Association of eNOS and Cav-1 gene polymorphisms with susceptibility risk of large artery atherosclerotic stroke

Endothelial nitric oxide synthase (eNOS) is localized in caveole and has important effects on caveolar coordination through its interaction with caveolin-1 (Cav-1), which supports normal functioning of vascular endothelial cells. However, the relationship between genotypic polymorphisms of e-NOS and Cav-1 genes and ischemic stroke (IS) remains lesser reported. This hospital-based case-control study aimed to determine the genetic polymorphisms of the eNOS (Glu298Asp) and Cav-1 (G14713A and T29107A) genes in association with susceptibility risk in patients who had suffered from a large artery atherosclerotic (LAA) stroke. Genotyping determination for these variant alleles was performed using the TaqMan assay. The distributions of observed allelic and genotypic frequencies for the polymorphisms were in Hardy-Weinberg equilibrium in healthy controls. The risk for an LAA stroke in the Asp298 variant was 1.72 (95% CI = 1.09–2.75) versus Glu298 of the eNOS. In the GA/AA (rs3807987) variant, it was 1.79 (95% CI = 1.16–2.74) versus GG and in TA/AA (rs7804372) was 1.61 (95% CI = 1.06–2.43) versus TT of the Cav-1, respectively. A tendency toward an increased LAA stroke risk was significant in carriers with the eNOS Glu298Asp variant in conjunction with the G14713 A and T29107A polymorphisms of the Cav-1 (aOR = 2.03, P-trend = 0.002). A synergistic effect between eNOS and Cav-1 polymorphisms on IS risk elevation was significantly influenced by alcohol drinking, heavy cigarette smoking (P-trend<0.01), and hypercholesterolemia (P-trend < 0.001). In conclusion, genotypic polymorphisms of the eNOS Glu298Asp and Cav-1 14713A/29107A polymorphisms are associated with the elevated risk of LAA stroke among Han Chinese in Taiwan.


Introduction
Despite the growing optimism due to recent advances in stroke therapy, strokes remain a major leading cause of disability and the third-leading cause of death among ethnic Chinese in Taiwan. An ischemic stroke (IS) is the most prevalent type of stroke, accounting for more than 80% of all stroke cases [1]. IS pathogenesis is believed to be multi-factorial, influenced by smoking, alcohol intake, hypertension, obesity, diabetes mellitus, and environmental factors as well as interactions with genetic components [2][3][4]. Recent progress in the understanding of the mechanism underlying the pathogenesis of cardiomyopathy has led to the discovery of therapeutic targets of nitric oxide (NO) with hopeful optimism [5]. Hence, gaining insight into the identification of genetic factors regarding NO has become an important issue to predict risk susceptibility and improve prevention strategies to reduce the human and economic burden of strokes.
Vascular endothelial NO synthesis is produced by the action of endothelial NO synthase (eNOS), whose encoded gene is located on chromosome 7q35. Endothelium-derived NO plays several predominant roles in arteries, including inhibition of the adhesion of platelets and leukocytes, vasodilation in the endothelium, and the reduction of the genesis of oxidized low-density lipoprotein to prevent atherogenesis [6][7][8]. In recent, eNOS has been reported to occur specifically in the subdomain within caveolin-1, encoded by the Cav-1 gene, which assembles as a coat and scaffolding protein in caveolae and is responsible for the multiple phenotypes in vascular endothelial cells. Membrane caveolae consists of caveolins, a cholesterol-binding family with 21-24 kDa integral membrane proteins [9][10][11]. Binding affinity of eNOS/Cav-1 has multifunctional signaling in NO release which may account for the modified effect of Cav-1 in association with eNOS activity on vasodilation [12].
An individual vulnerability of eNOS and Cav-1 polymorphisms to the pathogenesis among cardiovascular diseases has been studied [13][14][15][16][17], yet the correlation between genotypic polymorphisms of these two genes and the synergistic interaction of eNOS/Cav-1 on susceptible risk of large artery atherosclerotic (LAA) stroke has never been reported. Thus, this hospitalbased case-control study aimed to determine the comparative distribution of eNOS (G894T) and Cav-1 (G14713A and T29107A) polymorphisms in patients who had undergone an LAA stroke and control subjects free of neurological diseases. In addition, the modified effects of the traditional risk factors, such as hypercholesterolemia, alcohol drinking, and heavy cigarette smoking on the genetic predisposition to have a stroke were evaluated.

