INTRODUCTION
Helicobacter pylori is a Gram-negative bacterium that colonizes the stomach of half of the human population throughout the world [Reference Dunn, Cohen and Blaser1]. Although there is strong evidence that H. pylori infection increases the risk of gastric cancer (GC), the second most frequent cause of cancer-related death [Reference Suerbaum and Michetti2], the molecular mechanisms of H. pylori-associated gastric carcinogenesis remain undefined. Recently, a study provided evidence that H. pylori infection introduces DNA double-strand breaks (DSBs) in host cells and prolonged active infection leads to saturation of cellular repair capabilities, which may contribute to the genetic instability and frequent chromosomal aberration that are hallmarks of GC [Reference Toller3]. On the other hand, to date, more data suggest that severe H. pylori-mediated diseases are associated with not only persistent H. pylori infection but also genetic variants of the host [Reference Matsukura4–Reference Maran7].
Genetic variants could affect mRNA splicing, regulation of transcription, translation efficiency or mRNA stability, leading to altered polypeptide product levels, which could affect the function of proteins such as RAD51 [Reference Gray8–Reference Chistiakov, Voronova and Chistiakov10], which may underlie the variation in clinical outcomes from H. pylori infection. RAD51 is a well-known repair protein; it controls DSBs via gene conversion. It has been observed that the RAD51 G135C polymorphism increased activity of RAD51 [Reference Chistiakov, Voronova and Chistiakov10] and was associated with the development of GC [Reference Poplawski9].
In this study, to further explore whether the RAD51 G135C polymorphism was important in identifying patients who are at a higher risk of developing GC, we investigated the distribution of the genotypes and frequencies of alleles of this polymorphism in Bhutan, a country documented as having an extremely high prevalence of H. pylori infection (73·4%) and high prevalence of GC [Reference Shiota11, Reference Shiota12]. The age-standardized incidence rate of GC was reported to be high in Bhutan with 17·2 cases/100 000 per year (accessible at http://globocan.iarc.fr/). Indeed, when we performed a survey using gastroduodenal endoscopy in Bhutan in 2010, we found five cases of GC in 372 volunteers [Reference Shiota11, Reference Shiota12]. Therefore, we believe that the actual number of GC patients in Bhutan is higher than previously estimated.
SUBJECTS AND METHODS
Subjects
We recruited individuals with dyspeptic symptoms from Bhutanese volunteers. The surveys took place at the Jigme Dorji Wangchuk National Referral Hospital, Thimpu, Bhutan in December 2010. Written informed consent was obtained from all the participants. The protocol was approved by the ethics committee of Oita University Faculty of Medicine (Japan), and by the hospital where sample collection was performed.
During each endoscopy session, three gastric biopsy specimens were obtained from the antrum: one each for H. pylori culturing/DNA examination, rapid urease test, and histological examination. Clinical presentations included gastritis, duodenal ulcer (DU), gastric ulcer (GU), and GC. H. pylori-positive participants without peptic ulcers and/or gastric malignancy were defined as gastritis. We also collected blood from all volunteers on the same day of endoscopy to evaluate anti-H. pylori antibody in serum. Patients with a history of partial gastric resection were excluded. We also excluded patients with using proton pump inhibitors, but not other anti-reflux medication, during the last month. We further excluded patients with previous H. pylori eradication therapy (either successful or not).
H. pylori diagnosis
In this study, we selected H. pylori-positive cases from H. pylori-seropositive cases with additionally at least one more positive test result which included rapid urease test, histology or culture. Patients were considered to be H. pylori-negative when all the tests were negative as described in previous studies [Reference Shiota11, Reference Shiota12] or when only serology was positive.
Histopathology
Biopsy specimens for histology were fixed in 10% buffered formalin for 24 h, then embedded in paraffin. Serial sections were stained with haematoxylin and eosin and Giemsa stains, followed by evaluation by a single pathologist blinded to the patient's clinical diagnosis or the characteristics of the H. pylori strains. Histological analyses of the gastric mucosa were evaluated according to the updated Sydney system [Reference Dixon13]. The degree of inflammation, neutrophil activity, atrophy, intestinal metaplasia, and bacterial density were classified into four grades (0, normal; 1, mild; 2, moderate; 3, marked) [Reference Rugge and Genta14, Reference Rugge15]. All histological analyses were performed by one co-author (T.U.) without knowledge of clinical data or the experimental protocol for the respective tissue.
