Association of Hypertension with β2- and α2c10-Adrenergic Receptor Genotype
Abstract The adrenergic receptors have been implicated in the pathogenesis of essential hypertension. We hypothesized that hypertension is associated with variants at the β2-adrenergic receptor locus and at one of the α2-adrenergic receptor loci. In unrelated individuals, we measured untreated blood pressure and characterized each subject as hypertensive or normotensive. We then used genomic DNA to identify β2- and α2c10-adrenergic receptor restriction fragment length polymorphisms. In 175 subjects (49% with hypertension, 55% black), both hypertension and race were associated with genotype at the β2 locus (χ2 for hypertension=11, P=.004; χ2 for race=8.8, P=.012). The association with hypertension persisted in each race group separately (blacks only: χ2=9.6, P=.008; whites only: χ2=14.2, P=.001). This association persisted in a logistic model that controlled for race (P=.01). Genotype was also significantly associated with baseline systolic, diastolic, and mean arterial blood pressures (P=.05, .01, and .02, respectively). These data suggest that the β2-adrenergic receptor gene is a candidate gene for hypertension in blacks and whites. We also genotyped subjects at the α2-adrenergic receptor coded on chromosome 10. There was no association between hypertension and genotype at the α2c10 locus in the total group or in blacks, but there was significant association in whites (χ2=6.7, P=.03). These data suggest that the β2- and α2c10-adrenergic receptor genes may contribute, in a race-specific manner, to the inheritance of essential hypertension. Linkage studies in related individuals are needed to confirm these findings.
Hypertension is a complex trait with multiple environmental and genetic influences.1 Theoretically, any gene encoding a protein involved in blood pressure regulation is a candidate gene for essential hypertension. Among this wide array of candidate genes are those encoding the adrenergic receptors (ADRs). ADRs are located on vascular smooth muscle cells and in various cells in the kidney.2 They regulate blood pressure through effects on vascular tone, renal sodium excretion, and renin release.3 ADR abnormalities have been implicated in experimental and human hypertension, and the number and/or function of these receptors may differ in different racial groups. At the time the present study was initiated (1990), there were new reports of polymorphisms at two ADR loci—β2 on chromosome 5, and α2 on chromosome 104 5 —and technical resources were available at our center for the investigation of these genes. Therefore, we investigated the hypothesis that genotype at these ADR loci is associated with essential hypertension in unrelated blacks and whites.
Unrelated individuals with and without known hypertension were recruited through the Duke Hypertension Center, public media, and word of mouth. In addition, subjects were recruited from among participants in other ongoing research projects that did not involve the use of antihypertensive medication. Recruitment efforts were specifically designed to recruit approximately equal numbers of black and white individuals and equal numbers of women and men. Subjects were classified as “black” or “white” based on self-report. Individuals were considered black if they identified themselves as “black” or “African American” and non-Hispanic. Individuals were considered white if they identified themselves as nonblack and not Asian, Native American, Hispanic, or Arabic.
Subjects were eligible if they were at least 18 years of age and generally in good health. They were excluded if there was a history of malignant or accelerated hypertension; secondary hypertension; or cardiac, renal, or cerebrovascular disease. Subjects were also excluded if they were pregnant or nursing. This study was approved by the Committee for Clinical Investigation (IRB) of Duke University Medical Center, and all subjects gave signed informed consent.
Subjects were characterized as hypertensive or normotensive on the basis of the average of duplicate blood pressure measurements taken on two occasions at least 1 week apart. Hypertension was defined as an average diastolic pressure greater than 95 mm Hg. Normotension was defined as an average systolic pressure less than 140 mm Hg and average diastolic pressure less than 90 mm Hg in the absence of any blood pressure–lowering medication. Blood pressure was measured with subjects in the seated position by trained personnel using a standard mercury sphygmomanometer and an appropriately sized cuff. In 164 of 175 subjects, hypertension status was defined in this manner after the subject was free of antihypertensive or other blood pressure–lowering medications for at least 2 weeks. However, in the remaining 11 subjects, either severe and poorly controlled hypertension or target-organ damage (eg, history of cerebrovascular accident) prohibited medication withdrawal. These subjects were classified as hypertensive on the basis of elevated blood pressure while taking antihypertensive medication and/or clearly documented severe hypertension on more than one occasion before antihypertensive medication was initiated.
