Arg16Gly polymorphism of the β₂-adrenergic receptor is associated with differences in cardiovascular function at rest and during exercise in humans

Methods

Subjects

The protocol was reviewed and approved by the Mayo Clinic Institutional Review Board, all participants gave informed consent prior to participation, and all aspects of the study conformed to the Declaration of Helsinki. Caucasian subjects who were matched for age, sex and activity were recruited from a pool of subjects who were previously genotyped for Arg16Gly and Gln27Glu polymorphisms of the β₂AR. Supplemental recruiting and genotyping were performed as needed to recruit adequate numbers in each of the genotype groups.

Individuals who were homozygous for arginine (ArgArg, n= 16) or glycine (GlyGly, n= 31) or heterozygous (ArgGly, n= 17) at codon 16 agreed to participate in the study and had no exclusion criteria (history of cardiopulmonary abnormalities, pregnancy, inability to exercise). All subjects were healthy non-smokers, and not on medication. A larger number of GlyGly subjects were recruited in order have a sufficient number to subdivide this group according to position 27 of the β₂AR. Since all ArgArg subjects were homozygous for Gln at position 27, no further subdivision of ArgArg subjects was necessary.

Protocol

Subjects underwent baseline pulmonary function testing, an incremental cycle ergometry test to exhaustion, a blood draw for a complete blood count (to rule out anaemia) and, in women, a pregnancy test. The baseline exercise study served as an initial familiarization session, was used to determine work intensities for subsequent sessions, and acted as a screening study to rule out myocardial ischaemia and abnormal arrhythmias. Following these initial tests, subjects met with the General Clinical Research Center (GCRC) nutritionist and were put on a controlled sodium diet (3450 mg day⁻¹) for 3 days with a 24-h urine collection to confirm sodium intake (urine Na⁺ ArgArg = 77 ± 39, ArgGly = 57 ± 33, GlyGly = 69 ± 27 mmol l⁻¹, mean±s.d., P-ANOVA > 0.05). This controlled sodium diet was used because previous studies have suggested that the β₂AR may be sensitive to changes in dietary sodium (Skrabal et al. 1989; Kotanko et al. 1992). Subjects subsequently returned to the GCRC on two occasions for exercise testing while maintaining a salt neutral diet.

The next session consisted of a cycle ergometry test similar to the first visit but with the additional measurement of using a previously validated open-circuit acetylene uptake method (Johnson et al. 2000). This session served as a further familiarization with the measurements to be made on the final study day and also allowed for confirmation of workloads for the final visit.

On the last visit, prior to study, a 5 cm 20-gauge catheter (Arrow International, Reading, PA, USA) was placed in the radial artery after local anaesthesia with 2% lidocaine to assess catecholamines and direct arterial blood pressure. Resting measurements of , HR, SV and arterial BP were made. Subjects then exercised for 9 min at ∼40% and 9 min at ∼75% of their peak workload achieved during the initial exercise studies while measurements were repeated every 2–3 min. Nine minutes of exercise was performed because pilot data suggested that this was an adequate time frame to obtain three sets of measures and brought the subjects close to exhaustion with the higher workload.

Data collection

β₂AR genotyping. 

β₂AR genotyping was PCR based according to methods of Bray et al. (2000). Buffy coat, obtained from whole blood collected on EDTA, was extracted using the Gentra Puregene DNA Isolation Kit (Gentra Systems Inc., Minneapolis, MN, USA). The PCR reaction was conducted according to standard methods, using the following primer sequences (e.g. for Arg16Gly): (forward) 5′-AGC CAG TGC GCT TAC CTG CCA GAC-3′ (at −32) and (reverse) 3′-CA TGG GTA CGC GGC CTG GTG CTG CAG TGC-5′, resulting in a PCR product 107 base pairs in length. The reaction included 30 ng of DNA, 1.5 mm magnesium chloride, 0.5 U taq polymerase (Invitrogen, Carlsbad, CA, USA), 8.5% DMSO and standard concentrations of nucleotides and buffer in a 20 µl reaction volume. After initial denaturation at 94°C for 4 min, the fragments were amplified by 35 cycles of 1 min at 94°C, 1 min at 61°C, 1 min at 72°C, followed by 5 min at 72°C and 5 min at 98°C. The amplicons were then digested by exposure to 5 U of the restriction enzyme KpnI, followed by electrophoretic separation on 3% aragose gels, staining with ethidium bromide and visualization using UV light. The ArgArg homozygous genotype is represented by a single 107 bp band, the ArgGly group is represented by 25, 82 and 107 bp bands, and the GlyGly homozygous group by 82 and 25 bp bands.

