, was altered by Gly¹⁶Arg or Gln²⁷Glu polymorphisms. However, altered degrees of agonist-promoted down-regulation of receptor expression were described [Green, S. A. et al. 1994]. Specifically, the substitution of glycine for arginine at position 16 (Gly¹⁶) was associated with enhanced agonist-induced desensitization as compared to Arg¹⁶, and the substitution of glutamic acid for glutamine at position 27 (Glu²⁷) was associated with resistance to desensitization [Green, S. A. et al. 1995]. The resistance of the Glu²⁷ receptor variant to agonist-promoted down-regulation has recently been confirmed by performing cell-signaling studies in two HEK293 cell lines over-expressing similar levels of the Gln²⁷ or Glu²⁷ variant: The Glu²⁷ variant magnified the catecholamine-induced activation of ERK and p38, two mitogen-activated protein kinases that have been involved in myocyte hypertrophy [Iaccarino, G. et al. 2006].

In contrast to the two common polymorphisms, the rare Thr¹⁶⁴Ile polymorphism does alter ligand binding and G-protein coupling: In cells transfected with cDNAs that mimic this SNP, the Ile¹⁶⁴ receptor displayed a lower binding affinity for epinephrine, and functional coupling to G_s - as determined in adenylyl cyclase assays - was significantly depressed as compared to the wild-type receptor [Green, S. A. et al. 1993]. The 5’- leader cistron polymorphism Cys^-19Arg leads to altered promoter activity in recombinant cells, with reduced β₂-AR expression [McGraw, D. W. et al. 1998] and β₂-AR promoter-driven luciferase activity [Johnatty, S. E. et al. 2002] of the Arg^-19 variant as compared to the wild type.

IN-VIVO STUDIES

Vascular Function (Local Infusion Studies)

The main effect of β₂-ARs in the vasculature is to evoke marked vasodilation. For many years the cAMP pathway in vascular smooth muscle cells was considered to be the main dilating pathway activated by the β₂-ARs in skeletal muscle vessels. More recently it has been shown that β₂-ARs can evoke nitric oxide (NO) release from the vascular endothelium, and this mechanism likely contributes (~30-50%) to the dilator responses to both β₂-agonist infusions and mental stress [Cardillo, C. et al. 1997; Cardillo, C. et al. 1997; Eisenach, J. H. et al. 2002]. Several studies using local infusions of β-agonists into the brachial artery have demonstrated differences in vascular response dependent on β₂-AR genotype. Cockroft et al. found greater arterial vasodilator responses in Glu²⁷ versus Gln²⁷ homozygotes, and a trend toward greater responses in the Gly¹⁶ versus Arg¹⁶ group [Cockcroft, J. R. et al. 2000]. A larger study at our institution controlled for normal dietary sodium intake for 5 days (150 mmol/day), and observed greater arterial vasodilator responses in the Gly¹⁶ homozygotes compared to the Arg¹⁶ homozygotes [Garovic, V. D. et al. 2003]. After inhibition of NO synthase with L-NMMA these differences in vasodilator responsiveness were no longer apparent (see Fig. 4), suggesting that the augmented vasodilator responses seen in the Gly¹⁶ subjects are mainly due to β₂-mediated release of NO from the vascular endothelium [Garovic, V. D. et al. 2003].

Figure 4.

Forearm blood flow responses to graded infusion of isoproterenol in subjects homozygous for the β₂-AR Arg¹⁶ variant (n=18) or the Gly¹⁶ variant (n=23). Gly¹⁶ homozygotes demonstrated significantly greater forearm blood flow responses to isoproterenol than Arg¹⁶ homozygotes (P=0.02; upper panel). After inhibition of endothelial nitric oxide synthase with L-NMMA, the forearm blood flow responses to isoproterenol were significantly reduced in both the Arg¹⁶ and Gly¹⁶ homozygotes, and they no longer differed significantly between the genotype groups (P=0.27; middle panel). Consequently, the nitric oxide-dependent component of the vasodilator response to isoproterenol, calculated by subtracting the post-L-NMMA responses from the pre-L-NMMA responses, was significantly greater in Gly¹⁶ than Arg¹⁶ homozygotes (P=0.04; bottom panel). P_ANOVA is the P value for the contrast between genotypes from the repeated measures ANOVA; at each dose of isoproterenol, statistically significant differences between genotype groups are denoted by: *P<0.05; **P<0.01. From [Garovic, V. D. et al. 2003].