Study population
The sample size that was required in this hospital-based case-control study was computed by using "Genetic Power Calculator" (http://pngu.mgh.harvard.edu/) under the assumption of 80% statistical power, 5% variant allele frequency, 1:1 case/control ratio, and 5% error rate in an allelic test. The present study consisted of 229 consecutive patients who had undergone an LAA stroke and 243 age-and gender-matched healthy controls. IS was defined according to the World Health Organization definition as "rapidly developing clinical signs of focal (or global) disturbance of cerebral function lasting more than 24 hours with no apparent cause other than of vascular origin." Any patients with atypical symptoms, including transient ischemic attacks, intracranial hemorrhage, post seizure palsy, brain trauma, brain tumors and younger stroke (<50 years old) were excluded from the study. The age at onset is defined as the age at which the first symptom of IS become evident by patient recollection. IS patients underwent at least one brain imaging (CT and/or MRI) study. Based on the clinical features and the information from brain imaging, echocardiography, ultrasonography of extracranial and intracranial arteries, and angiography (MRA or conventional angiography), these IS patients were sub-classified according to the TOAST classification system [18,19]. The healthy control group was recruited from the same hospital in the health examination clinic during the same study period, and each of these controls was chosen on the basis that they have no evidence of any cerebrovascular disorders. Informed consent from each individual was obtained prior to specimen acquisition and the study design was approved by the Ethics Committee of Institutional Review Board of the Tri-Service General Hospital (TY101-11), Taipei, Taiwan.

Data collection
The information collected included demographic characteristics (ethnic background, residence area, and educational level) and age at diagnosis. The definitions of hypertension and diabetes mellitus (DM) followed guidelines of Hypertension Management Guide for Doctors [20] and guided by Diagnosis and classification of diabetes mellitus of American Diabetes Association [21], respectively. Patients with hypertension or DM were assured by the presence of an actual diagnosis as well as received antihypertensive/antidiabetic medication. History of alcoholism (defined as alcohol abuse disorder according to the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV) [22]), a current habit of daily cigarette smoking (heavy smokers consumed ≧10 cigarettes daily for at least 10 years), and the categorization of hypercholesterolemia as demonstrated in our previous report [23].

DNA extraction and genotype analysis
A sample of 10-ml of peripheral blood, collected in acetate-citrate dextrose, was obtained from a fasting venous blood sample drawn from each patient and control subject. Genomic DNA was extracted from peripheral blood lymphocytes using a QIAamp DNA blood mini kit (QIA-GEN Inc., Chatsworth, CA, USA) following the manufacturer's protocol and stored at -20˚C for subsequent genotype determination. These variant genotypes of interest were chosen for genotyping based on the following criteria: (i) the single nucleotide polymorphism (SNP) resulted in an amino acid substitution; (ii) the frequency of the variant allele was greater than five percent in the general population. Polymorphisms in the Cav-1 G14713A (rs3807987) and T29107A (rs7804372), and e-NOS Glu298Asp (rs1799983) genes were selected based on the information in the Chinese population arranged in the HapMap database (http://www. hapmap.org/). DNA sequences of probes and primers for the genotyping assays were chosen using the TaqMan Assays-by-Design™ supplied by Applied Biosystems (ABI, Foster City, CA, USA). Genotype analyses for these variant alleles were performed using the TaqMan assay and PCR amplification conditions on ABI Prism 7000 Sequence Detection System (Applied Biosystems, Foster City, CA). For quality assurance, independent positive and negative control samples were included and analyzed on each 96-well plate in each batch. To ensure that the observed polymorphisms were specific and not the result of experimental variations, all uncertain genotypic results were subjected to a repeating assay for confirmation.

Statistical analyses
The means and standard deviations or frequency and percent were calculated for continuous and categorical variables, respectively. Univariate and multivariate analyses were used to determine socioeconomic statuses and risk factors for IS. Hardy-Weinberg equilibrium (HWE) was assessed using a chi-square test. The allelic and genotypic frequencies of each polymorphism for these four genes were compared between the cases and the control subjects using chi-square or Fisher's exact test as appropriate. The adjusted odds ratios (aORs) and the corresponding 95% confidence intervals (95% CIs) for the association between polymorphisms in the Cav-1 and eNOS genes and LAA IS risk were estimated considering age and gender in multivariate logistic regression analysis. All statistical tests were performed with SPSS version 19.0 based on a 2-tailed probability and P-values less than 0.05 were considered statistically significant.

