Polymorphisms and Haplotypes of the Estrogen Receptor-beta Gene (ESR2) and Cardiovascular Disease in Men and Women

Abstract

Background

Cohort studies suggest an association between variation in the estrogen receptor-alpha gene (ESR1) and cardiovascular disease (CVD), but data are lacking for the effect of variation in the estrogen receptor-beta gene (ESR2).

Methods

Three polymorphisms of the ESR2 gene, and their associated haplotypes, were evaluated in 296 white women from the Women’s Health Study and 566 white men from the Physicians’ Health Study who developed CVD [myocardial infarction (MI) or ischemic stroke], each matched 1:1 to a member of the cohort study who remained free from CVD. Blood samples and cardiovascular risk information were collected at baseline.

Results

Women, but not men, who developed CVD or MI, but not ischemic stroke, were more likely to have the rs1271572 polymorphism variant T allele (p=0.05 and 0.02) and less likely to have the rs1256049 polymorphism variant A allele (p=0.003 and 0.004). No associations were observed for rs4986938. In conditional logistic multivariate regression, the rs1271572 variant was associated with increased odds of CVD [odds ratio (OR)=1.49, 95% CI: 1.10–2.01] and MI (OR=1.46, 95% CI: 0.96–2.23), whereas the rs1256049 variant was associated with decreased odds of CVD (OR=0.37, 95% CI: 0.17–0.79) and MI (OR=0.25, 95% CI: 0.09–0.73) in women. A common haplotype that included the rs1271572 variant was associated with a 7-fold increased risk of MI in women.

Conclusions

Two tightly linked polymorphisms of ESR2 were associated with risk of CVD, particularly MI, in women but not men. Further studies of ESR2 genetic variation and risk of CVD are warranted.

INTRODUCTION

The development of atherothrombotic cardiovascular disease (CVD) is likely polygenic. The cardiovascular effects of estrogen are mediated through binding to specific estrogen receptors at the cytosolic and nuclear level. Two separate estrogen receptors have been identified: estrogen receptor-alpha (ER-alpha) and estrogen receptor-beta (ER-beta). Although polymorphisms of the ER-alpha gene (ESR1) have recently received attention as possible contributors to CVD risk in men and women;(1–8), the relationship between genetic variation of the ER-beta gene (ESR2) and CVD has not been as well-studied.

Several lines of evidence support a potential role of ESR2 in CVD. ER-beta is highly expressed in endothelial and vascular smooth muscle cells (9). ER-beta has been associated with coronary plaque (9). In autopsy studies, ER-beta expression positively correlates with increased coronary artery plaque area in both women and men (10, 11). Polymorphisms of ESR2 have been associated with left ventricular mass and left ventricular wall thickness in women but not men (12). To the best of our knowledge, only one published study has evaluated ESR2 and atherosclerosis. In a Brazilian case control study, an ESR2 variant (rs4986938) was more common among cases with premature coronary artery disease than among controls (13).

ESR2, located on chromosome 14 q22–24, is comprised of 8 exons. We evaluated 3 polymorphisms in the ESR2 gene as well as their associated haplotypes, with risk of CVD (defined as myocardial infarction or ischemic stroke) among men in the Physicians’ Health Study and women in the Women’s Health Study, using a nested case-control design. Two of these polymorphisms (rs1256049 and rs4986938) were initially described as restriction fragment length polymorphisms(14) and have been investigated previously for association with coronary heart disease (13). The 3rd polymorphism was chosen because of its location in the promoter region (rs1271572).

METHODS

We employed a nested case-control design within the Physicians’ Health Study (PHS) and the Women’s Health Study (WHS). Initiated in 1982, the PHS is a randomized, double-blinded, placebo-controlled trial of aspirin and beta-carotene among 22071 male, predominantly white, US physicians, 40 to 84 years of age at study entry (15). Before randomization, 14916 participants provided an EDTA-anticoagulated blood sample that was stored for genetic analysis. The WHS, initiated in 1992, is a randomized, double-blinded, placebo-controlled trial of the administration of low-dose aspirin and vitamin E in 39876 female, predominantly white, US health professionals, 45 to 89 years old at study entry (16, 17). Before randomization, 14,916 PHS participants and 28,345 WHS participants each provided an EDTA-anticoagulated blood sample that was stored for genetic analysis.

