Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

HDAC9 Variant Rs2107595 Modifies Susceptibility to Coronary Artery Disease and the Severity of Coronary Atherosclerosis in a Chinese Han Population

  • Xue-bin Wang ,

    Contributed equally to this work with: Xue-bin Wang, Ya-di Han

    Affiliation Center for Gene Diagnosis, Zhongnan Hospital of Wuhan University, Wuhan, Hubei, China

  • Ya-di Han ,

    Contributed equally to this work with: Xue-bin Wang, Ya-di Han

    Affiliation Center for Gene Diagnosis, Zhongnan Hospital of Wuhan University, Wuhan, Hubei, China

  • Shrestha Sabina,

    Affiliation Center for Gene Diagnosis, Zhongnan Hospital of Wuhan University, Wuhan, Hubei, China

  • Ning-hua Cui,

    Affiliation Department of Clinical Laboratory, Children's Hospital of Zhengzhou, Zhengzhou, Henan, China

  • Shuai Zhang,

    Affiliation Center for Gene Diagnosis, Zhongnan Hospital of Wuhan University, Wuhan, Hubei, China

  • Ze-jin Liu,

    Affiliation Center of Clinical Laboratory, Wuhan Asia Heart Hospital, Wuhan, Hubei, China

  • Cong Li,

    Affiliation Zhongshan School of Medicine, Sun Yat-sen University, Guangzhou, China

  • Fang Zheng

    zhengfang@whu.edu.cn

    Affiliation Center for Gene Diagnosis, Zhongnan Hospital of Wuhan University, Wuhan, Hubei, China

HDAC9 Variant Rs2107595 Modifies Susceptibility to Coronary Artery Disease and the Severity of Coronary Atherosclerosis in a Chinese Han Population

  • Xue-bin Wang, 
  • Ya-di Han, 
  • Shrestha Sabina, 
  • Ning-hua Cui, 
  • Shuai Zhang, 
  • Ze-jin Liu, 
  • Cong Li, 
  • Fang Zheng
PLOS
x

Abstract

A previous genome-wide association study showed that a single nucleotide polymorphism (SNP) rs2107595 in histone deacetylase 9 (HDAC9) gene was associated with large artery stroke (LAS) in Caucasians. Based on the similar atherosclerotic pathogenesis between LAS and coronary artery disease (CAD), we aimed to evaluate the associations of SNP rs2107595 with CAD risk and the severity of coronary atherosclerosis in a Chinese Han population, and explore the potential gene-environment interactions among SNP rs2107595 and conventional CAD risk factors. In a two-stage case-control study with a total of 2317 CAD patients and 2404 controls, the AG + AA genotypes of SNP rs2107595 were significantly associated with increased CAD risk (Adjusted odds ratio (OR) = 1.23, Padj = 0.001) and higher modified Gensini scores (Adjusted OR = 1.38, Padj < 0.001). These associations remained significant in subtype analyses for unstable angina pectoris (UAP), non-ST-segment elevation myocardial infarction (NSTEMI) and ST-segment elevation myocardial infarction (STEMI). Subgroup and multifactor dimensionality reduction analyses (MDR) further found the gene-environment interactions among SNP rs2107595, body mass index, type 2 diabetes and hyperlipidemia in CAD risk and the severity of coronary atherosclerosis. Moreover, patients with CAD had higher levels of HDAC9 mRNA expression and plasma HDAC9 than controls. Subsequent genotype-phenotype analyses observed the significant correlations of SNP rs2107595 with HDAC9 mRNA expression and plasma HDAC9 levels in controls and patients with NSTEMI and STEMI. Taken together, our data suggest that SNP rs2107595 may contribute to coronary atherosclerosis and CAD risk through a possible mechanism of regulating HDAC9 expression and gene-environment interactions.

Introduction

Coronary artery disease (CAD), the leading cause of death and disability worldwide[1], is mainly caused by multiple interactions between genetic and environmental risk factors of atherosclerosis[2]. Atherosclerosis in the coronary arteries, is a continuous process of deposition of lipoproteins that results in the formation of atherosclerotic plaques, the progression of plaque instability, and eventually myocardial infarction (MI)[3, 4]. Therefore, the genetic and environmental determinants involved in CAD risk may vary across different stages of atherosclerosis, leading to the underlying heterogeneity between stable and unstable CAD[4, 5]. Moreover, this atherosclerotic pathogenesis also applies to the cerebral vessels and contributes to the development of ischemic stroke (IS), especially to the development of large artery stroke (LAS)[6]. Numerous studies have shown that CAD and LAS shared several conventional risk factors and susceptible loci[7, 8], such as smoking, hypertension and single nucleotide polymorphisms (SNPs) on chromosome 9p21, suggesting shared heritability for both diseases[6].

Recently, a genome-wide association study (GWAS) first reported that SNP rs2107595 in 3'-region of histone deacetylase 9 (HDAC9) gene was associated with LAS risk[9]. HDAC9 gene is located on chromosome 7p21.1, and encodes a member of histone decetylases, which is mainly responsible for the deacetylation of histones and subsequent gene transcription[10]. Accumulated evidence highlights the crucial roles of SNP rs2107595 and HDAC9 gene in the development of atherosclerosis. First, in a population-based cohort study, the minor allele of SNP rs2107595 was correlated with higher carotid intima-media thickness (CIMT)[11], which has been reported as a predictive marker for atherosclerosis[12]. Second, HDAC9 plays central roles in heart development[13], and is highly expressed in atherosclerotic plaques at human systemic arteries[11]. Finally, HDAC9 deletion could lead to down-regulation of inflammatory genes[14], up-regulation of lipid-associated genes[14], and dramatic reduction of atherosclerotic lesion size in mice[15]. Taking all these findings together, we hypothesize that SNP rs2107595 may also contribute to the development of coronary atherosclerosis and CAD risk by modifying HDAC9 expression.

