Association study of polymorphisms in the ABO gene with ischemic stroke in the Chinese population

Background Ischemic stroke is the main cause of mortality and disability in older people worldwide. Recently epidemiological studies indicate that ischemic stroke is a complex disorder with a strong genetic component. Genome-wide association studies (GWAS) identified several single nucleotide polymorphisms (SNPs) associated with coronary artery disease (CAD) and myocardial infarction (MI) locus in ABO gene. Our study examined the association between four variants in the ABO gene and the risk of ischemic stroke and its subtypes, large-artery atherosclerosis (LAA) and small-vessel diseases (SVD) in the Chinese population. Methods In this case–control study, we recruited 1897 subjects, including 979 healthy controls and 918 ischemic stroke patients (465 with LAA and 453 with SVD). We selected four single nucleotide polymorphisms (rs579459, rs651007, rs514659 and rs529565) of the ABO gene and performed genotyping assays to assess the association with ischemic stroke and its subtypes. Results We found three polymorphisms, rs579459 and rs651007 were significantly associated with LAA using additive model and rs529565 was significantly associated with LAA using additive and dominant models. And we failed to find any significant association between these SNPs and ischemic stroke and SVD in the Chinese population. However, after the Bonferroni correction for multiple comparisons, the P-values of these SNPs failed to exceed significant threshold under any models. Conclusion Our findings indicated that genetic variations of ABO gene may contribute to susceptibility of LAA but not ischemic stroke and SVD in the Chinese population. Our preliminary results should be further validated in prospective independent studies with expanded sample size.


Background
Stroke is one of the main causes of death and adult disability around the world [1,2]. There are two major categories of stroke: ischemic stroke and hemorrhagic stroke, and ischemic stroke constitutes over 80 % of total stroke in origin [3]. Increasing evidence indicated that ischemic stroke is a complex clinical syndrome resulting from environmental and genetic factors [4,5]. According to the modified Trial of Org10172 in Acute Stroke Treatment (TOAST) classification,ischemic stroke itself can be divided into five subtypes, LAA and SVD are two common etiologic subtypes of ischemic stroke [6,7]. Conventional risk factors such as hypertension, diabetes mellitus, dyslipidemia and smoking could not completely explain all ischemic stroke risk, family and twin-based studies demonstrated that genetic factors also play a key role in the development of ischemic stroke [8,9].
The ABO gene, located around 9q34.2, encodes glycosyltransferases, which catalyze the transfer to different carbohydrate groups onto the H antigen, thus forming A and B antigens of the ABO system [10,11]. In recent years, several studies have found that genetic variants of ABO gene were associated with several diseases. Previous studies have identified rs579459 in ABO as genetic variants associated with the risk of CAD in European descent [12]. Recent study also identified the ABO rs579459 polymorphism as genetic variants significantly associated with venous thromboembolism (VTE) [13]. However, fewer studies focus on the association between ABO gene and ischemic stroke [14][15][16]. Ischemic stroke and CAD have several common risk factors and genetic susceptibility, especially for LAA and CAD. Taking all these considerations together, the ABO gene may be a promising candidate gene of ischemic stroke. Recent GWAS showed that three SNPs (rs505922, rs643434, and rs651007) of ABO gene were associated with ischemic stroke and its subtypes in the European population [15]. However, there were no independent replication studies regarding the association between ABO gene and ischemic stroke in Chinese population.
Against this background, in the present study we aimed to investigate the association between four SNPs (rs579459, rs651007, rs514659 and rs529565) of ABO gene and ischemic stroke susceptibility in the Chinese population. University of Traditional Chinese Medicine as a leader of the organization jointed three other hospitals to collect the object of study. All case and control subjects were unrelated to one another and were recruited from the Chinese population. Clinical diagnoses of ischemic stroke were confirmed through computed tomography (CT) and/or magnetic resonance imaging (MRI) scans of the brain. The brain images were independently assessed by two welltrained technologist and physician. The common subtypes (LAA and SVD) of ischemic stroke were determined by the modified TOAST classification system [6]. Control subjects were recruited from the health management of the Second People's Hospital Affiliated to Fujian University of Traditional Chinese Medicine. Controls with stroke and other neurological diseases and cardiovascular diseases were excluded in this study. The questionnaire was designed to collect demographic characteristics, clinical vascular variables, and medical histories for both cases and controls. Conventional vascular risk factors including hypertension, diabetes mellitus, and dyslipidemia were evaluated through WHO/ISH criteria. This study was approved by the institutional review boards of all participating hospitals. Written informed consent and peripheral blood samples were obtained from patients and controls before they attended our study.

