In this study of a Japanese population, we have demonstrated for the first time that the -105G>A polymorphism of SEPS1 gene was associated with increased risks of intestinal type of gastric cancer and gastric cancer located in middle third part of the stomach. Although low selenium status has been associated with risk of human gastric cancer [18, 19], and high selenium diet inhibit the growth of H. pylori in Guinea pigs , human clinical trials have failed to demonstrate a benefit for selenium supplements in the prevention of precancerous gastric lesions . The -105G>A promoter polymorphism, which is located in an ER stress response element of the seps1 gene coding for SEPS1, was strongly associated with circulating levels of pro-inflammatory cytokines, such as IL-1b, IL-6, and TNF-a, and with SEPS1 gene expression levels in humans. The phenotypic consequences of this -105G>A polymorphism have been investigated in some diseases related to chronic inflammation [11–14]. The results have been inconsistent. One study in Finnish cohort showed the relation of other SEPS1 SNPs and coronary heart disease or ischemic stroke event, and one report in a large Norwegian case-control cohort showed the A allele of the SEPS1-105G>A polymorphism is a significant risk factor for preeclampsia. On the other hand, one report in Germany showed the SELS-105G>A polymorphism was not associated with IBD susceptibility and did not contribute to a certain disease phenotype or increased TNF-alpha levels in IBD patients. Also one report showed non-significant differences of SEPS1 allele frequencies between young stroke patients and healthy controls from Italy and Germany.
In our study, there was no association overall between the -105G>A variant and gastric cancer. However, in detailed clinicopathological analysis, we found significant associations between carrying the A allele and the odds of specific types of gastric cancer, with adjustment for age, sex, and H. pylori infection status. We have 74 healthy volunteer DNA (all are H. pylori negative). We examined these DNA about SEPS1 polymorphism -105G>A. As a result, the numbers of GG were 70, the numbers of GA were 4 and none of them was AA. From these results, we thought our comparison between gastric cancer patients and non gastric cancer patients was proper comparison for analyzing the association risk of gastric cancer and SEPS1 polymorphism other than the association of H. pylori.
The carcinogenic pathway for the intestinal type of gastric cancer mainly involves H. pylori infection. The infection causes inflammation and tissue regeneration, and these processes cause the deviation from the normal pathway of gastric differentiation to precancerous states[4, 22]. It has been suggested that the SEPS1 polymorphism may be involved in the pathway from chronic gastric inflammation to carcinogenesis. However, the AA genotype is very rare in the Japanese population, and thus, definite conclusions about the genotype-phenotype correlations of homozygous carriers of this SEPS1 promoter polymorphism can only be drawn from larger studies involving multiple hospital centers.
Among males, the SEPS1 -105A allele carriers had increased odds of gastric cancer compared with those with the GG genotype. There is one report showing gender-specific associations of the SEPS1 polymorphism with coronary heart disease. The authors of the study found relations between a polymorphism in another region of the SEPS1 gene and coronary heart disease risk in females. They suspected that the gender-specific associations with SEPS1 SNP were due to differences in disease etiology or in the hormonal milieu for men and women . In epidemiological studies, gender has also been related to gastric cancer [23, 24]. These epidemiological findings are consistent with the gender specific association between the -105G>A polymorphism and gastric cancer.
Based on the reports of the consequences of impaired SEPS1 gene expression and of the SEPS1 -105G>A polymorphism being functionally involved in inflammatory responses, we further assessed gastritis scores in the non-cancerous areas of the antrum in H. pylori positive patients. However, we did not find significant differences between SEPS1 -105GG homozygotes and -105A allele carriers (data not shown). It is possible that the number of cases for this analysis were relatively low. Also, this investigation was not a direct comparison of inflammatory cytokine levels by genotypes of this polymorphism as has been reported in the literature. Therefore, the exact role of selenoproteins in gastric carcinogenesis and inflammation has not yet been determined. Previous reports have found associations between cytokines and gastric cancer[15, 25]. In future studies, the analysis of these inflammatory cytokines in gastric cancer will be useful for understanding the mechanisms of carcinogenesis in the stomach.