This Article Abstract Full Text (PDF) Data Supplement All Versions of this Article: 49/1/96    most recent 01.HYP.0000252029.35106.67v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Rana, B. K. Articles by O’Connor, D. T. Search for Related Content PubMed PubMed Citation Articles by Rana, B. K. Articles by O’Connor, D. T. Pubmed/NCBI databases Gene GEO Profiles HomoloGene UniGene Compound via MeSH Substance via MeSH Medline Plus Health Information High Blood Pressure Related Collections Other hypertension

(Hypertension. 2007;49:96.)
© 2007 American Heart Association, Inc.

Original Articles

Population-Based Sample Reveals Gene–Gender Interactions in Blood Pressure in White Americans

Brinda K. Rana; Paul A. Insel; Samuel H. Payne; Kenneth Abel; Ernest Beutler; Michael G. Ziegler; Nicholas J. Schork; Daniel T. O’Connor

From the Departments of Psychiatry (B.K.R., N.J.S.), Pharmacology (P.A.I., D.T.O.), Computer Science and Engineering (S.H.P.), and Medicine (P.A.I., M.G.Z., D.T.O.), Polymorphism Research Laboratory (B.K.R., N.J.S.), and Center for Human Genetics and Genomics (N.J.S., D.T.O.), University of California at San Diego; the Department of Molecular and Experimental Medicine (E.B.), Scripps Research Institute, La Jolla, Calif; and Sequenom (K.A.), San Diego, Calif.

Correspondence to Daniel T. O’Connor or Nicholas J. Schork, Department of Medicine (0838), University of California at San Diego School of Medicine and VASDHS, 9500 Gilman Dr, La Jolla, CA 92093. E-mail doconnor{at}ucsd.edu or nschork@ucsd.edu

	Abstract

Top Abstract Introduction Methods Results Discussion References

 
The influence of genetic contributors, such as common single nucleotide polymorphisms, on blood pressure and essential hypertension may vary with the gender. We used the power of a large, community-based sample to probe whether gender interacts with genes in contributing to extremes of blood pressure in 611 male and 656 female age-matched white Americans within the top and bottom 5th percentiles of blood pressure among >53 000 people in a health maintenance program. This approach has >90% statistical power to detect genes contributing as little as 3% to trait (blood pressure) variation. We scored 60 000 genotypes in the subjects: 48 single nucleotide polymorphisms at 33 autosomal and 2 X-linked genes in adrenergic and renal pathways that regulate blood pressure. Six individual variants significantly affected blood pressure and demonstrated gene-by-gender interaction, yielding different effects of the single nucleotide polymorphism on blood pressure in males and females. In females, polymorphisms at ß₁-adrenergic receptor and _2A-adrenergic receptor contributed to blood pressure, whereas in men, polymorphisms at ß₂-adrenergic receptor and angiotensinogen were associated. An _2A-adrenergic receptor haplotype influenced blood pressure in women, whereas 2 angiotensinogen haplotypes were associated in men. We also detected gene-by-gene, gender-specific interactions (epistasis) in pathophysiological pathways. This study reveals gender-specific effects of single nucleotide polymorphisms, haplotypes, and gene-by-gene interactions that determine blood pressure in white Americans. Such genetic variants may define genetically and etiologically distinct subgroups of men and women with essential hypertension and may have implications for rational treatment selection.

Key Words: gender • polymorphism • epistasis • essential hypertension • blood pressure • adrenergic receptors

	Introduction

Top Abstract Introduction Methods Results Discussion References

 
Many common diseases exhibit gender bias, and gender differences in the development of common multifactorial (or complex) disorders, such as essential hypertension and other cardiovascular diseases, are the subject of considerable attention.¹ Earlier studies pointed to contributions of gender in vascular physiology² and in response to adrenergic and antihypertensive drugs, such as ₂-adrenergic agonists,^3–6 but the potential interaction of gender and heredity in the determination of elevated blood pressure (BP) is incompletely understood.

The gender of a subject is generally accounted for, usually as a covariate, in association studies, but more recently, studies have begun to identify both autosomal genes and genes involved in sexual dimorphism that differentially contribute to multifactorial traits/diseases or are differentially expressed in tissues in males versus females. For example, Nakayama et al⁷ identified a genetic variant in the follicle-stimulating hormone receptor that may contribute to essential hypertension in females. Peter et al⁸ identified evidence of gender-specific contributions of estrogen-related genes to BP variation. A study of black youth revealed that both systolic BP levels and diastolic BP (DBP) levels were lower in carriers versus noncarriers of the Met 235 allele of the angiotensinogen gene, and this occurred in a gender-specific manner.⁹ Studies in mice have shown that a large number of genes, both autosomal and on sex chromosomes, display gender-specific differences in their pattern of tissue expression.¹⁰ Given this growing body of evidence of a differential role of gender on genetic expression of complex traits, including BP, we systematically investigated the interaction of gender with genetic variants in contributing to BP.

To test the gender dependence of heritable influences on BP, we exploited the statistical power that is possible with a very large sample size. Accordingly, we selected individuals through a sampling design with the power to detect small contributions to BP variation, ascertaining individuals within the highest and lowest 5th percentiles of BP in a primary care setting in southern California with a large community database (>53 000 subjects) of men and women whose BP was measured at routine health maintenance visits.

We then tested the relationship of these BP extremes with 35 loci that have physiological roles in the regulation of BP through the kidney, by components of the renin–angiotensin system and by adrenergic pathways that are activated by catecholamines (Figure IA and IB, available online at http://hyper.ahajournals.org). We, thus, explored a role for gender in the contribution of polymorphism at key renal and adrenergic loci to elevated BP. The data demonstrate pronounced gender effects on the genetic contributions of haplotypes, as well as gene-by-gene (epistatic) interactions.