Catecholamines. 

Adrenaline (ADR) and noradrenaline (NA) were assessed according to methods developed in the Mayo Clinic GCRC immunochemical core laboratory and the methods of Sealey (1991). For ADR, our lab intra-assay coefficients of variation (CVs) were 12.2% and 3.6% at 13.8 and 242 pg ml⁻¹. Inter-assay CVs are 8.5% and 6.3% at 179 and 390 pg ml⁻¹.

Cardiovascular assessment during exercise

Gas exchange measurements. 

Oxygen uptake , carbon dioxide production , minute ventilation , breathing frequency (f_b) and tidal volume (V_T) were measured continuously during the various exercise tests and stages using a Medical Graphics (St Paul, MN, USA) metabolic cart interfaced with a Perkin Elmer mass spectrometer (Wellesley, MA, USA). This system has been validated against classic ‘bag’ collection techniques, and stability is verified by regular testing at standard exercise intensities by laboratory personnel (Proctor & Beck, 1996).

Assessment of , SV, BP, SVR and HR. 

Cardiac output was assessed using a 10-breath open-circuit acetylene wash-in technique as previously described (Johnson et al. 2000). Briefly, a pneumotachograph was connected to a non-rebreathing Y valve (Hans Rudolph, Kansas City, MO, USA) with the inspiratory port connected to a pneumatic switching valve (Hans Rudolph) which allowed for rapid switching from room air to the test gas mixture (filled in a large reservoir containing 0.7% C₂H₂, 21% O₂, 9% He, balance N₂). Gases were sampled using a mass spectrometer (Perkin-Elmer), which was integrated with custom analysis software for the assessment of using our previously described iterative technique. Stroke volume was calculated by dividing the cardiac output by the heart rate.

Intra-arterial BP was assessed using a SpaceLab 512D patient monitor (SpaceLabs Inc., Hillsboro, OR, USA). Mean arterial pressure (MAP) was calculated using the equation: where DBP is diastolic blood pressure and SBP is systolic blood pressure. Systemic vascular resistance was calculated from MAP using the equation: . Heart rate was assessed using 12-lead electrocardiography (Marquette Electronics, Milwaukee, WI, USA).

Statistical analysis

The demographic data were examined using an analysis of variance (ANOVA). To examine the group differences at rest and during exercise in and SVR, we performed an ANOVA using the SPSS statistical software package (Chicago, IL, USA) with a Tukey HSD test to determine differences between groups. All data were found to have homogeneity of variance prior to the ANOVA using Levene's test for equality of variance. The α levels for the ANOVAs and post hoc analyses were set at 0.05.

Results

Subject characteristics

There were no differences between the genotype groups in age, height, weight, body mass index (BMI), body surface area (BSA), or (Table 1).

Resting cardiovascular function

All resting measures were taken upright, following 5 min of quiet rest. At rest, the ArgArg group had a lower and SV when compared to the GlyGly group (P-ANOVA < 0.05, post hocP= 0.001, Fig. 1) but had a higher HR and a lower MAP when compared to both the ArgGly and GlyGly groups (SBP = 109 ± 9, 116 ± 10, 113 ± 11 mmHg; DBP = 69 ± 6, 78 ± 9, 75 ± 9 mmHg, for ArgArg, ArgGly, and GlyGly groups, respectively, P-ANOVA < 0.05). Stroke volume corrected for BSA (stroke volume index, SVI) was also significantly lower in the ArgArg subjects when compared to the GlyGly subjects (38 ± 2, 42 ± 3, 46 ± 3, for the ArgArg, ArgGly and GlyGly groups, respectively, P-ANOVA < 0.05, post hocP < 0.05 for ArgArg versus GlyGly) There were no differences in SVR between the genotype groups at rest.