In the 164 participants in whom all blood pressure–lowering medication was withdrawn for at least 2 weeks (94% of the study population), we estimated the baseline untreated blood pressure in one of three ways, depending on the subject’s route of entry into the study. In 121 of these subjects (74%) who were recruited specifically for this project, we performed ambulatory blood pressure monitoring for 6 morning hours (12 measurements per subject) and defined untreated blood pressure as the average of these measurements. In 28 subjects (17%) who were recruited through their participation in a related study of blood pressure regulation, baseline blood pressure was the average of four seated measurements obtained during a single morning with a Dinamap automatic device (Critikon Inc). In 15 subjects (9%) who were unwilling to undergo ambulatory monitoring, we defined baseline blood pressure as the average of duplicate seated measurements taken in the clinic on three separate occasions (total of six measurements). All baseline blood pressure measurements were used in regression analyses only (see below); the primary outcome variable was hypertension status, as defined above.
Demographic data were obtained by interview. Height and weight were measured with a standard stadiometer and digital scale, respectively. Each subject was asked to provide a 24-hour urine collection and simultaneous blood specimen for measurement of creatinine clearance. Twenty-four-hour urinary sodium and potassium excretions were measured to provide a crude estimate of dietary intake.
Genotyping was performed by identification of restriction fragment length polymorphisms. We obtained whole blood from each subject, from which we extracted genomic DNA with a Stratagene kit (No. 200600). Genomic clones containing the β2- and α2c10-ADR genes, obtained from the laboratory of Dr Robert J. Lefkowitz (Duke University Medical Center, Durham, NC), were used as probes. DNA was digested with commercially available restriction enzymes that had been previously reported to detect polymorphisms of the β2- and α2c10-ADR genes.4 5 The products were then separated by 1% agarose gel electrophoresis and submitted to Southern blot analysis.
Continuous variables were compared across race and hypertension status by the Wilcoxon rank sum test. (We chose this nonparametric method to avoid assumptions about distribution of these variables in our population; results with the unpaired Student’s t test were similar.) Proportions were compared across race and hypertension status by χ2 tests. To test for association between genotype and phenotype (ie, hypertension status or race), we calculated likelihood ratio χ2 statistics. (Similar results were obtained with the Pearson χ2.) We then tested for association between genotype and hypertension status, controlling for race, with logistic regression models. In addition, we considered blood pressure as a continuous variable and tested for association with genotype, again controlling for race, with linear regression modeling.
One hundred seventy-five individuals participated in this study. Eighty-six subjects (49%) were hypertensive, and 96 (55%) were black. The study was designed to have approximately equal numbers of subjects in each race/diagnosis category. In fact, there were 46 (26%) black hypertensive subjects, 50 (29%) black normotensive subjects, 35 (20%) white hypertensive subjects, and 44 (25%) white normotensive subjects. We were not quite as successful in achieving gender balance: 65 study subjects (37%) were male, and 110 (63%) were female. Table 1⇓ shows baseline characteristics for the total study population by hypertension status and race. Hypertensive subjects were older than normotensive subjects and as expected had higher blood pressure. By design, kidney function was normal in these subjects (serum creatinine ranged from 53 to 150 μmol/L [0.6 to 1.7 mg/dL]). Within the normal range, hypertensive subjects had significantly higher serum creatinine than normotensive subjects (93±18 versus 88±18 μmol/L, P<.05) but similar creatinine clearance. There were no statistically significant differences between hypertensive and normotensive subjects with respect to sex or race distribution, obesity index (body mass index), or urinary excretions of sodium and potassium.
Table 1⇑ also demonstrates that compared with white subjects, black subjects were younger and more likely to be female. There was no difference between black and white subjects with respect to the proportion with hypertension (48% of blacks, 51% of whites), mean blood pressure, renal function, and obesity index. Blacks had lower urinary sodium and potassium excretions than whites, which, given the imprecision of a single measurement, may or may not reflect differences in dietary intake.