	Methods

Top Abstract Introduction Methods Results Discussion References

 
Subjects
Subjects were sampled based on phenotype data from 53 078 individuals (27 475 women and 25 538 men) whose medical information was obtained through routine, yearly health appraisal visits to the Kaiser–Permanente medical group, a subscription-based, primary care, health maintenance organization located in San Diego, CA. A total of 81% of enrollees attend this health appraisal clinic, whereas 46% of eligible subjects gave informed consent and were, therefore, considered for enrollment; consenting subjects were slightly older (58±14 versus 51±16 years) and more likely to be men (50% versus 45%) than nonconsenting subjects but did not differ in the frequency of self-reported cardiovascular disease.^11,12 Genomic DNA samples from these subjects were collected from >21 000 individuals.

BP was measured in seated subjects using aneroid sphygmomanometry. If DBP was elevated, repeat measurement was obtained. Blood for preparation of genomic DNA was obtained with informed consent, and samples were anonymized.

Subject Selection Criteria
Men and women were selected from the highest and lowest (extreme) percentiles of DBP distribution. Subjects were ascertained by using DBP as the trait, because twin and family studies provide evidence that DBP is substantially heritable.^13–15 Ethnicity, defined by self-identification, including that of both parents and all 4 grandparents, was specified as white (European) ancestry. Subjects in the upper DBP percentiles were diagnosed with hypertension based on repeated BP measurements and did not have renal failure (serum creatinine concentration was 1.5 mg/dL in 98.6% of subjects). Ages of subjects at the lower extreme did not significantly differ from those selected from the upper extreme. Among the women, 311 with higher DBP (92 mm Hg) met the selection criteria and had values in the upper 4.5 percentile of the overall DBP distribution, whereas 345 were ascertained on the basis of a lower DBP value <60 mm Hg, representing the lower 4.8 percentile. Among the men, 343 were ascertained with higher DBP (96 mm Hg), representing the upper 4.9 percentile of the overall male DBP distribution, whereas 268 were ascertained with lower DBP values corresponding with the lowest 4.2 percentile (<61 mm Hg).

Elevated DBP values were verified by repeated BP measurements; 48% of these hypertensive subjects reported being prescribed and taking 1 antihypertensive drug. Subjects in the lower DBP group did not have histories of hypertension or use antihypertensive drugs. Table 1 summarizes the phenotypes.

View this table: [in this window] [in a new window]   TABLE 1. Phenotypes of Male and Female Subjects, Ascertained Based on DBP Measurements in the 4th to 5th Highest and Lowest Percentiles of DBP (as Described in the Text)

Genotyping
Single nucleotide polymorphism (SNP) sequences were obtained from the public SNP database (http://www.ncbi.nih.nlm.gov/SNP) or by sequencing (Table 2). Genomic DNA was typed using a matrix-assisted laser desorption ionization time-of-flight mass spectrometry system developed by Sequenom according to a published protocol.^16,17 Reproducibility of genotyping was verified with 50 blinded replicate samples.

View this table: [in this window] [in a new window]   TABLE 2. Candidate Gene Polymorphisms for Association With DBP Extremes

View this table: [in this window] [in a new window]   TABLE 2. Continued

Power Analysis
Power calculations were performed using the online genetic power calculator for quantitative trait loci (http://statgen.iop.kcl.ac.uk/gpc/)¹⁸ according to the method described by Schork et al.¹⁹ The power of an association study on the extreme samples was computed under varying disease allele frequencies for type I error rate of 0.05, 0.001, or 0.00000001 ("genome-wide" level), for recessive, additive, and dominant models of inheritance, respectively, using proportion of variance of BP explained by the locus at 0.5%, 1%, 2.5%, and 5% and assuming the marker locus is the actual trait locus (D'=1.0). We determined that this sample has >90% power to detect genotype association with a trait when the genotype contributes as little as 3% to the total variation in men; the power is even higher in the female cohort.

Statistical Analysis
Univariate ANOVA determined the significance of differences observed in DBP measurements between individuals from the high and low BP groups. Two-way ANOVA (covariates: age and body mass index) was used to assess interaction effects of gender and SNPs on DBP, as well as epistatic interactions of 2 genes. Phylogenetics and Sequence Evolution software package and SNPEM were used to infer haplotypes from diploid genotypes for each individual.^20–22

Resampling methods were used to compute the probability that observed associations were not likely because of chance given the number of statistical tests that we pursued.²³ DBP values were independently permutated 500 times to generate an empirically derived distribution of test statistics (ie, ANOVA F test statistics). The observed test statistics were compared with the empirically derived distributions of test statistics to derive P values.

	Results

Top Abstract Introduction Methods Results Discussion References

 
SNP Genotyping
We typed 48 SNPs at 33 autosomal and 2 X-linked (MAOA and MAOB) loci crucial for BP regulation by adrenergic and renal pathways. Table 2 lists the SNPs typed, their gene position, and the frequency of the minor (less frequent) allele.

Gender-Specific Allele Contributions to DBP: Gene-by-Gender Interactions
Of 48 SNPs, 2 (at ADRB2 and AGT) influenced BP only in men, whereas 2 (at ADRA2A and ADRB1) contributed only in women (Figure 1). All 4 SNPs demonstrated significant gender-by-gene interaction effects on BP. Two other SNPs, a second one at AGT (3' untranslated region [UTR] +453 ca) and one at PNMT (–184 ag) also showed significant gene-by-gender interaction, but their effects were not significant in either men or women alone.