Catecholamines

At rest, there were no differences in ADR between the groups (Fig. 2). With light exercise there were minimal increases in ADR; however, the ArgArg group had lower ADR than the GlyGly group (post hocP= 0.036). During heavy exercise there was a dramatic increase from baseline in ADR in all genotype groups (P < 0.01). No differences were observed between genotype groups in ADR with heavy exercise. Adrenaline values averaged 93 ± 36, 95 ± 40 and 103 ± 40 pg ml⁻¹ with light exercise and 386 ± 400, 277 ± 157 and 294 ± 193 pg ml⁻¹ with heavy exercise for the ArgArg, ArgGly and GlyGly groups, respectively. At rest and throughout exercise NA was not significantly different among genotype groups (rest = 279 ± 22, 301 ± 105, 317 ± 11 pg ml⁻¹; light exercise = 600 ± 41, 602 ± 22, 589 ± 36 pg ml⁻¹; heavy exercise = 2099 ± 252, 2171 ± 162, 2521 ± 145 pg ml⁻¹, for  the ArgArg, ArgGly, and GlyGly groups, respectively.

Cardiovascular response to exercise

All three groups exercised at similar relative and absolute workloads during light exercise and heavy exercise (37% and 75% of peak power, and 38% and 78% of peak HR, respectively, P-ANOVA > 0.05) (Table 2).

Light exercise

With light exercise the ArgArg group continued to have a lower than both the ArgGly and GlyGly groups (P-ANOVA < 0.05)(Fig. 3) and lower SV when compared to the GlyGly group (P < 0.03); however, SVR and HR were not different. The ArgArg group also had a lower MAP when compared to both the ArgGly and GlyGly groups (P < 0.05), which was influenced by differences in both SBP and DBP (SBP = 137 ± 19, 148 ± 18, 146 ± 20 mmHg; DBP = 71 ± 8, 68 ± 11, 74 ± 11 mmHg, for the ArgArg, ArgGly, and GlyGly groups, respectively, P-ANOVA < 0.05 for DBP). Cardiac output corrected for BSA (cardiac index, CI) and SVI remained lowest in the ArgArg group (CI = 5.3 ± 0.16, 5.7 ± 0.10, 6.0 ± 0.13; SVI = 42 ± 1.8, 45 ± 2.0, 46 ± 1.6, for ArgArg, ArgGly and GlyGly, respectively, P < 0.05 for ArgArg versus GlyGly). The ArgGly group had the largest change from baseline in both and SV with light exercise (, for ArgArg, ArgGly, and GlyGly, respectively, P-ANOVA < 0.05).

Heavy exercise

With heavy exercise the ArgArg group continued to have a lower and SV compared to both the ArgGly and GlyGly groups (P < 0.01) but there remained no differences between the genotype groups in SVR or HR. In addition, the ArgArg and GlyGly groups had a lower MAP than the ArgGly group (P < 0.05); however, there were no differences in SBP or DBP (SBP = 160 ± 22, 177 ± 23, 168 ± 22 mmHg; DBP = 72 ± 8, 60 ± 21, 68 ± 19 mmHg, for the ArgArg, ArgGly, and GlyGly groups, respectively). Cardiac index and SVI remained lowest in the ArgArg group (CI = 7.3 ± 0.20, 8.0 ± 0.15, 8.1 ± 0.17; SVI = 42 ± 1.8, 46 ± 2.0, 46 ± 1.6, for the ArgArg, ArgGly and GlyGly groups, respectively; CI: P < 0.05 for ArgArg versus ArgGly and GlyGly, SVI: P < 0.05 for ArgArg versus GlyGly). During heavy exercise the ArgGly group had the largest change from baseline in and SV (, for the ArgArg, ArgGly, and GlyGly groups, respectively, P-ANOVA < 0.05). In addition, the ArgGly group had the largest drop in SVR from baseline (ArgArg, −42 ± 3%; ArgGly, −51 ± 1%; and GlyGly, −44 ± 2%; P < 0.01).

Recovery

Five and ten minutes into recovery, the ArgArg group had lower MAP compared to both the ArgGly and GlyGly groups (P < 0.05) but no differences were observed in , HR, SV or SVR or the percentage change from baseline in these parameters.