At study entry, participants in both trials were free of prior myocardial infarction (MI), stroke, transient ischemic attacks, and cancer. Yearly follow-up self-report questionnaires provided reliable updated information on newly developed diseases and the presence or absence of other cardiovascular risk factors. History of cardiovascular risk factors, such as hypertension, diabetes, or hyperlipidemia, was defined by self-report of diagnosis at entry into the study. For all reported incident CVD events (MI or ischemic stroke) occurring after study enrollment, hospital records, death certificates, and autopsy reports were requested and reviewed by an endpoints committee using standardized diagnostic criteria.

The diagnosis of MI was confirmed by evidence of symptoms in the presence of either diagnostic increases of cardiac enzymes or diagnostic changes on electrocardiograms. In the case of fatal events, the diagnosis of MI was also accepted based on autopsy findings. Stroke was defined by the presence of a new focal neurological deficit, with symptoms and signs persisting for ≥24 hours or until death and was ascertained from blinded review of medical records, autopsy results, and the judgment of a board-certified neurologist. Stroke was classified as ischemic, hemorrhagic, or unknown on the basis of clinical reports and computed tomographic or MRI scanning, and only ischemic strokes were used in these analyses. For each case, a control matched by age, smoking history, and length of follow-up was chosen form among those individuals who remained free of vascular diseases. Genetic data were available for 566 case-control pairs.

White men and women who had DNA samples available and who developed CVD during follow-up (cases) were matched 1:1 to white participants who remained free of disease until the case date and were matched for age and smoking history as well as postmenopausal hormone therapy for women. For men there were a total of 566 case-control pairs and for women a total of 296 case-control pairs available for the current analyses.

The study was approved by the Brigham and Women’s Hospital Institutional Review Board for Human Subjects Research.

Genotype Determination

In 2004, 3 polymorphisms in the ESR2 gene were chosen for investigation, based on prior reports in the literature(14) and possible functional significance. Two of these polymorphisms were initially described as restriction fragment length polymorphisms (14), rs1256049, which represents a relatively rare synonymous G→A change at position 1082 in exon 5 (Rsa1 restriction site, also known as G1082A), and rs4986938, which is a G→A change at position 1730 in the 3’UTR of exon 8 (Alu1 restriction site, also known as G1730A). In addition, rs1271572, an A→C transposition in the promoter region, was chosen because its location suggests possible functional significance.

We performed genotype determination with an ABI fluorescence-based allelic discrimination method (Applied Biosystems). Each 10-mL amplification reaction volume contained 1X TaqMan Universal Master Mix (Applied Biosystems) and 10 ng of template DNA. Amplification reactions were carried out in duplicates on an ABI 7900HT Sequence Detection System according to the manufacturer’s specifications.

To confirm genotype assignment, scoring was carried out by 2 independent observers. Discordant results (<1% of all scoring) were resolved by a joint reading, and where necessary, a repeat genotyping. Results were scored blinded as to case-control status.

Statistical Analysis

We used χ²-analysis for comparison of genotype and allele frequencies in cases and controls and tests for Hardy-Weinberg equilibrium. Pairwise linkage disequilibrium (LD) was examined as described by Devlin and Risch (18). Haplotype frequencies were estimated from genotype data using the PHASEv2.1.1 algorithm (19, 20). Haplotype distributions between cases and controls were compared by global likelihood ratio test. In addition, the relationship between haplotypes and clinical outcome was examined by conditional logistic regression analysis using a baseline-parameterization procedure, conditional on age, smoking, and hormone therapy use (WHS only), with further adjustment for body mass index, diabetes, hypertension, and randomized treatment group. For each odds ratio, we calculated 95% CIs. A 2-tailed nominal p-value of 0.05 was considered to represent a statistically significant result. Main results are presented assuming a genotype model (or additive mode of inheritance) with a 3-level categorical variable.

RESULTS

Baseline characteristics of the 566 men who developed CVD (326 MI and 240 ischemic strokes) and 296 women who subsequently developed CVD (156 MI and 140 ischemic strokes) compared with those who remained free of CVD (controls) are shown in Table 1. As expected, both men and women who later developed CVD had a higher baseline prevalence of cardiovascular risk factors.