Thus, in this two-stage case-control study, we assessed the associations of SNP rs2107595 with CAD risk and the severity of coronary atherosclerosis, and further performed a genotype-phenotype analysis by detecting HDAC9 mRNA expression and plasma HDAC9 levels, followed by subgroup and multifactor dimensionality reduction (MDR) analyses to explore the potential gene-environment interactions. This is the first study to assess the associations of SNP rs2107595 with CAD risk and the severity of coronary atherosclerosis in the Chinese population.

Materials and Methods

Study population

This two-stage case-control study included a total of 2317 CAD patients and 2404 age- and sex-matched controls. For reducing the impact of population stratification, we only recruited unrelated ethnic Han Chinese in Hubei Province, central China. In the discovery set (Study 1), 1172 patients and 1086 controls were enrolled from Wuhan Asia Heart Hospital between March 2011 and June 2016. The replication set (Study 2) with 1145 cases and 1318 controls was recruited from Zhongnan Hospital of Wuhan University between July 2013 and June 2016. In these two sets, CAD was angiographically defined as stenosis of more than 50% in at least one major coronary artery or their main branches. Then, according to the American College of Cardiology/American Heart Association Task Force on Practice Guidelines[1618], CAD patients were further categorized based on the following clinical presentation: (1) stable angina pectoris (SAP); (2) unstable angina pectoris (UAP); (3) non-ST-segment elevation MI (NSTEMI); (4) ST-segment elevation MI (STEMI)(S1 Appendix). Finally, for each case, modified Gensini scores[19] were calculated by two independent angiographers to assess the severity of coronary atherosclerosis. The control groups included participants without lumen stenosis validated by coronary angiography and healthy subjects without cardiovascular diseases identified by physical examinations. The following data for each participant were collected: (1) clinical information such as blood pressure, body mass index (BMI), fasting plasma glucose (FPG) and lipid levels; (2) conventional CAD risk factors including smoking status, alcohol drinking status, and histories of type 2 diabetes mellitus (T2DM), hyperlipidemia, and hypertension (S1 Appendix). The following subjects were excluded: (1) individuals with systemic diseases such as cerebrovascular diseases, cancers, autoimmune diseases and renal or hepatic diseases; (2) subjects with cardiac diseases including myocardial bridge, coronary artery spasm and congenital or valvular heart diseases. This study followed the Declaration of Helsinki, and was approved by the Ethics Committees of Wuhan Asia Heart Hospital and Zhongnan Hospital of Wuhan University. Written informed consent was obtained from all participants.

Criterion of modified Gensini scores

In the modified Gensini scoring system[19], angiographic stenosis of each coronary segment was first scored according to the degree of luminal narrowing: 1 for 0–25% stenosis, 2 for 26–50%, 4 for 51–75%, 8 for 76–90%, 16 for 91–99% and 32 for 100%. Then a multiplier was assigned to each segment depending on the physical importance of the area supplied by that segment: 5 for main left coronary artery (MLCA), 2.5 for proximal left anterior descending coronary artery (LAD) and proximal left circumflex branch, 1.5 for mid-segment of LAD, 0.5 for second diagonal branch and posterolateral branch, and 1 for other branches. Finally, the weighted scores for each segment were added to give modified Gensini scores. As acute coronary occlusion mainly occurs in a previous angiographically non-critical lesion, acute total occlusion was scored as a non-critical lesion (0–5 score) in the modified Gensini scoring system [19].

SNP rs2107595 genotyping

Genomic DNA of peripheral blood leukocytes was isolated by a phenol/chloroform method. SNP rs2107595 was genotyped by high resolution melting (HRM) analyses using a LightScanner 96 system (Idaho Technology, Salt Lake City, UT, USA)[20, 21], as described in S1 Table. Repeated assays and DNA sequence analysis were used to validate the accuracy of HRM genotyping (S1 Fig).

Reverse-transcription quantitative PCR analysis of HDAC9 mRNA

Total RNA was extracted from peripheral blood leukocytes using the Trizol reagent (Invitrogen, Carlsbad, CA, USA). After eliminating DNA contamination using the RNase-Free gDNA eraser, 1μg of total RNA was prepared for reverse transcription using a reverse transcriptase kit (Takara Bio Inc, Kusatsu, Shiga, Japan).Then, to determine HDAC9 mRNA expression, the cDNA products were used for reverse-transcription quantitative PCR (RT-qPCR) analysis with the SYBR-Green kit on a CFX96 Touch system(Bio-rad, Hercules, CA, USA). RT-qPCR analysis for each sample was performed in triplicate and followed the MIQE guidelines [22]. The relative expression of HDAC9 was normalized to the expression of reference gene (GAPDH) using the 2 -△△Cq method[23]. Primer sequences and RT-qPCR conditions for HDAC9 and GAPDH were also summarized in S1 Table.

Assessment of plasma HDAC9 activity

Plasma samples were separated by centrifugation (at 2000 g for 10 minutes at 4°C) and stored at -80°C until detection. According to the manufacturer' s instructions, plasma HDAC9 levels were measured by an enzyme-linked immunosorbent assay (ELISA) (HDAC9 ELISA kit, Xinfan Biosystems, Shanghai, China),followed by quantification using a standard cure with the detection limit of 0.1 ng/mL.CV values for intra- and inter-assays were 4.7% and 6.5%, respectively.