Genotyping of the RAD51 polymorphism
Genomic DNA was isolated from biopsy specimens which was used for H. pylori culture using QIAamp DNA mini-kit (Qiagen, USA). The G135C single nucleotide polymorphisms (SNPs) of the RAD51 gene were determined by restriction fragment length polymorphism analysis of PCR amplified fragments (PCR-RFLP). RAD51 genotyping was analysed by PCR amplification of a 175-bp region around nucleotide 135 using the following primers: forward (5’-TGG GAA CTG CAA CTC ATC TGG-3’) and reverse (5’-GCG CTC CTC TCT CCA GCAG-3’) [Reference Blasiak16]. This region contained a single MvaI site that abolished the 135C allele; therefore only wild-type alleles were digested by MvaI resulting in 86-bp and 71-bp products. The digested fragments were separated and visualized by 15% acrylamide gel electrophoresis (Wako Pure Chemical Industries Ltd, Japan). Wild-type/mutant genotype was confirmed by automatic sequencing using the ABI-PRISM Big Dye™ Terminator v. 3.0 Cycle Sequencing Read Reaction kit (Applied Biosystems, USA). After purification, the sequencing products were visualized on an ABI-PRISM 310 Genetic Analyzer (Applied Biosystems).
Statistical analysis
The individual genotype and allele frequencies were analysed using a 2 × 2 contingency table with χ 2 test to examine the association between each genotype and clinical outcome. All determinants with P values of <0·10 were entered together in the full model of logistic regression, and the model was reduced by excluding variables with P values >0·10. Spearman rank coefficients (r) were also determined to evaluate the association between the different genotypes of the strains. A P value <0·05 was accepted as statistically significant. The SPSS statistical software package v. 19.0 (SPSS Inc., USA) was used for all statistical analyses. Hardy–Weinberg equilibrium was calculated for the genetic polymorphisms by χ 2 test.
RESULTS
Subjects
We selected the first 150 volunteers (mean age 39·0 years, range 16–79 years) who matched our criteria for enrolment in the study, including the criteria for diagnosing H. pylori infection, from 372 volunteers examined in our previous study, which was a survey of H. pylori prevalence [Reference Vilaichone17]. After exclusion of 12 subjects with failed genotype or undetermined H. pylori status (e.g. only seropositive subjects), we successfully identified RAD51 genotyping in 138 subjects (122 with gastritis, three with GU, seven with DU, two GC and four samples without diagnostic information). The characteristics of studying population are described in the Table 1. There was no difference in the male:female ratio in the study population.
Table 1. Characteristics of the subjects
For histological scores minimum to maximum (0–3) and mean (median) values are presented.
Allele frequencies of RAD51 G135C SNPs
All samples were divided into three genotypes of the RAD51 5’-UTR: G135C wild type (G/G), heterozygous (G/C) and mutant homozygous (C/C). The allelic distribution of the RAD51 gene G135C SNPs in Bhutanese subjects is given in Table 1. G135G genotype was more dominant than G135C, C135C genotype was rare and only one subject carried the genotype. Genotypic and allelic frequencies of RAD51 G135C were in agreement with Hardy–Weinberg equilibrium.
RAD51 G135C SNPs and clinical outcome
We examined the association between RAD51 genotypes and clinical outcome (Table 2). Probably due to the small number of DU, GU and GC subjects, the difference between clinical outcome and distribution of the RAD51 genotypes was not found in present study.
Table 2. Clinical and histological features of 137 subjects with the allelic distribution of RAD51 G135G and RAD51 G135C
n.s., Not significant (P > 0·05).
For histological scores minimum to maximum (0–3) and mean (median) values are presented.
† By Pearson's χ 2 test.
‡ By Mann–Whitney test.
* Significant at P < 0·05.
RAD51 G135C SNPs and histological findings
In all subjects, histological scores for activity, inflammation and atrophy were not different between G135G and G135C carriers both in the antrum and corpus (Table 2). However, G135C carriers showed significantly higher scores for intestinal metaplasia in the antrum than G135G carriers [mean (median) 0·33 (0) vs. 0·08 (0), P = 0·008] (Table 2). Interestingly, higher scores for intestinal metaplasia of G135C carriers compared to G135G carriers were also observed in H. pylori-positive volunteers [mean (median) 0·3 (0) vs. 0·1 (0), P = 0·002], and even limited to H. pylori-positive volunteers with gastritis [mean (median) 0·4 (0) vs. 0·1 (0), P = 0·002] (Table 3). By contrast, these differences were not observed in either H. pylori-negative volunteers or H. pylori-negative gastritis subjects. Regarding H. pylori-negative volunteers and H. pylori-negative gastritis volunteers, G135G genotype showed higher atrophy scores in the antrum compared to G135C genotype (P = 0·02 in both cases).
