C825T Polymorphism of the G Protein β3-Subunit and Antihypertensive Response to a Thiazide Diuretic
The T allele of the C825T polymorphism of the gene encoding the β3-subunit of G proteins has been associated with increased sodium-hydrogen exchange and low renin in patients with essential hypertension. To assess its association with blood pressure response to diuretic therapy, we measured the C825T polymorphism in 197 blacks (134 men, 63 women) and 190 non-Hispanic whites (76 men, 114 women) with essential hypertension (mean±SD age 48±7 years), who underwent monotherapy with hydrochlorothiazide for 4 weeks. Mean declines in systolic and diastolic blood pressures were 6±2 (P<0.001) and 5±1 (P<0.001) mm Hg greater, respectively, in TT than in CC homozygotes. Responses in heterozygotes were intermediate between the homozygous groups. Other univariate predictors of greater blood pressure responses included black race, female gender, higher pretreatment blood pressure, older age, lower waist-to-hip ratio, and measures of lower renin-angiotensin-aldosterone system activity. After the effects of the other predictors were considered, the TT genotype remained a significant predictor of greater declines in systolic and diastolic blood pressures. Thus, the C825T polymorphism of the G protein β3-subunit may help identify patients with essential hypertension who are more responsive to diuretic therapy.
Despite the availability of numerous classes of antihypertensive drugs that act on a variety of blood pressure–regulating systems, fewer than 40% of treated hypertensives have their blood pressures adequately controlled.1 Within each class of antihypertensive drugs, blood pressure responses vary considerably among patients, with a range of response that is several times greater than the mean response.2 This interindividual variation in response is primarily due to pharmacodynamic, not pharmacokinetic, differences3 and likely reflects variation in pathophysiological mechanisms that contribute to hypertension in individual patients.4
It has long been suspected that interindividual variation in drug responses may be influenced by genetic factors.5 Recently, a polymorphism (C825T) was described in exon 10 of the gene encoding the β3-subunit of G proteins (GNB3),6 and subsequently it was found to be associated with a shortened splice variant of the Gβ3-protein that gives rise to enhanced signal transduction via pertussis toxin–sensitive G proteins.7 The C825T polymorphism was originally identified through studies of lymphoblasts, derived from whites with essential hypertension, in which sodium-proton antiport activity was increased as a consequence of enhanced G protein–dependent signal transduction in response to a variety of vasoactive and growth-promoting stimuli.8
The present study was prompted by recently reported associations of the 825T allele with obesity9 and with low plasma renin10 that suggest that the 825T allele might also be associated with a volume-expanded, sodium-sensitive, and therefore diuretic-responsive form of hypertension.11 12 Hence, our objective was to determine whether the C825T polymorphism predicts interindividual variation in blood pressure response to diuretic therapy among subjects with essential hypertension. We tested this hypothesis in a community and clinic-based biracial sample of hypertensive women and men undergoing monotherapy with a thiazide diuretic.
The sample consisted of 197 unrelated black adults (134 women, 63 men) from Atlanta, Ga, and 190 unrelated non-Hispanic white adults (76 women, 114 men) from Rochester Minn, aged 30 to 59.9 years with previously diagnosed hypertension who were participants in an ongoing study to identify predictors of blood pressure response to diuretic therapy. In Atlanta, study candidates were identified through lists of registered voters in Fulton and DeKalb counties (70%); outpatient medical clinics of Emory University Hospital, Grady Memorial Hospital, and the Atlanta Veterans Administration Hospital (15%); and other community sources, including churches and public media (15%). In Rochester, candidates were identified through the Rochester Epidemiology Project13 and a diagnostic index maintained by the Mayo Clinic for all residents of Olmsted County. A letter was sent to potential subjects providing information about the study. Interested candidates who contacted the study centers were considered for recruitment. Before enrollment, blood was drawn for screening hematology and chemistry profiles. To qualify for study participation, subjects had to be in good general health, with systolic blood pressure <180 and diastolic blood pressure <110 mm Hg, and they had to have no evidence by medical record review, history, physical examination, or screening laboratory studies of secondary hypertension, renal or liver dysfunction, serious heart disease, diabetes, gout, or sulfa allergy. Subjects who had been taking diuretics and were found to be hypokalemic at the screening visit were allowed to participate (as described below); however, subjects with unexplained hypokalemia were excluded. Subjects had to be able to discontinue drugs that could influence blood pressure level or the renin-angiotensin-aldosterone system or that could antagonize the effect of thiazide diuretics; however, postmenopausal women were allowed to continue hormone replacement therapy. Participants were required to read and sign a written consent form. The institutional review boards of Emory University and the Mayo Clinic approved all procedures involving study subjects in Atlanta and Rochester, respectively. All study procedures were performed in each institution’s General Clinical Research Center (GCRC) in accordance with the respective institution’s guidelines.