Statistical analyses

For clinical characteristics, the differences in categorical and continuous variables between cases and controls were analyzed by the Pearson 2 test and the Student's t-test, respectively. For SNP rs2107595, Hardy-Weinberg equilibrium (HWE) was assessed by the Pearson 2 test. Allelic and genotypic association analyses were performed by logistic regression analyses with and without adjustment for age, sex, BMI, smoking status, alcohol drinking status and histories of T2DM, hyperlipidemia and hypertension. The homogeneity between the odds ratios (ORs) of two sets was evaluated by the Breslow-Day test. The association of SNP rs2107595 with modified Gensini scores was analyzed by the Mann-Whitney U test and logistic regression analyses. When samples were stratified, the multiplicative likelihood ratio test was carried out to test the possible gene-environment interactions in CAD risk and the severity of coronary atherosclerosis. The differences in HDAC9 expression and plasma HDAC9 levels between cases and controls, and the associations of SNP rs2107595 with HDAC9 expression and plasma HDAC9 levels were assessed by analyses of covariates (ANCOVA) after adjusting for covariates. The correlations of HDAC9 expression and plasma HDAC9 levels with modified Gensini scores were examined by the Spearman correlation test. All these statistical analyses were performed by SPSS 17.0 (SPSS Inc., Chicago, IL, USA) and P values of less than 0.05 (two-sided) were considered as statistically significant.

In association analyses for SNP rs2107595, we performed the Bonferroni correction and Monte-Carlo permutation tests to control for multiple comparisons. The Monte-Carlo permutation test can formally calculated an emprical P value by repeating permutations 100,000 times to randomly redistribute genotype counts of cases and controls [24]. An emprical P value of less than 0.05 was considered as a stable result for multiple comparisons. A statistical power was calculated by PS 3.0 program (Vanderbilt University, Nashville, TN, USA). To further assess the high-order gene-environment interactions in CAD risk and the severity of coronary atherosclerosis, multifactor dimensionality reduction (MDR) analyses were carried out using MDR 2.0 beta 8.4 program (UPenn, Philadelphia, PA, USA) [25]. In brief, this program first constructed all possible combinations of included variables. Then, by using 100-time cross-validation and 1000-time permutation tests, the best n-factor models for predicting CAD risk and the severity of coronary atherosclerosis were found with the maximal cross-validation consistency (CVC) and the optimal testing accuracy. Finally, for these best n-factor models, the interaction entropy graphs were applied to visually depict the univariate effect of each variable and the pairwise interactions. The univariate and pairwise effects were expressed as the percentage of entropy.

Results

Characteristics of study population

As summarized in S2 Table, in two cohorts of our study, there were significant differences in blood pressure, BMI, FPG and lipid levels as well as the frequencies of smoking, alcohol drinking, T2DM, hyperlipidemia, and hypertension between cases and controls. The genotype distributions of SNP rs2107595 were in accordance with HWE in two control groups (in discovery and replication sets, PHWE = 0.734 and 0.904, respectively).

Allelic and genotypic association analyses between SNP rs2107595 and CAD risk

In the discovery set, allelic association analyses showed that the minor allele A of SNP rs2107595 was significantly associated with increased CAD risk (OR = 1.19, P = 0.008).This significant association was further identified in the replication set with an allelic OR of 1.20 and a P value of 0.002. As the Breslow-Day test confirmed the homogeneity of ORs between the two sets (P = 0.864), a meta-analysis of these two cohorts was carried out. And the results indicated that the minor allele of SNP rs2107595 had a 1.19-fold (P = 6.08 × 10−5) increased risk of CAD in the merged set. All these associations remained significant after the permutation test for multiple testing and adjustment for covariates (Table 1). Given a minor allele frequency (MAF) of 0.3055 in controls, a MAF of 0.3442 in cases and the type I error of 0.05, the merged sample size could provide a statistical power of 81.0% to detect the association.

thumbnail
Table 1. Allelic and genotypic associations of SNP rs2107595 with CAD risk.

https://doi.org/10.1371/journal.pone.0160449.t001

To further explore the inheritance pattern of SNP rs2107595, genotypic association analyses were performed based on additive, dominant and recessive models (Table 1). In both discovery and replication sets, significant associations were identified between SNP rs2107595 and increased CAD risk under both additive and dominant models (Table 1). When we combined these two cohorts, the dominant model was considered as the most significant model with the largest OR of 1.28 and the smallest P value of 3.04 × 10−5. After adjusting for covariates, all these associations remained significant (Table 1). In both allelic and genotypic association analyses, all significant results in the merged set were robust enough to withstand the Bonferroni correction (Table 1).

Subtype and subgroup analyses for associations between SNP rs2107595 and CAD risk

To further evaluate the effects of SNP rs2107595 on different CAD subtypes and the potential gene-environment interactions, subtype and subgroup analyses were performed under a dominant model (Table 2). In subtype analyses, compared with the GG carriers, subjects with the AG + AA genotypes had a 1.32-fold (P = 7.75 × 10−4), a 1.53-fold (P = 1.20 × 10−5) and a 1.62-fold (P = 9.61× 10−5) increased risk of UAP, NSTEMI and STEMI, respectively,. In subgroup analyses, the association between the AG + AA genotypes and increased CAD risk was consistently significant in almost all subgroups, except for individuals with BMI ≤ 25 and participants without T2DM and hyperlipidemia. Furthermore, we observed the multiplicative interactions of SNP rs2107595 with BMI (Pinter = 0.009), T2DM (Pinter = 0.004) and hyperlipidemia (Pinter = 0.001) status in CAD risk. All the above associations were still significant after adjusting for covariates (Table 2). Based on the Bonferroni correction, we also found significant associations in subtype analyses for UAP, NSTEMI and STEMI as well as in subjects with BMI > 25, T2DM and hyperlipidemia (Table 2).