View larger version (38K): [in this window] [in a new window]   Figure 1. Gene-by-gender interactive effects on DBP: Single gene associations. Tests showed that (A) ADRB2 (Gly16Arg), (B) ADRA2A (–1297 cg), (C) ADRB1 (Ser49Gly), (D) PNMT (–184 ag), (E) AGT (Met235Thr), and (F) AGT (3'UTR +435 ca) are differentially associated with DBP in men versus women. Bars indicate average DBP (mm Hg, ±1 SEM) for each diploid genotype. The P values, obtained from 2-way ANOVA analysis, are given for each gender, as well as interaction (SNP-by-gender) effects. Numbers of individuals carrying each genotype are given in parenthesis.

View larger version (19K): [in this window] [in a new window]   Figure 1. Continued.

ADRB2 (Arg16Gly) showed a significant gene-by-gender effect on DBP (P<0.020). In men, the codon 16 Arg allele contributed to DBP (P<0.002) in an apparently recessive mode of inheritance: Arg/Arg homozygotes had 6 mm Hg greater DBP than either Arg/Gly heterozygotes or Gly/Gly homozygotes (Figure 1A). Women showed no significant contribution of this ADRB2 SNP to BP; indeed, the female DBP trend was directionally opposite from the male effect of the Arg allele (Figure 1A). Furthermore, permutation analysis was performed for each SNP association test to compute the probability that observed associations are not likely because of chance because of the large number of statistical tests conducted in this study. In the permutation test, DBP values were independently permutated 500 times to generate an empirically derived distribution of test statistics (ANOVA F test statistics) for each SNP. The observed test statistics from the single SNP association test were compared with the empirically derived distributions of test statistics to derive permutation test P values. For ADRB2 (Arg16Gly) in men, only 2 of the 500 permutations resulted in test statistics almost as extreme as our observed test statistic, providing a significance of P<0.004 through the permutation test.

A gene-by-gender interaction was also significant (P<0.03; permutation test P<0.015) for ADRA2A (–1297 cg). Men carrying 2 copies of the –1297 c allele tended toward higher DBP, whereas women with 2 copies of the other allele (–1297 g) had significantly (P<0.03; permutation test P<0.04) higher DBP than women carrying 1 or no –1297 g alleles (Figure 1B). A second SNP at ADRA2A, in the 3' untranslated region of the gene, ADRA2A (+427 ga), did not associate with BP in either gender.

A significant (P<0.02; permutation test P<0.01) gene-by-gender interaction on DBP occurred at ADRB1 (Ser49Gly). In women, individuals with 2 Gly alleles had significantly higher DBP (P<0.04; permutation test P=0.056) than individuals with 1 or no Gly alleles (Figure 1C). In men, a contribution of this SNP was not observed; indeed, the trend was directionally opposite.

We noted a strong (P<0.002; permutation test P<0.0001) gene-by-gender interaction on DBP for AGT Met235Thr and a moderate (P<0.04; permutation test P<0.02) effect for the 3'-UTR AGT (+435 ac) variant (Figure 1E and 1F). Although men carrying 1 or 2 copies of the codon 235 Thr allele had higher DBP than men with no copies of that allele, women showed an opposite trend with lower DBP when carrying 1 or 2 copies of 235 Thr. A trend for AGT (+435 a c) alleles contributing in opposing fashion in men and women was also observed (Figure 1E and 1F).

Effect of Antihypertensive Medication on Genetic Association
Because the use of antihypertensive medication may confound the association of genetic variants with BP, we explored the effects of medication use on our SNP association results: 48% (51% female and 45% male) of the high DBP subjects were taking 1 antihypertensive medications. We first reanalyzed for SNP association in individuals not taking antihypertensive medication: the ADRB2 (Arg16Gly) association remained significant (P<0.006) in men, with the codon 16 Arg allele contributing to DBP in a similar recessive mode of inheritance as when subjects taking antihypertensive medication were included. After exclusion of male subjects taking antihypertensive medication, the mean DBP for each genotype was reduced: average DBP in subjects carrying 2 Arg alleles was 80.2±2 mm Hg, 1 Arg allele was 72±2 mm Hg, and 0 Arg alleles was 75.6±1 mm Hg (compare data in Figure 1A). Women showed no significant contribution of this ADRB2 SNP to BP whether or not subjects treated for hypertension were included. We assessed the average DBP of men taking antihypertensive medication within each ADRB2 (Arg16Gly) genotype and found no significant difference in average DBP among the different genotypes with average DBP for each genotype 101±0.6 mm Hg. However, among the male subjects with high DBP, there were proportionately more (P<0.05, ²) subjects on antihypertensive medication that carried the Arg allele: 36%, 35%, and 29% of subjects on antihypertensive therapy were Arg/Arg, Arg/Gly, or Gly/Gly, respectively, whereas 29%, 29%, and 42% of subjects not being treated with antihypertensive medications were Arg/Arg, Arg/Gly, or Gly/Gly, respectively.

The observed association of the AGT Met235Thr SNP in men remained significant (P=0.039) when individuals on BP medication were excluded from the analyses: average DBP in men carrying 2 Thr alleles was 77.2±2 mm Hg; 1 Thr allele was 77.3±1 mm Hg; and no Thr alleles was 72.5±1 mm Hg. Among the 3 AGT genotypes, the proportion of men with high DPB who were or were not taking antihypertensives was not significantly different from those not being treated. No association at AGT was found in women whether we included or removed from the analysis individuals receiving antihypertensive medication.