In isolated hand vein experiments, Dishy et al. observed that the maximal venodilatory responses to isoproterenol were greater in Gly¹⁶ than Arg¹⁶ homozygotes [Dishy, V. et al. 2001]. However, the difference was attributed to effects of the Glu²⁷ variant rather than the Gly¹⁶ variant, because greater responses were present only in Gly¹⁶+Glu²⁷ homozygotes, not in Gly¹⁶+Gln²⁷ homozygotes, when compared to Arg¹⁶+Gln²⁷ homozygotes. In the same study, β₂-agonist-mediated desensitization in the vasculature was tested and interestingly, the findings differed from those obtained in vitro by Green et al. [Green, S. A. et al. 1995]: Namely, in healthy individuals, homozygosity for the Arg¹⁶ was associated with rapid agonist-mediated vascular desensitization, whereas homozygosity for the Glu²⁷ polymorphism was associated with enhanced agonist-mediated vasodilation [Dishy, V. et al. 2001]. Along these lines, Bruck et al. reported that terbutaline-induced venodilation in the dorsal hand vein following 2 weeks of oral terbutaline treatment evoked the greatest desensitization in the Arg¹⁶ homozygotes [Bruck, H. et al. 2005]. Thus the Arg¹⁶ variant of the β₂-AR could be considered a loss-of-function mutation because of the shorter duration of stimulation, and the Glu²⁷ variant could be considered a gain-of-function mutation because of the longer duration of stimulation.

The differences between in vitro and in vivo data may be explained by a dynamic model of receptor regulation, where receptor variants that are highly sensitive in vitro are ‘pre-desensitized’ in vivo by chronic exposure to endogenous agonists, e.g. catecholamines. These receptors may not become further desensitized by the challenge with an exogenous agonist, thereby producing an apparently paradoxical response [Liggett, S. B. 2002].

Vascular Function (Systemic Infusion Studies)

In contrast to the local infusion studies, results from systemic infusion studies of β₂-agonists in vivo paint a picture that is generally but not uniformly consistent with the cell signaling studies, with greater systemic vasodilation seen in Arg¹⁶ than Gly¹⁶ homozygotes [Gratze, G. et al. 1999; Hoit, B. D. et al. 2000; Snapir, A. et al. 2003] or no difference between genotype groups [Bruck, H. et al. 2003; Eisenach, J. H. et al. 2006]. Additionally, inhaled β₂-agonist produced a greater decrease in diastolic blood pressure in asthmatics homozygous for Arg¹⁶+Gln²⁷ [Lee, D. K. et al. 2004]. However, the methods used to assess cardiac output and peripheral vasodilation in these studies have significant limitations, as no efforts were made to account for the potentially confounding effects of cardiovascular baroreflexes on the hemodynamic responses to the infusions [Shannon, J. R. et al. 1998; Wilkins, B. W. et al. 2007]. Because the response to a vasoactive drug is mediated by the net effect of vascular sensitivity to the drug and the counter-regulatory baroreflex actions, the contrasting results from local versus systemic infusion studies may be explained by counter-regulatory baroreflex activation, (over-)compensating for augmented β₂-mediated vasodilation in carriers of the Gly¹⁶ or Glu²⁷ variant. Further studies – e.g. during disengagement of confounding compensatory baroreflex activity – will be needed to address this hypothesis.

Another group has focused on studying the time course of desensitization. They found that the Arg¹⁶Gly and the Gln²⁷Glu variants of the β₂-AR do not alter the extent of terbutaline treatment-induced desensitization of cardiac β₂-AR-mediated increases in heart rate and contractility. However, subjects who were homozygous for Glu²⁷ exhibited a slower time course of desensitization compared to subjects who were homozygous for Gly¹⁶ or homozygous for Arg¹⁶+Gln²⁷ [Bruck, H. et al. 2003]. This idea is generally consistent with studies suggesting that the Glu²⁷ variant is associated with resistance to desensitization and thus enhanced responsiveness, and larger group sizes are needed to further characterize the Glu²⁷ interaction with the polymorphism at position 16.