Table 2⇓ provides baseline data by hypertension status within each race group. These data demonstrate that women were particularly overrepresented among blacks without hypertension. The severity of hypertension was similar in blacks and whites.
Digestion with enzyme Ban I revealed two previously reported allelic forms of the β2-ADR gene: a 3.7-kb fragment and a 3.4-kb fragment.4 One hundred seventy-three subjects were genotyped at this locus. In the total study population, the frequency of the 3.7-kb allele was 0.46, and the frequency of the 3.4-kb allele was 0.54. The heterozygous 3.7/3.4-kb genotype was present in 73% of all subjects.
Table 3⇓ shows the distribution of β2-ADR genotype by hypertension status and race. Both hypertension status and race were associated with genotype (χ2 for hypertension status=11, P=.004; χ2 for race=8.8, P=.012). High blood pressure and black race were both associated with a relatively increased frequency of the heterozygous 3.7/3.4-kb genotype and decreased frequency of the homozygous 3.4/3.4-kb genotype.
In light of the association between β2-ADR genotype and race, we tested for association in blacks and whites separately. Table 4⇓ demonstrates that the association between β2-ADR genotype and hypertension persisted in both race groups (χ2 in blacks=9.6, P=.008; χ2 in whites=14.2, P=.001). Furthermore, when we controlled for race in a logistic model, the association between hypertension status and β2 genotype remained significant at the P=.01 level.
Interestingly, the difference in genotype distribution in hypertensive and normotensive subjects appears to be race specific. Hypertension in blacks was associated with increased frequency of the homozygous 3.7/3.7-kb genotype (17% in hypertensive subjects versus 6% in normotensive subjects) and decreased frequency of the homozygous 3.4/3.4-kb genotype (2% versus 18%), whereas the frequency of the heterozygous genotype was approximately the same in hypertensive and normotensive blacks (81% versus 76%). In contrast, among white hypertensive subjects, the 3.7/3.7-kb genotype was underrepresented (0% in hypertensive subjects versus 13% in normotensive subjects), as was the 3.4/3.4-kb homozygous genotype (16% versus 38%), and the heterozygous genotype was overrepresented (84% versus 49%). These patterns are graphically depicted in the Figure⇓.
The analyses discussed above are based on categorization of an individual as hypertensive or normotensive based on an arbitrary blood pressure threshold. In addition, we considered blood pressure as a continuous variable in linear regression models. In regression models controlling for race, baseline systolic, diastolic, and mean arterial blood pressures were significantly associated with genotype at the β2-ADR (P=.05, .01, and .02, respectively). This form of analysis also allowed us to consider the possibility that inequalities existed in the distribution of age, sex, and race with regard to hypertension status (ie, the fact that young black women were overrepresented in our population). In linear regression models that simultaneously controlled for age, sex, and race, mean arterial and diastolic pressures remained significantly associated with β2-ADR genotype (P=.04, and .03, respectively). These data consistently suggest a race-specific association between genotype at the β2-ADR locus and hypertension, suggesting a role for this gene in the inheritance of essential hypertension in both blacks and whites.
We also tested for an association between hypertension and genotype at the α2-ADR coded on chromosome 10 (α2c10). Digestion with restriction enzyme Dra I revealed two allelic forms of the α2c10-ADR gene: a 6.7-kb fragment and a 6.3-kb fragment. One hundred seventy-two subjects were genotyped at this locus (Table 5⇓). Table 5⇓ demonstrates that there was no significant association between hypertension status and genotype at the α2c10-ADR locus in the group as a whole (P=.16). Similarly, there was no association between α2c10-ADR genotype and baseline blood pressure as a continuous variable. However, there was an association between race and α2c10-ADR genotype (χ2=9.1, P=.01). When each race group was evaluated separately (Table 6⇓), the lack of association with hypertension persisted among blacks, but there was a significant association of genotype and hypertension among whites (χ2=6.7, P=.03). Among whites with hypertension, the homozygous 6.7/6.7-kb genotype was relatively overrepresented (82% in hypertensive subjects versus 62% in normotensive subjects), whereas the heterozygous 6.7/6.3-kb (18% versus 30%) and homozygous 6.3/6.3-kb (0% versus 8%) genotypes were underrepresented. Although there are limitations in the interpretation of association data, as discussed below, these data suggest that the α2c10-ADR locus is a candidate gene for hypertension in whites.