Because of the low frequency of the ADRB1 Gly allele (Ser49Gly) and the g allele of the ADRA2A (–1297 cg), we were not able to analyze the data set for these 2 SNPs after excluding female subjects on antihypertensive medication. However, we reanalyzed the data using the method of Cui et al^24,25 in which DBP is additively adjusted depending on the number of antihypertensive drug classes that a subject receives. Using this method, we found that women with 2 copies of the –1297 g alleles had significantly (P=0.026) higher DBP than women carrying 1 or no –1297 g allele, with results concordant with the findings above in which subjects were assessed irrespective of whether they received antihypertensive medication. Using the same approach and in accord with findings above, we found that women with 2 Gly alleles of the ADRB1 (Ser49Gly) SNP had significantly higher DBP (P=0.046) than individuals with 1 or no Gly alleles.

Haplotype and Gender Interactions
Two or more common variants were genotyped at the ADRA2A, ADRB1, PLCB1, AGT, GPRK4, and ROCK2 loci, allowing the inference of haplotypes by 2 different algorithms (PHASE and SNPEM). The 2 approaches gave congruent results: 3 common and 1 rare haplotype of the 4 (2²) possible haplotypes at PLCB1, 4 common haplotypes at AGT, and 3 common haplotypes at ADRA2A and ADRB1 (Table 3). These autosomal haplotype frequencies were not different in men and women. The AGT and ADRA2A haplotypes contributed significantly to DBP, showing significant gender-by-haplotype interactions at these loci (Figure 2). The PLCB1, GPRK4, and ROCK2 haplotypes did not contribute to DBP or systolic BP.

View this table: [in this window] [in a new window]   TABLE 3. Haplotypes in the Blood Pressure Extreme Samples: Haplotype-by-Gender Effects on DBP

View larger version (38K): [in this window] [in a new window]   Figure 2. Haplotype association tests reveal significant haplotype-by-gender interactions. (A and B) 2 common AGT haplotypes; (C) ADRA2A/A haplotype; and (D) ADRB1/C haplotype. Haplotypes are labeled as described in Table 3. Bars indicate average DBP (mm Hg, ±1 SEM) for individuals carrying 0, 1, or 2 copies of the haplotype. The P values, obtained from 2-way ANOVA analysis, are given for each gender, as well as interaction (haplotype-by-gender) effects. Numbers of individuals are given in parenthesis.

Haplotype analysis (Figure 2A and 2B) clarified the trend for association that was observed for the 2 AGT SNPs using single-SNP association tests (see Figure 1E and 1F discussed above). Men carrying the common AGT/B haplotype (Met at codon 235 and adenine [a] at 3'UTR [+435 a/c]), had lower DBP (P<0.012) in a gene dose-dependent manner (Figure 2A), suggesting an additive mode of inheritance for the AGT/B haplotype effect on DBP. By contrast, men carrying 1 or 2 copies of the common haplotype (AGT/A) with opposite alleles at the same SNP positions (Thr at codon 235 and cytosine [c] at the 3'-UTR [+435 a/c]) had significantly higher DBP (P<0.004) than men having no copies of this haplotype, indicating a dominant mode of inheritance (Figure 2B). Remarkably, the same AGT haplotype (AGT/A) that conferred increased DBP in men showed a trend toward lower DBP in women. The AGT/A haplotype demonstrated a strong (P<0.001) gender-by-haplotype effect on DBP, whereas the AGT/B haplotype showed a moderate gender-by-haplotype effect (P<0.020).

The ADRA2A/A haplotype had a significant (P<0.04) gender-by-haplotype effect on DBP (Figure 2C). Women with 1 or 2 copies of this haplotype had significantly lower DBP (P<0.04), implying a dominant mode of inheritance (Figure 2C).

The ADRB1/C haplotype demonstrated significant (P<0.02) gender-by-haplotype effects on DBP (Figure 2D). Similar to our observation with ADRA2A/A, women with 1 or no copies of this haplotype had significantly lower DBP (P<0.05). In men, the presence of 1 or 2 copies of this haplotype showed an opposite trend, that is, toward higher DBP.

Gene-by-Gene Interactions in Physiological Pathways: Epistasis
Of all 595 possible gene-by-gene pairs (n[n–1]/2=35x34/2=595), only 2 showed significant interactive contributions to DBP (Figure 3), and in each case the gene-by-gene interaction depended on gender stratification. Two of 595 represent fewer significant interactions than might be expected by chance alone at the =0.05 level (595/20=30; ²=13.4; degrees of freedom = 1; P=0.0003). However, our sample size was selected for adequate power to detect marker-on-trait associations¹⁹ rather than marker-by-marker interactions to affect traits; hence, our study may not have sufficient statistical power to detect all of the pertinent gene-by-gene interactions.

View larger version (34K): [in this window] [in a new window]   Figure 3. Gene-by-gene interactions (epistasis) in pathways subserving blood pressure. Numbers in each bar represent the number of individuals carrying the combination of genotypes. A, Pathways in series. Interactive effects of SNPs at REN and ACE in the renin–angiotensin pathway on DBP in men but not women. B, Pathways in parallel. Interactive effects of SNPs at ADRA1A (1A-AR) and GNAS1 (Gs) in adrenergic receptor signaling pathways on DBP in women but not men.

REN and ACE encode enzymes in series within the angiotensin pathway. Neither the REN (intron 4, –131 gt) SNP nor the ACE (intron 24, –6 ga) SNP alone contributed to BP in our population; however, the interaction of these 2 SNPs together contributed significantly (P<0.007) to DBP in men (Figure 3A) but not women (P=0.268). Specifically, the REN t allele seems to lower DBP, but only on a background of ACE a/a homozygosity. The interaction effect of gender with this SNP–SNP pair was also significant (P<0.01), suggestive of a gender-specific gene-by-gene interaction.