thumbnail
Table 2. Subtype and subgroup analyses for the association between SNP rs2107595 and CAD risk.

https://doi.org/10.1371/journal.pone.0160449.t002

Association of SNP rs2107595 with the severity of coronary atherosclerosis

By using a modified Gensini scoring system, we assessed the severity of coronary atherosclerosis in CAD patients and found that patients with the AG + AA genotypes had higher modified Gensini scores than the GG carriers in discovery (P = 0.028), replication (P = 0.001) and merged sets (P = 6.47 × 10−5)(Table 3). When we classified CAD patients into two groups according to the median (30) of modified

thumbnail
Table 3. Association of SNP rs2107595 with the severity of coronary atherosclerosis in 2317 CAD patients under a dominant model.

https://doi.org/10.1371/journal.pone.0160449.t003

Gensini scores, we also observed that the variant genotypes (AG + AA) of SNP rs2107595 were associated with a 1.38-fold (Padj = 1.68 × 10−4) increased risk of higher modified Gensini scores (> 30) in the merged set after adjusting for covariates. These associations remained significant in almost all subtypes and subgroups, except for SAP subtypes as well as subgroups of BMI ≤ 25, non-T2DM and non-hyperlipidemia (Table 3). Moreover, the AG + AA genotypes multiplicatively interacted with higher BMI (> 25)(Pinter = 0.043), and histories of T2DM(Pinter = 0.013) and hyperlipidemia(Pinter = 1.51× 10−4) to increase the severity of coronary atherosclerosis. After the Bonferroni correction, the associations in the merged set as well as in patients with BMI > 25, T2DM and hyperlipidemia remained significant (Table 3).

MDR analyses for assessing the high-order gene-environment interactions in CAD risk and the severity of coronary atherosclerosis

Based on the results of multiplicative interaction analyses, data on SNP rs2107595, BMI, T2DM and hyperlipidemia status were included in the MDR analyses to further explore the high-order gene-environment interactions. As presented in Table 4, the four-factor model including all variables was selected as the best predictor for both CAD risk and higher modified Gensini scores, because it had the optimal testing accuracy and the maximal CVC values. For this four-factor model, when evaluating the univariate effects of different variables, history of T2DM (0.58%)exerted the greatest independent effect on CAD risk (Fig 1A), while history of Hyperlipidemia (1.56%) was the strongest risk factor for higher modified Gensini scores (Fig 1B). When assessing the strength of pairwise effects, the strongest pairwise interactions were found between SNP rs2107595 and T2DM (3.66%) in MDR analyses for predicting CAD risk (Fig 1A), and between SNP rs2107595 and hyperlipidemia (2.89%) in MDR analyses for predicting the severity of coronary atherosclerosis (Fig 1B).

thumbnail
Table 4. The gene-environment interactions between SNP rs2107595 and conventional CAD risk factors in CAD risk and the severity of coronary atherosclerosis by MDR analyses.

https://doi.org/10.1371/journal.pone.0160449.t004

thumbnail
Fig 1. Interaction entropy graphs for the gene-environment interactions between SNP rs2107595, BMI status, and histories of T2DM and hyperlipidemia in CAD risk and the severity of coronary atherosclerosis.

(A)Interaction entropy graphs for the gene-environment interactions between SNP rs2107595, BMI status, and histories of T2DM and hyperlipidemia in CAD risk; (B)Interaction entropy graphs for the gene-environment interactions between SNP rs2107595, BMI status, and histories of T2DM and hyperlipidemia in the severity of coronary atherosclerosis. In these graphs, the univariate effect of each variable and the pairwise interactions are expressed as the percentage of entropy. The univariate effect of each variable is presented in a box with the percentage of entropy. The pairwise interactions are shown as connected lines accompanied by the percentage of entropy. Different colors on lines indicate the strength of the interactions, i.e. red, orange and yellow lines mean strong, moderate and weak interactions, respectively.

https://doi.org/10.1371/journal.pone.0160449.g001

The differences in HDAC9 mRNA expression and plasma HDAC9 levels between CAD patients and controls

We measured HDAC9 mRNA expression and plasma HDAC9 levels in 250 representative subjects (125 cases vs 125 controls, S3 Table) randomly selected from the replication set. In ANCOVA models adjusted for covariates, HDAC9 mRNA expression (1.20 ± 0.41 vs 1.06± 0.39, P = 0.031) and plasma HDAC9 levels (54.2± 9.9 vs 50.6± 8.7, P = 0.015) in CAD patients were significantly higher than those in controls. In subtype analyses, patients with NSTEMI and STEMI had higher levels of HDAC9 mRNA expression and plasma HDAC9 than controls (Table 5). Moreover, we also observed a positive correlation between plasma HDAC9 levels and modified Gensini scores (r = 0.184, P = 0.040).

thumbnail
Table 5. Associations of SNP rs2107595 with HDAC9 mRNA expression and plasma HDAC9 levels.

https://doi.org/10.1371/journal.pone.0160449.t005

Associations of SNP rs2107595 with HDAC9 mRNA expression and plasma HDAC9 levels

Assuming a dominant model, significant correlations were found between the AG + AA genotypes and increased HDAC9 mRNA expression in both cases (P = 0.011) and controls (P = 0.008), and between the AG + AA genotypes and higher plasma HDAC9 levels only in controls(P = 0.009) (Table 5 and Fig 2). When patients were stratified by CAD subtypes, we further observed the significant associations of the AG + AA genotypes with increased levels of HDAC9 mRNA expression and plasma HDAC9 in patients with NSTEMI and STEMI (Fig 2).

thumbnail
Fig 2. Associations of SNP rs2107595 with HDAC9 mRNA expression and plasma HDAC9 levels.

