Endothelial nitric oxide synthase polymorphisms and haplotypes in genetic epidemiology and pharmacogenetics: remarks regarding a lack of association on the risk of myocardial infarction

Abstract

To the Editor, Arun Kumar et al. [1] have recently reported a clinical study on the possible link between the association of 894G[T and -786T[C NOS3 (endothelial nitric oxide synthase 3) gene polymorphisms and the risk of myocardial infarction (MI) in the South Indian population. They have retrospectively studied a considerable number of patients and healthy controls, but found no association between these NOS3 polymorphisms and MI, even after adjusting for the confounding variables like age, sex, BMI, smoking, alcohol status and lipid parameters by logistic regression analysis. The authors stated that the observed results of lack of association could be due to the distribution of NOS3 variants among different ethnic groups, which may explain interethnic differences in NO-mediated vasodilation and response to drugs. Moreover, they addressed the importance to specifically characterizing South Indian variations to study gene-disease associations rather than comparing it with distantly related ethnic groups, as the South Indian population is different from the other ethnic groups around the world [1]. I would like to point out some issues that should probably have been taken into consideration by the authors. First, we know more about the genetics of some human populations as compared to others, as previously highlighted [2]. Therefore, in order to assess the distinctiveness of the South Indian population, I have compared the genotype and allele frequencies of 894G[T and -786T[C NOS3 polymorphisms for the healthy controls reported by the authors [1] with those reported for the populations of the HapMap Project [3] (Supplementary Material). Unfortunately, there are no data available for the Indian population in the third phase of HapMap project (Gujarati Indians in Houston, Texas). Nevertheless, I have found striking differences when comparing the South Indian healthy controls with other Asian (Chinese and Japanese), African and European populations from HapMap using v tests (P \ 0.0001; Supplementary Material). Indeed, we have previously reported a remarkable disparity in the distribution of NOS3 genotypes and alleles among Amerindians from three tribes of Brazilian Amazon when compared to black and white Brazilians [4, 5]. Taken together, these findings support the idea that the distribution of gene polymorphisms affecting disease susceptibility or drug sensitivity often varies among different ethnic groups [2, 4]. However, other important issues may be considered when interpreting the observed results of lack of association reported by Arun Kumar et al. [1]. Second, the authors should have estimated the haplotype frequencies, the combination of genetic polymorphisms within a specific gene, and performed association analysis with NOS3 haplotypes. It has been acknowledged that haplotype analysis can provide much more useful information than the information derived from single polymorphisms analyses in association studies [6–8]. For example, our group has previously reported that NOS3 haplotypes are associated with the development of hypertension, and that such findings would have been obscured if specific NOS3 genotypes alone had been considered [9–11]. Notably, we found that specific NOS3 haplotypes, but not genotypes, are associated with susceptibility to Electronic supplementary material The online version of this article (doi:10.1007/s11033-013-2928-1) contains supplementary material, which is available to authorized users.