GNAS and ADRA1A function in parallel adrenergic receptor pathways, and the pattern of epistatic effects also differed between men and women with a significant interaction of gender with this SNP–SNP pair (P<0.03). Although the GNAS (Ile131Ile) and ADRA1A (Arg347Cys) SNPs did not independently associate with DBP in women, the interaction of these SNPs significantly contributed to DBP (P<0.04) in women (Figure 3B) but not men (P=0.375).

	Discussion

Top Abstract Introduction Methods Results Discussion References

 
Problems in Genetic Association for Complex Traits
Association studies have provided insights into the genetic basis of Mendelian diseases, such as monogenic forms of hypertension,²⁶ but such approaches have been less successful in identifying genes responsible for more common disorders, such as essential hypertension, or for common interindividual variations in BP. Moreover, inconsistencies in association studies involving BP have been observed for many of the SNPs included in our study, which emphasized candidate genes encoding products with physiological roles in BP regulation.^27–31

Unique Strengths of This Study
To avoid such problems in identifying genetic variants associated with BP, we took advantage of a combination of several methods: population BP extremes, sampling of both men and women, haplotype structures, and physiological pathway approaches. The power of this study lay in the population-based ascertainment of a very large number of phenotyped individuals.

We sampled men and women from the extremes of diastolic BP among >53 000 people in a large community-based health maintenance system.¹¹ The use of extremes is a powerful approach both theoretically, as well as practically, in genetic association studies.^19,26,32 Schork et al¹⁹ proposed a statistical methodology in which one determines thresholds for sampling from phenotypic extremes and suggested a method for assessing the power of such samples for candidate loci association tests. Using this approach, we estimate that our sample had 90% power to detect loci that contribute as little as 3% to population BP variance.

Haplotype-Based BP Associations
If a disease-causal SNP is not in linkage disequilibrium (LD) with the marker/analyzed SNP, association may be missed by single SNP association tests. However, when multiple SNPs exist in a particular gene, haplotype-based studies can be more powerful than single SNPs to define genetic association.³³ We, therefore, genotyped 2 SNPs at 6 of the 35 candidate genes to infer haplotypes. Consistent with the theoretically demonstrated increased power for haplotype-based association tests, we observed that haplotypes formed by the 2 AGT SNPs showed stronger association with DBP (Figure 2A and 2B) than was observed with either of the 2 individual-SNP association tests (Figure 1E and 1F). Superior association by a particular haplotype might suggest either that both SNPs contribute to the observed effect or that another (as-yet unstudied) SNP in LD with the measured SNPs confers the effect on DBP. Our data cannot distinguish these 2 possibilities.

Role of Gender
Our results reveal a major role for gender in the genetic determination of BP. Consistent with known differences in BP between males and females,^34–37 we found a significant effect of gender alone on DBP in the population extremes. Of additional note is the observation that genes showing association with BP variation differ markedly in effect between males and females (Figure 1). Furthermore, we documented several examples wherein particular alleles were observed to show directionally opposite effects on DBP in males versus females (Figure 1A–1C, 1E, and 1F).

One of the most prominent gender differences in contribution to DBP was observed with the ADRA2A (–1297 cg) variant, located in the promoter 1297 nucleotides upstream of the start codon for ADRA2A, which encodes the _2A-AR (Figure 1B). Haplotype results (Figure 2C) were compatible with the single SNP results. Because single SNP analysis at ADRA2A (3'UTR +427 ga) did not show an effect on DBP, the effect might be attributed to either ADRA2A (–1297 cg) or another (as yet undetermined) variant in close linkage disequilibrium with –1297 cg. Indeed, 14 polymorphic sites in the white American population have been reported recently within the 5-kb region of the ADRA2A gene, including ADRA2A (–1297 cg) and 3'UTR +427 ga, and evidence suggests that haplotypes containing ADRA2A (–1297 cg) and 3'UTR +427 ga alter mRNA and/or protein expression.³⁸

The difference in association of the ADRA2A (–1297 cg) SNP with DBP in men versus women (Figure 1B) may help explain previous physiological/pharmacological observations on differential ₂-adrenergic control of the circulation. For example, King et al³⁹ demonstrated that men and women respond differently to a selective ₂-adrenergic receptor agonist, azepexole.³⁹ In that study, women showed a major "presynaptic" vasodilatory effect to agonist at therapeutic levels, whereas this effect was absent in men. Kneale et al⁶ have also demonstrated differential vasoconstriction and selectivity to adrenergic agonists in men versus women.

We also observed a role for gender in the contribution of an AGT SNP to DBP (Figures 1E, 2A, and 2B). The 235Thr variant was first implicated in hypertension by Jeunemaitre et al.⁴⁰ Our studies are consistent with that report: we found that the 235Thr variant, in either the homozygous or heterozygous state, contributes to elevated DBP in men (Figure 1E and 1F) and that both the 235Thr variant and the AGT haplotype containing 235Thr and the c allele of the 3' UTR +435 c a SNP had even stronger association with DBP; we observed a gender effect on the contribution of the 235Thr variant, either alone or as a part of the AGT/A haplotype, with a directionally opposite effect on DBP in men versus women (Figure 2A and 2B).

Epistasis/Pathways/Gene-by-Gene Interactions
Our candidate gene approach that tested loci encoding molecules in the adrenergic and renin/angiotensin pathways allowed us to assess for gene-by-gene (epistatic) interactions. Of 595 possible gene-by-gene combinations among variants at the 35 loci analyzed, only 2 interactions (REN and ACE and GNAS1 and ADRA1A) were identified, both with gender effects as well. REN and ACE illustrate an interaction between components of a pathway in series (Figure 3A): men carrying 2 t alleles at REN (intron 4, –131 gt) and 1 adenine allele at ACE (intron 24, –6 ga) had significantly lower DBP than other men. Other components of this pathway (AGT and AGTR1) did not display such gene-by-gene interactions, although our sample size was not sufficiently powered to permit analysis of 3-way (gene-by-gene-by-gene) interactions.