Results
Demographic characteristics of cases and control subjects are summarized in Table 1. There was no significant difference between the two groups in their age (stroke patients and controls were 73.2 ± 9.9 and 72.8 ± 7.9 years old, respectively) and gender. An increased LAA stroke risk was associated with risk factors, including hypertension (aOR = 3.17; 95% CI = 2.15-4.65), diabetes (aOR = 2.80; 95% CI = 1.83-4.28), hypercholesterolemia (aOR = 1.45; 95% CI = 1.01-2.08) using multiple logistic regression analysis after adjusting for age and sex. Besides, a significant risk elevation of a stroke was observed for patients with a cigarette smoking habit (aOR = 2.52; 95% CI = 1.62-3.94) ( Table 1).
The allelic and genotypic frequencies of e-NOS (G894T) and Cav-1 (G14713A and T29107A) polymorphisms are shown in Table 2. The frequencies of observed alleles and genotypes for these polymorphisms were in Hardy-Weinberg proportions in the controls. On the basis of inferred phenotypic manifestations [10,[24][25][26], subjects who carried rare alleles of the  (Table 2). We evaluated interactions between the variant alleles and atherosclerotic factors by determining the genotype frequency of the polymorphisms among the LAA stroke patients and the controls. As summarized in Table 3, a greater proportion of the eNOS 894GT and TT variants was seen in cases, showing a 3.23-fold risk increase in patients with hypercholesterolemia (95% CI = 1.70-6.12) and a 3.50-fold risk in patients with a long-term habit of heavy cigarette smoking (95% CI = 1.75-6.99). Besides, carriers with the Cav-1 G14713A polymorphism showed an elevated IS risk among those subjects within the same categorization of the clinical parameters, including hypercholesterolemia (aOR = 2.46, 95% CI = 1.42-4.27) and heavy cigarette smoking (aOR = 3.46, 95% CI = 1.75-6.84), Similarly, observations of the effects of hypercholesterolemia and heavy cigarette smoking on LAA stroke risk elevation were statistically significant in patients with the Cav-1T29107A variant genotype (TA/AA) (P < 0.05) ( Table 3). The eNOS/Cav-1 combination has been suggested to play a key function in endothelial hyperpermeability, and our hypothetic mechanism underlying ischemic contracture is attributed to Cav-1 acting on vascular hyperpermeability through eNOS inhibition to reduce NO levels, leading to vasoconstriction and ischemic damages of the vascular system. Support for this hypothesis came from the observation that multiple interactions of genes on LAA stroke risk elevation were present in patients carrying a combination of the susceptible genotypes. By using a dummy variable coding scheme [27], an increased IS risk was shown in individuals who carried eNOS Glu894Asp variants together with the Cav-1 allelic variants of G14713A and T29107A (aOR = 4.20, 95% CI = 1.66-10.63). A significant trend toward LAA stroke risk elevation was found in patients carrying one additional risk genotype measured by the β estimation (P trend = 0.002) ( Table 4).
Based on the hypothetic mechanism that these high-risk genotypes were associated with an increasing risk of stroke disease by mimicking ischemic contracture on reduced vascular hyperpermeability to inhibit NO generation, the modified effects of hypercholesterolemia, alcohol drinking, and heavy cigarette smoking on LAA stroke risk would be more significant in carriers of high-risk genotypes of eNOS and Cav-1. We thus examined the possible interaction between the eNOS Glu894Asp and Cav-1 G14713A and T29107A polymorphisms in conjunction with the clinical features using the joint and stratified methods. As a result, among eNOS Glu894Asp carriers, individuals having variant alleles of the Cav-1 G14713A and T29107A variants were associated with an increasing LAA stroke risk when compared to those , and heavy cigarette smoking (aOR = 8.61, 95% CI = 1.74-42.50); the P-value for the trend test in eNOS 894Asp carriers with one additional Cav-1 polymorphism was significant, as measured by the β estimates from the regression model (Table 5).