Participants had their antihypertensive medications withdrawn, and other contraindicated drugs were discontinued. If the participant had been taking a diuretic and their serum potassium level was <3.6 mEq/L at the screening visit, a potassium supplement (potassium chloride 20 mEq/d) was prescribed. A dietitian instructed each participant in a diet designed to provide a standard sodium intake of 2 mmol per kilogram of body weight per day. Compliance was monitored by 24-hour urine collections obtained every other week on the diet. Study subjects were seen every 2 weeks by the study nurse for blood pressure monitoring. At a minimum of 4 weeks after antihypertensive drug therapy was discontinued, subjects began taking 25 mg of hydrochlorothiazide orally each day for the next 4 weeks. Compliance was assessed by pill counts. Subjects’ weight, blood pressure, and serum potassium concentrations were remeasured after 2 and 4 weeks of diuretic therapy. Low-dose oral potassium supplements (20 to 40 mmol/d) were prescribed after 2 weeks of diuretic therapy if serum potassium was <3.4 mEq/L. At the end of the drug-free period and at the end of the diuretic-therapy period, subjects slept overnight in the GCRC. At ≈6 the next morning, blood for measurement of plasma aldosterone concentration, renin activity, and serum potassium concentration was drawn from subjects in the seated position after 30 minutes of ambulation.
Aldosterone and renin activity were determined by radioimmunoassays. Each sample was assayed in triplicate and the average used in the analyses. The C825T polymorphism was genotyped by polymerase chain reaction amplification of the relevant region of genomic DNA, followed by restriction enzyme digestion, gel electrophoresis, and ethidium bromide staining.6
For quantitative traits, data were summarized by calculating means and variances for each GNB3 genotype within each racial group. A 1-way ANOVA was used to assess differences in means among genotypes within each racial group. Relative frequencies of genotypes and alleles were calculated for each racial group. χ2 contingency tests were used to assess differences in relative frequencies between racial groups.
We used linear regression to assess whether variation in genotype made a statistically significant contribution to the prediction of systolic and diastolic blood pressure responses to hydrochlorothiazide. We first considered models that included only genotype and subsequently models that included genotype as well as covariate traits (including race, gender, pretreatment blood pressure levels, age, waist-to-hip ratio, and measures of the endocrine renin-angiotensin-aldosterone system). In preliminary analyses, we found no evidence of significant interaction between the effects of genotype and those of the other predictor traits.
Genotype and Allele Frequencies
Relative frequencies of the CC, TC, and TT genotypes differed significantly between racial groups (P<0.001), with the T allele significantly more frequent in blacks than in whites (76.1% versus 28.9%, P<0.001) (Table 1⇓). In blacks, genotype and allele frequencies did not differ significantly between genders (P>0.5). In whites, differences in genotype frequencies between the genders also did not reach statistical significance (P>0.06); however, the T allele was significantly more frequent in men than in women (33.3% versus 22.4%, P<0.03). In each race, observed genotype frequencies did not differ significantly from those predicted by the Hardy-Weinberg law.
Means for pretreatment systolic and diastolic blood pressure levels, waist-to-hip ratio, plasma renin activity, and plasma aldosterone concentration differed significantly between blacks and whites (Table 2⇓). However, only the lower means for waist-to-hip ratio and plasma renin activity in blacks were consistent across all genotype groups. Because mean waist-to-hip ratio was significantly lower in women than in men in both races (P<0.001, analyses not shown), lower mean waist-to-hip ratio in blacks could be attributed to the higher percentage of women among blacks (68%) than among whites (40%; see Table 1⇑). In contrast, because mean plasma renin activity did not differ significantly between black women and men (1.0±1.1 versus 1.1±1.1, P>0.5) and mean plasma renin activity was significantly lower in white men than white women (1.6±1.3 versus 2.1±1.4, P<0.04), gender differences could not account for lower mean PRA in blacks than whites.
In each race, means for the covariate traits did not differ significantly among genotype groups (Table 2⇑). However, because of racial differences in the genotype frequencies (see Table 1⇑), means for pretreatment systolic and diastolic blood pressure levels, waist-to-hip ratio, plasma renin activity, and plasma aldosterone concentration appeared to differ significantly among genotype groups when blacks and whites were pooled (Table 2⇑).