Another gene pair displaying both gene-by-gene and gene-by-gender interaction likely functions in parallel pathways: ADRA1A (_1A adrenergic receptor) and GNAS1 (Figure 3B). Although ADRA1A couples to G proteins other than Gs, these 2 gene products likely have related (albeit parallel) functional impacts in vascular smooth myocytes (online supplement Figure IA), wherein norepinephrine activates _1A adrenergic receptors to trigger vasoconstriction, and ß₂ adrenergic and certain other receptors promote vasodilation via their ability to activate GNAS1. Thus, either series (Figure 3A) or parallel (Figure 3B) physiological pathways can harbor genetic effects with epistatic contributions.

Conclusions and Perspectives
Although our phenotyping protocol was simple, this study coordinately harnessed several complementary approaches to identify polymorphisms contributing to BP variation: quantitative trait extremes in a large community-based population, physiological pathway genotyping, haplotyping, and gender representation. The results identified novel evidence of gender- on-gene contributions to BP, as well as pathway-dependent gene-by-gene interactions on the trait. Extensive candidate genotyping also identified novel variants not implicated previously in elevated BP. The results reinforce the pivotal role of gender in complex cardiovascular traits and illustrate how gender can act much like a prism to refract genotype in very different ways toward disease determination. The results imply that development of genotype-based diagnostic and therapeutic indices for hypertension must take gender into account to provide an accurate assessment of the role of genes in the origin, treatment, and consequences of this complex trait.

	Acknowledgments

 
Sources of Funding

This study was supported by grants HL58120, HL69758, and RR00827 from the National Institutes of Health.

Disclosures

None.

Received June 12, 2006; first decision July 5, 2006; accepted October 31, 2006.

	References

Top Abstract Introduction Methods Results Discussion References

 