Table 1.

Baseline characteristics of PHS and WHS study participants who subsequently developed CVD (Cases), and those who remained free of vascular disease during follow-up (controls). All participants were white men (PHS) or women (WHS).

	Male Controls	Male CVD Cases	P	Female Controls	Female CVD Cases	P
	(N=566)	(N=566)		(n=296	(n=296)
Age (years)	60.7±0.4	61.0±0.4	m.v.*	63.6±0.4	63.7±0.4	m.v.*
Smoking Status (%)
Never	42.2	42.2	m.v	40.5	40.5	m.v.
Past	41.5	41.5		41.2	41.2
Current	16.3	16.3		18.2	18.2
Hormones Use (%) – women only
Never	--	--		34.5	34.5	m.v.
Past	--	--		22.3	22.3
Current	--	--		43.2	43.2
Body Mass Index (kg/m2)	24.8±0.1	25.4±0.1	0.002	25.6±0.3	27.0±0.3	0.001
Hyperlipidemia†	15.3	22.7	0.002	50.2	58.2	0.07
Hypertension (%)‡	37.9	46.5	<0.0001	33.7	54.2	<0.0001
Diabetes (%)	2.8	8.9	<0.0001	4.0	12.0	<0.001
Parental history of premature CAD (%)§	8.9	11.6	0.14	9.9	17.0	0.03
Aspirin use (%)	47.4	45.2	0.49	50.7	44.6	0.14

Mean±SE unless otherwise stated.

^*m.v., matching variable.

^†Hyperlipidemia defined as self-report of history of total cholesterol >240mg/dL).

^‡Hypertension defined as physician diagnosis of hypertension or reported BP of >140 mmHg systolic or >90 mmHg diastolic blood pressure.

The observed genotype frequencies were in Hardy-Weinberg equilibrium among the controls. Among women but not men, the rs1271572 and rs1256049 polymorphisms were associated with CVD and MI (Table 2). In women, the T allele of rs1271572 was more common among women who developed CVD (47% of cases vs. 41% of controls, p=0.05) and MI (49% of cases vs. 39% of controls, p=0.02), whereas the A allele of rs1256049 was less common among women who developed CVD (2% of cases vs. 5% of controls, p=0.003) or MI (2% of cases vs. 7% of controls, p=0.004). No associations were observed between rs4986938 allele frequency and outcomes in either men or women. In men, however, the rs4986938 AA genotype was more common among MI cases than controls (19.9% vs. 12.2% with p=0.05).

Table 2.

Genotype and allele frequencies of ESR2 polymorphisms among men and women for cardiovascular disease (CVD), myocardial infarction (MI), and ischemic stroke.