Human essential hypertension is multifactorial and may be influenced by several genes with additive or interacting effects. A large number of candidate genes have been proposed in the study of essential hypertension, corresponding to the large number of blood pressure regulatory mechanisms. The candidates we selected are genes that encode ADRs. The ADRs normally regulate blood pressure through effects on systemic and renal vascular resistance, glomerular filtration, renin release, and tubular reabsorption of sodium.3 We chose these genes because of their location and function and because of data implicating receptor abnormalities in genetic hypertension.2
We found significant associations between β2-ADR genotype and both race and hypertension. The racial difference in genotype distribution did not account for the association between this locus and hypertension, which was present in each race group independently and persisted in models that controlled for race. The genotype pattern associated with hypertension in blacks differed from the pattern associated with hypertension in whites.
These race-specific associations do not necessarily mean that genetic variants of the β2-ADR locus lead to hypertension. Genetic associations in unrelated individuals could merely indicate that individuals of similar genetic background share environmental exposures that affect blood pressure. Alternatively, these associations could merely indicate that the β2-ADR gene is in genetic disequilibrium with a true hypertension gene. Genetic association must be confirmed in genetic linkage studies (in progress). However, even if genetic linkage is demonstrated in subsequent studies, a role for the β2-ADR gene (and not a gene “nearby” on the chromosome) in the development of hypertension requires the demonstration that a linked variant encodes an abnormal gene product and that the abnormal gene product promotes high blood pressure.
Despite the preliminary nature of our association data, the β2-ADR gene is a candidate for hypertension from a pathophysiological point of view. Alterations of β2-receptors have been reported in some models of hypertension, leading to the hypothesis that blood pressure elevation is the result of impaired receptor-mediated vasorelaxation.6 The data implicating this receptor are particularly intriguing with regard to hypertension in blacks. For instance, β2-ADR density is increased (perhaps reflecting decreased sensitivity) in skin fibroblast cultures from humans with a pressor response to sodium loading.7 These data implicate the β2-ADR in a subtype of genetic hypertension (salt sensitivity) that is present in 70% of hypertensive blacks.8 Similarly, a change in receptor density or sensitivity in vascular smooth muscle cells could lead to another subtype of hypertension frequently found in blacks, vascular hyperreactivity.9 However, the relevance of these findings to our data is purely speculative because we did not assess the degree of salt sensitivity or vascular reactivity in our subjects.
A role for β2-ADR abnormalities in hypertension in blacks is also supported by recent data concerning the effects of ethnicity and hypertension on β2-ADR function. Mills and colleagues10 demonstrated that black hypertensive subjects have more sensitive β-receptors (assessed by isoproterenol-stimulated cAMP in lymphocytes) and higher receptor density in lymphocytes than normotensive blacks and both normotensive and hypertensive whites. The increased receptor sensitivity was apparently intrinsic to the receptor itself (or its coupling to the Gs nucleotide) because there were no group differences in postreceptor adenylate cyclase activation, measured by forskolin-stimulated cAMP production.
It is unclear to what extent these findings in blacks are relevant to the observed association between β2-ADR genotype and hypertension in whites. Clearly, blacks comprise a heterogeneous population that is often the result of admixture of Caucasian and African ancestry, making it unlikely that blacks and whites represent genetically distinct groups. However, the race-specific patterns of genetic association we observed suggest that distinct ADR variants contribute to hypertension in different subpopulations. The functional significance of the genetic association, the extent to which our observations correlate with reported β2-ADR abnormalities, and the role of race in this process await further evaluation.