Discussion
In this hospital-based case-control study, we carried out a comparative analysis of the frequency distribution of the 894T allele in the eNOS gene in our control subjects with the reported frequency and confirmed a rare prevalence of the variant T allele, similar to the documented frequencies among Asian populations, including ethnic Taiwanese [28][29][30]. Our observation that a positive association between the eNOS Glu298Asp variant and LAA stroke is consistent with the findings that the eNOS Glu298Asp polymorphism is in association with  [31][32][33][34][35]. These results provide justification of the mutated 894T allele playing an independent risk factor for carotid atherosclerosis [25]. Further, because NO molecule derived from eNOS conversion in the vascular endothelium has the vasoprotective effects through elimination of superoxide radicals, reduction of platelet aggregation, anti-inflammation, and limiting the oxidation of atherogenic low-density lipoproteins [36]; thus, dysfunction of the endothelial NOS activity has an impact of oxidative stress on vascular endothelium in healthy smokers [37]. Therefore, our findings that individuals carried eNOS Glu298Asp genotype with hypercholesterolemia and with habits of drinking alcohol and heavy smoking cigarettes are interpreted by the presence of an individual impairment in the vasoprotective capability causally linked to the eNOS polymorphism. The caveolar coordination of eNOS, more specifically its interaction with Cav-1, plays a key role in normal eNOS activity to generate vascular NO, which is a bioavailable controller of shear-dependent NO release [38]. Knockdown experiments of Cav-1 have been shown to cause cardiohypertrophy, diabetes, and pulmonary and focal cerebral ischemia and reperfusion injury [39][40][41]. Recent genome wide association reports have shown for that Cav-1 SNPs are associated with stroke disease and that the Cav-1 G14713A elevates risk of coronary artery disease and myocardial infarction [24,[42][43][44]. However, some results are conflicting regarding the association between the Cav-1 polymorphisms and the risk of atrial fibrillation in cardiac arrhythmias [45]. In this study, we screened two SNPs, the G14713A and T29107A polymorphisms of the Cav-1 gene and examined their genotypic association with LAA stroke susceptibility (Table 2). Remarkably, based on a combinatorial study with genetic-environmental consideration, our findings provide realistic data regarding the risk susceptibility of an LAA stroke modified by gene-to-gene interactions; the two-locus risk genotype 14713A/T29107A at the Cav-1 showed the highest odds ratio in combination with the eNOS Glu298Asp genotype (Table 4). It is likely that polymorphisms in the Cav-1 gene increase high plasma triglyceride levels, resulting in endothelial function abnormality and ischemia [46]. It was reported that caveolae and caveolin-1 reverse cholesterol transport in atherosclerosis [47], and that the integration of eNOS and caveolin-1 principally participates in cholesterol trafficking, lipid homeostasis, and transmembrane signaling. An observation in vitro also indicated that the eNOS Glu298Asp variant alters caveolar localization in the microdomain of endothelial cells and function deficit in this eNOS corresponding enzyme casually linking Cav-1 highlights the relevance of interaction between eNOS and Cav-1 in reducing vasoprotection in arterial disease [43]. Therefore, molecular evidence of the interaction between eNOS and Cav-1 polymorphisms, such as 14713A, T29107A and 12759A alleles, on risk predisposition in LAA strokes in carriers who have hypercholesterolemia should be required to provide support in justifying our current association reports.
Notably, future molecular studies are required to reveal the important roles of eNOS Glu298Asp and Cav-1 14713A/29107A regarding the development of the coronary atrial disease. In the genomic era of personalized medicine, population-based genotyping association studies that are precise can be done if individual data with consideration of the traditional vascular risk factors [48], including hypertension, atrial fibrillation, obesity, diabetes, hypercholesterolemia, hyperuricemia, alcoholism, and heavy cigarette smoking, are available. In conclusion, individuals' susceptibility to an LAA stroke is influenced by genetic variants in caveolar genes in association with stroke-related risk factors. This study demonstrated that the eNOS Glu298Asp in combination with 14713A/29107A at the Cav-1 locus was associated with a significant risk of an LAA stroke.
Supporting information S1 Table. Demographical characteristics (age, gender, presence of hypertension, history of alcoholism and cigarette smoking) and polymorphism data of stroke patients and control subjects. (XLS)