Cardiac Function

Several studies have determined the influence of β₂-AR gene variation and ventricular function [Eisenach, J. H. et al. 2005; Iaccarino, G. et al. 2002; Tang, W. et al. 2003]. In an echocardiographic analysis of a biethnic sample of normotensives, the Gly¹⁶ homozygotes displayed greater fractional shortening, ejection fraction, midwall shortening, and stress-corrected midwall shortening compared to the heterozygous and Arg¹⁶ homozygous groups, independent of age, sex, ethnicity, heart rate, body mass index, systolic blood pressure, LV end-diastolic dimension, and field center [Tang, W. et al. 2003]. These genotype-dependent differences in resting cardiovascular function have been confirmed in studies using other techniques to evaluate ventricular function: Individuals homozygous for Gly¹⁶ have a greater cardiac output (assessed via the open-circuit acetylene wash-in method) and stroke volume compared to Arg¹⁶ homozygotes at rest. Furthermore, during both low- and high intensity exercise, the Gly¹⁶ group continues to have a greater stroke volume and cardiac output compared to Arg¹⁶ [Snyder, E. M. et al. 2006]. These findings suggest that the Arg¹⁶ genotype is associated with reduced baseline receptor function or density compared to the Gly¹⁶ genotype, and that these baseline differences contribute to the observed differences during exercise.

The Gly¹⁶Arg polymorphism also makes a difference in patients with heart disease: A collaborative group recently performed echocardiography in 95 heart failure patients (EF ≤ 40%) and found that Arg¹⁶ homozygotes had higher plasma norepinephrine and atrial natriuretic peptide levels, greater left atrial diastolic dimension, higher peak velocity of early/late diastolic filling ratio, shorter deceleration time, and reduced exercise tolerance when compared to Gly¹⁶ homozygotes [Wolk, R. et al. 2007].

Response to Sympathoexcitation

Laboratory-based measures of cardiovascular responses to sympathoexcitation provide intermediate physiologic characteristics of cardiac and vascular function that are useful determinants of broader phenotypes in humans, such as hypertension. Normotensive individuals who show a robust pressor response to sympathoexcitatory stimuli like mental stress are at increased risk for developing hypertension [Cardillo, C. et al. 1998; Flaa, A. et al. 2006; Matthews, K. A. et al. 2004; Murphy, J. K. et al. 1994; Murphy, J. K. and McGarvey, S. T. 1994; Treiber, F. A. et al. 1994]. Furthermore, increased job strain has now been shown to be predictive of future incidence of hypertension [Markovitz, J. H. et al. 2004; Matthews, K. A. et al. 2004]. In this context, adrenergic receptors play an important role in the arterial blood pressure responses to mental stress and other sympathoexcitatory maneuvers like isometric handgrip to fatigue [Eisenach, J. H. et al. 2005; Eisenach, J. H. et al. 2004; Freyschuss, U. et al. 1988; Halliwill, J. R. et al. 1997; Hjemdahl, P. 2002]. For example, during mental stress some vascular beds undergo β₂-AR mediated vasodilation while others undergo α-adrenergic vasoconstriction, and the balance between these responses probably contributes to the magnitude of the overall rise in blood pressure [Eisenach, J. H. et al. 2005; Eisenach, J. H. et al. 2004; Freyschuss, U. et al. 1988; Halliwill, J. R. et al. 1997; Hjemdahl, P. 2002]. This means that if β₂-receptor mediated dilation is blunted, the pressor response to mental stress is likely to be increased. Consistent with these general ideas are a number of observations primarily from the pre-genomic era: 1) The pressor response to mental stress is heritable [Busjahn, A. et al. 2000; Carmelli, D. et al. 1985; Carmelli, D. et al. 1991; Li, G. H. et al. 2001; McCaffery, J. M. et al. 2002]; 2) β₂-mediated forearm vasodilator responses to mental stress are blunted in subjects thought to be at increased risk for hypertension and in individuals with mild hypertension [Cardillo, C. et al. 1998; Cardillo, C. et al. 1998; Hughes, J. W. et al. 2003]; 3) Vasodilator responses to brachial artery, or systemic infusions of β-agonists are blunted in subjects thought to be at increased risk for hypertension, and in mild hypertension [Dimsdale, J. et al. 1988; Feldman, R. D. 1987; Feldman, R. D. 1990; Ford, G. A. et al. 1992; Halliwill, J. R. et al. 1997; Lang, C. C. et al. 1995; Sherwood, A. and Hinderliter, A. L. 1993; Stein, C. M. et al. 1995; Watkins, L. L. et al. 1995].