(A) Associations of SNP rs2107595 with HDAC9 mRNA expression in controls and patients with NSTEMI and STEMI; (B) Associations of SNP rs2107595 with plasma HDAC9 levels in controls and patients with NSTEMI and STEMI. ANCOVA models are used to assess statistical significance. Data are expressed as mean ± standard deviation (SD).

https://doi.org/10.1371/journal.pone.0160449.g002

Discussion

In 2012, the METASTROKE collaboration reported a significant association between SNP rs2107595 of HDAC 9 gene and LAS risk in Caucasians[9], suggesting an important role of this locus in atherosclerotic stroke. Subsequent GWAS datasets in Europeans identified that SNPs rs11984041 and rs2023938 in this gene were associated with the risk of LAS[26, 27] and CAD[6, 28], respectively. However, based on the data from a previous report[29] and 1000 Genome Database[30] (http://www.1000genomes.org/), these two SNPs in Caucasians are not polymorphic in the Chinese Han population, suggesting that the genetic architecture in HDAC9 region may differ across ethnicities. More importantly, no previous studies have assessed the effects of SNP rs2107595 on coronary atherosclerosis and subsequent CAD risk. Therefore, in the current study, by using a two-stage case-control design with a total of 2317 CAD patients and 2404 controls, we genotyped SNP rs2107595 and found that the minor allele A of this locus was significantly associated with increased CAD risk and higher modified Gensini scores in a Chinese Han population. Then, by using subgroup and MDR analyses, the gene-environment interactions among the AG + AA genotypes, higher BMI and histories of T2DM and hyperlipidemia were identified to increase CAD risk and the severity of coronary atherosclerosis. Finally, the genotype-phenotype analyses further suggested that SNP rs2107595 might be functional by regulating HDAC9 mRNA expression and plasma proteins.

The present study, for the first time, reported the significant associations of SNP rs2107595 with CAD risk and the severity of coronary atherosclerosis. These findings are further supported by the current result that the minor allele A of SNP rs2107595 was correlated with higher HDAC9 mRNA expression and the previous reports that HDAC9 expression was involved in the development of atherosclerosis[11, 14]. By predicting the potential functions of SNP rs2107595 in the F-SNP database [31] (http://compbio.cs.queensu.ca/F-SNP/), we found that this variant is located in an evolutionarily conserved region, suggesting the importance of its physical location. Subsequent bioinformatics using the MatInspector [32] (http://www.genomatix.de/index.html) and SiteGA databases [33] (http://wwwmgs.bionet.nsc.ru/cgi-bin/mgs/sitega/index.pl) further showed that the minor allele of this locus disrupted the binding sites of several E2F transcription factors, including repressor E2F-4 and -6, which have been reported to repress gene transcription by forming a heterodimeric complex with RB protein[34]. Although the exact mechanism needs to be further elucidated, the above evidence reinforces the possibility that SNP rs2107595 may regulate HDAC9 expression and protein levels, then contribute to the development of coronary atherosclerosis and CAD risk.

In a GWAS for Caucasians, SNP rs2107595 was first associated with LAS, which was considered as an atherosclerotic subtype of IS[35]. In the current study, we also found stronger associations of this variant with unstable CAD (UAP, NSTEMI and STEMI) than with SAP. In general, the continuum from SAP to unstable CAD mainly results from the progression of plaque vulnerability caused by the aggravated inflammation in plaques[36]. Notably, this inflammatory process can be further induced by the polarized switching of macrophages from an anti-inflammatory M2 phenotype to a proinflammatory M1 state[3739]. Recently, Cao et al.[14] have reported that HDAC9 deficiency could decrease the expression of M1 inflammatory genes and promote M2 polarization. More importantly, in the present study, besides the significant association of SNP rs2107595 with HDAC9 expression, we also found that HDAC9 mRNA expression in patients with NSTEMI and STEMI was significantly higher than that of controls. Taking all the evidence together, it is reasonable to hypothesize that SNP rs2107595 and HDAC9 expression may involve the progression of atherosclerosis, and thus exert stronger effects on unstable CAD than on SAP.

The above evidence also pinpoints the importance of subtype analyses when assessing the associations of susceptible loci with chronic, progressive diseases like CAD and IS. Recently, a case-control study by Su et al.[40] failed to find a significant association of SNP rs2107595 with IS in a southern Chinese population. Besides the relatively small sample size of their study (816 cases and 816 controls, a power of 37% and 60% based on the allelic ORs of previous GWAS (OR = 1.13) and the present study (OR = 1.19), respectively) and the genetic heterogeneity between central (a MAF of 0.306 in our controls) and southern Chinese (a MAF of 0.345 in their controls), another important reason for their non-significant results may be that Su et al.'s study did not define IS subtypes for their cases, and therefore could not detect the association of SNP rs2107595 with specific IS subtypes.