We also studied the α2c10-ADR gene. These receptors are located on vascular smooth muscle cells and various cell types in the kidney and are known to influence vascular tone (both systemic and renal), proximal tubular sodium reabsorption, and renin release.11 Therefore, this gene is also an intriguing candidate in the investigation of genetic hypertension. We found no association between hypertension and genotype in the group as a whole or in blacks but significant association in whites. These findings are inconsistent with two previous reports of no association in white and Japanese populations12 13 and one report demonstrating the presence of association in black subjects.14 None of these studies (including our own) is based on a random population sample, and therefore, the observations may be biased by nonrepresentativeness of the populations studied. In fact, Table 2⇑ demonstrates that age and sex are unevenly distributed across race and hypertension status. The effect of this distribution on the observed genetic association is unknown. In addition, a small proportion of our subjects had concomitant diabetes mellitus (<5%, without differences by race or hypertension status), confirming the selected nature of our study population. Alternatively, the lack of association we observed in blacks could simply reflect insufficient statistical power. Nonetheless, the association we observed in whites, presumably differing in numerous clinical and genetic aspects from populations studied previously, suggests that the α2c10-ADR gene may be important in an as yet unidentified subset of the affected population. Additional investigation of this locus is needed.
This study establishes an association between the genes encoding the β2- and α2c10-ADRs and essential hypertension. Our data are consistent with a role of genetic variance at the β2-ADR locus in the pathogenesis of hypertension in blacks and whites and indicate that the β2-ADR locus is a candidate gene for hypertension. In addition, our data suggest a role for the α2c10-ADR in the development of hypertension in whites. Evidence of association allows us to move forward into linkage analysis and further characterization of linked variants. The demonstration of genetic association is an important first step in defining the role of these ADR genes in the inheritance of essential hypertension.
This research was supported in part by Grant-in-Aid No. 92012180 from the American Heart Association and National Institutes of Health (NIH) grants RO1-HL-50176, R29-HL-46218, and P20-AG-12058. Additional support was provided by M01-RR-30 National Center for Research Resources, General Clinical Research Centers Program, NIH. Scientific guidance and adrenergic receptor probes were generously supplied by Dr Robert Lefkowitz, Duke University Medical Center. We thank LaVerne Johnson-Pruden for her assistance with preparation of the manuscript.
Reprint requests to Dr Laura P. Svetkey, Duke Hypertension Center, 3024 Pickett Rd, Durham, NC 27705. E-mail firstname.lastname@example.org.
- Received December 4, 1995.
- Revision received January 22, 1996.
- Accepted February 24, 1996.
Lander ES, Schork NJ. Genetic dissection of complex traits. Science. 1994;265:2037-2048.
Michel MC, Brodde O-E, Insel PA. Peripheral adrenergic receptors in hypertension. Hypertension. 1990;16:107-120.
DiBona GF. Sympathetic nervous system influences on the kidney: role in hypertension. Am J Hypertens. 1989;2:119s-124s.
Lentes K-U, Berrettini WH, Hoehe MR, Chung F-Z, Gershon ES. A biallelic DNA polymorphism of the human beta2-adrenergic receptor detected by Ban I-Adrbr-2. Nucleic Acids Res. 1988;16:2359.
Hoehe MR, Berrettini WH, Lentes K-U. Dra I identifies a two allele DNA polymorphism in the human alpha2-adrenergic receptor gene (ADRAR), using a 5.5 kb probe (p ADRAR). Nucleic Acids Res. 1988;16:9070.
Kotanko P, Hoglinger O, Skrabal F. Beta2-adrenoceptor density in fibroblast culture correlates with human NaCl sensitivity. Am J Physiol. 1992;263:C623-C627.
Weinberger MH, Miller JZ, Luft FC, Grim CE, Fineberg NS. Definitions and characteristics of sodium sensitivity and blood pressure resistance. Hypertension. 1986;8(suppl III):III-127-III-134.
Mills PJ, Dimsdale JE, Ziegler MG, Nelesen RA. Racial differences in epinephrine and β2-adrenergic receptors. Hypertension. 1995;25:88-91.
Umemura S, Hirawa N, Iwamoto T, Yamaguchi S, Toya Y, Kobayashi S, Takasaki I, Yasudi G, Tamura K, Ishii M, Sun L, Pettinger W. Association analysis of restriction fragment length polymorphism for alpha2-adrenergic receptor genes in essential hypertension in Japan. Hypertension. 1994;23(suppl I):I-203-I-206.