During emotional stress there is an increase in forearm blood flow [Abramson, D. I. F., E.B. 1940; Grant, R. T. P., R.S.B. 1938], and this reaction in humans is similar to the defense reaction in animals [Bulbring, E. and Burn, J. H. 1935; Folkow, B. et al. 1948]. Studies have demonstrated that the most likely mechanisms involved in the vasodilator responses to sympathoexcitatory maneuvers including mental stress center around sympathetic withdrawal, local release of NO, and β₂-mediated stimulation (via circulating epinephrine) acting on receptors on the vascular endothelium and smooth muscle [Dietz, N. M. et al. 1994; Halliwill, J. R. et al. 1997; Joyner, M. J. and Halliwill, J. R. 2000; Liu, Z. et al. 2006; Reed, A. S. et al. 2000]. As a result of these multiple mechanisms, the pressor response and forearm vasodilator response to mental stress demonstrates marked inter-individual variability. The key questions are, what is the role adrenergic receptor gene variation on these responses, and will gene variation ultimately influence the cardiovascular phenotypes?

The contribution of the β₂-AR polymorphisms to differences in the pressor response to certain sympathoexcitatory maneuvers has been evaluated, and in comparison to other polymorphic sites in the β₂-AR gene, the Arg¹⁶Gly polymorphism has been suggested to exert a dominant effect [Busjahn, A. et al. 2000]. An analysis of German Caucasian twins determined that the Arg¹⁶Gly polymorphism was associated with systolic and diastolic blood pressure at rest, during mental arithmetic, and during the cold pressor test, as well as the increase in diastolic pressure during both maneuvers, which was higher in the Arg¹⁶ homozygotes [Li, G. H. et al. 2001]. The Pittsburgh Twin Study revealed a higher resting diastolic BP in Gly¹⁶ homozygotes, but no association of the Arg¹⁶Gly polymorphism with cardiovascular responses to mental stress [McCaffery, J. M. et al. 2002]. The absent genotype effect on the pressor response to mental stress in the latter study is not readily explained.

Regarding regional vasodilator differences during sympathoexcitation, a study of Brazilian women found that subjects homozygous for Gly¹⁶ and Glu²⁷ demonstrated a greater forearm vasodilator response to mental stress and to isometric handgrip, an effect that was no longer present following brachial artery infusion of a β-blocker [Trombetta, I. C. et al. 2005]. These findings were attributed to depend on position 27, as Gly¹⁶+Glu²⁷ subjects had a greater FBF response than Arg¹⁶+Gln²⁷ and Gly¹⁶+Gln²⁷ subjects. Two studies at our institution confirmed genotype-dependent differences in response to handgrip; individuals homozygous for Gly¹⁶ had greater increases in heart rate and cardiac output, and a tendency for lower systemic vascular resistance [Eisenach, J. H. et al. 2005; Eisenach, J. H. et al. 2004]. The higher heart rate and cardiac output response during handgrip in the Gly¹⁶ homozygotes may be due to greater ß₂-AR-mediated vasodilator effect was evoked by circulating catecholamines in Gly¹⁶ homozygotes. This means that to achieve the same pressor response, greater increases in heart rate were needed in the Gly¹⁶ subjects. This interpretation is also consistent with findings in the previous studies demonstrating greater ß₂-AR-mediated NO-dependent vasodilator responses during local arterial infusion of ß₂-agonists in Gly¹⁶ subjects [Cockcroft, J. R. et al. 2000; Garovic, V. D. et al. 2003]. These findings underscore the need for further characterization of the interaction between positions 16 and 27, as well as additional coding or noncoding polymorphic variants in the ß₂-AR gene that may contribute to inter-individual differences. Further characterization of additional DNA sequence variation in the ß₂-AR gene may help refine understanding of these relationships