rs1271572	GG	GT	TT	P	G	T	P*
MEN
CVD
cases (n=566)	35.3	44.4	20.3	0.84	0.57	0.43	0.72
controls (n=566)	33.7	46.0	20.3		0.57	0.43
MI
cases (n=326)	35.7	43.2	21.1	0.64	0.57	0.43	0.47
controls (n=326)	32.2	46.2	21.7		0.55	0.45
Ischemic stroke
cases (n=240)	34.7	46.0	19.3	0.97	0.59	0.41	0.78
controls (n=240)	35.6	45.6	18.8		0.57	0.43
WOMEN
CVD
cases (n=296)	27.1	52.8	20.2	0.07	0.53	0.47	0.05
controls (n=296)	31.9	55.0	13.2		0.59	0.41
MI
cases (n=156)	23.9	53.5	22.5	0.02	0.51	0.49	0.02
controls (n=156)	31.7	57.8	10.6		0.61	0.39
Ischemic stroke
cases (n=140)	30.5	51.9	17.6	0.93	0.56	0.44	0.72
controls (n=140)	32.1	51.9	16.0		0.58	0.42
rs1256049	GG	GA	AA		G	A
MEN
CVD
cases (n=566)	94.6	5.4	0.0	0.79	0.97	0.03	0.54
controls (n=566)	93.9	5.9	0.2		0.97	0.03
MI
cases (n=326)	95.8	4.2	0.0	0.32	0.98	0.02	0.21
controls (n=326)	93.8	5.9	0.3		0.97	0.03
Ischemic stroke
cases (n=240)	93.0	7.0	0.0	0.71	0.97	0.03	0.66
controls (n=240)	94.1	5.9	0.0		0.97	0.03
WOMEN
CVD
cases (n=296)	96.2	3.8	0.0	0.01	0.98	0.02	0.003
controls (n=296)	90.4	8.9	0.7		0.95	0.05
MI
cases (n=156)	96.0	4.0	0.0	0.02	0.98	0.02	0.004
controls (n=156)	87.6	11.1	1.3		0.93	0.07
Ischemic stroke
cases (n=140)	96.4	3.7	0.0	0.41	0.98	0.02	0.30
controls (n=140)	93.6	6.4	0.0		0.97	0.03
rs4986938	GG	GA	AA		G	A
MEN
CVD
cases (n=566)	41.0	39.1	19.9	0.49	0.63	0.37	0.59
controls (n=566)	41.3	45.3	13.4		0.64	0.36
MI
cases (n=326)	41.0	39.1	19.9	0.05	0.61	0.39	0.10
controls (n=326)	43.0	44.9	12.2		0.65	0.35
Ischemic stroke
cases (n=240)	41.1	47.6	11.3	0.54	0.65	0.35	0.34
controls (n=240)	38.8	46.6	14.7		0.62	0.38
WOMEN
CVD
cases (n=296)	33.2	49.1	17.7	0.75	0.58	0.42	0.80
controls (n=296)	32.6	51.7	15.6		0.59	0.41
MI
cases (n=156)	31.6	51.3	17.1	0.56	0.57	0.43	0.40
controls (n=156)	33.8	53.6	12.6		0.61	0.39
Ischemic stroke
cases (n=140)	35.0	46.7	18.3	0.83	0.58	0.42	0.60
controls (n=140)	31.4	49.6	19.0		0.56	0.44

Genotype distributions were in Hardy-Weinberg equilibrium.

^*P-values for chi-square test.

To further investigate the observed associations, genotypes were modeled separately in a conditional logistic regression analysis, conditional on age, smoking, and hormone therapy use and further adjusted for randomized treatment group (aspirin and vitamin E in WHS, aspirin in PHS), body mass index, and history of hypertension and diabetes. Among men there were no significant associations between CVD, MI, or ischemic stroke and any of the polymorphisms. For women, the rs1271572 variant was associated with increased risk of MI and CVD (p=0.01 for CVD and p=0.02 for MI). The less common rs1256049 variant was associated with decreased risk of CVD and MI (p=0.01 for CVD and MI). No association was observed for these polymorphisms and ischemic stroke in women.

As shown in Table 4, the polymorphisms were in strong linkage disequilibrium with one another. Five haplotypes had a frequency greater than 1% in this population (Table 5). Further investigation of a possible association was performed using a haplotype-based conditional logistic regression, adjusting for potential confounders including body mass index, smoking, hormone therapy use, randomized treatment group, and history of hypertension and diabetes. A common haplotype (2-1-1) including the variant allele of rs1271572 was associated with increased odds of MI in women (OR =7.91 [95% CI, 1.19–52.4]; p=0.03) (Table 6).

Table 4.

Pairwise linkage disequilibrium (D’) analysis for ESR2 SNPs, for women and men.

	D’ (Women, Men)
	rs1271572	rs1256049	rs4986938
rs1271572	-------	1.00, 0.88	1.00, 0.85
rs1256049		-----	0.77, 0.91
rs4986938			------

Table 5.

ESR2 haplotype distribution by sex and disease status.

	Haplotype frequency (%)					P value*
	1-1-1 †	1-1-2	1-2-1	2-1-1	2-1-2
Men
CVD cases	0.19	0.35	0.03	0.41	0.02	0.62
CVD controls	0.19	0.35	0.03	0.40	0.03
Women
CVD cases	0.24	0.29	0.00	0.32	0.13	0.07
CVD controls	0.25	0.30	0.02	0.30	0.12

^*P-value for global likelihood ratio test

^†1 denotes major allele, 2 denotes minor allele at each site: rs1271572, rs1256049, rs4986938.