Table 3. The correlation of RAD51 G135C SNPs and histological score
n.s., Not significant (P > 0·05); IM, intestinal metaplasia.
For histological scores minimum to maximum (0–3) and mean (median) values are presented.
P values by Mann–Whitney test.
* Significant at P < 0·05.
DISCUSSION
This is the first study to show severe H. pylori-mediated diseases being associated with the RAD51 G135C polymorphism. A previous study showed a strong association between RAD51 G135C polymorphism and the occurrence of GC in individuals with high levels of oxidative DNA damage or impaired repair of such damage [Reference Poplawski9]. However, that study did not examine H. pylori status, which is well known as a pathogen that can cause DNA damage in host gastric epithelial cells and a source of reactive oxygen species (ROS). In addition, the ethnicity of the subjects enrolled in that study were not described in detail. In the present study, all the enrolled subjects were selected from volunteers with a low number of GC cases, which is a limitation of this study; thus the association between GC and RAD51 G135C polymorphism was not found. However, the association between gastric intestinal metaplasia and adenocarcinoma of stomach is well known [Reference Correa, Cuello and Duque18–Reference Shimoyama20] and we found a strong association between RAD51 G135C polymorphism and higher intestinal metaplasia score in the antrum in H. pylori-positive subjects in Bhutan. The correlation coefficient of G135C genotype with high score of intestinal metaplasia in the antrum were highest in H. pylori-positive gastritis subjects (r = 0·4, P = 0·001) followed by H. pylori-positive subjects (r = 0·3, P = 0·002) and gastritis subjects (r = 0·2, P = 0·007). Interestingly, this association was not found in H. pylori-negative volunteers. While G135G genotype was evenly distributed in H. pylori-negative and H. pylori-positive subjects (47% vs. 53%), prevalence of G135C genotype in H. pylori-negative subjects was less than twofold lower in H. pylori-positive subjects (31% vs. 69%). Compared to G135C carriers, G135G carriers only showed higher atrophy scores in the antrum in H. pylori-negative subjects with a very limited number of G135C carriers. Our findings reveal that a combination of H. pylori infection and RAD51 G135C genotype of the host showed an increasing intestinal metaplasia score, which might be useful for predicting the risk of GC.
The severity of H. pylori-induced gastric diseases are not only associated with the bacterium's persistence in the host but also with the adaptation of bacterium to host differences, which in turn is associated with the host's remarkable genetic variability. GC with H. pylori infection shows genetic instability, besides impairment of important DNA repair pathways [Reference Cervantes21–Reference Habano25]. A number of studies have demonstrated that H. pylori infection induces the synthesis of ROS and causes DNA damage [Reference Toller3, Reference Machado22, Reference Obst26] or decreases the activity of DNA repair pathways in the host [Reference Park27–Reference Machado29], and that a lengthy period of infection will increase the risk for development of GC. In response to DNA damage, human RAD51 has been described as a well-known protein functioning in DNA repair. RAD51 has been shown to be involved in the repair of different kinds of DNA lesions during replication and it promotes genomic stability in eukaryotic cells [Reference Lundin30, Reference Orre31]. The increased DSB repair capacity might be followed by RAD51 up-regulation. However, the overexpression of RAD51 also affects other cellular processes influencing cell survival, cell cycler progression or promotion of apoptosis in the cells [Reference Flygare32, Reference Richardson33]. The higher mRNA expression of RAD51 in tumour tissue of GC compared to normal tissue was observed in a previous study [Reference Borrego34]. Moreover, the RAD51 G135C polymorphism was expected to result in increased activity of RAD5 [Reference Chistiakov, Voronova and Chistiakov10]. These findings support for the hypothesis that the contribution of RAD51 G135C polymorphism and H. pylori infection might be responsible for the aberrant increase in RAD51 expression following increased risk of GC. In future work the importance of such polymorphisms for outcome of H. pylori-associated diseases should be studied in more detail using a larger number of H. pylori infection with GC.
In conclusion, our results reveal that RAD51 G135C SNPs might be an important predictor of GC in H. pylori-infected patients.
ACKNOWLEDGEMENTS
This report is based on work supported in part by grants from the National Institutes of Health (DK62813) (Y.Y.), and Grants-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan (24406015, 24659200, 25293104, 26640114) (Y.Y.), and Special Coordination Funds for Promoting Science and Technology from the MEXT of Japan. T.T.H.T. is PhD student supported by The Japanese Government (Monbukagakusho: MEXT) Scholarship Programme since 2011.
DECLARATION OF INTEREST
None.