In this study, by using subgroup and MDR analyses, we found the gene-environment interactions among SNP rs2107595, T2DM, hyperlipidemia and BMI in CAD risk and the severity of coronary atherosclerosis. Subsequent interaction entropy graphs further observed the greatest pairwise effect of SNP rs2107595 and T2DM on CAD risk, and the strongest pairwise interaction between SNP rs2107595 and hyperlipidemia in higher modified Gensini scores. For T2DM and SNP rs2107595, recent studies have shown that HDAC9 ablation mice exhibited improved glucose tolerance and insulin tolerance[41], and that HDAC9 up-regulation enhanced gluconeogenesis in vitro[42]. The important roles of HDAC9 gene in glucose metabolism, combined with the known impact of dysglycemia on CAD risk [43] as well as the significant correlation of SNP rs2107595 with HDAC9 expression, support that the interaction between SNP rs2107595 and T2DM may greatly increase CAD risk. In clinical studies, hyperlipidemia has long been associated with the severity of coronary atherosclerosis[44, 45]. For HDAC9 gene, functional studies in mice also suggested that increased HDAC9 expression could inhibit cholesterol efflux through reduction of histone acetylation at promoters of cholesterol efflux genes[14]. As the first stage of reverse cholesterol transport[46], cholesterol efflux plays crucial roles in lipid metabolism, and is inversely correlated with dyslipidemia[47] and the severity of coronary atherosclerosis[48]. Therefore, by increasing HDAC9 expression, the minor allele of SNP rs2107595 may inhibit cholesterol efflux, then interact with hyperlipidemia to aggravate coronary atherosclerosis.

In conclusion, by regulating HDAC9 expression and interacting with T2DM, hyperlipidemia and BMI status, the minor allele A of SNP rs2107595 increased CAD risk and the severity of coronary atherosclerosis. Functional studies are needed to explain the underlying mechanism.

Supporting Information

S1 Appendix. Supplementary materials and methods (Diagnostic criterion of different CAD subtypes and Definition of clinical characteristics).

https://doi.org/10.1371/journal.pone.0160449.s001

(DOCX)

S1 Dataset. Clinical and genetic data in our population.

https://doi.org/10.1371/journal.pone.0160449.s002

(XLSX)

S2 Dataset. Clinical and genetic data of participants that were randomly selected to measure HDAC9 mRNA expression and plasma HDAC9 levels.

https://doi.org/10.1371/journal.pone.0160449.s003

(XLSX)

S1 Fig. HRM and sequencing analyses for different genotypes of SNP rs2107595.

(A)HRM plots for different genotypes of SNP rs2107595. The normalized melting peaks are given in the left column, and the normalized melting curves are given in the right column. Arrows indicate the genotypes. Heterozygous samples are identified by a change in melting curve shape, and different homozygotes are distinguished by melting temperature (Tm) shifts;(B) Direct sequencing analyses for different genotypes of SNP rs2107595.

https://doi.org/10.1371/journal.pone.0160449.s004

(TIFF)

S1 Table. Primer details and PCR conditions for HRM, direct sequencing and RT-qPCR analyses in our study.

https://doi.org/10.1371/journal.pone.0160449.s005

(DOCX)

S2 Table. Clinical characteristics of participants in our study.

https://doi.org/10.1371/journal.pone.0160449.s006

(DOCX)

S3 Table. Comparative analyses for clinical and genetic characteristics between the randomly selected subjects and the whole samples.

https://doi.org/10.1371/journal.pone.0160449.s007

(DOCX)

Author Contributions

  1. Conceptualization: XBW FZ.
  2. Data curation: XBW CL.
  3. Formal analysis: XBW.
  4. Funding acquisition: FZ CL.
  5. Investigation: XBW YH SZ.
  6. Methodology: XBW YH.
  7. Project administration: FZ.
  8. Resources: ZL FZ.
  9. Software: XBW.
  10. Supervision: FZ.
  11. Validation: SS NC SZ.
  12. Visualization: XBW.
  13. Writing - original draft: XBW FZ.
  14. Writing - review & editing: XBW FZ.

