|Home | About | Journals | Submit | Contact Us | Français|
High levels of circulating heat shock protein 60 (Hsp60) and antibody to human Hsp60 have been associated with greater risk of coronary heart disease (CHD) in several studies, but associations between polymorphisms of the hsp60 gene and CHD risk have not been investigated.
By resequencing DNA from 30 unrelated Han Chinese and using HapMap Phase I Chinese data of hsp60 gene, we selected four tagging single nucleotide polymorphisms (tagSNPs) named rs2340690, rs788016, rs2305560, and rs2565163, and determined their frequencies in 1,003 Chinese CHD patients and 1,003 age- and sex-frequency-matched controls. Furthermore, we used PHASE 2.0 software to reconstruct haplotypes and logistic regression to control for potential confounders in multivariate analyses.
We found 13 SNPs in hsp60 gene (including four novel SNPs) in Han Chinese subjects. Our results showed no significant differences in four selected SNPs in patients with CHD and controls after adjusting for other conventional risk factors and stratifying by age, sex, smoking status, past history of hypertension and DM; however, our results showed that subjects with the GCTC haplotype had about twofold higher risk of CHD than those with the GTTC haplotype (OR=1.91, 95%CI: 1.26–2.89, P=0.002).
Our results suggest that the GCTC haplotype in the hsp60 gene is significantly associated with higher CHD risk in a Chinese population.
There is increasing evidence that autoimmune reactions play a pivotal role in the pathogenesis of coronary heart disease (CHD) (Hansson et al. 2002). The increased expression of heat shock protein 60 (Hsp60) in endothelial cells, macrophages, and smooth muscle cells (Kleindienst et al. 1993), may serve as an autoimmune antigen and accelerate the progression and development of atherosclerosis (Schett et al. 1995; Xu et al. 1999; Xu and Wick 1996). Human investigations have shown that high circulating Hsp60 levels (Pockley et al. 2000; Xu et al. 2000) and high levels of antibody to human Hsp60 have been associated with higher risk of CHD (Burian et al. 2001; Prohaszka et al. 2001; Veres et al. 2002; Xu et al. 1999; Zhu et al. 2001). Therefore, hsp60 may be an important candidate gene for CHD.
The hsp60 gene, which resides on human chromosome 2q33.1 and spans about 13.69 kb, includes 12 exons and 11 introns (Hansen et al. 2003). More than 80 single nucleotide polymorphisms (SNPs) in hsp60 are reported in the database of National Center for Biotechnology Information (NCBI) (http://www.ncbi.nlm.nih.gov/SNP). To the best of our knowledge, however, only two previous studies have examined the associations between the variability in hsp60 gene and human diseases. One French study indicated that the G+292A (dbSNP: rs41350646) polymorphism in hsp60 was associated with hereditary spastic paraplegia (Hansen et al. 2002), and another suggested that one MspI polymorphism was related to sudden infant death syndrome (Rahim et al. 1996). No study has investigated the association between hsp60 gene polymorphisms and CHD.
A haplotype is a combination of alleles of SNPs at multiple linked loci. With the completion of the International HapMap Project, it is possible to investigate disease-associated genetic polymorphisms through a whole-gene-based tagging SNP approach (Couzin 2006). In this case-control study, we resequenced the hsp60 gene in 30 unrelated Han Chinese and used HapMap Phase I Chinese data (http://www.hapmap.org) on the hsp60 gene and a linkage disequilibrium (LD)-based approach (Carlson et al. 2004). We selected four tagging polymorphisms of the hsp60 gene to examine their associations with CHD.
The study population was composed of 1,003 case patients and 1,003 age- and sex-frequency-matched controls. All enrolled subjects were unrelated ethnic Han Chinese. Case patients, who were from three hospitals (Tongji Hospital, Union Hospital, and Wugang Hospital) in Wuhan City, Hubei province were diagnosed as having CHD according to World Health Organization (WHO) criteria or by coronary angiography (significant coronary artery stenoses ≥50% in at least one major coronary artery). Myocardial infarction was diagnosed by a representative set of electrocardiograph (ECG), cardiac enzyme values, and typical symptoms. Angina was defined as use of nitroglycerine, experience of typical chest pain, or ECG changes compatible with ischemic heart disease. A total of 1,078 patients diagnosed as having CHD were recruited; 1,003 of them (93.0%) consented to participate in the study and provided questionnaire information and blood samples. After cases were diagnosed with CHD, they were interviewed in person within 3 days. The control subjects, residing in the same communities as the cases, were judged to be free of CHD and peripheral atherosclerotic arterial disease by medical history, clinical examinations, and electrocardiography. The response rate for the controls was 92.4% (1,003 of 1,085). Subjects with severe liver and/or kidney disease were excluded. Medical history, sociodemographic information, family history of cardiovascular disease, medication use, home environment, and lifestyle factors were obtained through questionnaire interview.