Modulation of β₂-AR function by Dietary Sodium Intake

Dietary sodium intake influences the pathogenesis and treatment of hypertension [He, F. J. et al. 2005; Meneton, P. et al. 2005], and the blood pressure response to changes in dietary sodium aggregates in families and is heritable [Miller, J. Z. et al. 1987]. In normotensives and individuals with hypertension, dietary sodium intake influences β₂-AR mediated vascular function [Feldman, R. D. 1990; Feldman, R. D. 1990; Feldman, R. D. et al. 1987; Feldman, R. D. et al. 1983; Feldman, R. D. et al. 1984; Lang, C. C. et al. 1995; Naslund, T. et al. 1990; Stein, C. M. et al. 2000; Stein, C. M. et al. 1995]. Evidence linking salt sensitivity to the β₂-AR locus has emerged from a linkage analysis of both hypertensive and normotensive siblings of hypertensive individuals [Svetkey, L. P. et al. 1997].

The effects of sodium intake on sensitivity to β₂-AR mediated vasodilatation have established that a normal response to increased sodium intake, consisting of increased sensitivity to β₂-AR mediated vasodilatation, is decreased in hypertensive compared to normotensive individuals [Naslund, T. et al. 1990]. Furthermore, β₂-AR mediated responsiveness is reduced selectively in peripheral veins of borderline hypertensive subjects, an effect which potentially is reversible by a low sodium diet [Feldman, R. D. 1990]. In this context it seems reasonable to postulate that variation in the β₂-AR gene would influence vasodilation in the setting of dietary sodium manipulation. Our group recently demonstrated that compared to Arg¹⁶ homozygotes, the greater β₂-mediated forearm vasodilator responses in healthy normotensive Gly¹⁶ homozygotes following a normal sodium diet are no longer present following a low sodium diet [Eisenach, J. H. et al. 2006]. Furthermore, in the Gly¹⁶ group, dietary sodium restriction increased resting systemic vascular resistance, decreased cardiac output, and tended to decrease resting stroke volume, whereas these indices were essentially unaffected in the Arg¹⁶ group [Eisenach, J. H. et al. 2006]. A concomitant study at our institution administered an acute intravenous sodium load (normal saline) to healthy individuals and found that mean arterial pressure was greater and urinary sodium uptake was greater (less urinary sodium excretion) in the Gly¹⁶ subjects [Snyder, E. M. et al. 2006]. Taken together, these findings provide early evidence that sodium influences the Arg¹⁶Gly β₂-AR genotype-dependent cardiovascular and renal intermediate phenotypes, and decreases in sodium balance may reduce ß₂-AR-mediated cardiovascular function for the Gly¹⁶ allele.

Interestingly, a recent multi-center trial examined blood pressure following a 2-week crossover of low (10 mmol/day) and high (200 mmol/day) dietary sodium intake. Genotype differences were not seen among normotensives (n=48); however, among hypertensives (n=171), Arg¹⁶ and Gln²⁷ – alone and in combination – displayed the greatest increase in mean arterial pressure from the low to the high sodium state, and the combination Arg¹⁶+Gln²⁷ was associated with low-renin hypertension, higher plasma aldosterone, lower renin, and lower potassium [Pojoga, L. et al. 2006]. Another investigation administered inhaled β₂-agonist to asthmatics and found a greater decrease in serum potassium in Arg¹⁶+Gln²⁷ homozygotes (n=8) vs. Gly¹⁶+Glu²⁷ homozygotes (n=8) [Lee, D. K. et al. 2004]. In view of these findings and the above findings from intravenous saline loading in healthy subjects [Snyder, E. M. et al. 2006], one may conclude that β₂-AR genotype-dependent intermediate physiologic traits may differ between healthy individuals and individuals with hypertension or asthma; furthermore, although unifying conclusions between these investigations are elusive, this nevertheless affirms the emerging theme that variation in the β₂-AR gene has powerful broad-based physiologic implications.