References

  1. 1. Mozaffarian D, Benjamin EJ, Go AS, Arnett DK, Blaha MJ, Cushman M, et al. Heart disease and stroke statistics—2015 update: a report from the American Heart Association. Circulation. 2015;131(4):e29–322. pmid:25520374.
  2. 2. Lanktree MB, Hegele RA. Gene-gene and gene-environment interactions: new insights into the prevention, detection and management of coronary artery disease. Genome Med. 2009;1(2):28. pmid:19341499.
  3. 3. Willer CJ, Sanna S, Jackson AU, Scuteri A, Bonnycastle LL, Clarke R, et al. Newly identified loci that influence lipid concentrations and risk of coronary artery disease. Nat Genet. 2008;40(2):161–9. pmid:18193043.
  4. 4. Andersson C, Vasan RS. Compiling the complement of genes implicated in coronary artery disease. Circ Cardiovasc Genet. 2014;7(6):738–40. pmid:25516622.
  5. 5. Iannaccone M, Quadri G, Taha S, D'Ascenzo F, Montefusco A, Omede P, et al. Prevalence and predictors of culprit plaque rupture at OCT in patients with coronary artery disease: a meta-analysis. Eur Heart J Cardiovasc Imaging. 2015. pmid:26508517.
  6. 6. Dichgans M, Malik R, Konig IR, Rosand J, Clarke R, Gretarsdottir S, et al. Shared genetic susceptibility to ischemic stroke and coronary artery disease: a genome-wide analysis of common variants. Stroke. 2014;45(1):24–36. pmid:24262325.
  7. 7. Cheng YC, Anderson CD, Bione S, Keene K, Maguire JM, Nalls M, et al. Are myocardial infarction—associated single-nucleotide polymorphisms associated with ischemic stroke? Stroke. 2012;43(4):980–6. pmid:22363065.
  8. 8. Kessler T, Erdmann J, Dichgans M, Schunkert H. Shared genetic aetiology of coronary artery disease and atherosclerotic stroke—2015. Curr Atheroscler Rep. 2015;17(4):498. pmid:25690589.
  9. 9. Traylor M, Farrall M, Holliday EG, Sudlow C, Hopewell JC, Cheng YC, et al. Genetic risk factors for ischaemic stroke and its subtypes (the METASTROKE collaboration): a meta-analysis of genome-wide association studies. Lancet Neurol. 2012;11(11):951–62. pmid:23041239.
  10. 10. Haberland M, Montgomery RL, Olson EN. The many roles of histone deacetylases in development and physiology: implications for disease and therapy. Nat Rev Genet. 2009;10(1):32–42. pmid:19065135.
  11. 11. Markus HS, Makela KM, Bevan S, Raitoharju E, Oksala N, Bis JC, et al. Evidence HDAC9 genetic variant associated with ischemic stroke increases risk via promoting carotid atherosclerosis. Stroke. 2013;44(5):1220–5. pmid:23449258.
  12. 12. Lorenz MW, Markus HS, Bots ML, Rosvall M, Sitzer M. Prediction of clinical cardiovascular events with carotid intima-media thickness: a systematic review and meta-analysis. Circulation. 2007;115(4):459–67. pmid:17242284.
  13. 13. Chang S, McKinsey TA, Zhang CL, Richardson JA, Hill JA, Olson EN. Histone deacetylases 5 and 9 govern responsiveness of the heart to a subset of stress signals and play redundant roles in heart development. Mol Cell Biol. 2004;24(19):8467–76. pmid:15367668.
  14. 14. Cao Q, Rong S, Repa JJ, St Clair R, Parks JS, Mishra N. Histone deacetylase 9 represses cholesterol efflux and alternatively activated macrophages in atherosclerosis development. Arterioscler Thromb Vasc Biol. 2014;34(9):1871–9. pmid:25035344.
  15. 15. Azghandi S, Prell C, van der Laan SW, Schneider M, Malik R, Berer K, et al. Deficiency of the stroke relevant HDAC9 gene attenuates atherosclerosis in accord with allele-specific effects at 7p21.1. Stroke. 2015;46(1):197–202. pmid:25388417.
  16. 16. Fraker TD Jr, Fihn SD, Gibbons RJ, Abrams J, Chatterjee K, Daley J, et al. 2007 chronic angina focused update of the ACC/AHA 2002 guidelines for the management of patients with chronic stable angina: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines Writing Group to develop the focused update of the 2002 guidelines for the management of patients with chronic stable angina. J Am Coll Cardiol. 2007;50(23):2264–74. pmid:18061078.
  17. 17. Anderson JL, Adams CD, Antman EM, Bridges CR, Califf RM, Casey DE Jr, et al. 2012 ACCF/AHA focused update incorporated into the ACCF/AHA 2007 guidelines for the management of patients with unstable angina/non-ST-elevation myocardial infarction: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. Circulation. 2013;127(23):e663–828. pmid:23630129.
  18. 18. O'Gara PT, Kushner FG, Ascheim DD, Casey DE Jr, Chung MK, de Lemos JA, et al. 2013 ACCF/AHA guideline for the management of ST-elevation myocardial infarction: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol. 2013;61(4):e78–140. pmid:23256914.
  19. 19. Montorsi P, Ravagnani PM, Galli S, Rotatori F, Veglia F, Briganti A, et al. Association between erectile dysfunction and coronary artery disease. Role of coronary clinical presentation and extent of coronary vessels involvement: the COBRA trial. Eur Heart J. 2006;27(22):2632–9. pmid:16854949.
  20. 20. Wang XB, Qiao C, Wei L, Han YD, Cui NH, Huang ZL, et al. Associations of Polymorphisms in MTHFR Gene with the Risk of Age-Related Cataract in Chinese Han Population: A Genotype-Phenotype Analysis. PLoS One. 2015;10(12):e0145581. pmid:26689687.
  21. 21. Wang XB, Han YD, Cui NH, Gao JJ, Yang J, Huang ZL, et al. Associations of lipid levels susceptibility loci with coronary artery disease in Chinese population. Lipids Health Dis. 2015;14:80. pmid:26209006.
  22. 22. Bustin SA, Benes V, Garson JA, Hellemans J, Huggett J, Kubista M, et al. The MIQE guidelines: minimum information for publication of quantitative real-time PCR experiments. Clin Chem. 2009;55(4):611–22. pmid:19246619.
