Prostate Stem Cell Antigen Gene Polymorphism Is Associated with H. pylori–related Promoter DNA Methylation in Nonneoplastic Gastric Epithelium

Gastric cancer is one of the most common malignancies and the second causal of cancer-related death worldwide (1). Epidemiologic studies have demonstrated that Helicobacter pylori (H. pylori) infection is a critical event in the development gastric cancer and its premalignant conditions (2), while a complex interaction of host genetic factors may also be relevant in such process.

Previous genome-wide association study identified two functional SNPs associated with gastric cancer. The one was a polymorphism (rs2294008) in prostate stem cell antigen (PSCA; ref. 3), and the other was a polymorphism (rs4072037) in mucin 1 (MUC1; ref. 4). Both polymorphisms significantly affect functional activity of targeted genes and confer susceptibility to gastric cancer especially the diffuse type (3, 4). Although, detailed mechanisms how these polymorphisms influence the risk of gastric cancer remains unclear, it is possible that polymorphisms alter inflammatory immune response in gastric mucosa under the H. pylori infection.

Background gastric mucosa with diffuse-type gastric cancer is characterized as severe gastric inflammation, which is associated with enlarged-fold gastritis and aberrant DNA methylation (2, 5–7). Aberrant DNA methylation is shown to be associated with gastric cancer and H. pylori–infected gastritis (8–10). The gastric mucosa of patients with enlarged-fold gastritis exhibits especially higher levels of promoter methylation in H. pylori--infected patients (7), suggesting that aberrant DNA methylation is an essential organizer of carcinogenesis in individuals at high risk for diffuse-type gastric cancer. Premalignant conditions in the gastric mucosa under the H. pylori infection involves complex biological processes that are still not completely understood. The aim of this study was to investigate a potential link between H. pylori–related aberrant DNA methylation, and PSCA and MUC1 functional SNPs in the human gastric mucosa.

The study population comprised of 410 cancer-free subjects, attending the Endoscopy Center of Fujita Health University Hospital (Toyoake, Japan) from January 2005 to January 2010. This cohort was partly recruited from previous study investigating the association between promoter DNA methylation and severity of chronic gastritis (11), NSAIDs use (12), and host genetic factors (13, 14), but was independent from our recent studies investigating the DNA methylation in the gastric mucosa after successful H. pylori eradication (15, 16). All participants underwent upper gastroscopy as part of a health check, as a secondary procedure following barium X-ray examination for suspected stomach cancer, or for the investigation of abdominal discomfort.

Patients who had severe systemic disease, or malignancy in the stomach or other organ, or who had a history of gastric surgery were not included in this study. Patients who had history of H. pylori eradication were also not included in this study. The study was conducted in accordance with Declaration of Helsinki and was performed after approval by an Institutional Review Board of Fujita Health University School of Medicine and written informed consent was obtained from all the subjects. During upper gastroscopy, biopsy specimens were taken from the antrum, as well as upper corpus along the greater curvature from grossly nonpathologic mucosa in all the patients. The specimens taken from the antrum were cut into two pieces, and one part of specimens were immediately frozen and stored at −80°C until use. Using the frozen specimens from the antrum, genomic DNA was extracted using standard protein precipitation method. Other specimens from both the antrum and the upper corpus were fixed in 10% buffered formalin and embedded in paraffin for microscopic histologic examination. H. pylori infection status was assessed by histologic analysis of both the biopsy specimens from greater curvature of the gastric antrum and upper corpus using antibody to H. pylori to avoid false negative of H. pylori infection. In case of negative of H. pylori infection according to histologic analysis, serologic and urea breath test were added, and subjects with at least one of the diagnostic tests was diagnosed as H. pylori infection positive. Through microscopic examination, hematoxylin and eosin–stained histologic slides were also scored for their histologic parameters: acute or chronic inflammation, atrophy, and intestinal metaplasia. The scoring was performed according to updated Sydney systems (17). This analysis was successfully performed for 275 of 410 subjects.

Bisulfite pyrosequencing was used to quantify promoter CpG island (CGI) methylation of three genes (IGF2, MYOD1, and SLC16A12). The selection of these genes was based on the frequency of methylation in H. pylori–infected gastric mucosa (10). Bisulfite-treated genomic DNA was used to evaluate the methylation status by bisulfite pyrosequencing. Bisulfite treatment of DNA was carried out using a BislFast DNA Modification Kit (TOYOBO, Co., Ltd) according to the manufacturer's protocol. Pyrosequencing was performed by using a PSQ24 system and the Pyro-Gold Reagent Kit (Qiagen). The PyroMark Q24 Software (Qiagen) was used for processing the results. The primers used for pyrosequencing are listed in Supplementary Table S1.

Using genomic DNA extracted from uninvolved mucosa of the gastric antrum or peripheral blood, PSCA rs2294008 C/T and MUC1 rs4072037 A/G polymorphisms were genotyped by the pyrosequencing. Pyrosequencing was carried out using a PSQ24 system with Pyro-Gold Reagent Kit (Qiagen), and the results were analyzed using PyroMark Q24 Software (Qiagen). The primers used for pyrosequencing are listed in Supplementary Table S1.