Association of Blood Pressure Response With Genotype
Variation in genotype was a statistically significant predictor of systolic and diastolic blood pressure responses to hydrochlorothiazide treatment (Table 3⇓ and Figure 1⇓). In the pooled sample, systolic and diastolic blood pressure responses increased progressively with the number of T alleles: mean declines in systolic and diastolic blood pressure were, respectively, 10.2 and 5.9 mm Hg in CC homozygotes, 13.6 and 7.8 mm Hg in TC heterozygotes, and 16.3 and 10.5 mm Hg in TT homozygotes. In pairwise contrasts between genotypes, the mean declines in systolic blood pressure were significantly greater in TC heterozygotes and TT homozygotes than in CC homozygotes (P<0.05 and P<0.001, respectively), and the mean decline in diastolic blood pressure was significantly greater in TT than in CC homozygotes (P<0.001). This same pattern of progressively greater systolic and diastolic blood pressure responses in association with the number of T alleles was observed in each race-gender group except black women (Figures 2⇓ and 3⇓). The differences in responses between TT and CC genotypes were greatest in white women (for systolic blood pressure response, 30.3 versus 10.0 mm Hg, respectively; for diastolic blood pressure response, 12.3 versus 5.8 mm Hg, respectively).
By itself, variation in genotype accounted for 3.1% and 4.5%, respectively, of interindividual variation in the systolic and diastolic blood pressure responses to hydrochlorothiazide (Table 3⇑). Other significant univariate predictors of greater systolic and diastolic blood pressure responses to hydrochlorothiazide included higher pretreatment blood pressure levels, black race, female gender, older age, and lower plasma renin activity and urinary aldosterone excretion (Table 3⇑). Lower waist-to-hip ratio and lower plasma aldosterone concentration were additional significant predictors of systolic and diastolic blood pressure responses, respectively.
In multiple regression models that included genotype in addition to the other predictor traits (which were considered one at a time in Table 3⇑), genotype remained a statistically significant predictor of systolic (P<0.04) and diastolic (P<0.01) blood pressure responses to hydrochlorothiazide (models not shown). In addition to C825T genotype, pretreatment blood pressure levels, plasma renin activity, and urinary aldosterone excretion also remained significant predictors of systolic and diastolic blood pressure responses, whereas gender and age remained significant predictors of the diastolic but not the systolic blood pressure response, and race and plasma aldosterone concentration made no additional contribution to the prediction of either systolic or diastolic blood pressure response. The models that included effects of all of the predictor traits (including genotype) accounted for 32% and 18%, respectively, of interindividual variation in systolic and diastolic blood pressure responses (P<0.001 for both models).
In separate multiple regression models that considered the effects of race, gender, and age, variation in genotype was not a statistically significant predictor of waist-to-hip ratio. In models that also included waist-to-hip ratio, variation in genotype was not a statistically significant predictor of pretreatment blood pressure levels, plasma renin activity, plasma aldosterone concentration, or urinary aldosterone excretion (models not shown).
The findings in this study suggest that among subjects with previously diagnosed essential hypertension, the C825T polymorphism of the gene encoding the β3-subunit of G proteins is associated with interindividual differences in blood pressure response to a thiazide diuretic. One possible inference is that the 825T allele may mark hypertensive individuals in whom elevation of blood pressure has a diuretic-responsive and therefore sodium- and volume-dependent component. It is appropriate to consider whether this observed association has a plausible mechanism or whether it might simply be explained by confounding factors.
One potential confounder is race. Consistent with previous reports, we found that the 825T allele was more frequent in blacks than in whites14 and that both black race and older age were also predictive of greater blood pressure response to hydrochlorothiazide.15 However, the relationships we observed between the C825T polymorphism and interindividual differences in blood pressure responses to hydrochlorothiazide were present in all race-gender groups except black women. Moreover, in multiple linear regression models, the C825T polymorphism remained a significant predictor of interindividual differences in blood pressure responses after the contributions of race as well as the other identified predictors of response were considered.