Subjects were identified as nonsmokers, former, or current smokers. Those who had smoked less than 100 cigarettes in the past lifetime were defined as nonsmokers; otherwise, they were defined as smokers. Those smokers who stopped smoking for more than 1 year were considered former smokers. Pack-years were calculated by multiplying the number of packs of cigarettes smoked per day by the number of years the person has smoked. Habitual physical activity was classified into four groups: little, light, moderate and vigorous. Body mass index (BMI) was calculated as weight in kilograms divided by the square of height in meters. Subjects were considered to be hypertensive if their systolic blood pressure was ≥140 mmHg and/or diastolic pressure ≥90 mmHg or they were already being treated with antihypertensive drugs. All subjects gave written consent after receiving a full explanation of the study. The Ethics Committee of Tongji Medical College approved this study.
Genomic DNA was amplified using Taq polymerase, followed by purification using the ethanol/NaAc method. Polymerase chain reaction (PCR) products were then used as a template for sequencing reactions with the BigDye Terminator kit v3.1 (Applied Biosystems, Foster City, CA, USA). Purified sequencing reactions were run on an ABI3100 genetic analyzer. Sequence analysis, SNP detection, and genotype calling were performed using DNAStar software. The entire 1–12 exon, the boundary region between exons and introns, and upstream promoter region of hsp60 gene were screened in 30 unrelated healthy Han Chinese. All primer sequences and reaction conditions are available on request from the corresponding author.
Using HapMap Phase I Chinese data (http://www.hapmap.org) acquired at the beginning of our work, according to the criteria of r2≥0.8 and minor allele frequency (MAF) ≥ 0.1, we extended 2000 bp in the 5′and 3′ end of hsp60 gene and selected 3 tagSNPs named rs2565163, rs2340690, and rs788016 to be analyzed (Table 1).
Furthermore, referring to our analysis of resequencing results for the hsp60 gene using htSNPer1.0 software, (Ding et al. 2005) we additionally selected another SNP, rs1116734, in the 5′ flanking region of the hsp60 gene, to analyze the association between hsp60 gene polymorphisms and CHD risk (Fig. 1). Because the allelic frequency of rs1116734 could not be detected by the Taqman SNP allelic discrimination method and PCR-restriction fragment length polymorphism method, we substituted another nearby SNP, rs2305560, for this analysis.
Fasting venous blood was collected in 5-ml heparin tubes, and genomic DNA was isolated with a Puregene kit (Gentra Systems, Inc., Minneapolis, MN, USA). Genotyping of hsp60 was performed with Taqman SNP allelic discrimination by means of an ABI 7900HT (Applied Biosystems, Foster City, CA), in 384-well format. PCR reactions were carried out in reaction volume of 5 μl containing 5 ng DNA, 2.5 μl 2× Taqman universal PCR Master MixNo AmpErase UNG (Applied Biosystems), 0.125 μl 40× Assay Mix. PCR conditions included 95°C for 10 min, followed by 40 cycles of 15 s at 92°C and 1 min at 60°C. Two blank controls (DNA hydration solution) and two replicate quality control samples were included in each 384-well format, and two replicate samples were genotyped with 100% concordance. The intensity of each SNP met the criteria of three clear clusters in two scales generated by SDS software (ABI).
Fasting blood glucose (FBG), total cholesterol (TC), and triglyceride (TG) were assayed using standard laboratory procedures in the department of clinical laboratory at the Union Hospital.