Discussion

This is one of the first studies specifically designed to examine the effects of an endogenously released agonist on the CV response to exercise according to β₂AR genotype. We have found that the Arg16Gly polymorphism of the β₂AR influences resting cardiovascular function in healthy adults which persists during light and heavy exercise when respectively minimal and marked increases in catecholamines occur. The ArgArg group had a lower and MAP compared to both of the other groups, and a lower SV compared to the GlyGly group at rest. During exercise, and into recovery, the ArgArg group continued to have a lower SV, and MAP relative to the other groups. These findings suggest that the Arg16Gly polymorphism of the β₂AR affects baseline cardiovascular function and that these baseline differences are sustained during short-term low and high intensity exercise. Despite the reduced cardiac function in the ArgArg subjects, was clearly in the normal range for healthy adults and there was no indication that these subjects had a limited ability to exercise. The exact mechanisms for the baseline differences are not known, but it is possible that subjects homozygous for Arg at amino acid 16 either have fewer β₂ARs or have sustained down-regulation of the receptors, either of which could explain the observed differences.

The β₂ARs and cardiovascular function

The β₂ARs have been shown to play a modest role in cardiac regulation. We hypothesized that the ArgArg group would have a blunted increase in and SV relative to the GlyGly group during heavy exercise but that there would be minimal differences at rest and during light exercise. In addition, one might expect increased vasodilatation and therefore decreased SVR as exercise progressed, based on previous research that has shown that the genotype difference at position 16 of the β₂AR is likely a result of differences in agonist-induced desensitization (Cockcroft et al. 2000; Garovic et al. 2003; Eisenach et al. 2004). We found, however, that there were baseline differences in cardiovascular function between the genotype groups which were maintained during light and heavy exercise. These differences are in agreement with recent work by Tang et al. (2003), who have shown that ArgArg subjects have attenuated left ventricular function (fractional shortening, ejection fraction, and midwall shortening) at rest when compared to GlyGly subjects suggesting that the ArgArg genotype is associated with reduced baseline receptor function when compared to the GlyGly genotype.

β₂ARs and blood pressure regulation

In the present study, the ArgArg group also had a lower MAP at rest, during exercise, and into recovery when compared to both the ArgGly and GlyGly groups. Previous studies have observed an association between the Arg16Gly polymorphisms of the β₂AR and elevated BP (Kotanko et al. 1992; Timmermann et al. 1998; Bray et al. 2000; Pereira et al. 2003), while others have shown no association (Xie et al. 2000; Herrmann et al. 2002). Specifically, the GlyGly allele has been associated with increased BP (Bray et al. 2000; Pereira et al. 2003). These findings are somewhat counterintuitive, however, because several studies in healthy subjects have shown that the GlyGly group tends to have greater agonist-mediated vasodilatation (Cockcroft et al. 2000; Dishy et al. 2001; Garovic et al. 2003; Eisenach et al. 2004). All else being equal, one would hypothesize that this improved vasodilatation would lead to a decreased SVR and lower BP. There are other possible β₂-mediated mechanisms involved in blood pressure regulation, however, that may cause the GlyGly group to have higher blood pressures (Snyder et al. 2005). The most likely cause of this elevated BP in the GlyGly group is the elevated when compared to the ArgArg subjects. This does not fully explain the difference, however, because the GlyGly group has a higher resting than the ArgGly group, but the ArgGly group has a higher resting mean arterial pressure than the GlyGly group (despite similar SVRs). Another possible candidate for the elevated BPs in the ArgGly and GlyGly groups is an increased renin activity that may be caused by stimulation of the renal adrenergic receptors (Koepke & DiBona, 1986).

In addition, the renal epithelial sodium channels (ENaCs) are mediated, in part, by the β₂AR. The β₂ARs are located along the nephron and the ENaCs have been localized to the distal collecting duct and inner medullary collecting duct (IMCD) and both probably play a key role in long-term Na⁺ and blood pressure regulation (Wallace et al. 2004). β₂AR stimulation in IMCD cells causes a significant increase in anion secretion which is abolished with β₂-blockade (Wallace et al. 2004). In vivo work has shown that β-blockade leads to diuresis and naturesis, and blocks the antinatriuretic response to environmental stress (Smits et al. 1982; Koepke & DiBona, 1986). If the GlyGly group has a greater β₂-mediated activity of the ENaC in the kidneys they would probably have greater reabsorption of Na⁺, and may have elevated blood pressure, even though there is evidence that these subjects have a greater agonist-mediated vasodilatation.