SNP selection and genotyping
We selected four SNPs of the ABO gene, including rs579459, rs651007, rs643434, and rs505922. For two SNPs, rs643434 and rs505922 failed assay primer design, alternative SNPs in complete Linkage disequilibrium (LD) were chosen. By HapMap, rs643434 and rs505922 were replaced respectively with rs514659 and rs529565 as in strong LD (r 2 = 1) in the Chinese population. The SNPs were genotyped using Sequenom MassARRAY platform (San Diego, U.S) at CapitalBio Corporation (Beijing, China) and the genotyping analysis was undertaken according to the manufacturer's protocol.
Genomic DNA was isolated from human peripheral blood samples of each individual through Wizard® Genomic DNA Purification Kit (Promega, Madison, WI, USA). DNA concentration was determined by DNA spectrophotometer (ND-1000, NanoDrop, Wilmington USA). Specific assays including a locus-specific PCR reaction based on a locus-specific primer extension reaction were designed using the MassARRAY Assay Design software package (v3.1). Mass determination was carried out with the MALDI-TOF mass spectrometer and Mass ARRAY Type 4.0 software was used for data acquisition.

Data analysis
Association analysis was performed with PLINK software using additive, dominant, recessive and genotype models [17]. Hardy-Weinberg equilibrium (HWE) were performed for each SNP. Logistic regression was used for risk stratification with or without covariate adjustments determined by significant differences between ischemic stroke patients and controls, such as age, gender, hypertension, and diabetes mellitus.

Clinical characteristics of total ischemic stroke patients and controls
The demographic characteristics and clinical vascular variables for the 918 ischemic stroke patients (51 % LAA and 49 % SVD) and 979 healthy control subjects used in this study are shown in Table 1. Distribution of age and gender between cases and controls were significantly different. Other risk factors such as smoking, drinking, and diabetes mellitus were found to be more prevalent in cases compared to controls. Hypertension was significantly lower in ischemic stroke cases than in controls. There were no significant differences in levels of triglyceride between the ischemic stroke cases and the controls, but total cholesterol, low density lipoprotein, and high density lipoprotein levels were significantly lower in ischemic stroke cases compared to controls.

Comparison of allele and genotype frequencies in ischemic stroke
The association between the four SNPs in ABO gene and the risk of ischemic stroke was analyzed using additive, dominant, genotype, and recessive models. However, the polymorphism rs514659 was not found to be in HWE in control subjects, therefore excluded from further statistical analyses. The observed allele and genotype frequencies for ischemic stroke cases and controls are shown in Table 2, rs579459 was significantly associated with the risk of ischemic stroke using dominant model (p = 0.04, OR = 1.22, 95 % CI = 1.00-1.48). The allele frequencies of the other two SNPs (rs651007 and rs529565) showed no difference between the ischemic stroke and the control group (p > 0.5). However, the association between rs579459 and ischemic stroke failed to remain significance after logistic regression analysis adjusting for age, gender, hypertension, diabetes mellitus, dyslipidemia, smoking and drinking status.