CLINICAL TRIALS / OUTCOME STUDIES

Evidence that β₂-AR gene variation has an important impact on clinical outcome is increasing; recent studies have suggested that the Gly¹⁶ and/or Glu²⁷ alleles may actually be favorable in cardiovascular health. The Cardiovascular Health Study (CHS), a population-based prospective cohort study with 7 to 10 years of follow-up, evaluated 4,441 white and 808 black participants. Irrespective of race/ethnicity, Glu²⁷ carriers had a lower risk of coronary events than Gln²⁷ homozygotes, and there was a suggestion of decreased risk among Gly¹⁶ carriers compared with Arg¹⁶ homozygotes [Heckbert, S. R. et al. 2003]. A subsequent report from the CHS demonstrated a higher risk of sudden cardiac death in Gln²⁷ homozygotes irrespective of race/ethnicity; similar findings were noted in the Cardiac Arrest Blood Study (CABS), presented in the same publication [Sotoodehnia, N. et al. 2006]. In another study, eight polymorphisms in the sympathetic nervous system and renin-angiotensin system were evaluated in 227 patients with heart failure. Of these, the Arg¹⁶+Gln²⁷ diplotype was the only genetic marker of increased risk of death or heart transplantation (see Fig. 5) [Shin, J. et al. 2007].

Figure 5.

Event-free survival (event defined as death or heart transplantation) in patients with heart failure. Possession of 2 copies of the Arg¹⁶ and Gln²⁷ allele was associated with a 90% increase in the risk of death or heart transplantation, after adjustment of gender, etiology, NYHA class, ejection fraction, serum sodium levels, creatinine clearance, and medications. Modified from [Shin, J. et al. 2007].

There is also growing evidence that genetic variation in the β₂-AR may predict the efficacy of therapeutic regimens [Iaccarino, G. et al. 2006; Kaye, D. M. et al. 2003; Lanfear, D. E. et al. 2005]. In eighty genotyped heart failure patients treated with carvedilol, individuals homozygous for Gln²⁷ represented a significantly lower proportion of ‘good’ responders (improvement in left ventricular function) than individuals who were homozygous or heterozygous for the Glu²⁷ polymorphism [Kaye, D. M. et al. 2003]. Another prospective cohort study followed patients with acute coronary syndrome: No mortality difference between genotypes was found among patients discharged without β-blocker therapy for neither the Arg¹⁶Gly nor the Gln²⁷Glu polymorphism. However, among patients treated with β-blockers, both polymorphisms – independently as well as combined – were predictive of survival: Patients homozygous for Arg¹⁶+Gln²⁷ had the worst survival, whereas patients homozygous for Gly¹⁶+Glu²⁷ had the best survival (see Fig. 6) [Lanfear, D. E. et al. 2005].

Figure 6.

Cumulative incidence of mortality in patients placed on β-blockers following an acute coronary syndrome. Individuals homozygous for both Arg¹⁶ and Gln²⁷ were at high-risk with a 3-year Kaplan-Meier mortality rate of 20%, whereas individuals homozygous for both Gly¹⁶ and Glu²⁷ were at low risk having a 3-year Kaplan-Meier mortality rate of only 6%. Modified from [Lanfear, D. E. et al. 2005].

Knowledge of the patient’s genotype may also be advantageous in choosing the optimal hypertensive therapy in patients with hypertension-induced left ventricular hypertrophy (LVH): Iaccarino et al. randomly assigned hypertensive patients with LVH to receive therapy with either a selective β₁-blocker (atenolol) or an ACE-inhibitor (enalapril) [Iaccarino, G. et al. 2006]. After 2-year follow-up, the patients carrying at least one Glu²⁷ allele showed a larger reduction in LVH when treated with enalapril, but not with atenolol therapy. A possible explanation for these data is based on the finding that the Glu²⁷ variant enhances the hypertrophic effect of the sympathetic system. Angiotensin-converting enzyme (ACE)-inhibitors are able to reverse LVH through the reduction of the hypertrophic effect of catecholamines [Trimarco, B. et al. 1985]. This property may be particularly relevant in patients carrying the Glu²⁷ allele, because by reducing sympathetic activation, it may prevent the more marked catecholamine-mediated hypertrophy stimulus induced by the Glu²⁷ variant [Iaccarino, G. et al. 2006].

HAPLOTYPES AND INTERMEDIATE CARDIOVASCULAR TRAITS

As mentioned above and shown in Fig. 3, the original description of β₂-AR haplotypes demonstrated the importance of interactions among multiple SNPs within a haplotype to influence physiologic function at a greater predictive power than individual SNPs. Specifically, the authors found that asthmatic patients with haplotype pair 2/2 as shown in Fig. 3 had a 50% greater bronchodilator response than those with haplotype pair 4/4; furthermore, this correlated with β₂-AR mRNA expression in HEK cells transfected with haplotype vector [Drysdale, C. M. et al. 2000]. Comprehensive analysis of 5’-flanking, coding-region, and 3’-untranslated sites of variation with lung function in asthmatics has been subsequently characterized by these authors [Hawkins, G. A. et al. 2006].