Continuous variables between two groups were assessed using the Student t test.

Categorical variables between two groups were assessed using the χ² test.

Univariate and multivariate analysis were used for the association between clinicopathologic factors, DNA methylation, and PSCA or MUC1 polymorphisms. A probability value of less than 0.05 was considered significant.

The characteristics of subjects are shown in Table 1. After upper gastroscopy, 82 participants were diagnosed as having active peptic ulcer diseases. PSCA rs2294008 C/T and MUC1 rs4072037 A/G polymorphisms were successfully genotyped in all 410 subjects. Prevalence polymorphisms were, 154 T/T (37.6%), 194 C/T (47.3%), and 62 C/C (15.1%) for the PSCA rs2294008, 268 A/A (65.3%), 129 G/A (31.5%), and 13 G/G (3.2%) for the MUC1 rs4072037 polymorphisms. Prevalence of both genotypes seems to be consistent with those reported in cancer-free subjects in other studies in Japan (3, 4). We chose promoter CGIs from three genes (IGF2, MYOD1, and SLC16A12) for evaluation of DNA methylation status of gastric mucosa. All genes were previously reported to be hypermethylated in H. pylori–positive gastric mucosa (10). Methylation levels of all three genes in both H. pylori–positive and -negative subjects seemed to be quite similar to those seen in other data set from our previous study (10). Methylation of all three genes and its mean methylation z-score were significantly higher in H. pylori–positive subjects compared with those without H. pylori infection (all P < 0.0001; Supplementary Fig. S1).

Figures 1 and 2 show the association between DNA methylation status of three genes in different PSCA and MUC1 polymorphisms. In overall subjects, we found that methylation of all three genes and its mean methylation z-score were significantly higher in subjects with PSCA rs2294008 T/T compared with the PSCA rs2294008 C/C (all P < 0.05). The subjects with rs2294008 C/T genotype seemed to have intermediate methylation status in all three genes and its mean methylation z-score but the methylation status was not significantly different in the comparison with PSCA rs2294008 T/T or C/C genotypes (all P > 0.05). When subjects were divided into H. pylori–positive and -negative cases, same associations were more enhanced in H. pylori–positive subjects (all P < 0.01), while no association was found in H. pylori–negative subjects (all P > 0.1; Fig. 1).

Concerning the MUC1 rs4072037 A/G polymorphism, we did not find any significant association between DNA methylation status and different genotypes in overall subjects (Fig. 2). No association was also found when subjects were divided according to the H. pylori infection status (all P > 0.1; Supplementary Fig. S2).

We next performed univariate and multivariate analysis to assess the factors related to DNA methylation statistically. Sex, age, H. pylori infection status, NSAIDs use, and presence of gastric and duodenal ulcers were included for the analysis with PSCA polymorphisms. The mean methylation z-score of three promoter CGIs in the gastric mucosa presented an approximately Gaussian distribution, with overrepresentation of methylation-high cases, we set cut-off value of 0.65 (mean methylation z-score) for the definition of methylation-high cases. The univariate analysis demonstrated that age [OR: 1.03; 95% CI (confidence interval), 1.01–1.04; P = 0.005], H. pylori infection (OR, 13.1; 95% CI, 6.14–27.76; P < 0.0001), and PSCA rs2294008 T/T genotype (OR, 2.36; 95% CI, 1.10–5.02; P = 0.027) were significantly associated with methylation-high gastric mucosa, while weak trend was also found between PSCA rs2294008 C/T genotype and methylation-high gastric mucosa (OR, 1.86; 95% CI, 0.88–3.92; P = 0.1; Table 2). We then set a threshold of P = 0.10 or less and performed multivariate analysis. The multivariate analysis demonstrated that age (OR, 1.03; 95% CI, 1.01–1.05; P = 0.004), H. pylori infection (OR, 14.4; 95% CI, 6.71–31.12; P < 0.0001), PSCA rs2294008 C/T genotype (OR, 2.37; 95% CI, 1.06–5.29; P = 0.035), and PSCA rs2294008 T/T genotype (OR, 3.2; 95% CI, 1.41–7.25; P = 0.005) were significantly associated with methylation high as independent factors (Table 3).