1. Mendelsohn ME, Karas RH. Molecular and cellular basis of cardiovascular gender differences. Science. 2005; 308: 1583–1587.[Abstract/Free Full Text]
2. Forte P, Kneale BJ, Milne E, Chowienczyk PJ, Johnston A, Benjamin N, Ritter JM. Evidence for a difference in nitric oxide biosynthesis between healthy women and men. Hypertension. 1998; 32: 730–734.[Abstract/Free Full Text]
3. Del Rio G, Verlardo A, Zizzo G, Marrama P, Della Casa L. Sex differences in catecholamine response to clonidine. Int J Obes Relat Metab Disord. 1993; 17: 465–469.[Medline] [Order article via Infotrieve]
4. Freedman RR, Sabharwal SC, Desai N. Sex differences in peripheral vascular adrenergic receptors. Circ Res. 1987; 61: 581–585.[Abstract/Free Full Text]
5. Kaiser J. Gender in the pharmacy: does it matter? Science. 2005; 308: 1572.[Abstract/Free Full Text]
6. Kneale BJ, Chowienczyk PJ, Brett SE, Coltart DJ, Ritter JM. Gender differences in sensitivity to adrenergic agonists of forearm resistance vasculature. J Am Coll Cardiol. 2000; 36: 1233–1238.[Abstract/Free Full Text]
7. Nakayama T, Kuroi N, Sano M, Tabara Y, Katsuya T, Ogihara T, Makita Y, Hata A, Yamada M, Takahashi N, Hirawa N, Umemura S, Miki T, Soma M. Mutation of the follicle-stimulating hormone receptor gene 5'-untranslated region associated with female hypertension. Hypertension. 2006; 48: 512–518.[Abstract/Free Full Text]
8. Peter I, Shearman AM, Zucker DR, Schmid CH, Demissie S, Cupples LA, Larson MG, Vasan RS, D’Agostino RB, Karas RH, Mendelsohn ME, Housman DE, Levy D. Variation in estrogen-related genes and cross-sectional and longitudinal blood pressure in the Framingham Heart Study. J Hypertens. 2005; 23: 2193–2200.[Medline] [Order article via Infotrieve]
9. Wang X, Zhu H, Dong Y, Treiber FA, Snieder H. Effects of angiotensinogen and angiotensin II type I receptor genes on blood pressure and left ventricular mass trajectories in multiethnic youth. Twin Res Hum Genet. 2006; 9: 393–402.[CrossRef][Medline] [Order article via Infotrieve]
10. Yang X, Schadt EE, Wang S, Wang H, Arnold AP, Ingram-Drake L, Drake TA, Lusis AJ. Tissue-specific expression and regulation of sexually dimorphic genes in mice. Genome Res. 2006; 16: 995–1004.[Abstract/Free Full Text]
11. Waalen J, Felitti V, Gelbart T, Ho NJ, Beutler E. Penetrance of hemochromatosis. Blood Cells Mol Dis. 2002; 29: 418–432.[CrossRef][Medline] [Order article via Infotrieve]
12. Waalen J, Felitti V, Gelbart T, Ho NJ, Beutler E. Prevalence of coronary heart disease associated with HFE mutations in adults attending a health appraisal center. Am J Med. 2002; 113: 472–479.[CrossRef][Medline] [Order article via Infotrieve]
13. Evans A, Van Baal GC, McCarron P, DeLange M, Soerensen TI, De Geus EJ, Kyvik K, Pedersen NL, Spector TD, Andrew T, Patterson C, Whitfield JB, Zhu G, Martin NG, Kaprio J, Boomsma DI. The genetics of coronary heart disease: the contribution of twin studies. Twin Res. 2003; 6: 432–441.[CrossRef][Medline] [Order article via Infotrieve]
14. Kupper N, Willemsen G, Riese H, Posthuma D, Boomsma DI, de Geus EJ. Heritability of daytime ambulatory blood pressure in an extended twin design. Hypertension. 2005; 45: 80–85.[Abstract/Free Full Text]
15. Snieder H, Harshfield GA, Treiber FA. Heritability of blood pressure and hemodynamics in African- and European-American youth. Hypertension. 2003; 41: 1196–1201.[Abstract/Free Full Text]
16. Buetow KH, Edmonson M, MacDonald R, Clifford R, Yip P, Kelley J, Little DP, Strausberg R, Koester H, Cantor CR, Braun A. High-throughput development and characterization of a genomewide collection of gene-based single nucleotide polymorphism markers by chip-based matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. Proc Natl Acad Sci U S A. 2001; 98: 581–584.[Abstract/Free Full Text]
17. Le Corre P, Parmer RJ, Kailasam MT, Kennedy BP, Skaar TP, Ho H, Leverge R, Smith DW, Ziegler MG, Insel PA, Schork NJ, Flockhart DA, O’Connor DT. Human sympathetic activation by alpha2-adrenergic blockade with yohimbine: bimodal, epistatic influence of cytochrome P450-mediated drug metabolism. Clin Pharmacol Ther. 2004; 76: 139–153.[CrossRef][Medline] [Order article via Infotrieve]
18. Purcell S, Cherny SS, Sham PC. Genetic power calculator: design of linkage and association genetic mapping studies of complex traits. Bioinformatics. 2003; 19: 149–150.[Abstract/Free Full Text]
19. Schork NJ, Nath SK, Fallin D, Chakravarti A. Linkage disequilibrium analysis of biallelic DNA markers, human quantitative trait loci, and threshold-defined case and control subjects. Am J Hum Genet. 2000; 67: 1208–1218.[Medline] [Order article via Infotrieve]
20. Stephens M, Donnelly P. A comparison of bayesian methods for haplotype reconstruction from population genotype data. Am J Hum Genet. 2003; 73: 1162–1169.[CrossRef][Medline] [Order article via Infotrieve]
21. Stephens M, Smith NJ, Donnelly P. A new statistical method for haplotype reconstruction from population data. Am J Hum Genet. 2001; 68: 978–989.[CrossRef][Medline] [Order article via Infotrieve]
22. Fallin D, Cohen A, Essioux L, Chumakov I, Blumenfeld M, Cohen D, Schork NJ. Genetic analysis of case/control data using estimated haplotype frequencies: application to APOE locus variation and Alzheimer’s disease. Genome Res. 2001; 11: 143–151.[Abstract/Free Full Text]
23. Good P. Permutation Tests: A Practical Guide to Resampling Methods for Testing Hypotheses. New York, NY: Springer-Verlag; 1994.
24. Cui J, Hopper JL, Harrap SB. Genes and family environment explain correlations between blood pressure and body mass index. Hypertension. 2002; 40: 7–12.[Abstract/Free Full Text]
25. Cui JS, Hopper JL, Harrap SB. Antihypertensive treatments obscure familial contributions to blood pressure variation. Hypertension. 2003; 41: 207–210.[Abstract/Free Full Text]
26. Lifton RP, Gharavi AG, Geller DS. Molecular mechanisms of human hypertension. Cell. 2001; 104: 545–556.[CrossRef][Medline] [Order article via Infotrieve]
27. Busch CP, Harris SB, Hanley AJ, Zinman B, Hegele RA. The ADD1 G460W polymorphism is not associated with variation in blood pressure in Canadian Oji-Cree. J Hum Genet. 1999; 44: 225–229.[CrossRef][Medline] [Order article via Infotrieve]
28. Fornage M, Turner ST, Sing CF, Boerwinkle E. Variation at the M235T locus of the angiotensinogen gene and essential hypertension: a population-based case-control study from Rochester, Minnesota. Hum Genet. 1995; 96: 295–300.[Medline] [Order article via Infotrieve]
29. Frossard PM, Hill SH, Elshahat YI, Obineche EN, Bokhari AM, Lestringant GG, John A, Abdulle AM. Associations of angiotensinogen gene mutations with hypertension and myocardial infarction in a gulf population. Clin Genet. 1998; 54: 285–293.[CrossRef][Medline] [Order article via Infotrieve]
30. Glorioso N, Manunta P, Filigheddu F, Troffa C, Stella P, Barlassina C, Lombardi C, Soro A, Dettori F, Parpaglia PP, Alibrandi MT, Cusi D, Bianchi G. The role of alpha-adducin polymorphism in blood pressure and sodium handling regulation may not be excluded by a negative association study. Hypertension. 1999; 34: 649–654.[Abstract/Free Full Text]
31. Morrison AC, Bray MS, Folsom AR, Boerwinkle E. ADD1 460W allele associated with cardiovascular disease in hypertensive individuals. Hypertension. 2002; 39: 1053–1057.[Abstract/Free Full Text]
32. Ranade K, Shue WH, Hung YJ, Hsuing CA, Chiang FT, Pesich R, Hebert J, Olivier M, Chen YD, Pratt R, Olshen R, Curb D, Botstein D, Risch N, Cox DR. The glycine allele of a glycine/arginine polymorphism in the beta2-adrenergic receptor gene is associated with essential hypertension in a population of Chinese origin. Am J Hypertens. 2001; 14: 1196–1200.[CrossRef][Medline] [Order article via Infotrieve]
33. Akey J, Jin L, Xiong M. Haplotypes vs single marker linkage disequilibrium tests: what do we gain? Eur J Hum Genet. 2001; 9: 291–300.[CrossRef][Medline] [Order article via Infotrieve]
34. August P, Oparil S. Hypertension in women. J Clin Endocrinol Metab. 1999; 84: 1862–1866.[Free Full Text]
35. Calhoun DA, Oparil S. The sexual dimorphism of high blood pressure. Cardiol Rev. 1998; 6: 356–363.[Medline] [Order article via Infotrieve]
36. Kennedy BP, Rao F, Botiglieri T, Sharma S, Lillie EO, Ziegler MG, O’Connor DT. Contributions of the sympathetic nervous system, glutathione, body mass and gender to blood pressure increase with normal aging: influence of heredity. J Hum Hypertens. 2005; 19: 951–969.[CrossRef][Medline] [Order article via Infotrieve]
37. Maris ME, Melchert RB, Joseph J, Kennedy RH. Gender differences in blood pressure and heart rate in spontaneously hypertensive and Wistar-Kyoto rats. Clin Exp Pharmacol Physiol. 2005; 32: 35–39.[CrossRef][Medline] [Order article via Infotrieve]
38. Small KM, Brown KM, Seman CA, Theiss CT, Liggett SB. Complex haplotypes derived from noncoding polymorphisms of the intronless alpha2A-adrenergic gene diversify receptor expression. Proc Natl Acad Sci U S A. 2006; 103: 5472–5477.[Abstract/Free Full Text]
39. King D, Etzel JP, Chopra S, Smith J, Cadman PE, Rao F, Funk SD, Rana BK, Schork NJ, Insel PA, O’Connor DT. Human response to alpha2-adrenergic agonist stimulation studied in an isolated vascular bed in vivo: biphasic influence of dose, age, gender, and receptor genotype. Clin Pharmacol Ther. 2005; 77: 388–403.[CrossRef][Medline] [Order article via Infotrieve]
40. Jeunemaitre X, Soubrier F, Kotelevtsev YV, Lifton RP, Williams CS, Charru A, Hunt SC, Hopkins PN, Williams RR, Lalouel JM, Corvol P. Molecular basis of human hypertension: role of angiotensinogen. Cell. 1992; 71: 169–180.[CrossRef][Medline] [Order article via Infotrieve]
41. Bonnardeaux A, Davies E, Jeunemaitre X, Fery I, Charru A, Clauser E, Tiret L, Cambien F, Corval P, Soubrier F. Angiotensin II type 1 receptor gene polymorphisms in human essential hypertension. Hypertension. 1994; 24: 63–69.[Abstract/Free Full Text]
42. Siffert W, Rosskopf D, Siffert G, Busch S, Moritz A, Erbel R, Sharma AM, Ritz E, Wichmann HE, Jakobs KH, Horsthemke B. Association of a human G-protein beta3 subunit variant with hypertension. Nat Genet. 1998; 18: 45–48.[CrossRef][Medline] [Order article via Infotrieve]