At the time of this review, we are not aware of any similarly comprehensive genotype-phenotype analyses with intermediate cardiovascular traits. This is most likely due to the large number of subjects required to achieve sufficient power with these more extended haplotypes. However, because of linkage disequilibrium, further inspection of the haplotypes in Fig. 3 reveals that individuals with haplotype 2/2 are homozygous for Gly¹⁶+Glu²⁷, and individuals with 4/4 are homozygous for Arg¹⁶+Gln²⁷. Furthermore, haplotype 4 (containing Gly¹⁶+Glu²⁷) also contains Cys¹⁹ (T-47C) in the 5’-leader cistron, which affects β₂-AR expression [McGraw, D. W. et al. 1998], and T-20C, which is completely concordant with Gln²⁷Glu. Taken together, these four sites generate haplotypes that account for approximately 94% of all haplotypes [Cerrone, G. E. et al. 2007; Herrmann, S. M. et al. 2002], and would explain why the majority of studies described in this review have examined SNP variation at positions 16 and 27, individually and in combination. Future investigations will take advantage of these tag SNPs to group patients according to haplotype and advance current understanding of the influence of β₂-AR haplotype on intermediate cardiovascular traits relevant to pharmacogenomic association studies.

PERSPECTIVES

Summarizing the clinical trials, the Arg¹⁶Gly and Gln²⁷Glu polymorphisms appear to have powerful pharmacogenetic and disease-modifying implications, with growing evidence that the Gly¹⁶ and/or Glu²⁷ alleles may be associated with favorable cardiovascular outcome. The causal contribution of each single polymorphism to the observed differences as studied in several in vivo studies is less clear so far; lack of statistical power and strong linkage disequilibrium between SNPs may account for inconsistencies in reported effects of isolated SNPs. The unique interactions of multiple SNPs within a haplotype ultimately can affect biologic and therapeutic phenotype, whereas individual SNPs alone may have poor predictive power as pharmacogenetic loci. Thus comprehensive haplotype analysis is needed to gain more precise information regarding the functional importance of genetic variation in the β₂-AR.

In the era of genome-wide association studies, it becomes increasingly clear that not only genetic variation within a single gene, but genetic variation throughout an entire pathway is important for pathophysiology and pharmacology and may have potential impact on clinical outcome. However, this makes defining clinically relevant polymorphisms much more complex. With increasing number of potentially important genetic subgroups (i.e. haplotypes, or even combinations of haplotypes of different genes), it will be increasingly difficult to perform randomized clinical trials with a sufficient number of participants to ensure appropriate power.

Nevertheless, more physiological studies are needed to answer the question how genetic variation in a crucial therapeutic target, the β₂-AR, affects cardiovascular responses, and more prospective randomized clinical trials are needed to find out the optimal therapy for a patient with a specific β₂-AR genotype. As recently reported, sequence variants in the β₂-AR can risk-stratify patients receiving β-blocker therapy and may predict β-blocker efficacy post-acute coronary syndrome [Lanfear, D. E. et al. 2005]. In the long term, reassessment of the benefits of β-blocker therapy within haplotype groups should be pursued, and extension of these findings into other disease states where β-blocker therapy or adrenergic stimulation is important (e.g. heart failure) should be considered. If future studies confirm a consistent and clinically relevant effect of genetic variation in the β₂-AR, then these findings could ultimately be used to tailor pharmacotherapy to the individuals` needs (personalized medicine).

LIST OF ABBREVIATIONS

ACE

angiotensin converting enzyme

Arg

arginine

β₂-AR

β₂-adrenergic receptor

CABS

Cardiac Arrest Blood Study

CHS

Cardiovascular Health Study

Cys

cysteine

Gln

glutamine

Glu

glutamate

Gly

glycine

HEK

human embryonic kidney

Ile

isoleucine

LVH

left ventricular hypertrophy

Met

methionine

NO

nitric oxide

Ser

serine

SNP

single nucleotide polymorphism

Thr

threonine

TPMT

thiopurin-S-methyltransferase

Val

valine