Interaction between position 16 and other sites of the β₂AR

A possible reason for the variable findings in previous research on genotype variation of the β₂AR may be the interaction of other sites encoding the β₂AR. Another site of the β₂AR which has been studied in depth is amino acid 27. Drysdale et al. (2000) have suggested the use of possible haplotype combinations of the β₂AR instead of SNPs because of site to site interactions. Combination of polymorphisms at positions 16 and 27 can be used to estimate haplotypes due to linkage disequilibrium throughout the SNPs and the known frequency of the polymorphisms in Caucasians.

Enough ArgGly and GlyGly subjects were recruited in the present study to perform an analysis on position 27 and probable haplotypes as described by Drysdale et al. (2000) (Table 3). They found that the haplotype had the smallest increase in forced expiratory volume in 1 s following administration of a β-agonist, while the group had the largest increase. We also found that the Arg16Arg/Gln27Gln subjects (haplotype) had the lowest SV and during light and heavy exercise. However, we found that the Gly16Gly/Gln27Glu subjects (haplotype) had the highest and the lowest SVR at rest and during light and heavy exercise and the highest SV during heavy exercise. Because of the variable cardiovascular response within position 16 based on position 27 (i.e. Gly16Gly/Gln27Gln, Gly16Gly/Gln27Glu, and Gly16Gly/Glu27Glu) and likely interactions between positions 16, 27 and other sites of the β₂AR, there is a clear importance for haplotype analysis of the β₂AR in future studies.

Study limitations

Our study design allowed for the assessment of the cardiovascular response to exercise at an intensity in which minimal catecholamine release occurred and one in which there was a substantial catecholamine response. There were no differences across genotypes in the change in cardiovascular function (, HR, SV, MAP, SVR) from the first measure during heavy exercise to the last measure during heavy exercise despite large increases in ADR, suggesting no group differences in β₂AR desensitization during this short-term exposure (Fig. 4). This is in disagreement with previous studies that have shown the ArgArg group to have enhanced agonist-mediated desensitization upon exposure to an exogenously administered agonist. However, our adrenaline exposure may have been too brief to induce receptor desensitization to a degree which markedly alters cardiovascular function.

Another possible limitation of the present study is the assessment based on one allele of the β₂AR. We found that there was significant variability in the cardiovascular response to exercise according to amino acid 16 which may also be related to amino acid 27 or other alleles in linkage disequilibrium with position 16. Although we provide data according to probable haplotypes as described by Drysdale, our haplotype groups were relatively small and were not carefully matched for age, sex, or fitness according to haplotype. Although difficult to perform invasive, hypothesis-driven studies according to haplotypes combinations, future studies are necessary and should include polymorphisms of the 3′ and 5′ flanking regions of the β₂AR.

We made every effort to match the groups according to age, sex, fitness and body mass index. There were, however, subtle non-significant differences between the genotype groups in weight and height which could influence baseline cardiac function. To account for these more subtle differences we indexed and SV according to body surface area and the results were similar to the non-indexed values, suggesting genotype difference, and not body size, account for differences in cardiovascular function. Although we selected only Caucasian subjects for the present study, there are possible differences in population admixture (ancestry) between the genotype groups which could influence the findings. All subjects were from Rochester, Minnesota and most were of European descent; however, the exact ancestry of these subjects was unknown.

Conclusions

Our findings show that subjects homozygous for Arg at amino acid 16 of the β₂AR have attenuated and SV at rest when compared to heterozygous subjects and subjects homozygous for Gly. Additionally, subjects homozygous for Arg at amino acid 16 also had a lower resting mean arterial pressure when compared to heterozygous subjects and subjects homozygous for Gly. These differences in , SV, and MAP were sustained during light exercise and heavy exercise suggesting differences in baseline receptor function or density which contributed to the observed differences during exercise.