Comparison of allele and genotype frequencies in LAA
As shown in Table 3, we also observed the association between the three SNPs of ABO gene and LAA occurrence under additive, dominant, genotype, and recessive models. Three polymorphisms, rs579459, rs651007, and rs529565 were significantly associated with LAA in both additive and dominant models. After adjusting for age, gender, hypertension, diabetes mellitus, and dyslipidemia by logistic regression analysis, two polymorphisms, rs579459 and rs651007 were remained significantly association with LAA using additive model and rs529565 was significantly associated with LAA using additive and dominant models. However, all P-values failed to reach Data were shown as mean ± standard deviation (SD) or as n (%). Significant differences between cases and controls were indicated with an asterisk (*) significance after the Bonferroni adjustment for multiple comparisons.

Comparison of allele and genotype frequencies in SVD
To explore whether the ABO polymorphisms are confined to a specific subtype, we also evaluated the association between the three SNPs of ABO gene and the risk of SVD. As shown in Table 4, no significant associations were observed for the allele and genotype frequencies between the cases and controls in all three SNPs.

Discussion
The ABO gene is located near the end of the long arm of chromosome 9 and encodes glycosyltransferases, which add sugar residues to the H-antigen producing A or B antigens of the ABO system. Previous studies found that the non-O phenotypes were more frequent in ischemic stroke patients than controls and was associated with an increased risk of MI and CAD [18][19][20]. In consistent with these results, previously study showed that compared with the O phenotype, non-O phenotypes associated with an increased risk of stroke [21]. In contrast, other study did not detect significant association between ABO blood group and ischemic stroke and any of the four main etiologic subtypes of ischemic stroke [14]. Ischemic stroke is a complex disease with different pathophysiology and risk factors. It is important to investigate risk factors in different etiologic subtypes. Some subtypes of ischemic stroke and CAD shared many common risk factors, for example, atherosclerosis plaque were observed in both LAA and CAD as a common pathophysiologic mechanism. Consequently, it was speculated that genetic variants of ABO gene associated with CAD [12], may be also associated with LAA. Previous studies have found that polymorphisms of ABO gene were associated with many diseases [12,13,15]. As reported previously that rs579459 in ABO gene was associated with CAD in Caucasians [12]. Fewer studies have investigated the genetic association between ABO polymorphisms and ischemic stroke. Early studies  reported that ABO gene variants (rs651007, rs643434, and rs505922) are associated with LAA and cardioembolic stroke in the European population [15]. However, there was no study reported the association between ABO gene and ischemic stroke in the Chinese population. Hence, in our case-control study, we investigated the association of four polymorphisms in ABO gene with the risk of ischemic stroke and its main subtypes. Our results supported previous observations that ischemic stroke in particular LAA and CAD share several risk factors. We found that rs579459 in ABO gene was associated with LAA. In line with previous study [15], our current study found significant associations between ABO SNPs (rs651007 and rs529565) and LAA, and these SNPs were failed to be associated with ischemic stroke and SVD, although the association disappeared after the Bonferroni adjustment, which was known to be one of the most stringent methods for multiple comparisons. The observed minimal differences between the results from our results and previous study may partly due to differences in genetic background and our study have a small sample size.
Our study had a number of limitations. First, we selected four SNPs of the ABO gene which showed association with CAD or ischemic stroke in the European population. These SNPs merely represented limited genetic variability of ABO gene. The future studies will be required to confirm the association between ABO gene and ischemic stroke and its subtypes by high density genotyping on SNPs of ABO gene. Second, limited size of the cohort might reduce the power to detect association. Thus, prospective independent studies with a comparatively larger sample size are required for validation in the Chinese population.

Conclusions
In the present study we aimed to investigate the association between ABO gene and ischemic stroke and its main subtypes in the Chinese population. We found that three SNPs (rs651007, rs579459 and rs529565) of ABO gene were significantly associated with LAA in the Chinese population, though not survived Bonferroni correction for multiple comparisons. Therefore, prospective studies with a comparatively large sample size are required to confirm the association between ABO gene and ischemic stroke in the Chinese population and to characterize the functional role of ABO underlying ischemic stroke or LAA.