This article has been cited by other articles:

				S. L. Demel and J. J. Galligan Impaired Purinergic Neurotransmission to Mesenteric Arteries in Deoxycorticosterone Acetate-Salt Hypertensive Rats Hypertension, August 1, 2008; 52(2): 322 - 329. [Abstract] [Full Text] [PDF]

				D. T. O'Connor, G. Zhu, F. Rao, L. Taupenot, M. M. Fung, M. Das, S. K. Mahata, M. Mahata, L. Wang, K. Zhang, et al. Heritability and Genome-Wide Linkage in US and Australian Twins Identify Novel Genomic Regions Controlling Chromogranin A: Implications for Secretion and Blood Pressure Circulation, July 15, 2008; 118(3): 247 - 257. [Abstract] [Full Text] [PDF]

				K. Norris and A. R. Nissenson Race, Gender, and Socioeconomic Disparities in CKD in the United States J. Am. Soc. Nephrol., July 1, 2008; 19(7): 1261 - 1270. [Abstract] [Full Text] [PDF]

				C. Zeng, V. A. M. Villar, G. M. Eisner, S. M. Williams, R. A. Felder, and P. A. Jose G Protein-Coupled Receptor Kinase 4: Role in Blood Pressure Regulation Hypertension, June 1, 2008; 51(6): 1449 - 1455. [Full Text] [PDF]

				P.-a. B. Shih and D. T. O'Connor Hereditary Determinants of Human Hypertension: Strategies in the Setting of Genetic Complexity Hypertension, June 1, 2008; 51(6): 1456 - 1464. [Full Text] [PDF]

				M. Coylewright, J. F. Reckelhoff, and P. Ouyang Menopause and Hypertension: An Age-Old Debate Hypertension, April 1, 2008; 51(4): 952 - 959. [Full Text] [PDF]

				O. Seda, J. Tremblay, D. Gaudet, P.-L. Brunelle, A. Gurau, E. Merlo, L. Pilote, S. N. Orlov, F. Boulva, M. Petrovich, et al. Systematic, Genome-Wide, Sex-Specific Linkage of Cardiovascular Traits in French Canadians Hypertension, April 1, 2008; 51(4): 1156 - 1162. [Abstract] [Full Text] [PDF]

				M. M. Fung, C. Nguyen, P. Mehtani, R. M. Salem, B. Perez, B. Thomas, M. Das, N. J. Schork, S. K. Mahata, M. G. Ziegler, et al. Genetic Variation Within Adrenergic Pathways Determines In Vivo Effects of Presynaptic Stimulation in Humans Circulation, January 29, 2008; 117(4): 517 - 525. [Abstract] [Full Text] [PDF]

				A. Y. Deng Genetic basis of polygenic hypertension Hum. Mol. Genet., October 15, 2007; 16(R2): R195 - R202. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Data Supplement All Versions of this Article: 49/1/96    most recent 01.HYP.0000252029.35106.67v1 A