More relevant to potential mechanistic pathways in patients with essential hypertension, the 825T allele has also been associated with obesity,9 low plasma renin, and elevated aldosterone-to-renin ratio,10 all of which are correlates of volume expansion and sodium sensitivity.11 12 However, we did not detect significant associations of the C825T polymorphism with measures of body fat or activity of the renin-angiotensin-aldosterone system. Additional potentially relevant associations of the 825T allele have been reported with enhanced in vitro sodium-hydrogen antiport and proliferation of immortalized lymphoblasts from subjects with essential hypertension.16 17 Unfortunately, we did not make direct measures of sodium ion transport or cellular proliferation in the present study. Because G proteins relay signals from a wide variety of hormones and neurotransmitters and transmit them to numerous effector systems,18 it is plausible that some unidentified “gain of function” related to the 825T allele may influence the balance of intravascular fluid volume and vasoconstriction that in turn influences the blood pressure response to a thiazide diuretic.
The contribution of the C825T polymorphism to the explanation of interindividual variation in blood pressure response to the thiazide diuretic was small. Nevertheless, the magnitude of its effect was similar to that of other previously identified predictors of diuretic responsivity, including race, age, and measures of the renin-angiotensin-aldosterone system.15 Only the predictive effects of pretreatment systolic blood pressure level substantially exceeded those of the C825T polymorphism. Insofar as these results are indicative of the expected magnitude of effects of variation at other single genetic loci, clinically useful prediction of complex drug-response phenotypes, such as the antihypertensive responses to a thiazide diuretic, would appear to require measurement of the variation in multiple genetic and environmental factors. Nevertheless, our present findings provide additional evidence that direct measures of genetic variation can improve the ability to predict responses to antihypertensive drug therapy.19
In summary, the present study suggests that measured variation in the gene encoding the β3-subunit of G proteins predicts interindividual differences in blood pressure response to a thiazide diuretic and that the predictive effects of the C825T polymorphism are additive and similar in magnitude to those of other predictors, including race, age, and measures of the renin-angiotensin-aldosterone system. Knowledge of genetic polymorphisms that are predictive of blood pressure response to antihypertensive drugs with known mechanisms of action could provide insight into the molecular mechanisms that contribute to hypertension and aid in the tailoring of effective antihypertensive drug therapy in individual patients.
This work was supported by US Public Health Service grants R01-HL53330 and funds from the Mayo Foundation. We gratefully acknowledge support from the General Clinical Research Centers at Emory University (M01-RR00039) and the Mayo Clinic, Rochester (M01-RR00585).
- Received October 26, 2000.
- Revision received December 11, 2000.
- Accepted December 29, 2000.
Meissner I, Whisnant J, Sheps S, Schwartz G, O’Fallon W, Covalt J, Sicks J, Bailey K, Wiebers D. Detection and control of high blood pressure in the community: do we need a wake-up call? Hypertension. 1999;34:466–471.
Laragh J, Lamport B, Sealey J, Alderman M. Diagnosis ex Juvantibus: individual response patterns to drugs reveal hypertension mechanisms and simplify treatment. Hypertension. 1988;12:223–226.
Motulsky A. Drug reactions, enzymes and biochemical genetics. JAMA. 1957;165:835–837.
Rosskopf D, Busch S, Manthey I, Siffert W. G protein β3 gene: structure, promoter, and additional polymorphism. Hypertension. 2000;36:33–41.
Siffert W, Rosskopf D, Mortiz A, Wieland T, Kaldenberg-Stasch S, Kettler N, Hartung K, Beckmann S, Jakobs K. Enhanced G protein activation in immortalized lymphoblasts from patients with essential hypertension. J Clin Invest. 1995;96:759–766.
Schunkert H, Hense H, Doring A, Riegger G, Siffert W. Association between a polymorphism in the G protein β3-subunit gene and lower renin and elevated diastolic blood pressure levels. Hypertension. 1998;32:510–513.
Siffert W, Forster P, Jockel K, Mvere DA, Brinkmann B, Naber C, Crookes R, Heyns A, Epplen JT, Fridey J, Freedman BI, Muller N, Stolke D, Sharma AM, Al Moutaery K, Grosse-Wilde H, Buerbaum B, Ehrlich T, Ahmad HR, Horsthemke B, Du Toit ED, Tiilikainen A, Ge J, Wang Y, Rosskopf D, Yang D, Husing J. Worldwide ethnic distribution of the G protein β3 subunit 825T allele and its association with obesity in Caucasian, Chinese, and black African individuals. J Am Soc Nephrol. 1999;10:1921–1930.
Rosskopf D, Fromter E, Siffert W. Hypertensive sodium-proton exchanger phenotype persists in immortalized lymphoblasts from essential hypertensive patients. J Clin Invest. 1993;92:2553–2559.