Table 6.

Odds ratios^* from conditional logistic regression according to ESR2 haplotype, for cardiovascular disease (CVD) and myocardial infarction (MI).

	Men			Women
Haplotype†	OR*	(95%CI)	P	OR*	(95%CI)	P
CVD
1-1-1	1.0	Referent		1.0	Referent
1-1-2	1.11	(0.82–1.50	0.51	0.77	0.39–1.56	0.47
1-2-1	0.80	0.41–1.57	0.52	--	--	--
2-1-1	1.15	0.85–1.57	0.37	1.33	0.65–2.74	0.44
2-1-2	0.92	0.38–2.24	0.85	1.73	0.90–3.33	0.10
MI
1-1-1	1.0	Referent		1.0	Referent
1-1-2	1.22	0.77–1.92	0.40	4.74	0.64–34.87	0.13
1-2-1	0.48	0.19–1.23	0.13	--	--	--
2-1-1	1.24	0.81–1.91	0.33	7.91	1.19–52.41	0.03
2-1-2	0.85	0.28–2.64	0.78	4.54	0.94–21.96	0.06

^*Conditional on age, smoking, and further controlling for randomized treatment group, BMI, history of hypertension, presence or absence of diabetes.

^†1 denotes major allele, 2 denotes minor allele at each site: rs1271572, rs1256049, rs4986938. Confidence level for haplotype estimation and inference was >=95%.

Discussion

We examined 3 polymorphisms of the ESR2 gene and risk of CVD, myocardial infarction, and ischemic stroke in both men and women. We found significant differences by sex, with no association for any of the polymorphisms or associated haplotypes and risk of CVD in men. In contrast, 2 of the ESR2 polymorphisms were associated with CVD, particularly MI, in women. Women with the variant T allele of rs127152 had significantly increased odds of CVD and of MI, whereas women with the relatively rare A allele of rs1256049 had decreased odds of CVD and MI. There were no associations for any of the polymorphisms and ischemic stroke in women. In haplotype analyses, the odds of MI were 7-fold higher for women with a common haplotype that included the variant allele of rs127152.

Limited, but accumulating, evidence supports a role of ESR2 and atherosclerosis in women. Particularly in women, ER-beta is the estrogen receptor that is predominantly expressed in human vascular smooth muscle (9). In autopsy studies, ER-beta expression positively correlated with increased coronary artery plaque area in both women and men (10, 11). Some studies suggest that increased ESR2 expression may be a secondary or compensatory phenomenon. In rat models, expression of ER-beta, but not ER-alpha, was induced after vascular injury (21). To the best of our knowledge, this study is the largest investigation to date of ESR2 and atherosclerotic CVD. In contrast to our study, a Brazilian case-control study of 153 men and women with premature coronary heart disease age <55 years) found no association between rs1256049 and CVD (p=0.24 for alleles, p=0.83 for genotypes), whereas the rs4986938 A allele was associated with increased risk of CHD; rs127152 was not examined.(13). Although in our study men who developed MIs had a slightly higher prevalence of homozygosity for the AA genotype of rs4986938 than controls (p=0.05), the results were not significant in conditional logistic regression. The rs1256059 polymorphism, with another tightly linked ESR2 polymorphism (1256031), has been related to left ventricular mass in hypertensive women (12). In another study, mean LDL and TC concentrations in women with higher estrogen exposure (premenopausal women or postmenopausal women using hormone therapy) were lower in those who were heterozygous for rs1256049 than in women who were wild-type homozygotes (22). The precise mechanisms by which variations in ESR2 might confer increased risk of CVD are unknown. Estrogen receptors, as members of the large superfamily of nuclear receptors, classically function as ligand-activated transcription factors, interacting with the estrogen response element in a large number of target genes (23). However, rapid nongenomic signaling pathways involving membrane-mediated signaling have also been described (24, 25). Polymorphisms of ESR2 have not been clearly associated with insulin sensitivity, metabolic syndrome (26), lipids(27, 28), or endogenous estradiol concentrations (28), although the ESR2 gene could theoretically affect all of these. One study found that homozygotes of the rs4986938 ESR2 variant had higher body mass index, serum triglycerides, and apolipoprotein B, while having reduced HDL-cholesterol (13). ESR2 was associated with vascular function and hypertension in animal models (29), and ESR2 variation (length of a cytosine-adenine repeat sequence) was associated with hypertension in postmenopausal Japanese women in one study (30). Whether the tested ESR2 polymorphisms have functional significance themselves or are in linkage with other functional genetic variants could not be determined in our study.