Continuous variables were reported as the mean value ± SD. Normal distribution of data was analyzed using the Kolmogorov–Smirnov normality test. Data with a normal distribution were compared by Student’s t test, and those with unequal variance or without a normal distribution were analyzed by a Mann–Whitney rank sum test. Categorical values were compared by the chi-square test, which was also used to test for deviation of genotype distribution from Hardy–Weinberg equilibrium. The htSNPer1.0 software was used to select tagSNPs in hsp60 gene (Ding et al. 2005) and PHASE 2.0 program (Stephens and Donnelly 2003) to infer haplotype frequencies based on the observed hsp60 genotypes. The most common haplotype among controls was used as reference in the logistic regression model. The associations between variants in the hsp60 gene and CHD risk were estimated by computing odds ratios (ORs) and 95% confidence intervals (CIs) from the multivariate logistic regression analyses. An unconditional logistic model was used to adjust for multiple cardiovascular risk factors. The probability level accepted for significance was P<0.05.
All data analyses were carried out with the statistical analysis software package SPSS12.0 (SPSS Inc., Chicago, IL, USA).
We found 13 genetic variants in the sequenced region of hsp60. Among them four SNPs had not been reported previously: 198189061 a/g (exon 2, het 0.06), 198184890a/g (exon 5, het 0.04), 198184774 ag/-(intron 5, het 0.5), and 198180643 ca /- (intron 7, het 0.5) in chromosome 2; another nine SNPs reported in NCBI and European Bioinformatics Institute (EBI) are rs1116734, rs3749095, rs1050347, rs8539, rs2340690, rs3214832, rs7585486, 198180659a/c (without rs ID), and rs788016 (Table 2). We found that rs1116734 and 198189061 a/g, and rs1116734 and rs788016 are not in LD, but SNPs such as rs1050347, rs2340690, rs3214832, rs7585486, and 198180659a/c are in perfect LD (r2=1). Other SNPs such as rs8539, 198184774 ag/-, and 198180643 ca/- are also in perfect LD (data not shown).
In the present study, cases were more likely than controls to be smokers, nondrinkers, and to have comorbidities such as hypertension and diabetes. Age and sex were similar for cases and controls, but cases were significantly more likely than controls to have a family history of CHD. FBG levels were significantly higher in cases than in controls (6.66±3.63 mmol/l vs. 5.32±2.00 mmol/l, P<0.01), however, TC levels were significantly lower in cases than in controls due to the use of cholesterol-lowering drugs by CHD patients (4.50±1.05 vs.5.08±1.30 mmol/l, P<0.01). There was no significant difference in TG levels and BMI between cases and controls (Table 3).
As shown in Table 4, the genotype frequency of hsp60 polymorphisms ranged from 7.6% to 58.8%. For each hsp60 polymorphism, there was no significant difference in genotype distribution between CHD and controls, and adjustment for the conventional risk factors such as age, sex, pack-years of smoking, drinking, activity, hypertension, DM, and family history of CHD did not appreciably alter the results (Table 4). Stratified analysis according to age (≤60 years and >60 years), sex, smoking status, past history of hypertension, and diabetes yielded the same results (data not shown).
The potential associations between the polymorphisms of hsp60 and anthropometric markers such as BMI, TC, systolic and diastolic blood pressure, and TG were also analyzed, but the results indicated that there were no significant differences in different genotype groups of four SNPs, and adjusting for conventional risk factors did not appreciably alter the results (data not shown).
All the pairwise LD measure D’ of the four investigated SNPs in hsp60 gene are above 0.87, whereas r2 ranged from 0.10 to 0.90 (data not shown). A total of 13 haplotypes were estimated in the CHD and control groups by using PHASE 2.0 software to reconstruct haplotypes based on the observed genotypes (Stephens and Donnelly 2003). Among the 13 haplotypes, five were >1.0%, namely GTTC, ACTC, GCCT, GCTC, and GCCC. The associations between the common haplotypes (covering 97.35% and 98.50% of allelic variance in CHD and controls, respectively) encompassing hsp60 polymorphisms and CHD were also examined. Compared with the highest-frequency haplotype of GTTC, the GCTC haplotype had 91% increased risk of CHD (OR=1.91, 95% CI: 1.26–2.89, P=0.002) (Table 5).
There is increasing evidence that elevated levels of Hsp60 and Hsp60 antibody are associated with CHD (Xu et al. 2000; Zhu et al. 2001). However, no previous study has examined the relationships between hsp60 genetic variations and the risk of CHD; specifically, none has investigated the association between the hsp60 gene and the risk of CHD using genetic markers to represent the overall variability of this gene. Therefore, the relationship between the variations in the hsp60 gene and CHD risk still remains unclear.