Because severity of histologic gastritis has been shown to be associated with DNA methylation status in the gastric mucosa (10, 11), we investigated whether PSCA polymorphism would be associated with severity of histologic gastritis. Detailed histologic information regarding acute or chronic inflammation, atrophy, and intestinal metaplasia in antral biopsies where methylation status was evaluated was available among 275 subjects. Among H. pylori–infected subjects, degree of atrophy in the antrum was significantly higher in subjects with PSCA rs2294008 T/T genotype compared with that in subjects with PSCA rs2294008 C/T genotype (P = 0.03). We also found that degree of intestinal metaplasia in the antrum was significantly higher in subjects with the PSCA rs2294008 T/T genotype compared with that in PSCA rs2294008 C/T genotype in all subjects (P = 0.04; Supplementary Fig. S3). To further evaluate whether PSCA polymorphism would be associated with DNA methylation in the presence of gastric atrophy or intestinal metaplasia, we performed multivariate analysis with adjustment for the presence of gastric atrophy and intestinal metaplasia in the same 275 subjects. For this analysis, age, H. pylori infection status, PSCA rs2294008 C/T genotype, and PSCA rs2294008 T/T genotype were also included in addition to the gastric atrophy and intestinal metaplasia. The PSCA rs2294008 T/T genotype (OR, 3.04; 95% CI, 1.12–8.26; P = 0.029) was significantly associated with methylation high as an independent factor, while only a weak trend was found between PSCA rs2294008 C/T genotype and methylation-high gastric mucosa (OR, 2.52; 95% CI, 0.97–6.55; P = 0.059), probably due to the low statistical power as a result of reduction of sample size (Table 4). Finally, we investigated whether PSCA polymorphism is associated with H. pylori infection. However, no association was found between PSCA polymorphism and H. pylori infection status (Supplementary Table S2).

We showed that PSCA rs2294008 polymorphism is associated with DNA methylation status in the gastric mucosa. Methylation levels of selected three genes (IGF2, MYOD1, and SLC16A12) and its mean methylation z-score were significantly higher in subjects with PSCA rs2294008 T/T genotype compared with that in the subjects with PSCA rs2294008 C/C genotype. The result was more enhanced in H. pylori–infected subjects. The multivariate analysis demonstrated that PSCA rs2294008 C/T and T/T genotypes were significantly associated with methylation-high gastric mucosa as independent factors.

From a functional perspective, PSCA rs2294008 polymorphism located in the promoter region of the gene and it also corresponded to the GeneHancer Regulatory Elements (https://genome.ucsc.edu/FAQ/FAQblat.html). The PSCA rs2294008 T allele has been shown to reduce transcriptional activity of an upstream fragment of the gene and confer susceptibility to gastric cancer especially the diffuse type histopathology (3). The three promoter CGIs analyzed in this study are frequently methylated in H. pylori–infected gastric mucosa and are associated with histologic degree of gastritis (10). Importantly, background gastric mucosa with diffuse type gastric cancer is characterized as severe gastric inflammation, which is associated with enlarged-fold gastritis and higher levels of H. pylori–related promoter methylation (2, 5–7). Our result provided evidence that PSCA rs2294008 polymorphism influence the susceptibility to gastric cancer through altering methylation status of gastric mucosa.

PSCA is expressed in differentiating gastric epithelial cells, has a cell proliferation inhibition activity in vitro and is frequently silenced in gastric cancer (3). Because the mechanisms related to methylation accumulation would be rapid cell proliferation (8, 10) or chronic inflammation (18), it is possible that PSCA polymorphism influence the cell proliferation or inflammatory response in gastric epithelial cells and modify methylation status in the gastric mucosa. Although we did not find a relation between PSCA polymorphism and degrees of acute and chronic cellular inflammation, PSCA rs2294008 T/T genotype was associated with more severe gastric atrophy and intestinal metaplasia, which was in line with other study (19). Gastric atrophy and intestinal metaplasia would be a final stage of mucosal change of the gastric inflammatory response resulting in gastric cancer, and it can also be noted that these are also closely associated with higher levels of H. pylori–related promoter methylation (10, 11). Our result suggests that PSCA polymorphism may be associated with methylation-related carcinogenesis modulating inflammatory response in the gastric mucosa. It is also possible that the polymorphism influences the expression of target genes by altering the methylation status in promoter CGI. More detailed mechanistic examination about how PSCA polymorphism influence the DNA methylation in the gastric mucosa need to be performed in the future study.

We did not observe any significant association between MUC1 rs4072037 polymorphism and methylation status in the gastric mucosa. The result suggests that underlying mechanisms for developing gastric cancer would be different among PSCA and MUC1 polymorphisms. Although the MUC1 polymorphism also increases the susceptibility to gastric cancer with diffuse type histopathology (4), mechanisms may not be related to methylation-related carcinogenesis.

No potential conflicts of interest were disclosed.

Conception and design: T. Tahara, S. Tahara

Development of methodology: T. Tahara, S. Tahara

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): T. Tahara, S. Tahara, M. N. Horiguchi, T. Terada, M. Okubo, D. Yoshida, K. Funasaka, T. M. Nagasaka, T. Tsukamoto, N. Ohmiya

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): T. Tahara, S. Tahara, T. Kato, Y. Shinkai, Y. Nakagawa

Writing, review, and/or revision of the manuscript: T. Tahara, S. Tahara

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): S. Tahara, H. Kurahashi

Study supervision: S. Tahara, T. Terada, T. Shibata, N. Ohmiya

JSPS KAKENHI Grant-in-Aid for Young Scientists (B) grant no. JP26870685 supported this study.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.