We observed possible sex differences, with an association present for women but not men, despite higher power for men than women. Other studies of ESR have revealed potential sex differences in disease associations. Two ESR2 polymorphisms (rs1256031 and rs1256059) were associated with left ventricular mass and wall thickness in women but not men (12). In addition, variation in the ESR2 gene has been linked to Alzheimer disease in women but not men (31). Similarly, the relationship between ESR1 and myocardial infarction seems to be affected by differences related to sex.(1, 3) The underlying mechanisms responsible for sex-related differences in association are unclear but might include sex differences in ESR2 expression, ligand-binding affinity, or ESR2 splice variants.

This study has several strengths and limitations. It is one of the largest reported studies of ESR2 polymorphisms and CVD. Cases and controls were drawn from the same population, minimizing bias in the selection of controls, and endpoints were carefully documented. Of note, our analyses were restricted to white participants and thus are not necessarily applicable to other racial or ethnic groups. Although we observed an association between these polymorphisms and CVD in women, we have no direct evidence of a causative disease relationship for these particular polymorphisms. The rs1256049 polymorphism is a silent synonymous change and therefore may be in linkage disequilibrium with another relevant ESR2 mutation, such as rs127152, or the codon change may lead to changes in mRNA folding and subsequently to differences in mRNA translation or stability (32). The rs127152 polymorphism is located in the promoter region and therefore may have functional implications, although this theory has not been proven. Other SNPs in the ESR2 gene were not available for a more detailed examination. The study had limited power in women, and we could not perform more detailed subgroup analyses. Multiple comparisons were made, and it is possible that these results represent false positives. Nevertheless, the consistency of significance and association for rs1271572 and 1256049 among women warrants replication by others.

In summary, we found evidence for an association of 2 ESR2 polymorphisms and their associated haplotypes with increased risk of incident MI in women. These results suggest that additional evaluation is warranted to investigate the association between ESR2 genetic variation and CVD.

Table 3.

Adjusted conditional logistic regression odds ratios^* for CVD, MI, and ischemic stroke for men and women according to ESR2 polymorphism.

	WOMEN			MEN
	OR*	95%CI	P	OR*	95%CI	P
rs1271572**
CVD	1.49	1.10–2.01	0.01	0.98	0.81–1.18	0.81
MI	1.65	1.07–2.57	0.02	0.95	0.75–1.20	0.67
Ischemic stroke	1.34	0.87–2.07	0.18	1.00	0.72–1.38	0.99
rs1256049†
CVD	0.37	0.17–0.79	0.01	0.84	0.47–1.48	0.54
MI	0.25	0.09–0.73	0.01	0.65	0.32–1.32	0.23
Ischemic stroke	0.59	0.18–1.94	0.38	1.36	0.48–3.83	0.56
rs4986938
CVD	0.98	0.74–1.29	0.87	1.00	0.81–1.25	0.98
MI	1.21	0.78–1.87	0.40	1.07	0.78–1.45	0.69
Ischemic stroke	0.85	0.58–1.24	0.39	0.95	0.70–1.30	0.74

^*Conditional on age, smoking, and adjusted for randomized treatment group, BMI, history of hypertension, presence or absence of diabetes. OR is for genotype/additive mode of inheritance.

^**P values of 0.08 and 0.10 for CVD and MI in dominant mode. P values= 0.01 and 0.02 for CVD and MI in recessive mode.

^†P values of 0.02 and 0.02 for CVD and MI in dominant mode. Recessive mode could not be calculated due to rarity of variant homozygotes.

Abbreviations

CVD

Cardiovascular disease

ESR2

estrogen receptor-beta gene

ESR1

estrogen receptor-alpha gene

MI

myocardial infarction

ER-alpha

estrogen receptor alpha

ER-beta

estrogen receptor-beta

PHS

Physicians’ Health Study

WHS

Women’s Health Study