  23. 23. Schmittgen TD, Livak KJ. Analyzing real-time PCR data by the comparative C(T) method. Nat Protoc. 2008;3(6):1101–8. pmid:18546601.
  24. 24. Kinnamon DD, Martin ER. Valid Monte Carlo permutation tests for genetic case-control studies with missing genotypes. Genet Epidemiol. 2014;38(4):325–44. pmid:24723341.
  25. 25. Hahn LW, Ritchie MD, Moore JH. Multifactor dimensionality reduction software for detecting gene-gene and gene-environment interactions. Bioinformatics. 2003;19(3):376–82. pmid:12584123.
  26. 26. Bellenguez C, Bevan S, Gschwendtner A, Spencer CC, Burgess AI, Pirinen M, et al. Genome-wide association study identifies a variant in HDAC9 associated with large vessel ischemic stroke. Nat Genet. 2012;44(3):328–33. pmid:22306652.
  27. 27. (ISGC). NSGNSISGC. Loci associated with ischaemic stroke and its subtypes (SiGN): a genome-wide association study. Lancet Neurol. 2015. pmid:26708676.
  28. 28. Deloukas P, Kanoni S, Willenborg C, Farrall M, Assimes TL, Thompson JR, et al. Large-scale association analysis identifies new risk loci for coronary artery disease. Nat Genet. 2013;45(1):25–33. pmid:23202125.
  29. 29. Qingxu G, Yan Z, Jiannan X, Yunlong L. Association Between the Gene Polymorphisms of HDAC9 and the Risk of Atherosclerosis and Ischemic Stroke. Pathol Oncol Res. 2016;22(1):103–7. pmid:26347468.
  30. 30. Abecasis GR, Altshuler D, Auton A, Brooks LD, Durbin RM, Gibbs RA, et al. A map of human genome variation from population-scale sequencing. Nature. 2010;467(7319):1061–73. pmid:20981092.
  31. 31. Lee PH, Shatkay H. F-SNP: computationally predicted functional SNPs for disease association studies. Nucleic Acids Res. 2008;36(Database issue):D820–4. pmid:17986460.
  32. 32. Cartharius K, Frech K, Grote K, Klocke B, Haltmeier M, Klingenhoff A, et al. MatInspector and beyond: promoter analysis based on transcription factor binding sites. Bioinformatics. 2005;21(13):2933–42. pmid:15860560.
  33. 33. Levitsky VG, Kulakovskiy IV, Ershov NI, Oshchepkov DY, Makeev VJ, Hodgman TC, et al. Application of experimentally verified transcription factor binding sites models for computational analysis of ChIP-Seq data. BMC Genomics. 2014;15:80. pmid:24472686.
  34. 34. Dimova DK, Dyson NJ. The E2F transcriptional network: old acquaintances with new faces. Oncogene. 2005;24(17):2810–26. pmid:15838517.
  35. 35. Bang OY, Saver JL, Liebeskind DS, Pineda S, Ovbiagele B. Association of serum lipid indices with large artery atherosclerotic stroke. Neurology. 2008;70(11):841–7. pmid:18160673.
  36. 36. Hansson GK. Inflammation, atherosclerosis, and coronary artery disease. N Engl J Med. 2005;352(16):1685–95. pmid:15843671.
  37. 37. Stoger JL, Gijbels MJ, van der Velden S, Manca M, van der Loos CM, Biessen EA, et al. Distribution of macrophage polarization markers in human atherosclerosis. Atherosclerosis. 2012;225(2):461–8. pmid:23078881.
  38. 38. Lee CW, Hwang I, Park CS, Lee H, Park DW, Kang SJ, et al. Macrophage heterogeneity of culprit coronary plaques in patients with acute myocardial infarction or stable angina. Am J Clin Pathol. 2013;139(3):317–22. pmid:23429367.
  39. 39. Shaked I, Hanna DB, Gleissner C, Marsh B, Plants J, Tracy D, et al. Macrophage inflammatory markers are associated with subclinical carotid artery disease in women with human immunodeficiency virus or hepatitis C virus infection. Arterioscler Thromb Vasc Biol. 2014;34(5):1085–92. pmid:24651679.
  40. 40. Su L, Shen T, Liang B, Xie J, Tan J, Chen Q, et al. Association of GWAS-supported loci rs2107595 in HDAC9 gene with ischemic stroke in southern Han Chinese. Gene. 2015;570(2):282–7. pmid:26093197.
  41. 41. Chatterjee TK, Basford JE, Knoll E, Tong WS, Blanco V, Blomkalns AL, et al. HDAC9 knockout mice are protected from adipose tissue dysfunction and systemic metabolic disease during high-fat feeding. Diabetes. 2014;63(1):176–87. pmid:24101673.
  42. 42. Chen J, Wang N, Dong M, Guo M, Zhao Y, Zhuo Z, et al. The Metabolic Regulator Histone Deacetylase 9 Contributes to Glucose Homeostasis Abnormality Induced by Hepatitis C Virus Infection. Diabetes. 2015;64(12):4088–98. pmid:26420860.
  43. 43. Ross S, Gerstein HC, Eikelboom J, Anand SS, Yusuf S, Pare G. Mendelian randomization analysis supports the causal role of dysglycaemia and diabetes in the risk of coronary artery disease. Eur Heart J. 2015;36(23):1454–62. pmid:25825043.
  44. 44. Xu RX, Li S, Li XL, Zhang Y, Guo YL, Zhu CG, et al. High-density lipoprotein subfractions in relation with the severity of coronary artery disease: A Gensini score assessment. J Clin Lipidol. 2015;9(1):26–34. pmid:25670357.
  45. 45. Mitsutake R, Miura S, Zhang B, Saku K. HDL-associated factors provide additional prognostic information for coronary artery disease as determined by multi-detector row computed tomography. Int J Cardiol. 2010;143(1):72–8. pmid:19237208.
  46. 46. Saleheen D, Scott R, Javad S, Zhao W, Rodrigues A, Picataggi A, et al. Association of HDL cholesterol efflux capacity with incident coronary heart disease events: a prospective case-control study. Lancet Diabetes Endocrinol. 2015;3(7):507–13. pmid:26025389.
  47. 47. Medina-Urrutia A, Juarez-Rojas JG, Cardoso-Saldana G, Jorge-Galarza E, Posadas-Sanchez R, Martinez-Alvarado R, et al. Abnormal high-density lipoproteins in overweight adolescents with atherogenic dyslipidemia. Pediatrics. 2011;127(6):e1521–7. pmid:21555497.
  48. 48. Khera AV, Cuchel M, de la Llera-Moya M, Rodrigues A, Burke MF, Jafri K, et al. Cholesterol efflux capacity, high-density lipoprotein function, and atherosclerosis. N Engl J Med. 2011;364(2):127–35. pmid:21226578.