In this study, we complement our resequencing SNP panel with validated markers from HapMap to provide additional coverage of the genetic variation within the hsp60 gene. In addition, we investigated the associations between polymorphisms of hsp60 and CHD risk in a relatively large case-control study in a Chinese population. None of the polymorphisms showed a significant association with CHD. In addition, we did not find any significant associations between hsp60 gene variations and BMI, systolic and diastolic blood pressure, TC, and TG levels. However, using a haplotype-based approach, we found that haplotype distributions between the cases and the controls were statistically different for GCTC, which increased the risk of CHD by 91%.
It has been documented that information on haplotype blocks can increase the power by 15–50% compared with single SNP analysis (Gabriel et al. 2002; Thomas et al. 2003). It may be more reasonable to explain the relationship between genotype and phenotype, although the exact underlying molecular mechanism still remains unclear.
Hsp60 is an important chaperone protein and can be induced to protect cells from injury under stressful conditions. Most of the known risk factors for CHD, such as oxidized low-density lipoprotein, hypertension, infections, and oxidative stress, can increase the expression of Hsp60 in endothelial cells and smooth muscle cells. Extracellular Hsp60, which can induce adhesion molecules and cytokine production, may further promote local inflammation and possibly activate the innate immune system (Kol et al. 1999, 2000). On the other hand, there is 50–70% amino acid sequence homology between human and microbial Hsp60 (Wick et al. 2001). Serum antibodies against microbial Hsp60 may cross-react with human Hsp60 expressed in and released from the stressed cells of the blood vessels. These cross-reactions to human Hsp60 may play an important role in the process of CHD (Mayr et al. 1999). In addition, complement activation plays an important role in the development of CHD, and human Hsp60 may activate the complement system by complexing of antibodies to human Hsp60 (Prohaszka et al. 1999). These findings suggest that Hsp60 might play a multifunctional role in CHD development.
The present study has several strengths. This is the first study on the association between hsp60 gene polymorphisms and CHD risk in a Chinese population. The large number of cases and controls provides sufficient power for us to detect effects of the genotypes. In addition, for the first time, we used tagging single nucleotide polymorphisms (tagSNPs) to explore the association between polymorphisms of hsp60 and CHD risk. This made it possible to detect all the genetic variants across a specific region by using a subset of tagSNPs, which infer the allelic state of all of the common polymorphisms and cover maximum genetic variability. The use of tagSNPs in genetic association studies has been considered an efficient method to narrow association signal and localize susceptibility variants (The International HapMap Consortium 2003). Third, our populations are racially homogeneous, with all the participants being Han Chinese, thus minimizing the possibility of population stratification.
Three limitations should also be acknowledged. First, like all other case-control studies, there may have been selection bias (inclusion of patients surviving CHD). Meanwhile, some of our controls may have undiagnosed CHD, which might bias the results. Because we selected controls with normal electrocardiograms and without history of CHD, but without performing coronary angiography on them, we could not exclude the possibility that some of them were affected by CHD. Second, the haplotype of the hsp60 gene associated with CHD in this study may be in linkage disequilibrium with polymorphisms of other nearby genes that actually contribute to the development of CHD. Finally, although we included >1,000 cases of CHD and a comparable number of age- and sex-frequency matched controls in the final analysis, the present study was limited due to exclusion of low-frequency SNPs, which might be important in CHD development.
In conclusion, this study provides evidence that the haplotypes of the hsp60 gene are genetic susceptibility factors for CHD in Chinese, but the underlying mechanism remains to be elucidated in further studies.
This study was supported by research funds from the National Natural Science Foundation (30525031 and 30430590) and the National Key Basic Research and Development Program (2002CB512905) of China. Dr. Frank B. Hu’s work was supported in part by the American Heart Association Established Investigator Award.
We are particularly grateful to all individuals who participated in the present study and to the medical personnel of Tongji Hospital, Union Hospital, and Wugang Hospital in Wuhan city, Hubei Province, China.
The first two authors contributed equally to this paper.
Frank B. Hu, Phone: +1-617-4320113, Fax: +1-617-4322435, Email: Frank.hu/at/channing.harvard.edu.
Tangchun Wu, Phone: +86-27-83692347, Fax: +86-27-83692560, Email: wut/at/mails.tjmu.edu.cn.